134 
DRS. L. MONO, W. RAMSAY, AND J. SHIELDS ON THE 
the very slow melting’ of ice or freezing of water inside the calorimeter. When the 
ice is first produced in the calorimeter the whole apparatus must stand for several 
days surrounded by ice before equilibrium between the ice and water inside is 
established. Owing to adventitious circumstances, atmospheric pressure, etc., the 
normal creepage of the meniscus sometimes amounted to several millims. an hour, 
but experiments were never conducted until this had diminished to one or two 
millims., whilst in most of the experiments the normal creepage was less than one 
millim. per hour. 
Fig. 2. 
After admitting hydrogen the normal creepage was never the same, as, even after 
three or four hours, hydrogen continues to be slowly absorbed. This difficulty was 
overcome in the following way :—After the last reading of the deflection was taken, 
the volume of hydrogen used was immediately noted, and by subtracting what was 
required to fill the experimental tube at 0°C. (the top of the tube and the taps being- 
covered with melting ice) the quantity of hydrogen occluded up to the time of the 
last reading of the deflection was found. The final occlusion of hydrogen was of 
course very slow, and the real calorimeter deflection was obtained by plotting the 
readings in a co-ordinate system as shown m tig. 2. 
The first part of the curve I, represents the normal creepage before the 
admission of hydrogen. The horizontal part IT, a large portion of which is omitted, 
shows the chief deflection due to the main occlusion of the hydrogen during the first 
few minutes (10-15 min.). The curve then bends upwards, III, but never becomes 
