THE CAUSE OF THE LIGHT OF FLAMES. 697 



the flame wherefrom it was deposited. It is well known that the tem- 

 perature of the flame varies in various places, and Magnus's experi- 

 ments have proved that from heavy hydrocarbons at a less high tem- 

 perature a hydrogenous tarry product besides hydrocarbon is also 

 eliminated. The soot, whose analysis I give, was obtained from a 

 bat's-wing burner by allowing a small silver basin, filled with water, 

 to dip for about two or three minutes in the flame. Benzine removed 

 traces of a solid yellow body, but the small amount of it prevented its 

 being further investigated. Alcohol, and alcoholic solution of caustic 

 potash, and dilute sulphuric acid, dissolved nothing. 



After being carefully and repeatedly washed with boiling water 

 and dried at 130, 0.206 yielded: Carbonic acid, 0.6985; water, 

 0.0195 ; ash, 0.0020 which amounts in 100 parts to. 



Containing Ash. Free from Ash, 



Carbon 96.446 97.390 



Hydrogen 1.051 1.061 



Ash 0.970 



Oxygen 1.533 1.533 



I attribute the presence of oxygen to a small amount of water, 

 which, even at 130, was still retained, and this, when deducted, gives 

 the composition of 100 parts of soot free from water and ash as con- 

 sisting of carbon, 99.095 ; hydrogen, 0.905. 



This analysis is in accordance with the chemical composition of 

 the soot of the flame, and with the well-known behavior of heated 

 hydrocarbons. 



2. " How could the light of a flame be as transpai-ent as in reality 

 it is, if it was filled with solid carbon-particles ? " asks Dr. Frankland. 



In reply to this, it must be admitted that one is able to read the 

 writing held behind the flame of a bat's-wing burner. It is, however, 

 easily observable that the flame is more transparent in the lower non- 

 luminous j)ortion. The reading becomes also more difficult through a 

 flame of greater thickness, and impossible through the flame of a 

 candle or petroleum-burner. If, as is proved hereby, the transparency 

 of a flame is only very limited, it may also be remembered that one 

 can also read the same writing through media which are known to be 

 filled with solid particles. The fact that solid bodies are by prefer- 

 ence apt to become light-radiating is not at all changed by this, and 

 thus far it is demonstrated only that there can be but one solid body 

 to which the luminosity of flame can be attributed. If we consider, 

 therefore, all the facts, we can draw only one conclusion, namely, that 

 the light of our illuminating flame comes from incandescent carbon- 

 molecules, and that the old view is still to be retained. 



Experience teaches that, for the artificial production of light, a high 

 temperature is requisite before all things. Temperature is, however, 

 that part of the total heat of a body which influences the surrounding 



