The site of Kuweikat (c. 30 dunams; 56 m asl; Fig. 1) lies on a low chalk outcrop at the eastern margins of the ʽAkko coastal plain. The western part of the site lies within the bounds of Kibbutz Bet Ha-ʽEmeq; and the eastern part is mostly covered by a planted forest and is cut through from north to south by Road 70. The Arab village of Kuweikat (Kuwaykat) occupied the site until it was abandoned in 1948. The village, with its well and spring, was described by nineteenth-century travelers and researchers (Van de Velde 1854:268–269; Guérin 1880:7, 29; Conder and Kitchener 1881:147, 154). Based on the village’s name, the site was identified with Crusader-period Cochetum or Coquetum, recorded in 1129 CE as belonging to the Hospitaller Order (Röhricht 1893:1135, 1187).

The area was surveyed as part of the Archaeological Survey of Israel, Map of ʽAmqa, and whilst ancient settlement remains are meager, the chalk outcrops contain many rock-hewn tombs dated to the Persian, Hellenistic and Roman periods (Frankel and Getzov 2012: Sites 176, 177), as well as many rock-hewn agricultural processing installations and quarries (Frankel and Getzov 2012: Sites 152, 153, 154). In addition, installations, including rock-hewn winepresses, quarries, and Roman-period kokhim burial caves, were recorded to the east of Road 70 (Smithline 1997; Frankel and Getzov 2012: Sites 155, 158, 180, 181).

About 500 m southeast of Kuweikat lies Tel ʽEmeq (Tell Mimas), also located alongside Road 70. This is a large multi-period hill site, where pottery and other finds from the Late Bronze and Iron Ages, as well as the Persian, Hellenistic and Roman periods, were identified (Frankel and Getzov 2012: Site 178). During a recent excavation, rock-cut tombs and structures dated to the Roman period and settlement remains dated to the late Byzantine and Early Islamic periods were exposed (Permit No. A-7446; Y. Shalev, pers. comm.). It is probable that the area of Kuweikat served as a burial ground within the peripheral margins of the contemporary settlements at Tel ʽEmeq (Frankel et al. 2001:13, Sites 34, 35).

 
The present excavation at Kuweikat comprised a long row of 22 single squares (Sqs 1–22; 4 × 200 m; c. 800 sq m), extending alongside the western flank of Road 70 (Figs. 2, 3). Along the entire excavation strip, the uneven chalk bedrock surface was exposed just below ground level. The principal elements exposed were three hewn tombs (Tombs 1–3), small clusters of stone quarries and a segment of a rock-hewn water channel, apparently part of the Hellenistic aqueduct leading to ʽAkko-Ptolemais. In addition, scant remains of an early twentieth-century building and an undatable hewn pit were unearthed.
 
Hewn Tombs
Tomb 1 (Sq 21; Figs. 4, 5), oriented east–west, comprised a deep, rectangular shaft with vertical walls (L129; 1.1 × 2.1 m, depth 2 m) and a steep step on the eastern end for access into the shaft. A low, arched entrance hewn in the western wall of the shaft led into a small, rectangular burial chamber, or loculus, with a low flat ceiling (L131; 0.9 × 1.8 m, height 1.1 m). The chamber’s dimensions suggest that it was intended for a single primary burial. The tomb was devoid of finds and was probably looted in antiquity. The shaft was blocked with ashlar blocks (Fig. 6), probably damaged blocks from the nearby quarry.
 
Tomb 2 (Sq 4; Figs. 7, 8) comprised a small, narrow north–south corridor, with two burial chambers or loculi at either end. The original access from the bedrock surface down into the tomb was not evident. The southern burial chamber (L130; 0.8 × 1.5 m, height 1.15 m) had an intact vaulted ceiling, whilst the ceiling of the central corridor (L126; 0.9 × 2.0 m) and of the northern chamber (L123; 1.2 × 2.2 m) was no longer extant, presumably removed by ancient or recent rock-cutting activities. The bedrock floor was uneven, with a low step ascending from the corridor floor into each of the two burial chambers. The entrances from the corridor into the chambers were blocked off by a combination of two or three stone blocks, carefully wedged with smaller stones. Here too, the small dimensions of the burial chambers suggest that they were intended for individual primary burials. The only artifact retrieved in Tomb 2 was a small fusiform unguentarium or perfume bottle (Fig. 9:1), found at the far end of the southern chamber (L130), probably overlooked by the tomb robbers. This fine, imported vessel with thin walls is similar to unguentaria from Dor, which were dated there to the Hellenistic period (mid-second century BCE; Guz-Zilberstein 1995:306, Fig. 6.27:1–4).
 
Tomb 3 (Sq 6; Figs. 10, 11) comprised a rectangular shaft and a burial chamber, both oriented east–west. The deep rectangular shaft (L127, L132; 1.2 × 2.8 m, depth 2.9 m) had vertical walls and exhibited a staircase of eight very narrow steps hewn out of the southern wall (Figs. 10: Section 2–2; 12). An arched entrance, with an ashlar stone threshold or step, was hewn in the western wall and led into the burial chamber (L133; Figs. 10: Section 3–3; 13). A couple of large hewn stones partially blocked the entrance to the chamber, and numerous medium-sized to large fieldstones were piled up against these large stones in the shaft. A few Hellenistic-period pottery sherds were found on the bedrock floor of the shaft (L132). These included the raised ring-base of a fine-ware small bowl with an incurved rim (Fig. 9:2). Similar bowls at Dor were dated to the third–second centuries BCE (Guz-Zilberstein 1995:306, Fig. 6.1:3, 4).
The shaft was found full of earth overlying the pile of stones. About halfway down were burial remains (L127; elevation 62.76 m asl): a storage jar base containing some fragmentary human bones, a cooking pot and an intact unguentarium near the shaft’s eastern wall (Fig. 14); and an amphora base containing fragments of a glass bowl (Fig. 9:6; below) and several small unguentarium sherds beside the southern wall of the shaft. These fragmentary remains belonged to either one or two burials that must have been secondarily interred after the burial chamber was blocked and the shaft was partially filled in. The intact unguentarium (Fig. 9:3) is elongated and fusiform, made of semi-fine ware with a thick base and thick walls. It is dated to the Hellenistic period and has parallels at Kedesh and Shiqmona, dating to the early third–early first centuries BCE (Berlin 2015: Pl. 6.1.20). The accompanying cooking pot (Fig 9:4) and amphora base (Fig 9:5), however, date from the Roman period. The first is a Kefar Ḥananya closed cooking pot Form 4C (although not made of Kefar Ḥananya ware), which is attributed a chronological range from the early second to the mid-fourth centuries CE (Adan-Bayewitz 1993:128–130). The second belongs to a conical and ribbed amphora made of a local coastal fabric, possibly of the type manufactured at Ḥorbat ʽUẓa in the third to early fourth centuries CE (Avshalom-Gorni 2009: Fig. 2.36:14).
The glass bowl (Fig. 9:6) is hemispherical and has rather thin walls, a slightly incurving, fire-rounded rim and a low concave base without a pontil scar. The glass is bluish green, with a blackish brown layer of weathering. The form of the bowl and the quality of the glass are characteristic of the Galilee in the third century CE. This bowl has many parallels in Roman-period burials in upper Western Galilee, for example in Cave D at Ḥurfeish, where it is dated to the third century CE (Gorin-Rosen 2002:153*–154*, Fig. 9:28, 29, and see there additional parallels from Yeḥiʽam, Kabri, Oshrat, Loḥame Ha-Geta’ot and Nahariyya).
The limited finds retrieved from Tomb 3 suggest that it was first used in the Hellenistic period (third–first centuries BCE) and that the secondary mid-shaft burial(s?) can be dated to the Late Roman period (third–early fourth centuries CE). The Hellenistic unguentarium found in the Roman secondary burial was probably removed from the original Hellenistic-period burial within the chamber.
 
Rock-hewn shaft tombs were a common burial form along the Phoenician coast throughout the first millennium BCE. Rock-hewn tombs with vertical, rectangular or square shafts, leading into single burial chambers were common in Iron Age II (ninth century BCE and later) at Akhziv on the Phoenician coast, and some of these tombs exhibited steps (Dayagi-Mendels 2002:3–4, 41, Plan 4.1). Rock-cut shaft tombs with a single burial chamber were also characteristic of the Persian period (fifth–fourth centuries BCE), and were found at Tell er-Ras, near Lohame Ha-Geta’ot (Onn 1999). Two rock-hewn shaft tombs with several burial chambers or loculi dating from the late Hellenistic and Early Roman periods were also exposed at Tell er-Ras (Rochman-Halperin 1999).
 
Quarries
Small clusters of rock-cuttings were found in the uneven rock surface in Sqs 2 and 3 (Fig. 15), Sqs 18 and 19 (Figs. 16, 17) and in Sqs 20–22 (Figs. 4, 6, 18)—the negatives of rectangular quarrying marks and severance channels for ashlar building blocks carved out of the nari layer overlying the chalk bedrock. In Sq 19, a single ashlar block (0.2 × 0.4 × 0.6 m) was still attached to the bedrock. Most of the cuttings were shallow (c. 0.2 m), whilst others exhibited two tiers of steps (c. 0.4 m), indicating that in some cases more than one tier of stone blocks was quarried. The quarrying remains in Sqs 20–22—two adjacent, slightly sunken rectangular surfaces from which stone blocks had been removed; and negative cutting marks visible around the shaft of Tomb 1—as well as the ashlar blocks filling in the shaft of nearby Tomb 1, indicate that the stone-quarrying activity was probably a preliminary stage in the hewing out of the shaft tombs. The evidence illustrates a symbiotic relationship between the stone quarrying and the tomb-hewing, possibly substantiating the premise that the quarrying activities also date to the Hellenistic period, or at least began in this period. A few small worn, ribbed pottery body sherds and very few small glass fragments from the Roman and Byzantine periods that were scattered on the bedrock surface may reflect some general presence or activity in the area during these periods.
 
Water Channel
A short segment of a rock-hewn channel (L125; length 4 m; Fig. 15: Section 1–1) was unearthed at the northern edge of Sq 3. At the outset, it seemed to be a simple, deep, quarry cutting on the sloping bedrock. However, after the square was expanded northward by removing the wide balk with a mechanical backhoe, it become evident that it was the south-sloping side of a deep, rock-hewn east–west channel (Figs. 19, 20). The carefully hewn channel had a narrow, leveled floor (width 0.7 m, elevation 62.19–62.16 m asl from east to west); its steeply sloping walls (1.5 and 1.7 m high) broadened out at the top to a maximum upper width of 1.9 m. Notwithstanding the short four-meter long exposed segment, the 3 cm lower elevation of the western end of the channel’s floor was sufficient to indicate that water in the channel flowed from east to west. The channel was found full of soil, but no sherds or other datable finds were retrieved.
The large dimensions of the channel indicate that it was almost certainly not hewn for domestic or irrigation purposes. Its original water source, its destination and the date of its construction can only be deduced from its orientation in the context of the several other water-channel segments previously exposed in the region. Such a comparison leads to the understanding that it was part of the Hellenistic-period aqueduct that supplied fresh water to the city of ʽAkko-Ptolemais.
Previous surveys, as well as salvage and research excavations, revealed several segments of ancient aqueducts within the area of the ʽAmqa Map(Frankel 2012; Frankel and Getzov 2012). The various aqueduct segments exhibit a variety of dimensions and construction techniques: some are stone-built, others are tunnels with vertical shafts, and yet others are rock-hewn open channels (Frankel 2012). Two main groups of aqueduct segments were identified: a western group, found in the coastal plain near Kibbutz Lohame Ha-Geta’ot and Moshav Nes ʽAmmim; and an eastern group, about 5 km east of the first group, on the lower slopes of Naḥal Bet Ha-ʽEmeq, east of Moshav ʽAmqa (Fig. 23).
The western group comprises a rock-hewn tunnel segment at Nes ʽAmmim (Frankel 1985:134–138; Frankel and Getzov 2012: Site 150) and a plastered stone-built channel segment and two underground rock-hewn tunnel segments with stone-built vaulted ceilings at Loḥame Ha-Geta’ot (Porat 2009; Frankel and Getzov 2012: Site 26). The Nes ʽAmmim segments were dated by a few pottery sherds, including a Hellenistic-period lamp. The location, orientation, dimensions, high construction standard and the surmised date led Frankel and Getzov to the clear understanding that the western group was part of the main aqueduct that supplied water to the city of ʽAkko-Ptolemais in the Hellenistic period (Frankel 2012:13*).
The eastern group comprises three parallel aqueduct segments—two channels and one tunnel—that ran close together for over 2 km along the lower slopes south of Naḥal Bet Ha-ʽEmeq. They were almost certainly constructed to transport water from the springs that emerged in Nahal Bet Ha-ʽEmeq, the main source being at ʽEn Yifʽam (also known as ʽEn Poʽem and ʽEin Majanuna), situated to the northeast of Yarka (Frankel 2012:12*–13*; Frankel and Getzov 2012: Site 159). Their construction was ascribed to the Hellenistic period based on several pottery sherds retrieved from the channels. This group exhibited technical differences and a lower standard of rock-hewing than the western channels. In addition, the rock-hewn tunnel—a short segment that includes vertical shafts excavated in Moshav ʽAmqa (Frankel and Getzov 2012: Site 159)—seems to have never been completed (Frankel 2012). It thus seems that these three aqueduct segments were hydrological engineering endeavors to complete the aqueduct leading to ʽAkko-Ptolemais.
Frankel and Getzov considered two possible explanations for the two groups of channels. First, they proposed that both the western and the eastern groups were part of the Hellenistic aqueduct for transporting water from the Naḥal Bet Ha-ʽEmeq springs to the Hellenistic polis of ʽAkko-Ptolemais (Frankel 2012:13*–14*). Their suggestion was lent credence by the large scale of the engineering project and by the understanding that only a ruling authority such as a polis, could have undertaken such a venture. The alternative proposal was that the western channel segments of the Hellenistic aqueduct to ʽAkko-Ptolemais were fed by the more northerly Kabri springs, as was the case two millennia later, when the eighteenth- and nineteenth-century CE Ottoman aqueducts to ʽAkko were constructed (see Fig. 23). According to this premise, the eastern group of Hellenistic-period channels and tunnels alongside Nahal Bet Ha-ʽEmeq provided the local villages with water for drinking and irrigation.
The short Kuweikat segment provides ‘the missing link’: it lies midway between the western and eastern aqueduct segments, and its east–west orientation aligns well with both groups. It thus indicates that both groups are part of the Hellenistic-period aqueduct that transported fresh spring waters along almost 16 kilometers, from the Yifʽam springs in Naal Bet Ha-ʽEmeq to the Hellenistic polis of ʽAkko- Ptolemais.
 
Additional Remains
The only building remains uncovered in the excavation are a couple of flimsy low stone walls cemented with primitive concrete that must have been part of an early twentieth-century CE building (L109; Sqs 5–6; Figs. 9, 10, 21). Near the quarry remains in Sq 2 was a small, roundish hewn pit (L124; Figs. 14, 22). The pit was found intentionally refilled with medium-sized to large fieldstones, but it contained no datable finds, and its function was not clarified.
 
The three simple, rock-cut shaft tombs unearthed in the excavation, had either one or two small burial chambers. Whilst no artifacts were extant to date Tomb 1, the earliest burials in Tombs 2 and 3 were dated to the Hellenistic period (second century BCE), and Tomb 3 was reused for burial in the Roman period (third or fourth century CE). The tombs were probably part of the Hellenistic and Roman burial grounds of the adjacent contemporary settlement at Tel ʽEmeq. The rock-cut shaft tombs exhibit the same tomb plans found along the Phoenician coast since Iron Age II (ninth century BCE), reflecting cultural, and possibly ethnic, continuity.
The clusters of negative quarrying marks discerned on the bedrock surface are evidence of ashlar nari quarries. Whilst no pottery dated the quarrying activity, the possible symbiotic association between the stone quarrying and the tomb hewing may support a Hellenistic-period date for the quarries as well.
The most significant contribution of the excavation was the short segment of the large east–west rock-cut water channel. This ‘Kuweikat Channel’ provides the ‘missing link’ connecting two groups of previously known aqueduct segments and confirms the hypothesis that the Hellenistic polis of ʽAkko-Ptolemais received its fresh running water supply via a c. 16 km-long aqueduct from the springs of Naḥal Bet Ha-ʽEmeq. The Hellenistic aqueduct to ʽAkko-Ptolemais is the earliest, large-scale aqueduct known in Israel.