1. It is well known that the molar theory of magnetism, in which the magnetic medium is regarded as being continuous in its ultimate structure and properties, is of very limited applicability to the elucidation of magnetic phenomena. The fact of the impossibility of isolating one kind of magnetism from the other by any process based on inductive action makes evident the essentially molecular nature of magnetism. This was recognised even in the early development of the theory by Poisson, who presumed that the molecules only became magnetic under the action of an external field. Weber recognised the necessity for the ascription of permanence to the magnetic quality of the molecule in order to account for the phenomena of magnetic saturation; and his theory was modified by Maxwell so that it might also include in its range the phenomena of residual magnetisation. In Maxwell's discussion of Weber's theory, we have the first explicit statement of the distinctive feature of the modern theory of molecular magnetism.