How I Killed Pluto and Why It Had It Coming - Mike Brown (2010)

Chapter 5. AN ICY NAIL

There is a critical tension in science between the very human desire to announce discoveries immediately (both because you are excited about them and because you don’t want to be scooped by someone else) and the very important need to carefully and systematically check and document your results. In some cases this documentation can take additional years of work. In the case of the new discovery that Chad had told me about, we were quite worried that someone else might stumble upon it in months or even weeks, so we put together a plan to try to learn everything we could about it in as short a time as possible, and we set ourselves a deadline of only four months to make an announcement, complete with a full scientific account of the discovery and anything else we could learn. For me, those months of trying not to tell anyone about our discovery was harder, even, than not telling Diane that I had had a ring in my pocket all along.

During this time, we decided that rather than repeating “the object that we just discovered,” we should give it a temporary name. We settled on Object X. “X” was for Planet X, for unknown, and, perhaps, for tenth planet.

As scientists, we were eager to know everything about Object X, but the first question on our minds, the one that would put everything else into context, was: What sort of orbit did Object X have? Did it go in a circle around the sun like the planets, or did it have an elongated orbit like Pluto and the other objects in the Kuiper belt? To answer this, we would have to track the object through its orbit and learn where it went. This would take time and patience. Pluto, after all, takes 255 years to go around the sun. Object X, farther away, would take even longer. But time and patience were two things we could not afford. Fortunately, though, we didn’t need to wait hundreds of years. We don’t actually have to follow an object all the way around its orbit to know where it is going to go (a good thing, since we’ve still watched Pluto for only a little more than a quarter of its orbit). If something is moving under the influence of gravity alone, we need only to know precisely where the object is, precisely how fast it is going, and precisely what direction it is moving in to know where it was at all times in the past and where it will be in the future.

Even if you don’t know how to work out the math yourself, your brain certainly does. Try this experiment. Stand in a field and have someone thirty feet away throw a ball in your direction (using a foam ball would be a good idea, as will become obvious). The second you see the throw, close your eyes and see if you can figure out where and when the ball is going to hit the ground. Chances are you’ll do pretty well. Your brain is instinctively trained to quickly estimate the three key things—where, how fast, which direction—and predict where a projectile is going to go. But chances are you will not be precisely correct. The ball will probably land a little to the side, or a little later than you predict. That will be because you looked at the ball for only an instant, and your brain could not discern the speed or direction or location as accurately as you needed. Watching the ball a little longer before you close your eyes would improve your predictions. In the end, closing your eyes is never a good way to actually catch the ball, because at that point you want your estimate of where the ball is going to land to be accurate to a few inches, but if you just want a good indication of the ball’s general movement, those first few moments of observing will suffice.

Object X is just like that ball being thrown. It is affected only by the force of gravity (the earth’s gravity for the ball, the sun’s gravity for Object X), so once we know where it is, how fast it is going, and the direction of motion, we know everything we need to know to be able to follow its orbit forever. Those first three hours that we had already seen, however, were like the very instant that someone threw a ball. If that’s all you get to see, your estimate of where the ball is going will not be very accurate. We needed to keep our eye on the ball for a little more time before we knew the actual orbit of Object X.

In general, to understand the orbit of something so far away takes about a year’s worth of precise observations. We couldn’t wait a year. While we tried hard not to lose sleep at night thinking about someone else discovering Object X while we were still studying it, I would pick up the newspaper almost every morning with dread in my stomach. We were determined to wait long enough to write an accurate and thorough scientific paper on Object X, but we wanted to wait not a minute longer, for fear of being scooped. Wait until next year? No way.

Luckily, we didn’t actually have to wait a year in the future. We could, instead, go back a year in the past. Many astronomers have taken many pictures of the sky over time, and perhaps we could find Object X there; by now, there were even online repositories of many of these images. Chad and I set to work in our separate offices across the hall from each other, probably looking at the exact same online pictures. I’ve heard stories of different parts of the same scientific team working in parallel on the same problem as a way of double-checking an important result, but I must admit, the fact that Chad and I were doing the same thing at the same time had nothing to do with double-checking. Looking back through the archive photos was simply so much fun that we both wanted to do it.

Here’s how it worked, at least on my side of the hallway. First, I did the best calculation of where Object X was going and predicted where it should have been on a particular date a few months earlier. I then searched the archive for images at that position. Not surprisingly, there were none taken on the particular date I was looking for, but there were some taken a few weeks earlier. I went back and calculated the position of Object X for a few weeks earlier, and luckily, the position was right on that image. I downloaded the image from the archive and displayed it on my computer. The picture was full of indistinguishable stars. How could I know which was Object X? The only way to distinguish our discovery from the many, many stars in the sky was to see it move. But there was only one picture from that night, so there was no way to see it move. I could, however, go back to the archive and find a picture of that part of the sky taken a year earlier. Object X was moving, so a year earlier it would have been somewhere else entirely. I compared the pictures from the night when Object X was supposed to be there to the earlier pictures. It’s easy on the computer; you just line the pictures up, press a few buttons, and the two pictures blink back and forth like a very short and repetitive movie. The two pictures were nearly identical. The stars and the galaxies had not changed at all over the year. But there, in the middle of the more recent picture, was a new starlike object that hadn’t been there the year before.

That was what I was looking for; I still couldn’t tell for sure that it was an object that was moving, but it certainly was one that hadn’t been visible a year earlier. There are many things in the sky that can appear where they weren’t seen before—stars that get brighter, stars that explode—so I didn’t know for sure if this was our Object X or not. But if I assumed it was, I could calculate a little better where the object was going. With this more refined calculation, I could figure out where Object X should have been yet another full year earlier. I then restarted the whole process. Look for a picture in the right place; realize it is not quite the right time; revise the time; find the place; find an earlier comparison; look for something new. There it was! Right where I had predicted! I ran across the hall to tell Chad I had found Object X from a year earlier. He had found it a few minutes before me and was already looking for pictures from two years earlier. We were racing down the right trail.

We quickly followed Object X back for about three years, which was the limit of the data we could find online in the archive. While we were sitting in my office pondering what we might do next, Chad wondered aloud if perhaps Object X might be found on Charlie Kowal’s plates. Ah yes. Charlie Kowal’s plates.

Most of us have a blind spot, something we can’t see even though it is right in front of us. Charlie Kowal’s plates were directly in my blind spot. I knew about them but preferred not to think about them. Why? Kowal had, years earlier, proved that there were no planets out past Pluto. Since this information did not fit well into my view of the solar system, I chose not to think about it.

Charlie Kowal was an astronomer who had worked at Palomar Observatory in the 1970s and 1980s. He had decided to do something that no one had ever tried before: use the Palomar 48-inch Schmidt Telescope to find a planet beyond Pluto. At the time, Planet X was generally expected to exist (this was the 1970s, before the alleged evidence for the influence of Planet X on the outer planets had been thoroughly discredited), the 48-inch Schmidt was designed to cover large areas of the sky, and no one had mounted a serious search since Clyde Tombaugh. Thirty years later I would tell other astronomers about my search for planets, and they would frequently look at me critically and say, “Charlie Kowal did that thirty years ago, and he showed there was nothing there.”

I had reasons for ignoring the critical astronomers. Kowal had, indeed, done almost exactly the same thing, but thirty years earlier he hadn’t had computers around to do all of the searching for him. He had to look at each pair of photographic plates by eye and slowly search for anything that looked as though it moved from one night to the next. This was the job that I had calculated would have taken me forty years to accomplish, yet Kowal had done it all in something like a decade and in his spare time. I was banking on the fact that the only way Kowal could have looked at so much sky was if he went very quickly and paid attention to only the brightest objects on his photos. The fainter objects might actually be on his photos, but they would have slipped through his net. Many of my fellow astronomers weren’t convinced by this argument and thought instead that I was off on a wishful-thinking fantasy chase. Chad’s discovery of Object X made it clear that they were wrong in principle, and we now had a chance to see if they were wrong in practice. From published records, we found that Kowal had pointed the telescope directly at the predicted position of Object X on the nights of May 17 and 18, 1983. If we could find Object X in those pictures, we would have a twenty-year-old position for Object X, and we would then know its full orbit exquisitely.

Kowal’s photographic plates—and all of the other plates from fifty years of historic photographic work at Palomar Observatory—should have been stored in the airtight humidity-controlled halon-protected vault in the basement of the astronomy building next door to me on the Caltech campus. I went down to the vault, opened the lock, and peered inside, not sure exactly how I was going to find the specific photographic plates I needed among the thousands that were in there. The vault was in general disarray—no one had really used the photographic plates for a long while—but after letting my eyes adapt to the dim lights I could see that the place was laid out like library stacks, with the photographic plates in large manila envelopes arranged like books on the shelves, but by date rather than by author. I excitedly walked down the rows until I found 1983, and then I ducked into the aisle and looked up to where May should be, anxiously wondering what condition the plates would be in. But there were no plates. There was nothing. May of 1983—and several months before and after—were blank spots on the shelves, with little more than years-old dust. If the plates were misfiled, or perhaps had never been filed, the chances of my randomly coming across them in the vast vault were essentially zero.

That night I called Jean Mueller at Palomar. Jean had been involved with the 48-inch Schmidt Telescope for so long that I thought she might remember the Kowal plates and might know if they had ever been stored. She told me that, by chance, she was going to be down in Pasadena the very next day and would be happy to take a look. That day, the two of us went down to the vault, opened the door, and let our eyes adjust.

“I was down here a while ago, and I think I came across them,” she said as she moved down the stacks. She quickly passed 1983.

“That’s where they’re supposed to be,” I pointed out.

She ignored me, kept walking, and four or five rows later, turned left into an aisle between shelves crammed with manila envelopes full of photographic plates. She walked ten feet, turned right, reached up to the second-to-top shelf, pulled down an envelope, and said, “I think they might be around here someplace.”

She wasn’t quite right. She had put her finger on the plates from May 3, 1983—two weeks earlier than I needed. Our plates were about twenty-two inches to the right.

“How are you going to look at them?” Jean asked.

“Um, well, I was just going to look.”

“You won’t see a thing. Here, you’ll want this.” And she led me back to the front, where some decrepit equipment lay in disarray from decades of neglect. She handed me a light box—an ancient wooden tabletop enclosure with a slightly unsafe-looking power cord that, when plugged in, illuminated a photographic plate placed on top of it so that someone could examine it.

“We used to have a blink comparator”—the same sort of device Clyde Tombaugh had used to discover Pluto—Jean said. “Kowal would have used it on these plates himself. But I think that disappeared twenty years ago. You’ll have to just look back and forth between the two plates and see what you see.”

I put the envelopes containing the plates and the light box on an unstable rolling cart and brought them back to my office, almost knocking them down only once, when I had to push them over the lip and onto the carpeting in my building. I set the box on my table, gingerly plugged it in (carefully moving anything flammable from the vicinity), and flipped the lights on.

The plates were initially deceptive. They are rather heavy fourteen-inch-square pieces of glass kept inside large paper envelopes. When I pulled the first plate out of its envelope, I could see nothing at all except a few little marks apparently made by Kowal himself twenty years earlier, perhaps indicating candidate Planet Xs that he wanted to double-check.

Had the plates turned black with time? Was something wrong?

No, when I put the plate on the light box, I could suddenly see hundreds of stars, with large blank patches between them. I leaned over, my eye a foot away, and realized that each little patch of the sky that had looked blank itself contained hundreds of stars. And when I leaned all the way down and put my eye right up to the plate, I could see, it seemed, the whole universe in a single square inch, with countless tiny stars like glints from diamonds and myriads of swirling galaxies. And on this whole expanse of photographic plate, one of those countless tiny stars was, I believed, not a star, but was Object X and was moving from one night to the next.

I laid the plates from May 17 and 18 next to each other. On the two plates were countless stars, in precisely the same spots from one night to the next. Hiding amid them I was looking for one faint blip—Object X!—that jumped slightly between the nights. Only then, looking at the plates, did I truly realize the enormity of what Clyde Tombaugh had accomplished seventy-two years earlier by picking out Pluto from the stars. My job was easier. I knew roughly where to look on the photographic plate. I compared some of the bright stars to a modern star map, zeroed in on the approximate location, and boxed the area on both nights in felt-tip pen (very erasable from the glass surface). I then pulled out a hand-sized magnifier that was designed to ride over the top of the plates, and I started looking. I would look at one field of stars from the first night and try to memorize where everything was before looking at the second night. Was that star in a different place? Oops, no, I had just not noticed it before. How about that? Nope. Just a scratch on the plate. It took me thirty minutes to search one square inch of the photographic plates—about one-third of 1 percent of the total area—before I finally saw it. A tiny star was there one night but missing the next. And a second tiny star appeared the second night in a place where there was nothing the first night. I let out a scream, and then I forced anyone who walked down the hallway in my building for the next half hour to come in to look at the two spots on the photographic plates and see Object X as it had appeared in 1983.

It really was not surprising that Charlie Kowal had missed this one in 1983. It was a barely visible smudge that had taken me half an hour to find when I knew where to look and knew that there was something there to be found.

We now knew where Object X had been twenty years before, which meant that we could compute a very precise orbit for it. Just as important, we demonstrated that our hunt was not in vain. There might be more things out there that Kowal had not seen on his plates.

But first, we needed to get back to Object X itself. The orbit that we found was surprising. Object X goes around the sun every 288 years in an orbit closer to circular than even most of the planets, but it is tilted away from the planets by 8 degrees. Eight degrees might seem small, but compared to the planets it is enormous. What was Object X? How did it get its almost perfect but slightly askew orbit?

Today we still don’t know the answer. We have elaborate theories of how the objects out in the Kuiper belt have been tossed around in their orbits by the giant planets, but all of this tossing both tilts and elongates the orbits. Tilted but circular? All but impossible. Finding out that something you have just discovered is considered all but impossible is one of the joys of science. It is an enormous clue to billions of years of the early evolution of the solar system. If only we knew what it meant. Eventually we’ll piece together enough other parts of the story so that the peculiar orbit of Object X will suddenly make sense.

With the orbit and the position of Object X determined, we could finally try to answer the one question that had been burning in the backs of our minds. How big was it really? From the day of discovery we were convinced that it was bigger than Pluto. But we didn’t actually know that for certain. Object X was so far away that, from our telescope, we couldn’t tell that it was anything other than a point of light. It looked like a star; it was starlike, an asteroid by the literal meaning of the word, though that literal meaning had long ago been forgotten. Object X was bright, but all that “bright” means is that it reflects a lot of sunlight. An object can reflect a lot of sunlight if it has a shiny surface—because it is covered in snow, for example—or it can reflect a lot of sunlight if it has a darker surface but is really big. You would have the equivalent problem if you were on the ground and someone was signaling to you with a mirror high in the mountains. You wouldn’t be able to tell the difference between someone with a small but highly polished mirror and someone with a larger but dirty mirror. Both would reflect the same amount of light in your direction. Both would appear as simple points of light from your distant vantage point.

There was, possibly, one telescope that could see the disk of Object X crisply enough that we might be able to directly measure its size. The Hubble Space Telescope orbits the earth high above the atmosphere and, now that the original defects in its mirror have been corrected, takes the sharpest pictures of anything around. Even the Hubble has fundamental limits—due not to defects but to the laws of physics—as to how tiny an object it can resolve, but I quickly calculated that if Object X was really the size of Pluto, then Hubble’s newest camera, recently installed by visiting astronauts, would have no problem seeing the tiny disk and allowing us to measure its size.

To use the Hubble Space Telescope you have to submit a lengthy proposal—which is accepted only once a year—detailing what you would like to look at and why; then a committee of astronomers looks over all of the proposals and selects those they believe are the very best. The next due date for proposals was not for about nine months. The earliest we could possibly hope to get a picture from the Hubble was in about a year. We seemed to have only two choices. We could announce our discovery quickly, tell everyone that we thought it was likely bigger than Pluto, and then wait for a year to confirm. But our estimate of the size really was just an educated guess. What if our object was actually smaller than Pluto? We didn’t want to have to be in the position to come back a year later and say that the thing we had called a new planet was actually smaller than Pluto after all. Our other option, though, was to wait a year so that we could announce the correct size when we announced our discovery. But we couldn’t delay the announcement of our discovery for a year; someone else might find it in the meantime and not feel the need to know how big it was before making it public. And even if we did delay until after we got images from Hubble, we didn’t think the secret would keep. Once the proposal was submitted, it would be read by dozens of people, and while proposals are ostensibly confidential, we were pretty sure that word would leak out quickly. Luckily, there was a third option.

It is understood that sometimes discoveries will be made that need pictures from the Hubble Space Telescope faster than the process will allow, so there is an official route by which you can appeal for data immediately. Even this route made me nervous. Many, many people would still be reading the request and learning about the object. So I went for an even more direct route. I sent a note to one person I knew who worked for the Hubble Space Telescope. I explained that we had just found something potentially bigger than Pluto and wanted to look at it with Hubble as soon as possible, but we were afraid to go through any of the official routes in case the information leaked. I attached a detailed proposal just like the one that I would have submitted, but requested that the fewest people possible know about it. I sent the note by e-mail and sat back to look at a few more images of the sky, but within about two minutes I had already gotten a reply: YES!

I quickly set to work trying to figure out the right time to target the Hubble. We wanted to make a very precise measurement of the size, so we knew we wanted to take the pictures just as Object X was moving close to a distant star to which we could compare it. I called up archival images of the sky, had the computer draw in the path that Object X was going to take through the stars, and looked for a good time. I found that in only three weeks the object was going to skim past a bright star; the timing would be perfect. I designed the precise sequence of pictures for the Hubble telescope to take and then sat back to wait the three weeks.

Normally that three-week wait would have driven me crazy, but I had a distracting trip planned. I was flying out to Hawaii to use one of the the Keck telescopes—the largest telescopes in the world—to take a first really good look at Object X. Just as with any of the other great telescopes in the world, getting to use a Keck telescope requires writing a detailed proposal explaining what you will use the telescope for and why it is a good use of the time. As usual, the proposal is read by other astronomers, and then three to nine months later you might find yourself assigned to a particular night at the telescope. Unfortunately for us, again, we didn’t know we were going to discover Object X ahead of time, so we couldn’t have already written the proposal. Luckily for us, though, I had written a proposal to do something else entirely at the Keck—to study the moons of Uranus for evidence of icy volcanoes—so I was scheduled to be at the telescope soon after our discovery. One of the unspoken rules of being at a telescope is that once you are there, the night is yours to do with what you want. Yes, I had planned to look for icy volcanoes, but looking at Object X would clearly be a much more interesting and pressing use of the time.

The Keck telescopes sit atop the currently dormant summit of the giant Mauna Kea volcano on the Big Island of Hawaii. At nearly 14,000 feet above sea level, the summit looks more like the sterile surface of the moon than part of a fertile tropical island. The only sign of wildlife I have come across up there was a mouse who must have hitchhiked up in an equipment shipment and who lived on the crumbs dropped by astronomers or others working inside the dome. If the mouse ever got itself locked out of the telescope, it would find nothing to eat for miles around.

While the majestic Hale Telescope at Palomar Observatory looks like part spotless battleship, part elegant WPA dam, and part nineteenth-century high-rise, the monster Keck telescopes look like nothing but high-strung engineering projects. The dome at Palomar is mostly empty space, with the smooth outlines of the telescope truss looming high above in the darkness. The domes at Keck are the same size, but the mirrors on the telescopes are four times as big, meaning that the telescopes are so tightly crammed into the domes that there is nowhere to stand to even get a good perspective on what the telescopes look like. If you take one of the elevators that goes midway up a dome and step outside onto the metal platform encircling the telescope, you can walk around and get some idea of the different components—white girders, sprawling wires and cables, massive industrial-sized cranes—and you will find yourself looking directly into one of the two biggest telescopic mirrors in the world. It’s not one mirror, though; it is a bug eye of thirty-six smaller hexagonal mirrors all arranged into a much larger, almost circular hexagon looking back at you. The mirror itself, all combined, has a square footage only slightly smaller than the house that I lived in.

Later that night, when we pointed the telescope at the faint dot in the sky that was Object X, the mirrors would concentrate all of the light from that immense area onto a tiny spot about the size of the period at the end of this sentence. Our goal was to take that concentrated light and pass it through a system that acts as a prism, to spread the light out, and then look at the different components. By looking at this spread-out light—the spectrum—I hoped that I could determine what was on the surface of Object X.

I was scheduled to be at the telescope for two nights. I arrived in Hawaii a day early to begin to shift my body to a nighttime schedule and to do final preparations far from the distractions of home (including planning a wedding that was now only seven months away). I stayed up late at the observatory’s headquarters refining calculations on the computer, and then I went to sleep with the hope that I would sleep until noon so I would be fresh for the long night ahead. Instead, I woke up before dawn. I tried to force myself back to sleep, but my mind was uncontrollably running through the plans for the night, how I would set up the telescope and instruments, what would be the best way to collect the most useful data possible. I gave up on sleep and walked over to the telescope control room to set up for the night.

The control room is arranged as a dense ring of desks around the center of the room, with an even denser ring of computer screens. At last count the room had something like twelve computer screens, all of which might be in use during the night. I checked the weather reports, the telescope reports, how things had gone the previous night. All of the nighttime staff from the observatory were still asleep, but there was plenty of preparatory work to do. At lunchtime, I walked to the shopping center to get some fresh Hawaiian poke from the grocery store.

Walked to the shopping center? No, there is not a shopping center on the desolate summit of Mauna Kea. I was in the little cowboy town of Waimea, only a couple of thousand feet above sea level and surrounded mostly by ranch land. To use the Keck telescope these days, astronomers rarely actually go up to the summit. Instead, we sit in the control room in Waimea and connect to the summit by a fast video and data link. We talk to the people there and control the instruments there, but we don’t go there ourselves.

The first time I used a telescope like this while being in a control room miles away, I felt strangely disconnected from what was going on. I couldn’t walk outside to feel the wind and humidity. I couldn’t check for cloudy patches or impending fog. I couldn’t hear the reassuring clanking of the dome and rumbling of the telescope. How could I do astronomy this way?

The answer is, nearly perfectly. Your brain doesn’t work very well in the sudden oxygen deprivation of 14,000 feet. Combine that with lack of sleep, and efficient work is extremely hard. Fish-eye cameras pointing at the sky are better at seeing clouds coming and going than your eye will ever be. Wind and humidity gauges work just fine. And the video link is so seamless that you almost forget that you’re not talking to someone sitting right next to you. Still, I always find it disconcerting when, on nights that I am working at the telescope and the sky at 14,000 feet is beautiful and clear and the humidity is low and we are collecting beautiful data, I think to look out the window and, outside the control room at 2,000 feet in Waimea, rivers of rain are being driven horizontally by gale-force winds.

Object X was going to rise above the horizon at about 8:00 p.m. I had finished setting everything up and was waiting anxiously to get started for the night. The crew arrived at the summit around 5:00 p.m., and we chatted over the video about the plans for the evening. When the sun went down, the big dome swung open and the thirty-six little hexagonal mirrors pointed together to begin collecting the light from my first target in the sky.

My first job was to do a very quick check of all of the systems. We swung to a nice bright star, focused the telescope, and put the light from the bright star down through the prism to see if everything worked. After a few minutes, the spectrum appeared on one of the computer screens in front of me. I typed a few commands to take a quick look; the spectrum of the star looked just as it was supposed to. I stored the data away to later compare it to Object X. Finally, it was time to find Object X. We turned the telescope in the right direction and took a picture to see what was there, and the picture that appeared a minute later on my screen showed that there were twenty stars more or less where I expected Object X to be. Which one was it? I knew how to find out: It would be the one that moved. We did a little more calibration, and then twenty minutes later we took another picture. At first glance, the picture looked precisely the same, but I lined up the two pictures on the computer screen and blinked back and forth between them. Nineteen of the twenty stars reappeared in exactly the same place. One of the stars had shifted slightly. It wasn’t a star. It was Object X.

Though we had been studying it and tracking it for more than a month now, my first view of Object X through the giant Keck telescope—or at least on the computer screen twelve thousand feet below the giant Keck telescope—still amazed me. I was about to get the first peek at the composition of something that might be bigger than Pluto, something that only a handful of people on the planet even knew existed. I shifted the telescope slightly to direct the light of Object X into the prism, and we were ready. Though Object X was the brightest thing beyond Pluto that had ever been seen, it was still faint. Even with the biggest telescope in the world, we had to collect a large amount of light before we had enough to be able to make a sensible analysis. We stared at Object X all night long, stopping every once in a while to be sure that the light was indeed going into the prism. I watched the data come in and obsessively checked the weather reports. Everything went perfectly. No clouds, no fog, no telescope malfunctions. Everything went so perfectly that it was, to be honest, an incredibly tedious night. I occupied myself with loud music, junk food, double-triple-quadruple-checking that everything was going perfectly, and speculating about what I might find.

The sky began to brighten with the rising sun at around 5:30 a.m., and I finally made my way back to my little room. I slept until almost 11:00 a.m., went back to the control room, and again began preparing for the night. The second night was almost exactly like the first. I went to sleep around 6:00 a.m., got up the next day at 10:30 a.m., and was on a flight back to LAX by 1:00 p.m., confident that I had collected exactly the data I needed.

Two nights at the Keck telescope will provide weeks’ or even months’ worth of data to pore over. Though totally exhausted, I got started on the five-hour airplane ride back home, trying to use all of the pictures and data to create one coherent view of what we had seen. First, I had to carefully remove any effects that were caused by the telescope or the prism or the earth’s atmosphere rather than by Object X itself; second, I had to figure out what we were seeing; and third, I had to figure out what it all meant.

It quickly became clear that we were seeing dirty ice. Perhaps that should not have been a big surprise for something so far from the sun. Ice was supposed to be one of the main components of Pluto, too, and it was on the surface of almost all of the big satellites of Jupiter, Saturn, Uranus, and Neptune. But in addition to the dirty ice, there appeared to be something that looked like frozen methane. Methane would perhaps not be surprising to find on the object’s surface, since it is one of the main components of the surface of Pluto, but it had never been seen anywhere else in the Kuiper belt, and the signature of methane was not overwhelmingly convincing. If methane was there at all, it was in extremely small amounts. A few years later another astronomer would suggest that perhaps there was no methane at all on Object X, but that what I thought looked like methane was actually evidence for the same icy volcanoes on Object X that I was supposed to have been looking for on the satellite of Uranus to begin with.

The methane on Object X (and it was methane, after all) never made sense until years later, when Emily Schaller, a graduate student of mine working on a Ph.D. dissertation about the methane clouds on Titan, walked into my office with an idea for why Titan and Pluto both had methane. Her final explanation was deceptively simple and explained not just these objects but the rest of the Kuiper belt as well. Object X, it turned out, formed with methane—as did Pluto and Titan—but Object X was just a little too small, so that its gravitational pull was not quite strong enough to hold on to the methane forever. With the Keck telescope we were seeing the very last remnants of frost on a cold, dying world.

While I was still working to understand the data from the Keck observatory, the Hubble Space Telescope snapped its sequence of pictures and transmitted them to the ground, where they were sent to my computer in Pasadena. Because the Hubble is totally automated and you design the entire sequence ahead of time, you can very easily lose track of when the telescope is actually looking at your target. The Hubble pointed at Object X on a Saturday, as I was having a housewarming party to welcome Diane as a new resident of my—now our—home. The house, with a square footage only slightly larger than that of the Keck telescope, was a bit of a tighter fit now. I didn’t make it to work until Sunday afternoon, after a long cleanup from the party. The new data would immediately tell us how big Object X was. Much bigger than Pluto? Only a little bigger? A tad smaller? When I first opened up the file that contained the image, I immediately closed it and double-checked what I was looking at. Clearly this was not Object X, the object potentially larger than Pluto—how could it be? But yes, the tiny dot that surely couldn’t be the tenth planet was, indeed, Object X. Object X, in the end, turned out to be only about half the size of Pluto.

How could this be? How could we have turned out to have been totally wrong? The answer, in a single word, is albedo. Albedo is a measure of how reflective something is. Freshly fallen snow has a high albedo, while coal or dirt has an albedo that is quite low. No one really knew what albedo to expect for things in the Kuiper belt, but back when the first object was found, everyone assumed that they were dark—as dark as coal or soot or ash. When we see an object out in the Kuiper belt, all we see is sunlight reflected from the surface. If that surface is dark and doesn’t reflect much light, the object needs to be big to reflect a lot of light, but if the surface is icy or shiny for some reason, it can reflect just as much sunlight while being smaller. It turned out that Object X was not as dark as coal or soot or ash; it was more like ice with a bit of coal or soot or ash thrown in. It was shinier than we’d initially guessed, meaning that it was smaller than we’d thought.

I was disappointed at the time, but only a little. We were just getting started, and we had planets in our sights.

Now that we finally knew how big it was—no planet for sure—it was time to give Object X a more dignified name. There are rules, decided upon by the International Astronomical Union, for the naming of most everything in the sky. Craters on Mercury have to be named for deceased poets; moon of Uranus are named for Shakespearean characters. For this type of object in the Kuiper belt, the rules said that the name had to be a creation deity in a mythology. After some quick thought, Chad and I decided that we should move from Old World mythologies, which have been traditionally used, to New World mythologies, in honor of where Object X was found. We even thought we might try to preserve the X. If you’re looking for New World mythologies and names that begin with X, you can do no better than the Aztecs. They were fond of X names—Xiuhtecuhtli is one of my favorites—but none of those felt quite right, or quite pronounceable. A little more Internet searching brought us to consider more local deities. Object X had been found at Mount Palomar, which is surrounded by Native American tribal reservations. Did the Pala tribe have deities? The Pechanga tribe? What gods did they worship in earlier days? We searched the Internet but couldn’t find any; our search brought up only early-eighties entertainers who were currently playing at their massive Harrah’s casinos, whose Las Vegas–style lighting is slowly ruining the view of the sky above the telescopes on top of Palomar. But we did find something even more local: The Tongva tribe, mostly known as the Gabrielino Indians because of their proximity to and assimilation into the San Gabriel Mission, had long been the inhabitants of the Los Angeles basin. In their mythology, the world was begun when their creation force—called Kwawar—sang and danced the universe into existence. It occurred to us, though, that there were actual members of the Tongva tribe around and that we really should ask their permission first.

We didn’t know anyone in the Tongva tribe, but Chad went to, found a phone number, and called it. The chief answered. Chad said something like, “Hi, I’m an astronomer from Caltech, and we just discovered something big in this region of space called the Kuiper belt and were hoping to name it after a Tongva creation myth and wanted to talk to you about it,” at which point the chief probably thought there was a pretty good chance that Chad was a lunatic rather than an astronomer from Caltech. Perhaps to hedge his bets, or perhaps just to get rid of Chad as quickly as possible, he gave the name of the tribal historian and chief dancer, who would be a better person to talk to about such matters.

Chad made the next phone call. After Chad convinced the tribal historian that he was not a crazy person but was indeed an astronomer who had found something half the size of Pluto that needed a name, the Tongva agreed that Kwawar—or rather Quaoar, their preferred spelling—was the appropriate name.

The correct pronunciation of Quaoar sounds like Kwa-o-ar, with a very soft W sound and a bit of a Spanish roll to the R, no doubt a product of the mission days. Simply saying Kwawar works fine, too. But when we picked the name, it didn’t occur to us that if you didn’t see it spelled Kwawar originally, as Chad and I had, the English language doesn’t give many clues on how to pronounce the word correctly. No word in the entire English language has that particular combination of four vowels: aoaa. People trying to pronounce it tend to start with the Q and then quickly trail off into nothingness.

With a name in place, we were now ready to announce to the other scientists and to the world what we had found. A large international meeting of astronomers was taking place in Birmingham, Alabama, just two hours from my hometown, and we decided to make the announcement there. Chad submitted a paper with the innocuous-sounding title “Large Kuiper Belt Objects.” In his talk, he discussed everything that we had learned: Quaoar’s oddly circular yet inclined orbit, its diameter about half the size of Pluto’s, its icy surface. All of the questions, though, had nothing to do with Quaoar. Most of the inquiries from the press that day and over the following weeks never even mentioned Quaoar itself. They just wanted to know one thing: what did this discovery mean for whether or not Pluto was a planet?

What, indeed? Even as more and more objects in the Kuiper belt were being found, Pluto still stood out as being significantly larger than any of the rest—but it was larger than Quaoar by only a factor of two. Was that enough to doom Pluto? In many ways, the answer was clearly yes. If after only nine months of looking, we could find something half the size of Pluto, how much longer would it take to find something the size of Pluto? We figured it was only a matter of months. For the confirmed Pluto fans, finding something smaller than Pluto meant nothing; Pluto was still the biggest, and thus they could go on calling it a planet. Yet it seemed that perhaps Pluto, while not yet dead, was on its deathbed. As The Birmingham News quoted me as saying later that day, Quaoar was a big icy nail in the coffin of Pluto as a planet.

The week after we returned from Birmingham, Caltech threw a black-tie dinner to announce the kickoff of an ambitious fund-raising campaign. Many of the people at the dinner were donors who had been with Diane on one of her many Caltech travel-study trips around the world. Having just been in the newspapers a week earlier for the discovery of Quaoar, I was a minor celebrity at the party. Being engaged to Diane, though, made me a major celebrity.

I spent the evening in a conversational loop: “You’re the person who discovered that thing out past Pluto?”

Yes, indeed.

“I want to introduce you to my friend—hey, do you know Mike Brown? He’s the guy who discovered the thing past Pluto.”

“Sure, I know Mike; he’s the guy who is engaged to Diane Binney. Hey Mike, I want to introduce you to my friend—hey, do you know Mike Brown? He’s the guy who is engaged to Diane Binney.”

“Sure, I know Mike Brown—he’s the guy who discovered that thing out past Pluto. Let me introduce you to a friend who is really interested in planets.…”