September 17, 1915. THE COLLIERY GUARDIAN. 585 The Science of Combustion.* By Prof. W. A. BONE, F.R.S. Gaseous Combustion. This is the centenary of Davy’s invention of the miner’s safety lamp, which formed the starting point of his brilliant researches upon flame, in which he disclosed and brought within the range of experimental enquiry most of the intricate and baffling problems connected with the fascinating subject of gaseous combustion. Also the ground on which we meet to-day is known to the whole scientific world as the place where, during more than a quarter of a century of continuous investiga- tion, a succession of Manchester chemists, led and inspired by Prof. H. B. Dixon, have devoted themselves to the elucidation of the many problems which Davy’s work foreshadowed. Therefore, both in point of time and place, the occasion is singularly appropriate for a review of recent advances in this important field of scientific enquiry. Ignition Phenomena. The first section of my 1910 report was concerned with “ Ignition Temperatures and the Initial Phases of Gaseous Explosions,” and it is in connection with ignition phenomena that subsequent progress has been most marked. For the ignition of a given explosive mixture it is necessary that the temperature of its constituents should be raised, at least locally, to a degree at which a mass of gas self-heats by combination until it bursts into flame, or, in other words, to a degree at which the chemical action becomes autogenous or self-propelling so that it quickly spreads throughout the whole mass. This particular degree, or in some cases range, of temperature is commonly spoken of as the ignition point of the mixture; but in using the expression certain qualifications should be carefully borne in mind. In the first place, as H. B. Dixon and H. F. Coward had shown in 1909 f whereas when certain combustible gases (such, for example, as hydrogen and carbon mon- oxide, the mechanism of whose combusion is probably of a fairly simple character) and air or oxygen are separately heated in a suitable enclosure before being allowed to mix, the temperature at which ignition occurs lies within a very narrow range, which is, within the limits of experimental error, the same for both air and oxygen (i.e., in the case of hydrogen it is 580 degs. to 590 degs., and for carbon monoxide, 640 degs. to 658 degs.), in other cases, where the mechanism of com- bustion is known to be very complex (i.e., hydrocarbons), the ignition range is either fairly wide, or is materially lower in oxygen than in air, or both, thus :— In air. In oxygen. Methane ... 650 to 750 degs. ... 556 to 700 degs. The explanation of such behaviour is probably to be sought in the known complexity of the combustion, and the marked tendency for appreciable and fairly rapid interaction between the inflammable gas and oxygen before the actual ignition tpoint is reached. If, by any means, such preliminary interaction could either be entirely suppressed, or on the. other hand, it is very rapid in character, the observed “ ignition range ” would be narrowed. There are two other means by which an explosive mixture may be ignited, one is by adiabatic compression, and the other, and most commonly employed of all, is by the passage of an electric spark. According to Dixon and Crofts’ recent determination of the ignition points of mixtures containing electrolytic gas, whereas successive additions of hydrogen or nitrogen progressively raise the ignition temperature of the undiluted gas by regular increments, as would be supposed, successive additions of oxygen, on the other hand, lowers it. The lowering effect of oxygen is indeed puzzling, and its meaning can only be conjectures. Dixon and Crofts have suggested that it may be due either to the formation of some active polymeride of oxygen under the experimental conditions, which seems doubtful, or that the concentration of oxygen in some way or other brings about increased ionisation of the combustible gas. This at once raised the larger question of whether or not ignition is a purely thermal problem, as until recently has generally been supposed. Prof. W. M. Thornton, J of Newcastle, has recently published some very suggestive work on the electrical ignition of gaseous mixtures which, quite apart from its theoretical interest, has an important bearing on the the safety of coal mines where electrical currents are used for signalling and other purposes. From Thornton’s work would it appear that a definite minimum of circuit energy is required before a given mixture at given pressure can be ignited by a spark. And, moreover, he finds that a circuit energy required for the spark ignition of a given mixture, say, of methane and air, is something like 56 times greater with alternating than with continuous current at the same voltage. From this he argues that the igniting effect cannot be simply thermal, but must be in part at least conic. This con- clusion he further supports by the statement that the igniting power of a unidirectional current is, in fact, proportional to the current in the case of many gaseous mixtures over an important part of their working range of inflammability. While there is much that is suggestive in Thornton’s work, there is also a good deal which seems very difficult to interpret from a chemical standpoint; I refer more particularly to his latter supposition of “ stepped ignition,” based upon certain observed abrupt increases * Abstract of the address to Section B (Chemistry) of the British Association, September 8, 1915. + Trans. Chem. Soc., 1909, vol. 95., pp. 514-43. 1 Proc. Roy. Soc., A., vol. 90 (1914), p. 272; ibid., vol. 91 (1914), p. 17. in the minimum igniting current required with con- denser discharge sparks as the proportions of combustible gas in the air mixture examined progressively increases. In other words, it is claimed that continuous alteration of the proportions of gas and air in an explosive mixture is, or may be, accompanied by discontinuous alterations in the spark energy required for ignition. I must con- fess that after careful examination of the published curves, I am quite at a loss to give them any chemical interpretation, and to being somewhat sceptical on the supposed “ stepped ignition.” A repetition and exten- sion of Prof. Thornton’s experiments would be most valuable as a means to a better understanding of the conditions of spark ignition. The Influence of Electrons upon Combustion. During the discussion upon my 1910 report, Sir J. J. Thomson reminded us chemists that combustion is con- cerned not only with atoms and molecules, but also with electrons moving with very high velocities. They might be a fact of prime importance in such intensive forms of gaseous combustion realised in contact with hot incandescent surfaces, as also in the explosion wave. It is known, of course, that incandescent surfaces emit enormous streams of electrons travelling with high velocities, and the actions of such surfaces may be due to the formation of layers of electrified gas in which chemical changes proceed with extraordinarily high velocities. Again, the rapidity of combustion in the explosion wave might conceivably be due to the molecules in the act of combining sending out electrons with exceedinly high velocities, which precede the explosion wave and prepare the way for it by ionising the gas. With regard to this interpretation of the action of surfaces, Mr. Harold Hartley carried out a promising series of experiments in my laboratory at Leeds University upon the combination of hydrogen and oxygen in contact with a gold surface, * which lends some support to the idea, but they require further exten- sion before it can be considered as finally proved. It is my intention in the near future to resume the systematic investigation of the matter as rapidly as circumstances permit, but the experimental difficulties are formidable, and the mere chemist working by himself may easily be misled. We badly need the active co- operation of physicists in elucidating the supposed role of electrons in combustion. Prof. H. B. Dixon and his pupils have, at Sir J. J. Thomson’s suggestion, recently tested the idea as applied to the explosion wave, with, however, negative results, f It is known, of course, that the motion of the ions can be stopped at once by means of a transverse magnetic field, in which they curl up and are caused to revolve in small circles, and a question which Prof. Dixon decided to put to the test of experiment was whether the damping of the electronic velocities in a powerful magnetic field would have any appreciable effect upon either the initial phase of an explosion or upon the high velocity of detonation. But although he employed a very intense magnetic field produced by powerful magnets specially constructed by Sir Ernest Rutherford for the deflection of electrons of high velocity, no appre- ciable effect was observed upon the character or velocity of the flame with any gas mixture at any stage of the explosion; and inasmuch as the high constant velocity of the explosion wave can be entirely accounted for on the theory of a compression wave liberating the chemical energy as it passes through the gases, there seems to be no grounds for attributing it to the ionising action of electrons. The Initial Period of “ Uniform Movement ” or “ Inflam- mation ” of Flame through Inflammable Mixtures, and Limits of Inflammability. An exact knowledge of the properties of flame pro- pagation during the initial period of uniform slow movement, as well as of the limits of inflammability for mixtures of various combustible gases and air, is very important from a practical point of view. Makers of apparatus for burning explosive mixtures of gas and air want to know the speed of flame propagation through such mixtures not only at ordinary temperatures and pressures, but also when the mixtures are heated and used at higher pressures. Also it would be important to know whether or not in the case of a complex mixture of various combustible gases and air when complete composition can be determined by analysis (as for example, coal gas and air) the velocity of flame pro- pagation can be calculated from the known velocities for its single components. Unfortunately, although more than 30 years have elapsed since Mallard and Le Chatelier’s work was published, the necessary data are still wanting to answer such questions, and anyone who will systematically tackle the question and carefully work it out in detail will be doing a real service to the gas-using industries. I am hoping shortly to make a beginning with such an investigation in my new depart- ment at the Imperial College, London, but the successful and rapid progress of such work will involve considerable financial outlay as well as organisation and expert direction. An accurate knowledge of the behaviour of methane and air mixtures under known variations of conditions is of prime importance, from the point of view of the safety of coal mines, and it is rightly occupying the attention of my friend and former collaborator, Dr. R. V. Wheeler, at the Home Office Experimental Station at Eskmeals. And from papers which he has already pub- lished, as well as from some unpublished results which * Proc. Roy. Soc., 1914. f Proc. Roy. Soc., 1914; sec. A, vol. 90, p. 506. he has very kindly permitted me to refer to in this address, it is now possible to correct certain errors in Mallard and Le Chatelier’s results, and to arrive at clearer view of the phenomena as a whole. In the first place, it would appear that the initial “ uniform movement ” or flame in a gaseous explosion, or in other words, propagation of the flame from layer to layer by conduction only (as defined by Le Chatelier) is a limited phenomenon and is only obtained in tubes of somewhat small diameter, wide enough, however, to prevent appreciable cooling of the flame, but narrow enough to suppress the influence of convectional currents. Moreover, ignition must be either at or very near to the end of the tube, or otherwise—particularly with the more rapidly moving flames—vibrations may be set up from the beginning. Whilst all methane air mixtures develop an initial uniform slow flame movement period when ignited at or near the open end of a horizontal tube, both its linear duration as well as the flame velocity are not, according to private information which Dr. Wheeler has sent me, independent of the dimensions of the tube. The speed of flame increases with the diameter of the tube, and the linear duration of the uniform period increases with both the diameter and length of the tube up to a certain maximum, after which increase in length makes no appreciable difference; also, for the same tube, it varies with the proportion of methane in the explosive mixture, being greater as the speed of the flame diminishes, until with the two “ limiting ” explosive mixtures it appears to last almost indefinitely. Dr. Wheeler’s recent redetermination of the velocities of the flame movement during this initial uniform period for mixtures of methane and air in varying pro- portions within the explosive limits has revealed serious errors in Mallard and Le Chatelier’s original results for horizontal tubes of the same diameter as those which Dr. Wheeler has employed. Moreover, Mallard and Le Chatelier’s method of determining the composition of the upper and lower limits of inflammability by extra polation from their curves has been proved to be unwarranted. Dr. Wheeler considers that the length of the tubes used by Mallard and Le Chatelier (one metre only) was insufficient to ensure that the speed measure- ments of the initial uniform flame movement period were unaffected by the subsequent “ vibratory period.” Also, the methane used by them, prepared as it was from sodium acetate, would obviously be impure. The most important difference between the latest results published by Dr. Wheeler and those originally determined by Mallard and Le Chatelier are as follow :— 1. According to Wheeler, the limits of inflamma- bility for horizontal propagation of flame in methane- air mixtures, at atmospheric temperature and pressure, correspond to 5-4 and 14-3 per cent, methane contents, respectively, whereas Mallard and Le Chatelier gave 5-6 and 18-7 per cent. 2. Whereas, according to Mallard and Le Chatelier, the flame velocities for mixtures nerr the upper and lower limits would gradually approximate the zero velocity ordinate, as the limiting composition was approached, according to Wheeler the velocities for both the upper and lower limiting mixtures are con- siderable (in each case about 36 cu. m. per second) and there is an abrupt change from these velocities to zero velocity as the particular limiting composition is passed. 3. Whereas Mallard and Le Chatelier found a maximum velocity of 63 cu. m. per second for a mixture containing about 12-2 per cent, of methane, and a rapid falling off in velocity as this particular composition is deviated from. Wheeler finds a maximum velocity of 100 to 112 cu. m. per second for a series of mixtures containing from 9-45 to 10-55 per cent, of methane. Such differences as are thus dis- closed only emphasise the need of a complete experi- mental revision of the subject. Messrs. Burgess and Wheeler have recently deter- mined the limits of inflammability of methane when mixed, at atmospheric temperature and pressure, with “ atmospheres ” or oxygen and nitrogen containing less oxygen than ordinary air (see below). From these results (see below) it would appear that as the oxygen content of the atmosphere is reduced, the limits of inflammability are narrowed until they coincide when the oxygen content falls below 13-3 per cent., which means that an atmosphere containing 13-3 or less per cent, of oxygen is truly extinctive for a methane flame at ordinary pressures :— Atmosphere. Methane per cent. Oxygen. t ( Nitrogen. Lower limit. Higher lij 20’90 .. .... 79-10 .. . .. 5’60 ... ... 14’82 17’OJ .. .... 83’00 .... .. 5’80 ... ... 10-55 1582 .. .... 84-18 .... .. 5-83 ... 8’96 14-86 .. .... 8514 .... .. 6-15 ... 8'36 13’90 .. .... 86’10 .... .. 6’35 ... 7-26 13-45 .. .... 86-55 .... 6'50 ... 670 My review of this part of the subject would be incomplete without a reference to some interesting observations which have been made by Dr. H. F. Coward and co-workers at the Manchester School of Technology upon the behaviour of weak mixtures of various inflam- mable gases and air, at or just below the lower limit of inflammability in each case. * Their principal experi- ments were carried out in a rectangular box of 30 cm. square section, and 1-8 m. length, with two opposite sides of wood. The box was placed in an upright position, the bottom being water-sealed, and the top closed, with a suitable igniting device near the bottom. They have shown that caps or vortex rings of flame may be projected for some distance upwards from the source of ignition, sometimes apparently for an indefinite distance, without igniting the whole of the combustible mixture. In such mixture there may be an indefinite upward slow propagation of flame, together with incom- * Trans. Chem. Soc. 1914, 105, p. 1859.