334 On the Methods of detecting [Mat, 



which closely surrounds the flame, and which is largest above it, 

 and at last passes into the smoke. This body is rather obscurely 

 seen round the flame of a candle in the night, in a room not much 

 illuminated by any other body; but it may be distinctly seen by 

 using the following means : — Hold the candle in one hand at about 

 a foot or 18 inches from the eye, interj^ose the other hand between 

 the flame and the eye at about one inch from the candle, and let 

 only a very small portion of the upper part of the flame be seen. 

 By this arrangement the vivid light emanating from the flame will 

 be intercepted, and the body in question observed for one-eighth to 

 a quarter of an inch, and sometimes higher, above the flame. By 

 changing the position of the hand, so that a small portion of the 

 flame round eitlier side may be seen, this body will have, from the 

 top to the middle of the flame, a yellowish-brown colour, about 

 one-sixteenth of an incli broad ; below the middle, the same colour 

 inclining a little to a purple ; and downwards it gradually passes 

 into an ultramarine blue, which is the colour of the flame at its 

 bottom. At that place too, this body joins the flame. Such is that 

 body which the miner calls the candle-top, as it appears when a 

 candle is burning in pure atmospheric air ; and as he uses only that 

 part of it which is above the flame, his term of top is very appli- 

 cable. 



Now there are certain differences in the top of a candle between 

 its burning in a mixture of atmospheric air and carbureted hydrogen 

 gas, and in the air only. These differences are proportionate to the 

 quantity of gas in the mixture, and point out to the miner when 

 the quantity is such that a very little more would cause the instant- 

 aneous combustion of the whole body of gas. The alteration in the 

 top consists of a change in colour, and of an increase in its dimen- 

 sions. These changes become greater as the quantity of gas in- 

 creases. The change of colour which is universally considered as a 

 sure indication of the existence of carbureted hydrogen gas, and of 

 course the most dangerous change, is from yelloivish-hrown to 

 greenish-blue ; especially vvlien it is accompanied with a rapid 

 assension of very small luminous bodies, generally called points, 

 which appear to be propelled upwards through the flame and top 

 with a very faint crackling noise. The precise dimensions of a 

 dangerous top is not the same in all coal-pits, nor in every situation 

 in the same pit, nor in every candle in the same situation : so that 

 the top which might be considered very dangerous, in point of size, 

 in one situation, is not so much so in another. Hence a great 

 latitude is left for the miner's judgment and prudence ; and if he 

 be experienced in his candle symptoms, and well acquainted with 

 local differences, he may, if necessary, venture mucli further in a 

 mixture of airs, than a stranger to that particular situation with the 

 same degree of judgment. I have seen the greenish-blue top one 

 inch and a quarter long above the flame of the candle. Other 

 miners say they have seen it longer; but others say the inflammable 

 air has fired when it was shorter. With respect to this variation in 

 1 



