194 CORRELATION OF PHYSICAL FORCES. 



er, of the cylinder and all surrounding bodies. The force ex- 

 pended in expanding this iron to a very small extent is equal 

 to that which expands the vapour to a very large extent : this 

 expansion of the iron is capable, in its turn, of producing a 

 great mechanical force, which is practically lost. Could all 

 the force be applied to the vapour, an enormous addition of 

 power would be gained for the same expenditure : and per- 

 haps even with our present means more might be done in 

 utilising the expansion of the iron. 



Another great difficulty in experimentally ascertaining the 

 dynamic equivalents of different forces arises from the effects 

 of disruption, or the overcoming an existing force. Thus, 

 when a part of the initial force employed is engaged in twist- 

 ing or tearing asunder matter previously held together by 

 cohesive attraction, or in overcoming gravitation or inertia, 

 the same amount of heat or electricity would not be evolved 

 as if such obstacle were non-existent, and the initial force 

 were wholly employed in producing, not in opposing. There 

 is a difficulty apparently extreme in devising experiments 

 in which some portion of the force is not so employed. 



The initial force, however, that has been employed for such 

 disruption is not lost, as at the moment of disruption the 

 bodies producing it fly off, and carry with them their force. 

 Thus, let two weights be attached to a cord placed across a 

 bar ; when their force is sufficient to break the cord or the 

 bar, the weights fall down and strike the earth, making it 

 vibrate, and so conveying away or continuing the force ex- 

 pressed by the cohesion of the bar or cord. If, instead of 

 breaking a cord, the weights be employed to bend a bar, their 

 gravitating force, instead of making the earth vibrate, pro- 

 duces heat in the bar, and so with whatever other force be 

 employed to produce effects of disruption, torsion, &c., so 

 that, though difficult in practice, the numerical problem of 

 the equivalent of the force is not theoretically irresolvable. 



The voltaic battery affords us the best means of ascertain- 



