Note on the relation between Kendall's tau and the Cronholm-Revusky test