215 
of nitrate of potash. This mixture is placed in a small 
copper tube, and ignited by a fuse prepared by soaking two 
or three strands of ordinary lamp wick in nitrate of potash 
solution and drying. About three-quarters of an inch of 
this fuse is placed upright in the mixture, the tube contain- 
ing which is put on a stand, surrounded with four upright 
brass springs (strips of brass fixed to the stand at the bottom 
and curved upwards and inwards). 
The fuse is ignited, and smoulders slowly down to the coal 
and oxygen mixture. In the meantime a cap or cylinder of 
copper, closed at the top (to which is attached a narrow 
copper tube furnished at the end with a stop-cock), is placed 
over the copper tube containing the coal, chlorate, and 
nitrate mixture by pushing it over the springs which serve 
to keep the copper cylinder in its place. The whole 
apparatus is now lifted by the narrow copper tube, and 
immersed in a glass cylinder about 18 Jin. long by 4|in. 
wide containing a large quantity of water (preferably 
either 2,000 or 1,934 cubic centimetres), the temperature 
of which has been accurately taken by a delicate ther- 
mometer. The mixture ignites and burns away like a 
squib, the hot gases bubbling through the water, to 
which it parts with its heat. When the combustion is com- 
plete, the stop-cock at the end of the narrow copper tube is 
opened to allow the residual gas in the apparatus to escape 
and to allow the water to take up the heat still retained by 
the copper tube and by the salts left from the combustion. 
It is admitted that even with the most careful use of this 
apparatus it is impossible to obtain accurate results, because 
there are so many sources of error. First, the chlorate of 
potash in becoming dissociated into chloride of potassium, 
and oxygen liberates a considerable quantity of heat; the 
nitrate of potash, on being decomposed, absorbs heat ; the 
oxygen, on being liberated from the solid condition and 
expanding to the ordinary pressure of the air, absorbs heat; 
