Relation between Uranium and Radium. 287 



must elapse before this can be decided. If we take the mean 

 of the first four observations as giving the quantity of radium 

 initially present, this is l'9x 10~ u gram, and the subsequent 

 increase is therefore not greater than 10 — n gram. 



Accepting Rutherford's value for the rate of disintegration 

 of uranium in 250 grams in 600 days 0'8 x 10 -7 gram of 

 uranium disintegrates. The amount of radium formed is 

 therefore not greater than one eight-thousandth part of the 

 uranium disintegrating. 



In this connexion may be mentioned an experiment of 

 Boltwood (Am. Jour. Sci. 1905, vol. xx. p. 239) who kept a 

 quantity of 100 grams of uranyl nitrate (=48 grams of 

 uranium), purified by repeated crystallization from water, 

 for 390 days and was unable, either initially or finally, to 

 detect any radium present. It was stated that 1*7 xlO -11 

 gram of radium could have been detected if present. Esti- 

 mating as before, the amount of uranium disintegrating is 

 10~ 8 gram, about 600 times the quantity of radium stated to 

 be detectable. The experiment is with too small a quantity 

 of material yet to afford much information. It corresponds 

 to a period of less than six weeks with a kilogram of material. 

 It is clear that repeated crystallization from water would 

 have a better chance of removing intermediate products than 

 precipitation with barium sulphate, but even if it had not 

 done so the effect sought, with the small quantity of uranium 

 in the time of the experiment, could only have been but 

 small. On the basis of the old experiment with a kilogram 

 of material, which gave a leak of 15 in 550 days, 100 grams 

 in 400 days would give a leak of 1, of the same order as the 

 natural leak of our electroscope, and as the limit, 1/7 x 10 -11 

 gram of radium, detectable. Making allowance for this limit 

 as stated being too high (Rutherford and Boltwood, Am. 

 Jour. Sci. 1906, vol. xxii. p. 1). and for the fact that 

 Boltwood boiled his uranium solution, it is probable that a 

 growth of radium, if it had occurred at the same rate as in 

 the sample purified by barium sulphate, would have been 

 detectable but only by a narrow margin. 



Experiment II. 

 This experiment is practically a duplicate of the last and 

 was designed to cover the possibility of one of the intermediate 

 disintegration products between uranium and radium being a 

 gas with considerable life -period, not itself radioactive but 

 disintegrating raylessly, ultimately generating radium. In 

 Experiment No. I. such a gas would be removed in the 

 periodic tests and lost, and the result on this account might 



