Recoil Atoms in Ionized Air. 725 



strength of the emanation — which served as a measure of the 

 ionization — is taken as 100 for the first experiment, and 

 the subsequent values were calculated from the decay of the 

 emanation. These values were, however, checked occasionally 

 by comparing the 7-radiation with that from a radium standard. 

 The number of actinium I) atoms on the plate, as measured by 

 the (3 radiation from the atoms disintegrating, was taken as 

 100 when the a-ray tube contained no emanation. 



It will be noticed that as the ionization decreases the 

 relative number of actinium D atoms drawn to the negative 

 plate through the ionized air increases. 



When no external ionizing agent was used, the a-particles 

 from the actinium active deposit (about 2 x 10 6 per sec.) 

 produce some ionization. Besides this ionization there is also 

 a considerable ionization produced by the recoil atoms 

 themselves *". When this ionization was increased by means 

 of a weak polonium plate (10 5 a-particles per sec), no per- 

 ceptible decrease in the amount of actinium D collected was 

 noticed. The 10 millicnries of radium emanation produced 

 about 10 14 ions per second in the air between the plates. 



When the potential difference was changed (the air 

 being stronglv ionized), an increase was accompanied by an 

 increase in the amount of actinium D collected, while a 

 decrease produced a decrease in the amount of actinium D. 

 It is quite evident that with a very high potential difference 

 the time for the ions and also for the recoil atoms to cross 

 the field is small and, no doubt, the path more rectilinear. 

 Hence the chance of recombination is smaller. 



Changes of Potential Difference and of Distance 

 between the Plates. 



Experiments were also carried out by changing the 

 potential difference between the plates keeping the distance 

 constant, and by changing the distance keeping the potential 

 difference constant. In these experiments no external 

 ionizing agent was employed. The results are given in 

 Table II. (p. 726). 



The results obtained are in good accord with the view that 

 the recoil atoms lose their charge by recombination like 

 positive ions. With no potential difference the actinium D 

 collected is no doubt due to diffusion. When the distance 

 between the plates was 12*3 mm., only a trace of actinium D 



* Wertenstein, Le Radium, ix. p. 6 (1912). 

 Phil. Mag. S. 6. Vol. 24. No. 143. Nov. 1912. 3 B 



