142 THE DESICCATION OF AFRICA. 



which may be covered with drifting sand, the " Ergs," or 

 they may be just bare stony ground, the " Hammada," or, 

 again, they may be covered with pebbles, when they are called 

 " Serirs." The deserts are in actual fact seamed by great 

 water-courses or wadys; some of these, like the Igharghar, 

 north of Tasili, can be traced as gorges through the hills and 

 as deep river-beds in the flats, though no water now flows 

 down their channels, and in places sand dunes entirely 

 obliterate all traces of them. Prehistoric stone implements for 

 the grinding of corn are found all over the western Sahara, 

 and the ruins of cities of stone-built houses occur, as at 

 Taodeni, north of Timbuktoo, where settled life is no longer 

 possible. River fish of the barbel family are found in pools 

 in the Tasili plateau, far from any river now flowing. All 

 these facts go to prove that the deserts are of recent origin 

 and were at no very distant date fair portions of the earth's 

 surface inhabited by man. Paradoxically enough, the very 

 fact that the desert is over large tracts covered with drifting 

 sand is proof, according to Captain Courbis, of underground 

 moisture, which, by capillary attraction, rises through the sand 

 and binds the grains together; were there no ground moisture 

 the sand would all blow away as in the stony deserts or 

 hammada. For the same reasons Foureau considered the 

 Ergs, or areas of drifting sand, to be the broad basins of a 

 former river system. 



The deserts and the peculiar courses of the rivers are 

 causally related, and we shall see evidence in the sequel to 

 show that the great African rivers originally ran through the 

 deserts, giving them the requisite moisture to make them 

 normal fertile regions, but that, owing to the diversion of these 

 rivers by others working backwards from the coast and 

 capturing the waters of the inland system, a large portion of 

 the continent has been bereft of its natural supply. 



Before we begin the detailed evidence of the capture of 

 the several great rivers, there are a few fundamental facts to 

 be mentioned with regard to the topography of Africa, which 

 perhaps are not generally recognised, but which will help us 

 considerably in understanding this question of desiccation. 



In the first place, Africa is divided into two halves, the 

 line of division being a great volcanic fissure commencing in 

 the Gulf of Guinea by the string of islands, Anobon, St. 

 Thomas, Price's Island and Fernando Po. Then inland, there 

 are the volcanic peaks of the Cameroons, Mt. Atlantica, south 

 of Yola, the Marra Hills in Darfur, and the volcanoes in the 

 Bayuda bend of the Nile, between Dongala and Khartoum. 

 The half of Africa north of this line lies for the most part 

 below 1,500 feet above sea-level and that south of this line 

 above 1,500 feet. 



