40 SECTIONAL ADDRESSES. 



lines can be similarly explained as due to singly ionised nitrogen and 

 singly ionised oxygen. 



The identification of these lines was made by ignoring so far as conveni- 

 ent the rules of the quantum theory which had been evolved from labora- 

 tory experience, and given some theoretical basis by Bohr and his followers. 

 These rules forbid certain lines which might occur according to the com- 

 bination principle. When a state of excitation of the atom is such that it 

 cannot directly pass to a lower state without breaking one of these rules, 

 that state is called metastable ; and this is the case which we have in the 

 nebular lines. I shall return presently to the consideration of metastable 

 states and ' forbidden ' lines. 



The Auroral Spectrum. 



The next cosmical problem I wish to refer to is the long outstanding 

 one of the green line of the aurora. This was first seen by A. J. Angstrom 

 at Upsala in 1868, and he recorded the observation in one of the supple- 

 mentary notes at the end of his great paper in which an extensive scale of 

 wave-lengths for the solar spectrum was first established. In this case the 

 enigmatic line is even more isolated than in the case of the nebulte, since, 

 except in the case of unusually bright auroras, one can see nothing else in 

 the spectrum at all. For some years I took every available opportunity 

 of looking at this spectrum, and never did so without a deep sense of 

 mystery. The origin of the line was not in this case in the depths of space, 

 but in our own atmosphere at the distance of a short railway journey from 

 the observer. Yet an apparently exhaustive study of the spectra to be 

 obtained from terrestrial gases by the combined efforts of very many 

 experimenters gave no clue to its origin. 



As is well known, the clue was eventually found by McLennan, who 

 was able to produce the line by heavy electric discharges in a mixture of 

 oxygen and helium, or, better, oxygen and argon. Oxygen is essential, 

 and there is now no doubt that the aurora line is an oxygen line, but the 

 function of the inert gas is not very clear, though various more or less 

 plausible guesses may be made. To have established that the line is due 

 to oxygen is an immense step forward. There is, however, yet more to 

 be done, for we do not know how to get the green line alone or with only 

 the negative nitrogen bands as we see it in the sky. In the artificial 

 spectrum the ordinary oxygen lines and the lines of the inert gas, helium 

 or argon as the case may be, are conspicuous. 



The wave-length of the auroral line could not be foreseen or calculated 

 from our present knowledge of the arc spectrum of oxygen. In this case 

 we have only a single line to deal with, and are thus without the invaluable 

 clue afforded in the'case of the nebulse'by the frequency separations of a 

 doublet or triplet. There is, however, no difficulty in finding a conjectural 

 place for it in the scheme of the oxygen arc spectrum as given by Hund's 

 theory. This theory, which may be regarded as a generalisation of all 

 our knowledge of line spectra, affords a kind of frame into which we may 

 confidently hope to fit new empirical knowledge as it accumulates. 



McLennan, arguing from the fact that nitrogen bands do not appear 

 in the spectrum of the night sky, which, however, shows the green line. 



