d later to
England, where he died in 1875.

[Illustration: Figure 35.--THIS GROVER AND BAKER CABINET-STYLE SEWING
MACHINE of 1856 bears the serial number 5675 and the patent dates
February 11, 1851, June 22, 1852, February 22, 1853, and May 27, 1856.
(Smithsonian photo 45572-F.)]

By the mid-1850s the basic elements of a successful, practical sewing
machine were at hand, but the continuing court litigation over rival
patent rights seemed destined to ruin the economics of the new industry.
It was then that the lawyer of the Grover and Baker company, another
sewing-machine manufacturer of the early 1850s, supplied the solution.
Grover and Baker were manufacturing a machine that was mechanically
good, for this early period. William O. Grover was another Boston
tailor, who, unlike many others, was convinced that the sewing machine
was going to revolutionize his chosen trade. Although the sewing
machines that he had seen were not very practical, he began in 1849 to
experiment with an idea based on a new kind of stitch. His design was
for a machine that would take both its threads from spools and eliminate
the need to wind one thread upon a bobbin. After much experimenting, he
proved that it was possible to make a seam by interlocking two threads
in a succession of slipknots, but he found that building a machine to do
this was a much more difficult task. It is quite surprising that while
he was working on this idea, he did not stumble upon a good method to
produce the single-thread (as opposed to Grover and Baker's two-thread)
chainstitch, later worked out by another. Grover was working so intently
on the use of two threads that apparently no thought of forming a stitch
with one thread had a chance to develop.

At this time Grover became a partner with another Boston tailor,
William E. Baker, and on February 11, 1851, they were issued U.S. patent
No. 7,931 for a machine that did exactly what Grover had set out to do;
it made a double chainstitch with two threads both carried on ordinary
thread spools. The machine (figs. 34 and 35) used a vertical eye-pointed
needle for the top thread and a horizontal needle for the underthread.
The cloth was placed on the horizontal platform or table, which had a
hole for the entry of the vertical needle. When this needle passed
through the cloth, it formed a loop on the underside. The horizontal
needle passed through this loop forming another loop beyond, which was
retained until the rede