390 
Proceedings of the Royal Society of Edinburgh. [Sess. 
sociation of potassium iodide in alcohol of different strengths of solution. 
As we have neither of these values with certainty, it is not possible to 
calculate out the values of the respective E.M.F.s for alcohol as was done 
for water. These matters may be treated of in a subsequent paper ; in the 
meantime I confine myself to testing how accurately the osmotic pressure 
equation for E.M.F. applies to alcoholic solutions of iodine, and whether the 
E.M.F.s obtained indicate the formation of KI 3 in the solutions. 
It is evident that, as in the case of water, if the potassium iodide is 
present in great excess the E.M.F. should be calculable from the iodine 
concentrations alone, and that, as the amount of iodine and potassium 
iodide become sensibly the same on one side of the cell, the E.M.F. should be 
higher than that due to the iodine concentrations alone if KI 3 is formed 
in the solutions. Moreover, if the ratio between iodine and KI is made 
sensibly the same as that formerly employed, we should be justified in 
calculating the E.M.F. of the cell with the largest ratio of iodine to KI on 
both sides, from the iodine concentrations alone. 
Four independent sets of solutions were made of potassium iodide and 
iodine in alcohol. The alcohol used was Kahlbaum’s absolute alcohol, 
which contains some *2 per cent, or '3 per cent, of water ; and while care 
was taken to expose the solutions as little as possible to the air while 
being prepared, yet probably traces of water were absorbed. When once 
the solutions had been transferred to the little stoppered electrodes 
immersed in absolute alcohol, no further absorption of water was to be 
feared. The probable errors introduced by traces of water are discussed at 
the end of the paper. The following were the results obtained at 25° C. : — 
Number of Iodine 
Molecules in 1000 c.c. 
E.M.F. Reading for 
Solutions 
Mean 
E.M.F. 
a. 
b. 
c. 
d. 
Solution containing '075 mole- 
( -01 against *001 
*0356 
•0365 
•0356 
•0353 
•0358 
cules of KI in 1000 c.c. 
\ *01 „ -0001 
•0652 
•0661 
•0653 
•0649 
•0654 
Solution containing *05 mole- 
( *01 against *001 
•0393 
•0391 
•0384 
•0388 
•0389 
cules of KI in 1000 c.c. 
t *01 „ *0001 
•0695 
•0689 
•0691 
•0684 
•0690 
Solution containing *025 mole- 
/ "01 against *001 
•0505 
•0499 
•0499 
•0501 
cules of KI in 1000 c.c. 
\ -01 „ -0001 
•0802 
•0812 
•0806 
•0807 
If we subtract the values obtained for *01 against *001 from those 
obtained for *01 against *0001, and arrange the resulting figures beginning 
with those containing the largest proportion of KI to I 2 , we get the 
following table, corresponding to the one obtained for water solutions : — 
