THE EXCRETION OF URINE 547 



very great, at least 500 liters per day, only about 13 per cent of the fluid contained in 

 the blood passing through the glomerulus would pass by filtration through the capsule of 

 Bowman. The fact that such a large amount of fluid would have to be reabsorbed from 

 the uriniferous tubules is a possible a priori criticism of the theory, but Cushney points 

 out that the amount each tubule would have to absorb per hour would be very small 

 (in his experiment on a cat amounting to less than 0.014 c.c. per hour). 



According to the modern view, there are therefore two fundamentally 

 different processes occurring' in the kidney; filtration in the capsule and 

 selective reabsorption in the tubules. It is important to consider some 

 of the evidence which is considered to indicate that each of these pro- 

 cesses occurs. 



The Filtration Process. The filtration of the protein-free blood fluid 

 through the renal capsule, like that through any other membrane, depends 

 on several factors. (1) There must be a difference in the pressure between 

 the blood and the urinary filtrate. In the laboratory the pressure used in 

 filtering is usually supplied by gravity, but in the case of the filtration of 

 the urine through the capsule the force is furnished by the pressure of blood 

 in the glomerular vessels. (2) The character of the filter determines what 

 substances shall pass. The renal capsule is a membrane normally im- 

 pervious to the proteins of the blood, but pervious to the other constitu- 

 ents. Under certain conditions it loses this character. (3) The char- 

 acter of the fluid determines how readily it will filter through the mem- 

 brane. If the fluid contains a substance which can not pass through the 

 filter and which exerts an osmotic pressure in opposition to the filtering 

 force, the rate of filtration as well as the amount filtered, will be reduced. 



If the capsule acts as a filter it should be possible to alter the rate of 

 urine excretion by varying any of the above mentioned factors, and 

 experimentally this is true. The factors can be varied in several ways. If 

 the blood pressure is raised by tying off several of the branches of the aorta, 

 the urine is appreciably increased, or if the blood pressure is lowered, as can 

 be done by cutting the spinal cord the amount of urine is decreased. In the 

 artificially perfused kidney, the fluid exuding from the ureter increases 

 as the pressure of the perfusion fluid is raised, and decreases as 

 the pressure is decreased. Whether changes in the pressure in the blood 

 are directly responsible for variations in the rate of urine excretion, or 

 whether they act indirectly by varying the rate of the bloodflow in the 

 kidneys, has been the subject of much debate. Probably both factors are 

 involved. Apparently, excretion can continue only as long as the col- 

 loids of the plasma are not notably increased, for, as the osmotic pressure 

 due to the indiffusible colloids rises, the pressure in the capillaries is no 

 longer able to oppose it. The same point has been shown by Star- 

 ling and his pupils, who found that the excretion of urine ceases when 

 the capillary pressure in the glomerulus fell below that exerted by 



