When a steady field is applied to a ferromagnetic, and also a strong alternating field, it is found that, if the strength of the alternating field is gradually reduced to zero, the final intensity of magnetization depends only on the strength of the steady field, being independent of the previous history of the specimen. In this way an "ideal" magnetization curve, with no hysteresis is obtained. (This is shown by the light curve of fig. 4o.) The interesting point emerges that the ideal curve apparently rises from the origin practically at right angles to the H axis, a high intensity of mag netization corresponding to an indefinitely small applied field. This is of considerable theoretical significance for it lends strong support to Weiss's theory of the spontaneous magnetization of ferromagnetics.
Magnetization is partly a reversible and partly an irreversible process, the irreversible process corresponding, on the Weiss the ory, to a sudden change in the direction of the spontaneous mag netization through a small domain. The irreversible changes are most prominent on the steep part of the magnetization curve (BC, fig. 4o). If the field is increased slowly and continuously, the intensity may increase discontinuously, the magnitude of the discontinuities depending on the size of the domains through which the sudden changes occur. Such discontinuities in the mag netization were first demonstrated by H. Barkhausen (1919) by
means of an amplifier the current pulses through a search coil, due to discontinuous changes in the induction through a specimen, were rendered audible. More recently the discontinuous changes have been more directly investigated using a quick acting mag netometer. R. Forrer has found that, for nickel wire which has been subjected to special thermal and mechanical treatment, the magnitude of the jumps may be a considerable fraction of the maximum magnetization. Estimates can be made of the size and shape of the domains through which the discontinuous changes occur. These cannot be identified with the actual crystalline grains of the material—they may be much smaller. Further in vestigations promise to throw much light on the real significance of the "domains" of the Weiss theory.
Most ferromagnetic materials acquire the "final" intensity of magnetization corresponding to the applied field with great rapid ity. The permeability of a specimen of steel, for example, was found to be constant in alternating fields up to frequencies exceed ing a million. For very soft iron, however, if a field is suddenly applied, the magnetization acquires a certain value, which, with the field remaining steady, increases gradually and appreciably during a time which may be as great as several minutes. The 20 20 40 60