CONTEMPORARY ADVANCES IN PHYSICS, XXIX 2^1 



that they are all discriminated from each other by something physical, 

 and yet they all have something in common which distinguishes them 

 as a whole from states of other sequences. As to the nature of these 

 "somethings," the original theory of Bohr is perfectly explicit. All 

 the states correspond to orbits of the valence electron; the different 

 P states correspond to orbits of different but definite sizes and shapes; 

 what all the P orbits have in common is, a common value of the angular 

 momentum of the electron revolving in its orbit. 



To repeat Bohr's argument here would be an unjustifiable use of 

 time and space. Let me merely recall that when it was applied to the 

 hydrogen atom, it led to several extraordinary agreements with the 

 data of experiment, to which others have since been added. To some 

 degree these are transferable to the sodium atom, especially since the 

 electric field in which the valence-electron of the sodium atom mostly 

 revolves is extremely like that in which the electron of the hydrogen 

 atom always revolves {i.e., beyond the "cage" to which I referred, 

 the ten electrons of the cage effectively cancel the influence of the 

 portion + lOe of the nuclear charge, leaving uncancelled only the 

 field of a nuclear charge + e which is the same as the charge of a 

 hydrogen nucleus). These agreements were the primary cause of the 

 enormous role which angular momentum has ever since been playing 

 in all atom-models. They were responsible also for the numerical 

 values now assigned to angular momenta which figure in atoms; for 

 according to the original theory, the orbital angular momentum of the 

 P states is KJijlir), and all the other values which it may take for 

 states of other sequences are other small-integer-multiples of {h/lir); 

 and while these values have since been somewhat altered without 

 impairing the numerical agreements on which they rested and on 

 which now the new ones rest, it remains true that all angular momenta 

 occurring in atom-models are expressed as multiples of ^(hjlir), the 

 multiplying factors being integers usually smaller than ten. 



As my words have already implied, there are various types of 

 angular momentum nowadays fitted into atom-models, the one already 

 described — hereafter to be called the "orbital angular momentum" — 

 being only one and the first. We turn now back to the principal 

 series of sodium to discover why another type is required. 



Examined with a sufficiently good spectroscope, each "line" of that 

 series is found to be actually a close doublet. These imply that each 

 of what I have been calling the P states is actually a pair of states.^ 

 (The confusion introduced into language by referring to one and the 



^ That it is not the normal state which is resolved into a pair is proved bj- various 

 facts which it is not necessary to mention here. 



