fsATTEm.Y-ELWORTHY] CANADIAN MINERAL SPRINGS 19 



Description of the Apparatus. 



Fig. 1 shows the ionisation vessel and the electroscope. The 

 former consists of a stout brass cylinder C of height 25| cms. and 

 diameter 14 cms. Its volume is 3,930 cubic centimetres. The 

 electrode A is a brass rod of 4 mms. diameter. This projects into the 

 cylinder through an insulation consisting of a quartz tube held in 

 place by a hard rubber collar 1 -4 cms. in diameter. The joint is 

 air tight. A reaches to within 2 cms. of the bottom of C. The gold 

 leaf L, about 35 mms. long and 1| mm. wide, is attached to the upper 

 end of the rod A. A cubical brass box S of about 6 cm. side surrounds 

 the gold leaf. Two small windows in opposite sides of S enable the 

 leaf to be illuminated and observed. The rate of fall of the leaf is 

 measured by means of a telemicroscope mounted horizontally in front 

 of the windows. The microscope used has a magnification of about 

 40. It is furnished with an eye piece scale and all leaks are expressed 

 in terms of the divisions of this scale. The vessel C is provided with 

 three openings to which are attached short pieces of thick rubber 

 tubing. One of these is connected to a mercury manometer G, 

 another O to a pump by means of which C can be exhausted and the 

 third, I, serves to admit the gas after it has passed through suitable 

 drying tubes. 



The vessel is firmly mounted on a wooden stand, uprights from 

 the stand gripping the vessel at the projecting tubes. The stand 

 also carries the support of the telemiscroscope. The gold leaf was 

 always charged to practically the same potential. The position 

 given by the zero reading on the microscope scale corresponded to 

 350 volts and the 100 reading on the scale to 200 volts. The rate 

 of leak was observed over this interval. The voltage is always such 

 as to ensure saturation for the small amounts of radioactive matter 

 dealt with, and the movement of the leaf was found to be practically 

 uniform over the range employed. The normal "air" leak i.e., the 

 leak when the ionisation vessel was filled with dry air was 0-02 scale 

 divisions per minute. The normal air leak is usually subtracted 

 from all other leaks in order to get the leak due to the radium emana- 

 tion contained in the air. 



Measurement of the Radium Emanation Dissolved in Water. 



To get the radium emanation out of a sample of water, nine litres 

 of the water was placed in a tinned copper boiler P (Fig. 2) connected 

 by a large rubber bung to the inner tube M, also of tin, of a condenser 

 C sloping upwards as in the figure so that the condensed steam might 

 run back into the boiler and thereby keep the strength of the sample 



