In [1], the author proposed a "rank sum" criterion for testing whether or not a sample was drawn from a population having a completely specified continuous distribution. This criterion under the null ...