﻿620 Mr. Bernard Cavanagh on 



We have then a* as a practical function of y, which is a 

 function of the concentrations of the solutes (see (44)), and 

 if we know the form of Gx s " we can readily complete the 



evaluation of \^d log (1 + ?77C) . 

 c=o 

 First, let us assume that ~ G Xs " can be neglected and write 



y = M s jirflog(l + mC). 



C^O 



From the previous paper we can then write, 



y=M s C[l-ia 1 C + la 2 C 2 -ia 3 C 3 ], 

 and from (43) and (47) we then get 



" S =« 50 -+&C + 7,C 2 + S S 3 . . . . . (48) 

 Where 



u So = M s ^ , y s = (b 2 M s * - iaAMJ*) , 



ft = M s 2 ^, 8. = (6 3 M S 4 - ^W + J a AM S 2 ) 



(49) 



which may be expected to hold up to a value of C rather 



higher than ^- (since y is considerably less then M S C at 



high concentrations), that is, in aqueous solution for example, 

 up to a total concentration (" true ") of at least twelve-molar 

 (12M.) when the solvate is a pentahydrate, or six-molar 

 (6M.) when it is a decahydrate. 



ryt 



When ~ Gr Xs " is not negligible the above will cease to hold 



x 

 exactly, but it may be possible to express ttQx s " (with a 



negligible residual error) as a function of C, the total solute 

 concentration, and if this function can approximately take 

 the form of a short series of integral powers we shall get a 

 result in the above form, (48), but with departures in the 

 values of the coefficients <y s and S s . 



When the solute considered forms several solvates, the 

 problem of obtaining a s as a function of the concentrations 

 is more complex. Equation (21) takes the form 



d log (1 -X s ) =Xx Sl [1* d log (1 -f roC) - g ^/'J ; (50) 



but we now have to use instead the series of equations (6) 



