Do Continents Really Move?

Look at a map of the Atlantic Ocean. The bulge of Brazil looks as if it would fit into the curve of Africa’s west coast, like two pieces of a jigsaw puzzle. People noticed this as soon as there were good maps of both coasts, in the 1500s and 1600s. But a fit could be chance. To show the continents were once joined, you need more than shapes.

The coast is not the edge

A coastline is not really where a continent ends. Around most continents the land carries on under the sea, as a shallow platform called the continental shelf, usually less than 200 m deep. At its outer edge the sea floor drops steeply, down the continental slope, to the deep ocean floor 4 or 5 km down. The true edge of a continent is the bottom of that slope, not the beach. Where the sea level is today is an accident: during the last Ice Age, 20,000 years ago, the sea was about 120 m lower and much of the shelf was dry land.

So the fair test is to fit the continents along the edges of their shelves. In 1965 the British geophysicist Edward Bullard and two colleagues did it on a computer, an early use of computers in geology. They fitted the edges where the sea is about 900 m deep (500 fathoms, as their charts measured it), halfway down the slope. Try it.

Along the edges of the shelves, the fit is astonishing: over thousands of kilometers of coast, the gaps and overlaps are mostly less than 100 km. A few pieces don’t fit, and they are the exceptions that prove the rule. The biggest overlap is the delta of the River Niger in Nigeria, which the river built out into the sea after the Atlantic opened, from mud it carried down. Pieces of land added later shouldn’t fit, and they don’t.

A shape that fits is a clue. Fossils are a stronger one, because living things don’t cross oceans easily.

Here are four of the living things whose fossils Wegener and others used. All four lived late in the Paleozoic era or early in the Mesozoic, from about 300 to 235 million years ago (you met the eras in Grade 8: The geological time scale).

None of them could have crossed an ocean. Mesosaurus and Lystrosaurus were land and shallow-water animals, and Mesosaurus is found only in rocks laid down in one shallow sea, on two coasts that are now 6,000 km apart: not in North America, not in Europe, not anywhere in between. Glossopteris seeds were not the kind that wind or birds carry across thousands of kilometers of ocean, and the plant is not found on the northern continents at all. And it isn’t just one fossil turning up in two places: whole sets of plants and animals match, layer after layer.

Now put the fossils on a map of the continents fitted together. Each one’s area runs across the joins, as if the continents had been torn apart after the animals and plants lived.

This great southern land is called Gondwana, after the Gond region of central India, where its rocks, coal and fossils were first studied, in the 1870s. India was part of it. Glossopteris is found just over Nepal’s eastern border, in the hills of Darjeeling and Sikkim.

Wegener compared it to a torn newspaper. If you tear a page in two, you can show the pieces belong together by fitting their edges. But if the lines of print also run straight on from one piece to the other, you can be sure.

The lines of print are the rocks. Different parts of a continent are made of rocks of different kinds and ages, in belts. If two continents were joined, the belts should run across the join.

Mountains match too. The old, folded mountains along the south of South Africa, the Cape Fold Belt, have a twin in Argentina, south of Buenos Aires: the Sierra de la Ventana, made of the same kinds of rock, folded the same way, at the same time, about 280–250 million years ago. Fit the continents together and the two ranges line up. In the north, the Appalachian Mountains of North America line up in the same way with the mountains of Scotland and Norway.

Ice in the tropics

The strangest clue is ice. About 300 million years ago, great ice sheets scraped across parts of South America, Africa, Antarctica, Australia and India. Where ice slides over rock, stones frozen into its bottom scratch long, straight grooves, and when it melts it leaves a jumble of clay and boulders.

Rocks like these are found in central India and in southern Africa, where it is hot today. And when the continents are put back together as Gondwana, the grooves on all of them point away from one region, as if one great ice sheet had spread out from a center, near where the South Pole was then. Coal from forests in Antarctica, and ice scratches in India: neither makes sense with the continents where they are now. Both make sense if the continents have moved.

Wegener’s clues showed that the continents had been joined. They didn’t show how they moved. That evidence was under the sea, and nobody could see it.

Until the 1920s the depth of the deep sea was measured by lowering a weighted rope until it touched the bottom: hours for a single measurement. Then ships began to use echo sounding: send a short pulse of sound down from the ship and time its echo from the sea floor. Sound travels through sea water at about 1,500 m/s.

By the 1950s, ships had crossed every ocean with their echo sounders running. Now it is your turn.

The sea floor is nothing like the flat, muddy plain people had imagined. Along both lines you found:

• Continental shelves and slopes at the edges of the continents, as in the jigsaw fit.
• Deep plains, flat and roughly 4 to 6 km deep, covered in mud.
• A mid-ocean ridge: a long range of underwater mountains, 1 to 3 km higher than the plains on either side (gentle and broad in the Pacific, steeper and more rugged in the Atlantic). In the Atlantic it runs down the middle, the same distance from both coasts and the same shape as them. Along its crest runs a deep crack, a rift valley. The ridges of all the oceans join into one chain about 65,000 km long, the longest mountain range on Earth, nearly all of it under the sea.
• In the Pacific, a trench: a long, narrow trough, much deeper than the plains, right beside a coast lined with mountains and volcanoes, the Andes. The deepest place in the oceans, the Challenger Deep in the Mariana Trench, is nearly 11 km down: Sagarmatha would fit inside with more than 2 km of water above its top.

On a map, the ridges are also cut across, every few tens or hundreds of kilometers, by straight cracks called fracture zones, where one stretch of the ridge is shifted sideways from the next, like the steps of a staircase.

A ridge in the middle of an ocean, the same shape as the coasts; trenches by the mountains; continents that fit. In 1962 the American geologist Harry Hess put the clues together, and in the late 1960s the whole idea was worked out. It is called plate tectonics (from the Greek tekton, a builder).

The Earth’s outer shell is not one piece. It is broken into about a dozen large pieces, and many smaller ones, called plates. A plate is about 100 km thick: the crust and the cold, stiff top of the mantle below it. The plates rest on hotter mantle rock that, over millions of years, can flow very slowly, like very stiff toffee. And the plates move, a few centimeters a year, carrying the continents and the sea floor with them. A continent doesn’t plow through the sea floor, as Wegener thought: it rides on a plate, like a passenger.

Where are the plates’ edges? Plates moving against each other make earthquakes, so the earthquakes show you.

Most of the Earth hardly ever shakes. Nearly all earthquakes happen along narrow lines, and the lines are the ridges, the trenches and the young mountain ranges you found on the sea floor and the continents. The lines join up into closed shapes: the plates. What happens at a plate’s edge depends on how the two plates move:

• Moving apart, at a mid-ocean ridge. Hot rock from the mantle rises into the gap, part of it melts, and the melt erupts as lava along the rift valley and cools into new sea floor. As the plates keep moving, the new rock is carried away on both sides, and more fills the gap. The sea floor is made here, like two conveyor belts running away from each other.
• Moving together. Where a plate of sea floor meets another plate, the old, cold, heavy sea floor bends down and sinks back into the mantle under it, making a trench. The sinking plate is followed down by earthquakes, and melting above it feeds a line of volcanoes, like the Andes. Where two continents meet, neither sinks back into the mantle, because continental rock is too light: they crumple and pile up into high mountains. That is happening in Nepal.
• Sliding past each other, along faults like the fracture zones, with earthquakes but no new or sunken sea floor.

So the Atlantic is growing. South America and Africa are on two different plates, moving apart from the Mid-Atlantic Ridge between them, and every year a little new sea floor is added along it. If that is true, the sea floor should be youngest at the ridge and older and older towards the continents. In 1968–69 the drilling ship Glomar Challenger drilled into the sea floor across the South Atlantic, near 30° S, and found exactly that: the oldest mud lying on the rock was 11 million years old 191 km from the ridge, and 67 million years old 1,270 km to the west. You can drill it yourself in Why Is the Sea Floor So Young? (Drilling for the proof), in Going Further 9–12.

Plates move about as fast as your fingernails grow: a few centimeters a year. That sounds like nothing. But the Earth has had a very long time.

To work out speeds over millions of years, a trick saves a lot of zeros. One kilometer is 1,000,000 mm, so 1 km per million years = 1 mm per year.

Today plates are measured directly, by GPS: the same satellites your phone uses, but with receivers fixed to solid rock that measure their position to within a few millimeters, day after day, for years. You met this use of GPS in Grade 9 Grade 9: Global Positioning System. The fastest plates move more than 10 cm a year; the slowest, a few millimeters.

Put the clues together and you can follow India’s journey. About 300 million years ago it was part of Gondwana, far in the south, next to Africa, Madagascar and Antarctica, with ice sheets scraping its rocks and, later, Glossopteris forests growing in its swamps. Gondwana began to break up about 180 million years ago. India broke away and moved north across what became the Indian Ocean, for a time at up to about 15 cm a year, one of the fastest journeys of any continent we know of. Ahead of it lay an ocean, the Tethys, whose sea floor sank under Asia.

About 50 million years ago India reached Asia. A continent can’t sink back into the mantle, so the edges of the two crumpled and piled up, and the Himalaya began to rise, carrying up the layers of mud and lime laid down in the seas along India’s northern edge. That is why the shaligrams that pilgrims find along the Kali Gandaki are fossil sea animals, ammonites, and why the rock at the top of Sagarmatha is limestone made on the floor of a warm, shallow sea more than 440 million years ago, with broken bits of sea lilies in it.

So most of the rock of Nepal was once part of India. As India slides north under Tibet, along a great fault called the Main Himalayan Thrust, slices of its crust are scraped off its top and stacked up, one over another, into the hills and the high Himalaya. The Terai lies on the Indian Plate itself, south of where that fault reaches the surface.

It hasn’t stopped. GPS stations across Nepal show India still pushing north, and about 2 cm a year of that push squeezes Nepal itself: southern Nepal is moving closer to Tibet, year by year. The rocks can’t squeeze for ever. The strain builds up, until a fault deep under the hills slips in a few seconds. On 25 April 2015, in the Gorkha earthquake, GPS stations in Kathmandu jumped about 2 m to the south and rose by more than 1 m.

• How Does One Rock Become Another? Plates make mountains and sink sea floor into the mantle: how the rocks of the Earth are melted, squeezed, worn away and made again.
• How Do Rocks Tell Time? How the layers of rock and the fossils in them give the order of the Earth’s history.
• Why Is the Sea Floor So Young? In Going Further 9–12: magnetic stripes on the sea floor, the drill holes, and why no sea floor is older than about 200 million years.