Prof. Owens on Thorium Radiation. 365 



was found to be practically the same as when air was drawn 

 directly from the room. 



When a large tube of P 2 5 was placed before the inlet 

 pipe, the pump being in operation, the current increased to 

 36. When a plug of glass wool was added, the value of the 

 current became 37. 



We thus see that scarcely any difference is noticeable 

 between moist air, dry air, and air free from dust. Indeed, 

 the small increase of current actually observed may easily be 

 accounted for by the diminished pressure and velocity of the 

 gas in the vessel occasioned by placing the drying-mixture 

 and wool plug before the inlet tube. The oxide was then 

 covered with a layer of common writing-paper and the con- 

 duction-current measured, with and without the pump in 

 operation ; in the former case the current was about one third 

 of its value in the latter. 



Layers of thin aluminium foil were then substituted for the 

 paper ; with one layer of foil the current with pump on 

 was 58 per cent, of current when air in vessel was at rest, 

 with two layers 69 per cent., and with three layers 75 per 

 cent. 



The effect is thus seen to be diminished as the surface of 

 the oxide is better protected from air-currents. The pro- 

 tection, however, was probably only partial, as both the paper 

 and foil were no doubt more or less porous. 



On covering the oxide with a mica sheet the actual conduc- 

 tion-current was greatly diminished, but remained practically 

 constant whether the air in vessel was at rest or in motion. 



Using a thin layer of oxide unprotected with paper or foil, 

 the current with pump on was 80 per cent, of the current 

 when the air was at rest, indicating that the radiations from 

 thin layers are much less susceptible to air-currents than those 

 from thick ones, or at least that the effect is closely related 

 to the thickness of the layer. Using a thick layer of the sul- 

 phate the diminution of current when the pump was operated 

 was about 15 per cent., and with the nitrate practically the 

 same, both much less than with the oxide. 



Substituting uranium oxide for the thorium salts no dimi- 

 nution in the conduction-current was observable when the 

 pump was in action. 



With thorium salts the same general effects of air-currents 

 were observed as different voltages were applied to the lower 

 plate, as also was the case when the distance between the plates 

 was varied. When oxygen and coal-gas were passed through 

 the apparatus instead of air, a similar diminution of the con- 

 duction-current when the gases were in motion was observed* 



