+-------------------------------------------+
    |                 Note:                     |
    |                                           |
    | = around word indicates bold =CAPSULE.=   |
    | _ around word indicated italics _Erebus_  |
    +-------------------------------------------+





    HARVARD PSYCHOLOGICAL
    STUDIES

    EDITED BY

    HUGO MÜNSTERBERG

    Volume II

    BOSTON AND NEW YORK
    HOUGHTON, MIFFLIN AND COMPANY
    The Riverside Press, Cambridge
    1906


    COPYRIGHT 1906
    BY THE PRESIDENT AND FELLOWS OF HARVARD COLLEGE
    ALL RIGHTS RESERVED

    _Published June 1906_




CONTENTS


    EMERSON HALL: Hugo Münsterberg.

    I. Experimental Psychology in Harvard                              3
    II. The Need for Emerson Hall                                      8
    III. Emerson as Philosopher                                       16
    IV. The Place of Experimental Psychology                          31
    V. The Psychological Laboratory in Emerson Hall                   34

    OPTICAL STUDIES.

    Stereoscopic Vision and the Difference of Retinal Images: G. V.
      Hamilton                                                        43
    Eye-Movements during Dizziness: E. B. Holt                        57
    Vision during Dizziness: E. B. Holt                               67
    Visual Irradiation: Foster Partridge Boswell                      75

    FEELING.

    The Expression of Feelings: F. M. Urban                          111
    The Mutual Influence of Feelings: John A. H. Keith               141
    The Combination of Feelings: C. H. Johnston                      159
    The Æsthetics of Repeated Space Forms: Eleanor Harris Rowland    193
    The Feeling-Value of Unmusical Tone-Intervals: L. E. Emerson     269

    ASSOCIATION, APPERCEPTION, ATTENTION.

    Certainty and Attention: Frances H. Rousmaniere                  277
    Inhibition and Reënforcement: Louis A. Turley                    293
    The Interference of Optical Stimuli: H. Kleinknecht              299
    Subjective and Objective Simultaneity: Thomas H. Haines          309
    The Estimation of Number: C. T. Burnett                          349
    Time-Estimation in its Relations to Sex, Age, and Physiological
      Rhythms:
        Robert M. Yerkes and F. M. Urban                             405
    Associations under the Influence of Different Ideas: Bird T.
      Baldwin                                                        431
    Dissociation: C. H. Toll                                         475

    MOTOR IMPULSES.

    The Accuracy of Linear Movement: B. A. Lenfest                   485
    The Motor Power of Complexity: C. L. Vaughan                     527

    ANIMAL PSYCHOLOGY.

    The Mutual Relations of Stimuli in the Frog _Rana Clamata_ Daudin:
      Robert M. Yerkes                                               545
    The Temporal Relations of Neural Processes: Robert M. Yerkes     575
    The Mental Life of the Domestic Pigeon: John E. Rouse            581
    Reactions of the Crayfish: J. Carleton Bell                      615




PLATES


    _Frontispiece_         3
       I                  60
      II                  64
     III                  78
      IV                  80
       V                 269
      VI                 271
     VII                 273
    VIII                 293
      IX                 295




EMERSON HALL

[Illustration: HARVARD PSYCHOLOGICAL LABORATORY]




EMERSON HALL

BY HUGO MÜNSTERBERG


I. EXPERIMENTAL PSYCHOLOGY IN HARVARD

On the 27th of December, 1905, Harvard University opened its new house
of philosophy, Emerson Hall. The presence of the American Philosophical
and Psychological Associations gave national significance to the
completion of this building.

The psychologist will find quarters in all parts of Emerson Hall. The
general courses in psychology will be held on the first floor in the
large lecture-room, which has nearly four hundred seats; and close by
are the psychological seminary-room and smaller lecture-rooms for the
advanced psychological courses. On the second floor the psychologist
finds his special library as a wing of the large library hall. But
the exclusive domain of the psychologist is the third floor,--a
psychological laboratory with twenty-five rooms. A large attic hall for
laboratory purposes on the fourth floor completes the psychologist's
allotment.

The work to be reported in future in the Harvard Psychological Studies
will be work done in this new building, and while the researches
reported in the following pages were completed in the smaller quarters
of the old laboratory, it seems natural that this volume, which appears
at this new epoch of our work, should give an account both of our
psychological past and of the development and purpose of Emerson Hall.

The Harvard Psychological Laboratory was founded in 1891 by Professor
William James, who had introduced some experimental features into his
psychological lecture courses for some time before the formal opening
of a regular workshop. Professor James started with two large rooms
on the second floor of Dane Hall, and secured an excellent equipment,
especially for the study of the psychology of the senses. He was
assisted by Dr. Herbert Nichols, and at once gathered a number of
graduate students for research.

In the following year Professor James withdrew from the experimental
work, and the conduct of the laboratory was given over to me. In the
years which followed, Dr. Arthur Pierce, Dr. J. E. Lough, and Dr.
Robert MacDougall were the assistants until three years ago, when the
development of the laboratory demanded a division of the assistant
functions; since that time Dr. E. B. Holt has been the assistant for
the work in human psychology, while Dr. R. M. Yerkes has had charge
of the work in comparative psychology. Since from the first I laid
special emphasis on research work, a greater number of small rooms
was soon needed. In the year 1893, we divided a part of the adjacent
lecture-room into four rooms for special investigations, and two years
later the larger of the two original rooms was divided into five. As
the lecture-room also was finally made part of the research laboratory,
we had at last eleven rooms in Dane Hall. The activities of the
laboratory, however, went far beyond the research work. We had regular
training-courses in experimental practice, and the lecture courses in
human and in comparative psychology drew largely on the resources of
our instrument cases. Yet the original investigations absorbed the
main energy of the laboratory, and demanded a steady expansion of
its apparatus. An illustrated catalogue of the instruments has been
published as part of the Harvard Exhibition at the Chicago World's Fair.

The participation of the students has been controlled by a principle
which has characterized our Harvard work through all these years,
and distinguished it from the methods of most other institutions. I
insist that no student shall engage in one investigation only, but
that every one who has charge of a special problem shall give to
it only half of his working time, while in the other half he is to
be subject in four, five, or more investigations by other members
of the laboratory. In this way each research is provided with the
desirable number of subjects, and all one-sidedness is avoided. Every
experimenter thus comes in contact with a large range of problems and
gets a fair training in manifold observations, besides the opportunity
for concentration on a special research. It is true that this demands
a complicated schedule and careful consideration of the special needs
of every research, but it gives to the work a certain freshness and
vividness, and banishes entirely the depression which is unavoidable
whenever a student is for any length of time a passive subject in one
psychological enquiry only. In both capacities, as experimenter and
as observing subject, only graduate students have been acceptable. In
this way about _one hundred investigations_ on human psychology have
been carried on, for most of which I have proposed the problems and
the special lines of work, taking care that the research of succeeding
years and of succeeding generations of graduate students should show
a certain internal continuity. Whenever the results seemed fit for
publication, the papers have been published under the names of the
students who had the responsibility for the conduct of the experiments.
Until three years ago the publication was scattered; most of the
papers, however, appeared in the Psychological Review. The Harvard
Psychological Studies, beginning in 1903, are to gather the bulk of our
material, although not a few of the researches of recent years have
been published in other places.

The laboratory has always sought to avoid one-sidedness, and this the
more as it was my special aim to adjust the selection of topics to the
personal equations of the students, many of whom came with the special
interests of the physician, the zoölogist, the artist, the pedagogue,
and so on. My own special interests may have emphasized those problems
which refer to the motor functions and their relations to attention,
apperception, space-sense, time-sense, feeling, etc. At the same time
I have tried to develop the psychological-æsthetic work, which has
become more and more a special branch of our laboratory, and there has
been no year in which I have not insisted on some investigations in the
fields of association, memory, and educational psychology. On the other
hand, in a happy supplementation of interests, Dr. Holt has emphasized
the physiological psychology of the senses, and Dr. Yerkes has quickly
developed a most efficient experimental department of animal psychology.

As the work thus became more manifold, the old quarters in Dane Hall
appeared less and less sufficient. And yet this laboratory development
has been merely parallel to the growth of general philosophical studies
in the whole University. The demand for a new hall, exclusively devoted
to philosophy, was thus suggested from many sides. The idea of linking
it with the name of Ralph Waldo Emerson has been for years a cherished
plan of Professor Palmer.

An especially appropriate time for the realization of such a plan came
in the approach of the hundredth anniversary of Emerson's birthday.
Almost two years before this date the Department took the first steps
in seeking to interest the members of the Visiting Committee for the
collection of the necessary funds. This Committee, consisting of Mr.
G. B. Dorr, chairman, Mr. R. H. Dana, Dr. R. Cabot, Mr. J. Lee, Mr.
D. Ward, and Mr. R. C. Robbins, showed not only warm interest, but
lent itself to the furtherance of the plans with such an energy and
devotion that the Philosophical Department owes to these friends of
philosophy in Harvard the most lasting gratitude. Various means were
taken by the Committee and by the Department to stir the interest of
the public, and soon the gifts began to come in, gifts of which some
were clearly given from sympathy with the work of the Philosophical
Department, some evidently in memory of Emerson. The original plans
of the architect called for $150,000 for the building. When, on the
25th of May, 1903, the hundredth anniversary of Emerson's birthday was
celebrated, the University had contributions amounting to more than
this sum, and given by one hundred and seventy persons.

It was soon found, however, that this sum was inadequate; yet we never
asked in vain. Additional gifts came in for the building fund, just as
later the generosity of several friends furnished the building with a
handsome equipment and the laboratory with new instruments. Mr. R. C.
Robbins gave the books for a philosophical library to be placed in the
new Hall.

The architect chosen was Mr. Guy Lowell, who has had to labor under the
difficulties involved in the fact that the best and quietest available
place was on Quincy Street opposite Robinson Hall. This spot demanded
that the new building be harmonized with Robinson and Sever Halls, two
structures most unlike in their architectural style. There was not
even the possibility of making it a companion to Robinson Hall, since
the latter has but two stories, while it was evident that Emerson
Hall needed three stories. The plan finally accepted, a Greek, brick
building with brick columns and rich limestone trimmings, provided for
the work of the whole Philosophical Division with the exception of
education. The Education Department, with its large library, will soon
need a whole building of its own, and has thus had no interest in being
housed under the roof of Emerson Hall. On the other hand, the building
was to give full space to that part of our Philosophical Division which
now forms, like education, an administrative unity,--the Department of
Social Ethics. A special library, museum rooms, etc., for social ethics
were planned for the second floor by the munificence of an anonymous
benefactor. Altogether we have six large lecture-rooms, two library
halls, two collection-rooms, a department-room, a seminary-room,
two studies and conference-rooms, twenty-five laboratory-rooms, all
connected by very spacious, well-lighted halls and broad, imposing
stairways. Surely never before in the history of scholarship has such
a stately house been built for philosophy. And while the nature of the
work is certainly not determined by the luxury of stone and carved
wood, teachers and students alike must feel these superb surroundings
as a daily stimulus to their best efforts.

At Christmas, 1905, the building stood ready for use, and Duveneck's
bronze statue of Emerson was unveiled in the entrance hall. At the
opening meeting, after short dedicatory orations by President Eliot
and Dr. Edward Emerson, a real exchange of ideas in a joint debate
of the Philosophical and Psychological Associations was substituted
for the usual formal exercises. The question debated was suggested by
the fact that Emerson Hall was to house the psychological laboratory.
Does psychology really belong to philosophy or rather to the natural
sciences? As the representative of Harvard, it was my part to open the
debate and to characterize the attitude of the Harvard laboratory.

My remarks on that occasion may thus serve as the most direct
introduction to our work. They are printed here, together with a short
sketch of the equipment of the laboratory. I venture to add also two
other papers, one of which points to the administrative, the other and
longer one to the philosophical background of Emerson Hall. Inasmuch
as I was Chairman of the Philosophical Department throughout the five
years in which the plan for Emerson Hall was growing and became finally
realized, it has been my official duty repeatedly to express our hopes
and ideals. Thus I had to formulate the wishes of the Department at
the outset in a letter to the Visiting Committee, a letter which was
used as a circular in asking the public for funds. Two years later when
Harvard celebrated the Emerson anniversary, I delivered an address on
Emerson as philosopher. This epistemological paper may seem far removed
from the interests of the Harvard Psychological Studies, and yet I am
glad to print it in this laboratory volume, and thus emphatically to
indicate that I for one consider philosophy the true basis for the
psychologist.

There follow thus, first, the letter to the Visiting Committee, with
which the Emerson Hall movement took its official inception in 1901;
secondly, the address delivered at Harvard on the celebration of the
Emerson anniversary in May, 1903; thirdly, the paper contributed to the
debate of the philosophers at the opening meeting in December, 1905;
and, finally, a description of the present status of the laboratory in
January, 1906.


II. THE NEED FOR EMERSON HALL

  [The letter addressed to the Visiting Committee of the Overseers of
  Harvard University, in 1901, reads as follows:]

Gentlemen,--The philosophical work in Harvard has in the last twenty
years gone through an inner development which has met with a hearty
response alike on the part of the University and of the students. The
students have attended the courses in constantly growing numbers, the
Governing Boards have provided the Division amply with new teachers,
steadily increasing the number of professors, instructors, and
assistants. The outer growth of the Division has thus corresponded
most fortunately to the internal development, by an harmonious
coöperation of the administration, the teachers, and the students of
the University. And yet there remains one other factor as an essential
condition for the healthy life of the Department, a factor which cannot
be provided by the University itself and for which the help must come
from without. Our work needs a dignified home where under one roof all
the varied philosophical work now carried on at Harvard may be united.
The need has been urgently felt for many years, but only with the
recent growth has the situation become intolerable. It is therefore the
unanimous opinion of the Department that we must ask the public for the
funds to build at Harvard a "School of Philosophy," in the interest of
the students and of the teachers, in the interest of the Department and
of the University, in the interest of culture and of scholarship.

The present work of the Division of Philosophy can be indicated by
a few figures. We entered the current year with a teaching-staff of
six full professors, two assistant professors, four instructors, two
teaching-fellows, and six assistants. The instruction of these twenty
men covers the ground of history of philosophy, metaphysics, theory
of knowledge, psychology, logic, ethics, æsthetics, philosophy of
religion, philosophy of science and sociology. Thirty-two courses
have been offered. These courses are grouped in three classes: the
introductory courses, intended primarily for Sophomores and Juniors;
the systematic and historic courses, planned for Juniors, Seniors,
and Graduates; and the research courses for Graduates only. But
the students whom we try to reach differ not only with regard to
their classes, their corresponding maturity, and their degree of
philosophical preparation, but also with regard to the aims and
interests for which they elect philosophical studies in the University.
The one group seeks in our field liberal education. The fundamental
problems of life and reality, and the historic solutions of them
which the great thinkers developed, the values of truth and beauty
and morality, the laws of the mental mechanism and of the social
consciousness, all these promise and prove to be incomparable agencies
for widening the soul and giving to our young men depth, strength, and
ideals. Not a few of the students who belong to this group remain loyal
to philosophy through three or four years. A second group of students
need our courses as preparation for divers scholarly or practical
aims. The future lawyer, teacher, physician, minister, scientist, or
philanthropist knows that certain courses in ethics or psychology, in
education or logic afford the most solid foundations for his later
work; there is hardly a course in our Division which is not adjusted to
some kind of professional study. The third group finally, naturally the
smallest, but to the teachers the most important, consists of those to
whom philosophy itself becomes a life's work. The Harvard Department
believes that there is nowhere else in this country or abroad such
an opportunity for systematic and all-round training for an advanced
student of philosophy as is offered here, covering easily a man's full
work for six years, advancing from the introductory courses of the
Sophomore year to the six seminaries of the graduate years and finally
reaching the doctor's thesis in the third year after graduation.

The extent to which the Harvard students make use of these
opportunities is to be inferred from the figures which the last Annual
Report of the President offers. These refer to the year 1899-1900; the
current year will show somewhat the same proportions, perhaps even
an increase of graduate work. The figures are necessarily too low,
inasmuch as they refer merely to those students who take examinations
in the courses and omit those who merely attend the lectures. The
attendance in the philosophical courses was last year over one thousand
students. They belonged to all parts of the University, 188 Graduates,
210 Seniors, 218 Juniors, 175 Sophomores, 59 Specials, 57 Scientifics,
55 Divinity students, and the rest from the Freshman class, the Law
School, and the Medical School. The introductory courses were attended
by almost four hundred students, that is, by a number corresponding to
the size of the Junior class. As, in spite of natural fluctuations,
this figure is pretty constant,--in 1897 reaching its maximum with
427,--it can be said that in Harvard under the system of absolutely
free election practically every student who passes through Harvard
required of himself at least a year of solid philosophical study.

An even higher interest, however, belongs to the figures which refer
to the most advanced courses offered, especially to the courses of
research. It has always been the most characteristic feature of the
Harvard Philosophical Department to consider the advancement of
knowledge as its noblest function. The productive scholarship of the
Department is shown by the fact that the last two years alone brought
before the public eight compendious scholarly works from members of our
Department, besides a large number of smaller contributions to science.
To train also in the students this highest scholarly attitude, that
of the critical investigator as contrasted with that of the merely
receptive hearer of lectures, is thus the natural aim of our most
advanced work; it is this spirit which has given to the Department its
position in the University and in the whole country. This prevalence
of the spirit of research is the reason why, as the Report of the Dean
of the Graduate School points out, the Philosophical Department has a
larger number of graduate students who have carried on graduate studies
elsewhere than any other Department of the University. The table of
the Dean which records these migrating graduate students who come to
us for advanced work after graduate studies at other universities,
is as follows: Mathematics 6, Natural History 7, Political Science
7, Modern Languages 11, Classics 14, History 15, English Literature
16, Philosophy 20. If we consider the whole advanced work of the
University, that is the totality of those courses which are announced
as "primarily for Graduates," we find that the following number of
graduate students, including the graduate members of the professional
schools, have taken part: Classics 103, Philosophy 96, English 75,
German 61, History and Government 52, Romance Languages 45, Mathematics
39, Economics 23, Chemistry 21, in the other departments less than
twenty. But this situation turns still more strongly in favor of
philosophy as soon as we consider the technical research courses, those
which in the language of the catalogue are known as the 20-courses, and
omit those graduate courses which are essentially lecture courses. In
these research courses the number of Graduate Students is: Philosophy
71, History and Government 34, Chemistry 13, Zoölogy 12, Geology 10,
and in the other departments less than ten.

These few figures may be sufficient to indicate not only the extent
of the Department and its influence, but above all the harmonious
character of this development. The most elementary courses, the solid
routine courses, and the most advanced courses, show equal signs of
growth and progress, and the whole work with its many side branches
remains a well-connected unity with a clear systematic plan. All this
must be understood before one can appreciate the striking contrast
between the work and the workshop. It is of course easy to say at once
that the truth of a metaphysical thought does not depend upon the
room in which it is taught, and that the philosopher is not, like a
physicist or chemist, dependent upon outer equipments. Yet, this is but
half true, and the half of the statement which is false is of great
importance.

The dependence upon outer conditions is perhaps clearest in the case
of psychology, which has been for the last twenty-five years an
objective science with all the paraphernalia of an experimental study:
the psychologist of to-day needs a well-equipped laboratory no less
than the physicist. Harvard has given the fullest acknowledgment to
this modern demand and has spent large sums to provide the University
with the instruments of an excellent psychological laboratory; the
one thing which we miss is room, simply elbow-room. Our apparatus is
crowded in the upper story of Dane Hall, and even that small story
must give its largest room for the lectures of other departments and
another room to a philosophical reading-room. The space which remains
for the psychological work is so absolutely out of proportion to the
amount of work going on that the problem how to bring all the men into
those few rooms has become the most difficult of all our laboratory
problems. During the current year, besides the training-courses,
twenty-three men are engaged there in original research, each one with
a special investigation and each one anxious to devote as much time as
possible to his research; only the most complicated adjustment makes
it possible at all, and yet the mutual disturbance, the necessity of
passing through rooms in which other men are working, and of stopping
the work when other men need the place interfere every day with the
success of the instruction. A mechanical workshop is an urgent need of
our laboratory, and yet we cannot afford the room; and while the only
desirable arrangement would be to have the psychological lectures in
the same building where the apparatus is stored,--as the instruments
are necessary for the experimental demonstrations,--there is no room
for the lectures under the roof of Dane Hall, which houses the Bursar's
Office and Coöperative Stores. The result is that the instruments
must be carried through the yard in rain or shine, an effective way
to damage our valuable equipment. But the evils connected with the
present locality of the psychological laboratory are not only such
as result from its narrowness. Its position on Harvard Square, with
the continuous noise and the vibration of the ground, is perfectly
prohibitory for large groups of psychological studies and disturbing
for every kind of work for which concentration of attention is a
fundamental condition. Finally a psychological laboratory, perhaps
still more than a physical one, needs in its whole construction a
perfect adaptation to its special purpose; the walls, the shape, and
the connection of the rooms, everything must be built, as has been
done in other universities, for the special end. We have merely the
rooms of the old Law School with thin partitions dividing them. In
short everything is in a state which was tolerable during the last few
years only because it was felt as provisional, but the time when the
psychological laboratory must have really adequate quarters cannot be
postponed much longer.

The needs of the psychological work can thus be easily demonstrated to
every beholder; but while perhaps less offensive on the surface, the
outer conditions of the other branches of the Philosophical Department
are not therefore less unsatisfactory. The advanced student of logic
or ethics does not need a laboratory, but he needs seminary-rooms with
a working library where his work may have a local centre, where he
can meet his instructors and his fellow students engaged in related
researches, where he may leave his books and papers. To-day all this
theoretical work has no home at all; the seminaries seek refuge in
an empty room of the laboratory at a late evening hour, in a chance
lecture-room, or in private homes; there is nowhere continuity, no
place to collect or to deposit, no opportunity to meet beyond official
hours, no feeling of coherence suggested by surroundings. The most
advanced research work of the country is thus done under external
conditions which suggest the spirit of a schoolroom, conditions which
deprive students and instructors equally of the chance to make our
seminaries the fitting forms for their rich content. But if all this is
most deeply felt by the advanced students, it is not less true and not
less deplorable for the undergraduate courses. There is nowhere fixity
of association between the work and the room. The philosophy courses
are scattered over the whole yard, wandering each year from one quarter
to the other, creeping in wherever a vacant room can be found, not even
the instructors knowing where their nearest colleagues are meeting
students. The dignity and the unity of the work are equally threatened
by such a state of affairs. There remains not even a possibility for
the instructor to meet his students before or after the lecture; his
room is filled up to the time when he begins and a new class rushes
in before he has answered questions. A business-like restlessness
intrudes into the instruction, and yet philosophy above all needs a
certain repose and dignity.

Thus what we need is clear. We need a worthy monumental building at a
quiet central spot of the Harvard yard, a building which shall contain
large and small lecture-rooms, seminary-rooms, a reading-room, and
one whose upper story shall be built for a psychological laboratory,
so that under one roof all the philosophical work, metaphysical
and ethical, psychological and logical, may be combined. Here the
elementary and the advanced work, the lecture courses and the
researches, the seminaries and the experiments, the private studies in
the reading-room and the conferences and meetings of the assistants
would go on side by side. Here would be a real school of philosophy
where all Harvard men interested in philosophy might find each other
and where the students might meet the instructors.

Such a home would give us first, of course, the room and the
external opportunities for work on every plane; it would give us
also the dignity and the repose, the unity and the comradeship of a
philosophical academy. It would give us the inspiration resulting from
the mutual assistance of the different parts of philosophy, which
in spite of their apparent separation are still to-day parts of one
philosophy only. All this would benefit the students of philosophy
themselves, but not less good would come to the University as a
whole. The specialization of our age has brought it about that in
the organization of a university, even philosophy, or rather each of
the philosophical branches, has become an isolated study coördinated
with others. The average student looks to psychology as to physics
or botany; he thinks of ethics as he thinks of economics or history;
he hears about logic as coördinated with mathematics, and so on. The
University has somewhat lost sight of the unity of all philosophical
subjects and has above all forgotten that this united philosophy is
more than one science among other sciences, that it is indeed the
central science which alone has the power to give inner unity to the
whole university work. Every year our universities reward our most
advanced young scholars of philology and history, of literature and
economics, of physics and chemistry, of mathematics and biology with
the degree of Ph.D., that is of Doctor Philosophiae, thus symbolically
expressing that all the special sciences are ultimately only branches
of philosophy; but the truth of this symbol has faded away from the
consciousness of the academic community. All knowledge appears there as
a multitude of scattered sciences and the fact that they all have once
been parts of philosophy, till one after the other has been dismissed
from the mother arms, has been forgotten. A school of philosophy as a
visible unity in the midst of the yard will renew this truth and thus
give once more to the overwhelming manifoldness of intellectual efforts
of our University a real unity and interconnection; the external
connection of administration will be reënforced by the inner unity of
logical interdependence.

The time is ripe for a school of philosophy to play this rôle and to
fulfil again its old historical mission, to be the unifying principle
of human knowledge and life. The second half of the nineteenth century
was essentially controlled by realistic energies, by the spirit of
analysis, by the triumph of natural science and technique. But a long
time before the century came to an end a reaction started throughout
the whole intellectual globe; the synthetic energies again came to
the foreground, the idealistic interests were emphasized in the most
different quarters; the historical and social sciences make to-day the
same rapid progress which fifty years ago characterized the natural
sciences, and everywhere in the midst of the empirical sciences
there is awakening again the interest in philosophy. In the days of
anti-philosophical naturalism scientists believed that philosophy
had come to an end and that an unphilosophical positivism might
be substituted for real philosophy; to-day the mathematicians and
physicists, the chemists and biologists, the historians and economists
eagerly turn again and again to philosophy, and on the borderland
between philosophy and the empirical sciences they seek their most
important problems and discussions. The world begins to feel once more
that all knowledge is empty if it has no inner unity, and that the
inner unity can be brought about only by that science which enquires
into the fundamental conceptions and methods of thought with which the
special sciences work, into the presuppositions and ultimate axioms
with which they begin, into the laws of mental life which lie at the
basis of every experience, into the ways of teaching the truth, and
above all into the value of human knowledge, its absolute meaning and
its relation to all the other human values--those of morality, beauty,
and religion. The most advanced thinkers of our time are working to-day
in all fields of knowledge to restore such a unity of human life
through philosophy. To foster this spirit of the twentieth century in
the life of our University there is no more direct way possible than to
give a dignified home to the philosophical work. Such a building ought
to be a Harvard Union for scholarly life.

The beautiful building which we see in our minds should not be
devoted to a single system of philosophy. In its hall we hope to see
as greeting for every student the busts of Plato the Idealist and
Aristotle the Realist, of Descartes and Spinoza, of Bacon and Hobbes,
of Locke and Hume and Berkeley, of Kant and Fichte and Hegel, of Comte
and Spencer, of Helmholtz and Darwin. The School of Philosophy will
be wide open to all serious thought, as indeed the members of the
Department to-day represent the most various opinions and convictions.
This ought never to be changed; it is the life-condition of true
philosophy. Yet there is one keynote in all our work: a serious,
critical, lofty idealism which forms the background of the whole
Department and colors our teaching from the elementary introductions
to the researches of our candidates for the doctor's degree. All the
public utterances which have come from the Department in recent years
are filled with this idealism, in spite of the greatest possible
variety of special subjects and special modes of treatment. Here belong
The Will to Believe and the Talks to Teachers, by William James, the
Noble Lectures and the Glory of the Imperfect, by George Herbert
Palmer, Poetry and Religion, by George Santayana, The Principles of
Psychology, and Psychology and Life, by Hugo Münsterberg, Jesus Christ
and the Social Question, by Francis Peabody, Educational Aims and
Educational Values, by Paul Hanus, Shaftesbury, by Benjamin Rand, the
Conception of God, and The World and the Individual, by Josiah Royce.

We have sought a name which might give symbolic expression to this
underlying sentiment of idealism and might thus properly be connected
with the whole building. It cannot be that of a technical philosopher.
Such a name would indicate a prejudice for a special system of
philosophy, while we want above all freedom of thought. It ought to be
an American, to remind the young generation that they do not live up
to the hopes of the School of Philosophy if they simply learn thoughts
imported from other parts of the world, but that they themselves as
young Americans ought to help the growth of philosophical thought. It
ought to be a Harvard man--a man whose memory deserves that his name
be daily on the lips of our students, and whose character and whose
writing will remain a fountain of inspiration. Only one man fulfils all
these demands perfectly: Ralph Waldo Emerson. It is our wish and hope
that the new, dignified, beautiful home of philosophy may soon rise as
the moral and intellectual centre of Harvard University and that over
its doors we shall see the name: Emerson Hall--School of Philosophy.


III. EMERSON AS PHILOSOPHER

  [The following address was delivered at Harvard University, May,
  1903, as part of the Emerson Celebration:]

At the hundredth anniversary of Emerson's birthday, Harvard University
is to take a noble share in the celebration. For years it has been one
of the deepest desires of the Harvard community to erect in the college
yard a building devoted to philosophy only. To-day this building is
secured. To be sure, the good-will of the community must still do much
before the funds allow the erection of a building spacious enough to
fulfil our hopes; but whether the hall shall be small or large, we know
to-day that it will soon stand under the Harvard elms and that over
its door will be inscribed the name: Ralph Waldo Emerson. No worthier
memorial could have been selected. Orations may be helpful, but the
living word flows away; a statue may be lasting, but it does not awaken
new thought. We shall have orations and we shall have a statue, but
we shall have now, above all, a memorial which will last longer than
a monument and speak louder than an oration: Emerson Hall will be a
fountain of inspiration forever. The philosophical work of Harvard
has been too long scattered in scores of places; there was no unity,
philosophy had no real home. But Emerson Hall will be not only the
workshop of the professional students of philosophy, will be not only
the background for all that manifold activity in ethics and psychology,
in logic and metaphysics, in æsthetics and sociology, it will become
a new centre for the whole University, embodying in outer form the
mission of philosophy to connect the scattered specialistic knowledge
of the sciences. Harvard could not have offered a more glorious gift to
Emerson's memorial.

But the spirit of such a memorial hour demands, more than all,
sincerity. Can we sincerely say that the choice was wise, when we look
at it from the point of view of the philosophical interests? It was
beautiful to devote the building to Emerson. Was it wise, yes, was it
morally right to devote Emerson's name to the Philosophy Building?
Again and again has such a doubt found expression. Your building, we
have heard from some of the best, belongs to scientific philosophy; the
men who are to teach under its roof are known in the world as serious
scholars, who have no sympathy with the vague pseudo-philosophy of
popular sentimentalists; between the walls of your hall you will have
the apparatus of experimental psychology, and you will be expected to
do there the most critical and most consistent work in methodology
and epistemology. Is it not irony to put over the door, through which
daily hundreds of students are to enter, the name of a man who may be
a poet and a prophet, a leader in literature and a leader in life,
but who certainly was a mystic and not a thinker, an enthusiast but
not a philosopher? Not only those who belittle him to-day and who
short-sightedly deny even his immense religious influence, but even
many of Emerson's warmest admirers hold such an opinion. They love
him, they are inspired by the superb beauty of his intuitions, but
they cannot respect the content of his ideas, if they do not wish to
deny all their modern knowledge and scientific insight. Yes, for the
most part they deny that his ideas form at all a connected whole;
they are aphorisms, beautiful sparks. Did he not himself say: "With
consistency a great soul has simply nothing to do. He may as well
concern himself with his shadow on the wall." And yet how can there
be philosophy without consistency; how can we interpret reality if we
contradict ourselves? If Emerson's views of the world did really not
aim at consistency and did really ignore our modern knowledge, then
it would be better to go on with our philosophical work in Harvard
without shelter and roof than to have a hall whose name symbolizes both
the greatest foe of philosophy, the spirit of inconsistency, and the
greatest danger for philosophy, the mystic vagueness which ignores real
science.

But Emerson stands smiling behind this group of admirers and says,
"To be great is to be misunderstood." Yes, he did say, "A foolish
consistency is the hobgoblin of little minds, adored by little
statesmen and philosophers and divines;" but he soon adds, "Of one
will the actions will be harmonious however unlike they seem." Emerson
despises the consistency of the surface because he holds to the
consistency of the depths, and every sentence he speaks is an action of
the one will, and however unlike they seem they are harmonious, and, we
can add, they are philosophical; and, what may seem to these anxious
friends more daring, they are not only in harmony with each other,
they are in deepest harmony with the spirit of modern philosophy, with
a creed which ought to be taught by the most critical scholars of
Harvard's Philosophy Hall.

What is the essence of Emerson's doctrine in the realm of philosophy?
It seems like sacrilege to formulate anything he said in the dry terms
of technical philosophy. We must tear from it all the richness and
splendor of his style, we must throw off the glory of his metaphor, and
we must leave out his practical wisdom and his religious emotion. It
seems as if we must lose all we love. It is as if we were to take a
painting of Raphael and abstract not only from the richly colored gowns
of the persons in it, but from their flesh and blood, till only the
skeletons of the figures remained. All beauty would be gone, and yet we
know that Raphael himself drew at first the skeletons of his figures,
knowing too well that no pose and no gesture is convincing, and no
drapery beautiful if the bones and joints fit not correctly together.
And such a skeleton of theoretical ideas appears not only without
charm, it appears necessarily also uninteresting, without originality,
commonplace. All the philosophies, from Plato to Hegel, brought down to
their technical formulas, sound merely like new combinations of trivial
elements, and yet they have made the world, have made revolutions and
wars, have led to freedom and peace, have been mightier than traditions
and customs; and it is true for every one of them that, as Emerson
said, "A philosopher must be more than a philosopher."

There are, it seems, three principles of a philosophical character
without which Emerson's life-work cannot be conceived. To bring
them to the shortest expression we might say, Nature speaks to us;
Freedom speaks in us; the Oversoul speaks through us. There is no
word in Emerson's twelve volumes which is inconsistent with this
threefold conviction, and everything else in his system either follows
immediately from this belief or is a non-essential supplement. But
that threefold faith is a courageous creed indeed. The first, we said,
refers to Nature; he knew Nature in its intimacy, he knew Nature in
its glory; "Give me health and a day and I will make the pomp of
emperors ridiculous." And this Nature, that is the assertion, is not
what natural sciences teach it to be. The Nature of the physicist, the
dead world of atoms controlled by the laws of a dead causality, is not
really the Nature we live in; the reality of Nature cannot be expressed
by the record of its phenomena, but merely by the understanding of its
meaning. Natural science leads us away from Nature as it really is. We
must try to understand the thoughts of Nature. "Nature stretches out
her arms to embrace man; only let his thoughts be of equal greatness;"
and again Emerson says, "All the facts of natural history taken by
themselves have no value, but are barren like a single sex; but
marry it to human history and it is full of life;" and finally, "The
philosopher postpones the apparent order of things to the empire of
Thought."

And in the midst of Nature, of the living Nature, we breathe in
freedom; man is free. Take that away and Emerson is not. Man is free.
He does not mean the freedom of the Declaration of Independence, a
document so anti-Emersonian in its conception of man; and he does not
mean the liberty after which, as he says, the slaves are crowing while
most men are slaves. No, we are free as responsible agents of our
morality. We are free with that freedom which annuls fate; and if there
is fate, then freedom is its most necessary part. "Forever wells up the
impulse of choosing and acting in the soul." "So far as man thinks he
is free." "Before the revelations of the soul, time, space, and nature
shrink away." "Events are grown on the same stem with the personality;
they are sub-personalities." "We are not built like a ship to be
tossed, but like a house to stand." This freedom alone gives meaning to
our life with its duties, and puts the accent of the world's history on
the individual, on the personality: "All history resolves itself very
easily into the biography of a few stout and earnest persons," and "An
institution is the lengthened shadow of a man."

Nature speaks to us, Freedom speaks in us, but through us speaks a Soul
that is more than individual, an over-individual soul, an "Oversoul,
within which every man is contained and made one with all others."
Now even "Nature is a great shadow, pointing always to the sun behind
her." Every one of us belongs to an absolute consciousness which in us
and through us wills its will; "Men descend to meet" and "Jove nods to
Jove from behind each of us." Yes, "Man is conscious of a universal
soul within or behind his individual life, wherein as in a firmament
justice, truth, love, freedom arise and shine." The ideals, the duties,
the obligations, are not man's will but the will of an Absolute.

Does not all this sound like a wilful denial of all that has been fixed
by the sciences of our time? Does not every Sophomore who has had his
courses in Physics, Psychology, and Sociology know better? He knows,
we all know, that the processes of Nature stand under physical laws,
that the will of man is the necessary outcome of psychological laws,
that the ideals of man are the products of human civilization and
sociological laws. And if every atom in the universe moves according
to the laws which physics and chemistry, astronomy and geology, have
discovered, is it not anti-scientific sentimentality to seek a meaning
and thoughts in the mechanical motions of the dead world of substance?
So the poet may speak, but we ought not to say that his fanciful dreams
have value for scholarly philosophers. The philosophy of the scientist
ought to be the acknowledgment that matter and energy, and space and
time are eternal, and that the smallest grain of sand and the largest
solar system move meaningless by blind causality.

And emptier still is the naïve belief that man is free. Do we not
profit from decades of psychological labor, whereby the finest
structure of the brain has been discovered, wherein the psychological
laws have been studied with the exactitude of a natural science,
wherein we have studied the mental life of animals and children, and
have observed the illusions of freedom in the hypnotized man and in
the insane? Yes, we know to-day that every mental act, that every
psychological process is the absolutely necessary outcome of the given
circumstances; that the functions of the cells in the cortex of the
brain determine every decision and volition, and that man's deed is as
necessary as the falling of the stone when its support is taken away.
Yes, modern psychology does not even allow the will as an experience of
its own kind; it has shown with all the means of its subtle analysis
that all which we feel as our will is only a special combination of
sensations which accompany certain movement-impulses in our body. Can
we still take it seriously, when the philosopher steps in and pushes
sovereignly aside all the exact knowledge of mankind, and declares
simply "Man's will is free!"

Finally, the claim for the over-personal, absolute consciousness in
man. It is a triumph of modern science to understand how the duties and
ideals have grown up in the history of civilization. What one nation
calls moral is perhaps indifferent or immoral for another people or
for another time; what the one calls beautiful is ugly for the other;
what one period admires as truth is absurdity for another; there is no
absolute truth, no absolute beauty, no absolute religion, no absolute
morality; and sociology shows how it was necessary that just these
ideals and just these obligations should have grown up under a given
climate and soil, a given temperament of the race, a given set of
economical conditions, a given accumulation of technical achievements.
Man has made his Absolute, not the Absolute made man, and whatever
hopes and fears make men believe, the scholarly mind cannot doubt that
these beliefs and idealizations are merely the products of the feelings
and emotions of individuals bound together by equal conditions of life.
Leave it to the raptures of the mystic to ignore all scientific truth,
to get over-soul connection beyond all experience. In short, to accept
Emerson's philosophy, the scientist would say, means to be a poet where
Nature is concerned, means to be ignorant where man is concerned, and
means to be a mystic where moral and religious, aesthetic and logical
ideals are concerned. Can such be the herald of modern philosophy?

But those who are so proud and so quick are not aware that the times
have changed and that their speech is the wisdom of yesterday. In the
history of human knowledge the periods alternate. Great waves follow
each other, and while one tendency of scientific thought is ebbing,
another is rising; and there is no greater alternation than that
between positivism and idealism. The positivistic period of natural
science has ebbed for ten or fifteen years; an idealistic one is
rising. Emerson once said here in Harvard that the Church has periods
when it has wooden chalices and golden priests, and others when it
has golden chalices and wooden priests. That is true for the churches
of human knowledge too, and for knowledge of all denominations.
Forty, fifty years ago, in the great period when Helmholtz discovered
the conservation of energy and Darwin the origin of species, one
naturalistic triumph followed the other, golden high priests of
natural science were working with wooden chalices in narrow, awkward
laboratories; to-day natural science has golden chalices provided
in luxurious institutions, but there are too many wooden priests.
The fullest energies of our time are pressing on to an idealistic
revival, are bringing about a new idealistic view of the world, and
turning in sympathy to that last foregoing period of idealism of
which Ralph Waldo Emerson was perhaps the last original exponent.
But also with his period idealism was not new. When he came to speak
on the Transcendentalist, he began, "The first thing we have to say
respecting the new views here in New England is that they are not new."
Yes, indeed; since the beginnings of Greek philosophy, more than two
thousand years ago, the two great tendencies have constantly followed
each other. Each one must have its time of development, must reach its
climax, must go over into undue exaggeration, and thus destroy itself
to make room for the other, which then begins in its turn to grow, to
win, to overdo, and to be defeated.

Glorious had been the triumph of Positivism in the middle of the
eighteenth century when the French encyclopædists were at work, those
men who wrote the decrees for the French Revolution. But before the
last consequences of the Positivism of the eighteenth century were
drawn, the idealistic counter-movement had started. Immanuel Kant
gave the signal, he fired the shot heard round the world; and Fichte
followed, whose ethical Idealism changed the map of Europe, and his
spirit went over the Channel to Carlyle, and finally over the ocean to
these shores of New England and spoke with the lips of Emerson. It is
unimportant whether Emerson studied the great transcendental systems
in the original; he knew Kant and Schelling probably at first through
Coleridge, and Fichte through Carlyle. But in the mean time Idealism
too had exaggerated its claims, it had gone forward to Hegel, and
while Hegelian thought, about 1830, held in an iron grasp the deepest
knowledge of his time, his neglect of positive experience demanded
reaction, a counter-movement became necessary, and in the midst of
the nineteenth century the great idealistic movement with all its
philosophical and historical energies went down, and a new Positivism,
full of enthusiasm for natural science and technique and full of
contempt for philosophy, gained the day. With logical consistency,
the spirit of empiricism went from realm to realm. It began with the
inorganic world, passed into physics, then forward to chemistry, became
more ambitious and conquered the world of organisms, and when biology
had said its positivistic say, turned from the outer nature of being
to the inner nature. The mind of man was scrutinized with positivistic
methods; we came to experimental psychology, and finally, as the
highest possible aim of naturalism, to the positivistic treatment of
society as a whole, to sociology. But naturalism again has overdone its
mission, the world has begun to feel that all the technique and all the
naturalistic knowledge makes life not more worth living, that comfort
and bigness do not really mean progress, that naturalism cannot give
us an ultimate view of the world. And above all, the reaction has come
from the midst of the sciences themselves. Twenty years ago scientific
work received its fullest applause for the neglect of philosophical
demands. Ten years ago the feeling came up that there are after all
problems which need philosophy, and to-day philosophers, with good or
bad philosophy, are at work everywhere. The physicists, the chemists
and the biologists, the astronomers and the mathematicians, the
psychologists and the sociologists, the historians and the economists,
the linguists and the jurists, all are to-day busily engaged in
philosophical enquiries, in enquiries into the conditions of their
knowledge, into the presuppositions and methods of their sciences, into
their ultimate principles and conceptions; in short, without a word
of sudden command, the front has changed its direction. We are moving
again towards philosophy, towards Idealism, towards Emerson.

Does all this mean that we are to forget the achievements of natural
science, and ignore the results of empirical labor, of labor which has
given us an invincible mastery of stubborn nature and an undreamed-of
power to calculate all processes of the physical and of the psychical
world? No sane man can entertain such a notion. Yes, such ideas
would contradict the laws which have controlled the alternation
of Idealism and Positivism through the ages of the past. Whenever
Positivism returned, it always showed a new face, and the teaching of
the intervening period of Idealism was never lost. The naturalism of
the middle of the nineteenth century was not at all identical with the
naturalism of the middle of the eighteenth; and so Idealism too, as
often as it returned to mankind after periods of neglect and contempt,
had every time gained in meaning, had every time found increased
responsibilities, had every time to do justice to the new problems
which the preceding period of Positivism had raised. If Idealism
to-day wants to gain new strength, nothing must be lost of all that
the last fifty years have brought us, no step must be taken backward,
the careful scientific work of the specialists must be encouraged and
strengthened, and yet the totality of this work must be brought under
new aspects which allow a higher synthesis; yes, a higher synthesis
is the problem of the philosopher of to-day. He does not want to be
ignorant of natural science and simply to substitute idealistic demands
in the place of solid, substantial facts; and he should feel ashamed
of the foul compromise with which half-thinkers are easily satisfied,
a compromise which allows science its own way till it comes over
the boundaries of human emotions, a compromise which accepts rigid
causality but pierces little holes in the causal world, making little
exceptions here and there that human freedom may be saved in the midst
of a world-machinery; a compromise which accepts the social origin
of ideals, but claims a mystic knowledge that just our own private
pattern will remain in fashion for eternity. No philosophy can live
by compromises. If natural science is to be accepted and Idealism is
to hold its own, they must be combined, they must form a synthesis in
which the one no longer contradicts the other. Just such synthetic
harmonization, and not at all a stubborn ignorance of the other side
or a compromise with cheap concessions, was the aim of the period from
Kant to Emerson. It is merely the naturalistic period which ignores its
idealistic counterpart, which delights in its one-sidedness, which is
afraid of harmony because it is suspicious of demands for concessions.
It is naturalism only which thinks that mankind can walk on one leg.

If we ask where such harmonization can be found, where the great
Idealists of the beginning of the last century have sought it, and
where our modern philosophy is seeking it again, well aware that by
the progress of science in the mean time the difficulties have been
multiplied, the logical responsibilities have become gigantic, we
cannot do more here than to point out the direction; we cannot go
the way. And it is clear, of course, too, that such an answer has its
individual shape, and that no one can promise to give a bird's-eye
view of the marching movement while he is himself marching among his
comrades. But the individual differences are non-essential. The one
great tendency, the Emersonian spirit, if it is rightly understood,
is common to them all. What has modern philosophy all over the world
to say about that threefold claim concerning Nature, Freedom, and
Oversoul? What has it to say when natural science has fully said its
say and had its fair hearing, and has been approved as sound and
welcome?

A philosopher might answer, perhaps, as follows: You Positivists have
done wonderfully with your microscopes and your telescopes, with your
chronoscopes and spectroscopes; you have measured and weighed and
analyzed and described, and finally explained the whole world which you
perceive, and there is nothing in space and time and causality which
can escape your search. But did not all that work of yours involve
certain presuppositions which you had accepted and which it was not
your business to look on critically, but which, nevertheless, may be
open to enquiry? Your first claims granted, all may follow; but how is
it with the first claims? You examine all that is in space and time,
but what are space and time? You examine the material substances and
the contents of consciousness, but what is consciousness, and what is
matter? You seek the special applications of causality, but what is
causality? Well, you reply, you give the facts just as you find them;
but do you do that really? And what do you mean by saying that you find
the facts? Let us look, at least for a moment, at the very simplest
facts with which your work begins. You say there are physical objects
made up of atoms, and you describe them as a physicist; and there are
mental ideas in consciousness made up of sensations, and you describe
them as a psychologist; and both, you say, you are finding. But what
does it mean, that you find the physical object outside there and the
mental idea of the object inside in you; is that really a statement
of your immediate experience? The physicist speaks of this table here
before me, outside of me; and the psychologist speaks of my idea of
this table, enclosed in my consciousness. Both may do well to speak
so; but will you make me believe that I find that doubleness in my
experience? If I see this table and want to use it, I am not aware
of one table of wooden stuff and another in me of mental stuff. I am
not aware of a two-ness at all, and if the physicist says that this
wooden table is made up of molecules and has in itself no color and
no continuity, and that the mental idea in me furnishes all those
qualities of color and smoothness, but has no solidity, then they
speak of two interesting worlds about which I am anxious to know, but
certainly neither of them is the world I live in. If I lean on this
table I am not aware of a table in my mind at all. I know the one table
only, and this one table has its color and its smoothness.

I know what you will answer. You will say, in your immediate experience
there are indeed not two worlds of objects, a physical and a
psychical; the real thing to which our interests in life refer is not
differentiated into a molecular object outside of us and a sensational
object in us, but it is clear that every real thing allows a kind of
double aspect; we can consider this table in so far as it is common to
all of us, in so far as it is a possible object for every one of us,
and in so far as it becomes an object for the individual, and we can
then call the objects, in so far as they are common property, physical;
and in so far as we take the aspect of individual relations, psychical;
and as it must be of the highest importance for our practical purposes
to discriminate between those two aspects, we have clearly the right
to consider the world from the point of view of both the physicist
and the psychologist. It is, of course, an abstraction if we leave
out in the one case the one side, in the other case the other side
of our objective experience; but we gain by that the possibility of
constructing two closed causal systems of which each one must have its
special conditions of existence, inasmuch as the one is conceived as
related to individuals and the other as independent of individuals.

Very true, we should answer. Something like that saves you completely,
justifies fully your claim to separate the physical and the psychical
worlds of objects, the world of matter and the world of ideas; but
can you deny that you have lost your case, are you not now yourself
in the midst of philosophical, methodological discussions, which your
physics and psychology themselves cannot settle, yet which must be
settled before they can enter into their rights; and above all, do you
not yourself see now that your whole physics, for instance, is not at
all an account of reality, but merely a certain logical transformation
of reality; that you do not find the world of physics at all, just as
little as you find the psychical ideas, but that you can merely work
over and reshape the reality which you find till you construct out
of it your world of matter and your world of consciousness? What you
believed you would find you have never found, while your construction
of physical things may have been most necessary for your purposes; but
do not deny that you have left reality far behind you.

And so it is with all your doings. You tell us proudly, for instance,
that you show us the deepest nature of the world by showing us the
elements which the object contains, and that you thus bring us at least
nearer to the essence of things; and yet if we begin to look into your
real achievements, we are disappointed again to find that you are far
away from even attempting anything of the kind. You tell us that water
is hydrogen and oxygen, and if we say "Prove it," you show us simply
that you can transform the water into hydrogen and oxygen, and that
you can transform these two elements again into water. Is that really
what you promise? We want to know what the thing is, and you show us
simply how the one thing can be transformed into another thing; and
whenever we turn to your wisdom, it is always the same story. You show
us always, and most nicely, how the one goes over into the other, but
you never show us what the one or the other really is in itself. For
your practical purposes the first may be the most important aspect,
but do not make us believe, therefore, that it is the only possible
aspect. In short, whether science describes or explains, it never
gives us what we find in reality, but makes out of reality a new ideal
construction in the service of certain purposes, and never gives us
the things as they are, but merely the effects and changes which they
produce. Are we still, then, to be deeply impressed with the claim
of the naturalist that he alone has the monopoly of knowing reality,
while we see now that every step of his leads us away from reality? And
have we still to be afraid to raise the voice as philosophers with the
claim that reality itself must find its expression, that there must be
a science which shall give account of reality as we really find it,
of nature before it is made up and repolished for the purposes of the
physicist? Only if we have such other account of nature, then only
do we speak of that nature in which we live and in which we act, and
compared with such an account of the fuller reality, the constructed
schematism of the physicist must appear, indeed, as Emerson said,
"barren like a single sex." Not the slightest result of natural science
is depreciated, not the slightest discovery ignored, if we insist that
all these so-called facts have a meaning only under certain artificial
conditions which set them apart from the reality of our life; and in
this reality lies the interest of the philosopher. We have thus no
reason to reproach the scientist so long as the scientist does not
fancy that his science gives an account of nature as it really is. Both
kinds of work are necessary, and the scientist may well speak, as the
squirrel in Emerson's poem:

    "Talents differ,
    All is well and wisely put;
    If I cannot carry forests on my back,
    Neither can you crack a nut."

Natural science has to crack our nuts, but philosophy has to carry
on its back the flourishing forests of life, in which we wander and
breathe. And if Emerson is right, to-day and forever, in claiming that
the facts of natural science are not expressions of reality, it is only
a small step to see that he was not less right in saying that man is
free. Consider man as a particle in the physical universe, consider his
actions from the point of view of a causal science, and there is no
possibility of escaping materialism and fatalism. We must understand
every activity as a necessary outcome of foregoing conditions.
Psychology must do so, and physics must do the same. The empirical
sciences would be disloyal to their own principles if they allowed
the slightest exception. The noblest gesture, the greatest word, the
bravest action, must be considered by them under the category of
causality. They are necessary effects of all the preceding causes. It
may be interesting, it may be fascinating to follow such lines with the
enthusiastic energy of scholarly research. But are we really obliged
to accept the outcome as an ultimate word concerning the meaning of
our freedom? "Forever wells up the impulse of choosing and acting in
the soul." Is it really merely an illusion? Has responsibility still
its moral value, are we the actors of our actions, are we still good,
are we still guilty, when every deed follows as necessary effect? Is
not, then, the whole constitution of the world, which has made us,
responsible whenever we move our hand for good or for bad?

But we know now where we are standing; we know now that the world
of objects, of psychical as well as of physical, is a constructed
world, constructed for the purpose of satisfying our demand for causal
connection; for that world holds causality because it is the world
seen from the point of view of causality; and just as there cannot be
anything in that world of physical and psychical objects which is not
causally connected, just so it cannot have any meaning at all to ask
for causal connection before the world is conceived in the service of
this artificial construction. Reality in itself is not causal, and to
ask for the causes of the real experience of our inner life has not
more meaning than to ask how many pounds is the weight of a virtue, and
how many inches is the length of our hopes. But we must go farther.
To apply the question of cause and effect to our real will means not
only that we apply to the real object a standard which belongs to
the artificial or constructed object, but it means above all that we
consider as an object something which in reality is not an object at
all. The will which the psychologist describes and must describe, the
will which has causes and which is thus not free, is a will conceived
as an object found in our mind like an idea, something of which we are
aware, something whose happening we perceive, and yet if anything is
sure it is the immediate experience that we are aware of our will in
a way which is absolutely different from the way in which we perceive
objects. We do not perceive our will at all, we will it, we strive
it, we fight it; yes, we feel ourselves, only in so far as we are the
subjects of will. Our will is our personality, which we do not find but
which we are, and which stands opposed and separated by the deepest
gulf from the world of objects. Those objects are means and purposes of
our will, are ends and aims and instruments; but they come in question
for us only as we will them, as we like and dislike them, as we approve
and reject them. And if we take this world of objects and reconstruct
it into the artificial world of physical and psychical things connected
by causality, in this very act of reconstruction we feel ourselves as
willing, deciding, approving, aiming personalities, whose wills decide,
who think the world as causally connected, whose freedom guarantees
the value of our conception of a world not free. There is no knowledge
but in our judgments; there is no judgment but in our affirming and
denying; there is no affirming and denying but in our will. Our will
chooses for its purposes to conceive reality as if it were unfree.
What a climax of confusion to think that this conception of an unfree
world, the conception of science, can itself now condemn the freedom
of the will which has chosen. "Freedom is necessary," said Emerson. We
can add, necessity itself is merely a purpose determined by freedom.
"Intellect annuls fate," Emerson says. We may add, fate is merely an
idea of intellect. Let us be psychologists if we want to analyze, to
calculate, to explain the unfree man; but let us be philosophers to
understand what it means to be a psychologist. Now the synthesis is
reached; the real world is free, but we choose for our purposes to
conceive the world as unfree, and thus to construct causal sciences.

And if we understand that in reality man is free and that the
psychological aspect of man as unfree is a special way of looking on
man for special purposes, then suddenly there opens itself before
us the vast field of history, and the historical life, which seemed
deprived of all interest by the psychological, iconoclastic mood,
suddenly wins again a new importance. We feel instinctively that
this free man of reality, this man who is a responsible actor of his
actions, he only is the agent of history; and history is falsified
and cheapened when it is brought down to a causal explanation of
psychological man instead of real man. History had become an appendix
of sociology, and what great historians aimed at in the interpretation
of the few "stout and earnest personalities" seemed lost in favor of a
construction in which the great man and the genius rank with the fool
as mere extreme variations of psychological averages. Now suddenly do
we understand that history has to deal with the world of freedom, that
it has not to explain, but to interpret, that it has not to connect
the facts by linking causes and effects, but by understanding the
meaning of purposes, their agreement and disagreement, their growth
and liberty. Now we understand why Fichte, why Carlyle, why Emerson
believes in heroes and hero-worship, why Idealism has been at all
times the fertile ground for writing history and for making history,
while Naturalism has made technique, and thought in an anti-historical
spirit. Our time begins again to think historically. It can do so
because it again begins to emancipate itself from its positivistic
disbelief in man's freedom and from its unphilosophic superstition that
causal science alone is science, that we know only when we explain.

And when we at last stand man to man in full freedom, no longer as
psycho-physical constructions but as free personalities, and when we
debate and try to convince each other, will you deny that Jove stands
behind each of us and Jove nods to Jove when we meet? Would it even
have a meaning for us to go on with our talk, should we try at all
to convince each other if you thought and I thought, each one for
himself, that our will is only our personal will, that there is no
over-individual will, no Oversoul behind us? Can we discuss at all if
we do not presuppose that there is really a truth which we are seeking
in common, that there are certain judgments which we are bound to will,
which we are obliged to affirm, which we will, but not as individuals,
and of which we take for granted that every one whom we acknowledge
at all as a personality must will them too; and if you come with the
flippant air of the sceptic and tell me, "No, there is no truth, all
is only as it appears to me, there is no objective truth," do you not
contradict yourself, are you not saying that at least this, your own
statement, expresses objective truth; that you will this with a faith
and belief that this will of yours is an over-individual will which
is, as such, a duty, an obligation for every one who thinks? Every
escape is futile. And all the over-individuality that lives in our
will towards truth comes to us again in our will towards morality. Do
not say sceptically that there is no absolute obligation, that you do
not feel bound by an over-individual will in your action, that you
will do in every moment what pleases you individually. You cannot even
speak this sceptical word without contradicting yourself again, as you
demand through the fact of your saying it that we believe that you
speak the truth and that you thus feel yourself bound not to lie. If
you leave us doubtful whether your word was not a lie, the word itself
cannot have any meaning. Do not try to dodge the Oversoul. Men live
and fight in its purposes, and men descend to meet. It is as Emerson
said, "At first delighted with the triumph of the intellect, we are
like hunters on the scent and soldiers who rush to battle; but when the
game is run down, when the enemy lies cold in his blood at our feet,
we are alarmed at our solitude." Let the sociologists triumphantly
reduce the ideals to necessary social products of evolution in the
same spirit in which the psychologist eliminates the freedom of the
individual; but let us never forget that such a social mechanism is as
much an artificial construction necessary for its purposes as is the
psycho-physical mechanism of individuality. In that reality with which
history deals, in which our freedom lies, there our over-individual
will comes from deeper ground than from the soil and the food and the
climate. Our logical obligations, our ethical duties, our æsthetic
appreciations, our religious revelations, in reality they do not come
from without, they come from within; but from within as far as we are
souls in the Oversoul. There is no duty in the world but the duty
which we will ourselves; no outer force, no training, no custom, no
punishment can make us have duties. Duty is our will, it may be the
duty to think for the ideal of truth, the duty to feel for the ideal
of æsthetics, the duty to act for the idea of morality, the duty to
have faith in the ideal of religion; but it is always our own will,
and yet not our fanciful, personal, individual will. It is a system
of purposes upon whose reality all knowledge of the world, and thus
the world as we know it, is dependent forever. The wave of Idealism
is rising. The short-sighted superstition of Positivism will not lurk
under the roof of a new hall of philosophy. To be a true student of
the most scientific, of the most scholarly, of the most insistent
philosophy means to respect and to study the sciences, the physical and
the psychical sciences, but at the same time to understand that natural
science is not the science of reality, that psychology does not touch
the freedom of man, that no life has a meaning without the relation to
the Oversoul. We cannot write a whole system and a whole text-book
on the front of the new building. It must be enough to write there a
symbolic word; happy, forever happy, the university which can write
over the door of its temple of philosophy the name: Ralph Waldo Emerson.


IV. THE PLACE OF EXPERIMENTAL PSYCHOLOGY

  [At the opening of Emerson Hall, December 27, 1905, the American
  Psychological Association discussed the relation of psychology to
  philosophy; I opened the discussion with the following remarks:]

From the whole set of problems which cluster about psychology and its
relation to neighboring sciences, this hour, in which Emerson Hall is
completed, and this room, in which I hope to teach psychology to the
end of my life, suggest to me most forcibly to-day the one question:
Have I been right in housing psychology under this roof? I might have
gone to the avenue yonder and might have begged for a psychological
laboratory in the spacious quarters of the Agassiz Museum, to live
there in peaceful company with the biologists; or I might have
persuaded our benefactors to build for me a new wing of the physical
laboratory. But I insisted that the experimental psychologists feel at
home only where logic and ethics, metaphysics and epistemology keep
house on the next floor.

I certainly do not mean that the psychologist ought to mix the records
of his instruments with the demands of his speculations, and that he
may seek help from the Absolute when the figures of the chronoscope
or the curves of the kymograph are doubtful. Experimental psychology
is certainly to-day and will be for all future an independent exact
discipline with its own problems and methods. No one can insist more
earnestly than I do on the demarcation line between the empirical study
of mental phenomena and the logical enquiry into the values of life.

Yet I deny that it is a personal idiosyncrasy of mine to try to combine
vivid interest in both. There is no antagonism between them; a man may
love both his mother and his bride. I am devoted to philosophy, just
as I love my native country; and I am devoted to psychology, just as I
love the country in which I do my daily work; I feel sure there is no
reason for any friction between them.

Of course, on the surface a psychological laboratory has much more
likeness to the workshop of the physicist. But that has to do with
externalities only. The psychologist and the physicist alike use subtle
instruments, need dark rooms and sound-proof rooms, and are spun
into a web of electric wires. And yet the physicist has never done
anything else than to measure his objects, while I feel sure that no
psychologist has ever measured a psychical state. Psychical states are
not quantities, and every so-called measurement thereof refers merely
to their physical accompaniments and conditions. The world of mental
phenomena is a world of qualities, in which one is never a multiple of
the other, and the deepest tendencies of physics and psychology are
thus utterly divergent.

The complicated apparatus is therefore not an essential for
the psychologist. Of course, we shall use every corner of our
twenty-four laboratory-rooms upstairs, and every instrument in the
new cases--and yet much of our most interesting work is done without
any paraphernalia. Three of the doctor-dissertations which our
psychological laboratory completed last year consisted of original
research in which no instruments were involved; they dealt with
memory-images, with associations, with æsthetic feeling, and so
on. Yes, when, a short time ago, a Western university asked me how
much it would cost to introduce a good practical training-course in
experimental psychology, I replied that it would cost them the salary
of a really good psychologist, and besides, perhaps, one dollar for
cardboard, strings, rulers, colored paper, wire, and similar fancy
articles at five cents apiece.

On the other hand, I do not know a psychological experiment which
does not need a philosophical background to bring its results into
sharp relief. Of course, you will say, the psychologist deals with
facts, not with theories, and has to analyze and to describe and to
explain those facts. Certainly he has to do all that; only he must
not forget that the so-called "fact" in psychology is the product of
complex transformations of reality. A will, an emotion, a memory-image,
a feeling, an act of attention, of judgment, of decision--these are
not found in the way in which stones and stars are noticed. Even if
I choose perceptions or sensations as material for my psychological
study, and still more when I call them _my_ perceptions and _my_
sensations, I mean something which I have found at the end of a long
logical road, not at its starting-point, and that road certainly leads
through philosophy. Emerson said wisely, "A philosopher must be more
than a philosopher;" we can add: A psychologist must be more than a
psychologist. First of all, he must be a philosopher.

What would be the result if our laboratory had moved to the
naturalistic headquarters? It would be the beginning of a complete
separation from philosophy. Our graduate students would flock to
psychological research work without even being aware that without
philosophical training they are mere dilettantes. And soon enough a
merely psychological doctorate would be demanded. I do not deny at
all that such pure psychologists would find enough to do; I should
doubt only whether they know what they are doing. There are too many
psychologists already who go their way so undisturbed only because they
walk like somnambulists on the edge of the roof; they do not even see
the real problem; they do not see the depths to which they may fall.

But does the laboratory itself gain from such divorce? Just the
contrary. It is evident that everywhere in the world where the
psychological laboratory turns to natural science, the experiments
deal mostly with sensation, perception, and reaction; while those
laboratories which keep their friendship with epistemology emphasize
the higher mental functions, experimenting on attention, memory,
association, feeling, emotion, thought, and so on. But is it not clear
that only the latter work gives to the psychological laboratory a real
right to existence, as the former is almost completely overlapped by
the well-established interests of the physiologists? If psychology
cannot do anything else than that which physiologists like Helmholtz,
Hering, Kries, Mach, Bowditch, and the rest have always done so
successfully, then experimental psychology had better give up the trade
and leave the study of the mind to the students of the organism.

I have said that we ought not to depend on authorities here. Yet
one name, I think, ought to be mentioned gratefully in this hour in
which the new psychological laboratory is opened for work. I think
of Professor Wundt of Leipzig. The directors of the psychological
laboratories in Columbia, and Yale, in Clark and Chicago, in
Pennsylvania and Cornell, in Johns Hopkins and Washington, in Leland
Stanford and Harvard, and many more are his pupils. Some weeks ago,
when I did not foresee our present discussion, I told him of Emerson
Hall; and a few days ago I got an answer from which, as my closing
word, I may quote in translation. Professor Wundt writes to me: "I am
especially glad that you affiliated your new psychological laboratory
to philosophy, and that you did not migrate to the naturalists.
There seems to be here and there a tendency to such migration, yet I
believe that psychology not only now, but for all time, belongs to
philosophy: only then can psychology keep its necessary independence."
Mr. Chairman, these are the words of the father of experimental
psychology. I hope they indicate the policy to which Harvard University
will adhere forever.


V. THE PSYCHOLOGICAL LABORATORY IN EMERSON HALL

A monumental staircase leads from the first--the lecture-room--floor
of Emerson Hall to the second, the library floor; at the two ends of
its broad corridor smaller staircases lead to the third floor, the
laboratory. Its general division of space is seen at a glance from the
sketch of the ground plan (opposite page 1). Eighteen rooms of various
sizes with outside windows form a circle around the central hall
which is well lighted by large skylights; but at each end of the hall
itself two large windowless spaces are cut off and each of these is
divided into three dark rooms. We have thus twenty-four rooms, besides
coat-room, toilet-rooms, etc. A further stair leads to the wide attic
which is mainly a store-room for the institution.

In order that the laboratory should be adaptable to the most diverse
purposes, the permanent differentiation of the rooms has been kept in
narrow limits. It seemed unwise to give from the first every room to a
special line of research, as the preponderance of special interests may
frequently shift; there are years when perhaps studies in physiological
and comparative psychology make the largest demand and others in
which studies in æsthetical and educational psychology stand in the
foreground. A thorough-going specialization, by which special rooms
are reserved for tactual studies and others for chronoscope work or
for kymograph researches, allows of course certain conveniences in the
fixed arrangement of instruments and a certain elaboration of equipment
that is built in, but it very much impairs the flexibility of the
whole laboratory, and has thus not seemed advisable for an institution
whose catholic attitude welcomes investigations as different as those
contained in this volume.

To be sure certain constant requirements have demanded a special
fitting up of one room as a workshop, one room for the more
delicate instruments, one for the beginning course in experimental
work, a lecture-room for the courses in comparative psychology, a
photography-room, a battery-room, a sound-proof room, the chief animal
rooms, and the dark rooms. We have seven light-proof rooms, finished
in black, of which two have outside windows for heliostats; of the
others, four can be used for optical research; the longest one contains
the photometer. Six other rooms, including the lecture-room, may be
darkened by opaque blinds. One contains a partition with door and a
grooved window-frame fitted with screens in which openings of any
desired size and shape may be cut. This window is opposite the main
door of the room, and opposite this, across the central hall, some
sixty feet away, is the door of another dark room; optical stimuli can
thus be given from this window to a subject over seventy feet away.

Several rooms are fitted up with special reference to the investigation
of the various forms of organic movement, animal behavior and
intelligence. As one result of several investigations in animal
psychology already pursued here, the laboratory has a considerable
number of devices for testing and making statistical studies of the
senses and intelligence, methods of learning and emotional reactions of
animals.

Adequate provision is made for the keeping of animals in a large,
well-lighted, and well-ventilated corner room. Instead of having
aquaria built into the room, an aquarium-table eighteen feet long has
been constructed to support moveable aquaria of various sizes. Whenever
it is desirable for the purposes of an investigation, any of these
aquaria may be moved to the research-room of the investigator or to
such quarters as the special conditions of the experiment demand.

The vivarium-room contains, in addition to provisions for
water-inhabiting animals, cages of a variety of forms and sizes. The
largest of these cages, six and a half feet high, six feet wide,
and four feet deep, may be used for birds, monkeys, or any of the
medium-sized mammals. Cages for rabbits, guinea-pigs, and other small
animals are arranged in frames which support four double compartments.
Similarly, small cages suitable for mice, rats, and other small rodents
are in supporting frames which carry four of the double cages, each of
which is removeable and may be carried to the experimenting-room at the
convenience of the experimenter.

In a large unheated room above the main laboratory are tanks for
amphibians and reptiles. These tanks, since they can be kept at a low
temperature during the winter, are very convenient and useful for
frogs, tortoises, and similar hibernating animals.

In view of the prime importance of electricity to a modern
psychological laboratory, a rather elaborate system of wiring has
been designed and built in. The unit of this system is a small
delivery-board six inches wide by eight inches high, which carries
the following five circuits: _a_, a time-circuit for running magnetic
signals; _bb_, two low-tension circuits for chronoscope, bells,
signals, etc.; _c_, a high-tension alternating current (110 v. and 60
phases) for alt. current motors, to be used where great constancy of
speed is desired; _d_, a high-tension direct current (110 v.) for dir.
current motors, where it is desired to vary the speed continuously (by
the introduction of resistance). Two such delivery-boards have been
set on opposite walls of all except the smallest rooms, which have
but one board. Circuits _a_ and _b_ are represented on the board by
binding-posts, while the high-tension currents, _c_ and _d_, appear as
flush, protected sockets that take a double-pole plug.

Circuit _a_ is a single circuit led from a time-pendulum permanently
set in the battery-room, and carried once around the laboratory. It is
connected with the _a_ binding-posts of the individual delivery-boards
in parallel. It follows that the time-circuit is alike for all the
rooms at any one time; but in different hours the pendulum can be
adjusted to give various impulse-rates. If an investigation requires
some special rate of impulse, the special time-apparatus is set up in
the investigator's room and current for it taken from one of the _b_
pairs of posts.

Each _b_ pair goes directly from the delivery-board to the battery-room
and ends at a double-pole (telephone type) socket on a large
switch-board. Thus every room has two or four direct and independent
connections with the battery-room.

The _c_ and _d_ circuits do not come from the battery-room, but from
their respective generators that are stationed outside of the building.
They are of course connected at the delivery-boards in parallel.

The large switch-board in the battery-room consists of an upper and a
lower part. The upper part bears the double-pole sockets from the _b_
posts in all the rooms; the lower part carries some fifty pairs of
single-pole sockets that are connected with the batteries stationed
near by. These pairs are labelled, and some give a current from cells
of the Leclanché type, others of a gravity type. The student has merely
to choose the kind and number of cells that he needs, from the lower
part, and connect them with one of the double-pole sockets of the
upper part which runs to a _b_ pair in his own room. By connecting
two double-pole sockets with each other, the student can establish a
circuit between any two rooms of the laboratory,--this for purposes of
telephonic or other communication. Since every room has two, and most
of the rooms have four of the _b_ circuits, the greatest variety and
elasticity of service is attained.

The large switch-board further carries a voltmetre and an ammetre,
both of the Weston make, which are reached (electrically) from
double-pole jacks (sockets) on the upper part of the board. Thus
before connecting the current with his room, the student can in a
moment measure its amount and intensity. These instruments are of the
flushface type, and dead-beat.

All of the rooms are lighted by electricity, and the lighting system
is independent of the delivery-boards. Nine of the rooms are provided
with soapstone sinks, and six (not including the lavatories and
service-room) with enamelled iron or porcelain sinks. All the sinks
have two taps and each of these ends with a screw-thread so as to take
a tip and rubber hose. The soapstone sinks were specially designed
with soapstone drip-boards. This is probably the best material for a
research-room, and the porcelain and enamel sinks were put only where a
neater appearance was desired, or where chemicals were to be frequently
used--as for instance in the battery and photographic rooms. Gas is not
used for illumination, but six rooms are provided with jets for the
smoking of drums, soldering, brazing, etc.

The instrument-room is equipped with large dust-proof cases for holding
the more delicate and valuable instruments. The larger unused pieces
are stored, out of sight but readily accessible, in an attic which has
a clear floor-space of something more than half the total area of the
laboratory. Dust-proof cases for demonstration and class-work material
are provided in the lecture- and class-rooms.

The shop contains a wood-working bench with two vices, tool-racks,
shelves, drawers, cupboards, and stock-racks, for the use of students;
and a 9-in. lathe, circular saw, grinding- and buffing-machine,
separate bench, vice, racks, and drawers for the use of the mechanic.
The machinery is run by a 5 h.p. electric motor suspended from one of
the outside brick walls, on brackets. One who selects the equipment of
such a shop has to weigh carefully the respective merits of circular
and band saws; the latter undoubtedly lends itself to a greater variety
of uses, but it is also a far more dangerous machine to have running
in a room to which students are to be given access. This latter
consideration determined in the present case the choice of a circular
saw. It is quite dangerous enough, and may be used only by, or under
the supervision of, the mechanic.

It has been stated on competent authority that a truly sound-proof
room cannot be built except under ground. This has not been
attempted, but the laboratory contains one room (no. 17) which is
virtually sound-proof. A double door separates it from the adjoining
experimenter's room, and double doors also separate this from the
main hall. The wall between these two rooms consists of two layers of
plaster with special deadening material inserted between. Two small
tubes, ordinarily stuffed with felt, connect these rooms. When the
acoustical stimulus is a tuning-fork, it is placed in a distant room,
connected with one of the _b_ circuits of the sound-proof room, and
then with a telephone receiver near the subject's ear.

The photographic-room contains the ordinary sink, red lights, shelves,
etc. The indirect entrance is light-tight when the door is not closed,
so that the experimenter may pass in and out even when developing is
going on. This room, like all the others which have no window (except
the sound-proof room), has forced ventilation.

The class-room is designed for the experimental training-courses. It
has eight of the regular delivery-boards, ten tables, instrument-case,
blackboard, and sink.

The lecture-room for specialized courses in comparative and
experimental psychology seats eighty students. It is provided with
two Bausch and Lomb electric projection-lanterns, horizontal and
vertical microscope attachments, and attachment for the projection of
opaque objects. On the lecturer's platform, besides the blackboard,
projection-screen, and chart-racks (capable of holding twenty charts),
is a large demonstration-table provided with a delivery-board, water,
gas, sixteen chart-drawers, two other drawers, and three cupboards.

As has been said before, the general psychology course of the
University is not given on the laboratory floor, but downstairs
in the large lecture-hall with about 400 seats. A number of large
demonstration instruments of the laboratory serve the special purpose
of this course; this hall too has its own stereopticons.

Our instrumentarium is, of course, in first line, the collection of
apparatus bought and constructed through the fourteen years of work.
Yet with the new expansion of the institute a considerable number of
psychological, physical, and physiological well-tested instruments has
been added. Especially in the departments of kymographic, chronoscopic,
and optical apparatus the equipment presents a satisfactory
completeness; its total value may be estimated to represent about
twelve thousand dollars. Yet the place of the laboratory which we
appreciate most highly is not the instrument-room but the workshop,
in which every new experimental idea can find at once its technical
shape and form. Whether those experimental ideas will be original and
productive, whether their elaboration will be helpful for the progress
of our young science, in short, whether the work in the new laboratory
will fulfil the hopes with which we entered it, may be better decided
as soon as a few further volumes of the Harvard Psychological Studies
shall have followed the present one, which is still from cover to cover
a product of Harvard's pre-Emerson-Hall period.




OPTICAL STUDIES




STEREOSCOPIC VISION AND THE DIFFERENCE OF RETINAL IMAGES

BY G. V. HAMILTON


The question which the Laboratory proposed to me for experimental
enquiry was one which demanded a definite reply of yes or no. The
positive answer seemed a necessary consequence of the traditional
psycho-physiological theories, while a certain practical consideration
seemed to suggest the negative solution. The question which seems
to have been overlooked so far was this: According to the theory of
stereoscopic vision two points of light which are seen by each of the
two eyes under the same angle appear to lie in the same plane; as soon
as the angle for the right eye is larger than that for the left, that
is, as soon as the two stimulated retinal points in the right eye are
more distant than the two retinal points stimulated in the left eye,
the right light-point seems to be farther away than the left one. If we
relate them to planes vertical on the ideal binocular fixation-line,
the right point lies in a more distant plane. This principle, which, of
course, controls all arrangements for stereoscopic effect, is deduced
from experiences in which the fixation-line is vertical to the line
that connects the nodal points of the two eyes; the plane in which
the equally distant points lie is then parallel to the forehead. If,
on the other hand, the eyes are turned to the side, that is, if the
ideal fixation-line forms an acute angle with the line connecting
the eyeballs, the two fixated light-points, which lie in a plane
perpendicular to the fixation-line, cannot be seen by the two eyes
under the same angle. Any object on my right side is somewhat nearer
to my right eye than to my left, and therefore must throw a larger
image on my right retina. The two light-points of a plane vertical
to the fixation-line give thus with the eyes turned to the right two
unequal pairs of retinal stimuli; and the difference of the retinal
stimulations is evidently just the same as if the eyes were looking
straight forward but the two lights were at different distances. If
difference of retinal images really produces the conscious experience
of seeing the lights in differently distant planes, vertical to the
fixation-line, it follows that with the eyes turned to the right,
lights which objectively lie in the same plane must appear subjectively
to lie in different distances. The question arises whether the facts
correspond to this conclusion. If we look with eyes turned sidewise
towards a plane vertical to the direction of seeing, do the points
of that plane remain in it for consciousness or do we see them in
different planes? We see that practical considerations suggest a "No"
to this question, because it would mean that everything which does
not lie exactly in front of us must change its plastic form, and
this the more strongly the more we see it on our right or our left,
and this of course again the more strongly the nearer it is to the
eyes, inasmuch as the relative difference of the retinal images must
increase with the nearness of the object. If a short-sighted person
fixates an object a few centimetres from the eyes strongly turned to
the side, the distances in the retinal image of the one eye may be
almost the double of those in the other. Under normal conditions the
differences would be smaller, but yet everything would be necessarily
distorted in its three-dimension shape as soon as it is seen in
indirect vision or with sidewise fixation. On the other hand, if the
objects keep their three-dimensional relations in spite of sidewise
movements, it is evident that the accepted psycho-physiological theory
of stereoscopic vision is incomplete and must be revised in a very
essential way. The experiment had to decide. Of course the question
might be approached experimentally in different ways. It would have
been possible, for instance, to study the stereoscopic synthesis of two
separate flat pictures seen with the eyeballs in different positions.
But we preferred the simplest possible way, seeking the threshold of
distance for two parallel vertical edges with eyes turned forward and
to the side. We chose edges instead of hanging threads for the purpose
of avoiding the possible influence of the apparent thickness of the
threads on the judgment of distance. Of course, distance is here never
distance from the one or the other eye, but from the centre of the
line which connects the two nodal points of the eyes; the two vertical
planes whose edges were to be compared stood always vertical on the
ideal line of fixation which starts from that central point between the
two eyeballs.

The apparatus used in these experiments consists of three parts, viz.:

(1) A plank 2.5 metres x 9.5 centimetres x 4 centimetres, set on edge
and screwed to a table at either end.

(2) A head-rest 45 centimetres high, 35 centimetres broad and 15
centimetres deep. Attached to the centre of the lower strip of the
frame is a concave trough for the chin. Another trough, shaped to
fit over the vertex and with a strip of wood fastened to the front
of it for the forehead, slides up or down within the frame. The
attachment for the forehead can be moved and fixed at various positions
antero-posteriorly. By means of these devices the head can be securely
fixed in any position desired without discomfort to the subject.

In order to have the eyes always in the same plane and at a known
distance from the apparatus at the other end of the plank, a hole was
made in either side of the frame with its centre at a level of the
eyes. Extending through the vertical diameter of each hole is a fine
wire. Fitted into the inner portion of each hole is a cardboard tube 10
centimetres long: the inner end of each tube contains a vertical wire
so arranged that the four wires all fall into a plane at right angles
to the long direction of the plank. A mirror at the outer exit of
either hole enables the experimenter to align the tips of the subject's
corneæ with the wires.

Two parallel strips of wood are so attached to the "head-rest" end of
the plank--one below and the other above it--that they can be rotated
laterally upon the plank, with the bolt which secures them to it for a
centre of rotation. Opposite this centre, and attached to the anterior
surface of the upper parallel strip is a wire needle 25 centimetres
long. By means of a quadricircular piece of cardboard attached to the
plank at the end of the needle, the extent of rotation to the right
or left can be read off in degrees. (The point midway between the two
corneal tips when they are aligned with the wires is in the same axis
of rotation as the head-rest.)

A vertical iron rod 50 centimetres long extends upwards from either end
of the parallel strips, and upon these rods the frame of the head-rest
can be moved up or down by means of thumb-screws upon which it rests.

(3) At the opposite end of the plank there is attached a flat board,
35 centimetres long and 30 centimetres wide. Attached to the edge of
the board which faces the head-rest is a piece of black cardboard
40 centimetres long by 35 centimetres broad. In the centre of the
cardboard is a rectangular aperture, 7 centimetres by 14 centimetres.
On the upper surface of the board are two slots, one at either side.
Sliding within each of these slots is a block of wood to which is
attached an upright sheet of black-painted tin, 15 centimetres wide and
20 centimetres high. The surfaces of these tins lie in planes parallel
to the plane of the four wires in the head-rest, when the latter is
at right angles to the plank. When their surfaces are equidistant
from the wires, the inner vertical edges of the tins are separated
from each other by 3 centimetres. The sides of the slots, in which
the blocks with their tins slide, are fitted with millimetre scales,
thus enabling the experimenter to determine the distance of the edges
from the corneæ. The point on the scale at which an edge was exactly 2
metres from the vertical plane of the wires was chosen as the "zero"
point, and if this distance was decreased by moving an edge forward,
the latter was said to stand at "minus" one, two, or more millimetres,
as the case might be. Likewise, an edge was said to stand at "plus"
the number of millimetres' distance beyond the zero point if it had
been moved at a greater distance than 2 metres from the wires. A piece
of ground glass attached to the distal end of apparatus enabled the
experimenter to secure a uniform illumination, the room being darkened
and the light coming from a 32-candle-power electric lamp set about a
metre and a half behind and on a slightly lower level than the glass.

It was found that by shading the lamp itself and admitting a dim
light to the room by means of drawing down only the ordinary thin
window-shades, the edges could be made to seem almost isolated in space
and to stand out in clear relief.

The subjects of the experiment were Messrs. Bell, Flexner, and Tait.
Each subject determined the equality-point and the threshold for the
normal primary position of the eyes, for the eyes in a lateral position
of 15° and in a lateral position of 30°, both to the left and to the
right.

Eyes at 0° means the following: that the most anterior part of the two
corneæ lies in a plane parallel to and two metres' distance from the
plane in which the two parallel edges lie at 0. Eyes at 30° to the left
means that a line drawn in front of the two corneæ intersects such a
line at an angle of 30°, the left eye being at the distal end of the
line. In calculating the visual angles 7.4 mm. are added in order to
compensate for the distance from the extreme anterior portion of the
cornea to the nodal point of the eye.

The results for Mr. Tait are as follows:

The position of eyes 0°. The right edge was moved, at first from an
evident + position to equality, then from equality to the - threshold,
then from an evident - position to equality, then from equality to
the + threshold. These four points were determined each fifteen times
and the average taken. Then exactly the same fifteen sets of four
determinations with the left edge moved. The averages of these 120
experiments are these: When the left edge is moved from + to =:-2.77,
from = to -:-6.97, from - to =:+0.77, from = to +5.93. When the right
edge is moved from + to =: +2.83, from = to -:-1.6, from - to =:+5.9,
from = to +:+10.53. The first equality-point appears thus when the left
edge is moved at -0.76, when the right edge is moved at +4.41, with a
threshold of about 5 in either case. With the normal eye-position the
edges must thus not be exactly in the same plane to appear equally
distant; at a distance of 2000 mm. the left must be about 2 mm. nearer
than the right to appear in the same plane, vertical to the line of
regard.

If the position of the eyes is 15° to the left, we have the following
results: When the left edge is moved from + to =:-4.17, from = to
-:-8.5, from - to =:-1.33, from = to +:+1; when the right edge is moved
from + to =:+4.17, from = to -:+1.17, from - to =:+4.5, from = to
+:+8.67.

If the position of the eyes is 30° to the left, we find when the left
edge is moved from + to =:-2.67, from = to -:-6.67, from - to =:+0.5,
from = to +:+3.33. When the right edge is moved from + to =:+2.33, from
= to -:-0.02, from - to =:+9., from = to +:+12.33.

If we take again the general averages, we have for the eye-position of
15° to the left an equality-point of -3.25 if the left edge is moved
and judged and +4.63 if the right edge is moved and judged. That is,
if the right edge stands at 2000 mm. the left edge must be moved to
1996.75, and if the left stands at 2000, the right must be moved to
2004.63. For the eye-position of 30° to the left, the equality-point
lies at -1.49 if the left edge is moved and judged, and at +5.91 if
the right edge is the variable. The threshold lies in all three cases,
for eyes at 0°, at 15°, and at 30°, at about ±5 mm.; the position of
the eyes has thus no influence on the threshold for the perception of
distance in the direction of regard.

But the point essential for our investigation is of course not
the threshold but the equality-point. To take the extremes of the
eye-positions 0° and 30° we find the equality when the left edge is
judged, at -0.76 for 0° and -1.49 for 30°, and when the right edge is
moved, at +4.41 at 0° and +5.91 at 30°; the middle is thus +1.82 for 0°
and +2.21 for 30°, that is a difference of less than 0.4 mm.

To understand this figure we must enter into the calculation of the
angles. We have an eye-distance of 60 mm., a distance of the edges from
the cornea 2000 mm., from the nodal points 2007.4 mm., the distance of
each edge from the median line 15 mm., the distance of the two edges
from each other thus 30 mm. as long as they are in the same plane. We
have to determine the angle under which each eye sees the distance of
the two edges. A simple trigonometric calculation gives the following
figures: If both eyes are in normal position, at 0°, and both edges
are in the same plane, 2000 mm. from the corneæ, the angle for each
eye is 51' 22". If the left edge is now moved to +5, the left eye sees
the distance of the edges at an angle of 51' 25", the right eye under
51' 10", the difference is thus 15"; if the left edge is at +10 mm.,
the left eye's angle is 51' 29", the right eye's angle 50' 59", the
difference 30". If the left edge is moved to -5 mm., the left eye's
angle is 51' 18", the right eye's angle 51' 33", the difference 15";
if the left edge is moved to -10 mm., the left eye's angle is 51' 14",
the right eye's angle 51' 45", the difference 31". Now we saw that with
normal eye-position when the left edge was moved the threshold was
+5.93 and -6.97; a difference of 15" to 20" between the visual angles
of the two eyes was thus amply sufficient to give a distinct experience
of different distance. When the left eye's angle was about 15" smaller
than the angle of the right eye, the difference of the retinal images
gave a sure impression of the greater nearness of the left edge.

If we now bring the eyes into the position of 30°, the angles are of
course different when both edges are in the same plane vertical to
the direction of regard. If the two edges are in the same plane, the
left eye's angle is 50' 59" and the right eye's angle 51' 45", the
difference thus 46". If we move the left edge to +5, the left angle
becomes 51' 1", the right angle 51' 34", the difference 33". If we move
the left to +10, the left angle becomes 51' 4", the right 51' 24"; the
difference is thus still 20", and we must move the left edge to +17 mm.
to get an equal angle for the left and the right eye. If we move the
left to -5, the difference becomes of course larger, the left eye sees
under 50' 56", the right eye 51' 55", the difference 59"; and at -10,
the left eye has the angle 50' 53", the right eye 52' 6", difference 1'
13". It is hardly necessary to state here the angles for the changes of
the right edge or for an eye-position of 15°, inasmuch as the maximum
differences bring out our case most clearly. With an eye-position of
15°, the edges at the same plane give angles of 51' 10" and 51'34",
that is, a difference of 24"; if the left edge is moved to -5 mm. the
difference becomes 38"; if it is moved to -10 mm. the difference is
54"; if the left edge is moved to +5 the difference decreases to 10"
and at +10 mm. to 6".

We have thus the following fundamental result: If the eyes are in
normal primary position, a movement of the left edge to ±6 mm. is
constantly apperceived at threshold of distance and this corresponds to
retinal images whose visual angles differ by about 17". A difference
of 17" in the visual angles of the two eyes produces thus under the
conditions of this experiment for this subject a strong stereoscopic
effect when the eyes are in primary position. If the eyes are in the
position of the head 30° to the left, the left eye thus much further
from the edges than the right eye, the visual angle of the left image
thus much smaller than that of the right image, we find the same
equality-point with the same threshold. We saw that in this position
the two visual angles would be equal if the left edge were moved to
+17 mm.; instead of at +17, the equality-point--when the left edge is
judged--lies at -1.49, that is, at a point at which the visual angle
of the left eye is more than 46" smaller than the angle of the right
eye. While in normal position a difference of the two retinal images of
17" constitutes a distinct threshold value; at a lateral position of
the eyes of 30° the great difference of 46" becomes necessary to give
the impression of equal plane, while a decrease of that difference to
30" gives a distinct feeling of greater distance. Equal retinal images
produce for the lateral eyes thus the same effect which for the normal
position very different images produce; and to get for the lateral eyes
the effect which equal images produce for the normal position, the
angles of the images must differ by 46".

The results for the second subject, Mr. Flexner, are practically the
same. With the position of the eyes at 0°, when the left edge is judged
and moved, we find the following averages: from + to =: +0.03, from =
to -:-3.8, from - to =:-0.7, from = to +:+3.93; when the right edge is
moved from + to =:-0.08, from = to -:-4.29, from - to =:+1.21, from =
to +:+4.08. It is evident that the difference between right and left
which existed for Mr. Tait does not enter into Mr. Flexner's results.
The equality-point as average of 120 experiments lies for normal
eye-position practically at zero, and the threshold is ±4 mm.; his
sensibility for differences of retinal images is thus still finer than
for Mr. Tait, as we saw that the threshold of ±4 mm. means a difference
of visual angles of less than 15". If Mr. Flexner's head is turned 15°
to the left, his left eye thus considerably farther away from the edges
than the right eye, the results are these: If the left edge is moved
and judged, we find from + to =:-0.02, from = to -:-3.17, from - to
=:0, from = to +:+4.67; if the right edge is moved from + to =:-0.01,
from = to -:-2.5, from - to =:-0.8, from = to +:+3.33. Experiments
with lateral movement of 30° were not carried through, as the subject,
accustomed to eye-glasses, became less accurate in the judgments; but
the experiments with the position of 0° and of 15° are unequivocal.
They show that the equality-point and the thresholds are exactly the
same for 15´ as for 0°. For the lateral position of 15° again the
average equality-point is exactly at 0° and the threshold at less than
±4 mm. We saw that for a lateral movement of 15° the difference of
the angles at the equality-point is 24". We find thus for Mr. Flexner
that with primary eye-position a difference of angles of less than 15"
gives a distinct stereoscopic effect, while with a lateral position of
the eyes a plane effect demands a difference of 24" for the two visual
angles.

Experiments with Dr. Bell finally showed a rather strong fluctuation
of judgments and the determination of the equality-point for normal
eye-position has not only too large a middle variation to be a reliable
basis, but is influenced by a constant tendency to underestimate the
distance of the edge moved. Yet the general result is the same as with
the other two subjects, that is, the equality-point is with him, too,
practically the same for the eyes in normal and in lateral position.

The general conclusion from the results of all three subjects is thus
evidently that the traditional physiological theory is untenable, the
stereoscopic effect cannot be simply a function of the difference
of the two retinal images. The same pair of unequal retinal images
which gives a most striking stereoscopic effect for eyes in primary
position, has no stereoscopic effect for eyes in lateral position and
_vice versa_. The stereoscopic interpretation is thus the function
of both the difference of the retinal images and the position of the
eyeballs. Of course the two retinal images are in any case never
felt as two pictures if they are not different enough to produce a
double image. With the primary position of the eyes as long as the
two different retinal views are sufficiently similar to allow a
synthesis in a three-dimensional impression of our object, we perceive
every point of the object not as double image but as one point of
a given distance. The distance feeling of the normal stereoscopic
vision demands thus itself more than the reference to the different
retinal images, and the only factor which can explain the phenomena
is the response of the eye-muscles which react on the double images
by increase or decrease of convergence. The distance of a point in a
stereoscopic image is determined by the impulse necessary for that
particular act of convergence of the eyeballs by which the two retinal
images on non-cor-responding points would be changed into images on
corresponding points. The different retinal images are thus ever for
the normal eye-position merely the stimuli for the production of that
process which really determines the experience of distance, that is,
the motor impulse to a change in convergence.

If thus the stereoscopic vision under normal conditions is ultimately
dependent upon the central motor impulses, it is not surprising that
a change in the psycho-physical conditions of movement produces a
change in the resulting impulses. Such a change in the conditions is
given indeed whenever the eyes are in a lateral position. Just as
the same stimulus produces a different response when the arm or leg
is in a flexed or an extended position, so the retinal double images
stimulate different responses according to the particular position
of the eyeballs. That pair of unequal retinal images that in primary
eye-position produce in going from one end of the object to the other a
strong increase of convergence and thus a feeling of greater nearness,
may produce with the lateral eye-position no increase of convergence
and thus a feeling of equal distance or even a decrease of convergence
and thus a feeling of removal. The psycho-physical system upon which
our three-dimensional visual perception depends is then much more
complex than the usual theory teaches; it is not the retinal image of
the double eye, but this image together with the whole distribution
of contractions in the eye-muscles, which determines the stereoscopic
vision: the same retinal images may give very different plastic
perceptions for different positions of the eyeballs.

The experiments point thus to the same complex connection which
Professor Münsterberg emphasized in his studies of the "Perception of
distance."[1] I may quote the closing part of his article to bring
out the intimate connection of the two problems. He reports his
observations on the so-called verant and insists that the monocular
verant almost as little as the ordinary binocular stereoscope can give
the impression of normal distance of nature. Professor Münsterberg
writes: "Whoever is able to separate seeing in three dimensions from
seeing in natural distance cannot doubt that in both cases alike we
reach the first end, the plastic interpretation, but are just as far
removed from the other, the feeling of natural distance, as in the
ordinary vision of pictures. The new instrument is thus in no way a
real 'verant.'

"The question arises, Why is that so? If I bring my landscape picture
on a transparent glass plate into such a distance from my one eye
that every point of this transparent photograph covers for my resting
eye exactly the corresponding point of the real landscape and yet
accommodation is excluded, as, for instance, in the case of the
short-sighted eye, or in the case of the normal eye with the verant
lenses, then we have exactly the retinal images of the real view of
nature and the same repose of the lens. Why are we, nevertheless,
absolutely unable to substitute the near object for the far one? This
problem exists in spite of all the theoretical assurances that the one
ought to appear exactly like the other, and I think that it is not
impossible to furnish an answer to it.

"If I am not mistaken, there is one point of difference between seeing
the mere picture and seeing the far landscape, which has been neglected
in the usual discussions. Every one knows, of course, that we see the
picture and the landscape normally with the help of eye-movements. The
eye moves from point to point; but psychologists have neglected the
consideration that the relation between eye-movement and retinal image
must be quite a different one for the landscape and for its photograph.
Let us consider the simplest possible case, the case of the myopic eye
without any lenses whatever, and without any need of accommodation for
a picture as near to the eye as 10 cm. If I take a small landscape
picture made with a camera whose distance from lens to plate is 10
cm., I have a splendid plastic view if I see it at a distance of about
10 cm. from my eye. I have before me just such a picture in which two
mountain peaks are, in the photograph, 1 cm. distant from each other.
If I now have my little picture at the distance of 10 cm. from the
eye, these two mountain tops correspond in their distance of 1 cm.
exactly to the retinal image which the two real mountains, which are
ten miles away and one mile distant from each other, produce in my
retina. The retinal image of the two mountain peaks in the photograph
is thus for my resting eye indeed identical with that of real nature.
Does that mean that I have to make the same eye-movement to go from
the left to the right mountain in the landscape as in the picture? Of
course, that would be so, the movement would be just as identical as
the retinal images if the nodal point of the light-rays were identical
with the rotation-point of the eyeball. But everybody knows that this
is not at all the case. The light-rays cross in the lens. The angle
of vision, and thus the size of the retinal image, are thus dependent
upon the distance of the lens from the retina. But the movement of the
eye is related to a rotation-point which lies about 13 mm. behind the
cornea, roughly speaking 1 cm. behind the nodal point of the rays. This
additional centimetre plays, of course, no rôle whatever, if I look at
my mountains in the real landscape; following with my eyeball from the
fixation-point of the left mountain to the fixation-point of the right
mountain, I make a movement whose angle can be declared identical with
the angle under which I saw the two mountains with the resting eye in
the first position. This angle of vision was determined by the distance
of the nodal point, which was in our case ten miles, while the angle
of eye-movement was determined by the distance of the rotation-point,
which would be ten miles plus one centimetre, and there is of course
no possible difference for practical discrimination between these two
distances.

"But the situation is completely changed if I turn to my little
picture 10 cm. distant from my eye. The angle under which I see my
two peaks is, of course, again the same under which I saw them in the
real landscape. It is determined by the distance of the picture from
the nodal point, which is in this case 10 cm. But the angle of the
eye-movement necessary to fixate first the left and then the right
peak is now a much smaller one because it is again determined by the
distance from the rotation-point, and that is in this case 10 cm.
plus 1 cm. With this short distance of the picture from the eye this
one additional centimetre is not at all the negligible quantity which
it was in addition to ten miles in the landscape. For the two real
mountains the angle of the eye-movement had a tangent of one tenth;
for the photograph mountains, in spite of their equal size of retinal
image, the angle of necessary movement would of course have a tangent
of one eleventh. Roughly speaking, we could say that the photograph,
in order to produce the same eye-movement which the mountains in the
landscape excited, would need a pictorial distance between the two
photograph mountains of 11 mm. instead of 10 mm. Of course if the
distance in the picture were made 11 mm. instead of 10, it would not
cover any more the mountains of the landscape. The retinal image would
thus be relatively too large and would not give us any longer the true
landscape. On the other hand, if we tried to correct it by bringing the
picture one centimetre nearer to the eye, then of course every retinal
image would be enlarged by that necessary tenth, and yet there would be
no help for the situation, as now again the eye-movement demanded by
the retinal image would be relatively increased too.

"We can put it in this way: _my real landscape demands a relation
between retinal image and movement which my picture cannot produce
under any circumstances whatever_. That which would be needed to
imitate the relations would be realized only if I had my retinal
images from the picture at a distance of 10 cm., and at the same time
the movements belonging to the same picture seen at a distance of 9
cm. That is of course unrealizable. We cannot see a picture without
having our movements constantly controlled by the size of the real
retinal images, as it is necessary that the distance seen in indirect
vision is the distance covered by the fixation-point during the
eye-movement. That demands, as we have seen, a different relation
between retinal image and eye-movement for near and far, and no verant
and no stereoscope can eliminate this factor. If a 10-mm. object in the
photograph demands an 11-mm. movement to give the impression of real
natural distance, then we have a condition which cannot be fulfilled.

"If we remember how extremely delicate is our normal sensitiveness for
retinal distances and how the newer studies in stereoscopic vision have
demonstrated an unsuspected delicacy of adjustment between retinal
images and motor responses, it is evident that this so far always
neglected relation must be an extremely important one. If we have
one adjustment of central reaction in which a certain eye-movement
corresponds to retinal images of one size, and another adjustment in
which the same movements correspond to retinal images which are ten
per cent larger, we can really not expect our judgment of distance
to neglect the difference between these two systems of relations. Of
course they represent two extreme cases. Every distance beyond 10 cm.
demands its special adjustment up to the point where the distance
becomes too large to be influenced by the distance from the nodal point
to the rotation-point. We must thus presuppose a sliding scale of ever
new adjustments for the different distances at which we see any object,
and we have, in this relation, probably not the least important factor
in the judgment of the third dimension for relatively near objects, and
probably even more important than the irradiation circles which control
the accommodation, as these circles must be the same for objects which
lie before and behind the fixation-point. Of course the whole system
of our localizing reactions becomes through these considerations more
complex by far than the schematizations of the text-books propose. But
physiological optics has shown at every point in its development that
mere simplification has not always meant a deeper insight into the real
relations."

It is evident that our studies in stereoscopic vision with lateral
eye-position involve exactly the same principle and reaffirm completely
Professor Münsterberg's theoretical views. In both cases, in the
monocular of the verant as in the binocular of our experiments, the
same retinal image has different psycho-physiological space-value on
account of the different motor situation.

FOOTNOTE:

[Footnote 1: Münsterberg: Perception of Distance, The Journal of
Philosophy, Psychology and Scientific Methods, vol. 1, p. 617, 1904.]




EYE-MOVEMENTS DURING DIZZINESS

BY E. B. HOLT


It is a familiar fact that when the head is passively turned about
its vertical axis, the eyes do not move with the head but lag behind,
keeping their fixation on that object toward which they were directed
before the head moved. The eyes move in their sockets in a direction
opposite to that in which the head has moved. Now it has been proved
beyond a doubt by the experiments of Mach,[2] Crum Brown,[3] and
Breuer,[4] that these lagging movements of the eyes are reflex and are
governed by the semi-circular canals, which are stimulated directly by
the motion of the head. Similar reflex eye-movements are found when the
head is turned about some other than its vertical axis, the direction
of such movements being always in confirmation of the theory. All
these movements, together with the theory, are well described in the
summaries of Peters[5] and Nagel.[6] The present paper deals solely
with the eye-movements that occur after rotation of the head about its
vertical axis.

The mechanism of these lagging, reflex movements is not, then,
identical with that which enables us, when the head is at rest, to
fix on and follow a luminous moving object,--the "pursuit movements"
of Dodge.[7] It is, however, identical with that of Dodge's "fourth
type"[8] and that of the compensatory eye-movements described
by Brown,[9] Nagel,[10] and Delage,[11] and recently studied by
Angier.[12] This function of the semi-circular canals was first
suggested by Goltz in 1870. Now if the rotary movement of the head is
prolonged, the eyes lag for a while on their first fixation-point, and
then dart suddenly forward to a new fixation-point on which they rest
for a while as before, until they dart forward again. Therefore if the
head continues to rotate, the eyes fall into a regular and well-marked
nystagmus. In this the lagging movements, or those opposite to the
direction of the head, are called "compensatory," and are relatively
slow and long. Their rate coincides closely if not exactly with that
of the head-movement. But the movements forward, in the direction of
the head-movement, are short and swift. Such are the facts during the
rotation of the head.

But if this rotation has been somewhat prolonged, the ocular nystagmus
continues after the head and body are brought to rest. But now its
phases are reversed, and the slower eye-movements are in that direction
in which the head has moved; while the swifter are in what before
was the lagging direction. These observations are in accord with
the semicircular canal theory, and are well established by various
investigators.[13]

This paper presents the results of a photographic study of the reflex
eye-movements following after rotation of the head (and body) about the
vertical axis.

The subject whose eyes were to be photographed sat in a chair placed
on a rotating platform, in such a position that the vertical axis of
rotation passed through, or just posterior to the nose. Rays from an
arc-lamp of 6 amperes, placed about 60 cm. from the subject's face,
were so converged by a lens that when the subject came to rest, after
the rotation, his two eyes were brightly illuminated. An adiathermal
screen consisting of a dilute solution of copper ammonium sulphate
kept the heat from being painfully intense on the eyes. The light fell
slightly from one side on the subject's face, when he was brought to
rest; and directly in front of him, at a distance of about 40 cm., was
a camera of which the lens was on a level with his eyes. The ordinary
ground-glass screen at the back of this camera was replaced by a
light-proof box, in the front of which, and in the plane which should
have been that of the ground glass, was a slit 55 mm. broad and 5 mm.
high. Inside the box was a Ludwig kymograph of which the drum rotated
on a horizontal axis: the circumference of the drum lay tangentially
to the front of the box, and the line of tangency passed horizontally
through the long axis of the slit. For each photograph a photographic
film of sensitometer 40 was fixed to the drum, as paper is ordinarily
fastened, and in moving, the drum carried this film upwards past the
slit. It follows from this arrangement that 5 mm. along the length of
this film were always exposed at once. The camera was so focused that
the images of both eyes were sent through the slit, and fell on the
film.

[Illustration: Figs. 1 and 2]

The subject's head was rigidly held by a rest: this rest was adjusted,
and the camera focused, before the rotation. The adjustment of the
head was greatly facilitated by fastening a fine black thread to pegs
that projected forward from the head-rest, on either side; the thread
was stretched horizontally, and at such a height that its image in the
camera coincided with the middle of the long (horizontal) axis of the
open slit. If then the subject, on seating himself in the chair, had
his head so adjusted that each eye was directly behind the thread,
each eye would certainly be imaged on the sensitive film. Neither the
shadow of this thread on the subject's face, nor its image on the film,
interfered in the least with the exposure that was made after rotation.
This thread was further found very useful by the subject himself,
who, after the rotation and just before the exposure was made, could
make sure by sighting on the thread that his eyes had not slightly
changed position during the rather protracted rotation. The subject was
ordinarily turned twenty-five times at about the rate of one turn in
two seconds. The kymograph was set in motion and the exposure commenced
as soon as the whirling chair was brought to a dead stop. This stopping
always took two or three seconds, at the very time when the nystagmus
was most pronounced, so that the photographs do not show the maximum
eye-movements. The exposure lasted through one rotation of the drum,
nine seconds.

In the strongest negatives the movements of the eyes can be fairly
well made out from the undulatory curve generated on the film by the
dark image of the iris as it oscillated from side to side. But this is
true only of the best negatives, and almost never of these if the eyes
photographed had the iris blue. In order to obtain better definition
in the photographs of the eye-movements, small flecks of Chinese white
were tried, as invented and described by Judd.[14] A small square of
white was laid with a brush on each cornea, on the side toward the
lamp, so that its image on the film should be as bright as possible.
The flecks were found to adhere to the eyeball even more perfectly than
Judd himself has claimed; and they produced so little discomfort that
the subject ordinarily forgot their presence on the eyes. Nevertheless
their image as produced on the negatives, although much better than
that of the iris, was generally not clearly readable, owing to the
brief exposure and the illumination by electric light. This light seems
not to be well reflected by the Chinese white: but in all cases where
daylight can be employed the use of these flecks must be eminently
satisfactory.

Thus it was found necessary to fall back on the image of the arc as
reflected from the cornea. This corneal image invariably traced a
clear, strong curve on the negative, and would have been appropriated
at the outset, were it not that its movements are not, as is well
known, a true register of the _amplitude_ of the corresponding
eye-movements; a fact that was shown clearly from a comparison in
these negatives of the curves produced respectively by the flecks
of Chinese white and by the corneal image. The former showed a much
greater amplitude of movement. But the corneal reflection is a perfect
register of the _time_ and _direction_ of the eye-movements; and in
the following tables these features alone are studied. This reflection
traced on the film a perfectly readable curve, although in some of the
films, owing to a shifting of the carbons in the lamp taking place
during the rotation, one of the eyes would be badly illuminated and a
good record would be obtained from the other eye alone.

The arc ran on an alternating circuit of 60 phases per second, and
owing to these interruptions of the illumination the curve of the
corneal image showed on the negative as a dotted line in which the
distance between any two dots represented one sixtieth of a second.
Since the constancy of this alternation in the current has been
measured in the Jefferson Physical Laboratory (of Harvard), and found
to vary within a few tenths of one per cent only, the spacing of the
dots on the negatives formed the most convenient possible means for
determining the durations of the nystagmiform movements. These dots are
shown in Figs. 3, 4, and 5 (Plates I and II).

[Illustration: PLATE I.

Fig. 3

(By an error Fig. 4 is shown reversed; the lettering is correct.)

Fig. 4]

Fig. 3 shows a portion of one of the films. The two curves are to be
read from below upwards; but at the bottom is a photograph of the slit
(showing a part of the subject's face) taken when the drum had made
a little over one revolution and had come back to rest. Hence below
the image of the slit, the curve of corneal reflection is doubled.
"Right" and "Left" refer to the subject's right and left sides, so
that the reader looks into the subject's face from in front. In the
picture of the slit, the place on the cornea of the corneal reflection
is shown; and also a minor reflection, which as may be seen traced no
curve, from some other source of light. The fine line that crosses
the slit horizontally is the image of the thread, above mentioned,
which was used in adjusting the head. The time-dots are seen to be
perfectly distinct, so that they could be accurately read with the
help of a jeweller's eyeglass. Fig. 4 shows another part of the same
negative, a portion subsequent to the single eye-curves of Fig. 3,
that is, a continuation vertically upwards of Fig. 3. The rotation
had been from the subject's left to his right, a direction that will
be termed "clockwise" throughout this paper, and it can be seen that
the quick eye-movements are toward the subject's _left_, while the
slow are towards his right: had the photograph been taken _during_ the
rotation, the directions of the quick and slow movements would have
been reversed. Two points may be observed in this figure which the
tables will also bring out,--that the two eyes move together, and that
as the nystagmus subsides the quick eye-movements become less frequent
but endure no longer, or in other words, the slow movements alone
increase in duration. The corneal reflection does not accurately show
the amplitude of the movements; but direct inspection of a subject's
eyes, as the nystagmus dies away, shows that generally (but perhaps
not always) the amplitudes of both quick and slow movements decrease
together. When this is the case, it follows that at the end of the
nystagmus the _rate_ of the slow movements decreases very much faster
than that of the rapid movements.

Readable negatives were obtained from four, out of six subjects. Of
such negatives there are fourteen, ten of which are of eye-movements
after rotation clockwise, and four after rotation anti-clockwise.
This distribution is accidental, for the rotations in each direction
were about equal in number. With the exceptions to be noted later all
the negatives exhibit the same features, so that of the fourteen only
four examples are given in full in the tables; while for the others
merely the averages of the duration of quick and slow eye-movements
respectively are given.


TABLE I

    KEY:
    1: Subject
    2: Film
    3: Eye
    4: Direction of the rotation.
    5: Slow movements toward Subject's
    6: Rapid movements toward Subject's
    7: Average duration in seconds of slow movements
    8: Average duration in seconds of rapid movements

     1   2    3        4          5        6       7       8

     C   1  left   clockwise    right    left     .32     .05
     "   2   "         "          "        "      .36     .06
     "   3  right  anti-clock   left     right    .26     .08
     H   1   "     clockwise    right    left     .54     .07
     "   2   "         "          "        "      .45     .07 }
     "   "  left       "          "        "      .45     .07 }
     "   3   "         "          "        "      .50     .08}
     "   "  right      "          "        "      .49     .08}
     "   4   "     anti-clock   left     right    .49     .07
     "   5  left   clockwise    right    left     .53     .06
    Ta   1  right      "          "        "      .73     .07
     "   2   "     anti-clock   left     right    .48     .10
    Tu   1   "     clockwise    right    left     .50     .06 }
     "   "  left       "          "        "      .49     .07 }
     "   2   "         "          "        "      .49     .12}
     "   "  right      "          "        "      .49     .12}
     "   3  left       "          "        "      .40     .07
     "   4  right  anti-clock   left     right    .58     .08

                                             Av.  .48     .08

Table I gives these averages for all the fourteen negatives. In four
of these (H 2, H 3, Tu 1, Tu 2) simultaneous curves for both eyes
were obtained. In every curve the slow eye-movements were in the
same direction as the previous rotation; the rapid in the opposite
direction. The very few single movements that are exceptions to this
are noted under Table II. Had the photographs been taken during
(instead of after) the rotation, the directions of rapid and slow
movements would undoubtedly have been reversed. It is to be noted
that when both eyes were recorded, their movements were generally
identical, within the accuracy of measurement (one sixtieth of a
second). There are a few exceptions to this. The averages of all slow
and all rapid movements merely show that in general, and for that part
of the nystagmus that was photographed, the slow eye-movements lasted
six times as long as the rapid ones. This ratio varies considerably
from one case to another, and at best throws little light on the whole
nystagmiform series, since during the very first instants after the
rotation the ratio of quick to slow movements would be less than one
sixth, and at the very end of the series would be considerably more;
this because toward the end the slow movements become much slower,
while the rapid seem to change very little. The variations from case
to case arise, at least partly, because in some cases the picture was
taken more promptly, after the rotation stopped, than in others.


TABLE II

                           All records in seconds.
    Subject C.       Subject H.              Subject H.         Subject Tu.
     Film 3.           Film 2.                Film 3.             Film 4.
  anti-clockwise.     clockwise.            clockwise.      anti-clockwise.
    right eye.  left eye.   right eye.   left eye.   right eye.  right eye.

   slow  fast  slow  fast   slow fast  slow   fast slow   fast   slow  fast
    m.    m.     m.    m.    m.     m.    m.    m.   m.    m.     m.    m.
    to    to    to    to     to    to     to   to   to     to     to    to
   lft.   rt.   lft.  rt.   lft.   rt.  lft.  rt.  lft.    rt.   lft.   rt.

   .26
         .08   .03          .05        .06         .08                 .06
   .2                .51         .58          .1          .1     .45
         .06   .06          .06        .06         .05                 .08
   .05               .16         .16          .13         .13    .75
         .08   .05          .06        .1          .13                 .06
   .19              1.01        1.05          .36         .33    .36
         .05   .06          .06        .1          .06                 .13
   .02               .26         .26          .28         .3     .48
         .05   .05          .06        .13         .13                 .05
   .21               .26         .26          .28         .23    .61
  (.19)        .06          .06        .11         .13                 .1
   .16               .55         .6           .35         .35    .41
       .05           .1          .1    .08         .1                  .05
   .03               .25         .25          .33         .3     .65
       .05     .06          .05        .06         .11                 .06
   .18               .33         .33          .25         .2     .51
       .06     .1           .1         .1          .1                  .05
   .03              1.65        1.65          .83         .83    .66
       .05     .06          .06        .1          .13                 .16
   .29               .26         .26          .63         .61    .66
       .11     .06          .08        .06         .06                 .08


TABLE II, _continued_.

                             All records in seconds.
      Subject C.       Subject H.            Subject H.         Subject Tu.
        Film 3.          Film 2.               Film 3.           Film 4.
    anti-clockwise.     clockwise.           clockwise.     anti-clockwise.
      right eye.   left eye. right eye.   left eye. right eye.   right eye.

  slow  fast   fast   slow  fast   slow  fast  slow  fast  slow  slow  fast
  m. to m. to m. to m. to  m. to m. to m. to m. to m. to m. to  m. to m. to
  lft.   rt.   lft.   rt.   lft.   rt.   lft.   rt.  lft.   rt.   lft.  rt.
  .29                 .78          .76          .45         .43    .68
       .03           .06          .06   .05          .1                .15
  .04                 .16          .2           .45         .43    .23
       .05    .1           .1           .06          .1                .11
  .25                 .38          .33          .4          .38    .36
       .15    .08          .11          .08          .06               .08
  .3                  .58          .56          .58         .56    .35
       .05            .06          .06   .06          .06               .06
  .33                 .78          .78          .58         .56    .38
       .05    .08          .1           .08          .08               .05
  .28                 .71          .71          .35         .35   1.78
       .11    .08          .06          .05          .06               .06
  .41                 .46          .45          .51         .5
        .06           .05          .06   .06          .05
  .43                 .56          .58          .6          .61
        .11    .06          .06          .08          .1
  .35                 .33          .31          .73         .68
        .05    .1           .1           .06          .06
  .23                                           .86         .86
        .15                              .08          .08
  .38                                           .21         .21
        .1                               .06          .05
  .43                                           .8          .81
        .11                              .08          .08
  .38                                          1.53        1.58
        .03                              .06          .05
  .53
        .2
  .23
  (.18)
  .36
        .03
  .23
        .06
  .45
        .11
  .25

  Averages
  .26   .08    .07    .45   .07    .45   .08    .50   .08   .49    .58  .08

Parentheses indicate time during which the eye did not move at all.

[Illustration: PLATE II.

Fig. 5]

Table II gives in detail the data yielded by four of the most
instructive films. C 3 is the longest record that was obtained; Tu 4
is among the shortest, though it is not the very shortest. H 2 and
H 3 show how nearly alike are the simultaneous movements of the two
eyes: .07 sec. is the greatest difference recorded on any film between
simultaneous movements. All four records show how much less the
duration of the slow movements is at the beginning of the record than
at the end, and how little the fast movements vary in this respect.

H 2 is given because it is not typical; and about one half of the film
itself is reproduced in Fig. 5 (Plate II). It will be seen that at four
points there intervened between slow movements (toward the right) a
rapid one that was also toward the right. This is the only record in
which such a thing happened: and its explanation is problematical. With
the subjects C and H, and only very rarely with these, a rapid movement
sometimes took the place of a slow one, that is, occurred in the same
direction as the slow movements (_e. g._, Table II, C 3). And a trifle
more often, yet very seldom, a rapid movement was relatively slow (_e.
g._, _ibid._). With every subject there are a few cases in which the
eyes stood still for a small part of a second (_e. g._, _ibid._), and
these moments of rest seem to come after a rapid or a slow movement
indifferently.

McAllister[15] and others have shown that the eyes are seldom at rest
even when voluntary fixation is attempted, and these anomalies in the
nystagmiform series may well be the result of such random factors,
which instead of being always inhibited by the afferent impulses from
the semicircular canals, which govern the nystagmus, operate along with
these latter, and sometimes even inhibit them. With the exception of
these anomalies, the movements recorded in the photographs confirm the
observations of Purkinje, Mach, Breuer, Delage, and other investigators.

In conclusion, the sensations of vertigo and of nausea seem not to be
essentially connected with the nystagmus. Several subjects were so
disagreeably affected by a preliminary rotation that it seemed best not
to continue the experiment with them. With those, however, whose eyes
were photographed, while they experienced a mild degree of vertigo and
nausea during and after the first few rotations, these sensations soon
wore off with further practice, while so far as could be observed their
eye-movements were as ample and rapid as at first. The introspection
of these subjects was that after the rotation the body seemed at rest
and the stomach quite settled, while the visual field alone whirled
rapidly in the direction opposite to that of the previous rotation.

FOOTNOTES:

[Footnote 2: E. Mach: Sitzungsb. d. k. Akad. d. Wissensch., Wien, 1874.]

[Footnote 3: A. Crum Brown: Proceedings of the Royal Soc., Edinburgh,
1874.]

[Footnote 4: J. Breuer: Med. Jahrb., Wien, 1874-75.]

[Footnote 5: W. Peters: Arch. f. d. ges. Psych., vol. 5, p. 42, 1905.]

[Footnote 6: W. Nagel: Handbuch d. Physiol. des Menschen, vol. 3, p.
762, 1905.]

[Footnote 7: R. Dodge: Amer. Jour. of Physiol., vol. 8, p. 317, 1903.]

[Footnote 8: _Ibid._ p. 327.]

[Footnote 9: A. Crum Brown: Proceedings of the Royal Soc., Edinburgh,
1895.]

[Footnote 10: W. Nagel: Zeitsch. f. Psych. u. Physiol., vol. 12, p.
331, 1896.]

[Footnote 11: Yves Delage: Arch. de Zoöl. expér. et générale, vol. 1,
1903.]

[Footnote 12: R. P. Angier: Zeitsch. f. Psych. u. Physiol., vol. 37, p.
235, 1905.]

[Footnote 13: See the summaries of Nagel and Peters, above referred to.]

[Footnote 14: C. H. Judd: Yale Psych. Studies, Psych. Rev., Mon.
Supplements, vol. 7, no. 1, p. 7, 1905.]

[Footnote 15: C. N. McAllister: Yale Psych. Studies, Psych. Rev. Mon.
Supplements, vol. 7, no. 1, p. 17, 1905.]




VISION DURING DIZZINESS

BY E. B. HOLT


During and after a prolonged rotation of the head, the visual field
seems to spin around before one's eyes,--a phenomenon that is
ordinarily called the "dizziness of Purkinje." Delage describes it as
follows:[16] "In the experiment of Purkinje, while we are rotating in
a positive sense, space seems possessed of a motion in the opposite
direction.... This phenomenon is explained _by the direction of the
nystagmus_."

"In the nystagmus," he continues, "the eyeballs execute two
well-differentiated motions: one, a compensatory, _relatively_ slow
motion, during which images pass across the retina so as to give the
appearance of a movement of space in the opposite direction; two, a
swift motion opposite to the slow one, and so rapid that the images
passing across the retina leave no sensation of their movement."

Now, in a previous paper[17] I have shown that there is a central
anæsthesia, or central inhibition of visual sensations, during about
the latter two thirds of the time occupied by every voluntary eye-jump;
and in view of this I was led to enquire whether in fact, as Delage so
confidently asserts, it is the _speed_ of these more rapid movements,
or some other factor, that causes them to leave no visual sensations.
There can be no doubt that they do leave none, since, aside from the
statement of Delage, in dizziness the visual field whirls always in
only one direction; whereas it should otherwise appear to swing now
to one side, now to the other, as the eyes move back and forth across
the objects. I have found but one other mention of this point in the
literature. In his Analyse,[18] Mach says, parenthetically, "(the
jerky eye-movement leaves no optical impression)"; but he does not
suggest that this is because of its greater speed.

In order to test this point, a 2 c. p. incandescent lamp was so
arranged that it could be moved vertically in front of, and about four
metres distant from, a rotating chair. Since after a rotation the eyes
are oscillating from side to side, if the lamp is moved up and down an
obliquely inclined after-image streak must be generated on the retina;
and clearly there are four possible positions in which this may lie, as
shown in Fig. 1.

The results were absolutely uniform (the author alone as subject);
the after-image streak always lay on that side of the moving light
_toward_ which the _slow_ eye-movements were directed, that is, the
lamp appeared to drift obliquely up or down and in a lateral direction
_opposite_ to that of the slow eye-movements. Apart from its vertical
displacement, then, the lamp behaved like the less intensely illumined
parts of the visual field, seeming to be totally invisible during the
swifter eye-movements. Now since the experiment was done in a partially
darkened room and the eyes were partly adapted to darkness, the lamp
should have been intense enough adequately to stimulate the retina
even during the more rapid movements, and might be expected to leave
an after-image streak on that side toward which these rapid movements
were directed, and differing only from the streaks seen during the slow
eye-movements in being inclined at a less angle from the horizontal.
Yet no such streaks were visible.

These observations were made at about the same number of seconds after
the rotation stopped, as the photographs were taken that are recorded
in the preceding paper of this volume. The rapid movements were
therefore about one sixth as long in duration as the slower ones. Since
the respective amplitudes of rapid and slow must average very nearly
the same, the rapid movements must have been about six times as swift
as the slow movements. It needs therefore to be shown beyond a doubt
that the 2 c. p. lamp _was_ bright enough, in view of the briefness of
stimulation of any one retinal element during the rapid eye-movement,
to be above the threshold of perception. For this reason the experiment
was not continued with other subjects.

The certainly adequate degree of illumination was realized during the
photography of the eyes described in the preceding paper. Here during
the post-rotary dizziness an arc lamp (of 6 amp.) was in front of
the face and but a little to one side of the primary line of regard;
it was 60 cm. distant from the eyes and on a level with them; a lens
condensed the rays on the two eyes, and the light was diminished
only just enough as not to be painful, by a dilute screen of copper
ammonium sulphate about 3 cm. thick. Of course such an illumination
must adequately stimulate each retinal element even during the most
rapid eye-movements. Nevertheless with the four subjects that were
photographed the arc lamp, like the rest of the visual field, seemed
always to swim in one direction, and that opposite to the slower
eye-movements. In one case where the eyes were photographed without the
adiathermal screen, and the light was rather painfully intense, the
lamp was still seen to drift in one and the same direction. There was
never any trace of its moving to and fro, as there should have been had
it been visible during both phases of the nystagmiform movements.

[Illustration: Fig. 1]

This absence of visual sensation during the more rapid eye-movements
might conceivably depend on either peripheral or central inhibitory
factors. But the anatomy and physiology of the eye offer no point of
support for the supposition that during such movements the irritability
of the rods and cones is momentarily reduced, or that the retinal
layers posterior to the rods and cones suffer an interruption of
function during a movement of the eyeball in its socket. Indeed, during
some such movements, the "pursuit" movements (Dodge's second type),
vision is unimpaired.[19] In view of these facts, and of the many known
cases of the mutual inhibition of sensations where undoubtedly the
process is a central one, it is by far most probable that this visual
inhibition is also a central process; as was certainly the visual
inhibition during voluntary eye-jumps, previously reported by me.[20]

The conclusion above reported that the visual inhibition during the
more rapid phase of the nystagmus in no wise depends on an inadequate
stimulation of the retina, due to the greater speed of the rapid
movements, and that the inhibitory process is purely central, is
further supported by the following phenomenon. If before the rotation
has commenced, the eyes are so strongly stimulated that a lasting
after-image is obtained, this after-image will, during the rotation,
always be seen to swim in the direction opposite to the rotation, that
is, _with_ the _slow_ eye-movements; but when the rate of rotation
begins to decrease, and as Mach, Breuer, and Delage have shown, the
slow eye-movements reverse their direction, the after-image also
reverses its direction, and now swims in the direction of rotation,
that is, _still with_ the _slow_ eye-movements. If the after-image
persists long enough, it may still be observed, after the rotation
has ceased, swimming in the same direction as the surviving slow
eye-movements. If, for instance, the slow movements are from left to
right, the after-image (best seen with the eyes closed) swims from the
left to the right hand side of the field and disappears, reappears at
the left and swims again toward the right, and continues to do this
until the nystagmus entirely ceases.

This experiment was repeated several times, with four subjects, and
with both clockwise and anti-clockwise rotations, and the results were
uniformly as described above. In order to see whether this motion of
the after-image really depended on the slower nystagmiform movements,
the following variation was tried. It will be recalled that if the
head is rotated not about a vertical (longitudinal) axis, but about a
transverse axis, as, say, one passing through the ears, a nystagmus
is produced in which during the rotation the slower eye-movements are
opposite to the direction of rotation, while when the rotation is
checked or stopped, the nystagmus, as before, reverses. The same is
true if the rotation is about a sagittal axis. These conditions were
approximately realized by having the subject sit as before on the
rotary chair, but during the rotation hold his head horizontally to
the right or left, forward or back. With any of these positions of the
head, however, the rotation produced, on all of the subjects tried,
extreme dizziness and a feeling of nausea that lasted in some cases
for several hours. This fact made it impossible to ask for a set of
the four possible positions of the head from any of the subjects. The
following are the records that were obtained:

  Subject Fl. Head horizontally to left; rot. anti-clockwise.
              During rot.; after-im. moved clockwise, _i. e._, from
                             subject's brow to chin.
                           Eye-mov. not observable during rot.
              After rot.;  after-im. moved anti-clockwise, chin to brow.
                           Slow eye-mov. anti-clockwise, chin to brow.
                           Vis. field clockwise, brow to chin.

  Subject H. Head horizontally to right; rot. anti-clockwise.
             During rot.; after-im. clockwise, chin to brow.
                           Eye-mov. not observable.
             After rot.;   after-im. anti-clockwise, brow to chin.
                           Slow eye-mov. anti-clockwise, brow to chin.
                           Vis. field clockwise, chin to brow.

  Subject H. Same repeated, with same results.
  Subject H. Same as case of Fl., with identical results.

  Subject K. Head horizontally to left; rot. anti-clockwise.
             During rot.; after-im. not observed.
             After rot.;   after-im. anti-clockwise, chin to brow.
                           Slow eye-mov. anti-clockwise, chin to brow.
                           Vis. field not observed.

So far as these records go, they entirely confirm the results of other
investigators as to the direction and the reversal of the nystagmus. In
each of the cases the after-image moved with the slow eye-movements,
reversing its direction with these slow movements, while the visual
field whenever it was observed (the eyes were kept closed during the
rotation) moved in the opposite direction to that of the after-image
and the slow eye-movement. It is well known that after-images move
_with_ every involuntary eye-movement, and although they disappear
during voluntary eye-jumps,[21] they reappear at the end of the jump in
a position that is related to the new fixation-point exactly as the old
position was to the former fixation-point. These after-images, then,
are seen during the slow eye-movements whose direction they follow;
but are not seen during the quick movements, when they must naturally
move in the direction of these quick movements. And aside from this it
is possible to observe introspectively that the after-image disappears
at that side of the visual field toward which the slow eye-movements
tend, and is for a moment invisible before it reappears on the other
side of the field. As was shown above, the visual field always moves
opposite to the direction of the slow eye-movements, as must of course
be the case if there is no inhibition of vision during these movements.
The simultaneous appearance of the after-image moving with, and the
rest of the visual field moving contrary to, the direction of the slow
eye-movements, with a uniform absence of the converse phenomena, seems
to prove that vision is unimpaired during these slow movements, while
it is completely inhibited during the rapid phases of the nystagmus.

Purkinje himself[22] called the slower phases "involuntary and
unconscious," meaning by "unconscious" not that the visual field was
not seen (for it just then is seen), but that the movement of the
eyeball during the slow phases was not felt. I have observed, with
the confirmation of several subjects, that _this_ movement can also
not voluntarily be inhibited; whereas the swift movement is so far
voluntary that it can be inhibited at pleasure. It is possible, that
is, to fix the eyes on that side of the field toward which the slow
movements are directed, but not on any point at the other side of the
field. The slow movements, then, during which vision is possible, are
purely reflex. These slow movements, purely reflex and yielding clear
vision, with the rapid movements, partly under voluntary control and
attended by an inhibition of vision, present a parallelism, that may
be not without significance, to the "pursuit" eye-movements (Dodge's
"second type"), that are likewise relatively slow, are reflex, and
yield remarkably clear vision, and the ordinary voluntary eye-jumps
(Dodge's "first type"), that are relatively rapid, and are, like the
rapid nystagmiform movements, attended by a central inhibition of
vision.

FOOTNOTES:

[Footnote 16: Yves Delage: Physiol. Studien über d. Orientirung
(Aubert's transl.), p. 100, Tübingen, 1888.]

[Footnote 17: E. B. Holt: Harvard Psych. Studies, Psych. Rev. Mon.
Supplements, vol. 4, p. 1, 1903.]

[Footnote 18: E. Mach: Analyse der Empfindungen, 2d ed., p. 98, Jena,
1900.]

[Footnote 19: R. Dodge: Amer. Jour. of Physiology, vol. 8, p. 317,
1903.]

[Footnote 20: E. B. Holt: Harvard Psych. Studies, Psych. Rev. Mon.
Supplements, vol. 4, p. 42, 1903.]

[Footnote 21: S. Exner: Zeitschrift für Psych. u. Physiol., vol. 1, p.
46, 1890; E. Fick and A. Gürber: Berichte d. ophth. Gesellschaft in
Heidelberg, 1889; E. B. Holt: _op. cit._ p. 4.]

[Footnote 22: Purkinje, 1825; reprinted in Aubert's Physiol. Stud. über
d. Orientirung, p. 117, Tübingen, 1888.]




VISUAL IRRADIATION

BY FOSTER PARTRIDGE BOSWELL


There are various kinds of visual irradiation, of which perhaps the
best-known variety is that which appears as the enlargement of a
brightly illuminated surface at the expense of a contiguous one of less
intensity. This has been until recently the only form recognized, and
until very lately the greater part of the literature has dealt with it
alone.

The whole subject was carefully investigated by Plateau in 1831,
and retinal irradiation extricated from phenomena which very often
accompany it. He showed that the extent of irradiation varies with
the intensity of the stimulating light and the time during which it
is allowed to act. He was also the first to call attention to the
phenomenon of so-called negative irradiation.

Somewhat later Volkman again called attention to negative irradiation,
while Aubert, in opposing the explanation advanced by Volkman, first
showed the relations existing between irradiation and contrast.

Dove was the first to investigate the influence of irradiation on
stereoscopic pictures, thus calling attention to the question of
binocular irradiation. Experiments in this direction, however, have
in general given negative results in so far as any enlargement of the
binocular portion is concerned.

Helmholtz examined the manner in which the stimulation at the
border-line between a light and dark field changes in intensity,
and drew a curve showing these modifications of intensity due to
irradiation. Hering showed that the form of the Helmholtz intensity
curve would be modified by the presence of other phenomena not strictly
those of irradiation.

De Roux demonstrated the difference in the extent of a real induction
on the foveal and the extra-foveal parts of the retina.

Charpentier has attempted to carry forward the general explanation
by saying that this spreading of neural excitation, the existence of
which he proves to be beyond question, takes the form of an undulatory
excitation in the free nerve-endings of the retina. Bidwell has
investigated more thoroughly in some respects than Charpentier the
phenomena of the after-images of moving sources of light, which have
bearing upon irradiation. The same is true with regard to McDougall,
von Kries, Hess, and others. Burch has instituted investigations
along these lines, especially concerning the inhibition of stimuli
on contiguous portions of the retina. Hess has worked carefully
upon the different phases of the stimulation derived from a moving
source of light, the differences in functioning of the foveal and
extra-foveal parts of the retina, the respective functions of the
rods and cones, and in connection with this, made investigations in
the visual perception of color-blind subjects. All these observations
have important bearing on irradiation, contrast, and theories of
color-vision.

In connection with some work which was being done upon the after-images
of moving sources of light in the Harvard laboratory in the early
winter of 1903, some phenomena were observed which I believe are due to
one form or other of visual irradiation. They can be seen in various
ways, perhaps most advantageously by observing with fixed eyes the
passage of a luminous image over the retina. What one sees as such a
figure moves by is a travelling band of light, its forefront somewhat
like that of the stimulating source, the rest composed of a long
train of after-images which differ very decidedly from one another
in intensity and color. The advantage of this well-known method of
observation lies in the fact that it enables one to translate the
temporal relations between the different phases of the stimulation
into spatial relations between the different portions of the moving
band of light. For since the figure moves across in a plane before the
observer, that which appears in his consciousness first in time will
likewise appear as foremost on the plane in space. Thus by observing
the train of images one practically sees the different phases of
the stimulation spread out in order before one. The new phenomena
we observed, however, have to do with but a single phase of the
stimulation, the extreme front of the stimulating image.

The intensity of light used varied considerably with the differently
colored images, and was regulated so as to give as well as possible
the phenomena we wished to study. With white light the intensity was
less than that of an eight-candle-power electric lamp placed about ten
feet distant from the observer. When colored light was employed it was
necessary to use a very much stronger source of illumination, since
the colored glass which was used absorbed a great deal of light and
in case of colors lying toward the violet end of the spectrum greater
luminosity seemed demanded.

The apparatus used consisted of a three-foot pendulum with a screen
attached. This screen swung with the pendulum. In the screen was an
opening about four inches wide and three inches high, into which
strips of cardboard or tin backed by a piece of ground glass could be
slipped. In these strips differently shaped holes were made through
which the light passed. In this manner an image of any desired form
might be used. Behind the screen, between it and the lamp, was a frame
in which other pieces of ground or colored glass were placed. These
pieces of ground glass would reduce the intensity of the light and
diffuse it evenly over the image. The observer sat ten feet away. When
the pendulum was set in motion, the image would appear moving back and
forth in an arc. In order to shorten this arc and to aid the observer
in keeping his gaze perfectly fixed, a second screen was placed before
and very close to the pendulum, between it and the observer. This
screen was stationary. In it was a hole six inches long and two inches
wide. The top and bottom of this hole were arcs of circles parallel
with the arc in which the pendulum swung. The ends were radii.

The screen was so placed with reference to the observer that the moving
image would pass directly across the middle of the opening, appearing
from behind one side and disappearing behind the other. In the centre
of the opening, directly in front of the place occupied by the moving
image when the pendulum was at rest, were two luminous fixation-points,
one above the other, below the path of the moving light. In order
to measure apparent spatial differences between the phases of the
stimulation, two wires were stretched vertically across the opening
in the stationary screen. These wires could be moved nearer together
or farther apart. Thus by measuring the apparent distances in space
between the different parts of the moving figure a measure could be
had of their temporal differences in coming into consciousness. The
luminous image moved, during the time it was visible, at a velocity of
about one and a quarter feet per second. Since the observer sat about
ten feet from the instrument, this would be at an angular velocity of
about seven degrees per second. In one experiment a higher and a lower
velocity were also employed.

It was of course very easy to change the figures and vary them widely
in form, color, and intensity. Most of those employed, however, were
rather small, subtending an angular distance of not more than one
degree. Since the whole opening did not subtend an angle of more than
three degrees or so, nearly all the phases of the stimulation occurred
at the fovea.

We noticed that the form of the stimulating images themselves seemed
to suffer modification as the light swung by, not only because of the
train of after-images which dragged behind them over the retina, but
in other ways as well. For instance, a circular image (Plate III, Fig.
1) appeared crescent-shaped, and its forward edge possessed greater
curvature than the segment of the circle which produced it. It was
longer also from horn to horn than the diameter of the generating
circle, and a faint haze surrounded the points extending outward and
backward until lost in the blackness of the background. Von Kries
remarks that a circular moving image appears cylindrical in form with
a concave edge behind. By using a little higher speed we observed this
phenomenon. At first we thought the crescent-shaped image to be due
merely to an intensely black after-process, which Bidwell describes
as following the positive image of a bright white light. This, taking
place before the circular disc of light had gone forward a distance
equal to its own diameter, would overlap the bright image from behind
and a crescent-shaped figure would result, but the increase in width
and convexity of the stimulating image as well as the laterally
trailing clouds of light remained to be explained, and as this could
not be done in terms of anything which might happen to the back of the
image, another explanation had to be sought. In order to determine
the effect of the form of the figure used as a source of light on the
form of the apparent image, several differently shaped figures were
employed. In place of the original circle, an oblong pointed at both
ends was tried (Plate III, Fig. 2). The front of this figure appeared
very convex indeed, while the ends, which, owing to the shape of the
figure, were very much less effective as a stimulating source, trailed
far behind the centre.

A crescent-shaped figure (Plate III, Fig. 4) gave rise to a very pretty
phenomenon. When it moved toward its concave side, it appeared very
much less concave on that side than the real figure, but when it moved
the other way, toward its convex side, it seemed very much more curved
than it was in reality.[23]

[Illustration: PLATE III.]

No. 3, a simple oblong figure, appeared curved like the others, almost
as perfect a crescent as any of them.

The idea occurred to me that perhaps all these modifications in the
curvature of the figures could be explained if we assumed two things:
First: that there is a spreading of excitation from one portion of the
retina to another. Each point will therefore be stimulated not only by
the light falling directly upon it, but it will also derive a certain
reënforcement of its stimulation from the points surrounding it. Thus
a point lying toward the centre of one of these figures would be more
favorably situated for receiving reënforcement than one located toward
the periphery, where there are few neighboring points, and those lying
mostly in one direction, namely, toward the centre.

This may be represented diagramatically, as in the illustration
(Plate IV, Fig. 10), where the horizontal coördinates represent the
spatial dimensions of an oblong image and the vertical coördinates
the intensity of the excitation due to direct stimulation and its
reënforcement by surrounding points at various portions of the
figure.[24] Secondly I assumed that the stimulation at one part of the
figure being thus rendered more intense, that part would appear in
consciousness more quickly than the others and cause a modification
in the form of the figure.[25] For example, in the case of the oblong
figure, the light would be rendered most intense at the centre and
less and less intense toward the ends, for the points in the centre of
the figure will have their intensity increased by nervous excitation
spreading to them from points lying toward the ends. Those toward the
ends will be reënforced by light coming only from toward the centre.
Thus the intensity of the centre of the figure will be increased, and
as the figure moves across before the observer, the centre, appearing
first in consciousness, would likewise appear foremost in space, the
points near the centre a little later and so on, until finally, the
ends being the last to appear, the whole front of the figure would take
the form of a convex curve, after the manner in which it was observed.
The back of the figure also appears curved, probably because of the
fact that the front of the negative after-image, which closely follows
it, is of the same shape as the front of the positive image, as was
shown in the case of the circular figure.[26]

It is of course a well-known psychological fact that a light of
greater intensity will take less time in coming into consciousness
than one of less intensity. In this case, however, it was necessary to
find some way of showing such differences between lights which were
very little different in intensity. For one is practically unable to
see any difference in intensity between the parts of a stationary
image. So unless it could be shown that a difference in intensity
between two sources of illumination, so small as to be imperceptible
to the observer, will nevertheless make its presence known by the
appearance of the brighter light in consciousness before the other,
the explanation which I have suggested for the curvature of the images
would have to be abandoned.

The following experiments do show, as I believe, that of two sources
of light not perceptibly different in intensity, the brighter will
appear in consciousness before the other, and that in the case of these
figures the curvature of the image is due to a heightened intensity of
the light in the centre through reënforcement of the excitation there
present by stimulation spreading from the ends.


EXPERIMENT I

In the first of these experiments three dots of about three sixteenths
of an inch were placed in a vertical row about three eighths of an inch
apart (Plate IV, Fig. 1). No change was then observed in the form of
the figure. The row of dots swung across the opening in a perfectly
vertical line one directly above the other (Plate IV, Fig. 2). They
were presumably too far apart for irradiation to take place between
them. When, however, another dot was interposed between each end dot
and the centre dot (Plate IV, Fig. 3), so that the excitement could
extend from one dot to the next, the front of the line of dots no
longer appeared vertical, but decidedly convex, the centre dot being
perhaps three eighths of an inch before the dots on the ends (Plate IV,
Fig. 4).

[Illustration: PLATE IV.]

Absolutely the only difference between the two cases was that in the
one, irradiation presumably could not occur, while in the other it
conceivably could.


EXPERIMENT II

In the second of these experiments the curvature of a line of dots was
observed and measured. Then the centre dots were slightly darkened
(Plate IV, Fig. 5) by shading lightly with a lead pencil the ground
glass which travelled with the pendulum and held in place the card
from which the dots were cut, until the front of the image lost its
curvature and appeared vertical (Plate IV, Fig. 6). The pendulum was
then stopped and the row of dots observed closely, in order to see
whether the dots in the centre were perceptibly of less intensity than
those on the ends. No perceptible difference was found.


EXPERIMENT III

All the dots were covered, except the shaded central and the two
unshaded end dots, in order that no irradiation might take place
between them (Plate IV, Fig. 7). The pendulum was again set in motion,
and the centre dot, instead of remaining co-linear with the dots on the
ends, appeared considerably behind them (Plate IV, Fig. 3). This would
show that irradiation must heighten the intensity of the excitation in
the centre of the figure--for the two cases just mentioned are alike in
every respect except that in the first (Fig. 6), where the dots were
near enough together so that irradiation might occur between them, the
intensity of the centre dot, which was objectively fainter than the
end dots, was heightened enough by this induced excitation to appear
in consciousness as soon as the two end dots, which were objectively
of greater intensity; whereas in the second case (Fig. 7), where the
dots were too far apart for irradiation to take place between them,
the centre dot, being objectively of less intensity than the end dots,
appeared behind them.

These experiments show that of two sources of light very little
different in intensity the brighter will appear in consciousness
before the other. Other things being equal, the difference in
intensity may even be so small as to be imperceptible by direct
comparison; it is able nevertheless to make its presence known by
the order in which the lights appear. Exner made some experiments in
1868 to determine the time necessary for the perception of lights of
different intensity. He used, however, stationary images of brief
duration and tried to eliminate the effects of the after-image
by flooding the visual field with light. This method has its
disadvantages. It is incapable of measuring the minute temporal
differences in latent perception of sources of light very slightly
different in intensity.

While my method does not give the absolute time taken by any one light
to enter consciousness, it is a very much more delicate method than
Exner's for measuring _differences_ in time of latent perception of
sources of light very close to one another in intensity. It would be
a very easy matter, having found the time of latent perception for a
light of standard intensity, to determine by this method the time of
lights of greater or less intensity.

These experiments also show that when irradiation is absent, the
curvature of the images is absent; when irradiation is presumably
present, curvature is present. For I find, not only in these, but also
in a number of other experiments, that under all conditions in which
the presence of irradiation is to be expected, the form of the images
tends to be modified in precisely the manner that the assumption of its
presence would lead one to anticipate. In all cases where irradiation
is presumably absent, the contour of the front of the moving figure
depends entirely on the amount of light proceeding from its different
parts.

It is next in order to say something of the physiological causes of the
phenomena we have been considering.

It is probable from what has been observed that in the case of the
curved figures we are dealing with a form of visual irradiation which
is due to the spreading of neural excitation over or through the layers
of the retina. It is also evident from the close connection between
irradiation and intensity that it must be of such a kind that the
excitation produced in one part of the retina may communicate itself
readily to another part. We have also seen in the case of the moving
line of dots that the several dots could remain distinct from one
another and yet could reënforce each other by means of communicated
excitation. It must also be a very rapid form of irradiation, for the
curvature of the figures does not increase very much during the time
they are visible.

I think that the demands made by these different facts are best met
by assuming that the spread of the nervous excitation which gives the
reënforcement takes place in one of the interconnecting layers of nerve
cells and fibres underlying the rods and cones. The line of dots which
appeared curved and yet perfectly distinct from one another could
very well communicate excitation to one another along these fibrils,
and the intensity of one part be raised by the excitation of the
near-lying parts. The fact that the dots remain distinct would not be
contradictory. For in that case very near-lying parts might communicate
excitation to one another without arousing to any very great activity
the nerves that lead to the brain from the small unstimulated portions
which lie between them. In this manner the intensity of the centre dots
could be heightened enough to make the row appear convex, without any
merging into one another on the part of the several dots. The fact that
the dots do not fuse shows that the curvature is not due merely to a
forward-spreading of the excitation in the retina. However, there is
always a certain amount of light visible between the dots, with all the
colors. This is especially noticeable with green light.

The fact that the elements of the retina form a kind of concatenated
series from without inwards, a number of rods and cones corresponding
to but one ganglion cell, furnishes a further bit of evidence in
support of the explanation just advocated, since the irradiated
excitation would tend to be "drained off" through the group of ganglion
cells corresponding to the most highly stimulated portions and leave
the intervening spaces comparatively free from centrally proceeding
excitation. Thus also the individual dots in the five-dot figures may
appear entirely distinct from one another and yet the centre ones be
reënforced enough by irradiation to appear in consciousness in advance
of the others.


SUBSIDIARY EXPERIMENTS

A number of other observations were made which present various
exemplifications of the principles we have considered.


EXPERIMENT IV

An oblong figure, all its parts objectively of the same intensity,
had its ends slightly darkened. When this was done the curvature had
increased from twelve sixteenths to fourteen sixteenths of an inch.

The pendulum was stopped, and a very slight difference was perceived
between the ends and the centre of the figure. This difference in
intensity was greater than in the dot experiment, when the image had
been darkened enough in the centre to make it appear vertical, because
in this case, when the ends were darkened the centre would still be
reënforced by irradiation from a considerable space which intervened
between the shading and the centre.


EXPERIMENT V

The centre of the oblong figure was considerably darkened so as to
counteract the effect of induction. By properly varying the amount
of shading, one may make the front of the figure appear less convex,
vertical, or even concave. This shows perfectly the effect of
differences in intensity upon the curvature of the figure, but does not
show so neatly as the similar experiments performed with the dots, the
influence of the presence or absence of irradiation upon the intensity
of the centre of the figure and so upon the curvature.

The illustration shows a case where the centre was too much darkened.

[Illustration]

The two ends were comparatively free from shading. In each end-part
irradiation took place. The points lying toward the centres of these
ends received reënforcement, both from points lying toward the centre
of the figure and from the extreme ends, and so the centres of the ends
of the image were considerably brighter than either the extreme ends
of the figure itself, or the sides of the end-parts toward the heavily
shaded centre of the figure. Accordingly each end appeared convex for a
short distance. The whole figure, however, being considerably brighter
at the two ends than at the centre, on account of the heavy shading,
the ends appeared in consciousness first and the centre afterwards, so
that the figure as a whole seemed concave.


EXPERIMENT VI

An oblong figure was shaded rather heavily at one end, gradually
becoming lighter toward the other, while about a third of the figure
was free from shading. The shaded end always seemed to lag behind. The
extreme front of the figure was at a point a little distance from the
other end, before the shaded portion began. So that the front of the
whole figure appeared, not like a segment of a circle, but like part of
an oval with the bulge toward the brighter end.

Beyond the ends of all these images faint clouds of light were seen, as
has been mentioned before, extending outward and backward, gradually
decreasing in intensity, until lost in the surrounding blackness of the
background.

Charpentier's bands, sometimes more and sometimes less in number, were
observable in all of my figures and with all colors. Very often they
appeared to be parallel to the forefront of the image, or even of a
slightly greater degree of curvature.


EXPERIMENT VII

It is a well-known fact that a rotating color-disc, having colors which
just fuse at a certain intensity, will show flicker at a slightly less
intensity.

A color-disc was set in motion and the speed found where the colors
were on the point of fusing. A piece of black cardboard, with a hole
about an inch in diameter, was held close to the screen.

Around the periphery of the hole flickering appeared, while at the
centre there was fusion. (The cardboard was held very close to the
disc, so that there would be no shadows on the disc near its edges.)
This fusion at the centre of the disc is probably due to the fact that
the centre of the field is of slightly greater intensity than the
edges, owing to irradiation. This difference in intensity makes the
difference between the fusion at the centre and the slight flicker seen
at the periphery.

Karl Marbe in a recent article mentions the difference in fusion
between a point in the centre of the disc and a point near its border,
and he thinks the increase of flickering in the latter is due to
some influence on the part of the moving edge which separates the
different parts of the disc. It would seem more probable from this
last experiment that the fusion at the centre of the field of view
was due to reënforcement of intensity by irradiation, and that the
flicker about the periphery of the field was due to the lack of such
reënforcement.


EXPERIMENT VIII

Three large dots were used and the centre one covered with tissue
paper. The two end dots then appeared ahead of the centre dots. They
were larger than the centre dot, due to irradiation over their borders.
But this increase in size did not account for their position ahead in
space. The centres of all the dots were not co-linear, but the middle
dot was behind the others, thus, of course, showing the greater time
necessary for the perception of the less luminous object.


EXPERIMENT IX

This was exactly similar to the preceding, except that the intensities
of the various dots were reversed. The end dots were covered with
tissue paper, instead of the centre one. Then the centre dot appeared
first and the end dots after it.


EXPERIMENT X

Professor Hess finds that an image which, compared to those we
used, was very long, subtending an angular distance of about thirty
degrees, and which extends entirely across the fovea and overlaps the
surrounding parts of the retina will appear curved backwards at the
fovea, owing to the longer time of latent perception of the fovea and
the macula. The accompanying illustration shows a modification of one
of Hess's figures, in which the presence of this phenomenon and that
of the convex image are both shown. The two phenomena were observed
when a two-inch image was observed at a distance of about fourteen
inches. The intensity of the light was that of an eight-candle-power
lamp with three pieces of ground glass in front of it. (Very many of
Hess's intensities are too great to give convex images.) Thus the image
would be about 12° in height. About 5/12 of the figure would then fall
on the macula and fovea and appear curved backwards in relation to the
ends. The ends where they fell on the extra foveal parts of the retina
appeared convex in front and concave at the rear as any small image
of the right intensity does which falls on a homogeneous part of the
retina.

[Illustration: Figure observed with centre curved backward at the
fovea, and ends curved forward owing to irradiation.]


EXPERIMENT XI

Charpentier, Bidwell, and others have made the observation that if
a small source of light be exposed for a brief interval, excitation
will proceed out in all directions over the retina, but if the light
be exposed for a slightly longer period, the excitation will contract
again and the light appear nearly its proper size and in its proper
location at the stimulated portion of the retina. Using variously
shaped figures we obtained analogous results, and the additional fact
appeared that the outgoing excitation proceeds from the borders of the
figures and that its form is somewhat determined by the form of the
figure. An oblong image appeared vaguely elliptical, a diamond-shaped
figure in the form of a more pointed ellipse, etc. These images were
exposed for only a small fraction of a second, by means of a shutter.
As the exposure grew longer the true form of the figures came out more
and more clearly. There thus seems to be a general spreading of the
stimulation in all directions over the retina from the borders of the
images. Then, upon a slightly longer duration of the stimulus, this
very rapid irradiation of excitation contracts and the irradiation
becomes confined within the borders of the stimulated portion and
affects the intensity of the different portions of the image. With
strong intensities and certain colors it is, however, never wholly
confined to the stimulated portion even of moving images. Charpentier
speaks of "clouds of light accompanying his figures." With green light
these clouds are especially noticeable. His "palm branch" phenomenon is
a good instance of the irradiation of stimulation.

Besides these experiments which I have just described, several
phenomena of a like sort were observed in connection with other
experiments which were being performed in the laboratory at the same
time. Dr. Holt was experimenting with a bright circular spot of light
about one half inch in diameter, surrounded by a very faint ring about
one half inch wide. When the whole image was moved about, the spot
would seem to go back and forth across the less intense part so that
the whole image looked like a jelly-fish swimming about in the water.

When the figure was allowed to remain stationary for a few moments it
would resume its natural shape. Otherwise the bright part would seem to
advance faster than the rest, sometimes even overlapping the border.
This phenomenon was due to the fact that a bright light requires less
time in coming into consciousness than a less intense one, and is, of
course, the same in principle as those which were performed with dots
when the bright dot moved ahead of the rest.

Another one of these phenomena occurred when an isosceles triangle
was moved in a direction parallel to its base. The side toward which
it moved appeared curved forward, with the apex bent backward. Toward
the bottom, where there was the best chance for irradiation to have
its effect, appeared the most advanced portion of the figure, while
the bottom corner, although objectively the most advanced part of the
figure, appeared rounded off and somewhat behind the part just above.

A narrow, vertical image with a large bulge behind the central part
appeared with a large portion of this bulge in advance of the centre of
the figure.

All these experiments show that a more intense object is, other things
being equal, always located ahead of other objects co-linear with it.
And I assume irradiation to account for the priority in localization
of parts of the figure which are not objectively of greater intensity
than others, but whose position makes them subject to reënforcement.
The localization itself may be a function of more central organs, and
not directly a question of the coming into consciousness more quickly
of a more intense stimulation, although that seems to be the simplest
explanation, but in any case priority of localization varies directly
with the degree of intensity.

If the light is not bright enough to produce much irradiation the image
will lose its curvature. If the light is too bright, although there
may be a maximum of irradiation aroused and the absolute difference in
intensity between the ends and centre of the image be at its greatest,
yet this difference may not be great enough in proportion to the
absolute intensity of the light to make the centre of the image appear
in advance of the rest.

The curvature also varies with the angle subtended by the image and the
portion of the retina upon which the image falls. If the image were too
long, although all the processes which produce curvature be present,
yet the front of the image would still appear vertical, because of the
fact that each point in this long line would not derive reënforcement
sensibly greater than that of the neighboring points. The best one
could expect would be that these long figures should have their ends
rounded off, which is usually the case. Most of the images which
Professor Hess used in his experiments were too long to appear curved.
All the images whose curvature we measured did not subtend an angle
greater than 1° 10´, and were all seen on the fovea.

An image which subtends an angle of more than about 2° will hardly
appear curved when it passes over the fovea.

We were sometimes able to see the curvature reversed. This happened in
my own case about once in a hundred times, usually when my eyes were
fatigued by the repeated passing of the moving light back and forth
over the same portion of the retina. With other observers it occurred
more often.

Slight vertical differences in fixation would cause the central part
of the path taken by the moving light to become more fatigued than the
edges and so to respond more slowly to the stimulation and reverse
the curvature. It may be that some brain process which has to do with
the apperception of the form and movement of visual objects becomes
fatigued or does not always function properly, and so the curvature
of the image may sometimes appear reversed. At any rate the more
usual cases are those in which the convexity is present. The others,
owing to the number of factors involved, and the vast majority of the
opposite cases, may be regarded as due to temporary defects in the
psycho-physical mechanism, which when properly working would give the
more usual result.


QUANTITATIVE EXPERIMENTS

The object of the following experiments was to measure the amount
of curvature produced by differing degrees of intensity of light at
different speeds. An oblong figure was employed one fourth inch wide
and two inches long. As has been mentioned, two vertical wires were
stretched across the path in which the light moved. As the light swung
by, it was attempted to get the wires at such a distance from one
another that when one appeared tangent to the curve at the front of
the figure the other would seem to cross the image at the point of
intersection of the curve with the rest of the figure, as indicated in
the diagram. (Plate IV, Fig. 9.)

The distance between the wires was then read off on a scale. Thus one
was able to obtain a measure of the curvature of the figure when it was
moving at different speeds and illuminated by different intensities of
light, and to compare the observations of different subjects. The mean
error in this work is surprisingly little, considering the difficulties
in making the judgment as the light passed rapidly by the wires.
Usually the moving light had to be observed several times before the
curvature of the front of the moving image could be measured exactly.
It would be perfectly obvious that the front was considerably curved,
but it would often be wholly impossible to tell just how much it was
curved, until the pendulum had swung back and forth four or five times.
Fatigue and darkness adaptation modify the judgments considerably. If
one's eyes were partially adapted to darkness some little difficulty
was experienced in seeing clearly the curvature of the image. Fatigue
comes on very rapidly indeed. Usually it was impossible to get more
than four judgments without resting, and often only two could be made.
It was sometimes impossible to measure the curvature at the exact point
when the light passed under the cross-wires, so the curvature had to be
observed carefully and compared with the distance between the wires,
and a judgment made when the wires were not superimposed upon the
image. With each intensity of light two judgments were taken, one when
the cross-wires had to be brought nearer together, the other when they
had to be moved farther apart. Several series of measurements were made
by different observers, and the results averaged up and compared.

The following curves and tables give the different observations for the
nine different intensities of white light,[27] and the three speeds
which were used. In the case of the high speed the light moved across
the opening in the screen placed before the pendulum at a velocity of
about 1.5 ft. per sec. The middle speed was about 1.27 ft. per sec. and
the low speed about .917 feet per sec. In all cases an oblong image was
used, 1/4 inch wide and 2 inches long. The numerals on the left of the
plotted curves give the apparent curvature of the image in sixteenths
of an inch, and were obtained by measuring the distance between the
cross-wires when this distance measured the apparent curvature of the
image in the way described above. The figures at the bottom designate
the different intensities of light which were used. Number one is the
greatest intensity, number nine the least; the others those in between.

_High Speed._ This curve shows very well indeed what seems to be
typical of the relations between the intensity of the moving light
and the apparent curvature of the front edge of the image. With the
lowest degrees of intensity the amount of the curvature is very little.
Sometimes it was difficult to measure it at all. The light was so
faint and the speed so rapid that probably very little reënforcement
or irradiation took place, although what did occur would show its
presence most prominently, since, on account of the high speed at
which the pendulum moved, any part of the image which should come
into consciousness ahead of the rest, even by a very little time,
would appear considerably in advance of the rest of the image in
space. Of course a certain amount of time would be required for the
stimulus to spread itself over the retina, since it has to overcome
a certain amount of resistance in the nerve-layers, and if this time
were not given, the curvature of the resulting image would be of course
decreased. As the light brightened, however, the curvature increased
rapidly, until finally, when the intensity of the light neared its
highest point, the curvature ceased becoming greater, and finally
decreased. The mean error in eight judgments taken by two people for
each intensity of light was about .099 in.

[Illustration: High Speed]

[Illustration: Middle Speed]

[Illustration: Low Speed]

The measurements with the middle speed were very similar. The curvature
with the lowest intensity of light was somewhat greater than when this
same light moved with the highest speed. The maximum point of curvature
was reached with a light of less intensity, and the curvature was less.
When yet higher intensities were used, the curve decreased rapidly. The
amount of curvature was also much less with the brightest light than
with the higher speed. The following table shows the judgments of three
observers for this speed:


MIDDLE SPEED

  Intensities.     1     2      3      4       5      6     7     8     9

  First Subject.   8    10      9     11      10      9     9     7     8
                  10    11     13     13      10     11    10     9     8
                  10    10     11     11       9      8     8     7     7
                  10    11     11     14      10     12     8     9     9
                  10    11     10     11       9      9     7    10     7
                  11    11     11     13      10     10     9    10     7

  Second Subject. 10     9     11     11      12      8    11     7     8
                  12    14     13     14      15     14    10     9     9
                  10    14     13     13      12     11    11    10     9
                  11    14     13     13      13     12    12    10     9
                  13    11     13     14      12     12    10     9     9
                  13    12     13     14      12     12    11    11     9

  Third Subject.   7    10      9     13      10      9     9     7     8
                   8    12     12     13      12     10    10     8     7
                   7    11     10     10      11     10    10     8     8
                   9    11     13     11      12     10    11     8     8
                   9    11     11     10      10      9    10     9     7
                   9    11     12     12      11     10    10     8     8

  Average.    9-15/16 11-1/3 11-1/2 12-5/18 11-1/9 10-1/3 9-7/8 8-5/8 8-1/9

                  Mean Error, .075 in.

The low-speed measurements show the same general tendencies except
that the curvature is smaller with this speed when the faintest lights
were used than with any of the others. The maximum curvature is also
less and occurs with a more intense light than with the middle speed.
These modifications offer no special difficulties. Since the light
moves slowly, although the centre of the image, which is reënforced
by induced excitation from the ends, does appear in consciousness in
time ahead of the rest of the image, yet it does not appear so far in
advance of the rest of the figure in space as it would if the light
moved with a higher speed. While the same difference in brightness
between the centre and the ends of the image should make one part
appear in consciousness just as far ahead of the rest in time with
this speed as with any other, yet since the speed is slow it would not
appear to be so far ahead in space. The fact that the maximum amount of
apparent curvature is less would also be explained in the same manner.

When the high and middle speeds were used the results were surprisingly
consistent and the variations between the observers not very great.
With the low speed the individual differences are very much more
prominent. Uncertainties and variations between observers and between
different observations of the same observer became greater and greater
as the intensity of the light decreased. One seemed to be approaching
the lower limit of induction, below which, even if the spreading of
light stimuli through the retina took place at all, it was to such a
slight extent that it made no very marked difference in the appearance
of the moving image. Individual differences are very great in this
respect. For instance, my own average measurement was 15/64 in. of
curvature for the image produced when a light of the lowest intensity
moved at the lowest speed. Mr. Vaughan's measurements averaged 30/64
in., a measurement of just twice as much for the same light at the same
speed.

So far there has been given only a general view of what happens when an
oblong moving image appears convexly curved. It may be well to consider
the different causes which determine a certain curvature of the image
and see how they are related.

As we have seen, the curvature is a function of the difference in
intensity of various excitations between the centre and the ends of the
retinal area excited. This difference is modified both by the objective
intensity of the light and by the speed at which the light moves. Its
efficiency to produce curvature is also modified by both these factors,
since it requires a certain small amount of time for the irradiation to
take place. If this time is not given by the too rapid passage of the
image over a certain part of the retina, the difference in intensity
will be lessened and the curvature therefore be decreased. On the other
hand, if the figure moves too slowly, although this difference in
intensity may be as great as possible for the brightness of the light
which is acting, yet the curvature may be lessened on account of the
fact that, although the centre of the image does appear ahead of the
ends in time, yet it does not appear so far ahead of the ends of the
image in space as it would if the same difference in intensity were
present and the image moved more rapidly.

It will be remembered that the intensity of the centre of the figure
owes its increase to reënforcement by excitation irradiating from the
surrounding points. It seems only reasonable to suppose that this added
intensity due to irradiation does not increase without limit, or with
exactly the required ratio to produce a curvature of the front of the
image which becomes continuously greater as the intensity of the light
increases, indefinitely.

If this be so, then, at a certain brightness, the difference in
intensity between the ends and the centre of the image will have
reached a maximum, and a further increase in brightness of the light
will not serve to increase the apparent curvature of the image, but
rather to decrease it.


COLOR IRRADIATION

The color of the image has a decided influence upon the amount of
perceived curvature, independently of the intensity.

The following experiments were performed with lights of different
colors, in order to investigate the relations between the kinds and
amounts of irradiation of the different colors by comparing the amounts
of the curvatures obtained. We encountered a good deal of difficulty in
fixing upon a proper method of comparison between the different colors.
Finally it was decided to use such intensities of light as would give
a maximum amount of curvature with each of the four primary colors,
to measure this amount in each case, and also to measure the amount
obtained when the intensity of the light was greater and less than that
required to give the maximum.

It was found that very different objective intensities of light were
required to give a maximum amount of curvature with the different
colors. The colored images were obtained by placing colored pieces of
glass in a frame which stood before the source of light. The intensity
of the light could be regulated by interposing or taking away pieces
of ground glass which rested in the frame between the light and the
colored glass.

The red glass gave a nearly saturated color, but its place in the
spectrum was rather nearer the orange than I could have wished. It
was a thick piece of glass and absorbed a great deal of light. A
32-candle-power light with four pieces of ground glass in front of it
gave a maximum curve for most observers.

The yellow gave a very well-saturated color with light from the
incandescent lamps which we used. The glass was thinner and absorbed
less light than the red. A 32-candle-power lamp with three pieces of
ground glass usually gave the maximum.

Two 32-candle-power lamps and one of 24-candle-power were required with
the green.

The green glass was not quite so saturated in color as the red or
yellow. It was a slightly yellowish green. Red and yellow rays were
visible through it to some extent when it was examined through the
spectroscope. It absorbed somewhat less light than the red and
decidedly more than the yellow. The maximum curvature was obtained when
the source of light was screened with four pieces of ground glass.

The blue glass was a bluish violet, very heavy, and absorbed a great
deal of light; it allowed many red and violet rays to pass through. It
was necessary to use with this glass two 32-candle-power lamps and one
100-candle-power. When the combined light of these lamps was reduced by
interposing three thicknesses of ground glass, the maximum curvature
was observed. The light which then appeared, however, seemed of greater
intensity than any other which gave a maximum.

The curvature of the white light was measured again in order to compare
it with the colored lights. This was necessary, since the work was done
with a different set of subjects and the former work showed individual
variations. An 8-candle-power light was used as before. This, reduced
by four pieces of ground glass, gave the maximum in most cases.

The following curves and tables show the average of the observations of
four subjects. In the table the figures under the columns numbered 1
and 2 represent the amount of curvature perceived when the intensity of
light was greater than that required to give a maximum under 4 and 5,
when the light was not strong enough to produce a maximum of curvature.
The columns numbered 3 represent the greatest amount of curvature
perceptible with each color.

The curves shown in the diagram represent these measurements plotted
out.


TABLE

    Intensities.        1       2       3       4       5

    Red.              13.50   15.20   16.85   14.06   13.46
    Yellow.           13.90   15.46   18.00   16.40   15.85
    Green.            15.66   18.00   19.86   18.32   18.00
    Blue.             13.00   14.15   16.85   15.50   14.09

    Average for all
      the colors.     14.00   15.70   17.90   15.90   15.35

The measurements were made in the same way as before, and are given in
sixteenths of an inch.

In the diagram the abscissas represent the different intensities, the
ordinates the amount of the curvature. To avoid confusion, the curve of
the average of all the colors is left out of this diagram.

[Illustration]

It will be noticed in these records that the different colors give very
different measurements of curvature. Green gives by far the largest,
being greater than any of the others at every point. Since the process
of obtaining the curvature was the same with all the colors, these
differences in curvature can only be due to inherent differences in
the processes which give the sensations of the different colors.
It cannot be due simply to one sort of intensity process, the same
for all the colors, otherwise the curvature of all the colors would
be the same. At the same time the curvature of the image is due to
differences in intensity of excitation between one part of the image
and another. There must be, therefore, a retinal excitation in some
respects different for each color, capable of its different degrees
of intensity. Of course these individual differences would have a
decidedly limited range, for, as every one knows, if the intensity of
a color be increased sufficiently its saturation vanishes and white
appears in its place, while if the intensity be decreased without
limit, black appears. It may be that different degrees of excitation in
the different processes have different rates of time in coming into
consciousness, so that an equal degree of difference in excitation
between the ends and centre of the green image and the ends and centre
of the red image would give decidedly different amounts of curvature,
if it took a longer time after the centre of the green image had
appeared in consciousness for the ends to appear than it did in the
case of the red.

The time-differences might be greater with the same differences in
intensities of excitation with one color and another. Or it may be
that the excitation spreads in a different manner with each one of the
colors, and therefore gives differing degrees of reënforcement with the
different colors, and thus produces different amounts of curvature.

It is noticeable also that the amounts of curvature are related to one
another in a peculiar way. Green has the greatest amount of curvature,
yellow the next. Red is greater than blue with the higher intensities,
they are equal at the maximum, and blue is greater than red when the
lower intensities are used.

When a spectrum showing a fair degree of saturation is observed, it is
seen that the point of greatest brightness lies in the yellows. As the
intensity is heightened, this point moves toward the red end, and as it
is lowered, it moves toward the blue.

It will be seen that the relation between the different amounts of
curvature for the different colors is the same as that between the
different degrees of apparent brightness when the intensity of the
colors in the spectrum is decreased. It is not that of the extreme case
of the phenomenon of Purkinje, but when the point of brightness has
moved from the yellow into the yellowish greens or decidedly to the
right of the place it occupies in the normal spectrum. In that case
yellow would be the color second in brightness. In our measurements
the amounts of curvature obtained from yellow images were next in size
to those of green. The red and the violet-blue which we used would
therefore be about equal. It is a noteworthy fact, however, that when
the intensities of light (1) and (2) are too great to give a maximum of
curvature the amount of curvature obtained with the red is greater than
that of the blue, while with the intensities which are too small (4)
and (5) to give a maximum the blue curve is greater than the red.

As yet it has not been possible for me to find an interpretation of
these facts which would seem to meet all the requirements, and I
should not wish to offer any explanation at present. The question
of the possible connection of this phenomenon with Purkinje's is
probably important for any explanation, though it is possible that the
arrangement of the curves is merely a coincidence, yet this hardly
seems likely, and it would seem as if an explanation of the connection
would involve an attempt at explanation of Purkinje's phenomenon, and
lead at once into the most doubtful problems of the theory of visual
sensations.

It is also noticed that the series of numbers obtained when the
amounts of curvature of the different colored images, at the different
intensities given in the foregoing table, are averaged up, and the
curve of the average of all colors is thus obtained, that this average
curve is very like that obtained for the white light. These curves and
the series of numbers which they represent are here given.

[Illustration]

    Average for all colors.   14.00    15.70    17.90    15.90    15.35
    Curve for white light.    11.50    15.00    17.80    15.75    14.25

It will be noticed, however, that the curve for the white light, while
nearly equal to that of the average for the colored lights at the
maximum point, nevertheless falls considerably below it at each end.
This may possibly be due to the fact that with white light it was only
necessary to use an 8-candle-power lamp as a source of light, so that
when pieces of ground glass were interposed in order to reduce the
intensity of this light, very much greater reduction would occur with
this comparatively weak source than would take place with an objective
source of light of far greater brilliancy, as was the case with the
colored lights. Hence there would be a greater difference in absolute
intensity between intensities 1 and 3 with the white light than between
intensities 1 and 3 with any of the colored lights, or that represented
by their average. Thus the falling of the curve of the white light
at each end may possibly be due to the fact that there is a greater
difference in intensity represented by these parts of the curve in the
case of the white light than is represented by the analogous portions
of the curve of the average of the colored lights.

It will be remembered that these measurements were obtained when
the image was upon the fovea, so that the white obtained was "cone
white," and not due in any way to the functioning of the rods. It is
interesting to note that the curve of the white is very near that of
the average curve of all the colors, though I should hesitate to draw
the conclusion from this that "cone white" is due to a mixture or
fusion of all the excitations corresponding to the different colors.

In regard, however, to relations of the amounts of curvature of the
images, there are several further considerations which ought to be
noted. In the first place all three measurements were made when the
images were entirely on the fovea. In the fovea there are no rods, so,
whatever the connection of these facts with Purkinje's phenomenon, it
is one which has to do with the functions of the cones alone.

Professor Hess, in his experiments upon totally color-blind subjects,
found that exactly the same oscillatory processes in the course of the
stimulation occurred with them as with normal subjects. He also found
that the difference in the time of latent perception between the foveal
and extra-foveal parts was the same for one set of subjects as for the
other. The sole difference seemed to be in the one fact of not being
able to perceive colors. From these facts it does not look as if the
difference between seeing colors and color-blindness were by any means
always due to the absence of cones in the color-blind eye. It may of
course be true that an eye which is deficient in cones or which has
a lesion of the fovea would have poor color perception. But it seems
also true that an eye which, in so far as the rods and cones and their
purely retinal processes were concerned, seems to be normal in every
way, except perhaps that somewhat different intensities were required
to give the same reactions (which might be explained by different
central processes), may nevertheless belong to a person who is totally
color-blind or totally unable to perceive colors with that eye.

If this should prove true, the cones would still be regarded as the end
organs of color perception, but the cones would only give sensations
of color when functioning in conjunction with some other more central
process. The usual cases of color-blindness would be attributable, not
to any deficiency in the cones or any other retinal process, but to a
defect in this more central process, which, working in conjunction with
the cones, gives us our sensations of color.

The usual views of the functions of the rods would not be affected by
these considerations. They would continue to be regarded as end organs
whose main business it is to deal with weak stimulations and to notice
movement in objects whose images fall upon the periphery of the retina.

But the main difference would be that all cases of partial
color-blindness and most cases of total color-blindness would be
explained by lesions in the brain rather than abnormalities of retinal
structure.


VARIOUS FORMS OF IRRADIATION

The endeavor to explain these phenomena of moving images which we
have been considering and an examination of the literature of the
subject have led me to conclude that there are five distinct types of
irradiation. These are:

1. Irradiation α. The very rapid spreading of the nervous excitation
over the retina, which extends far beyond the borders of the image
and which occurs immediately upon stimulation. It is most distinctly
observed with stationary sources of illumination of the briefest
duration perceptible. This kind of irradiation has been discussed at
length by Charpentier and Bidwell.

2. Irradiation β. As the apparent form of the moving image becomes
distinctly perceptible, such irradiation takes place within the
confines of the stimulated portion of the retina, so as to make the
excitation present at favorably situated localities more intense than
that of other places. The portions which are so situated as to receive
this reënforcement are the first to enter consciousness. The various
phenomena discussed earlier in this paper furnish examples of this
process, as well as the phenomenon of the curved image.

3. Irradiation γ. After, and in part during, the rise and development
of the reënforcing irradiation, emanations of nervous excitation of
small intensity proceed from the borders of the stimulated portions
and from the after-images, rapidly extending themselves over the retina
and gradually decreasing in intensity.

4. Irradiation δ. When two fields of different intensities are brought
into juxtaposition, the field having the greater intensity will enlarge
itself at the expense of the other. This constitutes what has been
usually termed irradiation, and is observable with stationary objects.
This enlargement varies with the time during which it is observed, the
absolute intensity of the light employed, and the relative differences
in intensity of the two fields. Its angular extent under determinable
conditions is constant, although it varies considerably from one
observer to another, and with the same observer at different times.
Its physiological explanation is probably similar to that of the
other kinds of irradiation, viz., the spreading of nervous excitation
over or through the layers of the retina, although various factors of
accommodation, dispersion, achromatism, astigmatism, etc., enter in
and modify the totality of the phenomenon. It will be noticed that
reënforcement occurs in this kind of irradiation as well as in certain
of the other forms. The sides of the dark fields upon which this
form of irradiation shows itself appear curved inward at the centre,
apparently showing the presence of a greater excitation in the lighter
fields next to the centre of the darker ones.

5. Irradiation ε. When a luminous object has been observed for a long
time (from thirty seconds to several minutes), the whole surrounding
field will be flooded by a faint haze of light, which within certain
limits increases in intensity the longer the stimulation is present.
This phase has many characteristics of the first and most rapid kind of
irradiation, and possibly represents a discontinuance of functioning,
through fatigue of certain nervous mechanisms which prevent the
spreading, or inhibit the perception, of this irradiatory excitation
after the form of the object is distinguished clearly. It is probably
largely due to such a mechanism that we are enabled to perceive as
clearly and sharply as we do the outlines of objects which differ
greatly in intensity from their backgrounds.

W. McDougall has developed a theory of inhibition[28] which he uses
to explain the more usual kinds of irradiation. This explanation
harmonizes very well with the results of my own experiments and helps
to explain all the kinds of irradiation we have distinguished. Briefly
stated the theory of inhibition is this: there is a transference of
nervous excitation or energy through the nerves and from one neurone to
another. This living nervous energy he calls neurin. The place where
it crosses from one neurone to another he calls the synapse.

Of course these conceptions are not to be taken too literally. They
seem to be rather, if they are to be of any value at all, a convenient
way of handling certain neurological processes of which at present we
know very little, but whose grosser modes of action are comprehended
more easily by the use of such terms as "resistance," "neurin," etc. It
is in this manner that I wish to be understood in the use I have made
of Dr. McDougall's valuable contributions to the methodology of the
subject.

Neurin is generated when a stimulus is applied to the afferent nerves.
When a strong stimulus is applied, neurin is generated rapidly, and
discharges across the synapse to the efferent neurone in a series of
very rapid discharges like the multiple discharge of a Leyden jar.
When the stimulus is weak the discharges take place more slowly.
Consciousness occurs at the time of the discharges and occurs in
pulses. When these pulses occur in very rapid succession we experience
a continuous sensation, when the discharges take place at a lower rate
we are conscious of a pulsative sensation, as for instance, in the
visual phenomenon of Charpentier's bands.

Continuance of stimulation continues to produce neurin, but the
multiple discharges caused by the incoming neurin cause fatigue in
the synapses, and the neurin seeks new paths of discharge through
unfatigued synapses.

The resistance of the synapses is first lowered by the incoming neurin,
then raised again through fatigue. When the resistance is first lowered
upon application of the stimulus, the neurin which might go through
other channels of discharge is "drained off" through the synapses which
have their resistance thus lowered, then as the resistance is again
raised through fatigue, it again seeks discharge through synapses which
are unfatigued.

Applying these conceptions to the different kinds of irradiation we
have distinguished, we can bring them all under one category. One
might remark in passing that, in so far as our purposes are concerned,
it makes very little difference whether we regard consciousness as
occurring upon the crossing of neurin from one neurone to another, or
upon the charging and discharging of a cortical cell, so long as the
conditions already referred to are maintained, viz., first, a lowering
of resistance as the incoming nervous excitation finds its way through
the cell or across the synapse, and then the gradual rise of resistance
and its conduction into new channels by fatigue of the synapse, or
exhaustion of the cell and a consequent turning of the excitation
through fresh cells across fresh synapses before its passage into the
efferent nerves.

When a light stimulation falls upon the retina, during the first one
hundredth or one fiftieth of a second the nervous excitation of neurin
will spread about generally through the retina for a considerable
distance from the point immediately excited. Thus by means of the
fibres of the retina faint excitations will go to the brain from all
these different points, so that one will perceive a faint cloud of
light, similar to that described under the first kind of irradiation.
Moreover, since the portion of the retina directly stimulated by the
light will have the most intense stimulation, this part will come to
consciousness somewhat more quickly than the outlying parts, so that
the cloud of light will first seem to spread outward from its source,
and then, as the resistance in the synapses is lowered through the more
intense stimulation of the part of the retina upon which the light
directly falls, the outrunning excitation will be "drained off" from
these portions of the retina outside of the borders of the image, and
the halo or cloud of light will appear to contract again. This was
observed by Charpentier and Bidwell, and in our own experiments.

Moreover, in case the synapses corresponding to the portions of
the retina indirectly stimulated should have themselves periods of
discharge and periods of charging, we might expect to see dark rings
upon this halo, this was also first observed by Charpentier and Bidwell.

Secondly, as the resistance is lowered in the central organs
corresponding to the end organs of the retina upon which the
stimulation falls, the image tends to assume its true form, but
irradiation has been, and probably still is, present through the layers
of the retina, so that certain favorably located portions of the image
secure reënforcement by means of this irradiation, in the manner
described, and these portions appear in consciousness sooner than
the others. This reënforcement, in the case of the travelling oblong
image, will make it appear convex. Moreover, since the resistance of
the synapses corresponding to the centre of the oblong images will be
less than those corresponding to the ends, there will be a certain
tendency to "drain off" the stimulation from the rest of the image, a
sort of reënforcement of the reënforcement, which will also help in
making the image appear curved. Of course all the conditions which we
found to modify the curvature of the images will still hold good, these
conceptions being used only to describe the course of events which
causes the image to appear convex. Thus a very weak or a very intense
or a very long or an excessively short image will not appear curved,
owing to a lack of difference in intensity between the ends and the
centre great enough to produce perceptible curvature.

As to the third kind of irradiation, that which proceeds from the ends
of the moving image over the unstimulated portions of the retina,
and which has the appearance of long streamers of light extending
outward and backward from the moving image, this may be regarded as
being in certain respects a form of the first and very rapid kind
of excitation. It may well be that all the outrunning excitation
which occurs immediately upon stimulation does not find its way to
the central organs through those nerve-paths which correspond to the
directly stimulated portions of the retina, even after the form of
the image may be very clearly determined, but that some excitation
proceeds outward from one retinal element to another, arousing fainter
and fainter excitation as it proceeds. This being the case, we should
expect to find these streamers of light from the ends of the image
extending outward and backward over the retina. Of course the faster
the image moved and the more intense it was, the longer then would be
these streamers. For if the image moved very fast, very much less of
the excitation would be "drained off" through the directly stimulated
portion, and thus more of the excitation would be left behind, so to
speak, by the image when it moved along rapidly, and this would appear
to drag farther and farther behind. Of course these streamers being
curved backward would appear more curved the faster the image moved,
and if the pulsative processes occurred with these stimulations which
occur in the course of other retinal stimulations, we should have
Charpentier's "palm-branch" phenomenon.

The fourth kind of irradiation which we have defined is of course
the best-known form, and is that which has been the most discussed
by the many writers on the subject. It will be remembered that this
form appears in stationary objects which have been observed for some
little time (from four to ten seconds), and consists in the apparent
enlargement of a more intensely illuminated portion at the expense of
a less illuminated one. This enlargement occurs after all trace of
the first kind of irradiation has vanished, and of course no trace of
the third kind comes in, since the object is stationary. The course
of events may then be somewhat as follows. In the first perception
of the object we have the wide-spreading irradiation described. Then
way is made through the synapses corresponding to the stimulated
portion of the retina, and the wide-spreading irradiation is drained
off through these open channels, so that the image contracts again to
its proper size. But at the same time it is not likely that there
will not be a slight irradiatory enlargement of the borders of the
image. For irradiation is present within the confines of the image.
This is shown not only in the case of moving images, but also in the
fact that the edges of the less intensely illuminated portions of the
field are curved inward, this being most probably due to the fact that
the centres of the contiguous luminous portions are reënforced by
irradiation proceeding from the direction of both the ends.

Not all of the excitation proceeds to the brain from the directly
stimulated portions of the image merely, but a little irradiates
over the borders and causes an apparent enlargement of the brighter
field. It has also been shown by Plateau and others that the amount
of irradiation increases both with the intensity of the stimulation
and with the time during which it acts. Of course, as to the intensity
there is no question. As to the time-element, it may be that the
excitation at the border spreads rather slowly outward after the
previous contraction of the image to its proper dimensions, which takes
place within a very short time after stimulation, until a sort of
balance is reached between the tendency of the image to enlarge itself
through irradiation and the tendency for this irradiatory excitation
to be drained off through the nerves corresponding to the stimulated
portion of the retina, after which no further apparent enlargement
takes place.

In some of our experiments with dots we found that after a dot of the
proper intensity of illumination had been steadily gazed at for some
time the centre would appear dark. This seems to be due to the fact
that the centre of such an image was reënforced by irradiation, so
that the nervous mechanism corresponding to it became fatigued more
quickly and the stimulation at the centre no longer gave such intense
sensations as the rest of the figure, but appeared darker.

Passing to the fifth and last variety of irradiation, this seems due
to fatigue in the inhibiting apparatus which reduced the spread of the
first kind of irradiation. Following out the scheme we have applied,
it would seem as if the channels which were first opened by the direct
stimulation became blocked through fatigue, and, therefore, the
excitation produced in the retina were forced to seek new paths through
to the brain by means of the nerves which proceed from the unstimulated
portions of the retina. Thus if the resistance through fatigue occurs
slowly, the excitation which spreads may increase in intensity and in
extent. So, as the resistance increased, a portion corresponding to the
directly stimulated portion and its slight irradiatory enlargement of
the borders would be surrounded by a cloud of light growing in size and
intensity.

Of course the limiting case would be when the external cloud of light
attained as great or even greater intensity as the stimulated portion,
but such a case would probably be impossible to realize because of
other conditions which would prevent.

It may be that this fifth variety is caused partly at least by a
cortical spreading of the excitation. But it seems to me more likely,
in view of the fact that we could find no irradiatory enlargement of
the binocular portions of stereoscopic images and for a number of other
reasons, that the induction is retinal in character and that after
the resistance through fatigue has arisen in the central organs the
stimulation spreads out over the retina to the unstimulated portions
of the field and proceeds from thence to the brain. This seems more
probable than that the stimulation continues to be confined merely
to the stimulated portion of the retina, but seeks passage from one
portion to another of the brain through fresh neurones which branch
only from those nerve-tracts which proceed from the directly stimulated
portions of the retina.

To conclude; we have seen that there are various forms of irradiation
which take place during the perception of stationary and moving sources
of illumination.

That there are certain modifications in the form of a moving image
which are probably due to one of these processes.

Concerning color irradiation it was found that the curvature of the
images varied with the color of the light, so that a figure illuminated
by a colored light of one intensity would not have the same curvature
as one illuminated by a light of the same intensity but of another
color. Green gives the greatest curvature, yellow the next, red and
blue about the same. In other words the differences in curvature of the
images follow the order of the brightness of the colors in a spectrum
the intensity of which is much reduced.

From a consideration of these phenomena we were led to discuss the
functions of the rods and cones in the retina of the eye, and the
suggestion was made that differences in color-vision were due to
central rather than retinal processes, and that in many cases of
partial or total color-blindness the retina would be found normal and
the defect in vision due to a lesion in some more central structure.

The various forms of visual irradiation which have been described
by a number of different writers we found to be all forms of one
rather simple process. Resistance, removal of resistance upon further
stimulation, and recurrence of resistance through fatigue in some part
of the optic tract, together with the spreading of stimulation over the
retina (probably through the molecular layers) from one afferent nerve
to another are assumed as the minimal requirements which are sufficient
to explain the five forms of irradiation which have been considered in
this paper.

FOOTNOTES:

[Footnote 23: Image no. 5 appeared with the concavity in front. In the
centre of the figure appeared a dark grayish splotch of light, very
much darker than the rest of the image. This is due, most probably, to
the presence of Charpentier's phenomenon of recurrent bands. If this
happens in this figure the ends of the recurrent bright images would
overlap while the centres would not, so that the black bands appearing,
as it were, through interstices in the central part of the figure,
would seem like a dark splotch, especially since the outlines of the
bands are vague and hazy. The back end of the figure had the effect of
being vertical or nearly so. This is probably due to the same cause as
that which made the circular figure to appear as it did, namely, the
negative after-image overlapping the positive after-image. The front
of this black image is usually of about the same shape as the front of
the real figure which it follows. If this is so, then, in this case, it
would make the back part of the image pretty nearly vertical.]

[Footnote 24: The intensity of the objective stimulation may be
represented by the line _AB_. If there were no reënforcement of
stimulation the whole figure would be flat on top and of this height.
The difference between _AB_ and _AC_, or _BC_, represents the increase
of intensity due to irradiation at the most favorably situated portion
of the figure. The other portions receive increments proportional to
their location, as indicated in the diagram.]

[Footnote 25: Favorable localization will of course depend largely on
the shape of the figure in which the point is situated. Thus one in the
angle of a triangle or at the horn of a crescent would have much less
reënforcement of excitation than another point, say halfway down the
side.]

[Footnote 26: Bidwell describes this "black process" or "negative
after-image" of a bright, white light as being of a blackness more
intense than the ordinary blackness of an entirely dark room. This is
perfectly true. The black image, however, lasts for a very much longer
time than the recurrent images of the same light. Often this velvety
black band would trail along behind the moving light for the distance
of a foot or more, gradually lightening into the darkness of the
background.]

[Footnote 27: The intensities used with white light are all less than
an eight-candle-power electric lamp placed at about a foot behind the
opening and covered with two pieces of ground glass.]

[Footnote 28: McDougall, 1901, 1903.]




FEELING




THE EXPRESSION OF FEELINGS

BY F. M. URBAN


The material of this paper was obtained by an experimental
investigation which was carried on in the Harvard laboratory from
February, 1904, till June, 1905. The immediate purpose of these
experiments was a study in the expression of the feeling-tone of
simple sense-stimuli. Breathing and circulation were the functions the
changes of which were observed by tracing the curves of thoracal and
of abdominal breathing and the sphygmographic curves simultaneously.
Acoustical, tactual, pain, and smell sensations were studied in this
way, special attention being devoted to the smell and pain sensations.
These stimuli have the advantage that the physiological reactions of
the subject are more uniform than the reactions to other stimuli.
The number of experiments performed in this investigation was large,
although a subject was never experimented on for more than forty
minutes, because the facilities of the laboratory allowed a continuous
experimenting for several hours a day on different subjects. All the
experiments were performed on trained subjects. Only the changes in
the form of the sphygmographic curve will be discussed in this paper.
The results of this observation confirm the observations of previous
investigators in so far as the same changes in the curves were observed
and the introspections of the subject were, on the whole, similar to
those obtained by other observers. It does not seem probable, however,
that a satisfactory discussion of the results can be given on the basis
of merely mechanical measurements of the curves, and it, therefore,
seemed necessary to reconsider the principles of the theory of the
sphygmographic curves.

There are two methods which can be applied to the study of the
psychology of feelings. They are called the method of impression and
the method of expression. The first is a purely psychological method,
while the latter is confined by its definition to the study of the
physiological changes which are the accompaniments of feelings. The
method of expression is never used as a pure method in investigations
which are carried on for psychological purposes, because the
introspections of the subject must be compared with the physiological
results. It therefore has the character of a mixed method. The first
experimental investigations into the psychology of feelings were
started by Fechner, who employed the pure method of impression. At
this time, however, the apparatus for studying the circulation had
been greatly improved and sooner or later these instruments were sure
to be used for a more exact study of the influence of feelings on
circulation. It was to be hoped that the crude observations on the
changes of the heart-beats and of the circulation under the influence
of feelings might be followed up in detail.

Darwin laid stress on the importance of certain bodily accompaniments
of feelings, and he inaugurated the genetic explanation in this
field. But even if the genetic explanation is successfully carried
through, human psychology remains unexplained, and, furthermore, those
emotional expressions which Darwin described form only a part of the
physiological accompaniments which may be observed with the instruments
now in use. The invention or at least the great improvement of these
instruments is due to the investigators in the middle of the last
century, and a more thorough understanding of the delicate changes of
respiration, circulation, and of temperature was not possible before
the construction of these sensitive recorders. It seems that Mosso was
the first to observe these small changes under the influence of mental
activity in general, and feelings in special; in this sense it may
be said that Mosso started the experimental physiology of feelings.
The discovery of the influence of feelings on circulation is very
important, and it is to be appreciated that Mosso saw these slight
changes which escaped an observer like Marey. In the Mémoire offered
to the Academy[29] on March 26, 1860, Marey gives a great number of
circumstances which influence the sphygmographic curve, but feelings
or mental phenomena are not mentioned in this list. It is true that he
speaks in a later publication[30] of the influence of "moral ideas" on
the circulation and makes the hypothesis that these ideas influence the
circulation in the same way as other disturbing influences, _i. e._, by
changing the peripheral resistance. At this time Marey was already in
possession of his sphygmograph, but nothing in this passage indicates
that he saw the influence of feelings on the tracings. On the contrary,
the words "Sans rien livrer à l'hypothèse" seem to indicate that Marey
had no other facts in mind than those commonly known. He certainly did
not follow up his observation, and his statement at this point does not
differ very much from the observations of the old psychologists, that
emotions change certain physiological functions, of which a more or
less complete list is frequently given.[31]

It certainly is a long step from this vague statement to Mosso's
experimental investigations. His new instruments, the plethysmograph,
and the balance, enabled him to study the distribution of the
blood,[32] and he observed the influence of mental phenomena on the
circulation,[33] on the bladder,[34] and on the temperature of the
brain.[35] His work, "La Paura," describes the physiological effects of
emotions somewhat in detail.

The way toward applying the method of expression to the study
of emotions was shown by the results of previous physiological
investigations. Casual observations of the influence of certain
sense-stimuli on respiration and circulation were made by Naumann,
Couty and Charpentier, Thanhoffer, Dogiel, Gley, Mays, Istomanow and
Tarchanoff, Féré, Delabarre, and others.[36] The changes of breathing
seem to be of greater importance, and some writers account breathing
the most delicate physiological index of feelings.[37] It seems,
however, that a satisfactory treatment can be obtained only by direct
comparison of the respiration and circulation, and it now but seldom
occurs that circulation is observed exclusively.

There are three different instruments for observing the circulation:
the plethysmograph, the sphygmomanometer, and the sphygmograph. Each
of these instruments allows one to observe a different feature of the
circulation. The sphygmomanometer records the pressure in the artery;
the plethysmograph records the volume of a certain part of the body;
and the sphygmograph records the movement of a certain part of the
arterial wall. The curves traced with the sphygmograph indicate to a
certain extent the pressure of the blood, and sometimes they are called
curves of blood-pressure to distinguish them from the plethysmographic
curves which are called curves of pulse-volume.

The invention of these instruments is due to physiological
investigations of the pulse. The problem of studying the pulse
by graphic, or at least experimental methods, begins with the
investigations of Hales and Poiseuille. The first great success in
this line was the construction of the "Kymographion" of Ludwig, but
this instrument had the disadvantage that it could be applied only
by scission of an artery. This circumstance, of course, confined the
application of the instrument to the study of the pulse of animals.
After several attempts by Hérisson, Chelius, and others, Vierordt
succeeded in constructing his sphygmograph, by which curves of the
normal human pulse could be obtained. Some years afterwards Marey
constructed his much more sensitive instrument, which was made still
handier by the use of air transmission. Buisson was the first to use
air transmission for sphygmography, but Upham had used it before for
similar purposes. A considerable number of sphygmographs has been
constructed since, and though they may show some improvements in
detail, the technique of the sphygmograph has made no marked progress
since Marey, and his instrument has been found by experimental tests
remarkably exact.

The curves traced with the sphygmograph are extremely variable in shape
and size. In almost every normal curve, however, a steep ascent may be
seen; it is called the up-stroke or percussion stroke, and this part
of the sphygmographic curve has the name of the anacrotic phase. This
line of ascent ends abruptly and within the limits of the usual speed
of the recording drum it goes over into the descent by a sharp angle.
The descending part of the curve is called the catacrotic phase. The
descent is not so abrupt and is not a more or less straight line, but
is interrupted by secondary elevations. The first secondary elevation
is the largest and is called the dicrotic.[38]

These secondary elevations were seen first by Chelius and Vierordt,
and from the beginning they aroused considerable interest. It was
known that sometimes during fever the pulse takes an abnormal form,
where two beats of the pulse, a strong one and a weaker one, may be
felt for every heart-beat (pulsus bis feriens). This form of the pulse
was thought to be entirely abnormal and it was therefore a great
surprise for the first modern investigators to find these secondary
elevations in tracings of the normal pulse curve. The conviction of
the abnormality of the dicrotic pulse form was so firm that Vierordt
always applied his instrument in such a way that it did not trace the
dicrotic elevation, although it was sensitive enough to trace the exact
form of the pulse curve. Marey, however, used his much more delicate
instrument and found the dicrotic elevation in most of the normal
pulse curves.[39] For this reason Marey's sphygmograph met at first
with considerable criticism (Meissner), but the critical examinations
by v. Wittich, Buisson, and Mach showed that the dicrotic elevation
could not be due to an error of the instrument, for so great an error
was out of question, and there no longer remained a doubt as to the
genuine existence of the dicrotic elevation in the normal pulse curve.
The sphygmograph, thus, had revealed two new and surprising features of
the pulse; (1) The ascent and the descent do not take place with equal
rapidity, the ascent being steep, the descent gradual;

(2) the descent is interrupted by secondary elevations. Neither of
these facts could be observed by applying the finger and it seemed
important to explain them. The explanation of the dicrotic promised to
be of special interest, as it was shown that abnormal dicrotism is in
close relation to the normal form of the pulse curve.

This caused considerable interest in the observation of the pulse,
and the sphygmograph was supposed to be of the greatest importance
for medical diagnosis. Burdon Sanderson,[40] Landois, Lorain,[41]
Ozanam,[42] Pfungen,[43] Riegel,[44] Roy and Adami,[45] and others
have studied the sphygmographic curve under abnormal conditions, and
wellnigh all diseases have been studied by these observers with the
sphygmograph. The results were ambiguous and did not seem to justify
the amount of work spent on these observations. The enthusiasm for
the sphygmograph subsided, and it was no longer expected to obtain
a diagnosis, or even, indeed, a prognosis of a disease from mere
inspection of a pulse curve. Later investigators, in fact, confined
their research to the proof of the ambiguity of the sphygmograms,
which could be valuable only in connection with other observations.
It could not be hoped that an explanation of the abnormalities of the
pulse curve would be found before an understanding of the normal form
was attained. It, therefore, seemed necessary to decide between two
theories of the origin of the normal pulse curve, which had opposed
each other almost since the discovery of the existence of the dicrotic
elevation. Both theories chiefly refer to the origin of the dicrotic,
and they agree on this, that the dicrotic elevation is due to a wave
travelling in the blood, but they disagree on the direction in which
this wave is moving. These two theories may be called the theory of the
peripheral, and the theory of the central origin of the dicrotic wave.

The theory of the peripheral origin of the dicrotic wave assumes that
the change of pressure which is indicated by the dicrotic elevation
originates somewhere at the periphery and travels through the arteries
towards the heart.[46] Commonly it is assumed that the dicrotic
originates in the arterioles. This theory has been mentioned first,
because it is the simpler in every respect, though the less probable.
The origin of the dicrotic wave according to this theory is similar to
the origin of the echo.

Buisson was the first who gave an explanation of the dicrotic elevation
by assuming a central origin of this wave. His theory was adopted by
Marey, who stated it in this way. The action of the heart causes the
blood to be pumped into the aorta with considerable strength. The blood
leaves the aorta by its inertia and expands the arterial system. In
the arterioles it finds an obstacle and being reflected it flows back
to the aorta. But there it finds the semilunar valves closed and a new
wave is produced by reflection. This wave has an effect similar to the
first, and this reflection of waves lasts until the valves are thrown
open again. The existence of several secondary waves is explained by
the great velocity with which the blood travels through the arterial
system.[47]

This theory is open to many objections. First, there is no reason why
the blood wave should not produce a dicrotic elevation when it flows
back to the aorta. Second, the narrow lumen of the arterioles cannot
be an obstacle to the flowing blood, because if an artery splits up
into small branches, the sum of the lumina of the branches is greater
than the lumen of the artery. Lack of space, therefore, cannot be the
cause of the reflection of the pulse wave. Marey, finally, is mistaken
in his conception of the effect of the blood pumped into the aorta by
the action of the left ventricle. He supposes that the entering blood
pushes before it the whole column of blood in the arteries. This view
is refuted by the actual measurements of the velocity of the pulse
wave, because if it were true the pulse would appear at the same moment
in every part of the body.[48]

These are the more obvious of the arguments against Marey's theory.
Other investigators have tried to state a more correct theory of the
central origin of the dicrotic wave. Landois's theory belongs to this
type of improved theories of the central origin. The action of the
left ventricle, according to Landois, causes the primary pulse wave
which travels down the arterial system, until it is extinguished in
the arterioles. The walls of the arteries are expanded by the arriving
blood wave, and, when the valves close, they force the blood onward by
their elasticity. There is a free way to the periphery, but the blood
pushed towards the heart finds the semilunar valves closed and is
reflected. In this way a new positive wave originates which may produce
in the same way a secondary or tertiary wave.[49]

It seemed necessary first to decide between the theories of the central
and of the peripheral origin of the dicrotic wave. Many investigations
have been carried on for this purpose, and some of them bear witness
to the high ability of the investigators. It is, however, remarkable
that the arguments which have been brought forward in favor of one
hypothesis chiefly consist in reasons why the other hypothesis should
not be accepted. These experiments can be divided into two classes.
The first class comprises all the experiments which study the relation
of the pulse curve to other functions, or its dependence on various
conditions. The above mentioned observations of the pathological
changes of the pulse curve belong to this class. The object of frequent
studies of this type has been the relation of the sphygmographic
curve to the curve of the apex beat. The papers of Otto and Haas,[50]
Garrod,[51] Traube,[52] Rosenstein,[53]

Maurer,[54] Gibson,[55] François Frank,[56] and Edgren[57] deal
with this problem. The curve of intraventricular pressure cannot be
studied in man for obvious reasons, and only in some cases has an
attempt been made to compare the sphygmographic curve with the curve
of intraventricular pressure obtained from animals. One of the most
interesting attempts in this line will be mentioned later.

To the second class belong all those investigations, by which
experimental evidence in favor of one or the other hypothesis has been
collected. The experiments which belong to this class are in so far
more decisive as the conditions of the experiments are better known
and, therefore, easier to interpret. Von Kries proved the existence of
the dicrotic in the femoral artery of an animal after having replaced
the heart by a bag filled with liquid.[58] Grashey[59] and Hoorweg[60]
have demonstrated the existence of secondary waves in models, on which
peripheral reflection was impossible. To the same type of experiments
belong Marey's[61] and Grashey's registration of the waves in elastic
tubes, and Mach's[62] tracings from a mechanical model on which the
resulting movement of two simple components could be registered.
Without giving any physiological theory Mach showed how curves similar
to the pulse curves can be obtained by the registration of a movement,
the mechanical conditions of which are known.

As the results of these investigations, we may state the following
facts as arguments against any hypothesis of the peripheral origin of
the dicrotic elevation.

(1) Automatic registration of the pulse wave shows that the dicrotic
appears sooner in the regions nearer to the heart than in regions
which are more distant. The opposite would be the case if the dicrotic
elevation were due to a wave travelling from the periphery to the heart.

(2) The dicrotic appears at the same time after the primary wave in a
dwarf as in a tall man. This would be impossible if the wave had to
travel so much farther.

(3) Inhalation of amyl nitrite makes the dicrotic almost disappear. The
adherents of the theory of the peripheral origin of the dicrotic wave
explain this fact by supposing that this drug dilates the arterioles
and makes little reflection possible. Their opponents say that the
action of the heart and the resistance of the system are so enfeebled
that the backward flow is slight and gives rise only to a small wave.

(4) If an artery is opened and the blood allowed to spurt on a
revolving drum of white paper a curve is obtained which shows the
dicrotic elevation (the hemautographic curve of Landois). The
resistance of the periphery is totally lacking in this case and the
dicrotic elevation could not appear if it were due to a wave reflected
at the periphery.

(5) The appearance of the dicrotic is not retarded if an elastic tube
is placed between the periphery and the place where the instrument is
adjusted. If the dicrotic were due to a wave reflected at the periphery
it would be retarded because the wave would have to travel a distance
so much greater.

These arguments prove the impossibility of the theory of the peripheral
origin of the dicrotic wave. Also the other hypothesis meets with a
number of serious difficulties, and we mention the following facts
which are arguments not against any special form of this theory, but
against any hypothesis which starts from the assumption that the
dicrotic elevation is due to a wave travelling from the heart to the
periphery.

(1) The descent of the catacrotic phase ought to be a succession of
diminishing waves, but not a slow descent with merely small elevations.

(2) This hypothesis accounts for none of the abnormal pulse forms.

(3) The blood ought to push against the semilunar valves with a
force not less than 1/2 - 2/3 of the force of the contraction of the
ventricle, because this is about the relative height of the first
secondary elevation with regard to the primary wave, which is due to
the contraction of the ventricle.

(4) It does not account for the disappearance of the dicrotic
elevation through lack of elasticity of the arterial wall: for the
dicrotic elevation is most marked in youth, becomes lower in old age,
and disappears in diseases like atheroma and arteriosclerosis, which
impair the elasticity of the arterial wall. Landois's theory overcomes
this theory only apparently, although the dicrotic would be absent, yet
in that case the descent of the primary wave ought to be as steep as
its ascent.

(5) This theory is refuted by the experiment of v. Kries, who proved
the existence of the dicrotic if the heart is replaced by a valveless
bag.

The obvious impossibility of making the theories agree with the facts
does not permit one to accept any of them. All of them are based on the
supposition that the dicrotic elevation is due to a wave travelling in
the blood, and this belief is founded on the following argument: If a
wave travels in the blood the sphygmographic curve shows an elevation;
the dicrotic elevation is an elevation in the sphygmographic curve.
Therefore, the dicrotic elevation is due to a wave travelling in the
blood. This fallacy is responsible for the astonishing fact that the
refutation of one of two apparently contradictory statements does not
prove the other. It is characteristic of the present state of the
problem concerning the origin of the dicrotic elevation, that a modern
writer[63] calls it "inextricably complicated."

The contradiction between the theories of the peripheral and of the
central origin of the dicrotic, however, is only apparent, and neither
may be true, because it might be that this elevation is not due to
a wave which travels in the blood. The experiments of the previous
investigators seem to point in this direction. The disappearance of the
secondary elevations when the arterial wall has lost the properties
of an elastic body, the above-mentioned experiments of v. Kries, and
the observations of Grashey and Marey on the movements of the walls
of an elastic tube indicate clearly that nothing but elasticity is
needed to produce these secondary or dicrotic elevations, for, in the
different experiments, they are produced as well when the heart and
its valves are replaced by a valveless bag as when the function of
the valves is unimpaired; as well with resistance at the periphery as
without, the only condition being that the walls are elastic. This
proves the importance of the elasticity of the arterial wall. The
experiments of the graphic registration of the movements of the walls
of an elastic tube, furthermore, indicate that the conditions of this
experiment are a close imitation of the mechanical conditions which
prevail in the arteries. It may be expected that the analysis of the
conditions of the experiment will give an insight into the origin of
the sphygmographic curves, because the tracings which Grashey and Marey
took from the walls of a rubber tube resemble closely the tracings of
the human pulse. This experiment, first, proves that the form of the
curve depends merely on physical conditions. The movement of a point
of the wall of the tube depends on the following four factors: (1) The
elasticity of the wall; (2) the incompressibility of the liquid; (3)
the form of the original wave, _i. e._, the way in which the liquid
is pumped into the tube; (4) the rate of outflow. If the process of
pressing liquid into the tube is repeated regularly, a stationary form
of movement will be obtained eventually; the amount of outflow for one
interval is constant in this case. This means that eventually a state
is attained where the same quantity of liquid which is pumped into the
tube at one end flows out from the tube at the other. The physiological
bearing of this result is that the turgor of an artery does not change
without a cause. Such a change would be indicated by the going up or
down of the base-line of the tracing.

The first two factors are, in physiology, studied with relative
ease. The elastic qualities of the arteries have been studied since
Poiseuille and John Hunter by Wertheim, Zwardemaaker, Marey, and
others, and they are more or less well known. The physical properties
of the blood are very nearly those of an incompressible liquid, and
this is certainly true for the small pressure to which the blood is
exposed in the arteries.

As to the initial form of the wave which the action of the left
ventricle produces in the arterial system, we get a hint from the
experiments of Grashey, v. Kries, and Marey, where the sudden
compression of a bag furnished the initial shock.[64] These changes of
pressure can be represented by a curve like that in Fig. 1.

So long as the contraction of the left ventricle lasts and the valves
are open, the action of the heart produces a certain pressure in the
aorta, but the influence of the intraventricular pressure is zero when
the valves are closed. The second phase of the curve Fig. 1, where the
pressure is zero, certainly gives the influence of the intraventricular
pressure during the diastole, because there is no communication between
the ventricle and the arterial system when the valves are closed. The
question is whether the rest of the curve can represent the changes of
the intraventricular pressure when the valves are open.

[Illustration: Fig. 1. Changes of pressure produced in a bag by sudden
compression.]

[Illustration: Fig 2. Decreasing amount of liquid in a tube when the
outflow is uniform.]

The first curves of intraventricular pressure were traced by Chauveau
and Marey. These experiments were made on a horse, and they have been
repeated since it was discovered that they can be performed also on
smaller animals. Besides Chauveau and Marey may be mentioned the names
of Fick, Huerthle, v. Frey, Rolleston, Bayliss and Starling. The curves
obtained by various observers belong to two types; one shows the so
called "plateau," the other does not. Recent experiments have proved
that this difference of results is due to a difference in methods. This
also is suggested by the fact that different curves have been obtained
from animals of the same species. Two methods have been applied lately
for testing these curves of intraventricular pressure. The first
was devised by Bayliss and Starling. It consisted chiefly in the
photographic registration of the movement of the liquid in a manometer
tube. The photographic registration is frictionless, and the mass of
the moving liquid was so small that vibrations by inertia were fairly
excluded for pressures which are not greater than the intraventricular
pressure.[65] The second method was used by Porter. The idea of this
method was to trace only a part of the curve, not the whole. The
writing lever, thus, has in the beginning of the tracing no inertia at
all, and the tracing may be overdrawn but is certainly correct in form
up to the next point of inflexion of the curve.[66] These tests and
the repeated experiments of Chauveau leave no doubt as to the existence
of the plateau.

The varying pressure from the heart which produces the pulse wave may
be described in this way: The pressure suddenly rises to a maximum
and maintains it for a certain time; when the semilunar valves close,
the pressure drops as suddenly as it rose, and remains at zero until
the valves open again. Such a function can be represented by a curve
like Fig. 1, and this is the reason why the complicated action of the
heart can be superseded by the compression of a bag without changing
the mechanical conditions of the problem. Of course it can not be
expected that a schematic curve will show all the details of the real
tracing. It is suggested, however, by Frank[67] that many of the small
irregularities of the curve of intraventricular pressure are due to
vibrations caused by the inertia of the apparatus and that the true
form of the curve of intraventricular pressure is very simple. This
remark is supported by Huerthle,[68] who tested the apparatus of Marey,
Knoll and Grunmach. Marey's tambour was found to be the most exact,
but even this instrument produces deformities in the tracings, though
the general outlines are exact. This would indicate that the schematic
representation of Fig. 1 is a very close imitation of the real form of
the curve of intraventricular pressure, although empirical tracings do
not show right angles and straight lines. It seems, however, that the
undulations of the plateau are genuine, since they are found in the
most reliable tracings, and it may be possible to explain them merely
on the basis of the physical conditions of the experiment.

The fourth factor of importance is the rate of outflow. We may
introduce the following assumption as to the rate of outflow of the
blood through the capillaries: The outflow through the capillaries
is uniform in the short time of one heart-beat. The fact has been
mentioned above that the quantity of outflowing blood must be equal
to the quantity of incoming, for any stationary form of the pulse
movement; this new hypothesis means that the velocity of the outflow
is constant. One might think that this assumption is warranted by the
law of Poiseuille that the amount of outflow through a horizontal
capillary filled with liquid under constant pressure depends on
the fourth power of the radius and on the difference of pressure
at the two ends of the tube, and is inversely proportional to the
constant of friction and to the length of the tube. This law has
been proved mathematically and tested physically only for horizontal
tubes and constant pressure. Neither of these suppositions holds for
the capillaries of the arterial system. The connection between the
hypothesis in question and Poiseuille's law is this. Let us suppose
that an artery splits up in a great number of arterioles which go
off in every direction. The amount of outflow is then a complicated
function, because the law of Poiseuille does not hold for every
direction of the capillaries; but it will be equal to the outflow
through a tube of certain radius and certain direction in the same
time. Our assumption says that the law of Poiseuille holds for this
typical but imaginary tube. The essential point of this hypothesis
is merely the supposition that the outflow of blood through the
capillaries follows α law.[69]

It is possible to show that the graphic registration of a movement
under these four conditions must give curves which correspond to
the pulse curves in every respect. The action of the left ventricle
causes the pulse wave which travels through the arterial system with
considerable velocity. This wave expands the arteries and the whole
system is filled with blood because the wave arrives by its great
velocity at the periphery before the contraction of the ventricle
is finished. The increased pressure forces the blood to enter the
arterioles, through which it passes at a constant rate. When the valves
are closed, the amount of blood decreases uniformly and the volume of
the blood contained in an artery can be represented graphically by a
straight line of more or less steep descent, as is shown in Fig. 2.
Now the walls of an artery have to a high degree the qualities of an
elastic body, and, therefore, they are forced back by elasticity after
being displaced from the position of equilibrium by the shock of the
arriving pulse wave. The movement of a point of the arterial wall,
therefore, results from two components: (1) From the movement which it
would perform if it were merely forced to remain on the surface of the
blood in the artery, and (2) from the movement due to the elasticity of
the arterial wall. Both movements have the same direction, because the
column of blood is enclosed in a cylinder the radius of which decreases
regularly, and the elastic force of the arterial wall is directed
towards the centre. The direction of both forces is in the line of the
radius, and the resulting movement of these two components, therefore,
can be found by simple superposition. Of the first component we know
that it can be represented graphically by a straight line.

An elastic force tends always to bring the body back to the position
of equilibrium; if the distance is not too great, the force is
proportional to the elongation. A physical body is always under the
influence of friction, the acceleration of which is opposite to the
direction of the movement, and therefore diminishes the velocity. The
form of the resulting movement depends on the amount of friction,
and, roughly speaking, we may distinguish two types of elastic
movements:[70] the first type is a periodic movement, the second an
aperiodic. Let us suppose that a body is carried from its position of
equilibrium by a sudden impulse, which transmits a certain velocity to
the body. Friction and elasticity diminish this velocity, and after a
certain time the body attains a maximum elongation, where the velocity
is zero. Then the body returns under the influence of elasticity and
under the retardation of friction. There are two cases possible, either
the elastic force is strong enough to overcome friction and to carry
the body over the position of equilibrium, or it is not strong enough.
In the first case, it is easy to see, the body repeats the same form
of movement on the other side of the position of equilibrium, and
the conditions being constant a vibratory movement results as the
stationary form. In the second case the body approaches the position
of equilibrium asymptotically. The first case may be illustrated by
the vibrations of a magnet needle suspended with little friction, the
second by the movement of a door which is regulated by a well-working
shutter.

These forms of the movement of a body under the influence of elasticity
and friction are illustrated in Fig. 3.

Curve 1 shows a movement where friction is so small that it can be
neglected; it is, of course, a simple sine curve. Curve 2 shows the
effect of friction on vibrations. The period of damped vibrations is
greater than in the frictionless movement, but the amplitudes are
smaller. The amplitudes of a damped vibration decrease constantly and
there is a simple relation between two subsequent amplitudes. The ratio
between them is constant, and, therefore, if one amplitude and this
constant ratio are known, all the other amplitudes can be calculated.
The amplitudes of such a movement decrease as the terms of a geometric
series. The dotted line in Fig. 4 represents the rapidity of this
decrease. It is obvious that the smaller the constant ratio of two
subsequent terms is, the more rapidly will the amplitudes decrease.
This ratio depends on friction, and becomes smaller when friction
becomes greater. A vibration under heavy friction dies out quickly.
Curve 3 shows a movement where friction is too great to allow any
vibrations. The body does not acquire a velocity which can carry it
over the position of equilibrium, but it approaches this position with
ever diminishing velocity.

[Illustration: Figs. 3 and 4]

These are the types of movement which the arterial wall can perform by
its elasticity in consequence of the shock of the arriving pulse wave.
The mechanical nature of the components on which depends the form of
the sphygmographic curve is, therefore, known. The constructions in
Fig. 5 show how the resulting movement can be found.

[Illustration: Fig. 5]

These curves are constructed in this way. The lines _AB_ represent
the time of the interval of one heart-beat. The straight line _EB_
represents the decreasing volume of the artery and the curves on
_AB_ represent the elastic movement of the arterial wall. Both are
synchronous movements, and a line perpendicular to _AB_ gives the
corresponding points. The points of the resulting movement are found
by arithmetical addition of the two ordinates. The results of these
constructions prove that the curves show the dicrotic elevation only
if the elastic force is great enough to make a vibratory movement
possible. Aperiodic movements do not produce this elevation. The
friction is always great for the movement of the walls of an artery,
and there are only the two possibilities, of a vibratory movement which
dies out quickly, and of an aperiodic movement. This accounts for the
fact that the dicrotic elevation may be missing sometimes, and that in
other cases several secondary elevations may be seen, the number of
which, however, is always limited, and their relative height rapidly
diminishing. It may be remarked that the length of the lines _AB_ seems
essential to the form of the resulting curve. Curves I and III differ
very much in the length of the lines _AB_, while the lines _AE_ are
equal and the vibratory movements are only slightly different. The
resultants, nevertheless, seem to differ very much. It is easy to see
that a different speed of the recording drum will have an effect on
the tracings which is similar to that of a change in the length of
the lines _AB_ in the constructions. This is one more reason why mere
inspection of the curves cannot give a satisfactory result.

These constructions show that the sphygmographic curves must show
great variations, since the amount of blood pumped into the system,
the elasticity of the arteries and friction of the surrounding tissues
are subjected very likely not only to individual but also to local and
temporal variations. But under given conditions only a certain form of
the pulse wave is possible, and this form does not change so long as
these conditions do not change. The sphygmograms in Fig. 6 show some of
the typical forms of the pulse curve.

[Illustration: Fig. 6]

No. I shows the influence of high arterial tension, and No. II of low
tension. The first corresponds to No. II in Fig. 5, the second to Nos.
I and III. Nos. IV and V of Fig. 5 show the effect of great friction
and small elasticity. The constructions differ in the form of the
elastic movement; the position of equilibrium is reached with different
velocity in both cases. The resulting movements differ slightly in the
form of the catacrotic phase. Both forms may be seen in No. III of Fig.
6. This sphygmogram was taken from an artery with low tension, and this
form of the sphygmographic curve is well known as characteristic of the
"soft" pulse. If the artery has lost to a large extent the qualities
of an elastic body, and if the outflow is very rapid, the pulse curve
shows nothing but the slight elevation of the travelling wave; No. IV
in Fig. 6 shows a curve of this character.

This theory explains many surprising facts which resisted every attempt
at explanation. The anacrotic part shows a steep ascent, because it
is due to the sudden arrival of the blood wave. It seems that an
interruption in the descent may be seen only in abnormal cases. The
sphygmograms of twelve normal individuals were observed regularly by me
during more than a year without once discovering an anacrotic elevation.

The hemautographic curve of Landois is produced in this way. The form
of this curve depends on the velocity of the escaping jet of blood. The
velocity of the blood flow depends on the resistance of the arterial
system in the sense that the velocity decreases when the resistance
increases. When the arterial wall is in the negative phase of vibration
the lumen of the artery is smaller, and, therefore, the velocity
smaller. This is confirmed by the actual tracings of the velocity of
the circulation by Marey.

It is also obvious that the dicrotic elevation never can arrive before
the primary wave, because the arterial wall cannot perform elastic
vibrations before it is expanded by the impulse of the arriving blood
wave. Neither is it surprising that the "dicrotic wave" seems to travel
in the same direction and with a velocity equal or almost equal to the
velocity of the pulse wave. Such a difference can be produced only by a
difference in the time of the vibrations of the arteries at different
points of the body. The time of one vibration is necessarily very
short, and the length of this interval depends on the circumstances
which determine the elasticity of the arterial wall and the friction.
These conditions may be subjected to local variations. If, therefore,
the time-interval between the primary and the secondary elevation
is measured at two different points (_e. g._, at the carotid and at
the radialis) a difference of time may be found. Starting from the
supposition that the dicrotic elevation is due to a wave travelling in
the blood, one could attribute this difference of time to a velocity of
the "dicrotic wave" which is slightly different from the velocity of
the primary wave. The fact that the dicrotic elevation appears later in
places farther from the heart was interpreted as a proof that the wave
travelled out from the heart. No theory which assumes that the dicrotic
elevation is due to a wave travelling in the blood can give a reason
why two waves of the same form and origin should travel through the
same liquid at different velocities.

At this point a theory must be mentioned, which was brought forward
recently, because it is based on measurements of the velocity of
propagation of the dicrotic wave. This theory is connected with Krehl's
theory of the function of the valves. The blood, according to Krehl,
enters the aorta through a small opening, and expanding in a large
space it produces fluctuations and eddies, which would close the valves
if they were not kept open by the blood which streams through under
high pressure. They must, therefore, close at the moment when the
aortic pressure is equal to the intraventricular pressure. This occurs
shortly after the moment indicated by the beginning of the decline of
the intraventricular pressure curve. Now the second sound of the heart
is heard somewhere in the descending part of the cardiogram[71] and
the measurements of Huerthle[72] have shown that the second sound is
heard 0.02" after the beginning of the descent of the cardiogram. This
seems to indicate that the second sound of the heart is in a temporal
relation to the closure of the valves. Many theories of the origin of
the sounds of the heart agree on this one point that the second sound
is due to a noise in the muscles. It therefore may be supposed that
the second sound is due to the tension of the valves when they close
or shortly afterwards. The problem now would seem to be to find an
elevation in the descending branch of the curve of intraventricular
pressure, or in the tracings of the apex beat, which could be
attributed to the closure of the valves. It was taken for granted that
the curves of intraventricular pressure and those of the apex beat
were identical. In many of these tracings an elevation was found which
may be called "the wave _f_." This elevation is not found in all the
tracings, and its position seems to be rather variable. Edgren[73]
remarks that the wave _f_ was always found near the abscissa no matter
whether the preceding decline of the curve was great or small. In some
of Chauveau's tracings the wave _f_ is missing or indistinct,[74] in
others it is very well marked and approximately in the middle of the
descending branch of the curve.[75]

Edgren made experiments on the temporal relation of the wave _f_ and
of the dicrotic wave, which to avoid misunderstandings he calls the
"wave _f_´." His experiments were made as follows. A sphygmogram from
the carotid and a cardiogram were taken simultaneously, the points
of the writing-levers being in the same vertical line. The wave _f_´
appeared a little after the wave _f_. The length of this interval
could be calculated by measuring the distance between these waves, as
the speed of the drum was known. From this was subtracted the time
of propagation of the dicrotic from the heart to the point where the
instrument was fixed. In this way it was found that the time between
the appearance of the wave _f_ and of the wave _f_´ was equal to
the time of propagation of the dicrotic wave from the heart. Edgren
concluded that the dicrotic wave is in close temporal relation to the
closure of the valves.[76] To this comes the supposition that the wave
_f_´ is due to a change of pressure proceeding from the heart. The wave
_f_´, therefore, could be attributed to the tension of the valves.[77]
Edgren and Tigerstedt are the chief exponents of this theory.

In so far as this theory assumes that the dicrotic elevation is due
to a wave travelling from the heart to the periphery,[78] it is
open to all the arguments against a theory of the central origin
of the dicrotic wave. Against the more special assertion that the
dicrotic elevation is in connection with the closure of the valves,
the following facts must be mentioned. We grant that the tracings
of the apex beat may be directly substituted for the curves of
intraventricular pressure, although this is by no means obvious, since
one tracing gives the form of the pressure changes and the other the
effect of the shock of the heart against the wall of the chest. It is,
furthermore, not proved that the wave _f_ is due to the closure of the
valves and that the waves _f_ and _f_´ correspond to each other so
closely as Edgren's experiments seem to indicate. His measurements of
the length of lines were made with an exactitude of 0.1 mm., but his
computations were carried to the third decimal place of a second. The
third decimal is generally inexact and the second in a large number of
cases. Experimental evidence, furthermore, directly contradicts the
statement that the dicrotic elevation corresponds to the wave _f_.
Fredericq[79] traced pressure curves in the ventricle and in the aorta,
and determined the points of equal pressure in both curves. He thus
found that a point near the beginning of the descent of the curve of
intraventricular pressure corresponds to the dicrotic. His experiments
are rather conclusive against the theory in question, since the wave
_f_ is very well marked in these tracings of Fredericq. The following
facts, however, are fatal for the theory that the closure of the valves
causes the dicrotic elevation: The dicrotic wave disappears in diseases
like atheroma and arteriosclerosis which do not impair the function
of the valves, but affect the elasticity of the arterial wall, and it
is not affected by valvular insufficiency. The independence of the
dicrotic from the function of the valves is conclusively proved by v.
Kries, who found the dicrotic elevation in the femoral artery of an
animal whose heart was replaced by a valveless bag.

All these facts, on the contrary, can be understood easily in the light
of the theory that the sphygmographic curve gives the movements of the
arterial wall, which movement is conditioned by the decreasing amount
of blood in the artery, and the elastic vibrations of the wall around
a variable position of equilibrium. In some cases the conditions of
the problem are rather simple, and admit an analytic treatment, the
results of which fit closely to the experimental facts. This part of
the theory, however, has merely physiological interest, and therefore
is discussed in a separate paper. It may be mentioned at this point
that this theory of normal dicrotism is essentially identical with the
theory of abnormal dicrotism as stated by Galen. He believed that the
second beat of the pulsus bis feriens was due to an elastic vibration
of the arterial wall. "Ex eodem genere sunt dicroti; nam arteria in
occursu quasi repellitur, moxque redit.... Neque enim tum arteria
contrahitur, sed quasi concuteretur, occidit; cuius delapsum a primae
distentionis termino nulla dirimit manifesta quies, ut animadvertitur
in contractione: sed simulatque attolli destitit, recidit atque ita
paulisper vibrata, mox occurrit iterum."[80] Galen, however, is
mistaken in his view, and in his observation that sometimes three or
more pulse-beats may be felt with the finger. No form of the pulse is
known where three or more beats may be felt for every heart-beat, and
the actual tracings exclude the possibility of this observation for
the pulsus bis feriens. The pulsus bis feriens is due to an increase of
the frequency of the heart-beats. If the new pulse wave arrives before
the vibrations of the arterial wall have had time to subside, the new
wave and the already existing vibration may interfere in such a way as
to produce this abnormal pulse form.

The form of a single wave of the sphygmographic curve may be influenced
by changes in the following conditions:

(1) The pulse wave may have an initial form which cannot be represented
by the schematic curve in Fig. 1. This may be due to an irregularity of
the function of the ventricle. The action of the heart has an influence
on the length of the waves, which length is determined by the rapidity
of the heart-beats. This influence has been mentioned before. A change
in the rapidity of the heart-beats has no great influence on the form
of the catacrotic part of the curve so long as the impact of the new
pulse wave does not arrive before the vibrations of the arterial wall
have had time to subside.

(2) Differences of the elasticity of the arterial wall affect
materially the form of the catacrotic part of the sphygmographic curve.
It has also some influence on the height of the curves, because the
amplitude of elastic vibrations depends on the elastic force for a
given force of the shock. The degree of elasticity of the arterial
wall is subjected to individual variations, and it depends in a given
subject on the state of innervation of the wall.

(3) The surrounding tissues have a certain influence, since their
resistance determines the friction opposing the vibration. This
accounts for the fact that merely local conditions, such as a change
of the position of the arm or the adjustment of the instrument, may
change the form of the pulse curve. For instance, if the sphygmogram is
taken from the a. radialis the instrument is placed between the styloid
process of the radius and the tendon of the flexor carpi radialis. In
the neighborhood of this place are two venae comites and a superficial
branch of the median or radial vein. A change in the position of the
arm will have a certain influence on the circulation in the veins, and
influence the turgor of these vessels. Increased turgor increases the
friction, and thus produces the different forms of the tracings.

(4) The changes of the turgor of the artery, moreover, cause a general
rise or lowering of the curve. This symptom is essentially ambiguous
for the turgor of the artery may be changed as well by an increase or
decrease of the amount of blood pumped into the arterial system as by
a decrease or increase of the amount of blood which passes through the
capillaries.

The influences mentioned under (1) may be seen in tracings taken from
cases of cardiac insufficiency, and have merely pathological interest.
All the other influences, however, can be observed in the curves which
are traced for psychological purposes. The changes in the general rise
or fall of the curves are not so very hard to observe,[81] and for the
observation of the rapidity of the heart-beats it is only necessary to
trace a time-curve and count the number of beats or measure the length
of every single beat. Also the changes of the height of the waves can
easily be measured. This has been done conscientiously by several
observers. It is by far harder to see the changes in the form of the
catacrotic branch, and only a few keen observers have seen them. These
changes of the pulse curve under the influence of feelings were proved
as facts by experiments, but their interpretation was doubtful. With
the exception of the rapidity of the heart-beats, which could easily
be observed in some other way, all the symptoms of the influence of
feelings on circulation are ambiguous. A difference in height of the
single waves may be due to a change in the amount of blood which is
pumped into the artery, but it also may be due to a change in the
amplitude of the vibrations of the artery. The form of the catacrotic
part of the sphygmographic curve may be changed by a different state of
innervation of the arterial wall, but it also may be due to an increase
or decrease of the friction of the surrounding tissues. The general
rise or fall of the curve may indicate a change in the amount of blood
which leaves the left ventricle, but it also may indicate a change in
the amount of the capillary outflow.

The problem, nevertheless, is fully determined, and a solution is
suggested by the constructions in Fig. 5. The form of the resulting
movement depends, first, on the length of the line _AB_, secondly, on
the length of the line _AE_, and thirdly, on the nature of the elastic
movement. An elastic movement is determined if three constants are
known, one of which is the amplitude, the second the friction, and the
third the elasticity. Only _AB_ can be measured directly, and there
remain four unknown quantities to be determined. Four measurements
must be sufficient for this purpose. It is obvious, however, that not
any four measurements will do, but a method can be devised by which
it is possible to determine each one of these four quantities. The
problem can be solved in every case provided that the sphygmogram is
trustworthy enough to justify the work. The length _AB_ is proportional
to the time of one heart-beat, and the length of the line _AE_ is
proportional to the amount of blood pumped into the arteries. The
successful analysis of the pulse curves, therefore, shows changes of
the action of the heart and makes it possible to distinguish them from
the changes at the periphery.

Besides the length of the heart-beats there are invariably these
four quantities which must be determined by the analysis of the
pulse curves: Amount of incoming blood, amount of outflowing blood,
elasticity of the artery, and friction of the tissues. These quantities
depend on the action of the heart, the peripheral resistance, and the
state of innervation of the artery. It is not possible to discuss here
the bearing of this theory and of the facts which may be connected with
it, on the different views of the localization and operation of the
centres which control these functions. Anatomical and physiological
evidence, however, leaves no doubt that the function of the heart and
the innervation of the arteries and capillaries _are_ under the control
of nervous centres. It may be supposed, therefore, that changes of
the pulse curve like those due to the influence of feelings are the
effect of the function of these centres. It is to be expected that the
detailed analysis of the pulse curves may give some indications as to
the nature of this influence, for it may be observed how the function
of these centres changes under the influence of mental processes.

A complete analysis of the physiological accompaniments of a
feeling process must give a description of the changes in the
function of the heart and the system, besides a description or at
least enumeration of the other changes which can be observed. By
a number of such investigations material for a general theory of
physiological accompaniments of feelings may be obtained, which
would not be void of interest for the psychology of feelings. Such
a theory must contain the answers to the following questions: (1)
How do the physiological reactions depend on the sense-stimulus? (2)
How many possible circulatory reactions are there? (3) What is the
location and interdependence of the respective physiological centres?
The first question cannot so far be answered in general, but it
will be possible to give a general answer when a greater number of
systematic investigations on the effect of sense-stimuli have been
carried on. Papers like those of Mentz may settle the question for
certain sense-stimuli. From the results which have been obtained so
far it comes out clearly that the reaction does not depend merely
on the nature of the stimulus, but that it depends largely on the
psychical and physiological state of the subject. The answer to the
second question may be given readily, but it seems advisable to give
it in connection with an experimental investigation. It may be said,
nevertheless, that the number of typical reactions is rather limited.
The third problem, by its nature, cannot be definitely answered before
the location of the respective centres is ascertained and their
interdependence explained.

It is, finally, a merit of this theory of the pulse curves that it
shows how the form of this curve may depend on central processes. The
problem of the mysterious influence of mental processes is thus reduced
to the analysis of merely physiological conditions. The theories on the
nature of this influence are so numerous that they may well be called
innumerable, and they vary from accepting a direct influence of ideas
on the circulation to considering the body as a sounding-board which by
every sensation is shaken in all its parts. Each one of these theories
is also a theory of feelings, and a more or less exact description of
these changes has been often taken for a descriptive psychology of
feelings. The example of the sounding-board is taken from one of those
papers which expound the theory that bodily changes follow directly
on the perception, and that our sensation of these facts is the
emotion. Every one of these bodily changes, whatsoever, is perceived,
acutely or obscurely, the moment it occurs. This theory is defended
by the argument that if we try to abstract from consciousness all the
sensations of our bodily symptoms, we find we have nothing left behind.
This argument, which may be found in almost every paper that deals
with this theory, is remarkable, because it sometimes is referred to
processes of every description, and thus comes into contradiction with
psychophysical parallelism which excludes the acceptance of psychical
states which have no physical correlate. This theory, as will have been
noticed, is the theory of feelings expounded by James, Lange, Ribot,
and others. It is widely accepted, and may be found also in books of
popular or semi-popular nature. Two observations must be made against
this view:

First, a perception of a bodily change which is felt in the moment
the change occurs exists only in the theory, every real process
needing a certain time. This point of the theory may be improved by
admitting that the afferent process lasts as long as any other of the
physiological processes of this kind. Either assumption, however, is
contradicted by the experimental evidence supplied by Lehmann that the
physiological changes occur after the beginning of an emotional state.

Secondly, if the theory refers only to those bodily changes which we
know, it certainly is not true, for emotional states are sometimes
observed without it being possible to find with modern instruments any
bodily accompaniments. If the theory refers to bodily changes of every
description, it is certainly true, or, better, it is beyond all attack
because it becomes identical with psychophysical parallelism. In this
general form this theory of feelings is as good as no theory at all,
because it refers to mental states of every description.[82]

This conception of emotional states of mind as perceptions of bodily
sensations would hardly have been promulgated, if the authors had tried
to base it on experiments performed in the laboratory. An emotion
but not the feeling-tone of a simple sensation may be mistaken for
the sum of bodily sensations. It is, furthermore, remarkable that
the promoters of this theory do not make a clear distinction between
sensation and feeling. They introduce an emotional element by calling
the perception of bodily changes a feeling of these changes. Only in
this way do they succeed in building up emotional states of mind out
of elements which are seemingly sensational. This does not succeed if
the word feeling is replaced by the word sensation. The failure of this
theory is due to two facts, first to the starting from a philosophical
doctrine, and second to the lack of a precise distinction between
feeling and sensation. It cannot be doubted after the above discussion
how a definition of this difference may be given which holds for every
empirical investigation.

A sense-stimulus produces a complex of nervous and central processes.
Among these is a certain group of processes which manifest themselves
by changing the innervation of the heart, the blood-vessels, the
lungs, and certain muscles. Another group is formed by those nervous
and central processes which are more or less immediate effects of
the sense-stimulation. The first group of processes is referred
subjectively to an emotional state of mind, and the second to a
cognitive process; the first group of processes is the physiological
accompaniment of feelings, the second that of sensations. The relative
independence of the first group from the second group is warranted
by the fact that the same processes are observed as accompaniments
of ideational processes. A strict limit between these two groups of
processes can be drawn when the central processes are better known,
because to the first group belong all those processes which are found
to be accompaniments as well of sensational as of ideational processes.
In different sensations the emotional process may be more or less
marked, and in others the cognitive process may be prominent, but it
seems that feelings are an invariable accompaniment of the sensation.
This suggests the definition of feelings as psychic processes, the
physiological accompaniment of which are central processes which depend
largely on the state of the organism, and which manifest themselves by
changes in the innervation of the heart, the blood-vessels, the lungs,
and muscles. The impossibility of directly comparing the sensations
of different subjects is recognized, and it is also impossible to
compare feelings, because in either case we are dealing with psychic
processes.

FOOTNOTES:

[Footnote 29: Comptes Rendus, vol. 50, p. 637, 1860.]

[Footnote 30: Comptes Rendus, vol. 53, p. 98, 1861. "Sans rien livrer à
l'hypothèse il est bien certain que des changements dans la circulation
périphérique arrivent souvent sous l'influence d'émotions morales....
En résumé d'après ce qui précède il nous semblerait illogique de faire
une exception pour les actions que les causes morales exercent sur les
battements du cœur, et nous pensons qu'elles doivent agir comme toutes
les autres influences, c'est à dire à la périphérie primitivement."]

[Footnote 31: We quote as an instance Lotze: Medicinische Psychologie,
p. 257, 1852: "Es ist wahr, dass Gefühle sehr lebhafte motorische
Ruckwirkungen äussern; wir sehen die Respiration in Unordnung gerathen,
den Druck der Arterienwandung auf das Blut bei heftigen Schmerzen
zunehmen, Erbrechen auf widrige Geschmackseindrücke, allgemeine
Muskelkrämpfe bei physischen Martern eintreten." All the principal
physiological features of feelings are enumerated here, but one hardly
will give any great credit for priority to Lotze. Such a general
statement, in fact, belongs to the class of easy observations from
which philosophical speculation often starts. G. L. Duprat shows
that the same "observations" underlie the theories of Aristotle,
Hippocrates, and Plato. (Duprat: La psycho-physiologie des passions
dans la philosophie ancienne, Archiv f. Geschichte d. Philosophie, vol.
18 (N. F. 11), pp. 395-412, 1905.)]

[Footnote 32: Application de la balance à l'étude de la circulation du
sang chez l'homme, Archives Italiennes de Biologie, pp. 130-143, 1884.]

[Footnote 33: Ueber den Kreislauf des Blutes im menschlichen Gehirn,
1880.]

[Footnote 34: Mosso et Pellacani: Sur les fonctions de la vessie,
Archives Italiennes de Biologie, vol. 1, pp. 97-127, 1882.]

[Footnote 35: La température du cerveau, Arch. Ital. de Biol., vol. 22,
pp. 264-311.]

[Footnote 36: For the literature, see P. Menz: Die Wirkung akustischer
Sinnesreize auf Puls und Athmung, Phil. Stud., vol. 11, p. 61, 1895.]

[Footnote 37: Meumann und Zoneff: Ueber Begleiterscheinungen
psychischer Vorgänge in Athem und Puls, Phil. Stud., vol. 18, p. 3,
1901; and Wundt: Physiologische Psychologie (5th ed.); vol. 2, p. 298.]

[Footnote 38: It has been pointed out that this terminology, which
is due to a large extent to Landois, presupposes a certain theory of
the origin of the secondary elevations. (Edgren: Cardiographische
und sphygmographische Untersuchungen, Skandinavisches Archiv. f.
Physiologie, vol. 1, p. 92, 1889.) It is not easy to change a
terminology, and the Greek terms, some of which were used by and
before Galen, are so indifferently connotative that they can be kept
without inconvenience. If one were to be rigorous, one would change
the name of the sphygmograph into palmograph, because this instrument
does not serve exclusively for the registration of the abnormal
pulse. It does seem, however, advisable to drop the terms recoil wave
(Rückstosselevation, Landois), and "onde de rebondissement" (Marey),
because they are taken from modern languages and directly suggest a
certain theory with which they are intimately connected.]

[Footnote 39: Recherches sur l'état de la circulation d'après les
caractères du pouls, Journal de Physiologie de l'Homme, vol. 3, p.
249, 1860; and La circulation du sang, p. 264, 1863. "Le dicrotisme du
pouls est un phénomène physiologique, on l'observe presque chez tous
les sujets; seulement il n'est sensible au doigt que dans les cas où
il est extrêmement prononcé." References for previous observations of
the slow descent of the pulse curve are given by Landois, Die Lehre vom
Arterienpuls, p. 36, 1872.]

[Footnote 40: Burdon Sanderson: Handbook of the Sphygmograph, 1867.]

[Footnote 41: Lorain: Études de médecine clinique; Le Pouls, 1870.]

[Footnote 42: Ozanam: La circulation et le pouls, 1886.]

[Footnote 43: Pfungen: Pulscurve der Arterien in Gad's Lexicon der
medicinischen Propaedeutik, vol. 3, pp. 544-642, 1895.]

[Footnote 44: Riegel: Ueber die Bedeutung der Pulsuntersuchung in
Volkmann's Sammlung klinischer Vortraege, Nos. 144, 145, 1878.]

[Footnote 45: Roy and Adami: Heartbeat and Pulsewave, The Practitioner,
vol. 1, pp. 81-94, 161-177, 241-253, 347-361, 412-425, 1890.]

[Footnote 46: Cf. Howell: American Text-book of Physiology, p. 436,
1897. Marey's first explanation of the dicrotic belongs to this type
(Comptes Rendus, vol. 47, p. 826, 22 Nov., 1858.) He supposed that
the dicrotic elevation was due to a wave reflected from the Iliacae
communes. He was led to this theory by the erroneous observation of
Beau, that abnormal dicrotism never occurs in the lower extremities.
Marey's own observations refuted this theory, since they show that the
dicrotic elevation is found also in the sphygmograms of the arteries of
the leg. (Marey, La circulation du sang, p. 274, 1863.)]

[Footnote 47: Marey, La circulation du sang, pp. 271, 272, 1863. "Dans
ces conditions, l'ondée lancée par les ventricules se porte vers la
périphérie, et par suite de la vitesse acquise, abandonne les régions
initiales de l'aorte pour distendre les extrémités du système artériel.
Arrêtée en ce dernier point par l'étroitesse des artères qui lui fait
obstacle, elle reflue vers l'origine de l'aorte; mais cette voie est
fermée par les valvules sygmoïdes. Nouvel obstacle, nouveau reflux, et
par suite nouvelle ondulation (où rebondissement). Ces oscillations
alternatives se produisent jusqu'à ce qu'une nouvelle contraction du
ventricule vienne y mettre fin en produisant une onde nouvelle."]

[Footnote 48: This view was held by Haller, Bichat, and Bourgelat and
goes back to Galen ("Omnes enim clare cernunt, omnes partes arteriarum
eodem distendi tempore," De causis pulsi, book 2, c. 8). The first who
saw that the pulse did not appear at the same time in all the parts of
the body was Josias Weitbrecht, but his observations were neglected
until E. H. Weber actually measured the velocity of the propagation of
the pulse wave. (His famous thesis of 1827,--"Pulsum arteriarum non in
omnibus arteriis simul, sed in arteriis a corde valde remotis serius
quam in corde et in arteriis cordi vicinis fieri.") For the results of
other measurements see Tigerstedt: Physiologie des Kreislaufes, p. 385,
1894. Some use of these measurements is made in the present writer's
L'Analyse des Sphygmogrammes, which is to appear in the Journal de
Physiologie et de Pathologie Générale for May, 1906.]

[Footnote 49: Landois: Human Physiology (English translation), p. 145,
1889, and Die Lehre vom Arterienpuls, p. 188, 1872.]

[Footnote 50: Otto und Haas: Vierteljahrsschrift f. praktische
Heilkunde, vol. 34, p. 41, 1877.]

[Footnote 51: Garrod: Journal of Anatomy and Physiology, vol. 5, pp.
17-27, 1870.]

[Footnote 52: Traube: Gesammelte Beiträge, vol. 3, p. 595, 1878.]

[Footnote 53: Rosenstein: Deutsches Archiv f. klinische Medicin, vol.
23, pp. 75-97, 1879.]

[Footnote 54: Maurer: Deutsches Archiv f. klinische Medicin, vol. 24,
pp. 291-341.]

[Footnote 55: Gibson: Journal of Anatomy and Physiology, vol. 14, pp.
234-240, 1879.]

[Footnote 56: Fr. Frank: Travaux du laboratoire Marey, pp. 301-327,
1877.]

[Footnote 57: I. G. Edgren: Skandinavisches Archiv f. Physiologie, vol.
1, pp. 67-152, 1889.]

[Footnote 58: v. Kries: Studien zur Pulslehre, p. 62, and M. v. Frey,
Die Untersuchung des Pulses, p. 164.]

[Footnote 59: Grashey: Die Wellenbewegung elastischer Röhren, p. 166,
1881.]

[Footnote 60: Hoorweg: Archiv f. d. ges. Physiologie, vol. 46, p. 143,
1890.]

[Footnote 61: Marey, _loc. cit._, pp. 267-271, and Traité de Physique
Biologique (publié par d'Arsonval, Chauveau, Gariel, Marey), vol. 1,
p. 390, 1901; these tracings are reproduced rather frequently; _e.g._,
Pfungen, _loc. cit._, p. 563, and Chapman: Human Physiology (2d ed.),
p. 270, 1899.]

[Footnote 62: E. Mach: Sitzungsberichte der K. Akademie der
Wissenschaften, vol. 47 (2), p. 43, 1863; and in the tables, figs.
48-53.]

[Footnote 63: L. Hill, in Schaefer's Text-book of Physiology, vol. 2,
p. 111, 1902. The same opinion maybe found in Hermann's Lehrbuch der
Physiologie (12th ed.), p. 79, 1900. "Ihre (der dicrotischen Wellen)
Erklärung ist noch nicht widerspruchsfrei gestellt." In the previous
editions different views were given, and this critical doubt may be
regarded as the final outcome of the investigations of almost half a
century.]

[Footnote 64: Sudden compression is the most convenient way of
producing a wave in a liquid which is enclosed in an elastic tube.
It was used already in the first experiments on the propagation of
these waves. (E. H. Weber: Anwendung der Wellenlehre auf die Lehre vom
Kreislaufe des Blutes und insbesondere auf die Pulslehre, Berichte
d. Kgl. Sächsischen Ges. d. Wissenschaften, Math.-Phys. Cl., p. 177,
1850.)]

[Footnote 65: Bayliss and Starling: On the form of the intraventricular
and aortic pressure curves obtained by a new method, Intern.
Monatsschrift f. Anatomie u. Physiologie, vol. 11, pp. 426-435, 1894.]

[Footnote 66: W. T. Porter: A new method for the study of the
intraventricular pressure curve, Journal of Experimental Medicine,
vol. 1, pp. 296-303, 1896. A similar method was used by O. Frank: Ein
experimentelles Hülfsmittel für die Kritik der Kammerdruckkurven,
Zeitschrift f. Biologie, vol. 35, pp. 478-480, 1897.]

[Footnote 67: O. Frank, _loc. cit._, p. 480.]

[Footnote 68: Huerthle: Beiträge zur Hämodynamik, VIII, Zur Kritik des
Lufttransmissionsverfahrens, Arch. f. d. ges. Physiologie, vol 53, pp.
281-331, 1892.]

[Footnote 69: The special assumption on the rate of outflow
is by no means essential for the following theory. Two other
possible assumptions are mentioned in the author's "L'Analyse des
Sphygmogrammes," and others may be found easily. Every one of these
theories is equally probable as long as no experimental evidence can be
brought forward. The assumption that the rate of outflow through the
arterioles is uniform has the merit that it is the simplest and that it
can be deduced from considerations of the average directions of tubes
which split up in "every" direction.]

[Footnote 70: A detailed discussion shows that four different cases are
possible, but this distinction is of minor importance for the purpose
of this paper. The distinction holds that the movement is either
periodic or aperiodic.]

[Footnote 71: Edgren: Kardiographische und sphygmographische
Untersuchungen, Skandinavisches Archiv f. Physiologie, vol. 1, pp.
88-91, 1889; Fredericq: Vergleich der Stoss und Druckcurven der rechten
Herzkammer des Hundes, Centralblatt f. Physiologie, vol. 7, p. 770,
1893; Einthoven und Geluk: Die Registrierung der Herztöne, Archiv f. d.
Ges. Physiologie, vol. 57, p. 631, 1894.]

[Footnote 72: K. Huerthle: Beiträge zur Hämodynamik, Archiv f. d. Ges.
Physiologie, vol. 60, p. 281, 1895.]

[Footnote 73: Edgren: _loc. cit._ p. 87.]

[Footnote 74: A. Chauveau: Inscription électrique des mouvements
valvulaires, Journal de Physiologie et de Pathologie Générale, vol. 1,
p. 388, fig. 4, 1899.]

[Footnote 75: _Ibid._ p. 391, fig. 6 (curve 5); and the same author's
La pulsation cardiaque, in the same Journal, vol. 1, p. 795, fig. 5,
and p. 796, 1899.]

[Footnote 76: Edgren: _loc. cit._ p. 114.]

[Footnote 77: The exposition of this theory may be found in R.
Tigerstedt: Intracardialer Druck und Herzstoss, Ergebnisse der
Physiologie, vol. 1, pp. 258-262, 1902. This theory, equally
remarkable for its logical beauty and for its confirmation by Edgren's
experiments, has not found its way into recently published text-books
of physiology, though Edgren's paper belongs to the most frequently
quoted publications on sphygmography and cardiography.]

[Footnote 78: Tigerstedt: _loc. cit._ p. 261.]

[Footnote 79: Fredericq: La pulsation du cœur chez le chien, no. 5.
La comparaison du tracé du choc du cœur avec celui de la pression
intraventriculaire, Travaux du Laboratoire de Liège, vol. 5, p. 67,
1896.]

[Footnote 80: Galenus: De pulsuum differentiis, lib. 1. c. 16.]

[Footnote 81: The amount of change in the base-line is the chief
difference between the sphygmograms and the plethysmograms. It was
stated recently that for this reason the plethysmograph could not be
used for psychological experiments. An analysis of the mechanical
conditions of these two instruments shows that also the sphygmogram
must show some plethysmographic influences, and the author supplied
experimental evidence for this result.]

[Footnote 82: More recent publications have taken this view. Cohn
speaks of "Organgefühle des Gehirns," approaching Meynert's view on
the causes of pleasure and pain. (P. Cohn: Gemüthserregungen und
Krankheiten, pp. 23 and 50, 1903.) Cohn's book shows clearly that
this theory belongs to the type of philosophical explanations. This
is also suggested by Duprat who remarked the parallelism between the
theories of James, Lange and Ribot, and the theories of certain Greek
philosophers. (Duprat: "La psycho-physiologie des sentiments dans la
philosophie ancienne," Archiv f. Geschichte d. Philosophie, vol. 18
(3), p. 395, 1905.)]




THE MUTUAL INFLUENCE OF FEELINGS

BY JOHN A. H. KEITH


The object of this investigation was to ascertain the mutual influence
of simultaneous stimuli that appealed to different senses with regard
to the _intensity_ of their feeling values. The investigation covers
combinations: (1) of colors and active touches, (2) of colors and
passive touches, (3) of tones and active touches, (4) of tones and
passive touches, (5) of colors and tones.

The basis of appreciation was a numerical scale[83] as follows:

    1. Very disagreeable.
    2. Disagreeable.
    3. Slightly disagreeable.
    4. Indifferent.
    5. Slightly agreeable.
    6. Agreeable.
    7. Very agreeable.

The color series began with the one hundred thirty-six colors as put
out by the Milton Bradley Co. This series consists of ninety pure
spectrum colors, ten whites, blacks, and grays, and thirty-six broken
spectrum colors. The colors were exposed at the back of a semicircular
black-lined box for about two seconds. The subject was seated at a
convenient distance, about three and a half feet, from the colors. In
order to have a constant light, all experiments were conducted in a
dark room with an electric light suspended over the subject's head.
The whole series was used for ten times in order to get the range of
judgments. Then twenty-eight colors, covering as fully as possible the
range from 1 to 7, were selected for further experiment in combination.

At the same time a series of thirty-six touches, from velvet to
sandpaper, was being employed as the colors were. From this number
fourteen were finally selected.

Similarly, by using a reed box, with reeds ranging from 128 to 1024
vibrations per second and separated from each other by four vibrations,
from a much larger series twenty-seven tone-combinations were finally
selected.

Moreover, from time to time, each selected series was given alone; and
on the basis of these readings, averaging from thirty to forty, the
"standard" for each stimulus was made. Tables I to III give a brief
description of the stimuli and also the "standards" for each of two
subjects, F. and M.


TABLE I. COLORS

    _No. of_        _Description._      _Standard_  _Standard_
    _Color._                             _for F._    _for M._

      1       Violet Red. Tint no. 1      5.90       4.00
      2       Red. Tint no. 1             6.00       6.00
      3       Red                         6.00       5.40
      4       Orange Red                  5.60       6.20
      5       Red Orange                  4.20       5.20
      6       Yellow Orange. Tint no. 1   3.10       4.50
      7       Yellow. Tint no. 1          2.30       5.20
      8       Yellow. Shade no. 2         2.00       4.00
      9       Green Yellow                2.80       5.60
     10       Yellow Green                4.45       6.20
     11       Green. Shade no. 1          5.70       4.40
     12       Blue Green. Tint no. 1      3.40       6.00
     13       Green Blue                  5.00       3.50
     14       Blue. Shade no. 1           5.40       2.10
     15       Blue Violet                 5.10       5.00
     16       Violet. Tint no. 2          4.20       5.30
     17       Violet. Shade no. 2         5.50       3.50
     18       Red Violet                  5.65       3.70
     19       Black                       4.00       4.00
     20       Green Gray. no. 1           2.20       4.20
     21       Green Gray. no. 2           2.00       4.00
     22       A-Red. Light                3.70       3.70
     23       A-Red. Dark                 2.10       2.70
     24       A-Orange. Dark              2.00       3.00
     25       A-Yellow Orange. Light      3.10       4.30
     26       A-Green Yellow. Dark        2.50       4.00
     27       A-Blue Green. Medium        2.40       4.30
     28       A-Violet. Medium            4.10       4.50

                   Totals               110.40     124.50
                   Average                3.94       4.44

In connection with this table it may be noted that the subjects agree
regarding 2, 19, and 22, red tint no. 1, black, and A-red light;
that M. estimates the colors higher than F. in seventeen cases; and
that F. estimates 1, 3, 11, 13, 14, 15, 17, 18, higher than M. does.
These individual differences are probably explicable on grounds
of association; they are not, however, connected with the problem
under consideration here, for we are concerned with the effect of
combinations with other stimuli.


TABLE II. TOUCHES

The various articles were fastened to small pieces of wood and placed
in small cardboard boxes. In active touch, the subjects were allowed
to stroke the object gently twice, always with a contracting movement
of the forefinger of the right hand. In passive touch, the operator
stroked the subject's forefinger with the object, twice as before. The
table explains itself.

                                                        _Standards._
    _No. of_                                      _Active._     _Passive._
    _Touch._          _Description._             _F._  _M._    _F._   _M._

     1    Thick napped velvet                    6.90   6.60    7.00   7.00
     2    Thin stretched rubber, such as is
            used on tambours                     5.70   5.40    6.00   6.00
     3    Glazed thin cardboard                  5.80   5.50    5.50   6.00
     4    White silk ribbon--always stroked
            with the ribs                        5.80   5.70    5.40   5.00
     5    Soft, split, rough leather             5.50   5.50    5.00   6.00
     6    Smooth polished cork                   5.70   5.10    5.40   6.00
     7    Glazed tin                             5.00   4.50    6.00   6.00
     8    Rough, tarred paper                    4.30   4.60    4.00   5.00
     9    Blue blotting paper                    4.50   5.10    4.20   5.00
    10    Sand paper no. 1, fine grained         2.20   3.50    2.00   4.00
    11    Shot no. 3, set in paraffine           2.70   4.90    3.00   3.50
    12    Sandpaper no. 2-1/2, coarse-grained    1.40   2.00    1.50   3.00
    13    A coarse, rough, ridged cotton cloth,
            always stroked across the ridges     4.00   2.40    4.70   3.50
    14    A thin, closely woven white muslin     4.20   4.60    4.20   5.00

                         Totals                 63.70  65.40   63.90  71.00
                         Average                 4.55   4.67    4.56   5.14


TABLE III. TONES

The table of tone-combinations shows that some are simply repetitions
of the same chord in a higher octave, as 1 and 9, 5 and 13, 15 and 14.
The first sixteen are harmonious; so also the twenty-sixth; the others
introduce beats and discords, some of which are agreeable, as 20 and
21, while others are disagreeable, as 19 and 27. Individual differences
appear in this as in the previous series. The totals introduced into
the tables simply go to show that in each series the total judgments
are not widely diverse.

    _No of_                         _Standards._
    _Tones._  _Description._       _F._     _M._

      1    256, 320                4.25    3.25
      2    256, 384                4.25    3.00
      3    256, 320, 384           5.60    3.00
      4    320, 384                4.30    3.50
      5    256, 512                4.65    5.50
      6    320, 512                4.10    5.00
      7    384, 512                4.90    5.00
      8    512, 640                5.40    5.30
      9    512, 768                5.50    5.80
     10    640, 768                4.40    5.00
     11    640, 1024               4.25    3.40
     12    768, 1024               4.80    5.50
     13    512, 1024               5.60    5.80
     14    512, 640, 768, 1024     6.30    5.90
     15    256, 320, 384, 512      6.50    4.00
     16    320, 512, 768           5.00    4.70
     17    136, 144                2.00    2.50
     18    156, 160                3.80    4.00
     19    136, 140                2.70    3.50
     20    440, 444                6.81    6.00
     21    504, 508                6.70    6.40
     22    148, 152, 156           1.70    2.50
     23    172, 296, 452           3.00    2.50
     24    180, 480, 768           3.40    2.50
     25    232, 328, 492           3.20    2.50
     26    256, 512, 768           5.00    5.50
     27    256, 240, 384           2.00    1.70

           Totals                120.11  113.25
           Average                 4.45    4.19

The following tables deal with the combined series.

Table IV shows that the appreciation of the colors was, in general,
lowered slightly by the combinations with the tones; and, also, that
the appreciation of the tones was lowered more than one point by the
combinations with the colors. By referring to Tables I and III, M.'s
average for the colors is 4.44 and for the tones 4.20.


TABLE IV. COLOR-TONE RESULTS. M.

           _No. of times color was_        _Total No. of points_
    _No. of_                       _Not_    _color was_ _Net result_
  _Color_  _Raised_ _Lowered_  _Affected_ _Raised_+ _Lowered_-  +    -
     1        3      15         9      3      15                    12
     2        0      22         5      0      23                    23
     3        0      27         0      0      32.2                  32.2
     4        0      27         0      0      20.4                  20.4
     5        0      27         0      0       8.2                   8.2
     6       23       4         0     12.5     2             10.5
     7       14      13         0     11.2     2.6            8.6
     8        6       5        16      6       5              1
     9       16      11         0      6.4     6.6                    .2
    10        1      26         0       .8    20.2                  19.4
    11        0      27         0      0      36.8                  36.8
    12        0      23         4      0      23                    23
    13        2      25         0      1      18.5                  17.5
    14        9      18         0      8.1     9                      .9
    15        1      12        15      1      14                    13
    16        1      26         0       .7    16.8                  16.1
    17        7      20         0      6.5    12                    5.5
    18       13      14         0      4.9    10.8                   5.9
    19        0       0        27      0       0
    20        1      26         0       .8     9.2                   8.4
    21        0       6        21      0       6                     6
    22        7      20         0      3.1    16                    12.9
    23       21       6         0      9.3     4.2            5.1
    24       12       2        13     17       3             14
    25        8      19         0      5       7.7                   2.7
    26        1       4        22      1       4                     3
    27       16      11         0     12       3.3            8.7
    28       13      14         0      6      10                     4

    Totals  175     450       132                           47.9   271.1

    Grand Total (28 × 27) 756.             Net lowered             223.2

    Average lowering of each color judgment,          223.2/756 = .295
            % of judgments of color lowered,          450/756   = 59+
            % of judgments of color raised,           175/756   = 23+
            % of judgments of color not affected,     132/756   = 17+


           _No. of times tone was_     _Total No. of points_
   _No. of_                     _Not_      _tone was_   _Net result_
   _Tone_  _Raised_  _Lowered_ _Affected_  _Raised_+ _Lowered_- +    -
     1        0      28        0       0      29                    29
     2        4       9       15       4       9                     5
     3        2      22        4       2      28                    26
     4        7      21        0       4.5    17.5                  13
     5        0      28        0       0      32                    32
     6        0      24        4       0      35                    35
     7        0      22        6       0      32                    32
     8        4      24        0       2.8    11.2                   8.4
     9        3      25        0        .6    37                    36.4
    10        0      17       11       0      25                    25
    11       15      13        0      16       7.2            8.8
    12        6      22        0       3      20                    17
    13       10      18        0       2      22.4                  20.4
    14        3      25        0       .3    37.5                   37.2
    15        0      27        1       0      55                    55
    16        4      24        0       1.2    38.8                  37.6
    17        2      26        0       1      22                    21
    18        0      27        1       0      66                    66
    19        0      28        0       0      56                    56
    20        1      19        8       1      23                    22
    21        6      22        0       3.6    13.8                  10.2
    22        0      28        0       0      37                    37
    23        1      27        0        .5    33.5                  33
    24       10      18        0       5      13                     8
    25        7      21        0       3.5    24.5                  21
    26        5      23        0       2.5    50.5                  48
    27        1      27        0        .3    18.9                  18.6

    Totals   91     615        50                             8.8  749.8

    Grand Total (27 × 28),  756.                Net lowered        741

        Average lowering of each tone judgment,               .98+
                % of judgments of tones lowered,     615/756 = 81+
                % of judgments of tones raised,       91/756 = 12+
                % of judgments of tones not affected, 50/756 =  6+


TABLE V. COLOR-TONE RESULTS. F.

          _No. of times color was_  _No. of points color was_ _Net result_
   _No. of_                    _Not_
   _Color_ _Raised_ _Lowered_ _Affected_ _Raised_ + _Lowered_ -   +      -
     1        9       18         0       10         19.2               9.2
     2        0       24         3        0         38                38
     3        0       26         1        0         31                31
     4        1       26         0        0.4       33.6              33.2
     5        9       18         0        7.2       16.6               9.4
     6        3       24         0        3.7       12.4               8.7
     7       13       14         0        9.1        4.2        4.9
     8        9        4        13        9          4          5
     9       21        6         0       12.4        5.8        6.6
    10       18        9         0       15.90       12         3.9
    11       19        8         0       10.7        5.6        5.1
    12        9       18         0        7.4       13.2               5.8
    13        5        7        15        6          7                 1
    14        7       20         0        6.2        4          2.2
    15        5       22         0        6.5        5.1        1.4
    16        1       26         0         .8       17.2              16.4
    17        2       25         0        1         13.5              12.5
    18        3       24         0        1.35      18.6              17.25
    19        0        3        24        0          3                 3
    20       18        9         0       14.4        2.8              11.6
    21        9        0        18        9          0          9
    22        5       22         0        4.5       26.2              21.7
    23        2       25         0        1.8        5.5               3.7
    24        5        7        15        6          7                 1
    25        1       26         0         .9       28.4              27.5
    26       16       11         0        8          5.5        2.5
    27       19        8         0       14.4        3.2       11.2
    28        7       20         0        6.3        4          2.3

   Totals   215      451        90                             54.1  250.95
             Grand Total       756                   Net lowered     196.85

           Average lowering of each color judgment,                .26+
                  % of judgments of color lowered,      451/756 = 59.6+
                  % of judgments of color raised,       215/756 = 28+
                  % of judgments of color not affected,  90/756 = 11.9+


           _No. of times tone was_    _No. of points tone was_ _Net result_
   _No. of_                      Not
   _Tone_ _Raised_ _Lowered_ _Affected_ _Raised_ + _Lowered_-    +      -
     1        6       22         0         3         20.5              17.5
     2        2       26         0         1.5       22.5              21
     3        2       26         0          .8       24.6              23.8
     4        2       26         0         1.4       16.8              15.4
     5        4       24         0         1.4       27.6              26.2
     6        3       25         0         6.7       26.5              19.8
     7        4       24         0         1.4       40.6              39.2
     8        1       27         0          .6       32.8              32.2
     9        1       27         0          .5       34.5              34
    10        8       20         0         5.8       25                19.2
    11        3       25         0         2.25      33.25             31
    12        4       24         0          .8       38.2              37.4
    13        2       26         0          .8       23.6              22.8
    14        2       26         0         1.4       25.8              24.4
    15        1       27         0          .5       32.5              32
    16        0       21         7         0         40                40
    17        4       10        14         4         10                 6
    18       18       10         0        22.6       14          8.6
    19       17       11         0        22.1       13.7        8.4
    20        6       22         0         4.8       22.6              17.8
    21       10       18         0         3         17.6              14.6
    22       18       10         0        10.4        7          3.4
    23        1       20         7         2         21                19
    24        0       28         0         0         20.4              20.4
    25        3       25         0         2.4       21                18.6
    26        0       22         6         0         39                39
    27        1       11        16         1         11                10

   Totals   123      583        50                              20.4  581.3
         Grand Total  756                             Net lowered,    560.9

          Average lowering of each tone judgment,                .742+
                  % of judgments of tones lowered,       583/756 = 77+
                  % of judgments of tones raised,        123/756 = 16+
                  % of judgments of tones not affected,  50/756 =  6+


Table VI. Tone-Active Touch Results. M.

         _No. of times tone was_     _No. of points tone was_  _Net result_
  _No. of_                      _Not_
  _Tone_    _Raised_ _Lowered_ _affected_  _Raised_ +  _Lowered_ -  +    -
     1        14        0          0          31.5         0       31.5
     2        14        0          0          29           0       29
     3        14        0          0          30           0       30
     4        14        0          0          32           0       32
     5         9        5          0           9.5         4.5      5
     6        14        0          0          25           0       25
     7        14        0          0          22           0       22
     8        14        0          0          22.8         0       22.8
     9        14        0          0          16.8         0       16.8
    10        14        0          0          25           0       25
    11        14        0          0          46.8         0       46.8
    12        14        0          0          21           0       21
    13        14        0          0          13.8         0       13.8
    14        13        1          0          13.4          .9     12.5
    15        13        0          1          23           0       23
    16        14        0          0          24.2         0       24.2
    17        14        0          0          25           0       25
    18        13        0          1          28           0       28
    19        14        0          0          30           0       30
    20        13        1          0          13           0       13
    21        13        1          0           7.8         0        7.8
    22        14        0          0          24           0       24
    23        14        0          0          21           0       21
    24        14        0          0          42           0       42
    25        14        0          0          38           0       38
    26        12        2          0          11           2        9
    27        13        1          0          22.9          .6     22.3

    Totals,  365       11          2                              640.5
           Grand Total, 378.                          Net raised, 640.5

                 Average raising of each tone judgment,         1.69+
                 % of judgments of tones lowered,        11/378 = .2+
                 % of judgments of tones raised,        365/378 = 97+
                 % of judgments of tones not affected,    2/378 = .5+


                _No. of times touch was_   _No. of points touch_
  _No. of_                      _Not_       _was_              _Net result_
  _Touch_   _Raised_ _Lowered_ _affected_ _Raised_ + _Lowered_ -  +     -
     1         27       0         0           10.8        0      10.8
     2          6      21         0            3.6       12.4           8.8
     3         10      17         0            5.8        8.5           3.5
     4          6      21         0            2.8       16.7          13.9
     5         14      13         0            8          6.5     1.5
     6          8      19         0            7.2        6.9      .3
     7         16      11         0           11          6.5     4.5
     8         22       5         0           13.8        3      10.8
     9          7      20         0            6.3        5       1.3
    10         26       1         0           23           .5    22.5
    11         21       6         0            2.1        5.4           3.3
    12         26       0         1           38          0      38
    13         27       0         0           48.4        0      48.4
    14         23       4         0           14.2        2.4    11.8

  Totals   239         138        1                             149.9  29.5
        Grand Total, 378                            Net raised, 120.4

      Average raising of each active touch judgment,                   .31+
                    % of judgments of active touch lowered,   138/378 = 36+
                    % of judgments of active touch raised,    239/378 = 63+
                    % of judgments of active touch not affected, /378 = .2+

Under combination influences, the average is reduced to 4.14+ for
colors, and 3.12 for tones.

The next Table, V, shows the color-tone results for F. as Table IV
showed them for M. F.'s average for colors (Table I) alone was 4.35,
and was reduced in combination with tones by .26, or to 4.19. So, also,
F.'s average for tones alone (Table III), 4.33, was reduced by .74+
to 3.59. The averages in both cases show the same general tendency
to a lowering of the appreciation in both series when the series are
combined, but the tones are lowered more than the colors.

Table VI shows the effect of combining tones and active touches as
reported by M.

The effect of this combination is clear and unmistakeable. The
appreciation of the tones is raised 1.71+ points; and of the active
touches, .31+ points. This result is the opposite of that shown
in Table IV, where colors and tones were combined. There is this
agreement, however, that the appreciation of the tones is changed more
than that of the other stimuli. Relatively, the appreciation of the
touches changes least.

Table VII shows the effects on F. of combining tones and active
touches. The same general tendencies appear as in the case of M.; but
the changes in appreciation are not so marked. This is not easy to
explain, for F. estimated both the active touches and the tones higher
when taken alone than did M.

Table VIII shows the effect on F. of combining tones and passive
touches. The same general tendency to increased appreciation appears,
but the tones are raised more and the touches raised less than in
Table VII. This may be explained, perhaps, on the basis of increasing
appreciation with increased participation.

Tables IX and X show the effect on M. and F., respectively, of
combining colors and active touches. M. estimates both slightly higher,
while F. estimates both slightly lower. This difference cannot be
explained by the standards for each subject. From Tables I and III we
get:

                  _Standards._

    _Colors._                _Active Touches._

     M. 4.44                            4.24
     F. 4.35                            4.55

    With M. the colors go up to            4.87
    With F. the colors go down to          4.22
    With M. the active touches go up to    4.41
    With F. the active touches go down to  4.17


Table VII. Tone-Active Touch Results. F.

            _No. of times tone was_
  _No. of_                    _Not_   _No of points tone was_ _Net results_
  _Tone_  _Raised_ _Lowered_ _affected_ _Raised_+  _Lowered_-     +     -
     1      11       3           0        9.25        1.75       7.5
     2      10       4           0        9.5         1          8.5
     3       7       7           0        5.8         7.2               1.4
     4      10       4           0       12           3.2        8.8
     5       9       5           0        9.85        4.25       5.6
     6       9       5           0       15.1         4.5       10.6
     7      10       4           0        8           6.6        1.4
     8       7       7           0        5.2         6.8               1.6
     9      10       4           0        8           9                 1
    10      12       2           0       14.2         2.8       11.4
    11      12       2           0       16           2.5       13.5
    12      13       1           0        9.6          .8        8.8
    13      13       1           0       12.2          .6       11.2
    14      10       4           0        7           2.2        4.8
    15       6       8           0        3           4                 1
    16       9       1           4       11           2          9
    17       9       0           5       14           0         14
    18      12       2           0       16.4         2.6       13.8
    19      12       2           0       27.6         2.4       25.2
    20      12       2           0        2.4         2.6                .2
    21      14       0           0        4.2         0          4.2
    22      14       0           0       17.2         0         17.2
    23       4       5           5        6           5          1
    24       0      14           0        0           9.6               9.6
    25       7       5           2        5.6         6                  .4
    26       6       1           7        8           1          7
    27      11       0           3       14           0         14

   Totals,  259     93          26                             197.5   15.2
          Grand Total, 378                         Net raised, 182.3

         Averaging raising of each tone judgment,                  .48+
                   % of judgments of tones lowered,        93/378 = 24+
                   % of judgments of tones raised,        259/378 = 68+
                   % of judgments of tones not affected,   26/378 = 7-


           _No of times touch was_
  _No. of_                     _Not_  _No. of points tone was_ _Net result_
  _Touch_ _Raised_ _Lowered_ _affected_ _Raised_+ _Lowered_-  +       -
     1       22        5         0         2.2       6.5             4.3
     2       16       11         0         8.8       8.7      .1
     3       20        7         0         7         5.6     1.4
     4       19        8         0         5.8      11.4             5.6
     5       18        9         0        14         7.5     6.5
     6       16       11         0         6.8       9.7             2.9
     7       13        5         9        13         5       8
     8       13       14         0        13.1      12.2      .9
     9       19        8         0        14.5       5       9.5
    10        9       18         0         8.2       7.6      .6
    11       22        5         0         8.6       3.5     5.1
    12       15       12         0        15         4.8    10.2
    13       22        1         4        35         1       34
    14       20        7         0        26         1.4     24.6

   Totals,  244      121        13                          100.9   12.8
            Grand Total, 378                             Net raised, 88.1

   Average raising of each active touch judgment,                   .23+
              % of judgments of active touch lowered,      121/378 = 32+
              % of judgments of active touch raised,       244/378 = 64+
              % of judgments of active touch not affected,  13/378 = 3+


TABLE VIII. TONE-PASSIVE TOUCH RESULTS. F.

       _No. of times tone was_
  _No. of_                  _Not_   _No. of points tone was_  _Net result_
  Tone_  _Raised_ _Lowered_ _affected_ _Raised_ + _Lowered_ -  +         -
    1       13        1       0          17.75       .25      17.5
    2       12        2       0          18         0         18
    3       13        1       0          11.2        .6       10.6
    4       14        0       0          14.8       0         14.8
    5       11        3       0           6.65      1.95       4.7
    6        9        5       0          12.1       1.50      10.6
    7       13        1       0           6.3        .9        5.4
    8       14        0       0          17.4       0         17.4
    9        6        8       0           5         4          1
   10       14        0       0          20.4       0         20.4
   11       14        0       0          12.5       0         12.5
   12       10        4       0           7         5.2        1.8
   13       14        0       0          10.6       0         10.6
   14       12        2       0           8.4        .6        7.8
   15        8        6       0           4         3          1
   16        1        2      11           1         2                     1
   17        4        0      10           4         0          4
   18        5        9       0           7         8.2                 1.2
   19        9        5       0           2.7       4.5                 1.8
   20       13        1       0           2.6        .8        1.8
   21       13        1       0           3.9        .7        3.2
   22       13        1       0           8.9        .7        8.2
   23        6        2       6          10         2          8
   24       12        2       0          15.2        .8       14.4
   25       11        3       0          15.8        .6       15.2
   26        0        6       8           0         8                   8
   27       10        0       4          13         0         13

  Totals,  274       65      39                                221.9   12.0
             Grand Total, 378                      Net raised, 209.9

                Average raising of each tone judgment,              .55+
                    % of judgments of tones lowered,        65/378 = 17+
                    % of judgments of tones raised,        274/378 = 72+
                    % of judgments of tones not affected,   39/378 = 10+


        _No. of times touch was_
  _No. of                     _Not_  _No. of points touch was_ _Net result_
  Touch_ _Raised_ _Lowered_ _affected_ _Raised_ + _Lowered_ -   +     -
     1      0        10       17          0           10             10
     2      4        13       10          4           16             12
     3     11        16        0          6.5         12              5.5
     4     13        14        0         12.8          9.6     3.2
     5     12         5       10         17            8       9
     6     12        15        0          7.2         12              4.8
     7      3        12       12          3           14             11
     8     14         2       11         17            2      15
     9     11        16        0         12.8         20.2            7.4
    10     19         1        7         19            1      18
    11      5         2       20          5            2       3
    12     27         0        0         34.5          0      34.5
    13     14        13        0          8.2          9.1             .9
    14     17        10        0         25.6          2      23.6

 Totals,  162       129       87                             106.3   51.6
    Grand Total,    378                       Net raised,  54.7

      Average raising of each passive touch judgment,       .14+
      % of judgments of passive touch lowered,     129/378 = 34
      % of judgments of passive touch raised,      162/378 = 42
      % of judgments of passive touch not affected, 87/378 = 23


TABLE IX. COLOR-ACTIVE TOUCH RESULTS. M.

            _No. of times color was_
  _No. of_                   _Not_ _No. of points color was_   _Net result_
  _Color_ _Raised_ _Lowered_ _affected_ _Raised_ + _Lowered_ -    +      -

     1         5        1        8         5         1           4
     2         0        7        7         0         7                  7
     3         3       11        0         1.8       4.4                2.6
     4         3       11        0         2.4       4.2                1.8
     5         3       11        0         2.4       2.2          .2
     6        12        2        0         6         1           5
     7         2       12        0         1.6       2.4                 .8
     8        12        0        2        12         0          12
     9        13        1        0        14.2        .6        13.6
    10         7        7        0         5.6       2.4         3.2
    11        11        3        0         6.6       1.2         5.4
    12         7        1        6         7         1           6
    13         9        5        0         5.5       2.5         3
    14        11        3        0        13.9        .3        13.6
    15        11        0        3        12         0          12
    16        10        4        0         8         1.2         6.8
    17        12        2        0        16         1          15
    18        14        0        0        24.8       0          24.8
    19         0        0       14         0         0
    20         9        5        0         7.2       1           6.2
    21         1        0       13         1         0           1
    22        14        0        0         5.2       0           5.2
    23        14        0        0        10.2       0          10.2
    24         8        0        6        14         0          14
    25         9        5        0         6.3       1.5         4.8
    26         4        1        9         4         1           3
    27        11        3        0         4.3        .9         3.4
    28        13        1        0        10.5        .5        10

  Totals,    228       96       68                             182.4   12.2
                Grand Total, 392                          Net raised, 170.2

                  Average raising of each color judgment,        .43+
                 % of judgments of colors lowered,       96/392 = 24+
                 % of judgments of colors raised,       228/392 = 58+
                 % of judgments of colors not affected,  68/392 = 17+


            _No. of times touch was_
  _No. of                    _Not_  _No. of points touch was_  _Net Result_
  Touch_ _Raised_ _Lowered_  _affected_ _Raised_ + _Lowered_ - +    -

     1      8       20           0        3.2        15               11.8
     2      4       24           0        3.4        11.6              8.2
     3     18       10           0       10           5        5
     4      8       20           0        3.4        23               19.6
     5      8       20           0        5          13                8
     6      2       26           0        1.8         8.6              6.8
     7     25        3           0       26.5         4.5     22
     8     21        7           0       10.4         4.2      6.2
     9      3       25           0        2.7         8.5              5.8
    10     27        1           0       33.5          .5     33
    11     23        5           0        3.3         4.5              1.2
    12     22        0           6       41           0       41
    13     26        2           0       42.6          .8     41.8
    14     19        9           0        9.6         8.4      1.2

  Totals  214      172           6                           150.2    61.4
              Grand Total, 392                     raised,    88.8

     Average raising of each active touch judgment,                .22+
     % of judgment of active touch, lowered,              172/392 = 43+
     % of judgments of active touch raised,               214/392 = 54+
     % of judgments of active touch not affected,           6/392 = 1.8+


TABLE X. COLOR-ACTIVE TOUCH RESULTS. F.

              _No. of times color was_
  _No. of_                   _Not_   _No. of points color was_ _Net result_
  _Color_ _Raised_ _Lowered_ _affected_ _Raised_ + _Lowered_ -   +      -
     1       13         1         0        8.3         .9       2.4
     2        5         4         5        5          6                 1
     3        2         5         7        2          5                 3
     4        9         5         0        7.6        3         4.6
     5        9         5         0       16.2        7         9.2
     6        0        14         0        0         20.4              20.4
     7        3        11         0        2.1        5.3               3.2
     8        1         3        10        1          3                 2.0
     9        5         9         0        3          8.2               5.2
    10        9         5         0       10.95       8.25      2.7
    11       12         2         0        8.6        1.4       7.2
    12        8         6         0        9.8        2.4       7.4
    13        4         1         9        6          1         5
    14       10         4         0       11          1.6       9.4
    15        2        12         0        2.8        3.2                .4
    16        2        12         0        1.6        7.4               5.8
    17        8         6         0        5          4         1
    18       11         3         0        9.85       3.95      5.9
    19        0         0        14        0          0
    20        1        13         0         .8       14.4              13.6
    21        2         7         5        2          7                 5
    22        8         6         0       11.4        9.2       2.2
    23        0        14         0        0         11.4              11.4
    24        0        10         4        0         10                10
    25        2        12         0        3.8       10.2               6.4
    26        0        14         0        0         11                11
    27        0        14         0        0          8.6               8.6
    28        2        12         0        1.8        6.2               4.4

  Totals,   129       209        54                            57     111.4
             Grand total, 392                             Net lowered, 54.4

               Averaging lowering of each color judgment,         .13+
               % of judgments of colors lowered,         209/392 = 53+
               % of judgments of colors raised,          129/392 = 32+
               % of judgments of colors not affected,     54/392 = 13+


           _No. of times touch was_
  _No. of_                   _Not_   _No. of points touch was_ _Net result_
  _Touch_ _Raised_ _Lowered_ _affected_ _Raised_ +  _Lowered_ -   +      -
     1      11        17          0         1.1        28.3            27.2
     2      14        14          0        11.2        13.8             2.6
     3      19         9          0         9.8         7.2       2.6
     4      11        17          0         2.2        25.6            23.4
     5      13        15          0        12.5        16.5             4
     6      12        16          0         4.6        12.2             7.6
     7      15         0         13        23           0        23
     8       2        26          0          .6        26.8            26.2
     9       6        22          0         4          25              21
    10       0        28          0         0          17.6            17.6
    11      19         9          0         6.7         6.3     .4
    12       3        25          0         3.8        10               6.2
    13       2        26          0         2          32              30
    14       8        20          0         6.4        19              12.6

  Totals,  135       244         13                           26.0    178.4
            Grand Total, 392                             Net lowered, 152.4

            Average lowering of each active touch,                 .38+
            % of judgments of active touch lowered,       244/392 = 62+
            % of judgments of active touch raised,        135/392 = 34+
            % of judgments of active touch not affected,   13/392 =  3+


TABLE XI. COLOR-PASSIVE TOUCH RESULTS. M.

             _No. of times color was_
  _No. of_                   _Not_   _No. of points color was_ _Net Result_
  _Color_ _Raised_ _Lowered_ _affected_ _Raised_ + _Lowered_ -     +     -
     1        9        1         4         10          1           9
     2        7        2         5          7          2           5
     3        7        7         0          4.2        2.8         1.4
     4        6        8         0          4.8        1.6         3.2
     5        9        5         0         11.2        1          10.2
     6       14        0         0         21          0          21
     7       13        1         0         14.4         .2        14.2
     8       11        0         3         14          0          14
     9       14        0         0         18.6        0          18.6
    10        7        7         0          5.6        1.4         4.2
    11       10        4         0          7          2.6         4.4
    12        3        3         8          3          3
    13       13        1         0         12.5         .5        12
    14       14        0         0         25.6                   25.6
    15        2        0        12          2          0           2
    16       10        4         0         11          1.2         9.8
    17       13        1         0         16.5         .5        16
    18       13        1         0         13.9         .7        13.2
    19        0        0        14          0          0
    20       13        0         1         12.4        0          12.4
    21        7        0         7          7          0           7
    22       13        1         0         11.9         .7        11.2
    23       14        0         0         18.2        0          18.2
    24       12        0         2         22          0          22
    25       14        0         0         15.8        0          15.8
    26        9        0         5          9          0           9
    27       12        2         0          9.4         .6         8.8
    28       10        4         0          6          2           4

  Totals,   279       52        61                                 292.2
             Grand total, 392                          Net raised, 292.2

               Average raising of each color judgment,        .75-
               % of judgments of colors lowered,      52/392 = 13+
               % of judgments of colors raised,      279/392 = 71+
               % of judgments of colors not affected, 61/392 = 15+


             _No. of times touch was_
  _No. of_                    _Not_  _No. of points touch was_ _Net Result_
  _Touch_ _Raised_ _Lowered_ _affected_ _Raised_ + _Lowered_ -   +     -
     1        0        3         25         0          3               3
     2        5       13         10         5         16               11
     3        5       10         13         5         12                7
     4        8        3         17         9          3         6
     5        6        2         20         6          3         3
     6        5       11         12         5         12                7
     7        1       19          8         1         27               26
     8       12        0         16        12          0        12
     9       22        0          6        23          0        23
    10       21        0          7        22          0        22
    11       28        0          0        24          0        24
    12       16        0         12        19          0        19
    13       22        6          0        32          3        29
    14       20        0          8        22          0        22

  Totals,   171       67        154                            160     54
             Grand total, 392                             Net raised, 106

       Average raising of each passive touch judgment,          .27+
       % of judgments of passive touch lowered,         67/392 = 17+
       % of judgments of passive touch raised,          171/392 = 43+
       % of judgments of passive touch not affected,    154/392 = 39+


TABLE XII. COLOR-PASSIVE TOUCH RESULTS. F.

           _No. of times color was_
  _No. of_                    _Not_  _No. of points color was_ _Net Result_
  _Color_ _Raised_ _Lowered_ _affected_ _Raised_ + _Lowered_ -   +       -
     1       14        0         0         12.4        0        12.4
     2        7        2         5          7          2         5
     3        1        1        12          1          1
     4        9        5         0          3.6        3           .6
     5       14        0         0         15.2        0         15.2
     6        0       14         0          0          2.4              2.4
     7       13        1         0          9.1         .3        8.8
     8        7        0         7          9          0          9
     9       11        3         0          5.2        2.4        2.8
    10       11        3         0         12.05       4.35       7.7
    11       13        1         0         14.9         .7       14.2
    12        4       10         0          2.4        6                3.6
    13        1        1        12          1          1
    14       11        3         0          6.6        1.2        5.4
    15        9        5         0          4.5         .5        4
    16        5        9         0          4          3.8         .2
    17       12        2         0          6          2          4
    18       13        1         0         12.55        .65      11.9
    19        0        0        14          0          0
    20       11        3         0          8.8         .6        8.2
    21       10        0         4         11          0         11
    22       10        4         0         14          1.2       12.8
    23        3       11         0          4.7        1.1        3.6
    24        2        0        12          2          0          2
    25        4       10         0          5.8        1          4.8
    26        7        7         0          4.5        3.5        1
    27       11        3         0          7.6        1.2        6.4
    28       12        2         0         10.8         .2       10.6

  Totals,   226      100        66                               161.6   6
             Grand total, 392                       Net raised, 155.6

         Average raising of each color judgment,                .40-
         % of judgments of colors lowered,             100/392 = 25+
         % of judgments of colors raised,              226/392 = 57+
         % of judgments of colors not affected,         66/392 = 16+


           _No. of times touch was_
  _No. of_                    _Not_  _No. of points touch was_ _Net Result_
  _Touch_ _Raised_ _Lowered_ _affected_ _Raised_ + _Lowered_ -   +    -
     1       0        8        20          0          8                8
     2       3       14        11          3         15               12
     3      23        5         0         14.5        2.5      12
     4      17       11         0         11.2        4.4       6.8
     5      21        1         6         27          1        26
     6      16       12         0          9.6        5.8       3.8
     7       4       10        14          4         10                6
     8      13        7         8         14          7         7
     9      17       11         0         18.6        3.2      15.4
    10      15        1        12         15          1        14
    11       0        1        27          0          1                1
    12      28        0         0         21          0        21
    13      21        7         0         18.3        7.9      10.4
    14      24        4         0         31.2        1.8      29.4

  Totals,  202       92        98                             145.8   27
             Grand total, 392                     Net raised, 118.8

         Average raising of each passive touch judgment,              .30+
         % of judgments of passive touch lowered,             92/392 = 23+
         % of judgments of passive touch raised,             202/392 = 51+
         % of judgments of passive touch not affected,        98/392 = 25+

Evidently no dynamic explanation of this difference is possible. It
has been impossible to reveal the cause of this particular difference;
in all other respects, however, the subjects agree in the general
tendencies revealed.

There remain the combinations of colors and passive touches. These
combinations are shown in Tables XI and XII.

In these tables the same general tendency to estimate both higher when
colors and passive touches are combined appears. M. raises the colors
more than F., and F. raises the touches more than M. This is perfectly
regular, as the following table shows:

              _Standards._                 _Raised in Combination to_
         _Colors._   _Passive Touches._   _Colors._   _Passive Touches._
    M.     4.44            5.14              5.16            5.40
    F.     4.53            4.56              4.75            4.85

The whole results are recapitulated for both M. and F. in Table XIII.


TABLE XIII. RECAPITULATION

               _Average_    _% of judgments of_ (----)  _Av. + or - influ._
      F.      _Standard_ _Raised_ _Lowered_ _Not affected_ _on each_ (----)
  1 Colors       4.35       28      59.6       11.9       (Color)   -  .26+
    Tones        4.33       16      77          6         (Tone)    -  .74+

  2 Colors       4.35       32      53         13         (Colors)  -  .13+
    A. Touches   4.55       34      62          3         (A. T.)   -  .38+

  3 Colors       4.35       57      25         16         (Colors)  +  .40+
    P. Touches   4.56       51      23         25         (P. T.)   +  .30+

  4 Tones        4.33       68      24          7         (Tones)   +  .45+
    A. Touches   4.55       64      32          3         (A. T.)   +  .23+

  5 Tones        4.33       72      17         10         (Tones)   +  .52+
    P. Touches   4.56       42      34         23         (P. T.)   +  .14+

     M.
  1 Colors       4.44       23      59          17        (Color)   -  .29+
    Tones        4.20       12      81           6        (Tones)   - 1.08

  2 Colors       4.44       58      24          17        (Colors)  +  .43+
    A. Touches   4.24       54      43           1.8+     (A. T.)   +  .17+

  3 Colors       4.44       71      13          15        (Colors)  +  .72+
    P. Touches   5.14       43      17          39        (P.T.)    +  .26

  4 Tones        4.20       99        .2          .5      (Tones)   + 1.71+
    A. Touches   4.24       63      36            .2      (A. T.)   +  .31+

From this last table certain conclusions may be drawn.

(1) When Colors and Tones were combined, both were lowered in the
appreciation of both subjects. The percentages show: (_a_) That about
the same number of colors was lowered, 59.6% for F. and 59% for M.
(_b_) That about the same average displacement of colors occurred, .26+
for F. and .29+ for M. (_c_) That about the same number of tones was
lowered, 77% for F. and 81% for M. (_d_) That the tones were lowered
more for M., -.74+ for F., and 1.08+ for M.

(2) When Colors and Active Touches were combined, for F. both are
lowered; for M. both are raised. The colors are lowered only very
slightly, .13 for F, while for M they are raised .43+; and conversely,
the active touches are lowered .38+ for F, and raised only .17+ for M.
Still, it appears clear that with F. there was an interference,--and
both colors and active touches are lowered while the same combinations
with M. are mutually reënforcing.

(3) When Colors and Passive Touches were combined, the appreciation of
both was raised for both F. and M. The result shows: (_a_) That the
percentages of colors and passive touches raised are practically the
same for both F. and M. (_b_) That the color displacement is greater
for M. than for F., being +.72+ for M. and only .40+ for F. (_c_) That
there is only a slight difference in the displacement of the passive
touches.

(4) When Tones and Active Touches were combined, the appreciation
of both was raised for both F. and M. It appears (_a_) that the
displacement for F. is very slight, .45+ and .23+, when compared
with the displacement for M., 1.71+ and .31+. But, (_b_) that the
displacement of tones is greater than that of the active touches for
both F. and M. (_c_) That this same relatively great displacement of
tones occurred in the opposite direction when colors and tones were
combined.

(5) When Tones and Passive Touches were combined, the appreciation of
both was raised for F,--the tones being raised more than the passive
touches. This combination was not tried with M. for lack of time.

From time to time some special tests of these general tendencies were
applied. From the results already set forth, one could predict that
there was a strong probability that when the tone-series was combined
with a constant touch-series, active or passive, the appreciation of
the tones would be raised. This was tried by allowing the subject M. to
rub his hand over the somewhat rough pillow of a tilting board. The
results showed that the appreciation of twenty-five out of twenty-seven
tones was raised. Other predictions were similarly verified.

Of course, any experiment of this nature is exposed to a great many
chances of error. The subjects may be fatigued, or depressed generally.
But the wide range of different readings taken is a reasonable
assurance that the chances of error are minimized. And it is to be
noted, also, that individual differences of appreciation do not vitiate
the results. In getting the "standards" no less than 2000 judgments
were given by each subject, while for the tables each subject gave
no less than 4000 judgments. The curves made from these data show
the effect of each separate combination by their variation from the
standard. The tables and analyses and conclusions already introduced
show, in general, that _our appreciation of each of several stimuli
in combination is different from our appreciation of the same stimuli
when taken separately_. The results show that this appreciation may be
either raised or lowered; that is to say, our feeling of values is not
constant for a given stimulus under all conditions.

From Table XIII I have found the average displacement of each series to
be as follows:

    Tones, .90.
    Colors, .37.
    A. Touches, .28.
    P. Touches, .23.

This shows that passive touches are subject to the least displacement,
while active touches, colors, and tones are respectively subject to
a greater variation. The sight-touch world is more stable than the
auditory world. With M. the tones go over a full point both below and
above the standard.

This report does not treat of the particular effects of a qualitative
nature that follow from the possible combinations of series of stimuli
of different feeling values,--such as the effect of an agreeable touch
upon a slightly disagreeable tone, or upon an indifferent color, and
so on. All such effects can be traced by rearranging the data already
collected, and this may be done in a subsequent paper which may enter
also into the theoretical discussion of the whole problem.

FOOTNOTE:

[Footnote 83: Attention may be called to the fact that this paper
arranges the conventional seven degrees of feelings in an order
opposite to that of the other papers of this volume; it follows still
the earlier traditions of our laboratory, while the more recent
investigations call very disagreeable 7 and very agreeable 1; the
indifference point remains the same.--EDITOR.]




THE COMBINATION OF FEELINGS

BY C. H. JOHNSTON


The problem at issue in the present investigation concerns the
combination of feelings. On the basis of theories as different in
many respects as those of Wundt, of Titchener, of Lipps, etc., the
feeling-state is always a unity. The affective process for Wundt
must be always "coextensive with consciousness." When he chooses to
speak of a "mixed feeling," it never for him signifies a "mosaic in
consciousness," but is always a new _Totalgefühl_, which "swamps
consciousness as a whole."

Titchener also believes that with every affective experience an
inevitable and pervasive "tilting" of the whole organism occurs. W.
McDougall, in his recent Physiological Psychology, makes the general
statement that always a "massive state of feeling results" when many
sensations are simultaneously excited, and that in such case we cannot
"introspectively distinguish the feeling-tone of each sensation."

We started to examine by experiment and by introspection which
feeling-effect really results from a combination of various impressions
with affective tone, and whether it is really impossible that various
feelings coexist and remain distinguishable. In case they can coexist,
the question arises: What mutual influences can be discovered?

For the main part of our study the simplest possible feelings were
chosen, because here presumably the subjects will not be forced to
grapple with complex personal psychoses, necessarily confusing from
their very richness. It was thought that here they could be more nearly
normal, naïve, less artificial, and able to a maximum to rid themselves
of preconceived personal opinions and unaccountable associations.
Here, with simplest stimulations, unsophisticated necessarily, the
hope at any rate is that work with a good number of subjects of
distinct emotional temperaments may bring to light certain fresh
simple introspective facts, which may in their turn offer valuable
considerations concerning the psychology of feeling.

Throughout the course of the experiment, except in the advanced stage
when more complex states were under consideration, sounds, colors,
odors, simple figures, and tactual surfaces were used. In the late
stage of the investigation sentences and pictures more or less morally
and æsthetically suggestive served to furnish for study the complex
feeling-states.

The progress of the investigation divides itself naturally into the
following four distinct parts.

I. From every experience of each individual the investigator sought to
obtain from the subject's own introspection at the time as adequate
a description as possible of the particular feeling provoked by the
chosen stimulus. The feelings studied in this first period of the
investigation are entirely those which heretofore have not been at
all classified except in terms of the objects which call them forth.
Part I is concerned with single stimuli affecting only one kind of
sense-organ, visual, tactual, auditory, or olfactory as the case may be.

Two requests were made of each subject, viz:

(_a_) To describe as clearly as possible how the particular experience
_felt_.

(_b_) To report always all the accompanying physiological or physical
processes which seemed to _mean_, to result from, or apparently
only accidentally to accompany the stimulus judged by him to have a
feeling-tone.

The work of the experiment covers a period of two years, and
fortunately several of the subjects were available for the whole
period. No subject was used for more than two hours each week. In the
preparatory training with sensations from only one sense-organ, the
range of colors, odors, etc., was chosen as follows: Twenty colors,
and as many tactual surfaces, etc., were presented in turn, and
each subject was requested to make his judgment as to the relative
degree of agreeableness or disagreeableness of the feelings arising
in the several cases. The scale of numbers from 1 to 7 served in
the traditional way to indicate approximately the hedonic value of
the feeling-tones, 1 signifying highest degree of pleasure, 2 very
pleasant, 3 slightly pleasant, 4 indifferent, 5 slightly unpleasant, 6
very unpleasant, and 7 the highest degree of unpleasantness. Though the
personal differences were in some cases rather striking, the individual
subject from day to day showed a relatively constant standard. This was
done in order simply to be able to choose approximately the stimulus
in the individual case likely to call up the _kind_ of feeling one
wished to study more in detail, and thus facilitate the progress of
the investigation. In this preliminary stage pleasant or unpleasant
seemed to the subjects more or less an exhaustive account of these
faint feelings. This was a means of eliminating practically indifferent
shades, as there is here no special interest in the psychology of color
as such.

       *       *       *       *       *

II. Following upon this preparatory training, the second part of the
experiment consisted in a similar study of the _mutual influences_
of _simultaneous feelings_ accompanying sensations from different
sense-organs. How does the feeling of pleasure obtained from contact
with a smooth surface influence the feeling occasioned by the _sight_
of a pleasant or unpleasant object? Here, for example, colors were
exposed in a large black frame manipulated by means of shutters easily
opened or closed, at the same time that a tactual surface was being
applied, or a tone from a tuning-fork was being sounded.

The introspection method was essentially the same here as in Part I.

(_a_) First, the subject was requested, without the necessary
distraction of directing his attention at all to the bodily processes,
to give himself up to the situation and to report as accurately as he
could the kind of affective state experienced.

(_b_) Next, as in Part I again, in a repetition of the same experience,
he was requested to be on the lookout for any and all accompanying
bodily changes. The problem here was to discover to what extent the
more complex state now in question would correspond to the specific
and noticeable bodily reactions such as were noted in Part I, where
single experiences presumably resulted. If different feeling-elements
are in experience at once, can one fix upon correspondingly different
suggested actions? Does the organism react to more than one situation,
or to two sources of stimulation at once? Is affection present only
when the whole organism is, to use Titchener's expression, "tilted"
one way? Is the Totalgefühl the single undifferentiated result
always, or can we here also detect such phenomena as summation,
fusion, inhibition, and partial or total mutual reënforcement of the
different feeling-components? Do the new reactions which seem to mean
the feelings always refer to actions so inclusive as to result in the
inhibition of any other tendencies to response, or is there sometimes a
clear strife between two simultaneously conflicting feelings, two kinds
of relatively self-dependent reactions both going on at once? Or again,
when the hedonic or algedonic characters of two given simultaneous
stimuli, such as a soft, soothing, pleasant touch with an irritating,
exhilarating, invigorating but also pleasant yellow color, do not
differ as to their pleasant-unpleasant character, must one be pale and
empty "intellectual perception" when the other is being enjoyed? These
are some of the questions that suggest themselves at once.

Not at this point, however, considering the dimensionality of feeling,
the four simple combinations were first studied; such, for example, as
(1) a pleasant color with a pleasant touch, (2) pleasant color with
unpleasant touch, (3) unpleasant color with pleasant touch, and (4)
unpleasant color with unpleasant touch.

       *       *       *       *       *

III. This part of the work was an attempt to estimate the average
time-interval in which feeling-tones develop, and what influences other
feelings given simultaneously or immediately beforehand have upon the
time-development of the feeling-tone in question. Will certain feelings
hasten and others retard a third feeling whose character remains
unchanged when it crosses the limen of awareness? Does the feeling, for
example, aroused by contact with a soft, soothing, yielding tactual
surface, put one in such a state that he will more or less quickly
obtain pleasure from the visual impression coming from a soft rich
red color? What effect will a feeling already aroused by a low tone
have upon the time development of the feeling one gets from looking
at a deep green color? Are there, again, pleasant feelings of certain
dimensions which will be hastened by other feelings, and still others
which, by the same means, will be retarded? If so, under what general
principle do they seem to fall?

Here the time-development of a certain feeling-tone is taken when
there are no other influencing factors. Then the comparison of this
rate is made with the later reported time-interval when that feeling,
again aroused, has been immediately preceded or accompanied by a
feeling-tone from another source of stimulation. Here also feelings
for colors presented in a frame, without any special suggestion of
form in connection with them, were in like manner compared as to
their time-development with the affective states arising from those
same colors presented again enclosed in cardboard frames of special
character. Some of these forms were very pleasing, such as upright
ovals, small circles, etc., while others, frames cut purposely into
irregular shapes, were to most observers decidedly unpleasant.

       *       *       *       *       *

IV. Here complex feeling-states were in question, and evidence was
sought as to how much could be detected here that would tend to
substantiate or to call in question what seemed to be the fundamental
principles of feeling-relations where the states are very simple.
Further complications, such as three and even more stimulations at
once, were tested. After this, feelings aroused by looking at pictures
of statues were studied and described as accurately as possible. "Perry
Pictures" were used. _Dying Alexander_, _Venus of Milo_, the _Dying
Gladiator_, the _Laocoön_ group, and _Apollo Belvedere_, served to
introduce sufficient variety. Then copies of these same statues were
cut out from the card and presented to the subject with the same colors
before studied used as background. These were allowed to play their
part in the feeling aroused therefrom.

After this, pictures more or less morally as well as æsthetically
suggestive were used. Millet's _Angelus_ and his _Shepherdess Knitting_
and Rosa Bonheur's _Horse Fair_ afforded suggestion hints as to the
contrasted motor significance of the complex states called forth. Here
the attempt was made to find out in how far the feeling when once
aroused is dependent upon the retained after-images or memory-images
of the original visual stimulation, and what sort of feelings tend the
longer to persist. Or again when both are taken in in quick succession,
what sort of imagery and associations result. Are the resulting
associations or images colored by both feeling-tones in any definite
way? And if the feeling itself persists despite the loss of imagery,
can it be referred merely to more internal sensations, or does there
seem to be a necessity to consider it of purely central origin?

Such, in brief outline, has been the proposed method of study. In an
experiment of this delicate nature there are clearly many things to
guard against. There is danger that the investigator will unwittingly
make suggestions to the subjects by his questions. There is a great
danger of auto-suggestion on the part of the subject. The likelihood
is also considerable that the subjects will fall into stereotyped
forms of expression and general listlessness in introspection, where
from week to week these simple experiences are being repeated for
closer and closer examination. Again the special mood of the day will
necessarily tend to affect all such feeling-attitudes toward slight
stimulations supposed to have a feeling-tone. These and other dangers
were recognized at the outset, and avoided as much as possible by
such legitimate variations as could be introduced without changing
the general purpose of the work. No subject was used when he felt,
for whatever reason, unable to adapt himself to the conditions of the
experiment. No subject knew anything of the recorded results of the
others, and it was constantly urged that each person should wholly
regard the present feeling in question, ignoring any remembered tone
which that special stimulation had before afforded him.

It very soon became evident that the variations among individuals,
especially as to the amount of feeling and the consequent ability to
fix upon the special physical processes involved, were considerable.
The subjects represent types. Hence, it seems necessary at once to
mention briefly some characteristics of the persons themselves who have
reported these various experiences. This was kept in mind throughout,
and seems of decided significance in the interpretation of the recorded
results. After an examination of the results of each individual,
whatever remains that is common to all will be briefly summarized.

All the subjects were graduate students in Harvard University or in
Radcliffe College. Seven of the twelve had had from one to five or more
years' training in laboratory investigations. Two subjects were ladies,
the rest were gentlemen.

Subject A was a man of bright cheerful pleasant even temperament,
responsive, very musical, alert, physically vigorous, very careful in
statement, and decided as to the distinctness of his emotional states.
He uses his facial muscles a good deal while conversing.

Subject B is musical, sings a good deal, is not especially
demonstrative, nor always able to become adapted to the necessarily
oft-repeated stimulations from the same colors and tones. This subject
is especially discriminating as to shades, and has decided preferences
for certain colors.

Subject C usually found it difficult to find any decided feeling-tone
for many of the stimuli used. This subject is rather reserved and
undemonstrative as a rule. He is not at all musical, nor does he care
for art. He is a rather cool but extremely careful observer, and is
always guarded in his introspection.

Subject D is impulsive habitually, flashy, responsive, especially to
any suggestion of an æsthetic nature, such as forms, and very decided
as to his experiences. He walks with a quick nervous step, is sprightly
always, vivacious in conversation and outspoken.

Subject E is rather non-emotional as he often says. He is very
energetic, full of life, quick but not precise in all his movements,
always on a tension, does not enjoy without effort anything so mild as
the stimulations here used, and finds introspection of this affective
nature difficult.

Subject F is careful, experienced in introspective work, musical, talks
a great deal, enjoys this kind of work, has decided preferences, is
athletic and energetic. This subject makes use of facial, arm, and
shoulder gestures quite freely in general conversation.

Subject G has a penchant for talking a great deal, is decided in his
likes and dislikes, musical, of an uneven temperament, sometimes
cheerful, often cross, but always animated.

Subject H confesses he does not ever especially enjoy colors, nor
respond with any sign of demonstration to any situations. He is steady,
calm, apparently unruffled, and not an especially acute observer of his
own states, proving in this experiment unusually undiscriminative as to
simple experiences.

Subject I is rather morose, claiming to be habitually unmoved by even
display of great passion or excitement. He finds it generally much
easier to call up unpleasant than pleasant experiences, this being
exceptional among the subjects. He is much slower than the average, and
his feelings are not easily aroused. He is deliberative and confident
as to his state of mind. He is nervous and often becomes fatigued
before the hour's introspective work is over.

Subject J is nervous, of an uneven temperament, emotional, and quick
to react to a situation of any kind, and, rather more than the others,
subject to suggestion.

Subject K responds very quickly always, is habitually prompt and clear
in statement, of an even temperament, and unusually interested in the
experiment.

Subject L is unexperienced in this particular kind of work, but slow
and careful. Though athletic, his movements are rather heavy. He is
deliberate in speech and of an even, though rather undemonstrative
temperament. He also is musical.

In order to verify my somewhat personal descriptions here recorded a
questionnaire was given each subject to fill out according to his own
personal judgment of his emotional disposition. This was done toward
the close of the investigation, and the answers agree in the main with
the descriptions offered above.

For the first month's preliminary practice, and with the purpose of
stimulating curiosity and interest, and of testing the comparative
richness of even slight feeling-experiences, a great variety of
stimulations was used. Twenty different tactual surfaces from softest
plush to very rough sandpaper served for the tactual impressions.
Twelve different odors, as many colors differing in saturation and
intensity, and tones from high and low tuning-forks, and noises
variously produced, were employed as stimulations for the other senses.
Besides these, circles, upright and horizontal ovals of various sizes,
imperfect circles and ovals, and other irregular shapes were all
presented in the same large black frame. When studied alone indifferent
gray fillings were used. When complex states were in question colors
served as fillings. When the subjects thus became accustomed to these
very simple but very definitely _felt_ experiences, in these for the
most part habitually ignored affective elements of ordinary sensations,
the investigation at once became narrowed to more careful and minute
attention to a few of these feeling-tones. It was soon found also that
odors could not easily be used in combination, since they effectively
effaced all feeling-tones for the simultaneously given colors or
touches. Five colors, fairly representative for all subjects of
different kinds of feeling-tones, were chosen, and were used throughout
the whole investigation. These were the following: a soft deep red,
light brilliant yellow, deep pure green, saturated blue, and a dingy
greenish-yellow. The dimensions of the exposed surfaces were six by six
inches.

For tactual impressions of approximately equal value soft plush,
velvet, and two kinds of sandpaper were used, and for tones high and
low tuning-forks. All the above-named forms were used in connection
with the chosen colors. The subjects differed considerably as to the
_amount_ of feeling that could be obtained from such material. The
variation of kinds and of intensity in the same subject was sometimes
noticeable from day to day, but not great. It is hardly necessary to
give detailed quotations from each subject. The following summary of
the results of the experimental work, however, contains nothing that
was not frequently reported by a majority of the subjects. This, then,
does not represent at all what was once or occasionally reported by
individual subjects, but what after training seemed to be reliable and
definite and constant feeling-states.


PART I

_Section A._ The following are the collected expressions which many
subjects used to describe the feeling for this _particular shade of
red_. It feels as if it would be soft. It suggests warmth. The feeling
is one of seriousness, pleasantness, quietness, of free repose,--a
full feeling of the sense of safety. It is soothing, rich, full of
strength, and inviting. One feels restful, grave, calm, appeased. There
is an agreeable longing and a tendency to lose one's self in the color.
The feeling is one of comfort, luxury, satisfaction, expansiveness,
tranquillity, and quiescence, with no accompanying feeling of weakness
by exertion of effort or energy. There is neither marked tension on the
one hand, nor collapse on the other. There is a sense rather of easy
self-control and command of one's body, but with no aggressive sorrow
nor joy element,--a feeling of being attracted, with nothing to suggest
any obstacle to the adaptation.

Occasionally to all subjects this color, and, indeed, all colors,
seemed "dead," arousing no feeling whatever. Here the color "ought
to be pleasant," but is only "for the time potentially not actually
actively pleasant." Still more rarely did this red appear to be
unpleasant. Some subjects thought that this afforded the greatest
_amount of sensual pleasure_. More than any of the other colors they
think it appears to "give you something." It does not so much stimulate
as furnish a content itself. It has a direct effect rather than a
tendency to make one wish to do something and thus give pleasure
from the activity itself. Only one subject failed to find this color
pleasant. His early association of it with blood and ghastly scenes
could not be overcome. Some others, when a glare or glaze appeared on
the red, found in it slight suggestions of stimulation and excitement,
but the general decision in the great majority of cases was that the
feeling was a sort of emotional massiveness compared with the effects
from other colors.

In marked contrast, for the most part, appears the characteristic
feeling-tone for our chosen shade of _yellow_. Almost universally
subjects find such words as these descriptive of the feeling here
in question. It is cheerful, brisk, pleasant[84] also, bright, gay,
light, sprightly, merry, jovial, easy to get, pleasantly irritating,
stimulating, stirring, spurring, thrilling, invigorating, and produces
agreeable discontent. It is jolly, nice, trim, neat, awakening, full of
the sense of motion, soaring, and arouses a feeling of welcome strain,
of pleasure in action, of alertness and self-assertion. Here, in
contrast for the most part to the red, there is no feeling of sinking
into the color. The impulse rather is to be free, to enjoy motor
expression, even if of some vague sort. There is a _felt_ necessity to
do something, a "joy of overflowing or of exuberance," it is called.
There is little present here of what we mean by a suggestion of sensual
richness found above in the feeling for red. Here there is less of
amount of pleasure, but much more of the general activity element. Some
subjects feel the demand for greater saturation, and occasionally it is
unpleasant for just this reason apparently. Subjects C and B frequently
reported this. They think the feeling would be more "stable" and
"grave" and "secure" and "soothing" and one would not feel "unruffled,"
if it could be "toned down." Most of the subjects, however, think that
it belongs to the ultimate elemental feeling for yellow that it should
have just this distinguishing characteristic.

It is more difficult to describe the feeling for _green_. It is almost
always agreeable. Two subjects, however, never like it. Sometimes it
is somewhat soothing in character, but more often it is exciting.
The feeling seems to be between that for red and that for yellow,
partaking on the whole of the characters of feeling for the latter
rather than the former. For all subjects associations tend to color
the feeling-tone for green especially, and hence introspection for
the feeling of pure color is doubly difficult. The most prominent
partial feeling-tone for it is "irritating."[85] The agreeableness
or disagreeableness of this stimulating character is particularly
inconstant, varying greatly for the same subject, as well as for
different subjects.

The feeling for the _blue_ seems still more to be dependent upon
the person. Many like it. Many others dislike it decidedly. When it
affords a pleasant feeling, it is described in some such terms as
these: The feeling is spiritual, lofty, beautiful, serene. The subject
himself feels immoveable. To other subjects it is too rich and intense
and painful. To one subject who heartily dislikes it always, it is
offensive or revolting, calling up a feeling akin to the emotion one
has toward insincerity in general. To none does this feeling seem
to have any great amount of sensual significance. Even when it is
called "too rich," the incongruity between the richness itself and the
ultimate qualitative significance of the blue is spoken of. Even when
pleasant, the feeling is of an "airy pleasure," volatile, unstable, and
not reliable, nor safe and secure as is the feeling for red. One feels
that it is always apt to vanish, vague, intangible, and with little
immediate definiteness of meaning. Subjects often desire to call it an
intellectual, æsthetic, or ideal sort of feeling.

No color was universally unpleasant. Two subjects found this
_greenish-yellow_ almost always mildly pleasant. For most of the
subjects, however, it was unpleasant. Here were reported feelings of
contraction, of withdrawal, of disgust, of doubt, of hesitation, of
stimulation without definiteness, dissatisfaction, slight feeling
of nausea, of sea-sickness, of opposition, and the general feeling
of offensiveness. The necessary, unpleasant aggressiveness, unrest,
or discontent characterizes this feeling. This unpleasant critical
attitude where a decision is wanted but not easily gotten, is called
often the feeling of uncertainty.

In no sense is this investigation a study of the psychology of color;
the only purpose here is to find certain clearly defined feelings
for slight stimulations, in order to find in what way they relate
themselves to other similarly simple feelings from a different source
of stimulation.

In a similar manner, then, the investigation was conducted in the
analysis and description of feeling-tones for _tactual impressions_.

For _plush_ there was a feeling of pleasure, ease, safety, and content.
The mood was one of a general enjoyment of sinking one's self into
the situation, an agreeable self-surrender. Here also is a feeling
of unbending one's self, of general expansiveness, of relaxation.
One is soothed, enjoys a suggestion of freedom from disturbance, of
a "regularity" of the experience, feels at the same time strength in
the suggested repose, responds to pleasant reverberating thrills by
the falling off from the accustomed muscular tonicity, and hence has a
decided feeling of satisfaction. To some subjects the feeling aroused
by the hard, polished, glazed tin surface, possessing no "yielding"
character, corresponded more nearly to the feeling for the yellow
color than for the red. To all red "went best" with the plush. No
tactual feelings offered such distinguishable elements for analysis,
nor were they as definitely described as the visual or olfactory or
auditory impressions. The sensational elements were in many cases more
pronounced. The feeling for the plush, however, much like that for
the red color, suggests a "settling down to," or a "dropping forward
toward," rather than an aggressive "taking in" of the feeling-material.

The feeling-tone with sensations from sandpaper is grating, irritating,
stirring, stimulating. The feeling is one of contraction, of
withdrawal, of uneasiness. One is full of "collapsing chills," of
minute little pains, and there is a decided call for an opposite kind
of behavior. The sense of weakness, of waste of power and energy, of
being penetrated, of strained expectation, of unwelcome tension, and of
slight "wasteful excitement" results. To some subjects, notably subject
E, at times the whole feeling of stimulation as such predominated,
and the total effect produced was agreeable, as it "satisfied a felt
need of waking up." Here again one subject, subject B, throughout the
whole period of two years, failed to find any element of pleasure in
any tactual sensation that was pronounced or prolonged sufficiently to
furnish material for introspection.

As regards simple _tones_ from tuning-forks the subjects find little
to say. All are pleasant, as a rule, and almost universally, _low
tones_ are most pleasant, richer in content, greater in amount of
"general appeal," more soothing, and pleasantly stimulating. The
feeling of the easy attitude called for contributes to the whole
feeling. _High tones_, calling for more activity on the part of the
subject, more strain, and greater stimulation, coupled with some rather
unprepared-for irritating elements, are less pleasant, and also more
limited in their general appeal to the whole organism. The noises
variously produced were at first unpleasant, and the only assignable
reason seemed to be that their suddenness came as a shock. If expected
or continued they too became pleasant very often.

Feelings for _forms_ seem to relate even more definitely to the
activity element. The pleasure for the most part is described as
being far less sensual, if indeed, so at all. Small _upright ovals_,
1-1/2 × 1 in., are most pleasant, because somehow they are "more
suggestive of definiteness." _Circles_ one inch in diameter are next
in order of value as to their feeling-tones. _Horizontal ovals_ are
less pleasant still, though for most subjects not unpleasant. Upright
ovals are best, as the kind of action apparently called for by the
aroused feeling is most agreeable and suitable to the subject's
natural upright position of body. An explanation of this general
result of introspection, as well as the preference for the particular
size chosen almost without exception, is attempted in another part
of this report, where are given in more detail the various kinds of
bodily accompaniments. The feelings for those ovals have also the
characters of stimulation, mild excitement, and a feeling of easy
freedom in a pleasing kind of activity. Tension is always present as
an agreeable element when reported at all. This element is coupled
with the "feeling of assurance of certainty" which the whole situation
calls for. It often seems clearly to suggest that one do something.
Circles tend more to suggest inner stability and completeness. They
stand on their own axes. Here there is a sense of satisfaction,
complacency, and sufficiency. The feeling here of a call for immediate
activity on the part of the subject is weak and indefinite, when not
altogether absent. The subjects do not use for this experience such
expressions as excitement, tension, irritation, quick contraction, or
the impulse to self-assertion. Horizontal ovals are least pleasant, it
seems to me, for obvious reasons. Here such noted elements as "felt
unnaturalness," "difficulty of adapting one's self," "wrong direction
of activity," which alone and in themselves would be unpleasant, are
nevertheless more than counterbalanced by other and pleasing elements,
such as symmetry, definiteness, partial stability, and other agreeable
features. Often these latter features are not pronounced, and then the
judgment is, that the total feeling is unpleasant.

Likewise as regards so-called bad forms, no single statement is
unqualifiedly true of any considerable number of subjects. The
decided feeling of irregularity, the "bulging-out" or the undesirable
"pushing-in" of the figure, the feeling of weakness in one's own body
corresponding, the feeling of instability which one tends himself to
imitate in various ways, the total effect of lack of poise, all tend
to make these figures on the whole unpleasant. But one cannot even
here count upon the constancy of the subjects' feelings. At times, due
perhaps to undercurrents of association processes of which even the
subject himself is not clearly aware, the figure suddenly looms up as
quite definitely pleasing, and full of vague suggestiveness and hidden
richness of content. These varying characters of the feelings for
forms come out interestingly later in the study of them when they are
presented as frames for the above described colors.

       *       *       *       *       *

_Section B._ The bodily processes noted by the subjects are numerous,
and here also, just as with the amount of feeling above, the personal
differences are striking. Some subjects detect a great many forms of
organic commotion, others rarely find anything that can be said to be
descriptive or explanatory of the feeling-state. To all of them at
first this looking for bodily accompaniments destroyed the feeling
itself. Only after considerable training was it possible for them to
find any physiological processes that seemed at all significant. As a
general statement the evidence would all tend to suggest that feelings
for color are most readily and directly referred to the head, face,
throat, and particularly to the forehead and to the eye-muscles. When,
however, the feelings are particularly strong, they tend to pervade the
whole organism. Red thus often brings about the suggestion of general
bodily comfort, and yellow, when very strong, arouses the impulses
calling for "spreading-out, aggressive movements," referred to arms,
shoulders, and chest. Tones have in general the same reference to
the head. Odors are always more organic, affecting more directly the
respiration, muscles of the abdomen, and the more internal apparatus
generally. Tactual impressions refer to the trunk rather than to
the face, hand, arms, or legs. Forms seem to call forth imitative
movements, and the actual or incipient motor impulses refer to the
action of the eyes in motion, the position of the head, of the whole
body, of the shoulders particularly, of the shaping of the cheeks,
lips, etc., and of the similarly imitative actions in the hands and
arms. The following is a list collected from the reported bodily
references given for the feelings described in Part I, Section A.

(1) Free full respiration and free activity of all voluntary muscles;
or, for other feelings, the checking of respiration and often the lack
of impulse to move at all, with no suggestion, however, in most cases,
of lassitude.

(2) Chest expansion and general relief pervading the whole body.
The expansion or contraction is further modified by the _degree_ of
_regularity_ and by the _rate_ of the movements involved, as also by
the ease or _difficulty_ in the performance. So also, in the cases of
feeling whose tone exists but is doubtful in character, the bodily
situation seems to mean "lack of movement or change in any definite
direction." The feeling-tone and its vividness are interdependent and
reported as closely connected.

(3) A cringing all over and a "holding up of all activities."

(4) Abdomen contraction, chest and shoulders drawn in, hands clenched,
and jaws set.

(5) A feeling at once in different parts of the body of both process of
contraction and expansion.

(6) An incipient feeling of nausea in the digestive tract.

(7) A tendency to incline the head forward or backward, or to keep it
rigid, or to turn it aside.

(8) For touch, waves, reverberations, pleasant penetrating thrills
in the chest and abdomen especially, less frequently in the limbs,
occur. Sometimes these suggest expansion of the whole frame; sometimes,
even when also pleasant, the tendency to contraction and tension is
noticeable, but in these latter cases the contraction seems to be
rather definitely the calling into action of those general innervated
muscles which refer to the bodily situation of one when he intends to
go toward the pleasantly stimulating object.

(9) For unpleasant touch the reference or localization of the bodily
response is, when reported definitely at all, generally in the back,
described as chills not thrills, contractions always, contractions also
which often suggest shivers of withdrawal. These feelings also are
referred to the situation of the trunk of the body, and are felt to
originate in the small of the back, and in the back of the shoulders.
For two subjects there occur twitchings in the tendons of the hips and
thighs, and movements of the knee-cap.

(10) The pervasive bodily collapse, which seems to accompany feelings
characterized as depressing, altogether unlike the soothing feeling of
unwearied repose given by certain soft rich colors or by low deep full
tones or smooth yielding surfaces, is another form of organic response
which is often spoken of.

(11) The direction of the stimulus with respect to the normal position
of the body also seems to have something to do with the regularity
of the response, and with the general forward or backward tendency.
Tactual surfaces applied or tones sounded behind the subject do seem
to make the bodily adjustment more confused, and less pleasant. All
that subjects could say was that the position was felt as abnormal and
correspondingly less pleasing.

(12) In many unpleasant feelings, where there was no specific
localization possible, the "stiffening tendency of hardening one's self
to a necessary experience" was frequently reported. In the case of
other states of undifferentiated pleasure a "consenting bending forward
of the whole body" was often detected.

(13) Many stimulations seem to demand that one draw one's self erect,
square the shoulders, and "assume the attitude of alertness."

(14) Certain colors for almost every subject independently hint at
sea-sickness. Others, as noted above, report the incipient suggestion
of nausea in the digestive tract. Indeed, abdominal references are
frequently reported by most of the subjects. The abdominal muscles
become "eased up," or again there is a "sucking-in of the belly."

(15) The feeling of "being natural," of regularity, a universally
popular feeling, is described as a pleasant relief from all tensions
and habitual inhibitions, or a dropping of one's characteristic
muscular tonicity.

(16) Other stimulations still, particularly certain delicate odors, for
men, subjects C and E for example, seem to suggest what they call the
"childish play impulse." They are called "simple, foolish, childish
pleasures," ignored in ordinary life. They are slightly pleasantly
irritating, and merely make one wish to do something. It is pure
bodily restlessness, a general kinæsthetic enjoyment. Three subjects,
especially, find here the frequent twitchings in the calves of the
legs, in the knee-cap, and the more decided innervations which contract
the tendons of the thighs and hips.

(17) Subject I frequently detected sensations of contraction in the
tensor tympani connected with the pleasure derived from high tones.
Others referred feelings for tones partly to the regions of the ears.

(18) The kinds of facial references are numerous. General contraction
or expansion around the eyes, forehead, temples, sometimes to the whole
head, and quite frequently it seemed as if the feeling referred to the
very inside of the eyeball, to the iris and accommodation movements.

(19) Subjects A, D, F, and K noted specific incipient tendencies to
smile, to smooth the brow, and to "unbend the face" as characteristic
descriptions of certain oft-repeated experiences.

(20) Introspections from subjects F and G quite constantly revealed
articulatory impulses vividly accompanying the feelings for many colors
and forms.

(21) A scowl and puckering of the lips was descriptive of the attitude
taken toward some unpleasant situations.

(22) A contraction or relaxation of the throat-muscles and of the vocal
chords generally was not infrequently noticed. The tendency to swallow
is spoken of. The throat is felt often to be "concave" when certain
bad feelings are sufficiently pronounced. A contraction in the mucous
membrane, with teeth on edge, such as one would experience in eating
something sour, is frequent. A twitching of the ears, squinting of the
eyebrows, and a "heavy feeling" through the neck and chest occur often,
or again a pressing hard of the tongue against the roof of the mouth.

(23) Forms suggested a shrinking in the volume of the face, sometimes
of the crown of the head, and even of the whole head.

(24) Upright or horizontal ovals especially provoked the impulse to
imitate the figure itself, either with the lips or with the hands and
arms. When the feeling was particularly strong, all these impulses
often occurred together and appeared mutually to reënforce, or to
intensify each other.

(25) Horizontal ovals gave one the feeling of being "flattened out,"
coupled with an impulse to adjustment altogether unlike the sprightly,
alert, airy feeling aroused by the "trim," upright figures.

(26) Occasionally when the irregular shapes were presented directly
after a subject had been enjoying one of the perfect figures, that
side of his face or body corresponding to the distorted portion of
the figure was felt to be in an abnormal and unpleasant position.
This "caving-in" or "bulging-out" sensation, which accompanies the
unpleasant feeling, happens when the whole muscular system at the time
for the subject seems inert or externally controlled.

All these sensations of bodily processes, taken from the introspective
descriptions given by the subjects, are distinctly reported by them
as very faint. They by no means detect them in every experience, nor
do they always seem to the subject himself to _mean_ the whole of
the feeling as experienced. Neither did any one subject find all the
concomitant processes recorded above. Subject H failed throughout the
whole period to detect anything whatsoever, except slight tendencies
to frown, smooth the brow, or to open wide the eyes. This subject was
unable to detect any special differences in his feelings, either in
variety or in amount. For him neither soft red nor brilliant yellow
was either exciting or soothing. They were and always remained for him
more or less vaguely pleasant, and this description for him was both
ultimate and exhaustive.

Subject B could get no kind of pleasant feeling from any tactual
surface, while to Subject E even the coarsest sandpaper usually
afforded pleasant stimulation. As spoken of above, articulatory
impulses were characteristic of the motor tendencies of Subjects D
and G. To Subject A the experiences seemed richest and fullest, and
the corresponding bodily processes were likewise more pronounced and
varied. In the great majority of the experiments, especially during
the period of training, the feeling itself vanished when the subjects
attempted to analyze the bodily processes. It was chiefly, however, a
matter of training, and this more and more ceased to be a disturbing
element.

Some subjects preferred often to speak of circulatory, or at least,
decidedly internal and usually involuntary changes in addition to, and
sometimes without, the controlled muscular actions. The mood of the
time affects the amount of feeling, and occasionally, but far less
frequently, the quality. The moral significance of the feelings was
most prominent when the subject felt most interested in the experiment,
as may be noted above in their descriptions of the feelings for red
and yellow. What may be termed the "_regularity element_" would seem
generally to serve as the test especially for the pleasant-unpleasant
character of the feeling-tone. The feeling of expansiveness never
accompanied unpleasant feelings. Feelings of contraction, on the
other hand, very often occurred when the feelings were not at all
disagreeable. In such cases there was a significance attached to the
direction or meaning of the adjustment.


PART II

_Section A._ Here simultaneous stimulations of different sense-organs
were given, and the situation became at once more complex. For some
time only colors and tactual surfaces were employed. Later tones from
tuning-forks and noises were added. Forms with different colors as
fillings still further complicated the experience. Odors as a rule were
unsatisfactory, being so strong as entirely to inhibit all noticeable
effects from the other senses.

For all subjects at first the feeling-tone related only to the one
object directly attended to. Some effort is required to detect the
feeling-tone for these slight stimulations, and while this is being
done, the feeling for the other sensation tends to vanish. If, while
enjoying the soothing contact with the plush, a chosen color is
disclosed in the frame and attended to sufficiently to obtain from it
a decided feeling, there is a distinct awareness of the dropping of
the feeling for the touch. To some subjects, whatever the combinations
used, this almost constantly occurred for perhaps a month. Often again
there seemed to result a total "feeling of the situation," when the
attention was on neither stimulating object.

Frequently, too, the attempted introspection at this point failed to
fix upon any feeling-tone at all definite. The condition was one of
confusion and bewilderment. The state of mind when one cannot feel at
all definitely seems to correspond closely to that state of mental
confusion when thought processes are in a jumble, with no path for
the moment leading anywhere. All these difficulties were overcome,
partially at any rate, by continued training. It was not as if
introspection revealed the fact that there was nothing to be found, and
this was frequently reported by the subjects. After some time the touch
character could be retained, and its peculiar value for feeling did
not disappear when other things came in and contributed an affective
element of their own. The old law of the opposition, or mutual
exclusiveness, of feelings would thus seem to mean little more than
that we, generally speaking, experience one thing at a time. Without
a special analytical purpose in view, we do not find many distinct
elemental feelings, as we do not, until we psychologize, find elements
of cognitive character separable. It has been, and is now, commonly
supposed that myriads of ideational elements, partially analyzable at
any rate, go to make up what we choose to call a single perception.
This experience as a whole is of some affective nature; but, as
generally stated, of one unanalyzable sort always. It is true just in
the same sense as in the cognitive state, perception. In the sense
that every perception is unique, in this sense every affective state
is likewise a unit. The evidence I submit, however, is that one may be
the subject of analysis into elemental parts just as much as the other.
Affection, as Titchener defines it at the beginning of his treatment
of feelings, is merely a "tilt of the whole organism." If this is the
ultimate statement, then there are no combinations, and no relations of
feelings except that of mutual exclusion from the field of awareness.
He has taken only one of the above possible attitudes toward affective
states. Geiger, in his study of very complex emotions, however, has
taken the other attitude, and bases his whole position upon it. This
present experimental test furnishes evidence that the latter position
is also a legitimate, and perhaps more desirable position, if feeling
shall have scientific analytical treatment.

In this investigation, after considerable training, the subjects,
with a single exception, _were all convinced that both feeling-tones,
for tactual and visual impressions, could be present at once_. When
three or more were given at once, confusion as to the state of feeling
was usually so great that valuable introspection was always rendered
exceedingly difficult. Impressions from the same field, as, for
example, colors presented in chosen forms as enclosures, were most
often taken as one object with one feeling-tone. This even was by no
means always the case. When it was thus taken, the experience was still
reported as more complex than either element alone had produced.

When the feeling-tones for simultaneous stimulations from two
different sources came out sufficiently clearly, the kind of feeling
was described in some such terms as indicated in Part I, Section A,
except that almost invariably the introspection was more difficult.
The relations of these various feeling-components of an affective
experience are numerous. There is a frequent tendency to read one
into the other. The soft soothing feeling coming from plush, if in
the particular experience the color be the more prominent partial
element, tends often to make the subject enjoy more the color, because
there seems to be added to it a soft yielding surface texture.
Frequently also, as in the case of red above, the warmth it suggests is
intensified.

In cases of feelings of opposite nature occurring together, the
stronger generally prevails, finally in most cases effacing all
specific tone for the weaker element. An odor, for example, even when
always unpleasant, becomes less so when one looks at a pleasant color,
when a feeling-tone can, or often even when it cannot, be detected for
the color at the time. Again, when a very unpleasant form or tactual
impression is being felt, a slightly unpleasant color tends to arouse
often in this situation, as if by contrast, a simultaneously pleasant
element in the total experience.

For many subjects frequently there results what I shall call a "Total
Mood." This, as to its feeling-character, can be merely different
from, more than, less than, or the same as either component or of
both together. To some the feeling is proportionate to the degree of
concentration of attention, and in all such cases rarely does the
whole complex situation afford a feeling equal to that given by either
component alone, the extra stimulation for the time being simply a
disturbing factor. To others the shifting of the focus of attention
from one to another of the external objects of interest, or from one
feeling-element to the other, is not at all disturbing, "any more
than is any general state of satisfied self-contemplation." This kind
of experience is often and distinctly reported, not as the enjoyment
of two where the discernible elements persist wholly unrelated, but
rather an enjoyment (or disagreeable experience as the case may be),
simply from two sources of stimulation, a total mood with similar or
harmonious constituents. The red color and the tactual feeling for
plush afford this. Similarly the unpleasant color above combines with
certain odors or with the sandpaper. Yellow, however, does not as a
rule produce a feeling that peaceably "falls in with" the tactual
impression brought about by the plush. Low tones tend to combine thus
with the red color or with the softest plush in the same kind of Total
Mood. The feeling-tones usually for pleasant high tones are described
as "falling in with" the feeling for yellow when the feeling exists as
described above, and as nearer to that of the feeling for the green
color than for the particular deep shade of red. What may be termed
the "_Congruity_ or _Incongruity of Feeling-Tones_" is perhaps a good
name to designate feeling-tone relations. It implies neither mutual
exclusiveness nor total fusion, and some such term is necessary.

The various phenomena of fusion, summation, partial reënforcement,
merely simultaneous, independent coexistence, partial and total
inhibition, of one by the other occur. The feeling-tone for yellow
tends most readily to fuse with the feeling-tone for high tones and
upright ovals. This is not so marked for the green, but more so for all
other colors than for the red. Red harmonizes and tends to fuse, for
most subjects, with the feeling-tones for soft plush, low tones, and
circular forms. This harmonizing, however, is not all that contributes
to the amount of feeling in these complex cases. Subjects often prefer
the low tones with yellow, even though there is less harmony. So
also upright ovals are in themselves generally so much more pleasant
than the circles that red is preferred thus presented, though its
feeling-character is more akin to that suggested by the circles. These
are cases where the intensity itself of the feeling-tone is preferred,
even though what is felt to be an harmonious combination is lessened.

When the situation admits of a complete fusion, the one resulting
feeling is almost always greater. When summation of unpleasant stimuli
occurs, the singleness of the attention process is not a prominent
feature of the experience. Rather each unpleasant element exists
throughout, each in turn intensifying the whole undertone of feeling,
but also remaining a feeling-tone of a particular kind. Partial
reënforcement is descriptive of that state when both feeling-tones
contribute to a feeling of the same kind, yet do retain some individual
characteristics which stand out for themselves. The general state of
pleasantness, for example, is increased by both elements contributed
by a low tone and the yellow color, yet one retains its soothing and
the other its exciting character. Again, the feeling-tone for green
may occur when its relation, on the other hand, to a pleasantly
sounding tuning-fork is not at all noticed. Subjects find in such
cases always more effort required to note both the feeling-tones, and
there is probably some diminution in quantity of feeling for each of
the simultaneous elements. Other subjects have preferred to call this
partial inhibition. Cases of total inhibition have been noted above,
and are by far the most frequent, as would naturally be expected. When
sandpaper is being applied, and no repose is felt in the body, a color,
suddenly presented, for a moment pleases the eye, but quickly loses all
feeling-character, and can only be "intellectually perceived."

Again, the way in which subjects will take certain combinations seems
to depend entirely upon the person. Beautiful colors, presented in
disagreeable forms, bring about for some a feeling altogether worse
than does an unpleasant color in the same form. To others there is
always the tendency to enjoy the color and to "reconstruct" the form,
or stress in it those elements only which do suggest symmetry and
definiteness. All feel, when two or more elements contribute to the
feeling-experience, that a total mood generally serves as the undertone
for them. When there is a clear strife between the two, they both can
exist as equal partial tones with an undertone of unpleasantness in the
failure to coördinate them. There are still other cases where the total
result cannot well be called a fusion or summation. For example, when
an unpleasant color in an unpleasant form, or for Subject D, a pleasant
color in an unpleasant form, is presented, the feeling for the whole
is often out of all proportion to the value of each alone, or of what
might be expected from the simple summation. The uncommon revulsion
here was frequently so striking that the subjects would afterwards
laugh heartily over the strength with which it first appeared.

       *       *       *       *       *

_Section B._ Introspection here as to the physiological accompaniments
referring exclusively to one of the two or more existing feeling-tones
is still more meagre, but at times very definite. When the elements of
a total feeling fuse there is of course no reference to the particular
processes which bring this about. It is then simply a general response
to a situation. _When, however, distinct, or opposing feeling-tones
are present and detected, they do often mean opposing inclinations
to action._ The yellow color can retain its exciting tone, and refer
clearly to such activities as opening wide the eyes, incipient
smiling tendencies, and general alertness of facial expression, when
a soothing touch is also felt as suggesting a toning-down of the body
and a general relaxation of the muscles of the abdomen. This is the
most frequently noticed effect. Tactual impressions are accompanied
by pervading organic feelings in the trunk, while visual and auditory
stimulations, in their incipient stages, at least, have the more
pronounced effect upon facial muscles of expression, and general
sensations in the head. When any of these feelings are particularly
strong, however, the sensations, whose feeling-tones seem to constitute
the feeling in question, tend to pervade the entire system and to usurp
the whole bodily activity. The motor tendencies noticed above for
the irregular forms are also reported when the color itself remains
pleasant. Yellow, possessing more of this activity itself, is least
pleasant when exposed in these forms. The opposition of tendencies is
noticed, yellow meaning its own peculiar kind of aggressive movement,
and the bad form at the same time calling for that irregular kind
of unpleasant adjustment. Red does not "intrude itself" nor demand
action, and is always less strikingly in opposition to the form than is
the case with yellow or green or blue. Forms, almost perfect, relate
themselves to feelings of tension. One feels that he cannot quite take
them as perfect figures, and this strain and inability to take them
for what they suggest provokes a decidedly unpleasant feeling. Very
irregular forms become "grotesque" or ludicrous, and the bodily change
is indicated as a "jumble of partially carried out reactions."

In many cases sensations or motor tendencies are noted all over the
body during the existence of these complex states. At such times
they are not recognized as referring to either feeling-tone in
particular. When also a favorite color is presented to a subject who
is experiencing a disagreeable feeling from sandpaper, the touch is so
pervasive usually that he feels that this "controls the whole response"
and inhibits any reaction, or even any suggested reaction to the
color. When there does fail to be even any possible incipient motor
suggestion, as a rule the feeling-tone for the object is extremely
vague if it exists at all, and the object appears for the time "dead"
or "valueless." Subjects speak of their own inability to respond
in such cases. It is not at all as if the color is definitely bad,
but rather as if one cannot do two different things with the same
muscular apparatus at once. As often, as has before been noted, does
the opposite occur. Colors in definitely characterized forms illustrate
the relations of similar activities when feeling-tones occur together.
Yellow is preferred in upright ovals, for both accentuate the same
demand for activity, and calling for the same kind of response, tend to
fuse into a single object. Yellow and plush do not harmonize, and in
many cases where both retain their feeling-tones, distinct activities
in different parts of the body are aroused simultaneously. With the
circles the feeling-tone for yellow does not agree with that for form.
The yellow becomes almost unpleasant at times. Circles are "heavy,"
"stable," "on their own axes." A yellow thus enclosed seems "too fat,"
too "unnaturally heavy," not free and light, and the effect is less
pleasing. Circles suit the red better than they do the yellow or the
green or the blue, and tend to be seen as one object, or to fuse, more
readily than red in an upright oval form. The feeling-tones for red and
for upright ovals are both very pleasant, but not as much in harmony,
and consequently usually taken as two different feelings.

As a general result of introspective analysis at this point, when
different feeling-tones did occur together, they were described in
terms similar to those used when each alone was experienced. The
bodily references, when found, were of the same character, the only
difference being that there was much difficulty in determining to which
feeling-tone the response referred. In many instances, however, again
it seemed quite certain that different kinds of adaptation in different
parts of the body were suggested which seemed to correspond to distinct
affective qualities. Also distinct feeling-tones, each of which alone
could call forth a similar kind of action, when given together tended
to accentuate the total unified response. The upright ovals mean
alertness and soaring motions with a general suggestion of drawing the
shoulders up. The yellow color accentuated this. The circle with the
soothing red, or the fusion of feelings for red and plush, pleases in
quite another fashion.


PART III

In attempting to measure the rate at which feeling-tones for
those slight stimulations develop when no disturbing factors are
consciously present, an interval of from one and one half to two and
one half seconds seemed to be required. At such a time the feeling
was experienced as having reached its maximum. There was no marked
difference for different subjects, nor any constantly noticeable
difference among the kind of stimulations used. A possible exception
was found for Subject I, but this was probably due, as he himself
thought, to his inability to adapt himself easily to the requirement of
the experiment.

After this was sufficiently tested, the interval which was required
for one feeling-tone to arise when another was already present, was in
the same manner tested. The interval in all cases was too long to be
measured by means of a chronometer. A stop-watch was used.

While the subject was consciously enjoying a sound from a tuning-fork
or a tactual impression from some chosen texture surface, one of the
colors was presented to him. The time-interval thus ascertained as
necessary for the new feeling-tone to reach its maximum was compared in
each case with the time-interval when the color alone was presented.
Various combinations were here employed also. Colors in forms in
addition were studied in comparison with the same colors presented
without regard to the enclosing forms. No definite results could be
obtained in most cases. It was thought that the repose one feels
for plush might appreciably hasten the feeling-tone for the red and
probably retard that for the more exciting yellow. The evidence is not
directly conclusive. This was not found to be the case in much more
than half of the tests. It did, however, in the great majority of the
cases with all subjects, retard the time-interval for the development
of the unpleasant character of ordinarily disagreeable colors. Given
at such times also the normally unpleasant colors not infrequently
appeared themselves as slightly agreeable. In these cases the interval
was also appreciably longer, suggesting evidence that new processes of
some sort were set up. A pleasant low tone hastened the arousal of a
pleasant feeling-tone for red quite perceptibly for three subjects, and
had no influence upon the other subjects. The feeling-tone for yellow
under the same circumstances was for two subjects retarded regularly,
with no marked effect either way for the others. The same low tone
retarded all the unpleasant colors, as did the plush, in many cases
causing them to appear as pleasant.

The effect of forms, as enclosures for colors, upon the time-rate was
more marked and constant. Subject I again was always disturbed when
colors were presented to him in definite forms. For him feeling-tones
never arose so quickly when the form-element entered. For the other
six subjects, available for this part of the work, upright ovals
considerably increased the whole state of pleasure whether or not
fusion of the different elements resulted. For them the feeling-tone
for every pleasing color was hastened from two fifths to four fifths
of a second. These same forms retarded the unpleasant colors whenever
one element of the experience seemed to be opposed to the other.
Occasionally here also the color appeared as itself directly and
unaccountably pleasant, the prepared situation of the subject being
such, apparently, that the ordinary character of the color did not
appear at all. This was very frequently the case for all subjects.

The irregular unpleasant forms generally retarded the feeling-tone
for the enclosed color when that color appeared to have lost some
of its accustomed agreeableness. When, however, the contrast in
feeling-character between the form-element and the color-element as
such was noted as marked, the feeling-character of the color was more
often hastened than retarded. These same forms in almost every case (of
nearly two months' work for seven subjects) hastened the feeling-tone
for the corresponding disagreeable color. Often again pleasant colors
changed the feeling-tones for these irregular forms. In such cases the
influence could not be attributed to the effect of unpleasant forms
upon feeling-tones.

Statistics alone seem insignificant here. Each variety of affective
experience in itself presents its own peculiar difficulties. In a
great number of tests the affective phases of the experiences were
all described in such terms as to suggest that too general a grouping
of them would not mean much. Often when one thought, after a careful
choice of the stimuli to be used, that the experiment would show that
feelings whose prominent characteristics were those of excitement or
tension, for example, were exerting an influence upon some other kind
of feeling, introspection would reveal the fact that altogether other
phases of the experience were the pronounced elements. Examples of what
at first appeared to be capricious results illustrate the baffling
nature of the problem here dealt with. Red is very pleasant. The oval
with the bulging side is repulsive. This combination caused no marked
retardation in the time required for a feeling-tone to develop. The
blue, not so markedly pleasant alone, with the same bulging oval as
its frame, had its feeling-tone changed, and the time-development
quite perceptibly hastened. This same blue color with an upright oval
as its frame produced a feeling-tone much more pleasant, also with
marked hastening of the speed-development of its feeling-character.
The pleasant-unpleasant dimension of the feeling clearly cannot alone
furnish one with an explanation of these different phenomena. The red
under normal conditions, _i. e._, if not influenced by either favorable
or unfavorable coexisting feeling-tones, aroused its peculiar and not
necessarily pervasive kind of physiological process. Likewise all our
evidence goes to show that the feeling for blue is correlated with a
peculiar physiological process, not so deeply seated in the organism,
and not so satisfactorily coördinated, or "definite." Now the specific
feeling-tone for forms arises when the imitative adjustment called
for is successfully accomplished. In the first combination cited
above the feeling-tone for red, being mild, soothing, more pervasive
than blue, but lacking in the exciting character, is correlated with
processes not so easily influenced by the reactions occasioned by
the presented forms. Subjects say that it does not call for "surface
reactions." It is less "intrusive." It does not "fall in with," nor
does it strikingly oppose, the necessary reaction to the forms. Its
influence upon the time-development of feeling-tones for accompanying
stimuli is consequently small. This is not the case with the blue.
The explanation, however, does not here differ in principle. This
"volatile, unstable, indecisive, thin, or shallow" feeling, can be more
easily influenced by the definite and decisive processes characteristic
of the forms. It, indeed, needs something to determine its character,
or coördinate its general reaction. Hence in both the above
combinations the development period of the new feeling-tone for blue
is shortened. The feeling reaches its maximum in either combination
more quickly than when it occurs alone. As one should expect, fusion
or mutual reënforcement quickens coördinated reaction; and partially
independent coexistence, except where the contrast is sharp, serves as
a condition for the lengthening of the latent period of feelings.


PART IV

It is beyond the province of this paper to report the accounts the
subjects have attempted to give of the complex feelings aroused by the
pictures of statues. The primary and limited purpose is to try to trace
out the influences of the feelings before dealt with for colors when
these are also present in some way related to the now complex æsthetic
states. The _Einfühlung_, often reported for the simple forms, is here
much more easily detected, if the statues arouse agreeable feelings.
They "work themselves into the statue," or assume the position, or
the facial expression if this is prominent, or feel very strongly in
their own body what seems to be the most prominent element in the
feeling portrayed by the figure. Few subjects liked all the statues.
Incipient if not actual tendencies to motion of some sort, with the
sensory counterparts to these situations called for when the subject
feels that he is in the "proper attitude" to get most feeling from the
presentations, chiefly constitute what was in different ways reported.

These statues presented on colors as backgrounds are variously and
interestingly modified. The feeling-tones for colors distinctly affect
the meaning of the statues. Of the above colors our shade of red is
preferred with Venus by all subjects. Here the feeling-tones more
nearly fuse. Always the feeling-character of the statue predominates,
and the other feeling-elements of the situation are accepted or
rejected in proportion as they harmonize or fail to harmonize with the
predominant partial tone. Red, for example, here adds to the "richness
and luxuriousness." It accentuates the strength, poise, grace, balance,
ease, rest, wisdom, composure, endurance, and dignity. It is more
soothing, and calls for no unnecessary action. One subject never liked
any color as a background. In this case colors were good in proportion
as "they kept out of the way." This is the reason for red being always
preferred to yellow or green. With these latter colors there is an
interplay of reactions not coöperating. The color-exciting element is
more immediate, tension is brought about, the color asserts itself, is
pleasant, and tends directly to inhibit the feelings for the statue.
The pleasure in the color is called "thin" in comparison, and the power
of sympathetic appreciation of Venus is lessened. There is suggestion
now in the statue still of its strength, but with no "enduring"
quality. It has become commonplace, merely a "pretty woman," jaunty,
self-sufficient, cynical, and with little dignity. The motion element,
now prominent, is not pertinent. The statue looks "cheap," and as a
figure is volatile and unsteady.

In a similar way one finds these feeling-tones for colors variously
playing their part. The statue of Apollo is not pleasing to some
subjects. They want it "toned down." The red effects this. To some it
is most pleasing by its suggestion of easy grace and springy, elastic
step. The yellow harmonizes and accentuates this chosen feeling. The
blue often destroys its moral meaning. The red hampers the feeling
for the _Laocoön_ group. They become listless, dead, and have still
strength, but no struggle. Yellow increases the amount of activity,
but often lessens the "serious despair." Fierceness is added, but
the liveliness thus furnished is at the expense of the necessary
balancing solemnity. The color again becomes intrusive. "The snakes
fairly dance," and the "flashing action behind" the statue is now too
prominent.

For the _Dying Gladiator_ or _Dying Alexander_ red is preferred. It,
however, as do all the other colors, often produces an overbalance
in the whole situation. Here it suggests no conflicting feelings. It
adds--often too much--to the hopelessness of the situation, and gives
to them an exaggerated solemnity and resignation, which emphasizes
a melancholy cast, not altogether called for. Green and yellow are
always incongruous. They tend to distract the attention to certain
particular muscles, thereby lessening the whole general effect. The
little "prettiness" they still retain is not called for, and is not "of
the right sort." They do not allow one to be sufficiently contemplative
or thoughtful. They have little depth, and cause inharmonious bodily
commotions, and too much intensify the life-struggle and anguish.

The general effect of the statues here is much like that of the simple
forms above. Both not only call for something to be done by the
subject, but some action more or less already definitely outlined.
The Einfühlung for little wooden figures, such as cones, columns,
pyramids, etc., was clear and decided. The tipping character or the
straight erectness caused a feeling which seemed describable in terms
of the way in which the bodily position, as one naturally adapted
himself to the object, took place. Statues afford richer experiences,
but the principle is not different. They seem full of suggestions of
abstractions, such as strength, wisdom, grace, beauty, power, and, in
general, what are often called spiritual feelings. These are not so
easily imitated in detail. Subjects have their own ways of adapting
themselves. They want to carry out the suggestion or impulse in their
own way. These impulses are projected into the figure, and all of the
vigor inhibited in one's own body becomes a living part of the figure.
The impulses thus from the colors may or may not be of such a character
as to bring about the same proportionate adjustment as a desirable
intensification equally of all the feeling-elements. Whether desirable
or not, however, these feeling-combinations furnish additional
illustrations of the various mutual relations of coexisting feelings.

The _Angelus_ or the _Shepherdess Knitting_ bring about feelings in
striking contrast to the feeling for the _Horse Fair_ picture. The
latter arouses suggestions of a tumultuous bodily condition, increased
muscular tonicity, muscles twitching everywhere, breathing heavier,
shoulders strained, and in comparison, great general innervation.
When the characteristic feeling had been aroused, the subjects were
requested to close their eyes and observe if possible to what extent
the feeling already aroused was dependent upon the retained images.
The results were clear. If the feeling is slight, as it is for some
subjects, the feeling tended to vanish and return with the recurring
images. If it is fairly strong, the feeling persists for some time
after visual imagery is lost. If very strong, the feeling is constant
for a still longer period and still less dependent upon the original
peripheral excitation. The feeling is always more constant than
any imagery. Not often for example is the whole picture retained.
Sometimes one prominent part only remains. Often again various kinds
of imagery aid in preserving the feeling. Besides visual, auditory,
the sounds of the horses' hoofs, of the tones of the Angelus bell, are
chiefly prominent in preserving the situation and the condition for
the feeling. Articulatory impulses again, in the tendency to repeat to
one's self such words as are descriptive of the moral meaning of the
pictures, offer sufficient clues to keep the desired feeling aroused.
When the feeling has "struck deep," subjects report motor imagery
pervading the whole system. In such cases the recurring visual imagery
has little effect upon the feeling. On the whole, the feelings for the
more quiet pictures last longer and are more easily retained than is
the case with the more exciting ones, if the original feelings are, as
to mere intensity, approximately equal.

The character and strength of these feeling-tones determine also to
a large extent the lines of association followed. Here the mutual
influences of feelings are clearly recognized. The character of the
new associated images and situations is colored by the feelings which
were connected with the original stimulations. The pictures, such, for
example, as the _Angelus_ and the _Horse Fair_, were presented to the
subjects in quick succession. These were to be merely starting-points
for association. For all the subjects who were able to report anything
definite, the feeling-tone for one was read into the associations which
were aroused by the other. The second of the two starting-points as
a rule controls the imagery. A few examples will illustrate how both
feeling-tones are retained. The _Angelus_ was presented first in these
cases, and the _Horse Fair_ second.

For subject K, the parts of the picture of the _Horse Fair_ remained.
The feeling of seriousness and quietness, foreign to it itself, was
projected into it. Solemnity and the feeling of strength and power was
accentuated. The gaiety originally present was very much lessened, and
finally not noticed at all.

For subject C, a sacred feeling was aroused. Wars of the Bible were
recalled. There was a fusion of the imagery. He saw the church on a
battlefield near a cavalry fight. The feeling of active earnestness and
the sacred moral character was reported as due to the retained feeling
first brought about by the _Angelus_. The influence of the other
starting-point is clear.

Subject B found the incongruity between the two feelings very strong.
The _Angelus_ was the stronger in influence. The other caused one to
stress the lighter, more trivial character of the former. Meadows,
streams, pools, and enchanted regions typified the fanciful mood thus
brought about.

It is not a question as to whether such trains of thought would have
occurred if only one starting-point had been used. It is rather that,
in such cases as the above, two distinct feeling-tones were actually
detected as playing their part in the resulting complex experiences.
It is with some effort that both feeling-tones can be thus at first
retained. The resulting undertone or general mood, however brought
about, colors and determines to a certain extent the associations which
follow. The feeling is more deeply seated than the image, and here also
it is retained longer.

The above recorded account of the behavior of simple feelings fairly
represents the accumulated data at our disposal. How they can be
adjusted to modern theories of the relation of consciousness to
movement may be briefly suggested. Yet the rudimentary state both of
the psychology and of the physiology of feeling makes the present
task a hazardous one. Psychologists are not agreed as to the best way
to conceive of the relation of feelings to sensations. Feeling-tone
is in some ways dependent upon sensations; and at the same time,
in comparison with other sensation attributes, it is relatively
independent. Physiologists are still farther from agreement with regard
to the nervous processes involved.

But the _deeply organic seat_ of feelings is unquestioned. However the
concept of feeling itself may differ, all are looking for corresponding
bodily processes by means of which to classify these affective states.
Clearly, to say that feeling is of such a nature that one need never
hope to be able to predict it from psycho-physical conditions, is
no more justified than to say that we can never predict exactly the
intensity nor the vividness of any stimulation. Feeling-tone is here
simply on a par with other attributes ascribed to sensation.

According to Münsterberg's Action Theory the intensity of the sensation
depends upon the strength of the incoming current. Its quality depends
upon the position or location of this current in its particular
neurone. The vividness depends upon the "openness" or "closedness" of
the neurone conditioning the _outgoing current_. And finally to the
feeling-tone shall correspond the local difference of this discharge in
outgoing currents. For instance, the pleasant feelings have, related to
them, central outgoing paths which lead to approach, and thus to the
continuation of the stimulus, and the unpleasant feelings have related
to them in turn central neurones which lead to withdrawal or escape,
and thus to the breaking-up of the stimulus.

Our empirical data gathered from the experiments above reported demand
not so much a modification as an elaboration of this theory. The
_tridimensionality_ of the feeling-tone itself must be physiologically
described. We must conceive the feeling-tone itself as possessed of its
own vividness, intensity, and quality.

It seems clear indeed that any explanation of the affective or
feeling-character of experience must be sought somewhere in the
outgoing currents from the motor region. This alone will serve to
account for the inevitable volitional or "intent" aspect which
invariably accompanies feeling, and I think may serve to account also
for the organic or necessarily coördinating or functioning aspect
required by some writers who so stoutly object to "barren atomistic or
structural" psychological explanations.

The Action Theory might then be specialized in the following way:

The _intensity_ of the feeling will depend upon the force or amount
of the outgoing currents from the motor cells. This would enable one
to explain that state of mind when a sensation only is experienced
from a stimulus which ordinarily has a characteristic feeling-tone,
but which feeling-tone in the special instance is lacking. Many cases
have been cited above where one feeling seemed to efface another. The
nerve-energy called for in arousing the unpleasant feeling-tone for
the sandpaper inhibited the process of the discharge from the cells
conditioning any response to the ordinarily pleasing red color. Others
again can reënforce or at least not seriously interfere with each
other. All cases already cited where two feeling-tones were detected
as existing simultaneously are examples in point. It is quite clear of
course that the intensity of feeling is not at all commensurate with
the intensity of sensation. Commotion is not the only condition for
emotion. Yet where there is no tendency to _do_ anything, as is so
noticeable in the reported introspections above, there is no feeling.
A mere shock, even though intense as a sensation, simply benumbs one.
In thus describing any feeling for a particular stimulation, one
should include, besides the original results of the chosen peripheral
excitation, all the reënforcing factors that accumulate by reason of
the sensory counterparts to this originally called-for movement. When
one is, for example, feeling sandpaper, the feeling for the soft red,
when it exists at all, is less intense. Subjects say, "It ought to be
more pleasant than it is. The trouble is in me, not in the color."
The suggested movement which conditions the intensity is lessened in
amount, or partially inhibited. One could scarcely say, so far as the
sensation is concerned, that it has lost some of its brightness, or
that it is not strong enough to arouse its customary feeling-tone. This
is distinctly reported as not the case. It is of course almost always
recognized as the same shade of color. The recorded examples, showing
that intensity of feeling is itself one dimension of a feeling-tone
in no way necessarily related to the intensity of the sensation, are
numerous.

The _vividness_ of the feeling-tone is likewise a relatively
independent phenomenon, and it, too, is not commensurate with the
vividness of the sensation as such, and hence demands a different
explanation. It can then be dependent upon the _actual stage in the
process of completing the movements suggested_ by the color or tone
or form in question. All feelings dealt with in this investigation
one can describe by relating them to the actual stage in the process
of completing the _coördinated adjustments_. Without some progress
in such a process no feeling would cross the threshold of awareness.
In Part III above are recorded many illustrations, where _degrees of
vividness_ for feelings are noted by the subjects. When they were
attempting to report the actual time when a feeling became definite
enough to be called such at all, there was much difficulty in knowing
just when to give the signal. Feelings develop much more slowly than
do perceptions. Subjects often give the signal too soon, at once
correcting themselves by saying that it was too vague at that moment.
It grows in definiteness, and has degrees of vividness. A movement
in the first stages of the process, before the feeling-tone has
sufficiently developed, is a state of vague feeling. Again, many states
of so-called indifferent feeling meant, according to the subjects,
not lack of feeling, but rather vagueness, lack of vividness. Three
or more stimulations from different sources resulted in confusion
where no feelings were vivid. When the color again, for example, is
pronounced "dead" so far as feeling is concerned, other feelings and
other movements are too prominent. The sensations are in such cases
unchanged. The intensity and vividness of the feeling-tone for the
color are at a minimum.

And thirdly the _quality_ of the feeling-tone must be dependent upon,
and must be described in terms of the particular kind of coördinated
movements suggested or actually carried out. Thus the characters of the
feeling-tones for the yellow color above described, for the upright
ovals, for the very high tones, for the _Laocoön_ group, and for the
_Horse Fair_, are in some respects alike. They have the same general
_Gefühlsgrundlage_. The qualities of the feelings for soft deep red,
for tactual plush, for low tones, and for the _Angelus_, and, for most
subjects, for Venus, would represent another class having the same
_Gefühlsgrundlage_. This admits of all the uniqueness specific feelings
may have, and at the same time permits of a general classification
and description. Some subjects, D and F, for example, may have a
feeling whose quality is disgust at some color-form combination. The
accompanying sensations may be localized, as they frequently are,
in the arms, with impulses to "ward off" the displeasing influence.
Subject B often for the same feeling finds sensations of contraction
in the throat most prominent, and subject A a stiffening of the
features and incipient scowl. The most prominent localization depends
upon the habits of the person and the habitual kind of reaction he
has acquired and developed during his lifetime. The localization of
muscular activity may differ, but _the kind of coördination does not,
so far as our introspection shows_. The _regularity_, the _rate_, the
_smooth light ease_, or the _heavy, ponderous, deep-seated character_
of the suggested responses indicate some of the terms which would serve
as aids in classifying kinds of processes which are physiological
conditions for feelings of definite character. Again, feelings of
pleasant repose, of depression, or of sudden collapse are still changes
also in innervation tonus. These are adaptations for situations just as
are the more positive or aggressive kinds illustrated above. Feelings
where quick collapse occurs differ in _quality_ from feelings of calm
repose. All can be conceived as kinds of adaptations or responses, and
clearly correspond to the characters of feeling-tones rather than to
any other dimension of feelings or sensations.

Certainly the central preparedness for discharge largely determines the
feelings. The external excitations are merely the clues. The internal
apparatus is set vibrating in a constant manner if no other external or
central stimulus is present to demand other adjustments or to intensify
the same kind. When such synergetic or antagonistic stimuli are also
present the mutual influences of feelings do seem to be, indeed, of
great significance.

FOOTNOTES:

[Footnote 84: As Mach long ago pointed out, _pleasant_ is a vague term
and in itself does not serve as a true descriptive term. _Pleasant_
here applies to both feeling for red and for yellow, but something more
is needed to distinguish these very different feelings.]

[Footnote 85: "Irritating" as I shall use the word has no hedonic or
algedonic significance.]




THE ÆSTHETICS OF REPEATED SPACE FORMS

BY ELEANOR HARRIS ROWLAND


Part I

The object of this paper is to discover some of the sources of our
pleasure in repeated space forms, and the laws which govern this
repetition. The repetition of an object, and its regular recurrence
subject to certain possible variations, is one of the basal principles
of art, and of architecture in particular.

It is necessary at the outset to define our use of the word
_repetition_ more exactly, for there are obviously different meanings
of the word, which may lead to confusion.

1. The term repetition may be applied to the existence of any two
objects similar to each other, whether they are near together or widely
sundered. Our pleasure in such a repetition would be merely that of
re-seeing and recognizing the two as counterparts of each other. This
kind of repetition I call _conceptual_, for it requires only that the
memory-picture of the object be held in mind and the two recognized
as similar when met again. This is not the kind of repetition which I
have in mind, and I shall never use the word in this sense during the
discussion.

2. In any one work of art there may be some feature repeated, some
motif which is taken up and carried out in different ways throughout
the whole, and these features we recognize as having an orderly
relation to one another in the unity. This might be termed repetition
of _content_, and be applied to the recurrence of some type of
decoration over a window or a peculiar arch taken up in various ways
throughout a cathedral. I do not use the word in this sense, but limit
it still further.

3. By _repetition_ is meant during this discussion the regular
recurrence of an object, and an equally regular recurrence of
intervals. The repeated object must come at uniform intervals, and
this restricts us to the consideration of that repetition alone which
consists of recurrence at regular intervals of some object more or
less beautiful in itself, and the description of the nature of our
æsthetic feeling in experiencing such a series.

Although this discussion is divided into the two divisions of
_experiments_ and _analysis_ of architectural examples, and the
experiments are described first, the investigation was not carried out
separately in this order. The two went along together, the art-analysis
suggesting experiments, and the experiments in turn throwing light on
the analysis. The two parts of the discussion are kept separate merely
for the convenience of the reader, and in the experimental discussion
all allusions to the art-illustrations are excluded in order to avoid
confusion. In reality the two went hand in hand, but the connection
between the experiments and art-analysis will be reserved for the
latter half of the paper.

The experiments were begun in the following manner: In a velvet screen
about a foot high was cut a window 460 mm. by 35 mm. in size. Behind
the window was a metre measure and a rod from which hung small strips
of cardboard 10 mm. wide. First two, three, and four strips were
hung behind the window, and the subjects were required to arrange
them at the intervals where they preferred to see them repeated. The
results were uniform in certain particulars and very suggestive. In
their arrangements of two, three and four strips, the subjects were
guided by considerations of symmetry or proportion. They insisted that
although they knew that the strips were repeated, they did not feel the
repetition, but the strips seemed like parts of some larger unity to be
arranged with reference to the unity of the whole. With the addition
of the fifth strip came a difference in their apperception. Instead of
the strips seeming parts of a whole including figure they seemed like
repeated units.


FAVORITE ARRANGEMENTS

         2 _strips_   3 _strips_     4 _strips_          5 _strips_

    J.    30 mms.     4 mms.   { mid. sp. =  25    any symmetrical
                               { ends     =  10    arrangement
                                                   better than
                                                   equality

    S.   170         12        { mid. sp. =  15     {mid. sp. = 40
                               { ends     =  12     {ends     = 30

    U.    40         20          30                35

    R.    30        130        { mid. sp. =  30
                               { ends     =  10    10

    L.    23         40          70                70

    W.    40         10          30                30

    V.    20         10        { mid. sp. = 100
                               { ends     =  60    15

It will be seen, from the table, that with two exceptions they
preferred five strips equally distant from one another, while with four
strips, four subjects had preferred a symmetrical arrangement. These
gave as their reason that with five strips the latter appeared more
definitely to be repetitions of one another, while the four strips
seemed more like parts of a whole which required symmetry in its
arrangement. Moreover the two subjects who preferred five strips in
symmetrical arrangement instead of at equal distances affirmed that a
distinct feeling of repetition came with five strips that had not been
felt before, only they did not enjoy this feeling of repetition as well
as one of symmetry. After having seen the five strips, some subjects
could feel the repetition with four strips, but none with three. The
question naturally arose, what is this _feeling_ of repetition which
makes one say that four or five repeated objects deserve the name,
while three or less are regarded in a different light? The analogy
between the apperception of this visual repetition and auditory rhythm
seemed so strong as to deserve attention.

In auditory rhythm it is necessary that there be recurrence of more
than two elements; they must come at a certain rate and within a
certain temporal space to seem connected with each other, and they may
be subjectively grouped in different ways. The apperception of both
kinds of repetition had so many analogies as to suggest that some of
the factors in both experiences were identical.

To focus the problem I took a definite thesis in regard to it. Our
apperception of repeated space forms is due to the rhythm of our own
motor adjustments which are excited in face of repetition, harmoniously
if they accord with certain rhythmic laws in us, inharmoniously if they
do not. It was then necessary to find what facts would support such a
thesis, to see if in reality such facts could be marshalled, and if the
explanation and support they offered was conclusive enough to make it
needless to look farther.

It would seem, if our pleasure in repetition depended on temporal
motor responses in us, that if the amount of time normally taken to
traverse a repeated series were shortened, or if the eyes were fixed
and not allowed to move over the field at all, our enjoyment would
cease altogether, or at least be seriously diminished. If we found it
impossible to enjoy the series except when seen for a certain time,
long enough for the eyes to go over it in the rhythm peculiar to each
subject, we should then conclude that our enjoyment did depend, to some
extent, on such temporal rhythm.

I experimented on this question with nine subjects, and the results
brought out different ways of apperceiving repetition, which divided
the observers into two rather well-marked types.

The apparatus was of the simplest, consisting of white silk strings
hung on a wire against a black background across one side of the room.
The strings were attached to the wire by little hooks, which enabled
one to change their position easily, while a cloth hid the weights on
the ends of the strings, so that nothing but the vertical white lines
were visible.

Fifty strings (50 mm. apart) were hung before the subjects, and they
were asked to survey the field and give a signal as soon as the
experience became pleasant. Then having found the average length of
time for each subject to enjoy these simple repetitions, a shorter
period was given when they were to shut their eyes at a given signal,
and see if in that shortened time they were still able to enjoy the
series. Next they fixated the eyes and kept the whole body rigid, to
see if pleasure was still possible when all outward motor response was
checked, so far as possible.

The results of this experiment were very suggestive. Of nine subjects,
all felt pleasure when allowed to move the eyes over the series at
random; with eyes fixed, five felt their pleasure much altered in
its quality as well as lessened, while with one it was altogether
destroyed. With four, however, although there was considerable
alteration in the quality of the pleasure, its amount was increased
rather than lessened.

    B. (1) Average time necessary to enjoy the series: 4.7 seconds.
       (2) Three-second exposure. No pleasure, needs more time during
            the movement.
       (3) Eyes fixed: 4 secs. = Av. time necessary to enjoy it. Lines
            bunch toward centre and fade away at sides, giving a kind of
            unity, but he feels constraint.

    R. (1) Av. time: 4.3. Sees them in pairs.
       (2) Two-sec. exposure. Very faint pleasure; feels that only a part
            is perceived.
       (3) Eyes fixed: 4.3. One pair fixated, the others fade away, making
            a kind of figure. Pleasure faint and constrained.

    L. (1) Av. time: 2.1.
       (2) 1-sec. exposure. Pleasure faint and incomplete. He feels the
            pleasure comes from memory of the previous experience.
       (3) Eyes fixed: 2.2. Great effort to find any pleasure. It consists
            mainly in seeing a few strings, and feeling there are others,
            even though they are not distinguished.

    V. (1) Av. time: 2.2. Sees them in pairs.
       (2) 1.5-sec. exposure. Enjoys the experience in memory after the
            eyes are shut again.
       (3) Eyes fixed: 1.9. Still sees them in pairs, but cannot see enough
            of them, hence they are less pleasant.

    W. (1) Av. time: 4.3.
       (2) 2-sec. exposure. Not enough time to feel any relation between
            the strings, most of the pleasure supplied by the memory.
       (3) Eyes fixed: 5.3. Pleasure is very faint, and consists in having
            the strings appear to converge to a central point, and fade at
            the sides.

    J. (1) Av. time: 2.3. Sees them in pairs.
       (2) 1.5 exposure. Less pleasant.
       (3) Eyes fixed: 2.7. Series seems more like a unity and he enjoys
            it more, since no time is spent in exploring the field, but it
            is one unified experience.

    U. (1) Av. time: 28. Only enjoys it by ignoring all except those in
            the centre--does not want so many.
       (3) Eyes fixed: 18 secs. Enjoys it when eye lights on one string,
            so that the others can fade away equally at the sides, in one
            figure.

    S. (1) Av. time: 5.
       (2) 3-sec. exposure. Less pleasant.
       (3) Eyes fixed: 8.8. Pleasure consists in converging of lines toward
            central point. It appears like one figure and is more intense
            than (1).

    H. (1) Av. time: 9. Sees them in pairs.
       (2) 1 sec. Just as pleasant as before.
       (3) Eyes fixed: 4.6. Pleasure in unity of whole series with centre
            of fixation emphasized. Only felt pleasure anyhow when the eyes
            had stopped moving, so now it comes all the sooner.

From these introspections it is obvious that there are two distinct
ways of apperceiving repetition: One in which the rhythmic element
is pronounced, so that when the time necessary for such a rhythm is
shortened, or by fixating the eyes the motor response is hindered, the
pleasure in the repetition is either altered or destroyed altogether.
The other type takes a repeated series in the sense of a unified
presentation and wants it all at once in a symmetrical whole. The
rhythmic factor is present in both, as is shown by the fact that the
quality of the pleasure was changed in every case when the time of
exposure was shortened. But in the latter type of subject the pleasure
felt in the presentation of the whole at once, and the feeling of
symmetry around a middle point, are more intense than a rhythmic
apperception. These two kinds of apperception remain fairly constant
throughout the experiments, and for convenience' sake we shall call
them spatial and temporal types. With the former, the value of the
experience consists especially in having a central fixation-point from
which the repeated elements fade away equally on the sides, making a
symmetrical whole. With the temporal type, the pleasure is felt by
means of the rhythmical passage from one element of the series to the
other. In passing from point to point the rest of the field still
remains in indirect vision, so to the distinctly temporal a distinctly
spatial factor is also present. For this reason the temporal type of
apperception is the richer of the two, and a description of it comes
more nearly to the essence of repetition as such.

Up to this time, the repeated element had always been a single string.
This was varied and the strings hung in pairs (50 mm. wide, 100 mm.
between pairs). When the strings had hung at equal distances from
each other, six out of the nine subjects had seen them in pairs while
enjoying them, and had found such grouping more or less essential to
enjoyment. In seven cases the pleasure was increased by this grouping.
They expressed their preference in various ways: "Easier to keep track
of where we are going." "Can go quicker over field, for repetitions
are more well-marked." "Single line is too thin to rest on, this gives
broader space for repose." All these introspections instanced the
necessity of the rhythm being marked and made plain, so that there
should be no confusion of point with point. The two who disliked this
grouping were of the spatial type, who found no pleasure in traversing
the field, hence too little content in it, in this arrangement, at
any one time. Grouping of some kind would seem to faciliate the
apperception to a certain class of subjects, while with others the
amount and quality of the content of the field is of more consequence.

Since accents are such an important factor of auditory rhythm, the
next experiment was to see if the apperception of a series of repeated
elements would be facilitated by accenting every other one.

[Illustration: Fig. 1]

Another string was hung in every other pair thus making it more
striking, but here came a difference between the feeling of accent in
auditory and visual rhythms. The subjects declared the pairs in which
a third line was hung were not intensified alone, as when a greater
stress is put on a tone in auditory rhythm, but the pairs were changed
qualitatively. The group of three became the repeated element, while
the pair was only an alternating figure different from the principal
unit. This unanimous testimony brought up a variety of questions.
1. Is any purely _intensive_ accent, without involving qualitative
changes, possible in visual repetition? 2. What factor makes us choose
one object rather than another as the repeated element? 3. What is
the value of the alternating figure in such a series? 4. What is the
value of the empty space between repeated figures, and does it have as
distinct a value as an alternating figure? 5. Are all the recurring
objects and spaces felt as separate repetitions; if so, how many can be
carried on at once?

These questions were put to the subjects in regard to the series just
described with considerable uniformity of answer.

(1) No such thing as purely stress accent seemed possible. The word,
signifying greater intensity without change of quality, did not apply.
If one attempted to _intensify_ the repeated object in any way, either
hanging another string, thickening the strings, or any similar device,
it ceased to be the old unit but became a new one, whose repetition
was followed for its own sake, while the weaker one retired into the
background, and was not _felt_ as the element repeated in the series.

(2) Any regular change of the element which made it more interesting or
caught the attention, fixed it as the chosen unit of the series, whose
repetition was followed.

(3) Concerning the exact _value_ of the alternating figure in the
series, there was great difficulty in introspection. They all "_knew_
the alternate figure was just as truly repeated as the principal
one, but could not feel it so." The three-group formed the unit of
the repeated series, and although the pair was clearly part of the
experience and distinctly perceived, for some reason it was not
felt as repeated in the same way as the other. It was merely an
alternate, a filling, which was essential to the other, but which had
no significance in itself as a repeated thing. Two subjects were able
(if they tried) to carry both repetitions along together, _i. e._, not
only feel the three-groups as coming at regular distances from each
other, but the pairs as forming another interlacing series. This kind
of apperception was very fatiguing, however, and they could not enjoy
it. For any pleasure to be derived, the pairs must retire into the
background, and attention be fastened on the three-group.

(4) If the alternating figure was to be so subordinate, was there any
difference between its significance, and that of an empty space? This
was everywhere answered in the same way. The alternating, or minor
figure, had a very distinct value, and any irregularity in it was even
more irritating than in the principal unit itself. When the space was
empty they thought nothing of it, the equality of the interspacing
was taken for granted; while if they felt an irregularity in it, it
destroyed their pleasure in the whole series. But there was no feeling
for the empty space until its regularity had been violated, while there
was a distinctly pleasant factor in the minor figure, even though
different in quality from the principal element.

In the foregoing experiment the differences between the spatial and
temporal type of observer were still strongly marked. The former type
invariably grouped the elements (usually with the three-group in the
centre and a pair on each side) and they took their pleasure in the
symmetry of each figure so made, moving from the centre of one to the
centre of the next adjacent. In this method of apperception there
was no empty space between repetitions, for the whole group of three
figures was taken as the repeated unit. The empty or rest-phase was
gotten in moving from the centre of one to the next, in which passage
the limiting pair was ignored. One spatial subject, finding the
proportions of this artificial grouping poor, got no enjoyment at all.

With the temporal type, the experience was quite different. They moved
across the field with the three-group as their stopping-points. These
principal elements were what they looked for, and their pleasure seemed
to consist in expecting and meeting it. What part the _pair_ played,
they had difficulty in analyzing. Some said that while the three-groups
occupied most of their attention, they gave a lesser degree to the
pair, so that the rhythm of the passage across was marked by heavier
and lighter beats. Another found the figure in the alternate space
only an obstacle, and felt he was hindered in the passage from unit to
unit, the only compensation being, that the "hindrance" came at regular
distances. Others felt that two repetitions were actually being carried
on at once. By this they meant that the two sets of elements were kept
distinct, although objectively combined, but the repetition of the pair
was subordinate in interest to that of the three-groups.

I tested the same thing (accents or major elements, and the value of
the alternating minor elements) by simply doubling every other string
of the series 50 mm. apart. In every case the effect was found poor.
It was "confusing," "too much work." They all felt adjusted to the
repetition of the double string and then encountered the single one,
which hindered them, and by trying to keep both elements going at once
they were fatigued. Most of them had a distinct feeling that they
wanted to swing from one element to the next, and were baffled by the
alternate. In this arrangement, even members of the spatial type who
had not been able to get any rhythmic feeling before, felt the movement
in the series as if they were going across, although they went (in
two cases) in groups instead of single elements. They all, however,
felt the single string as an obstacle which hindered their enjoyment,
whether the double string or a group was taken as element. This
suggested the question: What makes the difference between the minor
figure being an enrichment to the experience and being a hindrance?
They insisted some rest-period was necessary; some really empty space
between the repeated units, and when in place of rest they had more to
do, it spoilt the pleasure.

[Illustration: Fig. 2]

Next, two strings were hung at equal distances between the double
strings, and the latter put 100 mm. instead of 50 mm. apart. This was
liked better in every case, and the reasons given were much alike.
The double string was still the repeated unit, while the two strings
between did not _feel_ repeated. In spite of the obvious inconsistency
of the statement that they did not _feel_ the alternates to be repeated
even when they knew that they were so, just as much as the double
string, several subjects made the same remark. V. felt the series as
a rhythm, where the double strings were all he was interested in,
although he knew he should notice if the others were changed. B. could
not detect that they were of any importance, except as he imagined them
absent. B. could feel either the double strings or the two strings
between them as the major unit, only, whichever one he took, the
other retired into obscurity. He felt the minor units in a different
way as being repeated together with the majors, but very weakly, and
not at all unless he previously considered that he ought to do so.
L. said his attention was fastened on the double strings, but it was
the "effort or ease with which he passed over the alternates which
formed the pleasure." Another temporal subject felt the rhythm as the
others did, with more emphasis on the double strings. But the major
units seemed the _rest-phase_ in his rhythm, _i. e._, he paused here
in observation of the series, although his attention was most active;
_vice versa_, the minor units required little attention, but were the
active or moving parts of the rhythm. Since they all considered the
strings in the alternate space merely as steps or lesser beats on the
way to the major element which they sought, and as obstacles rather
than otherwise, why did they prefer the space with two strings rather
than one? This suggested that some factor in the alternate spaces was
important other than simply the amount of resistance to overcome in
getting past the two lines. In answer to this they could only say that
the two strings in the alternate space formed a pleasanter unity by
itself, although, as they went across, they did not think of it in
terms of unity.

What, then, is the real rest-phase of the rhythm of alternating
repeated objects?

In the beginning of the discussion, when the analogy between visual and
auditory rhythm was felt so strongly in a certain type of subject, they
had expressed themselves as if the object which they called the unit of
the repeated series were the active stimulated part of the experience,
while the alternate space was the rest-phase, valuable only as a period
of repose or blankness before the object was again encountered. But
in this case, although they felt they were putting no attention or
emphasis on this space, in reality they were keenly alive to what was
hung in it, even preferring more "hindrance" in the way of strings
than less, which suggested that the alternate space was of more value
than they were conscious of. (Some of this increase in pleasure was of
course due to the increased actual distance between the double strings,
but some also to the extra string.) The introspection on this question
as to which part of the rhythm was actually the rest-phase (if there
were any such) was difficult for them all. They felt they spent more
_time_ on the major element; that was what they looked for and found
pleasure in meeting again.

One said, "The unit is what I look for; as soon as I have it the
pleasure ends and I want to move on again. The pleasure does not
consist in resting on it after it is found, but in knowing I am going
to meet it again, and in doing so." As to the alternate spaces, he
could only say he was not consciously interested in them, he took
them for granted, but knew he should feel it, if they were changed.
His feeling for them was wholly negative. The other temporal subjects
agreed essentially with this. The alternate figures had to be passed,
but passing them was only of importance as it helped or hindered the
perceiving of the major elements. All agreed that any change was
noticed and felt irritating at once, although they could not understand
how it should, since so little attention was paid them normally.
One subject felt the alternating strings only as obstacles between
the doubles, and demanded an actual, empty rest-period between any
repeated units. When asked if it were really the rest-period _between_
elements or _on_ them, he said he felt there was a complete discharge
of attention on the major units, and an attempted one on the minor or
alternate units, and the attempted ones became confused.

These introspections would point to the fact that alternate minor
spaces while affording rest for the attention were periods of activity
of some other kind. The fact that no one could say _what_ kind, and yet
insisted on the feeling of its being important and distinctive, and
moreover repudiated the idea of change in a minor space even more than
in a major--this seemed to show that there was a value in the alternate
spaces quite aside from attention, but fully as distinct in its own way.

As might be expected, those of my subjects to whom rhythm was not a
conscious factor of the experience of repetition could not understand
exactly what was meant by the distinction between _rest-phase_ and
_emphasis_ of rhythm. In all the preceding cases where the temporal
type gave the introspection I have described, the spatial subjects
grouped the single lines, in Fig. 2, about the heavier pair as centre,
and moved from the centre of one such group to the next. The experience
then consisted of a succession of adjacent symmetrical groups,
connected by movement from centre to centre. When asked if there was no
pleasure in finding equal distances between their centres, _i. e._, any
temporal element whatever, they all denied feeling any. They could not
detect that they _felt_ the distances between their centres as equal,
although they _knew_ them to be. They spent so much attention on the
group that all feeling of the distance between its centre and the last
was lost before going on to the next.

These two marked types of apperception of an alternating series seem
varieties of _emphasis_, rather than of actual experience. It was
evident that those in the spatial type must have some recollection of
the amount of distance passed over between the various groups to feel
the whole series as connected in any way; while those of the temporal
type could not be wholly absorbed with the separate lines of the series
as they traversed it, but were distinctly conscious of the space
relations of those in the side of the field that they had just passed
or were coming to.

[Illustration: Fig. 3]

Next, I tried to see what were the different factors which made up the
value of the minor spaces. By varying both their size and filling, and
doing the same to the major element, I could judge the relative value
of these changes on the two, and their effect on the whole series.

The test was made in the following manner. The series as it stood
consisted of a double line alternating with a single one.

With every temporal subject the double line was conceived to be the
repeated thing, and the space between considered as an alternate,
with a repeated line of its own, to be sure, but not felt in the
same way as the other. With the spatial type, the single line was
merely the limiting edge of the symmetrical figure, with a double
line in the centre. One subject varied back and forth in his method
of apperception, and considered the richness and variety of these
different apperceptions as one of the chief sources of the pleasure
therein.

_Variation of alternating spaces:_ The minor spaces were varied by
hanging two strings in one, and one in the other, and subject asked how
such a change affected his feeling for them. The change was marked.

The spaces which had before been minor were so no longer. The alternate
space in which two strings were hung with the boundary-line of the two
double strings became the new element, and the alternate in which only
_one_ string was hung continued to be the alternate in the new series.
The whole series shifted itself, and settled into a new equilibrium.
Some of the subjects were able to feel all the former minor spaces
still as such, but only by a definite effort, and not while taking any
pleasure in it. The change in the alternates spoilt the whole scheme of
the repetition as it already stood, and made a regrouping necessary. I
next tried varying alternates by removing a string at intervals.

[Illustration: Fig. 4]

[Illustration: Fig. 5]

Since the strings were not removed in any regular fashion, and because
the subject could not find any possible consecutive way to group them
with the double strings, this variation was partially overlooked,
and although confusing the series somewhat, repetition of the double
strings could still be felt. Thus a mere _gap_ where the scheme
remained the same was not so disturbing as an extra feature inserted,
or one noticeably changed. Something could be supplied by the subject,
but not altered so easily. In these cases, however, the change was only
tolerated because it was ignored. They felt it as a mistake and so
overlooked it, but, accepted as a component part of the series, it was
impossible to feel it as a repetition or get pleasure from it.

[Illustration: Fig. 6]

The next variation was in the position of alternate figures. With
a three-group as the major element of the repetition and a pair of
strings in the alternate space, the size of the two minor spaces
was altered, thus making the distance between the three-group and
adjacent pair shorter than between that and the next three-group. This
immediately threw out the feeling for the old series and made a new
one. The new series thus formed varied with the different subjects,
although no particular difference was noticed between spatial and
temporal types. They all disliked the new arrangement, in whichever
of a variety of ways it was apperceived. (It will be noted the actual
distance between the three-groups was not varied, but the size of the
spaces each side of the minor figure, _i. e._, the minor figure was
shifted from its central position.) One typical spatial subject took
it in either of three ways: (1) He grouped the three-group and pair
nearest together, into the repeated element of the new series; (2)
he ignored the pair and regarded it as a repetition of three-groups;
or (3) ignored the difference in the division of the alternates, and
regarded them as alike. The artificiality of the latter methods of
taking the series is evident. What pleasure survived after such a
strain was very slight, and was moreover not of the series as given,
but as imagined differently, which was not a valid judgment. Most of
the subjects grouped both figures into one, and, finding the unity thus
made ugly and unsymmetrical, derived no pleasure from it. One tried to
keep both elements in separate series and have them go along together,
equally distant from those of their own kind, although not from each
other. This was, however, very fatiguing and unsatisfactory. Those
who grouped the different figures said they did so because they could
not help it, not because they liked it, and said it was impossible
to regard the alternate figure as such, if varied from its central
position. If they were all varied together, they were grouped, with
the major unit, into a new one. If varied irregularly the series was
spoiled--no rhythm whatever remained. It became a heap.

Next, I tried varying the size of the alternate spaces, keeping the
filling in its central position.

[Illustration: Fig. 7]

Here also it was universally regrouped. They found it more difficult
to feel the rhythm of the three-groups as separate elements than when
the minor spaces had remained uniform in size, but different in the
position of the filling. The alternate space, then, which had at first
seemed the unimportant part of the rhythm and for which no subject
could assign any conscious value whatever, was evidently a potent
factor of the experience, and when varied either in size or filling
(especially the former) it not only changed the feeling-tone, but
shifted the entire scheme of the rhythm, or broke it up altogether.

_Variation of major units:_ Was variation more allowable in the major
than in the minor unit of a series? This was tested first in the same
manner as for accents. In a series of which a double string was the
major element, a third string was hung with every such double, thus
changing the unit in both size and content.

[Illustration: Fig. 8]

The series immediately readjusted itself with the three-group as
element for most of the subjects, although one was still able to feel
them all as one unit, varied by the added string. Varying only the
_size_ of the major unit gave the same result.

[Illustration: Fig. 9]

The pairs, instead of remaining the same size, were made alternately
larger and smaller, and a new repetition was made, _i. e._, with the
larger pair as major element and the smaller one as minor. They all
agreed, however, that less change was made in these cases than when the
_minor_ spaces had been changed in size. In the latter case either a
regrouping was made, changing the whole character of the series, or it
was spoiled altogether. With change of _major_ units alone, however,
although a new element was made, it was still possible to take it in
the old way without much difficulty or change in feeling-tone.

It was then necessary to see how change of content would affect the
major unit, the size remaining constant. A group of two sets of double
strings 10 ccm. apart was taken as the repeated element, and these
groups placed at 10 ccm. from each other.

[Illustration: Fig. 10]

Within one element was hung one string, and within the next two, thus
varying the content while the size remained constant. In every case the
answer was the same. It was not so pleasant as when the filling was the
same, but the group still remained the unit of the repetition, and the
series essentially the same.

Several variations were made in this element. Instead of hanging
strings regularly (1 in one, and 2 in the next) they were hung
irregularly, _i. e._, an extra one here and there at intervals in no
special order. As long as the boundary-lines of each group remained
at the same distance from each other, and from the next group, thus
keeping the unit at uniform size, although the pleasure-tone varied,
the balance of the series was not changed. No regrouping or shifting of
the equilibrium resulted.

It would seem from the preceding experiments that in any series
variation of the _major unit_ was tolerated more than of the
_alternate;_ while in _either case_ variation in _content_ had less
influence than variation in _size_.

_Symmetry:_ In the previous experiment, three subjects had insisted on
symmetry as a necessary attribute both of the unit and its alternate.
U. (spatial type) described his experience as "a succession of
symmetrical experiences or states of equilibrium; when they are not so,
they must be regrouped, or pleasure is impossible." R. (temporal type)
insisted especially on the necessity of the alternate figure being
symmetrical as regards the major units, _i. e._, halfway between them;
and also on symmetry as regards itself. One temporal subject said
there was some pleasure in merely going from one unit to the next, even
though no repose was possible on each because of its asymmetry. This
suggested experiments on the importance of symmetry in repeated series.
Is it necessary that the separate elements of a series be symmetrical?
Must both major and minor element be symmetrical? Does this necessity
vary according to the temporal or spatial type of the subject, _i. e._,
is it more necessary to the spatial type, whose pleasure depends more
on repose in the unit, than to the temporal type, whose enjoyment rests
mainly in the rhythm of movement from one unit to the next? Or is it a
common demand? This experiment was begun in the following simple way.
The strings were hung in two group-forms; one with three and the other
with four.

[Illustration: Fig. 11]

This was a symmetrical grouping and uniformly pleasant. The series was
then changed by removing the second string in the four-group, thereby
making it unsymmetrical.

[Illustration: Fig. 12]

This change made the repetition less pleasant in every case, but did
not spoil it. Instead of the four-groups becoming more prominent they
seemed less so, and the three-group on account of its "compactness"
became in most cases the major element, thereby shifting the balance
of the repetition, but not detracting very much from the pleasure.
Next the three-group was changed by moving the middle string to the
left. By this means the group which had been minor in Fig. 11, became
unsymmetrical, while the four-group was regular.

[Illustration: Fig. 13]

This change was preferred to that in Fig. 12, although different
reasons were given. One said it was because this change in arrangement
made the elements more distinct, hence easier to keep apart, while in
Fig. 12 they were made more alike. Moreover, one element seemed as
important as the other. He did not class them as major or minor, so
he could not compare the relative values of symmetry in principal and
alternate units, for in this series he did not feel the distinction.
The other answers to this question were rather incoherent, but the
series did not seem to suffer much change, either pleasantly or
otherwise. Since lack of symmetry in _the element_ was at least
tolerated in the examples already given, would it be allowable so
to place the units that the two adjacent to any one unit should
lie unsymmetrically on either side, that is, may the elements lie
unsymmetrically with regard to one another? Suppose a four-group to be
repeated at regular intervals, and a three-group likewise; if the two
series were combined, must they occur halfway between each other? That
is, must they be symmetrically placed as regards the intervening space,
or could they be put to one side?

[Illustration: Fig. 14]

The subject was asked not to group them (as in previous similar
arrangements), but to keep them as separate repetitions if possible,
and to see if this equal distance was necessary to keep them apart.
The result was the same in all cases. The subjects could not help
grouping them, and found it impossible to keep them distinct unless
so much effort was put into it that no pleasure was left. They said
they "_knew_ each unit was as equally distant from the next unit in
its _own_ series, as if it did not come at unequal distances from the
units in the other, but they could not feel it so, and were obliged
to group the two together." For this reason the experiments did not
satisfactorily illustrate the point in question. It was necessary
to have a series of elements whose unity was more strongly marked,
and whose different parts would still remain one _whole_ even after
variations, instead of shifting into each other. It was suggested
by these imperfect experiments that symmetry was _not_ so important
a factor in the different units of a series as the subjects had
previously supposed; but that, on the other hand, the different units
must be placed at equal distances from each other, if they are to be
kept distinct either as two series or as one. Moreover, that _two_
series could not be kept distinctly in mind as separate, _anyway_,
without fatigue, the tendency being always to group them into one
series with a new repeated element, composed of a combination of the
other two. It was necessary, however, to test this more completely. By
a simple device the former series was changed radically, so that the
difficulties mentioned were overcome. The strings of both the three
and four groups were twisted together at the bottom, thus binding them
closely into separate unities. By remaining attached at the bottom,
whatever variations might occur elsewhere in the figures, they could
not lose their individuality and become merged in each other as before.
They remained distinct groups without effort on the part of the subject.

[Illustration: Fig. 15]

With this I began as in the previous experiment. The subject was asked
to look at the series of repetitions, enjoy them as much as possible,
telling what was the pleasant factor in the experience, and how he
apperceived the series. The subjects separated into types as before;
the spatial type immediately grouping the elements into a larger
unity and enjoying the groups more for their own sake than for their
repetition, while the temporal type went from one to the next in the
series, enjoying the _rhythm_ more than the elements as such. (It may
be remarked here that the subjects were perfectly naïve as to their
apperception. They did not know they were separated into types, nor
were they urged to be consistent. Even the experimenter did not know
of the distinctness with which these types separated themselves, and
consistently held to their own method of apperception, until looking
over the records afterwards.)

With the temporal type the four-group was the major group. Some
expressed its prominence in terms of _time_, _i. e._, they spent more
time on it, and less on the three-group. One felt it as emphasized,
because he moved from one four-group to the next like it, and at each
step moved back and forth from left to right, to see the alternate
three-groups on each side, always _resting_ on the four-groups.

It is noticeable with these subjects, in whom the rhythmic element
was more strongly developed, that although they admitted that the
language of "temporal rhythm" did not adequately cover their experience
(because the element did not disappear after perception as in auditory
rhythm, but remained in the visual field), still they could not express
themselves in other terms. L., the most extreme of this type, insisted
that the experience of repetition would be exactly as pleasant if he
saw the elements pass one by one behind a moving window, with never
more than one in the field at once. In other words, their temporal
relations were all he felt.

The others did not go so far as this, and agreed to the significance of
the whole field, even while especially interested in passing from one
to the next. B. partook of the characteristics of both types, and by
combining the apperceptions of both bridged the chasm between them.

With the four spatial subjects, the apperception showed its usual
divergence. Three grouped the elements, either with the three-group
in the centre on account of its being more compact and graceful, or
the four-group because it was heavier. One of them could group it
either way, distinguishing between the prominence in an element due
to _interest_ and due to _beauty_, _i. e._, he found the four-group
more noticeable and interesting on account of its size, while the
three-group was more beautiful as a unity, on account of its proportion
and grace. Therefore according as one factor or the other predominated,
one or the other figure was taken as the more prominent element, and
placed in the centre of the group. Sometimes they separated into two
series running along together, but this was not usual.

Having got these varied introspections, with yet a certain likeness
running through them, the balance of the four-group and of the
three-group were varied in turns, to see how the change in symmetry of
elements would affect the series; and the relative value of symmetry
for the major and minor units of a series.

First, the four-group was altered, by moving the second string further
to the left, while the three-group remained symmetrical.

[Illustration: Fig. 16]

In three cases this arrangement was preferred to the regular one
previous, and each time for the same reason. The four-group was
made more noticeable by being unsymmetrical, and hence more easily
distinguished from the other. The two were easier kept apart, and the
alternation between the two was made more clear-cut and obvious. With
others the change was unpleasant for the reason that it affected them
in an exactly opposite manner. The four-group _lost_ its individuality,
and, by separation into two unsymmetrical parts, could not be
distinguished so well from the three-groups as formerly, hence the
effect was spoiled.

A distinction was made between the relative _importance_ and _interest_
attached to the units, when symmetrical and when unsymmetrical.
Every one agreed that making the four-group unsymmetrical gave it
more prominence of a certain kind. With the first three subjects
mentioned, this prominence was enough to accent the rhythm still more
than before, and differentiate the two units more strongly. But with
the other there was a feeling that while it gained prominence and was
more _noticeable_, it lost coherence and interest, thence it could not
be kept as the principal unit, but the attention passed over to the
three-group which maintained its unity.

It would seem, then, that the mere fact of one unit in a series
alternating with another, and being more noticeable, taking up a larger
space, being more complicated, etc., did not insure its being the chief
unit in the series. One subject voiced essentially the feeling of all,
in his comment on the series: "There is a constant struggle between
the prominence which the four-group gains from size and eccentricity,
and the insignificance which it deserves on account of its looseness
and lack of unity; it cannot hold its own as one individual thing,
and because the three-group still does, it becomes in one way more
prominent, while the four-group remains so in another." Another subject
felt he gave more _time_ to the four-group than before, because being
separated it would not bind together again without effort. At the same
time the three-group gained in interest because it was easy to find
and did not vary. Another subject felt that the _time_ spent on a unit
had nothing to do with its rhythm; it was all a matter of interest and
attention. Often he looked a longer time at one unit, choosing another
for the chief element in his series, because it interested him more.

All this introspection brought out two things clearly:

(1) The apperception of a series of alternating units, whether of the
spatial or temporal type, is not fixed, but any variation of its unit
is liable to shift the emphasis. Thus, as in the present case, when a
symmetrical major unit is made unsymmetrical, it may not remain the
principal unit, but becomes the minor one, because the attention shifts
to the other which was before relatively unimportant.

(2) Whether either element shall be the principal one or not, does not
depend wholly on its objective prominence, but on the amount of beauty
or interest which it holds for the observer. Neither size, complexity,
nor eccentricity can force a certain unit to be taken as the major in a
series, unless it thereby presents an interest which makes the observer
choose it.

The next change was to vary the three-group in a similar way, by
pushing the middle string to the left, thereby making it unsymmetrical.

[Illustration: Fig. 17]

The responses were as follows: Five said it did more violence to the
series to have the alternate varied than the major unit; it was more
confusing. Three preferred it, giving as their reason that it made
the elements more different from each other than before, hence more
easily distinguished. The preponderance of evidence was, therefore,
that, although any variation from symmetry in a unit was likely to
be detrimental to the repetition, it was more likely to be tolerable
in the major unit than in the alternate space. In either case it was
demanded that the two units be distinctly different, and it depended
on the individual subject, whether in this experiment the variation
of one unit or the other brought out this distinction more obviously.
Aside from this consideration, however, it appeared that the alternate
spaces as such required equilibrium more than the principal unit.
Also, variation of symmetry in the major unit, while it made it more
prominent in the way of _eccentricity_, also made the symmetrical minor
unit more prominent in the way of _interest_. As one subject expressed
it, "Since the others vary, the attention requires something which does
_not_ vary, and forces prominence on the minor unit, because it remains
symmetrical"; and "The minor unit is too small to merit such prominence
as it gets by lack of symmetry. It is distorted, and has not enough
content to bear it."

These introspections from temporal and spatial subject alike, all point
to the fact of a certain _value_ attached to the alternating units in
a series. (1) The units must not be too much alike in _interest_, or
they rival each other. (2) They must not have more prominence given
them as regards the whole than they have interest to sustain. (3) There
must be a congruity between the two elements so that one shall not be
noticeable in one way, and one in another, thus carrying the attention
in two different directions.

One more thing was suggested by this experiment: (4) The subjects who
had invariably _grouped_ their different elements in other series found
it very difficult to do so in this, or wholly impossible. None of them
did so when taking the series naturally, but moved on from one to the
next just as the rhythmic type did. They felt "forced to move on," "no
place to rest," while one in whom the rhythmic feeling was weakest was
much fatigued by this movement, and insisted on having something stable
to rest upon if he was to gain any pleasure at all.

Next the series was varied by making both units unsymmetrical; first
with the balance tipping the same, and second in opposite ways.

Those who preferred this gave essentially the same reason. They agreed
that the unity of both elements was broken up by this change, and they
did not stand out distinctly from each other; but all felt a certain
_congruity_ in having both major and minor units follow the same scheme
in composition. They were not distinctly an alternating series, but
harmonized better as lines. The two spatial subjects, who disliked this
arrangement more than the other, gave the same reason: the unity of the
elements was spoiled, they did not "hang together." Their dislike was
similar in kind to that of the others, only the congruity which made
up for it with the former failed to satisfy these. With the symmetry
broken and the balance tipping in different ways, the feeling was not
strong in either direction. They still criticised it in the same terms
of congruity and distinctness, with no especial change on account of
this modification.

[Illustration: Fig. 18 A]

[Illustration: Fig. 18 B]

These experiments all pointed to the fact that (1) a certain amount
of congruity and equality was necessary between elements of a series
(although it did not establish what were the essential features of such
a harmony). (2) It is more pleasant, as a rule, to have the elements
symmetrical, although symmetry was not a necessity for an agreeable
series. (3) Provided the change in the symmetry of the units was not
enough to shift the whole order of the series, changing the major to
minor units (and _vice versa_), any varying of the symmetry of a minor
unit was more disturbing to the repetition than of the major, while
varying their symmetrical position, as regards the unit on either side,
was absolutely destructive to the order.

The next experiment dealt with a different side of the question. Since
the unit of a repeated series may be a group with repetitions inside
itself, does the repetition of lines or figures _inside_ the group
differ from the repetition of the groups as a whole? If so, how? That
is, in the enjoyment of a series of groups with repeated lines in the
group, in what respect does our apperception of the repetition differ
in the two cases? Or does it in reality differ at all?

To test this, the strings were arranged in the following way. 10 groups
of five strings were hung 100 mm. wide and 100 mm. apart. Each unit
had, then, five repetitions within it.

[Illustration: Fig. 19]

The arrangement was pleasant to all the subjects, and they described
the effect of the experience, falling at once into the spatial and
temporal types as before (this was wholly naïve, for the same questions
were asked of each, and they had no idea of being grouped in types).
The introspection of both types must be taken in some detail, to fully
analyze the experience.

_Spatial:_ J. felt he took in all groups at once. Each unit seemed like
a rich experience in itself, but he could not detect any rhythm in it,
nor in the whole series. The pleasure consisted in getting a number of
similar objects in the field at once, and enjoying the combination of
them all, feeling that they stretched away in each direction. H. and U.
grouped several unit-groups into a larger unity and enjoyed the cluster
as a whole. They did not group them in any particular system, nor could
they detect the slightest pleasure in moving from one such group to the
next. One found his enjoyment solely in the contrast effects in each,
while the other laid it to the space relations of each independently.
The pleasure only came when each group of groups was spread before
him. Those outside the immediate field meant nothing to him, and the
movement between them had absolutely no conscious interest for him.
S. said that enjoyment stopped altogether during motion of any kind,
and the experience was pleasant only during total repose, on whatever
happened to be in his field at once.

With the temporal type came a marked difference in apperception.

B. affirmed the pleasure to _consist_ in going from one cluster to
another, and to _begin_ just at the point where he meets the next
stimulation and feels it is _going to be_ the same as the one previous.
It is the _expectation_, rather than the _verification_ recurring at
intervals, which makes up the pleasure; not the actual movement, or
subsequent contemplation of a group. The pleasure came in pulses; in
knowing by seeing from the side of the eyes that the experience _is to
be_ repeated, and on reaching the edge of a new group, in the feeling
that the experience is just about to begin.

R. felt that she "wriggled around" in each group of lines, and that a
certain _feeling_ came from "wriggling" among the lines in a particular
fashion.

The pleasure consisted in having this feeling recur at regular
intervals. The repetitions inside the group and of the group as a whole
differed in this respect: For the separate unit-group, it apparently
consisted in repeated short irregular movements, back and forth, enough
to bring about a certain feeling which seemed pleasant and sufficient
unto itself. Repetition of the groups as a whole meant movement across
the field in one direction, for the purpose of meeting another group,
and getting the required feeling from it again. The pleasure was not in
the movement or in any repose (she could detect no repose at all), but
in experiencing the group again, feeling that it had been so before,
and would be again.

L. (the most extreme of the temporal type) agreed with R. that the
lines inside the group were perceived and enjoyed _temporally_, as well
as the groups as a whole. There was no experiencing the groups _at
once_. He felt that he moved regularly across the field encountering
five lines, one after the other, then an empty space, then five lines
more. The only meaning which the group as such had for him was the
five accents which came near one another in time. He could feel no
unity whatever apart from this. He was even certain that his pleasure
would be identical if in some mechanical way the same figure could be
pushed forward, so that the same amount of time and movement would be
necessary to reach it that was required to move from one figure to
the next on the field. The experience was in every way analogous to
auditory rhythm with him, and he was unable to express himself in other
than temporal terms. Immediate perception, repose on the object, or
groupings, had no significance for him.

The other two subjects were links between the extremes already
described. They could feel each group, and sometimes even the whole
series _at once_ apparently, and yet were all the time conscious of
a certain rhythm in going from one to the next. The whole experience
seemed immediate at first, but on reflection a certain alternate
rhythm was felt to be present, which was too rapid to take any
considerable time, but yet had to be included as a factor in the
experience. These introspections I believe to throw light on the nature
of the whole experience of repetition. Since there are two methods
of apperception so extreme, but moreover certain subjects partake of
the characteristics of both, it might seem that both types represent
but _one_ side of the experience. Since both are enjoying the same
objective series, but in their description of their feeling in face of
it emphasize such different sides (leaving at the same time the other
side unaccounted for), and since certain subjects share the experience
of both, it might be that the sum of both methods of apperception was
necessary to the fullest appreciation of the repetition in question,
only in certain subjects one aspect of it was so much stronger that the
other possible factors in the experience were overlooked.

It would tend to bear out this view, that when it was suggested to
those of the temporal type (always excepting L.) that according to
their description the other groups remaining in the field, after
having performed their part in a temporal series, _ought_ to have no
further influence in the repetition, whereas they did in reality, they
admitted the fact, but could not account for it. Moreover, those of
the spatial type admitted that their enjoyment in having spaces equal,
and in having repeated objects exactly like one another, had a certain
character which no other experience possessed. This did not seem
accounted for by any description they could give of its effect on them,
although they could not detect what this other elusive factor might be.

These introspections, therefore, and the confessions on the part of
both that there was a feeling of something _more_ which they could
not hold long enough to describe, suggested that both types were but
opposite ends of a series of possible apperceptive types, and that in
both cases certain essential features were emphasized at the expense of
the others.

After these experiments, the next step was to find how a series of
groups was apperceived when the lines in each group were arranged
symmetrically about a centre, as distinguished from their arrangement
at invariable distances apart. The same number and size of groups were
taken, but the arrangement of lines in each varied as stated above.

[Illustration: Fig. 20]

_Spatial subjects:_ J. felt a different kind of pleasure from that
felt with Fig. 19. Here the enjoyment was in each unit for itself, a
certain repose in its symmetry. Although he fixated on the centre of
the groups, and in going across the field moved from centre to centre,
there was no feeling of rhythm whatever, merely enjoyment of the unit
itself. Moreover, although he had _detected_ no rhythm in the previous
experience, this one seemed distinctly different in having lost a
feature that the other had. He felt by comparison that the other had
had a temporal character, some movement in the groups, which was wholly
lacking in this. This was more beautiful and restful, the other more
exciting and rich.

H. and S. both enjoyed this series better on account of possibility
of greater repose in the unit-groups. The pleasure was solely in each
unit for itself, not in their repetition, so the group which offered
most balance and equilibrium in itself was pleasantest. S. also found
enjoyment in slight variations in the groups (trifling difference in
distance, different light effect, etc.). It is noticeable that when
repetition alone was the main feature of the series, any variation was
either ignored or found unpleasant. But when the unit for itself is the
object of enjoyment, variation if slight is another element of pleasure.

There is also pleasure in the mere _repeating_ of symmetrical groups,
although, when the attention is turned to this feature, the feeling of
_symmetry_ is less felt. Even when attending chiefly to the repetition
of the groups, the symmetry of each is felt somewhat, which makes the
whole experience better than Fig. 19, but the two attitudes seem to
hinder, rather than help each other.

This introspection was suggestive, giving rise to two more questions:
(1) When is variation allowable, and when not? Is it adapted only to
objects when taken, as ends-in-themselves, and not when considered as
means to something else, _i. e._, as means to make a series or border,
or anything which takes attention from themselves as unities with
individual meaning? (2) Is a distinctly symmetrical group as adapted
to repetition, as such, as one with merely equal divisions? Does it
not tend more to repose in itself, instead of to the motion necessary
for the apperception of a repeated series? These questions will be
considered later.

_Temporal type:_ R. felt a difference in the movement across, in that
in Fig. 20 it was from the centre of one unit to the next, while in
Fig. 19 there was no regularity in the movement.

L. felt the difference between the two apperceptions very strongly. In
Fig. 20 the movements seemed organized. He felt as if his attention
(if not his eyes) went back and forth from edge to edge of the unit,
finally settling in the centre; while in Fig. 19 the very essence
of the apperception was that every line was compared with every
other, meaning a great number of movements in both directions, _not_
stopping in the centre. If he did rest finally in the centre, in the
unit of Fig. 19, instead of seeming evenly repeated, it too became
symmetrically perceived, but the usual way to get from one such unit to
the next was to move _from_ and _to_ any point in the next adjacent,
other than the central one. In either case the pleasure came in
identification of the second figure with the first, and the feeling, "I
have seen it before." The pleasure lay in the process of recurrence of
sameness.

V. also, who had not felt much motion in Fig. 19 at first, felt it
strongly now in comparison with Fig. 20. He said in the former,
although he did not make actual movements across (in fact his eyes
were plainly at rest), he was sure he felt "dispositions to do so"
which were lacking in Fig. 20. The pleasure came in the first moment of
repose after finding the new unit was the same as the old.

After we had investigated the different methods of apperceiving groups
of repeated lines, and compared the effects of different groupings,
and studied the feeling of one unit alternating with another, another
question arose.

These questions had all referred to the alternation of _two_ units;
either a unit with an empty space, or with a space of different
filling. How did the apperception differ when _three_ repeated units
alternated with each other? To test this, three spaces were taken
equally wide (110 mm.) and equally distant from each other (150 mm.),
but with three different designs within them. These designs were of
the same general character and importance although different, and
repeated themselves regularly.

[Illustration: Fig. 21]

The subjects were asked as before to describe their reaction on the
series. Not one of all the number was able to _feel_ the repetition
of the three units; what pleasure they got from the series (if indeed
they got any) was from other sources. The general type of answer was
formulated more fully by L. He saw 1, then looked for 2 and found it
different, but could have included it in the series if it were not for
3. That being still different sent 1 and 2 out of mind, so that he
could not _feel_ any repetition of 1 when he met it again. He felt a
certain sense of repetition in that the spacing and general motifs were
the same, but there was no pleasure in that. What pleasure he got was
wholly intellectual, not immediate, except for a slight pleasure in
their uniformity of position. In him the rhythmic feeling had always
been of the strongest, but he found in this experience none whatever.
It was simply impossible to keep the three units going at once. Another
temporal subject tried to group 2 and 3 as one element, with one as
an alternate, thus reducing it to a rhythm of twos. This process was
labored, but otherwise no enjoyment was possible. The spatial subjects
derived what pleasure they could, either from the units separately,
with no regard to their repetition, or from some method of grouping, by
which their difference could be overlooked. One expressed his pleasure
solely in terms of contrast of the white strings against a black
ground. Any immediate feeling for the repetition was impossible for
either type. It will be noticed that the _feeling_ of the repetition is
quite different from the _knowing_ it is there. They were all perfectly
conscious that 1 was repeated again after 3, but could not _feel_ it,
while repeated simply after 2, they _could_ feel it.

Next, the series was varied again. The size of the blocks, instead of
being alike, was varied three ways, while the designs remained similar.

[Illustration: Fig. 22]

The interspaces were 150 mm. in every case, but 1 was 150 mm. wide; 2 =
110 mm.; 3 = 70 mm. All the spatial subjects found Fig. 22 worse than
Fig. 21. The irregularity and general disorder was more pronounced.
Although the rhythm had never consciously given them pleasure, and,
when not violated, was never noticed, still the threefold difference
in size violates some feeling which they can only express in rhythmic
terms. Some tried to group the three units into a larger group, but
this being unsymmetrical displeased them. Others picked out the most
satisfactorily proportioned unit and ignored the others, but any
possible apperception was irritating. The temporal subjects found it
equally poor. They felt the continual dissatisfaction of having their
expectation, that the adjacent unit should be the same, disappointed.
They all said that they could carry the feeling of repetition over
_one_ dissimilar unit (_i. e._, in an alternating series of two
different units), but that the third difference completely upset
the scheme. When only the filling varied as in Fig. 21, it could be
partially ignored, but difference in size could not be ignored, and
only the equal distances apart kept them from being a heap. They could
not _feel_ the evenness of the empty interspaces, however. They were
not consciously present in the experience at all, they merely _knew_
they must be even. There was no feeling-tone whatever to the empty
alternates.

Only one subject preferred Fig. 22 to Fig. 21, and the reason was
obvious. 1, 2, and 3 appeared as the same unit where variation in size
was apperceived as due to perspective. Thus, instead of appearing as
three units repeated, they were one set which progressed by means of
"pulsations" or regular intervals of perspective. This gave an added
richness to the rhythm, and was very pleasant. As three separate units
of different size, there was no meaning in the series whatever.

It is evident from these introspections that, although the likes and
dislikes may vary, the principles on which they are based have much in
common.

The points on which they agreed unanimously were the following:

(1) There is no feeling of repetition for three separate units. The
series may be enjoyed by means of subjective grouping of one kind or
another, but as separate elements, the feeling of repetition is broken
by adding the third.

(2) There is a distinction between perceiving or _knowing_ a
repetition, and _feeling_ it. Even though a subject is equally
conscious that elements are repeated according to some scheme in two
different cases, he may feel it in one case and not in the other.

(3) The empty spaces between the elements have no conscious part to
play in the experience. Even when there is a figure in the alternate
space, it comes very little into consciousness as part of the
repetition, yet it is alterations in these alternates which make or mar
the feeling of repetition. A series may not be beautiful in itself, but
if the alternates are regular, it _feels_ repeated. _Vice versa_, the
units may be enjoyable in themselves, but they do not feel repeated
unless the alternates are regular and conform to certain requirements.
In the units lies the _meaning_ of the repetition, in the regular
alternates the possibility of its expression.

(4) The rhythmic character of repetition is not felt by a certain type
of subject, when it goes smoothly. When a variation is made which would
destroy any possible rhythm, its lack is felt, and its violation finds
expression only in rhythmic terms.

(5) More violation is done to a series to have the _size_ of units
varied than the filling. (This corroborates previous experiments.)

(6) A certain amount of ignoring and regrouping can be done by the
subject. The series is not taken exactly as given, but with selective
attention.

(7) In a series of different elements alternating, the most prominent
one is chosen for the major unit, and the others for alternates. This
prominence is more influenced by _size_ than any other factor, but may
be due to intrinsic interest of any kind.

(8) The major and minor elements must have a certain difference from
each other, both in _appearance_ and _interest_, and they must be
different enough for the difference to be easily perceived, but not
enough to be incongruous. They must differ in interest enough, so that
one is easily more prominent than the other, or may be made subservient
to the other, in the apperception.

(9) Variations are pleasant in the principal unit repeated, but not in
the alternating figure unless very slight indeed, or affecting only
secondary parts of the figure, not the main lines.

(10) Not the time actually spent on a unit makes it more or less
prominent, but the feeling of more or less "energy" expended on it.

_Ends:_ In an alternating repetition, must the series end on a light or
heavy beat? That is, must the major or minor unit be on the end?

To test this a series of strings was hung in which a group of three
alternated with a single string.

[Illustration: Fig. 23]

The subjects were asked to look at it with the three-groups on the
end, and with the single string. In every case the three-group ending
was emphatically declared the best. What reasons were given were much
the same, although most of them could give no explanation at all. S.
said the minor space on the end left him "hanging in mid-air, it needs
the heavy beat to land me again." Others said it was "ragged" unless
the three-group ended the series. R. said anything _interesting_ would
do on the end, as well as the larger-sized unit, it simply needed
something of sufficient interest to stop the rhythmic process and keep
one from going on.

It was impossible to describe the experience except in rhythmic terms,
and those in whom this sense was not strong could give no account
whatever for the difference in their feeling for _end_.

It will be remembered that some experiments were previously described
relating to the difference in apperception of a group of lines equally
distant from one another, and a group averaged at equal distances each
side of a middle point, but unequally from each, to emphasize the
bilateral symmetry. Two such series were now taken to find if there
were any difference necessary in appropriate endings. Since the two
types of groups differed so much in apperception, did that difference
so extend to the whole series that a different space was needed at the
end to finish them off?

The method of experiment was the following: Two series of repeated
groups were hung (100 mm. wide and 100 mm. between) with the design
of the groups varied as described. At the end of each a strip of
cardboard was hung, which the subject was asked to move so that it
bounded the amount of space at the end, necessary to finish the series
adequately.

[Illustration: Fig. 24 I]

[Illustration: Fig. 24 II]

Thus _b_ is the cardboard strip, and _a_ the space which was to be
varied according to his taste. The same experiment was tried with each
series, with the following results:

                 I                      II
    U.  _a_ =  96 mm.    _a_ = 90
    J.              33                   50
    S.              97                   90
    H.             109                  104
    R.             160                  150
    V.             170                  135
    T.             145                  125
    W.              80                   68

In the case of every one but J. the subjects preferred a longer end
space with I than with II. J. was, however, of the extreme spatial type
who gave as his explanation that with II, when the central line was
prominent, the end (_a_) must equal just the distance to another middle
line, while with I it must harmonize with the shorter distances in the
group, but not exactly equal them, for that would make it too narrow.

It would mean, then, that the apperception of the repeated group in
I (if it accords with the subject's own introspection) consists in
repeated fluctuations of attention over the five strings, with no
repose on any one more than another. The movement is back and forth
from edge to edge, and hence needs more of an end to finish it than in
a series of symmetrical units where the movement is not back and forth,
but balanced and resting on the central point. In other words, in
Group I there is a rhythm of movement within the group itself, as well
as of the whole, while in II it is balanced and coördinated from the
centre of each group, out and back, so that a longer, or at least more
important end of some description is necessary to break the rhythm, and
stop the series in I than in II.

It is noticeable also that H. and S. thought in both cases they were
making the end spaces equal to the interspaces; but after Series I,
_a_ was made 102 and 109; and after Series II 95 and 104 respectively.
This naturally raised another question: Does a series of groups, with
repetitions within each, tend to make one overestimate distances
between or at the end, or at least does one overestimate these
distances in comparison with a series of symmetrical units?

The subjects were so unanimous in preferring a larger "embankment"
after Series I than II, that it was useless to test them further on
that point, and the experiment was changed to the other question
according to the suggestion above.

A series of eight cards was prepared (125 mm. wide) on four of which
five heavy black strips were drawn equally distant from each other, and
on the others a much wider strip in the centre with another on each
side near the edges. Of the two series just made, one was composed of
what we have called "rhythmic units" and the other of "symmetrical
units."

[Illustration: Fig. 25]

The subjects were asked to arrange the two separate series so that
the interspaces should be exactly equal to the units. It will be
observed that the rhythmic unit had a black strip on each edge, thereby
apparently decreasing its size, while the edge of the symmetrical unit
was white. In this respect the comparison was hardly fair, but the
result was the following. The figures represent an average of two
trials, and stand for the size estimate of the interspaces for each
subjects respectively, in the two series.

               I                II
    J.        132.5            130
    S.        125              122
    U.        133              129
    H.        128              124
    R.        126              122
    W.        136              133
    V.        129              131

Average difference of estimate of both series = 2.64. Mean variation =
1.37.

It might be contended, however, that Series I is an example of optical
illusion, that the card was overestimated for that reason, and the
interspaces necessarily made wider. To avoid this difficulty another
series was made. Two sets of cards (125 mm. wide) were prepared; one
with five black strips at equal distances apart as before (excepting
that the strips were made heavier), the other with six strips. The card
with an odd number of strips had thereby a strip in the middle upon
which the attention could centre,--possessed a kind of balance. The
card with an even number of strips had, moreover, no such central line
but only a space, thus preventing repose of attention, and making the
unit more pronouncedly rhythmic. (It will be noticed in the foregoing
table that one subject, V, made narrower interspaces in I than in II.
He said he felt the units as centring around the group of three lines
in the centre, not as proceeding equally to the edge. The unit became
thus for him symmetrical instead of rhythmic, which could easily
account for the difference in estimation.)

The results in the present case are an average of three trials:

               I, 5 strips      II, 6 strips
    J.         129              137
    S.         125              129
    U.         132              133
    H.         125              131
    R.         127              138
    W.         126              133
    V.         123              129

Average difference in estimate of both series = 6.1. Mean variation =
2.1.

Since both these figures represent an effect usually explained by
optical illusion, that factor may be counted out, and the difference
in the estimate be accounted for by the difference in the rhythm of
the units. The difference in estimation between the two rhythmic
units, differing only in odd and even number of strips, is greater
than between the rhythmic and more strictly symmetrical, and yet the
two were more comparable in construction. It would seem, then, that
the greater overestimation of II is due to the rhythmic movement which
is not limited or driven back to a central line as in I, but, by
continuing over the limits, produces a greater feeling of breadth.

The same question was experimented on in another way. Smaller strips
of cardboard all 50 mm. wide, but with different designs, were hung
behind the narrow window previously used. Four of each set were hung
at a time behind the window, and subjects arranged them so that the
interspaces appeared to equal the strips. These designs were to
illustrate different points in question. The difference in estimation
for an empty card, and a filled one; the difference according to the
strongly centred, or rhythmic, or slant lines of the filling. These
experiments were not so complete as the former ones, since the subjects
were scattered; hence they represent only one trial or an average of
two. But the results conform with what we have been led to expect.

[Illustration: Fig. 26]

                       I     II     III    IV      V
    J.                49     57     52.5   52     53
    S.                51     56     50     49     51
    U.                53     54     49     50     48
    H.                       53     52     49     51
    R.                52     52     51.5   48     51
    W.                50     55     50     50     52
    V.                51     53     49     48     51

    Average =         51     54     50.6   49.4   51
    Mean Variation =   1.5    1.4    1.2    1.06    .9

These results point to the fact that there is a tendency to
overestimate the strips unless there is a strong central accent, which
draws the attention back to the middle of the strip, in which case it
is slightly underestimated. This would seem to be contradicted in I,
where the centre is strongly marked by slant lines coming toward it.
But the subjects, instead of taking the lines as pointing _towards_
the centre, in almost every case felt them as leading _away_ from it,
and the oblique lines gave an appearance of greater breadth, which
result was carried out by the greater overestimation of II. In this
case, in addition to slant lines, there was no central accent, and the
overestimation was proportionately large. III and IV were intended
to illustrate the difference in estimate of rhythmic and symmetrical
units, but although a slight difference is apparent, the subjects did
not feel III as strongly rhythmic, because the black lines on the
ends of the strip were ignored against the black background, and only
the two central lines were taken. This made it more a balanced than a
rhythmic unit, so it is not a fair type of the point in question.

We may say, in conclusion, that oblique lines (which involve a more
complex muscular adjustment to perceive them) give an impression of
greater distance traversed, hence are overestimated; of two rhythmic
groups, the one containing an even number is more overestimated than
the odd, since the movement across is unchecked, and not balanced
around a central line; a series of strongly centred groups is more
correctly estimated as to its interspacing, and even slightly
underestimated, because of the check imposed by the centre of fixation
in each group. Although these results are very uniform, a more complete
series of experiments should be done on this subject, to make the
conclusions thoroughly valid.

[Illustration: Fig. 27]

Another question was suggested by these results: Is it more agreeable
to have a series of repeated space forms nearer or farther apart
when a design is within? Does the design, by drawing attention to
itself (especially if it be markedly central), make the objects
demand narrower or wider interspacing? To test this question, four
blank strips of cardboard were hung behind the narrow window, and the
subjects arranged them at the distance apart which suited them best.
Then two other sets of cards, of the same size, but of different
designs, were hung successively the same way, and these arranged also
at the most agreeable distances. One decorated card had a circle within
a rectangle, the other a triangle of gilt stars. The judgments were
made in pairs, _i. e._ the blank cards and the one with the circular
design were arranged twice in succession; then the blank cards and the
star design. This gives three judgments for the blank cards, two for
the circular, and one for the star design, and the judgments are given
in full, since an average would disguise the point in question.

               I     II    III
    J.        50     50
              55     52
              55            50
    S.        35     70
              45     20
              55            35
    H.        85     75
              90            65
    U.        50     45
              50     48
              57            50
    R.        45     25
              15     10
              45            23
    L.        90     35
              65            65
    W.        20     10
              25     10
              25             0
    V.        45     60
              30     35
              30            25

Although the favorite arrangements varied somewhat on the different
days, the _filled_ cards were with only four exceptions preferred
nearer together than the _empty_ ones at any one trial, and two of
these put them equal. (The choice of V. was affected by the fact
that the circular designs produced such strong after-images that he
was obliged to put them farther apart, to avoid confusion with the
real design.) The reason suggested by the subjects for a narrower
interspacing with decorated cards was that, when they attended to the
design, they paid no attention to the actual edge of the card, but
the card ended so far as its interest was concerned with the design.
Therefore they had to be nearer together to bring the _designs_, not
the whole field of the cards, into a series. If, then, the design
extended over all the card, and its interest was no more in the centre
than the ends, would this difference in interspacing cease to be
demanded?

Another series of cards was hung with a design of oblique lines over
the whole field, and these arranged as the others were, at the most
agreeable distances.

           I             IV
    J.    53 mm.   J.    53 mm.
    S.    45       S.    47
    U.    52       U.    50
    R.    35       R.    37
    W.    23       W.    43
    V.    35       V.    34

[Illustration: Fig. 28]

If, for the sake of comparison, the average be taken of the favorite
arrangements of the four blank cards, and they compared with the
interspacing of the oblique line design, it will be seen they approach
each other closely, except in the case of W.

These experiments would seem to show that an empty space, or one
completely covered with decoration, is taken in its entirety when
repeated in the series. But when decorated, especially toward the
centre, the _design_, instead of the whole including space, is taken
as the repeated unit, and for this reason the different units must
approach each other to make a satisfactory series.

To what extent does change in _level_ and _plane_ affect the units of
a series? To test this, a series of diamond shapes was hung on the
same level and at equal distances, and the subjects enjoyed them as a
repeated series.

[Illustration: Fig. 29]

Then another row was hung above them, and halfway between.

[Illustration: Fig. 30]

The subjects grouped them either in twos or threes, thus transforming
them into one series of similar group-units of triplets and pairs. They
were asked if they could take them up and down, one after the other
without grouping, as they would have done when on the same level. With
a little practice two of them succeeded, but they found the series
tiresome when taken in this way, and deprived of much of its pleasure.

The series was then changed by hanging a smaller diamond between the
others, at the same level.

[Illustration: Fig. 31 A]

[Illustration: Fig. 31 B]

This was enjoyed even more than the other, as an alternating series,
but when the smaller diamond was hung between but on a higher level
although it could still be included if _grouped_ in some way with one
or two of the larger diamonds, it baffled all attempts to include it as
an alternate minor unit in the other series. The two series separated,
and one ran along above the other, or else a definite grouping took
place, so that the large and small diamonds made one group-unit which
was repeated. But to combine two different elements as major and minor
units of one series, when the two were on different _levels_, was
generally declared impossible.

Provided units stay on the same level, however, a difference in _plane_
does not prevent their being in one series, provided the plane varies
regularly, and the variation is not too great. The variation in plane
of a few inches, used with these shapes, did not prevent their being
taken as one series, although it much facilitated their being taken as
two, if desired.

These experiments have all pointed to the fact that our pleasure
in repeated space forms is an immediate experience. We do not look
over the series and merely recognize that regular repetitions occur,
but there is an immediate _feeling_ of repetition, analogous to our
feeling of auditory rhythm. This feeling does not always accompany
a recognition that certain repetitions occur, but is a distinct
experience in itself dependent upon certain conditions in the series.
The series excites a certain response in the observer, which, if it
corresponds with his rhythmic organization, is pleasant, and if not, is
otherwise.

With a certain class of subjects this rhythmic response is very
noticeable, and they feel it as a conscious part of the experience.
With others, the symmetrical properties of the series are the more
prominent, and they detect no rhythmic response until the necessary
objective conditions for such a rhythm are violated. Then they feel it
as keenly as the other type.

In a series of units, there is a rhythmic discharge of energy on each,
the fixed temporal intervals being regulated by the alternating space.
When the units are too near together, or when the alternating spaces
vary irregularly in size, this rhythm is disturbed. If the alternating
spaces vary regularly in size, a richer rhythm is introduced, which
increases the pleasure up to a certain point when more variation makes
it too complicated, and confusion results. When one element alternates
with another, the one on which more energy is expended for any reason
becomes the principal unit. The other has less significance as to
its content than as to its size, for on this depends the regularity
of the rhythm. Variations in the content of alternating figures must
be cautious not to disturb, by the extra attention necessary to take
them in, the rhythm of the whole. Variation in the principal unit may
take place almost without limit, provided there is an equal amount
of interest in each, thus making a rhythm of equal discharges. There
must be an alternation of _two_, _i. e._ of discharge of attention and
rest. However rich the rhythm is made by greater and lesser accents
or groupings, the rhythm must fall eventually into a discharge of
attention, and a rest-period. In the temporal type of subject, to whom
the actual _motion_ across the series is a necessary factor of the
pleasure, this discharge and rest-period of _attention_ is exactly
inverse to the _motion_ across the alternate and rest upon the unit.
That is, on the principal unit is the discharge of the attention, but
the rest-period of the motion across; while the alternate unit supplies
the field which is travelled over, but requires but little attention.

The rhythm of the series may be not only of the units, but inside the
units as well, in groups of elements. The rhythm inside such a group
may be of two kinds: (1) a rhythm, which is at the same time restrained
and coördinated about a central point or line, and (2) a rhythm which
goes back and forth from edge to edge, and has a tendency to overstep
its limits, thereby carrying on the series with more activity. The
former is more connected with odd-number groups, and the latter with
even, although modifications in their arrangement may reverse the
effect. Since the eye moves more slowly and intricately over a curve
than over a straight line,[86] it may be that that is the reason why
an arched series is taken as the unit of a series, rather than the
vertical supports (as it invariably is in some unfinished experiments
not recorded here), whereas in a series of lintels the horizontal line
of the lintels requires less muscular adjustment to perceive it than
the vertical support, therefore the latter are taken as the units.

In any case, the unit of the series which attracts the most attention
and interest, for whatever reason, is taken as the principal unit, and
may vary in details, while the alternate must be invariable, except in
certain cases where it meets other demands. There may be rhythm in a
series, and at the same time symmetry with respect to a middle point.
In such a case a balance must be obtained between the two different
motor responses.

A series of analogies between the rhythm of sounds and of visual
objects, will illustrate more forcibly the similar demands of both.

(1) Auditory Rhythm: Periodicity is necessary. Accentuation may be
supplied by the subject, but there must be fixed temporal intervals,
and if the temporal conditions are not fulfilled, no impression of
rhythm is possible.[87]

Visual Repetition: Alternate spaces must be of invariable size, or
the series is broken up. Different degrees of interest may attach to
the principal unit, or the subject may group them in different ways,
but the alternate spaces must be uniform, or there is no feeling of
rhythmic repetition.

(2) Auditory Rhythm: Sounds must recur within a certain rate. When
succession falls below, or rises above a certain rate, no impression
of rhythm results.[88] A certain voluminousness is necessary for very
slow measures, to make the separate elements connect themselves in a
series.[89]

Visual Repetition: Objects must recur at certain proportionate
distances from each other, to connect themselves into a series. Larger
objects may be at a greater distance from each other than smaller ones,
and still form a series.

(3) Auditory Rhythm: "Perception of rhythm is an impression, an
immediate affection of consciousness, depending on a particular kind of
sensory experience. It is never a construction or reflective perception
that certain relations of intensity, duration, etc., do obtain."[90]

Visual Repetition: The feeling of rhythm in a visual series is
immediate, and wholly distinct from the knowledge that certain
objects do recur. This is especially illustrated in repetitions of
three distinct units, when subjects saw and understood the scheme of
repetition, but could not feel it.

(4) The number of units in an auditory group depends on the rate of
succession, but any higher number of elements in a group than _six_
or _eight_ falls back into smaller groups.[91] _Eight_ is about the
highest number that can be held in a rhythmic group.[92]

Visual Repetition: _Eight_ is the highest number that can be held in a
rhythmic group, and some subjects can only hold seven. Many more units
can be felt in a group, when the size of the including space is taken
as the measure and compared; but no more than eight can be felt and
recognized as the number of units it is. (There may be exceptions to
this rule in either auditory or visual rhythm, but this is the norm in
both cases.)

(5) Auditory Rhythm: In all long series, there is a subordination
of the higher rhythmic quantities, so that opposition of simple
alternate phases tends more and more to predominate over triplicated
structures.[93]

Visual Repetition: However complicated the repetition becomes, with
regular variations of the size of major or alternate units, the units
tend to re-group themselves, and so resolve ultimately into a simple
alternate repetition of two group-elements.

(6) Auditory Rhythm: "The introduction of variations in the figure of a
group does not in any way affect the sense of equivalence between the
unlike units."[94]

Visual Repetition: Changes in the content of the major unit do not
affect the repetition provided the alternate space remains invariable.

(7) "Feeling of rhythm is more definite as we proceed in a verse, or
in a series of simple sound sensations. At first the cycle is not
perfectly adjusted, and complete automatism established."[95]

Visual Repetition: Observers often had to look over a series several
times "to learn it" as they said, before the rhythm was felt.

To these may be added several other analogies, which, owing to the fact
that visually repeated objects remain in the field, while auditory
rhythm is purely successive, have other features which somewhat confuse
the resemblance. The principle, however, is the same in both.

(8) Auditory Rhythm: "At the close of a period, we have a pause, during
which the tension between the two opposing muscle-sets dies out, and we
have a feeling of finality."[96]

Visual Repetition: An alternating series must end on the heavy unit,
but since one does not look at series from left to right any more than
from right to left, a heavy unit must be at both ends, not on one
alone. In auditory rhythm, this final pause is not a function of any of
the intervals of the period, for it comes at the end, when the sounds
are no more present. But in visual repetition, after feeling the series
rhythmically, it is still in the field, either as an unending series,
or as a whole, in which each part is equally related to all the others.
The final pause of a series that ends must be at _each_ end, and the
series perfect from either point of view. It therefore fulfils the
demands of symmetry as well as rhythm, but since symmetry in its strict
sense has no meaning for sound-series, this double finality of visual
repetition cannot be analogued in auditory rhythm.

(9) I have found no recorded experiments of rhythms of sounds of
different timbre and pitch, _i. e._, a regular rhythm of a bell with a
violin, a piano with a whistle, etc. It would seem, however, that such
a succession would produce the same irritation as a visual repetition
of incongruous elements; as a circle introduced into the Greek fret,
or a series of Renaissance columns and Egyptian monoliths. In both
cases, the whole set of adjustments for each element would be thrown
into confusion by encountering the next one, which would require a
different attitude. Such a readjustment would be impossible in the
space necessary for the perception of any rhythm, hence there must be
congruity in elements, either auditory or visual, to be in a series at
all.

(10) Auditory Rhythm: "If every alternate element of a temporally
uniform sound-series receive increased stress, the interval which
succeeds the unaccented sound will appear of greater duration."[97]

Visual Repetition: The distance between unit groups with a strongly
accented centre appeared shorter than between rhythmic groups where the
movement was not restrained at the centre, but went from edge to edge.
The principles which explain these similarities are, however, different.

In the auditory rhythm, the stronger accented beat excites a greater
response than the unaccented. This lasts over longer in consciousness,
and for that reason the interval after the accent seems shorter.

In visual rhythm, however, the symmetrically rhythmic group drives
the attention in toward the centre, and whatever excursion it makes
to either side, it returns finally to the centre. In the even-number
rhythmic group, there is no such central line to restrain it, and
as one goes across it one has less check at the edge, the rhythm
does not wholly finish, and the space is thereby overestimated. The
_overestimation_ is due to the rhythmic activity in the group which
oversteps its limit.

The essentially rhythmic character of the experience is, however, the
same in both. The experience of visual repetition is only one-sided
when the symmetry or proportion of a finished series is regarded as the
explanation of its essential character, and when the temporal rhythmic
factor is neglected.


PART II

The purpose of the latter half of this discussion of repetition is to
consider a certain number of examples of its use in typical buildings
of all the European styles of architecture from Greece down, and to
show that the principles laid down in the earlier half have been
expressed almost without exception in those of recognized merit. In
other words it is to show that the laws of repetition, which have been
brought out in the experiments of the first part, and which would of
necessity be true if that explanation were correct, have indeed been
exemplified in types of architecture universally accepted as beautiful.


The illustrations have all been drawn from architecture beginning with
Greece, and not from the older Eastern styles. Egyptian architecture,
although it recognized the importance of repetition to some extent, in
its colonnades, avenues of sphinxes, and hieroglyphic decoration, never
reduced it to any principle, nor adhered to any one scheme throughout
a piece of work. Supports of the same kind and diameter have no fixed
relation to each other, they may be of the same or different lengths,
and may vary in diameter as well.[98] Spaces between columns of one
size and design may vary considerably, and the entablatures be of
different proportions. The art of Egypt was not rhythmic.

The architecture of Assyria and Chaldea had even less of repeated
forms in its style. They made but little use of columns or piers, and
had few arches.[99] The bare Assyrian edifice was like a great box,
perpendicular to its foundations, and the long walls pierced by hardly
an opening in the way of windows or doors.

Persian architecture was noted for extreme nicety of execution, but
a monotony in all its forms, and conventionality about its use of
the column, which makes it little more fruitful for our study of
repetition as an artistic value. In its decorations of bas-relief, the
pose and gesture of each figure is so exactly similar that they appear
almost machine-made.[100] When a little variety is introduced, it is
evidently done with misgivings, and shows none of the spontaneity or
first-hand pleasure in either repetition or variation which would make
it profitable for illustration.

Such a lack of feeling for repetition is, indeed, according to the
peculiar genius of these styles of architecture, what might have
been expected. The ruling idea, especially in Egyptian and Assyrian
architecture, was ponderous strength. Everything was built with the
idea of remaining immoveable through centuries to come. The enormous
temples and tombs, the long palaces with their heavy walls without an
opening to relieve them, the pyramids themselves like mountains of
rock--all these meant strength and immutability, to which the motion
and rhythm involved in repetition was totally foreign in spirit. In
Persia indeed (as well as India and China, which will not be considered
here) there was a change in tone. The column was used, not the massive
one of Egypt, but a lighter shaft, which showed a tendency toward
other effects than immensity and strength.

With this change of ideal, repetition in some kind of system made its
appearance, but its variations were tentative. It had not become used
to its new sense, and it was left for Greece to develop the rhythm and
movement of repetition, and to combine it with proportion and symmetry
into its perfection.

The method of analysis employed has been to go through a certain
number of architectural photographs, picking out all the examples of
repeated forms of any description, and classifying them according
to the principles which they exemplified or seemed to violate. For
this purpose a collection of about five thousand photographs from
the library of Robinson Hall, Harvard University, was analyzed. The
photographs were taken in order of styles: Greek, Roman, Romanesque,
Gothic, Italian and French Renaissance, and modern. The examples of the
different points in question were taken as they came in the cataloguing
of the library stacks, without respect to whether they appeared to bear
out the previous conclusions or not.


VARIATION OF ALTERNATING UNITS

The first principle which we shall consider is the variation allowable
in the units of an alternating series. It will be remembered that
the principle was as follows: (1) In any series of two alternating
units, the one on which the most energy is expended is regarded as the
principal unit, the less important one as an alternate. Variation of
the principal unit is allowable, often desirable and even necessary;
variation of the alternate never allowable, unless other circumstances
change the situation. If the minor unit is changed, so that in interest
it equals the major unit, the rest-phase of the rhythm is destroyed,
the effect is of two rival repetitions going along together, and
fatigue results. If variation in the alternates exceeded that of the
principal units, the balance of the rhythm would change, the alternate
become the major unit, and a new series begin.

From the very nature of the case, then, it will be impossible to look
for variations in alternates, which make it exceed the principal units
in interest. We must investigate alternating series, in order to see if
_one of the elements remains the same_, while the other may or may not
vary. If this were true, a rest-phase for the rhythm would be assured
in the series, while the principal unit might vary, provided the same
amount of attention were required in each case. (2) It will also be
remembered that _size_ and _limiting shape_ were the factors that could
not vary without doing violence to the rhythm, while content might vary
almost without restriction. (3) The position of alternating units as
regards each other cannot vary; the two units are so dependent on each
other that the position of one must remain halfway between two of the
opposite kind. In other words, if the two series of units run between
each other, they form _one_ series or rhythm. Two rhythms cannot be
kept up alongside; so if one unit, however regularly placed with
regard to another of its own kind, recurs at unequal distances from
the _other_ units, the feeling of the repetition is lost, the rhythm
broken, unless the two units can be grouped into one, and so make a
single rhythm again.

We shall, then, look for alternating series, of which the two units are
at _equal_ and _invariable_ distances from each other; the variations
of content (if such there are) occur only in the major unit; and are of
the filling, not of the including shape or size.

It may be readily seen that there are difficulties in finding
alternating series which exactly illustrate this particular point, or
in reducing them to any system. It was necessary to look through many
photographs to find one that presented the required conditions (_i.
e._, two repeated series of units, alternating with each other), and
when found, they were of so many different varieties, from windows in
an apse to reliefs on a fountain, that each has had to be described
by itself, and any rigid classification was impossible. Moreover, it
was difficult to find a scale of judgment by which to decide whether
a series was really alternating or plain repetition. From one point
of view, _every_ repetition is alternating, that is, the repeated
unit always alternates with an empty space. Although such repetitions
bear out the theory still further, and emphasize yet more strongly
the invariability of alternates, and the possibility of variations in
the principal units, I have used the term in a stricter sense, and
only given illustrations of repeated objects, when one unit actually
alternated with another definite unit.

Had the other sense of the term been used, examples might have
been multiplied without limit, of slightly varying repeated units,
and unvarying alternate blank spaces. But it was felt that such
accumulation of illustration was unnecessary, and that what was true in
a stricter sense of the term would be recognized as true for the larger
number of cases that might be cited with a wider meaning. If the minor
units had a definite enclosing outline, they were counted even though
they were blank within, but without an enclosing outline, that is, if
they were mere spaces, they were not considered, although the fact that
there is such universal use of this type of decoration shows only more
conclusively how the necessity of the invariability of alternates is
taken for granted as an axiom of design.

Another type of alternate repetition was not included in the
illustrations, _i. e._, when two sets of units alternated, _without
variation in either one_. To this class belong all the conventionalized
designs used so much in all kinds of decoration, and of which a very
full account is given in Owen Jones's Grammar of Ornament.

These, to be sure, illustrate the negative points, viz., that size and
shape are unalterable for rhythmic repetition; that distances must
be equal and invariable; and that alternate units must not vary. But
since the principal units do not vary either, it seemed needless to
give them as examples of the point in question. A mention of this class
of alternating repetitions, of which there is such a great number, is
enough to show that they fall within the theory. But one example is as
good as a thousand, and their inclusion among the illustrations for
rhythmic alternates will be taken for granted without further mention.

We are left, then, to the consideration of those alternating
repetitions alone, where both have a definite outline, and one or
both varies to a greater or lesser extent. The effort will be to show
that the unit which for some reason is of principal importance in the
rhythm, is the one chosen to vary, and if not that the repetition
suffers thereby.


125 EXAMPLES

    A. Variations in Principal Unit alone: 87.

      I. Content alone:

        a. Metopes and Tryglyphs in Friezes: 9.
        b. Arches and Columns: 2.
        c. Statues in niches alternating with supports: 37.
        d. Windows alternating with supports or decorations: 12.
        e. Paintings or mosaics: 7.
        f. Carved designs in screens or ceilings: 17.

      II. Size and Content.

        a. Doors, paintings and reliefs on façades.
           Vary in size to emphasize symmetry: 4.
        b. Statues or shields over arcades.
           Vary in size to complicate rhythm: 1.

    B. Variations in Principal Unit AND Alternate: 37.

      I. Content alone:

        a. Windows and decorations on façades.
           Alternates vary in design to emphasize symmetry: 4.
        b. Windows and turrets. Vary alternately in design to complicate
            rhythm: 2.
        c. Reliefs alternating with tablets; reliefs or statues and
            pillars: Alternates vary in design to give richer effect: 11.
        d. Alternate unit is human figure: 9.
        e. Alternate unit varies in design, but is on a different level: 3.
        f. Irregular variation in alternates (windows, shields, and
            railings): 3.

      II. Of Size of Alternate Unit.

         a. Windows and supports. Vary in size to emphasize symmetry: 1.
         b. Statues and pillars; windows and pilasters.
            Alternates vary in size to complicate rhythm: 3.
         c. Vary in size irregularly. Disorder: 1.

    C. Variation of Distance.

      Row of windows--Distance between first two is wider: 1.

The 125 illustrations of alternating repetition which were taken at
random among 5000 photographs show a decided compliance with the
principles already laid down. But there are many divergences as well,
which it is necessary to consider, to see whether they are really
contrary in principle or fall under its wider application. Eighty-two
accord exactly with the principles with which we started. The distances
between each set of units are equal and invariable; one unit varies in
content but not in size or including shape; the alternating unit is
invariable.

There is an interesting modification of this principle in the case of
the metopes and triglyphs of the Greek friezes. Here the triglyphs are
unquestionably the principal units structurally, and to many observers
the principal beat of the rhythm when taken rhythmically. But the
triglyphs never vary and the metopes do, which would seem at first to
violate the rule that principal units alone, and not alternates, should
vary. This difficulty is obviated in two ways. With the spatial type of
observer, the triglyph is indeed the principal beat of the rhythm when
the series is at such a distance that the difference in the metopes (if
there is such) cannot be detected. When, however, the series is nearer
at hand, there ceases to be any rhythm, but each carved relief is
taken for itself without regard to the others. With the rhythmic type
of observer, if the triglyph has been the principal unit before, the
principal beat changes on nearer approach to the metope and the whole
series shifts its accent. It is impossible for any observer to keep
the triglyph as the principal unit of the rhythm, when so near that
differences in the metope are easily perceived.

There are still thirty-eight cases which vary from these rules, and
many of them vary in more than one respect. These exceptions fall into
several classes, quite distinctly marked off from one another, and will
be taken up in turn.

In five cases, the _size_ of the principal unit varies as well as the
content, but in four cases the variation of size is either at each end,
or in the centre unit, to emphasize bilateral symmetry of the series as
a whole. The series in this case is taken as a larger unity of which
the separate units are parts; and hence they are not only repeated with
respect to themselves, but are symmetrical with respect to the whole.
In the other case where the size of the principal unit varies, it
varies on _every other one_, thereby complicating but not confusing the
rhythm, _i. e._, a stronger accent comes on every other principal unit.

There are also five cases in which the alternate spaces vary in size.
Three vary regularly, thereby enriching the rhythm by introducing
alternate heavy beats, and one varies at each end of the series to
emphasize bilateral symmetry of the whole, with regard to the central
unit. In the other case the alternates vary in both shape and size,
with no regularity and from the point of view of repetition alone,
disorder is all that results. This is on the Palazzo Pretoria in
Pistoia, where carved shields occur at equal distances between windows.
These shields are not component parts of the building, but were added
with some other kind of significance; hence they express nothing so far
as repetition for its own sake is concerned.

The other variations are all in the _content_ of the alternate, minor
space. Four vary _symmetrically_ in the designs on each side of the
central point, so as to accent the bilateral symmetry of the whole
taken as a unity. Two vary rhythmically in design, _i. e._, there are
two sets of designs which alternate with each other in the unaccented
spaces. When they vary regularly in design, the rhythm of the whole
is _enriched_ not confused, provided there are only _two_ sets, not
three or more. The alternate spaces are passed over on the way to
the principal unit, but by having an alternating design between them
(varying only in detail, but of the same general character) a more
complex rhythm is introduced which is good, since in both cases the
alternates and principal units are so different they could not possibly
be confused with each other, even though both varied. (In one case,
turrets and statues vary with windows of the same shape but different
decorations; in the other, arched windows and arched spaces alternate
with statues.)

Eleven more cases of variation in the minor as well as major spaces
fall under another head. These do not vary with regularity, but are
different in each case--the detail of the design varying, while the
shape, size, and distance remain unchanging. It is interesting to
notice that these examples of variations of alternates were almost all
taken from examples of Renaissance architecture, where a richness of
effect was desired, even at the expense of regular rhythm. This could,
indeed, be attained in no other way so well as this. In all these
eleven cases, the conditions are alike: both of the repeated units
are enclosed by limiting lines of unchanging outline. The principal
units are more prominent than the others on account of greater size or
interest, but the alternates, instead of retiring entirely into the
background, have slight variations in decoration. This variation is
always only in detail: the tracery on the pillars of the tomb of Louis
XII, of the Loggia dei Novoli, in the Chiesi di Frari, Venice, etc. So
unimportant in fact is the variation that it is not observed until one
attends closely to it, and yet the rhythm is just enough disturbed by
its presence to give a feeling of _extra sensation_ or _luxuriance_
which cannot be attained through variation of the major units alone.
It is in the alternate spaces that the _feeling_ of repetition lies.
Any material change in them destroys the series, but a slight variation
in the lines of decoration, a little rearrangement of the conventional
curves in each alternate, gives, even though unattended to, in fact
partly _because_ unattended to, a vague feeling of variety, of some
superfluous sensation being brought into consciousness, although the
regular shape, size, and distance of the objects remains unchanged. It
is this feeling of superfluity and slight disturbance which constitutes
the peculiar richness of certain styles. These examples, then, far
from falling outside of the laws of repetition, owe their opulence of
sensations to the very principles of regular rhythm which they violate.

Another set of exceptions will involve more searching analysis. Nine
of the examples described have the human form for the alternate unit,
and in every case where this happens, the alternate varies. In the
majority of cases where statues of the human form alternate with any
other object, the statue is taken as the principal unit on account of
its superior interest, but this is not always the case. In the Padua
Basilica, and in the Church of St. Guistiana, cherubs alternate with
conventional decorations, but the latter are so much larger and more
elaborate that they would naturally be taken as the principal units. In
the other seven cases, statues alternate with bas-reliefs which also
have human figures in them; hence, since the bas-reliefs equal the
statues in interest and exceed them in size and importance, they are
taken as principal units.

It might at first be expected from the previous discussion that, in
order not to shatter the repetition, the alternate statues must be
alike, must be conventionalized into identity; but this is not the
case. Another principle now comes into play. We demand variation in
the human form whatever its place in art, even in the unimportant
position of alternate in a repetition, and although they are kept as
much alike in pose, size, level of head and feet, general character
(_i. e._, cherubs do not alternate with old men, nor draped figures
with undraped), yet there is some variation of pose or direction of
glance, to keep them from being duplicates. We should expect this
variety of repetition to be in danger of becoming fatiguing because of
its lack of an unchanging rest-phase, but this difficulty was evidently
felt in building them, for in every case _some unchanging element_
has been supplied to the series to bind it together and to keep the
constant changes of attention from upsetting the series. The cherubs
of the Padua Basilica are in high relief against a uniform rectangular
background which does not vary, and which furnishes an alternate
just in character with the principal unit, the bas-relief. In the
Cantoria of Donatello, although the dancing children move across the
whole space, uniform double columns occur at intervals, and supply an
unchanging alternate, while the children vary in position behind them.
Around the pulpit of Lincoln Cathedral, although both units, reliefs,
and statues vary, the pilasters behind the statues are invariable and
supply a constant, unchanging factor in the series. In the alternating
reliefs and statues of the Milan Cathedral or in the paintings of
different sizes in All Souls Church, Oxford, an unchanging element
is supplied in the frame, which is of like design in every case,
so that in passing from one to the other an unvarying alternate is
always present. In the Sienna font, and in the statue to Leonardo da
Vinci, which are types of a vast quantity of repeated forms, there is
uniformity in the minor pedestals and in the frames of the alternating
bas-reliefs, which supplies the unchanging factor.

Moreover, another factor is noticeable in this kind of repeated
series,--it is never long. The fatigue which would certainly result
from a too long continuance of varied alternates, even with unvarying
factors in the way of supports, pillars, and frames, is obviated in
various ways. The series is either short and the whole has a definite
bilateral symmetry, as in the Padua Basilica, and in the Oxford church;
or, as in a great number of cases, the series goes around a fixed
central point so that only three units are seen at a time. It is thus
especially that this method is used in fonts, pulpits, and monuments,
where from the circular arrangement enough can never be brought into
the field at once to fatigue the attention.

This consideration of alternates which vary widely, as do human
figures, even when they are alike in size, general shape, and
character, and, moreover, the discovery that there is almost without
exception an invariable element _between_ the other alternating units,
_i. e._, a third alternate; or _behind_ them, as in the case of the
pilaster behind the statue, may well bring up two questions:

(1) When the unchanging factor comes _between_ the other two units,
is not _it_ in reality the alternate, and the two other units either
variations of shape and size of one principal unit or two sets of
principal units? In other words, do we not actually apperceive the two
principal objects as the units of importance, and take the unvarying
factor which comes between, no matter how slight it may be, as the
alternate? Do we not demand the unvarying as our alternate, no matter
how many variations may be in the other figures?

(2) When the unchanging factor comes _behind_ the alternating statue,
in the same plane with the bas-relief, do we not inevitably take _it_
as the alternate in the series, and regard the statues more as episodes
or attendants on the series but with real values of their own? Is not
the fact that the unvarying factor and principal units are in the same
plane an indication that they constitute a real series, while the
statues or paintings which are in a plane by themselves make a series,
harmonizing with the other, it is true, and in part coinciding with
it, but felt in a different way? Therefore the actual repeated series
conforms to the given conditions and is made to do so in every case by
its unchanging alternate in the same plane; while human figures with
values of their own never can be considered quite as alternates, but
are really felt to be a series by themselves.

This introduces another question. In two more cases of varying
alternates, there was variation in decoration above the level of the
rest of the series. In the Borghese Casino, there is variation in the
busts placed _over_ the alternate windows. In the Venice rood-screens,
there is variation in the carving of the alternating supports, which
rise _above_ the rest of the series. Is that part of the series above
the level of the principal units really included in its perception? It
would seem rather that when the series as a whole is being taken, those
variations above the level of the main units (if they are not very
marked, and they were not in either of these two cases) are ignored
or only felt in a vague way as added richness. When, however, the
attention is turned toward them especially, they form a series of their
own, in which they become the principal units, and alternate with empty
spaces. There is no limit to the changes possible in apperception,
according to the _level_ and _plane_ of the alternating units.

There are three cases left; two where alternates vary in content
with no system, and one with variation in distance. The first two
are differently carved sections of railing on the side of Freiburg
Cathedral, and a differently decorated frieze of squares and circles in
the S. Lorenzo Cloisters, Rome. The effect is only of disconnected and
fragmentary series in both cases, and especially in the latter case it
is impossible to feel it as a repetition at all unless the variations
are ignored, and the attention fixed on the unvarying factor of size.

The variation of distance is in the Beauvais Palais de Justice, where
the first window is at an unequal distance from the others in the
series. The effect is only of disorder and accident.

We have, then, surveyed all our examples of alternate repetition, and
found that in the exceptions to the general principles laid down some
other effect than repetition as such was sought. Either (1) symmetry
for the series as a unity was required, which demanded variation of the
end or central units. In so far, then, as it fulfilled the requirements
of symmetry, those of repetition were disregarded.

(2) Richness of effect was accomplished by those slight variations
in decoration of alternates as well as the principal units. These by
their vague suggestions of different combinations of similar elements,
and minor differences felt but not attended to, gave a superfluity of
experience which made up its peculiar richness.

(3) When the human form (or any other form of especial meaning in
itself) makes the alternate unit, some variation is demanded as in
keeping with its own significance, since in proportion as a thing
has meaning in itself, it must not be exactly duplicated. But an
invariable alternate is always supplied in the way of a frame, or
background, which is felt as the real rest-phase of the rhythm, while
the varying alternate forms have a place in a series of their own.
Also, since such a complex attitude would be fatiguing, such series are
always short, or circular, so that few units are in the field at once.

(4) Regular variations in size or content, in either major or minor
which recur at fixed intervals, give a heightened rhythmical effect by
making certain beats heavier than the rest. As has been stated before,
the major unit holds within it the real significance of the _content_
of the experience; the minor unit holds the secret of the _rhythmic_
effect.

(5) Only 4 examples of the 125 were found to repeat themselves
alternately with irregular variation of alternates and violation of the
other principles laid down at the start. These can only be regarded as
accidents, as faulty examples of art, whose virtue lies in some other
part of the work as a whole, and not by any beauty they possess in
themselves as repeated series.


SUMMARY

                    125 _Illustrations._

      I. Variations in Principal Unit alone, 87

        82 Content
         5 Size          4 Symmetry
                         1 Rhythm

      II. Variations in Alternate Unit (and Principal Unit)
        32 Content      11 Richness
                         9 Human figure
                         2 Rhythm
                         4 Symmetry
                         3 Different level
                         3 Disorder
         5 Size          1 Symmetry
                         3 Rhythm
                         1 Disorder

      III. Variation in Distance
                         1 Disorder

Several questions have been raised in this discussion of variations,
but one which seems directly leading from it will be considered next.

When is variation _necessary_ in a repeated series? We have considered
the numerous cases where variation is _possible_, and the different
ways in which a series may vary according to the idea to be expressed.
Moreover, what appeared to be exceptions to the rule were shown to be
guided by a desire for some other effect than repetitions as such.

But when do we demand variation in a series? Is there any case where
variation of the unit is not only allowable, but positively necessary
to its æsthetic value?

There were no experiments on this question, for it will be seen from
what follows that they would have been impracticable. But observation
of several thousand photographs has made the following clear: When the
series consists of objects having an æsthetic significance of their
own, not depending on something else for their value, then variation
is demanded. In other words, when a thing is an end in itself, we do
not tolerate an _exact_ duplicate. It may have a place in a series
of others similar to it, but its own meaning loses force if another
is beside it precisely alike. When, however, an object has no great
significance by itself, or when however great its value, it be regarded
as means to something greater, hence not an end in itself, it may be
repeated without variation.

This principle may be stated from another point of view: Any work
of art, of the _highest_ significance in itself alone, must not be
repeated at all. There must not be even the suggestion of repetition.
The highest values are individual, and to have a copy or a series
defeats its whole reason for being. Thus, a second Sistine Madonna, or
a series of Venuses, would shock our whole æsthetic feeling. Moreover,
we do not want a _suggestion_ of repetition; even a series of different
Madonnas in similar frames would take away from the significance of
each, in so far as they were regarded as a series, and not as a mere
collection of detached units.

But grading down from these works of the highest value in art, there
comes a point where an object, although possessing considerable value
in itself, is not so intensely individual but that it can gain somewhat
by a place in a series of others like it _in some respects_, but
differing enough so that each still keeps its own meaning distinct from
the rest.

The balance between these two artistic aims, _i. e._, the significance
of the unit, and the rhythm of the series, must be adjusted with
great nicety, and certain principles obtain wherever such series are
found. It would be useless to cite the numberless cases where such
series occur. Many have already been given in the examples of statues
of saints, paintings on altar-pieces, and reliefs alternating with
statues. One such series is a type of all. The human form represents
that which has the most significance in itself, so when it is used
in a rhythmic series, its individuality must be toned down and
conventionalized; it must have no marked feature in one unit that does
not appear in another; the head and feet must be on the same level, or
vary with regularity; the general character and spirit of all must be
similar, but never identical.

The reducing to a common type is the demand of the rhythmic series; the
difference in attitude and arrangement of detail is the demand of the
unit.

Thus, the subjects chosen for repetition of this kind are in the
majority of cases apostles and saints, whose spirit and general
conception are the same; typical representations of abstract qualities,
such as Virtue, Courage, etc.; or conventionalized cherubs, and even
animals. As has been stated before, a long series of this kind is
impossible without fatigue. In proportion as the object is repeated
the individual units lose their own meaning, and they must have their
individuality definitely toned down and conventionalized to avoid the
clash between the two artistic values. Yet their essential peculiarity
must always be maintained, for we refuse to admit or allow the total
identity of any expression of living values, especially as expressed in
the human form.

It may be urged that statues are often arranged at regular intervals
around a building, where the effect of repetition is distinct, and
yet each statue is distinctly valuable for itself. But a distinction
must be insisted upon. The statues form a repeated series as regards
uniformity in position, height, pedestal, and color, so that the direct
sensuous effect may be called rhythmic. But as the attention fastens on
each for itself and takes it for its own meaning, it ceases to be part
of a series at all, but becomes a unit in a world of its own.

But what of the cases where the human form is repeated in a series,
and does not vary? Examples of this are rare, but they do occur, and
are interesting, since they throw light on what has been already said.
In the whole collection of photographs only two were found where a
series of identical statues of the human form occurred,--_The Porch
of the Maidens_ in the Erectheum of Athens, and the _Baths of the
Forum_ in Pompeii. In the former case the left knee of the caryatids
on the right of the centre, and the right knee of those on the left
of it, are raised a little; but aside from this slight variation the
six statues are exactly alike. In the latter case a row of titans all
around the interior bear the ceiling on their uplifted forearms and are
all alike. These two examples are very perfect of their kind, and, far
from offending us, are very satisfactory. The reason is obvious. In
both cases the statues are not the æsthetic end in themselves, but are
there for a purpose, namely, that of a support. They are not ends but
means to something else, and as soon as we feel _that_ in regard to any
work which would otherwise be of individual significance, it ceases to
be individual, or to demand a peculiar expression different from all
others, but may be duplicated without offence. Therefore, since the
support of the superstructure obviously is dependent on the maidens
in the one case and on the giants in the other, and since instead of
existing simply for their own value they are there to hold up the roof,
their artistic significance changes at once from _ends_ to _means_,
and variation is not required. Moreover, it will be found in the
majority of cases that we demand this invariability in actual supports.
Although we find but these two cases where caryatids are actually
identical, we find also that in most cases the caryatids do not really
uphold the weight, but a pillar or pier behind them supplies the real
architectural support, and, that although they have a place in front of
the pillar and give an apparent assistance in bearing the weight of the
roof, yet the eye is not deceived. We see that the work is really done
by the pillar behind them, so they that resume their place as artistic
ends demanding variation, and not as means to something else. The
following examples were found:

_Milan. Arca di S. Pietro Martire._ Pillars uphold the arch while four
statues of women stand just in front. The pillars bear the weight
although the statues add strength to the whole. The statues are varied.

_Dijon. House of Caryatids._ Piers behind the caryatids give real
supports to the roof, while the figures added for decoration are all
varied.

_Dresden. Zwinger._ Conventionalized figures ending at the waist are
put on the outside of unvarying piers which bear the actual weight of
the superstructure. The figures are all varied, but they cannot be
conceived as really bearing the strain, since they have no foundation,
but are merely added to the pier as a decoration.

_Rouen. Tomb of Duc de Brezé._ Four caryatids, all different, under
four jutting projections of the arch. These projections are built
securely into the rest of the structure and do not depend in the
slightest on the figures for support. The figures are not integral
parts of the whole architecturally, for the arch would stand exactly as
well if they walked away, which indeed they are apparently in the act
of doing.

_Toulouse. Hotel de la Borde._ Two caryatids under jutting projections
of a window. The projections are securely built into the lintel and
no weight rests on the caryatids nor even appears to. They are there
solely as decorations and are different.

_Paris. Hotel de Ville._ Two caryatids under jutting projection of
a window, again. Here is a very slight variation of the two female
figures. The position of each is reversed to accent the symmetry of
the whole. Very little weight is actually borne by them, but more than
in the former cases, and we find proportionately less variation in the
figures. They approach identity, but there is variation in detail.

These were the main instances found of the point in question, and are a
type of the other minor ones found in support of pulpits, choir-stalls,
and windows. It will be seen that in no case but the two classic ones
given at the beginning are the human figures architecturally necessary
to the structures, and in these cases they do not vary. In the other
cases they are more or less playful, and the effect of the whole would
be very unsteady did the superstructure actually depend upon them for
support; but since piers rise invariably behind them and bear the
weight, they fall into the sphere of decoration and from that point of
view they must and do vary.

We have, then, considered variation of units in a repeated series,
where they may vary and where they must, and we find the real value
of repetition to appear in inverse proportion to the individual
significance of the separate units; the more interesting or expressive
the unit is in itself with individual significance, the less do
we want it repeated; and so repetition of the human form must be
conventionalized to the type (or to the same unvarying features), with
enough individual differences still remaining to meet the demands both
of the series and the individual. What apparent exceptions we have
found to this rule have been shown to be meeting, in reality, another
artistic demand.


ENDS OF SERIES AND ARRANGEMENT OF REPETITIONS WITHIN THE UNIT

The next question to consider is the _ends_ necessary for a repeated
series. Do they end with a heavier or with a lighter unit than the rest
of the series, or with a unit of the same size? It will be remembered
in the experiments touching this point that the subjects, without
exception, preferred the series ending with heavier units. We should
then expect, in examples of repeated groups of posts, pillars, etc.,
alternating with wider or more prominent ones of the same kinds, that
the series would end with the heavier or more prominent one. Examples
of railings or balustrades alternating with heavier supports are so
common, and the supports come so invariably on the end, that repeated
examples seem almost unnecessary. But another question arose in
connection with this: Does not the apperception of a group of lines
equidistant from each other consist in going back and forth over them
from edge to edge, with no rest on one point more than on another;
while in a group of lines arranged at equal distances each side of the
centre but not from each other, to emphasize bilateral symmetry, does
not the attention rest on the centre, and move from the centre of one
group to the next?

Moreover, we found that a wider space or embankment of some sort was
necessary, to finish off a series of groups in which the separate lines
were equidistant from each other, than to finish the groups whose lines
were symmetrically arranged. This suggests that the activity which
goes back and forth in the former case, being less coördinated and
not bound to a middle point, needs more at the end to stop it than is
needed in the latter case, when the attention is more upon the centre
of each figure. It would seem, then, that the former arrangement would
be appropriate for railings and balustrades, where the effect is of
continuity either running wholly around the structure and into itself
again or where a continuity of parts is desired and a connected series.
The other arrangement divides the series into discrete parts. If the
attention is stopped at every central point, the effect is less of
continuity and more of separate unities bound together externally by
their equal distances. We should, then, expect such series of units
much less in continuous balustrades, but if they occurred at all, that
they would be in connection with separate unities that did not want
continuity or place in a series emphasized at the expense of their
individuality. All this we might expect from the experiments alone,
although whether such a refinement would have got into architecture
seems questionable. Moreover, the question whether a symmetrical
group of units needs a less heavy end to finish it than a group of
the equidistant type is even more difficult to illustrate. Although
the two types may be given under some conditions in experiments, in
actual architecture they never appear so, for the two types never
appear in the same buildings allowing them to be compared. Besides, few
photographs are taken exactly in front, and no two at just the same
angle. Any accurate measurement of such end piers and any comparison
of them is out of the question in the present methods of research.

One other question may be considered here. Does a series ever occur in
which three units are repeated regularly, instead of one or two? In
experiments we discovered that the subject found it impossible to feel
repetitions of three in a series, and the only way that such a series
was tolerable was when the three could be grouped somehow into one or
two units. Therefore we should not expect to find such repetitions
frequently, if at all.

To sum up: Do series always end with a heavier unit? Are units equally
distant from each other more adapted to continuous or run-on railings,
while units with symmetrical arrangements within themselves are found
more often where separateness of objects enclosed is more aimed at than
their connection? Is a less heavy end found after symmetrical series
than after the other kind? Are repetitions of three units used at all,
and if so in what way?

Obviously the only illustrations of these questions will be found
in the arrangement of posts and pillars in balustrades of whatever
description. In these cases alone do we find repeated series, with
repetitions within the unit, as well as of the unit as a whole. The
following examples have been taken by looking over about one thousand
photographs and by recording every instance that occurred.

          100 _Examples_

    A. 73 Continuous Railings: Balustrades across façades; around roofs; up
          flights of stairs; around towers and baptisteries.

        I. 57 Rhythmic Units:
           a. 31 Even numbers of units in group.        Support arches 5.
           b. 26 Odd number of units in group.
              11 Support even number of arches.
               6 More than eight units in group.
               1 Two sections of railing. Odd number in ends, decoration
                  in centre section to emphasize symmetry.
               4 No grouping. Too many to count.
               4 Other reasons not assignable.

       II. 16 Symmetrical units: Slabs with carved reliefs or plain; Carved
               scroll or diamond designs alternating with posts; Heraldic
               designs on shields;
               Conventional decorations in stone or wrought iron.

    B. 27 Detached enclosures: Separate windows and doors.
        I. 4 Symmetrical units:
       II. 23 Rhythmic unit-groups.
           a. 10 Odd number of units in group.
           b. 11 Even number of units in group.
               4 Support odd number of arches.
               4 Although before separate windows make a continuous row
                 across the side of the building.
               2 Three sections of railing. Odd number in ends, even in
                 centre section to emphasize symmetry.
               1 No reason assignable.
          c. Indefinite number in group. Iron bars in railings, and slender
              pillars on façade.

    C. 8 Do not end on the heaviest unit.

    D. No cases of regular repetition of three units.

Having 100 illustrations of repetitions of groups, with units repeated
equidistantly between them, and of elements distinctly symmetrical,
several new factors came to light. In all the one thousand photographs
looked over, not a single instance was found of unit-groups with the
units within, arranged at other than equal distances. There were many
variations in the number of units in the groups; but the number being
given, the units were arranged at equal distances from each other
wherever the effect desired was of detached sections or of continued
series. There are obvious structural reasons for this. Any repetition
of groups for a balustrade or protective railing, which is the almost
exclusive use of this variety of repetition, would be weakened by wider
apertures on either side of the centre. A reasonably enclosed space
is necessary to make the railing of value, therefore the specifically
symmetrical unit as opposed to the rhythmic unit was found always
in carvings, scrolls, bas-reliefs, etc., alternating with vertical
supports. We should expect, then, in general, that in railings where
an aspect of continuity of progress along some border or a tendency
to go around an enclosure was sought, the units would be rhythmic
in character, impelling one to motion and to carrying the eye and
general organism out of repose into movement. We should expect, on
the contrary, that symmetrical units would be found where repose or
partial distinctness of the separate elements enclosed was desired, and
where the attention was not to be carried away in so marked a degree.
Seventy-three of the one hundred illustrations were of balustrades
where the rhythmic factor was presumably aimed at.

The Rathaus at Braunschweig had a symmetrical design alternately
occurring, but with four in a section, so that the section as a whole
was not symmetrical and the attention was driven on, and in the other
cases some other effect than rhythm was obviously aimed at. The genius
of the structures was heavy and massive and the balustrade made in
keeping with them, since an effect of motion or rhythm would have
clashed with the spirit of the whole.

These examples have all been of the balustrades around enclosures,
balconies, etc. Since the rhythmic unit has been found more fitting for
them, we should expect, conversely, that in front of separate unities,
such as windows, doors, etc., the symmetrical unit would be more in
evidence. At first sight, the facts do not seem to bear us out in
this. Of twenty-seven examples of separate windows, doors, and gates
enclosed by railings, only four had distinctly symmetrical designs.
(Casa Palladio, Bergamo Chapel, Petit Trianon.) These are wrought-iron
designs in the centre with repeated rods on each side, or a row of six
pillars with the central two larger and more decorated. Twenty-three,
however, remain to be accounted for, and the solution of the difficulty
is observed at once in the distinction between _odd and even_ numbers.
As was previously suggested there are obvious difficulties in having
posts in a balustrade at any but equal distances, since the gaps
left by unequal distances from the centre would destroy their reason
for being. This difficulty can easily be overcome in wrought iron by
extra central decoration, although it is not always done by any means;
but in stone balustrades, unless there is carved open-work, or solid
reliefs, there is no other choice than repeated posts, either divided
into sections or continuous, and no variation is possible except to
have an even or odd number of them. We should then expect that there
would be an odd number in separate detached enclosures, bringing a post
in the centre to emphasize the balance, while in a continuous series
each group would have an even number, thus giving no centre to fixate
upon, but driving the attention on without repose at any one point more
than another. It might seem doubtful that any such refinement should
have been actually expressed in architecture, but examination of these
examples shows this treatment to be very general. Of the twenty-three
examples of separate enclosed details, eleven have an odd number of
posts. Of the ten that remain, _four_ are examples of windows along the
side of a building, with separate detachments of balustrade in front
of each. By having an even number of group-units the continuity of
the row is maintained in spite of a separation of the sections. _Two_
of the ten are sections of balustrade over the central doorway of a
building. These balustrades are divided into three sections, of which
the centre is widest and the ends only half as wide. Thus, although
there are six posts in the central section, the balustrade as a whole
is distinctly divided into a bilateral symmetrical arrangement. _Three_
of the others have an even number of pillars, but they support an odd
number of arches; and the arch, not the pillar, is taken as the unit
of the repeated series. (Arches will be discussed later.) The _one_
example unaccounted for represents a number of possible cases, where
for some reason, following out a general scheme of building, or what
not, the odd number is not insisted upon for separate clusters. But the
fact that only one out of twenty-three is thus unexplained shows an
unmistakeable tendency in the other direction.

A distinction between odd and even numbers cannot be felt above eight
repetitions without actual counting, and often not even then.

The two final exceptions are of a gate and a decoration over a door
(Fontainebleau, Piacenza) where there are nine or more units in the
group. It is impossible to feel the system of this arrangement, and
the result is proportionately confusing. A reservation must be made
here concerning iron railings. There is no discrimination between odd
and even in the number of iron rods in a section of railing and no
tendency to symmetrical designs rather than rhythmic before detached
enclosures. This is because from the nature of the case, there is no
distinction possible between odd and even in the number of slender iron
rods necessary to enclose a space with any security. There must of
necessity be so many of them that the difference cannot be perceived,
and so slight is the importance of each rod that the effect is more
of a variegated surface than of actual beats of a rhythm. As soon as
iron is wrought into large enough shapes, each repeated detail is of
the same importance as in stone, but the slender rods commonly used
in iron railings, although their repetition is rhythmic like all the
others, give too slight a motor impulse to carry the attention past the
heavy limits of whatever they enclose. They are found in front of many
windows, but on account of the lightness of their rhythm compared with
the solidity of limiting piers, no confusion results.

Having thus concluded that the odd numbers of units in groups is
more adapted for separate enclosures, is the opposite true? In the
continuous balustrade, previously discussed, are the units of groups
made up of an even number of elements? Of the fifty-seven examples
cited of continuous railings, thirty-one have an even number of posts
in their groups. These conform to the rule: but what will explain the
twenty-six remaining? It will be noticed that _six_ of these have too
many in a group for the eye to perceive any difference between odd
and even, since they range from nine to thirteen. When so many units
are in a group, the effect is always of the run-on type, whether the
actual number turns out to be odd or even on subsequent count. _One_
has a balustrade with only two sections on a side, each side of the
centre door. Seven are in each section, and since the appearance of
a symmetrical whole is the desired effect, an odd number is more in
keeping than an even; in fact, this example, Monte Berico, might
better come under the other head of separate enclosures, although it
partakes of the character of both. Another balustrade with three in a
section (Blois Château) is so heavy and massive in all its parts that
fixity and solidity is more in keeping with it than rhythm. _Eleven_ of
them, that is, the larger proportion of all those with an odd number
of pillars in a section, support arches, and the arch is taken as the
unit instead of the separate pillar; and we find an even number of
arch-units in each section, which is what we should have expected. It
is a noticeable fact, which was previously suggested in connection
with separate enclosures, that when a row of pillars supports a plain
lintel, the _pillar_ is taken as the unit of repetition. (When the
row is on the front of a building, temple, etc., the _opening_ may be
the unit, if the _purpose_ of the central door or the fact of _going
through_ is in the mind: but when the series stands for itself, the
_pillar_ is the unit.) When pillars support arches, the _arch_ is the
unit, unless it is very narrow as in the Moorish style, when the pillar
is often so high and the arch so narrow in comparison that its value is
weakened.

Of the thirty-one balustrades with an even number of parts in a
section, four sets of pillars bear arches, and make an odd number of
them. This would seem to make an exception to the rule were they not
so narrow in two cases that the pillar was still the unit, and in the
other two the motif of the arch was built around the intervening piers,
so that they did not seem divided into sections at all, but continuous.

We have thus surveyed the whole field of repetitions of rhythmic and
symmetrical units, and their difference in treatment according to
the end they serve, and the results bear out our expectations. The
symmetrical unit, as exemplified chiefly by an odd number of units in
groups, is more used for detached enclosures; and the rhythmic type,
with even numbers, is used more especially for continuous ones. In
the former case the motor tendency is toward the central balance,
while in the latter it is driven on out of itself through the series.
When pillars support arches, the arch is the unit; when they support
lintels, the pillars themselves remain the unit. Any number of units
over eight loses its value of odd or even, since the difference can no
longer be perceived and becomes rhythmic whether odd or even.

It must not be supposed that these rules are inevitably carried out or
that the effect is necessarily poor if they are not. It shows a general
æsthetic demand, however, which in individual cases may be modified by
other demands, or altered in parts to make a more unified whole. When,
however, the series is taken for itself, and judged entirely on its own
merits, these conclusions will be found generally valid.

We have still to consider whether series always end with a heavy unit.
All the series examined _do_ end in this way; in fact we feel the
necessity of this so clearly that one illustration would be as good as
a hundred. But there is a difference in the use of the end unit, which
is noticeable in any two series of symmetrical and rhythmic units. Of
the sixteen examples of continuous series whose units were distinctly
symmetrical instead of rhythmic, eight of them, although ending on
supports, do not end on the principal unit of the series. This can be
best shown by one or two examples. The Orvieto Cathedral has on the
façade a balustrade of rectangular reliefs alternating with supports.
The reliefs are undoubtedly the more interesting and important element
of the series, yet the series ends with the less important element, the
support or post, and we feel that it must do so. The Palazzo Contarini
has a balustrade on its façade in which carved wheel-like designs
alternate with supports which come at the ends. Why, in these cases,
do we feel it as inevitable that the heavier and more important unit
should _not_ come at the end, as with rhythmic units we feel that they
should? The answer to this is partly structural and partly æsthetic.
We must feel, first of all, that the series is properly supported,
that it will not fall away at the ends or down in the middle, and for
this reason support of some kind must come at the end to hold it up
and give a feeling of solidity and stability. But why are not these
supports made the more interesting and important unit so that they
might still bear up the superstructure and end the series as well? Here
the æsthetic demand appears. As soon as the object is regarded as an
æsthetic unity and care put upon it to make it beautiful for its own
sake, it must not be thought of as the _end_ of any series. It must
be cut off from the rest of the world by supports or framed in some
way, and while it still may have a place in a series, provided it is
sufficiently conventionalized and not too important in itself, it must
not be thought of as either ending or beginning, as depending on a
series to give it importance, or lending support to anything else. It
simply exists, cut off from the world, even though in the balustrade
not an integral part of it, and one ought to be able to remove it
without affecting the stability of the structure.

The question whether series of symmetrical units have less heavy ends
to finish them than series of rhythmic units cannot be settled by
these methods of analysis. While it seems certain that the rhythmic
series drives the attention on by its greater motor activity, and hence
would need more of an end to stop it, so many other factors enter in
of more importance, such exact measurements would be necessary (quite
impossible with the photographs of the scale here used), the refinement
would be so great, since the stone of which most of the examples are
made, by its own weight supplies a check to rhythmic activity, all
these considerations make it impossible to illustrate this conclusion
and it must remain an experimental result alone.

There remains one question: Is regular repetition of three units ever
found? They may be in combination of some kind so that they fall into a
rhythm of twos, but are they ever found repeated as three separate and
distinct units? The answer to this is without exception. Of the five
thousand photographs analyzed, not one instance of this kind of series
was found. In many cloisters the pillars are of different design, and
often one design is repeated through an otherwise varying series, but
their repetition is either without scheme of any kind, or in some
combination that falls into a rhythm of twos. No three-rhythm has been
used in art, any more than it has been found possible in experiments.


ARCHES

It has been noticed in the preceding discussion that when a series
of pillars supports arches, the arch, not the pillar, is taken as
the unit. If this is so, it would seem that the arch by binding two
pillars together with a curve awakens a more vigorous response than the
vertical line of the pillars, and this greater expenditure of activity
makes it to be taken as the element of repetition. It suggested that
the arch (like the rhythmic unit) tends to drive attention on out of
one unit to the next in the series. The outward thrust of the arch
arouses an outward-tending activity, and for this reason a row of
_arches_ would need, to give a finished, stable effect, a wider and
heavier embankment at the end than a series of lintels. The experiments
on this point were inconclusive owing to the difficulty of obtaining a
series of arches and of lintels which should be comparable in size.
For this reason the validity of this suggestion must depend upon the
actual treatment of arches in architecture. It would seem that the
arch would, like the rhythmic unit, be more appropriate for continuous
series than for detached short rows; or if the series were short, the
ends should be treated in some way, by reduction in size, change in
width of pillar, pier, or decoration, so that the outward-activity
might be counteracted by some inward thrust or some accentuation of
the centre. Thus the unity or balance of the series as a whole would
prevent the arches from seeming to "run away" which they might appear
to do without such treatment. We shall, then, look through photographs
of buildings where arches are used, to find if their treatment carries
out the supposition.

It may be seen at once that such a treatment of arches differs from
the arrangement necessary to make plain lintels effective. The pillars
on the front of Greek temples were indeed slightly farther apart at
the middle entrance, and the centre was moreover further accented by
the point of the pediment. But on the sides the rows of from thirteen
to sixteen columns had equal interspace and no noticeably heavier
columns or embankment of any kind at the ends, for none was necessary.
The series appeared ended whenever it stopped, and did not carry the
attention over, nor demand some finish to "hold it down," as does the
arch. The pillars, to be sure, completely surrounded the temple, and so
were, in name, continuous. But on a building with square corners, the
other sides do not carry the series on to the eye (with variations in
foreshortening of the ends) as in a circular structure, and the effect
of continuity is not immediate.

Many examples might be given of buildings with pillars and lintels
on the façade, which have no visible modifications of central or end
columns to give balance or symmetry to the whole, and yet which are
perfectly satisfactory as repeated series and do not demand either
such treatment or further continuation, but are complete and finished:
London, Trafalgar Square; Rome, Pantheon; Vienna, St. Karl, Barrome
Kirche; Berlin, Schillerplatz, etc. These have the centre accented by
the superstructure, but there is no discernible modification of the
series itself.

Examples might be multiplied, but there are sufficient to illustrate
the essential stability of repeated vertical units and to contract them
with the outward-tending, run-on effect of arches which need various
kinds of treatments to finish a series.

        165 _Arch Series._

    A.  45 Go completely around exteriors: Colosseum, arenas, baptisteries,
          towers, cloisters, courts, basilicas, tombs.

        59 _Series that end:_

    B.   I. 30 Central arch largest: triumphal arches, doors and windows on
               façades of churches.
        II.  1 Central arch smallest: doors on Peterborough Cathedral.
       III.  4 End arches larger: windows or decorative arches on the walls
               of buildings.
        IV.  6 End arches smaller: windows, decorative arches, or arches
               halfway around a court.
         V.  6 Arches go obliquely into higher central point and back:
               decorative arches running into the pointed roof on
               Romanesque façades.
        VI.  6 Central arch accented by decoration: windows and gates.
       VII.  6 End arches in different planes: doors on façades of
               buildings or in gates.

    C.  20 Arches go around interiors: up naves and across the apse of
           churches, halls, and loggias.

    D.  27 Around the outside of porches, apses, etc.; diminish in size
           at ends; are carried on in the transepts; motif is carried on,
           although whole arch is not; end arches are closed, or centres
           decorated.

                                   7 _Good_

    E.  14 Other arrangements: Roman aqueducts (endless); interlacing
           arches; filled with statues; finished by gables or turrets;
           bridges (land on each side a sufficient embankment); arches
           included in large ones.

                                   7 _Poor_

           Series not sufficiently finished at the ends; only two arches
           in series; three arches, with first arch different from the
           others.

Of one hundred and sixty-five examples of such series examined, only
seven do not conform to the principles we have considered, and these
are proportionately unsatisfactory. Forty-five illustrate buildings
where the arches go completely around the outside of a structure,
so that the series instead of requiring an end simply runs into
itself again. It will be noticed further, that unlike series of
columns around rectangular Greek temples, these are around circular
structures where the series does not change its direction suddenly
but by degrees. With the exception of courts and cloisters where the
observer stands within and sees the whole series, these are all around
domes, baptisteries, etc., where the end arches in the field at any
one point of view are seen in perspective gradually fading off and
yet leading attention on around the building. There may indeed be
arches which go across square-cornered buildings or even around them,
but in these cases some other device is necessary to make each side
a finished series in itself. The mere fact of its continuance around
a corner where it cannot be seen from the same point of view is not
enough. (These various arrangements of arches on a flat façade will be
taken up later.) Rows of arches are often used around towers square
as well as round, but towers from their very shape and size allow the
observer to see different sides from nearly the same point of view,
so the series is not broken up into sections on different sides of
the tower as it is in a larger building. Twenty more examples are of
arches in interiors and are all of arches down a nave, with either a
regular arch or an arch motif carried across the apse. It might be
supposed that an arrangement of arches in an interior would be more
difficult than on an exterior surface, since the genius of an arch
is its outward thrust and its tendency to run on. Without careful
treatment it would spoil the interior by trying to overstep its bounds;
by making certain walls look wider than others; the arched sections
utterly discrete in general character from the plain or otherwise
decorated section. In point of fact, the use of the arch-series in
interiors is quite conventionalized, and all the illustrations are of
loggias, or of churches where the arch goes down the nave and in a
more or less modified form across the apse. In the Sistine Chapel the
arched windows go down the side walls and across the end in a vaulted
double-arch. In some cases a series of Roman arches down the nave has
a more or less pointed arch across the apse, but in every case the
continuity has been kept in some way so that the series is unbroken.
Moreover the columns in the cathedral naves are often so high and the
arches so proportionally narrow that the pillar instead of the arch is
taken as the unit. This is somewhat true in St. Mark, Venice, also in
St. Sophia, Constantinople, where the large arches are divided into
sections of seven smaller ones, each one of which is so narrow that the
pillar is felt as the repeated unit instead of the arch; or if the arch
be taken, the narrow span prevents it from too great outward thrust.

Thirty of the arch-series are on façades of buildings or in structures
by themselves, as gates and triumphal arches, where the central arch
is larger than the other, thereby emphasizing the middle point and
drawing attention to it away from the ends. This centralizing a series
or balancing it as a whole may be accomplished in various ways. Two
examples make the central arch larger instead of smaller. Six make the
end arches smaller while four make them larger. It will be readily
seen that just which one of these variations is chosen for the series
depends on the function of the series. The central arch is wider, with
only one exception, when the series is of arched doors and the central
door is the main entrance; while the end arches are more apt to be
varied when the series is purely decorative and serves no function.
The central balance may be further gained by differences of level.
In the decorations of many façades, especially the early Romanesque,
rows of arches go obliquely into the point of the roof and by this
strong pointing toward the centre create an inward tendency. Six of
the illustrations have the central arch accented by decoration; seven
have heavier piers around the central and end arches; six have the end
arches brought out into a nearer plane which effectually finishes the
series. All these examples illustrate the necessary disposition of
arches on a flat wall or façade where the series in the field of vision
must end suddenly, that is, cannot gradually fade away around a corner.
The variety and yet invariability of these devices shows the need felt
for some finish at the end, some balance of the whole with the central
accent, which need, apparently, is not felt for pillars and lintels.

When the arch-series is on a circular structure, such as apses,
porches, and the like, even when it does not entirely surround it, as
an arena or spire, the regular diminishing of the series on either
side, owing to the curve, supplies the finish necessary, and the size
and arrangement of the arches need not vary otherwise. Twelve of the
examples illustrate such a use of the arch, and although in some cases,
Morano Cathedral, Nomantala Church, the arches are continued into the
transepts gradually tapering in size, or are modified in size growing
narrower from the centre, as in the Bergamo Church, such a treatment
is not necessary for finished effect. The difference in proportion
resulting from a curved series, or even on arches carried around a
square corner (as in porches on Goslar and Braunschweig Rathäuser),
where the series is open enough to clearly see its continuity as it
runs into the main building, will suffice to make a series finished
without modifications of the arch-units.

There are many instances of long rows of very narrow arches on
cathedral façades which are too narrow to give outward tendency, or
else they have statues within them which really take the attention
and form a series of vertical units in place of the arches. There
is also the common device of interlacing arches, where a supporting
pillar of another arch stands in the centre of every arch, thereby
always driving the attention backward and restraining it. Perhaps the
natural outward tendency of the arch-series and the necessity for its
limitation can be seen by violations of the principle. Seven of the
examples do not conform to any application of this rule and the results
are not satisfactory so far as the mere series itself is concerned.
Over the right and left doors of the Piacenza Cathedral are sections
of nine arches which end abruptly and do not even meet each other.
The Fredericksborg Schloss at Copenhagen has a row of fifteen arches
enclosing a court. These run into wings on each side, to be sure, but
all seen at once as they are and without central or end modification
they are too sharply cut off and inclined to overstep their limits. The
Loggia dei Lanzi at Florence, with its three wide arches and narrow
pillars, the William Tell Chapel in Switzerland, with only two arches,
illustrate forcibly the tendency of an arch to move outward, to appear
too wide for the superstructure and too "active" unless bound down in
some way. Four arches on the right and left of the façade of Marmonte
Church, but not across the centre, have the same unfinished effect. The
roman arch on one side of the St. Lo Cathedral façade with two gothic
arches on the other defy every principle of repetition and symmetry as
well.

From this survey of one hundred and sixty-five of arch-series we find
through a variety of means a uniformity of purpose in their treatment;
that all point to a common demand, however differently expressed,
according to the function of the series. The series must be prevented
from "running away." It must either run completely around a structure
into itself, or be balanced as a whole so that the attention which
naturally runs off the ends is driven towards the centre. This may be
accomplished by enlarging, decreasing, decorating, or pointing toward
centre of the arch by means of the obliquity of both halves of the
series. It may also be brought by enlarging, decreasing, changing the
plane of the end arches or altering the size of the limiting piers. The
essential value of the arch may be altered by narrowing it, by filling
it with something more important than itself, thereby making it only an
attendant series upon its content, by interlacing it, or by any device
that transforms or revises its outward tendency.

         165 _Examples of Arch-Series._

    45 Go around outside a circular structure.
    32 Go around interior and apses.
    30 Central arch largest.
     2 Central arch smallest.
     4 Ends largest.
     6 Ends smallest.
     6 Central arch accented by decoration.
     6 Central arch accented by upward incline of two halves.
     6 Ends in different planes.
     7 Different width of piers around centre and ends.
     5 Very narrow arches.
     9 Other reasons.
     7 Unaccounted for.

The question discussed in the experiments, as to whether narrower
interspacing was required between units decorated toward the centre,
and units blank, or covered entirely with non-centrally accented
decoration, could not be taken up in the latter analysis. To settle
such a point, illustrations would have to be found of blank and
decorated units of the same shape and size, in the same structure, and
their relative interspacing compared. But no such examples were found,
where the spacing was not regulated by some obvious structural reason
other than pure pleasure in the repetition. This must stand, therefore,
solely as an experimental result.

The use made of difference in plane or end, to facilitate two series
being taken along together, whereas they would be fatiguing if the
same in those respects, has been touched upon in the discussion of
statues and bas-reliefs, and other series of more complicated units.
Where the unit and alternate are both rich and significant, and would
tire the observer by following each other at the distances they are
obliged to be in a series, a slight difference in plane relieves the
situation, and is used largely in monuments, fountains, pulpits, and
such structures.

Many other questions have come up in the investigation which might be
discussed in the same manner as the preceding, but can only be hinted
at in conclusion:

Just what factors make an element and its alternate congruous? What
is the exact relation of lines, which makes the scroll decoration in
a balustrade alternate satisfactorily with an upright support, while
the alternation of the arches in the Colosseum with the Greek pillars
between them is incongruous?

In what does the pleasure in repeated series differ, when the observer
is not certain just what is the repeated element? May there be a bare
rhythmic pleasure, when the series is too far away to distinguish
what the elements are, or when they run together, so that no definite
demarcation is felt between them? Do such series excite a pleasure of
repetition without content as to elements, and does it differ from mere
variation and contrast?

The series of unsymmetrical units was found in the experiments to have
a peculiarly unstable run-on effect similar to that of rhythmic units
and of arches. Are they used in the same kind of cases as the others
were, when a particularly active effect is desired?

Must a space be wholly enclosed, to be taken as a unit?

In a series of projections along a wall, the _projections_ are taken as
the unit, even when they almost meet at the top of the alternate space.
When they actually do meet at the top, the enclosed space becomes the
unit instead.

These questions and others similar might be experimented upon, and
examples of their treatment analyzed, as in the previous questions
discussed.

[Illustration: PLATE V.]

FOOTNOTES:

[Footnote 86: Stratton: Eye-Movements and Æsthetics of Visual Forms,
Philosophische Studien, vol. 20, p. 350, 1902.]

[Footnote 87: MacDougall: Structure of Simple Rhythm Forms, Harvard
Psych. Studies, vol. 1, p. 321, 1903.]

[Footnote 88: MacDougall: Structure of Simple Rhythm Forms, _ibid._ p.
322.]

[Footnote 89: MacDougall: Structure of Simple Rhythm Forms, _ibid._ p.
322.]

[Footnote 90: MacDougall: Structure of Simple Rhythm Forms, p. 325.]

[Footnote 91: Bolton: American Journal of Psychology, p. 223, 1894.]

[Footnote 92: Külpe: Outlines of Psychology, p. 395.]

[Footnote 93: MacDougall: Structure of Simple Rhythm Forms, p. 348.]

[Footnote 94: MacDougall: Structure of Simple Rhythm Forms, p. 349.]

[Footnote 95: Stetson: Rhythm and Rhyme, Harvard Psychological Studies,
vol. 1, p. 455, 1903.]

[Footnote 96: Stetson: Rhythm and Rhyme, Harvard Psychological Studies,
vol. 1, p. 455, 1903.]

[Footnote 97: MacDougall: Structure of Simple Rhythm Forms, p. 377.]

[Footnote 98: Perrot and Chipiez: Art in Ancient Egypt, p. 100.]

[Footnote 99: Perrot and Chipiez: Art in Assyria and Chaldea, p. 126.]

[Footnote 100: Perrot and Chipiez: Art in Persia.]




THE FEELING-VALUE OF UNMUSICAL TONE-INTERVALS

BY L. E. EMERSON


Modern theories of melody start always with the presupposition that the
scale must be composed of tones having the simple mathematical relation
to one another of 2, 3, 4, 5, 6 (and by Meyer 7) and their multiples
in order to give pleasant combinations of successive tones. But the
question arises whether other tone-combinations which given together
appear disharmonious may not, by their mere acoustical difference,
similarity, and contrast, awake definite feelings of pleasure. And
if such feeling-tones exist independently from harmony it is evident
that they would enter into every melody in addition to the strictly
musical feelings of harmony and that they deserve consideration as a
factor of music. It would not even appear impossible that if every
successive tone-distance has its particular natural feeling-character,
the distances of successive harmonious tones might be only through
secondary factors as habit and training preëminent among the various
possibilities of combinations. A tone-consciousness, which under the
guidance of experiences of harmony has been trained in our musical
tone-relations, must give instinctive preference to such successions
as our melodies offer. But if we artificially inhibit the conscious
relation to our musical system by introducing a continuous tone-series,
or at least one of steps much smaller than musical intervals, do
we destroy the possibility of pleasure, and if not, do we find the
pleasure in the musical interval stronger than that in other instances?
That even the musical subject introduced into the realm of smallest
tone-steps can easily forget and inhibit his normal standards is well
known; the whole acoustical perspective seems changed by the new
intervals, and the subject begins at once to build up a new temporary
system of relations. The experiments in Wundt's laboratory have shown
that in such cases the theoretical judgment of distances is indeed
quite different from the standardized one; the octave may appear
equal to the higher fifth. I wanted to study in a similar way the
feeling-value in such a state of musical disorientation, when all
imaginative representations of our musical intervals are inhibited.

The instrument I used was an Appun Tonmesser giving reed-tones from 128
to 512 vibrations in intervals of 4 vibrations between adjacent tones.
The intervals with which I experimented varied from 4 to 88 vibrations
in steps of 4. The observers were all experimental psychologists, and
varied in musical discrimination from a very low to a very high degree
of natural ability and skill.

The observer reported his pleasure in the progression given, in the
traditional grades of 1 to 7, where 1 represents the greatest degree
of pleasure, 2 means very pleasant, 3 pleasant, 4 indifferent, 5
unpleasant, 6 very unpleasant, and 7 most unpleasant of all.

The immediate problem was: What is the relation between the width of
interval used and the pleasure got by hearing the motive _a-b-a_ and
_b-a-b_, where _a_ is always the lower tone. The method of procedure
was to take a fixed tone (460 vibrations in the first case) and get
a series of observations on successive progressions _b-a-b_ where
_a_ differed from _b_ by 4, 8, 12 ... 56 vibrations. The greatest
difference thus is approximately a musical whole tone. Then a series of
observations was taken on _a-b-a_ where _a_ similarly differed from _b_
by 4, 8, 12 ... 52 vibrations. The progressions were given in irregular
order, that there might be no chance of the observer getting into a
fixed habit of replying. The intimate relation between the pleasure
in successive musical tones and the pleasure in musical harmonies
suggested naturally the question whether the feeling-value of these
unmusical progressions was not somehow dependent upon the affective
character of the simultaneous presentation of the same tones. Therefore
after a progression had been given once and judgment recorded, the
two tones used were given as a "harmony," that is simultaneously,
and a judgment taken as to its agreeableness. This was immediately
followed by the same progression, thus giving opportunity to observe
the relation between the feeling-tone of the interval as it appeared in
successive and in simultaneous presentation.

The results of this part of the investigation are graphically
represented in the following plates. Tables I and II indicate the
feeling-value of _a-b-a_ where _a_, the lower tone, is 460 vibrations,
and _b_ is from 4 to 56 vibrations in addition, and the feeling-value
of _b-a-b_ where _b_, the higher tone, is 460 vibrations and _a_ is
from 4 to 56 vibrations less.

[Illustration: PLATE VI.]

The base-lines from which the vertical lines to the curves are drawn
represent the feeling-tone 4, the indifference-point. Above comes 3,
2, 1 and below 5, 6, 7; each square represents a unit. The horizontal
abscissæ represent the width of the interval; the arrows indicate the
musical intervals. The observers are given by initials. The first
evident fact for both average curves of Plate V is that the maximum
pleasure does not coincide with a musical interval, but comes with an
interval four or eight vibrations less than either the half or the full
tone of the musical scale. While in both cases the first elevation
of the curve comes before the semi-tone, _b-a-b_ shows a decrease of
pleasure as the whole step is approached while _a-b-a_ rises again. The
order _a-b-a_ is liked better than _b-a-b_.

Plate VI gives the "harmony" curve for the same tone-combinations, and
it is clear at the first glance that the curves for the simultaneous
tones do not correspond to those for the successive ones; in many
respects they are directly the opposite. The hypothesis that the
pleasure in such an amusical "melody" results from the resolution
of the corresponding "harmony" is thus untenable; both are highly
independent of each other. Yet, here too we notice the insignificance
of the musical interval, while the strong pleasure in the tones
different by 4 vibrations only refers probably to the complete fusion
of the tones; there arises a direct enjoyment from the four waves of
sound in every second, given by the beats. The pleasure-curve of these
simultaneous tones indicates of course that the inhibition of the
musical dispositions and expressions holds over from the successive to
the simultaneous series. The pleasure is thus clearly different from
that in real harmony.

Plate VII finally gives the "melody" curve for _aba_ and _bab_ with
changes from four to four vibrations when the interval started with is
larger than a full musical step. In _aba_ the _a_ is 384 vibrations
and _b_ varies from 436 to 516, the variations lying thus between the
musical Second and the musical Fourth. It is evident that here again no
feeling-preference is given to the musical intervals.

The question arises whether such small tone-intervals of amusical
character allow the construction of more complex combinations of
æsthetic value. Can we have amusical micromelodies with their own
completeness and feeling of end? The following experiments represent a
first step into this field. We used three tones only, _a, b, c_ in 26
different combinations, and each of the 26 variations with intervals of
4, 8 and 12 vibrations between _a-b_ and _b-c_. Each of the resulting
78 "melodies" was given repeatedly to six subjects in a time-order
which allowed one second for each tone. The subject had to judge on
the pleasantness of the whole progression and had further to judge
whether it produced a feeling of end or not.

The combinations followed in the experiments in this order: _abc_,
_cbabc_, _abcb_, _cba_, _abcba_, _cbab_, _bcba_, _cbabcb_, _ababc_,
_babc_, _abcbab_, _babcba_, _cbcba_, _bcbabc_, _abca_, _acba_, _acb_,
_cbac_, _abcab_, _cabc_, _cbacb_, _acbab_, _cab_, _bca_, _cabcb_,
_bac_. The lowest tone was varied between 200 and 444 vibrations; _b_
and _c_ were thus always still less distant than the next musical
tone. The chief results may be shortly characterized as follows.
There are hardly any judgments of indifference, the combinations are
always decidedly pleasing or unpleasing. Of course a certain training
in the apperception of such small-interval melodies preceded the
real experiments and produced an attitude of adjustment to amusical
relation. If we are in the midst of musical tone-relations and go
over directly to such miniature intervals, we are seeking for the
fulfilment of the habitual expectation and feel dissatisfied, or in
the best case the procession is an indifferent chance combination.
But as soon as a certain training with small intervals has inhibited
the strictly musical expectations, a new setting of judgments with
new standards comes in and a new source of pleasantness is opened.
Of course even then no extreme feelings are to be expected; while
the indifference-judgment 4 is lacking, the strong pleasure and
displeasure, the judgments 1 and 7 are completely lacking too; three
fourths of the judgments are 3 and 5. The pleasantness is decidedly
more frequent than the unpleasantness, and this relation increases
with the interval. The differences of four vibrations were especially
with the higher tones hardly distinct for some of the subjects. Among
288 judgments in each group there were 150 pleasant and 138 unpleasant
when the distances between _a-b_ and _b-c_ were four vibrations, 208
pleasant and 80 unpleasant when the distances were 8 vibrations, and
226 pleasant and 64 unpleasant when the distances were 12 vibrations.

The order of pleasantness expressed by the fraction of judgments of
pleasantness and unpleasantness is the following: the largest number of
pleasant feelings belonged to the figures _cbab_ and _bac_, immediately
followed by _abcb;_ the further order downwards in affective value was:
_cab_, _cbac_, _babc_, _abca_, _cbcba_, _ababc_, _abc_, _cabc_, _acba_,
_cbabcb_, _bcba_, _acb_, _abcba_, _cba_, _cbabc_, _abcab_, _babcba_,
_cbacb_, _acbab_, _bcbabc_, _abcbab_, and _cabcb_ as least pleasant.

[Illustration: PLATE VII.]

As to the feeling of end or æsthetic completeness the results are
similar and yet independent. In a few cases the answer was "doubtful,"
but in the overwhelming majority a definite reply was given; and
while the judgment of completeness was by far more frequent in the
pleasant combinations than in the unpleasant ones, yet often the
unpleasant processions appeared as complete and the pleasant ones
as incomplete. Here again the feeling of completeness grows with
the interval, being smallest for the figures with distances of four
vibrations. But most characteristic seems the fact that the feeling
of end is in no way as in music dependent upon the return to the
starting-point. The combinations which involved such return to the
"tonica" show in no way a preponderance of judgments of completeness.
If we order the results according to the number of this æsthetic factor
the figures _acba_, _cbac_, and _cabc_ stand very low, giving in the
majority of cases the suggestion of not-completeness in spite of their
return to the beginning, while the figures of the type _abcb_, _cbab_,
or _cba_, or even the complex _babcba_, suggest in a majority of
judgments the feeling of an end. The feeling of an end comes, according
to the subjective reports of the observers, with an "internal unity of
meaning" of the phrase given. This unity of meaning is here evidently
quite independent from any simple mathematical relation.

The music-like quality of the figures was emphasized frequently in
the subjective records. "I just enjoyed the progressions as music."
"The elements are the same as in music." A melody of 384, 392, 400
was called a "very mournful strain"; 444, 452, 460 "Wagnerian motive;
Tristan and Isolde"; and the same tones in another order "Very
pleasant; expressed a pathetic resignation," or "Sounds like a little
piece of music"; and so in most varied forms.

The basis of these experiments is of course by far too slender to
build on them a theory, yet our results suggest at least a greater
interest in the æsthetics of those tone-combinations which are
excluded from our regular music. This interest is reënforced by the
self-observations of all participants. They felt strongly that after
all our musical pleasure in melody does not belong intrinsically to
the tone-perception, but is learned and acquired like the grammar of
our mother tongue. Such grammar too controls completely our internal
demands for expression, and yet the learning of a different language
can bring a new adjustment and a new set of psychophysical dispositions
for linguistic demands. That whole apparently natural demand for the
tone-combinations which give fusion and consonance can be inhibited
during the listening to amusical combinations as soon as a short
training in miniature intervals changes the acoustical perspective.

The development of instrumental music demanded evidently the selection
of distinctly separated tones and of intervals which give harmonious
combinations. The external conditions of resonant chambers may have
reënforced this selective process of historical music. It is certainly
different with oriental nations, which produce music not in resounding
chambers but in the free air and who are singers and not players, using
instruments mostly for producing a mere body of tone as a background
against which the melodies move; their intervals appear to our musical
ear at first bizarre, and yet there too we are readjusted to the
new dispositions for satisfaction with unsuspected quickness. We
have no right to identify æsthetic pleasure in successive tones with
the pleasure in our conventional music with the simple mathematical
relations which alone give the pleasure of fusion; but being accustomed
to this system of harmonies and being trained to expect it also in the
resolved form of the melody, we need indeed an inhibition of habits and
a certain new training till the more modest pleasure in amusical tone
progressions comes to its natural right.




ASSOCIATION, APPERCEPTION ATTENTION




CERTAINTY AND ATTENTION

BY FRANCES H. ROUSMANIERE


The results of the experiments on the feeling of certainty which I
have conducted fall into two divisions--those on the nature of the
feeling itself, and those on the effect of voluntarily attending to
certain aspects of a total experience upon certainty in the judgments
as to the constitution of that experience. The problems of the first
division are: Are there different kinds of certainty? In any one kind
of certainty are there degrees, and if so, are these of a limited or an
unlimited number? Can certainty be analyzed into elements? The problems
of the second division are: Can it be said that in the report of any
experience the judgments made with the highest degree of certainty will
be confined to an attended-to group, and if not, will there be more
there than elsewhere? In such a report will the direction of voluntary
attention toward certain aspects materially alter the distribution
of the judgments of the highest order of certainty over the various
aspects of any given field?

These two divisions are so distinct in problem and result as to make
it seem best to describe them as independent experiments. As some
interesting results on the relation of error to the different grades of
certainty and to the effect of attention developed in connection with
this second division of the experiment, those results are given also.

In general the same subjects took part throughout the experiments.
One, an instructor in Harvard University, whom I shall call K, was
not subject for the second division of the experiment. Two others, E
and H, both graduate students in Harvard University, could not serve
as subjects in an important part of the first division. Of those
remaining, B was a student in Radcliffe College, F an instructor in
Harvard University, and A, C, and D graduate students in Harvard
University. These last five were my subjects for all parts of the
experiment.


I. THE NATURE OF THE FEELING OF CERTAINTY

The general method here was, of course, the method of introspection.
Situations were created about which the subject might be expected
to make judgments with different sorts or different degrees of
certainty, if such should be possible. He was then questioned as to
his experience. The method has the fault of all introspective methods,
viz., its results can in no case be verified. The results here are none
the less suggestive, and, for the second problem, at any rate, definite
enough to be convincing.

Most of the experiments were conducted in connection with visual
fields. In working at the first problem which we have now to consider,
however, the certainty connected with the dermal sensations and that
connected with the simple reasoning process of addition were also
examined. The apparatus used consisted of three sets of cards. On
one set were pasted geometrical shapes cut from colored paper, and
black and white letters or figures. Each of these cards was shown to
a subject for a second and a half, or two seconds. After the exposure
he told what he judged to be on the card, giving all that he could
about the nature of his feeling of confidence (or certainty) for each
judgment. On the second set of cards square pieces of tin, smooth
rubber, rough rubber, cotton, felt, undressed kid, leather, eiderdown,
flannel, coarse and fine sandpaper, and pricked paper were stuck, six
on each card. The experimenter passed these cards so that these bits
of material rubbed against the forefinger of the subject, while a
curtain kept the card and the hand hidden from the subject's sight.
Here, again, the subject judged of what had been on the card, just as
he had done after seeing each of the first set of cards. Small sample
cards, each having pasted upon it a piece of one of the substances
used, were also behind the curtain, and the subject was allowed to feel
of these as much as he wished while giving his report. Such sample
cards were required because of the underdevelopment of the association
of names of any kind with the dermal sensations. A single card with
three groups of figures for addition upon it made up the third part
of the apparatus. Here the subject was asked first to add the columns
rapidly and to introspect as to his certainty of the correctness of the
different results; then to go over the addition again, and yet a third
time, and to compare his feelings of certainty in the different cases.
The introspection was developed partly through the help of questions
put by the experimenter, but in asking these questions great care was
taken to prevent their influencing the judgment of the subject. Some
observations made by the subjects during the second division of the
experiment (also conducted in connection with visual fields) are, also,
introduced here. Apart from this, the experiments on the feeling of
certainty connected with this sense of sight were greater in number
than the other experiments; and it is those that have given us most of
the data for answering the second and third problems.

The subjects did not agree in their answers to the first problem. Some
found not only that the certainty connected with their belief in the
results of their addition seemed to be of a distinct type from that
connected immediately with the sense of sight, but also that there
were different sorts of certainty connected immediately with the sense
of sight itself. Others found but one kind of a feeling of certainty.
All agreed, however, that so far as the kind or kinds of certainty
associated with them was concerned there was no difference between
the sense of sight and the dermal senses, so that it would seem to be
true that any distinctions which are to be found within the feeling
of certainty will not be distinctions springing from the difference
in the sense-organs. Within the sense of sight, however, subjects B,
E, and F divided their feelings of certainty into two classes,--an
absolute feeling of certainty which they felt could not be shaken,
and a feeling of confidence which they would act upon but which they
felt might be shaken by questioning, and which seemed different by
more than degree from the feeling of certainty proper. Subjects A,
C, K and H found no such marked distinction between their feeling of
greatest certainty and all lesser feelings of conviction. Subject D
at one time felt that the distinction into two such distinct classes,
the definitely certain and the more wavering, fitted his experience,
and at another time said that it seemed to him that each degree of
conviction stood for an unique feeling of certainty and that any two
of them were as different from each other as any other two. A second
division of the feelings of certainty into two classes is to be found
with subjects A, F and H. This developed in connection with the
visual experiments again. The distinction here may be called one into
psychological and logical certainty. The latter rests on reasoning
either from the probable character of the field, or from a feeling
as to its general character, to the nature of some detail. We shall
notice the characteristics of these two classes later. One subject, A,
further distinguished as different the feelings of certainty connected
with the two methods of logical certainty just given. The others made
no such distinctions. In the experiment with the columns for addition
only six subjects, A, B, C, D, F, and K took part. Of these the two
who had made the distinction into psychological and logical certainty
with the visual experiments (subjects A and F) again made the same
distinction. Subject F, however, who had had occasion to do a good deal
of important work with statistics, found practically no element of
logical certainty in connection with his addition, though it seemed to
him that what confidence he felt in his result should be distinguished
from the psychological certainty he had had as to the character of
the visual fields. Subject B felt no certainty in her results except
as she could so hold the process together as to have what seemed to
her a simultaneous experience. When she had to judge of the results
of a set of successive experiences that could not be so unified, she
characterized her state of consciousness not as holding a feeling
of certainty or of uncertainty, but as simply lacking any feeling
of certainty. The other three subjects found no difference between
the feelings of certainty and uncertainty associated with visual
experiences and those associated with the process of addition. As a
whole, it seems then that we must answer our first problem by saying
that the case seems to be different with different individuals. With
some the highest grade of certainty associated with a sense-experience
is sharply distinct from the other grades, and with some again there
appear at least the two general classes of psychological and logical
certainty. On the other hand, there seem to be people for whom the
feeling of certainty has no such sharp distinctions of kind within it.

The results as to the second problem may be more briefly and more
distinctly given. No subject found any evidence that the number of the
grades of certainty which he could distinguish would be limited by
anything except his keenness in introspection, although in the simple
tests given for the experiment, four was the greatest number of grades
distinguished at any one time. Two of the subjects (B and F), who set
the highest grade of certainty apart from the judgments made with
lesser confidence, said that there might be degrees within that higher
grade as well as among the "uncertainties." There was no evidence that
logical certainty differed from psychological in respect of the grades
to be found within it, and some evidence that they were alike in that
respect, although logical certainty was less carefully examined. It
would seem, then, that our second problem is to be answered thus: There
are degrees present in some if not in all kinds of certainty, and there
is no evidence that the number of these degrees is limited.

It was not generally found possible to analyze the feeling of certainty
into a sum of elements, although certain characteristics seemed to be
persistent in it. Here again there is marked individual variation. The
general test used for the difference in degrees of confidence was the
question "On which judgment would you risk more?" This satisfied every
one as a true criterion for such distinctions, but subjects H and C
said that for them the feeling of certainty had a much more distinct
relation with the past than with the future. Perhaps for that reason,
subject H proposed the test "Which judgment could I be converted from
most easily and simply?" The distinctness of an image had something to
do with the feeling of certainty for subject C. Beyond this, he could
not characterize his feeling. Neither was he sure that the degree of
certainty varied exactly with the degree of distinctness. Subject D
found that all objects about which he made judgments of which he was
certain were present to his mind in the form of distinct images; but
did not feel that that covered all that was to be said of the feeling
of certainty. The number of images present, as visual and auditory,
seemed to increase the degree of certainty for him. Subject F could
give no characterization of his feeling of psychological certainty. His
feeling of logical certainty seemed to spring largely from a feeling of
consistency between the present experience and his past experiences.
With subjects A, B, and K the vividness of an image was a strong
determining factor in the degree of certainty felt in any judgment, but
again was not the whole story. Something they could not characterize
was also present for A and B, and, as well, a feeling of more or less
perfect congruence between an image and the general character of a
field. (This introspection developed in connection with the visual
experiments.) Among these eight subjects we have but one (K) who is
satisfied with reducing certainty to a set of elements.

To my mind the most valuable thing to be gained from this division
of the experiment is the suggestion that there are definite types
of certainty, and that people may be classified by these. There are
obviously marked individual variations as to the characteristics of
this feeling. I should expect from my work this year that two pretty
distinct types could be discovered. For one of these, certainty in a
judgment as to an experience would rest very largely upon the vividness
of an image; for the other, upon the congruence of an image with other
previously accepted images, that is, the absence of conflicting images
when the experience judged about is imagined part of a wide setting
of past experiences. I should not expect either element of certainty
to appear absolutely, without the other form. For many people one
element would predominate in certain fields, as in judgments regarding
sense-experiences, the other in the more logical fields. For some,
again, perhaps, the two would be nearly coördinate in every experience
of certainty. But for some subjects, as, I think, for subject K here,
the vividness of the image would always be the determining factor,
while for others, as for subject H, congruence with wider experience
would be much more important. This classification of subjects according
to their types of certainty might develop into a much more complicated
affair. The experiments described here have gone no farther than
to suggest lines along which it may perhaps run. There may be other
elements equally important with these two. A set of experiments
consisting of attempts to raise uncertainty to certainty would bring
out the essentials of certainty from a new point of view, and would,
perhaps, test this theory that individuals may be classified according
to the types of their certainty, in the most satisfactory manner.


  II. THE EFFECT OF VOLUNTARILY ATTENDING TO CERTAIN ASPECTS OF A TOTAL
      EXPERIENCE UPON CERTAINTY IN THE JUDGMENTS AS TO THE CONSTITUTION
     OF THAT TOTAL EXPERIENCE.

As has been said, judgments as to the elements of visual fields were
tested for this part of the experiment. The apparatus used was the
following: The subject was seated before a low table which was shut
from his view by curtains and boards. He looked down upon the table
through an opening into which a camera-shutter had been fitted. This
shutter was set for a two seconds' exposure and opened by means of a
bulb which the subject held in his hand. Just before each exposure,
the experimenter placed a card on the table below the camera-shutter.
The set of twenty cards so used were alike in that the background for
all was gray and the objects pasted upon the cards black letters and
numerals and simple geometrical figures of chosen shapes and colors.
No color was repeated on any one card. The cards were different in the
choice and arrangement and in the number of objects used. The number of
letters and numerals on any one card varied from two to five, the total
number of objects from eight to twelve. A white card on which were
pasted dark gray samples of each of the eleven shapes used, together
with a card of the background of those shown in the experiment on which
were pasted torn scraps of the eleven colored papers used, was always
in sight at the subject's side. A camera-shutter, experiment cards and
sample cards thus made up the apparatus.

The presence of the sample cards needs explanation. They stood for
the attempt to place the colors and shapes on the same footing as the
letters and numerals. Their presence, in the first place, and, as well,
the limitation of the number of letters and numerals used, did away
somewhat with the advantage that letters and numerals naturally have
for ease of naming. In the second place, the use of a new color for the
sample shapes and the absence of definite shape in the sample colors
helped to keep the colors and shapes more distinct. With the help of
these cards it seemed that we could properly hold we had a a visual
field of three very nearly coördinate sets of elements.

The experiment as a whole, as conducted, had four phases which, except
for one particular, were exactly alike. The subject's attention was
directed toward a certain aspect of the field by (1) asking him before
each exposure (or less often if that appeared unnecessary) to attend
to that aspect, as, for instance, to the colors present, and (2)
taking care that any questions asked should tend to strengthen rather
than counteract the effect of that voluntary attention. At a given
signal the subject pressed the bulb which opened the shutter. On the
closing of the shutter he reported what he had seen. This report the
experimenter recorded almost in the subject's own words, and later
tabulated in the manner described presently. So far as giving the
objects present was concerned, the report was given almost invariably
without any suggestion by the experimenter as to the possibilities of
the field. To help the subject distinguish the amount of confidence
which he had in the judgments that such or such objects were present,
however, the experimenter frequently asked such questions as, "Would
you risk more on the fact that there was a square in the field than
on the fact there was something blue there?" In giving his report the
subject pointed to the sample cards or spoke, as he might wish. He was
also allowed to be as leisurely or as rapid in giving it as he chose.
A half-minute interval elapsed between the end of each report and
the signal that the shutter be opened again. No persistent effort to
distract the subject's attention was made then, though conversation on
other topics was frequently carried on. The point in which the phases
of the experiment differed was in the aspect of the field to which
attention was called. In the first, this was the shapes, in the second,
the colors, in the third, the letters and numerals, and in the fourth,
the number of objects in the field. Fixing the attention upon the
number of objects in the field served to distribute it equally over all
the groups represented there. The general method of calling attention
to the different aspects and of learning the effect of such attention
was, as has just been said, the same for all phases.

As a preliminary to making up the tables here given, from which we are
to answer our problems, the experimenter first tabulated the reports
of the subjects in such a way as to show how many judgments (correct
and incorrect) of each of the four grades of certainty adopted for
this division of the experiment were made by each subject on each card
for each group on the card (shape, color, or letter or numeral). From
these tabulations the tables that follow were in turn compiled.

The number of grades of certainty adopted for this division of the
experiment is obviously decidedly arbitrary. Grades of certainty there
surely are. The introspection of the subjects develops that clearly,
as has been stated. But there is no reason in the conditions of the
case for holding to the number four, as is done here. In giving the
results for which the experiment was undertaken, I shall, indeed,
confine myself to studying the range of the judgments made with as high
a grade of confidence as the subject believed he should ever have. This
is called certainty (1) or certainty proper. But for the tributary
discussion on the relation of certainty and error, the consideration
of three other grades used in the report and early tabulation, is also
introduced. This lowest grade (4) might better be named "as complete
uncertainty as will admit of one's making any judgment." The other two
are intermediate. It was at first intended to give the results with
regard to the effect of voluntary attention upon the place of these
grades of certainty, also, but such a discussion has been omitted
because it promised to add very little more than complexity to the
report. Besides this, the classification into these lower grades is
too purely approximate to make the distinction there of great value.
For judgments of the order certainty (1) we have the test, "Are you
as certain of this as you can imagine being in an experiment of this
sort?" but no such test for the other grades could be found. Yet,
though he tended to omit judgments of the lower grades of certainty,
each subject seemed to find four grades a convenient number to use in
giving his report.

The number of experiment cards used varied with the subjects. E had so
clear a memory of the cards that after as many as ten had been shown,
he found difficulty in distinguishing his memory of the one which he
had just seen from that of others seen earlier. Ten cards only were
used in his case. A, B, D, and F showed signs of fatigue after fifteen
cards which made the value of any later results questionable. C and K
showed no such signs of fatigue. The same set of cards was, of course,
shown any one subject for all four phases of the experiment. Those
omitted were the last ten or the last five of the complete set as the
case might be.


TABLE I

  % of cards where  % of cards where   % of cards where   % of cards where
  certainty (1)     certainty (1)      certainty (1)      certainty (1) is
  appears in the    appears elsewhere  in attended-to     stronger outside
  attended-to       than in the        group only.        than within the
  group.            attended-to                           attended-to
                    group.                                group.

    A  91%                49%                49%                13.3%
    B  97                 91                 83                 28
    C  95                 67                 30                 16.6
    D  93.3               69                 24.5               11.2
    E  96.6               83.3               13.3               26.6
    F  88.8               30                 53.3                8.3
    H  91.6               70                 26.6               10

Table I answers the first part of our first problem promptly. Every
subject gave judgments of the order certainty (1) about groups other
than that attended to, in the case of a very considerable percentage of
the cards. True, again in the case of a considerable (though generally
smaller) percentage of those cards, each subject confined his judgments
to the group attended to. The fact of individual variation stands out
again here; and, moreover, the conclusions drawn should be qualified
slightly because of the fact that it was often possible for the
subjects to give all the letters and numerals on the cards, and still
have, as it were, some attention left over for the other, supposedly
non-attended-to groups. Such reaching beyond the properly attended-to
group never seemed to be possible with either shapes or colors. Aside
from this, however, it is clear that judgments of the highest grade of
certainty were by no means limited to the group attended to.

This same table answers, also, the second part of the problem. Each
subject found certainty of the highest grade sometimes stronger outside
than within the group which held his attention. It is, of course,
practically impossible to make absolutely certain that each subject's
attention was invariably held to the group toward which it was turned,
yet the percentage where certainty was stronger outside than within
such groups seems large enough, in some cases, at least, as with
subjects A, B, and H, to warrant our answering this second part of the
problem in the negative. I should feel, however, that this was answered
less definitely than was the first part of the problem. We may say,
then, that the judgments made with the highest degree of certainty
about a visual field will not be confined to the group attended to, and
that we have strong evidence pointing toward the belief that we cannot
expect there will invariably be more of such judgments within the group
attended to than outside it.


TABLE II

    # 1: % of judgments of certainty
    (1) given to each group in phase
    I (or when shapes were attended to).

    # 2: % of judgments of certainty
    (1) given to each group in phase
    II (or when colors were attended to).

    # 3: % of judgments of certainty
    (1) given to each group in phase
    III (or when letters and numerals
    were attended to).

    # 4: % of judgments of certainty
    (1) given to each group in phase
    IV (or when the attention was
    equally distributed over all the
    groups).

                                    1      2      3      4

    Subject     Shapes      (a)    94%    38%    18%    31%
       A        Colors      (b)     5     61     20     59
                Letters and (c)
                Numerals            0      0     61      9

    Subject                 (a)    48     39     21     33
       B                    (b)    43     60     29     38
                            (c)     8      0     51     28

    Subject                 (a)    88     15     26     34
       C                    (b)     8     77     28      8
                            (c)     5      7     47     58

    Subject                 (a)    60     13     11     40
       D                    (b)    19     67      2     37
                            (c)    19     19     86     23

    Subject                 (a)    51     15      0     37
       E                    (b)    22     56      8     29
                            (c)    26     28     91     34

    Subject                 (a)    66     12      0     50
       F                    (b)    14     77      0     27
                            (c)    19     10    100     23

    Subject                 (a)    43     21      7     24
       H                    (b)    29     52      4     19
                            (c)    27     26     88     57

The most interesting part of this division of the experiment is brought
out in Table II in answer to the problem, "Will the place of voluntary
attention materially alter the distribution of judgments of the
highest order of certainty among the given groups?" In every case the
percentage is affected, in most cases, greatly affected. Take the case
of subject A, for instance. Although, when his attention is equally
distributed over the field 59% of the judgments we consider were of
colors, yet when his attention was fixed on shapes and on letters and
numerals this fell to 5% and 20% respectively. When it was fixed on
colors, it rose, indeed, only to 61%. When, however, subject A fixed
his attention upon the letters and numerals, 61% of the judgments
were confined to the group attended to,--the same percentage as when
colors were the attended-to group,--although, when his attention was
distributed over the whole field, the percentage of these judgments
about the group of letters and numerals was 9% only. When shapes were
attended to, the 31% of the fourth phase of the experiment rose to
94%,--almost all of the judgments of the highest grade of certainty
that were given were judgments about shapes. A similar study of the
results given in the table can be made for the other subjects. The
degree of change varies with the subject and with the group, but
always there is some change, and often a very marked one. In this
experiment the place of voluntary attention clearly did alter, and
alter materially, the proportion of judgments of the highest order of
certainty made about any given group.

That, indeed, would seem to me to be the answer of this experiment to
the question as to the effect of voluntary attention upon certainty
in one's judgments. Every subject showed a tendency to have more
certainty in those judgments which were made about that aspect of the
field toward which his attention was directed. Yet, on the other hand,
this was a tendency only, one not strong enough to make it possible to
predict beforehand exactly how great a proportion of the judgments in
which he had the highest degree of confidence would be limited to that
field, or even to be sure in every case that the greater proportion of
those judgments would be so limited. The place of voluntary attention
has an influence upon the subject-matter of the judgments made with
certainty about a visual field just seen, but an influence of varying
and uncertain strength.


TABLE III

    1 = % of mistakes in judgments of certainty (1).
    2 = % of mistakes in judgments of certainty (2).
    3 = % of mistakes in judgments of certainty (3).
    4 = % of mistakes in judgments of certainty (4).
    x = no judgments of that kind given.
                                           1        2        3        4

    Subject A (in giving shapes)  (a)      7%      10%       0%      100%
              (in giving colors)  (b)      2       14       23         0
              (in giving letters
               and numerals)      (c)      0        0        x         x

    Subject B                     (a)      2       10       10         0
                                  (b)      3        6       20        25
                                  (c)      4       50        0         x

    Subject C                     (a)      1        8       10         0
                                  (b)      4       14       14         0
                                  (c)      1        0       16         0

    Subject D                     (a)      3        4        0        16
                                  (b)      1        6        0         0
                                  (c)      5        0        0         0

    Subject E                     (a)      2       10       25        50
                                  (b)      1       33       40         0
                                  (c)      0        0        0         0

    Subject F                     (a)      1       25        7        25
                                  (b)      4       15       29        26
                                  (c)      3        0        0         x

    Subject H                     (a)      3        6        5         0
                                  (b)      6        2       15        15
                                  (c)      4        0        0        20


TABLE IV

    Label 1: General % of mistakes
    in judgments
    of certainty (1).

    Label 2: % of mistakes in judgments
    of certainty (1) about
    attended-to groups.

    Label 3: General % of mistakes
    in judgments
    not of certainty (1).

    Label 4: % of mistakes in
    judgments not of
    certainty (1) about
    attended-to groups.

                   1        2        3        4

    Subject A      4%       1%      17%      27%

    Subject B      3        2       10        7

    Subject C      2        3        9       24

    Subject D      3        4        4        0

    Subject E      1        2       22       34

    Subject F      3        1       21       23

    Subject H      4        6        6       10

The results given in Tables III and IV were compiled from the same
records as those of the two Tables just discussed. They give the
relation of error to certainty and to attention, as that relation was
developed in this experiment. No experiments were conducted with these
relations of error primarily in view, but the results developed in
connection with the problem of the effect of attention upon certainty
in one's judgments.

Both Tables show again marked individual variation. They suggest to me,
in the first place, a further line of investigation in the same field
and for the same purpose as those investigations which L. William Stern
outlines in an article[101] entitled _Aussagestudium_. This further
line is the testing subjects to learn the probable relative correctness
of the judgments made with different degrees of confidence. Although a
comparison of the first and third columns in Table IV makes it clear
that the proportion of mistakes for the highest grade of confidence is
lower than for the other grades taken together, there is a very marked
difference among the subjects to be noticed. The difference in the two
percentages is, for instance, very slight in the cases of D and H, and
very great in the case of E. It is interesting to notice with regard
to E that while he has the lowest percentage of mistakes for certainty
(1), he has the highest percentage for the group of certainties (2),
(3), and (4). In the more detailed percentages given in Table III we
see further that in certain fields and sometimes in all fields (as
with subject C) judgments made with the lowest grade of confidence
were invariably correct. Such Tables might be of help in a case where
the evidence of eye-witnesses conflicted. We might perhaps learn that
witness N made a large proportion of mistakes where he was absolutely
certain, whereas witness M was seldom wrong in judgments in which he
had a low degree of confidence. Even when the probity of both was
unquestioned, we should not then assume that N was more probably right
because he had so much more confidence in his judgments than M had in
his. A much longer and more comprehensive set of experiments would be
necessary before we could feel that we had at hand a table from which
to work in this way.

The question of the effect of voluntary attention upon error, for
answering which Table IV was compiled, brings out again the marked
individual variation among these seven subjects which has shown itself
in practically all parts of the experiment. Some effect seems to
have been produced always, but this was sometimes to give a larger
percentage of mistakes in the attended-to groups and sometimes a
smaller. With A, B, and F the percentage of mistakes in certainty (1)
was lower for the groups attended to than for the total number of
judgments of that order. Only with subject B, however, is this true
of the group of lower grades of certainties also. On the other hand,
with subjects C, D, E, and H the percentage is greater for certainty
(1) in the groups attended to than for certainty (1) in the collection
of all the judgments of certainty (1) taken together. Here, too, in
the case of subject D, the results with regard to the lower grades of
certainty reverse those for certainty (1). Thus all four possibilities
as to the kind of influence of voluntary attention upon certainty
appear. We cannot say that the place of voluntary attention will tend
to affect the percentage of error in any given way. We can only say
that apparently it made some difference with each subject. It might be
found by further experimenting that the character of this difference
is associated with some other characteristic of either attention or
the feeling of certainty, as, for instance, with the ease with which
attention is held to the chosen field or with the type of the subject's
certainty.

Like all experiments, these open up further questions quite as much
as they answer those toward which they are aimed. To repeat something
of what has already been said, I feel that what it has established is
(1) that introspection develops distinct grades of certainty in the
case of every individual, (2) that the particular characteristics of
the feeling of certainty vary markedly among individuals; (3) that the
feelings of certainty associated with the different senses are not, as
feelings of certainty, to be distinguished from each other; (4) that
the judgments of the highest degree of certainty which are made about
the constitution of any visual field just seen will not be confined to
the group in that field toward which the attention is directed; and
(5) that such fixing of the attention will, nevertheless, materially
alter the subject-matter of such judgments of greatest certainty. The
rather vague statement that the percentage of error is not surely less
with the judgments of a group because attention is fixed on that group
may perhaps be added as a sixth conclusion. The most interesting and
promising of the problems which the experiments seem to me to raise
are: (1) the problem, are there such definite types of the feeling
of certainty that people may be classified according to their types,
and, if so, what are the types and what their relation to other
psychological characteristics of the individual? (2) the problem, what
will be the result of careful and trained introspection as to the
relation of so-called logical and psychological certainty and in what
fields do these appear for different individuals? (3) the problem,
how can a test for grading the probable percentage of error in the
judgments of different grades of certainty made by any one person be
constructed? and (4) the problem, how are such facts as those given
in Table IV to be connected with the effort required for attention,
the type of certainty of each subject, etc.? Other problems could, of
course, be suggested, but these, I feel, mark the steps that naturally
follow the experiments described here.

[Illustration: PLATE VIII.]

FOOTNOTE:

[Footnote 101: Stern: Beiträge zur Psychologie der Aussage, vol. 1, p.
46, 1904.]




INHIBITION AND REËNFORCEMENT

BY LOUIS A. TURLEY


Experiments made by Ranschburg[102] on the significance of similars
in the process of learning and remembering determined that when
duplicates occur within a series of stimuli, one either totally or
very greatly inhibits the perception of the other according as they
are contiguous or are separated by other stimuli. Dr. Yerkes,[103]
in testing the effect of auditory on visual and tactual stimuli in
frogs, found that if the auditory stimulus preceded another stimulus
by various time-intervals, it had an alternating reënforcing and
inhibitory effect. A similar result was obtained by Hofbauer[104] in
a similar experiment on human subjects. The question now arises,--if
the time-interval were increased between a stimulus and its duplicate
in a series would the inhibitory effect gradually approach zero where
all effect of the preceding stimulus ceased, to which Ranschburg's
experiments point, or would the inhibitory effect be alternated with
one of reënforcement as the experiments of Dr. Yerkes and Hofbauer
would indicate? This problem--the effect of a stimulus on its duplicate
in a succeeding series of stimuli--is the problem I undertook to solve.
For this purpose, it was necessary to introduce exactly determinable
time-intervals between the stimulus and its duplicate. Therefore I
used--as Miss Kleinknecht[105] did for other purposes--a stroboscopic
arrangement instead of simultaneous presentation which Ranschburg used.

My apparatus was Professor Münsterberg's Stereoscope without Prisms or
Lenses, a description and photograph of which was published in the
article by that title in Psychological Review, vol. 1; or rather, I
used Professor Münsterberg's attachment to Kohl's centrifugal machine,
since my apparatus was not identical, except in principle, with the
"Stereoscope." The "attachment" consists of two black discs about
thirty inches in diameter, mounted about eight inches apart on the
disc-shaft of the centrifugal machine. The back disc is of wood. The
outer three inches of its face is furnished with thirty-six equidistant
strips of black tin, one end of each of which is bent so as to grip a
groove in the rim of the disc, and the other end of each is gripped
by tiny thumb-screws so that the strips lie along radii of the face
of the disc. The front disc, slightly smaller than the back disc, is
of pasteboard. Between the two discs a stationary black screen with
a short narrow slit was placed so that the slit revealed only the
strip on the horizontal radius of the back disc. Behind this screen an
eight-candle-power electric light was placed to illuminate the back
disc,--as the experiment was carried on in a darkened room. By moving
this light I was enabled to vary the intensity of illumination to
offset the skill of the observer.

For my purpose, a small white figure--one of the ten characters of
the Arabic notation--was stuck on about the middle of each of the tin
strips on the back disc; and radial slits, one millimetre wide and an
inch long, were cut from one sixth of the circumference of the front
disc so as to come opposite six of the strips on the back disc. Similar
radial slits were cut at various intervals from the remaining five
sixths of the circumference of the front disc. These were covered by
small pieces of cardboard fastened to niagara clips, thus making them
readily removeable. By this means any desired figure could be exposed
in the same revolution with the series exposed by the six slits above
mentioned.

The thirty-six strips were divided into six series of six each,
indicated by chalk-marks on the disc. Each of the series was often
changed in whole or in part by shifting and interchanging the strips.

The figure on which the effect of a preceding stimulus was tested
occupied the fourth place in the series, since this is the place where
the greatest number of errors occur, as is shown by the experiments
of Ranschburg and previous investigators in the Harvard Laboratory.
In my experiment, 4, 5, 6, 7, 8, and 9 occupied the fourth place
in the 1st, 2d, 3d, 4th, 5th, and 6th series respectively, and the
effect of a preceding stimulus was tried on each of these figures
for each time-interval. The preceding stimulus in each case was a
duplicate of the fourth member of a series, and was a member of some
other series. Thus the fourth member of each series was at all times
fixed and constant while the preceding stimulus occupied successive
progressive positions round the disc. The other members of each series
were chosen at random, care being taken that the fourth figure was not
duplicated within its series, since it would then have taken part in
inhibition within the series.

[Illustration: PLATE IX.]

By adjusting the front disc, I exposed any one of the series desired,
and by removing the cardboard blind from one of the suggestion slits, I
gave a stimulus at the desired time-interval in advance of the fourth
member of the series. The first interval I used was 1.11 sec. as Miss
Kleinknecht had tried intervals up to 1 sec. My second interval was
1.39 sec., the third 1.8 sec., and then every .277 sec. up to 4.3 sec.
In performing the experiment I exposed alternately a series without
and a series with a preceding stimulus--taking from the observer three
reports of each--until the six series had been seen. I then repeated
this, exposing with a preceding stimulus those series that had been
exposed without preceding stimulus, and without preceding stimulus
those series that had been exposed with a preceding stimulus in the
first instance. In this way I equalized and minimized the effects of
novelty and memory.

At 1.11 sec. there was considerable inhibition in five out of six
cases. In the sixth case there was slight reënforcement at this
interval. With an interval of 1.39 sec., with one exception,--not the
exception above mentioned,--there was a stronger inhibition than at
1.11 sec. Inhibition in all cases began to decrease from 1.39 sec.
until it ceased at about 1.8 sec. The preceding stimulus then had a
reënforcing effect which reached a maximum in four cases at 2.08 sec.,
one at 2.36 sec., and one at 2.64 sec. Then, in all cases, there was
a decrease of the reënforcing effect which in three cases amounted to
inhibition. In the other three cases, the preceding stimulus had no
inhibitory effect for an interval greater than 1.8 sec. For one of
these, Fig. 5, the preceding stimulus had a reënforcing effect for
all the intervals beyond 1.8 sec. The second trough in the wave or
interval of maximum inhibition was at either 2.64 sec. or 2.92 sec.,
except for the person for whom there was constant reënforcement beyond
1.8 sec., in which case the first interval of least reënforcement or
second trough was at 3.19 sec. This was the second interval of greatest
enhancement, or second crest, for four of the others. Then followed a
third point of no effect or inhibition, which was 3.75 sec. or 4.03
sec. For the person for whom the preceding stimulus had least enhancing
effect at 3.19 sec., the second interval of greatest reënforcement
coincided with the interval of greatest inhibition for the majority
of the other observers. For four of the six observers, the third
interval of greatest reënforcement was 4.3 sec. In this, the observer
agreed for whom the last interval of greatest reënforcement was 3.75
sec. Thus while, for this observer, the first two points of greatest
reënforcement were separated by an interval of 1.11 sec., the second
and third points were separated by an interval of only .55 sec. This
same thing occurred in the records of two other observers, for one at
this point, and for the other at another point. Of the two dissenters
from the opinion of the majority that the third crest was at 4.3 sec.,
one was an erratic observer; and for the other, there was a slight
reënforcement at 4.03 sec. and no effect at all at 4.3 sec.

Fig. 1 represents the average of the records of the six observers.
The curve is based on the difference between the number of times the
fourth members of the series were seen with and without preceding
stimulus. The base-line represents the number of times the figure was
seen without preceding stimulus, taken each day as the normal for that
day. Figures above the base-line represent the greater, and those below
the line, the less number of times the figure was seen with preceding
stimulus, or reënforcement and inhibition, respectively. The first two
points are the average of fifty-four observations; each point beyond
the second is the average of 108 observations. Figs. 2, 3, 4, and 5
represent individual records constructed as Fig. 1, each point being
the average of eighteen observations.

The curve in Fig. 1 is somewhat misleading in showing points of maximum
reënforcement at 3.19 sec., 3.75 sec., and 4.3 sec. In no individual
case was this true. The reason for the crest at 3.75 sec., or at least
for its height, is that in two cases reënforcement was considerable
at this interval, and there was little inhibition to offset this in
the general average. At 3.19 sec., which was the second interval of
greatest reënforcement, for four out of the six observers, owing to
practice, the reënforcement was not great (Fig. 4), but in no case was
there inhibition at this point. Thus for the lack of strong positive
effect at 3.19 sec. and the lack of strong negative effect at 3.75
sec., the two crests are the same height, while the first represents
the maximum effect for four and the second for two observers.

From these results, taking everything into consideration, my
conclusions are:

(1) If a stimulus precedes at various time-intervals its duplicate
in a series of stimuli, it will alternately inhibit and reënforce the
perceiving of the duplicate stimulus.

(2) Within 4.5 sec. there are at least three points each of maximum
inhibition and maximum reënforcement.

(3) The points of maximum inhibition and likewise those of maximum
reënforcement are separated by intervals of from .55 sec. to 1.2
sec.--more often by one of the two extremes than by any mean.

(4) Up to 4.5 sec., as the time-interval increases, the maximum
inhibition generally decreases, while the maximum enhancement
correspondingly increases.

What the limit of this periodic effect is, I cannot as yet say, as up
to the present I have not used time-intervals beyond 4.3 sec. But from
the intensity of the effect at this interval, I do not expect the limit
to be within several seconds.

FOOTNOTES:

[Footnote 102: Ranschburg: Ueber die Bedeutung der Ähnlichkeit beim
Erlernen, Behalten und bei der Reproduction, Journal der Psychologie
und Neurologie, Bd. 5, p. 94.]

[Footnote 103: Dr. R. M. Yerkes: The Sense of Hearing in Frogs, Journal
of Comparative Neurology and Psychology, vol. 15, no. 4, 1905; also
this vol. Harvard Psychological Studies.]

[Footnote 104: L. Hofbauer: Interferenz zwischen verschiedenen Impulsen
im Central-nervensystem, Pflügers Archives, Bd. 68, p. 564, 1897.]

[Footnote 105: Kleinknecht: This volume.]




THE INTERFERENCE OF OPTICAL STIMULI

BY H. KLEINKNECHT


The purpose of this investigation is the determination of the location,
extent, nature, and cause of the interference of optical stimuli.
Ranschburg[106] studied the phenomena carefully in using optical
stimuli which were spread over the retinal field, for instance, a
series of letters or figures one beside the other. But if we are
to experiment on the inhibitory influence of a certain qualitative
impression, we must try to eliminate the local difference; the letters
or figures ought to be seen at the same spot.

This became possible by a stroboscopic arrangement, consisting of two
parallel circular discs one foot apart on the same axis, whose motion
was controlled by an electric current.

The discs were 60 cm. in diameter. Thirty-six radii were drawn
equidistant on the farther disc, and on these were clasped black tin
strips bearing letters or numbers or colors. The nearer disc was
similarly divided and an opening, 3 mm. in width, was cut at each
radius. This exposed the number. A cardboard placed between the discs
limited the range of vision, its opening being 4 × 5 cm.

The figures were 10 mm. high, white, and placed on a dark background.

Preparatory stimuli were given to enable the subject to adjust his
eye to the farther disc. They were so placed as to fall on different
retinal points, thus avoiding fatigue.

Many of the tests employed by Ranschburg were used again to ascertain
the influence of the change in method and with the hope that such
differences might throw some light on the nature of the interference.
At first there were six subjects, afterwards eight--all graduate
students and trained in laboratory work. The experiment was carried
on in the morning. Numbers consisting of six digits were exposed on a
dark background. The time of exposure varied with the subject, but was
constant throughout the experiment. The subject was asked to record
the number immediately after perceiving it, but in almost every case
it was read verbally (its retention being thus facilitated) and then
recorded.

For the first few weeks letters were used. But since subjects found
it very difficult to distinguish these, a change was made to figures.
For a month and a half numbers were given for the purpose of training
the subjects and of ascertaining the speed best adapted to each. This
varied from 5-1/2" to 8" a revolution, each figure being exposed from
115 to 166 sigma.

Three series of numbers were given: (1) Homogeneous, containing a
repeated figure, as, 495851. (2) Heterogeneous; as, 708654. (3)
Similar, that is, in construction; as, 813470 (8 and 3 being easily
substituted for each other). Other similars given by Ranschburg are 9
and 0, 9 and 6, 9 and 2, and 5 and 3.

In order to determine the place of greatest interference, the repeated
figures were located in all possible positions, while the preceding
and succeeding figures were left unaltered, so as to obviate any
new influences which might result from a change of relations. There
are fifteen possible variations of the series: _mabcdm_, _ambcdm_,
_abmcdm_, _abcmdm_, _abcdmm_, etc.

The following table, illustrative of the scheme _ambmcd_, will show the
character of the results obtained. Only the numbers in which errors
occur are here recorded, those figures which were incorrectly perceived
being printed in heavy type. The dash is used when the location of the
figure omitted is known, and the interrogation mark when the reply is
doubtful.

               8"       8"     5-1/2"   5-1/2"    5-1/2"     8"
               V.       R.       S.       M.       H.        E.
    708025   70625    76082    .....    70825    7082-5     70285
    958564   95584    .....    .....    95864    985 ? 4   958-54
    281845  281485    .....             20861     28185    281-54
    436392   43632    43636   436932   436924     43632     43632
    526273  526723     5257   572673    52763     .....     52623
    940469   94069   940465    .....    94069     94640    940-69

The interference may result in permutation, substitution, or
inhibition. The latter two may take several forms; as, inhibition of
identicals, of similars, of dissimilars, the location of the omitted
figure being known or unknown; also, substitution of an identical,
similar, or dissimilar figure which precedes or follows.

The homogeneous series (540 tests) gives results as follows:


HOMOGENEOUS SERIES

                     Inhibition of    Inhibition of     Inhibition
                      Identicals.       Similars.     due to Location.
       Location
        of the        Spot   Spot      Spot   Spot      Spot   Spot
      Identicals.    Known  Unknown   Known  Unknown   Known  Unknown

       _mabcdm_
     1   1=6                   1        1       4                5
       _ambcdm_
     2   2=6                   5                2         1      3
     3   3=6                   2                3                5
     4   4=6                   3        3                 4      2
     5   5=6           6      15(?)             4                5
     6   1=5           1       2        1       2         1      3
     7   2=5                   5                3                3
     8   3=5           3       7                          2      1
     9   4=5           2      21(?)             2         5      3
    10   1=4                   9                4                3
    11   2=4           3       9                4         1      7
    12   3=4           3      19(?)             1         2      4
    13   1=3              [107]5+3(?)            2         1     10
    14   2=3           1      11(?)             3                8
    15   1=2                   6(?)     1       4                9

    Total             19    48+75(?)    6      38        17     71


I. _Inhibition_

(1) There is considerable inhibition only when identicals are next to
each other.

(2) There is but little difference in the amount of inhibition when
identicals are removed two and when removed three places.

(3) The interference is greatest when 3d = 4th, 4th = 5th, and 5th =
6th figures, in which schemes it is almost equal in amount.

(4) When identicals are adjacent, it is impossible to decide whether
there be inhibition or fusion, _i. e._, whether one be inhibited and
the other appear, or whether the figure seen be a fusion of the two
(unless there is an omitted figure whose location is known to the
subject). Its intensity does not serve as a clue, for the perception of
the number demands the full concentration of the attention.


II. _Substitution_

When the interference is not sufficiently great to cause inhibition,
substitution may result.

(1) In the majority of cases the substituted figure is a dissimilar not
occurring in the number.

(2) A preceding figure is frequently substituted.

(3) Occasionally a figure is replaced by its similar, but this is not
true of the homogeneous element. (Cf. with Ranschburg.)

(4) Sometimes the next figure in the natural number series is
substituted; as, 9 for 8, 6 for 5.

(5) The figures containing straight lines (4, 7, and especially 1) are
less subject to illusion; likewise the smaller numbers (1, 2, 3, 4).


III. _Permutation_

The permutation represents the least interference.

(1) The 4th and 5th figures are most often exchanged.

(2) The figure is seldom permuted more than two places, and generally
but one.

The recording of the number was most interesting. Generally the
first few figures and the last were written without comment, but the
4th and 5th often called forth an expression of doubt, which was
immediately followed by an exclamation at the coming of the figure into
consciousness as if by "inspiration." The experience was extremely
peculiar. The figure, fully as distinct as those already perceived,
was always from 5″ to 10″ late, and seemed to "pop in unannounced"--to
"come from nowhere." A substitution or permutation occurred without
this lapse of time.


HETEROGENEOUS SERIES

(1) There are less than half as many inhibitions as in the homogeneous
series, the largest number being in the 4th and 5th places.

(2) The number of substitutions is decreased by a fourth, the
identicals and similars remaining the same.

(3) There are no fusions.

(4) Fewer permutations are found in this series. The 4th and 5th
figures are most often permuted. In a very few cases the figure is
permuted four and five places.

(5) There are an equal number of doubtful perceptions in both series.


SIMILAR SERIES

(1) There are few cases of inhibition, and even more surprising is the
small number of cases in which a figure is inhibited by its similar.

(2) There are more substitutions, 6 being very often substituted for
5, generally in the 6th place and when preceded by 0 or 9, often by
both. Similars are never replaced by identicals (69 by 66 or 99) as
Ranschburg found in his experiments.

(3) The fusion of similars equals that of identicals in the homogeneous
series.

(4) The number of permutations is the same as in the homogeneous series
and less than in the heterogeneous.

(5) The doubtful perceptions have decreased by half.

That there are fewer errors in this series than in the homogeneous
or heterogeneous, may be due to the fact that it was given last,
especially since one subject showed marked improvement in the entire
series and another during the last half. These subjects suddenly began
to see six figures, while previously they had seen but five and those
contained errors.

In the above 1620 tests, 9 and 0, and 8 and 3, are sometimes inhibited
by and substituted for each other, but the remaining similars mentioned
by Ranschburg seldom have any such effect.

It is impossible to determine definitely the nature of the
interference, the greatest uncertainty existing in the homogeneous
series when two identicals are adjacent. But the interference is
dependent not only upon the identity or similarity of the figures of
which the number is composed but also upon their location.


INHIBITIONS

                      1         2      3       4       5       6      Total
             Place  Place
             known unknown  K. U.   K. U.   K. U.   K. U.   K. U.    K. U.

  Homogeneous               1  13   2 28   9  41   15 31   15 44   42 157
                               6(?)   14(?)   19(?)   21(?)   15(?)   75(?)

  Heterogeneous 2       5   1   8   2 15   4  36    5 34    2 25  16 123

  Similar                       4      4       5    3  6    2 13   5  32

  Total
   excluding(?) 2       5    2  25   4  47  13  82   23 71   19 82  63 312

  Total of Known
    + Unknown       7         27      51      95      94      101    375

          (?) Inhibition or fusion.


SUBSTITUTIONS

                         1     2     3     4     5     6     Total

    Homogeneous          8    12    26    27    38    14      125
    Heterogeneous        4     4    14    21    30    20       93
    Similar              1     5     9    25    33    27      100

        Total           13    21    49    73   101    61      318


FUSIONS [See (?) under Inhibitions]

                         1     2     3     4     5     6     Total

    Homogeneous                1     2     1     3    10       17
    Heterogeneous                    0           0     0        0
    Similar                          3     3     6     6       18

        Total                  1     5     4     9    16       35

Note. There were no clear cases of fusion, but the evidence favored
fusion rather than inhibition.


PERMUTATIONS

                         1     2     3     4     5     6     Total

    Homogeneous    (a)   6    29    46    56    30            167
                   (b)         5    21    45    68    35      174
    Heterogeneous  (a)  15    25    60    62    28            190
                   (b)        14    23    51    82    44      214
    Similar        (a)  13    20    37    78    16            164
                   (b)        12    17    26    75    28      158

    Total          (a)  34    74   143   196    74            521
                   (b)        31    61   122   225   107      546

        (a) forward, (b) backward

Note. The permutation of an inhibited figure was not noted unless its
location was known: hence the difference in the number of forward and
backward permutations.

                         1    2     3     4     5     6    Total

    Total Interferences 54   160   323   509   524   300    1870
           %             3%    9%   17%   27%   28%   16%
    Absolute Errors
    (excluding          20    55   119   191   225   103     713
    Permutations)        3%    8%   17%   27%   31%   14%


ABSOLUTE ERRORS (excluding Permutations)

                    Homogeneous     Heterogeneous     Similar
    Inhibitions          199            139              37
    Substitutions        129             93             101
    Fusions               17                             18
        (?)               75

        Total            420            232             156
                          52%            29%             19%

Over 50% of the errors were found in the 4th and 5th places.

[Ranschburg: 90% of errors in right half--60% in 5th place, 30% in 4th,
few in 6th.]

In 1620 tests, the homogeneous series contained 52% of the absolute
errors, the heterogeneous 29%, and the similar 19%.


COLORS

In the hope that some light might be thrown upon the main question at
issue, the writer changed the stimuli, using colors instead of numbers.

It was important that the colors should be of the same or only slightly
varying intensity and that they should be easily distinguishable. In a
series of preliminary experiments in which red, blue, yellow, green,
brown, gray, pink, and violet were used, red was lost in 8% of the
tests, and gray in 25%.

Colors 1×4 cm. in size "ran into each other," while those which were
1×1 cm. remained distinct.

Here it was found necessary to distinguish between the various
factors which might cause inhibition. Three factors entered into each
test--perceiving, naming, remembering.

Four subjects found difficulty in naming, especially at first. The
various methods of naming are given below in detail. M. says: "The name
of the color is localized in my mouth. Generally there is no movement
of the tongue--an impulse only; and the name is felt in that part of
the mouth where the sound would be reflected, as, red in the upper
part, blue near the front, etc."

S.: "Usually there is no apparent tendency to pronounce. Occasionally,
naming them over inaudibly before recording is found advantageous."

E., V., and H.: "The naming is mental, but is accompanied by a slight
movement of the tongue and throat."

684 heterogeneous and 200 homogeneous tests showed that greatest
inhibition occurred in the following order: 4th place (27%), 3d
(26%), 5th (24%), 2d (11%), 6th (8%), 1st (4%). There was but little
difference in the 3d, 4th, and 5th places.

During first tests subjects were allowed only one exposure, but later
it was thought best to eliminate all omissions resulting from inability
to name colors perceived, and hence they were asked to record only when
able to name all colors perceived during that exposure. However several
required but one exposure.

Preliminary drill was given for two weeks. Since no clear cases of
fusion had been obtained in the entire number-series, the one aim
of the experimenter was to ascertain whether fusion of colors, even
though of heterogeneous, be possible. Eight hundred heterogeneous
tests gave 927 cases of inhibition, 7 of fusion, and 18 which, though
somewhat doubtful, yet gave more evidence of fusion than of inhibition.
Yellow (3d place) and brown (6th place) were seen as yellowish-brown,
brown and pink as pinkish-brown, etc. Gray was seen several times
instead of a color and its complementary when these were in immediate
succession. This was true of both red and blue. Half of the total
number of substitutions was due to the displacement of yellow by brown.
And a color not in the series was as likely to be substituted as one
preceding or following the displaced color.

Two hundred and fifty-two homogeneous tests showed that there is
greatest interference when identicals are in immediate succession, and
least, when removed two places. The doubtful (fusion?) cases number
_one third_ of the inhibited. The 4th and 5th colors are permuted most
often, as was found to be the case in the heterogeneous series also.
The element is generally permuted but one place.

The heterogeneous color-tests show three times as much interference as
the corresponding number-tests, and the homogeneous twice as much. The
discrepancy in the amount of variation may be due to the experiments
with the heterogeneous colors being earlier, when naturally more errors
would be made.

However, a comparison of 252 homogeneous with the same number of
heterogeneous tests, taken at the same time, shows that there is
a much larger difference in the number of absolute errors between
the heterogeneous and the homogeneous number-series than there is,
proportionately, between the two series of color-tests.

Lest the want of correspondence in the results might have been due to
the comparatively small number of immediately successive identicals in
the color-tests, 90 homogeneous tests, equally distributed among all
possible variations in the location of the identical elements, were
compared with 90 heterogeneous, and it was unexpectedly found that the
absolute errors as well as the permutations were almost equal in the
two series. Nevertheless, the validity of a conclusion based on so few
tests may well be questioned.

Ranschburg found that simultaneous homogeneous stimuli interfere with
one another; while simultaneous heterogeneous stimuli clear the way for
one another. On the basis of the experiments with numbers, the writer
would amend the conclusion reached in the earlier research to read
thus: Homogeneous optical stimuli, whether occurring simultaneously in
different positions, or in immediate succession in the same positions,
interfere with one another; while heterogeneous stimuli clear the way
for one another.

FOOTNOTES:

[Footnote 106: Ranschburg: Zeitschrift für Psychologie, vol. 30, 1902.]

[Footnote 107: Fusion or inhibition?]




SUBJECTIVE AND OBJECTIVE SIMULTANEITY

BY THOMAS H. HAINES


This investigation finds its starting-points in two widely separated
lines of experimentation in the problems of attention. These two lines
are the "scope-of-attention" experiment with the tachistoscope, and the
"time-displacement" experiment with the pendulum apparatus. It seems
to me these two can be brought into relation to each other to the help
of each of them individually, and that an investigation taking these
wide relations within its scope may reasonably be expected to throw
new light upon the manner in which mental processes are related to
each other when they are together in consciousness at the same time.
The first of these experiments (tachistoscopic) is concerned with the
number and relative clearness of the processes which go on at the same
time. The second (displacement) is concerned with the conditions of
the subjective displacement of one of two objectively simultaneous
stimuli with reference to the other. Its problem is the essential
psychological problem involved in the astronomer's error in transit
observations by the eye-and-ear method, for the personal equation
arising in these observations is more a matter of the reciprocal
relations among the processes which are together in consciousness at
the moment of observation than it is of mere reaction time. It is
primarily more a matter of relative clearness, as controlled probably
through interference of one with another, than it is of the more or
less temperamental facility of converting ideas into action.

The psychological question at the heart of the observation-error,
called the personal equation, is this,--What are the conditions which
hinder such a division of attention among the parts of the complex
operation of coördinating sense-stimulations, that the processes
which start simultaneously may proceed to equal clearness at the
same time, and so be perceived as simultaneous? The facts sought in
order to answer this question are the very same as some of those,
at least, demanded by the "scope-of-attention" investigation when
it really opens up to its true problem. W. Wirth[108] has recently
shown, in an exhaustive criticism of the tachistoscopic method, that
"scope of attention" is primarily concerned with the _relations_ of
the processes present together, and that this demands a previous
exhaustive study of their _relative clearnesses_. Earlier studies by
the tachistoscopic method, as, for example those of Cattell[109] on
the relatively short time for the perception of letters in words,
as compared with that for separate letters, and the overlapping of
processes in continuous reading, showed that the important question is,
_what_ are the processes which may go on at the same time. Leaving out
a statement of the nature of the processes is equivalent to leaving
out one of the dimensions when endeavoring to state the contents of
a solid. The scope of attention can be defined adequately only when
one knows fully _what_ the separate processes are as well as _how
many_ there are. This analysis, which the scope-of-attention problem
demands, cannot fail to be directly fruitful for the solution of the
time-displacement problem. The analysis of this larger problem directly
involves the former. Any attempt to investigate the time-displacement
of sense-impressions from simultaneous stimuli must inevitably place
the highest value upon the whole detailed analysis of any moment of
attentive effort.

The present investigation, starting with the facts of
time-displacement, and taking the hint offered by Gonnessiat,[110]
attempts to show, by a more complete analysis, the effects of the
various relations within each series,--the visual within which the
sounds are to be placed, and the auditory series itself, and also
relations existing between the two series. In other words, the
attempt is made to strip the "displacement" experiment until nothing
more remains to be coördinated than a single pair of simultaneous
stimuli. This was the experiment of Exner.[111] He investigated
the shortest discriminable interval marked off by various pairs of
stimuli, addressed to the same sense and to different senses. From this
coördination of a visual and an auditory stimulus, where the limits
of the "specious present" are obtained, I make a turn into the realm
of the scope of attention. By a new method, whereby impairment of
accuracy of processes is made the test as to whether the processes have
proceeded together, it is shown that two such perceptual processes can
go on just about as well at the same time as separately. Since this
test is subject to the objection that the visual and auditory processes
may really be successive, though seemingly at the same time, owing to
retinal inertia, the same question is removed to an entirely different
plane in a further and more detailed set of experiments where the
processes combined are _judgments of comparison_ based upon one and the
same visual sensation.


EXPERIMENTS IN TIME-DISPLACEMENT

The Leipsic Complication Experiment with the pendulum apparatus (for
description of this see Wundt's Physiol. Psy., 5th ed., vol. 3, p. 82)
was an early adaptation of the astronomers' eye-and-ear method to the
purposes of psychological experimentation. Instead of localizing a
visual stimulus (star on meridian) in an auditory series (clicks of a
chronoscope) as in the eye-and-ear method, this adaptation localized
an auditory stimulus (bell-stroke) in a visual series (successive
positions of a pointer on a graduated circle). This pointer passed
around to the right and to the left from the position of rest, in
which it pointed vertically upward, as the pendulum, to which it was
connected by clockwork, swung back and forth. By a simple adjustment
the bell-stroke could be made to come at any point in the complete
double swing of the pendulum, and so anywhere in the arc over which
the pointer moved. This machine makes an additional problem as to the
effects upon displacement of the increasing and decreasing speed. My
aim being to simplify as much as possible the displacement-error and
so reduce it to its elements, this feature was not only not of direct
interest, but it was very desirable to dispense with it altogether.
This was done by arranging the visual series so that the members were
shown in perfectly regular order, _i. e._ with equal time-intervals,
throughout the series. These equal intervals were secured by the
rotation of a disc at a uniform rate.

My method also gave a more distinctly serial character to the visual
stimuli, in that they were separated by blank periods. The series
consisted of letters in alphabetical order. Denison's smallest white
letters, about six millimetres in height, were pasted upon a disc of
black cardboard, near the circumference and perpendicular to radii, so
that they would appear in succession and right side up, to an observer
looking through a slit at the peripheral region of the disc, as it
rotated. The letters were placed in three concentric rows, so that as
the disc rotated they appeared in three different places. The disc was
56.5 cm. in diameter. As a further aid in securing separate exhibitions
of letters, another black disc of the same size as the one bearing
the letters, with radial slits 2 mm. wide and cut in from the edge 4
cm., opposite each letter on the other disc, was mounted on the same
shaft, six inches from the first, and between it and the observer. A
short observation-tube was placed at the same height as the axis of the
discs parallel to this axis, and opposite the slits when they were at
this elevation. Looking through this, as the discs were rotated, one
would see the letters right side up and in serial succession. Uniform
illumination was secured by working in a dark room with artificial
light. An electric lamp was hung between the discs. Uniform motion was
secured by an automatic control gravity motor, connected by belt with a
pulley on the disc-shaft.

The auditory stimulus, a click, adjustable to any part of the series,
was made as follows: A wooden shaft, mounted on the same axle as the
discs, and beyond the discs from the observer, could be rotated freely
around the axle when the nut securing it was loosened. This shaft
extended beyond the edge of the disc. It carried a copper wire which
was in contact with the axle. A mercury cup was placed on the table,
upon which the machine rested, in such position that the copper tip
passed through the mercury when the discs rotated. It was thus a very
simple matter to connect an electric sounder so that it would click
every time the circuit was made by the copper passing through the
mercury. And, by the adjustment of the wooden shaft, the click was
readily placed anywhere in the visual series.

As already suggested above, the length of interval between members
of the visual series, and also the time between clicks, seem to be
important factors in determining the amount, and perhaps also the
direction of the displacement. Bessel found his personal equation was
considerably diminished when he used a clock marking half-seconds
instead of one marking seconds. Wolf also diminished his error by using
a clock beating one hundred times a minute instead of one beating
seconds, which he was accustomed to use. Wundt found his customary
negative displacement on the pendulum apparatus (coördinating the
sound with a position of the index earlier than that with which it
was actually simultaneous) disappeared when he had members of the
visual series one thirty-sixth second apart and the auditory stimuli
one second apart. It seemed important at the outset, therefore, to
determine, if possible, the effects of each of these factors.


BOTH INTERVALS PROGRESSIVELY VARIED

In each experiment the observer was allowed to observe as many
complications (coincidences of click and letter) as he desired, in
order to assure himself of his judgment. The experimenter counted and
recorded the number observed in each experiment. Experiments were made
in series of ten. Six different combinations of intervals were used in
this first group of experiments. The auditory intervals (time between
successive clicks) and visual intervals (time between successive
members of the visual series) are given at the tops of the columns in
Table I. This table is a summary presentation of the results of this
group. There were three observers. During each hour of experimentation
with a given observer, at least one series with each of the first four
time-interval combinations was tried out. "Aver. num. Trials" means
the average number of complications observed in the whole number of
tests averaged. "Num. Series av." means the number of series of ten
experiments each averaged to give the displacement results below.
"Aver. Error" is the average of all the displacements of the auditory
impression, _irrespective of the direction of the displacement_. "Mean
Displacement" is the _actual mean displacement_ as obtained by dividing
the algebraic sum of all displacements, positive and negative, by the
number of experiments. The plus sign indicates a positive displacement,
and the minus sign, a negative. Negative and positive are here used in
the sense customary in similar experiments,--namely, the click, being
heard as simultaneous with a visual impression which actually came
before it, was said to be displaced negatively, and the click, being
heard as simultaneous with a visual impression coming in fact later
than it did, was said to be displaced positively. Average errors and
mean displacements are given in the table in thousandths of seconds.
Observers were asked to locate the click in the visual series in terms
of one tenth the distance or time between the letters.


TABLE I

     Aud. Interval (sec.)   1.28     2.56    4.04    8.40    1.28    2.02
     Vis. Interval (sec.)    .040     .080    .120    .260    .080    .120

  Obs.
  B  Av. num. Trials       13.9      5.8     3.8     2.1     9.8    13.9
     Num. Series av.        8       13      13       8       2       2
     Aver. Error (sec.)      .056     .064    .077    .164´   .045    .067
     Mean Displac'mt (sec.) +.045    -.040   -.067   -.152   +.045   +.067

  Bo Av. num. Trials        9.4      4.1     3.0     2.0     5.5     3.5
     Num Series av.         6        10      11       8       3       2
     Aver. Error. (sec.)     .114     .060    .054    .049    .05     .082
     Mean Displac'mt (sec.) +.114    +.045   +.033    .000   +.045   +.082

  M  Av. num. Trials        6.3      3.1     2.5     2.2     5.3     4.4
     Num. Series av.        9       12      12      10       3       2
     Aver. Error (sec.)      .09      .07     .076    .110    .067    .172
     Mean Displac'mt (sec.) +.089    -.058´  -.058   -.104   +.062   +.168

The first four combinations of intervals above, with which the major
part of the results was obtained, it will be noticed, are approximately
proportionate increases in each interval, column by column. These
conditions were planned with a view to revealing the conditions, most
favorable for coördinating the auditory and visual impressions, for
each observer, so that his displacement would disappear, or show a
tendency to disappear. So far as is shown by these results, there
are here two types of observer. Bo has no mean displacement for the
8.40-.260 sec. combination, and it steadily decreases toward this
point as the two intervals increase. Both B and M, on the other hand,
have a considerable positive mean displacement for the 1.28-.040 sec.
combination, and a considerable negative mean displacement for the
2.56-.080 sec. combination, and there is a further increase in the
negative displacement as the intervals increase from this point. It
seems as though these observers would give a mean displacement of zero
for some combination of intervals between these first two. It will be
noticed that the average number of trials is exceptionally large for
all three of the observers in the first combination. This seemed to
be pretty clearly due to the very short interval separating visual
impressions.


THE AUDITORY INTERVAL _alone_ VARYING

In order more certainly to isolate the influence of the time-interval
between successive auditory impressions, another series of experiments
was performed, in which this interval between clicks, alone, was
varied from series to series. The visual interval was kept at .083
sec. throughout. This seemed to be about the shortest time-separation
at which the successive impressions were perfectly distinct. The
auditory impressions were at 1, 1-1/2, 2, 3, and 4 sec. intervals.
The additional observer, H, was myself. I obtained these results by
experimenting alone. I adjusted the wooden shaft carelessly to a new
position and started the machine. When speed was attained, I would
make the observation just as an observer for whom the adjustment had
been made. I would have as little idea beforehand as he with regard to
the position of the click in the series of letters. Having made the
observation, however, I measured the actual place of the sound and
recorded it, as well as my judgment. In this way, of course, I had some
idea, all the time, as to what kind of displacements I was making and
how large. I was as careless of this knowledge as possible, and the
records were laid aside absolutely, until I was through with the whole
experiment. Terms used in Table II are the same as in Table I.


TABLE II

       Aud. Interval (sec.)        1      1-1/2       2        3        4
       Vis. Interval (sec.)       .083     .083     .083     .083     .083

    Obs.
    B  Av. num. Trials           8.5      6.8      6.2      4.8      4.8
       Num. Series av.            10       10       10       10       10
       Aver. Error (sec.)         .097     .108     .106     .097     .101
       Mean Displacement (sec.)  +.097    +.108    +.106    +.097    +.101

    Bo Av. num. Trials           6.0      5.0      4.2      3.2      3.1
       Num. Series av.            10       10       10       10       10
       Aver. Error (sec.)         .103     .080     .081     .092     .082
       Mean Displacement (sec.)  +.102    +.073    +.078    +.089    +.075

    M  Av. num. Trials           4.4      3.8      3.4      3.0      2.8
       Num. Series av.            10       10       10       10       10
       Aver. Error (sec.)         .088     .084     .081     .068     .052
       Mean Displacement (sec.)  +.086    +.079    +.072    +.051    +.048

    H  Av. num. Trials
       Num. Series av.            10       10       10       10       10
       Aver. Error (sec.)         .043     .036     .047     .040     .037
       Mean Displacement (sec.)  -.022    -.012    -.027    -.017    -.013

One series of ten of each of these combinations was given during each
hour of experimentation with each observer. These were also given
in a different order each day, so that no combination should have
the advantage, by practice or lack of fatigue, in the average of the
ten series. Here again it was evident, in the records of each of the
observers for whom the count was made, that the largest number of
trials was necessary in the 1-.083 sec. combination. It thus appears
that it was not the short visual interval, .040, in Table I, that was
responsible for the large number of trials necessary in the first
combination. Here, where there is the same visual interval of .083
sec. throughout, it must be the short auditory interval which makes
particularly difficult conditions for attention. This agreement between
the results in both groups of experiments seems to indicate unfavorable
conditions for accurate coördination at auditory intervals as short
as one second. The large changes in the mean displacement for B and M
between the first two combinations in the first group (Table I) was
kept especially in mind in planning this second series of combined
intervals. It was presumed from the results given by these observers
in Table I that they would each, with the range of auditory interval
presented them in these experiments, show a point of no displacement,
or a very slight one, and an increasing displacement on each side of
this point. They both seemed to indicate a time-interval favorable
for the "ripening of apperception" as Wundt and Von Tschisch call
it, and I planned these experiments especially to bring it out more
clearly. But there is far less indication of a time most favorable for
"ripening" than in the previous group of experiments. B and M both give
all mean displacements as positive, and decidedly small differences in
displacement for the various combinations. Results of Bo are, however,
entirely consistent with those of Table I. H gives a very small
negative mean displacement throughout. This, as well as the smallness
of the average error, may be due to the knowledge of results which I
had.

An examination of the detailed daily results, which cannot be exhibited
here, shows considerable change in the direction of the displacements
as the work proceeded. This is especially marked in the case of B,
who, during the first two hours of experimentation, gave only negative
displacements. Through the rest of the first group there was a gradual
increase of positive displacements, and in the last two hours about 90%
were positive. In the second group he did not give a single negative
displacement. The same change is manifested in the results of M for
the first group; but he did not change over nearly so completely.
In the five hundred experiments of Table II, for M, there are three
hundred and ninety-two positive, sixty-seven negative, and forty-one
_no_ displacements. Bo gave a number of positive displacements from the
start. These increased considerably in the second over the first group,
showing only thirty-seven negative displacements in the second group.
This change in the direction of the displacement, rather independently
of the intervals, is an interference with the main purpose of the
experiment. It may represent the effect of practice.

Angell and Pierce[112] found the same progressive change from negative
to positive displacements. They explained it as a change in the focus
of attention. The visual series is focal at first, and the sound
becomes focal in later experiments. Negative displacements result from
fixing the last possible point in the visual series before the sound
is heard, while positive displacements result from getting the first
letter possible after the sound. The method of my observers, with
the large numbers of trials at their disposal, was to "let the sound
_announce_ the letter" on the first trial, and then to "lie in wait for
the letter" so announced, and to "see whether it was too late or too
early." It was found to be too late usually, for this was the second
method of Angell and Pierce, which gave positive displacements.

So at the next trial the preceding letter would be waited for, and
tested in the same way. The first trial was thus auditory-visual
attention and the second was visual-auditory, and there was a striving
after a balance where neither auditory nor visual impression had the
preference.

As soon as adjustment to the conditions of a given combination had
been secured, it was a simple matter to anticipate, with a fair degree
of accuracy, both a given letter and the recurrence of the sound. The
attention could thus be pretty accurately divided between the two, and
a very small time-displacement was the result. When I was acting as
observer, a change of the auditory interval _upset_ the whole _plan_ of
procedure for a short time. I had to accustom myself to the new rhythm.
But as soon as this adjustment was made, it was just as easy to make
the judgment at one rate as at another, barring variations which might
be called fortuitous, since they were so small. This experience with
the conditions here under consideration, as well as the introspections
of the other observers, convinces me that the conception of an
apperception-ripening time has been overworked.

It is true that I find here, just as Pflaum[113] found, displacements
in both directions with every observer. It seems very doubtful to me,
however, whether these are in any sense due to what may be considered
a fixed apperception-time for a given observer, under fixed objective
conditions. The facility with which adaptation is made to the changed
conditions of a new combination of intervals, so that just as small
displacements are made under one as another, indicates to my mind that
one can control the conditions so that the apperception shall ripen
quickly or slowly, depending upon the warmth of the interest, and the
concentration or division of the attention,--that there is a capacity
in the ordinary individual so to adapt himself to the conditions as to
do equally good work in coördinating two sense-impressions anywhere
within a wide range of intervals. The influence of the length of the
interval separating succeeding clicks, in determining displacements,
has been considerably overestimated. I should state here that no
one of the three observers had any specific training to reduce the
displacement. The results were not discussed with them. They had no
means of knowing what displacements they were making. This certainly
adds strength to the inference, from these results, that there are
adaptable apperceptive conditions for coördinating sense-impressions.


THE INFLUENCE OF THE LENGTH OF THE SERIES OF VISUAL IMPRESSIONS

The next step in the analysis of the complication experiment, bringing
it into relation with the simple coördination of two disparate stimuli,
is to show, if possible, the influence of the _series_ of _visual_
impressions. This naturally divides into two lines, namely, (1) the
_length_ of the series as such, and (2) the relative influence, in
case of a given kind of displacement, of the part of the series coming
_after_ the auditory stimulus, and the part _preceding_ it. For the
first, I used in comparison, a series of twelve letters, a series
of three, and a single letter. For the second, the letter, whose
coördination with the click was set as the task of the observer, was
made successively the first, the last, and the middle member of a
series of five letters.

During each hour of experimentation, the observer was tested as to
his accuracy of localization of the click, (1) in a series of twelve
letters at intervals of .083 sec., (2) in a series of three at the
same interval, and (3) with reference to a single letter. The method
for the first two was exactly as in the preceding experiments. In the
case of the single letter, he was asked to localize as accurately as
possible in terms of the intervals as he remembered them from the
series. This introduced an element of uncertainty. One observer, St,
would not give any judgments as to time-differences in the case of the
single letter. Another method had to be adopted in order to obtain more
comparable results. These results (Table III) are presented as showing,
by comparison with the following table, the transition from one method
to the other. Clicks were at 2-sec. intervals. Each number in the
table is the average result of fifty or more experiments. They are in
thousandths of seconds, and the plus and minus signs indicate positive
and negative displacements.


TABLE III

    _Observer_   _Twelve Letters_   _Three Letters_   _One Letter._
        A           +.012 sec.         -.029 sec.       -.010 sec.
        G           -.022 sec.         +.004 sec.       -.004 sec.
        Sh          +.028 sec.         -.079 sec.       -.057 sec.
        St          -.050 sec.         -.036 sec.
        Bo          +.029 sec.         -.015 sec.       -.022 sec.

The method of right and wrong cases was used in the next group of
experiments, to secure the same conditions of making the judgment in
each of the three cases used above. Selecting a letter near the middle
of each series, I asked the observer, in each of these cases, just as
in that of the single letter, to say whether the click was before,
on, or after the letter. I worked _out_, in successive experiments by
successive adjustments, from the position of apparent simultaneity
of click and letter, in both directions, to a point where in 75% of
the cases the click seemed to come before; and also to one where it
seemed to come after, in 75% of the cases. So also I worked _in_ both
ways, by successive adjustments, from regions of clear discrimination
of time-difference and direction, to points where the time-relation
was uncertain or wrong in 75% of the trials. By averaging the just
perceptible and the just not perceptible, in each case, the thresholds
were obtained for "click first" and "click last." The time between
these thresholds I call the "range." It is really a measure of James's
"specious present" and of Stern's "Präsenzzeit." (An admirable
presentation of similar results by Wilhelm Peters[114] has appeared
since this work was done.) The best means of comparing these results,
for our present purposes, and also of bringing them into relation with
the complication-results already obtained, is to take the mean point
between these thresholds, and state its position, in time, relative to
the time of the visual stimulus (letter) just before or after which
the click came. This mean point is called the "Threshold Mean" in the
following tables. In Table IV, for example, "After Letter .026 sec."
means that the mean point between the thresholds, "click first" and
"click last" falls twenty-six sigmas after the time of the exposure of
the letter. These results are readily comparable with those of Peters.
By dividing the "range" by two, and adding the "threshold mean" to one
half, and subtracting it from the other, one has the total interval
between "click first" and "click last" and its place with reference to
the time of the visual stimulus.


TABLE IV

    _Obs._     _Twelve Letters_     _Three Letters_      _One Letter_

  A Threshold  Mean After                                 After
                    Letter .026 sec. On Letter .041 sec.  Letter .020 sec.
       Range               .093 sec.                             .062 sec.

  G Threshold Mean Before               Before            After
                   Letter  .015 sec.    Letter .020 sec.  Letter .062 sec.
       Range               .072 sec.           .083 sec.         .304 sec.

  Sh Threshold Mean Before              After             After
                    Letter .003 sec.    Letter .027 sec.  Letter .003 sec.
       Range               .172 sec.           .111 sec.         .241 sec.

It must be distinctly understood that these "threshold means" are
not displacements, and that the two cannot be compared as if they
were statements of the same facts. These _means_ indicate the centre
of gravity of the "click first" "click last" interval with respect to
the visual stimulus. Changes in this centre of gravity may reasonably
be expected to approximate a variation _inverse_ to that of the
displacements of the auditory stimulus. For example, any change in the
conditions which would tend to increase a _negative_ displacement would
tend also to put the centre of gravity of the "click first" "click
last" interval _after_ the visual stimulus, or, if it were already
after, to increase its time after. So also the _positive_ displacement
and the position of the threshold mean _before_ the visual stimulus may
be considered similar indications. For a click given at the time of the
threshold mean of a given observer, in connection with the same visual
stimulus, would certainly be judged by that observer as simultaneous
with the visual stimulus. If, then, this mean is before the visual
stimulus, the sound will be displaced positively, _i. e._, coördinated
with a visual stimulus coming later. If the mean is after the visual
stimulus, the sound will be displaced negatively, _i. e._, coördinated
with a visual stimulus coming earlier. The position of the mean of the
thresholds indicates a tendency toward the displacement of the auditory
impression in the opposite direction.

In Table III, three out of five observers, A, Sh, and Bo, show a change
from a negative displacement in the series of three to a positive
displacement in the series of twelve. If this were the effect of the
series, the same should show in the series of three as compared with
the single letter. Such a change is manifest in the results of Bo. It
is, however, very slight. The others increase the negative displacement
from the single letter to three letters. In Table IV, of the same three
observers represented, Sh changes the threshold mean from _after_ in
the three-letter series to _before_ in the twelve-letter series, and
A changes from _after_ in one letter to _on_ in three letters. These
changes correspond to changes from negative to positive displacements
for increase of series and introduction of series. G shows the same
change from one letter to three, in both tables. These changes, in
55% of the cases offered for comparison in the two tables, indicate
a _decrease_ of _negative displacement_ and an _introduction_ of
_positive displacement_ as the _effect_ of the _visual series_. The
visual element is made more focal in expectant attention as it is
more isolated, and so the tendency toward negative displacement and
increasing negative displacement as the serial character of the visual
impressions is stripped off. But there are strong counteractive
tendencies, which control the 45% of comparisons not mentioned above,
where the increasing series shows increasing negative displacement.

In the series all the observers adopted the method which has been
outlined above, that of letting the click pick out the letter, or
letting the letter announce itself. One said "the letter hits the
sound." After this sorting-out of the letter, they resorted to the
system of tests and counter-tests, in succeeding trials, to correct
the first impression. One can readily understand, then, that when
they were taken off the series altogether, an entirely different
kind of adjustment had to be made. G did not succeed in making this
new adjustment very well, as is shown by his exceptionally large
range under one letter. He could not get the two impressions to come
together. In attending to either one, he could not get the other in
relation to it. There was something in the visual series which enabled
him to get the visual impression in line with the auditory, and when
this was absent the same kind of work could not be done.

St had also a peculiar method, which was directly dependent upon the
serial character of the visual stimuli and impressions. He allowed the
series of clicks and the series of visual impressions to establish
themselves as a complex rhythm. Each series was rhythmic independently.
The two got connection by means of the click appearing as an
"after-strike," as on the piano, to a member of the visual series.
The letter "flashes out" for him as that of which the click was the
"after-strike." The click was thus between two letters. But there was
no amount of before or after about it. It was a general quality of the
whole complex which was taken to mean such and such a position of click
in the series. What he thus translated into temporal judgments, were
qualitative aspects of the rhythmic experience, to which he usually
attached no temporal meaning whatever. Learning how so to translate
them into temporal terms was a definite process of training for him.
Under these circumstances, he of course had an entirely new lesson to
learn when the visual series was taken away. In fact, it might be, he
would now find no visual impression to which the click could be an
after-strike, and so he would be entirely without material to translate
into temporal terms.

Under these circumstances it is not surprising to find G and St
exceptions to the majority of the observers in this experiment. This
makes more probable the effect of the series, inferred above for the
other observers,--namely, series decreases negative displacement.


THE INFLUENCE OF THE _Position_ OF THE _Series_ OF _Visual_
IMPRESSIONS

It was noticed in the series of the three letters, particularly, that
some observers were much more accurate in their work when the click was
near one end of the series. In this experimental group, the comparison
is between cases where the click is coördinated with (1) the first
member of a visual series of five, (2) the middle member of a series
of five, and (3) the last of such a series. The method was the same as
that used in obtaining the results of Table IV. H was the letter used
in each case for coördination. Results follow in Table V.


TABLE V

  _Obs._                           _H first_     _H middle_     _H last_
   A  Threshold Mean After Letter        After Letter      After Letter
                                .010 (sec.)      .021 (sec.)   .025 (sec.)
      Range                     .072 (sec.)      .085 (sec.)   .093 (sec.)
   G  Threshold Mean  Before Letter      On Letter        Before Letter
                                .007 (sec.)                    .007 (sec.)
      Range                     .124 (sec.)      .083 (sec.)   .151 (sec.)
   R  Threshold Mean After Letter        After Letter     After Letter
                                .032 (sec.)      .016 (sec.)   .042 (sec.)
      Range                     .464 (sec.)      .398 (sec.)   .369 (sec.)
   Sh Threshold Mean After Letter        After Letter     After Letter
                                .025 (sec.)      .015 (sec.)   .030 (sec.)
      Range (sec.)              .176             .176          .166
   St Threshold Mean After Letter        After Letter     After Letter
                               .050 (sec.)   .062 (sec.)       .078 (sec.)
      Range (sec.)             .140 (sec.)   .108 (sec.)       .108 (sec.)

Under these conditions, whatever the effect of the visual series, if
it has any effect, opposite tendencies in direction of displacement
ought to be shown in the "H last" from those in the "H first," results,
as each is contrasted with "H middle." Contrasted in this way, these
results, for A and St, show a relative approach of the mean to zero
for "H first," and a relative departure from zero for "H last," or a
_decrease of a negative displacement for "H first"_ and an _increase
of the same for "H last."_ In other words, the series draws the
displacement of the click toward itself. A _negative displacement is
increased by a series coming before the visual stimulus_ in question,
and _decreased by such a series coming after_. For R and Sh, the
negative displacement is increased in both H first and H last as
compared with H middle, but relatively the most for H last in both
observers. For G there is the same positive displacement introduced
by both H first and H last, but it is less than in any of the other
cases. The drift of the evidence here, then, is that the _visual series
draws the displacement in its own direction_. Each observer who has a
negative displacement (Threshold mean after) with "H middle" increases
this when the series all comes before (H last) and two decrease it
when the series comes after (H first).


THE EFFECT OF RHYTHM (_Repetition of Auditory and of Both Stimuli_)

It is very evident to any one who has worked at all in the complication
experiment, that rhythm plays an important part in the displacement.
Witness also the astronomers' experience cited above, St's waiting
for the rhythm to establish itself, and my own readjustment to the
new conditions when a new combination of intervals was given in the
experiment with varying auditory intervals. In order to show the part
played by rhythm, I tested each one of five observers on several
different days, to fix for each of them both the "click first" and
the "click last" thresholds, as above, under each of the following
conditions: (1) one visual (single letter) and one auditory stimulus
(one pair), (2) one visual (single letter) and many auditory stimuli,
and (3) many visual (single letter repeated) and many auditory stimuli
(many pairs). For visual fixation, the observer had a very dim light at
the end of the observation-tube. The visual stimulus was a flash of red
in the place thus fixated. It had a total duration of less than .005
sec. The surface exposed subtended a vertical visual angle of about
seven tenths of a degree. In the case of one visual and many auditory
stimuli, the visual stimulus was given when the observer had heard the
recurring auditory stimuli several times and had himself given the
"ready" signal. The results follow in Table VI.


TABLE VI

  _Obs._                                     _One Visual and_
                           _One Pair_        _Many Auditory_   _Many Pairs_
  A  Threshold
       Mean   After
               Letter (sec.)  .005  After Letter  .022   After Letter  .042
     Range (sec.)             .021                .024                 .039
  G  Threshold
       Mean   After
               Letter (sec.)  .022  After Letter  .009   After Letter  .005
     Range (sec.)             .078                .083                 .084
  H  Threshold
       Mean   Before
               Letter (sec.)  .011  After Letter  .006   After Letter  .012
     Range (sec.)             .035                .035                 .039
  Hy Threshold
       Mean   After
               Letter (sec.)  .034  After Letter  .030   After Letter  .046
     Range (sec.)             .089                .074                 .072
  St Threshold
       Mean   After
               Letter (sec.)  .054  After Letter  .037   After Letter  .041
     Range (sec.)             .096                .080                 .083

In this experiment, the observers A, H, and Hy, show an increasing
distance of the threshold mean after the visual stimulus, with the
successive introductions of the auditory series and the combined
series. In other words, the second column negative displacement is
larger than that of the first, and the third column has a still larger.
G and St are again exceptions, as they would be expected to be from
the above analysis of their methods. Each did his most accurate work
in a case where there was some rhythm present. St said in regard
to this work "the one pair abolishes the sound as a standard." The
rhythmic factor most missed by these observers, in the case of the
single pair, was the sound; for their results are almost the same in
the second and third columns. Introduction of the repetition of the
visual series does not make any decided difference. A, H, and Hy were
able so to adjust their attention as to get the best results in the
case of the single pair. The rhythm seemed to introduce for them a
subjective rhythm which upset the nice adjustment of attention and so
increased the displacement or the time between the threshold mean and
the visual stimulus. The negative displacement was increased under
these circumstances, probably as a result of the facilitation of the
auditory perceptive process. It has an _opened path_. It is a case of
pre-perception. Even when both were repeated (many pairs) the auditory
dominated, and so did the most at opening its path. But it seems more
likely to me that the rhythm, as such, whether auditory or auditory and
visual, claimed the attention and so proved a distraction from the work
of accurately discriminating the times of the impressions. And this
exaggerated the displacement or lack of discrimination in whichever
direction it was tending before.

In the successive stages of the investigation thus far, the
complication experiment has been stripped down by degrees to the
simple problem of the shortest possible interval between two disparate
stimuli,--in this case shortest auditory-visual and visual-auditory
intervals, as in the one-letter experiment of Table IV and the one-pair
experiment of Table VI. The various factors in the complication
experiment which have been successively analyzed out--the interval
between members of the auditory series, the length of the visual
series, the position of the visual series in relation to the auditory
stimulus, and the auditory series itself--have all been shown to be
factors intimately connected with the way the observer attends to
the stimuli in question. From the present standpoint, it may be said
they are all factors which, being introduced into the simple interval
discrimination experiment, modify the resulting judgment with regard to
the interval, by an interference with the normal attention-processes in
the discrimination of intervals.


INTERVAL DISCRIMINATION

The method of interval discrimination deserves special consideration.
Some of the introspective observations made by observers while engaged
in the work, already reported, are instructive in this connection. In
the case of a single pair, one observer said, "I know which is first
because it gets hit first." This remark is a very apt expression of
my own experience in trying to answer the same question. "Getting
hit first" clearly means, to my mind, some kind of _action_ on the
part of the observer. He was ready, in the moment of preparation for
the experiment, to see a flash of red with his right eye (either eye
could have been used) and to hear a click with his left ear. (The
stimuli were each produced 25 cm. from the respective sense-organs.)
His preparation consisted in securing the "hair-trigger" condition in
the two parts of the cortex and conduction apparatus immediately in
question in the sensing of the two expected stimuli, and other parts
are in a shut-off-from-discharge condition. This is the interpretation
which seems to me an appropriate explanation of the feeling of special
readiness to discharge in these two directions, when the expected
stimuli shall come. The eye- and ear-muscles, in such case, are held
tense on the sides (in the organs) where the stimuli are expected.
The breath is held, and the whole trunk is under a strain. All bodily
processes, in so far as they are controlled, are directed in such wise
as to get whichever of these expected impressions shall come first, in
as short time as possible, in order to know that it is first.

The reaction which gives the basis for the judgment may be a conscious
"hitting" of the first. Or it may be a reaction, ostensibly as a part
of the whole apperceptive process of which the auditory and visual
processes are parts. This reaction may be any one of many kinds.
Often it is a letting-go of the held breath. The exhalation or other
reaction comes in response to the whole stimulating or "setting-off"
process, and the one or the other of the two stimuli is judged to
be first by certain peculiar relations within the experience of the
moment. Such an explanation is in part suggested by the expression
of St, that the visual impression when it came before the auditory,
appeared as a "grace-note," and when it came after the auditory, as
an "after-strike." St played the piano. He himself thought that this
discrimination was a motor affair, _i. e._, a difference judged on
the basis of a difference in the motor response. The judgment of the
temporal order of the two impressions seemed to be an interpretation
or translation of the different motor responses.

A, whose method brought the shortest range in Tables IV, V, and VI,
said, "I hold my breath at the moment of expected stimulation, and it
goes at the first impression." At another time he said, "When I say
'click first' I have the feeling that the click is left, and when I say
'click last,' that the click is on the right." He interpreted this to
mean that when the click sounded first, he had moved slightly toward
it, that is, to the left, and that when the visual stimulus had come
first, he had moved slightly toward it (it was sensed by his left eye),
and this was rather away from the sound, which would have come before
the movement could have been more than initiated.

In my own case, I felt distinctly different motor responses in the
two cases. There was an immediate feeling of release in whichever
organ the stimulus first reached. A little involuntary jerk occurred
in the musculature of this sense-organ, and sometimes the head moved
slightly in the direction of the first stimulus. The condition of
the next moment from which the judgment proceeded seemed to be best
expressed thus, "I had it at a time when the other was not there." The
attention was accurately set for both. Right eye and left ear were both
distinctly innervated. The first stimulus "struck" the appropriate
organ, and the "set" of the organ was released.

I am persuaded that the difference in sensitivity to intervals between
auditory and visual impressions is due, in part, to a difference in
the power of "cocking the ear" to hear, as one fixates the eye to see.
The observers who got the smallest ranges between upper and lower
thresholds had the most distinct kinæsthetic sensations in the moment
of preparation, in the middle ear and about the external meatus.
All had some sensations from the side of the head in question. The
less accurate had a general feeling in the neck-muscles. Accuracy of
discrimination was in no wise connected with voluntary control of the
musculature moving the pinna. This was subject of careful enquiry with
all observers.

If this introspective evidence leads me aright, it seems that the
non-discriminable interval between auditory and visual impressions is
due principally to two things, (1) the impossibility of perfect balance
in the preparation of the attention for two expected stimuli, and (2)
the possible difference in time it takes to react to the different
impressions. The various complicating conditions which are added to
the simple interval discrimination in the cases of a complication
experiment, such as we started with in this investigation, are
chiefly interferences with the first-named factor. They disturb the
nice balances of attention. In this simple discrimination experiment,
under favorable conditions, a close approximation to a balance can
be attained. Any difference in the reaction-times to different
stimuli will remain as a constant error of displacement. It is well
known that reaction-times to visual stimuli are longer than those
to auditory. There is a retinal inertia which delays the perception
of the visual impression, in comparison with the auditory, coming
from exactly simultaneous stimuli. Having this physiological basis,
it will be relatively constant, as compared with the ever-varying
attention-differences.


THE COEXISTENCE OF MENTAL PROCESSES

Having given, then, these relatively fixed temperamental conditions of
reactions to different stimuli, which remain after practice (training
in the control of attention) has reduced the reactions to their
lowest terms, and has secured the conditions which are favorable for
the best balancing of the attention, there is yet one other question
very germane to the subject. It will have occurred already to any one
reading the above, that while the response to one stimulus is being
made, the other may be held in abeyance in the fringe region of the
attention-field, and that it is only brought up to clear perception
when the first has been disposed of. In other words, it may well be
that the first of two simultaneous but disparate stimuli, which gets
a start at setting-off its appropriate response in its sense-organ,
will bring out this response and be perceived before the other one
gets started,--that we do only one thing at a time,--that even in
such minute processes as this there is no possibility of division of
attention. It is hardly probable on the basis of the experimentation
already reported, that this is the case. There is some division of
attention. Otherwise there would be an equal certainty of judgment in
every case, no matter how small the separating interval. But still
the question as to how two mental processes, starting at the same
moment of time, do proceed, as compared to the progress of each of
the same processes when it holds the field alone, is very vital to
the understanding of the psychology of interval-discrimination. And
thus the question of objective time-relations is necessarily involved
in that of making judgments of the time-relations of simple mental
processes (subjective time-relations). The question is, Do these
processes, starting simultaneously, proceed just as freely as if they
were the sole occupants of the field of attention and so had the whole
energy of attention concentrated upon the single process, or do they
_interfere_ with each other?

Distribution of attention, of some sort, is granted. It is generally
conceded that there must be some sort of overlapping of the processes
in any complex mental operation. But there is the greatest lack
of information as to how this overlapping takes place,--as to the
mechanism of the distribution of attention. Fechner held to the notion
of a fixed maximum of available psychophysical energy. If this energy
is being consumed in a single process, that process is very vivid, and
all other processes are below the threshold. If, on the other hand, it
is distributed over several simultaneous processes, they are all of
diminished vividness. Distribution of attention always means diminution
of vividness, and concentration of attention, increase of vividness.
(See Elemente der Psychophysik, vol. 2, p. 451, 1860.) There is no
question of the truth of the last statement, and very likely Fechner's
fundamental concept is a true one; but there is need of more definite
data on the conditions and nature of simple mental processes occurring
at the same time, before it is considered proved.

Such researches as those of Paulhan,[115] Jastrow,[116] Loeb,[117]
and De Sanctis[118] all dealt with the combination of processes which
were themselves quite complex. It may well be that such processes as
reciting a poem, performing a subtraction or multiplication of long
numbers on paper, or keeping time with a metronome with the hand, seem
to go along together when combined, so that the time taken to do the
two of them together is much less than the sum of the times required
for their separate performance, and, in some cases, no greater than the
time required for either alone, and yet there may be no real proceeding
together. The apparent saving of time may be due, as Paulhan suggested,
to a rapid oscillation from one to the other of the two complex
processes which are largely automatic and can proceed, to such extent
as they are automatic, without any attention. This illustrates how
these investigations have probably missed the real point at issue with
regard to the division and distribution of attention. The attention
might be distributed over several of the minuter part-processes of
these processes so that many were proceeding at the same time, and yet
the method of these experiments would not reveal it. They were not
planned with sufficient precision. There is a problem in the division
of attention which they did not come within sight of, and this is the
real question of division in case of the simplest processes.

This problem is really that of the mechanism of mental assimilation.
The process it investigates is illustrated by the maturing collective
idea, as a melody or a spoken sentence. There is a gradual enrichment
or growth in meaning, as such a process goes on toward its completion.
At any instant during the process, implicit associative and nascent
perceptive elements are working together to their own mutual
clarification and explication. All focal content is the result of
complicated interworkings of such fringe material. It is impossible,
it seems to me, to question the causal relation of the fringe elements
or processes of one moment to the focal of the next; and it is equally
impossible to deny the complication of these same fringe processes.
They must go on at the same time in order to enter into one and the
same resultant process. The question of direct interest at this point
in the discussion is, To what extent do they proceed at the same time?

It would seem from the way in which this question, of the relationship
and interference of mental processes which proceed or start to proceed
at the same time, has come up in this investigation, that the natural
method of pursuing it would be that of comparing reaction times for
cognition reactions to the single and combined stimuli. But we are
warned against this by very clear inferences from an investigation of
Professor Münsterberg's in which he used the reaction method.[119]
By an ingenious use of the reaction experiment, the author shows
that two-part processes in a reaction, as, for example, a restricted
judgment of class and a subjective preference, occupy about the same
time when combined in a single reaction as when either is performed
in a separate reaction. In other words, two judgments of distinctly
different kinds can be made in the same time as either can be made when
it has all the attention concentrated upon it. The conclusion that
these elementary processes go on together--that at least there is some
degree of overlapping--seems unavoidable.

But when the first part of the same report is considered in relation
to the second, it is clearly shown that the reaction experiment is not
adapted to the finer investigation of this problem. For the first part
shows that no matter how much a _motor_ reaction is complicated by
choices or other judgments, it always takes place in just about the
same time as the simple reaction. The complications may be such as
actually to double the reaction time in the case of a sensory reaction,
and yet a motor reaction, under precisely the same conditions as far as
they may be the same for a motor, shows no increase in time. The "set"
of the attention in the motor reaction is, no doubt, such a change
in the order of succession of the parts of the process that some of
those, which come after the stimulus is received in the case of the
sensory reaction, are made to come before the stimulus in the motor.
When, however, it is found that the motor response to the question,
"Is this the name of a scientist, philosopher, poet, statesman, or
musician,--Sappho?" is made by the appropriate finger, as previously
agreed upon, in just as short a time as the observer can make a motor
response with any finger to a simple auditory stimulus, it indicates,
either that the whole of the choice judgment has been made before the
stimulus was received, or that the judgment itself is so automatic
that it is practically a reflex. This latter cannot be true. The
judgment, as conscious choice, cannot be made before the stimulus is
given, _i. e._, until the question is completed. And judgment cannot
be made automatic and yet be a judgment. In fact both alternatives are
untenable, and there is no other course than to hold the situation
which gave rise to them at fault. If the judgment process here required
previous to the reaction does take time apart from the processes of the
simple reaction, the reaction process is shown by these experiments to
be unable to exhibit it. A more microscopic method is demanded before
the matter can be settled.

The Leipsic method of measuring the scope of attention by means of
the tachistoscope is the standard means of securing data as to the
_number_ of elementary processes which can go on at the same time in
consciousness. The same question, with which we are here concerned,
grows directly out of the investigation of the number of processes
which can go on together. Wundt acknowledges the great difficulty which
inheres in the investigation of this problem.[120] Cattell's early work
with the tachistoscope showing the numbers of letters, syllables, and
words, which could be apperceived under the same objective conditions,
indicated the great importance of what we may best call _meaning_, in
apperception, and its influence on the number of different processes
which may proceed together. In fact the number depends upon the
definition of the unit with which the investigator starts out. Wirth,
the latest emendator of the tachistoscopic method, has shown,[121] in
a very thoroughgoing and genuinely constructive criticism of earlier
work, that the one question of primary importance in investigations
of the content of the moment of consciousness, _i. e._, scope of
attention, is to set forth the relative clearnesses of the elementary
processes there proceeding together. He shows that the different grades
of clearness which may present themselves in the field of consciousness
of a momentary act indicate, on the one hand, the impossibility of
sharply distinguishing the "scope of attention" from the "scope of
consciousness" as Wundt uses these terms, and, on the other, the
serious indefiniteness of any merely numerical statement of the scope
of attention. His main purpose is to set forth a method by which
this field can be enriched by exhaustive statements of the relative
clearnesses of the processes going on at the same time. All the work of
the present study had been performed before the publication of Wirth's
work. Otherwise some of his suggestions would have been used in the
plan of the experiments following. I may say, however, that I believe
the method here used has its own distinctive merits.


GENERAL METHOD FOR TESTS IN COEXISTENCE

Taking the suggestions offered by Professor Münsterberg's study of
apperceptive and associative processes, I selected simple _judgments
of comparison_ as the best means of trying-out this question of
coexistence. The perceptive act itself is made up of judgments, and
these may very properly be the processes studied in combination as in
the tachistoscopic experiment. But the judgment which has a previous
perceptive act as its condition, determining its start, seems to be
better under control. It is itself a central process, not dependent
upon the variations of the objective factors in sensation. My plan
was to have the stimuli so arranged as to give rise to two or more
perceived conditions at the same moment, and so have one or more
judgments of comparison between the perceived features made at the
moment the perceptions were completed and immediately stated. If one
makes two series of single judgments of comparison, and a series
wherein these two judgments are combined in a single act, all three
under precisely the same objective conditions, and the same subjective
conditions, saving only the necessary changes in the _direction_ of
the attention, and the percentage of correct judgments is recorded in
each case, providing always that in no single series of judgments were
the conditions such that all judgments could be correctly given, he
would then have reasonable grounds for making inferences with respect
to the _interference_ of simple mental processes going on at the same
time,--whether there is any, and, if there is, how much there is.
_Interference_ would be indicated by the _falling-off in percentage of
correct judgments as the combinations were increased_.

Such relative accuracy of judgments, single and combined, was the test
sought after and relied upon in the following experiments. It was very
necessary to have the objective conditions such that the results in
cases of single judgments, later to be combined, should be short of
absolute correctness, in order that interference from the combination
should show itself in impaired accuracy. Otherwise there might be some
_free energy of attention_, which could readily take up the extra work
when the judgments were combined, and so there would be no impairment
of accuracy. It was the aim to have the objective conditions, such as
duration and extent, so regulated that about ninety per cent correct
judgments resulted in the series of single judgments. If, then, when
two were combined, eighty per cent were given correctly, and when three
were combined, seventy per cent, the inference would seem reasonable
that this falling-off in correctness was due to interference. The
failure of the perceptive process, indicated by the ten per cent
incorrect judgments in the series of singles, would remain a _constant_
source of error throughout.

There is, however, one other source of increasing error, with the
increasing combination of judgments, supposed above. The judgment
processes might go on at the same time without any impairment of the
accuracy of the single judgment, and yet the results, as expressed,
might show a falling-off in accuracy. This imperfection would then be
due to a partial failure of the retentive and reproductive processes,
and not to the imperfection of the judgment processes. I have found
no sure means of separating this factor and excluding it. As the
experiments were arranged and conducted, though, I believe any
impairment in accuracy resulting from combination is more likely due to
interference of the judgment processes.

If neither of these factors is efficient, on the other hand, there will
result no falling-off in accuracy of results when single judgments are
combined. To be sure the conditions of the experiment, as outlined
so far, do not preclude the possibility of the combined judgments
occurring in succession, and so giving rise to as large a percentage
of correct results as when occurring singly. That is, while one of the
so-called combined judgments was in process, the latent conditions
of the other would remain for the moment as mere physiological or
possibly psychical dispositions, and to these one would "hark back" in
the next moment. Here the reaction method is suggested as the means of
assurance that this is not the case. But this method, we have already
seen, will not lend itself to work of such precision as this. The
probability of this succession of judgments is reduced to a minimum in
the experimental groups following.

It can be practically precluded by a prevention of all sensory images.
In these experiments every precaution was used to prevent them. In
all cases where visual stimuli were used, for instance, a brightly
illuminated blue field immediately succeeded the momentary stimulus,
while comparative darkness preceded it. I cannot be so sure that there
were no memory images functioning. But all observers were carefully
questioned on this point at frequent intervals during the experiments,
and no evidence of their existence was found in any case. I feel sure
sensory images were excluded and think memory images very improbable.


SINGLE AND COMBINED JUDGMENTS FROM VISUAL AND TACTUAL STIMULI

In this group of experiments, I used judgments from visual and tactual
stimuli, singly and in combination. Both stimuli were given by means
of a large pendulum in the Harvard laboratory, specially constructed
for Professor Münsterberg. This pendulum is about one and a half metres
in length. It is hung in a heavy steel frame which rests upon a large
table. A curved steel bar, concentric with the swing of the pendulum,
and ninety degrees in extent, is so set to the frame that it serves
as the attachment for an electro-magnet, at any point in the swing of
the pendulum. The pendulum-rod carries an armature which fits this
magnet. By means of this magnet, the pendulum may be held at any point
between the position of rest and forty-five degrees out in either
direction; and it may be released by _breaking_ the circuit through the
magnet. The pendulum also carries a segmental screen of about seventy
degrees extent. An opening about nine by eight centimetres near the
centre of the screen affords means of tachistoscopic observations. A
sliding shutter makes the slit as narrow as may be desired. In these
experiments a black tube was set up, at right angles to the direction
of the motion of the pendulum, and at the height of the slit in the
screen. On the other side of the pendulum screen, and directly opposite
the tube, was placed a support for holding the object to be shown.

The object for the visual stimulus was one of two light gray lines on a
black background. These lines were 4 mm. wide, and one 44 mm. long, and
the other 40 mm. The work was done in a dark room. The stimulus card
was illuminated by an electric light hanging between it and the screen.
Both cards were shown the observer several times, before experimenting,
till he was sure of their lengths. Upon one being shown, in experiment,
he was asked to say whether it was the longer or shorter. The touch
apparatus was so arranged that the experimenter could at will give
the observer one or two contacts on the back of his right hand. The
contacts were made by means of an electro-magnet. This was actuated by
a current which was _made_ by the closing of a switch which was secured
by a set-screw to the same curved steel bar as bore the pendulum
magnet. This switch was closed by the pendulum in passing. It was
adjustable on the bar. Another similar switch, opened by the falling
pendulum the next instant, removed the tactual stimuli. These switches
were so placed in the course of the pendulum fall that the tactual and
visual stimuli were exactly simultaneous. The tactual judgments were,
_one_ or _two_ points touched. Results are presented in Table VII for
three observers, A, B, and Bo. The number of series which were averaged
in each case is given, in order properly to weight the results.


TABLE VII

                                                        _Combined_
                         _Single_          _Single_     _Tactual_
    _Obs._               _Tactual_         _Visual_     _and Visual_
    A Number of series
        averaged             3                 3             3
      Per cent Correct      80                89            79
      Judgments              \-------\/-------/
         Average                     84

    B Number of series
        averaged             5                 5             5
      Per cent Correct      72                78            78
      Judgments              \-------\/-------/
         Average                     75

    Bo Number of series
         averaged            5                 5             5
      Per cent Correct      88                71            78
      Judgments              \-------\/-------/
         Average                     79

In Table VIII are given A's results for further experimentation under
the same conditions, also for a pair of visual judgments, a pair of
tactual, and all four combined. One series of each of the five was
given each hour of experimentation. For the additional visual judgment,
the observer was required to say whether the line was _high_ or _low_.
It was of two heights from the lower edge of the card, 17 mm. and 20
mm. For the other tactual judgment, he reported the point or points
touched on the hand, as on the _right_ or _left_ side. The middle line
was traced by the experimenter, before tests, as often as the observer
wished to be reassured of its position.


TABLE VIII

                                        _Tact._                 _Two Tact._
  _Obs._             _Sing._  _Sing._   _and_    _Two_    _Two_    _and_
                     _Tact_   _Vis._   _Vis._   _Tact._  _Vis._  _Two Vis._
  A Number of series
      averaged          6        6        6        6        6        6
    Per cent Correct
      Judgments        87       88       83        9       81       83

The next additional combination was a pair of judgments based upon
auditory stimuli. Four electric clickers were placed on the wall behind
the observer. Two were loud and two faint. Each pair was accurately
adjusted so they were of the same intensity and quality. One of each
pair, _i. e._, one loud and one faint, were hung about four feet to
the left of the observer's median plane. The other two were hung at
an equal distance to the right of this plane. The circuit making the
click was _made_ by a switch closed by the pendulum as it fell. The
experimenter by pressing any one of four buttons gave the one of
the clicks he desired. The observer's two judgments were as to the
_loudness_ and the _position of the click_.


TABLE IX

                           _Two Vis._     _Two Tact._     _Two Aud._
    _Obs._               _Lgth._ _Pos._  _Num._ _Pos._  _Inten._ _Pos._
    B  Number of series
        averaged            8      8       8      8       8       8
       Per cent Correct    82     94      89    100      94      92
       Judgments           \----------------\   /----------------/
           Average                           91.7

    Bo Number of series
        averaged           12     12      12     12      12      12
       Per cent Correct    77     91      87     97      81      82
       Judgments           \----------------\   /----------------/
           Average                           85.8

                                    _Six Judgments Together_
                             _Visual_       _Tactual_      _Auditory_
    _Obs._                _Lgth._ _Pos._  _Num._ _Pos._  _Inten._ _Pos._
    B  Number of series
        averaged             14     14      14     14       14     14
       Per cent Correct      89     97      86     96       70     86
       Judgments             \----------------\   /----------------/
           Average                             87.3

    Bo Number of series
        averaged             20     20      20     20       20     20
       Per cent Correct      77     93      85     98       81     80
       Judgments             \----------------\   /----------------/
           Average                             85.7

It is evident, on the face of these returns, that there is no positive
assurance of interference. Each of these observers had been in some
part of the complication work. And so the inference from lack of
evidence here can be carried back to that work, and we may rest assured
that the lack of accuracy in interval discrimination work by these
observers was due in minimal measure, if in any, to interference of
the mental processes, auditory and visual, tending to proceed at the
same time. Some parts of the results here presented look like evidence
for interference. But there is, on the whole, just as much evidence
of what one might call facilitation, in combination, as there is of
interference.

There is one source of possible explanation for the non-appearance
of evidence of interference in these results: that is the fact that
the stimuli are disparate, and so probably take different times for
maturing. Thus the judgment processes, so far as they thus start
from disparate sensations, may start at different times. There was
good reason for using disparate stimuli first for the combination of
two mental processes, as this was the closest related to the simple
interval discrimination experiment to which the complication experiment
had been reduced. But this objection is now easily overridden by making
the conditions of experiment such that all judgments start from one and
the same perceptual process.


ONE, TWO, AND THREE JUDGMENTS BASED UPON A SINGLE SENSE-PERCEPTION

The conditions here were such that the perceptual basis for any one of
the single judgments was at the same time the possible basis for any
other single judgment and also for any or all of them combined. What
judgment or judgments were given depended entirely upon the directions
given, and the consequent preparation of the attention. Under these
conditions, there could no longer be any doubt about the even start
of all judgments, so far as outer conditions were concerned. The only
remaining cause of an uneven finish--lagging of a process, as shown by
its increased inaccuracy when combined--must be interference with its
progress by other processes going on at the same time.

Visual stimuli were used. The objects to give the perceptual basis
for the judgments were small rectangular openings in cardboard seen,
on exposure, by _transmitted_ light. These rectangular windows in the
cardboard were 2 cm. by 1 cm. and stood in the vertical position 1 cm.
apart. The judgments were all based upon differences existing between
these rectangles as shown. One of these differences was in _length_.
They might be of the same length, or either the right or left might
be 2 mm. longer than the other. Another difference was in _shade_.
This was secured by different thicknesses of paper, pasted over the
openings. Two shades were used. The opening on one side might be shown
as either the same brightness, brighter, or less bright. The third
difference was in the _number of lines_ which crossed the rectangles.
Two or three wires were placed across them horizontally and about 5 mm.
apart. Thus they had the same number of lines, or one had fewer or more
than the other.

The same large pendulum was used in these experiments. The moveable
magnet on the curved steel bar was kept in one position throughout.
It held the pendulum, ready for release, at twenty degrees from the
position of rest. The adjustable weight on the pendulum was also kept
in one position. The only adjustment which was changed during this
series of experiments was the width of the slit in the window of the
screen. This was varied from one millimetre to five. The whole time
during which any part of the two rectangles was in view (the total
exposure) with a 5 mm. slit was .033 sec.; with a 3 mm. slit .031 sec.;
with a 1 mm. slit .029 sec. These times were measured with a Hipp's
chronoscope. The entire visual field, embracing the two rectangles, was
about 2 cm. by 3 cm., and was about three fourths of a metre from the
observer's eye. It could be accurately fixated beforehand and fully
exploited during the moment of exposure.

The observer was always instructed to give his judgments in terms of
one of the two rectangles. If, for example, length was in question, he
should say of the left-hand rectangle that it was longer, shorter, or
of the same length as the right-hand one. The process of expressing
the judgments was also facilitated by using the terms plus, minus, and
equal, for all three sorts of judgments. This was a special aid to
expression where two or more judgments were in question at the same
time. In these cases the observer was always given an order beforehand,
in which the judgments were to be given. This order for the three
combined, for example, was always, "length, lines, shade," as in the
following tables. If, then, the judgments were given "plus, minus,
minus," it meant that the left-hand rectangle was longer, had fewer
lines, and was less bright than the right. The process of making these
interpretations, as well as the order, was made automatic with the
observer, by practice, before experimenting.

Three observers, A, B, and Y, were used in this experiment. The
judgments were made in series of ten. Each hour's work was distributed
over (1) several series of single judgments, (2) two at a time, and
(3) three at a time, the aim being to get an equal number of judgments
of each kind, length, lines, and shade, under each of the three
conditions. The results are given as general percentages of correct
results. To properly weight these averages, the number of series (of
ten judgments each) which are included in making up any average, is
given just above the average.


TABLE X

    KEY:  Lth = Length
          Ln  = Lines
          Sh  = Shade

                  _Single Judgment_   _Two Judgments_   _Three Judgments_
    _Obs._         _Lth_ _Ln_ _Sh_    _Lth_ _Ln_ _Sh_    _Lth_ _Ln_ _Sh_
    A Number of
       series
       averaged      11   12   10       15   17   14       11   11   11
      Per cent
       Correct
       Judgments     95   93   96       90   91   83       94   91   89
                     \----\/----/       \----\/----/       \----\/----/
         Average          95                 88                 91

    B Number of
       series
       averaged       8    8    7        9   10    9        7    7    7
      Per cent
       Correct
       Judgments     93   80   90       76   77   90       75   70   75
                     \----\/----/       \----\/----/       \----\/----/
         Average          88                 81                 73

    Y Number of
       series
       averaged      12   13   13       19   19   16       13   13   13
      Per cent
       Correct
       Judgments     76   80   58       72   76   61       72   74   58
                     \----\/----/       \----\/----/       \----\/----/
         Average          71                 70                 68

Since these general averages for the single judgments are so close
to those in pairs, it seemed possible that the presence of objective
differences, other than the single one asked for, might be a
distracting agent, and really interfere with the judgment process
in question. For example, when judgment on length was in question,
it might be possible to give it correctly a larger number of times,
if there were no differences in shade or lines, than if these were
present. Some careful test experiments were made with a view to
clearing up this situation. The observers in no case knew the nature
of the investigation, nor were they aware that other differences
were absent in some of the cases. The results presented in Table XI
certainly show that the presence of other differences than the one in
question is no cause of interference.


TABLE XI

                        _Length_           _Lines_             _Shade_
                    _With_             _With_              _With_
    _Obs._         _Diffs._  _Alone_  _Diffs._  _Alone_   _Diffs._ _Alone_
    A  Number of
        series
        averaged      5         5        5         5         5        5
       Per cent
        Correct
        Judgments    98        96       96        96        98       94

    B  Number of
        series
        averaged      5         5        5         5         5        5
       Per cent
        Correct
        Judgments    90        92       96        94        78       84

    Y Number of
       series
       averaged       7         8        8         8         7        8
      Per cent
       Correct
       Judgments     73        76       75        64        88       67

Notwithstanding the precautions taken to secure the full energy of
attention for the single judgment process, as already indicated in the
discussion preliminary to these experiments,--namely, by making the
stimulation conditions so near the threshold that only a part of the
judgments could be given correctly,--there still appeared a probability
that there was free energy of attention during the single judgment
process. The observers seemed to do more work when more judgments were
asked for. If this is true, the results of Table X are not a true index
of interference. If there is free energy during the moment of making
the single judgment, this may readily be used for another process when
combined with the first, and so there will be no interference. This
is a sufficient proof so far as it has immediate bearing upon the
interval discrimination experiment, but the further question as to what
will take place if we can use this free energy, if it exists, in both
processes alike, is an important one for the question of the relation
of two processes going on together in consciousness.

To ascertain the fact in this matter, I performed a series of
experiments with the same observers, in which previous occupation
of the mind served as a distraction. The distraction consisted in a
simple arithmetical operation,--addition or subtraction. The moment
before giving the stimulus for the judgment processes,--in the place
of the "ready" signal, I would call out some numbers, as, for example,
"twenty-four from sixty-three" or "fifty-seven and fifteen," the
first indicating subtraction and the second addition. The answer to
the addition or subtraction was always given before the judgment or
judgments, to make sure that it was performed. And in any case where
the observer knew that the addition or subtraction was done before he
attended to the stimulus for the judgment, that particular test was
thrown out. The results are given in the same form as in Table X.


TABLE XII

(_Addition and Subtraction as a Distraction_)

    KEY:  Lth = Length
          Ln  = Lines
          Sh  = Shade

                  _Single Judgment_   _Two Judgments_   _Three Judgments_
    _Obs._         _Lth_ _Ln_ _Sh_    _Lth_ _Ln_ _Sh_    _Lth_ _Ln_ _Sh_
    A Number of
       series
       averaged       8    8    8       15   15   16       15   15   15
      Per cent
       Correct
       Judgments     79   84   67       66   68   66       53   69   59
                     \----\/----/       \----\/----/       \----\/----/
          Average         77                 67                 60

    B Number of
       series
       averaged       6    4    5        8    7   11        7    7    7
      Per cent
       Correct
       Judgments     57   70   60       66   44   53       51   50   44
                     \----\/----/       \----\/----/       \----\/----/
          Average         62                 54                 52

    Y Number of
       series
       averaged       8    8    7       15   16   15       14   14   14
      Per cent
       Correct
       Judgments     66   60   51       56   62   56       66   57   55
                     \----\/----/       \----\/----/       \----\/----/
          Average         59                 58                 59

These results (general average percentages) show, for observer A, a
more regular and somewhat larger falling-off with combination than in
Table X, for B and for Y, a diminished falling-off, and relatively less
for the three than for the two combined judgments. The percentages are
lower throughout. This is a result to be expected. But there is no
notable change in the relative lowering of two judgments in comparison
with single judgments, or of three in comparison with two, such as
should appear if, as supposed, in the experiment resulting in Table
X, there had been free energy of attention in the case of the single
judgment.

It was my aim in these experiments, with distraction through another
simultaneous process, to secure a uniform residue of attention for
the judgment processes, whether single, in twos, or in threes. The
arithmetical operations were therefore as uniform as possible. But
it may readily be that very unequal demands were made upon a given
observer by successive operations, one's automatisations in number-work
may be so various. These would no doubt tend to average up in the
course of the whole work running through several weeks. But in order
to make more sure of the point, I tried another means of using the
free energy of attention which may exist in the case of the single
judgment, namely, by suggesting a judgment or series of judgments just
before an exposure. It will be recalled that the order of judgments
was always the same as that of the tables, and that all were expressed
as minus, plus, or equal. So if the experimenter called out before
a three-judgment exposure, "plus, equal, minus," it would be in the
nature of a challenge to the observer to assure himself beyond a
doubt whether or not the exposure showed the left-hand rectangle as
longer than the right, having the same number of lines, and being less
bright. The so-called suggestion was a distinct factor in heightening
attention. This is shown especially in Y's case by the larger
percentage of correct judgments. Results are averaged in Table XIII.


TABLE XIII

(_Attention heightened by Suggested Judgments_)

    KEY:  Lth = Length
          Ln  = Lines
          Sh  = Shade

                  _Single Judgments_   _Two Judgments_   _Three Judgments_
    _Obs._          _Lth_ _Ln_ _Sh_    _Lth_ _Ln_ _Sh_    _Lth_ _Ln_ _Sh_
    A Number of
       series
       averaged       2    4    2        6    7    7        8    8    8
      Per cent
       Correct
       Judgments     90   89   85       87   81   76       92   72   77
                     \----\/----/       \----\/----/       \----\/----/
         Average          87                 81                 80

    B Number of
       series
       averaged       4    3    4        8    9    7        8    8    8
      Per cent
       Correct
       Judgments     85   70   92       84   80   83       84   74   81
                     \----\/----/       \----\/----/       \----\/----/
         Average          82                 82                 80

    Y Number of
       series
       averaged       6    6    5       11   13   12       16   16   16
      Per cent
       Correct
       Judgments     88   75   94       90   81   77       89   74   70
                     \----\/----/       \----\/----/       \----\/----/
         Average          86                 83                 78

An analysis of the results obtained from B to show the effect of the
suggestions is given in Table XIV.


TABLE XIV

                               _Single_      _Two_       _Three_
                             _Judgments_  _Judgments_  _Judgments_
    Per cent of right suggs.
      judged correctly           87           85           79
    Per cent of wrong suggs.
      judged correctly           73           77           80

The effect of the so-called suggestions in making for correct judgments
was then quite noticeable in the case of single judgments, less so in
two judgments, and none whatever in three. This observer was able to
overcome 73% to 80% of the wrong so-called suggestions. Now, when it
is considered that only 80% to 82% of all the judgments given by B
(see Table XIII) are correct, it is very clear that their action as
suggestions was very slight. They had an influence, however. It was
shown, as expected, in a heightened attention. This was especially the
case with Y. Compare his general averages in Table XIII with those in
Table X. This rise in general averages coincides with the impression of
the experimenter during the experiment. It seemed then that this was a
distinct challenge to keen attention on the part of Y. He is a man who
intends to make impartial observations for himself, and has no notion
of being told what he is to see. That his general averages of correct
judgments stand so much farther apart in this case with heightened
attention than in either of the others (see Tables X and XII) is
indicative of an interference of the judgment processes themselves.

Such a series of general averages as those of Y in Table XIII, as
those of B in Table X, or as those of A in Table XII, seem, in
themselves, and under the conditions of the experiment, to be pretty
clear indication of an interference of simple mental processes carried
on at the same time. The only other explanation is that suggested
above, namely, an interference of the processes of reproduction and
expression. The conditions of the experiment seem to reduce the
probability of this to a minimum. But the centre of interest, in
considering the results, does not lie in the question as to whether it
is interference of the judgment processes themselves or the processes
of their reproduction. The foreground is occupied by a prior question,
namely, whether there is any evidence here presented for interference.
For if there is interference of such processes, why does it not show
up in the results for each of the observers in each of the Tables X,
XII, and XIII? Of the nine cases here offered for comparison, only
the three above designated show what may be called clear evidence
of progressively increasing interference with increase of combined
processes proceeding at the same time.

Under these circumstances this cannot be accepted as indisputable
evidence of interference. Such results as those of A in Table X, where
correct judgments, two at the same time, are given in 88% of the cases,
and three at the same time, in 91% of the cases, stand directly opposed
to interference. They seem to show a facilitation by combination. This
is indeed possible where three and only three sorts of judgment are
worked with. It is the limiting case, and if more than one is asked
for it is really easier to give three than to select two. A himself
remarked that this was the case. A similar explanation holds concerning
the results of B in Table XII, and those of A in Table XIII. If such
an explanation is the true one, it is manifest that the "limit of
attention," of which mention has been made above, has probably not been
reached in any of these cases. On the whole, these experiments _seem
to indicate a small degree of interference of simple mental processes
going on at the same time_. But such interference _cannot be considered
proved_ by these experiments.


THE QUESTION OF SYNERGY

In connection with these last experiments, where the comparative
judgments all proceed from one definite perceptive act, and where
therefore the conditions are most accurately controlled for showing the
effect of interference, if it is a fact, there is yet another means
of looking into that question. This is afforded by the _similarity_
of the _means of expressing_ the different kinds of judgment. In
connection with the current emphasis given to the motor side of mental
processes, it is often urged that mental processes go on at the same
time when they are working together toward one and the same motor
out-go. Otherwise they are likely, at least, to hinder each other,
and to take their turns. If such is the case, the similarity of motor
out-go which is present in these cases, where all three judgments are
plus, or all minus, or all equal, ought to produce a larger percentage
of correct judgments than is found in cases where there are two or
three kinds of expression. Furthermore, if there is no interference
of the judgment processes as such, but, as supposed possible above,
the impaired accuracy of judgment in the combination of judgments is
due to the imperfection of the memory, this too will be diminished by
similarity of expression of the three judgments. In fact similarity
should reduce this source of error to a minimum. The results presented
in Tables X, XII, and XIII, for three combined judgments, were worked
over, so far as possible, and all cases where the three judgments, if
correctly made, would have been expressed similarly, were separated
out. The total number of such cases, and all those where the judgments
would have been properly expressed dissimilarly, are recorded for
each observer in Table XV. The number actually given correctly under
each class is also recorded, as well as the _percentage_ of correct
judgments in each class for each observer.


TABLE XV

    KEY:  TN = Total Number
          CJ = Correct Judgments
          N  = Number
          %  = Per cent

             _Judgments expressed_    _Judgments expressed_
                  _Similarly_             _Dissimilarly_
    _Obs._     _TN_      _CJ_            _TN_      _CJ_
                      _N_    _%_                _N_    _%_
     A         222    172    77          561    479    85
     B         195    153    78          498    356    71
     Y         372    277    74          863    572    66

If the results of B and Y were presented alone, they would seem to
indicate synergy of similarly expressed judgments. But those of A
are most strongly contradictory of such a working together of such
judgments. This is very surprising to me, as A had such a facility in
expressing these similar judgments, especially "equal, equal, equal,"
that it suggested this comparison. But the apparent facile expression
is here shown to have attended a diminished accuracy. No conclusion
can be drawn with respect to synergic influence from the similarity of
expression of judgments.


RELATION OF OBJECTIVE AND SUBJECTIVE SIMULTANEITY

Reviewing this work in combination of judgments with reference to its
bearing upon the complication results, and the interval discrimination
results, it seems that interference of simple mental processes going
on at the same time, though it appears to be a fact, showing itself in
impaired accuracy of processes combined, is yet quite inadequate to
explain the whole, or indeed, any considerable part of the synchronism,
as we may call the "click first" "click last" interval of Tables IV,
V, and VI. The slight amount of interference of such processes as
the auditory and visual perceptions, tending to proceed at the same
time, would tend to a very slight displacement of one with regard
to the other. It is true, for reasons already discussed, that this
time-difference is so slight and so difficult of seizure that it cannot
be measured, and so no measure is offered. We cannot, therefore, be
certain _how much_ of the non-detectable interval is due to this cause.
But the evidence offered in the above tables of results is ample
justification for the statement that _interference_ can be responsible
for only a _very small part_ of the "click first" "click last" interval.

In the case of this interval, as in that of an interval between any
disparate stimuli, a part of it must be due to the different resistance
or inertia of the sense-organs. The eye is undoubtedly slower than the
ear. This would at once suggest itself as the cause of the interval
between the threshold mean and the visual stimulus in the results
shown in Tables IV, V, and VI above. That is, vision being slower, an
auditory stimulus given at the same time as a visual will appear to be
earlier, and it may be given considerably later and yet appear earlier.
In general, therefore, so far as this cause is active, one would expect
that the interval, at which a sound must precede a visual stimulus
in order to be certainly distinguished as coming before the latter,
would be much shorter than the interval, at which a sound coming
after a visual stimulus could be unfailingly distinguished as coming
later. In other words, the centre of gravity of the "click first"
"click last" interval, so far as this visual inertia is the cause of
its displacement with reference to the visual stimulus time, will be
_after_ the visual stimulus.

In one case in my results, Table V, St, H middle, there is presented
an extreme where not only the centre of gravity (threshold mean) is
placed after the visual stimulus (letter), but the whole synchronous
period ("click first" "click last" interval) is after the visual
stimulus, so that a sound coming .008 sec. after the visual stimulus is
distinguished with certainty as coming before it. So also St, in Table
VI, one pair, the sound coming .006 sec. after is judged as coming
before the visual stimulus.

But the variety of displacements of the threshold mean in different
observers, and more particularly in the same observer under different
experimental conditions, indicates very clearly that there are factors
other than visual inertia which are quite as important, and perhaps
equally responsible for this displacement. In Table VI, H, one pair,
for example, the threshold mean is before the visual stimulus .011
sec. So in Table V, G, H first, and also H last, it is before the
letter .005 sec. In these cases there must be some factor or factors
quite as strong as this visual inertia, and counteractive to it. These
are, in part, the complex attention factors which have been referred
to already. Prominent among them are the rhythmic perception which
is so marked in St; the movement toward the first stimulus and the
"letting-go" of the breath, of A; the passive "striking" of the letter
by the sound, in the case of some of the observers; and the "cocking"
of the eye and the ear, of others. These all have to do with the length
and place of the "click first" "click last" interval quite as much as
does the visual inertia. But however this may be, of this inertia and
the other factors just now named, probably each has more to do with it
than does the interference of the perception processes themselves.

But after eliminating the parts played by each and all of these
agencies in the determination of the interval, there will remain a
period of "present time," in which there are no time-differences,
and no qualitative differences which lead the subject to suspect the
existence of time-differences. The mental content of this reduced
synchronous period in experience is _one experience_. The sound was
heard and the letter was seen, but they came _together_ as aspects of
_one_ experience. In the moment of perceiving either one, it was not
possible to say that the other was already a memory. In other words,
the _primary memory_ of either, whichever came first, had lasted over
into the perception of the second. There had been no perceivable
transformation of the first since the instant of its perception. At
the moment of the inception of the second process, the first was
still, to the perceiving subject, what it was at the moment of its own
inception. Though change was probably going on in the physiological
substrata of the mental process in question, in every minutest moment
of the interval, yet a certain amount of effect of this change had to
accumulate before the observer could become aware of the change, and so
be aware of the passing of time or of temporal difference. This was,
then, only a case of the working of the law of _relativity_. And the
perception of time is a function of the duration and amount of change
of mental process.

Looked at from this point of view, we see the whole explanation of the
existence, the amount, and the position of this synchronous period
under one rubric, if only we could grant the combination of mental
processes without interference. If mental processes go on together,
the sole ground of the imperceptibility of short periods of time
separating mental processes is in the fact that the first of these
processes _has not changed sufficiently to be known as different_, to
the perceiving subject. The minimal perceivable interval will vary
from man to man, and in the same man from time to time, inversely
as the amount of change per unit of time, in the process itself.
The same statement could be made in terms of vividness or relative
clearness. The more focal the idea or process, _i. e._, the more vivid
or relatively clear it is, the more rapid will be the changes and the
perception of those changes. Professor Münsterberg's physiological
explanation of vividness,[122] as due to the facilitation of the motor
discharge, has already found confirmation in the method of keenest
interval discrimination as outlined above. The more rapidly the first
process can get into action, the more is the discriminated interval
shortened. So in Exner's experiments, where it was known which of two
stimuli would come first, the interval was very much shorter than any
of my results, for the motor preparation could be made very complete
beforehand, as in a muscular reaction. Therefore the perceptible
change, upon perception of the stimulus, occurred in a shorter time.
Under any circumstances, the conditions, subjective or objective, which
make for rapid maturing (and by the principle of dynamogenesis maturing
means going over into action) of the mental process, make also for the
shortening of the least perceptible interval.

These conditions are as various as the gamut of human experience is
wide. There is nothing, from the primary temperamental characteristics
to the passing wave of feeling of the present moment, which does not
affect it. Most particularly, though, is it a matter of the relations
existing among the elementary processes striving to go on together.
Among the focal and fringe elements of a given moment of experience, no
matter how carefully the practised introspectionist may strive after
an ideal condition of monoideism, there is an incessant interaction.
There are all sorts of hindrances and facilitations. Herein is the
justification of Stern's statement that the "praesenzzeit," as he calls
it, "varies with the quantity and quality of conscious content, the
direction of attention, and the strength of psychical energy," and
that it cannot be assigned a maximal value but rather what he calls
an "optimal value." All that is included, in fact, in the complex
rubrics, _attention_ and _interest_, has to do with the length of this
indiscriminable interval.

Time-difference in consciousness is the very simplest thing in mental
life, for it is a case of the bare awareness of change. The elementary
time-judgment is mere judgment of change in content of consciousness.
In the experiment where one is asked to say which of two expected
stimuli comes first, however, the case is already complicated.
There must be a double preparation to react and to note the change
characteristic of each case, and so convert it into a time-judgment. In
the combination of two judgments, there is the same double expectancy,
preparation to react in two ways at once. In each experiment, the
preparation and shaping of expectation is the same as in reaction
experiments. In all reaction work, the short reaction comes as the
result of catching the attention wave at its most favorable point.
If the signal to react catches the idea of reaction in the mind of
the observer at the very focal point in consciousness, the shortest
reaction possible under the given conditions results. So in both the
combination experiment and the interval discrimination experiment, it
is very necessary to catch the attention wave, _equally prepared for
both or all the processes_, and at the highest crest of advancement.
Both demand the same preparation as a compound reaction. I believe it
is this inequality of balance of the attention between the various
processes that is responsible for the interference which is evidenced
in my results. This is my explanation of the appearance of impaired
accuracy for combinations for a given observer under some conditions
and the failure of any sign of impaired accuracy for the same
observer under other experimental conditions, or even under the same
experimental conditions at different times.

In the time-interval discrimination experiment the evenness of
balance in the attention wave will make for the shortest interval
discrimination, and the proportion between the two will be direct,
so far as other factors do not interfere. But there are special
interferences here. One of these is the fact that the two mental
processes do not set off at the same moment. No matter how even the
balance in attention at the moment of impact of the first of the two
stimuli, the preparation for the other, not yet set off, cannot be held
in equal readiness while this is going off. This discharge has already
disturbed the preparation to discharge in the other direction. In the
case of a given pair of stimuli of definite qualities and intensities,
the relation will be one of mutual facilitation for one interval of
separation and one of inhibition for another interval. In one case
the first opens the path for the second, being a case similar to the
summation of stimuli, and in the other, it draws all the available
energy in its own direction.

FOOTNOTES:

[Footnote 108: W. Wirth: Zur Theorie des Bewusstseinsumfanges und
seiner Messung, Philos. Studien, vol. 20, p. 487, 1902.]

[Footnote 109: Cattell: Ueber die Trägheit der Netzhaut und des
Sehcentrums, Philos. Studien, vol. 3, p. 94, 1886.]

[Footnote 110: Gonnessiat: L'Equation personnelle, Paris, 1892.]

[Footnote 111: Exner: Experimentelle Untersuchungen der einfachsten
psychischen Processe, Archiv. f. d. gesammte Physiologie, vol. 7, p.
601, 1873, and vol. 11, p. 581, 1875.]

[Footnote 112: Angell and Pierce: Researches upon Attention, American
Journal of Psychology, vol. 4, p. 528.]

[Footnote 113: Pflaum: Neue Untersuchungen u. d. Zeitverhältnisse der
Apperception einfacher Sinneseindrücke, Philos. Studien, vol. 15, p.
139.]

[Footnote 114: Peters: Aufmerksamkeit und Zeitverschiebung in der
Auffassung disparater Sinnesreize, Zeitschrift f. Psychologie, vol. 39,
p. 401, 1905.]

[Footnote 115: Paulhan: Revue Scientifique, vol. 39, p. 684.]

[Footnote 116: Jastrow: American Journal of Psychology, vol. 5, p. 239.]

[Footnote 117: Loeb: Archiv. f. gesammte Physiol., vol. 39, 1886.]

[Footnote 118: De Sanctis: Zeitschrift f. Psy. u. Physiol. d.
Sinnesorg., vol. 17, p. 205.]

[Footnote 119: Münsterberg: Willkürliche und unwillkürliche
Vorstellungsverbindung, Beiträge zur experimentellen Psychologie, vol.
1, pp. 64-188.]

[Footnote 120: Wundt: Physiol. Psy., 5th ed., vol. 3, p. 351.]

[Footnote 121: Wirth: Zur Theorie des Bewusstseinsumfanges und seiner
Messung, Philos. Studien, vol. 20, p. 487, 1902.]

[Footnote 122: Münsterberg: Grundzüge der Psychologie, vol. 1, p. 525.]




THE ESTIMATION OF NUMBER

BY C. T. BURNETT


I. There are situations not a few in life in which we find ourselves
estimating the number of objects in some group. Sometimes we desire
to know merely whether the group is large or small. Sometimes we try
to reach an absolute number that shall approximate roughly to the
real number. Sometimes, again, we only care to know whether the group
in question is more or less numerous than some other group that we
have before us or perhaps recall in memory. The public speaker finds
himself wondering whether this present scattering audience is larger
than the one that last night crowded into the front seats. The farmer
riding between adjoining orchards judges roughly the prospective yield
by a comparative estimate of the fruit in sight. The politician too
has an interest that is very notable indeed in such rough numerical
estimates. He asks himself, for example, whether the voters will be
more influenced by reports favorable to his party sent in from numerous
small towns or by such reports from a few large centres. Or perhaps he
is planning a demonstration in favor of his candidate. His problem then
is so to arrange his procession that five hundred men will look like
five thousand. Turning to another field, how is it that the enrolment
in some institutions of learning seems larger and the size of the
faculty more portentous than in other similar institutions that are
really of about the same size?

These examples bring to mind our interest in rough numerical estimates
and at the same time suggest the probability that we are swayed back
and forth in these estimations without ever a numerical difference
occurring in the objects of our judgment. These considerations lead us
on, then, to an enquiry about the factors that can thus influence our
estimation of number.


II. INFLUENCE OF FACTORS IN THE SAME SENSE-FIELD AS THE OBJECTS WHOSE
RELATIVE NUMEROUSNESS IS IN QUESTION.

The experiments described in the following pages are concerned with the
influence exerted on the judgment of a given factor by other factors
presented at the same time. The object of judgment in these studies is
visual number, which is to be submitted under varying conditions of the
objects whose number is in question, for example, varying conditions
of form, size, distribution, with the intent to discover whether
this judgment is a function of these other factors as well as of the
numerical. The scope of the enquiry includes both relative and absolute
number.

The objects chosen as a basis for the number-judgment were bits of
paper pasted in two well-defined groups side by side upon a background
of black cardboard. This card fitted into an upright frame where it was
held in place by a pivoted spring, which allowed easy adjustment and
removal of the card. The opening of the frame, 15×20 cm. was concealed
at will from the observer by a black wooden screen that played up and
down on guiding posts, when released by a cord and lever from the
catch that held it in place before the card. It fell by gravity upon a
cushion that deadened the sound; and it was restored to its position
by the operator's thrusting his fingers beneath and lifting it till
the catch above caught and held. The entire apparatus, as well as the
operator's movements, was concealed from the observer by a large black
cardboard screen resting upon a black-covered table. The one opening in
this screen was just large enough to allow a full view of the card when
the inner wooden screen fell from sight.

This apparatus which we will call the Two-Group Apparatus, admitted of
simultaneous exposure of the two groups of objects, and that only. At
first, to make successive exposure possible, a light wooden frame was
constructed in whose grooves two leaves of black cardboard ran like
sliding doors. By means of rods fastened to their outer edges these
leaves could be pulled apart or thrust together till their inner edges
met. When this apparatus was placed between the outer screen and the
frame bearing the card, and the inner wooden screen had been dropped
out of the way, this substitute divided screen was sufficient roughly
to accomplish the end in view.

With this apparatus the illumination was daylight, coming through a
very large window at the back of the observers. By means of a curtain,
marked variations in the light could be prevented.

For the length of simultaneous exposure of the groups the following
rule was adopted: Each observer was to be allowed time enough to get a
satisfactory feeling of relative number, but not time enough to admit
of counting. This time was kept constant during the work of any one
sitting. As the weeks went on, it was found possible, under the rule
laid down above, to shorten the time for some of the observers, and
to use with all the same length of exposure that had sufficed for the
speediest. The range of variation was from 1.2 sec. to 1.6 sec. Time
was measured by the ticks of a watch. Later tests showed for the time
studied that, where effective at all, the longer exposure diminished a
given tendency. Often it had no apparent effect.

The method of control already described is not only rather rough but
does not exclude the possibility of a space error. This possibility
proved actual by experiment. So an apparatus was contrived that should
present the groups in succession at approximately the same place and
should shorten the exposure, if desirable, to a small fraction of a
second.

This new apparatus, which we will call the One-Group Apparatus,
required artificial light and a dark room. By means of a 125 cp.
incandescent electric lamp, images of the groups of objects were
reflected through the lens of a camera and came to a focus upon its
ground-glass screen. A second screen of ground glass was placed in
front of the first and as close to it as possible, that an even
distribution of light might be obtained. The cards containing
the objects were of the same general character as in the earlier
experiments. They were held in a moveable slide whereby each group in
succession could be brought before the lens. When the slide was drawn
to the limit in one direction a single circle appeared in a black
field. This circle was used as a signal and a means for directing
the eye in the dark to that region where the groups were to appear.
The exposures were made with a camera bulb, the shutter being set
for instantaneous movement, with diaphragm 22 and length of exposure
1/26 sec. A shorter time was thought on trial to make perception too
difficult. The apparatus rested upon a table of special construction
and was enclosed as far as the glass screen with a wooden frame covered
with denim. Double curtains of this material formed this enclosure on
one side and made possible an easy adjustment of the cards between
exposures, as well as the admission of the operator's hand during a
given experiment for the adjustment of the shutter. This had to be
set, of course, before each of the three exposures constituting one
experiment. During its progress the hand was not removed at all, the
curtains falling about the arm in such a way that little light escaped.
The other hand managed the moveable slide from behind the enclosure.

Time was measured by watch-ticks. The three exposures--dot-signal,
Group 1, Group 2--were separated from each other by intervals of 1.6
sec. This was fixed upon as the minimum for convenient operation of the
apparatus.

In much of the experimentation on relative number two observers were
employed at once. Their chairs were placed closely side by side on
a line about 150 cm. from the plane in which the groups appeared.
These groups were not very far from being on a level with the eye.
Each observer recorded his own judgment, against the number of that
experiment. There were three possible kinds of judgments,--equality or
either group larger. If the judgment was of difference it was recorded
in terms of the larger.

When the dark room was used, special arrangements were required, for
convenience of the observers in making their record. After several
schemes were tested the following was adopted as least trying to their
eyes: A large, black-topped table was placed before them, bearing an
electric lamp enclosed in a black box with a small aperture that could
be closed at pleasure; or, if left open, did not let enough light
escape to disturb the perception of the groups.

The absolute number of objects in the groups was determined, first, by
the character of the problem, and then by convenience. If we are to
learn anything about the influence exerted upon the number-judgment
by other factors than the numerical, we must eliminate all influence
of the latter. Correct judgments may be determined by this factor
alone; erroneous judgments must have been otherwise conditioned; and
these conditions it is the task of our method to isolate and study, as
modifying factors. From correct judgments we learn nothing definite
about our problem, but from erroneous everything. Other things being
equal, it is preferable to eliminate from the results the influence
of this numerical factor, just as one handles any other disturbing,
unavoidable element, by equalizing the numbers in the two groups.

What may be called the standard number of objects in each is twenty.
This choice was governed by the purpose of using a number large enough
to make counting impossible in a brief time and yet not so large as
unnecessarily to increase the labor of preparation and the difficulty,
for the observer, of getting an idea of the groups as a whole. To the
cards containing equal groups, 20 to 20, were added others, 20 to 19,
19 to 20, for the purpose of easy variation in arrangement, by omitting
one object from a group, without making the actual numerical difference
easily perceivable. In later work these small objective differences
were dropped. Yet other cards, 23 to 17, 17 to 23, were added, to the
end that the observers might find unmistakeable number-differences, and
so not be bothered by the suspicion that the groups were all equal.
The reversal of the number-relations, as indicated above, was in the
interest of equalizing the influence of the actual numerical factor in
the two groups.

The following proportion was kept among the numbers of observations
made upon each kind of card: 1/2 upon groups objectively equal; 5/12
upon those differing by one from each other, where half each went to
(20 to 19) and (19 to 20); 1/12 to those showing the maximum objective
difference of six, where again half went to (17 to 23) and half to (23
to 17). Of course the observations upon cards of this last sort are
excluded from the tables.

As to the number of cards employed for each series of experiments,
it was found at first convenient to use seven,--3 (20 to 20), 1 (20
to 19), 1 (19 to 20), 1 (17 to 23), 1 (23 to 17). In each group the
arrangement of objects was irregular. The use of three of the first
sort was to encourage freshness of judgment, each having its particular
irregularity. Cards were but rarely remembered, practically never
except in the case of groups differing widely in number. So far as the
observers could tell, judgment was formed afresh in all these cases.
In later experiments eight cards were used. This number was in the
interest of avoiding the distribution-error. At first it was thought
sufficient that all the groups should be merely irregular. Later it
became evident that discrimination was very fine here and so that this
factor must be eliminated by the usual precise method.

The space- and time-errors, where likely to be present, were eliminated
in the usual way by performing an equal number of experiments with the
groups in reversed arrangement. Several methods of doing this were at
first tried; but these were all abandoned in favor of the following:
The experiments were arranged in sets of 24, in each of which the
proportion of kinds of cards was kept as indicated above. Each set
with one space- or time-order of the groups was repeated with that
arrangement reversed.

A word must be added as to the arrangement of results in the tables.
Judgments of equality upon objectively unequal groups are entered
as overestimations of the smaller groups. The per cent of correct
judgments is equally divided between the two other classes, and for
this reason that interest centres, not in correctness at all, but in
the difference between the tendency of error in one direction and that
in the other direction. No doubtful judgments were admitted, but in
such cases another trial was allowed later, usually when the observer
was not aware that he was being given a new chance. The subjects are
divided into three classes according as the results show a tendency to
favor one or the other group or no tendency either way. A difference of
10% is arbitrarily taken as significant.

1. _The Influence of Group-Area._ The Two-Group Apparatus was employed.
The four sets of experiments carried out with this factor differed
primarily in the material upon which the observer's judgment was based,
and secondarily in certain matters of method. The attempt in them all
was to approximate more completely to the isolation of the factor
under investigation. They are numbered in the order of approximation.
As marked results were obtained from each, they have all been offered
for consideration in the four parts of Table I. A description of the
material used in each case follows.

_A._ Squares (1 cm.) Neutral Gray no. 1. (Bradley), arranged
irregularly in two groups with irregular outlines on a background of
black cardboard. One group was large in area, the other small, the
attempt being made to fill each space homogeneously. Groups were not
proportional in shape of area.

_B._ As above, save that circles (11 mm. approx. in diameter) were
substituted for squares, in the interest of distinctness for the
several objects.

_C._ The area of the groups was oblong and regular, and the sides
were proportional. (Compact 72.5 mm.: 58 mm.; scattered 110 mm.: 88
mm. These relations were determined by the size of the frame that had
already been used and by the desire to make the difference in area
as marked as other necessary conditions would admit.) Each area was
marked by a circle in each corner. The color of the compact group
was the deepest shade of normal gray (Prang Normal Gray Darker); of
scattered group the next higher shade (Normal Gray Dark). These dark
grays were used in order to reduce to a minimum the tendency to produce
after-images. The difference in the shades of the two groups was in the
interest of avoiding the greater brightness due to the mass-effect of
the compact group.

_D._ As in C, except that India ink outline circles (1/3 to 1/2 mm.
line) were used on a background of granite cardboard. This change was
made to avoid, as far as possible, the greater mass-stimulation due to
the reënforcing effect of the compact arrangement. The size of circles
remained as before.


TABLE I

    KEY:  NS    = Number of Subjects
          AV%   = Av. % of difference in favor of

                     A               B               C               D
                    274             248             120             132
              _experiments_   _experiments_   _experiments_   _experiments_
               _with each_     _with each_     _with each_     _with each_
                _subject_       _subject_       _subject_       _subject_

               _NS_   _AV%_    _NS_   _AV%_    _NS_   _AV%_    _NS_   _AV%_

    Small        5    31.5      10    34.6       7    44.1      10    46
    Large        3    35.1       4    44.1       5    41.5       3    26.3
    No tendency  1     8.8       2     4.2       4     5.7       3     6.6

  The per cent recorded in the no-tendency class is an average of _all_
  per cents below 10, whether in favor of the one or the other of the
  two remaining classes. This is true for all the following tables.

The following facts are presented by the several parts of Table I: (1)
The large per cents of difference show that area is to a large extent a
determinant of the judgment of relative number. (2) Different subjects
show opposite tendencies. (3) A comparison of the results of individual
subjects through the four series shows that this opposition in tendency
occurs in the same subject at different times. The introspective notes
of one of these subjects show the internal process of change from one
tendency to the other. It consists in a gradual increase of analytic
activity toward the compact group. At first glance the composition of
the scattered group was more evident; but when attention was fairly
turned toward the compact, the inability to isolate objects made them
seem very numerous. The importance of a coöperating subjective factor
is here evident. (4) Out of a possible 57 cases there are but 10
showing no tendency.

2. _The Influence of the Internal Arrangement._ As before, the
Two-Group Apparatus was employed; and the factor was studied in three
aspects.

_A._ The material consisted of two groups of gray circles (Normal Gray
Darker, Prang) covering equal areas. In one group this area was filled
homogeneously, in the other the circles were gathered into nuclei. In
order that there might be exactly the same relation of parts when the
cards were reversed, each group was so arranged on a diagonal axis of
symmetry from upper left to lower right corner that each half repeated
the other in reverse order. Otherwise the arrangement of circles was
irregular.

Six cards only were used,--four (20 to 20), one (17 to 23), and one
(23 to 17). Slight differences among them occurred in the arrangement
of the equality-cards, which might help to counteract any incipient
reasoning from sameness of appearance to sameness of number. The large
increase in the difference-values is accounted for in part by the fact
that the cards (19 to 20) and (20 to 19) were omitted, and for this
reason: that when an observer tended largely to favor a particular
group, the introduction of a card in which that group was objectively
greater would mean an increase in the number of correct judgments;
whereas the introduction of two objectively equal groups for the others
would increase considerably the number of erroneous judgments.

_B._ The numerical character of the cards here used shows a return to
the usual. The material was like that of _A_, except for a new internal
arrangement. Here the area of one group was filled homogeneously, while
that of the other contained a pattern of this sort: An ellipse just
contained within the boundaries of the normal area; a circle in each of
the four corners of that area; and in the centre a diamond formed of
four circles. Numerical changes were always confined to the ellipse, as
less open to counting than the rest of the figure.

_C._ Material and method repeat _B_ but with another internal
arrangement. One group, as before, showed an area homogeneously filled,
and irregularly, as usual. The other group carries to an extreme the
distinction of open and filled space made prominent in the other groups
of this table by massing the circles in an outline completely enclosing
the area and in a diagonal from upper left to lower right corner. The
outline did not show even spacing; more circles were crowded in one
part than in others, that counting might be more difficult.

That there might be no attempt to remember cards, in all cases where
there were twenty objects in the homogeneous group the same kind of
irregular arrangement was repeated. This is a different method from
that employed in _A_. Since the length of exposure was so short and the
arrangement in the group irregular, either one is probably as good as
the other.

_D._ The material used for _B_ and _C_ had a kind of regularity,
since definite patterns were used. The introspection of the observers
showed, however, that the patterns as such were not in question in
the judgment, but rather the vacancies and the crowding. With the
Two-Group Apparatus an arrangement in parallel lines could rather
easily be counted, but with the One-Group Apparatus and its means for
instantaneous exposure this difficulty was to some extent overcome. The
arrangement of the objects in parallel lines was therefore adopted
and matched against the irregularity of an accompanying group. The
same size of group-area was kept, but the small difference-cards were
omitted. There was no other change beyond those made necessary by the
apparatus and already indicated on an earlier page. The length of
exposure was 1/25 sec. The bearing of this time-factor on the results
will be considered in a later section.


TABLE II

    KEY:  H  = Homogeneous
          N  = Nucleated
          P  = Pattern
          O  = Outlined
          R  = Regular
          I  = Irregular
          NT = No tendency

                     A               B               C              D
                    132             132             132             88
              _Experiments_   _experiments_   _experiments_   _experiments_
                _with each_     _with each_     _with each_     _with each_
                 _subject_       _subject_       _subject_       _subject_

                 H    N   NT      H    P  NT      H    O  NT      R   I  NT
  Number of
   subjects     10    2   2      10    1   3      7    5   2      2      1
  Av. % of
   difference
   in favor of 53.23 1.5 6.1    35.7 52.8 5.3   42.7 34.8 7.9   40.3    1.2

The several parts of Table II give us the following facts: (1) The
judgment of relative number is very markedly a function of the internal
arrangement. (2) The marked tendency among the observers to favor
the homogeneous in _A_ and _B_ meets a check in _C_. Recalling the
direction of difference between _C_ and the other sets, that in _C_
the gradually increasing contrast between the inner vacancies and the
filling reaches a maximum, we may suspect that these vacancies begin
to seem no longer a part of the group-situation, while the compactness
of the filling, where it does occur, is thrust prominently forward.
The notes of the observers confirm this suspicion. (3) The results of
the different subjects show that the shifting of tendencies occurs
as before. (4) As to the way in which regularity functions in the
judgment, the notes of one observer are very clear. The blank spaces
in the irregular are noticeable, he says, which is not true of the
regular, where, on the contrary, one has a feeling of compactness
of figure. I am able to confirm this character of the spaces by my
experience outside this experiment. A simple pattern is very easily
apprehended and irrelevancies of the background dismissed. Increase its
complexity to a maximum, as in the case of an irregular group, and I
am almost at a halt to isolate objects from their fellows and maintain
them apart, yet together. The background is hardly to be shut out. This
is probably due to the absence of a centrally excited image of the
group. The object and the not-object run together. (5) The position
of the single observer in the no-tendency class of _D_ was marked
subjectively by great difficulty in forming a judgment. The groups
seemed incomparable, the vividness of form excluding the perception of
number.

3. _The Influence of Complexity in Group-Composition._

Complexity of group-content was attained by introducing objects of
different colors; so there was not a clean isolation of factors. By
comparing these results with those recorded in Table IV, A, we shall be
able somewhat roughly to make allowance for the factor of mere color.

Sets of 132 experiments each from sixteen observers were obtained for
each of these factors. Unfortunately the distribution-error was not
eliminated. Later experiments showed the importance of this factor,
and, in consequence, the impossibility of interpreting the results
under consideration. So new experiments were performed under the proper
conditions, but at a time when only a few of the first observers could
be used. Their results from the earlier series are given in Table III,
A. The exclusion of the small-difference cards from the later series
(Table III, B) and the consequent increase of the number of experiments
on objective equality prevent comparison.

The material in _A_ consisted of two groups of circles of the usual
size, one Normal Gray (Prang), the other of three colors--Red, Yellow
Orange Shade 2 (Bradley), Light Blue Blue Green (Prang). The intent
was to equalize the two groups in brightness. When the observers were
questioned about the relative brightness, supporters were found for all
three possible opinions. So it seems probable that the groups did not
differ widely in this respect. As nearly as the number-condition would
allow, the three colors were represented equally in the group; and they
were distributed so as to make the whole as homogeneous as possible.

In _B_ the changes were the correction for distribution as described
in the introduction to this section, and the substitution of
equality-cards for those of slight numerical difference. In addition,
three other colors replaced those of _A_, in the interest of regulated
brightness and more pleasing æsthetic effect. These were, in the
Bradley system of broken spectrum scales, A-Red, medium; A-Yellow
Orange, dark; A-Blue Green, dark. With these exceptions _B_ was like
_A_. The observers were all inclined to consider the gray brighter than
the mixed. The Two-Group Apparatus was used.


TABLE III

    KEY:  G  = Gray
          MC = Mixed Colors
          NT = No Tendency

                           A                   B
                          132                 132
                     _experiments_       _experiments_
                      _with each_         _with each_
                       _subject_           _subject_

                     G    MC   NT        G    MC   NT
    Number of
     subjects        3         1         3         1

    Av. % of
     difference
     in favor of    17.7       2.2      19.9       6.8

The following facts may be gathered from Table III: (1) The tendency
to overestimate the gray is due in part at least to the additional
factor of complexity in the other group, as is shown by the markedly
changed tendencies in Table IV, A, where a solid color takes the place
of the mixed colors. The actual colors involved in the two cases are
different, to be sure, and necessarily so, and this difference may of
course be invoked as the cause, as well as a possible difference in
brightness between the gray and mixed in III, B. The introspective
notes help us here. One observer felt that he favored the gray
primarily because there was a tendency to consider but one color in the
mixed. Another was drawn toward the gray because it seemed definite and
consistent. For both of these observers æsthetic elements were involved
in favor of the gray. The latter found also that the greater brightness
of the gray gave it a larger area. With a third subject the fact of
variety was felt as decidedly important; but his notes show a conflict
between this factor and that of distribution which was the conscious
basis for his normal judgment.

4. _The Influence of Differences in the Kind of Objects._

_A. The Factor of Color._ The material in A 1 consisted of two groups
of circles of the usual size, one Normal Gray (Prang), the other
Red (Bradley). The attempt was made by this choice to equalize the
brightness. The size and shape of the group-area were those of the
smaller area of Table I, C and D. In A 2 the only changes were the
correction for distribution, as described in the introduction to this
section, and the substitution of equality-cards for those of slight
numerical difference. The Two-Group Apparatus was used.

The following results appear in Table IV, A 1 and A 2: (1) While
complexity seemed on the whole to diminish apparent number, red
noticeably increases it. Some of the observers report that group as
more vivid and interesting. One observer compensated by emphasizing the
gray in attention, as his results showed. (2) If one ask how the color
red functioned in the judgment, the reply must apparently be, by its
brightness and vividness. The mixed group functioned in a double way,
as vivid and so more numerous, as fragmentary and so fewer.


TABLE IV

               A1                                     A2
     132 _experiments_                      132 _experiments_
     _with each subject_                    _with each subject_

            _Gray_  _Red_  _No tendency_  _Gray_  _Red_  _No tendency_
  Number of
  subjects     3         1                   3          1

    Av.% of
    difference
    in favor of          18.3      5.4                 30.3        9


                           B                                    C
                    132 _experiments_                    132 _experiments_
                   _with each subject_                  _with each subject_

            _Large_ _Small_ _No tendency_ _Circles_ _Squares_ _No tendency_
    Number of
    subjects     8       3         5           8        4         4

    Av.% of
   difference
   in favor of 33.6  27.5        6          28.4     18.8       3.5


                         D1                                     D2
                 88 _experiments_                     132 _experiments each_
                _with two subjects_

                _44 experiments_
                _with two subjects_

                _exposure_ = 1/25 _sec._                _exposure_ = 1/4 _sec._

           _Simple_ _Complex_ _No tendency_ _Simple_ _Complex_ _No tendency_
    Number of
    subjects            3           1                     2

    Av.% of
   difference
   in favor of         30.7        5.6                   21.2


                          E
                    88 _experiments_
                   _with each subject_

                   _exposure_ = 1/25 _sec._

                _Bright_   _Dark_   _No tendency_
    Number of
    subjects                  3

    Av.% of
   difference
   in favor of               47.4

_B. The Factor of Size._ The Two-Group Apparatus was used, the material
consisting of India ink circles (1/3 to 1/2 mm. line) on a background
of granite paper. This paper was chosen here and for the experiments
of Table I, D, to get a suitable mean between too sharp contrast and
sufficient distinctness. The circles in the one group were ten mm.
in diameter; in the other seven mm. The two areas were approximately
equal, and of the same size as that of the more compact group in Table
I. This is in fact the standard size throughout these studies in
Relative Number, wherever area is not in question. The small-difference
cards were included.

Two sources of possible complication must be considered. It is
unavoidable that the factor of differences in compactness should
enter and that clean results on the basis of object-size be denied.
Our interpretation must not fail to consider this fact. Because of
this, it seems unlikely that a distribution-error should arise; so
the usual precaution to eliminate it was omitted both here and in the
study of area (Table I). Distribution affects the appearance of the
vacant spaces. When differences in the _amount_ are by the conditions
inevitably prominent, differences in the _conformation_ may be safely
regarded as of minimal vividness.

The following results appear in Table IV, B: (1) The illusion of
numerical inequality is marked for many subjects. (2) The judgment
is quite possibly a function of the two factors--object-size and
group-vacancies. If we recall the fact that the small-object group
is more scattered than the other, we shall note that the leading
class here is like the leading class in Table I, and we may fairly
reckon this factor as of importance in the issue. Of the incomplete
introspective notes on this question, those of only one observer speak
clearly for the size. He says: "There is an overpowering feeling of
predominance in case of the large and I must judge for them. The large
space covered seems an important factor. The longer I reflect upon
the relative numbers the more numerous seem the larger, that is, they
appear to increase over the small after the exposure. It is hard to
give judgments of equal in most cases."

_C. The Factor of Form._ The material consisted of a group of circles,
each of the same size as in former material; and a group of squares,
each approximately equal to a circle of the other group. These were
made of Prang's gray paper (Normal Gray Darker) and pasted upon a
black background. The areas of the two groups were approximately equal.
The squares were set irregularly except for those in the corners, where
the edges were placed parallel to the edges of the card. The Two-Group
Apparatus was used and the small-difference cards included.

In this material, again, the formal elimination of the
distribution-error was not attempted. The striking difference in
the conformation of the vacancies through the form-differences of
the objects probably makes the repetition of the exact positions
insignificant. Still the fact must be noted.

The results appear in Table IV, C. (1) The illusion of numerical
inequality is here again marked for many observers. (2) The
introspective notes are not on the whole very illuminating as to the
basis of judgment. One observer, who favored circles, found that the
appearance of more orderly arrangement in squares made them seem
few. Another, who favored squares, found, on the contrary, the more
regular the more numerous, and thought that the squares may have seemed
more regular. A third, who favored circles, found the squares better
individualized, with whom a fourth agreed in both respects, who also
was influenced by the apparently greater bulkiness of the squares.
Fewer could go into a given area. A fifth, on the other hand, who
found the circles better individualized, still favored them. So we
have these observers apparently doing the same thing under opposite
conditions, and the opposite thing under the same conditions. Here
indeed is a situation for any theory. So far as we can learn from the
foregoing, the form may influence the judgment merely through its
space-characteristics, but possibly also through the vividness of
intrinsic interest.

_D. The Factor of Complexity._ The One-Group Apparatus was used in this
work and results were obtained for two different lengths of exposure,
1/25 sec. and 1/4 sec. The material differed, in that to the centres
of the circles of one group were added small Red (Bradley) circles (6
mm.). With the apparatus used, the color was not very effective, the
brightness contrast between dark centres and white periphery being
chiefly prominent. The total group-brightness was of course diminished
by those centres. The small-difference cards were omitted.

The results are recorded in Table IV, D 1 and D 2. (1) The illusion
is apparently strong. (2) The amounts of the difference-values show
that the shorter exposure is more favorable to the illusion. (3) The
introspective notes indicate that both brightness and complexity
functioned in the judgment. Two observers, both of whom show large
tendencies, were not conscious of any influence of complexity. One of
these did find differences in brightness important; and in favoring
the darker group his results exactly coincide with those of Table IV,
E, where this factor is under direct consideration. A third found the
complex group interesting. With a fourth the complex group developed
in number amazingly during the few moments after exposure and had an
appearance of great intricacy, often seeming to be in active movement.
A fifth observer too felt that its numerical character depended on its
complexity.

_E. The Factor of Brightness._ Hitherto the absolute arrangement of
the objects in any two groups compared, where this factor has not been
the object of enquiry, has been in the two cases different, though
with respect to irregularity alike. This course was governed by a
desire to avoid the substitution of a form-judgment for one on number,
through recognition of the fact that both groups had identical forms.
The resulting distribution-error I tried to eliminate in the usual
way. Tests toward the end of these studies showed that there was no
danger from this source. Errors seemed about as frequent as before.
No observer made any comment on the fact, except one who through his
official connection with the laboratory work knew that the test would
be made sometime, but not exactly when. During many of the experiments
he did not perceive the likeness of form; and when he did the numerical
judgment arose without connection with that factor, as was shown by
the feeling that the two groups were unequal in number. He called the
relative fewness of the first group a case of "perspective effect."
This must have significance for any account of the time-error; but by
no means carries with it its own interpretation.

One welcome result of these tests was their assurance that I might
without fear further simplify the experimental conditions by avoiding
the possibility of a distribution-error. The material for these
experiments on brightness therefore profited by this possibility. Each
card had a different specific irregularity, but always in duplicate. In
choosing the degree of brightness-difference Prang's brightest shade of
normal gray was found as dark as could be conveniently perceived with
the artificial light of the One-Group Apparatus. The contrast between
this shade and white was quite evident enough for the purpose. The
small-difference cards were omitted.

Table IV, E, shows the decisive character of the results. The observers
fall all into one class in favoring the darker group, and by a large
difference-value. The following introspection of one observer shows
the extent to which the factors of brightness and number fuse: "I
frequently lose sight of time-order. It is a question of number and
not one of light-intensity, and if called upon to state which group
came first I might not be able to answer. In equality-judgments the
difference of light comes out distinctly."

5. _The Influence of Complexity of Environment._

The material prepared for these experiments certainly lays stress
upon _relative_, not _absolute_, complexity; for the conditions were
satisfied by placing 5 mm. strips of white paper, equal in length
to the width of a group, a few millimetres off at the top and the
bottom of the groups that were on one side of the cards. The One-Group
Apparatus was used and the small-difference cards omitted.


TABLE V

    44 experiments with two subjects.
    88 experiments with two subjects.

    Exposure = 1/25 sec.
                           _Simple_        _Complex_        _No_
                         _environment_   _environment_   _tendency_

    Number of subjects                         2             2

    Av. % of difference
      in favor of                            15.9            8.5

The results are recorded in Table V. (1) The drift of tendency is
toward the group with the more complex environment. No one markedly
favors the other group. (2) The notes of the observers indicate that
the added strips functioned through their effect upon the apparent area
of their group. The observers all found the dimensions increased; but
with some, apparently by contrast, the added height brought out sharply
the narrowness. One observer found this true in general; another,
when the barred group came first. The latter says: "The unbarred
group, coming first, appears to reflect its compact character on the
barred one, when it comes, so that it does not look so attenuated and
strange." Here the image brought over to the second took the width of
the second somewhat out of relation to its illusory height, whereas in
the reverse order the contrast relation was fully maintained.


III. THE INFLUENCE OF FACTORS PRESENTED IN OTHER SENSE-FIELDS BY THE
OBJECTS WHOSE NUMBER IS IN QUESTION

A very simple apparatus was employed. The objects whose number was in
question were bright steel balls (3/8 inch) thrown loosely into square
black frames, 13 cm. inside, placed side by side on a black-topped
table. The experiments were performed in series of 30. The groups were
kept equal numerically, with this exception, that into each series were
introduced four experiments where the groups were so unequal that the
observer could have no question as to the correctness of his judgment
and the existence of objective differences. This numerical superiority
was given to each group alternately, and judgments on it were, of
course, excluded from the results. The actual numbers employed in a
series varied between 35 and 60 in accordance with the following scheme:

    1.   50 each
    2.   45   "
    3.   50   "
    4.   55   "
    5.   60 to 40
    6.   50 each
    7.   45   "
    8.   40   "
    9.   45   "
    10.  50   "
    11.  55   "
    12.  60   "
    13.  40 to 60
    14.  50 each
    15.  45   "
    16.  40   "
    17.  35   "
    18.  40   "
    19.  45   "
    20.  50   "
    21.  55   "
    22.  60   "
    23.  60 to 45
    24.  50 each
    25.  45   "
    26.  50   "
    27.  60   "
    28.  55   "
    29.  50   "
    30.  45 to 60

The time of a single exposure--in this case two groups at once--was
3 sec. measured by a stop-watch. As to the arrangement of the balls,
care was taken that they should not be massed in one place, but
scattered somewhat homogeneously over the space within the frames. The
illumination was daylight, so managed that shadows cast by the balls
were reduced to a minimum. The observer sat close to the table with
the groups directly in front of him. He either kept his eyes closed
between experiments or held a small screen before them. Sometimes he
merely turned away. The operator worked from the opposite side of the
table, taking care to make the necessary noises as little suggestive
as possible. The observers agreed that they were not consciously
influenced by the manipulation.

The progress of these experiments disclosed an astonishing space-error.
So far as was conveniently possible the usual technique of elimination
was employed.

1. _The Influence of Active Pressure._

In this study the groups were differentiated in this way: With one hand
the observer rolled the balls of one group under his fingers, while the
other group was presented to vision only. The method of observation
consisted in rapidly and lightly rolling the balls under the fingers a
few times and then surveying both groups visually for the remainder of
the exposure, judgment being given on the visual number.

Evidently there is much that is rough about this procedure. Pressure
and kinæsthetic factors are lumped off together; the length of the
touch-stimulus was not exactly determined; and there is the possibility
that the visual stimulation from the group touched is weakened. To be
sure the method prevents any great difference in the latter respect;
and if we are guarded in our interpretation, something of interest may
be learned.

There appeared to be no convenient way to eliminate the space-error.
The right hand was used with the right group and the left with the
left. So here again interpretation must be circumspect.


TABLE VI

                         A                          B
               52 _experiments with_       260 _experiments with_
                  _each subject_           _one and_ 208 _with the_
                                              _other subject_

                      _No_     _No_                            _No_
            _Touch_ _touch_  _tendency_     _Uneven_ _Even_ _tendency_
  Subjects                        2                     1         1

  Av. % of
  difference
  in favor of                    7.6                    10        1

                 SPACE-ERROR

                                 _No_                          _No_
             _Right_  _Left_  _tendency_   _Right_  _Left_  _tendency_

  Subjects      1       1                     1       1

  Av. % of
  difference
  in favor of  69.2    23                    30.8    20.2

                    C
           145 _experiments with_
            _one and_ 260 _with the_
              _other subject_

                     _No Weight_   _No weight_  _No tendency_
   Subjects                                          2
   Av. % of difference
   in favor of                                       2.4

           SPACE-ERROR

                       _Right_   _Left_    _No tendency_
                                    1            1
  Av. % of difference
  in favor of                      32            4.4


Turning to the results in Table VI, A, we find the following: (1) The
influence of the pressure-kinæsthetic complex practically does not
appear; while the space-error shows a marked tendency that, for the
two observers, is in opposite directions. (2) On the other hand, the
notes of one observer show that in his case at least the face value of
the table is erroneous. To this effect he says in substance that he can
make a more accurate estimate of the number in the group touched. He
tries to ignore these sensations of touch, but with ill success in the
case of the left hand, where clumsiness not only makes it difficult to
touch the balls gently but also to keep them under the fingers, which
often feel the ground-space. For this cause the group seems small in
number. Clearly enough, then, it is the space-error that tells the
story of the effect of the added stimuli on this observer, only it
must not be interpreted as space-error. The pressure functioned in the
judgment through its numerical aspect. But the positive effect with the
right hand was turned to a negative with the left through its emphasis
of vacancies. The high difference-value in the space-column becomes
thus a striking evidence of the effect of pressure, and the results are
accounted for without reference to the kinæsthetic factor. The other
observer felt that the active pressure was relatively indifferent. (3)
The entire absence of correct judgments on the objectively equal groups
shows to what a surprising extent other factors have modified the
numerical.


2. _The Influence of Unevenness in Active Pressure-Feeling._

In the experiments of this section the groups differed in this way,
that one rested on the smooth table-top while the other had for its
bottom a coarse wire mesh covered with black cloth. The balls of the
one rolled smoothly beneath the fingers while the other balls moved
lumpily over their mesh. Both hands were used--each for the group on
its side; and the method of observation and length of exposure agreed
with those conditions in the preceding section, except that the balls
were rolled a little more vigorously that the factor studied might
come clearly into consciousness. The groups were interchanged for half
the number of series. This could not of course completely eliminate
the space-error, since kinæsthetic differences in the limbs remained
uncompensated. In general the criticism in the preceding section is
again applicable.

The results appear in Table VI, B. (1) One observer shows a tendency
to favor the smoothly rolling group, while the other again shows no
tendency. Both have large space-errors of the same character as in A of
this table. (2) The introspection of the observer showing no tendency
is to the effect that touch plays little or no conscious part in the
situation. The other's notes give no hint that the factor studied
here was influential; but to the effect on the judgment of touch in
general, especially with the right hand, they give clear witness. The
touch-sensations, he says, were difficult to ignore. Those from the
right hand were more vivid than those from the left; and the right
hand seemed more sensitive. Judgment was based on a general feeling of
"moreishness" which came promptly. There is nothing to contradict the
evidence of the earlier experiments that touch is again influential
through its numerical character. (3) Both observers regard factors of
distribution as of fundamental importance, though one was inclined at
first to insist that there was nothing but number in his judgment. The
significance of this unanalyzed feeling will appear in a later section.
(4) These results agree with the preceding in the approximate exclusion
of correct judgments.


3. _The Influence of Active Weight._

The variation here in question consisted in lifting one of the groups
during judgment of the relative number in the two groups. The apparatus
was made by transforming into trays the frames containing the balls,
by putting into these frames wire-mesh bottoms covered with black
cloth. They were set each upon four small wooden pillars so that the
hand could be easily thrust under the tray. At the signal a given tray
was several times raised a little way and lowered, and the judgment
formed on the same factor as before. Here again the space-error was not
entirely eliminated. Each hand was used with the group on its side, but
kinæsthetic differences peculiar to each of the limbs remained. There
was always some motion among the balls in the lifted tray, though the
gentleness of the lifting prevented the existence of much. This is a
radical defect, but one not easily avoided with maintenance of other
desirable conditions. Even more serious, as the issue proved, was the
failure to control the lifting impulse; yet, as it happens, we are not
prevented from getting an experimental answer to our question.

The results are recorded in Table VI, C. (1) They show no apparent
effect of the weight, and with one observer the further unusual fact
of no space-error. This error is marked enough in the case of the
other, and, conforming in direction to that of the preceding sections
of this table, allows us in so far to adopt the same interpretation of
his results. (2) The introspection of one observer was to the effect
that he felt a tendency toward a modification of the number-judgment
by weight. It was especially strong when the group was lighter or
heavier than was anticipated, the light group seeming less numerous,
and the heavy group more. Occasionally he caught himself weighing the
second group mentally; and sometimes he had to recover himself from a
tendency to make judgments on a wrong basis, presumably that of mere
weight. With such a conflict of tendencies the character of the results
is not surprising. Particularly important are the opposing tendencies
lying in the factor of weight itself. The other observer reported that
a very heavy weight exerted an influence that it was hard but not
impossible to ignore, while a lighter weight did not effectively enter
the situation at all. His earlier inclination to say that there was
nothing but number in his judgment inclines one to believe that fusion
of factors may have passed beyond the stage of ready analysis. (3)
Our analysis has given us reason to believe that active weight has a
definite tendency to modify the judgment of relative number.


4. _The Influence of Muscular Strain in Observation._

The study was made from the point of view of more than one set of
experimental conditions, viz.:

    (1) Equal strain (minimum).
         (_a_) Right--left.
         (_b_) Up--down.
    (2) Equal strain (maximum) eyes turned.
    (3) Strain _vs._ ease.
         (_a_) Head and eyes turned.
         (_b_) Eyes turned.

The conditions of (1) (_a_) were exactly those of the earlier
experiments with the exception that the groups were undistinguished
save by position. In (1) (_b_) one group was so placed between the
other and the observer that there might be as little increased effort
as possible in viewing the farther. In the up-down movement more
muscles are involved in the lift than in the fall of the eye. So really
we have here a case of (3) but not so marked. In (2) the groups were
put to the right and left at such distances that, when sitting between,
the observer could just take each one in without turning his head. This
brought a decided strain upon the eye-muscles. In (3) (_a_) the groups
were separated by the length of the table--a distance of 90 cm.; and
the observer placed in alternate series before each; as he was in (3)
(_b_) where the farther group was carried to the limit of vision to be
reached without turning the head. Here the strain was like that in (2),
but for one group only.

An incompleteness in experimental analysis lies in the impossibility of
separating the factors of distance and strain.


TABLE VII

          A - Baldwin 46 experiments
              Hutchinson 78 experiments
              Equal strain (minimum)

                           B - 52 experiments
                               with each subject
                               Equal strain (minimum)

                                            C - 52 experiments
                                               Equal strain (maximum)
                                               Eyes Turned

                                                   D - 104 experiments
                                                        with each subject
                                                      Head and eyes turned

                                                         E - 52 experiments

                    A              B           C          D          E

              Right   Left    Lower Upper   R.  L.   Ease Strain  E.   St.

  Subjects   Baldwin Hutchison        2        Bald.         2   Bald.

  Av.% of
  difference
  in favor of 30.4    28.2           68.3      71.2        52.4  80.8

Here are the facts of chief interest: (1) The following tabulation
gives us a ready view of the character of the results in Table VII; and
shows the extent to which they are consistent:

                         A       B        C         D         E

    Baldwin favors     right   {upper    left   {farther   {nearer
                               {strain          {strain    {no-strain
                                                 left       left

    Hutchison favors   left    {upper           {farther
                               {strain          {strain
                                                 left

(2) The only inconsistency in the strain-distance complex is with
Baldwin in E. He reported that the more distant group appeared rather
as an undifferentiated mass whose number was not so well obtained,
while in the near the individuals were significant. He seemed to
be in the midst of these. The case seems analogous to that of the
observer whose introspection was reported under Table I, and who at
first accepted what we may call the objective analysis, by which
the scattered group gave up more distinct objects than the compact;
but later attempting voluntarily to disintegrate the compact, found
a bewildering confusion in the task that made this group seem very
numerous, and brought about in the end an exact reversal of tendency.
(3) Can we now separate in the results between the influences of strain
and of distance? So far we have regarded them as one complex. But the
introspections speak merely of the space-characters of the objects,
Hutchison agreeing with Baldwin that the more remote group is judged
as an area rather than as a collection of definite objects. (4) The
almost entire absence of correct judgments in these experiments adds
new evidence to that of the immediately preceding experiments in proof
of the insignificance of the actual numerical relation for the judgment
of relative number.


IV. THE INFLUENCE OF FACTORS OUTSIDE OF THE OBJECTS AND IN OTHER
SENSE-FIELDS

The One-Group Apparatus was employed, and cards in general
corresponding to those where area was not in question,--white-circle
groups equal in size and irregular in inner distribution, which was not
duplicated on the same card, though the resulting distribution-error
was formally eliminated in the usual way. The usual care was taken
to fill the group-area homogeneously. The small-difference cards
were retained at first; but on later discovering the possibility of
duplication a few supplementary experiments were added.


1. _The Influence of Touch._

The apparatus employed to give the touch-stimulus consisted in a long
lever attached to the armature of a small electro-magnet. In the end
of the lever was inserted at right angles a wooden peg, cork-tipped.
In view of the other conditions of the experiment a convenient spot
for the application of the stimulus was found to be the forehead where
it curves backward above the right eye. The apparatus was supported
by rods and clamps upon a long upright steel rod set in an iron base
and placed behind the chairs of the observers. The same rod carried a
head-rest, designed not as a support but merely to show the observer
that he had returned to the original position after he had bent forward
to record judgment. Where two observers were used at once two sets of
this apparatus were employed, with the magnets in a single circuit
governed by a floor-button. The touch-stimulus was made to coincide as
closely as possible with the appearance of a given group.

In view of the practical remoteness of this factor from the object of
judgment the experimentation here took two forms,--one in which the
observer was passive toward the touch-stimulus; the other in which the
effort was made closely to associate the touch with the visual group by
imagining the group to be responsible for the touch. For the passive
method the touch was given irregularly now on the first and now on the
last, but as many times on one as on the other.

For the active method, it was given always on the last group. This
constancy was held to favor the active association of touch and
particular group. The constant time-error was guarded against by
experiments in which no modifying factor was introduced. A and B of
Table VIII present the results of the passive and active methods
respectively. C and D repeat A with duplication of groups,--C with the
usual (1/25 sec.), D with a longer, exposure. These last sets were
taken that the factor of touch might be studied when the objective
conditions of the strong distribution influence should have been
removed. It might prove that a factor swamped in the former situation
might emerge into effectiveness.

The following summary gathers the chief facts of Table VIII: (1)
Touch appears practically without effect in A. (2) In B, the
results for touch seem again insignificant; but comparison with the
control-results, to isolate touch from time-order, while it shows no
marked change for Angier, does show for the others that touch was
effective in determining the direction of error by difference-values,
in the two cases of 10.2 and 14 per cent. The active method seems to be
slightly more favorable to the influence of touch. (3) The duplication
of the groups in C gives a large increase to the apparent effectiveness
of touch, which is considerably diminished but not destroyed by
the lengthening of the exposure in D. (4) The introspection for A
indicates that touch under these experimental conditions has little
subjective importance for the judgment of number. It is sometimes quite
unnoticed. Angier made a possible exception in its favor in cases of
great hesitancy where it added "importance" to the group with which
it occurred. Usually he felt little doubt. With Shaw the touch was at
first distracting but later indifferent. Johnston's notes indicate
rather more effect. The touch prevented strict attention to the figure
impression whereby the space-intervals in that group lost in value.
Later it lost its confusing effect. Here seems to be subjective
tendency, but not enough to predominate in results.


TABLE VIII

    A
    88 _experiments
    with each of two
    subjects._ 132 _experiments
    with one
    subject._

                B
                198 _experiments
                with each of two
                subjects._ 110 _experiments
                with one
                subject._

                        C
                        44 _experiments
                        with each subject._

                                    D
                                    88 _experiments
                                    with one subject
                                    and_ 44 _with the
                                    other_
                                    _Exposure_ = 1/4 _sec._

                           A                              B
                               _No                            _No
            _Touch_ _No touch_  tendency_  _Touch_ No touch_   tendency_

    Subjects                    3                     3

    Av.% of
     difference
     in favor of                3.2                   5.7


                           C                              D
                                _No                            _No
            _Touch_ _No touch_  tendency_  _Touch_ _No touch_  tendency_
    Subjects   2                                 2

    Av.% of
   difference
  in favor of 26.1                              13.7

  Results in B for the subjects separately were as follows: Angier 4%,
  Johnston 8.2%, Shaw 5%, all, so far as they went, in favor of the
  touch-group. Control experiments to determine the time-error gave
  the following results: Angier 6.8% in favor of the group last seen,
  Johnston 2.2%, and Shaw 9% in favor of the first group.

Some further introspective evidence appears in connection with the
active method of B. Angier confirms his earlier account exactly.
Usually the factors of distribution practically associated with number
determine the judgment promptly; but in cases of doubt the touch is
felt to add to its group something that appears as number-value.
Johnston's subjective situation seems a little complicated. I may
summarize thus: (_a_) The connection between the touch and its group
being established, that group seems smaller, as being, together with
the touch, somewhere nearly equal to the first. (_b_) The connection
established and touch failing to come, that group seems smaller. (_c_)
The connection not established and attention being concentrated on
the visual impression, the touch-group feels much larger. The curious
attitude in (_a_) results in a discounting in advance of the actual
number. This done, the touch adds numerical value to its group. In
(_c_) the effort at abstraction appears to emphasize the second (touch)
group. Later, he reported similarly that the touch-stimulus seemed
to add to the number of circles in its group even when the judgment
favored the other group; and that "any outside stimulus connected
with the one of two exposures tends to lose its own significance and
be translated into number of dots to help the accompanying exposure
to equal or exceed the first." The touch-group is felt to have more
significance through association with an idea of superior energy or
greater motor impulse.

Of the character of the influence exerted by the touch, Shaw reported
that there seemed to be a diminution in the size of the first group
and something extra in the second. More specifically, this effect
appeared at times as an added circle at the right of the second (touch)
group. He thought that this effect was overruled by the real bases of
number-judgment which he summarized as "size, regularity, density, etc."

These notes show a definite tendency on the part of the touch-stimulus
to break in upon the course of the number-judgment ordinarily
determined by the practical association of a specific group of factors
with number. That this result gets no more marked registration in
the percentages is apparently due to the strength of these customary
associations.

(5) The extent to which the distribution-error complicates the
present study is shown by the prompt increase in effectiveness of the
touch-stimulus when the groups were duplicated, as in C and D.


2. _The Influence of Hearing._

The scheme of the experimentation upon this factor conformed in
general to that of Section IV, 1. But a new sort of differentiation
was possible in the auditory field, and one more readily suggestive
of numerousness, perhaps, in that by use of an electric bell a rapid
succession of sounds could be given with one group while with the other
a single sound could be produced. An actual numerical difference in the
auditory field might fuse with the factor of relative visual number
and determine the judgment to its direction. These results are set
down in B of Table IX. The same set of cards was used in the One-Group
Apparatus for these experiments as for those of Section IV, 1. In these
two sections of Table IX the observers did not know on which group
the sound or the particular sound would be given; but any possible
disturbing effect of this irregularity was formally eliminated as in
Section IV, 1. The experiments of Table IX, C, repeat those of A with
duplication of groups; and D repeats those of C with longer exposure.

The sound for A was that of a small organ-pipe (Ut 4) blown by mouth.
As in A of the preceding table the observers did not know in a given
experiment with which group the sound would be given, but, as before,
it was given the same number of times with the first as with the last.
For B the multiplied sound was produced by an electric bell with
a wooden gong. This was adopted in preference to metal because of
the prompt ceasing of the sound after the stroke,--a very necessary
condition when this sound accompanied the first group, that it might
be clearly connected with its own group. A metal gong was used for the
single sound, that the two might not be too unequal in loudness. Its
vibrations were deadened by a rubber band, and each bell was controlled
by a floor-button. For C and D a higher sound, from the same pipe
unstopped, was used in preference to the former, for the reason that in
certain experiments performed just previously the lower sound had been
used and was presumably very familiar. So in order that the sound might
be brought, if possible, afresh to the attention, the change was made.


TABLE IX


                  A                             B

           132 _experiments_             44 _experiments_
            _each with_ 3                 _each with_ 2
           _subjects_. 180               _subjects_. 88
           _with_ 1 _subject_              _with_ 1 _subject_

           _Exposure_ = 1/25 _sec._        _Exposure_ = 1/25 _sec._

          _Sound_\_No       _No          _Many     _One     _No
                   Sound_   tendency-   Sounds_   Sound_   tendency_
  Subjects                     4           1                  2

    Av.% of
    difference
    in favor of              5.4         18.2                2.2


                    C                               D

            44 _experiments each_                  88 _experiments each_

             _Exposure_ = 1/25 _sec._             _Exposure_ = 1/4 _sec._

        Sound_  _No Sound_  _No_            _Sound_  _No Sound_   _No_
                          _tendency_                             tendency_
  Subjects
          1                   1                                     2

        20.4                 4.6                                   2.2

The results of these experiments may be summarized as follows: (1)
The figures give evidence of but two cases out of eleven where sound
was influential. (2) Duplication of groups is not effective in
developing evidence of the influence of sound. (3) Increased length
of exposure works, as in former cases, to lessen the influence of the
modifying factor. (4) The introspections are to the effect that the
sound seems to be entirely without influence upon the judgment, beyond
the distraction it brings in the earlier stages of work. Sometimes it
dropped wholly out of consciousness. Sometimes the distraction seemed
to last longer. One observer reported, when D was taken, that he
felt as if the sound sometimes increased and sometimes decreased the
apparent numerousness. In some other experiments not directly upon this
point, but later to be reported, a sound was used; and one observer
reported that it seemed to become functionally connected with certain
gaps in the groups, as though the puff had blown a hole in the group.
Here its effect was of course to emphasize negative factors. It appears
thus that the sound might function in opposite directions at different
times, somewhat in accord with the particular character of the visual
presentation. We should expect, then, to have percentages that look
insignificant. (5) We shall not have failed to notice the difference
between touch and auditory stimuli in the feeling of influence upon
the number-judgment. If we seek a cause for the superior influence of
touch, we may perhaps find it in the fact that practical experience
has trained us to disregard in any case of judgment such simultaneous
presentations as were employed for auditory stimuli; while a definite
tap upon the brow is a rather unusual experience likely to attract
notice to itself in spite of attempts at abstraction. As one observer
said, who took part in both kinds of experiments, the touch seemed more
"intimate."


3. _The Influence of Kinæsthetic Impression._

The method consisted in the employment of active effort upon a fist
dynamometer or a wooden handle during the appearance of one of the
groups. The handle was preferable because noiseless. The effort was
made with the left hand because the right was used in recording. The
amount of it was left to the observer's regulation, with the one
instruction that its presence be made decidedly evident but without
too great fatigue. The cards of Section IV 1 and 2 were used in the
One-Group Apparatus. Similarly again the experiments were repeated
with the duplicate-group cards. I present no table here because the
figures show practically no influence of the effort. On one subject 176
experiments were made; on a second 132; on a third 88.

It is interesting to note here certain results obtained from one
observer when he was in what he described as an active attitude toward
the groups, in which he seemed to rouse himself to an unusual pitch of
concentration upon the visual situation. This was evidently a condition
of increased effort to abstract. Without abstraction he gave 26 to 6 in
favor of the effort while with abstraction this tendency had fallen off
to 30 to 17. The strength of the tendency is thus strongly indicated.
Another observer felt a kind of motor difference between the groups;
he expected the effort-group to look larger and felt additionally
excited, a scattered activity, while he was passive toward the other
group. Perhaps this account puts a little meaning into his small per
cent. That his power of abstraction was effective here is hinted by
his remark that he felt a difference in the groups even when he judged
them equal. The third observer found no subjective evidence that effort
modified his judgment.


V. THE "ERRORS" OF EXPERIMENTATION

Throughout the foregoing experiments has been involved the possibility
of some one of the three "errors" of experimentation, those of time,
space, and distribution, and sometimes all three. Their effect on the
results, if it existed, was, to be sure, eliminated in the well-known
way, but their existence, if actual, would raise an interesting
problem. It was possible, in the case of every group of experiments,
to rearrange the tables in such a way as to bring out the evidence for
any tendency to overestimate, for instance, the first group as against
the second, the right as against the left, or one kind of irregular
distribution as against another.

The distribution-error must have a word of explanation. It refers to a
tendency to give more wrong judgments in favor of one kind of irregular
distribution than of the other kind with which, in a given card, it is
mated. In the construction of a set of cards several forms of irregular
internal arrangement were used, in order that the judgment might not
be one merely of form, and of course on any given card the forms were
not the same. Elimination of the effect of these form-differences
from the results involved the appearance of any given one as many
times in connection with one of the two contrasting factors studied
in a given experiment as with the other. Thus two sets of forms were
carried through an experimental series--a source of error indeed, but
avoidable only by such means as were used to escape the effects of the
space-error. Analysis would show which, if either, of the two sets
received more judgments in its favor, resulting in further evidence as
to the extent to which the judgment of relative number is a function
of distribution, and as to the fineness of discrimination for such
differences.

Now the tables, when thus rearranged, show that these errors exist to
a surprisingly large extent. In many cases their causes, whatever they
are, seem to be the controlling factors in the judgment of relative
number.

Barring the experiments of Section III, in which the space-error has
largely been accounted for, I now propose to gather in one survey all
the results of those analyses that have given us the information of the
existence of these errors, and all the material of later tables that
bears on this point, and to test them by further experimentation. I
will begin with the space-error.


TABLE X

                      _Av. % of          _Av. % of         _Av. % of
                    difference in      difference in      difference in
                       favor of_         favor of_         favor of_

               _Cases_  _Right_  _Cases_ _Left_   _Cases_ _No tendency_

    Angier        1      25         6     17.9       9      5.7
    Davison       5      26.4       1     10.8       6      6.2
    Dunlap                          7     18.3       4      4.4
    Holt          2      13.6       7     17.6       4      3.2
    Hylan         8      23.6                        7      6.7
    Meakin        2      19.2       1     29.6       8      5.5
    Meriam        2      16.5       2     12.9       7      6.8
    Moore         2      11         3     16.6       7      3.1
    Peterson      1      13.6       1     12.2       9      4.3
    Rogers        4      21.9       2     15.9       6      4.5
    Rouse         3      15.5                        8      5.1
    Shaw          3      20.9       5     18         7      6.1
    Windate       1      22.8       4     19.5       7      4.5
    Yerkes                          6     26.6       8      5.2
    Henry         1      10         2     16.6       3      8.1
    Woods                           3     19.3       3      4.8

1. _The Space-Error._

Table X presents to us a summary of the values of the space-error
tendency. (1) Taken as a whole they fall into all the three classes
that are possible; (_a_) favoring right; (_b_) favoring left; (_c_) no
marked tendency. (2) There is no observer that does not at some time
show a fairly marked tendency. (3) All the observers fall into (_c_)
and all but five into both (_a_) and (_b_). (4) More favor the left
than the right,--50 to 35. (5) This survey makes it clear that the
observers agree neither with themselves nor with each other in the
direction of influence exerted by the causes underlying the space-error.

_a. Special Experiments to establish the Facts._ It might be suspected
that irregularities would be more apparent where other factors such
as we have been studying enter to complicate the situation from the
point of view of pure relative position of the two groups. Table XI
presents the answer to this query. The cards used contained groups
of gray circles (Gray Darker, Prang) arranged in equal areas of the
usual size and shape. The distribution-error was eliminated, though
not by duplication, and the small-difference cards were retained. The
Two-Group Apparatus was used, with an exposure of 6/5 sec.


TABLE XI

                   88 _experiments with each_

                           _Right_     _Left_     _No tendency_

    Subjects                 4           7             3

    Av. % of difference
      in favor of           25.6        23.1           3.4

The results give us again our inevitable three classes, and in many
cases a difference-value surprisingly large when we reflect on the
simplicity of the conditions. That the omission of complicating
features was of importance is shown by the fact that more of the
observers (11 out of 14) show a marked error than in any other case.
Clearly enough the various factors introduced tend to eliminate
the space-error, but when in any case it does enter, it is even
then capable of rising to as high a degree on the whole as in the
uncomplicated series, as is shown by the fact that in but four cases
does the new value surpass the best of the old, and in three of these
by a trifling amount.

It is interesting to note that the three cases of minimum space-error
show a well-defined tendency to be determined by distribution.

_b. Possible Bases of this Error._ The outcome of these special
experiments is that the factors found in the groups are at least not
directly responsible for the situation that we are considering. The
divergence among the observers shows this. In hunting after the cause
for this apparent influence of side, we look first for changes in
the peripheral, and then in the central, processes that precede the
judgment. The material used for the experiments of Table XI seems
approximately to have equalized all the objective factors in the two
groups. How could there be anything further in the peripheral process
whereby group could be differentiated from group? The most evident
thing is that the visual stimulus is received in a different way from
the two groups. There is a definite peripheral mechanism whose factors
seem essentially to be two, however variously they may be combined:
(_a_) The relative amount of time given to each group; (_b_) the order
in which the groups are viewed.

The observers were instructed and continually reminded to equalize the
amount of attention devoted to the group; but as this is not wholly
a voluntary matter, the possibility of failure to conform has to be
reckoned with. Experiment must therefore be employed to test the
influence of these factors before one can fall back upon a central
process as the cause for this tendency to favor a side.

_c. Its Relation to Differences between the Groups in Length of Look._
The material was the same used for the experiments of Table XI. The
method was the same with the following necessary exceptions: The longer
exposure was double the shorter (4/5 sec. to 2/5 sec.), and 2/5 sec.
elapsed between the two. Further, the experiments were so arranged as
to equalize the influence of the order of exposure with respect to both
side and relative length. The means for effecting successive exposure
took the earlier form described in the introduction to Section II.


TABLE XII

                       _88 experiments with each_

                          _Longer_    _Shorter_     _No tendency_

    Subjects                              6              7

    Av. % of difference
      in favor of                        18              6.5

These facts are yielded by Table XII: (1) There are but two classes of
observers, as no tendency exists to favor the group of longer exposure.
(2) The time-error shows a considerably more marked tendency than the
length of look, which is indeed somewhat surpassed by the space-error.
(3) The persistence of the space-error, even among those that reveal
a tendency in length of exposure, shows that the factor of relative
difference in length of look cannot account for it. The persistence
of it, too, when the order of exposure is controlled, even though the
conditions are not wholly adapted to the study of this latter factor,
suggest at least that the space-error is independent of even that
order; but into this we shall make special enquiry. (4) The judgment
of number is independent of the amount of eye-movement devoted to the
fixation of the objects in a group. This conclusion, so far as the
actual movement is concerned, is established by the fact that so many
observers favor the shorter look; and by all the experiments with the
One-Group Apparatus where an exposure of 1/25 sec. was used, since
that time was too short to admit of movement. That ideated movement is
likewise insignificant appears from the fact of marked error arising
in the material where the groups were duplicates. Here no motives to
different movements could lie in the material.

_d. Its Relation to the Order in which the Groups are viewed._ Table
XIII gives us the results of the enquiry. The experimental conditions
were not changed except as to the length of exposure. Each group was
given 3/5 sec.; and half the experiments were performed in the order
right-left and half in the reverse order.


TABLE XIII

                       _88 experiments with each_

                          _First_    _Last_     _No tendency_

    Subjects                 3         9             2

    Av. % of difference
      in favor of           17.5       28.3           1.7

The results may be thus summarized: (1) The order of exposure is
notably influential upon the judgment of relative number, giving the
usual three classes, with the tendency to overestimate the last group
well in the lead. (2) The persistence of the space-error under these
relatively simple conditions shows conclusively that it is not a
function of the order of exposure. The two are independent variables.

2. _The Time-Error._

In pursuit of our enquiry we must survey the facts as they are given
in the various experiments already reported and later to be reported.
These facts are gathered into Table XIV, which furnishes the following
items of significance: (1) All the observers, with the exception of
Rouse, show at some time a definite tendency. One case only is given
for him in this table, but other experiments not included in the
tables from which the present is drawn confirm this fact by the ratio
29 to 30. (2) There is a rather striking consistency in the several
observers. (3) The predominance of the last group is marked.


TABLE XIV

                      _Av. % of_          _Av. % of_            _Av. % of_
                     _difference_        _difference_         _difference_
                     _in favor of_       _in favor of_        _in favor of_

             _Cases_   _First_    _Cases_   _Last_   _Cases_  _No tendency_

    Angier                           7       19.1       4          4.9
    Baldwin     4       20.5                            1          3.4
    Bell                             1       18.2       4          6.3
    Davison                          2       31.2
    Dunlap      1       11.4                            1          2.2
    Holt                            10       19         2          6
    Hylan       2       16           5       25.4       5          4.9
    Johnston    2       25.2         4       35.5       4          4.2
    Meakin                           2       39.7
    Meriam      1       29.6                            1          2.2
    Miller                           3       17.1       4          5.5
    Moore                            1       11.4       1          2.2
    Olmsted     1       15.2
    Peterson                         1       17         1          9
    Rogers                           1       10.2       1          5.6
    Rouse                                               1          1.2
    Shaw        6       16.7                            5          7
    Windate                          1       11.4       1          1.2
    Yerkes                           2       42.7

_a. Relation of the Error to the Absolute Length of the Total
Exposure._ Table XV is set to answer this question. It is
unsatisfactory in that but two observers took part in both XIII and XV.
The material used for judgment consisted of the same cards used in the
earlier experiment, but presented now in the One-Group Apparatus. The
time of exposure was changed from 3/5 sec. to 1/25 sec. for each group.
As the space-error was eliminated, the tendency to a time-error, if
present at all, would presumably have freer play.

But the difference-values of the new table are for the most part very
small. We have thus the further fact about the time-error that, under
the conditions studied, it appears to be independent of the absolute
length of exposure, when the groups are equal in this respect. To
this we may add another fact drawn from Table XIV, that with the
One-Group Apparatus the time-error is greater on the whole where the
groups are differentiated by other factors. Thirdly, the values for
Table XIII show that with all complicating factors withdrawn, except
the differences in position, the error is at a maximum. This may
be significant of the effect of space-differences upon that error,
or, more probably, be due to the general difference between work by
daylight and work in a dark room by artificial light. We shall be
better able to consider this later.


TABLE XV

     88 experiments with each of two subjects.
    176 experiments with one subject.
    154 experiments with one subject.
     66 experiments with one subject.

      Exposure = 1/25 sec.

                          _First_   _Last_   _No tendency_

    Subjects                1                     4
    Av. % of difference
      in favor of          15.2                   6.8

3. _The Distribution-Error._

The last of our three "errors" of experimentation is now before us.
We may recall once more the meaning the term has had for us in these
studies. It points to a tendency discovered by the use of those
cards where all objective factors were in the course of a series
equalized,--a tendency to mass one's judgments in favor of a particular
arrangement of the circles; though each group had been constructed
with a view to filling the given area as homogeneously as an irregular
arrangement would allow.

As in the two "errors" preceding, so here we must get possession of
the facts that gave rise to the present enquiry. Table XVI presents
them to us, gathered out of all the tables wherein such a tendency has
been technically reckoned with. But first a few words of explanation
are needed to make the new table intelligible. Two sets of results are
found in its two parts. In each set the particular group-arrangements
employed and the frequency of their appearance are exactly the same.
The two sets differ, as their headings suggest, in that the material
for the second set was formed out of the first by replacing the
small-difference cards by those having equal groups. Such a change
as this might affect the proportion of judgments given in favor of
the two sets of arrangements in a particular series, and these new
results are, therefore, no longer fully comparable with the earlier
ones. In presenting the directions of tendency in the results, it is
impossible here, as in all the similar cases throughout the tables, to
name a factor as a standard in whose favor all the judgments in the
plus column should be understood as given,--impossible for this reason
that, because the very method by which the circles were distributed
in the groups, the experimenter was unable to satisfy himself as to
the significant differences in the arrangements. All the results,
however, when analyzed on this basis, were recorded consistently, so
that consistencies and agreements among the observers might be readily
apparent. We can now understand in part what Table XVI has to say to us.


TABLE XVI

                                  A
                                                           _No_
            _Cases_ _Class 1_ _Cases_ _Class 2_ _Cases_ _tendency_

    Angier                       4       19.1      2       6.3
    Davison                      3       23.1
    Dunlap                       2       13.7      1       2.2
    Holt      3      15.5        1       25        2       4.5
    Hylan     1      11.4        2       14        3       6.7
    Johnston                     5       30.4
    Meakin                       3       34.9
    Meriam                       1       16        2       6.8
    Miller                       5       32.1
    Moore     1      39.8                          2       5.7
    Olmsted                      1       27.2
    Peterson                     3       42.1
    Rogers                       1       11.4      2       5.7
    Rouse                        2       14.2
    Shaw                         6       29.1
    Windate                      3       30
    Yerkes                       3       17.8

                                  B
                                                           _No_
            _Cases_ _Class 1_ _Cases_ _Class 2_ _Cases_ _tendency_

    Angier                                         3       2.9
    Davison
    Dunlap
    Holt      1      22          1       21.6      2       3.1
    Hylan     2      50.8        1       11.4      1       8.4
    Johnston
    Meakin
    Meriam
    Miller
    Moore
    Olmsted
    Peterson
    Rogers
    Rouse
    Shaw                         2       26        1       0
    Windate
    Yerkes

  Here as elsewhere the per cents recorded indicate the average per
  cent of difference in favor of a given class.

Here are the facts, first of A: (1) The only lapses from consistency
are confined to two observers; and in both these cases there is but a
single break in a uniform trend. (2) With three exceptions all agree
in the trend of their difference-values. Of these three--Holt, Hylan,
and Moore--the last furnishes but one significant value, and so must be
left out of the reckoning on this point. (3) Of the 64 cases, 50 rise
above 10%, some far beyond, showing the importance for the judgment of
relative number of this factor of distribution. (4) Of the 50 cases,
45 agree in tendency. (5) That with this surprising agreement we
have still a few exceptions, adds another item to the growing array
of evidence on behalf of the importance of some subjective factor for
the number-judgment. As to the nature of this factor we are yet in
the dark. (6) To these facts B of this same table adds the further
information that the observers inconsistent in the old are not
consistent in the new, while the consistent still maintain their record.

_a. Analysis of the Experimental Conditions of Distribution._ At once
we are interested to enquire for the factors underlying these results.
To put ourselves upon the right track we must first consider what
factors are involved in any such arrangement of objects as we have used
in the material for these studies, and then, more precisely, we may ask
in what way such arrangements could differ significantly. Finally, by
an experimental trial-and-error process, we may solve our problem.

The groups of objects in our material were arranged in an area marked
out in each corner by a circle. Within this area the circles were set
irregularly, with the result that the group, as a mass of objects
distinguished from a homogeneous background, had a more or less
irregular outline whose irregularity varied with different internal
arrangements. Within its outlines this area presented a mixed pattern
of bright and dark. While the total enclosure marked off by the corner
circles was always the same and theoretically the relative amounts of
brightness and darkness in equal groups was likewise the same, yet
practically differences, more or less slight, might enter through the
changing character of the rude outlines whose ideal completeness could
scarcely be brought out of a black background by the uninitiated. The
amount of this difference is sometimes surprising to one whose chief
thought of the group has been as vignetted in process of construction.
As the objects are pushed toward the edges the central spaces open out;
as they are withdrawn toward the interior gaps appear in the margin.

It is not a very easy task to fill an area with objects in irregular
arrangement in such a way that no sections of vacancy or filling
stand out by contrast against the remainder of the same element. To
succeed in this is to fill the area homogeneously. But the chances
are good that some vacant patch will get slightly the better of its
neighbors or some section of circles will gather a little more closely
than the surrounding circles; or perhaps a gap in the outline will be
unexpectedly intrusive. Now in a given area the circles of one part
cannot become more thickly massed without a corresponding enlargement
of the vacancies of the other parts, and of course the converse is
as true; but this theoretical situation may be quite out of ken at
the moment when the group is seen. Either member of this pair of
complements may stand out vividly in the field and its fellow quite
escape perception. The very nicety with which in practical affairs we
have to make a reliable comparison of this sort shows what suspicion
of accuracy the off-hand judgment has bred. And further, the widening
of a gap or thickening of the filling in one small part of a group may
give a complementary loss to the rest of the group small enough to be
unperceived when distributed throughout the larger section.

Two factors must therefore be considered as possibly significant in
moving the judgment,--vacancies and filling; and with the former must
be reckoned indrawing of the outline. Psychologically, increase in
the prominence of either of these factors would be all one with their
objective increase. With respect to the direction of their influence
upon the judgment of number the increase of vacancies must signify the
waning, and the increase of filling the waxing, of the objective number
in the group.

It is in advance altogether probable that the results gathered into
Table XVI were brought about by these two factors, at least in large
part. And we have also in these factors the possibility of two types;
for as we saw above, increased vacancies in one part involves increase
of filling in another, and conversely. So the interesting question
turns upon the altogether disproportional representation of types.
Which is the type of the majority?

_b. Experimental Test of Hypotheses._ The question was put to the test
of experiment. This was done by using groups in which now vacancies
and now filling were objectively emphasized in contrast with the usual
homogeneous group. First the vacancies. A set of cards was prepared
after the method previously used to eliminate the distribution-error
without duplication of groups on any one card. (See Section II.) In the
present case, however, the two sets of arrangements were definitely
differentiated as already indicated. One set had a homogeneously
filled area, the other a prominent vacancy within or gap in the edge.
The size of these variations was kept pretty close to the limit of
noticeableness, that the increase in compactness of the other portions
might be as slight as possible. It was experimentally necessary to
free the material as far as might be from ambiguity, and practically
important to avoid rousing the suspicions of the observers and the
resulting reflections. It seemed very likely that the strength of the
tendency shown by the distribution-error was due to its appearance
in situations where the observers knew that other factors were being
tested.

The general method already described was used in preparing the groups
that gave objective prominence to compacted parts of the filling. To
fulfil the conditions outlined above was here even more difficult
than in the first set; and the cause will appear in the sequel. The
small-difference cards were omitted and the One-Group Apparatus used.

A further attempt was made to head off reflection by a subterfuge.
It had been found that, among the factors whose influence on the
judgment had been studied, hearing had been as little effective as
any. So the small stopped pipe used for those experiments was again
brought into service and the error resulting eliminated in the usual
way. Incidentally our new tables will thus give us further information
about the effect of this factor, though of course under conditions that
are theoretically highly unfavorable, since we are forcing upon the
attention of the observers other factors that experience has shown them
only too ready to seize upon. So if a tendency traceable to the factor
of hearing should appear, we ought perhaps to give it somewhat more
than its face value.


TABLE XVII

                             A.

                     _Exposure = 1/25 sec._

             _Homogeneous_   _Vacant_    _No tendency_

    Angier      [1]50
    Baldwin     [2]53.4
    Bell        [1]52.2
    Holt        [2]44.4
    Hylan       [2]51.2
    Johnston    [1]56.8
    Miller                                  [2]4.2
    Shaw        [1]29.6

                             B.

                     _Exposure = 1/4 sec._

             _Homogeneous_   _Vacant_    _No tendency_

    Angier      [2]51.2
    Baldwin     [2]55.6
    Bell
    Holt                     [1]13.6
    Hylan       [1]52.2
    Johnston    [2]62.6
    Miller      [1]25
    Shaw        [2]14.8

                             C.

                     _88 experiments each_
                     _Exposure = 1/25 sec._

             _Homogeneous_   _Compact_    _No tendency_

    Angier        39.6
    Baldwin       35.2
    Bell                                       3.4
    Holt                        27.2
    Hylan         39.6
    Johnston      44.4
    Miller                      16
    Shaw                                       2.2

    [1] 44 Experiments.
    [2] 88 Experiments.

  The per cents recorded indicate the average per cent of difference in
  favor of a given factor.

Now we are ready to inspect the results. Table XVII, A is the outcome
of the attempt to emphasize vacancies. Its experiments with 1/25 sec.
exposure were repeated with one of 1/4 sec. as Table XVII, B, shows.
In Table XVII, C, the emphasis of compactness is concerned.

For convenience we may again resort to a summary outline in extracting
the meaning from these tables. First Table XVII, A. (1) All the
observers but one agree in favoring the homogeneous, most of them
with very high difference-values. (2) Miller alone gives no tendency,
and his notes show a conflict between the increased vacancy and the
increased compactness. In other words, his discrimination was too keen
for the material. Under the circumstances he constitutes no exception
to the conclusion that the vacancy objectively emphasized was the cause
for an underestimation of its group.

From Table XVII, B, we learn the following: (1) All the observers save
one favor the homogeneous group, in most cases by large values. (2) The
difference in the length of exposure seems to have no significance for
this tendency, since, while Holt and Shaw decline, Miller rises in the
scale.

Table XVII, C, gives us these facts: (1) The difference-values have
noticeably fallen off. (2) We have again the customary three classes,
but with homogeneous leading as in the earlier tables. (3) By his
present favoring of the compact, Miller has now appeared in all three
classes, while Holt has developed the preference for the compact
that was budding in XVII, B. (4) The presence of four well-marked
preferences for the homogeneous shows that the vacancies in the
compact group were more significant for the number-judgment than
was the increased compactness of the filling, and that in spite of
the experimental effort to the contrary. (5) The decrease of this
tendency and the growth of the opposing, indicates that the judgment is
determined in either case by the more vivid factor.

The conclusions to be drawn from these facts lie close at hand. (_a_)
The results in Table XVI, with their disproportionate division into
classes, were evidently due to the tendency of three observers to note
the filling and of the rest to be concerned with the vacancies. (_b_)
The judgment of relative number under these conditions is primarily
a judgment of vacancies. (_c_) The subjective factor of vividness
determines the direction of error, and may attach to either vacancies
or filling, though it usually attaches to the former.

It may not be out of place here to speculate a bit as to the probable
cause for so close a dependence of the number-judgment upon what has no
number, so to say; upon an object that has no standing in the official
conclusion. The situation seems to be fundamentally based upon the
conditions that determine contrast. In a homogeneous field no part
stands out. Introduce a small object quite different in brightness or
complementary in color and the attention is drawn instantly to it,
but internal differences in its content are quite lost in the common
quality by which it differs from the ground. A case somewhat analogous
is furnished by our material, particularly in the One- and Two-Group
Apparatus. The small group is so unified by its contrast with the field
that internal differences must be made out with relative effort. Now
internal differences are necessary to the numerical character demanded
of it, and they can be brought out in no way save by attending to the
vacancies and so isolating parts in the threatening unity, each in a
kind of space-matrix. The most careful observer could not do better on
his way to truth; and that is why the error was so much larger when the
factor of space-differences was studied.

That group is normally the more numerous in which the vacancies are
less completely developed under observation. We say "normally" here
by virtue of the speculation just completed as to the best method of
attaining a judgment objectively true. For a man thus proceeding,
our proposition is a sound statement of fact, to which the following
results of our experiments bear witness. (_a_) The experiments
recorded in Table XII on Relative Difference in Length of Look shows
no exception of a value equal to 10% to the general statement that all
tendencies, when any existed, were in the direction of favoring the
shorter group. The shorter the time of exposure the less completely
would the vacancies develop. (_b_) Table IV, E, shows that without
exception the darker group tends to be judged the more numerous.
(_c_) Table XXI shows for each subject that in a shorter exposure the
absolute number seems considerably greater than in a longer exposure.

No comment seems necessary to concentrate the force of such evidence.
If we carry our proposition to the detailed results of our separate
studies in factors of distribution, we shall find that it helps us to
understand those few exceptions to the general trend of observers as
they appear in Tables II and XVII. The exceptions there favored the
groups in which compactness of parts went along with certain large
vacancies. Possibly enough they refused to fall in with the objective
analysis, and, disregarding the prominent vacancies, devoted themselves
to a development of the vacancies within the compacted parts.

_c. The Factor of Hearing._ The time-error analyses of the experiments
of Table XVII have already contributed their facts to the special
section dealing with that error. But one or two interesting facts have
remained unnoticed in the sound-analysis. In the experiments of Table
XVII, A, there is a single case of marked tendency to favor the sound
group. With the lengthened exposure of B, this tendency, as usual,
disappears; but returns in C to some extent and two other observers
share it. A fourth markedly favors the group without sound. So the
experiments of this last table present as marked external evidence as
we have for the influence of hearing upon the judgment. These facts are
presented in Table XVIII.


TABLE XVIII

                                 A
                       _44 experiments with_
                       _each of 4 subjects,_
                       _88 with each of 3._

                       _Exposure = 1/25 sec._

                              _No-_      _No_
                   _Sound_   _Sound_   _tendency_
    Subjects          1                    6

    Av. % of
     difference
     in favor of     27.2                  4.1

                                 B
                       _88 experiments with_
                       _each of 3 subjects,_
                       _44 with each of 3._

                       _Exposure = 1/4 sec._

                              _No-_      _No_
                   _Sound_   _Sound_   _tendency_
    Subjects                               6

    Av. % of
     difference
     in favor of                           5.9

                                 C
                       _88 experiments each_

                       _Exposure = 1/25 sec._

                              _No-_      _No_
                   _Sound_   _Sound_   _tendency_
    Subjects          3         1          3

    Av. % of
     difference
     in favor of     12.9      20.4        4.5

It is a further curious fact, well sustained by these same experiments,
that where there is some confusion, each of the factors present has a
better chance to determine the judgment. The values for both time-error
and sound rise higher for the majority in C than in A or B.


VI. THE INFLUENCE OF FACTORS IN THE SAME SENSE-FIELD UPON THE JUDGMENT
OF ABSOLUTE NUMBER

The nature of the enquiry that we have been pursuing through so many
pages is such that it may be raised exactly as well in the case of
absolute as in that of relative number. There appears to be no reason
why in this new field the results should not be exactly comparable
with those in the old, to be taken indeed as a kind of test for the
interpretation to be put upon the old. Without a single exception,
unless it were imposed by a technical difficulty, all the earlier
factors could be studied with the new purpose. Our practical interest
to go to such lengths would depend pretty largely upon the results of
first attempts. If wholly confirmatory, these would probably suffice.

The experimental conditions were of the simplest. The 3-8 in. steel
balls of Section III were again pressed into service as objects for the
number-judgment. They were thrown loosely into a fixed black frame, 20
cm. square. To avoid suggestive noises, its undersurface was made of
a thick piece of felt covered with black cloth; and the whole rested
of course on a black-topped table. The exposures were 2 sec. long,
timed by watch-ticks. Between experiments the observer held a cardboard
screen between him and the objects. When conditions were ready for a
new judgment, closing his eyes he lowered the screen, opening his eyes
again at the word of command and shutting them at the close of the
experiment.

Of course the observers felt that their judgments were for the most
part extremely vague. With small numbers they had a greater feeling of
confidence. Yet altogether it was surprising with what readiness an
absolute number-judgment would spring up in the presence of any given
collection whatever within the limits set by the experimental series.
Sometimes the observers thought that they made rough calculations on
the basis of the filling in a unit of area. So far as this held it
would tend to cut off the more astonishing departures from correctness,
and it would probably advantage the smaller groups more than the large.
Still it was entirely too rough a method to prevent the influence of
the factors introduced, as the results will show. There was no time
for systematic counting, which, in any case, the observers knew to be
forbidden.

The figures in which the observers reported their judgments of
absolute number have a value that is chiefly qualitative. The marked
inconsistencies and disagreements are our guarantee for this statement.
With all the observers there was but the loosest association between
group-appearance and number-name. The innumerable variations in
internal space-relations were of course responsible. For one observer a
particular name probably had a quantitative significance far in excess
of its value for another observer in this respect. To one man 100 might
have meant about the same as 60, for example, to his neighbor. On the
whole they were parsimonious; but Baldwin decidedly not.

A more or less constant influence was exerted on any given judgment by
the comparison of the presented group with the traces of the preceding
still in mind. The observers felt, however, that the judgment was
largely independent of such comparison, and its fluctuations give some
credence to this feeling.

The numbers chosen ranged by fives, from 25 to 100. In four cases a
number was immediately repeated that rough suggestions as to the
definiteness of the judgment and its dependence upon the actual number
might be gained. These were indeed but rough suggestions, since,
with certain exceptions to be noticed later, the arrangement was
disturbed between times; but they made possible a closer watch upon the
flickering of the judgment than could be kept by a mere repetition of
the series. In the latter case it might be unstable and yet relatively
firm in the other. A standard series is here recorded. Its order was
determined by drawing the numbers out of a heap, but the repetitions
were inserted arbitrarily.

     1. 95
     2. 25
     3. 35
     4. 65
     5. No change
     6. 30
     7. 90
     8. 85
     9. 45
    10. 100
    11. 50
    12. No change
    13. 60
    14. 40
    15. No change
    16. 70
    17. 80
    18. 55
    19. 75
    20. No change

1. _Absolute Number under Standard Conditions._

An indispensable preliminary for the present study is the establishment
of a standard. Unless we know something in advance about the
characteristics of the judgment of absolute number in relatively simple
conditions, we shall be unable to tell what influence, if any, to
attribute to the modifying factor in later experiments. Having then
decided as to the general conditions under which we will study the
problem we must make these the standard conditions of our work; and
having discovered the nature of the judgments given under them, measure
up to these results in all that is to follow. These standard conditions
have already been set forth in the introduction to this section. The
results are recorded in Tables XIX and XX.


TABLE XIX

    KEY:  St = Standard
          Sc = Scattered
          Co = Compact

                 _Subject_ = _Baldwin_    _Subject_ = _Miller_

    _Trials with_
    _each number_   6      3      3         4      2      3

      _Original_
      _Numbers_   _St_   _Sc_   _Co_      _St_   _Sc_   _Co_

         25         1      0     -6       -10     -6     -4
         30         3      5     -5       -12     -7     -6
         35        10      7     -3       -14     -4     -7
         40        10     13     -8       -15     -8     -11
         40        10      8     -8       -14    -10     -12
         45        10     18     -3       -21    -13     -10
         50        17     23      2       -19    -15     -17
         50        15     22      2       -17    -15     -14
         55        27     40      0       -18     -3     -12
         60        19     33      5       -20    -23     -17
         65        27     53      2       -18     -3      -7
         65        24     57      5       -13    -13       2
         70        38     77      0       -19      8      -8
         75        31     73      0       -24     10     -20
         75        28     78     -5       -24      8      -8
         80        36     83     -5       -14      5     -18
         85        54     85      5       -19     -8       0
         90        54     90      5       -13      8      -3
         95        48     73      5       -30     20     -15
        100        61     87      7       -13     10      -7

  The figures recorded are the average of the algebraic sums of other
  figures that represent the difference between the actual and the
  estimated number. Fractions are replaced by an added unit, if the
  value is 1/2 or over.

  Baldwin never underestimated the scattered group, and only once the
  standard; but 31 times the compact. Miller only once overestimated
  the standard group, and but 6 times the compact; but 13 times the
  scattered.

Turning to these tables we notice at once, as characteristic of all
the observers, the following facts: (1) Wide variation from objective
correctness. (2) A far wider discrepancy with the larger numbers than
with the smaller. Miller does not wholly agree here. His judgments by
series show inconstancy, tending at first to follow the rule, but in
the last two series to a maximum error near the middle. Certain remarks
of this observer suggest that possibly in the latter case reflection
as to the convenience of certain actual numbers for manipulation may
have had influence. The three earlier series of Hutchison conform to
the rule. The remainder, on the contrary, show no definite progression
in tendency. It should be noted here that both Miller and Hutchison
were more inclined than the other two observers to rough calculation.
The effect of its adoption or of increased practice in it is shown by
the disappearance of the characteristics of the earlier series. We
have thus in these two cases a doubleness of standard that we must not
fail to consider in our later comparisons. (3) There is a pronounced
instability of judgment, as shown by the fluctuations for the same
number in different series, and especially in successive judgments,
of the same in any given series. (4) There is a general tendency to
judge in multiples of five. That there should be any splitting of
fives, particularly in the large numbers, might be regarded as mere
caprice. Not so did it seem to the observers. They were conscious of
an apparent absurdity in it where judgments were necessarily so vague;
but they insisted that this stood for a kind of qualitative shading
in the perception which threw out the choice of the round numbers just
above and below. (5) The number is on the whole underestimated, three
observers agreeing in this respect; but the fourth shows a very large
and consistent tendency in the opposite direction.

In spite of the manifold special inconstancies and disagreements, these
general tendencies are decidedly well-featured in the results. We may
say that we have found a kind of standard illusion that will serve us
for a guide through our later studies.

2. _The Influence of Distribution._


TABLE XX

                  _Subject_ = _Hutchison_       _Subject_ = _Olmsted_
    _Trials_
    _with_
    _each_
    _number_                3                      2       4
                         _First_     _Second_
             _Original_  _Standard_  _Standard_  _Mixed_  _Small_
             _numbers_   _Series_    _Series_   _Sizes_  _Sizes_
                25          -5          -2         -3       -1
                30          -4           0         -5        0
                35          -6          -5         10        1
                40         -11          -5         -5      -13
                40         -12          -9        -13      -11
                45          -7          -4        -15      -10
                50         -14         -10        -20      -12
                50         -13          -8        -15      -15
                55         -10          -7        -20      -11
                60         -19         -18        -25      -20
                65         -15          -7        -20       -5
                65         -15         -10        -10      -18
                70         -14          -3        -15      -20
                75         -21         -17        -28      -23
                75         -25         -20        -20      -24
                80         -24         -17        -30      -19
                85         -17         -13        -35      -13
                90         -13          -7        -25      -20
                95         -25         -13        -40      -23
               100         -22         -13        -35      -24


                 6           2        3
                         _Mixed    _Small_
            _Standard_   _Sizes_   _Sizes_
                -8         -8      -12
               -11        -14      -16
                -8         -5      -14
               -18        -17      -20
               -13        -18      -22
               -18        -20      -18
               -19        -18      -23
               -22        -15      -27
               -24        -28      -23
               -22        -23      -27
               -28        -28      -33
               -22        -28      -23
               -27        -30      -28
               -32        -23      -27
               -33        -25      -32
               -27        -23      -33
               -27        -38      -38
               -33        -40      -37
               -39        -55      -40
               -36        -15      -37

  For the meaning of these figures see under Table XIX. Hutchison
  overestimated the standard group only 5 times, never the mixed-size
  group, and 9 times the small-size group. Olmsted never overestimates
  at any time.

The first of the modifying factors to be considered has to do with the
arrangement of the objects. Hitherto they had been thrown loosely into
the frame. Now in successive studies they were, first, well scattered
over the surface and, second, brought together into several compact
nuclei. The last arrangement was adopted in preference to that of a
single mass as being less open to comparison with preceding judgments
and to judgment on the basis of form and size of group.

The results are shown in Table XIX: (1) The effect of scattering the
objects is very markedly to raise the apparent number. Baldwin's
preceding overestimations soar still higher; while Miller's former
tendency to underestimation is checked to such an extent that 13
overestimations appear. (2) The effect of compacting the objects
is just as markedly in the opposite direction. Baldwin gives 31
underestimations, and Miller reverts in a measure to his former type.
(3) When similar arrangements were up for study in relative number we
found two classes of observers, one favoring the compact, the other the
scattered. The present results of Baldwin and Miller put them into the
latter class.

3. _The Influence of Complexity of Group-Content._

This new factor of complexity in the content of the group was realized
experimentally by making up the collection out of steel balls of two
sizes, 1/8 in. and 3/8 in. The former looked almost infinitesimal
beside the latter. The same objective numbers were still maintained and
divided between the two sizes except where in so doing a five must be
broken. In such a case the extra five went to the larger balls.

The results are found in Table XX. Olmsted shows no definite influence
of the new factor. Hutchison, however, shows a very evident decrease
in his estimations, when comparison is made with his second standard
series. With the earlier series the new results rather closely
correspond. That the latter are not simply a vacillating reversion
seems fairly clear from this observer's account of his method. The
small balls, he says, did not distinctly come in visually. To his
judgment of the large he added an amount based on a very insecure
estimate of the small. The number of the latter seemed from time to
time pretty constant.

This situation corresponds very fully to that in the investigation
of the same factor by use of a group of mixed colors, where relative
number was in question. (Section II.) The tendency there discovered was
to neglect the other colors in favor of one which thus surpassed the
others in vividness. There as here the mixed group seemed smaller.

4. _The Influence of Size of Objects._

A study of this factor was made possible by substituting for the
usual objects steel balls of a smaller size, 1-4 in. The results are
contained in Table XX. They are not so striking as those obtained in
our study of distribution. Still the influence of this new factor is
evident, in the reduction of the apparent number. Olmsted shows this
more generally for the smaller numbers. We find it in Hutchison when we
compare the new results with the second standard series. This tendency
to underestimation increases in the two final series of the present
set. At the beginning of these two he remarked that he thought he had
been overestimating the group. This tendency of smaller size to reduce
apparent number was found true for the majority of observers in our
earlier study of relative number.

5. _The Influence of the Length of Exposure._

I found that in relative number the shorter the look the more marked
was the influence of certain factors. Reports of the observers making
this seem highly probable happened in this way: When working with the
One-Group Apparatus in relative number the shutter of the camera would
occasionally stick, leaving a group exposed beyond its usual time.
The effect of this upon some but not all the observers was to cause a
noticeable shrinking in numerousness. Of those questioned, the only one
failing to notice this effect is included among the observers in this
new study.

To test this possibility resort was had to the One-Group Apparatus as
affording a more satisfactory means for getting different lengths of
exposure of small absolute magnitude. Cards were prepared containing
a single group of larger area (67 × 82 mm.) than had been used for
relative number. The objects were the usual white circles. Each corner
was marked as usual; and, by reason of the number involved, the outline
of the area was more regular than had been true in the earlier work.
The number of circles on each card varied by steps of two from 16 to
30, giving eight cards in all. The series was arranged irregularly as
before, and two of the cards repeated immediately upon their first
presentation, making ten experiments in one set. The order of the
series follows:

    1. 24
    2. 22
    3. 26
    4. No change
    5. 18
    6. 28
    7. 16
    8. 20
    9. 30
    10. No change

Two time-magnitudes were used for comparison,--1-25 sec. and 1 sec. The
latter was managed with bulb exposure. All the experiments with the
shorter time were made before those with the longer had been begun. The
results are given in Table XXI. So far as the material is comparable,
we may include in our comparison the standard experiments of Tables XIX
and XX with 2 sec. exposure.


TABLE XXI

    KEY:  A = 1/25 sec.
          B = 1 sec.

                          _Baldwin_   _Miller_   _Hutchison_   _Olmsted_
                _Actual_
                _Numbers_  A    B      A    B      A    B        A    B
    _Number of_
    _trials with_
    _each number_          5    6      5    5      5    4        4    4
                    16     6    9     -2   -3     11   -1        2   -5
                    18     8   10     -4   -1     12   -1       -3   -5
                    20    34   19      1    3     19    1        1    1
                    22    37   17      6    6     14    6        4   -2
                    24    49   26     12    5     14   10        5    0
                    26    70   33     15    8     12    9        5    4
                    26    79   37     21    9     14    4        9    4
                    28    93   42     29   11     18    7        8    7
                    30   106   52     38   12     18    5       19   10
                    30   103   50     45   11     20    4       19    4

    For the meaning of these figures see note under Table XIX.

The outcome may be thus summarized: (1) The apparent number is
inversely proportional to the length of exposure. The tables show a
perfectly clear progression from 2 sec. to 1-25 sec. All those that
formerly underestimated are brought into the opposite class. (2) The
results of the earlier experiments are confirmed on the whole with
respect to the occurrence of greater errors with the larger numbers.
(3) Baldwin's overestimation reaches astonishing heights. (4) These new
facts for absolute number are quite in accord with Table XII, where,
under the conditions of interpretation laid down, the tendencies were
wholly in favor of the shorter look.

The issue of these tentative experiments in absolute number confirms
the teaching of our studies in the related field. Absolute number, like
relative, has been found largely subject to a modifying influence of
certain factors. In the new field, too, distribution has asserted its
supremacy among these, and similar effects of shortening exposure have
been observed. There has been variation among the observers and some
shifting of tendency, both of which point as before to the coöperation
of some subjective factor in our results. Indeed the whole situation,
as opened by these preliminary studies, indicates a theoretical
interpretation that for both fields is at bottom one. So to an attempt
to reach such an interpretation the next section will be devoted.


VII. THEORETICAL DISCUSSION

1. _The Fact of Modification._

That such an influence upon the judgment of number should have been
exercised by the factors considered seems in many cases to receive
an adequate account on the principle of association. Our practical
experience in the simultaneous variability of number and certain other
characteristics of a group of objects has been such as to lead us
into illusions when the two no longer vary together. In such a case,
when we have no time to count, we are actually led to _see_ a group
as smaller or larger in accordance with the variations perceived in
the associated factor. This interpretation is supported by the fact
that on the whole the space-factors were more markedly influential in
creating illusions than were any others. For those cases, however,
in which the modification was effected by a factor unconnected with
number, as color, or the simultaneous stimulation of other senses by
irrelevant objects, it appears that the mere occurrence of greater
total stimulation during the appearance of one group is sufficient to
create illusion, either through failure of the observer to discriminate
between the relevant and the irrelevant, or because he is led through
fear of disturbance to overemphasize the other group.

2. _The Direction of Modification._

The foregoing account of the general fact throws no light upon the
_direction_ of the influence. Why should a given factor make a group
seem more numerous and not less? Why should it affect one man in one
way and his neighbor in another? Why should it vary with the same man
at different times? Appearances no less contradictory than these are
what we must face in carrying a theoretical account to completion.
The following propositions with appended commentary are offered in
satisfaction of these requirements.

_a._ Differences in vividness among the factors determine differences
in number.

Our study of the factor of distribution in Table XVII, where it was
possible in a measure to control the vividness, furnishes evidence for
this proposition. Introspective reports in other cases confirm this
view by showing that the direction of the attention, the popular way of
stating our proposition, was the determining feature. This will receive
further support in our discussion of the following proposition.

_b._ If the vivid factor or complex be positive, _i. e._, associated
in experience with the numerous, or if it be neutral, its group will
seem the more numerous. If negative, _i. e._, associated in experience
with the few, its group will seem the less numerous.

The experiments upon the effect of distribution support this
proposition, especially as set out in Table XVII. When the vacancies in
a given group were made vivid, the other group seemed more numerous;
when its filling surpassed in vividness, the judgment was given for
it. We have other confirmation in the fact that lengthening the time
of exposure reduced the absolute number. Take also this note of one
observer on the material in Table II, C:

"I noticed that I had set the open spaces in the outlined group over
against the lack of them in the homogeneous, without paying much
attention to the nearness together of the spots in the lines of the
outlined. Then for a time my attitude was quite vacillating. I found my
attention drawn to the nearness together of the spots in part of the
outlined group so strongly that if I did not turn it voluntarily to the
fact that the other was filled without any large open spaces, I was led
to call almost any outlined group the larger. Toward the end of the
experiment I got back into my original attitude, in which the outlined
group seemed to have its spots hardly more thickly arranged in any part
than the homogeneous, and to have also the bare spots and so to be the
fewer."

That the vividness of a neutral factor or complex increases the
apparent number was suggested by comments of the observers. One
observer reported of the material in Table IV: "The greater brightness
of red gave it more importance. The natural thing seemed to be to give
the red the judgment. The gray fought more for recognition." And again:
"The red seems a vitalized space and the dots more omnipresent, also
the red lasts longer in memory and is there more vivid, so that often
in cases of doubt, where the decisive comparison was made in memory,
the red may have been given the vote. Often there was an immediate
unanalyzed feeling that if the groups had been both of the same color,
the judgment would have been for the gray." In both cases his results
showed this tendency. Another observer, whose results agree with the
former, found that his eye was directed involuntarily toward the red.

This fact was put to a special test. In the material of Table II,
B, a card, in which the pattern group had an appearance strikingly
different from the normal, was introduced, the two groups being
objectively equal. With three observers the effect was overestimation
of this group, and with a fourth, the suppression to equality of a
previous overestimation of the opposite group. This fact, together
with the observers' comments, seems to justify the conclusion that the
vividness of a neutral factor or complex was the determining condition
of the judgment. That the observers did not all show positive results
in this experiment may be set down to the difficulty in controlling
the subjective conditions of vividness. Of course the space-relations
within the new pattern were different from those in the old. The only
justification for taking no account of these is the character of the
introspections themselves.

It should be said of the red group that beside its vividness it had
characters mentioned by other observers that might independently have
made its number seem greater. It was called "dazzling," "blurred," and
its area seemed increased. In this respect the effect of the color
should be discussed as a special case of distribution or object-size.

The vividness tested in this special way seems due to contrast, in
the one case with surroundings, in the other case with the expected.
Such a judgment is very far removed from the normal bases, rather more
so, it would seem, than even those where a group had sound or touch
accompaniment; for in the latter case there could be no question about
the "moreness"; the only doubt could be about its legitimacy. Of the
precise extent to which this cause of vividness has operated throughout
our studies, even where spatial differences have been concerned, we
cannot be sure. The patterns of the materials in B and C of Table II
seem to offer that possibility. That it should enter anywhere opens,
indeed, the entire field.

_c._ The observers fall into the following classes on the basis of the
character of the association:

(1) Relatively fixed association,

  (_a_) involving correct adjustment to objects (vividness of relevant
  factors);

  (_b_) involving incorrect adjustment to objects (vividness of
  irrelevant factors).

(2) Relatively unstable association.

It will be remembered that in the case of nearly every factor studied
under Relative Number, we found three classes of observers,--those
favoring one group, those favoring the other group, and a so-called
"no-tendency" class. The bases of classification were, first, the
relative constancy in the character of the error, and, secondly, its
direction. In this third or "no-tendency" class were really lumped off
two kinds of observers, not separated at the time because our special
interest did not demand it. Along with those that gave large errors
in both directions was a much smaller class that gave a relatively
large proportion of correct judgments; but could never claim any one
observer all the time. In the new classification of Proposition _c_
this mixed composition is recognized by dividing it between (1) (_a_)
and (2). The prime condition of correct judgment is asserted to be one
in principle with that of the illusion,--namely, vividness, but in this
case vividness of relevant factors. Our original "tendency" classes
both fall under (1) (_b_).

Proposition _c_ is merely an attempt to apply the principles of
association and vividness to an organization of our results. The types
in question have no hard-and-fast connection with any particular
observer; they rather represent a kind of ideal fixation of opposite
tendencies playing through all.

3. _The Time-Error._

So far the time-error has been left without interpretation. The chief
facts to be considered were: (_a_) Divergence of error and general
trend in favor of last. (_b_) Individual inconsistencies. (_c_)
Occasional absence.

We are in a position now to invoke at once the principle of vividness
to account for the existence of the error and the vividness of recency
to account for the predominant tendency to favor the last of the two
groups exposed. In this respect this error may be classed with the
effect of red. That is to say, a factor or complex, directly through
its vividness and not indirectly through its association with the
numerous or the few, draws the judgment after it. Here the content of
the group is the effective thing, not the character of the vacancies.

But the observers do not all agree in the direction of the time-error
nor are they always consistent. Here we shall get help from a
proposition offered in the discussion of the distribution-error in
which it is asserted that the group seems the more numerous in which
the vacancies are less developed under observation. We have already
noted a decided tendency to depend in judging upon the vacancies. Let
us suppose that the two groups presented in succession differ with
respect to the success of the observer in developing these vacancies.
If this be true, that difference may well depend upon the occurrence
of maximal attention during the exposure of but one of the groups. The
tendency of the majority to overestimate the second group suggests
that the attention is likely to be at a maximum when the experiment
begins. If, on the other hand, it ripen later, or if the observer seek
to rescue the second group from relative unclearness, then we should
have the first group overestimated. The time-error would disappear for
those that could attend alike to both. Clearly enough this account is
decidedly hypothetical.

4. _The Space-Error._

Our attempt to reduce this error to one of time in some form was
proven a failure. The facts brought out indicate that at the bottom
is some subjective factor thus far not isolated. This factor is not
a preference going directly with right- or left-handedness because
on the surface at least it runs in the observers independent of such
asymmetry. A single bit of available introspection would seem, however,
to point to some relation of that sort; for one observer, who favored
the left, felt that a group on that side gained an importance that was
somehow due to the greater absolute value of a weight in the left than
in the right hand. Even if this be decisive for him it will still be
inapplicable to errors in the opposite direction unless we assume that
with variations in bodily energy the emphasis is cast now in one, now
in the other, direction, after the analogy of those two types of man
to be found in our social experience, for whom respectively mountains
are molehills and molehills mountains. Such successive differences in
type in a single individual would then find an intelligible account in
the shifting tides of that bodily energy. It is to be noted that the
observer just quoted once, but only once, made a decisive reversal of
his error from left to right.

It may occur to some one that the use of two observers sitting side by
side may have given them a preference for one position of the groups.
In the first place care was taken that both groups should be as readily
seen from one point of view as from the other. Secondly and chiefly,
there is no regularity among the observers in this respect.

It is not unlikely that a chance aspect of a particular group develops
an emphasis that gives the mechanism of subjective adjustment a
particular bent that for a time is relatively independent of the
objective situation. Still the fact that there are some cases of
persistence in type is rather damaging to this assumption and speaks
rather for the earlier one. That one, if true, seems indeed adequate to
account for the situation. As an hypothesis it accords with analogous
physiological facts; but its weakness lies in imposing the burden of a
strong tendency upon asymmetrical differences that may be in comparison
relatively slight. Finally, these studies furnish no proof that the
bodily condition of an observer of a particular type corresponds to the
demands of the hypothesis.

_Summary:_ 1. The estimation of relative number in the visual field is
modified by group-area, internal distribution, order, and complexity in
group-composition; by the size, form, color, brightness, and complexity
of the individual members; and by the character of the environment.
It is further modified by factors contributed by the objects through
other senses, as in active pressure, special differences in pressure
character, active weight, and that complex of muscular and spatial
factors arising when a group is observed under the condition of
eye-muscle strain. The judgment is also influenced by factors outside
the group in the field of touch, but not in that of kinæsthetic
impressions.

2. On the whole the most influential factors were those lying in the
space-characters of the groups; while those of least moment were
contributed by other objects in other fields of sensation. Hearing was
very nearly ineffectual.

3. In very many cases the observers fell into three groups, one of
no-tendency, and a second and third showing opposite tendencies with
respect to the factor investigated.

4. With a majority of observers there is a tendency to underestimation
in the judgment of absolute number, though with a single observer the
tendency is directly the reverse. Scattering the objects increases,
and compacting diminishes, the apparent number. The smaller the size
of the objects the fewer, under conditions, do they appear; while
heterogeneity in group-composition lessens the number for one observer
and has no apparent effect upon the other.

5. The apparent absolute number of objects is inversely proportional to
the length of exposure of a group; and in relative number the influence
of a factor was on the whole greater for shorter exposures.

6. The marked tendency to a space-error was found to be independent of
differences between the groups in the length of look, and of the order
in which they were viewed.

7. The distribution-error is grounded in a fundamental tendency to base
the judgment of relative number upon the character of the vacancies
in a group; though a secondary tendency to depend upon the filling
was shown to exist. The subjective factor of vividness, attaching
now to one and now to the other of the foregoing factors, determines
which shall be operative, though it usually is joined to the first.
The ground for the primary tendency may very well be the necessities
imposed upon discrimination by the material. The contrast effect
between the large black background and the brighter objects tends to
unify the latter, which, to be discriminated as a number, must be split
up by an emphasis of the vacancies.

8. The time-error is possibly due to differences in power to dismember
the groups exposed in succession in one experiment, while its
variations in direction seem adequately accounted for by differences in
the time at which attention becomes maximal during the progress of a
single test.

9. The ground for these facts of modification is found in the strength
of the association between these several factors and the elements that
signify number.

10. The basis for the different tendencies found among the observers
is the differing vividness among several factors. If the vivid factor
is associated with the idea of numerousness, or is in this respect
neutral, its group will seem more numerous. If it has been associated
with the idea of fewness, its group will seem less numerous. The
difference between the two classes of "tendency" and "no-tendency" lies
in the fact that for the latter either only correct clues are vivid,
or else there is so frequent an alternation in vividness of opposing
incorrect clues that through any given series no tendency appears,
while for the "tendency" class misleading clues are without shifting in
the ascendant.

At the time when these experiments were completed, no work precisely
upon this problem had been published. Since then, however, Dr. J. F.
Messenger has issued a monograph entitled The Perception of Number
(Psych. Rev., Mono. Supp., vol. 5, no. 5), of which certain parts fall
within the scope of the present studies. He was concerned with the
estimation of absolute number and was primarily interested to discover
the nature of the number-judgment. The reader of both articles will
find agreement between the results and interpretations here recorded
and such part of Messenger's work as has been a common object of
study,--viz., the factors of distribution and size.

FOOTNOTES:

[Footnote 123: 44 experiments.]

[Footnote 124: 88 experiments.]




TIME-ESTIMATION IN ITS RELATIONS TO SEX, AGE, AND PHYSIOLOGICAL RHYTHMS

BY ROBERT M. YERKES AND F. M. URBAN


The desirability of a statistical study of time-estimation was
suggested to us by a note concerning "sex-differences in the sense
of time" which was published in Science recently by Prof. Robert
MacDougall.[125] By comparing the time-estimates of groups of men and
women consisting of fifteen individuals each, MacDougall discovered
that for intervals of from one quarter of a minute to a minute and
a half, the women exhibited a far stronger tendency to overestimate
than did the men, and were at the same time markedly less accurate.
The nature and extent of the overestimation discovered by MacDougall
are indicated by the results presented in the accompanying table from
Science. The numerals, 1, 2, 3, and 4 refer to different fillings of
the intervals (listening to reading, marking letters, etc.), the signs
+ and - to over- and under-estimation respectively.

            Period, One Minute.

    Sex      1       2       3       4
    Men     +29"   + 1.3"   +22"   - 3.5"
    Women   +66    +22.0    +80    +24.0

These apparent sex-differences in time-estimation demand further
attention, first, because the number of individuals studied by
MacDougall, as he recognized, is too small to establish the fact of the
existence of such differences, and, second, because if the differences
really do exist they should be studied in their relations to age and
the fundamental physiological rhythms.[126]

It seemed probable that further investigation of this subject might
reveal some important facts concerning the development of the ability
to estimate time in the individual, the significance of various
conditions for time-estimation, the psychology of sex, and the
relations of rhythms to personal affinities, antipathies, and motor
capacities.

In this report the results of a statistical study of the
sex-differences in time-estimation are discussed, and in later papers
we shall present the results of investigations of the relations of
time-estimation to age and to individual and sex rhythms, and attempt
to work out a convenient and serviceable rhythm-formula. The need of
such a formula for expressing individual rhythms is obvious, as is also
the need of comparative studies of individual and sex rhythms.


TIME ESTIMATION

    Name                              Place
    Age                               Date

    ORDER OF TESTS              TIME IN SECONDS

                      _Time of_          _Male_          _Female_
                     _Intervals._     (_17 years_)     (_17 years_)
                                         No. 1             No. 9

     1. Idleness        _108̋_            _70̋_            _120̋_
     2. Reading          _36_             _30_             _118_
     3. Writing          _72_             _36_              _60_
     4. Estimating       _18_             _15_              _30_
     5. Reading         _108_             _90_              _68_
     6. Idleness         _36_             _35_              _60_
     7. Writing          _18_             _10_              _10_
     8. Estimating      _108_            _100_             _125_
     9. Reading          _72_            _100_              _66_
    10. Idleness         _72_             _75_              _58_
    11. Writing          _36_             _25_              _22_
    12. Estimating       _72_             _60_              _60_
    13. Reading          _18_             _14_              _15_
    14. Writing         _108_            _130_              _59_
    15. Estimating       _36_             _30_              _41_
    16. Idleness         _18_             _10_              _18_

  How did you estimate the interval when you were asked to estimate it
  as accurately as you could?

  n o f e y m i q r s a d r g d e s t k n w e r a x u p x z y o n d f n
  o d c a e h p m a l g s r w y t b c k p s o n q a r v q c o m p v r i
  c p k t o s n q z r l x m i h u v o q g P p f u t o i c n g s c a r n
  o t c d a a o b i a r s a d e r w o a i e r g l c r t h f s o r a e n
  s i o c r b x g r z b h o w l t s

    Number of letters counted               85                 88
    Pulse-rate                              72                 81

The experimental data now to be considered were obtained as follows.
Record-sheets of the form reproduced above were printed, with blanks
for age and name of subject, place, date, for sixteen judgments of
time-intervals (numerals 1 to 16), for a statement of the subject's
method of estimating time, for the number of letters counted in thirty
seconds, and for the pulse-rate. Four intervals were used, 18, 36, 72,
and 108 seconds, and for each of these intervals judgments were taken
under the four conditions designated on the record-sheet as idleness,
reading, writing, and estimating. In the experiments the intervals
were not given in order of regular increase or decrease of the length
of interval, nor were all the judgments for any one interval taken
together, but instead, for the purpose of avoiding the influence of
expectation of a particular interval or filling, they were arranged
irregularly in the order of column two of the record-sheet. This
column, as also columns three and four, which are specimen series of
judgments for a male and a female respectively, of course were not
printed on the record-sheets which were supplied to the subjects.

The experimental procedure was as follows:

(1) Each subject was given a record-sheet.

(2) The experimenter was provided with a record-sheet on which the time
of the intervals numbered from 1 to 16 was given. Care was taken that
the subjects should not know the length of the intervals before the
experiments.

(3) The beginning of each interval was indicated to the subjects by the
word "start" uttered distinctly by the experimenter; the end, by the
word "stop."

(4) Before beginning the sixteen tests the experimenter gave a
thirty-second interval as a standard of judgment. The experiment then
proceeded with only sufficient pause between judgments to allow of the
recording of estimates by the subjects.

(5) During the filling called "idleness" the subject did not pay
special attention to the estimation of the time, but instead permitted
his attention to wander.

(6) During "reading" the experimenter read aloud to the subjects.

(7) During "writing" the subjects wrote from the dictation of the
experimenter.

(8) During "estimating" the subjects judged the interval as accurately
as they could, by whatever method they chose except the use of a
time-piece.

(9) Each subject recorded his judgment of the length of an interval in
seconds at the appropriate place on the record-sheet as soon as the
interval was ended.

(10) The question following judgment number 16 on the sheet was
answered as soon as the sixteen judgments had been recorded.

(11) The number of letters counted in thirty seconds was determined
by the use of the lines of letters at the bottom of the sheet. The
subjects began at the left of each line and counted singly as many
letters as they could between the "start" and "stop" signals of the
experimenter. They then marked the last letter counted and immediately
recorded, in the place provided on the record-sheet, the number of
letters counted.

(12) The pulse was counted by the experimenter immediately after the
experiment when possible and the rate recorded on the sheet.

(13) The experimenter avoided delays, interruptions, or other
irregularities in the course of the series of experiments.

The materials for our discussion of the sex-differences in
time-estimation consist of the judgments of 251 males and 274
females. The majority of the males were students in Harvard College,
the majority of the females, in Radcliffe and Smith Colleges. The
remainder of the records were obtained in Ohio State University, Pomona
College, and West Chester State Normal School. The authors gratefully
acknowledge their indebtedness for assistance in the obtaining of
records to Professors A. H. Pierce, T. H. Haines, W. H. Scott, D. R.
Major, A. M. Smith, H. A. Miller, and B. T. Baldwin. The males ranged
in age from 17 to 23 years, the females from 17 to 20. The total number
of judgments, the distribution of which among the various ages is shown
in Table 1, is 4014 for the males, 4375 for the females.

Despite the fact that our experiments are open to the criticisms of all
work done under variable conditions and by different experimenters, it
cannot be doubted that the results indicate certain sex-differences in
time-estimation which suggest additional problems. For the present we
refrain from interpretations for the most part and state merely the
statistical results of the investigation.

Previous studies of the "time-sense" and the conditions which influence
time-estimation suggested to us the desirability of examining our
data with reference to (1) sex-differences in estimates of intervals,
(2) age-differences, (3) the influence of different fillings, and (4)
differences dependent upon the length of the interval. The results have
been studied, therefore, with reference to the significance of sex,
age, filling, and length of interval, but as no marked age-differences
appeared, the detailed tables which were constructed to exhibit the
results for the subjects of each year of age have not been printed.

In all the tables the results for males and females are presented
separately. The judgments for the sixteen intervals are arranged with
reference to the length of the interval, not in the order in which they
were taken; all the 18̋ intervals, for example, are grouped (Table 2).
The letters I, E, R, W, refer to the fillings of the intervals.


TABLE 1

NUMBER OF SUBJECTS, AGE, SEX, AND NUMBER OF JUDGMENTS

                       _Males_

     _Age_    _No. of subjects_   _No. of judgments_

     17 yrs.        16                 256
     18             27                 432
     19             40                 639
     20             67                1071
     21             50                 800
     22             35                 560
     23             16                 256

    Totals         251                4014

    _Interval_                    _No. of judgments_

     18"                              1004
     36"                              1003
     72"                              1003
    108"                              1004

    Total                             4014

                    _Females_

     _Age_     _No. of subjects_  _No. of judgments_

     17 yrs.        73                1160
     18             57                 911
     19             64                1024
     20             80                1280

    Totals         274                4375

    _Interval_                    _No. of judgments_

     18"                              1092
     36"                              1094
     72"                              1096
    108"                              1093

    Total                             4375

A general survey of the individual records, all of which for any one
year and sex were tabulated, for convenience of examination, on a
single large sheet of coördinate paper, showed that the judgments vary
within a wide range and are very inexact. Table 2 exhibits the number
of correct judgments for each sex, interval, and filling. Of the
4014 male judgments only 96 (2.39%) were correct; of the 4375 female
judgments only 46 (1.05%) were correct. The number of correct judgments
decreases as the length of the interval increases. For the 18-second
intervals there were 7.37% for the males, 2.48% for the females,
while for the 108-second intervals there were only 0.10% and 0.37%
respectively.


TABLE 2

FREQUENCY OF OCCURRENCE OF CORRECT JUDGMENTS

                                     _Males_

             18̋           36̋           72̋          108̋
         I  E  R  W    I  E  R  W    I  E  R  W    I  E  R  W     Σ     %
         29 26 11  8   5  6  3  0    3  3  0  1    0  1  0  0    96   2.39
  Totals 74 = 7.37%   14 = 1.40%     7 = 0.70%     1 =0.10%      96   2.39

                                    _Females_

          7 15  1  4   2  4  1  0    2  5  0  1    2  1  0  1    46   1.05
  Totals 27 = 2.48%    7 = 0.62%     8 = 0.73%     4 = 0.37%     46   1.05

  _List of abbreviations which occur in the tables._

  Σ always designates the sum of the results of the column which it
  heads.

  I, E, R, W refer respectively to the intervals of idleness,
  estimating, reading, and writing.

  The % sign refers to the value of the result in question in terms of
  the total number of judgments.

  C refers to the results of the letter-counting test.

The male judgments for the 108-second intervals range from 11 to 300
seconds. If random guesses be made within these limits the probability
of the occurrence of right guesses (108") would be 1 in 290; therefore
among 1004 guesses (the number of male judgments for 108-second
intervals) 3.5 would be right. In the experiment only one judgment of
the 1004 was correct. For the other intervals, with the exception of
18 seconds, the number of correct judgments is only slightly greater
than random guessing would have given. Both males and females, however,
show considerably more correct judgments for 18-second intervals than
the number of probable right guesses. Within the range of the male
judgments and for their number 16.9 right guesses might be expected,
for the females 10.9. In contrast with these numbers the experiments
furnished 74 and 27 correct judgments respectively.

It is noteworthy that for those intervals which are most frequently
correctly judged, not only is the number of correct judgments greater
for the males than for the females (the ratio of the percentages is
about 3 to 1), but the ratio of the number of correct judgments to the
probable number of right guesses is also greater for the males.

The female judgments vary within a wider range and are less often
correct than the male. For the latter the total number of correct
judgments is more than twice that for the former.

Another interesting fact concerning the judgments of the time-intervals
is that certain numerals occur in the last place of a judgment more
frequently than we should expect if their occurrence depended on random
guessing. Tables 3 and 4 exhibit the results of an analysis of the data
made for the purpose of studying this fact. In Table 3 the frequency
of occurrence of the digits 0, 1, 2, 3, etc., in the last place of the
male judgments is given for each filling under the four intervals. For
example, the digit 0 occurred in the last place of the male judgments
for the interval reading 36 seconds 98 times, as we learn by referring
to the first line and third row of the second column of Table 3.

Examination of Tables 3 and 4 shows at once the marked preference of
the subjects for 0 and 5. The percentage of male judgments which end in
0 is 41.50; of female 58.51. Similarly the percentages of occurrence of
the digit 5 for the males is 24.41, and for the females 23.11. Only two
of the other digits (2 and 8) occur with a frequency of over 5%.

Among the 4014 male judgments 0 occurred as a final digit 1666 times,
5, 980 times. Among the 4375 female judgments 0 occurred 2560 times, 5,
1011 times. In the male judgments 0 occurred about four times as often
as it would in random guessing; in the female, almost six times as
often.

Comparison of Tables 3 and 4 indicates that the occurrence of 0 is
17.10% greater for the females, while that of 5 is 1.30% greater for
the males. The sum of the percentages of occurrence of 0 and 5 for the
males is 65.91, therefore the probability that a male judgment ends in
one of these digits is almost twice that in favor of any other digit.
For the females the sum of the same percentages is 81.62, and the
probability of occurrence of 0 or 5 is therefore more than four times
that of the other eight digits.

Tables 3 and 4 show that even numbers occur more frequently than uneven
as final digits. Of the total number of judgments 2461 (3063)[127] end
with even digits and 1553 (1312) with uneven.


TABLE 3. FREQUENCY OF OCCURRENCE OF THE NUMERALS (0 TO 9) AS FINAL
DIGIT OF THE JUDGMENTS COMBINED RESULTS FOR ALL MALES

                18̋                36̋              72̋
         I    E    R   W     I   E   R   W    I    E   R    W

    0   70   57   72  76   131  77  98  90   124  74  120  125
    1    5   11    9   9    10   9   8   7     1  18    5    2
    2   20   15   21  18     5  19  16  15    11  20    5   14
    3    9   17   14  10     8  11  10   8     6  14   12    4
    4    8   12   14   9     3  14  12   3     4  14    8    7
    5   67   58   56  63    62  54  60  85    72  59   62   73
    6   13   17   12  13     9  14  11   6    10   9    7    7
    7   15   22   14  12     6  15  11  10     5  14   13    4
    8   36   30   29  31    15  19  16  20    11  17   11   11
    9    8   12   10  10     2  18   9   7     7  12    8    3

                  108̋
            I    E   R    W       Σ      %      C

        0  151  121  127 153    1666  41.504  21.92
        1    4   11   10   2     121   3.014   7.57
        2    9   12    9   8     217   5.406  13.15
        3    3   12   14   3     155   3.861   7.17
        4    6   18    9   3     144   3.587   7.57
        5   54   45   52  58     980  24.414  10.36
        6    8    7    6   4     153   3.812   8.76
        7    4    7   10   8     170   4.235   7.17
        8    8   13    5   9     281   7.004   8.76
        9    4    5    9   3     127   3.164   7.57


TABLE 4. FREQUENCY OF OCCURRENCE OF THE NUMERALS (0 TO 9) AS FINAL
DIGIT OF THE JUDGMENTS COMBINED RESULTS FOR ALL FEMALES

               18̋                   36̋                   72̋
        I    E    R    W      I    E    R    W      I    E    R    W

    0  120  117  106  126    169  124  182  151    176  145  175  194
    1    4    8    3    2      6    7    1    3      2    3    2    1
    2   17   18   12   13      9   22    7   11      7   11    3    4
    3    4    9    9   11      5    9    3    7      6   15   10    4
    4    2   11   13   10      1    6    4    2      9   14    3    1
    5   89   63   91   85     62   61   60   79     56   55   65   63
    6    9   12   11    6      5   13    7    5      4    6    3    3
    7    5    5    7    7      4    9    5    3      2    7    3    3
    8   16   22   17   12     10   13    5   10     10    9    6    1
    9    5    9    5    1      3    9    0    2      2    9    4    0

                  108̋
            I    E    R    W       Σ        %         C

        0  226  197  196  186    2560    58.515    26.14
        1    1    5    2    0      50     1.143     9.09
        2    4   11    6    3     158     3.611     5.30
        3    3    7    4    7     113     2.583     5.68
        4    2    4    2    2      86     1.966    11.36
        5   32   44   49   57    1011    23.109    14.39
        6    0    9    3    1      97     2.217     6.44
        7    0    5    2    5      72     1.646     6.44
        8    5   11    8    7     162     3.703    10.61
        9    1   10    2    4      66     1.508     4.55

In order that the probability of the occurrence of even and uneven
numbers may be calculated, those judgments which end in 0 and 5 must be
subtracted from the total number of judgments, for the occurrence of
these two digits is apparently due to a constant influence. The problem
may be formulated thus. First, what is the probability that a judgment
is determined by the constant influence in favor of 0 and 5? Second,
what is the probability of even and uneven numbers, when the influence
in favor of 0 and 5 is eliminated? The calculated probability of 0
or 5 is 0.65919 (0.81622) and therefore the probability that a given
judgment is not determined by this influence is 0.34081 (0.18378). The
probable limits of these numbers are 0.00505 (0.00395).

After the subtraction of those judgments which end in 0 or 5, there
remain 1368 (804), of which 795 (503) are even and 573 (301) uneven.
The probability of an even number is 0.58115 (0.62562) and the inverse
probability of an uneven number is 0.41885 (0.37438). The probable
limits of these numbers are 0.00900 (0.01027). There are therefore even
chances that the percentage of occurrence of even numbers is between
the limits 57.215 and 59.015 (61.535 and 63.589), or outside these
limits.[128]

Statistical studies have already proved that in random guessing even
numbers occur somewhat more frequently than uneven. It is therefore
worthy of notice that in these results the frequencies of even numbers
are not uniformly greater than that of uneven; for with the exception
of the digit 6 in the female judgments, the digits next to 0 and 5, _i.
e._, 9 and 1, 4 and 6, occur with least frequency.

In the case of the number of letters counted in a half minute, also, it
appears (see last column (C) of Tables 3 and 4) that 0 and 5 occur more
frequently in the last place than chance would lead us to expect. In
contrast with the results for the time-judgments, in the same tables,
the percentages of occurrence of the various digits in counting present
less marked differences. For the males 3 and 7 occur least frequently,
for the females 2 and 9.

To sum up the results of our examination of the materials with
reference to the occurrence of digits in the final place of the
judgments, the order of decreasing frequency of the various digits is
0, 5, 8, and 2. Of the others 3 and 7 occur more frequently than 4 and
6, with one exception. The statement that even numbers in general occur
more frequently than odd must be modified by the statement that in
these results the digits next to 0 and 5, namely, 9, 1, 4, and 6 occur
with least frequency. These statements hold for both males and females,
but for the latter the frequency of occurrence of 0 is far greater than
for the males.

These results clearly indicate that the judgments are not random
guesses. In seeking further for some explanation of the surprising
frequency of occurrence of judgments which end in 0 or 5, we discovered
that certain numbers occur very frequently, namely, the multiples of
15, 30, and 60. In order to exhibit this tendency quantitatively Tables
5 and 6 have been constructed.

In these tables will be found tabulated the number of times 15 and
multiples of it which are not also multiples of 30 or 60 occur for any
given interval and filling. Likewise are tabulated the frequencies of
occurrence of 30 and multiples of it which are not multiples of 60, and
finally, of 60 and its multiples. The numbers as they occurred in the
three categories run as follows:

     15             30             60
     45             90            120
     75            150            180
    105            210            240
    135            270            300
    165            330            360
    195            390            420

Fifteen and its multiples, as given above, are arranged in one division
of the tables, thirty and sixty each in its own separate division.
The line at the bottom of the tables marked Σ gives the frequency of
occurrence of these three groups of numbers for all the subjects and
for each interval and filling.

As is shown by the percentage of frequency columns of the tables, in no
instance do the multiples of 15 constitute less than 19.52% of the male
judgments and 23.81% of the female judgments. The lowest frequency for
any of the four intervals is 20.32% of the total number of judgments.
The maximum frequency for the males (43.03%) and for the females
(56.57%) is for the interval idleness 108 seconds. That the male and
female maxima should fall on the same interval is interesting.


TABLE 5

FREQUENCY OF OCCURRENCE OF 15, 30, 60, AND THEIR MULTIPLES

_Males_

               15      30      60         Σ         %      _Average_

         {I    48       5       0        53       21.12
         {E    42       9       0        51       20.32
     18̋ {R    43       8       0        51       20.32      20.32
         {W    45       4       0        49       19.52

         {I    29      60       7        96       38.25
         {E    17      41       5        63       25.20
     36̋ {R    22      41       6        69       27.41      29.57
         {W    36      28       5        69       27.41

         {I    29      18      46        93       37.05
         {E    25       7      34        66       26.29
     72̋ {R    28      26      33        87       34.66      34.70
         {W    28      42      32       102       40.80

         {I    25      31      52       108       43.03
         {E    16      23      20        59       23.51
    108̋ {R    22      30      39        91       36.25      36.16
         {W    25      37      43       105       41.83

    Σ      480     410     322      1212
       %    11.96   10.21    8.02     30.12


TABLE 6

FREQUENCY OF OCCURRENCE OF 15, 30, 60, AND THEIR MULTIPLES

_Females_

               15      30      60         Σ         %      _Average_

         {I    58      30       0        88       32.47
         {E    30      35       8        73       26.64
     18̋ {R    56      26       3        85       31.02      28.48
         {W    41      24       0        65       23.81

         {I    31      55      39       125       45.62
         {E    21      36      17        74       27.21
     36̋ {R    31      63      33       127       46.35      38.86
         {W    41      45      13        99       36.26

         {I    29      28      60       117       42.70
         {E    17      18      52        87       31.75
     72̋ {R    26      37      52       115       41.97      41.60
         {W    30      51      56       137       50.00

         {I    12      45      98       155       56.57
         {E    21      27      58       106       38.83
    108̋ {R    23      36      84       143       52.19      47.83
         {W    24      31      64       119       43.75

    Σ      491     587     637      1715
       %    11.22   13.42   14.56     39.22

If no influence worked in favor of the multiples of 15 in these
experiments only one judgment in thirty would be 15 (3.33%) and only
one in sixty, 30 or 60 (1.67%). For the probability of the occurrence
of 15, 30, and 60 is 1/60 + 1/60 + 1/30 = 1/15, as is obvious from the
fact that among sixty consecutive numbers (1 to 60) there are four
which are multiples of 15. According to probability we should expect
multiples of 15, 30, and 60 to occur 268 times among the 4014 male
judgments and 292 times among the 4375 female judgments. As a matter of
fact there are 1212 such judgments for the males, 1715 for the females.
The probability that a male judgment is a multiple of 15, 30, or 60 is
0.3012 (probable error 0.0049); for a female judgment the probability
is 0.39199 (probable error 0.0050).

These statistics indicate that the subjects are constantly and strongly
influenced in favor of judgments which are simple fractions of a
minute. Closer inspection of the tables gives some suggestion of the
nature of this influence.

Comparison of the four intervals (Tables 5 and 6) with respect to the
occurrence of simple fractions of a minute shows that the frequency of
such numbers increases rapidly as the length of the interval increases.
The various percentages of frequency for males and females and for the
four intervals are again presented here for convenience of comparison.

                  18̋        36̋        72̋       108̋

    _Males_     20.32%     29.57%     34.70%     36.16%
    _Females_   28.48      38.86      41.60      47.83

As is obvious from these figures both frequency and rate of increase
are far higher for the females than for the males.

Examination of the percentages (totals) at the bottom of Tables 5 and
6 reveals another remarkable sex-difference; for the frequency of
occurrence of 15 and its multiples regularly decreases for males from
the 15 to the 60 class, whereas for females it regularly increases.

Undoubtedly the time-judgments of these experiments were strongly
influenced by thought of the conventional time-unit, the minute, for
in all quantitative work there are errors in favor of the standard of
measurement and simple fractions thereof. In the present instance this
tendency to favor the unit was strengthened, perhaps, by the giving of
a half-minute interval as a standard for comparison at the beginning of
the tests.

Two explanations of the sex-differences above mentioned are suggested
by our study of the data. One is the fact that the females are less
exact than the males; the other that they generally overestimate the
intervals, whereas the males often underestimate them. One's estimate
of an interval is determined partly by confidence of accuracy. The
longer an interval the less we feel able to estimate it accurately,
and, as a consequence, the more frequently it is judged as the same as
the time-unit or a simple fraction of that unit. The females are less
exact in their estimates than the males, and less exact for long than
for short intervals, and as an accompaniment of their inexactitude we
find the frequent occurrence of multiples of 15, 30, and 60.

But confidence of ability to estimate accurately must be considered in
connection with the fact which suggests our second explanation, namely,
that the female estimates are higher than the male. Tables 7 and 8
show that the females almost invariably overestimate the intervals
rather largely, while the males sometimes underestimate considerably.
The range of the male judgments is from 1 to 300, of the female from
1 to 400. Obviously the chance of occurrence of 15, 30, 60, and their
multiples varies with the range. The greater the range the greater
the probable frequency of 30 and 60 in comparison with 15. In random
guessing the probabilities of the occurrence of 15, 30, and 60 for long
and short intervals is the same, but our results show that this is not
true in the case of these time-estimation judgments. It seems possible,
therefore, that the sex-differences referred to are due to the fact
that the intervals seem longer to the females, and that, therefore,
a feeling of greater inexactitude than would be felt for shorter
intervals leads to the choice of simple fractions of a minute more
frequently than in the male judgments and more frequently for the long
than for the short intervals.

It is of interest in this connection to note that the length of a
second is usually underestimated by females, overestimated by males.
The average number of seconds counted in half a minute by twenty men
and twenty women was as follows:

    _Men._   M. 30.4, M.V.  8.7, R.V. 34.94.
    _Women._ M. 38.9, M.V. 10.6, R.V. 36.70

These figures would seem to indicate that the overestimation of the
intervals of these experiments by the females is due to the use of a
time-unit which is shorter than that of the males (although presumably
of the same length).

We cannot with certainty say whether inaccuracy of judgment stands in
the relation of condition or consequence of the occurrence of simple
fractions of a minute, but it would appear that the female tendency to
overestimate is responsible for the sex-differences already noted. For
whatever be the facts concerning longer intervals the second as judged
by the female is considerably shorter than that of the male.

Since a complicated periodicity of frequency in the distribution of
the judgments is exhibited in the results of Tables 3-6 it is obvious
that the distribution-curve will have a tertiary mode for each number
ending in 0 or 5, a secondary mode for 15, 30, 60, and their multiples,
and a primary mode which may or may not coincide with one of the
secondary or tertiary modes. Extreme irregularity is characteristic of
the distribution-curve. Different groups of judgments, as, for example,
those for the two sexes, those for the different intervals, etc.,
give somewhat different forms of distribution, for the frequency of
occurrence of 0 and 5, as well as of the multiples of 15, is variable.

These facts are important in connection with the selection of an
interval for the construction of the distribution-curves, in that they
indicate how large the interval or class of the distribution-curves and
tables should be.

It is clear from the results of Tables 3 and 4 that the smallest
interval which can be of value is 10 seconds, for a smaller interval
would necessarily exhibit irregularities due to the greater frequency
of 0 than of 5. The question is whether the interval can be so
enlarged, without the loss of all details of the nature of the
distribution, that every class will represent the influence of the
same conditions. For this purpose only three intervals are possible:
10, 30, and 60 seconds. Of these 30 and 60 are undesirable because the
interval 60 gives classes which are so large that all details of the
distribution are lost, while 30 exhibits only a few details without
doing away with the periodicity due to the preference for multiples of
60.

The further question remains, with which digit should the interval end,
in order that uniformity of conditions for the various classes may be
gained? Theoretically there are ten possibilities, but of these all
except two, 0 and 5, are excluded by reason of the unequal frequency
of the various digits already discussed. In favor of 0 is the fact
that all the classes thus formed are of equal size, _i. e._ 1-10,
11-20, etc., whereas for 5 the first class would differ from the others
in being only half as large, 1-5. This, however, is only a slight
disadvantage, for there are very few judgments which fall in this
class. On the other hand, since 0 is the final digit of most frequent
occurrence, classes ending in 5 have the advantage of placing the value
of greatest weight in the middle. On the whole it seemed desirable to
arrange the judgments in 10 second classes, beginning with the class
1-10. But for purposes of comparison the male judgments have been
distributed in classes of 10 seconds, which end in 5, 1-5, 6-15, 16-25,
etc.[129]


TABLE 7

DISTRIBUTION OF MALE JUDGMENTS IN 10" CLASSES

  _Classes_      18̋            36̋           72̋             108̋        C
             I   E   R   W  I   E  R  W   I   E   R   W   I   E   R   W

   1 - 10"  31  16  69 136  3   1  4 18   1   1   1   4   0   0   0   0   0
  11 - 20  174 179 142 108 40  34 45 97   4   2   5  23   1   2   1   7   0
  21 - 30   34  47  35   7 99 106 98 82  15   9  26  65   5   0   6  23   0
  31 - 40    7   7   3     56  69 69 28  25  16  34  47  10   2   6  31   1
  41 - 50    4   2   2     37  23 22 18  41  43  54  33  20   9  19  28   0
  51 - 60    1              9  13  9  6  64  68  53  35  31  17  27  32   7
  61 - 70                   5   3  0  0  37  42  28  18  27  26  22  19   8
  71 - 80                   0   1  3  1  23  34  16   9  33  34  35  20  34
  81 - 90                   1      1  1  15  15  15   8  30  40  43  29  58
  91 -100                   0             9   8   7   1  23  35  25  17  65
  101 -110                  0             2   6   3   3  10  26  17   6  33
  111 -120                  0             8   5   5   2  29  16  17  16  29
  121 -130                  0             3   1   2   0   5  15   7   8  10
  131 -140                  0             1   1   1   1   6   9   5   0   4
  141 -150                  1             0       1   0   5   5   8   4   1
  151 -160                                1           0   2   3   2   2   1
  161 -170                                1           0   2   1   0   2
  171 -180                                1           1   5   4   5   2
  181 -190                                                0   1   2   1
  191 -200                                                3   2   2   2
  201 -210                                                0   1   1   0
  211 -220                                                0   0   0   0
  221 -230                                                1   1   0   0
  231 -240                                                2   0   1   0
  241 -250                                                1   1       1
  251 -260                                                    0       0
  261 -270                                                    0       0
  271 -280                                                    0       0
  281 -290                                                    0       0
  291 -300                                                    1       1

  Totals
       251 251 251 251 251 250 251 251 251 251 251 250 251 251 251 251  251

As a result of these groupings of the male judgments it appeared
that the former method gives a far more regular distribution than
the latter. In view of this result and the above considerations,
Tables 7 and 8 were constructed by the use of 10-second classes,
beginning with 1-10. In these tables (column C) the distribution of the
letter-counting results has been included for convenience of comparison
of the two kinds of judgments as to form of distribution.

As instances of the general form of distribution of the judgments the
curves have been plotted for letter-counting, Fig. 1 _A_. (Males ----,
Females, ... ,) for idleness 36 seconds, Fig. 1 _B_, and for idleness
108 seconds, Fig. 2. The distribution of the letter-counting judgments
in classes of 10 is very regular in comparison with that of the several
time-estimation judgments. For idleness 36 seconds there are several
fairly distinct modes, and for idleness 108 seconds the modes are still
more numerous and more marked.

[Illustration: Fig. 1_A_. Distribution of the Results of the
Letter-Counting Test. Males ----, Females ....]

[Illustration: Fig. 1_B_. Distribution of the Time-Estimation Judgments
for the Interval Idleness 36̋. Males ----, Females ....]

[Illustration: Fig. 2. Distribution of the Time-Estimation Judgments
for the Interval Idleness 108̋. Males ----, Females ....]


TABLE 8

DISTRIBUTION OF FEMALE JUDGMENTS IN 10̋ CLASSES

                  18̋               36̋             72̋               108̋
  _Classes_
             I   E   R   W  I   E   R   W  I   E   R   W   I   E   R   W  C

    1 - 10  72  20  88 133  5   6   5  33  2   0   1   6   0   1   0   2  0
   11 - 20 114 119 109  94 33  29  40  83  7   5  15  25   2   2   7  10  0
   21 - 30  50  87  54  29 70  57  73  74 16  13  25  52   4   1   3  27  0
   31 - 40  17  17  10   6 51  87  53  30 24  21  24  30   2   1  11  14  0
   41 - 50  10  14   7   4 40  35  39  20 35  18  37  32  14   6  17  34  2
   51 - 60   2  10   3   1 43  30  40  14 48  44  47  53  30  18  35  36  5
   61 - 70   1   3   0   3 11   7   5   2 28  50  20  21   8  11  23  22 18
   71 - 80   3   2   1   2  5   8   2  10 29  32  23  14  14  21  20  14 57
   81 - 90   0   0   1   1  9   3   6   2 27  29  28  14  41  35  35  19 67
   91 -100   0   2   0      1   4   2   1 16  11  15   5  25  35  17  11 50
  101 -110   0       0      0   1   1   1  4   7   6   0   9  24   7   6 25
  111 -120   0       0      2   3   6   3 17  20  10   8  52  38  33  16 19
  121 -130   2       0      0   1   0      3   4   3   3   4  15   6   8 13
  131 -140           0      1   1   0      1   2   1   4   6   9   6   6  5
  141 -150           0      1   0   1      5   4   4   2  15  13  16   7  1
  151 -160           1      1   0   0      0   4   2   0   3   7   2   2  2
  161 -170                  0   0   0      1   1   0   0   3   2   0   1
  171 -180                  1   0   0      7   4   7   4  22  14  16  22
  181 -190                      0   0      1   0   1   0   2   2   3   0
  191 -200                      0   0      1   1   3   0   3   5   6   6
  201 -210                      1   0      0   0   0   0   1   0   1   0
  211 -220                          0      0   1   0   0   1   2   0   0
  221 -230                          0      0   1   0   0   1   0   6   2
  231 -240                          0      2   1   0   1   3   3   2   4
  241 -250                          0          1   0       0   3   0   0
  251 -260                          0              0       2   0   0   1
  261 -270                          0              0       1   0   0   0
  271 -280                          0              0       0   0   0   0
  281 -290                          0              0       0   0   0   0
  291 -300                          1              2       4   2   1   1
  301 -310                                                 0   0   0   0
  311 -320                                                 1   0   0   0
  321 -330                                                 0   0   0   0
  331 -340                                                 1   0   0   1
  341 -350                                                     0   0
  351 -360                                                     1   1
  361 -370                                                     0
  371 -380                                                     0
  381 -390                                                     0
  391 -400                                                     2

  Totals
        271 274 274 273 274 273 274 273 274 274 274 274 274 273 274 272 264

Tables 7 and 8 show that the range of the judgments increases with the
length of the interval judged, and that the modal class is always much
nearer the lower than the upper limit. Asymmetry is characteristic of
the distribution of organic data, and in certain instances, as for
example writing 18 seconds, males, the choice of a 10-second class
interval results in extreme asymmetry, and one is reminded of the
tables which Fechner gave as examples of his logarithmic method in
statistics.[130]

It is not to be expected that a method of grouping should be found
which will give regularity of distribution throughout, but it is
important that there should be regularity about the mode. In the table
of distribution for the males (Table 7) all the intervals from idleness
18 seconds to reading 72 seconds are regular.[131] The remaining
intervals, with the exception of estimating 108 seconds, are irregular.

Trial proves that for these intervals increase of the size of the
class to 30 seconds is sufficient to give regular distributions, as is
obvious from Table 9. Grouping by 30-second classes gives regularity
for most of the female judgments, but for idleness 108 seconds and
writing 108 seconds there are still slight irregularities, as Table 10
indicates.

Tables 7 and 8 show that the distribution is far less regular for
the females than for the males. The fact that it becomes regular
when the class interval is increased to 30 seconds suggests that the
irregularities of distribution which appear in the tables are due to
those influences which favor simple fractions of a minute and not to
the small number of judgments.

Having now noted certain important characteristics of the
time-estimation judgments and the nature of their distribution, we
may examine the arithmetical means and other statistical quantities
which have been determined for our data. Those quantities which have
been determined for the several ages, intervals, and fillings as well
as for the sexes are: (1) The Mean (_M._ in tables), (2) the average
variability (_M. V._), (3) the positive variability (+ _V._), (4)
the number of judgments with positive variation (No. + _V._), (5)
the negative variability (- _V._), (6) the number of judgments with
negative variation (No. - _V._), (7) the relative variability (_R. V._)
= _M_ + _V_/_M_ × 100.

Since the sums of the positive and the negative variations are equal,
it is possible to make certain of the accuracy of the means and
average variabilities by comparison of the + _V._ and - _V._ As this
was done in all cases we feel confident of the reliability of the
statistical quantities presented in the tables.

In Table 11 have been arranged the various quantities as determined for
the males and females for each interval. The values given in this table
are averages of the determinations made for the several ages separately.


TABLE 9

DISTRIBUTION OF CERTAIN MALE JUDGMENTS IN 30̋ CLASSES

    _Classes._  W. 72̋    I. 108̋    R. 108̋    W. 108̋

      1- 30       92          6          7         30
     31- 60      115         61         52         91
     61- 90       35         90        100         68
     91-120        6         62         59         39
    121-150        1         16         20         12
    151-180        1          9          7          6
    181-210                   3          5          3
    211-240                   3          1          0
    241-270                   1                     1
    271-300                                         1

                 250        251        251        251


TABLE 10

DISTRIBUTION OF FEMALE JUDGMENTS IN 30̋ CLASSES

       18̋              36̋              72̋             108̋
  _Classes_  I   E   R   W    I   E   R   W   I   E   R   W   I   E   R   W
    1- 30  236 226 251 256  108  92 118 190  25  18  41  83   6   4  10  39
   31- 60   29  41  20  11  134 152 132  64 107  83 108 115  46  25  63  84
   61- 90    4   5   2   6   25  18  13  14  84 111  71  49  63  67  78  55
   91-120    0   2   0        3   8   9   5  37  38  31  13  86  97  57  33
  121-150    2       0        2   2   1       9  10   8   9  25  37  28  21
  151-180            1        2   0   0       8   9   9   4  28  23  18  25
  181-210                         1   0       2   1   4   0   6   7  10   6
  211-240                             0       2   3   0   1   5   5   8   6
  241-270                             0           1   0       3   3   0   1
  271-300                             1               2       4   2   1   1
  301-330                                                     1   0   0   0
  331-360                                                     1   1   1   1
  361-390                                                         0
  391-420                                                         2

  Totals  271 274 274 273  274 273 274 273 274  274 274 274 274 273 274 272


TABLE 11

MEANS, ETC., FOR EACH SEX, INTERVAL AND FILLING

    KEY:  Ma. = _Males_
          Fe. = _Females_

                 M.           M. V.          +V.          No. +V.
             Ma.   Fe.      Ma.   Fe.      Ma.  Fe.      Ma.   Fe.

        {I  17.7  20.74    5.4  10.35     5.8  15.48    15.9  22.75
    18" {E  19.5  25.55    4.9   9.84     6.0  13.98    16.9  23.75
        {R  15.5  18.46    4.9   9.14     6.3  13.03    14.6  24.25
        {W  11.5  15.58    3.7   8.56     4.6  11.83    14.3  25.75

        {I  33.3  42.81    9.1  16.55    12.6  20.63    14.2  28.00
    36" {E  33.1  41.54    8.4  15.23    10.7  23.34    14.2  23.00
        {R  32.1  41.71    8.4  16.42     9.7  20.78    16.3  27.75
        {W  24.7  30.10    9.0  14.71    10.9  22.08    14.4  23.50

        {I  63.3  73.00   17.2  27.20    20.9  33.72    15.7  27.75
    72" {E  63.1  77.13   16.0  26.56    18.8  37.27    16.9  24.75
        {R  57.9  70.78   17.3  30.30    20.9  39.28    15.2  26.75
        {W  51.2  54.93   19.8  24.21    23.7  29.47    15.0  38.25

        {I  92.7 113.37   29.8  40.13    35.3  44.08    15.2  32.00
    108"{E  99.8 114.88   26.3  36.38    29.3  49.26    14.9  26.25
        {R  90.1 100.47   28.3  40.18    33.9  47.90    15.3  34.25
        {W  75.5  87.45   32.4  45.33    40.8  61.12    14.9  34.25

                -V.         No. -V.       R. V.
             Ma.   Fe.     Ma.   Fe.     Ma.  Fe.

        {I   5.7   7.78   20.0  45.00    30  49.83
    18" {E   4.4   7.69   19.0  44.75    25  39.10
        {R   4.4   7.07   21.2  44.25    31  49.26
        {W   3.2   7.14   21.6  42.50    33  54.55

        {I   8.1  14.01   21.7  40.50    27  38.32
    36" {E   7.2  11.57   21.7  45.25    25  36.66
        {R   7.5  13.84   19.6  40.75    26  38.84
        {W   8.7  11.24   21.5  44.75    36  48.64

        {I  15.5  22.83   20.2  40.75    27  36.87
    72" {E  14.7  20.98   19.0  43.75    27  34.43
        {R  15.7  24.91   20.7  41.75    30  42.77
        {W  16.0  21.26   20.9  36.75    37  43.84

        {I  25.9  37.87   20.7  36.50    32  35.34
    108"{E  24.4  29.51   19.9  42.00    26  31.57
        {R  24.8  34.60   20.6  39.75    31  39.95
        {W  27.5  35.86   19.9  42.75    42  51.67

The following facts are revealed by comparison of the results for the
two sexes.[132] Without exception the means for the females are larger
than those for the males. All but one of the sixteen intervals (E 18̋)
are underestimated by the males, whereas all but six are overestimated
by the females. The amount of over- and under-estimation is given in
Table 12. In every instance the females overestimate in comparison with
the males. The mean variability is very much greater for the females,
as is also the relative variability. If variability be taken as a
measure of reliability of judgment the males are far superior to the
females.

As is obvious from Table 12, both under- and over-estimation increase
with increase in the length of the interval. For 18̋ intervals they are
least, for 108" intervals greatest.

The influence of the fillings is marked. The writing intervals without
exception are judged as shortest; reading gives the next shortest
intervals, while sometimes idleness, sometimes estimating, comes third.
In order of increasing length of average estimates of the intervals the
fillings stand: writing, reading, idleness, estimating. As a rule the
averages for the idleness and the estimating intervals are nearly the
same, but it is worthy of note that the females always overestimate
to a greater extent when estimating than when idle. This is another
indication of the discrepancy between the female time-unit and the
objective unit.

About ninety per cent of the subjects estimated by some counting
method. The methods most frequently used were "counting seconds,"
counting "1 and 2 and 3 and 4, etc.," counting the swings of a
pendulum, tapping, and counting imaginary watch-ticks.

The above statements might suggest that the overestimation
characteristic of the female judgments is due to a more rapid counting
rhythm. This, however, is not true, for the letter-counting tests
indicate a slightly more rapid rhythm for the males, 93.42 as opposed
to 91.89.


TABLE 12

AMOUNT OF OVER- AND UNDER-ESTIMATION OF INTERVALS

       _Males._ _Females._    _Males._
      {I - 0.3 + 2.74    Age       17    18    19    20    21    22     23
      {E + 1.5 + 7.55 Over- or under-                                 -8.55
  18̋ {R - 2.5 + 0.46 estimation -3.60 -7.50 -13.02 -13.49 -8.71 -13.20
      {W - 6.5 - 2.42    No. +      6     2     2     0    2     2      2

      {I - 2.7 + 6.81    No. -    10     14    14    16   14    14     14
      {E - 2.9 + 5.54
  36̋ {R - 3.9 + 5.71
      {W -11.3 - 5.90     _Females._

      {I - 8.7 + 1.00
      {E - 8.9 + 5.13     Age              17     18      19      20
  72̋ {R -18.1 - 1.22     Over- or under-
      {W -24.8 -17.07      estimation   + 2.38 - 4.16  - 2.57  + 2.55

      {I -15.3 + 5.37     No. +          11      8       8      12
      {E - 8.2 + 6.88     No. -           5      8       8       4
  108̋{R -17.9 - 7.53         + = overestimation
      {W -32.5 -20.55         - = underestimation

In Table 13 are the means, variabilities, errors, etc., for the
letter-counting tests. In this table we have presented the values of
the various statistical quantities for the several ages of both males
and females, for there are certain interesting differences which
should be noted. Similar age-tables have been prepared for all of the
other results, but in no case have noteworthy differences appeared.
From Table 13 it will be observed that the males on the average
count more letters in thirty seconds than do the females, and at the
same time make more errors. The mean and relative variabilities for
the sex-groups are almost the same. Curiously enough the number of
letters counted as well as the accuracy of counting decrease for the
females with age (17 to 20 years being the age-limits of the group
under consideration). The males within the same age-limits increase in
rapidity of counting, but decrease in accuracy. In the examination of
Table 13 it is to be remembered that the 17 and 23 year groups of males
contain only 16 individuals each, and therefore cannot be compared to
advantage with the other groups. Both mean and relative variabilities
decrease from 17 to 20 years for the females, whereas for the males
there is no constancy in the direction of the change.


TABLE 13

MEANS, ETC., FOR LETTER-COUNTING

                                     _Males._

    _Age_      17      18      19      20     21      22      23  _Average_
    M.       92.81   92.48   94.00   95.87  92.66   97.06   89.06    93.42
    M. V.    10.29   15.31   16.50   12.90   7.81   13.04   19.20    13.58
    +V.      11.76   22.96   16.50   11.70   8.13   16.30   21.94    15.61
    No. +V.   7       9      20      37     24      14       7       16.96
    -V.       9.15   11.48   16.50   14.40   7.51   10.87   17.09    12.57
    No. -V.   9      18      20      30     26      21       9       19.00
    R. V.    11.1    16.5    17.6    13.5    8.4    13.4    21.6     14.59
    Errors    0.88    1.04    2.10    3.03   1.70    0.89    2.75     1.77

                                      _Females._

    _Age_           17            18            19          20    _Average_
    M.           97.74         95.51         89.21       85.09      91.89
    M. V.        17.06         14.56         11.30       11.28      13.55
    +V           18.67         18.49         12.06       14.56      15.94
    No. +V.      32            21            30          31         28.50
    -V.          15.71         12.14         11.66        9.21      12.18
    No. -V.      38            32            31          49         37.50
    R. V.        17.5          15.2          12.7        13.3       14.70
    Errors        1.35          1.46          1.49        1.65       1.47


SUMMARY

(1) The length of a second is slightly overestimated by men, greatly
underestimated by women.

(2) Intervals of from 18 to 108 seconds are usually slightly
underestimated by men (ages 17 to 23 years), and greatly overestimated
by women (ages 17 to 20 years).

(3) The time-estimates of women are far more variable than those of
men, and on the whole markedly less accurate.

(4) Both men and women favor estimates which end in 0 or 5, as well as
simple fractions of a minute, but the tendency is stronger in women
than in men. Over one third of the estimates reported in this paper
were 15 seconds or simple multiples thereof.

(5) In letter-counting the groups of subjects studied (251 men and
274 women) exhibited differences just the opposite of those in
time-estimation, for the men counted more rapidly and less accurately
than did the women.

(6) Of the four fillings for the intervals used in the experiments,
"writing" gave the smallest estimates of the intervals, listening to
"reading" next, while "idleness" and "estimating" were conditions in
which the intervals seemed much longer to both men and women.

(7) This preliminary study of sex-differences in time-estimation, by
which it has been learned that women overestimate and are notably
inaccurate in comparison with men, is to be followed and supplemented
by the results of an investigation now in progress concerning
the relations of sex-differences in time-estimation to age and
physiological rhythms.

FOOTNOTES:

[Footnote 125: Science, N.S., vol. 19, pp. 708-709, 1904.]

[Footnote 126: E. Mach (Analyse der Empfindungen, 1900, p. 161) thinks
that his physiological time-unit has become larger with age. And he has
also noted that the time-sense differs in animals of the same species
which differ in size.]

[Footnote 127: The unbracketed number is that for the males, the
bracketed that for the females.]

[Footnote 128: It may be asked in connection with the above statistics,
what is the probability that even and uneven digits have unequal
chances of occurrence? The analytic treatment of this problem
indicates that the probability is ½[1+φ(4.25)] = 0.99999999891 for
the males and ½[1+φ(15.93)] for the females. It is therefore almost
certain, theoretically, that even and uneven numbers have not the same
frequency. See Czuber, Wahrscheinlichkeitsrechnung, 1903, pp. 158-161.
The constants of this problem are in Czuber's denotation, _a_ = 52.308
and _b_ = 222.]

[Footnote 129: The judgments might be grouped in 10 second classes
beginning with the lowest number in the experiments, but this would
have the disadvantage of rendering the distribution tables or curves
for different groups of judgments incomparable.]

[Footnote 130: G. Th. Fechner, Collectivmaasslehre, pp. 338-346, 1897.
Compare the asymmetry of writing 18 seconds for both males and females
with Fechner's Table III, p. 442.]

[Footnote 131: The distribution may be called regular if the
classes increase to a maximum, the mode, and then decrease without
interruption. (See Fechner, Collectivmaasslehre, pp. 121-122.)]

[Footnote 132: The values of the means and variabilities was determined
to the second decimal for the females, but only to the first for the
males; consequently the relative variability could be computed exactly
to the first decimal for the females and to the unit only for the
males.]




ASSOCIATIONS UNDER THE INFLUENCE OF DIFFERENT IDEAS

BY BIRD T. BALDWIN


The purpose of the following investigation was to study the influence
of two or more starting-points on the train of associated ideas. It
was begun in October, 1902, and concluded in January, 1905. Nineteen
graduate students acted as subjects, and the experiments were conducted
with each individually for one hour per week, except in a few instances
where two subjects were present together. Occasionally the experimenter
acted as subject in order to get a clearer insight into introspective
data. There are recorded here thirteen groups of one hundred and eight
sections, including eight hundred and fifty-five separate experiments,
with a sum total of eleven thousand, one hundred and thirty-five named
associations.

So far the field has not been studied experimentally, although
Cordes,[133] while attempting to observe the effect of an unconscious
intermediating factor in 'mediate association,' noticed that the
accompanying factor (particularly with a tone) sometimes combined with
the starting-point in determining the associated series. The fact is
simply mentioned, and Scripture's[134] conclusions from the same are
inadequate when he states, "In general we may say that two simultaneous
ideas have an effect that depends on their relative masses; if one of
the ideas is over-poweringly weighty the next idea will be chiefly
influenced by it, but if the two are nearly balanced the next idea will
be the result of the two." Miss Calkins[135] makes a near approach to
the problem in a study of "Mental Combination," where auditory words
were given as nearly simultaneously as possible and the subject was
required to remain for four seconds in silence and then to write the
train of imagery which passed in six following seconds; the words as
auditory starting-points were finally excluded from the experiments
"because the first word pronounced tends often to establish itself so
firmly that its association-images are proof against the intrusion of
the second word, which has therefore no chance to be grasped with the
first." These results differ from the ones here recorded which were
obtained independently and under different conditions. In her valuable
monograph on Association[136] there is in one instance a suggestion of
the problem, but no results are given.

The two or more starting-points were nonsense syllables, concrete or
abstract words, pronounced by the investigator or shown on cards, or
finally, pictures. The associations continued for a period of fifty
seconds unless otherwise indicated. The same letters denote the same
subjects throughout all the experiments, K. and R. being women. In the
examples given, Roman numbers have been used, in place of the letters,
as it was found impossible to get representative series which were
entirely devoid of personal references.

After the associations had been "jotted down," the subjects in each
case copied them, noting also the suggesting ideas, and giving, in many
cases, a number of introspective notes. It is important to mention that
in all of the work these records of suggesting ideas were made by the
subjects without any questions on the part of the investigator.


GROUP I. TWO NONSENSE SYLLABLES

In order that a standard might be obtained that would serve as a basis
for subsequent experiments, two nonsense syllables were pronounced with
equal emphasis, as starting-points, and the subject was asked merely to
wait until both had entered consciousness and not to favor nor inhibit
either. If the subject stated that associations arose between the
pronunciation of the starting-points the results were discarded.

Turning to the notes we find that K. reported, "The first just hovered
around," and the associations followed the second, or as V. states
it, "Sof was carried subconsciously for a while without influence,"
while By. frequently noticed that "The first may be in the background
on the point of breaking in but be inhibited involuntarily," or in
another, "It was frequently present but exerting no influence." A.
"forgot what the other word was, but felt it trying to get a hand in."
Ro. goes further and states in terms similar to A., "The words which
might have been suggested by 'taf' were wholly inhibited, though I
am sure 'taf' was present to me throughout the series, but was not
efficacious as against the other syllable." It is unnecessary to
multiply instances to show that the wraith, so to speak, of one may
linger for a short time, or occasionally for the entire series, but
that the associations are determined by the other. Or the lingering
wraith of a starting-point may hover ineffectively in consciousness
and exert no determining influence on the series until toward its
close; then, however, although both starting-points still persist,
they exchange places so far as effectiveness is concerned, and the
former wraith becomes the primary factor in effecting the series of
associations. For example, "The two syllables were present all the way
through, but the former exhausted itself as a word-suggester after
a few words." Again Ro. gives examples where "One alone may be in
consciousness and determine the associations when the other may enter
without effect upon the series, or may determine the series, the first
persisting but losing its effect"; and a case of both ideas persisting
and each alternately exerting an influence was noted by M. These last
two conditions are exceptional.

Studying carefully the notes and graphic representations as given in
the tables, it is apparent that one starting-point may be followed
independently and exclusively; we find "'fef' suggesting 'theft,' and
the subsequent associations monopolizing the field of consciousness
and 'tuz' not again appearing." (V.) The results show that in the one
hundred and seven experiments one was followed exclusively only twenty
times. The following would be an example:


III. Naf-Tam

    --tambour            tam
    --experiment         tambour
    --psychology         experiment
    --laboratory         experiment
    --space              psychology
    --time               space
    --ego                space and time
    --Ebbinghaus         ego
    --discussion         psychology
    --posted notice      psychology
    --door               posted notice
    --gray               door
    --stairs             door
    --fishes             laboratory
    --water              fishes
    --decay              fishes
    --animal             fishes
    --meat               fishes
    --odor               meat

In other cases one starting-point may be followed and then the
other for a few or many associations, each "occupying the mind to
the exclusion of the other for a number of words, when suddenly the
other appears and suggests a new series," which V. calls "a curious
zigzagging in consciousness." This may take place at any point along
the series. As Ro. indicates, "There was rivalry here; 'yud' wholly
inhibited all the 'zid' associations, even 'zid' itself for the time,
then, 'yud' apparently being exhausted, 'zid' entered again into the
focus."

In contrast to the above forms, the tables indicate and the notes
verify that while the associations primarily follow one there can be
traced the modifying effect of the other; "The nature of the words
was influenced by a fringe of consciousness which contains the other
starting-point," M.'s comment runs, while J. adds, "The other helped
in the mental picture but had no special significance." There are no
experiments in the first group where this partial fusion took place
throughout the series, and there are only six examples where it took
place before the sixth word and continued to the end. While there are
no cases here of alternating partial fusion, yet it sometimes happens,
that is, the predominating influence is sometimes transferred from one
to the other. The modifying influence may reach such a degree at any
point in the series that the identity of each point of departure of the
association is lost and there is total fusion, resulting in one idea or
one train of thought. For example, 'rel' and 'mem' gave 'realm' for V.;
for Ht. 'fef' and 'tuz' gave 'fezz'; for M. they gave 'fez'; for Ro.
'fuz.'

Of the twenty-one experiments where one starting-point was followed
exclusively, there were two cases for the first and nineteen for the
second. There were but two cases of alternating groups, a third one
being unique in that after the first three words there was a persistent
alternation between two apparently independent series of inter-related
words and inter-related visual images. There were three cases of
complete fusion throughout the series and nine where total fusion
started before the fifth word and continued.

In a number of cases the initial starting-points were forgotten before
the end of the series and the subjects were sometimes unable to recall
them.

As the subjects were kept ignorant of the characteristic persistence
of either starting-point in forming separate associations, and of the
tendency towards fusion, as well as of the preponderance of influence
of one over the other, it is to the tabulated results that we must
look for quantitative measurement of these. 76.2% of the associations
are due to the independent influence of one or the other of the
starting-points; 23.9% are due to the combined influence; for the first
starting-point we have 18.7%, for the second 57.5%. In 15.3% of all the
associations the combined influence was the result of total fusion; in
4.1%, the result of a union where the first starting-point predominated
in influence; in 4.4%, the second predominated. The following graphic
representation indicates the order of influences for the first pair of
syllables. The table gives the results for the ten pairs.


TABLE I. TWO NONSENSE SYLLABLES SPOKEN

  Time--50 seconds.

  Characters: | represents the influence of the first, - the second, +
  complete fusion, α fusion with the first predominating, β fusion with
  the second predominating.

                    1) Taf--Coz

    M.   | - - - - | | | | | - - - -
    F.   - - | | + + + + | | | + + β β β β β β
    H.   - | | | | | | | | | | | | | | |
    R.   | - - - - - - - - - - - - - - -
    K.   - - - - - - -
    V.   | | - - - - | - - | + + + +
    Ro.  | | | - - - - - - - - - - - - - - - -
    Bl.  - - - - - - - - - - - - - - -
    By.  | - | | | - - - - - - - - - - - - - - - - - - -
    Bs.  - - - - - - - - - - - - - - -
    Ht.  - - | | | | | | | | | | | | |

    Starting-points:

     (1) Taf--Coz.
     (2) Cim--Bef.
     (3) Yud--Zid.
     (4) Sof--Deb.
     (5) Naf--Tam.
     (6) Fef--Tuz.
     (7) Sar--Nef.
     (8) Sek--Lub.
     (9) Hov--Bes.
    (10) Rel--Mem.

             |         -         +         α        β

    M.    { 20         69        26        4        3
          { 16.5%      56.6%     21.2%     3.3%     2.4%

    F.    { 28         22        37        0       39
          { 22.2%      17.5%     29.4%             30.9%

    H.    { 39         98         0       11        0
          { 26.3%      66.2%               7.5%

    R.    {  9        102        32        0        0
          {  6.3%      71.4%     22.3%

    K.    { 47         58         1       14        0
          { 39.1%      48.3%       .7%    11.9%

    V.    { 30         79        54        0        0
          { 18.4%      48.4%     33.2%

    Ro.   { 32        111         0        5        0
          { 21.4%      75.1%               3.5%

    Bl.   {  5        131        15       15        0
          {  3.1%      78.9%      9.0%     9.0%

    By.   { 43        102        23       15       22
          { 16.8%      52.6%     11.7%     7.7%    11.2%

    Bs.   {  9         72        31        0        6
          { 11.4%      58.5%     25.3%              4.8%

    Ht.   { 32         58        20        0        0
          { 29.0%      52.8%     18.2%

    Totals{294        902       239       64       70
          { 18.7%      57.5%     15.3%     4.1%     4.4%

  Number of subjects, 11; number of sections, 10; number of
  experiments, 110; number of associations, 1569.


GROUP II

Two concrete nouns with apparently equivalent connotation were
pronounced. In order that the subject get no clue as to the
preponderance of one starting-point over the other, the nine sections
were given at irregular intervals in connection with other experiments.

The tables show few cases of fusion, there being but one case in the
eighty experiments where all the associations were the result of the
combined influence of both starting-points. There was no other instance
where complete fusion took place before the sixth association and
continued throughout, and only six cases where any form of fusion took
place in the eighty experiments within the first five words (7.6%),
while in the one hundred and seven experiments of Group I there were
twenty-five (23.9%).

The tables show a tendency which was noted when two syllables were
used and which is emphatically brought out here, namely, it is the
position or sequence which determines which of the two equivalent
starting-points shall produce the greater influence. When the two
starting-points are given in immediate succession, it is the second
which predominates in influence. This preponderance may be clearly
demonstrated by the tables alone, though there are many notes which
show "the greater influence of one" (the second). A. seems to have
realized this when he wrote, "While the first was uppermost the second
hovered in subconsciousness, resting content, knowing that it would
have its turn soon." In corroboration the tables give--abstracting from
all cases of fusion--268 words (23.2%) for the first, 721 words (62.3%)
for the second. In fifty-five of the eighty experiments the last had
full control at the end of the series. Taking the cases where one
was followed exclusively throughout the series, starting between the
first and the sixth named association, we find eight for the first and
twenty-nine for the second. This greater influence is again to be found
in the cases of partial fusion; the first predominating in 2.6%, and
the second in 4.3%.

The results of both groups show that the starting-points tend in a high
degree toward independent influence, and also that such a method of
presentation is one of sequence rather than simultaneity, as the two
words show unequal influence. The former must in a manner be reproduced
to become a point of departure for associations, while the latter acts
directly. Is it then a condition of mind that when similar impressions
are presented the one in the presence of which consciousness is
reacting directly has a greater influence in arousing associations than
one which is just past? This we are forced to conclude is the case, but
the proof of the conclusion will be supplemented by later results.


TABLE II. TWO WORDS SPOKEN

    Time--50 seconds.
    Characters--same as Table I.

                      (1) Library--River

    M.    |  +  +  β  -  -  -  -  -  -  -  -  -  -  -  -
    F.    |  -  -  -  -  -  -  -  -  -  |  |  |  |
    H.    -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -
    Ro.   |  |  |  |  -  -  -  -  -  -  -  -  -
    J.    |  -  |  |  |  |  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -
    Bl.   |  |  -  -  -  -  |  |  |  |  |  |  |  |  |  -  -  /
    By.   -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -
    Bos.  |  -  -  -  -  -  -  -
    Bur.  -  +  -  -  -  -  -  -  -  -  -  -  -

    (1) Library--River.
    (2) Hat--Road.
    (3) Newspaper--Medicine.
    (4) Bicycle--Drum.
    (5) Theatre--Magazine.
    (6) Store--Church.
    (7) Soldier--House.
    (8) Ship--Boy.
    (9) Furniture--Tree.


TABLE II.--_continued_

                |      -      +      α      β

    M.        {32     66     26      0     13
              {23.4%  48.2%  18.9%          9.5%

    F.        {32     67     13      5      1
              {27.2%  56.7%  11.0%   4.3%    .8%

    H.        {59     79      5      0      0
              {41.3%  55.2%   3.5%

    Ro.       {44     85      0      0      0
              {34.1%  65.9%

    J.        {21    121      8      5      0
              {13.5%  78. %   5.3%   3.2%

    Bl.       {22     83      6      0      1
              {19.7%  74.1%   5.4%    .8%

    By.       {16     97     18     14     23
              { 9.5%  57.8%  10.7%   8.3%  13.7%

    Bs.       {26     64      3      7      1
              {25.8%  63.5%   2.9%   6.9%    .9%

    Bur.      {16     59      8      0     10
              {17.2%  63.4%   8.6%   0     10.8%

    Total    {268    721     87     31     49
             { 23.2%  62.3%   7.6%   2.6%   4.3%

  Number of subjects, 9; number of sections, 19; number of
  experiments, 80; number of associations, 1156.


GROUP III

For two syllables and like words the second strongly predominates in
awakening associations; will the same be true when two simple outline
pictures are shown in the same order? The following results show in
attestation of the above conclusion percentages remarkably similar
to those of the words. There are 23.7% for the first and 61% for the
second; 84.7% of the associations show no sign of fusion and only 15.3%
for fusion and the different forms of partial fusion, the second still
holding the ascendency. When the two pictures did fuse, R. tells us,
"All the associations were more elaborate pictures than when mere words
were given."

A comparative study, inadequate though it is, offers a partial
parallelism between the predominating memory type and stimulation. The
subjects who are preëminently of the visual, K., visual motor, M., H.,
V., and visual lingual motor, Ht., find the pictures more suggestive
than the syllables, while those of the pure motor, F. and By., reverse
the order and find that the syllables offer more numerous and "more
vivid" (F.) associations.

Common experience, however, immediately shows the limitations of
reaching conclusions, when considering prolificacy as criteria of
suggestiveness, inasmuch as starting-points which are abundantly rich
in associations tend to produce so many points of departure that they
tend to inhibit one another. Again there is to be considered the kind
of associations, those of the syllables being of a very elementary
character and in serial form.

A few experiments were given in which colored slips of paper were used
as starting-points. These proved very suggestive for the subjects. Also
a few tones were given, but these were soon discontinued, as other
tones, which could not be recorded, were frequently suggested. The
sentence which forms a very satisfactory starting-point where one is
used could not be used to an advantage where several were given. As we
are here interested in the mutual influence of the starting-points,
our remaining study will be confined to a quantitative and qualitative
variation of the forms used in the previous groups.


TABLE III. TWO PICTURES SHOWN

    Time--50 seconds.
    Characters--same as Table I.

            (1) Boy Rolling Hoop--Blacksmith

    M.  | | | - - - - - - | | | | |
    F.  | | - - | + +
    H.  - - - - - - - - - - - -
    R.  - - - - - - - - - - - - - - - -
    K.  - - - - - - - - - - - - - - - - -
    V.  - - | | + + + + + + + + + + + + + +
    S.  - - - - - - - - - | | | | | | |
    By. | - - - - - - - - - - - - - - - - - - - - -
    Bs. - - - - - - - - -
    Ht. | | - - - - - - - -

    (1) Boy Rolling Hoop--Blacksmith.
    (2) Old Man With Umbrella--Bird House.
    (3) Carpenter--Mower.
    (4) Children Playing--Boy with Basket.
    (5) Horse--Dog House.
    (6) Shoemaker--Fisherman.
    (7) Little Girl--A Chicken.
    (8) Boy--A Sheep.

             |      -      +         α      β

    M.     {41     65      2         0      1
           {37.6%  59.7%   1.8%              .9%

    F.     {11     42     21         0      5
           {13.9%  53.2%  26.6%             6.3%

    H.     {37     52     11        26      9
           {27.4%  38.5%  8.1%      19.3%   6.7%

    R.     {41     58      8         0     15
           {33.4%  48%     6.5%            12.1%

    K.     {32     82      0         0      0
           {27.1%  72.9%

    V.     {12      95    21         3      5
           { 8.8%   70%   15.4%      2.1%   3.7%

    S.     {36      44      0        0      0
           {45%     55%

    By.    { 8     111     12        0     13
           { 5.4%   77.2%   8.3%            9.1%

    Bs.    {13      40      7        0      0
           {21.7%   66.7%  11.6%

    Ht.    {22      72      7        0      4
           {20.9%   68.6%   6.7%            3.8%

    Totals {253    661     89       29     52
           { 23.7%  61%     8.2%     2.7%   4.4%

  Number of subjects, 10; number of sections, 8; number of experiments,
  76; number of associations, 1084.


GROUP IV

The previous experiments suggest the conclusion that that
starting-point in the presence of which consciousness is reacting
exerts a greater influence than one just past. The proof of this
anticipated result is supplemented by subsequent consectary data.

It is necessary to test the outcome when the nature of the impressions
is varied and the starting-points are given simultaneously. An outline
picture and a word may be so given. These experiments verify from a
different standpoint the statement that the picture establishes itself
more permanently and is more influential, there being 47.8% of the
associations produced by the picture alone, 14.8% by the words alone,
25% of a fusion of the two, and 12.4% of a uniting influence with
the first predominating. The subject R. is an exception in that she
favors the word when not favoring a combination of the two, a fact
which I am unable to explain except to add that the subject stated
that the æsthetic pleasure connected with the picture was sufficient
to inhibit the associations. K. has 100% for the pictures, which is
partially explained by the fact that she is a remarkable visualizer who
reproduces all situations in visual terms.


TABLE IV. PICTURE SHOWN AND WORD SPOKEN

    Time--50 seconds.
    Characters same as Table I.

              (1) (a rabbit)--Table

    M.  | - - | | | | | | | + - - - -
    R.  + + + + + + + + + + + + + + + + + +
    K.  | | | | | | | | | | | | | | |
    S.  | | | | | - | | | | |
    J.  | - - | | | | | | | | | - - - | | | | | | | | |
    Ht. | - | | | | | | | | | | | | - - - -

    (1) (a rabbit)--Table.
    (2) (horse)--Book.
    (3) (boy)--Dish.
    (4) (duck)--Iron.
    (5) (dog)--Paper.

              |       -       +        α         β
    M.     { 38      9       16        6         0
           { 55.0%   13.1%   23.3%     8.6%

    R.     {  1      20      55        0         0
           {  1.3%   26.3%   72.4%

    K.     { 65       0       0        0         0
           {100.0%

    S.     { 29      15       3       10         0
           { 50.8%   26.3%    5.3%    17.6%

    J.     { 75      19      14        9         0
           { 64.2%   16.3%   11.9%     7.6%

    Ht.    { 17       7      30       33         0
           { 19.5%    8.1%   34.5%    37.9%

    Totals {225      70     118       58         0
           { 47.8%   14.8%   25.0%    12.4%

  Number of subjects, 6; number of sections, 5; number of experiments,
  30; number of associations, 471.


GROUP V

The character of the starting-points determines which shall be followed
in a simultaneous presentation; can the character of two immediately
successive starting-points be so varied as to overcome the factor of
temporal difference? The few experiments in this group show an attempt
to shift and measure this influence, selecting as variants words of
general and specific character, the first being a term of everyday
experience with the subjects, of wide connotation, and therefore
presumably rich in associations, and the second, which had the
advantage of position, one of more limited and specific connotation,
thus in a measure making it possible to test the stability of the
second in establishing itself as a word-suggester. For the words
"Gymnasium--Stamens" the second still strongly predominates. There is a
slight increase in the amount of fusion (24.5% in all), the explanation
of which would probably lie in the fact that the first is more
assertive than before and offers more elements common to the two. There
was one case of total fusion, one case where fusion started before the
sixth word and continued, no cases of alternating fusion, four cases of
partial fusion, five cases where the second was followed throughout, no
case where the first was followed throughout, five cases of alternation
between the first and the second. In 75.5%, the influence was exerted
without fusion. Eight of the nine subjects favor the specific term.


TABLE V. GENERAL AND FAMILIAR WORD--SPECIFIC WORD SPOKEN

    Time--50 seconds.
    Characters--same as Table I.

                             (1) Laboratory--Rake

    M.   |  |  |  |  |  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -
    F.   |  -  -  -  -  -  -  -  -  -
    H.   +  +  +  +  +  +  +  +  +  +  +
    V.   |  -  -  -  -  -  -  -  -  -  -  -  -  -  |  |  |  |  |
    S.   -  -  -  -  -  -  -  -  -  -  |  |  |  |
    Bl.  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -
    By.  |  -  -  |  |  |  |  |  |  |  |  |
    Bs.
    Ht.  -  -  -  -  -  -  -  -
    J.   |  |  β β β β β β β β β β β β |  |  |  |  |

    (1) Laboratory--Rake.
    (2) Experiment--Butterfly.
    (3) Gymnasium--Stamens.


TABLE V.--_continued_

            { 22         25          4          6          0
    M.      { 38.6%      43.9%       7.0%      10.5%

            {  4         19          0          0          0
    F.      { 17.4%      82.6%

            {  4         25         13          0          0
    H.      {  9.5%      59.6%      30.9%

            {  7         38          2          0          7
    V.      { 12.8%      70.7%       3.7%                 12.8%

            { 11         34          0          0          0
    S.      { 24.4%      75.6%

            {  0         36          0          0          0
    Bl.     {           100.0%

            { 11         16          7          0          9
    By.     { 25.5%      37.1%      16.5%                 20.9%

            {  0          9         12          2
    Bs.     {            39.1%      52.2%       8.7%

            { 10         11         10
    Ht.     { 32.3%      35.4%      32.3%

            { 17          8         15          0         13
    J.      { 32.1%      15.1%      28.3%                 24.5%

            { 86        221         63          8         29
    Totals  { 21.2%      54.3%      15.4%       1.9%       7.2%

  Number of subjects, 10; number of sections, 3; number of experiments,
  27; number of associations, 407.


GROUP VI

What modification of influence takes place when an auditory impression
in the form of a word of abstract nature (the pure abstract words are
taken later) is given in comparison with a concrete noun? The tables
indicate that we are now able partly to overcome the disadvantage
of first position by the advantage of concrete content. 24% of the
associations are the result of total fusion; there are eight cases of
total fusion throughout the series, and seven where total fusion took
place before the third word and continued. There are two cases where
the first was followed throughout, and but one where the second was
followed exclusively. The tables indicate that there were very few
words. Some of the subjects claim, "These words did not seem equally
rich in associations. I was not at all conscious of the one while the
other was in consciousness." (H.) The most characteristic feature here
was, as Br. also indicates, "Great amount of rivalry at the beginning
of the series"; while Bs. states, "There was a lot of confusion and
a feeling of groping for words." Br. adds later, "For some seconds
association seemed obstructed. Then by an effort the process was
started which followed an involuntary course. A kind of confused
presence of both words." Another subject adds, "There was a long blank
after the words were said in which both words were balancing off in
the fringe of consciousness and the mind expectant, passively waiting
for an association to turn up. The hesitant period seemed marked by an
attempt at a synthesis of these two words in some way."


SUBJECT XIX

          Lamp--Justice.

    | chimney       lamp.
    | white         lamp chimney.
    | yellow        white chimney.
    | flame         yellow.
    | nickel        lamp plus the other words.
    - scales        justice.
    - purple robe   justice.

As the tables indicate, there was frequently a strong tendency here for
the abstract terms to fuse. As H. noticed, "It does not tend to call up
associations of its own stripe, but in some way becomes concrete."


TABLE VI. TWO WORDS: CONCRETE--ABSTRACT SPOKEN

    Time--15 seconds.
    Characters--same as Table I.

          (I) Desert--Hate

    A.  |  |  |  +  +  +
    M.  |  |  |  |  -  -
    F.  +  +  +
    H.  |  -  |  |  |  |  |  |  |
    Ro. |  |  -  -  |  |  |  |
    Bl. +  +  +  +  +  +  +
    By. |  |  |  +  +  +  +  +  +
    Bs. -  |  |  |  |  |
    Br. |  |  |  +  +

    (1) Desert--Hate.
    (2) Lamp--Justice.
    (3) Pen--Fatigue,
    (4) Gate--Fear.
                          TABLE VI--_continued_

              |       -      +        α       β
    A.     { 10       2      6        0       0
           { 55.5%   11.2%  33.3%

    M.     {  7       8     10        0       0
           { 28.0%   32.0%  40.0%

    F.     {  3       8      5        0       0
           { 18.7%   50.1%  31.2%

    H.     { 17      16      1        0       0
           { 50.0%   47.1%   2.9%

    Ro.    { 15      11      0        0       0
           { 57.7%   42.3%

    Bl.    {  4      15      7        2       0
           { 14.2%   53.5%  25.1%     7.2%

    By.    {  4      21      8        0       1
           { 11.8%   61.8%  23.5%             2.9%

    Bs.    {  5       1     15        0       2
           { 21.8%    4.3%  65.2%             8.7%

    Br.    { 11       8      2        0       0
           { 52.3%   38.1%   9.6%

    Totals { 76      90     54        2       3
           { 33.8%   40.0%  24.0%      .8%    1.4%

  Number of subjects, 9; number of sections, 4; number of experiments,
  34; number of associations, 225.

Reversing the order by placing the concrete noun second, it gains in
influence. We are told by the subjects at this point, "The choice seems
to be determined by the concreteness of the word." (H.) "The abstract
soon exhausted itself as a word-suggester." (Ro.) There was fusion in
24% of the associations of the first group, and 22.2% in the second.
There are six cases where the second alone prevails, two for the first
starting-point and four for the second. There are eleven cases of
fusion beginning before the sixth word and continuing throughout the
series. There is much partial fusion, with the second predominating in
influence.

The results again emphasize the fact that the influence is
transferable, also that normally the second has the advantage;
furthermore, they illustrate the preponderance of the concrete word as
a starter of associations, and that the abstract term when it exerts
an influence tends to fuse rather than persist in having separate
associations; all of which shows that concrete terms produce more vivid
impressions than abstract ones, and would, when it is possible to use
them, be of direct aid to the learner.


TABLE VII. TWO WORDS: ABSTRACT--CONCRETE SPOKEN

    Time--15 seconds.
    Characters--same as Table I.

         (1) Honesty--Tide

    A.   -  |  -  -  -  -
    M.   -  -  -  -  |  |  |  |
    F.   |  -  -  -  -  -
    H.   |  -  -  -
    Ro.  |  |  -  -  -  -
    Bl.  |  -  -  -  -  -  -
    By.  -  -  +  +
    Bs.  -  -  -  -  -  -  -
    J.   -  +  +  +  +  +  +

    (1) Honesty--Tide.
    (2) Skill--Coal.
    (3) Terror--Sky.
    (4) Refined--Flag.

               |       -        +       α        β
    A.     { 2        7        11       0        4
           { 8.4%    29.2%     45.8%            16.6%

    M.     { 4       11        10       1        5
           {12.9%    35.5%     32.3%    3.2%    16.1%

    F.     { 2        8         9       0        0
           {10.5%    42.2%     47.3%

    H.     { 8       11         1       3        0
           {39.2%    43.5%      4.1%   13.2%

    Ro.    {10       13         3       0        0
           {38.4%    50.0%     11.6%

    Bl.    {15       10         0       0        0
           {60.0%    40.0%

    By.    { 2       20         3       0        0
           { 8.0%    80.0%     12.0%

    Bs.    { 1       11         6       0        2
           { 5.0%    55.0%     30.0%            10.0%

    J.     { 6       16         6       0        0
           {21.4%    57.2%     21.4%

    Totals {50      107        49       4       11
           {22.6%    48.4%     22.2%    1.9%     4.9%

  Number of subjects, 9; number of sections, 4; number of experiments,
  36; number of associations, 221.

Increasing the disparateness by making the one a proper name and the
other a pure abstract noun, we find the name dominates consciousness,
almost to the exclusion of the abstract term. The tables confirm the
conclusions that the abstract term, even when given the advantage of
position, exerts little influence, for in the first group of eighteen
experiments of two hundred and eighty-four associations there are 13.4
times as many associations for the first as for the second, or two
hundred and fifteen words for the first (75.8%) and sixteen (5.5%) for
the second.

Reversing the order, the burden of influence swings back to
thirty-eight (12.8%) for the first and one hundred and fifty-six
(52.7%) for the second with an amount of fusion increased to
ninety-nine (33.4%).


TABLE VIII. TWO WORDS: PROPER NOUN--ABSTRACT NOUN SPOKEN

    Time--50 seconds.
    Characters--same as Table I.

                         (1) Lowell--Liberty

  M.   |  |  |  |  α  α  α  |  |  |  |  |  |
  R.   |  α  α  |  |  |  |  |  |  |  |  |  |  |  |  |
  K.   |  |  |  |  |  |  |  |  |  |  |  |  |  |
  J.   |  |  |  +  +  +  α  α  α  α  α  α  α  α  α  α  α  α  α
  S.   |  |  |  |  |  -  |  |  |  |  |  |  |
  Ht.  -  -  |  |  |  |  |  |  |  |  |  |  |  |  |  |

    (1) Lowell--Liberty.
    (2) Roosevelt--Fidelity.
    (3) Eliot--Integrity.

            |        -        +         α        β

    M.   { 29        3        2         7        0
         { 70.7%     7.4%     4.8%     17.1%

    R.   { 30        0       17         2        0
         { 61.2%             34.7%      4.1%

    K.   { 43        2        0         0        0
         { 95.5%     4.5%

    J.   { 42        0        8        13        0
         { 66.6%             12.7%     20.7%

    S.   { 40        2        0         0        0
         { 95.2%     4.8%

    Ht.       { 31         9          3          1          0
              { 70.4%     20.4%       6.9%       2.3%

    Totals    {215        16         30         23          0
              { 75.8%      5.5%      10.7%       8.0%

  Number of subjects, 6; number of sections, 3; number of experiments,
  18; number of associations, 284.


TABLE IX. TWO WORDS: ABSTRACT NOUN--PROPER NOUN SPOKEN

    Time--50 seconds.
    Characters--same as Table I.

                 (1) Individuality--Lincoln

    M.   |  +  |  +  +  +  +  +  +  +  +  +  +  +  +
    R.   -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -
    K.   -  -  -  -  -  -  -  -  -  -  -  -
    J.   |  -  -  -  -  -  -  -  -  -  -  -  |  |  |  |  |  |  |  |  |  |
    S.   -  -  -  -  -  -  -  -  -  -  -  -  -  -  -
    Ht.  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -

    (1) Individuality--Lincoln.
    (2) Brevity--Webster.
    (3) Justice--Hanus.

             |        -         +        α        β

    M.     { 4       19        25        0        0
           { 8.3%    39.6%     52.1%

    R.     { 5       31        12        0        0
           {10.4%    64.5%     25.1%

    K.     {14       31         0        0        0
           {31.2%    68.8%

    J.     {13       11        44        0        0
           {19.1%    16.1%     64.8%

    S.     { 2       40         0        0        0
           { 4.8%    95.2%

    Ht.    { 0       24        18        0        3
           {         53.4%     40.0%              6.6%

    Totals {38      156        99        0        3
           {12.8%    52.7%     33.4%              1.1%

  Number of subjects, 6; number of sections, 3; number of experiments,
  18; number of associations, 296.


GROUP VII

Occasional experiments of the previous groups indicated an abnormal
influence of those ideas which were accentuated by the special
interests of the subjects. We ask therefore: What is the relation of
the newly aroused associations to the present content of consciousness?
May the present content of consciousness be so varied as to reënforce
or inhibit the characteristic influence of the individual impressions
successively presented? In order to test this, words or a sentence were
pronounced just previous to the presentation of the two starting-points
related to them as follows; in the first divisions the preparatory
words lead to the second, in the second divisions to the first, in the
third divisions to neither, in the fourth divisions to both, and in the
last we have a slight change in the marginal setting and a sentence
leads to the second.

The following are examples of the first and second divisions:

    Quartz--Granite--Shale.
    HORN--SLATE.
    Dog--Sheep--Horse.
    SQUIRREL--TELEGRAM.


SUBJECT XVIII

    --slate                slate
    --slate ledge (Col.)   slate
    --Great Falls, Mont.   quartz, granite, shale
    --geology              quartz, granite, shale, G. F.
    --Will B.              geology
    --State University     Will B.
    --Sexton James         State University
    --Mrs. J.              Sexton James
    --C. S. J.             Mrs. J.
    --Indian School        C. S. J.
    --Fort Shaw School     Indian School
    --Mrs. C.              Ft. Shaw School
    --Mrs. E.              Mrs. C.
    --Seattle, Wash        Mrs. F.
    --Miss M.              Seattle, Wash.
    --Everyman             Miss M.
    --Cousin V.            Everyman
    --theatre              Cousin V.


SUBJECT XII

    | hunt             squirrel (and dog)
    | pasture          sheep, horse, and hunt
    | yard             squirrel and pasture
    | Harvard          yard
    | Freshman         Harvard
    | themes           Freshman and squirrel
    | hunting          themes, dog,     "
    | George           hunting
    | squirrel         George
    | creatures        squirrel
    | animals          creatures
    | activity            "      animals
    | agility             "         "
    | grace               "         "
    | cuteness            "         "
    | pets                "         "
    | civilisation     creatures
    | vs. cats            "
    | enemies          cats

That one starting-point establishes itself more firmly, and offers
more dominant associations with an increased degree of suggestiveness
is intimated by various expressions of the subjects who state, "the
preparatory words called forth associations connected with this or that
starting-point"; or "there was a felt fusion of all"; "a summation of
all"; or "the preparatory words had an influence throughout"; "they
strengthened this or that starting-point"; "they affected one and not
the other"; and many similar expressions.

The influence of these preparatory marginal settings was also indicated
very often by the nature or kind of words in the series.

Turning to the tables which collectively represent in a graphic and
quantitative form the notes of the subject, and which are in harmony
with the above, we find they demonstrate that there is a very intimate
and definite relationship. When the state of mind immediately preceding
the moment of the formation of the associated series is conditioned by
the preparatory words leading to the second, the amount to which that
starting-point dominates consciousness in arousing associations is
greater than in any previous case where the words of like nature are
pronounced.

With the preparatory words leading to the second starting-point, we
have 11.5% of the influence for the first, 69.2% for the second; with
them leading to the first, 58.2% for first, 17.2% for second; with
them leading to neither, 40.6% for first, 44.2% for second; with them
leading to both, 16.1% for first, 54.8% for second; with a preparatory
sentence leading to second, 12.9% for first, 59.7% for second.

A few experiments were made where one word was given as a setting and
four or eight starting-points were shown, but the starting-points were
so numerous that they tended to confuse the subject's introspective
account.


TABLE X. TWO WORDS SHOWN: THREE PREPARATORY WORDS SPOKEN LEADING TO THE
SECOND

    Time--50 seconds.
    Characters--same as Table I.

    Corn, Wheat, Oats.        (1) MOB--HAY

    M.  + + + - + + - - - -
    R.  - - - | + + α + β - - - - -
    K.  - - - - - - - - - - - - - -
    S.  + + + + - - - - - - - - - - - -
    J.  + + + + + + β β - - - - - - -
    Ht. - + + + + β β β β β β β β

    (1) Corn, Wheat, Oats.  MOB--HAY.
    (2) Cloud, Mist, Dew.  ORATION--CANOE.
    (3) Turtle, Fish, Frog.  KEY--NET.
    (4) Quartz, Granite, Shale.  HORN--SLATE.
    (5) Geometry, Plane, Rectangle.  CASE--CUBE.

               |          -          +          α          β

            { 10         41         17          0          7
    M.      { 13.3%      54.7%      22.7%                  9.3%

            { 15         37         21          1          4
    R.      { 19.3%      47.4%      26.9%       1.2%       5.2%

            {  0         55          0          0          0
    K.      {           100.0%

            {  8         64          4          0          0
    S.      { 10.5%      84.2%       5.3%

            { 17         69          6          0          2
    J.      { 18.1%      73.4%       6.3%                  2.2%

            {  2         45         17          0          8
    Ht.     {  2.8%      62.5%      23.6%                 11.1%

            { 52        311         65          1         21
    Totals  { 11.5%      69.2%      14.4%        .2%       4.7%

  Number of subjects, 6; number of sections, 5; number of experiments,
  29; number of associations, 449.


TABLE XI. TWO WORDS SHOWN. THREE PREPARATORY WORDS SPOKEN LEADING TO
THE FIRST

    Time--50 seconds.
    Characters--same as Table I.

    Dog, Sheep, Horse. (1) SQUIRREL--TELEGRAM

    M.   | | | | | | | | | | | | | |
    R.   | | | | | | | | | | | | | | |
    K.   | | | | | | | | | | | | |
    S.   | | - - - - - - - - - -
    J.   | | | | | | | | | | | | | | | | | | |
    Ht.  | | | | | | | | | | - - - - - - - -

    (1) Dog, Sheep, Horse.  SQUIRREL--TELEGRAM.
    (2) Pineapple, Banana, Orange.  FRUIT-STAND--ELECTRIC LIGHT.
    (3) Justice, Truth, Beauty.  CHARITY--PHOTOGRAPH.
    (4) Gold, Copper, Silver.  DIME--SKULL.

    M.      {28         3         15        0        8
            {51.8%      5.5%      27.9%             14.8%

    R.      {16         9         23        0        11
            {27.2%     15.3%      38.9%              18.6%

    K.      {56         2          0        0         0
            {96.5%      3.5%

    S.      {36        24          0        0         0
            {60.0%     40.0%

    J.      {38         4         24        1         7
            {51.4%      5.4%      32.4%     1.4%      9.4%

    Ht.     {37        20          0        0         0
            {64.9%     35.1%

    Totals  {211       62         62        1         26
            {58.2%     17.2%      17.2%      .2%       7.2%

  Number of subjects, 6; number of sections, 4; number of experiments,
  24; number of associations, 362.


TABLE XII. TWO WORDS SHOWN. THREE PREPARATORY WORDS SPOKEN LEADING TO
NEITHER

    Time--50 seconds.
    Characters--same as Table I.

    Botany, Statue, Postal.  (1) CHOCOLATE--BANNER

    M.  | | - - - - - - - - - - - - - |
    R.  - | | | | | | | | | | | | | | | | | | | |
    K.  | - | | | - | - | - | - |
    S.  | | - - - - - - - - - - - - - - - -
    J.  | | | | | | | - - - - - - - - - - - - - - -
    Ht. - - + + + - - - - - - - - - - -

    (1) Botany, Statue, Postal. CHOCOLATE--BANNER.
    (2) Idea, Proposition, Syllogism. SILVER--GAME.

              |         -          +        α        β

    M.      {3         28          0        0        0
            {9.6%      90.4%

    M.      {22         1          2        0        14
            {56.5%      2.5%       5.2%              35.8%

    K.      {28         5          0        0         0
            {84.8%     15.2%

    S.      {13        20          0        0         0
            {89.4%     60.6%

    J.      { 17         22          0          0          0
            { 43.6%      56.4%

    Ht.     {  0         14          8          7          0
            {            48.3%      27.6%      24.1%

    Totals  { 83         90         10          7         14
            { 40.6%      44.2%       4.9%       3.4%       6.9%

  Number of subjects, 6; number of sections, 2; number of experiments,
  12; number of associations, 204.


TABLE XIII. TWO WORDS SHOWN. THREE PREPARATORY WORDS SPOKEN LEADING TO
BOTH

    Time--50 seconds.
    Characters--same as Table I.

    Tree, Shrub, Grass.     (1) TEA--FERN

    M.      + + α α α α α α α - -
    R.      + + + + + + + + + + + +
    K.      - - - - - - - - - - - - - - - -
    S.      - - - - - - - - - - - -
    J.      | - - - - - - - - - - - - - - - - - - - -
    Ht.     + + α α α α α α | | | | |

    (1) Tree, Shrub, Grass. TEA--FERN.
    (2) Fox, Wolf, Moose. DEER--DOVE.

               |          -          +          α          β

    M.      {  2         12          4          8          0
            {  7.6%      46.3%      15.3%      30.8%

    R.      {  0          0         25          0          0
            {                      100.0%

    K.      {  1         30          0          0          0
            {  3.2%      96.8%

    S.      {  8         12          0          4          0
            { 33.3%      50.0%                 16.7%

    J.      {  8         26          0          0          0
            { 23.5%      76.5%

    Ht.     {  8         12          2          6          0
            { 28.5%      42.8%       7.2%      21.5%

    Totals  { 27         92         31         18          0
            { 16.1%      54.8%      18.4%      10.7%

  Number of subjects, 6; number of sections, 2; number of experiments,
  12; number of associations, 168.


TABLE XIV. TWO WORDS SHOWN (AUTHORS)--WITH A PREPARATORY SENTENCE
LEADING TO THE SECOND

    Time--50 seconds.
    Characters--same as Table I.

    "God's in His Heaven--all's right with the world."

          (1) BURNS--BROWNING

    M.    + - - - - - - - - - | | | | | |
    R.    | - - - - - - - - - - -
    K.    + + + + + + + + + + + + +
    S.    | - - - - - - - - - - - - - | | |
    J.    | | - - - - - - - - - - - - - - - - - -
    Ht.   - - - - - - - - - -

    (1) "God's in His Heaven--all's right with the world."
    BURNS--BROWNING.

    (2) "All that glistens is not gold."
    BYRON--SHAKSPERE.

    (3) "Hitch your wagon to a star."
    SPENCER--EMERSON.

             |       -        +          α       β

    M.     { 9       15       15         0       4
           {20.9%    34.8%    34.8%              9.5%

    R.     { 1       47        0         0       0
           { 2.1%    97.9%

    K.     {15        0       26         0       0
           {36.5%             63.5%

    S.     { 6       34        0         0       0
           {15.0%    85.0%

    J.     { 3       43       14         0       0
           { 5.1%    71.6%    23.3%

    Ht.    { 0       18       13         0       0
           {         58.1%    41.9%

    Totals {34      157       68         0       4
           {12.9%    59.7%    25.8%              1.6%

  Number of subjects, 6; number of sections, 3; number of experiments,
  18; number of associations, 263.


GROUP VIII

The aim here is to test the effect of interruption in the series of
associations, and to throw further light on the relation of the series
to the present content of consciousness when this content is a series
of associations and the new content is a pronounced word which is to
act as a point of departure for new associations.

There are three divisions. The time in all is fifty seconds; in the
first a word of general connotation is given and after an interim of
ten seconds a second more specific word is pronounced. In the second
three similar words are given at an interval of fifteen seconds; the
third four words with an interval of ten seconds. The following are
examples of the first and third divisions:


SUBJECT III

        Commencement--Sieve

        commencement
    | college       commencement
    | cap           college
    | gown          cap
    | boys          commencement
    | confetti      commencement

        sieve
    - holes         sieve
    - water         sieve
    - flour         sieve
    - space         holes
    - concept       space
    - Royce         concept
    - time          concept
    - eternity      time
    - damnation     eternity
    - Hamlet        "Consummation," etc
    - Shakspere     Hamlet


SUBJECT VII

          Wax--Jug--Tar--Sod

        wax
    | Charley              wax
    | picnic               Charley
    | horse                Charley
    | saddle               horse

        jug
    - ink                  jug
    - clay                 jug
    - Hegel                jug

        tar
    / 'old tar'            tar
    / ship                 'old tar'
    / Bermuda              'old tar'

        sod
    \ grave                sod
    \ graveyard            grave
    \ house                graveyard
    \ church               house
    \ music                church
    \ white                church

In all the experiments the subject simply knew that possibly more than
one starting-point would be given. There was of course the conscious
recognition on the part of the subjects that the pronounced words were
starting-points, which would imply an attentive consciousness, but they
were cautioned neither to favor nor inhibit the newly pronounced word
nor an association in progress.

The notes are uniform in showing that often one, two or three words
of the former association-series are written after the new word is
pronounced. "The momentum," says F., "was great enough to carry the
associations two or three words beyond the pronounced word"; while Bl.
found "a tendency for the trend of associations to persist, though
not strong enough to overcome the new influence." By.'s experience
was slightly different. As stated before, he often wrote the word as
it came into consciousness. "On hearing a new word it gets precedence
over the next associations not yet formed, and there is considerable
confusion and lost time unless the motor discharge of writing the
pronounced word is permitted to have free expression." The tables
verify the same, and also show that there are more associations during
the first interval.

Does a former starting-point regain its influence? In the first
division there are two cases where the first and second fuse, but no
place where the first independently forms an association; in the second
but one word for subject Bl. in "Quill--Bench--Chalk," and in the third
not any. There was a small amount of fusion in all, since but two words
are due to the combined influence of the first and third, five to the
combined influence of the first and fourth, with three starting-points,
and one to the combined influence of the first and third with four
starting-points.

The train of associations is inhibited by a new starting-point which
dominates in influence. No mention is made in any note that a former
starting-point remains in consciousness for the series, but M.
emphatically writes, "Absolutely no influence of the preceding word
or words when the next is taken up"; and later, "As soon as the new
one is pronounced the old word and the series it had brought up were
immediately suppressed." Bl. comments, "How remarkable it is that each
new word crowds the old trend of associations out and starts new ones";
and the graphic representation, one of which only is given here on
account of lack of space, shows that there is no return to the original
series.

The tables are indicative of the tendency. In the first division of
the group there are three possible lines of fusion, in the second six
possible lines, and in the third twelve possible lines, but we find
only 13.2% for all forms of fusion in the first, 7.9% for the second,
and 10.5% for the third. In the eighty-seven experiments of the series
there is but one absolute return to the previous starting-point. (See
Group IX, sec. 2, Bl.) The tables show the varying degrees of fusion,
and while the percentages have little meaning, as there is a variable
time-element, the numbers do show accurately the number of words and
the relative and continued influence of each starting-point.

We conclude that, when the present content of consciousness is a
series of associations, the newly given impression establishes itself
sufficiently to inhibit the associations of the previous series.


TABLE XV. TWO WORDS--GENERAL AND PARTICULAR--SPOKEN

    Time--50 seconds, with an interval of 10 seconds between 1 and 2.
    Characters--same as Table I.

                (1) Commencement--Sieve

    M. | | | - - - - - - - - - - - - - -
    F. | | | | | | + + + + + + +
    H. | | | | | - - - - - - - - - - - -
    V. | | | | | - - - - - - - - - - - -
    S. | | | | | | | | - - - -
    E. | | | - - - - - - - - -
    Bs.| | | - - - - - - - - - - - - - -
    Ht.| |   - + + + + + +
    By.

              |      -       +       α      β
    M.     { 3      14       0       0      0
           {17.6%   82.4%

    F.     { 6               7
           {46.2%    0      53.8%    0      0

    H.     { 5      12       0       0      0
           {29.4%   70.6%

    V.     { 5      12       0       0      0
           {29.4%   70.6%

    S.     { 8       4       0       0      0
           {66.6%   33.4%

    E.     { 3       9       0       0      0
           {25.0%   75.0%

    Bs.    { 3      14       0       0      0
           {17.6%   82.4%

    Ht.    { 2       1       6       0      0
           {22.2%   11.2%   66.6%

    By.      0       0       0       0      0


    Totals {35      66      13       0      0
           {31.4%   55.4%   13.2%

  Number of subjects, 9; number of groups, 1; number of experiments, 8;
  number of associations, 114.


TABLE XVI. THREE WORDS SPOKEN

  Time--50 seconds. Interval, 15 seconds.

  Characters: | first, - second, / third, + partial fusion between
  first and second, α partial fusion between first and second with
  first predominating, β partial fusion between first and second with
  second predominating, γ partial fusion between first and third, δ
  partial fusion between second and third, ε partial fusion between
  first and third with third predominating.

                            (1) Gun--Bug--Jaw

    M.  | | | | | | | | - - - - - - / / / / / /
    F.  | | | | | - - - δ δ δ
    H.  | | | | | | - - - - / / / / /
    V.  | | | | | | | | - - - - - - / / / / / /
    Bl. | | | | | | - - - - - / / / / / / /
    By. | | | | | | | | | - - - - - - / / / / / /
    Bs. | | | | | + + + + / / / / /
    Ht. | | | | | | | α | | | δ - - - -
    Ro. | | | | | | - - - - - - / / / / / /

                         (1) Gun--Bug--Jaw
                         (2) Quill--Bench--Chalk
                         (3) Hall--Moss--Leather

            |      -      /      +     α      β       γ      δ      ε

    M.     {20     14     18     1     0      0       0      0      0
           {37.8%  26.4%  33.9%  1.9%

    F.     {15     11     15     0     0      0       2      3      0
           {32.6%  23.9%  32.6%                       4.4%   6.5%

    H.     {19     11     15     0     0      0       0      0      0
           {42.3%  24.4%  33.3%

    V.     {24     15     21     1     0      4       0      0      0
           {36.9%  23.1%  32.3%  1.5%         6.2%

    Bl.    {15      9      9     0     0      0       0      0      0
           {45.4%  27.3%  27.3%

    By.    {23      8     14     3     0      4       0      3      0
           {41.8%  14.5%  25.5%  5.5%         7.2%           5.5%

    Bs.    {13     10     10     4     0      0       0      0      5
           {30.9%  23.8%  23.8%  9.6%                              11.9%

    Ht.    {23     11     10     1     1      2       0      1      0
           {46.9%  22.5%  20.5%  2.0%  2.0%   4.1%           2.0%

    Ro.    {15     15     24     0     0      0       0      0      0
           {27.8%  27.8%  44.4%

    Totals {167    104    136    10    1      10      2      7      5
           { 37.8%  23.5%  30.8%  2.3%  .2%    2.3%    .4%   1.5%   1.2%

  Number of subjects, 9; number of sections, 3; number of experiments,
  27; number of associations, 442.


TABLE XVII. FOUR WORDS SPOKEN

  Time--50 seconds. Interval, 10 seconds.

  Characters -- same as Table XVI, with addition of \ representing the
  fourth, ς partial fusion between first and fourth, η partial fusion
  between second and fourth, θ partial fusion between third and fourth,
  ☐ total fusion between first, second, third, and fourth, ι partial
  fusion between first, third, and fourth, κ partial fusion between
  second, third, and fourth, λ partial fusion between third and fourth,
  with fourth predominating, [ partial fusion between first, second,
  and third.

                (1) Den--Nag--Cot--Fan

    F.  | | | | | | - - - - - - δ - - \ \ \ \ \
    H.  | | | | | - - - / / / / / / \ \ \ \ \ \ \ \ \
    V.  | | | | |  - - - / / / / \ \ \ \ \ \ \
    Bl. | | | | - - - - / / / \ \ \ \
    By. | | | | | | - - - - - δ δ δ \ \ \ \ \
    Bs. | | | - - / / / \ \ \ \ \
    Ht. | | | | | - - - - - / / / / \ \ \ \
    Ro. | | | | - - - - / / / / \ \ \ \ \

    (1) Den--Nag--Cot--Fan
    (2) Tax--Fan--Map--Dog
    (3) Paw--Wand--Box--Mug
    (4) Bud--Car--Cub--Mat
    (5) Wax--Jug--Tar--Sod
    (6) Cur--Elk--Pug--Man
    (7) Rope--Wig--Ink--Grass

    (_Table on page 460_)


GROUP IX

The aim here was to see if it were possible to have the first
starting-point such that the conditions would be similar to the results
obtained by using preparatory marginal settings, but include rather
than inhibit the second starting-point. A review of the tendency toward
mental combination in the former experiments suggested that the words
be in the relation of the whole and part. The "part" was given the
position of predominating influence in order to see to just what extent
it would persist in combining.

              |     --     /     \      +      γ    δ    ς    η    θ

      F.  { 29     19     11     19     3      0    4    0    0    11
          { 30.3%  19.8%  11.4%  19.8%  3.1%        4.2%           11.4%

      H.  { 31     18     24     25     3      0    0    0    0     8
          { 28.5%  16.5%  22.1%  22.9%  2.7%                       73%

      V.  { 28     28     21     50     0      0    0    0    0     0
          { 22.0%  22.0%  16.6%  39.4%

     Bl.  { 18     17     17     29     0      0    0    0    0     0
          { 21.6%  20.5%  20.5%  34.9%

     By.  { 36     22      9     24     2     4     3    0    6     2
          { 32.1%  19.6%   8.0%  21.4%  1.8%  3.7%  2.6%      5.3%  1.8%

     Bs.  { 22     13     17     18     4     1     0    1    0     6
          { 24.2%  14.3%  18.7%  19.8%  4.4%  1.1%       1.1%       6.6%

     Ht.  { 30     18     10     21     3     0     0    0    2     1
          { 34.4%  20.6%  11.4%  24.2%  3.4%                  2.4%  1.2%

     Ro.  { 24     22     16     48     0     0     3    0    0     0
          { 21.3%  19.4%  14.2%  42.4%             2.7%

    Totals{218    157    125    234    15     5    10    1    8    28
          { 26.6%  19.2%  15.2%  28.5%  1.9%   .6%  1.3%  .2%  .9%  3.4%

             ☐    ι     β      κ     λ     [

      F.     0     0     0      0     0     0

      H.     0     0     0      0     0     0

      V.     0     0     0      0     0     0

     Bl.     0     0     0      2     0     0
                                2.5%

     By.    4      0     0      0     0     0

     Bs.    0     3      1     0      5     0
                  3.3%   1.1%  0      5.4%  0

     Ht.    0     0      0     0      0     2
                                            2.4%

     Ro.    0     0      0     0      0     0

    Totals{ 4     3      1     2      5     2
          { .4%   .3%   .2%   .3%   .7%  .3%

  Number of subjects, 8; number of sections, 7; number of experiments,
  52; number of associations, 819.

In the light of the interpretation of previous facts and the subsequent
results of this group of experiments, we are now in a position to
conclude that, if the present content of consciousness on the reception
of a new impression is such that the reactions are not antagonistic
but reënforce each other, the second will not persist in independent
influence, but will be rather included in and supplementary to the
influence of the first, which otherwise would be less assertive. The
results below show that while the first of the two starting-points has
a decided disadvantage of position, and therefore has little influence
in arousing associations, it here is responsible for 43.2% and the
second for but 9.5%, while there is a combined influence of 47.3%, the
first strongly predominating in partial fusion. There is but one case
where the second was followed exclusively.

The explanation of this, extended to an hypothesis, would rest in
the fact that each word has definite characteristic reactions and
that the fusion of two words or lines of thought means that the motor
accompaniments are such that they unite and reënforce each other, or
that the one includes the other. There are few antagonistic impulses.


TABLE XVIII

TWO WORDS--WHOLE AND PART--SPOKEN

    Time--15 seconds.
    Characters--same as Table I.

                         (1) Crowd--Man

    A.  |  +  +
    M.  +  +  +  +  +  +  +  +  +
    F.  +  +  +  +  +
    H.
    J.  |  |  |  |  |  |  +  +  +
    S.  -  -  -  -
    V.  +  +  +  α  α  α  α  α  α
    E.  |  +  +  +
    L.  |  |  |  |
    Bs. α  α  α  α  α  α
    Br. |  |  |  |  |  +  +

    (1) Crowd--Man
    (2) Ton--Pound
    (3) Moose--Horn
    (4) Engine--Whistle
    (5) Book--Page
    (6) Music--Octave

              |         -       +      α       β

    A.     {  4        3      24       0       0
           { 12.9%     9.6%   77.5%

    M.     { 13        0      27       2       0
           { 30.9%            64.2%    4.9%

    F.     {  7        3      22       1       0
           {  20.6%    8.8%   64.8%    5.8%

    H.     {  28       5       0       0       5
           {  73.6%   13.2%                   13.2%

    J.     {  21       1      17       6       0
           {  46.6%    2.3%   37.8%   13.3%

    S.     {  25      12       0       0       0
           {  67.5%   32.5%

    V.     {   4       2      16      10       0
           {  12.5%    6.3%   50.1%   31.1%

    E.     {  11       3       9       5       0
           {  39.2%   10.8%   32.2%   17.8%

    L.     {  14       6       4       0       0
           {  58.4%   25.0%   16.6%

    Bs.    {  15       2      11       6       0
           {  44.2%    5.8%   32.4%   17.6%

    Br.    {  25       0       9       9       0
           {  58.2%           20.9%   20.9%

    Totals { 167      37     139      39       5
           {  43.2%    9.5%   35.9%   10.0%    1.4%

  Number of subjects, 11; number of sections, 6; number of experiments,
  62; number of associations, 387.


GROUP X

In the last group the words were called whole and part, although
the reader no doubt observed that this was not always strictly the
case, but rather that the more complex was more influential as a
word-suggester than the simple and often tended to include it. Can
this be demonstrated in another way? Attention was called in Group of
Experiments IV (two pictures shown) that picture no. 3 (a man sawing
wood), although first, had more than the normal number of associations,
the explanation of which might lie in the fact that it contained more
objects. We here attempt to test this by comparing a comparatively
complex and a comparatively simple picture as starting-points. It has
been clearly proven that the second picture to be presented has a
decided advantage of position. The simple picture is here given the
advantage of position. The notes show the first to be more important,
while there is no suggestion of the first including the second, on
account of there being less fusion than for words and more independent
associations for the second starting-point; the averages are 38.5% for
the first, 30.5% for the second, 31.0% fusion.

    (a) Picture of Blacksmith shoeing horse
    (b) Picture of a Sheep.
     | shop                   (a)
     | nails                  shop
     | shoe                   shop and (a)
     | blacksmith             shop and shoe and (a)
     | Longfellow             blacksmith
     - pasture                (b)
     - sheep                  pasture
     - lambs                  pasture and sheep
     - grass                  pasture and lambs
     - hillsides              grass
     - brook                  pasture and hillsides
     | iron                   shop-shoe
     | hammering              iron
     | soldier                iron-hammering and first group
     | battle                 soldier
     | shoe                   battle
     | horse-shoe nail        iron, hammering, soldier, battle
     | anvil                  horse-shoe nail


TABLE XIX. TWO PICTURES SHOWN

    Time--50 seconds.
    Characters--same as Table I.

                   (1) (a) Blacksmith shoeing Horse
                       (b) A Sheep

    J.    |  |  |  |  |  -  -  -  -  -  -  |  |  |  |  |  |  |
    E.    |  +  +  -  -  |  |  |
    W.    |  -  |  -  -  -  |  |  |  |  |  |  |
    V.    |  |  +  +  +  +  +  +  +  +  +  +  +  +  +
    L.    -  |  |  |  |  |  |  |  |
    Ht.   -  -  α  α  |  |  |  |  -  +  +  +  +  +


    (1) (a) Blacksmith shoeing Horse
        (b) A Sheep

    (2) (a) Girl and Boy
        (b) A Bird

    (3) (a) Three Children
        (b) A Duck

    (4) (a) A Sower
        (b) A Dog

              |      -      +      α         β

    J.     { 30     23      9     18         0
           { 37.5%  28.8%  11.2%  22.5%

    E.     { 28     10      2      0         0
           { 70.0%  25.0%   5.0%

    W.     { 14     18      5      0         0
           { 37.8%  48.7%  13.5%

    V.     {  6     11     37      0        13
           {  8.9%  16.4%  55.3%            19.4%

    L.     { 27      4      1      0         0
           { 84.3%  12.5%   3.2%

    Ht.    { 16     30      6      6         0
           { 27.6%  51.8%  10.3%  10.3%

    Totals {121     96     60     24        13
           { 38.5%  30.5%  19.2%   7.6%      4.2%

  Number of subjects, 6; number of sections, 4; number of experiments,
  24; number of associations, 314.


GROUP XI

Three words spoken, one immediately following the other, are given
in this group in order to test the span of consciousness and the
influence of an immediate interruption, as one was given immediately
following the other in such a way that all associations were checked
until after the third starting-point. If we attempt to follow the
initial starting-points we see they may disappear after the first few
associations and reappear in the series; or each remain as initial
starting-points for a few associations; or one alone control the whole
series, while the others are present without influence; or there may be
an alternation of independent influences; or an influence which shows
a modifying effect of one or both of the other starting-points, which
may reach such a degree that all fuse, and in so doing get an advantage
over the single words. The starting-points are to a large extent
disparate and there is very little fusion; we find only .5% total
fusion and 18.7% partial fusion.

An example would be as follows:


SUBJECT VIII

FROG--ICE--TABLE

    - snow         ice
    | bench        table
    | fish         frog
    | pole         fish
    | boat         fish
    | line         fish
    | bait         fish
    | weeds        fish
    - skating      ice
    - sleds        ice
    - coasting     sleds
    - girls        coasting

There were no cases of the total fusion of all the starting-points
throughout the series, no cases of total fusion of the first and
second, and two cases for the second and third. There were two cases of
partial fusion of the first, second, and third throughout. The first
was followed exclusively in two cases, the third in one. The first and
third were followed intermittently in four cases, the second and third
in eleven, the first, second, and third in thirty-five.

The second starting-point has the ascendency of influence. This is
not due to habits formed in earlier experiments, as the two groups
of experiments with two words were given in connection with other
experiments. This group informs us that in immediate interruption the
new impression has not sufficient power to establish itself more firmly
than a past impression which is just past, and must in a manner be
reproduced to start associations. The influence of the first has been
somewhat destroyed, but the second is greater than the third.


TABLE XX. THREE WORDS SPOKEN

  Time--50 seconds.

  Characters--same as Table XVI, with the addition of [ representing
  total fusion between the first, second, and third, μ partial fusion
  between first, second, and third with first predominating, ν partial
  fusion between first, second, and third with first and third
  predominating.

                                   (1) Frog--Ice--Table

    A.  / - - - - -
    M.  - | | γ γ γ γ γ γ γ γ γ
    F. - - - - - / / | | | | / / / - - -
    H.  | - - - - - | | | | / / / - - -
    J.  | - - / / / / - + + + + + + + +
    Bl. - / | | | | | | - - - -
    By. | / / / / / / - - - - - - - - - - - - - - - -
    Bs. - | | | - - - / /
    Ro. | | - - / / / / / / / / / / /
    Br. - / / / δ δ δ δ δ δ

    (1) Frog--Ice--Table
    (2) Key--Shoe--Knife
    (3) Desk--Park--Glove
    (4) College--Church--Cafe
    (5) Boston--Elevator--Lake
    (6) Skate--Book--Theatre
    (7) Bridge--Sleigh--Ticket
    (8) Gun--Lamp--Watch

           |      -      /     +    α     β    γ     δ    [     μ    ν
  A.    { 26     32     30     0    0     0    0     0    2    10    0
        { 26%    32%    30%                               2%   10%

  M.    { 25     36     21     3    0     0   10     1    1     0    6
        { 24.3%  34.9%  20.5%  2.9%            9.8%   .9%  .9%      5.8%

  F.    { 15     23     24     8    11    0    7     6    0     0    0
        { 15.9%  24.5%  25.5%  8.5% 11.8%      7.4%  6.4%

  H.    { 31     68     31     0     0    0    0     0    0     0    0
        { 23.8%  52.4%  23.8%

  J.    { 19     18     48    15     0    0    0    22    1     0    0
        { 15.5%  14.6%  39.1% 12.2%                 17.8%  .8%

  Bl.   { 23     33     16     5     0    0    0    22    0     0    0
        { 23.2%  33.2%  16.0%  5.0%                 22.6%

  By.   { 13     56     48    15     0    1   17     2    0     0    0
        {  8.7%  36.8%  31.5%  9.8%       .6% 11.2%  1.4%

  Bs.   { 19     13     19     9     0    0    0    19    2     0    0
        { 23.4%  16.0%  23.4% 11.2%                 23.5% 2.5%

  Ro.   { 32     33     33     0     0    0    0     0    0     0    0
        { 32.6%  33.7%  33.7%

  Br.   { 12     24     38     0     0    0    6     0    0     0    0
        { 15.0%  30.0%  47.5%                  7.5%

  Totals{215    336    308     55   11    1   40    72    6    10    6
        { 20.3%  31.1%  29.1%   5.2% 1.1% .9% 3.6%  6.8%  .5%   .9%  .5%

  Number of subjects, 10; number of sections, 8; number of experiments,
  79; number of associations, 1060.


GROUP XII

In order to test the relative influence and to throw more light on the
problem of immediate interruption four words were shown and the subject
was directed to read from left to right. There were few experiments
and most of the subjects were new. The results show that the third
starting-point has a decided disadvantage, having an influence of only
7.8%. The second and fourth are almost equal, while the first is again
less.


TABLE XXI. FOUR WORDS SHOWN

  Time--50 seconds.

  Characters--same as Table XVII, with the addition of ξ partial fusion
  between the second, third, and fourth with the second predominating.

                  (1) Gun--Car--Ink--Fan

    M.  \ \ \ - - - - - / / / / / /
    R.  \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
    K.  \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
    S.  - | / - - - - - - - -
    J.  | | | - - - - - - - - - - -
    Ht. + + α α α | | | | | | | | |

                      (1) Gun--Car--Ink--Fan
                      (2) Brain--Umbrella--Telephone--Chain
                      (3) Book--Money--Hour--Chart

          |      -      /      \      +     γ     θ     α      κ     ξ
     M. {14      6      9      4      0     6     0     0      0     0
        {35.9%  15.4%  23.0%  10.3%        15.4%

     R. {16      0      0     30      0     0     0     0      0     0
        {34.8%                65.2%

     K. { 0     35      0     16      0     0     0     0      0     0
        {       68.6%         31.4%

     S. { 5     15      7      4      0     0     0     0      0     0
        {16.2%  48.4%  22.5%  12.9%

     J. {13     13      4     15      2     0     0     0      0     0
        {27.7%  27.7%   8.5%  31.9%   4.2%

    Ht. {15      9      0      5      2     0     7     3      1     1
        {34.8%  20.9%         11.6%   4.7%       16.3%  6.9%  2.4%   2.4%

  Totals{63     78     20     74      4     6     7     3     1      1
        {24.5%  30.4%   7.8%  28.5%   1.6%  2.4%  2.6%  1.4%   .4%    .4%

  Number of subjects, 6; number of sections, 3; number of experiments,
  18; number of associations, 257.


GROUP XIII

Four similar words were pronounced in immediate succession, the results
of which show that we are correct in calling the above a case of
succession and also to establish clearly and definitely that there is
a difference of degree between immediate and postponed interruption.
The third starting-point has again a decided disadvantage of influence
and exerts no apparent influence in thirty-seven of the fifty-two
experiments. The last starting-point exerts the greatest influence.


TABLE XXII. FOUR WORDS SPOKEN

  Time--50 seconds.

  Characters--same as Table XVII with the addition of ο partial fusion
  between first, second, and third with first and second predominating.

    (1) Cow--Roof--Fence--Girl
    (2) Cathedral--River--Elevator--Newspaper
    (3) Cane--Harness--Box--Coat
    (4) Book--Snow--Rope--Stone
    (5) Wire--Flower--Horse--Paper
    (6) Gun--Wharf--Chair--Stove

       |    -    /    \    +    λ     δ    η    θ    ☐  ι   ο   ε   β
  M. { 35    3    6    32    0   3    17    0   9    0   0   0   0   0
     { 33.4% 2.8% 5.8% 30.4%     2.8% 16.2%     8.6%

  F. { 16   11    1    35    0   3     0    0   1    2   1   0   5   0
     { 21.3% 14.6% 1.4% 46.6%     4.0%          1.4% 2.7% 1.4%   6.6%

  H. { 21   38    11    30    0   0     0    0   0   0   0   1   0   1
     { 20.6% 37.3% 10.8% 29.5%                                   .9% .9%

  V. { 16   16    0    36    2   2     9    0   1    0   1   0   0   0
     { 19.3% 19.3%      43.3% 2.4% 2.4% 10.9%     1.2%     1.2%

  Bl.{ 36    8   10    19    0   0     4    0   0    0   0   0   0   0
     { 46.8% 10.3% 12.9% 24.8%        5.2%

  By.{ 23   30    3    36    1   0     0    0   1    0   0   0   0   0
     {24.5% 31.9% 3.3% 38.3% 1.0%               1.0%

  Bs.{ 22   14    1    20    1   0     0    7   0    1   0   0   0   0
     {33.3% 21.3% 1.6% 30.0% 1.6%          10.6%     1.6%

  Ht.{  9   19    4    31    0   0     0    0   1   14   0   0   0   0
     {11.5% 24.3% 5.2% 39.8%                   1.3% 17.9%

  Ro.{ 26   23    6    16    0   0     0    0   0    0   0   0   0   0
     {36.6% 32.4% 8.4% 22.6%

  Totals
    {204   162   42   255    4   8    30    7  13   17    2   1   5    1
    {27.2% 21.6% 5.4%  33.9% .5% 1.1% 3.9% .9% 1.8% 2.3% .3% .2% .7% .2%

  Number of subjects, 9; number of sections, 6; number of experiments,
  52; number of associations, 751.

We ask finally how far our results and notes point to a theoretical
understanding of the mechanism of associations. Previous work,
especially that of James, Cordes, Calkins, and Scripture, as well as
the accumulated notes of my subjects, confirm that the transition
may be made by means of total, partial, and focal recall, and that
in partial and focal recall the prominent persisting elements are
surrounded in the formation of a new idea by other new elements.

If a latent idea remains in the margin of consciousness and exerts
an influence, which not merely modifies but determines the series of
associations, and leads up to the focalisation of the latent idea,
we have a case of predetermined association, which, when noted by
investigators, has invariably become confused with mediate association.
Here there is an element or group of elements, persisting in the margin
of consciousness, which is gradually maturing and becoming focalised
into groups of elements comprising an idea which ultimately dominates
consciousness. In some cases three, four, and five ideas have been
named before this takes place, and we have here a reversed form of
association. Four subjects noted the experience on different occasions,
and it is not to be confused with the common experience of apprehending
the present contents of consciousness as part of a larger whole where
we are conscious of its existence but not of what it is.

The notes further show that the common conscious elements may be
predominantly visual, auditory, olfactory, gustatory, or kinæsthetic,
or a complex or compound of these in character, while to this may be
added an indication of the fact that the transition, incipient as it
is, may in many cases be reduced to a condition which is in the last
analysis one of the motor nervous system. Ht., for instance, finds
that the words all pass over into innervations of the organs of speech
and "are accompanied by the impulse to make the sound," stating later,
"they hang on the tongue." The following is one of the series given
which represents rather an extreme case, Taft, taffy, toffy; tough,
rough, ruff; buff, bluff, tough; muff, duff, tuff. Br., who also gave a
large percentage of verbal associations, finds that "some part of each
word seems to linger on the tongue with motor sensations till the next
comes." "I am subject," he adds, "more or less frequently to verbal
automatism of this auditory incipient motor type." Ro., who has many
auditory associations, reports "they are always accompanied by motor
images, together with many associations." A changing of orientation is
a common accompaniment, with statements of the feeling of the impulse
to turn in various directions. For F., who is predominantly of the
motor type, we have an example where the rhythmic ticking of a clock
fades into the rhythmic watching of a boat rising and falling on the
water.

The notes would seem to indicate that there is no idea without a
motor fringe, and also that these elements of incipient impulses to
movement may accompany the elements of transition, and are observed
introspectively by the subjects. They are therefore data for
psychology. Do they influence or direct the associations? In short, are
they the processes which connect and which determine the associations?

F. states, "There seem to have been waves of motor sensations. Such
waves may start with a word and carry one in faint mimicry through
the whole succession of bodily sensations that one experienced in
that event, and then may come a relapse until other stronger currents
appear." Here we are face to face with the dynamics of association,
the most fundamental and important problem of brain association. Have
these phenomena of ideational images "acquired by contact a kind of
magnetism which causes the one to attract the other and have, so to
speak, become magnetic?" (Zanotti.) Or are they on the other hand
independent of all force and "merely ideas of antecedence and sequence
only?" (Mill.) While there is no mention of a magnetic force, the notes
and results all show that the ideas are systematically conditioned in
a way which cannot be explained by the contiguity of the objects. The
motor elements play the deciding role. Ht. emphasises the influence
of ideated movement when he writes, "Kinæsthetic. Slow regular
tramping on snowshoes brought up the characteristic swing of arms,
and therewith the idea (sensations of weight) of the stick (or stock)
which I have generally carried on Norwegian snowshoes. Transition from
Vermont to the Black Forest by association with snowshoeing in both
places. Real sensations in play were free breath, movements in chest
(kinæsthetic), fresh air (olfactory), cold (thermal), and emotion
of emotional strength." Again, "Looking up at sun suggested general
ideas of expansion of attention and with this breath comes the idea,
breezes"; another subject adds, "A tendency to imitate the sounds of
syllables and this leads on to a train of associations"; another, "A
slight feeling of sudden changed impulse"; another, "A sort of motor
after-image came back and took the foreground"; and F. goes further
when he states, "Ideationally my hand wandered to the upper right-hand
corner of the page, then suddenly the auditory image of 47 came up as
if whispered to me." All of which indicate that some ideas at least
depend for their entrance into consciousness upon motor reactions.

Passing to the more refined reactions expressed in emotions we find
that they are not merely accompanying coloring influences, but also
often actual determining factors. All of the subjects notice at some
time a coloring atmosphere from an emotion, but others find that "the
growing word is rather felt emotionally than definitely formulated,"
and we have "a nameless idea, largely feeling-tone" (Ht.); or the words
may "all come as parts of a growing feeling, an indistinct though
strong state of mind." (J.) The same subject observed, "The previous
word may create a mood or feeling which in the main determines the
associations; a group of words is dependent upon strong accompanying
feeling--there is a summation and a discharge while the next word has
been accumulating force" (J.), and we have a form of summation; or
in other words, "a general mood accumulated while several words were
in mind at once, then all dropped and another general feeling came
to the front with an accumulation of other words." (F.) Here we have
a typical example of constellation where all the words and ideas are
implicitly present as a total attitude or disposition, the elements of
which become successively focalised into a series of associated images.
The last subject finds that "the emotional atmosphere often controls
the associations." Indeed, it would seem that occasionally for some
subjects this strong accompanying undercurrent of undifferentiated
emotional feeling is capable of bringing about trains of thought
independent of any logical connection. K. finds "the feeling to carry
one on"; H. finds the "point of departure the interesting idea"; all
find that the words change with the disposition, as may be verified by
a study of the lists of associations.

We are forced to conclude that the impulses to movement or other
emotional attitudes may act as determining factors in association,
which extended to an hypothesis would mean that the mode of transition
in the associated series is in the last analysis to be found in
delicate incipient motor tendencies to action, the psychic concomitants
of which are observable; that psychic states are both as to their unity
and organisation consequences of motor reactions which are implicitly
present as parts of a total reaction to the present situation. It is
these motor tendencies to action which determine what idea shall enter
consciousness. Just in so far as they become released they become
prolonged, accentuated, and form a nucleus for the new idea. To speak
of association independent of motor elements is merely to make an
empirical classification of successive states of consciousness.

There remains a psychical phenomenon which must be satisfactorily
accounted for before we go farther. An element of an idea, an idea or
a series of ideas may occupy consciousness to the exclusion of others.
If the second starting-point were not given, the associations would
undoubtedly follow the given one. Inhibition must then be one form of
"obstructed association," the inhibiting ideas being present to the
exclusion of the inhibited. But are we thus forced to say inhibition
is the "negative side of the association process," claiming that
all ideas not in consciousness are inhibited, and thus being forced
to conclude the conscious idea is inhibiting an unconscious idea,
which cannot exist (by the very definition and presuppositions of
psychology) until it is an object of consciousness. This would mean
that content of consciousness and inhibition are identical. On the
other hand, the notes and exemplifying facts of the tables show Dr.
Breese's fallacious position when he concludes that "because, obeying
the laws of association, the train of ideas takes one direction
rather than another can hardly be considered sufficient ground to
hold that the other possible train of ideas is inhibited."[137] He
has overlooked the possibility of two or more trains of associations
having been started and the associations of one starting-point are
excluded from entering the focus of consciousness by the direction
of the given series. Inhibition would then be the negative side of
fusion. The explanation must, as has already been demonstrated, be
psycho-physical in character. If these impulses to action have actually
been observed by the subjects we are justified in concluding that just
as in physiological inhibition one action excludes another, so the
correlative tendencies to movement of one idea exclude others.

By. observed that the image of the starting-point lingered and
inhibited subsequent ideas. The implication here, from our previous
reasoning, would be that not the ideational images, as such, but the
physiological motor concomitants, persisted and excluded others, and
this is why disparate terms give a "shock to the nervous system"
(A.), "require different lines of expression" (A.); and "one has more
momentum," as so many report. This would explain why the associations
of a new starting-point inhibit the associations of a former one; for
as the motor nervous impulses tend to work themselves out into action,
the reaction of the previous impulse will be suppressed by those of a
new impulse which enters, by the conditions of these experiments, an
attentive consciousness. Thus the prepotent impulses to action are the
conditioning factors in mental inhibition.

All this indicates that the basis of habit which has been the universal
principle of explanation of associations is inadequate. As Münsterberg
has pointed out, contrary to what we mean by habit, either idea may
bring to consciousness the other, in a manner independent of the order
of the original presentation. Extending our hypothesis to include the
formation of associations, the conclusion will be that in order for
two ideas to become associated they must be together in consciousness,
each as parts of a total experience, a total attitude; the motor
reactions of the ideas must be parts of a more comprehensive reaction
which includes both as simultaneous correlated motor impulses: when,
in future time, the reactions of the one are reëxperienced, there is
a sequence of infinitely delicate and complex impulses to movement,
and any tendency toward such reaction tends to reproduce the whole of
which it is a part, as each reaction is more or less bound up in the
integrity of the whole central nervous system.

FOOTNOTES:

[Footnote 133: Cordes, G.: Experimentelle Untersuchung über
Association, Phil. Studien, vol. 17, p. 30.]

[Footnote 134: Scripture, E. W.: Elements of Experimental Phonetics, p.
142.]

[Footnote 135: Calkins, M. W.: Memory and Association, Psychol. Rev.,
vol. 5, p. 451.]

[Footnote 136: Calkins, M. W.: Association, Psy. Monograph, no. 2, p.
46.]

[Footnote 137: Breese, B. B.: On Inhibition, Psy. Rev. Monograph, vol.
3, p. 15.]




DISSOCIATION

BY C. H. TOLL


The purpose of this investigation, of which the following gives a
preliminary report, was to compare the tendency to associate by
contiguity, with the tendency to associate by similarity.

In every series of stimuli to which one gives attention there is
tendency to association by contiguity. But some similarity among
certain elements of the series may produce a dissociation of the given
elements into two series with some bond of similarity in each. This
is a matter of common experience, as when you find you can read your
newspaper and listen to your neighbors' conversation at the same time,
understanding both, although the actual order in which the several
words are perceived would form a meaningless mixture.

We may say dissociation is always accomplished by a tendency to
association by similarity overcoming the constant tendency to
association by contiguity. Study of the relative efficacy of the two
may therefore be called a study of dissociation. The tendency to
associate by contiguity might be measured in two ways.

First, when one attempts to learn a series in exactly the given order,
the number of errors in the series as recollected may be taken as an
inverse indication of the strength of association by contiguity. The
three kinds of error possible in nearly all of the experiments were
Omissions, Displacements, and Imperfections. All of these three have
been tabulated. But the number of elements omitted seems considerably
the most reliable as an indication of the degree of inadequacy of the
associative tendency. The cases of displaced or imperfect elements are
comparatively few: moreover, Displacements and Imperfections are not
mutually exclusive categories. A single element may be both imperfectly
recollected and wrongly placed in the recollected series. On the whole,
it seems that the number of given elements which were omitted in the
recalled series is the most positive and reliable of the errors. Our
conclusions are based on the Omissions.

Second, when one makes no attempt to learn the series, simply giving
attention to each element as it comes, and afterward lets the elements
recur spontaneously, the number of cases in which a recollected element
is followed by an element given contiguously may be taken as a direct
indication of the strength of association by contiguity.

Tendency to association by similarity can evidently be measured in
the same two ways, by counting errors when one purposes to learn the
series as two groups of similar elements, and by counting sequences of
similar elements when one avoids any effort to learn the series and
recollection is spontaneous.

In the first seven experiments we used the first method. The errors
made when the purpose is to associate by contiguity can then be
compared with the errors made when the purpose is to associate by
similarity, an equal number of series, given under the same conditions,
and of identical character, being given in each case.

In the last four experiments we have used the second method. The number
of sequences of elements given contiguously can then be compared with
the number of sequences of similar elements.

Five subjects have coöperated in this, but the experiments were
strictly individual, one observer being alone in the room with the
experimenter. Each test lasted about an hour. As a matter of course,
the results have been calculated for each of the five subjects and
their agreements and disagreements have been carefully considered. But
as this first report is to indicate merely the general tendency, we
give here at first only the average of the five persons.

The experiments have varied as to the kind of elements used, the manner
of presentation, the time allowed, and the manner of recording the
recollected series. But throughout each experiment the series were of
one identical type, while the individual elements were altered in each
series.

In the experiments where the series were to be learned, some in the
given order, some dissociated by similarity, it was found rather
confusing to turn from one method to the other; so several consecutive
series were learned by one method, and then several by the other, four
alternations being made each hour to neutralize any effect of practice
or of fatigue.

The series were of course different in kind in the several experiments,
but were usually of eight or of ten elements. Half of this number had
some distinct characteristic in common, the other half some other
characteristic. In some experiments these elements were alternated, in
some arranged irregularly.

In the first eight experiments the subject wrote down the elements
recalled, as soon as the series had been given. In the last three the
subject spoke the elements recalled.

In all cases where the first method of measurement was used, the time
allowed for learning the series was made a little too short to permit
of learning the series perfectly. Since comparison of the number of
mistakes was our method, we naturally had to make sure there would be
mistakes to compare.

The details of the several experiments were as follows:

(1) The elements were letters and numbers. They were about 12×8 mm. in
size and were printed on white cards 15×30 mm.

Five letters and five numbers were placed, alternately, in a straight
row on a sheet of white cardboard. The series was then exposed to
the subject by turning up the small tin shutter of a screen that was
clamped to the table-edge.

The time of exposure was measured with a stop-watch and was constant
throughout the hour for each individual subject. Four seconds proved
the best time for most of them, but in one case it was necessary
to allow only three seconds. Twenty series were presented during
each hour, ten for each method of memorizing. There were duplicates
of all the numbers, and of eight letters, but not more than two
of any element. Selection in forming the series was by chance. In
dissociating, the letters were separated from the numbers.

As soon as the exposure was ended, the subject wrote down the elements
recollected, trying to preserve their relative order. This recollected
list was then copied beside the operator's record of the given series,
so making the errors apparent.

(2) The elements were all letters, printed as before, and the alternate
cards were placed half their length out of alignment with the original
row.

The method of presentation, the length of exposure, the number of
elements presented, etc., were as in no. 1. In dissociating, the
letters on one level were separated from those on the other.

(3) The elements were all letters, printed as before, and five of the
ten elements presented were placed out of alignment. But the disaligned
cards were at irregular intervals and often in groups, and were only a
quarter of an inch out of alignment. This order was varied each time,
but without any system.

The other details were as in no. 1.

During this experiment I came to notice the effect produced by the
natural tendency to learn the five elements of the dissociated series
in a rhythmical form, thereby increasing the ability to retain them;
while there appeared to be no natural tendency to apply any such
inclusive rhythm to the ten elements of the series when learned in the
given order. To counteract this effect the subjects were instructed
to consider the series, when learned in the given order, as two
consecutive series of five elements each, and to use the same natural
rhythm in learning these as they did in the dissociating. But this
correction was not made in the first two hours, nor very perfectly in
the rest.

(4) The elements were all numbers, printed as before, five of the ten
being placed a quarter of an inch out of alignment, and in irregular
groups, precisely as in the last experiment.

The time was reduced to three seconds for some and two seconds for the
others. Details of presentation were as described in no. 1.

This time all the subjects tried to neutralize the effect of the
instinctive rhythm for the five-element series by learning the
ten-element series in two groups of five elements each.

(5) The elements were all nonsense syllables, each consisting of a
vowel between two consonants, printed on white cards 20×20 mm. Eight
of these were placed in an even row on a sheet of white cardboard, and
four of them were marked by laying a quarter-inch strip of blue paper
over the bottom of the card. The serial position of the marked cards
was irregular, and was altered each time.

Ten seconds was given to some subjects, eight to the others. Other
details of exposure, etc., were as in no. 1.

In learning the series in the given order, the blue markings were
ignored; but in dissociating, the marked and unmarked syllables were
learned in separate groups.

There seemed to be no rhythmical tendency; but to be safe the subjects
were instructed to learn the straight series in groups of fours.

Seven series were given to be learned in each method during the hour
with each subject.

(6) The elements were one-syllable nouns, alternated with nonsense
syllables, all spoken by the operator. The nonsense syllables were all
different from those used in the preceding experiment: the nouns were
ordinary words, and were so arranged as to avoid any obvious sequence
or relation among them. Very few, if any, were used twice in one hour.
Five nouns and five syllables were given in each series.

The elements were spoken at the rate of forty-six a minute, timed by a
metronome which was muffled in a heavily padded box so that its sound
was no disturbing factor. The speaker sat within three feet of the
subject and enunciated as distinctly as possible.

Dissociation was performed as previously: in each hour eight series
were dissociated, and eight learned in the given order.

(7) The elements were one-syllable nouns, spoken as before, alternated
with nonsense syllables, printed on small white cards. The nouns were
all different from those used in the previous experiment: the nonsense
syllables were the same, but were this time printed, in letters 10 mm.
high, on cards 40 mm. square. They were exposed by sliding them, one at
a time, in front of an opening in a cardboard screen which was fastened
to the table-edge.

The optimum rate for presenting the elements was found to be about
forty a minute, measured with the metronome.

Five nouns and five nonsense syllables were given in each series. Eight
series were given to be learned in the given order, and eight to be
dissociated into separate series of nouns and of syllables.

(8) The elements were names of mammals, alternated with names of cities
of the United States, all spoken. The names were all fairly familiar.
Ten elements were given in each series.

The interval in reading was planned to be long enough for some
appreciation of the meaning of the words, but not enough to permit
mental repetition of the preceding elements. Any mechanical
time-measurement was found impracticable.

The subjects were instructed to avoid any effort to memorize the
series, simply receiving each element as given.

After the last element there was a pause of about two seconds, to
decrease the mere sound-recollection of the last few elements. Then the
operator repeated, in an altered tone, one of the given elements. The
subject at once wrote down the first element that came to mind, then
the next, and so on.

In the seven preceding experiments the set of series presented had been
different for each subject, though of course identical in character.
But in this experiment and the following ones the lists of words were
identical as read to each subject. The same element was repeated for
each. Sixteen lists were given.

(9) The elements were nouns. In each series five names of similar
objects were alternated with five names of a different sort of objects,
_e. g._, names of fishes with names of poets. All were read, as before.
In each series new sorts of objects were chosen. The subject never knew
what sort of words were to be given; the subjects agreed this was not a
disturbing factor to them, and it obviated the tendency to think what
words would probably be given, as is natural when the general character
of the series is announced beforehand.

The subjects were instructed to be passive during the reading, and
during the four-second pause that followed, avoiding mental repetition
of the words. Then the operator gave a signal and the subject repeated
aloud the words as they happened to be remembered. The words being
numbered on the list from which they were read, the operator was able
to record the words as fast as spoken.

The subjects were instructed to give the word which they found to be
foremost after they had spoken the preceding one, rather than to try to
repeat a group of words which usually appeared simultaneously at the
first effort of recollection, but which usually faded while one of them
was being spoken.

The same sixteen series, of ten elements each, were given to each
subject.

(10) The elements were nouns, the ten presented in each series all
being names of similar objects, _e. g._, flowers. Five were spoken,
alternated with five printed on small cards which were shoved in
front of a 10×10 cm. opening in a cardboard screen fastened to the
table-edge. Cards were 40 mm. square, the words printed by hand, but
carefully, in letters 10 mm. high.

A series was given in about 13 seconds, but the time was not
mechanically measured; it was at a rate which some practice showed to
give a fair time to comprehend each element.

As before, the subjects were told to be passive until, after a
four-second pause at the end of the series, the operator gave a signal.
Then the recollected words were spoken.

The class of nouns was different in each series.

(11) The elements were nouns. In each series five of some familiar
class were alternated with five of some other familiar class. The
classes were different in each of the twelve series given.

From this regular series of ten, five were chosen irregularly, and
were printed on cards as in no. 9. The remaining five, of course also
irregularly placed in the series, were spoken. This irregularity was
different in each series. Thus some words of one kind were spoken,
the rest printed; some words of the other kind were spoken, the rest
printed.

The other conditions were exactly as in the last experiment.

       *       *       *       *       *

A table for the individual subjects, indicating not only the omitted
but also the displaced and imperfect objects would have, for instance,
the following character: C indicates that the effort was made to
associate by Contiguity, S by Similarity.


SPOKEN NOUNS, ALTERNATED WITH PRINTED NONSENSE SYLLABLES

             Nouns Omitted   Syll. Omitted   Displaced   Imperfect
                C     S         C     S       C     S     C     S
    Turley     13    16        21    14       7    13     6    10
    Emerson     4     5        26    16       7     4     4    13
    Miss Kent   5     8        15     8      18     5     9     4
    Flexner     4     6        10     9       7     3     8    16
    Toll        8     7         8     2      10    12     8     3

    Total      34    42        80    49      49    37    35    46

If we consider total results only, and among them only the omitted
elements, we come to the following percentages. They give the
percentage of the errors of omissions among the elements recalled.

    1. Letters and numbers alternated                        C 26.  S 10.8
    2. Letters, alternatingly disaligned                     C 21.2 S 15.
    3. Letters irregularly disaligned                        C 23.8 S 22.4
    4. Numbers irregularly disaligned                        C  7.  S 20.
    5. Nonsense Syllables, irregularly marked                C 27.5 S 27.5
    6. Nouns and Nonsense Syllables alternated, spoken       C 35.  S 37.2
    7. Nouns and Nonsense Syllables alternated, nouns spoken,
         syllables printed                                   C 28.5 S 22.7

In the second group, experiments 8 to 11, not the errors of omission,
but, as explained above, the different kinds of reproduced elements,
had to be analyzed with special reference to the question whether a
sequence linked two contiguous or two similar objects. In the following
table the total number of recalled sequences is taken as basis and
the different kinds of sequences are given in percentages of it. The
elements themselves are described above. B means a break, that is, a
sequence without similarity or contiguity.

   8. Dissimilar elements, similarly presented             S 45  C 28  B 28
   9. Dissimilar elements, different kind in each series   S 53  C 25  B 21
  10. Similar elements, dissimilarly presented             S 54  C 20  B 26
  11. Dissimilar elements, dissimilarly presented
                                            S (Meaning)      27  C  7  B  8
                                            S (Presentation) 13.

The results by the first method of measurement may be summarized
as follows, though the first and third conclusions are weakened by
disagreement among the individual subjects.

_A._ When the only dissociating factor is some slight unessential
feature (a bit of color on the card, a slight disalignment), this
similarity and contiguity are nearly equally efficient. No. 3 and no. 5.

As this unessential feature is made more striking (disalignment half a
card-length), the strength of similarity increases, only three fourths
as many errors being made in dissociation as in contiguous association.
No. 2.

The case of no. 4 (all numbers) is of little or no value. The time
allowed for learning had to be made short enough to ensure the
appearance of some errors; perfect recollection would obviously give
no basis for comparison. And the time had to be so short in this
case (only two seconds for some of the subjects) that the additional
eye-motions and adjustments necessary in dissociating took time enough
to spoil the results.

_B._ When the only dissociating factor is in the meaning of the
elements (letters and numbers), this similarity is stronger than
contiguity, only one half as many errors being made. No. 1.

The results of no. 6 do not support this proportion, but its results
are not consistent, while those of no. 1 are.

_C._ When both meaning and manner of presentation are combined as
dissociating factors (nouns and nonsense syllables, seen and heard),
this similarity is stronger than contiguity, only three fourths as many
errors being made.

But this method of measurement is not well adapted to series of
auditory elements, so this experiment is unsatisfactory. No. 7.

The results by the second method of measurement may be summarized as
follows:

_A._ When the only dissociating factor is in the meaning of the
elements (names of different sorts of objects), this similarity is
stronger than contiguity, twice as many similarity sequences as
contiguity sequences being recalled. No. 8 and no. 9.

_B._ When the only dissociating factor is in the manner of presentation
(to sight and hearing), this similarity is stronger than contiguity,
nearly three times as many similarity sequences being recalled. No. 10.

_C._ When both meaning and manner of presentation are dissociating
factors, these similarities are much stronger than contiguity, more
than four times as many similarity sequences being recalled.

_D._ When these two dissociating factors are opposed to each other:
(1) Four of the subjects show similarity of meaning much stronger
than similarity of presentation, from two to five times as many
similarity-of-meaning sequences being recalled. (2) One subject is
strongly and consistently otherwise, giving nearly three times as many
similarity-of-presentation sequences. No. 11.




MOTOR IMPULSES




THE ACCURACY OF LINEAR MOVEMENT

BY B.A. LENFEST


The starting-point for our investigation was the observation of
Woodworth[138] that there is a certain rhythm in which a certain
hand-movement is made with the maximum of exactitude, and which
represents thus an optimum for the periodical discharge of the
particular motor centre. Our question was whether this rhythm is a
constant one for all parts of the body, or whether different groups of
muscles produce the greatest exactitude in different periods; further,
whether secondary factors, like complexity of movement, resistance by
weight, fatigue, etc., influence this psycho-physiological optimum.

The investigation, however, showed soon the necessity to consider the
whole problem of the accuracy of rhythmical linear movements, and the
experiments are thus not always directly related to our starting-point.

There is very little material published that can be collected under
the subject head, accuracy of voluntary movement, and still less when
the enquiry is confined to the accuracy of straight lines or linear
movements.

The most suggestive contribution is that of Dr. Woodworth on the
accuracy of voluntary movement. He has collected consistently what can
be found up to the date of his publication, and the reader is referred
to pages 7-16 of his monograph for the most reliable collection of
authorities.

It must be said, as we run over the list from Goldscheider on the
threshold of perceptible movement, through the results of Hall,
Hartwell, Loeb, and Delabarre on "bilateral asymmetry" and comparisons
of right and left hands; consider Fullerton and Cattell in their
suggestive results, and Münsterberg's studies of movements; and finally
take the testimony of Bryan as to the growth of accuracy of movement
in children, that the vast accumulation of material bearing on reaction
time--and similar phenomena would be of more value if concerned more
with the accuracy and less with the production or perception of
movement.

A paper by Miss M. K. Smith, in the Philosophische Studien for 1900,
with the title, Rhythmus und Arbeit, concerns the influence of
rhythmical action upon the quality and quantity of work performed. The
method was to commit to memory nonsense syllables and letters.

The results show a tendency to take up a certain rhythm, especially
in the later results and after practice; easier memorizing if rhythm
is present; motor reactions, as tapping, nodding, or swaying of body
are noted frequently; the feeling of pleasure accompanies rhythmic
reactions. While there are no data as to accuracy, there is suggestive
matter bearing on the optimal rate and on the relations of compound and
simple movements of the hand.

As far as the writer knows, he is the first to present systematic
results as to the head and foot movement. The purposes of this enquiry
may be briefly stated as

(1) the collection of a large body of facts, bearing on the actual and
relative accuracy of straight-line movements possible with various
parts of the body, such as hands, arms, head, legs, and feet;

(Something like 340,000 lines have been drawn and calculated.)

(2) to introduce certain variations in the conditions attending the
production of ruled lines, such as

(_a_) to rule with the eyes opened and eyes closed, with other
conditions the same;

(_b_) to change the rate of ruling or interval between the production
of ruled lines; the rates chosen were 20, 30, 40, 50, 60, 70, 80, 100,
120, 140, 160, 180, and 200 beats per minute;

(_c_) to change the length of the normal or first line; the lengths
used were 14, 10, and 1 cm.;

(_d_) to impose a weight on the ruling hand to either retard or
accelerate the movement, choosing a weight of such magnitude that it
would be perceptible, but would not have mass enough to cause pain or
fatigue; 260 grams was used;

(_e_) to introduce a simultaneous movement of the free hand; _i. e._,
the one that did not carry the recording pencil, of a similar character
and extent but of opposite direction to the ruling hand;

(_f_) to record movements of both hands, of the head and of both feet;

(_g_) to conduct a series of experiments of similar character, as
regards time-rate and extent of movement, to the series presented by
Dr. Woodworth, with the idea of corroborating or disproving the results
of his investigations; lines of 140 cm. were accordingly chosen;

(_h_) to conduct a series of experiments where the subject chooses his
own rhythm or rate at which the easiest and best lines, subjectively
speaking, could be ruled;

(_i_) to find the rates of respiration and pulse-beats and find the
connection, if any, between them and the linear records.

(3) To examine, by variations of the number of lines ruled, the
questions of fatigue and persistence of the memory-image; series of
50 lines for the first year and of 20 lines for the second year, were
accordingly selected.

(4) To find the relations, if any, between constant errors and mean
variations, so called.


THE APPARATUS

It is proposed to give the briefest possible discussion or explanation
of the apparatus required for the investigation, it being desired at a
later stage to enter into a comparison of the method adopted here with
that of the only other investigation at all comparable to this one: the
research problem of Dr. Woodworth, already referred to.

The underlying principle has been to avoid complication in apparatus,
partly because of the delay and expense involved in working out, and
making up elaborate schemes for apparatus, but mainly because of the
advantage in duplicating this series of experiments, or of carrying
on related investigations, to be derived from a choice of such
parts, entering into the complete apparatus, as are at hand in any
psychological laboratory, or that can be obtained and set up at small
expense.

The use of smoked paper has been avoided, because a short preliminary
series, using the usual smoked-paper records, was found to give no
better results than did the method here adopted of ruling on white
paper with a soft pencil, and the labor was thus considerably reduced.

To the objection that the pencil-ruling is more difficult, and involves
more loss in friction and more complicated adjustments on the part of
the subjects, only one of fourteen subjects admits that this is the
case; and even if the testimony was unanimous as to the greater ease
of production of the smoked records, it would be no reason for its
adoption, since one of the first rules for all experimental work is
uniformity of conditions, and this is equally well attained in either
case.

The apparatus for free hand-movements and for the compound movements of
both hands consists:

(1) Of an adjustable wooden rest (see Fig. _A_) with a base (_a_) about
40 × 60 cm. hinged to a vertically adjustable flat board (_b_), called
the arm-rest, about 40 × 70 cm., and having on its upper edge two brass
pins or plates (_c_) about 30 cm. apart.

The pencil is started from one of these pins, depending on the hand
used, and moved until it comes in contact with a wooden rod that is
held against the opposite pin and which is of the right length to give
a movement of the pencil of 1, 10, or 14 cm., as desired.

The operator holds this rod in place for the first line ruled and then
instantly removes it, so that the second and all later lines are ruled
by memory of the first one, as closely in length to the first, or
so-called normal line, as is possible.

(2) The apparatus for actuating and taking care of the paper.

This consists of two drums (_d_ and _d´_, Fig. _B_) 20 cm. diameter
by 40 cm. wide, mounted on suitable supports about 1 metre apart, and
fastened to a table, with axes parallel.

The drum upon which the record is to be made (_d_) is adjusted close
to the arm-rest, so that each ruled line will be carried down and out
of sight before the next one is ruled, the pencil being held in the
position (_e_); note that the arrow shows the direction of rotation.

The second drum (_d´_) is actuated by a motor (_F_) through a round
belt (_g_), this motor being a clockwork type, with gear-changes and
adjustable vanes for varying the speed, and having the power derived
from a suspended weight (_w_).

The recording paper (_h_) transmits motion from (_d´_) to (_d_). This
paper consists of a strip about six metres long by twenty-eight cm.
wide, with one end pasted to (_d_), and then wound upon (_d_), leaving
enough to be carried to (_d´_) and pasted to the latter. As the paper
is unwound from (_d_), it is wound upon (_d´_), and, both to keep
the paper tight and to prevent too rapid unwinding of (_d_), it is
necessary to apply a friction-brake to the shaft of (_d_).

(3) A metronome, capable of being used for a range of 20 to 200 beats,
and a stop-watch, to enable the operator correctly to time the subject,
are in constant use.

[Illustration]

The metronome is set in vibration and the subject is permitted to take
his own time to start the ruling, the operator holding the wooden
rod in place with one hand, while the other hand holds the stop-watch
ready to start it the instant the subject's pencil is moved. There is
thus a personal equation for the length of period, but this is of no
consequence, as will be apparent when the method of calculation and the
use of the planimeter is considered.

In the series of records with the weight, it is impossible to run the
speed about 80 to 100 beats, unless the modification in apparatus shown
in Fig. _C_ is used; for the vibration of the string running from the
hand to the weight around a pulley is violent enough either to throw
the string off the pulley or cause the weight to jump so severely as to
render the records useless.

This is entirely obviated by the given method of using a heavy weight
acting with a small leverage (about 1 cm.) and thus moving only a short
distance, so that it is capable of operating at the highest speeds with
no perceptible shock or jump; the string is led to the hand or wrist
from a grooved pulley of about 12 cm. radius, so the highest velocity
of the weight is only about one twelfth that of the hand. This method
makes it possible to carry the weighted records to the highest speeds.

This same method is used for the head and foot records, with the
following additional apparatus; the string (Fig. _C_), shown leading to
the hand, is led horizontally over to and around a similar large pulley
on the opposite side of the table and either down to the foot or in a
diagonally upward direction to the head; so that movements of the head
or foot are faithfully recorded on the drum by means of a pencil held
in a block of wood, this block of wood being fastened on the horizontal
string in a suitable position for recording on the drum paper. The
pencil is kept against the paper by a light spring or elastic band.

The foot is connected to the string by a stirrup that prevents any
movement of the feet at all, unless the same is recorded by the pencil.

The head is furnished with a skull cap or harness consisting of
non-elastic webbing and stiffened, where the string is attached, by
a strip of sheet brass formed to fit the forehead or the back of the
head, as the case may be. The object of the brass strip is to prevent
a lost motion in the flexible webbing, that is found troublesome
otherwise.

It will be evident, then, that the weight is continually acting as an
accelerating or retarding influence in all records for head and feet,
but it is not considered objectionable, for it is a constant throughout
the series.

The other plan would require a circuit of cord leading in both
directions from the head or feet in a complete circuit, and would cause
in the opinion of the writer too much complication of apparatus.

The pulse-beats were taken by the stop-watch and wrist method so
familiar to the physician, while the respiration results were obtained
by the usual tambour apparatus for registering the chest expansion upon
smoked paper.


THE METHOD OF CALCULATION

Suppose that the drums have been set in rotation and that the paper is
unwinding from (_d_) and being wound on (_d´_), Fig. _B_, and suppose
that the subject has ruled series of 20 to 50 lines, as may be desired,
regulated by the stop-watch in the hands of the operator. The records
will appear much as Fig. 5 under the planimeter discussion, there being
for each speed one normal line to start and a series of lines following
and intended to be of the same length as the normal line. A series of
records, then, consists of 13 records of 20 or 50 lines, each running
from 20 to 200 beats per minute, the complete series having not less
than 260 and not more than 650 lines.

It should be added that the operator holds a pencil-point on the end
of each normal line just after the record of 20 or 50 lines is made
and turns the drum (_d_), thus marking a line nearly perpendicular
to the ruled lines and at the average or normal distance from the
starting-point; an absolutely correct record would show all ruled lines
ending on this line.

The calculation of this series of records by the ordinary method of
measuring each line, adding the lines of the series, averaging for the
constant error, and repeating the operation in a slightly different
form for the mean error or mean variation is of such enormous labor for
an extended investigation as to be beyond the capacity of one or of
several students; it is fortunate that the planimeter is at hand to be
employed in averaging each series, and this instrument has therefore
been selected as overcoming this difficulty.

It is desirable to consider the method employed by Dr. Woodworth to
overcome this danger of excessive computation, and it will now be
subjected to a critical and comparative examination.

He says, page 19 of his monograph on the Accuracy of Voluntary
Movement, that the subject's sole duty was to make the present line
equal to that immediately preceding, and the width of the slot was
so adjusted that the subject could see only the line just ruled.
After discussing certain matters of memory and its relation to the
memory-image, in the attempt to support this changing normal plan,
he confesses, on page 20, that this device is advantageous in much
simplifying the most tedious part of the graphic method, that of
computation.

While this is undoubtedly true, it needs careful scrutiny before
adoption, for, on the same page, he says that one source of error in
the method of making each line equal to the preceding one is that the
different movements in the same series are not comparable, but the
positive constant error is cumulative in its effect, and the normal
tends to become longer and longer.

Some relation between this source of error and such a record as shown
on page 29, Fig. 2, is evident, for, while it should be noted that this
cumulative effect is peculiar to a series of lines for one speed, it
has further a tendency to produce overruling at all speeds, and the
natural result is to increase the error unduly and unnaturally for the
higher speeds or as the speed increases, because there is then less
time for the discrimination and choice that will tend to shorten the
ruled line. It may be predicted, then, that Dr. Woodworth's method
will show a slight lengthening of normal between lines at slow speeds
and a much greater one at high speeds, the effect being to introduce
a variable factor that would have no existence were a better plan
adopted. The computation required for the average error is simple,
being dependent only on the first and last lines of a series, and it
is suspected that this very simplicity has led to its adoption and the
consequent neglect of certain serious sources of error.

He tells us, on page 20, that the constant and variable error may
well be isolated and studied separately, but indicates that they must
"somehow" be considered combined as nature has made them; that is,
analysis is desirable, but the synthetic method is more scientific.

This investigation will present data suggesting that

(1) Such a curve as that on page 29 of his monograph is not a
characteristic one and relations of length of ruled line, as well as
effects of weight, make it impossible to apply Weber's law or even the
law of Fullerton and Cattell in the way proposed by Dr. Woodworth.

(2) There is no relation, mathematical or other, between constant and
mean errors, and they not only may be but must be isolated and studied
separately, if an investigation is to be conducted in the interests of
scientific exactness.

It will be necessary to reject the method of Dr. Woodworth if the
most reliable results are desired, in which case the planimeter is a
necessity.

The theory of the planimeter cannot be developed at this place; every
physicist and engineer is acquainted with it. The writer believes he
was the first to apply the planimeter to the calculation of results
from psycho-physical data for averaging both mean and variable errors.
More than 340,000 lines were involved, each demanding two measurements.
The best type of planimeter for general use and the one used here is
the Amsler adjustable-arm form.

In Fig. _D_ is shown a record taken at twenty beats per minute that
will both explain the method of computation and show how the planimeter
has been used to find the constant and mean errors.

[Illustration: Fig. D

Hylan-20 beats. L.H.E.c.-2-13-Ό1.]

The record, as made and ready for computation, is not provided with the
line _cd_ or with the dotted lines that connect the ends of the ruled
lines. The line _ab_ is drawn by turning the drum of the apparatus with
a pencil held at the end of the normal or left-hand line _af_, which
was here 100 mm. long.

The tracing-point of the planimeter being placed at _a_, a reading
is taken, which was in this case 1486; after following with the
tracing-point the dotted path to _g_ and returning, via _gb_ and _ba_,
a second reading is taken, which was 1248; subtracting gives 238,
which should be read 2380 square mm. for the area of the space _agba;_
dividing by the distance _ab_, in this case 119 mm., gives the average
height, which is + 20.0 mm., the plus sign suggesting that the distance
thus found, which is the constant error for the series, be laid off in
addition to or beyond _Fa_.

This being done, a line _cd_ is drawn parallel to and 20.0 mm. from
_ab_, as the mean line of constant errors.

To find the mean error of the series a slightly different method is
necessary.

Place the tracing-point of the planimeter at _c_ and read vernier,
giving 1916; follow the dotted path from _c_ to _h_, the straight line
from _h_ to _i_, the dotted path from _i_ to _k_, the straight line
from _k_ to _l_, the dotted path from _l_ to _m_, the straight lines
from _m_ to _n_ and _n_ to _g_, the dotted path from _g_ to _m_, the
straight line from _m_ to _l_, the dotted path from _l_ to _k_, the
straight line from _k_ to _i_, the dotted path from _i_ to _n_, and the
straight line from _h_ to _c_, when a second reading is taken, which
was in this case, 1806. Divide the difference of these two readings,
1100 mm., by the length of _cd_, 119 mm., and the result is 9.1 mm., or
the mean error (mean variation).

It will be noted that this method gives the sum of the errors from
the mean line _cd;_ that is, the same result would be obtained if the
tracing-point were (1) carried from _c_ around all the area below _cd_,
and this area were calculated as before; (2) carried from _c_ around
all the area above _cd_ and the area measured as in other cases; and
(3) these two results added and averaged.

To apply the method for _ab_, or constant error computation, to _cd_
should give equal readings at _c_ or a 0 mean error, a result evidently
incorrect in the record selected.

After averaging results by the planimeter, the collection of data has
been arranged by months; the record for one month only can be presented
here, but the method of tabulation is the same throughout.

Each figure given for N, M, c and v, in the accompanying typical table
for the month of May, 1904 (pages 495-499), is the average from 20 or
50 lines, ruled as already shown, Fig. _D_.


RESULTS

It is necessary to observe that the limits of space imposed on the
writer preclude all but the barest outline of the deductions to be
drawn from the investigation, and to this fact is due whatever of
dogmatism is inherent in the argument; for it is manifestly impossible
to present all the material, and the writer asks, then, the indulgence
of the reader when he claims to have impartially examined and presented
the evidence.


HAND MOVEMENTS

    Simple movements
    Lines 14 cm. long.


TYPICAL SERIES FOR THE MONTH

    Key.
    v    = mean error.
    R.H. = right hand.
    R.F. = right foot.
    E.O. = eyes open.
    si.  = simple motion
    N    = normal line.
    Unit = 1 mm.
    L.H. = left hand.
    L.F. = left foot.
    E.C. = eyes closed.
    co.  = compound motion.
    M    = mean line.
    b    = beats per minute.
    c    = constant error.

                    See                              Beats per minute.
    Day. Subject. Key. 20   30    40   50    60   70   80    100   120


    6  Hylan.     N  10   10.5  11   11    10    10    10    12    10
       L.F.E.O.   M  16.1 13.5  12.1 13.8  13.1  10.0  10.0  11.2  12.5
                  c  +6.1 +3.0  +1.1 +2.8  +3.1   0.0   0.0  -0.8  +2.5
                  v   2.8  3.5   1.9  0.9   1.6   4.6   2.0   2.7   1.2

     140   160   180   200

      12    10    11    11
      12.0   6.1  11.0  14.21
       0.0  -3.9   0.0  +3.2
       1.0   1.3   2.1   4.0


       Hylan.     N  10   11    10   12    10    12    11    11    11
       L.F.E.C.   M   8.9 10.5  10.9 11.4  13.1  10.3  11.8  13.1  13.9
                  c  -1.1 -0.5  +0.9 -0.6  +3.1  -1.3  +0.8  +2.1  +2.9
                  v   2.7  6.4   2.2  1.4   2.7   1.6   2.3   1.6   4.9

        11    11.5  11    11
      15.8  11.0  15.0  17.8
      +4.8  -0.5  +4.0  +6.8
       2.9   2.3   2.3   2.5


       George.    N  10.5 10    10   10    10    10    10    10    11
       L.F.E.O.   M  14.6  9.7   7.1  6.4   7.7   8.3   8.0  10.0  11.0
                  c  +4.1 -0.3  -2.9 -3.6  -2.3  -1.7  -2.O   0.0   0.0
                  v   2.1  1.6   1.8  2.3   0.8   0.7   1.4   1.7   0.8

        10    10    11    11
       9.0  10.0   8.0  12.5
      -1.0   0.0  -2.0  +1.5
       1.0   1.3   1.6   0.4


       George.    N  10    9     9   10    11     9.5   9    10     8
       L.F.E.C.   M  11.7  9.2  10.2 12.6  13.2  11.5  12.2   6.6   5.0
                  c  +1.7 +0.2  +1.2 +2.6  +2.2  +2.0  +3.2  -3.4  -3.0
                  v   2.0  2.9   2.5  0.7   0.5   1.0   0.8   2.0   4.1

        10     9     9    10
       8.0   7.6   8.0  11.0
      -2.0  -1.4  -1.0  +1.0
       3.1   2.5   2.1   3.7


       Moore.     N  10   10    10   10    11    11    11    10    10
       L.F.E.O.   M  15.7 18.8  17.7 16.7  18.3  18.5  16.5  15.6  16.0
                  c  +5.7 +8.8  +7.7 +6.7  +7.3  +7.5  +5.5  +5.6  +6.0
                  v   3.5  3.8   1.1  0.8   3.6   2.1   2.7   0.6   2.4

        10    11    10    10
      16.2  16.3  16.4  18.8
      +6.2  +5.3  +6.4  +3.8
       0.5   3.5   2.2   3.8


                N   10     10     10     10     11     10     10      9
      Moore.    M   19.6   15.3   15.3   14.3   14.9   13.5    6.7   13.4
      L.F.E.C.  c   +9.6   +5.3   +5.3   +4.3   +3.9   +3.5   -3.3   +4.4
                v    2.6    2.5    2.4    0.4    2.0    6.4    0.7    4.1

       9.5   11     11     10     10.5
      20.0   14.3   14.5   16.9   17.9
     +10.5   +3.3   +3.5   +6.9   +7.4
       2.6    3.8    1.1    3.1    2.8


                  N   10     10      9     10     10      9.5   10     10
    9   Angier.   M   13.6   12.5   11.6   11.6   11.5   11.7   13.0   12.9
        R.F.E.O.  c   +3.6   +2.5   +2.6   +1.6   +1.5   +2.2   +3.0   +2.9
                  v    2.9    2.3    1.8    2.3    1.7    1.6    2.0    2.7

        11     10     10     10     10
      12.2   12.2   13.5   22.1   16.5
      +1.2   +2.2   +3.5  +12.1   +6.5
       3.3    1.5    1.5    6.7    1.5



                  N   10     10     10      9.5   10      9     10     10
        Angier    M   11.4    7.7   13.8    8.3   11.4   11.5   11.0   10.3
        R.F.E.C.  c   +1.4   -2.3   +3.8   -1.2   +1.4   +2.5   +1.0   +0.3
                  v    4.8    1.3    3.0    2.6    1.4    1.7    1.8    1.7

                                          140 beats

         9     10     10     11     10     10
       9.9   12.8   16.7   17.7   12.4   12.0
      +0.9   +2.8   +6.7   +6.7   +2.4   +2.0
       1.8    1.4    1.7    4.2    2.1    1.7


                  N   11     10     10      9     10     11     10     10
        Huggins.  M   12.5    8.5   12.6    9.7    9.7   16.6   15.7   18.7
        R.F.E.O.  c   +1.5   -1.5   +2.6   +0.7   -0.3   +5.6   +5.7   +8.7
                  v    3.6    2.6    2.7    2.8    3.3    3.2    3.3    3.6

        10     11     11     11     10
      13.6   15.8    9.3   18.3   14.3
      +3.6   +4.8   -0.7   +7.3   +4.3
       3.1    2.7    2.9    3.8    3.0


                  N   11      8     10     10     11     10     10     10
        Huggins.  M    6.5   10.5   10.7    8.8   12.6   12.9   12.2   22.3
        R.F.E.C.  c   -4.5   +2.5   +0.7   -1.2   +1.6   +2.9   +2.2  +12.3
                  v    1.4    2.1    1.8    1.6    3.7    3.1    2.0    4.1

      10.5   10     10     10     10
       9.9   21.8   12.0   12.5   16.3
      -0.6  +11.8   +2.0   +2.5   +6.3
       1.4    5.0    2.3    3.6    1.8


                  N    9      9      9      9      9.5    8      7     10
    13  Lenfest.  M   12.7   11.9   11.1   11.3   12.1    8.7   11.6   12.6
        R.F.E.O.  c   +3.7   +2.9   +2.1   +2.3   +2.6   +0.7   +4.6   +2.6
                  v    3.7    2.4    1.9    2.3    2.8    0.2    2.8    2.1

      10     10      9      8     10
       9.0    8.6   11.7    7.0   11.8
      -1.0   -1.4   +2.7   -1.0   +1.8
       2.0    2.1    3.2    1.4    4.1


                  N   12     11     11     10     10     10.5   11     11
        Lenfest.  M   14.9   14.7   12.2   13.5   12.6   11.9   12.6    8.7
        R.F.E.C.  c   +2.9   +3.7   +1.2   +3.5   +2.6   +1.4   +1.6   -2.3
                  v    3.3    2.2    3.5    3.5    3.2    3.0    2.6    1.7

      11     11     10     10     10
       9.1    8.9    8.4   14.1    6.9
      -1.9   -3.1   -1.6   +4.1   -3.1
       2.4    2.8    2.9    4.1    2.3


                  N   12     11     11     11     11     10     11     10
         George.  M    8.3    9.0    8.6    8.8    8.7    7.0    7.4    7.9
         R.F.E.O. c   -3.7   -2.0   -2.4   -2.2   -3.3   -3.0   -3.6   -2.1
                  v    5.0    3.3    2.3    0.1    3.3    2.9    2.4    3.1

      11     10.5   11     10     10
       6.8    9.7    9.0    5.8   11.0
      -4.2   -0.8   -2.0   -4.2   +1.0
       1.5    4.6    2.9    2.9    3.9


                  N   12     11     12     11     11     11     10.5   11
         George.  M    7.0    7.7   14.1    8.3   10.3    8.4    8.7    7.8
         R.F.E.C. c   -5.0   -3.3   +2.1   -2.7   -0.7   -2.6   -1.8   -3.2
                  v    2.5    2.6    2.4    3.7    2.9    2.3    1.6    2.1

      10     10.5   11     10     10
       6.4   10.0    9.0    7.9    7.3
      -3.6   -0.5   -2.0   -2.1   -2.7
       1.9    3.7    2.9    1.8    3.8


                  N   11     11     11     11     12     10      9     11
         Moore.   M   14.5   16.6   16.5    9.0   17.0   10.7   10.4   11.9
         R.F.E.O. c   +3.5   +5.6   +5.5   -2.0   +5.0   +0.7   +1.4   +0.9
                  v    1.7    2.2    1.9    1.8    1.3    2.2    2.6    2.2

      10     10.0   10.5   11      9.5
      11.5   12.5   12.5   15.1   12.6
      +1.5   +2.5   +2.0   +4.1   +3.1
       2.2    2.5    2.5    1.4    0.8


                  N   11     10     11     11     11     10     10      9
         Moore.   M   14.3   13.2   15.9   12.3   17.9   10.0   11.3   15.9
         R.F.E.C. c   +3.3   +3.2   +4.9   +1.3   +6.9    0.0   +1.3   +6.9
                  v    1.8    2.8    1.6    2.1    1.7    2.4    3.2    2.2

      11     11.0   10     10     10
      14.8   15.5   13.4   11.8   12.2
      +3.8   +4.5   +3.4   +1.8   +2.2
       2.7    1.4    2.2    1.8    2.2


                  N   11     11     10     11     10     10     10     11
      16 Angier.  M   14.3   13.5    9.3   14.4   14.1    9.5   15.1   12.7
         R.F.E.O. c   +3.3   +2.5   -0.7   +3.4   +4.1   -0.5   +5.1   +1.7
                  v    1.7    1.9    2.0    2.4    2.0    1.1    2.4    2.8

      10     11     10     11     11
      11.1   12.6   13.7   11.0   16.6
      +1.1   +1.6   +3.7    0.0   +5.6
       2.0    2.0    2.7    1.3    3.5


                  N   11     10     10      9     10     10.5   10     10
        Angier.   M   14.4    4.8    2.5    3.6    5.8    6.7    8.0   11.0
        R.F.E.C.  c   +3.4   -5.2   -7.5   -5.4   -4.2   -3.8   -2.0   +1.0
                  v    3.2    2.7    2.3    3.6    2.7    2.3    1.3    2.8

      11     10     11     11     11
      13.5   11.1   13.9   10.1   11.0
      +2.5   +1.1   +2.9   -0.9    0.0
       2.8    1.4    2.4    1.8    2.2


                  N   97     97     98     97     98     99     97     98
        Huggins.  M  107.6  107.8  100.4  114.2   98.0  110.3   89.4  102.5
        L.F.E.O.  c  +10.6  +10.8   +2.4  +17.2    0.0  +11.3   -7.6   +4.5
                  v    5.9    0.6    4.8    6.5    5.3    6.7    6.4    5.5

      95     98     99     99    100
      99.2  100.8  106.4  101.5  108.9
      +4.2   +2.8   +7.4   +2.5   +8.9
       7.8    6.0    4.7    7.0   12.7


                  N   97     98    100     97     95     95     99     99
        Huggins.  M  102.8  104.3  115.1  101.4   94.1  102.0  106.4   93.4
        L.F.E.C.  c   +5.8   +6.3  +15.1   +4.4   -0.9   +7.0   +7.4   -5.6
                  v    5.8    8.6   10.8    9.9    7.9   10.4    8.1    6.8

      97     98     99     99    100
      87.7   98.7  108.4   97.7  110.8
      -9.3   +0.7   +9.4   -1.3  +10.8
       9.4    5.8    6.9    5.3    6.1


                  N   11     11     10     10     10     10     10     10
     20 Lenfest.  M   16.1   13.5   11.6   10.9   10.9   14.2   11.7   10.5
        L.F.E.O.  c   +5.1   +2.5   +1.6   +0.9   +0.9   +4.2   +1.7   +0.5
                  v    2.8    3.0    3.0    3.0    1.2    1.9    1.0    1.6

      11     10     11     11     10
      12.4   11.1   12.2    8.7   12.9
      +1.4   +1.1   +1.2   -2.3   +2.9
       3.2    5.1    1.9    1.7    4.2


                  N   12     11     10     10     10     11     10     10
        Lenfest.  M   20.4   16.8   11.6   11.2   11.8   14.0    9.0    9.3
        L.F.E.C.  c   +8.4   +5.8   +1.6   +1.2   +1.8   +3.0   -1.0   -0.7
                  v    2.5    3.6    3.0    1.9    1.6    2.0    3.0    1.2

      10     10     11     11     10
       6.2   10.0    8.5    5.1    8.3
      -3.8    0.0   -1.5   -5.9   -1.7
       1.9    0.3    1.8    1.7    0.8



                  N   98     97     94     98     98     97     98     98
         George.  M  104.0  100.4  102.6   99.3  105.1  111.1  101.9  100.5
        L.F.E.O.  c   +6.0   +3.4   +8.6   +1.3   +7.1  +14.1   +3.9   +2.5
                  v    6.4   11.3    9.6    7.7    5.3    8.0    3.7    5.6

      97     96     98     98     97
      96.4  102.0   97.4   94.2   95.8
      -0.6   +6.0   -0.6   -3.8   -1.2
       5.7    5.7    9.1    7.8    3.6


                  N   98     97     94     97     97     98     98     97
         George.  M   93.6   81.2   94.7   92.7  104.2   99.3   93.4   89.6
        L.F.E.C.  c   -4.4  -15.8   +0.7   -4.3   +7.2   +1.3   -4.6   -7.4
                  v    9.1    8.9    5.5    4.6    5.8    6.9    4.9    6.6

      96     98     98     99    100
      96.4   93.0   87.3  101.8   87.4
      +0.4   -5.0   -9.7   +2.8  -12.6
       7.8    4.8    7.7    5.6    5.7


                  N   97     99     98     97     96     96     97     98
          Moore.  M  106.1  106.9  105.2  103.0  104.7  108.6  102.7  106.9
        L.F.E.O.  c   +9.1   +7.9   +7.2   +6.0   +8.7  +12.6   +5.7   +8.9
                  v    4.3    5.6    4.6    5.2    8.0    6.0    7.4    6.9

      99     98     99     99     99
     110.9  105.7  111.2   99.9   93.0
     +11.9   +7.7  +12.2   +0.9   -6.0
       4.8    9.0    4.7   11.4    1.8


                  N   96     97     97     97     96     97     96     97
          Moore.  M  119.2   79.8   87.3   79.6   81.6   91.2   96.5  106.2
        L.F.E.C.  c  +23.2  -17.2   -9.7  -17.4  -14.4   -5.8   +0.5   +9.2
                  v   16.1    6.7   11.6    7.2    7.5    4.1    4.0    6.6

      99     98     97     98     99
     100.4   91.9  104.0   88.6   80.9
      +1.4   -6.1   +7.0   -9.4  -18.1
       4.0    6.1    5.4    6.6    8.2


                  N   98     97     95     96     96     97     99     97
     23 Lenfest.  M  105.7  110.8   97.4   93.8   97.8  100.2   99.9   92.9
        L.F.E.O.  c   +7.7  +13.8   +2.4   -2.2   +1.8   +3.2   +0.9   -4.1
                  v    9.3    1.5    3.6    5.1    3.6    4.4    2.9    3.3

      97     98     98     98     99
      86.1   96.6   91.9   80.9   83.0
     -10.9   -1.4   -6.1  -17.1  -16.0
       6.7    5.0    8.2    6.0    7.8


                  N   96     94     96     96     96     94     96     97
        Lenfest.  M  111.3   95.0   98.3  101.7  101.7  110.1   98.8   79.0
        L.F.E.C.  c  +15.3   +1.0   +2.3   +5.7   +5.7  +16.1   +2.8  -80.0
                  v    7.1    5.1    5.8    2.7    5.4    7.8   12.1    3.7

      97     97    100    100    100
      88.9   72.7   88.4   69.7   80.8
      -8.1  -24.3  -11.6  -30.3  -19.2
       5.6    8.0    7.4    8.4    7.0


                  N   96     97     96     96     98     98     97     97
        Huggins.  M  113.5  105.7  105.5   97.7   92.5   98.4   92.5  102.0
        L.H.E.O.  c  +17.5   +8.7   +9.5   +1.7   -5.5   +0.4   -4.5   +5.0
                  v    4.2    2.4    3.6    5.0    3.5    3.1    4.2    5.8

      95     98     98     99     97
      95.0  112.3  104.5  108.7   95.7
       0.0  +14.3   +6.5   +9.7   -1.3
       4.7    6.4    5.8    7.1    4.4


                  N   98     97     97     97     97     99     99     97
        Huggins.  M  103.5   82.8   86.1   87.8   87.7   94.2   92.1   94.5
        L.H.E.C.  c  +5.5   -14.2  -10.9   -9.2   -9.3   -4.8   -6.9   -2.5
                  v   8.4     6.7    5.4    3.5    6.7    5.4    6.0    5.8

      97     97     98     98     96
      99.9  111.5  101.3  113.7   95.1
      +2.9  +14.5   +3.3  +15.7   -0.9
       4.8   13.3    4.7    9.6    2.6


                  N          96     95     96     96     96     97     96
     27 Lenfest.  M         101.6   93.4   91.2   90.0   97.5   93.2   93.4
        R.F.E.O.  c          +5.6   -1.6   -4.8   -6.0   +1.5   -3.8   -2.6
                  v           6.2    5.1    3.4    2.8    5.5    4.5    3.7

      96     99    102    101    100
      89.8   97.5   88.6   96.8   66.2
      -6.2   -1.5  -13.4   -4.2  -33.8
       5.5    3.6    4.3    4.9    8.1


                  N   96            96     97     97     97     98     98
        Lenfest.  M  103.8         101.2   94.0  100.3   96.4  101.2  105.0
        R.F.E.C.  c   +7.8          +5.2   -3.0   +3.3   -2.6   +3.2   +7.0
                  v    6.2           7.4    4.5    3.8    3.5    4.7    2.4

      95     98    100     99    100
      78.0   98.5   88.8   85.1   65.8
     -17.0   +0.5  -11.2  -13.9  -34.2
       4.7    4.3    7.0    6.9    5.9

The records are averaged for nine subjects, three of them being
left-handed. For the right hand we find, for mean error, a reduced
error with visual control.

For constant errors, a similar result is apparent; when following the
eyes-closed curve one may note a large negative error 20-50 beats, and
a similar but larger positive error 70-160 beats, with a falling to a
negative error again at 200 beats.

This may be interpreted to mean a groping for the correct length of
line at the lower speeds when some time for reflective processes is
allowed, and an inhibitory effect on the motor discharge; later the
speed prevents this discrimination, and introspective testimony goes
to show that a mental conception of a barrier, beyond which one cannot
carry the pencil, is set up and kept more or less constant through the
help of the joint and muscular sensations. It would follow, then, that
this muscular stop is overestimated where reflection is not possible.

Finally, the falling-off of the length of the line is probably due to
physical inability to rule a line of the full length of 140 mm. at 200
beats per minute and an examination of some individual cases confirms
this opinion, for the lines may be started some distance away from the
origin apparently in order to end them at the correct point.

Curve inclinations are upward, for mean errors, with visual control,
while the eyes-closed records show no increase in error for the
increased speeds.

For constant errors, with visual control, there is a similar
inclination downward for both hands, with a 0 error at about 120 beats.
It should be noted that this opposite tendency in mean and constant
errors suggests that they should be kept separate in all computation.

The left-handed subjects have much better control of their left hand
than have the right-handed subjects, and they may dispense with visual
control to a large extent.

On the other hand, for right-hand records we find much the same
increase in irregularity and error for both left- and right-handed
subjects; they all must depend on visual control for reduction of
errors.

It follows that the non-visual control exerted by the left-handed
subjects on the right hand is as good or as great as for the
right-handed subjects; while they have the hand in which they may be
expected to excel under much better control.

It is not intended to present this as an argument for teaching
left-handedness, but it is certainly suggestive when considering the
question that ambidexterity be taught in early life.

It should be noted that two of the three left-handed subjects might be
expected, because of special training, to show marked manual dexterity,
while only one of the four right-handed subjects has had special
training along this line.

No extended discussion is appropriate here as to the question of what
portion of this extra ability of the left-handed subjects to react
accurately is due to practice and habit, _i. e._, is automatic, and
accomplished without reference to the sensory motor by-path to the
cerebral cortex; and on the other hand, as to whether the direct
sensory motor path via spinal cord or medulla is not cut off entirely.

For 140 mm. averages and free motion, we find in general

(1) a reduced error and greater uniformity of result at all speeds
where visual control is added, in the case of both mean and constant
errors and for all subjects;

(2) the mean errors for visual-control records show a rise along a
line whose equation is approximately _y_ = _px_, or the equation of a
straight line, where _p_ is an undetermined constant.

On the other hand,

(3) the eyes-closed mean errors show no increase or decrease in value
during the entire series;

(4) the constant errors for visual-control records show a drop from
positive errors to negative errors, along a line whose equation is
approximately _y_ = _qx_ or the equation of a straight line, where _q_
is an unknown constant, somewhat less in value than _p_ in the case of
mean errors; the constant error becomes 0 at about 120 beats;

(5) the eyes-closed constant errors follow the same equation for
left-handed subjects, using the left hand, but all other cases suggest
a curve of the parabolic form, having 0 constant errors at 60 and 180
beats and being convex upward.

Considering individual records for 14 cm.

A general survey of the charts suggests certain irregularities that
call for explanation, for there will be sudden large increases in
errors, that are explicable on the hypothesis that the subject has
temporarily lost control of the moving hand, that is, that fatigue is
to be noted.

While the purpose of the investigation has been to allow no lines to be
ruled while the subject was conscious of any such feeling, there being
a pause of any desired length to permit time for rest, it is to be
noted that a considerable amount of recorded data as to fatigue shows
that it is an unconscious or subconscious phenomena.

Further, the series of records have been arranged to occur from 20
to 200 beats and never in the reverse order, because of subjective
limitations, so it is reasonable to expect that during the period of
twenty minutes to one half an hour required for a series of records,
there will be lapses of volitional control entirely beyond the ken of
the subjects. It is to this cause rather than to pure chance that the
results will be attributed. With this exception, the individual records
show close agreement with their average.

The results obtained from a consideration of free hand-movements of 1,
10, and 14 cm. length are:

For 14 cm. lines

for the average of nine subjects:

The mean errors,

(1) increase with speed for eyes open;

(2) do not change in error with speed-change for eyes closed;

(3) visual control reduces errors for right hand, but does not for the
left;

(4) right-handed subjects alone gain from visual control.

The constant errors,

(1) decrease with speed in visual cases;

(2) increase with speed to the middle and then reduce to 200 beats for
eyes closed;

(3) left-handed subjects are more accurate for the left hand and can
dispense with visual control;

(4) all subjects need visual control for the right hand;

(5) left-handed subjects show less error throughout for non-visual.

For the individual cases,

the mean errors

(1) show evidences of loss of control or fatigue for some speeds, and
the average results are confirmed.

The constant errors show that the average deductions are confirmed.

Lines 10 cm. long:

Averages for seven subjects as regards mean errors have especial
interest for l.h.e.c. records, which alone show a rise of error with
speed-increase.

Noting that the records are overwhelmingly averages of right-handed
subjects (six to one), it is of interest to examine this record.

We may say, then, that for right-handed subjects, the voluntary control
for the right hand is not much improved by the introduction of visual
assistance; it is more marked for speeds of 100 beats or less than for
the high speeds. And in the latter case, it is under 10%; but when the
left hand is considered, a marked gain or 40 or 50% is apparent, when
the eyes are used except for the two lowest speeds.

As far as it is possible to offer any hypothesis from the few
facts tabulated, it may be said that right-handedness implies a
high development of muscular control, but slightly improved by the
introduction of the visual element, as far as the right hand is
concerned; but for the left hand muscular control comparable to the
right-hand control can be obtained only with visual control; in short,
I fail to find evidences of cross-education, where the visual element
is absent, nearer than about 50% of the mean error.

No clearly marked gain through visual control can be pointed to in
the case of constant errors; there is, to be sure, a slight gain in
steadiness and error-reduction, where eyes help in the case of both
hands, but not 5% in magnitude of the difference noted with mean errors.

Individual records show fatigue-points at 40 to 80 beats and again
above 140 beats, but there is no perceptible loss of control during a
series of lines ruled at only one speed.

It is apparent, when comparing with the 140 mm. records, that there is
no physiological reason why the subjects may not rule the full length
of a 10 cm. line at 200 beats, and the limit of movement for high
speeds is probably between 10 and 14 cm.

The constant errors are in general positive and only Me. shows a
tendency to underrule lines at high speeds.

For 10 cm. lines,

for the average record,

(1) the mean-error curve is horizontal for l.h.e.o., but otherwise
rises with speed-increase;

(2) visual control is a 10% gain for right hand and 40 to 50% for the
left, so right-handedness is prominent for the eyes-closed series;

(3) the constant-error curve for r.h.e.c. rises, but all others show
reduction of error as rate increases;

(4) visual-control gains are not over 5%.

The individual curves show,

for mean errors,

(1) marked-fatigue points for l.h.e.o;

(2) r.h.e.o. curve is horizontal, but all others rise;

(3) l.h.e.o. curve is about the same as r.h.e.o., but there is more
loss for non-visual series with left hand.

For constant errors,

(1) r.h.e.c. curve rises, but all others are horizontal;

(2) the eyes reduce errors especially for the left hand;

(3) overruling is prominent even at high speeds, for there is no
evidence that the lines are shortened at high speeds.

Lines 1 cm. long:

Averages are made for nine subjects, three being left-handed.

The eyes are effective in reducing mean errors and to a less extent for
constant errors.

A noteworthy feature of the constant-error record is that the errors
are positive with one exception, that of 20 beats with l.h.e.c. and
even this curve jumps rapidly above the 0 line.

In all cases the motor discharge is of sufficient magnitude to cause
overruling in cases of normal lines of one cm. The assistance afforded
by the eyes is not marked.

Certain evidence of an introspective character, that most of the
subjects offer, is to the effect that, "when I would do good, evil
is present with me"; that, where there is a decided feeling that the
muscular limit, if such a term be permitted, is exceeded, yet the
subject's will-power is not sufficient to inhibit the overruling; there
is a more or less vivid conscious error in the 10 mm. series for the
hands.

In regard to the relation of mean and constant errors, there is more
close uniformity than with the 140 mm. lines, but it is to be noted
that there is no comparison to be drawn between maximum or minimum
points; for example, at 100 beats the minimum points for r.h.e.o. agree
closely, but the maximum constant error matches the minimum mean error
at 100 beats for r.h.e.c.

We cannot predict, then, that a subject capable of closely ruling to
the normal will be able also to rule each line of the same length as
the rest of the series, or _vice versa_.

As in the case of mean errors in general note that subjects show a less
constant error and more regularity for their more dexterous member;
it is not true for the left hand that for left-handed subjects visual
control is a hindrance for accurate work; otherwise the same gain, by
use of the eyes, is to be noted for the rest of the records.

Individual records show close correspondence with the average of
results, and the latter may be considered fairly representative.

Almost the whole series shows the constant error positive, the most
consistent example being for J. with l.h.e.o.; this tendency to
overrun the 1 cm. lines is consistently uniform and has been elsewhere
commented on, so it may be left with the observation that the log shows
that the subjects were frequently conscious of this overruling, but
confessed inability to correct it.

Only in the case of Y. for the three left-handed subjects and for W.
among the six right-handed men does the left hand show less mean error
than the right hand, and all other cases show such an interweaving of
curves as to render it difficult to perceive any advantage that the
more dexterous hand possesses on the score of accuracy.

For constant errors:

For individual cases it is to be noted that for the left-handed
subjects J. is better for the right hand, Le. is indifferent, and Y.
prefers the left hand; while three of the six right-handed subjects
prefer the left hand and one is indifferent; thus giving still further
proof that a more dexterous hand is a fiction on the score of the right
or left-handed theory, when accuracy of straight-line movement is to be
considered.

    For 1 cm. lines,
    for the average, note
    for the mean errors:

(1) visual control reduces errors;

(2) errors increase for eyes open, but decrease for eyes closed, as
speed increases when considering right hand, but left-hand errors are
constant;

(3) as left-handed subjects are better for r.h.e.o. than right-handed
subjects, but not for eyes closed, it is suggested that visual control
equalizes differences in the subject's less trained hand.

For constant errors:

(1) visual control reduces errors;

(2) curves are horizontal in all cases;

(3) all errors are positive, showing consistent overruling;

(4) as visual control of the left hand is a gain for right-handed but
a hindrance for the left-handed subjects, the more practised hand is
probably able to dispense with visual control, and depend largely on
the muscular sense.

Mean and constant errors are not comparable. For the individual
cases we find a corroboration of the above and for mean errors: more
dexterous hand does not excel, and evidence against ambidexterity is
conflicting; for constant errors: overruling is consciously done.

Constrained hand-movements for lines 14 and 1 cm. long and for the
weight, both accelerating and retarding the movement, are to be next
considered.

Constrained motions are of two general types as examined by the writer.
Series of the records for the hands were taken at 140 mm. and 10 mm.
bases with a weight hung on the finger or fingers of the hand under
investigation; in one series the weight acted as a pull or accelerating
effect on the ruled line and in the other series the weight was imposed
as a retarding effect, tending to restrain the movement of the hand.

This weight was in all cases 260 grams, this weight being chosen as of
sufficient amount to have a perceptible effect, but not large enough to
cause feelings of pain or fatigue in any case.

The average for seven subjects, three being left-handed, is as follows:

In general, the mean errors for the right hand are less, and less
variable as compared with the left hand. The left-hand records are very
close to the corresponding right-hand curves, especially the portions
of the eyes-closed records 20 to 120 beats.

This may be said for both mean and constant errors. In general, mean
errors are reduced, and curves are more nearly straight lines when the
weight is added; also the weight reduces constant errors, and gains
more regular records at all speeds. It is to be noted as a point of
unusual interest that there is no apparent shortening of the line ruled
when the weight is hung on the hand, for the negative errors are less,
not more when the weight is applied.

In general, then, the imposition of a weight that will be small enough
not to cause pain or fatigue shows that both mean and constant errors
are reduced; that the amount of error is less variable over the range
of speeds used; that the records show no retarding effect, but that
the subject is both able to move the hand just as far as without the
weight, and do it with much greater accuracy.

Individual records for 14 cm. and weight-retarding show a marked
reduction in both mean and constant errors, and a less marked gain
in uniformity in every case. This tends to confirm the introspective
opinion of W. subject that the imposition of a retarding weight tends
to reduce errors of both classes and to cause greater steadiness.

It should be added that there is evidence of an occasional letting-go
of voluntary control, so to speak, resulting in a large increase in
mean error, as already pointed out, or a large increase in negative
constant error, as shown on all individual records, and it would
seem then that the matter of cortical control is more vital and
indispensable for the restricted movements.

The effect of weight-retardation on visual records is to reduce the
error and steady the ruling of the less dexterous hand to a much more
marked degree than for the well-trained hand.

In the l.h.e.c. records the lack of corrective effect of visual
control is marked, as in the case of free movements, but the dip in
the curve at 30 to 70 is not noted in the free ruling and should be
considered as a distinct shortening due to weight-retardation before
discriminative processes have oriented the subject.

Without considering the accelerating weight-records in detail note that:

The effect of weights (less than that necessary to cause pain or
fatigue), either tending to accelerate or retard motions, is to reduce
both mean and constant errors and to render more uniform or more
uniformly increasing or decreasing such errors, except in the case
of l.h.e.c., where constant errors are greater positively with the
weight-pulling and greater negatively with the weight-retarding, than
for free motions; that is, the effect of the weight is natural, and
shows no signs of inhibition in this particular case.

There is no such marked fluctuation in error for the pull-records
as was noted for the weighted curves, and it is further noted that
the individual pull-records are more bunched or consolidated about
some mean than are the free-movement curves. This suggests that the
accelerating weight is a decided help for accuracy and regularity, and
it would seem to call for less voluntary control than for either of the
other movements.

Further, as the effect of pulling weights is to equalize the accuracy
of movement of the hands, the hypothesis is proposed that weights
either accelerating or retarding the movements of the hand tend to
equalize their accuracy or to promote ambidexterity as far as accuracy
of straight-line mean errors is concerned.

L.h.e.c. rise for Ha., are horizontal for J., W., and Y., and slope
downward for the other subjects, the net effect being a slight downward
slope. The loss of accuracy and regularity when the visual sense is
inhibited is to be noted in every case, it being especially marked for
Bo., Li., and W.

As compared with the r.h.e.c., there is not sufficient evidence to
lead to the conclusion that the right hand is a more accurate member
than the left, but on the contrary the left-hand record for non-visual
control is lower for both weighted series than is the right-hand curve.
Contrasting this with the eyes-open records for free and weighted
movements, the visually aided results show a greater accuracy and
regularity for the right hand.

This leads to a proposition that the greater dexterity on the line
of accuracy, of one hand, that is the right hand for right-handed
subjects, and the left hand for left-handed subjects, is a matter
of visual control and is in no sense due to the muscular sense or
to automatic action, for without eyes we are ambidextrous as far as
accuracy of linear movements is concerned; the proposition needs
careful scrutiny in application to the general question, but is held to
be correct within the range of experiments.

We are tempted to extend this matter somewhat in the following way,
by saying that there is no evidence deducible from this research that
there is hereditary preponderance of activity or accuracy of one hand
or one leg (as shown later) over its mate, and the baby is brought into
the world with an equal capacity of accuracy of both members.

It is, then, an evolutionary matter, not racial but individualistic,
and right-handedness or left-handedness is largely a development after
birth. Our system of education is responsible for the over-development
of one hand, and such a case as that of Dr. Anderson of the Yale
University Gymnasium, who in class demonstration cannot instantly tell
which hand is being used to actuate the chalk at the blackboard, is the
normal symmetrically developed man.

The school reform for ambidextrous training is radical enough, but
seems a logical conclusion of the argument. Apologies are appended for
driving the argument beyond the limits of the investigation, but it is
hoped that the enquiry is at least suggestive.

For 14 cm. lines,

weight-retarding movements:

For the average of nine subjects:

The weight reduces errors and promotes regularity in the case of both
mean and constant errors, nor does it tend to cause underruling, save
in the case of left-handed subjects for l.h.e.c. records. There is a
gain, in general, when the visual factor is introduced.

For mean errors,

(1) right-hand curves are horizontal, while the visual records show
increasing error and l.h.e.c. a reduction of errors;

(2) the right hand gives slightly better results;

(3) note that l.h.e.c. record is equally good for free or weighted
movements.

For constant errors,

(1) r.h.e.o. and l.h.e.o. curve downward, while both non-visual curves
slope upward;

(2) the left hand seems equally efficient, as compared with the right
hand.

For individual cases,

note (1) fatigue-spots are more numerous than for free movements,
especially for the left hand;

(2) weight reduces both mean and constant errors and to a less extent
even records.

For mean errors,

(1) visual control reduces errors;

(2) the weight tends to equalize the accuracy of the right and left
hands.

For constant errors,

(1) there is no general testimony showing shortening of lines at high
speeds;

(2) the less trained hand is more helped by the weight, especially for
non-visual work.

The evidence for right- and left-handed subjects is inconclusive, and
we cannot finally say that the more trained hand is capable of greater
accuracy.

Weight-accelerating movements:

The average of seven subjects:

The accelerating weight reduces mean and constant errors, and improves
regularity of curves, except for l.h.e.c. constant-error record. There
is some evidence that a pull causes overruling, while a retarding
weight causes underruling, but there are exceptions enough to warrant
care in finally accepting this statement. Visual control with
accelerating weight reduces error more than the weight acting alone.

For mean errors,

(1) weight reduces errors for r.h.e.o. and l.h.e.c. as compared with
free-movement records, while the other two curves are inconclusive;

(2) visual sense helps in accurate ruling;

(3) non-visual records are not reduced, as a rule, from the results of
free motion.

For constant errors,

(1) the accelerating weight tends to greater accuracy, with an
exception for the-non-visual records.

No testimony of marked importance is to be noted in comparison of
right-handed and left-handed subjects; the more trained hand shows
greater accuracy in some cases, but fails to excel in others; so the
data is inconclusive.

For individual cases we find:

(1) the acceleration records are more accurate and regular, and present
fewer lapses than the free or retardation results, suggesting greater
ease with weight assisting;

(2) visual control is prominent throughout, and evidence shows that
this sense is the greatest factor in the predominance of the more
trained hand; the non-visual records should and do show no marked
difference in the hands;

(3) a weight tends to equalize accuracy of hands;

(4) the overruling effect of weight is over-corrected in some cases for
constant error of low rates.

Constrained movements of 1 cm.:

The average is of seven subjects, three of them being left-handed:

With weight-retarding movement, there is no reduction of mean error
with visual control of right hand, but there is with the left. Constant
errors show little reduction for either hand with eyes open.

The facts would seem to warrant the hypothesis that, for the left hand,
a movement uncontrolled visually, whether restricted by a weight or
not, can be made with greater accuracy, when time is permitted for
discriminative and reflective processes and visual-control results in
about the same error whatever the speed, while the right-hand motions
show no such evening effect of visual control with the weight-records
or even reduction of error; the free movement, however, does show a
reduction of error.

A general statement may be deduced that, for lines of 10 mm. in length,
there is no difference in either mean or constant errors, when a weight
is imposed to cause retardation, provided the weight is not large
enough to cause pain or fatigue.

By separating the averages for right- and left-handed subjects, it may
be further said that:

Visual control is not efficient to reduce the error and no particular
gain in regularity can be noted. The left-handed subjects show, for the
left hand, much better results without visual control as far as the
free motion is concerned.

While somewhat contradictory, it may be stated that constant errors
are reduced by the weight addition, and there is some evidence leading
to the belief that the ruled line is shorter when the weight acts as a
retarding influence.


INDIVIDUAL RECORDS

Considering lines 10 mm. long with a retarding weight:

A glance over the seven individual records shows some considerable
increase in both constant- and mean-error irregularities, as compared
with the free-motion curves, as well as in actual errors; there are
distinct losses of volitional control for both classes of errors,
especially at or near the ends of the series.

There are cases of very low mean error to be found on all records,
where the value is 1/4 mm. or less, and, while the same phenomenon
is found with free motion, it is more marked here and occurs more
frequently; in most cases it seems as a drop from errors of larger
values rather than a gradual matter, as if the subject realized the
large error and exerted unusual volitional control to correct and
produce a very accurate record, but found that the attention needed was
beyond his will-power, as shown by the immediate lapse of accuracy.

There is an indirect confirmation of this view from the introspective
testimony of the subjects.

The visual element steadies but does not reduce mean errors when weight
is retarding.

The general shape of curve for eyes closed is downward 20-40 beats,
and rising for the rest of the series; it is less regular, but more
accurate than the visual results.

The fact that constant errors are mostly positive leads to a denial of
any inhibitory effect of the retarding weight.

For 1 cm. lines,

weight-retarding movements:

For the average of seven subjects we find:

(1) visual control does not improve accuracy or regularity as in free
movements;

(2) a retarding weight tends to make errors constant whatever the
speed-rate;

(3) the testimony goes to show that the free-movement records are more
accurate than the retardation ones.

Mean errors are:

(1) no more accurate and perhaps less regular, when the weight is
imposed;

(2) right- and left-handed subjects are equally accurate.

Constant errors:

(1) the more dexterous hand is superior for coördinations requiring
accuracy;

(2) ruled lines are slightly shortened in some cases;

(3) weight-records do not give more accurate results as compared with
free movements.

For individual records we find:

(1) retarding weights increase errors and irregularity;

(2) fatigue-points are more marked and frequent than for free movements.

Mean errors,

(1) the visual factor is of some value, but the testimony is varied;
right hand for increased regularity only, and left hand for greater
accuracy only;

(2) curves are horizontal or reducing with speed-increase.

Constant errors,

(1) the more dexterous hand coördinates better;

(2) all errors are positive;

(3) visual control helps only for regularity;

(4) curves are horizontal or rising.

With weight-accelerating movements, the average record shows a
sudden rise in mean error at both ends, not in evidence with free or
retardation results.

In general it is to be noted:

(1) that the visual element is of no value for reducing the error, and
of little value for promoting regularity;

(2) that the pull-records are closely comparable to the free-motion
records, and the accelerating influence of the weight is imperceptible;

(3) that the pull-records are more regular and closer to the
free-motion curves than are the weighted records, especially at the
ends of the left-hand curves.

For constant errors:

It is more in accord with the facts to say that the imposition of a
weight tends to reduce the constant error, and this is more marked when
the weight acts in pulling or to accelerate the motion.

Comparing with the weighted curve, we find the same general type of
rising curve, similarly located, and the same is true when compared
with the free-motion curve. Constant errors are reduced, but slightly,
and visual control is rendered nil, when the weight acts either to
accelerate or retard the movement, and of the two, the accelerating
effect is more marked, as reducing errors and promoting regularity.

There is no appreciable tendency for the weight to reduce the ruled
lines when retarding motion, nor is the weight as accelerating, able to
extend the line beyond the point set in the free motion.

When contrasting averages from right- and left-handed subjects it may
be said:

As compared with free motions there is a slight reduction of error and
irregularity more marked with the left-handed subjects, but a general
close correspondence of results.

The question is now appropriate, why should the right-handed men show a
reduced error for speed-increase, while the left-handed subjects show
the reverse? Bearing in mind that the right hand is the more dexterous
or better trained in the former case, it may be suggested that the
order of record from 20 toward 200 beats is such as to cause more
accurate results at the upper limit, in spite of the fact that less
time is allowed for discrimination and adjustments; on the other hand,
left-handed subjects have much less advantage of practice and habit in
their use of the right hand, and will show the predominance of error,
when the ruling is too rapid for careful discrimination.

It becomes a struggle between automatism, or semi-automatism, on the
one hand, and discriminative processes on the other.

Visual control is not an advantage in the case of accelerating weight,
and the large reduction in error with visual control for the free
movements is not evident with weighted motions.

For the left-handed subjects we find that the eyes-closed record shows
closer work than does the eyes-open curve; it is lower and nearer the
line of 0 error; in this respect, it shows the same effect as with
the free-motion curve, and to a less extent as for the retardation
weight-record. The accelerating record is, however, more accurate and
regular than either of the other curves.

It will be clear, then, as observed, that for constant errors, visual
control tends to reduce errors and steady records whatever the
speed-increase, as far as right-handed subjects are concerned, but this
effect is not noted for left-handed subjects using the right hand, and,
with their left hand, visual control is a disturbing element.

Further this erratic effect of visual control is less marked but clear
when the weight acts as a retarding factor, but is much more noticeable
for the free-motion record.

Individual records show few lapses of control for either errors.

The bulk of the evidence is that the weight imposition, whether acting
as a retarding or accelerating influence, is effective in rendering the
results more accurate and regular, though at least one subject exhibits
the opposite effect for the accelerating weight.

The left hand is better for J., Le., and W., but is less regular for
all subjects, save Le. and W., showing again a somewhat complex mass of
testimony, from which we may conclude that the right hand is the more
accurate member for right-handed subjects, and to a much less extent
the left hand is preferred by the left-handed subjects.

Visual control is to be noted as effective for accuracy and regularity,
except for Ha., where the curves closely intertwine, and for J., where
the eyes-closed record is much better.

Weight-accelerating movements:

For the average of seven subjects we find

visual control is of doubtful advantage, for left-handed subjects, but
shows a clearly marked reduction of error for right-handed subjects.

Mean errors are:

(1) similar in all respects to free-movement results;

(2) acceleration-curves are closer to free-movement results than are
retardation records;

(3) the more trained hand shows reducing error for speed-increase,
while the other hand shows increasing errors, because of superiority
of practice-effects over the native tendency to increase error as
speed-rate rises, for the more dexterous hand alone.

Constant errors:

(1) there is no tendency to overrule, as compared with free movements,
when weight acts to accelerate movements, for there are even cases of
lines being shortened with accelerating weights;

(2) a weight seems to negate the results of visual control, as a rule.

For individual records we find:

(1) fatigue-points, for the right hand only, are to be found in a few
cases;

(2) weight promotes regularity and accuracy;

(3) visual control is effective only for reducing variations of error;

(4) the better trained hand is the more accurate in the records, to a
slight extent;

(5) there are evidences of semi-hypnotic or dreamy states in the
non-visual series.


COMPOUND MOTIONS

Series of records were taken at 100 mm. and 10 mm. bases for the hands,
with what is called compound motion. This consisted in an additional
movement of the hand that was not ruling with the pencil, in a similar
manner, as regards the amplitude and general character of the motion,
but in an opposite direction.

For example, suppose the left hand is ruling a 100 mm. line outward,
or to the left; coincident with this movement would be a similar motion
of the right hand outward or to the right. The origin of both motions,
or the starting-ends of actual and imaginary ruled lines, was optional,
it being desired to bring out the effect of such additional motion, as
little complicated as possible with restrictions, as to its position
or extent. Actually this distance varied from about 10 mm. where both
motions were outward to 600 mm. for inward motions.

A comparison of such compound motions with single-hand records shows in
general the following:

For 10 cm. lines:

The case for mean errors may be summed up by saying:

(1) left-hand records are less accurate and regular than the right-hand
curves;

(2) visual control reduces error and irregularity in all cases, but is
more marked with the left hand;

(3) errors increase with speed-increase;

(4) compound-motion records show little increase in error or
irregularity, as compared with the simple motions.

For constant errors:

No marked peculiarities are to be noted, but in general,

(1) left-hand records are less accurate and regular;

(2) visual control reduces errors and irregularity;

(3) errors reduce with increase of speed, except for compound motion
uncontrolled visually;

(4) compound-motion errors are not much greater, nor is the
irregularity increased.


INDIVIDUAL CASES

100 mm. hand with compound motions.

A glance at the charts shows for individual records a few examples of
inhibition of voluntary control for both constant and mean errors, it
being much more marked in the case of mean errors.

These lapses of control appear for constant errors for A. with l.h.e.c.
at 160 and 180 beats, for mean errors for G. with l.h.e.c. at 200
and with r.h.e.c. at 180 beats; for Le. with l.h.e.c. at 50 and with
r.h.e.c. at 160 beats; for A. with l.h.e.c. at 200 beats; thus giving
evidence that the visual element has a steadying effect, and that the
left hand is less reliable save for Le.

There seems reason for contending that the compound motion can be
carried out, as arranged, without loss of accuracy or regularity on
the part of the ruling hand, and further that the subjects are pretty
generally apt to react to a given stimulus within certain rather narrow
limits of accuracy.

The evidence is here pretty conclusive that the right-handed subjects,
as a whole, show greater accuracy by about 25% for the more dexterous
hand; but it will be wise to consider the individual cases on this
point.

Greater regularity and accuracy for the right hand is attained by all
right-handed subjects, while the preference of Le. for the left hand is
clear but much less definite.

For individual cases:

The evidence again is fairly well marked that the more practised hand
will give a better account of itself even when visual control is not
called on.

The results for compound movements of the hand for 1 and 10 cm. lines
are summarized as follows:

It should be kept in mind that the compound records were in all cases
taken in connection with a duplicate series of lines for one hand, and
called simple movements. These simple movements correspond with the
free-movement records that have been considered already.

The purpose has been to bring out the modification of results that a
compound movement introduces, rather than to bear heavily on intrinsic
phenomena, _i. e._, comparison is deemed more important.

For lines 10 cm. long:

Average of seven subjects:

We find for mean and constant errors:

(1) left-hand records are less accurate and uniform;

(2) visual control increases accuracy and regularity, especially for
the left hand;

(3) there is an increase as the speed increases for mean errors, and a
decrease for constant errors;

(4) compound movements are practically as accurate and regular as the
simple ones for constant errors.

For individual cases:

(1) a few lapses of control or fatigue-spots more marked for mean
errors with the l.h.e.c., for the visual sense steadies ruling, and the
left hand is less reliable;

(2) compound and simple records show close agreement;

(3) the more trained hand reacts more accurately, and with greater
regularity;

(4) non-visual records show a cautionary shortening of line at low
speeds, and another at the upper limit, the latter being due to
physiological limitations.

For 1 cm. lines:

As far as averages are considered:

We may say, then, for mean errors:

(1) that visual control is effective for reducing errors, and
increasing steadiness in both sets of records, being more marked with
the less trained hand, the left;

(2) only in the case of the l.h.e.c. curves is the movement of the free
hand noted as appreciably affecting the accuracy or steadiness of the
record.

(3) eyes-closed records in general show a considerably greater error at
20 beats that practice rapidly reduces up to 40 to 60 beats.

We may note for constant errors:

(1) all errors are positive and confirm the earlier deductions on this
point;

(2) visual control reduces error and improves steadiness of record;

(3) the free hand-movement does not affect either the accuracy or
uniformity of results;

(4) errors do not increase with speed.

For individual records:

Comparing the individual cases of simple and compound movement, there
is no particular reason for concluding that the compound movement
is a disturbing influence as far as the records of all subjects are
concerned, save possibly the lapse of G. at 20 beats, and on the other
hand a case of greater accuracy and evenness for compound movements for
Mo. with r.h.e.c. constant error.

It is, then, possible to extend the conclusion of the 10 cm. records,
and say that both lengths of lines are ruled with a fairly constant
limit of error, whether the movement be simple or complicated by
movement of the free hand.

Individually there is testimony in favor of the gain in accuracy with
visual control for Hu., Hy., and Le., while the crossing of curves for
the other subjects shows that there is no difference in eyes-open and
eyes-closed results, the general conclusion being in favor of the value
of the eyes for accurate results.

For mean errors the right hand is more efficient in the case of A.,
Hu., Hy., and Me., while the reverse is the case for the rest, and the
evidence goes to suggest that greater accuracy can be attained with the
more practised hand.

For constant errors the right hand is more accurate in the case of A.,
G., Hy., only; Me. and Mo. are equally accurate with the hands, and the
rest show a marked preference for the left hand, the evidence being
thus conflicting, pointing to the theory of ambidextrous development on
the lines of accuracy.

L.h.e.c. records are horizontal for all subjects except Hu., Me., and
Mo., who show an upward slope to the curve. Evidences of visual control
as giving greater accuracy are noted in general above 70 beats and
individually for Hy. and Mo., only, the remaining records being so
intertwined that no difference can be noted, all suggesting that the
eyes are of but little assistance when the left hand is considered. The
right hand is preferred with eyes closed by A., Hu., and Le., while
four right-handed subjects testify that the less trained hand is more
accurate.

The testimony here seems conclusive as pointing to a denial of the
current notion as to the greater accuracy of the right hand for
right-handed subjects, and of the left hand for left-handed subjects,
and further suggests that visual control is a large factor in the
supposed superior excellence of the hand mentioned.


SUMMARIZING

For lines 1 cm. long:

Average of seven subjects:

It may be said that:

(1) visual control reduces both mean and constant errors, especially
for left hand;

(2) errors are constant whatever the speed;

(3) constant errors are positive showing overruling in all cases;

(4) there is no disturbance created by the second-hand movement, save
for r.h.e.c. mean errors, where the accuracy is less for the compound
records; this is probably due to the fact that this record shows the
least evidence of voluntary control, and is thus most subject to
disturbances;

(5) there is a marked reduction of mean error 20-50 beats, probably due
to practice.

For individual cases note:

(1) fatigue-spots for non-visual mean errors only;

(2) the equality of result of both types of movements is noted for all
cases;

(3) the non-visual right-hand records for some subjects are more
accurate;

(4) the more trained hand is not, as a rule and subject to exceptions,
the more accurate one, especially for the non-visual records; and

(5) there is evidence that the superior accuracy of the right hand for
right-handed subjects is largely a matter of visual control.


HEAD-RECORDS

There was no attempt made to differentiate the visual element because
the very movements of the head prevent the full use of the eyes; as a
matter of fact, the subject's attempt to make use of the eyes and the
aid is more marked at slow speeds and upon facing the apparatus. It
is to be noted here that the visual element, as reducing the error at
low speeds, is equally marked whether the eyes are directed toward the
recording pencil or not. This raises an interesting question as to the
direction the eyes must take for the optimal result; must the eyes be
fixed on the moving pencil, on its immediate surroundings, or may they
wander at will about the surrounding objects?

My own introspective testimony, corroborated by others, who have acted
as subjects for this investigation, is that the eyes are most effective
when gathering spatial relations in a gross way, and it may be expected
that the effects of visual control as reducing errors will be equally
efficient, whether the recording pencil be screened or visible,
provided it be possible to bring on the retina objects that are grouped
about the centre of attraction, the pencil, but not in its immediate
neighborhood.

The records show that there is underruling at the higher speeds because
of physiological limitations; but this shortening is greater for the
backward movements, for the position of the subject is such as to lead
to greater uncertainty as to the exact length of ruled line, and it is
probable that a cautionary or inhibitory feeling is the cause of this
shortening beyond what will be clearly due to inability to perform the
desired movement.

Further, visual control is effective, in the case of constant errors,
in lengthening the ruled lines at high speeds, and thus reducing the
negative constant error.

While the muscular control of the head is a constant, whether the
movement be forward or backward, it is less effective for constant
error reduction when the head is moved backward. Consequently, while
the backward and forward curves are fairly well in correspondence,
there is some reason for offering the proposition that either the
eyes are of assistance in forward movements to reduce mean errors at
high speeds, and they are of no such value for backward movements, or
the muscular control of the platysma myoides, trapezius and associated
muscles of the neck group is more nearly perfect for movements of the
head forward than for backward motions, the latter being to my mind the
better hypothesis.

The results for head-movements for lines 1 and 10 cm. long are
summarized:

For lines of 10 cm. length:

Average of six subjects:

For mean errors:

(1) the curve for head-forward and head-backward closely corresponds
to l.h.e.c. record; the errors increase by 50% with increase of
speed-rate, suggesting that

(_a_) visual control is negligible, as far as seeing the moving pencil
is concerned;

(_b_) control of head for forward equals that for backward movements.

For constant errors:

(1) there is underruling at high speeds because of the usual
physiological limitations, and this is more marked for head backward
results, suggesting that

(_a_) spatial relations are obtained, when the apparatus is visible,
that tend to correct underruling, or

(_b_) an extra inhibitory effect, due to lack of knowledge of spatial
relations, is added to the normal physical shortening and the subject
moves the head a less distance than is naturally possible; or

(_c_) the muscular control is less complete for movements of the head
backward.

For individual cases we find:

(1) fatigue-lapses are less in magnitude than for the hands, because
the head-movement can be only a fraction of the forearm-movement;

(2) mean errors increase and constant errors decrease with speed-rise;

(3) similarity of individual head-forward and head-backward curves is
suggestive, taken with the fact that no typical form of curve is to be
found;

(4) head-backward constant errors are greater and less regular in all
cases, suggesting that the eyes, in head-forward records, by getting
spatial relations, are more efficient.

For lines 1 cm. long:

Average of six subjects:

For mean errors note:

(1) the head-backward records are less regular than the head-forward
ones, and rise a little with speed-increase, showing visual assistance
for accuracy or better muscular control for the forward movements or
both;

(2) the constant errors show shortening of ruled lines at high speeds a
little more marked for the head-forward results;

(3) there is constant overruling.

Individual cases suggest:

(1) fatigue-spots are apparent, especially for head-backward movements;

(2) errors do not increase with speed;

(3) the movements of the head forward are under better control.


FOOT-RECORDS

10 cm. records show that

(1) the eyes are of no assistance as to increasing accuracy but help in
promoting regularity of error;

(2) a shortening of ruled lines with speed-increase is noticeable, and
is probably due to the usual physiological reason;

(3) the feet are capable of less accurate motion than the hands, but
show better results than the head;

(4) mean errors increase but constant errors decrease with
speed-increase.

Individual records show:

(1) less violent fluctuations of errors in all respects than do the
results of head or hands, for vertical foot-movements are of less
extreme extent than are arm- or head-motions;

(2) that for visual control with mean errors, no foot is the more
accurate, and there is no reason to believe that the feet are unequally
educated.

1 cm. records show, as far as mean errors are concerned, that:

(1) visual control is of no value as either reducing actual errors or
as effecting greater regularity;

(2) Errors for foot-movements are no less, but considerably more
regular than for head-motions;

(3) errors for hand-movements are more regular, and only 50% of the
results for either head- or foot-movements;

(4) all curves are horizontal;

(5) there is no appreciable advantage as to accuracy or regularity that
can be attributed to either foot. The evidence goes to show that the
subjects are ambipedalous, if it be permitted to coin such a word.

In general, we find that, as far as constant errors are concerned,

(1) visual control does not help to reduce actual errors or promote
uniformity;

(2) errors for foot-movements are less than the head records, and but
little greater than the hand results, while the regularity for the feet
is comparable to the hand, and much greater than for the head;

(3) all curves are horizontal;

(4) there is no particular advantage that either foot has over the
other either as to accuracy or regularity.

The evidence is that the subjects were ambipedalous, as far as ability
to reach a certain point equally well by either foot is concerned.
The popular notion has been to the contrary, and it is a point of
considerable importance to note the last point.

For example, in kicking, as developed by football trainers, it is
commonly assumed that the right foot for right-handed subjects should
be developed, and the opposite foot for left-handed men. Or again, in
the case of a person lost in the woods and walking in a circle, it is
observed that right-handed persons will turn to the left; probably
because of the pace of the right foot being slightly longer than the
left. My reply to this evidence will be that the data herein presented
is for vertical movements of the foot, starting from the floor in every
case, the subject being seated in a chair.

On the other hand, it is an entirely different movement, calling for a
much different and greater muscular control in the case of kicking or
walking that must be considered. For this reason the evidence, while
conclusive within its range, is not offered as more than suggesting
that the feet are equally well trained for the usual adjustments, and
only an exhaustive investigation covering all possible foot-movements
will settle the question.

The result for foot-movements for lines 1 and 10 cm. long is here
summarized.

For lines 10 cm. in length:

Average of seven subjects:

Note in general that

(1) l.f.e.c. mean error is most erratic, while the same curve is the
most accurate, as far as constant errors are concerned;

(2) the left foot mean and constant errors are slightly greater than
those for the right foot, for visual records;

(3) mean errors increase and constant errors reduce with speed-increase;

(4) the visual sense improves regularity, but does not reduce errors;

(5) there is a physiological reason for the shortening of lines at high
speeds;

(6) the feet are more under control than the head, but less than the
hands.

For individual records:

(1) fatigue-lapses, all for non-visual, are less numerous and of less
magnitude than for the hands and head, for the vertical movement of
foot is likely to be of less extent than that of head and hands for the
particular motion required here;

(2) there is no foot capable of being called more accurate than its
mate;

(3) the eyes appear to be of no value for reducing or regulating errors
for foot-movements.

For lines 1 cm. long:

Average of six subjects:

It may be said in general that

(1) the visual sense is valueless for promoting accuracy or regularity
of curve;

(2) errors of foot-movements are more regular and, for the constant
errors, more accurate than for the head-records;

(3) errors of foot-movements are less and less regular by 50% as
compared with records for the hands;

(4) errors do not increase or decrease with speed-changes;

(5) the feet are equally accurate.

For individual results:

(1) fatigue-lapses and cases of large error-increases are noted in a
number of subjects, both for visual and non-visual records;

(2) further, evidence is available as to the indifference to visual
control;

(3) no preference for either foot is to be discovered.

The results for individual choice of rhythm.

In this series of records, the metronome was dispensed with, and
the subject was permitted to react as he desired, taking the speed
preferred because of ease, pleasure, or other reason.

Records were obtained for six subjects for feet, head, and hands, both
single-hand and double-hand movements, all for lengths of line 1 and 10
cm. The charts for individual choice were plotted for a comparison of
speeds rather than for accuracy.

It was noted for the hands:

(1) that every subject reacts more rapidly with the left hand;

(2) the eyes had little effect as to changing the speed-rate;

(3) single and double hand-movements were equally rapid.

Some subjects, as A., react more rapidly for the shorter lines, though
no clearly marked evidence of this speed-increase is to be noted.

For the head, the results for both eyes opened and closed show the
impossibility of separating the optimal or preferred rate of speed on
the score of visual assistance or because of direction of head-movement.

There is a close agreement of the subject as to his best speed, and
this is independent of special conditions; for example,

A. selects 50-57 beats per minute for 1 cm. and 48-68 for 10 cm.; G.
has a preference for 61-66 and 56-71; Hu. rises to 103-125 for 10 cm.
and selects 68-82 for 1 cm.; Le. 52-55 for 1 cm. and 45-52 for 10 cm.,
and so on.

We may say, then, that free rate-choice for head-movements results in
a selection of some rate of speed that is not affected by the visual
sense or direction of movement, and is strictly individualistic,
covering a range of 50-130 beats per minute, and not increasing as the
amplitude of movement is reduced.

Turning to individual choice of speed-rate, for the feet it will be
seen that

(1) the non-visual records closely correspond as to chosen speed, and
there is a less close correspondence of visual speeds;

(2) the visual records are ruled at a lower rate in some cases, but A.,
G., and Mo. show little difference;

(3) there is a tendency to speed up as the series progresses;

(4) the shorter lines are ruled with greater speed as a rule, though G.
and Le. fail to show this phenomenon;

(5) The left-foot records show a higher speed-rate for all cases.

Among many interesting points that cannot be examined in this
connection, such as relation of voluntary choice of rate to the main
line of metronome records as regards accuracy, the fact of the higher
rate of ruling for the left hand and foot stands most prominent.

Whether a record of head-movements to right or left, or other devices
to compare the sides of the body or to contrast arm and leg speeds,
will bear out this testimony is as yet unknown, so that the writer
prefers to announce the result and not now fit theory to data. It may
be said that the records were taken in reverse order and rearranged,
as regards right and left foot or hand, and, in addition, the initial
foot-movement varied with the subject, some being right and some left.

We ask finally: Is the time in which the greatest exactitude is
produced, the same for every group of muscles; that is, has every
motor apparatus the same natural rhythm? and: Is this natural rhythm a
constant rapidity for all motor nerve-centres or does it depend upon
the complexity and character of the movement?

The comparison will fall first on the averages and finally on the
individual records.

The hand-movements show the following results:

Constant errors for 14 cm.:

For simple and weight accelerating and retarding motions, there is a
close agreement about 120 beats for the minimum error for visual and
right-hand non-visual records; left-hand non-visual records are spread
more, but will also average the same.

For 10 cm. simple and compound movements the visual minimum errors are
at 180-200 beats, while with the eyes closed the results are grouped
about 60 beats; one record, that for l.h.e.c., has two minimum points
at 60 and 180 beats, the latter being clearly a crossing of the 0
error-line, because of physiological limitations.

For 1 cm. simple and weighted minimum errors are grouped between 20 and
60 beats, while the simple and compound group show less regularity and
a tendency to group minimum errors at 100 beats.

The head-movements show for both 1 and 10 cm. lines a minimum error at
180-200 beats, there being, however, one exception at 100 beats for 10
cm. head-backward movements.

The foot-movements show minimum errors at 80 beats for the right
foot, and 180 and 100 beats for the left foot, visual and non-visual
respectively.

Bearing in mind for a moment the individual choice records, there seems
here a suggestion that the left foot is capable not merely of higher
speeds, but of minimum errors at the higher rates as compared with the
right foot.

No such differentiation of the hands can be discovered, however.

Mean errors:

For the hands:

For 14 cm. for simple and weighted results we find that the right-hand
and left-hand eyes-open minimum errors are at 180 beats, but the
non-visual left-hand minimums are at 30 beats.

For 10 cm. simple and compound records we find all minimum errors are
between 160 and 200 beats.

For 1 cm. simple and weighted results there is a scattering of minimum
errors from 20 to 200 beats, with a heavy preponderance at 200, and the
same is true for the simple-compound series.

The head-movements minimum errors are at 40 beats without exception.

The foot minimum errors are distributed from 20-30 beats for the left
foot to 160-180 for the right.

It is thus evident that each group of muscles and each motor centre has
its own optimum, and that the conditions of complexity, resistance,
etc., influence greatly the accuracy of the periodic movement impulse.

FOOTNOTE:

[Footnote 138: Woodworth: The Accuracy of Voluntary Movements,
Psychological Review Monographs, no. 13, 1899.]




THE MOTOR POWER OF COMPLEXITY

BY C. L. VAUGHAN


A. COUNTING OF SIMPLE AND COMPLEX VISUAL OBJECTS

If every sensory stimulus has a motor reaction, then a simple figure
perceived in any way ought to produce a somewhat different response
from a more complex figure similarly perceived. Of course if only one
figure of each kind is given it is difficult to measure in any way
this difference, since it is so small. But we might make it measurable
by multiplying the process. Therefore I have cut out a row of similar
figures in a strip of cardboard and on another strip another series of
a different pattern. Now if these rows are counted figure by figure
each figure has a certain motor effect which influences the speed of
counting, so that the time of counting (measured by the chronoscope)
should give some indication of the comparative motor power of the
figures in question.

In the accompanying illustration nine cards of various patterns are
shown. Cards 1, 2, and 3 are comparatively simple patterns while 4, 5,
and 6 are comparatively complex, Card 6 having the added complication
of different kinds of figures on the same card. Cards 7, 8, and 9 form
another group, Card 7 having the same letter throughout, Card 8 having
letters composing a sentence and Card 9 a series of the letters, mostly
consonants, mixed promiscuously. In order to prevent the subject from
knowing the exact number, and thus, perhaps, bring in another influence
at the end of the row, most of the different cards have different
numbers of figures, but this difference is not great and some cards
have the same number. The subject usually forgets, from one experiment
to the next, the number on each card.

At first the experiment was performed with the figures in a straight
row, instead of in the broken line which is seen in the illustration.
In counting the straight rows, the observers found it hard to keep the
place in the line. A subject would become confused and count some spot
twice or else he would omit it altogether. Furthermore this disturbance
was found to be much greater with some figures than with others, with
Card 1, for example, more than with Card 2. Therefore the device was
adopted of diversifying the line, both by placing some of the figures
above and some below the line and by making the distances from one
figure to the next, different in the different cases. And in order to
prevent the subject from associating any peculiar turn in the line
with a certain number counted, it was decided to have the arrangement
on the different cards different. But it was still necessary to have
the intervals between figures about the same in all the cards, and
therefore the row was divided into sections of six figures each and
these sections were used as units, variously arranged, in constructing
the other rows. For example the first unit of Card 3 is the same as the
second of Card 4. Sometimes this six-figure unit is turned end for end
or upside down, and thus, though the same spaces are used, the cards
appear dissimilar.

[Illustration: FIG. 1]

The subject would be seated at the table with one hand resting lightly
on the key which sets the chronoscope in motion, his eyes raised so
that the table in front of him is not seen. One of the cards would
then be put in the proper position in front of him (always the same),
and he is told that all is ready. He looks down at the card and as
soon as he begins to count the first figures in the line he presses
the chronoscope key, and when he has reached the end of the line he
releases the key. The time for the operation is then noted. The whole
series of cards is thus gone through. An extra card of which no record
is taken is used for the first few tests so that the subject may be in
the proper state when the first test to be noted down is taken. Also
the order of the series is changed from one experiment to the next,
each card taking its turn at being first and last. It was hoped in this
way to distribute among the different cards the effects of practice and
fatigue, and also to guard against any expectations on the part of the
subject as to the character of the next card.

The subject is told to count as fast as he can, with a reasonable
feeling of certainty as to his correctness, the main object being to
have a uniform principle, in counting the different series. Wrong
counts were excluded, but later on the same cards given again so as to
keep the tables even. Subjects were not allowed to count the figures
by groups, but one by one. At first a certain amount of difficulty was
found in the fact that subjects in counting would repeat the numbers
to themselves, and as they seemed to be retarded by this, especially
in those numbers whose corresponding names have 2 or 3 syllables, the
result was that we were getting the speed with which subjects could
count the numbers from 1 up to 38 or 39 and this would be practically
the same whatever the figure. But all the subjects were finally trained
merely to think the number, or at least to have as little vocal
adjustment as possible. When this was done the subject no longer felt
that it was the speed with which he could count that was being measured
but the rate at which he could take in the different figures on the
card, one at a time.

Between three and four hundred tests were made of the counting of the
figures on the nine cards, the work being divided among seven subjects,
though not in exactly equal amounts. Since the number of figures on the
different cards are different, I have found the time it takes to count
one figure by dividing the total time by the number of figures on a
card. The following table shows the average time taken by each subject
for one figure on each card, time given in thousandths of seconds.
_A.M.V._ stands for average mean variation.

       _A._ _A.M.V._ _B._ _A.M.V._ _C._ _A.M.V._ _D._ _A.M.V._ _E._
    1 279.69 11.47  186.87 13.22  247.62 14.89  193.08 12.38  262.77
    2 270.60 12.47  180.55 11.88  249.21 18.00  190.51 11.82  257.56
    3 274.43  9.87  180.89 11.57  247.59 15.51  192.07  7.87  259.96
    4 286.82 12.47  190.39 12.56  255.20 16.78  200.53 10.72  267.11
    5 290.29 11.89  195.41 12.36  262.27 19.73  199.89  9.27  271.06
    6 293.06 12.21  185.33 11.51  275.40 18.13  199.20  7.92  264.59
    7 273.32 15.54  192.23 13.73  229.26 19.30  185.60  9.83  265.56
    8 279.77 13.86  185.23 12.69  246.66 18.86  193.39  9.81  277.52
    9 285.09 11.97  197 04 12.28  269.96 19.95  186.30  9.05  259.72

            _A.M.V._ _F._ _A.M.V._ _G._ _A.M.V._
             20.20  217.00 12.32  442.63 36.51
             16.03  195.00  9.50  431.00 24.39
             17.41  191.50 11.03  434.83 24.28
             20.31  226.40 29.11  445.71 14.58
             20.86  233.50 20.46  459.17 18.92
             15.00  220.80 14.92  432.17 26.87
             19.20  189.20 30.31  402.13 21.34
             14.93  220.70 21.79  388.77 27.48
             14.27  210.34 11.71  419.57 18.22

A, B, C, D, E, F, G are the different subjects, and 1, 2, 3, 4, etc.,
refer to the cards with the different patterns. It is seen at a glance
that great differences exist between the rates with which the different
subjects count. Subject G had much fewer tests than the others, and
thus, not having as much training, his average is higher in comparison
than it would be had he had the same training.

Now if we compare the counting of the first three or relatively simple
patterns with that of the next three or comparatively complex ones, we
notice at once that the simple figures are almost invariably counted
in less time than the complex, there being only two exceptions. B
counts 6 a little faster than 1, and G counts 6 faster than 1 and 3.
Even these apparent exceptions are easily explained. As noted already,
subjects are much more apt to lose their place in counting certain
cards than others. This is especially true of Card 1 even after the
line is broken. Now Card 6 is arranged on a different plan from the
others, for it has many kinds of figures on it. This is a great help
in keeping one's proper place in the counting of the series, and since
wavering between two figures is avoided, the series is counted more
rapidly. But B is the most rapid in counting, of all the subjects, and
it is natural that any differences in the ease of keeping place should
show themselves here, since the more rapid the counting the easier it
is to lose the proper position. This cannot be said of G, who is a
slow counter, but on the other hand it may be noted that he had only
a few cases, and at first the ability to keep one's position is much
less than after considerable experience. So in Cards 6 and 1 there are
two conflicting principles, degree of complexity and tendency toward
confusion of position. Of course both these principles are present in
all the other cards, but they reach a maximum in 1 and 6, in 1 extreme
simplicity with difficulty in keeping place, in 6 extreme complexity
with ease in keeping place. Card 1, it will be seen, is with nearly all
subjects a little slower than 2 and 3, while 6 is generally faster than
4 and 5.

Therefore it would seem that the apparently small exceptions are not
real exceptions, but variations due to the presence of other factors
than mere differences in complexity of the figures used. In observing
the averages for 7, 8, and 9 we see that as a rule 7 is fastest, 8
next, and 9 the slowest. The tables are not quite so regular as for the
cards just given. B and G count 8 faster than 7, and E counts 9 faster
than 7. The most of these cards have on them 36, 37, 38, or 39 figures.
Card 8 has 43 letters. The subjects report that the last three on this
card are counted much faster. They know, as soon as they reach 40, just
how many there are, and it is hard to keep from counting the rest in
a group. Otherwise they do not feel any difference in counting Cards
8 and 9. Arranging the letters in words does not affect the speed of
counting, so far as they can see, for in counting they do not notice
the words at all.

When we average the records of all the subjects giving equal weight
to each subject, though the number of tests may be different with the
different men, we get the following table. Time given in thousandths of
seconds.

    (1)  261.38
    (2)  253.49
    (3)  254.54
    (4)  267.46
    (5)  273.08
    (6)  267.22
    (7)  248.19
    (8)  256.01
    (9)  261.15

It is seen, from looking at this table, that all divergences from the
general rule have stopped. Cards 1, 2, and 3 each take less time than
any of the 4, 5, 6 group, and 7 is faster than 8 and 9. So the evidence
seems very strong that it takes longer to count complex than simple
figures. Should one object that the difference is extremely small, a
few thousandths of a second, and that thus a slight error in one test
might invalidate the result, we reply that the time which is given
is the time in which we count just one figure of the given pattern,
and that thus of course the difference between counting two different
figures must be very small. Moreover there has been a remarkable
agreement of the tests taken at different times. It is not a case of
finding 1, 2, and 3 counted faster one day and 4, 5, and 6 counted
faster the next, but 1, 2, and 3 are counted faster nearly every time.
Occasionally 1 will take longer than one of the 4, 5, 6 group. And
extremely seldom is there a case where the average of 1, 2, and 3 is
not less than that of 4, 5, and 6.

The experiment seems to have proven that it takes a longer time to
count a row of complex figures than a similar row of simple figures.
_The complex figure exercises a retarding effect upon the eye as it
sweeps along._ There is a greater amount of sensory stimulation,
consequently a greater amount of motor excitement. This motor
excitement does not act in harmony with the motor activity which impels
the eyes along, but has a somewhat antagonistic effect. The eye is
held more by the complex figure; it is a greater effort to withdraw
the gaze to look at the next figure. A certain interest, as we say,
on the psychological side tends to hold one to the figure looked at.
This interest is greater (other things being equal) the greater the
complexity of the figure. The nervous processes involved in counting,
though admittedly in very small degree, are thus inhibited by the
complexity of the figure and act more slowly.


B. REACTIONS TO SIMPLE AND COMPLEX OPTICAL IMPRESSIONS

Since the preceding experiments seem to show that reactions on optical
impressions are different according as the figures are more or less
complex, it would seem that we ought to be able to measure by graphic
methods the reactions to visual fields of varying grades of complexity
and in this way to demonstrate their different motor powers.

A Porter kymograph was used on which to register the reactions.
Resting on the top of the drum, and revolving with it, was a circular
band of white paper, upon which were pasted the different figures
to be observed. A screen was placed in front of the kymograph, thus
concealing the figures; but at their level was a little square window
in the screen, which, when the eye was placed in the proper position,
allowed the subject to see one of the figures but nothing more. A few
inches in front of this window was an eye-rest which kept the eye
properly placed. A tambour received the movement from the subject and
communicated it to a straw which made a scratch on the smoked paper
which covered the drum.

The figures used in this experiment form two series, one, composed of
geometrical figures, varying in complexity from a circle to a very
complex figure consisting of many overlapping squares, triangles, etc.,
and the other composed of colored figures varying in complexity from a
simple square of one color to a very complex mixture of various colors.
The area of the visual field is about the same in all cases,--an
inch square. The geometrical figures were formed of black lines on
a white background. The figures used are shown in the accompanying
illustrations.

The subject would be seated in front of the screen, his eye at the
eye-rest a few inches in front of the window in the screen, and the
forefinger of the right hand on the tambour, which is to the right of
and behind the screen, and thus not seen while the eye is at the rest.
Then as the drum revolves and brings a figure in front of the window,
the subject observes this figure carefully, and when it is all in the
field of vision he presses down with his forefinger, thus producing a
curve on the drum surface. _He tries to make the same finger-movement
every time_, whatever the figure at the window may be. But his
attention is not to be too much taken up with the making of the
movement, for he must be closely observing the figure. If he looks at
the figure until he observes its characteristics clearly and then turns
his attention from this to the finger-movement, it is evident that the
optical sensation would not have much effect upon the movement. The
movement must be performed while his interest in the figure is highest.
Now, after a little practice, any one can accustom himself to make a
certain definite movement in about the same way every time, and he can
then agree that he shall make this movement as a reaction to a given
stimulation. Then when the stimulus comes he makes the movement without
any longer thinking of the character of the movement. It has become, to
a certain extent, automatic and can look out for itself.

This is the state into which I have tried to get my subjects. Their
whole attention is to be taken up with the seeing of the figures in
the window, and to these figures they are to react as automatically as
possible. Thus, though finger-movements are usually voluntary, all the
capricious character of voluntary action will be removed here, and if
the stimulus is the same in all cases, the reaction tends to assume
the form of a uniform movement. There is, then, a chance to see the
influence of different optical stimuli upon this action.

[Illustration: FIG. 2]

Six different geometrical figures were seen at each revolution of the
drum and six reactions given by the subject. Between figures a white
surface would occupy the field of vision. The simple and complex
figures were distributed so that the subject never knew what kind of a
figure would come next. The purpose of the experiment was kept as much
as possible from the knowledge of the subjects; but some, knowing my
general problem, surmised quite correctly my main object here.

Ten revolutions were made at each sitting, thus causing the subject to
react ten times to each figure. Then a new drum paper was taken and the
case with the colored figures placed upon it. This had five colored
figures, and ten revolutions were made also in this case. Thus, in all,
in any one day, the subject would make one hundred and ten of these
finger-movements.

Since we have in all these experiments tried to find out in the
different figures merely differences in the _amount_ of the reaction,
and not differences in the character of the reaction, we shall keep up
this method here. Now a stronger reaction makes a higher curve, and
since the drum is all the while revolving, and since the higher the
curve, other things being equal, the longer it takes, the stronger
reaction will also make a wider curve. So it would seem that if we wish
to observe the differences in the amounts of reaction the most natural
course to pursue would be to measure the heights and widths of the
curves we have registered. This accordingly has been done.

In our discussion of these measurements let us, then, first, take up
the curve heights, and of these, those of the geometrical figures
which we call _U_, _V_, _W_, _X_, _Y_, _Z_. The height is measured
from a base-line [drawn by revolving the drum after the subject has
taken his finger from the tambour] to the highest point reached. These
measurements are taken from two hundred reactions to each figure,
divided among seven different subjects.

                            _Heights of Curves_
                   _U_    _V_    _W_    _X_    _Y_    _Z_
    _Subject_  A   6.83   6.68   6.59   6.55   6.63   6.79
               B   8.64   7.26   6.41   7.79   6.39   9.75
               C   6.67   6.55   6.73   6.85   5.87   8.53
               D  21.35  21.26  21.46  21.90  21.33  21.31
               E  16.13  15.77  15.17  15.85  15.29  16.08
               F  16.90  16.97  16.14  16.52  15.81  17.91
               G  11.42  11.32  11.39  11.48  11.06  11.10

                  87.94  85.51  83.89  86.94  82.38  91.48

    _Average_     12.56  12.26  11.98  12.42  11.77  13.07

                   _Arranged in order of height of curve_
                   _Z_    _U_    _X_    _V_    _W_    _Y_
                  13.07  12.56  12.42  12.26  11.98  11.77

If we put the figures in the order of strongest reaction for the
different subjects we get the following table:

    _Subject_    A  _U_  _Z_  _V_  _Y_  _W_  _X_
                 B  _Z_  _U_  _X_  _V_  _W_  _Y_
                 C  _Z_  _X_  _W_  _U_  _V_  _Y_
                 D  _X_  _W_  _U_  _Y_  _Z_  _V_
                 E  _U_  _Z_  _X_  _V_  _Y_  _W_
                 F  _Z_  _V_  _U_  _X_  _W_  _Y_
                 G  _X_  _U_  _W_  _V_  _Z_  _Y_

It is seen from these results that, although the subjects differ, the
height of the curve varies directly with the complexity of the figure.
The order of the figures, which we get by measuring the height of the
curves and then putting that figure with the highest curve first, with
the next highest second, and so on, is exactly the same order in which
we should put them if we were asked to put the most complex first,
the next second, and so on. Though the individual subjects may vary
somewhat from this rule, when they are all grouped together there are
no exceptions.

The variations of the reactions with the different subjects may be
shown very clearly in the following way, where the different figures
are in the left-hand side arranged in order of descending complexity.
"1st place," etc., refer to the order of arrangement of the figures
by the different subjects as shown in preceding tables. Thus, _Z_,
3 times, 1st place, means that three subjects have in the average a
higher curve for _Z_ than for any other figure.

      1_st place_ 2_d place_ 3_d place_ 4_th place_ 5_th place_ 6_th place_

    _Z_  3 times   2 times    0 times    0 times     2 times     0 times
    _U_  2 times   2 times    2 times    1 time      0 times     0 times
    _X_  2 times   1 time     2 times    1 time      0 times     1 time
    _V_  0 times   1 time     1 time     3 times     1 time      1 time
    _W_  0 times   1 time     2 times    0 times     3 times     1 time
    _Y_  0 times   0 times    0 times    2 times     1 time      4 times

One can see at a glance from this, how, as the figures decrease in
complexity, they take their position further on in the series. If a
diagonal is drawn from the upper left-hand corner to the lower right,
it will pass through or near the larger numbers in the table, thus
showing that the figures belong in the ordered series in the places
already shown.

Next in order let us take up the measurements of the widths of curves
for the same geometrical figures which we have been considering.

                           _Widths of Curves in mm._

                     _U_    _V_    _W_    _X_    _Y_    _Z_
    _Subject_    A  20.83  20.59  20.93  21.22  20.21  21.89
                 B  11.18  10.77  10.46  10.31   9.92  10.79
                 C   4.28   4.43   4.10   3.78   4.95   4.70
                 D  21.08  19.36  18.33  18.75  18.17  21.09
                 E  14.22  13.85  13.40  13.56  11.96  14.13
                 F  17.00  15.26  15.92  16.52  14.52  16.47
                 G   5.25   5.19   5.30   5.08   5.11   5.37

                    93.84  89.45  88.44  89.22  84.84  94.44

    _Average_       13.40  12.78  12.63  12.75  12.12  13.49

    _Order_          _Z_    _U_    _V_    _X_    _W_    _Y_
                    13.49  13.40  12.78  12.75  12.63  12.12

If as before we take the orders for the different subjects, we get the
following table:

    _Subject_    A  _Z_  _X_  _W_  _U_  _V_  _Y_
                 B  _U_  _Z_  _V_  _W_  _X_  _Y_
                 C  _Y_  _Z_  _V_  _U_  _W_  _X_
                 D  _Z_  _U_  _V_  _X_  _W_  _Y_
                 E  _U_  _Z_  _V_  _X_  _W_  _Y_
                 F  _U_  _X_  _Z_  _W_  _V_  _Y_
                 G  _Z_  _W_  _U_  _V_  _Y_  _X_

Here, as before, in the case of the heights, it is seen that though
the order is different with the different subjects, yet the general
tendency is to place the most complex figures first and the simplest
last. The most simple figure _Y_ never comes in front of the fifth
place except with subject C, who places it first. This exception may be
ascribed to the fact that this subject, on account of his going away,
did not have so many tests. In fact only one day's work of 10 reactions
for each figure is recorded, and it is but natural that some variations
from the standard should occur in his case.

If now, as before, we investigate where each figure occurs in the
series for the different subjects we get the following table:

                                  _Times in_
     1_st place_ 2_d place_ 3_d place_ 4_th place_ 5_th place_ 6_th place_
    Z     3           3          1          0           0            0
    U     3           1          1          2           0            0
    V     0           0          4          1           2            0
    X     0           2          0          2           1            2
    W     0           1          1          2           3            0
    Y     1           0          0          0           1            5

Here we again see the large numbers on a line from the upper left-hand
to the lower right-hand corner.

Thus we get the following order from the geometrical figures as
measured by the height and width of the curves:

    _Height_     _Z_    _U_    _X_    _V_    _W_   _Y_
    _Width_      _Z_    _U_    _V_    _X_    _W_   _Y_

The only difference, it is seen, is that the positions of _V_ and _X_
are reversed in the two series. Such a change would on our principle
be fairly likely to occur, since _V_ and _X_ are figures near to each
other in complexity and the motor effects are very similar.

[Illustration: FIG. 3]

In the same manner, the following tables show the reactions to the
colored figures of different grades of complexity. And first, as
before, is the table of the heights of the curves for the different
subjects, given in millimetres. The numbers given represent the
averages of all reactions made. We will call the figures, for the sake
of reference, _L_, _M_, _N_, _O_, _P_.

                   _L_       _M_       _N_       _O_       _P_
    _Subject_  A   5.75      6.01      5.90      5.82      5.74
               B   6.72      5.56      6.35      7.53      4.94
               C  10.92     10.90     10.76     10.52     10.99
               D  25.49     25.42     26.23     25.89     25.52
               E  20.63     20.82     20.37     20.55     20.30
               F  15.67     15.23     15.15     15.98     14.51

                  85.18     83.94     85.26     86.29     82.00

    _Average_     14.20     13.99     14.21     14.38     13.67

Order arranged as before in a descending series according to height of
curve:

     _O_      _N_      _L_      _M_      _P_
    14.38    14.21    14.20    13.99    13.67

This is exactly, as I should judge, the order of the complexity of the
figures reacted to.

The arrangement by the individual subjects is as follows:

    _Subject_ A   _M_  _N_  _O_  _L_  _P_
                   B   _O_  _N_  _L_  _M_  _P_
                   C   _P_  _L_  _M_  _N_  _O_
                   D   _N_  _O_  _P_  _L_  _M_
                   E   _M_  _L_  _M_  _N_  _P_
                   F   _O_  _L_  _M_  _N_  _P_

We see that individual differences are stronger here than in the
geometrical figures, but that the same tendency to react more strongly
to the complex is present in nearly every case. This can be brought
to the eye more clearly if we observe the table in which is shown
the position of the different figures in the series of the different
subjects.

                                 _Times in_
        1_st place_  2_d place_  3_d place_  4_th place_  5_th place_
    O        2           1           2            0            1
    N        1           2           0            3            0
    L        0           3           1            2            0
    M        2           0           2            1            1
    P        1           0           1            0            4

_M_ here presents the principal exception, coming too often in the
first place.

Finally we give the tables for the widths of the curves for the colored
figures; and first the table of the averages of all the subjects for
all the figures:

                    _L_      _M_      _N_      _O_      _P_
    _Subject_ A    25.76    24.27    25.06    24.77    23.14
              B     9.42     9.49     9.06     9.84     8.11
              C     5.85     5.35     5.62     5.80     5.24
              D    13.68    13.53    13.18    13.26    13.38
              E    22.06    21.37    22.50    22.17    20.44
              F    16.30    15.08    16.65    16.76    15.13

                   93.07    89.09    92.07    92.60    85.44

    _Average_      15.51    14.85    15.35    15.43    14.24

Order, arranged in a descending series according to width of curve:

     _L_      _O_      _N_      _M_      _P_
    15.51    15.43    15.35    14.85    14.24

Here the order is not just the same as we got from a measurement of the
heights. The three complex figures have changed places somewhat, but
there is no exchange of a simple and a complex.

The arrangements by the individual subjects are as follows:

    _Subject_ A   _L_  _N_  _O_  _M_  _P_
            B   _O_  _M_  _L_  _N_  _P_
            C   _L_  _O_  _N_  _M_  _P_
            D   _L_  _M_  _P_  _O_  _N_
            E   _N_  _O_  _L_  _M_  _P_
            F   _O_  _N_  _L_  _P_  _M_

The three complex figures have different places with different
subjects, but very seldom is a simple figure found among the complex,
or _vice versa_.

This can be seen easily from the following table:

                                _Times in_
        1_st place_  2_d place_  3_d place_  4_th place_  5_th place_
    _L_      3           0           3            0            0
    _O_      2           2           1            1            0
    _N_      1           2           1            1            1
    _M_      0           2           0            3            1
    _P_      0           0           1            1            4

A theoretical word may close our report.

The growth of biology and physiology has tended to show that there
is no break in the nervous mechanism. The stimulus goes to the brain
and out through motor channels to muscles, glands, etc. The nervous
current does not wait in the brain for the permission of the mind to
leave on its journey to a muscle nor does it need mental reënforcement.
The nervous current as a whole is a unity. The nervous system is a
physiological instrument for producing the appropriate reaction to
a certain stimulus. In the unicellular organism there is no nervous
system, but the protoplasm receives the stimulus and produces the
reaction. As we go up in the animal series a differentiation is seen
to be present in the organism. Some parts are more concerned with the
receiving of stimuli and others with the approach toward or withdrawal
from the stimulating object. There is a division of labor. The nervous
system is developed as a means of rapid communication between the
different parts, but this communication is a physiological one. The
stimulus sets up a chemical action in the sensory organ which is
transmitted along the nervous path to the motor organ which is caused
to react. As we ascend the animal series the differentiation becomes
greater and greater, and consequently the means of communication must
become more and more complex. So trunk lines are formed which lead to a
centre, and from this centre again go out main lines which divide and
subdivide until the muscles are reached. The centre acts as a kind of
automatic switch-board.

Accepting such a view of the nervous system it must be granted that
different stimulations would produce different reactions. It was my
aim in the experimental work which has been described to show that
this is true. And while much work has already been done in showing
that different kinds, or different amounts of stimulation produce
differences in reactions, it seemed important to demonstrate also
that mere differences in the complexity of the stimulus bring about
differences in the reaction. So the experiment of counting figures of
different complexity was entered upon, and we found that it took longer
to count figures the more complex they were in spite of the fact that
the act of counting seems always the same. The question is how must the
fact that counting becomes slower and slower, as the figures become
more complex, be interpreted?

When we count a row of figures, the eyes do not move along at a regular
uniform rate, but make a quick jump from one figure to the next, halt
a moment, make another jump, and so on. Now, I think the principal
difference comes in with the figures of different complexity in the
time the eye halts at each figure. The halt is longer the more complex
the figure is. It is well known that any visual object which stimulates
the retina is brought by a reflex movement of the eye to the place of
clearest vision. Of two objects stimulating the eye at the same time,
the more pronounced one will produce the reflex and will hold the eye
longer than a weaker stimulus. Similarly here, the more complex figure
produces a stronger reflex and holds the eye longer than the simple
figure. This is repeated at every figure in the series.

The complex figures have more features about them, all of which by way
of the retina and optic nerve are represented in the cortex and thus
more cortical cells are involved, which in turn produce a stronger
stimulation of the muscles which move the eye in the proper way to see
the figure, and thus the eye is held more strongly by the complex than
by the simple figure.

Again in the second experiment, the subject reacts more strongly to
the complex as shown already in explaining the first experiment and
for the same reasons. It might be said that in looking at the colored
figures, _e. g._, that since the same amount of retina is stimulated,
the reaction ought to be the same. But we may presume that the complex
figure, on account of the different shapes and contrasts on its
surface, will more variously affect the same amount of retina and that
the nervous currents sent to the cortex will, many of them, be stronger
than those from the simple figure and will thus cause the cortical
cells to be more strongly excited, or by a process of irradiation the
stimulation will spread to adjoining cells and thus finally more cells
be stimulated. However this may be, the amount of discharge into motor
cells is certainly greater and the muscular reaction, therefore, also
greater.

The interesting side of our results is thus given in the fact that
we have here two activities--counting with highest speed and making
hand-movements of certain length--which are performed every time with
exactly the same intention and with the subjective impression of
equal result, and which yet show marked differences according to the
complexity of the psycho-physical stimuli. It is a new contribution to
our knowledge of the independent motor power of ideas.




ANIMAL PSYCHOLOGY




THE MUTUAL RELATIONS OF STIMULI IN THE FROG RANA CLAMATA DAUDIN[139]

BY ROBERT M. YERKES


I. ANIMAL BEHAVIOR AND THE SENSES

Since the behavior of an animal is conditioned by its senses, it is
extremely important that the comparative psychologist should have
accurate and detailed knowledge of the sense-impressions received by
his subjects. Knowledge of the so-called "special senses" does not
suffice for the satisfactory description of behavior, for there are
several other kinds of sense-data of equal or even greater importance
than those of the five special senses. As investigation of the subject
progresses the banefulness of the notion that all sense-experience is
summed up in "the five special senses" becomes more and more evident.
For the comparative psychologist the senses are not five, six,
eight, or ten, but as numerous as are the kinds of sense-data which
condition animal activities. There can be no doubt that many of the
lower animals are largely dependent upon senses which are not included
in the conventional special sense list. The contention that certain
organs which are commonly recognized as sensory in function, the cristæ
acusticæ of the ear, for instance, are merely reflex control organs,
has little weight in this connection, for every sense-organ is part
of a motor control mechanism, and so far as we are able to judge from
available evidence each has as an accompaniment of its functioning
a mode of sensation. If there are two kinds of peripheral organs in
connection with the afferent nerves, namely, those whose functioning
has sensation for an accompaniment and those in which motor control
is the sole phenomenon, it is high time that the fact were definitely
known.

Even a thoroughly accurate knowledge of the general condition of
the senses in a particular phylum, genus, or species may be of
trifling value in the study of the behavior of a given individual,
for within these groups the state of development and relative
importance of a sense may differ strikingly. Herrick,[140] in his
admirable investigation of the sense of taste in fishes, has rendered
comparative psychology an important service by showing that even a
highly developed sense may be of markedly different value in the
associative life of different species. The cat-fish, according to
Professor Herrick's observations, obtains its food primarily by the aid
of taste-impressions, the hake by the aid of touch, and the sea-robin
chiefly by means of vision. All three of the senses mentioned are
possessed by each of the fishes, yet their values differ so widely
that an understanding of the habits and associative processes of any
one of the species would be impossible except in the light of just
such facts as Herrick has discovered. Clearly, then, we must know the
relative importance of the various sense-impressions received by an
animal before we can discuss its behavior or psychic characteristics
intelligently.

Furthermore, if behavior is to be serviceably described in terms of
stimuli and physiological conditions it is necessary first of all
to recognize that an animal responds to a situation, not to any one
independent and isolated stimulus. Every situation, to be sure, may be
analyzed into its component simple stimuli, but the influence of each
and all of these stimuli is conditioned by the situation. Too often in
our accounts of an animal's behavior we name some one stimulus as the
condition of the reaction and entirely neglect the situation, without
which the stimulus would have been of quite different value to the
animal. For any given stimulus other external and internal stimuli
constitute an environment. The complete description of a reaction
demands knowledge of all the stimuli which enter into the situation and
of their mutual relations of interference or supplementation. A frog
which in its native habitat and undisturbed by an unusual situation
would react violently to the light touch of a stick may give no sign
of reaction to the same stimulus when a human being stands nearby.
The influence of the tactual stimulus has been changed entirely by
the simultaneous appearance of visual, olfactory, and possibly still
other sense-data (man). The animal reacts not to the touch alone, but
to this stimulus as part of a certain situation. The general effect
of a situation we often speak of as excitement, timidity, etc. These
are words for which must be substituted in our accounts of animal
behavior accurate descriptions of the situations. Experimental studies
prove that an animal must become thoroughly accustomed to the general
situation in which it is to be observed before the influence of any
particular condition can be studied to advantage.

Only a few of the important reactions of an animal to either external
or internal stimuli are visible to the casual observer, and many of
them can be detected only by the employment of indirect methods.
Frequently the lack of a visible motor response to a new situation is
good evidence of a fundamentally important reaction. The death-feigning
opossum, crustacean, or insect, truly reacts by becoming motionless. As
Whitman[141] has shown in the case of the leech it is as hazardous to
judge of the degree of sensitiveness of another animal solely on the
basis of our own as it is to maintain that lower animals possess only
the senses which are ours. Varied, indirect and delicate methods are
necessary in the investigation of the senses, as the results of the
experiments to be described below help to prove.

It is my purpose in this paper to call attention to and emphasize
the importance of studying stimuli in their mutual relations of
interference and supplementation. This I shall do by presenting the
results of an investigation of the behavior of the green frog. I shall
discuss briefly, first, the sense-data received by the animal, their
relative importance, significance, and mutual relations, and, next, the
phenomena of reënforcement and inhibition.


II. THE SENSORY REACTIONS OF THE GREEN FROG

The following sensory reactions have been observed in the frog, but
most of them have not been studied with care: olfactory, temperature,
visual, tactual, equilibrational, and auditory. It is my purpose to
investigate each of these senses in such fashion that we shall know the
receptive capacity of the animal. Thus far I have completed only the
work on auditory reactions, but the chemical and temperature senses and
vision will be discussed in similar fashion later.

At present there is little known concerning the chemical senses.
Unpublished observations made by Mr. Sherwin in this laboratory
indicate the existence of olfactory sensitiveness to camphor, iodine,
and several other strong stimuli. The reactions to the stimuli were
slow, however, and there is no reason to believe that the sense of
smell is of great importance to the animal. Of taste barely more is
known than that it is present.

There is marked sensitiveness to variations in temperature, as I
have demonstrated by preliminary test experiments, but the limits,
distribution, and significance of this sensitiveness remain to be
investigated. I am not aware that the existence of temperature spots
has been determined. In connection with a study of the reactions of
frogs to light Torelle[142] discovered that the animals suddenly become
inactive and usually attempt to bury themselves when brought into a
temperature of 8° to 10° C. This reaction is prompt and definite; its
value to an animal which hibernates is evident, yet one would scarcely
anticipate the suddenness and regularity with which it occurs.

My studies of habit-formation and reaction-time[143] have revealed
the importance of vision in the life of the frog. Perception of
movement appears to be of far greater value to the animal than
perception of form or color. The spectral colors are discriminated in
all probability, for the animals react very differently to those of
the blue end than to those of the red. According to Torelle blue is
preferred to red. There is evidence that red has a higher stimulating
value than blue, and the apparent avoidance of red in Torelle's
experiments may be due to this fact. None of the work with which I am
familiar demonstrates that the suspected color-reactions are due to
stimulation of the eye. They may be due to stimulation of the skin, for
Parker[144] has shown that the reactions of _Rana pipiens_ to light are
due to stimulation of the sink as well as of the eyes, or they may even
be due to intensity instead of color.

The tactual-auditory sense series is better known and also, it would
appear, better developed than the chemical series. A large portion of
the body surface of the green frog is keenly sensitive to mechanical
stimulation, and Steinach[145] by measurement of electrical changes in
the nerves of the skin has discovered the existence of "touch spots."
His method, which is ingenious, promises to be of considerable value in
the objective investigation of the senses, but it involves operations
on the subject which inevitably destroy the normal condition of the
sense.

According to Steinach we have in the negative variation in the
electrical condition of nerves during stimulation a phenomenon which
may be used in the determination of the threshold of stimulation as
well as in the investigation of irritability. In a previous paper[146]
I have discussed the associational rôle of tactual impressions as well
as the tactual reaction-time. All my observations lead me to believe
that touch is a highly developed and important sense in the green frog.

Of the senses intermediate between touch and hearing that of
equilibration has been most discussed. Certainly there is good reason
to suppose that the sense-organs of the semicircular canals of the ear
furnish the animal with impressions of position, movement, and possibly
also of direction. Further study of the tactual-auditory senses of
frogs may indicate the existence of conditions similar to those
discovered by Parker in certain fishes, in which, as he remarks, "the
skin, lateral line organs and ears represent, figuratively speaking,
three generations of sense-organs. The oldest is the skin stimulated
by varying pressures, such as are produced by irregular currents, and
capable of initiating equilibrational responses. From the skin have
been derived the lateral line organs stimulated by water vibrations of
low rate, and also significant for equilibration. Finally, from the
lateral line organs have come the ears stimulated by water vibrations
of a high rate and important for equilibration. The ear, unlike the
skin and lateral line organs, is differentiated for its two functions,
the sacculus for hearing, the utriculus for equilibration."[147]

The sense of hearing remains to be considered. My attention was first
drawn to this subject by failure to obtain motor reactions to sounds in
the investigation of the time-relations of the neural processes of the
green frog. Although a large number of sounds of different qualities,
pitches, and intensities were employed, no visible motor reactions
were observed. This led me to seek the significance of what appeared
to be either a surprising lack of sensitiveness to changes in the
environment which would naturally be expected to stimulate the animal,
or an interesting and important case of the inhibition of reaction to
auditory stimuli. This suggested the question, Are frogs deaf, or do
they under certain conditions completely inhibit their usual reactions
to sound?

In the literature on the senses and reactions of frogs I have found
nothing which contributes importantly to our knowledge of the sense
of hearing. Most of the investigations which deal with the ear are
concerned with the equilibrational and orientational functions of the
labyrinth organs, and have nothing whatever to say about hearing.
In the natural histories the existence of a well-developed sense of
hearing is usually assumed, and numerous instances of what are supposed
to be reactions to sound are cited. It is to be noted, however, that
none of the observations in these popular works furnishes satisfactory
proof of the exclusion of the influence of visual stimuli. Among the
few references to frog audition of which I have knowledge, the only
one which seems worthy of special notice is that of Gaupp in his
Anatomie des Frosches. Since his few paragraphs sum up the state of
our knowledge on the subject, while at the same time furnishing an
illustration of the assumption of hearing on the basis of analogy,
I present the substance of them in free and slightly abbreviated
translation.

"The labyrinth organ has an acoustic and non-acoustic (static)
function. For these two functions, according to the leading if not
generally accepted view, entirely different portions of the organ are
in question, and since the non-acoustic is attributed to the three
_Cristae acusticae ampullarum_ and the three _Maculae_ (_M. recessus
utriculi_, _M. sacculi_, _M. lagenae_), there remain for the acoustic
function only the _Papilla basilaris_ and the _Macula neglecta_. It is
not certain, however, that the non-acoustic organs do not participate
in the acoustic function.

"With regard to the acoustic sense of the frog nothing exact is known.
That it exists, and that in good development, is certain. The existence
of the drum and columella, and the fact that frogs have a voice are
unmistakeable proofs of hearing. The participation of the _Papilla
basilaris_ in acoustic functions is rendered certain by comparative
anatomical studies: the _Papilla basilaris_ is the nerve end-organ from
which, in the mammalia, the undoubtedly acoustic organ of Corti arises.
From analogy of structure we may also infer an acoustic function in
the _Macula neglecta:_ on this, as on the _Papilla basilaris_, there
is a simple tectorial membrane, and further the _Pars neglecta_, like
the _Pars basilaris_, has a strong thick wall which only in a limited
region, namely, where it approaches a part of the perilymphatic space,
is markedly thinner." (For fish Breuer (1891) has already stated that
if they really hear--which is proved--the _Macula neglecta_ alone can
come into consideration in connection with the function, for there is
no _Papilla basilaris_ in fishes, and the six other nerve end-organs
apparently serve the non-acoustic function.)[148]

As the green frog does not respond visibly to sounds under experimental
conditions, I found it necessary to employ indirect methods in the
study of audition. By observing the influence of sounds on respiration
and on the reactions to certain electrical, tactual, and visual
stimuli, I obtained results which, since they have already been
described in detail elsewhere,[149] may be summarized here as follows:

1. Observation of frogs in their natural habitat shows that they are
stimulated by sounds, but the sense of hearing apparently serves rather
as a warning sense which modifies reactions to other simultaneous
or succeeding stimuli than as a control for definite auditory motor
reactions.

2. Experimental tests prove that sounds modify the frog's reactions
to visual and tactual stimuli. When the sound accompanies the visual
or tactual stimulus it serves to reënforce the reaction to the other
stimulus, but when given alone it never causes a motor reaction.

3. The green frog responds to sounds made in the air, whether the
tympana be in the air or in water. There is some evidence that
the influence of auditory stimuli is most marked when the drum is
half-submerged in water. The influence of sounds upon tactual reactions
is evident when the frog is submerged in water to a depth of 4 cm.

4. Sounds varying in pitch from those of 50 to 10,000 vibrations per
second affect the frog. The most striking results were obtained by
the use of an electric bell with a metal gong. With this sound in
connection with a weak tactual stimulus a maximum reaction may often be
obtained even when either stimulus alone causes no perceivable reaction.

5. Sounds modify the reactions of the frog after tympana and columellæ
are removed. Cutting of the eighth cranial nerves causes disappearance
of the influence of sound. It is clear, then, that the reactions to
sounds are really auditory reactions and that the sense of hearing in
the frog is fairly well developed, although there is little evidence of
such a sense in the motor reactions of the animal.

6. Experiments during the spring months show marked influence of sounds
for both males and females, whereas experiments made during the winter
indicate a much diminished sensitiveness to auditory stimuli in both
sexes, but especially in the male.


III. THE MUTUAL RELATIONS OF STIMULI

In studying the various influences of complication of stimuli in the
frog, I have used two methods: the measurement of reaction-time and of
the amount of reaction. The reaction-time results will be presented
first.

Reaction-time to electric stimulation of the skin was studied with
special attention to the influence of other stimuli which were given
in definite temporal relation to the electric stimulus. A Hipp
chronoscope, controlled by a Cattell's falling screen, served as a
time-measuring apparatus. The other essentials of the apparatus were
a reaction-box, and devices for giving the stimuli and indicating the
reaction. On the bottom of the reaction-box a series of wires were so
placed that an electric stimulus could be given to the frog resting
upon them by the closing of a key in the hands of the experimenter.
In preparation for each experiment the frog was placed upon these
open circuit wires in such a position that the weight of its body
pressed upon a delicate spring in the floor of the box, thus causing
the chronoscope circuit to be completed. The forward jump of the frog
in response to stimulation caused the breaking of this circuit by the
release of the spring upon which the animal rested. When all was in
readiness for an experiment the chronoscope was started, and a key
closed which simultaneously gave an electric stimulus to the frog and
completed a circuit which caused the chronoscope record to begin. The
stimulus consisted of a current from one or more "Mesco" dry cells.
The motor reaction of the frog broke the chronoscope circuit, thus
causing the chronoscope record to stop. It was then possible for the
experimenter to read from the dials of the chronoscope the time, in
thousandths of seconds, intervening between stimulus and reaction
(reaction-time). In case of additional stimuli in connection with the
electric, various simple devices were introduced to meet the demands
of the experiments. These will be described in connection with the
statement of results in each case.

_Electric and photic stimuli._ A photic stimulus was given from one
to two seconds before the electric stimulus by the turning on of a
sixteen-candle-power incandescent light, which was placed thirty cm. in
front of the frog in the case of one series of experiments and fifteen
cm. above it in another. The light uniformly inhibited reaction to the
electric stimulus, as is shown by the results of Table 1.


TABLE 1

    Title of investigation,--Electric-Visual (Red Light).
    Experimented on,--Green Frog No. 4.
    Harvard Psychological Laboratory,--9.40 A.M., Feb. 28, 1902.
    Chronoscope control average, 189σ,--Electric stimulus, 1 Cell.

    NO LIGHT.

    _Number of_
    _Experiment._ _Reaction-time._
         1            152σ
         2            145
         3            221
         4            327
         5            263
         6            271
         7            329
         8            215
         9            225
        10            216

    LIGHT BEFORE ELECTRIC STIM.

        11            No reaction.
        12            No reaction.
        13            No reaction.
        14            No reaction.
        15            No reaction.

    NO LIGHT.

        16            216

The inhibitory influence of light depends upon the intensity of the
electric stimulus. Even a very strong light will not cause much
retardation of reaction to a three or four cell current. As the
strength of the electric stimulus decreases the delay of reaction
increases, until finally there is complete inhibition. At this point,
an electric stimulus, to which the frog would react almost invariably
when there is no disturbing condition, will fail to cause reaction in
the presence of a sudden increase in light intensity.

Merzbacher[150] states that the leg reflex of a frog, so placed that
its legs hang free in the air, is greater in response to a given
cutaneous stimulus in darkness than in daylight.[151]

_Electric and visual stimuli_ (moving object). For the purpose of
determining the effect upon reaction-time to an electric stimulus
of stimulation of the eye by a rapidly moving object, experiments
were made in which, as in the case of electric and photic stimuli,
reactions to electric stimulus alone and to the visual and electric
were observed alternately. Thus in the case of each pair of reactions
it was possible to note whether the visual stimulus shortened or
lengthened the reaction-time. The visual stimulus was given by quickly
moving a finger before a window in the reaction-box.

Two series of twenty pairs of reactions each were taken with each of
two frogs. In the first series the finger was suddenly moved across the
window and the electric stimulus was given either simultaneously or
a small fraction of a second later. It was impossible to arrange for
accurate measurement of the temporal relations of the two stimuli in
the case of these tests. In the second series the finger was moved back
and forth before the opening in the reaction-box for an interval of at
least a second before the electric stimulus was given.

These experiments, which were in the nature of preliminary tests,
yielded the following results. When the stimuli were given almost
simultaneously the visual reënforced the electric as was indicated by a
shortening of the reaction-time. As appears in the upper part of Table
2, the average time of forty reactions, twenty for each frog, to the
electric stimulus was 148^{σ},[152] and to the same stimulus when it
followed the visual 128^{σ}. Furthermore, examination of the several
pairs of reactions shows, as is indicated in the table, that there were
twenty-seven cases in which the visual stimulus caused shortening of
the reaction-time (reënforcement of the electric stimulus) to thirteen
in which it caused lengthening (inhibition). When the visual stimulus
preceded the electric by at least a second, the reaction-time to the
electric stimulus was greatly lengthened. The averages are 150^{σ}
for the electric stimulus alone, 178^{σ} when it was preceded by the
visual. In this series there are twenty-five cases of inhibition to
fourteen of reënforcement.

_Electric and visual stimuli_ (moving red disc). The indications
of the importance of the temporal relations of stimuli, so far
as reaction-time results are concerned, furnished by these crude
preliminary observations led to a more accurate study of the subject. A
revolving disc, which moved at the rate of one revolution per minute,
was so arranged that at a certain point it closed an electric circuit
in which a magnet had been placed. This magnet attracted a steel arm
at the end of which a disc of red cardboard 12 mm. in diameter was
suspended. With the making of the circuit the steel arm was drawn
downward suddenly and the red disc, by reason of the vibrations of
the arm moved rapidly back and forth in front of a window in the
reaction-box. In this way the moving object was exposed to view about
ten cm. to the right and three cm. in front of the right eye of the
frog. The revolving disc, a fraction of a second later, completed the
electric stimulus circuit. Thus both stimuli were given automatically,
at such an interval apart as the experimenter desired. In the two
series of results now to be described the intervals were 0.1 and 0.5
second respectively.


TABLE 2

Reaction-time to Electric Stimulation Alone, and to the Same when
preceded for 0.1, 0.5, or 1.0 Second by Visual Stimulus.

    Key:
    1   = Frog.
    2   = Electric Alone.
    3   = Visual 0.1" before elect.
    4   = Number Inhibited.
    5   = Number Reënforced.
    6   = Number Equal.
    7   = Visual 1.0" before elect.

    1      2      3       4     5     6     2       7       4      5     6

  Preliminary Series. Visual Stimulus Moving Finger.
                                                 Averages for 20 reactions.
  No. 5. 179ˢ    158ˢ     6    14     0   163ˢ     206ˢ     14     6     0
  No. 6. 116      98      7    13     0   136      150      11     8     1
  Gen.
   Aver. 148     128     13    27     0   150      178      25    14     1

  Visual Stimulus Moving Red Disc.
  Visual 0.1" before electric.      Visual 0.5" before electric.

  Series I.   Averages for 25 reactions.

  No. 5. 177     163     10    15     0   170      255      15     9     1
  No. 6. 148     112      6    19     0   115      178      18     7     0

  Series II.   Averages for 25 reactions.

  No. 5. 135     120      7    18     0   155      259      24     1     0
  No. 6. 128     111      6    19     0   132      227      17     7     1
  Gen.
   Aver. 147     126     29    71     0   143      230      74    24     2

These series consisted of twenty-five pairs of reactions each, with
two animals. The results of the series are presented separately, in
the lower half of Table 2, because the experiments which constitute
them were separated by a period of three weeks. It is to be noted
that these results agree fully with those of the preliminary series.
The visual stimulus of a moving red disc, given 0.1 second before a
2 cell electric stimulus, reënforces the electric reaction, _i. e._,
it shortens the time of reaction. The same visual stimulus given 0.5
second before tends to inhibit the electric reaction, _i. e._, it
lengthens the time of reaction.

_Tactual and auditory stimuli._ Since in the frog auditory stimuli
under experimental conditions seldom if ever cause visible motor
reactions, the study of the influence of this mode of stimulation
upon the reactions to other simultaneous or succeeding stimuli
is of special interest. In the investigation of the relations of
auditory stimulation to other forms of reaction _amount_ of reaction
instead of _reaction-time_ was taken as a measure of the influence
of the stimulus. By a method the details of which may be most easily
understood by reference to the plan of the apparatus in Figure 1, the
influence of auditory stimuli on the leg-movement induced by tactual
stimulation was observed.

In these experiments the frog sat astride a wooden support, held in
position by linen bands over the back and a wire screen cap over
the head. The hind legs hung free, and any movement of one of them
in response to a stimulus could be read in millimetres by reference
to a scale on the wooden support. This method of measuring the
value of a stimulus in terms of leg-reflex has been used by several
investigators--most recently by Merzbacher.[153] I have found it
desirable, as did Merzbacher, to observe the movements of a shadow of
the leg on the scale and thus read the amount of movement, rather than
to watch the leg itself and attempt to project it upon the scale.

As is indicated in Fig. 1, the auditory and tactual stimuli were given
automatically by means of a swinging pendulum, _P_, which was held in
position by the magnet _a_ until released by the experimenter. Early in
its swing the pendulum turned the key, _m_, thus completing a circuit
which caused the auditory stimulus to be given; later in the swing the
key, _n_, was turned, and the tactual stimulus thus given through the
magnetic release of the lever, _l_. The interval between the auditory
and the tactual stimuli could be varied from 0 to 2" by changing the
position of the key, _n_. For intervals over 1" it was necessary to
arrange this key so that the tactual stimulus was given at some time
during the return swing of the pendulum.

The auditory stimulus used was either the sound of a quick hammer blow
(momentary stimulus of Series I), or the ringing of an electric bell
for a certain length of time (prolonged stimulus of Series II). In Fig.
1 the bell is shown. It was placed eighty cm. from the frog, and in
order that the influence of vibration of the experiment table might
be avoided it was suspended from the pendulum frame. When the hammer
was used it was placed sixty cm. from the frog, on the pendulum table.
The holder for the frog and the tactual apparatus occupied a separate
table which was not disturbed by the jars of the pendulum table.

[Illustration: Figure 1. Auditory-tactual Reënforcement-Inhibition
apparatus. _P_, pendulum; _p_, contact point of _P;_ _b_, attachment
for electro-magnet, _a;_ _m_, key for circuit of electric bell, _B;_
_n_, key for magnet circuit of tactual apparatus; _K_, hand-key for
release of pendulum and temporary closing of electric bell circuit;
_k_{1}_, _k_{2}_, _k_{3}_, keys in circuits; _e_, _f_, _g_, magnetic
release for tactual apparatus; _l_, pivoted lever, bearing rubber cone,
_T_, and weights, _w_. (Drawn by Dr. Wm. E. Hocking.)]

The tactual stimulus was given by a rubber cone, _T_, two mm. in
diameter at its apex. This rubber point, after the electric release
of the lever to which it was attached, struck the frog at the middle
point of a line drawn between the posterior margins of the tympana. The
intensity of the stimulus could be varied by weighting the lever, _l_,
at _w_.

All experiments were made with the green frog, _Rana clamata_ Daudin.
The reactions were taken regularly at half-minute intervals in pairs:
first, a tactual stimulus reaction, then an auditory-tactual reaction.
Ten, fifty, or one hundred pairs constituted a series. So far as
the condition of the frog is concerned there seems to be nothing
undesirable in long series, for fatigue does not appear, and so long
as the animal is kept moist and in an unconstrained position, it
continues to react normally, and without frequent struggles to escape.
The advantage for the purposes of this investigation of taking the
reactions in pairs, rather than taking separate series of reactions
for each stimulus or combination of stimuli, is obvious. It enables us
to compare directly the reactions of each pair, in other words those
reactions which took place under most nearly identical conditions, and
to note at once whether the auditory stimulus reënforced or inhibited
the tactual reaction.

During a series the intensity of the tactual stimulus was changed as
conditions demanded, but for any one pair of reactions it was always
the same. It not infrequently happened that an intensity which at
first caused merely a slight movement of the leg, later in the series
uniformly brought about a maximal contraction, or the reverse might
be true, and inasmuch as a maximal reaction to the tactual stimulus
alone left no opportunity for judging of the influence of the auditory
stimulus, when it was given in addition to the tactual, it was always
necessary in such cases so to alter the intensity of the tactual
stimulus that a medium reaction resulted.

The frogs, after being placed in the saddle-like holder and held firmly
for a few seconds, seldom struggled very much, but if bound tightly
they became irresponsive to the stimuli.[154] It was, therefore,
necessary after they had quieted down to loosen the bands which held
them in position. For the purpose of excluding the influence of visual
stimuli a wire screen cap covered with black cloth was put over the
head; this served to keep the animal in position as well as to exclude
visual stimulation.

_a. Momentary auditory stimulation._ Four frogs were used for a study
of the influence of the momentary sound produced by a hammer blow,
and for each of these animals fifty pairs of reactions were recorded
in series each day. The temporal relation of the stimuli was changed
daily during a week of experimentation: the results therefore consist
of fifty pairs of reactions with each frog for each of the following
seven intervals: (1) Auditory and tactual stimuli simultaneous, (2)
auditory .25" before tactual, (3) auditory .45" before, (4) auditory
.15" before, (5) auditory .65" before, (6) auditory .35" before, (7)
auditory .90" before. The intervals were used in the experiments in the
above order to avoid the formation of the definite habits of reaction
which regular increase in the interval would have favored.

Typical of the results with all the animals are the following (Table
3) which were obtained with No. 1, a male. The figures in each case
indicate the average of fifty reactions. Reënforcement and inhibition
are expressed in terms of the tactual reaction, _i. e._, the
auditory-tactual reaction is so many per cent greater (reënforcement)
or less (inhibition) than the tactual. In the tables reënforcement is
indicated by the + sign; inhibition by the - sign. In the last column
of the table is given the number of reactions that were reënforced or
inhibited. This was determined by comparing directly the reactions
of each pair. Cases in which the two reactions were the same were
distributed equally between the two classes: tactual reactions
reënforced by auditory stimulus, and tactual reactions inhibited by
auditory stimulus. Assuming that the auditory stimulus was without
effect upon the tactual reaction, the number of reactions in these two
classes would be approximately the same, hence all auditory-tactual
reactions over half in a series, _i. e._, over twenty-five, which
are greater than the corresponding tactual reactions, are reënforced
reactions, and can be taken as a measure of the reënforcing influence
of the auditory stimulus. In the same manner all reactions over half
which show inhibition can be taken as a measure of the inhibitory value
of the auditory stimulus.

As preliminary tests described in an earlier paper[155] furnished
evidence of sex-differences, it is worth while to compare the results
given by the males and females in these experiments with momentary
auditory stimulation. For purposes of comparison I have presented
in Table 4 the reënforcement-inhibition values given by the males
and females for each interval. Column one contains the value of the
auditory-tactual reaction in terms of the tactual reaction; column two,
the number of reactions in excess of half which were reënforced or
inhibited.


TABLE 3. FROG NO. 1. MOMENTARY AUDITORY STIMULUS, HAMMER BLOW. WEIGHT
USUALLY 5 OR 10 GRAMS

               Reaction    Reaction    Amount of      Number of
                 to      to Auditory  Reënforc'm't    reactions
               Tactual   and Tactual      or         Reënforced
    Interval.   Stim.       Stim.     Inhibition.   or Inhibited.

      0"        6.84mm.    11.08mm.      +62.0%         +17.0
    .15        22.22       28.96         +30.3          +17.0
    .25        16.30       21.72         +33.3          +13.0
    .35        24.90       25.32         + 1.7          + 0.5
    .45        17.56       13.64         -22.3          -10.0
    .65        17.46       15.72         -10.0          - 6.0
    .90        31.26       31.48         + 0.7          + 0.5


TABLE 4. MOMENTARY AUDITORY STIMULUS, HAMMER BLOW

                       _Males_                 _Females_
                    Nos. 1 and 3.            Nos. 2 and 4.
             Per centum        No. of    Per centum    No. of
    Interval    Diff.          Reacts.      Diff.      Reacts.
      0"      +82.5% (Reënf't)   +17.5      +58.0%      +12.7
    .15       +58.1              +17.0      +25.4       + 8.5
    .25       +32.3              +12.7      +39.8       +12.7
    .35       + 4.0              + 1.2      - 9.7       - 3.2
    .45       -13.5 (Inhibition) - 7.2      -13.9       - 7.2
    .65       -12.5              - 6.2      -11.8       - 7.2
    .90       - 0.7              - 1.5      - 2.6       - 0.5

In these results two striking differences between the males and females
appear: first, the reënforcement is not so great for the females as
for the males; second, inhibition appears earlier and continues longer
with the females than with the males. The average reënforcement with
simultaneous stimuli is 82.5% for the males against 58.0% for the
females. Inhibition begins to appear in case of the females when the
interval between the stimuli is .25" to .35"; in case of the males it
appears between .35" and .45". Finally at .90" interval inhibition is
slightly greater for the females.

Although the exact significance of these facts is unknown, it is
not improbable that they are indicative of fundamentally important
sex-differences in reaction to sound. The males among frogs are
usually the vocalists, although in some species the females also
croak. Moreover, in case of the green frog the tympanum of the male
is much larger than that of the female. The results presented would
seem to indicate that certain sounds stimulate the males to activity,
whereas they inhibit activity in the females.

Graphically represented, the results of the momentary auditory stimulus
experiments with frogs Nos. 1, 2, 3, and 4 are as follows:

[Illustration: FIG. 2. Reënforcement-Inhibition curves for momentary
auditory stimulation, based upon amount of reaction. Male No. 1 ----
Male No. 3 ....]

[Illustration: FIG. 3. Reënforcement-Inhibition curves for momentary
auditory stimulation, based upon amount of reaction. Female No. 2 ----
Female No. 4 ....]

The curves are all plotted by the method which will now be
described in connection with Fig. 2. This figure presents the
reënforcement-inhibition curves for the males No. 1 (solid line in
the figure) and No. 3 (broken line). If in this figure we let the
zero-point on the ordinates represent the value of the reaction to the
tactual stimulus when given alone, then the value of the reaction to
the auditory-tactual stimuli would be represented at some point above
the zero-point if this reaction was greater than the tactual reaction
(reënforcement), and below the zero-point if the reaction was less
than the tactual (inhibition). Since one of our chosen measures of
reënforcement and inhibition is the amount, in per cent of tactual
reaction, by which the auditory-tactual reaction exceeds or falls short
of the tactual reaction, such a curve of reënforcement-inhibition
as that of Fig. 2 (solid line) can be constructed at once from the
data given in column four of Table 3. Here the auditory stimulus,
when simultaneous with the tactual, caused 62% reënforcement, as is
indicated in the figure. The figures in the left-hand margin of the
curves indicate amount of reënforcement or inhibition in per cent of
tactual reaction; those at the bottom of the curves mark the intervals.
On the curves dots indicate the intervals used in the experiments. Each
of the curves is plotted on the basis of 700 reactions.

[Illustration: FIG. 4. Reënforcement-Inhibition curves for momentary
auditory stimulation, based upon number of reactions. Male No. 1. ----
Male No. 3 ....]

[Illustration: FIG. 5. Reënforcement-Inhibition curves for momentary
auditory stimulation, based upon number of reactions. Female No. 2 ----
Female No. 4 ....]

In every way comparable with the curves for the males No. 1 and No.
3 in Fig. 2 are those for the females No. 2 and No. 4 of Fig. 3. The
similarity of the two curves in each figure is noteworthy. Inasmuch as
the conditions of experimentation were the same for all the animals
this would seem to indicate sex-differences which are worthy of further
investigation. The curves show clearly the greater reënforcement in the
males, and the greater inhibition in the females.

Figures 4 and 5 are the reënforcement-inhibition curves for the same
series of experiments plotted on the basis of the _number_ of reactions
in excess of half that were reënforced or inhibited. As there were
fifty pairs of reactions with each frog for each interval, uniform
reënforcement would be represented by twenty-five reactions above
the base-line; uniform inhibition by twenty-five reactions below the
base-line. The number of reactions is indicated by the figures in
the left margin; the intervals, by those below the base-line. As an
illustration of the application of the method of plotting, the curve
for male No. 1 (solid line) of Fig. 4 is constructed from the data
of column five of Table 3. With simultaneous stimuli 17 reactions in
excess of half, _i. e._, 17 + 25, or 42, were reënforced; at .35"
interval .5 of a reaction was the average amount of reënforcement;
at .45" interval 10 reactions in excess of half, _i. e._, 35, were
inhibited, therefore the curve falls to 10 below the base-line.

[Illustration: FIG. 6. Composite Reënforcement-Inhibition curve for
momentary auditory stimulation, based upon amount of reaction. Frogs
Nos. 1, 2, 3, 4. (Males and females.)]

[Illustration: FIG. 7. Composite Reënforcement-Inhibition curve for
momentary auditory stimulation, based upon number of reactions. Frogs
Nos. 1, 2, 3, 4. (Males and females.)]

Just as Figures 2 and 3 permit of direct comparison of the results
of the measurement of the _amount_ of reënforcement and inhibition
for males and females, so Figures 4 and 5 make possible comparison in
similar fashion of the _number_ of reënforced and inhibited reactions
for the sexes. It is to be noted that the two sets of curves, plotted
on the bases of _amount_ and _number_ of reaction, agree in all
important respects.

Figure 6 is the composite curve of amount of reënforcement-inhibition
for the four animals; Figure 7 is the composite curve of the number of
reactions reënforced and inhibited.

Summarily stated, the results of the experiments thus far described
are: (1) The auditory stimulus of a quick hammer blow produces the
maximum amount of reënforcement of tactual reaction when it is given
simultaneously with the tactual stimulus; (2) as the interval between
the auditory and the tactual stimulus approaches .35″ the amount of
reënforcement gradually decreases; (3) when given .35″ before the
tactual stimulus the auditory is practically without effect upon the
tactual reaction; (4) as the interval increases above .35″ inhibition
begins to appear; (5) the inhibitory influence of the auditory stimulus
is greatest when the interval is about .45″; (6) when the interval is
as long as .90″ the auditory stimulus is again ineffective. It thus
appears that the reënforcement-inhibition curve of this particular
stimulus under the conditions described is representative of a neural
process which completes itself, in passing through two phases, a
positive phase (reënforcement) and a negative phase (inhibition), in
about one second.

_b. Prolonged auditory stimulation._ The experiments previously
described have proved that a momentary auditory stimulus, which when
given alone never produces a visible motor reaction, either reënforces
or inhibits the reaction to a tactual stimulus which it accompanies or
precedes. The experiments now to be described were made for the purpose
of ascertaining whether reënforcement and inhibition occur in the same
way if the auditory stimulus is prolonged, instead of momentary.

In a trial series of experiments with frog No. 1, one hundred pairs
of reactions were recorded for each of six intervals of auditory
stimulation. The auditory stimulus was given by the ringing of an
electric bell. For all intervals the ringing of the bell continued
until the tactual stimulus was given. When the two stimuli were given
simultaneously the auditory stimulus was necessarily momentary, as
in the foregoing experiments, but for all other relationships of
the stimuli the bell rang for a certain length of time before the
tactual stimulus was given. The six relations of the stimuli were:
(1) simultaneous, (2) bell .2″ before and until tactual, (3) bell .6″
before, (4) bell 1.05″ before, (5) bell 1.5″ before, and (6) bell 2.0″
before. The other conditions of these experiments were the same as
those previously described, except that the auditory stimulus was here
given by the opening of the key which released the pendulum, instead
of being given by the turning of a key in the course of the pendulum
swing. This method of giving the auditory stimulus as the pendulum was
released was found unsatisfactory because of the irregularity of the
magnetic release; at one time the pendulum would start immediately, at
another time there would be a delay of as much as .1″.

The reënforcement-inhibition curve plotted on the basis of the 1200
reactions in this series is presented in Fig. 8. Before stopping to
consider the important features of this curve we should note the
results of certain more accurate experiments with prolonged auditory
stimulation.

[Illustration: FIG. 8. Reënforcement-Inhibition curve for prolonged
auditory stimulation, based upon amount of reaction. Frog No. 1.]

With two animals, No. 2, a female, and No. 3, a male, fifty pairs
of reactions were taken for nine different intervals (see Table 5)
of auditory stimulation. Each of the curves of Figures 9 and 10
is therefore based upon 900 reactions. The conditions for these
experiments were the same as those for the momentary stimulation
series, save that the electric bell took the place of the electrically
actuated hammer, as the mechanism for auditory stimulation.

[Illustration: FIG. 9. Reënforcement-Inhibition curves for prolonged
auditory stimulation, based upon amount of reaction. Female No. 2 ----
Male No. 3 ....]

The important facts exhibited by the results of these prolonged
auditory stimulation experiments in contrast with those with momentary
auditory stimulation are: (1) That whereas for the momentary
auditory stimulus of a hammer blow the reënforcement is greatest for
simultaneous stimuli, in case of the prolonged stimulation with the
electric bell, reënforcement increases during an interval of .25″
of auditory stimulation. Hence, the two conditions of stimulation
give us different types of reënforcement-inhibition curve. For the
momentary stimulus the maximum reënforcement appears at simultaneity,
and for the prolonged stimulus at .25″; (2) that the transition from
reënforcement to inhibition occurs at 1.2″ in the prolonged stimulation
curves, while in the momentary stimulation curves it occurs at .35″;
(3) that the maximum inhibition which appears in the curves under
discussion at about 1.5″ is less in comparison with the amount of
reënforcement than that of the momentary stimulation curves; (4) that
the auditory stimulus becomes ineffective when the interval during
which it continues before tactual stimulation is 2.0″. The curves of
Figures 8, 9, and 10 are then representations of a neural process which
passes through a positive and a negative phase in about 2″. The effect
of prolongation of the auditory stimulation interval is to lengthen
the period of reënforcement; the period of inhibition shows little
modification.

For the purpose of showing in greater detail the nature of the results
of this work the data from which the curves of Figures 9 and 10 were
constructed are presented in the accompanying Table 5.

[Illustration: FIG. 10. Reënforcement-Inhibition curves for prolonged
auditory stimulation, based upon number of reactions. Female No. 2 ----
Male No. 3 ....]

Having now presented the results of my own investigation I wish to
call attention to certain of their relationships to the work of other
investigators, and to discuss briefly their significance.


TABLE 5. PROLONGED AUDITORY STIMULATION (ELECTRIC BELL)

          _Frog No. 2._    _Female._    _Weight usually 25 grams._

                             Auditory                       Number of
                               and           Amount of      Reactions,
                Tactual       Tactual      Reënforcement    Reënforced
    Interval. Stimulation.  Stimulation.   or Inhibition.  or Inhibited.

       0″       9.20 mm.      12.12 mm.        + 31.7%         +10.0
     .25        4.56          11.88            +160.5          +22.5
     .45        8.94          16.94            + 89.5          +18.5
     .65       17.18          22.50            + 31.0          +15.5
     .90        9.42          13.32            + 41.4          +14.5
    1.20       10.54           9.64            -  8.5          - 2.5
    1.40       24.00          20.64            - 14.0          - 6.0
    1.58       19.16          17.80            -  7.1          - 7.0
    1.95       14.50          15.40            +  6.2          + 5.5

       _Frog No. 3._    _Male._      _Weight usually 5 or 10 grams._

       0″      14.92 mm.      25.20 mm.        + 68.9%         +11.0
     .25       15.88          38.54            +142.7          +24.0
     .45       13.48          26.02            + 92.9          +13.5
     .65       18.30          27.94            + 52.6          +13.0
     .90       20.94          29.06            + 38.8          + 9.5
    1.20       21.90          30.58            + 39.6          + 1.0
    1.40       19.18          18.34            -  4.4          - 5.5
    1.58       32.24          26.30            - 18.1          - 3.0
    1.95       13.86          14.14            +  2.0          + 2.0


VI. DISCUSSION OF LITERATURE AND RESULTS

The literature on reënforcement and inhibition is large, and even
that portion of it which deals especially with the importance of the
temporal relations of stimuli in connection with reënforcement and
inhibition is so extensive that it does not seem worth while to attempt
to give a systematic résumé of it for the purposes of this paper. I
shall therefore call attention merely to those investigations which
have contributed directly to the solution of the problems with which we
are now concerned.

Bowditch and Warren[156] discovered that knee-jerk in the human subject
is reënforced when an auditory, a visual, or a tactual stimulus
precedes the tendon blow by .1″ to .5″, whereas the same stimuli have
an inhibitory influence when they are given from .5″ to 1.0″ before the
tendon blow.

At the suggestion of Bowditch, Cleghorn[157] undertook to investigate
the influence of complication of stimuli upon voluntary movements. In
this research graphic records taken in connection with an ergograph
indicated (1) that "a sensory stimulus" applied just as the muscle was
beginning to contract (voluntarily) caused an increase in the height of
the contraction, and (2) that the relaxation following a contraction
with intercalated sensory stimulus is quicker and more complete than
when no stimulus is given (p. 344). Cleghorn did not give special
attention to the significance of the temporal relations of the stimuli
which he employed, and his work was limited to the phenomenon of
reënforcement of voluntary action by reason of the appearance, during
the progress of his research, of an excellent paper on the interference
of stimuli by Hofbauer.[158]

Hofbauer covered thoroughly the ground which Cleghorn had planned to
work over. The ergographic method was employed also by Hofbauer in
his very careful study of the interference of impulses in the central
nervous system of man. It was noticed that while the subject was
rhythmically contracting a certain group of muscles in response to
some prearranged signal (_e.g._, the sound of a metronome) the report
of a pistol caused the contraction which immediately followed it to be
much greater than the average of the rhythmic series, while the next
contraction was correspondingly less than the average. It thus appeared
that the sudden sound caused, first, reënforcement of the voluntary
movement, then, inhibition. The reënforcement is greatest, according
to Hofbauer, when the voluntary movement occurs immediately after the
pistol report. When the report precedes the metronome signal by .2″
reënforcement is still marked, but thereafter it decreases rapidly in
amount, until finally at .5″ inhibition appears. When the interval
between the two stimuli is 1.0″ the first stimulus has practically
no effect upon the voluntary movement in response to the second.
(Hofbauer, p. 558.)

What Bowditch and Warren, not to mention other students of the subject,
have described for reflex action in man, Hofbauer, Cleghorn, and others
have shown to hold true also of voluntary movements. Unfortunately my
own investigation was completed up to the point of the writing of this
paper before I read Hofbauer's work, so I have not followed methods
of dealing with my data which would make our results directly and
easily comparable. But, whatever may be the relations of our results in
detail, there can be no doubt that what he has demonstrated for man is
true in its important aspect of the reënforcement-inhibition phenomena
for the frog.

Important in their bearings upon the phenomena of reënforcement and
inhibition which we are now considering, are the various studies of
refractory period and rhythm of nerve cell and fibre. The existence of
a refractory period in neural substance, similar to that demonstrated
for certain kinds of muscle by Marey,[159] Englemann,[160] Kaiser,[161]
Cushny and Matthews,[162] Woodworth,[163] and many others, has been
proved by Broca and Richet.[164]

Broca and Richet found that in the normal dog the refractory period
of the nerve substance is too short to be easily detectable, they
therefore experimented with animals which were lightly chloralized and
kept at a temperature of 30 to 34° (the mean normal temperature of the
dog is about 39.5°). Under these conditions a dog, when two identical
stimuli (quality and intensity the same) were applied to the cerebral
cortex successively, exhibited the following reactions: (1) When the
stimuli were separated by .01″ they reënforced one another (addition);
(2) when the interval was .1″ they inhibited the reaction partially
(subtraction).

Concerning this phenomenon Richet writes in his dictionary of
physiology (p.5): "Marey showed, in 1890, that the heart of the frog,
at certain moments of systole, was inexcitable. Now our experiments
prove that the cerebral apparatus, a certain time after the excitation,
also ceases to be excitable: it then has a refractory phase, and this
refractory phase is much more prolonged than that of the cardiac
muscle." In a later publication Richet[165] makes the somewhat
startling statement that a refractory period is not exhibited by the
nerves of cold-blooded animals. In the tortoise, according to his
results, reënforcement occurs so long as the interval between the
two stimuli is not greater than 2″, while for longer intervals each
stimulus to all appearances works independently. Richet seems to have
generalized from a study of the tortoise. That his generalization is
unwarranted seems to me highly probable in the light of the results
of this paper, for there are many reasons for supposing that the
reënforcement-inhibition phenomena with which we have been dealing
in case of the frog are manifestations of the existence of the same
process in the nervous system which under somewhat different conditions
of experimentation exhibits itself in the so-called refractory period.

The researches of Richet and his students indicate that the time of
the process which conditions the phenomena of reënforcement-inhibition
is about .1″. Stimuli given at .1″ intervals do not interfere with one
another. That the process underlying the refractory period and the
reënforcement-inhibition phenomena of our experiments is a rhythmic
double-phase process is made still more probable by the following
results. Horsley and Schäfer[166] found that the rate of response of
the monkey to cortical stimulation was 12 per second, and Schäfer[167]
discovered that the maximum rate of volitional impulses in man is 10 to
12 per second.

It was shown by Exner that certain movements of the foot of a rabbit
could be produced by stimulating either the cortex or the skin of the
foot. Simultaneous stimulation of both regions gives reënforcement.
Stimulation of the cortex, if given not more than 3″ before subliminal
stimulation of the skin, renders the latter effective. When both
stimuli are subliminal each makes the subsequent one effective if
the interval between them is not over 1/8″ (Schäfer[168]). Similarly
for the dog Exner[169] proved that cortical and cutaneous stimuli
reënforced one another, when both were subliminal, if the interval
between them was not greater than .6″. Cortical and auditory stimuli,
and auditory and cutaneous (of the skin of foot) gave similar results.

Physiologists have long been familiar with several aspects of the
phenomena of reënforcement and inhibition in the frog, but I know of no
detailed study of the significance of the temporal relations of stimuli
in this connection. Goltz[170] called attention to the inhibition of
the croaking reflex by peripheral stimulation, as well as to several
similar phenomena. Nothnagel,[171] Lewisson,[172] and Wydensky[173]
further contributed to our knowledge of the interference effects of
stimuli in the frog. Wydensky proved that the application of an induced
current to a nerve-muscle preparation may result in either contraction
or relaxation of the muscle, according to the frequency of stimulation.

More recently Merzbacher[174] has dealt with the influences of
complication of stimuli in the frog with the purpose of ascertaining
the relations of the sense-organs to the reflex movements of the
animal. His first paper is concerned especially with the functional
importance of the eye in connection with reflexes. Unfortunately for
the demands of this research, he did not attend particularly to the
temporal relations of his stimuli. That a visual and a cutaneous
stimulus were given either "at the same time or within a short interval
of one another" (p. 250) is not the sort of information our problems
demand.

According to Merzbacher's very interesting results a visual stimulus
reënforces the reaction to a cutaneous stimulus. As the results of this
paper show, this is only half a truth, for the two stimuli may either
reënforce or inhibit one another's reactions. As Merzbacher observed
no evidences of reaction to auditory stimulation he presumably did not
attempt to study the influences of the ear in connection with reflexes.

There can be no doubt that the words reënforcement and inhibition
as at present used in connection with the functions of the nervous
system cover a multitude of widely differing phenomena. We can at
once distinguish at least two important kinds of reënforcement or
inhibition: first, that which is due to the functioning of special
augmentary or inhibitory portions of the nervous system; second,
that which is the result of the complication of stimuli. Any and
every process in the nervous system may have either a reënforcing or
an inhibiting influence upon simultaneous or succeeding processes;
doubtless most processes or impulses at various times have both
effects. The nervous system is constantly being modified by impulses
from many sources, which suppress or strengthen one another according
to their relative intensity, their temporal relations, and the motor
relations of the portions of the organism which they affect.

The existence of the so-called refractory period in brain cortex and
nerve indicates that every stimulus causes certain fundamentally
important changes in the condition of the neural substance. These
changes we may for convenience of illustration describe as modification
of excitability, or of the functional capacity of central or peripheral
tissues. Every stimulus causes a portion of the neural substance
to pass from its normal state through a condition of increased
excitability, which we may designate the positive phase, to a condition
of diminished excitability, the negative phase. There is first an
increase in the functional capacity of the tissues, then a decrease.
If during the course of the change produced by a given stimulus a
second stimulus becomes effective its result in reaction is determined
by the particular phase of the tissues upon which it intrudes. If the
nervous system is in the condition of increased excitability, and
the two stimuli act upon sensory regions whose motor connections are
not antagonistic, the reaction will be reënforced, as we say, by the
previous stimulus; if, however, the second stimulus falls upon the
negative phase of the nerve substance, the reaction will be partially
or totally inhibited.

The facts which are most prominent as the result of this investigation
are, first, that the temporal relation of stimuli is an important
condition of certain forms of reënforcement and inhibition; second,
that the interference effects of two stimuli cannot be studied to
advantage without attention to the relations of the forms of reaction
which are appropriate to each stimulus.


V. SUMMARY

1. Motor reactions of the green frog to electric stimuli are inhibited
either partially or wholly by photic stimuli. The visual stimulus of a
moving object has a like effect. It has been found, furthermore, that
the same visual stimulus may either inhibit or reënforce the motor
reaction in response to electric stimulation. When the two stimuli are
given simultaneously reënforcement occurs, when the visual stimulus
precedes the electric by half a second or more inhibition appears.

2. An auditory stimulus, which does not produce any visible reaction
when given alone, modifies respiration and the reactions to other
stimuli when given in connection with them.

3. The momentary auditory stimulus of a quick hammer blow when
simultaneous with tactual stimulation reënforces the reaction to the
latter stimulus. This reënforcement, or increase in the amount of
reaction, ranges from 50 to 100% of the average reaction to the tactual
stimulus alone. When the auditory stimulus is given before the tactual
reënforcement occurs in gradually decreasing amount until the interval
between the two stimuli reaches .35″; at this point the auditory
stimulus has no apparent effect upon the tactual reaction. As the
interval is still further increased inhibition appears and continues
for intervals between .35″ and .9″. Reënforcement is greatest when the
two stimuli are simultaneous; inhibition is greatest when the momentary
auditory stimulus precedes the tactual by .4″ to .6″. When the interval
reaches .9″ the first stimulus does not affect the reaction to the
second.

4. Reënforcement is greater for the males than for the females;
inhibition appears sooner and lasts longer in case of the females.
This apparently indicates that the males are stimulated to activity by
certain auditory stimuli, whereas the females are rendered passive by
similar sounds.

5. Prolonged auditory stimulation by means of an electric bell causes
reënforcement and inhibition, according to the temporal relations
of the stimuli, as does momentary auditory stimulation, with the
following differences: The maximum reënforcement occurs when the
tactual stimulus is given about .25″ after auditory stimulation has
begun; reënforcement continues for a period of 1.2″, _i. e._, when the
electric bell continues to ring until the tactual stimulus is given, it
reënforces the tactual reaction from simultaneity to 1.2″. Inhibition
then appears, and continues until 1.8″. Both momentary and prolonged
auditory stimulation cause first reënforcement, then inhibition of the
appropriate reaction to a tactual stimulus.

6. The reënforcement-inhibition curves for the frog are very similar to
those for man.

7. In case of the several pairs of stimuli whose interference effects
have been studied reënforcement-inhibition appears. The first stimulus
reënforces reaction to the second so long as the interval between them
is not more than about .4″, while it inhibits the reaction when the
interval is longer. Whether this reënforcement-inhibition curve as
given in the experiments described may similarly be obtained for any
and every pair of stimuli, no matter what their relation to reactions,
remains to be determined.

8. In connection with the study of the mutual relations of stimuli of
which this paper gives an account certain facts concerning the sense
of hearing have been discovered. A summary statement of the results on
hearing may be found on page 551.

FOOTNOTES:

[Footnote 139: The results brought together in this paper have been
published in part in connection with other work in the following
papers: Inhibition and Reënforcement of Reaction in the Frog, Jour. of
Comp. Neurol. and Psychol., vol. 14, p. 124, 1904. Bahnung und Hemmung
der Reactionen auf tactile Reize durch akustische Reize beim Frosche,
Arch. f. d. ges. Physiol., vol. 107, p. 207, 1905. The Sense of Hearing
in Frogs, Jour. of Comp. Neurol. and Psychol., vol. 15, p. 279, 1905.]

[Footnote 140: The Organ and the Sense of Taste in Fishes, Bulletin
U.S. Fish Commission for 1902, pp. 237-272.]

[Footnote 141: Animal Behavior, Woods Hole Lecture Series, p. 300,
1899.]

[Footnote 142: The Response of the Frog to Light, American Journal of
Physiology, vol. 9, p. 476, 1903.]

[Footnote 143: The Instincts, Habits and Reactions of the Frog, Harvard
Psychological Studies, vol. 1, p. 590, 1903.]

[Footnote 144: The Skin and the Eyes as Receptive Organs in the
Reactions of Frogs to Light, American Journal of Physiology, vol. 10,
p. 31, 1903.]

[Footnote 145: Ueber die electromotorischen Erscheinungen an
Hautsinnesnerven bei adaequater Reizung, Archiv für d. ges.
Physiologie, vol. 63, p. 503, 1896.]

[Footnote 146: Harvard Psychological Studies, vol. 1, p. 592, 1903.]

[Footnote 147: Abstract of paper read before Section F of American
Association for the Advancement of Science in Philadelphia, 1904.
Science, vol. 21, p. 265, 1905. See also Bulletin of the U. S. Fish
Commission for 1902, pp. 45-64, and the same for 1904, pp. 183-207.]

[Footnote 148: Anatomie des Frosches, VI, Lehre von Integument und von
den Sinnesorganen, pp. 751, 752, 1904.]

[Footnote 149: Journal of Comparative Neurology and Psychology, vol.
15, pp. 279-304, 1905.]

[Footnote 150: Ueber die Beziehungen der Sinnesorgane zur den
Reflexbewegungen des Frosches, Arch. f. d. ges. Physiol., vol. 81, pp.
222-262, 1900.]

[Footnote 151: "Blendung oder blosse Lichtentziehung erhoht die
Erregbarkeit für mechanische Reize" (p. 253).]

[Footnote 152: Thousandths of a second.]

[Footnote 153: Arch. f. d. ges. Physiol., vol. 81, p. 227, 1900.]

[Footnote 154: A case of inhibition.]

[Footnote 155: Arch. f. d. ges. Physiol., vol. 107, p. 213, 1905.]

[Footnote 156: Journal of Physiology, vol. 9, pp. 60, 61, 1890.]

[Footnote 157: American Journal of Physiology, vol. 1, p. 336, 1898.]

[Footnote 158: Arch. f. d. ges. Physiol., vol. 68, p. 546, 1897.]

[Footnote 159: Travaux du Laboratoire de Marey, 1876.]

[Footnote 160: Arch. f. d. ges. Physiol., vol 59, p. 309, 1894.]

[Footnote 161: Zeitschr. f. Biol., vol. 32, p. 1, 1895.]

[Footnote 162: Journal of Physiology, vol. 21, p. 213, 1897.]

[Footnote 163: American Journal of Physiology, vol. 8, p. 213, 1902.]

[Footnote 164: Comptes rendus, vol. 124, p. 573, 1897.]

[Footnote 165: Nature, vol. 60, p. 629, 1899.]

[Footnote 166: Journal of Physiology, vol. 7, p. 101, 1886.]

[Footnote 167: Journal of Physiology, vol. 7, p. 111, 1886.]

[Footnote 168: Text-book of Physiology, p. 841, London, 1900.]

[Footnote 169: Arch. f. d. ges. Physiol., vol. 28, p. 495, 1882.]

[Footnote 170: Beiträge zur Lehre von den Functionen der Nervencentren
des Frosches, p. 41, Berlin, 1869.]

[Footnote 171: Centralb. f. d. med. Wissensch., vol. 7, p. 211, 1869.]

[Footnote 172: Arch. f. Anat., u. Physiol., p. 259, 1869.]

[Footnote 173: Arch. d. Physiol. norm. et pathol., vol. 4, p. 690,
1892.]

[Footnote 174: Arch. f. d. ges. Physiol., vol. 81, p. 222, 1900.]




THE TEMPORAL RELATIONS OF NEURAL PROCESSES

BY ROBERT M. YERKES


Muscle contraction-time, according to the determinations of several
investigators, varies about .0035".[175] Sanderson states that the
time for direct stimulation of the muscle is approximately .0035" and
for indirect stimulation, by means of the nerve, .007". The rate of
nerve-transmission in the frog ranges from 25 to 35 metres per second.

Reflex reaction-time, as might be expected, varies widely with the
nature of the reaction elicited by a stimulus, the condition of
the animal, and the quality and strength of the stimulus. For many
of the simple motor reactions of the frog it ranges between 20 and
60^{σ}.[176] Whether reflex reaction-time is to be sharply contrasted
with instinctive and voluntary reaction-times, or whether they
indistinguishably merge into one another is a question of considerable
interest and importance for the student of the evolution of activity.

Voluntary reaction-time may be as short as 150^{σ} or as long as life,
in an animal capable of profiting by experience as does the frog. It is
preëminently the delayed type of reaction-time.

So much concerning the temporal relations of neural processes in the
frog being well established, the purpose of the present paper is
to call attention to some experimental results which indicate the
existence of clearly defined types of reaction, and suggest possible
values of reaction-time as a sign of mind.

The specific problems to be considered are: (1) Do reaction-times, in
any given animal, range with equal frequency of occurrence from short
to long, or are there certain modes (most frequented classes) which
indicate definite types of reaction, such, for example, as the reflex,
instinctive, etc.? (2) If there is distribution of the reaction-times
about one or more modes, what are the types of reaction indicated
thereby? (3) Finally, is reaction-time of service as a sign or measure
of consciousness?

I wish especially to call attention to the fact that this paper deals
with the reactions of the frog, not with animal reactions in general.


REACTIONS TO ELECTRICAL STIMULATION AND TYPES OF REACTION

Two years ago in connection with a discussion of the reaction-time of
the green frog to electrical and tactual stimuli,[177] I presented a
curve showing the distribution of 277 reaction-times to an electrical
stimulus. The curve exhibited two clearly defined modes: one at between
60 and 70^{σ} and the other at about 160^{σ}. There was further a group
of delayed reactions ranging about 500^{σ}. This form of distribution
was interpreted, at the time, as indicative of three types of reaction,
called, respectively, the reflex, the instinctive, and the delayed.

I have since obtained and examined with reference to form of
distribution the further data which are presented in this paper. The
reactions are all those of the green frog to electrical stimulation.
The stimulus was applied by means of wires on the reaction-board
on which the frog rested during the experiments. When reaction
occurred in response to the electrical stimulus a circuit through the
time-measuring apparatus was broken by the release of a delicate spring
which had been held in place up to the instant of reaction by the
weight of the frog. A Hipp chronoscope, controlled by a Cattell falling
screen, served as a time-measuring mechanism. Three intensities of
stimulus were used: (1) A current from one Mesco dry cell, (2) from two
cells, and (3) from four cells.

Of the reactions whose time was measured there are three series. Series
I is constituted by the recorded reaction-times in response to a
one-cell stimulus, Series II, those in response to a two-cell stimulus,
and Series III, those in response to a four-cell stimulus. The number
of reactions, range and mode of each series are as follows:

              Number of reactions     Range               Mode
    Series   I        193          161-798^{σ}         235^{σ}
    Series  II        288           41-647             235
    Series III        256           61-178             105

The distribution of the 481 reaction-times of Series I and II is shown
by Figure 1; that of the 256 reaction-times of Series III, by Figure
2. For both of these distribution polygons the reaction-times were
arranged in ten^{σ} classes, beginning with the class 41-50^{σ}[178] in
the case of the combined Series I and II and with the class 61-70^{σ}
in the case of Series III.

Series I exhibits a primary mode at 235^{σ}. There are no reflex
reactions in this series, unless it be maintained that the reflex
reactions of the frog may have a reaction-time of over 160^{σ},
but there are a number of delayed reactions, some of which have
reaction-times as long as 798^{σ}. This intensity of stimulation
(one cell) may be said to call forth prompt reactions, which we may
provisionally call instinctive, and delayed reactions, which have all
the appearances of voluntary acts. There are no reactions which come
within the range commonly considered as the reflex range of the frog
(20-60^{σ}), and there are relatively few delayed reactions: almost all
centre about the mode 235^{σ}.

[Illustration: FIG. 1, LOWER. FIG. 2, UPPER.]

Series II, in contrast with Series I, exhibits a secondary
mode at 65^{σ} in addition to the primary mode at 235^{σ}. The
stimulus-intensity of this series (roughly twice as great as that for
Series I) induces a variety of short reaction, which did not appear in
the case of the one-cell stimulus, and at the same time fewer delayed
reactions. The range of the reaction-times for the two series is about
the same, but the lower limits are markedly different.

Observation of the subjects during the experiments revealed two methods
of reaction to the two-cell stimulus: a locomotor reaction (jump)
which at once removed the animal from the source of stimulation,
and a twitch of the hind legs which was instantly followed by the
above-mentioned locomotor reaction. The leg reactions constitute the
reflex group of Fig. 1, the usual prompt locomotor reactions, the
instinctive group, and the slow locomotor reactions, the delayed or
voluntary group.

It is to be noted that the instinctive reaction-time mode is the same
for the two intensities of stimulation. This apparently indicates
that change in intensity of stimulation causes a change in the type
of reaction, not merely a gradual change in the position of the
mode. For example, the modal reaction-time of 235^{σ} given by a
one-cell stimulus did not shift to 200^{σ} or lower, as might have
been expected, but instead there appeared a new type of reaction. The
average reaction-times for the two series indicate a decrease in time
with increase in intensity of stimulation, but they give no indication
of the really important difference in the two series of reactions. The
great importance of the distribution of the data, in addition to the
common statistical quantities, is manifest.

Series III, whose reactions occurred in response to a very strong
stimulus, differs in several important respects from the other series.
Its range is much narrower, only 117^{σ}. Delayed reactions are
lacking, and so also, curiously enough, are the reflex reactions of
Series II. Instead of either or both of the modes of Series II, there
appears in Series III an intermediate mode at 105^{σ}.

Our interpretation of these facts is facilitated by results of
observation of the reacting subject. The leg reflex which frequently
occurred in response to the two-cell stimulus never appeared in
response to the four-cell stimulus. This in part explains the lack
of the short reaction-time mode of Series II; it does not, however,
account for the lack of delayed reactions. The latter fact may be
referred to the intensity of the stimulus. Another difficulty in
interpretation appears in connection with the intermediate mode,
105^{σ}. Is this to be considered an instinctive mode, as were those
at 235^{σ}, or a reflex mode? Where is the line between reflex
and instinctive action to be drawn? These results very clearly
indicate that no line can be drawn, except quite arbitrarily. Reflex
reaction-time, in the case of the frog, is continuous with instinctive,
yet for any given situation the reflex, instinctive, and delayed
(voluntary?) modes are likely to appear, as, for example, in the case
of the data of this paper. Our conclusion must be, therefore, that
although types of reaction are indicated by reaction-time results,
the mode for a given type varies too much in position with different
conditions to make it possible to say that a particular reaction-time
is that of a certain type.

We may safely say, then, that for any given subject, the muscle
contraction-time, nerve transmission-time, and simple sensory
reaction-time to the constant stimulus in question being known, we
should be able safely to interpret reaction-time records in terms of
reaction types. For reflex, instinctive, and voluntary are terms which
designate modes of reaction, albeit not isolated classes, for they
intergrade.

Whether there are more types of reaction than are indicated by the
data of this report does not concern us at present, for the practical
as well as the theoretical bearings of our conclusions depend upon the
existence of types, and not upon their number.


REACTION-TIME AS AN INDICATION OF CONSCIOUSNESS

Hesitation in reaction is commonly accepted as an important sign of
volitional consciousness in man; consequently delayed reactions in
lower animals are supposed to be indicative of psychic processes.
Granting this much, reaction-time may be used as a sign of
consciousness. It cannot be denied that the longer the reaction-time
of a given animal the greater the probability that the reaction is
conditioned by mental processes. Such a statement, it is true, has a
basis neither better nor worse than that of most of our inferences
concerning the nature of the actions of our fellow beings. As I have
already attempted to show in a discussion of criteria of consciousness
in animal psychology,[179] there is no one criterion of consciousness
which can be used alone satisfactorily, but instead there are numerous
signs of mind each of which has value according to the number and
variety of our observations concerning its occurrence in connection
with states of consciousness. The more of such signs we discover
and learn to evaluate properly in relation to consciousness in its
different grades and to one another, the safer will be our inferences
concerning the existence of mental processes in animals.

Reaction-time is presented in this paper as an additional sign of
mind. Like all other signs it is of value only if used as one of a
series of indications of mental life. For if we attempt to judge of
consciousness by reference to reaction-time alone, we may be seriously
misled, whereas if we use it in connection with docility, variability,
neural specialization, and other recognizedly valuable signs, we may
be greatly aided in our inference. As in juristic procedure judgment
is not based upon one bit of evidence nor even upon the evidence of
a single witness, but upon evidence accumulated from all available
sources, so in our attempts to judge of the existence of consciousness,
it matters not whether the being be human or infra-human, we should
make use of all phenomena which are recognized as signs of mind. The
chief task of comparative psychology at present is the discovery and
evaluation of signs of mind.

Reaction-time data, however, furnish another sign, or, as I prefer
to call it in this case, measure of the intensity of consciousness;
for variability of the time of reaction as well as its duration is
significant. Reflex reaction-time is relatively constant, instinctive
varies considerably, and the variability of voluntary reaction-time is
extremely large. Degree of variability of reaction-time may be used
as an indication of consciousness in the same way that variability in
the form of reaction is used. The higher the power of consciousness
the greater the variety in form of reaction and the variability of the
reaction-time.

Reaction-time studies, as well as introspection and the investigation
of animal behavior, indicate the importance of three activity concepts:
automatism, instinct, and will. The automatic act is quick and
relatively constant in form as well as reaction-time, while all signs
lead us to infer that consciousness, when it accompanies the act, is a
sequent phenomenon and not a condition of the act. The instinctive act
is both slower and more variable in form and time than the automatic:
consciousness is indicated as an accompaniment, and apparently it is
at times a condition of the act. The will-act is extremely variable,
unique in form, and almost without limits of reaction-time, for the
conscious organism may react to the present situation in a fifth of a
second, a day, or a year. Will is experience in action: it is our name
for individually acquired control, and voluntary action is above all
consciously conditioned activity.

Reaction-time, with respect to its two aspects of duration and
variability, may be used as a sign or criterion of consciousness, for
in accordance with the nature of these two sets of facts we classify
acts as reflex, instinctive, or voluntary.

FOOTNOTES:

[Footnote 175: Sanderson: Journal of Physiology, vol. 18, p. 147.
Tigerstedt: Archiv f. Physiologie, p. 111, 1895. Boruttau: Archiv f.
Physiologie, p. 454, 1892.]

[Footnote 176: ς = thousandths of a second.]

[Footnote 177: Yerkes: Harvard Psychological Studies, vol. I, p. 609,
1903.]

[Footnote 178: The last four classes of Fig. 1 are 100^{ς} classes,
401-500, 501-600, 601-700, 701-800.]

[Footnote 179: Yerkes: Journal of Philosophy, Psychology, and
Scientific Method, vol. 2, p. 143.]




THE MENTAL LIFE OF THE DOMESTIC PIGEON

AN EXPERIMENTAL STUDY OF CERTAIN EMOTIONAL AND ASSOCIATIVE PROCESSES

BY JOHN E. ROUSE


I. INTRODUCTION

Naturalists have observed the habits of pigeons, and physiologists
since Flourens have subjected them to numerous experiments, but so
far they seem to have received little psychological study. As a
contribution to this interesting field the present paper reports an
investigation of certain emotional and associative processes of the
domestic pigeon. Since the literature of the subject is meagre, I shall
state at the beginning a few related facts which I have gathered from
various sources; then I shall discuss in detail the problems, methods,
and results of my several experiments.

The brain of the pigeon is well developed, although the hemispheres
are unconvoluted. When they are removed, the animal retains unaltered
its reflex and vital activities, but ceases for a time at least to
show evidence of mental life, for example, memory and will.[180] In
the normal animal sight and hearing are acute, and touch seems keen,
although the claws are not used for grasping and eating, as in the case
of more intelligent birds, especially, parrots. There is considerable
sensitiveness to temperature changes. Taste, and probably smell, appear
to be deficient.[181] The "sense of support" is marked, even in the
young.[182]

Since the pigeon seems to dream and also to miss its absent mate, some
observers believe that imagery is present. There is certainly local
memory, and also capacity to observe. Various intelligent acts have
been reported.[183] The remarkable homing habits of the carrier pigeon
have received no satisfactory explanation. While Cyon[184] suggests the
stimulation of the nasal organs by air currents, Thauzièr[185] holds
to the electrical theory; they agree, however, that certain higher
psychical processes are probably involved.

Graber's[186] tests indicate that pigeons have no color-preference.
Beebe's[187] statement concerning birds in general is peculiarly true
of pigeons: "There are few species which do not show the emotions of
love and sympathy, and ... one will sometimes pine and die of grief at
the loss of its mate." After referring to their patient care of the
young, he adds: "Indeed, sympathy is the keynote in the development
of the higher mental faculties." These birds communicate, but their
language consists of comparatively few sounds. As in many other birds,
the play-instinct is highly developed.


II. PROBLEMS AND METHODS

My study of the pigeon's _emotional life_ had for its object certain
respiratory "expressions." These were investigated by means of a
pneumographic tracing, secured while the animal was comfortably
fastened in a shallow nest, partially open below. A small box was
placed over the bird, and apparatus was so arranged that the time
of giving various stimuli was recorded automatically on the smoked
paper of the kymograph drum, below the breathing-curve. A third line
indicated rate of drum movement. Although some interesting results
were obtained, the chief significance of the research consists in its
demonstration of the fact that this method of studying animal mind is
valuable.

In the study of _association_ I sought to determine the sense-data
which the process involves, its method of formation (with due regard
to social conditions), its rapidity, permanence, and modifiability,
and also its probable degree of complexity. Material contributing to
the subject was secured by observing the behavior of the animal when
seeking to obtain food by overcoming such obstacles as labyrinths with
wire passages, and latches, when the food was left in view, or by
finding it when out of sight. In the latter case it was placed in a box
occupying a customary place in a group of exactly similar boxes, or
else in a box of color or form unlike the other members of the group
and variously arranged, from time to time, with respect to them. When
the animals were learning the labyrinth habits, various stimulations
were given them; later the character of some of these was altered, and
the resulting changes in behavior were noted. After the habits had
been thoroughly learned, the birds were given a rest for some weeks,
and then tested again under the old conditions. A few trials were
arranged with special reference to the study of imitation; the animals
here were tested as to their ability to execute simple but unfamiliar
acts, after having only seen them performed by an animal previously
trained. Throughout the associational tests the animals very seldom
received food in their cages; but as they were tested daily and allowed
to satisfy their hunger completely at the last test, they were never
in a state of "utter hunger"--a condition which most experimenters
think best to avoid. A series of tests, given at the conclusion of the
investigation, indicated that the odor of the food had not assisted the
animals in reaching it.

In the two series of experiments (emotion and association) thirty-five
animals in all were used. They were confined in large cages in a fairly
well lighted and ventilated room, and were fed wheat, cracked corn,
and occasionally fruit, and kept well supplied with fresh water and
sand. They generally remained in a healthy condition throughout the
tests, especially during the winter. To exclude, as far as possible,
the disturbing influence of fear, they were usually handled only
after the room had been darkened. As the noise made by the curtains
was objectionable, the birds were tested with the room illuminated by
incandescent lamps; the light was turned off before the birds were
placed in position for the trials, and again before they were removed
from the apparatus to the cages. It is generally agreed that an
experimenter should be out of sight when giving a test. I am convinced
that it is important to avoid being seen by the animals at any time.
This involves great inconvenience, especially when one employs the
pneumographic method, but better results are thus obtained.

For practical suggestions as to apparatus and methods I am
greatly indebted to Dr. Robert MacDougall, at the beginning of my
investigation, and to Dr. Robert M. Yerkes, throughout. I also owe
much to the researches of Zoneff and Meumann,[188] Thorndike,[189]
Mills,[190] Small,[191] and Kinnaman.[192] Porter's[193] interesting
study of sparrows was made almost simultaneously with the investigation
here reported. Fewer animals were used by him, but in some instances
more tests were given.


III. INVESTIGATION OF EMOTION[194]

1. _Respiration in general._ The normal breathing-curve in pigeons
is quite similar in contour to that of the human subject, although
the rhythm is more rapid and the pauses are less pronounced. When
acoustical, visual, olfactory, or tactual stimuli are given, various
modifications appear, for example, quickening, deepening, and minor
irregularities. It was noticed that meaningless stimuli (pistol-shots)
quickly lose their disturbing influence, whereas the breathing remains
sensitive to those of a significant character, such as the noises
made by other birds. It was also found that a stimulus which no
longer affects the breathing will sometimes occasion disturbance if
accompanied by a second stimulus of another order, although of a weak
intensity (summation).

2. _Respiratory reactions to light._ As the easy control of conditions
makes vision an excellent field in which to work, light reactions were
investigated in detail. Two distinct series of tests were given. One
sought to determine the relation between quality of light and reaction;
the other, between intensity of light and reaction. Four colors of one
intensity and three intensities of one color, respectively, were used.
In the first series four stimuli, one for each of the colors, red,
yellow, green, and blue, were given daily; in the second series five
daily stimuli were given, of the same intensity for any one day, and
one minute apart; this made it possible to observe also the effect of
repetition. Each stimulus was given at the beginning of a respiration
and continued two seconds. When the tracings were studied, various
modifications were noted, but special attention was paid to alterations
in rate of breathing. In the case of both sets of trials an immediate
quickening usually occurred after stimulation, and occasionally
shallowing[195] and minor irregularities of contour.

In the first set of tests ten animals were used for twenty-five days.
Average results indicated that red and yellow are less stimulating
than green and blue. To secure data that would assist in the
interpretation of these results, an investigation was made of the
animals' color-preference. This was done by recording, at thirty-minute
intervals, the position of the birds when confined, singly, in a
box one half of which was illuminated (from the side) by light of
one color, and one half by light of different color but of the same
intensity. A water screen excluded the heat rays. After nine records
had been taken the colored glasses were interchanged, and the animal's
position relatively to the two colors was observed as before. This was
repeated with the other colors until each of the four had been used
with each of the other three. There were far more choices of green and
blue than of red and yellow, though none of the colors was avoided.
It seemed a question of _degree of liking_, rather than of liking or
disliking. As stated, Graber's experiments indicated that pigeons have
no color-preference, but his results are probably untrustworthy, since
he tested several animals at once and apparently was not careful to
change the colored glasses regularly. Putting together our two sets of
data (the latter stated first) we have the following comparison:

                           R      Y       G       B
    Color choices of }
     5 animals       }    72     129     167     172

    Breathing-rise of}
     10 animals      }   9.94%  10.39%  10.41%  12.11%

Although the proportions do not hold, there is a direct correspondence
between the two series of responses; hence it would seem that
_increased respiratory activity is an expression of agreeable feeling_
in pigeons, and this especially since the breathing, when varying at
all in amplitude, usually became shallower, and also showed certain
minor irregularities of contour, as often occurs in human respiration
during moderate stimulation of a pleasant character.[196]

In the second series of respiratory tests four animals were used
for fifteen days. Average results showed nothing as to the relation
between intensity of stimulus and amount of quickening, since the three
intensities used, 1, 2, and 4, produced reactions, respectively, as
follows: 6.6%, 4.3%, and 6.4%. This may have been because the three
intensities were employed each on different days. When the reactions
are averaged according to daily succession, without regard to the
intensities of the stimuli, we get the following results: first
reaction, 8.0% rise in rate; second, 3.7%; third, 4.1%; fourth, 5.7%;
and fifth, 6.9%. We should have expected the second daily response to
be less vigorous than the first, since the animals were perhaps better
prepared for the second stimulation. That the reactions increased
thereafter was probably due, partially to summation, and partially
to the fact that the short illuminations occasioned mental action
(_perception_ of interior of box, increased _desire_ to escape,
etc.) which involved heightened, rather than depressed, breathing
activity, and thus worked directly against the dulling tendency of
repetition.[197]


IV. INVESTIGATION OF ASSOCIATION

1. _Labyrinth experiments._ Four labyrinths in all were used (L, M, H,
O). Each was constructed by attaching moveable wire partitions in a
wooden box, covered with chicken wire. The pigeon was admitted through
a small entrance compartment which was fastened at one end of the box,
and which communicated with it by means of a lifting door, operated by
pulling a cord from behind the observation curtain. Food was placed
within the maze, and usually at the opposite end. Before beginning the
tests the bird was allowed to become thoroughly familiar with the box
without the partitions. After a few trials it learned to go to the food
immediately upon entering the box. The partitions were then put into
position, and the bird was tested as to the time required (except in
the case of labyrinth O) and as to the method employed in reaching the
food. The time was measured by means of a stop-watch, and the bird's
horizontal movements were recorded on a small plot of the labyrinth;
other general observations were added.


A. _Habits in Labyrinth L_

[Illustration: FIG. 1. Labyrinth L. BB, box 6 in. high; E, entrance;
F, feeding-place; a, b, c, and d, moveable partitions 6 in. apart; NN,
edge of pigeon cage; D, observation screen. The lamps which illuminated
the room hung directly above the apparatus.]

In this labyrinth (Fig. 1) six animals were tested once daily for
thirty days, and five of these again after two and six weeks,
respectively. On entering the labyrinth with the partitions in place
the first time, a bird started on its usual direct course toward the
food-box; running against the first partition it made vigorous efforts
to push through, flying at the wire and often clinging to it for a
short time; some of these random movements eventually brought it to
the left of the compartment, and thence, through the opening, into
the second compartment, and so on through the others, until finally
it reached the food by a series of fortunate accidents.[198] The same
general reaction was shown in case of the next few tests, except that
fewer and fewer useless movements were made, and that the right ones
were carried out with greater and greater precision. Later the animal
had no difficulty in reaching the food; it did not run against the
partitions, enter the blind alley, nor display such general signs of
uncertainty as pausing and looking about. The process of learning in
this case was obviously one of "trial and error," or the selection
of useful movements. From the mass of random movements constituting
the reaction to the unfamiliar environment, only those which enabled
the bird to reach the food were retained and improved; the others
gradually disappeared until finally the path taken became the shortest
one possible, and was entered upon and pursued without hesitation by
each animal as soon as it was allowed to enter the labyrinth. The time
required for the tests is given in Table I.

It will be seen at a glance that the absolute time required for
reaching the food varied for the individuals (see especially the
results given by different birds in the case of test 1), but that the
several periods for any one bird were relatively similar to those
for another; and also that the time was long at first, but rapidly
shortened from test to test, thus showing a steady advance in the
learning process. Various lapses occurred (for example, A, 8; C, 13; E,
10) after the habit had been fairly well fixed.

In tests 18-22 the time-shortening was due principally to quickening of
movements which had already become well defined. The great importance
of visual data is brought out by the abrupt lengthening of the periods
in the case of tests 23-25, and 26-30, where the light intensities
were decreased. The lengthening was roughly proportional to the change
of illumination. In the relative darkness the birds had to re-acquire
the habits. The same mistakes were made as at first (running against
partitions, and into the blind alley), yet here, as before, there was
a ready adjustment. That the food was out of sight, or at least very
much less visible, probably made no difference, since it was found
that the birds would readily go to the old place after both food and
food-box had been removed. In order to exclude the light entirely
without making their movements invisible to me, I blindfolded the birds
by means of a thin black hood, comfortably adjusted over their eyes and
top of head; as a result, none was able to make the course in twenty
minutes. The first turn, however, was usually made naturally, perhaps
because associated with certain non-visual sense-data (sound of the
lifting door, and perhaps tactual impressions of the close entrance
compartment, etc.). Rats[199] seem far less dependent upon visual data
than do pigeons. The great permanence of the pigeons' habits is shown
by comparing the periods for tests 31-3, given after two and six weeks
of rest, respectively, with those for tests 18-22.


TABLE I. TIME REQUIRED TO REACH FOOD IN LABYRINTH L

                                _Animals_

    _Trials_,   A       B       C       D       E       F    _Average_
    1 daily.
               ' "     ' "     ' "     ' "     ' "     ' "      ' "
          { 1)   28:50     :59   42:20   49:04   22:13    4:04    24:35
          { 2)    7:22     :22   25:47   10:17     :48    2:02     7:46
          { 3)    1:18     :12    8:29   12:35     :12    1:41     4:05
          { 4)     :32     :21   10:51    1:26     :19     :52     2:24
          { 5)     :24     :28    2:36    2:18     :12    1:33     1:15
          { 6)     :25     :26    1:10     :55     :12    1:50      :50
          { 7)     :15     :24     :28     :32     :15    2:09      :41
          { 8)    1:05     :23     :33    1:19     :17    1:46      :54
    1    {  9)     :16     :24     :57     :58     :10     :26      :32
          {10)     :24     :32    1:15     :51    2:12     :31      :58
          {11)     :12     :21    1:40     :30     :17     :54      :39
          {12)     :16     :32     :49    1:34     :22    1:18      :49
          {13)     :13     :18    2:30     :18     :10     :42      :42
          {14)     :29     :32     :27     :31     :25     :36      :30
          {15)    1:00     :24     :30     :31     :12     :35      :32
          {16)     :19     :52    1:10     :22     :17     :24      :34
          {17)     :14     :14     :31     :39     :13     :57      :28

          {18)     :13     :09     :29     :13     :16     :29      :18
          {19)     :10     :10     :36     :26     :07     :14      :17
    2    { 20)     :11     :15     :34     :17     :07     :10      :16
          {21)     :13     :14     :34     :16     :09     :21      :18
          {22)     :09     :16     :26     :14     :08     :11      :14

          {23)     :12     :42    1:29     :39    5:53     :13     1:31
          {24)     :20     :17    1:31     :33     :15     :13      :31
    3    { 25)     :15     :24     :40     :28     :19     :20      :24

          {26)    1:21   16:29    1:22   13:54    3:51    1:36     6:26
          {27)    3:36    4:45     :44    1:03    1:09    2:59     2:23
    4    { 28)     :51    1:24     :46    1:04     :56    1:09     1:02
          {29)     :51     :41    1:10     :40    2:17     :11      :58
          {30)    2:04     :18     :41     :14     :07     :19      :37

    5    { 31)     :08     :33     :25             :07     :07      :16
    6    { 32)     :09     :15     :20             :08     :31      :17

1: With 18-candle-power illumination of the room.

2: Same illumination; tests given after the animals had heard four
other pigeons pecking in the labyrinth.

3: With 2-candle-power illumination, other conditions the same.

4: With a slight illumination through single curtain, other conditions
the same.

5: After two weeks' rest, conditions as in 2.

6: After six weeks' rest, conditions as before.

Let us now notice the gradual progress of learning in three important
parts of the maze, as shown in Fig. 2. It will be seen that in the
beginning the animals started upon their usual course and pressed
against the first partition (stage 1), but that later they touched
it less and less (stages 2 and 3), and that finally they avoided it
entirely (stage 4). The adjustment here was fairly simple: the sound
made by the opening of the entrance door, and the glimpse thus given
of the labyrinth, gradually came to be conditions of the movements of
turning to the left, on emerging from the entrance, and passing along
the compartment toward the opening, where impressions, mostly visual,
in the same manner determined the movements of turning to the right and
entering compartment 2.

The blind alley was naturally a decided obstacle. The pigeons
learned to avoid this compartment by going around it only after many
unsuccessful attempts to go through it. During the first test the
animals entered it many times (stage 1); on emerging they returned to
the second or the first compartment, only to encounter the pen again
when they re-advanced toward the food; finally, on reëmerging from the
annoying enclosure, perhaps for the eighth or tenth time, they might
happen to turn to the right instead of going forward as usual toward
the entrance of the box, and thus make their way along the new passage
and reach the food. For the next few tests they usually entered the
blind alley, but less frequently, and they remained for shorter periods
(stages 2-4). Later they merely entered (stage 5); and still later they
passed very near the opening without entering, or only paused a moment
before it (stage 6); and finally they passed it without the slightest
hesitation, walking briskly, but with well-directed movements, midway
between the partitions (stages 7-8). The act of turning seemed to be an
especially important factor in this habit. We notice that it was a turn
to the right (most probably accidental) which first enabled the animals
to get beyond the opening of the blind alley; that this same act was
repeated in each successive trial until, by the gradual shortening of
the loop forming the path taken by the animals in passing into, and
from, the labyrinth it was finally reduced to a mere pause (stage 6);
and that this later disappeared entirely, leaving only the left turn,
which instead was now conditioned by the visual data at that part of
the labyrinth and carried the animal past the entrance of the blind
alley.

[Illustration: FIG. 2. Stages in learning Labyrinth L. The curves
indicate the pigeon's horizontal movements. Pauses are represented by
heavy dots.]

The animals did not come in contact with the second partition until
they had almost learned to pass the opening of the blind alley (see
stage 6). This was probably because the turn to the left which was
made on approaching _x_ was associated with visual data derived from
points farther along the course (_y_), and when the animals reached
_z_, compartment 2, these same data were received and were sufficient
to occasion the turn to the left there also, thus bringing the birds
against the partition. The adjustment was made principally on the basis
of new sense-data arising from running against the wire, looking at it
more closely, etc. For a few trials the birds made the turn at _x_ too
quickly, and thus failed to reach the third compartment. One of my most
intelligent subjects made this mistake in the third test, and again in
the sixth and seventh, and retained the act almost unchanged through
the tenth, eleventh, twelfth, thirteenth, fourteenth, sixteenth,
seventeenth, twentieth, and twenty-first tests, so strong was the
tendency to continue a movement once begun, though it was really
disadvantageous.[200]

[Illustration: FIG 3. Labyrinth M. E, entrance; F, food-box; height of
large box and width of passages the same as in Labyrinth L.]

After reaching the food and satisfying their hunger, the animals
often returned to the maze passages, seeking an exit; but they never
"explored" passages or showed other evidence of "free curiosity"
and "desire to know all their new surroundings," as Small reports
concerning rats.[201]


B. _Habits in Labyrinth M_

Five of the animals previously used were next tested twice daily,
forenoon and afternoon, for five days, in a larger, more complicated
maze. It had two blind alleys, and the food-box was near the centre
(see Fig. 3). The animals' general behavior was similar to that before
observed. The periods are given in Table II.


TABLE II. TIME REQUIRED TO REACH FOOD IN LABYRINTH M

                                _Animals_
    _Trials_,    A        B        C        E        F       _Average_
    2 daily.    ' "      ' "      ' "      ' "      ' "        ' "
     (1)       16:25     2:55     6:33     3:26     4:11       6:42
     (2)         :55     4:10     2:24     3:36      :23       2:18
     (3)        1:12      :55     8:27     9:06     2:07       4:21
     (4)         :48      :44     2:31      :34     1:04       1:08
     (5)         :27      :16      :14      :16      :14        :17
     (6)         :18      :32      :25      :15      :27        :23
     (7)         :14      :11      :31      :44      :30        :26
     (8)         :12      :16      :57      :16      :48        :30
     (9)         :12      :18      :23      :16      :16        :17
    (10)         :10      :19      :16      :15      :28        :18

Although this maze was much more difficult, it will be seen that
the animals learned the route to the food far more readily than
before. The first period in this series was only about one fourth as
long as the first period in the other, and the course was mastered
sooner (by the fifth trial instead of by the ninth). There was less
pressing against the wire than before, and unsuccessful movements were
sooner discontinued. This improvement was probably due entirely to
experience gained in dealing with the first maze. Thorndike speaks of
the gradually increasing ability of animals to deal with successive
contrivances.[202] The average results given in Tables I and II are
plotted in Fig. 4, next page.


C. _Habits in Labyrinth H_

Since hearing is an important sense in pigeons, we should expect
them to be capable of useful acoustical associations. Several things
occurred in the course of the two preceding experiments which seemed to
indicate that this is true; for example, although I could move about,
rather noisily, in the darkened room, without apparently disturbing any
of the birds, some few showed signs of fright (moving about restlessly)
on hearing the low, grating noises made by lifting the hanging door of
the cage, sounds which had always preceded the handling of the subjects
before experimentation, and which had probably become signs to them of
being taken.

[Illustration: FIG. 4. Learning-curves: A, in Labyrinth L; B, in
Labyrinth M. Divisions of ordinates indicate minutes; of abscissas,
successive daily trials. The rise on curve A for tests 23 and 26 was
due to diminution of illumination of room.]

To investigate this kind of association I constructed a labyrinth (see
Fig. 5) in connection with which sounds could be utilized as one form
of sense-data. The passages were so arranged that along the route
leading to the food there were three blind alleys, which the animals
would surely enter before mastering the course. In another part of the
room were placed, very close to each other, two electric gongs of the
same size, but of different material. One was of metal and gave a clear
ringing sound; the other was of wood and gave a low rattling noise.

When an animal was learning the route (as in the other two mazes) I
sounded the gongs, the metallic, as the bird approached and entered
the blind alleys, by openings _M_, _O_, and _R_, and the wooden,
as it emerged from them and proceeded along the proper course, and
occasionally after it had reached the food. The ringing sound was also
given after the animal passed _P_ and was approaching _Q_. When the
new route was fairly well learned, I changed the order of the sound
stimuli, ringing one gong at the places where the other had previously
been sounded, and compared the records thus obtained with those
obtained when the sounds were given in the original order. As an animal
is liable to become confused by the sounds, or else quickly accustomed
to them, I thought it best to give only a few trials, one trial with
the usual order of sound stimulations, the next immediately following
with the reversed order, and so on till four pairs of records had been
secured, the series of trials being completed in a single day. Four
animals were thus tested. The periods of the various trials are shown
in Table III.


TABLE III. SOUND ASSOCIATION, LABYRINTH H

_Time required to reach food under different sound conditions_

                 I
    _Order of gongs the same as when_
    _course was being learned._

           _Animals_
    _Trials_ A   B   E   G
             "   "   "   "
    (1)     13  19  24  17
    (2)     16  12  13  16
    (3)     10  16  20  14
    (4)     12  19  10  12

    Total,  51  66  67  59  "
                          243

                II
    _Order of gongs reversed._

           _Animals_
    _Trials_ A   B   E   G
             "   "   "   "
    (1)     14  16  18  27
    (2)     37  17  16  27
    (3)     14  19  26  11
    (4)     27  22  21  14

    Total,  92  74  81  79  "
                          326

In the case of thirteen of the sixteen tests given with reversed
acoustical conditions (see column II) the periods were longer than the
corresponding ones given alternately with them for comparison, and
there was an average time-lengthening of 5.2 seconds per trial, or
34.2%. The following is a short description of the animals' reactions
to the changed conditions. It corresponds to the time-values expressed
in column II of the table.

Bird A: test 1, animal undisturbed; test 2, drew back from _S_, turned
to the left and went toward _R_, but later returned and passed _S_
without pausing; test 3, paused at _O_ for a short interval, but did
not enter the blind alley; test 4, paused at _O_ again, later drew
back from _S_, turned to the left and entered blind alley 3; it soon
escaped, and this time passed _S_ without being disturbed, although it
paused at _T_ and _U_.

Bird B: tests 1 and 2, animal apparently undisturbed; test 3, a few
slight pauses at openings; test 4, drew back from _S_, entered blind
alley 3, but soon escaped and passed _S_ without hesitation.

Bird E: test 1, undisturbed; tests 2 and 4, paused at openings; test
3, turned back from _S_, entered blind alley 3, and paused at several
places later when passing toward F.

Bird G: test 1, many pauses; test 2, turned from _S_ and entered blind
alley 3; test 3, undisturbed; test 4, drew back from _S_, went toward
_R_, but did not enter the blind alley.

As the animals gave little attention to the wooden gong, but were
always sensitive to the metallic one, their observed movements probably
must be accounted for chiefly on the basis of certain visual and
organic sense-data _now with_, and _now without_, the ringing sound.
The data governing the start (as already noticed) were probably
sufficient for the avoidance of the first blind alley when the gongs
were reversed. In case of the other two, however, the birds had come
to depend upon acoustical data, and when these were lacking as they
approached the openings _O_ and _Q_, the left turn could not readily
be initiated, hence certain hesitations and misdirections of movement
frequently occurred. Experience with the blind alley in the first
experiment assisted the animals in dealing with the second blind alley
here, but mistakes were made. Visual data usually were sufficient to
produce the proper turn at _Q_, but when the ringing sound was given
just afterwards, it sometimes occasioned the left turn, thus bringing
the animals toward the opening of the blind alley. While the tests
given were not such as would indicate how far pigeons can discriminate
sounds, they certainly show that these birds are capable of useful
sound associations, although visual ones are evidently of greater
importance to them.


D. _Habits in Labyrinth O_

I next made tests in which tactual and electrical sense-data could also
be utilized. In one of the passages of a simple labyrinth was placed a
board 8 in. square and 3/4 in. thick, over which were stretched copper
wires which formed a series of interrupted electrical circuits. By
closing a key a bird could be stimulated whenever it stepped upon the
wire surface. A second key was connected with a metallic gong. When an
animal on its way through the maze first stepped upon the wire surface,
electrical and acoustical stimuli were given; later it was allowed
to walk across the board without being thus stimulated; afterward
acoustical stimuli were given it at various parts of the maze.

[Illustration: FIG. 5. Labyrinth H. For study of acoustical
association. Passages 6 in. wide, as before.]

Eight animals were used. All were found quite sensitive to the
electrical shocks, and when next tested they avoided the board,
especially if the gong sounded as they approached. Some would show
signs of uneasiness anywhere in the maze on hearing the gong. When
the board was so placed that they had to pass over it in reaching the
food, _when once on it_ they moved very leisurely, often lingering;
and if they stepped upon the wire surface in the darkened maze, they
showed no evidence of being frightened. Evidently no association had
been formed between the peculiar tactual stimulus of touching the
wires and the electrical shocks which had at first been given. But the
tactual stimulus may have been below the threshold. Yerkes[203] saw
evidence of association of this kind in the frog; this animal, however,
is probably much more sensitive to tactual stimulation received from
surfaces over which it passes than is the pigeon.

The results of these four experiments indicate that the pigeon easily
acquires complicated labyrinth habits; that these remain fixed for some
weeks at least; that acoustical, visual, and certain organic data are
the most important sensory factors; and that the process of learning
is one of "trial and error," in which the animal comes to form such a
close connection between the sense-data of the interior of the box and
those other sense-data arising from movements involved in reaching the
food, that when the box impressions are again encountered the other
sense-data are revived and readily condition the proper movements.
How much memory of eating was involved in these tests cannot be told;
but it was certainly not an essential part of the mental act.[204]
Proper guidance throughout the course was the main thing, and this was
determined by definite sense-data. That recognition, discrimination,
and perhaps choice were to some extent present seems likely from the
animal's hesitating movements at certain critical points. Thus it is
highly probable that when the bird approached a blind alley which it
had always entered before (see Fig. 2, stage 6), two alternatives
were recognized, to enter, as before, or not to do so, as was usual
thereafter, and that the pause had for its mental correlate a state
closely bordering upon what in us would be deliberation.

2. _Release experiments._ Under this heading are included certain cage
experiments in which some act, such as touching a lever, pecking, or
stepping upon a platform, resulted in the opening of the door, and thus
enabled the animal to escape and secure the food lying in view without.
The animal was admitted to the cage through an entrance compartment as
in the case of the maze trials. Before being tested it was allowed to
become familiar with the cage and to reach the food directly by passing
out through the open door. When first in the cage the animals did not
seem to notice the release apparatus, and hence they probably did not
begin learning the method of escape until later when they entered the
cage and found the door closed, and the ordinary exit thus obstructed.


A. _Latch Tests_

The cage here employed was an 18-inch cubical box. The top was of
chicken wire and the bottom and three sides of heavy boards; the fourth
side was formed by narrow vertical bars and a wire door which opened
inwardly and was held by a latch working on the outer surface of the
bars. At first a long wooden latch was used, which the animals raised
when reaching out for food. As this seemed an unnatural act, downward
pressure was substituted by attaching to the latch, now made smaller
and of brass, a string which ran over a pulley above the door and down
into the cage. As nooses did not seem adapted to the birds, the end of
the string was attached to a wooden lever which worked on the inner
surface of the bars, about three inches from the floor. Eight animals
were tested four times daily (twice in the forenoon and twice in the
afternoon) for ten days. The time required to escape and the animals'
behavior were recorded as in the case of the labyrinth tests.

[Illustration: FIG. 6. Curve of learning to operate latch, plotted from
last column of Table IV. Each vertical division indicates 30 sec. The
horizontal divisions represent successive days.]

When they first entered the box (singly as in the other experiments)
and found the usual exit closed they made various attempts to push
through between the bars, springing and often flying about with great
force and persistency. In course of their random movements they touched
the bar and opened the door and thus escaped. Later the unnecessary
movements were mostly dropped and the necessary ones became highly
specialized. The first association was established between the box
impressions received on entering and the movements involved in
approaching the front of the box and depressing the lever; later a
connection was formed between the sensations of touching the lever, of
hearing the sound of the opening door, of feeling the jar, etc., and
the movements of turning away from the lever and passing out. The sight
of the opening door seemed to be of less service to the birds than the
sound and jar. Each animal soon came to touch the bar at the point of
least resistance, and usually with considerable precision. The time
required by the several birds is shown in Table IV, next page. The
daily average results are plotted in Fig. 6, above.


TABLE IV. TIME REQUIRED TO ESCAPE FROM CAGE BY USING LATCH

                             _Animals_
    _Trials,_                                               _Daily_
    _4 daily._ A    B    C    E    F    G    H    I  _Av._ _Average_
              ' "  ' "  ' "  ' "  ' "  ' "  ' "  ' "   ' "    ' "
      (1)     :03  :06 3:20  :08  :03 1:10 4:30  :23  1:13
      (2)     :05 1:10  :59 1:50 2:00  :10 2:46 1:00  1:00
      (3)    5:28  :29 1:04  :21  :29  :52 2:33 4:05  1:55
      (4)     :31 1:45 3:52  :14  :06  :23 2:40 2:30  1:30   (1:25)
      (5)     :15  :10 2:25  :03  :10  :07  :31  :50   :34
      (6)     :17 2:00 5:03  :03  :05  :15  :39  :10  1:04
      (7)     :08  :20  :36  :31  :03  :11  :30  :46   :23
      (8)     :03  :54  :47  :28  :03  :55  :06  :47   :30    (:38)
      (9)    1:12  :34  :39  :17  :03  :04  :21  :20   :26
     (10)     :02  :19  :28  :07  :01  :03  :20  :47   :16
     (11)     :02  :51  :09  :14  :03  :02 1:46  :47   :29
     (12)     :02  :15  :24  :12  :02  :02 1:17  :27   :20    (:23)
     (13)     :03  :19  :22  :07  :06  :04  :48  :11   :15
     (14)     :02  :15  :22  :09  :02  :02  :20  :19   :11
     (15)     :08  :06  :12  :05  :02  :02  :03  :06   :06
     (16)     :03  :18  :44  :10  :02  :03  :02  :21   :13    (:11)
     (17)     :02  :17  :41  :03  :07  :07  :11  :06   :12
     (18)     :02  :42  :26  :05  :01  :03  :31  :15   :16
     (19)     :04  :13  :48  :04  :01  :02  :15  :14   :13
     (20)     :04  :32  :35  :04  :01  :02  :08  :22   :15    (:14)
     (21)     :03  :13  :10  :13  :09  :01 1:10  :16   :17
     (22)     :03  :05  :10  :11  :03  :02  :39  :26   :12
     (23)     :02  :07  :17  :05  :03  :15  :20  :26   :12
     (24)     :02  :14  :04  :02  :01  :02  :09 1:03   :12    (:12)
     (25)     :01  :03  :05  :04  :01  :02  :31  :24   :09
     (26)     :02  :04  :08  :01  :01  :05 1:34  :33   :19
     (27)     :02  :03  :03  :03  :10  :05  :15  :39   :10
     (28)     :02  :04  :03  :01  :06  :05  :23  :51   :12    (:12)
     (29)     :02  :04  :10  :03  :02  :06  :23  :21   :09
     (30)     :01  :03  :03  :02  :11  :02  :09 1:08   :12
     (31)     :21  :03  :11  :03  :01  :03  :28  :11   :10
     (32)     :03  :02  :04  :02  :01  :03  :11  :42   :09    (:10)
     (33)     :02  :02  :03  :10  :01  :03  :18  :17   :07
     (34)     :02  :02  :03  :02  :01  :03  :11  :11   :04
     (35)     :01  :02  :03  :02  :01  :02  :18  :09   :05
     (36)     :01  :03  :03  :03  :03  :02  :04  :10   :04    (:05)
     (37)     :01  :02  :07  :02  :04  :01  :11  :09   :05
     (38)     :01  :02  :02  :02  :02  :01  :08  :03   :03
     (39)     :02  :04  :11  :01  :02  :02  :14  :08   :06
     (40)     :01  :02  :03  :01  :01  :03  :10  :03   :03    (:04)

The periods here were similar to those given in the maze tests--the
time was long at first, then it shortened very rapidly for a few
trials, then more slowly but still constantly, until the act became
thoroughly familiar. The process was one of "trial and error"
throughout. As before, various lapses occurred, even although the
animals were as persistent as usual in their efforts to escape. When
the lever was moved to the side or back of the box, none of the animals
could escape. In general, pigeons show less ingenuity in dealing with
latches than do sparrows, according to Porter's[205] observations,
although in some other tests they are equally apt.


B. _Pecking Tests_

The preceding series of trials proved the animals' ability to utilize
certain touching or clawing movements, at first accidental, in making
an escape, and showed that these could become highly specialized.
Desiring to carry out similar tests in the case of pecking movements,
which are quite as natural to pigeons, I arranged a contrivance by
means of which the act of pecking at corn-grains fastened to a small
piece of cardboard (placed just outside the cage, but within easy reach
through an opening in the wire) would open the cage-door by making a
delicate electrical contact. Four animals were tested.

On entering the cage they endeavored to escape as before; failing in
this they began pecking about until they found the corn-grains and
made the contact which opened the door and allowed them to escape to
the food without. Three of the animals made their escape in this way
several times; but the habit seemed to be one that could not be readily
learned, as the successive periods showed little shortening.

Thinking that the pecking of things at a definite place perhaps
complicated the matter, I removed the electrical apparatus and arranged
to open the door myself by pulling a string whenever the pigeon pecked
anywhere upon entering the box. Preliminary experiments with Bird J
indicated some ability to profit by this kind of experience. As the act
of pecking could be used to advantage in a series of imitation trials,
this animal only was allowed to learn to escape by actually pecking;
the others were reserved for the imitation tests next to be reported.


C. _Imitation Tests_

In the experiments already reported the animals were used individually
and usually out of sight of the others, although in the same room
and within hearing of them. When efforts were made in some of the
experiments to test the animals in a separate room, signs of fear and
discontent were often noticed, and it was necessary to return to the
first room to continue the tests. Some instances were noticed in which
a pigeon would do what it saw another doing. For example, one of my
subjects would not eat one day, being ill apparently; but when I put
two others into the compartment with it, and they began eating the food
lying about, it also began pecking. Its act could not have been due
to its only then happening to see the corn, for it had before looked
toward the food when this was thrown to it.

Desiring to test, under definite conditions, the imitative ability of
these animals, I arranged trials in which birds were allowed to see a
useful, but simple act performed by another bird, and then were given
an opportunity to execute the act themselves. Using the animal which
I had trained for the purpose, I allowed its series of acts (entrance
to box, pecking, release, and food-eating without) to be observed by
another animal, confined in a small wire compartment (similar to the
entrance compartment before used) attached to the side of the large
cage. Care was taken to see that the confined bird was observing, or at
least was looking toward the acting one; in case of doubt, the trial
was repeated. Later the trained animal was replaced by the observing
one, and the latter's reaction was noted. Five animals, in all, were
tested, and each was given two opportunities to escape after having
seen the trained animal perform the act ten successive times. None of
them, however, showed any signs of trying to escape by repeating the
movements so often performed by the bird familiar with the act, but
each rushed against the sides of the cage and tried to push through at
various places, just as the trained bird had done when first learning
the habit.

As the act, or series of acts, was rather too complex to be easily
observed and utilized by the other pigeons, I arranged two much simpler
tests. In one case the leading bird was taught to open the cage-door
by stepping upon a platform (the lowering of which made an electrical
contact); in the other, to avoid a blind alley, enter a short passage,
and ascend a wooden plane (inclined at an angle of thirty degrees)
which led into another box containing food. In these tests it was more
difficult for the series of acts to be viewed, but the animals, singly
as before, were placed at a point of vantage and apparently saw the
movements of the other animals.

Of the five birds tested in the platform experiment, four utterly
failed to escape in the two trials given. The fifth, in its second
test, went to the platform promptly and thus made its escape, but the
success may have been accidental, or due to the animal's experience of
seeing and approaching the platform in its first test. In the labyrinth
experiment only one bird (second test) avoided the blind alley and went
directly up the inclined plane to the food; this success was probably
due to experience in the first test. There was certainly no evidence
that the animals had grasped the nature of the problem; nothing to
indicate that the performance of the trained animal had supplied data
for the guidance of future conduct, that is, for the conditioning
of the necessary movements, in this case those of pecking, stepping
upon a platform, or avoiding a blind alley and ascending an inclined
plane.[206]

These results are similar to those which other experimenters have
secured in the case of chicks, cats, dogs, monkeys,[207] and also
rats.[208] Although the method I employed is doubtless open to the
criticism of being artificial,[209] some value at least should be
attached to the results; if so it would seem probable that imitation in
pigeons is not above the "instinctive" stage, and that learning depends
entirely upon first hand experience, upon really doing the thing, and
not upon merely seeing it done.

3. _Position, Color and Form Tests._ The apparatus used in these
experiments consisted of small boxes, six inches in height, and open
at the top. Sometimes they were exactly similar, and sometimes they
differed in color or in form. They were moveably attached, six inches
apart, to a board which was placed in a large wire-covered box, having
an entrance compartment as in the case of the mazes and cages. Food
was placed in one of the small boxes, and the pigeons were allowed to
find it twice; later each bird was tested as to its ability to return
to the food by depending upon the position of the box in the group,
or upon its color or its form. Tests were given in series of six, and
the box which was first approached was recorded as the animal's choice
for that test. If it made a wrong selection, it was allowed to look
about until it found the box containing the food, but in no case was it
permitted to satisfy its hunger until the last test of the series. The
animal apparently did not see the food until it approached the box, and
subsequent tests demonstrated that it was not guided by odor.


A. _Position Tests_

In this series of trials I used at first six, later nine food-boxes,
four inches square and covered with dark gray paper. The board to
which the several receptacles were fastened was shifted at irregular
intervals to various oblique angles; this was done to prevent the
animal from being assisted by the position of the food-box in the
larger box rather than in the group of similar boxes. After a bird
had been tested sufficiently for one position it was then used, for a
week or so, in some other experiment, and thus given an opportunity
to forget, to some extent at least, the old experience before being
taught to find the food in a box placed elsewhere in the group. For
the positions 2, 3, and 4, in the group of six similar boxes, eight
animals were each given thirty tests in series of six, as stated above.
For the positions 5, 6, and 7, in the group of nine similar boxes, the
experiments were shortened. Six animals were given twenty-four trials
each, two animals for each of the three positions.

The animals quickly learned the position of the food-box and passed to
it promptly when released from the entrance compartment. Changing the
position of the board to which the food-boxes were attached did not
affect the animals' ability to reach the food readily. They usually
selected the proper box as before, although frequently they went around
the end of the board and approached the food from the opposite side.
The general distribution of choices in the case of positions 2, 3, and
4 is given in Table V; the rate of learning, in Table VI.

With a single exception (Bird B, box 3) the box containing the food
was far more often chosen than any one of the empty boxes, and usually
more often than all of them combined, the average right choices being:
position 2, 62%; position 3, 57.7%; position 4, 53.3%. It will be seen
that the animals were more successful in finding the food in the second
position than in either the third or the fourth, that is, positions
nearer the end were more easily located. But what is of greater
interest to us is the rate of learning to go to the right box. This
is indicated by the increasing number of right choices from series to
series. (Table VI.) That the increase was small was due to the fact
that the animals learned so quickly in the first series of six tests
that little improvement could be made thereafter; what they could
learn they acquired early in the experiment. There is some evidence
of improvement after the first series in the case of box 4, a more
difficult position, and the average for all three boxes shows a slight
improvement from series 1 to series 5, although a falling-off is seen
in the last series: 26, 27, 27, 31, 28. The same general features
appear in the case of the incomplete tests (positions 5, 6, and 7). For
each box there were, on the average, 26 right choices in the possible
48. There were more right choices in the case of box 7 than in the case
of either box 6 or box 5, box 7 being nearer one end (box 9). There was
little evidence of learning after the first series of trials.


TABLE V. ASSOCIATION OF POSITION: GENERAL DISTRIBUTION OF CHOICES

    _Choices of boxes 1 to 6 when food was placed in boxes 2, 3, and 4_

               _Food in box 2_      _Food in box 3_      _Food in box 4_
                   _Boxes_              _Boxes_              _Boxes_
    _Animals_ 1, 2,  3, 4, 5, 6.   1, 2,  3, 4, 5, 6.   1, 2, 3, 4,  5, 6.
       (B)    1  18  4  6  1  0    1  1   8  7  8  5    0  0  2  21  4  3
       (C)    2  20  6  1  1  0    0  9  19  2  0  0    0  7  2  15  6  0
       (E)    4  18  2  3  2  1    0  6  18  3  3  0    2  6  2  15  3  2
       (F)    5  20  1  2  2  0    1  6  18  3  2  0    0  2  5  14  8  1
       (G)    0  18  5  2  2  0    0  3  13  5  7  2    0  0  3  17  9  1
       (H)    0  22  4  3  1  0    1  0  18  4  6  1    0  4  8  15  3  0
       (I)    4  18  3  1  3  1    0  1  23  6  0  0    0  6  7  15  2  0
       (J)    1  17  7  5  0  0    3  0  21  3  2  1    1  4  3  16  6  0

      Total, 17 151 32 26 12  2    6 26 138 33 28  9    3 29 32 128 41  7


TABLE VI. ASSOCIATION OF POSITION: DISTRIBUTION OF RIGHT CHOICES

    _Choices from series 1 to series 5 in the case of boxes 2, 3, and 4_

                  _Box 2_            _Box 3_            _Box 4_
                 _Series_           _Series_           _Series_
    _Animals_ 1, 2, 3, 4, 5.     1, 2, 3, 4, 5.     1, 2, 3, 4, 5.
       (B)    4  4  5  2  3      2  2  1  1  2      4  4  4  5  4
       (C)    5  4  2  5  4      4  4  4  4  3      3  3  2  2  5
       (E)    3  4  4  3  4      4  4  2  5  3      3  2  2  4  4
       (F)    5  4  3  5  3      5  3  3  4  3      0  2  4  4  4
       (G)    3  4  4  3  4      2  2  4  2  3      3  3  2  4  5
       (H)    4  4  4  5  4      4  3  4  5  2      3  3  4  3  2
       (I)    1  4  3  5  5      4  4  5  6  4      2  1  3  4  5
       (J)    4  5  4  3  1      3  4  6  4  4      3  3  3  4  3

      Total, 29 33 29 31 29     28 26 29 31 24     21 21 24 30 32
            Average for the three boxes, 26, 27, 27, 31, 28.

The method of learning in these position tests was the same as
that noticed in previous experiments, namely, building upon chance
successes. When first admitted to the large box containing the row
of small ones at the farther end, the animal accidentally found the
receptacle containing the food, and later associated the movements
involved in reaching that position with various sense-impressions of
the box, especially those experienced upon entering--certain tactual
impressions of the small entrance compartment, sound of the lifting
door and sight given of the interior of the large box.

While the results clearly indicate that pigeons readily learn the
position of objects, nothing is proved as to "counting." Some
experimenters speak of similar trials as "number-tests," just as
they do of "form-tests," but this is probably going too far. To
investigate counting in animals, experiments should be arranged which
minimize spatial responses. These tests certainly show that pigeons
can discriminate positions readily, especially toward the ends of the
group, but little more is certainly indicated. Porter[210] says: "If
we do not find in birds the power to count, we have in their nice
sense for the location of a member of a series ... something of that
preliminary number-sense which Ribot describes as belonging to children
and savages."


B. _Color Tests_

To investigate the animals' ability to utilize colors[211] in finding
their food, I employed the same apparatus as before, except that
six boxes were used throughout and each was covered with paper of a
different color: red, yellow, green, blue (Bradley's standards, except
red, _RO_ being substituted), gray, and black. The boxes covered with
black and gray paper were employed merely to complete the group of
six. The same method as before was employed, except that the board to
which the boxes were attached was left stationary at the end of the
large box, and also that the position of all six boxes was changed
irregularly for each test.

The general behavior of the animals at the beginning of these tests
was quite similar to that shown in the preceding experiment; but it
was soon evident that colors occasioned them far more difficulty than
positions. The general distribution of choices is given in Table VII.
It will be seen that the proper box was usually chosen more often
than any one of the empty ones, but never oftener than the other five
combined, as occurred in the position tests; also that in the case
of each color there were instances in which another color was as
often, or more often selected. Yet it is clear that colors may serve
as valuable sense-data for these animals. In the first series of six
tests (see Table VIII) there were few right choices or none, but in
each succeeding series the number increased. The learning process was
evidently of the same type as before observed (selection, in this case
gradual, of chance but useful movements), and involved visual data
largely.


TABLE VII. COLOR ASSOCIATION. GENERAL DISTRIBUTION OF CHOICES

    _Choices of all 6 boxes when food was placed in red, yellow, green,
      or blue boxes_

                 _Food in Red Box_            _Food in Yellow Box_
    _Animals_ R,   Y,  G,  B, B'k.    G'y.    R,  Y,   G,  B, B'k, G'y.
      (B)     12   6   7   3   2   0        2  12   6   7   0   3
      (C)     15   8   1   4   2   0        3  12   1   6   3   5
      (E)     11   8   5   3   3   0        3  10   5   3   5   4
      (F)      7   5   9   6   2   1        6   6   5   6   3   4
      (G)      9   1   7   5   3   5        5  11   6   3   2   3
      (H)     13   3   5   3   3   3        5   9   6   3   3   4
      (I)     11   5   2   4   6   2        4  10   5   4   3   4
      (J)     10   0   6   8   4   2        5  10   3   3   5   4

    Total,    88  36  42  36  25  13       33  80  37  35  24  31

                 _Food in Green Box_           _Food in Blue Box_
    _Animals_  R, Y,  G,   B, B'k, G'y.    R,  Y,  G, B,  B'k, G'y.
      (B)      4   5  12   4   5   0            1   3   5  10   6   5
      (C)      3   4  14   5   1   3            3   1   2  13   2   9
      (E)      1   9   7  11   0   2            3   5   5  10   3   4
      (F)      0   8   9   6   0   7            0   3   6  11   5   5
      (G)      2   5  16   2   4   1            5   5   4   5   4   7
      (H)      1   3  15   5   6   0            7   5   5   6   4   3
      (I)      0   4  15   6   2   3            5   5   3   9   5   3
      (J)      4   1  12   9   1   3            6   4   4   7   5   4

    Total,    15  39 100  48  19  19       30  31  34  71  34  40


TABLE VIII. COLOR ASSOCIATION. DISTRIBUTION OF RIGHT CHOICES

    _Choices from series 1 to series 5 in the case of red, yellow, green,
       and blue boxes_

                _Red Box_            _Yellow Box_          _Green Box_
    _Animals_ 1,  2,  3,  4,  5,    1,  2,  3,  4,  5,   1,  2,  3,  4,  5,
      (B)     0   1   2   4   5     1   2   3   2   4    1   2   2   4   3
      (C)     2   2   4   3   4     1   2   3   3   3    3   2   3   4   2
      (E)     2   1   2   3   3     1   0   3   2   4    0   1   1   1   4
      (F)     0   1   1   2   3     0   2   2   1   1    0   1   2   3   3
      (G)     0   2   3   2   2     2   2   1   3   3    2   4   3   3   4
      (H)     1   1   3   4   4     0   2   2   2   3    2   3   4   3   3
      (I)     1   2   3   3   2     1   2   2   3   2    1   3   3   3   5
      (J)     1   2   2   2   3     1   2   1   2   4    1   1   3   3   4

    Total,    7  12  20  23  26     7  14  17  18  24   10  17  21  24  28

           _Blue Box_
          1,  2,  3,  4,  5,
          1   3   1   2   3
          3   2   2   3   3
          0   1   3   2   4
          1   1   2   4   3
          1   1   0   0   3
          0   1   2   2   1
          0   2   2   3   2
          1   2   1   1   2

          7  13  13  17  21

There is no evidence that the color-preference of the animals assisted
them in choosing correctly, in fact, they were rather less successful
in dealing with those colors for which they had previously shown
decided preference,[212] since the whole number of right choices was
less in the case of the green and blue boxes (85) than in the case of
the red and yellow ones (92), and since there was a relative diminution
in the rate of learning toward the last in case of the former
boxes.[213]

To test the animals' ability to discriminate shades of colors in
finding their food, two birds were used, with four boxes, each covered
with a different shade of red paper, and two with the boxes covered
with green paper. The brightness of the different shades was not
measured, but to the eye it seemed to be equal in each of the cases.
The food was placed in the box having the most nearly saturated color,
and twenty-four trials in series of six, as before, were given each
bird. The results were quite similar to those secured with different
colors. With the red shades there were twenty-two choices of the best
saturated shade to eight, ten, and eight, respectively, of the other
three; and with green, twenty-one to nine, ten, and eight. The 43
correct choices were distributed from series 1 to series 4 as follows:
7, 11, 12, and 13, which shows learning as before. The relatively large
number of right choices was probably due, partially to the fact that
fewer alternative choices were possible since only four boxes were
used, instead of six, and partially to the fact that the box containing
the food may have been slightly brighter than the others.

Throughout these trials the position-element was a decidedly disturbing
factor. When the animals were first learning to choose a box of a
definite color, some would show a marked tendency to approach a
receptacle occupying a certain position, and would persist in this from
series to series. Others at first showed no special preference for
certain positions, but, after happening to make a correct choice, they
would return to that same place the next time, and thus miss the right
box which had been changed for the new test.


C. _Form Tests._

In this experiment the six food-boxes were each of different form:
triangular, square, oblong, hexagonal, circular, and elliptical. They
were of the same capacity, and were covered with light-brown paper. As
in the preceding experiment, the birds were tested for only four of the
boxes, and were given thirty trials each. Six animals were used, and as
they were not the same as those previously employed, the square box
(which had always been used before) had no advantage over the others
in attracting the birds at the beginning of the trials. The tests were
given as in the preceding experiment, except that it did not seem
necessary to change the position of each of the six forms before giving
each test; it was thought sufficient to move the food-box, and, if a
wrong choice had been made in the preceding test, also the box wrongly
chosen. The results are shown in Tables IX and X.


TABLE IX. FORM ASSOCIATION. GENERAL DISTRIBUTION OF CHOICES

    _Choices of all 6 boxes when food was placed in Tri., Sq., Hex.,
      or Cyl. boxes_

                   _Food in Tri._              _Food in Sq._
    _Animals_ Tri. Sq. Ob. Hx. Cyl. El.   Tri. Sq. Ob. Hx. Cyl. El.
      (U)       8   6   5   5   2    4     5    9   6   1   6    3
      (V)       9   2   3   6   7    3     5    8   5   2   5    5
      (W)       8   7   3   4   5    3     5    8   5   3   3    6
      (X)      11   5   2   5   3    4     5    8   3   6   3    5
      (Y)      11   4   5   4   2    4     4    9   5   6   3    3
      (Z)       8   3   7   4   4    4     6   11   7   3   1    2

    Total,     55  27  25  28  23   22    30   53  31  21  21   24

                   _Food in Hex._              _Food in Cyl._
    _Animals_ Tri. Sq. Ob. Hx. Cyl. El.   Tri. Sq. Ob. Hx. Cyl. El.
      (U)       5   5   3  11   3    3     6    5   3   5   8    3
      (V)       4   4   5   8   5    4     6    6   3   4   7    4
      (W)       3   5   3  10   7    2     4    3   4   3  10    6
      (X)       4   5   4   7   6    4     2    5   4   4   9    6
      (Y)       3   4   2   8   6    7     3    2   5   6   9    5
      (Z)       5   4   2  11   3    5     3    5   4   4   9    5

    Total,     24  27  19  55  30   25    24   26  23  26  52   29


TABLE X. FORM ASSOCIATION. DISTRIBUTION OF RIGHT CHOICES

    _Choices from series 1 to 5 in the case of Tri., Sq., Hex.,
       and Cyl. boxes_

             _Food in Tri._  _Food in Sq._   _Food in Hex._  _Food in Cyl._
   _Animals_ 1, 2, 3, 4, 5,  1, 2, 3, 4, 5,  1, 2, 3, 4, 5,  1, 2, 3, 4, 5,
     (U)     2  0  2  1  3   1  2  2  2  2   2  1  2  3  3   1  1  2  1  3
     (V)     1  1  2  3  2   2  1  1  2  2   0  1  2  3  2   1  2  2  1  1
     (W)     1  2  1  2  2   2  1  2  1  2   1  2  2  2  3   1  2  2  3  2
     (X)     1  2  3  3  2   0  2  2  3  1   1  2  2  1  1   1  1  2  3  2
     (Y)     1  2  3  2  3   2  2  2  3  2   1  1  2  1  3   0  2  2  2  3
     (Z)     1  2  2  2  1   0  2  3  3  2   1  2  2  4  2   1  2  3  1  2

   Total,    7  9 13 13 13   7 10 12 13 11   6  9 12 14 14   5 10 13 11 13

It will be seen that each animal chose the right box oftener than any
other one box, but not oftener than all of them; also that there was a
small increase in the number of right choices from series to series. No
one of the four forms seemed better discriminated than the others if
we may judge from the practical equality of right choices made in each
case (55, 53, 55, 52) or from the similar increase in number of right
choices from series to series; the hexagonal and cylindrical boxes
received fewer choices in the first series than did the triangular
and square, but this was exactly counterbalanced in the last series.
The triangular box was more often confused with hexagonal and square,
and the square with triangular and oblong, than with the others. For
the hexagonal box the cylindrical was more frequently mistaken than
were the other forms, especially the oblong; and with the cylindrical
the elliptical was more frequently confused than were the others,
especially the oblong. In this series of tests nothing new as regards
general behavior or method of learning was observed.


TABLE XI. POSITION, COLOR AND FORM ASSOCIATION

    _Total Right
     Choices_[214]    _Right choices from series 1 to series 5_[215]
                             1         2         3         4         5
    _Position_ 57.9%     54.2%     55.6%     56.9%     63.2%     59.0%
    _Color_    35.3%     16.2%     29.7%     37.0%     42.7%     51.6%
    _Form_     29.8%     17.4%     26.4%     34.7%     35.3%     35.3%

If we compare the results obtained in these three experiments (see
Table XI and Fig. 7), we shall see that the pigeons were governed much
more by the position of the food-box than by either its color or its
form, and that color was better associated than form. Position was a
most important factor throughout, as was observed also by Porter[216]
in the case of the English sparrow. Porter[217] also found that his
sparrows could associate color better than form. In the position-tests
the pigeons showed very little improvement from series to series (see
table); almost all that the animals could learn was acquired at the
beginning. The more difficult color- and form-trials, however, showed
almost constant improvement, although we should have expected this to
be greater in the latter case than it was. When judged entirely by the
actual number of right choices in a given kind of tests, some of the
birds made a very poor showing; but from the standpoint of increasing
number of right choices they appeared in a wholly different light.

Thus, for example, bird F (Table VII) made only 33 right choices in a
possible 120, yet their arrangement is significant, being, from series
1 to series 5, respectively, 1, 5, 7, 10, 10. It is probable that there
would have been still greater improvement had the tests been continued;
perhaps the animal would have become as proficient in finding its food
by depending upon the color of the receptacle usually containing it, as
by relying upon the position of the box in the group.

[Illustration: FIG. 7. Position, Color, and Form Association. If
line S represent tests of a given kind, P, C, and F would represent,
respectively, the number of correct choices of position, color, and
form. The rate of learning in each case is shown by the corresponding
curves to the right, where vertical divisions each indicate 20%, and
horizontal divisions the successive series in which the tests were
given.]


V. SUMMARY

1. Respiration in pigeons is sensitive to various stimuli, and since
its alterations of rate, amplitude, etc., can be easily recorded
pneumo-graphically without frightening the animals, it may well serve
as a process through which to study their mental life.

2. By repetition meaningless stimuli, for example, pistol-shots,
quickly lose their disturbing influence; whereas the breathing remains
sensitive to those of a significant character, such as the noises made
by other birds.

3. Reaction to light of moderate intensity consists principally in an
immediate quickening, the amount varying with the color; since a direct
correspondence was found between color-preference and breathing-rate,
it would seem that here agreeable feeling involves increased breathing
activity.

4. Visual, acoustical, probably tactual, and certainly organic data,
are the principal sensory factors of the associations of pigeons.

5. The animals readily form useful associations by a method of "trial
and error," or the selection of successful movements which were at
first accidental.

6. Apparently a pigeon does not learn by merely seeing a new act
performed by another pigeon; yet there are instances of simple
("instinctive") imitation, and "trial and error" learning is not
wholly independent of social conditions, since it proceeds much more
satisfactorily if the animal is trained at least within hearing
distance of other pigeons.

7. When a habit is being formed, the "period" required for the first
test is usually very long, but learning proceeds quite rapidly during
the next few trials; later it is more gradual, but it continues till
the act becomes thoroughly familiar.

8. Associations are fairly permanent, and some remain practically
unaltered for at least six weeks. Modification is easily accomplished,
however, on the basis of new experience.

9. Pigeons differ widely both as to the ease with which they acquire
associations and also as to their permanence. Difference in activity
seems the chief reason for this.

10. While these birds seem mentally inferior to English sparrows and
to various mammals which have been tested in a similar manner, they
are capable of numerous ready adjustments. They discover circuitous
labyrinth passages, they learn to manipulate latch apparatus when
adapted to their natural habits and conveniently placed, and they
easily reach their food by depending upon the position, color, or
form of the box containing it. But the process is apparently simple
association throughout. There is no evidence of higher mental
activity--no looking the situation over and acting accordingly, no
"reasoning" in the proper sense of the word, but only blind movements,
some of which are retained and become highly specialized, merely
because successful.

FOOTNOTES:

[Footnote 180: D. Ferrier: The Functions of the Brain, p. 111, London,
1886.]

[Footnote 181: A. Hill: Can Birds Smell? Nature, vol. 71, pp. 318, 319,
1905.]

[Footnote 182: W. Mills: The Nature and Development of Animal
Intelligence, pp. 248, 250, New York, 1898.]

[Footnote 183: Hachet-Souplet: Examen Psychologique des Animaux, pp.
33-38, Paris, 1900. See also Riverside Natural History, vol. 4, pp.
240, 241, Cambridge, 1888.]

[Footnote 184: Orientation chez le pigeon-voyageur, Revue Scientifique,
vol. 13, pp. 352-359, 1900.]

[Footnote 185: Orientation du pigeon-voyageur, Revue Scientifique, vol.
2, pp. 417-420, 453-457. 1904.]

[Footnote 186: Grundlinien zur Erforschung des Helligkeits- und
Farbensinnes der Tiere, p. 102, Prague, 1888.]

[Footnote 187: Some Notes on the Psychology of Birds, Seventh Annual
Report of the New York Zoölogical Society, p. 154, 1902.]

[Footnote 188: Ueber die Begleiterscheinungen psychischer Vorgänge in
Athem und Puls, Philosophische Studien, vol. 18, pp. 7-14, 1901.]

[Footnote 189: Animal Intelligence, pp. 8-12, 31-36, 51-55, New York,
1898.]

[Footnote 190: Nature of Animal Intelligence and Methods of
Investigating It, Psychological Review, vol. 10, pp. 262-274, 1897.]

[Footnote 191: An Experimental Study of the Mental Processes of the
Rat, American Journal of Psychology, vol. 11, pp. 135-164; vol. 12, pp.
206-210, 1900-1901.]

[Footnote 192: Mental Life of Rhesus Monkeys in Captivity, American
Journal of Psychology, vol. 13, pp. 97-148, 180-210, 1902.]

[Footnote 193: A Preliminary Study of the Psychology of the English
Sparrow, American Journal of Psychology, vol. 15, pp. 313-346, 1904.]

[Footnote 194: For a more complete report of this special part III,
see the writer's paper, Respiration and Emotion in Pigeons, Journal of
Comparative Neurology and Psychology, vol. 15, pp. 494-513, 1905.]

[Footnote 195: If shallowing accompanies quickening, the respiratory
activity may be no greater than before; but since depth alterations
were seldom observed in these trials after the first day of
experimentation, the rise in rate may be taken as a fair measure of the
influence of the stimulus.]

[Footnote 196: P. Zoneff und E. Meumann: _op. cit._, pp. 57, 58.]

[Footnote 197: R. MacDougall: The Physical Characteristics of
Attention, Psychological Review, vol. 3, pp. 162, 176, 177, 1896.]

[Footnote 198: Thorndike: _op. cit._, pp. 13-15.]

[Footnote 199: Small: _op. cit._, vol. 12, pp. 236, 237.]

[Footnote 200: Small (_op. cit._, vol. 11, p. 146) states that in
his rats "the persistence of useless motor habits is striking" and
"explainable by the supposition that the movements are touched off
automatically."]

[Footnote 201: _Op. cit._, vol. 12, p. 214.]

[Footnote 202: _Op. cit._, p. 28.]

[Footnote 203: The Instincts, Habits and Reactions of the Frog, Harvard
Psychological Studies, vol. 1, pp. 591-593, 1903.]

[Footnote 204: Small: _op. cit._, vol. 12, pp. 230, 231.]

[Footnote 205: _Op. cit._, pp. 318, 319.]

[Footnote 206: C. L. Morgan: Animal Behaviour, pp. 179-193, London,
1900; Animal Life, and Intelligence, p. 453, Boston, 1891.]

[Footnote 207: Thorndike: _op. cit._, pp. 54, 56, 57, 60, 61; The
Mental Life of the Monkeys, Psychological Review, Monograph Supplement,
vol. 3, pp. 318, 319, 1901. Kinnaman: _op. cit._, pp. 198-200.]

[Footnote 208: Small: _op. cit._, vol. II, p. 160.]

[Footnote 209: W. Mills: Nature of Animal Intelligence and Methods of
Investigating It, Psychological Review, vol. 10, pp. 262-274, 1897.]

[Footnote 210: _Op. cit._, p. 335. See also C. L. Morgan: Introduction
to Comparative Psychology, p. 232, London, 1900.]

[Footnote 211: Cornish states (Animals at Work and Play, p. 30) that
hunters near the Caspian are able to decoy partridges by use of
brilliant colors.]

[Footnote 212: See the writer's paper, Respiration and Emotion in
Pigeons, _op. cit._, p. 502.]

[Footnote 213: One of Porter's sparrows was less successful with
yellow and red than with blue and green. He says: "This may be partly
explained from the fact that she was more afraid of these." _Op. cit._,
pp. 338, 339. See also E. L. Thorndike: Instinctive Reactions of Young
Chicks, Psychological Review, vol. 6, pp. 283-284, 1899.]

[Footnote 214: Tables V, VII, and IX.]

[Footnote 215: Tables VI, VIII, and X.]

[Footnote 216: _Op. cit._, p. 338.]

[Footnote 217: From the tables (_op. cit._, pp. 330-339) it seems that
the right choices for position, color, and form, were respectively,
40%, 58% and 20%. The comparatively small number of correct position
choices was probably due to his using ten boxes instead of six, as in
the other two series. My results given in Table XI were secured under
almost exactly comparable conditions. Compare results of Kinnaman in
case of the Rhesus monkey, _op. cit._, pp. 130, 131, 134, 141, and 177.]




REACTIONS OF THE CRAYFISH

BY J. CARLETON BELL


The crayfish has long been the typical Crustacean for anatomical
and physiological investigations, but it is only recently that its
reactions to sensory stimuli have been made the object of experimental
study. The purpose of this paper is to describe the reactions of the
animal to certain sensory stimuli under experimental conditions, and to
estimate the relative importance of these stimuli in the life of the
organism.


I. REACTIONS TO VISUAL STIMULI

Huxley[218] states that crayfish avoid direct sunlight, hiding under
stones during the day, and becoming active in the evening. On the other
hand, they are attracted like moths to fires lighted on the bank at
night, and may be scooped out by hand. Abbott,[219] giving an account
of the burrowing crayfish, _Cambarus diogenes_, states that it is very
difficult to observe the animals at work, since all their digging is
done at night. It would seem from the account of Miss Hoppin, quoted
by Garman,[220] that the blind crayfish, _Cambarus pellucidus_, is not
altogether insensitive to light, for, reporting on the fauna of the
caves of Missouri, she says that the crayfish are all found near the
entrance to the cave, where there is considerable light. In the dark
recesses there are only little white fishes. Blind fish and crayfish
are also taken from the wells in the neighborhood, where the crayfish
are found only in wells that are rather shallow and light; the fish, on
the other hand, are only obtained from deep, dark wells.

According to the above accounts it would appear that the crayfish is
negatively phototactic to direct sunlight or diffuse daylight, but
positively phototactic to a light at night, and moreover, that light
may influence the behavior of the animal even when the eyes have ceased
to function.

The directive influence of light upon the movements of the crayfish has
never been experimentally studied to my knowledge. Dearborn[221] thinks
that light has no effect upon the animals. Yerkes[222] and Towle[223]
have shown that Daphnia move toward the light. Bethe[224] finds that
Carcinus is negatively phototactic, and also shows a tendency to
hunt out corners. When the eyes are varnished with lampblack, the
phototaxis disappears, but the tendency to seek out corners still
remains. Bethe says that he has observed the same phenomenon in the
crayfish. Keeble and Gamble[225] discovered that _Hippolyte varians_
responds positively to light under all conditions, and Palæmon is just
as markedly negative. _Macromysis_, however, reacted now positively now
negatively, depending on the background. A black (absorbing) background
called forth a positive response, while a white (scattering) background
produced a negative reaction. Spaulding,[226] in studying the habits of
the Hermit Crab (_Eupagurus_), found that it is strikingly positively
phototactic. When animals are placed in an aquarium, one half of which
is shaded, none of them are ever noticed inside of the dark line.
Herrick[227] notes that lobsters are nocturnal, and avoid the light
when placed in a tank, and Bateson[228] says that prawns and shrimps
lie hidden during the day, and are active only at night. Parker,[229]
in a study of Copepods, finds that the females have a strong positive
phototaxis for light of a low intensity, while males show a weak
negative phototaxis. To light of over 100-candle power at a distance of
10 cm. or to direct sunlight the female Copepods are negative, while
the reaction of the males does not seem to be altered.

In his work on _Carcinus_, Bethe obtained retraction of the eye-stalks
by suddenly throwing a strong light on the eye by means of a mirror.
"Usually the eyes were quickly drawn in and protruded again, sometimes
several times in rapid succession, like a man blinking under a sudden,
strong light." When a dark object, the size of the hand, was moved just
over the water, the eyes were seldom retracted, but the antennules were
usually drawn in. Lemoine[230] observed that in Astacus retraction
was due to touch alone, and that no light, however strong, was able
to bring about such a reaction. Gulland[231] takes just the opposite
view with reference to Astacus, stating that there are no setæ of any
sort on the eye-stalk, and therefore it is insensitive to touch, but is
withdrawn only because the animal sees the object by which the stimulus
is given. If a curved needle is used, and the stimulus is applied
from behind, no retraction follows. Dearborn,[232] however, working
with Cambarus, agrees with Lemoine in saying, "Withdrawal of the
ophthalmites into their sockets occurs only on contact with some hard
object,--not from any light-stimulus of an ordinary sort." I may say in
passing that in none of the following experiments on Cambarus was there
ever a sign of retraction due to stimulation by light, the retraction
always taking place in response to a touch-stimulus.

Lyon,[233] in his study of compensatory movements of the eye-stalks,
found that when the eyes were painted with lampblack, the crayfish
showed a reduction of about 10% in the compensatory movements when
rotated in vertical planes, but the compensation remained the same for
rotation about the dorsi-ventral axis. On rotation in the dark the
compensatory movement of the eyes was found to be from 5° to 8° less
than in the light.


EXPERIMENTAL

In the investigations to be described, 58 crayfish of the species
_Cambarus affinis_ were made use of, and for identification the animals
were marked on the back with white enamel paint, the males receiving
the even numbers from 2 to 64, the females the odd numbers from 1 to 51.


1. _Reactions to White Light_

The questions proposed for investigation were, (_a_) How does the
crayfish react to diffuse daylight; (_b_) to reflected sunlight; (_c_)
to direct sunlight; (_d_) to artificial light of different intensities?
(_e_) What is the influence of previous conditions of exposure to light
upon the reactions of the animal? (_f_) Do changes of temperature
affect the reactions?

A wooden box, 80 cm. long, 25 cm. wide, and 20 cm. high, painted black
on the inside, and constructed so as to hold water, was covered with
a heavy black cloth to exclude the light from above. The front end of
the box was of glass, thus admitting the light from the end. In all
the experiments except those with direct sunlight, this glass end was
covered with black cardboard in which a hole 10 cm. long and 5 cm. high
had been so cut that the light was admitted at the middle of the bottom
of the glass. The direct sunlight was admitted through the whole of
the glass end. At the rear of the box a piece of black cardboard was
so arranged that an aperture was afforded for observing the animals
without admitting any appreciable amount of light, and this aperture
could be readily closed by a slide when not in use.

The method of experimentation was to place the animal in the box about
20 cm. from the glass end, and observe whether it went toward or away
from the source of light. The animals were experimented on in two
groups of five each, and one hundred observations were made on the
individuals of each group with each intensity of light, that is, twenty
observations on each animal. In order to check the influence of the
orientation of the animal at the time of exposure to the stimulus, the
following four positions for placing the animal were chosen: (1) Head
toward the light; (2) Head away from the light; (3) At right angles
to the light with right side toward it; (4) At right angles with the
left side toward the light. Thus five observations were made on each
animal of each group in each position, exposed to each of the different
intensities of light.

Seven different intensities of light were employed, and the results
have been arranged in eight sets, as follows: I. Diffuse daylight in
dry box, _i. e._, the animals were taken out of their ordinary medium,
water, and were exposed to the stimulus of diffuse daylight in the air.
The reactions under these conditions, however, were so slow and so
unsatisfactory that the test was abandoned after the first group, and
thus the second group has nothing to show for itself under this head.
The remaining seven sets of observations were made on animals placed
in 10 cm. of water at 15° C. II. Diffuse daylight. III. Reflected
sunlight. The box was placed near a window on a clear day, and the
sunlight was thrown in horizontally by means of a mirror. IV. Direct
sunlight. On a clear day the box was placed in such a position that
the sun shone in directly and illuminated the front half of it. V.
9-candle-power incandescent electric light. This lamp was marked 16
c., but it had been used a great deal, and on being tested with a
Lummer-Brodhun photometer showed only 9 c. VI. An incandescent electric
light of about 50 c. This lamp was marked 100 c., but had been used
considerably and was slightly smoked. Unfortunately it was broken
before there was any opportunity to test it. Judging from the fact
that another 100 c. lamp of the same manufacture, in slightly better
condition, measured 64 c., the estimate of 50 c. seemed a safe one.
VII. The incandescent electric light alluded to above, which measured
64 c. VIII. An arc light which varied in intensity from 150 c. to 250 c.

In intensities V and VI the lamp was placed 5 cm. from the glass end
of the box to allow the interposition of a heat-screen consisting
of an alum solution in a flat glass jar 5 cm. thick. Reckoned in
candle-metres, therefore, the intensity of the illumination at the
surface of the animal in V was 144 c. m., and that in VI was about 800
c. m. In VII two heat-screens were used, and between these was placed a
lens of considerable but not accurately determined curvature, so that
it is impossible to express the intensity in candle-metres. In VIII the
light was so variable that such an expression would mean nothing.

Unfortunately it was impossible to keep the two groups constant
throughout the whole series, owing to the death of two individuals in
each group during the experimentation. Group 1 was composed of nos.
1, 3, 4, 8, and 9, of which 1 and 8 were replaced by nos. 13 and 42
respectively. Group 2 was begun with nos. 23, 27, 32, 34, and 38, and
the vacancies caused by the death of 23 and 32 were filled by nos. 21
and 36. The following table exhibits the reactions to the different
intensities of light, + indicating an orientation toward the source
of light, - an orientation away from the light, and ± an indifferent
orientation, which usually means no movement at all.


TABLE I. SUMMARY OF REACTIONS TO WHITE LIGHT

               Group 1       Group 2       Totals
              +   -   ±     +   -   ±     +   -   ±
       I.    23  48  29                  23  48  29
      II.    49  47   4    19  81        68 128   4
     III.    40  60        30  70        70 130
      IV.    47  52   1    36  64        83 116   1
       V.    51  49        30  70        81 119
      VI.    39  61        39  61        78 122
     VII.    28  72        28  71   1    56 143   1
    VIII.    35  63   2    37  60   3    72 123   5

            312 452  36   219 477   4 = 531 929  40 = 1500

The following table gives the average time required for orientation
for each group. The time of each animal in seconds was noted with a
stop-watch from the instant the animal was placed in the box until
a definite orientation was assumed with reference to the light. If
no orientation followed within three minutes, the result was called
indifferent.


TABLE II. AVERAGE TIME OF ORIENTATION

                                  Intensities
              I    II    III    IV     V    VI     VII    VIII    Ave.
    Group 1 (144)  52     11   4-1/2   9   8-1/2    11     19    16-1/2
    Group 2         9      3   4       4   5         4      5     5

Inspection of these tables shows that when the animals were taken
from the water and placed in diffuse daylight in the air (I), their
movements were so sluggish that in twenty-nine cases out of one hundred
there was no orientation within the three-minute limit. Moreover,
in the seventy-one cases where there was definite orientation the
average time was over two minutes (144 seconds). While, therefore, the
conditions were quite different from the normal environment of the
animal, it is interesting to note that of the cases where orientation
did take place the negative reactions were more than twice as many as
the positive. In II, where the animals were under the same conditions
of diffuse daylight but in the water, a wide variation in the reactions
of the two groups is noted. In group 1 they are about equally divided
between positive and negative, while in group 2 there is the largest
proportion of negative reactions in the whole series. It will be
observed that the time for group 1 is extremely long compared with the
other averages, and this doubtless indicates a general sluggishness
and lack of sensitiveness to stimuli in the animals, which might to
some extent account for the difference in reaction. If we consider the
totals of both groups for each intensity, we are led to conclude that
there is no appreciable difference in the reactions of crayfish to
diffuse daylight, to sunlight, or to artificial light within the limits
here employed. A slight exception to this is found in VII, where the 64
c. lamp with the lens caused a somewhat more uniform negative reaction.
The action of direct sunlight in IV is rather remarkable in that with
the lowest proportion of negative reactions in the whole series we
also observe the shortest time-average, indicating that the animals
are the liveliest and most sensitive. This would seem to indicate that
while the animals are in general somewhat negatively phototactic to
all light-stimuli of moderate intensity the action of direct sunlight
tends to reduce the negative phototaxis to a minimum. If we consider
the totals of the two groups separately we observe that group 1 has
only 57% of negative reactions while group 2 shows 68%. This is rather
in accordance with what we would expect from the general time-average,
which is over three times as much for group 1 as for group 2. But
although we may in a general way connect rapidity of orientation with
a large percentage of negative reactions, it will not do to carry it
to individual cases, for it was observed that no. 27 showed 83% of its
reactions negative, yet its total time-average was 10 sec., the highest
in its group.

In general, then, we conclude that the crayfish is negatively
phototactic in the proportion of about two to one. This apparently
contradicts the statement made by Huxley that crayfish "are attracted
like moths to fires lighted on the bank at night." For surely if this
were the case some such tendency would have been observed in these
experiments. On the other hand, there is no such marked and definite
response to light as in the case of Daphnia or Hippolyte or Palæmon or
the Hermit Crab. The action of the stimulus is by no means mechanical
and constant, but there is wide variation in individuals.

As was mentioned in the description of the method of experimentation,
four different positions for placing the animal were chosen with the
idea that the initial position of the animal with respect to the
light might have some influence on the direction of its movement. To
determine what this influence might be, a careful record was kept of
the orientation with reference to each one of these positions, and the
following table gives a summary of these observations. In the table
position I is where the animal is placed with its head toward the
light; position II, with head away from the light; position III, at
right angles to the light with the right side toward it; position IV,
at right angles with left side to light.


TABLE III. INFLUENCE OF POSITION ON LIGHT REACTIONS

            Position I   Position II  Position III  Position IV   Totals
            +   -    ±    +   -   ±     +   -   ±    +   -   ±   +   -   ±

    Group 1 85  108  7    61 135  4     74 114 12    92  95 13  312 452 36
    Group 2 39  133  3    48 126  1     57 118       75 100     219 477  4

           124  241 10   109 261  5    131 232 12   167 195 13  531 929 40

Since the animals have been shown to be somewhat negatively
phototactic, we should expect that position II, with the head away from
the light, would show the largest number of negative reactions, and
this is what we find if we take the sum of both groups. But by the same
course of reasoning we should expect position I, with head toward the
light, to yield the smallest number of negative reactions, a condition
which prevails neither in the sum nor in either of the groups. On
the whole we can only say that difference of position seems to have
remarkably little influence on the orientation of the animals.

In his work on the eye of the crayfish, Parker[234] called attention to
the migration of the pigment in the retinular cells under the influence
of light. The question now arose, what influence, if any, does this
pigment migration exert upon the reactions of the crayfish to light?
The time required for pigment migration in the eye of the crayfish has
never been determined to my knowledge, but from the work of Parker[235]
on Palæmonetes it was thought that confinement in the dark for about an
hour would be sufficient to bring about a retraction of the pigment.
Accordingly, group 1 was kept in the dark for one hour, group 2 for
one hour and a half, before experimentation. A further test was made
to observe the effects of pigment expansion, both groups having been
exposed for one hour and a half to the rays of a 32 c. incandescent
electric light at a distance of 40 cm. The apparatus used in the
reaction-tests was the box described above, with the 64 c. light as a
stimulus. As to the method of observation, each group was placed in
the centre of the box at right angles to the horizontal rays of light,
and the position of each animal was accurately noted at intervals of
one minute for one hour. In reporting the results, all the observations
of animals in the half of the box nearest the light are denominated
positive, those in the half farthest from the light negative. The
results are given in Table IV, where line I indicates the reactions
after confinement in the dark, line II those after exposure to the
light.


TABLE IV. INFLUENCE OF PREVIOUS CONDITIONS UPON REACTIONS TO LIGHT

    Group 1    Nos. 13      3         4         9        42      Totals
               +    -    +    -    +    -    +    -    +    -    +     -
     I         54   6    60        5   55    48  12        60   167   133
    II         59   1    60            60     9  51    55   5   183   117

    Group 2    Nos. 21     27        31        36        38      Totals
               +    -    +    -    +    -    +    -    +    -    +     -
     I          5  55     5  55    7   53    10  50    4   56    31   269
    II         38  22        60   13   47    45  15   16   44   116   184

Let it be said at once that these results do not offer an altogether
satisfactory basis for an answer to the question proposed above. Some
of the animals would take up a position during the first ten minutes
and remain in it for the rest of the hour. Whether the position taken
was due solely to the light, or was owing to thigmotactic influences,
or whether it depended on the way in which the animal was released,
are questions which cannot be answered, and for this reason the
conclusions to be drawn from the table are tentative. If we examine
the table we find that the totals of both groups agree in manifesting
a decrease in negative results for line II, after exposure to the
light, as compared with line I, after confinement in the dark. This is
what we would expect from negatively phototactic animals. When taken
from the dark the pigment is retracted, and the sensitive retinal
substance is exposed to the direct action of a rather strong light. The
negative tendency of the animal we should expect to find accentuated.
The decrease in negative reactions is especially marked with group 2,
which was shown above to be much the livelier of the two, and all the
individuals except no. 27 share in the change. In group 1 the decrease
is not so striking, and is observed to be due to two individuals
solely. Inexplicable is the preponderance of positive over negative
reactions in the results for group 1.

All of the experiments thus far described were carried out in water at
approximately 15° C. The question naturally arises, what will be the
result of raising or lowering the temperature upon the reactions of the
animals to light? Unfortunately the experiments anent this question
are fragmentary and incomplete, but the results will be given for what
they are worth. The same apparatus and the same intensity of light (64
c.) were used as in the preceding paragraph. The results, presented
in Table V, are arranged in three sets, as follows: The line marked I
represents the results obtained from group 1 at a temperature of 5° C.
The animals were placed in the box one at a time, as in Table I, and
their orientation noted. They were set at right angles to the rays of
light, five times with the right and five with the left side toward
the source of the stimulus. No observations were made upon group 2
at 5° C., and those on group 1 are so few as to have a questionable
value. Line II gives the reactions of both groups of animals in water
at 25° C., and in this set the animals were placed in all four of the
positions indicated for Table III. Line III presents the reactions of
the animals in water at 25° C. by the method outlined for Table IV, _i.
e._, each group of animals was placed in the centre of the box, and
observed at intervals of one minute. To obviate the objection of the
animals remaining in one spot, they were reset every ten minutes in the
middle of the box, at right angles to the entering rays.


TABLE V. REACTIONS TO LIGHT AT DIFFERENT TEMPERATURES

    Group 1  Nos. 13     33         4        42         9       Totals
             +    -    +    -    +    -    +    -    +    -     +    -
      I       7   3    4    6    2    8     2   8    8    2     23   27
     II      16   4   14    6   11    9     9  11   13    4(±3) 63   34(±3)
    III      19  41   13   47   39   21    17  43   27   33    115  185

    Group 2  Nos. 21     27        31        36        38       Totals
             +    -    +    -    +    -    +    -    +    -     +    -
     II      12   8    5   15   12    8    15   5   13    7     57   43
    III      12  48   12   48   31   29    42  18   44   16    141  159

Judging from lines II and III we may say that there is a tendency
toward a decrease of the negative phototaxis with an increase in
temperature. It is true that group 1 in line III maintains the average,
62% negative reactions, but the others are much lower than this, line
II even going over to positive phototaxis in both groups. In line III
the animals of both groups were extremely active during the first
ten minutes, rushing about from one end of the box to the other,
pushing each other back and forth, and in general exhibiting great
restlessness. Some of the animals when first put into the water reacted
with a sort of cramp reflex, which was followed in a few seconds by
intense activity. After the first ten minutes the animals began to
grow more quiet, and in twenty or thirty minutes they had become quite
sluggish, scarcely moving out of the position in which they were reset.
During the period of restlessness the males showed marked sexual
activity, rushing up to the females, pushing them about, seizing them,
and trying to turn them over in spite of their vigorous resistance.
One of the males, no. 36, did succeed in turning a female on her back
twice, although she struggled violently to escape,--a thing which the
female never does in the ordinary sexual act. The rise in temperature,
therefore, seemed to stimulate the males to sexual activity, but not
the females.


2. _Reactions to Colored Light_

No observations have ever been made, so far as I know, on the reactions
of the crayfish to colored light. Lyon, in his work on compensatory
eye-movements, found that rotation in blue light gave a compensatory
movement only slightly less than that in white light, while in red
light the compensation was only a little larger than in darkness. In
some animals the interposition of an opaque object between the eye and
the source of light caused an elevation of the eye 1° or 2° toward the
vertical. Red glass acted like an opaque object, blue glass produced
no effect, _i. e._, blue light had the same effect as white light. To
observe whether the same thing applied to movement reactions was the
object of the following experiments.

_a. Reactions to Horizontal Colored Light._ The same apparatus was used
as in the previous experiments, viz., the dark box with light from
the 64 c. lamp entering horizontally at the end. Across half of this
end were placed pieces of colored glass of a saturated blue, green,
yellow, and red. The colored light obtained by this means was not
spectrally pure, but it was the nearest to it that could be obtained.
A more serious objection is that the intensities were not the same,
the red and the yellow being very appreciably brighter than the blue
and the green. In addition to observations with these colors, a piece
of black cardboard was introduced in the same position as the glass,
thus cutting off the light from that half of the box. This, to preserve
the uniformity of the series, is denominated black. The animals were
placed in the centre of the box, on the line separating the white from
the colored light, and were observed at intervals of one minute for
forty minutes, the position of each animal being accurately noted. At
the end of every ten minutes the animals were reset at the centre of
the box. The following table gives a summary of the results for each
individual. Here again it was impossible to keep the groups constant
owing to the death of individuals during the progress of the experiment.


TABLE VI. REACTIONS TO HORIZONTAL COLORED LIGHT

                      Group 1                         Group 2
  Animals 13 37 41 42 44 33  4  9 43 46  Sum   21 27 31 36 38 52  Sum  Sum
                                                                      Total
  Blue    12  9 21 30 18                  90   23  2 24 13 37      99  189
  White   28 31 19 10 22                 110   17 38 16 27  3     101  211

  Green   15        9    23 36 17        100   12 19  3 28 40     102  202
  White   25       31    17  4 23        100   28 21 37 12         98  198

  Yellow  20 10 31 28 19                 108   17 11 28 37 35     128  236
  White   20 30  9 12 21                  92   23 29 12  3  5      72  164

  Red     26        9    32 20 17        104    2 30 16 29 36     113  217
  White   14       31     8 20 23         96   38 10 24 11  4      87  183

  Black      24 27    12          27 26  116   15 18    26 25  5   89  205
  White      16 13    28          13 14   84   25 22    14 15 35  111  195

In this table the colored lights are arranged in the order of the
spectrum from blue to red. On the hypothesis that blue light has
practically the same effect upon animal reactions as white light,
while red is about the same as darkness, we might expect that the
reactions would be about equally divided between the blue and the
white, and that there would be a gradually increasing difference in
number as we go down the table, reaching the maximum with the last
pair, black-white. This, we see, however, is not quite the case. In
both groups the white has a slightly larger number of reactions than
the blue, while the pair green-white shows numbers more nearly equal.
In the sum totals the yellow shows a greater preponderance over the
white than any other color, and the black and white are very nearly
equal. Group 1, it is true, shows a fairly regular ascending scale in
reactions to the colored lights with the exception of the red, and the
same might be said of group 2 if it were not for the very low number
of reactions to the black and the exceptionally high showing of the
yellow. On the whole, however, the differences are so small and the
individual variations are so large that we can only conclude that for
these conditions colored light has little or no effect on the reactions
of the animals.

In the foregoing experiment the light came from a broad spiral coil
inside the bulb of the lamp, and the distance from it to the edge of
the glass was so small compared with the length of the box that there
was no sharp dividing-line between the colored light and the white,
but rather a wedge-shaped block of lessening saturation of the color,
and this wedge, having the point toward the light, took up the whole
of the box at the extreme farther end. Thus the imaginary central line
dividing the white light from the colored departed farther and farther
from the reality as the rear of the box was approached. To obviate
this difficulty and to get a check on the previous work, the following
series of experiments was undertaken.

_b. Reactions to Vertical Colored Light._ The same box was used as in
the previous experiments, but the end was closed with a black cloth,
and an electric light marked 32 c. but measuring only 22 c. was hung
exactly over the middle of the box, 40 cm. from the bottom. By means
of wires it was arranged that a plate of colored glass could be swung
in such a manner that all of one half the box (the whole of one end)
was illuminated with the desired color, while the other half was either
left white or illuminated with another color. In this way there was a
fairly sharp dividing-line between the two colors. The animals were
observed at intervals of one minute for 40 minutes, and reset at the
middle on the dividing-line every ten minutes as before. Table VII
gives the results of the observations.

In this set of experiments it was possible to keep the groups intact
except that in group 1 no. 13 had to be replaced by no. 46. If now we
conceive the colors arranged in the order of the spectrum with black
at one end and white at the other, and consider the black a lower
stimulus than the white, we have the ascending series black, red,
yellow, green, blue, white. Now since the animals have already been
shown to be somewhat negatively phototactic, we should expect them to
prefer a color of lower stimulus to one of higher. Turning to the sum
totals in the table we find that the first color of each pair (which
is always the lower stimulus) has the larger number of reactions in
every case but one, the first pair of red-blue. As was stated above,
it was impossible to secure colored light of the same intensity by
means of the glass at our disposal, and in the present case the red
was considerably brighter than the blue. Owing to the fact already
mentioned that different intensities of white light seem to have no
effect on the reactions it was thought that these differences in the
intensities of the colored lights might be overlooked. Since the only
thing that could be thought of to account for the anomalous behavior to
the red-blue was this difference in intensity, another experiment was
undertaken with the same animals under slightly different conditions. A
glass aquarium about 40 cm. long by 20 cm. wide was covered with black
cardboard and black cloth in such a manner that light could enter only
through a space 5 cm. wide at the bottom of each end. Each of these
ends was covered, the one with blue, the other with red glass, and 15
cm. from each end was placed an electric light marked 32 c. Later,
however, it was found that one of these lamps measured 30 c. and the
other 22 c. The red light was found to be much more intense to the
eye than the blue, so the former was damped down with tissue paper
until the two appeared to have the same intensity. The second pair of
red-blue in Table VII gives the results of the observations under these
conditions, and these are found to be in harmony with the rest of the
table, _i. e._,[236] the color giving the lower stimulus has the higher
number of reactions.


TABLE VII. REACTIONS TO VERTICAL COLORED LIGHT

                 Group 1            Group 2           Group 3           Sum
  Animal
     no. 13 37 41 43 44  Sum  21 27 36 38 52  Sum  54 56 58 60 62 Sum Total

  Blue   17 22 14 25 29  107  11 21 26 27 26  111  26 32 26 28 10  122  340
  White  23 18 26 15 11   93  29 19 14 13 14   89  14  8 14 12 30   78  260

  Green  17 26 11 21 29  104  21 27 30  7 26  111  29 31 26 10  9  105  320
  White  23 14 29 19 11   96  19 13 10 33 14   89  11  9 14 30 31   95  280

  Yellow 25 21 12 20 35  103  12 20 34 25 29  120  19 20  9 28 22   98  321
  White  15 19 28 20 15   97  28 20  6 15 11   80  21 20 31 12 18  102  279

  Red    37 33 23 22 28  143  21 30 30 33 22  136  32 37 26 33 28  156  435
  White   3  7 17 18 12   57  19 10 10  7 18   64   8  3 14  7 12   44  165

  Black  23 20  1 28 34  106   9 12 25 40 32  112  25 32 18 34 16  125  343
  White  17 20 39 12  6   94  31 28 15    14   88  15  8 22  6 24   75  257

  Black  20 15 13 27 37  112  16 11 16 27 24   94  36 40 18 29 29  152  358
  Red    20 25 27 13  3   88  24 29 24 13 16  106   4    22 11 11   48  242

  Black  20 25 26 20 32  123  22 15 39 40 32  148  33 22 34 34 11  134  405
        [236]
  Blue   20 15 14 20  8   77  18 25  1     8   52   7 18  6  6 29   66  195

  Red    25 17 28 18  8   96   3 15 21 28 20   87  21 13 15 14 28   91  274
  Blue   15 23 12 22 32  104  37 25 19 12 20  113  19 27 25 26 12  109  316

  Red    21 27 22 10 36  116  13 40 24 21 28  126  22 25 18 21 16  102  344
  [237]
  Blue   19 13 18 30  4   84  27    16 19 12   74  18 15 22 19 24   98  256

The most striking feature of the table is the marked predominance of
the red over the white. Here the red reaches 73% of the total number
of reactions, and inspection shows that this predominance is uniform
not only through the groups but even for the individuals. The constancy
of this reaction and the fact that it is so much more frequent than
the one to the black as compared with the white, would lead one to
expect that the red would have the higher percentage in the combination
black-red. Such, however, is not found to be the case, although it
does happen with one group. If the arrangement of our color-scale in
accordance with increasing intensity of stimulus were correct, we
should expect a gradually increasing predominance in reactions to
colored light over those to white in the first five pairs. Instead of
this we find that green and yellow stand nearest to the white, blue and
black come next and are almost equal, while red is very much higher
than any. In the pairs black-red and black-blue the red holds its
predominance over the blue at about the same rate as in the second pair
of the direct comparison, red-blue. The wide individual variations,
however, in all these reactions to colored light, except perhaps in
the case of red-white, indicate that there is nothing very regular,
stereotyped, or mechanical about them. The most that can be said is
that in a general way the red end of the spectrum furnishes a less
intense stimulus to negative reaction than the blue.

A tendency to habit formation was noticed during the course of these
experiments, and it is possible that this may have influenced the
results somewhat. Many individuals apparently formed a habit of going
to a certain corner as soon as they were reset at the centre. The
positions in which they were set were varied and they were headed in
different directions, but within a minute after they were released in
the middle of the box they would be found in their favorite corner.
This was especially the case with no. 38 in Table VI, and I think
accounts in some measure for the persistent avoidance of the white.
In no case did this continue throughout the whole series, but would
sometimes be noted for two or three days at a time in the case of an
individual. What were the controlling factors in this habit formation,
the means by which orientation and recognition were effected, I was
unable to determine.


3. _Reactions to Objects_

In no case did an animal give any sign of perceiving stationary objects
in its path or of avoiding them in any way that could be referred to a
visual stimulus. When the animal approached an obstruction there was
no hesitation in the movement until the object was touched. Usually
even when the antenna had touched the object the animal did not stop,
but continued until the contact of the chelæ or even of the rostrum
made further movement in that direction impossible.

With moving objects the case was quite different. Here the condition
and disposition of the individual animal seemed to be the deciding
factors. Often when the animals were trying to climb out of a shallow
pan in which they were kept in the experimenting-room, raising a
finger or holding out a pencil would be sufficient to make them stop
or even start back into the pan. Nor was this response occasioned by
any change in the intensity of light, such as that caused by a shadow
falling on the animal, for they would react to a movement made on the
opposite side of them from the window. In fact, no. 56, the most active
in response to moving objects, seemed to react more vigorously to a
motion made on the opposite side than when it was made between him
and the light. Whenever a person came near the aquarium he and one or
two others would take an attitude of defence, and would "face about"
to correspond to any movement the person made toward one side or the
other. When in the pan mentioned above, any movement of a person within
two or three yards of him usually called forth a reaction on his part,
and if the pan were placed on the table and the person moved slowly
round it, the animal turned with the person, making a complete circuit
of the pan.

Reaction to a smaller moving object, however, was not so marked. A
black object, 20×8×8 cm., was suspended above the middle of the pan
so that if set swinging it would just pass over the top. When it was
pulled to one side the animal responded slightly, but after the first
swing he seemed to pay no more attention to it. When the operator
stepped out from behind the screen, the animal was as keen in its
response as before. The experiment was now tried of allowing the object
to approach from one direction while the operator moved to a position
at right angles to its line of movement. Without hesitation the animal
moved so as to keep fronting the operator, without paying any attention
to the movement of the smaller object, although this was much nearer.

These observations on the reactions of the crayfish to stationary and
moving objects are in line with the conclusions of Plateau[238] and
Exner[239] drawn from observations on other Arthropods. It is Exner's
belief that the compound eye is a visual apparatus which is almost
worthless for detecting the _forms_ of objects, especially if these
objects are stationary, but that it may furnish a very keen perception
of _moving_ objects.


II. EXPERIMENTS WITH SOUNDS

Hensen[240] stated that Palæmon and Mysis reacted to sounds made by
striking a thin, resonant board floating on the surface of the water,
or by tapping the walls of the aquarium or of the room. Beer[241]
repeated Hensen's experiments, but denied that the Crustacea reacted to
sounds, and claimed that their movements were due to visual and tactual
stimuli. Prentiss[242] confirmed Beer's results on Palæmonetes, and
noted that the reactions were only slightly diminished by the removal
of the otocysts, but that removal of the antennæ and antennules caused
their almost complete cessation. More extended experiments were made
on the fiddler crab, _Gelasimus pugilator_, which is on land a good
deal of the time, and Prentiss's conclusions are: "(1) The reactions
formerly attributed to sound-stimuli are nothing more than tactile
reflexes. (2) The otocyst has little or no part in calling forth these
reactions. (3) There is no direct evidence to prove that decapod
Crustacea hear, and until such evidence has been obtained, we are not
warranted in ascribing to the otocyst a true auditory function."

The experiments performed on the crayfish in this connection
all resulted negatively and go to confirm Beer's and Prentiss's
conclusions. Rapping upon a board floating in the water, and tapping
the sides of the aquarium did not cause the slightest apparent reaction
in the animals under observation, even though the vibration of the
water could be plainly perceived by the sense of touch in the hand.
When a rather large electric bell was sounded just over the surface
of the water some reactions were observed which were evidently due
to the movements of the hammer, but there was nothing which could be
referred to the sound-stimuli. If the bell was held against the sides
of the aquarium, or in the water near the animals, the vibration could
be plainly felt by the fingers, yet no reactions on the part of the
crayfish were observed. A metal snapper making a crack like a small
pistol-shot was tried both in and out of the water but with no success
in producing a reaction. A large hand tuning-fork, when held with its
base pressed firmly against the glass walls of the aquarium, gave a
deep rich tone of great volume, or when lightly touched to the glass
produced a shrill, piercing, penetrating sound which was extremely
sharp and disagreeable. Here again the vibrations of the water were
quite perceptible to the hand at a distance of 10 cm., yet in neither
case was there a sign of a reaction. Finally two electric tuning-forks,
one of 256, the other of 512 vibrations per second, were tried on the
animals taken one by one, and especial attention was given to the
regular movement of the little thread-like appendages which keep up the
current of water to the gills, with the idea that perhaps their rate of
movement might be affected. In no case was there the slightest movement
that could be referred to vibration, although here again the tactile
stimulus was very perceptible to the finger. None of these experiments,
then, give any indication that the crayfish reacts to vibratory stimuli
which to the human ear produce sound.


III. ROTATION EXPERIMENTS

It has been found that the higher vertebrates, on being rotated on a
turn-table, exhibit all the symptoms which accompany the sensation
of dizziness in man. The question arises, to what degree and in what
manner do invertebrates respond to rotation? Schaefer,[243] the first
to take up this question, denied on rather meagre observations that
Crustacea respond in any way to rotation on the turn-table. Kreidl[244]
showed that this statement was altogether too sweeping, that Palæmon
reacts very definitely to rotation by running in the opposite
direction. Bunting[245] tried the crayfish, but all the rotation
experiments resulted negatively, so she was led to confirm Schaefer's
statement so far as the crayfish is concerned. Bethe[246] found that
Carcinus behaved in a very definite manner on being rotated, that
during the rotation the animals ran in the opposite direction to that
in which they were turned, and as soon as the motion ceased they began
running in the other direction. Finally Lyon,[247] while agreeing with
Bunting that adult crayfish do not react to rotation, discovered that
young animals two or three centimetres long react very prettily to the
movement, going in a direction opposite to the turn. To confirm and if
possible extend these observations on the crayfish was the purpose of
the following experiments.

It was soon found that a great deal depended on the method of
experimentation. None of the experimenters mentioned above gives any
detailed description of the manner in which the experiments were
carried out. One is left uncertain whether the animals were placed on
the periphery of the turn-table or over the centre, whether in the
former case they were set with their heads toward the centre or away
from it, or placed at right angles to a radius, or whether they were
merely set down in any chance fashion and whirled about. The same
indefiniteness exists in most of the accounts as to how fast they were
turned, and whether the experiments were performed in the air or in the
water. Finally it is not stated whether the rotation was always in the
same direction, or whether its direction was alternated.

The turn-table used in the following experiments was one that had to
be turned by hand, so that it was impossible to regulate the speed
accurately. The crank, however, was not attached directly to the
rotating board, but was connected with it by means of a gearing so
that one turn of the crank produced about ten turns of the table. This
gearing gave a steadying effect to the motion so that the speed could
be kept tolerably constant. A circular pan, about 15 cm. in diameter at
the bottom with the sides slightly sloping outward, was set so that its
centre coincided with the axis of the rotating table. It was in this
pan that all the experiments were tried. Through various preliminary
experiments to determine the most favorable speed, it was found that a
rotation rate of over one turn of the table per second produced such a
strong centrifugal force that unless the animals were set exactly over
the centre they were swept off against the side of the pan in such a
manner that it was difficult to decide whether the rotation as such had
any effect upon their movements. It was finally decided that the best
results were obtained from a rate of approximately one rotation in two
seconds.

It soon became evident that when the larger and more sluggish crayfish
were merely dropped in the pan and rotated there was no particular
reaction. This was true whether the animals rotated were in the air or
in the water. The smaller and more active crayfish, however, showed a
decided tendency to run either with or against the direction of the
rotation, especially when the experiments were carried on in the water.
In no case was there any tendency to go in the opposite direction when
the rotation ceased, except in so far as the animals were carried along
by the water. To get a quantitative expression for these tendencies a
more delicate method of experimentation was resorted to. If there was
a tendency on the part of the active animal to move either with or
against the rotation, such a tendency might also be supposed to exist
in the sluggish animal, only in the latter the inertia was sufficiently
strong to prevent its appearance. If, however, the animals should be
set radially to the periphery of the pan, the tendency to go with or
against the rotation would be exhibited in the direction in which they
turned out of the radial position. For it was found that no animal
would remain in that position for any great length of time. Two groups
of animals were used for these experiments, five animals in each group,
and the first group was selected from the smallest and most active
animals, the second from the largest and most sluggish. Each animal
was set in two positions, position I, with the head toward the centre,
position II, with the head away from the centre. Each animal was given
ten trials in each position, and the number of times it turned in a
direction _with_ the rotation is set down in the + column, the number
of times it turned _against_ the rotation is indicated in the - column.
In general from 5 to 15 turns were necessary for the orientation of
the animal, though sometimes the number ran up to 30 or 40. Each trial
was made in the opposite direction to the preceding one, in order to
avoid the formation of any habit in turning. All these experiments were
carried out in water, the depth of which in the pan was about 4 cm. In
order that there should be no difference between the velocity of the
water and that of the pan, the table was rotated a few times before
the animal was put in. As a check a series of experiments of 5 in each
position was performed in the air on the more active group.

The following table shows the results of these experiments in rotation:


TABLE VIII. REACTIONS TO ROTATION

                    In Water                  In Air
                                                                    Sum
                 I      II      Sum       I       II    Sum        Total
    Group 1    +  -    +  -    +   -    +   -   +   -   +  -       +   -
       44      5  5      10    5  15    1   4       5   1  9       6  24
       49      2  8    4  6    6  14        5   1   4   1  9       7  23
       56      3  7    8  2   11   9    5       3   2   8  2      19  11
       62      4  6   10      14   6    2   3   2   3   4  6      18  12
       64      2  8    9  1   11   9    1   4   1   4   2  8      13  17
      Sum     16 34   31 19   47  53    9  16   7  18  16 34      63  87
    Group 2
       21      5  5    3  7    8  12
       27      2  8    5  5    7  13
       36      7  3    8  2   15   5
       37      2  8    3  7    5  15
       54      5  5    6  4   11   9
      Sum     21 29   25 25   46  54                              46  54
    Sum Total 37 63   56 44   93 107    9  16   7  18  16 34     109 141

Examination of the table reveals great individual variation. Some
animals, as nos. 44, 49, and 37, turn rather constantly against the
direction of the rotation, while others, as nos. 56 and 36, are almost
as constant in their movement with the rotation. On the whole we
observe that for each group, and for Group 1 in both water and air,
there is a slightly greater tendency to go against the rotation than
with it. This tendency, strange to say, comes out much more clearly in
the air than in the water. It is evident, however, from the variation
exhibited that there is nothing very stereotyped or mechanical about
the reaction. Mention should be made of the fact that usually (though
not always) the animals not only oriented themselves with reference
to the rotation, but moved forward in that direction as long as the
rotation continued.


IV. GEOTAXIS, BAROTAXIS, AND TURNING

(1) _Geotaxis._ So far as my knowledge extends, no experimental work
has been done to determine the geotaxis of decapod Crustacea. Most
of the vertebrates are positively geotactic, while a great many of
the invertebrates, particularly unicellular organisms, larvæ of moths
and butterflies, slugs, etc., are negatively geotactic. Parker[248]
found that in the case of the Copepod, _Labidocera æstiva_, the
females exhibited strong, the males weak, negative geotaxis. In the
investigation of the geotaxis of the crayfish, two sets of experiments
were undertaken. In the first the method of procedure was as follows:

On a level table before a window a board was so arranged that it could
be set at an inclination of 5°, 10°, 15°, 20°, and 25° either toward
or away from the window. Starting, let us say, with the inclination
toward the window, each one of a group of five animals was placed on
the board with the right side to the window five times. The board was
then inclined the same amount away from the window and the process was
repeated. The same procedure was carried out with the animals set with
the left side to the window. The following table gives the results of
this set of experiments.


TABLE IX. GEOTAXIS IN FRONT OF WINDOW

            10         12         14         16         18       Totals
          +  -  ±    +  -  ±    +  -  ±    +  -  ±    +  -  ±    +  -  ±
     5°  14  5  1   11  7  2   11  9      13  7       8 10  2   57 38  5
    10°  13  7      12  7  1   13  6  1   16  4      14  6      68 30  2
    15°  17  3      14  6      16  3  1   17  3       9 11      73 26  1
    20°  12  8      17  3      18  1  1   19  1      14  6      80 19  1
    25°  18  2      19  1      19  1      17  3      16  4      89 11

From this table it appears that the crayfish is positively geotactic,
and that the positive geotaxis increases regularly with the increase
in inclination. As a check on these results another set of experiments
was undertaken with different animals under different conditions. The
board was placed on a level table in the centre of a darkened room,
and the operator stood behind a screen so as to be quite hidden from
the animals. In order to observe the orientation a 2 c. incandescent
electric light was suspended directly above the spot where the animals
were set, at a distance of 60 cm. above the board. Each animal of a
group of five was set five times in each of four positions, viz., head
down the incline, head up the incline, and at right angles to it with
first the right and then the left side down the slope. The results were
as follows:


TABLE X. GEOTAXIS IN DARKENED ROOM

            41         46         48         51         64       Totals
          +  -  ±    +  -  ±    +  -  ±    +  -  ±    +  -  ±    +  -  ±
     5°  12  4  4   13  4  3    8 10  2   14  5  1   11  7  2   58 30 12
    10°  14  5  1   13  6  1   11  8  1   13  6  1   14  3  3   65 28  7
    15°  16  4      15  5      13  7      11  6  3   14  2  4   69 24  7
    20°  16  4      19  1      16  4      20         17  2  1   88 11  1

It will be observed that Tables IX and X agree quite well in the main,
and we may conclude that the crayfish is positively geotactic and that
the positive reactions vary from 58% at 5° to 89% at 25°.

(2) _Barotaxis._ Verworn[249] uses the term barotaxis in an inclusive
sense to cover all pressure phenomena that can be classed under the
sub-heads of thigmotaxis, rheotaxis, and geotaxis. It seems preferable
to me to employ the term in a more restricted sense of reaction to
pressure other than the pull of gravity, the flow of a current, or
the contact with bodies. The following experiment with the crayfish
furnishes us, I think, with a case in point.

A glass aquarium, 54 cm. long and 28 cm. wide, was so inclined that
the water was 20 cm. deep in one end and 8 cm. deep in the other. A
board was so anchored that one end rested on the bottom at the shallow
end of the aquarium while the other end projected slightly out of and
above the deepest water. The board was about 45 cm. long, so that its
slope was very gradual. Nine animals were placed in this aquarium and
observed for three successive days. If we denote the bottom of the
deep end of the aquarium by A, the shallow end under the board by B,
the shallow end on top of the board by C, and the end of the board at
the surface of the water by D, the results of the observations were as
follows: On the first day 1 animal was found at D, 6 at C, and 2 at B.
On the second day 5 were at C, 3 at B, and 1 at A. On the third day 1
was at D, 4 at C, and 4 at B. Totals, 2 at D, 15 at C, 9 at B, and 1 at
A.

While these observations were too few to base very positive statements
on, the striking fact that only one animal was found at A, the deep
end of the aquarium, whereas 15 were noted on top of the board at C,
indicates strongly that the animals avoid the deeper water. That the
animals were found _on top_ of the board, not under it, indicates
that the observation is not to be referred to thigmotaxis, although
the latter is doubtless very strong, as we shall see later. It should
be observed that the negative barotaxis works against and overcomes
the marked positive geotaxis which, as we have just seen, the animals
exhibit in the air. Under the influence of the positive geotaxis,
we should expect to find the greater number of the animals at A,--a
condition which is speedily realized if we let the water run out
of the aquarium. We conclude, therefore, that at certain pressures
(specifically at the pressure exerted by water at a depth of 20 cm.)
the crayfish is negatively barotactic.

(3) _Turning._ In the experiments with light it was observed that very
seldom do the animals, when placed upon a surface, move off at once
in a straight line, but usually they first turn through an angle of
90° or more and then start off straight. This came out strongly in the
work on geotaxis, where oftentimes, when the animal was set with the
head up the incline, the reactions would be preponderantly positive,
whereas when set with the head down the incline the reactions were on
the whole negative. In other words, when headed up the incline the
animal would go down, and when headed down he would more often go up.
Some experiments were tried under various conditions to determine how
general this tendency is. The table presents the results in condensed
form.


TABLE XI. EXPERIMENTS IN TURNING

    Nos.   10 12 14 16 18 Sum       7  9  15  17  25  Sum
         I                     III
    90°-    8  5  6  8 10  37       2  4   7   5   5   23
    90°     4  6  7  2  2  21       3  1                4
    90°+    8  9  7 10  8  42       5  5   3   5   5   23
         II                     IV
    90°-    2  1  2     2   7          9   3   1   2   15
    90°     1     1     2   4       3  5   3   3   1   16
    90°+    7  9  7 10  6  39      12  1   9  11  11   44

    Nos.   41 43 46 48 64  Sum                             Sum
         V                                                Totals
    90°-    9 11 12 14  2  48                              130
    90°     2  1  2  1  3   9                               54
    90°+    9  8  6  5 15  43                              191

In this table the first line indicates the number of times each animal
turned less than 90° when starting off from the position in which it
was set, the second line the number of times the amount of turn was
practically 90°, and the third more than 90°. The five parts of the
table mark the different conditions; in Part I twenty observations were
made on each animal placed on a level board before the window, and set
now with the right now with the left side toward the window. In Parts
II and III the animals were set with the head turned now toward now
away from the window. In Parts IV and V the animals were placed on a
level board in the middle of a darkened room with a 2 c. light about
three feet above them. This was to exclude any possible directive
influence of light. In all cases the operator was concealed by a
screen. In Parts I and V twenty observations were made on each animal,
in II and III ten, and in IV fifteen.

Rarely the animal would turn completely round and start off in the
direction originally set, but usually the turn was between 90° and
180°. When once the animal began to move off, it would ordinarily keep
to an approximately straight line. How seldom this was observed when
the animals were first set down may be judged from the fact that out of
a total of 375 observations in only 18 did the animals move straight
ahead from their original position. From the table we observe that in
over 65% of the cases (a proportion of almost two to one) the animals
turned through 90° or more before starting off. At present the writer
has no explanation to offer for this phenomenon.


V. THIGMOTAXIS AND TOUCH REACTIONS

(1) _Thigmotaxis._ Experiment has shown that there are some animals
which tend to avoid contact with objects as much as possible, and
on the other hand there are animals that seek to get as much of the
surface of their bodies as possible in contact with objects. The former
are spoken of as negatively, the latter as positively thigmotactic.
Does the crayfish show any tendency in the one way or the other, and
if so is it positively or negatively thigmotactic? In a large glass
aquarium, 80 cm. long and 40 cm. wide, was a thin wooden box, 22 cm.
long and 16 cm. wide, set in one corner 4 cm. from the glass walls.
At the bottom of the box was an opening where the crayfish could
enter. The following table shows the disposition of the animals for
27 different days, on which one examination was made each day. Line I
indicates the number of times each animal was found against the walls
of the aquarium, line II in the 4 cm. space between the box and the
walls of the aquarium, line III inside the box against its sides, and
line IV resting freely in the middle of the aquarium or in the middle
of the box.


TABLE XII. THIGMOTAXIS REACTIONS

    Animal 4  5 9 13 21 27 29 31 33 35 36 37 38 39 41 42 43 44 45 46 47 48

      I    4  9 2 10  4 17 16  6     6 10  3     2 11  9  3  2  1  9  1 15
     II       5 2 13 18  9     5  2  4  5 22  1 14 11  3  6 11  1  9  4  7
    III       2 1  2  5  1  9  3  2  3 11  2 24  7  1     4 14     9  1  5
     IV    1 11             2     1           2  4  4  1  2


    Animal 49 51 52 54 56 58 60 62 64                              Sum
                                                            Totals
      I     6 11  3  1  2  4  7 10 12                              195
     II     8  3  7  5  8  1  3  2  1                              190
    III     1    16  9  5 10  3  2  2                              154
     IV     1  1           2  1                                     33
                                                                   572

In order to appreciate the significance of the figures in this table it
is necessary to consider the amount of lateral surface with which it
was possible to come in contact in each case. In IV of course it was
zero, in III it was 76 cm. with four corners in close proximity, in II
it was only 38 cm. and one corner, but the space was so narrow that
there was practically a contact-surface on both sides, and in I there
was 212 cm. of lateral surface with three corners. I mention corners
in this connection because they were almost invariably occupied. If
we examine the table with these facts in mind, we find, (1) that the
number of animals resting freely without contact with any lateral
surface is very small, only about 6% of the whole; (2) that the number
of animals found in the narrow space between the box and the walls
of the aquarium is very large in proportion to the length of the
space: indeed the animals were frequently found wedged into this space
three or four deep; (3) that the number of animals found in the box
was probably due largely to the fact that they found in it a greater
lateral contact-surface, particularly in the corners, than was possible
outside.

Two or three minor considerations are of interest. The animals were
frequently observed "on edge" about half out of the water, that is,
with the ventral surface of the body pressed against the vertical
surface against which they were resting. This was also observed where
the water was so deep that none of the members could touch the bottom.
It was perhaps on account of the quality of the surface affording a
rougher contact that so large a number of the animals were found in
contact with the wooden box rather than the smooth, slippery surface
of the glass. In the centre of the aquarium a wooden stopper 2 cm. in
diameter projected about 15 cm. above the surface of the water. Very
often a crayfish would be found almost at the top of this stopper,
completely out of the water. This tendency to climb was frequently
observed in the light-reaction experiments, where the animals would
climb up on any piece of wood that chanced to be left in the box. It
reminds one of the tree-climbing crabs of the West and East Indies.
Along the creeks of Ohio I have frequently seen crayfish that had
climbed up on logs or sticks that projected some feet out of the water.

In the table we see decided evidences of "habit" in the sense of an
animal returning to the same place which it had occupied. No. 5 has
almost half the observations in the open, nos. 21, 37, and 39 showed a
decided preference for the space between the box and the aquarium wall,
while nos. 38, 44, and 52 were more frequently found on the inside of
the box. This recurrence to a particular position also came out in the
light-reaction work, where an individual would return to the same spot
in the box for days at a time as soon as released.

From the above considerations we conclude that the crayfish is strongly
positively thigmotactic and that this thigmotaxis probably plays a most
important part in the life of the animal.

(2) _Touch Reactions._ Lemoine[250] investigated the reactions of
crayfish to touch-stimuli and found that the plates of the telson, the
sternal portions of the thorax, the abdominal pleopods, the chelæ, and
particularly the antennæ toward their points are especially sensitive,
but that nowhere, even on the back of the carapace, is a touch-stimulus
altogether devoid of reaction. Gulland[251] found that a needle could
be inserted between the tufts of setæ on the chelæ without causing any
reaction, but as soon as one of the hairs was touched, the chelæ closed
with a snap. Considering the setæ as the organs of touch, he claimed to
have found that the eyes, eye-stalks, and carapace (which he says have
no setæ) are impervious to tactile impressions. This claim of Gulland's
is strangely at variance with the facts. In no case have I been able
to bring about retraction of the eye-stalk by visual stimulation, but
a very light touch-stimulus on the eye itself or on the eye-stalk
or a stronger stimulus on some portion of the head will cause the
eye to be drawn in. It is true that after repeated stimulation the
eye is retracted no longer, and with a heavy bristle one can make a
perceptible indentation in the corneal surface without the eye being
withdrawn.

The antennæ, from their anatomical structure, their position, and
the manner in which they are carried, are generally considered the
special organs of touch. Nevertheless, as far as the reactions of the
animal are concerned, a stimulation of the antennæ by touch produces
a less decided response than almost any other portion of the body.
If the stimulus is very light no reaction at all is observed in most
cases, and if stronger the antennæ are moved away, but that is all. A
stimulation of the edge of the telson produces a more decided reaction.
Either the animal folds it under the abdomen at once or faces about
like a flash in an attitude of defence; frequently both reactions
occur. While the response to stimulation of the chelæ was decided,
that to touch on the first chelipedes was quicker and more accurate.
The mouth-parts are also very sensitive to touch. I cannot agree with
Gulland's assertions as to the insensitiveness of the carapace, for I
have been able to find no place upon it where a light-stimulation would
not produce a reaction. In this connection a curious phenomenon is
characteristic of the animal. If the carapace or the front portion of
the abdomen be lightly stroked with a solid object such as a pencil,
the animal will slowly turn toward the stimulus on its antero-posterior
axis. If, now, a like stimulus be applied on the other side, the animal
will roll back through the normal position to a like inclination
toward the stimulus on the other side. If the alternation be kept up
and the change made quickly, a continuous and curious rolling movement
is maintained, the animal growing more and more excited until it
scampers off with a kind of cramp-like motion. With some animals this
rolling reflex is more marked than with others, but in no case is it
altogether lacking. Some animals have been known to roll so far over
that they topple over on their backs. Dr. Yerkes informs me that he has
observed the same phenomenon in a less degree with turtles when the
edge of their shell is stimulated by scratching. The movement seems
to be caused by the reflex stimulation of the extensor muscles on the
opposite side of the body from the part stimulated. The thrust of the
legs thereby brought about raises that side of the body and thus causes
a rotation to some extent about the antero-posterior axis. But how was
this connection between the stimulation of one side of the body and
the contraction of the extensor muscles of the other side established?
I have no doubt that it is intimately connected with the positive
thigmotaxis described above. These animals live under loose stones for
the most part, and thus the carapace gets a great deal of stimulation.
If the animal is stimulated on one side, a contraction of the extensor
muscles of the opposite side tends to roll the animal toward the source
of the stimulus, and hence to increase the contact. In the race-history
of the animal this has doubtless been advantageous in enabling it to
escape the dangers of its habitat.

       *       *       *       *       *

In a succeeding paper the writer hopes to discuss the reactions of
the crayfish to chemical stimuli. In conclusion he desires to make
acknowledgment to Dr. Robert M. Yerkes of the Harvard Psychological
Laboratory for kindly suggestions and helpful criticism throughout the
course of the investigation.


SUMMARY

(1) Crayfish are somewhat negatively phototactic, going away from
rather than toward the source of light in the ratio of 62% to 38%. The
different intensities employed in this investigation produced very
little difference in the reactions. The average reaction-time was much
less for the group of animals which showed the highest percentage
of negative reactions, indicating a greater general sensitiveness.
Variations of the position in which the animals were set affected the
results very slightly.

(2) Previous confinement in the dark tended to increase slightly the
number of negative reactions, and previous exposure to strong light
tended to decrease the number, but the results were not constant. An
increase in temperature tended to decrease the number of negative
reactions to light, but here again the results were somewhat
conflicting.

(3) Reactions to horizontal colored light showed a tendency to go to
the colored light rather than to the white in the following order:
Blue 47%, green 50.5%, black (or the absence of light) 51%, red 54%,
yellow 59%. In the case of vertical colored light the comparison with
the white resulted somewhat differently, as follows: Green 53%, yellow
53.5%, blue 57%, black 57%, red 72.5%. In the latter experiments the
animals showed a marked and constant preference for the red.

(4) The animals showed no signs of reaction to static objects from
visual stimulation, i. e., there is no evidence of visual perception
of form in the case of stationary objects. Moving objects, especially
large ones, are plainly perceived and definitely reacted to.

(5) There were no reactions whatever caused by those vibrations which
to the human ear produce sound. So far as these experiments go, the
animals cannot be said to hear.

(6) In rotation experiments individual animals were rather constant in
moving either with or against the direction of the rotation, but no
definite tendency for all animals was observed.

(7) The pull of gravity was followed with constantly increasing
frequency from 58% at 5° to 89% at 25°. Therefore we conclude that
the animals are positively geotactic. They are negatively barotactic,
avoiding the pressure of water at the depth of 20 cm., and this is
sufficient to overcome their positive geotaxis. When placed upon a
level surface the animals show a peculiar tendency to turn through a
greater or less angle before starting out in a straight line. In only
18 out of 375 observations, or 5%, did the animals start straight, in
30% they turned through an angle of less than 90°, and in 65% they
turned through an angle of 90° or more.

(8) The crayfish is positively thigmotactic in a marked degree, as is
indicated by the fact that in only 33 out of 572 observations, or less
than 6%, were the animals found resting in the open, while in 190, or
33%, they were found in a narrow opening between two vertical surfaces.

(9) The animal is sensitive to touch over the whole surface of the
body, but especially on the chelæ and chelipedes, the mouth-parts, the
ventral surface of the abdomen, and the edge of the telson. If one side
of the carapace or of the dorsal surface of the abdomen be stimulated,
the extensors of the legs on the opposite side are contracted, and
the animal turns on its antero-posterior axis toward the source of
the stimulus. If opposite sides be stimulated alternately, a peculiar
rolling motion is set up.

FOOTNOTES:

[Footnote 218: An Introduction to the Study of Zoölogy, illustrated by
the Crayfish, Internat. Sci. Ser., 1880.]

[Footnote 219: How the Burrowing Crayfish works, Inland Monthly,
Columbus, Ohio, vol. 1, pp. 31, 32, 1885.]

[Footnote 220: Cave Animals from Southwestern Missouri, Bull. Mus.
Comp. Zoöl. Harv. Univ., vol. 17, pp. 225-239, 1889.]

[Footnote 221: Notes on the Individual Psychophysiology of the
Crayfish, Amer. Jour. of Physiol., vol. 3, pp. 404-433, 1900.]

[Footnote 222: Reactions of Entomostraca to Stimulation by Light, II,
Reactions of Daphnia and Cypris, Amer. Jour. of Physiol. vol. 4, pp.
405-422, 1900; Reactions of Daphnia pulex to Light and Heat, Mark
Anniversary Volume, pp. 359-377, 1903.]

[Footnote 223: Heliotropism of Cypridopsis, Amer. Jour. of Physiol.,
vol. 3, pp. 345-365, 1900.]

[Footnote 224: Das Nervensystem von Carcinus mænas, I, Arch. f. mikros.
Anat., vol. 50, pp. 460-547, 589-640, 1897.]

[Footnote 225: The Color Physiology of Higher Crustacea, Phil. Trans.,
London, Series B, vol. 196, pp. 295-388, 1904.]

[Footnote 226: An Establishment of Association in Hermit Crabs
(_Eupagurus longicarpus_), Jour. Comp. Neur. and Psych., vol. 14, pp.
49-61, 1904.]

[Footnote 227: The American Lobster; A Study of its Habits and
Development, Bull. U. S. Fish Comm., vol. 15, pp. 1-252, 1895.]

[Footnote 228: Notes on the Senses and Habits of some Crustacea, Jour.
Marine Biol. Assoc'n., Plymouth, N. S., vol. 1, pp. 211-214, 1889.]

[Footnote 229: The Reactions of Copepods to Various Stimuli, Bull. U.
S. Fish Comm., vol. 21, pp. 103-123, 1902.]

[Footnote 230: Recherches pour servir à l'histoire des systèmes
nerveux, musculaire, et glandulaire de l'écrevisse, Ann. des Sci. Nat.,
Series 5, vol. 9, pp. 99-280; vol. 10, pp. 5-54, 1868.]

[Footnote 231: The Sense of Touch in Astacus, Proc. Roy. Physiol. Soc.,
Edinburgh, vol. 9, pp. 151-179, 1886.]

[Footnote 232: _Loc. cit._]

[Footnote 233: Contribution to the Comparative Physiology of
Compensatory Movements, Amer. Jour. of Physiol., vol. 3, pp. 86-114,
1899.]

[Footnote 234: The Retina and Optic Ganglia in Decapods, especially
in Astacus fluviatilis Mitth. Zool. Stat. Neapel., vol. 12, pp. 1-73,
1895.]

[Footnote 235: Photomechanical Changes in the Retinal Pigment Cells of
Palæmonetes, etc., Bull. Mus. Comp. Zoöl. Harv. Univ., vol. 30, pp.
273-300, 1897.]

[Footnote 236: No. 46 substituted.]

[Footnote 237: The second pair of red-blue gives the results of an
experiment under somewhat different conditions as described above.]

[Footnote 238: Recherches expérimentales sur la vision chez les
Arthropodes, Mém. Corronnés de l'Acad. Roy. des Sci. etc. de Belgique,
vol. 43, pp. 1-91, 1889.]

[Footnote 239: Die Physiologie der facettirten Augen von Krebsen und
Insekten, 1891.]

[Footnote 240: Studien über das Gehörorgan der Dekapoden, Zeitsch. f.
wiss. Zool., vol. 13, pp. 319-412, 1863.]

[Footnote 241: Vergleichend-physiologische Studien zur
Statocysten-function, I. Ueber den angeblichen Gehörsinn und das
angebliche Gehörorgan der Crustaceen, Arch. f. d. ges. Physiol., vol.
73, pp. 1-49, 1898; Idem. II. Versuche an Crustaceen, Arch. f. d. ges.
Physiol., vol. 74, pp. 364-382, 1899.]

[Footnote 242: The Otocyst of Decapod Crustacea, Bull. Mus. Comp. Zoöl.
Harv. Univ., vol. 36, pp. 165-251, 1901. (Contributions, no. 123.)]

[Footnote 243: Das Verhalten wirbelloser Thiere auf der Drehscheibe,
Zeitsch. f. Psych. und Physiol. d. Sinnesorgane, vol. 3, pp. 185-192,
1892.]

[Footnote 244: Weitere Beiträge zur Physiologie des Ohrenlabyrinthes,
II. Mittheilung, Versuche an Krebsen, Sitzungsb. Kais. Akad. Wiss.,
Wien., vol. 102 (Part. 3), pp. 149-174, 1893.]

[Footnote 245: Ueber die Bedeutung der Otolithenorgane für die
geotropischen Functionen von Astacus fluviatilis, Arch. f. d. ges.
Physiol., vol. 54, pp. 531-537, 1893.]

[Footnote 246: Das Nervensystem von Carcinus mænas, I. Arch. f. mikros.
Anat., vol. 50, pp. 460-547, 589-640, 1897.]

[Footnote 247: Contribution to the Comparative Physiology of
Compensatory Movements, Amer. Jour. of Physiol., vol. 3, pp. 86-114,
1899.]

[Footnote 248: The Reactions of Copepods to Various Stimuli, Bull. U.
S. Fish Comm., vol. 21, pp. 103-123, 1902.]

[Footnote 249: General Physiology, English translation by Frederic S.
Lee, 1899.]

[Footnote 250: Recherches pour servir à l'histoire des systèmes
nerveux, musculaire, et glandulaire de l'écrevisse, Ann. des Sci. Nat.,
series 5, vol. 9, pp. 99-280; vol. 10, pp. 5-54, 1868.]

[Footnote 251: The Sense of Touch in Astacus, Proc. Roy. Physiol. Soc.
Edinburgh, vol. 9, pp. 151-179, 1886.]


    PRINTED AT THE RIVERSIDE PRESS
    H. O. HOUGHTON & CO.
    CAMBRIDGE, MASS., U. S. A.

       *       *       *       *       *
    +----------------------------------------------------------------+
    |                                                                |
    |                         Transcriber notes:                     |
    |                                                                |
    | Please note, _underscore_ indicates italics, equals sign around|
    |      the word indicates bold.                                  |
    | Fixed various punctuation.                                     |
    | P.47. 'to +5.93' may be 'to +:+5.93', however leaving it.      |
    | P.142. The Standard for F calculation should be 110.40 total,  |
    |  not 122.40, with the average 3.94, not 4.35. Corrected table. |
    | P.47. 'for eyes at 0' added a degree like the others.          |
    | P.272. 'abcab, cabe', changed 'cabe' to 'cabc 'as there is no  |
    |  'e' and it looks to be 'c'.                                   |
    | P.143. The Standards for Active M. Total calculation should    |
    |   be 65.40, not 59.40, with the average 4.67, not 4.24.        |
    |   The average for Total Passive M. should be 71.00 not 72.00.  |
    |   Corrected table.                                             |
    | P.144. The Standards for F. Total calculation should be 120.11,|
    |   not 117.04, with the average 4.45, not 4.33. The average for |
    |   M. should be 4.19 not 4.20. Corrected table.                 |
    | P.145. Correct the net results on table IV.                    |
    |   First table:                                                 |
    | Raised or lowered #25 2.1 to 2.7. #27 8.9 to 8.7. #28 3.5 to 4.|
    | Totals: 131 total is 132. 48.1 total is 47.9. 270 to 271.1.    |
    | Net lowered 221.9 change to 223.2.                             |
    | Average lowering of each color judgment, 221.9/756 = .293      |
    |        change to:  223.2/756 = .295                            |
    | % of judgments of color not affected, 131/756   = 17+ is       |
    |        132/756 = 17+.                                          |
    |                                                                |
    |   Second table:                                                |
    | Raised or lowered, #8 36.3 is 36.4. #14 3.72 is 37.2.          |
    | Total Raised 829.7 is 749.8, Net lowered 820.9 is 741.         |
    | Average lowering of each tone judgment,  1.08+ is .98+         |
    |                                                                |
    | P.146. Table V. 2nd.--Lowered #18 8.2 is 8.6, change the       |
    |        total to 20.4.                                          |
    | Net lowered change from 581.3 to 560.9                         |
    |                                                                |
    | P.147.                                                         |
    |1st table: The Net + raised should be 640.5, not 649.5; with    |
    |  the 'Average raising of each tone judgment,' being 1.69, not  |
    |  1.71. Changed in table.                                       |
    |2nd table: -3.3 Net result lowered in #11 moved to              |
    |      #12 and is now +3.3. No figures affected.                 |
    |                                                                |
    | P.149. Table VII. 1st table.                                   |
    | #15 Net result + 1 should be - 1. Moved. This does not affect  |
    |     negative figures.                                          |
    | However, the + figure total is 197.5, not 186.5.               |
    | Making the net raised 182.3.                                   |
    | Averaging raising of each tone judgment, .45 is .48+           |
    |                                                                |
    | P.150. Table VIII.                                             |
    | 1st table. #16. 1 Raised is 1 Lowered. Moved. #21 Raised 2.2   |
    |     is 3.2, changed.                                           |
    | Total Net Result needs to be changed from 11. to 12.           |
    | Total Raised net result needs to be changed from 211.7         |
    |   to 221.9, making Net raised = 209.9.                         |
    | Average raising of each tone judgment, .52 is .55+             |
    |                                                                |
    | P.151. Table IX.                                               |
    | 2nd table. #13. raised 41.6 is 41.8. Changed.                  |
    | Total for Net Result Raised 130.0 is 150.2.        .           |
    | Net raised Grand total is 88.8.                                |
    | Average raising of each active touch judgment, .17+ is .22+    |
    |                                                                |
    | P.152. Table X.                                                |
    | 1st table. #4 Net result raised, 4.3 is 4.6 = Total 57.        |
    | Net lowered is 54.4.                                           |
    |                                                                |
    | P.153. Table XI. First table. No. of colour 10-17 illegible.   |
    |        corrected.                                              |
    | Points raised total & net raised 282.2 is 292.2.               |
    | Average raising of each color judgment, .72- is .75-.          |
    |                                                                |
    | 2nd. table. #2. Net Result - 13 changed to 11.                 |
    | Total Net Result - 56 is 54.                                   |
    | Net raised, 104 is 106.                                        |
    | Avererage .26+ is .27+.                                        |
    |                                                                |
    | P.154. Table XII.                                              |
    | First table. Total of column Raised, 226 change  to 22         |
    |              Total of column Lowered, 100 change to 101        |
    |                                                                |
    | P.313. 'obseved' changed to observed.                          |
    | Table I. Obs B row, 13.9 sec., taken out sec.                  |
    |                                                                |
    | P.400. (b) 'irrevelant' changed to 'irrelevant'.               |
    | Footnote 128: 'Wahrscheinlichkeitseechnung' changed to         |
    |               'Wahrscheinlichkeitsrechnung'                    |
    | P.463. moved 0 & 0 up one column in L. row of table from the   |
    |   percentage line                                              |
    | P.499. Table. 23 Huggins. c. N. 98 first column '+55' changed  |
    |   to +'5.5'.                                                   |
    | P.550. Added closing quotation at: is markedly thinner."       |
    | P.551. 'under expermental' changed to 'under experimental'.    |
    | Footnote 141. 'Woods Holl' changed to 'Woods Hole'.            |
    | P.583.  'thoughout' changed to 'throughout'.                   |
    | P.594. Fig. 4. 'The rise curve' changed to 'The rise on curve'.|
    | P.606. 'mimimize' changed to 'minimize'.                       |
    | P.638. Table XI. No. II, first row, 90deg: the minus was added.|
    | Note: Carat followed by { } indicates superscription.          |
    |                                                                |
    +----------------------------------------------------------------+