124 Dr. W. Marshall Watts on the Spectra 



end at 3873 with a fairly strong edge. The fine shading lines 

 are closer together at the less refrangible edge 4315 than 

 those which form the more refrangible edge at 3871*. 



There are three marked lines at 4381*93, 4371*31, and 

 4365*01 superposed upon the first portion of this group/ 



These three lines seem to have been first noticed by Liveing 

 and Dewar, who attributed them to cyanogen, an opinion 

 which they retracted later f . In 1889 Kayser and Runge % 

 pointed out that they are probably part of the " Swan " 

 spectrum ; and adduced evidence from the relationship of 

 their wave-lengths to those of the other groups of the 

 " Swan " spectrum. 



Deslandres§ in 1891 re-discovers these lines and calculates 

 what their wave-lengths should be. He makes the erroneous 

 statement that they are not produced by the combustion of 

 hydrocarbons ||, but only in the arc and in the flame of 

 cyanogeu. Deslandres 5 paper is entitled "A new method 

 for the discovery of faint bands, &c." He says: " I have 

 studied the band-spectrum attributed to hydrocarbons or to 

 carbon itself by a new method by which I have been able 

 to complete it, and to add to it with certainty three new 



bands at 438*19, 437-13, and 436*5 The application of 



the general law of arrangement of bands which I gave in 

 1887 permits me to conclude that these three bands belong 

 certainly to the hydrocarbon bands." 



The calculations of Kayser and Runge (of 1889) are 

 criticised by Deslandres in the Journal de Physique for June 

 1891, who there repeats the statements quoted above from 

 the Comptes Rendus. 



There is no doubt that a numerical relationship exists 

 between the wave-lengths of the lines of the '* Swan " 

 spectrum, that is between C a C y C* O e and "the three"; 

 but the group /and more refrangible group will not fall into 

 line with the rest. Deslandres' formulae are approximately 

 correct, as may be seen from the following recalculation with 

 the more exact data now available ; but we are still in want 

 of sufficiently good measurements of the red lines and of 

 some others. 



* The fine-grained shading at the less refrangible edge and the coarse- 

 grained shading at the more refrangible edge are seen in the photographs 

 of the Bunsen spectrum given by Eder, Szb. Wien, xciv. p. 403 (1886), 

 and by Vogel, Szb. Berlin, xxi. p. 523 (1888), as well as in the repro- 

 ductions which accompany this paper. 



t Liveing and Dewar, P. R. S. xxxiv. p. 418 (1882). 



% Kayser and Eunge, Abhandl. Berlin Akad. 1889. 



§ Deslandres, C. R. cxii. p. 661 (1891). 



|| See the photograph of the spectrum of the Bunsen flame given by 

 Eder, Szb. Wien, xciv. p. 403. 



