JUST AS THE fall migration had captured the imagination of scientists and citizens for the better part of a century, maybe more, the spring movement of monarch b.u.t.terflies had also piqued their curiosity. Once the Urquharts found the Mexican overwintering grounds, they began tagging b.u.t.terflies there, hoping to recapture them once the colonies had broken up. The a.s.sumption was that the b.u.t.terflies went north; a recaptured monarch would be good evidence to support that hypothesis. But it was one thing to demonstrate that monarchs from the United States and Canada spent the cold months in the Mexican mountains, and quite another to say that the monarch b.u.t.terflies that had wintered in Mexico spent the spring and summer in North America. It was a third thing altogether to say-to demonstrate-that the b.u.t.terflies that made the trip south in October were the very same ones that flew north in March. That would mean that those b.u.t.terflies were living for six and seven months or more. It would mean either that they were recolonizing the entire northern range, moving up the lat.i.tudes like a ladder, breeding and laying eggs as they went, or that they were going part of the way, breeding, and then dying. In that case no individual monarch would make the entire round trip. In that case forget bird migration-the comparison would no longer be apt.

How did the spring leg of the migration work? Back in the nineteenth century, when Charles Riley, the Missouri state entomologist, was pondering the destination of the swarms of monarchs he was seeing each fall, another question was on his mind as well. Riley supposed that the b.u.t.terflies were heading south, like birds, but then what? Worn, tattered monarchs had been found each spring in the southern United States, but they had merely raised more questions than they answered. Who could know where those b.u.t.terflies had come from, and when? Even tagging data, when they were later obtained, were inconclusive. How could it be proved that a monarch b.u.t.terfly tagged in Minnesota in September and found seven months later in Texas, for example, had spent the winter in the Mexican highlands, unless it was recaptured there as well? And the chance of that happening, statistically speaking, was nil.

Enter-again-Lincoln Brower. While his sometime rival Fred Urquhart was occupied with his tagging project, Professor Brower was in his laboratory at Amherst, continuing the work on chemical defense in monarchs that he had begun as a graduate student at Yale. There he had shown that monarch b.u.t.terflies were distasteful to birds because of the toxicity of the milkweed they ingested as caterpillars. Now he and his colleagues turned that conclusion on its side and examined it from a different perspective. Since the monarchs stored the toxins-the cardenolides-in their bodies, and since different species of milkweed had different and specific concentrations of the cardenolides, Brower and his colleagues surmised that they should be able to determine which plant or plants a b.u.t.terfly had eaten in its larval stage. And, they reasoned, since the plants were geographically specific, growing exclusively in some places and not in others, knowing which plants it had eaten as a caterpillar would reveal where that b.u.t.terfly had come from. They called the process cardenolide fingerprinting. It did have that "the jig is up" quality to it.

To test their hypothesis, Brower and his a.s.sociates collected fall migrants, b.u.t.terflies at the Mexican overwintering colonies, springtime monarchs from Texas, and monarchs found in the northern United States in June. As they suspected, the first group, the fall migrants, had fed on Asclepias syriaca, the big, broad-leaved, common milkweed that grows north of the thirty-fifth parallel. No surprise there. It was the other groups that told them things they could have only guessed before. While both the winter monarchs and the faded ones found in Texas in the spring showed the syriaca pattern, those that had been captured in the North, where syriaca was prevalent, had the fingerprint of two southern milkweeds, viridis and humistrata. To Brower this was "definitive evidence" that the successive-brood theory was right: the migration was a kind of relay race in which fall migrants pa.s.sed the genetic baton in the spring to offspring whose offspring then continued moving north until they had colonized the entire range and it was time to head south again.

The evolutionary adaptation that had led to this kind of sequential migration had another interesting feature as well. Viridis and humistrata were both high in cardenolides. The monarchs that carried their fingerprints were extremely toxic to birds, while the ones that carried the syriaca pattern were less so. While this might seem to put the generation that left Mexico at risk, it did not. As these monarchs reached the northern tier, their predators had not yet fledged their own young; by the time that occurred, that generation of monarchs would already have reproduced, and the new generation, born to the southern milkweeds, would take its protection from them.

AT ABOUT the same time that Bill Calvert was writing from Angangueo that the spring fling had begun, Chip Taylor, sitting in his office at the University of Kansas, was banging out his own a.s.sessment. His was based not on firsthand observation but on what he had learned, over years of sorting through the anecdotal information that came his way, of the monarch's biological clock. This was the same clock that Lincoln Brower had referred to the previous year when he worried that the monarchs' departure from the overwintering site two weeks early was a sign not of an overeager b.u.t.terfly population but of habitat destruction that had served as an eviction notice.

On Tuesday, March 17, Taylor pointed out that "during most seasons, the majority of the monarchs leave the overwintering area during the last two weeks of March. Some pockets of monarchs, perhaps those in the cooler and most protected sites, remain in the overwintering areas through the first week of April. Worn and tattered monarchs should begin appearing in Texas and Louisiana in good numbers in the next four to eight days, weather permitting." It was a tense and expectant time for monarch watchers, who were waiting to see when, and where, the b.u.t.terflies would land. It was a time not unlike those few long minutes I remembered from my childhood, when the Apollo astronauts, tumbling to earth, would lose radio contact, and no one would know where they were, we could only imagine them falling, and imagine the heat, but not really know, and then someone would see it, a streak in the sky, and hear a voice against the silence, and the relative safety of an open parachute and finally the arms of the sea, reaching up, snagging it. Four days, eight days, and until then, where?

Three days after Chip Taylor's message, Gary Ross, a lepidopterist in Baton Rouge, posted one of his own: "First migrating monarch seen yesterday (3/19/98). It was a male [whose] wings were medium worn.... No others seen." Another lepidopterist, a young Canadian named Phil Shappert, who was attached to the Stengle-Lost Pines Biological Station near Smithville, Texas, reported finding a worn male a week later. Harlen Aschen, in Victoria, Texas, saw a tattered female around the same time. "No missing parts but seemed exhausted," he wrote. One after the other, reports like these started coming from Texas and Louisiana, just as Chip Taylor had predicted. If this had been ma.s.s transit, it would have had an excellent on-time rating.

As regularly as the migratory pendulum seemed to swing, the migration itself was different each year. The numbers were different, the pathways were different, the conditions were different. When the monarchs left Mexico two weeks early in 1997, not only were they early, they were ahead of the milkweed, causing some concern that they'd keep flying north until, not finding any, they died. The 1998 exodus adhered to a more typical calendar, though it was curious how the b.u.t.terflies abandoned their roosts to move down the mountains to form bud colonies at lower elevations. No doubt this had to do with water, which had to do with drought, which had to do with both weather and logging. But the spring of 1998 brought a related concern as well, one suggested by Betty Aridjis's anguished message from Contepec that past winter: fire.

"We witnessed many fires burning in forests all over the states of Mexico and Michoacn," Bill Calvert said in mid-March. "Fires were so frequent and dense that a permanent haze was evident in the sky. None of these were 'serious' fires such as the crown fires that we hear about in our northern forests. All were ground fires burning along the forest floors. They created a lot of smoke and, locally, a lot of heat. One such fire was burning near the Chincua colony located in the Arroyo Honda about five kilometers northwest of Angangueo. Although the smoke from this fire was clearly visible, it apparently has not affected the Chincua b.u.t.terfly colony."

Even so, the fires continued to burn all through the spring, some twelve thousand of them, almost all a result of the unhappy collusion of drought and slash-and-burn agriculture, producing smoke that crossed the border and covered parts of the U.S., too. In Mexico City the air rained soot day after day and residents were asked to stay inside with their doors and windows sealed shut. The monarchs, meanwhile, had little choice but to push through this gauntlet of particulates. Like the wind that carried it, smoke was not itself a predator, but for monarchs heading north it might as well have been one.

FOR THOSE WHO take delight in the sight of a monarch b.u.t.terfly coursing through the air or dipping into the still of blue asters, the first spring monarch is thrilling. Its now-you-don't-see-it-now-you-do trick, as it appears suddenly and out of nowhere, too small to have been picked up at a distance like the approaching ducks, and unannounced, unlike the jays and redwing blackbirds, can bring you up short. Where I live the first monarch may float in in late May or June, or July at the latest, and claim the black-eyed Susans and goldenrod. Others follow in a desultory way. There may be three in the yard one day, ten the next, then two, then fifteen.

Following the spring migration over the Internet was efficient, the whole map of the country progressively filling up with dots from bottom to top, each dot representing someone's story of a particular moment that connected his or her life to the life of a small winged creature. But following it over the Internet was frustrating, too. The monarchs could seem less real than symbolic, an icon of the natural world and its mysteries rather than the mystery itself. "As you wait for the monarch migration to reach your hometown, survey the area for milkweed plants," Journey North urged readers casting about for something to do. I didn't have to leave the house. There were a good four inches of crusty snow blanketing the ground. I knew there was milkweed under there, but it had not yet awakened.

"We are having very strong south winds, with gusts up to thirty-five mph," Bill Calvert reported from Austin on March 31. "These winds are undoubtedly driving the monarchs north." For some of us, though, not soon enough.

The ice went out of the pond two days later, retreating by the hour till the open water glowed like new skin and salamanders came to the surface, only to be caught by my daughter, who had fast hands. "I'm letting you go to see the whole wide world," she said to one of them, holding it aloft and turning a perfect circle. She had just turned five, and the pond and the ocean of land surrounding our house were her whole wide world.

On April 12 I surveyed the milkweed again. Nothing doing. The next day I flew down to Texas.

I think I expected to see scores of monarchs in Texas, faded and worn, the same ones, generationally speaking, that I had seen in the Adirondacks in September and in Mexico two months later. I also expected to see the b.u.t.terflies in droves, packs of them nectaring and laying eggs, together for the last time before fanning out and dying. I was wrong about that: they didn't come across the border in swarms and they didn't gather in groups. The spring migrants were solitary fliers, and though their general destination was known, their touchdown spots were scattered and diverse. To someone hoping to see them, their behavior seemed maddeningly random.

"SO DID YOU BRING the monarchs?" Bill Calvert asked as I slid into the pa.s.senger seat of his reliably cluttered pickup, pushing aside a package of pink hot dogs and a book on the philosophy of science as I did. The Bach tape was still on the dashboard, and maps were piled on the floor, and I knew without turning around that his battered net and makeshift extension pole were wedged in the back behind me. It could have been five months earlier, but for the odometer, Calvert's superego, which registered all the miles he had logged going back and forth to Mexico.

"I thought the monarchs were here," I said, "that's why I came."

Bill Calvert rubbed his mustache and looked amused. "Nope," he said.

"Where are they?" I asked. "Farther north?"

Calvert continued to look amused, maybe because we were in the vicinity of a Luby's cafeteria. "I'm remembering now," he drawled, "that you ask good questions." I was remembering, too: his penchant for answering them with a hapless, almost merry "I don't know."

It was hot in Austin. Bill Calvert was in his trademark jeans and scuffed brown oxfords and a T-shirt. The back of his neck was already deeply rouged by the sun. For me, flying south, the seasons had elided spring, and now it was summer. And somehow, though it was only mid-April, I had missed the monarchs. Calvert, who ran the Texas arm of Monarch Watch, said that most of the reports he was seeing placed the bugs in the northern parts of the state. But even there they were scarce.

By then we were off the interstate and on a two-lane county road that made the transition from suburb to farmland to scrub desert in less than half an hour. Cattle were grazing and oil pumps nodded like obsequious servants, but all in all it was pretty empty territory. We turned down a wooded driveway marked Masters School, and Calvert stopped the truck. The place was almost deserted except for a lone rooster hopping around its enclosure, making a racket.

"This is it," Bill said, making a broad sweep with his arms. His arms opened to a wide, gra.s.sy, unshaded field. "This," he said, "is my study site."

I looked around. A couple of buckeye b.u.t.terflies were ambling through the air, and a red admiral clung to a metal fence. Overhead a purple martin flew by, while a pair of yellow sulfurs mated close to the ground. Calvert was right: no monarchs.

He got out the hot dogs and his knife and handed them to me. "Cut these up," he said. "If you don't mind," he added. The hot dogs were slimy and warm, and I told him so. "You didn't think we were going to eat these, did you?" he asked.

"You're the guy who gets his vitamin A at Luby's," I said, slicing the hot dogs to fit into a container that looked like a medicine vial, the kind that pills come in.

"OK, here's the idea," Bill said conspiratorially. "We're going to put the hot dogs in the containers."

I nodded. This much I had figured out.

"Then we are going to lay them around each metal enclosure, inside and out." He pointed to the field. It was dotted with three circular fences, each about fifteen feet across. Inside each enclosure was gra.s.s, and milkweed. "We use the hot dogs to see if there are any fire ants here. They're bait-the hot dogs, not the fire ants. I think the ants are major monarch predators." We laid down the vials around each circle as if each one were marking off a piece of pie.

"Now we wait," Bill Calvert said. "I'm hoping that there will be a preponderance of ants on the outside of the enclosure and virtually none inside."

While he waited, Calvert took an inventory of the monarch larvae on the milkweed inside the enclosures, all the while talking into his tape recorder. "Four-inch shoot with nothing on it," he said. "About two-inch with fifth instar larva and fourth instar. Whoa, there's another one. I have an asperula, nine inches, with nothing on it. Sixteen-inch with mature buds."

When he'd finished inspecting the milkweed inside the fences, he turned to milkweed plants nearby but on the other side of the metal, flipping over each leaf to look for caterpillars.

He turned to me. "Well, this experiment is working well," he said. "I only have larvae inside the enclosures." The alarm on his watch began to beep and he bent over to pick up the first vial, which had been stationed outside the enclosure. "Oh, no," he said, "this is no good. There aren't any ants."

Calvert moved on, checking each vial in each enclosure and then the ones on the outside. At the second enclosure he held up the first vial. "No ants!" he called. He picked up the second, third, and fourth ones. "No ants!" he called again. "Only one more to go." He bent down and picked it up and waved it around. It, too, was ant-free. "Now this is what I like to see!" he exclaimed.

"Oh, no, this isn't good," he said a second later. "This is not good." He was kneeling in the third enclosure, frowning as he watched a gang of ants chew on a chunk of hot dog. Two of the five vials had ants in them. "This is not a good result. We're just going to have to nuke it," he said, taking out a foul-smelling poison and spreading it on the ground around the perimeter of the fence. "I hate to use it, but I need an experiment," he said, distracted for a moment as a single monarch, so worn it was almost translucent, came into range. "There's your monarch," he said to me, barely looking up.

The monarch flew around, it seemed, aimlessly. There were bluebonnets in bloom, and evening primrose, but neither seemed to do it for her. She was three feet off the ground and flapping.

"Why doesn't she do something?" I asked Bill Calvert, who looked at me like I was an idiot.

"What I'm saying is," I continued, "if it's the female's job to get through the winter in order to lay eggs, why is she flying around the field and not landing and not laying eggs?"

Bill Calvert, who was on his knees measuring the milkweed plants, looked up and gave me a big smile that, I have to say, made me feel like less of an idiot.

"What I'm saying," he said after a long pause, "is that I don't know."

I TOOK A WALK then and left Bill to his measuring. I watched him from a distance, the scientist at work, crawling around and checking the backs of milkweed leaves for caterpillars. "Got a big one!" he'd call, or "Lots of monarch bioma.s.s here!" Calvert was also checking to see which of the two milkweed varieties growing in the field was preferred by monarchs. It was too early to tell, but the results would dovetail nicely with Lincoln Brower's work on cardenolides. If given a choice, would the monarchs choose the milkweed that would give their offspring more protection? One might expect so, and Bill Calvert was eager to see what would happen. A half hour went by, then another. Calvert kept measuring and talking into his tape recorder. "This really gets exciting right around now," he called to me as I went over to the truck and tried to figure out a way to sit in its shade without attracting fire ants into my shoes and pants. "Who else would come and spend three hours in the hot sun except someone who is really excited?" he said.

THAT TRANSLUCENT MONARCH we saw noodling around Bill Calvert's study site was the last one of 1997 that I saw. Two other monarch b.u.t.terflies settled into the Indian paintbrush that afternoon, but they were bright orange-fresh ones-as were the few I found nectaring along the highway as I headed into Houston the next day. The fresh ones were the freshmen of 1998. By the time I'd start to see monarchs again in the Adirondacks a few months later, they would be three, maybe four generations removed-seniors, or postgraduates, to stretch the metaphor. But that a.s.sumed that each generation would do its part. That the chain would not be broken. Yet that, observers in the southern tier began to notice, was not what was happening. There were fewer monarchs, for one thing, or at least fewer monarchs being sighted, as Bill Calvert had told the readers of D-Plex just a few hours before I got off the plane in Austin.

"In general we are not seeing as many adults or eggs and larvae as we did last year at this time," he wrote. "The greatest numbers of adults were seen in late March and very early April. Since then numbers of adults have tapered off. We continue to see eggs, but not very many. It may be a ho-hum year for monarchs!"

But ho-hum it wasn't shaping up to be. The numbers were definitely down. A comparison of sightings reported to Journey North showed that during the first two weeks of April there had been fifty-nine sightings in 1997 and only twenty-three in 1998. And that was not all. Monarchs were showing up in strange places at strange times. One was seen crossing Bancroft Point on Grand Manan Island in New Brunswick, Canada, on May 4, nearly two weeks before a monarch had ever been seen there before. And then there were the fifty monarchs seen crossing the Gulf of Mexico, flying against the wind. That one caused a lot of speculation, but one man, David Gibo of Toronto, didn't find it peculiar at all. Gibo, as I was soon to find out, was a glider pilot with an encompa.s.sing interest in flight vectors. He was also a biologist, a professor at the University of Toronto at Mississauga. Those fifty monarchs had probably been riding thermals, Gibo supposed, and when a thermal that has formed over land drifts over water, the warm air feeding it is cut off and it breaks up-in this case causing the monarchs to descend and forcing them to fight their way back to land. "Were the b.u.t.terflies simply unlikely individuals that happened to have started too close to the coast when they picked up their first thermal of the day?" Gibo asked. "Probably. Were these sightings unusual? Not at all."

Even so, the numbers were down significantly, and no one knew why. Fire, drought, and logging were the obvious culprits, but so was nature itself. Population fluctuations had been observed for so long that some biologists, most notably Fred Urquhart, thought the monarchs were on a seven-year boom-and-bust schedule that reflected the rise and fall of certain parasitic killers. It wasn't so, at least not the cyclical part, though disease could be rampant. When I got home I called Bill Calvert to get his opinion.

"Why are the numbers down?" I asked him.

"Are they?" he replied.

"Well, the number of reported sightings is down," I said.

"Right," he said, making his point. I decided to change tacks.

"What if the numbers were down this spring? I mean, they were so high last fall, what if the spring population basically crashed? Why would that be?"

There was silence for a moment, and then, if a smile can be audible, I heard one traveling across Bill Calvert's face.

"That's a good question," he said.

Chapter 8.

THE MILKWEED BEHIND my house began to rea.s.sert itself, slowly at first and then with more vigor. Two inches in the beginning of May; ten more by month's end. Green with promise, it stretched skyward, adding leaves and stem, then more leaves and more stem, like a stunt done with mirrors. The raspberry canes that had lain brown and dormant among the milkweed began to wake, too. Horizontal on the twelfth of May, they had sprung some twenty degrees two weeks later and continued degree by degree, like someone rising from a long and deep sleep, till, heavy with fruit, they stood perfectly vertical amid the tangle of milkweed.

The birds came then, the thrushes and the warblers, and then the b.u.t.terflies and the bees. Traffic was heavy with all their comings and goings. Pollination, copulation, oviposit-ing, and predation were undertaken with such diligence that I began to think of the milkweed patch as a small enterprise zone or industrial park. There were beetles and earwigs, ants and wasps. And spiders and gra.s.shoppers. What there wasn't, though, was monarchs. The monarchs were conspicuously absent. In years past there had always been a few in early summer, outliers who had come up from the Midwest on a northern wind, second-generation migrants whose offspring might head up to Canada or stick around and breed the generation that would go down to Mexico. But this was not a typical year. The monarchs were missing, and not just from my backyard, and not just in May, and not just in June. Not one monarch b.u.t.terfly lit upon the joe-pye weed that grew casually around the perimeter of the property. Not one was sighted on the goldenrod in August. Their absence was a kind of visual silence-an anxious visual silence. The monarchs always seemed about to appear, but then they never did. And no one could say where they were.

But the world, even the world of my backyard, is a big place. It was possible that the monarchs that were not on my property were two miles over, on my neighbor's. It was possible that they were in Franklin County, not Warren County, or in Quebec, not New York. Or in Vermont, or Maine. Just because they could not be seen, it didn't mean they were not there. Empirical knowledge is tricky that way. Things unseen are not necessarily things not there.

Some days, walking around the yard, turning over the newest milkweed shoots to look for monarch eggs and caterpillars, I found it miraculous to think that I had seen monarchs (and eggs and larvae) there in the past. They were so small, and so scattered. But they had been there; against the vast backdrop of earth and sky, they had shown up. And if the consolation of empiricism is truth found in regularity, when things stop happening it is unsettling. This was in August. Cl.u.s.ters of mallards and mergansers had already taken to the air. I decided to, too.

"JUST SO YOU KNOW," I said to the man on the phone, in the interest of full disclosure, "I really don't like to fly."

"Many people are afraid of flying," he said in a soothing voice. "But they haven't really flown."

"I fly all the time," I said. "I can't tell you how many airplanes I've been in in the past year."

"Did the airplanes have engines?" he asked.

"I a.s.sume so," I said.

"So you haven't really flown, either," he concluded.

THE MAN ON the other end of the line was David Gibo, the University of Toronto biologist and glider pilot. Gibo had logged hundreds of hours in his Grob 103 ACRO, a two-seat, engineless fibergla.s.s aircraft with a wingspan twice as wide as a Piper Cub is long. A glider is lifted into the air by a tow plane, which unhooks it at two or three or five thousand feet or more, leaving it to find its own way back to the ground. It was in the air that Gibo began to understand how a monarch b.u.t.terfly could travel thousands of miles and end up on the side of a mountain in Mexico without apparent damage. But it was from the b.u.t.terflies that he first learned how to fly.

"One September day I glanced out of my office window and saw a monarch," he recalled. "I had just learned to glide myself and I was looking out the window to see if it was going to be a good weekend to drive out to the glider club. I saw this b.u.t.terfly coming toward the building and it started going up and got about a third of the way and then stopped flapping. From then on it went back and forth, turning figure eights or circles, I can't remember which, until it went up and over the building. I said 'Hmmn,' and I knew what it was doing, because if it wasn't flapping its wings it was coming down, unless it was in air that was going up. Buildings, like mountains, create lift."

This was in the mid-1970s. Fred Urquhart's National Geographic article extolling the seemingly valiant, enigmatic, long-distance journey of the monarch b.u.t.terfly had just been published. So it was known where the b.u.t.terflies spent the winter. What was not known was how they got there. And then David Gibo, a wasp guy by trade, saw the monarch catapult the South Building, and it gave him an idea.

"I decided to study flight tactics," Gibo said. "The solution for that seemed obvious. All I had to do was build a powered ultralight aircraft, add a few extra instruments, then fly in the vicinity of the migrants."

But it didn't work that way. Gibo built the plane and hauled it down to Texas at the height of the migration, taking off on a day when he could plainly see thousands of b.u.t.terflies overhead. See them, that is, till he was part of the sky himself. "All my equipment and theories about air currents could not help me see what amounted to a piece of paper set on edge against the horizon, and a camouflaged one at that," he said. "Look down, no b.u.t.terflies. Look ahead and to the side, ditto. I finally realized that I could see them if I looked up, but that isn't a very safe way to fly. Especially at low alt.i.tudes."

Gibo stopped looking for monarchs when he was in the air and concentrated on becoming an expert flier himself. If he understood aerodynamics better, he reasoned, he would understand how an insect with a three-centimeter-long body could travel forty-five hundred kilometers through the air-a feat, he calculated, akin to a six-foot-tall person's circling the globe eleven times in a row. But the monarch was not a person, it was a bug, a bug with a bug's brain. Or as David Gibo liked to say, it was aerial plankton with a guidance system. How complex could it be?

"Without a doubt, [the monarchs"] annual two-way migrations are among the most amazing accomplishments of insects," Professor Gibo a.s.sured readers of his flying, gliding, soaring, and science Web site, Tactics and Vectors, some of whom, he knew, ascribed spiritual dimensions to the unlikely, and seemingly miraculous, journey of the monarch, as if it were nature's vision quest. "Nevertheless, it can't be that difficult. We're talking about an insect. Like all insects, b.u.t.terflies are strong and resilient, but lack special (i.e., magical) powers and are p.r.o.ne to all the limitations that accompany small body size. Compared to migratory birds, migratory b.u.t.terflies are much slower [and] have an inferior capacity to regulate their body temperature and an absolutely ridiculous rate of fuel consumption during powered flight. In short, there seems to be nothing to recommend the b.u.t.terfly body plan, physiology, and nervous system for the task of making regular, long-distance, directed migrations.... Nevertheless, each year millions of b.u.t.terflies, who apparently haven't the good sense to recognize their serious design flaws, somehow manage to make their way across the continent. Apparently, we're overlooking something important here."

What he meant, and he wasn't being pejorative, was that it couldn't be all that complicated. Bugs were bugs. Biologically, physiologically, they were capable of only so much. They were scripted at birth. They followed a distinct genetic set of rules that directed them to fly to the overwintering sites and back. Gibo believed that those rules were discernible. He wanted to uncover them.

"Let's say that there is an international airport and we want to find the regulations that apply to different categories of aircraft," he said to me, reaching for an a.n.a.logy that he thought I might understand. "We also want to know which runways they take off from, and how it is aligned, and what the requirements for noise reduction are. If I just watched what was happening and recorded it, at the end of a year's observations I would say, 'OK, here are the regulations.' I'd be able to infer the rules."

What Gibo did not mean-and this was central to his way of thinking-was that the rules were about probability, the way, say, Adrian Wenner's dispersal model was, or about randomness and chaos, the way a model based solely on weather might be.

"Monarchs don't simply show up wherever the wind blows them," he pointed out. "They seem to show up where it is beneficial to them, at a higher lat.i.tude at the right time of the year, where the crops or fruit plants are. And this seems to be more predictable than what we expect from the weather patterns. I think that going and getting descriptive data and exploring them will give more insight into what's happening than using meteorological data and generating hypotheses that way."

IT WAS THE MIDDLE of August in the summer of the absent monarch b.u.t.terflies. In less than two hours I would be airborne with David Gibo, flying in a two-seat training glider above the farmland surrounding Arthur, a small agricultural village in southern Ontario, about an hour west of Toronto. We had driven up at midday, stopping for lunch at the local diner, a spare eatery with little to recommend it except that it was there and reliably patronized by members of the York Soaring a.s.sociation, Professor Gibo's glider club. Though he was so engrossed in a conversation with a fellow pilot that he kept forgetting to look at his menu, I was obsessed with the French fries and chocolate milkshake that I ordered with abandon on the theory that this might well be my last meal. And if we didn't crash, if I didn't die, I kept asking myself, would I be revisited by this food at fifteen hundred feet?

"In most planes, stability is a good thing," the other pilot was saying to David. Did that mean that gliders were not stable? I wondered. Flying in an airplane without an engine-this was definitely the most dangerous thing I'd ever done on purpose. And it was on purpose. I wanted to feel what it felt like to be carried along by wind. Or I had wanted to feel it. An hour, now, from takeoff, and I was no longer sure.

"Being stable requires less energy," the other pilot said loudly. He had an Australian accent and a confident manner. Gibo had told him I was going up for the first time, and I had the feeling that this speech about stability-a very calming word-was for my benefit exclusively.

"Monarchs are stable in their gliding configuration," David said. "If they have an active control system, their nervous system and muscles are going to be operating, and they'll use more energy."

This I understood. A monarch can carry only about 125 milligrams of lipids-its fuel-in its body. It takes just ten hours of powered flight-the kind of flying that is characterized by beating wings-to deplete that store.

"The idea is to get from here to there with as little energy used as possible," David said. "On the other hand, when they're attacked they have to go into violent maneuvers. They can flap and unflap their wings and beat them in different planes."

But what about us, flying in a craft with fixed wings and no fuel? What did we have to work with?

Not a lot, it turned out. At the gliderport David showed me our plane, a battered twenty-five-year-old Schweizer 2-33 trainer. Disproportionate to the airplane's body, like the arms of a rangy teenager that had grown faster than his torso, the wings spread out on either side of a remarkably small and compact hull. The hull was painted orange in a somewhat haphazard manner, the green and black of earlier paint jobs peeking out here and there. The steel housing was battered and pinged, and the effect was hardly rea.s.suring. It looked like a jalopy.

The glider's wings were orange as well, and rounded on top-airfoil wings. This, I knew, would help keep us aloft. As air moved over the top of the wing, the airfoil would slow it down and disperse it. With more pressure below the wing than above it, the airplane would be pushed upward. This was lift. A monarch's wings were orange with black, too, but all similarities ended there. b.u.t.terfly wings were flat. Lift came from flapping, from churning the air until it created a whirling ma.s.s that moved along the leading edge of the wing.

The gangly orange sailplane, though not nearly as elegant as an orange-and-black monarch, had certain mechanical advantages over a b.u.t.terfly. It had a rudder, located near the tail, that kept the fuselage aligned with the direction of flight. It had a pair of ailerons, one per wing, that controlled airflow and offered lateral control. It had elevators on the tail to make the nose point up or down. It had spoilers to reduce lift. Each of these was available to the pilot should he wish to change the airplane's flight angle, its alt.i.tude, or its direction.

Professor Gibo was explaining this as we hoisted ourselves into the rudimentary and snug c.o.c.kpit, me up in front, he in the back. Feet forward, I wriggled into place like a sausage being packed into casing. The plane-the inside of it, anyway-was narrow and tinny. There just wasn't much of it. And once the Plexiglas canopy was lowered, it seemed smaller still. Not a good place to be a claustrophobic, I was thinking, looking out the bubble overhead. Or to be a control freak, either, since there were almost no controls. Just the altimeter, rudder pedals, spoiler aileron, and elevator stick, and the tow-release k.n.o.b to disengage the umbilicus connecting the glider to the tow plane.

"I don't need to know about any of these, right?" I called to David, looking for a.s.surance that he, indeed, would be piloting the plane. But David couldn't hear me. The tow plane was buzzing up ahead and the yellow rope between us was losing its slack. After another second it grabbed the little orange glider as if it were a recalcitrant child and pulled it down the gra.s.sy runway. We clattered along, then lifted off the ground for a second like a kite on a short string, dipped back down, then took off again for real as the yellow rope stretched and grew taut. Twenty, fifty, one hundred feet and climbing. As I looked down at the receding ground, a line from a nameless poem went through my head: "From this there's no returning, none."

"Watch the tow rope," David called out to me over the loud, maddish complaint of an airplane being yanked through the air. "You'll know when we're going to hit a b.u.mp because you'll see it in the rope first." I guessed this made me feel more secure, though "more secure" might suggest that I felt somewhat secure, which at that moment I did not. We were bobbing around pretty regularly, mirroring the fits and starts of the tow plane but on time delay, like bad lip-synching. Still, knowing when it was going to happen let me tense up beforehand and brace myself.

"Five hundred feet," David shouted, a fact that I accepted ambivalently. Up was definitely better than down, but up meant we were moving farther away from the ground, from the world where gravity was as transparent and unthreatening as air. "Seven hundred feet!" I caught sight of the altimeter. It was inching up. But I didn't need an instrument to tell me that. The fields and farms below were growing distant, flattening the third dimension till it looked as if, really, there were only two.

As we approached a thousand feet, David asked me to call out the readings. "You have to release the rope at two thousand feet," he added.

"I do?" I yelled back. Apparently, I did. There was only one release k.n.o.b, David explained, and it was by my knees. I put my hand on it and looked straight ahead at the rope, which was sending a wave in our direction. b.u.mp.

"Fifteen hundred feet," I called. The higher we went, the more the sun bore through the canopy and spread heat and light relentlessly. No wonder David was in short sleeves and a wide-brimmed floppy hat. There was no escaping the sun. Off to our right I could see another glider, off its tow, making tight circles.

"It's in a thermal," David shouted, seeing it, too. "Don't worry. We won't hit that plane for the same reason you don't run into the car in front of you, or next to you, when you're driving."

Really? I knew plenty of people who had inadvertently rear-ended another vehicle. Didn't they count? "Seventeen hundred feet." The climb was steady, and the horizon stretched in front of us, and with no landmarks there was no way to distinguish eighteen hundred feet from nineteen hundred feet. But we had reached nineteen hundred feet and were pulling up to the invisible station. My hand tightened on the release k.n.o.b, independent, it seemed, of my reluctance to separate from the motor that was carrying us aloft. Something about letting go felt suicidal, like pulling the switch on one's own electric chair.

"Two thousand feet," I sang out to David.