Foundations
Lights out
The afterglow of lesson 1.4 was the last thing to happen for a long time. When the fog cleared, 380,000 years in, the universe was transparent for the first time and glowing orange-white at three thousand kelvin. Then it expanded, and the glow stretched: orange, then deep red, then, a few million years later, below the red heat that eyes can see. And there was nothing else. No stars, no galaxies, not a single point of light anywhere in a universe already millions of light-years across.
It stayed that way for something like a hundred million years. On the cosmic calendar, that is from a quarter past midnight on 1 January to about the 3rd or 4th. Astronomers call it the Dark Ages, and it is the last period of the universe’s history that we have never seen.
What was in the dark
The universe was not empty. It held gas, three quarters hydrogen and one quarter helium by mass, and nothing heavier, because nothing heavier had been made yet. It held about five times as much dark matter, the invisible stuff of lesson 1.7. And it held the ripples: the slight unevenness in the afterglow, one part in a hundred thousand, which was also an unevenness in how the matter was spread.
Gravity does not need light. Through the Dark Ages it worked on those ripples. Places slightly denser than average pulled in more matter and grew denser; places slightly emptier grew emptier. Dark matter, which feels nothing but gravity, fell together first, into clumps. The gas, which also feels pressure, followed more slowly, streaming into the clumps of dark matter that had already formed.
The first clumps
The smallest clumps in which a star could form were about a million times the mass of the Sun, a few hundred light-years across, and there were vast numbers of them. Gas falling into one heats up, and to collapse further it has to cool. Today’s gas clouds cool easily, by radiating away heat through carbon, oxygen and dust. The first clouds had none of those. Their only way of shedding heat was through molecules of hydrogen, and molecular hydrogen is a poor coolant. The clouds stayed warm, around two hundred kelvin, and warm gas resists being squeezed. So instead of fragmenting into many small stars, as clouds do today, the first clouds collapsed into a few enormous ones.
The first stars
Everything we can say about the first stars comes from simulation, because no one has yet seen one for certain. The simulations agree on their character. They were made of hydrogen and helium only. They were massive, tens to hundreds of times the Sun’s mass. They were hot, blue and violently bright, and they lived fast: a few million years, against the Sun’s ten billion. They died as supernovae, scattering the first carbon, oxygen and iron into the surrounding gas, or collapsed straight into black holes.
The moment the first one ignited, the Dark Ages ended. Somewhere in the universe, for the first time since the afterglow faded, there was light.
Cosmic Dawn
The first stars were few and far apart, and the universe stayed mostly dark for a while. But every star that died enriched the gas around it, and enriched gas cools easily and makes stars easily. Star formation snowballed. Clumps merged into the first small galaxies. Their ultraviolet light did something else, too: it stripped the electrons back off the hydrogen atoms between the galaxies, undoing what recombination had done. Bubbles of ionised gas grew around each galaxy, merged, and by about a billion years after the Big Bang, near the end of January on the calendar, the fog of neutral hydrogen had been burned away. Astronomers call this reionisation, and it is the last great transition of the early universe.
Seeing the dawn
We can see this happening. The James Webb Space Telescope has confirmed galaxies at a redshift of 14, seen as they were about 290 million years after the Big Bang, on the 8th of January. There are more of them, and brighter, than most models expected, which has sent theorists back to their simulations to ask how the first stars formed so fast. The end of reionisation shows up in the spectra of distant quasars, where neutral hydrogen blacks out a stretch of light for everything beyond redshift 6.
The Dark Ages themselves are harder. The only signal from that time is a faint radio whisper from hydrogen atoms, the 21-centimetre line, stretched by expansion to wavelengths of metres and buried under the far louder radio noise of our own galaxy. In 2018 an experiment in Australia claimed to have found it. In 2022 an Indian experiment, SARAS 3, floating on a lake in Karnataka to escape the noise of the ground, found nothing. The search is on.
On the calendar
A quarter past midnight, 1 January: the afterglow. A few minutes later: the glow fades from sight, and the Dark Ages begin. The 3rd to 6th of January: the first stars. The 8th: the earliest galaxies yet confirmed. The 27th: reionisation complete. And then the universe, lit and transparent, settles into the business of making stars, which is the next chapter.
Deeper
How fast the afterglow faded
The afterglow’s temperature falls as 1/(1+z). A glow becomes invisible to the eye below roughly 800 kelvin, dull red heat, which the afterglow reached when the universe had stretched about four times beyond its size at recombination, near redshift 300. In a matter-dominated universe time goes as (1+z) to the power minus three halves, so that was about three million years after the Big Bang. From then until the first stars, the universe was genuinely dark at every visible wavelength.
The gas gets colder than the light
After the afterglow was released, the gas and the radiation stayed in thermal contact for a while through the few remaining free electrons, then parted company around redshift 150. From then on the expanding gas cooled faster than the light, as 1/(1+z)², while the afterglow cooled as 1/(1+z). By Cosmic Dawn the gas was several times colder than the radiation passing through it. That difference is what makes the 21-centimetre signal an absorption feature: cold hydrogen absorbs the afterglow’s radio light. The predicted depth is a few tenths of a kelvin; the 2018 claim was about twice that, which is one reason it was met with doubt.
Why the first stars were giants
A gas cloud collapses when its own gravity beats its pressure, which requires a minimum mass, the Jeans mass, that rises steeply with temperature. Molecular hydrogen cannot cool gas below about two hundred kelvin, so the Jeans mass in the first clouds was hundreds of solar masses, against a fraction of a solar mass in today’s cold, metal-rich clouds. Simulations that follow the collapse find the gas accumulating into one or a few dense cores that grow to tens or hundreds of solar masses before their own radiation halts the infall. The exact mass distribution, and how often the clouds fragmented into pairs and clusters, is still debated.
Two kinds of death
A star above about a hundred and forty solar masses is predicted to be destroyed entirely by a pair-instability supernova, leaving no remnant and scattering all of its freshly made elements. Stars of a few tens of solar masses explode as ordinary core-collapse supernovae or collapse directly into black holes of tens of solar masses. Those black holes are one candidate for the seeds of the supermassive black holes found in the centres of galaxies less than a billion years later, which are otherwise hard to grow so fast.
Reionisation in numbers
The afterglow’s photons had a 5 percent chance of scattering off free electrons on their way to us, which fixes the midpoint of reionisation at redshift 7.7 ± 0.7, about 700 million years in. Quasars at redshift above 6 show the Gunn–Peterson trough, a near-total blackout blueward of the hydrogen line, while quasars at slightly lower redshift do not: reionisation ended between redshift 6 and 7. That the process took several hundred million years, rather than happening all at once, is itself evidence that it was driven by many faint sources rather than a few bright ones.
The 21-centimetre line
A hydrogen atom’s electron can flip its spin, emitting or absorbing radio light at a wavelength of 21 centimetres. From redshift 17 that arrives at 78 megahertz, in the middle of the FM radio band, and thousands of times fainter than the Milky Way’s own radio glow at that frequency. Separating the two requires modelling the sky and the instrument to a part in ten thousand. The disagreement between EDGES and SARAS 3 turns on exactly that modelling. Larger instruments, including the Square Kilometre Array, are designed to map the signal rather than just detect its average.
Frontier
The first stars, still unseen
No Population III star has been confirmed. Candidates have been proposed, in the form of pockets of gas near early galaxies that glow in helium lines without any sign of heavier elements, but none has held up unambiguously. Webb’s sensitivity makes a detection possible, either directly in a very early galaxy or through the peculiar element ratios of a pair-instability supernova. This is active research, and a detection would be the first sight of the end of the Dark Ages itself.
Too many, too early
Webb keeps finding luminous galaxies earlier than expected: the current record is spectroscopically confirmed at redshift 14.3. The Big Bang and its afterglow are untouched by this; what is under pressure is the assumed efficiency of star formation in the first few hundred million years. Whether early galaxies converted gas to stars unusually fast, formed unusually massive stars, or shone unusually brightly per star is being worked out now.
Cosmic Dawn on the radio
Whether the 21-centimetre signal from Cosmic Dawn has been seen is unresolved. EDGES says yes and too deep; SARAS 3 says no. Experiments on the far side of the Moon, shielded from Earth’s radio noise, are being proposed to settle it. A confirmed signal would open the last unobserved chapter of the universe’s history.
Transcript and shot list
Target runtime about 7 minutes at a measured narration pace, with room for the visual beats to breathe. 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. Timecodes are targets and will move to match the recorded narration.
[00:00] S01 · ANIM · A pure black frame. Nothing. After several seconds a faint temperature readout appears: 300 K.
This is what the universe looked like for a hundred million years. Not black because it was empty. Black because nothing in it shone. There were no stars. There had never been a star. If you could have floated there, you would have seen nothing at all, in any direction, for as long as you cared to wait.
[00:35] S02 · ANIM · Cosmic calendar; zoom from a quarter past midnight on 1 January to the 3rd, 4th and 5th.
On the cosmic calendar, this is the first three or four days of January. The afterglow at a quarter past midnight. Then darkness, until somewhere around the third.
[00:55] S03 · ANIM · The afterglow’s glow: orange-white, then red, then a dull ember, then nothing; the readout runs 3,000 K to 800 K.
Here’s how the lights went out. When the fog cleared, the universe glowed orange-white at three thousand degrees. Then it expanded, and the glow stretched: orange, deep red, a dull ember. A few million years in, it dropped below the red heat that eyes can see. And that was that. No other light existed.
[01:30] S04 · ANIM · A faint field of ripples; dark-matter clumps deepen; streams of gas flow into them; still no light.
The universe wasn’t empty. It held gas, three quarters hydrogen and a quarter helium, and nothing heavier, because nothing heavier had been made. It held five times as much dark matter. And it held the ripples from the afterglow: places a little denser than average, places a little emptier. Gravity doesn’t need light. For a hundred million years it worked in the dark, pulling the dense places denser. Dark matter clumped first. Gas followed, streaming into the clumps.
[02:20] S05 · ANIM · A gas cloud in a clump tries to shrink; a thermometer sticks at 200 K; the cloud stays big; a small caption: “only molecular hydrogen to cool with”.
Now the gas has a problem. To collapse into a star, a cloud has to cool, and today’s clouds cool easily, shedding heat through carbon and oxygen and dust. The first clouds had none of that. Their only coolant was molecular hydrogen, and it’s a bad one. The clouds stayed warm, and warm gas resists being squeezed. So they didn’t break into hundreds of small stars the way clouds do now. They collapsed into a few enormous ones.
[03:00] S06 · AI · A single blue-white star ignites in total darkness, blazing, alone. S07 · ANIM · Its size against the Sun; its lifetime against the Sun’s on a bar.
And then, somewhere, this. The first star. Made of hydrogen and helium only. Perhaps a hundred times the Sun’s mass, far hotter, blue-white and violently bright. It lived a few million years, against the Sun’s ten billion. Nobody has seen one. Everything we know about them comes from simulations, and the simulations agree on this much. The moment it lit, the Dark Ages were over.
[03:50] S08 · ANIM · The star explodes; new elements scatter into the gas; nearby clouds cool faster and light up; a small galaxy assembles.
When it died, as a supernova or by collapsing into a black hole, it scattered the first carbon, oxygen and iron into the gas around it. And enriched gas cools easily and makes stars easily. Star formation snowballed. Clumps merged into the first small galaxies. Light spread.
[04:20] S09 · ANIM · Bubbles of ionised gas grow around galaxies, merge, and fill the frame; a redshift counter runs from 20 to 6.
The light did one more thing. Ultraviolet from the first stars stripped the electrons back off the hydrogen between the galaxies, undoing recombination. Bubbles of ionised gas grew around every galaxy, merged, and by about a billion years in, the fog of neutral hydrogen was burned away. Reionisation. The last great transition of the early universe, finished around the 27th of January.
[05:00] S10 · REAL · A Webb deep-field image with a redshift-14 galaxy circled (NASA, ESA, CSA, STScI; confirm credit line).
We can watch it happen. The James Webb Space Telescope has confirmed galaxies at redshift fourteen, seen as they were 290 million years after the Big Bang, the 8th of January. More of them, and brighter, than most models expected, which has theorists back at their simulations asking how the first stars formed so fast.
[05:35] S11 · ANIM · A hydrogen atom flips its spin; a 21-centimetre wave stretches to metres; two instruments, one in Australia, one floating on a lake in Karnataka; one reads a signal, the other reads flat.
The Dark Ages themselves are harder. The only signal from that time is a whisper from hydrogen atoms, a radio line at twenty-one centimetres, stretched by expansion to metres and buried under the roar of our own galaxy. In 2018 an experiment in Australia claimed to have found it. In 2022 an Indian experiment, SARAS 3, floating on a lake in Karnataka to get away from the noise of the ground, found nothing. Active research, and some of it is happening here.
[06:20] S12 · ANIM · Labels: ESTABLISHED for the darkness and reionisation; ACTIVE RESEARCH for the first stars, the early galaxies and the radio signal.
The scorecard. That the universe went dark, and that light from the first stars and galaxies burned the fog away by a billion years in: established. What the first stars were, exactly when they formed, why the earliest galaxies are so bright, and whether anyone has heard the Dark Ages on the radio: active research.
[06:50] S13 · AI into ANIM · The single star becomes thousands, then a sky full; the calendar returns.
A hundred million years of darkness, and then a light that has never gone out. Every star that has ever shone descends from those first few, and so does every element in you heavier than helium. That is the next chapter. Next lesson, though, is the honest one: everything this chapter cannot yet explain.
Shot list
| Shot | Time | Type | What we see | Scene or asset | Status |
|---|---|---|---|---|---|
| S01 | 00:00 | ANIM | Black frame with a faint readout | NothingShines |
planned |
| S02 | 00:35 | ANIM | Calendar, 1 to 5 January | shared scene CosmicCalendarZoom |
planned |
| S03 | 00:55 | ANIM | The afterglow fades below red heat | AfterglowFades |
planned |
| S04 | 01:30 | ANIM | Gravity in the dark; clumps and streams | GravityInTheDark |
planned |
| S05 | 02:20 | ANIM | The cloud that cannot cool | WarmCloud |
planned |
| S06 | 03:00 | AI | The first star ignites alone | Higgsfield, style reference bh-cosmos-v1, 16:9, 8 s | planned |
| S07 | 03:20 | ANIM | Size and lifetime against the Sun | FirstStarScale |
planned |
| S08 | 03:50 | ANIM | Death, enrichment, first galaxy | Snowball |
planned |
| S09 | 04:20 | ANIM | Bubbles of reionisation | Reionisation |
planned |
| S10 | 05:00 | REAL | Webb deep field, redshift-14 galaxy | NASA, ESA, CSA, STScI; confirm licence and credit | planned |
| S11 | 05:35 | ANIM | The 21-centimetre line; EDGES and SARAS 3 | TwentyOneCentimetres |
planned |
| S12 | 06:20 | ANIM | Confidence labels | shared scene ConfidenceLabels |
planned |
| S13 | 06:50 | AI + ANIM | One star becomes a sky; calendar returns | Higgsfield 8 s clip, then SkyFills |
planned |
Two AI shots, ten code-rendered shots, one agency image. Estimated Higgsfield use: 3 to 6 generations including retries.