4:84 Prof. F. Soddy on the Relation 



that in preparation IV., show a somewhat closer corre- 

 spondence with theory than in the case of the latter 

 preparation. 



The natural leak of the electroscope used corresponds to 

 7xl0~ 12 gram of radium, and the extreme variation during 

 these sets of measurements has been only 6 per cent. It is 

 quite a mistake to suppose that quantities of the order 

 10~ 12 gram of radium are too small to be measured accurately. 

 Given due precautions against contamination, the percentage 

 accuracy of the measurements is at least as great when the 

 leaf takes an hour as when it takes a minute to cross the 

 scale. 



The results are given in the following tables which are to 

 be added on to those published *. Before leaving in 1915, 

 Miss Hitchins made a number of new radium standards and 

 determined the quantity of radium in the radium-barium 

 chloride preparation, referred to in the last paper (p. 217), 

 as 652 x 10~ 12 gram of radium per gram. The standards 

 employed were for the most part these, one new one being- 

 made from the same salt, which showed that the old ones 

 had not changed. The methods employed, the leak being- 

 taken with the leaf positively charged after three hoars,. 

 are the same as those previously used and have already 

 been fully described and illustrated. 



The constant for the electroscope, that is the number of 

 units of radium (10~ 12 gram) required to produce a leak of 

 1 d.p.m., for the first set of measurements is taken as 17*6 

 and for the second set as 16"83. Two measurements were 

 done on each of preparations Nos. IV. and III., and only one 

 on each of Nos. II. and I. These last two are open to some 

 doubt, owing to the presence of noticeable quantities of 

 nitrogen peroxide in the gas after introduction into the 

 electroscope, giving it a pronounced orange colour. Uranyl 

 nitrate preparations, purified by ether, are apt to generate 

 nitric oxide, which in absence of excess of oxygen, forms 

 nitrogen peroxide when mixed with air in the electroscope. 

 Unless the latter is perfectly dry the leaf system and 

 silvering are apt to be destroyed by this cause. 



At the end of the Table, the measurements from the start of 

 the growth of radium in the impurities, separated by ether y 

 from some 1*5 kilograms of commercial uranyl nitrate, used 

 for the earlier preparations, are included. It will be recalled 

 that the first evidence of the production of radium was obtained, 



* Loc. ciu 



