﻿and the Theory of Atomic Structure. 557 



For the elements Mo-In they are fairly intense lines, but 

 they are not very sharp ; their intensity seems to decrease 

 rather rapidly for elements with higher atomic number than 

 In. As has been pointed out in Part II. § 8, the frequency 

 differences @ n -/3 2 and /3 12 -/3 2 vary linearly with the atomic 

 number. If we assume, as has been done by Dauvillier * 

 and Wentzel f ? that the line /3 8 , found for a great number 

 of the elements Ta-U J is a double line, it is very probable 

 that this line is the same as the lines formerly denoted as 

 {3 n and /3 12 for the elements with lower atomic number. 

 The variation in intensity of these lines, however, is very 

 remarkable. As has been stated above, their intensity rela- 

 tive to that of the other lines is decreasing for the elements 

 with higher atomic number than In ; for the elements in the 

 neighbourhood of W these lines are extremely faint, whereas 

 their intensity seems to increase again for the elements in 

 the neighbourhood of Au. Besides, their frequency difference 

 with the lines /3 2 seems to vary for the heaviest elements 

 more rapidly than with the first power of the atomic number. 

 This made it somewhat difficult at first to recognize the line 

 /3 S of the heaviest elements in the lines /3 U and /3 I2 for the 

 elements Nb-Sb. 



Where two lines were expected nearly to coincide, in 

 most cases a broadening or darkening of the measured line 

 was actually observed. In. this case the wave-length given 

 is only a mean value. In Tables XIX. and XX. this value 

 stands in the column of the line which was considered to be 

 the most intense of the two coinciding lines. For Ho (67) 

 and Yb (70) the lines in question could not be measured, as 

 in this region copper K lines appear. It may be pointed 

 out that in general for the elements of Tables XIX. and XX., 

 with the exception of some copper and tungsten lines, no 

 other lines were found on the short wave-length side of /3 2 

 other than the lines appearing in the tables. 



§ 10. Table XXII. gives two very weak lines, which were 

 only found for some elements where very good plates were 

 obtained. y 7 is a non-diagram line, formerly given for the 

 elements Mo-Ba (Table X., Part II.) ; y 8 is a diagram line, 

 representing the transition 5 -L 2 . As is seen from the 

 table the frequency difference Ly 8 -L/3 7 j's equal t0 L/3!-La 2 

 within the limits of experimental error. The line y 8 was 



* Dauvillier, Comptes Rend us; clxxiii. p. 647 (1921). 

 t Wentzel, Ann. d. Physzk, lxvi. p. 437 (1921). 



X D. Coster, Zeitschr.f. Physih. 1. c. ; Dauvillier, 1. c. (Dauvillier calls 

 this line (B 2 "). 



