A re-determination of equilibria in the system MgO-H 2 O and comments on earlier work