716 Mr. Horace H. Poole on the 



per centimetre to remove any ions which might possibly 

 have got through the plugs. 



The testing vessels were interchanged to eliminate any 

 difference in the vessels themselves. Their natural leak was 

 found while passing a current of pure air at the same rate 

 as the emanation carrying current. In most of the 

 experiments this was about 20 c.c. per second. 



Preliminary observations with a pair of testing vessels 

 differing slightly from those described, indicated that the 

 difference between the activities of radium emanation at 

 ages of 10 seconds and 5 minutes was certainly small, and 

 apparently zero. It was found necessary to keep the platinum 

 strip at a bright white heat as otherwise, if its temperature 

 were raised, there was a large temporary increase in the 

 ionization currents through the vessels, indicating that 

 accumulated emanation was being evolved. At this high 

 temperature radioactive particles were apparently given off 

 by the strip, and were deposited on the w T alls of the containing 

 glass vessel and on the cotton-wool filter plug. Owing to 

 the use of insufficient filters in the first experiments the 

 testing vessels used became slightly radioactive and were 

 accordingly discarded in favour of those described above. 

 By the use of a sufficiently thick plug the radioactive dust 

 was prevented from reaching the testing vessels, but it of 

 course accumulated on the plug and on the walls of the vessel 

 surrounding the incandescent strip, where it apparently gave 

 off its emanation irregularly, thus causing variations in the 

 ionization currents through the testing vessels and so limiting 

 the accuracy attainable. 



When determining the natural leak of the testing vessels 

 the air current passed through the cotton-wool plug, but not, 

 of course, through the vessel containing the strip, so that the 

 natural activity included that due to emanation from any 

 radium on the plug, but was not affected by radium on the 

 walls of the vessel. 



The ionization currents through the two vessels were 

 measured alternately, the average rate of motion of the 

 electrometer needle during a period of about ten minutes 

 being determined for each vessel in turn. When this had 

 been done two or three times for each vessel, the connexions 

 of the vessels were changed so that the air current traversed 

 the vessels in the reverse order. This was done without 

 stopping the air current flowing over the incandescent strip. 

 Its constancy was tested by timing the revolutions of the 

 gas meter. More readings of the ionization currents were 

 then taken. 



