Absorption of X- Rays. 



473 



column 3, those found by the same scientists by character- 

 istic radiation from Cr, Fe, and Co ; and in column 4, those 

 found by myself, by using non-homogeneous rays from an 

 X-ray bulb with an anticathode of tungsten which have been 

 filtered through an aluminium screen : — 



Table I. 



Elements. 



1. 



2. 



3. 



Bragg and Pierce 



Barkla and Sadler 



with rays from 



with rays from 



Ag, Eh, Pd. 



Ag- 



Cr, Fe, Co. 



15-2 



14-3 



1-69 



1-30 



1-37 



1-22 



117 



1-15 



1 21 



113 



1-14 



1-22 



163 



163 



27-7 



7-64 



7-69 



1-61 



133 



1-37 



1-37 



105 



110 



[1-06] 



7-28 



7-86 



[IPS] 



4. 



Auren 

 with non- 

 homogeneous 

 ravs. 



Fe/Al. 

 Ni/Fe. 

 Cu/Ni 

 Zn/Cu 

 Pt/Al . 

 Pt/Ag. 

 Sn/Ag. 

 Au/Pt 

 Au/Cu 



13-4 

 1-30 

 1-10 

 1-15 

 156 

 1-76 

 1-04 

 1-02* 

 8-29* 



* Au=540 from curve, fig. 2. 



The agreement between the various experiments is, on the 

 whole, remarkably good, and where sensible changes have 

 been observed they may easily be accounted for. Within 

 the range of selective absorption, this law does not hold, 

 and, from what I have found, neither does it hold when rays 

 of greater wave-length (soft rays) are examined. Though 

 the agreement between the values of the first, second, and 

 fourth columns, in all cases but one, is good, the values of 

 the third column differ from those in regard to the combi- 

 nations Fe/Al and Pt/Al, which are too low. As for the com- 

 bination Pt/Ag, the values in the first and second columns 

 agree between themselves, but these values do not agree 

 with the values in the third and fourth columns, which, on 

 the other hand, closely agree between themselves. The 

 cause of the want of agreement between the said couples 

 certainly depends on the fact that one of the constituent 

 metals of the combination enters into the range of selective 

 absorption in using the radiation in question. As it appears 

 from the experiments of Barkla and Sadler, it is not here the 

 case that the relation between the absorption coefficients is 

 independent of the wave-length. 



The ratio of the atomic and molecular absorption coeffi- 

 cients respectively is, i'n the ensuing pages, expressed by 



Phil. Mag. S."6. Vol. 33. No. 198. June 1917. 2 L 



