A gemstone's wild ride: how diamonds erupt from deep within Earth.

Thanks to their hardness, durability, and rarity, diamonds are symbols of eternal love and wealth. A new analysis of how these gems emerge from the depths where they were forged billions of years ago suggests that they should also evoke images of diamond-studded fountains of gas and rock.

Almost all diamonds come from volcanic deposits called kimberlites. Tens of thousands of such deposits, some more than a kilometer across, have been found protruding through Earth's surface. However, only one in every 200 of them contains gem-quality diamonds. Increasingly, scientists are looking at kimberlites' odd mix of geological characteristics to discern how diamond deposits get there in the first place.

MYSTERIOUS MINERALS With or without diamonds, kimberlites are enigmatic, says James W. Head III, a planetary geologist at Brown University in Providence, R.I. The carrot-shaped deposits, which can extend to depths of 2.5 kilometers, are clearly volcanic in origin, but areas around them aren't marked by large amounts of lava. Furthermore, kimberlites contain large numbers of glass spherules that typically form from airborne droplets of molten rock Locked within kimberlites, however, the spherules apparently never make it above ground. Finally, as much as half the jumbled rocks in kimberlite deposits comes from Earth's mantle, the hot, viscous material that lies below depths of 35 km.

The presence of diamonds in kimberlites suggests that these volcanic deposits originate at depths of at least 250 km, where immense pressures keep the gems' crystal structures stable. Furthermore, the material in diamondbearing kimberlites must have risen from there to Earth's surface in just a few hours. That's because at shallower depths, where pressures are lower but temperatures remain high, the instability of diamonds would lead to their quick transformation into graphite.

In the May 3 Nature, Head and volcanologist Lionel Wilson of Lancaster University in England describe a model of a kimberlite eruption that explains the resulting mix of diamonds, spherules, and mantle rock in a carrot-shaped deposit.

At the beginning of an eruption, the researchers suggest, a wedge-shaped mass of carbon dioxide, pointing upward, forms inside a crack of some sort in Earth's semi-solid mantle. At such depths, carbon dioxide is typically a liquid.

Beneath the wedge is a large, buoyant mass of molten rock saturated with carbon dioxide. Its buoyancy forces the wedge of liquid carbon dioxide upward, further cracking the overlying rock. As the crack grows and the tip of the wedge races toward Earth's surface, pressures in the topmost layers of magma begin to drop, and some of the carbon dioxide in the molten rock fizzes out, creating a layer of foam.

Immense pressures in the mantle cause diamond-bearing chunks of rock to explode from the walls of the fissure and fall through the wedge of liquid carbon dioxide into the magma below. Once this rock-busting process begins, the wedge of carbon dioxide and the magma immediately below it race upward at speeds up to 180 kilometers per hour.

BREAKING GROUND When the wedge of carbon dioxide breaches the surface, pressure is released and the liquid evaporates explosively. This sends a jet of carbon dioxide gas, molten rock, and rubble skyward at a rate up to four times the speed of sound.

The rapid expansion of the carbon dioxide cools the molten rock, freezing it solid and plugging the eruption. Then, pressure builds and shock waves reverberate through the cavity, shattering its walls as well as the magma plug. The eruption begins anew, says Head.

Repeated eruptions of carbon dioxide, magma, and rubble blast a cone-shaped hole at Earth's surface. The series of eruptions runs out of steam probably no more than an hour or so after the carbon dioxide wedge breaches the surface. As the action wanes, the hole becomes filled with a porous mix of glass spherules, chunks of rock carried up from the mantle, fragments of volcanic ash and chilled magma, and, sometimes, diamonds.

The Head-Wilson hypothesis "very nicely explains how mantle rocks are transported to the surface," says Stephen E. Haggerty, a geologist at Florida International University in Miami. However, he notes, many questions remain about what happens deep within Earth at the beginning of an eruption.

For example: How might the viscous, nearly molten rocks in the mantle become cracked in the first place? At such high temperatures and pressures, why don't cracks in the rocks seal themselves?

"I like [Wilson and Head's] model, which takes a holistic view and tracks diamonds from their source to the surface," says Kelly Russell, a volcanologist at the University of British Columbia in Vancouver. Like Haggerty, Russell says that there are plenty of questions left to be answered by further analysis and fine-tuning of the model. For one thing, the role that water in mantle rocks might play in the eruption process isn't yet clear.

"This model makes a good target," Russell notes. "Now, the rest of us [scientists] can shoot at it"