96 THE EVOLUTION OF MATTER 



great. Were radium the first direct product, the 

 growth of radium in uranium, initially purified com- 

 pletely from it, could be observed in the course of an 

 hour, so excessively delicate are the radioactive tests 

 for this new element. Experiments were started in 

 1003 in London, continued on a very much larger 

 and more thorough scale in Glasgow, with the aid 

 of Mr T. D. Mackenzie. Yet in 1914 the expected 

 confirmation was still not clearly forthcoming. Long- 

 before that time it was known that radium was not 

 the direct product of uranium, and that another new 

 radio-element, ionium, intervened in the series. The 

 uranium changes into radium, via ionium, and this 

 ionium is an exceedingly slowly changing element in 

 comparison even with radium, not more than about 

 joo^ooth part changing every year. This retards 

 enormously the initial rate of growth of radium and 

 makes it proceed at first not linearly with the lapse 

 of time, but according to the square of the lapse of 

 time. That is, the growth after ten years would be 

 100 times, and after 100 years 10,000 times, that in 

 the initial year from purification. The oft-tested prep- 

 arations of uranium were transplanted to Aberdeen 

 in safety, and tests since carried out, in conjunction 

 with Miss Ada Hitchins, last year satisfactorily 

 established a growth of radium beyond all doubt in 

 the largest preparation, and showed that the rate was 

 proceeding as nearly as can yet be seen according 

 to the square of the time. The growth of radium 

 was not large. In three years it amounted to 

 iso.ooo.ooo.ooo.oooth of the quantity of uranium experi- 

 mented upon, and in six years to just four times 

 this quantity. The experiments gave, moreover, 

 indirectly a maximum estimate of the rate of change 

 of ionium as at most foo^ooth P art P er year. This 

 estimate has now been confirmed and made more 

 definite by some very fine direct work on ionium 



