392 



REPORT — 1886. 



may indeed turn ont to be proportional to m, the amount of salt added, 



but it is hardly likely to be exactly equal to it. 



Moreover, if two substances are active, we shall have to split up u into 



«! and U2, V into Vi and v^, as well as N into N, and Ng. 



N 

 If TO is the mass of salt contained in every gramme of solution, — ! is 



m 



the ' dissociation-ratio ' or ' activity-coefficient ' * of the salt, and 



N, 



-the 



1 — m 



dissociation-ratio of the water. And equation (1), written out fully, 



becomes — 



dv 

 ^i (^i+^i)+N2 (W2+V2)=A;>; -rr, • • • (3). 



The value of ^ to be used here is the simple straightforward conductivity of 

 the whole solution. There is to be no subtracting of hypothetical water- 

 conductivity in order to get at an inaccessible salt- conductivity ; the dis- 

 tinction between known and unknown quantities is perfectly definite. 



By a complete quantitative electrolytic experiment, such as Hittorf 

 first made, and Wiedemann a^d others have followed up, the four quantities 

 NjM,, NiiJ], N2W2J ^2^2 can all be obtained ; but it is not possible thus to 

 obtain the values u^, v,, u^, v^, separately, unless the dissociation-ratios 

 Ni/m and N2/I — m are known too. 



The mode in which I have begun to make complete determinations of 

 electrolysis may be stated for the case of copper sulphate. The cathode 

 vessel is in the form of a specific gravity flask with a long horizontal tube 

 Fig. 1. neck, which has a con- 



striction at one place, to 

 which it can be accu- 

 rately fiUed, and an open 

 mouth above the con- 

 striction for the anode 

 vessel to dip its beak 

 into. The cathode can 

 be a piece of platinum 

 fused through the glass, 

 or, more conveniently, 

 passed through a stop- 

 per. The anode vessel 

 may be like a tobacco-pipe with the anode immersed in its bowl, and its 

 stem recurved so as to dip into the mouth of the cathode vessel. Only one 

 of the two vessels is to have measurements made upon it, and the cathode 

 vessel is perhaps generally the more convenient. It can be kept immersed 

 up to its neck in a bath of water at constant temperature, as in fig. 10. 



The course of the experiment is as follows : — Fill the cathode vessel 

 with a standard solution, adjusting its level exactly to the mark at a 

 known temperature ; then weigh it. Pour a little more solution into its 

 mouth. Fill the anode vessel also with the same solution ; arrange as 

 suggested in figure 1, and pass a measured current for a known time, 

 with a silver voltameter in circuit. Then remove the anode vessel, re- 

 adjust the level in the other, and weigh again. Finally analyse the contents 

 of the solution in the cathode vessel, and weigh the cathode deposit or 

 voltameter plate. The necessary and sufficient data are these : — 

 ' See memoir of Arrhenius below, p. 364. 



f Full Size 



