342 BULLETIN OF THE UNIVERSITY OP WISCONSIN. 



of the solvent. 1 The idea underlying the attempt to bring this 

 hypothetical dissociating power into correlation with the dielec- 

 tric constant of the solvent, is that a high dielectric constant of 

 the latter would make it more difficult for the electrically 

 charged ions to neutralize their charges by reason of the electro- 

 static attraction existing between them. It is not claimed that 

 high dissociating power is proportional to the specific inductive 

 capacity of the solvent, but simply that it increases and dimin- 

 ishes with it, the exact mathematical relation being as yet un- 

 known. Since the j^ernst-Thomson rule has been put forth the 

 electrical conductivity of various salts in a goodly number of 

 solvents has been investigated. 2 While the dielectric constants 

 of all the solvents tested were not known at the time, so that a 

 comparison could be made in all cases, in the majority of in- 

 stances where the dielectric constants were known, the Nernst- 

 Thomson rule was indeed corroborated. 2 At the same time a 

 few striking exceptions were present. The relatively low dielec- 

 tric constant of liquid ammonia 3 and the high conductivity 14 of 

 solutions of salts in it, speak powerfully against the Kernst- 

 Thomson rule. It can hardly be argued that this high conduc- 

 tivity in liquid ammonia is due largely to the high rate with 

 which the ions move, because the conductivity of the solutions 

 has been examined at the boiling-point of the solvent, — 38° ; 

 for liquid ammonia has a specific inductive capacity lower than 

 that of alcohol, and yet alcoholic solutions at their boiling-point 

 have incomparably lower electrical conductivity than that ob- 

 served in liquid ammonia solutions. Another striking excep- 

 tion to the Xernst-Thomson rule has recently been pointed out 

 by H. Schlundt, 5 who found the dielectric constants of butyro- 

 nitrile and pyridine to be 20.3 and 12.4, respectively, and 

 pointed out that nevertheless the solutions in the latter solvent 



1 Compare for instance Nernst's table on page 365 of the 3d edition of his book, 

 Theoretische Chemie, F. Enke, Stuttgart (1900). 



2 Compare Kahlenberg and Lincoln, 1. c, also Lincoln, 1. c, and the additional 

 articles referred to in these papers* 



3 Goodwin and Thompson, Physical Review 8, 38 (1S99). 

 * Franklin and Kraus, 1. c. 



"Jour. Phys. Chem. 5, 157 (1901). 



