14

tween, and the heat communicated to, animmense number of surfaces.

My first safety lamps constructed onthese principles, gave light in explosive mix-tures containing a great excess of air, butbecame extinguished in explosive mixturesin which the fire damp was in sufficientquantity to absorb the whole of the oxygenof the air, so that such mixtures never burntcontinuously at the air feeders, which inlamps of this construction was important, asthe increase of heat, where there was only asmall cooling surface, would have altered theconditions of security.

I made several attempts to constructsafety lamps which should give light in allexplosive mixtures of fire damp, and aftercomplicated combinations I at length ar-rived at one evidently the most simple, thatof surrounding the light entirely by wiregauze, and making the same tissue feed theflame with air and emit light.

In plunging a light surrounded by acylinder of fine wire gauze into an explosivemixture I saw the whole cylinder becomequietly and gradually filled with flame, the