1026 M. L. Dunoyer and Prof. Wood on Photometric 



the spray of a very dilute solution of sodium chloride as a 

 source of light, and forming an image of it on the wall of the 

 bulb by an aplanatic condenser of very large aperture. The 

 sharpness of this image permits of a study of the phenomenon 

 of secondary resonance discovered by one of us in the case of 

 the vapour of mercury and described recently*. 



The surface of the bulb,, illuminated in the manner de- 

 scribed, becomes the source of a resonance radiation of 

 remarkable brilliancy, of a thickness too small to be observed; 

 as the duration of the phenomenon is ten or fifteen hours, it 

 may be investigated photometrically or spectroscopically 

 without difficulty. The preliminary study showed that the 

 intensity of the resonance is much greater if a flame relatively 

 poor in sodium is employed, than with a powerful flame 

 such as is obtained if a fragment of salt is placed on the 

 grill of a Meker burner. If the bulb is heated by a large 

 flame coloured only by the sodium in the air of the room 

 (previously charged by the operation of an intense soda- 

 flame for a few minutes), one observes the yellow glow of 

 the superficial resonance excited by the light emitted by the 

 flame used for the heating of the bulb. The flame must be 

 waved about rapidly over the surface of the bulb, in order to 

 secure a fairly uniform temperature. The phenomenon is 

 less marked if an intense sodium flame is employed. 



These experiments show that the greater part of the T) 

 light of the flame is inoperative in exciting the resonance. 

 Moreover, the intensity of the source appears scarcely 

 diminished if it is viewed through the bulb in which resonance 

 is taking place. In other words, it is only the central cores 

 of the D lines which are effective in exciting the resonance. 

 The same phenomenon was observed in the study of the 

 resonance of mercury vapour already alluded to, the lumi- 

 nosity (ultra-violet) excited by the 2536 line being enormously 

 greater when the exciting mercury arc was first started, than 

 after it had been in operation for a few seconds. 



In the present communication we shall discuss : — 



(1) The photometric study of the diffusing power of the 

 highly attenuated vapour for monochromatic light, 

 as compared with that of a white matt surface of 

 magnesium oxide. 



(2) The conditions under which all of the light removed 



from the exciting beam is re-emitted, giving us a 

 diffuse reflecting power equal to that of the magnesium 

 oxide. 



* Wood, Phil. Mag. May 1912. 



