370 THE CHEMISTRY OF THE SUN. [CHAP. 



the same way as with liquids, added media also affect the posi- 

 tion of absorption-bands in the case of gases, and that in this case, 

 as in the other, displacements of certain bands occur, while the 

 position of others remains unaltered." l 



Indeed the more this question of " shifts " is looked at the 

 more important seem to be the consequences which may follow 

 from its complete study. Lecoq de Boisbaudran and Ciamician 

 may be cited among those who have pointed out that cor- 

 responding lines and bands in the spectra of the elementary 

 bodies themselves may owe their origin to some such process. 



If all the information now available had been before me in 1878 

 it would have been necessary to point out, that with regard to the 

 position of lines due to the same substance, existing differently 

 compounded, or under different physical conditions, there were 

 two analogies open to us to throw light upon the possibility of 

 common lines being detected in the chemical elements. The 

 analogy of the salts of calcium above referred to, telling us 

 that such lines would be coincident ; the analogy as supplied by 

 the shifts telling us that such lines would probably not be coin- 

 cident, because they represent the vibrations of still finer mole- 

 cules which might have been shifted on the formation of the 

 element itself, and shifted differently probably in each element 

 if the same molecule entered into the formation of several. 

 If the evidence had been thus stated there would have been 

 ground, I think, to accompany it with the suggestion that the 

 shifts observed in the spectra of the elementary bodies would 

 not be so great as in the case of the higher complexes ; 

 in fact that lines apparently coincident with small dispersion 

 might be shown to be not truly coincident when higher dispersions 

 were employed. It may be urged that this would render very 

 doubtful any argument depending upon the basic-line part of 

 the hypothesis. It certainly requires the statement touching 

 basic lines to be modified, for now being in presence of two 

 1 Nature, vol. xxvii. p. 233. 



