Schmid's analysis of the rotational structure of a number of the ultraviolet C${\mathrm{O}}_2$ emission bands is discussed, and it is pointed out that in spite of uncertainties, some definite conclusions can be drawn, notably that the molecule in equilibrium is nearly linear, or probably strictly linear, in both initial and final states, that $B^{\prime }$ is nearly equal to $B^{\prime \prime }$, and that the values of both are approximately known (cf. Schmid). It is also shown that the bands are most probably of the type ${}_{}{}^1{}_{}{}^{}\Pi \to {}_{}{}^1{}_{}{}^{}\Pi$, but possibly ${}_{}{}^1{}_{}{}^{}\Sigma \to {}_{}{}^1{}_{}{}^{}\Sigma$ or (much less likely) ${}_{}{}^2{}_{}{}^{}\Sigma \to {}_{}{}^2{}_{}{}^{}\Sigma$ or some other type. If ${}_{}{}^2{}_{}{}^{}\Sigma \to {}_{}{}^2{}_{}{}^{}\Sigma$ the emitter must be C${\mathrm{O}}_2^{+}$, otherwise C${\mathrm{O}}_2$. It is proposed to designate by $\kappa$ the quantum number corresponding to the angular momentum of rotation of the carbon atom relative to the O atoms around the O-C-O axis. It is pointed out that in electronic bands, one expects predominantly $\Delta \kappa =0\mathrm{}$. This rule is then applied to possible interpretations of the band structures. Some suggestions are also made concerning the vibrational analysis. Evidence from the values of $B$ and ${\nu }_1$ is stated, which supports Smyth's interpretation of the $a$ and $c$ series as ${v_1}^{\prime }$ progressions. It is suggested that the isolated bands $\lambda \lambda 2896$, 2883 may be the (0,0) band of a ${}_{}{}^2{}_{}{}^{}\Pi \to {}_{}{}^2{}_{}{}^{}\Pi$ transition of C${\mathrm{O}}_2^{+}$ with $B^{\prime }$ and $B^{\prime \prime }$ almost equal.

• Received 17 September 1932

DOI:https://doi.org/10.1103/PhysRev.42.364