The rise and fall of Planet X: Neptune and Pluto were supposed to "fix" the weird orbit of Uranus. Now, it seems, the orbit wasn't "broke."

Natural History, June, 2003 by Neil deGrasse Tyson

Heard about Planet X lately? Probably not. It's dead--no matter what anybody has told you. Astrophysicists no longer need to postulate the existence of an "undiscovered" planet to explain the motions of the other planets in our solar system.

The rise of Planet X begins with the German-born English astronomer Sir William Herschel, who more or less accidentally discovered the planet Uranus on March 13, 1781. That episode was an exciting moment in eighteenth-century astronomy. Nobody in recorded history had ever discovered a planet. Mercury, Venus, Mars, Jupiter, and Saturn can each be seen relatively easily with the naked eye, and all were known to the ancients. So strong was the bias against finding additional planets that Herschel, even in the face of contrary evidence, assumed he had discovered a comet. Other eighteenth-century star watchers were in denial as well. Charles Messier, the French astronomer and consummate comet hunter, noted, "I am constantly astonished at this comet, which has none of the distinctive characters of comets."

Archival records of star positions show that several observers had seen Uranus before Herschel did, but each one had mistakenly classified the planet as a star. In an embarrassing example from January of 1769, the French astronomer Pierre Charles Lemonnier did not discover Uranus six times! When Herschel finally noted that the mysterious object moved, astronomers were able to calculate an orbit with good precision because of the availability of nearly a century's worth of "prediscovery" data on its position in the sky. Their calculations showed that the object's orderly, near-circular path, far from the Sun, had nothing in common with the eccentric trajectories of all known comets. At that point, you would have had to be both blind and boneheaded to resist calling the new object a planet.

But all was not orderly in the solar system. Uranus was behaving badly. The new planet was not moving through space the way astronomers expected it to. Its trajectory around the Sun was not following the path Newton's law of gravity would have it take. The historical observations fitted one orbit; the post-1780s telescopic observations fitted another.

Some astronomers suggested that Newton's laws might be invalid at such large distances from the Sun. That wasn't as crazy as it sounds--under new or extreme conditions the behavior of matter can and does deviate from the predictions of the known laws of physics. But only if Newton's theory of gravity had been nascent and untested would there have been good reason to doubt it. By the time Herschel discovered Uranus, however, Newton's laws had had a hundred-year run of successful predictions. The most famous of them was Edmond Halley's prediction of the 1758 return of the comet that would be named in his honor.

The simplest conclusion? Something else had to be out there, something yet undiscovered, whose gravity had not been accounted for in the predicted orbital path of Uranus.

In the life cycle of a physical theory, a scientist first makes a testable prediction about the world. Then a skeptical colleague runs a few actual experiments to see how well the prediction stands up to reality. The arithmetic differences between the theory's predictions and the experimenter's data are sensibly called "residual errors"--"residuals" for short--and they're the measure of a theory's success. Small residuals are good; big residuals are bad. If the theory describes nature accurately, and the experiment is well designed, the residuals are not only small, but they fluctuate between positive and negative values from one measurement to the next, yielding an average close to zero. If the average is anything other than zero, one can rightly say that crucial differences exist between the predictions and the measurements.

When that happens, it's not easy to assign blame. Maybe the theory needs to be modified, or maybe somebody blundered when the measurements were taken, or both. If your theory of gravity predicted that an object should fall upward when released, the theory would require significant modification, because the residuals between the predicted positions and the actual positions along the object's trajectory would be gigantic, and would not average to zero.

In the late eighteenth century the French mathematician Pierre-Simon de Laplace invented perturbation theory [see "Going Ballistic," by Neil deGrasse Tyson, November 2002], giving astronomers an indispensable tool for analyzing the small gravitational effects of an otherwise undetected celestial object. Encouraged by the expansion of their arsenal, mathematicians and astronomers across Europe continued to investigate what might be perturbing Uranus. In 1845 a young, unknown English mathematician, John Couch Adams, approached Sir George Airy, Britain's astronomer royal, with a request that he search a specific patch of sky for an eighth planet. But neither looking for planets nor following the leads of spunky young mathematicians was part of the astronomer royal's job description, so Adams's request was dismissed. The next year, the French astronomer Urbain-Jean-Joseph Le Verrier independently derived a similar prediction. On September 23, 1846, he communicated his prediction to Johann Gottfried Galle, who was then assistant director of the Berlin Observatory. Searching the sky that very night, Galle found the new planet, soon to be named Neptune, within a single degree of the spot Le Verrier had predicted. It took him only an hour to locate it.