Comments on a blood-bank inventory model of Pegels and Jelmert