The investigation here presented had for its object the demonstration of the relative importance of vascular injury, renal injury and hydremia in the production of edema, and was suggested by a recent theoretical consideration1 of the part played by chemical correlation in the pathologic conditions associated with disturbances of renal function.
Experimental evidence indicating the importance of hydremia or of vascular injury is not lacking, but, as it is based mainly on transfusion experiments, in which large amounts of fluid were used, and frequently experiments on dead or nephrectomized animals, it is not entirely satisfactory in that the conditions are too artificial. The recent observations on the edema of uranium nephritis point the way to a method of study which affords conditions more nearly in keeping with those associated with edema in man.
Uranium nephritis, in rabbits at least, is accompanied, when an excess of water is