144 



THE ROYAL SOCIETY OF CANADA 



electroscope E was used, the scattered rays entering and leaving the 

 electroscope through carbon plates equidistant from the radiator. 



A^, A,, A3 represent three positions of the active material. The 

 electroscope was screened as before. The absorption of primary rays in 

 the radiator was about the same in each case. The readings obtained are: 



Ai, angle of scattering 35°, scattered radiation =1'40 

 A,. " " " 90°. " " =0-61 



130°, 



= 0-63 



We should correct for absorption of the scattered radiation in the 

 radiator and in the carbon plate in front of the electroscope. Correcting 

 for absorption in the latter plate the numbers become 1 -77, 0-82, 0-99. 



®A, 



® A. 



A., 



Figure 2 



It thus appears^_;as if less radiation is scattered through 90° than 

 through other angles and as if very little more radiation was scattered 

 in the emergent direction than in the incident. The similarity to the scat- 

 tering of X rays will be noted and although the experiments are only 

 approximate we can say that the s(!attering of a beam of 7 rays is not 

 at all similar to that of a pencil of a or j8 rays. There is liittle doubt that 

 the scattering of X^ and 7 rays is the same in character there being 

 probably gradual changes in the quality and distribution of the scattered 

 radiation as the X or 7 radiation increases in penetrating power. 

 With the different types of 7 rays at our disposal there is probably 

 very little change in the distribution of the scattered radiation as the 

 penetrating power varies, although the softer the rays the more they 

 are scattered per unit mass. The primar}' rays in the present case 

 are always heterogeneous and if we suppose there is no change in quality 

 produced. by scattering and neglect absorption in the radiator, the 

 scattered rays Avould be softer on the whole but there would lie no change 



