Abstract
Whether the ratios Hz/Hy(t), Hz/Hy(a) are independent of the electromagnetic inducing period is a crucial property for studying two-dimensional induction anomalies. We have recalculated the two ratios over a range of period for a half-cylinder anomaly model (suggested by Summers) using both the integral equation method and the finite difference method. We have obtained results which are significantly different from Summers'. In particular, we find that Hz/Hy(t) is frequency dependent, implying that if we observe that Hz/Hy(t) is frequency independent, we may not use the local induction model to interpret the result. Besides, a reexamination of the situation is presented.
