General considerations on the theory of the separation of H
1
and H
2
by electrolysis of water
1. The Facts .―In a solution of NaOH (or KOH) with nickel (perhaps with other) electrodes and at high current densities of the order of 1 ampere per cm. 2 of electrode the observed rates of evolution E 1 and E 2 of H 1 and H 2 at the cathode obey an equation of the type E 1 /E 2 = q D 1 /D 2 , (1) where D 1 and D 2 are the relative concentraction of the H 1 and H 2 atoms in the water as a whole. The coefficient q has been shown to be independent of D 1 /D 2 over a very wide range of relative concentrations. Under the conditions most favourable to separation, which appear to be those just specified, q can have a value as great as 6, or perhaps 7. With other metals than nickel for the electrodes and with lower current densities the factor q may fall to a value well below 2 or even to a value so nearly unity that no effective separation occurs. The dependence of q (if any) on the temperature of the solution is not known. 1.1. The Arguments .―The facts have already been commented on by Polanyi from the theoretical point of view. He has been led to conclude from them that the separation is to be attributed to a difference of " over potential" for the deposition of H 1 and H 2 ions on the cathode, and consequently that Gurney's theory of the over-potential of the hydrogen-electrode at low current densities (as measured, for example, by Bowden) must be discarded. These conclusions if correct and unavoidable are of the greatest importance. It is most general way possible, in order to see that no types of mechanism have been overlooked which could lead to the observed results. When this is done it is found that the mechanism discussed by Polanyi which refers the separation to differences of over-potential is not the only mechanism which must be held possible a priori . The observed separation could perfectly well occur by a mechanism consistent with Gurney's theory. It does not yet appear to be possible to decide confidently between the two possibilities on experimental grounds.