100 



Dr. Karl Heumann's Contributions to 



The differences between these numbers and those actually ob- 

 tained represent the gain in intensity of light by heating the 

 gas. This gain is stated in column 9 in per cents, calculated 

 on an equal consumption of gas. 



I. 



II. 



III. 



IV. 



V. 



VI. 





1. 2. 



Intensity of light, 

 expressed in candles, 

 according as the gas 

 was passed through 



Cold tubes. Hot tubes 



0-4 

 2-1 



5-6 



0-8 

 20 

 4-9 



0-6 

 2-9 

 6-1 



1-5 



2-7 

 7-6 



Consumption of gas, in 

 litres. 



Cold tubes. Hot tubes. 



16 

 38 



83 



23 



41 



123 



15-5 

 37'5 



82 



23 



41 



121 



Increase 

 in intensity 

 of light, in 

 per cents. 



50'0 



38-1 



8-9 



87-5 

 35-0 

 55-1 



I. 



II. 



III. 



IV. 



V. 



VI. 



Ml 



Decrease in 



consumption 



of gas, in 



per cents. 



34 

 1-3 

 1-2 





 

 1-6 



Intensity of light 

 for the decreased gas-con- 

 sumption (column 4). 



Calculated. 



Cold tubes. 



0-3£ 

 20 

 5-5 



0-8 

 2-0 



4-8 



Found. 



Hot tubes. 



0-6 

 2-9 

 61 



1-5 



2-7 

 7-6 



Gain in in- 

 tensity of 

 light, in 

 per cents. 



57-9 

 45-0 

 10-9 



87-5 

 35-0 

 58-3 



These results indicate that a considerable increase in intensity 

 of light is brought about in every case by heating the tube of 

 the burner. This increase is much more marked in the case of 

 small flames than in those produced by using large quantities 

 of gas. The reasons for this latter fact are probably these : 

 the cooling action of the burner is more marked in the case of 

 small than of large flames ; and when large quantities of gas 

 are passed through a heated tube, there is every likelihood 

 that a portion of that gas only will be raised to the tempera- 

 ture of the tube. In the experiments in which the gas was 

 passed through a wide platinum tube, it is noticed that the in- 

 crease in intensity of light is very marked with a large con- 

 sumption of gas. I account for this apparently contradictory 

 result as follows. By comparing the numbers expressing the 



