CONTEMPORARY ADVANCES IN PHYSICS, XXIX 289 



as though it were moving through a magnetic field, and this is not so 

 easy to grasp, since we have postulated merely that it is revolving 

 in the electrostatic field surrounding the nucleus and the cage of inner 

 electrons. Yet if we were to shift our frame of reference and imagine, 

 not the nucleus standing still and the valence-electron revolving around 

 it, but the electron standing still and the nucleus revolving around 

 it— then we should have no difficulty in realizing that the revolving 

 nucleus, being a charge describing a closed path, and being therefore 

 equivalent to a current, would produce a magnetic field. Reverting 

 now to our original frame of reference, we may bring the magnetic 

 field back with us, and say that the spinning electron revolves both in 

 the electrostatic field aforesaid and in the magnetic field, and its 

 energy is influenced by both. This is not a very sophisticated way of 

 looking at the matter, and there is a tricky little relativistic detail in 

 the shifting of the frame of reference, which produces an error of a 

 factor 2 if disregarded ; but it serves to bring out the idea. The energy 

 of the valence-electron in its orbit is affected by this quasi-magnetic 

 interaction due ultimately to the fact that it is a magnet moving 

 through an electrostatic field ; and the value of the energy depends on the 

 orientation — on the angle d between the axis of the magnet-electron 

 and the normal to the orbital plane. One now sees readily that there 

 will be a minimum energy occurring when these two directions are 

 parallel, and a maximum energy occurring when they are anti-parallel. 

 But why then do we not find the individual P-state spread out into 

 a continuous band of states corresponding to all the energy-values 

 between these two extremes, and all the infinity of difi^erent orienta- 

 tions between the value 0° and the value 180° of the angle 61 



This is no question which classical physics can solve. The fact 

 that the individual P-state, or what would otherwise be the individual 

 P-state, is split into two instead of into an infinity — this fact implies 

 that only two orientations occur in nature, are "permitted," as the 

 phrase is; and this instance of quantization of direction, like all the other 

 instances of quantization, is a consequence of the quantum-mechanical 

 constitution of the world. More lucid instances occur when the atom 

 with its electrons is immersed in an applied magnetic field of known 

 intensity; that is, when sodium vapor is exposed to the measurable 

 field of a large-sized magnet, and its spectrum is observed. We will 

 take up some of these instances before beginning with the assignment 

 of spin to the nucleus. 



When a magnetic field of moderate strength is applied to sodium 

 vapor, each of the doublets of the principal series is split up into a 

 pattern of several lines or "components." I can no longer say that 



