April 13, 191 1] 



NATURE 



225 



These results are to some extent in keeping with those 

 >btained by the present writer on June 5, 1896, when he 

 nade the first experiments of this kind with mixtures of 

 :ombustibIe (in that case lycopodium) dust and inert dusts 

 chloride of calcium, dry clay, common salt) at University 

 "oUege, Cardiff, in the presence of Mr. Robson, then Chief 

 nspector of Mines for the South Wales district, and Mr. 

 raughan \ash. 



The third scries of experiments was made with the 

 )rincipal gallery, which in 1908 was 71 yards long, in 1909 

 vas lengthened to 2^0 yards, and in 19 10 to 328 yards, 

 rhe first 33 yards of its length is constructed with "ferro- 

 :oncrete specially strengthened with steel joists. Its form 

 s trapezoidal in cross-section, and its internal dimensions 

 ire 6 feet high, 4 feet 7 inches wide at the top, and 

 ; feet 3 inches wide at the bottom. This shape and these 

 limensions were chosen with the object of assimilating its 

 nterior to that of a roadway in a mine, and the similarity 

 s still further accentuated by moans of props and caps set 

 ip at the usual distances apart in its interior. In its final 

 orm, so far as one can gather from the descriptions, the 

 emainder of the g-allery is constituted by a sheet-iron 

 ylindcr, 6 feet in diameter, with one of its ends abutting 

 [gainst one end of the ferro-concrete section and its other 

 nd open. 



Two massive stone walls, one on each side of the free 

 nd of the ferro-concrete part of the gallery, extend back- 

 vards from the latter to a distance of 10 or 12 feet, and 

 onstitute supports to a vertical barrier of strong wooden 

 »eams, with which that end is closed. The cannon from 

 vhich charges of explosive are fired for the purpose of 

 gniting mixtures of firedamp and air, or of raising and 

 gniting coal dust in the interior of the gallery, can be fired 

 lorizontally at any desired height in the vertical centre- 

 ine of the latter, with its muzzle extending through a 

 lole in, and flush with the inner face of, the wooden 

 ' !. Its bore is 2 inches in diameter and 23J inches 

 md, except where specially mentioned to be otherwise, 

 . . ...vis was placed at a height of 2 feet above the floor. 

 I'he explosive employed was gelatine dynamite fired without 

 amping by means of an electric fuse, and the minimum 

 veight of charge that assured propagation of the coal-dust 

 lame under ordinary conditions was 160 grammes. 



A branch gallery built of masonry, connected at rigfht 

 ingles to the main gallery at a distance of about 17 feet 

 rem its closed end, serves the purpose of a channel, through 

 vhich air can be blown, by means of a ventilating fan, into 

 nd through that gallery, and also affords a means of 

 ngress to and egress from it. 



_ When an explosion is about to be produced, the connec- 



ion between the main and branch galleries is cut off by 



losing a strong door at their point of junction. 



; Tiiere are twelve plate-glass windows about the middle 



I eight of the ferro-concrete part of the gallery, through 



•hirh the progress of flame in its interior can be seen from 



nee. Its cylindrical prolongation, on the other hand, 



iiodded in the centre of a mass of debris, like a rail- 



mbankmont. about 6 feet high by 12 feet wide at the 



id with sloping sides. 



I : !• coal dust employed in the experiments is obtained by 



rinfling coal as it comes from the mine, first in a ball- 



lill. and secondly in an Alsing pulveriser. The degree of 



HPiuss attained in the latter depends upon the length of 



mr during which the grinding is continued. 



r having been first granulated in the ball-mill and 

 _;round in the Alsing pulveriser for the length of time 



I!! rl below, the following proportions of Li^vin coal are 



I' ~trd by a sieve of brass wire with 5625 meshes per 



ju.iif centimetre,* viz. : — 



Time 



o minute 

 115 

 30 

 45 

 60 „ 



Per cent 



remainins; 



on seive 



• 72-5 ■ 



• 3.vO 



• 12-5 



Some analyses of the same coal employed in the experi- 

 [ents are as follows : — 



f Comite Central des Houilleres de France Station d'Essais de Licvin- 

 OlMcme Serie d'Essais, p. 9(1910). 



NO. 2163, VOL'. 86] 



Per Per Per 



cent. cent. cent. 



Volatile matter (including moisture) ... 28-00 25-07 29-50 



Ash 8-57 i2-i6 5-97 



V'olatile matter (exclusive of ash ... 30-60 29-30 31-40 



The weighed quantity of dust employed in each experi- 

 ment was scattered uniformly over the floor of the gallery 

 by hand ; that remaining unconsumed after the experiment 

 was partly swept, partly blown out, by means of a strong 

 current of air from the fan ; and when it was desired speci- 

 ally to cleanse the gallery, jets of compressed air were 

 employed for the purpose. 



When it was desired to effect the ignition of the dust '^ 

 means of an explosion of firedamp and air, part of the 

 gallery next the wooden barrier was isolated by means of a 

 paper diaphragm in exactly the same way as was first done 

 for the same purpose in the Royal Society gallery of 

 1880-1, and afterwards in the Prussian gallery of 1884, 

 and the gas and air already mixed was introduced into it 

 in exactly the same way as an accurately measured quantity 

 of firedamp was introduced into the isolated part of the 

 Royal Society gallery, in which it was mixed with the air 

 by being drawn into the centre and expelled from the 

 periphery, of a rapidly revolving fan in the interior of the 

 gallery itself. 



The apparatus employed for measuring pressures is of 

 the "crusher type," such as is employed in testing 

 explosives. It consists of a cylinder, containing a hollow 

 piston with a block of lead in its interior, and a small 

 steel ball interposed between the block of lead and a fixed 

 support. The pressure acts on the piston which presses 

 the lead against the steel ball, and the latter, being pre- 

 vented from moving by the fixed support, penetrates the 

 lead to a greater or less depth. In spite of the extreme 

 accuracy with which it is professed that the depressions 

 produced in the lead block can be measured, a more clumsy 

 and probably inaccurate method of measuring the com- 

 paratively small pressures here requiring to be dealt with 

 could hardly well be imagined. 



The appliance for measuring velocity, which consists of 

 a counter marking fifths of a second, started at the moment 

 the explosion commences and stopped by an observer when 

 the flame appears at the end of the gallery, the length of 

 which for this series of experiments appears to have been 

 65 metres only, seems to be hardly less trustworthy than the 

 pressure recorder. 



On the other hand, the flasks for collecting samples of 

 the products of combustion immediately after the passage 

 of the flame, from which the air had been extracted 

 beforehand, and into which nothing could enter until a 

 sealed glass tube which communicated with their interior 

 had been broken by a detonator ignited by the flame of the 

 explosion, seem to be satisfactory. 



The firedamp employed in some of the experiments was 

 obtained from the pit near at hand, stored in a gasometer, 

 and mixed with air in the proportion of 9 or 10 per cent, 

 before being introduced into the isolated part of the gallery 

 in the manner already indicated. 



Of all the explosives tested, dynamite was found to 

 produce coal-dust explosions with the greatest facility. It 

 was found that the explosion of 8 cubic metres (282^ cubic 

 feet) of a mixture of firedamp and air, when ignited by 

 means of 100 grammes of black powder, easily gave rise to 

 a coal-dust explosion under favourable conditions ; but that 

 under less favourable conditions the superposition of a 

 firedamp explosion upon that of dvnamite actually diminished 

 the chances of propagation. M. Taffanel's attempt to 

 explain this phenomenon, by supposing that the large quan- 

 tities of carbon dioxide and water vapour projected into the 

 dusty atmosphere in consequence of the combustion of the 

 firedamp are responsible for this result, is altogether 

 erroneous. The true explanation is that the expanded wave, 

 following after the condensed wave in the cul-dc-sac con- 

 stituted bv the little gallery, overtakes and extinguisbrs the 

 flame. The present writer observed the same phenomenon in 

 his smaller Royal Societv gallery of 1877-8, and succeeded in 

 destroying the expanded wave and securing free propaga- 

 tion of the coal-dust explosion on every occasion by pro- 

 viding a flap-valve, opening inwards only, at the closed end 

 of the gallery, through which air was drawn with sudden 

 violence an instant after the firedamp mixture had exploded ; 

 and he has no doubt that M. TaflfancI would have exactly 



