334 Dr. Frankland on some Phenomena [March 8, 



light takes place in a less rapid ratio. One of the columns headed 

 " Calculated," in the above tables, exhibits the illuminating power 

 calculated from the constant just given, and it will be seen, that these 

 calculated numbers nearly coincide in most cases with the observed 

 amount of light. 



In explaining the cause of the above phenomena, the speaker 

 referred to the conditions upon which the light of ordinary flames 

 depends. He showed, that in these flames there are two sources of 

 light, viz. incandescent gaseous matter, and incandescent solid matter ; . 

 but that, practically, 99 per cent, of the light of such flames owed its 

 origin to the second of these sources. In gas, candle, and oil flames, 

 the incandescent solid matter consisted of carbon in a minute state of 

 division. The amount of light emitted by these flames depended, 

 within certain limits, first, upon the quantity of solid particles of 

 carbon existing, at any given moment, within the flame ; and secondly,, 

 upon the temperature to which these carbon particles were heated. 

 Now, the temperature of a flame might be affected by imperfect 

 combustion in rarefied air ; but it had been proved, by the analysis of 

 the products, that combustion was equally complete in the above 

 experiments under all pressures ; in fact, it was found that complete 

 combustion could be far more easily secured in rarefied air, than in air 

 at the ordinary atmospheric pressure. Other experiments also showed, 

 that the temperature of a flame was not materially affected by the 

 pressure of the air in which it was burning ; consequently, it was 

 inferred that the diminution of luminosity in rarefied atmospheres, 

 was not due either to imperfect combustion, or to reduction of 

 temperature. 



The diminution of light must therefore arise from the decrease of 

 the amount of solid carbon separated within the flame ; and this the 

 speaker believed to be due to the admission of oxygen in larger quan- 

 tities into the interior of the flame when the atmosphere was rarefied. 

 It was shown by experiment, that the admission of a comparatively 

 small amount of air, and consequently of oxygen, into the interior of 

 a gas flame, immediately reduced the illuminating power of the latter 

 to a very marked extent ; the carbon particles, instead of being 

 separated as such in the interior of the flame, being at once oxidized 

 to carbonic oxide. This increased access of oxygen to the interior of 

 a flame burning in rarefied air, was believed to be due to the greater 

 mobility of the particles of expanded gases, which enabled the gases 

 of the flame and the circumambient air to commingle more rapidly 

 than at ordinary atmospheric pressure. 



The cause of the less rapid decrease of the light of flames burning 

 in atmospheres below 14 inches of mercurial pressure was due to the 

 comparative prominence assumed by the light of the incandescent 

 gaseous matters of the flame at such high stages of rarefaction ; this 

 gaseous illumination being affected by pressure to a much less extent 

 than that afforded by incandescent carbon particles. 



In his celebrated researches on flame, Davy had not overlooked the 



