j ON ELECTROLYSIS AND ELECTRO-CUEMISTRY. 205 
» 
_ molecules as well as molecular aggregates, and that while the colour of the 
solution and a number of other properties depend upon the latter, the 
electricity may be conveyed by the former alone. It might even be sug- 
gested that if the temperature coefficient of absorption of a coloured 
solution were determined, it would be found to be closely related to the 
temperature coefficient of conductivity, and when allowance was made 
for the change of viscosity it might furnish the temperature coefficient 
of dissociation. 
An interesting point in connection with the determination of the 
ions is the question whether the ious are all of one kind in an electrolytic 
_ solution; in other words, whether the water conducts, or all the current 
_ is carried by the molecules of the dissolved salt. If an electrolyte be a 
_ mixture, as of HCl and H,O, do both compounds take a share in the 
conduction, or one only? Lodge (‘ Brit. Assoc. Rep.’ 1885) argued 
strongly in favour of a division of the conductivity between salt and 
solvent, and founded a theory of migration on that hypothesis; but the 
experimental evidence seems to have left the subject in the following 
state.! It is possible to obtain water with a very high degree of insulat- 
ing power, but, when it is pushed to the extreme limit, it is impossible 
to tell whether the conduction is due to water molecules or undetected 
impurity. Indilute solutions the increase of conductivity which is con- 
ferred upon the water by the addition of a small quantity of salt is due to 
the added salt alone, and the conductivity of a dilute solution containing 
the added salt may be deduced from the observed conductivity of the 
solution by subtracting the conductivity of the water of which the solu- 
tion was made; in other words, conductivity by water molecules forms 
no part of the added conductivity due to the salt.2 Thus water is 
regarded as a body of a special kind, which dissociates other salts and 
makes them conduct, but itself carries the current to no appreciable 
extent. 
The resulting chemical products are certainly different for different 
values of the current density. Ifa dilute solution of copper sulphate be 
subjected to electrolysis under the effect of a very high electromotive 
force, bubbles of hydrogen speedily make their appearance at the cathode, 
and it has been supposed that there is a limiting value of the current 
density beyond which the current ceases to traverse the salt solely, and 
an appreciable amount passes through the water. C. L. Weber, ‘ Zeitschr. 
fiir phys. Chem.’ vol. 4, p. 182, 1889, has employed this phenomenon to 
determine the absolute velocity of the ions. Itmay, however, be explained 
by the continued impoverishment of the solution in the neighbourhood 
of the cathode ; and, in fact, if the electrolysis be continued for some 
time between platinum electrodes, the whole of the copper may be 
abstracted from the solution. 
Ihave tried myself to ascertain whether the water took part in the 
conduction, by interposing a very dilute solution of copper sulphate 
between two much stronger ones, so that, if the water conducted, a layer 
of copper hydrate would be formed at the junction between the strong 
! The discussion has been somewhat lengthy. Finally Kohlrausch has admitted 
that an experiment of Faraday’s may possibly be explained satisfactorily by attri- 
buting a minute conductivity to the solvent. See Wied. Hlec. 2, § 583; Kohlrausch, 
aie 26, p. 161; Arrhenius, Brit. Assoc. Rep. 1886, p. 311; Hermann, Beidl. 
xi. p. 831. 
* F. Kohlrausch, Wied. Ann. 26, p. 190. 
