There is, at present, a discrepancy of 162×${10}^{-6\mathrm{}}$ between the temperature invariant part of the dielectric constant of HCl and the square of the index of refraction extrapolated to infinite wave-length; a difference which is more than 10 times the estimated uncertainty of the measurements, and which is usually assumed to be due to the contribution of infrared vibration bands to the index. In order to test the validity of the above assumption and also to investigate a seldom used method of determining the effective charge of molecules, the optical dispersion of HCl has been measured between 1 and 10μ. These measurements yield a value of (1.00±0.05) × ${10}^{-10\mathrm{}}$ e.s.u. for the effective charge of the rotator-vibrator; which shows, in agreement with measurements of intensity of absorption, that the contribution of the rotator-vibrator to $n_{\infty }$ is much too small to explain the above discrepancy. Since the uncertainty in ${n_{\infty }}^2$ as obtained from the refraction measurements is not more than 3×${10}^{-6\mathrm{}}$, this indicates that the uncertainty in the dielectric constant measurements must be somewhat larger than is ordinarily assumed. During the course of the investigation, it was discovered that the contribution of the pure rotation to the index of refraction was large enough to be measured with a fair degree of accuracy. Since most of the rotators are in excited states, these data furnish a quantitative check of the negative terms in the Kramers dispersion formula. They also furnish a new means of determining the electric moment; and the fact that the moment which is obtained agrees with that obtained from dielectric constant measurements indicates that the low absorption intensities observed by Czerny and by Badger must be in error.

• Received 11 September 1935

DOI:https://doi.org/10.1103/PhysRev.48.779