Magnetic Susceptibility

S. R. Williams. Magnetic susceptibilities, Science, 1921, pp. 54, 339,

The thoughts and conclusions in the reviewed survey article are, in their most general outlines, as follows. Despite the very numerous contradictions between Langevin’s and Weiss’s theories of magnetism and experiment, the basis of these theories—the hypothesis of elementary magnets that change their orientation under magnetization—should be regarded as firmly established, chiefly by the experiments of Barnett1, Einstein–de Haas2, and Stewart3, and, we may add, Beck4 and Arvidson5. The absence of any change in the X-ray diffraction patterns of ferromagnets upon magnetization, observed by Compton and Rognley6, if confirmed, should lead either to the rejection of the theory of magnetic atoms changing their orientation, or else to the transfer of the elementary magnets into the interior of the atom. The latter, for example, would contradict the paramagnetism of Fe and Ni salts and the diamagnetism of \(K_3FeCy_6\). However, the experiments under consideration were carried out only at a single field strength, and the result may have been accidental. Indeed, in the phenomena of magnetostriction—phenomena closely connected with the rotation of elementary magnets but little studied—one observes the fact that iron, after initially lengthening as the magnetic field is increased, then begins to contract again, finally assuming its original length. At this point the internal configuration is apparently close to the original one, and here Compton’s experiment may well give no result.

Ferromagnetism can hardly be explained by features of the crystal lattice; probably the main role is played by a greater or lesser freedom of rotation of the elementary magnets. Owing to an increase in this freedom, Mn, the nearest neighbor of the ferromagnets in the Mendeleev system, exhibits in Heusler alloys its hidden, ordinarily ferromagnetic character; similarly, Fe impurities in Cu increase the susceptibility of the latter far more strongly than follows from a percentage calculation; the same is true of an admixture of liquid \(O_2\) to liquid \(N_2\).¹

The atom is already magnetic as such: the sign of the susceptibility depends on whether or not the magnetic moment of the atom disappears in the absence of a field. Experiment teaches² that part of magnetism depends on the rotation of electrons; however, the dependence of the susceptibility on the atomic number \(N\), which stands out sharply in Dushman’s curves³ (abscissae—atomic \(N\), ordinates—\(\lg\) susceptibility), seems also to indicate the role of the nucleus.

B. Vvedensky.

¹ Perrier a. Kamm. Onnes, Proc. Roy. Ac. Amsterdam, 16, p. 901, 1914.
² See notes ¹), ³), ⁴), ⁵) on the preceding page.
³ Dushman, General El. Rev., May, Aug., Sept., Oct. a. Dec., 1916.

1. J. J. Barnett, Phys. Rev. 6, p. 239, 1915, and 10, p. 7, 1917. ↩

2. A. Einstein and W. Y. de Haas, Verh. d. d. Phys. Des. 17, p. 152, 1915. ↩

3. I. Q. Stewart, Phys. Rev. 11, p. 100, 1918. ↩

4. E. Beck, Ann. d. Phys. 60, p. 109, 1919. ↩

5. G. Arvidson, Phys. Zeitschrift, 21, p. 88, 1920. ↩

6. A. H. Compton and O. Rognley, Phys. Rev. 16, p. 464, 1920. ↩