12 ROYAL SOCIETY OF CANADA 



far, no clue has been obtained as to why this action should be so mani- 

 fest in thorium oxide. It appears as if in the pores of the thick layer 

 of thorium oxide some change takes place with time, which increasefs 

 the intensity of the radiation, and if the result of the action is con- 

 tinually removed, the intensity of the radiation is diminished. This 

 would explain why the action is shown chiefly in thick layers, and 

 depends on the current of air. All experiments so far tried show that 

 the action is not due to the removal of the ions between the plates by 

 the air currents, but is rather the result of a change in the condition of 

 the radio-active substance at or near its surface. 



The phenomena of electrical conduction produced by thorium 

 radiation are, as far as the investigations have gone, strictly similar to 

 those produced by Eontgen rays and uranium radiation, and obey the 

 same laws. For high pressures of air, thorium oxide behaves differently 

 from uranium, but the difference is probably closely connected with the 

 action of currents of air on the intensity of the radiation, which we have 

 already considered. A special apparatus was constructed to examine 

 the properties of uranium and thorium radiation at high pressures. 

 The investigation is not yet completed but has led to some interesting 

 results. It was found that the rate of recombination of the ions is very 

 rapid, especially at high pressures. If we take the case of thorium 

 radiation, the intensity is reduced to one-half, after passing through 

 10 mms. of air at atmospheric pressure, and assuming that the absorp- 

 tion is proportional to the pressure the intensity would be reduced to 

 one half in a distance of '5 mms. for a pressure of 20 atmospheres. The 

 ionization is thus very intense near the surface of active substances, and 

 the ions recombine with great rapidity after they are formed, so that 

 very large electromotive forces are required to carry them across to the 

 electrodes before any appreciable amount of recombination has occurred. 



The rapidity of recombination depends largely on the amount of 

 dust or nuclei present in the air through which the ions are passing. 

 A striking experiment to illustrate this action can be readily shown by 

 filling the vessel in which the radio-active substance is placed with 

 tobacco smoke. The rate of leak immediately falls to less than one- 

 tenth of its value. The intensity of the radiation emitted by the sub- 

 stance is unaltered, but a groat number of the ions in their passage 

 between the plates give up their charges to the nuclei in the smoke, 

 and the current is correspondingly diminished. 



