Here we found the small Kél-owí caravan encamped in the midst of a valley well covered with herbage, near the well Karáda. Our new companions were perfect specimens of the mixed Berber and Sudán blood, and, notwithstanding all their faults, most useful as guides. It was two hours after midnight when we arrived, and after a short repose we started again tolerably early the next morning. For the first hour we kept along the valley, when we began to ascend a narrow path winding round the slope of a steep promontory of the plateau. The ruins of a castle at the bottom of the valley formed an object of attraction. The ascent took us almost an hour, when the defile opened to a sort of plateau, with higher ground and cones to the left. After another ascent four miles further on, over a rocky slope about 180 feet high and covered with sand, we encamped at an early hour, as the heat was beginning to be felt, in a valley with sidr-trees and grass, called Erázar-n-Ákeru. A large basin of water, formed by the rains in a small rocky lateral glen joining the large valley on the west side, afforded a delightful resting-place to the weary traveller. The basin, in which the negro slaves of our Kél-owí swam about with immense delight, was about 200 feet long and 120 feet broad, and very deep, having been hollowed out in the rocks by the violent floods descending occasionally from the heights above. But on a terrace about 200 feet higher up the cliffs I discovered another basin, of not more than about half the diameter of the former, but likewise of great depth. All along the rocky slope between these two basins, cascades are formed during heavy rains, which must render this a delightfully refreshing spot. We soon emerged from the valley, and entered a district of very irregular character, but affording herbage enough for temporary settlements or encampments of the Imghád, whose asses and goats testified that the country was not quite uninhabited. Some people of our caravan saw the guardians of these animals—negroes, clad in leather aprons. Against the lower part of the cliffs, which rise abruptly on all sides, large masses of sand have accumulated, which, as in the case of the upper valley of the Nile, might induce the observer to believe that all the higher level was covered with sand, which from thence had been driven down; but this is not by any means the case.