318 
PROF. C. G. BARKLA ON X-RAYS AND THE THEORY OF RADIATION. 
is the atomic weight and e the electronic charge. (This nucleus, together with the 
number of electrons found from the scattering of X-rays, constitutes a neutral atom.) 
Bohr* on certain assumptions has shown that such an atom is capable of forming 
the basis of an explanation of the line spectra of hydrogen and helium, agreeing in a 
most remarkable way with experimental fact. 
Moseley, t too, has given further evidence in support of the conclusion as to the 
number of electrons per atom. His “ atomic numbers ” are identical with the values 
given above. 
This agreement is probably much more remarkable and significant than it at first 
sight appears. For apart from the orthodox electromagnetic theory of radiation, the 
intensity of the scattered radiation might have been anything between say xoW part 
and 1000 times what was experimentally observed; yet the value experimentally 
determined twelve years ago—long before the precise significance of the results could 
be realized—agrees within a very small fraction with that to be expected on the modern 
conception of the atom and the old electromagnetic theory of radiation. Originally, 
while assuming this theory to hold, I used it to determine the number of electrons 
per atom. It is, however, perfectly legitimate now that the conclusions have been 
confirmed as indicated above, to use this agreement as evidence in support of this 
theory of radiation. The chances of such an agreement being accidental are almost 
infinitesimal. 
Looking at the result in a somewhat different way, if for instance we imagined the 
gaseous ions to be the scattering units and calculated the number necessary in order 
to produce the scattering experimentally observed, we should find a number some 
millions of times the total number of molecules nresent in the scattering substance, 
that is, a number probably trillions of times too many. Yet applied to determine the 
number of electrons, the theory gives the number now accepted, certainly within a 
possible 5 to 10 per cent, error. 
The theory on which this is based is the spreading wave theory. It assumes that 
the scattered radiation is the radiation resulting from the disturbance in electrons 
while under the influence of—in fact while controlled byj—the primary radiation. 
Thus every electron in the matter traversed is influenced by every wave in the 
primary radiation passing over it, and receives a definite acceleration, f say, which 
is proportional to the electric intensity in the primary beam. The rate of radiation 
of energy from an electron with charge e and acceleration f is f/n'l/ '/V—as first shown 
by L armor. This summed up for all electrons gives the rate of emission of scattered 
radiation by those electrons ; and this is readily expressed in terms of the rate of 
transmission of energy of primary radiation through the substance, for the acceleration 
* ‘Phil. Mag.,’November, 1913. 
t ‘ Phil. Mag.,’ December, 1913. 
t At least the motion of the electrons resulting in radiation is not appreciably affected by atomic 
forces. 
