The building dated to the late Persian and early Hellenistic periods (fourth century CE) consisted a single room delimited by four walls (W11, W14, W19, W26; Fig. 3). The stones in W11 were arranged as headers and stretchers. A tamped-earth floor (L20) that abutted the walls was exposed inside the room. Pottery sherds from the late Persian and early Hellenistic periods, including jars (Fig. 4:1–6) and a juglet (Fig. 4:7), were discovered beneath the floor.
The building ascribed to the Early Roman period was constructed next to the earlier structure. It comprised an open-air courtyard and four rooms (1–4) adjacent to the south and east. The fieldstone constructed walls of the courtyard (W54, W56, W60) and of the rooms (1—W27, W55; 2—W40, W55; 3—W41, W42; 4—W57) were abutted by floors (L25, L35, L50, L59) made of firmly tamped light gray soil. A tabun (L48) was built into the floor in the courtyard’s southern corner. A section of a rock-hewn channel (L46; Fig. 5) that was apparently covered with stone slabs was revealed beneath a floor in the western corner of the courtyard. Room 1 was reached by way of a threshold stone that was fixed in the courtyard’s southern wall. A stone basin (L13; Fig. 6) and a grinding stone were discovered near each other on the western part of the floor in Room 1. Wall 57, which was part of Room 4, probably extended south and divided Room 1 into two parts. Pottery sherds from the Early Roman period (until the mid-first century CE), including bowls (Fig. 7:1, 2), cooking pots (Fig. 7:3–5), jars (Fig. 7:6–13), jugs (Fig. 7:14–20) and juglets (Fig. 7:21, 22), were discovered below the floors of the building. A stone weight (Fig. 7:23), a basalt grinding stone (Fig. 7:24) and flint items (below) were also found beneath the floors.
A rock-hewn winepress (L32), consisting a treading floor and a collecting vat, was documented east of the excavation squares. Two hewn cisterns (F61, F62) were documented just north of the winepress. The winepress and cisterns were evidently used during the same periods as the buildings. A wall (W63) and a cave (L64) were documented south of the winepress. 
The Flint Assemblage
Ronit Lupu and Jacob Vardi
The collected flint comprises 80 items dated to the Pre-Pottery Neolithic A and Middle Bronze Age II, including debris (chunks and chips), debitage, cores and tools (Table 1). The flint assemblage is small, perhaps because the excavated soil was not sifted. The items are made of two main types of raw materials: nodular flint and brecciated flint of the Mishash formation. The nodular flint is a uniform dark brown or dark gray color, and comes from the slopes of the Judean hills; it was primarily used in the Middle Bronze Age. The brecciated flint is gray or light gray and sometimes has a light patina or light colored dots on it. This flint was mainly used in the Pre-Pottery Neolithic A period. This raw material is well-known in the Modi‘in region, and its outcrops are found near the site, for example at Horbat Hadat, located c. 1 km northwest of the site, where items and tools from this source were discovered in the past (van den Brink 2007; Assis 2012).
Table 1. Flint artifacts
Debitage. The debitage consisted mainly of flakes (85%), about half dating from the Pre-Pottery Neolithic A period and half—from the Middle Bronze Age II. The flakes from the Pre-Pottery Neolithic A are made of brecciated flint, and some have transversal flaking. The Middle Bronze Age II flakes are made of brown or gray coarse-grain flint. They are square or trapezoidal, thick and broad, and are knapped from large blanks. The broken bulb of percussion on these items and the scarring pattern on the by-products of the flake production constitute additional confirmation that the assemblage dates to the later part of the Middle Bronze Age. The items are similar to those discovered at Mishmar David (J. Vardi, E.C.M. van den Brink and E. Yannai, pers. Comm.). In addition, three large core trimming items were made of dark gray raw material; they apparently date to the Pre-Pottery Neolithic A period. The debitage also contains two small narrow blades made of light gray flint, which probably date to earlier periods.
Cores. The cores are made of light gray or light brown raw material; most have several striking platforms and are fairly depleted. They are small and medium in size. Some are flake cores (Fig 8:1) and others are bladelet cores (Fig. 8:2). The cores should probably be attributed to the Pre-Pottery Neolithic A period. A small bladelet core (Fig. 8:3) was also discovered. It is made of gray raw material and is covered with patina; it has at least one depleted striking platform. The core is likely from the Epipalaeolithic period.
Tools (Table 2). Seven tools were found:
1. A backed sickle blade with bifacial retouch (a Beit Ta‘amir knife; Fig. 8:4). The glossed working edge is apparent. There is a break or a notch that was knapped at the base of the blade, possibly for a haft. The item is ascribed to the Pre-Pottery Neolithic A period.
2. A retouched bladelet (Fig. 8:5) knapped from gray raw material and covered with shiny patina. One side is delicately retouched and terminates with a truncation at the proximal part of the item. It is attributed to the Pre-Pottery Neolithic A period.
3. A thick broad adze knapped on a core trimming item. The item is attributed to the Pre-Pottery Neolithic A period.
4. Two end scrapers from the Middle Bronze Age II. One (Fig. 8:6) is knapped on a large blank. Its rectangular shape is characteristic of the blanks on which sickle blades were knapped in the second millennium BCE. The other scrapper is a thick, wide trapezoid; it too was knapped on a blank characteristic of the period, and there is a delicate lateral retouch on its cutting edge.
5. A retouched blade (Fig. 8:7). This implement is retouched along the entire length of one side. It has two notches opposite each other, making it narrow in the middle, possibly preparation of a base of a drill. The date of the item cannot be determined.
6. A blade with a retouched back. Its date cannot be determined.
Table 2. Flint tools
Retouched/backed blade/ bladelet
Sickle blade (Beit Ta‘amir knife)
End scraper
The flint items that were discovered in the excavation predate the buildings revealed in it; thus, the buildings may have been constructed on top of an ancient site. The site is located on a hilltop, and the items that were found there are thick and heavy; hence, they could not have been randomly swept there. It is possible that the flint items were brought there from an ancient site, together with soil that was used in the construction of the buildings. Surveys and excavations on the hills of Modi‘in revealed that the area is rich in prehistoric finds, particularly artifacts from the Pre-Pottery Neolithic A period (van den Brink 2005; Toueg 2009; Re’em 2012). However, the finds from the second and first millennium BCE are less common. The flint assemblage from the site supplements our information about the variety of known diagnostic items from the Modi‘in area and the distribution of the sites during these periods (Zbenovich 2006).