When combined, evidence from oral tradition, Arabic texts and archaeological sites indicates that ancient Mali's seat of government changed more than once during its imperial period from the twelfth to the sixteenth century. According to oral tradition, the town in which Sunjata spent his early years, and to which he returned from exile, was Dakajalan. This mansadugu or ‘king's town’ served as Sunjata's base of operations for his campaign against Sumanguru and may have continued for a time as both spiritual and military headquarters during the struggle for unification following the defeat of Soso. As Mande's core territory expanded into the beginnings of empire, the mansadugu was probably moved north-eastward, down the River Niger to take advantage of widening commercial opportunities and to govern an expanded population of imperial subjects who included large numbers of Muslims from the former Soninke territories. Niani was one of the oldest and most important cities of Mali, especially notable for its iron industry. If it served also as a political capital this would most likely have been in the sixteenth century under Niani Mansa Mamadu, a descendant of Sunjata's royal lineage.

The killing in May 1905 of Harry St. George Gait, a senior official of the Uganda Protectorate, has generally been treated in the literature as a political murder mystery. It can more usefully be seen as a window on two issues: the importance of clientship in relationships between agriculturalists and pastoralists in the kingdom of Ankole, and British reliance on pastoral allies to make real their power in Ankole. This paper suggests that Gait was killed by an agriculuralist frustrated by his own failure to advance in Ankole society; but that the repercussions of the killing were magnified by the fears and uncertainties of British officials on the spot over the reliability of their pastoralist allies. The British were, however, unable to dispense with these allies, and the crisis generated by Gait's death was resolved by a reaffirmation of the alliance between the British and the pastoralist elite, after the effective scapegoating of two minor chiefs.

The significance of dress in mediating social relations was deeply rooted in the Central African experience. In pre-colonial times, clothing, jewellery and insignia conveyed identity, status, values and a sense of occasion. Those with access to European trade cloth and second-hand clothes integrated them into their dress. Central Africans had a strong sense of the “politics of costume” long before new sources and ideas of clothing arrived with colonialism.

Brazzaville, the capital of French Equatorial Africa, then became the scene of opportunity, experimentation and choice. Foreign workers from West Africa, the French Antilles and the Central African coastal regions pioneered new styles which were quickly appropriated and adapted by other townspeople. Europeans, in their attire, also seemed to confirm the importance of dress and were a model for those who considered themselves évolués. In handing out clothing, European employers and missionaries had their own agenda, which was rejected by many townspeople as an autonomous fashion sense developed in Bacongo and Poto-Poto, the African districts of Brazzaville. With an entrenched monetary economy, cloth and clothing became widely available to all with cash. Styles, costs and values became issues of contention. Clothing not only symbolized change but became a vehicle for change.

In the late colonial period, the sources allow a deeper understanding of the relationship of dress to controversial social issues. Clothing became an arena for contesting and asserting class, gender and generational roles.

The professional Nigerian nationalist historiography which emerged in reaction against the imperialist Hamitic Hypothesis – the assertion that Africa's history had been made only by foreigners – is rooted in a complex West African tradition of critical dialogue with European ideas. From the mid-nineteenth century, western-educated Africans have re-worked European ideas into distinctive Hamitic Hypotheses suited to their colonial location. This account developed within the constraints set by changing European and African-American ideas about West African origins and the evolving character of the Nigerian intelligentsia. West Africans first identified themselves not as victims of Hamitic invasion but as the degenerate heirs of classical civilizations, to establish their potential to create a modern, Christian society. At the turn of the century various authors argued for past development within West Africa rather than mere degeneration. Edward Blyden appropriated African-American thought to posit a distinct racial history. Samuel Johnson elaborated on Yoruba traditions of a golden age. Inter-war writers such as J. O. Lucas and Ladipo Solanke built on both arguments, but as race science declined they again invoked universal historical patterns. Facing the arrival of Nigeria as a nation-state, later writers such as S. O. Biobaku developed these ideas to argue that Hamitic invasions had created Nigeria's proto-national culture. In the heightened identity politics of the 1950s, local historians adopted Hamites to compete for historical primacy among Nigerian communities. The Hamitic Hypothesis declined in post-colonial conditions, in part because the concern to define ultimate identities along a colonial axis was displaced by the need to understand identity politics within the Nigerian sphere. The Nigerian Hamitic Hypothesis had a complex career, promoting élite ambitions, Christian identities, Nigerian nationalism and communal rivalries. New treatments of African colonial historiography – and intellectual history – must incorporate the complexities illus-trated here.

The article explores a neglected aspect of West African history, namely the historiography of ‘stateless’ peoples in north-western Ghana. At the same time, it is a contribution to the recent debate on the role of history in the construction of new ‘tribal’ identities in Africa in the colonial and post-colonial periods. After discussing local oral patrician accounts of migration and settlement, and the historical imagination of colonial officers, I analyse histories of tribal origins written recently by Dagara intellectuals, which draw upon hypotheses, evidence, tropes and narrative models from both colonial and indigenous sources. The concern is not with a conventional history of ideas, but with showing how authors with new requirements and interests can fuse disparate elements into new accounts of the past whose underlying intentions differ considerably from those of their sources. More specifically, the article discusses the claims of intellectuals' and villagers' stories with reference to different underlying political agendas, suggesting that they constitute different historiographic genres appealing to different audiences. The former are basically concerned with establishing the Dagara as a political community within a modern state, the latter with providing charters for local boundaries and rights.

Despite his important political and literary activities, Blatta Gäbrä Egzi'-abehēr is almost unknown to scholars of Menilek's Ethiopia. This historical period is not particularly well researched, and the author stands out as one of the few Ethiopian intellectuals to have written such an important number of literary works focused on nationalistic and anti-Italian feelings. The Amharic/Ge'ez text under discussion, his letter to Menilek written in 1899, is a remarkable document from this point of view because it reveals a strong opposition to colonialism and the Italian occupation of Eritrea. This document is one of the first Ethiopian sources to testify to the growing nationalism and the growth of concepts of unity and independence. It allows us to consider more carefully the beginning of an Ethiopian secular ideology of the modern state. And such an ideology must be placed in the colonial context. The letter to Menilek raises some important questions regarding the new source material in the late nineteenth century available to historians of modern Ethiopia. A translation of the text is given as well as a comment on its historical significance.