Posted by Mars keeps two rocks in orbit, and neither of them looks like it belongs there. The Martian moons, Phobos and Deimos, are dark, lumpy and absurdly small: Phobos is about 22 kilometers along its longest axis, Deimos barely 12. Photographed up close they look like asteroids that wandered in from the belt and got stuck, which is roughly what most astronomers assumed for the better part of the twentieth century.
The trouble is that they do not move like strays. Both circle almost exactly in the plane of the Martian equator, on paths that are very nearly circular. Objects captured by a planet do not usually arrive that tidy.
The argument has been running for more than a hundred years. A spacecraft is being built to end it.
Two Nights in August 1877
Asaph Hall nearly quit before he found them. He was working the 26-inch refractor at the United States Naval Observatory in Washington during the opposition of 1877, when Mars swung unusually close, the same season that had Giovanni Schiaparelli in Milan mapping the surface markings he called canali. He was hunting for satellites most of his colleagues doubted existed, and the glare of the planet kept swamping the field. After a run of bad nights he was ready to give up. His wife, Angeline Stickney, a mathematician who had once been his teacher, told him to go back to the dome. On August 12 he caught a faint point of light trailing the planet. Fog closed in for several nights. When it cleared he had both.
The names came from Henry Madan, a science master at Eton, who reached for Book XV of the Iliad, where Phobos and Deimos, panic and dread, harness the horses of Ares. The largest crater on Phobos is named Stickney.
Asteroids That Refuse to Act Like Asteroids
Start with what the light says. Both moons are dark, throwing back only six or seven percent of the sunlight that hits them, and their spectra are a decent match for the reddish, featureless D-type asteroids of the outer belt and the Jupiter Trojans. Phobos also has a density near 1.9 grams per cubic centimeter, well below solid rock, which means a quarter to a third of its volume is empty space. It is a rubble pile, in the same family as Ryugu and Bennu. All of that points one direction: Mars grabbed a couple of passing asteroids.
Then the dynamics ruin it. Capture means bleeding off energy, and the mechanism usually proposed for Mars is drag through a puffed-up early atmosphere. Run the numbers and you can indeed snag an asteroid, but you end up with a stretched, tilted orbit that takes an awkwardly long time to round off, and you need the atmosphere thick at the right moment and thin shortly after. Two captures, both equatorial, both nearly circular, is asking a lot of luck.
The competing story is violence. Mars carries an enormous scar across its northern hemisphere, the Borealis basin, which many researchers read as the mark of a giant impact early in the planet’s history. A collision on that scale would have thrown a disk of debris into orbit. Work published by Pascal Rosenblatt and colleagues in 2016, and modeling by Robin Canup and Julien Salmon in 2018, suggests such a disk could build one large inner moon that later spirals in and dies, leaving Phobos and Deimos as small survivors out at the edge. Equatorial, circular orbits come free with this picture, because that is what debris disks produce.
The catch: material blasted off Mars ought to look like Mars, and the surfaces of the Martian moons stubbornly do not. Defenders of the impact model answer that the stuff has been cooked by radiation and micrometeorites for billions of years, or that the impactor’s own remains dominate the mix. Both are plausible, and both are very hard to disprove from a distance.
A third camp splits the difference. Amirhossein Bagheri and co-authors proposed in 2021 that both moons are fragments of a single earlier satellite that came apart one to three billion years ago, which would explain why their orbits are so nearly, but not exactly, aligned.
Nobody knows. That is not modesty, it is the state of the field.
Stickney, and the Scratches
Stickney is about nine kilometers across on a moon of twenty-two. It is as if something punched a hole a third of the way through Phobos and the moon simply absorbed it. Simulations of that impact tend to blow the target apart. That Phobos survived is itself an argument for the rubble-pile picture, since loose material soaks up a shock that would shatter a solid body.
Then there are the grooves. Mariner 9 saw them in 1971, and Mars Express and the HiRISE camera on Mars Reconnaissance Orbiter have since mapped them in detail: long parallel scratches, some running for tens of kilometers, cutting across large parts of the surface. Explanations are not in short supply. Chains of secondary craters from Stickney ejecta. Tidal stress fractures opening as Mars pulls at the moon. Boulders bouncing away from the impact site, plowing furrows through the dust. Debris kicked off Mars by other impacts, sandblasting Phobos in sheets.
Each idea handles some groove families and fails on others. They do not all radiate from Stickney the way ejecta should, and some cut across the terrain in a pattern that looks less like the moon was hit and more like it is being stretched.
Phobos Is Falling
Phobos circles Mars in 7 hours and 39 minutes, faster than Mars turns on its axis. That one fact decides its future. Our own Moon orbits more slowly than Earth rotates, so tidal friction feeds it energy and pushes it outward. Phobos runs ahead of its planet’s rotation instead, so the same physics works in reverse and drags it down, by roughly two centimeters a year. Deimos, further out and taking about 30 hours to go around, sits on the safe side of that line and creeps slowly away.
Two centimeters is nothing on a human clock. On a planetary clock it is a countdown. Within a few tens of millions of years Phobos will reach the distance where Martian tides overwhelm its own feeble self-gravity, and a rubble pile has very little to hold on with. Benjamin Black and Tushar Mittal argued in 2015 that it would not simply crater the planet but be torn into a ring of debris that then rains down over millions of years. Andrew Hesselbrock and David Minton went further in 2017 and proposed a cycle: ring forms, ring spreads, material re-accretes into a new moon, that moon spirals in, repeat. On that reading Phobos is not the original satellite at all. It is a late draft.
Mars may be between rings.
Ten Grams Would Settle It
Phobos has an unhappy record with visitors. The Soviet Phobos 1 was lost in 1988 to a faulty command sent from the ground. Phobos 2 reached Mars orbit and returned images before contact broke off in March 1989, days before it was to release two landers. Russia’s Phobos-Grunt, launched in November 2011 to bring a sample home, never escaped Earth orbit and came down in the Pacific two months later.
The current attempt is Japanese. JAXA’s Martian Moons eXploration mission, MMX, is built to survey both bodies, touch down on Phobos, and return at least ten grams of surface material to Earth in the early 2030s. It carries a small rover built by the French and German space agencies, plus a NASA gamma-ray and neutron spectrometer called MEGANE. The agency has done this before: Hayabusa2 delivered grains of the asteroid Ryugu in December 2020, and NASA’s OSIRIS-REx dropped its Bennu capsule into the Utah desert in September 2023.
Ten grams sounds trivial. It is not. Oxygen isotope ratios are a fingerprint of where in the solar system a rock formed, and they are measurable in milligrams. If the sample matches the Martian meteorites, the giant-impact model wins outright. If it matches primitive outer-belt material, the capture camp is vindicated. A genuinely mixed result would be the most interesting outcome and the most irritating.
There is a second prize in the same scoop. Every large impact on Mars flings rock into space, which is how ALH 84001 and the other Martian meteorites reached us. Some fraction of that ejecta lands on Phobos and stays, stirred into the regolith. That makes a Phobos scoop a lottery ticket for Martian crust as well, including material from epochs when the planet was wetter. Nobody expects biology up there. Billions of years of unshielded radiation are hard on organic molecules. But chemistry is tougher than people assume, and a sample in a curation facility can be re-examined for decades as instruments improve.
A Doomed Moon Is a Useful Moon
What makes this bigger than Mars is that satellites are not permanent. Planets acquire them, break them, swallow them, occasionally rebuild them, and we mostly study survivors. Phobos and Deimos let us watch the process mid-sentence: one moon on its way down, one drifting off, both holding onto whatever happened around Mars four billion years ago.
The answer is roughly a decade away, riding in a capsule that has not launched yet. If you want to follow the Martian moons from launch through landing to the laboratory, along with the rest of what we are learning about Mars, its neighbors and the search for life beyond Earth, SETIworld is a good place to read along and to argue about the results as they land.
