In 1859, French astronomer and mathematician Urbain Le Verrier detected something strange: Mercury deviated in its dance around the Sun, defying the orderly precession predicted by Newtonian physics.

This odd anomaly couldn't be explained by unknown planets tugging at Mercury's orbit; only by physicist Albert Einstein's 1915 general theory of relativity, which describes how gravity creates curves in the fabric of space-time.

Einstein's general theory has held strong in the century since, but there are a few things about the Universe his mind-bending model can't explain. It breaks down in the centers of black holes and at the dawn of the Universe, for example, and doesn't fit very easily with quantum mechanics, leading some physicists to ponder alternative takes on how gravity works.

While those ideas remain fringe theories, the discovery of gravitational anomalies in widely separated twin stars at infinitesimally low acceleration is once again challenging Einstein's general theory.

In a new study, astrophysicist Kyu-Hyun Chae of Sejong University in Korea has analyzed nearly 2,500 wide binary star systems observed by European Space Agency's Gaia space telescope, arriving at the conclusion that standard gravity is breaking down at certain points within them.

Chae first reported finding gravitational anomalies midway through 2023 in a study of the orbital motions of wide binaries, anomalies which he thought represented evidence of one theory of modified gravity, called modified Newtonian dynamics (MOND).

However, some physicists disagreed and instead suggested his sample had been 'contaminated' by pull of undetected close companions in the binary star systems. In other words, the larger-than-expected accelerations Chae observed in some wide binaries were more likely the effect of interlopers lurking in the shadows Chae had missed.

So the Sejong University physicist sought to test his methods again in a smaller, refined subset of 'pure' binary stars. Chae found that closely orbiting twin stars were behaving consistently with classical Newtonian dynamics, so no problems there.

But binary stars separated by more than 2,000 astronomical units appeared to get a velocity 'boost' at low accelerations, inconsistent with what classical mechanics predicts and regardless of whether hypothetical dark matter was included in the models.

"This gravitational anomaly implies a low-acceleration breakdown of both Newtonian dynamics and general relativity and so has immense implications for astrophysics, cosmology, and fundamental physics," Chae writes in his new paper.

"Thus, one cannot overemphasize the importance of confirming the claimed anomaly from as many independent studies as possible."

While two studies from the same researcher are light-years away from the independent verification theory-overturning results demand, Chae thinks his methods are solid. Although he does admit that theoretical interpretations of the reported anomaly are "wide open."

However, he also makes some big claims in his paper such as "the dark matter paradigm seems now doomed to be abandoned" and that "standard cosmology based on general relativity seems no longer valid, even in principle."

Those types of claims need ridiculously strong evidence to back them up, replicated multiple times over. Chae's paper will no doubt be scrutinized closely by his peers. Nonetheless, it's in findings like this that we might find a way to bridge our gaps in knowledge over gravity's remaining mysteries.

"The evidence for the gravity boost in the low-acceleration regime is now clear enough," Chae writes, "although the scientific community should keep gathering further evidence from future observations."

The study has been published in The Astrophysical Journal.