40 



surements of p for various substances; so also has H. W. 

 Schmidt; and the two sets of results do not agree very well. 

 This is not to be wondered at, for the quality of the secon- 

 dary radiation due to ^ rays is different to that of the inci- 

 dent, and depends on the nature of the reflector. For in- 

 stance, there is some very soft radiation, of which half is 

 absorbed in 25 cm. of air. The experimental value of p will 

 therefore depend on how close the ionization-chamber which 

 receives the secondary radiation is placed to the reflecting- 

 surface. It will be largest if the surface is actually within 

 the chamber ; and this is effectively the case in these present 

 experiments. 



Again, the incidence radiation should be somewhat less 

 than p times the emergence radiation ; and the words itali- 

 cized should apply specially to the case of the lighter atoms. 

 Referring back to the argument of our previous paper, it was 

 there pointed out that the cathode radiation, which originated 

 in any layer of the plate and was turned back, was scattered 

 and softened in the process. In determinations of p this effect 

 causes the result to be too high, because slow )3 rays produce 

 more ions per cm. than fast ones. At the same time, such rays 

 contribute less to the incidence radiation than they would 

 have done if they had possessed the same penetrative powers 

 as the ^ rays which go on and emerge from the plate. This 

 applies particularly to the lighter atoms, for it is in their case 

 that this scattering and softening effect is so pronounced. 

 (Quality of secondary radiation due to j8 rays. Trans. Roy. 

 Soc. of S.A., Oct., 1907, p. 300.) This seems a reasonable 

 explanation of the very large differences between the emer- 

 gence and incidence radiations of the lighter atoms. 



It is interesting to observe that the figures for the inci- 

 dence radiations were obtained by the use of a stream of y 

 rays fairly free from ^ rays. So far as we can discover, this 

 condition has not been realized previously. When we have 

 successfully applied the corrections described above, the results 

 should be of considerable interest. 



Having carried these experiments sufficiently far to show 

 clearly the want of symmetry between the secondary radia- 

 tions on the two sides of the plate, we put them aside for 

 the time, in order to compare the emergence radiations of 

 plates of different substances. At the beginning of this paper 

 it was stated that emergence radiations would be the sani<- 

 for all substances, if we could arrange to have (1) homoge- 

 neity of the y rays; (2) a density law of absorption for the 

 y rays; (3) a density lav; of absorption for the jS rays. All 

 these conditions cannot be realized ; but we can go a long 

 way towards satisfying the first two. Wigger has shown that 



