Produced by Chris Curnow, Chris Jordan and the Online
Distributed Proofreading Team at http://www.pgdp.net (This
file was produced from images generously made available
by The Internet Archive)









                               JACQUARD
                         WEAVING AND DESIGNING




                               JACQUARD
                         WEAVING AND DESIGNING


                                  BY
                              T. F. BELL

               NATIONAL SCHOLAR IN DESIGN (1875-8) AND
                 THIRD GRADE CERTIFICATED ART MASTER,
      SCIENCE AND ART DEPARTMENT, S. K.: MEDALLIST IN HONOURS AND
        CERTIFICATED TEACHER IN ‘LINEN MANUFACTURING,’ AND IN
 ‘WEAVING AND PATTERN DESIGNING,’ CITY AND GUILDS OF LONDON INSTITUTE


                                LONDON
                       LONGMANS, GREEN, AND CO.
                 AND NEW YORK: 15 EAST 16^{th} STREET
                                 1895

                         _All rights reserved_




PREFACE


The contents of the following pages have been derived from a long
course of art and technical training, together with a lengthened
practical experience in textile manufacturing and designing; during
which time I received much valuable assistance from many kind
friends--especially Mr. B. Ashenhurst, when master of the Belfast
Technical School, and his brother, of Bradford; also Mr. John Mitchell,
of Belfast and Glasgow--which I thankfully acknowledge.

To supply a treatise that would be alike suitable to the manufacturer,
the workman, and the student would be a difficult matter. To explain
each point so fully that it would be clear to every person would
make it wearisome to those having a knowledge of the subject. I
have therefore endeavoured to keep the descriptions as concise as
is compatible with a fairly clear explanation, which I hope will be
considered the wisest plan to adopt; and, while quite aware of the
many shortcomings of the work, I venture to hope that it will prove
a valuable assistance to those wishing to improve their knowledge of
jacquard weaving and designing.

                                                     T. F. BELL.
  Belfast, 1894.

It affords me much satisfaction to add a few words of introduction to
this work of my friend Mr. Bell. I cannot pretend to criticise his
explanation of technical processes, but, from an examination of the
proof-sheets, I am convinced that the book will be found an invaluable
aid to students of both art and technical schools. We already possess
numerous text-books on designing as an art, as also on the technical
processes of weaving, &c., and on the materials of manufacture. Mr.
Bell has aimed at bringing the technical and artistic sides of the
subject together in a practical form, and has thus provided us with a
valuable handbook.

Though we have a Government Department of Science and Art, a knowledge
of both branches is seldom united in the same individual. The artist
is too often deficient in the science of his craft; while the man of
science not unfrequently shows contempt for art. The author of this
volume is peculiarly qualified for the task he has set himself, being
a practical manufacturer as well as a designer of long experience, and
thoroughly learned in the science of his subject. His book will supply
a long-felt need.

                                              GEORGE TROWBRIDGE,
  _Head-Master Government School of Art, Belfast_.




CONTENTS


  CHAPTER                                     PAGE

     I. INTRODUCTION                             1

    II. JACQUARD MACHINES                       27

   III. FULL-HARNESS MOUNTINGS                  69

    IV. DESIGNING AND DRAUGHTING               110

     V. CARD-CUTTING AND LACING                139

    VI. SPECIAL JACQUARDS AND HARNESSES        155

   VII. GAUZE                                  196

  VIII. DOUBLE CLOTH                           233

    IX. TAPESTRY AND PILE WORK                 255

     X. CARPETS                                264

    XI. LAPPETS AND SWIVELS                    294

        INDEX                                  301




JACQUARD WEAVING AND DESIGNING




CHAPTER I

_INTRODUCTION_


Though the term ‘jacquard weaving’ is properly applied to work done by
the jacquard machine, it will here be taken to apply to all harness
weaving, or work that extends beyond the range of shafts, or leaves of
heddles.

The question arises, When is the limit to the number of shafts that
ought to be used reached? It apparently used to be when no more could
be got into the loom, as up to ninety-six shafts were used; and this
seems to be quite enough for any weaver to get the yarn through, or
for any loom to hold, but it must be remembered that at present the
appliances are much more suited to the work than they formerly were;
and now, except in woollen or worsted goods, where it is desirable to
use shafts on account of their firmness in comparison with that of a
harness, from twelve to sixteen shafts are as many as it is generally
thought desirable to have in a power loom. I have seen thirty-five
shafts, all in one tier or set, working diaper very conveniently in a
hand loom, and more than double that number of leaves working worsted
in a power loom; but whether the latter was desirable, or not, I must
leave to the judgment of the manufacturer who possessed it.

Many ingenious inventions have been made for the purpose of simplifying
the working of a large number of shafts, but as a description of them
would be out of place here, we may pass on to the draw loom, which
appears to be the first form of harness of which we have an accurate
description. How the cloths of Babylon were woven, in which

    Men’s figured counterfeits so like have been
    That if the party’s self had been in place,
    Yet Art would vie with Nature for the grace--

is not known, though in Gilroy’s report of Arphaxad’s description of
his loom to Deioces, king of the Medes, it is stated to have been
accomplished by means of carved blocks of wood acting on needles,
which wrought the harness or heddles and thus formed the pattern; but
as Gilroy has admitted that the introduction to his work on weaving
is a pure invention of his own, for the purpose of making it appear
that the Ancients were acquainted with motions similar to those on our
modern looms, or as a ‘take-off on those who angle hourly to surprise,
and bait their hooks with prejudice and lies,’ we need not dwell
further on the subject. In any case, figured cloths must have attained
considerable excellence in very early ages. The curtains of the
Tabernacle were embroidered with figures, and the veil of the Temple
was, according to Josephus, embroidered with all sorts of flowers, and
interwoven with various ornamental figures, the door curtain being
embroidered with blue and purple and scarlet. The ephod of the High
Priest was similarly embroidered.

[Illustration: Fig. 1]

The Egyptians worked coloured patterns in the loom so rich that they
vied with the Babylonian cloths, which were embroidered with the
needle. The method of working is unknown, but cloths taken from the
tombs in Egypt, which may be seen in South Kensington Museum and in the
Gobelins tapestry manufactory, Paris, appear to be made on a principle
similar to that of the Gobelins tapestry; the warp is of flax and the
weft of coloured wool: and the looms depicted on the catacombs in Egypt
are very similar in appearance to tapestry looms.

Embroidering was practised in Egypt prior to the Exodus of the
Israelites; and gold and silver threads or wires were used both for
embroidering and weaving, being known nearly 4,000 years.

The Babylonish garment taken by Achan, whose sin brought much woe upon
the Israelites, is said, by Josephus, to have been a Royal garment
woven entirely of gold; but it might only have been embroidered with
gold, and was probably wrought in the plain of Shinar, as it was not
till long after that Babylon was celebrated for its manufactures.

Pliny says that weaving cloth with gold thread was invented by Attalus,
an Asiatic king, and that the Babylonians were most noted for their
skill in weaving coloured cloths. This was in Homer’s time, about 900
B.C., when weaving and embroidering appear to have attained great
excellence, and to have been very gorgeous. At that time the labour
of the loom was considered an accomplishment, which ladies and even
princesses tried to excel in.

As before stated, the draw loom is the first form of machine for
figured weaving of which we have any record. It is not known where it
was invented, but it probably passed from China to Western Asia with
the silk manufacture. The ancient Egyptians, Greeks, and Romans do not
appear to have known it. The Chinese have still in use a draw loom in
which the drawboy stands on the top and draws up the parcels of twines
which have been previously arranged for him. After being established in
Damascus (hence the name damask), the draw loom passed on to Europe,
where the Chinese method of working was used till 1604, when M.
Simblot, in France, connected to the neck a separate series of cords,
called the ‘Simple’ (perhaps a corruption of his name), so that the
drawboy could work when standing at the side of the loom. It is said
to have been introduced into England in 1567. The next improvement was
to dispense with the drawboy’s services, and for this purpose a patent
was taken out in 1687, by Joseph Mason, for ‘a draw boy engine by which
a weaver may performe the whole worke of weaving such stuffe as the
greatest weaving trade in Norwich doth now depend upon without the help
of a draught boy.’ In 1779 William Cheape patented a plan to dispense
with the drawboy by having the ‘simple’ above his head, and drawing it
down with knots which were held in notches, as described in Fig. 2.

Before beginning to describe the draw loom it may be better first to
describe what it is required to do.

Its principal use appears to have been for the weaving of damask,
which is one of the simplest forms of figured weaving. Reduce a damask
texture to its elementary form, and it consists of twilling, or, more
correctly, turned or reversed twilling. If we take a common dice
pattern woven with shafts, it will easily be seen that one dice is
formed by a warp twill, and the next one by a weft twill, or that the
dices are formed by warp and weft twills alternately.

[Illustration: Fig. 2]

Now, what forms the pattern? The yarn may be all of one colour, the
threads may be so closely set together as to make them individually
invisible, or to appear as a plain surface, and yet the dices come
out distinctly in two shades of colour. The play of light on the
longitudinal and latitudinal threads produces this effect. The dices
formed by the latitudinal or horizontal threads will always appear
darker than the yarn in the cloth when the latter is placed between
the observer and the light, whether these threads be warp or weft, as
there is a certain amount of shade on each of them, and of shadow cast
by them, whereas the longitudinal or vertical threads are illuminated,
without any shade or shadow, and appear lighter than the yarn did
before being woven; and this is the reason why a good side light is
the best for showing up the pattern on damask, it developing the above
to the utmost. In a good material the difference of shade between the
ground and figure is very considerable, but in some thin, coarse goods
it is hardly visible, requiring them to be held in a favourable light
to show the distinction: the pattern will appear light on a darker
ground, or the reverse of this, according as the surface threads
forming it run across the light or in the direction of it.

This is the reason of the pattern appearing on the cloth; then it is
the business of the designer to regulate what form it is to partake of,
by preparing a suitable design; and according to instructions furnished
to him by the design, it is the duty of the drawboy to raise the warp
by regulating the cording of his harness, and drawing it so as to
reverse the twill from a weft one to a warp one wherever the figure is
to be formed on the cloth, and to do so in such order as to produce the
pattern required.

The draw-loom mounting consists of two parts--the drawboy mounting, or
the harness with its tail and simples, to be wrought by the drawboy;
and the shaft mounting, which is required to form the texture of the
cloth, or to interlace the warp and weft through both ground and
figures; the harness only interlaces them at the edges of the pattern,
or causes either warp or weft to be above, to form the figure _en
bloc_, but without interlacing them together.

It is therefore a ‘compound mounting,’ and is known as a ‘presser’ or
‘pressure’ harness. For simplicity’s sake let us suppose the principle
of the drawboy to be applied to shafts or healds, and take a simple
figure, as Fig. 1.

For it there are 5 parts, or it could be wrought with 5 leaves of
heddles with a straight draught. Fig. 2 shows the mounting; A A is the
back mounting, which in this case is a shaft mounting, but would be a
harness for a more extensive pattern. B, B are the pressure heddles
or front mounting. These are 5 in number, as the ground or texture is
taken as a 5-end satin or twill, C is the pulley box with the tail
cords, D, D, passing over the pulleys, and tied to the wall or to the
loom framing as at E. The knobs F hang over the weaver’s head, and are
attached by cords H, passing through a hole board G, to the tail cords,
D, D. There are heads on the cords H, and the holes in the hole board
are made thus [Illustration: keyhole]; so that when the weaver pulls
down a knob the bead can pass through the round hole, and the shaft
or shafts of the back mounting attached to it will be raised, and can
be kept in this position by drawing the cord into the notch or narrow
part of the hole, which the bead will not pass through. Any number of
shafts can be raised that are required to form the pattern, either by
pulling down the knob for each shaft, or by having the knots corded to
the shafts, so that each one will raise the proper number of shafts.
Thus, in the figure, each knob is only tied to one of the tail cords;
therefore a knob must be pulled down for each leaf of heddles to be
raised, but each knob might be attached to any number of the tail cords
according to the number of leaves of heddles it is required to raise,
so that pulling down each knob in succession will complete the pattern.
It might require too many knobs to do this, and then the former method
would have to be adopted. When the weaver begins to work he draws the
first figure shed with the back mounting by pulling down one or more
of the knobs as is required; he then works over the ground treadles,
b, b, till a change of pattern is required. Next he releases the drawn
shed by pulling the cord out of the notch in the hole board; draws
another shed, and works over the ground treadles as before. This gives
the principle of how the draw loom works, but the principle of forming
the texture with the back and front mountings combined will be fully
explained under ‘Pressure Harness.’

Fig. 3 gives the draw-loom harness; A, A is the carriage, or the rails
that support the harness, which rests on the capes or side rails of the
loom. Supported by the carriage is the pulley box P, which is a frame
fitted with small pulleys, and must be sloped at such an angle as will
allow the tail cords to sink when opening the sheds without obstructing
the pulleys underneath them. The neck twines extend from the figures
1 to 8 to the knots above the hole board D D. The cords which connect
the neck twines to the mails E, E are called sleepers, and those which
connect the mails with the leads F, F are called hangers. The hole
board is made of hardwood perforated with holes, which run from front
to back in diagonal rows from right to left; it should be a little
finer than the set of the reed, to allow for empty holes that are
sometimes caused by the tie of the harness ending with broken rows of
hooks in some or all of the repeats. In Scotland, for this reason,
when the reed is set on 37 in., the cumber board, or hole board, is set
on 36 in.

[Illustration: Fig. 3]

B, B are the tail cords, attached to the neck of the harness at one
end, and at the other end all of them are fastened to the tail stick
M, by means of which they are secured to the roof of the house. There
must be a tail cord for each part of the harness; here only eight are
shown for the front row of the harness, and if there were eights rows
of harness in the hole board, 64 tail cords would be required, and the
complete harness would be made up of several repetitions of the 64 neck
twines; four of these repetitions are here given for the first row of
the hole board. Of course there might be 400 to 600 tail cords in a
full mounting.

From each tail cord descends a vertical cord to the ground, as shown
at G, G. These are the simple cords, which, taken collectively, are
termed the ‘Simple.’ It is on these cords that the pattern is read, or,
rather, tied up. The simple cords are gathered together, according to
the pattern, by passing twines round them and forming the twines into
lashes or leashes, as shown at I, I. Heads of stronger cord, to which
the lashes are attached, are shown at N, N. The leashes or lashes are
made of cotton yarn No. 48, from six to eighteen plies of which are
moderately twisted together so that the twine will not curl; the heavy
twine is used for coarse work, where only a small number of lashes is
necessary. The length of the lashes is from 8 to 12 in., according to
the breadth of the simple. The heads are about 4-1/2 in. long, of good
cord, as foot twine, which is used finer or coarser according as more
or less heads are required. The heads are made with a noose on them
that will run up or down on the gut cord L, which is a strong cord,
generally extending from the ground to the roof of the house. K, K are
the bridles connected with the lashes, and used to draw them down in
succession as they are wanted by the drawboy. When there are a great
number of lashes, two gut cords are used, as shown at No. 1 (Fig. 3),
and the lashes are looped alternately on each and bridled accordingly.
In coloured work, where three or four draws are required for each weft
line of the pattern--that is, one draw for each colour--it is usual to
have two gut cords with cross bridles from the one to the other which
will slide up and down on them. On these, cross bridles the heads of
the lashes are fastened, about 1 in. apart, so that the drawboy can
take them in succession and draw the shed for each colour.

The method of preparing and mounting the draw-loom harness is much the
same as that now in use for jacquard harnesses, and, as it is entirely
out of use, it is unnecessary to describe it.

In order to make the neck twines draw evenly, rollers are placed
between each set of cords at the points 1, 2, 3, &c.; these rollers
keep the cords straight and make them all rise the same height at the
mails, which they would not do unless they all sloped to the hole board
at the same angle.


READING OR LASHING THE PATTERN

[Illustration: Fig. 4]

The pattern, painted on design paper, same as for pressure-harness
damask, is fixed upon a lashing frame, as shown in Fig. 4, and the
lower ends of the simples are passed over it and fastened to the
crossbar B. The simple cords are held in position over the design by
the comb C, C, which must be of such a fineness as to make each simple
cord stand directly opposite that space of the pattern to which it
corresponds, one simple cord being placed between each pair of teeth
of the comb. It will thus be seen that there must be a simple cord
for each vertical line on the pattern, or rather for each vertical
space between the black lines. In the same way, there must be a
head of lashes for each horizontal space, or line, as it is usually
called, and which would answer to a card for the jacquard or dobby.
The straight-edge E E is made so that it will slide up and down in
the frame, to mark the line on the design paper that is to be next
read by the lasher. Now refer to the line of the pattern above the
straight-edge, and it will be seen that the first square or check to
the left is blank, and it is accordingly passed over by the lasher;
the second and third checks are painted, and as the simple cords
corresponding to them have to be drawn to form the pattern, the lasher
twists one end of his lash over the pin G, and takes a turn of it round
the second and third simples, again passing it round the pin G. The
fourth check, being blank, is passed over, and a turn or tack of the
lash twine is taken round the fifth, sixth, and seventh simples, as the
checks on the design paper opposite these are painted. The reading and
lashing proceeds in this way till the line is finished, as shown in the
figure; then the two ends of the lash twine are tied together round
the pin G, which is then taken out, and the loop made round it by the
lash twine is twisted round and formed into a snitch for the purpose
of fastening it to the head. The lash is now pushed down behind the
board H, to make room for another; the straight-edge is then shifted to
the next line, and the lashing proceeded with as before. If too many
painted squares of the design paper come together, all the simple cords
corresponding to them must not be looped or lashed together, but can
be taken in two or more loops or tacks, never taking more than six or
seven simple cords into one tack of the lash twine.

It will be observed that the board H is rounded at the back; this is
for the purpose of having all the simples at an equal distance from the
pin G when they are tacked up by the lash twine, and consequently a
more regular shed will be produced when they are drawn in the process
of weaving.

The method of fastening the head to the lash is to loop the cord for
the head, which should be double, round the gut cord, then knot the
two ends of it together, and take this knot through the snitch formed
on the end of the lash, and when the snitch is drawn tight the knot
prevents the head from slipping out.

In weaving with the draw loom two persons are required--the weaver,
who works over the ground treadles, throws the shuttle, beats up the
weft, &c.; and the drawboy, who takes the lashes in succession as he
draws them down by the bridle, and by pulling out the simples raises
the harness and holds it in this position till the weaver has worked as
many shots as are required to be given to each draught.

When some thousands of twines were required for the harness, and with a
simple of three or four hundred cords, the weight and friction made it
very severe work for the drawboy. To assist him a fork, as shown at V,
No. 2, Fig. 3, was used. It was made to run to and fro on a carriage,
so that when the simples were drawn forward by the lashes, one spike of
the fork could be run in behind those drawn forward, while the other
spike was in front of them. When the fork was depressed, till the
handle T came to the position shown by the dotted lines, it drew down
the simple cords, and they could easily be held in that position till a
change of draught was required.

[Illustration: Fig. 5]

When the mounting of the draw loom was very extensive, it was necessary
to employ from two to ten pulley-boxes and as many drawboys, so that
it is not surprising that many endeavours were made to work without
the aid of a drawboy. One of these machines, known as the ‘Parrot’ or
‘Pecker,’ is shown in Fig. 5. It is wrought by the treadles T, which
are attached to the marches M, and these are connected by the pulley P,
on the rocking shaft R, by a cord which passes over and is fastened to
it. When one of the treadles is pressed down the pulley rocks and turns
the shaft to one side, and when the other treadle is pressed down the
shaft will rock to the other side. The parrot or pecker K is movable
on the rocking shaft--that is, it can slide along it--but it must rock
from one side to the other with the shaft. The cords C, C are passed
through holes in the boards B, B, for the purpose of keeping them in
position, and they have knots or beads on them at _m, m_, and weights,
W, W, at their ends to keep them in tension. Fastened to the cords
C, C is another set of cords, S, connected with the tail cords of the
draw-loom harness, and so corded or arranged as to draw those tail
cords required to be sunk to raise the harness to form the pattern. As
the shaft R is rocked from side to side by depressing the treadles in
succession, it carries with it the pecker K, and the groove or notch
at the point of the pecker, shown clearly in Nos. 2 and 6, coming
into contact with the knots or beads on the cords, draws them down
alternately, first at one side of the machine and then at the other,
until the pecker, as it slides along the rocking shaft, has passed
over all the cords; it is then released and drawn back to its original
position by means of the weight D, attached to it by the cord _e_. At
the end of the rocking shaft is a ratchet wheel, N. The cord _e_ passes
from the pecker through the segmental hole in the pulley P, as shown in
No. 4, and is fastened to a boss, O, on the inner side of the ratchet
wheel. This wheel receives its motion from a catch, as shown at E, No.
5, which is simply a pin fixed in a slotted piece of wood. The pin _y_
forms the catch, and the slot acts as a guide, which passes over the
edge of the ratchet wheel, and keeps the catch in position. The catch
is attached to the marches and works vertically. It is raised by the
treadles, and when released is drawn down by the weight attached to
it, the wire _y_ catching one of the teeth of the ratchet and moving
it round. There are two pins, _z_ and _t_, in the ratchet, as shown
in No. 3, and it is according to their distance apart that the length
of traverse of the pecker is determined. The bar _q_, shown in No. 3
attached to the pulley V, which is loose on the axle, is raised by the
stud _t_, as the ratchet wheel is advanced tooth by tooth, till it
comes into contact with the catch S, and raises it; this allows the
ratchet wheel to be reversed by the weight D on the end of the cord
_e_ till the stud _z_ comes round and draws away the bar, which allows
the catch to fall into position again and stops the pecker where the
pattern is to begin. Thus, the distance between the studs _z_ and _t_
must be arranged to suit the number of cords the pecker has to pass, or
to give the number of threads in the pattern. The teeth of the ratchet
wheel and traverse of the catch must be of such a pitch that as each
tooth is moved round the pulley will be turned the exact distance
required to move the pecker from one cord to the next one. No. 2, Fig.
5, is an elevation of a ‘parrot’ arranged for a single row of cords,
as these machines were first made. There was only one treadle, as
shown in section at M, No. 2, and the pecker only rocked to one side.
Either this or the double machine could be used for a shaft mounting
where a large number of treads are necessary. The cords C pass over
pulleys, and are fastened to long coupers or levers with their fulcrum
at the side of the loom, and to these the shafts are hung from jacks or
otherwise. These machines were made to work very exactly. In the double
machine the pecker would travel over three or four hundred cords in
consecutive order.

To avoid confusion the beads are not shown on the front cords in No. 1.


CROSS’S COUNTERPOISE HARNESS

About the year 1816 Mr. James Cross, of Paisley, invented a machine to
do away with the drawboys.

[Illustration: Fig. 6]

This machine is fully described by Murphy and Gilroy in their works
on weaving. Only the general principle of it will be given here, as
an introduction to the jacquard. The detail of drawing the lashes and
treading, though ingenious, is not of any practical importance now, and
it requires rather a lengthy description to explain it. The harness F
is the same as in the common draw loom till it reaches the tail cords,
where the counterpoise apparatus commences. The framing B B (Fig. 6)
of this machine is supported by the carriage A A, which rests on the
capes or top rails of the loom. In this frame are two boards, C and
D, perforated with holes corresponding in number with the tie of the
harness or cords in the simple. The top board is called the suspension
board, and is mortised into the bar E. From this board the harness
hangs, the neck being taken up through the holes in it, and fastened
above them. The lower board, D, which is mortised into the bar G, is
called the neck board, or directing board, as it keeps the harness in
its proper place. H and K are two other boards, perforated as shown
in Fig. 7, mortised into the sliding bars I and L respectively; these
are called the trap boards, M, M and N, M are four bars, called
pushers, which are fastened to the sliding bars I and L as well as
to the pulleys P, P, and when the pulleys are oscillated by means of
a treadle the sliding bars will be moved up and down. The knot cords
or tail cords O, O are fastened to the suspension board C, and pass
through the two trap boards, then through the neck board, and are tied
to the harness. Only two of these cords are shown tied to the harness,
to avoid confusion. These knot cords have knots or beads on them as
shown, and the round holes in the trap boards H and K, as shown in Fig.
7, must be sufficiently large (about 1/4 in. in diameter) to allow the
knots or beads to pass freely through. There are notches or saw-cuts at
the sides of the holes to admit the cords, but support the knots. T,
T^1 is the simple, extending horizontally through the knot cords. It
is fastened to the ceiling beyond T, and to the frame of the machine
beyond T^1. S is a half-leaf of heddles for the purpose of supporting
the simple cords. Each simple cord is tied to a knot cord, and beyond
T^1 the simple is lashed according to the pattern. In working the
machine the lashes are drawn by means of hooked levers, wrought by a
treadle connected with the pulley V, and those simple cords that are
drawn down pull the knot cords into the notches or saw-cuts in the trap
boards, so that when the trap board is raised the harness fastened to
those knot cords that are drawn into the saw-cuts will also be raised.
R is a set of cords with weights on their ends for the purpose of
drawing the knot cords out of the saw-cuts as soon as the simple is
released by the lashes. The two trap boards rise and fall alternately,
and this is why the machine is called a counterpoise.

[Illustration: Fig. 7]

Some time after an improvement was made upon this machine which was
known as


THE COMB DRAW LOOM

[Illustration: Fig. 8]

This machine appears to have been invented in both Scotland and
Ireland, as Gilroy describes it as an invention of Dr. McLaughlin, of
Ballyshannon, County Donegal; and Murphy describes a similar machine
invented by Mr. Bonnar, of Dunfermline. The machine is shown in Fig.
8. A, A are the posts of the loom, and B B the top rail; C C is
the framing of the machine. The harness G, G is suspended from the
suspension board D, and passes through the guide board E and the cumber
board F. The upper portion of the harness is composed of tail or knot
cords, as in Cross’s machine. From each of the tail cords a simple
cord, H, extends horizontally over the weaver’s head, and is fastened
to the board I. The lashes K hang from the simples over the weaver’s
head, and have a knob on the end of each, so that the weaver can catch
them and draw his own draught.

[Illustration: Fig. 9]

The lash cords have a knot or bead on them, so that when drawn they can
be held in the cuts of the board L, also shown in plan at L (Fig. 9).
M, N (Fig. 8) is a side view of the comb and handle, or lever, shown in
plan at M, N (Fig. 9). S is a cord or chain attached to the end of the
lever M, and passing down to a treadle. When the weaver draws one of
the knobs, the tail cords connected with the simples in this lash are
drawn between the teeth of the comb, as shown by the dotted lines in
Fig. 8. He then depresses the treadle, which raises the comb, and the
harness along with it; he holds the treadle down with his left foot,
and works over the ground treadles with his right one. The comb is
recovered or counterbalanced by the cord O, which passes from the comb
through the board P, and has a weight, R, suspended on it. D (Fig. 9)
is a plan of the boards D, E, F and I (Fig. 8).


THE BARREL OR CYLINDER LOOM

This machine was introduced by Mr. Thomas Morton, of Kilmarnock. The
harness and tail or knot cords are arranged similarly to those in the
comb draw loom; but instead of the simple cords for drawing out the
tail cords, each tail cord in the barrel loom passes through a slide,
or horizontal wire. The points of these slides are acted upon by the
pattern cylinder or barrel, and those held back press out their tail
cords from the others, and the knots on these tail cords are caught by
the teeth on the comb or roller, and the harness raised.

The pattern is arranged upon the barrel much in the same way as a
tune is arranged on the cylinder of a barrel organ or musical box. A
section of the barrel is shown at A (Fig. 10), with wire staples driven
into it to form the pattern.

[Illustration: Fig. 10]

Each of these staples represents so many lines of the design paper, or
so many lashes or draughts. The pattern is ruled out and painted on
the barrel, and staples are driven into it so as to cover the painted
squares of the pattern. The barrel is so arranged on the loom that
exactly the space of one line of the design paper is turned round for
each draught, and the slides are drawn back by cords attached from
their ends to a roller when the shift of the barrel is being made. B is
a section of the comb; it is a cylinder with teeth, C, like a parrot’s
beak fixed to it. The teeth are made of this shape to hold the knot
cords when they are caught by them, and they rise or fall as the roller
is rocked upwards or downwards by a treadle.

Whilst these improvements on the draw loom were being made in this
country for the purpose of producing a convenient method of harness
weaving, the French were endeavouring to obtain the same result, but on
a different principle, and their method has proved successful.

In 1725 M. Bouchon employed a band of pierced paper, pressed by hand
against a row of horizontal needles, so as to push back those which
were opposite the blank spaces, and thus bring loops on the extremities
of vertical wires into connection with a comb-like rack below, which,
being depressed, drew down the wires, pushed on the pins in it, and
raised the harness.

[Illustration: Fig. 11]

Fig. 11 is a sketch of a model of this loom in the Conservatoire des
Arts, Paris. A is the pulley-box with two rows of pulleys in it; B the
tail cords; C the simples, tied to rings on their upper ends, which run
on the tail cords at B; the other ends of the simples pass over a small
roller at D to prevent them rubbing against the side of the loom, then
down through the hole board F, under which they are tied to wire hooks
or loops, as shown under A (Fig. 11A). Next these wires pass through
the needle box G, also shown at B, Fig. 11A, and down to the comb H
(Fig. 11), and C (Fig. 11A). The perforated paper is rolled on the
roller E, and passing downwards, is pressed against the needles with a
hand bar, as shown at L, Fig. 11B. The roller K is for rolling up the
paper as it passes down from the upper roller.

Fig. 11B is a back view of the mounting. A shows where the simples are
connected with the tail cords P; B shows the connection of the tail
cords with the harness; C is the cumber board; D the mails and E the
leads. F and K are the two rollers for the paper, H the needle box, and
I the comb. This was the first attempt at forming the pattern by means
of perforated paper acting upon needles and wires.

In 1728 M. Falcon adopted a chain of perforated cards in lieu of the
perforated paper, and placed his horizontal wires or needles in several
rows or ranks, thereby admitting the use of a greater number of them
in a moderate space. He also used a square prism or cylinder, as it is
called, for the cards to pass over.

[Illustration: Fig. 11A]

[Illustration: Fig. 11B]

Fig. 12 is a sketch from a model of his loom, also in the Conservatoire
des Arts. The principle of it is much the same as the preceding. A is
the pulley box for four rows of pulleys, B the connection of the simple
with the tail cords, C the hole board for the simple to pass through
and also the support for the cylinder H, D is the needle box, E the
comb or griffe, F the levers for drawing down the griffe, and G the
treadle. The cards are laced in a chain and pass over the cylinders I
and H, but they are pressed against the needles by a hand bar, similar
to that used by Bouchon. There are two racks or receptacles for holding
the cards, as shown. The cylinders H and I are simply used as rollers
to support the cards, and not for pressing them against the needles, as
in the jacquard.

[Illustration: Fig. 12]

Figs. 12A and 12B give detailed views of the hooks, etc., for drawing
the harness: the letters in both refer to the same parts as are marked
with similar letters in Fig. 12. The simples B are tied to loops on
the hooks under the hole board C. In Fig. 12B it will be clearly seen
how the needles in four rows act upon the hooks. The griffe consists
of four round iron rods or wires set in the frame E, Fig. 12A, which
can be drawn down by the treadle G, connected with the levers F by the
cords 3 and 4. One of the levers has its fulcrum at one side and the
other has it at the opposite side, and the points of the levers are
tied to the griffe frame by the cords 1 and 2. The griffe, when sunk,
is brought back to its place again by weights hanging on cords running
over pulleys. It may be seen that the hooks hang in front of the rods
or bars of the griffe; therefore, those hooks will be pushed on the
griffe whose needles come against the solid portions of the cards, thus
acting similarly to Bouchon’s machine. The hand bar for pressing the
cards against the needles is shown at F. The needles can pass through
the slot in it, and, when a card is pressed against the needles, the
bar can be fastened with hooks for the purpose, so that the operator is
free to depress the treadle. There are no springs on the needles, but
a clap board comes behind them, which is pressed in by means of a cord
passing over each end of it and fastened to a spiral spring. There are
ten leaves of heddles in front of the harness. They are plain clasped
heddles, and apparently act as five, the front five being raised by the
top levers connected with the treadles, as shown in Fig. 11, and the
back five are held up by a set of levers at each side, with weights on
the ends of them, similarly to those now used for hand-loom damasks,
and tied to the treadles below, so that sinking a treadle would cause
one of the back leaves to sink and one of the front ones to rise. Only
the harness mechanism of Falcon’s loom is given in Fig. 12, the front
portion being similar to that of Bouchon, or to any hand loom.

[Illustration: Fig. 12A]

[Illustration: Fig. 12B]

In 1746 the accomplished mechanician, Vaucanson, altogether dispensed
with the cumbrous tail cords and simple of the draw loom, and made the
draw-boy machine completely self-acting by placing the hooks upright
on the top of the loom, and hanging the harness from them. This loom
may be seen in the Conservatoire des Arts, as well as a model of it on
the same scale as those already mentioned--viz., one-third the size.
The machine for drawing the harness is exactly like a small jacquard,
with two rows of hooks and two rows of needles, as shown in Fig. 13.
Instead of a square cylinder and cards, the pattern is punched on
a band of paper, which passes over a round or true cylinder. This
cylinder is fitted with a rack wheel, so that a tooth can be passed for
each change of pattern, the cylinder moving out and turning one tooth,
then pressing in against the cards again. The diameter of the cylinder
is about twelve inches. The hooks are raised with a griffe, similar
to that in a jacquard, which is fastened to a lever connected to a
treadle below. This treadle, and four others for working the heddles,
is wrought by tappets, made of wood, on a shaft running along the side
of the loom and turned by a wince handle in front. Altogether, the loom
is a nice mechanical contrivance, and a great step in advance of its
predecessors.

[Illustration: Fig. 13]

Joseph Marie Jacquard, a working mechanic of Lyons, having invented
a fishing net loom, turned his attention to improving the means of
drawing the harness in looms for figured weaving, about 1790. A model
of a machine by him, dated 1790, to dispense with the drawing of the
harness, is in the Conservatoire des Arts. It is made with cords and
rollers, and has no resemblance to the machine bearing his name. He was
brought to Paris to repair Vaucanson’s loom about 1804, and it appears
to be then that he combined the best qualities of the machines of his
predecessors, and produced the jacquard, a model of which, dated 1804,
is in the Conservatoire des Arts. This is very much like our present
jacquard, but with four rows of hooks and needles made similar to those
of Vaucanson, Fig. 13. He dispensed with Vaucanson’s cylinder and band
of paper, and used instead a square prism with a chain of cards passing
over it. The cylinder (or prism) he set in a frame or carriage, made to
run on four wheels or pulleys on the top of the frame of the loom. The
carriage is drawn out by depressing a treadle, and brought back again
to press the cylinder against the needles, by means of weights tied
to cords running over pulleys. The griffe is raised by means of two
levers, one at each side; one end of each is connected with the griffe,
and the others to a crossbar at the bottom of the loom, and this
crossbar is fastened from its centre to a treadle. In all these looms
the cards or paper hang at the side of the loom, the mounting being on
the principle known as the London tie. Jacquard was born in 1752, and
died in 1834. Vaucanson died in 1782.


FRENCH DRAW LOOM

Shortly after the introduction of Cross’s counterpoise harness, a
machine was imported from France, which is shown in Fig. 14, and
described in Murphy’s ‘Art of Weaving’ as a French draw loom. This
machine far surpassed any attempts at the improvements at the draw
loom that had hitherto appeared, in simplicity of construction and
operation. From the neck upwards the harness is similar in construction
to Cross’s counterpoise, having the knot cords arranged in the same
manner, but with only one trap board. Instead of the cumbrous tail, the
knot cords are acted upon by wires or needles, on each of which is a
loop, through which one of the knot cords passes. D is the cylinder or
barrel, perforated with holes, as in the common jacquard cylinder, and
C, C shows the chain of cards for forming the pattern; E is the lever
for raising the trap board, to which it is connected by means of pieces
of iron at each side, with a bar across between them, to the centre of
which the lever is connected with a piece of wire. O O are crossbars of
wood, with holes in their centres, through which run pieces of strong
iron wire, which are fixed into the trap board at each end to keep it
steady while in operation. There is no spring box for the needles as is
now used in the jacquard, but into the crossbar or frame F is inserted
a flat piece of wood moving on springs, which yields to the pressure
of the needles that are forced back by the barrel, and recovers them
again when the barrel is withdrawn. The lever E is drawn down by the
cord H, attached to a treadle, when the trap board is to be raised, and
the barrel is drawn back by the cord G, which is attached to another
treadle. The barrel is pressed against the needles by springs, and when
it is relieved by the treadle and is moving inwards, it is turned by
one of the catches shown in Fig. 14A. Either of these catches can be
brought into action, so as to turn the barrel either way, by raising
or lowering them with a cord. When at rest the knot cords stand in the
notches or saw cuts of the trap board, but when the cards are pressed
against the needles, except where there are holes in the cards, the
needles are pressed back and the cords are pushed out of the notches
so that the knots stand above the holes in the trap board, and pass
through them when the board is raised by depressing the treadle
connected with the lever E. The trap board is shown in Fig. 7.

[Illustration: Fig. 14]

[Illustration: Fig. 14A]

Whether this machine was Jacquard’s invention or not, I have not been
able to ascertain; but Gilroy states that Jacquard’s first machines
were made with cords and trap boards, like Cross’s counterpoise
machine. It is also recorded that William Jennings, of Bethnal Green,
invented a machine, similar to the above, about 1830, as an improvement
on the jacquard, on account of its simplicity, as the latter appears
to have given the weavers some trouble, and notice was taken of his
machine by the Society of Arts.

[Illustration: Fig. 15]

Machines similar to the above are still in use for hand-loom work, and
answer very well. They have also been used for power looms, but the
cords do not stand the friction long. As at present made about Glasgow,
the cylinders of these machines work on slide rods, and receive their
motion from the rising of the trap board, same as the cylinders in
jacquards are sometimes wrought by the rising of the griffe with what
is known as the ‘swan neck’ motion. The spring board at the ends of the
needles is held back by a spring and drawn in by the cylinder frame
as it moves outwards, which brings all the needles forward, and the
cords into the saw cuts in the trap board, or comb as it is frequently
called. The knots on the cords for these machines are made as at B,
Fig. 15, which, when drawn tight, are as shown at A. The cord is a
good twisted cord, such as whipcord, and the holes in the trapboard
are about 3/16 in. in diameter, which the knots on the cords must pass
freely through. The upper ends of the cords pass through a perforated
board on the top of the machine, and another cord is drawn through
loops on their ends, which prevents them from being drawn through the
holes.




CHAPTER II

_JACQUARD MACHINES_


The jacquard machine was introduced into England about 1818 and came
into general use from 1824 to 1832. It was introduced into Scotland
about 1824. Fig. 16 is a representation of the early form of jacquard,
and of course intended for hand-loom work. Although the present
machines for power-loom work are very different in make, nearly all
the working parts as here used are to be found in different machines
at present working, or still being made, though the best machine
makers have adopted newer and better principles for fast working and
withstanding wear and tear. Fig. 16A is a view of the interior of the
machine. The working of the machine will be explained further on;
only those parts that will not be given in the new machines will now
be noticed. The griffe or frame for raising the hooks is lifted by
the straps A, A, Fig. 16, which are attached to the pulleys B, B, and
a cord C over a pulley on the same shaft as B, B, is attached to a
treadle beneath the loom. As the weaver presses down this treadle the
griffe is raised, and when the treadle is released the griffe falls
of its own weight. The cylinder is moved out and in by the pulley E,
fastened on the bent iron bar, attached to the frame which carries the
cylinder, when the griffe rises and falls. D, D is a frame which lies
in the turned-up portion of the hooks; only a few hooks are shown, and
the outer bars of the frame. There should be a bar in the frame for
each row of hooks. This frame rises up and down with the hooks, the
turn on the ends of which must be of such length that when the frame
is raised by the hooks lifted by the griffe, it will not be raised
out of the turns on those that are left down. This frame is for the
purpose of preventing the hooks from turning round, so that the turns
on the upper end of them, as shown at A, A, Fig. 16A, cannot get out
of position to be caught by the blades or crossbars of the griffe. The
lower ends of the hooks rest on a perforated board as shown in Fig.
16A, and tail cords are looped on them and pass through the perforated
board as shown. To these cords is fastened the neck of the harness.
Machines with these perforated boards and frames to keep the hooks in
position are still (1890) being made in Yorkshire. In Fig. 16A is shown
a section through the spring box B. This box contains a spiral spring
for each of the horizontal needles, the ends of which press against the
springs, which allow them to yield or move back as the cards press on
their points; but recover them again when the card is moved away by the
cylinder. Similar spring-boxes are now used.

[Illustration: Fig. 16]

It might be interesting to describe the various changes and attempted
improvements that have been made on the original jacquard, but it
would take up too much space, and many of them are of more historical
than practical importance; some of them will be mentioned that may be
interesting from a mechanical point of view, in connection with the
descriptions of the machines, even though they have only been partially
successful. Though a very different machine from what it originally
was, the principle of the jacquard remains the same, and is not likely
to be altered or superseded till a revolution takes place in the
process of weaving.

[Illustration: Fig. 16A]

A jacquard machine is simply a shedding motion by which a great variety
of sheds can be formed; the larger the machine, or the greater the
number of hooks it contains, the greater is the variety of shedding
that can be produced by it.

If a jacquard is made small, with, say, from 16 to 48 hooks, it is
called a dobby or shedding motion, and is used for working shaft
mountings; but the ordinary jacquard machines have from 200 to 600
hooks, which have long cords, called the harness, connected to them, no
shafts being required, as each hook has only a few cords tied to it,
which can be raised independently of the others. The fewer the cords
that are tied to each hook, the greater is the variety of shedding that
can be made on the same number of warp threads, till, when there is but
one cord to the hook, any thread or any number of threads can be raised
independently of any of the others.

Jacquards may be divided into four classes--viz. single-acting,
double-acting lift, double acting with double cylinders, and twilling
jacquards; and besides these there are several other varieties made
for special purposes. The single-acting is the real jacquard, and much
the simplest machine. It has the disadvantage which all single-acting
shedding motions have--viz. that one shed must be closed before the
following one begins to open. This is on account of the same lifter
having to open each shed; it must bring down the set of hooks that are
raised, and then raise the next set.

This constitutes the true jacquard lift; and while it makes a clear
shed, and is desirable for some purposes, it is generally considered
the most imperfect form of shedding--that is, so far as the making
of a good cloth is concerned. It is not suitable for making a heavy,
well-covered cloth, nor for working at a high speed, 120 to 140 picks
per minute being a very good speed to drive it.

A single-acting jacquard is a very simple machine, and when properly
made should give very little trouble in working, particularly if the
motions are properly set in relation to each other, and if such methods
of working are adopted as will cause the least wear and tear on it.

In whatever way jacquards are made, the principle of working is much
the same. There are a number of upright hooks set in a frame; attached
to each hook is a horizontal wire or needle, one end of which is
pressed upon by a spring, which keeps both it and the hook steady and
in position, while the other end, or point, passes through a perforated
plate, beyond which it projects about half an inch (see Fig. 16A). To
lift the hooks there is a set of bars or knives arranged in a frame,
just below the heads of the hooks; this frame is called the ‘griffe’
or ‘brander,’ and if raised would draw all the hooks up with it. What
hooks will be lifted for each shed is regulated by perforated cards
being pressed against the points of the needles. A perforation in the
card allows the point of a needle to pass through and the hook to be
raised; but where there is no perforation the card comes against the
point of the needle, pressing it back, and holding the head of the hook
clear of the blade of the griffe, so that the griffe will pass without
raising the hook. It will thus be seen that any variety of shedding
can be made by punching the cards to suit it. Fig. 17 shows one of
the best makes of single-acting jacquards. One of the best methods
of driving is shown in this and the following figure. The griffe is
raised by means of the lever G, which is sometimes supported on a beam
fastened to the roof or pillars of the shed, or it may be supported by
an upright fastened on the frame of the loom. A portion of this upright
is shown in Fig. 17, and as well as being fastened to the loom frame,
it should be stayed to the top of the machine. One end of the lever
is fastened by a link connection to the centre of the bar across the
griffe frame, care being taken that the connection is so made as to
draw up the griffe vertically, and not have any strain on the slide
rods or spindles that are used for keeping the griffe horizontal when
rising.

[Illustration: Fig. 17]

The other end of the lever is connected to a crank on the crank-shaft
of the loom by a rod, E (Fig. 17); also shown at A (Fig. 18), where the
crank is also given.

[Illustration: Fig. 18]

In hand-loom machines the griffe is frequently pushed up from below
instead of being drawn up from the top as is shown in Fig. 17. The
method of doing this is similar to that given for lifting the griffes
of twilling jacquards.

F (Fig. 17) is called the cylinder or barrel, evidently taking its name
from the round cylinders or barrels used in the old machines, but is
in reality a square prism. It is made of wood, and perforated on each
side with a set of holes--a hole for each needle in the machine; its
use is to draw round the chain of cards and press each one against the
needles, or horizontal wires, in the machine. In order to keep the
cards firmly on the cylinder, flat steel springs are sometimes used, as
shown, attached to the top rail of the frame which holds the cylinder;
and there are also steel wires which pass down in front of the cylinder
over the ends of the cards. These springs are useful when only a small
number of cards is used and the machine driven quickly; with a large
set of cards, where there is plenty of drag on them from their own
weight, they are not necessary, and but seldom used; but they are in
common use in the Yorkshire districts.

It will be seen from the illustrations (Figs. 17 and 18) that the
cylinder hangs in a frame suspended from the top of the machine;
this is called the swing or batten motion, to distinguish it from
the horizontal or sliding motion which is shown at Fig. 20 (No. 1)
and in Fig. 27. The swing motion is the simpler of the two, and is
cleaner, requiring less oil; but the sliding motion is steadier, and
does not swing the cards so much, consequently is more suitable when
the cylinder has to travel quickly. The swing motion also requires the
machine to be higher; with a slide motion the frame is usually cut off
a little above the griffe.

The cylinder has to travel out and in when the machine is working, so
that it may be turned round and bring a fresh card against the needles
for each shot. There are many methods of accomplishing this, which may
be divided into two classes--viz. independent motions, or those which
are driven from the loom independently of the machine; and self acting
motions, or those which drive the cylinder out and in through the
rising and falling of the griffe. The latter are the simpler, but the
former are much better, causing less wear and tear on both the cards
and machine, as will be explained further on.

It will be seen in Fig. 18 that as the cylinder travels out it will be
caught by the hook K^1, and turned round; the head or lantern of the
cylinder is made of iron, as shown, so that the hook, or shears, will
take a firm catch on it. To prevent the cylinder from turning more than
one card at once, and to keep it steady so that it will always come in
fair against the needles, it is held firm by a hammer pressed on it by
a spring. This pressure is applied in different ways, one of which may
be clearly seen in Fig. 17, and another in Fig. 27. When taking out
the cylinder, or wanting to run it round quickly to draw over a number
of cards, the hammers can be held up by a hook or sliding catch, which
should be fitted to the machine for the purpose.

One of the best independent motions for driving the cylinder is shown
in Fig. 17; and that shown in Fig. 18 is also a very good one for small
machines, perhaps the most convenient that is made; but the former
is much stronger. In Fig. 17 a connecting-arm from the frame of the
cylinder is attached to the lever B. The connecting-arm should have a
slotted joint, so that it can be made shorter or longer, if required,
for the purpose of regulating the pressure of the cylinder on the
needles.

The lever B is on a horizontal shaft, bracketed to the frame of the
loom, or to the beams on which the machine rests; or some machines
have bearings attached to their framing for it. There is, of course, a
lever, as B, and a connecting-arm at each side of the machine. There
is another lever on the end of the shaft, at right angles to B, which
is connected with an eccentric on the crank-shaft of the loom by a
rod, C, in the same way as the rod C is connected with the eccentric
in Fig. 18. The eccentric can be set to bring the cylinder against
the needles at any required time, independent of the lifting motion
of the machine, which cannot be done when the self-acting motions are
used. The larger the eccentric, the greater dwell the cylinder will
have against needles. The method of working the cylinder in Fig. 18
is somewhat similar, and can easily be seen; but it will be observed
that a good deal of pressure will be put upon the studs on which the
cylinder frame, or batten, hangs, particularly when the cylinder is
being pressed in, as this is effected by drawing down the lever L;
however, in a light machine this does not matter much.

[Illustration: Fig. 19]

The principal feature in this motion is the escapement apparatus for
the purpose of disengaging the cylinder from its connection with the
driving eccentric when it is required to turn some of the cards back.
Fig. 19 (Nos. 1 and 2) shows this arrangement. The motion is not quite
the same as that given in Fig. 18, but is on the same principle, and
one may be easily understood from the other. In Fig. 18 the latch G
comes out of the notch in the quadrant F, when the handle E is pressed
close; the handle is on the lever D, having its fulcrum on the shaft N,
and the quadrant F is connected to the eccentric rod C. The quadrant
is loose on the shaft, and the lever is fast on it. Fig. 19 is a more
convenient motion. The two halves of the handle A are held apart by
the spring H, and this, through the hook B on the inner end of the
handle, presses the latch or catch on the slide D into the notch on
the quadrant C. The quadrant and handle are one piece, and are fast on
the shaft E, and a lever F on this shaft is connected to the batten of
the machine, in the same manner as shown by D and B in Fig. 18. When
there are two or more machines, one of these levers would be required
for each. The lever K is loose on the shaft E, and the rod G connects
it with the eccentric, same as is shown by the rod C in Fig. 18. In
No. 2, Fig. 19, the quadrant is left out to show clearly how the hook
B acts on the slide D, and also to show the lever K on the shaft. The
two halves of the handle, being pressed out by the spring, keep D in
position, firmly pressed inwards; but when the handle is pressed the
hook B presses the slide D outwards, leaving the quadrant free to pass
up or down; and by pulling down the cord H (Fig. 18), which raises
the shears K and K^1, the cards can easily be turned back by working
the handle up and down, as the under shears will catch the cylinder
and turn it the reverse way. The weaver must be careful not to jerk
the motion and throw the cards off the cylinder or damage them; but a
very little practice will enable anyone to turn them back quite easily
and quickly. This motion answers very well for one, or perhaps two,
small machines; for a 400 or 600 machine, or any smaller size working
spottings, &c., it is very convenient, but when large mountings are
required, as in 800 to 2400 machines, it is quite too weak for the
work; even if made strong enough it would not be satisfactory, as the
strength of spring that would be required to bear the strain and keep
the catch in the notch would make it a very difficult matter to use the
motion for reversing the cylinder. For heavy machinery the method of
working the cylinder shown in Fig. 17 is far preferable, and separate
motions for turning back the cards can be fixed on the machine. These
will be explained further on.

Instead of the eccentric and crank for driving the cylinder and raising
the griffe being as they are shown in Fig. 18, though a plan frequently
in use, it is neater, and perhaps better, to have the eccentric at the
back of the fly-wheel, and the fly-wheel either cast with one half
solid, or have a plate fastened across two or more of the spokes, to
which the connecting-rod can be attached with a bolt fastened in a
slot. The amount of lift can be increased or diminished, either by
shifting the top of the connecting-rod along the lever G (Fig. 17), or
by increasing or reducing the throw of the crank at the fly-wheel.

[Illustration: Fig. 19A]

Self-acting motions actuate the cylinder through the rising and falling
of the griffe without requiring any special connection from the
loom. One of the most convenient of these is that frequently used on
hand-loom machines, and known as the S iron or swan-neck motion. It is
shown in Fig. 19A (No. 1), and another form of it on a swing cylinder
motion is shown in Fig 16. D is the swan-neck or S iron. In the groove
in it a roller stud on the griffe frame travels, sliding in and out the
cylinder A as the griffe falls and rises. E is the slide bar, which may
be flat or round; if round, there must be some means of keeping it from
turning in its bearings, which is generally accomplished by having a
crossbar bolted across the two slide bars behind the machine.

No. 2, Fig. 19A, is a motion for the same purpose, but consists of a
series of levers; and No. 3 is an arrangement of a similar nature. B is
the fulcrum of the levers, or fast pin by which they are connected to
the machine. C shows the attachment of the levers to the slide rod of
the machine. A is the connection with the cylinder frame. As the griffe
rises or falls it will easily be seen that the cylinder will be driven
out or in.

The connecting-bar H is in two parts, slotted and bolted together at H
to admit of regulating the position of the cylinder.

No. 4 is a motion on a different principle; it is a French motion. It
will readily be seen that the cylinder is driven out and in by the
toothed wheel, which is wrought by a rack on the slide rod E. This
slide rod works outside the framing of the machine, as is common in the
French machines. One point must be observed about these motions--viz.
that they must have a certain amount of dwell at the bottom of the
stroke, or when the cylinder is in. The reason of this dwell will be
explained further on, but the method of obtaining it may be given
here. In the swan-neck motion (No. 1), any desired dwell can easily be
obtained at either top or bottom by the length of the slot that is in
a vertical direction, as when the stud is passing through this portion
of the slot no motion is given to the cylinder. In the lever motions
Nos. 2 and 3, as well as in No. 4, the dwell is got by the levers or
arms passing the centres; in Nos. 2 and 4 it is by the lever or arm H
passing the back centre, which may be considered as a crank; and in No.
3 it is the short lever C passing the centre that gives the dwell.

[Illustration: Fig. 20]

[Illustration: Fig. 21]

[Illustration: Fig. 22]

Fig. 20 is a view of the interior of a single-acting jacquard machine
with the framework removed; only one row of hooks and needles are
given, to avoid confusion. The blades or knives of the griffe, with
a support running across their centres, are shown at B. A is the
face-plate or needle board, sometimes made of iron, but better to
be of hard wood. C is the spring-box, the detail of which is given
in Fig. 24. E, E^1 are the hooks, and F, F^1 the needles. D is the
grating through which the hooks pass, and are supported by it. It will
be seen that the hooks and needles are arranged in rows of eight: a
400 machine would have 50 or 51 of these rows in it; 500 machines
are usually arranged in rows of 10, and 600 machines in rows of 12.
The hooks should be set perfectly upright or vertical, and should
be close up against the knives, but not pressing against them. There
should be a provision made for shifting both the grating and the
knives, so as to admit of both them and the hooks being properly set
in relation to each other; but if set correctly by the maker, which
they should be, no alteration is necessary. A (Fig. 21) is a side view
of four hooks and needles, with the ends of the knives of the griffe
shown at _a_, _a_, _a_, _a_. When the machine is working the needles
are acted upon by perforated cards cut from the pattern. Suppose we
take plain cloth--that is, a pattern in which each half of the warp,
or every alternate thread, is raised and sunk alternately; then, if
the first card acts on all the odd numbers of the needles, and the
second card on all the even numbers, this repeated would make plain
cloth. Whenever a hook of the jacquard is to be raised a hole is cut
in the card for the needle connected with that hook, and a card with
all the even numbers of holes cut in it will cause the griffe to
raise all the even-numbered hooks. Refer to _b_ (Fig. 21), where the
second and fourth holes are cut in the card. If the card is pressed
against the needles, as at _c_, the first and third needles will be
pressed back, and will push the first and third hooks back from their
position--shown by the dotted lines--to the position in which they are
shown in B (Fig. 21); but the second and fourth hooks are not moved,
as their needles pass through the holes in the card. If the griffe
is now raised, the blades or knives will pass the first and third
hooks, but will lift the second and fourth; and if the odd numbers
of holes are cut on the next card, the first and third hooks will be
lifted when it is pressed against the needles, as shown at C (Fig.
21), thus making the cross-shed; and this explains the principle of
working any pattern by the jacquard without taking into consideration
the intricacies of mounting, &c. In C (Fig. 21), it will be seen that
if the knives _d_, _d_ were upright instead of slanting, they would
come down on the heads of the hooks that are under them, but, being
slanted, their lower edges pass the heads of the hooks, and press them
away as the griffe descends. Sometimes, even with slanting knives, if
there is much vibration in the hooks, or if the loom ‘bangs off,’ some
of the hooks are liable to get under the knives and be ‘crowned,’ or
bent down. To avoid this deep blades are often used, principally in
double-acting machines, so that the lower edges of the blades will
not pass the bottom hook, as shown at D (Fig. 21). This prevents any
danger of crowning, but it darkens the machine a little--that is, makes
it more difficult to see down into it if any of the wires require to
be examined; it also requires the heads of the hooks to be somewhat
higher above the heads of the needles than is necessary with the narrow
blades. Another principle has been tried--viz. that of making the
heads of the hooks as shown in Fig. 22, and using narrow blades. This
effects its object well, but there is too much friction of the knives
against the hooks, and the latter are liable to wear out too quickly.
A good machine with firmly set hooks should work very well without
these protections if it is steadily placed above the loom, and it is
better not to be resting on the loom framing, if convenient to have
it so. It will be seen from the foregoing explanation that the proper
time, or, rather, the necessary time, for the cylinder to press against
the needles is just after the griffe begins to rise. When the griffe
is down the top edges of the knives should be 3/16in. or 1/4in. lower
than the turned points of the hooks, and before this edge rises up
to the hooks those that are not to be lifted should be full back, or
the cylinder should be close in, having the front of the head of the
hooks 3/16in. or 1/4in. behind the blades. The cylinder should have
a short dwell in this position; and if it has a longer dwell it may
assist to reduce the friction of the heads of the hooks against the
knives; but if it has too great a dwell it may have to travel out and
in too quickly to make up for the lost time, which will probably not
be compensated for by the advantage of the increase in the dwell. Now,
when the cylinder is driven with an independent motion, as in Figs. 17
and 18, it is easy to set the eccentric so that it can be brought in at
any required time; but when a self-acting motion is used, it is plain
that if the cylinder must press against the needles when the blades of
the griffe are passing the heads of the hooks in rising, it must also
press against them in the same position when the griffe is falling,
unless some special escapement motion could be devised to avoid it, and
this is where the dwell is required, and where the evil effect of the
motion takes place; and it is worse in a double-lift machine with one
cylinder, as the heads of the hooks in it are larger, or have a longer
turn on them. A little consideration will suffice to show that when the
brander or griffe is falling, say, with one-half of the hooks hanging
on its knives, and the cylinder is brought in against the needles
before the hooks are quite down, as it must be, it will either cause
the hooks to be shot off the knives, or will put a considerable strain
on them, as well as upon the needles and cards. It is for this reason
that these motions are objectionable, particularly in power-loom work,
where the speed is high and the hooks are strong. In hand-loom work
it is not so objectionable, as there is more spring in the wires, and
the heads of the hooks need not be too large, and, besides, the speed
is less and the wear and tear not so great; but, even with this, if a
hand-loom machine that has been in use for some time be examined, it
will be seen that the points of the hooks are considerably worn, and
that the edges of the knives are hollowed out like a coarse saw by the
friction of the hooks on them. This latter will partly arise from the
lifting of the hooks.

[Illustration: Fig. 23]

In the old Jacquard, given in Figs. 15 and 16, the hooks are shown
resting on a perforated board, and it was mentioned that in order to
prevent them from turning round a frame lay in the turned-up portion
of the hooks. The grating in Fig. 20, through which the turned-up
bottoms of the hooks pass, readily accomplishes this object. Sometimes
flattened hooks are used, as in C (Fig. 23), with the needles twisted
once or twice round the hooks; this makes a firm arrangement, but if
anything goes wrong with a hook it is not easy to get it clear of the
needle. When the needles were made with a full twist or loop on them,
as at B, the same was the case; they are now usually made as at A,
and if arranged in the machine as in Fig. 20, there is no danger of
the hooks sliding out of the recess in the needle, and if a hook gets
bent or broken, it can be taken out and replaced by a new one without
disturbing the needles.

Fig. 16 shows how the needles press against the springs in the
spring-box, which is much the same as that at present in use. Fig. 24
is the present arrangement. No. 1 gives a plan of the end of a needle,
B. C C is the horizontal wire which supports it as shown in section at
C C in No. 2. D (No. 1) is a section of the vertical wire shown at D
D (No. 2), which passes through the loops or eyes on the ends of the
needles, and keeps the springs from shooting them too far forward. A
(No. 2) is a wire which passes down at the outside of the box over the
ends of a row of springs, so that by drawing out this wire any of the
springs can be drawn out without taking off the spring-box, as the
springs pass through the box. The springs should be strong enough to
keep the hooks and needles steady, but if unnecessarily strong they
give the card and cylinder unnecessary work.

[Illustration: Fig. 24]

In some machines there is no spring-box. The hooks are made double, as
shown in Fig. 25, and rods, as _a_, _a_, run along between the rows
of hooks; the spring of the double wire keeps the hooks steady. There
is a clap-board used, similar to that in the French draw loom, (Fig.
13), which is pressed against the needles with springs; this board is
connected with the face-plate by a bar at each end, forming a frame.
The needles do not project much through the face-plate, but when the
cylinder is pressed against it, it slides back on the needles, and
presses the clap-board back, which also allows those needles to go
back which the card presses against. The needles are not looped on the
hooks, but have a turned catch in front of them as shown. The bottom of
the hook rests on a hole board, C, through which the tail cords pass;
and through the hooks at D are wires fixed in a frame which rises and
falls when the hooks are raised, and keeps them from turning round,
same as explained in Fig. 15. Machines of this description are at
present being made in France, and work with a rising and falling shed,
which will be hereafter explained (see Fig. 30).

[Illustration: Fig. 25]

[Illustration: Fig. 26]

[Illustration: Fig. 27]

In working, the card cylinder must be so set that it will come forward
fair on the needles--that is, that when it comes forward the points of
the needles will enter fair into the centres of the holes in it. For
the purpose of setting it there must be provision made in the fittings
so that it can be moved laterally or vertically. In the swing motion
the frame can be moved laterally by means of the two screw studs on
which it hangs. C, Fig. 27, shows the bearing on which the stud of the
cylinder revolves. This bearing can be raised up or down in the frame
R--a side view of which is given at S--by slackening the bolt B with
the wing nut A, and adjusting the bearing with the set screw D or E.

A method commonly adopted by tacklers or tuners to see that the needles
are perfectly fair in the centres of the holes in the cylinder, is to
rub their fingers on some dirty oil, and touch over the points of the
needles with it. They then bring in the cylinder against the needles
with a card on it, in which about half of the holes are cut. The points
of the needles mark the card where there are no holes, and it can
easily be seen whether the mark is in the centre of where a hole should
be, or not.

One of the best bearings and attachments for a cylinder with a
horizontal slide motion is given in Fig. 26. D is the bearing for the
cylinder E, and C the bolts for setting it. F F is the bracket which
holds the hammer and bearings, which can be set in position on the
slide bar B by the bolt A. I is the hammer held down by the spring H
attached to the rod G.

The cards are kept in position on the cylinder by pegs or studs,
originally made of wood, and driven into the cylinder. Now they are
made of brass, and set in a slotted bracket, so that they can be
shifted in order to have the holes in the cards corresponding exactly
with those in the cylinder. The pegs should also be set on springs, so
that if a card gets off them, and between them and the needle plate,
they will yield or sink into the cylinder, and not break the card. In
all good machines they are made in this way.

When the motion for driving the cylinder is not fitted with an
escapement for the purpose of turning back the cards, it is necessary
for the convenience of the weaver to have a motion on the machine for
the purpose.

[Illustration: Fig. 28]

Figs. 28 and 29 show two varieties of these motions. A is the cylinder
head; C, the catch for reversing the cylinder; F, the spring for
returning the catch to its position; E, a cord which hangs down, with a
knob on the end of it, in a convenient position for the weaver to catch
and work the motion. In Fig. 28 the motion is on the opposite side of
the machine to the shears, but might be on either side, and the weaver
has to raise the shears to turn the cylinder, which she can easily do
by catching the knob for raising the shears in one hand, and working
the reversing motion with the other. The cylinder must be full out for
this motion to turn it properly, and this prevents the weaver from
tearing the cards on the needles, as she might easily do by trying to
turn the cylinder when it is too close in. In Fig. 29, B is the shears
for turning the cylinder, and both it and the pushing catch, C, pass
through a keeper or bracket, D, on the side of the machine. There is a
rise on C a little back from the point, and when the cord E is pulled
down C is shot forward by the lever, which has its fulcrum at G, and in
going forward the rise on it comes into contact with the bend in the
shears, and raises them up so that it can turn the cylinder when it
catches on the head of it. Both these are good and convenient motions.
For the latter the cylinder should be about half-way out when the cards
are being turned back.

[Illustration: Fig. 29]

Sometimes the cylinder may not be completely turned by the shears when
the machine is working, by reason of the cards catching, or if the
shears are too long, or it may arise from other causes. In this case
the cylinder would come in with one corner against the needles, and
be pressed heavily against the needle plate. Some of the levers would
probably be broken, or the cylinder might be shot out of its bearings
and fall, breaking the yarn, or perhaps injuring the weaver. To avoid
this, small snecks, as at H, Figs. 28 and 29, are set so that when the
cylinder is square it will pass over them; but if turned angularly
its lower edge will catch on the point of the sneck, as the cylinder
is coming in, and turn it square. The sneck is held up with a spring
so as to allow the cylinder in turning to depress it. Sometimes the
sneck, as at H, is liable to cause broken shots; for if the weaver
turns back the loom, and the cylinder moves out sufficiently far to be
turned to its angular position, and remains there, it will, in coming
in, be turned square by the sneck, and thus a card would be passed
over without a pick being put in for it. This is sometimes remedied by
putting the sneck above the cylinder, instead of below it, which would
turn it the reverse way; but this might come wrong at other times. The
better plan is to keep it below, and let the weaver get to understand
what she is doing, as it is not very difficult to learn.

The setting of a jacquard machine for working consists in adjusting the
cylinder motion so as to bring in the cylinder at the proper time, and
press it sufficiently close against the face-plate to keep the hooks
clear of the knives of the griffe, without pressing it too close; and
regulating the lifting of the griffe to suit the time for shedding,
and to give the size of shed required. The shed must be open for the
shuttle to pass through; the time for picking is when the cranks of
the crank-shaft of the loom are at the bottom centre, therefore the
shed should be almost fully open at this time. The lifting of the
griffe can be made a little earlier or later to suit circumstances,
but very little alteration can be made, as it takes a full revolution
of the crank to raise and lower the griffe. Further consideration will
be given to jacquard shedding after double-acting machines have been
explained.

The motion for working the cylinder, if an independent one, should be
rigid and strong; for if there is any spring in it, though the cylinder
may be brought in sufficiently close when there is much cutting on the
card, if a blank card or one with very little cutting on it comes on,
the extra pressure on the needles, especially with a large machine,
may prevent it from getting in sufficiently close to clear the hooks
from the griffe. With self-acting motions there may sometimes be some
trouble in this way, as the weight of the griffe may not be sufficient
to press in the cylinder. In this case the griffe may be weighted, or
may be allowed to drop quicker, or the lifting rod and lever may be
made to assist in pressing it down somewhat.

Before starting the machine the needles should be examined to see
that they are all free, and that they will spring out easily after
being pressed back. The griffe should be perfectly horizontal and all
the knives properly set; the holes in the cards and cylinder should
exactly correspond, and when the cylinder comes in the needles should
be fair in the centre of the holes; if not, the cylinder must be set as
described (see description of Figs. 26 and 27).

The driving of heavy single-acting jacquards will be further considered
under Twilling Jacquards.

When any of the hooks or needles in a jacquard get bent or broken, they
can easily be straightened, or taken out and replaced by others. By
putting a thin blade of iron or wood down through the needles alongside
of the hook to be replaced, and springing open the passage, the old
hook may be drawn out and a new one put into its place. The tail cord
must, of course, be cut off the bottom of the hook, and a new one tied
on. For changing a needle take off the spring-box and draw up the pin
which fastens the row of needles at the back; then the needles in this
row may be taken out till the defective one is reached, and the row
made up again; or, the old one may be renewed without taking any of the
others out. A flat blade is used to slide through the hooks and keep
clear the place for the needle to be put in.

_Card Frames._--The cards for jacquard work are usually hung on a
frame as O, Fig. 18, wires sufficiently long to catch on both sides
of the frame being tied to the lacing of the cards. The number of
cards between each wire may vary to suit the space and the quantity
of cards. Sixteen to twenty suit very well, the former for small and
the latter for larger sets, and for very small sets twelve or fourteen
might be more convenient. The frame may be made of round iron rod, or
of flat or bar iron, and should be of the shape shown in the sketch,
and not semicircular, as is usually the case, which presses the cards
together in the centre; almost flat at the bottom, with just enough
of a slope to make the cards slide back, is much the best. Of course,
for a few cards it does not matter much what shape it is. P (Fig.
18) is the frame for the rollers over which the cards travel to the
cylinder. They should be so sloped as to make the cards travel up
nicely with sufficient drag on them, and not too much; on this depends
a good deal the proper working of the cards, particularly when the
machine is running at a high speed, and when springs are not used to
steady them on the cylinder. It is usual to have a roller below the
cylinder, attached to the frame which holds the cylinder, and the
falling cards pass over it. Sometimes the cards, when falling, are
shaken, so that some of the wires may not catch on the frame, but pass
through it, or the ends of some of the wires may be bent and cause
the same result; if the weaver neglects to put these up on the frame
before the cards work round to those that have fallen, the wire may
catch on it and prevent the cylinder from turning, or, perhaps, pull it
out. Sometimes a crank or bend is made in the frame at the outer end,
so that the wires can pass up through it without catching. The frame
should just be sufficiently wide to enable the cards to pass freely
through it with, say, one-eighth of an inch clearance at each side;
then, if three or four inches at the outer end is cranked, or set out
so as to be a little wider between the two bars than the length of
the wires, they will pass up through without catching. Frames of the
shape described are, of course, only suitable when the cards are to be
wrought forwards--that is, with the cards falling between the cylinder
and the machine; but if they require to be wrought both backwards and
forwards, as is sometimes the case, the card frame must be made more
of a semicircular shape, and the rollers must be set so as to give a
sufficient fall to the cards, and keep them firm on the cylinder.

As has already been said, the nature of the shedding of a single-acting
jacquard is objectionable for speed in working, for ease on the yarn,
and for heavy work, or for well-covered work. The jacquard harness
is levelled so that the yarn is all sunk, and the shed is entirely a
rising one; it can easily be understood that when the griffe rises to
open one shed, it must again fall before it can begin to rise to form
the next shed. Now, mostly all tappet motions, and a great many dobbies
or shedding motions, either have the yarn springing up and down from
the centre, or have one portion rising and the other portion falling
at the same time, so that in them the second shed could be open at the
same time that the griffe in the single-acting jacquard had fallen to
begin to rise the second shed; but as this would be much too soon,
they can take a greater time to accomplish the work, and thus have a
much slower and steadier motion.

It has been attempted to work the single-acting jacquard on the
centre-shedding principle, and machines are at present being made in
France of this class. It is only necessary to let the board on which
the upright hooks rest fall at the same time that the griffe is rising
in order to accomplish what is required, and the method of working is
good, and would suit well in dobbies where the heddles can be taken
firmly down. But in the jacquard the drawback is in the harness: the
constant rising and falling causes a vibration in it, and does not
admit of nearly so firm work as when the rising shed alone is used.

[Illustration: Fig. 30]

One of these machines is shown in Fig. 30; they are a very compact and
neatly made machine, and contain a much greater number of hooks than
one of the English machines. Fig. 26 gives one of the hooks in this
machine, and the way in which it acts is there explained. It will be
seen that the frame for the cylinder is inverted; F is the face-plate
or needle board as it rests when the cylinder is not pressing against
it, being held in this position by the spring H pressing on a stud on
the bar I, which extends from the face-plate to the clap-board G. The
cylinder frame is driven by a rod, C, connected with a lever. E, E^1
are two racks on the ends of the hole board on which the hooks rest,
and D, D are two levers with quadrant racks on each end, which work
into the racks E and E^1, and also into the racks on the slide bar
_a^1_. This slide bar is fastened to the griffe A A, which is raised by
a connecting-rod from a horizontal lever to the stud _a_. As the griffe
is raised one end of the racked levers, D, is also raised, and the
other end sinks, taking down the hole board by pressing on the racks E
and E^1. The hole board is fixed so that it will easily slide up and
down. If the griffe is raised two inches, the hole board falls rather
more than one inch.

This is a very good motion, perhaps could not be surpassed for the
purpose, but, as I said before, has the objection which all similar
contrivances for the purpose must have--viz., causing too much
vibration in the harness. All the cords constantly dancing up and down
causes an unsteadiness and swinging that is not to be found in the
ordinary jacquard harness. These machines have sixteen rows of needles
in them, which are much closer set together than those in this country,
and give a large number of hooks in a small-sized machine, which is a
great advantage when extensive patterns are required. In this machine
the half-card contains 440 holes, or 880 holes to the full or double
card. The pitch is shown at A, Fig. 31, which represents the end of
one of these cards, and a piece of one of the usual 8-row cards as
here used is shown at B for contrast, the black dots in both cases
representing holes.

These machines have not been adopted here, nor is it likely that they
will be, although they are said to work well in France. They have
been tried here, and the difficulty lay in setting the cards properly
so as to act correctly on the needles, and keeping them so for any
length of time. The least contraction or expansion of the paper, or any
irregularity in the cutting, any bend in the points of the needles, or,
in fact, anything but perfect exactness, interferes with the working:
there is too great compactness in the machine, and in the ordinary
wear and tear of work a little allowance is necessary for success.
Considering that these are working, it would appear that the pitch
and size of the holes in our cards are unnecessarily large, except
where small machines answer and saving of space is no consideration;
in that case the extra paper required for the cards might be more than
compensated for by the strength of the machine, and the saving of
trouble in attending to it.

[Illustration: Fig. 31]

The method already mentioned of getting over the disadvantage of the
single-acting jacquard is not likely to gain general favour, and is
not required, as it is surpassed by the double-acting jacquard, or
that in which there are two griffes, one rising when the other is
falling, forming a counterpoise as well, making the shed more after
the principle of ordinary tappet shedding. In a loom fitted with a
single-acting jacquard, if there is much weight to be lifted, it will
turn round the loom so that it will rest in no position but with the
griffe down, and this is frequently of so much annoyance to the weaver
as to necessitate a counterpoise being applied to balance it. Sometimes
weights are used, and sometimes springs. With the double-acting lift
nothing is required, and this was first used in Cross’s counterpoise
harness (about 1816).

[Illustration: Fig. 32]

[Illustration: Fig. 33]

In the double-lift machine there are double the number of hooks that
there would be in a single-lift machine of the same size, but the same
number of needles; each needle is connected with two hooks, as shown in
Fig. 32, which gives one row of hooks and needles for an 8-row or 400
machine. There are two griffes, one working above the other, as shown
at A and B, Fig. 33. E, E is the top griffe in both; C, C is the bottom
griffe, shown complete at A, but at B the side-bar of the frame is
omitted, leaving only the knives to show how they fall in between those
of the bottom griffe; D, D^1 are the two slide rods or spindles which
keep the griffes steady in their traverse. There must be a sufficient
space between the frame C of the lower griffe and that of the upper
griffe, E, to allow of the required draw being given to form the sheds
without them coming into contact. In these machines, although there
are 16 hooks in the row, they only act as eight, so far as forming the
pattern is concerned, as two hooks are governed by one needle; the
additional hooks are solely for the purpose of obtaining a rising and
falling shed. Each pair of hooks is connected together at the bottom
by a piece of strong cotton cord, called the tail cord or tug cord, as
shown in Fig. 35. The griffes are raised by two levers arranged side
by side, and similar to the one shown in Fig. 17. These levers are
wrought by rods connecting them to a double crank, or sometimes to a
tappet, on the tappet shaft of the loom. The common form of crank is
shown at A, Fig. 34. B is the tappet shaft, and on the end of it is
fastened a disc. A second disc, which carries the double crank, G, is
bolted against this one, and can be shifted to whatever position is
required to give the tread at the correct time. The cylinder is driven
from an eccentric on the crank-shaft as before described. One of the
griffes rises for each shot, the other falling at the same time; but
the cylinder must come in for every shot, and it is here that the
principal fault in this machine lies. Not only has the cylinder to
travel at a high speed, but when one griffe is at its highest position,
it has to press against the needles, so as to clear away those hooks
that are not to be lifted by the lower griffe, which should now be
beginning to rise; and in consequence of the needles being attached
to two hooks, one belonging to the upper griffe, and the other to
the lower one, it follows that those needles that are pressed in by
the cards have to spring back the hooks connected with them that are
raised by the upper griffe, which is a severe strain on both needles
and cards. To prevent the hooks being pushed off the upper griffe, they
must have larger turns on them than would otherwise be necessary.

[Illustration: Fig. 34]

Fig. 35 shows how the raised hooks can be allowed to yield to the
pressure of the needles. A, B, C, D are four hooks connected with two
needles, 1 and 2. When the hook A is raised, if B, which raises the
same warp--as will be seen by the connection of the tail cords with
the harness at E--is not to be raised for the next shot, the needle 1
is pressed back by the card, and presses the hook B back from its knife
as shown, the dotted line being its original position; the A hook,
being connected with the same needle, must also be sprung back, but
its head cannot get back, as it is held on the blade of the griffe, so
that the wire would require to spring, if some escape were not made for
it. The lower ends of the hooks are made V-shape in the grating, so
that when raised, if pressed on by the needles, they can move forward
as shown, the dotted line showing the original position of the hook
A. When the hooks fall, they fill the slots in the grating, and are
thus kept steady. G shows the tail cords as connected with the harness
when both hooks are down; E shows them when one hook is up and the
other down, and if the hook B was raised for the next shot, the tail
cord on it would be tightening up as that on A would be falling, and
the harness attached to them would be caught up a second time from the
middle position, thus forming centre shedding with any portion of the
harness that is raised several times in succession. With plain-texture
cards the upper and lower portions of the shed would pass each other in
the centre when the sheds are being reversed, all the even numbers of
hooks being on one griffe, and all the odd numbers on the other.

[Illustration: Fig. 35]

[Illustration: Fig. 36]

The tail cord consists of two pieces of cotton cord, one fastened
to each hook of a pair, then the two ends are together tied to the
bunch of neck twines that are to hang from these hooks, as shown in
Fig. 35, and at A, Fig. 36. When one hook is raised and the tail cord
drawn up with it, the other portion of the tail cord, which is tied
to the other hook, is slackened--as shown at A, Fig. 36--which causes
a certain amount of friction on them. Also, when one hook of the pair
is falling with the descending griffe, and the other hook rising, the
pluck occasioned thereby on the cords, when the hooks are passing at
the centre, has a tendency to wear and break them. Although this does
not occur when the lingoes are of a moderate weight (18 to 25 per lb.),
and when only a few neck twines are tied to each tail, yet when a large
number of neck twines (say 20) are tied to each tail, with weighty
lingoes, as may frequently be the case in weaving small patterns on
woollen and worsted goods, the breakage of the tail cords is a common
source of complaint, which not only gives the trouble of renewing
them, but is liable to cause defects in the cloth, by the weaver not
observing the breakage for some time, as one hook of the pair may be
raising the neck twines--that is, in case of the tail cord to only one
of the hooks breaking. To remedy this Messrs. Hancock, Rennie, and
Hudson have this year (1890) introduced a patent link connection for
joining the tail to the hooks, which only requires one cord, or double
cord, to be used, instead of two as before. This is shown in Fig. 36
at B. When one hook is raised and the other down, the link is in the
position shown at C. The old method is shown at A. When one hook is
rising and the other falling, the partial turning of the link causes
a loss of time equal to about a quarter of an inch of lift, and this
eases the sudden pluck on the tail cord. When both hooks are down there
is also a loss of a quarter of an inch in the lift when one hook begins
to rise, caused by the turning of the link.

[Illustration: Fig. 37]

[Illustration: Fig. 38]

This patent works very well, but, except when heavy weights are on the
hooks, is not likely to supersede the older methods. These links are
fitted to machines by makers in Manchester and Bradford.

Before the form of griffe shown in Fig. 33 was adopted the hooks were
made of two heights, as shown in Fig. 37, and one griffe wrought
above the other, instead of the one set of blades or knives passing
through each other; but this method was given up on account of the
vibration of the long hooks, which made it uncertain whether they would
remain on the knives or keep clear of them when required. Fig. 37
shows this arrangement of hooks for a double-cylinder machine; for a
single-cylinder machine with two griffes the heights of the hooks would
be similar, but the heads would all be turned in the same direction,
and the knives sloped to suit this, as is done when the hooks are all
of the one height; each needle would be connected to a short hook and a
long one. The short hooks give much more certain work, and can have the
cylinder set so as to press back the hooks only as much as is required,
whereas with long and short hooks allowance had to be made for the
uncertainty of the vibration and the difference in the length of the
two sets of hooks.

Fig. 38 is a view of one of the best makes of double-lift jacquards
with a single cylinder.

The machine is made by Devoge & Co., of Manchester, but is not here
given as being specially recommended in preference to others; it is
only given as an illustration. Those wanting to buy a machine had
better see what are in the market, and select what they consider most
suitable to their work and price. This applies to all the machinery
given in these articles. There is rarely a best machine for all
purposes.

[Illustration: Fig. 39]

These machines are much in use, and can be run at a high speed, say
160, or even 180, and by many are preferred to the double-cylinder
machines, as there is no danger of one cylinder getting before the
other, and the cards are all laced in one set; besides, it may be
more convenient for working, as some arrange their machines so that
the cards for one loom hang to the back, and those of the next to
the front. When these machines are to work at a high speed, the
slide-motion cylinder will probably be found the most satisfactory.
Fig. 39 illustrates one of these machines with slide cylinder motion
and the levers for driving it, the connections being as before given.
The levers for raising the griffes are also shown, but here go to the
back, whereas they are usually at the side.

The most perfect jacquard machine in the market is undoubtedly the


DOUBLE-ACTING JACQUARD WITH TWO CYLINDERS

[Illustration: Fig. 40]

The only drawback to this machine is, except what may be said against
the method of shedding, the liability of one cylinder to be turned out
of time, or get a shot or two before the other, so as to put the cards
off their proper rotation; but this is only a difficulty in the hands
of inexperienced weavers; nevertheless it exists. The effect will be to
spoil the pattern on the cloth, giving the twill a mixed or broken-up
appearance. There are motions in use for stopping the loom, unless the
cards come in rotation, but many prefer to work without them. Fig. 40
is a view of a two-cylinder machine made by Messrs. Devoge & Co., with
swing-motion cylinders, which are, perhaps, the best motions for these
machines, as they do not require to travel quickly. A very good speed
for the machines to work at is 160 to 180 or 200 picks per minute,
and the cylinders would only travel at half this speed. The cylinders
should be driven by an eccentric, same as given for the single-acting
machines, but instead of being on the crank shaft, it should be on the
tappet shaft, which runs at half the speed; and as the two cylinder
frames are connected together, when the one is going out the other is
coming in, so that one eccentric making a revolution for two beats of
the slay will drive both cylinders. Sometimes the eccentric is on the
tappet shaft, inside the framing of the loom, and is connected with the
top lever, as shown in Fig. 34 (B). C is the eccentric, D the fulcrum
of a short lever attached to it, and E the upright rod attached to
an arm or lever on a horizontal shaft supported by the machine, or
on the top of the loom. A lever from this shaft on each side of the
machine drives the cylinders. The eccentric may be on the end of the
tappet shaft, same as it is shown on the crank shaft in Fig. 18; but it
might not always be convenient to have it here, and perhaps the most
desirable way to have it at any time is to have a pinion on the crank
shaft with, say, twenty teeth in it, and a stud wheel alongside with
forty teeth gearing into it. On this stud the eccentric can be fixed,
and will give a very steady and convenient method of driving. The
griffes are raised in the same way as for the double-lift jacquard with
one cylinder. (See Fig. 34 (A)).

[Illustration: Fig. 41]

[Illustration: Fig. 42]

Fig. 41 shows the arrangement of a row of hooks and needles for a
double-cylinder machine. The top needle of the upper set and the bottom
needle of the under set are attached to two adjoining hooks, which
are connected together with the same tail cord. This arrangement is
to enable the cards when working at both sides to act on the correct
hooks, which will be better understood by referring to the description
of lacing cards for these machines. In Fig. 41 it will be observed that
all the hooks are vertical. Sometimes the hooks are slanted a little,
as in Fig. 42, to give more space between the hooks at the top, without
increasing the width of the machine, and there is a slight difference
in the arrangement of the hooks and needles, as is shown. Both work
very well.

Fig. 43 shows a two-cylinder machine, by Messrs. Devoge & Co., with a
slide motion for the cylinders, which would be driven in the same way
as the swing motion.

It has been said that the shedding of a single-acting jacquard is of
the worst description for general weaving. That of the double-acting
machine is by no means perfect either; some prefer the single-to the
double-acting for making fine damask. Jacquard shedding cannot be
regulated in the same way as tappets or the best shedding motions can.
In tappets the dwell can be regulated to suit the cloth required, and
the time of the shedding can be made early or late as desired. In
jacquards this cannot be done to anything like the same extent. The
shed must always be open in time for the pick, and the pick should
begin when the cranks are about the bottom centre, a little earlier or
later, as desired. The jacquard must have the shed open at this time,
and must keep it open till the shuttle passes through.

It has been said that the usual method of raising the griffe or griffes
is by a crank (or a stud in the wheel, which is practically a crank)
on the crank shaft for a single-acting machine, and by a double crank
on the tappet shaft for a double-acting machine. Now, a crank gives a
continuous eccentric motion with a slight dwell when it is at both top
and bottom centres. Sometimes a tappet is used to raise the griffes
instead of a crank, and of course any required dwell can be made on a
tappet; but then it must suit the jacquard, and the greater the dwell,
the less time is occupied in the rising and falling of the harness. If
the harness is plucked up or dropped down too quickly, the result is
a dancing or unsteadiness of the cords, and in a double-lift machine
there will be a considerable plucking when the hooks that are rising
take up the cords that are falling. In order to keep the cords as
steady and free from vibration as possible, if there is any swinging in
the weights or lingoes, it is a common practice to put a frame round
these, with wires run through it at whatever distance apart is thought
desirable, so as to partition them off in bunches and keep them from
swaying about.

[Illustration: Fig. 43]

The smoother and slower the harness can be raised and lowered, the
better. Therefore, to get a high working speed, the time or the portion
of a revolution of the crank shaft given to the rise and fall must be
as great as possible, so that very little could be gained by using a
tappet. If a tappet is to be used, a box tappet will be required, or
is more satisfactory for a single-acting machine, in order to make the
griffe in falling follow the tappet and avoid any plucking or jerking;
sometimes a fork lever, with the tappet or wiper working between the
prongs of the fork, is used for the same purpose. For a double-acting
machine double wiper tappets, acting on levers or treadles, are
sometimes used, the griffes falling of their own weight; the tappets
are nearly round eccentrics, or like plain tappets with a very short
dwell, not more than one-fourth of a revolution of the crank shaft. It
may therefore be considered that the harness should always be moving
either up or down, with a small pause when the griffes are at the top
and bottom, to admit of an easy turn and to allow the shuttle time
for its passage through the shed. With the crank drive the shed will
require to be opened a little wider than if the dwell was as great as
it should be, especially in wide looms, in order to let the shuttle get
through freely; but it would be more desirable, and a saving of strain
on the yarn, to have the dwell greater, and not open the shed any wider
than is necessary to admit the shuttle.

Speaking generally, the usual rule for the time of shedding may be said
to be to let the shed be closed when the cranks of the loom are at the
top centre, or perhaps one-sixteenth of a revolution farther forward,
and let the shed be full open when the cranks are about the bottom
centre. It therefore follows that the single-acting jacquard must open
the shed in a little less than half a revolution of the crank shaft,
and close it in the same time; but the double-acting machine takes
nearly a full revolution (three-quarters, or a little more, should do)
to either rise or let fall one of the griffes. It will be evident that
there is a considerable difference in the nature of the shedding. A
single-acting machine requires the weft to be beaten up on a closed
shed (the time of the crank in coming from the top to the front centre
being taken up with rising the griffe from the bottom up to catch the
hooks), whereas in a double-acting machine the case is different;
though the shed is closed at the same place, or in the same position
of the cranks, the griffes are in an entirely different position. The
closed shed with the single-acting jacquard is when the griffe is down,
but with the double-acting machine it is when the two griffes are on
a level--that is, halfway up. Of course, more or less of the yarn may
be at the bottom position, so that there may be no closed shed, or
in no position of the griffes may the yarn be all on a level, unless
none of the hooks are on either of the griffes, or if one griffe has
all the hooks on it. When the single-acting machine begins to open
the shed, the driving-crank of the griffe is upright or at the top
centre, and in the position to give the slowest motion to the griffe;
whereas for the double machine the lifting cranks are horizontal
when the shed is closed, and in the position to give the griffes the
quickest motion; therefore, when the lay gets to the fell of the cloth
in a single-acting machine, the shed is still close, whereas with a
double-acting machine it is fully half open. From this it follows that,
to get a close covered cloth, the double-acting machine is the better,
as the weft is beaten up in a crossed shed; but to get a clean-surfaced
fabric, with the weft lying straight between the two portions of the
warp, the single-acting jacquard is better. It is for this reason that
the single-acting machine is preferred by many workmen for making
fine damask, which does not require much covering to give it a good
appearance, and there is less chance of having cut weft and of looping
when striking on the open shed. Of course the weft may be cut in a hard
fabric by the reed having to strike too heavily against it, and in this
case striking on a crossed shed might prevent the cutting, by the weft
going on easier.

In hand-loom linen damask the shot is struck up when the shed is about
half closed; the warp is held so firm in the loom that there is no
spring in it, and the weft does not rebound. A cleaner surface is thus
made on the cloth than if the weft was struck up in a closed or cross
shed.

In a single-acting machine the faults in the cloth are more readily
seen than with a double-acting one, as, in case of a hook missing the
knife, in a single machine it would show in a short time, whereas in a
double-lift machine one hook of the pair might be missing the knife and
the other taking it, so that it might be some time before it would be
observed. One of the tail cords breaking might have a similar effect.

The foregoing gives a general description of working the jacquard; but
no hard-and-fast lines can be laid down--a little variation may be
necessary at any time, to suit circumstances and the class of work.


JACQUARD STOP MOTION

When working with two cylinders, one may happen to be turned at a
time when it should not, and thus put the cards out of rotation.
Many attempts have been made to overcome this difficulty by stopping
the loom when the cards get out of the proper order, but none of the
methods adopted have ever gained much favour. Recently a new motion has
been patented, and is being applied to machines by Messrs. Devoge &
Co., of Manchester. It is called the ‘Devoge jacquard stop motion.’

[Illustration: Fig. 44]

Fig. 44 shows how it may be applied to a machine. The hooks A and B
and the needles E and F are those here used to work the motion. They
may be at either side of the machine, but should be at the side of the
belt handle. One hook must belong to the front cylinder needles, and
the other to the back ones. The hook A requires a lingo attached to it
to draw it down after being lifted. The hook B is attached to a lever
connected to the side of the loom, so that when one end is raised a
hammer on the other end pushes off the belt handle (a lever and bracket
are supplied for the purpose). C C^1 is a wire bell-crank lever with a
turn or loop on it at D to act as a spring. The fulcrum is at L, on a
piece of iron bolted to the edge of the machine; the lower end of this
lever passes through an eye in the hook A at C^1, and the upper end
passes through an eye in the needle E. There is a spring on the point
of the needle E, between the face-plate and the eye, which the lever
goes through, and which holds the needle back as shown.

Each time the hook A is raised the lever presses forward the needle E,
and with it the hook B, which would then be raised by the lower griffe,
unless the card pressed the needle back again and pushed it off. Thus,
by having a hole cut in the cards for the needle F, and none for the
needle E, the hook B would never be raised; but if a hole is cut for E
in a card following one in which a hole was cut for F, the hook B would
be raised and the loom stopped. It is, therefore, only necessary to
arrange the cutting of the cards to allow the loom to work when they
are following each other in rotation; but as soon as one card gets out
of order the loom should be stopped, though, perhaps, not till it has
run for a few shots. Thus--

     Number of cards--

    1    3    5    7    9    11
    --   --   --   --   --   --
    O    ·    ·    ·    O    O

     Cut the large dots for the needle E.

    2    4    6    8    10   12
    --   --   --   --   --   --
    O    O    O    ·    ·    ·

     Cut the large dots for the needle F.

This gives a repeat of twelve cards; but any number to suit may be
used. Thus--

     Number of cards--

    1    3    5    7    9    11    13    15
    --   --   --   --   --   --    --    --
    O    ·    ·    ·    O    O     O     O

     Cut the large dots for the needle E.

Number of cards--

    2    4    6    8    10   12   14   16
    --   --   --   --   --   --   --   --
    O    O    O    ·    ·    ·    ·    ·

     Cut the large dots for the needle F.

This gives a repeat of sixteen cards. The even numbers of cards go to
the front or top cylinder, and the odd numbers to the low cylinder,
and it may be seen that a hole in an even-numbered card following one
in an odd-numbered card will not stop the loom; but a hole in an odd
number following one in an even number will stop the loom, as it is the
hook B rising after A that stops it; therefore any suitable rotation
of cutting may be adopted, and the stoppage can take place either at
short or long intervals, as desired, the principle being to raise the
hook A two or three times, and push the hook B back again by having no
hole cut for the needle E. Then leave A down for three or four shots,
and cut holes for E, which have no effect unless the cards get out of
rotation, and one of those with a hole cut to raise the hook A comes
before one with a hole cut for the needle E, when B will be raised and
the loom stopped. This is a good arrangement, and works very well.

Another motion for a similar purpose, invented by the writer, is shown
in Fig. 45. It is based on the following principle: Suppose a cord is
taken from any two hooks of the jacquard, and passed round a pulley on
the ‘hound tail’ or long lever of the weft fork motion; if the cord is
left slack, so that raising one of the hooks will just tighten it, then
raising both hooks together will lift the lever, and can be made to
stop the loom. The difficulty to be got over is that one of the hooks
must belong to one griffe, and the other to the other one, in order
to make the motion act with the two sets of cards. As the two griffes
pass each other at the centre, or at the half-lift, this must be taken
as the full lift, the cord must be stopped here, and not drawn any
farther; for the remaining portion of the lift the hooks must draw a
spring. This can be easily arranged by having loops on the cord passing
round wires in the cumber board, or by having the two ends of the cord
passing through a small hole board, and having knots or beads on them,
beneath it. Other methods may also be adopted.

One of the most desirable arrangements is shown in the figure: A, A
are the two hooks; B, B are two small springs by which the two levers,
C, C are attached to the hooks with cords; D is the frame for holding
the levers, and is fastened to the top rail of the loom, under the
jacquard, or in any convenient place. It will be observed that the
front bar of the frame passes above the levers, so that it will prevent
them rising above the half-draw of the hooks, in which position they
are shown. E, E are two cords connected with a jack or tumbler, F, on
the end of a bell-crank lever, G H L, having its fulcrum at H, which
may be on the same stud as the weft fork lever, or in any convenient
place.

[Illustration: Fig. 45]

The weight of F and G keeps the cords in tension, and the point L of
the lever is set behind the lever on the loom which carries the weft
fork, at such a distance from it that when one of the cords E is drawn
it does not act on it; but when both hooks are raised, drawing up the
two cords, the point L of the lever presses against the weft fork
lever, pushing off the belt handle and stopping the loom.

The lever G H L may act directly on the belt handle, if desired; in
this case it would be fixed outside the loom framing. The cards are cut
on the same principle as for the last motion, but the same holes will
do in both sets of cards, as the two needles are acted upon by the same
number of holes in both back and front sets; that is, for two hooks
coming beside each other.

The following order of cutting will answer:--

Number of cards--

     1   3   5   7      2   4   6   8
    --- --- --- ---    --- --- --- ---
     O   O   ·   ·      ·   ·   O   O

Cut the large dots on the number of cards given, and of course they
must be cut to suit the needles connected with the hooks used.

The above gives a continuous working of the motion, but it would be
sufficient for it to work at intervals having 8 or 10 shots between
them, as--

Number of cards--

   1  3  5  7  9  11  13  15  17  19  2  4  6  8  10 12  14  16  18  20
  ---------------------------------------------------------------------
   O  O  ·  ·  ·  ·   ·   ·   ·   ·   ·  ·  O  O  ·   ·   ·   O   O   ·

This will not allow the loom to run for more than 20 shots after the
cards get out of order.

Some other motions are in use, but these are simpler.

Before describing twilling machines or any special make of jacquards,
it may perhaps be better to explain the mounting of ordinary machines,
according to the usual methods adopted in some of the leading
districts.




CHAPTER III

_FULL-HARNESS MOUNTINGS_


The mountings that will be alluded to in this section are all intended
for ‘full harness’; that is, the ordinary method of jacquard weaving
when applied to such fabrics as damask, dress goods, and, in many
cases, to double cloths, handkerchiefs, &c.

_Full harness_, or, as it is termed in some hand-loom districts, shot
and draft, to distinguish it from the ‘pressure harness’ system, holds
the first place amongst mountings, as by it can be done what cannot be
done by any other method, and everything can be done by it that can be
done by any harness, or combination of harness or heddles; though it
does not follow that it would be desirable to adopt it in every case.
In a full harness each thread in any part of the tie can be lifted
independently of the others; in wide fabrics, or in order to reduce
expense in narrow ones, recourse is had to gathering or repeating the
tie, and like threads in each repeat must be lifted together; but in a
single tie any one thread can be lifted independently of any other. In
full harness there is one thread of warp drawn into each mail or eye
of the harness, and one shot is given to each change of card. Other
descriptions of harnesses are known as ‘half harness,’ ‘split harness,’
‘pressure harness,’ ‘gauze harness,’ double-cloth and quilt harnesses,
&c.

_Half harness_ is the term applied to that description of harness in
which the half of the warp (every alternate thread) is drawn into the
harness, and the other half passes through it loosely. This is mostly
to be found in the manufacture of gauze or leno curtains, and is
wrought with a shaft mounting in front of the harness.

_Split harness_ is applied when two threads are governed by each hook
of the jacquard, and the ground of the cloth is wrought by some other
means, as shafts through the harness. This is to be found in the silk
trade.

_Pressure harness_ is when several threads of warp are drawn into each
mail, and when the harness is drawn it remains stationary for several
shots of ground texture, for which the sheds are sprung or pressed open
by heddles.

_Gauze harness_ is a harness fitted up with doups for weaving leno and
gauze, and is mostly full harness.

_Double-cloth and quilt harnesses_ are for weaving these fabrics.

Mounting, or gaiting, is a term that may be taken to apply to the
building of the harness and the preparing of everything in connection
with it. The form of ‘tie’ to be adopted will depend to a great
extent on the nature of the fabric to be made, and to the style of
pattern which is to be applied to it. For instance, dress goods may
be required, and the pattern may consist of small sprigs or objects
repeated over the surface of the cloth, forming a simple repeating
pattern. Again, handkerchiefs, napkins, or table-covers may be wanted,
which will require a bordered ‘tie,’ and may have both single and
double mounting in them. Sometimes the manufacturer mounts his looms
to what he considers a desirable ‘tie’ (or arrangement of cords) to
admit of having a good variety of patterns wrought on it for whatever
class of goods he is likely to make, and then he has the patterns made
to suit the tie or mounting. Sometimes patterns are procured which
will give the best possible effect on the least possible machinery, or
the least number of hooks of the jacquard, and the mounting is then
arranged to suit the pattern. On this method often a very considerable
saving of machinery can be effected by turning over or gathering the
harness; or, by arranging a variety of turns over, or gathers, and
repeats, a very diversified effect can be obtained with a small number
of hooks, as those accustomed to the larger forms of shaft mountings
will readily understand. Thus, if we take 100 hooks of the jacquard as
equal to 100 shafts, it will easily be understood that a large variety
of beautiful patterns can be produced with either a straight or zigzag
draft, though they will be mostly of a set or conventional type.
The objection to mounting a loom in this way is, that if a change of
pattern to a free or running style is required, it is necessary to cut
down the harness and remount it, probably requiring new machinery as
well. It may, however, suit to adopt both methods; that is, to have a
few looms for working conventional patterns on small jacquards, say 200
hooks, and others mounted for free patterns requiring, say, 400 or 600
hooks in the machine. Of course this entirely depends on the nature of
the orders likely to be received, and manufacturers must use their own
judgment, in which, however, they are more likely to err on the narrow
than on the liberal side of the question, the result being cramped and
stiff patterns, with a probable loss instead of gain. Two styles of
mounting harness are in general use--one, known as the ‘London tie,’
being used in Spitalfields by the silk weavers; the other is called the
‘Norwich tie,’ as it was there adopted in the early days of weaving.
About 1830 the Norwich style was adopted in London, as the weavers’
houses were too low to admit of the jacquards being set high enough to
suit the London method, which, having a quarter-twist in the harness,
requires more height than is necessary for the Norwich system, in which
the harness passes direct from the jacquard to the cumber board in flat
rows without any twist.

_The Harness._--When about to mount a loom the first process is to
prepare the harness. A harness is built up of several parts, the
methods of preparing and building varying in different districts. A few
of the best methods will be given. Fig. 46, Nos. 1, 2, and 3, show a
complete cord of the harness in three methods of preparing it--A (Nos.
1 and 2) are the tail or tug cords looped to the hooks of the jacquard;
these cords are only necessary on double-lift machines, and are usually
put on by the machine-makers. They are made of twisted cotton, and
are soft and pliable. L shows the knot, known as the ‘tug knot,’ by
which the tail cords are looped on the hooks. B, B are two methods of
knotting the neck or body of the harness to the tail cords; that in
No. 1 is the usual method. C, C are the hecks, or guide reeds, through
which the harness passes. G, G are the lingoes, or leads, formerly made
of strips of lead, but now of wire. From the lingo to the mail or eye,
F, through which the warp passes, is a double cord called the bottom
piece, or hanger. From the mail, passing through the cumber board or
harness reed E, is another piece of double cord, called the top or
mid-piece, or the ‘sleeper’; to this is looped or tied the neck twine
in various ways, two methods of which are shown at D _d_ _d_^1 (No. 1)
and at D (No. 2); No. 3 shows levelling below the cumber board, with
two methods for tying H and H^1. M is the snitch knot, which is much
used for fastening cords that require careful adjusting. Sometimes it
is a matter of choice to adopt any method of mounting, and sometimes
one plan may suit circumstances better than another.

[Illustration: Fig. 46]

Preparing Lingoes.--A prepared lingo is shown at No. 4, Fig. 46, this
work being generally done by little boys or girls, or by old women.
One of the commonest methods of preparing them is as follows:--Having
prepared the mails, which are small eyelets of brass, copper, or steel
of the shape shown at F, sometimes with round holes in the centre and
sometimes with elliptical or long shaped ones, they are put into a dish
or pan; a boy takes a piece of wire and strings a number of them on it.
The wire is then fastened on a rack, or in any convenient place, by
both ends, in a horizontal position. A bunch of small pieces of twine
is tied up in a convenient place; these are to form the hangers, or to
connect the lingoes with the mails. Having the lingoes conveniently
placed, and being provided with a pair of shears, the boy sets to work.
Taking a piece of twine, and pulling it through an end hole of a mail,
he doubles it evenly, and, lifting a lingo, puts both ends through the
eye in it, and casts on a knot, as shown at N (No. 4). The loop on the
twine there shown requires to be pushed up over the top of the lingo,
then drawn tight, and the ends clipped off. This mail is then pushed
along the wire, and the others proceeded with. The pieces of twine for
forming the hangers, as well as the sleepers, are prepared by warping
them off spools round two pins, and then cutting them across. Of course
the pins must be set apart at such a distance as will suit the length
of the sleepers and hangers required. The usual length of the hanger
(when doubled) is 7 or 8 in., and that of the sleeper or mid-piece when
it passes through the cumber board, as in Nos. 1 and 2, Fig. 46, and as
shown at No. 4, is 15 or 16 in. When a quantity of lingoes are hung on
the mails, the top cords or sleepers may be put through the top holes
in the mails, and tied with a weaver’s knot, the ends being neatly
clipped off.

[Illustration: Fig. 47]

A quicker method of preparing lingoes is as follows:--Two upright
pieces of iron, as A, A, Fig. 47, are fastened on a board, or on a
bench or table. A cut is made in each of these, so as to hold a piece
of flat wire such as is used for coarse reeds. This wire is pointed
at one end, so that by taking a handful of mails they can easily be
gathered up on it. It is then placed in position, as shown at B; the
mails, D, are shown on it, and C is a spool containing twine for the
hangers. By rubbing the end of this twine with wax, so as to stiffen
it, and gathering all the mails on the wire evenly together, the end
of the twine may be run through the holes in all of them at once. This
end is fastened, and with a hook, as shown at F, the twine can be drawn
through the mails, beginning at the one farthest away from the spool,
and passing from one to the other in succession, looping the twine
round the pin E, which should be at such a distance from the mails as
is required to make the hangers the proper length. If many mails are to
be threaded at once there should be two or three pins as E, so as to
avoid too much slope from them to the mails, which would increase the
length of the hangers. The twine can be cut when the mails on the wire
are completed, and the other ends of the mails turned round and treated
similarly for the sleepers; but the pegs will require to be shifted, so
as to get the required length of twine. When finished, these can be cut
also, and the wire tied up in a convenient place, so that the lingoes
may be hung on, as before described. The sleepers can then be knotted,
if required to be.

After the lingoes are thus prepared the twine is frequently varnished,
and sometimes twisted so as to prevent that of the hangers from rising
up through the hole of the mail and getting in between the two halves
of the yarn when the shed is crossing. This frequently happens if
anything prevents the lingo from dropping. A common method of twisting
them is for a boy to give them a coat of flour-paste or a light coat of
varnish, then, having them hung on a rod which passes through the loops
of the sleepers, he takes a handful of the lingoes, and, placing them
on his knee, rolls them round with his hand till they are sufficiently
twisted, then lets them hang down, and proceeds with another handful.

When the paste or varnish on these is set they can be well varnished.
About two inches at the top of the sleepers must be left without
varnish, to leave it soft and pliable for tying the neck to; the knots
on all the sleepers must be brought up to about the point before they
are twisted or varnished. When the lingoes are thus prepared and dry
they can be tied in bunches, to be used as required.

Some do not varnish the twines till the harness is all mounted, and
then varnish it all over. Others do not varnish at all. In this case
the hangers are not twisted, and for light open sets of harness they
need not be twisted at any time.


LINGOES OR LEADS

For pressure harness work the weights for drawing down the harness
were originally made of lead, about the length and thickness of a thin
pencil, but tapering to the lower end. The weight varied with the
number of ends to the mail, the strength of the yarn, and the weight
of the cloth to be made, ranging from 10 to 18 per lb., or lighter if
for fine hand-loom work. For a linen damask as formerly made, with 6
or 8 threads in the mail, a very heavy lead was required, even in the
hand loom, whilst a fine cotton two-thread harness would do with a
comparatively light lingo. Strips of lead cut from sheets were used for
lighter weights; these were then reduced to the proper size for drawing
them through holes in steel plates. Now, lingoes for power looms are
made of wire, and for ordinary damask or full harness work should range
from 25 to 40 to the lb.; for heavier work, such as linen towels,
worsteds, &c., 16 or 18 per lb., according to the weight of the fabric,
are required. Lighter ones answer for hand looms: Murphy mentions as
light as 110 per lb. for the centres of shawls, but 35 to 60 per lb.
will be more frequently found. 25’s to 30’s are a good size for power
looms; 12 in. long for 25’s or 30’s, 14 in. for 16’s or 18’s, and 16
in. for 10’s or 12’s are good lengths.

When there is a great slope in the border twines of a harness, it is
often necessary to put heavier lingoes on it than on the centre, to
keep the warp down. Of course, the less weight that has to be lifted
is always a saving of power, and easier on the harness; therefore it
is better to have two sizes of lingoes than to have them all weighty
enough for the borders.


HARNESS TWINE

The size of harness twine used varies greatly; some prefer a fine, and
others a heavy, twine. For the sleepers and hangers, 4 ply of 22’s or 5
ply of 30’s linen yarn is a very good medium size, and 4 ply of 14’s or
5 ply of 18’s or 20’s is a very good size for the neck or body of the
harness.

Sometimes, for heavy damask, cable cord is used for the body of the
harness, and is a good wearing cord; it is especially suited when the
neck cords are fastened to the sleepers below the cumber board. This
cord is made of good flax yarn; 5 ply of 30’s are twisted together,
and three of these cords are then twisted together, or 3 ply of 18’s
afterwards made 3 ply. A better size for medium work is 5 ply of 35’s
made 3 ply, or 3 ply of 20’s or 22’s made 3 ply. Heavy harness twine
requires to have heavier lingoes, particularly on those parts of the
harness that are much slanted, in order to have the same effect on them
that they would have on light twine. Some go to the opposite extreme,
and use very light twine, such as 4 ply of 30’s, which is only fit
for very light work in a narrow loom, where there will be but little
friction on the cumber board, and where no heck is required. Lighter
twine will suit better for a hand loom than for a power loom; 4 ply of
30’s for the harness of a hand loom, with lingoes of 50 or 60 to the
lb., would do very well for a light cotton warp.


SETTING THE JACQUARD

The proper position for the jacquard, when only one is required on the
loom, is so that the centre hook in it will be above the centre hole
of the cumber board. This can easily be found by tying a plumb line to
the centre hook of the machine and moving it, if necessary, till the
plummet rests over the centre of the cumber board. In case the cumber
board is not fixed in position, that of the jacquard may be found by
having the plumb line to pass about half an inch more than the half
breadth of the harness at the cumber board behind the top rail of the
lay when it is full back, and it should also be at equal distances from
each side of the loom. When more than one jacquard is required, they
should be arranged evenly over the cumber board, and as close together
as possible.

Sometimes the jacquard may be set farther forward or back to suit
circumstances, such as getting card space, the only disadvantage being
that there will be more slanting of the cords at one side than the
other, and more friction on them in the cumber board, also more drag
on the hooks on this side if no heck is used, and if a heck is used
the cords will have to bear the friction on it. The more direct the
cords of the harness can be, the better, and the above setting of the
jacquard should be adhered to when possible, but it is not absolutely
necessary for working to have it so.

It is always well to have the jacquards so arranged that they can be
raised or lowered a little by having them resting on bars, which can be
raised or lowered with screws. This is in case the harness is levelled
higher or lower than might afterwards be desired; but if the breast
beam of the loom, with the lay and back rail, can be raised or lowered,
it will suit the same purpose, and it is better not to move the machine
once it is fixed and the harness tied up.

The height the jacquard is to be above the loom must in many cases be
regulated by circumstances; for instance, the height of the roof, or
if the beams of the roof interfere with the working of it; sometimes
the machines rest on the framing of the loom--that is, if the loom
is made for a jacquard--and sometimes plain looms are used, and the
machines rest on beams supported from columns, or from the roof of the
house. This latter is the best plan, as it keeps the jacquard free from
the shaking of the loom, which is particularly useful in the case of
looms fitted with the knock-off motion. A good height for a jacquard,
independent of circumstances, is to have 8 ft. or 8 ft. 6 in. from
the mails to the bottoms of the hooks for a 10/4 loom--that is, one
with about 100 in. reed space; 7 ft. to 7 ft. 6 in. for an 8/4 (82 in.
reed space), and 5 ft. 6 in. to 6 ft. for a 4/4 (40 in. or 42 in. reed
space), are very good heights. 5 ft. 6 in. is about right for a narrow
harness, say 20 in. to 27 in. wide; 6 ft. would do for 80 in. wide, and
7 ft. or 7 ft. 6 in. for 90 in. wide, if necessary.


HECKS OR GUIDE REEDS

A heck is a frame of hard wood with wires across it, a wire for each
row of hooks in the machine, from back to front. It should be made
to suit the machine, so that when the harness passes from the hooks
between the wires, it will go vertically down, and have no slant in it.
In narrow hecks, there is usually one cross-stay to support the wires,
but for broader machines there should be at least two. The wire should
be iron, as brass soon cuts with the friction of the cords, and then in
turn cuts the cords. For a single-acting machine the wires should be
loose, so that they can roll with the cords; but for a double-acting
machine this would be of no advantage, as a portion of the cords are
falling when the remainder are rising. If the machine is very wide,
or has to be set forward or back on the loom, there should be cross
rollers of hard wood above the wires, at right angles to them, to
prevent the bottoms of the hooks from being drawn either backwards or
forwards, which might push their heads on or off the griffe knives.

Sometimes glass rollers are used in the silk trade for this purpose,
with mountings of the London tie, and while they are very smooth
and polish the twines, they get too hot if used in warm power-loom
factories working at a high speed. Hecks are not required for very
narrow harnesses, as the cords do not diverge much from the vertical,
and the friction on the heck being saved, the cords wear much longer.

In a wide harness it is impossible to draw an even shed without a heck,
although in some districts they are hardly known, and the more any
portion of the harness diverges from the vertical, the more irregular
will the shed be.

[Illustration: Fig. 48]

Suppose we take an extreme case of a loom having three or four
jacquards on it, the harness to be 96 in. wide, and the pattern to
be for a table-cover; then, if we deduct 2 ft. from the width of the
harness, it is possible some of the border twines may have to slant
across 6 ft., or 72 in. Now make a triangle as No. 1, Fig. 48. Let the
base, A B, be 72 in., and take it as the level of the cumber board.
Take the vertical side, A C, as the height of the harness to the heck,
viz., 84 in. Now calculate the length of the hypotenuse B C by squaring
84 and 72 and finding the square root of the sum, and it will be found
to be 110·63 in. Again: Make another triangle, as No. 2, with base 72
in.; vertical side 3-1/2 in. longer than that of the former triangles,
viz., 87·5 in. (this 3-1/2 in. is to represent the draw or lift of the
harness). Calculate the length of the hypotenuse as before, and it
will be found to be 113·31 in. From this deduct 110·63 in., the length
of B C in No. 1, and the remainder, 2·68 in., equals the height that
the cord B C has been raised, while the cord A C, which is vertical,
has been raised 3-1/2 in.; and if we take into consideration that
the side draw of the sloping cords, as B C, will pull the tail cord a
little to one side and rise the vertical cords a little higher, while
the sloping ones remain proportionately lower, 1 in. may be safely
taken as the difference of the height that the two cords, B C and A C,
would be raised by the jacquard, and all the other cords in the harness
would vary, being less than this in proportion to their divergence from
the straight or vertical line. It can thus be seen how the shed would
require to be opened to let the shuttle through, and the irregular
strain that would be on the yarn; and for any cloth that requires a
fine surface, any irregularity of strain on the warp has a deleterious
effect, very well known by experienced overlookers.

Some consider that the London style of harness is more suitable for
working without a heck than the Norwich style, and adopt it to avoid
using one, as it is severe on the harness twines. Some raise the
machines very high to avoid using them, but for particular work with
border ties they must be used to give a proper working harness. With
the London mounting rollers should be, and are, used when there is
no heck. These rollers are set as a coarse heck in a frame under the
machine, and lie lengthways under it, just as they would do when used
above a heck. The heck should be about 3 in. below the knots which
fasten the tail cords to the neck twines. Some have the tugs, or tail
cords, coming down through the heck; in this case the heck only takes
the strain off the hooks of the machine, and has no effect on the shed,
though sometimes this is mitigated by having more than one tug or tail
cord, and the neck twines that slant in different directions are tied
to different tail cords. The only point in favour of this is that it
saves some trouble in tying broken harness twines when they begin to
wear away by their friction on the heck.


PREPARING THE NECK OF THE HARNESS

The twine for the neck, as well as that for the other portions of
the harness, is usually wound on spools, and when the neck is to be
prepared it is warped from these spools to the length required, either
on a hand warping mill or round two pins fastened in a wall or on a
bench, as far apart as the length of the harness; three or four spools
are put on pins, and the ends from them are taken and wound round the
pins fixed for warping them on. When warped the twines may be cut at
one end, and can either be tied in a bunch or stretched on a board and
tied down on it, so that they may be kept straight and admit of any
number of them being pulled out as they are wanted.

In some methods of mounting the neck twines are first tied to the
tail cords; perhaps this is the plan most commonly adopted with
double-acting machines. Sometimes they are tied to the sleepers first,
and fastening them to the tail cords is the last process. This is
called ‘beeting’ the harness. When single-acting machines are made with
the hooks resting on a perforated board, they have tugs on the hooks;
but when the wire hooks pass down through a perforated grating, tugs
are unnecessary, and the neck twines are fastened to the hooks. When
fastened to the hooks, or even when fastened to the tail cords, and a
heck is not to be used, the neck twines should be formed into heads,
either before they are tied up or afterwards, as may be desired. The
reason of this is, that when a number of neck twines are tied to a
hook or tail cord, and the hook is raised by the machine when working,
the twines will be slanting in different directions, and when coming
down again would be liable to catch on the knots of those that were
not raised; and the head is for the purpose of keeping them together,
so that they cannot separate for a short way down. One method of doing
this is, after the bunch of neck twines is tied to a tail cord or hook,
to take one of the twines and knot it round the others about 3-1/2 in.
below the tail cord, and the same may be done with a second twine, if
there are many in the parcel.

Sometimes the neck twines are all prepared in heads, and the following
is a convenient method of doing so:--Fasten four flat pins of wood,
as A, B, C, D, Fig. 49, on a board as F, or on the edge of a bench or
table. C and D are firmly fastened, but A and B can be turned round to
the position shown by the dotted lines when the screws holding them are
slackened. The distance from C to D must be the length required for the
neck twines. E is the twine coming from a spool on a wire. The pin A
is pushed round to the position shown by the dotted line, and the twine
is warped round C, B, D, passing over B and under C, so as to form a
lease. When a sufficient quantity is warped, the crossing of the lease
is pushed up towards C, and the pin A turned round into the centre of
the twines, taking the place of B, which is turned back. The use of
the two pins is to allow room for warping, C and A being too close
together. The distance from the outside of C to that of A should be
the length the head is required to be (about 3-1/2 in. or 4 in.). The
twines may now be cut at D, and a piece of cord looped round them and
tied to D, so as to hold them steady, but allow them to be drawn out as
required. The number of twines for each head can now be drawn out and
tied round the two pins C and A, and when a number are done they can
be slipped off the pins and put on cords or rods; the lease keeps the
heads in order, and the bunch can be hung up at the loom, and each head
be taken in rotation by the mounter. The number of twines for each head
is regulated by the tie of the harness.

[Illustration: Fig. 49]

Suppose a double-acting machine is to be mounted with the harness
similar to that shown in No. 1, Fig. 46. The tail cords are usually put
on the machine by the makers, but if they are worn out, or if new cords
are required, putting them on is the first process. This can be done
when the machine is on the ground, and the neck may also be tied to the
tail cords before the machine is put on the loom. Some mounters prefer
one method, and some the other, and the one which most conveniently
suits the circumstances may be used. When the neck is attached to the
machine before putting it on the loom, it is usual to turn it on its
side on blocks or on a table. If the machine is on the loom, a board is
tied up to make a seat for the mounter, and the bunch of neck twines
is fixed convenient to his reach; or he may have a boy to draw them
out for him, and hand him the number he requires. Beginning at the
first hook, he takes a pair of tail cords, and having drawn them down
straight and stretched them firmly, he loops the neck twines on them in
the way shown in Fig. 46, and draws the knot fast. The distance from
the bottom of the hooks to the neck twines should be 11 in. or 12 in.
The first of these in each row is measured, and the mounter can then
regulate the others so as to have all the knots in a line.

The number of neck twines that are to be tied to each hook is regulated
by the tie of the harness. For each time that any repeat or portion
of a pattern is to be repeated on the cloth there must be a twine
attached to those hooks of the machine that are to work this repeat.
For instance, if a 400 machine is used, and the pattern consists of a
simple repeat on these 400 hooks, and this has to be repeated six times
on the cloth, then six neck twines must be tied to each hook of the
machine. Again: The pattern might be made for a 400 machine, 200 hooks
to be repeated six times, 100 four times, and 100 three times; then six
twines would be tied to the first 200 hooks, four to the second set of
hooks (100), and three to the third set (100), and any mounting would
be regulated in a similar manner.

Full particulars of ties will be given further on.

If all the hooks in the machine are not required, any number of rows
or portions of rows can be left idle at one end, or at both ends if
desired, or even at the back or front.

When all the neck twines are tied up, the next process is to draw them
through the heck, which should be fastened firmly 2 in. or 3 in. below
the knots connecting the tail cords and neck.

The jacquard is supposed to have been levelled and set in its proper
position on the loom, and firmly fastened there, and the same may now
be done with the cumber board.


CUMBER BOARDS AND HARNESS REEDS

These are both for the same purpose--viz., that of regulating the space
which the harness is to occupy--and both answer the purpose equally
well, generally speaking. The reed being stronger is, perhaps, more
suitable for coarse work; and the cumber board, giving a more evenly
distributed harness, is perhaps preferable for fine work. Be that as
it may, both suit for any medium work, though some workmen are all
against the reeds, and others all against the boards. Harness reeds
are strong-made reeds of cane or iron, of the depth and fineness to
suit the harness. Bridges are fixed in them about 6 in. apart, with
perforations for stiff wire to be run through; as many wires as are
required to suit the number of rows of the harness. A wire should
be outside the harness at both sides, to prevent it rubbing against
the ribs of the reed. The wires are made straight, drawn tight, and
fastened at each end to holes in the yoke of the reed. The reed is set
in a frame of wood or iron, so that it can be bolted to brackets on the
loom.

Cumber or comber boards, also called hole boards, are made in various
ways. Sometimes they are of wood about 3/4 in. thick, and bored in a
piece. These boards are strong enough to bolt to brackets on the frame
of the loom. Sometimes they are thin, about 1/4 in. thick, bored in the
same way, and framed. The wood used is beech, sycamore, and sometimes
walnut. The objection to these boards, particularly the thin ones, is
that if the wood is not very well seasoned they are liable to warp and
split. A great many prefer to use what are called ‘slips’--that is,
small pieces of wood of the depth required for the harness, and about
1-1/2 in. long or broad. The length, or long way of a cumber board
or harness reed, is frequently called the width or breadth, same as
weaver’s reeds, as this is the width of the loom. The cross-way, or
from back to front, would, in the same way, be the depth. These slips
are made of boxwood, beech, or other clean hard wood. Some prefer beech
or sycamore to boxwood. The twines cut into them all in time, but the
hardness of the boxwood causes a very fine cut to be made in it, which
cuts the twine. The others will not cut the twine so readily, and by
the time the wood is cut too much the harness would require to be
renewed as well as the cumber board. These slips are made about 1/4
in. or 5/16 in. thick, and are set in a grooved frame. The usual way
of arranging the holes in a board is shown at A (Fig. 50), which is
for an 8-row harness. B shows a patent method, devised for the purpose
of giving more space between any two holes in each horizontal row. The
arrangement of the holes is in 4-shaft satin order, and gives double
the space between the holes of each horizontal line that A does, and
of course it would take so much longer for the twines to cut through;
but this would only be an advantage if the holes were set very close
together, as otherwise the harness would be considerably off the level
before the twines had cut from one hole into another. In any case,
the irregularity of the holes will likely prevent it from ever taking
the place of the older method shown at A, though it may be desirable
sometimes.

[Illustration: Fig. 50]

The cumber board is fastened on two brackets--one at each side of the
loom--and should be perfectly level and firm. It should be so far
advanced as to allow the harness to pass within half an inch of the
handrail of the lay when full back, and should be from 8 in. to 10 in.
above the mails, the lower the better, as it keeps them steady; but
the height may have to be regulated by the swords of the lay, as they
must be clear of it, and in some looms they are higher than in others.
When the cumber board is levelled and fastened, it should not be moved
after the harness is tied up, as any change in its position must alter
the level of some portion of the harness. This cannot in every case be
adhered to, as in some cases it may be necessary to move it. In a small
harness a good deal of shifting can be made that could not be attempted
in a large or intricate one. In many places the harnesses are built in
a separate room, or by the machine maker, and sent to the manufacturer;
but practical experience with particular work will teach anyone that
it is most desirable to have everything about the harness as level and
true as possible, and it takes a good deal of care to effect this, even
without any shifting.


SLABSTOCKS AND LEVELLING FRAMES

‘Slabstock’ is a name given to the board used for fastening the mails
on before they are levelled. It is about 6 in. deep and 1 in. thick,
and should be as long, or longer, than the loom is broad, according to
the way it is to be fixed in the loom, which is usually by being bolted
to brackets fastened on the sides of it.

[Illustration: Fig. 51]

There is a groove or rebate in the top edge of it for holding the
mails, as shown in Fig. 51 at A and B. Sometimes the grooves are at
both sides, as shown at A^1. All the mails required for the harness are
put on the one slabstock, which is set directly under the cumber board
when the harness is being levelled.

[Illustration: Fig. 52]

A levelling frame is better than a slabstock; it is for the same
purpose--that of keeping the mails firm and level when being tied to
the neck twines. It consists of two flat bars of iron, C, C (Fig.
52), which can be bolted together in the form of a frame by the use
of two stays, one at each end, and by the same screws be bolted
to slotted brackets, A A, B B, fastened to the frame of the loom.
Sometimes two hanging brackets, as D, are used, fastened from the top
or heddle-bearer of the loom. The levelling bars, C, C, rest in these
brackets, and can be fixed firm in them with a pin or cotter. If there
is any yield or spring in the brackets, they can be stayed from the
front of the loom as well, as the frame should be made very firm. The
distance between the two bars should be 4 in. or 4-1/2 in., which will
suit any harness; if only narrow harnesses, as 8-row, are required, it
may be an inch narrower. The length of the frame must suit the breadth
of the loom. It should be perfectly straight and level on the upper
edges, and should have two rows of holes bored in it, about 2 in.
apart, for pins and skewers to pass through above and below the wires
on which the mails are strung, so as to keep them firm.


LEVELLING THE SLABSTOCK OR FRAME

Levelling a harness, or rather levelling the mails of the harness,
is the most important part of the mounting, as, if the mails are not
levelled as true as a straight-edge, good work need not be expected
from the loom. The position in which they are levelled, in relation to
the breast beam and back rail, will depend upon the nature of the cloth
to be made, to a certain extent; but as a general rule, for ordinary
work, the eyes of the mails should be half the depth of the shed
required to be drawn below the level of the breast beam for hand loom
work, and a little more than that for power-loom work. This is to have
the top and bottom portions of the shed of an even tension, and to make
both sides of the cloth equally good. The position of the back roll can
be altered a little afterwards, if it is required to slacken either
the top or bottom portions of the shed. Raising the back roll throws
up the weft pattern to the under side of the cloth, and sinking the
back roll gives the upper surface of the cloth a finer appearance, by
tightening the top portion of the shed. A 3-1/2-in. draw of the harness
is sufficient for power-loom work, and perhaps less would sometimes be
preferable. Levelling the harness mails 1-3/4 in. for light work, and
2 in. for heavy work, below the breast beam should be a good average
standard for a 3-1/2 in. draft of harness. The levelling frame or
slabstock is levelled to this height by means of a spirit level and
straight-edge, making allowance for whatever spring there may be in the
harness after it is tied, and the wires drawn out of the mails. This
can only be ascertained by experience, and depends principally on the
tension the mounter puts on the twines when tying them; it will also
depend upon whether the harness twines have any spring in them or not,
but they should be well stretched before being used, to avoid this. A
mounter who ties slackly will generally make a more level harness than
one who ties tightly. About 1/8 in. may be allowed for the mails to
rise if the cords are slackly tied, and if tightly tied 1/4 in. may
have to be allowed, and sometimes more than this.

Sometimes the level is taken by placing a straight-edge on the race
of the lay, when the cranks are at the back centre; then the under
edge of the straight-edge should give the position for the tops of the
mails on the slabstock, when it is fixed in its position in the loom.
In levelling by this method the race must first be set at the proper
height, and it must also be properly bevelled. From 1/2 in. to 3/4 in.
below the breast beam is a good level at which to have the race, when
the cranks are at the top centre, the latter for a low harness. The
angle, or bevel, of the race may not at all times suit for levelling in
this way. If not, the race should be made correct when possible; if it
cannot be corrected, the harness must be levelled a little lower and
the back rail kept down for working. A loom with the race bevelled for
plain work is not suited for damask; the angle between the race and
reed should, for damask, be about 93°, whereas for plain 87° would be
more suitable.


TYING UP THE HARNESS

Having the neck attached to the tail cords and the levelling frame
and cumber board fixed in position, the next operation is to get the
prepared lingoes placed in the loom, whether on the levelling frame or
slabstock. Suppose the levelling frame to be used. According to the
number of rows in the harness or cumber board, pieces of reed wire are
procured, one for each row, and as many lingoes are strung on each of
these as will be required for each row of the harness, by running the
wires through the mails. These wires are then placed in the frame and
secured by skewers being put through the frame above and below them.
The frame is then levelled and firmly bolted to the loom. The sleepers
or mid-pieces of the lingoes are next drawn up through the cumber
board, one through each hole required to be used, any surplus holes
being previously marked out to suit the tie of the harness, which will
be afterwards explained. When all are drawn through, the tying of the
neck twines to the sleepers may be proceeded with.

This must be done in accordance with the tie of the harness. Thus, if
there are four cords tied on the first hook of the machine, these must
be taken down to the lingoes in the proper place in the cumber board,
which may be seen by examining the particulars of the mountings, to be
given further on. One method of fastening the neck to the sleepers is
shown in No. 1, Fig. 46, at D, _d_, _d_^1. The twine is put through
the loop of the sleeper at _d_^1 (or better to have a snitch on the
sleeper, unless the knots are to be varnished afterwards), then a loop
is thrown on at _d_, and the end tied at D, and clipped close. This
method suits very well when the harness is liable to be altered, for if
the knots are brushed over with paste they hold sufficiently firm, and
can be loosed again at any time, if necessary. If varnished, it also
answers equally well for any coarse harness; but if it is fine and much
crossed, especially if the twines are strong or coarse, the knots are
liable to catch on each other and raise more warp than should be when
the loom is working.

A second plan is shown in No. 2, Fig. 46. In this case there are
no sleepers on the lingoes; they are put on the levelling frame or
slabstock without them, the mails and lingoes being connected by the
hangers; the levelling frame is set in the loom as before, and the neck
twines must be long enough to go down through the cumber board, through
the top hole of the mail, and up again through the cumber board to
where they are tied. The mounter, when about to tie them, takes one of
the twines and casts a single knot on it, leaving the loop open, then
puts the end of it through a hole in the cumber board and through a
mail, and with a small wire hook draws it up again through the same
hole in the cumber board, and through the open knot or loop he made on
it, at the same time; then, turning round the awl or piercer in the
other end of the handle of the hook, he puts it through the knot and
runs it up about 6 in. above the cumber board, draws it tight, and
casts another knot above it with the end of the twine, as shown at D,
No. 2, Fig. 46. This makes a neat harness, and when a mounter gets
accustomed to it he can proceed very expeditiously. This method is used
in England; the former is Scotch. The instrument used for drawing the
twines through the cumber board consists of a wooden handle, in one
end of which is a hook or barbed wire, same as is used for drawing the
warp through mails, and in the other end is a round awl or piece of
steel wire, tapered to a blunt point, which is used for running up the
loop or knots so as to have them all about the same distance above the
cumber board.

_Beeting_ is another Scotch method of mounting. The harness may be
beeted either above or below the cumber board; beeting above it was the
old method, and single slabstock was used, as shown in Fig. 51 at A.
When preparing it for the loom, the lingoes and mails are connected by
the hangers in the usual way, and hung on the slabstock with the mails
in the groove in it, as shown at B, Fig. 51. A piece of flat wire (reed
wire) is run through the eyes of the mails, as many as are required for
the whole harness, and is then tied down by cords fastened round it
and the slabstock, at short distances apart. The slabstock is now put
into a frame, or rack, with a rail as high above it as is required for
the length of the sleepers. A spool of twine is fixed on a wire pin at
the side of the frame, and with a needle, or otherwise, the end of the
twine is run through a number of the mails and fastened. With a hook
the sleepers can be reeled up to pins in the rail above the slabstock,
on the same principle that they are done in Fig. 47. When the sleepers
are all finished they can be slipped off the pins and cut.

The slabstock is next fixed in the loom and levelled with the upper
edge of the rebate or groove touching the under edge of a straight-edge
placed on the race of the lay when it is full back (for power looms).

The sleepers are then drawn up through the cumber board without their
ends being knotted; they should be long enough to reach about 8 in.
above it, and say 7 in. below, making 15 in. for their entire length.

Now, to beet the harness: Say there are two beeters, standing on the
ground, with a supply of neck twines convenient to them; they pick up
the sleepers from the first set of holes in the cumber board--that is,
those that are to be connected with the first hook in the machine--and,
having tied neck twines to them, hand them to the harness tyer, who is
up at the machine. He takes the lot of twines and draws them all to an
even tension, then, having drawn down the tail cords firmly and evenly,
knots the neck twines to them.

One method of knotting the neck to the tail is to have the tail cords
tied together so as to form loops; then, having straightened both the
neck and tail cords, lay the neck against the loop of the tail, and
take both firmly between the finger and thumb of the left hand, being
careful not to let them slip, and casting a knot, as at C or D, Fig.
51, round the tail, with the ends of the neck draw it fast. In order
to have all the tail cords plumb and the knots of a uniform height,
the harness tyer sometimes has a cord tied across the top of the loom
frame, at each end of the machine, at the height the knots are to be.
On these he lays a straight-edge, marked as a guide for where the tail
cords are to hang, so that he can regulate those of one row and have
them plumb, and the knots even; when one row is finished, he moves
the straight-edge to the next, and goes on with it. A newer method
of beating is to prepare the lingoes with sleepers and hangers, the
sleepers to be about 5-1/2 in. long. The slabstock A^1, Fig. 51, is
used, or the other if preferred.

The mails are put upon flat wires as before, but instead of all going
on one wire, half the number is put on each of two wires, and one of
them is placed at each side of the slabstock, where it can be fastened
with small staples. To facilitate getting the mails on the wires, when
reeling the sleepers on the pin, as in Fig. 47, a lease can be made on
the loops by giving them a twist when putting them over the pin.

A piece of twine can be fastened in this case, and the loops
afterwards cut. The mails can then be taken off in order, and run on
the wire for the slabstock. When the mails are fixed on the slabstock
it may be laid on the ground, or set in a frame, and the neck twines
tied to the sleepers; and when all are tied, they can be drawn through
the cumber board or harness reed. Afterwards, all are taken and fixed
in the loom. The cumber board must be set so high above the knots on
the sleepers that they will not come into contact with it when the shed
is opened (that would be 4 in. or 4-1/2 in. above it). The neck twines
are then tied to the tail cords as before.

This makes a very good harness, and has the advantage of having
no knots on the neck twines above the cumber board, which is very
important in an intricate harness with the cords close together, as the
knots when varnished are liable to catch on each other, or on twines
slanting across them, and lift them as they are being drawn up. When a
harness is mounted in this way it is not easy to make any alteration
on it, or to re-level any portion of it if necessary. If the sleepers
are tied in loops, same as used in No 1, Fig. 46, they can be connected
to the neck twines by having the latter double and putting both ends
of the neck through a snitch on the sleeper, or the neck twines may be
double and the two ends of the sleeper when untied may be put through a
snitch on the looped end of the neck twine and tied; in this way they
could be altered or adjusted afterwards if necessary, but if they had
to be varnished the knots would be rather rough.

The following method of mounting is adopted in the damask hand-loom
districts of the North of Ireland, and is used for particular
power-loom work as well. It is a slower process than the preceding, but
cannot be surpassed for getting a level harness, and the mails can be
regulated as desired--that is, to have the back rows getting gradually
a little higher than the front ones, which can only be accomplished in
the preceding methods by tying them a little tighter, or by having the
frame sloped a little.

The lingoes may be prepared as before, with the sleepers 5 in. long
when tied and clipped. The neck twines are put down through the cumber
board and knotted loosely in bunches underneath.

The levelling frame is set in the loom, the top edge of it being at the
level that the eyes of the mails in the front row are to be hung.

No wires are required. The lingoes are taken in bunches and put astride
on the frame as required, and boys fasten them up to the neck twines
by throwing on a snitch and running them up to somewhere about the
height they will be wanted when level; in doing so the knots on the
sleepers must be kept up as close to the snitch as is convenient for
tying them, as, if left too low, they would interfere with the warp in
shedding. When all the lingoes are hung inside the levelling frame--or
they may be hung first, and the levelling frame put up afterwards and
levelled--the mounter may begin to level the mails.

He uses a fine piece of waxed cord with a small weight at each end,
which he lays across the levelling frame as a guide, and levels each
row from back to front in succession. The front mail may hang with
the top of its eye level with the cord, and the others rise a little
higher, till the back one is perhaps, with the bottom hole, level with
the levelling cord, or 1/8 in. higher than the front one. The levelling
can easily be accomplished by sliding the snitch up or down the neck
twines; and when the correct height for the mail is got, the neck twine
is tied as at H, No. 3, Fig. 46, when it is to be varnished, or as at
H^1 when it need not be varnished; in the latter case cable cord is
used for the neck, and it is split at the end, when drawn through the
snitch, and then knotted.


VARNISHING A HARNESS

Varnishing is for the twofold purpose of making the harness twines wear
better, and keeping them from being affected by the atmosphere.

Sometimes the harness is only partly varnished, particularly when it is
liable to be changed in a short time, and then it can be loosed down
and used again.

If it is to remain for a length of time--that is, for steady work--it
should be varnished all over, but care should be taken to get a good
varnish, as some of them destroy the twines, and others come off and
are useless.

The common varnishes are made principally from shellac, beeswax,
and turpentine, and can be procured at any heddle-maker’s. Sometimes
white wax dissolved in turpentine is used to rub the twine at the
cumber board, for fine harnesses for silk and other light work, the
remainder being left unvarnished. Sometimes the neck of the harness and
the sleepers are varnished, and the knots above the cumber board are
brushed with paste, as each row is tied, to keep them from slipping.

Perhaps the best varnish is boiled oil, which, when well dried, gets
very smooth after working for some time, and keeps the twines soft and
pliable. It takes some time to dry, and does not suit well for using
in a dusty place, but answers very well when the harness is built in a
room for the purpose, and is afterwards taken to the loom.

Sometimes a very small quantity of beeswax or white wax is added to the
oil to give it more firmness, and sometimes driers are used to make it
dry more quickly; but it is better to do without driers, as they harden
the twines and are injurious. Varnish should not be disturbed till
quite dry, as, if the twines are separated and the loom started when it
is soft, the outer surface will rub up and make a rough harness. The
twines are separated by running a wire skewer between each cross-row.
French chalk dusted down through the harness assists in smoothing it,
and prevents too much friction when it is being started to work.


TO ARRANGE THE TIE OF A HARNESS

What is known as the tie of a harness is the arrangement or manner
in which the harness twines are connected to the hooks of the
jacquard--_i.e._ the number of twines that are tied to each hook, and
the position they are to occupy in the cumber board. Ties for ordinary
damask work may be divided into three classes--viz. straight, or
single; lay over, or repeating; and gathered, or centred.

_Straight or Single Tie._--This tie is only required when there is
no repetition of any portion of the pattern, as in silk markers, and
occasionally in handkerchiefs, d’oyleys, tablecloths, &c. Only one
harness twine is tied to each hook of the jacquard, and there must be
as many hooks as there are threads of warp. The twines are tied to the
jacquard and taken down through the cumber board in regular order from
first to last, and the warp is drawn into the mails in the same order.
Any pattern that it is possible to put on the cloth can be wrought with
this method of mounting.

_Lay Over or Repeating Tie._--This is, perhaps, the commonest tie, and
is used for all such patterns as Fig. 53, where one small portion, as
that enclosed by the dotted lines, will, when repeated several times,
cover the entire surface of the cloth.

[Illustration: Fig. 53]

_Gathered or Centred Tie._--This tie can be used when both halves of
any figure or border, &c., are alike when turned over; it will only
require half the number of hooks to work a figure or border, when
gathered or centred, that would be required for a single pattern.

All other ties are combinations of these.

Fig. 53 shows a pattern for a lay over, or repeating tie, such as might
be used for dress goods, handkerchiefs, &c. The complete extent of the
pattern is enclosed by the dotted line; this is called one ‘repeat’
of the pattern, because this portion repeated over the cloth makes
up the entire figuring on it. To ascertain how this may be woven, or
what number of hooks of the jacquard will be required to weave it, the
number of threads contained in 1 in., or other given space, of the
cloth required must be known. If the pattern is to be on the cloth say
2 in. × 2-3/8 in., and there are to be 100 threads of warp and 120
threads of weft per inch, then 200 hooks of the jacquard and 240 cards
would be required to weave it, each hook representing one thread of
warp, and each card representing one shot of weft.

[Illustration: Fig. 54]

Figs. 54 and 55 show mountings on jacquards with 208 hooks in each,
the former tied on the Norwich principle, and the latter in the London
style. In both cases only 200 hooks are used, the remaining eight
hooks being left idle for selvages, if required. These mountings are
made for four repeats of the pattern, which would only make 8 in. of
cloth; but there may be any number of repeats, say 13, to make 27 in.
of cloth or 28-1/2 in. of warp in the loom, with, say, 92 threads per
inch. There must be a harness twine for each repeat of the pattern tied
to each hook of the machine; when four repeats are used, four twines
are tied to each hook, as in Figs. 54 and 55. The harness reed or
cumber board, G, must be of the same set or fineness as the weaver’s
reed (if finer, it can be used by passing over the surplus holes). It
is divided into four portions, leaving 200 holes for each repeat, and
the twines are, in Fig. 54, taken straight down from the hooks to the
holes in it, as can easily be seen. In this figure the cross-rows of
the cumber board at the first of each repeat are shown filled; but in
mounting, the back row, taken from the hooks A to B, would be filled
first. In Fig. 55 the reverse is the case, the cross-rows, as shown,
being the first filled; this is owing to the twist in the harness, as
the machine is sitting so that the cards will hang over the side of
the loom. The letters A, B, C, and D in both figures denote the same
corners of the machines, showing that in Fig. 55 there is a quarter
twist in the harness. In Fig. 54, the dotted lines from the 200th hook
show the last cord of each repeat; L is the first cord of the last
cross-row, and E and F are the same in Fig. 55. In these figures only
the skeleton of the mounting is given; it must, of course, be filled up
as the first rows which are given.

[Illustration: Fig. 55]

For this description of mounting the yarn is drawn into the mails in
regular order, beginning at No. 1, and proceeding with the numbers
as given. In this case the back hook to the left-hand corner must be
taken as the first hook of the machine, and the cards cut to suit this.
Sometimes the first hook to the front left-hand corner is considered
the first hook, but if so it must be remembered when cutting the cards.
Sometimes the mounting is begun at the right-hand side, but in any case
it is only necessary to see that the draft of the yarn and cutting of
the cards correspond with the order of mounting; the result of the work
should be the same in all cases.

[Illustration: Fig. 56]

Fig. 56 shows a pattern for a gathered border with a repeating centre,
which is a very common style of design. It will be seen that the two
halves of the border, A and B, are alike, if taken from the centre
outwards; or if one-half of the border were traced on tracing paper, it
would, if turned over, form the other half. For this reason it is also
called a ‘turn-over’ border. In the same way, if two cords are tied
to each hook of the jacquard for the border, and those to the first
hook taken down to the cumber board for the two outside threads of the
border (that is, one to the right and the other to the left), then
those from the next hook taken to the two next outer holes, and so on,
coming from the outsides to the centre, the first half of the pattern
would be repeated by the mounting in exactly the same way as by turning
over the tracing paper.

[Illustration: Fig. 57]

Fig. 57 shows a mounting for a gathered border and four repeats of
the centre. Eighty hooks are taken for the border, which are repeated
four times in the cumber board, as at 80_a_, 80_b_, 80_c_, and 80_d_,
and 120 hooks are allowed for the centre repeat (see also Fig. 58).
This mounting is made up for a 200 machine with 208 hooks, the extra
eight hooks being left for working the selvages. Only the front row of
the harness is shown, but they are all alike. By following the cords
from the hooks to the cumber board, it can be easily seen how they are
taken through it. The first eight hooks are left idle, for the selvage
to be fixed to if required. The next ten rows = 80 hooks, are for the
border; these 80 hooks, with one twine on each, taken down through the
cumber board at 80_a_, would work the first half of the border, as A.
The other half of the border, B, can be wrought by the same hooks, with
another set of twines tied to them, and taken through the cumber board
at 80_b_; but the twine from the first hook must be crossed over to
the right-hand side of the border, being the 160th twine in the cumber
board; the twine from the second hook is the 159th in the cumber
board, and so on, coming in towards the centre to meet the first set of
twines; hence the name, ‘centred tie.’

[Illustration: Fig. 58]

The border must also be repeated on the other side, and two more sets
of twines must be tied to the same hooks (four to each hook in all),
which are taken down through the cumber board at 80_c_ and 80_d_, to
work the two halves of the border marked G and H, Fig. 58. The next 15
rows of hooks = 120, work the repeat of the centre pattern, C, Figs.
56 and 58; and as this is repeated four times, at C, D, E, and F, Fig.
58, there must be four twines tied to each of the 120 hooks, which are
taken down through the four divisions of the cumber board marked 120;
one twine from each hook being taken through each division, beginning
at the left and working to the right side, alike in them all, as this
portion is a simple repeating tie. Fig. 58 gives a portion of the
pattern in Fig. 56 completed across, as it would be on the cloth by the
mounting in Fig. 57; the cumber board is laid across the top of the
pattern and marked, showing, in conjunction with the mounting, Fig.
57, how the harness repeats the pattern so that a large surface can be
figured with a small machine by adopting a suitable tie.

When drawing the warp into the harness, wherever there is a gather in
the tie, or a turn in the harness, there must be a turn in the draft
also; not that there is any real change in the draft, but when the
harness is turned in the direction it is drawn through the cumber
board, the draft must also be changed to follow the mails in regular
order. Fig. 59 shows a draft for a gathered border, with 48 hooks for
the border and three repeats of the centre. It will be seen that the
first half of both borders A and B and the three repeats of the centre
are drawn in the same direction, but that the two turned-over portions
of the borders C and D are drawn in the reverse direction. The numbers
on Figs. 62 and 62A may assist in explaining this, the border A in 62
and B in 62A being taken; in these it will be seen from the numbers,
which are those of the harness twines, that 1 to 8 in border A (Fig.
62) run in the reverse direction of 1 to 8 in border A (62A), and both
are from the same hooks.

[Illustration: Fig. 59]

[Illustration: Fig. 60]

When two borders are to be woven on any piece of cloth, though the
pattern for each border may be single in itself, as Fig. 60 (which is
single except the two outlying lines, A and C), they will to each other
form a gathered tie, provided they have to be turned in the same way
to the centre of the cloth--that is, if the inside of one border on
one side is to be the inside of the other border on the other side. In
such a pattern as Fig. 60 it would not matter much if the inside of
one border was the outside of the other, as the border is much alike
either way, and if wrought in this way the twill on the cloth would not
be reversed, as is always the case in a turned-over border. In many
patterns this would not do, and it is therefore necessary to gather
the tie and turn the twill. Fig. 61 shows how, in the London style of
tying, the harness can be turned so that a turn in the draft is not
necessary, but this will not alter the reverse twill on the cloth.
It will be seen that one of each of the two cords tied to the eight
hooks goes to one border, and the other to the opposite border, but
instead of both going to the same row of the cumber board, the cord
from the first hook goes to the back row in one border, and to the
front row in the other. All the cords follow in this way, which has
the same effect as turning the draft, as would require to be done in
the Norwich mountings, and could also be done on this principle. Fig.
62 shows the draft of the two borders for Fig. 61 mounting, A and B
being a plan of the warp passing through the mails. The numbers refer
to the hooks of the jacquard. C and D are the two innermost threads of
the borders, both on the 48th hook of the machine. Fig. 62A shows how
the draft would be if the cords were not brought from front to back of
the cumber board, as the numbers will show. No. 1 shows where the cords
from the first hook pass through the cumber board. Fig. 62A gives the
right-hand border only, the left-hand one remaining the same as in Fig.
62.

[Illustration: Fig. 61]

[Illustration: Fig. 62]

[Illustration: Fig. 62A]

In Fig. 60, the two bars A and C could be wrought from the same hooks,
and, unless they might be required to be different from each other on
another pattern to work on the same loom, it would be unnecessary to
allow machinery for the two; either the outer or inner one could be
tied up, and the other repeated from it, but the one that is tied up
must be painted on the design paper.

[Illustration: Fig. 63]

Fig. 63 shows the front view of a mounting for weaving a pattern such
as that given on Fig. 60, one machine being used for the border, and
another for the centre. Usually the border machine would be to the
left-hand side, to correspond with the pattern; but in Fig. 63 it will
be observed that No. 1 machine is to the right-hand side, and that it
works the borders. It does not matter which machine is used for border
or centre; the mounter can use the one he considers most suitable for
the borders, and the pattern may be painted and the cards cut without
any consideration of this, it being only necessary to put the border
cards to the border machine, and those for the centre to the centre
machine. The only consideration is to avoid any unnecessary slanting
of the harness twines, which the present arrangement does; and it
also keeps the turned portion of the harness (that for the right-hand
border) separate from the twines of the centre, thus avoiding any
unnecessary friction.


TO MARK OUT A CUMBER BOARD

[Illustration: Fig. 64]

Cumber boards may be either the exact fineness of the harness, or any
set finer; they are better to be a little finer, to allow for any
broken rows at the beginning or end of any of the repeats. When the
cumber board is finer than the harness, the surplus rows are left idle.
A convenient plan of marking off a cumber board is as follows:--Let
the pattern be as Fig. 60--that is, with a border at each side and the
centre to be repeated several times to make up the width of the cloth.
Let 24 cords be required for the selvage, 172 cords for the border, and
128 for a repeat of the centre; there are to be four repeats of the
centre. Now mark off on the cumber board whatever width the harness
is to stand, which should be the same or rather wider than the warp
occupies in the reed, and let it be so as to have the harness in the
centre of the loom. Suppose Fig. 64 to give the width of cumber board
required. The selvages are wrought from 6 hooks, and there are 8 hooks
in each row of a 400 jacquard; therefore 4 rows of the cumber board
will be required for each selvage. Mark these off by drawing the lines
A and B on the cumber board. Now find out the width the borders are
each to occupy, by calculating how wide the yarn will be in the reed,
and mark them off by the lines C and G; then mark off the width of each
repeat by the lines D, E, and F. Counting the number of holes marked
off for the margin, it will be found that there are 32; but as the
selvage is on six hooks, the two back holes of each row of the cumber
board are marked out, as shown, not being required. The portion marked
off for the borders contains 23 rows on each side, whereas only 21-1/2
are required for 172 cords. As the centre repeat is on even rows, and
it is usual to leave broken rows to the left-hand side of the painting,
the first 4 hooks of the border portion of the machine will be idle;
therefore the 4 back holes of the cumber board for both borders are
marked out; this still leaves a row too much, which may be marked out
at any place. For the repeats, 128 cords, 16 rows are required, and the
surplus rows are marked out, as shown. This mounting is supposed to
be on a 400 machine, leaving 100 hooks idle to the left-hand side of
the mounting, then using 1 row for the selvages, and the remaining 38
rows for the pattern, all but the half of the first row, which is not
required.

[Illustration: Fig. 65]

In marking out a cumber board for a lay-over, or repeating pattern--if,
say, 108 hooks are required for each repeat, which makes 13-1/2 rows
of an 8-row machine--if the broken row is to the left-hand side of the
painting, and the harness mounted from left to right, then the first
four holes of the first row of each repeat of the pattern are marked
out in the cumber board to be left empty, as shown in Fig. 65. When
mounting, no attention would be paid to the cords on the half-row of
the machine till the first four rows along the cumber board have been
filled, the mounter beginning with the back cord on the first full row
of the machine, which is the fifth hook of the tie. Three rows for the
selvage, which is on four hooks at the front of the harness, are shown
at S, Fig. 65.

The front of a jacquard is generally considered by workmen to be the
cylinder side. Of course, when there is a cylinder on each side there
is neither front nor back to it. Sometimes it is very confusing talking
of the front and back, one considering it to mean the front of the
harness, or front of the loom, and another taking it to be the front
or cylinder side of the jacquard, which is usually at the back of the
loom for Norwich ties. Of course, the front of the loom is where the
weaver stands, or where the cloth is made, and the front of the harness
is towards the front of the loom. It simplifies matters greatly by not
minding the back or front of the machine, but speaking of every part of
the harness, &c., towards the front of the loom as the front of it.

[Illustration: Fig. 66]

It frequently happens that figured stripes are required, running up
plain or fancy textured goods, either forming borders to the outer
edges, or merely ornamental stripes, say 3 or 4 inches from the edges
of the cloth, as in towels, toilet-covers, &c. For this class of
work it is generally most advisable to work the stripes with a small
jacquard, and the plain or fancy texture of the body of the cloth with
a shaft mounting. Fig. 66 shows a mounting of this class; the shafts
can be wrought by tappets in the usual way, and small cumber boards,
as A, A^1, are fastened to the top rail of the loom for the harness to
pass through. It is usual to keep the harness behind the heddles, and
it is levelled in the usual way. The yarn may or may not be on the same
beam for both portions, according to whether the take-up, or shrinkage,
will be the same for both portions or not. Sometimes the shrinkage may
be regulated to be alike in both by using different counts of yarn, or
drawing it thicker into the reed; but perhaps in most cases two beams
would be desirable, unless the texture is much alike for both the
stripes and body of the cloth.

The lease rods for the yarn in the harness will not, probably, do to
be the same as those for the body yarn, as the action of the heddles
will be different from that of the harness. If the same rods will not
suit, it is easy to have a pair for each warp, those for the harness
being farthest back. Stripes of this description are frequently made
in towels, glass-cloths, &c., for clubs or hotels, with the name of
the company woven into them. A gathered tie is mostly used for these
mountings, except when letters are required to be woven, in which case
a gathered tie is not suitable. (See Letters and Monograms.)


TO VARY THE SET OF THE HARNESS

It may often be that in weaving figured piece-goods it would be an
advantage to have a harness that would weave different degrees of
fineness, so that the loom could be utilised for one when the other
is not required. For weaving small quantities, for samples or special
orders, this is often an advantage.

The usual method of procedure is to draw the warp into a finer or
coarser reed if only a slight difference is required, and to pay no
attention to the strive in the warp between the harness and reed. The
better the quality of the warp, the greater the strive may be, but from
1 in. to 2 in. at each side is as far as it can safely be run when the
harness is wider than the yarn in the reed; and if the reed is coarser
than the harness, 1 in. at each side will probably be the maximum limit.

When changes of this kind are liable to take place, the cumber board
should not be too low down, nor should the mails hang too close to the
back of the slay; with the cumber board high and the harness a little
back the strive will not tell so severely on the warp. When the cumber
board is made of boxwood slips, they can be spread out a little in the
frame, but when there is much of a slant in the harness this would take
the mails off the level; though this can sometimes be counteracted a
little by raising the cumber board slightly higher at one end than at
the other.

The best plan when much of a change is required is to pass over the
surplus rows of mails in the harness, as in harness work, as well as in
shaft mountings, the mounting may be used for any coarser set of warp
than it is built for--of course, if the width is suitable. This would
necessitate the warp, or at least a part of it, being drawn out of the
harness, and also requires a new set of cards for the new pattern, or
for the same pattern on a new set of cloth. The surplus mails may be
cast out in rows across the loom if many are to be rejected, or in
rows across the harness (from back to front) if found more suitable,
or when a small number of mails are to be left idle. Suppose for an
8-row harness, 400 machine, with 6 repeats = 2400 mails on 30 in., or
80 threads per inch, and it is required to weave a piece of cloth on
this having 74 threads per inch. Drawing the yarn into a coarser reed,
and, if only 30 in. wide is required, throwing off the surplus yarn
at each side, would probably be the simplest method; but if the cloth
to be woven is to have 60 threads per inch, every four throw of mails
across the harness, or one-fourth of the harness, might be left idle,
or the two back rows of the harness may be left empty. In the latter
case the pattern could be painted on 6-row design paper as if for a
6-row machine, and in the former case the number of designs across the
painting would be reduced by one-fourth, the card-cutter taking care
that he omitted those rows on the card that have been thrown idle on
the machine.

In order to reduce this trouble and expense to a minimum Messrs. Devoge
& Co. have patented an expanding harness which is intended to admit of
cloths of different degrees of fineness being woven without any trouble
but a little setting of the cumber board and re-reeding the warp. The
method of doing so is as follows:--There are two cumber-board frames,
one above the other, as shown in the sketch Fig. 67; these frames are
filled with slips which can be moved along when desired. The harness
can be set for two degrees of fineness of warp, as is shown in the
sketch. The slips in the upper frame, B, are set midway above the two
positions for a slip in the lower frame indicated by the letters _a_
and _a^1_. The black lines show the position of the cords when weaving
the finer set. The only change required to be made is to shift the
lower slip, _a_, to the position _a^1_, and to re-reed the warp; other
changes can be made on the same principle, or by altering both sets
of slips in relation to each other. When the slips are moved in the
frames, they can be kept in position by interposing blank slips between
them.

[Illustration: Fig. 67]

The same cards will suit in this case for several sets of cloth, if
required. Of course it is plain that the friction on the harness twines
working through the two cumber boards must be greatly increased from
that of the ordinary method of working; but for light work, and when
the harness is not required to last long, it may prove very serviceable
when frequent changes are required, as it furnishes a ready means of
accomplishing the desired object, which cannot be done without cost in
some direction. For wide looms or intricate mountings it will not be
of much service. In the Figure only one hook of the jacquard is given,
for simplicity, but the harness would be filled up in the usual manner.
If the harness is built with long mails having several eyes in them,
the second cumber board is not required; the slips can be shifted and
the warp levelled by drawing it through suitable holes in the mails.




CHAPTER IV

_DESIGNING AND DRAUGHTING_


Designing is the composing, drawing, and colouring, if necessary, of
the sketches for jacquard patterns, and in some cases planning the
texture of the cloth; and draughting (or drafting) is the term usually
applied to the painting of the patterns upon design paper.

A designer should be a first-class draughtsman, and have a thorough
knowledge of the various styles of ornament; he should also be well
practised in drawing floral forms from Nature, without which there is
generally a stiffness and want of freedom in his style of work. For
coloured work, such as carpets, &c., skilful colouring is the most
essential point. No drawing will atone for bad colouring; at the same
time, good colouring should not be wasted on bad drawing.

A designer may draught his own patterns, and if he is an experienced
draughtsman with a good knowledge of weaving, perhaps this is the most
successful method of working; but if the greater portion of his time
is to be taken up with draughting, his talent as a designer will be,
to a great extent, lost, and he will probably deteriorate; besides, he
will not have the same opportunity for getting up a variety of designs
that he would have if he had nothing else to attend to. A designer for
any class of work should have a knowledge of the technicalities of the
material for which he is designing a pattern; at the same time, a very
slight knowledge is essentially requisite, provided the sketch be given
into proper hands to be worked out. It must be remembered, however,
that in this case a sketch may often have to be taken only as an idea
for a pattern, and not be handed by the manufacturer to the draughtsman
with directions to make a truthful reproduction of it on the cloth. A
draughtsman may be a skilful designer as well, or he may be a skilful
draughtsman and have very little powers of designing. When he is a
designer, the most successful method of getting a variety of patterns
would be for the manufacturer to purchase sketches for ideas and hand
them over to his designer, who will be considered to have a thorough
knowledge of the practical work, to prepare them to suit the fabrics
for which they are intended, perhaps completely altering them, making
two or three out of one, or combining two or more to make one pattern
if necessary--in fact, using them as material to work from. They may
then be handed over to the draughtsmen, or be sent to a designing
establishment for draughting and cutting. In case the manufacturer has
no designer in his place, but only a draughtsman, the sketches must be
bought prepared to suit his fabrics, or it would be better to have the
complete work done, in many cases, at a designing establishment. Of
course, in draughting, as in all classes of work, a great deal of the
less important portions may be done by junior hands, but a knowledge of
drawing is required, if only to guide the eye, by them all, except the
twillers.

Sometimes designers accustomed to selling sketches get them up in
a very deceptive manner in order to deceive manufacturers or their
agents, who may have very little knowledge of the work; but this can
easily be checked by submitting them to a practical designer before
purchasing them.

A design that may be very suitable for one class of cloth may be
equally unsuitable for another class, and in many cases fashion
regulates the suitability quite independently of its artistic merits.
For coloured work, designs should be of such forms as lend themselves
to the disposing of colours, no style being more suitable than Persian;
whereas in plain work, as damask, the flow of line or the variation of
the forms, assisted by portions of fancy shading or twilling, must give
the complete effect. Again, different degrees of fineness of cloth will
suit for different designs; of course, any pattern that can be put on a
coarse cloth can also be put on a fine one, but many patterns are very
suitable for fine work that could not be put on a very coarse fabric.
When it comes to fine coloured silks, anything that the artist can
paint can be fairly represented, as may be seen in the work from the
Coventry silk marker looms, so that any design can be put upon cloth;
but the question is, Is the cloth suitable for the design?

[Illustration: Fig. 68]

Fig. 68 shows a piece of 8 × 9 design or point paper; 8 × 9 meaning
that each of the large squares, or designs, contains 72 small checks or
squares, 8 in breadth and 9 in length, but the 9 should occupy the same
space as the 8. Ten by ten appears to have been the standard design,
as patterns are still counted by the 100 designs, each containing 100
checks, or 100 checks in length and the same in breadth: and in some
places machines are known as so-many-design machines; thus, a 400
jacquard is called a 40-design machine.

When each design contains as many checks in length as it does in
breadth, the paper is intended for work that is to have equal
quantities of warp and weft threads in it; thus 8 × 8, 10 × 10, 12 ×
12 paper would all suit for cloth with, say, 80 threads of warp and 80
picks of weft per inch, the difference being that 8 × 8 is intended for
a jacquard with 8 needles in the row, and 10 × 10 and 12 × 12 for 10
and 12 row machines respectively. It is not necessary to have different
papers to suit, as one could be used for all, and after the pattern is
painted it could be ruled in rows to suit the machine, but it is much
more convenient to get the correct size of paper.

If 80 threads of warp required to have 100 picks per inch, then to find
the size of the paper state as 80: 100:: the number of needles in one
row of the machine to the number of cards in each design, giving 8 ×
10, 10 × 12-1/2, or 12 × 15; but 10 × 12, or 12 × 14 would have to be
used for the last two, as a half could not be made, and 12 × 15 is an
unusual size. Either would do by drawing out the design a little when
enlarging it for the point paper. Square paper might also be used by
counting off the number of checks required, and drawing an elongated
pattern to cover them; but it is more desirable to have suitable paper
for work that is at all particular, and in case of such as 12 × 15
paper, 8 × 10 is the same proportion, and could be used for it, the
squares afterwards being ruled in 12’s for the card-cutter.

Each upright space on the design paper, between the lines, represents
a thread of warp or one hook of the jacquard, and each space between
the horizontal lines represents a pick, or shot, of weft, or one card
of the pattern; so that a painted pattern is a magnified view of the
texture of the cloth, in common jacquard work.

[Illustration: Fig. 69]


[Illustration: Fig. 70]

Fig. 69 is a small sketch for a border with spot filling, which is
shown on design or point paper in Fig. 70. The paper is 8 × 8; the
first four checks are used for a selvage, and the pattern is on 12-1/2
designs or 100 threads of warp, and 22 designs or 176 cards would be
required for the weft, the pattern being made about three times the
size on the design paper that it is on the sketch. The design paper
has 16 checks per inch, consequently the cloth would have about 48
threads to the inch to make the pattern appear as the sketch. If the
cloth were finer, say with 72 threads per inch, then 150 checks on the
design paper would be required for the pattern, which means that 150
hooks of the jacquard would be required for working it. Whatever width
the sketch occupies (one repeat of the pattern), multiply this by the
number of threads of warp per inch to be in the cloth required, and
the product will be the number of hooks required for the jacquard, and
the number of checks or spaces required on the design for the warp.
The number of cards is found similarly from the weft of the cloth. If,
on the other hand, a pattern is to be made for a jacquard, and it
is required to find what size of pattern will work on it, divide the
number of hooks in the jacquard by the number of threads per inch in
the cloth required, and the quotient will be the size of the pattern
warp-ways, in inches. The length or weft-way of the pattern can then be
arranged to suit the number of cards, or the pattern can be made any
length to suit the style of design. Fig. 71 is a pattern of the same
style as Fig. 69, and it might be wanted to use it instead of Fig. 69
for cloth of the same make, say 40 to 45 threads per inch. This could
not be done, as it could not be put on the design paper, but it would
suit very well if intended for cloth with 80 to 100 threads per inch,
which would admit of its being sufficiently enlarged (say four times
the size of the sketch) to be correctly represented by the checks. Fig.
69 would be better on cloth having 50 to 60 threads per inch, coarser
sets requiring larger forms.

[Illustration: Fig. 71]

In preparing a design for point paper, the sketch is usually first
made out and selected by competent judges as suitable for the fabric
required, as well as for the taste of the market in which the goods
are to be sold. When selected it has to be enlarged to suit the size
it is to cover on the point paper. Sometimes the enlargement is made
on another piece of paper, and is then transferred to the point paper;
this is perhaps one of the best methods of proceeding, but it is not so
quick as if the enlargement were made upon the point paper direct. It
is usual to rule squares on both the sketch and design paper, which
bear to each other the same proportion that the size of the sketch does
to the size of the design paper required to be covered. This guides the
draughtsman, as everything in the small squares on the sketch should
be put into the corresponding large squares on the enlarging paper or
point paper.

A pantagraph is sometimes useful, and proportional compasses are a
great assistance if very correct enlargements are required. When a
careful outline of the pattern is made upon the design paper, it is
then painted. Some painters dot round the outlines, and in large forms
leave them to be filled in by assistants; others paint in solid as
they go along. Vermilion and scarlet and crimson lakes are the paints
mostly used. The first is easily washed out, but the lakes are more
transparent, and admit of the checks on the paper being clearly seen
through it, which is a benefit to both twiller and card-cutter. Scarlet
lake, with from a half to a quarter its quantity of crimson lake mixed
with it, makes a very good paint. The ordinary water-colour cakes are
the best paints to use, but powdered colour is sometimes preferred on
account of its cheapness. In Fig. 70 the painting is all black, with
white twilling on the flowering.

Simple flat ornamental forms, if of sufficient size, are not difficult
to put on the design paper, but more intricate forms and shaded effects
require a considerable amount of skill, and are tedious, unless to
an experienced hand, the difficulty being to get the checks on the
design paper to express the figures in the best possible manner, and
frequently it is necessary to slightly alter the forms to make them
come nicely on the paper.

When the sketch has to be enlarged to, say, three or four times its
size, a slight inaccuracy in the painting will have but little effect,
as it will be reduced on the cloth. At the same time, advantage should
not be taken of this to employ inferior hands at the work, as, the
more correct the painting, the more correct will be the pattern on the
cloth, although it be reduced in size; and a ragged-edged painting will
never have a clear, defined appearance on the cloth. It is in coarse
coloured work, such as carpets, that the accuracy of the painting is
of the utmost importance, as the pattern on the cloth is as large as,
or may be much larger than, it is on the design paper; therefore all
inaccuracies are magnified, and no forms that do not come satisfactory
on the squares of the paper will have a good appearance on the cloth,
so that to a great extent the design must be made to suit the paper.
When the pattern is all painted on the design paper, both the ground
and pattern for ordinary full-harness work have to be dotted over
with the texture of the cloth. This is called twilling. In Fig. 70
the texture is a 5-end satin on both ground and flower. For twilling
the ground the same colour is used as the flower or figured portion
is painted with, and for the figured portion black is mostly used,
sometimes white. The twilling dots on the figure mean blanks, as if
these checks were left without any colour on them, or as the ground,
and they are passed over by the card-cutter when the cards are being
cut.

In twilling care must be taken not to run the dots up against the
edges of the flowering so as to injure the form of the figures. In
some cases, as at any flat portion, such as a horizontal or vertical
line, or any portion of one, this cannot be avoided, but the red
dots on the ground should here fall in against the black dots on the
figure. It is necessary to begin one set of dots against the other
set to carry this out as far as possible, and in some cases, when
they will not join, the dots are set out of their places to make them
come together. The two dots coming together bind the threads and keep
the last thread of the ground on one side, and of the figure on the
opposite side--according as it is the weft or warp that is forming the
line--from hanging loose or sliding out from the others. The direction
of the twill on the ground or flower may vary so as to suit the twill
used, and whether a satin or sateen effect is required. Various
kinds of twills are frequently used to give effect, but too great an
irregularity of texture should be avoided, though a plain ground, with
an 8-or 10-end satin for the flowering, may be used for handkerchiefs
with a good effect; for heavier work a 5-end satin ground and an 8-end
satin figuring may be used, but for firm, strong cloth an 8-end satin
for both ground and figure is much better, letting either warp or weft
predominate largely in the cloth.

Fig. 72 shows how a leaf or any piece of ornament may be shaded; care
should always be taken not to let the texture be too close at any part
in the shaded portion so as to make hard pieces in the cloth, as would
be the case if a plain texture were used in a firm cloth.

[Illustration: Fig. 72]

Fig. 73 shows how a flower and bud should be painted so as to give a
natural, or rather semi-natural, representation. The shading should be
made to express the form as well as possible; a variety may also be
made in the twilling on the flower, as may be seen on the front petal,
where a straight twill is used; this makes the cloth richer and the
petal come out fuller. This pattern is rather small on the design paper
to come well on the cloth; it would be better twice as large, as it may
be seen that in many cases single lines of the design paper have to be
used for divisions, and for full-harness work it is generally better to
have at least two lines, representing two threads.

[Illustration: Fig. 73]

This painting, omitting the twilling, would suit for a pressure
harness, with each line of the design paper representing two or
three threads of both warp and weft. No twilling is required on a
pressure-harness pattern, as the texture is wrought by the front
mounting; otherwise the pattern is the same, bearing in mind that as
each check represents two or three threads, it will not be reduced on
the cloth as a full-harness pattern would. The twilling of this pattern
is the 8-leaf satin, which is generally used for fine damask, giving a
much richer effect than the 5-leaf satin.

It will perhaps be well to make some reference as to the desirability
of using these natural forms for designs, though it would be useless to
enter into the oft-disputed point of whether it is correct or in good
taste to attempt to imitate natural forms upon cloth, or whether even
shaded or rounded forms should be treated upon a flat surface. To the
latter I would answer without hesitation, by all means do so, but make
the appearance satisfactory and keep them in good taste; and this can
be done if the subject treated is duly considered.

The former, it must be admitted, is rather pandering to the popular
taste than following the dictates of decorative art. One thing is
certain, that the ‘million’ are better satisfied with floral forms than
any other class of ornament, and the more loosely treated these are,
or the further they are from ornament, the more pleasing they are to
those uninitiated in the beauty of line, wanting which no ornamental
forms can be pleasing to those educated in art. However, these natural
forms suggest Nature, and their admirers have so little idea of
Nature’s beauties that they do not see the defects in the attempted
representations. At the same time, it must not be understood that a
pleasing design, or one in good taste, cannot be composed of natural
forms--quite the contrary; for small, simple patterns, principally for
light fabrics, they are perhaps more suitable than any other ornament,
giving lightness and gracefulness of appearance, as may frequently be
seen from the designs on the better class of dress goods, whether woven
or printed; but the natural treatment must be kept subservient to the
flow of the design. Generally, a moderately flat treatment without any
strong attempts at light and shade is most successful, the treatment
consisting of a graceful arrangement of forms drawn from nature.
Everyone knows the beauty and grace of the Japanese designs; but this
is not simply because they are natural forms: it is the arrangement
of the forms that gives the effect. Very effective patterns can be
made for dadoes of coloured curtains from natural objects, as here
they can have the upright forms suited to them; but for coloured work,
generally speaking, and for carpets in particular, natural forms,
unless treated flatly and conventionally, are rarely either pleasing or
in good taste. With them it is difficult to obtain that intermingling
of colours, be they bright or dull, that produces a neutral bloom and
a satisfactory result. Another important point in a pattern, and often
overlooked, or not understood, is that, unless the pattern, as a whole,
is satisfactory, no variety or beauty of detail can make it a good
design. The general character or lines of the pattern must first be
made satisfactory, after which the detail may be made as interesting
as possible, provided it does not injure the general effect, and
is not incongruous to the nature of the design. As to whether flat
surfaces should be treated with shaded ornament or not is a point that
will always be in dispute, but it may be asked why should a designer
be hampered with such restrictions? If he can produce a satisfactory
pattern by using either flat or shaded ornament, or both, why not let
him do so? The criterion should be whether the result is satisfactory
or not. Of course such barbarous work as representing leopards or
tigers prowling over carpets or hearthrugs, or even decorating them
with the more homely duck or drake nestling, or waddling through
water, cannot be too strongly condemned, highly though they were
once appreciated, and though they still find favour with many: even
shading forms or objects in such a manner as will produce a feeling of
weakness, insecurity, or danger, can never be upheld--that is, such a
practice as shading ribbons flowing over carpets or hearthrugs like
snares, and buttoned down at certain points; or making the surface
of the carpet appear very irregular, which has often been done, and
is often seen in coloured tile pavements. Of course these remarks
refer more or less to any fabric, but a little relief in a curtain
would not be nearly so objectionable as on a carpet, so that almost
everything turns upon whether the result produced is satisfactory or
not--of course, to those competent to judge, and who are not overcome
by biassed prejudices. Natural or semi-natural floral ornament will
rarely be injured by a little shading, whether on a white or coloured
fabric, but it should not be overdone--a moderate amount of shading
or a half-flat treatment generally giving a more pleasing effect.
Conventional floral ornament is, perhaps, best treated flatly, or with
a very little shading on some portions that require a little relief. A
richer and purer effect in colour can be got with flat than with shaded
colours.

[Illustration: Fig. 74]

Fig. 74 is an example of a small floral pattern of a class much used
for dress goods, either woven or printed, and gives a very pleasing
effect. One repeat of the pattern is bounded by the lines _a b c d_, or
the whole might be taken as one repeat, and instead of having a row of
pears on a line, have a pear and some other fruit alternately, and the
same with the apples. This pattern is a semi-natural floral treatment,
and if on a larger scale a little more shading on the fruit and a few
turned-over leaves or half-open flowers would not injure it, but would
add to its richness.

Fig. 75 is a sort of floral ornamental, or flatly and symmetrically
treated floral pattern, somewhat after the Persian style, and though
some would consider it stiff in comparison with the previous one,
it has a very pleasing effect, and could be used the size indicated
for dress goods, &c., or if enlarged to two or three times the size
it would make a very good curtain pattern. Small sprigs of flowers,
grasses, grain, &c., scattered about over the surface of the cloth make
generally a favourite style of pattern.

So far the patterns given are such as would require the designer to
have a good knowledge of drawing and of ornamental and floral forms,
and many consider that a designer must be a draughtsman; but such is
not the case.

[Illustration: Fig. 75]

Large quantities of patterns are made that do not necessarily require
the designer to have any knowledge of drawing; this is called ‘small
pattern’ or ‘texture’ designing.

A designer of this class should have a thorough knowledge of weaving
and of the effect which the patterns on design paper will produce on
the cloth. He must, of course, have a taste for the arrangement of
forms, and can produce considerable variety by a judicious arrangement
of fancy threads through the cloth, particularly in worsted or woollen
goods. The following examples will show how great a variety of patterns
can be produced without any knowledge of drawing. But a knowledge of
drawing will be a great assistance, as it trains the eye to correctness
of form. Fig. 76 is a pattern for dress goods; it may be made of cotton
only, or with a cotton warp and worsted weft, and a good effect is
produced by having the warp a different colour from the weft, say a
gold cotton warp and a brown or giraffe worsted weft. Warp 60 to 80
threads per inch, and weft about the same. This is a pattern of the
bird-eye class, and a very large range can be made in this style, both
bold and effective, or fine and neat, as may be desired. The smaller
patterns may, of course, be woven with shafts, but larger ones require
a jacquard.

Fig. 77 is another pattern, of the flushed stripe class, also suited
for dress goods; this is a silk handkerchief pattern for 90 threads
and 96 picks per inch; 60/2 China silk for warp, and 60’s single China
silk for weft. If the small dots on the pattern are cut on the cards,
a finer and closer pattern will be produced than if the black squares
only are cut, but of course it would be less effective unless on a
coarser fabric.

[Illustration: Fig. 76]

[Illustration: Fig. 77]

Fig. 78 is a pattern that would suit for dress goods if woven similarly
to Fig. 76; and if woven much coarser, say 25 to 30 threads per inch,
it is suited for quiltings. Great variety can be made in this style
of quilts, particularly when floral and other forms are interspersed
through the texture, and also when the honeycomb texture is employed,
an example of which is given in Fig. 79 in combination with diaper
or bird-eye figures; only a portion of this pattern is given, but
it shows all that is required. In either pure white, or with a warp
of one colour and a weft of another colour, these patterns are very
effective when made from good twist yarn and coarsely set in the
reed; the flushing might be over twice as many threads as is shown in
the patterns, and then more threads of warp and weft be used. It is
perhaps to the woollen and worsted trades we must turn to find the
greatest variety of patterns of the description we are now dealing
with. True, the greater portion of them can be wrought on shafts, and
perhaps a greater number of shafts are used in these trades than in
any other branch of textile work, as shafts produce a firmer cloth
than a harness; but when large fancy patterns are required, recourse
must be had to the jacquard. Fig. 80 is an example of a fancy twill
stripe which could be wrought on 48 shafts if desired, but might
also be wrought on a small jacquard with greater simplicity, unless
very heavy cloth is required. Patterns of this description require
considerable skill in designing, and are in some cases more difficult
as they increase in size. It may be seen that the pattern or figure
is composed of different textures, and to make a perfect cloth the
various textures should all work together at an even tightness, or
so that there will be an equal amount of take-up of shrinkage of the
warp for each. When checks or large patterns are to be formed this is
most essential, but in smaller patterns it can in a great measure be
remedied by letting the various textures follow each other, so that
if one tightens a little another will make a corresponding slackness,
and all together will balance each other. Another important point is
to join the edges of the various textures together so that there will
be no break, or floats of warp and weft greater or less than in the
textures themselves. In some cases tight and loose textures are wrought
together to give a special effect, and figures may be formed by the
close texture pressing the looser threads together.

[Illustration: Fig. 78]

[Illustration: Fig. 79]

Another class of pattern by which a very good effect can be obtained
by very simple means, is the hair line or tricot style. Fig. 81 is a
dice pattern on this principle. The texture is plain cloth throughout,
and the effect of pattern is obtained by using two colours of warp and
weft, say black and white, and by arranging the draught and treading so
that the black weft will pass over the black warp and the white weft
over the white warp on one dice, and the reverse of this for the next
one, a horizontal and vertical lined effect will be produced, giving
a subdued pattern. The dark checks on the design paper are the tricot
or horizontal line effect; the grey checks the vertical lines or hair
line effect. It will be seen at the edge of each dice how the colours
are changed in both warp and weft by the two dots coming together. The
crosses along the bottom and to the left side of the design show the
dark threads and dark picks coming alternately with the white ones.

Fig. 82 is a small figure pattern arranged in this manner, the
reversing of the colours being done by the jacquard, as may be seen
from the dots on the design. Almost any figure may be treated in this
manner, but simple patterns, not too irregular or broken in outline,
will perhaps be found most successful.

[Illustration: Fig. 80]

These fabrics may be made in either woollen, worsted, or cotton for
dress goods. About 36 threads of warp per inch of 2/30’s worsted, and
32 threads of weft per inch, is a good setting for them; and in any
mild contrasts of yarns, such as two shades of grey, they are very
pleasing.

[Illustration: Fig. 81]

[Illustration: Fig. 82]

Coming now a step further towards the damask or figure patterns, we
have still great scope for variety, without much, if any, knowledge of
drawing being required from the designer. This consists of spotting
small geometrical figures over the cloth, the body or ground of which
may be plain, twill, or any texture desired. Fig. 83 is a pattern of
this description with a plain ground. The figures are arranged in
satinette or 4-end satin order. Of course, the ground might be a twill
if desired, and the figures may be of any form that ingenuity can
suggest. They may be solid, and a twilled texture used to bind them
if the flushes of warp or weft are too great, and part of the figure
might be warp flush and part weft flush; in fact any variety that the
designer can arrange to produce a good effect on the cloth, which, if
the warp and weft are of different colours, may be made very effective.
Any arrangement for the figures may be used if suited to the size of
the figure and the space they have to occupy to suit the repeat of the
pattern. The 4-end satin gives two figures on one diagonal, and two
on the other diagonal alternately; one figure on each diagonal may be
used, and also three figures on each with equally good effect, subject,
of course, to some extent to the size and form of the figures. Five
and eight-thread satin arrangements are two of the best that can be
used if a greater number of figures are required in each repeat of
the pattern. To arrange a pattern in satin order, the usual way is to
rule the repeat into as many squares in length and breadth as there
are threads in the satin to be used. Thus, for a 5-thread satin, rule
the repeat into five squares in length and the same in breadth, or
twenty-five squares in all, then place one of the figures into each of
five of these squares, the order being that of the 5-end satin twill. A
draughtsman would either sketch these figures on the design paper, or
sketch one of them on a piece of plain paper, trace it, and transfer
it to the different positions it would occupy on the design paper,
and then paint each of them independently. But a designer, without a
knowledge of drawing, would mark out the spaces on the design paper for
each figure, and, having made one of the figures the required shape on
the checks or small squares, would copy the others from this. In Fig.
83 let 2A be the first figure formed, it being started in the second
square upwards and first row. Now, the second figure can be similarly
begun in 4-thread satin order, which would bring it to square No. 2,
and it is begun in the same position in this square that the first
one was begun in square No. 2A. The third figure falls to the third
square from 3A and 3, but as this figure is turned round for variety,
the point for beginning it at figures 1 and 2 must be marked, and an
imaginary square run round it, and it must be kept in its position in
the square, or the centre of the figure may be found by counting the
checks in either of the first or second figures and this figure wrought
from the centre; or, again, as the figure is six checks greater in
length than in breadth, it must be set three checks to one side, and
the top point should rise three checks higher up above the large square
of the design paper on which it should be started; but with a plain
ground it is sometimes necessary to move the figures one check out of
position to allow the plain to fall in regularly all round them. This
figure is set one check down. Fig. 4 is copied from Fig. 3, and is in
the same position. With a twilled ground or irregular figures, it will
not matter if the ground does not fall in quite regularly round them.

[Illustration: Fig. 83]

[Illustration: Fig. 84]

If it was required to make a pattern similar to Fig. 83, but with five
figures instead of four, and preserve the same density of ground
structure round them, it could be calculated as follows:--Fig. 83 is on
44 checks broad and 44 long. Now 44 × 44 = 1936, and on this there are
four figures, then 1936 ÷ 4 = 484. As five figures are required 484 × 5
= 2420 checks for design. If there are to be as many checks across the
design as up it, find the square root of 2420 for the number of checks,
as [sqrt]2420 = 50 nearly. If the pattern is not to be square, the
proportion may be found for each side by making two calculations, one
for the greater number of checks squared, and the other for the lesser
number of checks squared, and extracting the square roots for the two
sides of the required pattern; or, suppose the warp to be to the weft
in the proportion of 10 to 12, then state, as 12 is to 10, so is 50 to
42, the number of checks across the bottom of the pattern; and as 10:
12:: 50 to 60, the number of checks in the length of the pattern, and
42 x 60 = 2520, which is nearly the same as 50 squared = 2500 for a
square pattern. But if this is for adding extra weft to the cloth, the
size of the pattern would be reduced, which might bring the figures
too close together and would alter the shape of them; in this case the
addition should be made to the number of checks in length, making the
pattern 50 broad and 60 long, or to contain 3000 checks, and the length
of the figures should be increased in proportion. Of course this will
produce a finer cloth. Fig. 84 is a pattern arranged with five figures
on it.

A great variety of patterns may also be formed by arranging dice or
squares in various ways, some of which, as may be seen on the Indian
fabrics, make very pretty patterns. It will thus be seen that there
is plenty of scope for the ingenuity of a textile designer without
his being of necessity a draughtsman; but a knowledge of drawing will
generally be of assistance to him. For the more elaborate patterns, the
designer does not require to have the same knowledge of manufacturing
that is required for these small patterns.

[Illustration: Fig. 85]

_Twilling._--It has already been shown how the patterns are bound
or twilled to form the texture of the cloth for ordinary repeating
patterns on a straight over-tie; but when a turn-over or gathered tie
is used, the twilling of one half of the gathered portion will, on the
cloth, be the reverse, or run in the reverse direction to that of the
other portion, and this is often a disadvantage to this description
of tie, as one half of the border or pattern, whatever it may be, will
look coarser than the other. It frequently happens that to avoid the
stiffness of a clean turn-over pattern, which must always have a clear
cutting line through the centre of it, so that, if cut up, one half
would, if turned over, be exactly the same as the other, a portion of
single, or a turning portion, is placed in between the two turned-over
portions, which in many cases adds very considerably to the effect of
the pattern, though it is often almost useless. Fig. 85 is a pattern of
the turn-over type, and, but for the interlacing of the bands in the
centre, might be wrought with a simple gathered tie. For this pattern
the portion marked B would require to be single, and A turned over to
C. Perhaps it is in floral patterns that this form of tie is of most
service. When a single tie takes up too much machinery, and a double
over or gathered tie would make the pattern too stiff, a small portion
of single introduced, as at B, is of great service to the designer in
making a pleasing pattern.

[Illustration: Fig. 86]

When twilling a pattern of this description, the portion marked A would
be twilled straight across, and this twilling would be continued to
the centre of B, as shown in Fig. 86, which gives the idea, the spots
in the centre being taken as the single portion. The portion marked C
is repeated from the first part A by the harness, and does not require
to be put on the painting; in the figure it is shown (in different
type) as it would fall on the cloth. It may be seen that the twilling
dots on the last line of A and the first line of C are alike, and that
every pair of lines corresponds, moving from these to the edges of the
pattern. Now a turn must be made in the twill in the single portion so
as to make it join correctly with C, and this should be done in such a
manner as to prevent it from showing plainly on the cloth, or having
too long floats on some of the threads while others are too closely
bound. This must be done by setting the dots out of their places on
a few lines in the centre, so as to make the join produce as even a
texture as possible. Advantage should always be taken of any portions
of the pattern that come on this part, such as the spots in Fig. 86,
to turn the twill round the edges of them. By following the dots on
the design paper, it will be seen how they are brought in at the join.
Any twill may be broken similarly, but a different number of threads in
the single will alter the arranging of the dots to join the two twills
together; however, when the principle is understood, it can easily be
worked out.


LETTERS, MOTTOES, ETC.

[Illustration]

A turned-over mounting or gathered tie will reverse any ornament, but
while for borders it makes the ornament come correctly, it prevents
letters or words from falling in as they should be read. Thus, if the
word CLUB were woven at one side of a cloth, as at A, and the harness
mounted as a gathered repeat so as to turn over a similar border to the
other side, the letters would fall as at B, making them read correctly
to the under side of the cloth instead of to the upper side. Instead
of making a gathered tie, it would be better to make a plain repeating
tie for the second border, making the word CLUB on both borders as
either A or B, according as they are to read correctly on the under
or upper side of the cloth; but if it is desired to make the letters
read correctly from both sides or edges of the cloth, as at C, it is
plain that there must be hooks in the jacquard to work each border
independently, as one is not a repeat of the other. Of course letters
on damask will only read correctly on either the face or back of the
cloth, not on both. The same principle holds good when putting names
across the cloth, or for coats of arms and mottoes in the centre, as
is frequently done in quilts, tablecloths, &c., but in this case it
would be the cards that would do the work. Suppose the word HOTEL to
be put across a cloth at one end, and it was required to have the same
at the opposite end to read similarly from that end of the cloth, it
would require to be painted thus, [Illustration: Inverted and reversed
HOTEL], or the cards might be cut from the painting of the former,
turned upside down. Provided it was required to make these words read
correctly on the under side of the cloth in the loom instead of on
the upper side, then the first border would be painted and cut as
[Illustration: Mirrored HOTEL] and the second border as [Illustration:
Inverted HOTEL] This will be best understood by printing the letters on
tracing paper and turning it round into the position required. In case
of a motto or coat of arms wanted for a large cloth with two centres
in it, one centre to be correct from one end of the cloth, and the
other from the opposite end, as shown by the girdles at A and B, Fig.
87, the painting would be as in this figure if the upper side of the
cloth is to be the right side, but if the under side is to be the right
side, the first centre would require to be painted as shown in Fig. 88,
and the second one would be as this turned round, the top where the
bottom is, not turned over. It might be thought that the same painting
would do for both; either that cutting the cards from top to bottom,
or backwards, or lacing them backwards, or working them backwards on
the loom, would do for the second centre; but it would not unless the
cards were turned over as well--that is, to have the outsides of the
cards turned inwards, and the right-hand end to the left. This would
suit if it could be done, but could only be done when the jacquards are
made with an equal number of needles to each side of the machine, as
25 rows, or 200 needles, to each half of the card for a 400 machine,
instead of 26 rows to one side and 25 to the other, making 408 needles,
as is usually the case.

[Illustration: Fig. 87]

In hand-loom districts, where old makes of machines are used, it is
common to have 400, 500, or 600 needles to the machine, and the cards
can be turned on the cylinders if required, and are also wrought
backwards or forwards to suit. By cutting the cards from the painting
in the reverse order--that is, beginning at the end of the painting
and reading and fingering backwards--cards from a painting of a first
centre could be cut to suit for a second centre, as it has exactly
the same effect as turning over the card; but all the rows of needles
must be used except any left at the first, and these and any selvage
must be brought to the end of the painting when cutting the second set
of cards, so as to get them to the first end of the card. The cutting
begins, as is usual, at the numbered end of the card. Repainting the
centre to suit is the safest way to avoid mistakes.

[Illustration: Fig. 88]

When a name is running up the side of the cloth, lacing or working the
cards backwards will reverse the reading of the letters from one side
to the other, but not when they are running across the cloth. If the
loom is mounted with the cards hanging to the front instead of to the
back, the letters would also be turned from one side of the cloth to
the other, and to be correct they would have to be painted or cut the
reverse, unless the machine or mounting was built to suit.




CHAPTER V

_CARD-CUTTING AND LACING_


The mountings which have been given are of two descriptions--the
Norwich, with the straight harness, and the London or twisted harness,
with the cards hanging to either side of the loom, as the machine
can be turned either way to suit. For the Norwich ties the cards are
supposed to hang to the back of the loom, but it often happens that
there is not space between the looms to admit of all the cards being
hung at the back, and it is usual for one loom to have the cards to
the back and the next one to the front. When the cards hang to the
front of the loom the mounting is tied up so that when finished it
will be the same as if mounted for the cards to hang at the back; and
afterwards turned round so that the back will be at the front. In case
the back left-hand corner twine is the first cord of the harness with
the cards at the back, the front right-hand twine would be the first
if the cards are at the front, and the drawing in of the yarn must
begin at it. If the loom is mounted in this way, and the yarn drawn in
from the opposite side, as if the cards were to hang at the back, the
result will be a toothed or broken-up appearance round the edges of the
pattern. The same cards will do for either mounting, but the pattern
and twill will be reversed on the cloth--that is, they will run from
left to right on the one, and from right to left on the other. The
effect produced on letters was mentioned when speaking of them.

It was mentioned that jacquards are usually made with 51 rows of hooks
or needles, of 8, 10, or 12 in the row, being for 4^{00}, 5^{00},
and 6^{00} machines respectively, which are those most used; 2^{00}
and 3^{00} machines are used for small mountings, but they may be
considered as half 4^{00} and 6^{00} machines. Sometimes the machines
are made of any size required--that is, with any desired number of
rows of holes in the needle boards. The objection to this is that the
cards cannot be copied on the ordinary repeating machines; but piano
card-cutting machines can be made with the index to suit for cutting
any length of card. When using 4^{00}, 5^{00}, or 6^{00} machines, or
the double sizes, 8^{00}, 10^{00}, or 12^{00}, and more than one is
required for the mounting, a second must be used, such as an 8 and a 4,
or a 12 and a 6; a 12 and a 5 would not do so well, as the 12 has 12
needles in the row and a 5 has but 10. It would be better to use a 12
and a 6 and leave 100 hooks idle, if only 1700 are required. If a 12
and a 5 is to be used it may be done by having the cumber board made 12
in the row, but made finer in the proportion of 12 to 10; then, where
the 12-row harness is to pass through it, every sixth row of holes can
be left empty, and it will be all right for the 10-row harness from the
5^{00} machine.

Designs are painted upon the point paper for the purpose of enabling
them to be transferred to the cards, which act on the needles of the
jacquard, and cause the proper warp threads to be raised. The pattern
is put upon the cards by punching holes in them, one hole being punched
for each check on the design paper that is filled in with colour, or it
may be the reverse of this, or otherwise, according to circumstances.
The common method is to cut the red, and leave the ground or unpainted
portion, and the black, which stands for ground. There is a card for
each line of the design paper for damask patterns; the holes in the
cards represent the dots on these lines, if each line was cut off and
cut into separate designs or larger squares, then these squares set up
lengthways and laid together. Take the pattern, Fig. 70, and cut the
first line from left to right. The card is shown at Fig. 89 at A. Turn
the pattern upside down and examine the first line to the right-hand
side. The two first checks are empty; two holes in the card or two
punches of the piano machine are passed over; the next three checks are
painted, for these three holes are cut in the card, and the remaining
three checks in the design being empty are passed over. The next design
has the second and seventh checks filled and these are cut on the card,
and so on with the others. The two small holes at each end of the card
are for the lacing twine, and the large one at each end for the peg on
the cylinder of the jacquard. The intersections of the fine lines on
the card show where the holes fall when required to be cut.

This card would be for a 200 machine, the empty or unused needles being
left at the first half of the card; 26 rows of 8 needles to the card.
The card B given in this figure is the first of the heavy cutting, No.
17 on the pattern. Reading from the pattern as before, pass 2, cut 2,
pass one which is black, cut 3, and so on, cutting the red and passing
the black or twilling dots whether they be white, black, blue, yellow,
or green; they are only there for convenience, to avoid the trouble of
leaving the spaces empty when painting the pattern at first, it being
simpler to dot them over afterwards.

[Illustration: Fig. 89]

The cards are numbered at the selvage end, which goes to the sneck
side of the jacquard when working. The design may be read from left to
right, but the usual method is to turn it upside down, and read from
right to left, which comes to the same thing, beginning to cut from
the numbered end of the card. Some begin to cut at the other end of
the card, and read from right to left on the design without turning
it, which is still the same. If the loom is mounted or the yarn drawn
in, beginning with the front row of the harness instead of with the
back row to the left-hand side, the cards should be read the reverse
way from the design paper; or, what would be the same, turn them over
after cutting them. This changes the top row of holes in the cards to
the bottom, or from the back to the front of the cumber board. In case
of the loom being mounted from back to front, and the yarn being drawn
into the harness beginning with the front row in the cumber board, the
cards, if cut in the ordinary way, would give the pattern a broken or
toothed appearance. Turning the cards upside down would remedy this,
but would only answer if each part of the mounting consisted of full
rows of needles, for, if any part began or ended on a broken row,
turning the card would take the holes in it for this part row away
from the needles: that is, say if the harness was connected with four
needles at the top of the row and the card cut for these, the holes
would fall on the four needles at the bottom of the row, if the card
was turned upside down. The cards should be numbered on the side which
is uppermost when they are being cut, and this side should come against
the needles if the work is properly carried out.

[Illustration: Fig. 90]

The old method of cutting cards was for one to take the pattern and
call out the holes to be punched, and they were punched out by a second
person, a perforated plate over the card being sometimes used as a
guide to place the holes in their correct positions. A better method
was found in the punching plates shown in Fig. 90. A and B are two
plates hinged together. B No. 2 shows the full upper plate. The card is
placed on the lower plate, and the upper one closed down on it. Punches
are then filled by hand into the holes in the upper plate according as
the pattern is read from the design; then the plates are passed through
a roller press, which pushes the punches through the card. With small
machines this suited very well; but the filling in of the punches was
tedious for larger machines, and shortly after the introduction of the
jacquard by Mr. S. Wilson, in 1821, he introduced a method for filling
the plate with punches to correspond with each line of the design
paper, as follows:--Over, or in front of, the design a set of upright
cords is placed, as in reading the patterns on the simple of the
draw-loom. On these cords the pattern is picked, and weft threads drawn
in, one for each line of the design paper, for plain damask. When all
the design is finished the threads compose one repeat of the pattern
in a loosely-woven cloth. The upright cords or leashes pass through
eyes in a set of needles, arranged like those in a jacquard machine and
each weft thread put in, when picking the pattern, shows which of the
leashes are to be drawn. As each set of leashes is drawn the needles
connected with them come into contact with a set of punches arranged in
a frame or plate, and push them into a punching plate, which is then
placed over the card, as in Fig. 90, and passed through a roller press.
This method of picking the pattern for card-cutting is still in use,
and an improved method of using the plates and rollers is still found
very convenient for many purposes, such as copying cards and small
patterns, &c. (See Copying Machines.)

[Illustration: Fig. 91]

Card cutting is now almost entirely done by ‘piano machines.’ They take
the name of piano from the pins or keys which are arranged for the
cutter’s fingers, and to distinguish them from an older machine, which
is similar in make, but in it the pins or keys which act as locks on
the punches are wrought with cords instead of by the fingers. The cord
machine is similar in make to the piano machine; but the cutter usually
sits at one side, and has the board or stand for the pattern facing
him. The treadles are also turned sideways; but the cord is put in
at the end, as in the piano machine. Fig. 91 shows how the cords are
arranged. AA are the cords, 12 in number, which the cutter draws. BB
are another set from the former ones to the keys, passing under pulleys
at C. D is the punch-box. The keys have springs on them, which press
them in over the heads of the punches; but at the top of the cords are
indiarubber springs, fastened to the crossbar E on the frame which
supports the cords, and these springs are strong enough to draw out the
keys. When any of the cords AA are drawn down the cords BB connected
with them are slackened, and the keys spring in over the punches,
locking them, and punching holes when the treadle is pressed down.

The cords BB can be connected from the right-hand side of AA to the
right-hand side of the punch keys, and run from right to left, or they
may be the reverse of this, according to the way the cutter reads the
pattern. Sometimes piano machines are fitted to work with cords to suit
the cutter. In this case there is only one set of cords, as AA, set
behind the punch block, and the cutter sits at the end of the machine.
Instead of springs, weighted levers, or tumblers, may be used at the
top of the cords on the bar E. When springs are used the bar should be
lowered when the cutter is not working, by slackening the set screws
at GG, so as to relieve the strain on the springs. When working the
cord machine the cutter uses only one hand to draw the cords; but on
the piano machine he uses both hands, and can proceed quicker with the
cutting.

Fig. 92 is a view of Devoge’s piano cutting machine. In these machines
one cross row of the card is punched at each tread; the cutter sits
in front of the machine, with the design fastened on the frame before
him, as shown in the figure. The straight-edge, or ruler, across the
design is for the purpose of guiding the cutter’s eye along the line
of the design paper that he is cutting from, and he screws it up or
down a line, as the case may be, for every card he cuts. He keeps one
foot on each treadle and his fingers on the keys in the punch block,
the arrangement of which is shown in Fig. 93, and presses in a key to
correspond with each dot on the line of the design paper on one design,
or large square, for each tramp.

The numbered end of the card is put into the clip on the index
carriage, and the guide for the card set to suit the width of the card
used, so that the punch will cut the holes in the centre of the card.
The lacing and cylinder peg holes are first cut, by pressing in the
keys E for the peg hole, and for the lacing 2 and 7 for an 8-row card,
1 and 8 for a 10-row, and _a_ and _c_ for a 12-row. Any blank designs,
or rows of the card, are now passed over, by running back the index
carriage, if the design does not occupy the full length of the card,
it being the usual custom to leave all empty rows to the selvage or
numbered end of the card, though it may in some cases be advisable to
be otherwise. The selvage, if any, is usually cut first, and then the
pattern.

[Illustration: Fig. 92]

Pressing down the right foot treadle, when any of the keys are pushed
in, punches holes in the card with the punches locked, and the left
foot treadle, which works alternately with the other, is for moving the
index carriage, which shifts the card from one row of holes that have
been punched to the next one, so as to place it in position for the
punches. In Fig. 93 the black dots represent the heads of the punches,
12 in number. The keys 1 to 8 are used for cutting an 8-row card; when
any key is pressed in it passes over the head of its corresponding
punch, and locks it, so that when the cutter presses down the treadle
and the punch block is brought down on the card, the locked punch is
pressed through it, while those not locked rise up. The keys _a_ and
_c_ added to the 8 suit for cutting 10-row cards, and _b_ and _d_ added
to these suit for 12-row cards. The four keys _a b_ and _c d_ can be
pressed in with the two thumbs. E is the key for the peg hole.

A good cutter can cut from 100 to 150 cards per hour, but this is
laborious for constant work. Sometimes the cutting machines are driven
by power, but as yet (1894) this is not general.

[Illustration: Fig. 93]

Messrs. Devoge & Co. can attach an arrangement to their machines so
that they can be driven with a belt. The punch block is wrought by
a lever driven by an eccentric, and is kept constantly rising and
falling; the cutter fingers the pins so as to keep time with the punch
block, and should he wish to stop, or should anything go wrong, he can
stop the card at once by raising one treadle and pressing down the
other, which throws the motion for shifting the index carriage out of
gear, and the card stands stationary, and although the punch block
continues to work it takes no effect when the punches are not locked
by the keys. The machine works very smoothly, with very little noise,
and with a little practice any boy or girl accustomed to card cutting
can easily become acquainted with working it. The index of the piano
machine is generally made for 51 rows of holes in the cards, 400, 500,
and 600 machines being those generally employed: 51 rows of 8 = 408
needles; 51 rows of 10 = 510 needles; and 51 rows of 12 = 612 needles,
which, with a few odd needles on the 25-row side, represent the number
of needles in a 4^{00}, 5^{00}, and 6^{00} machine. If larger machines
with longer cords are required, the index of the cutting machine must
be altered to suit them.

Card paper should be of such a quality as will not be easily affected
by the atmosphere. When cards are for small machines, and are only
required to work for a short time, an inferior paper will do for
them, and save expense; but for large patterns which are intended
to work for years it will come cheaper in the end to get a good
quality of paper--in fact, the best that can be made. Specially
prepared oil-saturated paper is made for the purpose of preventing the
atmosphere taking too much effect on it. For hand looms, sometimes the
cards are painted with red lead, and sometimes with shellac varnish, or
shellac (best red) dissolved in methylated spirits.

Cards are made of different weights of paper; some prefer them thin,
and others thick. A moderately thin, tough paper is the best. About 16
to 18 cards to 1 lb., for 400 cards, is a good medium weight.

After the cards are cut they have to be laced into a chain to revolve
round the jacquard cylinder on the loom. The usual method of lacing is
to have a frame with pegs or studs set in it at a distance apart to
suit the size of the jacquard cylinder. The cards are spread on this
frame as shown at A, Fig. 94, and laced with a steel needle, thicker in
the centre than at the ends; it is about 2-1/2 in. long, and has an eye
in the centre, through which a piece of fine twine is tied in a loop,
and the lacing tape can easily be drawn through this. The cards at A
are numbered 1, 3, 5, but all the numbers would follow in rotation if
the jacquard had only one cylinder; if a double-cylinder machine, half
the cards would be laced forward, as at A, and the other half (the even
numbers) backward, as at B. The odd numbers usually go to the back of
the loom, and the even ones to the front.

Several kinds of cord and tape are used for lacing cards, but two
plies of round cotton cord, well soaped or waxed, is usually the most
satisfactory for wear.

When lacing cards, the lacings should be crossed over, each one passing
from left to right of the other in turn, at each hole and between the
cards. This is important, to keep the cards straight.

The cards should be tightly laced, as the lacing is almost certain
to stretch a little; so that they are better for being rather tight
for the cylinder at first, otherwise they are liable to become too
slack after a little wear, and if such is the case they will not fall
properly on the cylinder pegs. This and other causes--such as irregular
expansion or contraction of the cards, badly cut cards, and a long draw
on the cards--occasion considerable wear on the peg holes, and now
eyelets are being tried in them to avoid the breaking-up of the holes;
but it is probable that good paper and careful work will prove a more
suitable remedy. The card should not be deeper than the face of the
cylinder; better a trifle less, to allow for the lacing to pass between
them, and then if laced as tightly as possible they will still be
pliable and work freely. Cards are usually numbered with pen and ink,
but the numbers may be stamped on with an indiarubber stamping machine
similar to that used for paging books.

[Illustration: Fig. 94]

_Lacing Machines._--The Singer Sewing Machine Company brought out a
machine some years ago for lacing cards, and it performs its work very
well, though many object to it, as it stitches a lacing tape along the
cards, instead of lacing them in the usual way. The machine is a 3-or
4-head sewing machine, according as three or four rows of lacing are to
be used on the cards. The lacing is a tape; one tape is laid below the
cards at each row of lacing, and another above, and both are stitched
together through the cards.

Another machine was invented by Count Sparre more recently which laced
or sewed the cards through the ordinary lacing holes, much in the same
manner as hand lacing, except that the lacings are looped on each other
through the holes instead of passing up and down through alternate
holes.

A still later one, by Messrs. Reid & Fisher, of Dunfermline, and Mr.
Parkinson, of Bradford, has further improvements; this machine is now
working very satisfactorily. Other machines have since followed.

_Copying or Repeating Machines._--There are several varieties of
machines for copying or repeating a set of cards after they have been
cut by the piano machine, and it often happens that this is required in
case of several looms being put on the same pattern; or it may be that
a set of cards for one loom is made up of two or three parts repeated
several times.

One of the simplest of the repeaters is a small treadle machine in
which the card is repeated by rows, the holes of the cut card acting
upon needles that push in the pins above the punches, which are
similar to those in the piano machine; and each row is punched by
depressing a treadle. This is a slow method of repeating, but suits for
small factories where it is not worth while having a more expensive
machine. Any boy or girl can repeat the cards, instead of requiring a
card-cutter to read them.

A more useful machine is the railway press repeater; it is taken
from the old method of cutting described on page 142. The cards to
be repeated are placed on a cylinder the same as that used in the
jacquard, and pass down over the ends of long needles similar to the
jacquard needles, but arranged in a box, and with long helical springs
on them. In a thick plate, or box, in front of the needles is a set
of punches, and the stamping plate rests in front of the punch box,
close up against it. By means of a treadle the needle box can be drawn
forward, and where there are holes in the card to be copied the needles
pass through them, and, pressing on the punches, push them from the box
into the stamping plate. The punches have heads on them that will not
pass through the stamping plate, but will go into the holes in the
box or frame, and when they are pushed into the plate it can be drawn
away and turned over so as to prevent them from falling out. The plate
is then placed on a box-plate made to hold the card to be cut, and
somewhat similar to that shown in Fig 90. The box or hinged plates are
set on a slide frame with rollers to travel on, and at the end of the
frame is a strong roller press. The plates can easily be run forward to
the press, and the card punched by running them under the roller; it
may then be taken out, and as many impressions from the plate cut as
are required. When finished, the stamping plate is taken back to the
punch box and placed against it as before; then, with a comb, or block
of wood, set with pins, all the needles are pushed out of the plate
into the punch box. Another card can be brought before the needles, and
the plate filled as before.

[Illustration: Fig. 95]

[Illustration: Fig. 96]

This machine is very useful for copying damaged cards, even when a
repeater as those about to be described is at hand. A damaged card can
be held between the needles and punches, and the plate filled, same as
when copying a set.

Figs. 95, 97, and 98 show the principle of working three of the
best-known repeating machines, which will copy a set of cards at the
rate of from 40 to 60 per minute. Fig. 95 shows the arrangement of the
punches in Nuttall’s patent repeater. The punches are shown at A, and
the stoppers or keys at B; the cards to be punched pass in between
the blocks C and D. B^1 shows the shape of one of the stoppers which,
when over a punch, locks it, and causes a hole to be cut in the card.
The stoppers all stand over the punches when at rest; the cards to be
copied are hung on a cylinder, and act on a set of horizontal needles,
_g_ (Fig. 96); these needles have upright wires, _m_, _n_, passing
through eyes in them, and also through eyes in the stoppers _b_. The
upright wires act as levers, working on centres, _m_; and when one of
the horizontal needles _g_ is pushed back by the card to be copied,
the corresponding stopper _b_ is drawn from above the punch _a_, which
allows it to rise when the card to be cut is raised up against it by
the block D (Fig. 95), between which and the punch block _c_ the cards
to be cut pass, so that no hole is punched in the card; but where there
are holes in the card to be copied, the needles _g_ are not pressed
back, and the stoppers are allowed to remain over the punches, locking
them, and causing holes to be cut in the card pressed up against them.

[Illustration: Fig. 97]

This machine has been superseded by one in which upright rods act as
stays on the punches. These rods stand on the punches, which have
concave tops on them; the upper ends of the rods rest under the bars
of a grid, or rebated plate, shown in section at P (Fig. 95). When
in this position the punches are all locked. A set of needles are
connected with these rods, and when the card to be copied presses on
them, those that are pushed back move the top of the rods connected
with them from under the bars to the spaces between them, which allows
the rods to rise up with the punches when the card to be cut is pressed
up.

[Illustration: Fig. 98]

The principle of a repeating machine made by Mr. McMurdo, of
Manchester, is shown in Fig. 97. The cards to be copied hang on the
card frame of a jacquard, as shown at C. From the hooks of the jacquard
to the punches A are wires in the form of a single harness. The blank
cards D pass in over the block B, which is stationary. Enlarged front
and side views of a punch are given at G and H. At 1, 2, and 3 are
shown the notches cut in the punch, through which the slide bars E
and F pass. These slides are also shown at E^1 and F^1. The bar E is
stationary, acting as a stay for keeping the punches in the box, and
the notches in them, at 1, are long enough to let them slide up and
down on the bar. The slide F is drawn out before the jacquard is drawn,
and pushed in again afterwards; if any of the punches are raised by the
jacquard, the notch 3 will be raised to the position of the notch 2,
and the slide bar will pass through the lower notch, thus holding the
punch up, and when the punch block is pressed down, holes will be cut
in the card where the punches are locked in the upper notch. When the
punch block is being sunk to punch the cards, it is necessary to have
a spring in the wires that connect the punches to the jacquard hooks,
and this is effected by having the wires made with a sliding joint and
a small spiral spring on them, as shown at L.

The principle of Devoge’s repeater is shown in Fig. 98. This is one of
the oldest machines in the market, and there are various opinions as
to whether it is excelled by any of the others, or not. It is somewhat
on the principle of the French repeating machines, being wrought by a
jacquard and harness, E, to which a second harness or set of cards, F,
is attached, and passes over pulleys to the punches A. When the harness
E is drawn by the jacquard, the cords F are slackened, allowing the
punches to drop, and they are locked in the punch box by the slide bar,
H. The cards to be copied are shown at D, and those to be cut pass
through at C. The punch block is stationary, and the block B rises
against the punches with the card on it.

The blank cards require to be laced to make them ready for the
repeaters, and also to have the peg holes cut in them; blocks are
provided for the purpose of cutting the lace and peg holes. An enlarged
view of a punch is shown at P.


WIRING CARDS

After the cards are laced, if intended to be wrought on power looms in
the usual way, they require to be ‘wired’--that is, to have pieces of
straight stiff wire, about 2 in. longer than the cards, tied across
them to the lacing, so that they may be hung on the racks or frames
prepared for them on the looms. The wires should project 1 in., or
rather more, at each end of the cards; they should be tied to the under
side of the cards, so as not to come against the jacquard cylinder, and
it also bears them better, being under the lacing. A piece of waxed
twine is mostly used for tying them on.

The distance between the wires depends upon the number of cards, and
the height they are when on the loom. For a small set, a wire to
every fourteen or sixteen cards is a very good distance apart, but
for a large set a wire to every twenty or twenty-four cards will be
sufficient in order to give a longer drop and take up less lateral
space.




CHAPTER VI

_SPECIAL JACQUARDS AND HARNESSES_


CROSS-BORDER JACQUARDS

A difficulty has always existed in working cross borders on
handkerchiefs, cloths, &c., so as to avoid having to shift or change
the cards, and also to do without the expense of getting a complete
set of cards cut for the whole cloth. It may be also that different
patterns are required on the cloth, such as cross bands of small
figures, and if, say, twenty cards would make a repeat of one figure,
and it was required to weave, say, 100 repeats of this, and then change
to, say, fifty repeats of some other figure, the same difficulty would
arise as with cross borders. When there is only a short distance
between the changes of pattern, it will generally be found more
profitable, all things considered, to have the complete set of cards
cut, so as to require no changing, particularly if the pattern will
run for a length of time; but when the patterns or cloths are long,
the cards become very expensive, and even if it would be desirable to
get a complete set for the full cloth, it may be that there would not
be space on the looms to hold them, over 15,000 cards being sometimes
required to weave a complete tablecloth.

Many methods have been adopted to remedy the inconvenience and to
save expense. A few of them will be given, but it must be left to
manufacturers to judge whether they would prove an advantage to them or
not.

For weaving cloths or curtains it is a common practice to have the
cross-border cards and one repeat of the centre pattern cut. The card
rack is made as in No. 2 (Fig. 99).

When either the border or centre cards are working, they are put into
the hollow part of the rack or card frame, as at F, and the other set
is hung on the flat rest of the frame, at either D or E. When a change
is to be made, a card-shifter has to go up and change from one set to
the other, lifting the set that had been working out of the hollow
of the frame, and putting it back or forward on the rest D or E, and
bringing the set which is to work down into the place of those removed.
This, of course, is expensive, gives a great deal of trouble, and is
severe on the cards; at the same time, in many cases it is found to
answer as well as any method yet adopted, particularly where space is a
consideration, and when two or more sets of cards are used, in case of
two or more machines.

[Illustration: Fig. 99]

A better plan is to have a sliding-card frame, as shown in No. 1 of
the same figure. A is the frame for the rollers over which the cards
travel, B, B are two bars on which the card frame C, with the roller
frame A attached, travels. The card frame is only shown in end view.
Two of these must be for each set of cards, and the whole frame must
be fastened together complete for all the cards required, so that it
will all slide on the bars B, B. Suppose there are two machines working
there must be four card racks; on two of these the centre cards are
hung, and on the other two the border cards. Either set can be placed
opposite the cylinders of the jacquard, and wrought as long as is
required; then they must be taken off the cylinders, the card frame
shifted till the other set of cards comes into their place, when they
can be put on the cylinders and wrought over. This is a very good plan
for working long cloths, but takes up a good deal of space, and for
more than two machines might in many cases be unsuitable.

[Illustration: Fig. 100]

Another method is to have two sets of card irons, one above the other;
the border cards are hung on the lower frame, and the centre cards on
the upper one. The centre cards will work in the usual way, and when
the border is to be wrought the cards for it are taken round the centre
set, which, of course, are taken off the cylinder. The rollers must be
made to shift so as to direct the cards properly into their respective
frames; the top frame is smaller than the bottom one, and farther out
from the machine. If hooks and eyes are sewn on the first and last
cards of these sets, the weaver can hook them together when a fresh
pattern is required, so that the one set will carry the other round the
cylinder. Then, by drawing a cord, the rollers should change position
so as to direct the falling cards into the proper frame. The card irons
are of the usual shape, and the distance between the upper and lower
ones may be regulated by circumstances.

Another simple method is to have the card frame made with a double
curve in it, as shown in Fig. 100. One set of cards hang in each of
these curves, as at A and B. The set in the curve A can be wrought over
as long as required; then these cards can be taken off the cylinder
and the set at B put over it. One thing must be observed: that the
frame and rollers be so placed that the cards can be guided into the
receptacle B, and to do this there must be a considerable fall for
the cards, so that for a loom with a low framing this might not be
feasible. Otherwise, it is a simple method of arranging two small lots
of cards, and if one lot is much larger than the other, let A be a
narrow curve, and the small lot of cards hang in it, which will lessen
the slant of the cards falling to B.

Although such arrangements as the foregoing are required where there
is much machinery used, it is evident that for frequent changes they
would take up quite too much time. Several machines have been made
for the purpose of changing from one set of cards to another without
having to move the cards. The principle of these machines is to have
two cylinders, one at each side; one set of cards is placed on one
cylinder, and the other set on the other cylinder, so that by working
one or other cylinder, as required, either border or centre may be
woven. No. 1 (Fig. 101) shows a plan that was tried some time ago,
but did not work satisfactorily. The needles passed through a needle
board at each side of the machine, as at A and B, and had small spiral
springs fastened on them inside the needle boards. The border cards
could be put on the cylinder at A, and the centre cards on that at B.
The cylinder at A could be wrought as a single-acting machine till
a change of pattern was required; then the cylinder at B could be
brought into action, and that at A remain stationary and clear of the
needles. The heads of the hooks were made heart-shape, as shown, and
the blades of the griffe could revolve into the position shown by the
dotted lines, so as to suit the working of the cylinder at B; this was
effected by the weaver pulling a cord. This machine was given up, as it
was difficult to keep it in order.

Another method is shown at No. 2 (Fig. 101), where there are two
griffes--one at F, and the other at G; one cylinder works at E, and the
other at D. Only one set of springs is required on the needles--behind
the needle board E--as the spring of the hooks is sufficient to act in
the opposite direction. When the cylinder D is working the lower griffe
is stationary, and when the cylinder E is working the upper griffe
is stationary. The cylinders and griffes are brought into action as
required by suitable mechanism. This machine has not been considered
sufficiently successful to make it worthy of being persevered with.

[Illustration: Fig. 101]

[Illustration: Fig. 102]

The principle of Davenport & Crossley’s border jacquard is shown in
Fig. 102. The upright hooks are as in an ordinary machine. Four of the
needles are shown at _b_, and are made in the ordinary way, with the
exception of the eyes in them for the small upright needles _d_ to
pass through. These needles _d_ are connected with a set of horizontal
ones, as shown at _a_, and act as levers between them and the needles
_b_, the bars behind them serving as fulcrums; so that when the needles
_a_ are pressed back by the cards, the needles _b_ are also drawn back
through the medium of the uprights _d_. The cards least in action are
put on the cylinder for the needles _a_, and the others act on the
needles _b_. The cylinders can be brought in or out of action by the
motion shown at A, B, C, D, E, F, G. A and B are the cylinders, working
on swing motions; C A, D B, are connecting-rods for giving motion to
the cylinders; either of these can be brought into action as required.
G is a rocking lever carrying a stud on each end, which work in the
slots C and D; these slots are shown at D^1 where the notch at one side
of them will be clearly seen. When the stud on the bar G gets into
this notch, the cylinder will be driven out and in, but when not in
the notch the stud can slide along the slot, allowing the cylinder to
remain stationary. By pulling the cord H either cylinder can be wrought
as required through the connections shown, one of the connecting-rods
being raised and the other lowered. F, E are the levers for driving the
rocking bar G. This machine works very well, though the principle of
its construction is not all that could be desired, and it requires a
considerable space. It can be made a double-lift machine by having two
hooks to each needle and two griffes as in the ordinary double-lift,
single-cylinder machine.

[Illustration: Fig. 103]

The principle of Devoge’s border jacquard is shown in Fig. 103. C C is
the rocking bar, for driving the cylinders, working on the centre D.
It is driven by the levers F, G, and the connecting-rod E. The notches
A and B fall over the studs on the cylinder frame, and can either be
let down or raised, as required, by the cords K, L and the levers H, I.
The machine is an ordinary double-cylinder one, and either griffe can
be wrought to suit the cylinder that is in action. This machine can be
used as an ordinary double-acting one, as a single-acting one raising
both griffes together and bringing in both cylinders together, or as a
border machine working one griffe and one cylinder at the same time.

Another method, similar to the above, is to have a double-cylinder
machine with the cylinders wrought by the griffes, then have a stud
on the crank-shaft wheel for raising one griffe, as for an ordinary
single-acting jacquard, and the two connecting-rods from the griffes
made so that either will work on it. A hook or catch can be put in any
convenient place, so that when one connecting-rod is on the stud the
other can be hooked out of the way. The rods can be made to slide on
to the stud easily, and be fastened with a pin, so that the weaver can
readily change from one to the other.


OPEN-SHED JACQUARDS

To obviate the vibration of the harness as much as possible, as well as
to economise the wear and tear of it and to minimise the friction on
the warp threads, especially in weaving patterns requiring heavy lifts,
such as warp-faced patterns, Messrs. Priestley & Co., of Bradford,
patented an open-shed jacquard, the principle of which is shown in Fig.
104. It did not, however, prove successful, as the tacklers found some
difficulty in working it.

The machine is an ordinary double-lift jacquard with one cylinder, and
works in the ordinary way. The hooks are made as in Fig. 104, where
it may be observed there is a turn or catch on the lower portion of
the hook, at A. Above these catches is a set of bars, _b_, like a
stationary griffe, and when the hooks are raised by the upper griffe,
in the ordinary way, they spring over these lower bars and rest on them
when the griffe begins to fall. All the hooks raised would thus remain
up, were it not that the cylinder, pressing in for the second griffe
(the machine being double-acting), which begins to rise as the upper
one begins to fall, presses back those hooks that are not to be raised
for the following shot, and as they are held by the griffe above, they
spring back at the bottom, and, clearing the bars _b_, come down with
the falling griffe. In this way the hooks, when once raised, remain up
till pressed off by the card, so that in the case of working a warp
stripe with an 8-leaf satin binding the hooks would only fall for every
seventh pick.

[Illustration: Fig. 104]

[Illustration: Fig. 105]

Another machine has since been tried, but did not work satisfactorily.
In this machine the lifting hooks are formed with loops or hooks, in
addition to the top hooks. The hooks are lifted in the ordinary way
by the griffes; and upper knives, made to reciprocate horizontally,
catch the raised hooks and support them. By suitably timing the motions
of the card cylinder the hooks may be kept up as long as required by
being transferred from one set of knives to the other. The shape of
the hooks and needles is shown in Fig. 105. The needles may be as at
_b_ or _b^1_. This machine acts as an ordinary double-lift jacquard,
but has only one set of needles, which pass through needle boards at
each side of the machine, and are acted upon by two card cylinders. No
springs are required, the hooks acting as springs to keep the needles
in position. The cylinders are driven by a tappet, and can be held
against the needles as long as required; one cylinder must press on
them when the griffes are passing, so as to prevent those hooks which
are descending from being taken up by the ascending griffe. There are
two griffes; three of the knives of the one are shown at 1, 3, 5, and
three of the other at 2, 4, 6. When these knives lift the hooks by the
lower loops or hooks on them, they deposit them on to a set of knives,
_a_, _a_, _a_, _a_, in an upper grid or grating, which has a lateral or
horizontal motion.

The lower knives raise the tops of the hooks slightly above these upper
knives, so as to clear them, and the upper grid is then moved sideways,
taking the knives from under one set of heads of the hooks and placing
them under another set. Each time the griffe rises all the hooks not
acted upon by the cards will be deposited on the upper grating, and
will be transferred from one knife to another at each lift, till the
card acting upon the needles keeps them clear of the upper grating
and allows them to fall with the descending griffe. In this machine
the tugs or tail cords, as used with double-lift machines, are not
required, the double hooks serving for two single ones. Other efforts
have been made in this direction, but the plan of using small pulleys
on the tail cords, between the pairs of hooks, to which the neck twines
are hung, is the only one worth mention. The same principle will be
found for working the shafts in Fig. 115, and is more suitable for a
few shafts than for 400 to 600 hooks, or pairs of hooks.


THE VERDÔL JACQUARD

Everyone accustomed to work large patterns on jacquards, especially on
power looms, must have found the inconvenience resulting from a large
set of cards. The space taken up, the time and trouble required for
changing them, to say nothing of the cost, have been sufficiently felt,
but have hitherto had to be borne with as a matter of necessity.

For working fine tablecloths with from 150 to 180 picks per inch, or
for curtains where from 3 to 6 cards are required for each weft line of
the design, and when from 2 to 4 jacquards are used, the inconvenience
and cost of a set of cards ranging from 10,000 to 50,000 are very
considerable. The small jacquards described at Fig. 30 make a great
saving in the space taken up by the cards, as well as in the quantity
of card paper used; but it has been thought that a still greater saving
could be effected.

[Illustration: Fig. 106]

M. Verdôl, of Paris, has for a length of time been endeavouring to
perfect a system of substituting a continuous roll of perforated
paper for the cards, and has succeeded in doing so, though it is
doubtful if he will be so far successful as to supersede our present
method of working. These machines were taken up by Benson’s Patent
Jacquard Company, of Belfast, in the year 1876, but did not then give
satisfaction, the greatest obstacle in the way being the effect of
the atmosphere on the paper; and as the needles are very closely set
together, it was found impracticable to keep the paper set so as to act
correctly on them. This difficulty is now claimed to be overcome by
having procured paper which is said to be less liable to be affected by
the atmosphere than the brass plates of the machine. Several of these
machines have been tried in Scotland, but it will take some time to
prove them thoroughly and get the workmen accustomed to them, as they
are much finer than the jacquards in general use.

The Verdôl machine consists of a small jacquard, the ordinary French
make of machine, which is usually made with 440 and 880 hooks for the
single and double machines respectively. In addition there is a frame
or box attached to the front of the machine, against the needle board,
containing another set of horizontal and vertical needles or wires,
which act upon the ordinary needles of the jacquard, according as they
are acted upon by the perforated paper. Fig. 106, Nos. 1, 2, and 3,
shows the principle of the machine. B, B show the needles, and D, D
the uprights or hooks of the ordinary jacquard; only four of each are
given, but sixteen are used. N is the face-plate or needle board, and
O is the clap-board used for bringing back the needles, which have no
springs on them; but the hooks are turned up double, and the back or
turned-up portion of them, coming against horizontal wires or bars,
shown in section above the needles, acts as a spring to keep them
steady. This portion of the machine is a complete jacquard of the
ordinary French make. In addition to this there is the apparatus on
which the perforated paper acts, and which communicates this action
to the jacquard. It consists of a box or frame containing two sets of
wires--one set horizontal, as A A, about 6 in. long, and terminated at
one end by a small head, as shown separately at A^1, which are called
hitting wires. The other set is vertical, as C C, of fine wire, having
a loop on one end, by which they are suspended in a frame, and a loop
in the centre, which passes round the horizontal wires. These are
called the vertical needles or feelers; one of them is shown separately
at C^1. The points or lower ends of these needles pass through a
horizontal brass plate, G, called a guide-bar. The hitting wires pass
through a guide-plate, M, placed so that their heads will come exactly
against the points of the needles in the jacquard, as shown in the
figure. The other ends of the hitting wires lie loosely in the loops
of the vertical needles, C. Underneath the guide-bar G is the curved
brass plate I, which is in lieu of a cylinder. Between the cylinder and
the guide-bar G the perforated roll of paper I I passes; there is just
sufficient room between them for it to pass freely through, as it is
moved forward by the stud wheel K. The paper may pass from one cylinder
or reel to another, or may hang in folds in the same way as cards
usually hang. At the points of the hitting wires is a grid or pushing
frame, H, with an iron angle-bar across above each row of wires, as
shown in section in No. 1 at H. There is a clap-board above the small
vertical needles, which is shown raised at E in No. 1, and down in No.
2; this is raised and lowered with these needles, and serves to bring
them all down when raised. E^1 shows the way in which these needles
are hung in the frame that supports them. When the machine is working
the frame H has a horizontal traverse coming against the ends of the
hitting wires when they are pressed up by the card on the cylinder,
and, pressing them against the needles of the jacquard, push them back
as an ordinary card cylinder would. It may be seen in No. 1 that two
of the feelers are down and two are pressed up, as they would be if
two holes were cut in the paper and two uncut; the two that are up
raise the hitting wires connected with them, and it will be observed
that this raises the points of the hitting wires so as to come against
the angle-iron of the grid H when it is coming forward; whereas the
two hitting wires that are not raised will pass through, and not be
pushed back. This enables the card paper to act on the needles of the
jacquard in the same way as if coming directly on them, as cards do.
The cylinder has a slight vertical motion, rising about 1/8 in. The
needles, when the cylinder is down, do not pass through the plate G; it
is perforated through and raised by the cylinder ascending. When the
frame H is pressed forward the clap-board O is pressed back.

The card paper is of a special make, thin and tough, and is
strengthened along both sides and in the centre by strips of paper
pasted along it where the stud holes fall. No. 3 shows the motion by
which the clap board E is raised and lowered by the sliding of the
cranked bar F between the two studs 2. No. 4 shows the size and pitch
of the holes in the hard paper. These are traced from the paper,
and lie diagonally, two rows of 8 making one row of 16 hooks in the
jacquard.

These machines are working in large numbers in France, and appear to
give more satisfaction there than here, partly because they are better
known, and the French workmen are better adapted to handle the small
parts belonging to them. A coarser machine is more suited for the
ordinary workmen to be found in the factories of this country. But it
is strange if a machine of a pitch about half-way between the French
jacquards and ours would not be better than either; not taking up so
much room as ours, and not so minute as the French. However, in many
cases the machines we have in use are none too large to give sufficient
latitude for the errors often found cropping up in work, none of which
should occur, of course, though they often do, through carelessness
or otherwise. For instance, the card cylinder may be a little off the
pitch, and the cards may be affected by damp, or may not be correctly
cut to suit the cylinder; the points of the needles may be a little
bent by the cylinder, or by the carelessness of the weaver when turning
back her cards, and the cylinder may not be set quite fair for the
needles. Mr. McMurdo, of Manchester, has made a very nice machine of
the same pitch as the French ones, with 1,296 hooks, 16 row cards. It
works as a double-acting machine with two cylinders and two griffes,
but only one set of needles acting on the two sets of hooks. Two of the
hooks are shown in Fig. 107; each needle clasps two hooks and passes
through a needle board at each side of the machine. There is no spring
on the needle, the hooks forming the spring, the tops of the hooks
coming against the cross-rods at A A. Otherwise, this machine works in
the usual way, and is very compact where a large number of hooks are
required in a small space.

[Illustration: Fig. 107]

The card-cutting machine used for the French system of work is a very
elaborate contrivance, being more like a jacquard loom than any of our
card-cutting machines, at least in so far as the harness is concerned.
The principle of it is seen in Fig. 108. a shows the set of cords by
which the punches are wrought. The method of preparing the pattern
to attach to these cords has been referred to before, but will be
repeated again further on. These cords pass over the pulleys B and
down through a reed to the lingoes C, which keep them in tension.
Another set of cords pass from these over the pulleys D and down to the
punches F. When any of the cords at A are drawn the punches connected
with them are dropped, as raising the lingoes lets the cords pass over
the pulleys D. When a punch is let fall, it is caught by one of the
angle-irons of the grid G, which is pressed forwards when a card is to
be cut. The card paper is on the roller JH, and passes down in front
of the punches at I, which, when pressed forward, punch the paper
between them and the block K. Any punches that are held up pass through
the grid over the angle-iron bars. Only four punches are given, but
sixteen are used in the machine; they are here drawn out of scale to
show them more clearly. This portion constitutes the cutting machine;
it is worked by two persons, one behind, to draw the cords, and another
in front, to punch the paper by depressing a treadle each time a draw
is made. In addition to cutting from the pattern the paper can be cut
from other sets of cards, as in a repeating machine--cut either on
the English or French system. N is the French machine, and a set of
cards can be put on it and copied, the cords O acting on the punches;
English-cut cards can be put on the machine at L and copied, the cords
M acting on the punches. Of course, when one portion of the machine is
working, the other parts remain idle, and it might be better to have
them separate if much work is wanted.

The pattern is prepared for this machine by hand, working one repeat
of it as follows:--The design, painted on point paper, is placed on a
frame in an upright position, and over each line on it is stretched a
vertical cord, which is taken as warp. The picker then takes a bobbin
of weft and inserts it under every cord that passes over a painted dot
on the first weft line of the pattern, keeping it in front of all the
cords that pass over blank spaces, just in the same order as a shuttle
would pass through if the shed were to be opened by a jacquard. When
this is done the next line of the design paper is picked similarly,
and so on, till a complete repeat of the design is loosely woven with
cords, which are made of well-twisted harness twine. This process is
somewhat similar to reading the pattern for the draw loom, _which see_.
For coloured work a shot would require to be worked in for each colour
on any weft line, the same as it is to be woven. When this is prepared,
it is taken to the cutting machine and the warp tied to the cords A,
the weft showing the warp ends to be drawn for each card.

[Illustration: Fig. 108]


HALF HARNESS AND MUSLIN HARNESSES

So far we have been speaking entirely of ‘full-harness’ work. Though
the principle of mounting is much the same in all classes of harnesses,
there are certain deviations peculiar to each.

One drawback to the full harness is the amount of machinery required
to give a large extent of pattern, an upright hook and a needle being
required for each thread in one repeat of the pattern. In a half
harness only half the amount of machinery is required on the same
fineness of cloth to give an equal extent of pattern; but it is only
for a few classes of goods that this method of working is suitable,
leno curtains and muslins being the principal. Leno curtains will be
mentioned under gauzework, as they are wrought with a gauze mounting,
but on the half-harness principle. Figured muslin curtains, with a
plain ground and coloured flowering, are called ‘crêtes,’ or ‘crête
curtains’; these are now generally wrought in a full harness. Figured
muslins are much of the same class of cloth, but are not figured with
colour and tint for the half harness, known as the ‘common spotting
harness,’ or ‘book harness.’ For working them the harness is mounted in
the ordinary manner, but only half as many cords are required for it
as there are threads of warp. Only half of the warp is drawn into the
harness, the other half--every alternate end--passing through it, and
being drawn into a plain leaf of heddles, which is hung close in front
of the harness. In fine work two leaves of heddles may be used as one,
to prevent crowding.

The ground of the cloth is a plain texture, and was formerly wrought by
having two leaves of long-eyed heddles in front of the harness, into
which all the warp was drawn, this probably being more convenient for
the hand-loom weaver; and besides, a more even plain ground can be made
with the heddles than with the harness and one leaf of heddles. Two
shots of ground, or fine weft, are given to one shot of figuring, or
coarse weft, usually cotton rove; but to avoid using a pick-and-pick
loom when working by power, the coarse shot may be made by throwing two
finer ones into the same shed. In power looms the ground is wrought by
raising the plain leaf of heddles and all the harness alternately, and
the figuring shed is formed by drawing the harness with a card acting
on the machine. The pattern is painted solid, without any twilling or
binding on either ground or flower, so that the figuring shot will be
in a plain shed for the flower, but will be loose or unbound over the
ground, and is afterwards cut off in a cropping machine.

[Illustration: Fig. 109]

The plain shed, into which the thick weft is thrown, must also contain
the shot of thin weft thrown in either before or after it, according as
the figuring shed follows the plain shed made by raising the harness or
the heddles; for supposing the heddle shaft to be raised and a ground
shot thrown in, then the shed reversed by raising the harness and
another ground shot thrown in, then the figure shed formed by raising
50 or 100 hooks of the harness, this opens portion of the last shed,
and the figure weft will be thrown in along with the last shot of
ground. When both wefts are white this is of but little consequence,
but when using coloured wefts for the figure the ground weft would
be liable to show along with them and injure the pattern. This was
originally overcome by adopting the ‘paper harness’ from the shaft
mounting for weaving paper spots. In this mounting two sets of shafts
and two harnesses are used, one behind the other. Half of the warp--all
the odd numbers of threads--was drawn on the back harness, and the
other half--all the even numbers--on the front harness; this would be
sufficient to make the cloth, but a pair of leaves of long-eyed heddles
were used for working the ground, acting as pressers. On this method
of working, a ground and a figuring shot are thrown in alternately, as
either half of the warp can be raised by the harness for the figuring
shed, and a richer spot is thus given; but still, the rove or figuring
shot would fall into the same shed as the ground shot, which may be
seen by examining the two spots given in Fig. 109, A being wrought
on the common spot or half-harness mounting, and B on the paper-spot
mounting. The grey shots are the ground, and the black ones the figure.
However, in the paper-spot mounting it is not necessary to throw the
rove into a plain shed; any suitable twill may be used for binding the
figure, and will give a much richer effect on one side of the cloth
than plain, and tie down the ground shot, which will go into a plain
shed.

This principle of mounting is now done away with, as it has no
advantage over the ordinary full-harness mounting, which can also have
a pair of presser leaves of heddles in front for working the ground, if
desired; but they are unnecessary in a power loom, except when a very
level ground is required.

A twilled or flushed figure may be formed with the common half harness,
but only having half the warp for binding causes it to be loose and
ragged round the edges.

To work a large pattern, perhaps the most economical way, at least
in the hand loom, is to use a pressure harness with two threads in
the mail, and with four presser heddle leaves in front. The number
of leaves regulates the twill on the figure. The cloth may be woven
pick-and-pick, ground and figure, or 2 picks ground to 1 of figure.
When the harness is drawn for the figure all the heddles must be
sunk but one leaf, and sinking the two front and two back leaves
alternately, raising the back ones when the front ones are sunk, and
_vice versa_, will form plain cloth with the draft 1, 2, 3, and 4 over
the leaves. A 4-leaf twilled figure can be made with this mounting, but
still it will not have the advantage of the full harness, in which the
figure can be varied in twill, and bound round the edges.


SPLIT HARNESS

The term ‘split harness’ is sometimes applied to the pressure harness
when wrought with two threads in the mail, this constituting a splitful
of warp, so that each cord of the harness controls a splitful of warp;
but this is only the finest description of pressure harness.

[Illustration: Fig. 110]

The split harness, or ‘shaft monture,’ was invented in the silk
district of Bethnal Green, shortly after the introduction of the
jacquard, for weaving rich silks which have about 400 threads of warp
per inch, but much less weft--perhaps about one-fifth of that number
of threads. The harness is wrought with the ordinary jacquard, but
there are two mails and lingoes attached to each neck cord, as shown in
Fig. 110. A, A are the neck twines, B is the cumber board, and D the
mails. If this mounting were wrought with the jacquard only, it would
produce the same effect upon the cloth as if two threads were drawn
into each mail of an ordinary harness. This would, of course, take away
the fine appearance of the cloth. To avoid this a set of shafts, C, C,
are passed through loops in the twine above the mails, and are attached
to a set of hooks in front of the jacquard, or at each side of it, or
may be wrought by a separate dobby. Twenty-four shafts are mostly used,
and are usually flat enamelled iron bars. These shafts are for working
the ground of the cloth, and can raise each row of mails separately,
as shown at No. 1 leash, thus splitting the pairs of threads that are
connected with each hook of the jacquard. One or more of these shafts,
according to the texture required for the ground, are raised at the
same time that the jacquard draws the figuring shed. No ground texture
is put on the pattern for the cards, but the figure must be twilled,
or have the binding marks cut on the cards for it. The binding of
the figure will be in pairs of threads--that is, two threads of warp
must sink together under a weft shot, though on the design these two
threads will appear as one, as they are wrought by the same hook of the
machine. On account of the number of warp threads, rising and sinking
in pairs will not much affect the appearance of the figure; though it
must be coarser than if the threads were bound separately, and will
also have a rougher or more ragged outline; but to get a large figure
with so much warp would necessitate a great amount of machinery if the
threads were drawn into separate mails with one to each cord of the
harness.

This method of working is surpassed by the twilling machines now in use
(see Twilling Jacquards), though they are more intricate, and would not
suit well for a hand loom, as they are heavy to draw.


PRESSURE HARNESS

This is the old draw-loom system of working, and, old as it is,
is still in use on our modern hand and power looms on the finest
description of damasks, and is not likely to be surpassed for making
a good piece of cloth. Much firmer than a harness, and producing the
largest possible pattern with the least possible cards, mounting, and
machinery it took up its position in the days of the drawboy, and has
held its own, with the aid of the jacquard, against all the inventions
of modern times. The principal advantage of this system of working
is that a much larger pattern can be produced with the same quantity
of cards, mounting, and machinery, than by any other means; but the
pattern wrought by a pressure harness will not compare with one wrought
by a full harness for fineness of outline or detail. When weaving large
tablecloths with 100 or 120 threads of warp per inch, and from 50 to
72, or even 90 in. in the single pattern--that is, when the pattern
is all, or nearly all, single--the amount of machinery that would be
required to work it on a full harness could not be crowded on a loom,
the French system being the only one by which it might be attempted;
and even then the difficulty of keeping the machinery and cards in
perfect working order would be very great, not to mention the cost of
mounting, patterns, cards, &c. The quantity of harness and machinery
is reduced in a pressure harness by drawing two or more warp threads
into each mail in the harness. Suppose we take 60 in. of cloth with
100 threads per inch = 6,000 threads, and allow 3 threads to each mail
in the harness, or each hook of the jacquard, 2,000 hooks, or four 500
jacquards, would be required to work the pattern. As many as six 600
jacquards are sometimes required on this description of work when very
fine; but three or four 500 or 600 machines are more commonly used on
either hand or power looms. To work one of these finest patterns in
a full harness 10,800 hooks would be required--say, six 600 machines
with 3 threads to each mail = 10,800; and almost double this number
of cards, even with working them backwards and forwards, would be
necessary.

In early times, when the drawboy took the place of the jacquard, 6,
8, or 10 threads were put to each mail, or went to what would now be
one hook of the jacquard; and patterns were not usually so extensive
as they are now, so that the cords of the harness were greatly reduced
in number, but with a proportionately coarser effect of pattern. The
difference between a pattern wrought with a full harness and one
wrought with the pressure or any of the twilling harnesses, is that
the outline is clear and defined in the full harness, and the detail
and points of the figures can be turned on a single thread, whereas
in the others the edges of the figures will be jagged or in steps,
and the points must turn on whatever number of threads are lifted
together. The pattern on the cloth must therefore have somewhat of the
rough, square effect of the design on point paper, though, of course,
reduced in size, but will be worse in this respect--viz. that whereas
the edges of the figure on the point paper are clear and well defined,
on the cloth they are not so, the rough edges to some extent blending
the figure into the ground, and not giving the clean, sharp effect of
a full-harness pattern. Shaded effects are also coarse on a pressure
harness, but can be made effective if broadly treated. Cloth with 100
or 120 threads per inch does very well to have three threads to the
mail, and from 80 to 100 threads per inch suits very well for two to
the mail. Any coarser set than 80 threads per inch requires to be woven
in a full harness to produce good work, and for superior work nothing
less than 100 threads per inch should have two threads to the mail,
though 80 per inch does very well.

When several threads are put to the mail, it is also usual to put
several picks to each card. The fewer picks, the finer will be the
pattern; but a good method of regulating this is to make the checks
formed on the cloth square, a little more or less according to the
fineness of the pattern required. Thus, if the cloth is wefted square,
or a little over that--say, 100 warp by 100 to 110 weft threads per
inch--paint the design on, say, 8 × 8 or 8 × 9 paper, and give as many
shots to the cards as there are threads in the mail. If the cloth is
to be wefted one-half over square (100 × 150) the same pattern will
still do, but with half as many more shots to the card than there are
threads in the mail. If there are two threads in the mail there will
be three picks to the card, but if there are three threads to the mail
there must be four picks to one card and five to the next one. This
would be for a pattern on 8 × 8, or square paper. If painted on 8 × 9
paper, which would give an extra card to every eight, and if the weft
must not be increased, then four or five picks must be taken off the
number given to the eight cards and put to the ninth one, for four
cards with four picks to each and four with five picks to each = 36
picks, and thirty-six picks put to nine cards would allow four picks
to each. This would make the edges of the pattern a little finer; and
if the design was painted on 8 × 10 paper it would be finer still, as
there would be more cards to a given number of picks.

In this way any alteration required can be made on the number of picks
per inch given to cloth woven on a pressure or twilling harness,
without distorting the pattern by varying the number of picks given to
each card to suit the shotting. Neither is it necessary to have the
same number of threads in each mail; the warp might be mailed 2’s and
3’s or 3’s and 4’s, but the more regular they are, the better. If the
fineness of the cloth requires to be altered, it may be woven in the
same harness without any alteration by varying the number of threads
in the mails. For instance, a warp of ninety threads per inch mailed
3’s and one with 120 threads per inch mailed 4’s would work in the
same harness. Similarly, the same set of cards would suit for making
different widths of cloth by making the harness narrower in the cumber
board and altering the mailing so as to keep the cloth the same set, or
it may be made a finer set and not alter the mailing. In either case
the pattern would be reduced in size.

Fig. 111 shows a portion of a pressure-harness mounting which is
similar to that of the draw loom. It is mounted in the same manner
as described for full harness, only that the warp must be divided by
the number of threads to be drawn into each mail in order to find
the quantity of harness required. The kinds of mails used are shown
at A and B, Fig. 111; it does not do to have more than two threads
drawn into each eye of the mail, as they are liable to twist round
each other; two will separate easily, but a greater number will not.
Ordinary full-harness mails answer very well for a two-thread harness.
The mails are levelled in the same position as for full-harness work,
viz. about 1-1/2 in. below the level of the back and front beams for
hand-loom work, and a little lower for power looms. The lingoes are
heavier than those required for a full harness; the weight depends upon
the strength of the yarn and the number of threads in the mail; 10 to
12 to 1 lb. are used in power looms for linen damask with two threads
in the mail, having about eighty to one hundred threads of warp per
inch. For hand looms they are usually made of lead, thicker and shorter
than the wire lingoes, and called ‘leads’; 11 to 15 per lb. suits for
three- or four-thread harnesses. Light cotton work, such as muslin
curtains, only requires lingoes of about sixty or seventy to 1 lb. for
hand looms, but from twenty-five to thirty per lb. are used in power
looms for two-thread harnesses.

[Illustration: Fig. 111]

The principle of working the pressure harness is best explained by the
diaper mounting shown in Fig. 2, where a back set of heddles takes the
place of the harness, and a set of ground leaves of long-eyed heddles
stands in front of the back ones. After the yarn is drawn into the
harness or back mounting in the usual way, it must be drawn into the
front or presser heddles, using, generally, a straight draught for a
satin ground. If Figs. 2 and 111 are examined, it will be seen that
when the shed is opened by the harness or back mounting, the heddles in
front sink part of the raised warp and raise part of the sunk warp. The
mounting raises all the warp of the figuring portion on each card or
line of the design paper, and leaves down all the ground warp, so that
it is necessary to have the heddles in front to bind the warp and weft,
or form the texture of the cloth. For an 8-leaf satin eight shafts are
required; one of these must be raised and another sunk for each shot,
so as to raise the binding threads of the sunk warp and sink those
of the raised warp, the harness forming the outline of the pattern,
or raising the warp of it in a mass. The other six leaves of heddles
stand in a middle position, and the long eyes allow the warp drawn by
the harness to rise. When the machine is drawn it is held up till as
many shots as are to be given to the card are thrown in, but the shed
formed by the heddles must be changed for each shot. In working bars
up the cloth one card would be sufficient, and when the machine is
drawn it would be held so, and the cloth wrought with the heddles; of
course, no machine would be required in this case, one leaf of heddles
with the stripes of warp drawn into it would be sufficient for the back
mounting. For dices two leaves of heddles would be sufficient for the
back mounting, the warp of one dice to be drawn on one leaf, and that
of the other on the other leaf; then one leaf would be raised and held
up till one dice was wrought, when it would be lowered, and the other
one raised and held up for the other dice. For fancy dices and diapers
the plan of mounting in Fig. 2 is very suitable and simple, but for a
variety of figuring or flowering the jacquard is necessary.

It will be seen that the presser heddles have three positions, viz. a
sunk, a raised, and a middle position. The length of the eyes is to
allow the harness to open the shed when the heddles are stationary,
or in their middle position. They must be a little longer than is
required to open the shed at the back leaf of the heddles; for a
2-in. shed a 2-1/4-in. to 2-1/2-in. eye is used. When the heddles are
stationary the lower loop of the eye should be fully 1/8 in. under
the sunk warp, and there should be the same clearance at the top when
the shed is drawn; some allow more. The shed for a pressure harness
is usually very small in front of the reed, and requires a very small
shuttle to be used, from 1/2 in. to 1 in. deep being the usual sizes.
The depth of the shed that can be made depends principally on the
elasticity of the yarn. With a linen warp a very small shed can be
made, as the yarn has but little elasticity, and if overstrained will
hang slack. For it the distance between the harness and the back shaft
of the front mounting should be 10 in. to 12 in., and there should
be a stretch of 27 in. to 34 in. behind the harness. The draw of the
harness may then be 3 in. to 3-1/2 in., and the shed at the back shaft
will be 1-3/4 in. to 2 in.; this will allow a shuttle of 3/4 in. to 7/8
in. deep to be used. For hand looms the shed is about 1-1/4 in. at the
back shaft, and a shuttle of 1/2 in. deep is employed. The shed must
be made very clear and regular, and the smaller it can be kept, the
better. With a good cotton warp 7 in. is a sufficient distance to have
between the harness and heddles, and will admit of a larger shed being
formed; but it is not desirable to have too large a shed, as there is
a considerable strain on the yarn, and a small, clear shed is more
satisfactory. This would be assisted by bringing the harness as close
as possible to the heddles.

One drawback to this method of working is the distance which separates
the harness from the fell of the cloth, or even from the reed; and
if there is any obstruction to the warp rising or falling, such as
roughness in the heddles or reed, or lumps on the yarn, it will not,
unless very tight, fall into its proper place, and the shuttle may pass
over or under it when it should not do so, giving a picked or darned
effect to the cloth; slack threads may cause the same.

The warp must be kept as tight as possible, and all the threads should
be at a uniform tension, the heddles straining each thread alike; the
harness should be as close to the heddles as the yarn will permit it
to be. The space occupied by the harness, heddles, and traverse of
lay should be no greater than is necessary; then, with a small but
clear shed satisfactory work can be produced. When drawing a warp
into the harness, a boy or girl sits behind, and hands the threads to
the drawer-in, who takes them into the mails with a wire hook, and
then either hands them to a second drawer-in sitting in front of the
heddles, to be drawn into them, or passes them over and under a pair of
rods tied across the harness so as to form a lease as she draws them
in; and when she has all drawn into the harness, she begins to draw
them into the heddles, a straight draught being mostly used for a twill
or satin.

[Illustration: Fig. 112]

The front mounting of a damask hand loom is shown in Fig. 112. A, A
are the shafts, four in number, but eight are generally used for an
8-leaf satin; B, B are the jacks; C the lams, which require to be one
more in number than the leaves of heddles; D the treadles; E, E^1 the
upper marches or jacks; F, F^1 two sets of coupers or levers, with
their fulcrums at _f_, and loaded at the outer ends with the weights G.
Under the ends of these levers is a bar N, to which the weights draw
them, and keep the heddles up to their middle position. The cords I are
not fastened to the heddles, but pass down through them to the lams
C. The cords L connect the coupers with the heddle shafts. Each lam
has two cords tied to it, except the two outer ones--_i.e._ the first
and last one--which together act as one, to avoid crossing the cords.
One of the cords I is tied to each lam, and also one from the jacks B;
but the first and last lams have only one cord tied to each of them,
one having the cord I, and the other that from B. Both these lams are
connected with one treadle, and the others are each connected with a
treadle. Of course, the cording is made in the usual way, agreeably to
the pattern, two methods of twilling being shown in Fig. 113. When the
weaver presses down a treadle, one leaf of heddles is raised by the
cord I connected from one of the levers E, E^1 to one of the marches
or lams C, and one leaf is at the same time sunk by the cord connected
from one of the jacks B to the lam connected to the treadle. The other
treadles act similarly when corded for a twill or satin.

[Illustration: Fig. 113]

The cording generally used for an 8-shaft satin is shown at A, Fig.
113. Sometimes the twill is run in the reverse direction. In either
case it will be observed that the twill on both ground and figure run
in the same direction, which makes one a sateen, or coarse twill, and
the other a satin, or fine twill. In order to have both twills alike
they require to be run in the opposite direction on the cloth, as shown
at B, which will give a fine twill on both ground and figure, on both
sides of the cloth, with single yarn. This does not hold good with
every twill. When weaving, the weaver presses down the treadle which is
connected with the jacquard (another treadle being required for this
purpose) with his left foot, then works over the twilling treadles with
his right foot, holding down the machine treadle till he gives as many
shots to the cards as are required. When the card is to be changed the
weaver lets down the machine and draws another shed, striking up the
weft again without throwing in a shot or taking his right foot off the
treadle. This clears up the shed, and makes the yarn steady before he
springs another shed with the presser leaves, giving a regularity and
firmness to the work which it is impossible to get otherwise, but which
has to be done without in the power loom. The weft is struck up on the
open shot, or before the heddles have closed the shed. Sometimes a sort
of dobby is used below the lams, which enables the weaver to work the
heddles with one treadle. In power looms a similar principle of front
mounting is sometimes adopted, substituting a wyper tappet for the
weaver’s foot; but a better plan, as it avoids having the cords passing
through the yarn and heddles, is to have a box tappet, on the Woodcroft
style, made with solid plates, and struck to give the rising and
falling motion. The connections from the tappet to the heddle shafts
are made in the same way as the ordinary Woodcroft tappet. The jacquard
may either be a single or double-lift one. Some prefer the one, some
the other. The single-lift is more easily fitted up, especially if
the number of shots to the card varies. When the shots on each card
are alike, perhaps the steadiest method of lifting the machine is to
have a box tappet struck one up and three down, or one up and two
down, according to the number of shots on the card. This tappet acts
on a bowl on a treadle, to which the rod for raising the machine is
connected. For a double-acting machine there must be two treadles and
tappets acting alternately. The cylinder can be driven from the rising
and falling of the machine with a swan-neck or lever motion, or may be
driven from an eccentric on the loom with a pinion on the crank shaft
turning it one to three or four shots, as may be desired. When the
shotting to the card is irregular there are several methods adopted for
lifting the machine griffes, one of the best of which is shown in Fig.
114 for a single-lift machine.

A is the treadle to which the connecting-rod from the machine is
fastened by a bolt through the slot at H, or the slot may be in any
desirable place. B is a rack in which the end of the lever works, which
keeps the treadle bowl steady to the tappet. E is the tappet on the
tappet shaft of the loom, and is made so as to act at every shot. D
is the fulcrum of the lever, and C is the stand, which is bolted to
the ground and fastened to the side of the loom. This portion of the
motion working alone would raise the griffe for every shot, the same
as would be required for a full-harness, single-acting jacquard; but
when the griffe is raised by the tappet E, it can be held up as long as
is desired by letting the bell-crank catch F fall in over it as shown.
When the treadle A is in this position the tappet merely touches the
bowl, depressing it about 1/8 in., so as to clear it off the catch;
this is to allow the catch to be easily pushed off when it is required
to let the treadle up, or to drop the griffe. The catch is moved by the
cam or tappet G acting on the bell-crank F. This tappet can be driven
by a pinion on the crank shaft or by a catch on the slay. It may be a
tappet struck to suit, or a barrel with a set of lags or pins on it, so
that the machine griffe may be raised and lowered in any order that is
desired.

[Illustration: Fig. 114]

An ingenious method of working the heddles of a pressure harness by the
jacquard was invented in Bethnal Green shortly after the introduction
of the original machine. It was used for weaving the richest silk
damask, which had 400 threads of warp per inch and about one-fifth
that number of weft shots, so that five threads of warp might be drawn
into each mail without making the pattern appear any coarser in the
warp than in the weft. A sketch of this mounting is given in Fig 115.
A shows the hooks for working the heddles, and B those for working the
harness. Eight hooks are given for working four leaves of heddles. It
will be seen that a cord from two hooks passes round one of the pulleys
C, and each of the heddle shafts is attached to one of these pulleys.
These hooks may be raised by the griffe of the jacquard, which would
require to rise and fall for every shot, or the griffe may be held
up for the number of shots to the card, and the hooks for working the
shafts may be wrought by a small dobby. When one of each pair of the
hooks A is raised, the heddle shaft connected with it is raised to
the middle position, as shown by the pulleys 1 and 2, and the leaves
connected with them. If neither of the hooks is raised for any shaft,
it will remain sunk, as shown in No. 3; and if both the hooks are
raised, the leaf will be raised to its top position, as No. 4. The
leaves are drawn down by weights or springs.

[Illustration: Fig. 115]


TWILLING JACQUARDS

[Illustration: Fig. 116]

As has been stated before, one of the drawbacks to a pressure harness
is the strain which the warp has to undergo when forming the shed. This
necessitates having a good warp, which adds to the cost of the cloth,
and in low-class goods this consideration may hinder the sale. Many
methods have been adopted for working the ground of the cloth without
using a pressure mounting. Some of these are explained under ‘half
harness’ and ‘split harness,’ but none of these methods would produce
cloth like the pressure harness. A twilling jacquard to act similar
to the pressure harness, but without using the front mounting, or by
dispensing with the leaves of heddles, was patented by Mr. Shields,
of Perth, in 1859. This machine underwent several improvements, and
now there are two varieties of it in use, one known as the Irish or
Bessbrook machine, being patented by Mr. Barcroft, of the Bessbrook
Spinning Company Limited, county Armagh; the other as the Scotch
machine, the improvements being made by Mr. Shields and others. The
principal difference between the two machines is that the blades or
knives of the griffe have a horizontal or sliding motion in the Scotch
machine to enable them to get clear of the heads of the hooks, whereas
in the Bessbrook machine they turn out of the way or partly revolve.
A full description of the Bessbrook machine is here given. Fig. 116
is a view of the framing of the machine. K is the cylinder, which may
be wrought by the swan-neck motion, as shown, but it is better to be
wrought by a separate motion from the loom. D is a cord attached to
the handle of the shears for reversing the cylinder; E is a brass
bushing through which a shaft passes for raising the griffe. The
shedding of this machine is exactly the same as that of an ordinary
single-acting jacquard. The griffe rises and falls for every shot;
the cylinder travels out and in, but does not turn till two or three
or whatever number of impressions required are given by each card. To
prevent the cylinder turning it is only necessary to raise the shears
so that they will not catch it. This is done by means of the tappet A,
shown in Figs. 116 and 117. The roller H on the shears rests on the
tappet, which is turned by means of the rack wheel, which is fast on
it, and the lever L, which is loose on the stud. This lever is pushed
backwards and drawn forwards by the rod I, which is connected to an
arm, either on the shaft for raising the griffe, or on one for working
the cylinder, and the catch on the lever takes a tooth of the ratchet
wheel at each draw, so that with four divisions on the tappet, and 12
teeth in the ratchet, three shots would be given to each card. If it is
required to work the cylinder the reverse way, it is only necessary to
tie the cord D on the end of the shears to a spiral spring made fast to
any convenient part of the loom, then the roller C will be acted upon
by the underside of the tappet, and the under hook of the shears will
turn the cylinder when required. B is a spring acting as a pawl to keep
the ratchet from moving backwards, and there is also a spring shown
over the catch which moves it forward; both these springs serve to keep
the tappet steady. There are two sets of hooks and needles in this
machine--the ordinary set for working the harness, as shown at A (Fig.
118), and a set of twilling hooks, one row at each side of the machine,
which are much stronger than the others. The ends of the twilling hooks
are hooked round bars, which pass through the loops at the lower ends
of the ordinary hooks, as shown at C (Fig. 118). These bars are kept in
their places by a grid D, which is between the ordinary hooks and the
twilling hooks. Each of the needles of the jacquard is connected with
two or more of the ordinary hooks, as shown at A. The twilling hooks
have also needles on them, but they are only for the purpose of keeping
them in their places, and are not acted upon by the cards. The springs
of these needles are on the ends next the cards, or at the back of the
faceplate or needle, as shown at E. The centre support for the knives
of the griffe is shown at F, with an end view of two knives G G. The
ends of the knives enter the griffe frame at each side, so as to allow
them to oscillate or partly revolve. This motion is given to them by a
set of bars or flat needles, as shown at A (Fig. 119). A single needle
is given at A^1. These needles have each the notch in them over the
top edge of two or more of the knives, and are acted upon by a barrel
C with studs in it, set to the twill. This barrel rises and falls with
the griffe, and is turned from one row of pegs or studs to the next one
each time the griffe falls by the head or lantern on it coming down on
a finger, as shown at A (Fig 120). B in the same figure is a strong
spring to keep the barrel steady and make it turn the correct distance.

[Illustration: Fig. 117]

The number of knives in the griffe must be regulated to suit the twill
to be put on the cloth; they must be a multiple of the twill, and this
to some extent regulates the number of needles that must be in each
upright row. For example, an 8-leaf twill may have 16 or 24 knives,
which would be twice or three times over the twill. If there are 8 rows
of needles to 16 knives, or 16 rows of hooks, that would be 2 hooks to
each needle, or if there are 24 rows of hooks there must be 3 hooks to
the needle; but if only 2 hooks to the needle are required, there must
be 12 rows of needles to 24 rows of hooks. The same principle holds
good for a 5-leaf, or any twill. Eight or 12 rows of needles would
not be suitable for a 5-leaf twill, neither would 10 rows of needles
be suitable for an 8-leaf twill, if the same number of hooks have to
be connected with each needle; but 10 rows of needles with 20 rows of
hooks or 20 knives will answer for a 5-leaf twill with 2 hooks to the
needle, or would answer for an 8-leaf twill with 24 rows of hooks, 4 of
the needles to have 3 hooks to each, and 6 of them to have 2 hooks to
each.

[Illustration: Fig. 118]

Each of the twilling needles, A, Fig. 119, must be connected with as
many of the knives as there are repeats of the twill; thus, for 24 with
an 8-leaf twill, the 1st, 9th, and 17th knives would be acted upon by
the one needle, and so on with the others.

Now as to the action of the machine. When the pattern card presses upon
the needles the griffe begins to rise, and when rising it must lift all
the hooks required for the pattern except 1/8th part of them which
must be left down (in an 8-leaf twill) to form the binding or texture
of the cloth. In addition to this, 1/8th part of the ground warp must
be raised for the weft to pass under it and form the ground texture.
This is all accomplished by the one rising of the griffe. When the
griffe is down, one of the twilling needles, A, Fig. 119, is pressed
back by a peg or stud in the barrel, which causes the knives connected
with this needle to turn out of the way of the heads of the hooks, so
that when the griffe is rising these knives will pass clear of them,
leaving every eighth row of hooks down to form the binding of the
raised or pattern warp. In the Scotch machines the knives slide back
instead of revolving. As shown at H, Fig. 118, there are projecting
pieces of brass fastened on the knives between the twilling hooks.
These hooks stand clear of the knives, but when the latter revolve the
projections on them push the hooks forward to the next knife, which
lifts them when the griffe is rising. The hooks draw up the bars to
which their lower ends are looped, and raise the rows of ordinary
hooks through which the bars pass, thus forming the twill on the ground
in the same way, but in the reverse direction, that the knives passing
clear of the adjoining rows of hooks form the twill on the figure.

[Illustration: Fig. 119]

[Illustration: Fig. 120]

The working of the texture requires the griffe to fall for every shot,
which would be otherwise unnecessary, and the card must come in against
the needles each time to push off the hooks that are not to be raised.
This causes wear and tear, which cannot be avoided with this machine,
but they work very well, though they are not by any means perfect.
There is a good deal of friction on the needles, which causes them to
wear quickly; but being so much easier on the warp than the pressure
harness, and more easily managed, they are extensively used for large
patterns in the fine linen damask trade. They will not make so firm a
cloth as the pressure harness, and have the objection that all twills
formed with the harness have when there is a gathered tie--viz. that a
portion of the cloth will have the twill running in one direction and a
portion in the reverse direction.

Like all single-acting jacquards, these machines have no counterpoise
in themselves, and being very heavy require one added to assist the
loom to raise them. Sometimes this is accomplished by means of a
carriage spring placed on a beam or on the top rail of the loom frame;
and an arm from the shaft, which raises the griffe resting on the
spring, will form a sufficient counterpoise, the spring being made
as strong as is required for the purpose. Unless the springs, which
are made similar to those used for carriages, are nicely tempered,
and the different pieces made so as to slide freely on each other,
they are liable to snap when the loom is running quickly. The griffe
is generally driven from the fly-wheel on the crankshaft in the same
manner as for ordinary single-acting jacquards. It is, however, a
better plan to drive it from a crank on a stud wheel gearing into
the tappet-shaft wheel. This does away with the necessity for a
counterpoise unless the machinery is very heavy, in which case a
few strong spiral springs will suffice. The horizontal shaft for
raising the griffe, or griffes if two or more machines are used, must
be very strong, so as not to twist with the tortuous strain, which
is very great. Three of these machines, with 500 or 600 needles to
each, are required for tablecloth mountings when tied up for centred
patterns--one machine to work the border, one the centre, and one the
portion between the border and the centre; or two machines to work the
double portion, and one the single portion. Fig. 121 gives a view of
a mounting tied up in this way, Nos. 1 and 2 machines being double
mounting and No. 3 single.

[Illustration: Fig. 121]

_The Karl Wein Jacquard._--This machine is a twilling jacquard on the
same principle as that last described, but performs its work in a more
scientific manner. It was patented by Messrs. J. Tschörner and K. Wein,
Kesmark, Hungary, and was first introduced into this country at the
Glasgow Exhibition of 1888.

The following is a description of the machine exhibited there: The
principal feature of this machine is perhaps that each knife acts
independently in a grid, and is wrought by a tappet at the side of the
loom, so that any row of hooks can be raised or lowered at pleasure,
and this without any change of card. Fig. 122 gives an end view of
the framing, showing the upper and lower grids in which the knives or
lifters slide. An end view of four of the lifters is given at A, 1 and
3 belonging to the top set, which work in the upper grid, and 2 and 4
to the bottom set, which work in the lower grid.

[Illustration: Fig. 122]

A perspective view of eight of these lifters is given at B C, Fig.
123, with upright slide bars D attached to them, which slide in the
rack or frame E. These bars are for the purpose of making the lifters
rise steadily and horizontally, as (which may be seen) they are not
all lifted at their centre. Only one connection is fair in the centre,
which would no doubt be an objection with a heavy harness to lift; but
this arrangement is made to suit for the machine sitting across the
loom, for a London tie, and some method of making vertical connections
from the levers to the lifters is necessary. The above does very well
for a narrow harness if light.

[Illustration: Fig. 123]

[Illustration: Fig. 124]

If the machine was fixed on the loom for a Norwich tie, or with the
cards to hang over the back, then the connections from the levers might
all be at the centres of the lifters and the bars D would only be
required to steady them. The levers for raising the lifters are shown
at F, with their fulcrum at I. G is a spring, one of which is attached
to each lever to keep it down, as the tappet has not a positive
rising and sinking motion, only raising the levers and allowing their
own weight and the draw of the springs to recover them. One of the
connecting rods from the lever to the tappet is shown at I. Fig. 124
gives the principle of this tappet. A is the lever or treadle with
its fulcrum at A; the connecting rod B connects the point of it to
one of the top levers F (Fig. 123), and C is a bowl at the other end
of it which travels on the tappet plate D. The tappet is made up of
17 plates struck to suit; 16 of these are for working the lifters and
1 for the cylinder; it sits at the side of the loom like a Woodcroft
tappet. For an 8-end satin twill 16 levers are required, 8 for the
bottom set of lifters and 8 for the top set. Fig. 123A shows two
lifters B^1 and C^1 with hooks and needles. A hook and a needle on a
larger scale are shown in Fig. 125. The hooks are flat pieces of iron
cut to the shape shown, with small projecting pieces, as _a_, riveted
to them, by which they can rest on the bottom lifting knives, as at
C^1 (Fig. 123A), so that when any of these knives are lifted a row of
hooks will be raised by them. As before explained in reference to the
Bessbrook machine, it is necessary when the pattern card presses on the
needles for all the knives except one to act, and also for one of the
bottom lifters to rise to form the texture of the cloth; the same must
be in this machine, and can easily be regulated by the tappet, which
should hold up the warp required for the pattern and work the ground
texture as well. This is done in the following manner: Suppose there
are 16 lifters, 8 top and 8 bottom ones. When the card presses on the
needles, 7 of the top and 1 of the bottom lifters should be raised by
the tappet, and the top lifter, which is left down, should be raised
so far as to catch the heads of those hooks in the row belonging to it
that are not pushed back by the card. For the next shot this lifter is
raised and another one let down, but it is not let down far enough for
the hooks to get off it, stopping at the same height that the first
one was raised to, which is easily regulated by the tappet. If there
are 3 shots to the card, another similar change takes place; then for
the fourth shot all the lifters are dropped, the cylinder presses
in, and all but one of the top lifters and one of the lower ones are
again raised, and the twill proceeded with as before. It will thus
be seen that the shedding for the twill acts as in a double-acting
jacquard, but the lift at the change of card is similar to that of a
single-acting jacquard. Of course the tappet must be struck to change
the lifters according to the twill required on the cloth. In the
Bessbrook machine there could be 16, 24, or 32 rows of hooks to 8 rows
of needles by having 2, 3, or 4 hooks to each needle, or the number
of hooks to each needle might vary and any number of needles might be
used. There must be a knife for each row of hooks, but all the knives
belong to the one griffe. In the Karl Wein machine there must be 2
lifters for each row of hooks, and 16 of them are enough to have for
convenience. In the machine exhibited, 16 hooks, or 2 rows of 8, were
attached to 6 needles, 4 needles having 3 hooks to each and 2 needles
2 hooks to each, as shown at B (Fig. 125), the thick vertical lines
representing 1 row of hooks, and the thin lines the next row. There
were 6 rows of needles in the needle board, and the point of each
needle was cranked as shown at C; each row of holes for the needles in
the needle board, or face-plate, stood between 2 rows of hooks: the
first 3 needles were connected with 1 row of hooks and the second 3
with the next row, the cranks of the first and second set of needles
being turned in the opposite direction so as to enable them to fall in
with the rows of hooks. Of course there might as well have been 8 rows
of needles with 2 hooks to each, or 4 needles to each row of hooks.

[Illustration: Fig. 125]

If necessary to use more than one machine, some arrangement would
require to be made for lifting them. It might be done by using a double
set of levers with connecting rods similar to those used for twilling
looms.

The twilling of the ground might be wrought by a griffe and hooks as
in the Bessbrook machine, thus doing away with seven treadles and
seven tappet plates, but this would make the twilling a single-acting
shedding motion.

Since this was written the machine has been altered so as to make it
more suitable to the requirements of the work in this country, but
does not appear to gain favour, and this class of weaving is not very
extensively used.




CHAPTER VII

_GAUZE_


The principle of weaving gauze with shafts and healds applies equally
to making it with a harness, though perhaps the latter is simpler,
as there is not the same trouble in the arranging of drafts; and, in
intricate work, the number of doup leaves required for shaft mountings
adds considerably to the trouble of the weaver and tackler.

The simplest, or in any case the most ready, method of weaving gauze in
a harness, is to use an ordinary harness, and form the gauze by means
of one or more doups and standards in front of it. Suppose we have have
a 400 machine with the harness tied up in simple repeats of 400; any
desired pattern can be put upon the cloth to the extent of 400 hooks of
the jacquard as in ordinary harness work, but it cannot be made of any
texture desired--it must be made to suit the working of the doups in
front of the harness. If the open work is to be a plain gauze with one
shot into each shed, then the pattern must be plain cloth, as the doup
standard must be raised and sunk alternately for each shot; but any
desired form or figure can be made in plain texture on the 400 hooks,
and the cloth will consist of a plain figure on a gauze ground; or it
might as well be a gauze figure on a plain ground.

Those not acquainted with figured gauze work, have considerable
difficulty in understanding how to make plain or gauze as required on
the cloth.

It is easy to understand that working the harness without the doups
can make plain cloth, and also that working the doups and standards,
and portion of the harness, for alternate shots, can make gauze; but
the difficulty is to thoroughly comprehend how, when working gauze,
to neutralise the crossing of that portion of the warp that is to
form the plain figure. There are two ways of neutralising the crossing
effect of the doups on any thread or number of threads that may be
required for the figure: first, by not raising the mail or mails
carrying the crossing threads, or threads in the doups, but forming
the plain by raising the mails carrying the threads not in the doups,
alternately with the doup standard; and, secondly, if the threads not
in the doups are raised along with the doup standard, no crossing can
take place; but this would not suit for working plain cloth, as these
mails must be down when the doup standard is up in order to form plain
cloth.

[Illustration: Fig. 126]

If a different gauze is required, say one with 2 shots into each shed,
or between the crossings, then the solid texture might be a rib or mat;
or if it suited the pattern, as in the case of dice work, two doups
could be employed, so that one could rest when the other is working
the gauze. Fig. 126 is a plain gauze and plain texture, common dice
pattern, which could be wrought with shafts, but shows the principle
of working with the harness as well. This could be wrought with one
doup in front of the harness. To work the dice A the mails carrying
the threads not in the doups would be raised alternately with the doup
standard, and thus form plain cloth; and at the same time, to make
the dice B a gauze, the mails carrying the threads in the doups would
be wrought alternately with the doup standard. For the dices C and D,
this would be reversed. In case of the gauze having to be made with,
say, 3 shots into each shed, and 2 threads twisting round 2 to give
openness of texture, it will be seen that it could only be wrought with
one doup, provided 3 shots went to each shed of the solid dice, as the
doup would require to be up for 3 shots in succession, and down for the
same. If plain cloth were required it would be necessary to have a
doup and standard for each dice, so that one could remain down when the
other is raised for gauzing; this would allow the harness to work any
texture on the solid dice, the doup of which is not gauzing. If it were
required to work a diagonal stripe of plain and gauze, as shown in Fig.
127, with 2 weft shots coming in between each crossing of the gauze,
6 doups would be required in front of the harness, whereas if only
one shot came between the crossings only one doup would be required.
It will be seen that there are 6 splitfuls of warp to one repeat of
the pattern, and that the change from the plain to the gauze is at a
different time for each; therefore, on account of the doup standards
having to be raised for 2 shots in succession when a portion of the
cloth is plain texture, it is necessary for each splitful of warp to
have its own doup so that it can be crossed without interfering with
the plain texture in the other portion of the cloth.

[Illustration: Fig. 127]

This pattern would be better suited for a shaft mounting than for a
harness with doups in front, but could be wrought with a gauze harness
more easily than with either. The advantage of using a harness with
doups in front, in this case, over a shaft mounting, would be that the
plain stripe could be figured if desired, and the gauze stripe might go
in steps of two or three splitfuls alike, which would give much bolder
stripes.

When more than one or two sets of doups would be required to be used in
front of a harness, it is advisable to adopt a true gauze harness, as
the doup shafts come in the way of the weaver. A gauze harness has the
doups in the harness, so that they can be lifted independently, which
admits of indiscriminate figuring on either the gauze or solid portion
of the cloth.

When using doups in front of the harness it is not necessary that the
plain or solid cloth should be all a plain or unfigured texture; any
suitable warp figure may be formed on the plain by the harness, but
weft figures cannot be made on the upper side of the cloth, because,
though the harness can be raised for any number of shots in succession
to form a figure by the warp, having it sunk would not in the same way
form a weft figure, as in crossing the gauze a portion of the yarn
would be raised by the doup standard, which cannot be prevented, unless
by having more doups than one. A harness with doups in front is better
suited for working stripes than any other class of work, and it is in
this class of work that gauze can be made most effective.

When figuring with gauze, unless several shots are thrown into each
shed, or between the crossings of the gauze threads, and 3 or more
warp threads are twisted together, the openness of the texture is not
sufficient to make a good contrast between the plain and gauze unless
the fabric is very light; and even with 2 threads round 2, and throwing
4 to 6 shots into each twist, the contrast is not so great as might
be desired when the cloth is heavy. A portion of the weft, say every
second or third shot, might be made to pass either over or under the
gauzed texture and interwoven only with the solid portion of the cloth,
the flushed or loose weft to be cropped off afterwards. In stripes, not
only can this be done, but the warp may be set much thinner or wider
in the reed for the gauze: say every second or third split may be left
empty; or it may be of a different description of yarn--a fancy twist
if desired, or in threads of different colours. A fancy stripe with
three threads to the split in a reed, set 30 splits per inch, and every
second split of the gauze portion left empty, makes a very effective
pattern for fine work, with say 80 picks per inch; or 120 picks might
be used, and every third or fourth one not wrought into the gauze, but
flushed over to be cut off afterwards. There might be 2 thick threads
or 4 finer ones to each split of the gauze, and 4 to 6 shots into each
crossing, this, of course, to be regulated to suit circumstances and
taste; one thing must, however, be remembered--viz. that when made too
bold or open the warp ribs are liable to slide on the weft, and this
is the chief defect in gauze when made open, especially when a smooth,
clean yarn, as linen, is used.

[Illustration: Fig. 128]

[Illustration: Fig. 129]

[Illustration: Fig. 130]

When working stripes, any pattern, or any texture or style of cloth,
whether single, double, or pile work, may be used for the solid
stripes, or the whole cloth may consist of stripes made of combinations
of gauze and solid textures. One set of stripes may be wrought quite
independently of the other by having the yarn on separate beams, and
when only one set of stripes is gauzed, the others are not drawn into
the doup heddles, so are quite independent of them. The doup heddles
may be spaced or otherwise as desired. When working with doups in front
of the harness, there must be a slackener for the crossing threads for
each doup standard, same as for heddle work. Figs. 128, 129, and 130
show some of the methods of slackening in use. In Figs. 128 and 129,
A is the yarn beam, B the back rail, C and D the vibrating bars or
slackeners, C acts as a fulcrum; and in Fig. 128 the bar D presses the
crossing yarn down and is raised by the tappet E, which may be either
above or below a lever on the end of the bar C, a spring on the lever
regulating it, to hold it against the tappet. In Fig. 129 the bar D is
sunk to relieve the crossing yarn, by drawing up the cord G, which is
attached to the jacquard or dobby, the spring H holding the lever E
down when it is relieved. Fig. 130 is similar to Fig. 129, but for two
doups; it is shown both in perspective and in end view, and will be
clearly understood. C and D must be set at such a distance above each
other as will permit of the traverse of D without C being moved. Two or
more shafts of heddles might also be used which could hold the crossing
yarn either up or down, same as the bars, till they are acted on by the
jacquard or by a tappet.

[Illustration: Fig. 131]

[Illustration: Fig. 132]

_Designing._--When designing a pattern for gauze, with doups in front
of the harness, it is only necessary to set off a few hooks of the
jacquard to work the doups and slackener, and design the fancy portion
on the remaining hooks; or if the gauze and slackener are wrought
with tappets the full jacquard will be available for the figured
portion, the same as if no gauze were being made. If the doups and
slackener are to be wrought by the jacquard the pattern can be painted
on the remaining hooks, and the dots for working these can be added
afterwards, agreeably to the texture of the gauze required. The doup
standard and slackener should act together, and the doup or half leaf
should rise when the harness carrying the crossing threads rises. If
the doup standard is to be raised every second shot, as is required
when plain cloth has to be wrought with one doup, it might be wrought
by a tappet, and the doup or loose heald might be raised by the machine
for every shot, as it must rise with its standard as well as when the
harness raises any of the yarn drawn into the doup. The doup shaft is
best held down by a light spring. Indiarubber about 3/16 or 1/4 in.
in diameter suits very well. One spring at each side of either one or
two ply of this should be sufficient. If the number of rows in the
harness is such that it will correspond with the number of threads to
be twisted together, or be a multiple of them, it will be all the more
convenient, though it is not necessary for it to be so. For instance,
if 1 thread is twisted round 1, =2, an 8-row harness would give 4
repeats of this in each row; but if 1 thread were twisted round 2, =3,
then 1 row of the harness would give 2 repeats, or hold 2 splitfuls of
warp and 2 threads over, so that another thread would have to be taken
from the back of the next row. For this a 6-row harness would be more
convenient, which could be got by leaving 2 rows of the machine idle,
and designing the pattern on 6-row paper, or ruling 8-row paper to
6-row after the design is painted. To work fancy textures with gauze by
means of a harness with doups in front, both must be arranged to suit
each other, unless a large number of doups be employed. For instance,
a 2-and-2 twill could be wrought with 1 doup leaf in front of the
harness by making 1 thread of the gauze twist round 3, and throwing
2 shots into each crossing, as shown in Fig. 131. If 1 thread were
to be twisted round 1 with a 2-and-2 twill, and 2 shots to the gauze
crossing, 2 doup leaves would be required, as shown in Fig. 132. A
4-thread fancy twill can be wrought with 1 doup, as shown in Fig. 133,
1 thread twisting round 3 and 4 shots into each crossing of the gauze.
Satins or regular twills are more difficult to work, unless by making
an irregular gauze. Take a 5-end satin, which has the threads over 4
and under 1. The gauze might have 4 shots and 1 shot into each crossing
alternately, and be easily wrought with 1 doup; but if 2 or 3 shots had
to go to each crossing it would be different, as the doup would require
to sink for 1 shot out of 5, in order to let the threads drawn into it
bind in the twilled portion of the cloth, or otherwise the binding
would have to take place on these threads when the doup is down, which
it could do if it suited otherwise. But it must be remembered that
when solid cloth and gauze are working together, the former is made
by raising the doup standard and those mails which carry the threads
that are not drawn into the doup, leaving stationary the mails carrying
the crossing threads, except where gauze is being made; therefore, in
working a 5-end satin, if the doup held up any of the threads for,
say, 3 shots, and then sank for 3, there would be no way of holding up
these threads in the twill portion to make them pass over 4 shots, as a
5-end satin must do, for if the harness twines carrying these threads
are raised, they will cause a cross to be made instead of completing
the twill. Patterns of this description should all be wrought in a
gauze harness. In the figures given the gauze crossings are represented
by dotted lines where they are raised by the harness, for sake of
distinction.

[Illustration: Fig. 133]


HAND-LOOM GAUZE HARNESS.

[Illustration: Fig. 134]

[Illustration: Fig. 134A]

A large quantity of goods are still being made in the hand loom on
the gauze principle. These are mostly for curtains known as Cretes
or Madras muslins, and for antimacassars or chair backs. They are
chiefly woven in and about Glasgow, and in Newtownards in the North
of Ireland. These goods are made on the half-harness principle--that
is, half the warp is drawn into the harness and the other half passes
through the harness and is drawn into a set of heddles in front. This
mounting is shown in Fig. 134. It will be seen that every alternate
thread is drawn into one of the harness mails and all the threads are
drawn through a pair of clasped leaves of heddles A B; those that are
drawn into the mails are drawn through the under clasp of the leaf A,
and those that are not in the mails through the upper clasp of the
leaf B. The threads are next drawn through the doups as shown, those
in the mails being in the back set, or upper doups. These doups are
made different from those generally used, the doup or half-heald being
connected with two standards; this is for the purpose of throwing less
strain on the yarn. When the harness is drawn the shaft A is also
raised, B being sunk, as well as the front set of doups. As the yarn
in the harness is also drawn through the back set of doups, it follows
that they must be relieved to allow the harness to raise the yarn. This
might be done, and sometimes is done, by sinking the ordinary doup
shaft and letting the yarn draw the doup through its standard; but in
this case the standard C is raised, which allows the thread to rise
in the doup, as shown at Fig. 134A, without having to draw it up, and
at the same time all the doups are kept firm in their places by the
standard D, thus preventing any entanglement in a fine set of doups.
The distance between the harness and back leaves is 4 in., and between
the back leaves and gauze mounting 3 in. The cloth is a plain gauze,
and the figure is formed by throwing one or more shots of rove weft,
either white or coloured, into the shed formed by the harness flushing
over the ground and clipped off afterwards. To form the figure shed,
draw the harness, raise the leaf A and the standard C, and sink the
other leaves, making a rising and falling shed, or centre shedding. To
form the gauze, for the open shed raise B C D and F, and to form the
cross shed raise E and F, in both cases sinking the others. For these
sheds fine or ground weft is thrown in, the flowering shots coming
between them. The shafts A and B are for the purpose of keeping a clear
shed, but might be done without. These goods are now largely made by
power looms, principally with gauze reeds, which see.


GAUZE HARNESS

In a true gauze harness the doups are connected with the mails or
leashes of the harness, the harness twines, or leashes, acting as
standards for the doups, so that any splitful of warp may be crossed
or twisted differently from the others, and any kind of texture can
be wrought along with the gauze without having to consider the one in
connection with the other, as in the previous harness. Figs. 135 and
136 show how these harnesses are tied up; the letters refer to the same
parts in both figures. The harness is shown with four cords in the row,
for the back harness and the doup mail in front, and this could be
repeated twice for an eight-row harness. It is better to have the rows
of the harness regulated in this way to suit the gauze to be made, and
have the doup leashes in one or two rows at the front, but it is not
necessary for it to be so, as the doups could come on any of the mails
through the harness, if desired, in case of a harness being tied up,
and it being wanted to add the doup leashes for gauzing; but this is
rather confusing. In any case, this method of working is now superseded.

[Illustration: Fig. 135]

[Illustration: Fig. 136]

A is the doup or dead leash, B the doup standard, or what answers to
the doup standard in heald work. C and D are the leashes carrying the
plain or lying threads, and E the leash carrying the crossing thread,
the crossing in this case being one thread round two. F is the doup.
Fig. 136 shows the harness drawn for the open shed, and it will be
observed that the dead leash A is connected with the back harness
E, so that when this leash is raised to form the shed it draws up
the dead leash, also relieving the doup. There was one trouble with
these dead leashes--namely, that they were liable to twist round the
standards when close to them in the cumber board, sometimes being put
through the same hole or the next one; but by having them in front
and a little distance off this could be avoided. Wires or cords were
sometimes stretched across between the lingoes to prevent the twisting.
Fig. 135 shows the harness drawn for the cross shed, the doup standard
being drawn up, drawing the doup up with it. The doup was fastened
into the mails as shown at No. 1, Fig. 135; but a better plan is to
have double-eyed mails, and fasten the doup as at No. 2. This form of
harness is now superseded by one having all the doups on a shaft in
front of the harness, which consists of two parts--namely, the back
harness and the douping harness (which is about 3/4 in. to 1-1/2 in.
in front of the former); the doups are connected with this latter
harness, and all the doups may be on one shaft, whether for one, two,
or more rows of standards in the front harness. The doups on the shaft
are known as a ‘slip heald’ or ‘slip.’ This is a much simpler and more
convenient method of working than the preceding.

Fig. 137 shows a simple and effective gauze mounting for a 4^{00}
jacquard, the cords being carried up to the machine in the usual way. A
is the breast beam of the loom, B the reed, C the back rail, and N the
yarn beam. D is the slip heald, or doup, which is simply a half-leaf of
ordinary clasped heddles, made of cotton or worsted, fine or coarse, as
may be desired for the quality of work to be wrought. E is the front
or doup mounting, connected with the two front rows of the machine,
and passing through a small cumber board in front of that for the body
of the harness; it is bolted so that it can be shifted to regulate
the distance between the doup harness and that of the figuring. The
distance between the two harnesses may range from 3/4 in. to 1-1/2 in.,
or whatever is found most suitable to ease the strain on the yarn and
make a clear shed. G is the cumber board, and F the body of the harness
for working the figure.

[Illustration: Fig. 137]

The method of slackening the threads that are drawn into the doups
is perhaps the principal feature of this mounting. When one or more
slackening bars are used it is plain that all the threads that pass
over the bar that is oscillated will be slackened at once, though for
figured work it may be that only a very few of these threads should
be slackened, and if the remainder or any portion of them are raised
by the back leashes they should be kept tight at the back so as to
enable them to draw up the slack doups. It follows, therefore, that for
giving good work and keeping the threads at a regular tension, every
thread in the doups should have its own independent slackener. It is
not absolutely necessary that this should be so; strong twist cotton or
worsted yarn will work very well in a harness with one slackening bar
vibrating for each shot, though it must throw an irregular strain on
the yarn, and it requires to be tightly paced.

In Fig. 137 the slackening arrangement consists of a small back
harness, as shown at I, through which all the whip threads are drawn;
these threads then pass over one lease rod and under the next one,
as shown, and then on to the harness to be drawn in according to the
draught of the gauze. The harness consists of small mails attached
to lingoes with twines in the same way as the lower portion of the
ordinary harness is prepared, but the mails are rounded off above the
eye or centre hole without having the top hole for threading the upper
portion of the harness to. The cumber board of this harness must be
nicely set, so that the lingoes will hang on the yarn without drawing
the mails quite close on the cumber board, which soon would cut the
twines. To regulate this the cumber board and back rail must be set in
relation to each other and the mails of the harness so as to produce
the desired effect. A medium position for working may be taken as
follows:--Cumber board 12 in. behind harness and 6 in. in front of back
rail. Back rail 2 in. above the level of the mails in the main harness,
and cumber board of small back harness 3 in. below the level of the
mails, or 5 in. lower than the top of the back rail. The back lease
rod might then come in about 8 in. behind the harness. The weight of
the lingoes on the back harness must be regulated to suit the strength
of warp to be wrought; about 30 to 40 per lb. for, say, from 40/2 to
80/2 cotton would suit. The mails for the body of the harness may be
25’s or 30’s, or the same as for ordinary work, and those for the doup
standards heavier, say 18’s or 20’s. In addition to this harness for
slackening, or rather for keeping a yielding tension on the whip, there
is a slackening bar supported by a lever K L M at both sides of the
loom. This lever has its fulcrum at L, and is attached to the swords
of the lay at M, and therefore vibrates at every beat up. The slackener
acts as a positive motion, and keeps the threads from tightening up
or being drawn forward by the twisting; it acts on the body of the
threads, and the harness acts on them individually.

This, when properly set, is an easy motion, and produces a good firm
gauze. It is a Bradford patent. Now, in reference to the working of the
gauze: The harness is levelled much in the same way as for ordinary
work, the mails being 1-1/2 in. or 1-3/4 in. below the level of the
breast beam. The doup mails should be a little lower than those for the
harness, say 3/16 in. or 1/4 in.

The shed should be closed, and begin to open again when the reed is
at the fell of the cloth. Too large a shed should not be made with
the doups, on account of the strain on the yarn: a small clear shed
gives much more satisfactory work. The front cumber board, containing
the doup leashes, may be shifted a little forward or back if thought
desirable, so as to get the cross shed to spring as easily as possible,
the amount of elasticity in the warp varying this considerably. The
slip heald is raised for every shot by cording it to the brander
or griffe of the jacquard, which for gauze work is generally a
single-acting one. Springs are used for holding down the slip--a light
wire spring, or, better, a piece of round indiarubber about 3/16 in.
in diameter, doubled if necessary. Bridles should be attached from the
shaft of the slip heald at each end and fastened to the cumber board so
as to bear the draw of the springs when the harness is down; the cords
connected with the jacquard for raising this shaft may then be left
slack, and the size of shed may be regulated or made less than the draw
of the machine, if required.

It is very important to have the doups nicely adjusted; on this depends
their lasting capacity, and sometimes a very short time is sufficient
to cut them to pieces if badly set. But if correctly done they will
generally last for a length of time, particularly if occasionally
turned on the shaft, which is more essential in working with a linen
than with a cotton warp.

One of the doup standard mails with a doup through it is shown at No.
1 (Fig. 138), where it may be seen that the doup is simply drawn
through the eye of the mail, and not connected with it in any way. It
is only held in this position by the warp thread passing through it,
and whenever the thread breaks the doup falls out of the mail. This
is a trouble to the weaver, at least till she gets accustomed to it,
and mails, as shown at No. 2, have been made to hold up the doup. The
remedy here is often worse than the evil. The mails wear the doups,
particularly if the springs draw them tight down on the eyes of the
mails; and besides, they have the disadvantage that the slip heald has
to be built in them after the harness is mounted, whereas with the
open mails the slips can be procured anywhere quite independent of the
harness. The slip heald must be set so that when the harness is down
the yarn will not be drawn through the eyes of the mails by the doup
sinking too low. This may be guarded against by adjusting the bridles
at each side, between the slip shaft and the cumber board. The loops
of the doups should pass about 1/4 in. through the eyes of the mails
when at the lowest or bottom position, and the front mails being about
1/4 in. lower than the others allows the loops of the doups to be just
bearing lightly on the warp, all being held up straight. The slackening
bar must be arranged to work in time with the doup standards, and
should not slacken the yarn too much nor too little, just keeping it
at a uniform tension. The great point with tender or weak yarn is to
reduce all friction to a minimum, no matter by what it is caused, and
if this is done the yarn and doups, if they are of a fair quality to
begin with, should both work well; but a very slight difference in
setting the doup may make a great deal of difference in the working of
it. A doup that may be cut to pieces in working 20 yards of cloth might
work 400 or 500 yards with so little alteration as not to be noticeable
to those not accustomed to the work. The slackening bar must not rise
so high as to raise the yarn that is above it, nor must it sink so low
as to press the whip down on the back harness board, and it must be so
regulated that within these limits it will give the proper amount of
slackening, and prevent the whip warp from getting tight and drawing up
the small back harness.

[Illustration: Fig. 138]

Worsted makes the best weaving doups for strong work, but for lighter
work cotton is mostly used, being finer and cheaper. Sometimes
polished linen yarn is used and makes a very smooth doup, but does not
give the wear.

Small metallic rings are sometimes used on the loops of the doups, and
the warp drawn through these rings; this suits for yarn with much fibre
on it, as it keeps the fibre from getting twisted into the doups, but
the rings cut the doups faster than ordinary yarn working in them would.

Instead of a slip heald in front, doups with a lingo on each have been
used, with the mails for the standards as shown at No. 2, Fig. 138, so
that the doups cannot fall out of them, and the yarn when drawn up by
the back harness in shedding raises the doups and lingoes up with it.
The weight of the lingoes must be such as will suit the strength of the
warp to be used--not too heavy for it to lift nor too light, so that
in the cross shed the doups would be drawn through the mails. These
doups being separate, are easily built in the harness and very easily
repaired by the weaver when they break, but they break too frequently.
It is, of course, only for a strong warp that this method of working is
suitable, and even for it, although it works very well, the principle
is not good.

The gauze harness which will probably be most frequently met with, at
least in the cotton trade, is shown in Fig. 138. It is an older harness
than that described, having been patented in Macclesfield in 1876 in a
somewhat different form. The difference between this and the preceding
chiefly consists in the method adopted for slackening the yarn, and the
jacquard may also be specially built to suit for this. In setting this
loom the harness and mails are tied up and levelled as before, but when
working the back rail is kept down, as it usually is for gauze work,
say about 1/2 in. lower than the eyes of the mails. Each whip thread
has its own slackener, the slackening apparatus consisting of a harness
wrought either by the same jacquard that works the figure, or by an
extra one, when much pattern is required.

The connections of the needles with the hooks are shown at A, the
two bottom needles being connected with the two front and two back
hooks, the former for working the gauze and the latter for working the
slackeners. This harness is intended for a 1-round 3-gauze, and is
mounted on a 10-row machine with an 8-row cylinder on it. It is not
necessary that the back harness for slackening should be raised so high
as the figuring harness is, and for this reason, as well as that it
makes the harness more direct and avoids friction, it is tied to two
sets of levers as shown at G and G^1; they are made of brass and are
fastened on two rods, one about 4 in. above the other, so as not to
come into contact with each other when rising or falling. The points
of the levers are fastened to the two back rows of hooks as shown, and
the harness is tied to wire hooks on these levers; F is the neck of the
harness, D the cumber board, and N the lingoes, which are much heavier
than those used for the figuring harness, from 6 to 10 per pound being
the weights frequently used, according to the class or weight of the
work.

The mails used on the back harness are similar to those used for the
doups--that is, such as are shown at No. 1 (Fig. 138); but if these cut
with the warp sawing through them, glass mails should be used. L L^1
are two rods or bars which support the warp that is drawn down between
them by the slackening harness. The closer these bars are set together
the greater will be the amount of slackening given to the warp when the
harness is drawn; therefore these bars should be made to slide either
way so that the desired slackness is obtained by setting them. About 3
in. to 4 in. apart is a good working width; the tops of the bars should
be about 1 in. lower than the mails in the figuring harness, and the
tops of the eyes of the mails in the back harness should be 1 in. or
1-1/4 in. lower than the rods.

Another method of arranging the needles in a jacquard for a gauze
mounting is shown in Fig. 139, which is, perhaps, more convenient for
the designer, as will afterwards be seen. In this harness there are 10
rows of hooks, the same as before--6 for the figuring harness and 2 for
the doup standards. The doup standards are here on the hooks connected
with the fourth and eighth needles, and come after the threads that
the whip in them twists round, and fall in this position on the design
paper, instead of both being at the end of a design--that is, on the
seventh and eighth checks. For any order of twisting, the arrangement
of the hooks and needles might be made to suit on the same principle.

It will be seen that in the leno jacquard, shown in Fig. 138, a number
of levers and attachments are required to work the slackening harness.
In order to simplify this, Messrs. Devoge & Co. have brought out a
leno jacquard with which none of these levers, &c., are required, the
slackening harness being tied to the two back rows of hooks in the
machine. These two rows of hooks are lifted by a separate griffe, the
lift of which is much less than that for working the body hooks of
the machine, and which can be varied to suit the amount of slackening
required. This machine is shown in Fig. 140, a view of the motion for
raising the griffes of both back and body harness being given in Fig.
140A. _c d_ (Fig. 140A) is the link connection between the bar of the
griffe _a a_ and the top lever L (Fig. 140), somewhat similar to what
is used for an ordinary jacquard. To raise the small griffe at the
back, it will be seen that there is a cross lever _e_ from the links
_d_, _c_ to its fulcrum _h_ on a bracket _g_ attached to the top of the
machine. The small or back griffe _b_ is connected to this lever by the
link _f_. The amount of lift given to the small griffe in proportion
to that given to the large one will be as the length of the lever from
the stud in _h_ to that in _f_ is to its length from the stud in _h_
to that in _d_. The farther the fulcrum is pushed back in the slot in
which it is fastened, the more the back griffe will be raised, the link
_f_ being always vertical. The back griffe is arranged to commence
lifting slightly in advance of the front one, in order to lessen the
strain on the doups. The needles are connected with the back and front
hooks in this machine similarly to those in Fig. 138.

[Illustration: Fig. 139]

The draft of a gauze in the harness is just the same as regards the
doups as for shaft work; the crossing may be made from left to right,
or the reverse, but the pattern must be made to suit. Some mount the
harnesses and paint the patterns from right to left. About Bradford
this appears to be the usual method, and it does not matter provided
the mounting, drawing and designing be kept in accord; but here we
shall keep to the same principle as is recognised all through--viz.
mounting and designing from left to right, which is the correct method
so far as drawing is concerned, and equally convenient in other
respects.

[Illustration: Fig. 140]

[Illustration: Fig. 140A]

To design a pattern for gauze: if it is to be a simple geometrical
pattern it may be put direct on the design paper; but if to be a
figured pattern, a sketch should be prepared for it in the same manner
as for ordinary figured patterns, whether for dress goods, curtains,
or any other material. Any desired figuring can be made, providing
sufficient space be left between the figures to admit of the open work
being made. The figuring should be treated as the ornamentation of a
plain or short twill ground figure, which is woven on a gauze ground;
the figure may be entirely plain or twill upon a gauze ground, in which
case it must be sufficiently plain and solid to be effective, or it
may be treated as damask figuring on ornamental patches of plain or
twill on a gauze ground; or, again, the gauze may form the figure and
the ground be plain or twill. Any variety may be made that ingenuity
can suggest, and figuring with extra warp or weft may also be adopted
if desired. When about to put the pattern on design paper, it may be
observed that out of the 8 rows of needles in the machine only 6 are
used for figuring, the other 2 being employed for gauzing; therefore,
some means must be devised for getting the pattern on the paper so
as to suit this. The simplest plan is that adopted in America--viz.
instead of using two rows of hooks along the jacquard for the doup
standards, to leave as many hooks at the first of the machine as will
correspond with the number of doup standards required for one repeat of
the pattern. Suppose for the same ties as those given--100 hooks for
the doups and 300 for the figure--then the cords from the first or last
100 hooks would be taken down through the two front rows of the cumber
board, and those from the 300 hooks would make up the body of the
harness. When mounting in this way, the jacquard would be better turned
with the cards hanging over the side of the loom, as for the ‘London
tie.’ This method of mounting enables the pattern to be designed for
300 hooks and painted on the design paper independently of the gauze,
which can be filled in afterwards.

[Illustration: Fig. 141]

Fig. 141 is an example of designing in this manner, using 6 rows of
hooks, to suit the gauze. The pattern is put on the rows of designs
numbered 4, 5, 6, 7, and 8. No. 3 is for the selvage, and the working
of the doups is on Nos. 1 and 2. This is for a 1 round 2 gauze, 6
checks, or 2 splitfuls of gauze to the design; there would be 10 doups
required for the 5 designs in the repeat of the pattern, and for these
10 checks are used of designs Nos. 1 and 2. The rising of the back
leash of the harness to form the crossing is shown by the dots marked
on the design, and as gauze is formed by raising this leash and the
doup alternately, it is only necessary to take each line of the design
paper for the doups, and, following up the markings for the back
leashes, fill in for the doups to rise when the back leash is sunk for
the gauze, taking care to keep clear of the figuring. In making gauze
the doup should always rise immediately after the figuring ends, to
form the first crossing, and it should also be up immediately before
the next portion of the figuring begins, so that in an irregular
figured pattern the gauze might frequently require to be broken to
effect this, as may be seen in Fig. 141. Take the first upright line of
the pattern, design No. 4: the back leash is raised for three, then the
doup for three, and the figure follows. Passing on to the tenth line
it will be seen that the doup rises for three after the figure, and is
also up for three before the next figure begins; but, passing on to
designs 7 and 8, it will be seen that the risings of the doup and back
leash would not fall in regularly in threes against the figure, as it
would be better if they did, and therefore the gauze is broken so as
to fall in; but in small patterns of this description, when it is of
importance to keep the outline of the figure as regular as possible,
this can and should be remedied; instead of the outline of the figure
running diagonally on the design paper in steps of one check, it should
go in steps of as many checks as there are threads to the twist of the
gauze in the warp and shots in each twist in the weft. This is shown in
Fig. 142, where all the twisting falls in regularly with the figure.
This may to a great extent be carried out with irregular figures by the
designer exercising a little judgment, and when filling in the gauze
make up the plain edging of the figuring to fall in with it--that is,
only to break the gauze where there is room to do it effectively, and
where there is not, a few dots can be added to the plain. If Figs. 141
and 142 be compared, it may be seen how this could be done: examining
designs 4 and 8 above the figures, the latter below as well; but
Fig. 142 shows the proper method of treating these small geometrical
figures. Another point to be observed is whether all the doup leashes
are to be raised together to form the gauze, or whether it is to be
formed by raising every alternate doup leash and every alternate back
leash, as is done in these figures. The latter method is preferable
for working, as it makes an even tension on every shed, whereas if one
shed is a full cross shed formed by the doups, and the reverse of this
an open shed with no crossing, the cross shed will be the tighter, and
produce a certain amount of plucking, with a tendency to shire in some
patterns; but in case of working dices or any pattern where the top and
bottom of the figure extend some distance straight across the design
paper, it would be impossible for the first shed of the gauze to be
formed by the doup if only every alternate doup leash was raised. Some
judgment must therefore be exercised in these matters, and the best
effect obtained that will suit circumstances.

[Illustration: Fig. 142]

Now to put the patterns on design paper to suit the harness when the
two front rows of the jacquard are used to work the doups, as in Fig.
138, also when any other rows are used, as in Fig. 139: the simplest
way to do this--at least, for those not well accustomed to the work--is
to design them on 6-row paper, as Fig. 142, and afterwards transfer
them to 8-row paper, leaving the lines for the doups empty, the doup
lines being afterwards filled in, as in Figs. 143 and 144. Fig. 144
has the advantage that each splitful of gauze stands alone, whereas
when filling in the markings for the doups in Fig. 144, two splitfuls
stand together; but a little practice makes one method as simple as the
other. The dots represent the markings for the doups, and the shaded
squares those for raising the back leashes to form the open shed of
the gauze. The transferring of the designs from one sheet of design
paper to another is, of course, a matter of trouble and expense, and
may be avoided if an experienced card-cutter is to cut the cards. The
design need not be transferred, but have the marks for raising the
doups put on the same lines of the design paper as those for raising
the back leashes for the open shed. This is shown in Fig. 145, the
shaded squares being the same as in Figs. 143 and 144; but the dots,
instead of being put on lines left for them, as in these figures, are
put on the same lines as the shaded squares, and the card-cutter sees
to their being cut on the correct lines, as in Figs. 143 and 144. In
designing, different colours would be used. The shaded square should be
the same colour as the figure, as they are to be cut for the lines they
are on, but the dots should be of a different colour, to be cut for a
different line of the card from that on which they are placed on the
design--the dot on the first check of a design being the seventh hole
in a row of the card, and that on the third check of a design being the
last hole of a row on the card.

[Illustration: Fig. 143]

[Illustration: Fig. 144]

[Illustration: Fig. 145]

Fig. 146 shows a portion of an ornamental pattern as it would be
designed for a gauze ground of two threads of warp twisting round two
= 4 in the split and 4 picks into each crossing of the gauze. This
would require a 10-row machine, eight rows for the figuring harness and
two rows for the doup standards. When two threads are twisting round
two, it is not necessary to have a doup for each, though it may be
desirable; but it suits very well to have both threads drawn through
the same doup. It will be necessary to have a stronger doup in this
case, but if the warp is very light it would have so much more to do in
drawing up the heavy doup than if each thread was drawn through a very
light doup that the latter had better be adopted; otherwise a stronger
doup, with both threads through one, is simpler. On the pattern, the
doups are raised by the dots, which must be cut on the proper line
of the card, as before; both lines are dotted as if each thread had
its own doup, but only the front dots of each pair are necessary. The
shaded squares are for raising the crossing threads with the leashes
for the open shed of the gauze. In this pattern it is necessary that
the outline of the figure should go in steps of four warp ways, as
there are four threads to the twist; but it is not necessary that it
should go in steps of four weft ways. A better outline can be given
to the figure by not adhering to this, though the perforations in the
gauze will be to some extent sacrificed by being irregular; but this
would be so in any case in order to raise the doup standard after
the figure when the doups and leashes are gauzing alternately; they
might be regularly started at one side of the figure, but this would
probably make them more irregular at the other side. By following
round the edges of this figure it will be seen that by a little care
no blemish of any consequence need be left, though it will not be so
perfect as when calculated to fall in as Fig. 145.

[Illustration: Fig. 146]

Any variety of simple gauze texture may be used, as well as the plain
and honeycomb gauze, but unless there is plenty of space and the
figures are plain and large, no very intricate crossings should be
attempted, unless for stripes, as they would probably mar the outline
of the figure; but if not, they may be used where suitable.

Another important point in designing for gauze is that if the texture
of the solid cloth is loose and that of the gauze pretty firmly
twisted, the take-up of the warp may be very different when weaving. To
avoid this the gauze and figure should be well intermixed, and not be
in too great quantities in any one part--in fact, the gauze and plain
should be constantly interchanging, and when the pattern is such that
this cannot be, the textures must be regulated so that there will be
about an equal quantity of warp taken up for each. A little practice
will soon show this.

Nothing has been said about the description of jacquard used for
working gauze, but the same principle holds good as for working with
shafts. A gauze shed is taken to be a plain jacquard lift--that is, the
shed formed by the single-lift jacquard raises the yarn and lets it
down again before beginning to raise the following shed. This enables
the doup and doup yarn to fall into their places before being drawn
away to form a following shed. At the same time, we are not limited to
a single-lift jacquard. If you have a double-lift jacquard, and want
to work a small piece of gauze, do so; but do it correctly, or, in
other words, take means to make a good job of it. Before explaining
how to do this, it will be well to show where the difficulty exists.
In a double-acting machine one griffe is rising when the other is
falling, and in plain gauze the doup standard and leash carrying the
crossing thread are raised alternately, one shot being thrown into the
shed formed by each. Therefore, the same thread is up for every shot,
and the thread round which it twists is always down. Now, when one
griffe has raised the doup standard and the next one raises the leashes
carrying the same threads, both will pass in the centre of the shed,
and, as shown in Fig. 147, the crossing thread will be dragging up the
lying thread from the position shown by the under dotted line, thereby
causing considerable friction on the yarn, which must be detrimental to
a fine or weak warp, and is inclined to drag the doup through the eyes
of their standards; but as the harness in coming to the centre slackens
the warp threads, and also as the slackener begins to act when the
harness begins to rise, the threads will be further slackened in those
sheds where the doups follow the back leashes, and this reduces the
friction. It may, however, be entirely avoided by raising the lying
threads to the mid position to meet the descending griffe, and then
let down with it. A simple method of doing this is to have the leashes
carrying the lying threads through a separate slip of cumber board at
the back, and knots on them above it, so that when it is raised all the
lying threads will be drawn up with it.

The cumber board can be raised by a cam, or in any other suitable way.
(See Journal Weaving for Scotch Carpets.)

[Illustration: Fig. 147]

These leashes might also be raised by means of a shaft through loops in
them, or with bars under the jacquard hooks, in the same way as is done
in twilling harnesses (which see), but with a special motion to raise
and lower them. In Fig. 147, the dotted line A, B shows the middle
position or centre of the traverse of the yarn in shedding.

Although this is the proper means to adopt to work gauze satisfactorily
with a double-acting jacquard, or with any double-acting shedding
motion, it may be that small portions of gauze, such as a few stripes
up a web, are wanted without the trouble of making any special
arrangements for it. Good strong twist cotton yarn will work very well,
if nicely tempered, in a double-acting machine. The slackeners can be
arranged by leaving a few hooks apart to work them with, and cords
tied to these hooks and let down at the back, in the same manner as in
the gauze harness (Fig. 138), which will do very well. When three or
more shots are thrown into each shed of the gauze--that is, between
the crossings--there will be much less friction than when the crossing
takes place at each shot, as it is only at the crossing the friction
takes place.

One source of annoyance in working gauze is the wear on the doups,
which requires them to be changed frequently and adds cost to the
production. It has often been attempted to work the gauze by other
means, and for some classes of work this has proved successful. One
of the best arrangements, and one that is largely used, is the gauze
reed, which is almost entirely adopted about Glasgow for weaving leno
curtains in the power loom; it is also employed for other classes of
work, but is limited in its use. It will not supply the place of a
gauze harness.


THE GAUZE REED.

This is principally used for working leno curtains or Madras muslins in
power looms. The texture of these fabrics is a plain gauze ground, the
figure being formed by one or more colours of thick soft weft, thrown
in and bound for the figure, but flushed or floated over the ground,
to be cut off afterwards. The reed is wrought in conjunction with a
harness, the reed forming the ground of the fabric and the harness the
figuring. The harness is mounted in the same way as a half-harness, or
for half the warp, and the reed does the work of the doups. The gauze
reed will suit for gauze up to 55 threads per inch in the reed, but
is more used for fabrics ranging from 30 to 50 threads per inch. A
portion of the reed is given in Fig. 148, in which it will be seen that
it consists of the usual dents set in ribs, and a set of half-dents,
perforated at the points which are set into the lower rib of the reed
behind the others, and lie about half-way through the splits in a
slightly slanting position. The gauze reed works behind the slay and in
front of the harness. It should be about 1/2 in. behind the slay when
the cranks are at the back centre, so as to leave sufficient room for
clearance between them.

[Illustration: Fig. 148]

[Illustration: Fig. 149]

[Illustration: Fig. 150]

A good method of working the reed is shown in Fig. 149, which is a
back view. A is a rocking shaft supported by brackets, as _a_, on
the top rail of the loom; B is a lever projecting towards the front
of the loom, and the rod _b_ from it supports one end of the reed; C
is another lever on the shaft A, projecting backwards, and the rod
D connects it with the tappet treadles, so that when the treadle is
pressed down the reed rises; about 3-1/2 in. is the usual height for
it to rise. F is the upper rib of the gauze reed, and F^1 the lower
one. G is the yoke of the reed. On the yoke is fastened a bar P P,
connected with the strap which passes round the top rib of the reed.
I I are two projections on the bar P P, through which the slide rod E
E^1 passes and is fastened at the top to a bracket C on the top of the
loom, and at the bottom it fits into a socket or stand N fixed to the
ground; the reed slides up and down on this rod. H is a tube round the
slide rod between the projections I I, and is about 1/2 in. shorter
than the distance between them; to the bottom of this tube is fixed
the support H^1 for the dipping rod K, which, when the yarn under it
is very tight, can yield 1/2 in., along with the tube H. The reed is
not positive in falling, but when raised by the tappet it falls by its
own weight and that of the weight-box L, which slides on the rod E E^1.
The weight-box is fastened by a rod M to the crank on the bar which
supports the dipping rod, and weights it as well as the reed. Small
weights, as R, about 1/4 lb. weight, can be placed in the weight-box
till as much weight as is required is put on. Fig. 151 gives a view
of the gauze reed mounting, with the gauze or cross-shed open. Every
alternate thread of the warp is drawn into one of the gauze dents, and
every other one into the harness, as shown. The harness is levelled 2
in. or 2-1/2 in. below the level of the breast beam; or, having the
slay levelled with the race 1/2 in. below the breast beam when the
cranks are at the top centre, push back the slay till the cranks are
at the back centre, and place a straight-edge across the race, and the
under edge of the straight-edge where it passes the front row of the
harness will give the height for the eyes of the mails. The back shell
E should be set 1/2 in. below the level of the breast beam, and should
be made to vibrate with a tappet, which should lower it when the gauze
shed is opened, and keep it raised for the figuring shots. A shows the
position of the weaver’s reed, B is the gauze reed, C the dipping rod,
D the heck or ravel which is on the harness under the harness reed or
cumber board F; the ravel should be about 6 in. above the yarn and 3
in. below the cumber board, making the latter 9 in. or 10 in. above the
yarn. The ravel is like a small harness reed with only one rib on it,
which is placed behind the harness, being covered with an iron tube to
strengthen it; in front a flat iron rod is fastened above the ends of
the dents and serves as the second rib. Three wires are stretched along
the ravel to keep the harness steady, in the same way that they are
stretched along a harness reed, but three are sufficient, instead of
having one for each row of harness. The ravel is shown in full in Fig.
150; it is about 1-1/2 in. deep (or across it), and the cumber board 3
in. deep; it is for the purpose of drawing the yarn in the harness to
either side, so that when the gauze reed is rising the half dents may
pass to the right and left of the threads in the harness alternately;
it is this which gives the gauze twist. Two bars A A^1 (Fig. 150)
extend from the ends of the frame of the heck; these slide in brackets
C C^1 fastened to the sides of the loom. The lateral motion is given to
it by a lever B B^1 with its fulcrum at B^1; this lever has a bowl D
on its lower end, which is acted on by a wheel E driven at the proper
speed to suit the shedding. E^1 is a face view of this wheel, which has
a projection on it for shifting the lever.

Sometimes the motion is imparted to the lever so as to shift the ravel
just when it is required--that is, before the gauze reed has risen to
form the shed; and sometimes the ravel is shifted a part of the way
at each shot of figuring, and has completed its traverse before the
gauze reed has risen; this latter makes a more gradual pressure on the
harness, and avoids any sudden plucking. When the wheel E acts upon the
lever the ravel is drawn to the right, and when released the spring F,
fastened from the side of the loom to the end of the ravel, draws it
back again. The traverse of the ravel should be about 3/4 in. C^2 is a
separate view of the brackets C and C^1.

Sometimes the heck is on the yarn between the reed and the harness,
instead of being on the harness; in this case there must be about 8
in. between the reed and the harness, and the heck or ravel rests
about half-way between them. The harness is wrought with a double-lift
machine and three tappets. One tappet which is much like a plain wyper
tappet, but with less dwell on it, acts on one of the treadles, which
is connected by a rod and lever to one of the griffes: this raises
the flowering shed, the lift being about 3-1/2 in., and the flowering
shot is thrown in as in ordinary work, the sheds being close when the
cranks are at the top centre or a little over it. As soon as this shot
is thrown in, when the cranks are a little over the back centre, the
second griffe is raised by another tappet. As this takes place before
the cylinder is in against the needles, all the yarn in the harness is
raised; the reason for raising it is to get all the harness yarn clear
of the half dents of the reed, so that it can be shifted over them by
the ravel. The gauze reed is raised by the third tappet; it begins to
rise shortly after the second griffe, or when the cranks are a little
past the top centre during the same revolution. The yarn raised by the
first griffe begins to fall at the same time that the reed begins to
rise; the second griffe raises the bottom yarn till it meets the top
yarn of the first shed descending, at which time the ravel should be
fully shifted, and the points of the half dents rising should be 3/4
in. below the harness yarn. As the harness falls and the reed rises,
the yarn in the former will pass down at the opposite sides of the
half-dents to which it was before being raised; a ground shot is then
thrown in (this is the second ground shot), and the reed descends,
taking the yarn in it down level with that in the harness. The harness
is again raised as before for the next flowering shot.

When the reed is raised to form the shed for the ground shot no card
is required to pass over the cylinder of the jacquard, and to prevent
one passing the sneck or shears is raised for this shot, so that the
cylinder is not turned. The reed must be raised in time to have the
shed open for the passage of the shuttle, but keeping it a little late
requires less lift to be given to the second shed which is being raised
to get the bottom yarn clear of the half dents of the reed. The above
description is intended for weaving fabrics with one cover, or one
colour of flowering shot, as when the pattern is in white or a single
colour, as it often is; but when two or more colours of figuring weft
are used, the flowering sheds might be opened with one griffe working
the machine single-acting, and constructing and timing the tappets to
suit. One tappet would then open two or more sheds for the flowering;
the second and third tappets would act on the bottom yarn and reed as
before. It is, however, usual to work the machine double-acting, and to
construct the tappets to raise the griffes alternately for two covers.
No. 1 tappet would raise one griffe, No. 2 tappet would raise the
other griffe; then a small wing on No. 1 tappet would lift the bottom
yarn, and No. 3 tappet would raise the reed, the sneck being lifted at
the last beat of the cylinder to prevent it from turning. For three
or more covers the tappets would require to be arranged to suit in a
similar manner. Fig. 151A is a front view of the position of the yarn
in the reed when the flowering shed is drawn. Only a portion of the
harness would be raised for each shot, as, say, the threads 1, 1, the
remainder 2, 2, being sunk. After the flowering shots are thrown in,
the bottom yarn in the harness 2, 2 has to be shifted to the other
side of the half dents for the gauze shed. It is therefore raised up to
meet the threads 1, 1 coming down. All are then shifted together, and
when the reed is raised the position of the harness yarn is shown at
the crosses; 0000 shows the position it was in at the previous ground
shot. The dots on the half dents show the position of the yarn in them
in relation to that in the harness 2, 2 when the reed is at rest, the
yarn in both reed and harness being on a level. The reed could be
wrought turned upside down with a harness, but all the loose flowering
weft would require to be below. In this case there is no necessity for
getting the under shed clear of the gauze dents for changing it with
the ravel to form the ground shed; but this is an objectionable method
of working, particularly with a light pattern, on account of the heavy
shed that has to be raised. If the flushed yarn was thrown to the top
and the reed turned upside down, the only way to work the heck would be
on the harness below the mails, which would require it to be as deep
as a reed or cumber board, in order to let the lingoes fall, and this
would not answer very well.

[Illustration: Fig. 151]

[Illustration: Fig. 151A]




CHAPTER VIII

_DOUBLE CLOTH_


DOUBLE CLOTH AND QUILTING HARNESSES

A very great variety of cloths come under the heading of double cloth;
in fact, almost every description of coloured cloth figured in the
loom, excepting coloured damasks and the commoner descriptions of dress
goods, as well as many uncoloured fabrics, are generally woven more or
less on this principle. Quiltings are mostly made on the double-cloth
principle, and have nothing to specially distinguish them, except
that some makes of cloth appear to be more suitable for goods of this
description, though frequently other fabrics of a much more costly
nature are woven in the same way; as, for instance, matelasses may be
made in the richest silk and worsted for jackets, as well as in coarse
cotton for quilts. One of the largest sections of double cloth is
carpetings, but these will be described in a separate section.

For figuring double cloths the same principles of the structure of the
cloth, &c., hold good as when weaving them plain, or in fancy texture.
Any suitable pattern can be put upon any description of double cloth
with a full harness mounting mounted in the ordinary manner; but the
pattern must be spread over the design paper to suit the description
of cloth to be made, which often necessitates its being in a very
distorted form, unless it can be arranged in different colours so that
the card-cutter can cut several cards from each line of the design
paper. For instance, if three colours are required to form one complete
weft line, then three lines of the design paper would be taken up with
these, unless the colours could be arranged that the three cards could
be cut from the one line, as frequently can be done; and the same for
the warp when there is a face and back warp, or several colours to be
brought up for one warp line. When there are two warp threads, one for
the face and one for the back, two hooks of the jacquard are required
to weave them, and two lines on the design paper are also required,
unless two colours are painted on one line to guide the card-cutter in
cutting the cards.

It is a great saving of work to the designer when the pattern has
not to be spread over the paper so as to give each line separately,
as doing so generally puts the pattern so much out of shape that it
requires first to be designed square, or of the proper dimensions, on
suitable design paper, and afterwards transferred to the working design
for the card-cutter. When goods are being largely made this can often
be avoided by special mechanism or mounting, to act instead of the
cards, it being then worth the trouble of having a specially prepared
mounting, and it is in these cases where double-cloth mountings are
used. Perhaps the simplest class of the double-cloth range, though
not really a double cloth, is figuring with extra weft. The figuring
weft may be thrown in as a spotting intermittently, or may be every
alternate shot, a ground shot following each figure shot.

Suppose the cloth to be a plain ground and a twill figure, and that 500
cards would be required to weave the figure if it were plain damask; it
is then evident that 1,000 cards would be required if every alternate
shot is ground, and the others for figure, and that the 500 plain cards
for the ground would be two cards repeated 250 times. Now, instead of
this, if we use a double-cylinder jacquard, and put the figuring cards
on one side and four plain cards on the other side, we can save 496
cards, as the four cards on the cylinder will work the plain at every
alternate shot. Or suppose a single-cylinder machine to be used, then
shafts might be put through loops in the harness; or the cumber board
might be made in sections, with knots on the harness twines above it,
so that in either case the harness could be raised in rows to form
plain cloth.

The plain shed could be formed with a tappet and levers, as in journal
weaving, and the griffe would only be raised for every figuring shot.
Plain cloth might also be wrought with a twilling motion on the
machine, such as is used on twilling jacquards. A pair of pressure
heddles in front of the harness is a common method of accomplishing
this in the hand loom. The harness is drawn for one shot, then let
down, and one of the pressure heddles sunk and the other raised for the
second shot. For the third shot the harness is again drawn, and for the
fourth shot the plain shed is crossed with the heddles. A twill or any
figure may be used, as well as a plain ground.

The next step in double cloth is perhaps double-weft-faced cloth--that
is, when the surface of the cloth on both sides is formed by weft, the
warp lying in the centre, and merely acting as binders to bind the two
wefts together. The warp may occasionally be brought to the surface to
give additional effect, or to add a third colour; or the body of the
weft may be kept on one side, the warp forming the ground on the other
side, with a weft figure on it, producing a one-sided cloth.

This class of work is best wrought in a pick-and-pick loom; but a check
loom--that is, one with extra boxes at one side only--can frequently be
used, and gives much less trouble. For a check loom two fine shots may
be used as one, having each pair of cards alike, except at the selvage,
and throwing two shots into each shed. Of course this takes twice the
number of picks and cards (unless two picks be given to each card), but
in some cases, as for borders, this may be an advantage in comparison
with working pick-and-pick. Often two succeeding cards of the pattern
for the same colour are made to follow each other, as may be found in
tapestries, &c., and the weft knocks up into its place without showing
any defect; but in a heavy cloth, with a thick or stiff warp, the weft
could not be knocked close together on this principle. With a fine
binding, warp of cotton and a thicker weft of woollen or worsted, there
is no difficulty in this working.

For double-weft-faced patterns the simplest method of designing is to
paint the pattern as if for plain damask, and then cut two cards for
each line of the design paper, one card being cut exactly the reverse
of the other; that is, the cutter cuts the painted squares for one card
and the ground squares for the next card. In lacing these two cards
are kept together, a ground and a flowering card coming alternately.
There is, however, an objection to this method of binding the texture
of the cloth, though in many cases it works very well. The objection is
that the warp has to be brought from the face of one side through to
the face of the other side for the succeeding shot to form the binding,
instead of being brought from the centre only to the face as it should
be.

[Illustration: Fig. 152]

Fig. 152 shows a portion of a stripe pattern designed to produce 8-end
twilled stripes. If a black and a white weft are used pick-and-pick,
the white pick coming first, A would be a solid black stripe and B a
solid white stripe, or the appearance would be as on the design paper,
but much reduced in size. On the under side of the cloth the colours
would be reversed.

[Illustration: Fig. 153]

The first card would be cut as the painting, cutting black, and the
second card would be exactly the reverse of this, cutting the white or
ground squares on the same line of the design paper. The texture formed
on the cloth by these cards is shown in Fig. 153, each line of the
design paper representing one card.

[Illustration: Fig. 154]

The correct method of binding is shown in Fig. 154, where it will be
observed that the binding of two succeeding weft shots does not fall
on the same warp thread, and that the binding dot will be covered by
the floats of the preceding and succeeding weft shots when the weft is
knocked up close. There are other squares on which the binding dots
might be placed as well as those in Fig. 154, one check to the left of
where they are being frequently used. A stripe pattern, as Fig. 152,
can be put upon design paper with each line representing two cards,
and be twilled to give the texture in Fig. 154, as is shown in Fig. 155.

When cutting this the cutter cuts the shaded squares and black dots
for the first card, and for the second card he cuts white and crossed
squares on the stripe B, and the solid black squares on the stripe A,
which it will be seen is exactly the same as cutting the black for the
first two lines of Fig. 154.

[Illustration: Fig. 155]

When painting the design the shaded squares may be red, the black
ones black, the black dots yellow or white, and the crosses green;
or any other convenient colours may be adopted. It will be observed
that in Figs. 152 and 155, where the raised and sunk twills come
together at the junction of the two stripes, there are no binding
dots on the two threads lying beside each other. It is not necessary
to have any binding for these threads, as the two wefts crossing bind
them sufficiently. Any desired figure may be woven on cloth on this
principle, which is a very convenient one, as the colours can be so
readily changed, and three or more colours may be used as well as two;
but in this case there is generally a right and a wrong side to the
cloth, the body of the wefts being at the back and each colour brought
through to the face to form the figure. Instead of the body of the
weft, or that portion not required for the figure, being at the back
of the cloth, it may be in the centre of the warp, and an equally good
face made on both sides of the cloth; but this requires more wefting,
makes a thicker and harder cloth, and is more expensive and more
troublesome to weave. Warp figuring may be done in exactly the same way
as described for weft figuring, and the pattern turned sideways shows
how it would be for warp; the only difference in the work all through
is to treat the warp as explained for weft. Of course, in the designing
and card-cutting, cutting two cards, one the reverse of the other, from
a design painted as if for a damask, would not fall in, but the designs
that are wrought fully out on the design paper would be all right. A
contraction of the work similar to cutting two cards from one line of
the design paper would have to be sought for in the mounting in the
first place, and in the card-cutting, to suit it.

[Illustration: Fig. 156]

Suppose the design to be as in Figs. 152 and 155, and to be painted
for a 200-jacquard, then to allow for double the quantity of warp a
400-jacquard would be required, the machinery being increased instead
of the number of cards. Now, to mount a loom for this class of work
the simplest plan is to divide the machine and cumber board into two
divisions, one half being used for one warp, which may be called the
face warp, and the other half for the other warp, which we may call
the back warp. A mounting of this description is shown in Fig. 156, in
which the machine is divided into two portions, 1 and 2, with 200 hooks
in each; the cumber board is also divided into two sections, marked A
and B. Here only four rows of needles are given, but any number can
be used. The cords from the hooks 1 to 200 are taken through the back
cumber board A, and those from the hooks 201 to 400 through the front
cumber board B.

When entering the yarn, one thread is taken from the back warp and
drawn into the first mail of the back harness, and the next thread is
taken from the face warp and drawn through the first mail of the front
mounting. Now, in reference to the card-cutting: instead of cutting two
cards one the reverse of the other, one half of the card must be cut
the reverse of the other half, the card-cutter cutting the coloured
squares of the design (200 checks) on the first half of the card, and
going back again and cutting the blank squares on the second half of
the card; or, when there is any variation in the twill, the pattern for
the face may be painted on 200 checks, and that for the back on 200
checks, and when cut on the card the threads of the warp will be raised
in proper order by the mounting.

Fig. 157 shows this principle of mounting with a straight or Norwich
tie, for which it is not so suitable as for the London tie (Fig. 156)
on account of the crossing over of the harness, though it is not very
objectionable if a narrow cumber board be used. Instead of having two
separate cumber boards with four or eight rows in each, one broad board
with eight or sixteen rows is better for the Norwich tie; then let the
first row of the back harness pass down through the back row of the
cumber board, and the first row of the front harness pass through the
second row of the cumber board, and so on, the back harness filling the
odd rows and the front the even rows; and when drawing in the yarn,
taking an end from the back and face warps alternately and drawing
them regularly over the harness causes them to fall correctly on their
respective mountings. Instead of dividing the card in this way and
reducing the machine to half its figuring capacity, a double-cylinder
machine might be used, the hooks from one set of needles being used for
one mounting, and those from the other set of needles for the other
mounting; then cutting the face pattern on one set of cards and hanging
them on one cylinder, and using a second set of cards containing the
back pattern for the other cylinder, should attain the same results,
and give double the extent of pattern on the same size of machine, both
griffes being raised and sunk together, and both cylinders being also
brought in and out together.

[Illustration: Fig. 157]

[Illustration: Fig. 158]

A modification of cutting the back and face pattern on its own half
of the card is to be found in the double-cloth quilting mounting
common about Paisley. These quilts are generally of a plain texture,
and the figuring is formed by passing one cloth up and down through
the other. The machine generally used is an improvement on the old
French draw loom, and is shown in Fig. 158 as it is made, principally
of wood, for hand looms. In this sketch the needles and cords in lieu
of hooks are not shown, but one row of them is shown in Fig. 159 with
only four hooks to each section instead of eight. H and K are the two
lifting levers for raising the griffes or trap-boards A and A^1 (Fig.
159). Nos. 1, 2, 3, and 4, and 1_a_, 2_a_, 3_a_, and 4_a_ are the cords
in lieu of hooks to which the harness is connected. A and A^1 are
two perforated boards with the slotted holes through which the cords
pass, turned in the reverse direction in one from what they are in the
other. Each needle is connected to two of these cords--one through each
hole-board--and the knots on the cords above the holes stand over the
slots or saw-cuts of the holes in the board A and over the round or
open portion of the holes in the board A^1, clearly shown at F (Fig.
159), so that if the lever H (Fig. 158) were pressed down and the board
A raised without any card acting upon the needles, all the harness tied
to this portion of the machine would be lifted, and if the board A^1
were raised by the lever K, none of the harness tied to this portion of
the machine would be lifted. Now, if a blank card were placed on the
cylinder, the reverse of this would be the case, and anything between
these two extremes can be got by cutting the cards.

[Illustration: Fig. 159]

The cylinder is connected with the back griffe only, and remains in
for two shots. The griffe A is first raised, and a shot thrown in;
it is then lowered, and the griffe A^1 raised, which presses out the
cylinder and changes the card; a second shot is thrown in, and so on. E
and F (Fig. 158) are two slide rods or spindles to steady the griffes
or trap-boards when rising: G and G^1 are bottom hole-boards to steady
the cords; L is a clap-board for pressing back the needles instead of
springs (it is fastened on springs I I, which hold it out a little from
the needles, and these springs are fastened from the top by cords to
the cylinder frame, which slides out with the cylinder, and draws in
the clap-board, pressing forward all the needles). Now to mount this
machine for double cloth. Suppose each frame to have 400 cords, and
that there are 400 needles, the same as an ordinary 400-machine. Let
the texture of the cloth be plain, and to work this 80 needles may be
set off, leaving 320 for figuring. These 80 needles will be connected
to 80 upright cords in the trap-board A, and the same in A^1. The
harness for the face cloth is tied to the remaining 320 cords in the
trap-board a, and that for the back cloth to the 320 cords in the
trap-board A^1, and this portion of the harness works the figuring
only without the texture. The texture is wrought by the 80 cords in
each board connected to the first 80 needles. To each of these upright
cords a number of harness twines are tied, one of which is fastened to
each twine of the harness from the 320 cords, just above the cumber
board; but in doing so the twines from the 80 hooks on the cords in
the trap-board A must be connected with the twines of the harness from
the trap-board A^1, and the reverse for the other set. In this way
all the back harness could be raised with either griffe, and the same
with the front harness. Either griffe could raise one warp with the
plain harness, and the other with the figuring harness. When cutting
the cards plain texture is cut on each card for the first 80 needles,
and the figuring portion without any texture is cut on the remainder
of the card. The 80 cords open the texture or plain shed, and the 320
cords keep the two warps in their proper places, raising what should be
above the weft, and leaving down what should be below it for each shot
according to the pattern. In these mountings the cards hang over the
side of the loom.

The same method of working can be applied to a jacquard with upright
hooks, by having the heads of the hooks for each machine or set turned
in opposite directions, as shown in Fig. 160. One set of hooks stand
with their heads over the knives in the usual way, as at A, and the
others require to be pushed back by the card to bring them over the
knives, as shown at B, so that pushing any needle back pushes the
front hook connected with it off the knife and the back one over the
knife. The mounting may be made up as before, or a twill or any smaller
texture may be used instead of plain; or a different texture may be
made on face and back cloths.

[Illustration: Fig. 160]

[Illustration: Fig. 161]

[Illustration: Fig. 162]

This description of cloth is much used for other makes of goods as
well as quilts. Kidderminster or Scotch carpets are woven on the same
principle, and curtains, cotton vestings, &c., are frequently woven
with the same texture. It can always be woven with a full harness
mounting by preparing the design to suit it. Fig. 161 is a pattern for
a diagonal figure in plain texture for this class of work. The black
squares represent the face warp raised to the face of the cloth, and
the shaded squares represent the back warp raised to the face also.
The blank squares represent weft on the face of the cloth. This would
give a magnified view of the face of the cloth if it were woven with a
black face warp, a grey back warp, and a white or light-coloured weft.
If each warp had a weft of its own colour the effect would be as shown
in Fig. 162, which would be the way the pattern would be painted on
the design paper if intended for weaving on looms specially mounted,
or with special machines as described; but if a full harness is to be
used, then the pattern must be spread over the design paper so as to
give a line of the paper for each thread of warp and pick of weft.
Fig. 163 is the same pattern arranged in this way. The black and grey
squares represent the face and back warp raised for the pattern. The
crosses show where the back warp is raised for the texture of the back
cloth at the back of the face cloth, and the dots represent the face
warp raised for the texture where the face cloth goes behind the back
cloth, or where the back cloth comes to the face. Of course, on a cloth
of this kind the warps are alike, face and back.

[Illustration: Fig. 163]

[Illustration: Fig. 164]

It will be understood that, when the threads of the back cloth are
being raised to form a shed, all the face threads of the corresponding
portions must also be raised to clear them out of the way of the
shuttle, and the raising marks for these are the short dashes on the
painting. Fig. 164 is the same pattern arranged for two threads face to
one of back, which admits of a heavy backing with a fine face, and is
suitable for quilts, &c. The markings are similar to those in Fig. 163,
but the number of threads is not such as to admit of the plain texture
repeating on the back cloth. It is therefore broken round the edges of
the figure where it would least show.

Instead of having the extra harness for working the plain texture as
explained in connection with the quilt harness, Fig. 159, it might be
done with four leaves of long-eyed heddles in front of the harness,
or by having shafts through loops in the harness to act as heddles,
or by having the cumber board divided into strips for each row of the
harness, and knots on the harness twines above them, so that raising
any strip will raise a row of the harness, or the machine may act as a
twilling machine to work the texture.

A Marseilles quilt is a double cloth with a fine face and a coarse
back, both stitched together round the edges of the pattern, and a
wadding shot thrown in between them, which causes the pattern to be
embossed or to rise up full and rich. The face is usually a plain
texture, and the back the same; but the face has double the number
of threads in it that the back has, and the yarn for the face is
proportionately finer than that for the back. Two picks of fine and two
of coarse weft are thrown in alternately--one pick of the fine goes
into the face cloth only, and the other, also a face pick, stitches
the face to the back. One of the coarse picks is thrown in between the
two fabrics, and is called the wadding, and the other pick goes into
the back cloth. In these fabrics a very fine effect can be produced by
covering the ground with a small bird-eye pattern, stitching the back
and front closely together, then having a bold floral pattern for the
figuring, which, being only bound round the edges of the pattern and
along veins, &c., is comparatively loose and produces a rich embossed
effect. These fabrics may be woven with a twilled as well as a plain
ground.

Matelasses are a similar description of fabric, but usually more
ornamented, the face fabric being any fancy texture. When used for
ladies’ jackets or mantles they are made of fine worsted or silk for
the face, with perhaps a woollen back and a woollen or cotton wadding.
The binding of the face and back cloths together takes place round the
edges of the figure, and in any other portion, such as veins of leaves,
&c., that may be thought desirable. The binding may be effected either
by leaving down a portion of the face warp when the back pick is being
thrown in, or by raising a portion of the back warp when the face pick
is being thrown in, this depending upon the pattern and the counts
of yarn used. The finer threads should be used for stitching, so as
not to injure the face by bringing any coarse threads through to it.
If the face of the cloth is a warp pattern it will be best to stitch
by raising a back thread over a face pick, but if the face is a weft
pattern then sinking the face warp for the back picks will probably
be more satisfactory. If the cloth is made, as is frequently done,
with a mohair or lustre worsted warp face with a cotton weft, and a
cotton warp back with a woollen weft, and a woollen wadding, then the
binding would be effected by raising some of the back warp thread over
the face picks. One pick (woollen) for wadding would be thrown into a
shed formed by raising all the face warp; the next pick would be for
the back, thrown into a shed formed by raising all the face warp and
that portion of the back warp required to form the texture of the back
cloth; the third pick would be for the face, thrown into a shed formed
by raising the portion of face warp required for the pattern, as well
as those ends of the back warp that are to form the stitchings.

_Quiltings._--A great variety of the cheaper description of quiltings,
toilet covers, &c., consist of a face cloth with an additional thick
weft. This thick weft is woven into the ground, which may be a twill,
mat, &c., but flushed loosely at the back of the figure, which is a
plain texture. Sometimes there is a small quantity of thin warp for
binding the thick weft loosely at the back of the figure, making a
double cloth in this portion, but all working together for the ground,
making it a solid cloth.

Of more recent origin are the satin quilts made by Messrs. Barlow and
Jones, of Bolton, and some other firms. They consist of two plain
cloths, intimately bound together; one cloth has a fine warp and a
very coarse weft, and the other has a fine weft and a coarser warp.
When binding, the fine warp is made to catch on the fine weft, and
the thick warp and weft cover the bindings. Suppose the fine warp to
be white and the thicker warp to be blue, and let the warping be two
threads blue and one white. Let the white warp be wefted with a thick
twist weft--say four times as thick as the warp--and this coarse weft
will form the figure. Let the blue warp be wefted with a blue weft
perhaps double the count of its warp, and in binding this weft catches
on the fine white warp. The thick white weft, which is fully double
the thickness of the thick warp, effectually covers the tie, and when
the yarns are properly proportioned the pattern stands out, producing
a clear stamped or embossed effect. These cloths may be made with a
twilled ground, and either all white or white and coloured. They are
a good firm fabric and wear well, but are liable to have a coarser
appearance after being washed. (See also Terry-Pile Quilts.)

_Woollens and Worsteds._--Weaving woollen and worsted cloths in the
jacquard loom is merely an extension of the patterns produced with
shaft mountings, or sometimes the same patterns are woven on small
jacquards by those who prefer the jacquard to working a larger number
of healds when the cloth is not so heavy as to require the latter to
be used, healds as a rule making a firmer and heavier cloth than a
jacquard will. Light worsted goods for dress fabrics, &c., are figured
like damask, or as double weft or double warp-faced cloths, or may be
as double cloths.

_Curtains and Tapestries._--These, though sometimes of damask, are
usually made on some principle of double cloth, as indeed all cloths
must be when a rich brilliant effect of colours is required. One of
the simplest methods of making these fabrics, and which produces a
very good effect, is to employ a fine binding warp of twist cotton
and produce the pattern on it with two, three, or more coloured wefts
of worsted, mohair, or silk--say a spun silk ground weft and a mohair
or fine worsted figure. A small portion of a diamond pattern, greatly
reduced for want of space, is given in Fig. 165. It is intended for
three colours of weft--one for the ground and two for the figure. The
ground weft might be black or gold, and the figure wefts olive green
and dull red, or claret, the warp to be black or a deep navy blue.

For this pattern three cards must be cut from each line of the design
paper--one for the ground weft and one for each colour of the figure.
The design is painted in various colours, which are here represented
by different markings on the squares. Let the black squares represent
the olive and the dotted squares the red of the figure. The white
squares represent the ground weft, and the shaded squares on it are
for binding the face weft down, while the crosses are for binding up
the weft at the back. The cutting of the cards is as follows:--For the
ground shuttle cut all the shaded, black, and dotted squares--that is,
the shaded squares on the ground and all the figure. For the first
figure shuttle (olive) cut all the ground except the crosses (that is,
the white and shaded squares) and the dotted squares of the figure. For
the second figure shuttle (red) cut all the ground except the crosses,
as before, and the black squares of the figure.

[Illustration: Fig. 165]

In this example the figure is so small as not to require binding dots
over it, but for a large figure the ground weft threads would require
to be bound up at the back of the figure in the same way that the
figure threads are bound up at the back of the ground in the design
given. The figure wefts might also require to be bound at both face and
back. On the design these binding marks would be dotted over the figure
in the same way as they are dotted over the ground on the design given,
using any suitable texture that may be desired.

To make the cutting clear, five lines of the pattern (Fig. 165) are
put upon design paper in Fig. 165A, as they would be cut on the cards,
beginning with the ground weft, olive and red following, in all making
15 lines or 15 cards. This cutting would make the upper side of the
cloth in the loom the right side or face; but it may, in some cases, be
desirable to weave the cloth with the face down to avoid heavy lifts in
the harness. This pattern would suit for a warp of about 50 threads of
warp per inch, and say 60 to 80 shots per inch of each colour.

[Illustration: Fig. 165A]

For the convenience of the designer and card-cutter, a much better plan
of working this class of fabric is to have an additional binding warp,
which may be in heddles or in a front row of the harness. This warp
can work plain twill or satin as required, and, being light and openly
set, can bind the back weft up and the face weft down, or the face may
be bound by the warp in the harness and the back bound by the binding
weft; but in this case there should be a fine thread of weft thrown in
every fourth pick, which should pass over the warp in the harness, or a
portion of it, and under that in the heddles, so as to bind both warps
together. In this case the twilling dots on the design are omitted,
which simplifies the cutting and designing of the pattern. If the face
is to be twilled with the harness warp, the dots for binding must, of
course, be put on the design, or the figure only may be twilled on the
design, and let the ground be bound by the binders. Shafts mounted in
this way may be wrought from tappets if the loom is mounted with them,
or may be wrought by strong hooks at each side of the jacquard and sunk
with springs. The principle of working fabrics of this description has
been given in detail, as from it many other varieties can be wrought
out, and any number of shuttles used to suit the colours wanted.

Figuring with two warps and one weft is a common method of working, and
gives three colours on the face of the cloth. An extra warp of fine
yarn may be used for binding the wefts, admitting of the coloured weft
being more used for figuring. Extra warps cannot be used to so great
an extent as extra wefts, as they crowd up the reed and prevent proper
shedding, particularly when soft or hairy yarns are used; but when
properly suited to the reed they make a firmer and more regular cloth
than a weft cloth, and on account of less picking the weaver gets over
the work much faster. Two warps, each having its own weft, are a good
method of figuring, but unless for goods with a large number of threads
and picks per inch, so that fine yarn can be used, it makes a heavy
cloth. Two warps and two wefts, all of different colours, with a fine
binding warp in addition, to admit of the colours of the wefts being
kept comparatively pure, gives a still richer effect. In this case
the binding warp may be of fine black cotton, the two figure warps of
thicker cotton, and two threads drawn into each mail; the colours may
be, say, dark blue and dark citrine, or clear sage. The wefts are soft
wool or worsted, say light gold and deep dull red. In this method of
weaving, pure red and pure gold can be got from the wefts, as the black
binders show but little. Pure blue and pure citrine can be got in small
quantities by floating the warps, and an admixture of the warps and
wefts can be got in any place desired.

A very handsome curtain fabric can be made as follows:--There are 120
to 140 threads of warp per inch; every second or every fourth thread of
the warp is used for binding the ground, which may be a plain texture
or a three-shaft twill. The binders may be an extra warp, wrought with
heddles, forming plain all over the cloth, or may be in the harness
and be portion of the ordinary warp working plain for the ground, but
rising to assist in forming the figure, which may be bound as desired,
say 8-or 10-end satin. The warp is of fine silk yarn loosely twisted,
and may be one or more ply, of a rich olive-green or deep scarlet
colour. There are two wefts, one a rich silk, say a golden colour, of
twist yarn; the other is a backing weft of the same colour as the
warp, and of cotton yarn, about the same thickness as the silk weft.
There are from 50 to 60 threads of weft per inch of each colour. The
gold weft forms the ground of the pattern, and the warp the figure,
which is bound with an 8-end satin binding. The cotton weft goes to
the back when the silk weft is on the face, forming the ground of the
pattern, the warp lying between them; but when the warp is on the face
for figuring, the gold weft goes to the back and the cotton weft lies
under the warp face, binding it and giving an embossed effect to the
figure. The gold weft is bound up by sinking a portion of the warp. The
ground may be formed with the warp, and the figure with the weft, if
desired, producing a sunk figure on a raised ground; but this is just
a reversal of the process, or making the ground in this case as the
figure in the previous case. This makes a light, close, and very rich
fabric.

_Figured Poplins_ are among the richest description of curtain fabrics;
they may be made of all silk, or silk and fine wool, the latter forming
the weft. The ground is a clean, sharp cord running across the cloth,
and the figure is formed by flushing the warp over the cords, binding
it with a long twill or satin binding. The weft may also be used for
figuring; but in this case it should be a silk weft put in for the
purpose of figuring.

Two portions of patterns of good makes of figured poplins or repps are
given in Figs. 166 and 167. Fig. 166 is warped two threads of rich
crimson silk and one thread of rich golden yellow silk, 180 threads per
inch. The weft is pick-and-pick, 50 picks per inch, one thread a round,
firm cord of firmly twisted worsted of the same colour as the warp, and
the other yellow silk, the same as the warp. This thread passes over
the crimson warp and under the yellow, the crimson cord reversing this,
thus producing a very fine yellow line between each pair of cords.
The method of forming the figure can easily be seen from the pattern.
Instead of a thick cord being put in for the weft to form the rib,
several fine shots could be thrown in as one, as in repp figuring, and
these threads could be brought out for figuring as well as the warp;
but they would in this case require to be of silk, which would make the
cloth very expensive.

[Illustration: Fig. 166]

[Illustration: Fig. 167]

Fig. 167 makes a very rich and handsome fabric. It is warped all a rich
purplish-brown silk, 180 threads per inch. There are four weft picks
in the pattern. The first is a clear olive-green silk about double the
thickness of the warp silk, the second shot is a round, firmly twisted
worsted cord of the same colour as the warp, the third is the same as
the first, and the fourth is silk of the same colour as the warp. There
are 100 picks per inch, but the two green silks go along with the cord
to form the complete cord or ridge. These silk picks can be brought
out to the surface for figuring, as shown by the shaded squares in the
pattern, which are not cut on the cards. The black squares are the warp
figuring, the dots the warp raised for the ground, and the crosses the
warp of the binding threads raised. These binding threads rise over the
brown silk pick, and are similar to the yellow warp in the previous
pattern.

For richer fabrics than these we must go to pile work, which is,
perhaps, the most exquisite production of the loom. Curtains or
hangings can be made extremely rich by figuring a rich corded silk
ground with a pile of different lengths and colours. The long pile can
be cut to form a plush or velvet pile, and the shorter pile may be left
uncut to form a looped or terry pile. By a judicious arrangement of
colours and length of pile, fabrics of extreme beauty can be made (see
Pile Work). For curtains of a heavy description chenille is much used,
and makes a very rich, warm-looking fabric; but it is too heavy and of
too coarse a nature for small rooms unless they are very well lighted.
It is very suitable for screens, or curtains dividing two portions of a
room (see Chenille).




CHAPTER IX

_TAPESTRY AND PILE WORK_


_Tapestry._--What are generally known in trade as tapestries, are
figured fabrics for curtains, hangings, &c., not damask, which is
distinct, being woven with only one warp and weft. Real tapestry is
a hand-made fabric of very ancient origin, being in use since the
days of the ancient Egyptians, who wove or worked it in a manner very
similar to that employed at the present time. It does not belong to
‘jacquard weaving,’ but being the first in point of the excellence of
its patterns amongst figured fabrics, a short description of it may
not be out of place. Henry VIII. tried to establish tapestry weaving
in this country, but failed, but James I. had a flourishing factory at
Mortlake. Tapestry weaving appears to have been introduced into France
about the ninth century. The Flemish were celebrated for it from the
twelfth century. Arras work surpassed all the others, and tapestry
was commonly known as arras work. The sixteenth century gave a new
impulse to the trade in France. Francis I. founded the manufactory of
Fontainebleau, and Henry IV. re-established tapestry making in Paris
in the years 1595 to 1606. About the year 1666 Louis XIV. bought the
Gobelins Works (which take their name from the original owners, a rich
family of wool dyers), and established the Royal Gobelins Tapestry
Manufactory, which is now one of the sights of Paris.

There were two methods of working tapestries--one known as ‘basse
lisse,’ or low warp, the threads of warp being placed in a horizontal
position in the loom; the other was called ‘haute lisse,’ or high warp,
as the threads of warp were placed in an upright or vertical position.
It is the latter method of working that is now adopted.

The loom consists of an upright frame of wood of a size to suit the
cloth to be made. There is a strong roller at the top, which acts as
a warp beam, and another at the bottom for the cloth beam. Both these
rollers have ratchet wheels on one end, and are held with pawls or
catches so as to allow the warp to be wound off and the cloth to be
wound up when required, and then hold the stretch of warp steady to
be woven. The warp is coarse, but a clean, regular thread of twist,
cotton, or linen, about 12 to 18 threads per inch (more or less, as
desired,) and of such a thickness that the spaces between the threads
are somewhat less than the diameter of the threads. The weft is usually
of fine wool but a fairly thick thread, say 4’s or 6’s worsted. This
is generally used alone, of whatever shade of colour is required,
and there is no end to the shades used; but in some places, to give
brilliancy and richness of effect, silk is used along with the wool,
and sometimes tinsel or gold cord. These are put in separate threads
along with the woollen or worsted weft. The silk is much finer than the
worsted, say about equal to 12’s or 16’s cotton, and, of course, can be
regulated to give the effect required. The gold is sufficient to give
sparkle to the portion it is employed in.

After leaving the warp beam the warp is divided by two thick glass
lease rods. Then every alternate thread, those to the front of the
frame, is drawn through a doup of a half-leaf of heddles which is fixed
in a horizontal position above the weaver’s head. This enables a plain
shed to be formed, as the thick lease rod divides the threads and forms
one shed. Then, when the weaver, sitting at the back of the loom, draws
back the half-leaf of heddles, all the front threads or any portion of
them can be drawn back past the back set, and form a cross shed. As
the doups of the heddles are long and strong, the weaver can take any
number of them he requires and draw the front warp back; then putting
his fingers into the shed thus formed, clear it down to the fell of the
cloth and insert his bobbin.

[Illustration: Fig. 168]

The pattern is painted on design paper, shaded in the way it is to
appear on the cloth, and the weaver must have a bobbin of weft to
suit each tint or shade of colour on the design. The outline of the
pattern is traced on the warp to guide the weaver, and the painted
pattern is fastened up either before him or to one side of him, so as
to be convenient to read the stitches off it. When ready for work he
looks at the pattern and finds perhaps 20 stitches of blue; then he
selects a bobbin of the correct shade and passes it through the open
shed of the warp round 10 threads; then he draws the 10 doups of the
heddles that are round the 10 threads intermediate with and in front of
these, pulling them back through the others and forming a cross shed.
The same bobbin is now passed through this shed, and the weft pressed
down with a pin or comb, both of which articles the weaver uses for
straightening and beating up the weft. This covers 20 threads of warp
with blue weft. The texture of the cloth is plain throughout, the weft
being bent round the warp and pressed close together, forming a rib.
Now, instead of completing this weft line all along the cloth, as is
usual in coloured weaving, the weaver goes on working with the bobbin
he has taken up, following the colouring on the pattern, and may change
his bobbin when required for a new shade or colour, and go on working
this portion for some distance upwards, and then begin an adjoining
portion and work it up in the same way. This will come all right when
the outlines of the portion he is working run obliquely; but if they
should run upright for any distance in the same direction as the warp,
it is plain there would be two selvages coming together without any
connection between these portions of the cloth, as shown in Fig. 168.
When the pattern runs in a diagonal direction, the weft threads passing
across from one warp thread to the next one make a sufficient binding
or connecting link between the two colours, as may be seen at the upper
portion of this figure, which would be quite sufficiently connected
when it is considered that the weft threads are pressed so closely
together as to entirely cover the warp. To remedy the objection of a
slit or division between the two portions of a pattern in the case of
a vertical line, it is customary to tack the bobbin of one colour of
weft round the first warp thread of the adjoining colour, at intervals
of, say 3/16 in. This produces a toothed appearance, but is not very
remarkable. The only other remedy would be to link the two wefts on
each other occasionally. In many of the old tapestries it may be seen
that this was neglected, even when the distance was as much as 3 in.
or 4 in. Sometimes they were stitched afterwards in these places with
a needle and thread. When there are many long straight lines running
in one direction, as in architectural subjects, it is best to work the
pattern so that these will run in the direction of the weft, to avoid
any joining being required. Patterns can be wrought sideways or upright
on the cloth as desired to suit the size, subject, &c.

[Illustration: Fig. 169]

As the weaver sits at the back of his frame or loom when working,
he has to come round occasionally to examine the face of his work.
Sometimes he keeps a looking-glass in front to show him how he is
getting on. He has also a rough pattern on the back of the cloth
similar to that on the front, as the nature of the weave must give him
the same on both sides were it not for the loose ends at each change
of bobbin, which have to be tacked up at the back. Tapestry weaving is
a very tedious process of working, but in skilful hands very beautiful,
and artistic results can be produced.

_Figured Pile Fabrics._--When we come to pile work we have a great
scope for the ornamentation of fabrics, but the nature of the work is
such that it is only suited to heavy or moderately heavy cloth, and a
large number of either threads or picks per inch are required.

Forming stripes of plain or figured pile alternately with twilled
or figured stripes can easily be accomplished with either a weft or
warp pile, but when we come to figure indiscriminately over the cloth
there is more trouble. For a length of time there was a difficulty in
cutting weft-pile figured fabrics, as the knife or cutter would not
always enter the proper ‘race’ round the edges of the figure, and the
floats were cut irregularly. This has now been overcome by arranging
the design to suit. Fig. 169 is a common velveteen designed for a weft
pile figure. Only a portion of the pattern is given, but enough to show
the method of arranging the design. It will be seen that the figuring
is produced by flushing the weft on one side of the cloth, and for the
ground it passes to the other side. The edges of the pattern are moved
across the design paper in steps of two, and no flush or float is less
than over four threads. In addition to this the figure should be turned
in the centre on an odd number of ends, which in cutting keeps the race
end on the inside of the step into which the knife enters. The knife is
run up on all the odd numbers of ends moving across in steps of two,
and, leaving no flushes of less than four, gives it a better chance of
entering in the proper place along the edges of the pattern.

Fig. 170 is another make of velvet for a heavier fabric, with 60
threads of warp per inch, and from 350 to 400 picks. It makes a good
firm cloth, suitable for curtains or furniture coverings. The principle
of designing is the same as for Fig. 169. Both these patterns are
portions of a diamond figure.

When the figure is produced with a warp pile, it is only necessary to
raise the proper warp threads with the jacquard for the insertion
of each wire, but the difficulty to contend with is that there will
be an irregular take-up of the warp pile threads, and this can only
be remedied by having a number of small warp beams. In some cases a
separate spool for each warp thread may be required, which, of course,
for a fine velvet means a very intricate arrangement, though it suits
very well for pile carpets. Different heights of pile can be formed
through a pattern in this way--say, a short pile to be left looped, and
a longer pile to be cut, which, in suitable colours of pile and ground,
forms a very beautiful though costly fabric.

[Illustration: Fig. 170]

Another method of warp pile weaving, more easily accomplished, is
that known as terry, or Turkish towel weaving. It is largely used for
quilts, toilet covers, towels, &c., and any bold figure can be woven on
it with almost as much ease as on ordinary damask work.

There are two methods of forming figures on the cloth in this style of
weaving, one by having two colours of pile warp and changing them from
one side to the other to form the figure or ground--that is, supposing
the one pile warp to be red and the other white--pile would be thrown
up on both sides of the cloth at the same time; on one side the figure
would be formed by the red pile and the ground by the white pile, and
the other side would be the reverse of this. Fig. 171 is a portion of
a pattern of this style. The black squares represent, say, the red
warp-forming loops on the face of the cloth, and the shaded squares the
same for the white warp. The crosses are the ground or binding threads;
and the dots represent the pile threads of the colours they are on,
raised to bind with the weft, when these threads are forming pile on
the underside of the cloth.

[Illustration: Fig. 171]

The other method is to have only one colour of pile warp--in fact,
to have only one pile warp, it might be said, instead of two, as in
the previous case. The cloth consists of a pile figure and a plain or
solid ground on one side, and the reverse of this on the other side.
The pile and ground may be of the same colour, or of different tints
if desired--say a rich cream ground and a white pile. This style of
working is very suitable for quiltings, toilet covers, &c. Fig. 172
is a portion of a pattern for weaving in this way. The black squares
represent the pile-warp raised to form loops, and the round dots are
also pile-warp raised, but only for binding when the loops are being
formed at the other side of the cloth. The shaded squares show the
ground warp raised to form the body texture. This cloth has about 60
threads of warp per inch, and the same or a little more weft.

[Illustration: Fig. 172]

Figs. 173 and 174 are two examples of six-shot pile cloth for quilts;
it is made with 50 to 60 threads of warp per inch, and double that
quantity of weft. The same method of marking the design paper is used
as that for 172 pattern. It will be seen that in Fig. 173 the loops or
flushes of pile are over five and under one, and in Fig. 174 they are
over two, under one, over one, and under one, thus making a fast pile
fabric, whereas the floats of five in Fig. 173 would be rather loose
unless the cloth is over-wefted. Both these piles may be used in the
one cloth, one for the face and the other for the back, as is given in
the figures; the light portion of Fig. 173 being of the same texture as
the dark portion of Fig. 174. One pattern may be taken as the face of
the cloth and the other as the back. Of course they may also be used
separately if desired. Any full-harness mounting that will suit the
pattern will answer for these fabrics, the loops being thrown up in the
usual way, by leaving a few shots standing out from the fell and then
knocking all up, drawing forward the pile warp, which is slackened at
this beat, but sliding on the ground warp, which is held firm.

[Illustration: Fig. 174]

[Illustration: Fig. 173]

Another method of forming figure pile is to weave a plain pile, and
when cut press or stamp a pattern on it with heated blocks cut to the
pattern; the standing pile is then shorn off, and afterwards that which
was pressed down is brushed up again, forming a full-pile figure on a
shorn-pile ground.




CHAPTER X

_CARPETS_


The manufacture of carpets has been in existence since the days of
the ancient Egyptians, who made rugs and carpets of various kinds and
ornamented them with animals and various devices. The manufacture
passed on to the East, and we find India, Turkey and Persia celebrated
for the richness of their carpets in early times. The Moors of Spain
introduced the Oriental floor coverings into Western Europe, and the
Belgians and French took up the manufacture and produced excellent
imitations of them. Even in the Middle Ages carpets were only used
as a luxury by the rich. Queen Elizabeth had a carpet spread upon
rushes, while Queen Mary had rushes only. These rushes were the _Acorus
Calamus_ (sweet reed of Norfolk), which were the usual floor coverings
down to about this time.

Tapestry weaving was started in this country in the beginning of the
seventeenth century, but the first we hear of carpet manufacturing in
England was at Kidderminster about 1735, the carpets then made being,
no doubt, the Kidderminster or Scotch carpet. The making of Brussels
carpet was introduced about 1750; they were first made in this country
at Wilton, near Salisbury.

The Axminster carpets take their name from being first made in the
town of Axminster about 1755. They rivalled the work of the Eastern
looms for beauty, durability, and colour; but owing to the cost of
their manufacture, and the introduction of the cheaper Brussels and
patent tapestry carpets, the demand became very small, and the work was
removed to Wilton in 1835. The patent Axminster or chenille carpet was
brought out by Messrs. Templeton, of Glasgow, in 1839.

In the early ages carpets or rugs were used for spreading on the floor
or grass to lounge upon. It is recorded that in Babylon the guests of
despotic sovereigns lounged on rich carpets, and walked over priceless
works of textile art; also how figured carpets made of the finest wool
were strewed under golden couches, and rugs richly embroidered with
figures were spread over daïs, stool, and table.

The Egyptian carpets may have been tapestries and embroidered fabrics.
They also had a method of making tufted carpets by drawing a portion
of the weft threads out of a piece of coarse linen and sewing tufts of
coloured worsted to the warp threads, enough of the weft being left in
to bind the warp threads together.

[Illustration: Fig. 175]

Persian and Indian carpets were made of wool; the latter were
occasionally made of silk, and sometimes an inferior description of
carpet was made of cotton. These, as well as Turkey carpets, are made
on a very simple loom consisting of two posts fitted at a suitable
distance apart to form a vertical frame. There is a roller at the top
and another at the bottom between which the warp threads are stretched,
much in the same manner as in tapestry weaving. The weaver sits in
front of the loom with a design before him, and is provided with a
quantity of bobbins of the colours required for the pattern; he looks
to the design for the colour, and, taking a bobbin of the colour
required, forms a loop round two of the warp threads with the weft,
cutting it off as shown in Fig. 175. After having completed a row along
the cloth forming one weft line of the design, he opens a shed and
inserts a ground shot all across the web, each ground shot going into a
shed the reverse of the preceding one, forming a plain texture ground
with a row of tufts between the ground shots. The ground shots are
beaten down with a comb. Instead of going across in even rows of tufts,
where there is a patch of colour several rows of tufts may be put in at
this place with a ground thread between the rows, leaving the ends of
these threads projecting so that they can be carried across when the
remaining portions of the rows are completed. The ends of the tufts
are cut off roughly at first, and afterwards shorn level with a pair
of shears. Persian carpets are sometimes very costly, a small-sized
carpet, when made of fine cashmere wool, costing 500_l._ or 600_l._

Axminster carpets are made on the same principle, and other art carpets
are being made similarly. There are about sixteen or twenty warp
threads per inch of strong cotton or linen thread; the weft is of fine
wool, three or four ply being put into each tuft; the ground weft is
soft hemp or flax, eight ply of yarn going to each shot. This makes
a full soft cloth. These carpets are about three-quarters of an inch
thick, and, like tapestries, there is no end to the variety of pattern
or colour that can be introduced.

_Kidderminster or Scotch Carpet._--This carpet, also called ingrain
carpet, is the oldest machine-made carpet in this country; it was
originally made at Kidderminster, but the chief centre now is Scotland.
Originally it was a two-ply cloth, the pattern being formed by passing
the two cloths, which were of different colours, through each other.
Mr. Morton, of Kilmarnock, improved on this by making it a three-ply
cloth, which enabled it to be made a thicker cloth with a richer
pattern. It is now made both two-and three-ply, and when made of
all-wool is a rich and durable article, taking a position between the
jute and felt carpets and the tapestry and Brussels; it is, however,
sometimes made with cotton warps and woollen wefts, and is in this case
an inferior article.

Both two-and three-ply carpets may be woven with only one colour of
weft, in which case the pattern is formed by the warp threads, which
must be much thicker so as to close in over the weft and hide it as
much as possible; on the other hand, there may be two warps and four
colours of weft, two of which are the same colours as the warp, and in
this case the weft is much thicker than the warp. The best of these
carpets are made with as many colours of weft as warp, as, in order to
get a pure effect, it is necessary to have wefts traversing warps of
their own colour; but a variety of effects can be produced by using
additional colours in both warp and weft.

One advantage in this style of carpet is that it is reversible; for
when one cloth rises to the face the other passes to the back, making
the pattern on both back and face alike, but of different colours. This
is the case when the carpet is two-ply cloth woven in the ordinary way,
but sometimes the back has to a certain extent to be sacrificed to
give more ornamentation to the face, at least in the ordinary method
of weaving. Whether woven two- or three-ply, the principle of working
is the same. Take a two-ply carpet. The fabric is a double or two-ply
plain cloth, figured in the usual way by passing the two cloths through
each other, there being no binding between the cloths except what is
formed by the one passing through.

The warps are of different colours--as, say, scarlet and black,
green and black, &c.--and each warp should have a weft of its own
colour if pure effects in the cloth are required. Usually four sets
of warp threads are employed instead of two--that is, two colours,
end-and-end, for each warp--and each colour of warp has its own colour
of weft. Say we take red and black for the face or figuring warp,
and white and olive for the back or ground warp. Various effects can
be produced from this arrangement--viz. the effect of the figuring
cloth, formed by weaving the red and black warps and wefts together,
which in plain texture with one weft will produce a rich brown effect;
the ground cloth produced by the white and olive warps and the weft
will be of a light olive colour; then lined, or what are known as
‘shot-about’ effects can be produced by throwing in a light and a dark
shot alternately--as white and black, olive and red--and binding them
on the face by warps of the same colour. All the weft on the face of
the cloth should be bound by warp of its own colour in order to give
pure effects. When a thread of one cloth is raised, the corresponding
thread of the other cloth goes to the back (white and red and olive
and black correspond in this case), but the red weft will be bound
with black warp on the under side of the cloth, and the black weft
will be bound with red warp, and the same with the others, as in the
system of working the harness with journals to form the texture, all
the black warp is controlled by one of the journals and must all rise
with it. Therefore when the black shot is being thrown in, the black
warp must be raised to make the pattern correct on the face of the
cloth, and this will also cause it to be raised for that portion of
the cloth that goes to the back, and when the black is up the red is
down; therefore the black weft must be bound by red warp on the under
side of the cloth, and the same with the other colours. It is easy to
understand that if an end-and-end warp--black and red--be wrought plain
on two shafts or leaves of heddles, and shot pick and pick of the same
colours, when the black weft is bound with the black warp on one side
it must be bound with red (which is the other half of the warp) on the
other side, and if the red is bound with red on one side it must also
be bound with black on the reverse side. To apply this to the double
cloth or carpet weave it is only necessary to consider this piece of
cloth as passing up and down through another one of a different colour.
Had we the power of raising any portion of the black or red warps
required, we could bind a portion of the black weft with black on one
side, and any other portion of it with black on the other side; to do
this we require a full or thread harness.

[Illustration: Fig. 176]

A section through the weft of a piece of two-ply carpeting is given
in Fig. 176, showing the position of the threads when arranged to
give the different effects of face cloth up as at A, back cloth up
as at B, shot-about effect of red and olive on the face as at C, and
shot-about effect of white and black on face as at D. The warp threads
are marked W, O, R, B, the initial letters of the colours; and the
numbers 1 to 32 give the order of picking. The shot-about effect is
here produced with the opposite threads--that is, the first thread of
one warp and the second thread of the other warp--as white and black;
but it might be preferred to produce the effect the corresponding or
mate threads (white and red, or olive and black) would give. In working
with journals, two picks face and two picks back would be required to
obtain this, and the needles of the jacquards would require to be acted
upon by the cards for each shot, thus requiring double the number of
cards. With a full harness any effect desired can be produced. Extra
colours of weft may also be used (as drab) in addition to the white
and olive--to work, say, white, olive, drab, olive, &c.; and dark
green or blue in addition to the red and black, which would come in as
red, green, black, green, &c., a shot of the light and one of the dark
colours following each other in succession. Stripes of a bright colour
may be introduced into the warp or weft and kept in the back cloth,
only being brought to the face in small portions, at intervals, to
sparkle up the centres of flowers, &c.

Ingrain carpets are made 36 in. wide, and with, say, 832 or 1088
threads of warp, according to quality. The former can be woven on a 200
jacquard (say 208=416), as the machine is double; and as there are two
repeats of the pattern in the width of the cloth, 416 × 2 = 832 threads
of warp. For the latter 272 needles are required = 544 × 2 = 1088 hooks.

When woven with journals, the jacquards used for figuring these carpets
are similar to those explained under ‘Quilt Weaving’--that is, double
machines with trapboards and knot cords; and when the card is pressed
against the needles it is held in for two shots, one trapboard raising
the harness for the first shot, and the other reversing the shed for
the second shot. The journals are four sections of the cumber board,
as shown in Fig. 177; each section contains two rows of the harness
and one colour of warp. No. 1 journal controls the white warp, No. 2
the olive, No. 3 the red, and No. 4 the black; the draft is shown by
the numbers 1 to 8 alongside the harness cords. These eight cords only
represent half a row from each machine, as there are eight rows of
hooks to the machine. One row of the two machines fills two rows across
the journals, or 16 holes. A represents the trapboard of the machine
for the ground or dark warp, and B that of the machine for the figuring
or light warp.

The harness twines are furnished with large knots above the cumber
board or journals, so that when one of the journals is raised the
harness will be lifted with it. The journals may be raised by strong
cords or wires from the machine, or by a tappet or shedding motion
working either above or below them. If the machine is divided into two
parts, four hooks in succession going to each colour of warp for each
cloth, the cords from these four hooks making two rows of the journal,
there will be no crossing of the harness in a straight or Norwich tie.

[Illustration: Fig. 177]

The order of working for the section Fig. 176 would be as follows:--For
the first pick, which is white, raise the trapboard B and the first
journal; for the second pick, which is red, raise the trapboard A and
the third journal; for the third pick, raise the trapboard B and the
second journal; for the fourth pick, raise the trapboard A and the
fourth journal. This is the general order of lifting for a pattern; but
a simple lined effect across the cloth as the section could be wrought
with the journals without the machine, using them as heddle shafts.
It could also be wrought by the machine without the journals. The use
of the journals in conjunction with the machine is to pass the cloths
through each other and make a pattern. For figure work, when a card
is pressed in against the needles, all the warp for the figure on the
design paper is raised by the trapboard B, which clears this portion
of the red and black warps out of the way of the ground shed; and as
the first pick of the ground is white, the white journal is raised to
bind the white pick, leaving the olive warp down to bind it at the
back. For the next shot the same card acts, and the trapboard A is
raised, which lifts all the white and olive warp corresponding with
the portion of the red and black that was left down at the last shot,
so as to clear this portion out of the way of the face shed. As a red
pick follows white, the red journal (No. 3) is raised to bind the red
on the face, leaving the black warp down to bind it on the back. The
other picks follow similarly. This shows the defective binding common
to journal weaving, the white being bound with olive and the red with
black at the under side of the cloth. To have pure binding, instead of
all the white or red being raised by a journal as above, only as much
of either should be raised as is required to bind its own colour of
weft on the face of the cloth. The olive or black in respective cases
should be raised for the remainder of the shed, leaving the white and
red down to bind their own wefts where these colours come out on the
back or under side of the cloth, thus giving pure colours on both sides
of the cloth, and letting the mixed or impure binding be in the centre
between the two cloths. This, however, is beyond the range of journals,
and requires a harness.

[Illustration: Fig. 178]

Fig. 178 gives a neat pattern on a small scale for a two-ply cloth
with, say, a red and a pale olive warp for ground and figure cloths
respectively, the shaded squares being red and the ground or white
squares olive; the weft for each cloth to be the same colour as the
warp, or sometimes a tint one or two shades lighter and brighter,
or deeper and duller, according to the colours or shades of colours
used, gives a good effect. A further effect in this can be produced by
having the 6th to 10th and 18th to 22nd threads black, brown, or dark
green, end-and-end with pale olive, for the back warp, and picking
similarly on the 5th to the 9th picks for the ground cloth. This would,
of course, make a striped effect on the under side of the cloth, as
where the additional colours do not show above they must appear below.
A still better effect can be produced by having an additional weft
lying between the two cloths which can be flushed on the face in spots
as at picks 12, 13, and 14, and of course a three-ply cloth will be a
step farther in advance; lined effects can then be produced anywhere
desirable, in portions to suit the nature of the design, to vary and
enrich it, at the same time the breadth of effect is maintained by
working the greater portion of the pattern in masses of pure colour,
whether they be large or small. For full-harness work the harness is
divided into two sections, four rows for the back cloth and four rows
for the face cloth, on an 8-row machine; no journals are used, the
machine doing all the work. The harness is tied up in the usual way,
the cord from the first hook passing through the first hole of the
cumber board, and the cord from the second hook passing through the
second hole, and so on.

The harness is nominally divided into two sections, the front four rows
being for the face cloth and the back four for the back cloth; when
drawing the warp into the harness the mails are taken in the following
order: 1, 5, 2, 6, 3, 7, 4, 8, the back or ground warp coming on 1, 2,
3, 4 mails, and the face or figure warp on 5, 6, 7, 8.

Suppose the pattern to be a dice, as Fig. 179, the warps to be white
and olive, and red and black, with the same colours of weft. Each
upright line of the design represents two threads of warp, one of which
will be drawn on the front half of the harness and the other on the
back half; and each horizontal line of the design represents two weft
shots, for which there must be two cards cut. Let the order of drawing
the warp into the harness be as Fig. 180. The white is on the odd and
the olive on the even numbers of the four back rows of the harness, and
the red is on the odd and the black on the even numbers of the four
front rows.

Now to cut the cards:--Take the first line of the design Fig. 179, cut
the odd numbers of holes (that is, the first and third) on the first
half, or first four holes, of the card, and cut the white and olive on
the second half of the card. For the second card, from the same line of
the design cut red and black on the first half of the card, and the
odd numbers of holes on the second half of the card. For the third
card, take the second line of the design and cut even numbers of holes
on the first half of the card, and white and olive on the second half
of the card. For the fourth card, from the second line of the design
cut red and black on the first half of the card, and the even numbers
of holes on the second half of the card. Of course when there is none
of a colour raised, as in the first line there is no black, no notice
is taken of it. For the pattern in Fig. 178 the cutting would be the
same, except that there are only two colours of warp to be dealt with
instead of four. No. 1 card: cut odd numbers of holes on the first half
of the card, and olive on the second half of the card. No. 2: cut red
on the first half of the card, and odd numbers of holes on the second
half of the card, &c.

[Illustration: Fig. 179]

[Illustration: Fig. 180]

[Illustration: Fig. 181]

The cutting of four cards is given is Fig. 181. The crosses indicate
the cutting of the odd and even holes in the cards, and the dots are
the figure. It will be seen that the crosses in No. 3 card would
reverse the shed formed by those in No. 1 card, thus working plain
texture on the back cloth, and the same would take place with Nos.
2 and 4 cards for the face cloth. This is simply the action of the
journals included in the harness, which it will be seen requires double
the number of needles in the machine, and double the quantity of cards;
but an ordinary machine and mounting will answer, instead of having to
get a double machine and journals.

The above cutting acts the same as journals, and is all right in both
cases when one warp with its own weft is used for each cloth; but it
has been pointed out that when each warp is end-and-end of different
colours with wefts to match, the under side of the cloth is defective
when wrought with journals, and so it would be with the full harness
if the cards were cut as above. To get each colour of weft bound with
its own colour of warp on both sides of the cloth, the following change
must be made:--Take the small piece of pattern given in Fig. 182, and
let the warps be as before--white and olive, red and black. The card
cutting for this pattern will be as follows, the warp being drawn
through the harness as before:

For the first card, cut odd numbers of red and even numbers of white on
the first half, or first four holes, of the card, and white solid on
the second half. This will form a shed for the red shot.

Second card: Cut from same line of the design red solid on the first
half of the card, and even numbers of white and odd numbers of red on
the second half of the card. This is for the white shot.

[Illustration: Fig. 182]

Third card, for black shot: Cut from the second line of the design even
numbers of black and odd numbers of olive on the first half of the
card, and olive solid on the second half of the card.

Fourth card, for olive shot: From the same line cut black solid on the
first half, and odd numbers of olive and even numbers of black on the
second half of the card.

These four cards are shown in Fig. 182, from which the cutting may be
traced. It must not be supposed, however, that cutting odd numbers of
red and even numbers of white as on the first card, means that these
colours of threads are to be raised; the colour has no reference to the
threads--only to the coloured checks on the design paper.

For the first card the white checks on the design paper are cut solid
on the second half of the card; this raises the white and olive warp
corresponding to the white portion of the design. Then, to form the
plain shed in the other warp, the even numbers of the white checks and
the odd numbers of the red checks are cut on the first half of the
card, which acts on the other warp. As the black warp is drawn on the
even numbers of harness twines it will be raised for that portion of
the design that is white, and the red will remain down to bind the red
weft; but passing along the design till the red weft is above, it will
be seen that the red warp should here be up to bind it, and the black
should therefore be down, so, as the red warp is on the odd numbers of
the harness and on the second half of the card, the odd numbers must be
cut on the card for the red portion of the design. This gives a pure
binding on both sides: and the same applies to the other colours. It
will be seen that the first halves of the 1st and 3rd cards together
work a broken plain texture, and the second halves of the 2nd and 4th
cards do the same.

In case the card-cutter cannot follow this method of cutting, and if
the design is so varied in order of weaving that a gamut or index for
cutting cannot be arranged, then it will be necessary to paint the
full texture of the pattern on the design paper in the same manner
as is done for double cloth. The first row of designs of the pattern
given in Fig. 178 is painted out in full in Fig. 183. The shaded lines
underneath represent the back warp, and the white lines the face warp.
The black squares are the figure, and the large crosses the ground,
or the white squares in Fig. 178. The dots are the rising marks for
the texture of the back cloth, as would be formed by the back warp
journals, and the small crosses are the same for the face warp.

This method of weaving is not confined to carpets, but is equally
applicable to quilts, curtains, &c., and a plain texture need not be
adhered to; a twill or any simple fancy texture may be used. With 80 to
a 100 threads per inch, one warp peacock green in satin texture for the
figure, and the other warp gold for the ground in a crêpe or mottled
texture, all shot peacock, a handsome curtain can be made.

[Illustration: Fig. 183]

_Brussels Carpets._--Brussels carpets, with their less expensive
allies, the tapestries, are the commonest of the better class of
carpets. They are a loop-pile fabric, the pile being formed by the
figuring warp, which is wool, the ground warp and weft being hemp or
flax. As the pattern is formed by raising the figuring warp threads of
the required colours, and as any thread of any colour may require to be
raised, it follows that all the threads must be on separate spools or
small warp rolls, so that any one thread can be drawn forward without
slackening any of the others. Large frames are therefore made to hold
the number of spools of each colour, and these are placed one above the
other in a slanting position, at the back of the loom, and when filled
with spools the whole set of threads are brought forward to the harness
as if from a warping creel or bank. According to the number of these
frames used the carpet is styled a 3, 4, or 5-frame carpet, the greater
the number of frames the richer the carpet both in colour (generally
speaking) and in body of warp. Sometimes 6 frames are used; but 4 or 5
are more frequent, 3 and 4 frame being the lower qualities.

The texture of a Brussels carpet is shown in Fig. 184, which is a
section through the weft. A and B are two of the ground warp threads,
a pair of these coming between the rows of pile loops, a portion of
one row being shown in the section. The weft threads are shown in
section coming under the loops of the ground warp at the top, and above
them below. Both this warp and weft are of coarse hemp, about 3’s: the
weft is steeped in glue size before being thrown into the cloth, which
makes it stiff and firm when dry. The pile warp threads are numbered,
1, 2, 3, 4, and 5; they are of soft-spun worsted, about 2/24’s, 2
and sometimes 3 threads being drawn in together as one through the
mails and reed. These five warp threads go to form one line of loops,
being raised over the wires as is required for the pattern, and lying
straight between the weft threads when not raised. By considering this
it will easily be understood that each thread must be on a separate
bobbin so as to admit of being drawn forward independently of any
of the others. In low qualities of these carpets one or two threads
of hemp warp are used to each splitful or row of pile threads; the
hemp thread lies in a straight line below the worsted threads in the
same way that No. 5 thread is shown in the section. These stuffing
threads add to the weight and thickness when the worsted is reduced in
quantity, but make a harsher and stiffer carpet. This warp, when used,
is put on a beam or warp roll, as is also the ground warp of the carpet.

[Illustration: Fig. 184]

[Illustration: Fig. 185]

Fig. 185 is the plan of drawing-in and weaving a carpet similar to
that shown in section in Fig. 184. The draft is given at A, and the
treading and cording for a hand loom at B. A 10-row jacquard should be
used for a 5-frame carpet, as twice over the 5 threads makes one row
of the cumber board. Five threads are drawn into the harness, one from
each frame, then two ground warp threads are brought through and drawn
on two leaves of heddles. In weaving, one ground thread is raised and a
shot thrown in; then all the pile warp and the same ground thread are
raised for the next shot; the third tread raises the harness and a wire
is inserted to draw up the pile; the fourth tread crosses the ground
warp threads; the fifth tread raises this thread again, as well as all
the pile warp, and the sixth tread raises the harness for the insertion
of the second wire; this completes, the order of treading. For the
second and fifth treads it will be observed that all the warp in the
harness must be raised; this was usually done by having shafts through
the harness or by raising the cumber board to act as a journal in the
manner explained for Scotch carpet weaving; but now jacquards are made
with bottom boards or gratings that can be raised or lowered to form
these sheds.

The pile wires are now commonly inserted at the time the adjoining pick
is being thrown in, a double shed being formed, and both the wire and
shuttle passed across at the same time, the former being uppermost; in
power looms the wires are both inserted and withdrawn by machinery;
about twenty-five are inserted before any are drawn out, to prevent
the loops from slipping. Jacquard machines are made specially for the
purpose of forming this double shed, the grating raising the hooks from
the bottom to the centre position, thus raising the worsted or figuring
warp above the shuttle, while the griffe carries the figuring hooks on
to the top, making a second or upper shed for the wire. The old method
was to form the lower shed by raising the cumber board, and the upper
one by the griffe. Jacquards with cords instead of hooks are frequently
used.

Brussels carpets are made 27 in. wide, with nominally 260 loops or rows
of pile in the width; but 256 rows are commonly used instead of 260 for
the better classes of goods, and a further reduction of 30 or 40 loops
is made for lower classes, the reduction going to cheapen the goods
or to the credit of the manufacturer. Although only 256 loops are on
the surface of the cloth, it must be remembered that for a 5-frame
carpet there are 5 times this number of threads, and that 1,280 cords
will be required for the harness, and an equal number of needles in
the jacquard when the pattern is single, or not repeated on the cloth;
but the design only covers 256 spaces on the point paper, each space
or check across it representing 5 threads. In Scotland a reed with 350
splits on 37 in. is used, one row of pile warp with its ground warp
going into each split. A carpet known as Axminster Brussels is made in
Scotland. It is similar in structure to the common Brussels, but has
in addition a woollen weft, which is thrown in pick-and-pick with the
usual hemp weft. The addition of the wool weft adds to the softness and
thickness of the carpet, giving more elasticity or spring to it, and
making it wear better.

Designs for Brussels carpets are sketched out and coloured in the
usual way; the method of repeating the patterns is much the same as
is adopted for wall papers. Usually there is one repeat in the width
of the cloth (27 in.), and the length of the repeat may be shorter
or longer to suit the pattern, expense, &c.; 256 checks, = 3/4ths of
a yard, is a common length, and 1-1/2 yard for bolder effects. The
number of colours must be regulated to suit the frames to be employed
in the weaving, unless the frames can be made to suit the pattern. Say
five frames are to be used, this would suggest five colours; but it
does not follow that only five colours can be used. A clever designer
may work out most patterns so that six or more colours may be used on
five frames, by arranging the design so that two colours may be put to
one or more of the frames, which process is called ‘planting.’ But it
may be that an extra number of colours will not improve the design,
in which case they should not be used. The process of ‘planting’ or
striping the warp--which is on the same principle as ‘chintzing’ by
striping the weft in other makes of goods--will be best understood by
referring to Fig. 186, which is a Brussels carpet pattern on a small
scale for working with five frames. The colours are indicated by
different markings on the squares of the design-paper, as shown by the
index or gamut below the design. On the first line of the gamut there
are two varieties of markings, and as this line indicates reds, there
are two colours of red used on this frame. Following up the design
above these markings, it will be seen that one colour of red is used in
one of the sets or rows of octagonal figures, and the other colour on
the intermediate rows, and that the two colours of red are not required
in the same portions of the design, the gamut showing where each is
required. It is only necessary to arrange the spools on the frame in
the loom and draw the threads through the harness, as shown by the
gamut, to make this fall in correctly.

[Illustration: Fig. 186]

What the designer must guard against is giving the pattern a striped
appearance, which would be the case if the planted colours were
brought to the surface too frequently; but if they are judiciously
brought up at intervals, the effect of a 5-frame can be made almost
as good as that of a 6-frame pattern, and this can be carried still
further if desired; every frame may contain two colours, or one frame
may contain several colours.

Carpet patterns are sometimes painted on small-sized design paper--that
is, the ordinary paper with, say, two designs per inch; but before
working it is more satisfactory to have them put on large-sized paper,
so as to show the exact effect they will have on the cloth, the
design-paper giving the pattern full-cloth size.

When painting patterns that are as large or larger on the cloth
than they are on the design-paper, much more care and skill are
necessary than when the pattern is one that is much larger on the
design-paper than it will be on the cloth, as is usually the case;
and in consequence of the large size of the checks required for
carpet design-paper, it is evident that representations of any small
objects--unless those of a very simple nature--cannot be put on it
without being greatly enlarged. After the sketch is made the forms
should be carefully adapted to suit the paper, so as to give a
satisfactory effect on it, rather than be a rigid adherence to the
sketch. Natural floral forms should not be attempted unless to satisfy
the demands of trade, and when they must be used they should be made
on so large a scale as to be fairly represented in a semi-conventional
manner. Large wild-floral patterns are frequently to be seen on
carpets, but they are rarely, if ever, satisfactory. Persian, Indian,
and other geometric or conventional forms that lend themselves readily
to a harmonious admixture of colours so that the carpet will, when on
the floor, present a rich glow of colour instead of obtrusive masses
of floral misrepresentations, are much to be preferred. Dull colouring
is not essential to good taste--rather the reverse; but it is less
obtrusive and more easily managed, particularly in the hands of an
inferior artist. No matter how brilliant the colours are, if properly
blended and broken up into masses in proportion to their intensities
in the spectrum scale they need not be in the least offensive; but it
requires skilful hands to do this.

The pattern given in our last figure (186) is a simple repeating
pattern, so that two pieces of the carpet laid side by side will join
correctly together. Another method of arranging patterns so that they
will repeat is shown in Fig. 187. In this figure it will be seen that
the two sides will not join as an ordinary repeat, but if two pieces
are taken the right-hand side of one will, if turned round, join
against the left-hand side of the other, and two pieces thus joined
will form a repeat. This is objectionable in the case of cut pile,
which usually has a slope in one direction; and when two pieces are
joined, one running one way and the other the reverse, the effect will
be that of shaded stripes, somewhat similar to that of a grass field
when rolled. Another method of forming repeating patterns is by making
what is known as a ‘drop repeat,’ which is shown in Fig. 188; though
the two sides will not join in against each other, as in an ordinary
repeat, if two pieces of the carpet be laid side by side, and one of
them drawn a little down, the point A in the figure of one piece will
fall in against B and B^1 of the other piece, and repeat.

[Illustration: Fig. 187]

These show the usual principles on which repeats are based, and of
course it is for the ingenuity of the designer to form the best
arrangements of pattern he can to suit them, or any other method he may
think of.

When cutting cards for carpet patterns the usual method of picking
out the various colours for each cutting must be adopted by the
card-cutter. Refer to Fig. 186 for the pattern, and to Fig. 185 for the
draft. The order of drawing is green, blue, black, yellow, reds, which
gives five rows of holes on the card, and this twice over makes the ten
rows; therefore, green would be cut on the first and sixth rows; blue
on the second and seventh; black on the third and eighth; yellow on the
fourth and ninth; and reds on the fifth and tenth rows. These may be
all cut when passing the card once along the cutting machine, giving
more than one tramp, if necessary, to each shift of the card index; or
the cutter, if not accustomed to this, may pass the card along once for
each colour, working two of the punches at each passage of the card,
and of course remembering to select the punches which correspond to the
colour being read.

[Illustration: Fig. 188]

It must be remembered that two checks on the design-paper correspond
with a row of ten holes on the cards, as each check may have any of
five colours on it in the case of a 5-frame carpet pattern, and there
must be a space for a hole in the card for any of these colours;
therefore the designs or large squares on the design-paper would
consist of two checks each. Suppose one of these checks--the first
one--to be green, the first hole in the card would be cut for it; if
the second check is also green, the sixth hole on the card would be
cut; but if it is black, the eighth hole in the card would be cut, not
the third hole, as it, being the second check, comes on the second half
of the card. The first check stands against the first five rows of
holes in the card, and the second check against the second five rows,
each check on the design-paper representing five threads, five mails
and five needles, or five spaces on the card; but only one hole is cut
on the card for it, and that must suit the colour the check is painted.

[Illustration: Fig. 189]

_Wilton Carpets._--These carpets are made much in the same manner as
Brussels, but the pile is cut, and they are of a better quality of wool
with a longer pile, and altogether a superior article. As the pile is
cut, the method of binding it into the ground is somewhat different
from that adopted for Brussels, in order to secure a firmer binding. A
section of the cloth through the weft is given in Fig. 189, showing two
pile loops cut and one uncut, and also the method of binding the pile
warp threads Nos. 1, 2, 3, 4, and 5, with the ground warp A and B, and
the weft is shown in section, or end view.

[Illustration: Fig. 190]

The warp or pile threads are cut similarly to velvet by drawing the
wires out of the loops and having each wire furnished at its end with a
cutting edge. Fig. 190 shows the draft and treading of a Wilton carpet
arranged for a hand loom. The back is the harness, and Nos. 1, 2, and
3 the shafts. No 2 shaft is only required when an extra or silent warp
is used for the purpose of giving weight and thickening the fabric.
This warp, if used, lies straight between the weft, and must be on a
separate beam from the ground warp.

_Tapestry Carpets._--These carpets, though they do not require a
jacquard to weave them, are so closely allied to Brussels, and patterns
so similar can be produced on them, that it is thought desirable to
give a short description of them. Though figured weaving, they are
not so in the strict sense of the term, as the pattern is printed on
the warp instead of being formed by various coloured warps; they are
simply an imitation of figured weaving in colours. Any pattern that can
be put on a Brussels carpet can also be put on a tapestry, and with
further variation if desired. The effect is not so sharp and rich on
the tapestry, as the colours, being printed on the warp, run into each
other more or less, and produce a slight blurring round the edges of
the pattern; whereas the pattern on the Brussels is so sharp and clear
round the edges as to make it harsh if two very contrasting colours
come together, so that in some cases the tapestry patterns are softer
and more pleasing.

[Illustration: Fig. 191]

Fig. 191 shows the texture of a tapestry carpet. A and B are the ground
warp threads which bind the weft, the same as in Brussels; these
threads go on one warp roll. 1, 2, and 3 are the lying or filling warp
threads of linen, hemp, or jute, used for giving weight and thickness
to the cloth; these threads go on a second roll. P is the pile, which,
it will be observed, is of various colours and looped all along the
cloth, or passes over every wire that is inserted, and the pattern is
formed entirely by the colours that these threads are printed. The pile
goes on a third warp roll. All the pile can go on the one roll, as
every thread is looped alike by each wire inserted. A pair of shafts
for the ground warp--one for the lying warp and one for the pile--is
all that is required for weaving the cloth. The pile consists of two or
three ends of two-fold worsted yarn, according to the quality of the
carpet. As the fabric is made up to a great extent with the three lying
ends of stiffened hemp or flax to each row of pile loops, the carpet
is of a harder and less pliable nature, with less spring in it than a
Brussels (which has the five pile worsted threads in it), and it is
also less durable. The number of pile loops in the width of the cloth
is the same as for a low-class Brussels carpet, say 216 to 225, and the
width of the cloth is the same--about 27 in. Five threads go to each
split of the reed--viz. one pile, two ground warp, and the three lying
or filling ends.

In these carpets, as well as in Brussels, it will be seen from the
sections that one weft thread is above a lower one, with a straight
tightly-stretched warp between them. A method of weaving has been
adopted on the Continent in which both sheds are formed and both
wefts picked at the same time. One shed is raised above the centre
warp, and one is sunk below it, and both shuttles are driven through
simultaneously; then a pile shed is raised and a wire inserted; and
this repeated, reversing the ground sheds, makes the order of weaving
for tapestry carpets. Another method is to have double mails, one
mail above the other on the same harness cord or heald; the wool warp
goes to the upper mail, and the lying warp to the lower one. When the
mails are raised a double shed is formed, and a pile wire and a pick
can be inserted at the same time. The ground or binding warp is in
two additional leaves of healds. To apply the former method to Wilton
carpets, two picks are put in below to balance with the two above,
instead of one pick below, as shown in Fig. 189.

For tapestry carpets the pattern is prepared in the same way as for
Brussels. The repeat of the pattern may be of any length, but 27, 36,
or 54 in. are usual lengths. The pattern, though prepared as if for the
jacquard, is only a guide for the printer when printing the colours on
the warp.

The pattern is printed on the warp in an elongated form, the amount of
elongation coinciding exactly with the reduction the pile loops make in
the length of the warp when it is woven.

The pattern was originally printed on the woven cloth by means of
blocks; next it was printed on the warp in an elongated form in the
same manner; now it is printed by means of small rollers. The yarn for
each thread of the warp, or of a number of warps, is wound separately
on a large reel or drum, of such a size as will hold a suitable
quantity of yarn (about 18-3/4 feet in circumference). Six or eight
threads are wound on together, so as to lie closely together side by
side and not overlap each other, till the drum is full, or the required
length is wound on it. This has now to be printed with the pattern in
an elongated form.

The right-hand edge of the drum is divided into as many divisions as a
length of yarn equal to the circumference of the drum will make loops
on the cloth. This is called the index, and 648 is a common number
of divisions for it to have, or 864 for a larger size. Of course the
number of loops any length of yarn will make depends upon the size of
them, and the reduction in length from the yarn to the cloth depends
upon the height of the pile and the number of loops in a given space.
The number of loops per inch varies with the quality of the cloth; 7
or 8 is usual for loop pile, and 9 or 10 for cut or velvet pile. The
design must contain such a number of checks in length as will divide
evenly into the index number, such as 108, 162, 216, 324. When painted
the design is cut into strips in the direction of its length, one line
or row of checks in each strip. When the yarn is wound on the drum,
and the printer is ready to begin work, he takes one of these strips
and pins it up before him to guide him in the colours he is to print.
The printing is done by means of a trough of colour with a roller in
it, set on a carriage beneath the drum, so that when passed across
the roller will press firmly on the yarn. The printer finds the first
colour on the design, and setting the drum to the first tooth of the
rack or index, he passes a trough of the proper colour across the drum
and back again; if the second check on the design is the same colour,
he turns the drum round a tooth of the index and passes the same
colour across; if the third check is a different colour, this requires
a second colour trough, and passing another tooth on the index this
colour is passed, and so on with the remainder; or, all one colour is
printed first, then the drum revolved again with the second colour,
and so on. When this is all printed the yarn is taken off the drum and
marked No. 1 thread, and the beginning and end of the thread should
be marked in some way as a guide for the setters. A fresh lot of yarn
is wound on for the second thread, which is printed according to the
second strip of the design. This must be continued for all the threads
in the warp. After being printed, the colours are fixed by a steaming
process, and afterwards wound on bobbins and marked to their numbers.
The design may be cut into strips of two checks each, instead of one
check, which will be less liable to get torn. One side of the strip can
then be printed first, and the other after.

The warp is made up by ‘setters,’ who arrange the threads together
in proper order in a frame for the purpose, and set them so that the
colours of each thread come together at the proper place to form on the
warp a correct elongated copy of the pattern; when correct it is wound
on the beam for the loom.

In printing and steaming, the colours are liable to run into each
other; some colours are worse in this respect than others, but it
is reduced to a minimum by the use of an absorbent in steaming; and
sometimes the designer makes a little allowance on the design-paper for
such colours as he knows will be liable to run and injure the pattern.
The pile of tapestry carpets is usually left uncut; but sometimes it is
cut, forming a velvet pile. In this case the carpet is made of a better
quality, and the pile is longer.

The present method of preparing tapestry warps was invented by Richard
Whytock, of Edinburgh, in 1832, and perfected by Messrs. Crossley, of
Halifax.

_Axminster Carpets._--Real Axminster carpets, as already said, are
made by hand much in the same manner as Turkish or Persian carpets,
and attempts have also been made to produce them by machinery, several
patents having been taken out for the protection of the inventions.
These carpets, with other art carpets now being made, may be classed
with tapestry as works of art.

The Axminster carpets of commerce are the Royal Axminster or Moquette
carpets, and the patent Axminster or chenille carpets.

Royal Axminster carpets are made by a peculiar process of weaving and
tufting on a loom made specially for the purpose. A number of little
funnels carrying from spools threads of the different colours of yarn
required for the pattern are fixed on the loom above the reed, and
these threads, by suitable mechanism, are brought down and bound into
the backing cloth by the weft. These threads, which form the pile, are
cut off, and afterwards the surface of the carpet is shorn level.

[Illustration: Fig. 192]

The texture of Moquette carpets is shown in section in Fig. 192, two
varieties of texture being given, one at A, the other at B. In both
these textures the pattern shows through on the back of the cloth.

Another variety is given in Fig. 193. A is the section through the
weft, B the texture, C D E F shows the interlacing of each of the
four warp threads in the pattern with the weft, and G shows the pile.
Similar numbers in these figures represent the same threads in the
cloth.

[Illustration: Fig. 193]

_Patent Axminster Carpets._--These were invented by Mr. Jas. Templeton,
of Glasgow, about 1839, and are a description of chenille weaving. They
do not require to be wrought by a jacquard any more than the other
Axminster carpets, but as the figures formed on them are so similar to
those produced by the jacquard, and as the method of designing the
patterns is the same for both them and other classes of chenille as it
is for jacquard work, it may not be considered out of place to give a
full description of chenille here.

_Chenille_ is made by a double process of weaving. First the weft or
chenille thread is woven, if for a figured pattern in various colours,
which corresponds to the printing of the warp in tapestry carpet
weaving, which gives a warp figure, whereas chenille gives a weft
figure. When the weft is woven in a piece it is cut up into strips
and twisted, if for curtains or any fabric on which the chenille weft
is to show on both sides; but for carpets, where all the pile is
raised to one side, the weft is doubled up, bringing all the pile in
one direction. Twisting machines are now in use for making the weft
for plain chenille fabrics, but for figured ones it is still woven.
If twisted for figured work it would afterwards have to be printed
somewhat similar to tapestry carpet warps, but without elongating the
pattern, and the uneven surface is a difficulty in the way.

The method of weaving chenille weft is as follows:--A loom fixed for
working gauze is supplied with a thinly laid warp, which, according to
the class of chenille required, is drawn through the heddles either as
plain gauze, two threads working plain with one crossing them, or this
doubled, as is shown in Fig. 194. The two, three, or six threads of the
draft are drawn into one split of the reed, which is finer or coarser
to suit the fabric required. For shawls or curtains there would be six
to eight full splits of the reed per inch; say in a reed of thirty to
forty splits per inch, four splits empty to one full. For carpets there
would be two or three full splits per inch; say in a reed of ten to
fifteen splits per inch, four empty and one full, coarser or finer to
suit the length of pile required.

For plain work this is tied up and woven as gauze, the warp being
cotton of a dark colour, and the weft woollen, noil silk, or other
fibre if desired; but it should consist of several ends and be of such
a nature as will divide easily to form a rich pile. If for figured
work, the pattern must be woven in stripes across the piece, which will
be explained after the designing.

[Illustration: Fig. 194]

When the piece is woven it will be as shown at A, B, or C (Fig. 194)
and must next be cut into strips midway between the ribs. Sometimes
this is done in the loom as the cloth is being woven, by having cutting
knives arranged to divide it as it passes over the breast beam; but it
is usually taken off in the piece and cut afterwards in a machine for
the purpose. The piece is fed through rollers in the cutting machine
over a grooved roller, shown at B (Fig. 195), the grooves being at such
a distance apart that the ribs of the piece will fall into them. Above
this roller is a set of cutting blades fixed in a spiral form round a
roller; one of these blades is shown at A (Fig. 195), with a section
of the roller. The blades are set at such a distance apart that when
revolving they fall into the cuts in the roller B, and running at a
high speed they sever the piece into strips as it passes through the
machine.

So far the process is alike for all classes of chenille work. The weft
is now in strips similar to that shown at A (Fig. 196) if for carpets,
and much smaller if for curtains or shawls. For ordinary chenille the
weft is next twisted so as to make the projecting fibres stand out
round the rib or core, and when finished it has the appearance shown
at B (Fig. 196). If for carpets, the fibres are turned up so that all
will project in one direction, as shown at C and in the cross-section
at _d_. This is done by running the strips over a grooved roller, as
C (Fig. 195), which is heated with steam, and as the projecting ends
are doubled up by the grooves in the roller the heat sets them in this
position. The weft is now prepared, and is wound on bobbins and woven
in large shuttles.

The texture for chenille is the same as for plain cloth. A fine black
warp of cotton is used, with twenty to twenty-five threads per inch
(more or less, to suit the fabric required), and twelve to sixteen
shots per inch of the chenille weft are used, which must also be varied
to suit circumstances and the thickness of the weft used. The fibres
or points of the thread of weft project through the warp, and a pile
fabric is produced which should entirely conceal the warp on both sides
of the cloth. For carpets the pile is, as a rule, only allowed to
project on one side, though some rugs are made with a twisted chenille
weft, and the colours on it are shown partly through the backing as
the fibres of the thread get mixed up with the ground or backing in
weaving, but usually all the pile projects through to the face.

[Illustration: Fig. 195]

[Illustration: Fig. 196]

[Illustration: Fig. 197]

The texture for patent Axminster carpets is shown at A (Fig. 197),
and sections of the cloth through the weft are shown at B and C in the
same figure, B is a section of the texture A with two picks between the
pile or chenille picks, and C is for a coarser description of work with
four ground picks between the pile picks. In A only the ground texture
is given complete; the pile weft P lies over the picks _f f_, and is
bound down by a fine black stitching thread C, which passes over it and
under the picks _g g_. The different thicknesses of the threads are
represented in the drawing: C are cotton, and the others hemp, flax, or
jute. There are about ten thick and ten fine warp threads per inch, and
five binding threads, and about seven chenille picks per inch, with two
ground picks between these. When there are four ground picks between
the pile picks the pile is coarse, and only about four picks per inch
are used. The chenille weft is wound on large wooden needles like huge
netting needles, about 4 ft. long, so that it may come off without any
twist in it, and the weaver pushes these through the shed, laying the
weft nicely in and combing it forward so as to get it straight and even
and have all the pile standing upright; he then knocks in two ground
picks and puts in another pile thread, as before.

This is all hand-loom work, but power looms are sometimes used for the
purpose. The ground of the fabric can be woven with heddles, as before,
but the binding warp threads are through needles, somewhat like gauze
dents inverted, and not through the heddles; and the beam or spools
containing them is above the loom in front. The chenille weft is wound
on a reel, and is through a guide or carrier. When the chenille weft
is to be laid in, the needles carrying the binders are raised and the
guide passed along, laying in the weft. The loom stands stationary
for a short time to give the weaver time to comb up the pile, and
then moves on and throws in the ground picks. Everything is done
automatically but the combing up of the pile.

The foregoing is a description of chenille weaving provided no pattern
has to be attended to; we must now consider the pattern.

The design paper used is the same as for ordinary work, with a greater
number of warp than weft threads, but it is ruled on a large scale so
that the pattern will be exactly cloth size. Fig. 198 is a sample of
this design-paper for seven picks per inch. The small checks, or what
in ordinary weaving would represent the warp threads, have here no
reference to them, nor do the warp threads require any consideration
when preparing the pattern. The narrow way of the checks is a guide to
the weaving of the weft threads. This paper might have been square,
seven by seven, and would thus suit for such a pattern as is shown
in Fig. 199--supposing seven shots of chenille weft per inch to be
correct; but there is an advantage in many cases to have it as it is,
or even more off the square, say seven by twenty, which is one of the
papers used. This will be most readily understood by following the
working out of the pattern.

The pattern given (Fig. 199) is necessarily very simple for want of
space. The different markings on the squares are to indicate different
colours. Here nine are used, but there may be any number--say from
eight to twenty--the only restriction on the number of colours employed
is that so many shades of yarn are required to match them, and that the
weaver has a greater number of shuttles to pass over when weaving the
weft. When the pattern is painted it is cut cross-ways into strips, as
A, B, two rows of checks in each strip; one row of checks may be in
each strip if desired; but this is unnecessary, and besides, they are
more easily torn. These strips are a guide to the weaving of the weft,
just in the same way as the pattern in tapestry carpet work is a guide
to the printer.

[Illustration: Fig. 198]

[Illustration: Fig. 199]

The weaver takes a strip of the pattern and puts it through a widened
split in the reed and pins one end of it to the cloth; to the other end
he attaches a cord and light weight, which he hangs over the back rail
of the loom so that the strip of design will be held steady and move
forward as the cloth is drawn up. The strip must be pinned to the cloth
so that the first check of the design projects over the fell of the
cloth. Suppose the strip to be A, Fig. 199: C is the first check; the
weaver sees that this is black, and proceeds to weave the gauze with
black weft for the length of this check (one-seventh of an inch). The
next check, moving along the strip in the loom, or from left to right
on the design, is a different colour, say scarlet; the weaver changes
his shuttle and weaves as many checks as there are of this colour,
which in this case is only one. Another change of shuttle is made,
and so on till the end of the strip is reached; it is then turned, the
other end of it being brought to the fell of the cloth, and the checks
on the second line of the design are woven in order as before, only
they must run from right to left on the design instead of from left
to right, as the first line did, which will be caused by the turning
of the strip of paper; for it must be borne in mind that if the first
shot of chenille weft is put into the carpet from left to right, the
next one will be put in from the reverse side or from right to left, so
that all the odd numbers of lines on the design-paper should be begun
by the weft weaver at the left-hand side, and all the even numbers at
the right-hand side, provided the weaver works in this way; in any
case one must be the reverse of the other. When all the strips of the
pattern are woven, each rib of the piece of weft will be long enough to
weave as many shots as are in one repeat of the pattern, and if there
be 40 ribs in the piece, it would, when cut up, furnish weft for 40
repeats of the pattern. For weaving this weft large needles are used,
somewhat similar to netting needles, which are about 4 ft. long for
wide looms; the chenille is wound on these, free from twist, and must
be put on so that it will come off correctly to suit the pattern. The
strips of chenille are marked to show which end is to be put on the
needle first. These needles are pushed through the shed by the weaver
when weaving the carpet--put in at one side and drawn out at the other.
He lays in the weft as correctly as possible, combing it up to the fell
of the cloth, and makes the colours fall in nicely together, drawing
it tighter or leaving it slacker as is necessary. Now the advantage of
using the special design-paper will be more readily understood. It is
plain that if 14 by 7 design-paper be used, instead of 7 by 7, any of
the colours may be changed in 1/14 in. on the design; and also in the
chenille weaving, when it is desirable to do so, instead of in each
1/7 in. when 7 by 7 paper is used. And if 20 by 7 paper were used, the
changes might be still more frequent; besides, in flowing patterns the
advantage also comes in to assist in getting the forms more correct,
in the same way as paper over square in either warp or weft assists
in getting the form of the figure in damask designing over that where
only the coarser way of the paper square is used; though the effect is
not by any means so satisfactory as if the paper were finer both ways,
and square. Yet when one way is fixed by circumstances it is in many
cases preferable to alter it in the other way than to leave both alike
coarse; but not always so, as, for example, the pattern given in Fig.
199. When these carpets are woven they are passed through a shearing
machine, which makes the surface of the pile quite level. They are rich
handsome carpets, as well as very durable.




CHAPTER XI

_LAPPETS AND SWIVELS_


_Lappet and Swivel Weaving._--The only classes of figured weaving of
any importance not already mentioned are lappet and swivel work, both
of which are very limited, but produce good effects. With lappets,
figures can be put on light fabrics without requiring any cutting off
of the surplus yarn. It is principally in the muslin trade that they
are used, for figuring Indian fabrics with gay colours. The principal
seat of the manufacture in this country is about Glasgow. The figure
is formed by an extra warp drawn through one or more frames of gauzing
needles set in the lay in front of the reed, which can be moved to
either side as desired. The pattern is cut on a large wooden wheel
or disc, in the face of which a groove is cut to form the pattern,
and a traveller working in this groove, and connected with the needle
frame, slides the needles to either side and stitches the figuring warp
into the cloth, the needles rising and falling as the cloth is being
woven. The figure is rather loose and only suited for simple effects.
A description of the process of working can be found in Murphy’s or
Gilroy’s works on weaving, and is interesting, though rather out of
date. Swivel work is of a much better class, though not so suitable for
getting small figures, as spots short distances apart. But it gives a
much better bound figure--in fact, it may be bound in any way--and is
very suitable when a few are required on the fabric, such as crests, or
any figures in the corners or centres of handkerchiefs, napkins, etc.
These are weft figures and usually of a different colour to the ground
of the fabric, which may be either plain or figured. These figures
could be formed with an ordinary box loom, cutting off the loose weft
or binding it up to the ground cloth; but the latter is useless for
light fabrics, and the former leaves the figure unbound round the
edges. If the figure is large enough it can be made in an ordinary
check or pick-and-pick loom, and bound as with swivels. The swivels are
small shuttles similar to those used for weaving silk book-markers and
such like fabrics, and are usually attached to the hand rail of the
lay of the loom. The swivel attachment is fixed so that when a shed is
opened the shuttle falls to pass through it, or a second high shed may
be opened for the swivel shuttle. The shuttle is driven by means of a
rack and pinion, the former being in the shuttle, and the pinions fixed
to revolve in the frame. When the pinions are revolved the shuttle is
drawn across. The breadth of the figures must be considerably less than
the length of the shuttle, so that for large-sized figures this method
of working is not suited.


CONCLUSION

Since the foregoing was written up to the present time (Sept. 1894)
nothing new worth describing has made its appearance. A few card lacing
machines have been brought out and several patents have appeared in
print, but most of these disappear shortly after.

No mention has been made in this work of Count Sparres’ patent
card-cutting machine, which was to have done away with the necessity
of putting patterns on point paper by cutting the cards direct from
the sketch embossed on a metallic plate. This was a highly ingenious
machine, and under suitable conditions produced very fair work, but so
far it has proved rather expensive to the company that took it up.

Another process to obtain similar results, patented by Messrs. Dawson
and Adams of Macclesfield, was described in the ‘Textile Manufacturer’
in 1893. In this process a perforated plate is filled with small
pins; a piece of strong paper is laid over this, and over the paper
another plate. The pattern is formed by pushing the pins through the
paper between the plates. A large portion of this work can be done
by mechanical means. When all the pins required for the pattern are
pressed through the paper, the top plate with the sheet of paper is
turned upside down, the pins sticking in the paper. By running a
roller over the paper the pins can be pressed out of it and into the
plate. This plate is then put into a reading and punching machine,
and by suitable mechanism the pins can be brought to act on either a
card-punching apparatus or on the harness of a pattern loom; so the
manufacturer can have a sample of the cloth woven without any cards and
can have the cards cut for the factory loom afterwards if he requires
them. The cards can be punched at the rate of 2000 per hour.

By means of a pattern cut out of a plate of wax, and a reading in
machine, the filling of the pins into the plate, or the hand reading as
it is called, can be dispensed with.

It has since been stated that this firm are now trying to perfect a
process which was tried thirty years ago, viz. to paint the pattern
with electric paint which is to act on needles charged with electricity.

These are all highly ingenious inventions, and are interesting to those
who do not lose too much by them; but it will require a nice machine to
produce all the variations in a pattern that an experienced designer
can, though in many patterns this could be dispensed with, and suitable
mechanical means may yet be devised to take a share of the work.




INDEX


    Axminster carpets, 266, 288

    Barrel or cylinder loom, 17
    Beeting a harness, 89
    Bessrook machine, 185
    Border design, 97, 100, 115, 134
    -- mountings, 105
    Bouchon’s loom, 18
    Brussels carpets, 27

    Cards, 51, 147
    Card cutting, 139
    -- -- machines, 143
    -- -- -- (Count Sparres’), 299
    -- -- -- (French), 167
    -- frames, 47, 156
    -- lacing, 139
    -- -- machines, 148
    -- repeating machines, 149
    Carpets, 264
    -- chenille, 290
    -- Axminster, 266, 288
    -- Brussels, 276
    -- Indian, 265
    -- Ingrain, 269
    -- Kidderminster, 266
    -- patent Axminster, 289
    -- Persian, 265
    -- royal Axminster, 288
    -- Scotch, 266
    -- tapestry, 285
    -- Turkish, 265
    -- Wilton, 284
    Centred tie, 97
    Chenille, 290
    Comb draw loom, 16
    Comber boards, 82, 103
    Cross-border jacquard, 155
    Cross’s counterpoise harness, 14
    Cumber boards, 82, 103
    Curtains and tapestries, 248
    Cylinder motions (horizontal), 44
    -- (self-acting), 36
    -- (slide), 43, 57, 61
    -- (swing), 33, 56, 58
    -- to set a, 40, 43

    Designing and draughting, 110
    -- hints on, 120
    Design paper, 112
    -- to prepare a, 115
    Designs, 94, 97, 100
    -- for dress goods, 122
    -- for gathered borders, 97
    -- for single borders, 100
    -- (repeating), 94, 122
    -- (texture), 124
    Double cloth, 233
    -- -- jacquard, 241
    -- warp faced cloth, 237
    -- weft faced cloth, 235
    Draughting a pattern, 115
    Draw loom, 3
    -- -- French, 24
    Drawing in a warp, 99

    Falcon’s loom, 19
    French card, 51
    -- -- cutting machine, 167
    -- draw loom, 24
    -- jacquard, 49
    Full harness, 69
    -- -- mountings, 69

    Gathered tie, 97
    Gauze, 196
    -- to design a pattern, 215
    -- harness (hand-loom), 203
    -- -- (power-loom), 205
    Gauze, jacquard, 214, 223
    -- patterns, 197, 198, 217
    -- reed, 225
    Griffes, 37, 52

    Hair-line patterns, 129
    Half-harness, 69, 169
    Harness, drawing warp into, 99
    -- illustrations of, 96, 98, 102, 191
    -- to prepare a, 71
    -- to tie up a, 87
    -- (double cloth), 233
    -- (full), 69
    -- (gauze), 203, 205
    -- (half), 69, 169
    -- (muslin), 169
    -- (pressure), 70, 174
    -- (quilting), 233
    -- reeds, 82
    -- (split), 62, 172
    -- twine, 75
    Hecks, 77
    Hole board, 83
    Hooks for jacquard, 38, 39, 52, 59
    -- in double-acting jacquard, 55, 59

    Indian carpets, 265
    Ingrain carpets, 269
    Introduction, 1

    Jacquard’s machine, 23
    Jacquard machines, 27
    -- (cross-border), 155
    -- (double-acting), 52
    -- (double cloth), 241
    -- (double-acting with two cylinders), 58
    -- (French), 49
    -- (gauze), 214
    -- (Karl Wein), 192
    -- (leno), 214
    -- (open-shed), 161
    -- (quilting), 241
    -- (twilling), 185
    -- (Verdôl), 163
    -- cards, 50
    -- cylinder, to set a, 40, 43
    -- hooks, 38, 39, 52, 59
    -- interior of, 38
    -- needles, 38, 39, 52, 59
    -- shedding, 60, 62
    -- stop motion, 64
    -- to set a, 43, 46, 76

    Karl Wein jacquard, 192
    Kidderminster carpet, 266

    Lacing cards, 139
    -- machine, 148
    -- twine, 147
    Lappet weaving, 298
    Lashing a pattern, 9
    Leno jacquard, 214
    Letters, mottoes, &c., 136
    Levelling frames, 85
    -- the mails, 86
    Lingoes, 72, 74
    London tie, 96

    Marking off a cumber board, 103
    Marseilles quilt, 246
    Matelasses, 246
    Mottoes and letters, 136
    Mountings, 69
    Muslin harness, 169

    Neck of harness, 79
    Needles, 38
    Norwich tie, 95

    Open-shed jacquard, 161

    Paints for patterns, 116
    Parrot machine, 11
    Patent Axminster carpet, 289
    Pattern, to draught a, 115
    Patterns (carpet), 280, 282-3
    -- (gauze), 197, 217
    -- (hair-line), 129
    -- (spot), 130
    -- (texture), 124
    Persian carpets, 265
    Piano machine, 145
    Pile work, 259
    Point paper, 112
    Poplins, 252
    Pressure harness, 174

    Quilting, 247
    Quilting harness, 233

    Reading a pattern, 9
    Repeat of a pattern, 94
    Repeating machines, 149
    Reversing motions, 45
    Royal Axminster carpets, 28

    Scotch carpets, 266
    Set of harness (to vary), 106
    Setting a jacquard, 43, 46, 76
    Shading floral forms, 118
    Shedding of jacquards, 62
    Simple (draw loom), 3
    Slabstock, 85
    Sleepers, 73
    Slips (comber board), 83
    Split harness, 172
    Spot patterns, 131
    Spring box, 42
    Stop motion, 64
    Suggestions on design, 120
    Swivel weaving, 298

    Tail cords, 54
    Tapestry (real), 255
    -- carpets, 285
    -- curtains, 248
    Terry pile, 260
    Texture patterns, 124
    Tie of a harness, 93
    Trap board, 15
    Turkey carpets, 265
    Turkish towels, 260
    Twilling designs, 117, 133
    -- jacquards, 185
    Twills for damask, 181
    Tying up a harness, 87

    Varnishing a harness, 92
    Varying the set of a harness, 106
    Vaucanson’s loom, 22
    Verdôl jacquard, 49, 163

    Wilton carpets, 284
    Wiring cards, 153
    Woollens and worsteds, 248


                              PRINTED BY
                SPOTTISWOODE AND CO., NEW-STREET SQUARE
                                LONDON


Transcriber's notes:

In the text version, italics are represented by _underscores_, and bold
and black letter text by =equals= symbols. Superscripts are represented
by ^{} and subscripts by _{}.

Missing or incorrect punctuation has been repaired. Inconsistent
spelling and hyphenation have been left as printed.

The following mistakes have been noted:

p. 15. boars has been changed to bars. Misaligned typeface has half
duplicated the first letter of each line on this page.

p. 26. Both trapboard and trap board have been used in the text and
have been left as printed.

p. 35. cyiinder has been changed to cylinder.

p. 65. 2 4 6 8 13 12 changed to 2 4 6 8 10 12.

p. 134. coarses has been changed to coarser.

p. 132. fig. 3 has been changed to Fig. 3 as used in all other cases.

p. 141. cyclinder has been changed to cylinder

p. 208. tightly paced should probably read spaced or placed, but has
been left as printed.





End of Project Gutenberg's Jacquard Weaving and Designing, by T. F. Bell