December 17, 1915] 



SCIENCE 



857 



volved by the existence of branch products, 

 like actinium, and products of a rapid rate 

 of change, like the emanation and radium-A, 

 the diagrams represent the general condi- 

 tions and changes to be found in an old^ 

 uranium mineral. The chief relation of 

 interest shown by the diagram is that since 

 the radium changing to radio-lead can not 

 be determined experimentally with sufS- 

 cient exactness, it is equally satisfactory 

 and very much simpler to determine the 

 ionium changing to radium and compare 

 its quantity with the total amount of radium 

 in the mineral. As a matter of fact the 

 actual amounts of radium involved in these 

 two quantities need not be known, it is only 

 their relative values which are required, 

 since the value of the disintegration con- 

 stant is determined by the ratio of one of 

 these to the other. In this respect the 

 method is independent of any standard of 

 purity of radium preparations, an advan- 

 tage which is not possessed by other meth- 

 ods which have been used for attacking the 

 problem. Thus, for example, the estimate 

 of the half-value period of radium made 

 by Rutherford and Geiger as a result of 

 their experiments in 1908, had to be altered 

 from 1,760 years to 1,690 years, when in 

 1912 the present international radium 

 standard was adopted. 



The results of a number of experiments 

 conducted by the writer according to the 

 method just outlined were published in 

 1908. In the most satisfactory of these ex- 

 periments the mineral taken was a quantity 

 of pure, primary North Carolina uraninite, 

 almost free from secondary alteration prod- 

 ucts. About 40 grams of this material were 

 used and the ionium was separated (with 

 the thorium, which has identical chemical 

 properties) by the ordinary analytical 

 methods for the separation of thorium. 



T There are well-known examples of minerals too 

 young for a state of equilibrium to have yet been 

 reached between their radioactive constituents. 



The growth of radium in this preparation 

 of ionium was then measured over a period 

 of 147 days, and a rate of change for 

 radium corresponding to a half-value pe- 

 riod of 1,990 years was obtained. The re- 

 sults of the other experiments were in fair 

 agreement with this value, which was as- 

 sumed to be the most probable one. It is 

 interesting to point out that this estimate 

 was made between the time of Rutherford's 

 estimate of 2,600 years and Rutherford and 

 Geiger 's estimate of 1,760 years. 



In view of the disagreement of the value 

 obtained by the "growing" experiment 

 with the value as calculated from Ruther- 

 ford and Geiger 's work, it was reasonable 

 to suspect that in the "growing" experi- 

 ments all of the ionium was not separated 

 from the mineral. Such a suggestion was, 

 in fact, made subsequently by Rutherford. 

 A careful investigation of some of the con- 

 ditions associated with the usual methods 

 employed for the chemical separation of 

 small proportions of thorium from large 

 amounts of uranium, indicated that a com- 

 plete separation of the thorium under such 

 conditions was extremely uncertain if not 

 altogether impossible. The chemical be- 

 havior of uranium and thorium are stri- 

 kingly similar: in the case of the uranous 

 (UO2) salts the chemical analogy of the 

 two elements is such a close one as to make 

 any separation at all almost impossible. 

 Since an incomplete separation of the 

 ionium would lead to too small a produc- 

 tion of radium in the growing experiments, 

 under the assumption that the separation 

 was complete the calculated half-value pe- 

 riod of radium would receive too high a 

 value. It was therefore highly desirable 

 that the experiments should be repeated 

 under conditions which would avoid any 

 uncertainty, and which would give an al- 

 together trustworthy value for the life of 

 radium as determined by this method. 



