A previous communication described the peculiar motion of the plasma trapped between erythrocytes in a capillary (bolus flow). In this paper the effect of this motion on capillary resistance to flow, as well as on dissipative effects associated directly with the cells, are described. The resistance that would be associated with plasma in bolus flow at high Reynolds numbers (relative to a capillary value of 0.01) was studied in a model, in which air bubbles, separated by short segments of water, passed along a glass tube. The resistance to flow, especially with short boluses, was at least ten times greater than that associated with Poiseuille flow. In a second series of experiments at lower Reynolds numbers, a single bolus of liquid was forced by air pressure along a glass tube. In these latter experiments, which more closely simulate biological conditions, the mean resistance to flow was only 30 per cent greater than that associated with Poiseuille flow. In the final series of experiments human blood and plasma, diluted in acid-citrate dextrose (A.C.D.) in varying degrees, were forced through glass micropipettes of capillary dimensions. The mean apparent viscosity of whole blood was found to exceed that of plasma by only about 5 per cent, thus verifying a conjecture to this effect made by Fahraeus and Lindqvist in 1931.