96 Dr. Karl Heumann's Contributions to 



for heat are " altogether imaginary." If the iron burner 

 were not maintained at the initial temperature of the expe- 

 riment by extraneous means, its action in withdrawing heat 

 and so causing decreased luminosity, would of course be 

 much less marked than it was in these experiments. Never- 

 theless I consider that these experiments prove the fact that 

 a notable diminution of light-intensity is brought about by the em- 

 ployment of metallic burners. 



The cooling action of a metallic burner may be observed in 

 another and very instructive manner. If an ordinary burner 

 be carefully examined shortly after the flame has been lighted, 

 a small quantity of moisture containing tarry matter will be 

 noticed collecting about the orifice of the burner. If the 

 burner be artificially cooled, this phenomenon becomes more 

 noticeable. If the burner be not cooled, the little drop of 

 liquid is soon evaporated ; but a brown residue of tarry matter 

 remains as testimony to the cooling action of the burner 

 upon the issuing gas. 



It has been already shown that a metallic wire, when held 

 in a flame, causes a considerable diminution in the luminosity 

 of the flame, but that this decrease does not take place when 

 the wire is heated to redness. Hence we should infer that 

 the action of a metallic burner in reducing luminosity would 

 disappear if the burner were strongly heated. Unfortunately, 

 in carrying out an experiment to prove the truth of this sup- 

 position, it does not appear possible to avoid raising the tem- 

 perature of the gas itself: hence the increased luminosity 

 cannot be altogether referred to the increased temperature 

 of the burner. 



In considering the phenomenon of the space which is main- 

 tained between flame and burner in the case of coal-gas burn- 

 ing under pressure, we have taken into consideration the 

 cooling action exercised upon the flame by the entering cold 

 gas. This cold gas would tend to decrease the luminosity of 

 the flame. Let us suppose that the flame could exist for a 

 short time independent of the propelling action of the gas ; 

 the temperature of the flame would not then be uniform 

 throughout ; by reason of the air diffusing inwards, and also 

 because of conduction and radiation, the outer portions would 

 part with more heat than the inner ; the hottest part of the 

 flame would therefore be at, or about, the centre. Suppose 

 now that cold gas were suddenly admitted to the flame from 

 beneath, it is evident that the greatest absorption of heat 

 would take place in the lower parts of the flame, and that the 

 hottest point would no longer be at the centre, but towards 

 the upper extremity of the flame. At the same time every 



