2SS Mr. W. C. Dampier Whetham. [Nov. 24, 



Thus the value obtained for the specific ionic velocity is indepen- 

 dent of the E.M.F. applied, or the velocity of the ions is proportional to 

 the potential gradient. 



The experiments with the large E.M.F. were repeated with new 

 solutions 



! = 0*00046 cm. per sec. 



The specific velocity of the bichromate group is not given by 

 Kohlrausch, but ^can at once be calculated by his method from a 

 knowledge of the molecular conductivity (9'10 x 10~ 12 ), which was 

 determined by Lenz, and of the migration constant, which was given 

 by Hittorf as 0'502. The velocity comes out 



v, = 0-000473, 



a number identical with the mean value of the measurements 

 described above. 



In order to estimate the effect of a discontinuity of potential 

 gradient, another determination of the velocity of this same ion was 

 then made, the carbonate being replaced by potassium chloride, 

 whose conductivity is considerably greater than that of the bichromate 

 (11-13 X 10- 12 and 9'10 X 10~ 12 ). Two experiments were made : 



(i.) Upward velocity .... v l - 

 Downward velocity . . 



(ii.) Upward velocity .... Vj, - 

 Downward velocity . . 



These numbers show that the effect is to increase the velocity in 

 one direction, and to diminish it in the other, while (at all events, if 

 the difference of specific resistance is not great) the mean value gives 

 a fair approximation to that obtained when solutions of identical 

 conductivity are used. 



With solutions of different resistances, the junction is often ob- 

 served to become sharp when travelling in one direction, and vague 

 when travelling in the other. This is owing to the fact that any ion 

 which gets separated from the main body finds itself in a region 

 where the potential gradient is different. Its velocity is therefore 

 altered, and in the first case it rejoins the ranks, and in the second 

 it gets further and further separated from them. 



The investigation was also extended to the case of alcoholic solu- 

 tions. These possess a very much less conductivity than the corre- 

 sponding aqueous ones, and the question whether Kohlrausch's theory 

 still held good seemed of great interest. No data for the migration 

 constants are known ; hence a modification of the method was applied. 

 An experimental measurement of the velocities of both ions of some 



