Vol.. 7, 1921 
CHEMISTRY: J. KENDALL 
57 
Numerous attempts towards reconciliation having failed, the tendency- 
is now wide-spread either to deny the applicability of the law of mass 
action to the ionization of strong electrolytes or to modify the ionization 
equation itself. 
Complete ionization of strong electrolytes, with an electrostatic equilib- 
rium between "free" and "bound" ions, has recently been urged by 
Milner, Ghosh and others. The existence of inter-ionic attractive forces 
and the necessity for recognizing their effect in any accurate ionization 
formula cannot be questioned, yet there are several points of difficulty 
which the Milner-Ghosh conception still leaves unexplained. Among 
these may be mentioned the lack of a definite gap between strong and weak 
electrolytes and the fact that the heat of ionization (as ordinarily defined) 
does not agree at all with the work required for the separation of the charged 
radicals. It is quite possible, indeed, that the concordance obtained by 
Ghosh between experimental and calculated "ionization" values may 
ultimately prove to be due rather to convenience of constants than to the 
fundamental validity of the equations developed. Full discussion of 
these points must be reserved for a future article. A still more serious 
objection, however (which applies also to the original hypothesis of 
Arrhenius), is the neglect of the rdle of the solvent in ionization. 
The essential equivalence of solvent and solute in ionization equilibria 
has been elaborated upon in a previous paper. 1 We take two liquids; 
for example, water and (liquefied) hydrogen chloride. Both are highly 
associated, both have high dielectric constants, both possess practically 
zero conductivity when perfectly pure. We mix them and obtain a 
solution with very high conductivity, which we ascribe to the dissociation 
of the hydrogen chloride into its constituent ions. Is there really any 
reason why the conductance should be referred entirely to hydrogen 
chloride and not at all to water? The active part played by the solvent 
in ionization is unconsciously recognized by our admission that ions are 
"hydrated" in aqueous solution, since (for example) a "monohydrated 
chlorine ion" (C1,H 2 0)~ might just as conceivably deserve the alternative 
title of a "solvated hydroxyl ion" (HCl,OH)~. A scheme according 
to which water is the only "ionizing" substance in all aqueous solutions 
would, in point of fact, be quite as logical (and certainly as simple) as the 
scheme at present employed, which relegates water to the inferior roles of 
"dead space" and "envelopes." Much more satisfactory than either, 
however, would be a theory taking both components of a conducting 
solution equally into account. 
Considerable experimental work, upon which such a theory may be 
based, has been performed by the author and his collaborators 2 during 
the past seven years. The main results and the conclusions to which they 
lead are summarized below. 
At the beginning of the work, it was felt that the most familiar case of 
