In a previous paper the evidence on the excretion of carbohydrates by the toadfish, goosefish, dogfish, dog and man was reviewed and it was concluded that, until evidence to the contrary is adduced, all the available facts are best interpreted on the supposition that pure carbohydrates in general are not secreted by the renal tubules of any of the above-mentioned forms (Shannon and Smith, 1935).The evidence on creatinine points in the contrary direction. Although this substance is not normally present in significant quantities in the urine of aglomerular fishes, administered creatinine is copiously excreted by the toadfish and the goosefish (Marshall, 1930(Marshall, , 1934. Marshall and Grafflin (1932) have shown in the toadfish that the quantity of creatinine excreted per unit time does not increase in direct proportion to the quantity injected into the body. This failure to obtain direct proportionality between the quantity of substance administered and the rate of its excretion was first described for any substance by Marshall and Crane (1924) in their examination of the excretion of phenol red by the dog, and these investigators pointed out that the observed curvilinear relationship between plasma level and rate of excretion was incompatible with excretion by exclusive filtration. A similar phenomenon has subsequently been shown by Bieter (1933) and Marshall and Grafflin (1932) to occur in the excretion of phenol red in the toadfish.An analysis of the simultaneous excretion of creatinine and xylose (or sucrose) in the dogfish by Shannon (1934a) has disclosed that the creatinine clearance at low plasma levels of this substance is from 4.2 to 7.2 times as large as is the simultaneous xylose clearance. As the plasma level is raised the creatinine clearance is depressed, both absolutely and relative to the clear-1 This paper is based on a thesis to be presented to the Graduate School of New York University in partial fulfillment of the requirements for the degree of Doctor of Philosophy.ance of the sugar, until at plasma levels of 140 mgm. per cent the creatinine/sugar clearance ratio is less than 2. Entirely similar results have been obtained more recently in a comparison of the creatinine clearance with the inulin clearance (Shannon, 1934b). Phlorizin lowers the creatinine clearance both absolutely and relative to the sugar clearance, as first described by Clarke and Smith (1932), who ascribed this result to a depressing action of this drug on the secretory power of the tubules.MacKay has shown in one individual that, under controlled conditions in man, the rate of excretion of creatinine is directly proportional to plasma concentrations up to 15 mgm. per cent; and Cope (1931) has obtained a similar result in another individual up to 8 mgm. per cent. We have not reexamined the matter within this range, but the above data, combined with those of Jolliffe and Chasis (1933) who worked between 2.4 to 11.6 mgm. per cent in one instance, and 1.8 to 7.0 mgm. per cent in a ,second, would seem to indicate that in ma...