EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS 



977 



terms of the empirical equation 3.4 describing the 

 observed osmotic pressure-concentration curve for 

 albumin under the stated conditions (fig. 3.1). This 

 close correspondence between observed osmotic pres- 

 sure and the pressure calculated on the basis of 

 Donnan theory might lead one to believe that the 

 problem has been solved. Unfortunately, this is not 

 the case. The correspondence between theory and 

 fact only occurs at two particular values of charge, 

 one of which occurs (fortuitously) at physiological pH 

 and salt concentration. Figure 3.5 shows clearly the 

 discrepancy between observed osmotic pressure and 

 the theoretical pressure predicted on the basis of 

 Donnan theory and van't Hofl's law. This dis- 

 crepancy may be explained in part by the binding of 

 chloride ions to the albumin molecule (86, 315). 

 Binding of chloride (or other anions) to the protein 

 decreases the fraction of salt free to take part in the 

 Donnan distribution and at the same time causes an 

 error in the calculation of net charge from titration 

 data. For this reason, also, the isoelectric point differs 

 from the isoionic point (86). When corrections are 

 made for chloride binding it appears that the true 

 isoelectric point for albumin is closer to pH 4.2 than 

 to pH 5.4 as might be inferred from its titration curve. 

 As shown in figure 3.5, the osmotic pressure does ap- 

 proach the van't Hoff pressure at pH 4.2 as would be 



40 - TT, mm Hg 



35 



30 



25 



20 



15- 



+ 30 +20 +10 



Van't Hoff's Law 



-10 -20 Charge, Z 



4 2 4.5 4.7 5.4 6.8 8.0 pH 



OSMOTIC PRESSURE OF 6 % ALBUMIN 

 In .15 M NaCI (m s = .l5) at 37 C 



fig. 3.5. Illustrating the discrepancy between osmotic 

 pressure calculated on the basis of Donnan theory and protein 

 osmotic pressure measured experimentally. Agreement with 

 theory occurs fortuitously at pH 4.7 and pH 7.6. [Calculated 

 from data in reference (313).] 



predicted from the Donnan relation (equation 3.6) 

 for the case z = o. 



Even after corrections have been made for chloride 

 binding, however, the Donnan theory fails to account 

 quantitatively for observed osmotic pressure under 

 most circumstances. Other factors involved include 

 electrostatic interactions between protein molecules 

 and between protein and salt. These interactions are 

 presumably represented in the third term of the em- 

 pirical equations (e.g., equation 3.4) describing 

 osmotic pressure-concentration curves, but no precise 

 theoretical explanation of these forces can be offered 

 at the present time. 



4. INTERSTITIAL FLUID PRESSURE 



("tissue pressure"), P,, 



Capillary blood pressure is only one point on the 

 transmural gradient of hydrostatic pressure which is 

 concerned with the filtration of fluid through the 

 capillary wall. The other point is the pressure on the 

 interstitial fluid outside the capillary wall. Landerer 

 (197) in 1884 attempted the first direct measurements 

 of interstitial fluid pressure by introducing a fine 

 cannula or needle into the subcutaneous or cutaneous 

 tissues and recording the pressures required to cause 

 fluid to flow into the tissues. Landerer's figures, like 

 those of Hajen (139) and others (158, 252) proved to 

 be fallacious because too much fluid was injected be- 

 fore pressure was measured (238). Nevertheless these 

 experiments were significant because they showed that 

 considerable resistance must be overcome if the 

 volume of interstitial fluid is increased suddenly. 



Even when the volume of interstitial fluid was in- 

 creased more gradually and physiologically, evidence 

 of mounting interstitial fluid pressure appeared almost 

 at once, at least in some tissues. Drury & Jones (80) 

 used an ordinary plethysmograph to measure rates 

 of filtration in the leg during venous congestion, and 

 found this rate declined as filtration progressed, 

 though their figures were irregular because their 

 method measured changes of vascular volume and of 

 interstitial fluid volume together. A ''pressure plethys- 

 mograph" (188, 209) made it possible to exclude 

 changes of vascular volume and thereby to measure 

 changes of interstitial fluid volume more accurately. 

 Landis & Gibbon (209) found in the human forearm 

 that the filtration produced by given venous pressures 

 declined rapidly, beginning even during the first 10 

 min of filtration. With a venous pressure of 20 cm of 

 water, filtration ceased in 35 to 55 min after the 

 accumulation of less than 1 ml of filtrate per 100 ml 



