Quantitative stratigraphic models have been generated using either geometric constructs or diffusion algorithms to simulate the development and timing of sequence boundaries and how they vary in response to tectonics, eustasy, climate, and sediment supply. While diffusion models provide a physical basis for the generation of clinoforms, the resulting model predictions appear to do little more than track the transgressive and regressive movements of the shoreline. On the other hand, geometric models do a relatively good job in simulating the observed stratigraphy but offer little in actually ascertaining the physics controlling the processes of deposition and erosion. All these stratigraphic models fail to predict the most fundamental response of depositional systems, namely, the development of a downward shift in coastal onlap during a fall in relative sea level. This basinward shift and the development of an onlap surface are the critical observations that define chronostratigraphic sequences and are not simply a consequence of shoreline transgressions and regressions. A major assumption in these two-dimensional models is that the observed stacking patterns are solely the consequence of deposition processes and sediment transport within the modelling section, in marked contrast to the observation that the major sediment transport direction along margins is sub-parallel to the margin. We present a variation on the diffusion theme in which across-margin transport and deposition are dominated by advection-type processes and along-margin erosion and deposition processes are dominated by diffusion. With this combination of advection and diffusion it is possible to generate a lowstand system tract and sequence boundaries (i.e., onlap surfaces). In particular, the generation of the lowstand system tract is, in large part, the result of depositional processes operative along the margin (i.e., out of the section).