Remarks on the temperature dependence of the optical Friedericksz transition