In November 2014, a trial excavation was conducted in a private lot at the ‘Atarot industrial zone near Jerusalem (Permit No. A-7265; map ref. 221488–549/639642–748), prior to development work. The excavation, undertaken on behalf of the Israel Antiquities Authority and financed by Rami Levi, was directed by D. Yeger (field photography), with the assistance of N. Nehama and R. Abu Halaf (administration), A. Hajian, M. Kahan and M. Kunin (surveying), R. Cohen and B. Dwyat (archaeological inspection), D. Levi (GPS), D.T. Ariel (numismatics), S. Krapiwko (archive), I. Lidsky-Reznikov (pottery drawing), L. Kupershmidt (metallurgical laboratory), B. Dolinka (ceramics) and B. Touri (IAA district archaeologist).
Five squares were opened where archaeological finds were exposed during the antiquities inspections that preceded the excavation (Fig. 5).
Square 1. A wall (W1; length 6 m, width 1.8 m, height 0.8 m; Fig. 6), built of two rows of large, hewn stones, with a fill of densely packed small stones and earth (L2; Fig. 7) and preserved to a height of two courses, was exposed. South of the wall there is a collapse of large building stones (L3), probably from a toppled wall. Gray mortar on the stones indicates the construction technique. Brown soil and fieldstones (L27) accumulated north of the wall. The pottery that was recovered from the core of the wall dates to the Crusader and Ayyubid periods (second half of the twelfth–first half of the thirteenth centuries CE).
Square 2. A wall (W2; length 7.1 m, width 1.8 m, height 0.55 m; Fig. 8), built of two rows of large hewn stones with a fill of densely packed small stones and earth (L4; Fig. 9) and preserved to a height of three courses, was exposed. The wall was subsequently broadened by the addition of another wall along its northern face (W6; combined width 2.5 m). Wall 6 was built of large hewn stones, with a dense fill of small fieldstones (L16). Here too there was a collapse of large building stones on the southern side of the wall (L5).
Square 3. Only the eastern half of the square was excavated. A wall (W7; length 5.25 m, width 0.5 m, height 0.85 m; Fig. 10) built of a single row of fieldstones and medium-sized hewn stones, and preserved to a height of two courses, was uncovered. North of the wall, a coin dating to the reign of Herod (IAA 144718) was found on a surface of stone slabs and a threshold (L10; Fig. 11). A fill of small fieldstones (L26) was exposed south of the wall. The entire square had been covered with a layer of collapse, which was removed during the antiquities inspection, leaving remains on the edges of the square.
Square 4. The area was severely damaged as a result of modern construction; nevertheless, five phases were discerned in it:
Phase 1—a floor of yellowish crushed chalk (L22) on a bedding of small dense fieldstones and brown terra rossa, was exposed above the bedrock (L14, L23; Fig. 13).
Phase 2—An irregular-shaped refuse pit (L19; Fig. 14) containing a large quantity of Abbasid-period pottery (ninth–tenth centuries CE), was dug through the floor and bedding.
Phase 3—A wall (W8; length 3 m, width 1.8 m, height 0.4 m) was built of two rows of large hewn stones, with a fill of densely packed small fieldstones and earth (L24; Fig. 15). A Hasmonean coin (IAA 144720) and a Mamluk fals (IAA 144719) of the fourteenth century CE, were found in a foundation trench that was dug in dark-brown soil (L12, width 10 cm; Fig. 16) west of the wall. The foundation trench was identified by the later wall (W3), which seems to have been built against the western face of W8, thereby suggesting a corner, a pillar or a doorway at that spot. Wall 8 cut the southern side of the refuse pit (L19). A floor of rectangular hewn stone-slabs (L21; Fig. 17) was built on a bedding of small dense fieldstones, mixed with reddish brown soil (L14). A Mamluk fals (IAA 144721) dating to the fourteenth century CE (from the el-Iskandariya mint), was found between the slabs. Remains of the floor concentrated in the southwestern part of the square and in the northern part above the refuse pit (L11), and it seems that it abutted W8 from the southwest; hence they are contemporary.
Phase 4—A wall (W3; length 1.8 m, width 0.3 m, height 0.3 m; Fig. 18) built of soft limestone blocks, still retaining remains of white-gray plaster, was constructed above W8. Wall 9 was built to the east of W3, and parallel to it, and both were similarly constructed. Wall 9 abutted the eastern part of W8. Walls 3 and 9 seem to be a double wall with a core (width 1.6 m); a modern sewer pipe (L17) leading to a sewer box (L25) cut through both and through W8, on which they were built. Collapsed stones of W3 (L18) were visible on the floor (L21; Fig. 19), indicating that the floor and the wall were both in use when the wall toppled over. Wall 9 abuts W8, and the construction of W3 conformed exactly to the western face of W8, and did not continue farther north; therefore the courses of these walls should be considered as a technological stage in the construction of the building, or a later repair.
Phase 5—On the western side of the square were two modern foundations (L15), and together with the sewer box and sewer pipe, they represent the latest phase of the square which dates to the 1970s (Fig. 20).
Square 5. A wall (W4; length 1.3 m, width 1.6 m, height 0.6 m; Fig. 21) was built of two rows of medium-size hewn stones, with a fill of densely packed small stones and earth (L6). Wall 4 formed a corner with W5 (length 2.4 m, width 1.1 m, height 0.8 m), which was built in a similar manner and bonded to it from the south. Three courses were exposed of each wall (Fig. 22). Collapsed building-stones and stone-slabs on the southern and eastern sides of the walls, indicate the existence of a vault (L7). Limited dismantling of the collapse yielded pottery dating to the Roman–Byzantine periods, and a coin (IAA 144717) from the time of Justinian I (534–565 CE, Carthage) was found on the surface. Terra rossa fill (L8) was observed north of W4.
A very limited assemblage of Abbasid-period pottery (ninth–tenth centuries CE) consisting of storage and serving vessels, was found in the refuse pit in Sq 4 (L19): a large bowl with triangular rim above plastic decoration (Fig. 23:1), a bowl with ledge rim decorated with incised triangles and dots (Fig. 23:2), a basin with ledge rim (Fig. 23:3) and basins with inverted rim (Fig. 23:4, 5), holemouth jars with ridged rim (23:6, 7), a jar with straight neck and wavy combed decoration (Fig. 23:8), undecorated jars with straight neck (Fig. 23:9, 10) and a jar with grooved rim (23:11). Later pottery from the Crusader and Ayyubid periods was found in the fill in W1: a globular cooking pot with everted rim, of a type that may have been manufactured in Beirut in the twelfth century CE (Fig. 23:12) and a fragment of a bowl with transparent glaze and a yellow stripe, dating between the mid-twelfth and mid-thirteenth centuries CE (Fig. 23:13).
The size of the site is estimated as two dunams, on the basis of the trial trenches that were opened prior to the excavation. The limited trial excavation that was conducted in an area of 125 sq m, revealed good preservation of antiquities, in keeping with the documentation and description of the site in previous surveys. The purpose of the trial excavation was to determine the location and extent of the antiquities.
At this point it is possible to say that the massive walls which were found in the northern part of the site testify to an enclosed and planned building, a conclusion that is corroborated by the exposure of walls with identical orientation in the southern squares. Apparently it was a complex of residential buildings, such as a khan or a fortified farmhouse, c. 100 m long and c. 40 m wide. The plan corresponds to the building that is documented in the photographs in the British Mandate archive.
The finds indicate several periods of occupation at the site, some of them, e.g. the Hasmonean, Herodian and Roman periods, represented by meager ceramic and numismatic finds from layers of fill. The main periods of occupation are the Abbasid (ninth–tenth centuries CE), Crusader-Ayyubid (twelfth–thirteenth centuries CE) and Mamluk (fourteenth century CE) periods. The artifacts from the later periods were found between the walls of the buildings and their adjoining floors, and they reinforce the identification of the site with a khan from the Mamluk period, which may have had its origin in a structure dating to the Crusader period.