424. DETERMINATION OF THE FREEZING-POINT DEPRESSIONS 
of another flask of 200 ¢.c. capacity. Both flasks were calibrated 
at 0°C. The concentrations, in gramme-equivalents per litre of 
solutions at 0°C., were hence easily calculated. But in some of 
the cases I was not sure of the dryness of my salt, and hence 
determined their concentrations gravimetrically by precipitating 
the KCI of a known quantity of the solution with Ag NQ,. 
Measurement of Freezing-Points. 
The method used in determining the freezing-points was a 
modification of that used by Loumis.* The principal changes were: 
(1) the stirring was done mechanically and at a constant rate; 
(2) the stirrer was not simply of the ring form but had vanes 
attached to it; (3) the temperature surrounding the whole 
apparatus was kept within .56 degree of O°C. The work was 
done in a basement room of the college in which there was no 
source of heat, and experiments were only made when the 
temperature of the air could be kept at O°C. by raising the 
windows. 
The freezing-apparatus consisted of three parts as follows: 
(1) an outer bath containing ice and water; (2) a bath which 
stood in this called the protection-bath, and which contained a 
mixture of salt water and snow, at a temperature of .005 degree 
centigrade below the freezing-point of the solution, whose 
freezing-point was being determined; (3) the freezing-tubes 
immersed in this, which contained the solution under experiment. 
The freezing-tubes consisted of two glass tubes one inside the 
other, the outer being about 3.2 cm. in diameter. The 
space, of about 1 mm., between the two was kept uniform 
throughout by winding silk thread about the inner one at two 
distinct places. This air-space, together with one at the bottom 
of the tubes, caused by the bottom of the inner tube being 
re-entrant, kept an ice sheath from being formed on the wall of 
the tube. The two tubes were sealed together at the top, in 
order to keep moisture from getting in between them, and in 
order to have their relative positions always the same. These 
*Phys. Review, 1, 199, 1893, and 9, 257, 1899. 
