PROGRESS IN PHYSICAL CHEMISTRY. 301 



nitrate are almost the same in water and alcohol, but as 

 this is not the case for water and methyl alcohol he regards 

 the transference ratio as a function of the solvent. 



One of the most interesting questions in connection 

 with conductivity is : Why should certain solvents only give 

 solutions which readily transmit the current ? In answer to 

 this question, Nernst points out that if the ions be electrified 

 points the forces between them will be diminished the 

 higher the dielectric constant of the medium, and hence the 

 dissociation should be greatest for solvents having the largest 

 dielectric constant. The dissociation of hydrogen chloride 

 in hydrocarbons, ether, isobutyl alcohol, ethyl alcohol, 

 methyl alcohol and water, is shown to obey this rule, water 

 having by far the largest dielectric constant. This action of 

 the solvent was pointed out by J. J. Thomson in 1893, and 

 Whetham (122) has shown that the behaviour of solutions 

 in water, methyl alcohol and ethyl alcohol agrees quantita- 

 tively with the theory if corrections are introduced for the 

 viscosities of the solvents. 



Many observations on solutions point to the fact that water 

 is, to an extremely slight extent, dissociated into ions, and 

 the amount of the dissociation has been calculated in various 

 ways (123). The simplest method makes use of Kohlrausch's 

 numbers for the conductivity of water. This calculation 

 was made by Ostwald, who, by means of Nernst's theory, 

 also obtained an independent value from the electromotive 

 force of a voltaic cell with electrodes reversible as regards 

 hydrogen ions. Ostwald's new value has recently been 

 slightly modified by Nernst/ 



Arrhenius by using Shields' results for the hydrolysis of 

 sodium acetate by water, and Wijs from the velocity of 

 saponification of methyl acetate by water, have also obtained 

 measures of the dissociation, and Laar has tried to show 

 theoretically that Wijs' value is most probably correct. 

 The numbers thus obtained are as follows : — 



Kohlrausch, <cv6 x io -6 at 25 . 



Ostwald — Nernst, o"8 x io -7 at 18 . 



Arrhenius — Shields, i'i x io -7 at 25°. 



Wijs, i*2 x io -7 at 2 5 . 



