Chapter I: Part 1

THE NEGROLAND
OF THE ARABS.

THE NEGROLAND
OF THE ARABS
EXAMINED AND EXPLAINED;
OR,
AN INQUIRY INTO THE
EARLY HISTORY AND GEOGRAPHY
OF
CENTRAL AFRICA.

BY
WILLIAM DESBOROUGH COOLEY.

LONDON:
PRINTED BY JAMES HOLMES, TOOK’S COURT.
PUBLISHED BY J. ARROWSMITH, 10, SOHO SQUARE.
* * * * *
1841.

TO SEÑOR
DON PASCUAL DE GAYANGOS.

MY DEAR GAYANGOS,

The following Essay owes its origin wholly to the extracts from the writings of Ibn Khaldún and Ibn Baṭúṭah, which you had the kindness to communicate to me. Not only did you occasion the present developement of my speculations, but you have also at all times cheerfully aided me in the researches to which they gave birth; you have placed at my disposal, as it were, your perfect knowledge of the Arabic language, and, from the abundance of your learning, have in some measure made good my deficiency. To whom, then, can this little work be dedicated so justly, as to you? Nor, while acknowledging my obligations, can I forego the pleasure afforded by such an opportunity of expressing towards you the friendship and esteem of

Yours sincerely,
WILLIAM DESBOROUGH COOLEY.

_London, March_ 3, 1841.

PREFACE.

* * * * *

The following Essay has for its object to establish the early geography of Central Africa on a solid basis. It aims at offering a clear and well-grounded explanation of the geographical descriptions of Negroland, transmitted to us by Arab writers; and, by thus connecting the past with the present, at giving an increased value to the historical information derived from the same sources. The attainment of that end will throw a steady light on the past condition of a country now awakening a general interest. It will enable us to trace some important political revolutions; to discern the nations which have stood forth politically eminent, and to estimate correctly, by means of a lengthened and authentic retrospect, the progress of civilization in Africa.

The task here undertaken has more of novelty in it than may be at first suspected. Hitherto no attempt whatever has been made to explain the Arab geography of Negroland by treating it as a whole, and as the immediate subject of investigation. Yet no department of the wide field of literature stands more in need of critical labour, or appears more justly entitled to it. The Arabs in the Middle Ages were copious and circumstantial writers, though neither profound nor exact. Geography was one of their favourite studies. The interests of trade and religious zeal led them across the deserts of Northern Africa to Negroland, of which they have left us accounts bearing in every lineament the expression of unaffected sincerity. Yet such has been the difficulty found in recognizing the places described in those accounts, that, up to this day, scarcely any addition to our positive knowledge of Negroland has been derived from the writings of the Arabs.

Ibn S’aíd, a writer of the thirteenth century, has enumerated thirteen nations of Blacks, extending across Africa, from Ghánah in the west, to the Bojá on the shores of the Red Sea in the east. Yet it is not till we arrive at the tenth of these, or Kánem, that we are able to identify satisfactorily the nomenclature of Ibn S’aíd with that of modern geography. The first nine nations towards the west, or nearly three- fourths of the whole, remain undetermined.

The Arab geography of Africa lies, at present, a large but confused heap of materials, into which modern writers occasionally dip their hands, each selecting what appears to serve his purpose, and adapting it to his views by an interpretation as narrow and partial as his mode of inquiry. Modern geographers—D’Anville and Rennell not excepted—have allowed fancied resemblances of sound to lead them far away from fact and the straight path of investigation. They have, for example, unanimously assumed the Kanó of the present day to be the Ghánah of past ages. The disorder introduced into the early geography of Central Africa by this false method of proceeding, has deprived it of all its value. It seems incapable of combining with the results of modern discovery; and instead of the harmony which ought to subsist between our present information and the ample accounts of Negroland written five or six centuries ago, we find in almost every application of the latter, the jarring consequences of false assumptions.

To give a new value to such confused materials, we must have recourse to a new and improved method of treating them.

The course here followed is, to examine the Arab authors of greatest value, and to develope, as completely as possible, the information found in them, their meaning being collected altogether from internal evidence, and without any regard to extrinsic systems. Where their statements are clear, natural, and consistent, no attempt has been made to interfere with or strain them by arbitrary conjectures. Where, on the other hand, they are obscure, absurd, or contradictory, care has been taken to inquire,—1st, What were the sources or channels of the author’s information? 2ndly, How far it must be taken in strictness, or may claim the latitude allowed to the language of ordinary discourse? 3rdly, The state of knowledge, and prevalent geographical systems in the writer’s time? 4thly, What portions may be looked upon as original or authentic, and what as founded on inference or surmise?—The point of greatest difficulty, and which demands the utmost care and perspicacity, is to distinguish between the language of experience and that of system; to separate the original information from the supplementary additions made to it, for the purpose of filling up or rounding the description, or of reconciling it with theoretical conceptions.

In all parts of the world, and in every age, the human mind moves in a constant cycle. In like stages of its progress it occupies similar positions, and goes through the same round of error. This truth is exemplified in the history of Geography, as well as in that of every other branch of human knowledge. The corrections made in our maps of Africa during the last three centuries; those made by Ptolemy in the geography of the same quarter of the globe, written by Marinus Tyrius; and those of which the great Alexandrian himself stands in need, are nearly all reducible to one common rule. The errors to which systematical geography tends, while it is not as yet founded on science, are so fully shown by experience, that we can safely derive from our knowledge of them a principle of rectification, applicable to all the materials of unscientific geography, presented to our notice, and obviously needing correction. The endeavour to trace errors according to fixed analogies, will at least lead us from mere conjecture towards a rule of reason.

The most vexatious and frequently recurring hindrance in researches of the kind here undertaken, arises from the defects of the Arabic written character, and the uncritical servility of Arab copyists and compilers. In Arabic writing, some of the characters closely resemble one another, and are distinguished only by diacritic points: the vowels likewise are chiefly indicated by points, which, like those of the former kind, are often wholly omitted. Hence it follows, that proper names, the correct reading of which cannot be inferred from the context, easily sink, when written in such ambiguous characters, from corruption to corruption, till at last they altogether cease to be recognizable. Misnomers arising in this way were perhaps often adopted in discourse, till, in process of time, they became authorized by usage. Uncertainty with respect to the true reading of proper names, besides being in itself a source of great perplexity, has the ill effect of encumbering the truth with much preliminary discussion of an apparently trivial kind. Nor are our difficulties with respect to proper names confined to the writings of the Arabs. In some parts of Africa, different races are so commingled, that five or six languages may be heard spoken within a narrow extent of territory. Travellers arriving in such a country from different quarters, and associating with different portions of the population, will naturally report in different ways the names of places. If Europeans, they variously represent, each according to the genius of his own language, articulations which, being strange to their organs, they did not invariably seize correctly. Names have been also transferred from one language to another, without due regard to the sound originally sought to be represented. Here, then, are abundant sources of confusion, which, so far, at least, as it involves the genius and construction of African languages, we are not always in a condition to clear up. Yet it is necessary, as we proceed, to endeavour to distinguish between the more and the less doubtful; and if any one should feel surprised at the attention occasionally bestowed in the following pages on proper names, let him only reflect, that the errors arising from the neglect of so many inevitable causes of variance, would at least equal what may ensue from even the least skilful discussion of them.

It would have been easy to lengthen the list of Arab authors here made use of; but little would have been gained for Geography by reference to works equally deficient in judgment and originality. The authors chiefly relied on are, El Bekrí, Ibn Khaldún, and Ibn Baṭúṭah. El Idrísí is analysed in some passages, chiefly for the sake of explaining certain discrepancies between him and the first-named writer, and to estimate the weight due to his authority. The conjectures scattered through the following pages are much fewer than the nature of the subject and the course of investigation would have permitted; but it was thought advisable to keep conjectures within bounds, even where they could not have interfered with the reasoning. Reserve and circumspection are especially necessary for an author whose conclusions differ widely from those of his precursors in the same field of inquiry, and who aims at deciding definitively questions which have long divided the judgments of the learned. Reference to the systems of recent authors has been, in general, avoided, lest the frequent recurrence of the language of refutation might give the whole an air of controversy. Pains have been also taken to abridge whatever was not strictly geographical, the object here aimed at being properly the consideration of geographical questions, the careful and satisfactory discussion of which demands a special treatise.

The mode here adopted of writing the Arabic names, is fully explained in the Journal of the Royal Geographical Society, vol. VII. p. 245; but in its application will probably be found much unsteadiness and imperfection, arising partly from the want of familiarity with it, and partly from the difficulty of setting exact bounds to the employment of a foreign orthography, which jars disagreeably with the necessary reference to the orthography of our modern travellers. The attempt at a systematic reform of the mode of writing African names, has been avoided, and therewith the alternative also, of either multiplying without end the subordinate topics of discussion, or else of allowing much hypothesis to steal in under the mask of precision.

In conclusion, it may be remarked, that the attempt here made, however successful it may be deemed—and it cannot be denied that it broaches some truths, and discloses a new and logical method of treating an interesting subject—is yet but a sketch, which remains to be filled up, after a careful examination of the numerous Arabic MSS. preserved in the public libraries here and abroad, by some one better qualified for that labour, and enjoying fairer opportunities than the writer of these pages.

CONTENTS.

* * * * *

INTRODUCTION 1

GHÁNAH, AÚDAGHOST, AÚLÍL.

Támedelt — Route to Aúdaghost — Line of Drifting Sands — 5
Position of Aúdaghost — Route to Ghánah — The Desert — Tíser —
Azawad — The Zenágah — Genéwah — Lumtúnah — Goddálah — Aúlil
— Trade of Aúdaghost — Aúkar — Sínghánah — Tekrúr — Tádmekkah
— River of Ghánah — Samaḳanda — Ghaïárú — Bokmo — Manners of
Ghánah — Ghánah near the site of Tomboktú

EL IDRÍSÍ COMPARED WITH EL BEKRÍ.

Measures of El Idrísí — His delineation of the Great River — 48
History of the Maghráwah

MÁLÍ.

Ghánah subjected by the Molaththemún — The Ṣúṣú — The People of 61
Málí — The Kings of Málí — Table of their succession

IBN BAṬÚṬAH’S JOURNEY.

Tegháza — The Desert — Aïwalátin — Kársekhó — The Ṣanṣarah — 70
Málí — Manners of the Court — Departure from Málí — Position
of that capital — Mímah — Tomboktú — Kaúkaú — The Berdámah —
Tekaddá — Káhir — Limits of Málí — Múlí and Malla — Kombori —
Yúfí — Remarks on Ibn Baṭúṭah’s Journey — Kanó not Ghánah

TEKRÚR.

Zághah — The Zagháï — Sanghee — Sungai — Sokai — Zachah — 97
Eastward Movement of the Tekrúr — Abuse of the name Tekrúr

KÚGHAH, KÁGHÓ, KAÚKAÚ, KARKAR.

El Bekrí’s account of Kúghah and Kaúkaú — El Idrísí’s hesitation 103
— More than one Kaúkaú — Karkar — Gargari — Surmise as to the
present existence of Kághó

LEMLEM, REMREM, DEMDEM, EL-LÍMIYÍN,

Various names of one people — Originally one name — Its probable 111
origin — Refers to Houssa — The modern Yemyem — El-Límiyín

NEGROLAND DIVIDED INTO NATIONS.

Ibn Sʿaíd’s list — Makrízí’s additions — Ibn Sʿaíd compared 116
with Leo — Results of the comparison — Kingdom of Kúra — Lake
Kúra — Compared with the Kowára — Interference of theory —
Recapitulation

POSTSCRIPT.

REMARKS ON HOUSSA.

Routes from Ashantí to the Kowára — The letter r often changed 140
into a ḍ — The Kambaroa — The Kadunía — The River Kowára
probably so called from the country in which the Quorrama rises

CORRECTIONS.

Page 14, note 31—for Moallakah read Maḳámah.

Page 60, note 107—for A’walílí read Awalílí.

_London, John Arrowsmith, 10 Soho Square 1841._]

THE
NEGROLAND
OF THE ARABS.

* * * * *

INTRODUCTION.

Nature has marked out, in a plain and peremptory manner, the chief lines of communication between the maritime regions of North Africa and the fertile Interior beyond the great desert. The Oasis, or habitable tract of Fezzán, south of Tripoli, projects far into the barren waste, and a journey of not more than forty days conducts thence to Kánem or Bornú. In the west, a route of equal length connects the last traces of cultivation at the back of Atlas, with the Great River of Negroland, where, winding in a long circuit towards the north, it seems to drive back desolation, and narrows the limits of the desert. The two routes here described, are those chiefly frequented at the present day by caravans proceeding to Central Africa; and the preference given to them is due to their combining such advantages of convenience and security, as must have constituted them the chief routes in all ages.

The first of these, or the road between Fezzán and the interior, may be presumed to have been frequented by the ancients. If we assume that commerce spread westward from Egypt, or if we fix our eyes on Augila or the Greek colony of Cyrene, the eastern route will certainly seem entitled to be the channel of the earliest intercourse with the Blacks. But the discussion of such questions does not lie within the scope of our present inquiry; and it will be here sufficient to observe, that so far as the Arabs were concerned, the western route, though last reached, was the most frequently trodden and most diligently explored. The stream of Arab invasion in Northern Africa flowed rapidly to the west, till accumulating between the shores of the Ocean and Atlas, it pressed on the Berber clans inhabiting all the fertile recesses of this range of mountains. Continual wars thence ensued, in the course of which the discomfited party always fled to the desert, wherein they wandered to the borders of Negroland.

It is by the western route that we have derived, through Arab writers, the amplest and earliest accounts of Central Africa. For the Arabs in Spain who cultivated letters maintained a constant intercourse with their rude but active brethren of Western Barbary, whom trade and warfare alike occasionally impelled to visit the countries beyond the desert. In the beginning of the eleventh century of our era, the hills on the south side of Wád Nún and Darʿah, or the northern portion of the western desert, were occupied by the Lumtúnah, a tribe of the Zenágah.[1] Separated from them by an uninhabited tract, were the Benú Goddálah of the same nation, whose territory, comprising the southern zone of the Ṣaḥrá, extended eastward from the sea shore to the country of the Blacks.[2] The more sterile tracts of the desert in the interior, within the limits possessed by the Zenágah, were abandoned to the wandering tribe of the Benú Masúfah, by their more powerful brethren near the coast.[3] These three tribes, inflamed with religious zeal, to which their intestine feuds had given a martial character, shaped their course northward, and being united under the name of Al Morábiṭún, or Champions of the Faith, they subjugated the fertile countries on both sides of the southern Atlas, and founded, in 1073, the empire and city of Morocco. The Al Morábiṭún, or Morabites, subsequently extended their sway into Spain, in the history of which country they figure under the name of Almoravides. But long before they carried their arms into Europe, they corresponded intimately with the polished courts of Mohammedan Spain; and while they had not yet quite relinquished the desert, nor forgotten their acquaintance with the frontiers of Negroland, they communicated their information to the inquisitive, and, for that age, well instructed Spanish Arabs. The age immediately preceding the foundation of Morocco is that in which we should accordingly expect to find the most valuable accounts of the Western Desert, and of the Negro kingdoms contiguous to it. The events of that period were calculated to bring within the reach of literary activity, full, fresh, and authentic information respecting the interior of Western Africa. It is fortunate for us, therefore, that we possess an account of that country written in the very age referred to, by one who resided at the most accomplished court in Spain; whose station in society and official rank afforded him the amplest means of satisfying his curiosity; and whose perspicuity and good sense entitle him to a distinguished place among Arab writers. The author here alluded to, Abú ʿObeïdi-llah Abdullah el Ḳorṭoby, was son of the independent ruler of Huelva. He resided chiefly in Cordova, at that time the centre of Arab refinement, and filled the highest offices in that kingdom. His account of Negroland, entitled ‘Kitábu-l-mesálek wa-l-memálek,’ or the Book of Roads and Realms, was written in the year of the Hijra 460, or A.D. 1067, just fourteen years after the first rise of the Morabites, and six before the foundation of Morocco. We shall therefore take El Bekrí as our guide while endeavouring to determine the true position of Ghánah, in his age the principal kingdom of Negroland.[4]

[Footnote 1: Wád Nún is also called by early writers Núl, or Núl el aḳṣa. Darʿah دَرعَه is also written Dirʿah دِرعَه (MS. B.M. fol. 101)—Lumtúnah لمتونه—The Berber name Zenághah زناغه was corrupted by the Arabs, as Ibn Khaldún informs us, into Ṣinhájah صنهاجه, pronounced in the west Ṣinhágah.]

[Footnote 2: Goddála, so pronounced, though written by El Bekrí Joddála جدَّاله (MS. B.M. fol. 106); by Ibn Khaldún and others, Godálah ڭداله.]

[Footnote 3: Benú Masúfah بنو مسوفه.]

[Footnote 4: For an account of this valuable author, see the recently published History of the Mohammedan Dynasties in Spain, by Don Pascual de Gayangos, p. 324. The excellent MS. in the library of the British Museum (No. 9577) there described, shall be here briefly cited as MS. B.M. A translation of El Bekrí’s Book of Roads and Realms, by M. Quatremère, has appeared in the 12th volume of the Collection entitled ‘Notices et Extraits des MSS. de la Bibliothèque du Roi.’ But the Parisian MS. is so deficient in points, that the translator, notwithstanding his learning and acuteness, has not always been able to divine the true reading. A fragment of the same work, containing what relates to Ghánah, has been translated by M. Jaubert, to whom it was sent from Barbary, and inserted in the 2nd volume of the ‘Recueil des Voyages et Mémoires,’ &c. published by the Société de Géographie in Paris. Wherever this extract, which is taken from a MS. of inferior authority, is quoted alone in the following pages, the reader may understand that the readings so adopted agree in characters with the readings found in the MS. B.M., and have in addition the vowel points.]

GHÁNAH — Aúdaghost — Aúlíl.

Previous to the foundation of Morocco, all the trade of Negroland with Western Barbary was directed towards Sijilmésah, a town on the eastern side of Atlas, eight or ten days from Fás or Fez, and in the district which is now called Táfílélt.[5] From Sijilmésah, a two months’ journey southward conducted to the nearest kingdom of the Blacks, which was that called Ghánah.[6] But in propriety of speech Ghánah was the title of the king, whose dominion, anterior to the rise of the Al Morábiṭún or Morabites, extended to Aúdaghost, a town on the southern border of the great desert, and containing a Berber population.[7] Driven from Aúdaghost, the negro king fixed his residence at Aúkár, fifteen days’ journey south-westward from the former place, and not far from the great river of the interior, called by Arab writers the Nile of the Blacks.[8] But the new capital, as well as the kingdom, was still generally known by the name of Ghánah.

A desert of forty days’ journey in extent lay between Aúdaghost and Támedelt,[9] a town of Sús el Aḳṣa, on the verge of the desert, and eleven days from Sijilmésah. The two months’ journey between Ghánah and Sijilmésah, would, if literally interpreted, place the capital of the Blacks forty-nine days distant from Támedelt. But where could a route of forty-nine days southward from the borders of Sús el Aḳṣa meet the Great River, unless in the vicinity of Tomboktú?[10] This city is distant about two months from Táfílélt, and not more than fifty days from Sús el Aḳṣa.[11] It owes all its importance to its advanced position, near the very point where the river turns eastward, after repressing the desert in its northerly course, and making its nearest approach to Western Barbary. The advantages of such a position could never have been overlooked while caravans traversed the desolate plains of the interior. The site of Ghánah then, and Tomboktú, being equidistant from Sús el Aḳṣa, both in a southerly direction from it, and both in the vicinity of the Great River, which, within the distance of fifty days from Sús el Aḳṣa, washes the desert during only a short part of its course, cannot have been far asunder. This brief argument is in reality unanswerable. Within the assigned time, a caravan travelling at the ordinary pace, could reach the Great River nowhere but in the vicinity of Tomboktú. The site of Aúkár then being near Tomboktú, Aúdaghost fifteen days distant from it towards the north-east, must have been situate in the same tract as the modern town or encampment of Mabrúk.[12] Thus the first view of the routes to the chief towns of the ancient Ghánah suggests the approximate positions of those towns. Let their positions be assumed accordingly, so as to give distinctness to our conceptions, while we follow, step by step, the routes to them across the desert; and the examination of these routes in all their particulars will, in turn, illustrate and confirm our assumptions, if they be correct.

The position of Támedelt, the starting point of the caravans to Negroland, must be in the first place determined. That town is stated to have been five days westward (or, we must rather suppose, south of westward) from Darʿah, and six days south-eastward from Iklí, the capital of Sús, which town, situate on a river flowing northward, was two days from Mésah and five from Wád Nún. Now the capital of Darʿah was five days distant from Sijilmésah, which was nine ordinary journeys from Fez.[13] Támedelt was therefore twenty days from Fez, by the road on the eastern side of the mountains, and it was also six days south-eastward from Iklí, which town must accordingly be so placed within five days of Wád Nún, and two days northward from the river of Mésah, as to allow Támedelt a somewhat westerly bearing from Darʿah. These conditions being fulfilled, the position of Támadelt will be nearly in lat. 28° 45′ N., long. 7° 10′ W., and not far from the modern Tatta.[14] The position thus assumed, though not quite free from uncertainty, will yet involve no inaccuracy capable of endangering the argument depending on it.

The starting point being ascertained, there remains no difficulty in tracing the route to Aúdaghost. The first day’s journey from Támedelt conducted to a deep well called Bír el Jemmálín:[15] the second led through a narrow defile. Then for three days the road went over the mountains of Azawwar,[16] strewed with masses of iron-stone. These mountains extend, according to El Bekrí, ten days’ journey from the ocean to the road (from Támedelt probably) to Sijilmésah. They are evidently the same chain of mountains which caravans now pass at the distance of six days from Wád Nún. It is apparent also that the road to Aúdaghost must have crossed them at a distance not exceeding ten days’ journey from the sea; and therefore could have scarcely inclined towards the east. Three days beyond the mountains was the watering place of Tendefas, and three days further a great well called Weínhílún.[17] Then another three days led to a scanty spring named Tázḳa, or the House.[18] Four days further were the brackish wells of Weítúnán, and after another four days the watering place of Aúkázenta.[19] There ended the hard desert, and the region of loose sand-hills commenced, the passage of which presented the greatest difficulty to the traveller, and was fraught with danger.

The northern limit and general direction of the sandy region of the great desert are marked out with tolerable distinctness in the narratives of some modern travellers. It was on the twenty-fifth day of his journey from Wád Nún to Tomboktú that Sidi Hamed entered the region of drifting sand.[20] This must have been somewhere between the twenty- third and twenty-fourth parallels of latitude. In the same line, or twenty-two days from El Harib, between Tatta and Darʿah, Caillié left the sand-hills on his journey from Tomboktú northwards.[21] Laing, on his way from Twát to Tomboktú, entered the sands in lat. 23° 56′ N.; and Scott crossed a similar tract, eleven days’ journey in extent, apparently from the twenty-first to the twenty-third parallel inclusive, and not far from the ocean.[22] The vast expanse of light sand thrown up into wave-like hillocks, which change with every wind, is a remarkable feature of the great desert, depending not more on the mineralogical constitution of the country than on the excessive dryness of the climate. The permanence of its general position, therefore, in spite of local fluctuations, is as certain as its extension from west to east, between the zones fertilized by rains. Now the site of Támedelt was four or five days westward from El Harib; and since on the road to Tomboktú from the former of these places the loose sand is entered on the twenty- fifth day, from the latter on the twenty-third; while from Támedelt to Aúdaghost the sands were entered also on the twenty-third day, there is reason to infer, that so far, the direction of the road from Sús el Aḳṣa to Aúdaghost differed little, if at all, from the modern road to Tomboktú. The ancient and modern roads to Negroland ran nearly parallel between the hills south of Sús el Aḳṣa and the zone of drifting sands.

The most difficult part of the journey, according to El Bekrí, lay over the ridges of loose sand, on entering which there was no water to be found till, after four days’ toil, the traveller reached the wells of Wanzamín, where all the roads to Negroland met together.[23] The roads thus said to meet that from Sijilmésah must have been those from Wergelán, Twát, and Télemsén or Tremecen. Now the roads from these places to Tomboktú all unite at Telig, about thirty days from El Harib, and our hypothesis respecting the position of Aúdaghost derives no slight confirmation from the circumstance that it gives to the meeting of the roads to that place, twenty-seven days from Támedelt, a position which harmonizes perfectly with their present junction on the road to Tomboktú. Near the meeting of the roads was a mountain, the recesses of which concealed bands of various tribes of the Zenágah, as the Lamṭah, Gezúlah, Geráwah, &c., all from the south-western extremity of Atlas, who lay in wait to attack caravans.

From the wells of Wanzamín the road continued through sand-hills in the province of Wárán for five days to a large well belonging to the Benú Wáreth, thence in two days to Agharef, and in three more to Akríri, that is to say, the reservoir of water.[24] Near this place was a mountain named Azgúnán, where caravans were in danger of being attacked by the Blacks.[25] One day further, over hot sands, lay the brackish wells of Wárán, then, for three days, fresh water was found in wells belonging to the Zenágah, and another day led over the high mountain, at the foot of which stood Aúdaghost. This town was situate in a hilly country, within the limits of the rains, and does not appear to have had any water but that of wells. It can hardly be supposed that the limit of the rains in the interior of the African continent, at a distance from the sea coast, ascends beyond the twentieth parallel of latitude.[26] Now if a distance of forty days’ journey be measured from Támedelt, so as to intersect the twentieth parallel towards the interior, it will be found that fifteen days’ journey south-westward (according to the distance and bearing of Aúkar from Aúdaghost) measured from that intersection, will reach to the neighbourhood of Tomboktú.

The foregoing account is evidently that of a route frequented by caravans, and therefore the distances mentioned in it may be received with confidence. Its general direction, as Abulfedá tells us, was to the south. El Bekrí describes also the journey from Darʿah to Ghánah, in a brief and uncircumstantial manner; but his description, though wanting in the minuteness and precision derivable from the experience of caravans, is yet not wholly uninstructive. From Wádi Darʿah to Wádi Tárḳa, on the margin of the desert, was a journey of five days. Then the traveller entered the wilderness in which water occurred only every two or three days. Of the wells first met with, one was called Tezámt. Eastward of it were Bír el Ḥammálín, or the porters’ well, and another named Nálellí or Málekí.[27] “From these wells,” says our author, “to the country of Islám, is a distance of four days; and at an equal distance are the Adarérén Wazzél, or mountains of iron.[28] There a desert begins in which there is no water for eight days; it is, indeed, that which is emphatically styled The Desert. The water in it belongs to the Benú Yentesír, a tribe of Zenágah. The village of Moddúken, which is next arrived at, belongs to the same race. Thence to Ghánah is a journey of four days.”[29]

Wádi Darʿah was three days from Támedelt, probably east by north. Wádi Tárḳa may be conjectured to have been on the southern side of the mountains of Azawwar mentioned in the account of the route to Aúdaghost. But the particular here chiefly deserving of attention, is The Desert, in which no water was met with for eight days. The name of this desert, omitted by El Bekrí, is supplied by subsequent writers. We are informed that the merchants going to Ghánah passed through the desert of Tíser, a dry and desolate wilderness of sand, with only a few pools of bad water, the chief of which was that called the well of Tíser.[30] The extent of this desert is variously stated to be eight, ten, twelve, and even fourteen days’ journey.[31] At its eastern extremity stood Aúdaghost, and hence it is not surprising that the journey to that city should present diminished difficulty; though the loose, hot sands, and intervals of four or five days without water, in the latter half of the route thither, between the 23rd and 32nd days, clearly mark the continuation of the inhospitable tract.[32] On the west, the desert of Tíser adjoined the division of the Ṣaḥrá called Kamnúdíyah, which bordered on the ocean in the vicinity of Cape Bojador, as shall be shown hereafter. Unless we assign, therefore, to the territorial divisions of the Ṣaḥrá a disproportionate extension from west to east, the contiguity of the maritime district of Kamnúdíyah to the desert of Tíser, strengthens the presumption that the latter could hardly have reached beyond the twelfth meridian from the shores of the Atlantic; or, in other words, that it did not extend far east of the meridian of Tomboktú, and, consequently, that the road to this city from the north passes over the tract of desert anciently named Tíser.

But it will naturally be asked, is there such a tract of peculiarly arid desert on the road to Tomboktú? Certainly there is; a desert of like extent, presenting the same physical character, and occupying a similar position in the route. Ibn Baṭúṭah, on his way from Sijilmésah to Tomboktú, arrived at Tegháza, near the edge of the desert, in twenty- five days. In ten days more, he came to the wells, or rather muddy pools of Táserahlá, where the caravan halted to prepare for the march over the formidable waste of sand, which it required ten days to traverse in order to reach Aïwalátin, a town on the southern border of the Ṣaḥrá.[33] Had that traveller crossed obliquely from Táserahlá to Tomboktú, he might perhaps have found the waterless desert to exceed a ten days’ journey in extent. But towards the east its width seems to diminish; whether it be that the northward course of the Great River, or the chain of mountains determining that circuit of the stream, and stretching across into the desert, modifies the atmosphere so as to give a greater range to the periodic rains in that meridian, we have no means of deciding. But the whole of the western Ṣaḥrá and the peculiar tract under consideration, are described by Leo Africanus in the following manner:—

“To begin with the desert of Zenaga; this is a dry and barren tract beginning from the ocean on the west, and extending eastward to the salt pits of Tegaza. On the north it is bounded by Numidia; that is to say, by Sus, Acca, and Darah; and it extends towards the south as far as the land of the Blacks; that is to say, to the kingdom of Gualata and Tombutto. There is no water found in it, except at intervals of a hundred miles, and this, after all, is salt and bitter, in wells of great depth, particularly on the road from Segelmesse to Tombutto. There are many wild animals and serpents in it, as shall be related in the proper place. In this waste is found a desert very difficult and dismal, called Azaoad, where neither water nor dwelling-place is met with for two hundred miles, from the well of Azaoad to the well of Araoan, which is a hundred and fifty miles from Tombutto, and in which great numbers of men and animals perish of heat and thirst.”[34]

The same writer elsewhere informs us, that the desert of Azawad was so called from its barrenness and dryness. It is not unreasonable to suppose that when the local designation of Tíser fell into disuse, the epithet expressing the general aspect of the region took its place. The name Azawad still remains to the tract of desert northward of Tomboktú.[35] And it cannot escape attention that the deserts of Tíser and Azawad resemble each other not only in extent and physical character, but that they are also equidistant from Sijilmésah; that they are both on the road southward from it, and both reaching to the southern limit of the Ṣaḥrá. There is still another point of resemblance between them, which, of itself, is almost sufficient to prove their common identity. El Bekrí remarks, that in travelling from Sijilmésah to Ghánah, a desert of two months’ journey was traversed, in which there were no fixed habitations, and the only people met with were wanderers, such as the Benú Masúfah, a branch of the Zenágah. The Benú Masúfah then, were in the middle of the eleventh century the tenants of the inhospitable plains over which lay the road to Ghánah; and it is remarkable that, three centuries later, the same miserable tribe hovered over the road to Tomboktú. Ibn Baṭúṭah found them to be the regular, and, as it were, hereditary guides across that desert, with the intricacies of which, tradition as well as personal experience had made them perfectly acquainted. In a region where the natural landmarks are so broad and unchangeable, and where man is comparatively so weak; where there is so little to tempt ambition or to nourish caprice, and where the whole tribe laying claim to a long extent of territory, could never exceed a handful of individuals,—the fact that the roads to Ghánah and Tomboktú, traced over peculiarly arid and forbidding tracts, both passed through the encampments of the same tribe, is a strong proof of the proximity of those roads: for the occupiers of the most inhospitable region in which life can with difficulty be supported, are not likely to be disturbed in their possessions.

It has been seen that Leo Africanus represents the country of the Zenágah, or Ṣinhájah, to have extended from the sea shore as far eastward as Tegháza and Tomboktú. He obviously meant to intimate that the road to Tomboktú formed the eastern boundary of that nation or division of the Berbers. But his expressions are not such as require to be strictly interpreted; on the contrary, they have a claim to that latitude of explanation which reconciles them with the state of things at present, when the Zenágah occupy the country round Tomboktú and to a short distance eastward of it. His expressions, however, militate most strongly against any hypothesis which would place Aúdaghost remote from Tomboktú, since the road to the former of those cities lay wholly within the limits of the Zenágah. If these people held as large a share of the Ṣaḥrá in the beginning of the sixteenth century as in the middle of the eleventh,—and, since they figured as conquerors in the only revolution generally affecting them in the meantime, this can hardly be disputed,—it must follow that the road to Aúdaghost, being within their limits, could not have passed far eastward of Tomboktú.

The Zenágah extended southward, according to Leo, “to the country of the Blacks, where are the kingdoms of Gualata and Tombutto.” To the words, “the country of the Blacks,” Marmol, while borrowing largely from the Arab writer, adds the gloss, “which is called Geneúa.”[36] And herein he accords with all other authorities respecting the position of Genéwah, which region, however vaguely defined, is yet always placed on the frontiers of Negroland, westward from Tomboktú. In early ages however, before the Berbers had derived strength from Arab instruction, the Blacks probably possessed a larger share of the desert, or at least the name Genéwah reached further northward. An Arab writer informs us, that Genéwah extended from the ocean in the west, to the country of Wergelán in the east, and from Amímah in the south to Arkí and Núl el Aḳṣa in the north.[37] The longitudinal dimensions of this country are here expressed in terms of obvious inaccuracy. Wergelán, though far to the north-east of the country called Genéwah, is yet made conterminous with it, owing to the great share which that Berber state took in the commerce of Negroland, at that time concentred in Ghánah. But the line on which the breadth of Genéwah is measured, drawn from north to south, from Arkí to Amímah, may be presumed to mark its central or principal section. Now Arkí was on the hills of the Lumtúnah, seven days distant from Wád Nún, and Amímah stood at a short distance westward or south- westward from the future site of Tomboktú.[38] The mean position of Genéwah being thus indicated, the author adds, “and of its cities, is Ghánah.”[39] Here, then, is another proof that Ghánah was contiguous to the western desert and to western Negroland, and that the tract of country in which it stood and flourished, was the same in which Tomboktú subsequently rose into importance.

The country assigned to Genéwah, in the above-cited passage, really belonged for the most part to the Zenágah, who, anterior to their conversion to Mohammedanism, lived much intermingled with the Blacks. The latter may have had the upper hand; or, though in a servile state, they may have been the more numerous class of the inhabitants; or finally, the slave trade being carried on universally in the desert, the Arabs of Barbary may have easily confounded the country exporting slaves with that which produced them; and thus applied the name Genéwah to the deserts of which Berber tribes were, if not the sole occupants, at least the masters. But the movements of the Morabites revealed more completely the partition and social condition of the Ṣaḥrá. The northern portion of it, towards the ocean, was possessed by the Lumtúnah, whose dwellings covered a range of hills (probably those called by El Bekrí Azawwar) said to be six days’ journey in length, and to be shaded by 20,000 palm trees.[40] On these hills, and about seven days from Wád Nún, stood the fort or town of Arkí, the chief place of the Lumtúnah, whose flocks wandered from the shores of the ocean as far eastward as the road to Ghánah.

South of the Lumtúnah, but separated from them by an uninhabited tract ten days’ journey in width, were the Benú Goddálah, the most powerful of the Berber tribes. In the uninhabited tract ten days wide, it is easy to recognize the sandy region south of Cape Bojador, and forming the continuation of the desert of Tíser or Azawad. The country of the Benú Goddálah is said to have extended a two months’ journey in length and breadth, a description which is applied also to the whole western Ṣaḥrá, and to the dominions of Ghánah; and considering that in each case the Atlantic Ocean is taken as a boundary, it is evident enough that no exact limits were set to territorial possessions in the desert, and that claims of sovereignty often extended from opposite quarters over the same ground.[41]

In the country of the Benú Goddálah was a mine or natural deposit of salt near a town or place of fixed habitation on the sea shore, called Aúlíl. At this place was a point of land, or peninsula, insulated by the tide, but accessible on foot at low water. Close to it was the port. Ambergris was collected on it at the sea side, and turtle, which constituted the chief food of the inhabitants, were there so large, that fishermen, as our author assevers, went to sea in their shells. From Aúlíl salt was carried inland to Ghánah and other cities of the Blacks. The road from the same place to Wád Nún was a two months’ journey in length, going along the sea shore, where fresh water was found by digging in the sand when the tide was out.[42]

There is little room for doubt or hesitation in determining the position of Aúlíl. One point only on the whole coast of the Ṣaḥrá can be selected for it with any show of reason: and that point is in the bay of Arguin, where the natural deposits of salt, the little island or peninsula, and the abundance of large tortoises, are all found together; and exactly at a distance of two months’ journey from Wád Nún, along the shore.[43] At Arguin also existed, in the fifteenth century, a trade derived from the natural productions of the place, exactly similar to that ascribed to Aúlíl in the eleventh century. Since the shores of the Great Desert offer one locality, and but one, answering to the description given by El Bekrí of Aúlíl, we must necessarily conclude that this town was situate in that locality,—namely, near Cape St. Anne in the bay of Arguin.[44]

Thus it appears that the Benú Goddálah were the possessors of the maritime region of the Ṣaḥrá, from Cape Blanco southwards. They also extended far inland (a two months’ journey, as has been already stated), and were separated by a distance of only six days from the Blacks on the Great River, and in the vicinity of Ghánah. They were the possessors, therefore, of the vast country which is now divided among the Ludayas, Brebísh, Trarzas, Erghebat, and others. Though esteemed the most powerful of the Berber tribes, they were yet compelled to yield the pre- eminence to the Lumtúnah, in the wars which immediately preceded the coalition of both under the name of Morabites. It may be fairly assumed, therefore, that these great tribes which divided between them the entire breadth of the Ṣaḥrá where it was least inhospitable, constituted the main body of the Berber nation to which they belonged; and therefore that the road to Aúdaghost, which left them on the west, lay near to the limit of the country occupied by the Zenágah, and consequently passed through the same tract as the road subsequently traced to Tomboktú.

Having thus examined the routes from Sijilmésah to Ghánah, and briefly surveyed the state of the western deserts, with a view to illustrate the geographical position of the latter country, we may now proceed to consider also its internal condition and character; to inquire who were its neighbours in Negroland; how these were placed in relation to each other; and how far the accounts given of them, taken collectively and in the plainest acceptation, accord with the knowledge which we at present possess of the interior of Africa.

Aúdaghost, once the residence, as we are told, of the king styled Ghánah, was situate in a hilly country, on the margin of the desert, but within the limit of the summer rains. Its water was chiefly drawn from wells; its irrigated gardens had small extent, yet to eyes accustomed only to the monotony of the sandy waste, the groups of palm trees around it formed a luxuriant scene. Its population, gathered from various tribes in the Belédu-l-Jeríd, belonged chiefly to the Berber nation of the Zenátah.[45] Nor is it difficult to explain why an isolated Zenátah population should fix and maintain itself in immediate contact with the roving and predatory Zenágah; for during the period referred to in these accounts, Sijilmésah, with which Aúdaghost was connected by ties of commerce, belonged to the Zenátah. The latter town was, in reality, but a trading colony on the frontiers of Negroland; and its mercantile inhabitants, content with the town and trade, seem to have never affected political independence.

In the year of the Hijra 350 (A.D. 961), the king of Aúdaghost was Tín Yerátán, son of Wasenbú, of the Zenágah nation. His empire is said to have had an extent of two months’ journey in length and breadth, and more than twenty negro kings paid tribute to him. But, at a later period, Aúdaghost became tributary to Ghánah. This submission of a Berber people to a nation of Blacks and unbelievers, served as a pretext to the Morabites, who, in 446, (A.D. 1054, the same year in which they made themselves masters of Sijilmésah,) destroyed Aúdaghost, carrying off the women and children into slavery.[46] Arab writers, of a later date than the fifth century of the Hijra, still speak of Aúdaghost and its Zenágah rulers. It is not unlikely that, regardless of events, they only re-echo the words of ancient historians; otherwise, we must suppose that place to have revived for a short time under the descendants of its Morabite conquerors. But the same revolution which yielded up the old and circuitous channel of commerce to the owners of the Western Desert, must have tended to let it fall into disuse. Under the new state of things, the Lumtúnah and Masúfah would naturally take the nearest road to Ghánah, through their own country and over the desert of Tíser; and thus Aúdaghost would be forgotten.[47]

The trade centering in Aúdaghost embraced not only the gold and slaves of Negroland, but also the productions of the Western Desert, and of the shores of the Atlantic Ocean. The skins of the antelope called Dant, or Lant, were wrought into bucklers by its artisans; who also manufactured ambergris, their supplies of which, we are told, they owed to their vicinity to the sea shore.[48] From this it may be concluded, not that Aúdaghost was near the sea, but that, in those early times, the possessors of the Western Ṣaḥrá were generally in too wild a state to allow trade to be carried on through their country; and that consequently the maritime productions of Aúlíl passed eastwards through the hands of the Benú Goddálah (who, occupying a favoured tract, had acquired more settled habits) to Aúdaghost, and thus reached Sijilmésah, after making the circuit of the domains of the rude Zenágah. El Bekrí says also, that on the hills round Aúdaghost grew trees yielding the gum which was used in Spain to dress silks.[49] Though this statement is not improbable, yet it is more likely that the Benú Goddálah, while conveying their salt and amber to the interior, likewise carried thither the produce of their rich gum forests. Wealth in Aúdaghost consisted chiefly in slaves, of which single individuals sometimes possessed a thousand each. That mercantile spirit had there fixed its abode in the midst of natural sterility, is forcibly expressed in the acknowledgment that slaves were the only luxury of the place. Aúdaghost exhibited the extreme licentiousness of manners characteristic, as will appear from other examples occurring in these pages, of all the towns in the southern border of the great desert, where the traveller, just escaped from the perils of the wilderness, indulges in the pleasures offered by a degraded population; and where the recklessness usual in a seaport is increased by the opportunities of the slave-mart.

After the destruction of Aúdaghost by the Morabites, Aúkár, fifteen days distant from it towards the south-west, became the capital of Ghánah; or rather that capital was composed of two towns, viz.—Aúkár, inhabited by Mohammedans, and containing no less than twelve mosques; and Ghábah, where the king resided, in the midst of a black population.[50] Dark woods environed the latter town, and spread a gloom well suited to the pagan rites for the performance of which they were reserved, and which often involved the sacrifice of human victims. An interval of six miles, covered with habitations, separated the two towns. The inhabitants drew their water from wells. The climate was deadly to all but the natives.

El Bekrí’s description of Ghánah evidently brings us within the limits of Negroland. It exhibits to us, on the one hand, the Blacks summoned before their king by beat of drum, sprinkling dust on their heads, and prostrating themselves in his presence; or performing the rites of their cruel superstitions in the darkness of their woods: and, on the other, the Arabs or Berbers dwelling apart in a more elevated and open situation, and yet suffering from the noxious humidity of the air. But, it is to be remarked, that he makes no mention of running waters, the importance of which no Arab author ever overlooks; and indeed, it must be inferred from that author’s words, that there was no river—certainly no great river—in the immediate vicinity of the capital of Ghánah.

The Benú Goddálah, who possessed the southern portion of the Ṣaḥrá from the shores of the ocean eastwards, carried their salt and other merchandise to the Ṣínghánah, the nearest black nation, from whom they were separated by a distance of only six days’ journey, and whose chief city stood on both banks of the Great River, called the Nile of the Blacks. When it is considered that the Lumtúnah reached within ten days of Ghánah; that they were separated from the Goddálah by a broad tract of uninhabitable sand; and that all accounts agree in representing Ghánah as the most western of the kingdoms of the Blacks,—or, in other words, that the desert alone intervened between it and the ocean; it will be apparent that the Ṣínghánah, who were nearest to the Goddálah, lay towards the south or south-west from Ghánah.[51] South-westward from the Ṣínghánah, at no great distance, was Tekrúr, and a little further on stood Silla, both likewise on the Great River or Nile of the Blacks. The last-named place was twenty days’ journey from Ghánah; and, from what has been already said, it will be manifest that its bearing from that capital was between south and south-west.[52]