﻿612 Mr. Bernard Oavanagh on 



our results in the first place in terms of the "true con- 

 centrations," introducing the " experimental '" concentrations 

 afterwards. 



Of the total number n s o£ gm .-molecules of the solute, (s), 



put into the solution, let the fractions x Sl , x So1 , go to 



form solvates containing respectively M #1 , M S2 , , grams 



of solvent per gram-molecule, and let us write X s for %a Sl , 

 and a s for "S# Sl M Sl ; then clearly (from previous paper) our 

 " original " expression for yjr is 



f = M [<£m + R f Ld log (1 4- mC)] 



~c=o 



+ 2n s {(l-X 8 )U So -Rlog|^-^l 



m 

 + 2* s , [k- R log -^~] } + M G', (3) 



m 

 and is of the form 



the independent variables being # M , n s (l — X s ), n s x Siy 



n s x S2 , 



The "experimental*' expression for ty, now required, 

 must have the form 



tV^&tavg.- •. • • w 



the independent variables being M ', n s , 



In changing the variables we make use of the equations 

 given by the chemical equilibrium governing the various 

 stages of solvation, viz. : 



.M _ Ml _ m jW - o '* " 



3*_ <W „m M_n I ' * (6) 



~dn S2 ^n S0 iVJs2 dM - U ' j 

 etc., 

 whence, of course, 



y„ -M. y ^ * ^ _n m 



O^*! 0^*0 oM-o 



