NUCLEAR MAGNETIC RESONANCE 77 



Fig. 2 — Peak of nuclear resonance absorption. This is the first peak to be 

 published other than those obtained with molecular beams. It pertains to protons 

 in water. (Courtesy of E. M. Purcell). 



by Hi , is of the order of a small fraction of one gauss up to several gauss. 

 Its frequency must be of the order of tens of megacycles. To be more 

 specific, the effect that is sought with protons is located at 42.6 mc 

 when H is 10,000 gauss. 



Finally there must be circuits and detectors for measuring the ab- 

 sorption of the electromagnetic wave-energy in the sample. These are 

 well kno^vn to those proficient in the art: we pass them over. 



Now of the two quantities H and v either is to be varied while the 

 other is to be kept constant, and the absorption is to be measured. 

 Usually H is varied while v is kept constant, and the data consist of a 

 plot of absorption against H for a set value of v. 



When such a curve is plotted it proves to be, in the main, a smoothly- 

 sloping curve, of no interest in the present connection. What is of 

 interest is that it is interrupted by a magnificent peak of extraordinary 

 sharpness, deserving to be called a needle. Probably there is nothing that 

 can please an experimenter more than a curve with a fine sharp peak: 

 here he has it. Fig. 2 exhibits the first such peak on record. But neither 

 Fig. 2 nor any other picture can convey an adequate idea of the sharp- 

 ness of the peak, for the distance from this imposing feature to the 

 axis of the ordinates at field-strength zero may be, and often is, tens of 

 thousands of times as great as the breadth of the peak. (With the 

 induction-method, peaks have been distinguished from each other that 

 are separate by only a millionth of the value of H at which they are 

 found). This needle has the narrowness that is characteristic of fine 

 lines in optical spectra; and this is as it should be, for a spectrum-line 



