718 ANNUAL REPORT SMITHSONIAN" INSTITUTION^ 192 8 



hypothesis of Clausius on continual momentary separation of ions. 

 Hittorf's work on migration, Helmholtz's conception of the atomic 

 nature of electricity, the work of Kohlrausch on conductivity, were 

 all leading up to some definite comprehensive theory which in the 

 end was furnished by Arrhenius. 



Arrhenius completed his experimental work in the spiing of 1883 

 and wrote the theoretical part at hi,s home in the summer of the 

 same year. The memoir containing the results of his conductivity 

 experiments, and the conclusions he deduced from them was sub- 

 mitted to the Swedish Academy of Sciences in June, 1883, and pub- 

 lished in the following year. (Bihang, vol. 8, Nos. 13 and 14.) It is 

 in French and is entitled " Investigations on the Galvanic Conduc- 

 tivity of Electrolytes. Part I. Determination of the conductivity of 

 extremely dilute solutions by means of the' depolarizer; Part II. 

 Chemical theory of electrolytes." Arrhenius undertook the experi- 

 mental investigation of dilute solutions himself, for, although Kohl- 

 rausch had made similar measurements and had quoted some numeri- 

 cal data, the final publication of his results was delayed till 1885. 

 The depolarizer which Arrhenius used was an apparatus devised by 

 Edlund in 1875, and corresponds roughly to a hand-driven rotating 

 commutator. It is of interest to note that the conductivity cell which 

 bears Arrhenius's name is described in this paper. 



Arrhenius measured the resistance of a considerable number of 

 salts, acids, and bases at various dilutions, sometimes a,s high as 

 'y = 10,000. Unfortunately the actual dilutions are not given, so that 

 it is difficult to correlate the data of Arrhenius with those of other 

 authors. He tabulated his results so as to show in what ratio the 

 resistance of an electroljrte is increased when the dilution is doubled. 

 This ratio, as Kohlrausch had found earlier, is nearly equal to 2 for 

 most salts, i. e., specific conductivity is nearly proportional to con- 

 centration. Departure from this ideal value he took as a basis for 

 classification of the dissolved electrolytes, and showed that chemi- 

 cally similar substances fell into the same category when classified 

 according to dilution ratios. A discussion of the data led Arrhenius 

 to the conclusion that " if on dilution of a solution the conductivity 

 does not change proportionally to the amount of electrolyte, then a 

 chemical change has occurred on addition of the solvent." He exem- 

 plifies this by the consideration of potassium cyanide with the ab- 

 normally high dilution ratio of 2.14, which he attributes to the 

 partial splitting of the salt into acid and base on dilution with water. 

 The abnormal value,s obtained for soluble hydroxides and dilute solu- 

 tions of acids he attributes to the presence of small quantities of 

 ammonium carbonate in the solvent water. 



The importance of this paper, however, does not lie in the experi- 

 mental measurements or in the detailed deductions, but in the general 



