ON ELECTROLYSIS. Gols 
By the time one equivalent of each ion has been liberated against the 
electrode, the ions transmitted by the tube will be 2 and nw equivalents 
respectively: hence the anode vessel will have lost altogether 1 — nz 
equivalents of each ion, and the cathode vessel will have lost 1 —z 
equivalents. 
But z and nz together convey the whole current, soz + nz = 1. 
Thus, then, we have as the respective losses of the liberated elements. 
by the two vessels : 
Loss in anode vessel. . =1——” at ; 
ltn Il1+n 
? 1 n 
Loss in cathode vessel . = 1 — —— = ; 
lin tIl-+n 
Hittorf’s well-known expressions. 
Or consider it this way : 
By the time one equivalent of copper has been deposited, and one of 
SO, has attacked either the anode itself if of copper, or the water in its 
neighbourhood if of platinum, forming CuSO, in the one case, 
H,SO, in the other—no matter which :—by the time all this has hap- 
pened, 3% atoms of SO, have passed from neighbourhood of cathode- 
towards anode, and z of copper have gone the other way. 
The cathode liquid has thus lost 1 — « of copper, and so must have: 
lost 1 — x of SO, too, no more and no less, or else it would contain: 
isolated radicles. 
The anode liquid has gained these 1 — xatoms of SO,; and they 
may be also reckoned as 3z, as we just now saw; whereforea =. The 
cell happens not to lose any SO, by deposition, but, instead, it either 
loses an equivalent of O or gains an equivalent of Cu, according to the 
nature of the anode. 
So the final result in any case is a transfer of 2? equivalent of 
SO, from cathode vessel to anode vessel, combined with a transfer of + 
equivalent of Cu in the opposite direction: a net loss of ? equivalent of 
copper sulphate by the cathode solution : and, in the anode liquid, either a 
net gain of # equivalent of CuSO,, if the anode be copper ; or a gain of a 
whole equivalent of H,SO, combined with a loss of + equivalent of 
CuSO, and of a whole equivalent of oxygen, if the anode be platinum. 
All this is in fair agreement with experiment. The only obviously 
weak place is that relating to the production of free acid. According to- 
the above it ought to appear, if at all, only at anode; now in practice, if 
the experiment be tried, it will be found to appear at cathode too. 
Still, the local concentration changes are satisfactorily and very simply: 
explained by this customary migration hypothesis. 
But before we decide that the sufficiency of this explanation proves 
its necessity or actual truth, let us try whether we cannot get the same 
result on another, and not only plausible but really quite necessary, 
hypothesis, viz. the hypothesis that the solvent conducts some of the- 
current as well as the salt dissolved ; and let us see whether this alone 
will not account for the whole ‘migration’ phenomenon, even though: 
opposite corresponding ions be supposed to travel at equal rates. 
By drawing a scheme, say, for the electrolysis of CuSO, solution with 
platinum electrodes, we can soon see that this hypothesis will do what is 
wanted; for we find, that if CuSO, conducts the whole current, the- 
