After eighty-three years, France signs a treaty with the maroons

Girard treats marronage as a war the colonial state lost. A first expedition went into the Bahoruco region near the Spanish border in 1702 to break up a community of runaways and failed. For decades afterwards the Bahoruco bands raided neighbouring plantations and French columns kept marching into the mountains after them. Eighty three years later, in 1785, the French and their Spanish neighbours gave up and signed a treaty granting the maroons independence in return for an end to the raids. He describes the mountain strongholds with their moats and stake fences, the sympathetic plantation workers who fed runaways and misdirected search parties, and concludes that Haiti's rebels had worked out the principles of modern guerilla warfare well before the twentieth century, operating from a border zone, using rough ground, relying on local support and refusing pitched battle.

In 1757 an enslaved man named Medor went before the court at Cap Francais in Saint Domingue and confessed without being compelled. He said that he and many others used magical powders to persuade owners to grant them their freedom. His statement opened the enormous investigation that became the Makandal affair. Trevor Burnard and John Garrigus have since tracked the timing and geography of the deaths blamed on poison that year and argue that mycotoxins in spoiled flour, during a wartime blockade, are the likelier cause. On that reading Medor and the others who confessed to using poison and spells to tame their owners may genuinely have believed their taming had killed by accident. The same pattern turns up far away. In Suriname, Clarinda and Nero insisted that the mixture Nero made from bush herbs was meant to calm Clarinda's mistress, and Clarinda was alarmed when the woman fell ill and died.

Over the summer of 1742, after torture, an enslaved man named Baron confessed to causing his mistress's death. He said he had bought herbs from Abraham on another plantation, meaning to cool her temper and avoid future punishment rather than to kill. He also named Coffij, who belonged to the Society of Suriname itself. Coffij denied it and gave the court Askaan and Emanuiel for poisoning other enslaved people, which produced their trials, and Emanuiel's trial produced Ajuba's. Baron additionally accused Louis, a driver on his own plantation, of going about with poison, and Louis's case produced Sambo's. Suriname had sixteen poison cases in the whole of the 1730s and forty two between 1740 and the end of 1745, and eighteen of the forty two were tied to another case. Baron, Sambo, Abraham and Emanuiel were branded, pinched with hot tongs until they died, and their heads set on spikes in the Suriname river.

On 23 January 1745 a court in Orange County, Virginia convicted a woman named Eve of mixing poison into the milk of her owner, Peter Montague, who had languished four months before dying. The record calls the act feloniously and traitorously done, and she was burned at the stake. Virginia treated the killing of a master by a slave as a species of treason, on the model of the English rule under which a wife who killed her husband committed petit treason. Trial records from Virginia counties usually run to a few lines listing the accused, the charge and the outcome, so cases with this much detail are rare. Where a method is named it is almost always something put into food or drink, which is what European ideas about poisoning expected. In four trials between 1763 and 1798 the accused were said to have poisoned their owners while pretending to give them medicine.