80 THE ROYAL SOCIETY OF CANADA 
(2) Electroscope errors; including— 
(a) Faulty leaf. ‘ 
(b) Not continuously air-tight. 
(e) Variation in the potential. 
(d) Change in position of rest of leaf or rod, affecting the 
electrical capacity. 
(e) Not charging to the same potential. 
(3) Microscope errors. 
(4) Errors in calculation. 
(5) Heating effects. 
(6) Variation of air-pressure inside the electroscope. 
(7) Errors due to faulty stop-watch. 
(1) (a) The Prodecure in this case was to boil vigorously until 
the last bubble passed over. About one or two c.c. still remained 
in the bend of the delivery-tube. This was driven over by removing 
the flame for a moment until the flask slightly cooled, and the condens- 
ed water was allowed to follow the emanation up the delivery-tube, 
just beyond the rubber tubing leading to the flask. This ensured 
that the emanation did not escape into the room when the rubber 
tubing was removed for aninstant to allow the air torush in and dilute 
up the remaining emanation. It was now re-boiled for two or three 
minutes, after which the diluted portion remaining was negligible. 
A slight error was possible here in the earlier part of the work, owing 
to the bend of the delivery-tube being not rounded enough and also 
by not allowing the water to pass beyond the rubber tubing in re- 
boiling. 
(b) All vessels containing the emanation, and the electroscope, 
were sealed with tap-grease and vacuum-wax, and the stoppers and 
rubber connections wired. Some very low readings, however, were 
traced to the rubber tubing having deteriorated where it was in 
contact with the pinchcock. This accounted for very divergent 
readings and occurred in only two or three instances. 
(2) (a) The leaf was observed over the scale used, and did not 
show any sign of being faulty, such as giving a “kick” at a certain 
position. It was distinct at the edges. 
(b) The electroscope was tested to see if it was air-tight before 
each set of observations by means of a pressure-gauge. It never 
showed the least sign of leakage after it had once been made air-tight. 
(c) The potential could be varied by not having the inside of 
the electroscope earthed. This was carefully tested repeatedly 
and the earth-connections found to be working perfectly. Tinfoil 
from the inside of the electroscope was in direct connection with 
the earth-wire. 
