220 Mr. F. Soddy on the 



equilibrium value of the " extra " radiation due to uranium Y 

 (Curve A IV.) would be about 14 d.p.m. The radiation due 

 to 19 hours accumulation of uranium X is 22 d.p.m., so that 

 in equilibrium this radiation will be about 1000 d.p.m. The 

 ratio is 1*4 per cent. From Curve A III. (fig. 2) it follows 

 that the equilibrium " extra " radiation due to the uranium Y 

 is 1*6 d.p.m., whereas that due to the uranium X would be 

 about 100 d.p.m., in substantial agreement with the former 

 result. 



The soft /3-radiation of the uranium X in the A curves 

 (fig. 1) is about 3 d.p.m., corresponding with 136 d.p.m. in 

 equilibrium. The ratio of this to the 14 d.p.m. contributed 

 by the uranium Y is again of the order to be expected if the 

 latter is the parent of actinium in a branch claiming 8 per 

 cent, of the total atoms disintegrating, and if each atom of 

 uranium X and uranium Y gives, on disintegration, a similar 

 amount of ionization. The same result can also be deduced 

 from the recovery curves. The equilibrium soft /3-radiation 

 of a film of uranium oxide such as was employed in the 

 A curves, fig. 2, is about 12 d.p.m. 



The uranium oxide in this experiment (fig. 2, curves A) 

 weighed 1'96 g. or 8 mg. per cm. 2 In the hope of obtaining 

 clearer evidence of the increased initial slope of the recovery 

 curve, the curves were taken for another uranium oxide 

 preparation weighing only 0*3 g. or 1*2 mg. per cm. 2 . It is 

 shown in fig. 2, curves B I. and II. The hard /3-radiation is 

 thus reduced 6 or 7 times. The total radiation, however, 

 was so exceedingly small that though the curves were taken 

 with the utmost care, several hours being spent on each set 

 of measurements, the variations of the individual points from 

 the smoothed curve are too great to allow of the difference 

 curve being drawn with any certainty. But here also it can 

 be shown that a much more marked difference of slope than 

 that of A III., fig. 2, is not to be expected. For even in the 

 thicker film, weighing 8 mg. per cm. 2 , more than 50 per cent, 

 of a radiation, for which /x(Al) = 300 (cm.) -1 , would escape 

 absorption, so that for an infinitely thin film the relative 

 initial increase of slope could not be so great as twice that 

 shown in A III. (fig. 1). On the whole I am inclined to 

 believe that the recovery curves do show a real but very 

 small effect due to uranium Y, which is not much beyond 

 the possible error of measurement. Clearly it will be very 

 difficult to get more conclusive evidence on this point. 



Finally, a search was made for a- radiation by the method 

 formerly used (Phil. Mag. 1909 [6] xviii. p. 859), the 

 preparation being mounted on a rectangular slip of metal 

 placed between the poles of a powerful electromagnet in an 



