58 Dr. P. Zaeman on Doublets and Triplets in the 



Egoroff and Georgiewsky * that cadmium also exhibits the 

 broadening of the spectral lines under the action of magnetism. 



8. The method of experiment was principally the same as 

 that formerly used. As in my former experiments, I used 

 a Puhmkorff electromagnet. I missed, however, the beautiful 

 Rowland grating I used in the laboratory of Prof. Onnes. I 

 now only bad at my disposal a smaller one with a radius of 

 6 ft., but like the Le} r den one with 14438 lines to the inch. 

 The second spectrum was very satisfactory. For the cadmium 

 spectrum a spark was used between cadmium electrodes ; with 

 a lens, as often used in this manner by Lockyer, an image of 

 spark and electrodes was formed on the slit ; it is then easily 

 verified that the part of the spark just between the poles is 

 analysed. 



The above-mentioned cadmium line is especially sharp 

 at the violet side. 



Now I succeeded, indeed, in observing the expected pheno- 

 mena (doublet and triplet a). 



l>. For observing the doublet along the lines of force a 

 perforated pole was used. With a current of 30 amp., the 

 distance of the poles being as small as possible without 

 interfering with the cadmium electrodes, a perfectly defined 

 doublet was seen. One of the components appeared to be 

 left-handed, the other right-handed circularly polarized f over 

 the xoliole ividth. It might still be argued that the intervening 

 dark space between the components is caused by a reversal 

 and that it is not a doubling of the lines. However, then it 

 must be supposed that this absorption-line just intercepts that 

 part of a magnetically broadened line emitting unpolarized 

 light. Now with a smaller value of the magnetic force 

 the dark space becomes narrower. I grant that this is also 

 the case with the supposed absorption-line. However, then 

 a greater part of the broadened emission-line should emit 

 unpolarized light, the smaller absorption-line now inter- 

 cepting but a part of it. It appeared, however, that even 

 with very narrow intervening dark space, the components 

 of the doublet were entirely circularly polarized, and hence 

 the intervening dark space seems to be different from an 

 absorption-line. There is yet another argument against the 



* C. Jt. 1897, t. cxxiv. pp. 748 & 242. 



t The sign of the circular polarization is the same in the case of this 

 cadmium line and of the D-lines. I must, however, correct my state- 

 ment in § 24 of my former paper. I now see that if the lines of force are 

 running towards the grating, the right-handed circularly polarized rays 

 appear to have the greater period. Hence the radiation is due chiefly to 

 the motion of a neffativc4i/-ch.a.rged particle. Probably my mistake aro?e 

 from a faulty indication of the axis of the X/4-plate used. 



