¹ This concept is most cogently and economically discussed by Philip Curtin, Steven Feierman, Leonard Thompson and Jan Vansina in their collective text, African History (Boston, 1978), 215–19.Google Scholar

² Hopkins, A. G., An Economic History of West Africa (London, 1973), 87, 89Google Scholar; Ajai, J. F. Ade and Espie, I. (eds), A Thousand Years of West African History (Ibadan, 1965), 217–25.Google Scholar

³ Kiyaga-Mulindwa, D., ‘The earthworks of the Birim Valley, Southern Ghana’ (Ph.D. dissertation, Johns Hopkins University, Baltimore, 1978).Google Scholar

⁴ These were from sites in southern Ghana such as Akropong, Aburi, Odumasi, Amoaful, Anamabo, Axim and Prasu; see Burton, R. F., ‘On stone implements from the Gold Coast’, Journal of the Anthropological Institute, xii (1882), 449–54Google Scholar, and also from Bekwai, Asante; Nsuta; Monkey Hill, Obuasi, Asante; Abomponsu in Asante; Fomena in Adansi, and Akwatia, Manso, Topiramang and Boadua all in Akim; see Wild, R. P., ‘Stone artefacts of the Gold Coast and Ashanti’, The Gold Coast Review, iii, ii (1927)Google Scholar; and ‘Stone Age pottery from the Gold Coast and Ashanti’, Journal of the Royal Anthropological Institute lxiv (1934), 203–18.Google Scholar

⁵ Wild, , ‘Vestiges…’, The Gold Coast Review, v (1929–1931), 17.Google Scholar

⁶ Ibid. 3; Brown, J. P., Gold Coast and Asiante Reader, I (London, 1929).Google Scholar This is also what most people in the Birim Valley believe, as is in the case of Birikuku Hill near Akim Manso, D. Kiyaga-Mulindwa ‘Birim Valley Selected Historical Texts’ (1978) no. 41 (copies available at Department of Archaeology, University of Ghana, Legon; Eisenhower Library, Johns Hopkins University, Baltimore, Md., and with the author).

⁷ Wild, , ‘Vestiges’, 7.Google Scholar

⁸ Ibid. 17.

⁹ Kiyaga-Mulindwa, ‘The earthworks’ and Figs 1 and 2.

¹⁰ It should be noted that this possibility of ‘overlap’ of the two traditions is not indicated by the material evidence within the stratigraphy of excavated sites, but rather derives from margins of error in several of the carbon-14 dates obtained from these excavations.

¹¹ Looking at the map of the Birim Valley (Map 1) this area appears to have been enclosed by physical barriers and therefore in a way cut off from most of the north-south or west–east communication network of the pre-nineteenth-century period. There is the Atewa range rising up to about 800 metres and stretching from Asamankese to Anyinam in the north-east. Then from Anyinam to Nkawkaw, in a north-west direction, are the southern Kwahu Scarp and the North Fumang Scarp with peaks of about 720 metres high. From Nkawkaw westwards the Birim Valley is hedged by a series of high ridges such as the Jade Bepo (700 metres high); the Ajenjua Bepo (500 metres high) and the Miransa Hills (300 metres high). The river Pra which rises from the hills between Nkawkaw and the Jade Bepo Hows between the Miransa Hills and the Ajenjua Bepo to join its waters with those of the Anum to the south-west of Nkawkaw, thus becoming a torrent river and forming a formidable western barrier for the Birim Valley. Because of these barriers most of the trade traffic from and to further north appears to have snaked around rather than through the Birim Valley, mainly preferring the major trade routes along the western bank of the Pra to Elmina and Cape Coast, or the route through the Nkawkaw Gap and south to Accra.

¹² For examples of these early maps see Bosman, W., A New and Accurate Description of the Coast of Guinea (London, 1967)Google Scholar, Daaku, K. Y. and van Dantzig, A., ‘Map of The Gold Coast, December 25, 1629’, Ghana Notes and Queries, ix (1966)Google Scholar, and Atlas de Delisle, Carte de la Barbarie de la Nigritie et de Guinée, 1707, (Atlas 3.70.1).

¹³ Kiyaga-Mulindwa, D., ‘The Akan problem’, Current Anthropology, xxi, iv (1980), 503–6.CrossRefGoogle Scholar

¹⁴ Vogt, J. L., ‘The early, São Tomé–Principe slave trade with Mina, 1500–1540’, The International Journal of African Historical Studies iv, iii (1973).Google Scholar

¹⁵ Wilks, I., ‘Law, labour, capital and the forest kingdom of the Asante: a model of early change’ (unpublished paper, Northwestern University, n.d.).Google Scholar

¹⁶ Fage, J. D., A History of West Africa (London, 1972)Google Scholar, table A, 83.

¹⁷ Curtin, P., The Atlantic Slave Trade: A Census (Madison, 1969), 123, 150Google Scholar and table 43.

¹⁸ Macdonald, G., The Gold Coast Past and Present: A Short Description of the Country and Its People (London, 1895), 13, 28, 29Google Scholar; Ward, W. E. F., History of the Gold Coast (London, 1958), 414–19Google Scholar; Lawrence, A. W., Trade, Castles and Forts of West Africa (London, 1963)Google Scholar, map 5.

¹⁹ Curtin, et al. African History, 234.Google Scholar

²⁰ Donnan, E., Documents illustrative of the History of the Slave Trade to America (New York, 1969), 11, 109.Google Scholar

²¹ Burton, R. F., ‘Ocean highways’, The Geographical Review (Feb., 1874), 452.Google Scholar

²² Kiyaga-Mulindwa, ‘The Earthworks…’, appendix F.

²³ Fynn, J. K., ‘Oral Traditions of the Fante States’, nos. 1–7, Institute of African Studies, University of Ghana (Legon, 1975).Google Scholar Slaves were procured from the ‘north’ and these included some Twi-speaking people from the north and others. On panyaring or man-stealing, see Dupuis, J., Journal of a Residence in Ashantee (London, 1966), 36Google Scholar, and Ajayi, and Espie, , A Thousand Years, 358, n. 1.Google Scholar

²⁴ D. Kiyaga-Mulindwa, ‘The earthworks…’. See ground plan of MN₅ and the cross-section in figures 15 and 17 respectively. From this cutting it was found that the top of the bank had indications of postholes spaced out at almost equal intervals. A reconstruction based on these findings shows that the entire perimeter of the earthwork was surrounded by a barricade running along the top bank. This barricade was made of posts of about 15 centimetres in diameter spaced out at intervals of about 2 to 3 metres and these interspaced with smaller uprights, leaving gaps of about one metre in between. These uprights must have been interlaced with horizontal beams, thus giving the barricade its strength.

²⁵ Kea, R., ‘Firearms and warfare on the Gold and Slave coasts from the sixteenth to nineteenth century’, Journal of African History, xii, ii (1971).Google Scholar

²⁶ Kiyaga-Mulindwa, D. ‘Birim Valley Historical Texts’ (1978), 1–15.Google Scholar

²⁷ Hunter, J., ’River blindness in Nangodi, northern Ghana: a hypothesis of cyclical advance and retreat’, Geographical Review, lvi (1966), 398–416.CrossRefGoogle Scholar For another example of depopulation through slave trade see Azevedo, M. J., ‘Precolonial Sara Society in Chad and the Threat of Extinction Due to the Arab Muslim Slave Trade 1870–1917’, Journal of African Studies, vii, ii (1980).Google Scholar

²⁸ Boahen, A. A., ‘The effects of the Atlantic slave trade on West Africa. A rejoinder’, Ufahamu, vii (1966/1967), 120.Google Scholar

²⁹ Hair, P. E. H., ‘Ethnolinguistic continuity on the Guinea coast’, Journal of African History, viii, ii (1967), 247–68.CrossRefGoogle Scholar

³⁰ Kiyaga-Mulindwa ‘The Akan problem’.

³¹ Ibid. 503.

³² Wilks, I., ‘Akwamu, 1650–1750: a study of the rise and fall of a West African empire’, MA thesis, University College of Wales, Bangor (1959), 102–6Google Scholar; Furley Papers, N. 40, 1715–17; also Reindorf, C., A History of the Gold Coast and Asante (Legon, 1966), 70.Google Scholar

³³ Christian, L. A., ‘Akwamu remnants in Akim Abuakwa’, BA Graduating Essay, Department of History, University of Ghana, Legon, 1974Google Scholar; Reindorf, , History, 71.Google Scholar

³⁴ Donnan, , Documents, 109–12.Google Scholar

³⁵ Bosman, , New and accurate description, 76–7Google Scholar, and Daaku, K. Y., Trade and Politics on the Gold Coast 1600–1720 (Oxford, 1970).Google Scholar

³⁶ Curtin, , Atlantic Slave Trade 161, 185–6, 208–9.Google Scholar

³⁷ Parliamentary Papers, 1842, xi, 171.

³⁸ Dupuis, , Journal, 45.Google Scholar

³⁹ This tradition is also recorded by Hill, P., ‘The history of migration of Ghana cocoa farmers’, Transactions of the Historical Society of Ghana, iv (1959)Google Scholar, and Hunter, J. M., ‘Cocoa migration and patterns of land ownership in the Densu Valley, near Suhum, Ghana’, in Prothero, R. M. (ed.), People and Land in Africa South of the Sahara (London, 1972), 86.Google Scholar I suspect they got this information from Field, M. J., Akim Kotoku and Oman of the Gold Coast (Westport, 1970), 1.Google Scholar Unfortunately, carried away by oral tradition, all these sources create an impression that the forests were uninhabited, until the seventeenth century, which assertion is refuted by the archaeological evidence now at hand.

⁴⁰ Kusi tradition. Interview with John Oteng (alias Kwasi Kuma) and Ohene of Kusi (Kiyaga-Mulindwa, D., ‘Birim Valley Selected Historical Texts’, No. 034, 1978Google Scholar).

⁴¹ Nketia, J. H., Funeral Dirges of the Akan Peoples (Westport, 1969), 143.Google Scholar

⁴² Dupuis, Journal, part I, xxxii, and his map of Wangara. Also map facing p. 448 in Burton ‘Ocean highways’; Bowdich, T. K., Mission from Cape Coast Castle to Ashantee (London, 1819), 233Google Scholar, and Christaller's, J. introduction to A Dictionary of the Asante and Fante Language called Tshi (Basel, 1881).Google Scholar

⁴³ Mac Donald tbe cooast 152

⁴⁴ Christaller, A Dictionary; Wilks, I., Asante in the Nineteenth Century (Cambridge, 1975), 6, 365Google Scholar; Rattray, C., Religion and Art in Ashanti (Oxford, 1927).Google Scholar

⁴⁵ Kiyaga-Mulindwa, ‘Birim Valley’, texts 22, 29 and 30.

⁴⁶ This kind of settlement as well as the forces that brought it about were common in most of the West African forest during this period. For comparison see Mabogunje, A. L., Yoruba Towns (Ibadan, 1971).Google Scholar This point is also well illustrated by Tham, P., ‘Photo-grammetry in Archaeology’, Liberian Studies Journal, 1 (1972–1974), 7–16.Google Scholar

⁴⁷ Otwereso and Achiase are important towns in the state of Akim Abuakwa, but are situated in the south and southwest corners of the Birim Valley, several miles away from the capital Kibi, in the northwest corner of this valley. They are separated from their capital by extensive tracts of Akim Kotoku territory; Field, Akim Kotoku, 1–2, and her map of Western Akim.

⁴⁸ The idea of settling at crossroads was to avoid surprise attacks, which are difficult to guard against if one is entirely surrounded by forest on all sides.

⁴⁹ Gold Coast Government Census Returns for 1891 and 1911, Government Printer, Accra.

⁵⁰ These places are referred to by Wilks, ‘Akwamu, 1650–1750’, and Hill, ‘History of Migration’, as being Akwamu towns. Traditions from these towns do indicate that most of them were never vowed to Akwamu, and to prove this they point to the fact that they never migrated with the Akwamu when the Akwamu crossed the Volta. The exception to this claim is the town of Pankase, whose elders claimed that they were truly Akwamu until the coming of Kotokuhene when they were still at their old settlement of Akwasuo (Okuasuo). While there, their chief, Nana Nyama, a relative of Akwamuhene, Ansa Sasraku, was obliged to borrow money (3 mpredwan) to settle a debt in Kwahu. The Kotokuhene, on lending him the money, obliged Nana Nyama to pledge vassalage to him and his successors. Kiyaga-Mulindwa, D., ‘Birim Valley Historical Texts’, n. 53.Google Scholar

⁵¹ The agents of the Kotokuhene had another tactic of using fictitious abusua relationships as a means of infiltrating the flanks of Atvvea settlements that were not willing to budge in face of Kotoku pressure such as Abirem. See Ghana National Archives, Accra, ‘Proceedings of the Hearing re Karikari vs Abiremhene in the Native Court “A” of Akim Kotoku-Oda Eastern Province – Gold Coast’, 5/12/1949 ADM 11/1233. Financial obligations, as indicated in the case of Pankase (footnote 50 supra), also won the Kotokuhene those otherwise ‘unwilling partners’.

⁵² Fynn, J. K., Asante and Its Neighbours 1700–1807 (London, 1971), 45.Google Scholar

⁵³ Ibid. 69.

⁵⁴ Wilks, ‘Akwamu, 1650–1750’.

⁵⁶ Reindolf, , History, 81–2.Google Scholar This appears to have been a common practice followed by Asante after each successful conquest of a particularly stubborn rebellious people, right into the nineteenth century. See for instance Wilks, , Asante, 83–7.Google Scholar

⁵⁶ This was after the Akantamasu war (1826) mentioned in the Abirem Land Case (see n. 51)above. Johnson, M., ‘Migrants progress’, Ghana Geographical Association Bulletin, IX (1964), 8Google Scholar, dates this war to around 1860. Sarbah, Mensah, Fanti National Constitution (London, 1968)Google Scholar also mentions the war between Amoa Atta and Atta at around 1860.