The transverse bias susceptibility and the angular dispersion were measured for the purpose of testing the validity of the dispersion theory in polycrystalline ferromagnetic films. If each measurement is made on a localized small region of a sample, the field dependence of the transverse bias susceptibility

agrees very well with the theoretical prediction, where

is the bias field normalized by the anisotropy field. Thus it is possible to determine the random local anisotropy

or the structure constant

of the region giving rise to the magnetization ripple. The dependence of the dispersion anisotropy product
α90K
uon the structure constant

shows that the dispersion theory of Hoffmann was quite reliable. If similar experiments were performed for the whole area of each film, considerable discrepancy is found between theory and experiment in many films. This discrepancy can be interpreted as resulting from the long wavelength dispersion or the extrinsic dispersion.