310 Mr. H. Kamage. Spectra of Potassium, Rubidium, [May 14, 



The second Subordinate Series. 



In this series 



n = 22850-21-812 W + — , 



CO — f) ' 



where B is for potassium, 57*8; for rubidium, 238*0 ; and for caesium, 

 553*6 ; and 



/* = -7990 + 7984 W 2 x 10- 9 . 



Tt will be observed that the doublets in the second subordinate 

 series are more widely separated than those in the first series. It 

 would appear also that the two series do not converge towards the 

 same limit ; the difference between the limits, however, diminishes in 

 the different metals as the atomic mass increases. This is true on 

 the supposition that fx is constant, and not variable as in the formula 

 given for the principal series. Kayser and Eunge hold the view that 

 there are different limits for the two series, while both Eydberg and 

 Kummel favour the view that the limits are the same. 



The observed and calculated oscillation frequencies are as follows : — 



Potassium 



m. 



Oscillation 



irequencies. 



Differences. 



Observed. 



Calculated. 



3 



14407 



14417 4 



+ 10-4 



4 



17230 -8 



17229 -9 



- 0-9 



5 



18720 -9 



18720 -3 



- 0*6 



6 



19602 "7 



19603 -9 



+ 1-2 



7 



20167 



20170 '5 



+ 35 



8 



20562 



20555 -3 



- 6-7 



9 



20823 



20828 -6 



+ 5-6 



m. 



Oscillation 



frequencies. 



Differences. 



Observed. 



Calculated. 



3 



14462 



14475 -2 



+ 13-2 



4 



17288*1 



17287 -7 



- 0-4 



5 



18778 -9 



18778 -1 



- 0-8 



6 



19659 -4 



19661 -7 



+ 2 -3 | 



