228 THE CHEMISTRY OF THE SUN. [OH. xvi. 



One important side of this work was, that it enabled any 

 spectroscopist or any chemist who chose to take the trouble 

 and devote the time to it, to examine as to the existence of im- 

 purities in different substances : not to determine the absolute 

 amount of impurity, but enabling him to say that in specimen 

 A there is a greater impurity of x than there is in specimen 

 B, and so on. The statements were not absolute, they were 

 relative, but if only relative they were certainly a very great 

 advance on anything which had been done before, because until 

 this question of long and short lines was introduced it was 

 impossible to see how to eliminate impurities. 



I am the more anxious to insist on this work because it took 

 a very long time to execute, and was of a very rigid nature ; 

 and because, so far as I know, no other suggestion has been 

 made with regard to obtaining pure spectra ; and of course, if 

 we wish to study the chemistry of the sun, the first desideratum, 

 as Kirchhoff saw, and as Angstrom saw, and as we all see 

 now, is to have a series of maps absolutely and completely 

 beyond all suspicion. 



