Some comments on domain structures in block copolymers