A Stirring Tale from inside Earth.

In the quest to solve one of Earth's biggest puzzles, a team of geophysicists has proposed a new theory for how heat escapes from the planet's scorching depths. The hypothesis, backed up by fresh discoveries, has the potential to douse a debate that has burned since the concept of plate tectonics revolutionized earth science in the 1960s.

For decades, researchers have tried to discover exactly how heat leaks upward through the great rocky bulk of the planet, called the mantle. Although the mantle is solid stone, the intense heat causes the rock to flow slowly.

Most seismologists, who peer inside the planet by taking advantage of earthquake waves, see evidence that mantle rock mixes like boiling water in a pot. The hot rock, they say, rises from the bottom of the mantle to the top, cools off, and then sinks back down to complete a current of convection.

Researchers who study the chemistry of lavas, however, argue that the mantle resembles a double boiler, with separate upper and lower layers that each have their own systems of convection currents. In this case, the rock of the lower mantle, below 660 kilometers in depth, would not mix with that of the upper mantle.

In the new hypothesis, a team of researchers agrees with the double boiler system, except they picture the division much deeper and bumpier than previously thought. "The model reconciles the geochemical observations with the seismologic observations that have been difficult to reconcile for so long," says lead author Louise H. Kellogg of the University of California, Davis. Kellogg worked with Bradford H. Hager and Rob D. van der Hilst of the Massachusetts Institute of Technology. They publish their report in the March 19 Science.

According to Kellogg and her coworkers, the mantle has separated itself at a depth of about 1,600 km. Below that undulating partition lies a 1,300-km-thick layer of primitive mantle rock that represents what much of the planet was like in its infancy. Rock above the 1,600-km demarcation, however, has gotten stale, like old chewing gum. Over the 4.5-billion-year history of Earth, the upper half of the mantle has lost much of its allotment of gases and other important elements, which have congregated in Earth's crust and atmosphere.

Kellogg's team used a computer model to simulate double-layer convection, which carries heat to the surface. In their experiment, the two layers remain separate for billions of years--the geological equivalents of oil and water.

In the past, researchers have placed the hypothetical boundary at 660 km because that is a natural break point where rock gets squeezed into a more compact structure. Geochemists supported this concept because lavas suggest that the mantle contains a hidden reservoir of pristine rock. Seismic images, however, show pieces of ocean crust sinking well past the 660-km depth--leading the seismologists to reject the idea of a boundary at that level.

A deeper separation may prove more palatable. Although sinking ocean crust does breach the 660-km level, it meets some type of barrier below 1,600 km, according to a separate report in Science by van der Hilst and Hrafnkell Karason of MIT. Moreover, analysis of earthquake waves indicates that the lowermost mantle contains different rock than the upper two-thirds, they say.

In a third Science paper, a Japanese and a British researcher report finding a thin sheet of rock, which they interpret to be old ocean crust, sitting about 1,400 to 1,600 km below the Pacific seafloor. Kellogg notes that this depth is just above the proposed boundary between the primitive lower mantle and the stale mantle rock above. "You might expect to see old crust piled up there," she says.

The new hypothesis has given geoscientists a different target on which to focus their attention. "It is an important model that we need to test," says Kenneth C. Creager, a seismologist at the University of Washington in Seattle.

That sentiment is echoed by Albrecht Hofmann of the Max Planck Institute for Chemistry in Mainz, Germany. "As a geochemist, I'm quite excited about exploring the further consequences of all this."