736 Prof. Bragg and Mr. Kleeman on the 
Fig. 7.—Curves showing the two lowest corners : thick layer of 
radium bromide. 
4-5 
\ i } 15 fa 
F 
a 
sh ne = 
S oI | | 
H }a6 
3 NS) 2 
LEAK OF ELECTROMETER 1N MM. DIVISIONS OF SCALE PEP SECOND. 
DISTANCE 4N CM. FROM RADIUM TO SON/SATION CHAMBER. 
Becquerel isolated a very narrow sheet of rays by placing 
the radium in a narrow groove, and confining the rays by 
means of a fine slit. Suppose P 
to be one ray, that is to say the 
path of one set of particles of uni- 
form speed, coming from the top 
of the radium. Then Q will re- 
present the path of a similar set 
from a somewhat lower stratum, 
R a set from still lower, and so on. 
These paths are exactly alike in 
form except that they are shorter. 
If a wire were bent to the shape 
of PBA, and pushed downwards 
in the direction of its length, and kept always in contact 
with B and A, then PQR would be successive positions. 
Now the path of any one particle is more curved towards 
the end of its course, because its velocity diminishes. In the 
figure ABP is intended to be more curved at the upper end. 
But for this very reason the locus PQR is more curved at 
the lower end, as can be easily realized by means of the 
VO pH 
