In May 2007, a young star in Centaurus nearly vanished. For weeks something drifted across it, at times blocking 95 percent of its light, and the flicker inside the shadow looked like rings: more than 30 of them, in a system over 200 times wider than Saturn's. Astronomers have been searching for whatever wears those rings ever since. They haven't found it.
My favorite story tonight comes last, though. It's about a rock from Gabon that ran itself as a nuclear reactor two billion years ago and kept a quiet record of the laws of physics. On the way there, a Victorian train runs late because of the Sun, and a black hole loses its glowing crown and then starts to tick.
1. 🪐 Whose rings crossed J1407?
The star has a clumsy catalog name, 1SWASP J140747.93−394542.6, so astronomers just call it J1407. It's a young Sun-like star, about 16 million years old, in Scorpius–Centaurus, a loose family of young stars in the southern sky.
Its strange behavior turned up in old data. Eric Mamajek of the University of Rochester and his colleagues were going through J1407's record from SuperWASP, a survey designed to catch giant planets as they pass in front of their stars. A planet makes a small, tidy dip that lasts a few hours. J1407 had done something else. In May 2007 it dimmed in a long, ragged series of eclipses that went on for weeks. On some days 95 percent of its light was gone, and inside that shadow the brightness jumped up and down over just tens of minutes.
When they reported it in 2012, Mamajek's team proposed that a dusty disk around a young giant planet or brown dwarf had passed in front of the star, the kind of disk where moons are thought to form. Three years later, Matthew Kenworthy of Leiden Observatory and Mamajek modeled the light curve in detail. Each quick flicker became the edge of a ring. The model needed more than 30 rings, each tens of millions of kilometers across, and the whole system came out nearly 120 million kilometers wide. That's four-fifths of the distance from Earth to the Sun.
"If we could replace Saturn's rings with the rings around J1407b, they would be easily visible at night and be many times larger than the full moon," Kenworthy said. Mamajek called the ringed object "kind of a super Saturn." The dust in the rings added up to roughly the mass of Earth. One gap was clean enough that Kenworthy suggested a young moon, somewhere between the mass of Earth and Mars, had swept it out.
Nobody has seen the rings directly. They're an inference from the shape of the shadow, and the object at their center, named J1407b, has never been seen either. The 2015 estimate put it at roughly 10 to 40 times Jupiter's mass, on an orbit about a decade long. The team asked amateur astronomers to watch J1407 and report the next eclipse.
It never came. Kenworthy and his colleagues went back through 387 old photographic plates from Harvard and from the Bamberg and Sonneberg observatories in Germany, covering 1890 to 1990, and through new measurements from 2012 to 2018. Their 2018 search found no other deep eclipse anywhere, which ruled out a large share of the possible orbits between 5 and 20 years long. There were also doubts that rings so large could stay stable. Put together, the team suggested that J1407b might not orbit J1407 at all.
That opens a lonelier possibility: a ringed object drifting freely among the young stars of Scorpius–Centaurus, which happened to pass in front of J1407 once. Kenworthy's team had already pointed ALMA, the big radio telescope array in Chile, at the field in 2017. They found nothing at the star itself, but they did pick up a faint source nearby, at a spot that fit a free-floating ring system that had crossed the star in 2007 and kept moving. It could also have been a distant galaxy in the background.
A drifting object would have kept drifting, so in mid-2024 ALMA looked again, at the old spot and at the spot where the source should have moved to. There was nothing at either. The new data rule out the kind of thick disk, full of grains bigger than a millimeter, that the team had favored, whatever its mass. The 2017 signal was probably a noise spike, or something that has since faded.
The 2007 eclipse is still the only one anyone has seen. Whatever cast that shadow might be circling J1407 on an orbit longer than anyone has checked, or it might be out among the young stars of Centaurus on its own, carrying its rings with it.
2. ☀️ A train left Exeter 16 minutes late, in the wrong year
The 10:05 p.m. train out of Exeter, in the southwest of England, left 16 minutes late. According to an unsigned account published in the journal Nature in 1871, the culprit was a "very intense magnetic disturbance" that interfered with the railway's signals. The line used an electric telegraph to check that the track ahead was clear before a train could go, and that night the telegraph was misbehaving.
The account dated the delay to October 18, 1841. That made it famous. It was cited again and again as the earliest known case of space weather disrupting human technology.
The physics behind it is real. When the Sun throws a burst of charged gas and magnetic field at Earth, our planet's magnetic field shakes, and a changing magnetic field drives electric currents through any long conductor. Today those are power lines. In the 1840s they were telegraph wires, which were spreading fast. During the great Carrington storm of 1859, some telegraph lines carried so much of this current that operators kept sending messages with their batteries disconnected.
This year, Jim Wild of Lancaster University, who is also president of the Royal Astronomical Society, led a team from Lancaster, the British Geological Survey, Natural Resources Canada, Baylor University, RMIT University and RAL Space that went back to check the Exeter story. They found a problem straight away. The railway line in the account, from Exeter to Starcross, didn't open until 1846.
"We know it was a typo because the Exeter-to-Starcross train line mentioned in that paper didn't exist until 1846," said Brett Carter of RMIT.
A typo in which digit, though? The team worked through railway timetables, old newspapers, records of sunspots and auroras, and digitized measurements of Earth's magnetic field. The timetables turned out to be the sharpest clue. A 10:05 p.m. departure from Exeter only appeared for a little over a year, starting in 1848. One timetable printed in Trewman's Exeter Flying Post in October 1848 lists the train as the "10 5."
Then the sky gave its own evidence. On October 18, 1848, magnetic records show a strong disturbance, and people across Britain and Europe reported sunspots and auroras. The 1871 account had the day and month right. The year was off by exactly seven.
"Our research has a hint of a detective story," said Mike Hapgood of RAL Space, the UK's national space laboratory.
The correction costs Exeter its title. The earliest credible report now known is from the Midland Railway near Derby, where a geomagnetic disturbance interfered with the telegraph on March 19, 1847. Exeter's delay, a year and a half later, is still one of the first clear cases of the Sun disrupting a piece of critical infrastructure.
What I like about this one is that the decisive instrument was a train timetable, read closely enough to notice that a train couldn't run on a line that hadn't been built yet. The study appeared in the journal Space Weather in September.
Wild's point is that this kind of trouble is as old as electrical technology itself. "What this research highlights is that space weather is not a new threat but a long-standing natural hazard," he said. Since 1848 the wires have turned into power grids, satellites and navigation signals. The Sun hasn't changed, and its next cycle of activity begins in the 2030s.
3. 🕳️ The black hole that lost its crown
Ten thousand times dimmer in less than a year. That's how far the X-ray glow from the center of a galaxy called 1ES 1927+654 dropped in 2018, and it made no sense to the astronomers watching it.
"We expect that luminosity changes this big should vary on timescales of many thousands to millions of years," said Erin Kara of MIT. At one point its brightness changed by a factor of 100 in eight hours.
The galaxy is about 270 million light-years away, in the constellation Draco, and its central black hole weighs about 1.4 million Suns. Like other feeding black holes, it's surrounded by a disk of swirling gas, and close in, by a corona: a cloud of extremely hot plasma that shines in X-rays. Nobody knows exactly how a corona forms, but astronomers think it has to do with magnetic fields twisting, breaking and reconnecting in the inner disk.
The trouble started in March 2018, when the sky survey ASAS-SN saw this unremarkable black hole suddenly flare to about 40 times its usual brightness. Telescopes turned to look. One of them was NICER, a small X-ray telescope on the International Space Station, which could check in every day. "It's this little washing machine bouncing around the ISS," Kara said. Over the following months, NICER watched the corona vanish. "It became undetectable, which we have never seen before."
The team, led by Claudio Ricci of Diego Portales University in Chile, could only guess at the cause. One idea is that a star wandered too close and was torn apart. Its debris would have crashed through the disk, dumped much of the inner gas into the black hole and scrambled the magnetic fields that keep the corona going. Then, over a few months, the black hole rebuilt its corona, nearly back to its old brightness.
That would have been enough for one black hole. It kept going.
Going through data from 2022, Megan Masterson, a physics graduate student at MIT, and Kara noticed that the X-rays were flickering in a rhythm, brightening and fading every 18 minutes. Over the next two years the beat sped up to every 7 minutes. A speeding-up rhythm suggests something in orbit, falling inward as it loses energy to gravitational waves. If that was the case here, it was moving at about half the speed of light.
Masterson and Kara's best candidate for the orbiting object was a white dwarf, the dense leftover core of a star, with about a tenth of the Sun's mass, circling within a few million miles of the event horizon. White dwarfs are compact enough to survive that close. And as it sheds gas into the black hole, the loss could give it a small kick outward that keeps it from falling in. If that's right, the system should give off gravitational waves that LISA, a space observatory planned for the mid-2030s, could detect. The rhythm could also come from the corona itself, wobbling and shrinking. Another team argued for exactly that in a 2025 preprint.
Then came the third surprise. By 2023, rising X-rays and a strong radio flare sent Eileen Meyer of the University of Maryland, Baltimore County to the Very Long Baseline Array, a network of radio dishes across the United States. From February 2024, its images showed jets emerging on both sides of the black hole, moving outward at nearly a third of the speed of light and stretching about half a light-year in total. "The launch of a black hole jet has never been observed before in real time," Meyer said.
And the rhythm stopped speeding up. Kara had said that if the white dwarf idea was right, it might be "right at the turn around point." Monitoring through January 2026, reported in a preprint this April, shows the flicker still going strong, more than 900 cycles of it, but holding steady at about one beat every seven minutes. On each beat the X-rays still jump by as much as 80 percent. The team writes that these observations strain the models they tried, including the ones based on eclipses.
So a black hole lost its corona, grew it back, started ticking faster and faster, launched a jet and then settled into a steady beat, all in eight years. "The one thing I've learned with this source is to never stop looking at it because it will probably teach us something new," Masterson said.
4. ⭐ The new star of 1600 never went away
On August 18, 1600, the Dutch mapmaker Willem Janszoon Blaeu recorded a new star in the constellation Cygnus, the Swan. It had reached third magnitude, easy to see without a telescope, which was lucky, because nobody would point a telescope at the sky for another nine years.
Blaeu logged it as a nova, a "new star." It's now called P Cygni, and it never went away. It stayed bright for about six years, then faded until it was invisible to the naked eye by 1626. It came back in 1654 and shone at third magnitude again until 1659, flickered for a couple of decades, and then crept upward until it reached magnitude 5.2 by 1780. Today it sits at about magnitude 4.8, faint but visible from a dark site.
P Cygni is a luminous blue variable: a very massive star in an unstable phase late in its life, when it can throw off large amounts of itself in fits. It's about 5,500 light-years away, with about 75 times the Sun's radius and some 560,000 times its luminosity. Its wind carries away about 0.00003 solar masses of gas every year, roughly ten Earths' worth, at 185 kilometers per second. It's one of only two stars in our galaxy seen during a giant eruption of this kind. The other is Eta Carinae, in the 1840s.
The 1600 outburst left a shell of gas around the star that's still there. Nathan Smith and Patrick Hartigan measured it in infrared light and found about a tenth of a solar mass in it. That's more than the star's steady wind has blown away in all the centuries since, though the eruption carried far less energy than Eta Carinae's. Dust that formed after the outburst is thought to have hidden the star for a while. Smith and Hartigan suggested it has since been mostly destroyed, freeing iron that now glows in the shell.
The star's name lives on in a second way, and this is the part I find most satisfying. The first photographs of P Cygni's spectrum, taken in 1897, showed something odd. Each bright emission line had a dark notch on its blue side. At first astronomers read each one as two different lines blended together.
Around 1930, the astronomers McCrea and Beals worked out what was going on. The star sits inside an expanding envelope of gas. Gas on every side of the star glows, and that makes the bright line. The gas between us and the star is also moving toward us, and it absorbs some of the starlight passing through it. Because that gas is approaching, it absorbs at a slightly bluer color than the line itself, thanks to the Doppler effect, which cuts a dark notch into the blue edge.
Astronomers still call that shape a P Cygni profile. Spotting one tells you at a glance that a star is blowing off gas, and how far the notch is shifted tells you how fast. McCrea and Beals explained the same shape in novae and Wolf-Rayet stars, and it has turned up in many other outflows since.
There's one more thing hiding in Blaeu's legacy. Between 1700 and 1988, P Cygni brightened on average by about 0.15 magnitude per century. That's slow, but it's steady, and astronomers read it as the star changing as it evolves. It made P Cygni the first star to show the effects of stellar evolution in its own historical record.
Most stars change too slowly for us to notice in a lifetime, or in a civilization. P Cygni has been brightening by a sliver every century since Blaeu's day. With four centuries of notes, we can just about watch it grow old.
5. 🌌 Jets light a trail through the haze around galaxies
Galaxies are surrounded by more fuel than they use. Around a big galaxy like the Milky Way lies the circumgalactic medium, a thin halo of gas that reaches roughly 10 to 20 times farther than the galaxy's visible stars. Some of that gas can cool, sink inward and become new stars. Yet galaxies make far fewer stars than all that gas suggests they should, and something seems to stop it from cooling.
The leading suspect is the supermassive black hole at the center. It's an odd suspect, because of the scale. A big black hole is roughly the size of our solar system. The galaxy around it could hold about 100 billion solar systems. "The surprising question is: how can something so small energetically impact something so enormous?" said Namrata Roy, who did the work at Arizona State University and is now an assistant professor at the Raman Research Institute in Bengaluru.
Some black holes launch jets: narrow beams of hot plasma that shoot far beyond their galaxies and glow in radio waves. Roy and Sanchayeeta Borthakur of ASU, with Timothy Heckman and Tanmay Singh, wanted to know whether those jets touch the halo gas. If a jet heats or ionizes hydrogen, the gas should give off H-alpha, a particular shade of red that hydrogen emits when its electrons are knocked loose and then recaptured.
That glow is far too faint to see around any one galaxy, so they used a trick. They found radio galaxies with jets in the LOFAR Two-metre Sky Survey, made with a European radio telescope, and then looked for quasars, very bright distant objects, sitting behind each galaxy's halo in the sky. DESI, a big spectroscopic survey, had recorded the light of those quasars. Each quasar's spectrum also catches a little light from whatever glows in front of it. Adding up 324 of those spectra, lined up at each radio galaxy's own redshift, could bring out a signal that no single one could show.
The key was keeping track of direction. For each quasar, the team noted the angle between its line of sight and the galaxy's jet.
When they added everything together, they found nothing. But when they kept only the sightlines within 20 degrees of a jet, a strong H-alpha signal appeared, about 100 times brighter than in ordinary galaxy halos. The glow had two bright zones. One sits just outside the galaxy's visible body, where the jet first runs into the halo. The other lies far out, near the giant radio lobes where jets are thought to dump much of their energy.
That also explains why earlier searches came up empty. They averaged over every direction, and the faint glow along the jets washed out. A jet behaves like a searchlight in fog: only the droplets inside the beam light up.
There was a useful check. Magnesium atoms in cooler gas absorb a little of each quasar's light, and that absorption looked the same in every direction, near the jets and away from them. The cool gas seems to fill the halo all around, whatever the jets are doing. Where a jet passes, it raises the density, pressure or ionization of the clouds in its way until they glow.
Borthakur called it "a pathbreaking result that solves the long-standing mystery of how black holes influence galaxies." I'd put it more carefully. The study makes a strong case that jets reach and change gas far outside their galaxies. By heating and stirring that gas, they could make it harder for it to cool and fall in, and act as a brake on star formation. Showing that this is what actually keeps galaxies from forming stars is the next step. The results were published in The Astrophysical Journal Letters.
"A black hole is incredibly small compared to a galaxy, but its impact can reach hundreds of thousands of light-years," Roy said. "The jet carries the energy outward, and the gas lights up along its path."
6. 🧭 Quasars and the oldest light disagree about our speed
We're moving. The Sun, Earth and everything near us are traveling through the universe, and the oldest light in the sky shows it. The cosmic microwave background, the afterglow of the Big Bang, is uniform to about one part in 100,000, except for one large pattern: it's warmer by about one part in a thousand on one side of the sky and cooler on the other. Cosmologists read that as a Doppler effect from our own motion.
George Ellis and John Baldwin saw a test in this back in 1984. If we really are moving relative to the universe as a whole, the motion should also show up in how distant galaxies and quasars are scattered across the sky. Ahead of us they should look slightly more crowded, through an effect called aberration, the same one that makes rain seem to come at you from the front when you run. They should also look slightly brighter, so a few more of them would make it into any survey with a brightness cutoff. The microwave background tells you how big both effects should be. The sources have to be very distant, because nearby galaxies clump together and can fake the pattern.
The test needs huge catalogs covering almost the whole sky, so it had to wait. In 2021, Nathan Secrest and colleagues ran it on 1.36 million quasars seen in infrared by NASA's WISE satellite. The pattern in the quasars pointed in roughly the same direction as the microwave background's. It was more than twice as large. The odds of a fluke that big were about one in two million.
Surveys with ground-based radio telescopes, which have entirely different quirks, have found the same excess. In a 2025 review in Reviews of Modern Physics, Secrest and his colleagues wrote that the discrepancy now exceeds 5 sigma, the level physicists usually treat as a discovery. Either matter and radiation don't share the same "at rest" frame, or something in at least one of them is genuinely lopsided. Either way, they argue, it's a serious challenge to the cosmological principle, the assumption that the universe looks the same in every direction when you average over large enough scales. Modern cosmology is built on that assumption.
That's a big claim, and the pushback has been serious. In a paper accepted this year by The Astrophysical Journal, Masroor Bashir, Pravabati Chingangbam and Stephen Appleby reran the quasar test with detailed simulations. Their simulations included how quasars cluster and the exact patch of sky the catalog covers. The significance dropped from 4.9 sigma to between about 3.3 and 3.6 sigma, depending on assumptions. That's weaker, though they found that clustering alone can't make the anomaly go away.
Oliver Oayda and Geraint Lewis came at it from another side, in a study accepted by MNRAS this year. WISE scanned the sky in a fixed pattern, so its measurement errors vary from place to place, and that can bias which faint quasars make the catalog. When they modeled this, the dipole was still about twice what the microwave background predicts. But its direction now differed from the microwave background's at about 3 sigma, and the models that fit best were the ones in which the catalog's errors are bigger than stated. They read that as strong evidence that the errors are underestimated or that some unknown problem in the data is at work. They still concluded that the excess is a persistent problem for the standard model.
Here's where I land, for what it's worth. When modeling one catalog's quirks keeps moving the answer, I'd want a completely independent sky survey before rewriting cosmology. The good news is that several are coming. Euclid, SPHEREx, the Vera Rubin Observatory and the Square Kilometre Array will all count distant galaxies by the million, each with different blind spots.
If they all find the same oversized dipole, one of the review's authors has argued, there's no easy patch. The standard model of cosmology, and the simple picture of space and time beneath it, would have to be rebuilt from square one.
7. ⚛️ Oklo's reactor kept a two-billion-year-old record of physics
Uranium is the same everywhere. Dig it out of the ground anywhere on Earth, or find it in Moon rocks or meteorites, and 0.720 percent of it is uranium-235, the kind whose atoms split easily. The rest is almost entirely uranium-238.
In 1972, at a nuclear fuel plant in the south of France, a batch of ore from the Oklo mine in Gabon came in at 0.717 percent. That's a tiny shortfall, but uranium-235 doesn't go missing for no reason. The first suspicion was that the ore wasn't natural at all, and had already been through a reactor. Francis Perrin and his colleagues at the French Atomic Energy Commission checked and found that it was natural ore, and that it was full of the leftover products of fission. The uranium had split by itself, in the ground, about two billion years ago.
A reactor needs two things. The first is enough uranium-235, and two billion years ago there was more of it. Uranium-235 decays faster than uranium-238, so the further back you go, the bigger its share of natural uranium. The second is something to slow the neutrons down so they can split more atoms. At Oklo, that was groundwater. "Like in a man-made light-water nuclear reactor, the fission reactions, without anything to slow down the neutrons, to moderate them, simply stop," said Peter Woods of the International Atomic Energy Agency.
The reactor didn't run steadily. Physicists at Washington University in St. Louis studied xenon gas, made by fission, that was trapped in a mineral from the site. The mix of xenon isotopes, they reported in 2004, showed that the reactor ran in pulses: about 30 minutes on, then about two and a half hours off. The usual explanation is the water. As the reaction heated the rock, water would boil off and the chain reaction would stall, then restart once water seeped back in.
That alone would be a good story. What makes Oklo profound is a point made by a physicist named Shlyakhter: this rock can test whether the laws of physics have changed.
The test hinges on a rare-earth element called samarium. One of its forms, samarium-149, is extremely good at swallowing slow neutrons, but only near one particular neutron energy, a sharp sweet spot called a resonance. In the cores of the Oklo reactors, samarium-149 is almost gone. The neutrons burned it up. How fast that happened depended on exactly where the resonance sat, and its position depends in part on the electric repulsion between the protons packed into the samarium nucleus.
The strength of that electric repulsion is set by a number called the fine-structure constant, about 1/137. It governs how light and charged matter interact everywhere in the universe. If it had been even slightly different two billion years ago, the resonance would have sat at a different energy, and the samarium would have burned at a different rate.
Thibault Damour and Freeman Dyson reanalyzed the Oklo data in 1996, taking particular care over the reactor's uncertain temperature. They found that the fine-structure constant two billion years ago matched today's value to within about one part in ten million. Astronomers can test the same constant much further back in time, using absorption lines in the light of distant quasars. At the time, the best limits from those spectra were several hundred times looser than Oklo's.
Two billion years ago, there were no eyes on Earth to see any of this. Groundwater trickled into a seam of uranium, the rock warmed, the water boiled away, and the cycle started again, over and over. And inside every samarium nucleus there, the number that sets the strength of electric forces today was quietly doing the same job.
🌙 One last thought before you sleep
The J1407 search leaned on 387 glass plates exposed between 1890 and 1990 by people who couldn't have known that a young star in Centaurus would matter one day. Someone looked after them for a century, and someone else measured the star on every one. I find that more moving than any new telescope: a sky photographed for no particular reason, waiting for the right question to come along.
The next eclipse of J1407 might come in my lifetime, or never. Either way, I hope somebody is taking pictures.
That's all for tonight. Thanks for wondering along with me. Good night, and clear skies. 🌌
— Kasi