In a normal man the volume of blood cleared of urea per minute by the kidneys averages about 75 cc. In a given subject the physiological variation attributable to functional elasticity of the normal kidney may be as much as it= 30 per cent of the mean urea clearance.2 (Meller, McIntosh, and Van Slyke, 1928.) In a dog the variation produced by dietary and other changes may be much greater. Rise or fall of blood urea in a given subject does not alter his urea clearance; excretion rate rises parallel with blood urea 'concentration, so that a minute's excretion continues to represent the urea content of the same volume of blood. This was demonstrated by urea administration to normal men by Addis and Drury (1923), and to both normal and nephritic men by MGller, McIntosh and Van Slyke (1928). Drury (1923) found it possible to increase the blood urea of rabbits 5O-fold without affecting the clearance. Data in the present paper show that the clearance values for dogs are not affected by lo-fold increase in the blood urea. When the secreting elements of the kidney are inactivated or destroyed 1 Preliminary reports of part of the experimental results have been published in abstract by Rhoads, Van Slyke, Hiller and Alving (1931), and by Van Slyke, Rhoads, Hiller, and Alving (1931). 2 The term "urea clearance" is used, in the sense defined by Mgller, McIntosh, and Van Slyke (1928), to indicate the number of cubic centimeters of blood, the urea content of which is excreted in one minute (see also Peters and Van Slyke, Quantitative Clinica Chemistry, Vol. i, p. 345). The normal variation of &30 per cent found in the clearance in man is considerably exceeded in the dog; Jolliffe and Smith (1931) have found that dogs on high protein diets may show clearances twice as great as on cracker-meal diets. When the urine volume falls below a certain limit, about 2 cc, per minute in men, defined by Austin, Stillman, and Van Slyke (1921) as the augmentation limit, the blood urea clearance ceases to be independent of the urine volume, and begins to decrease with decreasing urine output. The present paper, however, is concerned only with conditions in which the urine volume is above the augmentation limit, so that diurcsis does not act as a factor influencing urea excretion. 336