The explosive mixture is prepared in the water reservoir, the glass bulb shown in the accompanying illustration being employed. This bulb is closed at its upper end by means of a cock and is tapered at its lower end. The gas or air enters at the top by means of a rubber tube and gradually displaces the water through the lower end. The bulbs have a volume varying from 200 to 500 cubic centimeters (12 to 30 cubic inches), and the error resulting from each filling of a bulb is certainly less than 15 cubic millimeters (0.0009 cubic inches). The contents are emptied into a bell by lowering the bulb into the water and opening the cock. If seven bulbfuls of air be mixed with one bulbful of gas, an explosive mixture of 1 to 7 is produced, this being the proportion commonly employed for street-gas. For producer-gases the preferred proportion is 1 to 1, oxygen being often added to the air in order to insure complete combustion.
The calorimeter, after having been filled, is placed in a vessel containing a liter (1.7598 pints) of water so that it is completely immersed. A spark is then allowed to pass. The explosion is not accompanied by any noise; the temperature rises a fixed number of degrees, so that the quantity of heat liberated can easily be computed. Each division of the thermometer is equal to 0.01502 C. The scale reading is minute, each interval being divided by ten, so that readings to the 1,500th part of a degree can be taken.
It should be observed that the mixture generated in the reservoir is saturated with water vapor at the temperature of the reservoir. Consequently, the vapor generated by the explosion must condense in the calorimeter if the final temperature of the calorimeter is the same as that of the water reservoir. If, on the other hand, the temperature be slightly different, a correction must be made; but the error is negligible for differences in temperature of from 2 to 3 degrees C. (3.6 to 5.4 degrees F.). This, however, is never likely to occur if the operation is conducted under favorable conditions.
This apparatus is exceedingly simple and practical. It does not require the manipulation of a pump. The pressure of the mixture is read off on the barometer; the calorimeter is entirely immersed in the water of the outer vessel, so that all corrections of doubtful accuracy are obviated. The method requires but a very slight correction for temperature. Air, alone or mingled with oxygen, or a mixture of air and oxygen, can be easily tested with.