﻿556 



Mr. D. Coster on the Spectra of X-rays 





weaker than /3 9 . They seem to correspond to transitions 

 which are inconsistent with the rules of selection given in 

 Part I. § 3. /3 9 should correspond to the transition Mj-L^, 

 /3 10 to the transition M 2 -L 3 . From this we must expect that 

 the frequency difference of these lines is equal to the fre- 

 quency difference Lc^-L^- Both differences are given in 

 the last columns of Table XX. They agree within the limits 

 of experimental error. Further, we must expect the follow- 

 ing relation to hold for the frequencies : 



L/3 9 = L^i + Ls — L 1? 



where L : and L 3 are the first and third absorption discon- 

 tinuities in the L-series. In Table XXI. the frequencies 



Table XXI. 

 Frequency of /3 9 . 





Calculated. 



Measured. 



47 Asc 



251-95 251-67 

 279-57 279-55 

 29312 29319 



48 In 



50 Sn 



51 Sb 



307-39 307*26 



55 Cs 



368-36 368-58 

 384-82 384-31 

 417-4 417-25 

 435-3 1 43480 

 453 3 452-99 

 490-5 490-44 

 757-52 758-06 



56 Ba 



58 Ce 



59 Pr 



60 Nd 



62 Sm 



74 W 









calculated in this way are compared with the experimental 

 values. The absorption edges of Ag have been taken from 

 Part I. § 9, Table XII., the edges for Sb, Cs, and Ba have 

 been taken from measurements by Mr. Lindsay, those for W 

 from Duane and Patterson. For In and Sn the frequency 

 difference 7 4 -/3 7 has been taken, which for these elements 

 must be nearly the same as L 3 -Li ; for Cd, Pr, Nd, and Sm, 

 Lj has been taken from the author's measurements (see 

 Table XVIII.) ; whereas the value for L 3 for these elements 

 has been calculated from fy 4 (it has been assumed that for 

 these elements the frequency difference L 3 ~y 4 is 2 units). 

 As is seen from the table the agreement between the measured 

 and calculated values is very good. 



§ 9. The lines /3 n and /3 12 have recently been measured by 

 the author for the elements Xb-Sb (see Part II. Table X.) . 



