100 Journal oe the Mitchell Society. {Nov. 



fiths would indicate that the divergence at the lower temper- 

 atures is but partly due to the vapor pressure used. 



The divergence in ethyl alcohol at 0° C is probably due to 

 its being the end point of Biot formula used. 



Methyl alcohol, 210° to 238.5° C, ethyl alcohol, 190° to 

 242.5° C, and propyl alcohol, 120° to 260°, evidently should 

 be grouped together. They are associated substances, as was 

 water, and the theory was not expected to hold for associated 

 substances, because the molecules may not be evenly distrib- 

 uted, the molecules are of different kinds, and the number of 

 molecules changes with changing temperature. The fact 

 that the theory does hold, to a very considerable degree, for 

 these substances also, is significant, and points strongly to 

 the conclusion that the cause of the molecular association in 

 these substances must be the attraction which we are discus- 

 sing and not some other attraction such as we might denote 

 by the term chemical affinity. And the fact that these asso- 

 ciated substances do not agree even more closely with the 

 theory may be entirely due to the supposition embraced by 

 the formula that the molecules are uniformly distributed 

 throughout the space occupied by them — a supposition prob- 

 adly untrue for associated snbstances. 



Bearing in mind the comments above, we conclude that the 

 divergences shown by di-isobutyl, chlor-benzene, brom-ben- 

 zene, iodo-benzene, and water, all occur at such points and 

 are of such magnitude (none greater than 5.4 per cent.) that 

 they may easily be due to errors of observation, or the multi- 

 plication of such errors by the calculations. Stannic chloride 

 fails to agree with the theory. Water, methyl, ethyl, and 

 propyl alcohols and acetic acid are associated substances to 

 which the theory is not applicable a priori, and yet these 

 substances, acetic acid excepted, appear within limits to 

 agree with the theory. 



In conclusion, to prevent further misapprehension, I would 

 notice a review by Mr. G. N. Lewis 1 of the former paper on 



1 Jour. Am, Ohem. Soc. 3, 107 (1903). 



