The authors discuss the direct capture theory pertaining to primary electric dipole (E1) transitions following slow neutron capture. For light nuclides that they have studied (including 9Be, 12C, 13C, 24Mg, 25Mg, 26Mg, 32S, 33S, 34S, 40Ca, and 44Ca), estimates of direct capture cross sections using optical model potentials with physically realistic parameters are in reasonable agreement with the data. Minor disagreements that exist are consistent with extrapolations to light nuclides of generally accepted formulations of compound nucleus capture. They also discuss the channel capture approximation which is, in general, a good representation of these cross sections in heavier nuclei, particularly if the scattering lengths are not different from the corresponding potential radii. They also draw attention to cases where the use of this formula leads to inaccurate predictions.