138 BELL SYSTEM TECHNICAL JOURNAL 



energy, with which it flies away. This is known positively; for the 

 extracted electrons can be observed, and their energy measured. 



Spontaneous transitions from each of the X-ray Stationary States 

 occur to some, but not to all, of the States of lesser energy. Some 

 are evidently inhibited; and it is possible to lay down rules of selec- 

 tion, distinguishing those which are permitted. The complicated 

 system of rules originally proposfed has yielded place to a much simpler 

 one, exactly similar to the one prevailing in the optical spectra. That 

 is to say: it appears to be possible to assign to each of the X-ray 

 Stationary States a certain value of a numeral k and a certain value 

 of a numeral j, such that the only transitions which actually occur 

 are those in w^hich k changes by one unit and j either changes by one 

 unit or does not change at all; while transitions between states in 

 both of which 7 = are specially excluded. Furthermore, the various 

 values of k and j thus assigned to the several X-ray Stationary States 

 are identical with those assigned to the several States constituting a 

 doublet system, such as we have met already in Section S, such as 

 the sodium atom possesses; so that there is a complete correspondence 

 between the system of X-ray Stationary States which every atom 

 rich in electrons possesses, and the doublet system of optical Station- 

 ary States which only certain atoms possess. A part of this corre- 

 spondence is expressed in the following Table: 



TABLE II 



No doubt the implications of this close correspondence are deep; 

 but just what they are is not yet obvious. 



The fact that the residue, left behind when an electron is extracted 

 from an atom, may exist in any one of several distinct States, is quite 

 naturally interpreted as meaning that the various electrons of the 

 complete atom are variously situated, or revolving in various distinct 

 orbits; so that the several X-ray Stationary States differ essentially 

 in this, that differently-located electrons have been removed, leaving 

 different places untenanted. This notion is easily combined with the 

 idea that an atom is formed, or at all events behaves as though it 

 had been formed, by successive self-annexations of electrons to a 

 nucleus originally bare. Suppose that an atom is made by con- 



