It started with a tooth. On Dec. 17, 1992, as Gen Suwa of the University of Tokyo walked across the badlands of Ethiopia in Aramis, 45 miles south of Lucy’s resting place, a glint from the ground caught his eye. “I knew immediately that it was a hominid”–a humanlike primate, he says. “And because we had found other ancient animals that morning, I knew it was one of the oldest hominid teeth ever found.” Over that winter and the next, the 20-person team uncovered additional specimens locked in 4.4 million-year-old sediment, coming up with teeth, arm bones and parts of a skull and jaw that turned out to belong to 17 individuals. To the scientists, they clearly represented a new species. White (codis-coverer of Lucy), Suwa and Berhane Asfaw of the Paleoanthropology Laboratory in Addis Ababa named it Australopithecus ramidus. Australopithecus means “southern ape”; ramidus means “root.”

The diminutive creatures–adults would have weighed 65 pounds and stood four feet tall–are so perfectly positioned between humans and apes that Darwin must be smiling in his crypt. Members of the new species, write its discoverers, “share a wide array of traits with A. afarensis”–Lucy. But ramidus lacked some of 3.2 million-year-old Lucy’s definingly human characteristics. It had canine teeth comparable to ancestral apes’, for instance, and an unmistakably chimpanzeelike skull. As paleoanthropologist Bernard Wood of the University of Liverpool puts it in an accompanying commentary in Nature, ramidus “lies [extremely] close to the divergence between the lineages leading to the African apes and modern humans . . . The metaphor of a ‘missing link’ has often been misused, but it is a suitable epithet for [this] hominid.” Not that there was much doubt, but ramidus shows unequivocally that, as White says, “Darwin was right: humans evolved from an African ape.”

More surprising is how recently that happened. Ever since Darwin, anthropologists have believed that the lineage leading to Homo sapiens and that leading to today’s African apes were rooted in a common ancestor. For decades they believed that our last common ancestor lived 15 million to 20 million years ago. But then molecular biologists horned in. They compared blood proteins in today’s apes to blood proteins in humans, found them strikingly similar and in 1973 pronounced humans a lot closer to their simian cousins: the two lineages could not have split more than 4 million to 6 million years ago. (If they had, the proteins would have diverged more.) This became known as the “Eve” hypothesis, for the woman who would have been the first humanlike child of that common ancestor. Anthropologists told these interlopers to crawl back under the test tubes where they belonged. But virtually every recent fossil find has put the naked ape and the hirsute one on ever-closer branches of the genealogy. Ramidus confirms once and for all that the common ancestor lived just a little more than 4.4 million years ago (chart). Humans are barely down from the trees.

Now the question is, what brought them down? There is wide agreement that walking upright, on two feet, was the fateful change that spurred the evolution of humans. It made possible quintessentially human traits such as tool use (easier if you’re not using knuckles to get around), a big brain (spurred by use of tools) and families (bipedalism allowed early humans to carry infants and food). Did ramidus walk upright? Although the fossil hunters have not yet found any hip or leg bones that would clinch the case, two other bones are suggestive. A forward opening for the spine at the base of the skull and canine teeth smaller than apes’ are both typical of later creatures that walked on two feet. Moreover, ramidus’s arm wasn’t built for swinging through trees.

If it turns out that ramidus walked upright, then the implications for human evolution will be dramatic. Aramis was a thickly wooded flood plain 4.4 million years ago, filled with the chatter of colobine monkeys, prowling saber-toothed cats, browsing kudu, pigs and bears. (The habitat may explain why humanlike fossils older than 4 million years are so rare: in a woodland, acid from vegetation destroys bones before they fossilize.) But the conventional wisdom holds that our ancestors did not walk upright–and thus take the first steps toward becoming fully human–until climate change turned forests into savannas. On these vast grasslands, say textbooks, natural selection favored creatures that could get around other than by swinging from vines. But there’s a rogue view, too. “I have argued that bipedalism could only have evolved in a protected environment, like a woodland, not an exposed one,” says anatomist Owen Lovejoy of Kent State University. “On a savanna, a creature just learning to walk, and therefore very slow, would have been exposed to predators.”

Since learning to walk in a carnivore-filled forest is not exactly a prescription for longevity, ramidus must have derived a different advantage from the new posture. A clue to what that might be lies in its teeth. “In apes, the sharp projecting canine teeth are used as weapons of threat and display,” explains White. But ramidus had low canines that wouldn’t have frightened a kudu. So females must not have chosen mates based on how convincingly they bared their teeth at rivals, argues Lovejoy. Instead, females sought out mates who could help care for offspring; the best caregivers may well have been those who walked upright, using their hands to carry food and infants. According to natural selection, protohumans that became bipedal would have been more apt to mate, launching their genes for that trait into the next generation. “Lately, people have claimed some pretty bizarre reasons for walking upright, such as reducing exposure to solar radiation,” says Lovejoy. “This discovery promises to wipe such claims off the slate.”

The fossil hunters are returning to the field in November. If further finds confirm that our oldest direct ancestor learned to walk in the primeval forest, it will push to center stage a very ’90s notion: the crucial spur toward becoming human was changing from a tooth-baring aggressive ape to one that carried home dinner and helped raise the kids.

The fossils just discovered in Ethiopia come from the oldest direct human ancestor known. The new species, Australopithecus ramidus, has feature midway between apes and humans. It promises to provide clues to still earlier evolutionary stages.

A. afarencis Brain size: 400-500 ml Discovery: 1974 Site: Hadar, Ethiopia Features: “Lucy.” Fully upright, lived in family groups throughout eastern Africa. A. africanus Brain size: 400-500 ml Discovery: 1924 Site: Taung, South Africa Features: Long arms, light build; descendant of Lucy who lived in social groups. A. robustus Brain size: 530 ml Discovery: 1938 Site: Kromdraai, S. Africa Features: Brawnier cousin of africanus; left no descendants. Homo habilis Brain size: 600-800 ml Discovery: 1960 Site: Olduvai Gorge, Tanzania Features: First known species in the “Homo” family of humans; first tool user. Homo erectus Brain size: 750-1,250 ml Discovery: 1891 Site: Trinil, Java Features: First species to move out of Africa, colonizing the Middle East and to China. Homo sapiens (modern) Brain size: 1,200-1,700 ml Discovery: 1868 Site: Cro-Magnon, France Features: Ponders its origins and evolution. (SOURCE: UNIVERSITY OF CALIFORNIA, BERKELEY)