SCHENECTADY — More than 300 years ago, someone carried a load of white cobblestones from the river, carefully stacked them on Jan Roeloffsen’s property and firmly cemented them into place with dirt.

Or at least, that’s what archaeologists now believe happened.

After speculating all winter as to whether the rocks they found were dragged here by a glacier or stacked by human hands, one day of old-school geology and fancy new radar technology appears to have solved the mystery.

In fact, New York State Museum archaeologist Julieann Van Nest didn’t even need her colleague’s radar to determine that the rock formation was probably man-made. If it is, it may be the oldest house foundation ever uncovered in the city’s historic Stockade section.

Van Nest spent several hours on Friday carefully studying the rocks at the bottom of a five-foot-deep archaeological dig while museum geologist Andy Kozlowski slowly maneuvered his $23,000 radar up and down the land nearby.

As he progressed, inch by inch, toward the stones, she was in the hole scraping at loose dirt and peering at the soil above the cobbles.

The answer, she said, was clear: the stones were placed there by hand.

If the stones had been carried by a river or a glacier, they would be stacked by size, with the largest dropping out first.

“You don’t often see cobblestones climbing above fine-grained material,” Van Nest said. “Also, you don’t have all the sizes. All the stones are about the same size.”

The stones are so large that they could only be carried by a strong river, she added. Typical traces of a riverbed — such as a gravel bar — are nowhere to be found.

“It’s hard to move rocks that big,” she said. “You’d have a gravel bar, but you don’t. This is isolated.”

She glanced back at Kozlowski, who was still painstakingly mapping out the back of the yard with his radar. He was at least half an hour’s distance from the stones. She climbed into the pit and started analyzing the dirt.

“The original soil is missing around the cobblestones,” Van Nest said. “We can infer it was taken away, possibly dug up. If you’re moving dirt, there’s some kind of earth-moving for construction.”

If the stones were dropped by a natural source, she added, there would have been no dirt pressed between each stone.

“That’s another argument for it being manmade. In nature, it is class on class,” she said. “I think it’s humanly made.”

Ron Kingsley, the archaeologist who discovered the stones and had believed they were most likely carried by glacier or flood, was impressed by her arguments. His students in the Community Archaeology Program at Schenectady County Community College, who had been rooting for a determination of manmade, were delighted.

They pointed out that although there were no artifacts in the soil directly above the cobblestones, the remnants of a burnt post are just two inches away, at the same soil depth. Kingsley thinks that post was part of the original building on the site, which was burned in 1690 during a raid by the French and Native Americans.

Now that it appears the cobblestones were placed there at the same time, he said the stones could have been used to support the corners of the building. It clearly did not have a full foundation.

“They may represent piles of cobbles that were supports for the sills of a building,” he said.

If that’s the case, his students said, the other piles may be under the existing house. But Kingsley was holding out hope that the radar would point him to at least one more pile.

Finally Kozlowski got to the cobblestones. Kingsley’s students laid out a new grid and Kozlowski began to walk his radar over the precarious surface, trying not to fall into the pit or trip over the rocks and tree roots next to it.

On his first walk-through, he was so busy watching where he was going that he couldn’t look at his computer screen. But when he checked the results, it was clear there was something underground. Lots of somethings.

“We’re going to do some more excavation,” Kingsley said firmly.

Kozlowski doesn’t yet know if the anomalies found by the radar are at the same depth as the cobblestones. He’ll determine that when he builds a map back at his lab.

But he agreed with the consensus: the cobblestones did not get there by chance.

“It’s very probable it’s manmade,” Kingsley said. “Further investigation of that will be taken this summer.”

He has a long way to dig in a very short time. This year is slated to be the last at the excavation site, which now extends through most of the side yard of the Yates House on Union Street.

It will be the seventh summer of dust and crowds of strangers for homeowner Paul Miodzianowski. When he opened his land to the archaeologists in 2002, he didn’t expect that a swimming-pool sized hole would soon develop right next to his door, but he hasn’t complained.

He also hasn’t said a word about the unsightly and ever-growing pile of dirt next to his lawn chairs. But Miodzianowski is looking forward to the end of the endeavor.

“We are ready to get rid of it. It’s time to get the yard back,” he said. “We thought we were going to get it back last year, but then they found the cobblestones.”

He agreed to endure just a few more months, and Kingsley intends to make the most of them to research the discovery that was almost found too late.

“We have probably one of the richest archaeological sites in the Stockade right here,” Kingsley said. “It’s been very exciting.”