Though they proceed at far less than a snail's pace, Earth's continents are constantly on the move. Indeed, some 200 million years ago, they had a radically different arrangement from today's far-flung configuration, lumped, as they were, into a single giant landmass known as Pangea. But exactly how the continents fit together back then has proved puzzling: whereas the geological evidence supports a widely accepted model dubbed Pangea A, records of the earth's magnetic field do not. Findings presented yesterday in San Francisco at a meeting of the American Geophysical Union, however, appear to reconcile those differences.

In the Pangea A arrangement, South America lies against the southern edge of North America; Africa, just east of South America, borders the Atlantic coast of North America and is southwest of Europe. The paleomagnetic data, on the other hand, indicate that "the southern continents should be a little farther north," says University of Michigan geologist Rob Van der Voo. As a result, some researchers have proposed models that place northwestern South America alongside North America's east coast, or even farther east, just south of Europe. Others insist that the geological data fail to support this configuration.

But Van der Voo and his colleague Trond Torsvik of the Geological Survey of Norway suggest the two data sets need not contradict one another. Scientists have generally viewed Earth's magnetic field as akin to that of a bar magnet, with north and south magnetic poles. In reality, the field does have some extra components, but because these vary over time, researchers assumed they cancel out in the long run. Yet according to Van der Voo and Torsvik this is not the case. Their analysis revealed long-term, non-dipole components in the magnetic field. Theoretically, if those non-dipole components did not get averaged out, continental positions indicated by paleomagnetic data would differ slightly from those based on a purely dipolar field. In fact, when the team estimated the continental configuration based on the revised magnetic field data, their results fit almost perfectly with the Pangea A model. "The broader implications of this study are that paleomagnetic results for other times and other continental configurations must now be re-evaluated with the new geomagnetic field model that should include some 10 percent non-dipole fields," Van der Voo says, "this will keep us busy for decades."