ended for publication.
Young Harrison now proceeded to copy them out, together with the
diagrams. Sometimes, indeed, he spent the greater part of the night in
writing or drawing.

As part of his business, he undertook to survey land, and to repair
clocks and watches, besides carrying on his trade of a carpenter. He
soon obtained a considerable knowledge of what had been done in clocks
and watches, and was able to do not only what the best professional
workers had done, but to strike out entirely new lights in the clock
and watch-making business. He found out a method of diminishing
friction by adding a joint to the pallets of the pendulum, whereby they
were made to work in the nature of rollers of a large radius, without
any sliding, as usual, upon the teeth of the wheel. He constructed a
clock on the recoiling principle, which went perfectly, and never lost
a minute within fourteen years. Sir Edmund Denison Beckett says that
he invented this method in order to save himself the trouble of going
so frequently to oil the escapement of a turret clock, of which he had
charge; though there were other influences at work besides this.

But his most important invention, at this early period of his life, was
his compensation pendulum. Every one knows that metals expand with
heat and contract by cold. The pendulum of the clock therefore
expanded in summer and contracted in winter, thereby interfering with
the regular going of the clock. Huygens had by his cylindrical checks
removed the great irregularity arising from the unequal lengths of the
oscillations; but the pendulum was affected by the tossing of a ship at
sea, and was also subject to a variation in weight, depending on the
parallel of latitude. Graham, the well-known clock-maker, invented the
mercurial compensation pendulum, consisting of a glass or iron jar
filled with quicksilver and fixed to the end of the pendulum rod. When
the rod was lengthened by heat, the quicksilver and the jar which
contained it were simultaneously expanded and elevated, and the centre
of oscillation was thus continued at the same distance from the point
of suspension.

But the difficulty, to a certain extent, remained unconquered until
Harrison took the matter in hand. He observed that all rods of metal
do not alter their lengths equally by heat, or, on the contrary, become
shorter by cold, but some more sensibly than others. After innumerable
experiments Harrison at length comp