Key points are not available for this paper at this time.
It is shown that the atmospheric oxygen absorption bands can be attributed to a ^3S^1S transition from the normal (^3S) to a metastable ^1S excited state of O₂. This accounts for all the strong lines, and explains missing lines, without conflict with existing theory. Certain very weak series such as the A^' band are, however, not yet explained. Of the three rotational levels for each value of j₊ in the ^3S normal state, the two for which j=j₊1 show only a very small separation, which increases slowly with j₊, while the third is separated from the other two by an interval of about 2 wavenumbers which does not change with j₊ (cf. Fig. 2 and Table I). The ^3S and ^1S states involved in the atmospheric bands may perhaps be attributed both to the same electron configuration, in agreement with a suggestion made in a previous paper. If this is the case, it is likely that a metastable ^1D state derived from the same configuration also exists, and that infra-red atmospheric bands corresponding to the transition ^3S^1D should be found.
Robert S. Mulliken (1928) studied this question.