OF DILUTE SOLUTIONS OF ELECTROLYTES.—HEBB. 429 
when the over-cooling is zero. For this purpose I took a solu- 
tion and found its depressions for different over-coolings. These 
depressions I plotted as ordinates against the over-coolings as 
abscissae. This gives practically a straight line which, if pro- 
duced to cut the depression axis, cuts off a portion from it 
representing the depression when the over-cooling is zero. 
Raoult has shewn that the following relation holds for solutions 
of different concentrations 
C!=C(1+KS) 
where C* is the observed depression for over-cooling S, C is the 
depression for over-cooling zero, and K is a constant. Hence, 
determining C1 and § for different solutions, and knowing K to 
hold for all solutions, we can find C ineach case. I determined 
K to have the value .02. Hence it can easily be seen that for an 
over-cooling of .1 degree the values of the depressions will be 
.02'/ too great. 
The ionization coefficients are taken from a paper by 
Whetham.* Since he only carried his concentrations to .03 
gramme-equivalents per 1000 grammes of solution, I have 
extended the curve under guidance of extrapolated values given 
by Dr. MacGregor.f He obtained his extrapolated values by 
plotting, alongside of one another, the ionization coeflicient- 
concentration curves for 0° and 18°—the latter being obtained 
from data given by Kohlrausch. 
In the following table the concentrations are given in 
gramme-equivalents per litre of solution at O°C. The depressions, 
which have been corrected for over-cooling, as pointed out, are 
given in degrees Centigrade. The ionization coefficients are for 
O°C., and the equivalent depressions are the depressions in 
degrees Centigrade divided by the concentration in gramme- 
equivalents per litre of solution at O°C. The letters 7 and e after 
the coefficients shew whether they were obtained by interpola- 
tion or extrapolation. 
* Loe. cit. 
+ Proc. and Trans. N.S. Inst. Sci., Vol. X. p. 218, 1899-90. 
