² The chief primary sources are the military reminiscences and campaign accounts of French oflicers such as General Arlabosse, ‘Une phase de Ia lutte contre Samory (1890–1892)—souvenirs du Général Arlabosse’, Revue d'Histoire des Colonies, v (09–10 1932), VI (11–12, 1932);Google ScholarBaratier, Lt.-Col., A Travers l'Afrique (Paris, Perrin, 1912);Google Scholarde Lartigue, R., ‘Le rapport sur les opérations de la colonne du sud: prise de Samory’, Bulletin du Comité de l'Afrique Française, Renseignements Coloniaux, VII (1899);Google ScholarFrey, Lt.-Col. H., Campagne dans le Haut Sénégal et dans le Haut Niger (Paris, Plon, 1888);Google ScholarGallieni, , Deux campagnes au Soudan Francais (Paris, Hachette, 1891);Google ScholarGouraud, General, Souvenirs d'un Africain, t. I Au Soudan (Paris, Pierre Tisne, 1939);Google ScholarPéroz, Etienne, Au Niger, récits de campagne (1891–1892) (Paris, Calmann-Lévy, 1894).Google Scholar The works of Capitaine Binger, , Du Niger au Golfe de Guinée (Paris, Hachette, 1892)Google Scholar and Péroz, Etienne, Au Soudan Français (Paris, Calmann-Lévy, 1896)Google Scholar fall into a different category. Both paid visits to Samori and saw the empire from inside. In the absence of other reliable accounts one is forced to accept their assertions, except where internal contradiction makes them suspect. On the basis of his work in the archives Yves Person has found them both imaginative and unreliable, Péroz more so than Bingr: this would affect in particular my conclusions on the ‘volunteer militia’, the military governments, and Samori's assumption of the title of amīr al-rnu'minīn. I am indebted for this information to Yves Person, whose work when published will doubtless correct these points. In addition, there are the local traditions, recorded for example in Kouroubari, Amadou, ‘Histoire de l'Imam Samori’ Bull. IFAN, Parts 3 and 4 (1959)Google Scholar, and Diem, Mamadou Suleymane, ‘Un Document Authentique sur Samory’ (ed. Holas, B.), Notes Africaines, LXXIV (04, 1957), as well as numerous secondary works.Google Scholar

³ Neither ‘state’ nor ‘empire’ are wholly satisfactory descriptions of many Sudanic regimes, especially when they are differentiated as is usual by a subjective criterion of size. However, until more satisfactory terminology is developed, there is no alternative but to use ‘empire’ as an indication that the régime encompassed peoples of different cultural traditions.Google Scholar

⁴ The sources for this description are, unless otherwise stated, Binger, Du Niger… 1, 100–5Google Scholar, and Péroz, Au Soudan…, 405–15.Google Scholar

⁵ For the linguistic information in this paper I am deeply indebted to Mr Charles Bird, student at University of California, Los Angeles.Google Scholar

⁶ The fact that the sofa travelled lightly meant that he could travel fast and still arrive fresh for the battle unlike the heavily laden tirailleur. See Mévil, André, Samory (Paris: Flammarion, 1900).Google Scholar Mévil says the sofa could travel 40–50 km. a day under these conditions: Binger gives estimates for various types of travel—(a) rapid courier: 80 km. a day or less; (b) with gun and food-bag: about 30km. a day; (c) with a load on head: about 20 km. a day; (d) with loaded animals: about 16 km. a day (Binger, Du Niger…, 1, 87–8).

⁷ See, for example, Meniaud, J., Les Pionniers du Soudan, I (Paris, Société des Publicadons Modernes, 1931), 132–3 n.Google Scholar

⁸ It is only very recently that Bambara has been studied as a tone language.Google Scholar

⁹ Whether the term was introduced with firearms, or was an adaptation of an earlier term would be interesting to find out.Google Scholar

¹⁰ De Lartigue, ‘…Rapport…,’ 131 n.Google Scholar

¹¹ Karamoko, at one time Samori's favourite son, for example. After his return from a mission to France (1886–87), where he had deliberately been shown French military strength, he began to advocate a more conciliatory policy towards them. He fell out with his father and then returned to favour for a while, Finally, perhaps at the instigation of Sarangye Mori, he was walled up alive. See also Binger, Du Niger…, 1, 109Google Scholar, on the execution of sofa leaders. Some of the French reports are exaggerated, though. Thus Famako, whom Péroz described as executed by Samori for paying an unauthorized visit to the French at Bamako, was seen some months later by Binger at Sikasso. He was indeed in disgrace, serving as a simple kurusutígi, but he was still alive (Péroz, Au soudan…, 167–72;Google ScholarBinger, Du Niger…, 1, 19, 96Google Scholar).

¹² Binger, Du Niger…, I, 33;Google ScholarArcin, A., Histoire de la Guinée Francaise (Paris, Challamel, 1911), 127.Google Scholar

¹³ Péroz, Au Soudan…, 365.Google Scholar

¹⁴ A group of eighteen ‘whom Samori calls his tirailleurs’. Only eight had even Gras guns or Chassepots. Binger also describes a group of 50 men from Konia, who had been with him since his first conquests and were very devoted to Samori (Binger, Du Niger…, I, 96–7Google Scholar).

¹⁵ Dc Lartigue, ‘…Rapport…’, 113 n.Google Scholar

¹⁶ As the reports of travellers and military commanders attest, firearms were well known throughout this part of the Western Sudan. Gallieni, , Voyage au Soudan Français (Paris: Hachette, 1885)Google Scholar, gives a representative survey of trade-routes, distribution, and the nature of the weapons. See in particular pp. 321, 435, 593, and end-paper map. The existence of some trade from the north is apparent from Caillié's account of muskets made in Tunis that he saw in Timbuktu. See Caillié, René, Travels through Central Africa to Timbuctoo (1824–1828), II (London: H. Colborn and R. Bentley, 1830) 47, 51.Google Scholar

¹⁷ During the second half of the nineteenth century the chief features of the modern non-automatic military firearm were introduced, namely, rifling, the true cartridge, breech-loading, and the magazine. Guns with one or more of these features had been produced earlier, but the average army weapon of 1850 had none of them, while that of 1890 had all. In the transition period a great variety of adaptive models were introduced.Google Scholar

¹⁸ Gallieni, Voyage…, 393, 424.Google ScholarFrey, Campagne…, 220–4, gives some idea of the extent of the trade.Google Scholar For Freetown taxes and later controls, see Fyfe, C., A History of Sierra Leone (London, O.U.P., 1962), 398, 399, 500ff.Google Scholar

¹⁹ Gallieni gives the composition of Sudanese gunpowder as 7 parts saltpetre to 2 of charcoal to I of sulphur, which may be compared with 6.29: 1.1: I (Encyclopaedia Britannica, ‘Gunpowder’) and 7: 1.5:1 (Singer, Charles et al. , A History of Technology, II (Oxford: Clarendon Press, 1954–1958), 381–2).Google Scholar In earlier times, says Singer, the composition in Europe was 4, 5, or 6:2: 1, which corresponds more closely to the African formula. Sulphur was imported to the Sudan but the other ingredients were obtained locally (Gallieni, Voyage…, 423).Google Scholar

²⁰ See Person, Yves, ‘La Jeunesse de Samori’, Revue française d'histoire d'outre mer XLIX (1962), 2.Google Scholar

²¹ Gallieni, Voyage…, 435;Google ScholarFyfe, History…, 499, 517.Google Scholar

²² For some of the documents concerning Samoris approach to Sierra Leone and the halt of his armies at British request, see Fyfe, C., Sierra Leone Inheritance (London, O.U.P., 1964), 196, 202, 238–42.Google Scholar

²³ The first British mission saw Samori in Bure some time prior to June i886. Subsequently Festing (188788), Garrett (1890) and Kenney (1892) visited the empire, and Liginfali paid a visit to Freetown in JuneJuly 1885. See Fyfe, History…, 452, 489, 504;Google ScholarFyfe, …Inheritance, 240–1;Google ScholarArcin, A., Histoire…, 427, 468.Google Scholar

²⁴ Fyfe, History…, 517;Google ScholarArcin, Histoire…, 523;Google ScholarPéroz, Au Niger……, 297, 303.Google Scholar Arcin mentions ‘La Maison Bolling’ as a supplier of Samori, and that in 1892 the French seized repeaters being transported from Freetown to Samori.

²⁵ Fyfe, History…, 500;Google ScholarArcin, Histoire…, 526–39;Google ScholarFyfe, …Inheritance, 248–51;Google Scholar de Lartigue, ‘…Rapport…’, 114. In March 1894, Kunadi Keleba, a Samorian general, was at N'Zapa but must have retired eastward shortly after, for later in the year he is near Korhogo. See Diem, ‘Document…’, which from comparison with other sources clearly begins from the rainy season of 1894.Google Scholar

²⁶ De Lartigue believes he was influenced by his counsellors, most of whom originated from this area. This, in itself, would seem to be too sentimental a consideration to influence Samori (De Lartigue, ‘…Rapport…’, 114).Google Scholar

²⁷ Péroz, Au Niger…, 170 n.Google Scholar

²⁸ Péroz, Au Soudan…, 410.Google Scholar

²⁹ Binger, Du Niger…, I, 90, 97.Google Scholar

³⁰ See Meniaud, J., …Pionniers…, II, 151–219 passim.Google Scholar

³¹ These included Winchesters, Mausers (some modified to fire the Gras cartridge) Chassepots (some modified to fire metal cartridges, probably Gras) and Gras guns of various models. See Meniaud, …Pionniers…, II, 220;Google ScholarPéroz, Au Niger…, 40, 41, 161, 307.Google Scholar The Chassepot was a breech-loader with paper and linen cartridge. Introduced into the French army in 1866, it was replaced in 1874 by the Gras gun which had bolt action. The ‘Kropatschek’ was a Gras gun with Kropatschek magazine system (tube magazine under the barrel) and was the first repeater in French military use. Although introduced in 187980 it was first used in the Sudan by Frey in 1885. The Gras/Kropatschek was replaced in the French armies of the Sudan by the Lebel model 1886, a magazine repeater, first used against Samori in the 1892 campaign. See Hicks, J. E. and Jandot, A., French Military Weapons 1717–1938 (Connecticut: N. Flayderman and Co., 1964), 27–30;Google ScholarFrey, Campagne…, 60–1;Google ScholarMeniaud, …Pionniers…, I, 59; II, 260.Google Scholar See also Mevil, Samory, 99;Google ScholarMeniaud, …Pionniers…, II, 261, 292Google Scholar, for different estimates of the number of Samori's repeaters. Samori's ammunition was largely of Spandau manufacture, and apparently in each box there was a small bottle of perfume for one of Samori's wives! (Arlabosse, ‘Une phase…’, 431)Google Scholar. The amount of ammunition captured at the Tukoro powder store is revealing: 70,000 cartridges, 80,000 cartridge cases, 300 kilos of bullets, 50,000 kg. of powder. Here I follow Arlabosse op. Cit. 493Google Scholar. Baratier, A Travers…, 106;Google ScholarPéroz, Au Niger…, 237, differ slightly.Google Scholar

³² Péroz, Au Niger…, 169.Google Scholar Throughout this work there is information on the rapid arms build-up during the November 1891April 1892 period.

³³ See Gallieni, Voyage…, end-paper map. Caravans to the Gambia took three or four months (Frey, Campagne…, 136).Google Scholar

³⁴ Compare Péroz, Au Soudan…, 410Google Scholar, with Gallieni, Voyage…, 393.Google Scholar

³⁵ Meniaud, …Pionniers…, II, 136.Google Scholar Samori had supposed that his treaty with the French would mean they would supply him with modern weapons. However they had no intention of doing so, although they began to provide Samori's adversary Tiéba with repeaters. See Collieaux, M., ‘Contribution a l'étude de l'histoire de l'Ancien Royaume de Kenedougou, 1825–1898’, Bull. Com. d'Etudes hist. sci. l'A.O.F. 1924, 150.Google Scholar

³⁶ Kouroubari, ‘Histoire…’Google Scholar, Bernus, E., ‘Kong et sa region’, Etudes Eburnéennes, VIII (1960), 270.Google Scholar According to Suret-Canale, a French officer was killed at N'Zapa in 8894 when he surprised an arms convoy destined for Samori from Monrovia (Suret-Canale, J., Afrique Noire: l'ère coloniale, 1900–1945 (Paris, Editions Sociales, 1964), 198.)Google Scholar

³⁷ Mévél, Sarnory, 122;Google ScholarBinger, Du Niger…, I, 316–17; II, 228Google Scholar. Kouroubari also says ‘to Kong from the coast’. See also Marty, P., Études sur l'Islani en Côte d'Ivoire (Paris: Leroux, 1922), 219.Google Scholar

³⁸ Meillassoux, C., ‘Social and Economic factors affecting markets in Goroland’, in P. Bohannan and G. Dalton, Markets in Africa (N.W. University Press, 1962), 286, 290.Google Scholar

³⁹ Binger, Du Niger…, I, 316–17; II, 102, 105;Google ScholarFreeman, R. A., Travels and Life in Ashanti and Jaman (Westminster, Constable, 1898), 180–1, 212, 477, 479 (map).CrossRefGoogle Scholar However, neither Binger nor Freeman seem to have seen guns on sale in Bonduku (Binger, Du Niger…, II, 164;Google ScholarFreeman, Travels…, 234ff.).Google Scholar

⁴⁰ Freeman, Travels…, 546.Google Scholar

⁴¹ Mévil, Samory, 163;Google Scholarde Lartigue, ‘…Rapport…’, 113 n., 131;Google ScholarClaridge, W. W., A History of the Gold Coast and Ashanti, II (London, John Murray, 1915), 426–9.Google Scholar

⁴² Péroz, Au Niger…, 219. See also ibid. 364.

⁴³ Ingold, , Samory: sanglant et magnifique (Paris, Scorpion, 1961), 51.Google Scholar

⁴⁴ Baratier, A Travers…, 84;Google ScholarKouroubari, ‘Histoire…’, 565.Google Scholar See also Marty, Islam en Côte d'Ivoire…, 208, who reports that when Samori was at Dabakala, almost all the blacksmiths there emigrated from Bonduku.Google Scholar

⁴⁵ Meniaud, …Pionniers…, II, 163.Google Scholar See also Mévil, Samory, 182;Google ScholarBaratier, A Travers…, 84;Google ScholarKouroubari, ‘Histoire…’, 564;Google ScholarMeniaud, …Pionniers…, II, 149.Google Scholar

⁴⁶ The citations from Ingold and Meniaud mention Gras guns, while the Mévil and Kouroubari accounts describe the weapons as Kropatscheks.Google Scholar

⁴⁷ Arlabosse, ‘Une phase…’, 431;Google ScholarPéroz, Au Niger…, 267.Google Scholar

⁴⁸ Mévil, Samory, 175.Google Scholar

⁴⁹ De Lartigue, ‘…Rapport…’, passim. is the source for this paragraph.Google Scholar

⁵⁰ Gouraud, Souvenirs…, 209.Google Scholar

⁵¹ Péroz, Au Soudan…, 412–13;Google ScholarArlabosse, ‘Une phase…’, 431;Google ScholarMeniaud, …Pionniers…, 1, 359.Google Scholar

⁵² See notes 37, 38.Google Scholar

⁵³ Péroz, Au Soudan…, 413.Google Scholar

⁵⁴ See Binger, , Du Niger…, I, 130.Google Scholar

⁵⁵ Ibid. 1, 100; Péroz, Au Niger, 293–4, 312–13.Google Scholar It does not seem profitable to discuss costs of firearms and horses outside the general framework of the long-distance trade, which cannot be entered into here. Roughly, however, a musket might cost from 12,000 to 25,000 cowries, and a horse from 200,000 to 400,000 cowries. In terms of captives, according to Meniaud, a repeater exchanged for 24, and a horse for from 4 to 52. See Binger, Du Niger…, 1, 27, 357; II, 102, 105;Google ScholarGallieni, Voyage…, 436–7;Google ScholarMauny, R., Tableau géographique de l'Ouest Africain an lnoyen age (Dakar, IFAN, 1961), 283–6;Google ScholarMensaud, …Pionniers…, II, 158.Google Scholar

⁵⁶ See Mauny, Tableau…, 283–6, for a general discussion of horses in West Africa.Google Scholar

⁵⁷ Meniaud, …Pionniers…, I, 65.Google Scholar See also Binger, Du Niger…, I, 130;Google ScholarGallieni, Voyage…, 449;Google ScholarPéroz, Au Niger…, 149.Google Scholar

⁵⁸ Meniaud, …Pionniers…, II, 136.Google Scholar

⁵⁹ Ibid. I, 65; Kouroubari, ‘Histoire…’, 555;Google ScholarBinger, Du Niger…, I, 316–17; II, 105;Google ScholarBernus, ‘Kong et sa region’, 265–75. Most of the Mossi horses came from the Yatenga region.Google Scholar

⁶⁰ See Collieaux, A., ‘Details rétrospectifs sur l'histoire des dernières operations contre Samori et la prise de l'Almamy, 1897–8’. Bull. Corn. d'Études hist. sci. de l'A.O.F., 1938, 296.Google Scholar

⁶¹ The figure of 5,000 attributed to Péroz by Frey is probably an over-estimate (Frey, Campagne…, 107–8).Google Scholar Archinard in 1895 reported that Samori was trying to increase cavalry strength to 2,000–3,000; certainly this force was never evident in campaigns (Meniaud, …Pionniers…, 158).Google Scholar

⁶² Binger, Du Niger…, I, 99–100, gives the strength at Sikasso. Binger was met elsewhere by a force of 32 cavalry, though he remarks that only 12 were of passable quality.Google Scholar (Ibid. I, 19–20).

⁶³ See Meniaud, …Pionniers…, I, 161; II, 183;Google ScholarMévil, Samory, 175;Google Scholarde Lartigue, ‘…Rapport…’, 115, 136;Google ScholarPéroz, Au Niger…, 12–39, 169, 177, 285. Diem describes however the capture of 2,500 horses in a campaign against Babemba in the dry season of 189495. (Diem, ‘Document…’). See also notes 72, 75.Google Scholar

⁶⁴ Binger, Du Niger… I, 27.Google Scholar

⁶⁵ For these traditions see Péroz, Au Soudan…, 388–400;Google ScholarBinger, Du Niger…, I, 140–50Google ScholarKouroubari, ‘Histoire…’, 544–6;Google ScholarKalil, Fofana, ‘Almamy Samori: l'homrne et son œuvre’, Recherches Africaines, I (05–03 1963), 7–14;Google ScholarDuboc, F. F., Samory le Sanglant (Paris, Sfelt, 1947), 197–200. The last citation is to a document by Mohamed Denfa received by the French presumably at some time between 1882 and 1887.Google Scholar

⁶⁶ Péroz, Au Soudan…, 392.Google Scholar

⁶⁷ For example, Amara Diali (a griot), Mori-fin-dian (counsellor and childhood friend), Nassikha Mahdi (a brother of Samori's who died in about 1886), Alpha Umar (a cousin of Samori's), Mori Ture (who had built up an army and territory in the Odienne area and then allied with Samori, marrying his eldest daughter). See Binger, Du Niger…, I, 132, 134;Google ScholarPéroz, Au Soudan…, 336, 397, 404–5;Google ScholarMarty, Islam en Côte d'Ivoire…, 845, 113.Google Scholar

⁶⁸ Péroz, Au Soudan…, 398.Google Scholar

⁶⁹ Such as Liginfali, Bilali, Sekuba and Kunadi Keleba. None is mentioned as a relative of Sainori. (There was almost certainly more than one general called Sekuba, though it is impossible to discuss this here.)Google Scholar

⁷⁰ Péroz, Au Soudan…, 376;Google ScholarBinger, Du Niger…, I, 122. The Binger calculation, lower of the two, is methodologically spurious. Péroz gives no basis at all for his estimate.Google Scholar

⁷¹ Péroz, Au Soudan…, 408.Google Scholar This figure parallels that given by Mohamed Denfa: when the army was divided up between several chiefs, he says, each had a force of 12,000 warriors. The two sources may not be independent (Duboc, Samory…, 197–200).Google Scholar

⁷² Frey, Campagne…, 107–8.Google Scholar The largest army was that of 20,000 which campaigned against the French at the beginning of 1885: its reserve of 5,000 was not used as such, even in the battle of Kokoro (Péroz, Au Soudan…, 12, 324, 408)Google Scholar. The cavalry strength may have been as great as 1,000 since the army of 10,000 at the battle of Kommodo had 400500 cavalry and also Samori is reported to have lost 400–500 horses in the campaign (Péroz, Au Soudan…, 12, 277;Google ScholarDuboc, Samory…, 61Google Scholar). There were certainly other forces in the field at the time: Amara Diali in the south-east and Liginfali in the south-west, and perhaps Bilali also (Binger, Du Niger…, I, 134, 749).Google Scholar

⁷³ Binger, Du Niger…, I, 96–99, gives a complete list. In the one encampment that he counted rather than estimated there were betweeen 320 and 340 troops under Fabu Toure, drawn from Ouolosegubu, Kangare and Faraba.Google Scholar

⁷⁴ See note 72.Google Scholar

⁷⁵ Frey, Campagne…, 106.Google Scholar Onesource reports that there were 1,000 cavalry with the army (Kouroubari, ‘Histoire…’, 549).Google Scholar

⁷⁶ Meniaud, …Pionniers…, I, 159–61, 177, 158;Google ScholarMévil Samory, 33;Google ScholarPeroz, Au Soudan…, 10.Google Scholar Estimates for Keniéra vary wildly from 400 to 10,000 men! Mévil's of 4,000 is probably most accurate, since there were four diassas (camps surrounded by palisades) around the town, each possibly containing a ‘battalion’ of 1,000.

⁷⁷ The only evidence found which differentiates between regular army and reservists is in Péroz's account of his visit to Kenieba-Kura for the treaty of i886. Behind Samori was his guard of 500 young men. On one side was Malinke Mon with 200 cavalry, and behind him five companies of zoo young men each (regular army from five provinces?). ‘On the other flank, rather scattered but in distinct groups, were ranged the troops called from neighbouring regions to assist at this imposing ceremony…’ (Meniaud, …Pionniers…. I, 254).Google Scholar

⁷⁸ See, for example, the description of the siege of Niafadié in 1885 (Péroz, Au Soudan…, 296–302).Google Scholar

⁷⁹ Ibid. 303–17.

⁸⁰ For the major engagements see Mévil, Samory, 99;Google ScholarMeniaud, …Pionniers…, II, 163;Google ScholarPéroz, Au Niger…, 160, 166, 285, etc.Google Scholar The numbers for the ‘guerilla’ detachments are often not given though it is evident they are small. See, for example, Péroz, Au Niger…, passimGoogle Scholar, and Arlabosse, ‘Une phase…’, passim.Google Scholar A clear instance was in the campaign against the Combes column (February–March 1893), when Samori retired east and south until the French were exhausted and, when Combes turned back, split the army into small groups to harass him (Mévil, Samory, 112–17).Google Scholar

⁸¹ Péroz, who had once written that Samori's guns ‘never fired at more than thirty paces with any chance of success’, now found himself fired at at a distance of 700 yards with considerable accuracy! (Péroz, Au Soudan…, 347;Google ScholarPéroz, Au Niger…, 211Google Scholar). Péroz and Baratier amongst others were high in their praise of the Samorian army.

⁸² See Meniaud, …Pionniers…, 11, 151;Google ScholarPéroz, Au Niger…, 40.Google Scholar Baratier cites the case of N'Golo, a tirailleur who became a Samorian sofa-kùn, and the more dramatic instance of Koruba Mussa, who deserted to the French at Kankan in 1891. Enrolled in the spahi force (cavalry) he disappeared after being sent on a mission from the battlefield. He delivered the message, which is ironic, and was assumed killed. However, in the campaign of 1892 the charging spahi force was brought to a halt in one battle by the voice of Koruba Mussa calling from the line of sofa. He had acquired valuable experience in the ‘winter school’! (Baratier, A Travers…, 80–4;Google ScholarArlabosse, ‘Une phase…’, passimGoogle Scholar). French commands and techniques were introduced to a considerable extent, as is evident from a comparison of the descriptions by Binger and Péroz in 1887 with that of the French mission of 1897. On occasion French commands and bugle calls would be used by the sofa to confuse their enemy (Binger, Du Niger…, I, 104ff.;Google ScholarPéroz, Au Soudan…, 411;Google ScholarMévil, Samory, 174–6;Google ScholarPéroz, Au Niger…, 259).Google Scholar

⁸³ The battles of Sombi-ko and Diaman-ko on so January 5892 are a good example.Google Scholar

⁸⁴ De Lartigue, ‘…Rapport…’, 113 n.Google Scholar

⁸⁵ Gouraud, Souvenirs…, 76;Google ScholarMeniaud, …Pionniers…, II, 251.Google Scholar The force had 300 repeaters and 40 cavalry at this time. The battle is described by Arlabosse and by Péroz (Arlabosse, ‘Une phase…’, 465–9;Google ScholarPéroz, Au Niger…, 28–39).Google Scholar

⁸⁶ Arcin, Histoire…, 483–4.Google Scholar

⁸⁷ Tauxier, L., Le Noir de Bondoukou (Paris, Leroux, 1921), 116–17;Google ScholarClaridge, History of the Gold Coast…, 427–8.Google Scholar Duboc is almost certainly mistaken in saying 700–800, since the French mission at Dabakala reported at this time that the major part of the army (i.e. more than 2,000 men) was with Sarangye Mon (Duboc, Samory…, 108;Google ScholarMévil, Samory, 175ff.).Google Scholar

⁸⁸ Meniaud, …Pionniers…, II, 177, 179. Kali Sidibé's cavalry included some of Ahmadu's ex-talibés.Google Scholar

⁸⁹ Diem, ‘Document…’.Google Scholar

⁹⁰ Mévil, Samory, 188–.Google Scholar

⁹¹ Baratier, A Travers…, 69–70.Google Scholar

⁹² Ibid. 72–3. See also Péroz, Au Niger…, 361–2.Google Scholar

⁹³ Kouroubari, ‘Histoire…’, 559Google Scholar

⁹⁴ See de Lartigue, ‘…Rapport…’, 113 n.;Google ScholarMévil, Samory, 189, 212–13.Google Scholar

⁹⁵ De Lartigue, ‘…Rapport…’, 114ff.Google Scholar