152 On Galvanic Polarization in Liquids free from Gas. 



produce this minimum, because gases move on with extraordi- 

 nary slowness in metal when they are not impelled by any ex- 

 terior electromotive force ; to remove the last remains of them 

 consumes on this account an extreme length of time. 



To show this slowness of the gas-motion I adduce the follow- 

 ing : — Polarized plates of platinum standing in the usual air- 

 containing liquids apparently lose their polarization in a few 

 hours, or even minutes, when in conducting connexion with 

 each other. Indeed on this circumstance depended, for instance, 

 the utility of the polarizable electrodes employed by E. du Bois- 

 lleymond formerly for experiments on animal electricity. On 

 the contrary, in a gas-free liquid polarization dies away at first, 

 it is true, quickly, but later very slowly. I have in such a case 

 observed in the multiplier the depolarizing current for sixteen 

 days. From the electrolytic equivalents of the current used for 

 the polarization of the plates and the depolarization-current 

 afterwards still existing it followed that months must elapse ere 

 so feeble a current as the last named was could remove the 

 residue of the gas-charge still present. 



The phenomena in polarization which are similar to those at- 

 tending the residual electricity of a Leyden jar, are accounted 

 for by the occlusion of the gases. When hydrogen is penetra- 

 ting a platinum plate, if the current be interrupted a few seconds, 

 the gas has time during this interruption to advance further 

 into the interior, and thereby diminish its density in the super- 

 ficial layers. If the circuit be closed again, the resistance to the 

 entry of fresh hydrogen is thereby diminished, the current may 

 become stronger. Conversely, the depolarization-current can 

 remove the gas which has been driven to the surface ; if we in- 

 terrupt this current, the hydrogen slowly pressing up from 

 within will be able to accumulate at the surface and reinforce 

 its polarization. It is known that in a platinum plate, behind 

 an oxygen polarization an older hydrogen polarization can exist 

 simultaneously, which latter comes to light when the former has 

 vanished, and vice versa. 



So far as I can see, the phenomena belonging here are without 

 difficulty explained if we assume for the onward motion of the 

 gases occluded in metals the same laws as for the conduction of 

 heat. 



Lastly, it is to be remarked that in these cases, after the cur- 

 rent of the condenser has passed (that is, the electricity attached 

 only to the surface of the plates is discharged), another current 

 can only set in in proportion as gas presses from the interior of 

 the metal to the surface. If this takes place only very slowly, 

 the current-intensity in the circuit becomes as good as inde- 

 pendent of the resistance of the latter; so that, in my experi- 



