PRESIDENTIAL ADDRESS. 437 
tenets of the dissoviation hypothesis be more often than not one which is 
inadmissible, little is gained by applying the speculation quantitatively. As 
already remarked, the only cases in which chemical and electrolytic activity 
can be compared by the methods proposed are those in which the comparison 
is made between solutions of comparable or equivalent strength—that is 
to say, between compounds arranged in vertical series in the order of their 
activity. Electrolytes are comparable in most, if not in all, their pro- 
perties when the comparison is made in this way; but order of activity is 
one thing, actual activity another. It is in this sense and this sense only 
that we may agree with Arrhenius in his statement, ‘ L’activité électro- 
lytique se confond avec l’ activité chimique.’ 
The ionic dissociation hypothesis is a beautiful mare’s-nest, which fails 
apparently to fit the facts whenever it is examined. ‘And the moral of 
that is,’ to quote the words of the classical Duchess so well known to 
children, we must not use the words ion and ionisation in any speculative 
sense but confine their application to cases such as were contemplated by 
Faraday when he introduced the term ion; the conception of activity, 
whether electrolytic or chemical, should alone be attached to such words; 
no idea of actual, separate, individual existence should enter into our 
minds in using them: the ion is to be thought of merely as the potentially 
active, transferable radicle in a compound, not as a separated particle 
enjoying independent existence. It is so easy to speak of dissociation when 
it is desired to give expression to the idea; the first thing the scientific 
speaker or writer should guard against is ambiguity. 
The subject of gaseous interchanges must not be left out of account, 
although it is impossible to do justice to it. Mendelejefi’s contention that 
gaseous interchanges are usually bimolecular has been defended by Dixon 
and Larmor of late. But the facts must be faced. The almost incon- 
ceivable frequency of the molecular impacts must not be forgotten. The 
extraordinary attractive power of the hydrone molecule is also to be 
remembered: this would tend to promote the formation of aggregates with 
which the necessary third substance would every now and then form a 
bimolecular system—which, however, would in reality be at least trimolecular. 
The proportion of hydrone molecules in a dried gas has probably been 
under-estimated—the density of hydrone being very low (9)—as no dehy- 
drating agent can be supposed to remove all such molecules or even nearly 
all; the hydrated substance must have a certain pressure of dissociation. 
Sir James Dewar’s appears to be the only method which is in any way 
deserving of the epithet absolute; the results he has obtained with helium 
in a radiometer are strongly in favour of my view. Lastly, the gradual 
growth in velocity of the explosive wave up to the point of detonation as 
the compression becomes greater is clear indication that reduction in volume 
and increase of opportunity for the formation of systems of the proper 
degree of complexity is a matter of great consequence. Even the behaviour 
of cordite is significant, particularly the projection of unburnt rodlets of 
the material from the gun: apparently it is not decomposed by shock 
intramolecularly but is broken down by heat into gases which interact 
explosively. 
Having dealt with the subjects of chemical change and the nature of 
* Solutions of acids and alkalies have maximum conducting power at certain 
relatively high degrees of concentration. Hydrolytic activity also increases 
steadily in the case of acids as the solution becomes more concentrated ; whether it 
attains to a maximum and whether this coincides with the conductivity maximum 
is uncertain at present; it is very difficult to decide this point experimentally, as 
the rate of change is so rapid in strong solution; moreover, the action takes 
another course in strong solutions, as compounds are formed by the interaction 
of the hydrolyte and hydrolyst, so that two changes are superposed which cannot 
well be followed separately. 
