Some observations on the role of oxygen in lubricated wear