Abstract
It has been shown by Bragg (1924) that the birefringence of anisotropic crystalline materials can be ascribed in part to the anisotropic dependence of the magnitude of the induced internal electric field on the electric vector of an incident light wave set at differing orientations to the crystalline axes. The internal field depends on positional correlation between pairs of particles, and if this is anisotropic the induced field depends on the relative orientation of the electric vector to the symmetry axes of the pair correlation function. The square, of the refractive index m of the material depends on the ratio of the induced electric field to the applied field, and, when this ratio depends on the orientation of the applied electric field vector, m2 will have tensor-like properties—at least in so far as it will have three (in general) principal axes and values. In condensed phases the spherical symmetry of individual isolated atoms is lost and a second source o f birefringence resides in the ordered orientation o f individually anisotropically polarizable particles. In so far as it is also mathematically convenient, when treating condensed systems, to deal with the polarizability of any group of atoms which retains its group structure over long periods of time as that of a single entity, birefringence must a fortiori be ascribed also to an intrinsic anisotropy of polarizability of individual particles. Nitta (1940) therefore described the observed birefringence in certain tetragonal crystals in terms of an anisotropically polarizable unit corresponding to the content of one unit cell localized on tetragonal lattice points