KAHLENBEBG — THEORY OF ELECTROLYTIC DI8S0CIATION. 341 



in which there are no grounds whatever for assuming the 

 presence of ions, are nevertheless exceedingly reactive chem- 

 ically, — take for instance, many well known explosives as a 

 striking example. And on the other hand it is a well recognized 

 fact that, in many cases at least, salts, acids and bases unite 

 chemically with water and other solvents, and that any re- 

 action which then takes place in solution is one between these 

 new products, and that therefore such reactions might easily 

 occur in presence of water, whereas the anhydrous substances 

 might not react at all. It might be helpful to some to use the lan- 

 guage of the dissociation theory and to say, for example, — the 

 analytical test for chlorine ions is silver ions; instead of saying, 

 — the analytical test for chlorine in the form of a soluble chlor- 

 ide is silver nitrate or any other soluble silver salt in which silver 

 is the base ; but that the new terminology, coupled with an at- 

 tempt to apply the law of mass action to electrolytes in a way 

 in which it certainly does not apply, forms a scientific founda- 

 tion for analytical chemistry, is a position that is clearly un- 

 tenable. The very fact that analytical chemistry has not 

 received much benefit from Professor Ostwald's little book on 

 "The Scientific Foundations of Analytical Chemistry," in the 

 way of improving existing analytical methods and discovering 

 new ones, speaks for itself. 



In connection with the dissociation hypothesis the solvent has 

 been considered as having a peculiar "dissociating power". 

 Water, then, yielding such excellent conducting solutions as it 

 does, would have a very high degree of this dissociating power. 

 The effort has been made by Nernst and by J. J. Thomson, prac- 

 tically simultaneously, to ascribe this strong dissociating power 

 of water to its high dielectric constant, which is about 80 at 

 room temperatures. At the time when this was done, relatively 

 few solvents had been investigated that yielded solutions having 

 considerable conductivity, and so it was perfectly easy to show 

 that in the case of water, formic acid, alcohol, ether and benzol, 

 for example, the conductivity of the solutions in which these 

 liquids are the solvents diminishes with the dielectric constant 



