[mclennan-keys] 



ELECTRICAL CONDUCTIVITY 



43 



The wires EA and FB were made about 60 cms long so that when the 

 lower ends were immersed in liquid air their insulating supports carried 

 by the rods G and H were sufficiently far removed to prevent moistm-e 

 depositing on them through their being cooled below the dew point and 

 so spoiling the insulation at G and H. 



ISi 



Fi'ïi re 1 



The electrode EA was connected to a battery of small storage 

 cells and the electrode FB to the free quadrants of an electrometer. 

 Before being used the electrodes and their supporting wires were all 

 carefully cleaned. 



In carrying out these experiments the liquid air was always care- 

 fully filtered and the measurements were made as rapidly as possible. 

 This was done because even when the case was supplied with 

 large quantities of P^Oj and concentrated sulphuric acid moisture 

 was always deposited on the glass and quartz tubes surrounding 

 the wires at a point a short distance above the mouth of the Dewar 

 flask, and whenever this deposit had gained a considerable thickness 

 it was found that small pieces of ice and snow always fell from it 

 into the flask and loweied considerably the insulating power of the 

 liquid air. 



In using the apparatus a set of observations was taken immediately 

 after the filtered liquid air was poured into the flask with a series of 

 increasing potentials applied to the electrode EA. In every case it 



