The sulphates of aluminium and magnesium are widely distributed in the oasis, though the deposits are seldom of sufficient extent to repay extraction. They occur as thin bands in the shales and sandstones of the surface-water series, seldom exceeding a few centimetres in thickness, and being generally of very limited horizontal extent. The alum, which is often found in beautiful fibrous or hair-like crystalline masses, is frequently of a delicate rose tint, and analysis shows the colour to be due to the presence of small quantities of cobalt. Epsomite (Epsom salts or hydrous sulphate of magnesia) is also found in fibrous crystalline seams, a remarkably pure deposit, containing several hundred tons, having been quite recently located a few kilometres west of Headquarters. In the foot-hills to the north of Bellaida, at the base of the eastern slopes of Jebel Tarif, and in the hills immediately to the west of Qasr Lebekha, ancient mines of the most extensive description are to be seen. The rocks in these localities consist of variegated sandstones and grits, many of the beds being so dark and ferruginous that the hills, as a whole, have a blackened and almost volcanic appearance. They are literally honeycombed with ancient workings, many of which, both in the form of large chambers and narrow tunnels, penetrate for long distances underground. Huge dump-heaps mark the entrances of the workings, and bear witness to the enormous quantities of ‘country’ rock removed in the winning of the useful mineral. The rude shelters built by the miners are to be seen in the immediate vicinity, and in many cases are still intact. [Illustration: A WADI IN JEBEL TARIF.] [Illustration: A RIVER OF SAND NEAR UM EL DABADIB.] The extent and magnitude of the underground workings prove that whatever the mineral mined, it was a substance of considerable value in those days; and an examination of the blind terminations of the tunnels occasionally reveals the presence of very thin seams of aluminium sulphate, which, in the absence of indications of other useful minerals, we must conclude was the substance sought. The bands of alum were evidently followed until they thinned out to a fraction of an inch, or disappeared altogether.