II.—ON THE DETERMINATION OF THE FREEZING-POINT DEPRES- 
SIONS OF DILUTE SOLUTIONS OF ELECTROLYTES —BY 
Tos. C. Hess, M. A., Dalhousie College, Hulifax, N.S. 
(Read March 10th, 1902.) 
Dr. MacGregor has shown,* that,if for any electrolyte curves 
be plotted using ionization coefficients as ordinates and the 
equivalent depressions of different observers as abscissae, the 
curves so drawn diverge, as dilution increases, some to the right 
and some to the left of what their general course is at moderate 
dilution. He has also shown, that this tendency to diverge is 
less for the results of Archibald and Barnes, both of whom 
worked in the Dalhousie Physical Laboratory. This indicates 
that the method they adopted was free from some source or 
sources of error which affected the methods of other observers. 
As they had not carried their observations to solutions of great 
dilution, he suggested that 1 should go over the ground with 
one or two electrolytes and see if, for higher dilutions, their 
results were borne out. This I have done for potassium 
chloride with the results given below. 
I chose this electrolyte for the following reasons:—(1) It is 
one of the salts for which Whethamf has made determinations of 
the ionization coethcients for dilute solutions at 0°C.; (2) The 
rightward or leftward tendency, above mentioned, is most 
marked for this electrolyte ; and (3) solutions of known con- 
centrations are quite easily made up. 
Salt. 
The salt was obtained as chemically pure from Merck. On 
testing it, however, I detected free acid. This disappeared on 
once re-crystallizing and heating to constant weight. Solutions 
* Proc. and Trans. N.S. Inst. Sci., Vol. X., p. 211, 1899-09. 
+t Phil. Trans., A, Vol. 194, 1900, pp. 321-360. 
(422) 
