VOLUME OF BLOOD 



25 



0.8 - 



••-- DYE CO. 1972 CC/MIN 

 MCT 15.45 SEC. 



-o- CELLS CO. 1938 CC/MIN 

 MCT 15.25 SEC. 



-L. 



SECONDS 

 U 



_1_ 



_1_ 



-L. 



-L. 



80 100 120 140 160 ISO 200 220 24C 

 FIG. I. Flow-dilution curves for P'^ cells and T-1824 injected simultaneously into right ventricle, 

 samples drawn from aorta at origin of brachiocephalic artery. Barbitalized dog. Periodic mixing 

 phase ends at about 60 sec (note change in time scale). Ordinates show percentage of injected label 

 present in i ml of whole blood at sampling site. A primary pulmonary transit curve and one re- 

 circulation curve are discernible for both labels, .■\lthough the mean circulation time (MCT) for 

 dye is 0.2 sec longer than for cells, the area of the primary curves for the two labels is nearly identical, 

 since they yield essentially the same value for cardiac output (CO.). No attempt has been made 

 to estimate disappearance slopes beyond the first minute, the horizontal lines being drawn through 

 the arithmetic means for the two labels in the aperiodic phase of mixing. 



label becomes gradually mixed in the circulation by 

 processes of flow fractionation into circuits with differ- 

 ent transit times, and recombination with blood 

 having different activity at circuit confluences. If 

 samples are drawn at suitably short intervals, label 

 concentration may be seen to rise to a peak as the 

 primary wave of label passes the sampling site, and 

 then to decline more slowly. Attenuated waves may 

 usually be observed on first recirculation, and some- 

 times on third or even fourth passage (fig. 1). Com- 

 plete disappearance of such waves in the central cir- 

 culation means only that there is no longer a front 

 of rising label concentration in the system consisting 

 of the heart and the most rapid circuits, and offers 

 no assurance that mixing is complete elsewhere. 

 Mixing is obviously incomplete even in the central 

 circulation during the periodic phase. It is not com- 

 plete in the peripheral circuits until they as a group 

 are returning label to the central circulation at the 

 same rate as it enters them. Until this happens, there 

 is a net loss of label from the central to the peripheral 

 circulation. Unfortunately, there is no simple way of 

 distinguishing such loss from loss to extraxascular 

 regions. 



Since samples drawn from the central circulation 

 show a declining label concentration for several min- 

 utes after permanently tagged cells are injected (106), 



there is no question that mixing contributes to the 

 fall observed when other labels are used. Mixing is 

 usuallv considered complete, for impermanent labels, 

 when the initial phase of rapidly falling concentration 

 is terminated, and a slower rate of disappearance is 

 begun (fig. 2). 



Correction for Loss of Label During Mixing Period 



All plasma labels continue to leave the central cir- 

 culation after circulatory mixing is complete, until 

 they finally disappear completely. Erlanger (64) was 

 the first to suggest, in 1921, that in the case of those 

 which disappear at a nearly constant rate, it might be 

 possible to extrapolate through the mixing period to 

 an intercept at the zero time ordinate. If the rate of 

 loss during mixing is the same as that observed later, 

 this intercept represents the concetitration which 

 would be produced by instantaneous mixing, with Q. 

 at its injected value. Most plasma labels disappear 

 at a nearly exponential rate after circulatory mixing. 

 It is accordingly customary to follow the procedure 

 introduced by Gibson & Evans (77), in which the 

 logarithm of label concentration is plotted against 

 time, for purposes of extrapolating. This method of 

 correcting for the unobserved loss during the mixing 

 period can obviously be only an approximation. If, as 



