Page:Proceedings of the Royal Society of London Vol 69.djvu/176

168 radiophone experiment I observed what appeared to be a lateral radia- tion of a clear green light by the vapour.

My first thought was that it was light scattered by the small particles of oxide which are always floating about. If this were case the spectrum of the light should be that of white light after trans- mission through sodium vapour. An examination with the spectro- scope, however, showed that this was not the case, for the spectrum consisted of a diffused red band, a very narrow bright band in the case where the D lines occur, and a diffused band in the green, in which flutings seemed to be present. The general appearance of the light reminds one of the fluorescence of uranium glass.

I am at the present time engaged in an investigation of the spectrum of this fluorescent light, but am not yet prepared to say whether the flutings in the green are discontinuities in the emission or due to the absorption of the non-luminous vapour surrounding the fluorescent vapour. A careful photographic study will be necessary before this can be decided, for which investigation I am employing an iron tube provided with copper jackets at each end, through which water circu- lates. (Fig. 7.) The light emitted laterally is observed through a second iron tube inserted at a right angle, and provided with a glass window. The sodium is placed at the junction of the tubes, which is then raised to a red heat with a blast lamp.

FIG. 7.

So far as I am able to judge by passing the light causing the fluorescence through colour screens, the wave-length of the emitted light corresponds very nearly to that of the exciting vibrations. Interposing a thin cyanine screen, which removes the light in the neighbourhood of the D lines, causes no apparent diminution in the intensity of the emitted light; but the spectroscope shows that the bright band at wave-length 5890 has disappeared. A film stained with aurantia, which removes everything below the " b " group, greatly diminishes the intensity of the fluorescent light, and changes its colour to red. A solution of cuprammonium of moderate strength removes from the light all of the wave-lengths which are capable of exciting the fluorescence. A more careful investigation will be neces-