Magnified 25,000 times under Drexel University's scanning electron microscope, a couple of flecks of dirt offer up a landscape full of crags, valleys, ridges - and, to Dee Breger's eyes, a window back in time.

The tiny grains came from the sea floor below the Gulf of Carpenteria in northern Australia, part of an underground layer dating to the first millennium. Breger and her colleagues believe the material holds signs that a fragment of a comet crashed to Earth during that period. Such an event might explain the months of cold summers and dark days that began in A.D. 536 and led to a well-documented period of famine and unrest.

And, they say, while such an event would have been catastrophic, it was not unique. By comparing the historical and archaeological records with hard-to-prove physical evidence, they are trying to make a case that rocks from space were responsible for altering human affairs in ways so huge that some have entered mythology.

It is an uphill battle.

"We're mavericks," says Breger, a microscopist who is not formally trained in science. Scanning her dirt sample on a nearby screen, she zooms in on what looks like a splotch of paint. "We call that a splat."

Breger instructs the machine to analyze the composition. Traces of some metals in the form of a splat can be a sign of a powerful explosion, she says - one you might get if a piece of a comet or asteroid slammed into the Earth.

Scanning further, she stops at a sphere, less than a hundredth of a millimeter across (much smaller than the width of a human hair). Under the electron microscope it resembles a planet or some exotic moon, the surface scarred with rifts and cracks, all suggestive of molten rock or metal that was blasted into the air and quickly cooled.

It was Dallas Abbott, a marine geophysicist at Lamont-Doherty Earth Observatory in New York, who brought these samples to Drexel. Breger teamed up with Abbott while working at Lamont-Doherty, first as a scientific illustrator and then as a microscopist. The former art student moved to Drexel four years ago for the chance to work with more powerful instruments.

"The lab here is state-of-the-art," she says.

They and a few colleagues in the Holocene Impact Working Group - named for the period covering the last 20,000 years - have been proposing for years that several large objects from space hit the Earth with enough force to influence global climate within human history. Abbott estimates this happened perhaps five times in the last 6,000 years.

Most sudden climate changes over the eons remain unexplained, and most scientists argue that a lack of convincing evidence for any theory is not much of a reason to support this one.

"Impacts are the solution of choice when you don't have any data," says astronomer Donald Yeomans of NASA's Jet Propulsion Laboratory.

Still, astronomers generally agree that the Earth has been smacked around by comets and asteroids, and that some altered the history of life. The most famous of them fell around 65 million years ago - kicking up enough debris, it now appears, to cool the planet and kill off the dinosaurs.

Much more recently, for 18 months around A.D. 536, a thick haze and freakish cold gripped Europe. As Byzantine historian Procopius of Caesarea put it: ". . . the sun gave forth its light without brightness . . . and it seemed exceedingly like the sun in eclipse, for the beams it shed were not clear."

Tree growth rings dated to that time appear narrow and sickly as far away as North America and Central Asia, an indication that the cold and darkness spanned the globe. The period coincided with devastating crop failures that some have linked to the fall of the Roman Empire.

There definitely was a climate anomaly at that time, says Richard Alley, a climate expert and glaciologist at Penn State University. But he, like most in the field, favors an alternative explanation: that a large volcanic eruption dimmed the skies. Not only are eruptions relatively common compared with large impacts, he says, there now appears to be a record stored in ice caps near Greenland and Antarctica.

Scientists drilling out deep cores can date ice laid down in that era to within a year or two, Alley says, and evidence of telltale volcanic sulfates in layers from A.D. 536 was published just last month.

Looking more broadly, Abbott and Breger are also seeking clues to a possible impact 4,800 years ago. Around that time, many different cultures advanced myths of catastrophic floods, Abbott says, including the biblical account of Noah's ark.

To find out what happened, she's gathering clues on scales small and large. With satellite images now widely available through Google Earth, Abbott is examining massive, V-shaped formations in northern Australia and, at the other end of the Pacific, in Madagascar, the island off East Africa. She argues that these were created by a giant impact in the Indian Ocean that sent a mega tsunami in different directions.

"The one in Australia rises more than 100 feet above sea level," Abbott says. Her critics contend these are just big sand dunes created by prevailing winds, an explanation that she says doesn't go nearly far enough.

And the actual point of impact?

With remote sensing technology, Abbott says she's found signs that might indicate the presence of an 18-mile-wide crater. But it's more than 1,000 feet below the surface and hard to confirm. She's seeking funding for a more thorough exploration.

If they put together enough lines of evidence for enough separate events, Abbott says, the work may back up an idea promoted by a minority of British astronomers that a few thousand years ago a massive comet swung inward from the fringes of the solar system, broke up near the Earth - and has been periodically dropping pieces on the planet ever since.

What's left, these astronomers proposed, is now called Comet Encke - a tiny chunk of ice and dirt that orbits the sun every three years.

Others say that while the idea that pieces of what is now Encke fall to earth is plausible, the evidence is unconvincing. "It's a little dinky comet right now," says David Morrison, a planetary scientist at NASA's Ames Research Center in California. "If Encke was responsible for something, it's aged quickly."

By tracking asteroids that cross Earth's orbit and observing craters on the moon's surface, astronomers are able to estimate how frequently they hit in the past. Potentially climate-changing impacts with asteroids average less than once in a million years, they say, and comet collisions even less often.

That doesn't mean there weren't recent major comet impacts, Morrison says. It's just very unlikely.

Proving that a given geological formation was caused by an impact is not a simple matter even when it's on land, says Jay Melosh, a planetary scientist at the University of Arizona's Lunar and Planetary Lab.

It's even harder under water. In that case, one of the best indicators is the concentration of rare elements, especially iridium, that would have come with the comet or asteroid and then blasted through the atmosphere on impact.

And it's one thing to find evidence of an impact, Melosh says, and another to demonstrate that it was big enough to lead to global climate change.

The only clear-cut case is the impact that marked the end of the dinosaur era 65 million years ago. Underwater remote sensing has found a massive crater under the Gulf of Mexico, and geologists have discovered more than 200 places around the globe where rock samples have yielded unusually high levels of iridium in the same layer of geological strata after which dinosaurs suddenly vanish.

Mainstream astronomers say asteroids and comets do hit often enough that we should worry about them. Those optimistic enough to think we'll survive more than a few thousand years say a rock or comet will eventually get us. Unless we can figure out how to deflect it, we will go the way of the dinosaurs.