Stratum I (Early Roman period). A hewn staircase of a poorly preserved ritual bath (miqveh; Fig. 3) was excavated. It seems that the remains of this stratum were largely removed; the extant remains, as in the case of the ritual bath, were re-used during the Late Roman and Byzantine periods (Stratum II). Nevertheless, numerous non-in-situ coins and ceramic sherds from the Early Roman period attest to the rather intense occupation of the site during this period.
Stratum II (Late Roman and Byzantine periods) yielded the remains of a large structure (min. reconstructed size 17 × 33 m), probably a Roman villa that housed a clay lamp workshop. The structure was built along general north–south and east–west axes. The peripheral walls of the structure were built directly on the bedrock, with a foundation course composed of a row of large semi-worked boulders with a small stone backing. Remains of a white plaster layer were found coating the inner face, built of small stones. The northernmost room of the structure was paved with a white mosaic featuring a central, colored panel with a geometric design (Fig. 4). The design and the color scheme are reminiscent of the mosaic found in a room further south, unearthed in previous excavations (Dagan 2010:236, Fig. 306.3). The walls in the southern part of the structure were poorly preserved, and in most cases could only be identified through their associated hewn foundations. Likewise, several hewn installations were uncovered within the southern part of the structure; although they are likely associated with the structure, this could not be concluded with any degree of certainty. The structure’s southern wall abuts a plastered channel, probably part of a gutter system that collected runoff from the structure’s roof.
Over 600 oil-lamp fragments were collected throughout the structure area. The lamp assemblage is composed almost entirely of Beit Natif types (Baramki 1936; Rosenthal and Sivan 1978:99) dated from the third–fourth centuries CE. The dominant subtype at the site features an ovoid- or pear-shaped body, a large filling hole, repeated geometric designs on the rim and a small pyramid-shaped handle at the back (Rosenthal and Sivan 1978: Nos. 406–410;Sussman 1985–1986:68; Storchan 2017: Fig. 5). The majority of the lamps were painted red or brown, and some were left unpainted. In addition to the lamps, a cluster of 15 limestone lamp molds was found on the floor of one of the rooms in the southern part of the structure (Fig. 5). The molds, alongside the exorbitant quantity of lamp fragments, indicate that the southern part of the structure functioned as a lamp workshop.
Stratum III (Ottoman period). The entire site was covered with an earthen fill and used as an orchard. Stones were heaped to form terrace and field walls; a large part of these walls was removed during the initial excavation season. The terrace walls left intact on the slopes to the north and east of the excavation form terraced agricultural plots.
The plan of the Stratum II structure resembles that of a Roman villa; it dates from the fourth century CE. While most of the structure probably served a domestic function, its southern part had an industrial function as a workshop for the production of Beit Natif lamps. This discovery at Kh. Shumeila comes nearly 80 years after the Beit Natif lamps were first identified following an excavation in the nearby village of Bet Natif, where two cisterns-turned-refuse-pits contained numerous discarded lamps and several stone molds from a lamp workshop (Baramki 1936). But whereas the Baramki excavations uncovered waste discarded from an unexcavated workshop, the Kh. Shumeila excavation has unearthed the first in-situ clearly identified lamp workshop in the Judean Shephelah.