70 



Dr. G. Gore. 



has been passing in the previous direction, the greater that period the 

 longer must it pass in the other direction. The greater difficulty of 

 producing lines with two liquids, which differ only in degree of con- 

 centration (see Experiments Nos. 24, 26, 27, 30, 31, 32, 33, 34, and 

 39), shows that the rapidity of production of the definite line is not 

 proportional alone to the amount of current passing. I did not ascer- 

 tain for how long a period the line could be maintained. 



The bulk, or form of the mass, of either the lower or upper solu- 

 tion, have little or no effect upon the phenomena, except so far as they 

 affect the electric conduction resistance. It was frequently observed 

 that the lower liquid increased in bulk during an experiment, probably 

 in consequence partly of heat of conduction resistance and partly 

 because of ordinary diffusion transferring more liquid from weak to 

 strong solution than from a strong to weak one, partly also because 

 electric osmose is nearly always more copious from a dilute to a con- 

 centrated liquid than the reverse (see " Experiments on Electric 

 Osmose " loc. cit.). I have not ascertained whether the usually greater 

 electric osmose is wholly or only partly due to ordinary diffusion. 



4. Experiments with single Meniscus Apparatus. 



In order to be able to more perfectly analyse the phenomena by 

 reducing them to their simplest state, I employed in the following 

 experiments an apparatus having only a single meniscus tube (see 

 fig. 4). 



Fig. 4. 



The outer vessel contained the heavier liquid and the inner cylinder 



