The Ionization Curves of Radium. 127 
the ionization currents, then the curve of radioactive material 
of sufficient depth is simply a straight line such as AB. 
The « rays are supposed to be all of 
one kind, i. e. of one initial velocity 
on their expulsion from the parent 
atom, though not, of course, of uni- 
Cai. velocity ates they emerge into P 
the space above the radium. But if 
the layer is thin, then the ionization 
curve must take some such form as 
APM, where PM is the range which 
the a particles from the lowest layer 
possess when they emerge. If ¢ is 
the thickness of the layer and p the 
ratio of its density to that of air, then 
pt is equal to the difference between 
AOand PM. Thus part of the ionization curve is vertical. 
Since this effect was found in the ease of the irregular 
fragment of glass tube on which radium was deposited, it 
seemed to us probable that layers prepared with greater care 
would furnish valuable information. This we have found to 
be the case. The curves are curious and interesting. They 
fully bear out, so far as we have been able to judge, the 
hy pothesis advanced in the paper already cited. They afford 
a novel method of examining the phenomena of radioactive 
change, and a confirmation of the theories and results de- 
scribed by Rutherford and Soddy. They have also brought to 
light a tact which we believe to have been hitherto unobserved. 
It is, that the & particle isa more efficient ionizer towards 
the extreme end of its course. 
As has been already stated, 
the curve of a thin layer emitting 
one setof a rays must be of the 
form ABLO. But if there are 
four sets, as Rutherford and 
Soddy have shown, there should 
be four such figures, and the 
ionization curve should be of the 
form ABCDEFGH. Assuming 
the lowest part of the curve to 
represent the « rays emitted at 
the first change, as was shown 
to be probable in the paper cited 
above, then the other three ought 
to be nearly equal in width. For 
they represent the emanation and two induced activities, the 
latter two following so quickly on the emanation that, unless 
B 
Bo Ei aa (N P 
