434 TRANSACTIONS OF SHCTION B. 
change contemplated be one involving development of energy. It is 
important to remember also that a change which could not otherwise take 
place becomes possible when a suitable depolariser is introduced into the 
circuit. 
The evidence that similar considerations apply to the gaseous and the 
liquid states cannot well be gainsaid. Before framing a theory of chemical 
change it is therefore necessary to formulate a definition of an electrolyte. 
It is doubtful if any single substance be an electrolyte; the conductivity of 
fused salts may well be and probably is conditioned by some admixture. 
Aqueous solutions of alkalies, acids, and salts without exception are elec- 
trolytes. Everything points to the fact that im such solutions the solvent 
and solute act reciprocally; the contention that the solute alone is active 
cannot be justified. As water is altogether peculiar in its activity as a 
solvent and is a solvent which gives rise to conducting solutions, an 
explanation of its efficiency must be sought in its own special and peculiar 
properties. 
Since 1886 this conclusion has been impressed upon me with indisput- 
able force, and I have frequently ascribed the effect produced by the one 
constituent upon the other in a solution to the residual affinity of the 
negative elements in the two compounds which act reciprocally. It was 
only recently, however, that I saw my way to postulate a complete theory 
which would serve to account for the properties of solutions and that I 
realised how the reciprocal effect might be produced. 
I would substitute for the-misleading conception that liquids are com- 
parable in their behaviour with gases the idea that the liquid state is one 
in which the residual affinity of the negative elements in particular always 
comes into play and causes the formation of molecular aggregates of various 
degrees of complexity; moreover, that the alteration in the properties of 
any given solvent by the dissolution in it of another substance is largely 
and, in some cases, mainly due to a disturbance of the equilibrium natural 
to the solvent by an alteration in the proportion in which the several 
aggregates are present. The alteration in some particular property pro- 
duced in a given mass of the solvent may, from this point of view, be taken 
as the measure of the activity of a substance, just as the alteration in the 
pressure of a particular volume is taken as the measure of the alteration 
produced in a gas. In the case of non-electrolytes, if only a small amount 
of the solvent be withdrawn by combination with the solute, the alterations 
may be regarded as almost entirely due to the ‘ mechanical’ interference 
of the substance introduced, opportunity being given for the simpler, more 
attractive molecules of the solvent to exist in greater proportion because of 
the diminution of the chance of reuniting which is conditioned by the 
presence of practically inert molecules of another kind; if a more or less 
considerable amount of the solvent become associated with the solute the 
conditions become more complex but similar considerations apply. From 
such a point of view a liquid is rendered more active by the addition of 
any soluble substance. Its vapour pressure is therefore diminished; the 
internal ‘ osmotic’ stresses are raised; its freezing-point is lowered. 
Although it is generally admitted that water is not a uniform substance 
but a mixture of units of different degrees of molecular complexity, the 
degree of complexity and the variety of forms is probably underestimated 
and little or no attention has been paid to the extent to which alterations 
produced by dissolving substances in it may be the outcome and expression 
of changes in the water itself. The attempt to extend the ‘laws’ which 
are applicable to the gaseous state to liquids has led us away from the 
truth by narrowing our conceptions. If the contention be justifiable that 
the alterations attending dissolution are very largely alterations in the 
character of the water, attention has been directed of late far too exclusively 
to the dissolved substance. 
