motion adaptable for all manner of work. The ingenious substitutes Watt
had to invent to avoid the obviously perfect crank motion have of course
all been discarded, and nothing of these remains except as proofs,
where none are needed, that genius has powers in reserve for
emergencies; balked in one direction, it hews out another path for
itself.

While preparing the specification for this patent of 1781, Watt was busy
upon another specification quite as important, which appeared in the
following year, 1782. It embraced the following new improvements, the
winnowing of numberless ideas and experiments that he had conceived and
tested for some years previous:

1. The use of steam on the expansive principle; together with
various methods or contrivances (six in number, some of them
comprising various modifications), for equalising the expansive
power.

2. The double-acting engine; in which steam is admitted to press
the piston upward as well as downward; the piston being also
aided in its ascent as well as in its descent by a vacuum
produced by condensation on the other side.

3. The double-engine; consisting of two engines, primary and
secondary, of which the steam-vessels and condensers communicate
by pipes and valves, so that they can be worked either
independently or in concert; and make their strokes either
alternately or both together, as may be required.

4. The employment of a toothed rack and sector, instead of
chains, for guiding the piston-rod.

5. A rotative engine, or steam-wheel.

Here we have three of the vital elements required toward the completion
of the work: first, steam used expansively; second, the double-acting
engine. It will be remembered that Watt's first engines only took in
steam at the bottom of the cylinder, as Newcomen's did, but with this
difference: Watt used the steam to perform work which Newcomen could not
do, the latter only using steam to force the piston itself upward. Now
came Watt's great step forward. Having a cylinder closed at the top,
while the Newcomen cylinder remained open, it was as easy to admit steam
at the top to press the piston down as to admit it at the bottom to
press the piston up; also as easy to apply his condenser to the steam
above as below, at the moment a vacuum was needed. All this was
ingeniously provided for by numerous devices and covered by the patent.
Third, he went one step farther