30 



HANDBOOK OF PHYSIOLOGY 



CIRCULATION I 



r is the radius of rotation in centimeters. But it is 

 obvious tliat r cannot be precisely defined for the cells 

 throughout their descent from the top to the bottom 

 of the centrifuge tubes. A close appro.ximation is the 

 distance measured from the center of rotation to the 

 middle of the final packed cell column. Precision could 

 undoubtedly be improved by determining the rela- 

 tionship between r as measured in this way and the 

 percentage of occluded plasma, applying to each 

 sample the correction derived from this relationship. 

 Most investigators prefer to use a centrifugal force 

 and duration sufficient to keep the average percentage 

 small, and to accept the modest error due to variations 



(177)- 



All the earlier studies, using hematocrits uncor- 

 rected for occluded plasma, overestimate cell volume 

 regardless of method employed. Since the error ap- 

 pears equally in all the measurements reported, how- 

 ever, the conclusions drawn from a comparison of 

 methods are perfectly valid. 



Although some increase in the mean corpuscular 

 volume may be expected with an increase in the CO2 

 content of blood, the difference between arterial and 

 venous red cells from this cause is hardly measurable. 

 An increase in Pcoj froni 5 to 700 mm Hg produces 

 only a 13 per cent increase in the volume of rabbit 

 cells, 9 per cent in the volume of human cells, and 7 

 per cent in that of ox and sheep cells (124). None of 

 these change measurably in size within the physiologi- 

 cal range of 40 to 60 ml of COo per 100 ml of blood 

 (124). It is thus immaterial whether the central circu- 

 lation is sampled from an artery or a vein. If changes 

 in the mean corpuscular volume are suspected, they 

 may be demonstrated by measuring the hemoglobin 

 content of packed cells, a procedure which is more 

 reliable than cell counting (191). 



The Error in Computing Cell Volume from Plasma 

 Volume and the Central Hematocrit 



Whipple and his colleagues (240) found in 1920 

 that when cell volume was calculated from a dye- 

 distribution value for plasma volume and the venous 

 hematocrit, dogs subjected to measured blood loss ap- 

 peared to have a reduction in circulating cells con- 

 siderably greater than the volume drawn. This calcu- 

 lation is obtained as 



CF* = PV X 



H. 



H, 



(3) 



by the same group of investigators revealed that Cr'j, 

 was approximately 20 per cent greater than the 

 volume of cells measured by hemoglobin washout, 

 and the suggestion was made that the discrepancy 

 might be due to the existence of cell-free plasma in the 

 small vessels (222). The over-reduction in CVh by 

 bleeding is readily demonstrated in splenectomized 

 dogs (223). When bleeding is done in intact dogs, 

 however, restoration of the central hematocrit by 

 cells from splenic stores often masks the error (173). 

 That the overreduction is due to a systematic error in 

 CVh is shown by adding cells to the circulation. There 

 is an excess increase in CV], comparable to the excess 

 decrease when cells are withdrawn (141). It is ap- 

 parent that a volume of cells, which is known to have 

 a certain value outside the circulation, is overesti- 

 mated as CF), when it becomes a circulating volume. 

 The discordant finding that cell volume as measured 

 by carbon monoxide agrees well with CV^ in normal, 

 splenectomized, and hemorrhaged dogs (201) is no 

 longer pertinent. It is now known that the carbon 

 monoxide method also overestimates cell volume 

 (200), agreement between the two invalid methods 

 being fortuitous. 



Definitive proof of the error in CI'/, was afforded by 

 the first measurement of cell volume as the distribu- 

 tion volume of injected Fe^'-tagged cells, in 1941. In 

 the dogs studied, the tagged cell distribution volume 

 averaged about 75 per cent of CI';, (104). Similar 

 comparative studies have since been done for other 

 species including man, using a variety of tagged-cell 

 methods (38, 80, 86, 153, 188, 194, 242). 



The Mean Circulatory Hematocrit 



When cell and plasma labels are injected simul- 

 taneously, and the virtual distribution volume is 

 calculated for each, the ratio cells: plasma may be 

 computed for the circulatory system as a whole. If 

 this is expressed as an hematocrit, which is customary, 

 it becomes the mean circulatory hematocrit, //„„ and 

 is calculated as 



H„ = 



CV 



CV + PV 



(4) 



where PV is the distribution volume of plasma label, 

 and He is the central hematocrit. Subsequent studies 



where CV is the distribution volume of cell label, and 

 PV that of plasma label. A large number of studies 

 have been reported on tiie relationship between //„ 

 and the central hematocrit He. If assurance could be 

 had that the relationship H„/Hc is constant, it would 

 be possible to use a single label, for either cells or 

 plasma, and to estimate the unlabeled compartment 



