IONIC VELOCITIES. 383 



ous curious phenomena (analogous to those observed by 

 Gore, Proc. R. S., 1880 and 1 881) were observed. Several 

 surfaces of separation often appeared, the liquid between 

 them being of various tints, ranging from red through 

 purple to blue, and occasionally some of these junctions had 

 their motion reversed, and either for a time or permanently 

 travelled in the wrong direction. 



This again favours the view that in such solutions com- 

 plex ions are formed. Whether any such phenomena occur 

 in the case of aqueous solutions is not yet clear, but experi- 

 ments in the velocities of the ions of electrolytes of abnor- 

 mally low conductivity may throw light on this question. 



It will be observed that nothing has been said about the 

 state in which the ions exist or the mechanism by which 

 they pass from place to place. The theory and its confir- 

 mation are thus quite independent of the view we may take 

 of the nature of the processes of solution and electrolysis. 

 The phenomena of chemical action and the agreement with 

 Ohm's law shown by electrolytes prove that a continual 

 molecular interchange of partners is going on whether or 

 not the current passes, but a knowledge of the particular 

 mechanism by which this is accomplished is not necessary 

 for the investigation of Kohlrausch's theory. Whether we 

 accept the view of Van t'Hoff, Ostwald, and Arrhenius, 

 which supposes that all but a small percentage of the num- 

 ber of molecules in an electrolytic solution are dissociated 

 into ions existing uncombined in the liquid, or the chemical 

 hypothesis which requires freedom only at the instant of 

 interchange of partners between two colliding molecular 

 structures, we are obliged to imagine the ions travelling by 

 some means under the influence of the electric forces and 

 attaining a definite average limiting velocity. But although 

 Kohlrausch's theory does not oblige us to accept either the 

 "dissociation" or the "chemical" view of electrolysis, the 

 fact that in dilute solution each ion has a definite velocity 

 independent of the other present seems to favour the idea 

 that electrolytes contain free ions. It is important to re- 

 member that, as J. J. Thomson has pointed out (Pki/. Mag , 

 1893, xxxvi., p. 320), two bodies charged with opposite kinds 



