A Note on Optimal Partial Triallel Cross Designs