When human beings arrived in North America about 30,000 years ago, they brought language, fire, flint tools and skin tents. They also brought something they didn't know they had--a tiny microbe called JC virus sequestered in their kidneys.

Their descendants--or most of them, at least--long ago shucked off Stone Age technology. But JC virus is still with them.

Researchers are now using the virus to bolster the hypothesis that American Indians are the distant offspring of those migrants, and that they came from eastern Asia on a land bridge across the Bering Strait.

The strain of JC carried by modern-day Navajos, as it happens, is nearly identical to that carried by modern residents of Tokyo. It's somewhat different from the virus carried by the Chamorro, the aboriginal inhabitants of the South Pacific island of Guam. It's very different, however, from the virus carried by West Africans, East Africans and Europeans.

All in all, the evidence suggests the Navajo and the Japanese are closely related to each other, and are related (but more distantly) to the Chamorro. All are descended from the same prehistoric population of eastern Asians.

None of these insights is especially new. Anthropologists reached them years ago from studies of human genetics and the archaeological record. What's new is to have them confirmed by a virus.

JC virus is something that seems unimaginable--a microbe that travels so intimately, and so ubiquitously, with human beings that it's essentially indistinguishable from the people themselves.

Nearly everyone on Earth acquires JC virus as a child, usually from a parent, less often from someone in the immediate neighborhood. It rarely causes disease. Most important, its genes are astoundingly resistant to mutation, a distinctly uncommon trait for a virus.

JC was first isolated from a human being in 1971. (The name, in fact, is the person's initials.) Only in the last few years, however, have scientists realized they can differentiate human populations from one another by examining the strains of virus they carry, and from that deducing their movement over thousands of years.

"One would have thought that the distinctions [between viruses] would have been hopelessly homogenized," said Gerald L. Stoner, a neurotoxicologist at the National Institutes of Health and a leading expert on the "ethnicity" of JC viral strains. "The surprising thing is that they have not. No one would have ever predicted that they would characterize populations like they do."

Anthropologists have studied biological "markers" for ethnic groups for a long time. Blood types and enzyme variants were the first to be used, but they provided only crude information about the biological underpinnings of ethnicity. With the advent of new technologies, researchers turned to human DNA, the coin of human inheritance, for determining the relatedness of human populations.

The JC virus now joins human DNA as another tool, not for tracing families but for tracing communities.

The theory behind using DNA to deduce human migration is this: A gene can be described precisely by naming the units of DNA and their order. Slight differences, like the similarities, are inherited. Thus, they tend to be shared by populations that tend to breed among themselves, which is the case with most populations on Earth.

When a few individuals leave a population and go somewhere new, they take with them their DNA, and all its variations. For a while, these migrants are genetically indistinguishable from the people they left. Over time, however, new variations creep into both their DNA and that of the "ancestral" population they left behind. Because the variations arise by chance--specifically, by random mutation--it's very unlikely that the same ones will creep into the DNA of both groups.

By studying the type and number of DNA variations that stud a common genetic background, scientists can estimate how closely or distantly two (or more) groups are related. By looking where different genetic "types" occur on the map, they can infer patterns of migration.

JC virus is useful for such comparisons because, like DNA, it's resistant to "recombination" events, in which paired strands of similar DNA trade pieces of themselves. The JC genome does this very rarely because its DNA forms a closed ring that is difficult to open, a prerequisite for recombination.

Consequently, an ethnic group tends to carry a single type of JC virus, which records all the mutations that have accrued over millennia, none having been traded away.

JC virus has other, less benign, characteristics as well. Although they have nothing to do with the microbe's newfound usefulness for anthropology, they're the main reason why this otherwise obscure virus is known to biologists.

In certain people with damaged immune systems, the virus loses DNA in a key regulatory region of its genome and becomes a fatal pathogen of the brain. What triggers this transformation is almost entirely unknown. The disease that results, progressive multifocal leukoencephalopathy (PML), is similar to multiple sclerosis, but much more rapid, killing patients in about six months. Once rare, PML has emerged with the AIDS epidemic, and now affects about 7 percent of people with that disease.

Stoner and a former post-doctoral fellow at the National Institute of Neurological Disorders and Stroke, Hansjurgen T. Agostini, believe JC virus may have been carried by humans since the dawn of the species. They have identified seven major types, and nearly two dozen subtypes, distributed around the world.

As with most genetic markers, JC virus's ethnographic usefulness is proving more complicated than originally envisioned.

For example, in a study published last year, the African scientist Sylvester C. Chima found that Pygmies carry three types of JC--one of which may have evolved when they lived in isolation in the Ituri rain forest south of Sudan, and another donated by the Bantu to the west, with whom they've had extensive contact for 2,000 years. So "one-population, one-virus" is clearly an oversimplification.

Nevertheless, anthropologists hope the microbe will shed light on such ethnographic uncertainties as the origins of the Basque people of Spain, or the events that led to the peopling of Polynesia. (Recent research on the latter by Jonathan Friedlaender, an anthropologist at Temple University, supports the idea that much of the region was settled quickly and recently by Southeast Asians, rather than slowly by indigenous Melanesians.)

JC virus is likely to become a hot new tool of physical anthropologists. Its genome is smaller than that of human DNA, and somewhat easier to work with. It's also extracted from urine samples, not blood, and consequently can be collected with relative ease.

Whether it proves truly informative, or only a curious example of the intimate bond between men and microbes, remains to be seen.

CAPTION: TRACKING HISTORY WITH A VIRUS (This graphic was not available)