ee re 
F ATMOSPHERIC ENGINES. 381 
selves, therefore, exclusively with the means by the aid 
of which Papin proposed to induce his oscillating motion. 
Let us imagine a large vertical cylinder, open at top, 
and its base resting on a metallic table, pierced with a 
hole that a cock can either close or open at will. 
Introduce a piston into this cylinder, that is to say, a 
circular plate, filling it entirely but movable, that shall 
exactly close it. The atmosphere will rest with all its 
_weight on the upper surface of this species of diaphragm, 
and will push it from the top to the bottom. The portion 
of atmosphere that occupies the lower part of the cylin- 
der will tend, by reaction, to produce an inverse move- 
ment. This second force will be equal to the first if the 
tap is open, because gas presses equally in all directions. 
The piston will thus find itself urged by two opposite 
forces, which will equalize each other. It will descend, 
however, though only by reason of its gravity. A coun- 
terpoise, slightly heavier than the piston, will suffice to 
raise it, on the contrary, up to the summit of the eylin- 
der, and keep it there. Suppose the piston to have 
reached this extreme position. Let us seek the means to 
make it descend from thence with great force, and carry 
it up again. 
Let us imagine after having shut the lower cock, we 
succeed in suddenly annihilating all the air contained in 
the cylinder,—in a word, to render it a vacuum. A 
vacuum once made, the piston not receiving any pressure 
but from the external atmosphere which presses it from 
above, will descend rapidly. On this movement being 
achieved, the cock will be opened. The air will thereby 
return underneath to counterbalance the upper atmos- 
phere. As at the beginning, the counterpoise will make 
the piston remount to the cylinder, and all the various 
