In order to convert this roving into a thread, it is necessary that the fibres, which are for the most part curled up, and which lie in all directions, should be stretched out and laid lengthwise, side by side; that they should be pressed together so as to give them a more compact form; and that they should be twisted, so as to unite them all firmly together. In the original method of spinning by the distaff, those operations were performed by the finger and thumb, and they were afterwards effected with greater rapidity, but less perfectly, by means of the long wheel and spindle. For the same purpose, Arkwright employed two pairs of small rollers, the one pair being placed at a little distance in front of the other. The lower roller in each pair is furrowed or fluted lengthwise, and the upper one is covered with leather; so that, as they revolve in contact with each other, they take fast hold of the cotton which passes between them. Both pairs of rollers are turned by machinery, which is so contrived that the second pair shall turn round with much more swiftness than the first. Now suppose that a roving is put between the first pair of rollers. The immediate effect is merely to press it together into a more compact form. But the roving has but just passed through the first pair of rollers, when it is received between the second pair; and as the rollers of the second pair revolve with greater velocity than those of the first, they draw the roving forwards with greater rapidity than it is given out by the first pair. Consequently, the roving will be lengthened in passing from one pair to the other; and the fibres of which it is composed will be drawn out and laid lengthwise side by side. The increase of length will be exactly in proportion to the increased velocity of the second pair of rollers.

Two or more rovings are generally united in this operation. Thus, suppose that two rovings are introduced together between the first pair of rollers, and that the second pair of rollers moves with twice the velocity of the first. The new roving thus formed by the union of the two will then be of exactly twice the length of either of the original ones. It will therefore contain exactly the same quantity of cotton per yard. But its parts will be very differently arranged, and its fibres will be drawn out longitudinally, and will be thus much better fitted for forming a thread. This operation of doubling and drawing is repeated as often as is found necessary, and the requisite degree of twist is given by a machine similar to the spindle and fly of the common flax-wheel.

The spinning-mule, invented by Samuel Crompton, carried the mechanism of the cotton-factory many steps in advance. Long after Crompton, came the self-acting mule. It is a carriage some twenty or thirty feet long, travelling to and fro, and drawing out the most delicate threads through hundreds of spindles, whirling at a rate which scarcely permits the eye to trace their motion. Mr. Whitworth says,—"So great are the improvements effected in spinning machinery, that one man can attend to a mule containing 1088 spindles, each spinning 3 hanks, or 3264 hanks in the aggregate per day. In Hindustan, where they still spin by hand, it would be extravagant to expect a spinner to accomplish one hank per day; so that in the United States [and in Great Britain also] we find the same amount of manual labour, by improved machinery, doing more than 3000 times the work."

Samuel Crompton, inventor of the spinning-mule.

Of the rapidity with which some portions of the machinery operate, we may form an idea from the fact that the very finest thread which is used in making lace is passed through the strong flame of a lamp, which burns off the fibres without burning the thread itself. The velocity with which the thread moves is so great, that we cannot perceive any motion at all. The line of thread, passing off a wheel through the flame, looks as if were perfectly at rest; and it appears a miracle that it is not burnt.

Cotton Mule-spinning.