FIELD. H GAUSS 8 cubic) have been developed, and their magnetic properties have been investigated by K. Beck (1918), W. L. Webster (1925), K. Honda and S. Kaya (1926) and W. Gerlach (1927). There is good agreement in the general character of the results, but differences in detail which are prob ably due to marked changes in the magnetic properties which may be caused by the presence of small amounts of impurities and by the mechanical treatment. Some magnetization curves obtained by Gerlach are shown in fig. 45.
There is no strongly marked direction of easy magnetization (as in pyrrhotite crystals) but along the tetragonal axis saturation is reached for smaller fields than along the diagonal axis. G. S. Mahajani has shown that the differences in the magnetic prop erties along different axes may be accounted for in terms of the purely magnetic interaction of the atomic magnets; but this does not account flr the attainment of strong magnetization in weak fields characteristic of ferromagnetics generally. For the iron crystals used by Gerlach the magnetization increases rapidly in very small fields, as will be seen from the curves; the initial permeability being very high. There is very little hysteresis loss,
the upward and downward curves practically coinciding, while the coercivity may be as small as .o5 gauss. The remanence also is small, but owing to the low coercivity, and the difficulty of estimating the demagnetizing factor, it is not possible to deter mine the true remanence with any certainty. The results are not incompatible with the Weiss theory of spontaneous magnetization throughout small domains. The magnetic properties of a single crystal depend on the degree of perfection with which the domains are joined together. It follows that the properties of an ordinary piece of iron are not dependent simply on the grain size (that is, the size of the constituent micro-crystals). The assumption of magnetic domains, throughout which there is spontaneous mag netization, certainly enables many of the facts of ferromagnetism to be simply correlated.
The further investigation of single crystals will do much to elucidate the character of these domains, for there are no com plications such as those which arise with an ordinary piece of iron owing to its being a micro-crystalline aggregate.