"Ludicrous results," UC Berkeley physicist Richard Muller observes, "require ludicrous theories." Such is the nature of scientific revolution. Continents drift? Space bends time? Man is an ape? Virtually everything we know, or think we know, was ridiculed as a far-fetched solution to odd bits of data when it was first proposed. "The fool will turn the whole science of astronomy upside down," said no less of an authority than Martin Luther about Copernicus' notion that the Earth revolves around the sun.

Take the idea that the dinosaurs were killed off by a massive comet or asteroid slamming into our planet 65 million years ago. Greeted with almost universal skepticism-and a good deal of ridicule-when proposed sixteen years ago by Berkeley physicist Luis Alvarez and his son, geologist Walter Alvarez, it has fairly quickly become the dominant theory on dinosaur extinction as more and more evidence of just such a cataclysmic event has been discovered. And not just evidence of one such catastrophic cosmic event. Richard Muller, a protege of Luis Alvarez, has his own corollary to the Alvarezes'' theory, which he expounds in his 1988 book, Nemesis: The Death Star. Since the Earth's mass extinctions appear to be periodic, every 26 million years (the "ludicrous results" with which Muller had been presented), so too must be the death-dealing extraterrestrial objects. Any such regular cycle must have an astro cause. And Muller believes that (the "ludicrous" theory) the most likely such astronomical cause would be a companion star (which he calls Nemesis) to the sun, with an orbit that would bring it close enough to the inner solar system every 26 million years to shower the Earth with a killer storm of comets.

Eight years later, Muller's star still hasn't been found. But then, the sky is vast. No one knows precisely where to look. And the technology for finding what might be a rather small, dim star-Muller has speculated it may be what is known as a red dwarf-is in a phase of rapid development. Nemesis remains in the not-proven, but by no means disproven, category.

Meanwhile, Muller-whose restlessly inquisitive mind long ago earned him a MacArthur "genius" grant, along with a number of more specifically scientific awards-has been exploring an number of other projects. He's been teaching his regular physics courses at Cal. He's been serving on international disarmament panels with US and Russian scientists. He's been investigating the rate of the creation of craters on the moon. He's been writing a novel about truth, deception, and the life of Jesus, and co-wrote, with Philip Dauber, the recently published The Three Big Bangs an easy-to-read account of cataclysmic physical evolution-from the first, beginning-of-everything Big Bang, to the second, creation of the stars, to the lesser big bangs that caused the mass extinctions. And he's spent the last three years fighting an uphill battle to get his latest "ludicrous" theory published. This one wouldn't as completely overturn our contemporary worldview as would his Nemesis theory, which would alter our fundamental understanding of the entire solar system, but it contradicts one of the basic premises of the study of ancient climates. If Muller's theory about the causes of the earth's periodic ice ages is true, it will alter the way we study the climates of the present and future, as well as the past, and will have an impact on our understanding of the processes that might contribute to major problems such as global warming.

It also could give catastrophe theorists and end-of-the-millennium doomsayers yet another threat from the skies to worry about-something in between the invisible carcinogenic ultraviolet rays bombarding us through our dissipating ozone shield and the massive asteroids hurtling our way: killer dust from outer space. That, actually, would be stretching Muller's theory considerably in the paranoid direction, and would require a major condensation of the time periods involved. But there is something in it. Richard Muller believes that the 100,000-year cycle of ice ages, and intervening warm periods, that has persisted for the past one million years may be caused by space dust. Where the long held dominant theory maintains that ice ages are caused by a periodic decline in the amount of sunlight being absorbed in the northern hemisphere, Muller posits that it's more likely that the Earth is passing through a hitherto unknown band of cosmic dust orbiting the sun.

The band of cosmic dust is hypothetical, but if it is there, it almost certainly orbits the sun on a constant plane-the same plane as Jupiter, the dominant planet in our solar system. But the existence of the dust is the only real hypothetical in Muller's theory. The ingenuity of his proposition is that it's based on applying previously known astronomical facts in an attempt to determine the cause of our cyclical ice ages, facts that previous theories simply haven't taken into account. Where the dominant paradigm -known as the Milankovitch theory-rests heavily on the combined effect of the tilt of the Earth and the eccentricities in its orbit around the sun (sometimes more circular, sometimes more elongated and oval), Muller brings in yet another known oddity about our orbit: its wobble.

Like a top spinning in space-where there would be no friction to bring it to a halt-the Earth's orbit keeps wobbling in and out of the almost fixed plane of Jupiter's orbit. That wobble, the inclination of the Earth's orbit, has been known and measured at least since the Copernican system became established. Sometimes, for periods of tens of thousands of years, we're more or less in line with Jupiter's orbit; for other long periods, we're at a greater degree of inclination from it. What Muller and his collaborator, UC San Diego's Gordon MacDonald, have done, however-and they seem to be the first to have done so-has been to determine the periodicity of that orbital inclination. They're certainly the first to have noticed that the Earth wobbles on a 100,000-year cycle, a cycle that exactly coincides with that of the ice ages.

"I'm learning that some of the most exciting problems in science are the ones that have fallen into the cracks between the fields," Muller says with barely contained enthusiasm. A young-looking 52, his dark mustache just tinged with the gray that accents his tousled mop of hair and fluffy sideburns, he launches into explanations of basic and arcane scientific principles with a boyish excitement that is instantly contagious. His large of fice, in the Lawrence Berkeley National Laboratory's sprawling Building 50 complex, is an almost classic image of a scientist's lair, albeit one with a breathtaking panoramic view of San Francisco Bay through winlong overdue for a cleaning.

Odd gizmos constructed of glass tubing and other materials hang suspended from thickly packed and cluttered bookshelves. Papers are piled high in cardboard file boxes and spill out of overstuffed file cabinets. Next to the standard academic blackboard, covered with scrawled formulae, is a poster of Van Gogh's Starry Night, retitled "The Cosmic Microwave Background." A computer terminal, occupying the place of honor on Muller's main desk, burbles occasionally as new e-mail messages flash across the top of its screen. The massive profusion of binders-with labels like "Nemesis," "The Three Big Bangs," "Pulsar in 1989A," 'Xenospherules," or "How to Get a Date"-testify to the broad range of Muller's interests and projects. Unlike many of his peers, Muller has chosen to range restlessly through the fields of physics, -rather than to specialize.

"I used to think that the great areas for science in the future were out on the extremes," he says, "either very large, out in space, cosmology, or the smallest kinds of matter. These fields have defined themselves very well, with all their various problems. But then there are these really important problems that require knowledge across several different disciplines, and not too many people work on them for some very good reasons. For one thing, you need the knowledge, and for another, you need the funding. It's very hard to go to an agency and say, 'I want to work on the Earth's climate.' they'll say, 'Get me a climatologist.'

"I have done a great deal of research in astrophysics, but I don't think of myself as an astrophysicist. I've also published research on the demise of the dinosaurs, better methods for doing radioisotope dating for archeology. I have patents in medical physics. So I tend to be rather diverse. The thing we're talking about, the theory about the ice ages, requires a knowledge of geology and geophysics, of atmospheric physics, and of several other things, oceanography in particular, and the chemistry of the upper atmosphere."

It doesn't require anything quite so advanced to understand the thrust of Muller and MacDonald's theory. It helps, however, to have a basic understanding of astronomy, or at least the regular workings of the solar system and Earth's place in it-which is a little more complicated than most of us manage to remember from day to day. Fundamental to an understanding of the periodic cycle of ice ages, after all, is the concept of time itself. And, as Muller points out. "The whole concert of time starts with astronomy."

"It was surprisingly obvious to people in the 1800s, when they discovered the ice ages," he explains, "that their cause must have something to do with the Earth's orbit and the sun. They knew the seasons have to do with Earth's orbit; day and night have to do with Earth's orbit. So it was natural to assume that the ice ages did too. And today I think there's a general consensus that that's probably true. The basic evidence is that the ice ages come on such a regular schedule, and they keep to that schedule so precisely that that by itself is the best evidence that it's ruled by astronomy. There is no other natural clock outside the atomic world and the astronomical world. The in-between world, the world of Earth, is quite irregular. It's only the atomic level and the astronomic level that keep time so precisely. And the reason it works is because the planets move basically without friction."

Muller grabs a pencil, leans over, and pushes it across the floor. Once he releases it, it soon stops moving. "Physics says an object set in motion will remain in motion," he proclaims. "It doesn't, does it? Physics is violated every day by everything we see around us. Because of friction. Have you ever taken an introductory physics course? Then you've studied as much about friction as the typical physics major. Because friction is never tackled again after the elementary level. Because it's so hard, so unpredictable. So what do we do? Avoid it. Physics works at the atomic level or in outer space, because they're frictionless. In between, we call in the engineers or the chemists. We wouldn't ask a physicist to design a bridge or a modem. We ask the engineers, the guys who can handle friction. But in space, the laws of physics really work.

"To move without friction is to move with extreme regularity, and something that unchanging has always been of great interest to human beings. Because you may not know when the next rain will come, or precisely when spring will come. But you can really count on when the stars appear, when the sun rises, when the moon waxes and wanes. The ancients were fascinated with this, because once you realize how regular it all is, you develop the calendar."

At the basis of our concept of time are the two fundamental movements of the Earth. The rotation of the Earth on its axis gives us day and night. The orbit of the Earth around the sun, combined with the tilt of the Earth's axis, gives us the seasons. The tilt of the Earth is a constant, not in relation to the sun but to the stars. "That's why," Muller explains, "the North Star always sits above the North Pole." When, as we orbit the sun, the northern hemisphere is tilted toward the sun, we have summer in the north and winter in the southern hemisphere. And vice versa. But our orbit, remember, is not a circle but an ellipse-varying to some degree over long periods of time. At some points on that elliptical orbit, the Earth is closest to the sun, at others, furthest away.

"The interesting thing," Muller observes, "is that we're not closest to the sun on the first day of summer. If we were, summer would be much more intense. Likewise, if we happened to be the furthest from the sun on the first day of summer, then the summer would be less intense. So this was part of the idea of how the climate might be changing. We know the orbit of the Earth is changing slightly. And if we happened to be tilted toward the sun the first day of summer, during the part of the orbit that we're furthest from the sun, then you're not going to have a hot summer. The glaciers persist through the summer. You have an ice age.

"These things all change by little bits. The eccentricity of the Earth's orbit is changing-it gets a little less oval, a little more oval; actually, right now it's a little more of a circle. That's because of the effect of Jupiter, which is great grist for the astrologers, by the way. The changes are by small amounts, about two percent. But let me put that in perspective. A two percent change in the eccentricity of the Earth's orbit might change the Earth's climate by a few degrees, five, ten, fifteen degrees Fahrenheit. That's enough to be the difference between its present climate and an ice age. Because an ice age is not that much different. It's not that suddenly we go from having seventy degrees weather to thirty degrees weather. The average temperature change might be ten degrees Fahrenheit, perhaps fifteen. It's not that easy to measure precisely what the temperature was 50,000 years ago."

Plotted out on a graph, the cumulative effect of these minor average temperature changes over the past million years exhibit both a major impact and an astonishing regularity. The line rises and falls in sharp, jagged peaks, rather like a seismograph during a series of tremors, with the highest peaks spaced 100,000 years apart, and the low points also occurring at 100,000-year intervals. The graph, however, is likely to be a bit misleading to the average layperson. With our normal bias for judging extremes by their deviation from a norm, it's only natural to assume that the ice ages are those periods where the average temperature falls below the median line, while the periods above it are much like today. In fact, it's only the very highest peaks that represent the climate to which we've grown accustomed, the brief-in geologic terms-interglacial periods. Everything else is ice age.

"If we look at the last million years," Muller affirms, "it seems to be a domination by ice with short periods of warm weather in between. In other words, we are in an abnormally warm period. This is the best it gets. Until about 10,000 years ago, we were in a period of ice for about 90,000 years. We tend to have these long periods of ice of varying severity. We have a period of warmth. Then it gradually gets colder, more and more ice-and more and more ice for 90,000 years or so. And then that suddenly comes to an end, characterized by these sudden terminations when the ice melts with remarkable rapidity. We believe the sea level rises by several hundred feet in a period of maybe 10,000 years as the ice melts. And then we have a warm period.

"What's truly remarkable," he adds, booting up the relevant data on a laptop, "is that for the last 10,000 years we've had an extremely mild and unusually constant climate. That last 10,000 years of beautiful climate, which happens to coincide with the development of civilization, has inspired a lot of people to think about the relationship of climate and civilization. And what happens at the end of this period? We would expect things to begin cooling off right about now and heading into another ice age. If things behave the way they have in the past, that's what we can expect."

Gradual as the descent may be, the idea that we should soon start heading into another ice age makes it seem all the more important that we understand what causes this longstanding temperature cycle. And if Muller is correct, the basis for most of the studies in this area for the past several decades is wrong. That basis is the Milankovitch theory.

"Actually," Muller says with a slightly impish smile, "the theory was derived in the late 1800s by James Croll. It was revived in the 1920s by Milutin Milankovitch, who popularized it and got his name associated with it. He made some additions to it, most of which I believe are mistaken. But in the late 1950s, early 1960s, a great deal of data was accumulated that suggested that changes in the Earth's orbit were related to the ice ages, which increased the prestige of the Milankovitch theory."

The cause of the ice ages, according to Milankovitch, is a matter of "insolation," the amount of sunlight absorbed by the Earth, particularly in the northern hemisphere. "That depends," Muller explains, "on the Earth's orbit. It particularly depends on whether the northern hemisphere is pointed toward the sun at the point in the elliptical orbit when the Earth is close to the sun or when it's far away. So the Milankovitch theory-or the Croll theory- calculates the climate of the Earth based on being able to calculate back in time the tilt of the Earth's spin, the daily spin, and how far it was from the sun, to derive how much sunlight hit the northern hemisphere during summer. We can do these calculations-determine where the Earth was many millions of years ago-for the same reason we can calculate where our satellites will go when we send them off to Jupiter, with enormous precision-because the solar system has zero friction. So these variations were calculated and they seemed to have roughly a 100,000-year cycle-which was believed to be responsible for the 100,000-year cycle of the ice ages."

Though increasingly accepted as the dominant theory in the field, and the basis for most of the research over the past two or three decades, Milankovitch has always had its share of critics and skeptics. Among them was geologist Gordon MacDonald, whom Muller credits with alerting him to some of its problems. "Many of us thought there was a problem," Muller explains, "because there is evidence that the warming period, based on Milankovitch, begins a few thousand years after the ice age is ending. It should be the other way around. But it turns out to be very difficult to get precise data. I would say that a substantial number of geologists and geophysicists felt that these problems were minor and would go away with further data."

Other problems, Muller felt, were less easily explained away. "If you did the calculations for how much the sunlight changed," he says, "although it had a 100,000-year cycle, the amount was so small it was hard to see how it could account for the ice ages. That was the primary problem. In addition, the same calculations would say that not only was there a 100,000-year cycle, there should also be a 400,000-year cycle of much more intense cold, and that wasn't observed."

Muller began searching for other possible explanations that might better fit the data. And one of the first things that occurred to him was the accretion of cosmic dust-"material, junk, garbage from space" -an idea he doesn't hesitate to attribute to the pioneering work of Luis Alvarez and the comet that killed the dinosaurs, not to mention his own subsequent Nemesis theory. "Lute Alvarez's work triggered a whole lot of interest, including the whole issue of nuclear winter, the question of global warming, of what happens with volcanic eruptions. The cosmic dust undoubtedly blew out of my work on Nemesis."

In person, as in his book Nemesis, Muller is quick to pay tribute to Alvarez almost every chance he gets. Indeed, much of the book is a chronicle of how the late Nobel Laureate inspired him and served as a mentor, first as the Alvarezes developed their dinosaur extinction theory, then as Muller worked out the Nemesis theory. "What it shows," Muller says, "is Luie Alvarez coming up with ten bad theories for the one that finally works. But he knocks out each one of those ten until he finds one he can't knock out. That's the story of science." Or, as one of Muller's favorite quotes has it, "Research is the process of going up alleys to see if they are blind." It's a process Muller has wanted to be a part of ever since he was a child.

"I was born in the deep South-Bronx," he says with the kind of pause that tells you he's used this line many times before. "I call it the deep South Bronx because when Carter and Reagan went to the South Bronx it was a mile north of where I lived. They didn't venture that far south."

Muller went on to Bronx High School of Science and then to Columbia. "In my junior year, three of my friends and I took a long drive around the country. We'd never seen it before. And when I saw Berkeley I fell in love with it-nice people, nice weather, and a great university. New York City has great universities, too, but it fails on the other criteria." So in the fall of 1964, Muller arrived at UC for graduate school, intent on a career in nuclear physics, became involved in the Free Speech Movement, and was arrested. "I believe I'm the only student arrested in the FSM who later became a tenured professor here," he says. He's almost definitely the only one who went on to work on national security issues, as in his arms control work. "I never thought I'd be able to do that," he admits. "I believed that my involvement in the FSM would make that completely impossible. It probably would've excluded me if Berkeley hadn't been followed by a decade in which every university in the country experienced similar events over the Vietnam War. Pretty soon the US government realized that if they wanted to exclude everybody with an arrest record, there wouldn't be many people left they wanted. So it ended up having very little effect."

It did have a greater effect on another aspect of Muller's life, his thirty-year (so far) marriage to architect Rosemary Muller. "She says she wouldn't have looked at me twice if I hadn't been arrested in the Free Speech Movement." The Mullers have two daughters, one at Berkeley High, the other at UC San Diego. They also have another joint venture in their past, the Inn Season restaurant, which they conceived on a backpacking trip in the Sierra and started in 1976.

"That was the craziest thing I ever did," he says. "I was still a research scientist, not yet a faculty member. My wife did the architecture. We ran it for six years and never really made money. By the way, I now have an informal consulting service. Anybody who wants to open up a restaurant, I will talk them out of it. I'm very good at it. I can tell you precisely what's going to go wrong and why it'll never make a profit. The problem is that you're competing with other restaurants that are losing money and don't know it. They don't know it because they don't have to pay their bills this month. They get the income from the customers immediately, but they don't have to pay their bills until next month. As long as the business is growing, they're covered, even if they're losing money on every meal. This goes on for a year, until business stops growing, and suddenly they can't pay their bills. At least we knew we were losing money, but we had other jobs."

But, that, as they say, is another story.

It was a combination of MacDonald's skepticism about the Milankovitch theory and an article in the New York Times that turned Muller's attention toward the problem of the ice ages. The Times article reported new data derived from the Greenland ice sheet. "The layers of the ice," he explains, "are like tree rings. You can go back and see what kinds of materials were trapped, what gases, and there are ways of estimating what the temperature was. And the truly remarkable thing is that, during the last ice age, not only did we have cold weather, but it was constantly and rapidly changing: cold/warm; cold/warm. It was flickering, that's the technical term: cold with brief periods of warm-that's brief in geologic terms, maybe a few thousand years, maybe less than a thousand.

"It was the flickering that made me think of a mechanism that people hadn't considered before-that not just the changing sunlight but maybe debris from space could be coming in and causing that kind of flickering. We have meteors. We have dust. We have all that other junk up there in space. I had an idea that the changing orbit of the Earth could bring us into contact with streams of meteors and dust that we're not in contact with at other times. I knew it had to be an orbital change because it comes so regularly.

"The first thing I checked was to see if anybody else had examined this theory. I spoke to Gordon MacDonald, who's something of a world expert in the field, and he said he was quite sure no one had ever explored this idea before. Well, that's always quite a shock. It's very hard to find ideas in science that someone hasn't investigated unless there's a very good reason. In this case, I believe the reason was the complete dominance of the Milankovitch theory. My first idea was to look at the eccentricity of the orbit. When the orbit is elongated, we could run into some material that we wouldn't run into when it was more circular. Maybe this would explain the inaccuracies in the insolation model. Because the eccentricity was a factor driving the cycle, but it was causing the ice ages through accretion.

"The next step was realizing that if changes in the eccentricity of the Earth's orbit could effect the amount of accretion, then so could changes in the inclination of the orbit-the wobble. That's the Earth's deviation from what we call the 'Invariable Plane,' the plane of the orbits of Jupiter and Saturn, the plane of the asteroids. The amount we're tilted out of that plane changes with time, and this oscillation had been analyzed, but it turned out that nobody had calculated the frequency of the oscillation. So I felt I had to do that calculation. Suppose the cycle of the inclination turned out to be 50,000 years? Suppose I looked at the data and there is no 50,000-year cycle? Then that's very strong evidence against my theory. And the standard in science is, you don't publish your theory and allow somebody else to point out your mistakes. You have to find all your mistakes yourself, and if there's strong evidence against your theory, you either don't publish or publish while stating that it has this major problem.

"So I did the calculation. And when I got the number, it came as a big surprise. It was 100,000 years."

The 100,000-year cycle of the inclination of the Earth's orbit matched the 100,000-year cycle of the ice ages. But so did the 100,000-year cycle of the eccentricity of the orbit. So Muller now had two possible explanations. Either the Earth's orbit at its most elongated ellipse was bringing us into contact with a band of space dust, or our orbit entered a band of space dust when it passed through the Invariable Plane (the plane of the asteroids), and only escaped the ice age-causing accretion when its orbit wobbled furthest out of that plane. "For the first time," he notes, "we had two viable sources for the 100,000-year cycle-both astronomical-and they were different."

The next step was to do a spectrum analysis to determine how well the frequencies in these three 100,000-year cycles actually matched each other. "This is the point at which I solicited help from Gordon," Muller says, "because he's far more experienced in spectrum analysis than I am. So we worked together on the spectrum analysis, and that's when we got the next shock. Let me show you what I call the spectral fingerprint."

A few commands later, Muller has called up a series of charts on his laptop, juxtaposing them against each other. "If you just look at this data by eye, you see these 100,000-year cycles. But what we want to do is look at them more precisely, to see if they stay precisely in time with each other over the last one million years." The ice age data shows a sharp, high peak, indicating a consistent 100,000-year cycle. The data from Muller's calculations on orbital inclination is called up. It shows an almost perfect match.

"Now what about the old Milankovitch theory?" Muller says, running it through the same spectrum analysis. "Now this came as a shock to me." The Milankovitch eccentricity data comes to what can only be described as a double-peak, a bifurcated point. Not only that, it's produced a second peak off to one side. "It doesn't match," Muller proclaims. "Why did they think it ever did? It's taken me a little bit of detective work to find out what the hell is going on here. The answer is, if you don't want to see a mismatch, there's a way of making them match each other. Let me do that now."

Muller programs another set of statistical analysis figures and calls up the results. "There it is. It's actually the same spectrum but it's just blurred out. It averages things together. It's a way of doing a calculation when you think that the data wasn't very accurate, for instance, if you thought the time scale wasn't accurate. Now let's do the same thing with the ice age data." The two graphs now look pretty much the same, but blurred-so blurred that the double peak of the Milankovitch theory has become an apparent single peak.

"They match, right? What about this other peak that isn't in the ice age data? Well, there must be something in the Earth that suppresses that peak. That's the 400,000-year peak. We don't understand why it's not there, but there's a list of plausible explanations, one for each PhD thesis that's been written in the field for the last twenty years. Of course, what we've done is like taking a fingerprint and smearing it around a bit. Then we compare fingerprints.

"That's why nobody's ever noticed this discrepancy between the narrow peak and this double line before. Or nobody's ever mentioned it in the literature. If you look at the papers, 95 percent of them use the smeared spectrum. I talked to one scientist who had read my paper who said, 'Y'know, I had noticed that discrepancy, and I didn't know what to make of it, so I went back to doing the old way of analysis.' Basically saying, 'I didn't understand it, so I ignored it.' Which is not that uncommon among scientists. But Gordon and I had an advantage coming into this that nobody else had before, which was that we had an alternative theory that predicted a single narrow line. So when we saw the contradiction, we had an alternative sitting in our pocket. We didn't have to ignore it."

It sounds like a scientific success story. But this kind of success carries within it the seed of its own problems. New ideas are particularly hard to accept when they overthrow the fundamental assumption on which most of the work in the field is based. As Muller, who has a nice way with an aphorism, wrote in Nemesis, "Old theories never die, only old theorists."

So far, Muller and MacDonald have been unable to get their full paper, detailing their work, published, despite their considerable credentials. It's been rejected by Science. It's been rejected by Nature three times-the third time as recently as June-though the editors did request, and publish, a shorter version summarizing their findings last November. Why? Muller pulls open a ]long file drawer, crammed with papers. "Here it is. Essentially everything that's been published for the last twenty years assumes the Milankovitch model. I think it's very hard for people in this field, and all the referees to whom our paper has been submitted are working in this field, to accept our paper. They'd have to say that most of their own work for the past twenty years is fundamentally flawed."

But if the paper remains unpublished, the idea is already being widely disseminated. For one thing, the full text is available on Muller's Web site (http://muller. Ibl.gov). For another, it's received significant support in the form of a study by Kenneth Farley at Caltech, who heard about the theory through Walter Alvarez. "Farley, testing our theory, looked for cycles of extraterrestrial dust in the ancient climate records," Muller says. "He was working with sea floor cores, measuring helium-3, a light isotope of helium which is particularly abundant in cosmic dust. He discovered a sudden increase in the dust a million years ago. Walter Alvarez thought that fit nicely with the work we had done, so he asked Farley whether he could analyze the dust and see whether he found a 100,000-year cycle. He didn't think he would, but scientists love to disprove theories, so he did the experiment. And to his surprise, he found that the extraterrestrial dust goes through 100,000-year cycles."

Farley submitted his findings to Nature, which published them in December -the apparent reason why the publication asked Muller and MacDonald for the abridged paper published the previous month. Science writer Charles Petit picked up on that paper and wrote a story published in the San Francisco Chronicle about this work in January. Wallace Broecker of Columbia's Lamont-Doherty Earth Observatory, one of the leading experts in the field-and a longtime Milankovitch skeptic-visited Muller at Berkeley, intent on challenging his results, and ended up inviting him to give a series of lectures on the theory at Lamont-Doherty this spring. That series, attended by geophysicists and paleoclimatologists from all over the country, led to an extended e-mail debate between Muller and John Imbrie, one of the most prestigious Milankovitch supporters.

"That was thrilling, actually," Muller observes. "Every time he emailed a question or an answer to me, he sent copies to 45 other prominent people in the field. And then I'd send my answer back-often after taking a half-hour off to do the calculation-with 45 people listening in. All on email. I've never seen anything like this before. The way this would typically happen in the past is Einstein would write a letter to Bohr and Bohr would write back, and they'd have this correspondence over two or three years. And then finally they'd get together in an international conference with all the top people in the field in the middle of a great discussion. And then everybody would go home. But this was all taking place in cyberspace. And it was better because we could take a half-hour to do a calculation before responding."

The Imbrie-Muller debate remains unresolved, even after Muller flew to Brown University-at Imbrie's invitation-for a concentrated, detailed discussion in June. Meanwhile, Muller says he can afford to wait for more evidence to accumulate and the scientific community to come around. "This is the reason I have tenure," he says, "so it can take me three years to publish a paper. Most of what we're doing is on the record. Everybody in the field has read the paper. Nobody's going to steal the idea."

The main thing he's concerned about is the key matter of understanding how the climate works. "What causes climate? What affects it? To what extent do we affect climate? To what extent are we disturbing it? These are important issues. I tend to be very passionate about this. When I think about it, I say: What is the effect if the temperature were to drop one or two degrees? The answer that I get is that areas which are good for agriculture will probably shift no more than three hundred miles. But when that has happened in the past, it has led to war.

"When people talk about global warming, we're talking about the great breadbasket becoming a desert, and perhaps the productive regions shifting north. So yes, it might be nicer if it were a little warmer in Berkeley during the summer, but that's not what we're talking about. And yes, if we had a global government and everybody loved everybody else and people were happy to migrate across national boundaries and could do so freely.... But I fear that changes in climate have been a major-if not the major-cause of war in the past. Climate change is potentially the biggest disaster that could occur. And if humans have taken over and are now the dominant driving force for climate- which I think might be the case-than we need to understand how it works."