674 BELL SYSTEM TECHNICAL JOURNAL 



these conditions the experimental data are not easy to interpret for 

 that purpose. Suppose, however, that eventually the data are proved 

 to decide for one of the new forms of statistics against the old: what 

 then? Probably we shall merely remove one of the theoretical 

 foundation-stones of the kinetic theory of gases and insert another to 

 take its place, meanwhile leaving practically intact the great super- 

 structure of formulae and equations whereby the kinetic theory makes 

 contact with experience. Happily this is an easier process in theo- 

 retical physics than in architecture. 



Custom has lately changed the meaning of the term atomic theory, 

 making it almost synonymous with theory of the structure of the atom; 

 but the province which this latter has taken for its own is one to which 

 its forerunner disclaimed all right of entry. It was never supposed 

 that all of the properties of a gas can be interpreted as statistical features 

 of a swarm of corpuscles. The earlier atomic theory conceded some 

 of them to the individual atoms, thus in effect renouncing the ambition 

 to explain them; and among these were the spectra. Where the 

 statistical theory left off, the builders of atom models took up the 

 work. Bohr, for example, designed a model for the individual hydro- 

 gen atom, competent — at least to a great extent — to explain the 

 Balmer series and the rest of the line-spectrum of "atomic hydrogen." 

 This model he constructed, following Rutherford, out of a pair of 

 corpuscles. What he and his successors thus developed was in a 

 way an atomic theory of the atom — a degree deeper, or further, or higher 

 perhaps, then the atomic theory of matter which had provided him 

 with the notion and the scale of the atom to begin with. 



What then distinguishes this new "atomic theory of the atom" 

 from its ancestor.'* Well, the major differences in method and in aim 

 are traceable to the fact, that in the later theory the number of 

 elementary particles which constitute the system is quite manageably 

 small, while in the earlier, it is inconceivably tremendous. 



Bohr constructed his model for the hydrogen atom with only a 

 pair of corpuscles, and those for all the other atoms out of not more 

 than a few dozen each. Now with a model consisting only of two 

 particles, one can specify positions and velocities for these with the 

 utmost of precision, and go merrily ahead predicting and describing 

 orbits with as much exactness as one cares to lavish. Even with 

 dozens of electrons and a nucleus one can attain at least a specious 

 accuracy of detail; remember the portraits of the electron-orbits of 

 massive atoms which six or seven years ago were so profuse. Perhaps 

 it is not wise to make such definite assertions; but it is feasible. Not 

 so, however, with the subjects of the older theory. 



