Prior to the excavation, a preliminary survey was performed and the plan of the camp, published by Negev, was updated. Following the survey, it was possible to discern the outline of the casemate rooms, living quarters and towers, so that each room was defined as an excavation unit. The rest of the area was excavated in squares. All the soil from the excavation – from surface to bedrock – was sifted. The excavation areas were systematically examined with a metal detector. To evaluate the destruction and the strata in each of the rooms and in the streets, balks with sections, from surface to the bottom of the excavation, were left in place.
The army camp is a square structure (c. 100 × 100 m) surrounded by a wall (thickness 1.8 m) that was partially preserved 3 m high (Fig. 1). Twelve towers jut out from the wall of the camp: four corner towers (length 9 m, width 6 m) and eight intermediate towers (length 5 m, width 6 m, wall thickness 1.2–1.5 m), including a pair of towers at the main gate. Thirty-two casemate rooms were constructed along the inside of the camp’s wall.
Four narrow corridors were built in the corners of the camp. Each one leads from the peripheral road to an anteroom, which in turn leads to the ground floor of one of the corner towers. Seven of the casemate rooms served as anterooms for the intermediate towers; they are slightly smaller than the rest of the casemate rooms and lack the pilasters that appear in all the casemate rooms. The camp has three gates. The main gate is in the center of the eastern wall; it is fortified with a pair of outer towers constructed adjacent to the gate. Secondary gates were built in the middle of the southern and northern walls. A guard room was built alongside of each gate. Eight residential blocks – barrack units are built inside the camp, including five that have ten rooms and three that consist of six rooms. These residential blocks are separated from each other by a network of roads, oriented north–south and east–west. Four of the residential units were excavated: two of the ten rooms (Buildings B and D) and two of six rooms (Buildings A and C). Of the total sixty-eight barrack rooms, thirty-two rooms of uniform size were excavated (each c. 20 sq m). The roof of the casemate rooms was reached by way of four staircases that were built in the corners of the camp, next to the casemate rooms. The main road of the camp, the via praetorian (width 10 m), was oriented east–west and led from the eastern gate to the headquarters building (principia) that was erected next to the western wall of the camp. Another main road aligned north–south, via principalis (width 8 m), led from the southern gate to the northern gate and intersected the via praetorian in the center of the camp.
The headquarters building (principia) is rectangular (length 32 m, width 16 m), with five rooms and a courtyard in front of them. The rooms next to the wall were rectangular. A plastered stone ledge (width 1.5 m, height c. 0.4 m) was built at the western end of each room. The commander’s building (praetorium) consisted of seven different size rooms and a small rectangular courtyard to their east; it was built south of and adjacent to the principia.
Most of the potsherds found on the floors and below them belonged to Nabataean and Herodian wares, which were very common to the first and second centuries CE. The glass vessels dated to the first century CE and none had postdated this period. Most of the coins on the floor levels and below them were Nabataean and Herodian.
A fragment of a Nabataean inscription in a tabula ansata was found in secondary use in one of the walls of the principia. The inscription has not yet been deciphered.
A few pottery vessels from the Late Roman period were discovered and coins of the emperors Diocletian and Constantine (end of the third–beginning of the fourth centuries CE) were scattered throughout the building; these are probably from a later phase when its stones were dismantled after the structure was no longer used.
Based on the finds from the floors of the rooms, the construction of the camp can be dated to the end of the first–beginning of the second centuries CE, probably after the Romans annexed the Nabataean kingdom in 106 CE. The camp seems to have existed for only a short period and was abandoned sometime in the second century or the beginning of the third century CE.
During the Late Roman period the walls of the deserted camp were dismantled and its stones were used to build the Late Roman settlement. The wall remains surviving at the site were destroyed in the earthquake that struck in 363 CE and sealed its floors.
An analysis of the camp’s plan and its internal division shows that the camp was a base for a cohort of the Roman army.