546 Proceedings of the Koyal Society of Edinburgh. [Sess. 
The difference between the last two values of A in the table is less than the 
error of observation, which in this case is greater than usual owing to the 
solution being so dilute and the wave-length being so far in the violet. A is 
therefore constant for small values of c, but increases rapidly as the solution 
approaches saturation. The change has thus apparently nothing to do 
with ionisation. The question of the cause of the change, together with the 
similar colour change of the cobalt salts, is being held over for more 
exhaustive investigation. 
The effect of temperature upon very dilute solutions of the sulphate and 
chloride has already been investigated with a spectrophotometer by Nichols 
and Spencer.* No change was found in the case of the sulphate, but in the 
case of the chloride in the middle of the spectrum a permanent increase of 
absorption was obtained. Their original value of A is, however, too high ; 
possibly the solution was slightly turbid to begin with, and this turbidity 
was increased permanently by the heating. 
* E. L. Nichols and Mary C. Spencer, “ The Influence of Temperature upon the Trans- 
parency of Solutions,” Phys. Rev., ii. p. 344 (1895). 
{Issued separately July 8, 1911.) 
