HEAT. 49 



power of reaction of the compressed against the dilated air. 

 If the heat due to compression be made to perform mechani- 

 cal work, it would pro tanto be consumed, and could not 

 restore the temperature to the dilated air ; but if it perform 

 no work, no heat is lost. Mr. Joule has experimentally 

 proved this proposition. 



In commencing the subject of heat, I asked my reader to 

 put out of consideration the sensations which heat produces 

 in our own bodies ; I did this because these sensations are 

 likely to deceive, and have deceived many as to the nature of 

 heat. These sensations are themselves occasioned by simi- 

 lar expansions to those which we have been considering ; the 

 liquids of the body are expanded, i. e. rendered less viscid by 

 heat, and from their more ready flow, we obtain the sensation 

 of agreeable warmth. By a greater degree of heat, their ex- 

 pansion becomes too great, giving rise to a sense of pain, and, 

 if pushed to extremity, as with the heat which produces a 

 burn, the liquids of the body are dissipated in vapour, and an 

 injury or destruction of the organic structure takes place. A 

 similar though converse effect may be produced by intense 

 cold ; the application of frozen mercury to the animal body 

 produces a burn similar to that produced by great heat, and 

 accompanied with a similar sensation. 



Doubtless other actions than those above mentioned inter- 

 fere in producing the sensations of heat and cold ; but I think 

 it will be seen that these will not affect the arguments as to 

 the nature of heat. The phenomenal effects will be found 

 unaltered : heat will still be found to be expansion, cold to be 

 contraction ; and the expansion and contraction are, as with 

 the two bladders of air, correlative i. e. we cannot expand 

 one body, A, without contracting some other body, B ; we 

 cannot contract A without expanding B, assuming that we view 

 the bodies with relation to heat alone, and suppose no other 

 force to be manifested. 



I have said that there are few exceptions as to heat being 

 3 



