In the main text we've told a very simple story - that sedimentary rocks are laid down in continuous, parallel horizontal layers. But it's actually more complicated than that. Consider a scree slope. The layers of debris making up the slope are laid down on the surface of the scree - inclined at tens of degrees down the slope. So a sequence of sedimentary rocks deposited as scree may contain layering that was inclined at the outset. It would be said to have a depositional dip.
Scree deposits are rather rare in the geological record. But other depositional environments can also break the rule of sediments accumulating horizontally. Take a sand dune. With each blow of wind sand grains are dislodged down the slope of the dune, forming layers that are inclined - another example of a depositional dip. The same thing can happen under water (sub-aqueous dunes). As the dune continues to form and advance sand from the top is moved by down the slope, eroding the old top of the dune slope. So over time a sedimentary deposit is created with inclined surfaces bounded on the top by an erosion surface. We can recognise this in the rock record as cross-bedding. The erosion surfaces represent the principal bedding planes and these are continuous. In contrast the planes of cross bedding terminate at the bedding planes.
The imaginary scree slope and dune we've used so far create sedimentary successions with depositional dips on the scale of each of these landforms. But on much longer length-scales we can also take issue with the notion of bedding as being originally flat. Imagine sediments accumulating today across a continental margin, say on the Atlantic edge of Africa. The sea bed is not horizontal - it has shallow areas (the shelf and the abyssal plain) and an inclined portion (the continental slope). So a batch of sediment raining down on this sea bed, perhaps blanketing the sea floor will pick up the geometry of the sea bed itself. In other words - our imaginary bed will have a complex shape moulded onto its substrate. And there's far more to this too - but we can resist opening the Pandora's box of stratigraphic complexity any further for now!
Our simple assumptions about bedding look to be seriously flawed - but fear not. Cross-bedding can be readily distinguished in the rock record as can scree slopes (characterised by large angular blocks). And it's not often that we are concerned with understanding the structural geology of a deformed continental margin all in one go. At the local scale the assumption of bedding being originally flat is a fair starting point.