340 Prof. Nagaoka and Mr. T. Takamine on Anomalous 



satellite, SX varies nearly as \/H, while for the —branch 



SX varies as H 2 . The last branch is easily recognizable in 



PI. V. fio'. 4. taken in heterogeneous fields. The line becomes 



... 

 gradually faint in high fields, and vanishes just like the tail 



of a comet. Gmelin's observation in a satellite of the 



yellow line 5790 is therefore not at all singular ; we shall 



perhaps meet with the same instance in different satellites. 



The ultimate position of the satellites in field of 30 kilogauss 



is as follows : — 



— 



branch of 



4-117 





4- 47 m 



.A.U. 



— 



>i !) 



4- 59 





+ 40 (?) 



>> 



4- 



4- 



}1 J) 

 J5 )J 



f- 46(?) 

 1 - 60 

 - 114 



}- 



-40 



- 57 



» 



The ^-components cannot be distinctly photographed, so 

 that the + and — branches of satellites thus far investigated 

 are not markedly different as in the ^-components. With 

 4047, the rate of separation becomes almost the same as that 

 of the principal line from field of a few thousand gauss. We 

 were not able to identify any except a single branch of —60 

 from the beginning, and it is difficult to say to which satel- 

 lite the different branches given in the figure belong, on 

 account of the intricate mixture of lines in weak fields. 

 Judging from the asymptotic course given in fig. 1, it 

 seems that 



4- branch of 4- 59 lies 4- 68 from P +1 , 



- „ „ 4-59 „ 4-63 „ P_i, 

 4- „ „ - 60 „ - 60 „ P v 



- „ „ - 60 „ - 58 „ P_ r 



In addition to these, there are the lines 



4- 38<?)„ P 

 4- 44 „ P_ 1 



- 76(?)„ p_; 



The existence of some of these lines is described by 

 Wendt * ; the mean position of neighbouring lines observed 

 by us coincides fairly well with his observation. 



The numerous satellites of the green line of mercury are 

 also characterized by the anomalous separation, but in high 

 fields the tendency is the same as for the violet line. Some 

 of the faint lines were difficult to observe, and the mode of 

 separation must be postponed for future discussion. 



* Weiidt, loc, cit. 



