Radiocarbon dates published by southern African archaeologists since the last review in this journal (in 1980) illustrate an increasingly complex record of population movement and interaction. Substantial gaps in the distributions of dates reflect the ebb and flow of people in response to changing environmental and social circumstances. More interesting perhaps is the range of intergroup relations now emerging from the last two millennia with the appearance of pastoralists and agriculturalists. Radiocarbon dates, taken along with spatial and formal patterns in the archaeological record, show clearly that the distributions between hunters, herders and farmers and between Stone Age and Iron Age communities were blurred and flexible.

This article is a partial answer to M. D. D. Newitt, who proposed that settled Maravi states were established only as a result of the rise of Muzura in the first half of the seventeenth century (cf. J. Afr. Hist., 1982, ii). Newitt thereby challenged the more orthodox view that a formal Maravi state system existed already by the middle of the sixteenth century, if not earlier. It is argued here that the orthodox view is still valid in the case of the Lundu state in the lower Shire valley, and perhaps also in the case of some of the neighbouring states. It is shown that around 1590 the then Lundu incumbent embarked on a course of strong state centralisation during which he appropriated the power of the traditional rain priests and thus became both the secular and the ritual leader of the country. It is also argued that this unusual degree of centralisation was achieved and could for a time be maintained with the help of the Zimba, an army of fugitives from the south bank of the Zambezi. However, the present article challenges Malawian historiographical orthodoxy on a very different point, by maintaining that Muzura is not to be identified with the Kalonga dynasty on the south-western shores of Lake Malawi, but with a separate state system in the western Shire Highlands, which gained prominence well before the Kalongas came to the fore.

The results of recent archaeological research in the Upper Nile basin are summarized and placed within the context of the anthropological-historical debate concerning the origins of the Nuer, Dinka and Atuot as distinct ethnic groupings. The archaeological evidence demonstrates a considerable antiquity for cattle-keeping in the region, the existence of what appears to be a very widespread cultural tradition in the late first millennium a.d. characterized by a distinctive form of burial, and a hiatus in settlement in the area east of Rumbek early in the present millennium, possibly around the time when humped cattle were introduced further north. The implications of these data for the explanation of the origins of the Luo migrations are discussed.

Between 1772 and 1780 Richard Miles, a servant of the Company of Merchants, kept detailed records of some 1,308 barters for slaves he made on the Gold Coast, along with a much smaller number of barters for gold and ivory. The lists provide indirect information about the Fante brokers with whom he dealt, and how they conducted their trade. The names that appear show that the Fante dealers at the waterside were numbered in the hundreds, and indicate that many of them operated on a small scale or sold slaves to supplement other forms of income. The fact that many of the élite, though friendly to Miles, preferred not to deal with him indicates a fear amongst them of becoming over-dependent on the Company of Merchants.

The lists, with their daily records of prices and price changes in trade ounces in all of the goods in Miles's assortments, illustrate how the Fante dealt with rapid changes in supply and demand by price alterations, by manipulating the content of assortments and by changing the value of the ounce of trade as against the ounce of gold.

During the first quarter of the eighteenth century, European merchants bought more slaves in the Bight of Benin than on any other part of the West African coast. From c. 1720 until 1727 much of their buying was concentrated in Savi, the capital of a small Aja state called Whydah. When the Dahomeans overran Savi in 1727 they stopped the inland slave suppliers from travelling to the coast, prevented the local Hueda from going inland to collect slaves, and insisted that the Europeans bought slaves only from Dahomean dealers. In an attempt to make sure that the Europeans had nothing more to do with their former trading partners the Dahomeans burned the factories in Savi and forced their European occupants to retire to Grehue, Savi's port, a spot on the coast where the Europeans maintained a number of fortified warehouses.

The middleman policy did not at first operate satisfactorily. There were two reasons for this. The first was that the Dahomeans were, in practice, unable to prevent the Europeans from continuing to trade with the Hueda. The second was that the inland suppliers refused to sell slaves to Savi's conquerors. The Dahomeans solved their ‘coastal’ problem in the 1740S by placing a garrison in Grehue. This garrison kept the exiled Hueda at bay and held the Europeans in what amounted to open captivity. The Dahomeans were never able completely to solve their ‘supply’ problem. In the 1730s and 1740S the inland merchants took their slaves to ports which opened up on the Bight to the east of Grehue. Only in the 1750s and 1760s did they channel substantial numbers of slaves through Dahomey. In the last decades of the century they again boycotted the Dahomean market. Dahomey therefore prospered as a middleman state only between c. 1748 and c. 1770.

An examination of their eighteenth century trading suggests that the Dahomeans were a slave-raiding community whose members realised in 1727 that they would soon run out of fresh raiding grounds. They appear to have introduced their middleman policy in an attempt to ensure that they would continue to profit from slave trading even after they had ceased to be able to take large numbers of captives themselves. Although the policy was by no means a complete success, it was important in that it seems to have led the Dahomeans to begin placing garrisons in the territories they ravaged. It appears, in fact, to have been the pursuit of their middleman goals that led them to begin creating the often described nineteenth century ‘greater’ Dahomean state. The middleman programme ceased to be of much importance after c. 1818, when the fall of Oyo enabled the Dahomeans to resume raiding widely in unexploited territory.

This article seeks to illuminate some important themes in nineteenth-century South African history by examining the fate of a small chiefdom of difaqane refugees who settled in 1833 just north of the Orange River under the patronage of a French Protestant missionary. It situates the history of the Tlhaping of Bethulie against the background of the expansion of white settlement north of the Orange River and the development of colonial capitalism in the larger region. The processes of white state formation north of the Orange in the middle years of the century, especially the seminal role of British intervention during the period of the Orange River Sovereignty, are examined. The corresponding rise of white elites and the varied primitive forms of capital accumulation employed by the emerging Boer notables are investigated. The article then seeks to provide a concrete study of these themes in a local setting. These encircling developments provided the context for the rising tensions and conflicts at the Bethulie mission station and in the Tlhaping community in the 1850s. The gradual alienation of the Bethulie lands to private ownership eventually led to the destruction of the territory and the break-up of the community. These processes are examined within the context of the rise of local Boer notables and the nature of state structures in the Orange Free State Republic established in 1854. In the end those who orchestrated and benefited from the dismemberment of the Bethulie territory were those who controlled the instruments of patronage and power in the local Boer state.

The Xhosa cattle-killing movement of 1856–7 cannot be explained as a superstitious ‘pagan reaction’to the intrusion of colonial rule and Christian civilization. It owes its peculiar form to the lungsickness epidemic of 1854, which carried off over 100,000 Xhosa cattle. The Xhosa theory of disease indicated that the sick cattle had been contaminated by the witchcraft practices of the people, and that these tainted cattle would have to be slaughtered lest they infect the pure new cattle which were about to rise.

The idea of the resurrection of the dead was partly due to the Xhosa belief that the dead do not really die or depart from the world of the living, and partly to the Xhosa myth of creation, which held that all life originated in a certain cavern in the ground which might yet again pour forth its blessings on the earth. Christian doctrines, transmitted through the prophets Nxele and Mhlakaza, supplemented and elaborated these indigenous Xhosa beliefs. The Xhosa and the Christian elements united together in the person of the expected redeemer Sifuba-sibanzi (the broad-chested one). The central beliefs of the Xhosa cattle-killing were neither irrational nor atavistic. Ironically, it was probably because they were so rational and so appropriate that they ultimately proved to be so deadly.

This article reassesses Sir William Mackinnon's role in the evolution of Victorian imperialism in Eastern Africa. It rejects the view that Mackinnon's activities in Eastern Africa were motivated by a desire for self-glorification and attempts, by contrast, to demonstrate the relevance of business considerations. A search for shipping subsidies and railway guarantees, spreading out from British India, accompanied the Mackinnon Group's development of steamshipping and mercantile interests in Africa, in support of investments in the Persian Gulf and western India. Promotion of these interests drew Mackinnon into schemes to lease the Sultan of Zanzibar's mainland territories and to consolidate British rule in the Transvaal by the construction of a railway from Delagoa Bay. During the 1880s the Group's shipping and commercial operations were threatened by the rise of foreign competition. Behind the formation of the Imperial British East Africa Company lay the hopes of Mackinnon and his business associates that public funds could be attracted to the defence of the Group's interests in Eastern Africa and to the reconstruction of its shipping services in the western Indian Ocean.

The European goods which Africans consumed in the slave trade era tell us much about the African societies which imported them. However the study of the subject has involved much confusion through the application of fragmentary evidence from different societies in different stages of development towards the fashioning of broad hypotheses about the impact of the trade on West Africa as a whole. It is important therefore, when the evidence is available, to study each society and each group of African middlemen individually as well as within the wider context.

The papers (especially the barter records) of Richard Miles throw a good deal of light on one such microcosm: the Akan people of the Gold Coast in the second half of the eighteenth century. The Fante middlemen with whom Miles dealt required, for virtually every barter, an assortment of goods from five major categories: hardware, currencies, textiles, luxury items, arms and ammunition. Though all these categories were necessary for the trade, it is notable that textiles were far and away the dominant commodity desired by the Akan. Guns were in surprisingly low demand during this period which suggests that the Akan slave producers (principally the Asante) had no difficulty raising slaves through tribute in peacetime and were not forced to rely on wars and slave-raids.

Miles's documents also make it clear that generalizations drawn from the Gold Coast in this period cannot be extended automatically to other areas; Akan history tells us that neither can they be extended on the Gold Coast into a different era.

Existing works on the colonial history of Northern Nigeria are generally agreed that the emirs who reigned during the colonial era were selected by traditional methods, that is to say, by kingmakers. This article attempts to show that in the case of Sokoto Province the emirs who were appointed during the period 1903–30, though they had traditional claims to their position, were chosen by the British and not by the kingmakers. It is suggested that during this period the British were so pre-occupied with the security of their rule that they would not leave the important function of selecting emirs to the kingmakers whom they still suspected could select anti-British princes as emirs. It is argued that this policy was largely dictated by the Administration's fear of Mahdism which, up to the end of the 1920s was seen as a real danger to British rule. Thus only overtly loyal princes were elevated to emirships, regardless of whether they had the kingmakers' support or not. The British were able to do this without causing serious political unrest because the emirates were basically ‘competitive monarchies’ which left the British room for manipulation. Finally, the article suggests that, as a result of increased confidence in the security of their rule and owing to the fact that unpopular chiefs had proved to be a liability to the government, in the early 1930s the British restored the kingmakers' right to elect emirs without overdue interference by administrative officers.

In the nineteenth century, the entrenched power of three Cabindan families, Nsambo, Npuna and Nkata Kolombo, was challenged by the rise of the Franques. The dominant figure, Francisco Franque, amassed wealth through a close alliance with Brazilian slave traders and through freighting goods and passengers in ‘coasting’ vessels which were locally built. At the same time he invested in a large household and attracted to his village dependents who provided labour and armed support for the expansion of his territorial base. Beyond the village, Franque, like other ‘big men’ at Cabinda, depended on an alliance with kinsmen for the defence of family interests. In the last quarter of the century, the Franques were weakened by the end of the slave trade, by disputes over inheritance rights, following the death of Francisco Franque, by the challenge of Manuel José Puna and by the emigration of junior family members in search of employment in the colonial economy of Angola and neighbouring territories. After 1885, under Portuguese colonial rule, the household was no longer a principal unit of production and family cohesion was no longer relevant. European settlers and companies moved into prime land and the emigration of workers, including women, intensified in the face of deteriorating economic conditions. Some individuals continued to have access to privilege, as far as that was possible in Angolan colonial society, through education. At Cabinda, the Portuguese authorities gave at least nominal recognition to some senior family members, for example at official celebrations. The name of the old families lived on through prominent individuals although their collective power and influence had been drastically undermined.

Traditionally, historians have viewed Queen Ranavalona I as being responsible for inaugurating an autarkic policy in Madagascar. Her expulsion of most foreigners from the country in 1835 is seen primarily as a reflection of her conservative and xenophobic attitudes. In this she is contrasted with her predecessor, Radama I, who is viewed as an enlightened and progressive monarch who, through wise domestic policies and an alliance with the British on Mauritius from 1817, built up an economically sound and prosperous empire. This paper challenges the traditional interpretation, arguing that in fact the Merina economy was in a dire condition from the second decade of the nineteenth century because the slave exports upon which it heavily depended were severely restricted in consequence of the British takeover of the Mascarenes. The subsequent alliance between Britain and Imerina totally prohibited slave exports. However, Radama I looked to Mauritius and British aid to promote legitimate exports and to help impose Merina rule over all Madagascar. Autarkic policies were initiated by Radama I in 1825–6 as a reaction against the failure of the British alliance to produce the anticipated results, and against the free trade imperialism that accompanied it. Convinced by 1825 that the Mauritius government meant to subordinate Imerina both economically and politically to British imperial interests, he reneged on the British treaty and adopted a policy designed to promote rapid economic growth within an independent island empire. Ranavalona I, far from implementing irrational and xenophobic policies, extended her predecessor's autarkic policies in a rational and systematic manner, and for precisely the same ends.

This article examines the transition from the slave trade to ‘legitimate commerce’ in Portuguese Guinea between 1840 and 1880. Peanuts became the principal export crop. They were cultivated on plantation-like establishments called feitorias located primarily along the banks of the Rio Grande and on Bolama Island. From the 1840s through the 1870s, Luso-African, other Euro-African and European traders built these feitorias. These traders depended upon both slave and contract labour to cultivate their export crop.

Although Portugal claimed Portuguese Guinea, French trading houses dominated ‘legitimate commerce’ in this West African enclave. The demand for increased peanut production came from the burgeoning French oil mills rather than from Portuguese industries. French merchants supplied the ships needed to transport the crop as well as many of the imported goods sold locally. By the 1870s the Portuguese realized they needed to break this French monopoly. By that time Europe was suffering from an economic recession, peanut prices were falling and cheaper oilseeds from India and America were entering the market. Portugal's attempts to establish commercial dominance met with little success.

The economic crisis of the 1870s not only created difficulties for feitoria owners and their workers, but also for Fulbe groups in the process of expansion. These Fulbe wanted to establish political control in order to reap the economic benefits the peanut trade offered — especially access to firearms and in turn, slaves. As peanut production fell from 1879 onward, Fulbe groups began fighting amongst themselves for control of shrinking resources. By 1887, the feitoria system and this phase of peanut production had ended. The Portuguese, like the Fulbe, had to look for new ways to survive economically.

The notion of capitalist relations in Ghanaian cocoa-farming is familiar, yet their development has been relatively little studied. In Amansie district, Asante, capitalist relations of production developed as a result rather than as a cause of the cocoa ‘take-off’, c. 1900–16. This paper examines their emergence, which occurred largely during the subsequent period of much slower growth and generally lower prices. The introduction and spread of regular wage-labour, the widening and deepening burden of rent on ‘stranger’ cocoa farms, the proliferation of ‘advances’, and the introduction of farm mortgaging are described, together with the accompanying decline of slavery, pawning, and other non-wage forms of labour. Colonial officials ineffectually deplored the growth of money-lending and, to a lesser extent, that of wage-labour. From the mid-1930s, however, the tendency towards greater separation of labour from control of the farm was partly reversed by a new insistence by northern labourers on the replacement of annual wage contracts by a managerial form of share-cropping. This demand was sustained against the opposition of farmowners and despite persistent unemployment, an achievement made possible by the migrants continued foothold in subsistence agriculture in their home areas. This case of migrant labourers successfully challenging the extension of wage relations raises questions concerning the relationships between commercial agriculture and ‘precapitalist’ social relations of production in Africa generally.

Despite the evident importance of fishing in Malawi, its role in the territorial colonial economy has been largely ignored. This paper focuses on the evolution of fishing and fish-trading at the south end of Lake Nyasa (Lake Malawi), emphasising the interaction between ecological change and changes in market opportunity. During the late nineteenth century, fishing played an important role in the economy of the Mang'anja people alongside agricultural production. Communual tasks such as the setting of nets or building of canoes were conducted by male members of an mbumba or matrilineage group who traded fish with the agriculturally productive highland regions nearby in exchange for maize and beans. Little changed initially with the estalishment of colonial rule, though some labour previously employed in fishing may have been diverted into cotton-growing which the Government encouraged in the Upper Shire Valley. The establishment of military camps during the First World War, combined with the sudden drying up of Lake Chilwa, the major source of fish in the Shire Highlands, created the opportunity for enterprising fishermen to start a regular trade in dried fish to Blantyre and Zomba from about 1917. This was stimulated in the 1920s by the steady rise of water levels on the Shire River which brought cotton production virtually to a halt making fishing an attractive alternative.

The advent in the 1930s of non-African commercial fishermen who used lorries to transport fresh fish to Blantyre and dried fish to Salisbury did not prevent a further expansion of African fishing and fish-trading, many of the traders using bicycles to extend their sales into the southern Malawian hinterland. Officials tended to side with African fishermen when their interests clashed with those of incomers, notably the Greek Yiannakis brothers. But they had little success in introducing new techniques to improve productivity and fell back in the 1950s On the prohibition of exports to the Rhodesias, a policy aimed at ensuring a regular supply of fish to workers on European estates within Malawi.

By the 1950s, European companies were recorded as being responsible for over half the fish caught in Malawi. African fishing had been affected by the emergence of a small group of capitalist entrepreneurs, most of them former labour migrants, who had invested their savings in imported nets and boats and employed labour on a regular basis. Mang'anja fishermen now faced competition from Tonga migrants using new technical and organisational methods. In contrast to under-development sterotypes, the indigenous industry continued to expand, with migrant workers playing an important role in the development of fishing.

This article attempts to investigate the role of women in rural society in Morocco, and by extension in the Muslim world of the Near and Middle East. It does so by examining the evidence thrown up by a major crisis, the Rif war of the 1920s. The mobilization and organization of tribal society by Muhammad bin ‘Abd al-Karī;m (Abdelkrim) to fight the war against the Spanish and the French extended to women as well as men, involving them in new tasks under new laws. In the end, however, the evidence points not so much to a revolution in women's lives as to the activation for the purposes of war of a traditional ‘female space’. In so doing, it points to the real importance of the women's sphere in a society which was sexually strongly segregated, confirming the impression derived from studies of more literate, urban and aristocratic Muslim societies of North Africa and the Middle East.

This article surveys research into the business history of Africa completed during the past decade, taking as a point of departure the author's previous essays, ‘Imperial business in Africa’, in this Journal (XVII, (1976), 29–48 and 291–305), and using as a point of reference the published proceedings of two conferences held in Paris and London in 1981 and 1983. It is apparent that knowledge of indigenous and expatriate business in the nineteenth and twentieth centuries has expanded considerably during the past ten years, and also that the studies produced by specialists on Africa have particular strengths: they remain integrated with other branches of history; they have illuminated the relationship between business enterprise and official policy; and they have been concerned to explore the wider social consequences of business activities and to relate historical research to current development issues. The literature reveals some characteristic weaknesses too, quite apart from limitations of source materials: the market for knowledge remains imperfect, and specialists often fail to incorporate work which is available; and their analysis is frequently limited by a reluctance to make use of theories of the firm and of accounting techniques. An explanation of these characteristics is offered, and it is concluded that once the present deficiencies have been recognized they can be overcome, and that the quality of research will improve still further as the subject continues to grow during the next decade.