286 Messrs. F. ISoddy and T. D. Mackenzie on the 

 Table.-— Experiment I. 



Date. 



Days since 

 start. 



Radium 



Leak. 



y-ray Leak. 



Radium 

 Xl0- n g. 



1. 31/10/-05 



7 



1-4 



X 



21 1 S 



2. 14/11/-05 



21 



1-0 



2-6 



1-5 Li 



3. 5/12/-05 



43 



1-2 



. * 



i-8 r 3 



4. 17/ 1/-06 



85 



1-5 



26 



2-2J S 



5. 28/ 2/-06 



127 



1-5 



21 



2-7 



6. 18/ 5/-06 



206 



1-2 



21 



2-2 



7. 27/ 6/-06 



246 



1-2 



20 



23 



8. 8/ 8/-06 



288 



1-4 



2-25 



2-4 



9. 29/ 8/-06 



309 



1-55 



21 



2-8 



10. 31/ 5/-07 



584 



1-75 



...t 





11. 25/ 6/-07 



609 



Lost 







12. 3/ 7/-07 



617 



2-0 



2-8 



2-7 



* It is probable that the y-ray leak of the electroscope in tests Nos. 1 and 3 

 was similar to that found in tests Nos. 2 and 4, but it was not actually taken. 



t Through a mistake the y-ray leak in this 10th test was inaccurate and it 

 has been omitted. 



Regarding first the results o£ the actual observations 

 recorded in the third column, it will be seen that the lowest 

 result is the second, 21 days from the start, and the highest, 

 which is just twice the lowest, is the last, 617 days from the 

 start. For the last six tests there has been a regular increase 

 from 1*2 to 2'0, and the greater uniformity of the latter half 

 of the test may be the result of the experience gained in the 

 former tests. The values corrected for the sensitiveness of 

 the electroscope as shown by the 7-ray test in the fourth 

 column, are recorded in terms of 10 — u gram of radium in 

 the last column, but these results are not so regular as the 

 actual observations. It is to be noted that a gradual increase 

 in the value of the 7-ray standard with time would operate 

 against and tend to mask a real increase in the quantity of 

 radium present. 



The long interval of 275 days between the ninth and 

 tenth tests was intentionally allowed, so that this experiment 

 might be as nearly as possible comparable, in point of lapse 

 of time between the tests, with the former experiment. 

 Considerations similar to those discussed in the next experi- 

 ment made it very desirable to prove that the positive result 

 before obtained was not due to the long interval that elapsed 

 between the tests at the end of the first year from purification. 



The results may be summed up by the statement that the 

 amount of radium produced from 250 grams of uranium in 

 600 days is too small to be sure of. It is probable that the 

 slight increase observed may prove to be real, but more time 



