Confirmation of an exception to the ‘‘general rule’’ of surface relaxations