THE EXCRETION OF URINE 387 



capsule must play an extremely important part in the secretion of urine. A 

 theoretical account of the kidney function must assign some purpose also for 

 the tortuous course of the tubules up to the point where they enter the col- 

 lecting tubules. 



This has been done in the view advanced by Ludwig and supported by 

 many experiments, namely that a process of filtration takes place from the 

 glomerulus into the capsule, and that the nitrate represents a very dilute 

 urine, which during its passage through the tubules becomes gradually con- 

 centrated by transfusion of water into the lymph bathing their outer surface. 



To test this view we have first of all to form some conception of the physico- 

 chemical processes necessary for the filtration of liquid through the capsular 

 epithelium. As Tammann was the first to show, the latter cannot be regarded 

 as a semipermeable membrane, for, if it were, a blood pressure sufficient to over- 

 come the osmotic pressure of the plasma several atmospheres would be required 

 to force the filtrate through. On this account Tammann considers the epithelium 

 completely permeable to all the crystalloids dissolved in the urine, and imper- 

 meable only to the colloids. In order to separate a proteid-free filtrate of the 

 composition of the crystalloids found in the plasma, the blood pressure need 

 only be high enough to overcome the osmotic pressure of the colloids occurring 

 there. The latter according to Starling amounts to about 25-30 mm. Hg. 



However, the osmotic pressure of the sugar in the blood (more than 100 

 mm. Hg.) is not taken into account in this calculation; whereas it may be 

 considered that the capsular epithelium is permeable to sugar just as to the 

 other crystalloids. It is possible too that the sugar is not free, but occurs in 

 the blood in chemical composition with other substances such as lecithin or 

 proteid. If this were true the osmotic pressure occasioned by the sugar would 

 of course be considerably lower than if it were dissolved as such in the plasma. 



The lowest pressure in the glomeruli at which a production of urine could 

 take place would thus be about 25-30 mm. Hg. And yet Gottlieb and Magnus 

 have shown that under the influence of diuretic substances, separation of urine 

 can take place with a carotid pressure of only 6-9 mm. Hg. 



Besides this difficulty we have another just as little to be explained from 

 the standpoint of the filtration hypothesis, and that is this : if the supply of the 

 blood to the kidney be completely interrupted by clamping the renal artery 

 for a short time, say one and one-half minutes, the formation of urine stops 

 and is resumed again only after a considerable time (as much as forty-five 

 minutes). The properties of a perfectly inert filter could scarcely be changed 

 to such an extent by anemia of so short a duration. 



The second part of Ludwig's theory likewise meets with several difficulties. 

 If the glomerular filtrate as above assumed has the same composition and there- 

 fore the same osmotic pressure as the blood plasma (with the exception of the 

 colloids) then only the osmotic pressure occasioned by the colloids can have 

 anything to do with the concentration of the filtrate. Whether or not this force 

 is sufficient to produce the necessary transfusion of water back into the lymph 

 has not yet been decided. Moreover, the proportion of crystalloids in the urine 

 is quite different from that in the blood. It requires more than the mere absorp- 

 tion of water to get a fluid with the properties of the urine out of such a filtrate 

 as we are able to suppose this to be; we must assume an unequal absorption of 

 different constituents. Finally, it would be impossible for this fluid to be ab- 

 sorbed back into the lymph by purely osmotic processes, for the osmotic pressure 

 of the urine as a rule is higher than that of the blood plasma. 



