128 



Profs. G. D. Liveing and J. Dewar. 



[June 1 5, 



fectly pure Iiydrocarbon gives the series of bands we attribute to 

 cyanogen.* 



Observations on Flames. 



The temperature produced by the combustion of hydrocarbons and 

 other non-nitrogenous organic bodies well supplied with oxygen, is not 

 sufficient to induce the combination of nitrogen with carbon, so that 

 the cyanogen spectrum is absent from such flames. We know, how- 

 ever, that hydrocyanic acid is often produced in the oxidation of 

 organic bodies containing nitrogen, and that ammonia reacts with 

 carbon at a white heat, producing hydrocyanic acid and hydrogen. 

 Such actions led us to expect that the cyanogen spectrum ought to 

 appear in the flames of organic compounds containing nitrogen, pro- 

 vided the temperature were sufficient to render the radiation of this 

 substance sufficiently intense. Our first experiments did not succeed. 

 The most careful examination by the eye of the spectrum of a 

 hydrogen flame which had passed through a solution of hydrocyanic 

 acid or of a flame of alcohol containing nitrobenzol or nitrite of ethyl, 

 did not result in any recognition of the strong cyanogen groups. 

 This failure led to a chemical examination of the composition of the 

 gases withdrawn from the interior of such flames, in order to ascertain 

 the combustible mixtures which react during combustion to produce 

 hydrocyanic acid. The gases w r ere extracted from the flame with 

 an apparatus similar in principle to that employed by Deville in 

 his " Chemical Researches on Flame." When coal gas is passed 

 through a solution of ammonia and burnt, the flame gases contain 

 hydrocyanic acid and acetylene, but if oxygen is well supplied to the 

 flame no cyanogen reaction is given by the extracted gas. Car- 

 bonic oxide mixed with ammonia in the same way gave no trace of 

 hydrocyanic acid during combustion : even when a large quantity of 

 the mixture was burnt and the flame gases continuously withdrawn 

 no appreciable cyanogen reaction could be detected. Similarly hydro- 

 gen mixed with a little carbonic acid and ammonia gave no cyanogen 

 reaction. When hydrogen is passed through ammonia solution 

 mixed with chloroform, tetrachloride of carbon, bisulphide of carbon, 



* It is worthy of note that the strong carbon line wave-length 2478*3 present in 

 both tbe arc discharge and in the spark discharge in carbon compounds at atmo- 

 spheric pressure, is not found in the spectrum of tbe spark in cyanogen at low 

 pressure. We have tried to obtain a photograph of it from a " Plucker " tube 

 fitted with a quartz end, and placed end-on in front of the spectroscope, but 

 found no trace of it. As this line appears in the spectrum of the name of cyanogen, 

 its absence from the spark discharge in cyanogen of low tension seems intelligible 

 only on the supposition that the discharge is selective in its course, and lights up 

 only certain of the substances present, or else that the quantity of carbon vapour 

 present at any instant is so minute, as to produce no sensible effect on the photo- 

 graphic plate.— July 10, 1882. 



