1. In the United States during the first decades of the nineteenth century the number of blacks, slave and free, in the urban areas increased but decreased after 1840 (see Curry 1981). For a theoretical model to explain this phenomenon see Goldin (1986). But the model is hard to apply to Lima because it supposes that slave owners controlled the allocation of slave labor. [BACK]

2. The census was recently uncovered and analyzed by Gootenberg (1991, 109-158). [BACK]

3. Clavero (1885, 48; original emphasis) added more detail about the area, "which forms three zones. The first including the area between the mountains and the river; the second between the Rimac valley and the boundary starting at the monument of the Dos de Mayo and ending at the Dos de Mayo Hospital; and the third between the supposed border and the southern end of the city, 'Circle 5.'" It is very risky to correlate space to population, given that—as Gootenberg also points out (1991)—there were "tricky boundary problems in Lima's censuses." Even so we need a rough idea of the spatial distribution to comprehend countryside-city relations. [BACK]

4. Haitin (1983, 140 ff.) summarizes the difficulties of specifying the number of production units in estimates of the number of hectares per hacienda. [BACK]

5. To calculate these figures, I divide the sum of male and female slaves by the sum of Spaniards (citizens of European ancestry) of both genders in barrios and on estates. Citizens (with or without occupations) included indigenous persons, some descendants of free blacks, and mestizos . [BACK]

6. Cushner (1972, 180) reveals that on the Jesuit haciendas, located in Lima's hinterland and stretching from Pisco to Huaura on the northern and southern coasts, the average number of slaves per hacienda between 1665 and 1767 grew gradually, from 98.8 to 256. For the total period Cushner (1975, 180) gives an average of 162.7 slaves per productive unit. Overall, the Jesuit haciendas relied on more slaves than Lima's haciendas. [BACK]

7. For 1839 a figure of 7,922 slaves was recorded for Lima, divided between 4,792 in Lima proper and 3,130 in the valleys (Aguirre 1990, 178). In these later and more global figures the proportion of urban to rural slaves is lower than the one found for San Lázaro in 1813. This discrepancy is difficult to resolve because of the geographic imprecision in all the censuses. We must

take the figures as rough indicators of long-term trends. Perhaps the San Lázaro parish had a higher concentration of slaves than other Lima parishes. [BACK]

9. The figures recorded for the Villa hacienda in 1813 contradict the statement its proprietor made after the struggles for independence. In fact this owner, Juan Bautista Lavalle, complained in a letter to Flora Tristán about seeing the number of slaves reduced, from no less than 1,500 to 900 (cited in Aguirre 1990, 144). We might wonder whom Lavalle wished to impress; what we have recorded before the wars of independence is a notable resurgence of the slave population, somewhere on the order of 300—not 900—slaves. Haitin (1983, 141) indicates that around 1773 Carabayllo, Surco, and Bocanegra were among the most productive areas, together producing 50 percent of the total production recorded for Lima. Magdalena, Maranga, and La Legua were the least productive areas, at the other extreme: some fared better than others, discrediting the exaggerations that owners often invoked to bolster the slave trade. [BACK]

10. For Magdalena's haciendas here are the numbers of married slaves (with percentages in parentheses):

San Cayetano: 12 out of 40 adult slaves (30.0)
Maranga: 58 out of 100 (58.0)
Matalechuzas: 22 out of 42 (52.4)
Desamparados: 2 out of 7 (28.6)
Mirones: 4 out of 8 (50.0)
Cueva: 26 out of 42 (61.9)
Oyague: 22 out of 36 (61.1)
Borda: 16 out of 18 (88.9)
Orbea: 20 out of 31 (64.5)
Concha: 22 out of 53 (41.5)
Pando: 4 out of 34 (11.8)
Ríos: 12 out of 36 (33.3)
Buena Muerte: 14 out of 23 (58.3)
Ascona: 2 out of 23 (8.7)

Of Magdalena's two remaining haciendas, Palomino had no female slaves and Aramburú no married slaves. These figures could not be reproduced for the haciendas of Miraflores: marital status was recorded but the slaves were enumerated in a continuous series. In the other parishes the name of one spouse was noted next to the name of the other.

On many haciendas an unmarried adult female slave population appeared, which indicates that the percentage of slave couples was underestimated. An evaluation of the ratio between the sexes, above all in geographic areas that extended beyond the borders of the hacienda, could suggest that, perhaps as a result of the abolition of the slave trade, escapes, and self-manumissions, slave owners on the haciendas tried to contain and augment an increasingly scarce labor force. [BACK]

11. The percentages of the male and female populations of Magdalena and Miraflores closely correspond to the figure from the census of 1792 that Haitin cites (1986, 167): Lima's hinterland had 4,402 slaves, 63 percent of them men. This statistic would indicate that the situation in Surco and Chorrillos represented an exception—a notorious exception, because within Surco and Chorrillos were the two largest haciendas, the Villa and the San Juan, on which even more female than male slaves worked, both in 1790 and in 1813. [BACK]

12. One annulment, involving illicit copulation (that is, sexual relations prior to marriage and between individuals of prohibited degrees of kinship) was AA, NM, L 59 (1810—1819), Hilaria Josefa y José Nasario, esclavos de la hacienda San Javier en San Juan Bautista del Ingenio perteneciente al General D. Tomás de Arias. Another case, of parental opposition, was AA, CN, L 35 (1799-1814), Juana Bordon, mulata libre, y Joaquín Lara, sambo esclavo del Director General de Reales Rentas Estancadas, 1803. In this case Juana's mother was the one who opposed the marriage. [BACK]

13. Several works point out the gradual transition from slave to free labor arrangements (Aguirre 1990, Burga 1987, Engelsen 1977; Flores Galindo 1984; Haitin 1986; Macera 1977). Perhaps the strongest evidence has been put forth by Gootenberg (1991). For Brabón see also the document cited by Vivanco (1990, 49): AGN, L 82, C 1007. In the 1837 listing eompiled by Aguirre (1993, 52), 58.6 percent of the haciendas have no slaves (Figure 1). [BACK]

14. AGN, CS, CCI, L 288, 1842, Josefa Aparicio contra su arno D. Manuel Aparicio por sevicia. Here, we may remember Manuelita's birth as a quarterona . [BACK]

15. This assertion refers to a reading of Lima's recorded wills in the AGN. [BACK]

16. For a study of differing treatment by size of the unit of production in Martinique see Tomich ( 1990, 243). For a Brazilian study that contrasts with my analysis see Schwartz (1985, 390). [BACK]

17. The data in Table 6 show that even in valleys adjacent to Lima (including the Rímac valley), small properties accounted for 53.4 percent of all properties (342), and large ones for 8.7 percent. [BACK]

18. AGN, Cabildo, CCI, L 16, C 255, 1809, Autos seguidos por Romualda Tavira contra Doña Elena Maldonado sobre su libortad. [BACK]

19. CDIP XXVII:3:227-235. [BACK]

20. In the 1820s one dollar was equivalent to one Peruvian peso. The annual income, 55,870 dollars, represented a huge fortune, though some dowries were twice as large; it equaled the purchase price of 186 slaves. [BACK]

21. The sale of illegitimate slave children would help explain why there were so few minors on the hacienda and also perhaps why some haciendas had fewer male than female slaves. [BACK]

22. Schwartz describes similar conditions on the properties of the Benedictines in Bahia between 1652 and 1710 (1985, 355-356); it is difficult to assess to what degree Stevenson's description or that of his informants (the hacendados ) reflects this widespread legal and moral code. Aguirre also discusses the treatment of slaves on Lima's haciendas (1992, 57-74). [BACK]

23. AGN, Cabildo, CCI, L 33, C 562, 1816, Autos seguidos por José Chala, en nombre propio y de los demás esclavos de Don Manuel Menacho, sobre sevicia y para que los venda. [BACK]

24. AGN, RA, CCR, L 95, C 1161, 1802, Causa seguida por Don Antonio Ramón de Paranas contra Doña Jacoba Rubio, instigadora de la sublevación de esclavos occurida en la hacienda Punta. Violación de domicilio y otros excesos. [BACK]

25. AGN, RA, CCR, L 92, C 1129, 1801, Cuaderno incompleto de la causa seguida contra Don Juan E. Theves por suponerle intervención en el suicidio (homicidio) del esclavo Gabriel. For several similar episodes see Aguirre (1990), Flores Galindo (1984, 1990), and Blanchard (1991, 1992). [BACK]

26. The wife of the administrator of Rentas Unidas in Andahuaylas was indigenous. She ordered a zambo to make some candles and out of carelessness he let the dog eat the wick and then replaced it with a stick. ''What the said angry zambo replied was that he did not have any reason to obey an Indian woman." Later he stole her jewelry, for which offense he was beaten to death (AGN, RA, CCR, L 97, C 1185, 1802, Causa seguida contra Don José de Campo, teniente administrator de Rentas Unidas de Andahuaylas por la muerte de su esclavo Gerónimo a quien mandó azotar hasta matarlo). [BACK]

27. With one exception, the administrators of the haciendas of Magdalena that recorded ethnic figures were nonwhites. [BACK]

28. In 1801 the priests of Nuestra Señora de la Buena Muerte declared themselves mined, despite the fact that they had several properties. One of these, the most important, the Quebrada hacienda (which the 1803 figures probably refer to), was in secular hands "and headed toward total collapse." In 1808 twenty-five priests remained in this order, and of this number only nine were not incapacitated, as the priests themselves said; this number was not sufficient to fulfill their essential duty of providing spiritual assistance to those dying in jails, hospitals, and private homes. Founded by San Camilo de Luis, the order had obtained an operating license from Gregory XIV and Clement VIII between 1590 and 1591. It appeared in the colonies in 1735 and apparently could never solve its economic problems; impoverishment also affected the order in La Paz, Arequipa, Trujillo, and Cuzco (AA, Sección Convento de la Buena Muerta, L7). [BACK]

29. AA, Buena Muerte, 1808-1819, letter written by Juan Sánchez Quiñones to the viceroy in Lima, 18 April 1809. [BACK]

30. AA, Convento de la Buena Muerte, L 7, 1809, Expediente promovido por el Superior Govierno por los PP. de la Buena Muerte sobre que se les conceda permiso para la elección canónica de un prelado interino según lo determinado en sus sagradas Constituciones, yen que incide la solicitud del Provisor Síndico General de la Ciudad de Arequipa sobre remición de quatro o seis Religiosos Agonisantes a dha. Ciudad para el cumplimiento en ella en su Santo instituto, y dos mas para la Ciudad de La Paz. Lima, Octubre de 1808. [BACK]

31. The slaves of Guaca acted as did many maroon slaves and others who had some "sin" or disobedience looming over their heads to forestall their master's punishment; similarly, Antonio contacted Manuela's former master after she had been put in the panadería . On the Gualcará case see AGN, RA, CCR, L 114, C 1382, 1808, Causas seguidas contra José Espinoza y otros salteadores de caminos, Cañete; AGN, RA, CCR, L 119, C 1446, 1810, Autos contra Gavino Zegarra, Juan el Portugez, esclavos del Sor. Marqués de Fuente Hermosa en la hacienda Gualcará, Villa de Cañete por vagos, ladrones y salteadores en el Partido de Cañete. [BACK]

32. See AGN, Cabildo, CCI, L 36, C 588, 1817. In this case a slave couple fled from the Pativilca hacienda because of maltreatment by the owner, Doña Severina Alfaro. [BACK]

33. This case corroborates our observation (derived from analysis of the ethnic composition of owners, mayordomos and caporales on the haciendas of Magdalena and Miraflores) that the whiter and less remote a hacienda's owner, the tighter the mechanisms of control. [BACK]

34. AGN, EJ, CCI, L 96, 1830, Expediente que sigue Juan Castro con sus Esclabos Pedro José y otro pot cantidad de pesos. [BACK]

35. AGN, Cabildo CCI, L 5, C 51, 1802, Autos seguidos por Tiburcio María, esclavo de D. Vicente Salinas, sobre que lo venda. [BACK]

36. Reporting similar traffic in Brazil, Karasch (1987, 157) assures us that "some rural slaves moved between countryside and city as frequently as did their owners." [BACK]

37. AGN, Cabildo, CCI, L 22, C 369, 1811. The granddaughter of Augustina Carrión, slave of Don Fulgencio Guerrero and Vásquez, was Luisa Guerrero, slave of Doña Nicolasa Guerrero y Vásquez. Upon attempting to escape Luisa fell and the female owner, with some assistance, shackled and carried her to the Pescadería panadería . The judge decided that Luisa should return to her owner and not delay her owner's journey back to Ica. [BACK]

38. AGN, Cabildo, CCI, L 22, C 370, 1811, Autos seguidos por Tiburcio Arroserena, parda libre, contra D. José Martín de Toledo sobre procedimiento arbitario contra su hijo Juan Bautista, esclavo de Fr. Silvestre Durán. [BACK]

39. See Proctor (1825, 113). [BACK]

40. Blanchard (1992, 26) also quotes this passage. [BACK]

41. AGN, RA, CCI, L 70, C 720, 1807, Fransisca Suazo. [BACK]

42. AGN, Cabildo, CCI, L 11, C 143, 1806, Autos seguidos por Bernardina León, que rue esclava de D. Dámaso Jáuregui, sobre su libertad, dispuesta de su finado amo. [BACK]

43. AGN, Cabildo, CCI, L 19, C 310, 1810, Manuel Fuente, esclabe de Josefa Chabes solicita herencia de su padre natural Bartolomé de la Parra. [BACK]

44. AGN, CS, CCI, L 565, 1854, [s.t.]. [BACK]

45. Aguirre (1991, 122) notes that slave children were included in these averages, which helps explain the depreciation in slaves' value over the two decades before the abolition of slavery. [BACK]

46. This preference for female labor has also been noted by Reddock (1985, 64) for the Caribbean at the end of the eighteenth century and begining of the nineteenth, which she attributes to the higher mortality rate for males, as noted by Patterson (1967), Craton (1978), and Dunn (1972). [BACK]

47. See for example AGN, CS, CCI, L 162, 1835, Escrito de D. Fco. Chacón, albacea de D. Manuel Reyna y de su esposa Da. Mercedes Mori y tutor y curador de su hija Da. Petronila Reyna. [BACK]

48. See the proposals of Mintz (1979) and Scott (1985, 31), who interpret the system of day labor not as an indicator of the slave system's collapse but as successive contradictions within it. [BACK]

49. AGN, Notario Manuel Suárez, Protocolo 881, fs. 920, 1826. [BACK]

50. AGN, CS, CCI, L 578, 1854, Bartola Cisneros (esclava) contra Da. Isabel Bentín por su libertad. [BACK]

51. The difference—according to the protectorial decree of 24 November 1821—between patronage and slavery rested in the monthly payment of eight reales a slave received in patronato even though the slave was still property, as in slavery. [BACK]

52. AGN, CS, CCI, L 66, 1840, Patricio Negrón contra Da. Estefa Palacios y D. Carlos Relaysa por azotes inferidos en su persona. See also CDIP XV:1:173, sesión del 2 de noviembre de 1822. [BACK]

53. This dual situation was recorded for slaves in other coastal cities as well. See ADLL, Justicia Colonial, Intendencia, CCI, L 306, C 273, 1793. [BACK]

54. AGN, CS, CCI, L 576, 1854, Sebastiana García con el Síndico de la Molina sobre su libortad. For similar cases, see AGN, CS, CCI, L 574, D. Ignacio Palacios sobre libortad de su sobrina Francisca, 1854: AGN, CCI, L 560, Mariano Salazar, 1854: AGN, CCI, L 560, Lorenza García sobre la variación de dominio de su hija Fortunata, 1854. [BACK]

55. AGN, EJ, CCI, L571, 1854 [s.t.].

56. Ibid. [BACK]

55. AGN, EJ, CCI, L571, 1854 [s.t.].

56. Ibid. [BACK]

57. AGN, CS, CCI, L 569, 1854, Variación de dominio del esclavo Francisco Mansilla. [BACK]

58. Using the 1813 census of the district of Miraflores, we can check the possibilities that women would be the first to leave the hacienda. If we compare the numbers of married men and women to see which group was larger, we find among the collection of haciendas in Miraflores a surplus of fifteen married men and of seven married women. Thus the probability that married women would abandon the hacienda first was approximately 2:1. [BACK]

59. AGN, EJ, CCI, Autos seguidos per D. Manuel Esteban de Arsola sobre la propiedad de los esclavos Leandro y Ebaristo Arsola. Also of interest in this case is the mayordomo of Bocanegra's payment for the breast-feeding of children after the slave mothers had left the hacienda; it shows his determination to keep up the numbers of slaves on the hacienda and marks the boundaries between the responsibilities of master and slave concerning children. The November 1821 decree required owners to feed rather than pay wages to libertos during lactation (until they reached the age of three). [BACK]

60. AGN, Cabildo, CCI, L 35, C 578, 1817. [BACK]

61. AGN, EJ, L 92, 1830 [s.t.]. [BACK]

62. AGN, EJ, CCI, L 571, 1854 [s.t.]. [BACK]

63. These calculations are based on three charts preserved in the BN, D 8525, Cuerpos de Militias Provinciales Disciplinadas y Urbanas de Caballería

en el Virreynato del Perú con expresión de los partidos e Intendencias a clue pertenecen; D 8526, Cuerpos de Dragones de Milicias Provinciales Disciplinadas y Urbanas; D 8527, Cuerpos de Infantería Provinciales Disciplinadas y Urbanas, Lima 31 de diciembre de 1816. [BACK]

64. CDIP V:1:250-251, Oficio de Juan Delgado a Bernardo Monteagudo, Secretario de Guerra y Marina, Sayán, 3 de marzo de 1821. [BACK]

65. CDIP V:1:154-155, Circular enviada al comandante D. Francisco Aldao, Febrero de 1825. [BACK]

66. AGN, RA, CCR, L 106, C 1287, 1806, Autos de oficio seguidos contra Agustín Guerrero, Juan josé Ortiz y otros por los delitos de robo y asaltos en los caminos. [BACK]

67. AGN, RA, CCR, L 114, C 1382, 1808, Causa seguida contra José Espinoza y otros salteadores de caminos, Cañete.

68. AGN, RA, CCR, L 104, C 1263, 1805, Autos criminales seguidos de oficio por la Real Justioia contra el esclavo Antonio Caballero y otros por los delitos de haberles encontrado en su poder armas prohibidas y por ladrones en el camino de Chillón. [BACK]

67. AGN, RA, CCR, L 114, C 1382, 1808, Causa seguida contra José Espinoza y otros salteadores de caminos, Cañete.

68. AGN, RA, CCR, L 104, C 1263, 1805, Autos criminales seguidos de oficio por la Real Justioia contra el esclavo Antonio Caballero y otros por los delitos de haberles encontrado en su poder armas prohibidas y por ladrones en el camino de Chillón. [BACK]

69. Proctor (1825, 215-216). [BACK]

70. The notarial record books charted the methods of manumission and recorded how slaves had been acquired and where they had been born (and thus whether they had belonged to a hacienda). I consulted all the notarial record books in the AGN for 1830, 1840, and 1850; for 1840 I found only scanty documentation. It is possible that notarial record books were lost, or that during the period's usual political turbulences no notarial records were made; perhaps the transactions actually diminished, since at the end of the 1830s haciendas tried to reimpose control over the slave population. [BACK]

71. For 1836 the total number of slaves amounted to 5,971, and for 1845 it was estimated at 4,500 (Jacobsen 1974). [BACK]

72. Between 1560 and 1650, Bowser (1977, 363-364) notes, 33.8 percent of Lima's slaves were liberated unconditionally, and of this population, 92.2 percent were women and children under the age of fifteen. As we have seen, in 1850 (two hundred years later) this type of freedom accounted for only 9.1 percent of the cases of manumission, including conditional grants by owners. In all Latin American slave centers, self-manumission was central. For statistical evidence see Klein (1986, 221 ff.). [BACK]

73. The cases studied by Flores Galindo (1984) refer to the period 1760-1809, and those by Aguirre (1990) to 1836-1839. [BACK]

74. For rural-urban relations see the significant analysis by Fields (1985) in her work on nineteenth-century Maryland. [BACK]

75. Goldin (1976, 51 ff.) records movement in the United States between 1850 and 1860 but in the opposite direction, and largely determined by slave owners' interests. [BACK]

76. Haitin (1983, 177) claims that the archdiocese of Lima produced 37 percent of the tithes in the viceroy, and that tithes for Lima between 1774 and 1779 amounted to 126, 546 pesos per year, and 148, 886 pesos annually between 1790-1794. Despite the risks of using information about tithes, we must conclude that not all productive units followed the same destiny over the course of this long cycle. [BACK]