104

of the limits of the combustibility of gaseousmixtures of coal gas and air by increase oftemperature. For this purpose, I introduceda small wire-gauze safe-lamp with some finewire of platinum fixed above the flame, intoa combustible mixture containing the maxi-mum of coal gas, and when the i nflammationhad taken place in the wire-gauze cylinder,I threw in more coal gas, expecting that theheat acquired by the mixed gas in passingthrough the wire-gauze would prevent theexcess from extinguishing the flame. Theflame continued .for two or three secondsafter the coal gas was introduced; and whenit was extinguished, that part of the wire ofplatinum which had been hottest remainedignited, and continued so for many minutes,and when it was removed into a dark room,it was evident that there was no flame in thecylinder.

It was immediately obvious that this wasthe result which I had hoped to attain byother methods, and that the oxygene andcoal gas in contact with the hot wire com-bined without flame, and yet produced heatenough to preserve the wire ignited, and tokeep up their own combustion. I proved