﻿and the Theory of Atomic Structure. 571 



found some atoms for which the M-shell has one electron 

 more than usual. The principal discontinuity should be 

 connected with the removal of an electron from the K-shell 

 of a "normal" atom, the anomalous discontinuity with the 

 removal of an electron from the K-shell of an atom whose 

 M-shell has one electron more than usual. 



I, however, suggest the following explanation. We will 

 assume that the atoms of the same element have all the same 

 initial state. The removal of one electron from the K-shell 

 by absorption of X-ray energy may, however, happen in 

 different ways. The electron may be removed wholly to the 

 outside of the atom, a process which should correspond to 

 the principal discontinuity, or the electron may be removed 

 from the K- and enter the M-shell, which for the elements 

 in question is in a state of development*. The latter process 

 should correspond to an anomalous absorption-line lying on 

 the long wave-length side of the normal absorption-edge. 

 Besides, we may assume that, if in the latter case the re- 

 generation of the K-shell takes place by an electron of the 

 M-shell, this should give rise to a line which should have 

 exactly the same wave-length as the corresponding absorp- 

 tion-line. Such a line has actually been found by Hjalmar t 

 (compare Part V. §5). If this explanation is correct, we 

 should have in this line an example of a line which appears 

 at the same time as emission and as absorption-line. A line 

 of this type is only possible in a region of the periodic 

 system where the initial level corresponds to a shell which is 

 in a state of formation. 



Now T we may assume that in the rare-earth group where 

 the N-sbell is in a state of formation we should have the 

 same phenomena. For these elements an electron may be 

 removed from the L-shell under two different conditions : it 

 may be removed wholly to the outside of the atom, or it may 

 be transferred from the L-shell to the N-shell. The first 

 transition should correspond to the normal absorption-edge, 

 the second to an absorption-line lying on the short wave- 

 length side of the edge. Besides, the inverse process of the 

 second transition should give rise to an emission-line lying 

 on the long wave-length side of a diagram-line, and having 

 exactly the same wave-length as the corresponding absorp- 

 tion-line. 



Anomalous satellites lying on the long wave-length side 

 of diagram-lines have actually been found in the emission 



* Compare N. Bohr, Zeitschr.f. Physik, vi. p. 1 (1922). See especially 

 p. 60, where the possibility of such a transition has been suggested, 

 t Hjalmar, Phil. Mag. xli. p. 675 (1921). Hjalmar calls this line K/3'. 



