2IO 



NA TURE 



[August i i, 1923 



for radiothorium. For all practical purposes the two 

 uranium are as non-separable by this method as if they 

 were actually successive products. I spent many years, 

 before this part of the scries was at all well known, 

 lookinj; for the product of uranium X, and separated 

 this constituent from 50 kilograms of uranium nitrate 

 repeatedly in the attempt. I was looking for a growth 

 of a-rays concomitantly with the decay of the /?-rays 

 of the uranium X. If the product had been ionium, 



a a fi a 



as at first thought (UI — ^UII— ^UX — ^lo — >), 

 it should have been just possible to detect it; but 

 since it is the 30 times longer-lived uranium II, the 

 attempt is hopeless, especially as uranium X and 

 ionium are isotopes, and therefore the uranium X 

 separated must always possess a certain initial a- 

 activity due to ionium. 



The Absolute Chemical Identity of Isotopes 

 AND its Implications. 



The years 1908-10 were productive of many pro- 

 longed and serious efforts to separate isotopes by 

 chemical means. In 1908 Boltwood discovered ionium 

 and showed that it resembled thorium. Keetman, 

 who with Marckwald discovered ionium independently, 

 tried twelve good methods all known to be effective 

 in the purification of thorium in the attempt to separate 

 the ionium from thorium, completely without success. 

 Auer von Welsbach, on a technical scale, separated 

 the ionium and thorium from 30 tons of pitchblende 

 and tried fresh methods in the hope of separating 

 them, but failed. It was with this preparation that 

 Exner and Haschek tried without success to find the 

 ionium spectrum ; and Russell and Rossi confirmed 

 their result, that the spectrum of ionium was that of 

 pure thorium. When later I had determined beyond 

 doubt, from measurements of the rate of growth of 

 radium from uranium, that the period of ionium was 

 100,000 years, and that Welsbach's preparation must 

 have been approximately 30 per cent, ionium and 

 70 per cent, thorium by weight, it followed that the 

 spectra of isotopes must, like their chemical character, 

 also be identical. The difference, if any exists, is 

 almost beyond the limit of detection by the most 

 powerful methods. 



Similarly, the chemical identity of radium D and 

 lead was established as a consequence of very prolonged 

 and refined chemical examination. Paneth and Hevesy 

 established upon this their well-known method of 

 using radioactive isotopes as indicators for elements 

 in too small quantity to be dealt with except by such 

 methods. On the principle that wherever the radio- 

 active element is there will its inactive isotope be also, 

 provided that they have once been properly mixed, 

 many difficult or uncertain chemical analyses may be 

 converted into simple radioactive ones. 



In 1909 Stromholm and Svedberg made what was 

 probably the first attempt to fit a part of the dis- 

 integration series into the Periodic Table, and although 

 the effort in itself was in an important respect erroneous, 

 in their paper is to be found the first anticipation that 

 the chemical non-separability found for certain pairs 

 and groups of radio-elements may also apply to the 

 non-radioactiye elements. Remarking on the fact that 

 there are three parallel and independent radioactive 



NO. 2806, VOL. 1 12] 



series, they suppose this to proceed down through 

 the Periodic Table, " but that always the three elements 

 of the different genetic series, which thus together 

 occupy one place in the Periodic System, are so alike I 

 that they always occur together and also have not 

 been able to i)e apprecial)ly separated in the lalM>rator>'." 

 They point out also, this idea would explain the excep- 

 tions to the Periodic System " if the elements of the 

 scheme were mixtures of several homogeneous elements 

 of similar but not completely identical atomic weight. " 



In the next year I arrived independently, and 

 without in the least postulating any continuance of 

 the genetic .series beyond the radio-elements, at a 

 similar view. Marckwald and I found independenth 

 that mesothorium i was chemically similar to radium, 

 a fact undoubtedly known to Hahn and those engaged 

 in the technical extraction of mesothorium, but kept 

 secret. It was known from some work of Boltwood 

 that precipitating barium sulphate in a solution con- 

 taining mesothorium removes it, but it was thought 

 that the action of the barium sulphate was similar 

 to that in removing uranium X, for which it had Ion- 

 been used, namely, a simple adsorption. I wa- 

 surprised to find it absoluiely different. The removal 

 of the barium from the mesothorium, as from radium, 

 could only be accomplished by the fractional cr)-stallisa- 

 tion of the chlorides. In this fractionation the radium 

 and mesothorium remained together and behaved as 

 a single element. Within the limit of error of tht- 

 most careful radioactive measurements, there was no 

 change in the relative proportion of the two elements 

 at the end of the process from that in which they 

 exist in the original mineral. ■ 



Chemistry has many cases of elements similar in f| 

 chemical character, but nothing approaching this. 

 For we know, beforehand, that we are dealing with a 

 mixture of two substances and can estimate accurately 

 the proportion of each individual. Yet to all chemical 

 operations they behave as a single substance. The 

 differences of atomic weight are considerable, two 

 units in the cases of mesothorium and radium, and of 

 ionium and thorium, and four units in that of radio- 

 thorium and thorium. It was certain that if isotopes 

 existed in the case of the ordinary' chemical elements 

 the absence of a second radioactive nature independent 

 of the chemical nature would make it impossible for 

 them to have been recognised. Hence the implication 

 followed that any supposed element may be a mixture 

 of several chemical identities of different atomic 

 weight, and any atomic weight might be merely a 

 mean number (Ann. Reports, Chem. Soc, 1910, 286). 

 There is an element of tragedy in this. The lifetime 

 labours of the chemists who, since the time of Stas, 

 have devoted themselves to the exact determination 

 of atomic weights appear to have as little theoretical 

 interest now as if you sought to determine the average 

 weight of a collection of beer bottles, all exactly alike 

 but not all quite full. 



The Radio-elements and the Periodic Law. 



The years from 1911-13 were crowded with im- 

 portant advances, and to do the exact histon.- justice 

 would take an undue share of the available time. 

 In 191 1 the chemistry of most of the a-ray- giving 

 members was sufficiently known for it to be seen 



