148 On Irreciprocal Conduction. 



carbon electrodes, the non-formation of the film with platinum 

 electrodes in various oxidizing agents, and other minor pro- 

 perties, which will be found in the previous part of the paper, 

 have led us to doubt the accuracy of the oxide theory in the 

 case of platinum or carbon electrodes, and to provisionally 

 adopt another theory, as indicated below. 



The behaviour observed at the anode in the case of the 

 electrolysis of sodium benzoate solution suggests that the 

 badly-conducting sulphuric acid which appears, in electro- 

 lyzing that liquid, at the anode, may under certain conditions 

 be able to stop the current. In discussing such a hypothesis 

 we have to remember : — 



(1) That the specific resistance of H 2 S0 4 increases very 

 rapidly with the degree of concentration. 



(2) That the concentration of the acid at the anode during 

 electrolysis is partially prevented (a) by the evolution of 

 oxygen (as we have pointed out in Phil. Mag. April 1888, 

 p. 283), and also (b) by its dilution, the rapidity of which 

 depends on the viscosity and strength of the acid solution 

 used. 



(3) That a certain amount of oxygen forms a gaseous layer 

 on the anode. 



Bearing these points in mind, we will examine whether 

 they are sufficient to explain our experiments. 



Firstly. With low current-densities the concentrated acid 

 will be formed in quantity too small to withstand the two 

 causes tending to cause its dispersion. 



Secondly. With greater current-densities the rate of forma- 

 tion of the concentrated acid will overcome the rapidity of its 

 dilution, and the resistance of the layer of acid round the 

 anode may increase so rapidly as to form an insulating film, 

 before enough gas has been evolved to cause its destruction ; 

 the gas which has been produced becoming entangled in the 

 viscous concentrated layer, which, once formed, will not be 

 easily removed by friction, &c. 



Thirdly. With still higher current-densities, though there 

 will be a greater tendency towards the formation of this layer 

 immediately the circuit is made, yet it is conceivable that 

 more gas will be produced in the same time than can be pre- 

 vented from escaping, and thus destroying the non-conducting 

 layer ; for it would be expected that the tendency of the gas 

 to escape would increase at a greater rate than the tendency 

 of the acid layer to prevent the escape. An upper limit to 

 the current-density ought therefore to be discoverable beyond 

 which no film can be produced. In the case of foil elec- 

 trodes placed horizontally with only the lower surface exposed 



