HEAT. 53 



those of ordinary chemical decomposition where the affinity of 

 other substances is brought into play. 



In considering the effect of heat as a mechanical force, it 

 would be expected, a priori, and independently of any theory 

 of heat which may be adopted, that a given amount of heat 

 acting on a given material must produce a given amount of 

 motive power ; and the next question which occurs to the 

 mind is, whether the same amount of heat would produce the 

 same amount of mechanical power, whatever be the material 

 acted on or affected by the heat. I will endeavour to reason 

 this out on the view of heat which I have advocated. Heat 

 has been considered in this essay as itself motion or mechanical 

 power, and quantity of heat as measured by motion. Thus, if 

 by a given contraction of a body (say hot mercury) air within 

 a cylinder having a movable piston be expanded, the piston 

 moves, and in this case the expansion or motion of the material 

 (say iron) of the cylinder itself and of the air surrounding it is 

 commonly neglected. As the air dilates it becomes colder ; 

 in other words, by undergoing expansion itself it loses its 

 power of making neighbouring bodies expand ; but if the 

 piston be forcibly kept down, the expansive power due to 

 the mercury continues to communicate itself to the iron and to 

 the surrounding air, which become hotter than they would 

 if the piston had given way. 



Now, in the above case, if the air be confined and its 

 volume unchanged, will the expansion of the iron, assuming 

 that it can be utilised, produce an exactly equivalent mecha- 

 nical effect to that which the expansion of the air would pro- 

 duce if the heat be entirely confined to it ? 



Assuming that (with the exception of bodies which expand 

 in freezing, where, through a limited range of temperature, the 

 converse effects obtain) whenever a body is compressed it is 

 heated, i.e. it expands neighbouring substances ; whenever it 

 is dilated or increased in volume it is cooled, i.e. it contracts 

 neighbouring substances the conclusion appears to me in- 

 evitable that the mechanical power produced by heat will be 

 definite, or the same for a given amount and intensity of heat, 

 whatever be the substance acted on. 



Thus, let A be a definite source of heat, say a pound of 



