[ 626 ] 



LXIV. The Absorption Coefficients of X Rays. By W. H. 

 Bragg, M.A., JJ.Sc, F.R.S., Cavendish Professor of Physics 

 in the University of Leeds, and S. E. Peirce, B.Sc, 185 L 

 Exhibition Scholar, Sydney, Neio South Wales *. 



THE new method of analysis of X rays by reflexion from 

 crystals gives us the power of isolating for examination 

 homogeneous pencils of definite wave-length. It is important 

 to take advantage of this possibility in order to examine the 

 relation between wave-length and absorption coefficient. 



Barkla has already made accurate and comprehensive 

 measurements of the absorption coefficients of X rays of 

 various qualities by substances of various composition. He 

 has defined the quality of the X ray by reference to the 

 substance which emits that particular ray when duly stimu- 

 lated. In his actual experiment primary X rays from an 

 X-ray bulb were allowed to fall upon a sheet of zinc, silver, 

 arsenic, or any other substance, and the secondary rays 

 emitted by the substance were examined in respect to their 

 power of penetrating various screens. A summary of the 

 work of Barkla, and of others who have extended his inves- 

 tigations, is given on pp. 132, 133 of Kaye's recently pub- 

 lished book on X rays f. 



The older method of defining the quality of X rays must 

 naturally give place to the more quantitative and direct 

 definition in terms of wave-length. This is all the more 

 necessary because it has been found that the secondary rays 

 emitted by any one substance are not entirely homogeneous, 

 as was first supposed, but can be analysed into line spectra 

 of greater or less complexity. 



In this paper we give some results obtained by the use of 

 the strictly homogeneous beams of X rays which are now 

 available. They are restricted to a small part of the large 

 field which is open to exploration. The wave-lengths 

 examined are those emitted by anticathodes of silver, palla- 

 dium, and rhodium, each of which substances yields a spectrum 

 constituted mainly of two well-defined and intense lines. 

 Each wave-length is isolated in turn by reflexion from a 

 crystal of rocksalt, and the absorption coefficients are 

 measured in the usual way. The difficulties of the experi- 

 ment are mainly two, the inconstancy of X-ray emission 

 and the need of allowing for the effects of the u general" or 

 " white " radiation. In the cases here considered the effect 

 of the latter difficulty is not very great, because the " lines " 



* Communicated by the Authors, 

 t Long-mans, Green & Co., 1914. 



