SOME CONTEMPORARY ADVANCES IN PHYSICS— X 129 



introduced into a magnetic field, each of its Stationary States is modi- 

 fied into one or another of several new States, differentiated from one 

 another and from the original State to a small but appreciable extent. 

 This might arise from some distortion or internal alteration of the 

 atom by the field; and it will probably be necessary to adopt this view 

 in some cases. But there is also a simpler effect which the magnetic 

 field may have upon the apparent energ>-\-alues of the Stationary 

 States, an effect not involving any deformation of the atom by the 

 field^ — to wit, an orientation-effect similar to that which was assumed 

 to account for multiplets. This we proceed to examine. 



If an atom which is a magnet is floating in a magnetic field, it 

 experiences a torque which tends to orient it parallel with the field. 

 By saying that an atom is parallel or oblique to the field, I mean 

 that the magnetic moment of the atom and therefore also its angular 

 momentum, are directed parallel or obliquely to the field; and this 

 usage will be maintained. Owing to this torque it is endowed with 

 energy due to the field, in addition to its own intrinsic energy; this 

 additional energy, which depends upon the inclination of the atom to 

 the field, I shall call its extra magnetic energy. If the atom turns in 

 the field, the amount of its extra magnetic energy changes; and if its 

 magnetic moment suddenly changes, its extra magnetic energy also 

 changes unless it simultaneously turns by just the right amount to 

 compensate the change. If at the moment of passing over from one 

 of its Stationary States to another, its inclination or its magnetic 

 moment or both are changed; the amount of magnetic energy which it 

 gains or loses will be added (or subtracted, as the case may be) to the 

 amount of energy which it gains or loses because of the transition. 

 The frequency of the radiation sent out or taken in by the atom will 

 be equal to l/h times the sum of two energy-changes of distinct 

 kinds — not, as in the absence of magnetic field, to l/h times the 

 energy-difference between the two Stationary States alone. Thus 

 the effect of a magnetic field upon spectrum lines might be ascribed, 

 not to any deformation of the atom by the field, but to changes in the 

 orientations or in the magnetic moments of the atoms occurring 

 at the instants when they make their transitions. The question for 

 us now is, whether the actual details of the observed effect can be 

 interpreted in this manner. 



Expressing the foregoing statements in formulae, in which M 

 denotes the magnetic moment of an atom, H the magnetic field, and 

 a the inclination of the atom to the field, we have for the torque which 

 the field exerts upon the atom 



T = MH sin a (8) 



