94 Dr. Karl Heumanu's Contributions to 



the surface of the burner is large, and when the burner itself 

 is made of thick metal : this is especially marked with bur- 

 ners in which the gas issues from a number of small holes 

 pierced in a thick plate of iron. If the supply of gas to 

 such burners be small, the light effect is almost nil. The 

 heating effect of such (roasting) burners is also, I think, small 

 compared with the consumption of gas. If the supply of gas 

 be increased in such burners, the tops of the flames become 

 luminous, but the lower parts always remain blue. 



If a burner causes decreased luminosity by acting as a 

 withdrawer of heat from the flame, one would expect that 

 burners made of material which is a good heat-conductor 

 would exercise a greater action in this respect than those made 

 of material having a lower conductivity for heat. Neverthe- 

 less I am assured that in practice the only fault found with 

 metallic burners is the readiness with which they undergo 

 oxidation ; and a Commission by the English Board of Trade * 

 has reported unfavourably as regards the use of porcelain or 

 steatite burners, so far as increased luminosity is concerned. 



The differences in luminosity obtained by the use of differ- 

 ent burners may be so small as only to be distinguished by 

 very careful experiments. I found that the ordinary bur- 

 ners were of no use in performing comparable experiments, 

 a small difference in the cavity of two burners, or in the 

 position of the stopcock &c, often vitiating the experiment. 

 I therefore had burners of different material specially pre- 

 pared : these were exactly similar in construction, of the same 

 size, and pierced with the same instrument with a cross-shaped 

 opening ; they only differed in the material of which they 

 were constructed. These burners were attached to similar 

 tubes ; and the light obtained from each under exactly the 

 same conditions was tested by means of the photometer. 



Intensity of light, Gas-consumption, 

 in candles. in litres. 



Steatite burner . . 0-9 36 



Iron burner ... 0*85 37 



Besides the errors inherent in the photometrical methods, it 

 was found that other errors rendered the results in the case 

 of the iron burner of little value. Within the first few minutes 

 after lighting the gas, the intensity of the light, as also the 

 consumption of gas, showed most remarkable variations. 

 These were evidently to be traced to the heating of the burner, 

 whereby the opening was decreased in size, and hence the 

 velocity of the issuing gas was also decreased. In order to 

 * Polyt. Centralblatt, 1872, p 138, &c. 



