180 BUXLAKD AND MASON [CHAP. 10 



originally for the detection of submarines, are still widely employed, particu- 

 larly in the United States. The first measurements with a towed magnetometer 

 were made by Miller and Ewing (1956) with a fluxgate instrument. 



The fluxgate method will not be described in detail here. It depends on 

 applying an alternating field to a rod of high-permeability alloy. The alter- 

 nating field is sufficient to drive the material to saturation in both direc- 

 tions. If the steady field along the rod is zero, the wave form wiU be symmetrical 

 and free from even harmonics ; otherwise even harmonics are generated in the 

 energizing coil wliich are proportional in amplitude to the value of the steady 

 field acting along the rod. It is, therefore, possible to determine when the rod is 

 set perpendicular to the earth's field. 



The fluxgate magnetometer can be used to measure specific components of 

 the field, but because of the need to operate it in a towed fish, and the con- 

 sequent difficulty of maintaining stable reference directions, it is generally 

 arranged as a self-orienting total-field instrument. This is accomplished by 

 mounting three fluxgate units perpendicular to one another in a motor-driven 

 gimbal. Two of them are made to orient the third along the earth's field by 

 means of servo-loops, which control the orienting motors in such a way as to 

 maintain zero field along their axes. The even harmonic signal from the third 

 unit is then a measure of the total force of the earth's magnetic field. 



The accuracy and sensitivity of the method can be increased by reducing 

 the field acting along the third rod to zero by passing a measured direct current 

 through the energizing coil, using the fluxgate as a null detector. The accuracy 

 of measurement then depends on the stability of the coil and the accuracy with 

 which the current in it can be controlled and measured. There is no difficulty in 

 controlling the current to one part in 10^ for periods of a few hours, representing 

 better than 1 y ; the main source of error lies in the difficulty of stabilizing the 

 current and the dimensions of the coil over a long period. For a few hours a 

 stability of the order of 1 y is possible, but it is doubtful whether, even with 

 frequent comparison with a proton magnetometer, an absolute accuracy much 

 better than 10 y can be expected. In practice, it is usual to adjust the current 

 in discrete steps, corresponding to uniform increments of field, the signal due 

 to the residual field being applied to a recording instrument. 



The basic disadvantage of the fluxgate, compared with the proton precession 

 method, is that it is not absolute, and, because of the instability of the coil 

 system, it requires frequent calibration. It possesses one advantage in that it is 

 more suitable for measuring vector components of the field. 



Instruments developed in the United States have been described by Rumbaugh 

 and Alldredge (1949) and Schonstedt and Irons (1955) and Russian instruments 

 by Logachev (1955), Finger et al. (1961) and Kudi'evsky (1961). 



The proton magnetic resonance method was first used to measure the earth's 

 field by Packard and Varian (1954). An instrument for use on land has been 

 described by Waters and Phillips (1956) and by Waters and Francis (1958), 

 and one used in a rocket by Heppner, Stolarik and Meredith (1958). Hill (1959) 

 and others have used the instrument at sea. 



