LYMPH ABSORPTION 609 



of the capillary blood pressure. Direct measurements gave 

 40 mm. H 2 for the liver and 140 mm. H 2 for the skin. If the 

 capillary pressure fell considerably below that of the tissue spaces, 

 backward nitration might theoretically take place for a time. If, 

 however, this relative increase of the extravascular above the 

 intracapillary pressure were communicated to the outside of the 

 veins, they would be strangled, and the circulation through the 

 part cease. Starling has pointed out that this result might not 

 ensue, if the vessels were bound to the surrounding tissues by 

 jadiating fibres whose pull would tend to keep the vessels patent. 

 He found that the injection of normal salt solution under high 

 pressure into the subcutaneous tissue of a dog's leg greatly 

 decreased the venous outflow. The same was true of the sub- 

 maxillary gland and tongue. In these regions, at any rate, it 

 would appear to be impossible for backward filtration to take 

 place. 



Osmosis with Diffusion. Starling has shown that blood serum 

 as against a non-proteid salt solution, otherwise isotonic with it, 

 possesses an osmotic pressure of about 3 mm. Hg when separated 

 by gelatine between two layers of peritoneal membrane. He 

 supposes that the capillary wall, like such a dead membrane, is 

 more or less impermeable to proteids, that there is in consequence 

 a less concentration of proteid outside than inside the capillary, 

 and that the osmotic pressure of proteids, although trifling when 

 compared with that of crystalloids, can attract water from the 

 tissue spaces into the circulation. 



The absorption of an isotonic salt solution from the sub- 

 cutaneous tissue or serous cavity would take place according to 

 this view in the following way. Water would be attracted into 

 the capillaries by the proteids, and as this would raise the 

 concentration of the diffusible substances in the solution above 

 that in the blood, they would diffuse into the blood until the 

 fluid was again isotonic. Then osmosis would again come into 

 play, and alternate with diffusion until the whole fluid had been 

 absorbed. It has been shown by experiment that, when an 

 isotonic solution containing more proteid than the plasma is 

 introduced into the peritoneal cavity, an isotonic fluid passes from 

 the blood into the cavity until the concentration of proteid in 

 the cavity has been reduced to that of the plasma. According to 

 Starling, therefore, while " capillary blood pressure determines 



2Q 



