Chinese pharmacists have used fossils - which they call dragon teeth and dragon bones - as ingredients in potions intended to cure ailments ranging from backache to sexual impotence. The fossil-rich caves of southern China have been, and still are, sedulously mined by farmers, who sell these medicinal treasures to apothecaries in the cities. In just such a pharmacy, in Hong Kong in 1935, the German paleoanthropologist Ralph von Koenigswald came across a large fossil primate molar that did not belong to any known species. Over the next four years he searched further in Hong Kong and Guangzhou (Canton) and found three more of the oversize teeth, thereby establishing the existence of an extinct ape, the largest primate ever to roam the earth. He named the genus Gigantopithecus, meaning "gigantic ape," and the species blacki, in honor of his late friend and colleague Davidson Black.

At the time of the discovery, during the 1930s, von Koenigswald was working primarily in Java, unearthing fossils of human ancestors and their relatives. China's unique fossil shops had already played a major role in tracking down Homo erectus, which lived in Asia between about one million and 300,000 years ago. Homo erectus remains were first unearthed in Java in the 1890s, but pursuit of the source of dragon bones subsequently led to a system of fossil-filled crevices and caverns near the town of Zhoukoudian (Choukoutien), thirty miles from Beijing. There, in 1929, a team of Chinese and Western scientists discovered the first of a series of Homo erectus skulls that became world famous as "Peking man."

The original fossils of Peking man disappeared during the confusion of World War II - fortunately, after they were described and cast by anatomist Franz Weidenreich. The war also caught up with von Koenigswald, who was taken prisoner by the Japanese in Java. His precious collection of Gigantopithecus teeth - at that point, the only known specimens of the fossil ape - spent the war years in a milk bottle buried in a friend's backyard on the island.

Photographed at the American Museum in the 1940's, German paleoanthropologists Ralph von Koenigswald, left, and Franz Weidenreich, right, pose with the skulls of apes, Homo erectus, and modern humans. The first scientist to discover teeth of Gigantopithecus, von Koenigswald correctly observed that they belonged to an ape, while Weidenreich argued for their humanlike characteristics.

Meanwhile, however, Weidenreich, who had retreated from Beijing to the American Museum of Natural History in New York, set about studying plaster casts of the four teeth. Because of the unusually large size of a few of the Homo erectus specimens from Java, Weidenreich came up with the notion that there had been a period of gigantism in human evolution, and that modern humans were the diminutive descendants of these giants. In Apes, Giants, and Man, published in 1946, he argued that the Gigantopithecus teeth were humanlike, and that von Koenigswald had been mistaken in considering the animal an ape rather than a member of the human family tree.
During von Koenigswald's wartime internment, Weidenreich's views became widely accepted. To end the controversy that arose, more complete specimens of Gigantopithecus had to be found, a task only the Chinese could undertake, for the country was closed to Western scientists. In the 1950s, with the establishment in Beijing of what is now the Institute of Vertebrate Paleontology and Paleoanthropology, Chinese paleontologists began to search for the source of the Gigantopithecus fossils. Two veterans of the Peking man expedition, Pei Wenzhong and Jia Lanpo, headed a team that visited the warehouses that supplied all the apothecary shops in China with dragon bones and dragon teeth. They found vast quantities of fossils in Nanning, the capital of Guangxi Province. From there, they divided into two teams: one, led by Pei, headed north; the other, led by Jia, went south.

Jia's paleontological detective work took him to southernmost Guangxi, a karstic, or eroded limestone, region of great rock towers riddled with caves. In the town of Daxin, which the local people said was the source of all the fossils, they were directed to an old woman who had, in her house, a bamboo tray full of fossils. One of them was a Gigantopithecus tooth. She pointed out a very tall rock tower, described by Jia as "a hundred meters straight up - almost falling over, it was so steep." The mouth of a cave was clearly visible behind a screen of brush.

A cave near the top of the rounded limestone tower at Liucheng, China has yielded three Gigantopithecus jawbones and nearly a thousand teeth.

Although it was four in the afternoon and raining hard when they arrived, Jia says, "We were young, and couldn't be restrained. We climbed straight up to that cave." That very day, Jia himself found a Gigantopithecus tooth embedded in a hard, reddish matrix, the first time that a paleontologist had discovered a fossil of Gigantopithecus in a geological context.
Meanwhile, Pei was making a more momentous discovery to the north. Word had reached the scientists of a giant jawbone discovered by an old farmer in 1956 at a cave site called Liucheng. When Pei saw the fossil, he was able to identify it at once as the jawbone of Gigantopithecus, because it had all but three of its teeth still attached. On a second visit, in 1957, Pei's team discovered the first Gigantopithecus jawbone in place, in a very hard deposit resembling red clay. Another was excavated in 1958. One of the jawbones was extraordinarily large; presumably, it belonged to an adult male, while the other two were thought to be from an adult female and a juvenile.

In addition to the jawbones, Pei's group discovered nearly a thousand Gigantopithecus teeth and numerous other mammalian specimens, including some unusual dwarf varieties. Among them was a short-muzzled panda half the size of the living giant panda. Chinese scientists have recently suggested that this dwarf species was a direct ancestor of the modern one.
The next development came in 1965 with the discovery of twelve Gigantopithecus teeth at Wuming, a few hours' drive north of Nanning. These teeth were significantly larger than their counterparts from Liucheng, and the other animal fossils found with them suggested that the site was considerably younger (current estimates are that Liucheng is one million years old and that Wuming is between 300,000 and 400,000 years old). This suggested, first, that Gigantopithecus was around as a species for a considerable period, and second, that it may have become larger as the species evolved. This is a trend seen in other large mammals that evolved during the Pleistocene epoch, 1.8 million to 12,000 years ago.

A striking confirmation of both points was the discovery three years later that a smaller, earlier form of the giant ape had once inhabited northern India. In 1968, a farmer came forward with three pieces of a jawbone he had found twenty-four years before, when he was a boy of twelve working in his father's field. The specimen was identified by primatologist Elwyn Simons as belonging to a distinct species, Gigantopithecus giganteus, about half the size of Gigantopithecus blacki. The new species was not only smaller but also more ancient, coming from sediments that have been dated (by paleomagnetic reversals) to about 6.3 million years ago.

The discovery of the jaws resolved, at least for most scientists, any doubts that the creature was apelike and not, as Weidenreich had argued, humanlike. Based on the fossils, Gigantopithecus is now placed among the Asian apes, a descendant, along with the orangutan, of the earlier ape ancestor Sivapithecus, best known from an 8-million-year-old skull discovered in Pakistan. Its size and ape affiliation suggest Gigantopithecus was a ground-dwelling, fist-walking creature.

While more teeth of the extinct ape have been found, no other bones have turned up. Based only on the jaws and teeth, however, an attempt can be made to reconstruct both the animal and its way of life. The jaws are deep (top to bottom) and very thick. The molars are low-crowned and flat, with very thick enamel caps suitable for heavy grinding. The premolars are broad and flat and resemble molars. The canine teeth are not sharp and pointed but shaped more like what one would expect premolars to look like, while the incisors are small, peglike, and closely packed. The canines and incisors together form a specialized cutting tool, most similar to what is found in some present-day tree sloths and in the extinct giant ground sloth. The features of the teeth, combined with the massive, robust jaws, lead to the inevitable conclusion that the animal was adapted to the consumption of tough, fibrous foods by cutting, crushing, and grinding them.
As a rule, large herbivores subsist on diets of coarse leaves and grasses, which are low in nutritional value but typically available in very large quantities. (Large animals succeed with this regime partly because their metabolic requirements are relatively low, in terms of energy required per unit of body mass.) One suggestion is that Gigantopithecus, or at least the larger species in China, was adapted, like the giant panda, to a diet of bamboo, the giant grass abundant in the region. The jaws of Gigantopithecus and the giant panda, if set side by side with the jawbones of, say, the gorilla and the grizzly bear, appear thicker, deeper, and more massive. These differences reflect the specialized diet of the panda (and, by inference, of Gigantopithecus) compared with the much more general diet of the gorilla and grizzly.

A further similarity between Gigantopithecus and the giant panda is a high incidence of tooth cavities. Wu Rukang, in an encyclopedic survey of the Gigantopithecus teeth in China, found cavities present in 11 percent of them - an unusually high rate for an ape, but more or less equivalent to the rate of dental cavities in the fossil remains of the giant panda. Another Chinese researcher, Zhang Yinyun, has reported a high incidence of hypoplasia - pitting in the tooth enamel that indicates periods of arrested development. These may be a result of disease or food shortage. While no certain conclusion may be drawn, we do know that bamboo is subject to periodic die-offs, which produce food shortages that threaten the survival of the giant panda.
A more direct line of evidence that could be pursued regarding the diet of Gigantopithecus was pointed out to me by Bob Thompson, a graduate student in New World archeology, who attended one of my lectures about the extinct ape. He suggested we might look at the teeth for adhering phytoliths, microscopic pieces of silica found in many plants. The existence of phytoliths has been known since the early nineteenth century, and scientists had already successfully looked for them on stone tools, to which they apparently bond physically by the combined action of friction and moisture. But it was the first time, as far as I know, that anyone had suggested looking for them on fossil teeth.

Four teeth were borrowed for study from the British Museum (Natural History) and the Senckenberg Natural History Museum in Frankfurt: an upper incisor, lower canine, lower premolar, and lower molar. After the teeth were cleaned to insure that what we found was definitely part of the fossils, they were examined under a scanning electron microscope at the University of Iowa by Smithsonian paleoecologist Dolores Piperno. At least thirty phytoliths were found on the teeth, most of them on the molar. We also detected tiny scratches apparently left by phytoliths, which are harder than tooth enamel. In one case, we found a phytolith sitting astride the end of the track it had plowed into the tooth - like a sled stopped in its path in the snow.

More than half of the phytoliths we observed were long and needlelike and could be attributed to the vegetative part of grasses, possibly bamboo. The rest were conical or hat shaped, attributable to the fruits and seeds of dicotyledons. Piperno tentatively identified them as fruits from a tree of the family Moraceae, quite possibly durian or jackfruit, both of which are common throughout tropical Southeast Asia. This proved that Gigantopithecus had a varied diet, although we still suspect that bamboo was its staple food.

What other conclusions can be drawn about the extinct ape? An outstanding characteristic of giant herbivores is their extreme slowness. They have no particular need of speed: their size and thick skins protect them from predators, and of course their feeding habits require no more of them than that they move from place to place as they systematically denude the landscape of vegetation. Furthermore, they are usually stuffed full of bulky food to digest, which tends to produce inertia. Gigantopithecus probably followed this pattern.

Finally, the adult males of the giant ape were much larger than the females. Australian anatomist Charles Oxnard statistically analyzed 735 teeth of Gigantopithecus that were complete enough to be measured accurately. He found that they divided neatly into two size groups of equal number, which he interpreted to represent the males and females in the population. The contrast was greater than that seen in any living primate species, including the gorilla and the orangutan, two species in which the male is substantially bigger than the female. In Gigantopithecus, the difference in tooth size between the sexes may represent strong competition among males for mates - a clue to the species' social behavior.

The largest of the jaws, along with some of the teeth, are compared at with modern human remains.

To gain a more complete image of what the giant ape looked like, we sought the help of Bill Munns, who creates highly realistic, life-size models of existing endangered primates - gorillas, orangutans, and the Chinese golden monkey - for zoos and educational institutions. Based on the jaws and teeth, and using the proportions of the skulls of existing great apes, we estimated that the average male Gigantopithecus had a skull that measured eighteen inches from the bottom of the jaw to the highest point of the sagittal crest (a male gorilla, for comparison, has a skull ten inches high).

The next step was to project a hypothetical skeleton from the hypothetical skull. For this purpose Munns used as references two of the largest terrestrial primates known: one modern, the gorilla, and one from the fossil record, the extinct giant baboon Theropithecus oswaldi. In determining the size of Gigantopithecus, we felt it necessary to scale the body back a bit, so as not to be influenced too much by the giant ape's extraordinarily deep and thickened mandible. Nevertheless, given that the average male silverback gorilla is about six feet tall (standing erect) and weighs in at 400 pounds, Munns calculated that the average Gigantopithecus male was more than ten feet tall and weighed as much as 1,200 pounds - comparable to a large male polar bear.

Bill Munns stands next to his model of a Gigantopithecus male, a quadrupedal, fist-walking creature that also could have stood erect, as bears do.

One intriguing question is what contact our remote ancestor, Homo erectus, may have had with the giant ape. That the two coexisted for some time in the same region is supported by direct evidence. In 1965, Vietnamese paleontologists discovered the remains of both creatures at Tham Khuyen, a cave site in Lang Son Province, near the Chinese border. Chinese excavators followed suit, excavating Gigantopithecus and Homo erectus side by side in Hubei Province in 1970 and more recently, in 1987, in Sichuan Province.

Gigantopithecus was native to southern Asia, while Homo originated in Africa about 1.6 million years ago and migrated eastward, finally arriving in what is now Southeast Asia about one million years ago. The opportunity to explore this nexus attracted archeologist John Olsen and me to Vietnam. One reason we did not choose to go to China was that all the promising sites had been reserved by Chinese paleoanthropologists, and we doubted we would find a new site in a region that had been so thoroughly mined. In contrast, Vietnam had no history of exploiting fossil-rich caves for dragon bones. And so in January 1989 we found ourselves probing four caves at the base of a karst tower near the hamlet of Lang Trang, about 100 miles southwest of Hanoi, as part of a joint American-Vietnamese expedition.

The caves had seemed promising in our preliminary survey the previous May, and as we began work, even local children brought us fossil mammal teeth (although we tried to discourage them), which they retrieved from an underground stream by squeezing through a crevice in the cave we called Lang Trang I. Meanwhile, we began cutting out blocks of breccia, the sediment typical of caves, which is gradually formed by material washed or otherwise transported into a cave and cemented with limestone dissolved from the cave walls and ceiling.

The fourth day of our dig, Friday the thirteenth, turned out to be a lucky one: within the main deposit I found a lens-shaped vein of dark, sandy sediment that was unusually rich in fossils. The material had probably washed into the cave from the nearby Ma River, which in ancient times meandered right alongside the karst tower. Perhaps a violent monsoon had caused the river to overflow its banks and flood the cave. After the waters receded, the slow process of breccia formation began again, sealing the sandy lens within Lang Trang I.

We immediately set to work cutting out hunks of the sandy deposit, revealing a small chamber that we surmised was the source of all the fossils the children had been bringing us. Our finds included barking deer, a musk deer the size of a big dog; sambar, a large deer with three-pointed antlers; wild boar; and giant panda. A huge, ridged molar, weighing several pounds and belonging to Stegodon, an extinct relative of the elephant, assured us that we were dealing with a Pleistocene site that might also contain Homo erectus and Gigantopithecus. One softball-sized sample of this deposit was later analyzed at the University of Iowa, revealing that it also contained some small teeth and fragmentary limb bones of a diverse microfauna, including rodents, reptiles, fishes, and riverine sponges. These fossil fragments were about the same size as the coarse sand particles they were mixed with.

Then, on January 18, 1989, Nguyen Van Hao made a key discovery: in the floor of the fourth cave he found a premolar of Homo. Since it was an isolated tooth, we found it difficult - impossible, really - to identify the species. Since then, four additional teeth of Homo have been recovered from caves I, II, and IV. Subsequently, a boar tooth from cave I has been dated (by a method called electron-spin resonance) to about 480,000 years ago. Given this preliminary date, the specimens should be assigned to Homo erectus. The discovery helps fill the gap between Zhoukoudian, in northern China, and Java, more than 3.000 miles to the south.

We now have a fairly complete picture of the Pleistocene environment of Lang Trang. The jungle vegetation would have been more lush, but not startlingly different. The fauna, however, would have been striking, with huge beasts of all kinds dominating the landscape. Carnivores such as the tiger and leopard were much more common then and competed for food with species, such as the Asiatic black bear, that have entirely disappeared from Vietnam. And they all competed with the wolf and the Asiatic wild dog in preying on the dozens of bovid and cervid species (cowlike and deerlike mammals). Also present were the rhinoceros and elephant (both now rare) and the stegodon, as well as the orangutan and tapir, both now extinct in Vietnam. The giant panda, also now vanished, chomped its way through the bamboo stands. Taken in this context, Gigantopithecus was no freakish monstrosity, but simply the primate example of a Pleistocene phenomenon.

Primates make up 13 percent of the total fauna in our collection. At least five genera are accounted for: two types of macaque monkey, orangutan, langur monkey, gibbon and Homo. So far we have been disappointed only by the absence of Gigantopithecus.

Sometime near the end of the middle Pleistocene, perhaps 200,000 years ago, Gigantopithecus became extinct. The animal had flourished for at least six million years, quite a respectable figure, but it went the way of a great many genera of every shape and size. At about the same time, the giant panda disappeared from much of its original territory, notably insular southeast Asia, until it now survives only in the cold upland regions of Sichuan Province. The best guess as to what caused the panda's extinction in Southeast Asia is human hunting: even now the animal is hunted for food and for pelts, despite the best efforts of the Chinese government to discourage the practice. Similarly, human hunting may have led to the demise of Gigantopithecus.

Environmental change may also have been a contributing factor, just as the bamboo die-off in China in the 1970s nearly wiped out the remaining population of giant pandas, with fewer than a thousand estimated to have survived. Or by eating the tender bamboo shoots and exploiting the plant for other purposes, including toolmaking, humans may have outcompeted the giant ape for this critical resource. The competition from both humans and the giant panda may have been too much.

Gigantopithecus is gone. Or is it? Following the publicity about our research in Vietnam, I have received several letters from veterans who say that they came face to face with huge, hairy apes in the Southeast Asian jungle when they were posted in Vietnam. And of all the theories advanced to provide a zoological identity for Bigfoot, the Abominable Snowman, and other elusive creatures, perhaps the most popular is that they are none other than Gigantopithecus, still alive in relict populations (relict populations of Neanderthal man run a close second). While these contemporary reports are probably false, we can contemplate the time when our remote ancestors did encounter the giant of all apes in the tropical rainforests of Southeast Asia.