30 On Lockyer's Theory of Dissociation. 



able case not unfrequently occurs, that with the increase of 

 the dispersion of the solvent certain absorption-bands are dis- 

 placed, and others not. Hagenbach observed, for instance, 

 that bands I., III., and IV. of chlorophyll in an alcoholic 

 solution are situated more towards the red than those of chlo- 

 rophyll in an etheric solution, while band II. shows the same 

 position in both solutions ; and I have observed similar in- 

 stances in connexion with protosalts of uranium* and combi- 

 nations of cobaltf . 



Now Kundt has already (Joe. cit.) remarked that the same 

 rules hold for the absorption-spectra of gases as for those of 

 liquids. It is true that he adds, " Only it is questionable 

 whether, when hyponitrous acid gas, for example, is mixed 

 with various other transparent gases, the displacements of the 

 absorption-bands are so considerable as to be perceptible;" 

 this doubt, however, does not concern the rule above-men- 

 tioned, but only the possibility of testing it by experiment^:. 

 Hence it is an admissible assumption that, in like manner as 

 in liquids, so also in gases, media mixed with them act upon 

 the position of the absorption-bands, and that, in these as in 

 those, displacements of single bands may occur while the posi- 

 tion of others remains unchanged. 



If, therefore, in the sun-spots some lines undergo displace- 

 ment and others in the same place do not, motion is not the 

 cause, but the mixture of a foreign strongly dispersing gas, 

 which acts upon the lines displaced and not upon the others. 

 From this it further follows that curving of absorption-lines 

 of the sun-spots must not by any means always be interpreted 

 as motion of the absorbing gases in the direction of the line of 

 observation, but only when all the lines of a substance partake 

 of the curvature. 



That also bright lines of luminous gases under similar cir- 

 cumstances may suffer displacement " through mixture of 

 another, not luminous, or a vapour giving a continuous spec- 

 trum," Kundt has already pointed out (loc. cit. p. 620). 



Berlin, October 1882. 



* Vogel, Tract. Spectralanalyse, Nordlingen bei Beck, p. 248. 



t Monatsb. der Akad. der Wissensch. Berlin, May 20, 1878. 



\ Kundt previously doubted also the possibility of demonstrating an 

 anomalous dispersion in gases and incandescent vapours; but he has 

 recently succeeded in demonstrating it in the vapour of sodium (Wiede- 

 mann's Annalen, x. p. 321). 



