Abstract
Néel has attributed magnetic hysteresis to random perturbations of the properties of the magnetic material. There has been some question as to whether these properties should be regarded as varying more or less smoothly through the material or whether they should be thought of as highly localized ``pinning centers'' or ``defects.'' Most experimental results can be interpreted equally well using either point of view. However, for one experiment, the predictions of the two theories are distinctly different. This is the dependence of wall velocity on magnetic field when the wall is subjected to a step function of field. In this paper it is shown that the prediction of theory is at variance with experiment if the perturbation is a smoothly varying function of position whereas it is in agreement if the perturbations are localized. In this latter case, wall pinning and irreversible snapping free can occur. Account is taken of wall flexibility, including magnetostatic effects, in the calculations.