In the general article given under this head in the body of the work, the reader will find a pretty accurate digest of the information collected by modern travellers in regard to this obscure division of the globe; but as that article either does not touch upon, or but slightly alludes to various disputed questions, of considerable interest, connected with its ancient and modern geography, it hence becomes necessary that we should here make some remarks upon the points in question, as well as upon the additional information which recent enterprise and inquiries have supplied.*

I. The extent of the knowledge of this continent possessed by the ancients, forms one of the subjects of inquiry to which we allude, and upon which much learning and ingenuity have been employed.

1. The first point to be considered under this head is, whether Africa was ever circumnavigated in ancient times? Herodotus states, that Necho, king of Egypt, fitted out an expedition under the direction of certain Phoenician navigators, for the purpose of circumnavigating it; and that these navigators asserted that they had succeeded in this adventurous undertaking. By some of the ancient writers, the account which Herodotus gives of this voyage was held altogether fabulous. Among the moderns, Huet (Commerce et Navigation des Anciens, 34, 275.), the Abbé Paris (Academie des Sciences, VII. 79.), Montesquieu (Esprit des Lois, b. xxi. ch. 10.), and some others, have contended that it was really performed; while Vossius (in his Notes to Pomponius Mela), and D'Anville have intimated a strong degree of scepticism. But the most elaborate discussion of the question has taken place between Major Rennell on the one side, and Dr Vincent and M. Gosselin on the other. The opinion of the two latter is also moderately espoused by Gibbon, in an essay just published in the fifth volume of his posthumous works.

Major Rennell (Geog. Herodot. 672-93) observes, that, though the ancient mode of sailing, confined to a very small distance from the coast, presented difficulties and obstructions unknown to modern navigation, it yet afforded resources for overcoming these difficulties, wholly foreign to the modern practice of the art. Navigators were thus enabled to discover all the creeks and harbours which any coast presented, and which escape vessels that steer at a distance from the shore. The construction of their ships, with very flat bottoms and low masts, enabled them to keep much nearer to the land, than modern vessels of the same magnitude. They could even, when the exigency required, be drawn up along the beach; and instances are quoted, in which large fleets were thus placed in a state of security. And as to the period of three years, specified as having been spent in the voyage,—this may be sufficiently accounted for by the slow rate of ancient sailing—by the difficulties of an un-

* The reader will be assisted in the perusal of this article by occasionally referring to the accompanying map of Northern and Central Africa. We may observe, that there is no precise register of the time spent in sailing from the Straits to Cerne. But another and later Periplus, bearing the name of Scylax, (Geog. Graeci Minores), supplies this deficiency, by stating that distance at twelve days sail. From Cerne to the Southern Horn, twenty-six days are counted by Hanno. The whole number thus appears to be thirty-eight.

Bougainville (Academie des Sciences, X. 45.) contends, that the space which he has supposed to be passed over in the above period of time, though considerable, does not exceed what might reasonably be expected from the most skilful navigators of antiquity. He instances the squadron sent, in 1641, to establish the fort of Elmina, which sailed from Lisbon to Arguin in twelve days, and thence to Cape Three Points in twenty-six; this, by a singular coincidence, being the precise period of time employed by Hanno. It is impossible, he continues, till we arrive at the Senegal, to find any great river which abounds in the wild amphibia mentioned in the journal. In that part only of the coast, have travellers observed those vegetable conflagrations, which might occasion the appearance of a flaming mountain. There only are found the oran outangs, undoubtedly pointed at by the name of gorille. Cerne has been considered as Arguin by D'Anville, and all the most eminent modern geographers. The approach to it by sailing two days west and one day east, and its situation in a recess of the sea, exactly correspond to the Periplus. The case is otherwise as to its magnitude, the circumference being stated at half a mile, while that of Arguin is five or six miles. But this may be an error in the text. Cornelius Nepos states it at twenty stadia, or two miles. Hanno calculates that Cerne is opposite to Carthage, which he explains, by saying that the distance from Carthage to the Straits is equal to that from the straits to Cerne. The latter space is considerably greater; but the estimate might very probably be made by following, according to the ancient mode of navigation, all the windings of the coast. It will then be found so much more deeply indented on the Mediterranean than on the Atlantic, that if the two spaces be thus examined, they will almost exactly coincide.

Major Rennell, (Geog. Herodot. sect. 19, 25), in Rennell, treating of the important positions of Arguin and the Senegal, follows nearly the footsteps of Bougainville. But he ridicules the idea of ancient vessels sailing at the rate of seventy geographical miles in the day, which would be required, in order to reach, in twenty-six days, from Arguin to the Gulf of Benin. Such a rate would be considered rapid by modern ships, with every advantage of wind and tide. In the Gulfs of Bissao and Sherbro, he finds all the islands mentioned by Hanno, which cannot be satisfactorily placed upon any other system. He derives a curious argument, by applying, according to the literal sense of the Periplus, the term Horn to the bay itself, and not, as had usually been done, to the bounding cape. This form is certainly presented by the Gulfs of Bissago and Sherbro, and it does not occur on any other part of the coast.

In the voyage between the Straits and Cerne, Major Rennell justly reproaches the idea of five cities being founded beyond Cape Bojador, on the desert coast of the Sahara. Being unwilling, however, to admit the immense spaces of Bougainville, he is himself somewhat embarrassed by Scylax's estimate of twelve days sail from the Straits to Cerne, which would give a rate of 104 miles per day.

The extent of M. Gosselin's scepticism will appear from looking at the table. Instead of a space of 3000 miles, which is assigned by Bougainville, and 2200 by Rennell, he does not extend the voyage beyond 640 miles; instead of carrying Hanno into the heart of Guinea, he scarcely allows him to have passed the limits of Morocco. In support of this opinion, he makes the following observations. (Geog. des Anciens, I. 60—164). Hanno's rate of sailing could not but be slow. He was upon an unknown sea, where numerous precautions were requisite,—upon a voyage of discovery, which called upon him to examine minutely every part of the coast, and to sail only during the day. He was attended also by a large and encumbered fleet, in

which the slowest sailing vessel necessarily fixed the rate of the others. M. Gosselin thus conceives twenty miles a-day a very fair and probable average. He appeals to the difficulty which the Portuguese experienced in doubling Cape Bojador. After a series of fruitless attempts, they succeeded only by the aid of the compass, which enabled them to stand out to sea. Destitute of that instrument, Hanno, he insists, never could have overcome the formidable obstacles which there presented themselves. The important position of Soloe is distinctly stated by Herodotus as the western extremity of Africa, and the point at which the coast begins to take a southerly direction. This clearly points out Cape Spartel. The island of Fedala has precisely the dimensions assigned to Cerne, while those of Arguin totally differ. As to its equality of distance with Carthage from the Straits, he accounts for it by supposing, that on the known coast of the Mediterranean, vessels proceeded direct from cape to cape, while in an unknown sea, they would necessarily follow all the windings of the shore. The other islands mentioned by Hanno, he conceives to be merely shallows, inundated at high tides, as is proved by the salt water remaining upon them. To the objection that there are neither Ethiopians (negroes) nor crocodiles, nor hippopotami, nor ouran outangs now to be found in the kingdom of Morocco; he answers, that the term Ethiopians was not necessarily confined to negroes, but was extended to all people of a deep brown or olive complexion, and that the wild animals above mentioned formerly existed in this territory, whence they were expelled as the country became peopled.

In order to support this system, M. Gosselin has drawn forth and illustrated another curious document. This is the Periplus of Polybius, inserted by Pliny in the sixth book of his Natural History. The names are almost entirely changed from those of Hanno, which renders it difficult to compare the two together. A part of the distance, however, is given in Roman miles, which amount to 813, between the Straits and the Great Atlas. M. Gosselin then undertakes to prove, that this is not only the entire space passed over by Polybius, but that it equals the whole extent of the voyage of Hanno. Of this circumstance, indeed, Polybius himself is by no means aware. He states a great distance from the termination of his voyage to the Chariot of the Gods, and ten days from thence to the Western Horn. Our author, however, supposes, that not recognizing in the objects before him the wonders described by Hanno, he vaguely referred the latter to remote and still unexplored regions.

M. Gosselin has finally to combat the tables of Ptolemy, with which Major Rennell's ideas pretty nearly coincide, and which delineate an extent of Africa coast, three times greater than he is willing to admit as having been known to the ancients. He contends, that, in all the ancient narratives, the names of places were assigned in a rude and careless manner, and that they were almost entirely different in different journals. He therefore infers, that Ptolemy might have perused three different voyages along precisely the same shores, and, from the names being different in each of the narratives, might thus be induced to delineate an extent of coast three times greater than was actually explored. Nothing can better evince the obscurity which hangs over the writings of the ancient geographers, than that it should be possible to produce even plausible evidence of an error so singular.*

* Without venturing upon any very decided opinion as to the comparative merits of these three systems, we would observe, that Bougainville appears to have given an extravagant extension to the voyage of Hanno, and that we think the system of Rennell the most conformable of any to the features of the coast as delineated in the Periplus. We confess, at the same time, that the short diurnal progress contended for by M. Gosselin appears better suited to all the circumstances and difficulties under which the voyage was performed; but this difficulty would be removed, if, as we incline to think, a day's sail among the ancients was always fixed to a certain extent, like a day's journey by land. | Description of the Coast | Peripius | Ptolemy | Gosselin | Vincent | |--------------------------|----------|---------|----------|---------| | The cape at which the coast turns to the south | Aromata | Aromata | C. Guardafui | C. Guardafui | | A promontory and mart | Tabai | Pano Viens | C. d'Orfui | C. d'Orfui | | A considerable mart | Oponè | Oponè | Bandel Cans | Bandel D'Agua | | A promontory | Zengifa | Phalangis mons | Morro Cobir | Morro Cobir | | A mountain with three summits | Apokapa the less | Apokapa | C. Delgado | Zorzema | | Two successive gulfs | Apokapa the greater | Southern Horn | Bandel Cans | Cape Baxas | | Azania | Little coast | Little coast | No objects | No objects except Magadexo | | A port | Great coast | Great coast | Unknown | Brava | | A promontory and port | Serapion | Serapion | C. Baxas | Mouths of the Obii, or Quillimanci | | A harbour | Nicon | Nici | Formed by its branches at Melinda Mombasa, &c. | | Anchorages at mouths of rivers | Seven in succession | None mentioned | No traces to be found | | | Islands merely named | The Pyralaan | Occurs afterwards | Unnoticed | | | A low wooded island 300 stadia from land | Eitenmediom—Menuthias (Menuthias) | Misplaced | Monfia | | A promontory and great emporium (termination of the Peripius) | Rhapta | Rhampton | Bandel veillo (mouth of the Doara) | Quiloa | | An island 5° W. L. from Prasum | Menuthias | Magadexo | Madagascar | | | A promontory, port, and river, the limit of ancient knowledge on this side of the continent | Prasum | Brava | Mosambique | |

In forming his calculations, M. Gosselin (Geographie des Anciens, I. 165-198) considers that he is fully justified in resting chiefly on the authority of Ptolemy. That writer, from his residence at Alexandria, had an opportunity of examining personally the pilots who sailed upon those seas, and appears to have diligently employed his means of information. Following him, M. Gosselin states, that, owing to the variable nature of the winds, navigation is extremely slow in those latitudes; and that the space passed over in a night and day did not exceed 400 or 500 stadia. When a day's sail only is mentioned, he therefore estimates it at not more than 250 stadia, or 25 miles. Upon this principle, he finds 7500 stadia from Aromata to Rhapta. Examining the distances along the coast, he finds 8000 stadia between Cape Guardafui and the mouth of the Doara,—a very close approximation on so great a distance. Ptolemy, indeed, reckons Rhapta at about eight degrees of south latitude; but, in the first place, it is universally acknowledged that he has placed Aromata seven degrees too far south. Again, M. Gosselin conceives it to have been established that Ptolemy, following the error of Marinus, composed his degree of only 500 stadia, instead of the real number of 700. The degrees between Aromata and Rhapta must, therefore, be reduced two-sevenths. After these deductions are made, Rhapta will fall short of the equator, and nearly at the point to which we are conducted by the other species of measurement.

Dr Vincent, on the other hand, (Peripius, I. 69-166), prefers the Peripius to any other source of intelligence. It is evidently, he says, the testimony of an eye-witness who performed the voyage; it contains nothing marvellous or improbable, and its delineation of the coast corresponds better with that of modern navigators than any other account. He estimates the day's sail in that journal at 500 stadia, or 50 miles. Admitting Ptolemy's error as to the latitude of Aromata, he conceives that he corrected it in the course of the delineation. At Essina, under the line, the difference from Brava is only 1°. At Rhapta, the difference from Quiloa is not quite 10°. The graduated distance from Aromata to Rhapta, according to the Peripius, is 19° 45′. From Guardafui to Quiloa, according to D'Anville, is 20° 15′,—a coincidence almost too close. Still Dr Vincent rests his main argument on the correspondence of the features of the coast. He appeals, in particular, to the mention made in the Peripius of seven rivers, each at the distance of a day's sail from the other. This remarkable coincidence can be found nowhere, unless at the mouths of the great river Obii or Quillimanci, which lie far beyond M. Gosselin's limits. This proof does indeed seem pretty decisive; nor does that ingenious writer meet it with his usual intrepidity. He says, "The author of the Peripius speaks of seven other harbours which he does not name. No trace of them remains." But the words in the Peripius are express: Ἱπεραίνι Πλάσιοι (Geog. Grac. Minor, I. 90.) The most formidable objection to this theory appears to be this, that the ancients should have passed all the three Zanzibar islands without making the slightest allusion, unless to one. This circumstance may have led Mr Pinkerton to fix upon Pemba as Menuthias. Mr Salt, who, with classical knowledge combined opportunities of observation which fall to the lot of few, expressed to us decidedly the same opinion. Pemba alone, he observed, corresponds exactly as to the distance, the others being nearer. Mr Salt concurred also with D'Anville in viewing Cape Delgado as the Prasum of Ptolemy, and the farthest limit of ancient navigation.

4. We have now to consider the knowledge of the ancients relative to the interior of this great continent.

The coast of Africa, situated on the Mediterranean, was more completely and intimately known to the ancients than it has been to the geographers of modern Europe. Its connection with the states of Greece and Rome was so close, as to make it form only part, as it were, of the same system. As we advance into the interior, this knowledge becomes gradually more imperfect. Still, with regard to the countries watered by the Niger, it was perhaps, on the whole, superior to that possessed by Europeans down to the commencement of the present century.

So far back as the time of Herodotus, we find a notice respecting the eastern course of the Niger. That writer believed it to be the same river with the Nile,—an opinion still prevalent in Africa, though it appears to be erroneous. He also states a distinction between Africans and Ethiopians, which appears to coincide with the modern one between Moors and negroes. Beyond these few facts, the knowledge of Herodotus respecting the interior appears to have been exceedingly limited. War and commerce, however, gradually opened routes into these regions; and Alexandria, which became a great seat of geographical knowledge, afforded ample facilities for augmenting the stock of information. The best ancient description of the interior, is to be found in Ptolemy. It appears, however, that he was unacquainted with the extent of the Sahara; and, that he brought the territory of Nigritia much too near to the coast of the Mediterranean. His ideas, likewise, do not seem to have been very precise, in regard to the Garamantes, whom Rennell has upon good grounds placed in Fezzan, though we suspect the ancients must have extended the term to the nations inhabiting the valley of the Gir. But his geography displays an extensive and accurate acquaintance with the rivers of Africa. He has fixed the head of the Nile too far south indeed, and even beyond the equator; but he assigns to it its real origin among the Mountains of the Moon. It was by the data which he afforded, that D'Anville was enabled to ascertain the true sources of that great river. He has shewn that he was acquainted with the whole course of the Niger, and even with that of the Gambia and Senegal. He has not, however, thrown much light upon the celebrated question relative to the source and termination of the first mentioned stream: Major Rennell claims his suffrage to its eastern course, while Mr Pinkerton, perhaps rather more plausibly, charges him with giving it a western direction. So far as we can discover, Ptolemy did not pay the smallest attention to this circumstance, but considered merely the line of river-course. His general statement is, that the Niger "joins together mount Thala and mount Mandrus." (Lib. IV. cap. 6.) If Ptolemy annexed any precise meaning to this very singular description, it could only, we apprehend, be that of two rivers, flowing from opposite points into one common receptacle. He delineates also the course of another river called the Gir, almost equal to the Niger, and well known in ancient times. From its course and relative position to the Niger, there seems no room to hesitate in adopting the opinion of D'Anville and Rennell, that it is the river of Bornou, united, as we apprehend, with the Misselad of Browne, which we have no doubt is the same river, and forms certainly one continuous river-course.

II. Among those branches of knowledge which were particularly encouraged by the Arabian princes, geography seems to have held a conspicuous place; and as Africa contained some of the chief seats of Musulman power, that continent attracted a great share of their attention. The Arabian geographers were thus led to give particular descriptions of some regions in the interior, which have never been visited by Europeans, and, in regard to which, they still remain almost our only authorities. These considerations have induced us to examine their writings with much care; and not having met with any systematic attempt to estimate the extent and value of their knowledge respecting central Africa, we hope that it will not be unacceptable to our readers to follow us into some details upon this curious subject.

We would observe in general, that the works of the Arabian geographers exhibit a remarkable mixture of correct and solid information, with meagre, unsatisfactory, and even fabulous details. The causes of some of these defects are clearly enough pointed out by themselves. Thus, Abulfeda complains, that it is "only the Musulman provinces which they describe with any degree of accuracy." (Geogr. Graci Minoris, III. 21.) This, which to Abulfeda is a subject of regret, forms the glory of another Arabian geographer. "Of Christians and Ethiopians, says he, I have spoken little; for my innate love of justice, religion, and good government, made it impossible for me to find in these people any thing deserving of praise, or even of mention." (Ibn Haukal, Oriental Geography.) In fact the Saracens were in a state of habitual warfare with every other people. They were separated from them by a political and religious antipathy so deadly, as to exclude all interchange of ideas. The limit therefore of Moslem dominion will be always found to be likewise that of clear and accurate knowledge; after passing which, it gradually becomes faint, and gives place to fable and uncertainty.

A general impression seems to have prevailed, that Nubia was the point whence the information of the Arabians relative to central Africa was chiefly derived. This opinion rests solely on the belief of Sionita and Erpenius, that Nubia was the native country of Edrisi, the most copious of the Arabian writers on the subject of Africa. On that presumption, they published his work under the title of the Nubian Geography, which it still retains. This idea, is treated by Gibbon with great contempt, and completely refuted by Hartmann. (Edrisi, Introd. 49. &c.) In fact, there is scarcely any district of Africa, which is described in a manner so meagre and unsatisfactory by the Arabians, and above all by Edrisi himself. Nubia was Christian; * and was therefore a country shut against Mahometan Egypt. The merchants of each country, instead of travelling into the territory of the other, brought their commodities to the opposite sides of the great cataract, where, having landed and made the exchange, they immediately reembarked. (Edrisi, 72.) But in the absence of this medium of information, the Arabians had opened for themselves another, nearly unknown to the ancients. At some period, not quite ascertained, they had penetrated across the great desert to the eastern shores of the Niger. Attracted by the fertility of its banks, and above all by its golden sands, they were not long of forming permanent settlements. Emigration and revolution continually swelled the tide; and in the course of one or two centuries, several great Mahometan kingdoms were established upon the Niger. Among these, Ghana, or Gana, held a decided pre-eminence. It was the nearest point to Northern Africa; and it bordered immediately on Wangara, whose soil the overflowing of the Niger annually impregnated with gold. A regular commercial intercourse was soon established with the other Musulman states; and caravans, in every direction, traversed the vast expanse of the desert. Gana became their point of rendezvous, and the grand emporium of central Africa. The splendour of its court is described in glowing terms by the Arabian writers. The palace, built upon the lake, was adorned with masterpieces of painting and sculpture, and with glass windows, then a rare object of luxury. The pomp of the sovereign is said to have surpassed that of any other Musulman potentate; and the circumstance, which above all excited admiration and envy, was a large mass of native gold, which was placed immediately under the throne. Many of the neighbouring states, among which Wangara is particularly mentioned, were numbered as his subjects or tributaries. (Edrisi, 41-7. Ibn-al-Vardi, Notices, II. 37.)

From these circumstances, it appears evident that Gana was the point whence chiefly emanated the knowledge of the Arabians relative to central Africa. It did not extend very far; for their settlements were then confined within narrower limits than now. Westwards, at the distance of about 400 miles, Tocrur and Sala are described as flourishing commercial cities, situated on the Niger, and governed by a Mahometan sovereign. They seem to have formed the limit of precise knowledge; but the necessity of procuring a supply of salt had given a further line of sixteen days' journey to Ulil; which Edrisi describes as an island situated in the sea, at one day's sail from the mouth of the Nile of the negroes, or Niger. It is now a very general, and, we apprehend, just opinion, that this sea of Edrisi must in fact have been an inland lake; and it seems, indeed, to be indicated as such by Scheabeddin. (Notices, II. 156.) The precise situation of this lake, and of the island of Ulil, will be a subject of future discussion. Meanwhile, we shall observe, that no supposition has extended it much to the west of Tombuctoo. The countries beyond, watered by the Senegal, the Gambia, and the part of the Niger visited by Europeans, were entirely unknown, and were viewed as a wide expanse of trackless desert.

To the east of Gana, the Arabs particularly describe Wangara, then celebrated over all Africa by the appellation of the "country of gold." They represent it as nearly surrounded by branches of the Nile, which overflow it during the rainy season; and as containing two lakes adjacent to the cities of Reghehil and Semegonda. Cauga (now Fittre) they appear also to have been well acquainted with, though there is some difference of opinion, whether it belonged to Wangara or to Canem, (now Bornou.) Eastward from thence, their knowledge becomes indistinct; and, before reaching Nubia, seems to cease entirely.

The settlements of the Arabs, as well as their geographical knowledge, extended but a short distance to the south of the Niger. The Sahara they seem to have traversed in almost every direction, setting out chiefly from its western extremity. Gadamis, Segelmessa, and Vareklan, (the last uncertain, but westmost of any,) are described as the customary routes, which led them to Gana, and the "land of gold." (Bakui, Notices, II. 400, 405. Edrisi, 133-135.)

In regard to the origin of the Nile, and the first part of its course, the Arabians were excluded from the opportunities of direct observation. We cannot, however, agree with Hartmann, in thinking that they derived it from Abyssinia, a country to which they were peculiarly strangers. We rather concur with Rennell and Pinkerton, who apprehend that their opinion nearly coincided with that of Ptolemy. The Arabians committed one very curious error, of which modern geographers have not been fully aware. They certainly considered the river of Bornou and the Nile of Egypt to be one and the same. Hartmann has remarked this in the case of Scheabeddin; but has not observed, that all the rest entertained the very same idea.† Nor will this appear very wonderful, when all the circumstances are taken into consideration. The Arabs knew little of Nubia, nor had traced the course of the Nile above its confluence with the Tacazze. (Hartmann's Edrisi, 75, 76.) Observing then a great river running through Barbon northwards, which could reach the sea only by uniting with the Nile, it was not a very strained supposition, that it might, at some point, take a direction westward, and effect the supposed junction. In consequence, however, of this opinion, the Arabs lost sight entirely of the Gir of Ptolemy, and recognized in Africa only two great rivers, the Nile of Egypt, and the Nile of the Negroes.

From the circumstances now stated, we may form Parallel of

---

* See Bakui in Notices des Manuscrits de la Bibliothèque du Roi, II. 396. Ibn-al-Vardi. ib. 38. † Abulfeda makes the Nile of Egypt pass through Zagava; for the position of which see Hartmann, 66-8. Ibn-al-Vardi (Notices, II. 36.) carries it through Canem, a district of Bornon. Compare also Hartmann. 64-5, and 56-7.

an estimate of the knowledge of the Arabian geographers compared with that of Ptolemy. The ancients evidently penetrated through Nubia, up the Nile, and thence westward along the Niger. Hence the great desert of Africa remained unexplored; for if an occasional adventurer made his way across it, no regular communication was ever established. But they enjoyed the opportunity of tracing the origin, course, and termination of all the great African rivers. Generally speaking, their range was wider, as it was not hemmed in on every side by religious antipathies. The Arabs had a different and more limited sphere of observation; but within it, their knowledge is more copious, precise, and satisfactory. Ptolemy's general ideas of central Africa are good; but he does not afford the means of fixing a single position; while the materials furnished by Edrisi have been familiarly used by the best modern geographers in the construction of maps of Africa. This does not, however, arise from any merit in the astronomical part of their geography, which, on the contrary, is extremely defective. But they investigated very diligently the distance of one place from another, by the reports of the caravan merchants—a mode of intercourse which had already been established in Africa to such an extent, and conducted in so systematic a manner, as to afford very ample means of information. Edrisi's work, in this view, is particularly valuable. Major Rennell observes, that it is surprising how nearly his distances agree with those furnished by modern observations and inquiries.

The Arabs made little progress along the western coast of Africa; nor did their knowledge on that side equal that of the ancients. But the eastern coast, to an extent beyond any supposed limits of ancient discovery, was not only explored, but colonized by them. Flourishing settlements were established at Melinda, Mombaza, and at Sofala, called commonly the Golden Sofala. Vakvak, or Ouac Ouac, forms here the indistinct limit of knowledge. The Arabs not only had never passed the Cape of Good Hope, but had not even any idea of its existence. It appears, on the contrary, from the curious map of Edrisi, published by Dr Vincent, that, like Ptolemy, he extended it to the east, till it became contemporaneous with India and China. This explains his placing the islands of Vakvak in the sea of China—a circumstance which causes much perplexity to Hartmann.

III. The period from the tenth to the fourteenth century, may be considered as the most flourishing era of Arabian science. From that time downwards, geographical discovery has been exclusively the boast of Europeans. The fifteenth century was a splendid era in maritime enterprise. Africa was then circumnavigated, its form ascertained, and its coasts visited and colonized. But the geography of its interior, instead of advancing, became decidedly retrograde. The Portuguese, who almost wholly engrossed this department of inquiry, penetrated only at the two opposite points of Abyssinia and Congo. By applying their observations upon these regions to almost the whole of Africa, and combining them with the often misunderstood notices of Edrisi, they formed a system materially erroneous. The Nile, instead of being traced, as in Ptolemy, to its real source, was supposed to rise in Abyssinia, and its southern and largest, as well as longest branch, which descends from the Mountains of the Moon, was thus overlooked. They did not, however, on that account, abridge in any degree the course of that celebrated river. They had extended Abyssinia immensely, so as to make it pass the southern frontier of Congo and Monomotapata, while the Mountains of the Moon were fixed mid-way between the tropic of Capricorn and the Cape of Good Hope. The lake Zambre (Dembca) was thus placed about 30° (upwards of 2000 miles) south of its real position,—was made the common source of the Nile, the Congo, and the Zambezi,—and was described as "the great mother and chief lady of all the African waters."* The Nile being thus derived from sources far beyond the equator, was considered as holding the longest course of any river in the known world. (Ramusio, I. 261.) The Niger still retained a mighty name among the streams of this continent; but no part of it had yet been exposed to European observation. Proceeding on the principles of Leo and Edrisi, the Gambia and Senegal were generally believed to be the two mouths, by which, after traversing nearly the whole breadth of Africa, it fell into the ocean. Some, following Edrisi's idea of its common source with the Nile, fixed upon the lake Dembea; the greater number, with Leo, derived it from a large lake to the south of Bornou; while some "holy men" are said to have numbered it as one of the branches of the Gion, which took this direction after issuing out of Paradise. (Ludolfi Hist. Ethiop. Cadamosto in Ramusio.)

About the commencement of the eighteenth century, the English and French began to form settlements on the Gambia and Senegal, and to ascend these rivers. This afforded to Europe a new channel of information relative to the interior of Africa. The most noted English traveller into the interior during that age was Francis Moore, an intelligent man, who, about 1720, spent several years on the Gambia. Fully impressed with the above-mentioned theory relative to the Niger, he pretends to find even on the Gambia, Edrisi's principal stations. Ulil is Joally, an island at the mouth of the river, which, it seems, really supplies its banks with salt; Sala is Barsally; Gana is Yani; and Wangara alone is higher than Europeans have yet reached. He was much disconcerted, however, by the intelligence brought down by Captain Stibbs, who had sailed sixty miles above Barraconda. This person reported, that the Senegal above Gallam became very small, while, as to the Gambia, twelve days above Barraconda, the natives assured him that "fowls walked over it." He therefore infers, that the source of these rivers is "nothing near so far in the country as by the geo-

* See Map in Purchas's Pilgrims, IV. p. 738; also descriptions of Lopez, Merolla, &c. Ibid IV. 773, 774; II. 1021. It is remarkable that Sanson's map, 1696, is constructed nearly on the above principles. graphers has been represented." As for the Niger, on being closely pressed, he fairly stated his opinion, "that there was no Niger at all." This decision, however, is very ill received by Moore, who asks if it is probable that both Leo and Edrisi "should affirm that there was so great a river as the Niger when there was no such river." (Moore's Travels, London, 1738, p. 298—306. App. p. 6.)

The French meanwhile were more active both in forming settlements and in making discoveries. Before 1714, they had gained such intelligence, as induced Delisle to construct his map of the world, with the Niger as a separate river from the Senegal, and flowing westwards. But it is to the Memoirs of D'Anville, read before the Académie des Sciences, about the year 1754, that we are to look for the first approaches to correct geography, in regard to the rivers of interior Africa. By combining the scanty notices of Ptolemy and Edrisi, with the narratives of modern travellers, he succeeded in forming an almost correct outline of African hydrography. He showed the error into which the moderns had fallen, of conceiving the river of Abyssinia to be the Nile, and proved, both from the magnitude of the stream, and from ancient authority, the superior claims of the Bahr-el-Abiad. With regard to the Niger, instead of one great stream, running across Africa from east to west, and falling into the Atlantic, he recognizes three rivers, 1. The Senegal, well known, and distinct from the Niger; 2. The Niger itself; of which he says,—“We are actually informed, that the river of Senegal, considered hitherto as the lower part of the Niger, and its outlet into the ocean, is different from another river that lies deeper in the interior of Africa; and it is even inferred, from the report of the negroes, that this river flows in a direction contrary to that of the Senegal, or from west to east.” He finds no ancient authority upon which to rest this opinion, except the single testimony of Herodotus. Upon this principle, however, he constructs his map, fixing the commencement of the Niger at the Nigritis Palus of Ptolemy, and its termination at the lake of Reghebil, in Wanga.

3. Another river, flowing in a contrary direction, on the opposite side of Africa, known now as the river of Bornou,—the Gir of Ptolemy, and, as D'Anville erroneously supposes, the “Nile of the Negroes” of Edrisi. He seems strongly impressed with the idea of a communication between this river and the Egyptian Nile, which, however, Browne's route has proved to be a mistake. D'Anville fails chiefly in regard to the geography of the Arabians, none of whose writings, except those of Edrisi, had been at that time translated. He unaccountably supposed that Edrisi describes two rivers under the names of the Niger, and the Nile of the Negroes; whereas, in fact, the latter is the term by which he designates the former; the word Niger nowhere occurring in his work. Hence, while D'Anville has designated one river (that of Bornou) as Edrisi's Nile of the Negroes, he has placed upon another all the countries and cities appropriated to it by that writer. Upon the whole, however, his map of central Africa is an admirable performance, and formed a new era in the geography of this continent.

From this time, D'Anville's system, both with regard to the Nile and the Niger, was generally adopted by the higher class of geographers. Differences of opinion, however, still prevailed, and, in Britain particularly, a stand was long made for the system, which made the Gambia and Senegal the channels by which the Niger fell into the ocean. Kitchin, nevertheless, composed his map upon D'Anville's basis. In 1788, we find a Mr Barnes reporting to the Committee of Council on Africa, that “the Niger rises on the eastern extremity of the mountains of Govinea, and discharges itself into a large lake, the name of which he does not recollect.” He was then asked as to “the transport of European goods up the Senegal and down the Niger, or of African goods up the Niger and down the Senegal” (Reports concerning the Trade to Africa, and particularly the Trade in Slaves, 1788. Part I. Slaves); and thus it appears that the real state of this question was perfectly understood by the committee. During this period, however, an extreme apathy prevailed respecting the exploration of interior Africa; nor were any important contributions made to it, except those afforded by the travels of Mr Bruce. In this respect the African Association operated a most auspicious change, not only by their direct exertions, which were great, but by the enthusiasm which they kindled, and which led even private individuals to collect and communicate information on the subject. At the time when our former article was written, the result of this new impulse had already appeared in the travels of Park, Hornemann, and Browne, of which we there gave so copious an abstract, as to render any repetition unnecessary. It only remains to take a view of the discoveries, by no means unimportant, which have been since made; after which we shall be prepared to investigate those points of African geography which still remain unsettled and obscure.

IV. 1. Some interesting information has been recently laid before the public concerning that part of coeries in Africa which lies northwards from the Cape of Good Hope, beyond the country of the Hottentots and Caïres. It was first visited by two gentlemen of the names of Trutter and Sommerville, the substance of whose observations is given by Mr Barrow in the Appendix to his Voyage to Cochinchina. The next visitor was Dr Lichtenstein, who communicated the result of his observations, first in the Weymar Ephemerides, and since more fully in the second volume of his Travels in Africa. Lastly, Dr Campbell, a missionary traveller, who penetrated to Leetakoo, has just given to the public a volume of travels in these regions.

The Boshuanas, Bectjuanas, or Bootchuanas, are a Boshuanas, very numerous race, stated by Dr Lichtenstein to extend from the 25th to the 20th degree of south latitude, or, as he elsewhere expresses it, thirty or forty journeys north from the river Kuruhman. In fact, however, the northern limit has never been explored. They are evidently of the same stock with the Caïres, but do not quite equal them in bodily size and strength. On the other hand, they are much superior in civilization and the arts of life. They reside in towns of considerable magnitude. Leetakoo, ca- pital of the Matchappin tribe, has been most particularly described. Mr Barrow's informants reckoned the population at from 10,000 to 15,000 souls. Lichtenstein mentions, that, in consequence of the secession of part of the tribe, it had been reduced to less than half that number, and did not exceed 5000. Mr Campbell, however, counted 1500 houses, which, at five to each house, would give 7500, besides 1000 houses situated in the outposts. This calculation seems to have been very carefully made.

The Boshuanas are governed by a king, whose power is hereditary, and not subjected to any regular limits, though it is very mildly exercised. Quarrels are commonly decided by single combat, without his interference; but if the case is referred to him, he not only takes cognizance of it, but inflicts the punishment with his own hands. He commands in war, and his sons are employed in all the embassies to neighbouring states.

The Boshuanas pay considerable attention to agriculture; their fields are commonly fenced round; they cultivate the Caffre millet, two sorts of beans, gourds, and water melons. The labour, however, as usual among barbarous tribes, is performed by the women, who dig the ground with iron spades; the men merely guard the flocks, hunt, and go to war. Fish, though abundant, are viewed with abhorrence as an article of food. Their clothing consists entirely of the skins of animals, carefully tanned; and the greatest part of their body is covered. Their houses are circular, formed of the trunks and branches of the Mimosa, cemented within by a mixture of clay and cow-dung. There is an inner apartment for the family, and an outer one for slaves and domestics. Polygamy is generally practised; and it is common for the rich to have four or five wives. Mr Barrow asserts that there are no slaves; but this is expressly contradicted by Lichtenstein, who states that they are in considerable numbers, and assist the women in their labours, but are mildly treated.

The nation among themselves seem tranquil, mild, and courteous; but they are fond of war, which they carry on with ferocity. Their great object is to seize the cattle of their neighbours. Mr Campbell having, with a view to religious instruction, asked, "For what end man had been made?" the answer was, "For plundering expeditions." After having killed an enemy, they carry off a small part of the flesh, which, at a solemn festival, they afterwards devour.

The tribe immediately south of Leetakoo, is the Wanketzens, said by Campbell to be more numerous, and to carry agriculture to higher perfection. Lichtenstein gives rather an opposite statement. The Morolongs, the Chojas, and the Marootzees, are also numbered among the most powerful and numerous of the Boshuan tribes. The Macquanas, however, the most distant, appear to surpass all the rest. They were described to Lichtenstein as like the sand in multitude, and their city is said by Mr Campbell to be three times the size of Leetakoo. They are particularly skilled in works in iron, and in the manufacture of arms. It was through them that the inhabitants of Leetakoo had first heard of white men;—probably the Portuguese on the east coast.

On the first news of these interesting discoveries, Lord Caledon, then governor of the Cape of Good Hope, took measures to extend them. With laudable zeal, he sent a party, consisting of Dr Cowan and Mr Denovan, and twenty other persons, to explore Africa as far as Mosambique. The first accounts from them were favourable; but a silence of seven years afforded room for apprehensions, which were confirmed by the information collected by Mr Campbell at Leetakoo. After leaving that place, they had proceeded into the country of the Wanketzens, by whom they were at first well received. That people, however, whose character seems marked by ferocity and treachery, were only watching the opportunity of attack. Dr Cowan and Mr Denovan having imprudently gone to bathe, leaving part of their companions at the waggons, and another to guard the cattle, the natives took advantage of their separation; they attacked, first the party at the waggons, then those who guarded the cattle, and, last of all, the two chiefs of the expedition. The whole perished on the spot, with the exception of one, who also was afterwards taken and put to death.

In returning from Leetakoo, Mr Campbell crossed to the west coast, along the banks of the Orange river, a tract which had hitherto remained unexplored. He found it another Sahara, a complete desert of sand and rock. The rocks rise often in a perpendicular form resembling walls. One wall extended about thirty miles without interruption. A few kraals of Coranas are thinly scattered over this desolate region. The river seems on no occasion to overflow its banks, so as to communicate any fertility to this sandy waste. The country, for an unknown extent northwards, exhibits a similar aspect.

2. The recent work of Mr Salt has thrown considerable light upon the geography of Eastern Africa, respecting which nothing authentic had been published for several centuries. That traveller visited the Portuguese settlements of Mosambique, Mesuril, and Sofala, which appear to have greatly declined from the importance which they once possessed. Mosambique is situate upon a small island, lying directly across the mouth of a deep bay. The situation is advantageous, and the place must have been once very strong; but the eighty pieces of cannon by which it is now defended are marked with the rust of antiquity; and the garrison consists of a few confined felons, and invalids. The population of the town of Mosambique is reckoned by Mr Salt at 2800, of whom 500 are Portuguese, 800 the descendants of Arab settlers, and Banians, and the rest Negroes. So slender and mixed a population is altogether inadequate to preserve the colony in a respectable state. The appointments of the governor and inferior officers are greatly too small, and expose them to the temptation of engaging in traffic, and other transactions ill suited to their station. Upon the whole, the state of affairs is such, that a trader at Mocha declared, that, with a hundred stout Arabian soldiers, he could dispossess the Portuguese of the colony.

There is also a government-house at Mesuril, a pleasant village, situated on the peninsula of Caboceiro, which lies opposite to the island of Mosam- bique; and supplies the town with most of the provisions consumed in it.

Sofala is described as a miserable village; but the country round is extremely fertile, and supplies exquisite fruits to the inhabitants of Mosambique. Small forts are also kept up at Inhambane and Cape Corrientes, for the purpose of collecting the ivory, which is abundantly supplied in the neighbouring forests.

The native race in the immediate neighbourhood of Mosambique is that of the Makooa, or Makoonana, already alluded to. It comprises a number of very powerful tribes, extending southwards to Melinda, and north to the mouth of the river Zambezi. They are described as an athletic race of people, very warlike, and extremely hostile to the Portuguese. They sometimes carry their incursions even into the peninsula of Caboceiro. The Portuguese maintain their ground chiefly by alliances with other native tribes who are hostile to the Makoonana.

The principal settlements of the Portuguese in the interior are on the great river Zambezi, the mouth of which is three or four days' sail from Mosambique. Sena, the capital of these settlements, lies about two hundred and forty-seven miles up the river, on its southern bank. It contains about two thousand inhabitants, and is protected by a strong fort. The principal mart for gold, however, is at Manica, about twenty days' journey south-west from Sena. About sixty leagues above Sena is Tete, which contains a depot for merchandise, and is considered the best regulated settlement on the Zambezi. A month's journey beyond Tete is Zumbo, the remotest settlement of the Portuguese, and where it is only by the permission of the natives that they maintain a small factory.

The trade of Mosambique has declined. Its exports consist of gold, ivory, and slaves. The annual number of the latter is stated at four thousand. This traffic, however, has diminished, in consequence of the British prohibition, which has shut the markets of the Cape of Good Hope, the Isles of France and Bourbon, and Batavia. Ivory and gold bear very high prices. The former is stated at from Ls. 24 to Ls. 21, 15s. per hundred. Gold dust is at Ls. 3, 5s. the ounce avoirdupois. Provisions are in abundance, and at a moderate rate. The imports consist of what is usually required in a tropical colony; articles of dress, furniture, and ornament; arms and ammunition; teas; and cloths. Four or five vessels come annually from the Portuguese settlements in Malabar; and it was thought that a small British cargo might also be disposed of in the months of April, May, or June. The duty on imports is twenty per cent. ad valorem; with other charges, which amount to about five per cent. more.

We were informed by Mr Salt, that the present maps and descriptions of the eastern part of the continent in general are almost wholly imaginary. Thus, the great empire of Monomujii, or Nimeamay, which is made to fill nearly the whole breadth of Africa, is derived from a tribe behind Mosambique, called the Monjous. That chain of mountains, which makes so conspicuous a figure in our maps, under the title of the Spine of the World, is not believed by Mr Salt to exist. He saw persons who came from the lake Maravi, without having encountered any mountains on their route. In general, Mr Salt observed that it was usual in this quarter to extend every district far beyond its actual limits. The names of small towns and villages are currently transferred to a vast extent of territory behind them. In this manner have been formed the kingdoms of Sofala, Inhambane, Mosambique, and Querimba. The kingdom of Adel, which makes so conspicuous a figure in the history and geography of Africa, does not appear to have any existence. Its alleged site is occupied by the tribes of the Somauli, concerning whom Mr Salt obtained much information, and who appear to have made considerable advances in civilization. The name of Adel seems derived from the tribes of the Adiel, who inhabit, not the place assigned to that kingdom, but a more northern position on the shores of the bay of Zeila. Contiguous to them, on the west, is the kingdom of Hurur, which seems generally associated with the Somauli in the wars carried on against Abyssinia.

3. Dr Seetzen, a German physician residing at Alexandria, had lately an opportunity of receiving some information from two natives of the interior of Africa, who were on a pilgrimage to Mecca.* One of them, Abdallah, was a native of Bornou; the other, named Hassan, was from the western part of Darfoo. The former gave a very magnificent description, probably a good deal exaggerated, of his native country. The city of Bornou he represented to be of such magnitude, that "Cairo was a trifle compared to it;" and that a day was insufficient to travel from one end to the other. This great extension is probably produced by the irregular mode of building, not in streets, but by single houses, with gardens, and even fields intervening. The mosques are innumerable, built of stone, and mounted with towers, seven of which appear on the principal one. The schools are numerous in all the cities, particularly in Bornou. They are held in the mosques, and are supported at the expense of the Sultan. His palace is of immense extent, and he goes every Friday in state to the principal mosque, where he distributes alms to the indigent. Within a mile of Bornou there flows a river, which Abdallah described to be as large as the Nile, and, like it, to overflow its banks. It is navigated by vessels of considerable dimensions, carrying sails and oars. The soil is sandy, and requires irrigation; but it produces in abundance rice, and a great variety of vegetable productions. The forests are very extensive, and contain lions, orang-outangs, camelopards, and other wild animals. There are numerous negro and Abyssinian slaves, who, whenever they arrive, are converted to the Mahometan religion. This is effected by flagellation, which is continued until the person cries out, "Mahomet is the prophet of God;" when he is immediately ranked among sound Musulmen. Abdallah asserted also, that there were a few Christian slaves at Bornou.

Although Abdallah shews an evident disposition

* See Malte-Brun Annales des Voyages, XIX. 164—184. XXI. 145—179. to exalt his native country, there seems no doubt that Bornou is the most extensive and powerful kingdom of central Africa. Bergoo, Begherme, Wangars, and Cashna, are tributaries to its Sultan. He appears however, to boast of sway over regions to which his dominion has never really extended; as Hassan mentions among his subject countries Darfoor, Senaar, and part of Barbary.

From the native of Darfoor, some information was received concerning Bergoo, (which he called Mobba,) an extensive country between Bornou and Darfoor. It appears to be considerably inferior, both in populousness and civilization, to Bornou. The capital, Hara, or Wara, is about three times the size of Bulak, a suburb of Cairo. The houses are built of wood and earth; only some of the Darfoor merchants have them of stone. The inhabitants are chiefly Mahometan negroes. Natron and sal gem are produced in considerable quantities, and exported to Cairo. No grain is raised, except durra and millet, with each of which they make an intoxicating liquor. Three days' journey to the west of Wara, there passes a river, (doubtless the Misselad of Browne,) said to be larger than the Nile, and, like it, to overflow its banks.

Begherme is an extensive country, situated to the west of Bergoo, and, till of late, governed by its own Sultan, though dependent upon that of Bornou. According to the statement of Hassan, this monarch had espoused his sister. The Sultan of Bornou, scandalized by this incestuous connexion, sent orders to renounce it, or to await his vengeance. An insulting answer being returned, the offended sovereign immediately ordered his vassal, the Sultan of Mobba, (Bergoo,) to march his troops into Begherme, and annex it to his territories. The enterprize succeeded; the Sultan of Begherme was taken prisoner, and his dominions annexed to those of Bergoo.

4. The account of the Gold coast, published by Mr Meredith in 1812, exhibits a view of some important changes in the aspect of that part of Africa. The Fantees were known to be the principal people upon this coast; but immediately behind are the Ashantees, a more numerous and powerful nation, hitherto known only by report. It appears that they had long entertained a desire to open a communication with the coast, for which an opportunity was at last afforded. In 1806, two chiefs belonging to the kingdom of Assin, a dependence upon Ashantee, rebelled, and, being defeated, took refuge in the Fantee country. The Ashantee monarch pursued them; at the same time intimating, that he entertained no hostile views against the natives. The Fantees, however, joined their force to that of the fugitive chiefs; and a battle ensued, in which they were completely routed. The Ashantees marched down in triumph upon Anamaboe, where the English have one of their principal forts. In their way they destroyed the town of Cormantine, and possessed themselves of the Dutch fort there. The Anamahoes, meanwhile, used no means of defence, and suffered the Ashantee army to approach within three miles. The alarm being then given, all their forces were hastily collected; and the English, from the fort, witnessed, with no little anxiety, the action which followed. It quickly appeared that the Fantees were inferior, both in numbers and skill; and their army fled into the town in the utmost confusion. The Ashantees followed, and pursued them to the beach, where they vainly endeavoured to gain their vessels. A dreadful slaughter ensued, in which upwards of eight thousand are supposed to have perished. The English had imagined, that the mere discharge of a few cannon over their heads would have intimidated and deterred the Ashantees from entering the town; but, after accomplishing its destruction, these barbarians, undismayed, and upwards of twenty thousand strong, turned their arms against the fort. The garrison did not exceed fourteen men, and was soon reduced to eight. Yet such was the good condition of the fort, and the firmness of this small body, that, at the close of the day, they were still able to oppose a stout resistance; and the enemy, overawed by this vigorous defence, did not renew the attack on the following morning, nor even oppose the entrance of a small reinforcement from Cape Coast. A negociation having been opened, a treaty was soon concluded, and amicable visits interchanged. The Ashantees were found to be a people greatly superior to the Fantees, both in numbers and in civilization; and they seemed particularly anxious to open a commercial intercourse with the English. Want of provisions, and the inroads of disease, obliged them soon to return to their own country. They have since made several inroads, and particularly one, on a great scale, in 1811; but the Fantees, though always routed in the field, have hitherto succeeded, by their knowledge of the country, and other causes, in preventing them from forming any solid establishment.

5. The fullest description which has yet appeared of the great central city of Tombuctoo and its vicinity, is that appended to Mr Grey Jackson's Account of the Empire of Morocco. This narrative, perhaps too highly rated by some eminent critics, is regarded by others as utterly fabulous. The able editor of the lately published account of Park's last unfortunate expedition has wholly rejected Mr Jackson's details, upon the ground that they are solely derived from native travellers and merchants; and he broadly lays down the principle, that African evidence is not to be admitted in geographical disquisition. Were this rule admitted, all modern knowledge of central Africa, and, in particular, all the information collected by the African Association, with the single exception of Mr Park's communications, would fall to be thrown entirely aside; but we confess, that we cannot discover any good reason for such a sweeping proscription. The travelling merchants of Africa, though destitute of education, yet visiting a variety of countries, and being frequently placed in critical exigencies, acquire a cultivation of mind superior to that possessed by the same class in Europe. The general tests of evidence are, no doubt, to be carefully applied in regard to their communications. Attention must be paid to the qualifications of the narrator, to his means of knowledge, and to his having no interest in disguising the truth. It must be very desirable also to have more than one independent witness to the same fact. But, upon the whole, we scarcely recollect an instance in which the testimony of respectable Africans has proved materially erroneous.

Tombuctoo, according to the information collected by Mr Jackson, is situated on a plain, surrounded by sandy eminences. The houses are spacious, and built in the quadrangular form. They have no upper apartments, nor even any windows, being lighted by throwing open the doors, which are wide and lofty.† The inhabitants are said to possess an elegance and suavity of manners not known in the north of Africa; and their attachment to their country is such, that they never fail to return to it as soon as circumstances permit. The police is excellent; it is managed by a divan of twelve Alemma, who are elected every three years, and then fall into the mass of the people. Toleration is said to be exercised towards every class except the Jews, who are either excluded from the city, or obliged, while in it, to adopt the outward profession of Mahometanism.

Among the commodities brought to Tombuctoo, Mr Jackson mentions gold rings of Wangara—the first modern notice that has occurred to us of a trade so celebrated in the time of the Arabians. Mr Jackson mentions the countries of Lamlam and Melli, not in the erroneous position of Leo, but as fixed by the Arabs; and adds, that they are reported to be inhabited by one of the lost tribes of Israel. This report, which certainly savours of the marvellous, derives, however, some sort of countenance from the statements of Kircher and Edrisi,—purporting that a colony of Jews, which first settled on the shores of the Niger, afterwards removed to this country; flying probably before the arms of the Saracens. (Hartmann, 37.)

In regard to the sovereignty of Tombuctoo, Mr Jackson has committed at least an anachronism. He states, that, in 1800, it was subject to Woolo, king of Bambarra. But from Park it appears, that, in 1796 and 1805, the king of Bambarra was Mansong, and Tombuctoo independent. Isaac's journal, to be afterwards noticed, indeed mentions, that Woolo (or Wolloo) was predecessor to Mansong, and a distinguished warrior; so that he may at one time have held Tombuctoo in subjection; but it could not be in 1800.

We shall conclude this part of our subject with some account of the last expedition of the great explorer of the interior of Africa. After a number of delays, occasioned by ministerial changes and official difficulties, Mr Park set sail on the 30th January 1805. On the 28th March he arrived in the road of Goree, and having made the necessary preparations, and engaged a body of men from the garrison to accompany him, he set out from Kayee on the 27th April following. He was obliged to remain six days at Pisania, in order to procure some articles, the necessity of which had not been foreseen. Proceeding thence on the 11th of May, he reached Madina, capital of the kingdom of Wooli, and on the 15th was on the banks of the Gambia. The weather had as yet been favourable; and no obstacle presented itself, unless from the eagerness of the natives to obtain a larger share than he could prudently give, of the valuable articles which he carried along with him. By a happy mixture, however, of firmness and moderation, he always succeeded in extricating himself from these embarrassments without coming to extremities. He left Bammakoo on the 13th of May, and for some time held a prosperous course. The men continued in good health, and two who had been affected with the dysentery, were fast recovering. But the rainy season at length approached, and those who were on the recovery speedily relapsed. In a few days twelve were on the sick list. No sooner was the expedition perceived to be in a sickly state, than the natives conceived hopes of turning it to their advantage. Clouds of thieves hovered round, and, at every moment of distress, rushed forward to snatch whatever had been left exposed. By the 6th of July the whole party were in a state of sickness, and several affected with mental derangement. On the 27th July four men lay down, and refused to go farther. Mr Park was himself very sick and faint, and seems to have been on the point of yielding to despair, when he was cheered with the view of some distant mountains to the south-east. "The certainty that the Niger washed the southern base of these mountains made him forget his fever; and he thought of nothing but how to climb over their blue summits." Three weeks more of incessant effort and distress brought him to the wished for object; for coming to the brow of a hill, he once more saw the Niger "rolling its immense stream along the plain."

In this journey, Mr Park, after leaving the village of Fankia, changed his former route for one more to the north, through Konkodoo and Fooladoo, near the southern frontier of Kaarta. He does not state the reason, but his editor conjectures, with great likelihood, that it was to avoid the Jallonka wilderness. This new route gave him an opportunity of observing the process of washing for gold. A woman collected in a calabash a quantity of sand and water; and having cleared it of the large pebbles, agitated the whole with a rotatory motion. The baser particles being thus thrown out, a black substance remained, resembling gunpowder; this is called gold rust. On its being further agitated, yellow specks began to appear, which were the grains of gold. In two pounds of sand, twenty-three of these particles were found. The quantity of gold rust is in general forty times that of the gold. It was stated that pieces were sometimes obtained as large as the fist. Mr Park had afterwards an opportunity of seeing the mode of smelting and forming it into rings. It was performed by the mere action of fire, without any flux or mixture whatever.

This, of course, applies only to facts which have fallen under their sphere of observation; not to opinions, (such as that of the identity of the Nile and the Niger,) which relate to distant objects, and would require a large mass of evidence to enable them to decide.

† Captain Blankett states, on the contrary, that they are two or three stories high. Report of Committee of Council on Africa, &c. 1788. Mr Park was particularly struck with the scenery of Dindikoo. The villages on the mountains, he says, are romantic beyond anything he ever saw. "They are built in the most delightful glens; and while the thunder rolls in awful grandeur over their heads, they can look from their tremendous precipices over all the wild and woody plain, which extends from the Falme to the Black river."

Mr Park arrived on the Niger in a situation, and with prospects, far different from those which he had fondly anticipated. Of forty-four persons whom he had brought with him from Pisania, there remained only six soldiers, and one carpenter, all in the most infirm state of health, and one of them deranged. Such circumstances might well have deterred the boldest spirit from plunging farther into the depths of unknown and hostile regions. But Park never seems to have felt a moment's hesitation. As soon as a canoe could be hired, he set sail for Marraboo, where he arrived on the 23rd. He then sent forward his guide Isaaco to Sego, to request of Mansong, king of Bambarra, permission to pass through his territories. Mansong immediately dispatched his prime-minister Modibinne, to ask his motives for coming into Bambarra. Park made a very judicious reply, in which he particularly set forth the advantages which Bambarra would receive from the direct importation of English goods, instead of receiving them by the circuitous route of Morocco and Tombuctoo. Mansong, whose whole conduct seems to have been liberal, soon returned an answer, in which he gave full permission to travel through his dominions in any direction, and to build boats at whatever point might be deemed most advisable.

No wish, however, being expressed for a personal interview, Park immediately proceeded to Sansanding, which he fixed upon for the construction of his vessel. Here he resided for two months, and enjoyed a better opportunity than had ever before occurred, of observing the economy of an African city, and the mode of conducting trade in this part of the continent. The result exhibits a favourable view of its progress in civilization. The arrangements for the convenience of trade, and the division of employments, appear to be carried to greater perfection, than in any European town of the same magnitude. The market-place is a large square, and the stalls on which the different articles are exposed, are shaded by mats from the heat of the sun. In general, each stall is appropriated to a single commodity. Some contain beads; others indigo; others wood-ashes; and one was observed with nothing but antimony in small bits. Other articles enumerated are salt, scarlet, tobacco, amber, silks from Morocco, sulphur, copper, and silver rings, and bracelets. A table is also given of the prices which different goods bore at this market, and it is so novel and curious, that we shall make some extracts from it.

| Value in Cowries. | |------------------| | A musket, | 6 to 7000 | | A cutlass, | 1500 to 2000 | | A flint, | 40 | | Gunpowder, one bottle, | 3000 | | Amber, No. 1 to 6, | 60 to 1000 |

Coral each stone, 60 An Indian haft, 20,000 Scarlet cloth, 10 spans, 20,000 If sold to the Karankeas in retail, 30,000 Light yellow cloth, nearly the same as scarlet; blue, not so high. Paper, per sheet, 40

AFRICAN PRODUCE. A minkalli of gold (12s. 6d. Sterling) 3000 Ivory, the very largest teeth, each, 10,000 Ivory, the medium size, 7000 —the smaller, 3 or 4000

Mr Park here learned, that a large river called the Ba Nimma, from the Kong mountains, falls into the lake Dibbie. It is not quite half the size of the Joliba, and receives a smaller stream, called the Ba Miniana. Jimmie is not situated on the Niger, but at the confluence of the Nimma and Miniana. He learned that the northern bank of the river was occupied by the Moors at Tombuctoo only, and in other places by native tribes, called the Soorks, Mahinga, and Tuarick.

In the course of his residence here, five more of the party died, among whom was Mr Alexander Anderson, his near relation and intimate friend, whose fate he deplores in the most pathetic terms. "No event," says he, "which took place during the journey, ever threw the smallest gloom over my mind, till I laid Mr Anderson in the grave. I then felt myself as if left a second time lonely and friendless among the wilds of Africa." He had now with him only Lieutenant Martyn, and three soldiers, of whom one was deranged; yet he writes to Lord Camden: "Though all the Europeans who are with me should die, and though I were myself half dead, I would still persevere."—"I shall set sail to the east with the fixed resolution to discover the termination of the Niger, or perish in the attempt." One circumstance which tended to confirm him in this enthusiastic determination, was an opinion which he had eagerly imbibed, that the Niger was the same river with the Congo, and that a few months navigation down its stream would bring him to the Atlantic.

Park sailed from Sansanding; and accounts of his progress were long anxiously looked for. It was not till 1806, that some unfavourable rumours reached the coast by means of the native traders. As these gradually increased, Governor Maxwell, with a laudable anxiety, began to consider the means of inquiring into their truth; and he was fortunate enough to engage the guide Isaaco, to proceed on an expedition of inquiry. It so happened, that Isaaco, near Sansanding, met with a person named Amadi Fatouma, whom Park had taken with him as a guide down the Niger. From him he received a journal containing the particulars of the death of Mr Park. The correctness of this narrative, as well as the fidelity of the guide, have been called in question on very plausible grounds. But the great length of time which has now elapsed, renders it vain to doubt that this illustrious traveller must have perished on his voyage down the Niger.

V. We shall now take a view of the questions of the which yet remain to be solved in regard to the geography of this continent. These may be chiefly classed under the three following heads: 1. The course of the Niger beyond Tombuctoo, and its termination. 2. The course of the river, called by Ptolemy, the Gir. 3. The precise source of the Egyptian Nile, and its communication, if any, with the Niger.

1. The Niger, according to the information of Mr Park and Major Houghton, rises at Sankari, in the high country on the frontier of Manding. Thence it is ascertained, by ocular observation, to pursue a course of about three hundred miles to Silla. Then, by undoubted information, it flows four hundred miles farther to Houssa. Positive testimony here deserts us, unless at the single point of the ferry in Cassina; but all accounts, both ancient and modern, seem to agree in stating, that there is a continuous river-course from thence to the eastern extremity of Wangara, being nine hundred and seventy geographical miles; which makes an entire course of sixteen hundred and seventy. (Rennell, Illustrations of Park, chap. vi.) In Wangara it forms several lakes, and that country is entirely surrounded and intersected by its branches. It would appear from Edrisi, as if there were a communication between these waters and the lake of Kauga, undoubtedly the same described to Mr Browne under the name of Fitré.

There is thus no doubt of a continued stream from the source of the Niger in Manding, to the eastern extremity of Wangara. The next question is as to the direction in which it flows. This is established to be from west to east, by ocular observation, as far as Silla; by highly probable evidence as far as Houssa; and Major Rennell undertakes to prove that it follows the same course as far as Wangara. He quotes the testimony of a Moorish merchant, who had visited Houssa, and told Mr Beaufoy, that persons sailed thence to Ghinny (Gana) still with the stream. The description given by Edrisi, of Wangara, and partly also of Gana, is that of a country environed, intersected, and, during the rainy season, inundated by the waters of the Niger. These waters, therefore, spread over this extensive surface, and partly formed into lakes, may, he conceives, be entirely evaporated. It is possible, however, that a part may flow still farther eastward, and be lost in the lake of Fitré. (Illustrations of Hornemann, chap. 3.)

This opinion, at one time generally received, has of late been strongly controverted, and other conjectures of a very opposite nature brought forward. One, which seems very prevalent in Africa, is, that the Niger flows eastward, till it joins the Egyptian Nile, with which it forms one and the same river. We shall reserve, till we treat of the Nile itself, the reasons which seem to throw just discredit upon this opinion. But another hypothesis, which has just attracted great attention, is that so zealously adopted by Mr Park,—the hypothesis that it joins the Congo, or Zaire. This opinion certainly deserves consideration, both from the great name by which it is supported, and from the influence it may have upon future enterprizes of discovery. We may premise, that Mr Park's opinion was not derived from any facts observed by himself, but was adopted from a Mr Maxwell, an African trader, who had examined the Congo, and made a chart of its lower extremity.*

The first ground on which the Congo is identified with the Niger is its magnitude, which seems indeed to be stupendous. It is stated to be ten miles broad near its mouth, and the flood which it discharges into the sea to freshen the waters for upwards of thirty miles. These facts, which had been nearly forgotten in Europe, were well known and repeated, even with exaggeration, by the Portuguese, during their era of discovery. Lopez and Merolla describe it as almost thirty miles broad at its entrance, and as freshening the waters to the distance sometimes of eighty miles. (Atley's Collection, III. 296.) It is farther stated by Mr Maxwell, that it swells considerably, some time after the Niger is in flood, and before any rains have fallen to the south of the equator. Hence, he infers, that its sources lie far to the north; and this, combined with its extraordinary magnitude, and the mystery which involves the termination of the Niger, is conceived to establish a strong probability, that the two rivers are one and the same.

To this hypothesis it is objected by the able editor Objections of Park's last Journal, that we should thus assign to this Hypothesis the Niger a course of upwards of 4000 miles; a highly improbable length, and which would make it by much the largest river in the known world. But this argument does not appear to us decisive, since a great river, once formed, must continue to flow till it finds a receptacle. We attach much more importance to the next observation,—that the Niger could not reach the Congo but by crossing that great central chain of mountains, which, according to various concurrent testimonies, extends across the whole breadth of Africa.† It cannot be denied indeed, that so powerful an agent might effect a breach even in the most formidable mountain wall. Both the Indus and the Ganges cross the liminary chains of Indostan in their way to the ocean. But the important consideration appears to us to be, that so great a chain as that which crosses Africa is always found to rest on a very elevated base. The Indian rivers have their rise and early course along the highest table-land of Asia, which leaves only the mountain barrier to penetrate. It seems inconceivable, how the Niger, after descending for more than a thousand miles along the level plain of Nigritia, should ever reach the elevated

---

* For a full account of this hypothesis, and the observations and reasonings by which it is supported, see Quarterly Review, Vol. XIII. p. 120-51.

† The observations of Mr Park, corroborated by the succession of rivers from the south falling into the Joliba and Dibbie, extend it from Sierra Leone to nearly opposite Tombuctoo. Thence the evidence collected by the Association, (1790, ch. 8.) carry it west as far as Cassina. Ptolemy, Leo, and Edrisi, unite in describing Wangara as barred on the south by very lofty mountains. We are thus led to that portion, the most elevated of all, which gives rise to the Nile, and extends through Abyssinia to the Arabian Gulf. plateau on which such a chain must rest. We may likewise observe, that there is a total absence of all testimony, even the faintest rumour, in favour of this hypothesis. Yet both Ptolemy and the Arabians were well acquainted with the rivers of Wangara, and the extensive knowledge possessed by the former renders his silence, of itself, nearly decisive as to this question.

Let us now consider, whether this hypothesis is calculated, or is necessary, to account for the flooding of the Congo. It supposes, that the autumnal swell of this river is produced, like that of the Nile, by the rains of the northern tropic. Congo, too, being about the same distance with Egypt from the region of these rains, (which are bounded on one side by the equator, on the other by the tropic) it follows, that the Niger, in Congo, ought to exhibit very nearly the same phenomena as the Nile in Egypt. The magnitude of the rise, and its period, ought to be as nearly as possible the same. The average rise of the Nile is stated at thirty feet, which is also that of the Ganges, the Senegal, and, we believe, of all the great rivers fed by the rains of the northern tropic. The Niger-Congo, which has three-fourths of its immense course through the region of these rains, ought unquestionably to experience an equal rise. We learn, on the contrary, from Mr Maxwell himself, that it never exceeds nine feet, not a third of that which takes place in the other streams. It thus produces no remarkable effect on the aspect of the river, which in September, when at the lowest, exhibits all the appearances of full flood. There seems no conceivable reason, why the same cause, acting in circumstances so nearly similar, should produce an effect so greatly different.

Let us now consider the period. The tropical rains commence early in June. On the 17th the Nile begins to rise, and by the beginning of August all Egypt is laid under water. The Niger-Congo, indeed, would double the length of the Nile, and, though both its portions are much more rapid, we are willing to allow it double the period. The current of the Congo exceeds six miles an hour,—the usual rate, we believe, of flood current. This would carry the rise, from its source to its embouchure, in somewhat less than a month. The fact, however, is, according to Mr Maxwell's statement, that not the slightest rise takes place till the end of September, nearly four months after the northern rains have set in, at a period when these rains have entirely ceased, and when the Nile, after three months of inundation, has begun to retire. The discrepancy, therefore, is as complete with regard to the period as to the magnitude; and the Congo, instead of exhibiting, as has been imagined, phenomena which imply a connection with the Niger, presents none that do not run directly counter to such a supposition.

Having thus proved that the floods of the Congo bear no correspondence to those of any river passing through the northern tropic, it can scarcely be required, that, with such imperfect means, we should attempt to investigate their real causes. Were we called upon to form a conjecture, we should, with great diffidence, state the following: The perennial fulness of the Congo may easily be accounted for by supposing it to be fed from the region of perpetual rain immediately under the line. In fact, its direction from the E. N. E. would lead our views to this very quarter, and particularly to the southern side of that mighty chain, which pours down the Abiad, the Misselad, and other great streams of northern Africa. This would give also a very long course to account for its magnitude. The floods and tornadoes, which, in the region above mentioned, are never wholly intermittent, experience an increase of violence, as the sun passes the equinox, which seems sufficient to account for the slight swells which, in March and September, take place upon the Congo. Their period and magnitude do not seem to indicate the action even of the south tropical rains. The Portuguese, indeed, describe the main branch of the Congo as flowing from the south; but probably upon very imperfect information; perhaps solely from their erroneous ideas relative to the position of the Dembea, its imagined source. D'Anville, in his map prefixed to Labat's Ethiopic Occidentale, observes, that the course of all the rivers which combine in forming the Congo or Zaire, is extremely uncertain.

Upon the whole, then, when we consider that this hypothesis rests upon no evidence whatever, but is contradicted by the statements of well-informed writers; that it is inconsistent with the probable structure of the African continent, and completely so with the actual phenomena of the Congo, as reported by the author of the hypothesis himself; we cannot see any ground on which it can be retained, even as a plausible conjecture, in geographical science.

Another hypothesis, nearly similar, is advanced by M. Reichard, an eminent German geographer, who makes the Niger to fall into the sea, through a number of estuaries, not yet explored, in the Gulf of Benin. But this, though a more plausible notion, is wholly destitute of any positive evidence in its favour; it is equally liable with the former hypothesis to the objections arising from the direction of the great central chain of mountains; and besides, the supposition that the estuaries in question form branches of one great river, is purely conjectural.

Reverting then to Major Rennell's hypothesis: He supposes, that the Niger, after a long easterly course from Tombuctoo, loses itself among the lakes of Wangara. To this account of its termination it has been strongly objected, that there is no receptacle or lake in Wangara, sufficient for the vast mass of waters which must thus be supposed to flow into it. But that lakes do exist in that country, is made certain by the accounts of the Arabian writers; and as none of these writers, whatever M. Reichard may imagine, give any precise statements as to their magnitude, there is thus room to suppose, that some of them may be large enough to receive the waters of this mighty river. Were we therefore obliged to make a choice among these hypotheses, we should undoubtedly prefer Major Rennell's, as on the whole, the most probable. Our researches, however, have led us to form a hypothesis of our own; and we shall now proceed to explain it, in the hope of at least being able to bring into view some curious facts which seem to have been overlooked in the discussion of this interesting question. Briefly, then, our idea is this—That the great river-course which stretches across Africa consists in fact of two rivers, to both of which the name of Niger has been given; that one of these flows eastward by Sego and Tombuctoo, the other westward through Wangara and Cassina; and that these two rivers, at some intermediate point, not far from the modern position of Houssa, unite in a common receptacle.

The first question, then, relates to the course of that portion of the Niger which passes through Cassina, a country whose limits include the kingdom so celebrated by the Arabians, under the appellation of Gana. The flourishing states which the Arabs had founded in these regions, and the extensive trade which they carried on, gave them copious means of information. Now, these writers decidedly and repeatedly state, conformably to our hypothesis, that the great river flowing through this part of Africa, and which they call "the Nile of the Negroes," runs westward. Major Rennell indeed supposes, that this idea may have been formed by unduly extending some rivers to the west of Nubia. But we have already shewn, that Nubia was very imperfectly, and the countries to the west of it scarcely at all, known to the Arabians. Gana, the metropolis of their empire, the emporium of their commerce, itself situated on the Niger, was doubtless the central point whence their information respecting that river would be derived. Abulfeda uniformly calls it the Nile of Gana. It appears, therefore, very improbable that they should not have been aware how the fact stood at Gana; and we cannot see that it ought to discredit their authority, though they should make it flow in the same direction, through countries not Mahometan, and thus lying beyond their sphere of observation.

The only modern testimony of an eye-witness, to the course of the Niger through Cassina, is that of the Shereef Imhammed, from whose information chiefly Mr Beaufoy compiled his account of Bornou and Soudan, published in the Proceedings of the African Association, 1790. From the circumstances there mentioned, as well as from the coincidence of his statements with the most authentic of those received from other quarters, there seems every reason to consider his authority as respectable. As this person travelled in the countries south of the Niger, he must have passed it at least twice, in going and returning; and the prodigious rapidity which he ascribes to the current, would render a mistake as to its direction impossible. His report is decided, that its course is from east to west.

Niebhur, while in Egypt, collected some valuable information from the mouth of a native African, which he published in the German Museum; but as that work does not seem to have reached this country, we know it only by the extracts of Hartmann. It appears, however, that he is referred to only for the eastern part of the Niger, and that he states, as beyond all doubt, the western course of that river.

To this weight of ocular and historical testimony, nothing stands opposed, except the hearsay, somewhat vague, of a Moorish merchant, who had not travelled beyond Houssa. The Niger probably flows east for some distance after passing that city; a circumstance which might easily give rise to the report which the merchant received.

In support, however, of his opinion, that the Niger flows eastward through Cassina to the lakes of Wau-gara, Major Rennell urges an argument which certainly calls for consideration. If this river of Cassina flows westward, there must be a common receptacle between that place and Tombuctoo; and "we have not," says he, "heard of any such." (Illustrations of Park, chap. 2.) It may first be observed, that the tract is so completely unknown, that a receptacle might very well be supposed to exist there without any report of it being received. But we are moreover convinced, that, by diligently comparing ancient and modern accounts, we shall arrive at almost complete evidence of the existence, in this quarter, of a great lake, or inland sea, sufficient to form such a receptacle. The subject being curious, both in itself, and as connected with the question of the Niger, our readers may not be displeased to see these testimonies collected.

It seems to be now a general opinion among the best informed geographers of the present day, that the sea of the Arabians, into which they describe the Nile of the Negroes to fall, must have been a great lake, (Hartmann, 30. Pinkerton, II. 773.) In fact, the Arabic term rendered sea, signifies merely great water. It may therefore have been used in the latter signification, and was certainly so understood by one Arabian writer, (Scheabeddin, Notices, II.

---

• Edrisi, and Abulfeda, passim. Scheabeddin in Notices, Vol. II. 156.

† A very recent confirmation has been afforded in the concurrence of his description of Ashantee, or Asente, with that contained in the valuable work of Mr Meredith.

‡ This person's testimony would fall entirely, if it should prove true, that there is no such city as Houssa. (See Rennell on Hornemann, 185.) We must confess, however, our own opinion to be in favour of its existence, which seems supported by ocular testimony, while it is denied only by persons at a distance. But whenever Houssa is mentioned in this discussion, the site fixed by Major Rennell is meant, whether there be a city upon it or not.

§ To conceal nothing, we incline to suspect that this was not the case with Edrisi, but that, finding a great expanse of water at the extremity of African knowledge, he concluded it to be the ocean. He says, the Niger flows to the extremity of the west, p. 12. See also Dr Vincent's map. Ulil, a point situated on the opposite coast, might then be mistaken for an island. This will agree with Ibn-al-Vardi, a subsequent writer, who makes it a great city on the shore of the sea (or lake). It is certain that Edrisi gives land and caravan routes from Ulil to the principal cities of Africa. But, as he was not a person of very great There has not hitherto, however, been any attempt made to fix the position of this lake, according to the materials furnished by the Arabians, although their information seems sufficiently clear and precise. As Edrisi has given the routes from Ulil to several of the great cities of Africa, this will furnish the means of an approximation to its site. The route from Gana, westward, is given as follows:—To Berissa, twelve days' journey; to Tocurur and Sala, twelve; to Ulil, sixteen; in all, forty. Ulil is also said to be at the distance of a day's sail (or 100 miles) * from the mouth of the Nile. (Edrissi, 29-35.) The forty days from Gana, extended in a direct line, would reach to the vicinity of Tombuctoo, but would fall short of the eastern point of the lake Dibbie, much more of a day's sail beyond it. But a variety of details given by the Arabian writers fix the position much farther to the east, and even somewhat short of the reported site of Houssa. Ulil is said to be the capital of Maghara, or Mekzara, an extensive tract of desert, south of the Niger, immediately bordering on Lamlam, and even on Gana.† Schebebeddin states the Niger to flow not to the ocean, but to the extremity of the inhabited part of Djennawa (Gana.) At this point, accordingly, Major Rennell, in his map 1790, has laid down a desert of ten journeys or more, on the authority of Imhammed. Ulil, however, is still more precisely fixed by the assigned distance (Edrissi, 40) from Audagost (Agades) of one month, which, in the most direct line, would not reach beyond Houssa, and supposing a probable deflexion to the south, would fall several days short of it. There remains, indeed, the difficulty of the forty days from Gana, which, as already observed, would, in a direct line, reach considerably farther west. But routes along rivers are naturally winding; and the present being apparently a land route, must, in that case, have made a considerable circuit round the shores of the lake. Agades being an ascertained position,‡ the distance from it appears to be decisive.

From these collected statements, we may consider it as the clear and united testimony of the Arabian writers, that, in an extensive tract of desert, immediately to the east of the modern position of Houssa, there exists a great lake, or inland sea, which receives the eastern branch of the Niger.§ We shall now advert to the modern authorities. Down to the period of the African Association, report is silent, which cannot at all be wondered at, as no traveller ever approached that part of Africa; and it was only learned by imperfect rumour that Tombuctoo was still in existence. It is remarkable, however, that, in the old maps by Purchas and Dapper, (followed in this respect by Sanson, and by Delisle in his earlier maps) there is uniformly laid down, at the very point we have traced, an immense lake, under the appellation of the Lake of Guarda. Upon what authority this lake is so placed, has completely eluded our research. It certainly was not that of the Arabians, whose sea was then universally understood to be the ocean. It therefore must have been originally laid down by the Portuguese, and it seems improbable that they should have done so without some kind of authority.

In consequence of the active spirit of discovery which has been roused by the Association, more distinct notices have begun to be received. Ben Ali, from whose information partly, the report 1790 was drawn up, heard it stated at Tombuctoo, that the Niger terminated in "a lake in the desert;" which is elsewhere expressed by saying that it was "lost in the sands to the south of Tombuctoo." (Report 1790, p. 222, and 1798, p. 143.) Major Rennell endeavours, consistently with his hypothesis as to the termination of the Niger, to apply this to Wangara; but no country on earth bears so little analogy to sand or desert; and it is far too distant to make the term "south of Tombuctoo" at all applicable. The report, on the contrary, agrees exactly with the Arabian notices.

Mr Jackson expressly states, that "fifteen journeys to the east of Tombuctoo, (that is, three or four beyond Houssa) there is an immense lake called the 'Bahar Soudan, or Sea of Soudan.'" In describing the boundaries of Tombuctoo, he again mentions it as "a lake formed by the waters of the Nile Abeede, of which the opposite shore is not visible." He mentions a number of particulars respecting it; describes it as inhabited by a particular race of people, to be presently noticed, who navigate it with large decked vessels, containing from 100 to 150 men, which have sometimes made their appearance at Tombuctoo.

---

* So understood by D'Herbelot, Bibliothèque Orientale, art. Ulil. See also Hartmann, Introd. 119. † Edrissi, 28, 36, 41. Ibn-al-Vardi, Notices, II. 351. ‡ See the data on which Major Rennell fixes it, Afric. Assoc. 1790, p. 119, 120. § The salt-pits of Ulil have been a subject of perplexity to modern geographers, as Nigritia has always been dependent for that article upon the interior of the desert. Considering that the knowledge of the Arabians expired here, it appears to us a very easy supposition, that Ulil, a great commercial city, (as it is described by Ibn-al-Vardi) might, in consequence of becoming a mart for the salt of the desert, be supposed to produce it. Cadamosto, in fact, describes a great salt trade upon this very line, from Tegazza, by Tombuctoo, towards Melli. Ramusio, I. 100. He adds, that the exchange was conducted on the bank of a great water, which might have been supposed the sea, had it been salt; that large boats came from certain islands, &c. The period of thirty days' travelling from Tombuctoo is too great; but when we consider that the salt was carried on men's shoulders under a burning sun, they may well be supposed not to exceed half the rate of camel travelling. There seems, therefore, on the whole, a great coincidence with Edrisi.

|| The barche grandi of Cadamosto? We find no mention of any other kind of vessels than canoes on any other part of the Niger. Mr Park, during his first journey, received no accounts of any lake beyond the Dibbie. But during his last residence of two months at Sansanding, he obtained the following information: "One month's travel south of Baedoo, through the kingdom of Gotto, will bring the traveller to the country of the Christians, who have their houses on the banks of the Ba Sea Feena; this water they describe as incomparably larger than the Dibbie, and that it flows sometimes one way, and sometimes another." (p.168).

The expression "south of Baedoo" must either be a mistake, or understood in a very qualified sense, as the kingdom of Gotto extends eastward along the course of the Niger. Now, from Sansanding to Silla, two days; to Tombuctoo, fourteen; to the sea of Soudan, according to Jackson, fifteen; in all thirty-one: thus, we have Park's month; and the two statements coincide surprisingly.

Our readers are probably startled by the expression "country of the Christians." Jackson, however, makes a similar report. He says, "On its eastern bank (of the sea of Soudan) begins the territory of white people, denominated by the Arabs, (N'sirrath) Christians, or followers of Jesus of Nazareth." He adds some peculiarities of their dress and customs, and observes, that they are neither Arabs, Moors, Negroes, Sheliuhs, nor Berebbers. Considerable light is thrown upon this subject by Hornemann, who, it appears, had heard much of a white and Christian nation as being in the vicinity of Soudan and Tombuctoo. He ridicules the idea of their being actually Christian; and states, that though the word Nazari has properly that signification, yet the Arabs familiarly give it to all who are not Mahometans. The nation in question he describes to be a branch of the Tuarick, called Tagama, who are white, (this term is probably comparative), and Pagans. (p. 110—19.) Upon Hornemann's data, Major Rennell has placed the Tagama Tuarick near the very point where Jackson places his Sea of Soudan. It is thus clear that there is, in this very quarter, a nation known in Africa by the term usually appropriated to Christians, which removes every fabulous appearance from the statements of Jackson and Park, and establishes a striking coincidence between them.*

Neither of these travellers mention any particular eye-witnesses; but their language implies, that they had met with such, and both allude to the existence of this inland sea, as a fact of complete certainty and notoriety. So grand and obvious a natural feature, which required only eyes to see it, seems to leave no room for mistake. Neither Park nor Jackson had the least idea of any coincidence between their respective statements, and still less with the Arabian writers; so that we have thus three independent testimonies from opposite quarters, meeting exactly in the same point. Nor does there, so far as we know, exist any evidence at all respectable to the contrary.

Applying then the existence of this lake to the question of the course of the Niger, we may observe, that every one who describes it, seems to labour for words to express its magnitude; which is farther manifested by its having been repeatedly mistaken for the ocean. If the sail of a hundred miles to Ulil, be supposed to give the breadth, this will equal it to the Caspian; even half that space will place it on a level with the Aral. These seas, though placed in a comparatively cold climate, where evaporation must be less active, receive, however, some of the greatest rivers of Europe and Asia. In the present instance, the power of evaporation must be extraordinary, as every account represents the heat to be intolerable, even to those who are accustomed to the burning sands of the Sahara. (Afric. Assoc. 1790, p. 124. Cadamosto, as above.) There seems, therefore, sufficient provision made for disposing, by evaporation, of the two great rivers, which, under the common name of Niger, seem to discharge themselves into this receptacle.†

2. The river, to which Ptolemy has given the name of Gir, has not attracted the same attention among modern geographers, as among the ancient. Mr Pinkerton supposes it to be the Bahr Kulla of Browne, or the river, which, flowing from the westward, meets the Niger, and is lost in the same receptacle. On examining, however, the map, and descriptions of Ptolemy, we do not find that he considers these rivers as forming one line of stream, or even as communicating together; but as quite distinct, both in their source and termination. This circumstance, joined to their relative position, appears to leave no room for hesitation in adopting the opinion of D'Anville and Rennell, who identify the Gir with the river of Bornou. To this, we apprehend, must now be added the Misselad, which, from its relative course, must certainly communicate with the other. Mr Beaufoy's informants, indeed, clearly identify them, by stating the river of Bornou to rise from the same source i.e. vicinity with the Egyptian Nile. Major Rennell seems to sup-

* Mr Barnes, formerly quoted, states, that the Niger discharges itself into a large lake; that he has heard from the black traders, that there are white inhabitants upon the borders of this lake; and has been told by people who have seen them, that they dress in the style of Barbary Moors, and wear turbans (which agrees with Jackson, p. 262.) but do not speak Arabic. Report of Committee of Council, 1788.

† We have not included Ptolemy among our authorities, for reasons already stated; but it may be expected that we should prove that he advances nothing inconsistent with the above hypothesis. We observe, that the Lake of Lybia, which D'Anville and Rennell make the termination of the Niger, is not placed by Ptolemy upon the Niger at all, but upon one of its derivations or adjuncts, IV. 6. He gives nothing, therefore, which can be a termination, except the lake of Nigritia, to reach which it must flow westward. If this lake be supposed to be the Sea of Soudan, then the northern derivation, on which is placed Tuccaba (Tombuctoo?) will be the Niger known to Europeans. We repeat, that we lay very little stress upon this interpretation; but if anything can be made of Ptolemy, it seems to be this.

‡ Generally speaking, all rivers which arise in the same chain of mountains, are considered, in Africa, as having one source. pose that they run in different directions, and meet in the lake of Fittré. But the informants, both of Mr Beaufoy and Dr Seetzen, state distinctly that the river of Bornou flowed from south to north; to which the former added, that it was lost in the sands of the desert of Bilma. We have here the testimony of three respectable eye-witnesses, to which nothing stands opposed. The lake Fittré, besides, being in its permanent state only sixty or seventy miles in circumference (Hornemann), is ill fitted to be the receptacle of two such rivers. Considering then the whole as one, it will form a course from S. E. to N. W. of not much less than a thousand miles in direct distance. No portion of this mighty stream has ever been explored by any European; yet its course, and above all its termination, must be an object of great geographical curiosity. All that seems now ascertained, is, that, after traversing the empire of Bornou, it enters the vast desert of Bilma, out of which it never issues.

3. With regard to the question whether the Niger and the Nile form a junction, Mr Hornemann mentions, that all the persons with whom he conversed, informed him, that the Niger flowed by Darfoo into the Nile, or Bahr-el-Abid, and was, in fact, the same river: "he could not find a single person who said to the contrary." (p. 115—17.) Mr Jackson says still more pointedly: "In the interior of Africa, there is but one opinion as to the Nile of Egypt and the Nile of Tombuctoo, and that opinion is, that they are one and the same river." He adds: "The Africans express their astonishment whenever the Europeans dispute the connection of these two rivers." (p. 264.) Notwithstanding such positive assurances, it seems incontrovertibly proved by Major Rennell, that the opinion in question is totally erroneous. Herodotus, indeed, mentions it, but Ptolemy, whose information was greatly superior, clearly fixes the source of the Nile among the mountains of the Moon, in a direction nearly south from Egypt. The same information was received by Mr Browne, during his residence in Darfoo, at a few hundred miles from these sources. He learned also, that all the rivers, for a great distance southwards, flowed, not east towards the Nile, but quite in the opposite direction. What is still more important,—his route intersected the only line by which there was any likelihood of the Niger joining its waters to those of the Nile. A very little way south, the country begins to rise into the great chain of the Jibbel-el-Kumri, forming a complete barrier to the passage of a river which must previously have descended to the lowest level of the plain of Soudan.

But, although this point may be considered as settled, it remains still an interesting question, whether the two rivers may not communicate, by some intermediate channel, so as to form a continued navigation from Tombuctoo to Cairo. Mr Jackson positively asserts, that he knew a credible witness, who boasted of having actually performed this voyage. It is probable, that both the Misselad and the Kulla rise very near to the Nile, and afterwards communicate with the Niger; and it is possible, that at some point, particularly during the wet season, an union may take place. It is possible also, and perhaps more probable, that the two streams, like the Senegal and the Niger, may be considered as one, merely from approaching very near to each other. More precise testimony is necessary, in order to determine the question.

In taking leave of these disputed points of African geography, we may observe, that the tract which extends southwards from the Mountains of the Moon, to the vicinity of the Cape of Good Hope, forms the greatest mass of terra incognita which now exists on the surface of the globe. Reckoning from the southern frontier of Darfoo and Abyssinia, we find an extent of nearly forty degrees, or two thousand four hundred miles in length, and nearly half that average breadth, which not only has not been visited by any European, but concerning which no detailed or authentic particulars have ever been received. Yet there is no reason to suppose, that it is either very thinly inhabited, or by people in the lowest stage of civilization. On the contrary, most of the limiter districts rank with the most civilized and populous portions of this continent. On the west are the kingdoms of Loango, Congo, and Angola; on the east are Monomotapa, Zanguebar, and the Somalics. Even the heights of the Jibbel-el-Kumri are found, when crossed by the Ashantee Caravan, to abound in cattle, provisions, and people. The immediate vicinity of the Cape is barren and savage; but, in proportion as we advance northward, an improvement takes place. The tribes of Boshuanas recently visited, and still more those lying more remote, hold a high rank among the nations of Africa. Doubtless, there are deserts here also, and particularly a very extensive one, reaching north from the Orange river. But there appears no probability that they should bear any proportion to those which exist on the northern side of the continent.

VI. We shall conclude this article, with some miscellaneous details, regarding the present state and future prospects of this quarter of the globe. There is scarcely any question involved in greater uncertainty than the extent of its Population. The few estimates which have been attempted are founded chiefly on random conjecture, and exhibit an extraordinary discrepancy of results. M. Golberry, who seems to have particularly attended to this subject, hesitates not to carry the amount to one hundred and sixty millions. (Voyage en Afrique.) Mr Beaufoy seems to have entertained nearly the same idea, as he estimates the interior at a hundred millions. (African Assoc. 1790.) Mr Pinkerton, on the other hand, does not conceive that it can exceed thirty, perhaps not twenty millions. M. Malte Brun names seventy millions as the utmost conceivable limit. (Precis de la Geographie. II. 560).

The only principle on which any thing like an approximation can be founded seems to be this,—that the state of industry and civilization throughout cultivated Africa is so nearly the same, that the ratio of the ascertained density of any one portion may be extended to the whole. Golberry has formed a calculation, apparently pretty correct, by which he makes the square surface of Africa to amount to one million six hundred thousand square leagues, of twenty-five to a degree, which is equal to upwards of nine millions of square geographical miles. These, at a moderate European rate, would give a population of four or five hundred millions. There are many circumstances, however, which must materially reduce the African standard. The most prominent is the immense extent of its deserts. The Sahara, with all its adjuncts, equals perhaps the whole of Europe. Its Oases, even the larger ones of Fezzan and Darfour, though they relieve the aspect of sterility, do not materially change its general character. The former is calculated to contain only seventy-five thousand souls, the latter two hundred thousand; which will not exceed four or five to the square mile. Even the most fertile regions of Barbary and Nigritia are deeply indented with desert. We cannot, therefore, consider as exaggerated the estimate of Mr Pinkerton, that a full third of its extent is of this description. Of the remaining six millions, we think we may allow a third for those savage tracks, which, being devoid of cultivation, do not materially affect the general populousness of the continent. There will then remain four millions of square miles for cultivated Africa.

Mr Penny, a trader, who was intimately acquainted with every part of the African coast, being asked, in the Committee of Council (Report on Africa and the Slave Trade, 1788,) his ideas upon its populousness, answered as follows: He thought every one who had visited Africa and America must have been struck with the appearance of nearly an equal population on the coast of Guinea, and in the State of Virginia. Virginia contained then eight hundred thousand inhabitants, upon thirty thousand square miles, which is $26\frac{3}{4}$ to the square mile.

M. Golberry is the only African traveller who has furnished any precise data upon this head. In the course of his work, he gives the population of a number of states upon the Gambia and Senegal, as ascertained by the French colonial administration; and we shall throw the results into the following table:

| Districts | Vol. p. | Sq. Leagues | Inhabitants | |-----------|---------|-------------|------------| | Cayor | II. 102 | 2000 | 180,000 | | Sin | II. 109 | 140 | 60,000 | | Salum | II. 112 | 1500 | 300,000 | | Barra | II. 158 | 252 | 200,000 | | Bambouk | I. 432 | 1000 | 60,000 |

This gives 163 to the square league, or $26\frac{3}{4}$ to the square mile;—a result, the coincidence of which with the estimate of Mr Penny, is somewhat remarkable. Probably, therefore, we may not err very widely in fixing this as the average amount of population along the coast of Guinea, nor even in extending it to the whole of cultivated Africa. If some districts fall short of this standard, there are others of considerable extent,—as Egypt, Houssa, Tombuctoo, Whydah,—which must greatly exceed it. Four millions, then, of square miles, at the above rate, would give a population of a hundred and six millions; to which some addition must be made, though it would be difficult even to form a guess as to its amount, for the more desert and savage districts.

The Commerce of Africa has always presented Commerce some peculiar features. The first and most prominent is the immense scale of land conveyance. Africa is perforated by no arms of the sea; and her rivers, though large, are imperfectly navigable, and do not flow in a convenient direction. It is by land chiefly that the merchant must find his means of exchange; and his route must lie, not on smooth roads through smiling countries, but over an immeasurable expanse of trackless desert, where whirlwinds of sand threaten to overwhelm him at every step. To guard against these dangers, as well as to enliven the dreary scene, merchants were naturally led to form themselves into companies, which are known under the appellation of Caravans. The camel, an animal which nature seems to have formed expressly for travelling through the desert, is universally employed. The rate of progress is three miles an hour, and seldom more than seven or eight hours in the day. At each of the oases, a stay of at least two days is made, for the purpose of obtaining fresh supplies of water and provisions. The following are some of the principal routes pursued by these caravans: 1. From Mourzouk, the capital of Fezzan, to Cairo; a route of about forty days. The halting places are Siwah, Augila, and Temissa. 2. From Mourzouk to Boronou; a journey of fifty days. The track is through the deserts of Bilma and Tibesti; and the principal halts are at Temissa, Domboo, and Kanem. 3. From Mourzouk to Cashna. This occupies sixty days, through Hiatts, Ganatt, and Agades. 4. From Fez to Tombuctoo, a journey of fifty-four days; but as sixty-five besides are spent in rest, the whole number of days required is one hundred and twenty-nine. The stations are Akka or Tatta, the general rendezvous; Tegazza, and Arawan, or Arean. Another route, along the sea-coast, leads to the same point by Wedinoum, Cape Bojador, and Guatala. 5. and 6. The caravans from Sennaar and Darfour to Egypt. These do not travel so regularly as the others; and an interval of two or three years often elapses. The caravan with which Mr Browne travelled consisted of five hundred camels; but the number often reaches two thousand.

With regard to the objects of trade, and particularly of export, these have presented a remarkable similarity, in all ages, and over every part of this continent. The first, and by far the most considerable, has always unhappily been the trade in the human species. Africa has in all ages been ransacked for those unfortunate beings, whose degradation was to be the instrument of pleasure or avarice to the lords of the other portions of the globe. Besides the European part of this atrocious traffic, (ere long, it is to be hoped, to be fully, and for ever relinquished,) there has long been a similar trade for the supply of Barbary and some countries of the East.*

* The history and actual state of the trade in slaves, for European purposes, will be given under the head Slave Trade in this Supplement. Africa. times are drawn chiefly from the barbarous tribes inhabiting the central range of mountains. The whole number for this supply is estimated at twenty thousand. Part of these captives undergo a shameful mutilation, with the view of being placed in the seraglios of the East. In other respects, their lot is much milder than that of their West Indian brethren. They are used as domestics, are well treated, and often raised to posts of distinction.

Next to the human species, the article of exchange which Africa produces most abundantly, is gold. It seems probable that its mountains contain mines of this coveted metal, more ample even than those of the new world. All the rivers which descend from the Jibbel-el-Kumri, on both sides, and throughout its whole extent, roll down sands of gold. The great repositories are Wangara, the country near the sources of the Senegal and Niger, and that immediately behind the Gold Coast. Its abundance in the districts along the Niger is said almost to surpass conception. Mr Jackson states, we suspect with exaggeration, that at Tombuctoo, it is often exchanged for its weight in salt, tobacco, or other valuable commodities, (p. 260.) About the beginning of the present century, the value of that exported from the gold coast, is said to have amounted to between two and three hundred thousand pounds. (Wadstrom on Colonization.) We have not been able to obtain precise information respecting the present state of this trade; for the prejudices of the mercantile system prevent gold from being classed with other commodities in the amount of exports and imports. The slave captains in their examinations before the Committee of Council, (1788) generally state, that it was too costly to be an object of trade; that it sold at L.4. an ounce, and that they did not find it their interest to procure it, unless as a medium for the purchase of slaves. They admit, however, that the Dutch exported it; and the truth seems merely to be, that they found the other trade more profitable.

Ivory. Next to gold, ivory has always been the great subject of African export. Like gold, it is produced almost solely in the interior, and thence brought down to the coast. Where those vast meadows lie which feed such herds of elephants, has never been ascertained; but Europe derives from Africa its whole supply of this valuable article.

Gum. Gum, particularly Gum Senegal, forms an important branch of trade. The southern parts of the desert of Sahara contain vast forests of that species of acacia, from which this substance exudes. In the month of December, a dry and piercing wind bursts the bark, and the juice flows out. The neighbouring Moorish tribes then leave their habitations, and employ themselves for six weeks in collecting it; after which they repair to the banks of the Senegal, to exchange it for European commodities. The price is first fixed by a meeting of the merchants with their kings and chiefs. This point being settled, in a few days the Moors appear in vast bands, with their wives, children, and cattle; and for some days, a tumultuous market is held. M. Golberry calculates, that two millions of pounds of gum may be drawn from this part of Africa. (Travels, chap. vi.)

Among the valuable productions of this country, we may also mention hides and skins, particularly goatskins dyed red or yellow. These are brought by the caravans from central Africa to Morocco, whence they are exported to Europe, chiefly from the port of Mogador. Africa exports likewise ornamental and dye woods, particularly red or cam wood, to a very considerable extent. Raw hides and bees wax have, within the last few years, become considerable articles.

There are doubtless many other minor and local objects; but those now enumerated may be considered as forming the basis of African export. It is remarkable that, with the exception of a small quantity of leather, the whole consists, not merely of the unmanufactured, but of the spontaneous produce of the land. So small is the progress yet made by cultivation throughout this great continent. The imports are limited, by the limited wants of the climate and state of society. The manufactures most in demand are checked cottons, light coarse cloths, red woollen caps, a few linens and silks, and a large proportion of India piece-goods. Spirits, chiefly rum, are too much in demand; guns and gunpowder, hardware, chiefly in the form of knives, sabres, blades; brass, which is manufactured into ear-rings; coral, beads, looking-glasses, and other articles of ornament. The scarcity of salt causes a certain import from Britain, even of that bulky article.

The following tables are made up from information laid before the House of Commons, 13th June 1812:

| ARTICLES | 1805 | 1806 | 1807 | 1808 | 1809 | 1810 | |-------------------|--------|--------|--------|--------|--------|--------| | Elephants' Teeth | 10,285 | 11,331 | 10,693 | 14,314 | 16,270 | 16,488 | | Gum Arabic | 3,947 | 4,190 | 4,333 | 6,203 | 3,536 | 8,420 | | —— Senegal | 16,223 | 17,797 | 14,455 | 12,270 | 21,550 | 46,384 | | Hides, raw and tanned | 1,257 | 990 | 2,472 | 2,308 | 4,143 | 12,714 | | Skins and Furs | 10,838 | 8,318 | 34,124 | 36,401 | 23,800 | 76,437 | | Wax, Bees | 2,009 | 3,436 | 8,297 | 13,817 | 20,436 | 12,996 | | Red Wood | 51,311 | 52,160 | 27,753 | 15,908 | 47,791 | 26,058 | | All other articles| 10,960 | 17,926 | 19,921 | 42,055 | 47,125 | 57,790 | | Official value | 106,845| 115,948| 122,948| 143,276| 184,651| 257,387| | Real Value | 193,034| 226,396| 242,747| 374,306| 383,926| 535,577|

11 ### EXPORTS

| ARTICLES | 1805 | 1806 | 1807 | 1808 | 1809 | 1810 | |-------------------|--------|--------|--------|--------|--------|--------| | Brass and Copper | 12,085 | 10,315 | 6,496 | 6,979 | 3,755 | 3,231 | | Cotton Manufactures| 285,408| 456,661| 303,106| 168,910| 305,652| 196,214| | Gunpowder | 27,154 | 38,179 | 21,092 | 6,567 | 8,453 | 7,887 | | Guns | 48,500 | 57,685 | 35,698 | 7,452 | 14,251 | 6,393 | | Iron and Steel | 17,703 | 44,464 | 26,870 | 28,721 | 30,302 | 19,139 | | Woollens | 78,394 | 124,510| 58,788 | 51,955 | 76,430 | 68,402 | | India Piece Goods | 324,087| 390,605| 175,963| 72,444 | 84,472 | 59,967 | | Brandy and Geneva | 8,490 | 38,109 | 15,853 | 5,453 | 7,499 | 1,722 | | Rum | 56,181 | 43,855 | 10,636 | 12,066 | 11,663 | 7,551 | | All other articles| 182,573| 298,769| 143,308| 172,894| 163,522| 113,579| | Official value | 990,575| 1,433,152| 797,740| 532,841| 705,979| 484,082| | Real value | 1,156,985| 1,655,042| 1,022,745| 820,194| 976,872| 693,911|

On adverting to the future prospects of Africa, it is proper to say something, in the first place, as to the means of its more perfect exploration. It seems, then, to be now universally admitted, that no reliance can be placed on the common plan of solitary enterprise, or where the traveller has only one or two attendants. Nor is the device of assuming the character of a Mahometan likely to be repeated. Mahometanism is not a mere creed; it is a system, which regulates the whole train of thoughts and habits, and which moulds from infancy even the outward aspect and deportment. It is vain to expect, that the study of Arabic at an European seminary, or even a few months residence in the East, could mould all the habits of an European into a shape, which would enable him to deceive the watchful eyes of Musulmen. It appears also, that, while the character of a professed Christian is sufficiently detested, that of one who assumes a false semblance of Mahometanism is viewed with a much higher degree of reprobation. To the disguise, therefore, which Hornemann assumed, it is probable that this unfortunate traveller was partly indebted for his melancholy fate. The only mode, then, it would appear, of penetrating into Africa, with any prospect of success, is to carry an armed force, which shall not be so large as to excite alarm, and yet sufficient to guard against desultory or sudden attack. Such an expedition would the less tend to excite jealousy, that Africa is habitually traversed by parties of this description for commercial purposes. Yet it is remarkable, that the two attempts which have been made upon this plan have terminated still more fatally than those in which no such precautions had been taken,—we allude to the last expedition of Park, and that sent by Lord Caledon from the Cape of Good Hope. These failures, however, we regret to say, seem to have been the result of such obvious imprudence, as not to authorize any inference against the general superiority of this plan. Park, by unluckily exposing his party to be destroyed by the rainy season, reduced himself to the same unprotected state with former travellers. The expedition from the Cape was too small; and yet, till they gave way to supine security, and sacrificed all the advantages of their strength by dividing it, no hostile attempt appears to have been made.

The editor of Park's last journal recommends an expedition on a large scale, to be composed chiefly of African soldiers enlisted at Sierra Leone; but we cannot help feeling some doubts as to the safety of an expedition, in which the majority should be Africans. And as this writer suggests the propriety of employing a traveller from the East, we would add, that his attendants should rather be Scapoys than Africans.

The plan upon which the main reliance for the civilization of this region seems at present to be placed is, that of an extensive system of colonization. It is observed by Mr Macaulay, Ex-secretary of the African Institution, that it is "necessary to the rapid growth of improvement, that men should be brought together in considerable bodies; that they should be protected by just laws; and that they should enjoy the means of instruction." This, he conceives, can only be effected in Africa by means of a colony, of some extent, composed of natives. The difficulty of collecting a large native population is admitted, indeed, to be very great. The Africans, from many causes, and particularly from superstition, will not voluntarily resort to such a colony. A plan had, therefore, been suggested, in reference to the co-

---

*These tables exhibit a curious view of the effects of the abolition. Almost all the articles of import have remarkably increased, and their general amount has more than doubled. The exports have necessarily suffered a considerable reduction, which, however, has taken place almost wholly in the articles of guns, spirits, and (we know not exactly why,) of India piece-goods; while British manufactures maintain nearly their former amount.*