SEWING MACHINE thread No. 1 on the back side. While the old loop thus twisted and spread out is held open on the body of the hook, the point 19 enters the new loop and carries it into the old one, as seen in fig. 16; and as the hook continues to revolve the cast-off, 20, passes out of the old loop and leaves it to be drawn up to the under side of the fabric, as in fig. 17, which completes the stitch. One feature of the ma- chine consists in the application of what is called an automatic tension, lately devised and patented by Mr. Charles H. Willcox. Instead of subjecting the thread to a continuous ten- sion produced by partially confining it in a SEXTANT 803 FIG. 16. FIG. IT. groove or clamp through which the motion of the machine draws it, it is made to pass between two disks held together by a spiral spring firmly enough to hold the thread in- flexibly and draw it through the fabric to a definite distance until more is required to make a new loop, when it is instantly relieved by a little piston rod striking against the lower end of a rod attached to the upper disk. A uni- formity in the drawing up of each stitch is thus secured, and as the necessity for change in tensioD when different sizes of thread or thicknesses of cloth are used is done away with, no provision is made for change by the operator. The Grover and Baker machine makes what is called the double-loop stitch, employing two threads. Instead of passing a single thread through the loop formed by the eye-pointed needle in piercing the cloth, as in the lock-stitch machines, a loop made by a circular horizontally moving needle (fig. 18) below the cloth is passed through the former loop, and the needle at its next descent passes through the loop formed by the lower needle, so that the stitch is really formed by passing a loop of one of the threads through that of the other. The construction of the stitch thus made will be understood by inspecting fig. 19, where the threads are wide- ly drawn apart for the purpose of showing it. When the loops are properly drawn to their places underneath the cloth, a durable elastic stitch is produced, which is very desirable in elastic as well as other goods, and may be used with advantage in embroidery by using differ- ent sized threads. In this article it has been sought to explain the actions of some of the principal sewing machines, without entering into the details of their construction or a de- FIG. 18. scription of the various attachments for hem- ming, felling, ruffling, cording, &c., or of but- tonhole machines, which are either attachments Fio. 19. to sewing machines, or made specially for this work, and which are employed with more or less perfection of execution. SEXTANT (the sixth part), an instrument used in nautical observations and in hydrographic and land surveying for measuring the angu- lar distance between objects. Its principle and the manipulation of it, in its ordinary form, are explained in the article QUADRANT, which instrument it has entirely superseded for observations at sea, partly on account of its greater portability, but mainly because of its capability of measuring a wider range of angles. For important land surveys a full circle is preferred, of which there are the fol- lowing forms: 1. A simple reflecting circle, made by extending the arc of the sextant to a whole circumference and producing the index arm so that it can carry a vernier on each ex- tremity. Observations with this are free from the error of eccentricity, and in part from the accidental errors of graduation and of reading, since they are derived from the mean of two readings at opposite divisions of the arc. 2. The repeating reflecting circle, which differs from the first only in having its horizon glass and telescope fixed to the arm which revolves about the centre of the instrument, instead of permanently attached to the frame. By ta- king a sufficient number of cross bearings with this, errors of reading, eccentricity, and imper- fect graduation are essentially eliminated. In theory, therefore, the repeating circle is very nearly perfect, capable of eliminating its own errors. But as we cannot pretend to measure " what we cannot see," the refinement of the circle is really thrown away so long as an op- tical power is used so feeble as that of the tele- scope now employed ; for in fact its results do not surpass those of the common sextant so much as was expected from its theoretical per- fection. 8. The prismatic reflecting and re- peating circles, which differ from the above only in substituting for the horizon glass a glass prism, which is fixed on the line of sight behind instead of in front of the index glass. The advantage of this arrangement is that an- gles of all magnitudes can then be measured. These are the transformations which the sex- tant has undergone to adapt it for the varied requirements of the practical astronoiner and surveyor on land ; but to satisfy the demands of the nautical astronomer and hydrographic surveyor, changes even more radical and varied
