On the other hand it has been known since the work of Fahraeus and Lindqvist
(4) that the apparent viscosity of blood decreases as the radius of the vessel
decreases. Thus the pressure drop in the capillary circulation depends jointly upon two
quantities, namely the cross-sectional area of the capillary bed and the apparent
viscosity of blood. Landis (5) postulated that the apparent viscosity is greater in
the capillaries than in the arteries, whereas Fahraeus and Lindqvist (4) postulated
a smaller apparent viscosity for blood in capillaries than in arterioles. No theoretical
basis for the postulate of Landis has been advanced. But for the opposite postulate
of Fahraeus and Lindqvist, in so far as large tubes (i.e. wide enough for several
cells abreast) are concerned, at least two theories have been advanced (6); one
theory attributes the decreased viscosity to the fact that the "integration" in the
derivation of Poiseuille's law should be replaced by a "summation," the other attributes
it to the formation of a cell-free zone near the wall. Evidently neither of
these theories is applicable to flow through those capillaries which permit red cells
to pass only in single file. There is abundant experimental evidence for the decrease
in apparent viscosity of blood as the radius decreases, but only down to radii of
about 25 microns. However, Bayliss (7) has reported a few measurements made in
tubes of about 10 microns diameter.