218 
The water is then allowed to enter the tube and come in 
contact with the hot crucible and tobacco-pipe support and 
entrance tube for the gas, to abstract the heat left in them ; 
the whole of the water is then well mixed and the tem- 
perature again taken, the difference between the two 
temperatures being the heat given to the water by the 
combustion of the coal. I have found that the temperature 
of the water is practically not altered by passing about three 
gallons of air or oxygen through it, that being in excess of 
the quantity required to burn the coal, between IJ and 2 
gallons being actually required. 
By this method it is not necessary to deduct or add to 
the result obtained. It is more convenient than Lewis 
Thompson’s method, inasmuch as it is not necessary to 
have the coal in such a fine state of division, whilst the 
drying and weighing out of the nitrate and chlorate of potash, 
and incorporating the coal with them, is dispensed with, 
and when care is taken, the results obtained by a series of 
experiments on the same coal is the same for each experi- 
ment. The rise for each gramme of good coal is somewhere 
about 6 to 7 degrees Fahrenheit for the 1,934 grammes 
of water, that being equivalent to about 11,500 to 18,500 
units of heat. 
Graphite burns away quite easily in the oxygen apparatus. 
During the time the experiment is being made I find it 
necessary to have the cylinder containing the water resting 
on three pieces of cork in a loosely fitting vessel of bright 
tinned iron plate, having a slit Tin. long and lin. wide cut 
down one side, through which the combustion can be 
observed. This vessel practically prevents loss of heat from 
the water if it is above the temperature of the surround- 
ing air, and vice versa if the temperature of the water be 
lower than that of the air, but I prefer to have at hand a 
large supply of water which has been exposed to the 
atmosphere for some hours, in order that its temperature 
may become as nearly as possible the same as that of the air. 
