PROF. C. G. BARKLA ON X-RAYS AND THE THEORY OF RADIATION. 
323 
Tims the total absorption E may be written 
E = S + Ej + E k +E l +. 
in which all the right-hand terms appear to be independent quantities. 
Corpuscular Electronic Radiation .—In addition it may be shown that the corpus¬ 
cular radiation ( i.e . the radiation consisting of high speed electrons emitted from 
substances exposed to X-rays) may be divided into several distinct groups, each 
associated with the emission of a fluorescent X-radiation of a particular series ; thus the 
total corpuscular radiation C = Cj + C K + C L +... where Cj, C K , C L , are corpuscular 
radiations definitely associated with the emission of J, K, and L characteristic 
radiations respectively. This we shall proceed to show. 
C. T. ft. Wilson’s condensation experiments* show that X-ray ionization is due to 
the corpuscular radiation which the X-rays excite in the gaseous substance ionized. 
(All other investigations of less direct nature which at first appeared to indicate this 
only to a partial extent may now be shown to confirm this conclusion as accurately as 
the possible errors of experiment allow.)! Wilson’s experiments indicate, too, that 
there is little or no variation in the velocity of the electrons ejected by X-radiation of 
one wave-length. For the length of path of an electron was shown by Whiddington 
to vary as the fourth power of the velocity; consequently an electron with half the 
maximum velocity of ejection would have a path only one-sixteentli that of the other. 
If then the velocities of ejection ranged from zero to the maximum, Wilson’s 
experiments would show a very large number of short paths. Simple observation of 
the photographs showing the trails of the ejected electrons is sufficient to convince 
one that little variation of initial velocity exists, and that such variation as is 
observable in the lengths of the trails is probably due either to the heterogeneity of 
the primary beam, or to fore-shortening of the trails. Ionization of air by homo¬ 
geneous X-rays is therefore through the agency of electrons emitted with one velocity, 
approximately, if not accurately. 
From experiments on ionization it may readily be shown that the corpuscular 
radiation consists of several independent groups each associated with the emission 
of a particular fluorescent X-radiation. For the ionization of a substance by X-rays 
of varying wave-length varies with the wave-length of these rays in a manner similar 
to that of the absorption. Thus the ionization in a substance E. say, is proportional 
to the ionization in a second substance for a range of wave-lengths not near to a 
spectral line of either on its shorter wave-length side. As the wave-length of the 
primary X-radiation is diminished beyond that of a spectral line of R, there is a 
sudden rise in the ratio of ionizations, and the ratio gradually approaches a constant 
* ‘Roy. Soc. Proc.,’ 1912. 
t This refers to the investigations of Barkla, Bragg, Beatty, Barkla and Philpot. Bragg was 
the first to insist that ionization was entirely due to the high speed corpuscles. 
VOL. CCXVII.—A. 2 Z 
