On the Ionization of Atmospheric Air. 



159 



Experiments were now made on the variation of the rate of leak 

 with pressure. The measurements were made at a temperature of 

 about 15° C. Each experiment gave the leakage in a period varying 

 from six and a half to twenty-four hours. The silvered glass apparatus 

 was used. 



The following results were obtained : — 



Pressure in 



Leakage in 



Leakage 



millimetres. 



volts per hour. 



pressure. 



43 



0-22 



0-0052 



89 



0-53 



0-0058 



220 



1 -14 



0-0052 



341 



1-59 



0-0047 



533 



2-30 



0-0043 



619 



2.40 



0-0039 



635 



2-65 



0-0042 



731 



2-78 



0-0038 



743 



2-99 



0-0040 



These numbers show that the leakage is approximately proportional 

 to the pressure. While the pressure is varied from 43 mm. to 743 mm., 

 the ratio of leakage to pressure only varies between 0*0038 and 0'0058. 

 Since the individual measurements of the leakage at a given pressure 

 differed among themselves by as much as 10 per cent., it would hardly 

 be safe until more accurate experiments have been performed to 

 base any conclusions on the apparent departure from exact propor- 

 tionality between leakage and pressure. From these results one would 

 infer that it should be impossible to detect any leakage through air 

 at really low pressures. This is in agreement with the observations 

 of Crookes,* who found that a pair of gold leaves could maintain their 

 charge for months in a high vacuum. 



Experiments were now carried out to test whether the continuous 

 production of ions in dust-free air could be explained as being due to 

 radiation from sources outside our atmosphere, possibly radiation like 

 Rontgen rays or like cathode rays, but of enormously greater penetra- 

 ting power. The experiments consisted in first observing the rate 

 of leakage through the air in a closed vessel as before, the apparatus 

 being then taken into an underground tunnel and the observations 

 repeated there. If the ionisation were due to such a cause, we should 

 expect to observe a smaller leakage underground on account of absorp- 

 tion of the rays by the rocks above the tunnel. 



For these experiments a portable apparatus had to be made (shown 

 in fig. 2). It differed from that already described (fig. 1) in the 

 following respects : — The vessel, of thinly silvered glass as before, was 

 inverted and attached directly to the sulphur condenser, its neck 

 * ' Roy. So?. Proc.,' vol. 23, p. 347, 1879. 



