1910-11.] Temperature Coefficient of Concentration Cells. 395 
available. Calorimetric experiments might therefore enable us, along 
with the E.M.F. measurements, and ionic velocities to calculate the total 
amount of water plus alcohol carried by the ions. 
In the case of two liquids that do not mix, the ion has to part with 
its combined liquid at the margin of contact and take up the new liquid 
into combination. If we imagine the case of a cell with liquids that do not 
mix and in which the dilutions are sufficient to give complete ionisation on 
both sides, then the only phenomenon with which we are apparently dealing 
is the heat absorbed or set free by the combinations of the ions with the 
solvents. In the case of water and alcohol and KI, when iodine ions are 
travelling from water to alcohol and the K ions from alcohol to water, 
the summation of these combinations results in the absorption of heat. 
An approximate calculation of this heat absorption can be made from the 
measurement of the E.M.F. and the temperature coefficient of the cell in 
which *001 molecules of KI in alcohol was measured against ‘001 molecules 
of KI in water (with the assumption that the velocity of the potassium and 
iodine ions in alcohol is practically the same, just as it is practically the 
same in water). The result is to show total heat absorption of 9000 
calories. Thomsen gives the latent heat of solution of KI in water at 
5100 calories, but it must be remembered that he is here measuring the 
nett result, from which the heat of ionisation has to be subtracted. While 
I do not wish to lay much weight on this calculation, it certainly suggests 
that the heats involved in the very dilute solutions are much higher than 
those obtained in the ordinary way in the calorimeter. When both 
solutions are so dilute as to be completely ionised, then the partition 
coefficients are ionic partition coefficients ; and if K x and K 2 are the ionic 
partition coefficients at temperatures T x and T 2 , the general Van ’t Hoff 
equation, of which the one already used is a particular case, would surely 
apply to these solutions ; and if X is the nett result of the heats absorbed 
or set free by the combination of the respective ions with the two solvents, 
then the value of X can be obtained from the general Van ’t Hoff equation, 
log — ! = A ( J_ _ JL\ 
b r vt x t J 
To sum up, the following, as far as I am aware, new results have 
been obtained : — 
First, the demonstration of the effects of change of temperature on these 
cells. 
Second, the demonstration of the connection between the latent heats 
of solution and the observed results. 
Third, the demonstration that the condition of electrical equilibrium, 
