Abstract :
The first portion of the paper deals with the properties required in a permanent magnet. It is shown that in all cases, provided the magnet is properly proportioned to its work, the true criterion is the (BH)max. value. The relationship existing between the values of saturation density, remanent density, coercive force and (BH)max. is then considered, and a modified form of Kennelly´s or Lamont´s law is developed, which enables this relationship to be expressed both analytically and graphically. It is shown that in actual practice magnets of homogeneous composition obey this law very closely indeed. The paper next deals with the case of magnets which are microscopically not of uniform composition, and which contain two principal constituents. The relation between the magnetic properties of the magnet and the grouping of the constituents is considered, in addition to its relation to their magnetic properties. The constituents may fall under three headings :(a) non-magnetic, (b) magnetic and possessed of high coercive force (magnetically hard), and (c) magnetically soft, i.e. magnetic but of low coercive force. Evidence shows that the highest values of (BH)max. are obtained when the magnet, consists uniformly of a material of class (b), and that with the right steels and appropriate heat treatment other constituents are absent. The constituent in question is, however, only a transition product between the other two classes (a) and (c), and its presence in the pure state is therefore not always easy of attainment. Data taken chiefly from modern cobalt steels are given in support of the theories put forward, and some suggestions are made for further research and development work in this connection.
