Excited Activity of Thorium. 635 



with the formation or! a volatile compound, is still possible. 

 This was first tested by replacing the air in the discharge- 

 tube by hydrogen : the loss of activity was again from 30 to 

 55 per cent. A tube of a different shape was then used 

 (tig. 2), The discbarge passed between the terminals A and 



Fiff. 2. 



To ^ump 



B, bo that the active disk D was out of the direct line of dis- 

 charge, while the connecting tube E was wide enough to 

 allow free intermixture of gases between the lower branch 

 and D : with this arrangement there was no loss of activity. 

 When C was made anode, A or B being cathode and D in- 

 sulated, about 10 per cent, of the activity was lost; this may 

 easily be due to the anode acting temporarily as cathode, for 

 a considerable amount of reversal always occurs with a coil- 

 di-eharge. To test this point, C was made anode and B 

 catbode for the Wimshurst discharge : in this case there was 

 no loss of activity. With D as anode and A cathode, there 

 was about 20 per cent, reduction. This is considerably lesa 

 than the average value with D anode and C cathode, and this 

 (together with the low values for the deflected discharge) 

 - support to the view that some part of the effect is due 

 to a direct action of the rays. It was also observed that 

 \\ hen the discharge took place at a pressure which was too 

 high to give any cathode rays, but would not be likely to 

 alter any chemical actions occurring, there was no loss of 

 activity. All these experiments >how that the action cannot 

 be a chemical one. 



(/) Thar the effect is due to the presence of free ions in 

 _ number- near the active surface is also improbable, in 

 view of these results. This point was further tested by 

 negatively electrifying an active disk close to a Bunsen flame: 

 the flame-gases, though strongly ionized, caused no diminution 

 in the activity. 



