﻿Nature of the Ionization hij X-rays. 649 



Mao-. Sept. 1910) was a rougher estimate which Dr. Barkla 

 criticises also. Two years earlier still J made attempts whicli 

 were, I fear, even less accurate. But I hold that under the 

 circumstances all these were right and proper. 



The main part of Dr. Barkla's paper is concerned with 

 experiments on the ionization of gaseous mixtures, and 

 deductions therefrom. The latter may be summed up in 

 three statements. 



(1) Ionization by X-rays is not purely atomic. It depends 

 to a certain extent on the chemical combinations in the 

 molecule. 



This is quite in accord with all similar work on the new 

 rays. The action of the molecule on the ray is an additive 

 effect, but not the action of the ray on the molecule. If the 

 action of the X-ray is assumed to be indirect, the extension 

 of the general principle to the case of X-rays might be 

 inferred from the knowledge that it holds for /3 rays. 



2. The relative ionizations produced by equal absorptions 

 of X-rays in the gases experimented upon are the same as 

 those produced by the corpuscular radiations. 



This is exactly what should happen if the ionization is all 

 indirect. 1 have used the corresponding argument to prove 

 the indirect character of the ionization by the 7 rays (Phih 

 Mag. Sept, 1910, pp. 406-407). 



3. When a mixture of two gases is traversed by the 

 X-rays the observed ionization is the sum of the ionizations 

 that would be observed if the same pencil of X-rays went 

 through the gases separately. 



Dr. Barkla considers that this result proves that much 

 of the X-ray action is direct. Perhaps the point is best 

 understood when the argument is put into mathematical 

 form. 



Let k 1 and k 2 be the mass absorption coefficients of the 

 X-rays in two gases : in many of the examples given by 

 Dr. Barkla the rays were Ag X-rays and the gases were 

 ethyl bromide and air. Let \ 2 and \ 2 De f ne corresponding 

 coefficients in the case of the cathode rays due to Ag X-rays. 

 Let a\ and iv 2 be the masses of gas and air per c.c, and let 

 the ionization in the gas be fi times the ionization in the air 

 for the same expenditure of X or cathode ray energy. 



The ionizations in gas and air would separately be pro- 

 portional to kiwi/A and k 2 io 2 : and Dr. Barkla finds that when 

 gas and air are mixed the ionization is k^p + k 2 w ai on the 

 same scale. 



But if the X-ray action is indirect, we should first suppose 

 that the production of cathode rays is proportional to 



