166 



If, now, a hemispherical dome of Pb 4 mm. thick, with 

 its central portion removed so as not to intercept the origi- 

 nal stream of y rays, be placed in the position shown by h,hy 

 fig. i., and a set of observations be taken as before, sub- 

 tracting the reading with the radiator in from that with it 

 not in position, the curve B is obtained. Again, a maxi- 

 mum is reached for about the same thickness of radiation 

 as before, but the effect is for the different thicknesses of 

 radiator about 60 per cent, of its previous value. We may 

 conclude that it is y radiation which is being intercepted 

 by the dome, as in neither case can any but a very small 

 amount of ^ radiation pass obliquely through the thickness 

 of 1'25 mm. of lead forming the sides of the ionization 

 chamber ; that from the radiator are proceeding second- 

 ary emergence y rays which produce ionization in the test- 

 ing chamber, as shown in previous papers by means of — 



fa) "tertiary emergence" )3 rays as at I; 



(h) tertiary incidence ^ rays as at k; 



(c) ^ radiation produced in the gas of the chamber from 

 the y rays as they pass through it ; 



fcl) possibly some ions produced directly in the gas by 

 the y rays ; and 



(e) subsequent ^ and 7 ray effects of a higher order. 



