Relation between Uranium and Radium. 293 



present bad increased from 5*7 to 14'2 ( x 10 -9 gram). 

 Rutherford observed an increase in the amount of radium in 

 an actinium solution corresponding to 3 x 10 _9 gram radium 

 per year in 0*5 gram of actinium (activity 250 times uranium). 

 He refers also to experiments with the residual excited 

 activity of actinium which was obtained from a very active 

 actinium preparation, and dissolved off the platinum plate 

 on which it was deposited by acid, and then tested for the 

 growth of the radium emanation. He records a real growth 

 of radium in the solution, which was much less than what is 

 theoretically to be expected if actinium B changes directly 

 into radium. In the later communication Rutherford finds 

 that the parent of radium can be separated from actinium 

 and is a separate substance. He considers it not improbable 

 that actinium may have no direct connexion with the parent 

 of radium. If this is the case the residual activity of actinium 

 should not generate radium. 



Experiment VI. 



If the early positive result was correct, the residues from 

 the uranium should contain the radium-producing substance 

 which has lately been found by Boltwood and Rutherford in 

 preparations of actinium, and this substance should not be, 

 at least wholly, removed by the old method of precipitation 

 with barium sulphate. The residues were therefore purified 

 as carefully as possible from radium by this method. They 

 contained between 100 and 200 grams of uranyl nitrate, and 

 were dissolved in about two litres of water containing excess 

 of sulphuric acid. The liquid was precipitated in the cold 

 four times, each with one gram and once with 2*5 grams of 

 barium nitrate, the barium sulphate being removed after each 

 addition. It was then evaporated to 300 c.c. and tested by 

 the bubbling method, which showed that the quantity of 

 radium present was still greater than 10 -10 gram. The 

 solution was again diluted and precipitated hot five times 

 with one gram and once with two grains of barium nitrate. It 

 was finally sealed up in a glass flask provided with inlet and 

 outlet tubes, sealed into the neck, on June 17th, 1907, and 

 first tested on June 20th, 1907, in the manner described in 

 the case of the samples of uraninite. No appreciable quantity 

 of radium was present. The increase of the electroscope 

 leak was only 023. A second test on July 3rd, 1907, in 

 an electroscope giving 3*1 to the 7-ray test, gave a corrected 

 equilibrium leak of 0*33, corresponding to the presence of 

 only I x 10~ 12 gram of radium. 



It is hoped in the course of time that this experiment and 



