CONTEMPORARY ADVANCES IN PHYSICS 325 



tlon-bands of oxygen, photographed with the brightest Hght and the 

 thickest layer of oxygen on earth — the rays of the decHning sun, 

 shining obUquely through the air. Aston has since observed them in 

 mass-spectra. They are probably the most unwelcome of all isotopes, 

 since they necessitate an extra precaution in comparing chemical 

 atomic weights with physical measurements of the masses of isotopes. 

 The chemists' unit of atomic weight is one sixteenth the weighted 

 mean of the masses of the oxygen isotopes, while the physicists' unit, 

 as I have said so often, is one-sixteenth-of-0^«. The difTerence between 

 the two, according to the latest estimates of the relative abundances 

 of the three isotopes, is about 125 parts in a million. O^^ is the 

 presumable residue of the transmutation of N^* by impact of an alpha- 

 particle, which frees a proton and fuses with what is left. This is 

 the most completely analyzed of all transmutations, and Blackett, 

 who first observed it in detail by the expansion-chamber method, 

 might be regarded as the discoverer of O^^ 



Mass-number 19 belongs to fluorine. The mass is given as 19.0000 

 ± .002, another remarkable example which might convince one of the 

 objective existence of the unit of atomic mass. 



As one goes onward along the list (which space forbids our scruti- 

 nizing henceforward in such fullness), one meets a novelty at atomic 

 number 18. Here begin overlappings of the atomic masses of different 

 elements: the isotope A^o of argon (Z = 18) is heavier than K^'^ of 

 potassium (Z = 19), and K^^ is heavier than Ca'*'' (Z = 20). The 

 former of these overlappings is responsible for the formerly very sur- 

 prising "inversion" whereby the chemical atomic weight of argon 

 (39.44) is greater than that of its immediate follower potassium 

 (39.10). Another inversion (involving tellurium and iodine) occurs 

 farther along in the list and is similarly caused, and there are many 

 other overlappings which do not produce so drastic a result. 



Mass number 40 is shared by two atoms of different atomic number, 

 different elements therefore, argon and calcium. The reader can 

 pick out other examples from Figs. 6 and 7. There is even an 

 instance of three "isobares," as atoms differing in Z but not in A are 

 called: this is at yl = 124, the three, elements being tellurium, tin 

 and xenon. (There is probably another at ^ = 96, but it is question- 

 able as yet, as of the three in question (Mo9^ Zr««, Ru»«) the two last 

 are not positively affirmed by Aston.) It will be interesting to find 

 whether measurements of mass can be pushed to such a degree of 

 accuracy as to disclose small differences between isobares. Aston 

 gives 79.926 and 79.941 for Se^" and Kr^", but adds "the difference is 

 too near the possible experimental error [one part in 10^] to be of 



