ON GASEOUS EXPLOSIONS. 217 



luminous. These variations are compared in the accompanying table 

 with the approximate composition in volumes of air to one of gas before 

 ignition. The form of the curve depends to some extent on the shape, 

 size, and nature of the flame. It would not be the same for a bat's-wing 

 or Argand flame. The rate of gas consumption was maintained approxi- 

 mately constant, and the size of the flame varied with the strength of 



mixture. 



Total radiation of Meker burner per cent, of heat of combustion : 



Total Radiation per cent. . . 105 123 140 159 141 146 17 

 Ratio Air/Gas by vol. ... 5 4 3 2-5 15 1 



The gas was in all cases completely burnt. The ratio of air to gas 

 before ignition merely describes the nature of the flame. Mixtures in 

 these proportions, if burnt without further addition of air, would not, 

 of course, radiate the same per cent, of heat. With the ratio air 

 to gas = 5, the duration cf the luminous flame was estimated at about 

 a fiftieth of a second. 



Intrinsic Radiance of Flame. 



The intrinsic radiance of a flame has the same meaning in respect 

 of total energy of radiation that intrinsic brilliance or brightness has 

 in respect of luminosity. It may be measured by the radiation emitted 

 per unit area of surface, but in the case of a flame which is more or 

 less transparent the radiation comes from a finite thickness, and must 

 be measured per unit of solid angle subtended. This measurement 

 may conveniently be effected by means of a total radiation pyrometer 

 of any kind in which an image of the flame is formed on a radiometer 

 or bolometer. A Fery mirror pyrometer was employed for this purpose, 

 the instrument being focussed on the flame at a height of 4 to 5 cm. 

 above the grid, where the flame was steady and sensibly homogeneous. 



With this restriction it was found that the intrinsic radiance of the 

 Meker burner did not vary materially as the air supply was reduced 

 from that necessary for complete combustion, until the inner cone 

 became so large that the flame could no longer be regarded as sensibly 

 homogeneous. This showed that the increase of total radiation simul- 

 taneously observed was due chiefly to increase in the size of the flame, 

 and that the increase of thickness of the flame was compensated either 

 by fall in temperature or by increase in absorptive power. The thick- 

 ness of the flame at the height focussed in the pyrometer varied from 

 2*8 cm. with full air supply to 3'6 cm. when the inner cone was 3 cm. 

 high and just cleared the area focussed. 



In order to determine the manner in which the intrinsic radiance E 

 varied with the thickness x of the flame in the line of sight, and to 

 measure the coefficient of absorption, six precisely similar burners were 

 mounted in a row along the axis of the radiation pyrometer, which 

 was focussed in such a way that the reading was the same for any one 

 of the burners singly or for any combination of the same number of 

 burners at different distances. The pressure of the gas supply was 

 regulated to a constant value, and care was taken to prevent the air 

 in the laboratory from becoming contaminated, which produced a notable 



