Observations concerning the validity of the ventricular gradient concept