362 EEPOET— 1886. 



abstract of them would repay perusal, inasmuch as the author allows himself 

 rather too freely the use of hypothesis concerning wholly unknown molecular 

 interactions.' Perhaps it wUl he fairer if I give the resume in full. 



Resume. 



' In the first six sections of the present work we have described a new method 

 of measuring the resistance of electrolytic conductors. In this method we made 

 use of rapidly alternating currents, produced by a depolariser constructed for the 

 purpose by M. Edlund. We have tried to show the use of this method, and to 

 make clear the practical advantages which it possesses. 



' In the next part we have treated the process of making the observations, and 

 of calculating the results. Then we have displayed the figures found for very 

 dilute solutions of forty-five different bodies. Finally, we have discussed pre- 

 liminarily the figures obtained, with regard to the exponent of dilution, the mole- 

 cular conductivity, and the temperature-coefficient. 



' In Chapter III. we have, guided by the data of MM. Kohlrausch and Hittorf, 

 laid down the proposition of the proportionality between the conductivity and the 

 number of electrolytic molecules of a dUute solution, as well as two other propo- 

 sitions, according to which the figures of the preceding part were calculated 

 (1) (2) (3). Moreover, we have shown that if these propositions are not applic- 

 able, it is necessary to suppose that by dilution of electrolytic solutions 

 chemical reactions are set up, (41 (5). Proceeding from these different proposi- 

 tions, we have shown that all salts, properly so called, in solution are composed 

 of complex molecules, which are partly destroyed by dilution. We have also 

 indicated the mamier in which these complexes are formed. 



' By aid of this conception the properties of salts at all dilutions have been 

 explained, as well as the properties of all electrolytes at considerable concentration. 



' On the other hand, hydrates,^ and salts which partially transform themselves 

 into hydrates, manifest other properties when much diluted. We have shown that 

 this singularity can be explained by the action of impurities which accompany the 

 water used to dissolve them. 



' By some considerations as to the nature of galvanic resistance we have been 

 brought to the conclusions numbered (7) (8) and (9), of which the two latter com- 

 plete the first : these indicate the long-known relation between gah'anic resistance 

 and internal fiiction. The two propositions (8) and (9) are also in agreement 

 with published data.' 



Thus much is a meagre account of the first part of the complete memoir. I 

 now pass to the far more striking second part, and shall find it necessary to give a 

 much fuller, and in many places verbatim, rejjort. 



I am not able to judge as to how much is original and new, as I am but 

 slightly acquainted with the work of previous writers on similar subjects, nor am I 

 at all confident how far the hypotheses made by the author are perfectly legitimate. 

 So far as I am able to judge, however, and making allowance for possible inade- 

 quacy of data and somewhat hasty generalisation, the paper seems to me to be 

 a distinct step towai-ds a mathematical theory of chemistry. 



The title affixed to it is ' The Chemical Theory of Electrolytes,' but it is a 

 bigger thing than this : it really is an attempt at an electrolytic theory of chemistry. 



PAET II. — Them-ie Chimique des Electrolytes. Par Svante Arrhenitjs. 

 (89 Pages). Abstract and traiislation by Oliver Lodge. 



[Remarks by the abstracter are enclosed in square brackets.] 



§ 1. Ammonia considered as an electrolyte. 



Kohlrausch has shown that a solution of ammonia, as regards conductivity, 

 behaves differently fi'om all other bases. It is a much worse conductor than potash, 



• See § 6 of letter on p. 386. 



^ By hydrates the author alwaysjmeans hydrogen compounds like acids and bases. 



