4 
on a a 
+ 
Absorption of a Rays. 725 
radium should be of simpler character. This was found to be 
the case. The results are plotted in curve D,, and reduced, 
as regards abscisse, to curve Dj. The cone used was in this 
ease rather wide. It is clear that the most energetic « 
particles are almost, perhaps entirely, absent, and the first 
breakdown of the radium atom is responsible for the @ 
particle of perhaps the least range of the four. 
A thin layer traversed by the particle should reduce the 
ranges of all of them by the same distance. This effect is 
shown in curve ©, which shows the result of interposing a 
thin film of goldbeaters’ skin. With this exception the 
arrangements were the same as in the case of curve A. 
The effect is simply to reduce all the ordinates by the same 
quantity. 
It should be added that the @ and y rays were not 
eliminated, but were found by frequent trial to be of small 
and practically constant effect at all ranges. Also the 
meshes of the gauze did not interfere, fer when the gauze 
was hung by long silk cords and set swinging in its own plane, 
the general effects were exactly the same. 
It thus appears that there are several classes, perhaps 
four, of « rays, which may be distinguished from each other 
by their difference in initial energy. The slowest are 
probably due to the first act of disintegration, and this is in 
accordance with Rutherford’s experiments. Also the results 
go to show that the e particles are never deflected, but are 
‘“‘absorbed”’ only because they spend their energy on 
icnization. 
Finally, we may calculate the ionization that should on 
this hypothesis be produced in a chamber such as Rutherford 
used. If the film of radium be supposed very thin; if an 
obliquity factor cos 6 be introduced, and if the chamber be 
deep enough to absorb all the rays, the ionization should be 
approximately :— 
*cos ws 
{ “ 2a sin 6 cos 0 (a—pd sec 0)d@=x(a—pd)?/a, 
- JO 
where d is the thickness of metal traversed, and p is the 
ratio of the densities of metal and air. Thus the curve 
for simple substances like uranium and polonium should be 
parabolic with respect to d, and this is nearly the case. For 
radium the curve should be much more complicated, and 
might well approach the exponential form. 
