206 THE QUANTUM [xv. i 
field due to a number of collinear magnetic molecules. An 
account of his work and of the other early investigations is to 
be found in Zeeman’s Magneto-optics (1913). 
The normal Zeeman resolution may be deduced by decom 
posing the motions of the electrons in the atoms into three simple 
motions : a vibration parallel to the lines of force—this component 
will be unaltered by the magnetic field—and two circular motions 
in opposite senses in planes normal to the lines of force. This 
mode of resolution is legitimate if we assume the motion of the 
electron to be harmonic. 
Take a set of right-handed rectangular axes, OX, OY, OZ, 
as in Fig. 27. Let us suppose that there is a magnetic field of 
strength H in the direction of the axis of Z. The motion of the 
radiating electron is to be decomposed into a linear vibration 
Fig. 27.—Zeeman Effect. 
along OZ and two circular motions perpendicular to OZ, i.e. 
parallel to the plane of XY. These circular motions are in 
opposite directions (a and b). This resolution is clearly shown 
in a model constructed by Zeeman and illustrated in his book 
Magneto-optics. The central rod (Fig. 28) is parallel to the mag 
netic field, and we may take the straight arrow as indicating 
the direction of the magnetic field. One vibration is in the 
direction of this arrow. The other, circular, motions are shown 
by the circular arrows on the two discs. The effect discovered 
by Zeeman can be predicted very simply from this model, as 
will be shown later. 
Larmor * has given a very important theorem which may be 
stated as follows : 
Let a system of electrons describe orbits under the action of 
their mutual repulsions and of any other forces the directions 
* Larmor, Aether and Matter, p. 343.