Some Remarks on Herbach's Paper, "Optimum Nature of the F-Test for Model II in the Balanced Case"