Gadara PDF Print E-mail
Written by Micah Key and Stephen Langfur
Article Index
Water from afar
"Athens of the East"
Demons into Swine

Water from afar

Sources: English. German.

In 2004, engineer Matthias Döring was summoned by German archaeologists to a Bronze Age site in Jordan named Zeraqon, about 15 miles ESE of Gadara, to investigate a stepped shaft that led deep into the earth. Sixty meters below the surface was a section of plastered tunnel. The workmanship and the waterproof plaster indicated the Roman period. Exploring the area nearby, Döring found more such shafts, leading to horizontal sections of tunnel that were all at the same level. There was no large city from the Roman era in the vicinity, so Döring hypothesized that these sections must have belonged to a long-range system. Knowing that the Romans liked to keep the gradient of their aqueducts (the angle at which the water descends by force of gravity) down to around 1%, he took a topographical map and saw three cities that might be candidates: Adraa on the King's Highway and Abila and Gadara on the link road. He sketched out a possible line and went with a team in search of more shafts. Within weeks they had discovered 30. By the end of their 2006 campaign, the number topped 300, many of which had been filled. (Since the shafts are cut at distances of 30 to 50 yards, the final figure will probably exceed 2900.) The shafts led to three separate underground tunnels. The longest stretched from Wadi Shellala, a tributary of the Yarmuk, to Gadara. Its length is 58 miles.

To give an idea what this means: The tunneling was through limestone. It had to be kept around the height of a man, so that the workers would have air. A laborer wielding an iron hammer and chisel could advance four inches per day. That's 58 miles at 4 inches per day. At the bottom of each shaft, two laborers would have chipped in opposite directions, with the hope of meeting their colleagues from shafts about 50 yards distant—not bypassing them, not cutting too high or low.

Lines left by the water in the plaster indicate its height in the tunnel. In a section near Abila, the lines varied between 50 cm and 80 cm. According to Döring, this would indicate a flow of 40,000 to 60,000 cubic meters per day. Where would so much water have come from? The source must have been a lake (later a swamp, now dry) some 37 miles north of the first tunnel, near the road to Damascus, at Dille in Syria today. Portions of an above-ground aqueduct have been found, leading from there to Adraa. The duct would have incorporated springs en route.

We have, then, the following picture. First, an above-ground duct from Dille to Adraa. Then the duct curves west toward Abila and Gadara. It must somehow cross Wadi Shellala, a chasm 600 feet deep. Moreover, the land on the east side of the wadi is 40 m. higher than on the west, but the gradient must be kept moderate. (Too steep a drop would destroy any aqueduct.) Tunneling began, therefore, east of the wadi and south to a point where it was shallower. Here the Romans built a bridge 20 meters high, 100 meters long. On the west side began the 58-mile tunnel, winding often to skirt valleys. The total length from Dille to Gadara: 105 miles.

The great aqueduct system, 105 miles long

The bridge at Wadi Shellala, a tributary of the Yarmuk

Although the Romans measured carefully, using levels and plumb lines, there were mistakes. Sometimes the chippers passed each other and "then"—"then" meaning months of work— had to make corrections. Sometimes they didn't get the levels right. Toward Gadara, the accumulation of errors resulted in failure. The last phase of the tunnel was never plastered, an indication that it wasn't used. But the usable part had reached the spring of Ain Turab, and it is possible that the Roman engineers may have led the new duct's water into the older, deeper duct that supplied Gadara from here.

The work cannot yet be dated, except to say "Roman." Radiocarbon testing suggests the 2nd and 3d centuries AD. However, because the proportion of C-14 in the atmosphere varies at different places and times, the dates are not firm until calibrated by means of dendrochronology, still patchy for the Middle East before 362 AD