Abstract
It is shown that theoretical calculations which use the exponential model μ=μ0(rr)×exp(rρ) for the variation of the dipole moment μ with internuclear separation r can be forced into agreement with experiment only if one uses dipole parameters μ0, ρ which are in serious disagreement with values computed by the molecular Hartree-Fock approximation. On the basis of a Drude model calculation, it is concluded that the exponential model is an inadequate representation of the true physical situation because of its neglect of electron correlation effects.