ON ELECTROLYSIS. 729 
Finally, let the water conduct ae and the salt = th of the cur- 
a 
rent, then the ratio of the free acid formed at anode to that formed at 
cathode will be «—1. 
Now I have made many preliminary experiments ; avoiding porous 
partitions of course, as introducing electric endosmose, capillary forces, 
and all kinds of unknown disturbances; using only two vessels con- 
nected with a syphon U tube, and other such arrangements. I find in 
all cases acid at both poles, though, with platinwm electrodes, usually most 
at anode. I do not press these preliminary results; a great deal of easy 
quantitative work suggests itself in this connection, and it is necessary 
to specify and to vary the concentration of solution and the strength of 
current used. The method is the important thing, and I should like to 
have it either approved or condemned. 
If one uses copper electrodes, the results are somewhat different. On 
the hypothesis that the salt alone conducts, no free acid should be 
formed anywhere; the solution should become impoverished at cathode 
and concentrated at anode—phenomena which are well known to occur 
to a greater or less extent. If, on the other hand, the water alone 
conduct, free acid should appear at cathode but not at anode; the solu- 
tion near cathode shonld again become impoverished of copper, though 
not of SO,; while upon the anode is liable to form a coat of oxide 
which there is no sufficient supply of free acid near to dissolve off. And 
all these phenomena are also found to occur. The deposit of oxide on a 
copper anode is with intense currents most marked, so much so as almost 
to stop them. It is most instructive to put some dead-beat instrument 
like Ayrton and Perry’s Ammeter into circuit with a copper voltameter 
too small for the current, and watch the needle descend from the stops 
quickly almost to 0.1 Free acid does make its appearance at the cathode 
—and, with intense currents, plenty of it. 
Roughly one may judge perhaps that the salt conducts about three- 
fourths of the current and the water one-fourth, but I have no certain data 
yet for any such statement; and the proportion may well vary, for all I 
know, with the current intensity. One may, however, definitely accept 
hypothesis 3 or 4 as certainly truer than 1 or 2. 
The fact that acid is produced at anode, and alkali at cathode, in the 
decomposition of, say sodic sulphate, proves that the salt conducts at least 
some of the current; and an estimation of the amount of acid and alkali 
respectively generated would decide what proportion was so conducted. 
Again, experiments with electrolytes in series, as made by Hisinger, by 
Berzelius, and by Davy, are interesting in their bearing on this point.? 
Thus, take three vessels containing, say, HCl, HNO;, and AgNO,;; or 
Na,SO,, HCl, and BaCl,, respectively; and the formation of a preci- 
pitate is a good identification of a true ion. ‘The locality of such preci- 
pitate is still more instructive, since it conveys information as to the 
relative rates at which the opposite ions travel, and so leads us on to 
the next question, concerning what is called ‘ the migration of ions.’ 
' Hence, in arranging such voltameters, the anode should be larger than the 
cathode. 
? An old experiment of Faraday’s (Exp. Res. 494) on the formation of magnesic 
hydrate at a junction of Epsom salts and water when a current passes, is now ac- 
<epted, even by Kohlrausch, as proving that water may in some cases conduct. 
