Working in an archaeological trench in Romania, I was on my knees for eight hours a day. We dug in teaspoon-sized scoops so that we didn't miss any small coins or bone fragments. My hands ached, and dirt stuck to my sweaty skin. Every once in a while, we had to chase off a stray cow or horse.

My fellow excavators and I were digging up the long-buried Roman city of Porolissum. I had signed up for a foreign archaeological field school with no training or experience. I was with 20 strangers in a country where horses and buggies are as common as automobiles, and where the language was foreign to me, ready to spend a month doing hard labor in the sweltering heat - for fun. I received no pay, no credit and no letter of reference.

But we uncovered structures that proved the city was bustling long after it was believed abandoned.

We were briefed on the town before we began digging. Porolissum was conquered by the Romans in 106 A.D., and it defended the Roman Empire at its very edge. Over time it expanded into a civilian town.

The Romans withdrew in 271 A.D., and many scholars believe that once they abandoned the area, everyone left, including the indigenous people. But the Romanian director of our dig, Alexandru Matei, believed that people might have continued living there afterward.

We dug in Porolissum's forum, which is similar to a present-day town center. It held government buildings, shops and possibly temples. These buildings were set up in a rectangle around a central courtyard. We were attempting to define the dimensions of the forum and get a better understanding of life at Porolissum.

Early in the dig, in the first trench that we opened, we found a rubble layer composed of stones and roof tiles, indicating that a building had once stood there.

After we documented the rubble layer, we removed it to see what was beneath. We found what we were looking for: an intact wall that, we guessed, outlined the courtyard of the forum.

As we dug deeper, we found another wall at a perfect right angle to the original wall.

In the corner of these two walls, we found a pile of domestic animal bones. The building rubble we found earlier must have constituted the remains of the town butcher shop, which would have been located along the courtyard of the forum.

By analyzing those courtyard walls, we eventually proved correct Matei's theory about habitation after the Roman withdrawal. It was obvious that buildings had been converted for use other than their original purpose.

The walls of these buildings were not at perfect right angles to the original walls, which they would have been if the Romans had built them. They were also built in a different style.

The Romans used large stones and mortar to build walls, but people from later periods also used cement and small bits of marble.

Finding these two types of walls conjoined shows that the later people used the Roman walls to help in their construction, but changed the building plans for their own uses.

We also determined that most of the trading was done locally, not with major centers such as Rome or North Africa. This was rare because almost all of the Roman provinces traded solely with Roman commerce centers.

Throughout our dig, we found many ceramics that, according to past archeological findings, were fired in Romania.

At one point, one of our Romanian workers tossed a piece of this ceramic into a pile with a shovelful of dirt.

We saw the shard of orange amid the flying dirt, and we panicked. Two of us climbed out of the trench and raced over to the pile to dig through it and look for this artifact.

We found it, but from then on we kept a closer eye on the workers.