Recent proposals suggest types of inverse-square—law failures which can fit known astronomical data but which exhibit readily observable effects at laboratory dimensions. It is pointed out that there is no plausible way to infer from astronomical data that Newtonian gravitation applies in the laboratory. The experimental data are examined and it is found that past $G$ measurements in the laboratory set only very loose limits on a possible variation in $G$ and that present technology would allow considerable improvement.

• Received 27 August 1973

DOI:https://doi.org/10.1103/PhysRevD.9.850