734 Prof. Bragg and Mr. Kleeman on the 
perceptible. It is also conceivable that the particle, as its 
speed approaches the critical value below which it loses the 
power of penetration, may leave its rectilinear path and be 
buffeted about, causing a considerable amount of ionization 
without getting much Marthe away from its source. _ 
The experiments here described show, therefore, that the 
a particles from the first change have a range of about 
3°5 cms. in air at ordinary pressures and temperatures: those 
from the second induced activity have a range of about 6°7 cms. 
in similar circumstances, whilst the rays from the other two 
changes have ranges of about 4°5 cms. and 4-0 cms. respectively. 
These values may, however, be modified by the state of the 
air, by the humidity ‘of the radium, and by the distribution of 
the radioactive products in the radium. 
The emanation cannot be uniformly distributed through the 
radium ; it must be somewhat less at the surface than else- 
where, for it is always leaking into the atmosphere. The 
substances of “induced activity ” are formed from the ema- 
nation, and their distribution must be affected by the distri- 
bution of the emanation. 
To this cause we are inclined to attribute small irregu- 
larities in the curves which often appear at the heads of the 
various sections; and which are not experimental errors 
because the same effects are sometimes found unchanged on 
going over the curve again. 
In the paper referred to previously, a curve was given 
showing the ionization of a thick layer of radium, in radio- 
active equilibrium, at various distances from the source. 
Since the radium was not a thin layer, the curve ought to bea 
polygon with four corners. 
The curve showed two well-marked corners and traces of 
probably two others. 
We have repeated this experiment, using spherical ioni- 
zation chambers. Ifa cone of rays from a smnall quantity of 
radium R is isolated by a suitable 
stop S, as in the figure, the upper 
boundary of the radiation is a por- 
tion of a spherical surface. IJfan 
ionization chamber of about the 
same radius is used, it is plain that 
the entry of the rays into the 
chamber and the consequent ioni- 
zation within it will be mores 
sudden and more marked than if 
the chamber is made of parallel 
plane sheets. 
R 
With a chamber of 6 cms. radius we obtained the curve 
