On top of the device a small gray screen that looked something like an electronic fish finder showed a pattern of blurry, bumpy horizontal lines. As Thacker proceeded, the lines did too, slowly traversing the screen. Suddenly he stopped.

"See that?" he said. The horizontal lines on the screen had grown ragged and in one area plunged in a steep dip.

"That shadow there, I don't think that's a grave shaft," Thacker said. "But this one ..."

He reached down and planted a small red flag in the ground at his feet.

All around him, red flags dotted the earth. Each one marked an underground anomaly that Thacker or his colleagues had detected; each one, in other words, represented a possible grave.

The Preservation Society of Chapel Hill, working with grants from the Kelly Webb Trust and the Strowd Roses Foundation, initiated the project to try to discover, map and, if possible, ultimately identify long-lost graves in the black sections, known as Sections A and B, of the cemetery. The society contracted with Environmental Services Inc., last week to use ground-penetrating radar - the device Thacker was using - and electrical resistivity techniques to locate probable grave sites.

"We wanted to determine how many graves are in these sections," said Ernest Dollar, executive director of the Preservation Society. "By doing this we hope to give respect to their finally resting place and give the community back a piece of mind about their ancestors."

Black Chapel Hillians were buried there beginning more than 150 years ago - the oldest known grave in that section is that of a slave named Ellington Barnett, who died in 1853 - but very little remains to mark their resting places. Some of them may never have had headstones, and many of gravestones that were once there have been destroyed, moved or otherwise lost through vandalism or neglect; untold markers were broken or buried in a single day in 1985, when football fans attending a UNC-Clemson game at Kenan Stadium used sections A and B of the cemetery as a makeshift parking lot.

Now, only a few rough fieldstones and even fewer formal markers remain, scattered across the patchy grass under the pines. The number of red flags in the ground last week indicated the likely presence of dozens of unmarked graves.

"This is the first step," Dollar said, watching Thacker maneuver the ground-penetrating radar device and team of three manipulating the resistivity apparatus. "Once the readings are done, we will generate a map plotting the probable graves and work with the community to try to identify as many as we can. Then, eventually, we'd like to mark the graves again, which raises a whole series of questions -- what kind of markers, and who pays for them?

"But those are all still to come. The first step is finding the graves."

Thacker and the resistivity crew moved methodically across the tract in a linear pattern, like a farmer mowing a field. After they finished plotting the sections, he said, another team would come in and probe the sites with long fiberglass probes.

"The probes will sink right in where grave shafts were dug," he said. "The surrounding soil will be much firmer; although the older the grave the more subtle the difference. That way they can mark the four corners of each shaft."

Finding the graves is the easy part; identifying who is buried where will be more difficult.

"That's where we ask the public for help," said Steve Moore, chairman of the Cemetery Advisory Board. "The written historical records are very spotty. Once we know what we have here, we'll let the public know what we know. People might know they have ancestors buried in this cemetery, and they may have family histories, family Bibles, verbal histories or other information that could help us match up names with grave sites."

Chapel Hill has two other little-known, much smaller cemeteries with unmarked graves, Moore said - the Barbee-Hargraves Cemetery on Greenwood Road, and the West Chapel Hill Cemetery near the railroad tracks and Carrboro town line. The Preservation Society hopes to conduct similar studies at those sites.

"There are no records at all for those cemeteries," Moore said. "Hopefully, after this one, we'll be able to do the same thing there. This is the prototype, the first attempt to use these different methods - science, the historical record and family histories - to pull together all the information we can gather about these sites. This is just the beginning."