THEORIES OF ELECTROLYSIS. 



OUR knowledge of the phenomena which accompany 

 the passage of currents of electricity through liquids 

 has lately been much extended, and our theoretical ideas 

 about the nature of electrolytes have undergone at least an 

 equal development. 



At the present moment it seems worth while to state 

 the actual condition of the subject, and to trace the line 

 between ascertained facts and the necessary consequences 

 of those facts on the one hand, and the hypotheses which 

 have been framed in order to explain them on the other. 



The fundamental phenomenon from which we shall 

 start is the truth of Ohm's law for electrolytes. This is an 

 indirect deduction from the work which has been done by 

 Kohlrausch and others on the measurement of the con- 

 ductivity of liquids, and the accuracy with which the law 

 holds has been investigated by Fitzgerald and Trouton and 

 shown to be very great. 



Considering the whole of a circuit containing an electro- 

 lytic cell, it is obvious that Ohm's law cannot express the 

 facts, since no current will pass till the reverse electromotive 

 force of polarisation is overcome. But polarisation only 

 exists at the electrodes, and, once the current gets inside the 

 liquid, its intensity is strictly proportional to the effective 

 electromotive force which there acts. It follows that no 

 chemical work can be clone by the electrical forces acting 

 within the liquid, so that either interchanges between the 

 parts of the molecules must already be going on, or else the 

 work done in tearing asunder one molecule must be given 

 bu.ck bv the formation of the next. 



Since the former of these hypotheses is supported by the 

 facts of chemical double decomposition, it is generally 

 adopted, and we hence conclude that molecular inter- 

 changes are always going on, whether or not a current 

 passes — the effect of the electromotive force being to carry 

 a positively electrified stream of ions in one direction, and a 



