Abstract
A comparison is made of the recent systematic solution of the mean-field equations for liquid crystals by Shen et al. with that of McMillan. It is shown that the discrepancies in McMillan's phase diagram are due to an undesirable chain-length factor in his potential and not due to the truncation of his Fourier expansion of the spatial potential function. The rectified phase diagram is shown to be identical to that of Lee et al.