An umayyad-period aqueduct for the irrigation of farmland south of Ramla

A 150 m long section of an aqueduct was exposed approximately 100 m south of Ramla (map ref. NIG 18650-80/64725-40; OIG 13650-80/14725-40), aligned northeast-southwest. The aqueduct was discovered in a preliminary survey and excavated in a salvage excavation prior to the construction of Route 431. Parts of the aqueduct had been damaged by plowing in the twentieth century; therefore, in most places the state of preservation was poor, and only the base of the aqueduct survived. Excavations were carried out in four sections (T1-T4; Fig. 1; Plans 1, 2). The aqueduct had a U-shaped cross section. To enable the construction of the aqueduct, a foundation trench was dug in the hamra soil (L106; c. 2 m wide), and a layer of small limestone stones was laid in it. Above this foundation, the aqueduct walls (0.3 m wide, 0.35 m high) were constructed of medium-sized stones (0.1 × 0.2 m), bonded with gray mortar. The walls, base and foor of the aqueduct were plastered with three superposed layers of plaster (L104, L107): a layer (4 cm thick) of gray mortar, mixed with pieces of charcoal, topped by a layer of broken potsherds; and above it-a coating of hydraulic plaster (0.2 m thick). A thin layer of travertine had built up over the hydraulic plaster, testifying to water fowing through the aqueduct in ancient times. One covering stone was preserved in situ (L300; Plan 2; Fig. 3) in Section T3 (30.5 m long; Plan 2; Fig. 2). This covering stone indicates that the entire aqueduct was originally covered in a similar manner. Further along this section (T3), two water outlets were revealed in the southeastern side of the aqueduct, one of them leading to a secondary section of the aqueduct (L114; Plan 2; Fig. 2). At the end of this outlet, a square stone trough was installed, made of a single block of limestone (L203). The sides of the trough were plastered with hydraulic plaster. A shallow trench with a V-shaped cross section (0.1 × 0.2 m) was cut into the northwestern side of the stone basin, through which water had fowed from the aqueduct into the basin. The connection between the aqueduct and the basin was not preserved. Another outlet was revealed approximately 4.5 m south of this branch (L301), containing a saqiye jar in secondary use, in the base of which an opening had been made to enable it to be used as a clay pipe (Plan 2: Section 2-2). The vessel had been incorporated into the side of the channel, forming an outlet for water from the channel. The broken potsherds, which were incorporated in the plaster lining the aqueduct, included Gaza-type jars (Fig. 4:1-4). These vessels have been dated to the sixth-seventh centuries CE, giving a suggested date for the aqueduct's construction. Pottery vessels (Fig. 4:5-8) from the fll inside the aqueduct were mostly attributed to its fnal phase of use, during the eighth century CE. The excavated section appears to have been part of the aqueduct's water conduit system, built at the beginning of the Umayyad period. Similar aqueducts have been recorded by Yosef Porath at Khirbat al-Mefjar (eighth century CE) and Paza'el (Umayyad period). The excavated aqueduct may have been used to irrigate felds of fax, and the stone basin may either have been used to distribute water or as a trough for soaking the fax. The aqueduct may, however, already have been built at the end of the Byzantine period, and have provided water for a farm belonging to the rural hinterland of Lod. Based on the ceramic fnds, the aqueduct probably continued to be in use until the eighth century CE.