430 PROCEEDINGS OF THE AMERICAN ACADEMY. 



Let us see how this idea bears upon the deviations from the mass law, 

 K Ci = 2 2 . Here K is a quantity which does not vary with the con- 

 centration in a given medium. In different media however, K has 

 different values, as shown by experiment, and as we should also expect 

 from theory, whether the mass law be derived from the principles of 

 probability or from thermodynamics. If therefore with changing con- 

 centration there is an essential change in the character of the medium, 

 K no longer remains a constant but must change as the concentration 

 changes. 



Iu order to give a complete qualitative explanation of the phenomena 

 under consideration we have only to make the assumption that the 

 addition to any solvent of a better or a worse electrophile than itself 

 respectively increases or decreases the electrophilic character of the 

 solvent. 



Thus for example, water and acetic acid are both medium good elec- 

 trophiles. If small quantities of acetic acid are added to water the 

 electrophilic character of the medium is not materially changed and 

 therefore for dilute solutions of acetic acid the mass law holds. 



If on the other hand we add to water even a small quantity of a very 

 strong electrophile like potassium chloride the dissociating power of the 

 medium is increased, K increases with the concentration of potassium 

 chloride and we have the familiar deviation from Ostwald's dilution law. 



An extraordinarily good confirmation of the view that such deviations 

 are due to a change in the dissociation power of the medium is given by 

 Arrhenius, * who showed that the dissociation of weak acids in water is 

 considerably increased by the addition of small quantities of inorganic 

 salts. In other words a tenth or even hundredth normal salt solution is 

 an appreciably better dissociating solvent than pure water, f 



If finally we add to water a substance that is a much poorer electro- 

 phile than water itself we should expect a deviation from the dilution law 

 in the opposite direction. That is, the dissociation constant should de- 

 crease with the concentration. Such a phenomenon has never been 

 observed, nor, perhaps, is it likely to be until our experimental methods 

 are improved, for to discover this phenomenon it would be necessary to 

 investigate accurately the dissociation of extremely poor electrolytes. 



* Zeit. phys. Chem., 31, 197 (1899). 



t It is likewise probable that the dissociation of water itself is considerably in- 

 creased in strong salt solution. An experimental proof of this would be interest- 

 ing and important. 



