Edrisi says that the navigation from Ghana to Tirka (which latter is in the way to Wangara, admitted by the same authority to lie to the east of Ghana[149]) is with the stream of the Niger:[150] and if this be true, it ought unquestionably to have the same direction all the way from Houssa.
To these notices, of which the most full and positive, is that of an intelligent person who had visited the spot; are to be opposed the reports of Edrisi and Abulfeda, who wrote at a distance, and from the information of others. As to Leo, although his declaration is in favour of the two Arabian geographers, yet his authority loses all its weight, by his saying that the river runs to the west, by Tombuctoo; a fact which, I presume, no one will be hardy enough to contend for. And it will be found, that his descriptions do most completely do away his declaration: so that his testimony is turned against himself by the very context. For after saying that it runs towards the kingdoms of Ginea and Melli, he says also that they lie to the west, in respect of Tombuctoo. Now the contrary has already been made apparent, in [page lxv,] et seq.; so that in fact, Leo’s descriptions go rather to prove, that the course of the Niger is to the east, than to the west. But after all, his descriptions are the result of hearsay, rather than of observation: and it is plain, that his idea of the course of the Niger, was regulated by the supposed situation of the countries it ran through. Nor had he in his mind the Coast of Guinea, according to our acceptation of the term, when he spoke of the country of Ginea: for in his description of Nigritia he says, that the sea on the south, was unknown[151] to him. Thus the testimonies appear to be clearly in favour of an easterly course of the Niger from Houssa to Wangara. I next proceed to the question respecting its termination.
3. Concerning the Termination of the Niger.
Mr. Beaufoy’s Moor farther says, that “below Ghinea, is the sea, into which the river of Tombuctoo disembogues itself.” This may therefore be considered as the prevailing idea at Houssa and Tombuctoo, at which places he had resided, altogether, about 12 years. By the word sea, it is well known, the Arabs mean to express a lake also; (and even sometimes a river.) Edrisi and others describe large lakes in Ghana and Wangara.[152] And when Leo says that the Niger falls into the sea which borders on Ginea, it is not improbable that the lakes of Ghana and Wangara are meant; and that he was under the same mistake here, in supposing Ginea to be in the neighbourhood of the sea, as in what relates to the position of Ginea itself. In other words, that hearing from the natives, that the Niger expanded itself into lakes below Ghana (or Ginea), he supposed the western ocean to be meant. For it appears (p. 2.) that he had heard, that the Niger had its source in the mountains on the west, and running thence to the east, expanded itself finally into a vast lake: but misled by the supposed situation of Ginea and Melli, he disregarded the information.
He also describes Ginea to be a country annually overflowed by the waters of the Niger, but omits to say the same of Wangara, to which the description more particularly applies. It may be, that as Wangara in more early times formed a part of the empire of Ghana (or Ginea), his ideas might have been collected from some history of those times. I therefore consider his description of Ginea (p. 248), to include both Ghana and Wangara.
Edrisi describes three large fresh water lakes in Wangara, and one in Ghana.[153] The description of Wangara appears to be that of an alluvial country, environed and intersected by the branches of the Niger, and annually overflowed in August. Perhaps August was the time of the highest flood: for Leo says that Ginea (apply this to Wangara, also) is overflowed in July, August, and September; which is indeed the season of swelling of the rivers of the tropical regions, generally.[154]
From this description may be inferred the very low level of the countries of Ghana and Wangara; which level or hollow forms a receptacle for the surplus waters of the Niger, collected during the rainy season:[155] and whose permanent lakes, apparently form receptacles for its waters, during the dry season also. The country of Wangara alone, is said by Edrisi and Ibn Al Wardi to have an extent of 300 miles by 150 (i.e. Arabic miles, of 56⅔ to a degree); and Edrisi’s statement of the distances through it, proves that its length lies in the same direction with the course of the Niger; that is, from west to east[156] Now I have no kind of difficulty in supposing that any river may be evaporated, provided it is spread out to a sufficient extent of surface: and it may be that the level, or hollow, of Wangara and part of Ghana, may present an extent of surface sufficient to produce this effect.[157] And hence these countries must be regarded as the sink of North Africa, at all seasons. No doubt the inhabitants are amply repaid by the fertility produced by the deposition of the waters: but besides this, in the southern quarter of Wangara, they collect an incredible quantity of gold sand, after the waters are gone off, which is carefully sought after, as soon as the rivers regain their beds.[158]
It may be proper to observe, that, according to the estimation which we ought to make, of the quantity of water collected into the Niger, it ought not to bear a proportion to that, collected into the great tropical rivers of Asia; since it receives no branches, but on one side. Of course, it does not drain so great a surface of country, as those which receive them on both sides. Moreover it drains only the tract situated to leeward of the great chain of mountains, which opposes the main body of the clouds; so that more water is discharged by the south, by the rivers of the Coast of Guinea, than by the inland rivers; or by those of Senegal and Gambia.
Ben Ali reported to Mr. Beaufoy, that “it was believed, that the Tombuctoo river terminated in a lake in the Desert.”
On the whole, it can scarcely be doubted that the Joliba or Niger terminates in lakes, in the eastern quarter of Africa; and those lakes seem to be situated in Wangara and Ghana. That it does not form the upper part of the Egyptian Nile, may be collected from two circumstances: first, the great difference of level that must necessarily exist, between the Niger and the Nile, admitting that the Niger reached the country of Abyssinia. For by that time, it would have run at least 2300 G. miles, in a direct line; and near 2000, after it had descended to the level of Sahara, or Great Desert. And the Nile, at the point where the White River (which, alone can be taken for the Niger, if the idea of a junction be admitted) falls in, has more than a thousand such miles to run, before it reaches the sea; and has moreover two or more cataracts to descend, in its way. Besides, Abyssinia is positively a very elevated tract. Mr. Bruce, (Vol. iii. p. 642.) inferred from his barometer, that the level of the source of the Nile, in Gojam, was more than two miles above the level of the sea: and this is repeated in pages 652, and 712; where he says “fully” two miles.