At first, all Sgt. Paul Dostie had were handfuls of bones — fragile, gnawed-on human bones.

The animal-ravaged remains, found in a shallow grave in the Inyo National Forest in May 2003, told him little. Dostie knew only that the victim was a petite woman who wasn't dressed for the rugged Sierra Nevada, judging by her lacy blouse and flimsy jacket.

Was this the third murder this ski resort town had seen in a quarter-century?

• Click here to visit FOXNews.com's Evolution & Paleontology Center.

Bones don't talk, and the 20-year police veteran realized that cracking this case would take more than old-fashioned detective work.

Over the next 3½ years, Dostie combed the Internet for scientists who helped him extract information from the remains. Won over by Dostie's dedication and aw-shucks good nature, they contributed their expertise, often for free.

Dostie slowly compiled details of the victim's life story: where she was from, what she ate as a child, what she looked like, where she spent her last few months — everything but her name.

The search consumed him. He scoured scientific papers and attended conferences of forensic experts in search of new technologies.

"I probably know more about her and how she lived and died than anyone else out there," he said.

Dostie now believes he's weeks away from confirming the victim's identity. Only then will he be able to start the investigation he's waited years to pursue: the search for her killer.

___

It began with a hiker walking his dog in the national forest. Something in the bushes grabbed the dog's interest — a human skull.

Police searched for other remains, but found nothing until few days later, when a hunch led a sheriff's deputy up a nearby hill.

There, beneath the pines, Dostie was introduced to the victim who would define his career. Her cheap watch was still ticking, though it had spent the winter under snow.

The case got off to a good start: An employee of the Mammoth Lakes Visitor Center soon came forward, saying she remembered a small woman who'd come in the previous fall. She had prominent cheekbones and straight black hair flowing past her shoulders.

While her male companion was getting camping information, the woman told the employee in accented English said she was afraid of him — a heavyset white man with a mustache. The employee handed her a card from a local women's shelter, and the couple left.

The medical examiner had said the victim might be Asian, which seemed to fit the employee's recollection. He asked about the couple at local campgrounds, distributed fliers with the woman's description, and placed ads in Asian-language newspapers as far away as Los Angeles.

But it was a dead end. A year later, he was still empty-handed.

Dostie was casting about for new ideas in May 2004 when he heard about a Florida company called DNAPrint Genomics that searches a person's DNA for clues about their racial makeup. He sent a bone sample.

"It was 100 percent Native American," said Matt Thomas, the company's senior scientist. "I don't see that many samples that are that clearly Native American."

The finding still left a range of possibilities — native peoples with similar genetic markers are found throughout the Americas. But it gave Dostie something to work with.

"I took anthropology in college," he said. "I knew it was the key."

He turned to the Internet and found Philip Walker, then president of the American Association of Physical Anthropologists.

"Phil opened up a whole new world to me," he said. "His life's work is looking at bones."

Walker researches patterns of violence in ancient cultures, and he often puzzled over the same questions vexing Dostie: Who is this person? How did they die?

Walker recruited an orthopedic surgeon, a forensic pathologist and an anthropologist. Together, they gave Dostie a clearer picture of the victim.

She'd been repeatedly stabbed — a fact that escaped the medical examiner. She was likely a Native American from Mexico or Central America, between 30 and 35; and very small, no taller than 4-foot-9 and no more than 90 pounds.

Scarring on her pelvis meant she'd delivered at least one child, and the poor state of her teeth told Walker she'd never seen a dentist. The bones and muscles in her shoulders pointed to a life of hard physical labor.

"This is clearly a disenfranchised person who was vulnerable," said Walker, explaining his motivation to work on the case for free.

Walker wanted to know more about her diet and the water she drank — clues to her ethnic background and geographical origin.

They turned to Henry Schwarcz, a geologist who analyzes the chemical composition of ancient human remains. Remains that were puzzling to Dostie spoke clearly to the Canadian professor.

"In her childhood, she had been living mostly on corn — cornmeal, tortillas, up to a level that would be almost nutritionally unhealthy," Schwarcz said.

He looked for oxygen atoms in her teeth. These are absorbed from the water a person drinks as a child, and since most drinking water comes from local rain, they can be a good indicator of a person's origin. She seemed to have been raised in southern Mexico, or even farther south.

Schwarcz also looked at her bones and hair — cells that regenerate over the years, incorporating new information throughout a person's life.

These told a different story: In the last 18 months of her life, the woman's protein intake was like that of a typical North American. There was variation in the oxygen isotopes, suggesting she'd moved around a lot.

Walker also recommended a look at the woman's mitochondrial DNA, genetic material that only holds information about a person's maternal line. The sequence was sent for comparison to two scientists who manage databanks mitochondrial DNA. One confirmed she could be from southern Mexico.

The other had a hit: Among 3,000 specimens in his databank at the University of California, Davis, David Glenn Smith found a maternal relative of the victim: a Zapotec Indian living in the southern Mexican state of Oaxaca.

___

The only problem was that Dostie had never heard of Oaxaca, and didn't speak Spanish, much less Zapotec.

He went back to the Internet.

"I'm not that smart myself," he said, "but I can find a lot of people who are."

A UCLA linguist connected him to a Oaxacan graduate student, who introduced Dostie to the man who would guide him through the next phase of the investigation.

Ray Morales, president of the Oaxacan Business Association, was the perfect link. He spoke English, Spanish and Zapotec, and ran a business that delivered cash from immigrants in California to their families in Oaxaca.

Morales was impressed by Dostie's perseverance.

"This is a guy who doesn't know Oaxaca, who doesn't speak the language, taking a case he could have easily filed away," Morales said.

In May 2006 Morales went to Oaxaca.

"Oaxaca can feel pretty small, the communities are pretty tight knit," said Morales. "I thought it would be pretty easy."

He spread word of the missing woman through the local media. He carried fliers with her picture, and made a big splash.

Morales found the DNA donor, but was surprised when the woman claimed she didn't know the victim. And no one in the woman's village seemed to know of a missing woman who matched the description.

By now, Morales felt a sense of responsibility — to the woman, who seemed to have no one else, and to the detective who had brought the case this far.

"Why is science pointing to this town, but no one is filing a missing person report?" he asked.

Morales went back to Oaxaca.

This time, he went quietly. He spent time in the village and got to know the residents. Slowly weaving together rumors and implications, he formed a picture of a woman who might be the victim.

Her mother died when she was young, and she'd left the village for a nearby town. She'd returned about 10 years ago, then made a scandalous exit to the United States with help from a married man in Southern California.

Certain he was onto something, Morales collected DNA from an uncle and a half brother. But the results were inconclusive.

He needed DNA from a maternal relative. The woman's sister lived in another part of Mexico, and one of the victim's own children was said to be living in another Oaxacan village. Morales planned another trip.

Then violence erupted in Oaxaca. A teachers' strike evolved into mass protests involving leftists, Indian groups and students all calling for the governor's resignation.

Morales waited out the unrest, which lasted six months. He now has a ticket to return on Dec. 20.

Dostie is excited at the prospect, but he's also patient.

"We've been in this over three years, just trying to get to day one — to the day we can start figuring out who knew the victim, who could have killed her," he said. "We'll get there."