Keck Spectroscopy Finds Third Dark-Matter–Deficient Galaxy in NGC 1052 Field

A new spectroscopy study using the W. M. Keck Observatory reports that NGC 1052-DF9, or DF9, is a third galaxy in the NGC 1052 field whose internal stellar motions can be explained without invoking a normal dark matter halo. In an arXiv preprint revised June 15, 2026, astronomers Michael A. Keim, Pieter van Dokkum, Zili Shen, Shany Danieli and Imad Pasha say DF9 appears, like the better-known galaxies DF2 and DF4, to be dark-matter-deficient.

From integrated starlight measured with Keck’s Cosmic Web Imager, the authors report: “Using Keck/KCWI we find that DF9's stellar velocity dispersion is 6.5^{+3.9}{-4.3} km s^{-1}.” Velocity dispersion is a measure of how fast stars move inside a galaxy and is used to infer how much mass is present. Keim and colleagues say that value is consistent with the 8.3^{+0.9}{-1.4} km/s expected from DF9’s stars alone, for a stellar mass of 1.4 × 10^8 times the mass of the sun. It is far below the 24 ± 3 km/s they say would be expected if DF9 sat inside a more typical dark matter halo of about 1.4 × 10^10 solar masses.

That matters because DF9 is not just another oddball dwarf galaxy. The paper places it on the same proposed “trail” of faint galaxies in the NGC 1052 field as DF2 and DF4, following a linear relation in both position and radial velocity. If three galaxies along that same structure all show unusually low internal motions, the case grows stronger that they may share a common origin rather than representing isolated anomalies.

The question has drawn attention because dwarf galaxies are usually thought to be strongly dominated by dark matter. DF2 was first reported in 2018 as a galaxy whose motions could be explained largely by its stars alone, and DF4 followed in 2019. In 2022, researchers proposed that both might belong to a trail of galaxies created in a rare collision event dubbed the “bullet dwarf” scenario. In that picture, a high-speed crash between gas-rich dwarf galaxies separates gas from dark matter and leaves behind a trail of newly formed galaxies with relatively little dark matter. Keim and colleagues interpret DF9 as further evidence consistent with that idea.

But the new result comes with important caution. It is a preprint in an area where both distances and very low-dispersion kinematic measurements remain under close scrutiny. Measuring such small velocity dispersions is technically difficult, and DF9’s uncertainty bars are relatively large.

Distance, in particular, has remained a live issue in the NGC 1052 field. A separate June 2026 preprint by Tang and colleagues used Hubble Space Telescope surface-brightness-fluctuation distances for several trail candidates and a James Webb Space Telescope tip-of-the-red-giant-branch, or TRGB, distance for DF2. As Tang and co-authors wrote: “We find that the dwarfs are all at ~20 Mpc, and are not associated with the foreground NGC 1035 group. However, for DF2, we derive an SBF distance of 17.7±1.4 Mpc, inconsistent with the published HST TRGB distance (21.7±1.2 Mpc). Meanwhile, JWST observations of DF2 offer a second, and potentially more accurate, TRGB distance of 17.6±0.6 Mpc.”

Those revisions matter because changing a galaxy’s distance changes the inferred stellar and dynamical masses. Keim and colleagues say in their appendix that DF9’s immediate result is not sensitive to the assumed distance of the trail, but the broader interpretation of the group still depends in part on where these galaxies sit in space.

For now, the DF9 measurement adds another piece of evidence for the idea that DF2, DF4 and DF9 belong to the same unusual structure. But because this is still a preprint-era result in a technically demanding corner of astronomy, additional confirmation of both distances and kinematics across the group will shape how broadly researchers interpret it.