Outside the narrow group of elements from beryllium to fluorine, the assumption of an approximately common radial function to be multiplied by a sum of angular iunctions belonging to different shells is not at all plausible. The apparent paradox in the behaviour of anti-bonding d-like orbitals in transition group complexes and the intensities of absorption bands (which are unexpectedly high in almost quadratic chromophores formed by copper [II] and palladium [Ill] are discussed and compared with simpler diatomic molecules.