M. Arago's Biographical Memoir of James Watt. 245 



Conceive a wide vertical cylinder open above, and repos- 

 ing at its base upon a metal table pierced with a hole which 

 a stopcock can at will shut and open. Let us now introduce 

 into this cylinder a piston, that is to say a large and moveable 

 circular plate, which accurately fits it. The atmosphere will 

 press with all its weight upon the upper side of this kind of 

 diaphragm, and will tend to push it from above downwards. 

 That part of the atmosphere again which fills the lower part of 

 the cylinder, wUl tend, by its reaction, to produce the inverse 

 movement. This second foi'ce will be equal to the former if 

 the stopcock be open, for gases press equally in all directions. 

 The piston will thus find itself operated on by two opposing 

 forces, which will produce an equilibrium. It will neverthe- 

 less descend, but only in virtue of its own gravity. A slight 

 counterpoise somewhat heavier than the piston, will suffice to 

 draw it contrariwise to the top of the cylinder, and to keep it 

 there. Suppose the piston arrived at this point, we have 

 now to seek for the means of making it forcibly descend, and 

 then ascend again. 



Suppose that, after having shut the lower stopcock, we 

 should succeed in annihilating suddenly all the au' contained 

 in the cjlinder, — in a word, in making a vacuum ; the piston 

 receiving only the action of the external air, pressing from 

 above, would rapidly descend. This movement accomplished, 

 we might then open the stopcock, the air would then enter 

 from beneath, and would counterbalance the pressm-e of the 

 atmosphere above the piston. As at the commencement, the 

 counterpoise Avould now raise the piston to the top of the 

 cylinder, and every portion of the appai-atus would be found 

 in its original state. A second evacuation, or we may call it 

 abstraction of the internal air, would make the piston again de- 

 scend, and so on successively. The true moving power of this 

 machine would here be the weight of the atmosphere. And 

 let no one suppose that because we walk and even run with 

 facility through the an-, the atmosphere must therefore be very 

 feeble as a moving power. With a cylinder of two metres 

 in diameter the pressm-e of the piston in descending — the 

 weight it might raise throughout the whole height of the cylin- 



