of Edinburgh, Session 1879-80. 691 
mixture. The differences between this temperature and each of the 
other two temperatures, viz., the high temperature corrected for 
cooling, and the low temperature, gave the fall and rise of tempera- 
ture respectively. Calling now w the weight of water in the beaker, 
w 1 the weight of water added, t the rise of temperature, t 1 the fall of 
temperature, and c the thermal capacity of the beaker, we have 
wd\ - rot 
c — — — 
t 
The mean of a set of five experiments made to determine c gave 
almost exactly 12 as its valued 
The thermal capacity of . the heater was next determined. This 
was done by filling the vessel with the same volume of water as I 
intended to use of the solutions, and finding the rise of temperature 
produced in the water by nearly the same change of temperature 
in the heater as was to be used in the experiments. Putting w' for 
the weight of water, t' for the rise of temperature, t\ for the fall of 
temperature of the heater, and e' for its thermal capacity, we get 
= (w + c)t' 
or c = (w' -f c) t\ . . . . (2) 
It is manifestly of great importance that the value of c should be 
known with accuracy, and accordingly the experiments for determin- 
ing it were made with the utmost care. As the following results of a 
series of experiments do not vary by one per cent, from the mean 
result, we may conclude that the mean result is true to a fraction of 
one per cent. 
Number of 
Capacity of 
Experiment. 
Heater. 
1 
22-521 
2 
22-402 
3 
22-457 
4 
22-501 
5 . 
22-373 
Mean result, 
22-451 
If now W be the weight of an equal volume of the solution, and 
* The mass of this vessel was 70 grammes. 
