392 JAMES WATT. 



which is open at the top, in proportion as it expands 

 itself over their entire surface. This cooling is not com- 

 pensated during the whole ascension of the piston, with- 

 out the expense of a certain quantity of steam. But 

 there is no loss of this sort in the engines modified by 

 Watt. The atmospheric action is totally eliminated by 

 the following means : 



The top of the cylinder is closed by a metal cover, 

 only pierced in the centre by a hole furnished with 

 greased tow stuffed in hard, but through which the rod of 

 the piston has free motion, though without allowing free 

 passage either to air or steam. The piston thus divides 

 the capacity of the cylinder into two distinct and well- 

 closed areas. When it has to descend, the steam from 

 the caldron reaches freely the upper area through a tube 

 conveniently placed, and pushes it from top to bottom as 

 the atmosphere did in Newcomen's engine. There is no 

 obstacle to this motion, because whilst it is going on, only 

 the base of the cylinder is in communication with the 

 condenser, wherein all the steam from that lower area 

 resumes its fluid state. As soon as the piston has quite 

 reached the bottom, the mere turning of a tap suffices to 

 bring the two areas of the cylinder, situated above and 

 below the piston, into communication with each other, 

 so that both shall be filled with steam of the same degree 

 of elasticity, and the piston being thus equally acted upon, 

 upwards and downwards, ascends again to the top of the 

 cylinder, as in Newcomen's atmospheric engine, merely 

 by the action of a slight counterpoise. 



Pursuing his researches on the means of economizing 

 steam, Watt also reduced the result of the refrigeration 

 of the external surface of the cylinder containing the pis- 

 ton, almost to nothing. With* this view he enclosed the 



