NUCLEAR MAGNETIC RESONANCE 93 



the stopping of the oscillating field; and what gradually adjusts itself is 

 the distribution of the protons between the up and the down orienta- 

 tions. Now I will describe an experiment such as has been performed 

 on protons in water. 



Let the sample be placed in the big field some time — several hours 

 will always be ample — before the experiment is to begin. The experi- 

 menter should know in advance the frequency of the Larmor precession, 

 so that he can apply the oscillating field of proper frequency as soon as 

 he and the sample are ready. The sample then enters into what I will 

 call "the state of resonance." The experimenter is to measure the height 

 of the peak as soon as the oscillating field is switched on: I call this 

 initial stature Ao . The big field and the alternating field are now both 

 to be kept on. The height of the peak, A, is to be measured from time 

 to time, say once every tenth of a second (this has been done with 

 movie techniques). It is found that A is a declining function of time; 

 the peak is shrinking. 



After a while, let the alternating field be switched off while the big 

 field continues to be on. The state of resonance is now suspended. Again 

 the height of the peak is to be recorded every tenth of a second. Need- 

 less to say, the alternating field must be on while the record is being 

 made, but it shall be off all of the rest of the time, which is most of the 

 time. It will be found that the peak is growing again. It is, in fact, 

 trending back to its initial stature Ao , and the law of its rise is the 

 exponential law: 



A = Ao[l -exp(-t/TO]* (11) 



The constant Ti , which this experiment determines, is called the "spin- 

 lattice relaxation-time." "Lattice" will be recognized as a term ap- 

 propriate to crystals: in the literature of this subject it is however 

 applied to all solids and liquids. Its meaning in this field may be put 

 as follows: the "lattice" is all of the sample except the nuclear spins. 



The actual experiment is not usually done quite as I just described 

 it. The alternating field does not have to be switched on or off, because 

 if its frequency is far from the Larmor frequency it is practically in- 

 effectual. If the observer wants to end the state of resonance, he dis- 

 places H or V away from the resonance-value; if he wants to restore it 

 he brings H or v back to the resonance-value. By modulating the big 

 field with say a 60-cycle frequency, he may pass the system briefly 



* This formula implies that A = when i = 0; the reader can recast it to cover 

 the general case in which 0<A<Aoat^ = 0. 



