ON CONSISTENCY OF BAYES PROCEDURES.