Page:The American Cyclopædia (1879) Volume IV.djvu/29

 CARPET 21 produced, he approximated more nearly than the hand weaver to a perfect match in the figure ; and this he effected by taking up the woven cloth by a regular and positive motion which was unerring, the same amount for every throw of the shuttle and beat of the lathe. As the weft threads are not spun regu- larly, and the weaving in of the warp threads and passing the different colors from the upper to the lower ply or cloth to produce, the figures require sometimes more and sometimes less to make a given length, he determined to regulate the delivery of the warps as required by their tension, thereby throwing the irregularities into the thickness, where they cannot be noticed, instead of into the length, where they would destroy the match of the figures. He accom- plished this by suspending a roller on the woven cloth, between the lathe and the roll- ers that take up this cloth, so that when the cloth was being woven too short, which in- dicates a deficient supply of warps, the roller would be elevated, and by its connection in- crease the delivery motion to give out more warps, and vice versa. Still this served only to prevent the further extension of a fault already incurred. The roller, to perfectly ac- complish its purpose, should have been applied to the unwoven warps, which seemed then impracticable ; for when the lathe beats up the weft, these must be rigid to resist the beat, and no Avay was apparent to make the roller sensi- tive to detect and indicate the amount taken up. The warps, moreover, are necessarily all rolled up on the warp beam with equal tension, and so can only be given out equally. The improvement was afterward perfected by Mr. Bigelow in the following manner : Each warp thread in the usual way passes through a loop called a mail, attached to a card suspended from the jacquard, and each card has suspended to it a weight, all the weights being equal. The two trap boards of the jacquard move simultane- ously, one up and the other down, and in these movements they catch or trap such of the cards (determined by the combination of cards) as are required to bring up the proper warp threads at each operation to produce the figure, leaving down such of them as are not required at that particular operation ; and when the two trap boards are on a level, and all the warp threads connected with them are in a horizontal line, and those not connected with them hang down with the suspended weights, the lathe beats up the weft thread, which lies between the warps that are in a horizontal line, at the same time exert- ing a force on the weft threads previously thrown, and beating them up more closely. Now, as the warp threads are all connected at one end with woven cloth, and at the other with the beam, it follows that those which are hanging down in a bent line will receive a greater proportion of the force of the beat of the lathe than the others ; and as all the warp threads in succession take this position, and all iave an equal weight, it follows that each suc- cessively receives the same pull at the time the lathe beats up ; thus the tendency to irregular- ity of surface from the varying lengths of warp threads taken up in ingraining is counteracted. The selvage was made smooth and even by a contrivance which regularly gave a pull to the weft thread after the shuttle was thrown. Mr. Bigelow at last, by these improvements and others which he introduced, brought the loom to average from 25 to 27 yards a day of two- ply, and from 17 to 18 yards of three-ply car- pet. His improved method of producing fig- ures that will match was afterward introduced, and patented in 1845. The same machinery was found to be applicable to the manufacture of Brussels and tapestry carpets, the weaving of which otherwise than by hand was before generally considered a mechanical impossibility. With the hand loom they were made at the rate of three or four yards per day ; but with the improved loom the production was increased to 18 or 20 yards per day. The carpets, too, were made more exact in their figures, so that these perfectly matched, and their surface was smooth and regular. They surpassed, indeed, in their quality the best carpets of the kind manufac- tured in any other part of the world. The looms of Mr. Bigelow were introduced into factories built at Lowell, Mass., and Thompson- ville and Tariffville, Conn., for their use, and others were established at Clinton, in Worces- ter co., Mass., where carpets are now made to the annual value of about $1,000,000. Brussels carpet is so named from Brussels in Belgium, whence the style was introduced into England in the last century. It is made upon a ground of linen weft, which is con- cealed by the worsted threads that are in- terlaced with and cover it. The threads are commonly of five different colors. In the weav- ing these run the length of the web, and are so managed that all those required by the pattern are brought up together across the line of the carpet; before they are let down, a wooden instrument called a sword is passed through to hold up the threads ; this is replaced by a round wire, which, being at last removed, leaves a row of loops across the carpet. In a yard length the number of successive lifts of the sets of colors required is sometimes as many as 320, each of which forms a row of loops. Four colors must always lie beneath the fifth, which appears on the surface, and thus the carpet, with its linen weft too, is thick and heavy. The Wilton carpet, the moquette of the French, differs from the Brussels in the loops being cut before the wire is removed, a groove in the flat upper surface of the wire admitting of their being cut by passing a knife along the surface. The soft ends give the carpet a rich velvety appearance. In the imperial Brussels carpet the figure is raised above the ground of the pattern, and the loops of this are cut, but not those of the ground. Various methods have been devised of simplifying the processes of making Brussels carpet. Richard Whytock of Edin-