312 EEPORT— 1886. 



although I have not had the power of setting forth with sufficient clearness what 

 is abundantly evident to myself. 



With assurances that I shall be grateful to learn youi- views, &c., &e. 



SVANTE ArKHENIUS. 



EiGA, June 8, 1886. 



Dear Sir, — ITianks for note and British Association circulars. Since you speak 

 of communicating something of my views to the Committee, I will/to avoid 

 misunderstanding, just try in a few lines to explain my position. In my small 

 notice on the conductivity of gelatinous solutions, I am led to the view that internal 

 friction (viscosity) e.xerts no influence on conductivity. By experiments on the 

 conductivity of mixtures, with which I am now working, it appears, however, that 

 gelatinous solutions (and probably other pseudo-solutions) form an exception; since 

 for other (actual) solutions a very close relation exists between conductivity and 

 limpidity. It seems as if, when one adds to water a liquid, or in general any 

 foreign body (with the exception of all the best conducting salts), the limpidity of 

 the solution becomes less than that of pure water, no matter whether tlie limpidity 

 of the added body be less or greater thaji the water. It appears as though the 

 friction experienced by a molecule travelling through a liquid greatly depends on 

 the heterogeneity of the liquid. One could propel a water molecule more easily 

 through pure water than through water with which some other substance had been 

 mixed. But a water molecule consists of the ions H and OH, and what is valid 

 for the wliole molecule must be valid also for a part of it. If this is correct, the 

 advance of an ion through a liquid whose molecules have this ion as a constituent 

 must be opposed by a smaller resistance (friction) than if the molecules of the 

 liquid had not, or only partly had, this ion as a constituent. This .serves as a sort 

 of explanation of the greater velocity of the ions H and OH in aqueous solutions 

 in comparison with the velocity of other ions. I regard it as certain that the ions 

 H and OH travel quicker than other ions, which go at a pace pretty nearly equal 

 (not quite equal) among themselves. But the influence of heteroEreneity on the 

 internal friction and also on conductivity appears to diminish with increasing 

 temperature. So it is very possible, and indeed probable, that at some temperature, 

 higher than any hitherto employed in such observations, the ideal case might be 

 reached when all ions should go at the same rate — which would mean that at that 

 temperature all electrolytes, in extremely dilute solutions, should conduct equally 

 well ; just as a gas's obedience to Boyle's law approaches exactitude at low 

 pressures and high temperatures. 



This will probably be the tendency of the results of my not yet concluded 

 investigation on this interesting subject. It is impossible for me to give you the 

 experimental proof for the view above expressed, though I should have had much 

 pleasure in doing so. I should be obliged by your sending me the results and 

 report of the Committee, and on my side I will willingly, if you regard this as not 

 wholly without value, communicate to the Committee the results of my experiments 

 now being carried on. 



Professor Ostwald is sending you some of his papers. 



Tom's &c., 



De. Svaxte Arehenips. 



On the Accuracy of Ohm's Law in Electrolytes. 

 By Professor G. F. Fitzgerald, F.B.S., and Mr. Teouton. 



Some preliminary experiments were begun in the spring of this year, and have 

 been carried on from time to time since, with a view to determining how far Ohm's 

 law may be relied on in the case of electrolytes. 



Though as j-et no very high limits of accuracy in the experiments have been 

 attained, owing to causes which will be later on explained, still it is hoped 

 that from the experience already gained most of the difficulties have been got over, 

 and that soon results of a much greater degree of accuracy wiU be arrived at. 



