CONTEMPORARY ADVANCES IN PHYSICS 65 



in their original ratio; electrolysis does not favor one above the other, 

 nor does osmosis, and if a mixture of the two elements or of salts of 

 theirs is presented to an absorbent or an adsorbent or a solvent, it 

 accepts them in the same proportions as they are presented, while any 

 element willing to react with either reacts in precisely the same degree 

 with the other.9 Similarity such as this goes far beyond the interre- 

 semblances of the alkali metals, for instance, or even those of the rare- 

 earth elements, difificult as the task of separating these latter from one 

 another sometimes proves; it is not similarity merely, it amounts to 

 identity. 



As for the presumption that radioelements sharing the same atomic 

 number should be alike in what are loosely termed the "physical 

 properties," it is more difficult to test. In fact, there seems to be no 

 instance of two such elements, each radioactive and each obtainable 

 quite unmixed with the other in quantities large enough for such ex- 

 periments. The three elements sharing atomic number 86 are all 

 gaseous at ordinary temperatures, but they are too scanty and two of 

 the three are much too fugitive for making accurate comparative 

 measurements of such qualities as viscosity or elasticity or ionizing 

 potential. The elements sharing atomic number 82 are, as I shall 

 presently bring out, mostly stable, and upon them it is possible to test 

 the expected coincidence in optical line-spectra and X-ray spectra, 

 which is verified except for certain very minute (but unexplained!) 

 differences in the wavelengths of certain lines. The band-spectra of 

 these elements (more precisely, of their compounds) display slight dif- 

 ferences which are beautifully explained by the contemporary theory 

 of band-spectra, involving as it does a participation of the nucleus 

 with its mass in the production of the bands. ^^ Mixtures of two of the 

 elements sharing atomic number 90 (thorium and ionium) display pre- 

 cisely the same optical spectrum as pure thorium. ^^ In addition to 

 these observations, a great many have been made upon the physical 



^ There is a huge literature of the attempts to separate elements of identical 

 atomic number and to discriminate between their chemical properties, for a review 

 and bibliography of which I refer again to von Hevesy and Paneth {I.e. supra, 

 chapter XII). 



10 E. S. Bieler, Nature, 115, p. 980 (1925). 



_" This is vividly illustrated by a passage in the classical treatise upon radioactivity 

 which Rutherford wrote in 1912. Boltwood had isolated from uranium ores a sample 

 of thorium oxide which emitted, along with the alpha-particles from the thorium, a 

 considerable number coming from ionium. Russell and Rossi produced its arc 

 spectrum and "the spectrum of thorium was obtained, but not a single line was ob- 

 served that could be attributed to ionium. On the assumption that ionium has a life 

 of 100,000 years, the preparation should have contained 10% of ionium. Since 

 probably the presence of 1% of ionium would have been detected spectroscopically, 

 it would appear that the ionium was present in small amount, indicating that the life 

 of ionium must be much less than 100,000 years." As a matter of fact there was 

 probably more than 10% of ionium in the mixture; but its spectrum lines were 

 identical with those of thorium. 

 5 



