Results of an investigation of magnetic flux change in the direction perpendicular to the direction of the applied field indicate that the maximum flux in the transverse direction does not occur at the time when the sample is half reversed. From these data, taken on thin evaporated films of iron‐nickel, it is concluded that a coherent rotation theory which assumes a single magnetization vector to represent the entire sample is not adequate to explain the results. It is suggested that, when a small transverse field is present, the magnetization starts to reverse by coherent rotation but soon loses phase throughout the sample so that the reversal is completed by noncoherent rotation. With no transverse field, the coherence is never established and the reversal is completely by noncoherent rotation. Added evidence for this suggestion is presented in the form of a comparison of thin films with ferrite and Permalloy tape cores. In both the normal and transverse directions, the observations are similar, supporting the view that the reversal of films and cores is by the same general mechanism.