Area A (Figs. 3, 4)
A section of a massive wall (W144; exposed length 10 m) built of rectangular kurkar ashlars, oriented north–south, was exposed. Two or three carefully built courses were preserved in the northern part of the wall (Fig. 5); the collapse of the wall’s upper courses was discovered above them. A single course of stones constituting part of the wall’s foundation was all that was preserved of its southern part (Fig. 6). The narrow sides of the stones were plastered, perhaps indicating that they had been dismantled from a building nearby and were in secondary use. Whitish mortar was noted on the underside of the stones in the western side of the wall. The mortar was used to stabilize the stones in the soil and prevent water from penetrating into the wall’s base. Running the entire length of the wall, to its west, was a level of light brown sandy soil (L129), at a higher level than the foundation course of the wall and clay soil (L115, L146) was discovered to its east. A few fragments of mostly abraded pottery vessels were found; only one of the sherds, from L129, was identified: the rim of a Late Roman-period cooking pot (second–third centuries CE) (Fig. 7:1). At some point, perhaps during the Byzantine period, the wall ceased use and most of its stones were robbed. A field wall (W131; Fig. 3: Section 2–2) exposed above W144 was built haphazardly of stones removed from W144 and small fieldstones. In the northern part of the area, c. 4 m north of W131, was a section of another wall (W141), also poorly constructed. Wall 141 was oriented along the same axis as W131 and it therefore may be the northern continuation of that wall. The continuation of W144 and W131 to the south was severed, probably a result of later agricultural activity. A small fragment of a slightly concave stone object (Fig. 7:2) with chisel marks on one of its sides was found below the surface. This fragment probably belonged to an ossuary lid.
 
Area C1 (Figs. 8, 9)
In five of the eight squares, an irrigation system dating to the Roman–Byzantine periods (second–fourth centuries CE) was revealed, built in dark clay soil that included a section of a saqiye (water-wheel) well (W359), pillar remains of a saqiye installation (W261, W377), a foundation of a water channel (W326) and a square water reservoir. The complex was poorly preserved.
 
Saqiye Well. The northeastern part of a wall of an elliptical saqiye well (W359; inner diam. 2.7 m, excavated depth 2.5 m; Fig. 10) was discovered. The wall was built of an outer face of partially worked medium-sized kurkar stones and an inner face of ashlars (0.25 × 0.40 × 0.60 m) with fill consisting of small fieldstones bonded with mortar. The upper part of the wall did not survive. Only two stone courses were uncovered from the ashlar inner face of the wall, and apparently, the rest of the ashlar stones  were robbed after the well ceased use. Small and medium-sized fieldstones (L362; Fig. 8: Section 1–1) and dark damp soil containing fragments of saqiye jars (L375; below) were discovered in the collapse inside the well. Excavation in the well was suspended due to technical limitations.
 
Remains of the Saqiye (Water-Wheel) Installation. Remains of stone pillars of a saqiye (water-wheel) installation (L361, L377) were discovered north and south of the well; the installation itself was probably made of wood and did not survive. A single course of two roughly hewn kurkar stones (0.13 × 0.40 × 0.45 m) was preserved from the northern pillar (L361). A pavement of smoothed kurkar stones (L370) was set on a bedding of kurkar gravel mixed with plaster and soil abutted the western side of this pillar. On the floor was a broken, crumbling jar, in situ (L378). A layer of clay that resembled mud-brick (L383) was exposed below the floor bedding; neither the origin nor the use of the layer were ascertained. Only one hewn kurkar stone was preserved from the southern pillar (L377). A bedding of kurkar gravel mixed with light gray plaster and sections of pinkish plaster, shells and pottery sherds (L356) abutted the pillar from the east. Another section of this bedding (L376) was discovered southeast of Pillar 377, and it adjoined the western wall of the water reservoir (W320). Fragments of saqiye jars and other pottery vessels (L347, L350; below) were found above Foundations 356 and 376.
 
Water Channel Foundation (Fig. 11). The foundation of a water channel (W326; length 2 m) built of four large, roughly hewn kurkar stones, was exposed between the wall of the saqiye well and the reservoir. The foundation level of the channel was one meter above W359 of the well. The foundation collapsed toward the west, in the direction of the well, and therefore its western part was lower than its eastern part. Apparently, the collapse of the foundation to the west exerted pressure on the eastern side of the well’s wall, which toppled into the well. The two eastern foundation stones of the channel (0.5 × 0.5 m each) were discovered in situ next to the wall of the water reservoir (W320). The water raised from the saqiye well by the water-wheel flowed into a channel that conveyed it to the reservoir. A habitation level (L323) exposed north of the channel’s foundation included potsherds, saqiye jars, shells, plaster and a Roman imperial coin (IAA 156155) from the last quarter of the third century or the first quarter of the fourth century CE.
 
Reservoir. A square reservoir (6 × 6 m; Fig. 12) was exposed. The floor (L315), paved with small, white-plastered fieldstones (thickness 5 cm), was cracked and poorly preserved. The reservoir was enclosed within four walls (W309, W316, W320, W330; width 0.6 m) built of small fieldstones (0.12 × 0.16 m) bonded with white mortar; they were preserved to a height of 0.2–0.3 m. Wall 330 was broken in several places and the corner it formed with W316 had collapsed to the northwest, causing the floor in that spot to settle (Fig. 13). A probe (L327) conducted west of W320 revealed the wall foundation at a depth of 0.3 m; it was built of small fieldstones, similar in size to the stones of the walls, but without mortar (Fig. 14). The walls were treated with white plaster; their inner face was smooth while their exterior was rough. The corners of the walls were thick and straight on their inside, probably to strengthen them against water pressure (E. Ayalon, pers. comm.), rendering the reservoir an octagonal appearance. A circular profile (rolka) at the joint between the walls of the reservoir and Floor 315 was intended to improve the seal and strengthen the contact between the sides of the well and the floor of the reservoir. A pillar base built of small fieldstones bonded with white mortar was exposed on the outer part of Walls 316, 320 and 330. Two pillar bases and a channel for a water outlet (L369) were revealed in W309. The plaster coating the channel was identical to that of the floor of the reservoir. No irrigation channels were exposed outside the reservoir and it therefore seems that these channels had been dug in the ground and were not preserved. A square settling pit (L384; 0.48 × 0.64 m, depth c. 0.65 m; Figs. 15, 16), also white-plastered, was installed in the center of the pool. In a later phase, the pit was coated with a layer of light pink hydraulic plaster, characteristic of the Byzantine period; grooves were incised in the surface of the earlier plaster to ensure that the new layer would firmly adhere to the sides of the pit. A repair in the southern part of the reservoir’s floor may also have been carried out in this phase. The eastern side of the settling pit protruded slightly above the reservoir floor and was covered by the light pink plaster.
After the reservoir went out of use, probably at the beginning of the Byzantine period, fragments of the plaster that lined the settling pit, especially its southern wall, fell to the bottom of the pit. A layer of burnt soil (L352) containing animal bones (Marom, below) and fragments of pottery vessels above the fragments of plaster indicates that the pit was used for cooking during this phase.
In the last phase, the settling pit was filled with small fieldstones without mortar and was leveled to that of the reservoir’s floor. Burned animal bones and pottery sherds were discovered above the fieldstones, indicating another cooking phase here (L344).
A coin from the Mamluk period (fourteenth–fifteenth centuries CE, IAA 156156) was discovered on the surface, 10 m southwest of the reservoir.
 
Ceramic finds, mainly dating to the Late Roman–Byzantine periods (second–fourth centuries CE) and some dating to the Byzantine period (fourth–sixth centuries CE), were discovered in the irrigation complex. The finds from the second–fourth centuries CE included a cooking pot with a gutter rim (Fig. 17:1) dating to the second–third centuries CE, two lids (Fig. 17:2, 3), a jar rim (Fig. 17:4) and a juglet rim (Fig. 17:5), as well as eighteen rims of saqiye jars, most of them found in the well, twelve of which were drawn (Fig. 17:6–17). Ayalon (2000:221–225) distinguishes between saqiye jars from the middle of the Roman period (late second–early third centuries CE) and those of the Late Roman and Early Byzantine periods (fourth–fifth centuries CE). Apparently, the saqiye jars found in the excavation belonged to the type from the fourth–fifth centuries CE, carinated toward the bottom of the body, like vessels discovered in nearby Tel Qasile. The finds from the fourth–sixth centuries included a small bowl(?) (Fig. 18:1), two cooking pots (Fig. 18:2, 3), a lid (Fig. 18:4) and two Gaza jars (Fig. 18:5, 6), discovered above the reservoir floor after it ceased use, which were common in the fourth–sixth centuries CE.
 
Flint Assemblage
Maya Oron
 
Knapped flint items (Table 1; N=90), many of which were worn and covered with thick patina, were discovered in Areas A and C, apparently transported there and not discovered in situ. Most of the assemblage (80%) consists of knapping debitage, mainly flakes and non-diagnostic primary elements, several items knapped using Levallois technique, characteristic of the Middle Paleolithic period, and core-trimming elements deriving from bladelet production, which perhaps may be attributed to the Epipaleolithic period. The assemblage also includes twelve cores (one is broken)—mainly typical Epipaleolithic bladelet cores (N=9; Fig. 19:1–3) and several cores for non-diagnostic flakes (N=2; Fig. 19:4), tools (N=6), including ad hoc items (retouched flakes and a retouched blade), a scraper (Fig. 19:5) and two burins.
Strata that yielded flint tools characteristic of the Kebaran and Geometric Kebaran cultures of the Epipaleolithic period were exposed during an excavation conducted in 1996 at a nearby site on the banks of Nahal Ahiyya (Marder et al. 2001). Most of the flint items in the assemblage, including the bladelet cores, the core trimming elements, the scraper and the burins, presumably come from these strata. The origin of the other items, including the Levallois flakes, is unknown.
 
Table 1. Flint Items
Type
N
%
Chunks
6
6.7
Chips
10
11.1
Primary elements
18
20
Flakes
27
30
Blades
5
5.6
Core trimmings
3
3.3
Levallois flakes
3
3.3
Total debitage
72
80
Cores
12
13.3
Tools
6
6.7
Total
90
100
 
Archaeozoology
Nimrod Marom
 
Several animal bones, probably from the Byzantine period, representing activity carried out in the reservoir after it was no longer in use, were discovered in the fill inside the reservoir in Area C1. The bones (N=7) were mainly of cattle (N=5; L339, L351, L352): two radii fragments (a fused distal end with a knife-incised cut mark above it), a metapodial fragment and a permanent tooth from an upper jaw. These bones represent one or more individuals. A tooth of a sheep/goat (L342) and a fragment of a dog’s jaw (L402) were also discovered. The way the cattle bones were broken shows that most of them (four out of five) were split open when fresh; only one bone has a break that was caused when the bone dried out. Except for the dog’s jaw, all the bones show no signs of weathering due to exposure to climatic conditions and the sun—evidence indicative of quick burial. The butchering signs on the cattle bone and the breaking of the fresh bones seem to infer that these bones were remains of a meal discarded into the pit.
 
Remains of a building from the Late Roman period that were mostly robbed were discovered in Area A, apparently built of stones in secondary use taken from a nearby structure. After the wall ceased use and was plundered (during the Byzantine period?), a field wall was constructed above it that was apparently meant to delineate an agricultural area. An irrigation system, including a saqiye well dating to the Roman–Byzantine periods (second–fourth centuries CE), was exposed in Area C1. Numerous saqiye wells operated in the country from the Roman period onward, but only a few have been studied. To date, two such wells from the Late Roman period have been published, both rock-hewn (Milo 2001: Table 2): at Khirbet Ibreiktas (Kletter and Rapuano 1998) and at Kafr Manda (HA 1962). The excavators of the well at Kafr Manda dated the installation to the Mamluk period, but after examining the saqiye jars its date was revised to the Late Roman period (Ayalon 2000:223). Saqiye wells dating to the Byzantine period were discovered in Kefar Sava (Ayalon 1998:109–110), Yavne-Yam and Tel Ashdod (Ayalon and Drey 2005:245). A saqiye well from the Byzantine and Early Islamic periods was recently discovered at Ramat Ha-Hayyal (Haddad, Itach and Shiff 2017). Wells dating to the Early Islamic period were also recently exposed near Ramla (A. Gorzalczany, pers. comm.). The most common pottery vessels in the excavation are saqiye jars used in the installation above the well to draw water. The water collected from the well was spilled into a channel that conveyed it to a nearby storage pool. This system was utilized by the public to irrigate agricultural crops and provide drinking water. This water system was situated only 300 m northwest of the cemetery from the first–fourth centuries CE. Kaplan (1964:163), who excavated the cemetery, suggested that it belonged to the nearest settlement located northeast of Khirbat el-Aurah, which has not yet been excavated. It is also possible that the well belonged to the residents of this settlement. The excavation revealed that the water system ceased use at the beginning of the Byzantine period, possibly due to a lack of groundwater and perhaps due to the neglect of the system. Apparently, after the system ceased use, the water reservoir, especially the settling pit in it, were used by passersby for burning and cooking. The poor preservation of the reservoir’s floor with its many wide cracks that settled in several places and the wall that shifted, as well as the collapse of the water channel’s foundation wall, may have stemmed from the instability of the clay soil and its tendency to swell during arid weather, and possibly from the earthquake of 363 CE.