Chapter VII: Part 7

Some of the systematic Arab geographers divided Africa into three great regions, viz. Genéwah, Kaúkaú or Karkar, and Habesh or Abyssinia; others into four, Genéwah, Nubia, Habesh, and Zinj.[223] Genéwah, or the western division, was disproportionately enlarged, owing to the protraction of the Great River, the incurvations of which were overlooked, and because, in speculative geography, the known has a constant tendency to encroach upon and narrow the limits of the unknown. Zinj, on the other hand, must have been diminished, since Ibn Baṭúṭah believed Sofálah to be but a month’s journey distant from Yúfi (Núfí), on the left bank of the Great River, before it turned towards Nubia. The centre of the continent, where those divisions met, was occupied by Lake Kúra, whence issued the Niles of Ghánah, Egypt, and Maḳdishó. The shores of the lake were inhabited by the Demdem or (in the Arabicised plural) Demádem, who therefore stood, as occupants of the remote interior, in a defined relation with the coasts to which those rivers descended. When Arab writers, therefore, in speaking of the eastern coast of Africa, state that the interior is possessed by the Demádem, who invaded Abyssinia and Nubia in the early part of the thirteenth century, it is obvious that they speak the language of system (the name Demdem or Demádem being in reality unknown on the eastern coast), and hypothetically trace the course of the invaders from the shores of Lake Kúra and the sources of the great rivers.[224]

There is no injustice done to the Arabs in thus ascribing altogether to theory a positive statement made by many of their best authors. It is in the highest degree improbable, that with little or no knowledge of the various Black nations inhabiting the eastern coast of Africa, they should have had any accurate acquaintance with the remote interior: and besides, the acquiescence in system here imputed to them, is no greater than must have inevitably arisen from the imperfect state of their knowledge. Little more than a century ago, European geographers represented Abyssinia as occupying nearly a fourth of the African continent; on its eastern borders they placed a great lake, from which issued the Egyptian Nile, and all the great rivers of Southern Africa.[225] The maps of Africa of that date exhibit less vacant space than they do at the present day. The improvement of geography, with respect to that quarter of the globe, has consisted chiefly in reducing what is known within its proper limits. Distant nations were of course as easily brought together and united as distant countries. The different African tribes which, in the course of the sixteenth century, devastated the widely-separate coasts of Sierra Leone, of Angola, and of Melinda, were, by a sweeping generalization, all supposed to be one and the same people, and were furthermore identified with the Agows and Gallas of Abyssinia.[226] Vestiges of these ideas still remain in our treatises of geography, and in some of the latest maps, nor is the system of thinking from which they emanated yet quite obsolete.[227] But the close resemblance of European theories respecting the mysterious interior of Africa to those of the Arabs, is strikingly manifest in the following words of the Portuguese historian, Da Couto:—“About the year 1570, a horde of barbarians, like locusts, issued from the heart of Ethiopia, from the great lake whence flows the Cuama, the Zaire, the Rhapta, and the Nile.”[228]—Here then we have the exact counterpart of Lake Kúra and the Demdem or Demádem. The subsequent history of the horde referred to by Da Couto is taken up by other learned writers, who affect to describe its march southwards from Mombása to the Cape of Good Hope; thence to Angola, whence it spread to Sierra Leone and elsewhere: so that not even the Demádem were ever carried by conjecture so far from their native homes.[229] Thus it appears that the theories ascribed above to the Arabs, much excelled in sobriety, while they were exactly parallel in design with the geographical speculations of a later age.

The position of the kingdoms of Negroland enumerated by Arab writers having been now discussed and determined, and the efforts of Arab theorists to mould into unity and form the isolated facts before them, having been traced out, our task is at an end. The demonstration of the fact that Ghánah lay between the desert and the Great River near Tomboktú, at once reconciles with nature and probability, the history of the constant intercourse of that state with Sijilmésah. The nation whose language is spoken in the most important part of Negroland, is now brought into light. The Tekrúr have been traced from the vicinity of Silla to the eastern bank of the Kowára. The History of Málí has been made known, and the limits of that empire partially determined. It has been clearly shown that Kághó was also called Kaúkaú, but that the application of the latter name to one or more other places, further east, has caused incurable confusion. The ignorance and erroneous hypothesis of the Africans respecting the course of the Kowára, have been detected in the ancient accounts of Lake Kúra; and the limits of the positive knowledge of the Arabs have been ascertained in the fact that their theoretical geography embraced that lake, and the Demdem who inhabited its shores.

It will not be necessary to dwell here on the general harmony and widely-extended coincidence attending the conclusions arrived at in the preceding pages. A long series of inferences, each stamped with the character of likelihood, and all agreeing perfectly among themselves, yet obtained independently of one another, not by straining arbitrarily selected texts, but by eliciting and examining each author’s fullest meaning, and which form together a complete whole, reconcileable not only with geographical facts, but also with that speculative mind, which in the history of human knowledge is itself an incontestible though not easily seized fact;—such a series of inferences, we say, carries with it an internal evidence of truth not easily impugned. It remains therefore only to recal attention to the chief historical revolutions brought to light in the course of our inquiries. The wars and conquests of the Morabites eventually opened the Western Desert to commercial enterprise. The impulse given by the religious enthusiasm of the same people to Tekrúr, spread rapidly through western Negroland, till at length the wave recoiling on the desert, the Ṣúṣú first, and then the people of Málí, became masters of Ghánah, and reckoned some of the Zenágah tribes among their tributaries.[230] The outlines of the history of Málí deserve particular attention. The establishment of extensive empires in the early stages of society, almost always give rise to a better order of things, by breaking down the obstructions to general intercourse, and allowing free scope to aspiring industry. The progress of Tekrúr eastwards, the foundation of Aghades, and the change effected by both these events in the condition of Houssa, have been already pointed out, and need not be further insisted on.[231]

It is impossible to deny the advancement of civilization in that zone of the African continent which has formed the field of our inquiry. Yet barbarism is there supported by natural circumstances with which it is vain to think of coping. It may be doubted whether, if mankind had inhabited the earth only in populous and adjoining communities, slavery would have ever existed. The Desert, if it be not absolutely the root of the evil, has, at least, been from the earliest times the great nursery of slave hunters.[232] The demoralization of the towns on the southern borders of the desert has been pointed out; and if the vast extent be considered of the region in which man has no riches but slaves, no enjoyment but slaves, no article of trade but slaves, and where the hearts of wandering thousands are closed against pity by the galling misery of life, it will be difficult to resist the conviction that the solid buttress on which slavery rests in Africa, is—The Desert.

[Footnote 195: At the present day the servile and perhaps most numerous class of the population of the southern shores of the Persian Gulf, are Zinj, or Blacks, originally from Zinjibar (corrupted into Zanguebar), or the eastern coast of Africa. In Zinjibar, that is, the country of the Zinj, on the other hand, the rulers and upper classes are chiefly Arabs from ʿOmán and the Persian Gulf. The tribe of the Lámí, who have given their name to Lámú, near Patta, are originally from the neighbourhood of Baṣrah. The event related in the text ceases to appear improbable when the nature and antiquity of the intercourse between Zinjibar and the Persian Gulf are considered.]

[Footnote 196: The Berbers here spoken of are the inhabitants of the country called by the Greeks in general Βαρβαρία, and by the Arabs Al-ʿajemí—that is, foreign; which latter name has been converted by ancient geographers into Azania (Ptolemy; and Arrian, Perip. Mar. Eryth.), and by moderns into Ajan. The name Berber, in this as in most other instances, originated in the commercial and diplomatic language of the Roman Empire. The east-African Berbers are now called Somáli; but their ancient designation still remains to Berberah, a town or rather encampment opposite to ʿAden. The tribe who possess Maḳdishó (the Magadoxa or Magadocia of our maps) are the Bajúna or Bagúna, called by the Sawáḥilí, or natives of the coast of Zinjibar, Wagúña. They are the Baẓúnah of El Idrísí (Jaubert’s Idrísí, I. p. 55, where مدونه is read for بذونه). It will be shown further on, that the Demádem have been transferred to the eastern side of Africa by an ordinary effort of speculative geography.]

[Footnote 197: The title of the Emperor of Abyssinia was Negusa Negast, or King of Kings. (Ludolf. Comment. p. 11.) The city called by the Arab writer Kʿaber was Ankó-ber (or the Pass of the Ankó, a tribe formerly occupying that tract, but now removed further north), at present the capital of Shoa. The Arabs and Abyssinians in ancient times were intimately connected. The language of Tigré, or Northern Abyssinia, is of Arabic origin, and even the Amharic is thought by Gesenius (Ersch and Gruber’s Encyklopedie, art. Amharische sprache) to be an older offset of the same stock.]

[Footnote 198: The Bojá or Bogá are the Βουγαείται of the Greek inscription of Axum, copied by Salt (Trav. p. 410). Under that general name was included all the tribes of the desert between Abyssinia and Egypt; the Blemyes of ancient geographers, and the Bisharee or Bishareen of modern travellers.]

[Footnote 199: By the expression that the Nubians are brethren of the Zinj, it must be understood that they are of negro origin; and indeed there is little reason to doubt that the Nubians on the Nile were originally a servile population, the progeny of the Nubah of Kordofán, who, in the course of events, became sole possessors of their master’s domains. Their emancipation was forwarded by powerful external causes (see Edinb. Rev. No. 125. p. 297), and does not appear to have been accomplished by means of revolt or invasion, as was the case with the Zinj in Baṣrah, and the Funj in Sennár (Bruce’s Trav. vol. VI. p. 370). When the Arabs conquered Nubia, they exacted an annual tribute of slaves, which was called _Bakt_ (Quatremère, Mémoires sur la Nubie, II. p. 42), a word evidently derived from the ancient Egyptian language, in which _Bok_ signified a slave.]

[Footnote 200: The family of Abú Ḥafṣ, of Berber origin, rose to the sovereign power in Tunis, in the early part of the thirteenth century. (Makrízí in Hamaker, Spec. Cat. p. 105.)]

[Footnote 201: The names of nations from Kaúkaú westwards, enumerated by Ibn Sʿaíd, are written as follows by Ibn Khaldún (MS. B.M. fol. 90) and Makrízí (Hamaker, Spec. Cat. &c. p. 107, whose orthography is here retained) respectively: ?aghárah ٮغارة I.K.; Baḳárah بقارة M.—At-Tekrúr التكرور I.K. & M.—Kimí كمي I.K.; Nama نمي M.—?emyem ٮميم I.K.; Temím تميم M.—Ḥáyí (?) حايى I.K.; Já جا M.—Kúra كورى I.K.; omitted by Makrízí.—Inkizár انكزار I.K. & M.]

[Footnote 202: The comparison of texts made in the preceding note proves that Makrízí borrowed from Ibn Sʿaíd, but it is not easy to define the extent of his obligations to that writer. The Tunisian dynasty of Abú Hafṣ, the wars of the Zagháwah with the Wathekú (the opponents of that dynasty), and the invasion of Mábiná by the King of Kánem in 1252, referred to by Makrízí, all belong to the age of Ibn Sʿaíd, from whom he probably obtained his knowledge of them. But, on the other hand, Makrízí names the King of Kánem reigning in A.D. 1398, a century later than Ibn Sʿaíd. It appears more probable that his list of the Black nations near Kánem was the fruit of his own inquiry, than a transcript from an earlier writer.]

[Footnote 203: As names changed from Arabic to European writing are apt to acquire thereby a more determinate form than properly belongs to them, those mentioned in the text shall be here represented in their original character, that the reader may be enabled to appreciate our conjectures respecting them. Iklí اكلي; Afnú افنوا; Mambó منبو. Caancouma (in Hamaker) كانكوما is evidently Kátakúmá كاتكوما wanting a point. In like manner Hamaker’s Caancou must be changed into Kátakú. It may be thought that there is not sufficient authority to prove that Kátakúmá and Kátakú are distinct countries. But Burckhardt (Trav. in Nubia, p. 433) has stated the position of the latter, and the districts comprised in it (nearly all pointed out by Denham) with so much precision, that his testimony, corroborated by that of Mohammed Miṣrí (Journ. of the Roy. Inst.), decisively separates Kátakú from the Katagum (Kátakúmá), which was visited by Clapperton.—Ibkarem ابقرم—Rábúmá رابوما—Haúdama هودمي—Ankarar انكرر is probably written by an error of the pen for Ankarú انكرو.]

[Footnote 204: Shádí شادي; Mábiná مابنا; Abham ابهم; Atʿaná اتعنا; Yáfalam يافلم; Mekba مكبا; Kálkín كالكين.]

[Footnote 205: Aderma ادرما; Dafúmú دفومو; Abkalá ابكلا we have ventured to change into Ankalá انكلا; Túkámá توكاما.]

[Footnote 206: Afnú is the name given by the people of Bornú to Houssa (Lucas in Proc. Afr. Assoc. I. p. 165), or the eastern part of it. Einsiedel (Cuhn’s Merkw. Reisen. III. p. 439) understood that Hafnou (Afnú) lies between Bornú and Zegzeg. Abdu-r-Raḥmán Aga, Niebuhr’s informant (Walck. Rech. p. 72) also uses the name Afnú as equivalent to Houssa. The Sultan of Tekrúr, he says, who possessed Mara (Marra), was tributary to the Sultan of Afnú, residing in Zamfara. Seetzen also (Von Zach’s Monatliche Correspondenz, vol. XXI. 1810, p. 152), places Affano immediately to the west of Bornú. See also the Bulletin de la Soc. de Geogr. de Paris, tom. VI. p. 169, where Kachenah is stated to be the capital of Afnú. It is remarkable that in Bornú, and the adjoining deserts, the Arabic expression Súdán (country of the Blacks) is always given to Afnú or Houssa (Lucas, as above; Denham’s Discoveries, &c. II. p. 85), a strong proof that it was the country of the Remrem or Demdem, and the point to which the slave merchants directed their march.]

[Footnote 207: Kátakú comprises Mandara, Musgow, and the other provinces on the west of the River Shary, which are therefore not named. It is not to be ascribed to mere chance, that two names are changed, by the addition of a single point to each, into Kátakúmá (the Katagum of Clapperton) and Kátakú, the Katákó of Burckhardt, the Kotoko of the native of Bornú cited in the preceding note (Bullet. Soc. Geogr.), and the Kotko of Seetzen (p. 153).]

[Footnote 208: Umburm is in the country of the Yemyem near Jacoba (Clapp. Sec. Exped. p. 250). In Sultan Bello’s account of Baúshí (Denham and Clapp. Disc. II. p. 451), he mentions a province of that country called Aádám. We cannot venture to say whether this is the root from which Adamawa is derived, but it might be easily changed in discourse into Haúdama. Angarú (Ankarú) is three long days’ journey west of the capital of Bornú (Mohammed Míṣrí, in Jour. Roy. Inst.), and within the dominions of Bello (Clapp. in Denh. Disc. II. p. 313). It is the Ungura of Hornemann, which was supposed to be identical with Wanghárah (Proc. of Afr. Assoc. II. p. 200).]

[Footnote 209: One of the natives of Houssa, interrogated by M. Menézes de Drummond (Hertha, July, 1827, p. 12), mentioned the Schadŭh (Shádí) among the tribes depending on Zegzeg. Can the name Mábiná be the same word as Foobina, said by Mohammed Másíni (Clapp. Sec. Exped. p. 335) to be sometimes affixed to Adamawa? The name Bobyra, given in the Quarterly Review (No. 77, p. 178), on the authority of Clapperton, in whose published Journal it nowhere occurs, might easily have its origin in Fobina, or even Mabina ill-written in Arabic. According to Abdu-r-Raḥmán Aga, the King of Tekrúr possessed Marra and Adana. One of the Itineraries collected by Dupuis (Resid. in Ashantee, App. p. 129,) places an Etana on the river west of Marra.]

[Footnote 210: Clapperton found the plains of Zamfara covered with a chain of lakes which are connected in the rainy season; and Lyon (Trav. in N. Afr. p. 151) was told that the country between Kanó and Zegzeg is annually covered with water.]

[Footnote 211: The name Angalawha, occurring on the northern shores of Lake Chad, is easily traced by an analogy of the Bornowí language indicated by Denham (the tree Kuka being called also Kukawha,) to Angala, which name also occurs on the southern shores of the lake. The town or station of Togáma is seven days distant from Kachenah, on the road to Aghades (Lyon’s Trav. p. 131). Hornemann (Proc. Afr. Assoc. II. p. 300) gives some account of the tribe so called.]

[Footnote 212: If we suppose the word Kissour pronounced N’Kissúr with the nasal sound, which among the Africans so often precedes the letter _k_, then its affinity with Inkizár becomes more apparent. The Portuguese, like the Arabs, employ an initial vowel in prefixing the nasal; thus for N’Yáka, N’Yambána, N’Góla, they write Inhaqua, Inhambana, Angola. Of the guttural pronunciation which seems to foreign ears to confound the _a_ and _u_, many examples might be given. Leo Africanus says that one language (which he calls Sungai) extended from Málí to Kághó; and as we know that the Kissour, commencing at Jenni, now extends at least to Tomboktú, we are justified in concluding that it is the Sungai, or the language to which the Zagháï, the chief inhabitants of Inkizár, lent their name. We have seen that the word Daḳno, the name of the ordinary beverage of the people from Jenni downwards, was in use below Tomboktú in the fourteenth century. (See above, p. 84.) It avails little against this, that the Sungai language was also spoken in Málí: for what is more natural than that the language of the most populous and industrious part of an empire should be generally spoken in its capital; and that a Moorish merchant should give little attention to the language of the lower classes?]

[Footnote 213: Two intelligent natives of Kanó, who were in London a few years ago, when interrogated respecting Wanghárah, agreed in stating that it is “behind Ako,” or Yariba. In the same vague manner probably, Leo Africanus, little acquainted with the interior, conceived it to be behind Zamfara. But his description of Wanghárah (pt. VII. c. 14), the nature of the journey to it, its trade, and its fear of Tomboktú, leave no doubt as to the country intended by him. The meditated invasion of Wanghárah by the King of Bornú, may indeed provoke scepticism; but let it be considered that the historical traditions related to Clapperton (Second Exped. p. 102, 103) by the King of Boussa (Busá), testify the former conquests of Bornú on the western side of the Kowára. Leo had a very inadequate idea of the extent of Negroland south of the Great River. He even speaks of the ocean encircling the desert from Cape Nún to Gaoga (pt. I. c. 2). He could not, consistently with such views, place the distant and populous country of Wanghárah south-westwards from Zamfara.]

[Footnote 214: Ibn Sʿaíd died A.D. 1286, at an advanced age. Yaḳút, of whose Geographical Dictionary the Bodleian Library possesses a copy, flourished somewhat earlier. Both these writers are quoted by Abú-l- fedá, who died A.D. 1331. Yaḳút and Abú-l-fedá cite Ibn Fáṭimah with no other variance than is usual in different MS. copies of the same work. The Jábí of Abú-l-fedá is clearly preferable to Yáḳút’s Ḥání (see Note 201); but the Bedí of the former and the Yuthí of the latter are probably equally erroneous. It may seem a bold emendation to alter them into Yúfí or Núfí; but let it be considered that the country now called Núfí or Níffí may have changed its name with its population; that Ibn Baṭúṭah clearly means Núfí when he speaks of Yúfí; and that the name written Yúfí in the Gayangos MS., is in other MSS. written Yuwí (Lee and Kosegarten, Lee’s Ibn Baṭúṭah, p. 238), and in others Buwí (Burckhardt, Trav. in Nubia, p. 491; and Lee). Now Bedí بدي and Yuthí يُذي lie, with respect to Buwí بُوي, Yúwí يُوي and Yúfí يُوفي, within what may be called reasonable limits of corruption, and the proposed change brings all into order.]

[Footnote 215: Abú-l-fedá and Yakut wrote Kúra كُورَى; in one of the Routes (No. 4) published by Dupuis the river is called Koara كُوَرَا, though had the points been correctly written, we should probably have had Kúrá. Bello writes in his map Kowára كوارَ, or, as our travellers have called it, Quorra. In Brahima’s Itinerary (Bowdich, Mission, &c. p. 491), and in another translated by De Sacy, (Walck. Rech. p. 453), the Great River is named Lake Koad or Caudh كوض, which ought rather to be read Kúḍa. Further on we shall show that in these Itineraries the Arabic letter Dád ض is substituted for _r_; so that Lake Kúra is here intended. The Kowára, Kúra, or Quorra is frequently styled by the natives a sea or lake, according to some accounts, of forty-eight days’ sail in extent (Ali Bey Badia’s Travels, I. p. 338). Clapperton (Denham, Disc. II. p. 269) was told that the river Kowára falls into the sea (of Nyffi or Núfí) at Raka, where it is as wide as from Kano to Katagum, or about 150 miles. But not to multiply authorities, it will be sufficient to observe that Sultan Bello believed Raka (Rághá),—which has been recently reached by Mr. Jamieson’s steamer Ethiope,—to be a sea-port, and represented it as such in his letter to the King of England.]

[Footnote 216: According to Ibn Fáṭimah, “when any one among these people dies, they cast the dead body to their neighbours, and their neighbours do the like for them.” So Sultan Bello related (Clapp. Sec. Exped. p. 251) that in Umburm, where those who ail are killed at once, for economy, “the person falling sick is requested by some other family, and repaid when they have a sick relation.”]

[Footnote 217: Kimí might without much violence be changed into Límí, and thus explain the name Al-Límiyín. At-Tekrúr we may assign, on the authority of Abdu-r-Raḥmán Aga, to Marra, which probably extended from Zamfara westwards between Guber and Kábí. The Baghárah or Baḳárah were probably a tribe of the desert. The Kaúkaú of Ibn Sʿaíd is too far east to be the city of that name on the Great River; we must suppose him therefore to extend this name to Karkar.]

[Footnote 218: Leo’s Gago seems to be identical with the Caugha of Hornemann, or Kaúka of Burckhardt (Trav. in Nubia, p. 436). Its empire extended, he says, from Bornú to Nubia. The use of the name Bornú, in former times, however, and the modern geography of the countries round Lake Chad, are involved in an obscurity which it does not lie within the scope of this essay to dispel. Leo’s text offers not only the names Gago and Gaoga, but also, in two instances, Gaogao. He says (pt. VII. c. 14), that while he was in Negroland, the King of Bornú marched against Wanghárah; but learning, on his way, that Omar, King of Gaogao, meditated an attack on his dominions, he turned back, and Wanghárah was saved. But, in this passage, Gaogao is a misprint for Gaoga, as Leo himself discloses by naming the King of Gaoga “Omar chi oggidí regna.” In the other instance (pt. VII. c. 1), his Gaogao is justly changed by Marmol (vol. III. fol. 21) into Gaoga. Leo (pt. I. c. 7) having enumerated the fifteen kingdoms of Negroland visited by him, adds, that there are three times as many, sufficiently well known, lying to the south of the preceding; and names five of them, viz. Bito, Temiam, Dauma, Medra, Goran. Marmol (vol. I. fol. 15), in copying this passage, omits Dauma, and substitutes for it Mandinga. But Leo had no idea of increasing the kingdoms of Western Negroland. His Bito is the Bede of Denham and Clapperton, adjoining or comprised in the modern Bornú. Einsiedel names together, Schikou—the Schaïkou of Lyon (Trav. in N. Afr. p. 126), two days from the capital of Bornú—Bitou (Leo’s Bito), and Engar (Angarú). Temiam may be an error for Yemyem: Dauma is probably the Doma of our maps, or the country on the right bank of the River Chadda. Medra seems to be Mandara, one letter being obliterated in the Arabic MS. Goran (in Marmol Gorhan), which is often referred to by Leo, is evidently the Desert of Kordofán. This name كردفان might easily become, in negligent writing, Korhán كرهان; or as Leo, uniformly writing _kef_ with a _g_, and omitting the aspirates, would represent it, Goran. Another region often named by Leo, may be fitly considered here. In the Desert of Seu, south of Bornú (pt. VII. c. 15), and environing an immense lake (pt. I. c. 27), called the Lake of the Desert of Gaoga (pt. I. c. 2), he places the sources of the Niger (pt. I. c. 3). It is obvious that the lake alluded to is Lake Chad, and that the name Seu is the root of the appellative Showy, and the name Shouaa, respectively given by Denham to a town on the Shary, and the Arab tribes inhabiting the adjacent country.]

[Footnote 219: Yaḳút and Abú-l-fedá both mention the towns of Maghzá and Jájah in the vicinity of Lake Kúra. Al-Maghzá, according to the former writer, was the port in which were fitted out the fleets of the King of Tekrúr, “who wages perpetual war with the infidels to the south of his states.” Al-Maghzá signifies the place whence invaders sally forth. Abú- l-fedá however differs from Yaḳút in assigning both Maghzá and Jájah, not to Tekrúr, but to Kánem. Jájah (perhaps the Gagai of Clapperton, Sec. Exped. p. 174) was the capital of a petty state situate probably between those two kingdoms. It was remarkable for its fertility and variety of its productions; among other things for its spotted sheep (described by Lander in Clapp. Sec. Exped. p. 259-60). But it must be observed that the Arab geographers, in describing the bearings of those places, particularly in reference to the lake, speak in general the language of misconception.]

[Footnote 220: Bowdich remarks (Mission to Ashantee, p. 478 note), that the name of the river written Kúra, Kúda, &c. was always pronounced Quolla by the natives in their conversations with Mr. Hutchinson. Ignaz Pallme (in the Athenæum, 1840, p. 54), a traveller in Kordofán, relates that the natives of that country think that the Bahr el-Abiad may be followed westward through Baghermi, Kúḳo (Kouka), and Niero (Naroo, the hilly country north-east of Zegzeg); and “further on (he says), in Kúla (Kúra) flows a river not identified” (that is, different from the Nile). Browne also (Travels in Africa, p. 254) heard in Darfúr of Darkulla (the Land of Kúla or Kúra), where pepper was in abundance, and the rivers were navigated in large canoes. He indeed supposed Darkulla to lie towards the south. But his map exposes his mistake; for the rivers Bahr Wullad Ráshid, B. el Salamat, and B. Heimad, crossed on the route to Darkulla, and which he places to the south and west of Baghermi, bear the names of Arab tribes dwelling in Wadaí and on the shores of Lake Chad (Burckhardt, Trav. in Nubia, pp. 433, 436). The route therefore went westwards.]

[Footnote 221: The Kiliwah (the Kalawa of Capt. Lyon, Kolluvi of Hornemann), a Tawárik tribe, are masters of Asben, or the territory between Houssa and Aghades. Their town in Guber is called by Clapperton Killiwawa or Calawawa, by the Tatar merchant Wargee, Galibaba. Clapperton frequently mentions also the Kilghí (whom he calls Killgris), another powerful tribe of the same nation. Their territory is the kingdom called by Bowdich (p. 208), Kallaghee, fourteen days’ journey from Gamhadi (Kambari), or from the Quolla, crossing the Gambarou (Kamba-róa, or Kamba water) on the tenth. Kamba is apparently the name given by the indigenous population to Kábi, or a part of Yaúrí (Dupuis, Append. 85). The Gambarou of Bowdich is the Gulbi Kambáji, or river of Kambáji or Konbash of Dupuis’ Itineraries (App. 126 and 192). The name Kilghí is changed by the Blacks into Kilinghi (see Note 162), whence comes the title Kilinghiwa given to the King of Kachenah (Walck. Rech. p. 451).]

[Footnote 222: Hají Hamed (Quart. Rev. 1820, No. 45, p. 232), among others, bears witness to the course of the Great River from the Sea of Nyffé to Egypt by Kachenah and Kano. Capt. Lyon’s informant, however (Lyon’s Trav. p. 142), traces the stream from Funda to Katagum, while Ben Yusuf, Hornemann’s son (Denham, I. p. 334), and Mohammed Miṣri (Jour. Roy. Inst. 1823, p. 5) are equally positive in making it flow through Adamawa. Much has been said of the unanimity of the natives in connecting the waters of Lake Chad with the River Chadda, but they agree only in the vague outlines of a theory, not in facts; they are unanimous in making the Kowára flow into the Chad, and not the Chad into the Kowára.]

[Footnote 223: The author of the Kitábu-l-Jʿarafiah divides Africa into three parts, one of which is Karkar: Shehabeddin (Not. et Extr. tom. II. p. 156), adopting the same division, writes Kaúkaú. The division into four parts is frequently referred to by Marmol (tom. I. pp. 18, 21, 31), who follows probably Ibn Gezzar.]

[Footnote 224: El Bekrí probably wrote Remrem; though El Idrísí, copying him, writes Demdem; the latter author names also the Lemlem. Ibn Sʿaíd may be conjectured to have written Yemyem, but the doubtful text of the MSS. leaves the point undecided. Abú-l-fedá mentions not only the Demdem, but also the Nemnem, which latter people he places south of Saharte (the most eastern district of Tígré in Abyssinia) and of Samhar (the Dankali coast), and consequently in what is now called the Taltal country. All those names, Nemnem excepted, refer to the same people.]

[Footnote 225: In the maps of Forlani and others of the sixteenth century, the Nile, Zaire, Cuama, and Spirito Santo, were all made to flow from Lake Zambere. Sanson however (1650) allowed that lake no outlet towards the east, but Hollar (London, 1667) still joined the river of Kílwah with Lake Zaflan, which, as well as Lake Zambere, was connected with the Nile. In all these maps Abyssinia extended to lat. 18° S. Delisle was the first who reformed these absurdities.]

[Footnote 226: Labat (Rel. Hist. de l’Ethiopie Occidentale, II. p. 90), copying, but not faithfully, Cavazzi de Montecucoli, states, with surprising coolness, as matter of history, the supposed origin of the Jagas in the country of the Monoemugi (Monomoézi). The country of the Jagas, that is to say, of the chieftains so entitled, lies immediately behind Angola, perhaps not above 250 miles from the sea coast, and there is nothing in the history of their followers calculated to show that they come from the remote interior. Andrew Battel, who was seized on the coast and carried off by the Jagas, with whom he spent above a year, says (Purchas’ Pilgrims, II. p. 973), that they told him they come from Sierra Leone. This absurd statement shows that Battel had got into his head some of the geographical speculations of his day. The Jesuit Sandoval (Hist. de Ethiop. p. 48) thus abridges the information of the missionaries: “About ninety years back, a nation called in their own country Gangedes, in Congo, Jagas, in Angola, Guindes, in India (Eastern Africa), Zimbas, in Ethiopia (Abyssinia), Gallas, and in Sierra Leone, Zumbas (Cumbas, in Jarric, probably for Çumbas), which name they changed for Manes, and who lived on human flesh, issued forth,” &c. Finally, Anguiano (Epitome Historial &c. del Imp. Abyss. 1706, p. 8), speaking of the Agows, assures us that the names Agáo, Agag, and Giagos, or Giacos, are all the same.]

[Footnote 227: A writer in a popular journal, says of the Zoolus (properly Amazúla,) near Natal,—“They extend much further northward, where they are found under the names Sualies and Gallas.” (Quart. Rev. Febr. 1837, p. 178.) The Arabic word Sowáḥilí means “inhabiting the coasts.”]

[Footnote 228: The portion of Da Couto’s History here quoted (Decade X. lib. 6, c. 15) has never been printed, but the Library of the British Museum possesses two MS. copies of it. The lake here referred to is called by De Barros Zambere, a name copied servilely by all succeeding writers, though it was doubtless a misprint for Zambeze. Cuama is the name given to the lower portion of the Zambeze, which river is so named according to Dos Santos (Ethiop. Orient. p. 44), because, on quitting the Great Lake, from which proceed the chief rivers of Southern Africa, it flows through a territory inhabited by a people of that name.” The people alluded to are the Ambios of Da Couto, the Movíza of the Portuguese of the present day, but who call themselves M’Bíza. The true name of the river, therefore (and that intended also to be given to the lake), is Zambíza. N’yassi, or _the sea_, as this lake is called by the natives (whence D’Anville’s Massi, by mistake for Niassi), is commonly but erroneously designated in our maps Lake Marávi. The Marávi country, that is to say, the country in which the chieftains bear the title of Marávi, extends from the Zambíza to the Livúma behind Cape Delgado, and touches but does not encompass the lake. Da Couto, following De Barros, borrows the name Rhapta from Ptolemy. It serves to indicate eruditely rather than clearly what the Arabs call the River of Maḳdishó, that is, the Juba.]

[Footnote 229: Cavazzi de Montecucoli, a laborious and sincere writer, relates (Istorica Descrittione de tre Regni, &c. 1690, book II. c. 3) that a chief named Zimbo raised an army in Congo, with which he invaded Melinda on the opposite coast. Being there defeated, he retired towards the Cape of Good Hope, and afterwards attacked Angola, &c. Zimbo’s marches equalled those of Tamerlane. The enormous exaggerations and mistakes of the Catholic Missionaries respecting the interior of Southern Africa, still retain their places in works of geography.]

[Footnote 230: A Mandingo warrior named Abba Manca (Mança?), conquered Bambúk in the beginning of the twelfth century, and compelled its inhabitants to adopt the Mohammedan rites (Golberry, Fragmens d’un Voy. I. p. 419). Silla was one of the first converted of the negro towns; and as, in the Mandingo language, the word Silla means a way, road, pass, or ferry, and might therefore have been naturally employed to designate a town situate on the line of traffic, it may be conjectured that Silla belonged to the Mandingoes from the beginning. It may be here observed that the termination _boo_ (see Note 52), characterizing the names of villages in Bambara, signifies a hut. (Dard. Dict. Wolofe, pp. 19, 22; Caillié, III. p. 301).]

[Footnote 231: The Sultan of the Fellátah dynasty in Houssa styles himself Sultan of Tekrúr. The Fellátah conquerors of Houssa issued, in the beginning of the present century, from the province of Ader, adjoining Kábí on the north, and where they may have been long established. It is possible then that the kingdom of Tekrúr, mentioned to Niebuhr by Abdu-r-Raḥmán Aga, and also heard of by Hutchinson (Bowdich, p. 483), may have belonged to the Fellátah dynasty now dominant in Houssa.]

[Footnote 232: The Garamantes, a Libyan nation, chased, in chariots drawn by four horses, the Ethiopian Troglodytes (Herodotus, book IV. c. 152). El Idrísí (in Jaubert’s Trans. I. p. 110,) remarks that stealing children to sell them to strangers for a trifle, is a general practice in the desert, and “no one there sees harm in it.”]

POSTSCRIPT.

* * * * *

REMARKS ON HOUSSA.

The route from the capital of Ashantí, through Gonja, to the Kowára or Quorra opposite to Yaúry, is determined as satisfactorily as can be expected from native information. Three different itineraries of that route agree in the chief points, conducting from Lake Buro, or from Salagha (which is the same as Dagwumba), through Tonuma, Jabzogho, Ghofil, Zogho, Jambodú, Suso and Kúka to Nikki, the capital of Borghú. Beyond this place is the town of Rugha, the River Wori, and the mountains over which lies the road to Sholo on the banks of lake Kúra (the Quorra). But in two of the itineraries the names Rugha, Wori, and Kúra are written Ḍugha, Woḍi, and Kúḍa.[233]

From Sholo the route conducts by Kambashi, Ghúnti, and Yendukka to Kachenah in twenty-five days. Other native routes, passing northwards of this through the mountainous country of Fagh, cross the river at Gongo (_i.e._ the ferry,) to Múrí (Múli), and then proceed through Kábí, over the Gulbi Kambaji to Raka, and thence through Kotú-n-kúra to Marki, and crossing the hills near Surami to Kachenah. The Gulbi Kambaji (river of Kambaji) of Dupuis, it has been already observed, is the same as the Gambarou (Kamba-roa, or Kamba water) of Bowdich. The latter writer places to the north of it, Mallowa, Kallaghee (Kilghí), Barrabadi (Beráberi), and Kachenah. On the south side of his Gambarou, or between it and the Quorra, he sets Gauw, Gamhadi (Kombori), Fillani, Goubirri, Zamfarra, Yaoura, and Noufí. If a partial error in the position of Mallowa be allowed for, and a more decided one in the case of Gauw, it will be evident that the river dividing the countries here enumerated is the Quorrama, or river of Kábí.[234]

Mohammed Masíní, describing the Kowára, says, “this great river issues from the Mountain of the Moon; and what we know of it is, that it comes from Sookan (Sókaí) to Kiya (Kiyaú, the Gauw of Bowdich), to Kabi, to Yaouri, to Boossa, to Wawa, and to Noofee; but in that place there is another river that springs from Zirmá, to Ghoober, to Zeffra, to Kory or Koora, and then enters Noofee; its name is Kaduna. On the north of it Kanbari lies; on the east is Kory; on the south are Cankan and Kafath; and on the west is Bassoa or Bashwa (Busawa, the territory of Busá). About the centre of it is the kingdom of Noofee, with that of Abyou (Abbiwa).”[235]

The river here described under the name of Kaduna, as running first northwards from Zirmá to Guber, and then southwards to Núfí, is apparently the same described by Clapperton in these words: “This stream rises only a day’s journey in the mountains or hills south of Guari, runs through part of Zamfrá, and divides in one part the states of Katongkora and Guari, and enters into the Kodonia in Nyffé.” The Kaduna or Kadunia enters the Kowára in the vicinity of Raka and Rabba, perhaps a little above the latter place. According to the native accounts, it is during part of the year a great river, navigated in canoes made of a single trunk of a tree, yet large enough to carry nine horses, but in the summer it is quite dry.[236]

The country called by Mohammed Masíní, Kory or Koora (Kúra), is either the Guari of the maps, or else Kotú-n-kúra. The resemblance of this name to that of the great river (Lake Kúra) deserves attention. It appears probable that the natives of Houssa name the Great River from what they conceive to be its sources in their own country, tracing it from Kowára (the Guari of Clapperton), through Zamfara and Kábí, down to the sea of Kúra or of Núfí.[237] Hence it is not surprising, that, while Ibn Sʿaíd mentions Kúra among the kingdoms of Negroland, Yaḳút and Abú-l-fedá should apply that name to the great water which there received the Nile of Ghánah.

LONDON:
J. HOLMES, TOOK’S COURT, CHANCERY LANE.

[Footnote 233: One of these routes is given by Dupuis (App. 124), another by Bowdich (p. 491), and a third, translated from the original Arabic by De Sacy, is inserted in Walckenaer’s Recherches (p. 453). The comparison of this last route with the fragment (No. 11) in Dupuis’ Appendix (p. 135), shows that its author was Mohammed al Marrawi, the servant of Ibrahim, from whom Bowdich derived his information. The Sholo of the Itineraries is the Sooloo of Lander (Exped. to the Niger, II. p. 28), who means by the expression “the left bank of the river,” the bank on his left hand, and contra-distinguishes it from the east bank, on which his horses were. The systematic substitution of _d_ for _r_ by the natives of a part of Houssa, is manifest from these itineraries. The words Barrabadi and Gamhadi, for Beráberi and Kombori, are other instances of the same change; and probably the savage and naked people called Maradi, said to inhabit the country between Kachenah and Guber (Lyon, p. 140), are no other than the Marrí, or people of Marra. The deposed King of Houssa lives in Maradi (Lander, Sec. Exped. p. 63, 153). Dupuis describes a route from Salagha, a little eastward of north, to what he calls the great city of Andari, respecting which there is room for some curious conjectures. But it will be sufficient for the present to remark that his Andari عنظر (Itin. No. 6) and Fadaly فضلى (No. 10), on the importance of both of which places he dwells with complacency, are in reality one and the same.]

[Footnote 234: For the Gharanti and Yendoto of Dupuis, Ghúnti (the Gonde of Clapperton) and Yendukka (the Yendukwa of that traveller, and Yendakka of Lyon,) are here read, the change in Arabic writing being extremely easy. The name Kandashy, which Mr. Dupuis gives to a part of Houssa, originated in a mistake. He joined the word Dúshi, signifying the hills or mountains, to a part of some preceding word. Marina, Kandashy, ought to be Markí, and Dúshi, or the hills. Mr. Dupuis continues, “Great Souy is the name of the adjoining country; there is also another Souy, where the water is very broad, and bears the name of Boromi Mághami.” For Souy سُوْوِ in this place we must read Surmi سُرْمِ, a name variously altered into Solan, Zulami, Zurmie and Zirmie. The situation of Surmi, or rather Surami the less, is pointed out by Clapperton (Sec. Exped. p. 164). Lander crossed the river of Makamie (Mághami). The word Boromi apparently signifies river, in some dialect the use of which extends into Bornú and Kánem.]

[Footnote 235: The Cankan of Mohammed Masíní is probably the Ghana- ghanah of Wargee. (Asiat. Journ. 1823, vol. XVI. p. 23.)]

[Footnote 236: There is, no doubt, some difficulty in believing that the mouth of the Kadunia is above Rabba, when Lander states so clearly that it is below Fofo; yet a MS. sketch of the route from Kano to the Kowára in Núfí, drawn by Clapperton from native information, places it between Rabba and Leechee, and directly opposite to Raka. Bakani, the capital, is one day from the mouth of the river; it seems to lie directly in the road from Kanó and Zegzeg to Raka and Katunga (Hertha, July 1827, Geogr. Zeit. pp. 11, 14). From one of Clapperton’s MS. vocabularies, communicated to me by my invaluable friend, the Rev. G. C. Renouard, it would appear that the word Kaduna, in the Houssa language, signifies _little_. If this be correct, which appears very questionable, there are probably several rivers of that name. Here it may be remarked, that when Dupuis (pt. II. p. 100) speaks of Saghona, the capital of Yekoo (Ako or Yariba), he means Raka, which is also called Saguda (Clapp. p. 60).]

[Footnote 237: Clapperton (Sec. Exped. p. 232) says that the capital of Zegzeg is called Quorra. He probably means to speak of the town which he elsewhere calls Guari, but the name of which, in the native maps brought home by him, is written Ḳowárah. When he speaks of Kóra (p. 133), he appears to have in view the capital of Kotú-n-Kúra. This name is, through obvious mistake, written Kotunfauda in Bello’s map. Kotú is a name of frequent occurrence, as in Kotú-n-karafi (in the maps Cuttum Curraffee), “where there is a copper mine” (Bello in Denh. Disc. II. p. 451). Karafi means metal in general. The country named Kúra is mentioned, together with Niffi and Raka, by the Kaíd ben Yusuf (Denh. Disc. I. p. 334). The slaves Boniface and Francisco agreed in representing to M. Menézes de Drummond (Hertha, pp. 13, 14), that the Kowára rises in the centre of Houssa, and that it takes its name from the country named Kuara (Kowára), through which it flows.]

Transcriber's note:

Vowel diacritics in Arabic names have sometimes been adjusted to
match their transliteration.

Unpointed Beh-shape letters in medial or initial form have been
represented with the Alef Maksura (ى) character. Other unpointed
consonants are shown as printed.

Changes in the CORRECTIONS have been done, as well as:

pg xv, Changed: "GHÁNAH, AUDÁGHOST, AÚLÍL" to: "AÚDAGHOST"

pg 6, footnote 9, Changed: "Támedelt تَلمدَلْت (MS. B.M.)"
to: "تَامدَلْت"

pg 12, footnote 24, Changed: "Agharef أَغَوَفْ MS. B.M." to: "أَغَرَفْ"

pg 111, footnote 185, Changed: "(Walcknenaer, Rech. p. 450)." to:
"Walckenaer"

pg 124, Changed: "it it manifest that Ghánah coincides" to: "it is"

pg 134, Changed: "thirteenth century, it it obvious" to: "it is"

Other spelling inconsistencies have been left unchanged.