388 Proceedings of the Royal Society of Edinburgh. [Sess. 
has shown that this silver is quite reliable. The wires were iodised by 
being dipped into a strong solution of iodine in alcohol, washed with 
absolute alcohol and allowed to dry by waving them about in the air. 
The alcohol used was a fresh supply of Kahlbaum’s absolute alcohol. 
The solubility of pure potassium iodide in this alcohol at 25° C. was 
determined as '0955 molecules. For another purpose, Dr Denison had, 
after drying this alcohol over lime and distilling off calcium, determined 
the solubility of KI at 25° C. to be ’0957. This alcohol of Kahlbaum’s 
may therefore be regarded as sufficiently dry for the purposes of these 
experiments. 
Changing the wires about and reiodising them made no perceptible 
change in the E.M.F. 
Every solution before being used was mixed with pure dry silver iodide 
and kept in the thermostat for some hours, with occasional shaking so as 
to saturate it with silver iodide. 
In order to test the method, the following experiments were made : — 
A concentration cell of *01 molecules KI against ‘001 molecules KI with 
the silver iodide electrodes gave an E.M.F. at 25° C. of ‘0552 volts. 
According to Kohlrausch’s latest results, ‘01 KI is ionised 94 per cent. 
Plotting his results and continuing the curve graphically, ‘001 KI would be 
ionised about 96 per cent. Therefore, the E.M.F. of this cell should be 
‘0591, =RT (log ‘0094 — log ’00096). But a small correction has to be 
made in this for the diffusion E.M.F. 
Taking the values of U and V from Whetham’s Theory of Solution , 
we get for the theoretical E.M.F. the value ‘0577. 
Similarly an experiment was made with a ‘I, ‘01 concentration cell. 
The measured E.M.F. was ’0536 and the calculated was ‘0551. It will 
be noted that in both cases the measured E.M.F.s are about ‘002 volts 
below the calculated. This agrees with the results obtained by J ahn * for 
silver chloride, potassium chloride cells. It is not necessary here to discuss 
the conclusions at which Jahn arrives from these discrepancies. It is 
sufficient for our purpose that the readings obtained agree so closely with 
his, and that therefore the whole method is reliable. 
It is now necessary to describe the experiments made with the alcohol, 
water, potassium iodide cells with silver silver iodide electrodes. 
After the preliminary experiments, two solutions were compared con- 
taining each ‘001 molecules of KI and saturated with silver iodide. The 
result of my measurements with this cell was to establish ‘08 volts as the 
most probable E.M.F. Occasionally this would capriciously change to a 
* Zeit. phys. Ghem ., xxxiii. p. 370. 
