Fig. 1478 represents a similar section of Brussels carpet with the shed in like manner open for the insertion of a new wire. The texture of the fabric is very similar to the preceding, so far as the interweaving of the threads gees. The body-warp b serves the same purpose as in the preceding case, but is differently composed, there being three, four, or five times as many threads in the pile warp as before, and only or of these are raised at one time to form the pile. The remainder take the position of the filling-warp in the tapestry carpet. In Brussels carpet, the pile-threads are iodividually of one colour only, and not printed to form the design as in tapestry, the jacquard lifting the coloured threads to form loops as required by the nature of the design. Some of the pile-threads being required to be raised very frequently, and others only at long intervals, necessitates the use of a creel or bank in which bobbins containing only one thread each are placed, in this differing from the tapestry carpet. The pile-threads are all wound upon one beam. The embodiment of such a large proportion of the pile-threads in the filling-warp causes the colour of the threads to appear at the back—indubitable evidence of its being a true Brussels texture.
Fig. 1479 illustrates the construction of the web of a velvet or cut-pile carpet, which is almost identical with the Brussuls in texture, except that two threads of weft c are used for bindine the pile. The pile wires e ale also elliptical in the form of their section, thus making a deeper pile when inserted ; and, being furnished with a knife-cdge at the distant extremity, cut the loops when withdrawn, as illustrated in Fig. 1479. Fig. 1480 is a full-size illustration of the wire used for weaving a cut-pile fabric, g being the blade or cutting end.
Figs. 1481, 1482, 1483 exhibit the order of " shedding," by which the Brussels carpet is com posed. Fig. 1481 is for the passage of the shuttle aud the reception of the weft. All the pile threads a have been raised, and also one-half of the body-warp b, the remainder being left down. The shuttle is then thrown through the open shed, leaving the pick of weft that shows itself at the back of the fabric. The next movement opens the shed shown in Fig. 1482, which is also for the passage of the shuttle, and the reception of a thread of weft. In this, all the pile-threads are down, and the two portions of the body-warp b remain as before, the shuttle traversing the shed again puts in the top pick or thread of weft, which is the binding pick by which the loops formed upon the wire last inserted are tied securely in position. Fig. 1483 shows the last shed necessary for the formation of this particular fabric. It is for the reception of the wire used in the formation of the loops ; all the body-warp threads b are left down, also all the pile-threads a not required for the design ; those needed to form the pattern are lifted by the jacquard, and, on the insertion of the wire, are retained in loops by it until a binding thread of weft has tied them down. To keep them securely fixed, the wires are left in until a sufficient length has been woven to prevent them being drawn backwards.
As observed before, the problem to be solved in connection with the carpet-loom was the invention of an automatic plan of inserting and withdrawing the wires to form the loops. As in connection with nearly all ether mechanical inventions, the present approximately perfect system has been the result of growth, the first rude effort serving as the foundation upon which succeeding inventors have built. An arrangement was first constructed by which from a bundle of wires a single ono was successively drawn as required, and, by means of a pair of nippers, fixed at the end of a reciprocating red, was carried into the shed ; and after having been woven into the fabric and subsequently withdrawn from the loops, the wires were retnroed by the hand of tbe weaver to the bundle. An improvement followed this by which the whole operation was rendered self-acting. This consisted of dropping wires successively from a hopper into a longitudinal groove
in a rod, which was carried into the shed in guides, and was then caused to make a half-revolution by means of a screw-incline on the rod, by which tho wire was dropped into its place in the shed. The wires were successively- withdrawn from the fabric by reciprocating nippers, and carried up again into the hopper by endless chains. The next step was to simplify this by a plan of placing the wires separately in a groove from which in succesidon they were put:hed into the shed, and boihg constructed with a hook at the back cnd, each was subsequently, at the proper moment, withdrawn and transferred to the feeding-trough for use again.
In these first attempts, various rueana wcre used to support the forward extremity of the w irea, as it was assumed that they could not be otherwise held with sufficient rigidity to secure their correct insertion without being thus sustained. This plan was, however, nfterwards abandoned, as by the aid of further itnprovernents it was shown to be unnecessary.
Some years ago, W. WeiId, of Manchester, invented a system of inserting the wires in carpet looms which, with subsequent itnprovements, has remained in favour up te the present time. This plan is knovrn as the roller wire-motion, and is fixed to one side of the loom. Figs. 1484, 1485, 1486, show front elevation, plan, and section of this attachment. In the two first figures, a wire a is shown being inserted into the shed of the fabric. In the plan, Fig. 1485, is shown a portion of the woven fabrio, and the reed with which the weft and the wires are premed to the fell of the fabric—the part Nat woven. Fig. 1486 is a transverse section of the roller r in its casing e and the slide t for inserting and withdrawing the wires. The roller and its casing e are rather longer than the width of the carpet to be woven, and the upper part of the casing is cut away through the extent of one-sixth of the circumference of the roller, as shown in the figure. In the aurfuee of the roller r six longitudinal grooves are cut, of a width and depth corresponding to tbe diameter of the wires. The roller enclosed in its case is fixed with its end about 9 in. from the edge of the carpet, and with its upper surface parallel and in the same plane as the " " of the carpet. At the end of the casing nearest the warp, a recess is formed between the end of the roller r and a hoop i, Fig. 1485, concentric with the roller. In this recess, which extends through * of the circumference of the roller, the heads of the wires are held when inserted in the fabric as shown in the plan. The hoop keeps the heads of the wires down in their places, and prevents them being pushed too far into the warp, and at the same thne prevents their being withdrawn before the proper moment. A projection upon the end of a spring j attached to the side of the casing keeps the heads of the wires erect, pressing them against eaoli other,.and against the end of the recess during the whole time that the wires remain inserted in the carpet. At the opposite end of the roller, is a projecting collar formed all round the roller, between one end of the easing e and a hoop carried by a bmcket. This collar contains aix grooves in which to receive the heads of the wires, and retain them in their correct radial position, whilst being carried round by the roller, in the direction ahown by the arrow in the sectional view. When the roller is stationary, two of its six grooves are opposite the two extreme wires in the recess i, Fig. 1485,one being opposite the wire last inserted, and the other opposite the wire next to be withdrawn froru the fabric. The other four grooves h have each a wire lying in them as shown in the section, so that each time the roller makes * rev., one wire is brought round to the place for insertion into the shed of the warp, and the wire last drawn from the carpet is carried away.