HEAT. 53 



cold, both tending to approximate the particles, and such sub- 

 stances solidify at a higher temperature in proportion as the 

 pressure is greater ; so that we might expect a body of this 

 class, which under the ordinary pressure of the air is at a 

 temperature just above its freezing point, to solidify by 

 being submitted to pressure alone, the temperature being 

 kept constant. 



A similar class of exception to the general effect of 

 heat in expanding bodies is presented by vulcanised caout- 

 chouc. This has been observed by Mr. dough, and, in- 

 deed, was pointed out to me many years ago by Mr. 

 Brockedon to be heated when stretched, and cooled when 

 unstretched. 



Mr. Joule finds that its specific gravity is lower when 

 stretched than when unstretched, and that when heated 

 in its stretched state it shortens, presenting in this par- 

 ticular condition a similar series of converse relations to 

 those which are presented by water near or at its freezing 

 point. 



With the exception of this class of phenomena, which 

 offer difficulties to any theory which has been proposed, the 

 general phenomena of heat may, I believe, be explained upon 

 a purely dynamical view, and more satisfactorily than by 

 having recourse to the hypothesis of latent matter. Many, 

 however, of the phenomena of heat are involved in much 

 mystery, particularly those connected with specific heat or 

 that relative proportion of heat which equal weights of differ- 

 ent bodies require to raise them from a given temperature to 

 another given temperature, which appear to depend in some 

 way hitherto inexplicable upon the molecular constitution of 

 different bodies. 



The view of heat which I have taken, viz. to regard it 

 simply as a communicable molecular repulsive force, is sup- 

 ported by many of the phenomena to which the term specific 

 or relative heat is applied ; for example, bodies as they in- 



