“Famines gather history around them,” we are told, even more so, it seems, with high numbers of dead. These numbers are treated sometimes like monuments for famines, increasing over time according to utilitarian concerns. Sources for a famine on Madagascar show that though high numbers may be useful in drawing attention to a calamity, people closer to the event may not locate this history or situate their memory via numbers. Emphasizing numbers in lieu of other ways of remembering and also forgetting a calamity appear not to be very good guides to this history.

The killing famine that struck southern Madagascar in 1930–31 attracted substantial written comment among the French. Everyone seemed to have an opinion about this famine, which followed the surprising and dramatic killing of the predominant species of prickly pear cactus by cochineal insects in the late 1920s. A large area, a seventh of the island (approximately a sixth of France), with a population at the time of around a half million people and perhaps two million head of cattle, was effected by the biological war on cactus. “Cactus pastoralists” were suddenly without a very resourceful plant. It had provided thick fences of protection to these herders and their cattle; its fruit and water a mainstay for people; its singed cladodes a critical source of water and sustenance for cattle.

The “furnace of contamination”—the rapidly reproducing cochineal choking to death their cactus hosts—started in 1925 at the southwest provincial center of Toliara and spread to the east, north, and south at a rate of 100 kilometers per year.