Manchester Memoirs, Vol. Lx. ( 1 9 1 6), No. 11. 31 



The current was then cut off for half a minute to allow 

 surface hydrogen to escape by diffusion, and the electrode 

 then connected with a mercurous sulphate electrode 

 through a galvanometer reading io _G ampere. A current 

 of 30 x jo -6 amp. passed at first and fell rapidly, but after 

 20 mins. was still 2 x io~ 6 amp. This means that about 

 IO" 2 coulombs was given by the hydrogen alloyed with 

 the mercury, or about 1 cu. mm. H^ per sq. cm. of surface. 

 A similar experiment with two copper electrodes gave a 

 still higher result, the current remaining at 30 x io~ 6 amp. 

 even after the electrodes had been taken out of the liquid, 

 replaced, and short circuited for 20 seconds. Further 

 experiments on these lines will be carried out later. 



A strip of thin sheet-iron when used as cathode in 

 dilute sulphuric acid becomes much stiffer and almost 

 brittle. When used as anode in the same liquid, this 

 stiffness disappears and the metal becomes very soft and 

 pliable. This again suggests alloy formation on the surface. 



Electrolytic oxidations and reductions may be pro- 

 duced (i.) by the action of the electrode alloy, (ii.) by 

 the action of the discharged ion, (iii.) by the action of 

 electricity itself. 



(i.) is evident from the fact that in cases where the 

 alloys are obtainable in the free state, they are able to 

 produce similar effects to those obtained electrolytically. 

 Thus palladium-hydrogen or platinum-hydrogen alloys 

 combine directly with free oxygen giving water and the 

 free metal. Also platinum oxide is a very powerful 

 oxidising agent and is probably formed at a platinum 

 anode giving rise to the very high anodic overvoltage of 

 0-86 volt. 



(In J.C.S., 105, 1914, 2422 and 2423, the anodic over- 

 voltage of platinum is given as 0*56 volt. The error was 

 not observed until after the paper was printed). 



