426 DETERMINATION OF THE FREEZING-POINT DEPRESSIONS 
well stirred, and any change in temperature was quickly recoided 
by the thermometer. 
It was so arranged that both these stirrers—viz., the one in 
the protection bath and the one in the freezing-tubes—should 
have the same stroke. Hence they were both fastened to a 
slider on a vertical guide-post. This slider was worked by a 
crank vertically above it. The axis carrying the crank was. 
turned by a hot-air motor. The stroke of the crank was equal 
to the stroke required by the stirrers. 
The two thermometers—viz,, the one in the protection bath 
and the one in the freezing-tubes—were both of the Beckmann 
form, and were graduated to .01 degree. The one used in the 
freezing-tubes had been calibrated at the Physikalisch-Technische 
Reichsanstalt, Berlin. The value of its degree—its length being 
about 5.4 em.—was given to the third place of decimals. As, 
however, it had been tested with its bulb at O°C. and its scale 
at 15°C., I had to make a correction due to the fact that I used 
it with its scale also at O°C. In the corrected form the value 
of the degree was correct. For some time before it was used, 
and while it was being used, it was kept hanging in a vertical 
position with its bulb and scale approximately at zero. This. 
precaution is indispensible, as the constancy of the thermometer 
depends on it. This thermometer was read by means of a. 
microscope, which was firmly mounted on an adjustable stand. 
The eye-piece of the microscope contained a micrometer scale, 
thirty-seven divisions of which corresponded to .01 degree. As. 
half divisions were easily estimated I could read to .0001 degree. 
To get a clear imagine of scale and mercury, a small incandescent 
lamp,driven by a current from several Samson cells, was placed, 
when a reading was being taken, directly behind the thermometer. 
As, however, the mercury and scale are at different distances 
from the microscope, one cannot focus the both at once. Hence. 
I always made a reading with the mercury focussed, for it was 
quite easy to estimate the centre of the blurred image of the. 
scale line. In the course of my experiments, I found out how 
important it was to have the microscope always inclined at the. 
