Читать книгу Ouidah - Robin Law - Страница 12

1

Origins

Ouidah Before the Dahomian Conquest

The history of Ouidah is intelligible only by reference to its geographical situation, which has however often been misunderstood and misrepresented in accounts of the operation of European trade in West Africa. It is commonly referred to as a ‘port’, but this is strictly inaccurate, indeed positively misleading.¹ Although it became an important centre for European maritime trade from the seventeenth century onwards, it is not in fact situated on the coast, but some 4 km inland, actually to the north of the lagoon which in this area runs parallel to the coast, and so separates Ouidah from the seashore. The slaves and other commodities exported through Ouidah had therefore to be taken overland and across the lagoon to the beach, rather than being embarked directly into European ships. At the coast itself, moreover, there is no ‘port’ in the sense of a sheltered harbour, but only an open roadstead. Indeed, heavy surf along the beach, and on sandbars parallel to it, makes it impossible for large vessels to approach close to the shore. European ships trading at Ouidah had therefore to stand 2–3 km off, and to communicate with the shore through smaller vessels, for which purpose African canoes were normally employed. The town’s relative isolation from the sea, combined with its proximity to the coastal lagoon, played a critical role in shaping its historical development, during as well as prior to its involvement in the trans-Atlantic trade.

Although ‘Ouidah’ is the spelling of the town’s name current nowadays, it occurs in European sources between the seventeenth and nineteenth centuries in various other forms: in English most commonly ‘Whydah’, in Dutch ‘Fida’, in French ‘Juda’, and in Portuguese ‘Ajudá’. All these are attempts to render an indigenous name that would be more correctly written, by modern conventions, as ‘Hueda’ (or in a dialect variant ‘Peda’). Strictly and originally, Hueda was not the name of the town nowadays called Ouidah, but rather of the kingdom to which it belonged in the late seventeenth and early eighteenth centuries, whose capital was Savi, 11 km further north.² The people of Hueda belonged to the same linguistic group as the Fon of Dahomey, although historically distinct from them; this language family is nowadays generally called by scholars ‘Gbe’ (but formerly commonly ‘Ewe’ or ‘Adja’, and in French colonial usage ‘Djedji’).³ In contemporary sources, the name Hueda may have been noted first by Spanish missionaries visiting the kingdom of Allada to the north-east in 1660, who recorded it (apparently) in the form ‘Jura’ or ‘Iura’;⁴ more certainly, it enters the historical record in 1671 (as ‘Juda’), when the French first established a trading factory there.⁵ In 1727 the Hueda kingdom was conquered by the inland kingdom of Dahomey. As a political unit, it thereafter survived only in the form of a minor successor-state, formed by refugees from the Dahomian conquest, on the western shore of Lake Ahémé (Hen), about 20 km west of Ouidah, this relocated kingdom being distinguished as Hueda-Henji, ‘Hueda on [Lake] Hen’.⁶ However, the name Hueda (in its various European misspellings) continued to be applied to the coastal town, now subject to Dahomey. In the present work, to avoid confusion, the form ‘Hueda’ is used only to refer to the pre-1727 kingdom, and after 1727 to the successor kingdom-in-exile established to the west, while the modern form ‘Ouidah’ is used of the town.

Strictly, although the town could properly be described as ‘[in] Hueda’, the use of this name to designate the town specifically is in origin a foreign, European terminology; and in local usage even today ‘Ouidah’ remains its normal name only in French. The correct indigenous name of the town, which is still usually used by its inhabitants when speaking in the local language, Fon, is Glehue (in French spelling, ‘Gléhoué’). This name also regularly occurs in contemporary European sources from the seventeenth century onwards. The earliest extant document written from Ouidah, a letter from an English trader in 1681, is dated, quite correctly, at ‘Agriffie in Whidaw’, i.e. Glehue in Hueda.⁷ Later, Europeans used versions of the name Glehue interchangeably with, although less commonly than, Ouidah: for example, in English ‘Grigue’, ‘Griwhee’ or ‘Grewhe’; in French sometimes ‘Glégoué’ or ‘Grégoué’, but most commonly ‘Gregoy’.⁸

The French trader Jean Barbot, who visited the Hueda kingdom in 1682, gives the coastal village that served as its commercial centre a further different name, ‘Pelleau’.⁹ This name does not occur independently in reference to Ouidah in any later source, and Richard Burton, who enquired about it at Ouidah in the 1860s, found it ‘now unknown’.¹⁰ What seem to be versions of this name do occur, however, in European sources earlier in the seventeenth century, applied to a place on the coast between Popo (nowadays Grand-Popo, 30 km west of Ouidah) and Allada (whose principal coastal trading outlet was at Godomey, 30 km to the east): ‘Fulao’ and ‘Foulaen’.¹¹ From the situation indicated, this was presumably also identical with the later Glehue/Ouidah. The names ‘Pelleau’, ‘Fulao’ and ‘Foulaen’ probably represent Hula, or in an alternative form Pla, which is the name of an ethnic group (whose language belongs, like Fon and Hueda, to the Gbe family) which according to tradition originated in Grand-Popo (whose correct indigenous name is, in fact, Hula) and migrated east to settle at various places along the coast, including in particular Jakin (modern Godomey).¹² ‘Offra’, the name given by Europeans to their principal place of trade in Allada during the second half of the seventeenth century (which was situated close to, though distinct from, Jakin), is clearly another variant of this name. The application of this name to Ouidah presumably reflects the fact that an important, perhaps originally the dominant, element in its population was Hula rather than Hueda.

The foundation of Ouidah

Stories of the foundation of Ouidah are in fact contradictory. The original settlement, which predated European contact, is generally identified today with the quarter called Tové, on the north-eastern side of the town; and this is consistent with a report of the early eighteenth century that the indigenous village of Glehue was situated to the east of the French and English forts there.¹³ There is also, however, a compound called ‘Glehue Daho’, i.e. ‘Great Glehue’, to the west of Tové (nowadays considered to fall within Fonsaramè, the Dahomian quarter of the town); although now occupied by a Dahomian family, Nassara, this is also sometimes claimed to represent the original pre-Dahomian settlement, as its name implies.¹⁴

The founder of Ouidah is regularly named in local tradition as Kpase (in French spelling, ‘Passè’), who is in consequence the subject of a cult in the town to the present. After his death, he is said to have metamorphosed into a tree that still survives as the focus of his shrine, in what is known as Kpasezun, ‘Kpase’s Forest’, located in Tové quarter, or, rather, originally in the bush beyond Tové, but nowadays absorbed within the town.¹⁵ In contemporary sources, however, the earliest reference to the story of Kpase and his cult in Ouidah is only from the 1840s.¹⁶ The inhabitants of Tové are said to have been dispersed in the Dahomian conquest of 1727, but subsequently resettled there under Dahomian rule; they were led in this resettlement by a nephew of Kpase called Tchiakpé, who founded a family that still exists in the quarter.¹⁷ The dominant family in Tové in recent times, which also controls Kpase’s shrine, called Adjovi, rose to prominence only in the nineteenth century, but claims descent from Kpase (although this claim is disputed by others in the town).¹⁸

Kpase is normally supposed to have been a king of Hueda,¹⁹ usually identified as its second ruler, son and successor to the founder of the kingdom, who is named as Haholo.²⁰ While this has become the canonical version, however, a different account of the origins of Ouidah is given in the traditions of the Hula kingdom of Jakin, whose capital was originally Godomey but was removed, after the destruction of that town by the Dahomians in 1732, further east to Ekpè, and subsequently (after the destruction of the latter in turn in 1782) to Kétonou. These recount the migration of the Hula founder-king, called Kposi (‘Possi’), from Grand-Popo to settle at Glehue, which by implication he founded. This account envisages a period when Glehue was independent of the Hueda king at Savi, with whom Kposi is said to have delimited a frontier. However, subsequently the Savi king is said to have made war on Kposi, driving him to move east to settle at Godomey.²¹ Although the traditions state that this displacement occurred in the reign of Hufon (Houffon), the last Hueda king before the Dahomian conquest (reigned 1708–27),²² it is clear that if historical it must in fact have been earlier; Glehue was evidently already subject to Savi by 1671, when the French established their trading factory there, since they negotiated with the Hueda king for permission to settle it.²³

The names ‘Kpase’ and ‘Kposi’ are sufficiently similar to raise suspicions that they might be variants of a single name, and I suggested earlier that Kpase/Kposi was originally a figure in Hula tradition, whose co-option into the list of Hueda kings is spurious.²⁴ But the two names are understood locally to be philologically distinct. At the very least, however, some degree of confusion (or conflation) between the two figures is indicated by traditional stories relating to the arrival of the first European traders in Ouidah. These agree in attributing the first contact with Europeans to a man called Kpate (‘Patè’), who is said to have been collecting crabs on the seashore when a European ship was passing, and raised a cloth on a pole as a makeshift flag to attract their attention;²⁵ in contemporary sources, Kpate’s name and story were first recorded in the 1860s.²⁶ Like Kpase, Kpate is worshipped as a deified hero. The office of priest of Kpate, or Kpatenon, remains hereditary within a family that claims descent from him, resident in Docomè, the quarter of the Portuguese fort. In different versions of his story, Kpate is associated either with Kpase, the Hueda king at Savi (to whom he allegedly introduced the European traders), or with Kposi, the Hula king settled locally (in whose entourage he originally arrived in Ouidah).²⁷ The former version, it may be noted, implies that Europeans were hitherto unknown; whereas the latter explicitly states that Kposi and Kpate were familiar with them already, from earlier experience at Grand-Popo. There is also a parallel (and evidently related) ambiguity about Kpate’s own ethnic affiliation. Some versions claim that he was, like Kpase, a member of the Hueda royal family;²⁸ current tradition in the Kpatenon family denies this, but agrees that Kpate was Hueda.²⁹ But other accounts state that he was Hula.³⁰ These two traditions, of foundation by the Hueda Kpase and the Hula Kposi, may perhaps be regarded as complementary rather than contradictory, since Ouidah clearly included both a Hueda and a Hula element: the different stories may therefore relate to the origins of different elements within Ouidah, rather than strictly representing alternative traditions of the foundation of the town as a whole.

The Hula element in Ouidah is represented today most visibly by the cult of Hu, the vodun (god) of the sea, who was in origin the national deity of the Hula people. The priest of the cult, the Hunon (Hounon), nowadays has his compound in Sogbadji, the quarter of the English fort, which was established only in the 1680s;³¹ and one of the oldest-established Hueda families in this quarter, called Déhoué, claims to have invited the first Hunon to settle there, implying Hueda priority of settlement.³² However, the traditions of the Hunon priesthood itself claim that Déhoué was instrumental, not in the Hunon’s original settlement in Ouidah, but in his resettlement there after fleeing from the Dahomian conquest in the 1720s.³³ In any case, there is an older shrine of Hu, located in the area called Adamé, which is now included within the Maro quarter of Ouidah, but before the nineteenth century was beyond the south-western limits of the town; and it is this earlier shrine which is said to have been established by the Hula founder-hero Kposi. The cult was certainly established locally already in the seventeenth century, since European accounts of the Hueda kingdom in the 1690s refer to the worship of the sea, to whom offerings were made for calm weather to facilitate the operation of the European trade.³⁴ Hu’s importance was presumably enhanced, as these accounts imply, by the development of the trans-Atlantic trade, but he functioned as patron of watery spaces more generally, including the coastal lagoon; among lesser deities associated with him (and represented as his children) was the goddess Tokpodun, who was linked with the lagoon (and identified with the crocodile).³⁵ The Hula identity was in fact defined by their occupation of the lagoon environment, rather than by their connections with the sea as such. Certainly, the traditions of the Hunon priesthood claim that it was established in Ouidah already before the arrival of the first European traders, and insist that Ouidah was in origin a Hula settlement, in distinction from the Hueda town of Savi.

The question of priority of settlement as between the Hueda and Hula is difficult to resolve, but Hula claims to precedence are supported by evidence relating to the hierarchy of status among the gods worshipped in Ouidah. The national deity of the Hueda was Dangbe, the royal python, originally associated with agricultural fertility, who was incarnated in actual snakes that were maintained in his shrines.³⁶ Dangbe remains today one of the most important vodun of Ouidah, with his principal shrine located in the centre of the town.³⁷ Local tradition nowadays asserts that the cult was instituted in Ouidah from its beginnings by the Hueda founder-hero Kpase.³⁸ In the Hueda kingdom as a whole, as reported in the 1690s, first rank among the gods was held by Dangbe, to whom the sea-god Hu was considered a ‘younger brother’.³⁹ The principal shrine of Dangbe at this period, however, was located at the Hueda capital Savi, rather than in Ouidah; its relocation in Ouidah being a consequence of the destruction of Savi in the Dahomian conquest in the 1720s.⁴⁰ In Ouidah itself in recent times, it is in fact Hu rather than Dangbe who has been regarded as first in status among local vodun; in contemporary sources, the primacy of the Hunon within the priesthood of Ouidah was first recorded in the 1860s.⁴¹ In local tradition, this reordered ranking of the vodun is linked to the Dahomian conquest in the 1720s, the Hunon then being given ‘a special delegation of the royal authority of Abomey’ over the priests of the other cults, including that of Dangbe.⁴² It seems likely, however, that in this the Dahomians were merely recognizing and confirming the pre-existing local hierarchy, the point of their edict being probably to maintain the local primacy of the Hunon, in spite of the removal of Dangbe’s principal shrine into the town.

The Hula connection might also help to resolve a puzzle about the name of the town, Glehue, commonly explained as meaning ‘Farmhouse’.⁴³ It has been argued that the form of this name is linguistically Fon, rather than Hueda; and this has led to the suggestion that it represents a Fon ‘translation’ of a hypothetical original Hueda name, Single.⁴⁴ This, however, seems improbable, since, as has been seen, the name Glehue was already in use before the Dahomian conquest, being attested in contemporary European sources from the 1680s onwards; and it is difficult to understand why Europeans should have adopted a Dahomian form of the name, rather than the one current locally. Possibly the name was originally Hula rather than Fon, since names of this form also occur in Hula country;⁴⁵ however, even if Ouidah was not a Hula settlement, the Europeans, approaching it from the sea and therefore via the Hula, might have employed a Hula version of its name.

The principal local history of Ouidah, by Casimir Agbo, dates the foundation of the town by Kpase to ‘around 1550’.⁴⁶ This date is evidently based upon an earlier suggestion, by the French administrator Gavoy, that the encounter of Kpase and Kpate with the first Europeans to visit Ouidah occurred around 1580, with an allowance added for Kpase’s rule prior to this event.⁴⁷ The traditions of the priesthood of Kpate, however, give an alternative and earlier date for his encounter with the Europeans, 1548.⁴⁸ It may also be noted that the traditions of the Hunon priesthood give a list of eight predecessors in the title prior to the present incumbent, for whom dates of tenure are supplied which indicate that the first took office in 1452; but since this involves an improbably long average tenure of over 50 years each (and a term of office for the first Hunon of over 120 years, 1452–1581), this should evidently be taken symbolically, as an assertion of antiquity (and by implication, priority) of establishment, rather than literally.⁴⁹ There is no reason to suppose, however, that any of these dates have any firm basis. Gavoy’s date of c. 1580, for example, although sanctified by frequent repetition, was by his own account merely a speculative estimate made on the basis that the last king of Hueda, Hufon, displaced by the Dahomian conquest of 1727, was the third successor to (and great-grandson of) Kpase, on the assumption of an average length of reign/generation of 30 years (though the mathematical calculation is bungled).⁵⁰ However, it is known from contemporary sources that the king of Hueda recalled in local tradition as the son and immediate successor of Kpase, Agbangla, was reigning from the 1680s, dying in 1703.⁵¹ This might be held to suggest that Kpase and his foundation of Ouidah belong rather to the middle of the seventeenth century. But this is surely to take too literalistic a view of traditional history, and in particular of the remembered royal genealogy, which may well be telescoped, even if not in part fictitious.⁵² All that can be said with confidence is that the settlement at Ouidah predated the beginnings of European trade there in the seventeenth century.

In the long run, it may be hoped that archaeology will provide more concrete evidence on the early history of settlement in Ouidah. But to date no excavation has been undertaken in the town, apart from limited exploratory work within the courtyard of the former Portuguese fort during reconstruction works there in 1992;⁵³ more systematic excavation was concentrated at Savi, the former capital of the Hueda kingdom, rather than at Ouidah itself.⁵⁴

Environment and economy

The name Glehue, ‘Farmhouse’, is usually explained in local tradition as reflecting the fact that the town was originally established by Kpase as a farm.⁵⁵ Although this story may be no more than an inference from the name, the suggestion that Ouidah was originally an agricultural settlement is consistent with its location, some distance north of the coastal lagoon, on permanently dry and therefore cultivable ground. However, Ouidah’s proximity to the coastal lagoon clearly also played an important role in its early development, and it is likely that the settlement was sited with this also in mind.

The configuration of the lagoon system is complex and varies seasonally with the level of the water, becoming more extensive during the rainy seasons (April to July and October/November). It has very probably also changed over time, through processes of silting and erosion. In recent times, the only permanently continuous waterway in the Ouidah area has been the lagoon immediately behind the coast, called locally Djesin (‘Salt water’). Early European sources, however, speak of two major ‘rivers’ in the Hueda kingdom. The second (called by Europeans ‘Euphrates’), to the north of the capital Savi, is evidently the more northerly ‘lagoon’ called locally Toho, which runs south-eastwards by Savi before turning east into Lake Nokoué; this is nowadays for most of its length no more than a marshy depression, but was presumably a more substantial watercourse in earlier times.⁵⁶ In addition, the area between Ouidah and the coastal lagoon is low-lying and swampy and subject to seasonal flooding, temporarily creating additional watercourses. The only significant area of cultivable land south of Ouidah is around the village of Zoungbodji, halfway towards the beach. Tradition suggests that Zoungbodji is of comparable antiquity to Ouidah itself, attributing its foundation to a man called Zingbo (or Zoungbo), who is regularly linked with Kpate in the story of the arrival of the first Europeans; in the usual version, Zingbo fled in fright at their approach, leaving Kpate to make the first contact.⁵⁷ In contemporary sources, however, the settlement of Zoungbodji is not documented until after the Dahomian conquest in the 1720s, when it became important as the location of a Dahomian military garrison. The only other substantial settlement is Djegbadji, situated on a group of islands in the lagoon to the south-west of Zoungbodji. This was in origin a settlement of the Hula people, although Hueda and later (after the Dahomian conquest of Hueda in the 1720s) Fon elements also settled there subsequently.⁵⁸

In recent times, the lagoon has been an important source of fish, which are caught in static traps, as well as with lines and nets both from the shore and from canoes.⁵⁹ In the nineteenth century, it was noted that fish, rather than meat, formed the staple diet of most of the inhabitants of Ouidah; and dried fish was also traded from the coast into the interior, as far as the Dahomian capital Abomey.⁶⁰ In the late seventeenth century European visitors already noted that the ‘rivers’ in Hueda produced large quantities of fish;⁶¹ and there is no reason to suppose that this tradition of fishing did not date back earlier, before the arrival of the Europeans. Although the main centre of fishing in the area was presumably Djegbadji, families in Ouidah itself were also involved: the Déhoué family of Sogbadji, who claim to have been settled there prior to the establishment of the English fort in the quarter (in the 1680s), were traditionally canoemen and fishermen.⁶² One early eighteenth-century account also noted the existence to the south of Ouidah of salt-works, at which salt was obtained by boiling sea water in jars and was traded into the interior.⁶³ This evidently refers to Djegbadji, which remains a centre of salt production to the present day (as reflected in its name, meaning ‘On the salt marsh’). The reference to the boiling of sea water is inexact since later accounts make clear that salt was extracted in this area from the water of the lagoon.⁶⁴ Concentrations of salt are formed through the evaporation of shallow pools at the borders of the lagoon by the heat of the sun; earth is collected from these, the salt leached out by straining water through it, and it is the resulting highly saline water which is then boiled to produce the salt.

Although Ouidah’s main commercial function in early times was probably in retailing the produce of the lagoon, fish and salt, overland into the interior, it probably also acted as an intermediary in trade conducted by the lagoon itself, which offered a medium of lateral communication and trade along the coast. In recent times, the lagoon has normally been navigable by canoe as far as Porto-Seguro (Agbodrafo), in modern Togo, 70 km west of Ouidah, while to the east it is navigable as far as Godomey, where a brief overland portage can be made to Lake Nokoué, from where navigation continues further east to Lagos and beyond, in modern Nigeria. Some nineteenth-century sources claim that the navigation along the lagoon to the east was originally continuous from Ouidah into Lake Nokoué, the interruption at Godomey being due to recent silting.⁶⁵ However, it was reported already in the seventeenth century that the lagoon was ‘lost in the earth’ at Jakin (Godomey);⁶⁶ recollections of uninterrupted travel by canoe eastwards seem to relate to an artificial clearing of the northern branch of the lagoon, the Toho, which was only temporarily effective.

It is not strictly accurate to describe Ouidah as a ‘lagoonside port’, any more than as an ‘Atlantic port’,⁶⁷ since it is in fact situated over 3 km north of the permanently navigable waterway. Although the width of the lagoon varied both seasonally and from year to year, and in times of very heavy rainfall (as happened, for example, in 1686) the intervening land might be flooded, permitting canoes to carry goods over part of the distance to the town, this was clearly exceptional.⁶⁸ Nevertheless, Ouidah was sufficiently close to the lagoon to be able to benefit from the canoe-borne traffic along it: in the 1680s, for example, an English trader at Ouidah noted that trade could be done with Little Popo to the west for slaves, locally made beads and corn, communication being ‘by the river’, i.e. the lagoon.⁶⁹ Slaves were also supplied to Ouidah from Offra to the east, although it is not specified that these were brought by canoe.⁷⁰ In the nineteenth century, communication between Ouidah and Godomey was more usually on foot, although the journey was sometimes made by canoe along the lagoon.⁷¹ Beyond Godomey, Lake Nokoué and the lagoons further east provided a continuous navigable waterway, which was regularly used for trade. For example, in the mid-seventeenth century salt manufactured in the coastal area of Allada was being taken by canoe to Lagos, and thereby to ‘Lukumi’, or the Yoruba interior, from which locally made cloth was brought in exchange; and later Yoruba cloth was also taken further west to Ouidah.⁷² Very probably, such trade had also existed prior to the arrival of the Europeans, although its scale was certainly increased by their presence, Europeans purchasing African-made cloth and beads (both for resale on the Gold Coast to the west) and corn (for the provisioning of slave ships), brought along the lagoon, as well as slaves.⁷³

The importance of trade along the lagoon also afforded opportunities and temptations for piracy, although here again this would presumably have become more profitable after the initiation of the European maritime trade and the stimulus it gave to the lagoon traffic. Burton in the 1860s was told that Ouidah had been ‘originally a den of water-thieves and pirates’.⁷⁴ This is corroborated by a contemporary account of the mid-seventeenth century, relating to ‘Foulaen’, which as noted earlier seems to be identical with Ouidah, which reports that it was accustomed to send ‘robbers’ to raid the coastal towns of Allada to the east.⁷⁵

In contrast to the lagoon, the sea beyond it can have played only a marginal material (as opposed to religious) role in the life of early Ouidah. Unlike on the Gold Coast to the west, the inhabitants of the Slave Coast did not venture onto the sea prior to the arrival of the European traders. This was evidently due, on the one hand, to the greater difficulty of navigation on the sea in this region, due to the heavy surf and dangerous sand bars noted earlier, and, on the other, to the availability of the much easier facility for fishing and canoe-borne communication afforded by the lagoon. Indeed, the local people largely continued this avoidance of the sea even after the initiation of the European maritime trade; as will be seen hereafter, European ships trading at Ouidah had to bring both canoes and canoemen with them from the Gold Coast to the west, in order to communicate with the shore. Even after this introduction of seagoing canoes, little or no fishing in the sea was done at Ouidah, the canoes being employed only in servicing the overseas trade.⁷⁶ It has sometimes been suggested that African merchants from the Gold Coast may have conducted a canoe-borne maritime trade with the Slave Coast even before the arrival of the Europeans; but there is no evidence for such a trade before the mid-seventeenth century, and it is more likely that such contacts were initiated by the Europeans, and only subsequently imitated by the Gold Coast merchants.⁷⁷ Earlier, interest in the sea was probably restricted to foraging along the shore, for crabs, as recalled in the traditional story of Kpate’s meeting with the first European visitors cited above.

Both the role and the importance of Ouidah were, however, transformed by the arrival of the Europeans and the initiation of maritime trade, which until the mid-nineteenth century was primarily in slaves. The Portuguese first explored along the Bight of Benin in 1472, and a regular trade began during the second half of the sixteenth century; from the 1630s the Portuguese monopoly of this trade was challenged by the Dutch, joined in the 1640s by the English and in the 1670s by the French. European trade was initially located at Grand-Popo, west of Ouidah; but by the beginning of the seventeenth century had shifted east to the kingdom of Allada, where the principal centre of the trade, and the site of the European factories, was initially at Offra.⁷⁸ In 1671, however, the French West Indies Company transferred its factory from Offra to Ouidah, initiating the latter’s rise to become the pre-eminent slave port within the region.

The French establishment at Ouidah in 1671 is often assumed to mark the beginnings of European trade there.⁷⁹ However, it should be noted that King Hufon of Hueda in 1720 said that the Portuguese had been the first Europeans to trade in his kingdom.⁸⁰ Local tradition in Ouidah also generally identifies the first European traders welcomed by Kpate and Kpase as Portuguese, while the French are said to have arrived only subsequently. According to one (no doubt apocryphal) story, the first Portuguese in Ouidah buried an inscribed stone to commemorate their visit; and when they returned, to find the French now in residence, they were able to disinter it to establish their claim to precedence.⁸¹ The dates of 1580 or 1548 assigned locally for the arrival of the first Portuguese, as noted earlier, are merely speculative, but it must have occurred sometime during the late sixteenth or early seventeenth century. There are, indeed, a couple of earlier references, in the 1620s, to Portuguese trade at ‘Fulao’, which as noted earlier was probably an alternative name for Ouidah.⁸² But it is likely that any such early contact was not sustained, and that therefore the French establishment in 1671 remains significant as marking the beginnings of continuous European trade at Ouidah.

The French move from Allada to Hueda was soon followed by the other principal European nations engaged in the trade, the English and Portuguese in the 1680s and the Dutch in the 1690s, leaving Ouidah as the dominant ‘port’ in the region by the end of the seventeenth century. The slave trade through Ouidah had reached a volume of probably around 10,000 slaves per year by the 1690s, and attained its all-time peak in the years 1700–13, when probably around 15,000 slaves annually were passing through the town;⁸³ at this period, indeed, Ouidah may have been accounting for around half of all trans-Atlantic exports of African slaves.⁸⁴

The European forts

As far as the record goes, the first permanent European trading post in Ouidah (or indeed, anywhere in the Hueda kingdom) was established by Henri Carolof (Heinrich Caerlof), a German in the service of the French West Indies Company, as has been seen in 1671.⁸⁵ One version of local tradition claims that even before this, in 1623, a Frenchman called Nicolas Olivier had settled in Ouidah, and founded the quarter of the town called Ganvè, to the west of the site of the French fort.⁸⁶ But this story is certainly spurious: contemporary sources do not support the suggestion of any French activity at Ouidah before the 1670s. The Olivier (or d’Oliveira) family of Ganvè is in fact descended from a man who was director of the French fort in Ouidah at a much later date (1775–86); the attempt of the d’Oliveiras to claim priority of settlement may derive from rivalries for the leadership of the ‘French’ community in Ouidah in the nineteenth century.

The French factory was abandoned when it was destroyed in a local war in 1692; a French captain who visited the Hueda kingdom in 1701 requested its re-establishment, but the king was initially willing only to allow the French a lodge in his capital Savi.⁸⁷ However, in 1704 a visiting French expedition secured permission not only for the re-establishment of a lodge nearer the coast, but also for its fortification; the Hueda king supplied over 400 men and women for the construction.⁸⁸ It subsequently became known officially as ‘Fort Saint-Louis de Gregoy [i.e. Glehue]’. The local traditions nowadays current of the establishment of the French fort, which attribute it to the reign of the fourth king of Hueda, Ayohuan (or Hayehoin), clearly relate to this subsequent refoundation, rather than to the original settlement in 1671, since Ayohuan is evidently to be identified with the king known to contemporary Europeans as ‘Aisan’ or ‘Amar’, who reigned in 1703–8.⁸⁹ The French fort was owned by a succession of trading companies until 1767, when it passed into the authority of the French crown. It was abandoned in 1797, but reoccupied by private French merchants (of the firm of Régis of Marseille) from 1842. The building, however, no longer survives, having been demolished in 1908; its site is now a public square, which is still however called ‘La Place du Fort français’.⁹⁰

Assuming that the fort built in 1704 was on the same site as the earlier French factory, at the time of its original foundation in 1671 the latter must have been physically separate from the indigenous settlement of Tové, since the later fort was on the opposite (west) side of the town from Tové, the intervening space being occupied by the quarter of the English factory, Sogbadji, which was established later. This is consistent with the account of Barbot in 1682, which describes the French and English factories at Ouidah as situated ‘near to’, rather than actually in, the indigenous village.⁹¹ This arrangement was seemingly also paralleled in the kingdom of Allada earlier, where the original site of the European factories, Offra, was distinct from although close to the town of Jakin, although Offra eventually developed into a substantial and autonomous indigenous settlement also.⁹² This suggests the policy widely attested elsewhere in West Africa of segregating foreign traders in distinct quarters, on the outskirts of the indigenous towns. This practice is most familiar from the colonial period, when in Nigeria, for example, southern immigrants in northern cities were regularly segregated into ‘new towns’ (sabon gari), while in the south northern merchants settled in separate quarters called ‘zongos’.⁹³ (There is, in fact, a Zongo quarter in Ouidah itself, on the north-east of the town, which dates from the period of French colonial rule after 1892.) But in this colonial practice clearly followed indigenous pre-colonial precedents: in towns in the Borgu region in the north of modern Bénin, inland from Dahomey, for example, foreign Muslim merchants likewise formed their own quarters, such as the Maro quarter of Nikki and the Wangara quarter of Djougou.⁹⁴ This arrangement probably also accounts for the location of the principal market in Ouidah, called Zobé, which is still today situated south-west of Tové quarter, and between it and the quarters of the former English and French forts to the west.⁹⁵

The second of the European factories to be established was the English. The Royal African Company, which held a legal monopoly of English trade in West Africa at this time, first projected a factory at Ouidah in April 1681; but this was abortive, the factor left there being recalled four months later to take over the company’s factory at Offra to the east.⁹⁶ Later in 1681 a factory was established in Ouidah by an English trader called Petley Wyburne, who was not an agent of the Royal African Company but an ‘interloper’, that is, a trader operating independently of the company and in breach of its monopoly, and this was maintained until Wyburne was forcibly removed by agents of the Royal African Company in 1686.⁹⁷ The company itself had meanwhile established a factory in the Hueda kingdom, at a second attempt, in July 1682; but this was located not at Ouidah, but at the capital Savi.⁹⁸ In April and May 1684, however, this factory suffered two serious fires, which evidently effectively destroyed it, since later in 1684 the local chief factor reported that he was ‘busied about building a house’;⁹⁹ and the new factory now built was evidently situated at Ouidah.¹⁰⁰ The English factory was situated east of the French factory (at least, east of where the later French fort was established in 1704), between it and the indigenous settlement of Tové. It was fortified with earthworks, mounted with cannon, for defence against its French neighbour, in 1692, and a moat was added in 1694;¹⁰¹ it was later known as ‘William’s Fort, Whydah’, alluding to King William III (1689–1703), the English monarch in whose reign it was fortified.

From the Royal African Company William’s Fort passed into the possession of the Company of Merchants Trading to Africa, which replaced it in 1752, but it was abandoned in 1812, following the legal abolition of the British slave trade. It was reoccupied by a private British merchant, Thomas Hutton, operating from the Gold Coast to the west, from 1838, and was later occupied by a British vice-consulate (1851–2), by a British Methodist mission (1856–67), and again by a different British trading firm (F. & A. Swanzy, also of the Gold Coast) in the 1870s. It was sold off to a German firm (C. Goedelt, of Hamburg) in the 1880s, but was confiscated as enemy property by the French colonial authorities in the First World War and passed back into the hands of another British firm, John Walkden of Manchester, who remained in occupation until 1963 and with whose name it is still locally associated. Redevelopment had destroyed its appearance as a fortification by 1890, when the building occupied by Goedelt was described as ‘an ordinary house’; the moat was filled in in 1908.¹⁰² The only material remains of the earlier fort visible today are a few cannon scattered around its courtyard.¹⁰³ In local usage, the area nevertheless remains ‘Le Fort anglais’.

The third and last of the European forts in Ouidah was the Portuguese. Some accounts date the foundation of the Portuguese fort to 1680;¹⁰⁴ but although a Portuguese factory was indeed established in the Hueda kingdom around this time, it appears that this was at the capital Savi rather than at Ouidah, and in any case was ephemeral, or at least not continuously occupied.¹⁰⁵ The Portuguese fort in Ouidah was in fact built in 1721;¹⁰⁶ it was later known as ‘Fort São João Baptista de Ajudá [= Hueda]’. It was situated east again from the English factory, and adjoining Tové quarter on the south; a contemporary account of its establishment indicates that, unlike the earlier factories, it was constructed within an already built-up part of the town, which had to be demolished in order to clear the site, the construction employing over 500 persons for 30 days.¹⁰⁷ Unlike the other two forts, the Portuguese was from the first a possession of the Portuguese crown, and under the immediate authority of the viceroy of Brazil at Salvador, Bahia; and later (after the Brazilian capital was removed to Rio de Janeiro in 1763) under the provincial governor of Bahia. It was abandoned after the legal abolition of the Portuguese slave trade north of the equator in 1815, but the Portuguese claim to it was maintained. It was reoccupied by the Portuguese government in 1844, this time administered from the local Portuguese headquarters on the island of São Tomé, off the West African coast. This renewed Portuguese presence was at first tenuous and intermittent, and possession of the fort was briefly usurped by Roman Catholic missionaries of the French Société des Missions Africaines in 1861–5. But it was definitively reoccupied by Portugal in 1865, and remained an anomalous Portuguese enclave within the French colony of Dahomey throughout the colonial period, its evacuation being forced by the newly independent Republic of Dahomey only in 1961. Alone of the three European forts the Portuguese retains its character as a fortification, though the present layout of the buildings appears to date from the reoccupation of 1865 rather than from the original period of occupation in the eighteenth century.¹⁰⁸ It now houses the Historical Museum of Ouidah.

Local tradition speaks of the existence of Dutch and Danish forts also in Ouidah, and even indicates their supposed sites.¹⁰⁹ Memory of a Dutch fort that had allegedly existed earlier is already attested in Ouidah in the 1860s, when it was said to have been on the site occupied since the 1820s by the Brazilian slave-trader Francisco Felix de Souza, which remains today the de Souza family compound.¹¹⁰ But the existence of such a Dutch fort is not corroborated by earlier contemporary evidence. The Dutch West India Company did contemplate establishing a factory in Ouidah, after their existing factory at Offra was destroyed in the war of 1692, when the Hueda authorities offered them the factory formerly occupied by the English interloper Wyburne.¹¹¹ Although the Dutch factors were in the event evacuated, some of the company’s African employees apparently remained behind; in 1694 it was noted that there was a settlement of ‘Mina’ people (i.e. from Elmina, the Dutch headquarters on the Gold Coast), half a mile from the English factory at Ouidah, who assisted Dutch ships trading there.¹¹² But it does not appear that this establishment was maintained. A Dutch fort is mentioned at Ouidah, alongside the French and English, in one source of the early eighteenth century; but this seems to be a simply a mistake.¹¹³ The West India Company did maintain a factory (which was not fortified) in the Hueda kingdom between 1703 and 1727, but this was situated at the capital Savi, rather than Ouidah.¹¹⁴ Although the company’s local chief factor in 1726 obtained permission to build a lodge ‘where the other nations have their forts’, i.e. in Ouidah, this was not in fact carried out;¹¹⁵ and after the destruction of Savi, including the European factories there, in the Dahomian conquest of 1727, the Dutch company no longer maintained any establishment in the area. Although it is of course possible that some later individual Dutch trader maintained a factory in Ouidah, this can only have been ephemeral, and the authenticity of the ‘traditional’ site is suspect.

The idea of a Danish fort in Ouidah seems to have arisen from a misreading of the same early eighteenth-century account that wrongly mentioned a Dutch fort there.¹¹⁶ Contemporary sources, again, do not corroborate this; the factor of a Danish ship that traded at Ouidah in 1784–5 lodged in the English fort.¹¹⁷ Recent tradition asserts that the Danish fort was on the site occupied in the late nineteenth century by the French firm of Cyprien Fabre, immediately east of the English fort, but this is probably a confusion; contemporary evidence suggests that the former occupant of this building was the British trader Hutton, who after relinquishing the English fort to the British vice-consulate in 1852 occupied premises east of the fort, which after his death (in 1856) passed into the possession of a Spanish merchant, and by the mid-1860s into the hands of the Dahomian crown.¹¹⁸

The European forts were distinctive within Ouidah by being built, in part, in two storeys, being consequently known locally as singbo (or singbome), ‘great houses’.¹¹⁹ Nevertheless, it should be stressed that the Ouidah forts were structures of much less military strength than the better-known examples on the Gold Coast. Unlike the latter, those at Ouidah were built in local materials – in mud rather than in brick or stone, which left them subject to rapid dilapidation if not regularly maintained, and with thatched rather than tiled roofs, which made them more vulnerable to fire. Their relative weakness is demonstrated by the fact that the Dahomians were able to capture and/or destroy forts in Ouidah on three occasions, the Portuguese in 1727, the French in 1728 and the Portuguese again in 1743; on the latter two occasions at least, the destruction of the forts was due to the buildings catching fire and causing the explosion of stores of gunpowder. Moreover, the Ouidah forts were located a considerable distance from the sea, and therefore their cannon could not, like those on the Gold Coast, command the landing-places for their own supplies; in consequence, as Europeans explicitly recognized, even if they could defeat direct attack they could be starved into surrender.¹²⁰ In the early eighteenth century, both the English and the Dutch pressed for permission to build forts at the seaside, but the Hueda king, Hufon, refused, precisely because he was aware of the power that English and Dutch forts exercised on the Gold Coast, and feared that the erection of such a fort at the shore would make the Europeans ‘masters of his port’. When the trade at Ouidah was disrupted by the activities of European pirates in 1719–20, Hufon did authorize the construction of a stone fort on the beach by the French, in order to protect ships trading in his dominions, but, again, the idea was not pursued.¹²¹

The forts in Ouidah operated as secure places of storage for goods and slaves, rather than exercising any serious military power over the local community. The concept of an ‘enclave-entrepôt’, which has been applied to coastal towns on the Gold Coast in which Europeans settled, such as Elmina, does not seem applicable to Ouidah, which was in no sense an enclave of European authority, or even of their informal predominating influence.¹²² In Ouidah, there was never any question that the European establishments were in the final analysis subject to local control, rather than representing independent centres of European power.¹²³ This was explicitly expressed in the policy of the Hueda kings of forbidding fighting among Europeans in the kingdom, even when their nations were at war in Europe, which was formalized in 1703, when the king obliged the local agents of the Dutch, English and French companies to sign a treaty prohibiting hostilities in the Ouidah roadstead, or within sight of the shore, on pain of payment of damages to the value of eight slaves;¹²⁴ one chief of the English fort was deported in 1714, after a fracas with his French counterpart, which was deemed to be a breach of this treaty.¹²⁵

Each of the three European forts in Ouidah became the centre of a quarter of the town, occupied by persons in the service of the forts. These were commonly called the French, English and Portuguese quarters (or, in contemporary European sources of the eighteenth and nineteenth centuries, ‘towns’, ‘villages’ or ‘camps’), or in local parlance Zojage-ko, Glensi-ko and Aguda-ko (ko meaning ‘quarter’).¹²⁶ Of these names, ‘Glensi’ is merely a local version of the name ‘English’, but the others call for further comment. ‘Zojage’ is explained as meaning ‘Fire has come to earth’, which is said to have been an exclamation of wonder uttered by Zingbo, the companion of Kpate, upon sight of the first Europeans to land at Ouidah (alluding to their ‘red’ skin colour, which was thought to resemble fire).¹²⁷ This story implies that it was originally applied to Europeans in general (these first visitors being in fact Portuguese), but it was subsequently restricted to the French in particular; when first documented in a contemporary source, in a vocabulary collected in Brazil among African slaves from the Dahomey area, it already designated the French specifically.¹²⁸ ‘Aguda’ is a term of uncertain etymology, which is commonly understood nowadays to mean ‘Brazilian’, and when first attested, in the same vocabulary, was applied specifically to Bahia, as opposed to Portugal; but in West Africa in the nineteenth century its reference was national rather than geographical, applied to ‘Portuguese’ in general, including Brazilians, rather than to Brazilians as distinct from Portuguese.¹²⁹

In recent times, the three ‘European’ quarters of Ouidah have more commonly been known by other names, that of the French fort being called Ahouandjigo, that of the English Sogbadji, and the Portuguese Docomè, these names being first attested in the contemporary record in the 1860s.¹³⁰ The name ‘Ahouandjigo’ is translated as ‘where war cannot come’, and is usually explained by tradition as recording an undertaking by the Hueda king Ayohuan not to make war on the French fort there;¹³¹ it seems more likely, however, that it alluded to the policy of the Hueda kings of forbidding fighting among Europeans at Ouidah, which was reaffirmed in the treaty signed in 1703, the year before the establishment of the French fort. That of ‘Sogbadji’ for the English quarter refers to So, the vodun of thunder, meaning ‘So’s enclosure’; it is said to have been the place where the bodies of persons killed by lightning were taken, from which they could be redeemed for burial only on payment of a fine.¹³² The etymology of the name ‘Docomè’, ‘Do quarter’, is uncertain.¹³³

Although their origins are understood at one level to be connected with the establishment of the European forts, it is noteworthy that all three of these quarters celebrate indigenous Hueda persons rather than Europeans as their actual founders; by implication, these were already settled locally before the arrival of the various European groups whom they welcomed. Ahouandjigo claims to have been founded by a prince of the Hueda royal family called Agbamu (in French ‘Agbamou’); a prominent family in the quarter, that of Agbo, claims descent from him.¹³⁴ Docomè is said to have been founded by a man called Ahohunbakla (‘Ahohounbacla’), who belonged to the same family as Kpate, the hero who welcomed the first Europeans, who is also sometimes said to have belonged to Hueda royalty.¹³⁵ Both Agbamu and Ahohunbakla are also said to have survived to lead their quarters in resistance to the Dahomian conquest of the town in the 1720s–40s. The details of these traditions are suspect. The claim in both cases that the founder of the ward was also its leader against the Dahomians, although chronologically possible (at any rate, if Agbamu is assumed to have been associated with the building of the French fort in 1704, rather than the original establishment in 1671), may be doubted; it seems more likely that Agbamu and Ahohunbakla are composite or symbolic figures, into whose careers as recorded in the traditions events from different epochs have been telescoped. Indeed, as will be seen in the next chapter, it is clear that the historical Agbamu cannot have been either the founder of Ahouandjigo or its leader against the Dahomians, since he was in fact a king of the Hueda in exile, two generations after the Dahomian conquest. It may also be suspected that the name ‘Ahohunbakla’ is a variant or corruption of that of Agbangla, the Hueda king who died in 1703 and who is said by tradition to be buried in Docomè quarter.¹³⁶ The appropriation of such founding ancestors from among Hueda royalty is, however, significant as a claim of indigenous legitimacy, which was probably asserted against Dahomian overlordship, as well as and probably more than against European primacy. In Sogbadji, however, the claimed indigenous founder is of non-royal Hueda stock. The founder is usually named as Zossoungbo, said to have been a hammock-bearer to the Hueda king at Savi, whose descendants claim the hereditary headship of the quarter; but another Hueda family in Sogbadji, that of Déhoué, disputes priority of settlement with the Zossoungbos.¹³⁷

The personnel of the ‘European’ forts was in fact predominantly African. The English fort in the 1700s, for example, was manned by only 20 white men, with 100 ‘gromettoes’, or African slaves.¹³⁸ The ‘European’ quarters also included free Africans who were either in place before the establishment of the European factories or were attracted into their service subsequently. To the present day, these quarters are largely occupied by descendants of persons associated with the forts, including some Europeans who fathered families by local women, but mainly free African employees and slaves. Some of these families claim to be descended from persons employed in the forts before the Dahomian conquest in 1727, although most of those in place today seem to have arrived later, in the period of Dahomian rule. In addition to indigenous Hueda (and Hula) families, the populations of the ‘European’ quarters also included a large non-indigenous African element. Many of the fort slaves employed in Ouidah were from the Gold Coast to the west: in 1694 it was noted that ‘most’ of the slaves employed in the English fort were ‘Gold Coast negroes’, who were considered superior soldiers to the local Huedas; likewise in 1716 the slaves of the English fort at Ouidah (and also of the Dutch factory at Savi) were ‘almost all inhabitants of the Gold Coast, or Minas’.¹³⁹ Conversely, it may be noted, slaves purchased in Ouidah and Allada were commonly employed by Europeans in their factories on the Gold Coast;¹⁴⁰ this being reflected to the present day in the existence of ‘Alata [i.e. Allada]’ quarters in the Dutch and English sections of Accra.¹⁴¹ The logic of employing such ‘foreign’ Africans was, explicitly, that such outsiders were less liable to run away than slaves recruited locally, whose homelands were more accessible. Other fort slaves employed in Ouidah were imported from the interior; in effect, a portion of those purchased for export through the town was retained for local use. The slaves of the French fort in the eighteenth century were generally called ‘Acqueras’, a usage already established by the 1710s, and this was in origin the name of a specific ethnic group, reported to be located in the far interior, from which presumably many of the French fort slaves were derived.¹⁴² In 1723 the French fort reported that it had purchased ‘Chamba [Tchamba]’ slaves, this being another ethnicity in the interior (in northern Togo).¹⁴³

Other African foreigners settled in Ouidah as free immigrants, attracted there by the opportunities for employment in the European trade. The most prominently visible category among such incomers were canoemen from the Gold Coast. As noted earlier, the indigenous people of Ouidah, although using canoes on the inland lagoons, had no tradition of navigation on the sea, whereas on the Gold Coast the inhabitants had employed canoes for sea-fishing and coastwise communication even before the arrival of the Europeans. Since at Ouidah (and elsewhere on the Slave Coast) European ships were unable to approach close to the shore (owing to the dangerous bars and surf), they regularly bought canoes and hired crews of canoemen on the Gold Coast on their way down the coast, in order to land goods and embark slaves.¹⁴⁴ During the second half of the seventeenth century, indigenous Gold Coast merchants also began to travel to the Bight of Benin by canoe, to trade independently of (and in competition with) the Europeans, for cloth and other goods for resale on the Gold Coast: in 1688, for example, it was noted of the trade in African cloth at Ouidah that ‘the Blacks come with canoes there to trade in them, and carry them off continuously’.¹⁴⁵

Most of the canoemen who came to Ouidah from the Gold Coast returned home on completion of their contracts, but some settled permanently in Ouidah. At the end of the seventeenth century, Cape Coast, the English headquarters on the Gold Coast, was said to be visibly depopulated because of the recruitment from there of canoemen by English ships trading at Ouidah, ‘after which they liking the place, live there, and seldom remember to come home again’.¹⁴⁶ Some of these immigrant canoemen entered the service of the European factories in Ouidah on a long-term basis, as free employees, while other canoemen were recruited as pawns (bound to work while paying off debts) or slaves. In the 1710s, it was noted that the English and Dutch factories enjoyed the services of canoemen recruited respectively from Cape Coast and Elmina, whereas the French, having no Gold Coast establishments of their own, were at a disadvantage in this respect, and French ships had to hire canoemen on their voyage down the coast.¹⁴⁷ By the mid-eighteenth century, however, the French fort also had its own corps of canoemen, who in this case were slaves.¹⁴⁸

The cosmopolitan character of Ouidah, arising from its coastwise connections with the Gold Coast, is illustrated by an incident in 1686, when one Gold Coast man, described as from Kormantin but ‘an ancient inhabitant here’, was murdered by another, from Elmina, the latter having been sent to collect a debt owed by the first man to a third party in Elmina.¹⁴⁹ Some of these Gold Coast immigrants became prominent people in the local system: in the 1690s the official who served as interpreter to the English factory in Ouidah, who was also a substantial trader in slaves, called ‘Captain Tom’, was in origin from the Gold Coast.¹⁵⁰ This Gold Coast element in the population is reflected in the currency of local versions of the personal names used in the Akan languages of the Gold Coast which allude to the day of the week on which a person was born: as for example, Kwadwo, Kwamina, Kwaku and Kofi (given to boys born respectively on Monday, Tuesday, Wednesday and Friday), which occur in Ouidah in the forms Codjo, Comlan, Cocou and Coffi.

Early Ouidah

By the 1720s, Ouidah was thus made up of the combination of Tové, the original Hueda village, with the three quarters associated with the European forts, Ahouandjigo, Sogbadji and Docomè. The size of the settlement, along what was presumably its longest axis, from the French to the Portuguese fort, was only around 1 km in length. In comparison, the Hueda capital Savi was larger, being estimated in the 1720s to have a circuit of over 4 miles, or 6 km; while the Allada capital was said to have a circuit of 3–4 Dutch miles, that is 12–16 English miles, or around 18–24 km.¹⁵¹

The population of Ouidah at this time is a matter for speculation. The combined personnel of the European forts cannot have been more than a few hundred. In the 1710s the French fort had a total of 160 African slaves, including children as well as adults; by the 1770s this had grown only slightly to between 180–200, who were said to comprise 50 separate ‘families’, each living in its own ‘hut [caze]’ near the fort. The ‘European’ quarters also included free families whose members were employed by or provided services for the forts, who were perhaps roughly as numerous as the fort slaves; by c. 1789, when the French fort was reported to have 207 slaves, the total population of the French ‘village’, including free persons outside the fort, was thought to be nearly 500.¹⁵² The only scrap of evidence for the population of the settlement as a whole is an account of the establishment of the Portuguese fort in 1721, which refers to its location as being in a ‘quarter’ containing 300 households (‘hearths’), all the inhabitants of which were employed in the service of foreigners trading in the town.¹⁵³ This high figure seems likely to refer to the town as a whole rather than the two pre-existing ‘European’ quarters only, and suggests (on the analogy of 200 persons in 50 ‘families’ reported for the French quarter in the 1770s) a population of between 1,000–1,500; the addition of Docomè quarter with the construction of the Portuguese fort would have raised this figure, but the total population of the town is still unlikely to have reached as high as 2,000. In comparison, while no figures are available for Savi, the Allada capital in 1660 was thought to have 30,000 inhabitants.¹⁵⁴

In addition, there was a substantial transient population, in the form of African officials and merchants from Savi, as well as Europeans from visiting ships, and especially slaves in transit to embarkation from the shore. The total number of slaves annually passing through Ouidah, which peaked at around 15,000 in the early eighteenth century, was in fact substantially higher than that of the resident population. Although many of these slaves passed through the town rapidly, significant numbers might be held for some time locally, in the European factories: the English factory in 1687 was said to have space to lodge between 600 and 800 slaves.¹⁵⁵

In its spatial organization, Ouidah clearly differed radically from towns further inland that served as capitals of states, such as Savi and Allada (and later, Abomey), which were centred around the royal palace.¹⁵⁶ Ouidah was multi-centred, focused on the three European forts; in so far as it had a single centre, this was perhaps the Zobé market.¹⁵⁷ However, to what extent Ouidah yet formed a coherent community, rather than an assemblage of discrete settlements, is doubtful. In the early eighteenth century, the indigenous ‘village of Grégoué’ (i.e. Tové quarter) was still described as separate from the French and English forts, which were ‘a very short distance’ away.¹⁵⁸ The establishment of the Portuguese fort in 1721, immediately south of Tové and east of the English fort and Sogbadji quarter, produced greater contiguity of settlement, grouped around the market of Zobé; but the French fort with Ahouandjigo to the north-west remained physically distinct. In fact, it is not clear whether, within the Hueda kingdom, the town was administered as a unit or, perhaps more likely, the three European forts were individually responsible to the king of Hueda, and separately from the local indigenous authorities. There was a Hueda chief called ‘Prince Bibe’ or ‘Captain Bibe’, who is named along with the king as negotiating to permit the establishment of the French at Ouidah in 1671 and who in 1682 seems to have been residing at Ouidah, rather than at the capital Savi.¹⁵⁹ In the early eighteenth century, a list of Hueda chiefs who served as ‘governors’ of ‘provinces’ within the kingdom includes one entitled ‘Gregoué Zonto’, who was presumably the governor of Glehue;¹⁶⁰ and maybe this is the title which ‘Prince Bibe’ held.¹⁶¹ But whether he had overall administrative responsibility for the town, including the European forts, is doubtful; more probably, he was governor of the indigenous ‘village’ only.

The operation of the European trade in the Hueda kingdom gave rise to a number of new official positions. Most important was that of ‘Captain of the White Men’, as Europeans correctly translated the indigenous title Yovogan, or Yevogan, which is already attested in the 1680s,¹⁶² and from the 1690s was held by a man called ‘Carter’.¹⁶³ There was also an assistant to the Yovogan called ‘Agou’, and separate ‘captains’ for the European nations with factories in Hueda: the French (whose captain was called ‘Assou’), English (served in the 1690s by ‘Captain Tom’), Dutch and Portuguese.¹⁶⁴ But these officials who dealt with European traders did not, like their counterparts under Dahomian rule later, form a local administration for Ouidah; in fact, there is no evidence that they even resided there. In 1718 the Yovogan Carter was reported to be building a new house outside the capital Savi, where he was in consequence now expected to attend less regularly, but this residence was not at Ouidah; a later map of the Hueda kingdom shows Carter’s village situated to the east of Savi, while that of the ‘French captain’ Assou was even further from Ouidah, to the north-west of the capital.¹⁶⁵

Ouidah in this period did not engross the conduct of the European trade, since much of the business of European traders had to be transacted at the capital Savi. The emphasis in some modern accounts on the ‘separation of the political and commercial capitals’ as a feature of the organization of the slave trade in Hueda is in fact misconceived – or, more precisely, it incorrectly reads back into the period of the Hueda kingdom a distinction that emerged only after the Dahomian conquest.¹⁶⁶ When the English Royal African Company first sent a ship to trade in Hueda, in 1681, its chief factor ‘bought slaves at Sabba [Savi], the king’s town’, while his assistant was ordered to ‘Agriffie [Glehue], the lower town [i.e. Ouidah]’, where he presumably managed the landing of goods.¹⁶⁷ This pattern continued even when permanent factories were established in Ouidah. The French trader Barbot in 1682 thus noted that ‘it is with the king that you do the trade’, i.e. at the capital Savi, while the goods were ‘brought from the vessel to the lodge’ at Ouidah; and the chief of the English factory at Ouidah in 1685 reported that he ‘went up to the king’s town’ to buy slaves for an English ship trading there.¹⁶⁸ It has been suggested that, although this was the practice earlier, by the first decade of the eighteenth century the trade had been localized at Ouidah, rather than Savi; but there is no basis for this in the contemporary evidence.¹⁶⁹ In 1716, for example, it was still explicitly noted that slaves brought for sale were lodged in prisons ‘in the place where they are traded, this place is Xavier [Savi], the king keeps his residence there’.¹⁷⁰ At this period the Dutch West India Company, as noted earlier, actually maintained its factory at Savi rather than at Ouidah. By 1716 the English and French companies, in addition to their forts in Ouidah, also maintained lodges at Savi; in fact, the local French and English directors normally resided in Savi, leaving the forts at Ouidah under subordinate officers.¹⁷¹ When the Portuguese established their fort at Ouidah in 1721, they also maintained a subsidiary lodge in Savi.¹⁷² The factories at Ouidah served only as storehouses for goods and for slaves in transit to and from Savi; an English trader in 1694 noted that the factory at Ouidah ‘proved very beneficial to us, by housing our goods which came ashore late, and could not arrive at the king’s town (where I kept my warehouse) ere it was dark’, and also when slaves could not be embarked owing to bad weather.¹⁷³

European activities in Ouidah provided economic opportunities for the local inhabitants mainly in the form of employment in ancillary services, such as the supply of provisions and firewood, and especially as porters and canoemen. It has been argued that European trade in Africa, even at the height of the Atlantic slave trade in the late eighteenth century, was simply too small in scale, measured by per capita export earnings, to have had any major impact on indigenous societies.¹⁷⁴ While this may have some a priori plausibility for West Africa as a whole and for many particular societies in its interior, it is clearly not applicable to coastal communities such as Ouidah which were heavily involved in the trade. The value of slave exports through Ouidah by the end of the seventeenth century was enormous, in relation to the population of the town; 10,000 slaves per year in the 1690s, when the price of an adult male slave in the local currency of cowry shells was 10 ‘grand cabess [large heads]’ (40,000 shells), equivalent to £12.10s. (£12.50), would have represented (allowing for lower prices paid for women and children) the value of around 320 million cowries (£100,000), at a time when the wage for a porter per journey (in effect, per day) in Ouidah was 3 ‘tockies’, or 120 cowries (9 pre-decimal pence, £0.033/4).¹⁷⁵ But most of the income from this trade would have accrued to officials and merchants in the capital Savi (and beyond, to the officials and merchants of states in the interior from where many of the slaves were purchased), rather than to the inhabitants of Ouidah. It is difficult to estimate what sums would have been expended on goods and services in Ouidah itself, but some indication is provided by the statement of Barbot in 1682 that every ship had to pay the value of five to six slaves for the carriage of goods from the shore to the factory in Ouidah, and the same again for the canoes that landed goods and embarked slaves at the beach.¹⁷⁶ In the case of the canoemen, since many of these were hired from the Gold Coast rather than permanently resident in Ouidah, these earnings were partly, indeed perhaps predominantly, repatriated to the Gold Coast, rather than being expended or retained locally.¹⁷⁷ But, assuming 40 ships per year, and the price of a slave to be that of an adult male, payments for porters alone at this rate would have amounted to around 8.8 million cowries (£2,750) a year, representing wages for over 73,000 person-days, or, assuming that workers took one day of rest in each four-day ‘week’, continuous year-round employment for around 270 porters.

Porterage and other services for the European trade, of course, represented only one of the major sources of income for people in Ouidah, along with fishing, salt-making and agricultural production. The continuing importance of fishing, in particular, is attested by a European account of the town in the early eighteenth century, which observes that its wealth derived as much from the fact that its inhabitants were ‘all fishermen and canoemen’ as from the presence of European factories.¹⁷⁸ There is little basis on which to estimate the relative importance of the European trade in comparison with other sectors of the local economy. One stray figure recorded in the 1690s is that the revenue derived by the king of Hueda from a toll levied on fishing was the value of 100 slaves, presumably annually, from each of the two principal ‘rivers’ in the kingdom.¹⁷⁹ This was not far short of the revenue from the royal duty on slave exports, levied at 5 ‘galinas’ of cowries (1,000 shells) per slave, which would have yielded (on exports of 10,000 slaves per year) the value of 250 slaves annually, although the king’s total revenue from the slave trade, including ‘customs’ (in contemporary parlance, payments for permission to trade) and the proceeds of his own sales of slaves, was much higher than this, estimated in this period at around 30 slaves’ value per ship.¹⁸⁰ It is not specified at what rate the toll on fish was levied, but if it was comparable to that levied generally on sales in the kingdom, which was one-tenth,¹⁸¹ this would suggest a total output for the fishing industry of the value of 2,000 slaves annually, which was around a fifth of slave exports through Ouidah in the 1690s. It was also very substantially greater than the total paid for portering, as calculated earlier (and, if incomes from fishing were comparable to the wages paid to porters, would suggest that fishing employed around 2,400 persons). Although this figure for fish production is for the Hueda kingdom as a whole, rather than for Ouidah alone, any reasonable assumption of the share to be assigned to Ouidah would still leave fishing as a larger sector of its economy than portering.

The rise and fall of the Hueda kingdom

In a recent study, David Eltis suggested that, despite considerable research on the history of Hueda, ‘the question of why this small African state was so dominant in the slave trade still has no answer’.¹⁸² This seems an unduly pessimistic assessment of previous research. It is true that Ouidah had no obvious geographical advantage over other ports in the region, either in terms of coastal harbour facilities (of which, in common with its rivals, it had none) or of access to inland waterways (for which it was no better positioned than its competitors to west and east). As regards its geographical situation, the principal advantage possessed by Ouidah was its proximity to and accessibility from the powerful kingdoms in the immediate interior which were the principal suppliers of slaves to the coast, initially Allada (and, after 1727, Dahomey). This advantage, of course, was equally shared by Offra to the east, which in fact preceded Ouidah as the main outlet for Allada’s slave trade. The story of Offra’s displacement by Ouidah in the 1670s, however, is straightforward (and well-known). The initial diversion of trade from Offra to Ouidah in 1671 was due to a rebellion by Offra against Allada’s authority, which closed the paths between the two; and difficulties in relations between Allada and Offra recurred through the following two decades, culminating in the war in which Offra was destroyed in 1692, an event that decisively confirmed the commercial supremacy of Ouidah. In the early eighteenth century the kings of Allada sought to redivert trade away from Ouidah through a ‘port’ under their own control, now Jakin, which inherited the commercial role of Offra, and initially with some success. But this process was overtaken by the Dahomian conquest of both Allada and Hueda in the 1720s, which served to remove obstacles to the passage of trade to Ouidah; and, even more critically, by the subsequent Dahomian destruction of Jakin in 1732, which reconfirmed the concentration of trade at Ouidah.¹⁸³

A second factor in the shift from Offra to Ouidah is implied in a letter from the English Royal African Company to the king of Hueda in 1701, which states that the English had moved there because of the ‘ill treatment’ which they had received in Allada.¹⁸⁴ One aspect of the better treatment offered at Ouidah was that the Hueda kings accepted lower levels of ‘customs’ for permission to trade. In the 1680s the French at Ouidah were paying the value of 25 slaves per ship for customs, whereas at Offra they had paid 50; total charges, including the hire of porters and canoemen, came to between 32 and 35 slaves per ship at the former, but between 70 and 80 at the latter.¹⁸⁵ By the 1690s, the rate of customs at Ouidah had fallen further, to only 8 slaves, 6 to the king and 2 to the ‘caboceers’, or chiefs.¹⁸⁶ A second aspect was the problem of theft, and the failure of the Allada authorities to repress it; another observer in 1701 explained that no trade was done any longer at Allada, because of the reputation of its people for ‘cheating and stealing’.¹⁸⁷ Correspondence from the English company’s Offra factory in the early 1680s also alludes to problems in recovering debts.¹⁸⁸

To the extent that the greater attractiveness of Ouidah over Offra related to the policies pursued by their rulers, these were subject to change over time. By the 1710s the differential in customs between Ouidah and its rivals had been substantially reduced, although not entirely eliminated; the charges for permission to trade at Jakin, now the principal port of Allada, being only 12 slaves in all (6 to the king, 4 to the governor of Jakin, one each to two other chiefs), doubtless driven down by the need to compete with Ouidah.¹⁸⁹ By the 1690s also, Ouidah itself had acquired a reputation for the prevalence of petty theft, both by porters carrying goods up from the beach and from the European warehouses, the people of Ouidah being considered worse thieves than those of the Gold Coast.¹⁹⁰ There is no record, however, of problems at Ouidah over issues of credit: an account of trading in the 1690s stresses rather the meticulousness of the African merchants there in the prompt settlement of their debts.¹⁹¹

That the trade nevertheless remained concentrated at Ouidah was in part due, as noted earlier, to the political history of the region, as this affected relations between Allada (and later Dahomey) in the interior and Ouidah and Jakin at the coast. But in addition, it seems likely that Ouidah’s continued dominance reflected in large part the forces of inertia. It is noteworthy that the pattern of concentration of the slave trade at a single port within the Bight of Benin is paralleled in other regions of western Africa, as for example the dominance of Bonny in the Bight of Biafra; and indeed it also mirrors the situation in Europe, where a single port was likewise commonly dominant within the slave trade of each European nation, for example Liverpool in England and Nantes in France. As a recent analysis co-authored by Eltis himself has suggested, this pattern of concentration probably reflects, in both Europe and Africa, cost savings arising from access to market information in established ports.¹⁹² On the Bight of Benin itself, on the European side of the trade, existing investment in the forts in Ouidah also operated to discourage relocation elsewhere.

Karl Polanyi linked the concentration of the slave trade at Ouidah to its status as a ‘port of trade’, which was politically ‘neutral’ and militarily ‘weak’, and therefore acceptable to traders of the different nationalities involved.¹⁹³ This idea of a ‘neutral’ place of trade does, indeed, have considerable plausibility, not only in abstract theory but also in the documented history of this particular region: in the later eighteenth century, the lagoonside port of Abomey-Calavi, on the shore of Lake Nokoué, east of Ouidah, was described as a ‘neutral place’, which functioned as ‘a sort of free fair, where the different nations go to trade’.¹⁹⁴ It is doubtful, however, whether this analysis has any bearing upon the rise of Ouidah as an Atlantic port in the late seventeenth century. Ouidah was indeed maintained by the Hueda kings as a ‘neutral’ port as regards traders of different European nations, as decreed in the treaty of 1703 forbidding hostilities among Europeans, and this policy was clearly among the factors that attracted European traders there: as the director of the French fort noted in 1716, it was ‘the only place of neutrality where the vessels of every nation are bound to find a secure retreat in time of [intra-European] war’.¹⁹⁵ This, however, was a neutrality imposed upon Europeans by the strength of the Hueda state, rather than a reflection of the latter’s ‘weakness’; and Hueda was not in any sense recognized as a ‘neutral’ area by its African neighbours. Although Hueda was certainly a small and therefore relatively weak power in relation to the interior state of Allada, which was its principal supplier of slaves, its independence of Allada was sustained in the face of systematic opposition from the latter (including an actual military invasion in 1692), rather than accepted as serving its convenience.¹⁹⁶

Eltis also observes that ‘the question of why, with all the income from its Atlantic trade and a significant military capacity, Whydah was unable to maintain its independence or, indeed, become a major power has yet to be addressed’, and suggests that this might indicate that ‘the transatlantic slave trade was not the central event shaping African economic and political developments’.¹⁹⁷ This also seems an idiosyncratic reading of the literature. Hueda did, in fact, develop as a significant regional ‘power’ in the coastal area in the late seventeenth century, extending its authority at least briefly over Grand-Popo to the west and disputing control of Offra with Allada to the east;¹⁹⁸ and contemporary Europeans, at least, thought that its military power was enhanced by the wealth it derived from the Atlantic trade, although it is noteworthy that the factor they stressed was not Hueda’s ability to import European firearms but the financial resources that enabled it to hire mercenary soldiers from other African communities for its wars.¹⁹⁹ In relation to the dominant hinterland state of Allada, however, its potential power was limited by sheer demographic weight: it was estimated (probably with some exaggeration) that Allada had twenty times the population of Hueda, a disproportion beyond what could be offset by greater financial resources.²⁰⁰ Also, Hueda’s potential power was compromised in the early eighteenth century by chronic internal divisions, which may well have been exacerbated by competition over revenues from the Atlantic trade, although this was certainly not their only cause.²⁰¹

The ultimate fall of Hueda to conquest by Dahomey in 1727 on the face of it also presents no mystery. In addition to Hueda’s persisting internal divisions, which critically undermined its efforts at self-defence on this occasion, the Dahomian forces were decisively superior qualitatively, Dahomey having developed both a superior military organization and a more systematic military ethos than any other state in the region. This does not, however, demonstrate the irrelevance of the Atlantic slave trade to questions of military power, since the militarization of Dahomey was itself a consequence of the impact of that trade, Dahomey having been a major supplier of slaves for sale at the coast since at least the 1680s and its forces being by the 1720s equipped with imported European firearms obtained in exchange for such slave exports. The imbalance of military power between Hueda and Dahomey reflected the division of labour that operated within the African section of the slave trade, in which coastal communities such as Hueda operated mainly as middlemen, while the actual process of violent enslavement was left to military states in the interior such as Dahomey.²⁰² The Dahomian conquest of Hueda is thus an illustration, rather than a refutation, of the importance of the Atlantic slave trade in shaping local economic and political developments.

Notes

1. Finn Fuglestad, ‘La questionnement du “port” de Ouidah’, in Oystein Rian et al. (eds), Revolusjon og Resonnement (Oslo, 1995), 125–36.

2. For the history of the Hueda kingdom, see esp. Robin Law, ‘“The common people were divided”: monarchy, aristocracy and political factionalism in the kingdom of Whydah, 1671–1727’, IJAHS, 23 (1990), 201–29; Gilles Raoul Soglo, ‘Les Xweda: de la formation du royaume de Sayi (Saxe) à la dispersion, XVe–XVIIIe siècle’ (Mémoire de maîtrise, UNB, 1994/5).

3. Ewe and Adja are properly the names of particular subgroups of the linguistic family (in eastern Ghana and Togo), while Djedji derives from the name given to speakers of these languages in Brazil; ‘Gbe’ is a neologism, derived from the word for ‘tongue’ (and hence ‘language’) in these languages.

4. Basilio de Zamora, ‘Cosmographia, o descripcion del mundo’ (MS of 1675, in Bibliotheca Publica do Estado, Toledo, Collecçion de MSS Bornon-Lorenzo, no. 244), 47; Joseph de Naxara, Espejo mistico, en que el hombre interior se mira prácticamente illustrado (Madrid, 1672), 278.

5. As related retrospectively (1688) by Jean Barbot: Paul Hair et al. (eds), Barbot on Guinea: The Writings of Jean Barbot on West Africa, 1678–1712 (London, 1992), ii, 635–6.

6. This Hueda successor-state has been little studied; but see Soglo, ‘Les Xweda’, 70–78.

7. Robin Law (ed.), The English in West Africa 1681–83: The Local Correspondence of the Royal African Company of England 1681–99, Part 1 (London, 1997: hereafter cited as English in West Africa, i), no.476: John Thorne, Glehue, 24 May 1681.

8. The suggestion of Burton, Mission, i, 61–2, that the name ‘Glehue’ was given to the town only after the Dahomian conquest in 1727 is clearly incorrect.

9. Barbot, On Guinea, ii, 635.

10. Burton, Mission, i, 108.

11. ‘Fulao’, e.g. in Alonso de Sandoval, Naturaleza, policia sagrada i profana, costumbres i ritos, disciplina i catechismo evangelico de todos Etiopes (Seville, 1627), 51; ‘Foulaen’, in Olfert Dapper, Naukeurige Beschrijvinge der Afrikaensche Gewesten (2nd edn, Amsterdam, 1676), 2nd pagination, 115.

12. For the Hula, see esp. A. Félix Iroko, Les Hula du XIVe au XIXe siècle (Cotonou, 2001), which concentrates on the original Hula homeland to the west. For traditions of Hula migrations to settle at Godomey and other places to the east, see Thomas Mouléro, ‘Histoire et légendes des Djêkens’, ED, ns, 3 (1964), 51–76.

13. Jean-Baptiste Labat, Voyage du Chevalier des Marchais en Guinée, isles voisines et à Cayenne (2nd edn, Amsterdam, 1731), ii, 34. Here as often, the published version of this work includes material not in the original manuscript: ‘Journal du voiage de Guinée et Cayenne, par le Chevalier des Marchais’ (Bibliothèque Nationale, Paris: fonds français, 24223).

14. Fieldwork, Glehue Daho compound, 3 Dec. 2001; Sinou & Agbo, Ouidah, 225.

15. Described in Sinou & Agbo, Ouidah, 223.

16. Brue, ‘Voyage fait en 1843, dans le royaume de Dahomey’, RC, 7 (1845), 55 (giving the name as ‘Passi’).

17. Reynier, ‘Ouidah’, 47–8.

18. Ibid., 47; but see, for example, Agbo, Histoire, 203, who describes the claim as ‘hazardous’.

19. This version first in Gavoy, ‘Note historique’, 48. But the other early recension of local tradition, by Reynier, ‘Ouidah’, 47, is vaguer: Kpase merely ‘belonged to the Pedah [Hueda] family of which the head was the King of Savi’. The earliest recorded reference to Kpase, in the 1840s, presents him as a purely local figure, ‘cabocir [chief] of a small hamlet in the vicinity of Grégoué [Glehue]’: Brue, ‘Voyage’, 55.

20. So Gavoy, ‘Note historique’, 52, and later sources deriving from him, which list only five kings down to and including Hufon (1708–27). However, other versions of the Hueda king list include several additional names: one lists 13 kings of whom Kpase is the eighth, another 14 with Kpase the fourth; for discussion, see Soglo, ‘Les Xweda’, 47–51. Some of the additional kings listed (Yé, Amiton) appear in fact to be persons who ruled over sections of the Hueda in exile after the Dahomian conquest of the kingdom.

21. Mouléro ‘Histoire’, 43–4; also Reynier, ‘Ouidah’, 51–2.

22. Hufon’s attack on Kposi is linked by tradition to his war against King Agaja of Dahomey (in 1727), but accounts differ in detail: Mouléro says that Hufon attacked Kposi because he refused to assist him against Agaja, but Reynier says that Hufon’s attack on Kposi came first, and Kposi incited Agaja to attack Hufon in revenge.

23. Barbot, On Guinea, ii, 636.

24. Robin Law, The Kingdom of Allada (Leiden, 1997), 42.

25. E.g. Gavoy, ‘Note historique’, 48–9; Reynier, ‘Ouidah’, 38.

26. Burton, Mission, i, 146. An earlier (1840s) version of the tradition of the arrival of the first Europeans mentions only Kpase, not Kpate: Brue, ‘Voyage’, 55.

27. For the former version, see Gavoy, ‘Note historique’, 48–9; for the latter, Reynier, ‘Ouidah’, 51–2.

28. Reynier, ‘Ouidah’, 38, 47.

29. Fieldwork, Kpatenon compound, 3 Dec. 2001. This version claims that Kpate was settled in Ouidah even before Kpase, and gave him land to settle there.

30. E.g. Merlo, ‘Hiérarchie fétichiste’, 6.

31. The Hunon’s compound in Sogbadji is described Sinou & Agbo, Ouidah, 201.

32. Reynier, ‘Ouidah’, 36; fieldwork, Déhoué compound, 9 Jan. 1996. Another account claims that the Hunon settled in Sogbadji only during the reign of King Glele of Dahomey (1858–89): K. Fall et al., ‘Typologie des cultes vodoun’, in Sinou & Agbo, Ouidah, 72.

33. Fieldwork, Dagbo Hounon compound, 18 Jan. 1996.

34. Thomas Phillips, ‘A Journal of a Voyage made in the Hannibal of London’, in Awnsham Churchill & John Churchill, Collection of Voyages and Travels (London, 1732), vi, 226; William Bosman, A New and Accurate Description of the Coast of Guinea (London, 1705), 383.

35. Melville J. Herskovits, Dahomey (New York, 1938), ii, 155. The worship of the crocodile was also noted in the 1690s, and the name Tokpodun first recorded in the 1860s: Phillips, ‘Journal’, 223; Burton, Mission, ii, 148.

36. For the Dangbe cult, see esp. Christian Merlo & Pierre Vidaud, ‘Dangbe et le peuplement houeda’, in François de Medeiros (ed.), Peuples du Golfe du Bénin (Paris, 1984), 269–304.

37. Described in Sinou & Agbo, Ouidah, 195–7.

38. E.g. Agbo, Histoire, 15–16, who says that Kpase ‘consecrated his town to the fetish Dangbe’.

39. Bosman, Description, 368a, 383.

40. The main Dangbe shrine was in fact located outside Savi, according to European accounts of the early 18th century 3/4 or 1/2 a league (1-1/2–2 miles/2–3 km) away: ‘Relation du royaume de Judas en Guinée’ (MS. of c. 1715, ANF, Dépôt des Fortifications des Colonies, Côtes d’Afrique 104), 60; Labat, Voyage, ii, 154. However, an earlier (1690s) source gives a much greater distance, about 2 [Dutch] miles (= 8 English miles/12 km), perhaps a different site: Bosman, Description, 370. Some versions of local tradition maintain that the earliest shrine of Dangbe was in a forest outside Ouidah to the north, near the modern Roman Catholic seminary, which might be the location indicated: Sinou & Agbo, Ouidah, 195.

41. Burton, Mission, ii, 141.

42. Merlo, ‘Hiérarchie fétichiste’, 4. Other local accounts claim that the primacy of the Hunon was established only during the colonial period: Fall et al., ‘Typologie’, 72, n. 31. But this is refuted by Burton’s earlier evidence.

43. An alternative etymology, however, posits a founder called Gle: I. Akibode, ‘De la traite à la colonisation’, in Sinou & Agbo, Ouidah, 31.

44. Hazoumé, ‘Aperçu historique’, 2nd part, no. 5 (19 Nov. 1925), 8–9. Tradition in Zoungbodji recalls the original Hueda name of the town as ‘Glesinme’: fieldwork, Zoungbodji, 11 Dec. 2001.

45. For examples, see Iroko, Les Hula, 56–8.

46. Agbo, Histoire, 15.

47. Gavoy, ‘Note historique’, 48–50.

48. This date is given on painted and appliqué cloths commemorating the event kept in the Kpatenon compound, as observed in fieldwork, 3 Dec. 2001.

49. These dates are painted on a wall in the Hounon compound, observed in fieldwork, 18 Jan. 1996.

50. In fact, 120 years (four reigns at 30 years each) backwards from the end of Hufon’s reign in 1727 would indicate a date for the beginning of Kpase’s reign of c. 1610, rather than c. 1580.

51. See esp. ANF, C6/25, Du Colombier, 10 Aug. 1714 (giving the name in the form ‘Bangala’).

52. For variant versions of the Hueda king list, see n. 20 above.

53. Alexis Adandé, ‘Buried heritage, surface heritage: the Portuguese fort of São João Baptista de Ajudá’, in Claude Daniel Ardouin & Emmanuel Arinze (eds), Museums and History in West Africa (Oxford, 2000), 127–31.

54. Kenneth G. Kelly, ‘Transformation and continuity in Savi, a West African trade town: an archaeological investigation of culture change on the coast of Bénin during the 17th and 18th centuries’ (PhD thesis, University of California at Los Angeles, 1995); idem, ‘Using historically informed archaeology: seventeenth and eighteenth century Hueda/European interaction on the coast of Bénin’, Journal of Archaeological Method & Theory, 4 (1997), 353–66.

55. This explanation first in Gavoy, ‘Note historique’, 48.

56. See esp. des Marchais, ‘Journal’, 40v; Labat, Voyage, ii, 10–11. Des Marchais describes this river as running by the Allada capital, evidently conflating the Toho with a watercourse that runs into it further east.

57. However, a variant recorded in Zoungbodji itself claims that it was Kpase who accompanied Kpate to the shore and fled, and that Kpate then took the Europeans to meet Zingbo at Zoungbodji. These discrepancies evidently relate to disputes about seniority/primacy; the Zoungbodji version also claims that Zingbo was the first settler in the area, and gave land to both Kpase and Kpate. Fieldwork: Zoungbodji, 11 Dec. 2001.

58. Sinou & Agbo, Ouidah, 173.

59. For the fishing techniques employed in recent times by Hueda displaced by the Dahomian conquest of their kingdom in 1727, on Lake Ahémé to the west, see R. Grivot, ‘La pêche chez les Pedah du lac Ahémé’, BIFAN, Série B, 11/1–2 (1949), 106–28.

60. Burton, Mission, i, 33, says that ‘many’ preferred fish to meat; cf. 136–7, where he states that only the rich ate meat. For dried fish in the Abomey market, see ii, 243.

61. Phillips, ‘Journal’, 214, 221; Bosman, Description, 362a.

62. Fieldwork, Déhoué compound, 9 Jan. 1996.

63. ‘Relation du royaume de Judas’, 2, 75.

64. E.g. Duncan, Travels, i, 190–1. For salt production in modern times, including reference to Djedgbadji, see Josette Rivallain, ‘Le sel dans les villages côtiers et lagunaires du Bas Dahomey’, Annales de l’Université d’Abidjan, série I (Histoire), 8 (1980), 81–127; also A. Félix Iroko, ‘Le sel marin de la Côte des Esclaves durant la période précoloniale’, Africa (Rome), 46 (1981), 520–40.

65. E.g. Pierre Bouche, Sept ans en Afrique occidentale (Paris, 1885), 8, 320.

66. Barbot, On Guinea, ii, 621.

67. As suggested in Law & Strickrodt, ‘Introduction’, in Ports of the Slave Trade, 3.

68. Robin Law (ed.), The English in West Africa, 1685–88: The local correspondence of the Royal African Company of England, 1681–1699, Part 2 (London, 2001; hereafter English in West Africa, ii), no. 817: John Carter, Ouidah, 7 June 1686. On this occasion the level of the water had risen ‘4 or 5 feet higher than ever I saw them, and flooded the dry ground about a mile in breadth’.

69. Law, English in West Africa, i, no. 495: Petley Wyburne, Ouidah, 26 June 1683.

70. Robin Law (ed.), Correspondence from the Royal African Company’s Factories at Offra and Whydah on the Slave Coast of West Africa in the Public Record Office, London, 1678–93 (Edinburgh, 1990), no. 7: William Cross, Offra, 13 June 1681; Law, English in West Africa, i, no. 490: Timothy Armitage, Ouidah, 5 Dec. 1682.

71. PRO, FO2/886, Louis Fraser, Journal, Ouidah, 30 July 1851: ‘messengers go from here to Godomey by land, the rest of the route by canoe as far as Lagos’. But for an instance of travel by canoe from Ouidah to Godomey, see WMMS, William West, Cape Coast, 6 June 1859.

72. Dapper, Naukeurige Beschrijvinge, 2/118–19. The English fort at Ouidah in 1723 reported explicitly that the ‘Whydah cloths’ purchased by Europeans were not made locally, but in ‘Lucamee’: PRO, T70/7, Baldwyn, Mabyn & Barlow, Ouidah, 9 Aug. 1723.

73. For the impact of the European trade in stimulating the expansion of the lagoon traffic, see more generally Robin Law, ‘Between the sea and the lagoons: the interaction of maritime and inland navigation on the precolonial Slave Coast’, CEA, 29 (1989), 220–4.

74. Burton, Mission, i, 62. Burton assumed this tradition to refer to the period prior to the rise of Ouidah as a centre for the European trade in the seventeenth century, but this may have been a misunderstanding on his part, since an account by the Roman Catholic missionary Francesco Borghero, who was one of Burton’s principal informants, recorded stories of Hueda piracy along the lagoons in relation to a later period, after the Dahomian conquest in 1727, when the section of the Hueda people now in exile to the west recurrently raided Ouidah itself: ‘Relation sur l’établissement des missions dans le Vicariat de apostolique de Dahomé’, 3 Dec. 1863, in Journal de Francesco Borghero, premier missionnaire du Dahomey, 1861–1865 (ed. Renzo Mandirola & Yves Morel, Paris, 1997), 240.

75. Dapper, Naukeurige Beschrijvinge, 2/120.

76. ‘Relation du royaume de Judas’, 17; des Marchais, ‘Journal’, 60v; Labat, Voyage, ii, 20.

77. For discussion, see Law, ‘Between the sea and the lagoons’, 229–31.

78. Law, Slave Coast, 118–27.

79. E.g. Akinjogbin, Dahomey, 31.

80. King of Hueda to viceroy of Brazil, 26 Oct. 1720, in Pierre Verger, Flux et reflux de la traite des nègres entre le Golfe de Bénin et Bahia de Todos os Santos du XVIIe au XIXe siècle (Paris, 1968), 132.

81. Gavoy, ‘Note historique’, 49–51. However, one recorded version of the story of Kpate’s encounter with the first Europeans identifies them as French rather than Portuguese: Reynier, ‘Ouidah’, 51–2.

82. De Sandoval, Naturaleza, 51. In 1625 Dutch pirates took a Portuguese ship at ‘Fulao’: Beatrix Heintze (ed.), Fontes para a história de Angola de seculo XVII (Wiesbaden, 1985–8), ii, no. 53, Fernão de Souza, n.d. [c. 25 March 1625].

83. These figures are derived from Eltis, Rise of African Slavery, 166 (Table 7.1). See. also the earlier estimates of Patrick Manning, ‘The slave trade in the Bight of Benin, 1640–1890’, in Henry A. Gemery & Jan S. Hogendorn (eds), The Uncommon Market: Essays in the Economic History of the Atlantic Slave Trade (New York, 1979), 107–41, who gives a slightly higher figure for the 1700s; and also contemporary estimates cited in Law, Slave Coast, 163–5 (which are somewhat higher again). The estimates of Eltis and Manning relate to the Bight of Benin as a whole, but in this period slave exports from the Bight were almost entirely through Ouidah. Exports from the Bight continued at a comparable (or slightly lower) level in the 1710s, but from 1714 a significant proportion of the trade was diverted through rival ports to the east and west of Ouidah.

84. Eltis, Rise of African Slavery, 182.

85. The date is commonly given in local sources inexactly as 1670. Local ‘tradition’ (here as often, manifestly conflated with material from published contemporary sources) also attributes the founding of the Ouidah factory to Delbée: first in Gavoy, ‘Note historique’, 50. This is a confusion: Delbée had served in the expedition that established the original French factory at Offra in 1670, but did not accompany Carolof on the second expedition that transferred the factory to Ouidah in the following year.

86. This story first in Gavoy, ‘Note historique’, 50, 66.

87. ‘Relation du voyage d’Issyny fait en 1701 par le Chevalier Damon’, in Paul Roussier (ed.), L’Établissement d’Issigny 1687–1702 (Paris, 1935), 106 (where ‘Janire’ is clearly a miscopying of ‘Savire’, i.e. Savi).

88. Journal du corsaire Jean Doublet de Honfleur (ed. Charles Bréard, Paris, 1883), 253–6.

89. Gavoy, ‘Note historique’, 50–1. Gavoy himself assumed that the tradition related to the original establishment in 1671, with consequent distortion of his chronology of the Hueda kings. Agbo, Histoire, 40, supposed that ‘Amar’ was a distinct person from Ayohuan, and his immediate successor, and this is generally followed by subsequent writers. But since Ayohuan is described as son and successor to Agbangla, who is known from contemporary sources to have died in 1703 (and these also make clear that there was no other king intervening between Amar and Hufon, the last Hueda king, who succeeded in 1708), it is clear that the two are identical.

90. Description of the site in Sinou & Agbo, Ouidah, 127–9.

91. Barbot, On Guinea, ii, 635.

92. Law, Kingdom of Allada, 18–19.

93. Akin L. Mabogunje, Urbanization in Nigeria (London, 1968), 118, 205, 283; Abner Cohen, Custom and Politics in Urban Africa: A study of Hausa Migrants in Yoruba Towns (London, 1969). Hausa quarters in Yoruba towns are generally called sabo, but in Ghana they are called zongo: see Nehemiah Levtzion, Muslims and Chiefs in West Africa (Oxford, 1968), 23.

94. Denise Brégand, Commerce caravanier et relations sociales au Bénin: les Wangara du Borgou (Paris, 1998), 81–96.

95. Local tradition maintains that the location of the Zobé market predates the Dahomian conquest of Ouidah in the 1720s: Agbo, Histoire, 105–6.

96. Law, English in West Africa, i, nos 476–8: John Thorne, Glehue, 24 May 1681; William Cross, Offra, 18 Aug. 1681; Thorne, Offra, 19 Aug. 1681.

97. Ibid., no. 479: Thorne, Offra, 4 Dec. 1681. In 1686 Wyburne was arrested and taken prisoner to the Royal African Company’s headquarters at Cape Coast Castle on the Gold Coast, and from there shipped back to England: PRO, T70/11, Henry Nurse et al., Cape Coast, 19 March 1686; T70/12, Edwyn Steede & Stephen Gascoyne, Barbados, 27 April 1686. For the location of Wyburne’s factory, in ‘the Lower Town’, i.e. Glehue, as opposed to the royal capital Savi, see Law, English in West Africa, i, no.487: Timothy Armitage, Ouidah, 24 Oct. 1682. The English factory in Ouidah noted by Barbot in April 1682 was evidently Wyburne’s: Barbot, On Guinea, ii, 635.

98. Law, English in West Africa, i, no. 492: Thorne, Offra, 28 Jan. 1683; also enc. 2 in no. 487: Accounts of John Winder, Hueda, July–Oct. 1682. The location of this factory at Savi is implied in a reference to the removal of goods from it ‘to the Lower Town’, i.e. Ouidah: ibid., no. 487: Armitage, Ouidah, 24 Oct. 1682.

99. Law, Correspondence from Offra and Whydah, nos 16, 18: Carter, Hueda/Ouidah, 26 May & 11 Dec. 1684.

100. In 1685 the factor referred to having gone ‘up to the king’s town’, i.e. Savi: Law, English in West Africa, ii, no. 812: Carter, Ouidah, 19 Sept. 1685.

101. Robin Law (ed.), Further Correspondence of the Royal African Company of England Relating to the ‘Slave Coast’, 1681–1699 (Madison, 1992), no. 63: John Wortley, Ouidah, 5 Jan. 1692; Phillips, ‘Journal’, 215.

102. Édouard Foà, Le Dahomey (Paris, 1895), 417; Gavoy, ‘Note historique’, 30.

103. Description of site in Sinou & Agbo, Ouidah, 133.

104. Augusto Sarmento, Portugal no Dahomé (Lisbon, 1891), 52; Agbo, Histoire, 30.

105. Law, Slave Coast, 134–6.

106. A.F.C. Ryder, ‘The re-establishment of Portuguese factories on the Costa da Mina to the mid-eighteenth century’, JHSN, 1/3 (1958)’, 160–1; Verger, Flux et reflux, 132–9.

107. Verger, Flux et reflux, 136–9.

108. Description in Sinou & Agbo, Ouidah, 139–47.

109. Agbo, Histoire, 31–2; Sinou & Agbo, Ouidah, 148–9.

110. Borghero, Journal, 44 [20 April 1861]; also Burton, Mission, i, 67 (who probably had the information orally from Borghero). More recent tradition, however, has moved the site further west, to the area called Adamé.

111. Law, Further Correspondence, no. 63: Wortley, Ouidah, 5 Jan. 1692.

112. Phillips, ‘Journal’, 228.

113. Charles Davenant, Reflections upon the Condition and Management of the Trade to Africa (1709), reproduced in The Political and Commercial Works of Charles d’Avenant (London, 1771), v, 226. Also repeated in the English version (published 1732) of Jean Barbot’s work (see On Guinea, ii, 644), and from Barbot by Burton, Mission, i, 84.

114. ANF, C6/25, ‘Mémoire de l’estat du pays de Juda et de son négoce’, 1716; des Marchais, ‘Journal’, 29; Labat, Voyage, ii, 35.

115. Albert Van Dantzig (ed.), The Dutch and the Guinea Coast 1674–1742: A Collection of Documents from the General State Archive at The Hague (Accra, 1978), no. 250: Agreement with the King and Grandees of Hueda, 12 Nov. 1726.

116. First in Burton, Mission, i, 84, citing Barbot (in turn repeating Davenant), but the Danish fort mentioned by the latter was clearly at Accra, not Ouidah: see Barbot, On Guinea, ii, 644. Burton makes clear that in the 1860s there was no local recollection of this supposed Danish fort, which was ‘now quite forgotten’; very probably, the later ‘tradition’ derives from his work.

117. Paul Erdman Isert, Letters on West Africa and the Slave Trade (trans. Selena Axelrod Winsnes, London, 1992), 96; see discussion in Klavs Randsborg, et al., ‘Subterranean structures: archaeology in Bénin, West Africa’, Acta Archeologica, 69 (1998), 219.

118. Burton, Mission, i, 83.

119. Agbo, Histoire, 32–3; Forbes, Dahomey. i, 224; Burton, Mission, i, 156. However, Agbo’s statement that the Ouidah forts were the first ‘storey houses [maisons à l’étage]’ on the Slave Coast is not accurate, since already in 1670 part of the royal palace at Allada was ‘raised in two storeys’: ‘Journal du voyage du Sieur Delbée’, in J. de Clodoré (ed.), Relation de ce qui s’est passé dans les isles et terre-ferme de l’Amérique pendant la dernière guerre avec l’Angleterre (Paris, 1671), 418–19.

120. ‘Relation du royaume de Judas’, 67; William Snelgrave, A New Account of Some Parts of Guinea, and the Slave Trade (London, 1734), 128.

121. ANF, C6/25, ‘Mémoire concernant la Colonie de Juda’, 1722.

122. For Elmina, see Feinberg, Africans and Europeans.

123. Robin Law, ‘“Here is no resisting the country”: the realities of power in Afro-European relations on the West African “Slave Coast”’, Itinerario, 18 (1994), 50–64.

124. Van Dantzig, Dutch and the Guinea Coast, no. 115: W. de la Palma, Elmina, 10 Oct. 1703, with no. 121, copy of agreement, 25 Apr. 1703; text of the treaty also in des Marchais, ‘Journal’, 29–30v; Labat, Voyage, ii, 88–91. The date given in the latter, 8 Sept. 1704, is that of a subsequent renewal of the treaty.

125. ANF, C6/25, Du Colombier, Hueda, 4 Feb. 1715; PRO, C113/276, Randle Logan, 20 Feb. 1715.

126. Agbo, Histoire, 41.

127. Gavoy, ‘Note historique’, 49.

128. Obra nova de lingua geral de Mina de António da Costa Peixoto (ed. Luís Silveira & Edmundo Correia Lopes, Lisbon, 1945), 20 (giving the term in the form ‘sujaquem’).

129. Burton, Mission, i, 65, n. The suggestion sometimes made that ‘Aguda’ derives from ‘Ajudá’, the Portuguese version of the name Hueda, is unlikely on both linguistic and historical grounds.

130. First in ibid., i, 64–5.

131. Gavoy, ‘Note historique’, 50–51.

132. Fall et al., ‘Typologie’, 66; also fieldwork, Zossoungbo compound, 9 Jan. 1996.

133. Local informants offer derivations from doko, a form of bean cake supposedly sold in the area, or from the male personal name Dosu, supposedly a member of the founding family. These look like imaginative speculations.

134. Reynier, ‘Ouidah’, 32–3; also fieldwork, Agbamou compound, 11 Dec. 2001.

135. Reynier, ‘Ouidah’, 38–9; cf. Gavoy, ‘Note historique’, 66; also fieldwork, Sebastien Amoua, 11 Dec. 2001. The name is alternatively spelt ‘Ahomblaca’ or ‘Baclahahoun’. Reynier says that Ahohunbakla was a nephew (son of a sister) of Kpate, but Sebastien Amoua says that Kpate was his son.

136. Merlo, ‘Hiérarchie fétichiste’, 16–17, who gives the name in the form ‘Ahoho Agbangla’ (ahoho meaning apparently ‘old king’); however, this identification is not recognized in local tradition.

137. Gavoy, ‘Note historique’, 67; Reynier, ‘Ouidah’, 35; fieldwork, Zossoungbo compound, 9 Jan. 1996; Déhoué compound, 9 Jan. 1996.

138. Davenant, Reflections, in Works, v, 226.

139. Phillips, ‘Journal’, 228; ANF, C6/25, ‘Mémoire de l’estat du pays de Juda’, 1716. The term ‘Minas’, although in origin referring specifically to Elmina (originally ‘A Mina’, ‘The Mine’ in Portuguese), the Dutch headquarters on the Gold Coast, was frequently used in a wider sense, of persons from the Gold Coast (called the ‘Costa da Mina’ in Portuguese usage) in general.

140. Law, Kingdom of Allada, 90–1; cf. Eltis, Rise of African Slavery, 226.

141. Parker, Making the Town, 10–14.

142. Already in the earliest documents surviving from the French fort: ANF, C6/25, ‘Estat ou mémoire de la dépense nécessaire pour relever le fort de Juda et pour l’entretien du directeur et des employés’, enc. to Du Colombier, 10 Aug. 1714. For ‘Acqueras’ as an ethnonym, see Law, Slave Coast, 189–90.

143. ANF, C6/25, Levesque, 4 April 1723. One early eighteenth-century account refers to the slaves of the French fort in Ouidah as ‘Bambaras’, which is a name usually given to slaves from the interior of Senegambia; but this is presumably used here in a generic sense, transferred from Senegambian usage, for slaves employed as soldiers: Labat, Voyage, ii, 34.

144. See, for example, Barbot, On Guinea, ii, 529; Phillips, ‘Journal’, 210, 228–9.

145. ‘Relation du voyage de Guynée fait en 1687 sur la frégate “La Tempeste” par le Sieur Du Casse’, in Roussier, L’Etablissement d’Issigny, 15.

146. Bosman, Description, 50–51.

147. ANF, C6/25, Du Colombier, 10 Aug. 1714. This disadvantage of the French is mentioned in several later documents, and was an argument regularly used in support of proposals for the French to acquire a fort on the Gold Coast.

148. ANF, C6/25, Pruneau & Guestard, ‘Mémoire pour servir à l’intelligence du commerce de Juda’, 18 March 1750.

149. Law, English in West Africa, ii, no. 819: Carter, Ouidah, 22 Nov. 1686.

150. Phillips, ‘Journal’, 219; Bosman, Description, 375.

151. William Smith, A New Voyage to Guinea (London, 1744), 192; Dapper, Naukeurige Beschrijvinge, 2/115.

152. ANF, C6/25, ‘Le Fort de Juda, Côte de Guinée’, n.d. [1714?]; C6/26, Baud-Duchiron, ‘Exploration et construction du comptoir de Juda’, 1 Sept. 1778; C6/27, Gourg, ‘Mémoire pour servir d’instruction au Directeur qui me succédera au comptoir de Juda’, 1791.

153. Verger, Flux et reflux, 136.

154. Law, Kingdom of Allada, 14.

155. Law, English in West Africa, ii, no. 822: Carter, Ouidah, 6 Jan. 1687.

156. For Abomey, see Sylvain C. Anignikin, ‘Etude sur l’évolution historique sociale et spatiale de la ville d’Abomey’ (Ministère de l’Equipement et des Transports, République Populaire du Bénin, 1986).

157. Although it should be noted that there is no reference to a market in Ouidah (as opposed to the capital Savi) in contemporary sources prior to the Dahomian conquest.

158. Labat, Voyage, ii, 33.

159. Barbot, On Guinea, ii, 635, 642.

160. Des Marchais, ‘Journal’, 41v; Labat, Voyage, ii, 12. The second element in the title seems to be a generic term for provincial governors, given in the form ‘onto’ in another source: ‘Relation du royaume de Judas’, 27. Perhaps ‘zonto’ is a miscopying of ‘honto’, i.e. hunto, which meant in origin ‘ship’s captain’ (hun = ‘boat’) but, at least by the nineteenth century, was also applied to indigenous officials: see Burton, Mission, i, 121, n.

161. This now seems to me more likely than that he held the title Yovogan, as suggested in Law, Slave Coast, 214.

162. Law, English in West Africa, i, no. 478, enc., Accounts of Thorne, Glehue, 20 April 1681, giving the title in the form ‘Captain Blanko’ (from Portuguese branco, ‘white’). The indigenous title is given (as ‘Lievauga’) in ‘Relation du royaume de Judas’, 27.

163. Bosman, Description, 359.

164. See Law, Slave Coast, 207.

165. Van Dantzig, Dutch and the Guinea Coast, no. 234: Diary of Ph. Eytzen, 22 April 1718; map of the Hueda kingdom in Labat, Voyage, ii; des Marchais, ‘Journal’, 40.

166. Polanyi, Dahomey, 118, 123, 126.

167. Law, English in West Africa, i, no. 476: Thorne, Glehue, 24 May 1681; see Thorne’s accounts in the ‘warehouse’ in ‘Agriffie’, in no. 478, enc.

168. Barbot, On Guinea, ii, 637; Law, English in West Africa, ii, no. 812: Carter, Ouidah, 19 Sept. 1685.

169. Polanyi, Dahomey, 123.

170. ANF, C6/25, ‘Mémoire de l’estat du pays de Juda’, 1716.

171. Des Marchais, ‘Journal’, 28v, 40v; Labat, Voyage, ii, 34–5.

172. There was a Portuguese factory at Savi by 1727, when it was destroyed in the Dahomian conquest: Smith, New Voyage, 190.

173. Phillips, ‘Journal’, 215–16.

174. As argued notably by David Eltis: e.g. ‘Precolonial western Africa and the Atlantic economy’, in Barbara L. Solow (ed.), Slavery and the Rise of the Atlantic System (Cambridge, 1991), 97–119.

175. For slave prices, see Law, Slave Coast, 178; the average price is assumed to be 80% of that for adult males. For wage rates, see Robin Law, ‘Posthumous questions for Karl Polanyi: price inflation in pre-colonial Dahomey’, JAH, 33 (1992), 415–16.

176. Barbot, On Guinea, ii, 637.

177. Canoemen hired from the Gold Coast were paid half of their wages in advance on recruitment, in gold, and the remainder at the termination of their employment, in goods, which they would presumably carry back with them to the Gold Coast: Phillips, ‘Journal’, 229.

178. Labat, Voyage, ii, 33.

179. Bosman, Description, 362a.

180. For the latter estimate, see ibid. (given in the original in the form $1,500 per ship, around £375); see also analysis (with slightly different calculations) in Law, Slave Coast, 213.

181. Des Marchais, ‘Journal’, 49; Labat, Voyage, ii, 81.

182. Eltis, Rise of African Slavery, 184.

183. Law, Slave Coast, 127–41.

184. PRO, T70/51, Royal African Company to King of Hueda, 12 Aug. 1701.

185. Barbot, On Guinea, ii, 636–7, 658. Barbot actually says that the French in 1671 had paid 100 slaves in ‘customs’ at Offra, but this figure was for two ships, the rate per ship being 50: see Delbée, ‘Journal’, 439–40.

186. Phillips, ‘Journal’, 227. The term ‘caboceer’ (Portuguese cabeceiro, ‘headman’) was commonly applied to African officials.

187. Damon, ‘Relation du voyage d’Issyny’, 107.

188. Law, Correspondence from Offra and Whydah, esp. nos 4, 6, John Mildmay, Offra, 13 Oct. 1680 (referring to the possibility of recovering ‘your old debts from [the King] and some of the chief captains’); William Cross, Offra, Feb. 1681 (an unsuccessful attempt to refuse further credit to the governor of Offra).

189. Law, Kingdom of Allada, 96–7.

190. Bosman, Description, 348–9; cf. Phillips, ‘Journal’, 225–6.

191. Albert Van Danztig, ‘English Bosman and Dutch Bosman: a comparison of texts – VI’, HA, 7 (1980), 284 [passage omitted in the English translation of Bosman’s work].

192. Eltis et al., ‘Slave-trading ports’.

193. Polanyi, Dahomey, ch. 7; see also Arnold, ‘Port of trade’. For a critique of Polanyi’s analysis with reference to market centres in the West African interior, see Paul E. Lovejoy, ‘Polanyi’s “ports of trade”: Salaga and Kano in the nineteenth century’, CJAS, 16 (1982), 245–77.

194. De Chenevert & Bullet, ‘Réflexions sur Juda’ (MS of 1776, in ANF, Dépôt des Fortifications des Colonies, Côtes d’Afrique 111), 5.

195. ANF, C6/25, Levesque, mémoire (responding to a proposal to abandon the Ouidah fort), 6 July 1716.

196. For details, see Law, Slave Coast, 238–42, 245–7, 252–60; idem, Kingdom of Allada, 52–61.

197. Eltis, Rise of African Slavery, 184.

198. Law, Slave Coast, 243–4, 247, 249.

199. ‘Relation du royaume de Judas’, 13.

200. N****, Voyages aux Côtes de Guinée et en Amérique (Amsterdam, 1719), 121; see also Bosman, Description, 396.

201. Law, ‘“The common people were divided”’.

202. Robin Law, ‘Warfare on the West African Slave Coast, 1650–1850’, in R. Brian Ferguson & Neil L. Whitehead (eds), War in the Tribal Zone (Santa Fé, 1992), 103–26.