280 



Messrs. F. Soddy and T. D. Mackenzie on the 



is always perfectly air-tight, and in all our later measure- 

 ments the air-tightness o£ the instrument has been tested 

 immediately before each test. A small leak which may exert 

 no appreciable effect in 10 minutes might have an important 

 influence in several hours. In point of gain of time, both in 

 the actual measurement and the quicker recovery by the 

 electroscope of its normal natural leak after the emanation 

 is blown out, the advantages of the short method of working- 

 are very great, The curve shows the period of temporary 

 maximum after 10 minutes very clearly. It was obtained in 

 an electroscope of sensitiveness 3*2 to the 7-ray test, by the 

 introduction of the emanation re-accumulating in six days in 

 a solution of 3* 5 milligrams of uraninite. This curve is essen- 

 tially the inverse of the decay-curve of the excited activity 

 of radium produced by short exposure to the emanation 

 (Rutherford, 'Radioactivity/ 1905. p. 307). 



48 

 46 

 44 

 4£ 

 40 

 38 

 ce 36 



32 

 30 

 28 

 26 











































































































































. . 



1 





) 



i 

















































































































1 





















































































1 









































j 

















J~U7U 



>.27 A 



/SIC/7 



6Z 



fro 



aa/s 



771S-5 



■zcaz 

 mm 



vzu2 



ns 



\gfl 

 ■trra 



~?naT 



latio 

 te. 



72, 





f 























Sen 



ritire 



less c 

 S-Za 



tJEtet 



ZZSrp 



frose 

 rmii 



ope j 

 v. 



yrafy 



'/est) 





1 



f 



















































































1 











































14 16 18 20 22 



Minutes. 



2+ 26 28 30 32 34 36 38 40 42 



In the previous work the instrument was standardised in 

 terms of radium by direct comparison with a known small 

 fraction of the emanation from a weighed small amount of 

 pure radium bromide, and it was stated that a leak of 1 

 corresponded to the emanation from about 10~ 10 gram of 

 radium. The new method of calibration based on the work of 

 Boltwood and Rutherford (Am. Jour. Sci. 1906, xxii. p. 1), and 

 used by the former, was employed in the present work. The 

 radium is weighed out as uraninite, a sample of which is 



