The sandy soil that once nurtured the trees and now was claimed by the settlers, was part of a long, broad levee parallel to the river, a summit built over thousands of years by floods depositing sediments. A ridge slightly higher than the rest of the levee gave an ideal vantage point. It was along the crest of the ridge that people built their homes.

Next, they pulled the tree tops towards the center of the oval and tied them together with vines. To finish the house, the women helped cover the frame with deer hides, plant thatch, and bark strips taken from the big trees. The shelter was only partly effective for keeping out the rain and cold, but inside there was a smoldering fire within a stone circle that could be stoked for more warmth, as well as for cooking.

Near the fire in the dirt floor were two shallow pits where food wrapped in animal skins was kept, safe from scavengers and ready to be prepared for the next meal. The dwelling floor and ground around the shelter were swept clean of debris. Outside the house, several more fires burned nearby and there were also more pits. These were about a foot-and-a-half to three feet wide and about eight inches deep. Some, like the holes inside the shelter, were used to store food, while others were for cooking.

Animal skins, scraped with sharp rocks until free of flesh and hair, were loosely stretched across the cooking pits. The edges of the skins were pegged securely to the ground with sharp sticks. A stew of meat from whitetail deer, roots, and herbs simmered in one of these skin containers. From time to time, one of the women tending the stew went to a nearby fire where she used a long stick to pull out a soapstone slab with a hole in it from the flames. By pushing the stick further through the hole, she was able to carry the hot rock to the cooking pit where she dropped it into the stew. The liquid rose to a fast boil from the added heat, then eventually returned to a simmer, which lasted for some time.

She repeated the process many times, walking to the fire, built far enough away to avoid ash blowing into the cooking food, pulling out a dangerously hot rock, then depositing it into the stew, until she was satisfied with the results. Her young daughter shadowed her footsteps, watching and learn-mg.

Behind the shelter and down the slight slope from the ridge crest, another group of fires was scattered from five to ten yards away. Here the men and boys gathered. Their laughter mixed with the sounds of preparing stone weapons and their recalling the day's successful hunt. One man fed green wood onto a fire to cause smoke to waft over deer meat and fish arranged on wooden frames close by. When the food was thoroughly dried and smoked, he would give it to the women to save for other days when the hunting and fishing did not go so well.

Between some of the fires, rough poles of hickory and oak were stuck in the ground and draped with animal skins stretched between them to block the cool breeze. These baffles also helped keep the smoke from escaping before it served its purpose of drying the fish and meat. The sharp sound of rocks hitting rocks punctuated the air, and splinters flew dangerously as some men made weapons and tools. Stone debris was everywhere underfoot.

Furious growling caught everyone's attention. Two dogs were wrangling over a bony piece of meat, dragged away from the fires when nobody was watching. The fire tender picked up a sharp rock and threw it at the dogs, who momentarily stopped their battle, then quickly resumed it.

* * * * * *

About 5,000 years ago, not far from Anderson, South Carolina, people gathered in such a place and likely engaged in such activities. Their existence came to light during excavations guided by Dean Wood, who named the site Sara's Ridge after the newborn daughter of two of his crew.

Before Sara's Ridge, only three other sites in Georgia and South Carolina had shown any sign of Late Archaic housing, again detected by postmolds. Stains indicating a possible lean-to were found in the Georgia Coastal Plain near the Savannah River, while in South Carolina, a small, D-shaped postmold pattern was uncovered on Hilton Head Island, and potential signs of a structure about three yards long were discovered on Sol Legare Island. But with the possible exception of the Sol Legare site, these other postmolds indicated insubstantial structures, making the find at Sara's Ridge all the more important. (Since completion of the Russell Reservoir studies, archeologists have uncovered evidence of several other Late Archaic houses in Georgia.)

The discoveries on the levee crest at Sara's Ridge might never have happened with different weather and less skillful researchers. The crew gathered on the ridge in February. They were young, mostly in their twenties and thirties, a mix of professional archeologists, students, and interested amateurs. Some of them camped on the site during the dig, with a few trying their hands at fishing in the river and cooking their catches over campfires.

A bulldozer equipped with a sharp blade prepared the site by clearing away surface soil. Then the crew marked the ground into precise meter squares (about a yard wide) by inserting stakes, creating a grid for pinpointing exactly which spot might later yield a discovery. Patiently, workers dug with sharp, flat shovels inside the squares, scooping out about four inches of dirt, then sifting the soil through a screen stretched across a wood frame suspended from saplings. Dirt could pass through the quarter-inch mesh, but most stone artifacts would remain on the screen.

The discolorations would have gone unnoticed by most because postmolds are often only faintly perceptible from surrounding soil. Even trained eyes may overlook the stains if a hot sun hits the soil surface because the heat and exposure can quickly erase them from view. But fortunately, a recent rain had dampened the earth at Sara's Ridge and prevented the postmolds from vanishing before they could be mapped.

Analyzing their findings, the archeologists later speculated that most of the stains represented benches, drying racks for fish and meat, and baffles to block the wind. But the most exciting discovery was a distinct oval of postmolds about seven-and-a-half yards long and five-and-a-half yards wide. This had to be the outline of a shelter.

The Savannah was relatively unique among Southeastern rivers because of its use by migratory fish. In the years before dams blocked their way, the fish swam in the Savannah's waters far inland, deep into the Piedmont, providing a reliable, annual catch in a season when other resources were somewhat scarce.

Fewer than one percent of the eggs produced hatchlings. These and surviving adult fish eventually swam back to the ocean where they stayed until the younger generation was eventually driven back to the river to fulfill its role in the cycle.

While intercepting the fish as they swam past the rocky shoals would have fed many on the levee called Sara's Ridge, capturing them in nets or weirs would have required many cooperative hands. The abundant signs of human habitation at the site suggest such a possibility, although a firm conclusion about the number of people who spent time there is impossible.

Undoubtedly, cooking was one of their major preoccupations. Two possible hearths existed within the oval shelter, along with several more hearths nearby. In the work area on the slope behind the house, evidence of 18 more hearths was uncovered. So many fires suggest that more than one dwelling existed, although there was no definitive sign of others.

Possibly the site was a seasonal gathering place for people who normally lived apart. Because there were no plant-processing tools uncovered, such as pitted rocks, the residents' purpose in choosing the spot perhaps was quite specific-to fish, hunt, or both. While they might have used wood tools to prepare plants, the wood would have disintegrated, leaving no trace. Probably, however, stone tools were preferred. Stone tools, apparently used for processing plants, were found at another site nearby.

An examination of Late Archaic tools at Sara's Ridge disclosed they were made of chert, quartz, and metamorphic rocks. The metamorphic rocks came from at least 25 miles to the south, perhaps even farther. Most major sources of the stone are north of Augusta, Georgia, more than a 50-mile trip for the people of Sara's Ridge if the stone wasn't obtained through trade.

Soapstone slabs with holes drilled in them found at the site, coupled with others from a nearby excavation at Paris Island South, provided new information about the possible function of these artifacts. Previously, most archeologists had assumed these perforated rock slabs were "netsinkers" used, as their name implies, for weighing down fishing nets. But their frequent discovery near hearths at Sara's Ridge and Paris Island South suggests instead that they served as boiling stones.

Like Sara's Ridge, the Late Archaic site at Paris Island South was located on a levee overlooking the Savannah River. The excavation there, organized by Dean Wood and Dan Elliott, indicated that housing was probably once located on the crest of the levee, but any signs of a structure had been washed away from the island. Left intact, however, was an area for making stone tools located on the slope beneath the crest. This work area was conveniently close to probable dwellings but enough removed from the main living area to protect bare feet from sharp rock. The accumulation of stone debris there was much heavier than at Sara's Ridge.

Refuse of all kinds, which archeologists call midden, was once so plentiful at Paris Island South that as it decomposed it stained the earth a faint black, in some spots up to 16 inches thick. The discoloration, although welcome proof of human existence, interfered with the process of locating storage and cooking pits and postmolds. However, a few posts had been driven deep enough beneath the stained soil to confirm the probable location of drying racks or other sorts of frames. The postmolds appeared near charcoal and fire-cracked rock, indicative of hearths.

This prehistoric garbage dump, stretching over 16 yards long and more than six yards wide, yielded as many as 760 flaked stone tools, including spearpoints and knives. The huge volume and variety of artifacts were pivotal because they revealed that people along the river were now staying longer in one place. Paris Island South was either a year-round residence or at least a semi-permanent home for a family or small extended family.

Two pitted stones were also unearthed. Seven pits in one of these stones imply extensive use, perhaps for cracking nuts. There were also three hammerstones.

Other artifacts prompt an intriguing notion that the islanders may have engaged in a cottage industry of sorts, producing the curious perforated soapstone slabs for trading. Paris Island South is not far from a soapstone outcrop just across the river in Elbert County, Georgia, making the raw material readily available. Among the persuasive discoveries favoring the idea of prehistoric commerce was a profusion of these artifacts which were previously considered to be netsinkers. Piecing 424 scattered fragments together in a technique called cross-mending, researchers ended up with 46 complete perforated slabs. Less than half-an-inch thick, they averaged about six-and-a-half inches long and about four inches wide. The slabs were mostly shaped into rectangles or pentagons. Smooth surfaces were predominant, but several had thin lines or small notches etched in the surfaces, perhaps for decoration.

More findings in support of a possible manufacturing center were 23 stone drills that were once attached to wooden handles. The drill points matched the sizes of holes in the slabs. There were also some partially drilled slabs and a grinding stone used to shape the slabs, along with many stone scrapers, once attached to handles, used to smooth the soapstone.

There was probably a receptive market for the prepared slabs. Studies by Dan Elliott and Kenneth Sassaman have shown that people living farther away from soapstone sources tried to salvage broken perforated slabs by drilling new holes in them, indicating they valued the artifacts. But Paris Island South residents apparently never bothered to redrill holes, maybe because more soapstone was so easy to get.

Other illustrations of stone working on the island displayed the first hint in the Russell studies of another, distinctively human trait-the desire to ornament the body. A highly polished stone bead of reddish brown was found, as well as two smooth, small slabs of stone, possibly pendants or gaming pieces.

Still more artifacts, stone cobbles with abraded edges, suggest use in forming wood tools. Also found were stone abraders with grooves which were used to sharpen bone, antlers, or wood into points for use as fish-hooks, chisels, awls, and spearpoints.

The Rocky River excavation, located about a half mile from the Savannah River in Abbeville County, South Carolina, featured an expansive stone working area on the back slope of a slight ridge. This work place was also part of a thick prehistoric garbage dump or midden, a strong sign of extensive human use.

Researchers Charles Cantley and Andrea Lee Novick found that all stages of spearpoint production took place at Rocky River. Interestingly, however, larger flakes from the beginning stages of spearpoint manufacturing showed up mainly on the outskirts of the midden. Perhaps toolmakers periodically swept their primary work area clean.

Maybe those within the Russell area influenced others farther south in their decision to stay for longer periods along the Savannah River, a shift destined to have dramatic consequences. On Stallings Island near Augusta, Georgia, and at neighboring places, prehistoric people were about to make a discovery that would change life throughout much of the eastern United States.