The place is so inhospitable, in brief, that scientists must base their knowledge of the center of the Earth on indirect evidence. Like physicians who look at an embryo with ultrasound waves, earth scientists study the structure of the Earth by analyzing seismic waves that radiate from earthquakes. On their way through different layers of rock, seismic waves change in a way that gives clues about the structures they have passed through.

Basing their work on such observations, earth scientists have come up with a model of the planet that resembles the layering of an egg. The Earth’s crust is as delicate and brittle as the shell. Underneath the crust is Earth’s mantle, which corresponds to the egg’s white. The upper part of the mantle is partially molten, whereas the lower part consists of solid rock. The iron core in Earth’s center is the yolk. At the boundary between the mantle and the core, about 1,800 miles beneath the surface and roughly half-way to Earth’s center, the physical properties change abruptly. The swirling iron of the outer core has the consistency of water, and its temperature exceeds that of the neighboring solid, rocky mantle by more than 3,000 degrees Fahrenheit. Finally, the inner core, Earth’s center, consists of a solid iron ball about 1,500 miles in diameter.

Now, seismologists Sebastian Rost and Justin Revenaugh from UC Santa Cruz have found new evidence that the boundary between the solid mantle and the liquid outer core is not as sharply defined as scientists once believed. From the seismic fingerprints of seven earthquakes beneath the Pacific Ocean, Rost and Revenaugh resolved a very thin patch of the core-mantle boundary about the size of Santa Cruz that has both mantle-like and core-like properties. "It is a very flimsy sponge of solid material with a lot of liquid iron in it," Revenaugh says.

"People have assumed for a long time that the core and the mantle are completely separate reservoirs of material with no interchange across the boundary at all," he adds. "Now people are starting to think that there is some communication."

So far, Rost and Revenaugh can only speculate about what this odd patch does. And because they haven’t yet found it in other places, they can only guess that such patches cover the entire core-mantle boundary. Its location at precisely that boundary suggests, however, that it is part of the process that regulates the heat flow between the hellish furnace of the core and the cooler rocks of the mantle above. Volcanoes, hot springs, and geysers are the most conspicuous signs of this escaping heat. Less evident is that this heat is the force behind the slow drift of continental plates.

The radical notion of drifting continents dates back to 1912, when Alfred Wegener, a German meteorologist, suggested that the continents once clustered together to form one vast landmass. Wegener was struck by the fact that South America’s east coast and Africa’s west coast fit together like two puzzle pieces. Wegener’s ideas, however, weren’t seriously considered until the 1960s — mainly because earth scientists didn’t see a plausible mechanism that could have broken up the super-continent.

The skepticism vanished when scientists found evidence that vast plumes of hot rock coming from the core-mantle boundary surge toward Earth’s surface like air bubbles in a glass of water. Millions of years ago, such plumes first lifted and then eventually broke the monolithic continental landmass into huge pieces. These pieces now slowly drift about as today’s continents. According to measurements of rock formations in the Atlantic Ocean, the gap between South America and Africa widens by as much as half an inch every year.

The separation of continents and the collision of other continents remodeled the appearance of Earth profoundly. Oceans formed and mountain ranges rose, lush jungles turned into deserts, and land once flooded fell dry. Animal and plant species colonized and adapted to these newly formed ecological niches. The diversity of life forms widened.

"In terms of the history of life on Earth, super-continent break-ups are very important. If that’s related to plumes, then we have this tie to the base of the mantle," Revenaugh says. "It is kind of a neat thing, that the history of life might be tied to these little patches 1,800 miles below us."

Rost and Revenaugh discovered the patches about 900 miles north of New Zealand. They analyzed seismic waves of earthquakes originating beneath the islands of Tonga and Fiji. Like any other earthquake, they spawned two types of seismic waves. The pressure or "P" waves propagate in a sort of push-pull manner, like a crawling earthworm. The shear or "S" waves vibrate back and forth perpendicular to their travel path, like a fast-moving snake. Generally, P waves travel faster than S waves. And unlike S waves, P waves can travel through liquid. Both seismic waves either bend or reflect when they encounter a layer of rock with different density — much like a ray of light that passes from air into water or vice versa.

Both P and S waves reflect off the core-mantle boundary and zip all the way back to the surface. P waves might reflect as P waves, or they can convert into S waves. S waves behave likewise. From the changes the waves undergo during the rebound, scientists learn a great deal about the boundary’s structure.

Rost and Revenaugh found odd looking ScP waves — S waves that are converted into P waves — that they didn’t understand at first. But then Rost, a postdoctoral fellow in Revenaugh’s lab, discovered that these ScP waves must stem from a structure predicted by Bruce Buffett from the University of British Columbia, Canada, and his colleagues on theoretical grounds more than one year ago. "The [patches] were postulated before, and I knew about the models, but I never thought about detecting them," Rost says.

Buffett came up with the idea of iron-rich patches at the core-mantle boundary when he and two colleagues, Edward Garnero from Arizona State University and Raymond Jeanloz from UC Berkeley, looked for a way to predict the wobble of Earth’s rotation axis.

This so-called nutation — a sort of nodding motion of the rotation axis — is caused by the gravitational pull of the sun and the moon. Astronomers are especially interested in Earth’s motion in space. When they track a distant spacecraft with telescopes, they have to know when and by how much the Earth reorients itself in space, Buffett says. If they don’t correct for Earth’s nutation, they lose the spacecraft pretty quickly.

Buffett, Garnero, and Jeanloz knew that the way in which the Earth responds to this celestial pull is strongly affected by the fact that the liquid iron sloshes back and forth in its interior. The amount of sloshing scientists agreed upon at the time was quite vigorous. When Buffett and his colleagues put this variable into their calculation to predict Earth’s nutation, they merely came up with a close approximation. When the scientists assumed a less vigorous liquid iron core, however, they succeeded. But what force would be strong enough to slow down thousands of cubic miles of sloshing iron? Buffett suspects Earth’s magnetic field.

The moving iron in the core generates the magnetic field, the powerful natural force that drives both compass needles and the northern lights. Buffett’s model suggests that conductive iron patches at the core-mantle boundary deflect the magnetic field that passes otherwise unhindered through the mantle. "The magnetic field threads out from the core through the mantle and we ultimately see it at the surface," Buffett says. "When it passes through this layer of conductive material [the iron-rich patches], it tends to connect the fluid a little bit more tightly with the mantle." Thus the patches slow down the whirling iron.

How do these patches emerge in the iron core? Nobody knows for sure. Buffett and his colleagues believe, however, that lighter elements such as oxygen, sulfur, silicates, or carbon are dissolved in the liquid iron to the point of saturation — like a glass of water that contains so much dissolved salt that any more salt simply sinks to the bottom. As Earth’s core cools, some of the liquid iron solidifies and amalgamates with the solid inner core. Subsequently the concentration of these lighter elements increases until they precipitate out and float as sediments on top of the iron soup like foam on root beer. This froth accumulates in pockets at the core-mantle boundary. During that process, liquid iron is trapped and incorporated.

Other scientists find the research intriguing. "It is another important piece of information that tells us the earth isn’t as simple as we so commonly portray it to be in our introductory textbooks," says Edward Garnero of Arizona State University. "It is important because it opens up our perspective that we might have a scum collecting in places where once we assumed it was all just homogenous liquid iron."

But Garnero cautions that this is only the beginning. "In any first study like this, it is rare that everybody says ‘You showed us the truth!’" he says. Now that the scientists have an idea what these patches look like in their earthquake data, they should try to find them in other places too, Garnero suggests. A finding is generally accepted as the ‘truth’ when enough evidence accumulates and points in the same direction, he says.

The knowledge we have of the structures deep in our planet is still sketchy. And due to the elusive nature of inner Earth, scientists will never know for sure whether their models reflect the truth. All they can do is gather bits and pieces of information and then try to come up with a plausible explanation. "It is detective work," says Bruce Buffett of the University of British Columbia. "But if you like mysteries and detective stories, this is the business to be in."