1917] on The Complexity of the Chemical Elements 1^51 



made the experimental proof of the production of radium from 

 uranium such a long piece of work. Previously only negative results 

 were available. One could only say, from the smallness of the 

 expected growth of radium, that the period of average life of ionium 

 must be at least 100,(M)(> years, forty times longer than that of radium, 

 and, therefore, that there must be at least forty times as much ionium 

 by weight as radium in uranium mmerals, or at least 13 "(i grams per 

 iboo kilos, of uranium. Since then further measurements, carried 

 out by Miss Hitchins last year, have shown definitely for the first 

 time a clear growth of radium from uranium in the largest prepara- 

 tion, containing 3 kilos, of uranium, and this growth, as theory 

 requires, is proceeding according to the square of the time. In three 

 years it amounted to 2 x 10" " grams of radium, and in six years to just 

 four times this quantity. From this result it was concluded that the 

 l^revious estimate of 100,000 years for the period of ionium, though 

 still of the nature of a minimum rather than a maximum, was very 

 near to the actual period. 



Joachimsthal pitchblende, the Austrian source of radium, contains 

 only an infinitesimal proportion of thorium. An ionium preparation 

 separated, by Auer von AVelsbach, from 30 tons of this mineral, 

 since no thorium was added during the process, was an extremely 

 concentrated ionium preparation. The atomic weight of ionium — • 

 calculated by adding to the atomic weight of its product, radium, 

 four for the a-particle expelled in the change — is 230, whereas that of 

 thorium, its isotope, is slightly above 232. The question was whether 

 the ionium-thorium preparation would contain enough ionium to show 

 the difference. Honigschmid and Mile. Horovitz have made a special 

 examination of the point, first redetermining as accurately as possible 

 the atomic weight of thorium and then that of the thorium-ionium 

 preparation from pitchblende. They found 232*12 for the atomic 

 weight of thorium, and by the same method 231*51 for that of the 

 ionium-thorium. A very careful and complete examination of the 

 spectra of the two materials showed for both absolutely the same 

 spectrum and a complete absence of impurities. 



If the atomic weight of ionium is 230, the ionium- thorium pre- 

 paration must, from its atomic weight, contain 30 per cent of ionium 

 and 70 per cent of thorium by weight. Professor Meyer has made a 

 comparison of the number of a-particles given per second by this 

 preparation with that given by pure radium, and found it to be in 

 the ratio of 1 to 200. If 30 per cent is ionium, the activity of pure 

 ionium would be one-sixtieth of that of pure radium, its period some 

 sixty times greater, or 150,000 years. This confirms in a very satis- 

 factory manner our direct estimate of 100,000 years as a minimum, 

 and incidentally raises rather an interesting question. 



My direct estimate involves directly the period of uranium itself, 

 and if the value accepted for this is too high, that for the ionium 

 will be correspondingly too low. Now, last week. Professor Joly was 



K 2 



