April 25, 1913. _______________________________________________________________________________________________________ THE COLLIERY GUARDIAN. 849 before being thrown down the tube, the effect was j that the lamp-flame had very little action per se on the more marked, and the whole tube was filled with gas mixture passing through or by it, but that the solid flame.” I particles heated by their passage through the flame These results were obtained with proportions of gas caused chemical combination to begin after they had corresponding to from 5 to 6 per cent, of firedamp. They confirm the results obtained by M. Abel; extremely fine dusts raised to incandescence by a flame can produce by their contact the inflammation of mixtures poor in firedamp, but it must be added that the flame thus produced is not propagated automatically in a similar mixture beyond the paths traversed by the incandescent particles. Examination of Abel’s Theories. The statement made in Abel’s report on the Seaham* dusts that a current of air of more than 200 ft. per minute containing only 3*5 per cent, of firedamp could propagate flame is at variance with the results obtained by practically all other observers.* Assuming that the particular sample of firedamp used by Abel contained as its combustible constituent methane only, the smallest quantity of that gas that can form an inflammable mixture with air, at atmo- spheric temperature and pressure, is, according to our experiments, 5’6 per cent. It has been suggested that the “sharp” or “silver” gas from the Wigan Nine- foot seam used in the Garswood Hall experi- ments might have contained a considerable per- centage of ethane and thus had an inferior “ lower limit ” of inflammation than pure methane. But such a high proportion of ethane (nearly 50 per cent.) as would be necessary to account for the lower limit found by Abel was certainly not present in the gas, since the analyses made by Dr. W. Kellner of the gas employed have been published in detail,f and from these it is evident that the combustible portion of the firedamp (varying from 84’2 to 88’9 per cent, of the gas) must have been practically all methane. The important point, however, in the Garswood Hall experiments is not the 44 lower limit ” of inflammation of the gas mixture, but the fact that the addition of •calcined magnesia (and other incombustible dusts) to a non-inflammable mixture of the pit gas and air brought about its general inflammation either immediately or after a few seconds. The great reputation enjoyed by Sir F. Abel as an experimenter, and his long experience in dealing with explosives, have invested his conclusions with regard to the danger of incombustible dusts in mines with an .authority that hitherto has not been seriously questioned dn this country. In order that the point emphasised by Abel may be understood, it is necessary to recall the arrangement of the apparatus used by him and the effects he observed. It appears that the gas flame by which the mixture was ignited was 12 ft. in front of the hopper by which the dust was introduced, and that the current con- taining a non-explosive mixture of gas and air ran past the hopper towards the flame. As soon as incom- bustible dust was introduced from the hopper it was •carried forward as a cloud by the current towards the gas flame, and on reaching the lamp it first caused flares of flame in the current which had passed the lamp, but ^finally caused a general ignition of the gas in the whole ;gallery—as well that behind the flame as in front of it. The ignition of this cold, non-explosive gas mixture behind the flame is what evidently appealed to Abel as ^a proof of the power of non-combustible dust to ignite otherwise non-inflammable mixtures of gas and air. In his first explanation Abel appears to have thought * We are indebted to Dr. H. F. Coward for the following list of determinations of the lower limit of methane that /have been made by different observers:— passed beyond the flame. It is undoubtedly true that solid incandescent particles can form a centre of chemical action in mixtures below the “ lower limit,” and when such particles are wafted into a region of the unburnt gas mixture (which has escaped direct contact with the flame) they will be surrounded with an aureole of burning gas. Their effect will be the same as if the dimensions of the lamp-flame had been enlarged, but only in so far as they are carried or projected away from the original envelope of gas in which they were heated up will they produce any appreciable chemical effect—for the gas stream, which carried them through the flame, will itself have been heated up to a temperature high enough for combustion to begin (independently of the dust), and this combustion is visible in the cap streaming from the flame. If incandescent particles fall by gravity through a gas mixture just below the lower limit, as in Mallard and Le Chatelier’s experiment, each particle will be surrounded by a cap, showing that combustion is taking place round its surface. Any flame, or other source of heat, moving through a gas mixture causes this kind of combustion in its immediate neighbourhood; but this does not render explosive the rest of the mixture away from the source of heat. It is difficult to imagine how, in Abel’s experiment, the flame could strike back into the region where the particles were cold, if the flame could only exist by reason of the caps formed round previously heated particles. It would be necessary to conceive the dust particles, not as floating in a current, but as endowed with proper motions of their own. In Abel’s second explanation, viz., that the dust particles act catalytically—inducing chemical action on their surface—as finely-divided platinum induces the combination of hydrogen and oxygen—we have a similar difficulty. It is not suggested that the dust acts in the cold, but only after being heated by passing through the lamp flame. Why, then, should the flame strike back ? But granted that a catalytic action might begin even round cold particles, it is difficult to understand how such action could result in a general inflammation or explosion. Platinum produces an explosion in hydrogen and oxygen because the chemical union whereby the platinum is raised in temperature only uses up a small proportion of the combustible gases; the uncombined remainder will still explode, in spite of being mixed with some steam—the products of the flameless combustion. But in the Abel experiment an amount of combustible gas, which is insufficient to propagate flame in the mixture, is supposed, by suffering partial combustion, to bring the remainder into an explosive state. In other words, the running down of a portion of the chemical energy into heat is supposed to confer on the remaining energy a greater power of producing an explosion than was originally present. Of course, the heating of a gas mixture by an outside source of heat increases its explosive power: it is easy to show, as we have shown, that a mixture of gas and air that will not propagate flame when a spark is passed at the ordinary temperature, will do so if heated up to 100 degs. Cent, by a steam jacket. But if the source of heat is derived from the gas itself this is no longer true. It seems to us, therefore, that the only explanation of Abel’s observations which can be afforded by “catalytic action ” is the supposition that the finely-divided dust particles so increase the rapidity of chemical action in The Lower Limit of Inflammability of Methane and Air Mixtures. Observer. Reference. _______ _________ Davy .................... Wagner.................. Coquillon ................ Mallard and Le Chatelier ... _______________________________________ Phil. Trans., 1816.............................. Bayer Industrie und Gewerbeblatt, 1876, 8, 186 ... C.R., 1876, 83, 709.............................. Ann. des Mines, 1883, [8], 4, 347 ................. Lower limit. Remarks. Wiillner and Lehmann..... .............. Broochmann............. Le Chatelier........... j Roszkowski ............. | Clowes................... | Bunte................... Ber. der preuss. Schlagwetterkommission, 1886, B, 3, 193 Jour, fur Gasbeleuchtung, 1889, 32, 189 ......... Ann. des Mines, 1891, [8], 19, 388 .............. Zeitschr. Physikal. Chem., 1891, 7, 485 ........ “Detection of Inflammable Gas and Vapour/’ 1896 Ber., 1898, 31, 19 ............................ 6’2-6’7 5’9-6’25 5’9-6’2 5’6 i 5’9 i Deduced from the measure- I ments of the velocity of ; flame in different mixtures. Couriot and Meunier....... Le Chatelier and Boudouard Eitner.................... Burgess and Wheeler ..... C.R., 1898, 126, 750 and 901 ...................... C.R.., 1898, 126, 1510 ............................. Habilitationsschrift, Munchen, 1902 __............. Trans. Chem. Soc., 1911, 99, 2013 ............... : Dry. , Moist. I Ignition from below. ; Ignition from above. Deduced as by Mallard and i Le Chatelier. Ignition from below. Ignition from above. Ignition in Bunte burette. | Central ignition in large globe. _______________________________________________________________________________________________________ f Final Reports of the Royal Commission on Accidents in Mines, 18S6, Appendix V., p. 144. the gases surrounding them that a temperature and pressure are produced sufficiently high to bring the adjacent gas to its ignition temperature—since the quicker the combustion the less the loss of heat by radiation and conduction. It is known that increased pressure lowers the ignition temperature of a gas mixture, and that an increase of temperature and pressure lowers the lower limit of inflammation. It is also known that the sudden shock of an explosion wave will traverse and burn mixtures that are not ignited by a spark or an ordinary flame. In like manner it is just conceivable that incombustible dusts, by hastening the burning round them by means of catalytic action might cause a flame to propagate itself through a mixture otherwise uninflammable. We therefore thought it desirable not only to repeat Abel’s experiments under conditions as nearly as possible similar to those which he employed, but also to measure the rapidity with which flames are spread through gaseous mixtures with and without the presence of finely-divided dusts. (i) Experiments in a Rectangular Iron Gallery 1 square foot in Cross-section. The gallery, 14J in. high and 10 in. broad, was 24 ft long. At one end it was open; at the other it was con- nected by a short fan drift to a Sirocco fan, which drew the air and gas from a chamber provided with suitable openings for their admission. The gas was admitted into this chamber through a 3 in. main fitted with a long-handled stop-cock. Into the roof of this chamber, and therefore also on the intake side of the fan, an arrangement for introducing dust was fixed.* This arrangement con- sisted of a vertical tube, 3 in. in diameter and 3 ft. long, fixed near the fan-casing; its lower end was closed by a cone (4 in. in diameter at its base) attached to a rod 3J ft. long, passing upwards through the tube. The tube was filled with the required dust, which, when the cone was allowed to drop, fell into the intake and was carried by the air-current through the fan and along the gallery. A second (inverted) cone attached to the top of the rod prevented air from entering through the dust tube after the dust had fallen. By experiments made with an anemometer in the gallery, the velocity of the current was determined for any given speed of the fan. These measurements were repeated at frequent intervals during the experiments. It was found that the addition of magnesia dust to the current made no measurable difference in the velocity of the current. A sampling apparatus was fixed in the centre of the gallery and arranged so that a sample could be drawn off continuously through a narrow tube during a test. By drawing off samples at the top and bottom of the gallery it was shown that half-way along the gallery the mixture of air and gas was perfect at all velocities of the current. The Sirocco fan acts as a very efficient “ mixer ” of gases, which are drawn into it together. At a point along the gallery 16 ft. from the fan-drift a Davy lamp, with the gauze removed, was placed with its flame protected in some cases by a low glass cylinder, in others by a small metal shield on the windward side. A large thick plate-glass window allowed a full view of the gallery where the lamp was placed, and of some distance on either side. The combustible gas employed was either pentane, previously made into an air-gas and stored in a 1,500-ft. holder, or coal gas. The mode of experimenting was as follows:—The lighted lamp being placed in position the fan was set going until the desired rate of current was nearly attained. The gas supply was then turned slowly on until the observer watching the lamp saw the flame begin to “ tail ” and form a cap or flare. The gas flow was then kept constant for one or two minutes, while the lamp flame was observed and a sample of the mixture collected. Calcined magnesia powder was then introduced into the chamber on the intake side of the fan, whence it was drawn with the gas and air and carried forward as a cloud through the gallery. During the passage of the dust-cloud a second sample of the gas mixture was collected. In the preliminary experiments mixtures of coal gas and air were used. Owing to the distance of the gallery from our large gasholder, it was difficult to obtain sufficient gas to form explosive mixtures with a high velocity of current. With velocities upto 600 ft per minute, however, we could obtain a “ flare.” On adding calcined magnesia or stonedust to the current, no effect was produced, except an increase in the luminosity of the flare. In our systematic experiments one observer attended to the fan motor and kept it constant, a second regulated the gas supply, while a third watched the lamp flame. The gas supply was brought from a holder * The apparatus for dropping dust is shown in fig. 2,