The refractive index dispersion and polarization of gases

Abstract

In recent years, values of the refractive index of liquids, and sometimes of gases, have been extensively used for determining dipole-moments. Measurements of dielectric coefficients give the sum of the orientation, electronic and atomic polarizations P 0 + P e +P a , while the electronic polarization P e can be calculated approximately from the refractive index. If the dispersion is known, P e should be given with some accuracy by calculating the refractive index at zero frequency. In order to obtain P 0 , it is necessary to know P a , a quantity which has only been determined for a few substances, the infra-red spectrum of which has been examined in detail. It is, however, often assumed that P a is negligible or, at least, small in comparison with P e . In some cases, an approximation on an empirical basis has been made by assuming that P a is proportional to P e , or that P a + P e is equal to the value of P e when calculated from the refractive index for the yellow sodium line. Such assumptions can hardly be regarded as satisfactory. In liquids, determination of the polarization is complicated by the necessity of employing a solvent, but, with gases, no such difficulty exists. P a can therefore be determined with some accuracy by measuring the refraction for different wave-lengths, and the dielectric coefficient at different temperatures. Numerous measurements of the refractive index of the commoner gases have been made, but the results of different observers are not always in agreement and frequently the dispersion has not been determined. The dielectric coefficient of the same sample of gas seems never to have been measured.