Existence of Uranium Y. 2.19 



The difference curves (A IV., B IV.), obtained by extra- 

 polating the observed curve for the total radiation, after the 

 decay had become normal, back to the origin and subtracting 

 this extrapolated curve, were roughly exponential. The half- 

 valne periods for different curves and lor different parts of 

 the same curve varied somewhat widely, between 20 and 

 33 hours, but was, as a mean, nearer 24 hours than 36 hours. 

 Naturally, since one is dealing with only a 16 to 30 percent, 

 difference between an exponential and extrapolated curve, a 

 considerable variation may be made in the individual values 

 of the halt-period, according to the way the points are chosen. 

 Curves A IV. and B IV. were obtained from actual large 

 seale pencilled curves and not from the necessarily rougher 

 ink diagrams used for fig. 1. The results were corrected in 

 terms of the hard /3-rays of a standard preparation of uranium 

 oxide, but it is clear that barometric and temperature changes 

 must exert a considerable effect on the soft /5-radiation 

 which it is not easy to correct for. This probably accounts 

 for the difficulty of getting the curves for the uncovered 

 preparation as regular as those for the covered. Fig. 1 

 shows the decay curves for two preparations of uranium X 

 separated from the purest uranyl nitrate, after a time of re- 

 accumulation (A) of 19 hours by the ferric method, (B) of 

 23 hours by the barium sulphate method. The ordinates 

 for the latter preparation have been doubled. 



With regard to the recovery curves, there was some slight 

 evidence of an initial slope greater than the theoretical. 

 The smoothest and most regular curves, obtained from the 

 purest uranium preparations, are shown in fig. 2 A. Curves 

 A I. and A II. represent the curves for this preparation, bare 

 and covered with the celluloid screen respectively. Curve 

 A III. was obtained by subtracting from A I. a curve derived 

 from All. by increasing the ordinates by the factor 1*15 to 

 allow for the absorption of the hard /3-rays in the celluloid. 

 This factor was determined experimentally, and the result 

 agreed with that calculated from the absorption coefficient. 

 It will be seen that A III. shows a distinct though small 

 increase of slope over the first 100 hours during which 

 uranium Y, if it exists as a separate product isotopic with 

 uranium X, would be attaining the equilibrium value. This 

 increase of slope, though hardly beyond the region of ex- 

 perimental error, is of the right order to be expected from 

 the curves fig. 1. 



Taking the A curves of fig, 1, the time of accumulation 

 was 19 hours. If the period of uranium Y is 1*5 days, 

 0*41 of the equilibrium amount would be present and the 



