Before leaving the subject of water-supply we must briefly refer to two or three remarkable instances where water is found at very high elevations. The most important of these is on the upper road between Kharga and Dakhla, where a spring, known as Ain Amûr, occurs near the summit of the plateau at about 460 metres above sea-level. The water, which is quite potable, occupies the bottom of a hole about 3 or 4 metres in depth, at the base of a clump of palm-stubs, a few paces to the south of a solitary tree; it does not run, and the water-level is said to fall considerably in the summer. A small patch of green rushes lies a few paces to the west, and there is a good deal of scrub in the neighbourhood, from the position of which one is led to infer that there are several small springs thrown out along one of the bedding-planes, at the summit of the marly and clayey beds which underlie the limestones forming the uppermost portion of the cliff. The high-level spring on the eastern wall of the oasis, in the neighbourhood of Beris, consists of a pool of clear sweet water at the base of a large fallen block of limestone, in a desolate rocky dingle. Ball determined its height as being 180 metres above Beris, or 260 metres above sea-level. The pool itself is overgrown with weeds, and, scattered about on the sides and bottom of the valley, there is a good deal of vegetation, mostly in the form of coarse grasses, prickly scrub, and tamarisk bushes. Geologically, the spring appears to be similarly situated to Ain Amûr—that is to say, it emerges near, or at the summit of, the argillaceous Exogyra Series, and below the overlying Danian limestones. The water obtainable at Nakhail, 60 kilometres S.S.E. of the oasis, also appears near the summit of the Exogyra Series, and may therefore be considered to have a similar origin to the springs just described. These elevated occurrences, thrown out along more or less definite geological horizons, must be regarded as natural springs, quite distinct from the artificially-made wells of the oasis-floor. As the springs lie several hundred feet above the static head of the artesian water, and are separated from the water-bearing sandstones by a great thickness of argillaceous impermeable strata, it is perhaps permissible to assume that their waters are derived from an entirely different source. The occurrences known are of very limited number, and the water only appears in small quantities, so that it is not unlikely that it is derived from the very occasional rains which fall on the plateaux. A portion of these rains would doubtless find its way downwards through fissures in the limestone, and, in areas where the dip of the beds was towards the oasis, might travel underground and occasionally be thrown out as springs on the walls of the depression, at the junction of the limestones and underlying clays. CHAPTER XII THE ANCIENT SUBTERRANEAN AQUEDUCTS