Abstract
In rabbits, after rapid intravenous injection of Mg28, the plasma and urinary specific activities, corrected for transit time, are at all times in equilibrium. Injection of Mg28 during stop flow does not reveal any tubular zone permeable to magnesium. On the other hand, the magnesium behavior is in many ways comparable to that of a glomerular substance. These three arguments lead to the view that urinary magnesium is of essentially glomerular origin, and rule out any mechanism of tubular exchange or secretion.