The Atlantic demographics of Africans in Rio de Janeiro in the seventeenth , eighteenth , and nineteenth centuries : some patterns based on parish registers *

The article presents serial data on African Atlantic demographics in Rio de Janeiro in the seventeenth, eighteenth, and early decades of the nineteenth century, highlighting parish death, marriage, and baptismal registers. These represent partial, local findings from a broader demographic study underway on various urban and rural regions of slaveholding Brazil, based on a variety of historical registers, which are used in an analysis of sociodemographic patterns (age, occupation, kinship, disease, gender, price) and variations in African ethnonyms from 1650 through 1870.

T his article examines patterns of classification for enslaved Africans in Rio de Janeiro in the seventeenth, eighteenth, and first decades of the nineteenth centuries.The main idea is to assess the methodological potential for using ecclesiastical sources (parish registers) to explore variations in the nomenclature used to identify Africans (Klein, 2009).
Scholars of slave societies in different parts of the Americas have addressed the topic of the identities of Africans and their descendents.In the case of Brazil, the topic has also been approached from renewed perspectives (Parés, 2005;Reis, 1997;Slenes, 1999Slenes, , 1991Slenes, -1992)).A number of studies and researchers have focused on the terminology, classifications, and nomenclature used in the Atlantic trade and in slave societies to identify, situate, and indicate the geographic origins of Africans.Drawing from designations suggested by the slave trade (based on names of ports of embarkation and African locations); from characterizations published in newspapers about slaves for rent and runaway slaves; and from parish registers, bills of sale, inventory lists, and classifications of returned Africans, researchers have investigated the methodological potential of using these various designations to understand identity formation and identification systems for enslaved Africans.
In Brazil, parish registers provide information on the first generations of Africans and on 'nationals' who were baptized, interred, and married in regions that were the settings for early economic expansion and increased use of African labor, in addition to indigenous labor.In the case of Rio de Janeiro, these regions lay at some distance from the sugar plantations of sixteenth-and seventeenth-century Pernambuco and Bahia.While the sugar plantations in the captaincy of Rio de Janeiro, especially around the Guanabara Recôncavo, did not display the economic and demographic power of their counterparts in the Northeast, there were a substantial number of them in the late sixteenth century and throughout the seventeenth. 1An important study by geographer Maurício Abreu shed extraordinary light on the topic (Abreu, 2006).His key goal was to map the establishment of the first sugar plantations throughout the captaincy.Based on research conducted at more than twelve archives in Brazil, Lisbon, and Rome, Abreu produced an empirically cogent analysis of sixteenth-and seventeenth-century sugar plantations in Rio de Janeiro.His exploration began with colonial chroniclers like Friar Vicente do Salvador -who stated that in 1627 Rio de Janeiro had 40 sugar plantations, although it produced more "flour for Angola than. ..sugar" 2 (Salvador, 1982) -and also included Antonil, who in the early decades of the eighteenth century commented on Rio's economic prosperity and said there were 136 such plantations. 3Boxer indicated that an average of 20 to 25 caravels loaded with sugar left Rio de Janeiro for Portugal between 1638 and 1642 (Boxer, 1973).
According to Abreu (2006), it was not always possible to follow the history of Rio de Janeiro sugar plantations via documents.Some of these estates were built and quickly abandoned while others came to temporary standstills or saw their buildings undergo modification over the years.They were located not far from the city, near the urban core or else in the regions of Jacarepaguá, Inhaúma, Ilha do Governador, Irajá, Campo Grande, and Guaratiba, parishes on the semi-urban edges of the city.They also lay in the regions of Pilar, Marapicu Iguaçu, Suruí, Guapimirim, Magé, Inhomirim, and Jacutinga, in the The Atlantic demographics of Africans in Rio de Janeiro in the seventeenth, eighteenth, and nineteenth centuries Guanabara Recôncavo.Sugar plantations could likewise be found in the so-called 'banda d'além', an expression used in the sixteenth and seventeenth centuries to refer to land in São Gonçalo, Santo Antônio de Sá, Itaboraí, Itaipu through Maricá, Saquarema, and Cabo Frio.At the close of the seventeenth century, there were 131 sugar plantations in the captaincy of Rio de Janeiro.The few estates established in the late sixteenth century expanded in both territory and captives the following century, and in the eighteenth century they multiplied further, acquiring new owners and more land. 4

The first generations of slaves
What about labor on the sugar plantations and in the fields of the captaincy of Rio de Janeiro?Little is known about the impact of African demographics in this region in the seventeenth century and first half of the eighteenth.As part of this methodological exploration, I conducted an analysis of parish baptismal, marriage, and death records, examining the emergence of patterns of classifications and names that suggest a number of mechanisms for the identification and identity of African slaves and those born in the colony, especially the first generations.I have compiled aggregate data on the geographic origins of Africans in Brazil from the seventeenth through nineteenth centuries based on serial sources.Here I present more indices from the 'Nações Atlânticas' (Atlantic nations) database (phase Rio de Janeiro IV), with parish registers (Gomes, 2011). 5Baptismal records were rare for adult Africans in the seventeenth century, based on at least nine parishes: Candelária (1638-1709), Cabo Frio (1675-1700), Irajá (1677-1693), Jacutinga (1686-1690, 1691-1695, and 1696-1700), Magé (1677-1693), N.S. do Bonsucesso de Piratininga (1680-1700) 6 , São João Batista de Niterói (1671-1675, 1666-1670, and 1676-1680), and Sé (1700).Even if we bear in mind that forms of settlement were less dense and inconstant, that indigenous slavery was in play as well, and that the supply of African labor was precarious, the number of adult baptisms is limited, totaling only 97 registers.However, if we include baptisms of children and the classifications assigned to mothers, the figure reaches 1,200 (Table 1).The low number of baptized adults in Rio de Janeiro during the seventeenth

Flávio Gomes
century suggests that the new African arrivals had been baptized earlier, at trading posts on the African coast. 7Of the 92 baptized African adults, 13% are classified on their baptismal records as 'gentio de Guiné' (literally, heathens from Guiné) or Guiné and 87% as Minas. 8 Taking into account the origin of the mothers of the baptized children, there is a growing classification pattern of 'gentio de Guiné'/Guiné, not to mention a high index of undefined identification (20.2%), that is, baptized children of female slaves where no indication is made of the mother's origin.Given this rate of non-identification, the figures may refer either to the children of enslaved Africans or to captive indigenes who were still baptized.Mothers classified as Guiné/'gentio da Guiné' and mothers whose place of birth is not indicated represent 70.4% and 27.2%, respectively.For those born in the colony, classifications like Carijó and 'gentio da terra' (literally, heathens of the land) tally a maximum of eight records.It is likely that there were also few baptisms of enslaved adult indigenes, given that there was greater control in this region, and even bans, over the 'tropas de resgates' that were otherwise generally allowed to capture natives; strict control was also exercised against sending into this region indigenes who had been captured and baptized elsewhere, such as the old captaincy of São Vicente.So large numbers of indigenes born in captivity in the colonial setting may have been baptized and the origin of their mothers not mentioned.Another argument to be raised is that the baptismal records of slave children where the mother's place of origin is not specified might have camouflaged a growing number of captive children born to African women in the colony.As evidence against this argument, I would underscore the fact that a large rate of mothers were classified as Guiné/'gentio da Guiné' (Table 2).In considering this generic nomenclature, we must ask why the African mothers of some slave children were identified -even if only generically -but others were not (Lara, 1997).In the case of identified African mothers, a few are listed as Mina and two as Angola and Ganguela, terminologies used for West and Central Africans, respectively.The question is knowing which Africans -in terms of regions, territories, and provenance -were classified as Guiné/'gentio da Guiné'.
Available studies have strengthened the argument that they were Central Africans, especially from Loango/Luanda, a hypothesis that is reinforced by data from Behrendt, Eltis, and Richardson (1999) and by the research of Miller (2004).
The data on deaths (Table 3) confirm some of these hypotheses although the sampling comes from only three parishes: Cabo Frio (1695-1716 and 1678-1729), Jacutinga (1689-
The Atlantic demographics of Africans in Rio de Janeiro in the seventeenth, eighteenth, and nineteenth centuries  1721), and N.S. do Bonsucesso de Piratininga (1692-1700). 9There are 45 records for interred adults, 30 men and 15 women.The majority again consists of adult Africans classified as 'gentio da Guiné' (96%), of which 71.4% were men.The larger numbers of African men buried evince greater numbers of adult men embarked to Brazil.As to buried children, there is a concentration of Africans and a low index of 'crioulo' (born in Brazil), indigene, and 'mestiço' mothers.A total of 153 children of African women were buried, revealing a pattern of 94.7% 'gentio da Guiné' (Table 4).
Although the rate of Africans generically called Guiné/'gentio da Guiné' holds steady, marriage and death records display lower rates of unidentified origin.While the unidentified index reaches 20.2% for baptisms, it is 0% in the case of deaths, with marriages at 11% (Table 6).Some hypotheses can be raised about the naming of Africans, which may be more evident on death and marriage records for Africans who had spent more time in the  Diaspora, as compared with the baptismal records for new arrivals.In any case, we cannot forget the differences within this sample, which comprises a number of parishes and registers scattered among record books with fragmented periodicity.The sampling of baptisms encompasses a much broader number of parishes than the sampling of deaths.
In addition to differences across space, there may be temporal changes related to the parish priests who made the notations.Consistent patterns of classifications, changes, underreporting, and so on might also be associated with certain parish priests and specific ecclesiastical concerns, which are contexts that would interfere with the process of assessing changes in the forms of naming and classifying Africans and 'crioulos' in these parish registers.In terms of characteristics, not many differences were noted between more centrallying parishes in the city of Rio de Janeiro, those in semi-urban areas and the Guanabara Recôncavo, and others located farther away, such as in Cabo Frio and Maricá.In terms of patterns of classifying African 'nations', it is true that the data for Rio de Janeiro suggest that there was little definition, considering the designations 'gentio da Guiné'/Guiné.One hypothesis that cannot be discarded is that there was a predominance of Central Africans, based on trade out of Luanda, and of West Africans, coming from Upper Guinea.The possibility that West Africans set sail for Brazil out of Senegambia or The Atlantic demographics of Africans in Rio de Janeiro in the seventeenth, eighteenth, and nineteenth centuries Cape Verde and arrived directly or indirectly in Rio de Janeiro warrants further investigation, given the trade networks still in place in the first half of the seventeenth century (Table 7).These networks supplied the sugar regions of Pernambuco and Bahia as well as part of the Prata region and might also have sometimes supplied Southeast Brazil and the incipient economic life in the Guanabara Recôncavo and semi-urban areas of seventeenth-century Rio de Janeiro.The high unidentified rate also prompts me to consider an incidence of second-generation indigenes born in captivity and working alongside Africans on sugar plantations, cattle ranches, and cropland used for food production.New research on the slave trade from the seventeenth to the eighteenth century could confirm or disprove these hypotheses. 10 While all West Africans are called Mina, Central Africans -always limited in numberare classified as Angola, Congo, Ganguela, Luanda, and Monjolo.Classifications for the colonial population include Carijó, 'gentio da terra', 'índio', 'mameluco', 'mulato', and 'pardo'.The indigenous population accounts for 34.5% while mestiços represent 50%.

Demographic transformations
We do not know if Atlantic demographics changed very much in the seventeenth century in Rio de Janeiro, despite the impact of the slave trade, which involved numbers, provenance, and origins of Africans.Furthermore, local realities -with diverse Atlantic trade networks -may have produced different scenarios.Denise Demétrio analyzed slave families in the Guanabara Recôncavo from the turn of the seventeenth century to the eighteenth and pointed to a slave trade conjuncture quite similar to the one I have suggested for various regions of Rio de Janeiro: a small number of baptisms of adult Africans, a general tendency for Africans to be registered under the classification Guiné or 'gentio da Guiné', and a substantial number of West Africans identified as from the Mina nation.
Still, some colors of the African landscapes in the Recôncavo may have changed in the eighteenth century.Based on a sample of 1,167 baptismal records for adult Africans and children born to African mothers in the eighteenth century in the parishes of Campo Grande, Guaratiba, Ilha do Governador, Inhaúma, Irajá, and Jacarepaguá (Table 9), some trends in the eighteenth-century African pattern can be observed in areas near the Guanabara Recôncavo.
The low register of Central Africans (Angolas and Benguelas) is evident.There is likewise a relatively low register of West Africans, called Minas, perhaps baptized in urban parishes of Rio de Janeiro near the Valongo slave market, the starting point from which many were sent to mining regions, especially in the first quarter of the eighteenth century.There continue to be records of a significant number of Africans called 'gentio da Guiné', which might be either West or Central Africans, as pointed out for the seventeenth century.

Flávio Gomes
Ignoring the variable 'gentio da Guiné', we have 52.5% Angolas and 40.4% Minas.The limited evidence of baptisms of Africans (only 18.25%) suggests that the Rio ranchers and farmers who bought captives for their land either baptized them at urban parishes in Rio de Janeiro -a hypothesis in part confirmed by the studies of Mariza Soares (2000) -or bought many who had already been baptized at African trading posts.Large numbers of Africans were brought into sugar plantations in the captaincy of Rio de Janeiro in the late seventeenth century and early eighteenth, as reflected in the pattern of 'crioulo' baptisms (children of African women), which come to 81.75%.
The Atlantic demographics of Africans in Rio de Janeiro in the seventeenth, eighteenth, and nineteenth centuries In the case of patterns for 'crioulos' whose African mothers have been assigned a classification, a large number continue to be identified as 'gentio da Guiné', but we do see an increase in Angola and Benguela Africans -thus, in women not baptized in these parishes but who went there to baptize their children.Among other African nations, women called Congo, Ganguela, and Rebolo also stand out.Not counting women registered as 'gentio da Guiné', we observe patterns of women (mothers) identified as Angola and Benguela at 61.5% and 16%, respectively, while Mina African women (mothers) tally 14.6%.For the eighteenth century, in urban parishes of Rio de Janeiro, Soares' study offers baptismal rates for Africans and 'crioulos' from a sample of 9,269 baptisms, of which 28.7% involved adults.For the parish of Sé, during 1718-1726 and 1744-1750, West Africans varied from 86.2% to 88.6% of adult African baptisms.Dividing the records for the so-called 'gentios majoritários' (Soares, 2000, p.105) -or 'heathen majorities', that is, the most common classifications in the registers (Guinés, Minas, and Angolas) -we observe 40% Guinés, 23.51% West Africans (Minas), and 36.49%Central Africans (Angolas) in the parishes of Sé, Santa Rita, Candelária, and São José from 1718 to 1760.As the classifications Angola and Mina appear more often, registers of adult Africans baptized as 'gentio da Guiné' decrease, especially in the latter half of the eighteenth century.For these same urban parishes from 1751 to 1760, among baptized 'crioulos', African mothers are recorded as 68.9% Angola and 31.1% Mina (Soares, 2000, p.80, 84, 105, 126).In comparative terms, the indices of West African women (mothers) are smaller than the index of baptized adult West African women and overall baptisms of West Africans.This may be explained by the fact that West Africans were not only fewer in number but also married less and their women had fewer children.
Let us now look at some data for the parish of Campo Grande, comparing 'crioulos' and Africans who appear in parish marriage registers.There are 533 Africans and 417 'crioulos', including 'pardos' and 'cabras'.The high rate of Africans stands out, suggesting that the slave trade had an impact between 1710 and 1770, which was likely a period of demographic recomposition of some sugar mills and plantations in the Recôncavo.In any case, the rate of non-identification is significant, corresponding to one-fifth (20%) of the marriages.Associating gender with place of birth, we observe that the number of African women (268 and 50.2%) surpasses the number of African men (265 and 49.8%).Given that the bulk of the Africans brought over by slave ships were male, the numerical superiority of women is striking.Yet we cannot forget that the slave population in question here is not the absolute adult slave population for the parish but rather the population listed in the marriage records for slaves, that is, only those slaves whose marriages were sanctioned by the church, and always in pairs of men and women.In the case of this study, we only know how many Africans and 'crioulos' married at the church in a given period.
The designations for Central Africans -Angola, Benguela, Ganguela -rank first, reaching 44.7%.There are also varied numbers of Central Africans from the areas of North Congo (especially Cabinda and Congo), and from northern Angola (Cassange, Luanda, Monjolo, and Rebolo). 11A large number of Africans are classified as 'gentio da Guiné' (241 and 45.2%), but there is no way of knowing whether these were Central or West Africans (Table 10).When we exclude the variable 'unidentified', Central Africans break down into 57.2%Angola, 18.8% Benguela, and 4.8% Ganguela.For Central Africans from the north, 3.8% are Congo.West Africans total 8.6%.
The limitations of this type of serial source notwithstanding, the death records do allow us to glimpse transformations in the system of classification and nomenclature, and thus of African demographics.These would allow us to identify the profile of the African who, although baptized in Africa or during the Middle Passage, had recently The Atlantic demographics of Africans in Rio de Janeiro in the seventeenth, eighteenth, and nineteenth centuries disembarked and/or fallen victim to the high mortality rates typical of the post-disembarkation period.We can also take into account the redistribution of Africans and their permanence in different areas -for example, new arrivals who were baptized in Rio's urban parishes but then sold and transferred to areas of the Recôncavo, semi-urban regions, or the urban center of the city of Rio.
Let us first examine the death records for the parish of Candelária.There are 262 records (192 men and 170 women) for 1724-1736 (Table 11) and 87 records (51 men and 36 women) for 1793-1800 (Table 12).
Considering the totality of classifications and identification of all Africans (separating out those generically identified as Guinés, i.e., 18.8%), there are 33.3%West Africans, of which 88.7% are identified as Mina while the others are designated as Cabo Verde or São Tomé.It is interesting to note that the number of Mina women is higher, that is, 60.9%.During this period, East Africans account for a very small percentage: 1.8%.Central Africans total 64.9%, of which 74.8% are classified as Angola and 17.3% as Benguela.Another 4.1% are called Congo, while the others are listed as Ambaca, Ganguela, Loango, Monjolo, Quissama, Rebolo, etc.
Although the sample was much smaller, death records for Africans in the Candelária parish were calculated for the late eighteenth century, during the period from 1793 to 1800 (Table 12).

Total records 87
A comparison of Tables 11 and 12 initially reveals a tendency for the classification Guiné to disappear, accounting for only 4.6% versus 25.9% during the previous period, that is, 1724-1736.West Africans are also underrepresented, at 4.6%.Classified as Angolas, Central Africans continue to predominate, at 41.4%, while Africans from the southern part of this region (Benguela and Ganguela) tally 28.7%.Monjolo and Rebolo Africans total 6.8%; in 1724-1736, they accounted for 1.5%.
The most significant differences appear in the semi-urban parishes of Engenho Velho and Jacarepaguá (Tables 13 and 14) at the beginning and end of the eighteenth century.
We have two very limited samples for the first quarter of this century, with a total of 4 records from Engenho Velho for 1725-1735 and 12 records from Jacarepaguá for 1730-1738.West Africans total 54.5%.Angolas account for only 20% of Central Africans while those classified as Congos represent 60%.There is a notable paucity of Africans classified as Guinés (only 1).
For the second half of the eighteenth century, the African profile changes on the death registers from these two parishes.For Engenho Velho, we have samples (380 records) that reflect these variations.The difference between men (269 and 70.8%) and women (111 and 29.2%) in the records for Africans buried at this parish depicts the flow of the Atlantic trade patterns, with 7 men for every 3 women.The classification Guiné remains in frequent use, at 24.6%, a figure that was 75% lower in the 1789-1800 period, compared to urban parish registers.West Africans -most of whom were classified as Mina -tally 8.8%, while there is no mention of East Africans.Angolas represent 58%, and Africans from the south total 19.8%, between Benguelas and Ganguelas.Africans from North Congo (Congo) account for 7.8%.
We observe a similar African profile in the parish of Jacarepaguá for the late eighteenth century; what differs is that the classification Guiné all but disappears (only 1) among the 204 records (Table 14).West Africans tally 4.4%, including therein Africans from Upper Guinea (classified as Cachéu).The largest category is Africans classified as Angola: 86.6%.
The data in Table 15 allow us to infer more about general patterns and tendencies in the use of classifications like Guiné and 'gentio da Guiné' (which may simply refer to an African from any region) in the first half of the eighteenth century and the continued use of the classification Mina for West Africans and Angola for Central Africans, with this use increasing towards the close of the century.If some classifications and nomenclature are still used, albeit less frequently -like Cabo Verde, Quissama, Monjolo, Rebolo, and Caçanje The Atlantic demographics of Africans in Rio de Janeiro in the seventeenth, eighteenth, and nineteenth centuries -others appear more rarely, like Ambaca, Loango, Luanda, and Libolo, not to mention nations like Bojolo, Bamba, Mogumbe, and Bababuína.Comparing African patterns of identification -nomenclature and classifications -for urban and semi-urban parishes in the eighteenth century, 15.9% (only a slight change) are categorized as Guiné/'gentio da Guiné'.The largest differences appear for West Africans, which display a pattern of 11.9%; in urban parishes they reach 20.3% but only 5.7% in semi-urban (rural) parishes.This evidence points to a larger concentration of Mina Africans in the central areas of the city, engaging in urban services.
Based on death records from the urban parishes of Candelária and Santíssimo Sacramento, Sweet (2003, p.45-46) identified 258 names of nations, not taking into account those reported as 'gentio da Guiné'.Central Africans account for 66.7% (primarily Angolas, Benguelas, and Ganguelas), and West Africans (mostly Minas), for 27.5%; the latter pattern confirms the hypothesis that the West Africans who passed through Rio de Janeiro ports in the first half of the eighteenth century were re-exported to mining areas and also sold to residents of the city of Rio.
All data presented so far suggest a pattern of marked incorporation of Africans -especially Central Africans -into the Rio de Janeiro Recôncavo in the second half of the eighteenth century.With more series or other available sources that could afford a more thorough examination of this process, it might be possible to suggest that the waves of Africans brought to Rio's farmlands at the close of the seventeenth century and first decades of the eighteenth might have been both West Africans (including those from Upper Guinea) as well as members of more scattered groups (some areas) of Central Africa.After 1740, Central Africans eventually predominated, in more concentrated groups.By 1760, a sizeable first generation of 'crioulos' -children of these Central Africans -prevailed among rural slaves in the Recôncavo.This is not to say that the routes and networks linking Rio de Janeiro to the Mina Coast -via Salvador or not -had vanished.Although fewer in number, West Africans continued to arrive in Rio de Janeiro in the second half of the eighteenth century, staying in the city but also being transferred to the captaincy's rural regions. 12

Nineteenth-century impact 13
There was an even greater impact from the Atlantic slave trade to Rio in the first quarter of the nineteenth century.According to Goulart (1975), 570,000 Africans disembarked in Rio de Janeiro from 1801 to 1830 alone.Based on Karasch's calculations, this figure -for the period between 1800 and 1843 -surpassed 600,000 (Karasch, 2000, p.67 and ff.).Revising calculations from Philip Curtin's classic studies, Eltis emphasizes that between 1811 and 1830 alone, the number of Africans who entered the port of Rio de Janeiro reached about 470,000 (Behrendt, Eltis, Richardson, 1999, p.21-32;Eltis, 2000, p.224-257).In addition to Eltis's revised figures, considering the volume of the slave trade to various regions of the Americas, the most conclusive findings for Rio de Janeiro are those in the studies and quantification methodology originally developed and presented by Klein (1973Klein ( , 1978) ) and only revised by Florentino (1995) based on records of ship disembarkations.
Mary Karasch's pioneering study remains the most comprehensive examination of the origins of the Africans who came to nineteenth-century Rio de Janeiro. 14In a sampling The Atlantic demographics of Africans in Rio de Janeiro in the seventeenth, eighteenth, and nineteenth centuries based on customs notes, documentation on captured slave ships, and tax, arrest, and burial records from Santa Casa de Misericórdia, she divided African origins into several focus areas, classifying them as West, Central-West, and East African, in addition to those considered to be of 'unknown African origin'.Drawn from a variety of sources, her data projects the concentration of some major groups of Africans in Rio de Janeiro.The main groups are Congos, Angolas, Cabindas, Benguelas, Caçanjes, and Moçambiques, among others.West Africans accounted for 1.5% to almost 7% (depending on source and period) of the total number of Africans.Central Africans (divided into North Congo, northern Angola, and southern Angola) represented 79.7% and East Africans, 17.9%.Karasch also noted 37 names of unknown origin.There are also about 500 ethnonyms for various origins of Africans disembarked in Rio de Janeiro (Karasch, 2000, p.35 and ff.).
The period of 1801 to 1830 was analyzed in this review of these patterns, based on over 36,000 records of adult baptisms and 18,000 baptisms of 'crioulo' children of African mothers in the urban parishes of Candelária, Santíssimo Sacramento, São José, Santana, and Santa Rita.Using parish registers -excluding incomplete information and omissions -the universe comprised 30,629 baptized Africans, including some 17,000 Africans who were identified by nation.The impact of African demographics on Rio de Janeiro was first evaluated by considering the baptisms of adult Africans.East Africans -that is, Moçambiques, Quilimanes, and Inhambanes -accounted for 34.5%, which confirms Karasch's observations about the 'nations of Rio' and more recent studies on the slave trade that show the predominance of the trade from the Indian coast to Rio de Janeiro in the nineteenth century.What was surprising were the West African demographics.In some studies available on the slave trade -paradoxically, considering only the last decade of the eighteenth century, that is, from 1790 to 1830 -the demographic representation of West Africans varies as much as 2.9%, based on the arrival of slave ships.In the records of burials, arrests, and seized slave ships studied by Karasch, they reached 6.34% by 1852.But the baptismal registers for Africans that I culled from urban parishes change this picture considerably, since the number of West Africans almost doubles Karasch's highest estimates made on the basis of other sources.In other words, roughly 11% of Africans baptized in Rio were West Africans (mostly identified as Mina or Calabar).
The explanation for this difference and increase can be explained by intraregional trade in enslaved Africans in the first half of the nineteenth century, especially of West Africans transshipped in Bahia (Salvador) and sold in Rio de Janeiro (Graham, 2002).As an initial hypothesis, I posit that this increase in the number of West Africans suggests that, in addition to those slaves who entered Rio via the Bahian trade, we should also consider ships from the Mina Coast whose final destination and point of sale was Rio de Janeiro, whether or not they stopped in Salvador.This means that even after the West African trade was banned (1815), West Africans continued to arrive in Rio de Janeiro and be baptized.They may not necessarily have been sold to slave owners who were established in Rio de Janeiro but may have been re-exported to southern ports, especially Desterro, in the state of Santa Catarina, Porto Alegre and Rio Grande, in Rio Grande do Sul, or Banda Oriental (future Uruguay) in the first decade of the nineteenth century.This would mean that a direct slave trade continued between Rio de Janeiro and the Mina Coast in the nineteenth century, supplying southern regions and later, in the 1820s, the coffee-and sugar-growing areas of the Paraíba Valley and the northern area of the captaincy of Rio de Janeiro.
Although all available studies depict Central Africans as the majority of the nations of Rio, there are interesting differences.Central Africans were also among the majority of those baptized but at only 52.5%; this lower rate of baptisms may hide the underreporting of Africans in the city, as Angolas and Benguelas account for just 2.4% and 2.8% of adult Africans baptized.At the same time, rather than suppose that they are underrepresented here, we might suggest that many Central Africans embarked in Luanda and Benguela reached Rio already baptized, whether at trading posts, in African coastal villages, or even aboard slave ships.This certainly does not appear to have been the case with Central Africans from areas of North Congo.Out of every five Central Africans baptized, four were classified as Cabindas.
Table 16 shows the percentages of baptized adult Africans (sample = 17,113), grouped according to the nation by which they are identified in parish registers.Those called Moçambique and Cabinda stand out, accounting for over 60% of all Africans baptized in urban parishes, perhaps evincing a process of generalization of the classifications of nations among baptized slaves, in addition to the strong impact of the slave trade from Central and East Africa, given the differences that appear in data from postmortem inventories.In any case, considering Central Africans alone, 76% are from North Congo, 5.8% from southern Angola, and 18.2% from northern Angola.The Atlantic demographics of Africans in Rio de Janeiro in the seventeenth, eighteenth, and nineteenth centuries With fewer differences than in the eighteenth century, nineteenth-century death records for urban parishes from 1810 to 1830 confirm baptism patterns.Death records for 829 adult Africans interred at the central parish of Candelária were examined, revealing a pattern of 64.8% men and 35.2% women (Table 17).The nomenclature or classification 'nação/de nação' (nation/of a nation) (275 records), the absence of any classification at all, identification as new African arrival (52), and generic designations such as Coast of Africa, East Coast, and Coast of Guinea (10) accounted for 40.65% of African death records, contributing to a non-definition of African patterns in the nineteenth century.Still, records for 492 interred Africans have some form of identification.At 4.9%, the number of West Africans comes close to the average for patterns found by Karasch, Klein, and others.East Africans seem to be underrepresented, at 5.5%.An expressive number of Central Africans are from southern Angola (Benguelas), that is, 27.2%.There are also a number of Africans from North Congo, with Cabindas and Congos accounting for 34.7%.Overall, and including non-identification, underreporting, and generalizing terminologies ('of a nation' only), there is a preponderance of Central Africans of the classifications Benguela (14.5%),Congo (13.9%),Angola (6.3%), Cabinda (4.6%), Moçambique (3.2%), Rebolo (3.1%), and Mina (2.1%).
Total records 829 557 (absolute number) These patterns change little if we analyze the death records of 'crioulos' born to African mothers.There are 557 records, of which 66.9% are unidentified (missing data or listed as 'de nação').In those cases where the nation has been identified, West Africans account for 10.9%; East Africans, for 6%; and Central Africans, for 83.1%, with the largest groups among the latter being Benguelas (36.4%),Cabindas and Rebolos (12.5% each), and Angolas (9.2%).Table 17 shows the percentages of Africans and African mothers, taking into account all records (including those lacking any identification).
Data on nations obtained from the inventories (Table 18) suggest a greater balance in representations of African groups in the city and its surroundings, perhaps revealing that part of these classifications or terminologies only emerged in the dialogic experience of Africans and in the sociodemographic context of which they were part.The percentage of West Africans drops to 4%, approaching the variations found from other sources by Karasch and others.However, I do not know if these Africans were redistributed to other regions, although baptized in Rio de Janeiro.New destinations could include the south or interior regions of the province, such as Campos dos Goytacazes, or food-producing areas that supplied the city of Rio de Janeiro, such as Itaguaí and Mangaratiba, or even major coffeegrowing regions like Vassouras, Valença, and Paraíba do Sul.The number of East Africans is also substantially lower: the figure surpassed 30% in baptismal registers but in inventories it is only 10%.These East Africans -the overwhelming majority of whom were men -were 1 9 The Atlantic demographics of Africans in Rio de Janeiro in the seventeenth, eighteenth, and nineteenth centuries most likely being re-exported to the Paraíba Valley and especially to southern Minas Gerais, including cities like Juiz de Fora, Mar de Espanha, Campanha, and Barbacena.Central Africans increase to 86% in inventories, and their trade areas appear to be better distributed, with preponderance not only of North Congo (Cabindas) but also northern Angola, especially Rebolos and Caçanjes. 15There is also a higher incidence of Central Africans identified simply as Angola.Their demographic representation jumps from 2.4% for those baptized to 13.2% of those inventoried.An even larger difference is noted for Benguelas, where the percentage climbs from 2.8% to 22.2%.Paradoxically, the classification Cabinda drops sharply among inventoried Africans, standing at only 9%.
What is most notable is the increase in the number and scope of the classifications or terminologies of nations in the inventories, compared to what is observed in parish records.The classifications Ambaca, Baca, Bambuíla, Ambuíla, Camundá, Luanda, Moange, Mocumbe, Mogumbe, Mofumbe, and Quissama account for only 27 records of baptized Africans but surpass 300 in inventories, corresponding to 5% of Africans.It should be noted that the classifications Moange, Mocumbe, Mogumbe, and Mofumbe (which I have classified as 'other Central Africans') -are virtually absent in baptismal registers but reach 3% in inventories.West Africans called Mina tally only 3.3% while Rebolos rise from 1.6% in parish registers to 8.3% of inventoried Africans.For southern and northern Angola, after those classified as Angolas the main groups are Rebolos, Caçanjes, Monjolos, and Cabundás (Table 18).Comparing the nations that appear in the parish registers and inventories with Karasch's pioneering study and the numbers based on ship arrivals, significant differences can be observed (Table 19).Let us look at East Africans.According to Karasch's data, this figure could range from 2.3% to 26.86% between 1790 and 1850 in samples of slave ships, reaching 17.9% of the total number of Africans in the city.In samples of burial records from Santa Casa de Misericórdia for the years 1833, 1838, and 1849, East Africans reach 18.4%.Tax records from the Rio Police for 1832 indicate 26.37%.Clearly there are significant variations in Karasch's own data.The data presented here on baptisms are closer to those found in tax records, that is, 34.5%.However, data from inventories show a marked reduction in the number of East Africans.I suggest that the following hypothesis might account for this: parish registers capture the heavy arrival of East Africans to the Rio de Janeiro market, where they were baptized in urban parishes and generally described as Moçambique.Data from slave ships and mortality rates strengthen this hypothesis.In the case of mortality, their higher death rate also reflects the logic of the slave trade, which meant longer voyages and a higher mortality rate.So greater numbers died during the Middle Passage and certainly the weakest were the most likely victims of mortality in Rio de Janeiro, which killed many enslaved Africans.Their low numbers on inventories reflect the workings of the slave market.Not all East Africans (and others) baptized in the city of Rio were sold and remained there.Although they were baptized there -or buried there after dying during the Middle Passage -a considerable number were sent to the Paraíba Valley and southern Minas Gerais.Another hypothesis to explain the low index of East Africans in the inventories is that the sample for the years 1801 to 1830 did not record owners who bought slaves from 1820 on, when the importation of slaves grew, along with slave ownership by small farmers in urban and semi-urban areas.This younger generation of slave owners who purchased a good number of Africans (perhaps East Africans) may only have been recorded in sample inventories for the decades of the 1840s and 1850s.
For West African patterns, I posit other hypotheses.They total only 1.5% in the customs records for the slave trade but reach over 6% of interment and tax records.The latter percentage resembles the pattern for slave ships from areas of embarkation in West Africa that landed in Rio de Janeiro from the late eighteenth century through the first quarter of the nineteenth.However, they represented 11% of baptisms, despite representing just 4% of inventoried slaves (and in the case of the African mothers reported on the baptisms of 'crioulos', the figure is 7.1%) (Table 20).The first of the hypotheses I raise is the underreporting of arrivals of slave ships, which, following the 1815 ban, continued to dock in Rio de Janeiro with West Africans aboard, without any indication of African port of origin but only of their last port of call.Furthermore, vessels that first docked in Bahia but only disembarked Africans in Rio de Janeiro may also have been underreported.The 4% figure observed in the inventories could also be explained by the combination of urban estates and rural properties in semi-urban areas and the Guanabara Recôncavo, complicating the characterization of a more urban African scenario where West Africans stood out, both in Rio and in Salvador, as opposed to more rural settings where the West African presence may have been less dense.
For data on Central Africans, a comparison of sources where they appear as nations is even more indicative.In Karasch's data, the figure ranged from 96.2% to 66.59%.According to Florentino's data, Central Africans represent 82%.But in the parish registers, they constitute just 52.5% and in the inventories, 86%.The key is the emergence of variations in the terminology of nations; many of these are barely visible or are not mentioned in baptismal records but are plentiful in inventories, as we have seen.Karasch also found a large number of nations in burial registers, not to mention the documentation of captured slave ships, where dozens of ethnonyms appear.In my view, these variations could be due The Atlantic demographics of Africans in Rio de Janeiro in the seventeenth, eighteenth, and nineteenth centuries slightly more than 5% of baptized Africans, probably because they were christened on the African coast or during the Middle Passage, but they account for some 46% of African women who are the mothers of 'crioulos' baptized in Rio's urban parishes.At the same time, there is a considerable decrease in the number of East African mothers, at just 6%, while East African women surpass 24% of baptized adults.The explanation could be an even sharper disproportion between men and women brought from East Africa by the slave trade (Florentino, 1995, p.239).Rebolos account for 1.6% of baptized Africans but 9.9% of African mothers.Probably by 1850, a significant portion of 'crioulo' slaves made up the second generation of enslaved Africans in Rio.Based on parochial records from the urban parish of São José, it was shown that 'crioulos' were more likely to choose their children's godparents and they did so both from the nearby slave community and from among the free populations.In the case of marriages, there is an emphasis on the relationship between the movement of the slave trade and intra-and inter-ethnic unions.
During the period from 1790 to 1830, however, we find that this trend reverses.As the volume of the slave trade increased, there was a tendency for Africans of varied origins to intermarry (Ferreira, 2001, p.56-57).

Final considerations
In different contexts, Africans invented themselves in terms of nations and other arrangements, forming networks in an extraordinary Atlantic setting.Bodies, languages, and minds were indelibly re-marked in social and ethnic terms.The identities of Africans -and of 'crioulos' -were not single but manifold (Farias, Gomes, Soares, 2005).For The Atlantic demographics of Africans in Rio de Janeiro in the seventeenth, eighteenth, and nineteenth centuries example, underlying the generic category of Central African, there were hundreds of thousands of Africans and groups who embarked in the south, north, and center of Central Africa -taken from the coast or hinterland -and disembarked and lived in Rio de Janeiro, where they discovered "new social identities beyond such sites, and already multiple ones that formed in their agonizing progress toward the coast" (Miller, 2008, p.53).
For the Atlantic demographics of Africans in the Americas, there could have been different combinations of names, terminologies, and classifications that would appearor disappear -in more merged or converging forms, or more scattered and isolated, depending on source and period.The concentration of more general classifications in records of ship arrivals or adult baptisms therefore differed from burial records and even the evaluation of inventories, which show not only new African arrivals but those who had spent some time in the Diaspora, and it is possible that either they themselves had come up with or reported their nation or someone else had, like their master or an African peer, unlike the case of recent arrivals for whom we basically have generic information -in the case of the baptized -passed along by slave traders or by direct intermediaries in the slave trade.In part, this is what Midlo Hall suggests, likewise based on a broad database: "The Louisiana Slave Database 1719-1820, a sophisticated and detailed tool about slavery, shows us that when Africans were first sold, their ethnicities were rarely indicated, and therefore slaves only began to identify themselves ethnically after several years of experience in America.The longer the Africans remained in Louisiana, the more often their ethnic groups were identified" (Hall, 2005, p.31-32).
Ecclesiastical registers, not just of baptisms but also of deaths and marriages, can be used in aggregate analyses to understand both the classification system for Africans (nomenclature) as well as differences across space and time.By cross-referencing sources and compiling databases through academic, and transnational, cooperation, more panoramic approaches to systems of identification and identities of Africans in the Americas can be achieved, based on a variety of sources and their respective methodological natures.

Table 1 : Nations of baptized captive adults and baptized children in Rio de Janeiro, seventeenth century Nation African adults Men Women African mothers of baptized 'crioulos'
Source: Various parish registers (Arquivos da Cúria Metropolitana do Rio de Janeiro, Cúria de Niterói e Cúria de Nova Iguaçu).

Table 4 : Africans (mothers and children) on death records, Rio de Janeiro, seventeenth century
Source: Various parish registers (Arquivos da Cúria Metropolitana do Rio de Janeiro, Cúria de Niterói e Cúria de Nova Iguaçu).

Table 3 : Nations of adults and mothers of children interred in Rio de Janeiro, seventeenth century Source
: Various parish registers (Arquivos da Cúria Metropolitana do Rio de Janeiro, Cúria de Niterói e Cúria de Nova Iguaçu).

Table 10 : Nations of Africans from the parish of Campo Grande, based on parish marriage records, 1745-1797
Source: Slave marriages in the parish of Campo Grande, 1745-1797 (Arquivo da Cúria Metropolitana do Rio de Janeiro).

Table 12 : Africans on death records from the parish of Candelária, Rio de Janeiro, 1793-1800
Source: Slave death records from the parish of Candelária, 1793-1800 (Arquivo da Cúria Metropolitana do Rio de Janeiro).

Table 15 : Nations of Africans on death records from urban and semi-urban parishes in Rio de Janeiro, eighteenth century
Source: Slave death records from the parishes of Candelária, Engenho Velho, and Jacarepaguá, eighteenth century (Arquivo da Cúria Metropolitana do Rio de Janeiro).

Table 17 : Nations of Africans and 'crioulos' (born to African mothers) on death records from the parish of Candelária, 1810-1830
Source: Slave death records from the parish of Candelária, 1810-1830 (Arquivo da Cúria Metropolitana do Rio de Janeiro).

Table 20 : Distribution of African mothers on baptismal records for 'crioulos' baptized in urban parishes, 1801-1830
Source: Slave baptismal records for the parishes of Candelária, Santana, Santíssimo Sacramento, São José, and Santa Rita, 1801-1830 (Arquivo da Cúria Metropolitana do Rio de Janeiro).theconnectionsbetweenclassificationsused in the slave trade and the trans-ethnic montages of initial experiences in the Diaspora, in the case of Rio de Janeiro.Angola, Benguela, Congo, Mina, Cabinda, and Moçambique could sometimes appear as Caçanjes, Rebolo, Monjolo, Ganguela, Songo, Calabar, and Quilimane, and later as Camundongo, Cabundá, Camundá, Muxicongo, Cabo Verde, São Tomé, and Inhambane or Luanda, Ambuíla, Massangano, Baca, Mofumbe, Nagô, Macua, Sena, etc.I have already mentioned the Rebolos, whose number varies over 500% in baptisms and inventories, but there are other indicative examples.The Luanda nation hardly appears at all in baptismal registers but in inventories represents 1.2% of Central Africans and 3.2% of Africans from northern Angola.Records of Rebolos, Monjolos, Cabundás, Camundongos, Camundas, Quissamas, and Caçanjes reach 1,369 in inventories, that is, 20.3% of Africans and 23.5% of Central Africans.However, in baptismal records, out of 1,127 records, they tally only 6.6% of all Africans and 12.6% of Central Africans, while Camundongos and Camundás do not appear in baptismal records at all.The most emblematic case is the Cabundás, who represent 2% of Central Africans in inventories but only 0.4% of baptisms.Libolos account for 0.4% of all Africans baptized while they do not even appear in inventories, unlike the case of Ganguelas, which tally 0.25% of those baptized but 3.9% of inventories (Table21).In short, there are some nations in the inventories that are almost never mentioned in the baptismal records of nearly 35,000 Africans in Rio's central parishes, such as Baca/Ambaca, who account for 1.4% of Africans from northern Angola in the inventories.We cannot forget the 'crioulos', and any sort of generalization could pose a trap.The baptismal records for 'crioulos' born to African mothers from 1801 through 1830 (Table20) show a balanced distribution between the percentages of African women who were baptized and those who appear in inventories.Combined, Angolas and Benguelas represent to

Table 21 : Comparative distribution of nations from baptisms of African women and 'crioulos' (men and women) born to African mothers in urban parishes in Rio de Janeiro, 1801-1830 Nation African women baptized African mothers of baptized 'crioulo' children
Source: Slave baptismal records from the parishes of Candelária, Santana, Santíssimo Sacramento, São José, and Santa Rita, 1801-1830 (Arquivo da Cúria Metropolitana do Rio de Janeiro).