726 ANNUAL, REPORT SMITHSONIAN INSTITUTION, 192 8 



But since i is much greater than unity, the natural explanation is 

 to say that NaCl is partially dissociated, just as we say that at 

 high temperatures L is dissociated. Now this assumption might 

 be deemed very rash, were it not that on other grounds we are led to 

 look upon electrolytes as partially dissociated, for we assume that 

 they decompose into their ions. But as these ions are charged with 

 very great quantities of electricity of opposite sign, conditions are 

 such that we can not in all cases treat a solution of NaCl as if it 

 simply consisted of Na and CI. The pressure on the walls can not, 

 however, be appreciably affected, so that in this case the solution 

 acts as if Na and CI were free. And when we consider which sub- 

 stances (according to Kaoult's experiments) are abnormal, it is not 

 the inorganic (e. g., not HgCL, CO,, HgS, etc.), but the electrolytic 

 substances (i. e., substances which are conductors of the same order 

 as salts) even when they are organic, e. g., oxalic acid. Trichloro- 

 acetic acid and sulphonic acids must show this still more clearly when 

 thej^ come to be investigated. Since according to the above assump- 

 tion electrolytes decompose into their ions, the coefficient i must lie 

 between unity and the number of the ions. This in reality holds 

 good; for example, the coefficient nearly reaches 2 for NaCI, KCl, 

 KNO„ NaOH, etc., which have two ions; for Ba(0H)2, CaCL 

 K0SO4, etc., which have three ions, it almost approaches 3, and so 

 on. . . . From the above assumption we can even calculate the value 

 of i from the conductivity, and this I shall probably soon carry out ; 

 till now time has failed me. What I called in my paper ' Sur la con- 

 ductibilite ' active molecules, are thus the same as dissociated mole- 

 cules. One of the propositions which I then put forward would 

 now be written : In all probability all electrolytes are completely dis- 

 sociated at the most extreme dilution." Here we have the first ap- 

 pearance of the theory of electrolytic dissociation. If we can not fix 

 its birthday, at least we can its birth month. It is clear, definite, 

 and concise, and all Arrhenius's previous theoretical treatment can 

 easily be translated into terms of it. Van 't Hoff accorded the new 

 idea a favorable reception. He replied on April 7 : " Your statement 

 that the number of ions roughly keeps pace with the values of /, 

 and that the conductivity also increases with z, agrees with most of 

 the cases known to me." . . . He remarks that he had always thought 

 of the dissociation into ions as being confined to an extremely small 

 portion of the salt, but confesses that he sees no grave difficulty in 

 assuming a greater dissociation. Arrhenius in a letter dated April 

 13 is pleased to learn that this is van 't Hoff's view, and states that 

 Emil Fischer, with whom he had discussed the matter in Wiirzburg, 

 although he was friendly to the idea, was of opinion that most 

 chemists would be opposed to such far-reaching dissociation. Arr- 

 henius continues : " It is true that Clausius had only assumed that a 



