748 Di\ S. R. Milner on the Effect of Interionic 



to determine the osmoti 

 (6) would be satisfied by 



to determine the osmotic pressure P. If -^ were zero, 



2RT 

 r ~ Y 3 



the gas law for a completely dissociated electrolyte. When 



L_ is not zero, we may therefore take P to be of the form 

 dV J 



P == 2|T + 0(V,T), « . . . . (7) 

 and determine the function c/>. Substituting (7) in (6) gives 



whence 



dT ^"~~dN~ dV 



J T d dT 



there being no constant of integration since <f> vanishes with Ji. 

 This expression can always be integrated when, as is here 



the case *, A is cf the form F(V)/T, for take the -j^- outside 



dh dT ® * 



the f , and substitute j for -pn, when we get 



* Since K the s.i.C. of water, varies to some- extent with the 

 temperature, h is not strictly proportional to 1/T. If, however, we 

 were to try to take into account here the variation of K with T, it 

 would lead to a discrepancy between the expressions for the pressure 

 obtained by the kinetic and by the thermodynamical calculations. Such 

 a discrepancy cannot, however, really exist, for the virial equation is 

 consistent with thermodynamics. The fact is that K has already 

 virtually been assumed constant in determining the virial (if it were not 

 constant the electrical force between two ions would depend on their 

 kinetic energies as well as on their positions), and to obtain agreement 

 the game assumption must be made here also. 



