It is argued that the Nilotic contribution to Bantu Africa consisted essentially in the infiltration of Early Iron Age communities practising a mainly agricultural economy in certain specially favourable environments by a new mode of food production in which cattle-keeping played an important part. This new way of life made possible the food-producing occupation of the drier parts of Bantu Africa, and the interaction between the new and old kinds of food-producers was the essential feature of the Later Iron Age. Wherever the new elements were present, settlement patterns became more dispersed, iron-working and pottery became the occupations of specialists, systems of inheritance changed from matrilineal to patrilineal forms, and political systems tended to become more centralized through the domination of settled cultivators by more mobile and warlike pastoral elites. Although chronological data for the Later Iron Age is as yet greatly inferior to that for the Early Iron Age, it seems likely that the process of Nilotic infiltration began during the last two or three centuries of the first millennium a.d., and spread southward through the centre of the subcontinent, so as to reach the western parts of Zambia and Zimbabwe by the tenth century. The subsequent interaction between the two kinds of food-production came about more gradually over two or three further centuries, and involved many varieties of local evolution in different areas. In southern Zaire, Zambia and Zimbabwe the local development of metallurgical techniques during this period was as significant as the introduction of pastoralism. But, overall, the Nilotic contribution provided the most essential features of the Later Iron Age.

Le présent article fait la synthèse pour l'Afrique centrale des recherches archéo-logiques, et plus particulièrement des datations absolues, effectuées ces quatre dernières années. Comme les autres articles de cette série, seuls sont pris en considération les travaux se rapportant à la fin de l'âge de la pierre et à l'âge du fer. Eu égard à l'immensité de l'aire envisagée, l'archéologie apparaît, en dehors de quelques zones limitées, comme encore fort peu développée. Certains résultats récents méritent cependant d'être notés.

Au Cameroun on dispose maintenant de dates qui permettent pour la première fois de situer les débuts de la production de nourriture ainsi que l'apparition de la céramique dans la frange nord de la forêt équatoriale. Plus au sud, à Kinshasa, il a finalement été possible de dater l'âge du fer ancien. Vers l'est, au Rwanda, de nouvelles dates pourraient confirmer que l'âge du fer a débuté dans la région interlacustre un ou deux siècles plus tôt que ce que l'on admet généralement. Dans ce pays et au Burundi l'étude systématique des anciens fourneaux de fonte du fer a été poursuivie. L'apparition des vestiges caractérisant la seconde période de l'âge du fer doit également être reculée dans le temps. Au sud-est, à Lubumbashi, une date conduit à vieillir encore le début de la fonte du cuivre. Dans la dépression de l'Upemba la fouille de plusieurs sites et de nombreuses nouvelles datations permettent de suivre avec précision pendant près de deux millénaires l'évolution sociale, politique, économique et technique de cette important zone de peuplement. Un éclairage nouveau est ainsi porté sur l'origine de l'Etat luba. Différentes recherches en rapport avec les traditions orales et l'histoire récente ont aussi été effectuées au centre du Zaïre, en Angola, au Cameroun au Rwanda et au Tchad.

This article surveys the latest archaeological research and dating results for West Africa. For the Iron Age, recent fieldwork has been spread widely: especially noteworthy is that bearing on the history of ancient Ghana and Mali. Work on the Late Stone Age appears by contrast to have been rather patchy lately, although various palaeoecological researches continue to improve our understanding of the changing environments affecting West African populations over the last 10,000 years. In south-central Niger, moreover, remains of copper-smelting by a stone-using community are dated to around 2000 B.C. From the same region, as also from northern Ghana, comes further evidence for the inception of the Iron Age during the first millennium b.c.

The article is prefaced by some critical comments on the citing and interpretation of radiocarbon datings in historical discussions, and on the meaning of ‘corrected’ and ‘calendar dates’.

This article seeks to re-examine the contemporary evidence about the early history of the Maravi chieftaincies of northern Zambesia. It challenges the view that the Maravi were long established in Zambesia and became aggressive and expansionist in the late sixteenth century when the Portuguese interfered with the ivory trade. Instead it suggests that these chieftaincies were only formed in the second half of the sixteenth and the first half of the seventeenth century by a number of different groups who migrated into the region north of the Zambezi and who conquered the existing population. Some of these intrusive groups crossed the Zambezi and attacked the wealthy Karanga states to the south, but these invasions were frustrated by the Portuguese who blocked all attempts to settle south of the river. Defeated in their efforts, the Maravi chiefs turned aside to invade the Mozambique lowlands. It was one of these incursions that in the 1580s gave rise to the legend of the marauding, cannibal Zimba.

The establishment of settled Maravi states occurred only as a result of the rise of Muzura in the first half of the seventeenth century. He defeated his rivals with Portuguese aid but was, in his turn, frustrated in his attempts to interfere in Karangaland, and in the end was badly defeated by the Portuguese in 1632. Thereafter he concentrated his activity on developing the ivory trade with the Portuguese on the Zambezi and in Mozambique Island, while the region south of the Zambezi became subject to Portuguese warlords and the trade monopoly of the captain of Mozambique.

Le trafic des esclaves en Libye au XVIIIe siècle représents une branche de commerce relativement importante dans un pays aux faibles ressources économiques.

Les esclaves provenaient du Soudan d'où ils étaient acheminés à travers le désert. La plupart rentraient en Libye par le Fezzan et, dans une bien moindre mesure, par Ghadamès. Le Fezzan jouait le rôle de plaque tournante. Une partie des caravanes était orientée vers la Cyrénaïque, puis l'Egypte, mais nous sommes fort peu renseignés sur leur importance. L'autre partie était acheminée à Tripoli. Les apports annuels d'esclaves dans cette ville, ont varié au cours du siècle. D'une moyenne de 500 à 600 au début, ils sont montés à 2,000 au milieu du siècle, pour redescendre à 1,000 et remonter à la fin à 1,500.

A Tripoli, un nombre réduit d'esclaves se trouvaient retenus sur place. La plupart étaient exportés au Levant, en particulier Constantinople, les régions en bordure de la mer Egée et la Crète. Les traitants tripolitains nolisaient à cet effet des navires européens, surtout français.

Coins and weights provide evidence which can throw light on the origins of the trans-Saharan gold trade. Such a trade does not seem to have existed before the end of the third century a.d., but from 296 to 311 an irregular gold coinage was issued at Carthage, and by the end of the fourth century there were significant changes in the North African tax system to enable more gold to be collected. The solidus, a coin first issued in 312, provided the standard used for weighing gold-dust in the trans-Saharan trade, while copper, a major item of merchandise in that trade, was being imported to Jenne-Jeno by a.d. 400. This strongly suggests that the gold trade first assumed significance in the fourth century. The trade was evidently flourishing before the Arab conquest, for the Byzantine mint of Carthage produced a copious output of gold between 534 and 695. For weighing gold-dust, the standard based on the Roman ounce and the solidus was retained by the Arabs, and survived until the nineteenth century in the Western Sudan. It was also adopted by the Akan of Ghana and Ivory Coast, who made it the basis of their weight-system from about 1400 to 1900.

The Sēfuwa dynasty seized power in Kānem around 1075, but it was only in the beginning of the thirteenth century that the rulers of Kānem were able to extend their authority over Bornū. Prior to this move small groups of Saharan speakers had already established themselves among the Chadic speakers of the Komadugu Yobe valley. Towards the end of the reign of Dūnama Dībalāmi (c. 1210–48) the court of the Sēfuwa itself was shifted to Bornū, mainly as a result of disturbances in Kānem. Indeed, according to oral traditions of the sixteenth century, the Tubu, in alliance with certain members of the Sēfuwa aristocracy, staged a major rebellion against the central government, apparently attempting to resist the strict application of Islamic principles of government by Dūnama Dībalāmi. Towards the end of the thirteenth century powerful rulers were again able to establish the authority of the Sēfuwa on firm grounds: in the east, even on the fringes of Kānem, they brought the situation under strict control and in the west they extended – or confirmed – the political influence of the Sēfuwa dynasty over the focal points of interregional trade which began to rise in Hausaland. Thus Bornū became the central province of the Sēfuwa Empire in spite of the fact that several kings continued to reside temporarily in the old capital of Djīmī situated in Kānem. This major shift of their territorial basis affected the position of the Sēfuwa in their original homelands. Written sources from the end of the fourteenth century show that the increasing involvement of the Sēfuwa in Bornū and its western border states must have changed their attitude towards the people living east of Lake Chad: after having acquired the character of an autochthonous (or national) dynasty of Kānem – in spite of their foreign origin – the Sēfuwa progressively became an alien power in this major Sudanic state, even though the people of Kānem and Bornū were closely related. Furthermore, the rise of a powerful kingdom in the area of Lake Fitrī under the rule of the Bulāla became a serious threat to the Sēfuwa in their original homelands as the warrior aristocracy of the Bulāla state – which must have been of Kanembu origin – remained closely connected with the sedentary population of Kānem. When finally during the reign of 'Umar b. Idrīs (c. 1382–7), the Sēfuwa were forced by the Bulāla to withdraw their forces from Kānem, this territorial loss did not affect the future development of the Empire to the extent that has formerly been supposed, since losses in the east were largely compensated by earlier gains in the west.

Some 170 references to drought and disease along the south-western coast of Central Africa between 1550 and 1830 suggest that climatic and epidemiological factors motivated the farmers and herders of West-Central Africa in historically significant ways. Nearly all references come from documentary sources and so bear primarily on conditions in the drier and less fertile areas near Luanda and to the south, where African reactions would have been strongest.

While minor shortages of rain occurred too frequently to receive much explicit attention in the documents, longer droughts spread more widely every decade or so and attracted notice. Major periods of dryness, extending for seven years or more and touching all parts of the region, occurred perhaps once each century and produced comments throughout the documentation.

Localized minor droughts hardly disrupted the lives of Africans, who had presumably devised agricultural and pastoral strategies to take account of such ordinary climatic variation. Two-or three-year rainfall shortages produced banditry and warfare that often attracted Portuguese military retaliation. Major droughts disrupted polities and societies and hence coincided with major turning points in West-Central African history in the late sixteenth and late eighteenth centuries. In the earlier case, agricultural failures produced the famed ‘Jaga’ or Imbangala warriors, who elevated pillage to a way of life and who joined the Portuguese in establishing the Angolan slave trade. The later, protracted drought from 1784 to 1793 coincided with the historic peak of slave exports from West-Central Africa.

Nuer–Dinka relations are usually described as being based on constant mutual hostility. An examination of Nuer–Dinka relations along the Sobat and Zaraf valleys since the beginning of Nuer eastward expansion in the nineteenth century reveals a different pattern. Conflict during the immediate Nuer conquest of Dinka territory was followed by assimilation of individual Dinka into the Nuer social and political system. Through kinship ties connecting both communities a number of Nuer segments began to have closer relations with various Dinka groups than they had with related but more distant Nuer segments. Such relations inhibited the expansion of hostilities, and Nuer raids against the Dinka became increasingly selective. By the end of the nineteenth century a number of religious leaders deliberately fostered the growth of a Nuer–Dinka community. With the arrival of the Anglo-Egyptian administration an attempt was made to separate the Nuer and Dinka into discrete political units, and this policy exacerbated tensions between some of the neighbouring Nuer and Dinka settlements. The policy of separation ultimately failed due to the strength of the ties linking the two peoples. The model of structural opposition between Nuer and Dinka which Evans-Pritchard presented is based on data gathered during the early 1930s when separation was in force. It must now be modified by an examination of evidence covering a longer historical period.

Until 1471 the Wangara had enjoyed a monopolistic position in the Akan gold trade, consigning bullion to the markets of the Western Sudan (see the first part of this paper). The Portuguese entered the trade in the late fifteenth century, but experienced difficulty in deflecting gold to the coast. The strong demand for labour in the Akan country obliged them to import slaves from other parts of West Africa in order to achieve competitiveness. In 1477 Jenne – the principal northern outlet for Akan gold – fell to the Songhay, and for a time the Wangara were induced to do at least some of their business with the Portuguese. Subsequently the Wangara found markets for their gold as far west as the Gambia river. In the mid-sixteenth century the ruler of Mali, his frontiers crumbling on all sides, made a bid to take control of the Wangara gold trade. His troops occupied Bitu or Bighu, the Wangara entrepot on the edges of the Akan forest country, and he may (just possibly) have ordered an attack on Elmina, the principal Portuguese post on the coast to the south. If there was such a move against Elmina, it certainly failed, and at least some of the troops in Bitu did not return to Mali but set up their own state locally: Gonja. The developing Atlantic economy, built around new supplies of gold and new demands for slaves, eclipsed the older Mediterranean economy of which West Africa had been a geographically peripheral but commercially central part. The decline of Mali, and with it Bitu, was irreversible.

Recent archaeological excavations have revealed two distinct pottery traditions in the Birim Valley, southern Ghana. These have been classified as the ‘Earthworks Ware’ and the ‘Atwea Ware’. In certain archaeological contexts, the ‘Atwea Ware’ succeeds ‘Earthworks Ware’, and it also continues into present-day ethnography. The discontinuity between these two pottery traditions suggests a change in population. It is therefore suggested here that the population of ‘Earthworks Ware’ makers was one of the early victims of the Atlantic slave trade from about the mid-sixteenth century and that they were replaced in this area of the Birim Valley around a.d. 1700 by the Atweafo, a Twi-speaking group, whose descendants live there to this day. From the eighteenth century until close to the end of the nineteenth century a number of Denkyira, Asini and Asante migrants also moved into this valley. During this time the militarily weak Atweafo lived at the mercy of four major powers – the Asante, Akim Abuakwa, Akim Kotoku and the Akwamu. However, the Atweafo found means to survive under what seems to have been a highly volatile political environment by shifting their loyalty amongst these powers as situations dictated.

In late medieval and early modern times West Africa was one of the principal suppliers of gold to the world bullion market. In this context the Matter of Bitu is one of much importance. Bitu lay on the frontiers of the Malian world and was one of its most flourishing gold marts. So much is clear from sixteenth- and seventeenth-century writings, both African and European. A review of this body of evidence indicates that the gold trade at Bitu was controlled by the Wangara, who played a central role in organizing trade between the Akan goldfields and the towns of the Western Sudan. It is shown that Bitu cannot be other than Bighu (Begho, Bew, etc.), the abandoned Wangara town lying on the northwestern fringes of the Akan forest country, which is known (from excavation) to have flourished in the relevant period. In the late fifteenth century the Portuguese established posts on the southern shores of the Akan country, so challenging the monopolistic position which the Wangara had hitherto enjoyed in the gold trade. The Portuguese sent envoys to Mali, presumably to negotiate trade agreements. The bid was apparently unsuccessful. The struggle for the Akan trade in the sixteenth century between Portuguese and Malian interests will be treated in the second part of this paper.

This article provides a synthesis of the various studies which attempt to quantify the trans-Atlantic slave trade. Since the publication of Philip D. Curtin's pioneering estimates in 1969 (The Atlantic Slave Trade: A Census), there have been numerous revisions of different sectors of the trade, and some scholars – notably J. E. Inikori and James Rawley – have argued that Curtin's global estimate for imports into the Americas is too low. When the revisions are examined carefully, however, it is apparent that Curtin's initial tabulation was remarkably accurate. The volume of exports from Africa across the Atlantic is here calculated at 11,698,000 slaves, while imports into the Americas and most other parts of the Atlantic basin are estimated to have been 9·8–9·9 million slaves – well within range of Curtin's original Census. Many of the revisions are based on shipping data by national carrier, rather than on series derived from estimated imports into different colonies in the Americas. Hence it is possible to substitute new data for much of the eighteenth and nineteenth centuries for many of the import-derived series used by Curtin. The results of these substitutions shift the distribution of slave exports over time but do not affect estimates of the relative scale of the trade by more than 2–3 per cent – hardly significant considering the quality of the data. Inikori and Rawley have failed to distinguish clearly between imports by colony and exports by national carrier; hence their global estimates have resulted in double counting. Further revisions are likely, nonetheless, but until the completion of detailed research comparable to the studies of David Eltis, Roger Anstey, Johannes Postma, and a dozen other scholars it is not possible to estimate the extent of future modifications. In the meantime, the current state of research on the volume of the Atlantic slave trade is summarized in a series of tables which analyse the export trade by time period, national carrier, and coastal origin. It is expected that the present synthesis will challenge historians to examine the impact of the slave trade on different parts of Africa, both to test the regional breakdown of slave exports and to assess the demographic, political, economic and social repercussions on Africa.

The mechanics of food production by peasant cultivators have received relatively little attention from historians of colonial Africa, and yet a knowledge of food production systems and the labour they require is crucial to any understanding of rural change and stratification in the colonial period. On the Shire Highlands of Southern Malawi, ecological disturbance and the alienation of land in the early years of the twentieth century meant that an intensification of labour on food production was needed if hunger was to be avoided. The differential success of various groups in holding on to their family labour was a major factor making for economic differentiation in this period. Full-scale famine was avoided by the ability of some groups to adopt new cropping patterns, intensify labour, and thus continue to produce a food surplus. A degree of land shortage could be accommodated as long as labour could be intensified on the available land. In the Upper Shire valley at the same time, there was little shortage of land, but the concern for food security there was important in determining the outcome of the introduction of cotton as a cash-crop. The attainment of food security in this area, which was prone to drought, involved the deployment of family labour on food production over much of the year, including the planting of second crops and back-up crops in the dry season. The introduction of cotton was largely a failure because the returns did not compensate for the loss of labour on food production.

The colonial authorities perceived Upper Volta as the most important labour reservoir in French West Africa. Koudougou, the most populous administrative subdivision of what was Upper Volta during the colonial period, was probably also the most densely settled. Various types of colonial labour policies are examined. These include forced labour in its forms of annual ‘prestations’ and longer-term, large-scale recruiting for major public works, as well as military conscription and the use of the ‘2è portion’ in the public and private sectors. Based on archival research in West Africa, this study traces changes in the size of the labour force and the impact of the forced extraction of labour on the African populations. Finally, the wider significance of colonial labour policies is analysed.

During the Second World War the British West African colonies supplied raw materials and manpower for the war effort. The small peacetime army of the Gold Coast increased to nearly 70,000 men, including technical and service corps, and was used in overseas campaigns. Most soldiers were drawn from the supposed martial peoples of the Northern Territories but recruiting was extended to Asante and the south in mid-1940. Although formal conscription was only applied to drivers and artisans, a large number of recruits were forcibly enlisted through a system of official quotas imposed on districts and through chiefs. Opposition to military service, especially for overseas compaigns, was widespread and is indicated by the attempts to evade recruiting parties and also the large number of desertions. In order to release labour for the military and also conserve scarce supplies of raw materials, some gold mines were closed. Wartime shortages, inflation and the lack of jobs after the war led to discontent in the Gold Coast but there is little evidence to indicate that this resulted in a significant number of ex-servicemen being drawn into political activity.

This paper traces the process of the decline of African middlemen in Eastern Nigeria from wealthy entrepreneurs of the late nineteenth century to petty traders after 1930. Four phases are identified in their decline. During the first phase, 1900–5, the middlemen lost only the political control of their trading areas but benefited commercially. The establishment of colonial rule expanded their market. And with the reluctance of expatriate firms to move into the interior, the continued ignorance of the natives of the actual prices of their produce and of imported goods, and the encouragement from the colonial administrators, African middlemen prospered.

These advantages were lost during the second phase, 1905–16. As the firms began to move inland from 1905, they traded with the natives and fostered a new group of smaller and dependent middlemen. The middlemen's market began to contract and their wealth declined. Their fortunes worsened during the third phase, 1916–30. With the opening of the Eastern Railway to traffic in 1916 and the increased construction of roads during this period, the firms intensified their penetration of the interior, swallowing up what remained of the middlemen's market. The introduction of produce inspection in Eastern Nigeria in 1928 added more hardships for the middlemen, putting many out of business. And by 1930 the trust system, upon which most middlemen depended after 1916 for raising their trading capital, collapsed, leaving most of them impoverished.

Thus, after 1930 African middlemen were no more than petty traders, trading with little capital and making marginal profit. They became incapable of challenging the expatriate firms in the import–export trade as their predecessors had done in the nineteenth century. The firms employed various trade malpractices to ensure that the African traders retained this status until the 1940s.

In the sixteenth century most iron used in west Africa was produced within the region. Extra demand may have been met from the newly established European factors on the coast. By the end of the nineteenth century, in contrast, it was the residue in demand that was satisfied from local sources, the main bulk of iron being imported via the coast and transported inland. For the larger part of this 400-year period imported iron was cheaper than locally-produced iron. What was remarkable, then, was not that iron smelting eventually died out, but that it survived for so long and could be studied in detail in the second half of the twentieth century.

It is argued that, although the decline can be related to production constraints, such as the availability of charcoal, influences originating from the rest of the community can be seen to have prolonged the survival of local iron. The organization of labour of both the iron-smelting and blacksmithing processes, together with the way in which iron was marketed, are central to the analysis. In addition, consumption factors are of the utmost importance. Apart from the prejudice against innovation, the fact that imported iron was plainly not as suitable as local iron for the purposes to which it was put, weakened its impact. Also the ritual attitude to local iron has to be taken into consideration. The present universality of non-local sources has resulted in a change in the regard paid to the metal, but it is argued that the position of the smith is unlikely to alter significantly, since it is more related to his crucial role as supplier of tools for other essential activities such as farming, than to the production of iron itself.