On the night of September 23, 1846, Johann Gottfried Galle pointed the telescope at the Royal Observatory in Berlin in the general direction of Capricorn, stopping at star after star. His companion, Heinrich Ludwig d'Arret, pored over a sky map, ticking off each object as Galle called out its brightness and position. Sometime between midnight and 1 a.m., Galle registered one more mote of light invisible to the naked eye: right ascension 21 h, 53 min, 25. 84 seconds.  

d'Arret checked the chart.  “That star is not on the map!"

True stars remain mere points even in powerful telescopes. Galle’s mystery dot did not, showing instead an unmistakable disk, a full 3.2 arc seconds across.  That tiny disk settled the matter: Galle had just become the first human being to see a previously unknown planet whose existence was predicted in advance. Indeed, it had appeared right where the great French mathematical astronomer Urbain Jean Joseph Le Verrier had told him it would be. Within weeks, the newest member of the solar system would have its name: not Le Verrier, as he had hoped, but Neptune.

Last week, two Caltech astronomers, theoretician Konstantin Batygin and observer Michael Brown published a predictionnot yet a discoveryof another new planet lurking at the far periphery of the solar system. This (potential) new member of the solar system, dubbed Planet Nine has (possibly) materialized in the same way that Neptune did. The astronomers’ task was, Batygin says, “qualitatively the same” as the one Le Verrier solved, “an attempt to reproduce the orbit of an unseen planet deduced solely by its gravitational effects on other objects.”  

In both cases, there was an anomaly, an oddity in the motion of known objectsthe planet Uranus for Neptune; six distant bodies in the Kuiper Belt for Planet Nine. The Kuiper Belt is a swarm of dwarf planets and even smaller objects beyond Neptune of which Pluto is the most famous (and contentious) member.  Le Verrier had shown that Neptune could settle Uranus’ books, while Batyagin and Brown account for the peculiar behavior of the relevant Kuiper Belt objects (KBOs) by invoking a planet roughly ten times the mass of Earth.

In 1846, the discovery of Neptune turned Le Verrier into a celebrity; for a time, he was the most famous man of science in the world. He went on an international tour and seized the moment to rise to the top of power in the highly contentious and hierarchical world of French astronomy. Batygin and Brown are taking a much more measured tack with Planet Nineand for good reason.  “We felt quite cautious about making the statement we made,” Batygin says.  Why such concern? Because, he says, “immediately after the detection of Neptune spurious claims of planets in outer solar system began to surface. We didn’t want to be another red herring.”

There aren’t any obvious errors in Batygin and Brown’s gravitational argument, but nature has plenty of ways to fool astronomers into seeing planets where there are none. Any mass exerts (as Newton saw it) a pull on everything else, and Newton’s universal law of gravitation describes how strong that tug will be, and what motion would result. In the case of Neptune and, presumptively, Planet Nine, undiscovered objects reveal themselves in the unexplained residues of motion of what’s already been observed, once all the known gravitational influences have been tallied up.

That may sound simple, but gravitational glitches can be deceiving. Le Verrier himself was one of the first to be fooled by the impeccable logic of Newton’s theory.  After Neptune, he turned his attention to the inner solar system, and he showed (correctly) that Mercury’s orbit wobbles in a way that a Newtonian accounting couldn’t explain. With his own recent triumph so fresh in hisand everyone else’smind, the explanation was obvious: There must be a planet hidden in the ferocious glare of the sun, one soon dubbed Vulcan. Within months of making that prediction, Le Verrier trumpeted as a settled discovery an amateur astronomer’s claim to have seen Vulcan crossing the face of the sun. Over the next two decades, at least a dozen other reputable observers reported similar sightings.  

And yet, Vulcan doesn’t exist. Le Verrier’s argument for why it should have been real is perfectly consistent, and it matched the state of knowledge at the timebut it turns out that Mercury travels close enough to the great mass of the sun to be subject to the effects Albert Einstein would describe in 1915, with his General Theory of Relativity. The moral of this story isn’t that scientists make mistakes that get resolved through further work. Rather, it lies with the fact that so many observers persuaded themselvesand Le Verrierthat Vulcan was so clearly necessary that it had to be real. That same treacherous combination of hope and expectation has bedeviled planet hunters searching the space beyond Neptune.

Shortly after Galle and d’Arret first glimpsed Neptune, major planets began appearing, mirage-like, in the distant solar system. An early estimate of Neptune’s mass seemed to suggest that there was still elements of Uranus’ trajectory that demanded some other source of gravityyet another hidden planetto explain. Close analysis of several comet tracks led to the suggestion that not one but two such planets might exist. Various astronomers published predictions for a first planet at distances that clustered around fifty times the Earth-sun distance (a measure known as the Astronomical Unit, or AU), and another from sixty to seventy AUs. When more theorists took up the challenge, still more orbits were proposed. One astronomer, Harvard University’s William Henry Pickering, would offer up seven possible trans-Neptunian planets.

Finally, early in the twentieth century, the American Percival Lowell, best known for suggesting that canals and hence civilization might exist on Mars, began a new search for what he dubbed Planet X. In 1915, he published a prediction of a seven-Earth-mass planet orbiting at about 43 AU.  

Lowell died in 1916, but the observatory he had founded in Flagstaff, Arizona resumed the search for his trans-Neptunian planet in 1927. No results appeared for the next two years, and in 1929, the task was handed off to Clyde Tombaugh, a 22-year-old amateur astronomer who had just joined the Lowell Observatory staff. He made pairs of astrophotographs, images of the same patch of sky taken two weeks apart. He then compared those images to see if any object had moved over the intervening time. After a year of intense, monotonous, meticulous labor, finally on, Feburary 18, 1930, Tombaugh found a mote of light that shiftedjust enoughon his pictures.

Lowell Observatory

Tombaugh’s discovery was real. Named Pluto,  it would take its place, for a time, as the ninth planet. But from the start, it was obvious that something wasn’t quite right. Pluto orbits at a sharper angle to the eclipticthe plane described by the Earth’s track around the sunthan the first eight planets do. And crucially, while estimates of its mass would be revised down for decades, it was clear even in the 1930s that the new planet was too small to produce enough gravitational pull to deal with the apparent residue of the motion of Uranus.

And so the hunt for Planet X continued. As with Vulcan, the logic seemed inescapable. The explanation that settled Mercury did not apply to Planet X: As described by General Relativity, the sun’s great mass warps nearby space and time  enough to affect Mercury’s track, but in the outer reaches of the Solar System, relativistic effects are tiny enough so that Newtonian gravity dominates.  If Uranus wobbled, then something must be pushing it.

That was still a mistakebut of a different kind than that which snared the Vulcan-hunters. Those pursuing Planet X were overly confident about the precision of the measurements on which they relied. Neptune’s mass was first measured by analyzing the interaction between the planet and its large moon, Triton. But in 1989, the Voyager 2 spacecraft flew by Neptune, collecting data that suggested the planet was slightly lighter than previously believed. Recalculating with the new value gave a perfectly consistent orbit for Uranusand with that, the necessity of Planet X disappeared.

This long history of missed predictions and overconfidence haunts the astronomers who have proposed Planet Nine’s existence. “We spent a year making mathematical models, squeezing every possibility out of the solar system,” Batygin says. “In the end it was an act of desperation to introduce Planet Nine.” But then, “when we did, all of the features we saw in the Kuiper Belt became reproducible in our artificial solar system.” That, says Batygin, evoked, “Simultaeneously a feeling of relief and concern.”

The most useful result, from Batygin’s perspective, was one that went beyond the question he and Brown had set out to answer. Their initial problem had been to explain what could drive the paths of six very distant, recently discovered KBOs.  But as they developed their model, it consistently produced a few objects that moved on a completely unexpected path, perpendicular to the plane of the solar system. As he worked on the calculation, Batyagin hadn’t known of any real objects that behaved that way, but the model kept demanding them. This, he thought, “must be strong counter evidence for Planet Ninebecause we would have seen those orbits.”  That’s when he asked Brownthe observer of the pairwhat he thought. Brown produced the data for a KBO with an odd track. “We plotted the real data on top of the model” Batyagin recalls, and they fell “exactly where they were supposed to be.” That was, he says, the epiphany. “It was a dramatic moment. This thing I thought could disprove it turned out to be the strongest evidence for Planet Nine.”

Such storiesthe sudden, unanticipated match between a new, untested idea and the real worldare hallmarks of successful science. When Isaac Newton mapped the observed path of the comet of 1680 onto the track he calculated, using his then brand-new theory of universal gravitation, he declared that a theory that so perfectly corresponded with reality “cannot fail to be true.” Albert Einstein told friends that as the orbit of Mercury simply appeared as he calculated with his still-not-quite complete general theory, he felt heart palpitations, that he was so excited he couldn’t work, that he was “beside himself with joy.” That was the point he became utterly convinced that his new theory of gravity was true, the first such advance since Newton himself. The stakes with Planet Nine are orders of magnitude smaller than those predecessors, of course, but that moment when abstract mathematical thought syncs up with the real world is always a thrill.

Crucially, though, such joys are not quite proof, not the final confirmation required to turn a beautiful, plausible, perhaps even an obvious proposal into the stuff of nature. For General Relativity, that final step came when astronomers measured the bending of starlight around the sun during an eclipse that came four years after Mercury had appeared in Einstein’s equations. For now, Planet Nine seems the only satisfactory explanation for everything now known about the outer suburbs of the solar system. “If Newton is right, then I think we’re in pretty good shape,” says Batyagin. “We’re after a real physical effect that needs explanation. The dynamics of our model are persuasive.” And yet, he adds, that’s not enough. “Until Planet Nine is caught on camera it does not count as being real. All we have now is an echo.”