[gray] rays of radium 143 



(d) Intensity of primary rays, 19-(). Lead L^, 2-94 cms. thick. 



Thickness of lead M. 

 0-0 



Intensity of scattered rays 



Carbon 

 0-56 



These figures show at once that the usual explanation of the 

 scattering of rays cannot be correct. Take table 2. The scattered 

 radiation is reduced to about 12% of its value by a lead plate 3 mms, 

 thick. If these rays were due to the softer rays being scattered through 

 larger angles, we would expect that a lead plate 3 mms. thick placed 

 before the active material would cut down the scattered radiation to 

 the same extent. It takes, however, about 3 cms. of lead to do this. 

 Similar results are to be obtained from Table 1. It is therefore evident 

 that some change must take place in the quality of the y rays when 

 they penetrate matter, either direct or by scattering. The possible 

 production of characteristic radiations is not considered as the quality 

 of the secondary rays is very nearly independent of the nature of the 

 radiator. There seem to be two possible explanations. 



(1) 7 rays become softer the more matter they penetrate, not 

 taking into account any possible effect due to scattering. We would 

 then have a continuous production of soft 7 rays and the former 

 explanation of scattering might hold good. 



(2) When homogeneous 7 rays are scattered there is a change 

 in quality, the scattered rays being softer the greater the angle of 

 scattering. 



With respect to the first possibility there is not evidence of such 

 a transformation. Experiments on characteristic X radiations indicate 

 that the rays are homogenous and suffer no change in direct transmis- 

 sion through matter except by being diminished in intensity. Even 

 admitting this possibility there are other difficulties. In the experi- 

 ments of Madsen and Florance the scattered rays had to pass through 

 a certain amount of lead before entering the electroscope. As lead 

 absorbs the incident rays so much more quickly than the harder emerg- 

 ent rays their results may not give a fair comparison. To test this 

 the following experiment was made. 



Readings were taken of the relative intensities of the radiations 

 scattered from a carbon radiator in directions making angles 35, 90, 

 130° with that of the primary rays. The carbon radiator was the one 

 used previously in the scattering experiments (Table 1). The same 



