at constant volume and high pressure. 
87 
ture and time taken as in (1). As the heat capacity was reduced, 
owing to the removal of the gas, the rate of rise of temperature 
was greater. 
(3) The bulbs were next filled with water equal in weight to 
the gas originally present. This was effected by melting the 
solder and boiling the bulbs in water. On cooling, water entered, 
and its amount was adjusted so that the bulbs and water weighed 
the same as the bulbs did originally when filled with gas. The 
rise in temperature again was observed on the passage of the 
current, the rate again differing from that recorded in (1). As 
the heat capacity was now greater, the rate of rise in temperature 
was less. The experiment usually started at a temperature of 20° 
to 25° and was continued until it had risen to over 60°, but the 
actual rise made use of in the calculation was that from 36° to the 
temperature attained after an interval of ten minutes, 36° being 
chosen as a convenient point above the critical temperature, so 
that the actual measurement was made upon gas. As the current 
was kept constant during the three experiments, and the period 
of flowing was the same, it may fairly be assumed that equal 
amounts of heat were absorbed by the calorimeter and the con- 
tents. As the final temperatures were not far removed from each 
other, the very small difference in the heat developed owing to 
alteration in the resistance of the wire, was considered too small 
to have any very definite effect. 
We have now three heat capacities to consider : — 
(1) The calorimeter, the metal composing the bulbs, the 
paraffin and the gas. 
(2) The calorimeter, bulbs, and the paraffin. 
(3) The calorimeter, the bulbs, the water, and the paraffin. 
The heat capacities being inversely proportional to the eleva- 
tion of the temperature, from experiments (2) and (3), (since that 
of the water was known), that of the paraffin and metal could be 
deduced. Using the results obtained, in experiment (1), the heat 
capacity of the gas could be calculated. 
The following is the result of a typical determination : — 
Weight of bulbs full of gas = 206*6 
Weight of empty bulbs = 154*3 
Weight of gas used = 52’3 
Weight of calorimeter + bulbs, gas and paraffin = 617'2 
Weight of calorimeter + empty bulbs = 564*9 
Weight of calorimeter + bulbs filled with water = 617*2 
The curves plotted from this observation were quite regular, 
and it is of interest to note that when the curve is plotted for 
the gas from 20° to 60°, no break of continuity occurs between 
