730 Prof. Bragg and Mr. Kleeman on the 
in diameter and 2 cms. high. The readings A, B, and C 
represent values of the ionization when a piece of paper was 
placed over the radium so as to cut offall a rays. The most 
part of the leak when this was done was due to natural leak 
of the electrometer. This may have been due to very small 
amounts of induced radioactivity deposited by emanation within 
the apparatus ; for when we carefully excluded all radioactive 
material from the working room, and carried out all prepa- 
rations elsewhere, the leak was greatly diminished. Much of 
it was of course dueto B andyrays. It will be seen from the 
set of readings given later and from the curves drawn in the 
figures that the leak was very constant during any one set of 
experiments ; and it may be considered as certain that all to 
the right of the straight line drawn in the figures represents 
the ionization due to @ rays. 
Fig. 3.—Ionization curves of dish prepared (I.) 20 hours, (I1.) 90 hours, 
(III.) 140 hours before. 
rd 
Oo) 
Oo 
p 
W 
DISTANCE tN CM. FROM RADIUM 70 1ONISATION CHAMBER. 
LEAK OF ELECTROMETER 1 N MM. DIVISIONS 
OF SCALE PEP SECOND 
It is plain that the curve PQS does very fairly approxi- 
mate to the theoretical form. The gradual approach to this 
form is shown in fig. 3, where I., IL., III. show the ionization 
of dishes at various times after preparation. In this case 
the portion P is not fairly represented because we employed 
| 
| 
: 
