¹ The term ‘Middle Belt’ is used here to refer to that area of savanna on either side of the Niger–Benue axis including the present-day Kwara State, Niger Division (formerly Niger Province) of North-West State, and Benue Plateau State. While there is no exact definition of the term, its use here would most probably be regarded as acceptable by others who have used it, except as a political term.

² In his discussion of this region, Harrison Church points out, ‘This belt … suffered from slave raiding from both north and south…’. Harrison Church, R. J., West Africa (1966), 167–8.Google Scholar Hance, in his general work on Africa writes, concerning the same area, that ‘the low densities of the middle belt are explained by early slaving in the area from both north and south. …’ Hance, W. A., The Geography of Modern Africa (1965), 164.Google Scholar Buchanan and Pugh, co-authors of the standard work on Nigerian geography, accept that ‘ruthless slave raiding continuing in many areas until the late nineteenth century decimated the population of much of the Middle Belt…’. Buchanan, K. M. and Pugh, J. C., Land and People in Nigeria (1962), 62.Google Scholar The most recent general work dealing with the geography of West Africa points in the same direction. Its authors, Morgan and Pugh, refer to ‘…the depopulation of the West African “Middle Belt” by slave raiding from north and south …’ Morgan, W. B. and Pugh, J. C., West Africa (1969), 176.Google Scholar It is Gourou who gives the matter of slave raiding its full wide screen effect. ‘In the dry season’, he relates, ‘… the [Fulani] horsemen ravaged the plains of central Nigeria, massacring or carrying off the inhabitants as slaves. These raids increased in intensity during the eighteenth century [sic] the Ilorin Yorubas and the Nupe were plundered, weakened and made to submit to Fulani authority, whilst the Benue plains were even more fiercely devastated; the Jukun kingdom was not even converted into a vassal stage, but was ruthlessly destroyed; its inhabitants were dispersed and the area became a dry-season grazing ground for Fulani herds’. Gourou, P., The Tropical World (1968), 114. I am grateful to Mr J. G. U. Adams of the London School of Economics for these references.Google Scholar

³ Clapperton and Lander, who visited Nupe in 1825, were the first to record their impressions of the country. See Clapperton, H., Journal of a Second Expedition.… (1832),128.Google Scholar

⁴ Flint, J. E., Sir George Goldie (1960), 246.Google Scholar

⁵ The Incas are said to have justified their conquests by denouncing the barbarous customs of the tribes they subdued, and the Spaniards made a great deal of the human sacrifices they found in Mexico—as the Romans did long before of those of Carthage’. Kiernan, V. G., The Lords of Human Kind (1969), 36.Google Scholar

⁶ This view is put forward by Fage, J. D., ‘Slavery and the slave trade in the context of West African history’, J. Afr. Hist. x, 3 (1969).Google Scholar

⁷ Khadduri, M., War and Peace in the Law of Islam (1955), 105.Google Scholar

⁸ Amana is a word of Arabic origin used in both Hausa and Nupe to mean ‘entente’, often with the implicit meaning of submission of one party to another.

⁹ In 1864, for instance, the Nupe Etsu (king), Umaru Majigi, allied himself with his Muslim neighbour, Ilorin, and non-Muslim Ibadan, against the intractable Igbiras of the Kabba area. See Dupigny, E. G. M., Gazetteer of Nupe Province (1920), 18.Google Scholar CMS CA₃ M₃John, to Crowther, , 7 08 1875,Google Scholar end, in Crowther, to Lay Secretary, 27 08 1875 and CMS CA₃ M₃ ‘Annual Report of Bishop Crowther.…1875’. An Ilorin version is given in Ta'lff akhbār al-qurûn miii umarā' bilād Ilurīn by Ahmad ibn Bakr. I owe my thanks to Dr B. Martin and Dr P. Morton-Williams for this reference.Google Scholar

¹⁰ The attack of the allies against the Igbiras (see footnote 9 above) seems to have met with little success. Between 1878–81 the same Nupe ruler fought against the Akokos, whom he failed to subdue. See CMS G A 3/0109 ‘Report of the Lokoja Station… by Rev, T. C. John.

¹¹ Ajelai is the plural of ajele. Nadel (A Black Byzantium) derives the word from the Arabic ajala, ‘to dispatch’. Awe adduces that it is a Yoruba word meaning ‘owner of the land’.Google Scholar (Awe, B., ‘The Ajele system: A Study of Ibadan imperialism in the nineteenth century’, Journal of the Historical Society of Nigeria, vol. 3 no1 (12,1964).) In the Nupe context, an ajele was the representative of the central government in a conquered territory outside of Nupe proper.Google Scholar

¹² The Igbiras of Kabba throughout the nineteenth century, successfully resisted the pressure of the Nupe armies. See footnote 9 above.

¹³ For the Bassas see N(ational) A(rchives) K(aduna) SNP 393 P/1918. The southward emigration of the Yagbas is mentioned in NAK SNP 17/2 3965. The Kukurukus are mentioned in NAK SNP3875/1912.

¹⁴ See NAK Lokprof10/1918, NAK KABbadist Kabba and NAK SNP393P/1918.

¹⁵ See Rhodes House, MSS Afr. a. 101. Walter, Watts to Council, Royal Niger Company, 10 11 1895.Google Scholar

¹⁶ The Royal Niger Company and its official successor, the Protectorate government, clamoured loudly about the evils of ‘slave raiding’ and, by implication, the benefit which would be derived by the destruction of the states engaged in it. Thus their replacement by more benevolent regimes under the ultimate direction of enlightened British officials might be justified on humanitarian grounds. Goldie, for instance, in his preface to Vandeleur's account of the British invasions of Bida and Ilorin, wrote of the ‘appalling destruction of life resulting from slave-raiding’, pointing out that ‘under reasonable conditions of security the existing population might soon be trebled…’ Bida armies, he noted, ‘held down’ vast regions, ‘treating the less warlike inhabitants as cattle to be raided when wanted’. See Vandeleur, S., Campaigning on the Upper Nile and Niger (1898), xviii.Google Scholar The same theme was elaborated on in Lugard's 1901 report to the Colonial Office. ‘There is, probably’, he wrote, ‘no part of the “Dark Continent” in which the worst forms of slave-raiding still exist to so terrible an extent, and the slave raiders are not even provident of their hunting grounds, for those who are useless as slaves are killed in large numbers, the villages burnt, and the fugitives left to starve in the bush.’ Further evidence against the slave raiders was presented in Lugard's 1904 report, when he pointed out that in the fifty years since Barth had visited Northern Nigeria the population of the region had dropped from between thirty to fifty to less than nine million. See Annual Reports, Northern Nigeria, 1900–1911, 10–lI, 292.Google Scholar

¹⁷ See Perham, M., Lugard… II, 43, who, quoting Lugard, implies that the whole Middle Belt was ‘static and impoverished’ as a result of ‘perpetual man-hunting’.Google Scholar

¹⁸ For Yagba, see NAK SNP 558P/1913, for Kukuruku, NAK SNP 3875/1912 and for Bunu, NAK SNP 12P/1914.

¹⁹ See NAK SNP 558P/1913.

²⁰ For reference to Middle Belt slaves landed in Sierra Leone, see Curtin, P. and Vansina, J., ‘Sources of the nineteenth century slave trade’, J. Aft. Hist. v, no. 2 (1964), 195–6.Google Scholar Accounts of slaves of Middle Belt origin are also suggested in De Castelnau, F., Renseignments sur L'Afrique Central et sur une nation d'hommes à queue, (1851),Google Scholar and Etienne, I., ‘La Secte Musulmane des Malés et leur revolte en 1835’, Anthropos, IV (1909), 99, 414.Google Scholar I am grateful to Mr John Layers of Abdullahi Bayero College, Kano, for these references. Nupes in Cuba are mentioned in Bascom, W., ‘The Yoruba in Cuba’, Nigena, no. 37 (1951), 27.Google Scholar

²¹ Clapperton, H., Journal of a Second Exhibition (1829), 128.Google Scholar

²² Ibid. 138.

²³ Ibid. 122.

²⁴ Lander, R. and Lander, J., Journal of an Expedition (1832), I, 233.Google Scholar

²⁵ Ibid. 11, 298, 313.

²⁶ Ibid. 111, 18–19, 27.

²⁷ Ibid. 1, 28.

²⁸ Ibid. 1, 194.

²⁹ Ibid. 11, 260. That urban concentration was a corrolary of rural insecurity in some instances is likely, although in some areas, notably Egga and Bini, especially after 1857, the rise in urban population was more likely due to other factors. While heavily populated towns may have developed at the cost of rural populations, insofar as the rural peoples moved into local towns there would have been no loss to the area as a whole.

³⁰ Ibid. 11, 312.

³¹ Ibid. 11, 36.

³² Laird, M. and Oldfleld, R. A. C., Narrative of an Expedition (1837), 23, 11.Google Scholar The en slavement of the Kakandas is reflected in the numbers of these people in Sierra Leone and the date (1836) of the capture of Koelle's informant. See Curtin, and Vansina, ,Google Scholaribid. 200.

³³ Schon, J. F. and Crowther, S., Journals.… (1842), 158–9.Google Scholar

³⁴ Whitford, J., Trading Life in Western and Central Africa (1967 ed.), 236.Google Scholar

³⁵ Schon and Crowther, ibid. 180.

³⁶ Crowther, S. and Taylor, J., The Gospel on the Banks of the Niger (1859), 71–2.Google Scholar

³⁷ Whitford, ibid. 251.

³⁸ Schon, and Crowther, ,Google Scholaribid.324.

³⁹ FO 2/34 Glover, to Murray, , 28 02 1860.Google Scholar

⁴⁰ FO 97/34 Baikie, to Russell, , 13 03 1862.Google Scholar

⁴¹ Information on Nupe is largely derived from research for my Ph.D. thesis, which is in progress. See also Nadel, S. F., A Black Byzantium, 11, 181.Google Scholar

⁴² Bida seems to have grown prodigiously. In the latter half of 1857 it had an estimated population of 60,000. See CMS C A 3/02 ‘Extracts of a letter from Dr Davies to Consul Campbell…27 September 1857’. Just over two decades later another estimate put it at 100,000. MMS Yoruba 1877–82, Milum, to Kilner, , 24 10 1879.Google Scholar

⁴³ FO 2/27 May, to Malmesbury, , 13 11 1858.Google Scholar

⁴⁴ Flegel, E., Mittheilungen der afrikanischen Gesellschaft in Deutschland, III (1882), 60–63.Google Scholar

⁴⁵ Thomson, J., Good Words (1886), 253.Google Scholar

⁴⁶ NAK Lokprof 296. ‘Notes on Igbira history’. Under their leader, Omadibi, the Igbiras of what is now Kabba Division, Central West State, ‘withdrew from the outlying parts of their territory and consolidated their positions in various strong centres at or near…hills…. See also NAK SNP 304P/1916 and NAK SNP 12 P/1914. Another people who became more centralized as a result of Muslim attacks were the Madas, who were invaded from Keffi. See Ames, C. G., Gazetteer of Plateau Province (1932), 256.Google Scholar

⁴⁷ See, for instance, NAK SNP 17/2 3965, which notes: ‘A close connexion was kept up between the Bunus and the West of Yagba, whence came a constant stream of immigrants to the new settlements. Most of the immigrants during the nineteenth century were Yagbas seeking to escape the Nupe raids.’

⁴⁸ See NAK SNP 558P/1913, which comments: ‘From the day they reached there [i.e. Bida] until the year it was broken, we find Yagbas dribbling back to their homes, and those whose return is recent are those who were sold into slavery further north…’.

⁴⁹ Thomson, ibid., 254.

⁵⁰ See Heath, F. (trans.), A Chronicle of Abuja (1962), 22–4; Church Missionary Intelligencer XLVII, (1896), 355; CMS G3 A3/1896/70; ‘A journey to Koeffe’, L. H. Nott.Google Scholar

⁵¹ Tiv traditions themselves claim that the area into which the Tivs moved was but sparsely populated. Clashes with the Idomas, for instance, did not come until the nineteenth century, when the Raav and Mbakpa clans were forced, by other clans, north and west. I am grateful for this information to Mr D. C. Dorward of S.O.A.S.

⁵² By 1890 both Keffi and Nasarawa had crossed to the south bank. See FO 84/2109 ‘List of Documents…’ by Major C. M. McDonald.

⁵³ It may be, of course, that the present population density of Yoruba is a product of relatively recent times, although this seems never to have been suggested. Further, it is difficult to find any nineteenth century visitor to the area commenting on the sparse population of central Yorubaland. Adams's description, however exaggerated, of the Oyo army at the turn of the eighteenth century as having 100,00 men would suggest that Yorubaland was fairly populous even before the break-up of the Oyo Empire. Captain Jones reported 17,000 soldiers in the Egba army in the 1850s. See Ajayi, J. F. A. and Smith, R., Yoruba Warfare in the Nineteenth Century (1964), 10, 139.Google Scholar Iboland, another area which, even if its internal wars were not on the scale of those of Yorubaland, certainly produced a large number of slaves, seems, on botanical evidence also to have been densely populated in the nineteenth century. See Allison, P., J. Afr. Hist. III, no. 2 (1962), 243.Google Scholar

⁵⁴ The problem of Middle Belt population densities as well as the bio-geographical factors influencing them have been brought to my attention by Mr J. G. U. Adams of L.S.E., to whom I would like to express my gratitude.

⁵⁵ Buchanan, and Pugh, Google Scholar, ibid. 62.

⁵⁶ F.A.O., Agricultural Development in Nigeria, 1965–1980, 1966, 9.Google Scholar

⁵⁷ F.A.O. Ibid.

⁵⁸ Buchanan and Pugh, Ibid.