Gravity's force: chasing an elusive element.

Determining the values of fundamental physical constants has long served as a test of both physical theory and measurement technology. Now, experiments by three independent groups have produced values for the strength of the gravitational force (G) that disagree significantly with the currently accepted number and with each other (see table).

The teams involved in these experiments reported their results at last week's American Physical Society meeting, held in Washington, D.C.

"Each of these groups has done a careful job, but G is an extremely hard number to nail down," says George T. Gillies of the University of Virginia in Charlottesville.

The problem stems from the fact that gravity is much weaker than the other forces of nature. Moreover, it's always attractive, and there's no way to shield an experiment from external gravitational influences.

Mark P. Fitzgerald and his coworkers at the Measurement Standards Laboratory in Lower Hutt, New Zealand, used a torsion balance, which consists of a copper bar dangling horizontally from a thin, tungsten wire fastened to the bar's midpoint (SN: 3/12/94, p.376).

Normally, the gravitational force of two large masses brought near the bar's ends causes it to rotate, slightly twisting the wire. To avoid complications due to wire twisting and bar oscillations, the researchers applied an electric field to attract the copper bar and balance the

gravitational force, keeping the bar still at all times. By carefully measuring the compensating electric force, they could deduce the gravitational force and calculate G.

Winfried Michaelis and his group at the Physikalisch-Technische Bundesanstalt in Braunschweig, Germany, also used a torsion balance in which an electric force compensates for the gravitational force between pairs of masses (see diagram). In this case, the researchers eliminated the complicating effects of the wire by hanging metal cylinders from either end of a beam and letting the beam's support float in a mercury bath.

Hinrich Meyer and his colleagues at the University of Wuppertal in Germany adopted a different approach. This group used microwave technology to measure minute changes in the oscillation frequencies of a pair of pendulums disturbed by the movement of large masses in their vicinity.

The fact that these three, carefully performed experiments give different results "is truly a scientific mystery," comments Eric G. Adelberger of the University of Washington in Seattle. However, it doesn't necessarily imply that the physical theory is faulty.

"Each apparatus has its own idiosyncrasies," Gillies notes. "The

differences among these numbers probably can be explained by extraneous gravitational effects that weren't properly accounted for." But no such confounding factors have yet been identified.

It remains unclear in what way the accepted value of G will change to reflect the new findings. Meanwhile, Gabriel G. Luther and his colleagues at the Los Alamos (N.M.) National Laboratory are preparing for a new measurement of this elusive constant.