COERELATIVES OF WATER CONTENT AND EXCHANGES 257 



tures a different series (Meek and Eyster, '21; Blalock, '27; Gesell 

 et al., '30) . With more connecting arrows, and tipped at both ends, 

 a description of interrelations results that may help to visualize the 

 number of relations present, for it suggests the equal importance 

 of many coordinated variables. In the absence of criteria by which 

 to decide which of these are "directly" related to the restoration 

 of blood volume, all appear equally concerned in what goes on. 

 Changes in "blood" pressure are often believed to have a special 

 relation to changes of "blood" volume, since it is a "hydrostatic" 

 pressure having the dimensions of a force that may move fluids. 

 That the "capillary blood" pressure minus mean extravascular 

 pressure may ordinarily balance the effective "osmotic" pressure 

 between blood and extravascular spaces (Starling, 1896) has been 

 a useful hypothesis. Adding inferences, it is supposed that physi- 

 ologically effective gradients of osmotic pressure are modified (in 

 parallel with mean capillary blood pressure) and at balance always 

 equal the "colloid osmotic" pressure as arbitrarily measured in 

 vitro. 



A quantitative relation between capillary blood pressure and 

 rate of water exchange obtains for single minute blood vessels in 

 data of Landis ('27). How to fit the measured net "colloid os- 

 motic" pressures with the rates of passage of fluid is unknown; 

 it is believed and hoped that the net rate will be zero when this 

 pressure equals the net capillary blood pressure. These "forces" 

 might even be plotted in a diagram that is analogous to an equili- 

 bration diagram, showing forces in place of rates. But its descrip- 

 tive value, showing how blood volume is equilibrated in an average 

 capillary, is little at present, for the inferences involved in its con- 

 struction seem to me more presumptuous than when rates of fluid 

 exchange are correlated with increment of blood volume. 



(3) While the hypothesis of Starling is conceived as applying 

 to the maintenance of blood volume, the adjustment of tissue cell 

 volumes is quite generally supposed to occur at rates proportional 

 to the difference of ^Hotal osmotic" pressure between inside and 

 outside. No assurance is at hand that either all this or only this 

 force is at work. 



Within extravascular (interstitial) compartments it is com- 

 monly believed that effective "osmotic" pressure and effective 

 "hydrostatic" pressure are both lower than inside blood vessels. 

 "Capillarity" and other "surface" forces have occasionally been 



