734 REPORT—1885. 
Result of Double Electrolysis on the liquid in anode and cathode vessels, by 
the time one equivalent of an ion belonging to each substance has been 
deposited. 
Substance = pee Loss in anode vessel Total loss 
A 3A _ 3A 0 
A’ An! 1—$0' 1 
C 1—}a 3A 1 
cy — in’ 2A! 0 
AC 1—3a 1x 1 
A'C! ax! 140’ 1 
AC’ =X —i’ =i 
The facts of migration are thus perfectly accounted for, and a good deal 
beside. Thus free acid is shown to appear both at anode and cathode in 
the electrolysis of sulphate of copper with platinum electrodes: the amount 
at anode being 4 of that at cathode. Hencea determination ofthe rela- 
tive amount of free acid in anode and cathode vessels at once gives a 
means of determining X : ’, the proportion in which the current is shared 
by the two substances, if the above hypothesis be true. 
It is easy enough to put Hittorf’s migration number x into the above 
table instead of X\: one has only to write $A = lagen but only three lines 
of the table have any meaning on the Hittorfian hypothesis—viz. the first, 
‘third, and fifth lines—relating to the loss of the compound AC in the 
respective vessels ; the solvent, or other compound, A/C’, is ignored, and 
the production of free acid near cathode is not supposed to occur. But 
it does occur, as some preliminary experiments I have made prove. 
Our hypothesis makes the formation of acid and everything else 
independent of intensity of current. Experiment may not confirm this; 
but then we must remember that we have assumed no mizture af ions to be 
liberated, and it is known that with intense currents some hydrogen is 
given off at cathode as well as copper deposited. If this be taken into 
-account the reasoning must, of course, be generalised. 
A more formidable objection, however, may be made to the theory, 
viz. that it virtually assumes the conduction by the two sets of atoms to 
be independent; for instance, the » portion of current starting fairly 
away from the neighbourhood of electrodes via CuSO, is supposed to go 
through CuSO, all the way, and not to come across patches of water or 
to take either ingredient of the liquid at random. 
Consider for a moment what would happen if this did occur. Take 
a chain, CuSO,, CuSO,, H,O, H,O, H,O, H,O, CuSO,, CuSO,, H,0. 
Imagine the usual Grotthus action to go on through the chain: the 
result is 2H,SO,and 1CuO. This is an unstable condition, or condition 
of non-minimum energy; so one H,SO, will combine with the CnO; 
but it need not combine at once, because there is no need for the two 
compounds to be formed close together. Hence we must expect free 
acid sometimes to make its appearance in the middle of the liquid away 
from either electrode, and also for CuO to appear (in the form of some basic 
salt I suppose) too. Choosing proper materials, an action of this kind 
might be readily looked for, but I have never heard of its being detected, 
