Air in Closed Vessels and its Causes. 563 



Haying arrived at the conclusion that the ionization in a 

 closed vessel is partly doe to a penetrating radiation from 

 without, there remains the further question of the part, if 

 any. played by the walls. The question is one of the very 

 greatest interest and importance from its bearing on the 

 question of the radioactivity of ordinary materials. 



( '. T. 1\. AVil>on"* measured the ionization of various gases 

 in small sealed ves>el>. and found that, except in the case of 

 hydrogen, the ionization is approximately proportional to the 

 density. This is what would happen if the radiation were a 

 penetrating one from outside, or one from the walls, of 

 sufficient penetrating power to pass from side to side of the 

 vessel without complete absorption. Patterson found that in 

 a large iron vessel the ionization varied directly as the 

 pressure for pre-sures below 80 nuns., while it was practically 

 independent of the pressure for pressures greater than 

 300 mms. If the whole of the ionization in a closed vessel 

 is due to a very penetrating radiation either from outside or 

 from the walls, this result would be difficult to explain : but 

 if we suppose that such radiation is responsible only for a 

 part of the ionization, and that the remainder is due to a 

 radiation from the walls of comparatively feeble penetration, 

 the results can be very simply explained. For pressures 

 greater than 300 mms. ihe gas is sufficiently dense to absorb 

 all. this feeble radiation, and so no increase of pressure, with 

 eon-equent increase of density, will affect the amount of 

 ionization which this radiation will produce. As the pressure 

 is reduced, however, a point will be reached when the 

 radiation will be able to traverse the vessel without complete 

 absorption, and below this pressure the absorption and, pre- 

 sumably, the ionization produced will be proportional to the 

 pressure. The conclusion that the ionization is partly due to 

 a radiation from the walls was further supported by some 

 experiments of Struttf. He measured the ionization in a 

 vessel lined with different materials, and found that it varied 

 considerably, being large for lead and tin, and small for zinc 

 and aluminium. This pointed to the existence in each case 

 of a radiation peculiar to the material. 



The experiments next undertaken had a twofold object : — 

 (1) To ascertain whether samples of the same metal- as those 

 used by Strutt would give value- of the ionization similar to 

 tho>e obtained by him ; and | '1 ) to ascertain if possible whether 

 the radiation from the walls of the vessel was primary or 

 secondary in character, whether it was an intrinsic property of 



* Proc. Roy. Soc. lxix. p. -Ill (1901 ). 



t Phil. Mag. June l!>0:}. 



