KAULENLERG — THEORY OF ELECTROLYTIC DISSOCIATION'. 349 



sion of a gas, — the analogy is at once apparent. In seeing it 

 with the mind's eve one abstracts from the water and centers the 

 attention entirely on the sugar. But it must be remembered 

 that a gas will expand readily in vacuo, and that it will mix with 

 any other gas or mixture of gases, while on the other hand, the 

 lump of sugar will not dissolve when, for instance, benzene or 

 absolute alcohol is poured over it. The process of solution of 

 a substance and the expansion of a gas, then, while possessing 

 analogy, are in reality very different processes. And right here 

 lies the difficulty with the theory of solutions. It neglects the 

 all important role of the solvent. It fails to emphasize the fact 

 that the process of solution takes place because of a mutual at- 

 traction between solvent and dissolved substance, and that this 

 mutual attraction, which is a function of the chemical nature of 

 both solvent and dissolved substance, is the essence of the so- 

 called osmotic pressure. 1 It is true that the thermodynamic 

 considerations of van't Horr will hold whether the osmotic pres- 

 sure be considered as the outcome of a mutual attraction of solv- 

 ent and dissolved substance, or as resulting from the bombard- 

 ment of the molecules of the dissolved substance against the 

 semipermeable membrane. But if we choose to use the gas equa- 

 tion in working with solutions it is very evident that the factor 

 i must never be placed equal to unity (unless direct experi- 

 mental evidence justifying this step is at hand) if exact results 

 are desired, no matter whether we work with electrolytes or non- 

 electrolytes. The dissociation theory, as has been shown, does 

 not furnish a satisfactory explanation of the significance of i 

 in the case of electrolytes. This value of i varies in all cases 

 with the nature of the solvent and also, in general, with the 

 strength of the solution ; and it does not always vary to the same 

 extent nor even in the same sense in different solutions. It is 

 clear then, that when the simple equation PV=RT is applied to 

 a solution, only approximate results are obtained at best, unless 

 experimental data are at hand showing that in the particular 



1 Compare here the warning words of Lothar Meyer, in his article,— Das Wesen 

 des osmotischen Druckes. Zeit. phys. Chem. 5, 23 (1890). Also the reply of van't 

 Hoff, IMd., p. 174. 



