The treatment of nephrosis has been especially interesting to clinicians because in this disease edema is usually found without complicating cardiac or arterial lesions. Comparatively recently it has been realized that in nephrosis the kidney need not be severely or irreparably damaged and that, even in the stage of greatest disturbance in water excretion, tests for renal function may give approximately normal results. Such observations have stimulated search for therapeutic agents acting elsewhere than on the kidneys in patients with this condition. Thus, on the hypothesis that nephrosis might be due to a metabolic disturbance involving the thyroid gland, thyroxine and thyroid extract have been used, with results which at times have appeared promising. Likewise, because it was conceived that a sort of "diabetes albuminuricus" might explain the phenomena of nephrosis, high protein diets were tried, with enough success to warrant their use in most clinics. The theory has been