It is shown in the paper that the two fundamental assumptions (that of equality of particle and fluid diffusivity and that of purely inertial coasting within the viscous sublayer) on which existing deposition models are based, are untenable under most actual conditions. The concept of Stokes stopping distance, in particular, is shown to be invalid, since the effect of shear-flow-induced transverse lift force, which heretofore has been disregarded, is not negligible when considering the passage of a dense particle through the viscous sublayer. Due to the action of this lift force much lower radial velocities are required at the edge of the sublayer to insure particle deposition on the wall than would be the case if Stokes drag were the only force present. This explains why deposition models based on Stokes stopping distance concept must resort to the use of unrealistically high radial velocities within the sublayer to insure agreement with experimental data.