SECT. 1] THE MAGNETIC FIELD OVER THE OCEANS 181 



The method depends on the interaction of the magnetic moment of protons 

 with a magnetic field. A magnetic field of about 100 gauss is applied to a sample 

 of water. This causes about 10~'^ of the protons in the water to become oriented 

 along the field. If the applied field is removed the protons precess and produce 

 an alternating e.m.f. in a coil surrounding the sample. 



The phenomenon can be described from the point of view of either classical 

 or quantum mechanics (Andrew, 1956; Abragam, 1961). Both give the fre- 

 quency, /, of the proton precession as 



where y is the gyromagnetic ratio of the proton. This has been determined by 

 the measurement of the frequency in a known field. The result obtained by 

 Driscoll and Bender (1958, 1958a) and adopted by the International Union of 

 Geodesy and Geophysics (Nelson, 1960) is 



y = 26.751.3 ± 0.2 gauss-i sec-i. 



For another determination see Vigoureux (1962). The value of y required is the 

 observed value and not that corrected for screening by the extra-nuclear 

 electrons, which is what is usually given in tables of physical constants. 



It is necessary to leave the polarizing field on for a few seconds to allow the 

 orientation of the protons to reach thermal equilibrium. On turning off the 

 field, the protons will precess about the earth's field and will produce an 

 external alternating field for a time depending partly on the time taken to 

 reach a new thermal equilibrium and partly on a tendency for the protons to 

 exchange energy and get out of step. In practice, the signal decays to 1/e in a 

 few seconds. If the field is not constant over the sample, the rate of precession 

 of the protons will differ in different parts of it, and the signal will decay owing 

 to their getting out of step. An inhomogeneity of 1 in 10,000 would cause the 

 protons to get out of step in 5 sec and would have an appreciable effect on 

 the decay time ; it is, therefore, desirable that the field should be uniform over 

 the sample to within a few y. This condition is easily satisfied in a towed instru- 

 ment, but usually prevents the instrument working with the sample aboard 

 ship or in a building. 



The proton magnetometer requires no calibration. The relation (1) between 

 frequency and field depends only on the gyromagnetic ratio of the proton and 

 is independent of all instrumental and environmental factors. Also, the fre- 

 quency is independent of the orientation of the coil, and no stabilization is 

 necessary. The signal is a maximum if the applied field and the pick-up coil are 

 at right angles to the Earth's field, and decreases as sin^ d if this angle is reduced. 

 This is not a serious limitation in practice, since, over most of the earth, the 

 field makes a large angle with the axis of a horizontal coil and in the equatorial 

 regions a vertical coil can be used. 



Abragam et al. (1957) have described a method by which the proton resonance 

 may be made to control the frequency of a continuously running oscillator, but 

 this method has not been used at sea. 



