"The shelter of the uniform": the Brazilian army and runaway slaves, 1800-1888.

Antonio de Moura's luck ran out at 3:00 pm on 28 November 1863. Three months earlier he had voluntarily enlisted in the Brazilian Army's Eighth Infantry Battalion, then stationed in Salvador, the capital of the province of Bahia. Shortly thereafter, a company captain withdrew Moura from regular duty to serve as his orderly. Unbeknownst to the officer, however, his new batman was a runaway slave. That fateful afternoon, while the two were returning from Fort Sao Pedro to the captain's house, they encountered Jose da Veiga Ornellas who, recognizing Moura, accused the captain of harboring his fugitive slave. The officer later reported: "Upon hearing this, [Moura] responded to the young fellow [Ornellas] that he was mistaken, that he was not and had never been his slave, that this was a plot just like the one he had perpetrated on his sisters when their mother died." Despite this implicit recognition of Ornellas and the accusation that Ornellas had once sought to enslave him, Moura steadfastly denied knowing the man.

To placate Ornellas, the captain explained the proper procedure for reclaiming fugitive slaves from the army. He then questioned his orderly privately about the allegations, going so far as to promise him help in securing his freedom, were he really a slave. Mourn insisted upon his free status and named two former employers who could attest to it. Satisfied with Moura's offer to supply references and convinced that Ornellas's claim had been an error, if not a deliberate falsehood, the captain sent Moura to do his chores. Shortly thereafter, the battalion's adjutant arrived with orders to take Moura back to the barracks because he was being claimed as a slave. Moura overheard this and made good his escape, scrambling over the back wall of the garden, with the officers in pursuit. He threw off his pursuers in the woods on the outskirts of the city; seven weeks later, the authorities captured him. Facing desertion charges and the prospect of a return to slavery, Moura broke out of the barracks lock-up and disappeared in April 1864.(1)

The paper trail on Private Antonio de Moura, allegedly the slave of Jose da Veiga Ornellas, ends here, leaving numerous unanswered questions. Was he really a slave? Or had Ornellas tried to enslave him? If he were a fugitive slave, why did he join the army? In Moura's case, there are no straightforward answers to these questions; at the very least, however, he did not lack company. In nineteenth-century Brazil, slaves routinely ran away to join the army as volunteers while others were impressed, to the dismay of their owners, who were then forced into often long and cumbersome legal and administrative proceedings to reclaim their property. The documents left by 277 of these cases, originating primarily in the northeastern sugar-growing province of Bahia, raise important questions about the nature of military institutions - in this case, the army - in slave societies. Furthermore, the 276 men who moved between the status of slave and soldier (one of them joined the army twice) exemplify a liminal world between slavery and freedom, where the fortunate might escape bondage while the unlucky slid back into it. Analyzing the strategies of slaves in this gray area and, in particular, their artful use of the contradictions in the Brazilian state apparatus is one of the purposes of this article.

The second concern of this paper is the army's policy toward slavery. Here I take issue with a broad scholarly consensus that views the nineteenth-century Brazilian army as a "progressive" institution with strong abolitionist sympathies. Whether attributed to the increasingly middle-class origins of the officer corps or to a fundamental contradiction between a "professionalizing" army and the slave society that surrounded it, army-officer abolitionism is presented as an important contribution to Brazil's ending of slavery in 1888.(2) While it is true that some officers actively campaigned against slavery in the 1880s, in its dealings with runaway slaves the army exhibited far more complex and even contradictory attitudes. In principle, fugitives such as Antonio de Moura were to be returned to their masters, once they demonstrated proof of ownership. Until the 1880s, the Brazilian government and army thus upheld property rights; however, the army bureaucracy's stubborn legalism frequently vitiated this intent and produced unexpected outcomes, to the benefit of individual fugitives, a few of whom actually gained their liberty through enlistment. Nevertheless, officers' willingness to uphold the law by returning fugitives - there is no evidence that they complained about discharging slaves - raises grave doubts about the portrayal of the Brazilian army as an abolitionist institution.

The first section of this article sets out the legal principles that governed recruitment and slavery, as well as the administrative procedures that the Brazilian government and army developed to deal with runaways in the ranks. The complications introduced by the inability to distinguish slaves from free men at the margins of slavery and the nature of recruitment - primarily impressment - are the subject of the second section. A third section sketches a brief quantitative profile of the runaways while, in the fourth part, I turn to the strategies of slaves in the army. A concluding section returns to the question of the army's attitude toward slavery and sets the Brazilian experience in the context of other slave societies.

Recruitment and Slavery: The Legal Principles

As in all Western slave societies, Brazilian bondsmen could not serve in the army nor could they be conscripted. No law explicitly mandated this exclusion, which was little more than common sense for slaveholders. The jurist of Brazilian slavery, Agostinho Marques Perdigao Malheiro, supplied the full legal argument in 1866: because slaves were not citizens and military service fell by law only on Brazilian citizens, they could not enlist.(3) Furthermore, the drafting of slaves violated the constitution's guarantee of property rights, a theme that recurs in owners' petitions for the return of impressed slaves.(4) To nineteenth-century Brazilians, both of these arguments were patently obvious, the latter so obvious that Malheiro did not bother to mention it.

In wartime, American slave societies frequently abandoned these principles. From the competing offers of liberty in return for military service of the British government and its rebel North American colonies in the 1770s and early 1780s, through Simon Bolivar's decrees of the 1810s that simultaneously freed and drafted slaves in Venezuela and Colombia, to the Cuban patriots' enlistment of slaves in the late 1860s, the independence wars of American colonies are replete with instances of slave recruitment.(5) Even in such military emergencies, however, American governments generally acted cautiously when enlisting slaves. Loyal slaveowners could usually rest assured that their property would not be touched by their government.(6) The province of Buenos Aires, for example, expropriated a few thousand able-bodied slaves during the 1810s to fill the ranks of its armies but took care to compensate the owners.(7) The British government, unable to acquire sufficient recruits for its West India Regiments during the Napoleonic Wars, purchased African slaves for these units instead of drafting the island planters' property.(8) In short, wartime recruitment of slaves in the Americas rarely implied a complete rejection of slavery and usually acknowledged masters' rights over their property.

In this regard, nineteenth-century Brazil was no exception. On neither of the two occasions that the government recruited slaves did it formally challenge slaveowners' rights. After the Independence War in Bahia (1822-1823), the new empire bought out the claims of owners whose slaves had enlisted against the Portuguese and, during the Paraguayan War (1864-1870), the government compensated slaveowners who voluntarily freed their property on the condition that the freedmen immediately enlist.(9) To be sure, these episodes undermined slavery but that was not the government's intent.(10) Owners were, in all cases, to receive compensation and, during the Paraguayan War, the government returned at least 36 able-bodied fugitives to owners who did not want to part with their human property, thus recognizing the primacy of masters' rights over the claims of the state.(11)

Although Brazilian law clearly and unequivocally distinguished between slave and free when it came to formal military service, the legal status of soldiers and the nature of recruitment combined to blur this distinction in practice. Upon swearing their oath to the flag, Brazilian soldiers entered a partially autonomous legal realm. Obligated to complete their enlistment term, usually six or eight years, soldiers could not be discharged without express orders from the president, the highest provincial civilian authority. After two months of service, only the War Ministry could authorize such discharges.(12) While Brazilian soldiers did not enjoy the full legal privileges of the fuero militar familiar to students of colonial Mexico, they nevertheless responded first to military law.(13) Thus, on two separate occasions, owners solicited imperial pardons for their slaves who, after enlisting, had deserted three times and were serving six-year terms at hard labour. Although the highest military court recognized that the improper enlistment of slaves rendered all subsequent actions - including the sentence for desertion - null and void, it nevertheless held that these men could not be discharged and returned to slavery before completing their sentence unless an imperial pardon overturned the court-martial conviction.(14) In short, the military jealously defended its legal authority over enlisted men.

As a result of this legalism, discharges, whether of free men or of slaves, required proof that the soldier had been improperly impressed. In the case of runaways in the ranks, the onus lay on masters to demonstrate their ownership to the satisfaction of civil and military authorities. Owners' petitions would thus include copies of the slave's registration, receipts for taxes paid on him, bills of sale, baptismal certificates, or the relevant articles of partilhas (judicial divisions of inheritances) in which the slave was assigned to a given heir. If any doubt remained as to the man's identity, the claimant had to prepare a justificacao, a deposition sworn before a judge by three to five witnesses that the individual in question was, in fact, the missing slave. The presidency then reviewed the documents and submitted them to the commander of arms (the chief military authority in the province) who evaluated them and had the slave questioned. If both authorities were satisfied, the president would order a discharge.

Not all masters could, however, supply sufficient documentation. Joao Helling, a German-born founder in Santo Amaro, a town near Salvador, lacked proof of his ownership of Luiz de Moura because, he claimed, his title to the slave had been stolen. Not even the testimony of seven witnesses in two justificacoes convinced authorities that Luiz de Moura was indeed the slave whom the German had purchased many years earlier.(15) Delays in submitting documents to the presidency might exhaust the two-month time limit during which discharges could be issued in Bahia, obliging the owner to file a new claim with the War Ministry in Rio de Janeiro.(16) Well-connected individuals, on the other hand, could dispense with formalities. When a brigadier discovered his runaway slave as a soldier in 1842, the slave's commander, the Baron (later Duke) of Caxias, advised that the claim, backed only by a testimonial, should be accepted, for the document's four signatories were men "of recognized uprightness."(17)

The demand for detailed proof of slave status does not reflect any incipient abolitionist or emancipationist sentiment in army and government circles, however much it may have complicated slaveholders' lives. Authorities recognized the gravity of returning a soldier, by definition a free man, to slavery and claims which turned out to be false or cases of mistaken identity were not unknown.(18) Far more important, however, were the army's reluctance to part with a soldier under any circumstances and the institution's stubborn legalism - improperly impressed free men seeking discharges faced legal difficulties similar to those that plagued slaveowners.

Two additional legal and fiscal considerations affected slaveowners when they sought to reclaim their property. The military recognized the old Roman-Law principle that slaves who performed service to the state as soldiers should be freed.(19) While never formally expressed in Brazilian law - it would have been an open invitation to slaves to flee and join the army - the army nevertheless did not return fugitives who had distinguished themselves while in the ranks. Thus, the final legal opinion on the claim over Joaquim, a slave from Bahia who had enlisted in 1824 in Rio de Janeiro, held that, although he had not fought in the Independence War, he should not be returned to his owner, for he had served in the force that suppressed a republican rebellion, earning the campaign medal.(20) Instead, the government compensated the owner, thus quietly upholding property rights and preventing the return to slavery of a potentially rebellious slave. This principle, repeated in an 1842 decision on a slave who belonged to a government estate and had joined the army, crept into a manual of army legislation published in 1874. By this time, an emancipation fund, set up under the terms of the 1871 Free Womb Law, provided the means to compensate the owners of slaves who had distinguished themselves as soldiers.(21) These provisions, however, remained limited in practice; as late as 1880, the Council of State recommended the return to captivity of a fugitive on the grounds of inadequate service.(22)

While compensating the owners of the slaves that it kept as soldiers, the state in turn demanded reimbursement from those whose slaves it returned.(23) In other words, even though authorities had illegally seized property through impressment or had failed to verify the free status of volunteers, the state's fiscal interests led it to require owners to pay for the upkeep of men dismissed as slaves. While slaveholders, on occasion, argued that they were more properly owed compensation for lost labor time, most quickly paid the bill for the value of rations, uniforms, salary, and enlistment bonuses.(24) This principle could be carried to curious extremes. Maria do Rosario Ursulina de Jesus doggedly pursued her claim over Corporal Zacharias Jose de Miranda, impressed in 1871. By 1877, she was willing to accept 1,500 milreis (about $750) and the commander of arms agreed that her rights should be bought out, given Miranda's lengthy service and "exemplary conduct." Referring to calculations done in May 1876, however, he estimated that the army had already spent more than 1,700 milreis on the soldier, thus more than liquidating Maria do Rosario's claim!(25)

Recruitment in a Slave Society

Military recruitment and the social composition of the rank and file conspired with this bureaucratic labyrinth to confound slaveowners. Although no exact proportions can be established, most Brazilian soldiers were forced into service by the justice system as punishment for petty (or even major) crimes, vagrancy, or violations of moral standards. Press gangs picked up others during periodic recruitment drives. Most scholars conclude that military service therefore fell upon the poorest of the free poor, primarily young men of color.(26) The free poor, however, were a social category into which the slave class blended, sometimes imperceptibly. Slaves who worked on their own account, remitting only a portion of their wages to their owners; slaves on errands for their masters; conditionally-freed slaves struggling to earn the balance owed on their freedom or fulfilling testamentary provisions to accompany and serve their masters' heirs; and runaways seeking to make new lives for themselves; all were indistinguishable from the free population from which soldiers were forcibly drawn.(27)

Two examples illustrate these points. In 1868, a press gang seized Manoel Pereira de Santa Anna, the cashier of a commercial establishment near Salvador's docks, as he was closing the shop. A woman immediately filed for his release on the grounds that he was her slave, conditionally freed in 1858 to serve during her lifetime. She had duly registered him and had paid all the taxes required of a slaveowner. His only obligation to her was a monthly payment; she considered him typical of slaves who through "vanity ... wish to appear as free [men]." The National Guard captain who had ordered the arrest doubted that Santa Anna was a slave, for men of such low condition could not be clerks and would not have been tolerated as such by free men.(28) Unfortunately, the outcome of this case is not known, but Santa Anna's responsible occupation, requiring literacy and the confidence of his employer, makes his an exceptional case of a slave who merged into free society. Another more humble individual, the forty-year-old illiterate Joze Luis de Souza Reis, a small farmer in Salvador's suburban parish of Brotas, arrested in 1870, turned out to be Felipe, slave of the late Maria Theodora das Virgens, the owner of a farm in a nearby county. In his confession, he recounted 25 years of life on the run from the day that, as a fifteen-year-old, he had been arrested while on an errand for his mistress. Sent to Salvador and drafted into the navy, he adopted his new name and kept silent about his condition; epilepsy earned him a discharge for medical reasons after three years at sea. Later, he volunteered for the army, but was again dismissed on medical grounds. After the outbreak of the Paraguayan War, the authorities disregarded his discharge papers and sent him to the front; the army finally returned him to Bahia after he suffered a seizure during a battle.(29)

Manoel Pereira de Santa Anna and Jose Luiz de Souza Reis exemplify the many slaves who passed, sometimes undetected, through the army. Repeated orders not to accept men of color in the ranks without first verifying their free status had little effect, and the 277 cases that I have located are probably but a small proportion of the total.(30) In twenty-two and a half months between 1 March 1841 and 19 January 1843, the army discharged 146 men in Bahia, of whom fully 14 (almost 10 per cent) were slaves.(31) For the longer period of 1841 to 1845, I have located more detailed references to only five slaves in the army (Table 1).

Of these 277 slaves, only about 55 per cent actually became soldiers. Owners discovered the rest as recrutas, impressed men or volunteers awaiting formal enlistment; indeed, the distribution of these cases reflects the intensity of recruitment (Table 1). When authorities stepped up recruitment efforts, they were more likely to impress slaves. Enlistment drives began with public appeals for volunteers, thus inadvertently advertising the opportunities to would-be runaways. The two years in which the largest numbers of claims are recorded - 1865 with 40 and 1867 with 30 - nicely illustrate this point. The former was the first full year of recruitment for the Paraguayan War, while the latter was the first year in which the government "purchased" slaves for the war effort. Overall, a similar pattern holds true, with peaks in the 1820s and 1860s, reflecting slaves reclaimed after the independence war and recruitment drives during the Cisplatine (1826-1829) and Paraguayan Wars. The suspension of recruitment in the aftermath of the Cisplatine War and the reduction of the army to less than half its previous size caused the trough of the early 1830s. In response to internal revolts, army strength rose and recruitment intensified in the second half of the decade. The decline in the 1880s follows reductions in army strength in 1877 and 1880 and, of course, coincides with the final collapse of slavery in 1888.(32)

Table 1:

Distribution of Cases by Year and Status of Slave

Years                Status of Slave             Total
              Enlisted Man      Recruta(*)

1782               1                0               1
1816-20            2                0               2
1821-25            9                2              11
1826-30            4                8              12
1831-35            1                1               2
1836-40            4               11              15
1841-45            5               10              15
1846-50            8               10              18
1851-55            7                7              14
1856-60           14                9              23
1861-65           37               22              59
1866-70           30               33              63
1871-75           13               10              23
1876-80           11                3              14
1881-85            5                0               5

Total            151              126             277

(*)Unenlisted Impressed Man or Volunteer
Sources: 277 cases of slaves claimed from army.

The Cases: Some Generalizations

Taken together, the 277 cases analyzed here represent, with some restrictions, a cross-section of nineteenth-century Brazilian slavery. The diversity of the 93 owners who can be identified in some way confirms the depth of slaveholding in Brazilian society; people of virtually all social classes owned slaves.(33) Many of the occupational designations are vague, such as "proprietor" and "businessman," and say little about the individual's economic activities. They range from barons (five) and sugar planters (two), through priests (four), an engineer, and a civil servant, to women who described themselves as poor widows (ten), freed-persons (four, including three Africans), and even one slave. The predominance of presumably small-scale slaveowners (such as poor widows and freedpeople) among the 93 identified is consistent with the patterns of recruitment and the social origins of the rank-and-file described in the previous section, for their slaves were far more likely to engage in occupations that brought them into free society than those belonging to large owners such as sugar planters.

Most striking about the ethnic data on the 276 slaves (counting only once the slave who enlisted twice) is the presence of only two Africans (Table 2); one of them had been imported as a child and spoke Portuguese so well that he could pass for a creole.34 Until the suppression of the illegal slave trade in the mid-1850s, Africans comprised a majority of Brazil's slave population.(35) Freed Africans were not considered citizens and, especially after the 1835 slave revolt in Salvador, they were treated as dangerous aliens, subject to deportation.(36) Their consequent exclusion from the military served Amaro Jose Correia well. An African freedman, he found himself impressed in 1868. All that he needed to obtain his release was a statement from an officer that, "by the marks on his face and his accent" he looked to be a Nago (Yoruba).(37) However much Brazil may have needed men to fight against Paraguay, it would not require the African-born to serve.

Table 2:

Race and Ethnic Origin of Slaves Claimed from Army

Race/Ethnic Designation     Number      Per Cent

African                        2           0.7
Creole                        65          23.6
Mulatto                       80          29.0
Caboclo                        2           0.7
Unknown                      127          46.0
Total                        276         100.0

Note: One creole enlisted twice but is only counted once.

Sources: 277 cases of slaves claimed from army.

Occupation and residential data are available on but a handful of these slaves, for owners tended to mention only trades or skilled occupations (Table 3). Domestic servants, agricultural workers, or unskilled urban laborers are rarely listed as such. Nor is it easy to distinguish between rural and urban residents. Presumably urban slaves who enjoyed the greater freedom of the city had more opportunities to run away to the army or to suffer impressment than those restricted to plantations. Even rural slaves might nevertheless frequent cities. The owner of a farm on Itaparica Island regularly sent his slave, Andre, across the bay to fetch goods in Salvador, despite the fact that he had run away and joined the army in 1870. On 7 December 1873, Andre disappeared again, only to surface in the Sixteenth Infantary Battalion in March 1874.(38) Andre's two stints as a soldier bring us to the question of slaves' use of the complex institutional structure sketched out in the previous sections, to which we now turn.

Table 3:

Occupations of Slaves Claimed from Army

Occupation                            Number

Tailor                                  4
Shoemaker                               4
Mason (including one apprentice)        4
Carpenter                               5
Cabinet Maker                           1
Painter                                 1
Cigar-maker                             1
Cashier                                 1
Baker                                   1
Barber                                  1
Lackey                                  1
Sailor                                  1
Cook and Sailor                         1
Domestic Service                        2
Agriculture                             6
Unknown                               242
Total                                 276

Note: The slave who enlisted twice, employed in agriculture, is
counted only once.

Sources: 277 cases of slaves claimed from army.

Slave Strategies

Regardless of their occupation or the status of their owners, slaves pursued numerous strategies to better their lives, from foot-dragging, through the maintenance of autonomous cultures, to outright rebellion, in a constant process of conflict and negotiation with their owners.(39) Joining the army formed part of that process. The 54 slaves who voluntarily enlisted and the 3 substitutes (Table 4) actively sought "the shelter of the uniform," as one owner accused his slave of doing, when the latter volunteered in 1877.(40) Many of the 151 impressed men, often already runaways, remained silent about their condition, as Jose Luiz de Souza Reis had done three times. In contrast, others requested return to their masters, claiming their slave or conditionally-freed status as soon as they were impressed. The enormous variety in these cases and the fact that much of the documentation is incomplete makes generalization difficult, but several broad slave strategies can be discerned.

Virtually all of the men who sought to escape slavery by enlisting changed their names. One officer declared in 1824 that searching for runaways in the ranks by name was fruitless, for all adopted aliases. Therefore, unless the owner could identify the suspected slave, he could do nothing.(41) In an age before photography, a simple name change established a new identity, as long as the fugitive avoided contact with people who had known him as a slave.

To this end, the army offered runaways an effective means of putting distance between them and their owners. Notices of runaways frequently mentioned the possibility that slaves would seek to enlist.(42) Pedro and Benedicto, slaves of different owners in Alagoas, together fled to the neighboring province of Sergipe in 1860 and enlisted in the company stationed there. As volunteers, they had the right to choose their unit and opted for transfer to Rio de Janeiro. En route, however, the army landed them in Salvador and enlisted them in the local garrison, where their owners found them.(43) When discovered in the main barracks in Rio de Janeiro sometime during the Paraguayan War, Geraldo, then known as Jose, promptly requested and received a transfer to the front.(44) Others were less fortunate and found that the army moved them closer to their former owners. Luiz de Moura, whose master, the German founder, lacked adequate documentation to prove his ownership, was assigned almost immediately after his impressment to the detachment in his home town. Neighbors recognized him while he stood guard in the main square. Asked about his uniform, the mark of his status as a soldier, the hapless Luiz responded that it was none of the questioner's business.(45)

Table 4:

Enlistment Status of Men Claimed as Slaves

Status                 Number        Per Cent

Volunteer               54             19.5
Impressed Man          151             54.5
Substitute               3              1.1
Unknown                 69             24.9
Total                  277            100.0

Sources: 277 cases of slaves claimed from army.

Besides distancing themselves from their owners and establishing new identities - strategies common to virtually all runaways - slaves who joined the army took advantage of the institution and enlisted it as an unsuspecting ally in their struggles with their masters. For runaways, slaves who considered themselves free on the basis of oral promises, or even men who feared enslavement, joining the army literally brought them the protective cloak of the uniform. They gained a patron who might look after their interests, as the captain had promised Antonio de Moura. Thus, in 1854, the mulatto Francisco de Macedo reported to army headquarters in Salvador and declared that he wanted to enlist. When questioned about his status, he explained that, after having traveled overland from his native Ceara to Salvador in the company of one Jose Pereira de Castro, he now feared that Castro was about to sell him.(46)

If Macedo's case looks clear enough - he sought to enlist to avoid enslavement - others are more murky. In the dim areas where slavery merged into freedom and slaves into the ranks of the freed and free poor, the status of individuals could not always be determined. Here men and women could pass from slavery to freedom or slide back from a precarious freedom into slavery. Antonio de Moura considered himself a free man and accused Ornellas of attempting to enslave him, as he had apparently done to Moura's sisters after their mother's death. Luiz de Moura told his commanding officer that he had been "born of a free womb and entrusted to the said Joao Helling to learn the trade of founding."(47) The unlucky Moura considered himself an apprentice while the founder considered him a slave. When discovered in 1876, Alexandre Gomes da Silva, formerly Ephiphanio, a slave from a ranch on the coast, did not contest his identity, but sought to hold his owner to her oral promise of freedom. He emphatically declared that he would not return to her and that he had enlisted because she had "always told him that he was a freedman and that he could go where he pleased."(48)

We cannot know the truth in these cases. Anything is possible at the margins of slavery, where bondage and liberty merged. Here, more than anywhere else, individuals in the patronage-based society of nineteenth-century Brazil needed reliable protectors. The "patronless poor" risked all, even enslavement.(49) If Luiz de Moura really was an apprentice, then Joao Helling was no longer a reliable patron and Moura did what he could - run away to find another, hopefully more reliable, patron. The prominent abolitionist poet, Luiz Gama, did exactly that. In 1880, he recounted finding life-long patrons while serving as an enlisted man. The son of a freed African and an impoverished "nobleman," he was sold into slavery in 1840 by his father to settle gambling debts. More fortunate than most slaves, the young Gama learned to read and write and made himself useful to his new owner's slave-dealing business. After secretly obtaining proof of his free status, he ran away and joined the army in 1848. During his six years of service, he reached the rank of brevet corporal and, more important, he caught the attention of officers. The major in charge of the department in which he worked as a scribe during his spare time "became my friend;" in 1880, this patron held a senior bureaucratic post. From a career magistrate whom he served as an orderly, Gama "earned esteem and ... protection," as well as useful lessons in "high culture [letras] and civics, of which I am proud."(50)

Gama's experience is, of course, unique - literacy alone gave him enormous advantages enjoyed by only a tiny minority of slaves - yet his account of his passage from slavery through the army to freedom renders explicit the strategies that even illiterate slaves could employ. Benedicto, one of the two slaves who had fled from Alagoas to Sergipe to enlist, claimed that he had done so "to free himself from his master who mistreated him without pity [and] did not permit him to seek another master."(51) The justification for his flight, with its appeal to notions of moral behavior on the part of owners, points to implicit understandings about the "legitimacy" of slavery. When masters failed to live up to the paternalist standards with which they justified their dominion, slaves claimed a right to seek new patrons. Denied the opportunity to search for a better master, Benedicto went one step further and sought his new patron in the army.

More fortunate than Benedicto, whom the army sent back to his abusive owner, Arsenio Teixeira dos Santos played on the institution's legalism in his fight to pass from slavery to freedom. For his troubles, he nevertheless spent five years in army prisons. Days after Santos had volunteered for service in 1860, Sebastiao Jose Lopes identified him as his slave and declared that Santos had enlisted because he did not want to subject himself to domestic service. Santos categorically denied this, claiming to be a freedman. His birth certificate, which he had entrusted to an officer, listed him as freed at birth, but it lacked some necessary legal formalities. Lopes, on the other hand, could only present documents to prove his ownership of Santos after his marriage to the woman who had freed him. She was now dead, and Santos could not locate his godfather who might have clarified the circumstances of his baptism. In Rio de Janeiro, the adjutant general recommended resolving the case by buying out Lopes's claim, but nothing came of this and Santos was, despite petitions on his part, still in prison in July 1865.(52)

For men like Gama, Santos, the two Mouras, and Macedo, joining the army meant gaining the protection of the corporation to confirm their tenuously-held freedom. On the other hand, many a slave who found himself impressed refused to stay in the ranks. Contemporary observers and modern scholars have often concluded that conditions in the army were so bad that slaves preferred slavery to military service.(53) Enough evidence to the contrary has been adduced here but, in at least 17 cases, impressed slaves spontaneously confessed their condition to avoid enlistment. One or two of these confessions were spurious, as free men sought to avoid army service, but the rest were truthful. We can really only speculate why. Slaves who had inched their way into free society might well have judged their semi-free status preferable to confinement in army barracks. Conditionally-freed slaves, with the typical obligation to serve during their owners' lifetimes, might have looked forward to the predictable death of an elderly owner. Joining the army meant leaving behind friends, family, and loved ones. When Luiz Antonio de Oliveira fled his master's ranch to volunteer at the outbreak of the Paraguayan War, his tearful mother sought out a neighboring planter to tell him that her son had enlisted.(54) While there is no indication that her distraught condition affected Oliveira's decision to run away, the case does point to the importance of affective ties among slaves and, incidentally, the networks of allies that some slaves could build beyond the borders of the plantation. Finally, slaves involved in lawsuits for their freedom did not want their cases jeapordized by their impressment. Damiao Antonio do Sacramento, imprisoned as a recruta, petitioned the Chief of Police for his release in 1873, on the grounds that the courts were still ruling on the validity of his late mistress's three wills, thus reinforcing his new owner's claim.(55) In the first of three wills dictated in the last month of her life, she had freed the 69 slaves who worked her cane farm near Salvador. On the day before her death, the heirs who would have been dispossessed by this will prevailed upon her to revoke it but, just before passing away, she dictated a third testament, substantially similar to the first. Sacramento's hopes for a favorable judgment were dashed in 1874 when the Supreme Court upheld the validity of the second will.(56)

As most of these cases have demonstrated, the shelter of the uniform was but precarious cover. Officers who applied the rules with all their contradictions may have inadvertently helped runaways, but civilian masters did not lack allies. A glance at the moments when slaves were discovered reveals a society in which most free men collaborated in the search for runaways. Many a claim for the return of a fugitive began after chance discovery by a relative, acquaintance, or neighbor of the slave's owner. If army service provided opportunities for a slave to distance himself from his owner, it might also bring him together with someone who had known him years earlier. Luiz ran away from Serrinha, in the interior of Bahia, in 1861. Eight years later, a corporal from that town met the fugitive, now a soldier, in an army hospital in occupied Paraguay. In a letter to his mother, the corporal mentioned Luiz among the other soldiers from Serrinha whom he knew to be alive and well. The letter was read in public - an indication of the hunger for news from the distant front - and Luiz's owner thus discovered the whereabouts of his slave. As demobilized veterans returned during the following year, he called upon them to testify in two justificacoes backing up his claim for Luiz's value.(57)

Constantly facing the possibility of discovery, life must have been tense for the runaways in the army. Little wonder that many confessed their condition when caught.(58) Antonio de Moura lost his composure when discovered by Ornellas but still managed to deny his slavery, as did Luiz dos Santos. Discovered in 1874, he solemnly declared before officers and the tutor of his new owner, his late owner's grandson, that he was not a captive and that he did not recognize his new owner.(59) None, however, protested more desperately than Jose Joaquim de Santa Anna, who volunteered in 1865. When a "citizen" arrived at the barracks and identified him as a slave, the company commander had him imprisoned pending a formal claim. Jose Joaquim, however, had prepared for this eventuality, and took the arsenic secreted in his pocket. Despite ministrations of purified oil to dilute the poison, he was dead before the officer could write his report.(60)

Armies and Slavery

What does the treatment of runaways tell us about the Brazilian army's position on slavery? At the risk of digressing somewhat, we can reemphasize that the Brazilian exclusion of slaves from formal military service stands squarely within the Western legal tradition that identifies such service with citizenship and denies both to slaves. With a multi-racial army forcibly drawn from a free and freed population visibly indistinguishable from slaves, however, the Brazilian army inevitably faced the problem of slaves - whether runaways or men inadvertently impressed - in the ranks. In contrast, the other great slave power of the nineteenth-century Americas, the United States, maintained a color bar that effectively excluded blacks, both slave or free, from the pre-Civil War army.(61) Other slave societies resolved the issues posed by slavery and military service in very different manners. Imperial Russia recruited its rank-and-file among the unfree population of serfs until 1861. Facing a dilemma that would have been familiar to Brazilian officers - what to do with unfree men who had served as soldiers? - the Russian army avoided the problem of returning veterans to serfdom or sending them to their villages as free men by setting the enlistment term at 25 years, effectively for life.(62) The Islamic institution of military slavery stands at the opposite extreme of the American experience of slavery and military service. Mamluks, the quintessential Islamic soldiers, formed a slave caste whose members, despite their unfree status, often came to dominate Middle Eastern states.(63) In a society with very different concepts of slavery, freedom, and citizenship, Islamic military slavery is a world far removed from nineteenth-century American slave societies, where military service implied a movement away from servitude.

Given that the enlistment of slaves raised fundamental questions about citizenship and the nature of Brazilians' relationship to their government, the army's treatment of runaways in the ranks is an ideal issue on which to reassess the institution's alleged abolitionism. No evidence can be found in these 277 cases to suggest that officers disagreed in principle with returning fugitives to their owners. If officers had private opinions about slavery, they never slipped them into their correspondence about the runaways. Indeed, the evidence for the army's anti-slavery stance is rather thin. The most frequently cited incident, the Military Club's 1887 request that the army no longer be employed to pursue fugitives in Sao Paulo, where slaves were in open revolt against planters who still opposed emancipation,(64) apparently announced an eleventh-hour conversion. I can locate only one public manifestation of abolitionist sentiment among officers in Bahia, a province that had, admittedly, a weak abolitionist movement.(65) On 27 March 1883, the officers of Bahia's garrison honored Marshal Hermes Ernesto da Fonseca, the commander of arms, with a formal dinner. They presented him with an oil portrait and solemnized the occasion by presenting "Feliciano, grown gray in the shackles of captivity, father of Lance Corporal Manoel Simoes dos Reis and Private Pedro Manoel Florencio," with his letter of liberty.(66) Freeing an old slave, who had performed notable service to the state by fathering two soldiers and was probably of little value to his owner, a typically selective manumission, cannot be seen as a bold step. Furthermore, in 1884 and 1885, Marshal Hermes himself, despite his earlier involvement in anti-slavery masonic lodges, continued to advise the president on the correct procedure for returning slaves from the army to their owners.(67)

Arguments explaining an alleged early pro-abolition stance in the officer corps on the basis of its middle-class origins rest on the false premise that the nineteenth-century Brazilian "middle class" did not hold slaves. We have already noted the widespread nature of Brazilian slaveholding, and the data on army officers confirm it (Table 5). Although six tenths of the slaves owned by officers belonged to the two men who owned sugar plantations, almost two-thirds of the remaining 74 officers owned slaves. In every decade prior to the 1880s, a majority of these men owned at least some slaves. Qualitative evidence reinforces the data. In 1831, when the Portuguese-born commander of arms was unceremoniously expelled from the province, he embarked with his wife, two daughters, a servant, and five slaves.(68) Slaveholding was not restricted to senior officers. Corporal Lino Pereira Reboucas, who had inherited Manoel Christino from his father, faced the ironic prospect of having his slave, impressed in 1860, join him in the enlisted ranks.(69) The last officer with slaves recorded in his inventory in Salvador died in 1887 before completing the process of manumitting a domestic servant and receiving compensation from the emancipation fund.(70)

Slaveholding reinforced officers' commitment to their obligation to uphold the law by returning fugitive slaves. Contradictions which scholars now perceive between their efforts to "professionalize" or "modernize" their institution while backing slavery rarely bothered most officers. Slaveholding officers stood shoulder to shoulder with sugar planters whose wealth and status depended on their masses of human property and with poor women whose humble respectability or desperate survival rested on the possession of a single slave. The important contradictions lay at the heart of the institutions of the state apparatus charged with defending the social order and the nature of the population subject to recruitment. Sweeping through the racially-mixed free and freed lower classes, the nets of impressment inevitably caught slaves, while the identification of military service with freedom attracted runaways, as did the possibilities of using the army to distance themselves from owners. Once in the ranks, institutional pressures - the army's need for manpower, the state's fiscal concerns, and the legalistic bureaucratic culture of the Brazilian government - tended to hold slaves in the army, despite their formal exclusion from it. To the frustration of their masters, slaves demonstrated a shrewd understanding of these contradictions and turned the shelter of the uniform to their advantage in their unceasing struggles with their masters.

Table 5:

Slaveholding among Army Officers in Salvador, 1800-1888

Period      Number          Number         Number    Average Size
          of Officers   Leaving Slaves   of Slaves    of Holding

1800-29        6              5             303          60.6
              (5)            (4)           ( 84)        (21.0)
1830-39        7              5              25           5.0
1840-49       14             10             100          10.0
1850-59       13              9             343          38.1
             (12)           ( 8)           ( 27)        ( 3.4)
1860-49       17             12             104           8.7
1870-79       14              8              45           5.6
1880-88        5              1               1           1.0

Total         76             50             921          18.4
             (74)           (48)           (386)         (8.0)

Notes: Includes slaves liberated by testamentary provisions and
runaways. Figures in brackets exclude the two largest
slaveholders, sugar planters as well as army officers, with 219
and 316 slaves respectively. Sources: Seventy-five probate
inventories of army officers in the judicial district of Salvador,
Arquivo Publico do Estado da Bahia, Secao Judiciaria, Inventarios
e Testamentos; and inventory of Pedro Labatut, Revista do
Instituto Geografico e Historico da Bahia 68 (1942), 179-203.

Department of History Vancouver, BC, Canada V6T 1Z1

ENDNOTES

Earlier versions of this paper were presented at the Brazilianists Committee Meeting of the Conference on Latin American History, San Francisco, 7 January 1994, and the Race and Slavery in the Americas Working Group of the University of Texas at Austin, 28 October 1994. I thank the participants at these meetings, as well as Peter Beattie, Richard Graham, Aline Helg, and Sandra Lauderdale Graham, who provided helpful comments on earlier drafts of this paper. Joao Jose Reis, Celia Rodrigues, and Walter Fraga assisted my work as benign latter-day slave hunters by calling my attention to additional runaways hidden in the archives. Research materials were drawn from the following archives: Arquivo Historico do Exercito, Requerimentos (AHEx/RQ); Arquivo Nacional, Rio de Janeiro, Secao de Poderes Executivos (ANRJ/SPE); Arquivo Publico do Estado da Bahia (APEBa), Secao de Arquivo Colonial e Provincial (SACP) and Secao Judiciaria, Inventarios e Testamentos (SJ/IT); Arquivo da Sexta Regiao Militar (ASRM); Biblioteca Nacional, Rio de Janeiro, Secao de Manuscritos (BNRJ/SM). Decretos and decisoes are drawn from the Colleccao das leis do Imperio do Brazil (Rio de Janeiro, 1809-1890).

1. Felix Jose da Silva to Lieutenant-Colonel Commanding, Eighth Infantry Battalion, Salvador, 15 January 1864; and Commander of Arms to Vice-President, Salvador, 16 January 1864 (Secret), APEBa/SACP, maco 3409; Commander of Arms to President, Salvador, 5 April 1864; and Relatorio de prevecao, Antonio de Moura, 12 April 1864, ibid., maco 3418.

2. John Henry Schulz, "The Brazilian Army and Politics, 1850-1894," (Ph.D. diss., Princeton University, 1973, chap. 3); Paulo Mercadante, Militares e civis: a etica e o compromisso (Rio de Janeiro, 1977), 106-107; Wilma Peres Costa, "A espada de Damocles: o exercito e a crise do imperio," (Ph.D. diss., Universidade de Sao Paulo, 1990). The principal expositor of the middle-class interpretation of the Brazilian army is Nelson Werneck Sodre, Historia militar do Brasil (Rio de Janeiro, 1965). These arguments have much in common with army propaganda which presents prominent officers as early abolitionists and the institution as a progressive force in Brazilian society, Adalberto Martins da Silva, "O ideal abolicionista nas forcas armadas," in Arno Wehling, ed., Abolicao do cativeiro: os grupos dominantes, pensamento e acao (Rio de Janeiro, 1988), 94-101; Claudio Moreira Bento, "O exercito e a abolicao," in ibid., 83-93; Brazil, Estado Maior do Exercito, Historia do Exercito Brasileiro: perfil militar de um povo, 3 vols. (Brasilia and Rio de Janeiro, 1972), 2:668-673; Aurelio de Lyra Tavares, "O exercito e a abolicao: uma visao retrospectiva," Revista do Exercito Brasileiro 25:2 (April-June 1988), 8. An English-language work imbued with these arguments, although it glosses over abolition, is Robert Ames Hayes, The Armed Nation: The Brazilian Corporate Mystique (Tempe, AZ, 1989).

3. Agostinho Marques Perdigao Malheiro, A escravidao no Brasil: ensaio historico-juridico-social, 3 vols. (Rio de Janeiro, 1866-1867), 1:2-3; Petition of Victoria Maria de Jesus to President, [Caravelas], c. 1852, APEBa/SACP, maco 2885.

4. For examples, see Petitions of Maria Florinda de Sao Jose to Government of Bahia, Salvador, c. 1823, APEBa/SACP, maco 2889; and Felicia Rosa do Amor Divino to Emperor, Rio de Janeiro, 8 June 1847, AHEx/RQ, F-1-8. A full exposition of the legal basis of property rights over slaves and the conditionally freed is supplied in Dezembargador Procurador da Coroa to President, Salvador, 16 September 1865, APEBa/SACP, maco 3432.

5. Sylvia R. Frey, Water from the Rock: Black Resistance in a Revolutionary Age (Princeton, 1991); Benjamin Quarles, The Negro in the American Revolution (Chapel Hill, 1961); Nuria Sales de Bohigas, "Esclavos y reclutas en sudamerica, 1816-1826," in Sobre esclavos, reclutas y mercaderes de quintos (Barcelona, 1974), 85-102; Rebecca J. Scott, Slave Emancipation in Cuba: The Transition to Free Labor, 1860-1899 (Princeton, 1985), chap. 2. For an exhaustive but superficial account of the military roles of slaves and free and freed blacks in the colonial Americas, see Peter M. Voelz, Slave and Soldier: The Military Impact of Blacks in the Colonial Americas (New York, 1993).

6. The Congress of Gran Colombia, for example, undermined Bolivar's decrees, and slavery persisted in Colombia and Venezuela until the mid-1850s, Bohigas, "Esclavos y reclutas," 99-102. In contrast, after 1870, the Cuban insurgents formally proclaimed themselves to be abolitionists; nevertheless, they hedged up restrictions about the freed-men and thus sought to reproduce the hierarchies of slavery, Scott, Slave Emancipation, 48-62.

7. George Reid Andrews, The Afro-Argentines of Buenos Aires, 1800-1900 (Madison, 1980), 116-117.

8. Roger Norman Buckley, Slaves in Red Coats: The British West India Regiments, 1795-1815 (New Haven, 1979). Similarly, the French government purchased slaves to fill the ranks of its West African troops in the early nineteenth century; Myron Echenberg, Colonial Conscripts: The Tirailleurs Senegalais in French West Africa, 1857-1960 (Portsmouth, NH, 1991), chap. 2.

9. Decisao 113 (Imperio), 30 July 1823. On the freeing of slaves to serve in the Paraguayan War, see the polemical account in Julio Jose Chiavenato, O negro no Brasil da senzala a Guerra do Paraguai (Sao Paulo, 1980), 194-207; and the more considered assessments of Ricardo Salles, Guerra do Paraguai: escravidao e cidadania na formacao do exercito (Rio de Janeiro, 1990), 63-77; Jorge Luiz Prata de Souza, "La Guerra del Paraguay en el contexto de la esclavitud brasilena," (M.A. thesis, Universidad Autonoma de Mexico, 1990); Peter M. Beattie, "Transforming Enlisted Army Service in Brazil, 1864-1940: Penal Servitude versus Conscription and Changing Conceptions of Honor, Race, and Nation," (Ph.D. diss., University of Miami, 1994), chap. 2; and Hendrik Kraay, "Soldiers, Officers, and Society: The Army in Bahia, Brazil, 1808-1889," (Ph.D. diss., University of Texas at Austin, 1995), chap. 10.

10. Joao Jose Reis and Eduardo Silva, Negociacao e conflito: a resistencia negra no brasil escravista (Sao Paulo, 1989), 79-98; Dale Thurston Graden, "From Slavery to Freedom in Bahia, Brazil, 1791-1900," (Ph.D. diss., University of Connecticut, 1991), 169-170; and his "Voices from Under: The End of Slavery in Bahia, Brazil," Review of Latin American Studies 3:2 (1990), 150.

11. Manuela Carneiro da Cunha, "Silences of the Law: Customary Law and Positive Law on the Manumission of Slaves in 19th Century Brazil," History and Anthropology 1:2 (1985), 427-443.

12. Decreto 2171, 1 May 1858, Art. 23; Commander of Arms to War Minister, Salvador, 4 July 1842, ANRJ/SPE/IG1, maco 252, fol. 336.

13. Lyle N. McAlister, The "Fuero Militar" in New Spain, 1764-1800 (Gainesville, FL, 1957). On the extent of Brazilian soldiers' legal privileges, see Antonio Manoel da Silveira Sampaio, Instruccoes para o uso dos officiaies do exercito nacional, e imperial nos processos de conselhos de guerra (Rio de Janeiro, 1824), 7-10; and Antonio Jose Amaral, Indicador da legislacao militar em vigor no exercito do imperio do Brasil organizado e dedicado a S.M.I...., 2nd ed., 3 vols. (Rio de Janeiro, 1870-1872), vol. 1, part 1, pp. 275-283.

14. Imperial Pardon of Manoel Luiz Claudino, Rio de Janeiro, 12 December 1863, ANRJ/SPE/IG1, maco 587, fol. 34r; Parecer, Conselho Supremo Militar, Rio de Janeiro, 27 October 1863, ibid, fols. 37r-38r; Parecer, Conselho Supremo Militar e de Justica, Rio de Janeiro, 21 November 1863, ibid, fols. 42r-43r. For references to the other case, see Petition of Ildefonso Moreira Sergio to President, Salvador, 9 June 1865, APEBa/SACP, maco 2886; and Commander of Arms to President, Salvador, 10 June 1865, ibid., maco 3444.

15. Justificacao, Joao Helling, Juizo Municipal, Santo Amaro, 1857, APEBa/SACP, maco 2896; Commander of Arms to President, Salvador, 10 October 1857, ibid., maco 3389; Justificacao, Joao Helling, Juizo Municipal, Santo Amaro, 1858, AHEx/RQ, JJ-94-2587; Adjutant General to Minister of War, Rio de Janeiro, 19 July 1858, ibid. Avoiding responsibility, the War Minister authorized the president of Bahia "to examine the case carefully and ... turn him over to his owner if he is a slave," ibid. In Bahia, the president took these to be orders to return Moura to Helling, as can be inferred from Commander of Arms to Chief of Police, Salvador, 9 August 1858, APEBa/SACP, maco 6457.

16. For example, see Petition of Joaquina Simoes to Emperor, Salvador, 9 September 1873, AHEx/RQ, JZ-5-159.

17. Petition of Antonio de Sampaio de Almeida to Emperor, n.p., c. 1842, with marginal comments, Baron of Caxias, c. October 1842, AHEx/RQ, A-169-4358.

18. Commander of Arms to President, Salvador, 17 May 1839, APEBa/SACP, maco 3374; and 23 June 1863, ibid., maco 3417; Lieutenant-Colonel Commanding to President, Headquarters of Artillery Brigade [Salvador], 9 May 1825, BNRJ/SM, II-33, 31, 4, number 5, doc. 19.

19. Malheiro, Escravidao, 1:179.

20. Francisco de Paula e Vasconcellos to Adjutant General, Rio de Janeiro, 9 October 1825, AHEx/RQ, JJ-237-5790.

21. Decisao 18 (Fazenda), 21 February 1842; Manoel Joaquim do Nascimento e Silva, Synopsis da legislacao brasileira ate 1874 cujo conhecimento mais interessa aos empregados do Ministerio de Guerra, 2 vols. (Rio de Janeiro, 1874), 1:460. Resolucao, 15 May 1872, in Manoel Joaquim do Nascimento e Silva, ed., Consultas do Conselho de Estado sobre negocios relativos ao Ministerio da Guerra ... 1867-1872 (Rio de Janeiro, 1885), 518.

22. Parecer, 3 February 1880, in Manoel Joaquim do Nascimento e Silva, ed., Consultas do Conselho de Estado sobre ngeocios relativos ao Ministerio da Guerra ... 1878-1886 (Rio de Janeiro, 1887), 158-161.

23. Silva, Synopsis, 1:456; Parecer, 11 January 1858, in [Candido Pereira Monteiro!, ed., Consultas do Conselho de Estado relativamente a negocios do Ministerio da Guerra desde o anno de 1843 a 1866 ... (Rio de Janeiro, 1872), 125-126.

24. Petition of Baron of Traripe to President, Salvador, 11 December 1867, APEBa/SACP, maco 1886. The baron probably authored or arranged the publication of the condemnation of the practice of charging for maintenance of slaves that appeared the next day, "Uma injustica," Jornal da Bahia, 12 December 1867, p. 1, col. 1. For a more typical example of prompt payment, see Commander of Arms to President, Salvador, 29 October 1850, APEBa/SACP, maco 3387.

25. Commander of Arms to Adjutant General, Salvador, 9 February 1877 (copy), APEBa/SACP, maco 3436. This bit of creative accounting did not resolve the case and, in 1880, the Council of State insisted that she supply further proofs of her ownership and of Miranda's identity, Parecer, 3 November 1880, in Silva, ed., Consultas ... 1878-1886, 233-235.

26. On recruitment, see Michael C. McBeth, "The Brazilian Recruit during the First Empire: Slave or Soldier?" in Dauril Alden and Warren Dean, eds., Essays Concerning the Socioeconomic History of Brazil and Portuguese India, (Gainesville, FL, 1977), 71-86; Joan E. Meznar, "The Ranks of the Poor: Military Service and Social Differentiation in Northeast Brazil, 1830-1875," Hispanic American Historical Review 72:3 (August 1992), 335-351; Beattie, "Transforming," chaps. 1-4; and Kraay, "Soldiers," chap. 6.

27. On the blending of slave and free populations in late nineteenth-century Brazil, see Sidney Chalhoub, Visoes da liberdade: um historia das ultimas decadas da escravidao na Cone (Sao Paulo, 1990), 212-248; and Luiza Rios Ricci Volpato, Os cativos do sertao: vida cotidiana e escravidao em Cuiaba, 1850/1888 (Sao Paulo, 1993), 198-228.

28. Petitions of Maria Theresa do Sacramento to President, Salvador, 7, 17, and 24 November 1868; Captain Commanding, sixth Company, sixth Infantary Battalion, National Guard, to Captain Acting Commander, Salvador, 12 November 1868, APEBa/SACP, maco 2886.

29. "Perguntas feitas ao crioulo Felippe ...," Salvador, 10 June 1870, APEBa/SACP, maco 6464.

30. Raimundo Jose da Cunha Mattos, Repertorio da legislacao militar actualmente em vigor no exercito e armada do Imperio do Brazil, 3 vols. (Rio de Janeiro, 1834-1842), 1:229; Silva, Synopsis, 1:219. Other military institutions had similar problems with runaway slaves. In the course of my research, I came across several cases of slaves reclaimed from the militia, the navy, and the Bahian police. Luiz R. B. Mott located a case of a slave enlisted in the National Guard; Sergipe del Rey: populacao, economia e sociedade (Aracaju, 1986), 71; and Thomas H. Holloway notes the recurrent problem of runaways in the ranks of Rio de Janeiro's police, Policing Rio de Janeiro: Repression and Resistance in a 19th-Century City (Stanford, 1993), 173-174.

31. "Mappa demonstrativo do numero dos individuos que ... tiverao baixa do servico nesta Provincia desde 1.o de Marco de 1841 ate 19 de Janeiro de 1843," Salvador, 20 February 1843, ANRJ/SPE/IG1, maco 117, fol. 373r.

32. For a table of authorized army strength from 1830 to 1889, see William Sheldon Dudley, "Reform and Radicalism in the Brazilian Army, 1870-1889," (Ph.D. diss., Columbia University, 1972), 244-247.

33. Stuart B. Schwartz, Sugar Plantations in the Formation of Brazilian Society: Bahia, 1550-1835 (Cambridge, 1985), 439-467.

34. Petition of Antonio Pereira dos Santos to President, Salvador, c. 1850, APEBa/SACP, maco 2883.

35. In 1835, Africans represented 63 per cent of Salvador's slave population but, in 1872, they only accounted for 6 per cent of Bahia s slaves; Joao Jose Reis, Slave Rebellion in Brazil: The Muslim Uprising of 1835 in Bahia, trans. Arthur Brakel (Baltimore, 1993), 6; Brazil, Directoria Geral de Estatisticas, Recenseamento da populacao Brazil a que se procedeu no dia 1.o de agosto de 1872, 21 vols. (Rio de Janeiro, 1873-1876), 3:511. Africans comprised the overwhelming majority of Rio de Janeiro's slave population during the first half of the nineteenth century; Mary C. Karasch, Slave Life in Rio de Janeiro, 1808-1850 (Princeton, 1987), 8.

36. Reis, Slave Rebellion, 223-230.

37. Petition of Amaro Jose Correia to President, Salvador, c. 1868, and enclosed atestado, Felisberto Coelho dos Santos, Fort Sao Pedro, 13 January 1868, APEBa/SACP, maco 3491.

38. Justificacao, Marcolino Dias de Andrade, Juizo de Direito, 1.a Vara Civil, Salvador, 1874, APEBa/SACP, maco 2886.

39. In Brazilian historiography, this approach to slavery is exemplified by Reis and Silva, Negociacao e conflito; Silvia Hunold Lara, Campos da violencia: escravos e senbores na Capitania do Rio de Janeiro (Rio de Janeiro, 1988); and Chalhoub, Visoes; and has been forcefully criticized by Jacob Gorender, A escravidao reabilitada (Sao Paulo, 1990).

40. Petition of Jose Manoel de Araujo Goes to President, Salvador, 21 June 1877, APEBa/SACP, maco 2897.

41. Major Acting Commander to Adjutant General, Deposito de Recrutas, Praia Vermelha, [Rio de Janeiro], 24 March 1824, AHEx/RQ JJ-237-5790.

42. Commander of Arms to Chief of Police, Salvador, 7 April 1884, APEBa/SACP, maco 6465. Further examples of such notices can be found in Reis, Slave Rebellion, 144.

43. Commander of Arms to President, Salvador, 20 October 1860, APEBa/SACP, maco 3407.

44. Testimony of Joze Theodoro dos Santos, 15 July 1869, Justificacao, Joaquim Joze Gaiozo Sa Barretto, Juizo dos Feitos da Fazenda, 1869, AHEx/RQ, JJ-148-3816. Curiously, Joze was not imprisoned immediately after his discovery, as was standard army practice.

45. Testimony of Antonio Baptista Pereira Marques, 26 February 1858, Justificacao, Joao Helling, Juizo Municipal, Santo Amaro, 1858, AHEx/RQ, JJ-94-2587.

46. Commander of Arms to Chief of Police, Salvador, 21 January 1854, APEBa/SACP, maco 6461.

47. Lieutenant-Colonel Commanding, Corpo de Guarnicao Fixa, to Commander of Arms, Salvador, 19 June 1858, AHEx/RQ, JJ-94-2587.

48. See testimony in Justificacao, Francisca Alexandrina de Vasconcellos, Juizo dos Feitos da Fazenda, 1876, APEBa/SACP, maco 2897. The presidency accepted Vasconcellos's claim, but it is not clear whether the government bought out her rights or had the slave returned.

49. Patricia Ann Aufderheide, "Order and Violence: Social Deviance and Social Control in Brazil, 1780-1840," (Ph.D. diss., University of Minnesota, 1976), 101. On the importance of patronage in nineteenth-century Brazil, see Richard Graham, Patronage and Politics in Nineteenth-Century Brazil (Stanford, 1990), 20-23.

50. Luiz Gama to Lucio de Mendonca, n.p., 25 July 1880, in Evaristo de Moraes, A campanha abolicionista (1879-1888) (Rio de Janeiro, 1924), 251-256, quotes, 253,256.

51. Commander of Arms to President, Salvador, 20 October 1860, APEBa/SACP, maco 3407.

52. Petition of Sebastiao Jose Lopes to Emperor, [Capim Grosso], c. 1860; Birth Certificate of Arsenio, Se Parish, Salvador, 25 January 1839 (copy); Captain Acting Commander, Seventh Infantry Battalion, to Commander of Arms, Salvador, 14 February 1860; Colonel Commander, Seventh Infantry Battalion, to Commander of Arms, Salvador, 10 March 1860; Adjutant General to Minister of War, Rio de Janeiro, 9 August 1860, AHEx/RQ, S-17-523; [Unsent] Petition of Arcenio Teixeira dos Santos to Emperor, Salvador, 19 June 1863, APEBa/SACP, maco 3405; Commander of Arms to President, Salvador, 18 July 1865, ibid., maco 3438.

53. Schulz, "Brazilian Army," 66-67; Karasch, Slave Life, 338.

54. Testimony of Salvador da Rocha Lima, 21 March 1865, Justificacao, Francisco Joaquim Esteves, Juizo Municipal, Santo Amaro, 1865, APEBa/SACP, maco 3412.

55. Petition of Damilao Antonio do Sacramento to Chief of Police, Salvador, c. 1873, APEBa/SACP, maco 6459; Commander of Arms to President, Salvador, 22 January 1873, ibid., maco 3430.

56. This case can be followed in Inventario, Antonia Teixeira do Sacramento, APEBa/SJ/IT, 03/1158/1627/11; and Inventario, Manoel Jose Teixeira Barbosa, ibid., 07/3023/08, to which the court's final decision is appended.

57. Francisco Borges Ribeiro to Agostinha Maria de Jezus, Humaita, Paraguay, 16 April 1869; Justificacoes, Jose Joaquim de Araujo, Juizo Municipal, Santo Amaro, 1869; and Juizo dos Feitos da Fazenda, 1870, AHEx/RQ, JJ-259-6322.

58. For examples, see Commander of Arms to President, 28 April 1851, APEBa/SACP, maco 3384; and Petition of Jose Manoel de Araujo Goes to Emperor, [Salvador], c. 1859, AHEx/RQ, JZ-108-3241.

59. Commander of Arms to President, Salvador, 12 June 1874, APEBa/SACP, maco 3431. In response to Santos's declaration of freedom, the tutor pulled from his pocket the slave's birth certificate and registration. Observant offers noticed a minor discrepancy between the two documents, one listing Santos as a creole, the other as a mulatto, and held up the claim. Further evidence later satisfied the Commander of Arms and he recommended the return of Santos, Commander of Arms to President, Salvador, 23 July 1874, ibid., maco 3456. See also, Inventario, Leonarda Maria dos Anjos Monteiro, APEBa/SJ/IT, 05/2124/2593/04.

60. Alferes Commanding, First Company, Second Battalion, Voluntarios da Patria, to Commander of Arms, Salvador, 10 February 1865 (copy), APEBa/SACP, maco 3423; Chief of Police to President of Bahia, 14 February 1865, ibid., maco 2969.

61. The partial lifting of this bar during the Civil War thus marked an important change in the United States, although it stands within the long tradition of emergency wartime recruitment of slaves, Leon F. Litwack, Been in the Storm So Long: The Aftermath of Slavery (New York, 1979), 64-103; Ira Berlin, et al., Slaves No More: Three Essays on Emancipation and the Civil War (New York, 1992), 187-233.

62. Daniel Field, The End of Serfdom: Nobility and Bureaucracy in Russia 1855-1861 (Cambridge, MA, 1976), 37. Even so, rumors that military service would earn them their freedom prompted tens of thousands of serfs to flee to the colors during the Crimean War, Peter Kolchin, Unfree Labor: American Slavery and Russian Serfdom (Cambridge, MA, 1987), 284.

63. For an introduction to Islamic military slavery, see Bernard Lewis, Race and Slavery in the Middle East: An Historical Enquiry (New York, 1990), 62-71; and David Ayalon, "Preliminary Remarks on the Mamluk Military Institution in Islam," in V. J. Parry and M. E. Yapp, eds., War, Technology and Society in the Middle East, (London, 1975), 44-58. On its origins, see Daniel Pipes, Slave Soldiers and Islam: The Genesis of a Military System (New Haven, CT, 1981); and Patricia Crone, Slaves on Horses: The Evolution of Islamic Polity (Cambridge, 1980). On the impact of military slavery in Northeast Africa, see Gerard Prunier, "Military Slavery in the Sudan during the Turkiyya, 1820-1885," Slavery and Abolition 13:1 (April 1992), 129-139; and Douglas H. Johnson, "The Structure of a Legacy: Military Slavery in Northeast Africa," Ethnohistory 36:1 (Winter 1989), 72-88.

64. Robert Conrad, The Destruction of Brazilian Slavery, 1850-1888 (Berkeley, 1972), 251-252; Emilia Viotti da Costa, Da senzala a colonia (Sao Paulo, 1966), 446; Rebecca Baird Bergstresser, "The Movement for the Abolition of Slavery in Rio de Janeiro, 1880-1889," (Ph.D. diss., Stanford University, 1973), 61-71.

65. Luiz Anselmo da Fonseca, a Bahian abolitionist, complained in 1887 that his native province was "ultra-slavocrat," A escravidao, o clero e o abolicionismo, facsimile ed. (Recife, 1988), 134.

66. Commander of Arms to Delegado do Cirurgiao-Mor do Exercito, Salvador, 2 April 1883, Ordens recebidas, DCMEx, fol. 96r, ASRM. Hermes Ernesto da Fonseca should not be confused with his son, the future president of Brazil, Hermes Rodrigues da Fonseca.

67. Silva, "Ideal abolicionista," 96-97; Commander of Arms to President, Salvador, 8 November 1884, APEBa/SACP, maco 3443; and 8 June 1885, ibid., maco 3447.

68. "Rellacao das pracas, pecoas de Familias pertencentes as mesmas, Criados, Camaradas, e Escravos embarcados ...," 7 April 1831, BNRJ/SM, 1-31, 15, 19.

69. Petition of Lino Peteira Reboucas to President of Bahia, 25 July 1860, APEBa/SACP, maco 3424. Because Manoel Christino had not yet enlisted, he was promptly released, Commander of Arms to President of Bahia, Salvador, 31 July 1860, ibid.

70. Inventario, Francisco Antonio de Souza, APEBa/SJ/IT, 07/2915/01, fol. 20r.