56 . president's address — section a. 



It follows, from the foregoing discussion, that the electrolytic 

 solution pressure must not be identified with the osmotic pressure of 

 the positive ions within the metal. Though the behaviour of aqueous 

 solutions tends to suggest that this osmotic pressure is one of its factors, 

 there is evidently another factor due to the nature of the solvent and, 

 moreover, to the unionised part of the solvent, for we have shown that 

 the ionised part cannot account for the experimental facts. But, we 

 may be asked, if the unionised part of the sohevt has anything to do 

 with the matter, why not the unionised part of the meial also ? Why, 

 in a word, need we invoke the agency of the ions at all in order to 

 account for the solution of a metal ? It is much easier to ask such a 

 question than to answer it. We must, however, remember that ionisation 

 — whatever we may think of ionic dissociation — is an experimental fact. 

 Ions are material bodies which we can weigh and measure ; we can even 

 follow their movements in the voltameter. Ionic partial pressure — 

 whatever we think of the expression — is the name assigned to a 

 measurable physical stress. Since the presence of these bodies and the 

 application of this stress at a metal-liquid boundary are simple facts — 

 whether the hypothesis to which they owe theii" names be true or false 

 — it seems hardly reasonable to deny them a share, at least, in the 

 processes of metallic solution ; I am inclined, therefore, to think that 

 translation into a less hypothetical language and an extension of scope, 

 rather than annihilation, is the future which awaits Nernst's theory. 

 For the present we may sa}^ that the idea of solution pressure is a very 

 valuable one, provided we do not attempt to define it too strictly — 

 provided, i.e., that we give up Nernst's attempt to unify the theory 

 of electromotive force in terms of the ionic hypothesis alone. 



We must, then, abandon the idea of a table of solution pressures 

 and of a voltaic series which is the same under all conditions, and be 

 prepared to recognise that the voltaic series depends on the nature of 

 the solvent used— to admit, in short, a different electrochemistry for 

 every solvent, just as we were driven, 20 years ago, to admit a different 

 voltaic series for every gaseous medium. There is, after all, nothing 

 strange about this. It is no more unreasonable to ascribe to a metal 

 as many solution pressures as there are solvents than to do the same 

 thing for a non-metallic solute, which in fact everyone does. Of 

 course, this amounts to an acceptance of Kahlenberg's position — how- 

 ever he arrived at it — that the relative chemical affinities of the metals 

 are functions of their environment as well as of their own nature. All 

 the same, we have not yet heard t'le last of Nernst's theory. It is still 

 as capable as ever of acting as a guide to research, for the next great 

 step to be taken must be the studv of the electromotive force relations 

 of solvents other than water, in order to see whether Nernst's formulae 

 are equally applicable to them. It seems not unlikoly that a comparison 

 of such results, derived from a large number of solvents, may be the 

 necessary preliminary to the establishment of an adequate general 

 theory alike of electromotive force and of chemical affinity. But we 

 must keep an open mind on the subject, as it may be that the path to 

 such a theory lies in another direction altogether ; as old Homer was 



so fond of remarking, " -ravta Qewv eV ^ovvaat K6?Tat." 



