of Sodium Vapour excited by One of the D Line*. 459 



when this method was used. Both D lines appear except at 

 the place where the exciting light is diffusively reflected 

 from the horizontal grey strip, and there only one line is 



Fiar. 1. 



^^ 



°2, Pi 



i'S23 



B 



recorded. If the appearance of both lines resulted from 

 reflexion of stray light from the glass, there would be no 

 break in the D x line. 



In the early part of the work we obtained very conflicting 

 results, for on some plates we found no trace of the line which 

 was absent in the exciting light, and in others both lines were 

 present. We finally found that bulbs very carefully freed 

 from hydrogen, and operated at a comparatively low tem- 

 perature, gave always a pure resonance radiation, i. e. one 

 line only appeared. If hydrogen was present, or if the 

 temperature (and consequently the density) was raised both 

 lines were recorded. 



The effect of a change in the density of the pure sodium 

 vapour was first investigated. 



A bulb carefully freed from hydrogen was used for three 

 exposures to resonance excited by D 2 at temperatures of 210° 

 270°, and 340°, keeping all other conditions constant. J) 1 was 

 absent in the case of the exposure made at 210°, it was present 

 though faint at 270°, while at 340° it was nearly as strong 

 in respect to D 2 as in the case of a feeble sodium flame for 



which the intensity ratio ~ =2. 



In making our exposures we always made a preliminary 

 exposure of ten minutes on the MgO patch, following this 

 with the resonance exposure, and ending with a third 

 exposure on the MgO. This furnished a check on the purity 

 of the exciting light, and showed that no change had occurred 

 in the adjustments of the polarizing separator. In our 



