338 BULLETIN OP THE UNIVERSITY OP WISCONSIN. 



regularly with the dilution; but such cases do exist. So, for 

 instance, the molecular conductivity of aqueous solutions of the 

 alkaline hydroxides first increases with the dilution and then 

 decreases as the dilution increases 1 . 



Prom the facts here presented in the case of aqueous solutions 

 and those detailed above in the case of non-aqueous solutions, 

 it follows that there is no such connection between the freezing- 

 points and boiling-points of solutions on the one hand and their 

 electrical conductivity on the other, as is claimed by the theory 

 of electrolytic dissociation of Arrhenius. In numerous cases 

 not even a qualitative agreement exists. A careful scrutiny of 

 Arrhenius' original table reveals the fact that while in a certain 

 number of instances the agreement of the values of the factor 

 i, as computed from cryoscopic measurements and from con- 

 ductivity determinations, is acceptable, in many cases it is 

 poor, and again in others it is entirely wanting. To be sure, 

 Arrhenius bases this table upon measurements which he deemed 

 might in some cases contain large errors; but later experi- 

 mental work has really only served to emphasize the presence 

 of such discrepancies. Further we have seen that investiga- 

 tions of non-aqueous and of aqueous solutions have in reality 

 yielded a host of facts which even the most ardent adherents 

 of the dissociation theory have not been able to harmonize with 

 it. For some facts, to be sure, explanations have been offered 

 upon the basis of auxiliary assumptions which frequently had 

 no root in facts, having been created simply to save the dis- 

 sociation theory from being pronounced untenable. 



The attempt has been made to explain all of the various 

 properties of electrolytic solutions upon the basis of the theory 

 of electrolytic dissociation. To these efforts attention must 

 now be directed. The various additive properties of salt solu- 

 tions as mentioned by Arrhenius and as detailed by Ostwald 

 and by Nernst in their text-books, are presented as supporting 

 the dissociation theory. But it is clear that even if these ad- 

 ditive properties could be explained on the basis of the disso- 



*The determinations of Kohlrausch show this phenomenon. See also the 

 measurements of Kahlenberg and Lincoln, Jour. Phys. Cheni. 2, 87 (1S98). 



