Lecture 23: The Galaxy Zoo
Classifying Galaxies, Meeting Quasars, and Previewing the Cosmic Web
The Big Idea
The Milky Way is one member of a vast population of galaxies that comes in a small number of recognizable shapes. Those shapes are not arbitrary — they trace the star-formation history, gas content, and dynamical state of each galaxy. Many galaxies host a supermassive black hole at their center, and when that black hole is fed, it becomes one of the brightest objects in the universe — a quasar. On the largest scales, galaxies are not scattered randomly but arranged in a cosmic web of filaments and voids. To make sense of any of this, we need distances — which is the problem Lecture 24 will solve.
Observable: Galaxies have different shapes, colors, gas contents, spectra, central activity, and clustering patterns.
Model: Stellar populations, gas dynamics, black-hole accretion, and gravitational instability connect those observables to physical histories.
Inference: A galaxy’s appearance is not just a picture; it is a record of star formation, mergers, central black-hole growth, and dark-matter-guided structure formation.
Main uncertainty: Classification is a starting point, not a full history. Galaxies can merge, quench, reignite star formation, or host active nuclei for only part of their lives.
This page answers four questions:
- What kinds of galaxies are there, and what do their shapes tell us?
- How do we know galaxies are far away — and why is that measurement hard?
- What are quasars, and why do they matter?
- How are galaxies arranged in space on the largest scales?
Punchline: Galaxy morphology is a stellar-population readout. Quasars are SMBH accretion signatures. The cosmic web is the shape of the matter distribution, and stars are how we see all of it.
This reading is a guided tour of the galaxy population and its central engines.
Default expectation (best): Read the whole page before lecture. Pause at each Check Yourself prompt.
If you’re short on time (~20 min): Focus on:
- The Big Idea above
- The Hubble Sequence section (three morphologies, three stellar stories)
- The Quasar section (SMBH + accretion + Eddington limit)
- The Distance Problem preview
Goal after 20 minutes: You should be able to describe the three main galaxy types, explain what a quasar is and where its energy comes from, and articulate why measuring galaxy distances is central to the rest of Module 3.
Reference mode: The Galaxy Reference Box and the Glossary are the study references.
What to notice: Module 3 uses the same evidence chain over and over. Observables such as motions, spectra, standard candles, supernova brightness, the CMB, and element abundances become physical claims only after a model translates them. (Credit: Illustration: A. Rosen (SVG))
Learning Outcomes
By the end of this reading, you should be able to say:
Beyond the Milky Way
Lecture 22 ended at the Galactic Center, 8 kpc from Earth. Step outward. At 50 kpc you reach the Large Magellanic Cloud, a small companion of the Milky Way. At ~780 kpc (~2.5 million light-years) you reach Andromeda (M31), our nearest large neighbor. Beyond that, the universe opens up: somewhere between a few hundred billion and a few trillion galaxies lie within the observable universe. The exact count depends on how you model the faint end of the galaxy luminosity function, and recent measurements of the cosmic optical background (e.g., New Horizons’ LORRI data) have pushed the estimate toward the lower end of that range. Either way: a lot.
What to notice: the Milky Way is one member of a small gravitational neighborhood. Andromeda, Triangulum, the Magellanic Clouds, and many dwarf galaxies together make up the Local Group. (Credit: Course-provided figure)
The Local Group is the first place where “galaxy” stops meaning only “the Milky Way” and starts meaning a population. It contains a few large galaxies, a handful of intermediate companions, and many small dwarf galaxies whose faintness makes them hard to census completely.
Andromeda is the nearest large spiral in that neighborhood. Because it is close enough for Hubble to resolve individual stars across its disk, it lets us connect the stellar-population tools from Module 2 to the galaxy-scale questions of Module 3.
What to notice: Andromeda is close enough for Hubble to resolve enormous numbers of individual stars. That makes it both a neighboring galaxy and a bridge between stellar astronomy and galaxy astronomy. (Credit: NASA/ESA/Hubble)
Before we jump to deep fields, pause on the nearby universe. The Local Group is not a representative sample of all galaxies, but it is the place where distances, sizes, and morphology are easiest to compare directly.
What to notice: the Local Group contains a few large galaxies and many much smaller companions. Galaxy size varies enormously even before we leave our own neighborhood. (Credit: Course-provided figure)
What to notice: the Local Group is the nearby knot in a larger galactic neighborhood. The next galaxies are already millions of light-years away, so distance changes the scale of the story quickly. (Credit: Course-provided figure)
What to notice: a deep survey image is a population sample. Each galaxy is both an object with a morphology and a distance marker waiting for the ladder to place it in three-dimensional space. (Credit: Rubin Observatory/NSF/AURA)
What to notice: nearly every tiny smudge is a galaxy. A deep field turns an apparently empty patch of sky into evidence for a universe filled with galaxies at many distances and lookback times. (Credit: NASA/ESA/Hubble)
Use the exposure-time slider to watch a dark patch of sky turn into a field of galaxies. Start at the shortest exposure and ask: which specks would you trust as real? Then move right and notice how repeated exposures separate faint sources from noise.
What to notice: “empty sky” is often an exposure-time statement, not a physical statement. Deep fields are why we know the galaxy population includes faint, distant systems whose light has been traveling for most of cosmic history.
Deep-field galaxies are also multiwavelength objects. A galaxy that looks like a faint blob in visible light may reveal hidden star formation at radio wavelengths, where dust is less of an obstacle.
What to notice: distant galaxies are often multiwavelength measurements, not simple photographs. Visible light shows the stellar body, while radio emission can flag dust-obscured star formation. (Credit: Hubble Space Telescope and Karl G. Jansky Very Large Array)
What you would notice looking at those galaxies is that they are not random. Most of them fall into a small number of shape categories that were first systematized by Edwin Hubble in 1926 — the so-called Hubble sequence, or Hubble tuning fork.
What to notice: these labels are visual starting points. A galaxy’s type is a clue about stars, gas, dust, and interaction history, not a complete biography by itself. (Credit: Course-provided figure)
The Hubble Sequence
Galaxies come in three broad morphological classes:
- Ellipticals (E): Smooth, featureless, football-shaped systems ranging from nearly round (E0) to elongated (E7). Dominated by old red and yellow stars. Little cold gas, little ongoing star formation. The most massive galaxies in the universe are giant ellipticals, sitting at the centers of galaxy clusters.
- Spirals (S): Disk galaxies with a central bulge and spiral arms tracing out from the center. Classed from Sa (tightly wound arms, large bulge, less gas) to Sc/Sd (open arms, small bulge, gas-rich, active star formation). The Milky Way is a barred spiral (SBb or SBc). Spirals host a mix of old bulge stars and young disk stars.
- Irregulars (Irr): Galaxies with no clear symmetry. Often small and gas-rich. The Large Magellanic Cloud is a classic example.
What to notice: the Hubble tuning fork is a classification map, not an evolutionary conveyor belt. The images here are infrared views of dust in nearby galaxies, so morphology is being connected to interstellar material and star formation. (Credit: C. North, M. Galametz & the KINGFISH Team)
The three classes are not just aesthetic — they are stellar-population classes. An elliptical is a galaxy that stopped making stars long ago. A spiral is a galaxy that is still making them. An irregular is often a galaxy in a disturbed dynamical state, where a recent interaction has stirred things up and triggered new star formation.
What to notice: the Sombrero Galaxy mixes a bright bulge with a thin dust-rich disk. Real galaxies often sit between tidy textbook categories, so classification is a starting point for physical interpretation. (Credit: NASA/ESA/Hubble)
One important warning: the Hubble sequence is a classification scheme, not an evolutionary path. Ellipticals do not automatically turn into spirals, and spirals do not naturally march along the tuning fork from one subtype to another. The sequence helps us sort what galaxies look like today; the physical history has to be inferred from stellar populations, gas, dynamics, and environment.
Use this as a corrective to the tempting but wrong idea that galaxy type is a fixed identity. Follow the projected Milky Way-Andromeda interaction from separate spirals to a merged elliptical-like remnant.
What to notice: the stars mostly pass by one another, but the gas does not. Gas compression triggers star formation, and later gas depletion helps leave behind a more mature, less star-forming system.
The Whirlpool Galaxy is a concrete example of why morphology is historical evidence. Its spiral structure and companion are not independent facts; the interaction helps shape the very pattern we classify.
What to notice: the Whirlpool Galaxy turns interaction into structure. Its companion helps organize spiral arms, dust lanes, and star-forming regions, reminding us that morphology records environment. (Credit: NASA/ESA/Hubble)
NGC 602 is a star-forming region in the Small Magellanic Cloud, a satellite galaxy of the Milky Way. Use the exposure-time slider to connect two ideas at once: longer exposures reveal fainter structure, and young massive stars sculpt the gas around them.
What to notice: the same image is both a galaxy-population clue and a stellar-lifecycle clue. The blue young stars, glowing gas, and dusty pillars are the local evidence behind the phrase “gas-rich galaxies are still forming stars.”
M82 pushes the same idea into the extreme. It is a starburst galaxy: gas is being converted into stars so rapidly that feedback from massive stars and supernovae drives material out of the disk.
What to notice: starburst galaxies convert gas into stars at an extreme rate. The red outflowing material in M82 is feedback: young massive stars and supernovae pushing gas out of the galaxy. (Credit: NASA/ESA/Hubble)
Use the wavelength buttons to inspect a real galaxy collision in radio, infrared, visible, and X-ray light. This is a better mental model for galaxy evolution than a static tuning fork.
What to notice: different wavelengths reveal different stages of the same physical event. Radio and infrared trace gas and dust; visible light traces stars and star-forming regions; X-rays reveal compact remnants and hot gas from earlier waves of massive-star evolution.
What to notice: the Antennae collision is both a stellar event and a gas event. Blue starlight shows young populations, while the radio and submillimeter gas overlay marks where compressed material can fuel new star formation. (Credit: ESO/ALMA/Hubble)
Galaxy interactions are not just visual distortions in starlight. They also move the cold molecular gas that will make the next generation of stars. In a merger, the gas can be driven into the center, pulled into tidal features, or compressed into off-center star-forming knots.
What to notice: galaxy interactions rearrange cold molecular gas in many different ways. The gas distribution is the clue to where future star formation can be triggered, suppressed, or concentrated. (Credit: Course-provided figure)
| Type | Shape | Dominant stars | Gas/dust | Star formation |
|---|---|---|---|---|
| Elliptical (E) | Smooth, ellipsoidal | Old, red | Little | Very low |
| Spiral (S/SB) | Disk + bulge, arms | Mixed (old bulge, young disk) | Moderate, organized | Ongoing in arms |
| Irregular (Irr) | Disorganized | Often young, blue | Gas-rich | Often high, bursty |
Mnemonic: Ellipticals are done. Spirals are working. Irregulars are messy.
Check Yourself 1: Reading a Galaxy
You look at two galaxies. Galaxy A is a smooth red ellipsoid with almost no cold gas detected in 21-cm. Galaxy B is a blue-tinged pinwheel with prominent pink H II regions along its arms.
Which galaxy is forming stars today? Which has the older average stellar population? What is the key observational clue that distinguishes the two?
Galaxy B is forming stars today. The pink H II regions are ionized hydrogen around young, hot, massive O/B stars — a direct signature of ongoing star formation. The blue tint of the light also points to young, hot stars dominating the integrated light. Galaxy A, by contrast, is red because it is dominated by old cool K/M stars, and has no cold gas left to form new stars from. The key observational clue is the color (blue = young and hot; red = old and cool) plus the H II region signature (ongoing ionization requires young massive stars). This is the Module 2 H-R diagram applied at the galactic scale.
The Distance Problem
The Hubble sequence is easy enough to see. What is not easy is to know how far away each galaxy is. And without distance, we cannot convert what we see (apparent brightness, apparent size) into what we want (luminosity, physical size, mass).
Recall the situation for stars (Lecture 15): within ~100 pc, parallax gives distance directly. Beyond that, parallax fails, and we have to use standard candles — objects whose true luminosity \(L\) is known. From the inverse-square law,
\[F = \frac{L}{4\pi d^2} \quad \Rightarrow \quad d = \sqrt{\frac{L}{4\pi F}}\]
so measuring flux \(F\) gives distance \(d\) if we know \(L\).
What to notice: Each rung calibrates the next. Parallax (geometry) → Cepheids (standard candles) → Supernovae (Chandrasekhar limit) → Hubble Flow (cosmology). We infer the infinite from the infinitesimal. (Credit: (A. Rosen/NotebookLM))
For galaxies, the nearest are at ~1 Mpc (a million parsecs), a thousand times farther than any star we can parallax. We need a standard candle that is bright enough to be seen at that distance and reliable enough that its luminosity does not vary much from object to object. That is the distance-ladder problem, and Lecture 24 is dedicated to it. For now, note:
- Henrietta Leavitt’s Cepheid variables (discovered in the 1910s) are our first rung beyond parallax — they obey a period-luminosity relation that lets us read their \(L\) from their pulsation period.
- Type Ia supernovae (Lecture 25) take us further — they are so bright that we can see them across billions of light-years.
What to notice: Standard candles work because physics predicts their luminosity. Measure flux (F) + know luminosity (L) → calculate distance (d). (Credit: (A. Rosen/NotebookLM))
Both of these are stellar objects. Stars are what let us measure cosmic distances. Keep this thread in mind through the rest of the module — every distance beyond the Milky Way ultimately traces back to a star whose luminosity we have calibrated.
Check Yourself 2: Why “Standard”?
Why must a standard candle have a reliable, predictable luminosity \(L\)? What happens if two objects that look like the same kind of standard candle actually have luminosities that differ by a factor of 4?
From \(d = \sqrt{L/(4\pi F)}\), a factor-of-4 error in \(L\) makes a factor-of-2 error in \(d\) (since the square root of 4 is 2). Every derived quantity — size, mass, velocity — then inherits that error. A standard candle must have small intrinsic scatter in \(L\), and its scatter has to be calibratable, so that the distance uncertainty is bounded. “Reliable” here means small and well-characterized dispersion, not “literally identical every time.”
Active Galactic Nuclei
Lecture 22 ended with a supermassive black hole at the center of our Galaxy — Sagittarius A*, weighing about \(4 \times 10^6 \, M_\odot\). It is quiet. It barely shines.
Most SMBHs in the local universe are like Sgr A* — dormant. But in some galaxies, the central black hole is actively feeding on surrounding gas, and when it does, it produces one of the most extreme sources of light in the universe. These galaxies are called active galactic nuclei (AGN), and their most luminous members are quasars.
What to notice: a quasar is an accretion engine, not a star. The bright disk and jets are powered by gravitational energy released near a supermassive black hole. (Credit: Course-provided figure)
What to notice: an active nucleus is embedded inside a host galaxy. Dust can hide the central engine in some directions while jets carry energy far beyond the black hole’s immediate surroundings. (Credit: ESO)
The key idea is compactness. The galaxy can be tens of thousands of light-years across, but the power source sits in the central light-hours to light-days.
What to notice: the AGN engine is compact. A black hole plus accretion disk sits inside a dusty torus, and some systems launch jets perpendicular to the disk. (Credit: Course-provided figure)
This schematic is useful as a vocabulary map. Read it with one caution: the event horizon, disk, photon sphere, and jets are physically meaningful regions, while the singularity marks where our classical theory stops being enough.
What to notice: the event horizon, accretion disk, photon sphere, and jets are different physical regions. The singularity label is a signpost for where classical general relativity reaches its limit, not something directly observed. (Credit: ESO)
Use the wavelength buttons to build a multiwavelength picture of an active galaxy. First look in visible light and ask what is hidden; then compare radio, infrared, and X-ray views.
What to notice: the active SMBH is not just a bright point. Radio and X-ray light reveal jets tens of thousands of light-years long, while infrared and visible light reveal dust and stars from a merger history. This is the AGN engine embedded in a real galaxy.
The same lesson shows up in Hercules A. In visible light, the host galaxy does not announce the scale of the engine. Radio and X-ray views reveal the enormous structures inflated by the active nucleus.
What to notice: visible light alone makes Hercules A look fairly ordinary, but radio and X-ray views reveal enormous jets and lobes powered by the central AGN. (Credit: NASA, ESA, NRAO, and STScI)
An AGN has four components:
- The supermassive black hole — usually \(10^6\) to \(10^{10} \, M_\odot\).
- The accretion disk — gas spiraling into the black hole, heated by viscous friction to temperatures of \(10^5\) K or more, radiating across the ultraviolet and X-ray.
- The dusty torus — a ring of colder gas and dust surrounding the disk, responsible for much of the infrared emission.
- Relativistic jets (in some AGN) — collimated streams of plasma launched perpendicular to the disk at nearly the speed of light.
Where the Light Comes From
The energy source of a quasar is not nuclear fusion. It is gravitational potential energy released as matter falls toward the SMBH. As gas spirals inward through the accretion disk, viscous friction converts its gravitational energy into heat, and the heat radiates away. Up to ~10% of the rest-mass energy (\(E = mc^2\)) of the infalling material can be converted into light for a non-rotating (Schwarzschild) black hole, rising to ~30 – 40% for a maximally rotating (Kerr) black hole. Compare this to hydrogen fusion in a star, which converts only ~0.7% of mass-energy into light. Accretion onto a black hole is, pound for pound, the most efficient known sustained power source in astrophysics.
What to notice: the light source is not fusion in stars. Gas releases gravitational energy as it spirals through a hot accretion disk close to a supermassive black hole. (Credit: NASA/JPL-Caltech)
A quasar can outshine its entire host galaxy of \(\sim 10^{11}\) stars — from a region roughly the size of our Solar System.
Check Yourself 3: A Solar-System-Sized Lighthouse
A typical luminous quasar has \(L \sim 10^{13} \, L_\odot\). Observationally, quasars show flux variations on timescales as short as hours to days, which constrains the emitting region to be at most \(\sim 1\) light-day (\(\sim 2.6 \times 10^{13}\) m) — smaller than the outer Solar System. Using energy-reservoir reasoning (not luminosity density), explain why fusion cannot sustain this luminosity from such a small region, and why accretion onto a supermassive black hole can.
The argument is about total energy available per gram of fuel, combined with the small size of the emitting region.
- Fusion converts at most ~0.7% of rest-mass energy (\(E = 0.007 \, mc^2\)) into radiation. To produce \(L \sim 10^{13} \, L_\odot \approx 4 \times 10^{39}\) W via fusion, the mass-consumption rate would be \(\dot{m} = L / (0.007 \, c^2) \approx 100 \, M_\odot\)/yr. Sustaining that for even ~1 Myr requires ~\(10^8 \, M_\odot\) of hydrogen to be packed into a region smaller than the Solar System — a gravitationally self-destructive configuration. Stars that massive do not exist and could not sustainably fuse.
- Accretion onto a supermassive black hole converts up to ~10% of rest-mass energy. The required mass-consumption rate drops to \(\sim 1 - 10 \, M_\odot\)/yr. A \(10^9 \, M_\odot\) SMBH occupies a Schwarzschild radius of only \(\sim 20\) AU, naturally fitting inside a ~1-lightday region, and can plausibly accrete this much material from the surrounding galaxy over Myr – Gyr timescales.
- Rapid variability ($t $ hours) implies the emitting region cannot be larger than \(c \, \Delta t \sim\) a few light-hours, which is consistent with the size of an accretion disk around a \(10^9 \, M_\odot\) SMBH but inconsistent with any stellar-fusion engine.
The combination of small size, enormous luminosity, and high variability is exactly what accreting SMBHs produce and what no fusion-powered source can match.
The Eddington Limit
A feeding SMBH cannot be arbitrarily luminous. As the accretion luminosity rises, radiation pressure pushes gas outward, opposing gravity. When the outward radiation force equals the inward gravitational force, further inflow stalls. This critical luminosity is the Eddington luminosity:
\[L_{\text{Edd}} \approx 1.3 \times 10^{31} \, \text{W} \times \left( \frac{M}{M_\odot} \right) \approx 3.3 \times 10^4 \, L_\odot \times \left( \frac{M}{M_\odot} \right)\]
For a \(10^9 \, M_\odot\) SMBH, \(L_{\text{Edd}} \sim 3 \times 10^{13} \, L_\odot\) — close to the luminosity of bright quasars. That is not a coincidence: the brightest quasars are radiating at or near their Eddington limits. This gives us a lower bound on the SMBH mass for any quasar we see. Since \(L_{\text{Edd}}\) scales linearly with mass, a quasar of luminosity \(L\) must have
\[ \frac{M}{M_\odot} \gtrsim \frac{L}{1.3 \times 10^{31}\,\text{W}}. \]
The units matter: the right-hand side is a pure number, and multiplying by \(M_\odot\) gives the minimum mass.
Worked Example: Minimum SMBH Mass from Eddington
Given: A quasar with bolometric luminosity \(L \approx 10^{40}\) W (about \(3 \times 10^{13}\,L_\odot\)).
Relation: A source cannot radiate above its Eddington luminosity for long without blowing itself apart, so \(L \le L_{\text{Edd}}(M)\) with \[L_{\text{Edd}} \approx 1.3 \times 10^{31}\,\text{W} \times (M/M_\odot).\]
Rearrange to isolate the minimum SMBH mass: \[M_{\min} \gtrsim \frac{L}{1.3 \times 10^{31}\,\text{W}} \, M_\odot.\]
Substitute: \[M_{\min} \gtrsim \frac{10^{40}}{1.3 \times 10^{31}} \, M_\odot.\]
Evaluate: \[M_{\min} \gtrsim 7.7 \times 10^{8} \, M_\odot \approx 10^9 \, M_\odot.\]
Interpret: A galaxy hosting a \(10^{40}\) W quasar must have at its center a SMBH of at least ~\(10^9 \, M_\odot\) — two to three orders of magnitude more massive than Sgr A. This is how distant quasars anchor the high-mass end of the SMBH census: the brightness alone*, independent of any direct orbit measurement, sets a floor on the black hole mass.
The SMBH–Galaxy Connection
Here is a surprising observational fact: the mass of a galaxy’s central SMBH is tightly correlated with the mass (or more precisely, the velocity dispersion σ) of its central bulge. This correlation is the M–σ relation:
\[M_{\text{SMBH}} \propto \sigma^{\alpha}\]
with \(\alpha \approx 4 - 5\) from observations. A galaxy with a more massive bulge has a more massive black hole, and the correlation is so tight (scatter of only a factor of ~2 – 3) that it cannot be an accident.
What to notice: central black-hole mass increases with the random motions of stars in the galaxy bulge. The plot is not saying the black hole directly controls the whole bulge; it is showing that the black hole and bulge somehow grew in linked ways. (Credit: Astronomy: Roen Kelly, after Kayhan Gultekin et al.)
Read this as an evidence plot, not as a magic rule. The x-axis is a stellar measurement: the random speeds of stars in the bulge, inferred from Doppler broadening. The y-axis is a black-hole mass inferred from stellar or gas dynamics near the center. The trend says those two measurements know about each other.
This is strange. The bulge contains \(\sim 10^{11}\) stars and extends over ~kiloparsecs. The SMBH sits at the center and is, by galactic standards, tiny — \(R_S \sim 10\) AU even for the biggest SMBHs. The SMBH’s direct gravitational influence extends only over parsecs, not kiloparsecs. Yet they grow in lockstep.
The leading explanation is AGN feedback: when a SMBH becomes active, its radiation and jets heat and expel gas from the host galaxy, quenching both further star formation and further SMBH accretion. The SMBH is a thermostat for its bulge. But be careful: the correlation alone does not prove the mechanism. AGN feedback is compelling because it connects the correlation to a plausible energy source and to observed outflows, not because one plot can settle causality by itself.
What to notice: AGN energy does not stay near the event horizon. Jets, winds, and cavities connect light-hour scales near the black hole to galaxy-scale environments. (Credit: NASA)
This is a very active research frontier, and it is one reason AGN matter for more than just being extremely bright point sources — they are possible regulators of galaxy evolution.
We measure a galaxy’s bulge mass using — you guessed it — stars. Bulge velocity dispersion σ comes from spectroscopy of the integrated stellar light: Doppler-broadened absorption lines reveal the range of stellar velocities. Bulge mass comes from dynamical modeling using stellar orbits. Every point on the M–σ diagram is anchored in stellar kinematics.
Check Yourself 4: What Is Being Measured?
The M–σ relation connects a central black-hole mass to the velocity dispersion of bulge stars. Which part is the observable, which part is the model-dependent inference, and why does that distinction matter?
The observable is the broadening of stellar absorption lines, which tells us the range of stellar line-of-sight velocities in the bulge. The model-dependent inference is the bulge’s dynamical state and mass, plus the SMBH mass inferred from stellar or gas motions near the center. The distinction matters because the relation is not based on “seeing” a black hole grow with a galaxy; it is built by translating light into velocities, then velocities into gravitational mass.
The Cosmic Web
Zoom out one more time. Beyond individual galaxies is galaxy clustering. Galaxies are not scattered uniformly. They cluster into groups (like our Local Group of ~80 galaxies), then into clusters of hundreds to thousands of galaxies (like Virgo, Coma), and then into superclusters stretching tens of megaparsecs.
Between the clusters are vast voids — regions tens of megaparsecs across with very few galaxies. Connecting clusters are filaments — long, thin structures of galaxies and gas.
Together, this arrangement is called the cosmic web.
What to notice: a galaxy cluster is not a single object but a gravitational city of galaxies. The Antlia Cluster lets students see the step from individual galaxies to environments where galaxies live together. (Credit: NSF NOIRLab)
What to notice: every point is a galaxy with a measured redshift. The cosmic web is not a drawing; it is the pattern that appears when distances are added to sky positions. (Credit: NSF NOIRLab/DESI Collaboration)
The cosmic web is what dark matter looks like, traced by galaxies. In the early universe, tiny density fluctuations (seen as temperature fluctuations in the CMB; Lecture 26) grew under gravity. Regions with slightly more dark matter attracted more gas, which cooled, formed stars, and became galaxies. Regions with slightly less dark matter emptied into voids. The web is the gravitational sedimentation of the early universe’s initial conditions. We will see how those initial conditions are encoded in the CMB in Lecture 26.
Dark matter is not just a Milky Way problem. It shapes the entire cosmic web. The filaments and voids you see in maps like the Sloan Digital Sky Survey and DESI are the tracks left by dark matter as it gravitationally collapsed into structure over 13.8 billion years. Galaxies are the luminous decoration on a dark-matter skeleton.
Distance Ladder Check-In
Where are we in the distance ladder?
| Rung | Method | Reach |
|---|---|---|
| 1 | Radar ranging (Solar System) | ~AU |
| 2 | Stellar parallax (Gaia) | ~1 kpc |
| 3 | Spectroscopic parallax, main-sequence fitting | ~100 kpc |
| 4 | Cepheid variables | ~30 Mpc |
| 5 | Type Ia supernovae | ~thousands of Mpc |
| 6 | Hubble’s law (redshift → distance) | cosmological |
Rungs 4 – 6 are the subject of Lectures 24 – 26. By the end of Module 3, we will see how stars — specific, carefully calibrated kinds of stars — anchor the entire cosmic distance scale, and how that scale lets us measure the age, size, and expansion rate of the universe.
Deep Dives (Optional)
AGN come with a confusing zoo of historical names — Seyferts (Type 1 and 2), radio galaxies, quasars, BL Lacs, blazars. The Unified Model proposes that most of these differences come from viewing angle: the same basic object (SMBH + accretion disk + torus + jets) looks different depending on whether you see it face-on (quasar, blazar), edge-on through the torus (Type 2 Seyfert), or somewhere in between (Type 1 Seyfert). Some differences are real (e.g., radio-loud vs. radio-quiet AGN differ in jet power), but orientation does much of the zoological work.
What to notice: many AGN names partly encode viewing angle. The same basic engine can look like a Seyfert galaxy, quasar, or blazar depending on how the disk, torus, and jets point toward us. (Credit: NASA)
Cepheid variables are pulsating giants whose pulsation period is set by their mean density, which is in turn set by their luminosity. More luminous Cepheids pulsate more slowly. This is why \(L\) can be read off from the period \(P\) — an extraordinary fact that was discovered empirically by Henrietta Leavitt in 1912 and later understood theoretically as a consequence of stellar structure. Lecture 24 develops this argument in detail.
Misconceptions
WRONG. A quasar’s luminosity comes from accretion onto a supermassive black hole — gravitational potential energy converted to heat and light as gas spirals inward. Fusion is far too inefficient and would require an impossible amount of mass in an impossibly small volume to explain quasar luminosities.
MOSTLY WRONG. Ellipticals are dominated by old stars today, but they often reached that state via mergers of spiral galaxies, which turned disks into ellipsoids and used up available gas. Many giant ellipticals are dynamically young even if their stars are old. Morphology is a current state, not a birth certificate.
WRONG. The cosmic web is primarily a dark matter structure. Galaxies are tracers — they form along dark-matter filaments like frost condensing along the lines of a window’s temperature gradient. The bulk of the mass in the web is dark matter, not galaxies or hot gas.
Practice Problems
Solutions are available in the Lecture 23 Solutions.
Core Problems (Start Here)
Problem 1: Classify It. You are handed three galaxy images. Galaxy X is a smooth ellipsoid, red, with no visible dust lanes or H II regions. Galaxy Y has tightly wound spiral arms, a large central bulge, and some pink H II regions along the arms. Galaxy Z has no symmetry, is blue, and is full of bright knots of young stars.
Classify each as E, S(a/b/c), or Irr, and predict which has the most ongoing star formation.
Problem 2: Why Is Parallax Not Enough? The nearest large galaxy (Andromeda) is at ~780 kpc. Gaia parallax reaches a useful precision out to ~1 kpc. Explain, in two sentences, why Cepheid variables (and not parallax) are the tool for the Andromeda distance.
Problem 3: Quasar Luminosity. A quasar has an observed flux of \(F = 10^{-12}\) W/m² and (by redshift distance) is at \(d = 3\) Gpc (\(\approx 9 \times 10^{25}\) m). Compute its luminosity \(L\) and compare to \(L_\odot = 3.8 \times 10^{26}\) W. How many Suns is that?
Problem 4: Eddington Mass Lower Bound. The quasar from Problem 3 has \(L \approx 10^{39}\) W (check this once you’ve computed it). Using \(L_{\text{Edd}} \approx 1.3 \times 10^{31} \times (M/M_\odot)\) W, estimate the minimum SMBH mass required to power it.
Problem 5: Reading a Cosmic Web Map. Look at a survey map like SDSS or DESI (use the reference figure in class). Identify (a) a galaxy cluster, (b) a void, and (c) a filament. In 1 – 2 sentences, describe the dark-matter distribution you would expect in each region.
Challenge Problems (Deepen Your Understanding)
Challenge 1: Fusion vs. Accretion Efficiency. Hydrogen fusion converts ~0.7% of rest mass to energy. SMBH accretion can reach ~10%. Compute how much mass per second a star must fuse to sustain \(L_\odot = 3.8 \times 10^{26}\) W. Then compute how much mass per second a SMBH must accrete to sustain \(L = 10^{13} \, L_\odot\). Compare.
Challenge 2: M–σ Scaling — and an Outlier. If \(M_{\text{SMBH}} \propto \sigma^4\), and one galaxy has σ = 100 km/s while another has σ = 200 km/s, what is the ratio of their SMBH masses?
Now apply the scaling to the Milky Way (σ ~ 100 km/s, \(M_{\text{SMBH}} \approx 4 \times 10^6 \, M_\odot\)) and predict the expected SMBH mass in M87 (σ ~ 380 km/s). Compare to the observed value, \(M_{\text{SMBH}} \approx 6 \times 10^9 \, M_\odot\).
You will find that the prediction undershoots the observed M87 mass by roughly an order of magnitude. M87 is a well-known outlier on the M–σ relation — giant ellipticals at the centers of galaxy clusters tend to host overmassive SMBHs relative to their bulge σ. In 1 – 2 sentences, propose a physical reason this might be (think about how a galaxy ends up at the center of a cluster).
Challenge 3: Feedback as a Thermostat. Write a one-paragraph argument explaining why AGN feedback could naturally produce a tight M–σ relation. What happens if a SMBH tries to grow faster than the bulge? What happens if a bulge tries to grow faster than its SMBH?
Reading Summary
- Galaxies fall on the Hubble sequence (E, S/SB, Irr); morphology is a stellar-population readout encoding star-formation history and gas content.
- Measuring extragalactic distances is hard because parallax runs out at ~1 kpc. The fix is standard candles — objects of known \(L\) whose observed flux \(F\) gives \(d\) via \(L = 4\pi d^2 F\). Cepheid variables (Lecture 24) and Type Ia SNe (Lecture 25) are the key ones.
- Quasars and AGN are powered by accretion onto supermassive black holes, not fusion. Accretion reaches ~10 – 40% rest-mass efficiency (Schwarzschild vs. maximally rotating Kerr); fusion maxes out at ~0.7%. Rapid variability plus enormous \(L\) is uniquely consistent with an accreting SMBH.
- The Eddington limit \(L_{\text{Edd}} \propto M\) sets a lower bound on SMBH mass from observed luminosity alone; the brightest quasars need \(M \gtrsim 10^9 \, M_\odot\).
- The M–σ relation ties central SMBH mass to host-bulge velocity dispersion, implying co-evolution (most plausibly via AGN feedback). M87 and other brightest cluster galaxies are known overmassive outliers.
- Galaxies are arranged in a cosmic web of filaments, walls, and voids — the visible decoration on a dark-matter skeleton whose initial conditions we will read from the CMB in Lecture 26.
Glossary
No glossary terms for lecture 23.
Looking Ahead
Next lecture (Lecture 24) is where we close the distance-ladder gap that this reading opened. We will walk through Cepheid variables in detail (Henrietta Leavitt’s period-luminosity relation), connect them to the Type Ia supernovae that will appear in Lecture 25, and use them to derive Hubble’s law — the relation between a galaxy’s distance and its recession velocity that is the observational foundation of modern cosmology. We will also meet the Hubble tension, one of the most interesting open problems in modern astrophysics.
After that, Type Ia supernovae give us the accelerating universe and dark energy (Lecture 25), and the cosmic microwave background takes us all the way back to 380,000 years after the Big Bang (Lecture 26).
Dark matter is the first of three hidden things (Lecture 22, and already showing up in the cosmic web here). Dark energy is the second (Lecture 25). The origin of the elements — including the ones in your body — is the third (Lecture 26). Every one of them depends on stars.