﻿618 Mr. Bernard Oavanagh on 



and since 



whereas I dx n will be only some fraction — generally, as 



we shall see, not large — ©f x Sl , a simple graphical considera- 

 tion will suffice to show that (41) will generally be much 

 greater than (40). 



The practical significance and value of the quantitative 

 statement (39) is further illustrated in Appendix II. 



The analogy between the modification of the " general " 

 terms, expressed in (39), and that found necessary in the 

 previous treatment of "complex solvents" is plain. In 

 the latter we kept 



Gs = G/, 



whereas in the present case we have kept 



Gm = Ctm'j 



but, of course, since we started in the present case from the 

 results of the previous paper (and G 7 in this paper corre- 

 sponds to G in the previous paper), the two modifications of 

 the " general " terms in the " molecular " expression are, as 

 it were, superimposed. 



More detailed treatment than the above of the " general " 

 terms will be possible only when something is known about 

 their form (see, for instance, Appendix II.). 



For the evaluation of the linear terms we observe that 

 u s must be expressed as a function of the concentrations of 

 the solutes. 



An Expansion of ol s . 



Consider, first, a solute which forms only one solvate. In 

 this case we can write x s , M s , etc , instead of x sv M Sl , etc.,, 

 and, moreover, we have 



ot s =M S X S ^\ 



X s =x s I ..... (43) 



and equation (21) assumes the form, 



R^l g(l_^) = ^^p t ^Vlog(l+mC)-^G x / / ], (44) 



