334 
PROF. C. G. BARKLA ON X-RAYS AND THE THEORY OF RADIATION. 
Table IV. 
Approximate 
wave-length 
of primary 
radiation. 
Number of L, M, ... electrons emitted by Br 
Number of K electrons from Br 
Number of K, L, M, ... electrons emitted by air 
for equal absorptions of primary radiation. 
Number of K, L, M,... electrons from air 
for equal absorptions of primary energy. 
i 
i 
cm. 
1-54 xlO-s 
1-17 
1-17 x10 -8 
1-11 
1-10 x10- 8 
1-10 
0-87 xlO" 8 
0-49 
0-71 x10~ 8 
0-59 
0-56 x10 -8 
0-72 
0-49 x10~ 8 
0-81 
0-355 x 10~ s 
- 
0-82 
This shows the numbers of electrons emitted by ethyl bromide and by air to be 
about equal when the primary radiation is of longer wave-length than that 
characteristic of bromine (K ai = 1'035 x 10~ 8 cm.). # 
First Approximation. 
The significance of these results is most easily understood if at first we express 
them approximately. 
Let the wave-length of the primary radiation be m, and that of the K fluorescent 
(characteristic) radiation of the substance bromine be /jl k . Then when /u is greater 
than n K , for a given absorption of the primary radiation, the same number of electrons 
is emitted by ethyl bromide (L, M, N, &c., electrons) as by air (K, L, M, N, &c., 
electrons), as shown in Table IV., column II. 
When /a is just less than /m K , the additional (K) absorption in ethyl bromide is 
accompanied by the emission of only half as many electrons—that is, only half the 
K energy absorbed is re-emitted as energy of K corpuscular radiation (column III.).t 
* It might be suggested that the sudden decrease in ionization is possibly due to a sudden decrease not 
of the number of electrons emitted, but of the total ionizing power of each electron when its velocity is 
increased beyond a critical value—-that necessary to excite the K fluorescent radiation. There is, however, 
plenty of evidence against this—(, a ) theoretically it is highly improbable; (6) such changes in relative 
ionizing power with velocity have not been observed; and conclusively (c) the decrease above described in 
this case is true when the electrons are absorbed not in the substance itself, but in another gas (see p. 336). 
f We here assume that the energy of an X-radiation is approximately equal to that of the full corpuscular 
radiation which it excites in a substance completely absorbing it, when the conditions are such that little 
energy is re-emitted in the form of either scattered or fluorescent (characteristic) X-radiation. The 
evidence for this is very weighty and may be summarized as follows :— 
When during the transmission of a primary radiation through matter there is very little transformation 
into corpuscular radiation, practically the whole of the loss of energy is accounted for by the process of 
