374 Prof. Smithells on the Source of Light in Flames. [March 12, 



Some light is no doubt due to the completion of the oxidation, the 

 carbon monoxide forming carbon dioxide and the hydrogen forming 

 water, but the intensity of this portion of the light is inconsiderable 

 in the spectroscope, and in the visible spectrum not characteristic. 



The flame of cyanogen presents special points of interest. It has 

 been shown that the sharp differentiation of the flame into an inner 

 rose-coloured cone and an outer blue one, corresponds to the combus- 

 tion of the gas in two steps, the first being the oxidation of carbon, to 

 carbon monoxide, and the second the oxidation of carbon monoxide to 

 carbon dioxide.* Admixture of air with the gas before combustion 

 renders it possible to separate the two parts of the flame in the cone 

 separating apparatus, and when the distance between them exceeds a 

 certain limit and the gases are dried, the outer cone is quenchfid wlien 

 a bottle of dried air is held over it. [Experiment shown.] This 

 behaviour accords with the well known experiment of Prof. Dixon on 

 the combustion of carbon monoxide. According to the view which has 

 been developed in the foregoing, it would be expected that the light 

 emitted by the inner cone of a cyanogen flame should be due to the 

 carbon monoxide which is produced there, and if the Swan spectrum 

 be really due to that substance then the Swan spectrum should be 

 Been. As a matter of fact, the inner cone of a cyanogen flame gives a 

 brilliant spectrum, in which, however, only one band of the Swan 

 spectrum is distinctly developed. It is possible that the liberation of 

 nitrogen from cyanogen during its combustion may have a disturbing 

 influence. In any case it is very striking that when cyanogen is 

 burnt in oxygen instead of air the Swan spectrum is seen to be com- 

 pletely and brilliantly developed, and on the whole the evidence 

 derived from a cyanogen flame appears to strengthen tlie view which 

 associates the Swan spectrum with the production of carbon monoxide. 



Keviewing the evidence which has been oifered,it appears that the 

 primary source of light in flames is to be found in the intense vibra- 

 tory motion which is determined by the act of chemical union. This 

 is seen in the phosphorescence of phosphorus, in the flame of hydrogen, 

 and at the base of the flames of the hydrides of silicon and carbon. 

 A secondary source of light arises when the temperature effect of 

 the primary combustion causes the glow of a product or partial 

 product of combustion. This is seen in the white flame of phos- 

 phorus, in the brightest part of the flame of silicon hydride, and 

 in the bright yellow-white part of ordinary hydrocarbon flames. 



The question of the luminosity of flames containing the vapours 

 of salts introduces new problems, the elucidation of which is far from 

 beiug complete. This question, however, cannot be considered on 

 the present occasion. FA S 1 



Smithells and Dent, Journ. Chen). Soc. 65, p. 603 (1894). 



