IBERIAN DOMINANCE AND THE INTRUSION OF THE NORTHERN EUROPEANS INTO THE ATLANTIC WORLD: SLAVE TRADING AS A RESULT OF ECONOMIC GROWTH?

1 I would like to thank Leonardo Marques, Mariza de Carvalho Soares, Robert Slenes, and participants in the Universidade de Fluminense seminar for comments on an earlier version of this essay, as well as Gladys Sabina Ribeiro, Jonis Freire, and Carlos Gabriel Guimarães and the Centro de Estudos do Oitocentos, who made this event possible. The publication of the Forum was supported by the CNPq Universal Project 2016, n. 424255/2016-2, coordinated by Gladys Sabina Ribeiro (A publicação do Fórum contou com o apoio financeiro do Projeto Universal do CNPq de 2016, n. 424255/2016-2, coordenado por Gladys Sabina Ribeiro).


David Eltis
Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Fórum
Unfortunately, the history of the United States still tends to be written with insufficient awareness of what was occurring in the rest of the world. The US slave population probably did not surpass that of Brazil until 1815; at this point, slavery's role in the foundation of US industrialization must have been well established 6 . Furthermore, we now know that 4.9 million captives arrived in Brazil from Africa as opposed to only 390,000 to the North American mainland. Slavery was thus more important in Brazil than in the US until just half a century before the institution was abolished, and there were always more slaves in (as well as a vastly larger transatlantic slave trade to) Brazil than in the British, French and Dutch territorial possessions. Moreover, despite containing a vast indigenous population, some of whom were subjected to mita labor, Spanish America brought in four times more African slaves than the US after 1500.  Amsterdam, v. 67, p. 1-17, 2018. Attempts to link the slave trade to economic growth are particularly problematic given the 2 million colonial US population at the time of the Revolution and the average of 33 small US ships a year that bought slaves in Africa in the 40 years between 1768 and 1807 (http://slavevoyages.org/voyages/9kcmRQ4K). Scholars who try to link Dutch economic growth to the Dutch slave trade between 1670 and 1793 have the same problem, given that an average of eleven slave voyages a year left the Netherlands, compared to 750 Dutch voyages per year passing through the Danish toll alone (see <http://slavevoyages.

David Eltis
Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Fórum
Spanish meant that their possessions in America had higher value exports than either the US or any of their European rivals from 1500 to 1800. If a large slave sector (and the attendant exports of produce) was a feature and a pre-requisite of industrial capitalism, how very odd that Brazil, Portugal and Spain experienced the latter after both north Europeans and north Americans 7 .
In this article I would like to propose turning these arguments upside down with particular reference to the slave trade. In short, slavery and the slave trade in the Atlantic world were dominated by the Iberian nations, and it was the economic growth of northern Europe (and its offshoots in the Americas) that allowed the nations of that region to break into a well-established slave trade system. Given space constraints, the argument will focus on the slave trade rather than slavery itself, but it will also address the significance of this northern intrusion for the African experience of the slave voyage.
Allow me to conform to the norm and begin the analysis with a little micro-history. For five decades after 1808, Freetown in Sierra Leone -the African center of British efforts to suppress slavery -hosted a floating population of former slaver crews looking to return to their Brazilian and Cuban home ports after the British had detained their vessel. In 1849 five sailors from a recently condemned slave ship stole a new 29-foot open boat just delivered from London for the use of the British officials of Mixed Commission Courts. As the British Commissary judge reported, the men "pulled up to Rio Pongo where they either kidnapped or purchased five or six slaves, with whom… they sailed to Brazil and arrived there in safety." Certainly, a 29-7 TEPASKE, John J. A new world of gold and silver. Leiden: Brill, 2010, p. 315, indicates average annual output of bullion alone of 3.9 million pesos in the decade, 1791-1800, almost all of which was exported. US Historical Statistics: From Colonial Times to 1970(Washington, 1974, p. 885 shows US merchandise exports of $27.9 million (or $27.5 using the pesos-dollar exchange rate of 1.016 US in SUMMER, W. G. The Spanish dollar and the colonial shilling. American Historical Review, Oxford, v. 3, n. 4, p. 617, 1898. For a similar argument on the importance of Spanish colonies in the late eighteenth century, see CUENCA-ESTEBAN, Javier. Statistics of Spain's colonial trade, 1747-1820: new estimates and comparisons with Great Britain. Revista de Historia Económica, Cambridge, v. 26, p. 323-354, 2008.
David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?
Fórum foot boat was equipped with some sails, but the letter also specifies rowing. For this case -and possibly other open launches in the database -, perhaps slaves helped propel themselves to the Americas 8 . At the other end of the spectrum of the transoceanic slave experience were twenty-three steam-powered slave ships in the voyages database, averaging 361 tons and disembarking 1,004 captives on average -slave trading in the industrial era. And then there are the clipper ships of 600 tons or more. The Orion (ID 4807) had, according to the arresting officer in 1857, "the finest slave deck I have ever seen being about 8 feet in height and clear fore and aft" 9 . In addition, blockades by the British squadron had forced it to leave the coast with only two--thirds of its intended captives on board 10 . Compare these with the iconic and highly misleading image of the Brooks (ID 80666; Figure  1) from 1788 and the very accurate, by contrast, Marie-Séraphique (ID 30910; Figures 2 and 3) from two decades earlier, and it is hard to believe that all these vessels were in the same transatlantic business within a short 85-year period. Recognizing, or learning to recognize, an East India man, a whaler, a Dutch fluyt trading for grain in the Baltic Sea is easy, but not a slave ship 11 . David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Figure 1
Description of a Slave Ship (London, 1789) David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Figure 2 [Rene L'Hermite and Jean-Baptiste Fautrel Gaugy ?], Detail from Plan, Profil, et Distribution du Navire La Marie Séraphique de Nantes… [1770 ?]
Reproduced with permission of the Musée d'Histoire de Nantes, France David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

David Eltis
Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?
In short, the Atlantic slave trade ranged over 95 degrees of latitude and was shaped by so many environmental factors, regulatory regimes, and changes in technology over time, that no less than 74 types of ship (mainly different rigs) can be identified in the Trans-Atlantic Slave Trade Database (TSTD) 12 . Within a given type (or rig) ships also varied in size. Slavers leaving the North American mainland and the Caribbean before the nineteenth century averaged just half the capacity of their counterparts leaving Europe and Brazil, and on the African coast, vessels trading in Upper Guinea were somewhat smaller than those buying slaves east and south of Cape Palmas. Almost every port worthy of the name in the Atlantic-facing world organized and dispatched a ship to Africa for slaves, but in the age of sail, shipbuilding centers and the skills that supported them were even more broadly distributed geographically than were ports and merchants participating in transoceanic trades. Sailing ships were cheap to construct when compared to the costs of fitting out and the value of the cargoes they carried. Before the adoption of copper-sheathing technology, a hull could survive six or seven roundtrips in tropical waters. TSTD records the place of construction for the vessels used in 9,155 voyages. These locations ranged from Calcutta, India to Portobelo, Panama, to Quebec City and Mahone Bay in Canada to Point Askaig in the Western Isles of Scotland and Kronstadt in Russia. From this sample, seven hundred had been taken as prizes in wars, and most had never previously been employed in the slave trade. When looking for a vessel to send to Africa before the nineteenth century, investors could quickly adapt almost any ship under 400 tons for the purpose. Owners often built vessels specifically for a slaving venture, but these, too, were of a wide variety of rigs and sizes 13 . In a sense then, there 12 The five most common in Slavevoyages (henceforth, TSTD) -bergantim, curveta, schooner, brig and ship do account for close to two fifths of our sample of 23,500 records, but within these rigs, there was a wide variation in size (see also the discussion in RODRIGUES, Jaime. De costa a costa: escravos, marinheiros e intermediários do tráfico negreiro de Angola ao Rio de Janeiro, . São Paulo: Cia. das Letras, 2005, p. 121-133. 13 To put the main point here in a different context, any kind of transatlantic mercantile business (manufactured goods from Europe, produce from the Americas), drew on the same range of David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?
Fórum was no such thing as a slave ship, except insofar as it was a ship that carried slaves, but analysis of the records of 36,000 slave voyages and the approximately 9,000 ships that made these voyages, nevertheless make possible new generalizations. These in turn allow us to understand more fully both the patterns of the transatlantic slave trading, and the slave experiences that historians have missed.
As is well-known, the transatlantic traffic developed as an extension of the slave trade from Africa to Europe -often called the Old World slave trade. The very first captives brought into Portugal via the Atlantic Ocean were Guanches, the indigenous people of the Canaries -in the early 1420s -and then Moors, more precisely blackamoors, or Berber peoples from as early as 1441. After 1450 the traffic drew on sub-Saharan peoples with numbers carried off to Europe approaching 5,000 a year by the second half of the 1510s. Ivana Elbl has tracked the value of gold, ivory and other African produce exports to Europe in these early years, and supports the earlier assessment of A.G. Hopkins that "the Europeans who came to Africa in the fifteenth and sixteenth centuries were interested in goods other than sla-ves…and this commerce continued even after the overseas slave trade was well underway." 14 Commodities arriving from Atlantic ports prior to Columbian contact comprised sugar from Madeira (and later São Tomé) as well as mainland produce. While there are records of shipping types and vessel size. In the illegal slave trade era, Cuban merchants referred to a slave in their correspondence as a "bulto" or in English a "package", in a crude attempt to disguise their activities in the event of capture. The eighteenth-century counterpart to the modern container unit was the barrel. From the standpoint of the slave trading community -indeed society at large before the nineteenth century -a slave was a package or a barrel, albeit a dangerous package that might explode into revolt and destroy the ship, and one that, unfortunately, had to be fed and guarded. But it seems that if a vessel could carry barrels -the container unit of the early modern era -then it could also carry slaves, as long as it was modified by the carpenter on its way to Africa. If the specifics of ship size and type varied, this was largely on account of the differing coastal environments of the major African embarkation regions, not because of what the ship carried.
14 HOPKINS, Anthony G. Economic History of West Africa. London: Routledge, 1973, p. 89 Fórum ships arriving in Portugal with more than 100 slaves, such numbers were exceptional. The samples of voyages scattered through Ivana Elb's work indicate a mean of less than 100 15 . Captives were carried on merchant vessels and, if sufficiently numerous and comprising a threatening mix of gender and age, they were undoubtedly manacled and chained, but in these early years such vessels probably carried more crew than slaves.
The early transatlantic traffic presents a similar pattern. For the first quarter of the sixteenth century, the galleons that left Spanish ports for the Americas carried European merchandise, approved emigrants, and slaves, most, but not all, of the latter being of African descent. When African slaves began to arrive in the Americas directly from Africa rather than Europe in 1520, the mix of slaves and commodities continued. The mean number of captives on board of 131 vessels arriving in San Juan, Puerto Rico between 1520 and 1546 was just sixteen, with only four carrying more than ninety captives. Ships collecting slaves from the offshore bulking centers of Arguim Island and Santiago in the Cape Verde Islands, not only freighted merchandise for the Americas, they also picked up gold and perhaps sugar from Madeira, African spices and Ivory. Sixteen slaves carried on just one of the three legs of a voyage from Seville to Africa and the Americas and back, could not have been the main preoccupation of investors. In the 1550s detailed records of two large transatlantic vessels suggest that each disembarked over one hundred slaves in Hispaniola and Vera Cruz respectively. But on arrival they were also found to be carrying huge volumes of wine, olive oil, and a wide range of manufactured goods 16 . Even when the Iberians began to seek slaves south of Senegambia, first, at São Tomé in the Gulf of Guinea, and then West-Central Africa, commodities remained of central importance. Quite apart from African gold that at one stage made up ten percent David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?
Fórum of global supplies, São Tomé was Europe's major source of sugar in the second half of the sixteenth century. The ten slave voyages in TSTD that carried captives from that island to the Americas between 1526 and 1592, were likely carrying sugar from Africa as well as slaves. It is impossible to identify the point at which the value of slaves exceeded that of commodities on a "typical" transatlantic venture, but the main point is that slaves required space not only for themselves, but for provisions and water too. European merchants obviously found transporting merchandise to the Americas and African produce brought to Europe via the Americas just as profitable as slaves. The value of captives may not have exceeded that of commodities until the 1560s or 1570s. What we can say is that the very earliest transatlantic captives must have travelled under conditions that their successors could not have imagined possible.
The TSTD shows the average number of captives per vessel increasing after 1550 but given the extent of smuggling into the Spanish Americas, these data -drawn from official records -are not very reliable. Fortunately, between 1575 and 1637 seventy very well documented cases of vessels exist because the colonial authorities suspected that the captains were bringing in more slaves than that declared on board. The extensive investigations that followed revealed that, on average, the seventy vessels actually disembarked 278 captives (SD = 124.3) as opposed to the 167 that the captains claimed they had imported. After allowing for voyage mortality these ships would likely have embarked between 350 and 400 captives on the African coast. There is no statistically significant difference between the true average of 278 slaves disembarked by this sample and the mean of 263 that TSTD generates for the 10,163 British, French, Dutch, Danish vessels carried to the Caribbean over the whole period of the slave trade 17 . The conclusion from these comparisons is that a transatlantic slave trade of a type that most scholars would recognize, and most slaves had to endure, was fully established sometime between 1560 and 1590. By David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?
Fórum "type" it is meant ships carrying several hundred slaves under intensely crowded conditions, the value of whom far exceeded that of all other commodities on board. Such a system had not existed previously. Thus, the question of a ship built for the primary purpose of slave trading did not arise for over a century after Europeans began carrying slaves from Africa.
However, there is also need to re-examine and re-group the 35,800 voyages that occurred after 1560. Since the publication of Daniel Domingues' 2008 article, it has become standard practice to analyze the traffic in terms of the northern and southern gyres that essentially shaped the routes of transatlantic sailing ships 18 . While the gyres were indeed the most fundamental environmental factor, a close study of ship types, trading practices, shipboard rebellions, and voyage durations points to the emergence of a Portuguese system of acquiring and moving labor across the Atlantic that no other Europeans were ever able to emulate. This system dominated the Atlantic world until 1867, and became internationalized after 1807, even though Portuguese nationals retained a critical organizing role beyond the point at which their main market in the Americas -Brazil-was closed to the transatlantic business. The gyres clearly determined the maritime routes, but there were more important political and cultural factors that ensured Portuguese dominance in Atlantic slaving in the more than four centuries after 1450 -a dominance that remained unchallenged except for the British interlude in the half-century before abolition.
These same factors also ensured that African slaves and African slave traders experienced the slave trade very differently from their counterparts who engaged with and helped to sustain what is identified here as the northern European system of slave trading. The French, British, Dutch and Danes are best viewed not only as latecomers to the slave trade but as intruders, condemned to fall back on se-David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?
Fórum cond best options in the business. The fact that they were rarely able to adopt the preferred system of the Portuguese ensured that they faced inefficiencies and higher costs in accessing slaves, costs that were absorbed ultimately by their planters and consumers. We must revise our view of how the history of the Atlantic World evolved. As already suggested, it was not the slave trade, or more generally, global markets, that stimulated British and Dutch economic growth, but rather economic growth within Europe that allowed first the Dutch and later the British to intrude into Portuguese (and more broadly Iberian) domination of the Atlantic slave trade and the exploitation of the Americas.
At the very outset of their slave trading activities the Iberian slave traders -mainly Portuguese -traded from their ship, but soon established permanent factories, some like those at Mina and Axim dealing in gold. The early slave trade to Europe came to rely heavily on what Curtin called "bulking centers", or assembly points for captives 19 . As the early direct trade in slaves from Africa to the Americas developed, the Spanish and Portuguese used several such stations. All were on islands under the full control of Iberians. Tenerife in the Canaries and Santiago in the Cape Verde islands were added to Arguim with the first two being the most important down to the end of the sixteenth century. Thus, Tenerife and Santiago pulled in Africans from regions ranging from the Senegal river to what became Sierra Leone, and when the Southern Rivers emerged as the most important source in the region some transatlantic vessels collected their complement of slaves directly from Cacheu rather than Santiago. Luanda, founded in 1575 and later to become the single largest embarkation point in the whole Africa, was also first settled on an island, initially supplying slaves to São Tomé. Until 1720, Benguela dispatched all its slaves destined for the Americas to the "bulking center" of Luanda, before the volume of this traffic warranted direct trade David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?
Fórum connections with Brazilian ports 20 . Later in the century, as wars and St. Domingue and the ensuing international conflict erupted, French withdrawal from the Indian Ocean allowed Mozambique island to assume its major role in the transatlantic business. The three southern bases of Luanda, Benguela and Mozambique, became well-fortified Portuguese enclaves that together accounted for three-quarters of all captives carried off from Africa south of the Congo River. At each location, when sufficient numbers of captives were acquired, the Portuguese would embark the full complement of captives at once. A Capuchin friar in Luanda described the scene in 1708 as he boarded a slave vessel to Bahia: There was assembled there not only a large number of whites, but also a multitude of blacks who were to be embarked for America. On this island, all these blacks were reviewed, and a census of them was taken by the superintendent and ministers of the royal revenues. On my return the boarding began of these blacks, whose numbers amounted to 742. 21 20 FERREIRA, Roquinaldo Amaral. Slavery and resistance to slaving in West Central Africa. In: ELTIS, David; ENGERMAN, Stanley L. (Eds.

Fórum
Slave ship captains could thus receive all their captives much more quickly than in West Africa. The exceptionally large barracoons (or quintais) in Luanda and Benguela were noted by several observers 22 . For Bissau and Cacheu, Antônio Carreira makes no reference to shipboard revolts, or measures taken to suppress them, but does reproduce a document cataloguing 100 deaths and escapes per year "antes do embarque" (prior to boarding) in the factories over the period of 1768 to 1777, many of them the result of "uprisings" 23 .
Portuguese ship trade, or the gradual accumulation of slaves aboard the vessel over several weeks, became largely confined to the smaller coastal craft that connected the bases with the adjacent African mainland. In Angola, the bulking centers drew on land-based networks, rather than coastal seaborne and river traffic. The Portuguese were able to extend a degree of sovereignty -or at least influence -inland via warfare and treaties, maintaining a handful of forts and factories strategically located on trade routes. In all other parts of sub-Saharan Africa, European authority extended only as far as a cannon shot, and on most of the West African coast the cannon was located on a vessel. Even the famous Gold Coast forts amounted to no more than a large anchored ship with a well-armed crew; there were few mulattoes, and no European factories in their hinterlands. By contrast, Luanda had a population of 5,000 at the end of the eighteenth century, one quarter of whom were whites and mulattoes. Though Benguela and Mozambique were half the size of Luanda, compared to the Gold Coast forts, they constituted major communities under European jurisdiction that anchored trading routes extending into the inland 24 .

Fórum
From the slave traders' viewpoint, the success of this system was less due to the island bases than to the interactions between the Portuguese and Africans. These were of a kind that no other Europeans were able to replicate. The links between coastal bases and the sources of captives in the interior were established and maintained by lançado traders of mixed Portuguese and African heritage. There is no need here to re-enter the old debates over attitudes toward race in the different European and African Atlantic worlds, but differences there were and still are 25 . The British, French, Dutch and Danes had few equivalents to the lançados. Some families on the Upper Guinea Coast -the Ormonds and the Lightbodies in the Pongo River, for example -descended from minor Liverpool slave ship owners and captains, but the extensive documentation of slave purchases by the northern Europeans on the West African coast overwhelmingly points to the sellers of slaves being African, not some equivalent of lançados 26  David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Fórum
Nigeria, but followed this with the imposition of a century of colonial rule. Yet they have left little genetic trace in the modern populations of these countries. Nor have the northern Europeans left anything comparable to the Afro-Brazilian presence in say Porto Novo -where commercial contact with Bahia began only in 1760 27 . The Portuguese thus had perhaps the most complete plantation system in world history in place by the early seventeenth century. While, for the most part, they purchased captives rather than enslaved them, those purchases took place much closer to the point of enslavement than was ever to be possible for other Europeans. From the late sixteenth century forward the Portuguese were able to oversee most of the stages that turned labor into the consumption of sugar in Europe.
If the pre-1560 era of non-specialized slave trading constituted the first type of transatlantic slave trading, and the direct links between the African source of labor, the Brazilian engenhos and the European consumer comprised the second type, a third method emerged as other European nations entered the business from 1640. The English and Dutch had initially built -or in the Dutch case, taken control of -forts on the Gold Coast to facilitate trade in African produce -especially gold -rather than slaves. Together with the French, they gradually wrested some of the smaller Caribbean islands and South American mainland from Spanish control to establish their own plantation systems and slave trades. The forts did eventually become assembly points for slaves, but the major source of slaves was to the east and south of the Gold Coast, extending to the mouth of the Congo River. Except, possibly, for eighteenth century Anomabu, Gold Coast forts were never able to play the role of Tenerife, Santiago, Luanda, Benguela and, later, Mozambique. Apart from Ouidah -and even there, the Kings of first, Ardra, and then, Dahomey, were always very much in control -no European power was able to establish a permanent foothold. The Dutch attack on Luanda, temporarily successful in the David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth? Fórum 1640s, was a recognition of the advantages of system two over system three. If Dutch lançados (truly, an oxymoron) had existed, the occupation might have endured. With the Brazilian re-conquest of Luanda, the massive British, Dutch and French slave trades between Cape Palmas and the Congo River that developed after 1640 were, of necessity, ship based.
Thus, over a wide range of the coast the slave ship -which had hitherto been a means of transportation -was, with arrival of the northern Europeans, now adapted to additional roles. The ship ferried merchandise in one direction and slaves in the other, but it also became a trading platform where wares could be displayed and clients entertained, as well, most importantly, as a floating stockade or barracoon. A vessel might now expect to spend many months accumulating slaves, usually at more than one location. In Curtin's phraseology, bulking costs were now transferred from the coastal hub to the vessel itself. The implications of this for the design and construction of slave vessels, for slave resistance, for the African experience of transportation to the Americas, and more broadly yet, for imperial rivalries and economic development in the Atlantic world were not trivial.
For several thousand cases in TSTD, it is now possible to calculate the duration of different phases of a slave voyage. As we might expect, roundtrip voyages from Brazil during the eighteenth century -when data first become available -were 21 percent shorter than those setting out from Europe and selling slaves in the Caribbean (n = 5,707). But the key question in terms of shipping cost is how long the slaves were on board. The passage from Africa to the Americas accounts for much of the difference in roundtrip times. Voyages from both the Bight of Benin and West Central Africa to the Caribbean were 50 percent longer than those sailing to Brazil (n = 2,378). Nevertheless, the major factor in separating the Portuguese from the northern Europeans was what happened before the transoceanic trades began. In West African ports, between 1640 and 1807, Dutch, French, and English vessels received their first captive 80 days after leaving their home port on average. Accumulating a full complement of captives via trading from the ship took a further 140 days. Therefore, the very David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth? first captive purchased would typically be on board of the vessel for seven months, and severe crowding below decks usually began a month or so before departure when the captain began to pay higher prices for slaves in order to get off the coast as quickly as possible 28 . There are many voyages in TSTD where captives spent a year aboard the slave vessel between first boarding and disembarkation. By contrast, most vessels bound for Brazil had slaves on board for just a few days more than the duration of the crossing -less than two months. Despite this, Portuguese and northern European ships spent similar amounts of time on the African coast; but the Portuguese vessels were frequently empty during this time or had very few slaves on board. The competition between captains in places such as Luanda was over the order of embarkation, not about who should claim the last few captives that would make up a full complement.
It might seem that for the captive there was little to choose between a stockade and the confined hold of a sailing vessel. Both were unhealthy environments that made extensive use of shackles and offered few opportunities for escape. Given that the slave ship captain had the power to refuse captives who were not in prime condition, it could be argued that shore-based imprisonment was more lethal. Yet, as discussed below, a slave deck was without parallel in the annals of inhumanity; a tightly restricted space with some gratings for light, occupied for months by hundreds of people so that no part of the deck or hull was visible. Such conditions were never replicated in any other transoceanic movements of people whether convicts, troops, indentured servants, Asian contract workers, Irish famine migrants, or modern illegal migrants. In the northern European trade, the closest land-based parallel to the Portuguese system was the holding cell of a Gold Coast fort, something that relatively few of those entering the transatlantic slave trade experienced, though we do have instan-David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?
Fórum ces of the Danes, for example, boarding a complete complement of enslaved people from a fort just prior to departure 29 .
For the slave trader the difference was very much a matter of cost. A ship that was expected to imprison captives for months instead of weeks required additional fortifications, crew and armament. We have surprisingly little information on the design of slave vessels before 1670, but for the long eighteenth-century the sources provide relatively abundant data and several pictorial illustrations. For vessels trading in West Africa in the second half of the eighteenth century -when the transatlantic traffic reached its apogee -, the slave deck contained two key features designed to control resistance. First, the spiked barricado, a huge timber barrier straddling mid-ships, dominated the slaver sailing from northern Europe. If it was not already in place from a previous voyage, then the ship's carpenter would build it on the way to Africa. Typically, it was topped with swivel guns and pierced with pea-shot loaded cannon, all pointing directly at the men as their space in front of the barrier filled up during loading 30 . The barricado's chief purpose, as surgeon Alexander Falconbridge and Captain John Newton pointed out, was to stymie revolts, but it also served to separate the men from the women and child captives 31 . A second security feature comprised two chains running fore and aft along the deck to which the men were always shackled in fair weather daytime hours. At night, captives would be forced below -the men unshackled -, but only two at one time to pass through the hatch way, whereupon they would be immediately re-shackled. Women and chil- Fórum dren were held on the quarter deck during daylight hours, and below deck at night, normally without restraints at any time, but always divided from the men by the barricado. The quarter-deck, of course, was the command center of the vessel, it abutted and overlooked the barricado and the shackled men beyond. Yet on South Atlantic routes to Brazil, such a barrier is never mentioned in the sources, and a 1743 painting of the Nossa Senhora de Nazaré e S Antônio (ID 8148; Figure 4) with a full complement of slaves on board shows no sign of this large defensive work 32 . The shorter voyages around the South Atlantic gyre together with the practice of embarking captives quickly made it unnecessary. Robert Walsh's drawing of the slaver from Bahia, Veloz (ID 1126), made just prior to the 1830 ending of the legal trade to Brazil, also shows no barricado although it does have a cannon -presumably loaded with pea-shot -pointing at the slave deck 33 .

Fórum
And so, we come to resistance -a topic now at the core of modern historiography on slavery. It is sufficient here to note that many historians view the St. Domingue revolution as the trigger point for the ending of slavery in the Americas and, ultimately, elsewhere. The events in what became Haiti and the massive half-century long naval campaign to suppress the slave trade were both dramatic reversals of well-established historical trends. There is no doubt that the overthrow of the richest colony in the Americas has had a greater impact on shaping the modern world than did the destruction of 1,600 slaving vessels and the liberation (however qualified the resulting "freedom") of the over 100,000 Africans found on board. The first was the result of the actions of the 400,000 enslaved people who had created St. Domingue's riches, while the second was effected by the navy of the greatest imperial power of the nineteenth century and, moreover, gave that power more than a head start in the later partition of Africa 34 .
One of the widely cited findings to emerge from the first appearance of the TSTD on a 1999 CD-Rom, was that African resistance to the slave trade as expressed through on-board slave revolts displayed a strong regional bias. Rebellions, in short, were much more likely to occur on vessels leaving from Upper Guinea -the coast ranging from modern Senegal to the western limits of Côte d'Ivoire -than from the more southerly regions of sub-Saharan Africa. Moreover, compared to these other regions, Upper Guinea was not a major source of captives, supplying fewer than 12 percent of total embarkations over a period of 350 years 35 . The 20 years of research since 1999 that have both lengthened and thickened the TSTD have simply reinforced this finding. While Upper Guinea may have supplied only 12 percent of all David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth? captives carried off, more than 40 percent of the vessels experiencing revolts came from that same area. Were Europeans avoiding what to them were troublesome sources of slaves and thus tolerating longer and more costly voyages to the south and east to reduce the incidence of rebellion? Were enslaved Africans therefore helping to shape the distribution of the transatlantic slave trade? There can be no doubt that for merchants, shipboard mortality was the single most important determinant of losses in any slave trading venture. And no doubt, too, that Africans influenced the size and direction of the traffic. But the easy association between revolts and the coastal origins of slaves carried off is now looking more complicated.
As Marcus Rediker commented -"the greater the number of people in the plot, the greater the chance of success, but at the same time, the greater the chance that someone would snitch." In short, there was a collective action problem, where, because of the high risk and the horrendous consequences of failure, individual captives had an incentive not to participate in a revolt. It turns out that insurrection was more likely on vessels containing smaller numbers of slaves, spending more time on the coast, and carrying a higher ratio of males. These were all characteristic of slaving ventures to Upper Guinea markets lacking the networks that allowed relatively quick turn-around times in places further south 36 . Particularly striking is the relative absence of rebellions in that portion of the slave trade that operated between Bissau and Cacheu to Amazonia after 1750 -a branch of the trade under the Portuguese flag. But there is evidence of runaways and rebellion in the barracoons of both these locations, including 31 "killed as a result of uprisings" in Cacheu in 1772 37 . Shi- Fórum pboard rebellions may have shaped the transatlantic slave trade, but the causes behind them had little to do with social structures, religion or values of the societies from which captives were drawn. Instead, they had more to do with the way the slave trade was organized.
As noted above, there was one area where the Portuguese slave trade overlapped with that of the French, British and Dutch -and that was the Bight of Benin. During the whole period of the slave trade, the Portuguese obtained over one million slaves there, about 18 percent of all the Africans they carried off to the Americas. Here, the Portuguese came closest to the ship-based slave trade of their northern competitors and began to face the same problems. According to TSTD, before 1760 this trade was confined to just three locations -the Dutch castle of Elmina, Epe and Ouidah -locations with rapid turn-around times, especially for Brazilian vessels offering preferred trade goods in the form of gold and tobacco rolls 38 . Nevertheless, immediately before Brazilian gold began to arrive on the coast, the Portuguese lost four ships to slave rebellions in four years in the Ouidah road, presumably while they were assembling their full complement of captives 39 . Moreover, this is the one branch of the Portuguese traffic where an illustration of a slave ship complete with a northern 38 Available in: <http://slavevoyages.org/voyages/EqJ9oq0n>; Vessels from Bahia typically called at Dutch Elmina to pay a levy amounting to ten percent of their outbound cargo. According to TSTD, between 1720 and 1740 such vessels also took on board an average of 95 slaves at Elmina (n=42), before proceeding to Epe or Ouidah in the Bight of Benin to purchase a further two or three hundred captives. Fórum European barricado has survived 40 . From the late seventeenth century on, it would seem that a fragment, at least, of Portuguese slave trading was ship-based 41 .
As well as a regional bias in resistance, a second major finding to emerge from the CD-Rom version of TSTD was the strong temporal pattern in shipboard rebellions and attacks from the African shore. More than one quarter of the total number of people carried off from Africa embarked after British and US abolition of the slave trade in 1807. Yet fewer than five percent of the 565 resistance incidents recorded in the current TSTD occurred in this same period 42 . Part of the explanation for this extraordinary decline is simply the southward shift in the center of gravity of the nineteenth century traffic 43 . Upper Guinea outlets to the Atlantic, along with those on the Gold Coast, were among the first to withdraw from the business -either because of British naval pressure, or, according to some scholars, an independent Islamic and African-based abolitionist impulse 44 . By 1830, almost the whole slave trade drew on the eastern and southern coast of Dahomey through to the Makonde region of what is now northern Mozambique. But by now, this vast region with a community of slave traders had embraced the Portuguese system. David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Fórum
Thus, the nineteenth century brought major changes to the way slaves were carried off from Africa. The French transatlantic slave trade ended with the outbreak of war with England in 1793 and Dutch involvement, already in decline since the 1770s, followed suit with the French invasion of 1795. The Danes, British and US had made participation of their own citizens illegal by 1807. While the French rejoined the trade in 1813 for a further 18 years, harassment from first the British and then the French authorities, meant that the northern European practice of trading from the ship gradually disappeared. The presence of slaves on board a vessel was usually incriminating in the era of illegal slave trading. Slaves were therefore increasingly held on shore until the vessel was ready to depart -the system that the Portuguese had established in the 16 th century. Thus, in the final half-century of the traffic, slaves were spending much less time on the vessel. Table 1 compares the average number of captives found aboard slave ships captured off Africa with the average carried off by vessels that escaped capture. In the early phase of suppression, the detained slave ships contained relatively few captives. At this point captains in West Africa were still trading in the northern European style by putting people in the slave hold as they purchased them. At the point of detention, vessels were partly loaded only. Over the next few decades, the number of captured vessels aboard rose steadily as traders held their captives in shore-based establishments until all were ready to board. In other words, an increasing share of the vessels that the British captured contained a full complement of slaves on board because of the widespread adoption of the Portuguese system. The fact that column 1 never quite matched column 3 is explained by several detentions taking place just as the slaves were in the process of embarking. In the resulting confusion traders were able to abort embarkation or land some of the captives.
But the major explanation for the apparent decline in resistance lies with an unrecognized and unexpected impact of British naval tactics. The first anti-slave trade patrol began in 1808 but comprised just two warships that cruised only as far south as the southern limit of what is today Sierra Leone. From 1810, however, the Admiralty in-David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth? creased the size of the squadron and extended the patrol range to north of the equator, partly in response to a misinterpretation of a clause in the 1810 Anglo-Portuguese treaty 45 . Where the British detained five slave vessels north of the line in 1809, they took 29 in 1810 and over the next quarter century averaged eighteen detentions per year, despite the fact that for most of this period, only Spanish, Portuguese and Dutch vessels with slaves on board were liable to capture 46 . These initiatives had little impact on the volume of the transatlantic traffic, which reached its highest annual level ever in 1829, but it did induce slave traders to spend less time on the West African coast. Compared to the long eighteenth century (1701-1809), the number of days spent on the coast in the following quarter century (1810-1834) declined by one quarter. In effect, African sellers of slaves were holding enslaved people in barracoons on shore for longer periods of time and were thereby absorbing the risks of rebellion (and escape) that had previously been borne by the owners, captains and crew of the slave ship.
Given that vessels would land an agent, unload merchandise and then put out to sea again before returning to collect enslaved persons, a significant part of time on the coast was now spent not at anchor, but in cruising without slaves (and sometimes without slave trading equipment), while the vessel's intended "payload" was assembled on shore. The opportunities to stage a shipboard rebellion were now restricted to just the length of the middle passage. But this, too, was sharply cut from 68 days between 1701 and 1809 to just 40 between 1835 and 1866, as copper sheathing on the hull, improved design, iron fittings replacing wood, and even steam propulsion, all came into use. As we might expect, advances in maritime technology impacted legal and illegal activities alike 47 . If slave ships in the nine- Fórum teenth century spent far less time sitting on the coast accumulating a full complement of captives than their eighteenth-century predecessors, then the collective action problem identified by Rediker and other scholars moved from the slave deck to the stockade. In fact, the British navy helped reduce the incidence of shipboard rebellions by reducing the opportunities for them to occur -West African slave traders adopting the Portuguese system in response to British naval pressure after 1809.
The trend in sex and age ratios also contributed to a decline in the incidence of revolts. TSTD contains 4,273 voyages with some information on the proportion of men and children carried during the last two centuries of transatlantic slave trading. Prior to serious attempts to suppress the traffic, men had typically comprised half of those on board. After 1809, this ratio declined to 42 percent by mid--century. The sharp increase in the share of children was more important, from 16 percent before abolition to 32.5 percent from 1810 to the Anglo-Spanish treaty of 1835, and then a further jump to 42.2 percent in last three decades of the business. Several voyages carried only children in this era 48 . The only eighteenth-century parallel to such a pattern was the small and specialized traffic in children to serve as house slaves, which operated down to mid-century from the Gambia and the Southern Rivers of Senegambia to Lisbon. The causes of the nineteenth century decline in average age remain unclear. Perhaps slave traders were attempting to make their ventures more secure, or possibly, as with non-human commodities, more efficient transoceanic transportation allowed the movement of lower-cost items -in this case, from a slave merchant's perspective, children. Either way, opportunities for revolt were far fewer after 1809 than before. America, 2010, p. 47-49. On voyage times it should be noted that while the Mozambique to Brazil voyage was longer at 66 days (n=284), it was still ten days shorter than its West African-Caribbean counterpart of the previous century.

David Eltis
Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Fórum
The word "barracoons" (shore-based stockades) appears infrequently in the English and French records before 1810; just as the word "barricado" disappears after 1809. Paintings and sketches of captured slave ships constituted a sub-genre of maritime art in the aftermath of abolition. Not one of these illustrations -usually executed by a naval officer present at the time of detention -shows a barricado. The 1835 Anglo-Spanish convention was the single most important slave trade treaty of the post-abolition era in terms of interfering with the traffic as opposed to making it formally illegal. It allowed the legal detention of vessels on the grounds of equipment rather than slaves on board. Article of the treaty specified that the presence of any one of nine items was sufficient grounds for condemnation. These included shackles, open gratings, an excess of water barrels, bulkheads, or spare planks, but the purpose of the "spare planks" was described as "for[the] laying down of…a slave deck" not a barricado 49 . All nine items specified would have been familiar to crews and captives from the very beginning of the traffic, but the wording made no reference to what had been the main defensive work of West African slave vessels prior to abolition. Adult males remained in chains, within the field of fire of heavy armament when above deck, and, like women and children, men would be moved below and re-shackled at night and during inclement weather. But as the time spent bringing captives on board and crossing the Atlantic fell, along with the average age of captives, West African slave traders dispensed with the barricado. Paradoxically, in contrast to the pattern in the plantation Americas, abolitionist pressures were apparently associated with a decline in African resistance rather than an increase.
Shorter voyages and fewer slave rebellions gave the Portuguese an advantage over their northern competitors, not least because they resulted in lower shipboard mortality. Slave trading profits in the transatlantic business -and probably in the history of slave trading everywhere -depended heavily on the mortality and morbidity of David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth? captives 50 . In addition, Portuguese slave ships were much less susceptible to capture. The eighteenth-century wars that severely disrupted the slave trade of the northern powers at regular intervals were rarely fought in the South Atlantic -where the Portuguese predominated. Between 1660 and 1807 TSTD contains records of 5,535 Portuguese slaving ventures. Only 47 of these terminated in capture by a pirate, or by another European power -usually the Dutch -, yielding an extraordinarily low capture ratio of less than one percent. By contrast, frequent wars further north ensured that the French and English lost seven percent of their ventures to capture over the same period 51 . As one might expect from these findings, the crews of Portuguese slavers were smaller than average, and, as Table 2 shows, they carried more slaves per crew member than did their French, English and Dutch counterparts between 1750 and 1810 (the only years for which comparative crew data are available). Nothing approaching the rich accounting data available for the Dutch Middelburg Company, the London-based Royal African Company, and some Liverpool and Bristol slave traders, has survived for Portuguese and Brazilian slavers. We cannot therefore make direct cost comparisons across national flags, much less create a comparative price series for slaves, but all the physical (as opposed to financial) productivity data -voyage length, shipboard rebellions, capture ratios, crew sizes -suggests that the many private slave trading merchants in Brazilian ports -and at least one of the several Portuguese monopoly trading companies -were able to deliver captives to planters in Brazil at a lower cost than could northern Europeans to their own Caribbean colonies 52 .

David Eltis
Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Fórum
What this suggests is that Iberian domination of the Atlantic world lasted beyond the break-up of the Iberian union in 1640. When the most powerful state in Europe first came to the Americas, it did not choose to settle lightly populated Patagonia, Roanoke Island, or the St. Lawrence River. It went straight to the most heavily populated (and thus resource-rich central America) and quickly took over the two most powerful polities in the Americas (the Aztec and Inca empires). From this point until at least 1800 Spanish America was the largest, richest, most heavily populated, as well as the most urbanized European imperial domain in the New World, stretching eventually from California to Buenos Aires. As the first on the scene with transoceanic transportation facilities, the Portuguese found themselves in a similar position in Africa. No European power at the time could hope to conquer an Old World sub-continent, especially well-populated West Africa, but they were able to select the African regions with the combination of off-shore islands and hinterlands amenable to modest military interventions and cultural interactions via lançados. The Portuguese thus developed a method for obtaining slaves in other parts of sub-Saharan Africa which ensured steady supplies at minimum shipping costs and with little interference from European competitors. They may not have conquered Africa, but for several decades they had the best trading goods -Brazil's gold and rolled tobacco. The northern powers were left with the ship-based trading option which meant more crew per slave, more elaborate ships, heavier mortality, and the additional armament necessary to both suppress revolts among slave captives and to ward off attacks from their European competitors.
The British and Dutch were the most successful of the northern nations, with the former taking over from the Portuguese temporarily as the leading transatlantic slave trading nation between 1751 and 1800. But the British never succeeded in displacing the Portuguese David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?
in Upper Guinea, the Bight of Benin, and Angola. The secret to the partial British and Dutch success was their strong economic growth, and their associated ability to make the metals and some of the textiles that Africans wanted. Ivana Elbl notes that the Portuguese were actually importing iron from Sierra Leone before 1520 53 . The north Europeans were able to tap into Brazilian exports of tobacco rolls and gold by trading with Portuguese ships on the West African coast in order to create the right kind of assorted cargo. There may have been few traders of Anglo-African descent in Africa, but the English developed close friendships and business relationships with key Africans. In the seventeenth and eighteenth centuries the British and Dutch used their rising naval and economic powers to attempt to break into Iberian domination of the Americas and of the slave trade. In the end, northern Europeans had to make do with the temperate Americas (plus Jamaica and some of the minor Antilles) and with a slave trade in Africa based on ships rather than bulking centers.
By the nineteenth century, the British at last achieved the power to seriously invade the Portuguese slave-trade system, but apparently no longer had the will to do so. Indeed, they spent most of that century trying to suppress the slave trade. In a counterfactual world, one can easily imagine an Anglo-Portuguese treaty as the British fleet carried off the Bragança royal family from Lisbon in November 1807 that would not only have allowed British goods into Brazil, but would have seen English investors based in Liverpool and London become the dominant slave traders supplying nineteenth century Bahia and Rio de Janeiro with captives. After all, French and London-based English slave traders had been trying to ship slaves to both ports in the previous century 54 . If Fórum argued, Brazilian slave owners and slave traders squandered their gains pursuing the "aristocratic idea", an influx of British investors and capital just as the slave-enabled coffee revolution was gathering speed would surely have set Brazil on to a radically different path of development as well as improving the fortunes of the British 55 . But it was a lost opportunity, given that 1807 was also the year that the transatlantic slave trade became illegal for British subjects. The actual Anglo-Portuguese treaty -in 1810 -that followed on from the displacement of the Royal family to Rio de Janeiro contained an anti--slave trade clause. The relationship between capitalism and slavery, it seems, was indeed more complicated than the proponents of both the old (Williams) and the new (Baptist, Beckert et al.) histories of capitalism are able to recognize. Too many regions have enforced the most exploitative forms of slavery for long periods of time without developing industrial capitalism (e.g. Portugal, Spain, Brazil), and too many others became capitalist without chattel slavery. Of the three industrial powerhouses of the early twenty-first century, Germany, the US and China (two of whom also dominated the twentieth), only one ever had a significant slave population. Northern Europeans were latecomers to Atlantic slavery; it was their manufacturing capabilities, financial intermediaries and state support that allowed them to take a few islands from the Spanish and to break into the Atlantic slave trade. Rather than the slave systems of the Americas enabling economic growth, it was economic growth that allowed the North Atlantic countries to develop a West African alternative to the Portuguese slave procurement system.
Perhaps the new historians of capitalism should return to the long-running debate among Marxists that Maurice Dobb initiated in 1946. In its early stages, the debate was highly Eurocentric and focused on industrialization, yet it completely ignored Capitalism and Sla-David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?
Fórum very. Ironically, the latter was recognized only when the World-Systems school attempted to redefine "capitalist agriculture" to include what was going on in eastern Europe and the colonial Americas. The exploited classes were thus not just landless laborers in Europe, but also coerced peasants and slaves. The narrower definition of class struggle -between owners of capital and the social group of whom they intended to become their factory laborers -makes more sense because of the timing issue 56 . Barbadian output in the late seventeenth century was about the same as a small English county and Jamaica's output matched that of a larger county half a century later as industrialization was getting underway 57 . Neither was sufficient to trigger massive structural change in Britain's economy. Except for St. Domingue, the large slave empires were enabled by economic growth rather than the cause of it, and in St. Domingue's case one can hardly point to evidence of industrialization in pre-revolutionary France.
The Dobbesian view is not as popular as it once was and does not feature in the new history of capitalism literature. But it does have the advantage of eliminating the need to shoehorn awkward facts about slavery and the slave trade in the Atlantic world into a developmental model of history. Applied to the US, Dobb's position has its own problems. Now that everyone agrees that slaves were highly productive -and many who were skilled -, it cannot account very easily for abolition and the US Civil War. Black abolitionists, black regiments, and slaves fleeing north have at long last been given their due in recent decades, but no one could have expected that blacks in the US emulate St. Domingue in their attack on slavery. The lead up to US abolition was not a conflict between classes, but between sections of the country. The more important question is why, if slavery was so efficient and profitable, would it be abolished at all? An ensla-David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?
Fórum ved factory labor force would have made as much economic sense as an Anglo-Portuguese treaty allowing Liverpool slave traders into the Brazilian slave trade in the early 1800s. The recent literature pretends the question does not exist.

***
Finally, we turn to the implications of the slave trading patterns described above for enslaved Africans, for whom the intrusion of the Northern Europeans into the slave trade had profound consequences through to the end of the slave trade. Over seven million Africans embarked on slave ships between 1701 and 1809, and almost half as many again from 1810 through to the end of the trade. A large literature exists on crowding, mortality, and the horrors of the middle passage, most of it derived from the extensive investigations of a committee of the British Privy Council in the late 1780s, narratives of the crews of anti-slave trade cruisers, and the massive seven volumes of evidence collected by the British Parliamentary Select Committees on the trade, between 1847 and 1850 58 . This literature presents a single middle passage experience comprising physical violence, shortage of water and food, pestilential disease, periodic cataclysmic revolts, wastage of lives of all on board, and circulating sharks awaiting the next corpse to be dropped overboard. Above all, the image of the Brooks slave ship is still viewed as a reflection of reality rather than as an exaggerated, highly schematic, and very successful attempt to attract public attention. Scholars and public alike have failed to recognize how captive experiences varied markedly over time and space. In short, despite the extensive -and mainly abolitionist -sources, many aspects of the typical lived experience of a slave transported to the Americas are not yet represented in the now extensive scholarship.

David Eltis
Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Fórum
How different was the slave experience on Portuguese vessels? Alonso de Sandoval collected information from hundreds of Africans in early seventeenth century Cartagena and wrote what is probably the best ethnological treatise on early modern Africa. Scholars have taken his work to mean that adult males were held below deck throughout the voyage, which an abundance of later evidence makes extremely unlikely 59 . We know that a key feature on all slave ships, as well as the shore-based holding areas, were long chains running across the space of imprisonment to which captives were shackled. But as noted above, the barricado was absent from Portuguese slavers. Obviously, Africans spent a significantly shorter period of time in such conditions in the Portuguese trade, but no Portuguese slaver has ever attracted the attention generated by their British and French counterparts. The Nossa Senhora de Nazaré e S Antônio from the mid-1740s is a poor substitute, though it is probably more accurate than the Brooks. Almost all surviving images of Portuguese slavers are from after 1807, an era when slave ships comprised a very wide range of differing shapes, sizes, rigs, and lay-outs. The best-known image in Brazil is a lithograph of the hold of a slave ship created by Johan Moritz Rugendas in the early 1820s 60 . Unfortunately, there is no firm evidence that the engraver had set foot on a slave ship. Numerous paintings hanging in maritime museums, often show a slaver accompanied by its naval captor, but at least we know the paintings were executed from life. Three such illustrations allow us a closer examination of changes in slave accommodation during the nineteenth century. They are the Isla de Cuba (1859, ID 4961 in TDTD), the Albanez (1845 and ID 3483) and the Diligente (1838, ID 2588).
While it would be impossible to visually capture this variety in shipping practices, analyzing three images does give some sense of David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth? Fórum differences to the conditions shown in the Brooks and Marie-Séraphique images 61 . Slave vessels had never been large by transoceanic sailing ship standards. But in the quarter century after 1820 the average standardized tonnage of a slave ship declined by 23 percent. Slave traders not only abandoned the barricado, some even abandoned slave decks after treaties allowing detentions based on the presence of slave trading equipment such as slave decks and excess water casks came into effect. Figure 5 shows a plan of the hold of the Isla de Cuba (1859 and ID 4961), which displays the barrels of water and provisions that occupied most of the space in a slave ship. Here the slave deck is in the form of planks stacked on either side of the hold lying ready to be laid 62 . Attempting to escape a conviction for slave trading, the captain argued in a US court that the planks were intended for sale on the African coast rather than for use as a slave deck, and that the barrels of water were for ballast only 63 . A similarly sketched plan of the hold of the ironically named Legítimo Africano (ID 3049), detained in 1835 -not shown here -shows a 50-ton vessel without a slave deck (or the planks to make one) and built for speed. Yet, it carried 190 people in an area of 400 square feet and with a deck height of just 1'8". How was this possible? Instead of a deck, the captain formed a makeshift platform by filling the spaces between the casks with bags of provisions. The key element that made this possible was that all but one of the 190 people on board were children. Such a pattern 61 Not included here is the well-known image of the Vigilante from 1822. The unknown draftsman populated the vessel with images of Africans copied directly from the famous Brooks' poster except that their distribution is limited to mid-ship. As a guide to how Africans travelled on a slave vessel, the published image was thus just as misleading as that of the Brooks. Nevertheless, the Vigilante's sketch became accepted as authentic. Fórum meant a dramatic reduction in security costs 64 . The British found an identical below-deck environment on the 45-ton Jesus Maria (ID 2071) with 246 surviving Africans on board of whom only five were adultsall, unsurprisingly, women 65 . Naval officers reported at least a dozen such cases after 1835, describing sand ballast or firewood filling the spaces between casks, or sometimes simply "hides laid on the tops of leaguers." 66 Scholars have addressed the issue of children in the slave trade by focusing on shifting cultural patterns within Africa, especially the large regional and ethnic variations in child ratios within Africa. However, the jump in the proportion of children carried from all regions in the nineteenth century was most likely in part a slave trader response to naval activity 67 .

David Eltis
Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Figure 5 [Anon], Plan of the Hold of the Barque "Isla de Cuba," 1859
Reproduced with permission of the National Archives, United Kingdom, FO84/1100, f. 174

David Eltis
Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Fórum
The painting of the slave hold of the Albanez (ID 3483) shortly after its interdiction in the Congo River in 1845 captures how Africans may have been transported in vessels lacking a fixed slave deck (Figure 6). The painter, Francis Meynell, is probably sitting on the forward stairs looking aft and Africans can be seen on the casks. Frequently reproduced, the image is certainly authentic, and, in this respect, it matches the drawing of the Marie-Séraphique. The painting is not intended as a depiction of conditions on the Middle Passage because the apprehending cruiser -HMS Albatross -had taken on board many of the captives prior to the long voyage to adjudication in Freetown. Thus, the viewer sees the real below-decks of a slaver, but only some of the captives. Even so, the image evocatively captures the chaos of a dimly lit slave deck. Africans are spread uncomfortably across the tops of barrels, some on mats, some on the bare wooden hoops; one captive sits on a latrine in the foreground, wrapping himself with his arms. Above the barrels, slaves perch and lie on wooden beams, some with their legs dangling over the ledge. Many more captives are crammed together on platforms running along the vessel's side, one of the only commonalities with the Brooks and Marie-Seraphique. Light pours in from the ceiling but only illuminates the captives in the center of the image. The fact that it is daylight gives some sense of how the Africans would have experienced the Middle Passage: packed below deck on whatever few inches of space they could find 68 .

David Eltis
Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Lt. Francis Meynell, Slaves below deck, 1845
While the Albanez and Isla de Cuba give some sense of the holds of illegal slavers, the recently unearthed painting of the Diligente (Figure 7) reveals the sheer mass of humanity that slave traders crammed onto their vessels. The Diligente (ID 2588) was a 174-ton brig depicted leaning in slightly toward the painter. It carried 475 Africans -survivors of 520 embarked at Lagos. Detained on its way to Cuba in 1838, Reproduced with permission of the National Maritime Museum, United Kingdom, A1818 David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?
Fórum the Diligente provides, at first glance, the most accurate depiction of the fair-weather day-time experience of captives for any period; not even the image of the Marie-Seraphique (Figure 3) provides such a view. But things are not quite what they seem. The deck shows fewer than half the number of captives that we can document as disembarking a few days after the detention. Furthermore, the seven blue-jacketed figures can only be the prize crew from HMS Pearl, not the original slave ship crew. Thus, the artist is probably on the quarter-deck of HMS Pearl as the naval vessel conducts its prize to Nassau in the Bahamas -the capture having taken place in the Caribbean, not off the African coast. The missing two hundred or so Africans are probably below deck as can be seen in the open hatch beside the mainmast. But here, as with all the other Portuguese vessels mentioned here, there is no sign of a barricado. There is no illustration of, nor indeed any documentary reference to, such a structure in the illegal era, something that makes sense given the need to conceal the vessel's intentions from the British Navy. Crewmen likely enforced the separation of male and female slaves through restraints and violence or, alternatively, kept captives below deck for the voyage. To derive a perspective of crowding on the Diligente we need to imagine double the number of figures depicted in the painting, crammed below deck.

David Eltis
Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Figure 7 Lt. Henry Samuel Hawker, The Portuguese slaver Diligenté captured by H.M. Sloop Pearl with 600 slaves on board, taken in charge to Nassau, 1838
Reproduced with permission of the Smithsonian National Museum of African American History and Culture, United States.
David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?

Fórum
Three illustrations cannot encapsulate the experience of captives in the nineteenth century slave trade, but they can indicate change in that experience over time. The illegal phase of the trade as represented in the voluminous reports of British naval officers communicates a sense of the wild west where, especially after the 1835 equipment clause, almost anything was possible. Apart from the open launches discussed earlier, slave traders used other strategies to economize on small spaces that fundamentally altered the African experience of the Middle Passage. In 1842, a 29-ton vessel bound from Ambriz to Brazil took off with 127 captives -over half of them were children. The height between water casks and the underside of the deck was just one foot two inches, and "one half of the slaves were obliged always to be on deck where they were so confined that every foot of the deck was occupied, while the reminder below were squeezed to excess." 69 Such variation in shipping practices does not permit easy distillation of the average experience, but scholars and the general public alike should look beyond the Brooks.
In summary, European and African interaction on the coast generated three broad categories of slave experiences over 370 years. The first typically involved vessels carrying fewer than 100 captives and shipped considerable produce to Europe from Upper Guinea, or, if going to the Americas, European merchandise and migrants. Confined mainly to the first half of the sixteenth century, the transatlantic voyages might have obtained their captives from the Iberian Peninsula, or they might set out from Iberian ports and collect captives from the Canaries or Cape Verde Islands en route. In such cases the slave experience would have included an additional voyage of several hundred kilometers from the mainland to these off-shore islands -as well as detention in barracoons. The vessels involved were caravels or galleons little different from their counterparts that plied the Atlantic and Pacific without slaves on board. The second category evolved from the first, the main difference being that as demand for David Eltis Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth? slaves increased, the vessel became a recognizable slave ship complete with dedicated slave deck, and more or less permanent shackles for the men. Crowding became intense, but time spent on the vessel was not much different. This system continued throughout the slave trade era, though it became increasingly a defining characteristic of the South Atlantic slave trade. The third category, associated with the northern European incursion into the business, comprised the same dense crowding, but a more heavily fortified vessel, larger numbers of crew, long periods spent on the ship both before and after leaving Africa, and greatly increased risk of slave revolts. We cannot be sure of the start date, but it was probably in effect for only 160 years and it is likely that fewer than half the 12.5 million captives carried off from Africa experienced it.
This third category is certainly the best known of these experiences -partly because of the images of the Brooks and the Marie-Séraphique. Perhaps, in the end even accurate contemporary depictions have to give way to 3D visualizations. The surviving plans of L' Aurore, published in 1984 by Jean Boudriot furnish an opportunity to move beyond the contemporary illustrations referenced in this essay. The 2019 version of slavevoyages contains a video that begins to show what is possible.

David Eltis
Iberian dominance and the intrusion of the northern europeans into the Atlantic World: slave trading as a result of economic growth?