What you actually see in the fossil record, Thorne insists, is a smooth, continuous transition. “There’s a famous skull from Petralona in Greece, dating from about 300,000 years ago, that has been a matter of contention among traditionalists because it seems in some waysHomo erectus but in other waysHomo sapiens . Well, what we say is that this is just what you would expect to find in species that were evolving rather than being displaced.”

One thing that would help to resolve matters would be evidence of interbreeding, but that is not at all easy to prove, or disprove, from fossils. In 1999, archeologists in Portugal found the skeleton of a child about four years old that died 24,500 years ago. The skeleton was modern overall, but with certain archaic, possibly Neandertal, characteristics: unusually sturdy leg bones, teeth bearing a distinctive “shoveling” pattern, and (though not everyone agrees on it) an indentation at the back of the skull called a suprainiac fossa, a feature exclusive to Neandertals. Erik Trinkaus of Washington University in St. Louis, the leading authority on Neandertals, announced the child to be a hybrid: proof that modern humans and Neandertals interbred. Others, however, were troubled that the Neandertal and modern features weren’t more blended. As one critic put it: “If you look at a mule, you don’t have the front end looking like a donkey and the back end looking like a horse.”

Ian Tattersall declared it to be nothing more than “a chunky modern child.” He accepts that there may well have been some “hanky-panky” between Neandertals and moderns, but doesn’t believe it could have resulted in reproductively successful offspring.49“I don’t know of any two organisms from any realm of biology that are that different and still in the same species,” he says.

With the fossil record so unhelpful, scientists have turned increasingly to genetic studies, in particular the part known as mitochondrial DNA. Mitochondrial DNA was only discovered in 1964, but by the 1980s some ingenious souls at the University of California at Berkeley had realized that it has two features that lend it a particular convenience as a kind of molecular clock: it is passed on only through the female line, so it doesn’t become scrambled with paternal DNA with each new generation, and it mutates about twenty times faster than normal nuclear DNA, making it easier to detect and follow genetic patterns over time. By tracking the rates of mutation they could work out the genetic history and relationships of whole groups of people.

In 1987, the Berkeley team, led by the late Allan Wilson, did an analysis of mitochondrial DNA from 147 individuals and declared that the rise of anatomically modern humans occurred in Africa within the last 140,000 years and that “all present-day humans are descended from that population.” It was a serious blow to the multiregionalists. But then people began to look a little more closely at the data. One of the most extraordinary points—almost too extraordinary to credit really—was that the “Africans” used in the study were actually African-Americans, whose genes had obviously been subjected to considerable mediation in the past few hundred years. Doubts also soon emerged about the assumed rates of mutations.

By 1992, the study was largely discredited. But the techniques of genetic analysis continued to be refined, and in 1997 scientists from the University of Munich managed to extract and analyze some DNA from the arm bone of the original Neandertal man, and this time the evidence stood up. The Munich study found that the Neandertal DNA was unlike any DNA found on Earth now, strongly indicating that there was no genetic connection between Neandertals and modern humans. Now this reallywas a blow to multiregionalism.

Then in late 2000Nature and other publications reported on a Swedish study of the mitochondrial DNA of fifty-three people, which suggested that all modern humans emerged from Africa within the past 100,000 years and came from a breeding stock of no more than 10,000 individuals. Soon afterward, Eric Lander, director of the Whitehead Institute/Massachusetts Institute of Technology Center for Genome Research, announced that modern Europeans, and perhaps people farther afield, are descended from “no more than a few hundred Africans who left their homeland as recently as 25,000 years ago.”

As we have noted elsewhere in the book, modern human beings show remarkably little genetic variability—“there’s more diversity in one social group of fifty-five chimps than in the entire human population,” as one authority has put it—and this would explain why. Because we are recently descended from a small founding population, there hasn’t been time enough or people enough to provide a source of great variability. It seemed a pretty severe blow to multiregionalism. “After this,” a Penn State academic told theWashington Post , “people won’t be too concerned about the multiregional theory, which has very little evidence.”

But all of this overlooked the more or less infinite capacity for surprise offered by the ancient Mungo people of western New South Wales. In early 2001, Thorne and his colleagues at the Australian National University reported that they had recovered DNA from the oldest of the Mungo specimens—now dated at 62,000 years—and that this DNA proved to be “genetically distinct.”

The Mungo Man, according to these findings, was anatomically modern—just like you and me—but carried an extinct genetic lineage. His mitochondrial DNA is no longer found in living humans, as it should be if, like all other modern people, he was descended from people who left Africa in the recent past.

“It turned everything upside down again,” says Thorne with undisguised delight.

Then other even more curious anomalies began to turn up. Rosalind Harding, a population geneticist at the Institute of Biological Anthropology in Oxford, while studying betaglobin genes in modern people, found two variants that are common among Asians and the indigenous people of Australia, but hardly exist in Africa. The variant genes, she is certain, arose more than 200,000 years ago not in Africa, but in east Asia—long before modernHomo sapiens reached the region. The only way to account for them is to say that ancestors of people now living in Asia included archaic hominids—Java Man and the like. Interestingly, this same variant gene—the Java Man gene, so to speak—turns up in modern populations in Oxfordshire.

Confused, I went to see Harding at the institute, which inhabits an old brick villa on Banbury Road in Oxford, in more or less the neighborhood where Bill Clinton spent his student days. Harding is a small and chirpy Australian, from Brisbane originally, with the rare knack for being amused and earnest at the same time.

“Don’t know,” she said at once, grinning, when I asked her how people in Oxfordshire harbored sequences of betaglobin that shouldn’t be there. “On the whole,” she went on more somberly, “the genetic record supports the out-of-Africa hypothesis. But then you find these anomalous clusters, which most geneticists prefer not to talk about. There’shuge amounts of information that would be available to us if only we could understand it, but we don’t yet. We’ve barely begun.” She refused to be drawn out on what the existence of Asian-origin genes in Oxfordshire tells us other than that the situation is clearly complicated. “All we can say at this stage is that it is very untidy and we don’t really know why.”

At the time of our meeting, in early 2002, another Oxford scientist named Bryan Sykes had just produced a popular book calledThe Seven Daughters of Eve in which, using studies of mitochondrial DNA, he had claimed to be able to trace nearly all living Europeans back to a founding population of just seven women—the daughters of Eve of the title—who lived between 10,000 and 45,000 years ago in the time known to science as the Paleolithic. To each of these women Sykes had given a name—Ursula, Xenia, Jasmine, and so on—and even a detailed personal history. (“Ursula was her mother’s second child. The first had been taken by a leopard when he was only two. . . .”)

When I asked Harding about the book, she smiled broadly but carefully, as if not quite certain where to go with her answer. “Well, I suppose you must give him some credit for helping to popularize a difficult subject,” she said and paused thoughtfully. “And there remains theremote possibility that he’s right.” She laughed, then went on more intently: “Data from any single gene cannot really tell you anything so definitive. If you follow the mitochondrial DNA backwards, it will take you to a certain place—to an Ursula or Tara or whatever. But if you take anyother bit of DNA, any gene at all, and traceitback, it will take you someplace else altogether.”

It was a little, I gathered, like following a road randomly out of London and finding that eventually it ends at John O’Groats, and concluding from this that anyone in London must therefore have come from the north of Scotland. Theymight have come from there, of course, but equally they could have arrived from any of hundreds of other places. In this sense, according to Harding, every gene is a different highway, and we have only barely begun to map the routes. “No single gene is ever going to tell you the whole story,” she said.

So genetic studies aren’t to be trusted?

“Oh you can trust the studies well enough, generally speaking. What you can’t trust are the sweeping conclusions that people often attach to them.”

She thinks out-of-Africa is “probably 95 percent correct,” but adds: “I think both sides have done a bit of a disservice to science by insisting that it must be one thing or the other. Things are likely to turn out to be not so straightforward as either camp would have you believe. The evidence is clearly starting to suggest that there were multiple migrations and dispersals in different parts of the world going in all kinds of directions and generally mixing up the gene pool. That’s never going to be easy to sort out.”

Just at this time, there were also a number of reports questioning the reliability of claims concerning the recovery of very ancient DNA. An academic writing inNature had noted how a paleontologist, asked by a colleague whether he thought an old skull was varnished or not, had licked its top and announced that it was. “In the process,” noted theNature article, “large amounts of modern human DNA would have been transferred to the skull,” rendering it useless for future study. I asked Harding about this. “Oh, it would almost certainly have been contaminated already,” she said. “Just handling a bone will contaminate it. Breathing on it will contaminate it. Most of the water in our labs will contaminate it. We are all swimming in foreign DNA. In order to get a reliably clean specimen you have to excavate it in sterile conditions and do the tests on it at the site. It is the trickiest thing in the world not to contaminate a specimen.”

So should such claims be treated dubiously? I asked.

Harding nodded solemnly. “Very,” she said.

If you wish to understand at once why we know as little as we do about human origins, I have the place for you. It is to be found a little beyond the edge of the blue Ngong Hills in Kenya, to the south and west of Nairobi. Drive out of the city on the main highway to Uganda, and there comes a moment of startling glory when the ground falls away and you are presented with a hang glider’s view of boundless, pale green African plain.

This is the Great Rift Valley, which arcs across three thousand miles of east Africa, marking the tectonic rupture that is setting Africa adrift from Asia. Here, perhaps forty miles out of Nairobi, along the baking valley floor, is an ancient site called Olorgesailie, which once stood beside a large and pleasant lake. In 1919, long after the lake had vanished, a geologist named J. W. Gregory was scouting the area for mineral prospects when he came across a stretch of open ground littered with anomalous dark stones that had clearly been shaped by human hand. He had found one of the great sites of Acheulean tool manufacture that Ian Tattersall had told me about.

Unexpectedly in the autumn of 2002 I found myself a visitor to this extraordinary site. I was in Kenya for another purpose altogether, visiting some projects run by the charity CARE International, but my hosts, knowing of my interest in humans for the present volume, had inserted a visit to Olorgesailie into the schedule.

After its discovery by Gregory, Olorgesailie lay undisturbed for over two decades before the famed husband-and-wife team of Louis and Mary Leakey began an excavation that isn’t completed yet. What the Leakeys found was a site stretching to ten acres or so, where tools were made in incalculable numbers for roughly a million years, from about 1.2 million years ago to 200,000 years ago. Today the tool beds are sheltered from the worst of the elements beneath large tin lean-tos and fenced off with chicken wire to discourage opportunistic scavenging by visitors, but otherwise the tools are left just where their creators dropped them and where the Leakeys found them.

Jillani Ngalli, a keen young man from the Kenyan National Museum who had been dispatched to act as guide, told me that the quartz and obsidian rocks from which the axes were made were never found on the valley floor. “They had to carry the stones from there,” he said, nodding at a pair of mountains in the hazy middle distance, in opposite directions from the site: Olorgesailie and Ol Esakut. Each was about ten kilometers, or six miles, away—a long way to carry an armload of stone.

Why the early Olorgesailie people went to such trouble we can only guess, of course. Not only did they lug hefty stones considerable distances to the lakeside, but, perhaps even more remarkably, they then organized the site. The Leakeys’ excavations revealed that there were areas where axes were fashioned and others where blunt axes were brought to be resharpened. Olorgesailie was, in short, a kind of factory; one that stayed in business for a million years.

Various replications have shown that the axes were tricky and labor-intensive objects to make—even with practice, an axe would take hours to fashion—and yet, curiously, they were not particularly good for cutting or chopping or scraping or any of the other tasks to which they were presumably put. So we are left with the position that for a million years—far, far longer than our own species has even been in existence, much less engaged in continuous cooperative efforts—early people came in considerable numbers to this particular site to make extravagantly large numbers of tools that appear to have been rather curiously pointless.

And who were these people? We have no idea actually. We assume they wereHomo erectus because there are no other known candidates, which means that at their peak—theirpeak —the Olorgesailie workers would have had the brains of a modern infant. But there is no physical evidence on which to base a conclusion. Despite over sixty years of searching, no human bone has ever been found in or around the vicinity of Olorgesailie. However much time they spent there shaping rocks, it appears they went elsewhere to die.

“It’s all a mystery,” Jillani Ngalli told me, beaming happily.

The Olorgesailie people disappeared from the scene about 200,000 years ago when the lake dried up and the Rift Valley started to become the hot and challenging place it is today. But by this time their days as a species were already numbered. The world was about to get its first real master race,Homo sapiens . Things would never be the same again.