548 ‘Mr. C. G. Barkla on Energy of 
that for a primary beam coming from a copper plate, for the 
radiation from metals is more heter ogeneous than the primary 
radiation producing it. The two methods, however, would - 
give different results if want of homogeneity affected them 
to an appreciable extent. 
Both experiments gave within the limits of possible error 
the same absorption for the secondary as for the primary, 
showing that even for these very easily absorbed rays, ihe 
character of the secondary radiation differs little from that 
of the primary producing it. 
The possible error was in this case naturally much gre ae 
than in the previous experiments, and amounted to fully 1 
per cent. of the absorption coefficient. 
Comparisons were then made between the rates of ioniza- 
tion produced in the two electroscopes when neither primary 
nor secondary beam was intercepted by absorbing plates, 
when similar aluminium plates were placed in the primary 
and secondary beams just before the electroscopes, and when 
an aluminium plate was placed in the primary beam at the 
aperture P. As stated previously, the same ratio was given 
(1) when no plate intercepted either beam as (2) when plates 
of equal thickness were placed before each electr oscope. (3) 
When a plate was placed in the primary beam at P, the ratio 
of the rates of ionization in the two electroscopes secondary 
to primary was increased. 
In these experiments the ratio was increased by about 10 
per cent. when an aluminiun plate ‘04 cm. thick was placed 
at P. This plate reduced the ionization in the primary electro- 
scope to about 36 per cent. of its initial value, while the 
jonization produced by the transmitted radiation was reduced 
to about 54 per cent. by a second similar plate. 
Experiments (1) and (3) showed that the rays which got 
through the aluminium produced a greater proportional 
secondary effect than the whole direct primary beam. This 
result must have been due either to a change in the intensity 
or in the character of the primary beam in its passage through 
aluminium. There was a possibility that the intensity of 
secondary radiation was not proportional to that of the 
primary radiation, even though the radiations were of the 
same character; and thus that the diminished intensity 
accounted for the proportional i increase in the ratio of secondary 
to primary ionization. A given bulb was therefore worked 
in very different ways, currents of various strengths being 
passed through, so that in some cases the intensity of radiation 
was four or five times that in others. The differences in 
penetrative powers in the different experiments were: small. 
