776 Mr. F. Soddy on the Production 



into the electroscope. Care was taken to use fresh burettes, 

 rubber tubing, mercury, &c. for each dilution, as it was 

 found that otherwise the results were quite untrustworthy. 

 In one experiment, the emanation had been allowed to 

 accumulate for 16 days in the radium solution. In this time 

 about 94 per cent, of the maximum or equilibrium quantity 

 accumulates. It was diluted with air to 73 c.c. and 0*71 c.c. 

 taken. This was diluted to 92 c.c. and 0*042 taken. The 

 last quantity was made up to 50 c.c. and 20 c.c. introduced 

 into the electroscope. After 5 minutes the corrected rate of 

 leak was 24 divisions per minute. 



Several determinations gave results of the same order of 

 magnitude, although the individual determinations varied over 

 an extreme range of 1 to 4. From the mean of the results 

 it was deduced that the equilibrium quantity of emanation 

 from about 10~~ 10 gram of radium (element) caused a leak of 

 1 division of the scale per minute. 16*5 divisions per minute 

 may be taken as the measure of the equilibrium amount of 

 emanation accumulating in the uranium solution after 567 

 days. This corresponds to the presence of 1'6 x 10 -9 gram 

 radium. If the uranium nitrate is assumed to contain 50 per 

 cent, of uranium, and the radioactive constant of the latter 

 is taken as 6xl0~ 17 (sec. -1 ), then, on the assumption that 

 there is a direct change of uranium into radium, the quantity 

 of the latter that would be produced from a kilogram of 

 uranium nitrate in 567 days is l"5x 10 — 6 gram. The effect 

 of the uranium X stage would be practically inappreciable 

 after a period of one and a half years. So that the result 

 shows that the production of radium from uranium during 

 the first one and a half years is only about one thousandth of 

 the quantity calculated on the assumption of a direct change. 



IV. — From the figure already given it will readily be seen 

 that the effect of the uranium X stage is to reduce the initial 

 rate of production of radium below the final rate during an 

 initial period corresponding in order of magnitude to the 

 period of average life of the uranium X atom. The same 

 consideration would hold true for any other unknown inter- 

 mediate products between uranium X and radium. So that 

 the low result obtained would be explained if one or more 

 intermediate forms existed between uranium X and radium, 

 for which the period of average life was large compared 

 with the time of experiment. Before considering some 

 independent evidence in favour of this view, another possi- 

 bility has first to be considered. It seemed not unlikely that 

 the radium, under the conditions in which it was produced, 



