﻿■ M the 



Molecular Thermodynamics. 615 



terms is different in the two expressions for ^- 

 difference being the quantity 



G s -G, ' = f2*, l rf[G. 1 '--G,/-M, l G M ']. . (22) 

 Jc=o 



The form (18) will probably be the more generally useful, 

 possessing the advantage that the " linear " and " general " 

 terms separately are connected by the Gibbs fundamental 

 relation with those (respectively) in the simple form 



of** 

 SMo" 



afc-St" [**+*.$ i*w+™>}+<v. (23) 



c=o 

 For since (cf. (3), (4), and previous paper) 



M dGK' + %n,(l-X,)dG;J + $Sn s a: Sl dG Sx '= . (24) 



and 



M = M '-Zn s * s , (25) 



clearly 



M ^Gm' + ^/ l (1-X,)^G So ' + S^ 1 ^G,/-«^Gm']=0, (26) 



i.e. Mo'dG M , + Xn s dG s = 0; ... (27) 



so that, if G is a function of Ci,c 2 ', ("experimental" 



concentrations), such that 



M 'G = M 'Gm'4 2»,G. s , .... (28) 

 we have 



a, -(»#)'•-!» (29) 



V 0"s fyy OC s 



G -'=( W)„ [" G * (Bay)]=«-2c;|«. (30) 



Now G / is a function of e±, c 2 , and also of the various 



fractions x\ v #i a , , ar #l , x St1 If we take the quantity 



yr-, G f , and eliminate the " true " concentrations (C, c s ) by 



* This common usage of the suffix outside the bracket was mentioned 

 in the previous paper, n here indicates constancy of all the n's and x 

 (later) constancy of all the x's. 



