28 



CLAUSIUS ON THE MECHANICAL EQUIVALENT 



of both batteries, that is WjS and n<^s, be denoted by Sj and S5 

 (38) then passes into — 



'^-Mtm- 



and instead of this. Dove gives the formula 



C=:A. 



^Si . S, 



(39) 



(40) 



The results of his experiments agree well with this formula, but 

 the later results of Riess, on the contrary, agree better with 

 mine, as is shown by the table below. 



By both observers n^ was first preserved constant, and n^ so 

 varied that 



Wj = W2, =2?Z2, =3w2 and ^An^ 



successively ; in the next place Wj was preserved constant, and 

 Wg so varied that 



Wgss^ij, =2Wi, =3;ii and =4Wi 



successively. In order better to compare the results, I have 

 taken as unit in both cases the heat observed in the first expe- 

 riment, where ^^ = ^2, and reduced the other quantities of heat 

 generated to this. In the case of Riess, who gives two values 

 for each experiment, I have taken the mean of both. 



I. ^1 variable, n^ constant. 



Pogg. Atm. vol. Ixxii. p. 419. 



