﻿Magnetic Rotation Spectra of Sodium Vapour. 519 



circumstance that the lines of the different series get into 

 and out o£ step periodically: they may thus be considered 

 analogous to the bands seen when two diffraction-gratings 

 of slightly different spacing are superposed. 



Barium Excitation. 



The fluorescence excited by the barium arc appears to be 

 due chiefly to the line 4934, which coincides with one of the 

 extra lines in the magnetic spectrum. Line 4932 o£ the second 

 magnetic spectrum is also very near it; and we find that the 

 fluorescence spectrum contains lines which coincide with the 

 magnetic lines of the second series, as well as lines which 

 coincide with some of the extra magnetic lines. The barium 

 arc contains a good many other fainter lines which may give 

 rise to some of the fluorescent lines. It will be necessary 

 to repeat the experiment with the 4934 line isolated. 



Sodium Excitation. 



As I showed in the earlier paper referred to, if we stimulate 

 the vapour with intense sodium light, we obtain a yellow 

 fluorescence which the spectroscope shows to be made up of 

 two lines in the position of the D lines. AVe have here a 

 re-emission of light of the same wave-length as the exciting 

 light, and nothing else. This I have called resonance radiation, 

 as we may find that it is different from fluorescence, though 

 the two are doubtless intimately related. As there are a 

 number of pairs of lines in the ultra-violet which belong to the 

 same series as the D lines, it seemed of great importance to 

 determine whether these appeared in the spectrum of the 

 fluorescence excited by the sodium flame. The sodium tube 

 was provided with a quartz window, and the light of the 

 oxyhydrogen flame, heavily charged with sodium, focussed 

 upon the aperture of the retort with a glass lens. White 

 light from the arc was also used, as this excites the D-line 

 vibrations in the fluorescence. The spectrum was photo- 

 graphed with a small quartz spectrograph, and though the 

 D lines were greatly over-exposed, no traces of any of the 

 ultra-violet doublets were found on the plate. Conversely, 

 illumination with ultra-violet light in the region of the first 

 ultra-violet pair of lines failed to produce any visible fluor- 

 escence. It was hoped that a faint yellow fluorescence 

 might be produced in this way, due to emission in the region 

 of the D lines. I have not yet tried stimulation with D x 

 and D 2 , alone, to see whether both D lines appear in the 

 fluorescence. This will be a very difficult experiment, and I 



