Gravity gets measured to greater certainty.

Despite more than 200 years of experiments, the best laboratory determinations of the force of gravity have yielded no better than a ballpark figure. Adding to the haziness, meteorologists have scarcely probed the force's strength at distances of 1 millimeter or less. There, gravity may vary drastically from Newton's law, possibly signaling the presence of unseen extra dimensions, some theories suggest (SN: 2/19/00, p. 122).

Penetrating the fog, researchers in Seattle now report preliminary results from two separate gravity experiments. One set of findings offers the most precise value thus far for the gravitational constant, so-called Big G, which relates mass and distance to the strength of gravity in Newton's law.

The other experiment provides the first measures of gravity between objects separated by as little as 0.2 mm. No deviation from Newton's law turned up. Both University of Washington teams unveiled results at the American Physical Society meeting last week in Long Beach, Calif.

The new precision comes at a time when determinations of G disagree notoriously (SN: 4/29/95, p. 263). In particular, a 1994 German figure lies far out of line with most other recent values, including this latest one. Consequently, in a just-completed reevaluation of fundamental physical constants, the Committee on Data for Science and Technology (CODATA) of the International Council for Science in Paris has increased G's uncertainty by a factor of 12.

In typical experiments designed to find G, researchers hang a dumbbell horizontally from a fiber and measure the weight's twist when it is near objects of accurately known mass. In 1995, however, Kazuaki Kuroda of the University of Tokyo pointed out that mechanical stresses in the. twisting fibers can introduce errors.

Washington researchers Jens H. Gundlach and Stephen M. Merkowitz have used lasers to monitor the twisting of a pendulum as its housing slowly rotated near up to eight heavy spheres (SN: 5/18/96, p. 319). The experimenters sped or slowed the rotation to precisely counter any torsion of the fiber. "Since we don't twist the fiber, we do not have the uncertainty," Gundlach says.

"It's an excellent experiment," comments Riley D. Newman of the University of California, Irvine.

From the turntable accelerations, the team calculated a new, preliminary value for G: 6.6742 x [10.sup.-11] [meter.sup.3]/kilogram-[second.sup.2]. Its uncertainty is 0.0015 percent--a mere hundredth of CODATA's uncertainty and a tenth that of the next most precise experiment. But the new figure does little to reduce G's overall error bars because no one has found a flaw in the 1994 result, Gundlach says.

Probing gravity at short distances, Eric G. Adelberger, Blayne R. Heckel, and other Washington researchers, including Gundlach, suspended a top hat--shaped pendulum with holes in its brim over a rotating metal disk with identical holes in its outer edge. The disk's rotation created a time-varying gravitational tug on the pendulum that peaked when the holes aligned, Heckel explains. A metal membrane between the pendulum and disk screened electric forces.

At pendulum-disk spacings down to 0.2 mm, the team found that Newton's law, predicting a force inversely proportional to the square of the distance between two masses, holds firm.

While apparently confining extra dimensions to less than 0.2 mm, the new result "does not by any means rule anything out," argues Nima Arkani-Hamed of Lawrence Berkeley (Calif.) National Laboratory. He and coworkers had predicted deviations somewhere between about 1 mm and 10 micrometers.