NUCLEAR MAGNETIC RESONANCE 



87 



Fig. 1 in one important way: it is non-uniform, increasing in strength 

 from (say) the bottom to the top. In one respect the protons behave 

 just as they do in a sample in a uniform field: roughly half of them are 

 pointing up and the other half pointing down. But in the non-uniform 

 field the ''up" protons experience a net force pushing them upward 

 and the "down" protons a net force pushing them down. (Visualizing 

 each of the protons as a tiny bar-magnet, one sees that the fieldstrength 

 is bigger where the upper pole of the magnet is than where the lower 

 pole is). The beam is parted into two diverging pencils, the one con- 

 taining the "up" protons only and the other the "down" protons only; 

 I call the first the "up" pencil and disregard the second. 



The "up" pencil now passes through a region just like that implied 

 in Fig. 1 : a magnetic field which is big and vertical and uniform — H will 

 stand for its strength — and an oscillating field with the magnetic vector 

 parallel to the x-direction. If in this second region some of the protons are 

 turned by the oscillating field into the "down" orientations, that will make 

 no difference to their course across the remainder of the second region where 

 H is uniform. But beyond the second region lies a third where again there 

 is a big field that is non-imiform. In this third region the "up" protons go 

 one way and the "down" protons go another. The detector lies athwart the 

 first way; the "down" protons will miss it. 



The detector-reading is plotted against H for a set value of v. One 

 might think that two curves would be plotted, one with the alternating 

 field off and the other with it on, and that the latter would be sys- 

 tematically lower than the former. But the latter will be lower than the 



^95 



z 



HI 



cc 



LLI 



z 



^ 80 



to 



z 

 f^ 75 



2 70 

 < 



65 



60 



1^ • . 

 • i 



• 



-• ■■ 



• 



•._« 



• 



112 113 114 115 116 117 118 119 120 

 MAGNET CURRENT IN AMPERES 



Fig. 6 — Negative peak or valley of nuclear resonance absorption obtained by 

 the molecular-beam method. It pertains to lithium nuclei in lithium chloride 

 molecules. This was the first experimental evidence of nuclear magnetic resonance. 

 (I. I. Rabi, J. R. Zacharias, S. Millman and P. Kusch). 



