Effect of a weak magnetic field on the polarization of the fluorescent light from mercury and sodium vapor.—In explanation of the observations of Wood and Ellett, it is suggested that, in accordance with Larmor's principle, the field causes a precession of the electric vibrators with an angular velocity $\frac{1}{2}\left(\frac{e}{m}\right)H$ and this precession is shown to be quite competent to explain the observation if we assume a proper duration of the excited state or proper damping constant $\alpha$. The experimental results for mercury $\lambda 2536$ give a value for $\alpha$ of 1.14×${10}^7$ per sec., in good agreement with other values. This classical theory is substantially the same as the classical theory of the Zeeman patterns, and does not account for the behavior of lines showing a more complicated Zeeman pattern such as the sodium lines studied by Wood and Ellett. The fundamental fact of the fluorescent light being plane polarized is hard to reconcile with the assumptions of the quantum theory of atomic radiation.

• Received 29 March 1924

DOI:https://doi.org/10.1103/PhysRev.24.234