I. According to the oldest of these opinions, and that which is supported by the greatest authorities (being the opinion not only of some of the principal Geographers of antiquity, but of D'Anville and Rennell among the moderns), it is supposed, that the Niger has an inland termination somewhere in the eastern part of Africa, probably in Wangara or Ghana: and that it is partly discharged into inland lakes, which have no communication with the sea, and partly spread over a wide extent of level country, and lost in sands or evaporated by the heat of the sun.

[Footnote: Proceedings of the African Association, vol. i. p. 535.]

The principal ground of this supposition is, the opinion of some of the best informed writers of antiquity on the geography of Africa, and a sort of general persuasion prevalent among the ancients to the same effect; circumstances, it must be acknowledged, of considerable weight in determining this question: since there is good reason to believe, that the knowledge of the ancients concerning the interior of Africa was much more extensive and accurate than that of the moderns. It is justly observed by Dr. Robertson, that the geographical discoveries of the ancients were made chiefly by land, those of the moderns by sea; the progress of conquest having led to the former, that of commerce to the latter. (Hist. Of America, vol. ii. p. 3l6, 8vo.) Besides which, there are several distinct and peculiar causes which have essentially contributed to our present ignorance respecting the interior of Africa; namely, the great prevalence of the slave trade, which has confined the attention of European adventurers exclusively to the coast; the small temptation which the continent of Africa held out, during the continuance of that trade, to internal commerce; and the almost impenetrable barrier raised up against Europeans in modern times, by the savage intolerance of the Moors.

The ancient opinion, respecting the termination of the Niger which has just been alluded to, receives a certain degree of confirmation from the best and most authentic accounts concerning that part of Africa, in which the Niger is supposed to disappear. This is represented by various concurrent testimonies to be a great tract of alluvial country, having several permanent lakes, and being annually overflowed for three months during the rainy season.

Against the hypothesis of an inland termination of the Niger, several objections have been urged, which are well deserving of attention. They are principally founded on a consideration of the vast magnitude which the Niger must have attained after a course of more than 1600 geographical miles, and the difficulty of conceiving so prodigious a stream to be discharged into lakes, and evaporated even by an African sun. To account for such a phenomenon, a great inland sea, bearing some resemblance to the Caspian or the Aral, appears to be necessary. But, besides that the existence of so vast a body of water without any outlet into the ocean, is in itself an improbable circumstance, and not to be lightly admitted; such a sea, if it really existed, could hardly have remained a secret to the ancients, and entirely unknown at the present day.

It may just be observed, that D'Anville, following Ptolemy and other writers whom he considers as the best informed on the internal geography of Africa, is satisfied that there are two considerable rivers, the Niger and the Gir; both of which are said to terminate in the same quarter of Africa, and precisely in the same manner. The Gir, totally unknown at the present day, is familiarly mentioned by Claudian, who, however, it may be recollected, was a native of Africa:—

'Gir, ditissimus amnis
'Aethiopum, simili mentitus gurgite Nilum.'
Carm. 21. v. 252.

In some MSS. it is notissimus amnis; but the other reading is more probable.

'Domitorque ferarum
'Girrhaeus, qui vasta colit sub rupibus antra,
'Qui ramos ebeni, qui dentes vellit eburnos.'
Carm. 47. v. 20.

II. The second opinion respecting the Niger is, that it terminates in the Nile. In other words, this hypothesis identifies the Niger with the great western branch of the Nile, called the White River, which D'Anville traces from a source very far SS.W. to its junction with the Nile near Sennaar. He likewise accurately distinguishes this stream from the eastern branch, which is much shorter and of inferior magnitude, and which takes its rise in the mountains of Abyssinia. This opinion is maintained by Mr. Horneman, Mr. Grey Jackson, and several other modern travellers; and it is slightly sanctioned by Strabo and Pliny, who speak of the sources of the Nile as being reported by some to be in the farther parts of Mauritania. But it may be affirmed with great confidence, that of all the hypotheses respecting the termination of the Niger, that which supposes it to be a branch of the Nile, is the most unfounded, and the least consistent with acknowledged facts. It is indeed rather a loose popular conjecture, than an opinion deduced from probable reasoning; since nothing appears to be alleged in its support, except the mere circumstance of the course of the river being in a direction towards the Nile; and a few vague notions of some of the African natives with regard to this subject, which are unworthy of the smallest attention.