ON ELECTROLYSIS AND ELECTRO-CHEMISTRY. 201 
to note that the definition of the electrolytic property by the dissociation 
of the electrolyte seems applicable not only to liquids but also to solids ! 
and gases.” 
§ b.— What are the ions in any Electrolytic Decomposition ? 
In Part II., Section b, we have seen that the process of conduction 
through an electrolyte consists in the motion of ions in opposite directions, 
each carrying a definite charge of electricity. It is a matter of great 
interest to identify the ions in any particular case, and in the years fol- 
lowing Faraday’s researches in electrolysis attempts to identify ions by 
the methods of chemical analysis were very numerous, and the interest 
in them was increased by the fact that the results arrived at were entirely 
opposed to the Berzelius theory of salts. This department of the subject 
is most conspicuously represented by a series of well-known papers by 
Hittorf, Pogg. ‘ Ann.’ €9 p. 177, 98 p. 1, 193 p. 1, 106 p. 337, and 
p. 513 (1853-59). 
The ions are deposited at the electrodes, where they are either set free 
or take part in secondary actions, and the first step towards the identifica- 
tion of the ions is to determine the primary chemical result of electrolysis 
from the final results which are due to secondary actions. Thus when a 
solution of KHO is electrolysed the obvious products set free are H and 
O, but the analysis of the liquid shows that both these products are 
secondary, and are due to the action of the primary products, K and HO 
respectively, upon water. 
In order to determine the primary results of electrolysis from the 
secondary products, the division of the cell ideally represented in Part I. 
has been adopted, but the simple divisions there mentioned as used by 
Daniell and Miller do not serve the purpose of ideal separation; the 
arrangements necessary for this purpose are described in Wiedemann’s 
*Electricitit,’ 2, § 549. Hittorf’s arrangements are described in the 
same volume, § 550, and an apparatus used recently by Loeb and Nernst 
3 figured and described in p. 950 of vol. 2 of the ‘ Zeitschrift fiir phys. 
hem.’ 
It might be supposed that the results of the analysis of the liquid 
contained in the anode and cathode vessel respectively would give the 
amount and nature of the decomposed compound, and the amount and 
nature of each of the products, and hence that the ions would be statistically 
determinate ; that is, without giving any information as to whether all the 
ions were of the same kind or not, a result would be obtained which 
would strictly represent the average process. According to the general 
view, however, chemical analysis fails to give this conclusive evidence, 
the original electrolytic process being complicated by electric endosmose, 
and the unequal dilution in solutions, mentioned in Part I. (6) (c), as 
the following example will show :— 
In the electrolysis of a solution of copper sulphate containing 
3793 grammes of copper in 100 c.c. of solution, while one gramme 
equivalent (4 Cu) of copper is being deposited in the cathode, the total 
gain of copper in the cathode vessel, taking account both of the 
deposited and the still dissolved metal, is *75 of an equivalent, and 
» Van ’t Hoff, ‘Ueber feste Lésungen und Moleculargewichtsbestimmung an 
festen Kérpern,’ Zeitschr. fiir ph. Chem. 6, p. 322, 1890. 
2 J. J. Thomson, Phil. Mag. 358, 1890. 
