PROF. C. G. BARKLA ON X-RAYS AND THE THEORY OF RADIATION. 
329 
observation of the energy transformations of an exciting primary radiation through a 
long range of wave-lengths ; (3) the energy of fluorescent radiation from bromine is 
not far removed from the limiting value observed ; the behaviour of bromine is thus 
probably typical of the greatest number of elements, and for this reason the experi¬ 
mental results are the least open to question. 
mi i , • i , K fluorescent radiation (energy) i ,, L . 
the relation between —- 7 -,- 7 - f - ' ■ and wave-length of primary 
K. primary absorption (energy) 
radiation, when bromine is the fluorescent substance, is given in fig. 4, along with the 
position of the bromine spectral lines of series K, and is reproduced on a different 
scale in fig. 5, curve F. The continuous portion of the curve we may thus regard as 
obtained by direct experiment. It should be pointed out that the wave-length of 
the K a , line has been adopted throughout. 
There is much evidence, too, that the curve must turn through the origin as 
indicated by the broken line (fig. 5). This represents a small, decreasing, and 
ultimately vanishing amount of fluorescent X-radiation emitted by a substance 
exposed to a primary radiation of small and decreasing wave-length. 
Thus (l) Barkla and Philpot # showed that when a primary radiation is of much 
shorter wave-length than any characteristic radiation it excites, its absorption results 
in the emission of approximately the same number of electrons, whatever the absorb¬ 
ing substance; it follows from the complementary property of the characteristic 
(fluorescent) X-radiationf that there cannot be varying amounts of fluorescent 
radiation unless this is so weak that variations would not be observable. This means 
that all the curves must pass through the same point; the only point through which 
all the lines in fig. 4 can .turn consistently with the experimentally observed results 
is the origin, or a point near to it. 
( 2 ) There is little doubt that air like other substances emits a K radiation which 
is of very long wave-length compared with an ordinary primary beam of X-rays; yet 
C. T. R. Wilson’s cloud experiments show few, if any, very short tracks such as 
would be produced by the re-absorption of such a radiation if emitted in appreciable 
intensity. We conclude that with a primary radiation of comparatively short wave¬ 
length, the fraction of energy transformed into fluorescent X-radiation becomes very 
small. 
(3) Again, in the most accurate measurements by Beatty and by Bragg of the 
corpuscular radiation from metal sheets, the possibilty of the existence of such soft 
fluorescent radiations in any large quantity has been entirely neglected ; yet the 
results show sufficient consistency to indicate that the characteristic radiations of 
longer wave-length must account for only a small portion of the energy of primary 
radiation absorbed. 
* ‘ Phil. Mag.,’ June, 1913. 
f See p. 336. 
