Physicists at the US National Institute of Standards and Technology (NIST) have reported a new measurement of the gravitational constant, G, after a decade-long experiment designed to limit unconscious bias. The team found a value of 6.67387×10−11 m3/kg/s2, a result that differs from the 2007 measurement by the International Bureau of Weights and Measures (BIPM) by 0.0235%.
The gap is approximately 7.3 standard deviations, exceeding the uncertainty estimates attached to the two results. It leaves researchers with an unresolved question that has persisted for over 225 years: whether difficult experiments contain an unrecognised source of error, or whether the disagreement points to something unexpected in gravity itself.
G governs the strength of gravity between masses. It is among physics' fundamental constants, yet remains the least precisely known. The discrepancy is small enough that it does not threaten the use of G in most calculations. Its importance lies in the challenge it presents to efforts to establish a reliable value for a basic quantity in physics.
The NIST experiment was led by physicist Stephan Schlamminger and drew on the approach used in the 2007 BIPM measurement. Its torsion balance used eight cylindrical metal masses, a design with roots in Henry Cavendish's landmark 1798 experiment, which first determined G experimentally.
The apparatus measured the gravitational torque produced by the masses. NIST used an electrostatic force to counter that torque on the torsion balance. The team also carried out measurements using copper masses and sapphire masses. It found virtually identical results with both materials, reducing the prospect that the choice of mass material caused the difference.
Researchers also built a safeguard against expectations influencing the analysis. Patrick Abbott introduced a concealed variable into the calculations before the experiment's result was known. He subtracted a number, known only to himself, from the weights of select masses.
The correction was sealed in an envelope while the team completed its work. Schlamminger opened the envelope at the annual Conference on Precision Electromagnetic Measurements in Aurora, Colorado, on July 11, 2024. He described the correction factor inside as “large and negative”.
That blinded process was intended to prevent researchers from adjusting choices in ways that might unconsciously steer a result towards an expected value. Once the concealed correction was revealed, the team could produce its final measurement. Its report was later published in Metrologia in 2026.
The new result adds to a field in which separate experimental campaigns have repeatedly produced values that fail to align as closely as their reported uncertainties suggest. Researchers say the discrepancy could arise from experimental errors that have yet to be identified in one or more measurements, including the NIST or BIPM work. The difference from another prominent result also exceeds what the stated uncertainties would predict.
There is no established explanation for the conflict. It remains unknown which measurement, if either, is closer to the true value of G. Physicists say progress will depend on further measurements using independent methods, which could identify a hidden systematic effect or show whether the disagreement persists across different experimental designs.
A continuing mismatch would raise a more fundamental possibility: that present experiments are exposing an aspect of gravity that is not yet understood. For now, the NIST measurement has sharpened a long-running scientific puzzle rather than resolved it.
Comments (0)