COMPLEX SOLUTIONS.—ARCHIBALD. 41 
temperature throughout the whole range of concentration 
observed, the amount of the decrement diminishing with increase 
of concentration, until at a concentration of 0.7 it is only 0.6 
per cent. 
Method of measuring Depression of the Freezing-pornt. 
Of the different methods described for the determination of 
the freezing point of salt solutions, that of Loomis* appeared to 
me the best, and to give the most concordant results. His 
method was accordingly followed in making the measurements 
below. 
As it is most essential that the temperature of the room 
where the observations are made should be near zero and as. 
constant as possible, the measurements were carried out during 
the winter months in a basement room of Dalhousie College 
building, where it was found possible to keep the temperature 
below 2°C and constant to within 0.5 of a degree for a couple of 
hours at a time. No measurements were made while the tem- 
perature of the room was above 2°C. 
The thermometer was of tke ordinary Beckman form, 
graduated to 0.01 of a degree. No reading microscope being 
available, I had to be satisfied with the use of an ordinary hand 
lens for this purpose. Nevertheless, as the divisions of the 
scale were about 0.6 mm. in length, I am quite satisfied that 
I was able to read the temperatures to at least .001 degree. 
The following readings of the thermometer made in an experi- 
ment for determining the freezing point of water would seem to 
imply that I succeeded in reading even more closely :—2.3415, 
2.3410, 2.3420, 2.3415, 2.3420. Mean reading, 2.3416. Greatest 
divergence from mean, .0006. 
The thermometer had never been calibrated, and as appara- 
tus for this purpose was not available, I did not attempt to 
ealibrate it myself. The length of scale used for the following 
measurements, however, was less than what corresponded to 
1.4 degree, and for the more dilute solutions, say below 0.1 
* Phys. Review, 1, 199 and 274 (1893) and 3, 270 (1896). 
