TRANSACTIONS OF SECTION B. 659 



the two fiiiicls respectively : but this current very rapidly diminishes to infinitesi- 

 mal proportions. The usual explanation of this given in the textbooks is that 

 during the passage of the current electrolysis takes place, causing the development 

 of free hydrogen on the surface of the plate in the acid fluid, and of oxygen on 

 that of the other plate, so that an inverse E.M.F. is set up by the incipient ga3 

 battery thus produced, which by-and-by becomes practically equal to the E.M.F. 

 due to the chemical action of neutralisation. No direct quantitative proof of this 

 electrolysis, however, appears ever to have been given as regards the hydrogen ; 

 whilst as regards the oxygen the main evidence is the observation of Becquerel, 

 that if nitric acid be the acid and caustic potash the alkali, a tolerably constant 

 current is developed, the flow of which is accompanied by continuous evolution 

 of oxygen from the alkali plate, whence the term ' pile a oxygene ' applied by him 

 to the combination, the nitric acid being simidtaneously reduced to lower oxides 

 of nitrogen. 



Whilst studying the action of various gas-film cells, we had occasion to make 

 some experiments with * neutralisation cells,' i.e., cells in which one essential 

 feature is tliat an acid and an alkaline fluid are used, which neutralise one another 

 during the action; thus in the case of Becquerel's ' pile a oxygene/ the scheme 



2IIN0, I 2KNO3 I 2K0II 



Hj I 2iN03 K 1 2NO3 K 1 OH., I 



represents the primary action, the hydrogen, of course, not appearing as such, but 

 reducing the nitric acid. We found it difficult to obtain by titration very sharp 

 figures proving that one equivalent of acid disappears on one side and one of alkali 

 on the other for every equivalent of silver thrown down by the current in a silver 

 voltameter ; but our results always at least approximated to this. On the other 

 hand, we found no difficulty at all in proving that in Becquerel's ' pile a oxygene,' 

 8 milligrams of oxygen (5-6 c.c.s. at 0° and 760 m.m.) are evolved for every 108 

 milligrams of silver deposited. Moreover, we found that various oxidising fluids 

 can be substituted for nitric acid without affecting this result ; thus solution of 

 potassium permanganate or ferricyanide acidulated with sulphuric acid; chromic 

 acid dissolved in sulphuric acid ; hydrochloric acid saturated with chlorine ; or 

 dilute sulphuric acid saturated with bromine, all cause oxygen evolution (though 

 not all as rapidly as nitric acid) in precisely the quantity equivalent to the silver 

 deposited. 



It occurred to us that by suitably modifying the liquids used we might obtain 

 similar quantitative evidence as regards the hydrogen. If nitric acid or other 

 oxidising agent can be reduced by the nascent hydrogen whilst the oxygen escapes, 

 it might be expected that if some highly oxidisable substance is dissolved in the 

 caustic alkali used, the oxygen might be arrested whilst still nascent, whilst con- 

 versely the hydrogen escaped free at the other side ; and this, in fact, we idtimately 

 succeeded in doing. Several oxidisable substances, however, proved too weak ; 

 thus sulphites and ferrocyanides caused no distinct hydrogen evolution; but 

 caustic soda containing hyposulphite (Schiitzenberger's ' hydro.sulphite ') or 

 pyrogallol caused continuous evolution of hydrogen in quantity strictly propor- 

 tionate to the cm-rent passing, i.e., 11-2 c.c. normal for e^ery 108 milligrammes of 

 silver deposited. 



Precisely the same result was obtained when certain metals not ordinarily 

 soluble in cold acid or alkaline fluids, and simple caustic soda solution were used 

 instead of these alkaline oxidisable fluids with platinum plates. Thus tin or 

 lead immersed in caiistic soda and opposed to platinum in dilute sulphuric acid 

 dissolved freely, producing copious evolution of hydrogen from the platinum plate 

 surface. Th6 same result was obtauied on substituting copper and ammonia solu- 

 tion for tin and caustic soda ; whilst by employing strongly alkaline potassium 

 cyanide solution, mercury, silver, gold, and palladium were readily dissolved with 

 hydrogen evolution from the opposed platinum plate immersed in dilute sulphuric 

 acid. In all these different cases the amount of hydrogen obtained was exactly 

 proportionate to the current flowing, 11-2 cc, being collected for every milligram 

 equivalent of silver deposited in the voltameter. 



