Analysis of the available structural and optical data on calcite and aragonite type materials suggests a relationship between their birefringence, anion size and cation electronegativity.The structural dependence of the birefringence of calcite and the carbonates and nitrates isostructural with calcite, has been the subject of a number of investigations over the past fifty years. The latest and most comprehensive study is that given by Lo (1973).It was on the subjects of calcite and aragonite that Bragg, in his now classical (1924) paper, first succeeded in correlating birefringence with crystal structure, using the pointdipole theory as developed by Lorentz (1909) andEwald (1921). In his approach, however, Bragg considered optical anisotropy to arise from different dipole-dipole coupling along different crystallographic directions, and neglected the possibility of intrinsic anisotropy in the constituent atoms, particularly oxygen. The converse was done by Bhagavantum (1942), while Lawless & Devries (1964) took into consideration both sources of anisotropy, but evaluated only nearest-neighbour interactions in their analysis. In his study, Lo (1973) computed, in a manner not subjected to the above criticisms, the cation and anion polarizabilities for all the carbonates and nitrate isostructural with calcite. Lo, in assessing the results of his analysis, considered there to be no established trend between birefringence and cation/anion polarizations or with the cation species. He summarized his conclusions as 'within the validity of the point-dipole approximation the present investigation indicates that the polarizability of polyatomic anions in crystals is just as dependent on the nature of the constituent atoms as on the crystalline environment in which they are found.' In all the above mentioned studies it is evident that, without exception, the authors assumed that the dimension of the planar anion, CO 2-, NOr, is invariant and independent of the associated cation species. The reason for this assumption is perhaps obvious, namely that in very few cases have any of these structures been determined with adequate precision to evaluate the C-O or N-O bond lengths. However, the field of a dipole decays as a function of the inverse cube of the distance, and this brings in two considerations. Firstly, it is apparent that the induction effect on an oxygen atom by other atoms will be stronger for atoms in the same planar group than for those in neighbouring groups. Secondly, relatively small changes in C-O or N-O bond lengths could have a marked effect on the resultant polarization of the oxygen atoms in the anion and thus the ultimate birefringence of the crystal.