One stormy night near the turn of the millennium, a friend and I were sitting beneath the massive Hale Telescope at Palomar Observatory near San Diego. The telescope was standing idle inside its closed dome because of bad weather, and so we passed the time discussing recent discoveries about the edge of the solar system. I stated my conviction that beyond Pluto—the most distant of the known planets since its discovery in 1930—more planets were waiting to be found. In fact, I was so certain that I was willing to bet on it. We set a deadline of Dec. 31, 2004, for the discovery of a 10th planet. Then we did something that might seem surprising. We chose our own definition for the word planet. In 1999—and even as of mid-2006—astronomers had not yet agreed on an official definition for this familiar object. We did not think we could calculate immediately how big any newly discovered object was. As a result, we decided our criterion should be the amount of sunlight the object reflected, a rough estimate of size.

The idea that there might be planets beyond Pluto had been percolating in astronomers' minds since 1992, when David Jewitt of the University of Hawaii at Honolulu and Jane Luu, then at the University of California at Berkeley, spotted a small, icy object circling the sun well beyond Pluto's orbit. The object they discovered was the first identified member of a vast population of such objects orbiting the sun in the outer reaches of the solar system. This area is called the Kuiper (pronounced KY pur) belt, for Dutch-born American astronomer Gerard P. Kuiper, who in 1951 theorized that the belt may be the source of short-period comets (comets that take fewer than 200 years to orbit the sun). Astronomers believe that objects in the Kuiper belt are “building blocks” left over from the formation of the planets 4.6 billion years ago. By the end of 2000, astronomers looking at small areas of the sky had discovered nearly 500 Kuiper belt objects (KBO's), with the largest being about one-third the size of Pluto. Pluto's diameter is about 2,300 kilometers (1,430 miles). Because we were looking at much larger segments, of the sky, it seemed just a matter of time—and hard work—until we would find more objects at least that size.

My confidence about the discovery of new planets was bolstered by the project that I was working on at the time. Chad Trujillo at the Gemini Observatory on Mauna Kea, Hawaii; David Rabinowitz at Yale University in New Haven, Connecticut; and I had just begun the broadest search ever made of the Kuiper belt to find new large bodies in the solar system. Our work involved taking pictures of the skies using the 1.2-meter (48-inch) Samuel Oschin Telescope at Palomar Observatory. Night after night, we searched a different section of the sky that—from Earth—covered an area about the size of a hand extended at arm's length. We were looking for faint objects moving at a certain speed across the background of stars so distant that they appear to be fixed (unmoving).

Every object orbiting the sun moves across the sky. Objects close to the sun—such as the terrestrial planets Mercury, Venus, and Mars and the small, rocky asteroids orbiting between Mars and Jupiter—move across the sky quickly because of their relatively short orbits. Between the orbits of Jupiter and Neptune, in the thinly populated realm of the solar system's giant planets, objects move at more moderate speeds. The objects we were seeking were the most distant—and, therefore, the slowest moving—of all. We took three pictures of each area three hours apart and then compared them (using sophisticated computer programs) with images of millions of fixed stars and galaxies. Occasionally, we found what we were looking for—a faint point of light that looked like any other star in the first picture but had changed position slightly in the second picture and moved again in the third. About once a week, our photographs captured a dim, slow-moving object that we could add to astronomers' collection of known KBO's.

Most of the KBO's we found were small—perhaps one-fourth the size of Pluto. However, on a few occasions, we found an object brighter than anything else in the Kuiper belt. Each time we found such an object, we thought that we had finally discovered something larger than Pluto. But after more detailed examinations using other telescopes, we realized that none of these objects was larger than two-thirds the size of Pluto. In late December 2004, we discovered our largest object yet, but even that body was only about three-fourths of Pluto's size. On Jan. 1, 2005, I sent a note to my friend conceding the bet. My prediction that new planets would be found within five years had not come true.

Our search did not end, however. Four days after losing the bet, I was examining an unremarkable part of the sky in the constellation Cetus, the sea monster. Suddenly, the computer picked out a bright object that looked like a star in the first picture but had shifted its position slightly in the second picture. In the third picture, it had shifted again. This object was one of the brightest that we had ever seen in motion. Moreover, it was moving more slowly than any other KBO we had ever seen, meaning that it was also the most distant. The fact that it was so bright despite being so far away meant that the object was big—bigger than any of the KBO's we had seen to date. A quick calculation on my computer verified what I had initially guessed. This object was not only bigger than Pluto but also the largest object discovered in the solar system since 1846. I had lost the bet four days earlier, but the discovery confirmed my belief that our solar system had more than nine planets. I temporarily nicknamed the object Xena, after the Greek warrior princess in the 1990's television series. But had we really found a new planet? Some astronomers thought not.