48 COMPARATIVE PHYSIOLOGY 



tubule " has focussed an immense amount of research on the 

 attempt to define the role of these two structures in the process 

 by which non-volatile waste products are eliminated from the 

 body. Since the secretion includes practically all the crystal- 

 loidal constituents of the blood plasma, though these are 

 neither individually nor collectively present in the same con- 

 centration as that in which they occur in the blood, it is 

 attractive to explore the possibility that the whole process is 

 composed of two distinct phases, one consisting of the separa- 

 tion of a fluid identical in crystalloidal constitution with the 

 blood — a deproteinised plasma filtrate formed by exudation 

 from the capillary walls of the glomerulus, and a subsequent 

 process involving the specific activity of the glandular epi- 

 thelium of the tubule and resulting in the differential con- 

 centration of each of the crystalloidal constituents, either by 

 secretion or by reabsorption or both. 



Some colour is lent to this interpretation by a line of 

 experimentation which is possible owing to the peculiar 

 vascular arrangements which exist in the amphibian kidney, 

 where the glomerular blood supply is derived from the aorta 

 (via the renal arteries), while the renal portal veins only supply 

 the tubules. It is thus possible in the frog, to cut off the 

 glomerular blood supply while leaving intact that of the tubules, 

 or to perfuse separately the vessels of the capsules and renal 

 tubules. 



When the glomerular blood-supply is cut off the secretion 

 of urine stops. It is still possible, by injection of a solution of 

 urea, to induce a flow of acid urine containing chlorides, 

 sulphates, and urea, but the quantity is small compared with 

 the diuresis produced by the same method in the intact kidney. 

 Perfusion of the arterial (glomerular) supply of the kidney 

 with Ringers' solution at normal aortic pressure (20-24 cm. 

 in the frog) induces a copious flow of urine (of somewhat 

 more dilute concentration than the perfusion fluid). Perfusion 

 of the renal portal system at normal venous pressure does not 

 result in production of urine. 



Further support for the belief that the initial stage in the 

 process of renal secretion is the separation of a deproteinised 



