Diamagnetism and crystal structure

Prof. Ehrenfest has suggested (Physica, vol. 5, p. 388, 1925) that the high diamagnetic susceptibility of bismuth is to be ascribed to the existence in the metallic crystal lattice of electron orbits of large area including several atoms within their radius. There seems good reason to extend Ehrenfest's hypothesis to the case of carbon as well, since it affords an illuminating insight into the magnetic behaviour of the different forms of this element. It is known that graphite possesses a high specific susceptibility, which according to the most recent measurements of Vaidyanathan with carefully purified samples, is -5.1 &#215; 10<SUP>&#8722;6</SUP>, that is, quite ten times larger than the specific susceptibility of diamond (-0.49 &#215; 10<SUP>&#8722;6</SUP>), the latter being practically the same as that of carbon in organic compounds as found from Pascal's additive law. The abnormal susceptibility of graphite becomes intelligible in terms of the peculiar structure of the substance and its electrical conductivity, if we assume that there are electron orbits circulating round the plane hexagonal rings of carbon in the crystal-lattice. This fits in with the known fact (observed by Honda and Owen) that the susceptibility of graphite is six or seven times greater normal to the planes of cleavage than parallel to them. Diamond, on the other hand, being a dielectric would naturally not show the abnormal susceptibility