The long-awaited results from a NASA satellite sent up to test Einstein's theory of gravity will have to wait a few months longer. Although researchers have spotted one of the two spacetime-bending effects they were looking for with the Gravity Probe B (GP-B) relativity mission, they are still working to pry the second, fainter effect from the data.

Researchers gave their colleagues a peek at the analysis this weekend at the annual meeting of the American Physical Society. "We see the larger of the two relativity effects completely visibly in the data," says physicist and mission leader Francis Everitt of Stanford University. "We're making steady progress towards the full determination."

Einstein's theory of general relativity, put forth in 1915, holds that space and time are united as one malleable entity, spacetime. Decades in the making, GP-B was launched in April 2004 to sensitively measure two properties of spacetime: the geodetic effect, a 2.8-centimeter (1.1-inch) contraction in the circumference of Earth under its own gravity; and frame dragging—the weaker of the two and GP-B's main quarry—in which Earth twists spacetime like a ball rotating in honey.

Few experts would predict that general relativity would make the wrong prediction. Nevertheless, Everitt says, "testing Einstein is exhilarating."

Unfortunately, the $760-million GP-B got scooped in 2003 when researchers reported seeing frame dragging by comparing the heights of two satellites in Earth orbit. But that experiment, called LAGEOS (Laser Geodynamics Satellite), was not very precise, says physicist Clifford Will of Washington University in St. Louis, a member of GP-B's external review board. "To me, [GP-B is] still an astounding achievement in its own right," he says. "This is a unique experiment," he adds, and may yield a more precise answer than LAGEOS.

GP-B consists of four rotating, ping-pong ball–size spheres that act as sensitive gyroscopes. As Earth drags spacetime, it drags each sphere's spin axis with it. A telescope measures the change by comparing the axis with its alignment relative to a star in the Pegasus constellation.

To detect frame dragging, researchers have to identify a 0.000011-degree shift in the spin axis, which roughly equals the width of a human hair as viewed from a quarter mile away. "We're seeing glimpses of that," Everitt says, but the group still has to account for other influences on the spheres.

GP-B stopped collecting data in September of 2005. Everitt and colleagues had originally planned to announce their results last summer but unexpected subtleties arose, including electric fields that changed the spheres' orientations. Everitt says the team is now sorting these effects from the true spacetime signal and might be finished by December. When asked about the delay, he laughs: "You can put it down to my naiveté … that we thought we were going to do it quicker than was realistic."