PICK-AND-PICK LOOM.

The majority of box looms are made with movable boxes at one side of the loom only, so that single picks of any colour cannot be put in the cloth at will. As it is very desirable in many fabrics to use single picks of a colour or count of weft, it is necessary to have movable boxes at both sides of the loom, and where this is the case it is usual to have picking mechanism which will allow of several picks being made in succession from either side of the loom. If the matter be carefully thought over, it will be easily apparent that even with drop boxes worked quite independently of each other at both sides of the loom, if the picking mechanism is of the ordinary kind—viz. to pick once from each side alternately—it would be impossible to obtain that variety of changes in the shuttles which is in many cases necessary. In a loom with two boxes at each side worked independently, it would be impossible to obtain single picks alternately of two colours or counts. But by being able to pick twice in succession from each side this can be done. By going through all the changes possible with a given number of boxes, the advantage of this kind of picking arrangement will be very apparent, in the command it gives over any shuttle in the series for any pick. It is therefore necessary to have the picking mechanism aforementioned in order to allow of all the boxes being emptied at one side if required. A loom of this character is called a “pick-and-pick” loom; the picking motion is sometimes called a “pick at will” motion. The loom which we take as an example is one on the Diggle’s chain principle. There are two chains, placed at one side of the loom for convenience. Both chains are on one barrel, A ([Fig. 82]). The chain for working the boxes at the right-hand side of the loom operates the lever B, and the left-hand chain operates the lever C, the fulcrum of both levers being at D. When these levers are lifted they lift the levers E and F, and when E is lifted it lifts the boxes at the right-hand of the loom, and when F is lifted it lifts the left-hand boxes. The connection of the left-hand boxes with the lever F is shown at [Fig. 83]. The shaft G is placed under the loom, and the left-hand boxes are connected to the lever H, which is fast to the shaft G. The lever F is also fast to the shaft G, but the lever E rides loosely upon this shaft, which merely serves as a fulcrum for E. From these two figures it will be clearly understood how the boxes may be worked independently at both sides of the loom by two chains placed side by side upon the barrel A.

FIG. 82.

FIG. 83.

FIG. 84.

FIG. 85.

In order to make each link in these chains represent any number of picks, and thus prevent long cumbersome chains, the mechanism shown at Figs. [84] to [87] is employed. The barrel A in [Fig. 84] is the same as barrel A in [Fig. 82], and carries the chains for lifting the levers B and C. At the end of the barrel the star wheel I is fixed, and this star wheel is turned by the pins J. These pins are worked by a clutch motion shown at [Fig. 86], by which they can be withdrawn from gear with the star wheel as desired. The pins KK are fixed, and turn one tooth of the star wheel Y every pick, the wheel M having twice the number of teeth contained in L, which is on the crank shaft of the loom. The star wheel Y is fast to the end of a small octagonal barrel, which carries a pattern chain N composed of small metal cards, and we have seen how this barrel is turned one division every pick. Above this pattern chain N a finger, O ([Fig. 86]), is placed, and is lifted up against a spring every pick by the cam P on the face of the wheel M. When the finger is up, the pins JJ are taken inside the wheel M, as shown at [Fig. 86]. The cam P only raises the finger a sufficient length of time to allow the barrel Y to be turned round, and if there is a blank in the cards opposite the finger O when it is let down by the cam it will still keep the pins J inside the wheel, and will thus prevent either of them from engaging with the star wheel I, and will leave the boxes unchanged. This can be repeated any number of picks. If a change is required in the boxes, a hole is placed opposite the finger O, and when it is let down the pins J project through the wheel M (as indicated by the dotted lines in [Fig. 86]), and the star wheel I will thus be turned one tooth, and the chain can make the change required in the boxes.

FIG. 86.

FIG. 87.

[Fig. 87] is another view of the cam-shaped projection P, which raises the finger every pick, and [Fig. 85] is another view of the chain barrel A. The letters in the six Figs. [82] to [87] inclusive refer to the same parts in each case.

In this way the chains on A are rarely required to be very long, as one link may be made to represent any number of picks from one upwards. Of course a separate card on Y is required for each pick, but these are very small, only about 1½ inch in length, and a large pattern can be made with very little trouble.

When a Jacquard is used on one of these looms it is sometimes necessary to work the pattern from the Jacquard cards. This can be done in a very simple manner by covering the hole in the barrel carrying the cards N with a metal plate, which is held over the hole by a spring. When a change is required in the boxes, a Jacquard hook pulls the plate from over the hole, and allows the finger O to drop, and thus causes the star wheel I to be engaged by the pins J.

The picking mechanism in a pick-and-pick loom may be either over or under pick. In the former the picking tappets are sometimes moved on the shaft by a clutch arrangement. In the latter the top of the picking treadle is movable. As the under pick is perhaps the best adapted for this loom, we will describe it.

FIG. 88.

[Fig. 88] is a side view of the loom, and the top of the picking treadles consists of a metal plate with the “shoe” S of such a shape as to give the required force and character to the pick. This metal plate works round a pivot, P. The treadles at both sides of the loom are the same in this respect. At the back of the loom a rod, R, is connected to the extreme ends of the loose plates or the treadles, and when one plate is on the treadle, the other is fixed off its treadle, as shown in [Fig. 89]. The consequence is that when the picking bowls come round (there are two bowls on the bottom shaft at each side of the loom) the loom will pick always from that side where the loose plate is on the treadle, and at the other side, where the plate is off, the bowls will pass over the treadle without touching it. By moving the rod R sideways, the plates may be moved alternately off and on their treadles.

FIG. 89.

FIG. 90.

If the loom has four boxes at each side, it may be necessary to pick four times in succession from one side of the loom, and by a simple arrangement the picking can be regulated at will. The mechanism for moving the rod R sideways is shown at [Fig. 90]. Inside the loom framework a lever, L, is centred at C, and by a combination of levers is connected to the rod R, which is the rod referred to in the previous diagrams. A strong spring keeps the plates right for picking from one side, but when it is required to pick from the other the lever L is lifted, which moves the rod R sideways and moves the plate off one treadle and on the other. A chain is used for lifting the lever L, and the star wheel A is turned by two pins on the wheel B on the bottom shaft of the loom, or by one pin if on the crank shaft, thus causing the star wheel to be turned one division every pick. The loom may thus be made to pick four times from the right side, three from the left, twice from the right, and so on, of course always taking care that the shuttles are there to be picked.

CHAPTER IV
DOBBIES

THE tappet shedding motion is the simplest and most perfect for a small number of shafts. They may be made to work an indefinite number of shafts, but it is seldom that above eight or ten are worked with ordinary tappets, and about sixteen with Woodcroft’s or other plate tappets.

With dobbies, a higher number of shafts may conveniently be worked, but it is not only for this reason that dobbies are so extensively used. They are extensively used for weaving twills, satins, and other simple weaves, on four or five shafts. The chief advantage they possess is that any number of shafts within their capacity may be used without extra trouble or cost; whereas ordinary tappets have to be made specially for each pattern; whilst section tappets, and oscillating tappets, are inconvenient.

Dobbies are made to weave up to 40 or more shafts, but 16 and 20 are the commonest numbers. Most dobbies now used are on the double-lift principle; indeed, the single-lift dobby or witch machine is almost obsolete in cotton weaving. The chief kind of double-lift dobby is the “Hattersley” or “Keighley” dobby. The principle of this machine was invented by Messrs. Hattersley & Hill, of Keighley, Yorkshire; hence its name. Since the original patent rights have expired, almost all loom makers have their own particular form of this dobby, embodying many more or less minor improvements on the original. The principle of this dobby will be understood from the lecture diagram, [Fig. 91]. The dobby is placed at one side of the loom, and is therefore in a convenient position for being attended to. The upright rod R is connected to a crank on the bottom shaft of the loom, and therefore the rocking lever AB, centred at C, will make one complete movement to and fro, every two picks. The knives D and E slide along, always retaining a horizontal position, one going inward as the other comes outward.

FIG. 91.

The shaft or stave is connected to the jack lever FGH at F, and the upright MN is fastened to this lever at H, the fulcrum being at G. The upright MN has two hooks, P and Q, connected to it at opposite ends, and suppose that when the knife D is in its innermost position, as in the diagram, the hook P is dropped on to the knife; when the knife begins to move it will take the top of the upright MN with it until MN assumes the position indicated by the dotted line M′N, and the stave is lifted. If it is required to lift the same stave for the succeeding pick, the bottom hook is then dropped on to the knife E, which at that moment will be in its innermost position just commencing its outward movement, and is taken forward by it until the upright MN assumes the position indicated by the dotted line MN′; and it will easily be seen that as the top of the upright is moving back from M′ to M whilst the bottom of the upright is moving forward from N to N′, the centre of the upright H remains stationary at H′, with the exception of a slight movement caused by the knife going further back than the hooks, and thus the stave remains up all the time. The character of the shed is, therefore, what is termed “open shed”—that is, if a stave is required up for several picks in succession, when it is lifted it remains up until it is required to come down again. This is what is meant by “open shed” as compared with “centre shed,” the characteristic of which is that the lifted stave, instead of remaining up, is let down halfway every pick and taken up again if required.

The method of dropping the hooks is as follows:—Two levers, S, T, of different shapes are employed for each pair of hooks; these levers are centred on a rod, X. One of the levers, viz. T, is bent from the fulcrum to touch the bottom hook, and the lever S projects straight out from the fulcrum, and an upright needle O rests upon it, the top hook resting upon the upright needle. When the lever SY is lifted a little at Y it will drop the top hook, and when TY is lifted at Y it will drop the bottom hook.

In a 16 shaft dobby the parts shown in the diagram are duplicated sixteen times—that is, there are sixteen uprights MN, each with two hooks, sixteen levers SY, sixteen like TY, and sixteen of other parts. The levers SY and TY are operated by lags pegged so as to lift the staves to give the required pattern. These lags work round a cylinder or barrel, which is turned round the space of one lag every two picks intermittently. Each lag operates the hooks for two picks, one row of pegs operates the top hooks P, and the other row of pegs the bottom hooks Q. The method of pegging the lags will be understood from [Fig. 92], where two lags are shown with the pegging for a two and two twill. Of course care must always be taken that the pegs are put opposite the proper levers, as when only a portion of the jacks are used, say eight, it is often preferred that the staves be connected to eight jacks in the middle of the machine.

FIG. 92.

FIG. 93.

Dobbies constructed with single jacks, as indicated in [Fig. 91], are only suitable for narrow looms. Those constructed with double jacks are preferable for wider looms, as they not only keep the healds under better command, but they also move them in a perfectly vertical plane without the tendency to a slight side movement such as occurs when the healds are controlled by single jacks, in consequence of the ends of these describing an arc of a circle as they rise and fall. One of the best adaptations of double jacks to the Keighley type of dobby is that exemplified in the “Climax” dobby made by Lupton and Place, Burnley. This is represented in [Fig. 93], in which A and B are complementary jack-levers operated from the same baulk-lever, J, controlled by hooks P and Q, to govern the same heald. The distinctive features of this dobby are the construction of the outer jack-lever A in one part, instead of two parts, and its attachment with the inner jack, B, by means of a link, C. This modification is a great improvement on double-jack dobbies in which the connections are made with streamer hooks or rods, or those in which the jacks are geared by means of toothed segments, as these increase the number of parts that are liable to wear and to get out of order.

The Keighley dobby is decidedly the most popular one at the present time, but what is known as the “Blackburn” dobby is preferred by some. This is a double-lift dobby, which gives a centre shed—that is, the staves which are required up for a number of picks in succession are let down halfway every pick and taken up again. The principle of this dobby is illustrated at [Fig. 94]. The staves are lifted by the two jacks A and B; when B is lifted it causes A to lift the same distance. There are two hooks, D and F, for each double jack, and the lags are divided into two parts, all the odd numbered picks being fastened together, and the even picks forming another chain. The pegs in the lags press back the hooks, the back part of each of which forms a spring, so that when the hook is pressed back it leaves the stave down.

FIG. 94.

The knives lift alternately. When one is going up the other is going down, and when one hook of a pair is lifted, as in the diagram, a lag operates the other hook, and if the same stave has to be lifted for the next pick, the hook is left over the knife, and the second hook will be taken up whilst the stave is being let down, and will catch it halfway and take it up to the top again. This is the advantage of all double-lift machines over single-lift. The staves which are required up for a certain pick are being taken up whilst those which were up for the previous are coming down. A saving of time is thus effected, and the looms can be run quicker than with single-lift machines. A crank L on the end of the bottom, or picking, shaft is connected by means of a lifting-rod E to the end of a horizontal arm M, mounted on a shaft G, which constitutes a fulcrum for the arm. On one end of shaft G, at the rear of the machine, there is fixed a toothed quadrant H, which, through the medium of a small wheel, transmits motion to a similar toothed quadrant H′ fixed on the end of a shaft G′; and each quadrant is connected, by means of rods K and K′ to the rear ends of the respective griffe-bars D and F. On the opposite ends of the shafts G and G′, there are fixed, one on each shaft, two short arms which are respectively connected to the fore ends of the griffe-bars. Therefore, as the crank L revolves, the griffes are raised and depressed alternately, and in a contrary manner.

Another thing to be borne in mind is that in a single-lift machine all the staves come to the bottom every pick, and therefore the character of the shed is different from that of a double-lift. In double-lift machines there are the “open-shed” like the Keighley dobby, and the “centre shed” like the Blackburn dobby. It is important to remember these points, as the cover and appearance of the cloth is affected by the beating-up being done in different kinds of sheds.

FIG. 95.

The loom crank is usually set at the top centre, or thereabouts, when the rising and falling staves are level, so that the shed will be partly open for the next pick by the time the loom crank gets to the front centre. In single-lift dobbies the beat-up is made when the shed is closed, and so the warp has not the same chance of being spread as with the timing of double-lift dobbies. This difference in the character of the shed when the beat-up is made is caused by the fact that in a double-lift machine the knives, being in the middle of their stroke, are moving at their quickest speed when the shed is closed, and in a single-lift the knife is almost stationary when the shed is closed. The same thing occurs in Jacquards, and the matter may be better understood by a reference to the chapter on Jacquards.

Dobbies can be made “positive” in various ways. Keighley dobbies are made with a pin fixed on the upright MN ([Fig. 91]) at the point H. A wire is hooked on to this pin and connected to an L lever at the side of the loom opposite the dobby; this is connected to a lever at the bottom of the loom. By connecting the bottom of one stave to this lever the stave will be pulled down as the upright MN is taken forward, and so the knife whilst taken one stave up is pulling another down, rendering the dobby positive. This, of course, will only do for certain simple patterns, such as twills, satins, and similar weaves, and would not do for patterns where different numbers of staves are lifted every pick. Positive dobbies are not much used in the cotton trade.

The ordinary form of dobby is non-positive, the stave being kept down by springs in some form or other. One reason which may be urged against ordinary springs, or “jack boxes,” is that the pull on the heald increases as the stave is lifted and the spring opened. It is obvious that this is just the reverse of what is required, as the stave is lifted positively, and the pull on it may therefore conveniently be decreased as it is lifted, and the healds would last longer. The use of the spring is to keep the stave down, and therefore it should exert its greatest force when the stave is at the bottom. A simple method of accomplishing this has long been in use. Something on the same principle has been used on hand looms for generations, and very cheap and convenient undermotions of this kind have long been available for power looms; but, strange to say, cotton manufacturers have been very slow at adopting them. An undermotion of this kind is illustrated at [Fig. 96]. The spring is fixed at A, and a wire hook connects the spring with the quadrant at B. It will easily be seen that as the stave lifts, the direction of the pull of the spring is gradually moved over the centre of the quadrant at C. If the stave were lifted until the spring was in a direct line from A to C, the pull on the stave would be nil, as all the force would be exerted on the fulcrum. Each stave is connected at both sides in the same manner, the springs and other parts are all arranged in a very compact manner, and the cost is very small.

FIG. 96.

Another form of undermotion on the same principle is much used in Yorkshire in the woollen and worsted trades. This is illustrated at [Fig. 97], and is known as Kenyon’s undermotion. In this the springs are arranged horizontally, and therefore longer springs can be used. The quadrant is centred at C, and a strap is fast to the quadrant at D. The spring is connected with the quadrant at F. The strap passes from the quadrant under the bowl B, and then to the stave. Another quadrant serves in the same manner for the opposite side. The spring is fastened to a bar at E, and as the stave is lifted, the pull of the spring is gradually moved over the centre C, and therefore the pull on the heald gradually decreases as it is lifted.

FIG. 97.

CHAPTER V
MISCELLANEOUS

WHEN two or more pieces are woven in one width and afterwards cut or torn apart, if there are not a few leno ends to divide each piece the warp threads have nothing to stop them from coming out at the cut sides. In light fabrics this is a greater disadvantage than in heavy and finely picked ones, such as velvets, and therefore in the former it is usual to weave a few ends leno to keep the edge firm. There are various kinds of motions for effecting this object, one of the oldest being that illustrated at Figs. [98] and [99]. This is for an ordinary plain loom, and the crossing end is taken through the back stave and through a loop from the top of the front stave. This loop is often formed of a small fine pliable chain, as it wears longer than worsted. [Fig. 98] shows the back stave lifted, and [Fig. 99], the front stave up, when it will be seen that the crossing end is brought up on the opposite side from the previous pick.

FIG. 98.

FIG. 99.

FIG. 100.

FIG. 101.

Another, and perhaps a better, method, is Shorrock and Taylor’s patent, shown at Figs. [100] and [101]. For a plain loom the two straps A and B are fastened to a drum on the top roller of the loom. In these straps are the small eyes C and D, and through these eyes the crossing ends are taken. The “standard” ends, round which C and D are crossed, are drawn through the fixed eyes EF, immediately above the small bobbins MN. The straps pass round the bobbins and up to the elastics X, which are fastened to a hook, L, at the top of the loom. The top roller is rocked to and fro by the ordinary staves, and when rocked in the direction against the elastics the crossing threads are brought up inside as shown at [Fig. 100], and as the roller rocks back the elastics pull the eyes C and D completely round the bobbins and take the crossing threads up the other side of the “dummy” or “standard” ends, EF. The selvedge formed is thus like that shown at [Fig. 102].

FIG. 102.

There are many patents taken out every year for split motions, but the simple old forms still keep their place.

Another invention of a totally different kind may be mentioned. In this, the weft is cut between the two cloths every pick as it is being woven, and the loose end is then turned round and taken into the cloth at the next pick, thus forming a practically perfect selvedge; indeed, it would be impossible for any one to find out the difference without being told or making a very close examination. For about half an inch at the inner side of the cloth there are double picks, but this is scarcely noticeable. The practical utility of this invention is yet to be proved, and one thing to militate against its general adoption is its cost, which is several pounds per loom, whereas some of the ordinary split motions cost only a few shillings per loom. With Jacquards or dobbies it is an easy matter to arrange an ordinary doup heald to form a split, but the arrangements before mentioned are used for plain looms, where it is not so easy to get the required lift. The twist used must be very strong, as no slackener is used. Usually it is a three-or four-or six-fold cotton thread.

FIG. 103.

Another kind of selvedge motion is that used for producing a plain selvedge on a loom weaving satteens with tappets. The fact of the ordinary satteen being five picks to the round, and a plain selvedge being a necessity, causes either the tappets to be made ten to the round, working the plain selvedges by tappets on the same shaft, or the selvedge ends must be worked from another shaft. In what is known as Smalley’s satteen motion the former principle is acted upon: the tappets are ten to the round, and the plain is worked from the same shaft.

A more ordinary form is that shown at [Fig. 103]. A small tappet, A, is fitted on the bottom shaft (or picking shaft), and this acts upon a lever, B, to which the bottom of one set of harness threads containing, say, the odd-numbered ends of the plain is connected, the other, or even-numbered, ends of the plain being connected to the elastic E, the bowl F at the top being used for working round. When the tappet presses down the lever B, it will take half the plain ends down and the other half up, and the elastic will pull back again as the tappet allows it. In this way a plain selvedge is obtained in a five-shaft satteen.

FIG. 104.

Another method of effecting the same purpose is shown at [Fig. 104]. A shaft A is placed under the loom, and this shaft is made to rock to and fro to work the mails B and C alternately up and down. The picking shaft of the loom has a crank M fastened to it, and a strap S is taken from this crank to the small drum H on the shaft A, and is wrapped round it. As the crank M revolves it will pull the shaft A in one direction until the crank gets to the top, and when the crank has passed the top of its stroke the spring X will pull the shaft back to its original position, and thus the required reciprocating motion is given to the shaft A and to the mails B and C.

Double-beat Slay.

A double beat is sometimes required to be given to each pick of weft. This is done in weaving some of the heaviest kinds of sackings, carpets, and similar fabrics. [Fig. 105] shows how this is effected. AB is the slay, and is movable about B as a centre; EC is in two pieces, viz. ED and DC, and these are fitted loosely on a pin at D. It will be obvious that when the crank occupies either position QP or QP′, the slay will be at the front of its stroke, and as the crank is moving from P to M it will pull the slay back a little, and in moving from M to P′ a second beat-up will be made. Whilst the crank is moving from P′ to P the shuttle is passed through the shed. It is obvious that a beat-up of this kind will enable the weft to be beaten well up into the cloth, and more to be put in than with a single beat. The force exerted is often so great that the looms have to be very firmly fastened into the floor on which they stand, or they would move.

FIG. 105.

CHAPTER VI
JACQUARD WEAVING

THE Jacquard machine was the invention of a Frenchman of that name, who exhibited the machine about the year 1800. It was introduced into this country about twenty years later. The chief advantage of the machine is that a large number of warp threads can be operated separately, and a larger figure be produced than with a shaft harness. The chief ideas in the machine are that each mail is connected separately to its hook, and the use of perforated cards to leave any hook over the griffe if it is required to be lifted, or to push it away from it if the hook is required to be left down in the shed.

The original Jacquard machine was a single-lift, and although many minor improvements have been made in it, the main features are practically the same to-day as in the earliest machines introduced into this country. At the present day the single-lift is comparatively little used in cotton manufacture owing to the increased speed at which double-lifts can be worked, but it is still preferred in silk manufacture for several reasons. One reason is that the character of the shed when beating up in a double-lift machine is essentially different to that produced by a hand-loom, where of course a single-lift is always used, and as hand-loom fabrics have a finer touch and appearance than power-loom fabrics, the object is to imitate the hand-loom production as nearly as possible. The cause of this difference in the character of the shed when beating up will be explained later in this chapter. Another reason is that silk-looms could never be run at any speed higher than that of which a single-lift machine is capable, and therefore the advantage of increased speed of the double-lift is of no use.

Double-lifts, owing to the counterpoise and the division of the work on to two knives, are undoubtedly steadier in working, and this is an argument decidedly in their favour. Single-lifts are still used in the manufacture of figured lenos, as no shaking motion has yet been successfully adapted to enable the crossing ends to cross with a double-lift machine.

FIG. 106.

A single-lift Jacquard for weaving a pattern which occupies 400 ends in a repeat consists of 400 hooks and 400 needles, with an extra row of eight hooks for selvedges, or other auxiliary use. The hooks are arranged in eight rows with 51 hooks in a row. A cross section of this Jacquard is shown at [Fig. 106], where the uprights are the hooks and the horizontal wires the “needles.” A is the “needle board,” and this is a perforated board through which the needles pass. The bottom needle B is twisted or looped round the back hook D, and the connection of the other needles and hooks is shown. At the back of each needle a small spring made of fine brass or steel wire is placed. These springs are held in position in the “spring-box” S. There are, therefore, 408 springs required for the 408 needles. The hooks rest on the grate G, but in some makes of machine the grate is not used and the hooks rest upon a “bottom board.” In this case the hooks are very liable to turn round, and thus cause annoyance. To prevent this, flat hooks have been used, and the needle loop was shaped so as not to permit the hook to turn within it. The eight knives form the griffe. These knives are all fastened together, and are moved up and down from the crank-shaft of the loom. The illustration shows the knives at the bottom of their stroke, and at this point, or immediately after the griffe begins to move upwards, the card on the perforated cylinder E is pressed against the needles, and if there is a hole in the card, the needle directly opposite the hole will pass through it and into the perforation in the cylinder, and the knife will take up the hook to which this needle is connected. If the card is blank opposite any needle it will press back the hook, and as the knife lifts, the hook is left down. Thus it is possible to lift any of the 408 hooks in the machine for any pick. When the cylinder is taken away from the needles the hooks are forced back into their original position by the small springs in the spring-box S.

It will be noticed that the knives are leaning a little, and the reason for this will be apparent, as if they were not leaning they would catch the tops of the hooks in coming down, and would break or bend them. The sloping position enables the knives in coming down to press back the tops of the hooks and so get under them, ready for the next card to be pressed against the needles. The knives should come down low enough to be quite clear of the hooks, and therefore in this machine there is a considerable dwell when the shed is closed.

The harness for a straight-over pattern is mounted as shown at [Fig. 107]. In order to prevent confusion the connection of the cords to the machine is not shown, but the numbers on the line A represent the hooks in the machine to which the cords are to be attached. The “comber-board” or “cumber-board” B is a frame into which perforated slips are fitted. These slips are perforated to different degrees of fineness, the fineness being regulated by the number of ends per inch required in the cloth to be woven. The lingoes, L, are metal weights, and serve the purpose of keeping the mails down. MM are the mails, through which the warp threads are drawn in the order shown by the numbers, beginning at the back left-hand corner. The draft in straight-over patterns is always taken in this way in Jacquard weaving, although it is not compulsory. The harness is built with linen thread, and the method of tieing the lingoes to the hooks will be understood from the diagram.

FIG. 107.

When one lingoe has been connected to each of (say) 400 hooks, the first pattern is complete. Supposing there are 100 ends per inch, the pattern will occupy 4 inches, and therefore if cloth is required 28 inches wide in the harness, there must be seven lingoes attached to each hook, making seven patterns, or seven repeats of the pattern, in the width of cloth. Thus when one lingoe has been tied to each hook, beginning with the first and ending with the 400th, another is connected to each hook, beginning again with the first; and when this is done other patterns are formed in the same manner until the required number is complete.

FIG. 108.

It is important to have a clear understanding as to which is the hook which lifts the first end in the draft. This hook is the one connected with the bottom needle in the last row on the 25-side of the machine. As we stated previously, a 400s machine has 400 hooks arranged in 50 rows of 8, or 8 rows of 50 hooks, and in addition there is always a spare row of hooks, making 51 rows in all. As it is necessary to lace the cards in the middle as well as at the sides, a space has to be allowed for the lace holes, and therefore the machine is divided into two parts by a space between the 26th and 27th rows.

A plan of a card is given at [Fig. 108]. The length of the card between the two peg holes A and B is nearly 14½ inches, and the distance between the centre of the top needles and the bottom needles is 1⅞ inch exactly. This holds good for all English-made machines, but the American index is different.

It will be seen that there are 26 rows on the right of the machine and 25 rows on the left, and one is called the “26-side” and the other the “25-side.” The cards are always numbered at the “26-side,” and the cutting is commenced at this end. It may be as well to explain here the order in which the holes are cut from the design, as it will assist in following the point paper design to the loom. The cutting is usually done in a “piano” cutting machine, which will be explained more fully later on. By this machine one row of eight holes can be cut by operating eight punches and pressing down the right-foot treadle of the machine.

The number end of the card is gripped by the machine, and at the first stroke of the right foot, the lace holes EF and the peg hole A are cut, then one stroke of the treadle is made without cutting, and the pointer of the machine arrives at the 1st or spare row. If the selvedges are worked from this row, holes are cut accordingly. Then the pointer comes to the 2nd row, and in this row the cutting from the design is commenced.

At [Fig. 109] the design is made on point paper, as it is required to appear in the cloth right side up, with the twill in the ground running in the same direction as shown on the design. When cutting, the design is usually turned round, as shown at [Fig. 110], and the cutting commences from the top right-hand corner A. To show the matter clearly, the first row of holes cut are numbered, both in the design and the card, in consecutive order.

FIG. 109.

FIG. 110.

The first hole cut in the card is operated with the little finger of the right hand. Following this hole to the loom, we find it operates the last or 400th end in the draft, and that the hole cut last on the card (numbered 400) operates the first end in the draft. This is the hole which operates the bottom needle in the last row on the “25-side” of the Jacquard machine, which, as was previously stated, is the hook from which the draft begins.

Following out the operation of cutting the card. When the 26th row has been cut, the lace holes MN ([Fig. 108]) are cut, and then the cutting is again straight-forward to the 50th row. The piano machine is so constructed that with the same stroke of the treadle which cuts the 51st row the peg hole D is also cut, and then follows a stroke without cutting, after which the two lace holes T and Y are cut. This makes 56 strokes of the foot for each card.

It is usual, in order to economise space, for the Jacquards with straight, or “Norwich,” harnesses to be placed on the loom, so that on one loom the cards hang over the weaver’s head, and on the next the cards are at the back of the loom. In both cases the harnesses are built the same way, but in one case (cards over weaver’s head) the thread operated by the bottom needle on the “25-side” will be at the back of the comber-board, at the left hand; and in the other case (cards behind loom) the same thread will be at the front of the comber-board at the right-hand side.

As previously stated, the single-lift Jacquard for cotton weaving is not often employed except for special purposes, such as figured leno weaving. The advantage possessed by the double-lift Jacquard as regards speed is so very considerable that its adoption for ordinary forms of cotton weaving has become universal; and the advantage of speed is not the only advantage it possesses, as will be pointed out shortly.

A double-lift machine with one cylinder for a 400 end pattern consists of 800 hooks and 400 needles. Each needle is twisted or bent round two hooks, as shown at [Fig. 111]. The hooks are connected together in twos by neck cords, which are usually strong whipcord, as will be seen from the illustration. It will be seen that the bottom needle is bent round the back pair of hooks, the next needle round another pair, and so on. Each needle has a spring behind it, as in a single-lift machine.

FIG. 111.

There are two griffes, which work oppositely—that is, as one goes up the other comes down. The griffes (or knives) are worked by a double crank on the bottom shaft of the loom, so that each griffe moves from the bottom to top of its stroke in one pick, and from top to bottom in another pick.

The principle of the double-lift will be understood from [Fig. 112]. One knife, A, is at the top, and the other knife, B, is down. One hook of the pair is lifted, and therefore the ends in the mails connected to the neck cord at C will be lifted. Suppose now it is required to lift the same ends of warp for the next pick: a card is pressed against the needles, and if there is a hole in the card opposite the needle E, it will leave the needle and the hook N where they are, and as the knife B is lifted, the hook N will be taken up as the hook M is coming down. The hooks will cross at about the middle of their stroke, and the weight of the ends and lingoes on the cord C will at that moment pass from the hook M to the hook N. In the diagram the cord attached to the hook N is slack, but when this hook is lifted the cord will gradually tauten until it bears all the weight, when the cord from the hook M becomes slack. We thus have the ends for the second pick lifted whilst the ends which were up for the previous pick are coming down. This is where the advantage of the double lift lies. In a single lift the knife must lift the hooks up and then come down to the bottom before another card can operate the needles, whereas in a double lift the card for a second pick can be brought against the needles as soon as the ends which were up for the previous pick are ready to come down.

FIG. 112.

It is obvious that in the position shown in [Fig. 112], when one knife is up and the other down and the needle pressed back by the card, that the hook M will also be pressed back, as shown by the dotted line. The bend of the hook over the knife, therefore, must be sufficient to prevent the hook being pushed off the knife, and it will be noticed that the hooks in this class of machine are bent more than the hooks in a single-lift machine. The hooks rest on the grate G, [Fig. 111], and the shape of the hook at this point acts as a spring to straighten the lifted hooks after the pressure of the card has been taken off the needles. A machine of this kind can be run at a speed of about 160 or 170 picks a minute, as compared with the 130 or 140 picks of a single-lift.

A double-lift machine on another principle is illustrated at [Fig. 113]. This is a two-cylinder machine, and to weave a pattern repeating on 400 ends this machine requires 800 hooks and 800 needles. The cylinders work at opposite sides. The hooks are placed as shown in the diagram, the hooked parts facing each other in pairs, and by following carefully the manner in which the needles are twisted round the hooks it will be seen that there are really two single-lift machines placed together, alternate rows of hooks representing each machine. There are two griffes, as in the double-lift single-cylinder machine, and the griffes are worked in the same manner.

FIG. 113.

The cylinders work alternately, the cards being laced in two sets, all the odd numbers being together in one set and the even numbers forming another set. Immediately one knife is at the top and the other at the bottom, one cylinder is pressed against the needles, and it will be noticed that the hooks which each cylinder operates have the hooked parts in the direction of the cylinder. When the hooks operated by one cylinder are at the top the other cylinder is pressed against the needles, and thus the work done by one cylinder in [Fig. 111] is divided between two in this machine. The advantage of this machine is in the lessened speed of the cylinders. The vibration caused by the cylinder working at a high speed in a single-cylinder machine is so great that the limit is reached at about 170 picks per minute, whereas a double-cylinder machine can be run up to 200 or sometimes even more picks per minute, though perhaps 180 is a more advantageous speed. The top set of needles project a little further through the needle board to compensate for the difference in leverage on the hooks.

Besides the advantage of speed, double-lifts have an advantage in the counterpoise obtained by one set of hooks going up as the other comes down. This causes a more even motion and steadier working. Another advantage possessed by double-lifts is that the beating up of the weft is effected in a crossed shed, thus enabling more weft to be put in than in a single-lift, where the beat-up is done with a closed shed. This beating up in a crossed shed also spreads the warp better, and prevents the reed marks from showing, for the same reason as was given when referring to the spreading of the warp in the tappet loom.

In silk weaving a single-lift machine has an advantage in imitating more closely hand-woven goods, as hand-loom weavers usually beat up in a closed shed. This causes the weft to be put in straighter—that is, less wavy, which is very desirable in silk fabrics.

The cause of this difference in the shed when beating up in the two kinds of looms will be understood by following the relative positions of the griffes and the loom crank throughout its revolution.

In a single lift the time allowed for opening and closing the shed must be used to the best advantage; that is, as much time as possible must be given for this purpose. On this account it is necessary to pick the moment the slay is sufficiently far back to allow the shuttle to enter the shed—that is, when the slay is half-way back, or the crank at the bottom centre. The griffe is worked by a crank on the top shaft of the loom, and there is no actual dwell of the griffe or of the ends when the shed is open; therefore the shed must be opened a little wider than would otherwise be necessary for allowing the shuttle to pass through.

The shed must be sufficiently open to allow the shuttle to enter when the loom crank is at the bottom centre. This regulates the timing of the other parts. [Fig. 114] will make this quite plain. The shed must be nearly fully open when the crank is at the bottom centre to allow the shuttle to enter; and when the loom crank is at A the griffe must be nearly at the top. When the crank is at B the griffe will be at the extreme top, and when the crank is at the top centre, or C, the griffe will be as near the bottom as it was to the top when the loom crank was at A. As was previously pointed out, the griffe must go further down than the hooks to allow another card to operate the needles, and therefore it is when the loom crank has arrived at C that the knife is leaving the hooks resting on the grate, or bottom board. The griffe will be at the extreme bottom when the loom crank is at D, and when the griffe is up at the hooks again the crank is at the front centre, or E. Thus the shed has the fraction of a revolution between B and C to close in, and between E and B for opening. The shed remains closed for the quarter of a revolution, C to E.

FIG. 114.

FIG. 115.

In a double-lift the warp is much more leniently dealt with. As we have said, the shed must be open for the shuttle to enter when the loom crank is at the bottom centre. Therefore the griffes should be in their extreme position—one up and one down—when the crank is at the bottom centre.

The timing of the parts in a double-lift will be seen at [Fig. 115]. The cranks that work the griffes are on the bottom shaft, which of course makes a revolution every two picks. These cranks will be perpendicular when the shed is fully open; therefore when the loom cranks are at the bottom centre the cranks which drive the griffes must be in the position AB. If they are so set they will be in the position CD when the loom crank reaches the back centre, and in the position EF, or horizontal, when the loom crank arrives at the top centre, when the shed will be closed. We have thus a closed shed when the crank is at the top centre, as in a single-lift; but in this case when the shed is closed the griffes are moving quickly, whereas we have a quarter of a revolution dwell after the loom crank reaches the top centre in a single-lift. This causes, as we shall see, a difference in the shed when the slay beats-up, or is at the front centre. When the griffe cranks are in the position GH, the loom crank will be at the front centre, and thus the shed will be partly opened for the next pick when the reed comes in contact with the cloth.

Jacquards are made in various sizes. 100s, 200s, 300s, 400s, and 600s, are the most common. 100s are arranged in rows of four; 200s and 400s are in rows of eight; 300s and 600s in rows of twelve.

There are two distinct kind of harness mounting, the London and Norwich systems. In the former the Jacquard is placed with the narrow end towards the front of the loom, thus causing the cards to fall at the side. In the Norwich system, or “tie,” the machine is placed with the broad side facing the front of the loom, thus causing the cards to hang either over the weaver’s head or at the back of the loom. On this system, as there are eight rows in a machine, by taking the comber board eight rows deep the harness becomes what is called a straight neck. With the London system, the end of the machine facing the weaver, there must be a twist in each pattern in the harness. There is not much to choose between the two systems. Some prefer the London tie, as they say the twist in the harness causes the harness threads to support each other, and so last longer. The Norwich system is the more common, especially in the cotton trade.

FIG. 116.

[Fig. 116] shows the method of tying up the harness on the Norwich system for a bordered fabric, such as handkerchiefs. In these goods it is usually preferred that both borders should point inwards, as in the sketches Figs. [116] and [117].

FIG. 117.

The hooks to which the harness threads are attached are numbered on the line A, and it will be seen that the draft begins in the left-hand corner at the back of the comber-board, the lingoes being numbered in the order of the draft. The cords are tied up just as for an ordinary straight-over harness for the first 400, or one full pattern of the machine, but then, instead of commencing with the first hook again, the 201st lingoe is tied to the 201st hook, and the second half of the pattern is repeated. This forms the middle of the handkerchief, and it must be repeated over a sufficient number of times to give the required width of cloth after allowing for the trimming and border. In Figs. [116] and [117] nothing but a border and middle are shown, but sometimes a trimming of another small weave is required outside the border, and this, which is usually on a small number of hooks, is repeated over in the same way as the middle. In [Fig. 118] only two repeats of the middle are shown; but supposing that the harness had 100 ends per inch, and that the handkerchief was required to be twenty-four inches wide excluding the border, there would be twelve repeats of the middle required. When the middle has been repeated over a sufficient number of times, the other border must be tied up, and to obtain the reverse position of the figure the draft must be reversed. By tying the next lingoe to the 200th hook, and going backwards with the draft, it can easily be seen that the same figure will be woven at this side of the harness as at the opposite side; the only difference will be that the figure will point to the left, as will also the twill in the ground, if it is a twill. This system of tying up is compulsory in the Norwich system, as it is usual to keep the harness straight—that is, the harness threads from each of the eight rows in the Jacquard each form a separate row in the comber-board. We have thus eight rows in the machine and eight rows deep in the comber-board, and it would not do to have a thread taken from the front of the comber-board at one border and from the back of the comber-board at the other border to the same hook.

FIG. 118.

If the harness is a “London tie” it necessitates a half-turn in each pattern, as the machine is at right angles to the comber-board. Therefore the draft may be continuous, as shown at [Fig. 117], where, after the middle has been repeated a sufficient number of times, finishing with a thread from the 400th hook at the front of the comber-board, the next one is taken from the 200th hook through the back of the comber-board, and the border will finish with a thread from the first hook going through a hole at the front of the board—just the reverse to the other side.

Bordered goods are often made with two borders at each side, and sometimes the borders are repeated a few times. The number of hooks taken for the border and middle respectively vary according to requirements. Sometimes, in a 400 machine, 300 will be taken for the border and 100 for the middle, and so on. The cross-border must of course be designed, and the cards cut. The number of cards in a set in these goods is often very large, as the middle must be repeated over the required number of times, and there will be as many cards used in the set as there are picks in the handkerchief.

In designing for the mounting given at [Fig. 117], the design would be made on 400 ends: 200 for the border and 200 for the middle, and the cards would be cut just in the ordinary manner. The cross-border would also be designed in such a manner as to harmonize with the side borders. The portion to be designed is enclosed by the dotted lines.

Centre ties or point ties are another class of harness in regular use. This is really the two borders of a bordered harness joined together. [Fig. 118] shows how the tying up is done for a pattern of this kind. The first 400 threads are connected as usual, the draft being from back to front. When the 400th has been reached, the draft is reversed until No. 1 is arrived at again. The same effect is obtained as in a point of V draft in a shaft harness. The pattern must be of such a character that one half is the exact reverse of the other. This kind of harness is used for weaving large damask figures, and it is obvious that the effect produced is really that of a figure on 800 ends, or twice the size of the machine. Designs of this character are of course rather stiff, but are suitable for damasks, and similar fabrics.