134 ROYAL SOCIETY OF CANADA 



To accomplish this the electroscope was taken to the attic of the 

 building and set up beneath a small opening in the roof. The receiver 

 was placed over this opening on insulating supports on the roof and the 

 electrode which it carried was joined to that of the leaf system by the 

 insertion of a light metal rod 1 2-3 metres in length, A guard tube 

 was also added to protect this added length to the electrode. 



In setting up the apparatus this time a lining of thick sheet 

 zinc was inserted in the receiver which had been carefully abraded 

 with fine glass paper and then thoroughly washed in turn with dilute 

 hydrochloric acid, ammonia, methyl alcohol, and distilled water. 

 Fresh filtered air was drawn into the receiver and it was then her- 

 metically sealed. Readings were taken immediately afterwards. In 

 this case also it was noticed that the conductivity of the air 

 gradually decreased and ultimately assumed a steady value. In the 

 initial measurements the conductivity corresponded to as high as 19 

 ions per cc. per second, but in the steady state it approximated to 

 the generation of only 11.0 ions per cc. per second. 



This low value illustrates the great differences which exist in the 

 radioactivity of different metals. With the galvanised iron cylinder 

 alone it will be remembered that the conductivity corresponded to the 

 production of about 35 ions per cc. per second. The insertion of the 

 zinc lining it will be seen therefore cut off by far the greater portion of 

 the radiation from the galvanised iron. The reduction in the con- 

 ductivity of the enclosed air necessarily made that portion contributed 

 by the penetrating radiation of greater relative importance and it was 

 thought that a more severe test would thus be available for the detec- 

 tion of diurnal variation. It was found, however, that the insertion of 

 the long electrode greatly reduced the sensibility of the electroscope. 

 By the addition of this the capacity of the instrument was nearly 

 trebled, and consequently although the displacement of the gold leaf for 

 a given movement of the condenser tube could be increased by adding 

 to the voltage of the latter still this device had the defect of adding to 

 the time of taking a reading, and as it was desirable to have the read- 

 ings taken as frequently as possible it was found necessary therefore to 

 apply only a moderate voltage to the condenser. For this purpose a 

 potential of 21 volts was maintained on the condenser tube, and this 

 permitted readings to be taken, though with diminished accuracy, 

 every 15 or 20 minutes. 



