THE DESICCATION OF AFRICA. 149 



invaded the drainage area of the river, and in the tributaries, 

 the Bafing, the Bakoy and another river of the name Baoule, 

 we see clearly by their courses that they once belonged to the 

 larger river. The Senegal River originally began m the 

 Tamboura Mountains, near Bafoulabe, but the short, straight 

 course, combined with the easily-weathered, horizontal sand- 

 stones of this area, allowed it to eat into and rob the Niger 

 of a considerable portion of its waters. As it is, the Senegal 

 River is steadily gaining ground in the natural basin of the 

 Tankisso, and at some future date this great tributary of the 

 Niger will be lost to it. 



The floods of the Niger extended more than a hundred 

 miles north of Timbuktoo within the memory of man, but 

 these inundations are far more restricted nowadays; this is 

 usually ascribed to the drying up of the continent, but it is 

 far more probably due to the fact that every year sees many 

 square miles of country which formerly contributed its waters 

 to the Niger, drained by the encroaching coastal rivers. The 

 Senegal is the chief thief on the west, and on the south there 

 are the Liberian rivers, the Sassandra, the Bandama and the 

 Comoe. Thus, waters which originally went to fertilise the 

 inland districts are now drained rapidly to the sea, and their 

 beneficial effects are lost. 



Some 300 miles above Timbuktoo the Niger enters a great 

 level plain which is liable to inundations, the waters at such 

 times forming a lake i/o miles long and 80 miles broad. At 

 ordinary times the waters are restricted to a reticulating 

 system of channels and one more or less permanent lake, Lake 

 Debo. A further system of shallow depressions extends 

 northwards, west of Timbuktoo, the largest being Lake 

 Faguibine and Lake Horo. All the wells dug north of 

 Timbuktoo as far as Arouan, 200 miles north of the Niger, 

 are in fiuviatile deposits, full of shells of Melania, Physis and 

 Planorbis, showing the former extension of the flood-lake 

 system. A vast region like this, with practically no fall, 

 indicates a temporary phase in the evolution of a river system; 

 the Niger below the Debo Swamp has not cleared its course 

 sufficiently to accommodate itself to the new conditions. 



At one time it was thought that round about Timbuktoo 

 there had been an inland sea like the Black Sea, which 

 communicated with the Atlantic through a channel in the 

 Adrar Highlands ; Chudeau even designated this channel as 

 the depression El Khat (lat., 19° N.). According to Lemoine, 

 however, all the evidence for such a sea lies solely in the 

 presence of countless shells of Margmella, which are found 

 strewn about in the soil round Timbuktoo. These sea-shells 

 were at one time used as money and were probably the 

 "cowries " which poor Mungo Park was presented with by the 



