PROF. C. G. BARKLA ON X-RAYS AND THE THEORY OF RADIATION. 349 
Consequently the number of quanta of radiation absorbed per atom (i.e., per 
absorbing atom) varies from about two to just more than one as the wave-length of 
primary radiation diminishes from just less than /x K to a very small quantity. A 
quantum is a quantity of radiation evidence of the absorption of which has not been 
obtained. Frequently however—when /u is a small fraction of /u K —the absorption per 
atom approximates to the energy of a quantum. 
It should be pointed out that the last two conclusions concerning the quantum 
theory, although expressed in precise terms, are in no way dependent on a close 
agreement between the experimental results and those given as a first approximation. 
A New Series of Characteristic Radiations. (J Series.) 
A marked deviation of the experimental results from what was expected on the 
above theory has led to the discovery of a new series of characteristic X-radiations.* 
From light elements these newly-found radiations are of only moderate penetrating- 
power, but are much “harder” (of shorter wave-length) than the K radiations 
emitted by these same elements ; they are of the same order of wave-length as 
the K radiations from the heavier elements—silver, tin, &c. The irregularity was 
found in the relation between the energy of K corpuscular radiation and the 
K absorption of the primary beam (p. 331). Thus it appeared that the fraction 
K corpuscular radiation (energy) • , 
— 77 —- 7 - 7 - - -increased with the diminution ot wave-length 01 
K primary absorption (energy) 
the primary radiation only up to a certain wave-length ; beyond this there was a 
sudden diminution in its value. This result, however, was based on the assumption 
that the ionization in air was produced by electrons of a definite number of groups, 
K, L, M, &c., throughout the whole range of wave-lengths employed : in other words, 
it was assumed that within this range of wave-lengths there was no radiation 
characteristic of the elements nitrogen and oxygen, consequently that no fresh 
group of electrons was emitted as corpuscular radiation as the wave-length of the 
primary radiation was reduced. 
Thus it will be observed that the values given in column IX. of Table III. 
ionization coefficient in air 
depend upon the ratio 
But the value of the absorption 
absorption coefficient in air 
coefficient in air for the radiations of shorter wave-length were obtained by simple 
proportion from those in aluminium after the subtraction of the scattering coefficient, 
* There has been some evidence of this for many years. The writer in 1903 found that unless the 
primary radiation was kept “ soft ” an additional radiation of some kind was emitted by substances of low 
atomic weight. It was also observed that to obtain evidence of the most perfect scattering from light 
elements as shown by polarization experiments “ soft ” radiation had to be used. Again, experiments on the 
distribution of scattered radiation showed that this approximated to that for the fluorescent radiation 
when the primary radiation was not very “ soft.” These and other experiments suggested the possible 
emission of a characteristic radiation of another series. 
3 c 2 
