A twelve year holy war burns the libraries of Christian Ethiopia

In March 1529 Ahmed Gran, who led the armies of the Muslim sultanate of Adal, destroyed the Ethiopian army at Chembra Koure. He then worked through the country year by year, starting in the east and moving north and west, targeting churches and monasteries. Dabra Libanos went up in flames on 14 July 1532 and a number of its monks threw themselves into the fire. The cathedral of St Mary of Zion at Aksum, which probably dated from the sixth century, was reduced to rubble. Early in 1540 the royal fortress of Amba Geshen fell with the treasury and the state archives inside it. Tens of thousands of people accepted Islam and many more were marched off into slavery. Almost no Ethiopian manuscript survives from before 1530. By 1540 Gran considered himself king of the country and much of the population obeyed him. His soldiers had brought their wives from Adal and settled.

Mocombe opens his historical chapter with Columbus claiming the island for Spain and the Spanish exploiting its gold mines with Taino labour. He puts the pre-contact Taino population at between three and four million and states that after fifty years of Spanish rule they had been destroyed by the hardship of enslavement, organised massacres and imported disease. He calls this one of the most brutal exterminations on record. The mechanism he then traces is the substitution that followed. A Spanish priest protested against the massacre of the indigenous population and demanded an end to the injustices against them, and advocated importing Africans to work the mines and plantations as a way of ending Indian slavery on the island. Mocombe treats the two populations as directly linked by that argument rather than as separate chapters. Where this account departs from the standard one: Mocombe adopts a pre-contact Taino population of three to four million, at the very top of a range where most current estimates for Hispaniola are far lower.

During the 1530s Spanish colonists built the first real sugar works on Hispaniola: cane fields, a mill driven by oxen or water to crush the stalks, and a boiling house to reduce the juice. Cane had been growing on the island since 1505, though a mill is a different order of commitment. It costs a great deal, it cannot be moved, and it earns only if a large workforce is available in the same place every year for decades. That requirement is what fixed African slavery on the island. Indigenous labour was gone, and it had in any case been held under grants that were legally temporary and personal. Planters wanted a workforce they owned outright and could replace, and the Atlantic trade could supply one. Spanish Hispaniola never became a great producer; the mills were modest and the colony was soon overshadowed by Mexico and Peru. The system built here was carried to Brazil and later to the French west of the island.

By the 1530s merchants in Lisbon and Seville were organising the Atlantic crossing as a single commercial operation with three legs. A ship left Europe loaded with textiles, iron bars, brandy, firearms and cowrie shells. On the African coast that cargo was exchanged for captives. The ship then crossed to a Spanish American or Brazilian port, sold the people it carried, took on sugar, hides or silver, and returned to Europe. Every leg paid. Nothing sailed empty. This is the structure every later colonial power adopted, the Dutch, the English and the French copying an Iberian design with little modification, and it explains how fast the traffic grew. An investment with a calculable return, financed by banks, covered by insurers and protected in law, expands in a way that raiding never does. Understanding Saint-Domingue in 1789 means seeing its plantations as the middle stage of a machine that had already been running for two hundred and fifty years.