A Third Galaxy Missing Dark Matter along a Trail of Galaxies in the NGC 1052 Field

Abstract

While most dwarf galaxies are strongly dark matter dominated, two remarkable objects in the NGC 1052 field, DF2 and DF4, appear to lack dark matter. DF2 and DF4 were recently found to be part of a trail of low luminosity galaxies that follow a linear relation between their position on the trail and their radial velocity. If the other galaxies on this trail formed together with DF2 and DF4, e.g., from gas that was separated from dark matter through a 'bullet dwarf' collision, they may lack dark matter as well. Here we constrain the dark matter content of DF9, the galaxy on the trail that most closely resembles DF2 and DF4. Using Keck/KCWI absorption line spectroscopy we find that DF9's stellar velocity dispersion is $6.4^{+4.0}_{-4.3}$ km s$^{-1}$. This is consistent with the $8.3^{+0.9}_{-1.4}$ km s$^{-1}$ dispersion that is expected from DF9's $1.4\times 10^8$ M$_\odot$ stellar mass alone, and we conclude that -- like DF2 and DF4 -- dark matter is not required to explain the kinematics of DF9. The dispersion is far below the $27\pm3$ km s$^{-1}$ expected if DF9 fell on the stellar mass--halo mass relation. Our results are further evidence that the trail of low mass galaxies in the NGC 1052 field formed together in a unique galaxy formation channel, and are consistent with the prediction of the bullet dwarf scenario that other trail galaxies should show the same lack of dark matter as DF2 and DF4.

Figures (6)

• Figure 1: HST/ACS image of DF9 shown below a DECaLS image of the NGC 1052 field. The kinematically-connected trail of galaxies, including DF2 and DF4, is indicated in blue boxes. The field-of-view for KCWI pointings is shown in red, and the two clusters with known radial velocities are circled in yellow. Both images are oriented along the trail axis 2025ApJ...988..165K.
• Figure 2: The extracted 1D spectrum of DF9 (black), with two fit templates, each with ages and metallicities taken from the central results of past literature (2025ApJ...978...21T in blue, 2023MNRAS.524.2624G in green), over-plotted. Flux is given in units relative to the continuum, and regions with strong sky line residuals have been masked (dashed lines), with the sky spectrum plotted in the bottom panel (red).
• Figure 3: Velocity dispersion as measured from diffuse stellar light (green) and globular clusters (blue) as compared to that expected for a normal dark matter halo (black) and the stars alone (orange), for DF2 (top), DF4 (middle), and -- new in this work -- DF9 (bottom). All three galaxies have velocity dispersions consistent with that expected from their stars alone.
• Figure 4: A comparison between DF2, DF4, and DF9 (filled red circles) and a wider population of Local Group (open brown circles) and Virgo Cluster dwarf galaxies (open blue circles), both in terms of measured velocity dispersion versus size (left, with stellar mass indicated by markersize in logscale), and inferred halo mass versus stellar mass (right, with unreliable halo mass estimates given in light shading). Galaxies with similar velocity dispersions to DF2, DF4, and DF9 tend to have $>$100$\times$ lower stellar masses and 2$-$6$\times$ smaller radii, while galaxies with similar stellar masses tend to have $>$100$\times$ more massive dark matter halos. Even at the upper-bounds of their uncertainties, DF2, DF4, and DF9 fall below the universal baryon fraction (gray line); galaxies should not lie in this region, even if they convert all their baryons into stars.
• Figure 5: A demonstration of how the radial velocities of two globular clusters alone can indicate the presence of dark matter in a galaxy in the same manner that 1983ApJ...266L..11A used to show the Draco dwarf spheroidal contains dark matter. This is markedly not the case for DF9.