Production and Origin of Radium. 737 



in good agreement with that determined later for a similar 

 preparation. 



§ 3. Experimental Methods. 



Before considering further experiments, a brief description 

 will be given of the experimental methods employed to 

 determine accurately the quantity of radium in the various 

 preparations. The preparation was obtained in the form of 

 a solution and placed in a glass flask. The solution was then 

 boiled to expel completely all the emanation, and the exit- 

 tube sealed before the flask had cooled. After a definite 

 time-interval, the flask was opened and the air, mixed with 

 emanation, was expelled by vigorous boiling and collected 

 over water. Boltwood has shown that boiling is the only 

 satisfactory method of expelling all the radium emanation. 

 The air plus emanation was then introduced through a drying- 

 tube of phosphorus pentoxide into an exhausted electroscope 

 of capacity greater than the volume of air to be introduced. 

 Air was then let in to fill the electroscope to atmospheric 

 pressure and the electroscope closed. The electroscope em- 

 ployed in most of these experiments consisted of an Erlenmeyer 

 flask of about 1 litre capacity silvered on the inside, in which 

 the gold-leaf system was insulated after Wilson's method by 

 a sulphur bead. The motion of the gold-leaf was read 

 through openings in the silvering by a microscope with a 

 scale in the eyepiece in the usual manner. The natural leak 

 of the electroscope was small and corresponded to '14 division 

 per minute of the scale in the eyepiece. Since the emanation 

 from 10 -9 gram of radium in equilibrium produced a move- 

 ment of 11*5 divisions per minute, the natural leak was 

 equivalent to that produced by 1'2 x 10 -11 gram of radium. 

 As is generally observed, the natural leak increased gradually 

 for several days on standing, probably due to a small trace 

 of radium present, but always came back to the same value 

 if the electroscope was exhausted and refilled two hours before 

 observations were begun. Since the vessel was always ex- 

 hausted to introduce fresh emanation, this peculiarity of the 

 electroscope was no disadvantage. The natural leak tested 

 in this way was remarkably steady, and never varied more 

 than ten per cent, over the course of several months. 



Readings of the movements of the gold-leaf were com- 

 menced three hours after the introduction of the emanation. 

 At this time, the active deposit is nearly in equilibrium with 

 the emanation, and there is only a very slight change of the 

 rate of movement for several hours. The electroscope was 

 calibrated and standardized by means of the emanation from 



