of the impressed magnetic field and that the phase is slightly frequency dependent beyond 10 Hz. The signal from the monitoring fluxgate was fed to a dc power amplifier and the phase was advanced by about 8 regardless of frequency. The current from the power amplifier was fed to three compensation coils connected in series. The configuration of the system is shown in Fig. 1. Results are shown in Fig. 7 for impressed magnetic fields of 10 and 30 Hz both having an amplitude of 10 T. The shielding factor was increased by a factor of 20 at the center of the shell for a 10-Hz impressed magnetic field and 10 for a 30-Hz case with the compensation system on. The same experiment was also done for other frequencies keeping the strength of the impressed magnetic field the same. The following results were observed: 55 times for a 2-Hz impressed magnetic field, 32 times for a 5-Hz impressed magnetic field, and five times for a 50-Hz impressed magnetic field. The reason for this decrease in the shielding factor with increasing frequency of the impressed magnetic field can be attributed to a phase mismatch between the impressed and the compensating magnetic field. The lag in phase in the output (see Fig. 6) of the monitoring magnetometer is most responsible for this. This problem can be improved with a simple phase equalizer for several degrees.