the Theory of Luminous Flames. 97 



part of the flame must participate in the cooling effected 

 by the entering cold gas ; those portions whose temperature 

 was already but very little above that at which separation of 

 carbon occurs would be cooled below this point, and would 

 therefore be rendered completely non-luminous, whereas the 

 hotter portions of the flame would have their luminosity de- 

 creased, but not entirely removed. Hence it follows that the 

 luminosity of the whole flame will be decreased the more the 

 colder was the entering gas. The total luminosity, on the 

 other hand, would be increased, if the cooling action of the 

 entering gas were neutralized by raising this gas, before it was 

 allowed to enter, to the temperature of the flame. 



I had reached this point in my theoretical considerations, 

 when I met with a paper on the increase of luminosity occa- 

 sioned by heating the gas, by A. Vogel *, and the Report of 

 the English Board of Trade, already referred to, the results 

 of which are contradictory to those of Vogel. Vogel con- 

 ducted the gas through a V-tube placed successively in ice, 

 freezing-mixtures, boiling water, and heated paraffine, and 

 used a steatite burner attached to the V-tube. Represent- 

 ing the luminosity at 18° as 100, he found that at 0° this was 

 reduced to 76 and 85, at -20° to 33 and 44, at +100° to 

 104, and at +160° to 118. The influence of decreased tem- 

 perature is here very marked ; that of increased temperature 

 is somewhat less so. 



The results of the Board-of-Trade experiments are directly 

 opposed to those of Vogel ; they indicate no difference in lumi- 

 nosity for the temperatures of 0° and 145°. So far as I know, 

 Vogel has not answered these statements ; and although I have 

 unfortunately been as yet prevented from repeating the experi- 

 ments, I nevertheless venture to put forward the following sug- 

 gestions as perhaps likely to throw light upon the subject 

 when it is regarded from a general point of view. 



And first as regards the experimental data in Vogel's paper. 

 When we consider the small conductivity for heat of gas, it 

 can scarcely be admitted that the gas which passed through 

 the V-tube maintained at0°, -20°, 100°, and 160° was really 

 itself at these temperatures. A tube bent repeatedly might 

 have been advantageously substituted for the V-tube, and a 

 small thermometer inserted near the burner in the tube con- 

 ducting the gas would have enabled the experimenter to read 

 off the actual temperature of the gas just before it reached the 

 burner. 



In my own photometric observations I found it of the 



* Bayerisches Industrie- vnd Gewevbeblatt, 1869, p. 124; Wagner's 

 Jahresbericht, 1869. 



Phil Mag. S. 5. Vol. 3. No. 16. Feb. 1877. H 



