FLAME. From the earliest times the phenomenon of flame became invested with a peculiar element of mystery, and little progress was made in elucidating it until a late stage in the history of human knowledge. To people unable to discriminate between things material and immaterial, flame seemed to be a radiant nothingness, a sort of all-devouring element of divine or diabolic significance. At last the prosaic idea gained currency that flame was merely burning smoke. Newton, among others, expressed this view. The fact that when the flames of oil lamps and candles are extinguished a visible smoke trails in the region just where the flame was, and that this smoke can be lit to re-establish the flame, no doubt gave rise to the idea that flame is burning smoke. Very little alteration has to be made in this conclusion in order to render the modern view, namely, that flame is burning gas. The visible smoke of any extinguished flame we now know to be due to the fact that the hot invisible vapour (gas) of the liquid or solid combustible which was feeding the flame gets chilled, and so produces a visible cloud of little drops or solid particles, just as invisible water vapour does when it is chilled. A flame may con tain solid matter as well as its burning gas, but it is the burning gas that is the basis of the flame. No doubt the term flame is occasionally used for gases put into a state of glow by some other stimulus than that of the burning process, but in this article we shall use the term flame in the ordinary sense and in reference to the burning of coal-gas and other common combustibles.
The fact that flame is burning gas leads at once to the question of what is meant by "burning," or by the process of combustion. It will suffice here to say that by combustion is usually meant a chemical action between two different substances A and B, in which a large amount of heat is evolved owing to the union of A with B to form a compound AB, or, if A and B are composite bodies, to some exchange of chemical partnership. The heat so gen erated gives hotness to the gases of the flame, and it might seem also that the light of the flame would be due to the glow of the hot gases. The explanation of the light of flames, as will be seen later, is, however, not quite so simple. The fact that a flame is ordi narily the outcome of a chemical transaction between two gaseous substances A and B gives us the key to explaining the structure of flames. In the first place A and B are equal partners in the transaction, but practically all terrestrial flames are transactions in which the oxygen of the air takes part ; thus we get flames by leading other gases into our oxygenated atmosphere and lighting them where they enter. It is easily conceivable that on another globe what we on earth call combustible gases might form the atmosphere, and in that case flames would be obtained by leading in oxygen or air and igniting it as it entered such an atmosphere. This inversion of the ordinary way of getting a gas flame can easily be shown in a well-known laboratory experiment. It will be seen then that the terms "combustible" and "supporter of com bustion" are in some degree misleading, as they suggest a differ ence of function.
In order that we may have a flame the gas A and the gas B have to become admixed and the temperature has to be raised until ignition takes place. The mixing may occur spontaneously, as when one gas streams into the other, or by the use of some special device. The temperature is usually raised not gradually but by the use of a match or a spark much hotter than is really necessary for the purpose. It is true of all such gas mixtures that there are limits to the proportions within which the mixture will burn and also that the rate of burning will vary according to the proportions of the mixture. These facts are now well-known among those who drive motor-vehicles. In the case of average coal-gas and air at ordinary atmospheric temperature and pressure, mixtures of the two will not inflame if the proportion of coal gas is below about 7% or above 3o%. The maximum rapidity of burning—a travel of the flame through the stationary mixture at a rate rather more than one metre per second—is found in a mixture of about i 8 % of gas and 8 2 of air.
The fact that flames—at any rate large flames—are hollow was discovered long since and is strongly emphasized by Francis Bacon who, by thrusting an arrow into a flame and finding it charred only where it passed through the edges, concluded "that flame burneth more violently towards the sides than in the midst ; and, which is more, that heat or fire is not violent or furious but where it is checked and pent." Proofs of the hollowness of such flames and their content of unburned gas are easily made. A flame of carbon monoxide gives a uniform blue cone of flame, and if a match head be passed through the cone to the middle, the stick of the match will catch fire whilst the head remains for a con siderable time unignited. A simpler experiment is to bring a card horizontally down on to a candle flame until it nearly touches the wick; it will be found that the upper side of the card shows a charred ring. Another simple proof is to lead off the gas from the middle of the flame by an inclined glass tube and to light it at the projecting end. The appearance of some flames, however, is decep tive. The flat flame given by the old fashioned gas-burner con sists of thin sheets of flame, whose hollowness can be demon strated without much difficulty. The bright yellow part of a candle flame looks as if it went right through the flame but is really an exceedingly thin sheet.
Apart from the question of shape, the actual appearance of a flame varies according to the nature of the gases A and B that are used in producing it and the variety of chemical changes taking place. In the flame of hydrogen we have no reason to believe that there is any difference in the chemical processes taking place in different parts of the flame. Hydrogen and the oxygen of the air are throughout entering into chemical union to form water. With carbon monoxide the same chemical transaction —the union of the gas with oxygen to form carbon dioxide—is taking place everywhere in the flame. These flames therefore have the appearance of uniformity. But in the case of combustible gases of a more complex character, where the process of burning may take place in stages and where incidental chemical changes ensue, the flame loses its uniformity and becomes differentiated into distinct zones. The flame of coal gas, oil, tallow, wood, paper, string and in fact of all ordinary combustible bodies shows a differentiated structure. There is a distinctly blue part of no great luminosity and a distinctly yellow part of considerable brightness, to which the light is mainly due. The fact that the combustible substance in these cases contains at least two com bustible elements, carbon and hydrogen, the first of which can burn in two stages, and the further fact that the gases enclosed in these flames undergo chemical changes when subject to a mere baking process from the burning walls, account, as we shall see, for the complexity in the structure of these flames.
In using the blowpipe for producing a hot flame we force in just enough air for complete combustion ; more air would only reduce the temperature. Less air would do the same, but it also gives a character to the flame which is desired when the blowpipe is being used for some special chemical or industrial purposes.
When the blowpipe supplies oxygen instead of air we get rid of the diluting atmospheric nitrogen, which has four times the bulk of the oxygen and takes no part in the combustion. The flame therefore be comes still smaller and much hotter, the combustion being, as it were, still further condensed. The hottest flame we can pro duce on a practical scale is that pro duced by burning with oxygen in a blow pipe the gas acetylene (q.v.), and this flame is now largely used as a source of intense heat for welding and cutting iron and steel.
A third type of flame is that associated with the name of the great German chemist, R. W. Bunsen, who in 1855 devised his well-known burner for use in chemical laboratories. The object was to mix coal-gas before it issued from the burner with sufficient air, not for complete combustion, but enough to make it burn blue. The construction is shown in fig. I, which is reproduced from Bunsen's original publication. The gas is led by a side-tube to the burner where it issues through a small punctured nipple into an upright tube. At the base are circular openings which in the usual form of burner can be varied in size or altogether closed by turning a metal collar. The passage of the jet of gas past these openings sucks in air, and if a light is applied at the top of the burner a flame is given which, as the saying is, "burns blue" and does not blacken objects to which it is applied. This principle has been applied to the construction of many different shapes of burners used with the gas mantle for lighting, and for innumerable purposes of heating in gas fires, cookers, furnaces and a great variety of scientific and industrial appliances. They are often called "atmospheric burners" to distinguish them from burners where the gas is burned without air being admitted to it in the burner tube.
In burners of the Bunsen type it will be noticed that when the air-ports are closed the flame has the same character as that of a candle, while when they are gradually opened the flame loses its yellow luminosity and soon becomes wholly "blue." This blue flame does not at first show a very clearly differentiated structure, but as the air-ports are opened more widely the flame acquires a plainly evident double structure. There is a distinct inner cone of flame, and outside this a sheet of flame of a slightly bulged conical form; with further increase of air the inner cone becomes smaller, more sharply defined and greener in colour. If the air supply is car ried beyond this stage the inner cone shows signs of instability in a tendency to retreat into the burner-tube and in the end will actually do so, and passing down the tube will light the gas at the nipple. The burner is now out of order and the gas is burning within the tube with an amount of air insufficient for its complete combustion. There may or may not be a second flame at the top of the tube where the combustion is completed. If there is not, the half-burned gas escapes and the malodorous and poisonous constituents contaminate the surrounding air. When the burner is in this condition the gas must be turned off and only re-lighted after the air-ports have been adjusted to give a smaller supply of air. The retreat of the inner cone down the burner tube ("lighting back,'.' or "back-firing") in the case of long or large burners such as those of an ill-adjusted gas fire has the character of a small explosion. The principles of the Bunsen burner and the characteristics of the flames of atmospheric burners can be readily understood by use of the "Flame Separator," shown in fig. 2.
This apparatus consists of two tubes (best made of silica ware) arranged telescopically so that the wider tube will easily slide, gas-tight, up and down the narrower one. When the sepa rator is fixed to a Bunsen burner with closed airports and the gas is ignited, a luminous flame is obtained, and if the ports are gradually opened this acquires a clearly two-coned structure. With further air supply the rate of inflammation of the mixture will eventually reach a state at which it is greater than the rate of up-flow of gases in the wider tube. The inner cone accordingly travels downwards against the stream until it reaches the mouth of the narrower tube, where the gas stream is more rapid, and there will remain if the air supply be kept steady. The outer cone remains at the mouth of the wider tube. On sliding down the wider tube until its mouth is below that of the narrower one, the normal two-coned flame is established at the mouth of the nar rower tube.
It is clear that in the atmospheric burner combustion takes place in two stages. If more air be admitted to the flame in the stage last described, the rate of inflammation will soon exceed that of the upward velocity of the gases and the inner cone will enter the narrow tube, travel down it, and ignite the gas at the nipple. A further phase of the flame cannot be obtained with the ordinary Bunsen burner, but by supplying air under pressure and gas by means of a T-piece to the separator the supply of air may be made greater than is necessary for complete combustion. The rate of travel of flame through the mixture then begins to fall and reaches a point at which it is no greater than the rate of upward flow of the gases. At this point a stable flame may be obtained at the mouth of the separator, in which complete com bustion is taking place independently of the outside air. The flame is therefore single-coned and of the blowpipe type. A further supply of air merely dilutes and cools the flame until it is finally extinguished. The first phases of flame are illustrated in figs. 4-8 of the Plate.
Much ingenuity has been expended on the construction of at mospheric burners so as to adapt them for special uses, and several types have become well-known by the names of their inventors. The aim may be to adapt the burner for use in lighting with a Welsbach mantle, in which case the hottest possible shell of flame in which the mantle can be bathed is desired. The best known upright burner in this connection is known as the "Kern," but it has been largely superseded by the inverted burner in which the flame is projected down into a mantle (fig. 3). By this inversion a higher duty is obtained from the gas and the inconvenience of any shadow of the burner is avoided. In designing burners for heating purposes, as for a gas fire, care has to be taken to obtain a set of noiseless flames, which are of the same dimensions and adjustable so as to give the right degree of aeration. The fire-clay "fuel" fits over the flames so as to be bathed in flame without actual contact with the inner cone.
It is important to remember that in an atmospheric burner consuming coal-gas the region between the two cones of the flame is rich in carbon monoxide, and that if in the use of such flames there is any intrusion into the inner cone by the object which is being heated, there is liability of the escape of carbon monoxide, so that if there is no flue provided there may in such circumstances be very undesirable and dangerous contamination of the air in an apartment.
In a candle flame the combustible gas is supplied by the wick which brings up from the little pool at the top of the candle a steady supply of melted paraffin wax. The wick being in the form of a flat band bends over so that its tip gets into the air and burns with a red glow. It thus remains constant in size and does not require the periodical snuffing which was one of the troubles of the old straight-wicked tallow candles. In lighting a candle the heat of the match flame melts and vaporizes a little of the paraffin and the flame thus pro duced continues to maintain a pool of melted wax at the top of the candle. The flame of the candle exhibits three distinct re gions; within the flame is what we may call "candle-gas" stream ing from the wick and becoming mixed on its outer borders with the atmospheric air. When the mixture has attained certain proportions it becomes ignitable; this occurs long before the admixture of air is sufficient for complete combustion.
The region where the candle gas is first ignited is clearly evident at the base of the flame where a thin sheet of bright blue colour is situated like the calyx of a flower ; it shades off about a quarter of the way up the flame. Outside this bright blue calyx there will be seen a fainter region of flame, and in this the incompletely burned gas is finding air sufficient for the completion of the burning process. This faint sheath or mantle extends round the whole flame from top to bottom. It is not easy to see it all in a candle unless the glare of the yellow part is masked by a screen, but it can usually be well seen in the flame of a piece of string. These two parts of a candle flame really correspond to the two cones of a Bunsen flame; the combustible gas is burning in two stages or in two sheets, which are quite close together in the case of the candle. Within this "blue burning" part of the flame is candle-gas streaming from the wick and being subject to a roasting process as it ascends. Now it may be said generally of car bonaceous gases that when they are strongly heated in absence of air they deposit charcoal, so that in this we find a reason why, as the candle-gas gets heated within the blue burning walls of flame, it begins to yield particles of carbon. Not only are these particles hot but as soon as they reach enough air they burn with a bright glow, and thus we have a sheet of bright yellow lu minosity which shows a maximum glow about half-way up the flame. Where the particles are sparse the glow is feeble; where they are superabundant, as may be the case near the tip of the flame, the glow is dull and merely ruddy, or indeed some of the particles may escape unburned as smoke.
This explanation of the structure of a candle flame may be verified by simple observations of a small coal-gas flame as it is turned down until the yellow part of the flame has entirely dis appeared, or when, before the yellow part has disappeared, a ring of wire is placed in the blue walls. The cooling effect at once suppresses the separation of carbon and the yellow patch of light disappears. (See also COMBUSTION and SPECTROSCOPY.) (A. SM.)