173 



The Zn radiator was 12 mm. thick, the Al 17 mm. 

 Measurements were also made with radiators of other thick- 

 nesses, but the results were not sufficiently accurate to enable 

 one to say whether the quality of the emergence radiation 

 varied with the thickness of radiator. 



The results are plotted in fig. v., from which it will be 

 seen that the radiation from Zn and Al is far from homo- 

 geneous when tested by Pb domes. 



The emergence radiation from Zn and Al appears to 

 consist of two sets of y rays — one of much the same quality 

 as those which are produced from Pb, the other a very much 

 softer bundle. A rough analysis of either the Zn or Al 

 curve, if such is legitimate, would suggest that it may be 

 derived from a hard and soft bundle of y rays, for which the 

 values of A. by Pb are 1'3 and 5 respectively, the effect of 

 the hard bundle being initially almost twice that of the soft. 



It will be seen from the figure that mass for mass Al 

 and Zn domes are not able to bring out so clearly as Pb 

 domes the distinction between the quality of the secondary 

 radiations from Zn and Al as compared with that from Pb. 



This, again, is much the same as is found when one at- 

 tempts to sort out the original hard and soft y rays of Ra 

 by such screens. 



However, the radiation from Pb always appears some- 

 what harder than that from Zn and Al, no matter what 

 screen is used. 



This result has been checked on different occasions with 

 care, and is of importance, as the results obtained by Ivlee- 

 man from the secondary incidence y rays, and the theory of 

 selective absorption founded thereon, and extended to the 

 original y radiation, should apparently apply also to the 

 secondary emergence rays. 



§n. 



To investigate the secondary incidence y rays the ioni- 

 zation chamber, fig. i., was inverted, and the large plates 

 which serveql as radiators were laid directly upon the ioni- 

 zation chamber as at p,2). This enabled the secondary inci- 

 dence rays from the radiator to enter the ionization cham- 

 ber, not only by the sides i,i but also through the 

 top of the chamber, and should more than compensate for 

 any decrease in effect due to the primary y rays falling upon 

 a larger area than they did in the case where the emergence 

 radiation was measured. 



