52 president's address — section a. 



It is when we pass on to the consideration of cells of less common 

 type, those, namely, in which the metals are not in contact with solutions 

 of their own salts, that the arguments of Nernst's opponents begin to 

 have something like force. Their objections practically amount to three. 

 I may say at once that I regard one of these as probably unimportant 

 and another as merely demanding a revision of the theory on an un- 

 essential point, but the third as proving that the theory lacks generality. 



The first objection may be stated thus : — The electromotive force 

 between a metal and a solution of an electrolyte which is not one of 

 its own salts depends on the nature of the anion ; the solution pressure 

 of the metal must, therefore, do the same. This has been proved by 

 Paschen (u) and, independently of him, by Bancroft, {v) to be true in 

 several important cases, though not in all, or even the majority ; in 

 most cases it is, as Nernst assumed, untrue. Taylor (w) has supplied 

 an explanation of this difficulty, by showing that, when there is any 

 tendency to form double salts at the common boundary of two electro- 

 lytes, the potential differences neglected in deducing the formula 

 become very large, (x) and suggests that this explanation may be 

 general. I cannot find that he has proved it to be so ; but it is at 

 least a suspicious circumstance that the metals which behave abnormally 

 are actually the ones — such as platinum — which do tend to form double 

 salts. Taken by itself, this difficulty does not seem to amount to much. 



The second objection to the theory is more serious. It is to the 

 following effect : — The potential difference between a metal and an 

 electrolyte depends on the nature of the cations of the electrolyte — 

 whether they be identical with the metal's own ions or not ; further, 

 if we add to a sokition of one of the metal's own salts a salt of some 

 other metal with the same anion, we shall at once alter the potential 

 difference between the metal and the solution, and therewith change 

 the computed solution pressure. This statement Kahlenberg (y) has 

 unquestionably demonstrated for a large number of solutions, aqueous 

 and otherwise. It is true that Bancroft (z) had previously asserted 

 the reverse ; but it must be remembered that his calculations are based 

 on electromotive force measurements by different observers, which 

 may not, therefore, be fairly comparable with each other, to say nothing 

 of the fact that some of them hardly seem to be exact enough to settle 

 a question of this kind. Kahlenberg, to whom the very word ion is 

 anathema, sees in his own work the downfall of Nernst's theory, and a 

 strong argument against the hypothesis of ionic dissociation. But I am 

 not prepared to follow him in this contention, as in my judgment the 

 whole difficulty is due to an assumption, which Nernst certainly made, 

 but which can, I think, be sho^vn to be needless, arbitrary, and not 

 prescribed by the ionic hypothesis. 



Though there seem to be two cases for consideration, they can be 

 treated together by asking the question — Is the potential difference 



(u) Wied. Ann., XLTII. (v) Phys. Rev., III. 



(w) Journ. Phys. Chem., I. 

 {x) The explanation is the same as that suggested above for the case of the in- 

 soluble electrolyte ; but the reason for it is quite different. 

 {y) Journ. Phys. Chem., III., IV. (2) Loc. cit. 



