THE SECRETION OF THE URINE 493 



injection of urea, alone or in combination with dextrose, phlorhizin, 

 or di-sodium hydrogen phosphate (Na 2 HPO 4 ). In all the cases the 

 urine contained urea, chlorides, and sulphates, and was acid to 

 phenolphthalein. In one case after injection of urea and dextrose, 

 and in another after urea and phlorhizin, the urine reduced Fehling's 

 solution, and therefore presumably contained dextrose (Beddard 

 and Bainbridge). When the frog's kidney is perfused in situ with 

 oxygenated salt solution a certain flow of urine takes place. Sub- 

 stitution of non- oxygenated saline markedly slows the flow (Cullis). 



Apparently, then, the tubules have the capacity to secrete prac- 

 tically all the constituents of urine, and when the flow of urine is 

 small, probably most of it comes from the tubules. When, as in 

 the diuresis produced by salt solutions, large quantities of water 

 and salts have to be rapidly excreted, the bulk of the liquid comes 

 from the glomeruli, but also by a process of secretion. 



Lindemann has endeavoured to exclude the glomeruli in mam- 

 mals by injecting oil through the renal artery. After a short time, 

 according to him, the oil emboli clear away from practically all 

 parts of the kidney except the glomeruli, which remain plugged. 

 If indigo-carmine be subsequently injected into the blood, it is not 

 only taken up from it by the embolized kidney as well as by a normal 

 one, but is excreted. The quantity of urine is much diminished, 

 and its specific gravity increased, but its composition is not essen- 

 tially altered. He infers that the tubules are in a high degree 

 independent of the glomeruli as an apparatus for the secretion of 

 urine. More conclusive observations have lately been reported in 

 which the tubules were eliminated by producing an artificial nephritis 

 in rabbits by the subcutaneous injection of sodium tartrate. Tar- 

 trates act almost specifically upon the tubules, causing no noticeable 

 effect upon the glomeruli. After the intravenous infusion of a 

 solution containing sodium chloride and urea during pronounced 

 tartrate nephritis, all the chlorine appears in the urine within forty- 

 eight hours, but little, if any, of the urea. In the light of the 

 histological findings, this is interpreted to mean that under normal 

 conditions chlorides and water are passed through the glomerular 

 mechanism, and urea through the convoluted tubules (Underbill, 

 Wells, and Goldschmidt). These results constitute a direct and 

 striking confirmation of the Bowman hypothesis. 



As regards our first two questions, we may conclude that there is 

 no good evidence that reabsorption of water or other constituents of the 

 urine in the renal tubules plays an important part in the preparation 

 of that secretion. Many facts favour the conclusion that the glomeruli 

 and the renal epithelium act as distinct, although, of course, mutually 

 supplementary mechanisms, the glomeruli separating the larger portion 

 of the water and salts, the epithelium the larger portion, if not the 

 whole, of the characteristic organic constituents. 



