exposed, for if the horizontal line, V, L, Fig. 13, were carried back beyond the point _x_, to meet the prolongation of _b_, then these Lower Tertiary strata would not only be intersected by the line of valley level, but would form a much smaller angle with the plane V, L, and therefore spread over a larger area than where they crop out on the side of the hills. The foregoing are the three most general forms of outcrop, but occasionally the outcrop takes place wholly or partly on the summit of a hill, as, near the Reculvers in the neighbourhood of Canterbury, of Sittingbourne, and at the Addington Hills, near Croydon, in which cases the area of the Lower Tertiary is expanded. When the dip is very slight, and the beds nearly horizontal, the Lower Tertiary sands occasionally spread over a still larger extent of surface, as between Stoke Pogis, Burnham Common, and Beaconsfield, and in the case of the flat-topped hill, forming Blackheath and Bexley Heath, as in Fig. 14. Favourable as such districts might at first appear to be from the extent of their exposed surface, nevertheless they rarely contribute to the water supply of the wells sunk into the Lower Tertiary sands under London, the continuity of the strata being broken by intersecting valleys; thus the district last mentioned is bounded on the north by the valley of the Thames, on the west by that of Ravensbourne, and on the east by the valley of the Cray; consequently the rain-water, which has been absorbed by the very permeable strata on the intermediate higher ground, passes out on the sides of the hills, into the surface channels in the valleys, or into the chalk. Almost all the wells at Bexley Heath, for their supply of water, have, in fact, to be sunk into the chalk through the overlying 100 to 133 feet of sand and pebble beds, _b_. [Illustration: Section at Blackheath and Bexley Heath. Fig. 14.] Thus far we have considered this question, as if, in each instance, the outcropping edges of the water-bearing strata, _b_, were laid bare, and presented no impediment to the absorption of the rain-water falling immediately upon their surface, or passing on to it from some more impermeable deposits. But there is another consideration which influences materially the extent of the water supply. If the strata _b_ were always bare, we should have to consider their outcrop as an absorbent surface, of power varying according to the lithological character and dip of the strata only. But the outcropping edges of the strata do not commonly present bare and denuded surfaces. Thus a large extent of the country round London is more or less covered by beds of drift, which protect the outcropping beds of _b_, and turn off a portion of the water falling upon them.