It is not necessary to investigate as far as possible – as I did – on the basis of x-rays of the human structure. Four points of anatomical geometry will suffice to suggest that both the possible rotations of the arms around the shoulders and the legs pivoted on the pelvis identify four circles independent of the navel, two by two symmetrical but differently diverging from the central axes of the body.
They may be located at more or less unequal distances, according to various radiological samples or qualified schemes, but it is the circumference that must be adapted to these, and not the other way around.
In other words are the virtual circle arcs described by the physiologically possible rotations of the limbs, to delimit the maximum extension of the entire human body in every direction; and this is what must be contained by the circumference introduced by Vitruvius
In a basic scheme, we are facing two arcs centered on the hips, whose radius is determined by, and is evenly carrying the standing position lenght of the legs, not with the extended feet;
even if they overstep immediately the given circumference, at each leg abduction.