The original version of this story appeared in Quanta Magazine.
Late in the evening of October 5, 1923, Edwin Hubble sat at the eyepiece of the Hooker telescope at the Mount Wilson Observatory atop the mountains overlooking the Los Angeles basin. He was observing an object in the northern sky. To the unaided eye, it was visible as a faint smudge. But through a telescope it sharpened into a brilliant ellipse called the Andromeda Nebula. To settle a debate about the size of the Milky Way—which was then thought to be the entire universe—Hubble needed to determine Andromeda’s distance from us.
In the telescope’s field of view, Andromeda was a giant. Hubble patiently captured several exposures covering many glass photographic plates, and in the early hours of October 6, he made a 45-minute exposure on a small glass plate and scrawled “N” where he saw three new stars, or novas. But when he compared his image with photographs captured by other astronomers, he realized that one of his new novas was actually a Cepheid variable star—a type of star that can be used to measure astronomical distances.
He scratched out one “N” and wrote “VAR!”
Hubble used this pulsating star to calculate that Andromeda was 1 million light-years from Earth, a distance much greater than the diameter of the Milky Way (he was slightly off; Andromeda is about 2.5 million light-years away). And he realized that Andromeda was no mere nebula but an entire “island universe”—a galaxy distinct from our own.
With the cleaving of the cosmos into a home galaxy and a larger universe, the study of our finite home—and how it exists within that universe—could begin in earnest. Now, a century later, astronomers are still making unexpected discoveries about the only cosmic island we’ll ever inhabit. They may be able to explain some of the Milky Way’s characteristics by reimagining how it formed and grew in the early universe, by scrutinizing its uneven shape, and by studying its ability to form planets. The latest results, amassed over the past four years, are now painting a picture of our home as a unique place, at a unique time.
We have been lucky, it seems, to live near a particularly quiet star on the calm fringes of a middle-aged, oddly tilted, loosely spiraling galaxy that has been largely left alone for most of its existence.
From the Earth’s surface—if you are somewhere very dark—you can only see the bright stripe of the Milky Way’s galactic disk, edge-on. But the galaxy we live in is so much more complicated.
A supermassive black hole churns at its center, surrounded by the “bulge,” a knot of stars containing some of the galaxy’s oldest stellar denizens. Next comes the “thin disk”—the structure we can see—where most of the Milky Way’s stars, including the sun, are partitioned into gargantuan spiraling arms. The thin disk is encased in a wider “thick disk,” which contains older stars that are more spread out. Finally, a mostly spherical halo surrounds these structures; it is mostly made of dark matter but also contains stars and diffuse hot gas.
To make maps of these structures, astronomers turn to individual stars. Each star’s composition records its birthplace, age, and natal ingredients, so studying starlight enables a form of galactic cartography—as well as genealogy. By situating stars in time and place, astronomers can retrace history and infer how the Milky Way was built, piece by piece, over billions of years.
The first major effort to study the primordial Milky Way’s formation began in the 1960s, when Olin Eggen, Donald Lynden-Bell and Alan Sandage, who was Edwin Hubble’s former graduate student, argued that the galaxy collapsed from a spinning gas cloud. For a long time after that, astronomers thought that the first structure to emerge in our galaxy was the halo, followed by a bright, dense disk of stars. As more powerful telescopes came online, astronomers built increasingly precise maps and started refining their ideas about how the galaxy came together.
Everything changed in 2016, when the first data from the European Space Agency’s Gaia satellite came back to Earth. Gaia precisely measures the paths of millions of stars throughout the galaxy, allowing astronomers to learn where those stars are located, how they move through space, and how fast they are going. With Gaia, astronomers could paint a sharper picture of the Milky Way—one that revealed many surprises.
The bulge is not spherical but peanut-shaped, and it’s part of a larger bar spanning the middle of our galaxy. The galaxy itself is warped like the brim of a beat-up cowboy hat. The thick disk is also flared, growing thicker toward its edges, and it may have formed before the halo. Astronomers aren’t even sure how many spiral arms the galaxy really has.
The map of our island universe is not as neat as it once seemed. Nor as calm.
“If you look at a traditional picture of the Milky Way, you have this nice spherical halo and a nice regular-looking disk, and everything is kind of settled and stationary. But what we know now is that this galaxy is in a state of disequilibrium,” said Charlie Conroy, an astronomer at the Harvard-Smithsonian Center for Astrophysics. “This picture of it being simple and well ordered has been really tossed out in the past couple of years.”
Three years after Edwin Hubble realized Andromeda was a galaxy unto itself, he and other astronomers were busy imaging and classifying hundreds of island universes. Those galaxies seemed to exist in a few prevailing shapes and sizes, so Hubble developed a basic classification scheme known as the tuning fork diagram: It divides galaxies into two categories, ellipticals and spirals.
Astronomers still use this scheme to categorize galaxies, including ours. For now, the Milky Way is a spiral, with arms that are the main nurseries for stars (and therefore planets). For a half-century, astronomers thought there were four main arms—the Sagittarius, Orion, Perseus, and Cygnus arms (we live in a smaller offshoot, unimaginatively called the Local Arm). But new measurements of supergiant stars and other objects are drawing a different picture, and astronomers no longer agree on the number of arms or their sizes, or even whether our galaxy is an oddball among islands.
“Strikingly, almost no external galaxies present four spirals extending from their centers to their outer regions,” Xu Ye, an astronomer with China’s Purple Mountain Observatory, said in an email.
To trace the Milky Way’s spiral arms, Ye and colleagues used Gaia and ground-based radio telescopes to look for young stars. They found that, like other spiral galaxies, the Milky Way only has two main arms, Perseus and Norma. Several long, irregular arms also wind around its core, including the Centaurus, Sagittarius, Carina, Outer, and Local arms. It seems that, at least in shape, the Milky Way may be more similar to distant cosmic islands than astronomers thought.
“Studying the spiral-shaped Milky Way may reveal whether it is unique among the billions of galaxies in the observable universe,” Ye wrote.
Hubble’s study of Andromeda and its variable star stemmed from his fierce rivalry with another famed astronomer at Mount Wilson, Harlow Shapley. The Harvard astronomer Henrietta Swan Leavitt had pioneered the use of Cepheid variable stars to measure distances, and using her method Shapley had calculated that the Milky Way was 300,000 light-years across—an astonishing claim in 1919, when most astronomers believed the sun was at the galaxy’s center, and that the whole galaxy spanned 3,000 light-years. Shapley thus insisted that other “spiral nebulae” must be gas clouds and not separate galaxies because their sizes would mean they were inconceivably far away.
Hubble in turn wrote up his variable star measurements and convinced everyone that Andromeda was indeed a separate galaxy. “Here is the letter that destroyed my universe,” Shapley reportedly said after seeing Hubble’s data.
In terms of astronomical distances, however, Shapley may not have been that far off. In the intervening century, astronomers have calculated that the Milky Way’s bulge is about 12,000 light-years across, that the disk spans 120,000 light-years, and that the halo of dark matter and ancient star clusters extends hundreds of thousands of light-years in every direction.
A recent observation found that some halo stars are scattered as far as 1 million light-years away—halfway to Andromeda—which suggests that the halo, and therefore the galaxy, is not quite an island universe unto itself.
Astronomers led by Jesse Han, a graduate student at the Harvard-Smithsonian Center for Astrophysics, recently determined that the stellar halo is not spherical, as was long assumed, but shaped like a football. In work published on September 14, Han and his team also showed that the dark matter halo might be tilted by about 25 degrees, causing the entire galaxy to look warped.
And while that might seem weird enough, the tilt itself may be evidence of the Milky Way’s violent past.
Eons before Hubble sat at the eyepiece, ages before the sun was born, long before the Milky Way existed, the Big Bang wrenched apart all matter and indiscriminately scattered it throughout the newborn cosmos. The first galaxies eventually formed from bits of random detritus, starting a 13-billion-year sequence that led to us. Astronomers debate the intricacies of how those events unfolded, but they know that the galaxy we now inhabit grew through a complex process that included mergers and acquisitions.
Throughout the universe, galaxies collide and combine in unimaginably humongous calamities. The telescope named for Edwin Hubble captures these cosmic pileups all the time. And even though it’s relatively placid today, the Milky Way is no exception: By sifting through the archaeological records kept by stars, streams of gas, so-called globular clusters of thousands to millions of stars, and even the shadows of devoured dwarf galaxies, scientists are learning more about how the Milky Way evolved.
The first hints of violence came when astronomers peering through the storied 200-inch telescope at Palomar Observatory (which Hubble was the first to use) found evidence in 1992 that the Milky Way was ripping apart some of the globular clusters in its halo. The Sloan Digital Sky Survey confirmed that observation, and radio telescopes later found that the galaxy was also inhaling streams of nearby gas.
By mid-2018, astronomers figured that the Milky Way had merged with a few small galaxies throughout its lifetime, but that most of these were minor events. The largest recent merger, 10 billion years ago, was thought to involve the Sagittarius Dwarf Elliptical Galaxy, which donated streams of gas and groups of stars to the Milky Way’s stellar halo. But astronomers didn’t fully understand these objects until the Gaia satellite released its second data set in 2018.
As astronomers pored over the detailed motions and positions of about a billion stars, signs of a major disturbance in the galaxy emerged—they saw galactic wreckage in the halo. There, some stars orbit at extreme angles and have different compositions than others, suggesting that they originated somewhere else.
Astronomers took these oddball stars as evidence of a titanic collision between the Milky Way and another galaxy. The merger, which probably happened between 8 billion and 11 billion years ago, would have catastrophically disrupted the young Milky Way, ripped the other galaxy to shreds, and sparked a firestorm of new star formation.
The colliding galaxy’s remains are now called Gaia-Sausage-Enceladus, a result of two teams independently discovering the remnants of the merger. One team named it after the Greek deity Gaia, primordial mother of Earth and all life, and her son Enceladus. The other noticed that the remnants looked like a sausage. (Some astronomers dispute that the incoming galaxy was the only one involved, suggesting instead that many smaller collisions over a longer period could have resulted in the structures we now see.)
The merger changed everything: the course of the Milky Way’s halo, inner bulge, and flattened disk.
Now, astronomers are using various tools to understand the timing of the Gaia-Sausage-Enceladus pileup and how the infant Milky Way grew up as a result.
In March 2022, Maosheng Xiang and Hans-Walter Rix of the Max Planck Institute for Astronomy started by defining Milky Way 1.0, the proto-galaxy that existed before any mergers. They did this using ancient subgiant stars that are smaller than the sun, and which have used up their hydrogen fuel and are now growing puffy. A subgiant star’s brightness corresponds to its age, and its light serves as a fingerprint of its birth material. When Xiang and Rix used those clues to infer the migration histories of a quarter of a million subgiant stars, they found that the thick disk formed earlier than expected in galaxy formation theories—13 billion years ago, barely an eye-blink after the Big Bang.
Popular cosmological theories suggest that it should have taken longer for such large, well-defined structures to form after the Big Bang. And yet they keep cropping up in the James Webb Space Telescope’s observations of distant galaxies, said Rosemary Wyse, an astrophysicist at Johns Hopkins University.
“You can tie together how we think our galaxy formed with what JWST is seeing. Can we have a coherent picture of how a galaxy formed? Is our galaxy typical?” she said.
The thick disk might have existed before the main merger, but the thin disk coincided with the arrival of Gaia-Sausage-Enceladus, Xiang and Rix found. This two-pronged assembly process, which produces distinct stellar disks, may be common, and it could be crucial for sparking star formation. Birth rates have been declining since that frenzy, but the Milky Way still makes about 10 to 20 new stars a year.
Yuxi (Lucy) Lu, who just moved from Columbia University to the American Museum of Natural History, wanted to understand the history of the galactic disk and how it has changed over time. To do that, she studied how chemical changes over stars’ lifetimes could help identify their birth locations. She focused on similar puffy, subgiant stars, and in new, unpublished work she found that metal-rich subgiants—those with an abundance of elements heavier than helium—began growing in earnest around the time of the Gaia-Sausage-Enceladus merger, between 11 billion and 8 billion years ago.
The evidence for Gaia-Sausage-Enceladus continues to pile up. But what astronomers still don’t understand is why things have been calm ever since. The Milky Way’s chemical history and structural history seem atypical, Lu said.
Andromeda, for instance, has a much more violent history than the Milky Way. It would be odd for our galaxy to be left alone so long, considering other galaxies’ histories and the prevailing cosmological model that says galaxies grow by smashing into each other, Wyse said. “The merging history is unusual, and the assembly history. Whether we are actually unusual in the universe … I would say is still an open question,” she said.
Even as astronomers piece together the galaxy’s past, others are studying how the galaxy’s neighborhoods may be as different from one another as cities and suburbs—a possibility that raises the question of how planets (and maybe life) are distributed throughout the galaxy.
Here, around one particular star on the Local Arm, eight planets formed around the sun—four rocky and four gaseous. But other arms may be different. Those environments might produce different populations of stars and planets in the same way that specialized flora and fauna evolve on continents with distinct biospheres.
“Maybe life can only arise in a really quiet galaxy. Maybe life can only arise around a really quiet star,” said Jessie Christiansen, an astronomer at the California Institute of Technology who studies galactic conditions and their effects on planet-building. “It’s so hard with this statistical sample of one; anything [about our galaxy] can be important, or nothing can be important.”
A century after Edwin Hubble scrawled “VAR!” on a glass plate, the panoply of galaxies resolving in JWST’s field of view is changing what we know about the cosmos and our place in it. Just as we can use the Milky Way as an astrophysical observatory to understand the broader universe, we can also use the broader universe and its billions of galaxies to understand our home and how we came to be.
Astronomers continue to take a page from Hubble’s playbook and scrutinize Andromeda, the faint ellipse in the northern sky. As Gaia has done closer to home, the Dark Energy Spectroscopic Instrument at Kitt Peak National Observatory will measure individual stars in Andromeda and scrutinize their motions, ages, and chemical abundances. Wyse is also planning to study individual stars in the galaxy next door, using the Subaru Telescope on Mauna Kea.
Doing so will provide a new view of Andromeda’s past and a new comparison for our own galaxy. It will also offer a faint glimpse of the very distant future. Our galaxy will eventually destroy two small nearby galaxies, the Large and Small Magellanic Clouds, which are screaming across space in our direction. Our galaxy is already beginning to digest them.
“If we were observing all this a billion years from now, it would look much messier,” Conroy said. “We just happen to be at a time when things are relatively quiet.”
Next, Andromeda too will come join us. The galaxy spanning Edwin Hubble’s glass plates will be an island universe no more. Andromeda and the Milky Way will spiral in toward each other, their stellar halos swirling together. Over timescales that defy comprehension, the disks will also combine, heating cold gas and causing it to condense and ignite new stars. On the edges of whatever structure is built next, new suns will emerge, and with them new planets. But for now, all is quiet, here on the Local Arm of the only galaxy we will ever know.