appearance; that the bell crank arms, the heart cams, the weaver's
shuttles, the spring "take ups," rectangular needle bars, and gear wheels,
have developed into very different devices for performing the various
functions of those several parts.

The shuttle is perhaps the most important part of a lock stitch machine.
But what is a shuttle? So many devices for performing the functions of the
early weaver's shuttle have been introduced of late, that the word shuttle,
if it be used at all, must not be accepted as meaning "to shoot." We have
vibrating shuttles, which are, strictly speaking, the only surviving
representatives of the weaver's shuttle in these new orders of machines;
and stationary shuttles, oscillating shuttles, and revolving shuttles,
besides the earlier rotating hook, in several new forms, difficult to name.
But the general acceptation of the word shuttle, as indicating those
devices that pass bodily through the loop of upper thread, is, I venture to
think, sufficiently correct.

Many changes have been effected in the form, size, and movements of the
shuttle, and we may profitably inquire into the causes that have induced
manufacturers to abandon the earlier forms. The long, weaver's kind of
shuttle, originally used by Howe and Singer, had many drawbacks. Mr. A.B.
Wilson's ingenious device, the lock stitch rotating hook, was not free from
corresponding faults. The removal of these in both has led to the adoption
of an entirely new class of both shuttles and revolving hooks. It is well
known that the lock stitch is formed by the crossing of two threads, one of
which lies over, and the other under, the cloth to be sewn. This crossing
point, to insure integrity of the stitch, must occur as nearly as possible
in the middle of the thickness of the fabric. The crossing must also be
effected while a certain strain, called tension, is imposed upon both
threads. If the tension of one thread should outweigh that of the other,
the locking point becomes displaced. If the tension be insignificant, the
stitches will be loose. If the tension should vary, as in the long shuttle,
there will occur faulty points in the seam.

In the earlier rotating hook the tension depended upon the friction
developed between the spool and the hook. This tension, therefore, varied
in proportion to the speed of the latter, and could never be constant. This
was quite apart from the frictional resistance offered to the upper thread
in passin