728 REPORT—1885. 
cyanide solution of copper and zinc. True the combined resistance is not 
the semi-harmonic mean of the separate resistances, but there is no 
earthly reason why it should be. It is a mere superstition to expect such 
a result. Even if the substances were really mixed and not combined, 
it would not be so; unless there were a boundary surface between the 
media made wholly of stream-lines, i.e. a surface across which no elec- 
tricity flowed.! 
Another mode of discriminating between the hypotheses is to use 
intense currents.? For with intense currents you are more likely to 
get off the real ions: secondary actions are hurried, and cannot go on 
properly. With intense currents the ions are very apt to come off ina 
very self-combined and energetic state (as ozone instead of oxygen, &c.), 
and an extra polarisation force accounts for this extra energy; but 
they do not so easily combine with other matter in the solution. 
Now Magnus showed that from a solution of CuSO, weak currents 
deposited copper only, while strong currents liberated hydrogen as well. 
In fact the observation is now a commonplace. But it is much easier to 
turn copper out of combination than hydrogen. So if copper were the 
real primary ion there is no call for hydrogen to appear. It looks there- 
fore as if hydrogen were at least a primary ion, and possibly the only one. 
But then, as pointed ont by Smee, local exhaustion of the liquid near the 
cathode may cause such a deficiency of copper in the liquid touching the 
plate, that, if the current is too intense for diffusion to keep up the supply, 
hydrogen must perforce be liberated from the simple acid coating of the 
cathode. Hence in repeating any experiment of this kind it will be 
necessary to stir or scour the plate constantly and vigorously ; and even 
then there would be some uncertainty about the matter, supposing hydro- 
gen persistently appeared. 
Is there any more direct and simple mode of answering the question 
as to whether the salt or the water or both conduct the current P 
Yes, what seems to mea very simple and satisfactory one; by determin- 
ing at which pole free acid appears ; and, if it appears at both, by deter- 
mining how much appears at either. 
For the sake of clearness it may be well to point this out in detail, 
though indeed it is so simple as hardly to be necessary. 
Consider a solution of copper sulphate in water; decompose it with 
platinum electrodes, and first make the assumption that the sai¢t conducts 
the whole current, or that the real ions are Cu and SO,. At the cathode 
Cu appears and is deposited—nothing else happens. At the anode SO, 
appears, decomposes water, forming sulphuric acid, and setting free oxygen. 
Thus on this hypothesis all the free acid appears at the anode. 
Next make the assumption that water conducts the whole current, 
so that hydrogen and oxygen are the true ions. At the anode oxygen 
appears and is liberated—nothing else happens. At the cathode hydrogen 
appears, decomposes CuSQ,, forming sulphuric acid, and setting free 
copper. So on this assumption all free acid appears at the cathode. 
} This fact, that no crossflow between ‘multiple arc’ conductors must occur, if 
their combined conductivity is to be the sum of their separate conductivities, is 
sometimes overlooked. The covering on ordinary wires supplies the necessary con- 
dition ; but if the wires touch the law breaks down. To the parts of one conductor 
the ‘ divided circuit law’ is by no means necessarily applicable. 
? Meaning always by ‘intensity’ of current, strength of current per unit area, or 
what is sometimes inconyeniently styled ‘ density’ of current. 
