Secondary Réntgen Radiation. 549 
It was found that the ratio of the rates of ionization was 
practically the same in each case, showing it to be independent 
of the intensity of primary radiation. 
The increase in the ratio of secondary to primary joniza- 
tion by placing an absorbing plate at P was therefore due to 
the change in the char acter of the primary radiation by 
transmission through the plate. But it was shown to be 
independent ot the position of the plate in the primary beam, 
as the ratio was unaltered when the plate was moved from 
Pto Q. Hence the effect of secondary radiation from alu- 
minium was negligible. 
The conclusion was then that the rays of higher penetrating 
power produced a greater secondary ionization in proportion 
to the primary ionization produced by them than the more 
absorbable rays. This, again, could not have been due simply 
to scattering of these constituents in different proportions, for 
the results of experiments (2) and (3) would then have been 
identical. We thus arrive at the conclusion that there must 
have been a greater transformation* of the penetrating than 
of the more absorbable rays, the increase in ionizing power 
accounting for the proportional increase in the secondary 
ionization. 3 
It should be noticed that experiments (2) and (3) do not 
in any way give a comparison of the absorption of the primary 
and secondary beams. If the primary beam were homo- 
geneous in character, 2. e. consisted of rays of one penetrating 
power, then the ratio of the secondary ionizations in the two 
cases would be the ratio of the fractions of primary and 
secondary radiations transmitted by a plate of the thickness 
used. Such a test would, however, be no more delicate than 
that given by experiments (1) and (2). If the different con- 
stituents of the primary beam were transformed equally, then 
again this test would be no more delicate than that given 
by (1) and (2). 
The result depends merely on the difference in the trans- 
formation of the penetrating and of the more absorbable rays. 
If the absorption by aluminium of different radiations 
were proportional to the ionization produced by these radia- 
tions in a given volume of air, then the absorption by 
aluminium would have been easily measured. 
We thus conclude that for the rays experimented upon a 
* The measure of transformation used is the ratio of the ionization 
produced in a given volume of air by the transformed beam to that pro- 
duced by the primary beam if of the same intensity. This is not the 
ratio of the absorption coefficients for the rays by air, as the total ab- 
sorption is not proportional to the ionization, but it is the ratio of what 
may be called the ionizing powers. 
