﻿318 Prof. Barkla and Mr. Simons on Ionization 



raise 104 per cent. to about 130 per cent. There are possibilities 

 of much greater error, but this is one which may be easily 

 corrected, and it shows that the conclusion from the experi- 

 ment is not what at first sight it appears. 



Again, in a later paper by Bragg and Porter (Proc. Roy. 

 Soc. A. vol. lxxxv. 1911), results are given with the object of 

 showing that all the ionization produced in oxygen by X-rays 

 can be accounted for by the secondary corpuscular radiation. 

 A serious error, however, completely changes the significance 



of the results. The mass absorption coefficient (-1 for Sn 



X-rays in oxygen is found by experiment to be *425, and it 

 is assumed that all of the energy so accounted for is truly 

 absorbed in the oxygen. As a matter of fact, *2 of the '425, 

 or nearly one half, is scattered*, and in the experiments de- 

 Fcribed practically all this was lost. If, then, we accept all 

 the experimental data given, the final comparison is no longer 

 between 34 and 36'3, but between 34 and G8'4 ; in other 

 words, only half the ionization in oxygen is accounted for. 



We do not venture to use these results as an argument 

 against the hypothesis of ionization by secondary corpuscular 

 (electronic) radiation. We do say, however, that these ex- 

 periments furnish no direct evidence in favour of it. For 

 the present we must leave the question open, with the ex- 

 pression of opinion that on the whole the balance of evidence 

 is against. 



Beatty's experiments f probably give us the most exact in- 

 formation on the point. He concluded that fractions varying 

 up to 75 per cent, of the total ionization in seleniuretted 

 hydrogen were due to corpuscular radiation. But, again, these 

 results give only the order of magnitude. A small error in 

 his assumptions might easily change 75 per cent, into 100 

 per cent. Neither experiments on gases at various pressures 

 nor in chambers of various lengths show the effect of 

 extremely easily absorbed corpuscular radiations such as 

 might be and in all probability are set up by very soft 

 fluorescent X-radiations produced by the primary X-rays. 



It was thought that experiments made upon the ionization 

 produced by X-rays in certain gaseous mixtures would 

 probably throw some light on the processes involved, for, as 

 will be seen, we can to some extent distinguish between the 

 direct and secondary effects, possibly even though the secon- 

 dary rays be so soft as to be absorbed by the molecules in the 

 immediate neighbourhood of the radiating atoms. 



* Barkla, Phil. Mag. May 1904; Mav 1911. 

 f Proc. Roy. Soc, A. vol. lxxxv. 1911. 



