Empire of Brazil

It matters for three reasons. First, it was the only monarchy established by the independence movements of Latin America that lasted: Mexico's two short imperial experiments collapsed within years, and Brazil's lasted 67. Second, it kept Portuguese America whole at a moment when Spanish America fragmented into a dozen republics, so that the map of South America was decided less by geography than by the survival of a single government in Rio. Third, it was the last state in the Western Hemisphere to abolish slavery, in 1888, and the way it did so — gradual legislation, then a sudden collapse under pressure from below — shaped Brazilian society long after the monarchy itself was gone.

The empire was also, for most of its existence, a working parliamentary system in a continent of caudillos and constitutions on paper. Cabinets rose and fell, elections were held (narrowly), a press operated with surprising freedom, and the emperor governed through parties rather than against them. That combination of monarchy and parliament, of slavery and citizenship, is what makes the period worth studying rather than merely narrating.

At a glance

How did a Portuguese colony become a monarchy?

The empire began as an accident of European war. In November 1807, with a French army under Junot about to enter Lisbon, the Prince Regent João ordered the Portuguese court — the royal family, the treasury, the archives, the law courts, several thousand nobles and officials — aboard Portuguese ships escorted by the British navy and carried across the Atlantic to Brazil. It was a decision taken under British pressure and British protection, formalised in a secret convention of 22 October 1807 in which Britain renewed its guarantee of the Braganza dynasty and of Portuguese territorial integrity, now explicitly including the court in Brazil.

The fleet arrived in January 1808, first at Salvador da Bahia and then at Rio de Janeiro. Before disembarking at Salvador, João issued the decree opening Brazilian ports to the trade of all friendly nations, ending a three-century Portuguese monopoly in a stroke. The consequences were immediate. The number of ships entering Rio de Janeiro in 1808 was more than four times the figure for 1807, and most of them were British. In February 1810 two treaties with Britain followed: an alliance, and a navigation and commerce treaty giving British goods a maximum tariff of 15 per cent, the right of British residents to own property, their own churches and cemeteries, and their own special magistrates to hear cases involving British subjects. Brazil, in other words, was opened as an informal British commercial dependency at the very moment it ceased to be a closed Portuguese colony.

Brazil was formally elevated on 16 December 1815, when the colony became a kingdom and the Portuguese monarchy turned itself into the United Kingdom of Portugal, Brazil and the Algarves. Rio de Janeiro acquired the institutions of a capital: a bank, a printing press (which published the first Brazilian newspaper, the Gazeta do Rio de Janeiro), a military academy, a museum, a royal library. The court brought perhaps 15,000 people into a city of some 60,000, and the arrival permanently shifted the centre of gravity of the Portuguese world from the Tagus to Guanabara Bay.

The reversal came from Portugal itself. In August 1820 a liberal revolution in Oporto demanded a constitution and the return of the king, and the Cortes elected in Lisbon made clear that they intended to reduce Brazil to its colonial condition again — to abolish the Rio institutions, to end the special status of the kingdom, and to station troops in Rio. João returned to Lisbon in April 1821, leaving his son Pedro as prince regent with instructions that amounted to: hold the country together, and if the constitution cannot be avoided, swear to it. Within months the Cortes had ordered Pedro home too, stripped Brazil's provinces of direct representation, and begun issuing decrees that Rio read as a programme of recolonisation. The Portuguese garrison in Rio grew restive. On 9 January 1822 Pedro refused to sail, announcing that he would stay — the Dia do Fico, the day of "I remain" — and the break became a question of timing rather than principle.

How was independence actually achieved?

Independence was declared twice, and fought for more than a year. On 7 September 1822, riding from Santos back to São Paulo after settling a quarrel between local authorities, Pedro received at the Ipiranga stream a batch of despatches from Rio reporting further Lisbon decrees and demanding his return. He tore off his blue-and-white Portuguese cockade, drew his sword and shouted that Brazil was henceforth independent — the Grito do Ipiranga, the cry of Ipiranga. The scene as every Brazilian schoolchild knows it, with a white horse rearing, a scarlet uniform and a guard of dragoons, is a painting made 66 years later by Pedro Américo for the São Paulo provincial government; the reality was a small escort, a man in travelling clothes, an upset stomach and a mule.

The declaration was not a victory. The war of independence was fought in the north and north-east, where Portuguese garrisons held Belém, São Luís, Salvador and Montevideo, and it was decided at sea and in sieges. The imperial government hired a British admiral, Thomas Cochrane, to command a scratch navy, and hired British and French officers to man it — a mercenary fleet that was paid partly in prize money. Salvador held out until 2 July 1823, when the Portuguese commander made a deal to evacuate and the Brazilian army entered the city; 2 July is still a civic holiday in Bahia, and it is a better answer to "when did Brazil become independent" than 7 September. Belém fell in August 1823, Montevideo in March 1824, and the last Portuguese resistance ended with the surrender of the garrison of the Cisplatine province.

The political consolidation was equally contested. Pedro was acclaimed emperor in Rio de Janeiro on 12 October 1822 — his 24th birthday — and crowned on 1 December. But the provinces had to be persuaded, and some had to be coerced: the republican experiment in Pernambuco in 1824, the Confederation of the Equator, was suppressed by imperial troops after six months of fighting. Recognition abroad came at a price. Britain, which had mediated, made clear that recognition would not come without an end to the slave trade; Portugal recognised its former colony only under the treaty signed in Rio on 29 August 1825, by which Brazil agreed to pay compensation of £2 million — money raised largely on the London market, where Brazilian loans of £1.7 million in August 1824 and £2 million in January 1825 had already been floated. Treaties with Britain followed: the drafts Stuart signed in October 1825 were rejected in London as too generous, and the binding instruments of 1826 committed Brazil to treat the traffic in Africans as illegal from March 1830 — an obligation the country then spent 20 years evading.

So independence was not a single battle, and it was not a national uprising. It was a secession of the Portuguese court's own apparatus in America, led by the heir of the Portuguese crown, supported by the Brazilian landowning classes and the Rio merchant community, financed in London, fought with hired ships and completed by the expulsion of garrisons. That origin — a monarchy created by a prince against a parliament in Lisbon — explains much of what followed.

What did the 1824 constitution create?

The constitution of 1824 was written in reaction to a failure. Pedro had convened a Constituent Assembly in 1823, and it had produced a draft that limited his authority, subordinated the executive to the legislature and, in the eyes of the court, threatened to make Brazil a republic in all but name. In November 1823 troops surrounded the assembly building and arrested several deputies; the body was dissolved. A committee of the Council of State then drafted a text which the emperor granted — outorgou — in March 1824. It was never submitted to any elected body for approval, a fact its defenders and critics have argued about ever since.

The constitution of 1824 created four political powers, not three. Alongside the legislative, executive and judicial stood the Moderating Power, defined in article 98 as "the key to the whole political organisation" and delegated exclusively to the emperor as "Supreme Chief of the Nation and its First Representative", so that he might "ceaselessly watch over the maintenance of independence, equilibrium and harmony of the other political powers". The emperor's person was declared inviolable and sacred, and he was answerable to no one. Through the Moderating Power he appointed senators from triple lists, summoned and prorogued the General Assembly, dissolved the Chamber of Deputies "when the salvation of the State demands it", appointed and dismissed ministers at will, suspended magistrates, granted pardons and amnesties, and sanctioned or vetoed legislation. The executive power was also his, exercised through ministers.

The parliament was not a decoration. The Chamber of Deputies was elected for four years; the Senate was appointed for life from lists of the three highest vote-winners in each province. Deputies had the exclusive right to initiate taxation and recruitment, and ministers were answerable to the Chamber for their conduct. The budget had to be voted annually. The constitution guaranteed freedom of the press, the inviolability of the home, protection against arbitrary arrest and, in a clause that mattered enormously in practice, that no one could be persecuted for religious belief provided he respected the state religion.

But the political nation was tiny, and it was defined by property, gender and, in a way that is easy to miss, by freedom. To vote in the parish assemblies a man needed a net annual income of 100 mil réis; to be an elector, 200 mil réis; to sit as a deputy, 400 mil réis. Freedmen were expressly excluded from the electoral colleges. Non-Catholics could not be deputies. Women, the enslaved, the illiterate poor and the indentured had no vote at all. Catholicism was the religion of the empire, and other faiths were permitted only as domestic or private worship, in houses without the external form of a temple — which is why Protestant foreigners in Rio worshipped in buildings that looked like warehouses.

The result was a system that combined genuine parliamentary competition with a very narrow electorate and an umpire with a reserve power of dissolution. Brazilian politics for the next six decades consisted of two parties — the Conservatives, nicknamed saquaremas, and the Liberals, nicknamed luzias — alternating in office, competing for the favour of the Crown, and agreeing, whatever they said in public, on the essentials: property, slavery, order and the exclusion of the poor from politics. That agreement is what historians mean when they speak of the empire's "conciliation" or of a "party of order".

Why did Pedro I abdicate?

Pedro I's reign lasted nine years and ended in humiliation. The first crisis was constitutional: he had granted a charter, dissolved an assembly and then fought a war against a republic in the north-east within 18 months of his coronation. The second was territorial. In April 1825 the Thirty-Three Orientals, supported from Buenos Aires, raised the Banda Oriental against Brazilian rule, and the United Provinces of the Río de la Plata annexed it. Brazil declared war. The Cisplatine War lasted nearly three years, was fought on land in the south and at sea in the Plata estuary, and was ruinously expensive: the imperial treasury borrowed, the navy hired, the army was expanded, and the province was never recovered. British mediation produced the treaty signed in Rio de Janeiro in August 1828 creating the independent republic of Uruguay as a buffer between Brazil and Argentina. Brazilians read it as a defeat. The Rev. Robert Walsh, an Irish chaplain in Rio, recorded that the treaty was received with "universal disappointment and discontent".

The third crisis was financial and political. The war, the court's expenses and the collapse of the Banco do Brasil in 1829 produced inflation and a depreciating currency. Pedro's foreign policy — a disastrous expedition to the Río de la Plata, a treaty with Britain that bound Brazil to abolish the slave trade it depended on — made him look, to Brazilian landowners, like a Portuguese prince spending Brazilian money on Portuguese quarrels. His retention of Portuguese advisers, and the death of his empress in 1826 followed by a scandalous affair with his mistress, the marquise of Santos, damaged him further. In 1826 he had also become king of Portugal on the death of João VI, and though he abdicated the Portuguese throne in favour of his daughter Maria da Glória, the Portuguese succession dispute dragged Brazil into Iberian politics.

The end came in April 1831, after months of riots in Rio between Portuguese-born and Brazilian factions, the Noite das Garrafadas ("night of the broken bottles"), and the collapse of a ministry that could not hold the Chamber or the street. On 7 April Pedro abdicated in favour of his five-year-old son, Pedro II, handed over a letter asking Brazilians to love the boy, and sailed for Europe aboard a British warship. He went to Portugal, fought his brother Miguel for his daughter's throne, and died of tuberculosis at Queluz in September 1834, aged 35.

The abdication fixed the pattern of the regime's first decade: a monarchy whose legitimacy was personal, whose finances were fragile, and whose political class had learned that it could force an emperor out. It also created the problem that dominated the 1830s — how to govern in the name of a child.

What was the Regency, and why did it nearly break the country?

Between 1831 and 1840 Brazil was governed by regents. The constitution had provided for a regency in the event of a minority, but it had assumed that a member of the imperial family would take the post. Pedro II had no adult relatives in Brazil willing and able to serve, so the country improvised: first a provisional triumvirate of deputies, then a permanent triennial regency of three elected regents, then, after the constitutional reform of 1834, a single elected regent with a four-year term.

The Regency was the empire's laboratory of experiment and its moment of maximum danger. In 1834 the Ato Adicional, passed as a constitutional amendment, created elected provincial legislative assemblies with the power to legislate on provincial taxes, public instruction, civil and judicial division and municipal police, suppressed the Council of State, and replaced the triennial regency with a single elected regent. The intention was to hold the empire together by giving the provinces a stake in it. The effect was to legitimise provincial politics at exactly the moment when the central government was weakest, and to provide the institutional scaffolding for the regional revolts that followed.

The revolts came in a cluster, and each had its own social base. In Pará, the Cabanagem (1835–1840) began as a quarrel among provincial notables and became a rising of the riverine poor — cabanos, the inhabitants of the thatched huts on the water — who took Belém in January 1835, killed the province's president and the military commander, and held the city for months. The repression that followed was a war on the river, and the death toll is usually estimated at around 30,000 out of a provincial population of perhaps 150,000, making it proportionally one of the bloodiest upheavals in the history of the Americas. In Salvador, in January 1835, the Malê rising brought together African-born slaves and freedmen, many of them Hausa and Nagô Muslims, in a carefully planned night attack on the city's barracks and churches; it was betrayed, fought out in the streets within hours, and crushed, and the trials that followed produced executions, floggings and deportations. It was the only slave revolt in Brazilian history to reach the scale of an urban insurrection, and it fixed the example of Haiti in the mind of the Bahian elite for a generation.

In Rio Grande do Sul, the Farroupilha (1835–1845) was a separatist war of the southern ranching elite against Rio's taxation and appointment of provincial presidents. The rebels proclaimed the Riograndense Republic in 1836, held the interior for nine years, and were defeated as much by exhaustion and internal division as by imperial arms; the peace of 1845 granted an amnesty and a protective tariff on the salted beef that was the province's export. In Bahia again, the Sabinada (1837–1838) was an urban, middle-class and professional revolt that declared a Bahian republic to last until Pedro II came of age, and was crushed after four months of street fighting in Salvador. In Maranhão, the Balaiada (1838–1841) began among cattle drovers and cotton workers, was joined by escaped slaves under a leader called Cosme, and required a military campaign by the Marquis of Caxias, the empire's most effective soldier, to end.

The Regency's response to all this was to build up the instruments of central power that the constitution had left weak: the National Guard, created in 1831 as a militia of propertied citizens and used as a rural police; the Criminal Procedure Code of 1832, which gave local magistrates and elected justices of the peace wide powers; and the practice of appointing provincial presidents from Rio. The regresso — the conservative reaction after 1837 — reversed part of the Ato Adicional through the Law of Interpretation of 1840, restoring central control over the provinces' police and officials.

The Regency ended by a parliamentary manoeuvre. In July 1840 the Liberals, out of office and fearing a Conservative consolidation, pushed through the Chamber a declaration that Pedro II had reached his majority at 14. The "majority coup" gave the country an emperor, a court and a focus of loyalty, and the boy — tall, bookish, myopic, and already trained to hours of study — was crowned in July 1841. He would reign for 49 years.

How did the Second Reign hold together?

The Second Reign's answer to the Regency was order, and order was built out of three things: patronage, a national guard, and a party machine. The Conservatives, in office for most of the 1840s, constructed what the historian Jeffrey Needell calls the party of order — an alliance of coffee and sugar planters, magistrates, merchants and provincial bosses, held together by appointments, honours, contracts and the disciplined use of the ministries' power to make and unmake elections. The Liberals did the same when they had their turn. Elections in imperial Brazil were administered by the government of the day, which is why the result was usually known in advance and why the fraud was so institutionalised that Brazilians coined the phrase eleições a bico de pena — elections decided by the stroke of a pen.

The system was corrupt and it worked. It gave the empire something the Spanish American republics mostly lacked: a political class that could lose office without losing its property, and could therefore accept losing office. Cabinets changed, policies shifted at the margin, and the underlying structure — slavery, export agriculture, a narrow franchise, a centralised monarchy — went untouched. Graham's study of patronage showed how thoroughly the state was woven into private life: appointments, pensions, exemptions, military commissions and honours were the currency of politics, and the deputy who could deliver a post for a constituent was the deputy who was re-elected.

There was one serious attempt to break the system from the left. In Pernambuco, in 1848, the Praieira revolt combined Liberal factions, urban radicals and some abolitionist and republican ideas, and was put down by force in 1850. It was the last armed provincial challenge to the empire: after 1850, and for the rest of the reign, no regional movement seriously threatened the state until the army itself moved in 1889. What replaced the revolts was a system of provincial government through appointed presidents, electoral management and the distribution of offices — less violent, and rather more effective.

The 1850s were the empire's high noon. The slave trade was abolished in 1850, freeing capital and credit that had been sunk in the traffic. The Conciliation cabinet of 1853–1857 brought Conservatives and Liberals into the same ministry and passed a programme of administrative reform, banking legislation and railway concessions. Mauá built the first Brazilian railway, from Rio towards Petrópolis, in 1854, and Rio de Janeiro acquired gas lighting, steamships, a daily press and a theatre. The empire defeated Rosas's Argentine Confederation in 1852 alongside Argentine exiles and Uruguayan Colorados, and emerged as the dominant power in the Río de la Plata. The census of 1872 would show a country of nearly ten million people, tied to the coast by steamship and to Europe by coffee.

What kind of economy did the empire run on?

The empire was a slave economy with a single export staple, and both facts shaped everything else. Coffee accounted for something between 60 and 70 per cent of exports by the 1870s, having overtaken sugar in the 1830s and 1840s as the Paraíba valley and then the interior of São Paulo were planted. Sugar survived in the north-east, cotton had a boom during the American Civil War and collapsed after it, tobacco, hides, cacao, mate and, from the 1880s, Amazonian rubber completed the list. Brazil was the world's largest coffee producer by mid-century and remained so; the coffee frontier moved westwards across São Paulo in a sequence of booms and exhausted soils, dragging railways, immigrants and, eventually, free labour behind it.

The internal economy mattered more than the old plantation model allowed. Recent work on Bahia has shown that the great sugar estates bought their cassava flour, their cattle and their timber from small farmers, tenants, freedmen and free peasants, and that this internal market was a condition of export agriculture rather than a residue of it. Brazilian historians of the 1960s and 1970s wrote of a "plantationist" past in which everything was latifundium and monoculture; the evidence from inventories, parish records and municipal accounts shows a more varied countryside, with a large free population of poor farmers, artisans, drovers, fishermen and itinerant traders, many of them black or mixed-race, living alongside the slave plantations.

Slavery was the labour system of the export sector and, in the towns, of much else besides. Rio de Janeiro in mid-century was a city in which 40 to 50 per cent of the inhabitants were slaves, and where the streets were full of enslaved men and women hired out as porters, water carriers, seamstresses, cooks, boatmen and prostitutes, paying their owners a weekly wage and keeping the difference. That practice, the escravidão de ganho, gave urban slaves a measure of autonomy and a route to purchase their own freedom, and it created a class of freed people — libertos — who were legally free and socially marked. By the 1870s the internal slave trade had moved perhaps 5,000 to 10,000 people a year from the north-east and Minas Gerais to the coffee zones, breaking families and emptying whole provinces of their enslaved population.

The state was small, even by the standards of the time. Public spending was low, direct taxation fell mostly on imports and on the slave trade, and the empire borrowed steadily in London: £1.7 million in 1824, £2 million in 1825, £3.8 million in 1863, £7 million in 1865 to pay for the Paraguayan War, and more in 1871, 1875, 1883, 1886 and 1888. At the end of the empire 98 per cent of the public debt, £33.6 million, had originated in London loans. Railways, gasworks, docks, tramways and banks were largely British-financed and British-built: the San Paulo Railway, "A Inglesa", climbed the escarpment from Santos to São Paulo in 1867. Britain took 28 to 37 per cent of Brazilian exports during the empire, but much of that was re-exported to the continent and the United States, because, as Bethell observed, the British simply did not drink enough coffee.

How did slavery end?

Slavery ended in Brazil in stages that took 38 years, and every stage was a compromise that slaveholders believed they could control. The first was the abolition of the traffic. Under the treaty of 1826 the trade became illegal in March 1830, and a Brazilian law of 7 November 1831 declared all Africans imported thereafter free. Neither was enforced. Something like 750,000 Africans were landed in Brazil during the 1830s and 1840s, most of them on beaches between Bahia and Rio, in a traffic so open that the British parliament passed the Aberdeen Act in 1845 authorising the Royal Navy to treat Brazilian slavers as pirates. In 1850, after British warships entered Brazilian ports and waters in pursuit of slavers, the Brazilian government passed the Eusébio de Queirós Law and, for the first time, enforced it. The trade ended within a year and a half.

That changed the arithmetic of slavery. The enslaved population, which had been between two and two and a half million in 1850, began to fall: 1.7 million by the estimate of Perdigão Malheiro in 1866, about 1.5 million at the census of 1872 — 15 per cent of the population — 1,240,806 in 1884 and 723,419 at the registration of 1887. The fall was produced by high mortality, a skewed sex ratio inherited from the traffic, low fertility and manumission, and it was accelerated after 1850 by an internal trade that moved the surviving enslaved population south into the coffee zone.

The second stage was the Law of the Free Womb of 28 September 1871 — the Lei do Ventre Livre, or Rio Branco Law. Its first article was short: "The children of slave women born in the Empire from the date of this law shall be considered of free condition." The rest of the law was a set of arrangements that kept those children at work. Until the age of eight they stayed with their mother's owner, who was obliged to feed and raise them; at eight the owner could choose between receiving 600 mil réis in government bonds from the state, or using the child's labour until the age of 21. In practice almost every owner chose the labour. The law also created an emancipation fund, financed by a slave tax, lotteries and budget allocations, to buy freedom; it allowed slaves to hold property and to purchase their own freedom, with the price fixed by arbitration if there was no agreement; and it required a national register of slaves. In its first decade the fund freed fewer than one per cent of the enslaved. No slave was emancipated by the law, and Joaquim Nabuco later called it "imperfect, incomplete, unjust, even absurd" — while conceding that no one could ever again be born a slave in Brazil. It passed the Chamber 61 votes to 35 and the Senate with four dissentients, and was signed by the Princess Isabel, acting as regent while her father was in Europe.

The third stage was the Saraiva-Cotegipe Law of 28 September 1885, the Sexagenarian Law, passed after a decade of parliamentary failure and public agitation. It freed slaves over 65 without compensation, but required those between 60 and 65 to work three more years, and it set the compensation values for manumission so high that self-purchase became harder. Fewer than 20,000 people were freed under it in its first year, and The Times calculated that at that rate slavery could legally last until 1935.

The fourth stage was collapse. From 1886 the abolitionist movement moved out of the chambers and into the streets. Provincial emancipation — Ceará in March 1884, Amazonas in July, Rio Grande do Sul in September — had shown that slavery could be ended locally; the army's junior officers, many of them positivists or republicans, began to refuse to hunt fugitive slaves; and in 1887 and 1888 enslaved people left the coffee plantations of São Paulo in tens of thousands, with the connivance of abolitionists who organised trains, shelters and "underground railway" routes to the cities. The São Paulo planter class, having finally secured a mass alternative in Italian immigration — 32,000 immigrants arrived in the province in 1887 alone — abandoned slavery rather than lose the labour force entirely. On 13 May 1888 the Princess Isabel signed a law of two articles: "Slavery in Brazil is declared extinct" and "provisions to the contrary are revoked." There was no compensation. The Chamber had passed it 83 to 9, the Senate 46 to 6, and the remaining enslaved population — somewhere between 250,000 and 500,000 people — was free.

What did the Paraguayan War change?

The war of 1864 to 1870 was the empire's largest and last foreign war, and it did more than any other event to undermine the regime that won it. It began in Uruguay. Brazil had maintained a shadow protectorate in Uruguay since 1852, with Brazilian ranchers owning perhaps 30 per cent of the land and forming 10 to 15 per cent of the population, and when the Blanco government of Bernardo Berro began restricting Brazilian property and cross-border trade, Brazil backed a Colorado rebellion under Venancio Flores. Paraguay, which had its own claims and its own fear of Brazil and Argentina, came in on the Blanco side, seized a Brazilian ship in November 1864 and invaded Mato Grosso in December. Brazil, Argentina and Uruguay signed the Triple Alliance treaty in May 1865, and the war that followed was fought up the rivers of Paraguay for five years.

The scale was extraordinary. Brazil expanded its standing army from about 18,000 men to 60,000 or 70,000 within a year, by transfers from the National Guard, by the volunteer corps known as voluntários da pátria (some more voluntary than others) and by forced recruitment, including of freedmen and of Africans illegally imported and held in semi-slavery. In all, some 140,000 Brazilians were mobilised. The navy destroyed the Paraguayan fleet at Riachuelo in June 1865 and blockaded the rivers; the allies took Humaitá in August 1868 after a siege of more than a year; Asunción fell on 1 January 1869; and Francisco Solano López was killed at Cerro Corá on 1 March 1870. The war cost, on Bethell's reckoning, some 200,000 lives, mostly Paraguayan and Brazilian, from combat, cholera, typhus, dysentery and starvation — the bloodiest interstate war in Latin American history and, apart from the Crimean War, the bloodiest interstate conflict anywhere in the world between 1815 and 1914.

For Brazil the consequences were political before they were territorial. The war was financed by loans and by paper money, and it tripled the public debt and inflated the currency. It gave the army a sense of its own importance and a grievance: officers had fought and bled while the politicians quarrelled, and after 1870 they began to see themselves as a corporation with interests distinct from the throne. It exposed the empire's dependence on slave labour at the precise moment when the United States abolished slavery, making Brazil the only independent slave state in the hemisphere and the target of an international abolitionist campaign. And it made the emperor, who had visited the front once, in 1865, appear as a distant figure in a war fought in his name. The war did not cause the fall of the monarchy, but it changed the terms on which the monarchy was judged.

Why did the empire fall?

Three long-run tensions and one short-run accident brought the empire down. The first was the alienation of the Church. In 1872 two bishops, of Olinda and of Pará, ordered the suspension of priests who were Freemasons; the government, whose own ministers and emperor were Masons, had them tried and, in 1874, imprisoned. The "religious question" ended in an amnesty in 1875, but it taught the ultramontane wing of the Brazilian clergy that the empire was not the Church's friend, and it cost the monarchy the unconditional loyalty of the one institution that reached every parish.

The second was the alienation of the army. After 1870 a generation of officers, influenced by positivism and by the Prussian model, resented the small size of the army, its poor pay, its subordinate place beside the National Guard, and the habit of governments of using troops as police. The "military question" of the 1880s, in which officers publicly defended colleagues against ministerial discipline and were backed by the press and the street, established a principle that no monarchy can survive: that the army answered to its own honour rather than to the government.

The third was the alienation of the slaveholding class, and here the story is more subtle than the simple claim that the planters avenged abolition. The coffee barons of São Paulo had been converting to immigrant labour since the mid-1880s and had less and less to lose from abolition; the die-hard slavocrats of the Paraíba valley had everything to lose and never forgave the Crown. The Republican Party of São Paulo, founded in 1873 at a convention in Itu, drew on both: planters who resented the centralisation of the empire, its taxes and its appointments, and who saw a federation as a way to keep more of the coffee revenue. José Murilo de Carvalho, following the contemporary critic Eduardo Prado, argued that the republic was in part the revenge of a landowning class that had been dispossessed of its human property by an emperor it had never controlled.

The accident was the succession. Pedro II had no male heir. His daughter Isabel was unpopular with the political class, and her husband, the French count d'Eu, was widely disliked as a foreigner; the prospect of a "third reign" under Isabel and d'Eu concentrated minds in 1889, when the emperor was 63, diabetic and visibly failing. On 15 November 1889 a column of troops under Marshal Deodoro da Fonseca — a monarchist general who had been persuaded that he was to be arrested — occupied the Campo de Santana in Rio, the cabinet resigned, and that evening the republic was proclaimed from the balcony of the city hall by a handful of civilians and officers. There was no fighting. The navy did not move. The National Guard did not turn out. Pedro II, offered resistance by his aides, refused to allow it: he would not shed Brazilian blood for his throne. On 17 November the imperial family sailed for Europe, and the emperor died in a modest hotel in Paris two years later.

What did the empire leave behind?

The empire's most important legacy was the map. Brazil is the only state in the Americas that kept the territorial shape of a European colonial unit, and it kept it because the monarchy provided a single focus of loyalty through the fragmentation of the 1820s and the revolts of the 1830s and 1840s. Had the Farroupilha succeeded, or the Cabanagem produced a durable separatist state in the Amazon, South America would look different today. The empire's diplomats spent the 1850s and 1860s fixing frontiers on the principle of uti possidetis, and the boundary disputes with Argentina, Bolivia, Peru and Paraguay that were settled in the early 20th century all began from positions staked out under the monarchy.

Its institutional legacy was more ambiguous. The empire created a professional bureaucracy, a diplomatic service, a supreme court, a national library and archive, a geographical institute, and a political culture of constitutional government that survived the republic. It also created the machinery of patronage, the practice of government by the ministry of the day, and a rural oligarchy whose power was reorganised rather than broken by the republic. The 1891 constitution, drafted by Rui Barbosa on the American model, gave the states the autonomy the empire had denied them, and the result was the politics of the coronel, the local boss, which lasted until 1930.

Its social legacy was the most consequential. Abolition freed several hundred thousand people without land, without schools and without compensation, into a society in which the vote was restricted to the literate and the police were controlled by their former owners. Nabuco, who had done more than any other politician to achieve abolition, wrote that the abolitionist movement "stopped on the day abolition was decreed and retreated the day after". The empire's ideologues had long claimed that Brazilian slavery was milder than its northern counterpart and that racial mixing had dissolved prejudice; the census of 1872, in which 42.7 per cent of the free population were black or mixed-race, was used as evidence. Twentieth-century scholarship, and above all the work of Sidney Chalhoub, has shown how much that story was built on silence — on the refusal to discuss race as a political question, and on the law's habit of granting freedom in principle while withholding it in practice.

Historians still argue about the regime itself. Carvalho's studies of state-building treat the empire as the crucible of the Brazilian state and of a citizenship that was always partial. Needell's work on the Conservative party treats it as a machine for holding together a slaveholding elite in a hostile world. Barman's biography of Pedro II argues that the emperor was neither the ornamental sage of monarchist legend nor the "Pedro Banana" of republican satire, but a working politician of extraordinary endurance who governed by controlling the agenda rather than by decreeing it. Bethell's work on Britain and the slave trade treats abolition as the product of external pressure and internal calculation. What none of them doubts is the scale of the thing: a monarchy that outlasted every republic around it, that ended slavery later than anywhere else in the hemisphere, and that fell, when it fell, in an afternoon.

Common myths about the Empire of Brazil

Timeline of the Empire of Brazil

What the map shows next

At 1831 on the map the empire is at its weakest: a child on the throne, a regency without precedent, and within four years revolts in Pará, Bahia and Rio Grande do Sul. The borders drawn for that year already show the shape the empire would keep — the Amazon basin, the north-east, the centre-south and the southern plains — but not the authority to govern all of it.

At 1871 on the map the map shows the empire after the Paraguayan War, with the Free Womb law just signed, the coffee frontier pushing west across São Paulo and the slave population falling. By 1889 on the map, its last year, the empire covers almost exactly the territory of modern Brazil, and the Brazilian Republic is about to inherit that map intact.