the Theory of Luminous Flames. 95 



remove this source of error, the tube to which the iron burner 

 was attached was caused to pass through a piece of glass tubing 

 closed at the bottom with a cork and at the top with a watch- 

 glass, the top of the burner just appearing above the latter. 

 This tube was filled with water, which was maintained at 15°. 

 The experiments, the results of which are detailed, were con- 

 ducted with every precaution, and each set under exactly 

 similar conditions *. 





Steatite burner. 





Iron burner. 



Intensity 





Gras-con- 





Intensity 





Gas-con- 





of light, in 



Mean. 



sumption, 



Mean. 



of light, in 



Mean. 



sumption, 



Mean. 



candles. 



0-121 

 012 j 





in litres. 





candles. 





in litres. 





012 



14-51 



14-t ; 



14-6 



0-1(3 1 

 0-18 I 

 017 J 



0-17 



19-5 -] 

 20-0 

 19-0 

 19-5 J 



19-5 



0-3 1 

 0-3 j 





221 





0-31 





25-5 1 





0-3 



21 



21-3 



0-3 I 



0-3 



26-0 L 



25-7 





21 J 





0-3 J 





25-5 J 





0-9 ] 





39] 





0-91 





411 





1-0 



0-96 



39 [ 



39 



0-8 I 



0-87 



42 L 



41-7 



0-98 J 





39 J 





0-9 J 





42 ) 





It was not considered advisable, on account of the arrange- 

 ment of the testing-apparatus, to further increase the supply 

 of gas. 



From the second and third series of numbers it is at once 

 apparent that the steatite is preferable to the iron burner, inas- 

 much as an equal intensity of light is obtained with a less con- 

 sumption of gas. The first series of numbers does not, how- 

 ever, appear to bear out this result, inasmuch as the intensity 

 of light obtained from the gas burning from the steatite 

 burner is rather less than that obtained in the experiment 

 with the iron burner for an equal gas-consumption. 



If, however, the results obtained from the iron burner be 

 compared among themselves, it will be seen that this in- 

 crease of light does not always bear the same relation to 

 increase of gas consumed. The reason for this is to be found 

 in the fact that the cooling action of the burner is greater 

 in the case of small than of large flames. 



If the results be plotted in curves, the difference between 

 the two burners is very apparent. 



I therefore feel justified in contradicting the statement of 

 the Board-of-Trade Commission, to the effect that the advan- 

 tages of burners made of material of low conducting-power 



* In the original paper the precautions are detailed. 



