Intensities, corrected for absorption, of the principal spectral lines of wave-lengths 2259 to 11,289A are given for 4-ampere mercury arcs operated at vapor pressures of 0.03, 20, 450, and 500 mm (Hg). The variation across the arc of the intensity per unit area is given for twenty of these spectral lines. Plots of spectral intensity data were made on two alternative hypotheses: (1) That the concentrations of atoms in excited states are in a Boltzmann equilibrium with one another and that transition probabilities are proportional to $\frac{w_{\mathrm{AB}}}{\left[{\lambda }^2\right({n_A}^{*}+{n_B}^{*}\left){({n_A}^{*}-{n_B}^{*})}^2\right]}$, where $w_{\mathrm{AB}}$ is the Kronig a priori probability, $\lambda$ is the wave-length of the emitted radiation, and ${n_A}^{*}$ and ${n_B}^{*}$ are the effective quantum numbers of the upper and lower levels; (2) that all downward transitions are radiative and that the rate of excitation to a given level is proportional to the product of the a priori probability of the level and the concentration of electrons with energies which do not differ from the minimum for excitation by more than a constant amount. A Maxwellian distribution of velocities was assumed. The method of plotting is such that, if either hypothesis were correct and if the corrections for absorption were of the proper magnitude, the points on the corresponding plot would all fall on a straight line. Characteristic deviations of the points from a linear relationship are found on each type of plot. Both types of plot indicate that the electron temperature varies little with distance from the axis. This is in disagreement with Elenbaas' theory of the mechanism of a high pressure mercury arc.

Close

PrevNext

View figure in article