NUCLEAR MAGNETIC RESONANCE 95 



direction to the down direction, the alternating field also helps protons 

 to turn from the do\\Ti to the up direction. Processes of the first kind 

 involve absorption, as we already know; processes of the second kind 

 involve release of energy. What the detector receives, and what the 

 peak makes manifest, is the net of the absorptions over the releases. 

 (This effect of the alternating field in helping protons from the down 

 to the up direction is called ''stimulated emission"). 



Now we must scrutinize equation (13) more closely. It is evident 

 that T stands for an absolute temperature: the question is, what is it 

 the temperature of? 



One supposes perhaps that T is the temperature of the sample — that 

 is to say, the temperature which would be shown by a thermometer 

 stuck into the sample or possibly into a surrounding bath. And this is 

 indeed what is supposed w^hen the peak has the stature Aq , signifying 

 that the sample has stood long enough in the big field undisturbed by 

 resonance or anything else. When A is Aq and T is the temperature of 

 the sample, (13) is right. But when A is less than Aq because the peak 

 is falling, has fallen or is rising, we must choose between saying that 

 (13) is not right, and saying that (13) defines a temperature which is 

 to be called the temperature of the spinning nuclei, or the "spin tempera- 

 ture" for short. 



The second choice is made; and this is the most vivid language in 

 which to describe the situation. In this language we say that the reso- 

 nance elevates the spin-temperature, or heats up the spins; and that 

 after resonance ceases, the spins cool down to the temperature of the 

 lattice. Thus the study of relaxation becomes the study of the heating 

 and the cooling of the spins with respect to the lattice — ''lattice" being 

 defined, I recall, as everything in the sample except the spins. 



Recorded values of Ti range from times of the order of hours down 

 to times of the order of ten-thousandths of a second. The highest are 

 exhibited by protons in ice at extremely low temperatures; protons in 

 water have Ti = 2.33 seconds; the lowest values are found in the presence 

 of "magnetic impurities." The typical dependence of Ti on temperature is 

 represented by a curve ^vith a single minimum. 



The importance of "magnetic impurities" derives from the agent of 

 relaxation. Relaxation is operated normally by the varying magnetic 

 fields whereby the nuclei act on one another; these vary, as I shall 

 presently say more fully, because the nuclei are wiggling in thermal 

 agitation. But the magnetic fields of nuclei are comparatively small, 

 and therefore normal relaxation is comparatively slow. Much bigger is 

 the magnetic field of an electron, for the magnetic moment of an electron 



