On the physiological significance of the pre-ferred body temperatures of reptiles