450 MEASUREMENT OF ELECTROMOTIVE FORCES. 



But Berthelot has shown* that Wcestyn's law only holds through- 

 out a series of transformations as long as they do not bring about 

 changes of state. In the contrary case, the heat of combination is 

 moreover changeable with the temperature. It is just those ele- 

 ments with a solid depolarizer, such as mercuric sulphate, mer- 

 curous chloride, and argentic chloride, which are most sensitive 

 to changes of temperature, and which diverge most widely from 

 Thomson's law. The two orders of phenomena are therefore 

 connected : if the law of thermal capacities is verified, the chemi- 

 cal heat and the electromotive force are equal to each other, and 

 are independent of the temperature ; in the opposite case, they are 

 unequal, and vary with *the temperature. 



Numerous experiments have been made by Crapskif and Gockel| 

 to verify the conclusions from Von Helmholtz's theory. The values 



of E and of have been measured for a great number of elements. 



DE 

 If the values obtained for the expression T be compared with the 



difference between the chemical energy and the electromotive force, 



it is found that the verifications always succeed as regards the 



t 

 direction of the phenomenon, but not when we come to compare 



BE 



the numbers. The quantity of heat represented by T is never 



more than a fraction of the positive or negative difference between 

 the chemical energy and the electromotive force. It is scarcely 

 possible, in most cases, to ascribe these divergences to experimental 

 errors; and these are some obscure points which the theory must 

 clear up. 



* BERTHELOT. Essai de Mecanique Chimique, Vol. I., p. no. 1879. 

 f CRAPSKI. Wiedemann's Annalen, Vol. xxi., p. 209. 1884. 

 J GOCKEL. Wiedemann's Annalen, Vol. xxiv., p. 612. 1885. 



