Produced by Chris Curnow, Joseph Cooper and the Online
Distributed Proofreading Team at http://www.pgdp.net.









  CONTRIBUTIONS FROM
  THE MUSEUM OF HISTORY AND TECHNOLOGY:
  PAPER 19


  ELEVATOR SYSTEMS
  OF THE EIFFEL TOWER, 1889

  _Robert M. Vogel_


  PREPARATORY WORK FOR THE TOWER             4

  THE TOWER'S STRUCTURAL RATIONALE           5

  ELEVATOR DEVELOPMENT BEFORE THE TOWER      6

  THE TOWER'S ELEVATORS                     20

  EPILOGUE                                  37




ELEVATOR SYSTEMS of the EIFFEL TOWER, 1889

By Robert M. Vogel

     _This article traces the evolution of the powered passenger elevator
     from its initial development in the mid-19th century to the
     installation of the three separate elevator systems in the Eiffel
     Tower in 1889. The design of the Tower's elevators involved problems
     of capacity, length of rise, and safety far greater than any
     previously encountered in the field; and the equipment that resulted
     was the first capable of meeting the conditions of vertical
     transportation found in the just emerging skyscraper._

     THE AUTHOR: _Robert M. Vogel is associate curator of mechanical and
     civil engineering, United States National Museum, Smithsonian
     Institution._


The 1,000-foot tower that formed the focal point and central feature of
the Universal Exposition of 1889 at Paris has become one of the best known
of man's works. It was among the most outstanding technological
achievements of an age which was itself remarkable for such achievements.

Second to the interest shown in the tower's structural aspects was the
interest in its mechanical organs. Of these, the most exceptional were the
three separate elevator systems by which the upper levels were made
accessible to the Exposition visitors. The design of these systems
involved problems far greater than had been encountered in previous
elevator work anywhere in the world. The basis of these difficulties was
the amplification of the two conditions that were the normal determinants
in elevator design--passenger capacity and height of rise. In addition,
there was the problem, totally new, of fitting elevator shafts to the
curvature of the Tower's legs. The study of the various solutions to these
problems presents a concise view of the capabilities of the elevator art
just prior to the beginning of the most recent phase of its development,
marked by the entry of electricity into the field.

The great confidence of the Tower's builder in his own engineering ability
can be fully appreciated, however, only when notice is taken of one
exceptional way in which the project differed from works of earlier
periods as well as from contemporary ones. In almost every case, these
other works had evolved, in a natural and progressive way, from a
fundamental concept firmly based upon precedent. This was true of such
notable structures of the time as the Brooklyn Bridge and, to a lesser
extent, the Forth Bridge. For the design of his tower, there was virtually
no experience in structural history from which Eiffel could draw other
than a series of high piers that his own firm had designed earlier for
railway bridges. It was these designs that led Eiffel to consider the
practicality of iron structures of extreme height.


[Illustration: Figure 1.--The Eiffel Tower at the time of the Universal
Exposition of 1889 at Paris. (From _La Nature_, June 29, 1889, vol. 17, p.
73.)]

[Illustration: Figure 2.--Gustave Eiffel (1832-1923). (From Gustave
Eiffel, _La Tour de Trois Cents Mètres_, Paris, 1900, frontispiece.)]


There was, it is true, some inspiration to be found in the paper projects
of several earlier designers--themselves inspired by that compulsion which
throughout history seems to have driven men to attempt the erection of
magnificently high structures.

One such inspiration was a proposal made in 1832 by the celebrated but
eccentric Welsh engineer Richard Trevithick to erect a 1,000-foot,
conical, cast-iron tower (fig. 3) to celebrate the passing of the Reform
Bill. Of particular interest in light of the present discussion was
Trevithick's plan to raise visitors to the summit on a piston, driven
upward within the structure's hollow central tube by compressed air. It
probably is fortunate for Trevithick's reputation that his plan died
shortly after this and the project was forgotten.

One project of genuine promise was a tower proposed by the eminent
American engineering firm of Clarke, Reeves & Company to be erected at the
Centennial Exhibition at Philadelphia in 1876. At the time, this firm was
perhaps the leading designer and erector of iron structures in the United
States, having executed such works as the Girard Avenue Bridge over the
Schuylkill at Fairmount Park, and most of New York's early elevated
railway system. The company's proposal (fig. 4) for a 1,000-foot shaft of
wrought-iron columns braced by a continuous web of diagonals was based
upon sound theoretical knowledge and practical experience. Nevertheless,
the natural hesitation that the fair's sponsors apparently felt in the
face of so heroic a scheme could not be overcome, and this project also
remained a vision.




Preparatory Work for the Tower


In the year 1885, the Eiffel firm, which also had an extensive background
of experience in structural engineering, undertook a series of
investigations of tall metallic piers based upon its recent experiences
with several lofty railway viaducts and bridges. The most spectacular of
these was the famous Garabit Viaduct (1880-1884), which carries a railroad
some 400 feet above the valley of the Truyere in southern France. While
the 200-foot height of the viaduct's two greatest piers was not startling
even at that period, the studies proved that piers of far greater height
were entirely feasible in iron construction. This led to the design of a
395-foot pier, which, although never incorporated into a bridge, may be
said to have been the direct basis for the Eiffel Tower.

Preliminary studies for a 300-meter tower were made with the 1889 fair
immediately in mind. With an assurance born of positive knowledge, Eiffel
in June of 1886 approached the Exposition commissioners with the project.
There can be no doubt that only the singular respect with which Eiffel was
regarded not only by his profession but by the entire nation motivated the
Commission to approve a plan which, in the hands of a figure of less
stature, would have been considered grossly impractical.

Between this time and commencement of the Tower's construction at the end
of January 1887, there arose one of the most persistently annoying of the
numerous difficulties, both structural and social, which confronted Eiffel
as the project advanced. In the wake of the initial enthusiasm--on the
part of the fair's Commission inspired by the desire to create a monument
to French technological achievement, and on the part of the majority of
Frenchmen by the stirring of their imagination at the magnitude of the
structure--there grew a rising movement of disfavor. The nucleus was, not
surprisingly, formed mainly of the intelligentsia, but objections were
made by prominent Frenchmen in all walks of life. The most interesting
point to be noted in a retrospection of this often violent opposition was
that, although the Tower's every aspect was attacked, there was remarkably
little criticism of its structural feasibility, either by the engineering
profession or, as seems traditionally to be the case with bold and
unprecedented undertakings, by large numbers of the technically uninformed
laity. True, there was an undercurrent of what might be characterized as
unease by many property owners in the structure's shadow, but the most
obstinate element of resistance was that which deplored the Tower as a
mechanistic intrusion upon the architectural and natural beauties of
Paris. This resistance voiced its fury in a flood of special newspaper
editions, petitions, and manifestos signed by such lights of the fine and
literary arts as De Maupassant, Gounod, Dumas _fils_, and others. The
eloquence of one article, which appeared in several Paris papers in
February 1887, was typical:

     We protest in the name of French taste and the national art culture
     against the erection of a staggering Tower, like a gigantic kitchen
     chimney dominating Paris, eclipsing by its barbarous mass Notre Dame,
     the Sainte-Chapelle, the tower of St. Jacques, the Dôme des
     Invalides, the Arc de Triomphe, humiliating these monuments by an act
     of madness.[1]

Further, a prediction was made that the entire city would become
dishonored by the odious shadow of the odious column of bolted sheet iron.

It is impossible to determine what influence these outcries might have had
on the project had they been organized sooner. But inasmuch as the
Commission had, in November 1886, provided 1,500,000 francs for its
commencement, the work had been fairly launched by the time the
protestations became loud enough to threaten and they were ineffectual.

Upon completion, many of the most vigorous protestants became as vigorous
in their praise of the Tower, but a hard core of critics continued for
several years to circulate petitions advocating its demolition by the
government. One of these critics, it was said--probably apocryphally--took
an office on the first platform, that being the only place in Paris from
which the Tower could not be seen.


[Illustration: Figure 3.--Trevithick's proposed cast-iron tower (1832)
would have been 1,000 feet high, 100 feet in diameter at the base, 12 feet
at the top, and surmounted by a colossal statue. (From F. Dye, _Popular
Engineering_, London, 1895, p. 205.)]




The Tower's Structural Rationale


During the previously mentioned studies of high piers undertaken by the
Eiffel firm, it was established that as the base width of these piers
increased in proportion to their height, the diagonal bracing connecting
the vertical members, necessary for rigidity, became so long as to be
subject to high flexural stresses from wind and columnar loading. To
resist these stresses, the bracing required extremely large sections which
greatly increased the surface of the structure exposed to the wind, and
was, moreover, decidedly uneconomical. To overcome this difficulty, the
principle which became the basic design concept of the Tower was
developed.

The material which would otherwise have been used for the continuous
lattice of diagonal bracing was concentrated in the four corner columns of
the Tower, and these verticals were connected only at two widely
separated points by the deep bands of trussing which formed the first and
second platforms. A slight curvature inward was given to the main piers to
further widen the base and increase the stability of the structure. At a
point slightly above the second platform, the four members converged to
the extent that conventional bracing became more economical, and they were
joined.


[Illustration: Figure 4.--The proposed 1,000-foot iron tower designed by
Clarke, Reeves & Co. for the Centennial Exhibition of 1876 at
Philadelphia. (From _Scientific American_, Jan. 24, 1874, vol. 30, p.
47.)]


That this theory was successful not only practically, but visually, is
evident from the resulting work. The curve of the legs and the openings
beneath the two lower platforms are primarily responsible for the Tower's
graceful beauty as well as for its structural soundness.

The design of the Tower was not actually the work of Eiffel himself but of
two of his chief engineers, Emile Nouguier (1840-?) and Maurice Koechlin
(1856-1946)--the men who had conducted the high pier studies--and the
architect Stéphen Sauvestre (1847-?).

In the planning of the foundations, extreme care was used to ensure
adequate footing, but in spite of the Tower's light weight in proportion
to its bulk, and the low earth pressure it exerted, uneven pier settlement
with resultant leaning of the Tower was considered a dangerous
possibility.[2] To compensate for this eventuality, a device was used
whose ingenious directness justifies a brief description. In the base of
each of the 16 columns forming the four main legs was incorporated an
opening into which an 800-ton hydraulic press could be placed, capable of
raising the member slightly. A thin steel shim could then be inserted to
make the necessary correction (fig. 5). The system was used only during
construction to overcome minor erection discrepancies.

In order to appreciate fully the problem which confronted the Tower's
designers and sponsors when they turned to the problem of making its
observation areas accessible to the fair's visitors, it is first necessary
to investigate briefly the contemporary state of elevator art.




Elevator Development before the Tower


While power-driven hoists and elevators in many forms had been used since
the early years of the 19th century, the ever-present possibility of
breakage of the hoisting rope restricted their use almost entirely to the
handling of goods in mills and warehouses.[3] Not until the invention of a
device which would positively prevent this was there much basis for work
on other elements of the system. The first workable mechanism to prevent
the car from dropping to the bottom of the hoistway in event of rope
failure was the product of Elisha G. Otis (1811-1861), a mechanic of
Yonkers, New York. The invention was made more or less as a matter of
course along with the other machinery for a new mattress factory of which
Otis was master mechanic.


[Illustration: Figure 5.--Correcting erection discrepancies by raising
pier member--with hydraulic press and hand pump--and inserting shims.
(From _La Nature_, Feb. 18, 1888, vol. 16, p. 184.)]

[Illustration: Figure 6.--The promenade beneath the Eiffel Tower, 1889.
(From _La Nature_, Nov. 30, 1889, vol. 17, p. 425.)]

[Illustration: Figure 7.--Teagle elevator in an English mill about 1845.
Power was taken from the line shafting. (From _Pictorial Gallery of Arts_,
Volume of Useful Arts, London, n.d. [ca. 1845].)]


The importance of this invention soon became evident to Otis, and he
introduced his device to the public three years later during the second
season of the New York Crystal Palace Exhibition, in 1854. Here he would
demonstrate dramatically the perfect safety of his elevator by cutting the
hoisting rope of a suspended platform on which he himself stood, uttering
the immortal words which have come to be inseparably associated with the
history of the elevator--"All safe, gentlemen!"[4]

The invention achieved popularity slowly, but did find increasing favor in
manufactories throughout the eastern United States. The significance of
Otis' early work in this field lay strictly in the safety features of his
elevators rather than in the hoisting equipment. His earliest systems were
operated by machinery similar to that of the teagle elevator in which the
hoisting drum was driven from the mill shafting by simple fast and loose
pulleys with crossed and straight belts to raise, lower, and stop. This
scheme, already common at the time, was itself a direct improvement on the
ancient hand-powered drum hoist.

The first complete elevator machine in the United States, constructed in
1855, was a complex and inefficient contrivance built around an
oscillating-cylinder steam engine. The advantages of an elevator system
independent of the mill drive quickly became apparent, and by 1860
improved steam elevator machines were being produced in some quantity, but
almost exclusively for freight service. It is not clear when the first
elevator was installed explicitly for passenger service, but it was
probably in 1857, when Otis placed one in a store on Broadway at Broome
Street in New York.

In the decade following the Civil War, tall buildings had just begun to
emerge; and, although the skylines of the world's great cities were still
dominated by church spires, there was increasing activity in the
development of elevator apparatus adapted to the transportation of people
as well as of merchandise. Operators of hotels and stores gradually became
aware of the commercial advantages to be gained by elevating their patrons
even one or two floors above the ground, by machinery. The steam engine
formed the foundation of the early elevator industry, but as building
heights increased it was gradually replaced by hydraulic, and ultimately
by electrical, systems.


THE STEAM ELEVATOR

The progression from an elevator machine powered by the line shafting of a
mill to one in which the power source was independent would appear a
simple and direct one. Nevertheless, it was about 40 years after the
introduction of the powered elevator before it became common to couple
elevator machines directly to separate engines. The multiple belt and
pulley transmission system was at first retained, but it soon became
evident that a more satisfactory service resulted from stopping and
reversing the engine itself, using a single fixed belt to connect the
engine and winding mechanism. Interestingly, the same pattern was followed
40 years later when the first attempts were made to apply the electric
motor to elevator drive.


[Illustration: Figure 8.--In the typical steam elevator machine two
vertical cylinders were situated either above or below the crankshaft, and
a small pulley was keyed to the crankshaft. In a light-duty machine, the
power was transmitted by flatbelt from the small pulley to a larger one
mounted directly on the drum. In heavy-duty machines, spur gearing was
interposed between the large secondary pulley and the winding drum. (Photo
courtesy of Otis Elevator Company.)]

[Illustration: Figure 9.--Several manufacturers built steam machines in
which a gear on the drum shaft meshed directly with a worm on the
crankshaft. This arrangement eliminated the belt, and, since the drum
could not drive the engine through the worm gearing, no brake was
necessary for holding the load. (Courtesy of Otis Elevator Company.)]


By 1870 the steam elevator machine had attained its ultimate form, which,
except for a number of minor refinements, was to remain unchanged until
the type became completely obsolete toward the end of the century.

By the last quarter of the century, a continuous series of improvements in
the valving, control systems, and safety features of the steam machine had
made possible an elevator able to compete with the subsequently appearing
hydraulic systems for freight and low-rise passenger service insofar as
smoothness, control, and lifting power were concerned. However, steam
machinery began to fail in this competition as the increasing height of
buildings rapidly extended the demands of speed and length of rise.

The limitation in rise constituted the most serious shortcoming of the
steam elevator (figs. 8-10), an inherent defect that did not exist in the
various hydraulic systems.


[Illustration: Figure 10.--Components of the steam passenger elevator at
the time of its peak development and use (1876). (From _The First One
Hundred Years_, Otis Elevator Company, 1953.)]


Since the only practical way in which the power of a steam engine could be
applied to the haulage of elevator cables was through a rotational system,
the cables invariably were wound on a drum. The travel or rise of the car
was therefore limited by the cable capacity of the winding drum. As
building heights increased, drums became necessarily longer and larger
until they grew so cumbersome as to impose a serious limitation upon
further upward growth. A drum machine rarely could be used for a lift of
more than 150 feet.[5]

Another organic difficulty existing in drum machines was the dangerous
possibility of the car--or the counterweight, whose cables often wound on
the drum--being drawn past the normal top limit and into the upper
supporting works. Only safety stops could prevent such an occurrence if
the operator failed to stop the car at the top or bottom of the shaft, and
even these were not always effective. Hydraulic machines were not
susceptible to this danger, the piston or plunger being arrested by the
ends of the cylinder at the extremes of travel.


THE HYDRAULIC ELEVATOR

The rope-geared hydraulic elevator, which was eventually to become known
as the "standard of the industry," is generally thought to have evolved
directly from an invention of the English engineer Sir William Armstrong
(1810-1900) of ordnance fame. In 1846 he developed a water-powered crane,
utilizing the hydraulic head available from a reservoir on a hill 200 feet
above.

The system was not basically different from the simple hydraulic press so
well known at the time. Water, admitted to a horizontal cylinder,
displaced a piston and rod to which a sheave was attached. Around the
sheave passed a loop of chain, one end of which was fixed, the other
running over guide sheaves and terminating at the crane arm with a lifting
hook. As the piston was pressed into the cylinder, the free end of the
chain was drawn up at triple the piston speed, raising the load. The
effect was simply that of a 3-to-1 tackle, with the effort and load
elements reversed. Simple valves controlled admission and exhaust of the
water. (See fig. 11.)


[Illustration: Figure 11.--Armstrong's hydraulic crane. The main cylinder
was inclined, permitting gravity to assist in overhauling the hook. The
small cylinder rotated the crane. (From John H. Jallings, _Elevators_,
Chicago, 1916, p. 82.)]


The success of this system initiated a sizable industry in England, and
the hydraulic crane, with many modifications, was in common use there for
many years. Such cranes were introduced in the United States in about 1867
but never became popular; they did, however, have a profound influence on
the elevator art, forming the basis of the third generic type to achieve
widespread use in this country.

The ease of translation from the Armstrong crane to an elevator system
could hardly have been more evident, only two alterations of consequence
being necessary in the passage. A guided platform or car was substituted
for the hook; and the control valves were connected to a stationary
endless rope that was accessible to an operator on the car.

The rope-geared hydraulic system (fig. 13) appeared in mature form in
about 1876. However, before it had become the "standard elevator" through
a process of refinement, another system was introduced which merits notice
if for no other reason than that its popularity for some years seems
remarkable in view of its preposterously unsafe design. Patented by Cyrus
W. Baldwin of Boston in January 1870, this system was termed the
Hydro-Atmospheric Elevator, but more commonly known as the water-balance
elevator (fig. 12). It employed water not under pressure but simply as
mass under the influence of gravity. The elevator car's supporting cables
ran over sheaves at the top of the shaft to a large iron bucket, which
traveled in a closed tube or well adjacent to and the same length as the
shaft. To raise the car, the operator caused a valve to open, filling the
bucket with water from a roof tank. When the weight of water was
sufficient to overbalance the loaded car, the bucket descended, raising
the car. On its ascent the car was stopped at intermediate floors by a
strong brake that gripped the guides. Upon reaching the top, the operator
was able to open a valve in the bucket, now at the bottom of its travel,
and discharge its contents into a basement tank, to be pumped back to the
roof. No longer counterbalanced, the car could descend, its speed
controlled solely by the brake.

The great popularity of this novel system apparently was due to its smooth
operation, high speed, simplicity, and economy of operation. Managed by a
skillful operator, it was capable of speeds far greater than other
systems could then achieve--up to a frightening 1,800 feet per minute.[6]


[Illustration: Figure 12.--Final development of the Baldwin-Hale water
balance elevator, 1873. The brake, kept applied by powerful springs, was
released only by steady pressure on a lever. There were two additional
controls--the continuous rope that opened the cistern valve to fill the
bucket, and a second lever to open the valve of the bucket to empty it.
(From _United States Railroad and Mining Register_, Apr. 12, 1873, vol.
17, p. 3.)]


In addition to the element of potential danger from careless operation or
failure of the brake, the Baldwin system was extremely expensive to
install as a result of the second shaft, which of course was required to
be more or less watertight.

Much of the water-balance elevator's development and refinement was done
by William E. Hale of Chicago, who also made most of the installations.
The system has, therefore, come to bear his name more commonly than
Baldwin's.

The popularity of the water-balance system waned after only a few years,
being eclipsed by more rational systems. Hale eventually abandoned it and
became the western agent for Otis--by this time prominent in the
field--and subsequently was influential in development of the hydraulic
elevator.

The rope-geared system of hydraulic elevator operation was so basically
simple that by 1880 it had been embraced by virtually all manufacturers.
However, for years most builders continued to maintain a line of steam and
belt driven machines for freight service. Inspired by the rapid increase
of taller and taller buildings, there was a concentrated effort,
heightened by severe competition, to refine the basic system.


[Illustration: Figure 13.--Vertical cylinder, rope-geared hydraulic
elevator with 2:1 gear ratio and rope control (about 1880). For higher
rises and speeds, ratios of up to 10:1 were used, and the endless rope was
replaced by a lever. (Courtesy of Otis Elevator Company.)]


By the late 1880's a vast number of improvements in detail had appeared,
and this form of elevator was considered to be almost without defect. It
was safe. Absence of a drum enabled the car to be carried by a number of
cables rather than by one or two, and rendered overtravel impossible. It
was fast. Control devices had received probably the most attention by
engineers and were as perfect and sensitive as was possible with
mechanical means. Cars with lever control could be run at the high speeds
required for high buildings, yet they could be stopped with a smoothness
and precision unattainable earlier with systems in which the valves were
controlled by an endless rope, worked by the operator. It was almost
completely silent, and when the cylinder was placed vertically in a well
near the shaft, practically no valuable floor space was occupied. But most
important, the length of rise was unlimited because no drum was used. As
greater rises were required, the multiplication of the ropes and sheaves
was simply increased, raising the piston-car travel ratio and permitting
the cylinder to remain of manageable length. The ratio was often as high
as 10 or 12 to 1, the car moving 10 or 12 feet to the piston's 1.

In addition to its principal advantages, the hydraulic elevator could be
operated directly from municipal water mains in the many cities where
there was sufficient pressure, thus eliminating a large investment in
tanks, pumps and boilers (fig. 14).

By far the greatest development in this specialized branch of mechanical
engineering occurred in the United States. The comparative position of
American practice, which will be demonstrated farther on, is indicated by
the fact that Otis Brothers and other large elevator concerns in the
United States were able to establish offices in many of the major cities
of Europe and compete very successfully with local firms in spite of the
higher costs due to shipment. This also demonstrates the extent of error
in the oft-heard statement that the skyscraper was the direct result of
the elevator's invention. There is no question that continued elevator
improvement was an essential factor in the rapid increase of building
heights. However, consideration of the situation in European cities, where
buildings of over 10 stories were (and still are) rare in spite of the
availability of similar elevator techniques, points to the fundamental
matter of tradition. The European city simply did not develop with the
lack of judicial restraint which characterized metropolitan growth in the
United States. The American tendency to confine mercantile activity to the
smallest possible area resulted in excessive land values, which drove
buildings skyward. The elevator followed, or, at most, kept pace with,
the development of higher buildings.


[Illustration: Figure 14.--In the various hydraulic systems, a pump was
required if pressure from water mains was insufficient to operate the
elevator directly. There was either a gravity tank on the roof or a
pressure tank in the basement. (From Thomas E. Brown, Jr., "The American
Passenger Elevator," _Engineering Magazine_ (New York), June 1893, vol. 5,
p. 340.)]


European elevator development--notwithstanding the number of American
rope-geared hydraulic machines sold in Europe in the 10 years or so
preceding the Paris fair of 1889--was confined mainly to variations on the
direct plunger type, which was first used in English factories in the
1830's. The plunger elevator (fig. 16), an even closer derivative of the
hydraulic press than Armstrong's crane, was nothing more than a platform
on the upper end of a vertical plunger that rose from a cylinder as water
was forced in.

There were two reasons for this European practice. The first and most
apparent was the rarity of tall buildings. The drilling of a well to
receive the cylinder was thus a matter of little difficulty. This well had
to be equivalent in depth to the elevator rise. The second reason was an
innate European distrust of cable-hung elevator systems in any form, an
attitude that will be discussed more fully farther on.


THE ELECTRIC ELEVATOR

At the time the Eiffel Tower elevators were under consideration, water
under pressure was, from a practical standpoint, the only agent capable of
fulfilling the power and control requirements of this particularly severe
service. Steam, as previously mentioned, had already been found wanting in
several respects. Electricity, on the other hand, seemed to hold promise
for almost every field of human endeavor. By 1888 the electric motor had
behind it a 10- or 15-year history of active development. Frank J. Sprague
had already placed in successful operation a sizable electric trolley-car
system, and was manufacturing motors of up to 20 horsepower in commercial
quantity. Lighting generators were being produced in sizes far greater.
There were, nevertheless, many obstacles preventing the translation of
this progress into machinery capable of hauling large groups of people a
vertical distance of 1,000 feet with unquestionable dependability.

The first application of electricity to elevator propulsion was an
experiment of the distinguished German electrician Werner von Siemens,
who, in 1880, constructed a car that successfully climbed a rack by means
of a motor and worm gearing beneath its deck (figs. 17, 18)--again, the
characteristic European distrust of cable suspension. However, the effect
of this success on subsequent development was negligible. Significant use
of electricity in this field occurred somewhat later, and in a manner
parallel to that by which steam was first applied to the elevator--the
driving of mechanical (belt driven) elevator machines by individual
motors. Slightly later came another application of the "conversion" type.
This was the simple substitution of electrically driven pumps (fig. 21)
for steam pumps in hydraulic installations. It will be recalled that pumps
were necessary in cases where water main pressure was insufficient to
operate the elevator directly.

In both of these cases the operational demands on the motor were of course
identical to those on the prime movers which they replaced; no reversal of
direction was necessary, the speed was constant, and the load was nearly
constant. Furthermore, the load could be applied to the motor gradually
through automatic relief valves on the pump and in the mechanical machines
by slippage as the belt was shifted from the loose to the fast pulleys.
The ultimate simplicity in control resulted from permitting the motor to
run continuously, drawing current only in proportion to its loading. The
direct-current motor of the 1880's was easily capable of such service, and
it was widely used in this way.


[Illustration: Figure 15.--Rope-geared hydraulic freight elevator using a
horizontal cylinder (about 1883). (From a Lane & Bodley illustrated
catalog of hydraulic elevators, Cincinnati, n.d.)]

[Illustration: Figure 16.--English direct plunger hydraulic elevator
(about 1895). (From F. Dye, _Popular Engineering_, London, 1895, p. 280.)]


Adaptation of the motor to the direct drive of an elevator machine was
quite another matter, the difficulties being largely those of control. At
this time the only practical means of starting a motor under load was by
introducing resistance into the circuit and cutting it out in a series of
steps as the speed picked up; precisely the method used to start traction
motors. In the early attempts to couple the motor directly to the winding
drum through worm gearing, this "notching up" was transmitted to the car
as a jerking motion, disagreeable to passengers and hard on machinery.
Furthermore, the controller contacts had a short life because of the
arcing which resulted from heavy starting currents. In all, such systems
were unsatisfactory and generally unreliable, and were held in disfavor by
both elevator experts and owners.


[Illustration: Figure 17.--Siemens' electric rack-climbing elevator of
1880. (From Werner von Siemens, _Gesammelte Abhandlungen und Vorträge_,
Berlin, 1881, pl. 5.)]


There was, moreover, little inducement to overcome the problem of control
and other minor problems because of a more serious difficulty which had
persisted since the days of steam. This was the matter of the drum and its
attendant limitations. The motor's action being rotatory, the winding drum
was the only practical way in which to apply its motive power to hoisting.
This single fact shut electricity almost completely out of any large-scale
elevator business until after the turn of the century. True, there was a
certain amount of development, after about 1887, of the electric
worm-drive drum machine for slow-speed, low-rise service (fig. 19). But
the first installation of this type that was considered practically
successful--in that it was in continuous use for a long period--was not
made until 1889,[7] the year in which the Eiffel Tower was completed.

Pertinent is the one nearly successful attempt which was made to approach
the high-rise problem electrically. In 1888, Charles R. Pratt, an elevator
engineer of Montclair, New Jersey, invented a machine based on the
horizontal cylinder rope-geared hydraulic elevator, in which the two sets
of sheaves were drawn apart by a screw and traveling nut. The screw was
revolved directly by a Sprague motor, the system being known as the
Sprague-Pratt. While a number of installations were made, the machine was
subject to several serious mechanical faults and passed out of use around
1900. Generally, electricity as a practical workable power for elevators
seemed to hold little promise in 1888.[8]


[Illustration: Figure 18.--Motor and drive mechanism of Siemens'
elevator. (From Alfred R. Urbanitzky, _Electricity in the Service of Man_,
London, 1886, p. 646.)]

[Illustration:

  _Morse, Williams & Co._,

  BUILDERS OF
  PASSENGER
  AND
  FREIGHT
  ELEVATORS.

  ELECTRIC ELEVATOR.

  Write us for Circulars and Prices.

  Main Office and Works, 1105 Frankford Avenue,
  PHILADELPHIA.


  New York Office,      108 Liberty Street.
  New Haven  "          82 Church Street.
  Pittsburg  "          413 Fourth Avenue.
  Boston Office         19 Pearl Street.
  Baltimore  "          Builders' Exchange.
  Scranton   "          425 Spruce Street.

Figure 19.--The electric elevator in its earliest commercial form (1891),
with the motor connected directly to the load. By this time, incandescent
lighting circuits in large cities were sufficiently extensive to make such
installations practical. However, capacity and lift were severely limited
by weaknesses of the control system and the necessity of using a drum.
(From _Electrical World_, Jan. 2, 1897, vol. 20, p. xcvii.)]

[Illustration:

  MILLER'S PATENT
  LIFE AND LABOR-SAVING
  SCREW HOISTING MACHINE,
  FOR THE USE OF
  Stores, Hotels, Warehouses, Factories, Sugar Refineries,
      Packing Houses, Mills, Docks, Mines, &c.
  MANUFACTURED BY
  CAMPBELL, WHITTIER & CO., ROXBURY, MASS.
  _Sole Agents for the New England States._

The above Engraving illustrates a very superior Hoisting Machine, designed
for _Store and Warehouse Hoisting_. It is very simple in its construction,
compact, durable, and not liable to get out of order. An examination of
the Engraving will convince any one who has any knowledge of Machinery,
that the screw is the only safe principle on which to construct a Hoisting
Machine or Elevator.

Figure 20.--Advertisement for the Miller screw-hoisting machine, about
1867 (see p. 23). From flyer in the United States National Museum.]

[Illustration: Figure 21.--The first widespread use of electricity in the
elevator field was to drive belt-type mechanical machines and the pumps of
hydraulic systems (see p. 14) as shown here. (From _Electrical World_,
Jan. 4, 1890, vol. 15, p. 4.)]




The Tower's Elevators


A great part of the Eiffel Tower's worth and its _raison d'être_ lay in
the overwhelming visual power by which it was to symbolize to a world
audience the scientific, artistic, and, above all, the technical
achievements of the French Republic. Another consideration, in Eiffel's
opinion, was its great potential value as a scientific observatory. At its
summit grand experiments and observations would be possible in such fields
as meteorology and astronomy. In this respect it was welcomed as a
tremendous improvement over the balloon and steam winch that had been
featured in this service at the 1878 Paris exposition. Experiments were
also to be conducted on the electrical illumination of cities from great
heights. The great strategic value of the Tower as an observation post
also was recognized. But from the beginning, sight was never lost of the
structure's great value as an unprecedented public attraction, and its
systematic exploitation in this manner played a part in its planning,
second perhaps only to the basic design.

The conveyance of multitudes of visitors to the Tower's first or main
platform and a somewhat lesser number to the summit was a technical
problem whose seriousness Eiffel must certainly have been aware of at the
project's onset. While a few visitors could be expected to walk to the
first or possibly second stage, 377 feet above the ground, the main means
of transport obviously had to be elevators. Indeed, the two aspects of the
Tower with which the Exposition commissioners were most deeply concerned
were the adequate grounding of lightning and the provision of a reliable
system of elevators, which they insisted be unconditionally safe.

To study the elevator problem, Eiffel retained a man named Backmann who
was considered an expert on the subject. Apparently Backmann originally
was to design the complete system, but he was to prove inadequate to the
task. As his few schemes are studied it becomes increasingly difficult to
imagine by what qualifications he was regarded as either an elevator
expert or designer by Eiffel and the Commission. His proposals appear,
with one exception, to have been decidedly retrogressive, and, further, to
incorporate the most undesirable features of those earlier systems he
chose to borrow from. Nothing has been discovered regarding his work, if
any, on elevators for the lower section of the Tower. Realizing the
difficulty of this aspect of the problem, he may not have attempted its
solution, and confined his work to the upper half where the structure
permitted a straight, vertical run.


[Illustration: Figure 22.--Various levels of the Eiffel Tower. (Adapted
from Gustave Eiffel, _La Tour de Trois Cents Mètres_, Paris, 1900, pl.
1.)]


The Backmann design for the upper elevators was based upon a principle
which had been attractive to many inventors in the mid-19th century period
of elevator development--that of "screwing the car up" by means of a
threaded element and a nut, either of which might be rotated and the other
remain stationary. The analogy to a nut and bolt made the scheme an
obvious one at that early time, but its inherent complexity soon became
equally evident and it never achieved practical success. Backmann
projected two cylindrical cars that traveled in parallel shafts and
balanced one another from opposite ends of common cables that passed over
a sheave in the upperworks. Around the inside of each shaft extended a
spiral track upon which ran rollers attached to revolving frames
underneath the cars. When the frames were made to revolve, the rollers,
running around the track, would raise or lower one car, the other
traveling in the opposite direction (fig. 23).


[Illustration: Figure 23.--Backmann's proposed helicoidal elevator for the
upper section of the Eiffel Tower. The cars were to be self-powered by
electric motors. Note similarity to the Miller system (fig. 20). (Adapted
from _The Engineer_ (London), Aug. 3, 1888, vol. 66, p. 101.)]


In the plan as first presented, a ground-based steam engine drove the
frames and rollers through an endless fly rope--traveling at high speed
presumably to permit it to be of small diameter and still transmit a
reasonable amount of power--which engaged pulleys on the cars. The design
was remarkably similar to that of the Miller Patent Screw Hoisting
Machine, which had had a brief life in the United States around 1865. The
Miller system (see p. 19) used a flat belt rather than a rope (fig. 20).
This plan was quickly rejected, probably because of anticipated
difficulties with the rope transmission.[9]

Backmann's second proposal, actually approved by the Commission,
incorporated the only--although highly significant--innovation evident in
his designs. For the rope transmission, electric motors were substituted,
one in each car to drive the roller frame directly. With this
modification, the plan does not seem quite as unreasonable, and would
probably have worked. However, it would certainly have lacked the
necessary durability and would have been extremely expensive. The
Commission discarded the whole scheme about the middle of 1888, giving two
reasons for its action: (1) the novelty of the system and the attendant
possibility of stoppages which might seriously interrupt the "exploitation
of the Tower," and (2) fear that the rollers running around the tracks
would cause excessive noise and vibration. Both reasons seem quite
incredible when the Backmann system is compared to one of those actually
used--the Roux, described below--which obviously must have been subject to
identical failings, and on a far greater scale. More likely there existed
an unspoken distrust of electric propulsion.

That the Backmann system should have been given serious consideration at
all reflects the uncertainty surrounding the entire matter of providing
elevator service of such unusual nature. Had the Eiffel Tower been erected
only 15 years later, the situation would have been simply one of
selection. As it was, Eiffel and the commissioners were governed not by
what they wanted but largely by what was available.


THE OTIS SYSTEM

The curvature of the Tower's legs imposed a problem unique in elevator
design, and it caused great annoyance to Eiffel, the fair's Commission,
and all others concerned. Since a vertical shaftway anywhere within the
open area beneath the first platform was esthetically unthinkable, the
elevators could be placed only in the inclined legs. The problem of
reaching the first platform was not serious. The legs were wide enough and
their curvature so slight in this lower portion as to permit them to
contain a straight run of track, and the service could have been designed
along the lines of an ordinary inclined railway. It was estimated that the
great majority of visitors would go only to this level, attracted by the
several international restaurants, bars and other features located there.
Two elevators to operate only that far were contracted for with no
difficulty--one to be placed in the east leg and one in the west.

To transport people to the second platform was an altogether different
problem. Since there was to be a single run from the ground, it would have
been necessary to form the elevator guides either with a constant
curvature, approximating that of the legs, or with a series of straight
chords connected by short segmental curves of small radius. Eiffel planned
initially to use the first method, but the second was adopted ultimately,
probably as being the simpler because only two straight lengths of run
were found to be necessary.

Bids were invited for two elevators on this basis--one each for the north
and south legs. Here the unprecedented character of the matter became
evident--there was not a firm in France willing to undertake the work. The
American Elevator Company, the European branch of Otis Brothers & Company,
did submit a proposal through its Paris office, Otis Ascenseur Cie., but
the Commission was compelled to reject it because a clause in the fair's
charter prohibited the use of any foreign material in the construction of
the Tower. Furthermore, there was a strong prejudice against foreign
contractors, which, because of the general background of disfavor
surrounding the project during its early stages, was an element worth
serious consideration by the Commission. The bidding time was extended,
and many attempts were made to attract a native design but none was
forthcoming.

As time grew short, it became imperative to resolve the matter, and the
Commission, in desperation, awarded the contract to Otis in July 1887 for
the amount of $22,500.[10] A curious footnote to the affair appeared much
later in the form of a published interview[11] with W. Frank Hall, Otis'
Paris representative:

     "Yes," said Mr. Hall, "this is the first elevator of its kind. Our
     people for thirty-eight years have been doing this work, and have
     constructed thousands of elevators vertically, and many on an
     incline, but never one to strike a radius of 160 feet for a distance
     of over 50 feet. It has required a great amount of preparatory study
     and we have worked on it for three years."

     "That was before you got the contract?"

     "Quite so, but we knew that, although the French authorities were
     very reluctant to give away this piece of work, they would be bound
     to come to us, and so we were preparing for them."

Such supreme confidence must have rapidly evaporated as events progressed.
Despite the invaluable advertising to be derived from an installation of
such distinction, the Otises would probably have defaulted had they
foreseen the difficulties which preceded completion of the work.


[Illustration: Figure 24.--General arrangement of Otis elevator system in
Eiffel Tower. (From _The Engineer_ (London), July 19, 1889, vol. 68, p.
58.)]


The proposed system (fig. 24) was based fundamentally upon Otis' standard
hydraulic elevator, but it was recognizable only in basic operating
principle (fig. 25). Tracks of regular rail section replaced the guides
because of the incline, and the double-decked cabin (fig. 29) ran on small
flanged wheels. This much of the apparatus was really not unlike that of
an ordinary inclined railway. Motive power was provided by the customary
hydraulic cylinder (fig. 26), set on an angle roughly equal to the incline
of the lower section of run. Balancing the cabin's dead weight was a
counterpoise carriage (fig. 27) loaded with pig iron that traveled on a
second set of rails beneath the main track. Like the driving system, the
counterweight was rope-geared, 3 to 1, so that its travel was about 125
feet to the cabin's 377 feet.


[Illustration: Figure 25.--Schematic diagram of the rigging of the Otis
system. (Adapted from Gustave Eiffel, _La Tour de Trois Cents Mètres_,
Paris, 1900, p. 127.)]


Everything about the system was on a scale far heavier than found in the
normal elevator of the type. The cylinder, of 38-inch bore, was 36 feet
long. Rather than a simple nest of pulleys, the piston rods pulled a large
guided carriage or "chariot" bearing six movable sheaves (fig. 28).
Corresponding were five stationary sheaves, the whole reeved to form an
immense 12-purchase tackle. The car, attached to the free ends of the
cables, was hauled up as the piston drew the two sheave assemblies apart.

In examining the system, it is difficult to determine what single element
in its design might have caused such a problem as to have been beyond the
engineering ability of a French firm, and to have caused such concern to a
large, well-established American organization of Otis' wide elevator and
inclined railway experience. Indeed, when the French system--which served
the first platform from the east and west legs--is examined, it appears
curious that a national technology capable of producing a machine at such
a level of complexity should have been unable to deal easily with the
entire matter. This can be plausibly explained only on the basis of
Europe's previously mentioned lack of experience with rope-geared and
other cable-hung elevator systems. The difficulty attending Otis' work,
usually true in the case of all innovations, lay unquestionably in the
multitudes of details--many of them, of course, invisible when only the
successfully working end product is observed.

More than a matter of detail was the Commission's demand for perfect
safety, which precipitated a situation typical of many confronting Otis
during the entire work. Otis had wished to coordinate the entire design
process through Mr. Hall, with technical matters handled by mail.
Nevertheless, at Eiffel's insistence, and with some inconvenience, in 1888
the company dispatched the project's engineer, Thomas E. Brown, Jr., to
Paris for a direct consultation. Mild conflict over minor details ensued,
but a gross difference of opinion arose ultimately between the American
and French engineers over the safety of the system. The disagreement
threatened to halt the entire project. In common with all elevators in
which the car hangs by cables, the prime consideration here was a means of
arresting the cabin should the cables fail. As originally presented to
Eiffel, the plans indicated an elaborate modification of the standard Otis
safety device--itself a direct derivative of E. G. Otis' original.

If any one of the six hoisting cables broke or stretched unduly, or if
their tension slackened for any reason, powerful leaf springs were
released causing brake shoes to grip the rails. The essential feature of
the design was the car's arrest by friction between its grippers and the
rails so that the stopping action was gradual, not sudden as in the
elevator safety. During proof trials of the safety, made prior to the
fair's opening by cutting away a set of temporary hoisting cables, the
cabin would fall about 10 feet before being halted.


[Illustration: Figure 26.--Section through the Otis power cylinder.
(Adapted from Gustave Eiffel, _La Tour de Trois Cents Mètres_, Paris,
1900, pl. 22.)]

[Illustration: Figure 27.--Details of the counterweight carriage in the
Otis system. (From Gustave Eiffel, _La Tour de Trois Cents Mètres_, Paris,
1900, pl. 22{4}.)]


Although highly efficient and of unquestionable security, this safety
device was considered an insufficient safeguard by Eiffel, who, speaking
in the name of the Commission, demanded the application of a device known
as the rack and pinion safety that was used to some extent on European cog
railways. The commissioners not only considered this system more reliable
but felt that one of its features was a necessity: a device that
permitted the car to be lowered by hand, even after failure of all the
hoisting cables. The serious shortcomings of the rack and pinion were its
great noisiness and the limitation it imposed on hoisting speed. Both
disadvantages were due to the constant engagement of a pinion on the car
with a continuous rack set between the rails. The meeting ended in an
impasse, with Brown unwilling to approve the objectionable apparatus and
able only to return to New York and lay the matter before his company.

While Eiffel's attitude in the matter may appear highly unreasonable, it
must be said that during a subsequent meeting between Brown and
Koechlin, the French engineer implied that a mutual antagonism had
arisen between the Tower's creator and the Commission. Thus, since his own
judgment must have had little influence with the commissioners at that
time, Eiffel was compelled to specify what he well knew were excessive
safety provisions.

This decision placed Otis Brothers in a decidedly uncomfortable position,
at the mercy of the Commission. W. E. Hale, promoter of the water balance
elevator--who by then had a strong voice in Otis' affairs--expressed the
seriousness of the matter in a letter to the company's president, Charles
R. Otis, following receipt of Brown's report on the Paris conference.
Referring to the controversial cogwheel, Hale wrote

     ... if this must be arranged so that the car is effected [sic] in its
     operation by constant contact with the rack and pinion ... so as to
     communicate the noise and jar, and unpleasant motion which such an
     arrangement always produces, I should favor giving up the whole
     matter rather than allying ourselves with any such abortion.... we
     would be the laughing stock of the world, for putting up such a
     contrivance.

This difficult situation apparently was the product of a somewhat general
contract phrased in terms of service to be provided rather than of
specific equipment to be used. This is not unusual, but it did leave open
to later dispute such ambiguous clauses as "adequate safety devices are to
be provided."

Although faced with the loss not only of all previously expended design
work but also of an advertisement of international consequence, the
company apparently concurred with Hale and so advised Paris.
Unfortunately, there are no Otis records to reveal the subsequent
transactions, but we may assume that Otis' threat of withdrawal prevailed,
coupled as it was with Eiffel's confidence in the American equipment. The
system went into operation as originally designed, free of the odious rack
and pinion.

That, unfortunately, was not the final disagreement. Before the fair's
opening in May 1889, the relationship was strained so drastically that a
mutually satisfactory conclusion to the project must indeed have seemed
hopeless. The numerous minor structural modifications of the Tower legs
found necessary as construction progressed had necessitated certain
equivalent alteration to the Otis design insofar as its dependency upon
the framework was affected. Consequently, work on the machinery was set
back by some months. Eiffel was informed that although everything was
guaranteed to be in full operation by opening day on May 1, the
contractual deadline of January 1 could not possibly be met. Eiffel, now
unquestionably acting on his own volition, responded by cable, refusing
all payment. Charles Otis' reply, a classic of indignation, disclosed to
Eiffel the jeopardy in which his impetuosity had placed the success of the
entire project:

     After all else we have borne and suffered and achieved in your
     behalf, we regard this as a trifle too much; and we do not hesitate
     to declare, in the strongest terms possible to the English language,
     that we will not put up with it ... and, if there is to be War, under
     the existing circumstances, propose that at least part of it shall be
     fought on American ground. If Mr. Eiffel shall, on the contrary,
     treat us as we believe we are entitled to be treated, under the
     circumstances, and his confidence in our integrity to serve him well
     shall be restored in season to admit of the completion of this work
     at the time wanted, well and good; but it must be done at once ...
     otherwise we shall ship no more work from this side, and Mr. Eiffel
     must charge to himself the consequences of his own acts.

This message apparently had the desired effect and the matter was somehow
resolved, as the machinery was in full operation when the Exposition
opened. The installation must have had immense promotional value for Otis
Brothers, particularly in its contrast to the somewhat anomalous French
system. This contrast evidently was visible to the technically
unsophisticated as well as to visiting engineers. Several newspapers
reported that the Otis elevators were one of the best American exhibits at
the fair.

In spite of their large over-all scale and the complication of the basic
pattern imposed by the unique situation, the Otis elevators performed well
and justified the original judgment and confidence which had prompted
Eiffel to fight for their installation. Aside from the obvious advantage
of simplicity when compared to the French machines, their operation was
relatively quiet, and fast.

The double car, traveling at 400 feet per minute, carried 40 persons, all
seated because of the change of inclination. The main valve or distributor
that controlled the flow of water to and from the driving cylinder was
operated from the car by cables. The hydraulic head necessary to produce
pressure within the cylinder was obtained from a large open reservoir on
the second platform. After being exhausted from the cylinder, the water
was pumped back up by two Girard pumps (fig. 31) in the engine room at
the base of the Tower's south leg.


THE SYSTEM OF ROUX, COMBALUZIER AND LEPAPE

There can be little doubt that the French elevators placed in the east and
west piers to carry visitors to the first stage of the Tower had the
important secondary function of saving face. That an engineer of Eiffel's
mechanical perception would have permitted their use, unless compelled to
do so by the Exposition Commission, is unthinkable. Whatever the attitudes
of the commissioners may have been, it must be said--recalling the
Backmann system--that they did not fear innovation. The machinery
installed by the firm of Roux, Combaluzier and Lepape was novel in every
respect, but it was a product of misguided ingenuity and set no precedent.
The system, never duplicated, was conceived, born, lived a brief and not
overly creditable life, and died, entirely within the Tower.

Basis of the French system was an endless chain of short, rigid,
articulated links (fig. 35), to one point of which the car was attached.
As the chain moved, the car was raised or lowered. Recalling the European
distrust of suspended elevators, it is interesting to note that the car
was pushed up by the links below, not drawn by those above, thus the
active links were in compression. To prevent buckling of the column, the
chain was enclosed in a conduit (fig. 36). Excessive friction was
prevented by a pair of small rollers at each of the knuckle joints between
the links. The system was, in fact, a duplicate one, with a chain on
either side of the car. At the bottom of the run the chains passed around
huge sprocket wheels, 12.80 feet in diameter, with pockets on their
peripheries to engage the joints. Smaller wheels at the top guided the
chains.

If by some motive force the wheel (fig. 33) were turned counterclockwise,
the lower half of the chain would be driven upward, carrying the car with
it. Slots on the inside faces of the lower guide trunks permitted passage
of the connection between the car and chain. Lead weights on certain links
of the chains' upper or return sections counterbalanced most of the car's
dead weight.


[Illustration: Figure 28.--Plan and section of the Otis system's movable
pulley assembly, or chariot. Piston rods are at left. (Adapted from _The
Engineer_ (London), July 19, 1889, vol. 68, p. 58.)]


Two horizontal cylinders rotated the driving sprockets through a mechanism
whose effect was similar to the rope-gearing of the standard hydraulic
elevator, but which might be described as chain gearing. The cylinders
were of the pushing rather than the pulling type used in the Otis system;
that is, the pressure was introduced behind the plungers, driving them
out. To the ends of the plungers were fixed smooth-faced sheaves, over
which were looped heavy quadruple-link pitch chains, one end of each being
solidly attached to the machine base. The free ends ran under the cylinder
and made another half-wrap around small sprockets keyed to the main drive
shaft. As the plungers were forced outward, the free ends of the chain
moved in the opposite direction, at twice the velocity and linear
displacement of the plungers. The drive sprockets were thereby revolved,
driving up the car. Descent was made simply by permitting the cylinders to
exhaust, the car dropping of its own weight. The over-all gear or ratio of
the system was the multiplication due to the double purchase of the
plunger sheaves times the ratio of the chain and drive sprocket diameters:
2(12.80/1.97) or about 13:1. To drive the car 218 feet to the first
platform of the Tower the plungers traveled only about 16.5 feet.

To penetrate the inventive rationale behind this strange machine is not
difficult. Aware of the fundamental dictum of absolute safety before all
else, the Roux engineers turned logically to the safest known elevator
type--the direct plunger. This type of elevator, being well suited to low
rises, formed the main body of European practice at the time, and in this
fact lay the further attraction of a system firmly based on tradition.
Since the piers between the ground and first platform could accommodate a
straight, although inclined run, the solution might obviously have been to
use an inclined, direct plunger. The only difficulty would have been that
of drilling a 220-foot, inclined well for the cylinder. While a difficult
problem, it would not have been insurmountable. What then was the reason
for using a design vastly more complex? The only reasonable answer that
presents itself is that the designers, working in a period before the
Otis bid had been accepted, were attempting to evolve an apparatus capable
of the complete service to the second platform. The use of a rigid direct
plunger thus precluded, it became necessary to transpose the basic idea in
order to adapt it to the curvature of the Tower leg, and at the same time
retain its inherent quality of safety. Continuing the conceptual sequence,
the idea of a plunger made in some manner flexible apparently suggested
itself, becoming the heart of the Roux machines.


[Illustration: Figure 29.--Section through cabin of the Otis elevator.
Note the pivoted floor-sections. As the car traveled, these floor-sections
were leveled by the operator to compensate for the change of inclination;
however, they were soon removed because they interfered with the loading
and unloading of passengers. (From _La Nature_, May 4, 1889, vol. 17, p.
360.)]


Here then was a design exhibiting strange contrast. It was on the one hand
completely novel, devised expressly for this trying service; yet on the
other hand it was derived from and fundamentally based on a thoroughly
traditional system. If nothing else, it was safe beyond question. In
Eiffel's own words, the Roux lifts "not only were safe, but appeared
safe; a most desirable feature in lifts traveling to such heights and
carrying the general public."[12]

The system's shortcomings could hardly be more evident. Friction resulting
from the more than 320 joints in the flexible pistons, each carrying two
rollers, plus that from the pitch chains must have been immense. The noise
created by such multiplicity of parts can only be imagined. Capacity was
equivalent to that of the Otis system. About 100 people could be carried
in the double-deck cabin, some standing. The speed, however, was only 200
feet per minute, understandably low.

If it had been the initial intention of the designers to operate their
cars to the second platform, they must shortly have become aware of the
impracticability of this plan, caused by an inherent characteristic of the
apparatus. As long as the compressive force acted along the longitudinal
axis of the links, there was no lateral resultant and the only load on the
small rollers was that due to the dead weight of the link itself. However,
if a curve had been introduced in the guide channels to increase the
incline of the upper run, as done by Otis, the force on those links
traversing the bend would have been eccentric--assuming the car to be in
the upper section, above the bend. The difference between the two sections
(based upon the Otis system) was 78°9' minus 54°35', or 23°34', the
tangent of which equals 0.436. Forty-three percent of the unbalanced
weight of the car and load would then have borne upon the, say, 12 sets of
rollers on the curve. The immense frictional load thus added to the entire
system would certainly have made it dismally inefficient, if not actually
unworkable.

In spite of Eiffel's public remarks regarding the safety of the Roux
machinery, in private he did not trouble to conceal his doubts. Otis'
representative, Hall, discussing this toward the end of Brown's previously
mentioned report, probably presented a fairly accurate picture of the
situation. His comments were based on conversations with Eiffel and
Koechlin:

     Mr. Gibson, Mr. Hanning [who were other Otis employees] and myself
     came to the unanimous conclusion that Mr. Eiffel had been forced to
     order those other machines, from outside parties, against his own
     judgment: and that he was very much in doubt as to their being a
     practical success--and was, therefore, all the more anxious to put in
     our machines (which he did have faith in) ... and if the others ate
     up coal in proportions greatly in excess of ours, he would have it to
     say ... "Gentlemen, these are my choice of elevators, those are yours
     &c." There was a published interview ... in which Eiffel stated ...
     that he was to meet some American gentlemen the following day, who
     were to provide him with elevators--grand elevators, I think he
     said....


[Illustration: Figure 30.--Upperworks and passenger platforms of the Otis
system at second level. (From _La Nature_, Aug. 10, 1889, vol. 17, p.
169.)]


The Roux and the Otis systems both drew their water supply from the same
tanks; also, each system used similar distributing valves (fig. 32)
operated from the cars. Although no reports have been found of actual
controlled tests comparing the efficiencies of the Otis and Roux systems,
a general quantitative comparison may be made from the balance figures
given for each (p. 40), where it is seen that 2,665 pounds of excess
tractive effort were allowed to overcome the friction of the Otis
machinery against 13,856 pounds for the Roux.


THE EDOUX SYSTEM

The section of the Tower presenting the least difficulty to elevator
installation was that above the juncture of the four legs--from the second
platform to the third, or observation, enclosure. There was no question
that French equipment could perform this service. The run being perfectly
straight and vertical, the only unusual demand upon contemporary elevator
technology was the length of rise--525 feet.

The system ultimately selected (fig. 37) appealed to the Commission
largely because of a similar one that had been installed in one tower of
the famous Trocadero[13] and which had been operating successfully for 10
years. It was the direct plunger system of Leon Edoux, and was, for the
time, far more rationally contrived than Backmann's helicoidal system.
Edoux, an old schoolmate of Eiffel's, had built thousands of elevators in
France and was possibly the country's most successful inventor and
manufacturer in the field. It is likely that he did not attempt to obtain
the contract for the elevator equipment in the Tower legs, as his
experience was based almost entirely on plunger systems, a type, as we
have seen, not readily adaptable to that situation. What is puzzling was
the failure of the Commission's members to recognize sooner Edoux's
obvious ability to provide equipment for the upper run. It may have been
due to their inexplicable confidence in Backmann.


[Illustration: Figure 31.--The French Girard pumps that supplied the Otis
and Roux systems. (From _La Nature_, Oct. 5, 1889, vol. 17, p. 292.)]


The direct plunger elevator was the only type in which European practice
was in advance of American practice at this time. Not until the beginning
of the 20th century, when hydraulic systems were forced into competition
with electrical systems, was the direct plunger elevator improved in
America to the extent of being practically capable of high rises and
speeds. Another reason for its early disfavor in the United States was the
necessity for drilling an expensive plunger well equal in length to the
rise.[14]

As mentioned, the most serious problem confronting Edoux was the extremely
high rise of 525 feet. The Trocadero elevator, then the highest plunger
machine in the world, traveled only about 230 feet. A secondary
difficulty was the esthetic undesirability of permitting a plunger
cylinder to project downward a distance equal to such a rise, which would
have carried it directly into the center of the open area beneath the
first platform (fig. 6). Both problems were met by an ingenious
modification of the basic system. The run was divided into two equal
sections, each of 262 feet, and two cars were used. One operated from the
bottom of the run at the second platform level to an intermediate platform
half-way up, while the other operated from this point to the observation
platform near the top of the Tower. The two sections were of course
parallel, but offset. A central guide, on the Tower's center-line, running
the entire 525 feet served both cars, with shorter guides on either
side--one for the upper and one for the lower run. Thus, each car traveled
only half the total distance. The two cars were connected, as in the
Backmann system, by steel cables running over sheaves at the top,
balancing each other and eliminating the need for counterweights. Two
driving rams were used. By being placed beneath the upper car, their
cylinders extended downward only the 262 feet to the second platform and
so did not project beyond the confines of the system itself.[15] In making
the upward or downward trip, the passengers had to change from one car to
the other at the intermediate platform, where the two met and parted (fig.
39). This transfer was the only undesirable feature of what was, on the
whole, a thoroughly efficient and well designed work of elevator
engineering.


[Illustration: Figure 32.--The Otis distributor, with valves shown in
motionless, neutral position. Since the main valve at all times was
subjected to the full operating pressure, it was necessary to drive this
valve with a servo piston. The control cable operated only the servo
piston's valve. (Adapted from Gustave Eiffel, _La Tour de Trois Cents
Mètres_, Paris, 1900, p. 130.)]

[Illustration: Figure 33.--General arrangement of the Roux Combaluzier and
Lepape elevator.]

[Illustration: Figure 34.--Roux, Combaluzier and Lepape machinery and
cabin at the Tower's base. (From _La Nature_, Aug. 10, 1889, vol. 17, p.
168.)]


In operation, water was admitted to the two cylinders from a tank on the
third platform. The resultant hydraulic head was sufficient to force out
the rams and raise the upper car. As the rams and car rose, the rising
water level in the cylinders caused a progressive reduction of the
available head. This negative effect was further heightened by the fact
that, as the rams moved upward, less and less of their length was
buoyed by the water within the cylinders, increasing their effective
weight. These two factors were, however, exactly compensated for by the
lengthening of the cables on the other side of the pulleys as the lower
car descended. Perfect balance of the system's dead load for any position
of the cabins was, therefore, a quality inherent in its design. However,
there were two extreme conditions of live loading which required
consideration: the lower car full and the upper empty, or vice versa. To
permit the upper car to descend under the first condition, the plungers
were made sufficiently heavy, by the addition of cast iron at their lower
ends, to overbalance the weight of a capacity load in the lower car. The
second condition demanded simply that the system be powerful enough to
lift the unbalanced weight of the plungers plus the weight of passengers
in the upper car.

As in the other systems, safety was a matter of prime importance. In this
case, the element of risk lay in the possibility of the suspended car
falling. The upper car, resting on the rams, was virtually free of such
danger. Here again the influence of Backmann was felt--a brake of his
design was applied (fig. 38). It was, true to form, a throwback, similar
safety devices having proven unsuccessful much earlier. Attached to the
lower car were two helically threaded vertical rollers, working within
the hollow guides. Corresponding helical ribs in the guides rotated the
rollers as the car moved. If the car speed exceeded a set limit, the
increased resistance offered by the apparatus drove the rollers up into
friction cups, slowing or stopping the car.


[Illustration: Figure 35.--Detail of links in the Roux system. (From
Gustave Eiffel, _La Tour de Trois Cents Mètres_, Paris, 1900, p. 156.)]

[Illustration: Figure 36.--Section of guide trunks in the Roux system.
(From Gustave Eiffel, _La Tour de Trois Cents Mètres_, Paris, 1900, p.
156.)]


The device was considered ineffectual by Edoux and Eiffel, who were aware
that the ultimate safety of the system resulted from the use of supporting
cables far heavier than necessary. There were four such cables, with a
total sectional area of 15.5 square inches. The total maximum load to
which the cables might be subjected was about 47,000 pounds, producing a
stress of about 3,000 pounds per square inch compared to a breaking stress
of 140,000 pounds per square inch--a safety factor of 46![16]


[Illustration: Figure 37.--Schematic diagram of the Edoux system. (Adapted
from Gustave Eiffel, _La Tour de Trois Cents Mètres_, Paris, 1900, p.
175.)]

[Illustration: Figure 38.--Vertical section through lower (suspended)
Edoux car, showing Backmann helicoidal safety brake. (Adapted from Gustave
Eiffel, _La Tour Eiffel en 1900_, Paris, 1902, p. 12.)]


A curiosity in connection with the Edoux system was the use of Worthington
(American) pumps (fig. 40) to carry the water exhausted from the cylinders
back to the supply tanks. No record has been found that might explain why
this particular exception was made to the "foreign materials" stipulation.
This exception is even more strange in view of Otis' futile request for
the same pumps and the fact that any number of native machines must have
been available. It is possible that Edoux's personal influence was
sufficient to overcome the authority of the regulation.


[Illustration: Figure 39.--Passengers changing cars on Edoux elevator at
intermediate platform. (From _La Nature_, May 4, 1889, vol. 17, p. 361.)]

[Illustration: Figure 40.--Worthington tandem compound steam pumps, at
base of the Tower's south pier, supplied water for the Edoux system. The
tank was at 896 feet, but suction was taken from the top of the cylinders
at 643 feet; therefore, the pumps worked against a head of only about 250
feet. (From _La Nature_, Oct. 5, 1889, vol. 17, p. 293.)]

[Illustration: Figure 41.--Recent view of lower car of the Edoux system,
showing slotted cylindrical guides that enclose the cables.]




Epilogue


In 1900, after the customary 11-year period, Paris again prepared for an
international exposition, about 5 years too early to take advantage of the
great progress made by the electric elevator. When the Roux machines, the
weakest element in the Eiffel Tower system, were replaced at this time, it
was by other hydraulics. Built by the well known French engineering
organization of Fives-Lilles, the new machines were the ultimate in power,
control, and general excellence of operation. As in the Otis system, the
cars ran all the way to the second platform.

The Fives-Lilles equipment reflected the advance of European elevator
engineering in this short time. The machines were rope-geared and
incorporated the elegant feature of self-leveling cabins which compensated
for the varying track inclination. For the 1900 fair, the Otis elevator in
the south pier was also removed and a wide stairway to the first platform
built in its place. In 1912, 25 years after Backmann's startling proposal
to use electricity for his system, the remaining Otis elevator was
replaced by a small electric one. This innovation was reluctantly
introduced solely for the purpose of accommodating visitors in the winter
when the hydraulic systems were shut down due to freezing weather. The
electric elevator had a short life, being removed in 1922 when the number
of winter visitors increased far beyond its capacity. However, the two
hydraulic systems were modified to operate in freezing
temperatures--presumably by the simple expedient of adding an
antifreezing chemical to the water--and operation was placed on a
year-round basis.

Today the two Fives-Lilles hydraulic systems remain in full use; and
visitors reach the Tower's summit by Edoux's elevator (fig. 41), which is
all that remains of the original installation.


BALANCE OF THE THREE ELEVATOR SYSTEMS

_The Otis System_

Negative effect

  Weight of cabin: 23,900 lb. × sin 78°9' (incline of upper run) 23,390 lb.
  Live load: 40 persons @150 lb. = 6,000 × sin 78°9'              5,872
                                                                 ------  -- 29,262 lb.

Positive effect

  Counterweight: 55,000 × sin 54°35' (incline of lower run)
                 ------------------------------------------
                            3 (rope gear ratio)                  14,940 lb.

  Weight of piston and chariot:  33,060 × sin 54°35'
                                 ------------------
                                     12 (ratio)                   2,245

  Power: 156 p.s.i. × 1,134 sq. in. (piston area)
         ----------------------------------------
                     12 (ratio)                                  14,742     31,927 lb.

Excess to overcome friction                                                  2,665 lb.


_The Roux, Combaluzier and Lepape System_

Negative effect

  Weight of cabin: 14,100 × sin 54°35'                           11,500 lb.
  Live load: 100 persons @150 lb. = 15,000 × sin 54°35'          12,220
                                                                 ------  -- 23,720 lb.

Positive effect

  Counterweight: 6,600 × sin 54°35'                               5,380

  Power: 156 p.s.i. × 2 (pistons) × 1,341.5 sq. in. (piston area)
         ------------------------------------------
                         13 (ratio)                              32,196     37,576 lb.
                                                                 ------     ----------
Excess to overcome friction                                                 13,856 lb.


_The Edoux System_

Negative effect

  Unbalanced weight of plungers (necessary to raise full
    lower car and weight of cables on lower side)                42,330 lb.

  Live load: 60 persons @150 lb.                                  9,000
                                                                 ------  -- 51,330 lb.


Positive effect

  Power: 227.5 p.s.i. × 2 (plungers) × 124 sq. in. (plunger area)           56,420 lb.
                                                                            ----------
  Excess to overcome friction                                                5,090 lb.




Footnotes:

[1] Translated from Jean A. Keim, _La Tour Eiffel_, Paris, 1950.

[2] The foundation footings exerted a pressure on the earth of about 200
pounds per square foot, roughly one-sixth that of the Washington Monument,
then the highest structure in the world.

[3] A type of elevator known as the "teagle" was in use in some multistory
English factories by about 1835. From its description, this elevator
appears to have been primarily for the use of passengers, but it
unquestionably carried freight as well. The machine shown in figure 7 had,
with the exception of a car safety, all the features of later systems
driven from line shafting--counterweight, control from the car, and
reversal by straight and crossed belts.

[4] The Otis safety, of which a modified form is still used, consisted
essentially of a leaf wagon spring, on the car frame, kept strained by the
tension of the hoisting cables. If these gave way, the spring, released,
drove dogs into continuous racks on the vertical guides, holding the car
or platform in place.

[5] A notable exception was the elevator in the Washington Monument.
Installed in 1880 for raising materials during the structure's final
period of erection and afterwards converted to passenger service, it was
for many years the highest-rise elevator in the world (about 500 feet),
and was certainly among the slowest, having a speed of 50 feet per minute.

[6] Today, although not limited by the machinery, speeds are set at a
maximum of about 1,400 feet per minute. If higher speeds were used, an
impractically long express run would be necessary for starting and
stopping in order to prevent an acceleration so rapid as to be
uncomfortable to passengers and a strain on the equipment.

[7] Two machines, by Otis, in the Demarest Building, Fifth Avenue and 33d
Street, New York. They were in use for over 30 years.

[8] Although the eventually successful application of electric power to
the elevator did not occur until 1904, and therefore goes beyond the
chronological scope of this discussion, it was of such importance insofar
as current practice is concerned as to be worthy of brief mention. In that
year the first gearless traction machine was installed by Otis in a
Chicago theatre. As the name implies, the cables were not wrapped on a
drum but passed, from the car, over a grooved sheave directly on the motor
shaft, the other ends being attached to the counterweights. The result was
a system of beautiful simplicity, capable of any rise and speed with no
proportionate increase in the number or size of its parts, and free from
any possibility of car or weights being drawn into the machinery. This
system is still the only one used for rises of over 100 feet or so. By the
time of its introduction, motor controls had been improved to the point of
complete practicability.

[9] Mechanical transmission of power by wire rope was a well developed
practice at this time, involving in many instances high powers and
distances up to a mile. To attempt this system in the Eiffel Tower,
crowded with structural work, machinery and people, was another matter.

[10] According to Otis Elevator Company, the final price, because of
extras, was $30,000.

[11] In _Pall Mall Gazette_, as quoted in _The Engineering and Building
Record and the Sanitary Engineer_, May 25, 1889, vol. 19, p. 345.

[12] From speech at annual summer meeting of Institution of Mechanical
Engineers, Paris, 1889. Quoted in _Engineering_, July 5, 1889, vol. 48, p.
18.

[13] Located near the Tower, built for the Paris fair of 1878.

[14] Improved oil-well drilling techniques were influential in the intense
but short burst of popularity enjoyed by direct plunger systems in the
United States between 1899 and 1910. In New York, many such systems of
200-foot rise, and one of 380 feet, were installed.

[15] An obvious question arises here: What prevents a plunger 200 or 300
feet long and no more than 16 inches in diameter from buckling under its
compressive loading? The answer is simply that most of this length is not
in compression but in tension. The Edoux rams, when fully extended,
virtually hung from the upper car, sustained by the weight of 500 feet of
cable on the other side of the sheaves. As the upper car descended this
effect diminished, but as the rams moved back into the cylinders their
unsupported length was correspondingly reduced.

[16] M. A. Ansaloni, "The Lifts in the Eiffel Tower," quoted in
_Engineering_, July 5, 1889, vol. 48, p. 23. The strength of steel when
drawn into wire is increased tremendously. Breaking stresses of 140,000
p.s.i. were not particularly high at the time. Special cables with
breaking stresses of up to 370,000 p.s.i. were available.




Transcriber's Notes:

Passages in italics are indicated by _underscore_.

The original was printed in two columns per page.

Illustrations have been moved to the nearest paragraph break.

The following misprints have been corrected:
  "Trevethick's" corrected to "Trevithick's" (page 5)
  "then" corrected to "than" (page 14)
  "smiliar" corrected to "similar" (page 31)