NOTES 293 



scattered rays, their distribution in the plane containing the 

 primary ray is (except in directions nearly coinciding with 

 that of the primary ray) exactly that required by the pulse 

 theory and not in accord with the particle theory. Later, 

 diffraction experiments by other observers have placed beyond 

 all doubt the identity of X-rays and light, but the difficulty of 

 the energy relations still remains, and exists also in the case 

 of light, where a satisfactory harmonising of the undulatory 

 and quantum theories is one of the outstanding problems of 

 physics. 



The fluorescent X-rays are one of the most important of 

 Professor Barkla's discoveries. They form the main bulk of 

 the secondary X-rays from substances with atomic weights 

 from 50 to 100, and differ from the scattered rays in being 

 unpolarised and equally distributed in all directions even 

 when the primary ray is polarised. They have a definite 

 absorption coefficient depending on the nature of the secon- 

 dary radiator, and not at all on the hardness of the primary 

 beam. Thus successive similar aluminium sheets absorb the 

 same proportion of the rays falling on them, while in the case 

 of the scattered rays successive sheets absorb less and less as 

 the softer constituents of the rays are filtered out. The 

 fluorescent rays are analogous to monochromatic light, the 

 copper rays being distinguished by a definite absorption co- 

 efficient in aluminium, just as sodium light has a definite 

 refractive index in glass ; but the phenomena in X-rays are 

 much simpler than in light, for the hardness of the rays steadily 

 increases as the atomic weight of the substance producing them 

 rises. At first it was thought that the hardness was a periodic 

 function of the atomic weight, increasing from chromium to tin, 

 then decreasing, to increase again from tungsten to bismuth ; 

 but Barkla soon realised (1909) that in the case of the latter 

 elements the rays belong to a different softer series (the L series), 

 so that it would be possible for one metal to give off two dif- 

 ferent sets of fluorescent rays, just as hydrogen gives off three 

 different monochromatic radiations in the visible spectrum- 

 That this was actually the case was experimentally verified 

 (1910), the hard K and soft L rays being obtained from silver, 

 antimony, and tin. Lately (191 6) Barkla has proved the 

 existence of a third, still harder J series of rays from aluminium, 

 oxygen, and carbon. The term "fluorescent " is applied to all 



