Draft. This lesson has not completed review. Claims marked below may still change.

Lesson 2.4

Galaxies

Cities of stars, dark matter and the cosmic web.

Updated 2 September 2026 CosmologyPhysics Video: 6 to 7 minutes, planned
When The first galaxies by 8 January on the cosmic calendar; the Milky Way's oldest stars date from about 20 January; its disc took shape through March and April
How long Galaxies take billions of years to assemble and are still merging. The Milky Way's next big merger may be four to five billion years away, or may never happen
How big The Milky Way's disc is about a hundred thousand light-years across and its dark halo ten times wider; the web of galaxies spans the whole observable universe
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC 21 January
The misconception

A galaxy is just a very big cluster of stars.

Video In production. The transcript and shot list are at the end of this page.

Video plan · production notes, hidden once published

Title
A Galaxy Is Mostly Not Stars
Cold open
“Everyone thinks a galaxy is a very big crowd of stars. The stars are the part that shines. They are not the part that matters.”
Shape
Take apart the crowd-of-stars picture, rebuild the galaxy as a system, then keep zooming out until the galaxy itself is one dot in a web.
Thumbnail
Andromeda with the visible disc dimmed and the dark halo drawn around it, captioned THIS IS THE GALAXY.
Runtime
6 to 7 minutes, planned

Shorts cut from this lesson

  1. The Andromeda collision might not happen — A 2025 analysis makes it close to a coin flip. Textbooks say certain. (S08)
  2. We ate a galaxy, and Gaia found the crumbs — A quarter of the Milky Way's mass, swallowed ten billion years ago. (S07)
  3. Spiral arms are traffic jams — The pattern persists while every star passes through it. (S05)

Foundations

Bigger than a cluster

A galaxy looks, in a photograph, like a very large crowd of stars, and it is tempting to think that is all it is. It is not. A galaxy is a system: stars, yes, a few hundred billion of them in ours, but also gas and dust between the stars, a black hole of millions of solar masses at the centre, and, holding all of it together, an invisible halo of dark matter that outweighs everything else five to one. The stars are the part that shines. They are not the part that dominates.

When the universe got large

Until a hundred years ago it was reasonable to believe the Milky Way was the whole universe. The faint spiral smudges seen in telescopes might be nearby clouds or might be distant “island universes”; astronomers formally debated the question in 1920 and could not settle it. Edwin Hubble settled it. Between 1923 and 1929, using a kind of pulsating star whose brightness reveals its distance, he showed that the Andromeda spiral lies far beyond the Milky Way’s edge, a galaxy of its own, and that the sky is full of them. The universe became, in a few years, a hundred thousand times larger than anyone had known.

Our own

The Milky Way is a barred spiral: a flat disc about a hundred thousand light-years across, with a bar of stars through its centre and arms winding outward, wrapped in a faint sphere of old stars and globular clusters. The Sun sits in the disc, about twenty-seven thousand light-years from the centre, orbiting once every two hundred and thirty million years. It has made about twenty circuits since it formed. At the centre, hidden behind dust, is a black hole of four million solar masses, Sagittarius A*. Stars have been watched orbiting it for thirty years, and in 2022 its shadow was photographed by a telescope the size of the Earth.

The disc is where the action is: the gas, the dust, the star formation of the last chapter, at about one and a half solar masses a year. The spiral arms are not fixed groups of stars but patterns moving through the disc, like a traffic jam that persists while individual cars pass through it. Gas piles up in the arms, and that is where the bright young stars are born, which is why the arms shine.

The invisible part

Measure how fast the disc rotates at different distances from the centre and you find, as Vera Rubin did for other galaxies in the 1970s, that the outer parts move as fast as the inner parts. The visible matter cannot do that; there must be far more mass, spread far beyond the stars. The Milky Way’s dark halo extends to perhaps a million light-years and holds about a trillion solar masses. The galaxy sits inside it like a city inside a fog. Every galaxy does.

How galaxies grew

Galaxies were not born whole. In the Dark Ages, the ripples of the afterglow grew into small clumps of dark matter, gas fell into them, and the first stars formed. The clumps merged into larger clumps, and those into larger still, over billions of years: galaxies were built from the bottom up. The Milky Way’s own history is written in the motions of its stars. The Gaia satellite, which has measured the positions and speeds of over a billion of them, found in 2018 the scattered remains of a galaxy about a quarter of the Milky Way’s mass that it swallowed some ten billion years ago. Smaller galaxies are being absorbed today; streams of their stars still trail across our sky.

The next big event is a matter of debate. Andromeda, our nearest large neighbour, is approaching at a hundred and ten kilometres per second. For a decade the standard prediction was a collision in about four and a half billion years. A 2025 analysis that took the uncertainties seriously found only a coin-flip chance of a merger within ten billion years. Whether the Milky Way ends its life as part of a giant elliptical or drifts on alone is open.

Where we live

The Milky Way belongs to the Local Group, a loose family of about a hundred galaxies dominated by Andromeda and ourselves. The Local Group lies on the edge of the Virgo cluster, a swarm of a thousand galaxies fifty million light-years away. Virgo, in turn, is part of Laniakea, a supercluster of a hundred thousand galaxies half a billion light-years across, named in 2014 with a Hawaiian word for “immense heaven”. And Laniakea is one knot in a web: galaxies strung along filaments and sheets around enormous empty voids, a pattern that grew from the ripples in the afterglow and stretches across the whole observable universe. Counting the faintest, there are hundreds of billions of galaxies, perhaps two trillion, each a system like ours.

Deeper

The Milky Way by the numbers

Distance from the Sun to the Galactic centre: 8.18 kiloparsecs, measured directly in 2019 from the orbit of a star around the central black hole. Rotation speed at the Sun’s radius: about 230 kilometres per second, giving an orbital period near 230 million years. Stellar mass: about 6 × 10¹⁰ solar masses, with a total virial mass of roughly 10¹² solar masses within about 200 kiloparsecs. Star formation rate: 1.65 ± 0.19 solar masses per year. Central black hole: 4.15 million solar masses. The number of stars is quoted anywhere between a hundred and four hundred billion because it depends on how many faint red dwarfs are assumed; the mass is far better known than the count.

Hubble’s fork

Hubble’s 1926 scheme sorts galaxies into ellipticals, from round to cigar-shaped, spirals with and without bars, and irregulars. He drew it as a tuning fork and wondered whether galaxies evolved along it; they do not, but the classes are physically meaningful. Ellipticals are old, red, gas-poor and dominated by random motions, typically the products of major mergers. Spirals are rotating discs with ongoing star formation. About two thirds of large disc galaxies, including the Milky Way, have bars. Dwarf galaxies, the most numerous kind, mostly fall outside the scheme.

Rotation and the halo

For a disc of stars and gas alone, orbital speed should fall with radius beyond the bulk of the mass, as the planets’ speeds do in the solar system. Observed rotation curves stay flat to the last measurable point, implying enclosed mass rising in proportion to radius, and hence a density falling as one over radius squared: the signature of an extended halo. In the Milky Way the halo is traced by the motions of globular clusters, satellite galaxies and stellar streams out to hundreds of kiloparsecs.

Density waves and star formation

Lin and Shu proposed in 1964 that spiral arms are quasi-stationary density waves, regions of slightly enhanced gravity that orbit the centre at a fixed pattern speed while stars and gas pass through them. Gas entering an arm is compressed, cools, and forms stars; the brightest of those die before they leave, so the arms are outlined by young blue stars and glowing gas. Modern simulations show arms that are more transient and recurrent than the original theory, but the essential picture, that arms are patterns rather than material structures, holds.

Hierarchical assembly

In the cold dark matter model, small halos collapse first and merge into larger ones, and galaxies follow their halos. Mergers between galaxies of comparable mass destroy discs and build ellipticals; minor mergers thicken discs and add to the halo of old stars. Gaia’s measurement of stellar motions revealed a large population of stars on highly elongated orbits, chemically distinct from the disc, that share a common origin: a galaxy of about 10⁹ solar masses in stars, now called Gaia-Enceladus, accreted roughly ten billion years ago and responsible for much of the inner halo and the thick disc.

Structure on the largest scales

The Local Group’s galaxies orbit a common centre of mass; the Group falls toward Virgo; Virgo and its neighbours flow toward the Great Attractor region; and the whole assembly, Laniakea, is defined by the region within which galaxy motions converge. On still larger scales galaxy surveys map filaments hundreds of millions of light-years long, and the statistical pattern of that web, including the baryon acoustic scale from lesson 1.5, matches what gravity should have grown from the afterglow’s ripples in a universe with the measured amounts of dark matter and dark energy.

Frontier

The next merger

Whether Andromeda and the Milky Way collide depends on the sideways component of Andromeda’s motion, on the masses of both galaxies and their halos, and on the third large member of the Local Group, the Triangulum galaxy, whose gravity tugs on the pair. The 2025 analysis found that plausible ranges of these quantities give outcomes from a merger in five billion years to no merger at all. Gaia’s continuing measurements of Andromeda’s stars are narrowing the sideways velocity. Active research.

Too many satellites, or too few

The cold dark matter model predicts thousands of small halos around a galaxy like ours; only a few dozen satellite galaxies are seen. Either most small halos never formed stars, which the physics of reionisation and stellar feedback can explain, or the model needs adjusting on small scales. The count of faint satellites found by deep surveys has been rising, and the tension has eased but not vanished.

How galaxies stop

Large ellipticals stopped forming stars long ago, and their central black holes are suspected of doing the stopping, heating or expelling the gas that would otherwise cool and form stars. How a black hole a billionth of the galaxy’s size regulates the whole system is one of the central questions in galaxy evolution.

Transcript and shot list

Target runtime about 6 minutes. Built to take apart the crowd-of-stars picture, rebuild the galaxy as a system, then keep zooming out until the galaxy itself is one dot in a web. Shot types: ANIM is rendered from code with Manim, AI is a generated shot from Higgsfield with the shared style reference, REAL is agency imagery with credit, CAM is the narrator.

[00:00] S01 · REAL · The Andromeda galaxy, wide field (NASA, ESA, or an observatory image; confirm credit line).

This is Andromeda, two and a half million light-years away, the most distant thing you can see with your naked eye. A century ago nobody knew what it was. And most people, looking at it now, think they know exactly what it is: a very big crowd of stars. It isn’t. The stars are the part that shines. They are not the part that matters.

[00:30] S02 · ANIM · Cosmic calendar; markers at 8 January (first galaxies), 20 January (Milky Way’s oldest stars), March to April (the disc forms).

The first galaxies were assembling by the 8th of January on the cosmic calendar. The Milky Way’s oldest stars date from around the 20th. Its disc, the part we live in, took shape through March and April.

[00:47] S03 · ANIM · Typographic: 1920, the Great Debate; then a pulsating star in Andromeda with a distance readout jumping to millions of light-years; the universe expands outward a hundred thousand times.

In 1920 astronomers held a formal debate about whether the Milky Way was the whole universe, and couldn’t settle it. Edwin Hubble did. Between 1923 and 1929, using a kind of pulsating star whose rhythm reveals its distance, he showed that Andromeda lies far beyond the Milky Way’s edge. A galaxy of its own. And the sky was full of them. In a few years, the universe became a hundred thousand times bigger than anyone had known.

[01:23] S04 · ANIM · The Milky Way from above: bar, arms, disc; the Sun’s position 27,000 light-years out; its orbit traced; a counter of twenty circuits; zoom into the centre and the black hole’s shadow.

Our own galaxy is a barred spiral: a disc a hundred thousand light-years across, a bar of stars through the middle, arms winding out. The Sun sits here, twenty-seven thousand light-years from the centre, going round once every two hundred and thirty million years. About twenty laps since it formed. At the centre, hidden behind dust, a black hole of four million solar masses. In 2022 its shadow was photographed by a telescope the size of the Earth.

[01:59] S05 · ANIM · A spiral arm as a traffic jam: stars pass through it; gas piles up; bright blue stars ignite inside the jam.

The arms aren’t fixed groups of stars. They’re patterns moving through the disc, like a traffic jam that persists while every car passes through. Gas piles up in the jam, and that’s where the bright young stars are born. That’s why the arms shine.

[02:19] S06 · ANIM · A rotation curve: expected falling, measured flat; the galaxy shrinks inside a vast translucent halo.

Now the part you can’t see. Measure how fast the disc rotates at different distances and the outer parts move as fast as the inner ones. The visible matter can’t do that. There must be far more mass, spread far beyond the stars. The Milky Way’s dark halo reaches out perhaps a million light-years and holds about a trillion solar masses. The galaxy sits inside it like a city in a fog. Every galaxy does. The stars are the part that shines. They’re not the part that dominates.

[02:00] S07 · ANIM · Small clumps of dark matter merge into larger ones over a clock of billions of years; a galaxy assembles; a smaller galaxy is swallowed and its stars scatter into a halo.

Galaxies weren’t born whole. In the Dark Ages, small clumps of dark matter gathered gas and made the first stars, then merged into larger clumps, and those into larger still, over billions of years. From the bottom up. The Milky Way’s history is written in the motions of its stars. The Gaia satellite, which has measured over a billion of them, found in 2018 the scattered remains of a galaxy a quarter of the Milky Way’s mass that we swallowed ten billion years ago. Streams of stars from smaller victims still trail across our sky.

[03:43] S08 · ANIM · Andromeda approaches the Milky Way at 110 km/s; two futures branch: a merger into an elliptical, and a near miss; a coin flips.

The next big event is a live argument. Andromeda is approaching at a hundred and ten kilometres per second. For a decade the prediction was a collision in four and a half billion years. A 2025 analysis that took the uncertainties seriously found only a coin-flip chance of a merger within ten billion. Whether we end as part of a giant elliptical or drift on alone: open.

[04:14] S09 · ANIM · Zoom out: the Local Group; the Virgo cluster; Laniakea’s flow lines converging; then the cosmic web of filaments and voids across the observable universe.

Zoom out. The Milky Way belongs to the Local Group, a loose family of about a hundred galaxies dominated by Andromeda and us. The Local Group sits on the edge of the Virgo cluster, a thousand galaxies fifty million light-years away. Virgo is part of Laniakea, a hundred thousand galaxies half a billion light-years across, named in 2014 with a Hawaiian word for immense heaven. And Laniakea is one knot in a web: galaxies strung along filaments around enormous empty voids, a pattern grown from the ripples in the afterglow, across the whole observable universe. Hundreds of billions of galaxies. Perhaps two trillion. Each a system like ours.

[05:04] S10 · REAL · A galaxy-survey map of the cosmic web (SDSS, DESI or a simulation frame; confirm credit line).

This is a map of a slice of it, every dot a galaxy. The web is not a metaphor. It’s a structure, the largest there is, and the afterglow’s ripples are its seed.

[05:19] S11 · ANIM · Labels: ESTABLISHED for what a galaxy is and how it grew; ACTIVE RESEARCH for the Andromeda merger, the missing satellites and how galaxies stop forming stars.

The scorecard. What a galaxy is, that most of its mass is dark, that it grew by mergers from the bottom up, and how the web is arranged: established. Whether Andromeda hits us, why there are fewer small satellite galaxies than the model predicts, and how a black hole a billionth of a galaxy’s size can shut off its star formation: active research.

[05:48] S12 · AI into ANIM · Andromeda again, now with the Milky Way’s disc in the foreground; the calendar returns.

So look at Andromeda again. A few hundred billion stars, a black hole, gas, dust, and a trillion solar masses of something invisible, arriving at a hundred and ten kilometres per second, maybe. Everything we know about it we learned from its light. Next lesson: how to read starlight.

Shot list

Shot Time Type What we see Scene or asset Status
S01 00:00 REAL Andromeda, wide field NASA, ESA or observatory image; confirm licence and credit planned
S02 00:30 ANIM Calendar markers for galaxy formation shared scene CosmicCalendarZoom planned
S03 00:47 ANIM The Great Debate and Hubble’s distance UniverseGetsLarge planned
S04 01:23 ANIM Milky Way from above; the Sun’s orbit; the black hole OurGalaxy planned
S05 01:59 ANIM Spiral arm as traffic jam TrafficJam planned
S06 02:19 ANIM Rotation curve and the halo CityInAFog planned
S07 02:00 ANIM Hierarchical assembly and Gaia-Enceladus BottomUp planned
S08 03:43 ANIM Andromeda, two futures, a coin CoinFlip planned
S09 04:14 ANIM Zoom out to the cosmic web ZoomOut planned
S10 05:04 REAL Galaxy-survey map SDSS, DESI or simulation frame; confirm licence and credit planned
S11 05:19 ANIM Confidence labels shared scene ConfidenceLabels planned
S12 05:48 AI + ANIM Andromeda with our disc in the foreground; calendar returns Higgsfield 8 s clip, then ArrivingMaybe planned

One AI shot, nine code-rendered shots, two agency images. Estimated Higgsfield use: 2 to 4 generations including retries.

Connected across time

Key terms

Full glossary
Galaxy
A gravitationally bound system of stars, gas, dust and dark matter, from a few million to several trillion stars. The Milky Way is one of hundreds of billions.
Milky Way
Our galaxy: a barred spiral with a disc about a hundred thousand light-years across, a few hundred billion stars, a central black hole of four million solar masses, and a dark halo holding most of its mass.
Hubble sequence
Hubble's 1926 classification of galaxies into ellipticals, spirals and irregulars, still used, though it is not the evolutionary sequence he once suspected.
Dark matter halo
The invisible envelope in which every galaxy sits, extending far beyond its stars and holding most of its mass. Galaxies form inside halos, not the other way round.
Supermassive black hole
A black hole of millions to billions of solar masses at the centre of nearly every large galaxy. The Milky Way's, Sagittarius A*, was photographed in 2022.
Local Group
The small cluster of galaxies the Milky Way belongs to: Andromeda, Triangulum and dozens of dwarfs, spread over about ten million light-years.
Galaxy cluster
A gravitationally bound swarm of hundreds to thousands of galaxies in hot gas and dark matter. The nearest large one is in Virgo, about fifty million light-years away.
Cosmic web
The largest structure there is: galaxies strung along filaments and sheets around vast empty voids, grown by gravity from the ripples in the afterglow.
Hierarchical formation
Galaxies grew from the bottom up, small clumps merging into larger ones over billions of years. The Milky Way's own history of mergers is written in the motions of its stars.
Density wave
A spiral arm is not a fixed set of stars but a pattern, a traffic jam moving through the disc, in which gas piles up and new stars are born.

Check yourself

1. What makes up most of a galaxy's mass?

Show the answer

Dark matter. Stars, gas and dust together are a minority. The dark matter halo in which a galaxy sits holds about five sixths of its mass and extends far beyond the visible disc.

2. Until when was the Milky Way believed to be the whole universe?

Show the answer

The 1920s. The question was formally debated in 1920. Between 1923 and 1929 Edwin Hubble measured the distance to Andromeda using Cepheid variable stars and showed it lies far outside the Milky Way, a galaxy in its own right.

3. How do galaxies grow?

Show the answer

By merging with smaller galaxies and drawing in gas over billions of years. Galaxies form from the bottom up. The Milky Way swallowed a galaxy of about a quarter of its then mass around ten billion years ago, and is still absorbing dwarfs today.

Claim by claim

Every key claim in this lesson, with how confident science is about it and where it comes from.

Established measured and confirmed many times over Active research well supported, still being tested Open question not answered by current science

Sources

  1. Hubble, E. (1929). A spiral nebula as a stellar system, Messier 31. The Astrophysical Journal, 69, 103–158. doi.org/10.1086/143167
  2. Hubble, E. P. (1926). Extragalactic nebulae. The Astrophysical Journal, 64, 321–369. doi.org/10.1086/143018
  3. GRAVITY Collaboration (2019). A geometric distance measurement to the Galactic center black hole with 0.3% uncertainty. Astronomy & Astrophysics, 625, L10. doi.org/10.1051/0004-6361/201935656
  4. Event Horizon Telescope Collaboration (2022). First Sagittarius A* Event Horizon Telescope results. I. The shadow of the supermassive black hole in the center of the Milky Way. The Astrophysical Journal Letters, 930(2), L12. doi.org/10.3847/2041-8213/ac6674
  5. Bland-Hawthorn, J., & Gerhard, O. (2016). The Galaxy in context: Structural, kinematic, and integrated properties. Annual Review of Astronomy and Astrophysics, 54, 529–596. doi.org/10.1146/annurev-astro-081915-023441
  6. Licquia, T. C., & Newman, J. A. (2015). Improved estimates of the Milky Way's stellar mass and star formation rate from hierarchical Bayesian meta-analysis. The Astrophysical Journal, 806(1), 96. doi.org/10.1088/0004-637X/806/1/96
  7. Rubin, V. C., Ford, W. K., Jr., & Thonnard, N. (1980). Rotational properties of 21 Sc galaxies with a large range of luminosities and radii. The Astrophysical Journal, 238, 471–487. doi.org/10.1086/158003
  8. Lin, C. C., & Shu, F. H. (1964). On the spiral structure of disk galaxies. The Astrophysical Journal, 140, 646–655. doi.org/10.1086/147955
  9. Helmi, A., Babusiaux, C., Koppelman, H. H., Massari, D., Veljanoski, J., & Brown, A. G. A. (2018). The merger that led to the formation of the Milky Way's inner stellar halo and thick disk. Nature, 563(7729), 85–88. doi.org/10.1038/s41586-018-0625-x
  10. van der Marel, R. P., Besla, G., Cox, T. J., Sohn, S. T., & Anderson, J. (2012). The M31 velocity vector. III. Future Milky Way M31–M33 orbital evolution, merging, and fate of the Sun. The Astrophysical Journal, 753(1), 9. doi.org/10.1088/0004-637X/753/1/9
  11. Sawala, T., et al. (2025). No certainty of a Milky Way–Andromeda collision. Nature Astronomy, 9. doi.org/10.1038/s41550-025-02563-1
  12. Tully, R. B., Courtois, H., Hoffman, Y., & Pomarède, D. (2014). The Laniakea supercluster of galaxies. Nature, 513(7516), 71–73. doi.org/10.1038/nature13674
  13. Conselice, C. J., Wilkinson, A., Duncan, K., & Mortlock, A. (2016). The evolution of galaxy number density at z < 8 and its implications. The Astrophysical Journal, 830(2), 83. doi.org/10.3847/0004-637X/830/2/83
  14. Planck Collaboration (2020). Planck 2018 results. VI. Cosmological parameters. Astronomy & Astrophysics, 641, A6. doi.org/10.1051/0004-6361/201833910

Go deeper