where D is grain diameter, u and p are sediment and fluid densities respectively, g is gravitational acceleration, tan a is Bagnold's dynamic friction coefficient, V, is shear velocity, U, is mean forward speed of the bedload grain, and C is a drag coefficient. The term a U, is the flow velocity at a distance of one to two grain diameters above the bed, that is at the mean level of the (saltating) grain. Engelund and Fredsoe (1976) take a between 6 and 10, whereas Bagnold (1956) takes a value of 8.5. The weight force could be effectively corrected for a small lift force by changing the value of C, however Bagnold's (1974) results would deem this unnecessary. The downslope gravity component on the grain is justifiably assumed negligible.The resulting expression for mean grain velocity is where 8 is the dimensionles bed shear stress, and 8, = 4 tan a/3a2C. 8, and U,, must of course be the values of 8 and U, at the threshold of grain motion because when .8 = 8, or U, = U,,, then U, = 0. Equation (2) has been shown to fit experimental data very well if a lies between 9 and 10, and observed values of 8, are used. Abbott and Francis (1977) also support equation (2) but take a as about 14. They also