KAHLENBERG — THEORY OF ELECTROLYTIC DISSOCIATION. 343 



are known to be the Letter conductors. Again liquid S0 2 , 

 which Walden 1 has shown to yield excellent conducting solu- 

 tions, has a low dielectric constant 2 that would not lead one to 

 expect such a behavior. The dielectric constant of liquid HCIST 

 was found by Schlundt 3 to be 95 ; this would lead one to expect 

 this solvent to have a dissociating power greater than that of 

 water. Preliminary tests which Mr. Schlundt and I have made 

 show that solutions in liquid HCN are very much poorer con- 

 ductors of electricity than corresponding aqueous solutions. 4 

 On the other hand I have found that amyl amine with a specific 

 inductive capacity of the order of that of chloroform 5 yields 

 fairly good conducting solutions. More striking proofs that 

 the ISTemst-Thomson rule is untenable could hardly be pro- 

 duced. 6 The fact that the Nernst-Thomson rule can not be 

 maintained takes away another pillar upon which the dissocia- 

 tion theory has been resting. 



That the "dissociating power" of solvents is dependent upon 

 the polymerization of their molecules, as claimed by Dutoit and 

 Aston, is not in harmony with the facts in many cases, has 

 clearly been shown by Kahlenberg and Lincoln. 7 Again it has 

 been maintained by Briihl 8 that "dissociating power" is pos- 



*]. e. 



2 The exact value has been determined in this laboratory by Mr. Schlundt who 

 will report upon the same in connection with numerous measurements made on 

 other solvents, among which are practically all the solvents investigated by 

 Walden, 1. c. Some older determinations of the dielectric constant of S0 2 have 

 been found in the literature by Mr. Schlundt, which he will present in connec- 

 tion with his own work. 



*1. c. 



*We hope soon to be able to publish exact conductivity measurements of solu- 

 tions in liquid HCN and also of solutions in liquid cyanogen. 



6 The exact value of the dielectric constant of amyl amine will be reported by 

 Mr. Schlundt, whose list of determinations includes a fairly complete series of 

 the substituted ammonias of both the fatty and aromatic series. 



•Nernst himself (Theoretische Chemie, 3d Edition, p. 365) has realized the dif- 

 ficulty of harmonizing his rule with the far less striking instance that NaCl, 

 KBr, etc., in formic acid (dielectric constant 62) solutions conduct almost as well 

 as the corresponding aqueous solutions, in which connection he remarks that 

 other "specific influences" come into play. He states, "Wahrscheinlich steht 

 hier in erster Linie eine Association der Ionen mit Molekiilen des Losungsmit- 

 tels". It is interesting to compare in this connection the treatment which the 

 hydrate theory receives at the hands of the same writer, page 491. 



n. c 



8 Zeit. phys. Chem. 18, 514 (1895). Ibid. 27, 317 (1898). Berichte deut. chem. Ges. 

 30, 163 (1897). 



