the Theory of Luminous Flames. 105 



These burners, however, heat the gas but slightly ; more- 

 over, from their construction, the heat used for increasing the 

 temperature of the gas-stream is withdrawn from the flame 

 itself, which is thereby cooled. The flame is caused to heat 

 the tube of the burner, which then parts with its heat to the 

 cold gas, being itself thereby reduced in temperature, and 

 therefore brought into a condition in which it may exercise a 

 cooling action upon the flame. Heat is withdrawn from the 

 flame to be restored in diminished quantity, through the 

 agency of the stream of gas. The logic of such burners is 

 extraordinary. 



I will not, on a priori grounds, deny that these burners 

 may be useful under certain circumstances. In these cases, 

 actions other than the heating of the gas are probably to be 

 regarded as the true causes of the increased luminosity. 



Burners whose effect depends upon their power of heating the 

 gas before combustion, must be constructed so that the heat is 

 not derived from the luminous flame itself, but either from the 

 products of combustion or from an external source. Whether 

 the quantity of gas or other materials consumed in the latter 

 method would be repayed by the increased luminosity, is a 

 purely economical question into which I cannot enter. 



It might be supposed that the increase in luminosity of a 

 flame, brought about by heating the burner-tube, was to be 

 traced to the higher temperature to which the particles of 

 separated carbon are raised. (I have already shown that the 

 increase is not to be attributed to chemical decomposition 

 within the burning gas.) But on looking more narrowly at 

 the burning gas it will be seen that the flame-mantle becomes 

 larger, that it stretches downwards towards the orifice of the 

 burner, and that the lower portion, which was formerly blue, 

 begins also to become luminous. 



The conclusion which I draw from these phenomena is, 

 that the heated burner causes an earlier separation of carbon in 

 the flame than would otherwise take place. Increase in the 

 intensity of the light emitted by the glowing carbon particles, 

 and earlier separation of these particles in the flame, are the 

 actions brought about by heating the tube of the burner. The 

 total increase in luminosity registered by the photometer is 

 the sum of these two actions. 



The character of a flame is often conditioned, not only by 

 the total luminosity, but also by the intensity of the light — 

 that is, by the quantity of light emitted from a single con- 

 stituent part of the luminous flame. I think that too little 

 attention has hitherto been paid to this important fact. Al- 

 though the photometer is unable to distinguish between a 

 small but intense, and a larger but less-brilliant flame, never- 



