Astronomers using the Hubble Space Telescope, the Euclid observatory, and the Subaru Telescope have identified a ghostly galaxy in the Perseus cluster that is roughly 99% dark matter, with visible stars contributing just 1% of its total mass. Dubbed Candidate Dark Galaxy-2, or CDG-2, the object sits about 300 million light-years from Earth and was found through a method never before used to detect a galaxy, tracking a tight grouping of four globular clusters that betrayed the presence of an otherwise invisible structure. The discovery sharpens a long-running debate about whether dark matter truly exists or whether modified theories of gravity can explain the cosmos without it.
A Galaxy Found by Its Star Clusters Alone
CDG-2 is extraordinarily dim. Its total luminosity is equivalent to about 6 million Suns, a tiny fraction of what even a modest galaxy typically produces. Roughly 16% of that feeble glow comes from the four globular clusters themselves, meaning the remaining starlight is spread across an extremely faint diffuse halo barely distinguishable from the cosmic background. According to the peer-reviewed analysis published in The Astrophysical Journal Letters, CDG-2 qualifies as a gravitationally bound galaxy detected entirely through its globular cluster population, marking a first in observational astronomy and demonstrating that star clusters can serve as beacons for almost invisible systems.
The European Space Agency described the detection sequence in its institutional release: observers first noticed the unusually tight grouping of four globular clusters in the Perseus cluster, then used deep imaging from all three telescopes to tease out the faint extended emission surrounding them. That glow confirmed the clusters were not free-floating but embedded in a real, if almost invisible, galaxy. The fact that stars supply only 1% of CDG-2’s mass while the rest is inferred to be dark matter makes it one of the most dark-matter-dominated objects ever cataloged, and its location in a massive cluster offers an unusually clean laboratory for testing how dark matter behaves in dense environments.
Why Modified Gravity Struggles With Dark Galaxies
Modified Newtonian Dynamics, commonly called MOND, was proposed decades ago as an alternative to dark matter. Instead of invoking unseen particles, MOND tweaks the equations of gravity so that the observed motions of stars and gas in galaxies can be explained without extra mass. The approach works surprisingly well for individual rotating galaxies, where flat rotation curves follow from the modified force law. But galaxy clusters have long been a weak spot. Foundational theoretical work on how MOND-like modifications relate to gravitational lensing signatures showed that clusters bend light more strongly than modified gravity alone predicts, implying that some form of unseen mass is still required even under MOND.
CDG-2 intensifies this problem. A galaxy that is 99% dark matter by mass and was identified solely because its globular clusters revealed a bound structure demands an enormous reservoir of invisible material holding everything together. MOND proponents have sometimes proposed that ordinary neutrinos with masses around one electron-volt could supply the missing cluster mass. But a dedicated analysis combining strong and weak lensing data from CLASH clusters found that the phase-space constraints on such neutrinos are inconsistent with observational requirements. In other words, even the most popular patch for MOND at cluster scales does not hold up under precise lensing measurements, and CDG-2 adds yet another data point that is hard to reconcile without invoking a genuine dark matter component that dominates the galaxy’s gravitational potential.
Cluster Lensing Keeps Tightening the Vise
Separate from CDG-2, a recent synthesis used gravitational lensing-derived mass profiles, including strong and weak shear and magnification, for massive CLASH clusters to quantify MOND’s residual missing-mass problem. The results showed that even after applying modified gravity corrections, clusters still require substantial additional unseen mass that MOND cannot account for internally. This line of evidence has been building for years: as lensing reconstructions grow more detailed, the discrepancy between the visible matter and the total gravitational mass becomes harder to explain by tweaking gravity alone, especially on the largest bound scales in the universe.
CDG-2 slots neatly into this broader picture. The galaxy resides in a cluster environment where lensing already demands extra mass on large scales, and now its internal dynamics and extreme mass-to-light ratio point in the same direction on smaller, galactic scales. In effect, the cluster’s lensing map and the dark galaxy’s structure both testify to a dominant non-luminous component. For theorists, this dual constraint means any alternative to particle dark matter must simultaneously reproduce the cluster-wide lensing signal and the existence of hyper–dark-matter-dominated galaxies like CDG-2, a combined challenge that has so far proven formidable.
Echoes of NGC 1052-DF2 and the Dark Matter Debate
CDG-2 is not the first galaxy to challenge assumptions about dark matter’s distribution. In 2018, Hubble observations of NGC 1052-DF2 revealed a galaxy that appeared to lack dark matter almost entirely, with its formation possibly linked to powerful winds from a young black hole or tidal interactions that stripped away its invisible halo. That finding was initially controversial, but subsequent Hubble measurements produced a more accurate distance estimate for the galaxy and strengthened the case that its internal motions truly reflect an unusually low dark matter content rather than a simple mismeasurement. The result suggested that dark matter can be redistributed or removed under extreme conditions, complicating the once-simple picture that every galaxy must share a similar dark-to-luminous mass ratio.
CDG-2, by contrast, sits at the other extreme of the spectrum, with stars contributing only a percent-level fraction of the mass. Together, DF2 and CDG-2 illustrate that dark matter need not track starlight in a rigid way: some galaxies can be nearly devoid of it, while others are almost entirely composed of it. For dark matter models, this diversity implies that baryonic processes, environment, and merger history can dramatically reshape how invisible mass and stars are apportioned. For modified gravity, however, the existence of both dark-matter-poor and dark-matter-rich systems within the same overarching framework of cluster lensing is harder to accommodate, because the theory must reproduce a wide range of dynamical behaviors without the flexibility of adding or subtracting unseen mass.
What CDG-2 Means for Future Dark Matter Tests
The discovery of CDG-2 arrives at a moment when large surveys and powerful telescopes are beginning to map low-surface-brightness structures in unprecedented detail. Instruments like Euclid and Subaru are designed to probe weak lensing and faint galaxies across wide areas of sky, making them ideal for uncovering additional dark matter dominated systems. As more objects like CDG-2 are found, astronomers will be able to compare their globular cluster populations, internal kinematics, and spatial distribution within clusters, building a statistical picture of how often such extreme galaxies form and what conditions favor their emergence. That, in turn, will feed back into simulations of structure formation in a universe dominated by cold dark matter.
On the theoretical side, CDG-2 underscores the importance of cross-checking dynamical inferences with independent lensing and clustering measurements, a task that relies heavily on open-access data and preprint dissemination. Platforms such as community-backed repositories allow teams to share lensing reconstructions, simulation outputs, and follow-up analyses quickly, enabling rapid testing of ideas about dark galaxies and modified gravity. As researchers refine models of how dark matter clusters and interacts (if at all) with ordinary matter, objects like CDG-2 will serve as crucial benchmarks. Whether dark matter ultimately turns out to be a new particle species or a sign that gravity itself needs revision, the ghostly galaxy in Perseus ensures that any successful theory must account for a universe where most of the mass can remain almost entirely unseen.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
