What Built the Himalayas and the Alps

You have probably held the Matterhorn in your hand. It is on the Toblerone box, a silhouette with the bear of Bern hidden inside the rock, and it was there for decades until 2023, when production moved to Bratislava. Swiss rules brought in in 2017 say you cannot put a Swiss landmark on a bar made elsewhere, so the mountain was swapped for a generic peak and the wrapper stopped saying Swiss chocolate and started saying created in Switzerland.

Which is funnier than the lawyers intended, because the mountain was never entirely Swiss to begin with.

Above about 4,200 metres, the Matterhorn is gneiss belonging to the Dent Blanche nappe. That rock came from the Apulian plate, a piece of continent that broke away from Africa around 100 million years ago. Below 3,400 metres you are standing on the Tsaté nappe, which is ophiolite and sediment, the remains of the floor of the Piedmont-Liguria ocean. There is no such ocean now.

So the mountain on the chocolate is African rock, resting on a dead seabed, resting on Europe. That is not a metaphor. It was carried there and put down on top.

The man who worked it out was laughed at

Alfred Wegener stood up in Frankfurt on 6 January 1912 and said the continents move. He published the full argument in 1915. Most geologists thought he was a meteorologist with no business in their subject, and the criticism was withering.

He died in November 1930 on the Greenland ice sheet, on his fourth expedition there, and his body was not found for six months. The argument was not settled until the 1960s, when palaeomagnetism and seafloor spreading arrived and made it obvious.

What he was right about is that the outer shell of the Earth is broken into plates that grind around at a few centimetres a year. Roughly the rate your fingernails grow. Over fifty million years, that is thousands of kilometres.

Why some collisions build and some swallow

Everything else follows from one property, which is density.

Oceanic crust is basalt, and it is heavy. Continental crust is granitic, and it is not. Where an oceanic plate meets a continental one, the ocean floor loses and slides underneath, which is why the Pacific rim is a ring of trenches and volcanoes rather than a wall of rock.

But when two continents arrive at the same place, neither one will go down. Both are too buoyant to be pushed into the mantle. The ocean that used to sit between them gets consumed first, and then the collision has nowhere left to send the material except up, and sideways, and back over itself.

Every large mountain range on Earth is the record of two pieces of continent that both refused to go under.

India arrived faster than anything else we know of

India broke away and started north around 100 million years ago, and it moved at something like 20 centimetres a year, quicker than any plate motion since measured. One likely reason is that the Indian plate is thin, around 100 kilometres, roughly half the thickness of the other fragments of Gondwana, so there was less of it to drag.

It reached Eurasia about 55 million years ago. The Tethys Ocean that had been in the way was gone, and the crust began to pile up and double in thickness.

The proof is at the top. Everest stands 8,848.86 metres, and the rock at its summit is the Qomolangma Formation, an Ordovician limestone. When geologists put samples from near the summit under a microscope, they found the crushed remains of trilobites, crinoids and ostracods. Lower down, marble collected at about 8,300 metres turned out to be as much as five percent the ghosts of crinoid stems.

The highest point on the planet is a shallow tropical seabed, complete with the animals that used to live on it, lifted almost nine kilometres into the air.

The Alps did the same thing untidily

Europe's version was slower and much messier. Africa and its Apulian promontory closed on Europe over roughly thirty million years, and rather than folding into one clean arch, the rock came off in sheets.

Geologists call them nappes. Slabs of crust tens of kilometres across, sheared off their base and shoved over the top of whatever was in front, sometimes folded right over so the layers end up upside down. The Alps are a stack of these, laid one on another like badly dealt cards.

Which is why the Matterhorn reads the way it does. Bottom to top, you are climbing through the collision in the order it happened.

Most of a mountain is underneath it

There is a piece of this you cannot see, and it was worked out by accident.

In the nineteenth century, surveyors measuring India kept getting readings that did not add up. A plumb line hung near the Himalaya should be pulled sideways by all that rock, and it was, but nowhere near as much as the size of the range predicted. Surveyors in the Andes had run into the same thing.

The answer is that the crust floats on the denser, slowly flowing rock beneath it, the way an iceberg floats on water. Pile material up on the surface and it does not simply sit there. It presses down and the whole block settles deeper, so most of the thickening goes downward. On the standard treatment a mountain's root runs about five times its height, meaning eight kilometres of Himalaya sits on something like thirty-two kilometres of crust pushed into the mantle, against roughly forty kilometres for ordinary continental crust.

The mass was never missing. It was below the surveyors' feet.

The mountains went on to change the weather

In 1992 Maureen Raymo and William Ruddiman published an argument in Nature that mountain building is not just a product of the Earth's machinery but an input into its climate.

Push up a range that size and you expose an enormous quantity of fresh silicate rock. Weathering that rock consumes carbon dioxide and locks it away in the ocean as carbonate. Do it continuously for forty million years, across the Himalaya and the plateau behind it, and you draw down enough to help tip the planet into the cycle of ice ages we are still technically in. The same uplift also gave Asia its monsoon, by putting a wall in the path of the air.

This is genuinely argued over. Recent work questions whether the erosion and the cooling line up as neatly as Raymo and Ruddiman proposed, and the size of the effect is still being fought about in the journals. The part nobody disputes is that a range on that scale rearranges the atmosphere around it, and keeps doing so for as long as it stands.

It has not stopped

India is still pushing into Eurasia at around 17 millimetres a year, and parts of the Himalaya are still going up by as much as 10 millimetres a year. The strain does not release smoothly. In April 2015 a section of that collision slipped near Gorkha in Nepal and killed roughly nine thousand people.

So the Matterhorn on the wrapper is not a finished object, and neither is anything else on a map. It is a slow-motion photograph of two continents in the middle of hitting each other, and the only reason it looks permanent is that we are not around for very long.

Sources

• Wikipedia, "Matterhorn" (the 4,478 metre height, the Dent Blanche nappe gneiss above 4,200 metres from the Apulian plate, the Arolla series beneath it, and the Tsaté nappe ophiolites of the Piedmont-Liguria ocean below 3,400 metres).
• Wikipedia, "Toblerone" (the Matterhorn silhouette with the hidden Bern bear, and the 2023 move to Bratislava that forced the change under the 2017 Swissness rules).
• Wikipedia, "Alfred Wegener" (the 6 January 1912 Frankfurt presentation, the 1915 book, the hostile reception, his death in November 1930 on the Greenland ice, and acceptance in the 1960s).
• Wikipedia, "Indian Plate" (the separation around 100 million years ago, the roughly 20 centimetre a year peak rate as the fastest known plate motion, the plate's unusual thinness, and collision from about 55 million years ago).
• Wikipedia, "Mount Everest" (the 8,848.86 metre height, the Qomolangma Formation as Ordovician limestone, and the trilobite, crinoid and ostracod fragments found in summit samples and in the Yellow Band marble at 8,300 metres).
• Wikipedia, "Geology of the Himalaya" (uplift rates approaching 10 millimetres a year, and the present convergence of about 17 millimetres a year).
• Wikipedia, "Isostasy" (the floating-crust model, the nineteenth-century survey anomalies in India and the Andes that revealed mountain roots, and the roughly five-to-one root-to-height ratio against about 40 km for average continental crust).
• Wikipedia, "Nappe" (what a nappe is and how thrust sheets stack and overturn).
• Raymo and Ruddiman, "Tectonic forcing of late Cenozoic climate", Nature, 1992 (the argument that uplift-driven silicate weathering drew down carbon dioxide and cooled the Cenozoic).
• Wikipedia, "April 2015 Nepal earthquake" (the Gorkha earthquake and its death toll).