Magnetism

The result was unknown to Faraday when, in 1845, he discovered that magnetic properties were not restricted to the iron group of elements, but that all substances were influenced by a magnetic field, though to a much smaller extent than iron. (Iron, nickel and cobalt, with some other substances, may be placed in the special class of ferromagnetics.) Faraday distinguished between diamagnetics, which were repelled by a magnetic pole and, in gen eral, tended to set themselves at right angles to the lines of mag netic forces; and paramagnetics which, like iron, were attracted and tended to set themselves along the lines of force. He rather favoured a conduction theory of magnetism, according to which the magnetic lines tended to crowd into a substance whose con ducting power was higher than that of its surroundings. Para magnetics offered an easy passage to the lines, while the conduct ing power of diamagnetics was low.

The alternative view that the molecules were polarized was developed by Wilhelm Weber Ampere had made the brilliant suggestion that "molecular currents" might give rise to molecular magnets. Weber showed that if the existence of molecular circuits, without ohmic resistance, was assumed, the in duction of currents in them, when placed in a magnetic field, would account for the characteristic properties of diamagnetics.

In paramagnetics, however, it was necessary to suppose that there were permanent molecular currents, making the molecules per manent magnets. The old two fluid theory suggested no explana tion of the two types of magnetic behaviour, and also it could not account for the tendency of the magnetization of substances like iron to approach a saturation value with increasing field. To explain the fact that the elementary magnets of iron did not set themselves all parallel to each other when the magnetizing field was small, Weber assumed that any displacement of the mole cule was resisted by a restoring couple tending to restore the mole cule to its original position. This, however, was incompatible with the fact that iron may retain a certain amount of magnetization when the field is removed, and with the general phenomenon of the lag of the magnetization behind the field, that is with residual magnetism and hysteresis. Maxwell suggested that there were several possible equilibrium positions. The much less artificial theory that the effects observed are due to the mutual magnetic action of the molecular magnets has been developed with consid erable success by J. A. Ewing (589o), but there are still many baffling problems in connection with the magnetization of ferro magnetics.

The introduction of accurate methods of measurement for ferromagnetic substances was due to H. A. Rowland (1873), and important work in this direction was done by Ewing, the variation of susceptibility (the ratio of the magnetization to the magnetizing force) and hysteresis being investigated in a comprehensive man ner. In an investigation of great importance P. Curie measured the

susceptibilities of a large number of substances over a wide range of temperatures (1895). He found that while diamagnetism was usually independent of temperature, paramagnetic susceptibility over wide ranges—as was tested most carefully for oxygen— was approximately inversely proportional to the absolute tem perature, a result embodied in Curie's law.

The Electron Theory and the Quantum Theory.

The dis covery of the electron, as a discrete unit of negative electricity, and the measurement of its charge and mass, gave reality to the electron theory which had been extensively worked out notably by H. A. Lorentz; and it was on a firm experimental basis that P. Langevin, in 1905, developed a theory of magnetism which ac counted in a satisfactory manner for the general character of Curie's results. The molecular currents of Weber could be re garded as electrons circulating in molecules. Each circulating elec tron would have a magnetic moment associated with it, and a sub stance would be diamagnetic or paramagnetic according as to whether the magnetic moments of the electrons in the molecules compensated each other or not. By applying a statistical treat ment to investigate how the thermal rotational motions would affect the orientation of the molecules in a field, Langevin showed that the susceptibility of a paramagnetic gas should vary in versely as the temperature. From the expressions Langevin derived it became possible to deduce the magnetic moments of the paramagnetic molecules. Shortly afterwards (1907) P. Weiss extended Langevin's theory .by taking into account the mutual action of the molecules by the introduction of a molecular field, proportional to the magnetization acquired. In this way many of the properties of paramagnetics generally and of ferromagnetics were correlated, but the exact nature of the molecular field is very obscure. Chiefly from investigations on ferromagnetics, Weiss was led to the conclusion that atomic moments were all integral multiples of a fundamental unit, the Weiss magneton, a view which has been of considerable value, though it requires profound modification. The electronic orbit theory of atomic magnetism was open to the grave objection that, according to the classical theory, the circulating electrons should radiate, and gradually lose energy. This was part of a much wider problem. The classical theory was entirely unable to account for the upbuilding of stable atoms, or to show how they could give rise to the spectra observed.