A careful investigation of the deposits over a large area, however, shows this view to be untenable; for while in most of the isolated conical hills within the depression, the beds in question cap the lower sandstones and clays (No. 7) of the Cenomanian, on a portion of the edge of the western plateau they cap the bed of limestone (basal member of No. 6) which itself caps the lower sandstones and clays. (Plate I). There must therefore be an unconformable overlap below the two series. This is further borne out by the facts, first, that the ferruginous silicified grit has never been observed to pass under the limestone in the walls of the oasis, although found capping hills in close proximity; and secondly, that the limestone never occurs below the ferruginous silicified grit in the isolated conical hills. The former would happen if the deposit in question represented the top of the sandstones and clays, the latter if it represented the next bed above the limestone. Furthermore, these beds are found capping hills close to the oasis-wall, and occurring on exactly the same level as the limestone capping the latter, suggesting at first sight some sort of connection between the two; the beds are, however, so entirely different that it is not possible to imagine the one to be an altered condition of the other; moreover, if such were so, the gradual passage of limestone to the ferruginous beds should be visible, but such has never been observed. The only possible view of the origin of these beds, consistent with the above facts, is that they represent a far younger deposit than the strata on which they lie, a deposit formed in fact in a slight depression in the Eocene and Cretaceous rocks, long anterior to the time when erosion was carving out the area to its present form. The pisolitic character of the iron-ore of Jebel Horabi and the usually large amount of ferruginous material, as well as the general character of the beds, indicate shallow-water lacustrine deposition and precipitation. No organic remains have as yet been observed in these deposits. It should be mentioned here that it is frequently impossible to draw any sharp of line of demarcation between these deposits and the undoubted Cenomanian sandstones below, when they rest on the latter in the hills within the depression. The sandstones themselves are frequently ferruginous and limonitic in their upper layers. This, however, is easily explained on the supposition that there would have been considerable infiltration into these porous sandstones forming the lake-floor, with consequent deposition of ferruginous material.