In previous papers (Proc. R. S. Edin., 1905, 1907, 1912) on the subject of crystalline magnetic quality, all questions of temperature were left aside. Molecular heat motions were regarded as non-existent, and the magnetic atoms or molecules were considered as representable by small ideal magnets centred upon a geometrical space lattice. No other mode of representation is possible until explicit notions regarding the origin and seat of molecular magnetic quality become definitely possible; and the general nature of magnetic actions and properties indicate that the general nature of the conclusions thus arrived at will not be fundamentally altered by particular assumptions regarding the nature of the magnetic molecules themselves. The crystalline quality of the magnetisation is due to the geometrical spacing of the atoms or molecules, and is manifested under the control afforded by the magnetic fields originating in surrounding molecules. This is essentially, when freed from the restriction to crystals, the statement of the modern Theory of Molecular Magnetism, which was definitely put forward by Weber in 1852, in all its features, and was recognised by Maxwell, who, in the state of experimental knowledge at the time, contented himself with an addition to Weber's representation of the law of force by postulation instead of by direct deduction. The first explicit applications to crystalline, or randomly crystalline, solid materials were made by Ewing, who obtained approximate quantitative results, relatively to ordinary magnetic materials such as iron, by neglecting the effect of all but closely neighbouring molecules, a condition which, though only strictly justifiable under the postulate of small pole clearance, is sufficient to give a good-elucidation of the effects. The action of his models was not so narrowly restricted.
