Abstract
It is pointed out that the Seitz model may be interpreted as a molecular‐orbital model. This enables us to avoid some difficulties encountered in the interpretation within a Heitler‐London framework. It is demonstrated that the observed positions of the A, B, and C absorption bands and the observed ratio of the dipole strength of the C band to that of the A band are consistent with each other in the molecular orbital model for KCl:Tl, KI:Tl, and also for KCl:In with some reservation.