Demographic change in a post-export boom society: the population of Minas Gerais, Brazil, 1776-1821.

Introduction

The social and economic history of Minas Gerais in the late 18th and early 19th centuries is extraordinary in comparative perspective. It is perhaps the only known example of a large-scale Latin American slave system which successfully made the transition from a slave-based export economy dependent on foreign markets, to a more diversified agricultural and cattle economy oriented almost exclusively to local or regional markets within Brazil. What makes the case of Minas Gerais so unique is that through this economic transition dependence upon slave labor remained central to Mineiro economy and society. Although there was a decline in the slave population in the immediate aftermath of the export economy's collapse in the mid 18th century, during the 19th century the slave population increased dynamically and Minas Gerais became the largest slave-holding province of the Brazilian empire. The documentary evidence suggests that this later renewed growth of slavery was based in part upon natural demographic increase rather than solely on imports from Africa or the inter-regional Brazilian slave trade, another phenomenon which may have been unique in the history of slavery in Latin America and the Caribbean. In nearly every other Latin American slave system the demise of export-based economic cycles heralded the long-term decline of slavery as well. This article will consider the process of demographic readjustment occurring in Mineiro society in the aftermath of the 18th-century gold mining boom. This was a critical period in the history of Minas Gerais and it provided the foundations upon which the province's economic and social order would be built during the 19th century.

Aspects of Mineiro demographic history in the 18th and 19th centuries have been considered in a number of studies over the past twenty years.(1) These were undertaken as part of a broader historiographical consideration of the region's social and economic history, and a great many have logically focused on transformations occurring within the slave labor force. The large scale importation of African slaves followed the discovery of alluvial gold deposits and diamonds in the 1690s and early 1700s and their age and sex characteristics shaped the region's demographic evolution in the first half of the 18th century.(2)

Social and economic structures during the gold cycle were defined by the export of primary products and the importation of labor and essential manufactured goods. Minas was a classical export economy which developed in a frontier region lacking a population base to sustain economic growth, and without an industrial or agricultural infrastructure to provide the fundamental products needed to support the mining complex, although these gradually emerged. By the 1740s mineral production began to contract and after 1750 the boom was over, although mining activities did not entirely disappear. During the second half of the 18th and through the early 19th centuries a more diversified agricultural and animal raising economy emerged, supplemented by the growth of small-scale manufacturing which had appeared during the gold boom. Mining remained, but at substantially lower production levels. This changing economy initially had a large subsistence sector but it was also geared toward provisioning the region's urban markets which remained as a legacy of mining and the bureaucratic structures developed by the Portuguese Crown to administer mineral production and levy taxes. Southern regions in Minas not only provisioned the capitania's urban areas, but also began to forge commercial linkages with the large Rio de Janeiro market for foodstuffs.(3)

After more than a half century of gradual transformation, these post-mining boom social and economic realignments were in the process of consolidation when European political events forced the Portuguese Crown to flee Lisbon. The arrival of the Corte in Rio de Janeiro in 1808 heralded all kinds of changes for colonial Brazil. The perspectives of now resident Royal authorities regarding Brazilian economic potential were shifted, and a series of reforms ensued which stimulated economic expansion.(4) The Brazilian agricultural export sector had begun to slowly grow again with the Pombaline administrative changes initiated in the 1750s, the rapid recovery of markets for sugar in the aftermath of the Haitian slave revolt of 1791, and the development of new crops, the most important of which were cotton and coffee.(5) After the Crown's arrival in 1808 and the opening of Brazilian ports to free trade, export oriented economic growth was accelerated and this resulted in large-scale slave imports to Brazil.(6)

The impact of the revived 19th-century Brazilian export economy and the renewed expansion of the slave trade to Brazil on Minas Gerais has been the subject of a lively scholarly debate.(7) Although significant research has been conducted on various aspects of the history of Minas Gerais during the Brazilian empire, the post-mining boom years of transition considered here have been less studied. The reaction of Mineiro economic and social structures to the end of the 18th-century export boom is central to an understanding of the provinces revitalization during the 19th century.

Sources

Four population censuses conducted in 1776, 1786, 1808, and 1821 were used for this article and these are described below. There was no consistency in the way data were collected or presented from one census to another, and this means that in some cases estimates of different population sectors or regions had to be made with inevitable margins of error.(8) Some of these census counts were incomplete, although their numbers have long been accepted by historians of Minas Gerais. This is the case with the census of 1776 which was published in the Revista do Arquivo Publico Mineiro (hereafter RAPM) in 1899.(9) This census lists the population of the capitania's four administrative districts or comarcas: Vila Rica, Rio das Mortes, Sabara, and Serro Frio. [ILLUSTRATION FOR MAP 1 OMITTED]. It divides the population into three racial categories - whites, pardos (mulattos), and blacks, and lists these categories by sex for each comarca. However, it does not indicate slave or free population data.

The 1786 population census was also published in the RAPM but the criteria for enumeration were different from those of 1776.(10) The population was divided into slave and free, further delineated by color using the same white, pardo, and black divisions, and then each of these categories was listed by sex. However, the statistics presented are aggregated for the capitania as a whole and there are no listings of population by comarca.

Summary data for the 1808 and 1821 censuses were published in the same issue of the RAPM in which the 1786 census appeared. However, more complete versions of these two censuses are available. The manuscript data for the 1808 census were located in the Arquivo Publico Mineiro (hereafter APM) in Belo Horizonte.(11) This is the most detailed extant population census for Minas Gerais in the early 19th century. The administrative organization of the capitania had been changed by further dividing the four comarcas into 14 municipal districts (termos) and data were collected by these local level administrative units. The summary manuscript data for 13 of the 14 termos were found.(12) The census divided the population of each district into slave and free and listed these categories by color and sex. Furthermore, detailed age profiles were provided for each population sector, and this permits the age structure of the different components of the Minas Gerais population to be reconstructed and analyzed for the first time.

The detailed 1821 data were published in the late 1830s by Raimundo Jose da Cunha Matos.(13) Although these data were nominally listed for 1821, in reality the census returns for each municipal district were for varied years from as early as 1812 through 1820.(14) The administrative subdivisions of the capitania had shifted somewhat. A new comarca, Paracatu, had been created in 1815 by dividing the old comarca of Sabara into two. Additionally, four new municipal districts had been created in areas where population was growing, bringing the total number of Minas Gerais termos to 18.(15)

In order to compare regional population changes between the different census years, the administrative divisions of the comarcas of 1776 (Vila Rica, Serro Frio, Sabara, and Rio das Mortes - [ILLUSTRATION FOR MAP 1 OMITTED]) have been reconstructed for some of the analyses presented here for 1808 and 1821. There are no local level data which would permit this for 1786.(16) These are important reconstructions since they permit the changes in population characteristics in each of the four principal regions of the capitania to be charted, and by inference some of the regional socioeconomic changes occurring between 1776 and 1821 may be deduced.

One important unknown factor, which hinders a complete understanding of demographic change in Minas Gerais in the period under consideration, is the absence of information on immigration, emigration, and birth or death rates. There are no data on the slave trade to the province which are critical for an understanding of changes in the structure of the slave population. Nor are there data on manumission rates which would be central for a more complete analysis of shifts among the slave and the free black and mulatto populations. However, through the use of other demographic measurements, such as changing sex ratios by population sector and child/woman ratios, important suggestive conclusions on population movements can be drawn.

General Population Characteristics

Population growth in Minas Gerais between 1751 and 1821 is indicated in Table 1.

These population statistics include a substantial revision for 1776. The published data in the RAPM indicated a total population of 319,769 and this figure has been ritually repeated by every historical consideration of Minas Gerais. There was, however, an evident undercount of the population in the comarca of Serro Frio which contained two municipal districts: Serro and Minas Novas. The population of Minas Novas was apparently omitted from these summary data because of a long-standing jurisdictional dispute with the capitania of Bahia which for some time claimed to control this mining district in the northeastern region of Minas Gerais.(17) These revised population estimates for 1776 are indicated in table 2.(18)

The annual rates of demographic growth reveal that as the mining boom waned after mid-century, the rate of increase among the overall population of Minas Gerais slowed from an average annual rate of 1.6% between 1751 and 1776 to 1.4% between 1776 and 1786.(19) However, it was after 1786 that the economic crisis effected by the collapse of mineral production in the late 18th century took its toll on the demography of the capitania, whose rate of population growth fell to .4% yearly between 1786 and 1808. It is probable that this very sharp decline in the rate of population expansion was the result of the temporary end of the slave trade to Minas and the end of free migration to the region owing to the contracting economy.

The recovery of the annual population growth rate to 1.3% between 1808 and 1821 is suggestive. Changing sex ratios in favor of more balanced male to female ratios within all sectors of Mineiro society indicate increased reproductive potential by the late 18th and early 19th centuries. The demographic data on the slave population provide very little evidence of any large scale revival of the slave trade to Minas between 1808 and 1821 which could have significantly contributed to increased population growth rates, although small numbers of slaves continued arriving in particular regions within Minas Gerais. Nor is there any evidence of renewed immigration of free people on any significant scale.(20) In all likelihood, population expansion after 1808 resulted from the end of out migration and the generalized increase in natural reproduction among slave and free rather than migration to the capitania. Whether there was a general decline in death rates is unknown.

Table 1

Population of Minas Gerais, 1751-1821

                                Annual Rate of
                               Population Growth
Year        Population        from Preceding Year

1751          226,666
1776          341,869                 1.6%
1786          393,698                 1.4%
1808          433,049                 0.4%
1821          514,797                 1.3%

Sources: For 1751 see Annuario Estatistico Anno II (1922-1925)
(Belo Horizonte, 1929), p. 9. The 1776 population totals which
appeared in RAPM, Ano IV, No. 3, (1899), p. 511 have been revised
(see discussion in footnotes below). For 1786, 1808, and 1821
summary data (no age information) see RAPM, Ano IV, No. 2, (1899),
pp. 294-295; for 1808 detailed data including ages, APM, SG, Caixas
Avulsas, Caixa 77, Docs. 61, 63, 65, 69, 72, 74, 75, 77-80, 82,
84. For 1821 detailed data, Raimundo Jose da Cunha Matos,
Corografia Historica da provincia de Minas Gerais: 1837, 2 vols.
(Belo Horizonte, 1981), Vol. 2, pp. 45-49. All subsequent tables
and figures are derived from these sources unless otherwise noted.

Note: Annual population growth rates have been calculated
throughout this article by using the formula R = ([LN x
(Pe/Po)]/T) x 100 where R is the annual rate of population growth,
LN is the natural log, Pe is the population in the end year, Po is
the population in the beginning year, and T is the number of years
between census dates.

[TABULAR DATA FOR TABLE 2 OMITTED]

Regional Population Change

When we disaggregate the population of Minas into its four regions as defined by the comarcas of 1776 indicated in Map 1, contrasting demographic patterns may be observed, and these reflected the social and economic transformations in each area. The most striking was the growth of population in the southwestern comarca of Rio das Mortes between 1776 and 1821. The population of this old mining zone, which was transformed into a predominantly agricultural and animal raising region during the second half of the 18th century, increased at an annual rate of 2.0% between 1776 and 1808 and 2.3% between 1808 and 1821. This was significantly greater than in any other area in the captaincy. (See table 3 for these and following data summaries) Rio das Mortes also increased its share of the total Minas Gerais population from 24% (1776) to nearly 41% (1821) of all inhabitants. Although the areas around Sao Joao del Rei and Sao Jose del Rei (Tiradentes) had been centers of population dating from the first half of the 18th century when gold deposits were first discovered, the southwestern and far western regions of Rio das Mortes were virtual frontier areas at the beginning of the 19th century.(21) Land availability, soil fertility, and the potential of developing subsistence and commercialized agricultural activities and animal raising attracted migrants to this area of the capitania which was linked by old caminhos to the markets of Sao Paulo and more importantly to Rio de Janeiro.(22) These areas became the principal centers of grain production, sugar cane cultivation and aguardente manufacture, cattle, mule, and pig ranching, cheese manufacture, and tobacco production in Minas Gerais.(23)

While these regional data indicate internal migration to frontier zones in Rio das Mortes, demographic stagnation or decline typified the traditional mining districts. In the comarca of Vila Rica population contracted by 0.3% yearly between 1776 and 1808 and continued to shrink at the same annual rate between 1808 and 1821. In 1776 23% of all Mineiros lived in Vila Rica, but this declined to 13.6% in 1821. Serro Frio's population also contracted between 1776 and 1808 at an annual rate of 0.5% although between 1808 and 1821 there was renewed growth of 1.3% yearly. In the case of Vila Rica population decline reflected the contraction of the mining sector and less opportunity and potential for agricultural development. Vila Rica (which included the termos of Ouro Preto [TABULAR DATA FOR TABLE 3 OMITTED] and Mariana) had been heavily populated from the 18th century, and land availability was minimal compared with western frontier regions. The economic evolution of Serro Frio was very different, and this is why demographic patterns diverged after 1808. Large unsettled regions contiguous to Bahia and Espirito Santo attracted population because of land availability. The development of cotton cultivation on a significant scale within the northeastern district of Minas Novas in the late 18th century also drew migrants seeking economic opportunities.(24) Additionally, diamond production continued in the late 18th century and grew significantly in the early 19th century in Diamantina, while placer gold mining was maintained along the Arassuai river.(25) Thus, Serro Frio's economy was revived to some extent after 1808 although its relative population declined from nearly 24% of the Minas Gerais total in 1776 to 16% in 1821.

The comarca of Sabara (1776 definition) encompassed a vast area in the northwestern region of the captaincy. While the city of the Vila Real (Sabara) was one of the great centers of Minas gold production and population concentration during the early 18th century, the area far to the northwest beyond the headwaters of the Rio Sao Francisco was largely unpopulated. Land availability attracted migrants seeking subsistence alternatives with the decline of mining, and the most obvious indication of this was the colonial government's decision to create the new comarca of Paracatu in 1815. Gold mining continued on a small scale around Paracatu itself; and closer to the Sao Francisco River the production and trade in salt formed an important activity. The Sao Romao area became a large scale cattle raising zone from which hides and live cattle were exported. Unlike Vila Rica and Serro Frio, the comarca of Sabara's population increased steadily at 1% annually between 1776 and 1808 and .8% from 1808 to 1821. It is also conspicuous that Sabara was the only Minas Gerais comarca which maintained its relative population within the captaincy with roughly 29% of the total in both 1776 and 1821. These regional population data are summarized in table 3.

Demographic Change by Racial and Legal Categories

The well-known decline in mining activity, and the resulting demise of the Minas slave trade, resulted in a sharp decrease in the capitania's slave population from 1786 (188,698) to 1808 (148,772) - a decline of over 20% in total at an annual rate of population loss equaling 1.1%.(26) Not only did the Mineiro slave population fall significantly, but Minas Gerais effectively ceased being a society defined so emphatically by slavery, although its legacy would remain central to the social and racial characteristics of the region to the present. Slaves declined from 48% to 34% of the population between 1786 and 1808, and they were replaced by free blacks and mulattos as the dominant racial and legal sector. Free peoples of color increased by 33% between 1786 (133,506) and 1808 (177,593), at an annual rate of 1.3%, and grew from 34% of the total populace in 1786 to 41% in 1808. Thus, between 1786 and 1808, while Minas Gerais continued to be defined by its population of African origin, there were more free people of color than those who were enslaved by 1808, and this pattern would remain central to the demographic history of the future province throughout the 19th century. The extent to which this population growth was the result [TABULAR DATA FOR TABLE 4 OMITTED] [TABULAR DATA FOR TABLE 5 OMITTED] of positive reproductive rates among free peoples of color, or an increase in manumissions during the period of economic contraction, is unknown. It is probable that reproduction was the most important contributing factor. Those classified as white increased from 18% to 25% of the population in the same period at a rate of 1.8% annually. (See table 4 for these data)

Table 6

Sex Ratios of Minas Gerais Population by Racial Category and Legal
Status, 1786-1821

                       1786     1808     1821

Whites                  121      103      105
Free Blacks              83       94       95
Free Mulattos            94       99       93
Total Free Colored       90       98       94
Black Slaves            225      188      156
Mulatto Slaves           94      100      109
Total Slaves            201      175      148

Total Population        138      120      112

Source: RAPM, Ano IV, No. 2 (1899), pp. 294-295.

These racial and legal categories remained relatively stable between 1808 and 1821, although it is evident that the capitania had now entered a new period in its demographic history. The post-mining boom population readjustments had been consolidated in an epoch of economic reorganization characterized by a major geographical shift away from the old mining centers to new southwestern regions of agriculture and cattle raising. After 1808, all sectors of the population grew steadily, and available data indicate that these increases were related to natural demographic growth rather than migration, forced or voluntary. It should be emphasized once again that information on birth or death rates is non-existent for Minas Gerais, and there are no reliable data on the slave trade to Minas in the late 18th and early 19th centuries. However, changing sex ratios (number of males per females) among the various population groups, indicate that the potential for natural population reproduction evolved significantly after 1776. If we examine sex ratios by race between 1776 and 1821, the white and black populations were transformed from racial groups formed primarily by immigration, to sectors where natural reproduction was increasingly responsible for sex distribution patterns.(27) Among whites there were 144 males for every 100 females in 1776, and this continually declined to 1808 when the sex ratio was 103. Between 1808 and 1821 sex ratios among Mineiro whites stabilized (105 in 1821), indicating the virtual absence of immigration and a population which was growing naturally.

Among the black population, which was predominantly enslaved, the decline in the sex ratio was even more dramatic between 1776 and 1821.(28) In 1776 there were 235 black males for every 100 females in Minas Gerais but this ratio fell continuously to 136 by 1821. (See table 5) This is very clear evidence that the slave trade, always predominantly male, diminished significantly after 1776. Even during the period of revived economic growth in the late 18th and early [TABULAR DATA FOR TABLE 7 OMITTED] 19th centuries the slave trade to Minas Gerais was in all likelihood marginal to the demographic history of the capitania as a whole, although there were important regional variations.(29)

The demographic image of the Minas Gerais slave population sharpens after 1786 when detailed data on race and legal status become available. These reveal with more clarity the decline of the slave trade as a factor influencing the demographic development of the capitania between 1786 and 1821, and that the growth of the Minas slave population indicated from 1808 to 1821 (at a rate of 1.1% annually) was due principally to reproduction, although slaves continued to arrive in small numbers. Among black slaves the sex ratio dropped from 225 in 1786 to 156 in 1821, although among mulatto slaves it increased from 94 to 109 in the same period. This may be indicative of increased manumissions among mulatto male slaves during the period of economic reorganization, or the purchase of mulatto slaves from other Brazilian regions. The total slave population experienced a sex ratio decline by over 25% between 1786 and 1821 from 201 to 148. (See table 6).

These shifts in sex ratios are convincing indicators of the demographic transformations occurring throughout the capitania as a whole, but they must be examined from regional perspectives for a more complete image of the changes which were taking place in Minas economy and society. Although the boom period in mining was long over by the early 19th century this does not mean that mining activities disappeared, nor that the association of limited slave imports with mining waned as significantly as in areas where mining was less important. Changing regional sex ratio profiles between 1808 and 1821 indicate how differences in economic structures shaped regional patterns of demographic change among the slave population.

Table 8

Percentage of the Population by Race and Legal Status in Each
Reconstructed Comarca, 1808-1821

                     Whites       Free Colored       Slaves
                  1808    1821    1808    1821    1808    1821

Vila Rica         17.4    14.8    46.3    47.0    36.3    38.2
Rio das Mortes    32.9    39.4    28.3    26.6    38.8    33.9
Sabara            16.8    18.7    48.5    49.3    34.7    31.9
Serro Frio        15.8    14.5    49.0    55.7    35.7    29.7

[TABULAR DATA FOR TABLE 9 OMITTED]

It is evident that in the old comarca of Serro Frio there were still slave imports when these had virtually ceased elsewhere. This was because of labor demands related to the persistence of diamond production in Tejuco (Diamantina), placer gold mining activities around Serro itself, and the expansion of cotton cultivation in the late 18th century in the Jequitinonha and Arasuai river valleys. Although there are no statistical data, it is certain that these imports were fractional compared with the first half of the 18th century, and they were overshadowed by the reproductive process among the slave population occurring elsewhere. Nevertheless, it is important to note that the sex ratio among slaves in Serro Frio in 1808 (209) remained comparatively much higher than in the other regions of the captaincy. Vila Rica, Rio das Mortes, and Sabara (1776 comarca definitions) experienced constantly declining slave sex ratios after 1786, but in 1808 the Serro slave sex ratio still remained higher than the captaincy-wide ratio of 201 which had prevailed in 1786. This was most extreme among Serro Frio's black slaves whose sex ratio was 244 in 1808 compared with 185 in Vila Rica, 188 in Rio das Mortes, and 198 in Sabara. These data indicate that slave imports continued into Serro Frio, although there is no information on whether slaves were transferred from other regions within Brazil or Minas Gerais, or if they were recent African imports.

However, after 1808 the economic activities which sustained continuing slave imports into Serro Frio evidently declined. The sex ratio among the overall Serro Frio slave population fell significantly from 209 in 1808 to 143 in 1821, although among the small mulatto slave population in 1821 (5,501) it was 176, indicating the low-level importation of mulatto male slaves into the region after 1808.(30) Yet, the overall decrease in the sex ratio among Serro Frio's slaves paralleled trends among slaves in Rio das Mortes where the sex ratio was 175 in 1808 and 147 in 1821; and in Sabara where it was 188 in 1808 and 143 in 1821. Only Vila Rica maintained its slave sex ratio of 1808 (170) in 1821 (171), indicating the possibility that some slave imports may have continued to this area of the capitania. These data on the regional sex ratios of the Minas Gerais population in 1808 and 1821 are summarized in table 7.

Regional variations in the Minas Gerais slave labor force must be seen in perspective. Slaving may have continued at minimal levels to Serro Frio and on an even smaller scale to Vila Rica in the early 19th century, but we must keep in mind that the most rapidly growing region in the capitania, Rio das Mortes, had a slave population in 1821 (71,147) which was larger than that of Vila Rica and Serro combined (51,590). Here the sex ratio had steadily declined to 147 by 1821.

Rio das Mortes was not only a region where the slave population had greater reproductive capabilities because of declining sex ratios, but its racial configuration was very different than that of the other three comarcas. Nearly 62% of the Minas Gerais white population resided in Rio das Mortes in 1821 and what is even more striking is the fact that in the same year 39% of the population living there was classified as white. This compares with only 15% in Vila Rica and Serro Frio, and 19% in Sabara. This proportionally greater white population of Rio das Mortes does not mean that slavery was any less important to the regional economy and society in the early 19th century. Nearly 34% of the population was enslaved in 1821.

The greatest difference was the comparatively smaller population of free blacks and mulattos found in this southwestern area of Minas Gerais. It is striking that 27% of the Rio das Mortes population was comprised of free persons of color compared with 47% and 49% in Vila Rica and Sabara respectively, and 56% in Serro Frio. This does not mean that rates of emancipation in the old Rio das Mortes mining centers of Sao Joao del Rei and Sao Jose del Rei were lower than in the other comarcas. Nor does it imply that reproductive rates among free peoples of color in Rio das Mortes were lower than elsewhere in Minas Gerais. What it does suggest however, is a much greater level of spatial mobility among whites in the capitania than among free blacks and mulattos. The yearly population increments found in Rio das Mortes among whites between 1808 and 1821 (3.7% annually) indicate substantial migration to the region, although it is uncertain whether white migrants originated in other areas of Minas Gerais, or came from contiguous capitanias.(31) Since the white population grew so much faster because of migration, it is logical that whites would outnumber other population sectors which did not experience similar levels of immigration. The free black and mulatto population increased annually at 1.9% between 1808 and 1821, indicating comparatively lower rates of migration to Rio das Mortes. (Table 8 shows population percentages by race and legal status between 1808 and 1821 and table 9 changing sex ratios among different population sectors by municipal districts.)

The process of white migration to Rio das Mortes was most extreme in southwestern agricultural regions which were distant from the traditional mining and population centers in the eastern zone of the comarca. In the older mining regions of Sao Joao del Rei, Sao Jose del Rei, Tamandua, Queluz, and Barbacena, 33% of the total population was white in 1821, 38% was enslaved, and 29% were free blacks and mulattos. In the frontier regions comprising Campanha, Jacui, and Baependi (the Campanha boundaries of 1808) the population was 49% white, 24% free colored, and 27% enslaved in 1821. These municipal districts, agricultural and animal raising areas where land was available in the early 19th century, had the highest levels of white population concentration in the capitania, and clearly were major destinations of white migrants who probably originated in the traditional mining areas where land was unavailable.

Demographic Profiles of the Minas Gerais Population in 1808

The 1808 manuscript census found in the APM provides the most complete information on the population of the capitania in the early 19th century. As has been indicated, detailed data on 13 of the 14 termos are available for sex, race, and legal category, and these are the only known documents which provide age data for all sectors of the population in the period under consideration. Until now I have arrived at conclusions on migration to or within Minas Gerais, and [TABULAR DATA FOR TABLE 10 OMITTED] [TABULAR DATA FOR TABLE 11 OMITTED] reproductive possibilities among the different sectors of the population, based on general demographic information and changing sex ratio patterns. The 1808 data permit a more thorough consideration of these indicators by examining child/woman ratios and age pyramids of the slave and free populations.(32)

Among Minas Gerais slaves in 1808 the child/woman ratio (expressed as children per thousand adult women) suggests high fertility rates. These complement the data on declining sex ratios among Minas slaves and support the conclusion that slave reproduction was having a significant impact on the demography of the capitania's slave population by the early 19th century. Among all slaves the ratio for children ages 1-10/women ages 20-50 was 1,233 and 980 for children ages 1-10/women ages 15-50. By way of comparison Francisco Luna found a lower child/woman ratio of 792 (children ages 0-9/woman ages 15-44) in Sao Paulo in 1804 when the sex ratio there was 121.(33) This was before the development of coffee culture on a large scale and prior to significant slave imports. By 1829, after the development of coffee cultivation stimulated the slave trade to Sao Paulo, Luna and Herbert Klein found a ratio of 560 for children ages 09/women ages 15-49 in three Sao Paulo slave-holding regions. Stuart Schwartz noted an average ratio of 597 (children ages 0-9/women ages 15-45) in three Bahian sugar producing parishes in 1788.(34) By way of comparison in the U.S. South, where slave reproduction was solely responsible for slave population increases after the end of the slave trade in 1808, in 1860 the child woman ratio was 1,056 for children ages 1-9/women ages 15-45.(35)

Among whites, the child woman ratio was 2,252 for children ages 1-10/women ages 20-50 and 1,646 for children ages 1-10/women ages 15-50, indicating greater birth rates and impressive rates of population expansion. And among free blacks and mulattos, these ratios were 1,578 (children ages 1-10/women ages 15-50) and 1,238 (children ages 1-10/women ages 15-50).(36)

Figure 1 presents images indicating the age structures of the different sectors of the Minas Gerais population in 1808. These reveal with graphic clarity the disequilibrium in the sex ratio among black slaves who were 89% of the capitania's total slaves. But even in this most unbalanced population sector the age pyramid image suggests that the process of reproduction had been well underway. Over 35% of all black slaves in Minas Gerais were 20 years of age or younger. They had been born after 1788 during a period when slave imports to Minas had been minimal, and this means that almost all were Brazilian, rather than Africans.(37) It is unfortunate that the 1821 census data have no age information. However, it is probable that the age structure of the slave population was moving toward a pyramid with a wider base of younger slaves, and one that was more similar to the white and free colored populations. The age pyramids of the free population of Minas Gerais indicate populations formed almost entirely by reproduction in 1808. (Tables 10 and 11 include complete population summaries by comarca and termo for 1808 and 1821.)

Conclusions

The end of the mining cycle affected the Minas Gerais population in several ways. The immigration of free people seeking economic opportunity virtually ceased between 1786 and 1808 when the annual rate of population growth in the capitania declined precipitously. This was also paralleled by the end of significant slave imports, at least until the 1820s and possibly thereafter.

The slave population declined by over 20% between 1786 and 1808. The degree to which this resulted from greater deaths than births, the sale of slaves to other capitanias, or increased manumission rates is unknown. However, while Minas slaves were contracting both in absolute terms and as a percentage of the total population between 1786 and 1808, important changes were occurring in the sex ratio and in the age structure of the slave population. The number of males per females continually fell, resulting in a demographic situation which was favorable for slave reproduction. The age pyramid of 1808 indicates that this had begun, and it may very well be that the renewed growth of the Minas slave population occurring between 1808 and 1821 (+ 15%) was the result of natural population increase, although the small-scale importation of slaves should not be discounted.

The conclusion that the slave population was reproducing is sustained by a number of other studies on Minas slavery and for other regions of Brazil as well.(38) In Parana between 1800 and 1830 even lower sex ratios were found than those prevailing in Minas Gerais, and these are also indicators suggesting slave reproduction.(39) Downward trends in slave sex ratios in regions which experienced few or no African imports such as Minas Gerais, can be contrasted with Brazilian regions where the slave trade was active. In Sao Paulo sex ratios increased markedly in the early 19th century as sugar and coffee production developed. In 1777 the Sao Paulo slave sex ratio was 116, and this increased slightly to 121 by 1804. But by 1829 the sex ratio stood at 159, a clear indicator of slave imports between 1804 and 1829.(40) In Bananal, located in the eastern Sao Paulo Paraiba valley coffee region contiguous to Rio de Janeiro, the change in the slave sex ratio accompanying African imports was even more dramatic, rising from 139 in 1801 to 218 in 1829.(41) If slave imports had been renewed to Minas Gerais during the early 19th century, there would have been a corresponding rise in the sex ratio as was the case in Sao Paulo. That the sex ratio continually fell is convincing evidence that the increase in the number of Minas slaves after 1808 was mainly due to the reproductive patterns within the extant slave population of the capitania.

The contraction of the slave population between 1786 and 1808 was accompanied by the increase of whites and free blacks and mulattos. By the early 19th century free peoples of color had replaced slaves as the most numerous population sector, and their annual rates of demographic growth would assure population predominance in Minas Gerais from then on. Until future research uncovers data on births, deaths, manumissions, and migration it will be nearly impossible to compare with precision demographic trends among free peoples of color and slaves. However, it is conspicuous that growth rates among slaves (1.1% annually) and free peoples of color (1.2% annually) were so similar between 1808 and 1821. If it is assumed that declining slave sex ratios, increasing child/woman ratios, and a more balanced age pyramid in 1808 indicate the demise of slave imports, then it is clear that when Brazilian slaves were not demographically disturbed by large scale African arrivals, they reproduced at rates which were similar to those of the free colored population.

These population changes must be interpreted within the context of the economic reorganization which was taking place in Minas Gerais with the end of mineral production as the dominant activity. The economic history of the capitania in this period of transition has yet to be written. However, it is clear that a more diversified agricultural and animal raising economy was emerging, and that these sectors were supplemented by continued small-scale gold and diamond production, as well as cottage industries which were producing manufactured goods such as cotton textiles and aguardente.(42) These may have been largely subsistence activities in the initial aftermath of the mining collapse occurring in the 1740s and 1750s. However, there were remaining urban markets throughout Minas Gerais which had to be provisioned, while commercial linkages with the growing markets of Sao Paulo and Rio de Janeiro were progressively forged in the second half of the 18th century that developed more extensively after 1800. The growth of export-oriented sugar and coffee production in Sao Paulo and the beginning of the coffee cycle in Rio de Janeiro and southeastern Minas provided significant commercial opportunities for Mineiro farmers and ranchers by the beginning of the 19th century.(43) The migration of the Minas population to the southern and western regions of the comarca of Rio das Mortes, whose population increased two and a half times between 1776 and 1821, is indicative of both land availability and commercial opportunities provided by proximity to expanding markets. The relative population decline of the old mining centers and their shrinking importance in the regional economy paralleled these changes.

It is unfortunate that none of the four census reports indicates the birthplace of the Minas Gerais population. Immigrants coming to Minas during the gold cycle, whether free and voluntary or forced and enslaved, were heavily male and predominantly adults. The declining sex ratios and relatively high child/woman ratios among all racial and legal sectors of the population, strongly suggest that the vast majority of Mineiros residing in the capitania in the early 19th century were Brazilian born. In a recent study of the slave population of the mining municipality of Mariana, I found that between 1795 and 1808 nearly 60% of all slaves appearing in the inventories of deceased property owners were born in Brazil. This represented a significant shift from the 1750 to 1794 period when African slaves were the great majority of all Mariana slaves.(44)

With the consolidation of its post-mining boom economic and population readjustments, Minas Gerais had already entered a new phase in its demographic history by the first decades of the 19th century. The capitania's regional population distribution patterns had been redefined, its configuration was reorganized with the growth of the free black and mulatto population, and its slave labor force was stable and apparently reproducing. As the age of coffee emerged, Minas was poised to play an important role as a producer of agrcicultural and pastoral products in the new economic order which would develop during the Brazilian empire.

Department of Latin American and Puerto Rican Studies Bronx, NY 10468

ENDNOTES

1. For a summary of demographic research on Minas Gerais to the early 1980s see Herbert S. Klein, "The Population of Minas Gerais: New Research on Colonial Brazil," Latin American Population History Newsletter 4:1-2 (1984): 3-10. Among the most important works on the demographic history of Minas in the 18th and 19th centuries are: Francisco Vidal Luna and Iraci del Nero da Costa, Minas Colonial: Economia e Sociedade (Sao Paulo, 1973); Francisco Vidal Luna and Iraci del Nero da Costa, "Demografia Historica de Minas Gerais no Periodo Colonial," Revista Brasileira de Estudos Politicos No. 58 (1985): 15-61 and "Algumas Caracteristicas do Contingente de Cativos em Minas Gerais," Anais do Museu Paulista Vol. 29 (1979): 79-97; Wilson Cano and Francisco Vidal Luna, "La reproduccion natural de los esclavos en Minas Gerais: una hipotesis," Revista Latinoamericana de historia economica y social (Lima) 4:2 (1984): 129-135; Iraci del Nero da Costa, Vila Rica: Populaao (1719-1826) (Sao Paulo, 1981); Herculano Gomes Mathias, Um Recenseamento na Capitania de Minas Gerais (Vila Rica 1804) (Rio de Janeiro, 1969); and Clotilde Andrade Paiva and Douglas Cole Libby "The Middle Path: Alternative Patterns of Slave Demographics in Nineteenth-Century Minas Gerais," El poblamiento de las Americas, 3 vols. (Vera Cruz, 1992) vol. 1, pp. 185-232.

2. See the classic study of C. R. Boxer, The Golden Age of Brazil, 1695-1750 (Berkeley, 1962), A. J. R. Russell-Wood, The Black Man in Slavery and Freedom in Colonial Brazil (New York, 1982), and Russell-Wood, "The Gold Cycle, c. 1690-1750" pp. 190-243, in Leslie Bethell, ed. Colonial Brazil (New York, 1987).

3. See Kenneth R. Maxwell, Conflicts and Conspiracies: Brazil and Portugal 1750-1808 (New York, 1973), pp. 87-91 for a brief description of economic changes.

4. For example, in 1785 a Portuguese royal edict banned manufacturing of a variety of products including cotton textiles, a mainstay of certain Mineiro regions. This ban was repealed in 1808 as well as restrictions on trade. For a discussion of this see Fernando A. Novais, Portugal e Brasil na Crise do Antigo Sistema Colonial (1777-1808) (Sao Paulo, 1981), Segunda edicao, pp. 268-285. Also see "Documentos officiaes ineditos relativos ao alvara de 5 de Janeiro de 1785, que extin Brasil todas as fabricas e manufacturas de ouro, prata, sedas, algodao, linho, la, etc.," Revista Trimensal de Historia e Geographia (Rio de Janeiro, 1870): 213-225. For a brief review of the economic impact of the arrival of the Corte see Leslie Bethell, "The Independence of Brazil" pp. 3-44, in Leslie Bethell, ed. Brazil: Empire and Republic, 1822-1930 (New York, 1989).

5. See the discussion of the "Agricultural Renaissance" in Dauril Alden, "Late Colonial Brazil, 1750-1808" pp. 284-343, in Bethell, ed. Colonial Brazil, and the general consideration of the Pombaline era in James Lockhart and Stuart B. Schwartz, Early Latin America: A History of Colonial Spanish America and Brazil (New York, 1983), pp. 383-404.

6. The latest data generated by David Eltis, Economic Growth and the Ending of the Transatlantic Slave Trade (New York, 1987), p. 243 indicate that average yearly slave imports soared from about 20,000 per year from 1781-1808 to 37,000 yearly from 18091830.

7. See Amilcar Martins Filho and Roberto B. Martins, "Slavery in a Non-Export Economy: Nineteenth-Century Minas Gerais Revisited," Hispanic American Historical Review 63:3 (1983): 537-568, and the critical comments on this article by Robert Slenes, Warren Dean, Stanley Engerman, and Eugene D. Genovese, pp. 569-590 of the same volume. The Martins replied to these critiques in "Slavery in a Nonexport Economy: A Reply," Hispanic American Historical Review 64:1 (1984): 135-146. Also see Roberto B. Martins, "Growing in Silence: The Slave Economy of Nineteenth-Century Minas Gerais, Brazil" (Ph.D. dissertation, Vanderbilt University, 1980), and a more recent updated argument found in Roberto Borges Martins, "Minas e o Trafico de Escravos no Seculo XIX, Outra Vez," Belo Horizonte, UFMG/CEDEPLAR, 1994 (Texto para discussao No. 70) and Robert W. Slenes, "Os Multiplos de Porcos e Diamantes: A Economia Escravista de Minas Gerais no Seculo XIX," Cadernos I FCH/UNICAMP, no. 17 (June 1985). The Martins have argued that the slave trade to Minas was quite substantial during the 19th century while challengers have pointed to natural demographic growth as a significant factor in provincial slave population expansion.

8. The methods employed in making different population estimates are described in detail when estimates are presented.

9. See RAPM, Ano IV, No. 3 (1899), p. 51. The original manuscript upon which this publication was made has not been located. The total population of 319,769 listed in this census is an undercount because of the omission of the district of Minas Novas in the comarca of Serro Frio. This will be discussed below.

10. RAPM, Ano IV, No. 2 (1899), pp. 294-295. The original manuscript of the 1786 data has also not been located.

11. These data are contained in section SG, Caixas Avulsas, Caixa 77, docs. 61, 63, 65, 69, 72, 74, 75, 77, 78, 79, 80, 82, and 84.

12. These include the following: In the comarca of Vila Rica were the termos of Mariana and Vila Rica; in the comarca of Rio das Mortes were the termos of Barbacena, Queluz, Sao Joao del Rei, Sao Jose del Rei, and Tamandua; in the comarca of Sabara were the termos of Sabara, Paracatu, Pitangui, and Nova Rainha; and in the comarca of Serro Frio were the termos of Serro and Minas Novas. The missing district was Campanha in the comarca of Rio das Mortes, but the population of this district in 1808 has been ascertained by examining a number of other documents. This will be discussed below.

13. Raimundo Jose da Cunha Matos, Corografia historica da provincia de Minas Gerais: 1837, 2 vols. (Belo Horizonte, 1981), Vol. II, pp. 45-49. Again, the manuscript census materials upon which these data were based have not been located.

14. The Cunha Matos data have been cited as if they were actually for 1821 even though this is not accurate. See, for example, the summary statistics of the 1821 population census listed in RAPM, Ano IV, No. 2 (1899), p. 295 which notes a population total of 514,797 for Minas Gerais compared with the Cunha Matos total of 514,780 (a difference of 17) for the various census years between 1812 and 1820. (See table 11 for exact census years.)

15. In the comarca of Rio das Mortes two new termos were created by subdividing Campanha into Jacui, Baependi (which were both created in 1814), and Campanha. In the comarca of Paracatu two new administrative districts were created: Sao Romao and Desemboque. These were in addition to the termos existing in 1808. For the history of administrative subdivisions in Minas Gerais see Joaquim Ribeiro Costa, Toponimia de Minas Gerais (Belo Horizonte, 1970).

16. Each of the 14 termos which had been created by the 1808 census fits into one of the four still extant comarcas, and by aggregating their data the population of the four 1776 comarcas could easily be ascertained. For 1821, the same method was used for the 18 termos and since the fifth comarca Paracatu was a subdivision of Sabara (1776, 1786, and 1808 definition) it was not difficult to reconstitute the old 1776 divisions.

17. The omission of Minas Novas in the 1776 population census was detected in an anonymously authored document titled "Memorias Historicas da Provincia de Minas Geraes," RAPM, Anno XIII (1908): 523-639. The editors of the journal extracted this from a manuscript source located in the APM (whereabouts unknown) and estimated that it was written between 1849 and 1851. On p. 639, while discussing the 1776 census the author observed: "A populacao comprendida nos limites da diocese, e capitania, exceptuado o territorio das Minas Novas do Arassuahy, abrangia o total de 319:769 pessoas em todas as classes do brancos, pardos e pretos, no anno de 1776." (Translation: "The population found within the limits of the diocese and the capitania, with the exception of the territory of Minas Novas do Arassuahy, includes a total of 319,769 people of all classes of whites, pardos, and blacks in the year 1776.") The original document for the 1776 census has never been located.

In all probability this was because Minas Novas was part of Minas Gerais from an administrative perspective, but ecclesiastically it was part of Bahia during the 18th century. Cunha Matos, Corografia Historico, Vol. 1, pp. 156 and 165; and in Vol II, p. 161 describes this situation. On p. 41 Vol. II he also notes that in a 1768 "Memoria sobre a criacao dos Bispados do Brasil" written by Veloso de Oliveira (exact citation not listed), a population of 400,000 was estimated for Minas Gerais.

18. The population of Minas Novas was estimated to have been approximately 22,100 in 1776, and this figure was added to the published total to arrive at the 1776 population indicated in Table 1. This is an important correction because of its implications for the analysis of overall and regional population changes after 1776.

The Minas Novas population estimate of 22,100 for 1776 was derived in the following way: The population of the municipal district of Serro is known for 1776 (58,794) and 1808 (50,838). The geographical limitations did not change between these years, and the termo of Serro included the important population center of the Arrial do Tejuco (Diamantina). Between 1776 and 1808 the population of Serro declined at an average yearly rate of -0.45%. We also know that in 1808 the population of Minas Novas was 19,136 according to the manuscript census. If we make the assumption that the annual rate of population decline in Minas Novas was the same as for Serro (-0.45%) then projecting backwards the population would have been 22,100 in 1776. There is a small unknown margin of error here. All subsequent tables which refer to the 1776 census include this estimate.

The number of blacks, whites, and pardos by sex were also calculated for Minas Novas in 1776. These were included within the comarca of Serro Frio, and in the general population figures for the capitania presented in subsequent tables. These were calculated in the following way: The percentage of each population category (white males, white females, black males etc.) was calculated for the termo of Serro in 1776. It was assumed that roughly the same percentage of the population would have fallen into these sex and race categories in Minas Novas. These percentages for Serro were then multiplied by 22,100 (total estimated population of Minas Novas) to determine race and sex profiles of Minas Novas. The totals were added to those indicated in the 1776 census for Serro to determine the total population of the comarca of Serro Frio and the capitania of Minas Gerais. The Minas Novas calculations yielded 3,347 white males; 1,789 white females; 3,077 pardo males; 2,670 pardo females; 8,384 black males; and 2,833 black females.

19. These rates of annual population increase also lend credence to the fact that the 1776 census was incomplete. If we calculated the annual rate of population growth between 1776 and 1786 using the published 1776 population of 319,769, a yearly rate of 2.1% would result. This would make no intuitive sense given the 0.4% annual rate between 1786 and 1808. The Minas Gerais economy contracted steadily after 1750 and there are no reasons why population growth would be over 2% yearly between 1776 and 1786, when rates were so different shortly thereafter. The only explanation is that the 1776 population was undercounted.

20. Demographic data on sex ratios and rates of population growth among slave and free will be indicated below.

21. For a detailed description of this area in the early 19th century see RAPM, Ano I (1896) "Memorias Municipaes" Campanha, pp. 457-647 and Almanack do Municipio da Campanha (Campanha, 1900).

22. See the descriptions of Rio das Mortes in the well-known chronicle of Auguste De Saint-Hilaire, Viagem Pelo Districto Dos Diamantes E Litoral Do Brasil (Belo Horizonte, 1974 edition) (originally published in 1833), pp. 106-107; 111-112. Saint Hilaire traveled through Brazil between 1816 and 1822.

23. Rio das Mortes was described in the late 18th century by Jose Joaquim Rocha as a region which was "a mais abundante de toda a Capitania, por quanto della se sustentao a major parte dos habitantes das mais comarcas, principalmente de toucinhos, gados, queijos, milho, feijao, e arroz...." (Translation: "the most abundant of the entire Captaincy, from which the greatest share of the population of the other districts are supplied with bacon, cattle, cheese, corn, beans, and rice ... ") See RAPM, Ano II (1897): 478. The area around Baependi (part of Campanha until created in 1814) was a major tobacco producing zone. See "Memorias Municipaes" in RAPM, Ano II (1897): 247.

24. See the descriptions of the Minas Novas region made by Johann Baptist von Spix e Martius, Viagem Pelo Brasil, 1817-1820, 2 vols. (Belo Horizonte, 1981) (originally published in 1823), vol. 2, pp. 58-59. On cotton cultivation in Minas Novas in the late 18th century see Daniel de Carvalho, Noticia Historica Sobre O Algodao em Minas (Rio de Janeiro, 1916), p. 14 and Roberto Borges Martins, "A Industria Textil Domestica de Minas Gerais no Seculo XIX," Il Seminario Sobre a Economia Mineira (Diamantina, 1983).

25. See the data on diamond production in W. L. von Eschwege, Pluto Brasiliensis, 2 vols. (Belo Horizonte, 1979) (originally published in 1833), pp. 117-123. These indicate that from 1801 to 1816 production increased impressively, and then declined to the late 1820s.

26. The 1776 census only included information on sex and the color categories into which the population was grouped, white, pardo, and black.

27. I am examining only race here because the 1776 population data did not include divisions of the population into slave and free. The assumption is made that immigrant populations were heavily male, and that distorted sex ratios in favor of males are indicative of immigration. Changes in sex ratios among slave and free by race will be analyzed from 1786, when data become available, through 1821.

28. In 1786 nearly 80% of all blacks in the capitania were enslaved, and while there are no data for 1776 it is likely that the percentage was even higher.

29. Among the mulatto population. sex ratios did not change substantially between 1776 and 1821. This is a population with its origins among slaves able to acquire their freedom through various mechanisms. Since in nearly all slave societies female slaves were more frequently freed than males, among mulattos more females than males were found. See the survey by Orlando Patterson, Slavery and Social Death (Cambridge, MA, 1982), Chapter 10 "Patterns of Manumission," pp. 262-296. This allowed for demographic growth through natural reproduction, and this process had the tendency to even out the sex ratio among mulattos over time.

Although the slave trade ceased as a significant factor influencing the demographic development of the Minas Gerais black population, there were regions where slaves continued to be imported to 1808. This will be discussed below. This finding contradicts the assertions by Roberto Borges Martins that the slave trade escalated sharply between 1808 and 1819 when he estimated over 4,000 slaves yearly imported into Minas. See "Minas e o Trafico de Escravos No Seculo XIX, Otra Vez," p. 10. Unless there was some radical change in the sex distribution of the slave trade away from its predominantly male character, Martins is absolutely wrong in his speculations. If there had been renewed slave imports to Minas the sex ratio would have increased, not decreased, as was the case from 1776 through 1821.

30. Even though the sex ratio among mulatto slaves in Serro Frio increased between 1808 and 1821 it should be stressed that it declined significantly among black slaves (244 to 136) and among the total slave population of the comarca (209 in 1808 to 143 in 1821).

31. This migration to Rio das Mortes seems to have been family migration rather than the heavily male migration which characterized the mining boom period among all sectors of the population. This is indicated by the age structure data found in the 1808 manuscript census to be discussed below.

32. It is unfortunate that there are no other chronological reference points in the late 18th and early 19th centuries to which we can compare the population data of 1808. The interpretation of dynamic change through static data sets such as a census report is always problematic. However, the data presented below are very suggestive.

33. See Francisco Vidal Luna, "Caracteristicas Demograficas dos Escravos de Sao Paulo (1777-1829)," Estudos Economicos 22: 3 (1992): 443-483, Tabela 6.

34. In the absence of 5 year divisions in age structure after 10 years of age in the data, this was calculated by adding all children under 10 years of age and dividing by the number of women between 20 and 40 years old plus 1/2 the number of women in both the 10 to 20 and 40 to 50 age categories. There is a small margin of error in these calculations.

35. I am examining here the entire slave population. There may be a small distortion in interpreting fertility rates because mulatto slaves may have had white, free or slave mulatto, or free or slave black fathers. However, the distortion is limited since only 10.6% of the slave population of 1808 were mulattos. The 15 to 50 total for adult females used here was an estimate derived by adding all female slaves ages 20 through 50 plus 1/2 of all female slaves ages 10 through 20. There is an undetermined, but very small, margin of error.

Francisco Vidal Luna and Herbert S. Klein, "Slaves and Masters in Early Nineteenth-Century Brazil: Sao Paulo," Journal of Interdisciplinary History 21: 4 (1991): 558. Stuart B. Schwartz, Sugar Plantations in the Formation of Brazilian Society: Bahia, 1550-1825 (Cambridge, 1985), p. 359, table 13-9. The 597 figure is the average of the child woman ratios for the three Bahian parishes. Richard H. Steckel, "Children and Choice: A Comparative Analysis of Slave and White Fertility in the Antebellum South" in Robert William Fogel and Stanley L. Engerman, eds., Without Consent or Contract: Conditions of Slave Life and the Transition to Freedom: Technical Papers (New York, 1992), Vol. 2, p. 371. In 1860 the U.S. south slave child/woman ratio was 1,056.

36. These ratios among the mulatto population are not accurate measurements of fertility among a specific population group, since either parent of a mulatto child could have been white, mulatto (slave or free), or black (slave or free). Nevertheless, these data suggest fairly high fertility rates. Of the free colored population 27% was black and 73% mulatto.

37. These graphic age pyramid images for Minas Gerais can be contrasted with the age pyramid images presented for Bananal in Sao Paulo by Jose Flavio Motta, "A Familia Escrava e a Penetracao do Cafe em Bananal, 1801-1829," Revista Brasileira de Estudos de Populacao 5: 1 (1988): 83. Bananal was a region where slave imports were important in the formation of the coffee labor force, and the age pyramid reveals the expected small base of younger slaves in 1817 and 1827.

38. See footnote 1 for some of the most important works on Minas Gerais slave demographics.

39. See Horacio Gutierrez, "Demografia Escrava numa Economia Nao Exportada: Para-na, 1800-1830," Estudos Economicos 17: 2 (1987): 297-314 who found a slave sex ratio of 104 in 1804 and 103 in 1830.

40. See Luna, "Caracteristicas Demograficas dos Escravos de Sao Paulo (1777-1829)," Tabela 1.

41. See Flavio Motta, "A Familia Escrava e a Penetracao do Cafe em Bananal, 18011829," p. 76.

42. The most inclusive economic history of colonial Minas Gerais is found in an unpublished doctoral dissertation submitted to the Faculdade de Filosofia, Letras e Ciencias Humanas of the University of Sao Paulo. See Joao Antonio de Paula, "O Prometeu no Sertao: Economia e Sociedade da Capitania das Minas Dos Matos Gerais" (Ph.D. Dissertation, University of Sao Paulo, 1988).

43. See, for example, the export data taken from the registros of Minas Gerais in 1818 and 1819 by Guilherme Eschewege in "Noticias Reflexoes Estadisticas da Provincia de Minas Geraes," RAPM, Ano 4 (1899): 737-773. Cotton, toucinho, cheeses, live cattle and pigs and a variety of other products were being exported in significant quantities to the Rio de Janeiro market in 1818.

44. See Laird W. Bergad, "After the Boom: Demographic and Economic Aspects of Slavery in Minas Gerais, Brazil. Mariana, 1750-1808," Latin American Research Review 31: 1 (1996): 67-97. Francisco Vidal Luna and Iraci Del Nero da Costa also found that in 1804 approximately 60% of the slave population in three Vila Rica districts were Brazilian born. See Francisco Vidal Luna and Iraci Del Nero da Costa, "Estrutura da Massa Escrava de Algumas Localidades Mineiras (1804)," Revista del Instituto de Estudos Brasileiros Vol. 23 (1981): 138.