388 
Proceedings of the Royal Society of Edinburgh. [Sess. 
Table II. 
Calculated 
E.M.F. 
Observed 
E.M.F. 
•344 molecules potassium iodide in 1000 c.c. 
•05 molecules of iodine in 1000 c.c. against -005 molecules 
•0341 
•0326 
•23 molecules potassium iodide in 1000 c.c. 
•05 molecules of iodine in 1000 c.c. against *005 molecules 
•0371 
•0352 
T15 molecules potassium iodide in 1000 c.c. 
*05 molecules of iodine in 1000 c.c. against '005 molecules 
•0471 
•0444 
It is evident from these results that the actual ratios between iodine and 
potassium iodide in these solutions differ sensibly from those calculated 
from the formula = '0014, although the law is being approximately 
ixi 3 
followed. 
It might be suggested that these solutions are too strong to obey the 
gas laws accurately. I therefore took the solutions ‘05 iodine, ‘115 KI, and 
*005 iodine and ’115 KI, and diluted them both to one-tenth the strength. 
The E.M.F. readings were not sensibly affected by this dilution. 
Jakowkin shows that for stronger solutions of potassium iodide than 
those so far considered the constant alters for the stronger concentration of 
iodine, probably owing to the formation of sensible quantities of higher 
complexes. It seemed, therefore, of interest to see how far the E.M.F. for 
such solutions would agree with those calculated from the same equations, 
but allowing for the changing constant. In the table selected Jakowkin 
makes the amount of potassium iodide normal, and the amount of iodine 
ranges from '5066 molecules per litre to '0141 molecules per litre, the 
constant changing from '000773 to '001365. If, then, the weakest solution 
of iodine, namely, '0141, be taken for one electrode, and the other stronger 
solutions be taken in turn for the other electrode, the electromotive force 
can be calculated by applying the appropriate constant to each solution. 
The table on p. 389 gives the results of these calculations and experiments. 
Here also it will be noted that the results are very close where the 
amount of potassium iodide is in large excess, but only approximately for 
the other solutions. 
It seemed of interest to investigate the question whether KI S was 
stable at higher temperatures, and this can be done as follows : — 
If we consider the cell '115 KI, 05 I 2 against '115 KI, '0005 I 2 , it is evident 
