CONTEMPORARY ADVANCES IN PHYSICS 227 



victory. Our theory of spectra reciuires that atoms, or some of them 

 at any rate, should be magnets, and they are magnets. Moreover it 

 fixes the magnetic moments which certain atoms ought to have, and so 

 far as our experiments go, the atoms do have these moments. More- 

 over it imposes angular momentum on these atoms, and fantastic as 

 the consequences are, experience bears them out. Logically, then, I 

 should begin the main part of my talk by showing how the magnetic 

 moments and the angular momenta of atoms and of molecules are cal- 

 culated from their spectra by atomic theory. This, however, would by 

 itself require several lectures, and very difficult ones at that.* I must 

 therefore simply ask you to believe that the magnetic moments of 

 atoms are inferences from fundamental theory, not mere ad hoc as- 

 sumptions; and now I will explain what I had in mind when I wrote 

 down the word orientation to designate one of the topics of this article. 

 It is one of the best-known facts of physics that the magnetization 

 of a substance is not fixed and constant, but increases with the strength 

 of the magnetic field which is acting on the substance. By the way, be- 

 fore going any further I must definitely exclude the so-called "diamag- 

 netic" substances. That exclusion being made, we do not assume that 

 the magnetic moment of the individual atom increases similarly with 

 the field strength. People did not make that assumption, even in the 

 days before the fundamental theory was developed. Had they done 

 so, it would have been just as silly as saying that a marshmallow is 

 made of soft white sticky atoms, and calling that an atomic theory. 

 They supposed, and we suppose, that the moments of the individual 

 atoms remain practically the same whatever the field strength; what 

 changes is the average inclination of these moments to the field. The 

 atomic moments are vector quantities pointing in various directions, 

 different from one atom to the next. The magnetization of the sub- 

 stance as a whole is the resultant of all these myriads of tiny vectors 

 pointing in their various directions. If they all pointed the same way 

 the substance would be completely and perfectly magnetized, with a 

 moment equal to the total number of the atoms multiplied into the 

 moment of any one. This state of saturation is not, however, to be 

 attained, not even to be approached without a rare and felicitous con- 

 course of a favorable substance, a very low temperature and a very 

 strong field. Much easier of attainment is the opposite extreme, when 

 the vectors are pointing all ways at random and the magnetization is 

 zero. This happens with nearly all substances when there is no field 

 applied, and it seems quite natural. But when even the smallest field 



* This subject was partially treated in "Contemporary Advances in Ph\sics, 

 XXIX . . . ," April 1935 Bell Sys. Tech. Jour. 



