Radiation in a Magnetic Field. 



Ill 



of Fievez, that the effect of a magnetic field is to produce 

 reversals. The appearance of the figure itself seems to 

 indicate a true separation ; the depression bearing a much 

 larger proportion to the whole area than we are accustomed 

 to observe in reversals ; but this by itself would hardly be 

 considered conclusive. It was thought, however, that if 

 there were a true separation, the distance between the 

 components should vary with the strength of the field, 

 whereas in the case of a reversal one would expect only an 

 increase in the darkness of the absorption. 



Accordingly a series of observations was made with varying 

 strength of field, the results of which are shown in the 

 visibility curves of fig. 1, and the corresponding intensity 

 curves of fig. 2. The strength of field in the order of the 

 letters was 0, 5, 7, 11, 16, 20. It appears from the figures 

 that up to a strength of field 11, which is about 2000 C.Gr.S., 

 the principal effect is a doubling of the line ; but beyond this 

 the component lines are broadened as well as separated. It 

 is also clear that the separation is nearly proportional to the 

 strength of field. Thus, assuming this law to be true, the 

 following table shows the agreement between the observed 

 and the calculated distances. F = strength of field ; A = dif- 

 ference of path in millimetres corresponding to visibility 50 

 for single source ; S = corresponding half- width of the source 

 on a scale of 100 for D x — D 2 ; D = period of the coincidences 

 in millimetres due to the doubling, and a = the corresponding 

 distance between the components. 



Table I. 



A 



F. 







5 



7 



11 



16 



20 



A. 



20 

 20 

 20 

 20 

 14 

 10 



5. 



D. 



OS. 



a. calc. 



•66 

 •66 

 •66 

 •66 

 •94 

 1-32 



GO 



58 

 38 

 28 

 18 

 14 



0-0 

 10 

 1-6 

 2-2 

 3-3 

 43 



o-o 



1-0 

 15 

 23 

 3-4 



4-2 



B 



C 



D 



E 



F 





The next substance examined was cadmium. Cadmium 

 filings were enclosed in an end-on vacuum tube which was 

 placed between the poles of the electromagnet. The results 

 are shown in figs. 3 and 4, and prove that there is scarcely 



