Experiments on Polarized Rontgen Radiation. 793 



about 12 cm. from the radiator R 2 , two similar electroscopes, 

 Ai horizontal, and A 2 vertical, were placed so as to subtend 

 the same solid angle at the plate R 2 , the face of which was 

 equally inclined to vertical and horizontal lines drawn from 

 the centre of the radiator to the electroscopes K x and A 2 . 



V::- T 



A^f Dotted liTies) vertically above R 2 



The substance which was being tested as regards its power 

 of rotating the plane of polarization was placed at XX, so 

 that the secondary partially polarized beam of rays from R x 

 passed directly through it before striking R 2 . A full account 

 of this apparatus and the conditions which it must satisfy is 

 given in the original paper *, where it is demonstrated that 

 while there is but little polarization in the primary beam, 

 yet in the secondary beam there is a considerable amount of 

 polarization. This follows from elementary considerations 

 of the usual theory of scattering. 



In the present experiment, the primary and secondary 

 beams were horizontal. When this was the case, and both 

 the radiators at R x and R 2 consisted merely of a scattering 

 agent such as carbon, it was found that the horizontal 

 tertiary beam passing into A x was of much greater intensity 

 than the vertical tertiary beam passing into A 2 . When, 

 however, the carbon plate at R 2 was replaced by a plate 

 capable of giving a preponderating characteristic radiation, 

 which has been shown to be distributed equally in all 

 directions f, the intensities of the radiation reaching the 



* Barkla, Proc. Roy. Soc. Series A, vol. lxxvii. pp. 247-255. 

 t Barkla and Ayres, Phil. Mag. Feb. 1911, pp. 271-278. 



> 



