498 JOURNEY TO CENTRAL AFRICA.

side pillars with a stick. It was repeated three times with an interval between, and was so clear and distinct that I imagined it to be Achmet, following me. I called, but on receiving no answer, went out, and was not a little surprised to find no person there or within sight. The temple stands at a considerable distance from any dwelling, and there is no place in the smooth sands on all sides of it where a man could hide.

When I- mentioned this circumstance, on returning to the boat, Achmet and the rais immediately declared it to be the work of a djin, or afrite, who frequently are heard among the ruins, and were greatly shocked when I refused to accept this explanation. I record the circumstance to show that even iu the heart of Nubia there are mysterious rappings.

Beyond Derr I entered the mountain region of granite,, sandstone and porphyry, which extends all the way to Assouan. As I approached Korosko, which is only about twelve miles farther, the south-wind increased till it became a genuine khamseen, almost blotting out the landscape with the clouds of sand which it whirled from the recesses of the Biban. We were obliged to creep along under the bank till we reached Korosko, where we ran up to the same old landing-place at which I had stopped in December. The bank was eight feet higher than then, the river having fallen that much in the mean time. There was the same house, open on the riverside, the same old Turk sitting within, the dark sycamores shading the bank, the dusty terrace with the familiar palms tossing their leaves against the wind, the water-mill, the white minaret at the foot of the mountain, and, lastly, the bold, peaked ridge of Djebel Korosko behind. There was the very spot where my tent had stood, and where I first mounted a