AND ON PREVENTING ITS EXPLOSION. 43 
top. This is a yellowish-coloured diffusion of the light round 
the flame of the candle, rising higher, and assuming a 
greenish-blue colour when’ the quantity is considerable, and 
when it is still larger, giving rise to a rapid succession of lumi- 
nous points or flashes. ‘The miner, when judging from this of 
the presence of fire-damp, advances cautiously in the mine, ob- 
serving the appearance on his candle, by raising it slowly from 
a certain height from the floor,—a circumstance I may remark 
which shews very well the tendency to the accumulation of the 
inflammable gas in the roof, and the comparative purity of the 
air below. He thus advances, as far as he can with safety, and 
_it is singular to what length the miners sometimes proceed with 
this trial... This peculiar appearance seems to arise partly from 
the extinguishing effect of the reduced air on the flame of the 
candle; for a similar effect is produced in immersing a lighted 
candle slowly in any gas unfit to sustain combustion ; and as it 
extends so high, it must also partly arise from the imperfect 
ascension of the inflammable gas. If any inflammable air 
were to enter the lamp I have described, this appearance, or 
something similar to it, would take place, even sooner, from 
the rarefaction of the air, and probably the flame would be en- 
tirely extinguished before any explosion occurred. It would 
therefore give an equally sure indication with much more safe- 
ty ; though it is also probable, that no air, so far impregnated 
with carburetted hydrogen, could ever enter from the floor of 
‘the mine while the atmosphere above was such that it could be 
breathed. es 
_ When the danger is suspected to exist, the state of the air in 
any part of the mine may be accurately examined with great 
facility, and entire safety. With a moveable lamp and tube, 
such as has been described, or, if it were thought to give more 
safety in extreme cases, with light obtained by the collision ot 
; F2 steel 
