﻿Absorption of Rout gen Rays in Air. 683 



so well suited for attaining the greatest precision at ordinary 

 temperatures. The results at high temperatures are con- 

 firmed by the expansion of silica, and may probably be 

 accepted with confidence. The apparent discrepancy at low 

 temperatures is receiving further attention, but may be due 

 simply to anomalies in the expansion of solids which are here 

 most in evidence. We make no claim to infallibility, and do 

 not expect anyone to adopt our " comprehensive tables " to 

 the exclusion of others which they may consider more reliable. 

 We ask only that our work may be regarded as an honest 

 attempt to solve a difficult experimental problem, and we take 

 this opportunity of thanking our fellow-workers, especially 

 M. Pierre Ghappuis, for the generous appreciation which 

 they have so freely expressed of our labours. 



LXV. On the Absorption of Rontqen Rays in Air. By 

 A. S. Eve, 2J.Sc, and F. H. Day, M..8c. } McGill Uni- 

 versity, Montreal ** 



Tf^HE coefficient of absorption of the Rontgen rays in air 

 _l was first determined by Rutherford f for a very soft 

 bulb, and the value found was A = '001. A thorough in- 

 vestigation of the absorption of the very penetrating- 

 radiations from a hard bulb was subsequently made by 

 Rutherford and McGlung J, who found A,±=*000279. Two 

 pencils of rays passed down two long cylinders to suitable 

 testing-vessels connected to an electrometer. The pressure 

 in one cylinder was maintained constant, and in the other 

 cylinder it was varied, and in this way the coefficient of absorj> 

 tion by air and by various gases could readily be determined. 

 It has been pointed out by one of us § that, with any 

 ionizing point-source, the coefficient of absorption in the 

 atmosphere near the earth's surface can easily be obtained 

 by varying the distance between the source of the rays and 

 a testing electroscope. If I is the ionization current in the 

 electroscope at a distance r from the source of the rays, I 

 varies as e~ Kr lr 2 * If distances r u r 2 give ionization currents 

 Ii, I 2 , it follows that X may be found from 



log; Ii r^.-log, I 2 r 2 2 =\{r 2 -r,). 

 Preferably a number of distances and ionization currents 

 are measured, and the line 



log e L >2 = A— \r 

 is plotted with r as abscissa and log; I/' 2 as ordinate. The 



* Communicated by the Authors. 



t Phil. Mag-. Jan. 1899. 



| Phil. Trans. Roy. Sbc» (A) cxcvi. p. 25« 



§ Phil. Mog. July 1911. 



