A Galaxy Without (Much) Dark Matter

Astronomers have found a peculiar object: a fluffy galaxy that has little to no dark matter.

You can’t have a cup of coffee without the cup. Astronomers have long thought the same goes for galaxies and dark matter: the dark matter forms the cup, and the galaxy coalesces from the gas that pours into that cup, pulled there by gravity. ­

galaxy without dark matter

This Hubble Space Telescope image of NGC 1052-DF2 show's the galaxy's diffuse, see-through nature — it has so few stars and is so sparse, background galaxies are visible through it.
P. van Dokkum / R. Abraham / STScI

The fluffy galaxy NGC 1052-DF2 makes no sense in this picture: it’s the coffee without the cup. Using an impressive collection of ground- and space-based observations, Pieter van Dokkum (Yale) and colleagues have found that this diffuse galaxy has at most 1400 as much dark matter as expected, based on other systems of similar mass. In fact, the best explanation may be that there’s no dark matter at all.

The peculiar object is one of a collection of galaxies dominated by the beefy elliptical NGC 1052, which lies some 65 million light-years away on the border of the constellations Cetus and Eridanus. Astronomers already knew DF2 existed, as it showed up as a collection of dots in survey images. But when the team pointed the Dragonfly Telephoto Array at it, they instead saw a big blob.

Intrigued, they looked at it with several other instruments. They discovered that the galaxy is so sparse that it’s see-through, and its dots look like globular clusters, albeit strangely bright, similar to the Milky Way’s Omega Centauri cluster. They’re also about twice as large and squashed as globulars of similar brightness in our galaxy.

These ancient balls of stars loop around their parent galaxies with speeds determined by the galaxy’s total mass. Measure the globulars’ velocities, and you’ll measure the mass enclosed within their orbits — which usually turns up evidence for a whole lot of unseen matter.

Not this time. The 10 globular clusters the team clocked all move at most a third as fast as expected for galaxies of this mass (about 200 million Suns, or about a tenth that of the Small Magellanic Cloud). Their motions imply that the enclosed mass is roughly equal to the mass visible as stars.

In other words, there’s essentially no dark matter.

Dragonfly team

Roberto Abraham (left), Pieter van Dokkum (right), and several graduate students from the University of Toronto and Yale pose with half of the 48-lens Dragonfly array in New Mexico. The students are coauthors on the paper reporting that the galaxy NGC 1052-DF2 seems to contain no dark matter. There are five additional coauthors not pictured here, including another grad student.
University of Toronto

NGC 1052-DF2 is not the first tenuous galaxy to turn up. Astronomers have seen thousands of so-called ultra-diffuse galaxies (UDGs) since Van Dokkum, Roberto Abraham (University of Toronto), and their team first uncovered them with the Dragonfly array in 2014. But none of the others shows this paucity of dark matter.

How did something like this form? The team makes a couple of suggestions in the March 29th Nature, but nothing fits perfectly. For example, NGC 1052-DF2 could have formed from gas wrenched out of the nearby elliptical, which does show signs of a recent merger. But the UDG’s characteristics don’t match what astronomers would expect from that situation.

Another possibility is that DF2 has been stripped of its dark matter by the elliptical. Simulations by Alyson Brooks (Rutgers) and others indicate that when dwarfs pass through a big galaxy’s outskirts, they lose dark matter. “You can lose about 90% of the dark matter mass before any stars are stripped from the galaxy,” she says. That stripping is likely to cause the stars to “puff up,” she adds, perhaps creating an object like the diffuse, dark matter–less NGC 1052-DF2. But she’d like to know more about DF2’s trajectory before arguing the case.

Van Dokkum and his team don’t see signs of disturbance around the galaxy, which they think disfavors the flyby solution. It’s possible that any tidal features have faded away beyond detectability, but it’s unclear if there’s been enough time for that, Van Dokkum says.

That the coffee can exist without the cup indicates that the cup and coffee are both real, separate entities, the team concludes. If the presence of dark matter were only an illusion, arising because we’re using the wrong theory of gravity, then we’d always see signs of it in galaxies. But if the dark matter can sometimes be there and sometimes be absent, then dark matter exists.

Stacy McGaugh (Case Western Reserve University) isn’t so sure. He’s spent several years studying the strengths and weaknesses of modified Newtonian dynamics (MOND), a theory of gravity that does without dark matter by suggesting gravity works differently when low accelerations are involved. He agrees that NGC 1052-DF2’s globulars should move faster in MOND than they do. But he also points out that the galaxy is weird regardless of whether you accept dark matter’s existence or not.

“So yes, I find it troublesome for MOND,” he says. “But I don’t understand it any better in terms of dark matter.” Both dark matter and MOND have observational points in their favor, he explains, but he’s squeamish about upping or lowering the fraction of dark matter in any given galaxy just to make sense of stars’ motions — that’s an inference, not a prediction that can be proved or disproved, he says.

The Dragonfly team continues to look for more galaxies like NGC 1052-DF2. Of 23 other diffuse galaxies the astronomers are analyzing, three are potentially similar. With more than one galaxy in hand, astronomers might be able to say more about how these galaxies form and what they mean for dark matter.



Pieter van Dokkum et al. “A Galaxy Lacking Dark Matter.” Nature. March 29, 2018.

Pieter van Dokkum et al. “An Enigmatic Population of Luminous Globular Clusters in a Galaxy Lacking Dark Matter.” Posted to arXiv.org March 28, 2018.

4 thoughts on “A Galaxy Without (Much) Dark Matter

  1. Hendrik

    Nobody provides coordinates for this galaxy. I’ve searched through a lot of articles.
    Is this galaxy also known as PGC3097693 – LEDA 3097693?
    ICRS coord. (ep=J2000) : 02h 41m 46.5s -08° 24′ 08″ (from Simbad)


    1. Monica YoungMonica Young

      The Nature paper identifies the galaxy with coordinates right ascension α = 2 h 41 min 46.8 s, declination δ = −8° 24′ 12″. (Very close to what you found.) The large elliptical galaxy nearby, NGC 1052, is at 02h41m04.8s -08d15m21s, according to NASA’s Extragalactic Database.

  2. Howard RitterHoward Ritter

    The possibility that this object is so diffuse because it lacks the constraining gravity of a large mass of dark matter seems pretty obvious, like coffee poured onto a counter instead of into a cup. (Nice analogy, BTW. Is it original with you?) That raises the question of whether a DM deficit could be a common finding in UDGs in general. Your article makes the comment that “none of the [other UDGs studied] shows this paucity of dark matter”, but doesn’t detail whether they were amenable to, or whether they received, a comparable degree of scrutiny. I’ve been slogging my way through the Nature article online trying to understand this better, but I wonder if you know how thoroughly DM estimates have been made among other UDGs. In any case, I’d guess there are researchers around the globe who are putting together proposals to look at a *lot* more UDGs!

  3. mmaronati@aol.com

    Maybe it is not the Dark Matter proof and instead it is a missing black hole (or fainted for sporadic reasons) in the center of the NGC 1052 the reason that the galaxy is falling apart with no gravitational structure.

All comments must follow the Sky & Telescope Terms of Use and will be moderated prior to posting. Please be civil in your comments. Sky & Telescope reserves the right to use the comments we receive, in whole or in part, and to use the commenter’s username, in any medium. See also the Terms of Use and Privacy Policy.