§ 14 OF ARTS AND SCIENCES. 4II 



The general truth of the first equation was anticipated 

 in . Miller's Chemical Physics, where, oh page 342, he 

 gives a table of the Latent heat for equal Volumes. 

 The cause of the variation of these latent heats, tosrether 

 with that of their approximate agreement, is now ex- 

 plained, being due to the fact that the coefficient of 

 expansion varies through narrow limits for all ordinary 

 liquids. 



The average value of the product of the internal latent 

 heat by the molecular weight is easily found from Miller's 

 Table to be about 7,700. The average value of the 

 coefficient of voluminal expansion at the boiling point (say 

 80°) is not far from .0012;^ the product is therefore about 

 9.2, and the second term is equal to 5.9 -|- ^t-Zi ^^ 9-2 

 about, which sufficiently establishes the general truth of 

 the formula. 



The second formula has been already discussed; the 

 third and the two followino; are of interest as erivino- the 

 absolute thickness of a molecule. 



We have for water in equation III. 



/=: — - — - — -^ — = about ten thousand-mil- 



575 X 42,000,000 X I 



lionths of a centimetre, instead of fortv-six, according to 

 Riihlmann. Professor Cooke has suggested five thousand- 

 millionths as the most probable number.f 



The formulae for the difference of the internal specific 

 heats of the liquid and vapor are easily shown to agree, as 

 nearly as could be expected, with the observations of 

 Regnault, in which these two specific heats are in no case 

 determined for the same range of temperature. 



The specific heat of ether vapor is, for instance, 0.48, 

 nearly, from which, deducting 0.03 for external work, we 



* See Tables for expansion, Sharpies, pages 68-72. 

 t In the New Chemistry, page 34, fin. 



