Relation between Uranium and Radium. 291 



and after the test, and an increase in the rate of leak of the 

 electroscope looked for. The initial test for actinium was 

 carried out on August 31st, 1906, 17 days from the purifi- 

 cation, but no evidence of the presence of actinium was 

 obtained. Nine subsequent tests have been made at intervals, 

 the last being done on July 1st, 1907, 301 days from the 

 purification. Usually there has been a very slight temporary 

 increase in the leak of the electroscope with the wire after 

 the experiment, compared to the leak before, but it has not 

 yet exceeded in any case the possible error. In all cases of 

 course the measurements have been continued to see if the 

 added increase decays with time, and in this way the possible 

 error due to the contamination of the electrode with uranium 

 during the test has been eliminated. 



On June 5th, 1907, a test for actinium in the uranium 

 solution was made by simply exposing overnight a negatively 

 charged wire near the surface of the solution contained in a 

 large open beaker lined inside with tinfoil kept positively 

 charged with respect to the wire. The result was the same 

 as previously. 



The uranium solution contains 278 grams of uranium 

 (element). 



The amount of actinium detectable is necessarily much 

 greater than in the case of radium. The period of actinium A, 

 the longest-lived of the disintegration products of actinium 

 used in the detection of the latter, is about 170 times less 

 than that of the radium emanation, and for similar quantities 

 of the parent elements the amount of disintegration products 

 accumulating, and therefore the effect sought, must be 170 

 times less in the case of actinium than in that of radium. 

 This is on the assumption that the a. particle from the actinium 

 series is as efficient an ionizer as that from the radium series. 

 Under the most favourable circumstances only a fraction of 

 the actinium emanation generated can be made to produce 

 actinium A on the wire used in the test. As a rough 

 estimate, the amount detectable may be taken to be not less 

 than a thousand times greater than in the case of radium. 

 This is about 3 x 10 9 gram. In 278 grams of uranium in 

 300 days 4*6 x 10~ 8 gram of uranium disintegrates. 



Experiment Y. 



If actinium is the parent of radium it is to be expected 

 that the actinium emanation should, like radium itself, leave 

 behind a long-lived feeble residual activity after the ordinary 

 excited activity has disappeared, and this should prove to be 



