1912] on the Origin of Radium. 411 



growth is due to ionium initially present, the period of ionium must 

 eertiiinly be greater than the longest of these periods. We may 

 safely conclude, if ionium is the only intermediate member, that its 

 period is at lead 100,000 years. This is 40 times longer than the 

 period of radium itself. 



Entirely independent confirmation of this conclusion was obtained 

 in another way. The gap in our knowledge is, strictly speaking, not 

 between uranium and ionium, for the direct product of uranium is 

 well known and is called uranium X. It gives /3-mjs alone in 

 disintegrating and has a period of only 35*5 days, so that in all the 

 ])receding work it has not been necessary to take it into account. It 

 would retard the growth of radium inappreciably. But, if the view 

 is right, the product of uranium Xmust be ionium, which gives a-rays. 

 Concomitantly with the rapid decay of the intense y8-rays of uranium X 

 there should occur a growth of a-rays due to the ionium produced. 

 Whether these a-rays can be actually detected will depend on the 

 period of ionium. Fifty kilograms of pure uranyl nitrate were pre- 

 sented to me by Dr. Beilby, and the uranium X out of this quantity 

 was separated, in the first place by fractional crystallization of the 

 material. The experiment consisted in watching for the growth of a 

 very feeble a-radiation during or subsequent to the decay of the 

 (3 rays. Owing to the fact that the /3-rays are readily deviated by a 

 magnet, the intense ;8-radiation of the uranium X was largely sup- 

 pressed, and any a-radiation remaining could be measured from day to 

 day, as the y8-rays decayed. This experiment has been repeated many 

 times, but with completely negative results. There always was a 

 feeble a-radiation, due to impurities, but it remained constant 

 throughout. Neither during, nor yet, several years after the separa- 

 tion of the uranium X, has there been any detectable growth of a-rays. 

 Actinium was also looked for, as the origin of this element, like that 

 of radium, remains unexplained, but the result also was completely 

 negative. The product of uranium X remains unknown, and if it be 

 ionium, the period of the latter substance must be enormous. The 

 minimum period it could possess, calculated from these negative 

 results, is 30,000 years. This, then, confirms the conclusion of the 

 first totally distinct set of experiments, that ionium must have a 

 period very much greater than that even of radium. 



Marckwald and Keetman have since concluded that uranium X is 

 identical in chemical behaviour with thorium, and therefore with 

 ionium. This conclusion has been confirmed by Mr. Fleck in my 

 laboratory after an exhaustive series of tests, not yet published. 

 Hence in the repeated separation of uranium A' from the large quan- 

 tity of uranium the very best method was adopted unwittingly for 

 removing ionium at the same time. This accounts for the gratifymg 

 result with preparation No. IV, which was the purest fraction of the 

 large quantity obtained after numerous crystallizations of the original 

 material. 



