338 ON THE CONSERVATION OF FORCE. 



ment, the apparatus may be arranged more simply. At 

 C is a glass globe rilled with dry air, which is placed in 

 a metal vessel, in which it can be heated by steam. It is 

 connected with the U-shaped tube, s s, which contains a 

 liquid, and the limbs of which communicate with each 

 other when the stop-cock R is closed. If the liquid is in 

 equilibrium in the tube ss when the globe is cold, it 

 rises in the leg s, and ultimately overflows when the 

 globe is heated. If, on the contrary, when the globe is 

 heated, equilibrium be restored by allowing some of the 

 liquid to flow out at K, as the globe cools it will be drawn 

 up towards n. In both cases liquid is raised, and work 

 thereby produced. 



The same experiment is continuously repeated on the 

 largest scale in steam engines, though in order to keep 

 up a continual disengagement of compressed gases from 

 the boiler, the air in the globe in Fig. 44, which would 

 soon reach the maximum of its expansion, is replaced by 

 water, which is gradually changed into steam by the 

 application of heat. But steam, so long as it remains 

 as such, is an elastic gas which endeavours to expand 

 exactly like atmospheric air. And instead of the column 

 of liquid which was raised in our last experiment, the 

 machine is caused to drive a solid piston which imparts 

 its motion to other parts of the machine. Fig. 45 re- 

 presents a front view of the working parts of a high 

 pressure engine, and Fig. 46 a section. The boiler in 

 which steam is generated is not represented ; the steam 

 passes through the tube z z, Fig. 46, to the cylinder A A, 

 in which moves a tightly fitting piston c. The parts 

 between the tube z z and the cylinder A A, that is the 

 slide valve in the valve-chest K K, and the two tubes d 

 and e allow the steam to pass first below and then above 

 the piston, while at the same time the steam has free 

 exit from the other half of the cylinder. When the 



