¹ Cooley, W. D., The Negroland of the Arabs (London, 1841), 79–82.Google Scholar

² Binger, L. G., Du Niger au golfe de Guinée (Paris, 1892), i, 56–7.Google Scholar

³ Delafosse, M., Haut-Sénéegal-Niger (Paris, 1912), ii, 180–2.Google Scholar

⁴ Vidal, J., ‘Au sujet de l'emplacement de Mali (ou Melli), capitale de l'ancien empire mandingue], Bull. Com. Et. hist. et sci. de l'A.O.F. viii (1923), 251–68 and 606–19 (‘Un problème historique africain: le véritable emplacement de Mali]).Google ScholarGaillard, M., ‘Niani, ancienne capitale de l'empire mandingue’, B.C.E.H.S.A.O.F. viii (1923), 620–36.Google ScholarDelafosse, M., ‘Le Gâna et le Mali et l'emplacement de leurs capitales], B.C.E.H.S.A.O.F. ix (1924), 479–542.Google Scholar

⁵ The Ta'rikh al-Fattāsh was compiled in the mid-seventeenth century by a certain Ibn al-Mukhtār from traditions he gathered and earlier writings of members of the Ka'ti family (sixteenth and early seventeenth centuries). It states (p. 38) (text) = p. 66 (trans.): ‘The town [or district—balad[ in which was the abode of sovereignty (dār al-imāa) of Mallj-koj [the ruler of Mali] was named Jāriba and [he had] another called Yani’. ‘See also p. 56 (text) = p. 108 (trans.) where Yani' alone is named as the town of the Malli-koi. The latter passage occurs in only one manuscript.Google Scholar

⁶ Niane, Dj. T., Recherches sur l'empire du Mali au moyenâge (Conakry (Mém. Inst. Nat. de Recherches et Documentation no. 2) 1962), 46–50.Google Scholar

⁷ See Filipowiak, M., ‘L'expédition archéologique Polono-Guinéene à Niani¯ en 1968’, Africana Bull. (Warsaw) xi (1969), 107–17. He lists the following points which correspond with features mentioned (or implicit) in Ibn Battūta's narrative: a (possible) mosque site; an open space near the latter; an area where an ‘Arab’ community is said to have lived— Larabouro; a cemetery with ‘pronounced Muslim characteristics' at the north-west of the town where a track comes in, and some faint traces of what might have been a fortification. These sort of characteristics, it would seem to me, would have been present in almost any town of the Western Sudan of any significance.Google Scholar

⁸ ‘Ré-examen de l'itinéraire d'Ibn Battuta entre Walata et Malli’, paper presented to the Conference on Manding Studies, School of Oriental and African Studies, June 1972, being a revised and expanded version of his article ‘L'itinéraire d'Ibn Battuta de Walata a Mali’, J. Afr. Hist. xiii, 3 (1972), 389–95. My article published now was also presented as a paper to the Conference on Manding Studies and subsequently revised to take into account the hypotheses put forward by Meillassoux.Google Scholar

⁹ Khaldūn, Ibn (text). Kitāb al-'ibar, ed. de Slane, Bn. M., 2 vols. Alger, 1847–1851. (trans.)Google ScholarHistoire des Berbères, trans. de Slane, Bn. M., new ed., 4 vols., Paris, 1925–56. i, 267 (text) = ii, 116 (trans.).Google Scholar

¹⁰ Africanus, Leo. Description de l'Afrique, trans. Epaulard, A., 2 vols., Paris, 1956, ii, 466.Google Scholar The Italian text was first published by Ramusio, Giovanni Battista in his Navigationi et Viaggi. See 3rd ed., Venice, 1563, i, f. 78, for the section on Mali.Google Scholar

¹¹ Al-Umari, (text), al-Munajjid, S., Mamlakat Mālī 'ind al-jughrāfiyyin al-muslimīn, Beirut, 1963, vol. i, 43–70. p. 52, n.l.Google Scholar

¹² Al-'Umari, , L'Afrique moins l'Egypte, trans. Gaudefroy-Demombynes, , Paris, 1927. pp. 43, 44, 46.Google Scholar

¹³ op. cit. pp. 53, 57, 60.Google Scholar

¹⁴ Ibid. p. 52, n.l.

¹⁵ op. cit. i, 267 (text) = ii, 116 (trans.) So also the Beirut edition of the text (1956–61), Vi, 419.Google Scholar

¹⁶ Al-Qaiqashandī, , Subh al-a 'shā fī sinā'at al-inshā' (Cairo, 1912), V, 283.Google Scholar

¹⁷ op. cit. p. 53 (continuation of n.l. to p. 52).Google Scholar

¹⁸ Subh, v, 283, n.l.Google Scholar

¹⁹ 46 (text) = 60 (trans.)Google Scholar

²⁰ See Dozy, R., Supplément aux dictionnaires arabes (Leiden, 1881), under the root b.r.d. The barīd of the East was twice the length of that of the West.Google Scholar

²¹ (al-Mukhtār, Ibn), Ta'rikh al-Fattāsh, ed. and trans. Houdas, O. and Delafosse, M., Paris, 1913–1914. 38 (text) = 66 (trans.)Google Scholar

²² A1-Sa'di, , Ta'rīkh al-Sūdān, ed. and trans. Houdas, O., Paris, 1898–1900. 10 (text) = 19 (trans.)Google Scholar

²³ Battūta, Ibn, Tuhfat al-nuzzār: Voyages, ed. and trans. Defréméry, C. and Sanguinetti, B. (1854), reissue with preface and notes by V. Monteil, 4 vols., Paris, 1969. iv, 397.Google Scholar

²⁴ Ibid. iv, 390–1.

²⁵ Ibid. iv, 385, 387.

²⁶ Ibid. iv, 424.

²⁷ Ibid. iv, 394.

²⁸ Barth, H., Travels and Discoveries in North and Central Africa (New York, 1857–1859, repr., London, 1965), iii, 696–7.Google Scholar

²⁹ His strictures on the town's laxness of Islam are not echoed by Ibn Battūta, who describes its people as ‘long-standing Muslims, pious and studious’.Google Scholar

³⁰ See al-Sa'dī (text), 10, 72, 179, 271 and Ibn al-Mukhtār (text), 61, for Zāgha, and al-Sa'dī, 77 and Ibn al-Mukhtār, 39 for Zāra.Google Scholar

³¹ Boyer, G., Un peuple de l'ouest soudanais: les Diawara, Mém. I.F.A.N., xxix (1953), 34.Google Scholar

³² See, for example, Lenz, O., Timbouctou (Paris, 1886), ii, 221, where he describes the extensive millet fields stretching some hours journey from Kala, though the only important trade item he noted was slaves. Marty presents a much more glowing picture; a large commercial town, dating back to the fifteenth century, where caravan routes converge from Walata, Néma etc. in the north and Sansanding, Ségala and Monimpé in the south. ‘All the Moors and Fulbe of the north who come into Bambara and Mossi country in search of cotton strips for tents, grain, kola nuts etc., pass through Sokolo. All the Futanke Dyula who go up to the Fulbe of the north pass there also.’ (Etudes sur l'Islam et les tribus du Soudan, Paris, 1920, 1V, 95–6.)Google Scholar

³³ This Kābara is, of course, not to be confused with Kabara, the port of Timbuktu. From al-Sa'dī (p. 13 text p. 25 trans.) one gathers that the territory of the ruler of Kābara bordered the western limits of the territory of Jenne and other references (text 16, 48, 61 = trans. 29, 78, 99) indicate that it was an important centre of Islam from as early as the fourteenth century.Google Scholar

³⁴ op. cit. iv, 395.Google Scholar

³⁵ For Kābara see n. 33 above. Al-Sa'dī (text, 10 = trans. 19) states that the first of the rulers in Kala territory across the river (Niger), the Kūkiri-koi, governed a territory which bordered the western edge of the land of Zāgha.Google Scholar

³⁶ His remarks on the abundance of provisions on the route are, of course, post facto.Google Scholar

³⁷ op. cit. iv, 394–5.Google Scholar

³⁸ He uses the phrase ‘I went down to the ‘Nile’ [at Kārsakhū] one day’, which seems to indicate a stay longer than a mere halt. As argued above, he was there long enough to find out valuable information about the course of the Niger and the town of Zāgha.Google Scholar

³⁹ op. Cit. iv, 397.Google Scholar

⁴⁰ Ibid. iv, 425.

⁴¹ Sokolo, which lies some 30 feet below the level of the Niger at Sansanding, was once in the area of an ‘overflow’ lake of the Niger.Google Scholar

⁴² op. cit. iv, 429–30.Google Scholar

⁴³ Ibn Battūta went all the way from Walāta to Mali on a camel and he says: ‘I dismounted at the cemetery of the town.’ One can scarcely imagine the caravans which Ibn Khaldūn says came to Mali crossing three major rivers to reach it. Ibn Battūta approached Mali during the flood season, but perhaps in winter (the usual season for big caravans) even the Sansara would have been fordable. In the discussion referred to in the postscript, Paulo Farias also observed that it is most unlikely that the main route from Walāta to the Mali capital would have crossed the Bambuk gold fields, as seems to be implied in Meillassoux's hypothetical route. All indications are that it was Malian policy to keep the location of the sources of gold a secret from the outside world.Google Scholar

⁴⁴ See Park, Mungo, Travels in the Interior Districts of Africa (London, 1835), i, 344–5. He took four days to travel from Sansanding to Niamina and at a distance of two days' journey beyond that he came to a river (which he calls the Frina), flowing into the Niger from the north, which was so deep that he had to hire a canoe to ferry himself and his horse across.Google Scholar

⁴⁵ It seems unlikely that if Mali was ruled from beyond the Gambia river it could have conquered territory as distant as Gao (or even beyond), or that the Songhay could have occupied the capital as they did in 1545–6. (See al-Sa'dī, 98 (text) = 161 (trans.).)Google Scholar

⁴⁶ This, as I understand it, is the view of Ivor Wilks, at least as far as Beetuu is concerned (see his chapter ‘The Mossi and Akan States 1500–1800’ in Ajayi, J. F. and Crowder, M. (eds.), History of West Africa (London, 1971), i, 355–7) in contrast to the view of Yves PersonGoogle Scholar (in Vansina, J., Mauny, R. and Thomas, L. V., the Historian in Tropical Africa (London, 1964), 330–1). In the case of the ‘Bītu’ of the Arabic texts (and its possible transformations into Bī'u, Bīghu, Bayku etc.), it may be that the name was originally given to a location in, or near, the gold-producing area of Bure and subsequently applied also by the Dyula to a new settlement founded by them far to the south. The references to Bītu in the Ta'rikh al-Sūdān (Arabic text pp. 11, 16–17, 21) merely indicate that it was a gold- producing area where Dyula lived. Bīt is also mentioned in the Fath al-Shakūr of al-Bartilī (written in 1799/1800), where it is describcd as the southern limit of the ‘land of Takrūr’. All other references to the ‘land of Takrūr’ indicate that this term was applied only to the predominantly Muslim sahel and savannah lands of West Africa, and never to the mainly non-Muslim forest zonesGoogle Scholar (see my article in B.S.O.A.S. xxvii (1964), 572).Google Scholar