114 THE ROYAL SOCIETY OF CANADA 



tenable. For in some experiments by McLennan and McLeod^ it 

 has been shewn when working with clean air contained in an air-tight 

 Wolff Electrometer of zinc placed in a water-tight box of aluminium- 

 bronze that a value of 4-81 ions per c.c. per second was obtained for 

 the conductivity of the air at the surface of Lake Ontario, while a 

 value of 4-77 ions per c.c. per second was obtained when the box con- 

 taining the electrometer was immersed to a depth of 8 metres, this 

 would go to shew that the conductivity of the air was practically 

 the same in the two positions, which would indicate that air and water 

 had the same coefîficient of absorption for this hypothetical extremely 

 penetrating radiation. Since from two to three metres^ of water 

 will entirely cut off all known gamma radiations emitted by radium, 

 the penetrating powers of this radiation would have to be so extra- 

 ordinarily great, that in the absence of any collateral corroborative 

 evidence one hesitates to believe in the reality of its existence. In 

 considering the hypothesis of ionisation by collisions due to thermal 

 agitation Langevin and Rey^ have shewn from theoretical considera- 

 tions on the basis of exceptional collisions that the ionisation in air 

 should increase very rapidly with a rise in temperature. Patterson*, 

 however, has shewn experimentally that the conductivity of air re- 

 mains practically the same from 0°C. to 180°C. Again on the basis 

 of tangential collisions Wolfke^ has shewn that if the conductivity 

 of air be taken to be 4 ions per c.c. per second, at room temperatures 

 it should be represented by 2 ions per c.c. per second at 130°C. and by 

 about 6 ions per c.c. per second at — 20°C. As this result is also in 

 contradiction to the measurements of Patterson, it would , appear 

 that the residual ionisation in air cannot be explained on any theory 

 of collisions as yet brought forward. There remains, therefore, the 

 possibility of the ionisation being brought about by a spontaneous 

 breaking up of the atoms or through the agency of a radiation from 

 the walls of the containing vessel. 



With a view to examining the validity of these two hypotheses 

 some measurements were made on the residual ionisation in a number 

 of gases including air and the following paper contains an account of 

 these experiments. The investigation was begun by one of us in the 

 summer of 1913 (see Nature, Dec. 11, 1913), but had to be discon- 

 tinued through an accident to the electrometer. The apparatus hav- 

 ing been repaired the investigation was resumed last autumn and from 



1 McLennan and McLeod. Phil. Mag. Oct., 1913, p. 749. 



2 McLennan. Phys, Rev. Vol. XXVI, No. 6. June, 1908, p. 530. 

 'Langevin and Rey. Le Radium. April, 1913, p. 142. 



* Patterson. Phil. Mag. 6. p, 231, 1903. 

 ^Wolfke. Le Radium. Aug. 1913, p. 265. 



