1882.] On the Influence of Coal-dust in Colliery Explosions. 489 



same source as that used in most of the former ones. It had been 

 produced in the operation of grinding coal for coke-making purposes, 

 and had floated in the air and been deposited in a still atmosphere 

 ■within the building containing the machinery. I could not obtain the 

 results described in my previous papers, and in the present one, 

 with dust taken from screens, or from the roadways of mines ; for, on 

 the one hand, the dust from screens seems to have lost its finest 

 particles before being deposited ; and, on the other hand, the feeble 

 nature of my explosion prevented me from attempting to perform the 

 winnowing operation to which I imagine the dust in mines is subjected 

 before it is required to propagate the flame of an explosion. 



It is an indispensable condition that the dust be dry, and I have 

 found also that the best experimental results can only be obtained in 

 fine dry weather. For instance, on the 20th of July, when it was 

 warm and dry, I succeeded in producing flames of coal-dust and pure 

 air varying in length from 108 to 147 feet ; but on the 22nd, only two 

 days afterwards, when it rained occasionally, and the air contained 

 very finely divided moisture in the form of a light and hardly percep- 

 tible mist, I could not produce a flame even 36 feet long. 



In making the series of experiments corresponding to the ones we 

 are now discussing, which were described in No. II paper, I found 

 that when the gallery was air-tight, or nearly so, the flame could not 

 be propagated further than 30 or 40 feet from the origin ; whereas, 

 when the seams between the boards were open, the flame travelled 80 

 or 90 feet under the same conditions as before. The dry weather 

 which had prevailed before the 14th of July last had dried up the 

 timber of which the gallery was formed to such an extent as to pro- 

 duce twelve open seams in it longitudinally, varying from |-" to \" 

 wide ; and it was recognised after the seventeenth experiment had 

 been made, that this circumstance allowed the force of the fire-damp 

 explosion to be dissipated, without forming an air .wave of sufficient 

 energy to raise the coal-dust in the gallery. Accordingly, strips of 

 wood were nailed along the seams on the top, on the doors, and on the 

 back, while some of the succeeding experiments were in progress ; 

 strips of canvas were nailed along the seams on the floor ; the spaces 

 between the ends of the sections were carefully closed, and canvas 

 was nailed along the joints between the doors and the gallery, with the 

 object of making it as air-tight as possible everywhere. This opera- 

 tion was completed as far as the end of the sixth section, when the 

 thirty-eighth experiment was made. The experience in this case was 

 that the more air-tight the gallery could be made the better were the 

 results. 



When the twenty-eighth experiment was made, it first became 

 evident that the gas and air in the explosion chamber had not been 

 sufficiently well mixed in the preceding experiments to produce the 



