428 PROCEEDINGS OP THE AMERICAN ACADEMY. 



and 3, especially the former, which embraces the region of dilute solu- 

 tions. Here we may regard M as proportional to the molecular con- 

 ductivity and, assuming the mobility of the ions to be constant, to the 

 degree of dissociation. The latter quantity, instead of rising with in- 

 creasing dilution and approaching a constant maximum as in the case of 

 water solution, has a maximum in the neighborhood of C = 5, and from 

 that point decreases rapidly with increasing dilution. 



This behavior has been previously noticed by a number of observers in 

 the case of certain organic solvents and other solvents of small dissociat- 

 ing power. It is our belief that this is an extreme manifestation of a 

 phenomenon which is common to all solutions and which has been 

 frequently observed in aqueous solutions. 



M 



i'luuiiis 3. 



The mass law applied to the dissociation of a binary electrolyte gives 

 the equation K C x = Co 2 where 1 is the concentration of the undis- 

 sociated substance and C 2 that of one of the ions. It is well known, 

 however, that this equation is not satisfied in the case of all water solu- 

 tions. Bancroft * has shown that the empirical equation K C^ = C 2 ", 

 where n is a specific constant depending on the nature of the electrolyte, 

 expresses satisfactorily the behavior of any aqueous solution. This 

 quantity n varies from 2 in the case of weak acids and bases down to 

 1.36 in the case of potassium chloride. If now we could find some still 

 "stronger" electrolyte than potassium chloride, the value of n would 

 doubtless be still smaller. If a substance could be found with a value of 

 n less than 1 the degree of dissociation and consequently the molecular 

 conductivity would increase with the concentration as they do in the case 

 we have been studying. 



* Zeit. phys. Chem., 31, 188 (1899). 



