Every major galaxy near the Milky Way, except Andromeda, is moving away from us. This has been known for decades. In a universe dominated by gravity, nearby galaxies should show complex motions — some falling inward, some orbiting, some receding. Instead, the local Hubble flow is quiet. Galaxies within about 30 million light-years match the cosmic expansion rate more closely than expected, as if the Local Group's gravity were somehow weaker than its mass implied.
Wempe, Helmi, and colleagues (Nature Astronomy, January 2026) found the geometry that explains it. Using Bayesian reconstruction of the initial density field from the cosmic microwave background — a method called BORG — they simulated the evolving mass distribution around the Milky Way. The result: the Local Group sits embedded in a sheet of dark matter stretching at least 32 million light-years. The sheet's central plane has roughly twice the cosmic average density. Above and below it lie substantially underdense voids.
A sheet does not pull like a sphere. A spherical mass distribution draws objects inward from all directions equally. A flattened distribution pulls objects perpendicular to the plane — compressing matter into the sheet — but exerts far less force along its surface. Galaxies embedded in the sheet can flow outward along the plane with little gravitational resistance, while the voids above and below provide nothing to slow them. The galaxies are not fleeing the Local Group. They are sliding along a surface that gravity built but does not strongly constrain in the lateral direction.
The quiet local Hubble flow was not anomalous. It was the expected behavior of objects embedded in a planar mass distribution, misinterpreted because the standard assumption was a roughly spherical local environment. The motions looked strange only against the wrong geometry.
The structural observation: the mystery was generated by the model, not by the data. The galaxies' recession rates were measured correctly. The Local Group's mass was estimated correctly. The error was in the shape assumed for the mass distribution. A sphere of the same mass would produce the inward pull that was expected and not observed. A sheet of the same mass produces the anisotropic gravitational field that matches the observations. The answer was not new physics or missing mass but the correct geometry — and the correct geometry was invisible from inside, because an observer embedded in a sheet cannot measure its extent or orientation without simulating the external view. The structure that explained the puzzle was the one that contained the observer.