CONTEMPORARY ADVANCES IN PHYSICS, XXIX 309 



in this case — (7 — ^), also corresponds to two different values of Mz 

 which are equal and opposite, and each of which at zero field-strength 

 (but not for H — 0) coincides with one of the previous two. Now 

 consider any two of these values of M^ which correspond to equal and 

 opposite values of m and coincide with one another at // = 0. As // is 

 increased from zero, these two are altered in opposite senses, and one 

 of them actually passes through zero and then reverses at its sign at a 

 certain value of field-strength (m and hence a are negative, and Mz 

 vanishes when x — a) while the other is shifted in the opposite sense 

 and never vanishes (m and a are positive).^ We shall presently see 

 (page 310) that this behavior is the basis of one of the methods of 

 evaluating /. One more peculiarity of these equations must be 

 stressed: the magnetic moment of the nucleus nowhere appears in 

 them! This becomes evident when the field-strength is put equal to 

 zero and x vanishes, for then the several values of the right-hand 

 member of (16) become simply the projections, upon the field-direction, 

 of the magnetic moment of the extra-nuclear electron-system. Thus 

 we have the paradox that in these experiments the magnetic field gives 

 us information about the nucleus by virtue of the force which it exerts 

 upon the atom, and yet this force is exerted practically upon the 

 electrons alone, and not to any perceptible extent upon the nucleus. 



The laws expressed in equation (16) have thus far assisted in 

 three ways in the study of the nucleus: 



First, in respect to the experiment which I was describing (page 306) 

 when I began on this detour: as H is decreased from what I called the 

 "customary" magnitude, the (2/ -f 1) levels constituting each of the 

 there-mentioned groups draw gradually apart — i.e. they differ more 

 and more in respect of the value of the component of the magnetic 

 moment along the field-direction, which is what controls the deflec- 

 tion. The experiment must therefore be performed with field- 

 strengths H which are sufficiently low, much lower than those cus- 

 tomarily employed in the Gerlach-Stern experiment or in spectroscopy; 

 and this is one of the distinctive features of the technique of Rabi and 

 his school. Narrowness of the beam is all the more required, since 

 dll/dz must be large enough to produce considerable deflection, and 

 if both its value and the breadth of the beam in the s-direction were 

 large, H could not be small in every part of the beam. The beam 

 must also be made nearly homogeneous in speed, and this is done by a 



' Exception must be made for pairs of values of M^, both members of which corre- 

 spond to m — and vanish at H = 0; each member of such a pair departs farther 

 and farther from zero, to equal extents in opposite senses, as H is increased from 

 zero. 



