Einstein was right -- again.

Using a rare alignment of Jupiter against a far-off quasar, scientists for the first time have succeeded in measuring the speed of gravity, a fundamental constant of physics described by Albert Einstein in his general theory of relativity.

The new number announced yesterday will curb some of the more exotic notions of theorists working to formulate a "theory of everything" that unifies concepts of particle physics with scientists' understanding of gravity, and presents a coherent portrait of nature from the tiniest subatomic particle to the most titanic structures in the cosmos.

Isaac Newton "thought that gravity's force was instantaneous," said Sergei Kopeikin, a physicist at the University of Missouri-Columbia who led the observing team. "Einstein assumed that it moved at the speed of light, but until now, no one had measured it."

At a meeting of the American Astronomical Society in Seattle, Kopeikin and colleagues announced their finding that the force of gravity propagates outward from a source at 1.06 times the speed of light (186,000 miles per second in a vacuum), with a 20 percent margin for error. If confirmed, the number rules out theories based on the notion that the speed of gravity is, as Newton conceived it, infinite or instantaneous.

In Einstein's conception, gravity is a warping of the space-time continuum, sometimes depicted graphically in terms of balls rolling around in a net. A massive object such as a bowling ball -- or the planet Jupiter -- makes a bigger hole, or "gravity well," in the webbing -- the geometry of space -- than a less weighty object, say, a baseball.

In everyday terms, gravity keeps people, cats and other untethered objects from flying off the planet. The sun's gravity keeps Earth and the other planets from flying off into space. If the sun suddenly vanished from the sky, it would take about eight and a half minutes for the last particle of sunlight to cross the 93 million-mile distance to Earth, leaving the planet in the dark. Based on the new number for the speed of gravity, at about the same moment the light went out, Earth would "feel" the end of the sun's gravitational influence and skid off into the cold void, said Ed Fomalont of the National Radio Astronomy Observatory in Charlottesville, who worked with Kopeikin on the project.

The scientists used the continent-wide Very Long Baseline Array, a radio-telescope system with sites in the continental United States, Virgin Islands and Hawaii, along with a 100-meter radio telescope in Effelsberg, Germany, to achieve the effect of a huge single instrument capable of a very precise observation. They were precise to the width of a human hair seen at 250 miles, Fomalont said.

"We had to make a measurement with about three times more accuracy than anyone had ever done," Fomalont said.

They took advantage of the once-in-a-decade alignment of Jupiter against a grouping of brilliant quasars billions of light years away. The alignment occurred Sept. 8. The scientists fretted about weather on Earth and possible electromagnetic storms on Jupiter, which could have ruined their plan. But they got lucky.

They compared their experiment to one done in 1919.

Einstein had predicted in 1915 that starlight passing the sun should be bent gravitationally by a certain amount. Four years later, a team led by Arthur Eddington, a noted British astrophysicist, verified the prediction with measurements made during a solar eclipse. Eddington found that the position of stars in the sky appeared to move depending on whether the sun was near their line of sight -- because of the deflection of the light by the sun's mass.

Kopeikin and Fomalont used Jupiter, instead of the sun, because Jupiter moves relative to Earth in a way that the sun does not, and this motion was crucial to their experiment. Jupiter passes close enough to the path of radio waves from a bright-enough quasar only about once a decade, Kopeikin calculated. (Quasars are the most luminous and energetic objects known, brilliant beacons in far-off regions of the cosmos, and are believed to be galaxies with active black holes at their core.)

"Because Jupiter is moving around the sun, the precise amount of the bending depends slightly on the speed at which gravity propagates from Jupiter," Kopeikin said.

The effect they measured was a very slight bending of the radio waves emanating from the background quasar, caused by the gravitational influence of Jupiter. If the speed of gravity were infinite, the amount of bending would have been different.

A researcher in Japan has contended that the technique actually measured the speed of light, not gravity, but his interpretation is not attracting support.

Kopeikin and Fomalont spent considerable time conducting tests of the system on other days, to calibrate their margin of error and to account for the effects of Earth's atmosphere on the observations.

"No one had tried to measure the speed of gravity before, because most physicists had assumed that the only way to do so was to detect gravitational waves," Kopeikin said. Large instruments have been constructed to detect these gravitational ripples, believed to flow from events such as the collision of neutron stars.

"Our main goal was to rule out an infinite speed for gravity, and we did even better, Fomalont said. "We can confidently exclude any speed for gravity that is over twice that of light."