² Cohen, Abner, Custom and Politics in Urban Africa. A Study of Hausa Migrants in Yoruba Towns (London, 1969);Google ScholarWilks, Ivor, ‘The Transmission of Islamic Learning in the Western Sudan’, in Goody, J. R. (ed.), Literacy in Traditional Societies (Cambridge, 1968), 162–97;Google Scholar and Meillassoux, Claude (ed.), The Development of Indigenous Trade and Markets in West Africa (London, 1971)Google Scholar, see especially Meillassoux, ‘Introduction’, 49–86;Google ScholarIzard, Michel, ‘Les Yarse et le commerce dans le Yatenga pré-colonial’, 214–227;Google ScholarCurtin, Philip D., ‘Pre-colonial Trading Networks and Traders: The Diakhanké’, 228–38;Google ScholarBoutillier, Jean-Louis, ‘La cité marchande de Bouna dans l'ensemble économique Ouest-Africain pré-colonial’, 240–63;Google Scholar and Amselle, Jean-Loup, ‘Parenté et commerce chez les Kooroko’, 253–63.Google Scholar Also see Lovejoy, Paul E., ‘Long-Distance Trade and Islam: The Case of the Nineteenth Century Hausa Kola Trade’, Journal of the Historical Society of Nigeria, v, 4 (1971), 537–47;Google ScholarAbir, M., ‘Southern Ethiopia’. in Gray, Richard and Birmingham, David (eds.), Pre-Colonial African Trade (London, 1970), 120–37;Google Scholar and Sundstrom, Lars, The Trade of Guinea (Uppsala, 1965), 45–50.Google Scholar

³ Cohen, Abner, ‘Cultural Strategies in the Organization of Trading Diasporas’, in Meillassoux, Trade and Markets in West Africa, 266. The term ‘blueprint’ is Cohen's.Google Scholar

⁴ Wilks, ‘Islamic Learning’, 162–97;Google Scholar and Cohen, Custom and Politics, 161–80.Google Scholar

⁵ Nevertheless, large gaps still exist in our knowledge of some commercial networks, particularly the Jellaba merchants in Wadai, Darfur, and the Nile valley, the Dendi of Borgu, and the Kanuri.Google Scholar

⁶ Lovejoy, ‘Long-Distance Trade and Islam’, 537–7;Google Scholar and Cohen, ‘Cultural Strategies’, 271.Google Scholar

⁷ For a complete discussion of the origins, development, and commercial organization of the Hausa kola trade, see my Ph.D. thesis which is to be presented to the Department of History at the University of Wisconsin (referred to as Lovejoy, forthcoming).Google Scholar

⁸ The material upon which this paper is based was collected during my residence in Kano from July 1969 to June 1970. A Fulbright-Hayes Fellowship made the research possible, and I wish to express my sincere appreciation to the Program for its support. All conclusions, however, are my own and in no way reflect the views of the Program. I also wish to thank Professor Allen Isaacman of the University of Minnesota for his criticism of an earlier draft. I am further indebted to Professor Philip D. Curtin for his guidance in the conduct of my research and his years of inspiration as a teacher.Google Scholar

⁹ The history of the kola trade fits into a pattern which has occurred frequently in the Hausa states. For centuries the Hausa country attracted immigrants from other parts of West Africa. Sometimes individuals fled their homelands under pressure, but more often people settled among the Hausa in order to take advantage of greater opportunities. Their settlement has accounted for numerous, political, social, and economic developments which have profoundly influenced the course of Hausa history. The introduction and early spread of Islam was one example of this impact. Muslim inmtigrants of Mande origin and North Africans connected with the famed Islamic scholar, al-Maghili, formed the nuclei of several early Muslin communities in the Hausa states. Both groups arrived in the last two decades of the fifteenth century and became influential in the establishment of Islam as a court religion. The early introduction of Islam to Kano and other cities was an important precondition for the jihad of Usuman d'an Fodio, for the call to revolution depended upon a foundation of earlier Islamic teaching. For the Muslim immigrants at Kano in the fifteenth century, see Al-Hajj, Muhammad A., ‘A Seventeenth Century Chronicle on the origins and Missionary Activities of the Wangarawa’, Kano Studies, I, 4 (1968), 8–16;Google Scholar and Lovejoy, Paul E. ‘The Mande Impact on Kano: Notes on the Wangarawa Chronicle’, Kano Studies (forthcoming).Google Scholar Emigrant Hausa in turn helped extend Hausa political and social influence beyond the frontiers of the early Hausa states. Such dynasties as the royal family of Yauri and the Kanta family of Kebbi were immigrant Hausa from Katsina. The consolidation of their rule was part of a general expansion of Hausa culture and the Hausa language. A similar process associated with the creation of towns and villages among non-Hausa operated in the region to the south of Zaria. For Yauri and Kebbi, see Adamu, Mahdi, A Hausa Government in Decline: Yawuri in the Nineteenth Century (M.A. thesis, unpublished, Ahmadu Bello University, 1968), 47, 57–8;Google Scholar and Alkali, Muhammad Bello, A Hausa Community in Crisis: Kebbi in the Nineteenth Century (M.A. thesis, unpublished, Ahmadu Bello University, 1969), Chapter I.Google Scholar

¹⁰ Bornu was particularly important as the home of people who emigrated to the Hausa country. Some did so for commercial purposes, particularly in connexion with the trade in potash. Others moved for political reasons. One Kanuri group arrived in Kano sometime in the fifteenth century. It may have been an administrative or diplomatic corps which was part of Bornu expansion in the west, but, more likely, it was an exiled force which fled in the wake of a political crisis. No matter what the reasons for the emigration, the group was influential at Kano and may have been responsible for the introduction of such Bornu titles as Galadima and Chiroma into the Kano administrative structure. Other Bornu settlers established themselves elsewhere, although most were not as important as the Kano immigrants. In all cases the number of settlers was small. Many undoubtedly were unimportant economically or politically and only added to the population of local communities. Nevertheless, they represented a feature of Hausa history which helps explain why the region between Bornu and the Niger developed into an area of relatively homogeneous culture and dense population. All the groups were easily and rapidly assimilated into Hausa society, and in many cases the immigrants went on to contribute to the expansion of the economy and polities centered in the Hausa-speaking region. The head of the Bornu immigrants at Kano in the fifteenth century was Dagaci, who came with ‘many men and mallams’; see ‘Kano Chronicle’ in Palmer, H. R., Sudanese Memoirs (Lagos, 1908), III, 109–10.Google Scholar Other examples of Bornu immigrants in the Hausa country include an immigrant community at Dallol Maouri (Chatelain, M., ‘Traditions relatives à l'établissement des Bournouans dans le Dallol Maouri et le pays Djerma’, Bulletin du comité d'études historiques et scientifiques de l'Afrique Occidentale Française [1917], 358, 361)Google Scholar, and the Burmawa of Sokoto claim Kanuri origins (Johnston, H. A. S., A Selection of Hausa Stories [London, 1966], 113–14).Google Scholar

¹¹ A distinction should be made between the residents of the towns along the trade routes and the professional traders from the Sokoto Caliphate. See Lovejoy, ‘Long-Distance Trade and Islam’, 537–47.Google Scholar

¹² The oral data from which most of this paper is drawn are on deposit at the Centre for Nigerian Languages, Abdullahi Bayero College, Kano, and with the Northern Nigerian Research Scheme, Ahmadu Bello University, Zaria. The principal accounts of Kambari history are the testimonies of Malam Basharu na Mahu, aged 90, interviewed at Gummi on 7 Mar., 1970 (tapes Nos. 20 and 21); Malam Umaru Muss, aged 76, interviewed at Gummi on 8 Mar., 1970 (tape No. 21); Alhaji Audu Ba'are, aged 100, interviewed at Kano on 1 Jan., 1970 and 18 Jan., 1970 tapes Nos. 9 and 12); Alhaji Bak'o Madigawa, aged 82, interviewed at Kalgo on 1 Dec. and 7 Dec., 1969 and 17 Jan and 18 Jan., 1970 (tapes Nos. 6 and 12); Malarn Muhammadu D'an Turmi, aged 60, interviewed at Sokoto on 22 Feb., 1970 (tape No. 18); Malam Muhammadu and Malama Hauwa, aged 54 and over 80 respectively, interviewed at Kalgo on 27 Feb., 1970 (tapes Nos. 59 and 20); Malam Muhammadu D'an Amarya, aged 78, interviewed at Gummi on 5 Mar., 1970 (tape No. 18); Malam Sa'idu, aged 78, interviewed at Sokoto on 22 Feb., 1970 (tape No. 18); and Marasa Ulu, aged 55, interviewed at Sokoto on 22 Feb., 1970 (tape No. 18). All ages are in Muslim years.Google Scholar

¹³ List of Mais of Bornu down to A.D. 1808 with approximate dates, in Palmer, H. R., Sudanese Memoirs, III, 45.Google Scholar For the Kambari traditions, see the accounts of Malam Basharu na Mahu (tapes Nos. 20 and 21); and Malam Umaru Musa (tape No. 21); Alhaji Bak'o Madigawa (tape No. 6). Also see Lucas, S., ‘Mr Lucas' Communications’, Proceedings of the Association for Promoting the Discovery of the Interior Parts of Africa (Hallet, Robin, ed.), (London, 1967), I, 153.Google Scholar I wish to thank Mr John Layers of Abdullahi Bayero College Kano, for this and several other references.

¹⁴ Malam Basharu na Mahu stated that D'an Toga had 200 slaves (tape No. 20). He accounted for the movement of the immigrants into Gummi because of the danger of attack from a nearby town, Birnin Tudu. Malam Musa Umaru stated that Warn wanted to build up Gummi because the town was so small (tape No. 21). Malarn Muhammadu D'an Amarya also stated that Warn wanted to increase the population of Gummi town. D'an Amarya said that D'an Toga had ‘many people’ with him, and they stayed at Kali for three years (tape No. 20). Also see the account of Lamu, Malam Audu, aged 42, interviewed at Gummi on 5 Mar. 1970 (tape No. 20).Google Scholar

¹⁵ For the history of Gummi, see Harris, T. G., Sokoto Prornncial Gazetteer (Kaduna, typescript, 1938), 129, 142, 314, 316, 318–19.Google ScholarAlso see H. S. Edwardes, Assessment Report, Gummi District, Sokoto Province, Nigerian National Archives, Kaduna (NAK), SNP 10/5, 542p/1917;Google ScholarKrieger, Kurt, Geschiehte von Zamfara, Sokoto-Provinz, Nordnigeria (Berlin, 1959), 92–3; and personal communication from Malam Muhammad Bello Alkali, 23 Feb., 1970.Google Scholar

¹⁶ Malem Bashru na Mahu and Malam Umaru Musa (tapes Nos. 20 and 21). Also see Alkali, Kebbi, 139;Google ScholarMischlich, A. and Lippert, J., ‘Beitrage zur Geschichte der Haussastaaten’, Mittheilungen des Seminars für Orientalische Sprachen, VI (1903), 210–11;Google ScholarLast, Murray, The Sokoto Caliphate (London, 1967), 14–15.Google Scholar

¹⁷ Ribats were built along the Kebbi frontier in 1821, and Jega was one of these. This policy stabilized the region. Sokoto was initially the site of Muhanirnad Bello's camp and only later became a capital. For the events of the jihad, see Alkali, Kebbi, 156–68, 169, 171–3;Google ScholarLast, Sokoto Caliphate, 23–40;Google Scholar and Johnston, H. A. S., The Fulani Empire of Sokoto (London, 1967), 47–59.Google Scholar

¹⁸ See Lovejoy (forthcoming).Google Scholar

¹⁹ Aihaji Audu Ba'are (tape No. 9);Google Scholar Alhaji Bak'o Madigawa (tape No. 6). Also see Ilorin, Sule, Tarihin Goro (Zaria, 1958), 2; and Lovejoy (forthcoming).Google Scholar

²⁰ See Chapters 2 and 3 of Lovejoy (forthcoming).Google Scholar

²¹ Alhaji Audu Ba'are (tape No. 12); Alhaji Bak'o Madigawa (tape No. 6); and Malam Musa Umaru (tape No. 21).Google Scholar

²² Barth, Heinrich, Collection of Vocabularies of Central African Languages (Gotha, 1862–1866), 240–1;Google ScholarBenton, P. A., The Languages and Peoples of Bornu (London, 1968), II, 98.Google ScholarLukas, Johannes, A Study of the Kanuri Language (London, 1937), 227.Google Scholar The Yoruba and Bariba of Borgu call the Hausa gambari, probably because the first Hausa they knew were traders. A relationship between the Kanuri term kombali and the Yoruba gambari seems possible. For the word gambari, see Gunn, and Conant, , Peoples of the Middle Niger Region, Northern Nigeria, 21 (London, 1960);Google ScholarMarty, Paul, Études sur l'Islam au Dahomey, Le Bas Dahomey—Le Haut Dahomey (Paris, 1926), 179.Google Scholar It should also be noted that Gambarawa is a ward in Katsina City and appears to have been a centre of foreign trade since before the jihad. The significance of the ward's name has yet to be explored.

²³ The Kambarin Beriberi were not related to the non-Muslim Kambari people who lived between Kontagora and Yauri to the south of the Zamfara River Valley. The Kambarin Beriberi point to their relatively recent arrival at Gummi, the incorporation of ‘Beriberi’ in their asali, and their facial markings as proof of the distinction. Alhaji Audu Ba'are (tape No. 12) tried to explain the similarity in names through the assertion that his ancestors and the Kambari people of Yauri were both of Bornu origin but that the Yauri Kambari abandoned Islam and such alleged attributes of Muslim civilization as clothing when they settled to the south of the Zarnfara River. This explanation cannot be accepted as anything more than an attempt to rationalize the existence of similar names. For a discussion of the Kambari of Yauri and Kontagora, see Gunn, Harold D. and Conant, F. P., Middle Niger Region, 21–9.Google ScholarThe authors note ‘The exact relationship between the names Kambari and “Kambari(n) Beriberi,” and that between the peoples so termed, is not known precisely. The latter term is generally applied to peoples widely scattered, but generally in the Niger and Benue Valleys, who emigrated from Bornu somewhat over a century ago….’ (p. 21). Gunn and Conant further note that another group of people in Lafia Emirate, in Benue Plateau State 50 miles north of the Benue, called themselves Kambari Beriberi. There does not seem to be any connexion with the Kambarin Beriberi of Gummi (p. 113).Google ScholarTemple, O. and Temple, C. L. (eds.), Notes of the Tribes, Provinces, Emirates, and States of the Northern Provinces of Nigeria (London, 1922, 1965), 199, 218Google Scholar, also accept the distinction between the Kambari of Yauri and Kontagora and the Kambarin Beriberi. One of these accounts is based on a tradition collected by E. J. Arnett. Also see Adamu, Mahdi, Yawuri in the Nineteenth Century, 31–5;Google Scholar and Matthews, A. B., Anthropological and Historical Report on the Kamberri NAK, SNP, 17/8, K. 2105.Google Scholar Matthews also accepted the distinction between the Kambarin Beriberi of Gummi and the Kambari of Yauri. For an account of the Kambari of Yauri in 1830, see Richard, and Lander, John, Journal of an Expedition to Explore the Course and Termination of the Niger (New York, 1858), I, 299.Google Scholar Another tradition collected earlier in the century claimed that ‘kambari’ was an archaic Hausa word meaning ‘partly’ or ‘half’, so that the asali Kambarin Beriberi indicated people who were partly of Beriberi origin. The tradition is from Palmer, H. R., The Bornu Sahara and Sudan (London, 1936), 108–10Google Scholar, and it may represent nothing more than his own speculations:‘There is another class of people in the regions which extend as far west as Sokoto, called Tozer, or Tozamawa, who further east are called Tazar,… who originally worked the salt, called mangul…. Later, when they began to spread west to Katsina, their chief was a Bornu [man]…. During the time when the Katsina kingdom was powerful…the Tazar spread greatly and became the chief agents in the trade to Gwanja (Gold Coast) [sic]. As they settled down and intermarried with the Katsinawa and other Hausa peoples they acquired the name Kam-bari, i.e. persons of partly Bari-Bari (Kanuri) descent. Trade to the west was so entirely in their hands that Kambari (Gambari) [sic] in the west caine to mean ‘Hausa”.’The possible connexion with mangul salt miners bears further investigation, but many other of Palmer's conclusions and suggestions cannot be considered seriously. Palmer's dating for the movement of the Gunimi immigrants is too early, while the Kambarin Beriberi were never so numerous for their name to be adopted by the Bariba and Yoruba as a general term for all Hausa. Nor did the Kambarin Beriberi play such a major early role in the development of trade to Gonja and Asante.

²⁴ For Gamji, see the accounts of Alhaji Bak'o Madigawa (tapes Nos. 6 and 12); Malam Audu Lamu (tape No. 20); and Malam Umary Musa (tape No. 21). On Hungumi, see Malam Basharu na Mahu (tape No. 21); Marafa Ulu, Hakimin Hungumawa, who is Fulani, claimed Hungumi meant the “man from Gummi” (tape No. 18). Malam Muhammadu D'an Turmi (tape No. 18) claimed Hungumi was a madugu, hence the possibility that Garnji and Hungumi were the same person.Google Scholar

²⁵ Accounts of Alhaji Audu Ba'are (tapes Nos. 9 and 12); Malam Bawa, aged 70, interviewed at Kano on 21 Jan., 1970 (tape No. 12) and Mai Unguwar Mararraba Ibrahim (tapes Nos. 9 and 12). For the Kambarin Beriberi in Sabon Sara ward, see account of Alhaji Garba, aged 67, interviewed at Kano on 21 Jan., 1970 (tape No. 13). The Kambarin Beriberi in Sabon Sara ward belonged to the family of Malam Bawa, but Alhaji Garba was unable to provide much information on the family. Alhaji Bak'o Madigawa (tape No. 12); and also Alhaju Audu Ba'are (tape No. 12) could not explain the way (hanya) the Sabon Sara Kambari came to Kano, except that all Kambarin Beriberi ultimately came from Gummi. For the Katsina Kambari, see the account of Malama Zainabu, aged about 80, and who was born in Katsina, interviewed at Katsina on 3 Feb., 1970 (tape No. 16). Also see the accounts of Malama Mairamu and Malam Sule (tape No. 16), interviewed at Katsina on 5 Feb., 1970. Malama Mairamu was born around 1882 in Bornu while her father was on an expedition. Her father and uncle were the first to settle in Katsina. Her grandfather, Auta Mai Gari, is remembered as being an important trader who lived at Gurnmi, and who sent her father to live in Katsina. Maama Mairamu also knew the name of her great-grandfather. It was Usman, and he lived at Gummi as well. When they came to Katsina, they brought a letter from the sultan of Sokoto instructing the Sarkin Katsina to build them a house. Also see the account of the elderly Malama A'isa, whose father and grandfather lived in Katsina. She was interviewed at Katsina on Jan., 1970 (tape No. 11). Malam Muhammad Maikaka, aged 75, recounted the story of another Kambari family at Katsina. The family had come from Gurnmi as well (tape No. 11). For Zaria, see the accounts of Alhaji isa Madigawa (tape No. 9); Alhaji Bak'o Madigawa (tape No. 12).Google Scholar

²⁶ The main accounts for the Madigawa Kambari are those of Alhaji Bak'o Madigawa, Alhaji Ma'azu, and Alhaji Sule (tapes Nos. 6 and 12). Also see Alhaji Isa Madigawa (tape No. 9).Google Scholar

²⁷ For a complete dicussion of masuganci, see Lovejoy (forthcoming).Google Scholar

²⁸ Alhaji Bak'o Madigawa (tapes Nos. 6 and 12); Malam Audu Lamu (tape No. 20); and Malam Umaru Musa (tape No. 21).Google Scholar

²⁹ See account of Malam Muhammadu na Kalgo, whose father married Mijinyawa's daughter. The chief sources, however, were Malam Muharnmadu Mijinyawa and his mother, Malama Hauwa (tapes Nos. 19 and 20). Malama Hauwa was married to one of Mijinyawa's sons and knew the madugu's widows. She is related to the Kambari of Macligawa ward in Kano, and she stated that at least two of Mijinyawa's sons married daughters of the Madigawa Kambari. For the possible connexion between Mijinyawa and D'an Toga, see the manuscripts concerning the destruction of Mijinyawa's caravan in Borgu, in Heepe, M., ‘Gottlob Adolf Krauses Haussa-Handschriften in der Preussischen Staatsbibliothek, Berlin’, Mitteilungen des Seminars für Orientalisthen Sprachen, XXXI, 1928, xliii–xliv.Google Scholar The manuscript begins ‘Mijinyawa baban d'an D'an Toga ne’, or ‘Mijinyawa was the big (important) son of D'an Toga’. This implies a direct genealogical connexion with D'an Toga, although Mijinyawa could not have been a son of the Kambari founder. Since Mijinyawa was a contemporary of Madugu Umaru Sule of Kano, Mijinyawa could have been the great-grandson of D'an Toga. For another account of the death of Mijinyawa, see ‘Labarin Madugu Mijinyawa-maiakokari, da Madugai Duka’, in Edgar, Frank (ed.), Litafi na Tatsuniyoyi na Hausa (Lagos, 1924), 1, 239–41.Google Scholar A translation of this account is being published by Skinner, Neil in Hausa Tales and Traditions (London, forthcoming).Google Scholar

³⁰ All these people were in close personal as well as commercial contact. Three examples demonstrate this. Both Alhaji Gago, aged 72, and his father were born in Hungumawa ward at Sokoto, but Gago's grandfather was born in Madigawa ward in Kano (interviewed in Sokoto On 22 Feb., 1970, tape No. 18). Malam Basharu na Mahu of Gummi was born in Sokoto in Hungumawa ward, and his family stayed with the Kambari of Mararraba ward in Kano when they travelled to that city (tapes Nos. 20 and 21). Malam Audu Lamu, aged 42, was born in Gummi along with his father, but the family had a house in Mararraba ward in Kano (tape No. 20). Marriage ties also linked the various communities together. Audu Lamu was related to Madugu Umaru Sule of Madigawa ward, since the madugu was his mother's grandfather. These kinds of marriage connexions could be cited indefinitely. For Mijinyawa's donkeys, see Krause manuscript, xliii.Google Scholar

³¹ Malam Umaru Musa noted that the Kambari even wore alkyabba, a type of expensive gown which only the aristicracy could afford (tape No. 21).Google Scholar

³² Alhaji Bak'o Madigawa (tape No. 6) and Malam Audu Lamu (tape No. 20).Google Scholar

³³ The average was five lorries per Kambari owner, a figure which was higher than the average for any other group. The breakdown on lorry ownership for the Kambarin Beriberi was as follows: two individuals owned one lorry each, three individuals owned two lorries each, one individual owned three lorries, two individuals owned five lorries each, two individuals owned six lorries each, one individual owned ten lorries, one individual owned sixteen lorries. The only other group where close to half its members owned at least one lorry was the Agalawa. Twenty-six of fifty-three Agalawa owned a total of 105 lorries, an average of four each. These figures revise the preliminary findings published in my article, ‘The Wholesale Kola Trade of Kano’, African Urban Notes, v, 2 (1970), 139. The statistics are based on a survey of 196 kola wholesalers at Ujli, the wholesale kola market at Kano, during Mar. and Apr. 1970.Google Scholar