Travel back to a time 10,000 years before the present to a place yet to have a name. Latitude 41 degrees north, longitude 74 degrees west -- one day, New York City will stand on this place. But on this cold Pleistocene morning, the landscape is absent the slightest premonition of this future, minus even the Hudson River.

Sighting along all points of the compass, the view to the horizon is the same. The Earth is smothered beneath a great plain of glacial ice. The ice is as endless as a continent, as deep as a solid sea. For the prior million years, glaciers have come and gone across the face of the Earth. Ten times they have halted their spread, briefly retreating for up to 10,000 years. And each time, the glaciers have returned, relentlessly advancing for some 90,000 years, inexorably tripling their domain, devouring 30 percent of the land on the planet.

On this morning -- at a place absent forests, trees, or one single green weed -- the glaciers again are in motion. Sounding as they retreat, the glaciers drip, melt, and fracture, projecting a thunderous crack that rumbles over the landscape. Once again, their time is about to pass.

Now the year is 1993, the late Holocene. Less than two centuries have lapsed since humans discovered that ice once had covered much more of their planet. By the mid 20th Century, humans know that they live in a time of rare warmth, that these glacial epochs have recurred repeatedly. As is their nature, people wonder why. Why do Ice Ages occur? Why over the past million years have glacial and interglacial epochs alternated with such uncanny regularity, like a pendulum swinging back and forth once every 100,000 years?

In 1864, James Croll, a janitor and later a fellow at the Royal Society of London, was among the first to provide a theory to explain Ice Ages. By the turn of the century, his theory was all but forgotten. But in 1912, Yugoslav geologist Milutin Milankovitch revised Croll's idea, called it his own, and published a book, "Canon of Insolation." The book created a sensation, inspiring comic book covers of giant glaciers looming over the skyscrapers of New York City. Decades passed, and Milankovitch's theory of the glacial cycles became the textbook explanation.

By 1993, Berkeley Lab scientist Rich Muller has his doubts. An eclectic thinker with the courage to challenge conventional scientific wisdom, Muller suspects the textbooks are wrong. The "Canon of Insolation" relies on an astronomical cycle that causes the amount of sunlight on Earth to vary. According to Muller's friend, geophysicist Gordon MacDonald (now with Austria's International Institute for Applied Systems Analysis), the effect described by Milankovitch is too weak. Muller's curiosity is aroused. He comes up with an alternate explanation, a theory that owes a debt to the demise of the dinosaurs, and to his mentor, Berkeley Lab's Luis Alvarez. Almost two decades previous, Alvarez and his son, geologist Walter Alvarez, first proposed that the dinosaurs had been killed off by a massive comet or asteroid that had slammed into the Earth. In the years since, Muller has investigated the causes of periodic mass extinctions on Earth and he has learned a great deal about cosmic debris. Could cosmic debris and dust play a role in the cadence of the glaciers?

Muller begins writing a paper to explore an idea that picks up where Milankovitch left off. Milankovitch attributed the rate of glaciation to an astronomical rhythm that he believed to be a match. "Eccentricity" -- a very gradual change in the Earth's egg-shaped orbit around the sun that completes a cycle at the pace of 100,000 years -- that accounts for the Pleistocene glaciations, claimed Milan- kovitch. Eccentricity indeed does exist. But, maintains MacDonald, it does not alter the sunlight on Earth that dramatically. Eccentricity alone won't make an Ice Age.

Muller hypothesizes that eccentricity somehow changes the rate at which cosmic dust falls into the upper atmosphere, thus altering climate. As he works on the paper, Muller realizes that the Earth has yet another orbital cycle that likewise could alter dust flows. He wonders if a change in the Earth's orbital plane -- a phenomena known since the 1600s -- might somehow be involved. Though this oscillation has been described since the time of Copernicus, no one has ever calculated the time length of this cycle. Mainly to exclude the idea, Muller cranks up his desktop computer. Using data provided by MacDonald, he does the calculations. To his utter amazement, the cycle is 100,000 years, a match to the march of the glaciers.

In the ensuing months, MacDonald joins Muller, and together they examine the two different orbital cycles, comparing both to the details of the geological record. Using a geological fingerprinting technique, they find that the climatic changes recorded in the rocks matches their theory, but not that of Milan- kovitch.

Eccentricity changes the Earth's average annual distance from the sun and slightly alters the amount of sunlight hitting our planet. Milankovitch seized on this as the cause for the recent deep freezes on Earth. MacDonald, however, says the theory has a fatal problem. The temperature variation due to eccentricity is about 50 times less than that necessary to cool the planet into a glacial epoch.

To visualize the cycle that Muller and MacDonald say precisely matches that of the climatic record, imagine a flat plane with the sun in the center and nine planets circling close to the plane. In fact, all the planets orbit the sun near such a fixed orbital plane. Fixed, that is, except for the wobble. The Earth's orbit slowly tilts out of this plane and then returns. As Muller first calculated, the cycle of tilt repeats every 100,000 years.

Muller and MacDonald write their paper outlining what amounts to a maverick theory for the glaciations. Challenging Milankovitch, whose theory forms the foundation for a substantial edifice of subsequent research, they run into a brick wall of scientific dogma. Their paper is rejected again and again. In 1995, Nature publishes an abridged article in which they link glacial cycles and orbital tilt. The article contends that changes in tilt alter the influx of extraterrestrial debris. If cosmic dust indeed accounts for the glacial cycles, then the geologic record should contain signs of an overlapping dust cycle. Look at samples of sedimentary rock, they urge.

Kenneth Farley of Caltech, performing research in an altogether different field, happened to be doing analysis of core material from the sea floor. Farley detected a sudden increase in space dust a million years ago. Learning of the new theory on glaciation, he decided to analyze the record for the past million years. To his surprise, Farley discovered that the extraterrestrial dust has accumulated in 100,000-year cycles.

Chalk one up to Muller's side, but hold the champagne. As he explained during a January 1996 interview, "The problem is that we don't know of the existence of an actual cloud of dust or meteors that would account for the data." Muller began thinking about how scientists might detect a possible ring of dust around the sun. Later that year, he learned that he had been wasting his time. In 1988, NASA's IRAS satellite had detected a ring of dust. It lies along the orbital plane of the planets.

The Earth continues to race through its paths in the heavens, and now it is 1997. At the dawn of a new millennium, a global debate heats up over the extent to which humans have altered the atmosphere in ways that could influence climate. The natural climate cycles that have ruled the Earth for the past million years continue, running at a pace still imperceptible to modern science. None of the present climate models include the effects of extraterrestial dust and debris. And yet, Muller and MacDonald's data, published in Science magazine in July 1997, suggest that debris has played the dominant role in the climate for the last million years.

Muller continues to investigate why these natural climate cycles repeat. Relatively little of the dust in the atmosphere comes from space. The scientific community is in agreement about that. Why then would this 100,000 year gentle rain of cosmic dust trigger glaciation? Could the cooling be the result of some kind of chemical domino effect in the atmosphere? And so Muller takes up yet another scientific discipline, atmospheric chemistry. Despite the present lack of an understood trigger mechanism, he well may have identified the dominant factor that will determine the climate of the upcoming millennia. He is convinced he has found the engine that powered the Earth's climate for the past million years. The clock on Muller's cycle shows the Earth at the end of its latest 10,000 year respite of warmth. On this clock, it is time for the return of the glaciers.