produced in the fire, and not that of the coal itself. Only gases burn with a flame. In those cases where a solid or liquid seems to do so, it will he found that it is either volatilized or decomposed by the heat of the combustion, and thus converted into gas before it burns.
Let us now begin our study of the phenomena of flame by considering the flame of hydrogen. Suppose we have hydrogen-gas flowing from a round jet. Just over the opening we have a round column of pure hydrogen. This gas, being lighter than air, and being forced out under some pressure, rises. As it rises it mixes with the air, and we immediately have, surrounding the jet of hydrogen, a layer of mixed hydrogen and air which is inflammable. If we now apply a light this mixture takes fire. The hydrogen unites with the oxygen of the air, forming steam, which is carried away by the current of hot gases; more hydrogen is continually supplied from the jet, and more oxygen from the atmosphere: and thus we have a continuous formation, as fast as it is burned, of this inflammable mixture of hydrogen and oxygen around the central column of hydrogen.
Evidently, then, the flame must be hollow. That it is so may be shown by the familiar experiment of quickly inserting the phosphorus end of a match into the center of the flame, where it may sometimes be held until the wood of the match is burned through without taking fire. The flame can not spread inward, because there is nothing there to support combustion; nor outward, because there is nothing there to burn. The flame is simply that part of the current of gases where the chemical action takes place, and where, consequently, the heat is produced. It is more nearly a place than a thing. If we leave out of account for a moment the chemical changes, we may compare the current of gas which flows from the tube to a metallic rod which is being slowly pushed through a fire. The portion in the fire glows, and, as the rod moves on, different portions of it glow, while the glowing spot, which may be compared to the flame, remains stationary. In the hydrogen-flame the rod is of gas and invisible. We see only the spot which glows, and, as this is stationary, we are apt to regard it as an object by itself, instead of considering it as a spot in a constantly flowing stream.
The hydrogen-flame gives out very little light. If burned from a metallic jet, the flame becomes almost invisible. What little light such a flame does emit, it emits because the gases of which it is composed are hot; but ordinary gases, when heated to the temperature of a flame, emit very little light.
To render a flame luminous in the sense in which an ordinary gas flame is luminous, we must introduce into it some solid which is not converted into gas at the temperature of the flame. The light of a common gas-flame is due to innumerable small particles of carbon, which are separated from the gas in the interior of the flame and are heated white hut while passing through the flame. Illuminating gas
