Regular Reflexion of Light by an Absorbing Gas. 337 



Polarization Experiments. 



Renewed attempts have been made to detect traces of 

 polarization in the scattered resonance radiation, but without 

 success. Even when the vapour is illuminated with plane 

 polarized light, and the spot of surface luminosity (which we 

 have at say 100° ) is photographed through a Savart plate or 

 Fresnel double prism of R. and L. quartz with a polarizing 

 analyser of quartz and Iceland spar, no trace of the fringes 

 appear. The same thing occurs in the case of sodium vapour 

 illuminated with a sodium flame. This is very remarkable, 

 since strong polarization has been found associated with the 

 stimulation of the frequencies corresponding to the band or 

 channeled spectra of sodium and iodine vapour, the polari- 

 zation showing not only in the line directly excited by the 

 monochromatic light (resonance radiation), but also in all of 

 the other lines (lines of the resonance spectrum*). 



In our present search for traces of polarization we employed 

 the Fresnel double-image prism of right- and left-handed 

 quartz. This gave, with polarized green mercury light, on 

 analysation, six black horizontal bands, while with the light 

 of the 2536 line about thirty bands could be counted. They 

 were very distinct, however, in the photographs made with 

 the quartz camera. As an analyser we used a double-image 

 prism of Iceland spar and quartz, which we made by grinding 

 and polishing a prism of about 8° from a small piece of spar, 

 securing direct vision and fair achromatization by com- 

 pensating it with a small quartz prism of about the same 

 angle. Both prisms were ground and polished in less than 

 an hour. 



The double-image prism was mounted in front of the lens 

 of the camera, and the aperture considerably reduced by a 

 diaphragm to secure sharp definition of the fringes. In 

 some of onr experiments we employed a small Foucault 

 prism as a polarizer, but the loss of light was considerable, 

 and we accordingly cut a 60° prism of quartz perpendicular 

 to the axis; this, when mounted about 10 cm. behind the 

 monochromator, in the converging beam of 2536 light, gave 

 two brilliant polarized images on the uranium glass plate, 

 separated by a distance of about 3 cm. One of these was cut 

 off by a screen, and the rays diverging from the other 

 illuminated the prismatic plate of the bulb. 



♦ R. W. Wood, Phil. Mag. July 1908, and Oct. 1911, p. 480. 



_ 



