BUENOS AIRES, JUNE 22 -- Clyde Snow, an American forensics anthropologist, arrived in Argentina three years ago to teach people how to dig up the dead. His lessons have helped this young democracy expose the grisly atrocities and identify many of the victims who disappeared during the military repression of the last decade.

Returning today to the United States, Snow leaves behind a team of Argentine investigators he personally trained and organized to track the grim legacy of the military government.

The vivid story of Snow's skillful search through the burial grounds of Argentina is based on the application of scientific techniques now commonplace in U.S. police work but previously untried here.

In the heady days following democracy's return in December 1983, some of the suspected graves of victims of the so-called "dirty war" were haphazardly overturned by inexperienced diggers and mauled by bulldozers. The scattered bones were stuffed into plastic bags that still sit jumbled in various morgues.

Since then, Snow has shown how careful and methodical exhumations can help identify anonymous skeletons and provide incontrovertible proof of atrocities. By demonstrating the use of archaeology and anthropology in forensic science, he has broadened this field for Argentina and increased the powers of local sleuths.

His efforts have drawn requests for similar assistance from human rights groups in the Philippines, Chile, Bolivia, Peru and Honduras.

In Argentina, the 59-year-old Snow has spent countless hours combing cemeteries for John Doe graves and exhuming bones, then painstakingly matching his findings with data about the 9,000 or more victims of the repression. His efforts contributed to the conviction of five former junta members in 1985 and to the preparation of other cases against military officers.

He has not been the only American bringing science to bear in helping Argentina expose its dark recent past. A U.S. geneticist has developed a system here for tracing the heredity of babies born in clandestine detention centers and given surreptitiously to military and police families or others to raise.

As a result of the efforts of Mary Claire King of the University of California at Berkeley, local doctors have begun gene-typing relatives of the disappeared, establishing a permanent data bank intended to reunite kidnaped babies, if found, with their rightful parents, if still alive.

Both Snow and King were recruited by Eric Stover, a staff member of the American Association for the Advancement of Science. Argentine human rights groups had asked the AAAS for assistance after the military's fall.

Snow was a logical choice. After a 19-year career with the Federal Aviation Administration's medical institute in Oklahoma City, he became active in 1979 as a consultant in forensic anthropology.

He knew little about Argentina and spoke no Spanish. And it wasn't easy getting started here. Snow had trouble finding professional anthropologists in Argentina to form a team. Many feared getting involved. Finally, a handful of university students offered to work with the hefty, pipe-smoking American investigator.

"They had some concerns, though," Snow recalled in an interview, "like what shape the bodies would be in. I told them that after 10 years in the ground, they'd be nice, dry skeletons.

"Some were worried, too, about how depressing it would be digging up the coffins. These kids were in a way the younger brothers and sisters of the disappeared."

A Buenos Aires judge assigned the fledgling team its first grave to unearth.

"The way exhumations were being run down here before, you'd wind up with a crowd," Snow explained. "Everyone would be standing around, and the gravediggers would be down there throwing up bones, right and left. Unwittingly or not, they ended up destroying the evidence.

"We arrived out there {in a Buenos Aires suburb} to meet the same sort of gaggle. I laid out a perimeter, then asked everyone, including the police, to stand outside it.

"I had on a badge belonging to the Illinois Coroners' Association. It's big and vulgar as hell. But I always carry it around because whenever you get into a confrontation with police, the guy with the biggest badge wins."

The skeleton uncovered that day belonged to a woman in her early thirties. She had died from a bullet to her head.

Snow left Argentina in May 1984. He returned in early 1985 with several other U.S. specialists to teach a course in forensics.

In April 1985, Snow testified at the trial of the three military juntas charged with overseeing what is often referred to as the war on leftist subversion.

Snow chose to illustrate the case of Liliana Carmen Pereyra, a 21-year-old bank employe last seen in October 1977. Officials had informed her parents that she had been killed in a shootout with security forces.

Pereyra was five months pregnant when she vanished. Other victims claimed to have seen her at the Navy Mechanics School in Buenos Aires, among the most notorious of several hundred clandestine torture centers known to have existed. They also said she had given birth.

Snow testified using slides. The first one he flashed on the screen was Pereyra's fractured skull, which he had dug up in the coastal town of Mar del Plata.

He told the court not only that the young woman had died from buckshot to her head -- he found seven pellets among her remains -- but also that she had given birth before the murder. That was evident from the changed shape of her pelvic bones.

"It was perhaps a bit dramatic," Snow said. "But on the other hand, there's a tendency by these lawyers and judges to see the victims as paper people. When they see the bones and the bullets, they realize they're dealing with real people."

Snow left Argentina again, but came back in April 1986 with a grant from the Chicago-based J. Broderick MacArthur Foundation. He planned to help set up a special technical commission and laboratory under the jurisdiction of the government's undersecretary for human rights. The commission would be responsible for collating all available personal information about those who had disappeared.

The government still had little idea how many victims had been placed in anonymous graves. Questionnaires were sent to 125 municipalities in Buenos Aires province asking for information on such graves -- designated "NN" in cemetery files, for "ningun nombre" in Spanish, or "no name" -- between 1970 and 1984.

Replies came in from 110 municipalities, indicating that the NNs rose steeply above the historical average in the mid-to-late 1970s, the period in which the repression had peaked.

The official death certificates that accompanied many of these coffins also were revealing. They showed a statistically significant increase in gunshot wounds and drownings and in such suspicious diagnoses as cranial traumas or pulmonary edema with congestive heart failure (induced, for instance, when a poisoned plastic bag is put over a victim's head).

For all their determination, Snow and his team of six young medical and anthropology specialists have dug up only 64 bodies so far. Just half a dozen or so of these have been identified. The group still lacks an adequate lab or even a permanent office. Its hopes for a central data bank have yet to materialize.

Yet the political will for such projects may be weakening. This month, the Argentine Congress passed a law blocking further prosecution of most military officers accused of "dirty war" crimes.

Snow has urged the government to continue searching for the dead, out of respect for surviving relatives and as permanent testimony to the atrocities committed. As he said, reciting an old law enforcement rule in his slow-spoken Oklahoma drawl, "You never close the case on a homicide."