Spectrum produced by External Magnetic Forces. 5 7 



and 5, forms intermediate between the unchanged spectral 

 lines and the doublets and triplets may be expected. With 

 the doublet only one particular intermediate form exists, 

 viz., a line the edges of which are circularly polarized, the 

 central part emitting unpolarized light. This case I described 

 in my former paper. 



Of triplets two intermediate forms may be distinguished. 



I will call a line " Triplet a," if the magnetic change corre- 

 sponds to somewhat more than half the width of the unchanged 

 line. In this case the broadened line is composed in the 

 followiug manner : the central part will emit horizontal 

 vibrations, at both its sides bands of (chiefly, but not entirely) 

 unpolarized light border it, which again are enclosed by bands 

 of vertically vibrating light. The accompanying diagram 

 gives a rough scheme of the constitution of the line. 



Polarized light. 



,Ct Vibrations vertioal. 



, - ' i 



' ,-'"\~o Vibrations horizontal. 





= 



m 



1 v 

 1 x 



1 



1 

 1 



- , 



- + 

 l 



— 91 



Unpolarized light. 



Original width. 



the magnetic change amounts to less than 

 the original line. This case I described 



In " Triplet b 

 half the width of 

 in my cited paper. 



7. Hence we may expect that if Ave can just resolve the 

 doublet, the magnetic force also is sufficient for triplet a. 



As with sodium (cf. § 11) I did not succeed completely with 

 the means at my disposal in observing the expected doublets and 

 triplets, so with other substances I looked in the first place for 

 magnetic doublets. We might expect then that further inquiry 

 would show something about the triplets. At last I succeeded 

 in observing with the blue line (A, =4 80 //.//) of cadmium the 

 doublet and the triplet a. It has already been remarked b}' 



