Something is out there.
It throws no light you could ever see. It carries no name that anyone spoke before 1984. And if the rocks are telling the truth, it has been coming home to us on a schedule since long before the first animal hauled itself out of the sea.
Twenty-six million years. That is the interval. Not roughly, not loosely. Twenty-six million years, again and again, stamped into the fossil record of this planet like the tick of something enormous and patient.
Every time it comes, the world empties out.
Nobody went looking for it. Two men found it by counting.
The ledger
Picture a man alone with the fossil literature of the world.
J. John Sepkoski Jr. spent years on it. By hand, family by family, he went through the published record of ancient marine life and wrote down two dates for every one of them. When it first appeared. When it last did. First breath, last breath, thousands of times over, until he had a register of the dead going back a quarter of a billion years.
He was asking a dull question about diversity. He got something else.
In February 1984, Sepkoski and David Raup published what the ledger had given them in the Proceedings of the National Academy of Sciences. The title was flat: "Periodicity of extinctions in the geologic past." The finding was not.
The dying was not random. The peaks came on a schedule. Roughly every twenty-six million years, across 250 million years of stone, the oceans of this planet went quiet at the same interval.
Something was keeping time.
Only one thing keeps time like that
Think about what a rhythm that long actually demands.
Volcanoes do not run on a schedule. Neither do ice ages, nor sea levels, nor the chemistry of the oceans. They surge and they fade, and they do it when conditions push them, not when a clock says so. Nothing on Earth is known to hold a beat for a quarter of a billion years.
But one thing in nature keeps perfect time, and it has done since the solar system condensed out of dust.
An orbit.
That is the thought that turned a fossil study into a manhunt. If the dying is periodic, something in the sky is periodic. And whatever is up there keeping the count is an object. Objects can be found.
The dark companion
Two months later, the answer arrived twice.
The April 1984 issue of Nature carried two separate papers, written by people who were not working together, reaching the same conclusion. One came from Daniel Whitmire and Albert Jackson. The other from Marc Davis, Piet Hut and Richard Muller.
The Sun is not alone.
Picture what they described. A cold ember of a star, a red dwarf or a brown dwarf, so dim it could sit in our own sky and never be remarked on. It rides an orbit so stretched that it swings a light year out into the dark and then comes back. And on its way in, every circuit, it passes through the Oort cloud, the immense shell of frozen bodies wrapped around everything we know.
It never has to touch us. It only has to lean.
One pass is enough to shove a swarm of comets off their paths, and some of them fall inward, toward the light, toward the planets, toward here. What arrives at Earth is not a single strike. It is a rain of them, spread across a million years, on the companion's schedule.
Muller named it Nemesis, after the Greek goddess who comes to settle accounts. The press had a better ear. They called it the Death Star.
The name held. Muller published Nemesis: The Death Star in 1988. Raup had already written his own account, The Nemesis Affair, in 1986.
Then the idea grew a second head. Whitmire, working with John Matese, suggested the thing doing the stirring might not be a star at all, but a planet, far beyond Neptune, working the comet reservoir on a similar cycle. The Planet X literature and the Nemesis literature have been tangled together ever since.
Why the world listened
This was not a fringe idea whispered at the edges. It went off like a flare, and the reason is timing.
Four years earlier, in 1980, Luis and Walter Alvarez and their colleagues had published the iridium findings that pinned the death of the dinosaurs on something falling out of the sky. Before that paper, blaming space for a mass extinction was a career risk. After it, it was a live hypothesis with a chemical signature behind it.
Raup and Sepkoski's beat landed in a field that had just been taught to look up.
And look at the speed. Paleontology in February. Astronomy in April. Same year, biggest journal there is, written by working physicists, and inside a few years there were telescopes hunting for it.
The case against
Nobody has found it. That is the blunt version, and it has to be said plainly.
The searching has been real. Infrared surveys have swept the sky again and again since the 1980s, each one cutting deeper into the dark than the last. In 2014 the astronomer Kevin Luhman worked through data from the WISE spacecraft and reported nothing Saturn-sized out to 10,000 astronomical units, and nothing Jupiter-sized out to 26,000. That is a lot of empty volume, though it is not the whole of the orbit Nemesis was given.
Then there is the physics. Conventional celestial mechanics holds that anything orbiting a light year out is barely held at all. Passing stars pull at it. The tide of the galaxy pulls at it. Across 250 million years its period would wander, and a wandering period cannot carve a clean beat into rock.
And there are people who doubt the beat itself. Coryn Bailer-Jones went back over the impact crater record in 2011 and argued the rhythm there was better explained by which craters survive and get noticed. Others point out that the geological timescale is a human invention, revised over and over, and that errors in dating could lay a false rhythm across data that never had one.
The part that will not go away
Here is where it gets strange.
The star never turned up. The signal never left.
In 2010 Adrian Melott and Richard Bambach went back to the whole question with fossil databases that did not exist in 1984, and with a much longer stretch of time to work across. The beat was still there. Roughly 27 million years, reaching far deeper into the past than Raup and Sepkoski could ever have seen. Two independent datasets carried it.
And then they said the thing that keeps this alive.
The rhythm is too regular for Nemesis.
A loosely bound companion drifts. This signal does not drift. It keeps better time than any comet-shower model can produce.
Sit with that, because it cuts both ways at once. The physics argument that kills Nemesis kills every other wandering object in the outer dark along with it, because they would all drift. And the statistical argument now has to explain a beat that shows up in separate compilations, assembled by different people, from different fossils, decades apart.
Neither answer covers the whole of it.
What remains is a number
Twenty-six million years. Maybe twenty-seven.
It is written into the extinction record of this planet. It holds its period across hundreds of millions of years. And nothing anyone has found could be keeping the count.
Either it is a ghost in the data, an artifact of how human beings built a timescale, and it will dissolve as the dating sharpens.
Or something out there has been keeping time since long before there were eyes here to notice.
The rock has not said which.