1876.] of Coal-dust in Colliery Explosions. 357 



passed over the naked flame at B, and thence into the upcast. The 

 flame of the naked light seemed to lengthen somewhat under these cir- 

 cumstances ; but on no occasion did the mixture take fire, or show any 

 symptoms which led me to suppose it might possibly do so. I repeated 

 this experiment many times, both with steam-coal dust and with the 

 finest and driest dust of the No. 3 Rhondda seam, which is an excellent 

 coking coal, and invariably obtained the same result. The air was charged 

 at different times with every proportion of dust, from a thin, scarcely 

 visible cloud to one which extinguished the flame of the lamp. 



The following analyses of the two kinds of coal-dust employed in these 

 experiments have been kindly made by Dr. Erankland : — 



Steam-Coal dust. Bituminous-Coal dust. 



Carbon • 85-295 82-570 



Hydrogen 5-040 5-400 



Oxygen (by difference) . . 1-261 6-030 



Nitrogen " -608 1-050 



Sulphur -692 -800 



Moisture -614 -670 



Ash 6-490 3-480 



100-000 100-000 



In some of the experiments with coal-dust and air alone, the dust of 

 the bituminous coal was carefully dried and sifted through muslin before 

 being used. 



The results of these experiments, and of others which I have made 

 since, seem to indicate very conclusively that a mixture of air and coal- 

 dust is not inflammable at ordinary pressure and temperature. 



"When, however, a small quantity of firedamp was allowed to pass into 

 the apparatus at E, the valve E being partly opened to determine an air- 

 current from E towards B, the mixture of air, firedamp, and coal-dust 

 was found to be somewhat explosive. 



This experiment was conducted in the following way : — A safety-lamp 

 was placed in the apparatus at B, and the quantity of firedamp was regu- 

 lated until the flame of the safety-lamp showed no indication of its pre- 

 sence. The safety-lamp was then removed and replaced by a naked 

 light, and coal-dust was admitted at C. The effect was instantaneous ; 

 for, as soon as the cloud of coal-dust touched the flame, an explosion took 

 place, and the box was filled with a red flame, which continued to burn 

 until the supply of coal-dust was cut off or the air-valves closed. The 

 naked light inside the apparatus was raised up to the top by means of a 

 wire, just before the coal-dust was admitted, in order to ascertain that 

 there was not an explosive mixture of gas and air at that point. 



I have extracted the following analysis of the firedamp employed in 

 this and all the experiments that follow from a paper by Mr. J. W. 

 Thomas " On the Grases enclosed in Coals from the South- Wales Basin, 



