In 2016, astronomers led by Pieter van Dokkum of Yale University published a bombshell paper claiming the discovery of a galaxy so dim, yet so broad and heavy, that it must be almost entirely invisible. They estimated that the galaxy, dubbed Dragonfly 44, is 99.99% dark matter.
A heated debate ensued about Dragonfly 44’s properties that remains unresolved. Meanwhile, more than 1,000 similarly big but faint galaxies have turned up.
Dragonfly 44 and its ilk are known as ultra-diffuse galaxies (UDGs). While they can be as large as the largest ordinary galaxies, UDGs are exceptionally dim — so dim that, in telescope surveys of the sky, “it’s a task to filter out the noise without accidentally filtering out these galaxies,” said Paul Bennet, an astronomer at the Space Telescope Science Institute in Baltimore. The bright star-forming gas that’s abundant in other galaxies seems to have vanished in UDGs, leaving only a skeleton of elderly stars.
Their existence has caused a stir in galactic evolutionary theory, which failed to predict them. “They didn’t turn up in simulations,” van Dokkum said. “You have to do something special to make a galaxy that big and faint.”
Wild new theories have emerged to explain how Dragonfly 44 and other UDGs came about. And these giant smudges of light may be providing fresh evidence of dark matter’s invisible hand.
Too Much Dark Matter
As gravity brings clumps of gas and stars together, their combined energies and momentums cause the mashup to inflate and rotate. Eventually a galaxy emerges.
There’s just one problem. As galaxies rotate, they should come apart. They don’t appear to have enough mass — and thus gravity — to stick together. The concept of dark matter was invented to provide the missing gravity. In this picture, a galaxy sits inside a larger conglomeration of nonluminous particles. This dark matter “halo” holds the spinning galaxy together.
One way to estimate a galaxy’s rotation speed, and thus its dark matter content, is by counting its spherical clusters of stars. “We don’t know why, from a theory point of view,” Bennet said, but the number of these “globular clusters” correlates closely with those harder-to-measure properties. In the 2016 paper, van Dokkum counted 94 globular clusters inside Dragonfly 44 — a number that implied an extraordinarily large dark matter halo, despite how little visible matter the galaxy has.
No one had ever seen anything like it. Van Dokkum and co-authors suggested that Dragonfly 44 could be a “failed Milky Way”: a galaxy with a Milky Way-size dark matter halo that underwent a mysterious event early on that robbed it of its star-forming gas, leaving it with nothing but aging stars and a giant halo.
Or No Dark Matter
The object attracted the interest of another camp of astronomers who argue that dark matter doesn’t exist at all. These researchers explain galaxies’ missing gravity by tweaking Newton’s law of gravity instead, an approach called modified Newtonian dynamics, or MOND.
According to MOND, the modified gravitational force for each galaxy is calculated from the mass-to-light ratio of its stars — their total mass divided by their luminosity. MOND theorists do not speculate as to why the force would depend on this ratio, but their ad hoc formula matches the observed speeds of most galaxies, without the need to invoke dark matter.
When news broke about Dragonfly 44, MOND advocate Stacy McGaugh, an astronomer at Case Western Reserve University, calculated from its mass-to-light ratio that it should rotate more slowly than van Dokkum’s initial estimate indicated. The MOND calculation didn’t seem to fit the data.
But then in 2019, van Dokkum’s group downgraded Dragonfly 44’s rotation speed using improved data. MOND was vindicated. “Dragonfly 44 is an example of how these data evolve to agree with MOND,” said McGaugh.
Still, for the majority of astronomers, who believe in dark matter, the slower rotation speed just implied that Dragonfly 44’s halo is smaller than they thought. In 2020, an independent group further downsized the halo by counting dramatically fewer globular clusters, but van Dokkum disputes this result. Though the halo’s size remains uncertain, it may be less massive than initially supposed, suggesting that Dragonfly 44 isn’t a failed Milky Way after all.
Big Old Galaxy
A newly discovered oddity has compounded the mystery.
In a paper published in August, van Dokkum’s group found Dragonfly 44 to be extremely ancient, having formed between 10 billion and 13 billion years ago.
But such an old galaxy should not be as large as Dragonfly 44 is. Early-universe objects tend to be more compact because they formed before the universe’s rapid expansion.
Moreover, such an old, threadbare galaxy should have been completely torn apart by now. That Dragonfly 44 has held together implies that it has a hefty dark matter halo after all — potentially restoring the “failed Milky Way” hypothesis. “That’s a really fun explanation, so that’s why I like it, but I don’t know if it’s right,” said van Dokkum.
Another explanation, the “high spin” hypothesis, posits that two small galaxies merged while rotating in the same direction, such that the resulting galaxy, Dragonfly 44, acquired the angular momentum of both. This caused it to rotate more quickly, puffing it out and blowing out its star-making material.
Dazzlingly Diverse UDGs
Amid the scrutiny of Dragonfly 44, astronomers have also cataloged a vast and diverse collection of other ultra-diffuse galaxies. The findings are forcing them to conclude that galaxies form in more ways than they knew.
Some newfound UDGs seem to lack dark matter entirely. Van Dokkum’s group identified one such galaxy in 2018, then spotted a trail of others nearby. This May, the team conjectured in Nature that the trail formed in a long-ago collision of two galaxies. The collision slowed down the flow of the galaxies’ gas, but their dark matter kept going as if nothing had happened. The gas then compressed into clumps of stars, eventually forming a string of dark matter-free galaxies.
Meanwhile, Bennet discovered two UDGs in 2018 that point to a different formation theory. In each case, tidal forces from a heavy nearby galaxy seem to have ripped through the UDG, puffing it out and stealing its gas. (This can’t explain Dragonfly 44, which sits too far from heavy galaxies.)
Puzzlingly, a September paper reported recent star formation in a UDG, contradicting the idea that they only harbor old stars.
Such a range of UDGs that look the same outwardly but differ internally may validate dark matter theory over MOND. “If the stars are moving very fast in one galaxy, and very slowly in the other, that’s a big problem for those alternative theories,” van Dokkum said.
McGaugh agreed that if there are “genuine outliers” among the UDG population, “that is indeed a problem for MOND.” However, he added, “that doesn’t automatically make dark matter a better interpretation.”
Definitive answers will require new telescopes. The newly operational James Webb Space Telescope has already spotted distant galaxies as they appeared when they were forming in the early universe, which will help test and refine the nascent ideas.
“The big takeaway is that we still don’t know what’s out there,” van Dokkum said. “There are galaxies that we haven’t discovered that are very big, very close by, and have unusual properties, and they are not in our current catalogs even after all these decades of studying the sky.”