ON ELECTROLYSIS AND ELECTRO-CHEMISTRY. 209 
observations on mixed solutions, but the results are not arranged in the 
same form as those of Arrhenius, and the isohydric law is at any rate not 
apparent. (See also Ewing and Macgregor for resistances of mixtures of 
ZnSO, and CuSO, solutions, ‘Trans. R.S.E.’ 27, p. 51, 1873, and Bonty, 
*C.R.’ 104, p. 1699, 1887, ‘ Beibl.’ 11, p. €50.) Bouchotte, Paalzow, and 
Klein are also referred to by Arrhenius. 
§ c.—The Williamson-Olausius Hypothesis. 
We have seen in Part II. } (p. 189) that the transfer of electricity 
_ through an electrolyte is convective. If we consider, on the well-known 
hypothesis of Grotthuss, a chain of molecules of the electrolyte connecting 
the anode and cathode, the separation of an ion of each kind at the two 
electrodes respectively is associated with the simultaneous interchange 
of partners throughout the whole length of the chain. This assumption 
is sufficiently natural, for if the molecule at one end of the chain, at the 
anode suppose, be the one decomposed by the current, the anion remains 
at the anode, but the other part of the molecule, the cation, has to appear 
at the cathode, so far as we know, simultaneously. Now, on the assump- 
tion mentioned above, the time required for the transfer will’ be the same 
for long chains as for short ones (since every pair of ions into which the 
molecules are resolved will be under the action of equal separating 
forces), and is merely the time required for the separated ions to pass 
over the distance intervening between a single pair of molecules, and may 
well therefore be too small for measurement. 
| The interchange of ions between molecules has indeed long been an 
accepted notion in electrolysis, and requires no defence. And from the 
fact that the smallest electromotive force produces a current through an 
_ electrolyte, and that the physical properties of the liquid are, so. far as we- 
know, identical in every respect, when conducting the current and when. 
not, if also seems natural to suppose that the interchange of ions between: 
the molecules of an electrolyte is constantly going on whether a current. 
is flowing or not, but that the direction of the interchange is fortuitous. 
The idea of the dissociation and reformation of molecules constituting a. 
dynamical equilibrium of a chemical compound was originally suggested 
by Williamson! to account for etherification, and the explanation of 
electrolytic action by the same idea is due to Clausius,? who suggested 
that the effect of electromotive force was to determine the direction of 
the average motion of the respective ions, and not itself to produce the 
dissociation and recombination. 
It would follow that the work required to produce electrolytic decom- 
position is wholly spent in setting free the ions at the electrodes. 
Whatever representation may be made of the state of the molecules: 
of an electrolyte when no current is passing, it must be so arranged as to 
take account of the fact that when a current passes the dissociation and 
recombination are attended with the development of a quantity of heat in 
accordance with Joule’s law ; whereas when no current passes no heat is 
developed ; and the mere irregularity of direction of motion would not 
dispose of the heat production because that is independent of the direction 
of current and depends merely on the magnitude. Professor Fitzgerald 
1 Liebig’s Annalen d. Chem. u. Pharm. vol. 1, p. 37, 1851. 
2 Pogg. Ann. 101, p. 338, 1857. ; 
1890. P 
