Second Spectrum of Hydrogen. 335 



" Le spectre de notre melange gazenx a presente : 



" (1) Les raies brillantes de l'hydrogene. 



" (2) Les raies et les bandes lumineuses du carbone, con- 

 fonnes au spectre de l'oxyde de carbone dessine par MM, 

 Piiicker et Hittorf et an spectre du carbone de M. Morren. 

 Nous avons verifie Inexactitude remarquable de ces dessins en 

 operant sur line tube de Piiicker remplie d'oxyde de carbone 

 rarefie. 



" (3) En outre nous avons reconnu l'existence d'un groupe 

 particulier de bandes et de raies, qui n'ont ete signalees a 

 notre connaissance par aucun observateur *, En*effet, depuis 

 le jaune jusqu'au vert on apergoit une multitude de bandes 

 etroites et brillantes equidistantes ou a peu pres separees par 

 de fines raies noires. Le tout oftre l'aspect d'une serie de 

 cannelures delicates et extremement reserrees : elles sont 

 surtout manifestes a partir de la division 25 de notre micro- 

 metre et jusque vers la division 65. La portion jaune du 

 spectre, voisine de la raie du sodium, les presente avec le plus 

 grand eclat." 



As far as the first point is concerned, there is no doubt of 

 its correctness, since the mixture contained more than 98 per 

 cent, of hydrogen. But it is different with the second. The 

 bands mentioned here can no more be ascribed to the element 

 carbon, but form the spectrum common to all hydrocarbons, 

 which was first exactly studied by Swan f, and has since 

 formed the subject of investigation by many spectroscopists. 

 In my paper " On the Spectra of Comets, &c." % I have given 

 the history of this spectrum somewhat fully, and have given 

 reasons (chiefly from the researches of Angstrom and Thalen §) 

 why the above interpretation appears to me the right one. 

 Since, according to these investigations, which have been 

 recently confirmed by the exhaustive researches of Liveing 

 and Dewar ||, the acetylene which always forms in hydro- 

 carbons under the influence of the electric current is to be 

 regarded as producing these radiations, its appearance in the 

 spectrum of the mixture in question was not to be expected, 

 but forms a direct proof of the correctness of the explanation 

 of the spectroscopic phenomena observed with hydrocarbons 



o 



first given by Angstrom and Thalen. 



But if, as, strange to say, is still the case with most spectro- 



* It is remarkable that the passages treating of this point in Piiicker 

 and Hittorf s paper should have escaped the notice of the authors, since, 

 from the remarks made above under No. 2, we must assume that this 

 paper was known to them. 



t Edinb. Trans, vol. xxi. 



\ Memoires de V Acad, de St. Petersboury, vol. xxviii. No. 2. 



§ Acta UpsaL ser. 3, vol. ix. || Proc. Roy. Soc. No. 201 and 205, 1880. 



