( 442 ) 



SATELLITES. 



I shall just add a few words on the spectrum of Aluminium. 

 None of the foi-niulae given as yet represents the first associated series of 

 this element at all satisfactorily ; nor is a satisfactory result attained 

 with my formula. In the beginning of this paper I have pointed out, 

 that very deviating values for N^ were calculated from three of the 

 1^^ Unes of the series. 



The formula runs : 



109675 



w = 48287,9 — 



/ 1,038060\' 

 ( m + 0,89436 + j 



