382 
Proceedings of the Royal Society of Edinburgh. [Sess. 
thermostat at 25 c C. and allowed to separate, and introduced into their 
respective electrodes. 
No E.M.F. could now be detected in the electrometer in accord with 
Luther and Abel’s theory. 
This cell was now raised to 30°, and at once a positive E.M.F. was 
developed, and in cooling to 12 *5 a negative E.M.F. was developed. 
In each case these E.M.F.s amounted to some one or two hundredths 
of a volt. 
This cell is evidently therefore the same in its temperature reactions 
as the alcohol cell already described. That is, when the E.M.F. is reversed 
from positive to negative the cell has a negative temperature coefficient 
and becomes exothermic. 
Experiments made with alcohol water cells with potassium iodide in 
solution and silver silver iodide electrodes gave the same results. When the 
solutions were of equal strength, a considerable E.M.F. was obtained in 
the direction of transferring potassium iodide from alcohol to water, and 
the cell was endothermic. On increasing the concentrations of the water 
solution until the E.M.F. was reversed, the temperature coefficient became 
negative and the cell exothermic. 
The experiments with these potassium iodide cells will presently be more 
particularly described, as the attempt was made to get certain quantitative 
measurements with them ; but in the meantime the above result is sufficient 
for our purpose. 
In the following mathematical investigation, although the alcohol cell 
is being used as an illustration, the nitrobenzene cell is a more strictly 
typical cell for our purpose, as in the alcohol cell we have the complication 
of the interdiffusing alcohol, setting free heat and precipitating salt from 
solution. The effect of this on the results will ultimately have to be 
considered. 
In these cells, then, we have a very neat form of heat engine. If the 
salt is in partition equilibrium at a particular temperature, then on raising 
the temperature of the cell above this amount an E.M.F. is developed 
and heat is absorbed. On now cooling below the neutral temperature, 
the current is reversed, the chemical changes are replaced, and heat is 
set free. 
Mr Ernest Gibson, in a letter to me, has developed the theory of these 
cells directly from Abel’s equation in the following manner : — 
Taking the Abel equation in the following simple form, 
E = ET | log - log A- 1 = 0 (1) 
