The afterglow in ionized mercury vapor was studied as a recombination phenomenon, to determine how its intensity varied with concentration of electrons and of positive ions, vapor pressure and the mean energy of the electrons, the last expressed as a temperature, $T$. The afterglow was observed in a region outside of the arc to which positive ions and electrons were carried by a stream of vapor, auxiliary electrodes being used to control the electron temperatures. The total number of electrons, $N$, was assumed closely equal to the number of positive ions. The intensity of the visible part of the afterglow radiation was shown to be closely proportional to the total intensity under the experimental conditions, and was therefore used as a measure of the rate of recombination. This intensity, $I$, was found to be proportional to the first power instead of to the square of the electron concentration, with possible exceptions at low concentrations and low electron temperatures. $I$ varied with vapor pressure, increasing roughly as the fourth power of the pressure. With increase in electron temperature from 1300° to 5600°K, $I$ decreased by a factor of 2500. From this rapid change in $I$ it is concluded that the afterglow cannot be due to a simple recombination process, since work of other authors with caesium and helium has shown that $I$ should then vary not more rapidly than the reciprocal of the temperature. The results were found to fit either of the following empirical equations: $I=\mathrm{const} \frac{N^2}{N_0}$ or $I=\mathrm{const} \frac{N^2}{N_1}$, where $N_0$ is the number of fast electrons having energies greater than a determined value and $N_1$ the number having a definite energy. The values of the critical energies corresponding to these two equations were found to be 1.15 volts and 1.35 volts, respectively.

The results suggest that the recombination in mercury vapor takes place in two stages, of which the second is responsible for the radiation of the series lines, and that the effect of the fast electrons is to reionize from the first stage.

Considerable difficulty was experienced during measurements as the result of changes in the contact potential of the exploring electrodes used for measuring electron temperatures and concentrations. These changes were apparently due to traces of oxygen.

• Received 6 December 1930

DOI:https://doi.org/10.1103/PhysRev.37.182