94 



ROYAL SOCIETY OF CANADA 



which gradually disappeared, but which in doing so brought irregular- 

 ities into the results and often vitiated them. 



In taking the readings given below, exceptional care was therefore 

 taken to make certain that all the insulating supports were entirely free 

 from electrification. The electrometer was also exceptionally well set up, 

 the needle being so carefully adjusted and the suspension wire kept so 

 free from torsional strain that when both pairs of quadrants were joined 

 to earth no displacement from the zero took place for hours. 



With the apparatus in this condition the test was made for a charg- 

 ing effect with the potassium chloride, first when the needle of the elec- 



TABLE V. 



Time in Min. 



Deflection in Miû. on Scale. 



trometer was positively charged to 240 volts, and secondly when it was 

 negatively charged to the same voltage. The sensibility of the electro- 

 meter was such that a potential difference of 1 volt between the quadrants 

 gave a deflection of 600 mm. divisions on the scale. In making the test 

 the pressure was lowered to .003 mms., and maintained at this low pres- 

 sure throughout the measurements. 



The results obtained with the needle positively charged and also 

 when it was negatively charged are given in Tables V and VI, and curves 



