96 THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1953 



is 660 times as great as that of a proton. Now, it is indeed true that all 

 atoms and molecules contain electrons; and one may properly wonder 

 why they do not always dominate the relaxation. The reason is that 

 in most atoms and most molecules the electrons are paired ''anti- 

 parallel" so that their magnetic moments neutralize each other. (We 

 are to see in Part II that this confines the electronic type of resonance 

 to certain very specialized types of substances). One can however in- 

 troduce into a substance, water for example, atoms or ions for which 

 the neutralization is incomplete. These have much bigger magnetic 

 moments and magnetic fields than any nucleus, and they speed up the 

 spin-lattice relaxation. There is one special case which I treat in more 

 detail, because of its relevance in this connection and its importance in 

 solid-state physics. 



There are crystals, of fluorite for instance, which occur colorless in 

 Nature. These may be colored by exposing them to X-rays, or in other 

 ways which we pass over here; and colored examples may also be found 

 in Nature. Solid-state physicists have long been acquainted with these 

 colorations, which they ascribe to what they call "F-centers.** Various 

 lines of reasoning have converged on the conclusion that an F-center is 

 a cavity in the lattice (now I am using "lattice" in the normal sense, 

 that of the crystallographers) in which a free electron is batting around 

 like a wild animal in a cage. If this is so, then coloration of a colorless 

 crystal by X-rays or otherwise should reduce its relaxation-time, and 

 naturally-colored crystals should have lesser values of Ti than those 

 that are colorless. Experiment has ratified these inferences, and thus 

 nuclear magnetic resonance has come to confirm the theory of the 

 F-centres. So also has electronic resonance, since the F-centres display 

 it with extreme clarity; but this is a topic for Part II. 



The cause of relaxation has now been identified as the thermal agita- 

 tion of the substance, working through the variation-in-time of the 

 magnetic fields which act on every nucleus, weakly from its neighbor 

 nuclei and strongly from any uncompensated electron that happens 

 to be in the vicinity. In gases and liquids the nuclei cruise around, and 

 so do the "magnetic impurities" if there are any; fieldstrengths change 

 swiftly and relaxation tends to be rapid. In solids the atoms and their 

 nuclei vibrate around fixed positions, and thermal agitation has come 

 to be interpreted in the following way. 



Nowadays one thinks of the solid sample as quivering with compres- 

 sional waves, and perhaps torsional waves as well. These constitute the 

 thermal agitation of the sample, and their various frequencies form its 

 elastic spectrum. From this broad band of frequencies we isolate, in 



