For most of human history, the continents were assumed to be permanent features of the Earth, fixed where they had been since creation. The modern theory that they are in fact drifting slowly across the planet's surface - colliding, separating and occasionally destroying one another - was not accepted by mainstream geology until the second half of the 20th century, and even then only after decades of sceptical resistance.
The idea was first seriously proposed in 1912 by the German meteorologist Alfred Wegener, who was struck by the remarkable way in which the coastlines of South America and Africa seemed to fit together like pieces of a puzzle. He gathered further supporting evidence from matching rock formations on either side of the Atlantic and from fossils of species that appeared to have lived on both sides of the ocean at the same time. Wegener called the process "continental drift." Most geologists of the period rejected his theory, however, because he could not explain what force could possibly move whole continents.
The missing mechanism only emerged in the 1950s and 1960s, when new oceanographic instruments revealed the geology of the sea floor. Scientists discovered long underwater mountain ranges, running for thousands of kilometres, where molten rock was continuously rising from below and pushing apart the crust on either side. The surface of the Earth, it turned out, was not a single shell but a set of enormous rigid plates floating on a deeper, partly liquid layer.
The modern theory, known as plate tectonics, explains a wide range of phenomena in a single framework. Earthquakes and most volcanoes occur along plate boundaries. The great mountain ranges of the world formed where plates collided. Even the shape of coastlines and the distribution of animals and plants can be understood in terms of a planet whose surface is in constant, if extremely slow, motion. Wegener, who did not live to see his theory vindicated, is now considered one of the founders of modern earth science.