Boltwood — Production of Radium from Uranium. 241 



The third and more elaborate article by the same author 

 appeared in the June number of the Philosophical Magazine. 

 The data briefly given in the earlier articles are here treated 

 at greater length and a closer insight can be gained of the 

 experimental methods and the results on which the author's 

 later conclusions are based. Although it is stated in this paper 

 that observations had been taken occasionally over a period of 

 eighteen months and that these observations indicated a grad- 

 ual growth of the emanating power of the uranium solution, 

 the only definite and directly comparable numbers are restricted 

 to a total period of about three weeks (Dec. 17, 1904 to Jan. 

 9, 1905) and include only four measurements conducted at the 

 close of the period of observation. 



Without entering into a discussion of various minor details 

 in Mr. Soddy's papers, it is desired to call particular attention 

 to the following important considerations in relation to the 

 experimental data submitted : 



First. No conclusive evidence is brought forward to show 

 definitely how much or how little radium was present in the 

 uranium solution at the commencement of the experiment* 



Second. It appears extremely possible that the increase in 

 the content of radium which Mr. Soddy believes he has observed 

 in his uranium solution may in fact have been due to the acci- 

 dental and unconscious introduction of radium salts during the 

 tests conducted at the end of the twelve months period. 

 According to his own statements these tests were carried out 

 in a laboratory notably contaminated with various radio-active 

 products, and the accidental introduction of the sub-micro- 

 scopic quantity of material (l'6xl0~ 9 gram.) which was after- 

 wards detected would account for the later positive results. 

 The liability of contamination from an extraneous source is 

 strongly suggested by the behavior of Mr. Soddy's electro- 

 scope, in which the normal air leak has risen from 0*048 divi- 

 sion per minute to 1*56 division per minute, an increase of 

 over thirty times, during the period covered by his experi- 

 ments. 



The conditions essential to the elucidation of the question of 

 the actual production of radium in uranium compounds would 

 seem to be : 



* The writer of the present paper convinced himself at the beginning of 

 his own experiments that the method of procedure followed by Mr. Soddy 

 in testing his solutions for radium emanation is entirely unsuited for the 

 determination in question. A concentrated solution of incompletely puri- 

 fied uranium nitrate containing traces of radium gave up only a fraction of 

 the total radium emanation generated within it when the solution was 

 allowed to stand for days in contact with a small air space and air was 

 bubbled through it. It was speedily found that only by boiling the solution 

 vigorously for about fifteen minutes could the total emanation present be 

 positively separated. 



