PROF. C. G. BARKLA ON X-RAYS AND THE THEORY OF RADIATION. 
355 
the primary radiations were of longer wave-length than these characteristic radiations. 
This denotes that the number of J electrons emitted from air bears the same ratio to 
the total number of electrons (K, L, M, &c.) emitted by air as is the case from 
aluminium or sulphur. 
Fluorescence Method. 
The more direct method of demonstrating the existence of a characteristic radiation 
is to excite it by a primary radiation of shorter wave-length and observe it in the 
secondary radiation emitted. The secondary radiation emitted by the substance and 
detected by an electroscope in many cases is very approximately a homogeneous 
radiation ; in others a second (softer) characteristic radiation must be filtered out by 
absorbing sheets before the homogeneity of the radiation can be demonstrated; in 
other cases again, the scattered radiation produces considerable effects and cannot 
be filtered out, but in this the effect of the scattered radiation can be calculated and 
eliminated. The last method was used by Barkla and Nicol, # and was afterwards 
applied by Chapman t to substances of highest atomic weight. In this case the 
whole of the secondary radiation emitted by a substance was compared with that 
emitted by another substance whose secondary radiation under the condition of the 
experiment was almost purely a scattered radiation. 
In experimenting upon light elements this scattered radiation becomes a very 
important factor, and the elimination of the effect of this from the whole secondary 
radiation becomes very difficult. For our present purpose it is sufficient to show 
that in addition to scattered radiation a characteristic X-radiation is emitted by the 
light elements. As scattering substances, paper, paraffin wax, and aluminium were 
used. It had previously been shown that X-radiations differing considerably in wave¬ 
length are scattered to an equal extent by these light elements, and that scattering 
takes place without appreciable change of type. The introduction of a fluorescent 
characteristic radiation into the secondary radiation from one of these light elements 
would be shown by an increase in the total intensity of secondary radiation, and by 
a change in its penetrating power. 
A primary radiation of varying wave-length was directed on to a sheet of paper, 
paraffin w r ax, or aluminium, and the ratio between the intensities of the primary and 
secondary radiations was observed by means of two electroscopes, while the absorb¬ 
abilities of both primary and secondary radiations were tested by screens of aluminium. 
It was found (l) that the intensity of the secondary radiation rose and fell as the 
wave-length of the primary radiation was diminished in the same way as did that of 
a characteristic radiation ; (2) the penetrating power of the secondary radiation 
varied as it would by the superposition on the scattered radiation of a homogeneous 
characteristic radiation of intensity varying as above; (3) in aluminium the 
* ‘ Proc. Phys. Soc. Lond.,’ September, 1911. 
t ‘Roy. Soc. Proc.,’February, 1912. 
3 D 
VOL. CCXVII.—A. 
