By Curt Suplee
Washington Post Staff Writer
Monday, May 1, 2000; Page A11 

It's been more than 300 years since Isaac Newton--reportedly inspired by falling fruit--explained the force of universal gravitational attraction. Yet scientists around the globe are still trying to decide how strong it is.

 Strange as it may seem, that basic value, called the gravitational constant, is quite poorly known.

 "It's very embarrassing for modern physics," said Jens H. Gundlach of the University of Washington, who on Saturday proposed a new, slightly higher and unprecedentedly accurate value at a meeting of the American Physical Society. "It's such a fundamental constant that nature gave us, we just ought to know it."

 Worse yet, many of the latest attempts to determine the value of "Big G," as it is fondly known, disagree with one another to an outrageous extent. As a result, in 1998 the international committee that serves as the ultimate arbiter of such standards actually increased the official uncertainty in the value of G by a factor of 12, to a humiliating 1.5 parts in 1,000.

 That is definitely not close enough for government work. By contrast, the Planck constant (which defines how much energy a bit of light contains) or Avogadro's number (the number of molecules in a given quantity of a compound) are known to about eight parts in 100 million. Some benchmarks, such as the speed of light in a vacuum, are virtually exact.

 "Gravity is the most pervasive of the fundamental forces, yet it is the worst known of the fundamental constants," said Barry Taylor, manager of the fundamental constants data center at the National Institute of Standards and Technology.

 The reason is that, as fundamental forces go, gravity is a real wimp. It may not feel that way when you fall out of bed. But consider that a puny magnet the size of a pencil eraser can lift a thumbtack off a table--despite the Earth's entire 6 billion trillion tons of mass pulling the other way.

 Even in state-of-the-art labs, G is infuriatingly hard to measure. By one calculation, the gravitational attraction between two identical 400-pound lead balls whose centers are 2 feet apart is a whopping .00002 ounce. And the normal masses used to determine Big G are vastly smaller. In a typical device, Gundlach says, the total gravitational force to be detected is "only equivalent to the weight of a bacterium. And that small force must be measured very precisely."

 That's not exactly easy when a measurement can be spoiled by the motion of a single person 75 feet away, or a temperature difference of 1/10th of one degree within the apparatus, according to experts.

 And yet it has to be done. It is not enough to know, as Newton revealed, that the attraction between any two objects is proportional to the product of their masses, and inversely proportional to the square of the distance between them. To make that relationship into an equation that will yield a specific magnitude for the gravitational force, it has to be multiplied times Big G.

 In Newton's day, there were no instruments available to get close to an accurate figure. The first scientist to weigh in heavily on the subject was Newton's countryman, Henry Cavendish (1731-1810). In 1797 and '98, Cavendish used an ingenious gizmo called a torsion balance, designed by another English scientist, John Michell, in which a dumbbell-shaped weight with a small lead sphere on each end was suspended horizontally at its center by a thread. [See illustration.]

 Two larger lead spheres, the attractor masses, were placed near the lobes of the dumbbell, which then rotated toward the attractors, twisting the thread in the process.

 By measuring the amount of force that it took to twist the string, Cavendish calculated Big G to an astonishing precision--within 1 percent of the modern value. Thereafter, it was a fairly simple matter, using Newton's equations, to determine the mass of our planet. Cavendish called this "weighing the Earth."

 For the past 200 years, torsion balances have remained the favored instruments for studying Big G, although several other designs have proved useful. Some measure the distance that light travels between two reflector plates before and after they are exposed to attractor masses. Some involve dropping weights and recording their velocity. Others entail weighing a mass by itself and then in the presence of a sizable attractor.

 All of them were agreeing pretty well, and by 1986 the international Committee on Data for Science and Technology (CODATA) had published a value with an uncertainty of 0.013 percent. Not great, but certainly respectable.

 (After all, there are no day-to-day practical uses for Big G, although it does determine how brightly stars shine and, Einstein noted, how much space curves.)

 Then in 1994, after 15 years of work, the German equivalent of NIST came up with a new figure for Big G that was more than half a percent larger than the accepted value.

 "It was very credible, but it had this gross discrepancy from previous figures," said Taylor, a former chair of the CODATA committee, which in 1998 took account of the German result and others by increasing the uncertainty in Big G. "Since then, there has been a tremendous amount of activity" at a dozen labs around the world "to find out exactly what's going on."

 One of those labs is at the University of Washington, where Gundlach and colleagues recently completed a dozen three-day runs with a super-tech version of the Cavendish device (described at right). Their preliminary figures are a bit higher than the current CODATA standard, but have the lowest claimed uncertainty ever: 15 parts in a million, or 0.0015 percent.

 The present CODATA value (see www.physics.nist.gov on the Web) is 6.673 X 10-11 (that is, times 1/100,000,000,000) cubic meters per kilogram per second squared, with an uncertainty of 10 in the last two digits. Gundlach's team came up with 6.67390, with an uncertainty of 10 in the last two digits. Using that value, and satellite measurements, the Washington group has recalculated the mass of the Earth (at 5.9725 x 1021 metric tons).

 "We are confident," Gundlach stated, "that we know the mass of our home planet Earth now more precisely than it has ever been known to mankind."

 Those numbers may be revised, and the entire result is subject to peer review prior to publication. But Gundlach is optimistic that his groups's findings will agree with a number of new Big G figures expected soon. "The proof is in the pudding," he said.