Symmetrical n ARY unequal block arrangements and optimal n ARY block designs for comparing test treatments with controls