548 Mr. ~N. K. Campbell on Radioactivity 



this window in use. It is probable that the effects were 

 caused by the indifferent conductivity of the paper. 



§ 6. The window was formed of sheets of tinfoil '0013 Cm. 

 thick, equivalent in surface-density to '0035 cm. aluminium 

 or 7*35 air. The effect of heat was of the same nature as in 

 case 1, sometimes an increase, sometimes a decrease; but the 

 absence of dust did not seem to diminish the effect. No 

 change occurred while cooling down after heating. 

 Actions of Class A gave no effect. 



Actions of Class B gave a small effect of the same sign as 

 the heat effect. 



Actions of Class C, i.e. actions evolving gases (Zn + HCl, 

 Cu + HN0 3 , Fe + H 2 S0 4 , CaC0 3 + HCl) gave a slight in- 

 crease occurring about a minute after the action began. It 

 is only reasonable to suppose that this was due to the dif- 

 fusion of the ionized gases evolved through the small holes, 

 which could not be avoided, in the tinfoil window. 



§ 7. Window of aluminium-foil '00053 cm. thick sup- 

 ported on wire netting. 



The effect of heat was of the same nature as in case 3, 

 sometimes an increase, sometimes a decrease ; but fortunately 

 the cases in which the heat decreased the leak were far the 

 most numerous. Only in these cases were the effects of 

 chemical action tried. 



Actions of Class A gave no effect. 

 Actions of Class B gave a small decrease. 

 Actions of Class C gave an increase beginning some time 

 after the action, and greatly diminished by blowing a rapid 

 current of air over the surface of the reagents. It was also 

 found in this and in every case that the increase caused by a 

 chemical action was in no way diminished by placing over 

 the dish in which the action was proceeding a sheet of zinc 

 "08 cm. thick. If the effect had been due to rays emitted by 

 the reacting substances, it is impossible that these rays should 

 not have been diminished in intensity by absorption in the 

 zinc. If the increase were due to heat or to the evolution 

 of ionized gases, the zinc sheet might be expected to have 

 little influence. 



§ 8. We thus find :— 



(1) That chemical actions which evolve no heat cause no 

 increase in the leak. 



(2) That chemical actions which evolve heat sometimes 

 cause an increase and sometimes a decrease; the sign of the 

 change being the same as that of the change due to heat. 



(3) That chemical actions which evolve gases cause an 

 increase, but that increase is more readily explained by the 

 ionization of the gases than by a process of ray-emission. 



