216 REPORT—1890. 
per cent., and that of acetic acid only 1 per cent., HgCl, only 8 per cent., 
is not what one would expect @ priori ; but the general agreement of the 
results is so close that it can hardly be explained away. The theory is 
further supported in Arrhenius’s original paper by the consideration of a 
number of properties which are additive in dilute solutions; that is to say, 
the numerical values of these properties can be regarded as the sums of 
the values corresponding to separate parts, namely, the solvent, and the 
component ions into which the molecules of the salt are separated. A 
well-known example is that of electric conductivity,! which, for a very 
dilute solution, can be numerically regarded as made up of numbers 
corresponding respectively to the solvent and the several ions. 
The other properties of dilute solutions which Arrhenius mentions in 
this connection are the heats of neutralisation,” specific gravity and spe- 
cific volume,’ specific refractive power,’ depression of the freezing-point * 
and other properties connected with it, diminution of vapour pressure, 
osmotic pressure, and isotonic coefficient.” These additive properties 
have of themselves suggested the more or less complete dissociation of 
salts.© Perhaps the most striking corroboration of Arrhenius’s theory is 
that the cases in which the additive law is not satisfactorily made out, 
are precisely the cases in which the dissociation ratios deduced from the 
resistance measurements are considerably less than unity, even in dilute 
solutions. 
Against this formidable array of reasons in favour of the dissociation 
hypothesis, Armstrong’ has urged a number of considerations, among 
which are the following: There are difficulties from the chemist’s point 
of view, which dispose him to reject the idea that electrolysis is primarily 
an affair of atoms ; ‘ peculiarities and relationships which are patent to the 
chemist,’ but which ‘it is impossible at present to quantify.’ Moreover, 
it seems to be difficult to accept the idea that an electrolyte can be decom- 
posed by an infinitesimal electromotive force unless further proof is forth- 
coming ;*® and, again, there are anomalies that the dissociation theory 
does not explain, as, for instance, the conductivity of fused silver iodide in 
face of the non-conductivity of water and of pure hydrochloric acid, the dis- 
sociation of hydrochloric acid by water without a corresponding dissocia- 
tion of the water, and the more complete dissociation of what have always 
been regarded as the more stable compounds. The parallelism of diffu- 
mation of double molecules, even in dilute solutions, and in the case of CaCl, on 
account of the formation of CaCl (Van ’t Hoff and Reicher). 
at 
Those cases in which the ratio 7s considerably less than 1 in strong solution 
can be explained by ascertaining the formation of double molecules in the stronger 
solutions, 
1 Kohlrausch, Wied. Ann. 6, ps 167 (1879); 26, pp. 215, 216 (1885); Ostwald, 
Aeitsehr. fiir ph. Chem. 1, pp. 74 and 97 (1887). 
2 Ostwald, Lehrbuch der allgemeinen Chemie, p. 1250; Arrhenius, /.c. p. 643. 
3 Valson, C.R. 73, p. 441 (1871); Ostwald, Lehrbuch, i. p. 384. 
4 Raoult, Ann. d. Ch. et d. Phys. [6] 4,p. 401 (1885). 
5 De Vries, Pringsheim’s Jahrbiicher fiir wiss. Bot. 14, p. 519 (1883). 
® Valson, C.R. 73, p. 441 (1871); 74, p. 103 (1872), 75, p. 1330 (1872); Raoult, 
Ann. de Chim. [6] 4, 401, 426. 
7 Proc. Roy. Soc. 1886, p. 268 ; Electrician, Aug. 26, 1887. 
8 See a paper by Ostwald and Nernst, Zitschr. fiir ph. Chem. 3, p. 120, 1889, ‘On 
Free Ions,’ in which it is shown, on the assumption that the energy developed by the 
discharge of a conductor ina liquid is proportional to the square of the loss of elec- 
tricity, that no work is done by the electromotive force in separating the molecules 
into ions. 
