Boltwood — Ratio of Badium to Uranium. 103 



The chief sources of error in the method as described are 

 due to two causes. One of these lies in the fact that the heat- 

 ing of the tube connecting the bulb A with the bulb B natu- 

 rally warms the gases contained within them and causes a slight 

 alteration in their relative capacities. This difficulty was 

 avoided as much as possible by wrapping both bulbs in wet 

 filter paper before the tube connecting them was heated. The 

 other possible error lies in the fact that some of the minerals 

 tested, particularly JSTos. 7 and 8, contained a considerable per 

 cent of sulphides. On treating these minerals with the aqua 

 regia necessary to decompose them, the sulphur of the sulphides 

 was in part oxidized to sulphuric acid and a noticeable quantity 

 of lead sulphate separated from the solution. Although radium 

 sulphate is quite soluble in the strong aqua regia used, it is not 

 impossible that slight traces of this compound were carried 

 down by the precipitated lead sulphate. This would explain 

 the lower values of the ratios obtained with 7 and 8. 



I am particularly indebted to Prof. H. A. Bumstead of Yale 

 University for valuable assistance and advice given in connec- 

 tion with this research, and to Prof. S. L. Penfield of Yale 

 University and Dr. Joseph Hyde Pratt of Chapel Hill, N. C. 

 for most generously supplying me with the minerals used in 

 these experiments. 



Conclusions. 



The quantities of radium present in the uranium minerals 

 which have been examined are apparently directly proportional 

 to the quantities of uranium contained in the minerals. 



Since it has been suggested by J. J. Thomson and Ruther- 

 ford* as very probable that radium is formed by the breaking 

 down of the uranium atom, and if such were the case a final 

 state of equilibrium and a definite proportion between the 

 uranium and radium present in minerals would be expected, 

 these results seem to be of value in furnishing experimental 

 evidence of the actual existence of this fixed relation. It is 

 planned to extend the work to other minerals containing a 

 smaller per cent of uranium and to introduce certain modifica- 

 tions in the method which will considerably increase its accu- 

 racy. 



New Haven, Conn., June 29, 1904. 



* Kadio-activity, Cambridge University Press, 1904. 



