The Diffusion of Imprisoned Resonance Radiation in Mercury Vapor

Abstract

The rate of diffusion of imprisoned resonance radiation in mercury vapor.---The imprisonment of resonance radiation in mercury vapor was studied by measuring the rate at which resonance radiation emerged from one face of a slab of vapor after the exciting light, incident upon the other face, was cut off. The radiation was found to fall off exponentially, and the exponential constant of decay was measured for vapor densities ranging from 0.77\ifmmode\times\else\texttimes\fi{}${10}^{15}$ atoms per cc to 29\ifmmode\times\else\texttimes\fi{}${10}^{15}$ atoms per cc, corresponding to temperatures ranging from 60\ifmmode^\circ\else\textdegree\fi{}C to 130\ifmmode^\circ\else\textdegree\fi{}C. Slabs of two different thicknesses were studied, one 1.95 cm, and the other 1.30 cm. It was found that for vapor densities lower than about 4\ifmmode\times\else\texttimes\fi{}${10}^{15}$ atoms per cc, the exponential constant varied approximately inversely as the square of the thickness of the slab, in qualitative agreement with the theory of the diffusion of imprisoned radiation as worked out by Milne.Extension of Milne's theory: broad exciting line; collisions of second kind.---In this region of vapor densities it was found also that the exponential constant decreased with the vapor density. The failure of Milne's results to give quantitative agreement with this result is discussed and it is suggested that this discrepancy is due to the fact that the exciting light in this experiment was a very broad spectral line, and that postulated by Milne of very narrow width. A rough method of extending Milne's theory to include the absorption and imprisonment of frequencies larger and smaller than the heart of the 2536.7 line is discussed, and it is shown that the experimental results can be explained on this basis. At higher vapor densities, the decay constant increased almost linearly with the number of absorbing atoms per cc. It is shown that this result is in accordance with the theory of imprisonment when extended to include impacts of the second kind. The probability of an impact of the second kind between a normal and an excited mercury atom is calculated from the experimental data and found to be approximately 9\ifmmode\times\else\texttimes\fi{}${10}^{\ensuremath{-}4}$.