Some observations concerning "laminarization" in heated vertical tubes