726 REPORT—1885. 
bably a fall quota of decomposition, but, unless special precautions are 
taken, recomposition is just as rapid. 
The same thing goes on far more slowly in dilute acid, especially if 
it contain dissolved gases; and the experiments of Helmholtz with an 
ingeniously contrived gas-free cell, as related in his Faraday Lecture of 
1881, may be taken as settling the quantitative question whether this 
decay of polarisation is any of it due to slight metallic conduction, or 
whether the whole of it is rigorously due to solution and diffusion of the 
semi-liberated ions. It is of course true that these experiments do 
not settle it for all electrolytes, but for dilute sulphuric acid at any rate 
they prove that the diffusion and indirect recombination of the ions 
accurately account for everything; and that, when these are stopped a 
voltameter behaves as a good and non-leaking condenser up to a certain 
E.M.F., beyond which it conducts with visible decomposition. 
It may be as well, perhaps, to avoid misconception by stating that, by 
the solution and recombination spoken of above, I do not absurdly sup- 
pose that nascent O and nascent H, respectively dissolved in the liquid, 
travel to meet each other and recombine. Upon such direct action as 
that the presence or absence of dissolved air could have no influence. 
What one supposes to go on is this. Semi-liberated hydrogen, finding 
dissolved oxygen in its neighbourhood, combines with it; the corre- 
sponding oxygen simultaneously liberated at the other pole dissolves in 
the liquid, and, so replenishing it, keeps the action continuous at a slow 
rate regulated by the rapidity of diffusion. 
I see nothing, however, to prevent a liberated ion from dissolving in 
the liquid, or from diffusing across to the other electrode, where it may 
combine with the nascent ion, and so reduce the apparent out-put of 
the cell; and this action, having nothing to do with dissolved air, 
might go on even in Helmholtz’s gas-free cell. But it could not begin, 
I imagine, unless the H.M.F. were sufficient to set both ions free ;' conse- 
quently, by never applying more than a feeble H.M.F., this disturbing 
action may be obviated. 
When dissolved oxygen already exists, neither ion has really to be set 
free; consequently even a feeble E.M.F. is sufficient to maintain a weak 
current. 
In cases where the polarisation test for electrolysis thus fails, Helm- 
holtz proposes another, viz.: Hlectrolytes do not fall into Volta’s tension 
series. An assemblage of metals at constant temperature can give no 
permanent current: introduce an electrolyte into the series, and they can. 
On this principle Helmholtz considers he has proved that the con- 
duction of glass is not metallic ; for he makes a Daniell cell with a com- 
plete glass partition in place of a porous cell, and finds that the glass 
does not stop the current by polarisation, as a metal partition would, but 
that the cell behaves exactly like any other cell with an enormous internal 
resistance. 
One more test, those of us who believe in the electrical theory of 
light may suggest, viz. transparency. A conductor, if transparent, must 
be an electrolytic conductor. Thus a fused salt, if it be clear and trans- 
parent and yet conduct well (as argentic iodide does for instance), may 
be assumed to be an electrolyte, despite the absence of products of decom- 
1 The E.M.F. required to set an ion free enough to enter into solution is almost 
enough to set it free altogether. The only difference, I suppose, is pdv; and at 
ordinary pressures this is negligible. 
