1882.] On the Origin of the Hydrocarbon Flame Spectrum. 411) 



Magnesium," wherein the following passage occurs: — " The chemical 

 atoms of magnesium are either themselves capable of taking up a 

 great variety of vibrations, or are capable by mutual action on each 

 other, or on particles of matter of other kind, of giving rise to a 

 great variety of vibrations of the luminiferous ether; and to trace 

 satisfactorily the precise connexion between the occurrence of the 

 various vibrations and the circumstances under which they occur, will 

 require yet an extended series of observations." (" Proc. Hoy. Soc," 

 vol. 32, p. 203.) 



Specific spectra have been satisfactorily proved to emanate from the 

 compound molecules of cyanogen, water, and magnesium-hydrogen, so 

 far as we can interpret in the simplest way the many observations pre- 

 viously detailed. The fact that a fluted spectrum is produced under 

 certain conditions, by a substance which does not give such a spectrum 

 under other conditions, is of itself a proof that the body has either 

 passed into an isomeric state or has formed some new compound ; but 

 we are not entitled to assert, without investigation, which of these two- 

 reasonable explanations of the phenomena is the true one. There is, 

 however, a spectrum to which we have had occasion to refer in the- 

 papers on the spectra of the compounds of carbon, which closely 

 resembles that of a compound substance, and which we, in common 

 with some other spectroscopists, have been led to attribute to the 

 hydrocarbon acetylene, without, however, being able to bring forward 

 such rigid experimental proofs of its origin as we have adduced in 

 the case of the three substances above referred to. In other words, 

 the experimental evidence that the hydrocarbon flame spectrum is 

 really due to a hydrocarbon was always indirect. Thus, we showed 

 that many flames containing carbon, such as those of hydrogen mixed 

 with bisulphide of carbon or carbonic oxide, and the flame of 

 cyanogen in air, did not give this spectrum, and these particular 

 ilames are known, from the investigations of Berthelot, to be in- 

 capable of generating acetylene under conditions producing incom- 

 plete combustion. On the other hand, we found that a flame of 

 hydrogen mixed with chloroform, which easily generates acetylene, 

 gives the hydrocarbon flame spectrum in a very marked manner, and 

 it is known that the ordinary blow-pipe flame, in which the same 

 spectrum is well developed, contains this hydrocarbon. 



These and other experiments point to the intimate relation of 

 hydrogen and carbon in the combined form of acetylene to the pro- 

 duction of this spectrum during combustion. In our various observa- 

 tions on the spectrum of the electric arc taken in different gases, the 

 flame spectrum was always noticed, and seemed to be independent of 

 the surrounding atmosphere. In the mode in which those experi- 

 ments were conducted, it was easily shown that the carbons were 

 never free from hydrogen, and that the gases always contained traces 



