ON PRILL'S PROBLEM