The first excavation was conducted in a Roman-period burial cave hewn on the northern bank of a stream, c. 100 m northeast of Horbat Sa‘adon. The second excavation was conducted in the southwestern church of the Byzantine-period settlement, c. 300 m southwest of the burial cave.
 
The Burial Cave
A loculi burial cave (5 × 5 m) was entered via an opening in its southeastern wall; the opening, which widens inwards (external width 1.3 m, internal width 2 m; Figs. 3, 4), was revealed c. 2 m below the present surface—probably the ancient level of the banks of the stream, which have risen over the past 1600 years due the presence of terrace walls in the streambed. The cave comprises a central chamber and seven well-hewn arched loculi (average dimensions 1 × 2 m, height 1.26 m; Figs. 5–7): six along the northwestern wall, and one, whose quarrying was never completed, along the northeastern wall. The arched opening of one loculus was set within a carved frame, and the letter T was carved on the upper part of the northwestern wall, between two loculi. A standing pit in the center of the chamber was surrounded by a rock-hewn bench; it was found filled with mud and debris, and was not excavated. Chisel marks were discerned on the ceiling (Fig. 8). The burial cave’s plan is similar to that of the en-Nusra burial cave at ‘Avedat, dated to the Middle Roman period (second century to mid-third century CE; Negev 1997:79–88), albeit its smaller size.
 
The Church
A previous excavation at the site (Hirschfeld 2006:21–24) revealed a small basilica church (5.5 × 15.0 m; Figs. 9, 10): two rows of columns dividing the isles from the nave, a round central apse (diam. 1.8 m, preserved height 2.8 m) in the eastern wall, and adjacent to it, to its north, an arched prayer niche with a small reliquary niche beneath it. Illicit diggings in the church destroyed this reliquary niche and the wall below the prayer niche. These illicit diggings also revealed another arched prayer niche to the south of the apse, with a damaged wall below it (Fig. 11), and two openings along the southern wall of the church. It was consequently decided to excavate these elements of the church.
The excavation revealed three phases: (1) construction and occupation (fifth to mid-seventh centuries CE); (2) damage caused by an earthquake and subsequent repairs and down-sizing (mid-seventh century CE); and (3) abandonment, dismantling of the pavers and architectural elements and the writing of Arabic graffiti on the walls (eighth–ninth centuries CE). The excavation yielded pottery sherds and glass fragments dated to the middle and late phases of the Byzantine period (450–650 CE).
 
Phase 1. Well-preserved, hard limestone flooring slabs were found throughout the nave, but the floor of the bema, where an altar table and the chancel screen would have stood, was largely robbed out in antiquity. A square column found c. 1.1 m north of the southern wall of the church (Fig. 12) seems to have belonged to a row of columns which ran along the southern side of the nave. Pieces of painted plaster and stone, as well as larger, flat stones with painted decorations in black, red, white, green and yellow, were found in the area of the bema, near the eastern wall (Figs. 13–16). These are of rather good workmanship and seem to belong to the earlier phase of the church.
Numerous architectural elements, which probably belonged to the first stage of construction, but were later robbed from their original location, were found discarded in the church: one chancel column, carved from local limestone, was found in an undisturbed area in the nave; another limestone chancel column was discarded by the looters south of the church (Fig. 17); two carved limestone chancel screens (Figs. 18–20)—one of which was broken, but an engraving of a running gazelle and a cross could still be discerned (Figs. 19, 20)—were discovered near the eastern opening (width 1.2 m) in the southern wall; a single round stone column drum (Figs. 21, 22), which may have originated in a room or atrium south of the southern wall, was also revealed in front of this opening; and a decorated door-post base (Fig. 23) was discarded by looters inside the nave.
 
Phase 2 represents repairs in the church following damage which was apparently caused by an earthquake. These seem to include stonework repairs in the northern wall of the church, along the bema (Fig. 24); blockage of the western of the two openings in the southern wall; and possibly the perforation of one of the stones in the doorframe of the eastern opening (Fig. 25). The simple, whitish yellow plaster, still adhering to large parts of the eastern and southern walls (Fig. 26), is of a much lower workmanship quality compared to the painted plaster and stone, and thus seems to also belong to Phase 2, when the church was repaired.
The 1.1 m width space between the square column, still standing along the southern side of the nave, and the southern wall was blocked with a wall built of stones and covered with the same whitish yellow plaster (see Fig. 12). This blockage effectively reduced the size of the church in Phase 2. An external addition which seems as a revetment was built during this phase against the outer face of the northern wall of the church; it was discerned at the western end of the wall (Fig. 27) but was not excavated. Revetments of this type, supporting the walls of buildings, were found throughout the site, including the structure immediately east of the church (Fig. 28), and around the Northern and Western churches at nearby Rehovot-in-the-Negev (Tsafrir 1988:77).
 
Phase 3. After the church was abandoned, architectural elements were dismantled and robbed from their original location. In the area of the bema, most of the floor was removed, and the chancel screen and columns were dismantled. Charcoal graffiti were written on the walls of the church. These include Arabic inscriptions and several figures (Figs. 26, 29): traces of Arabic writing and a depiction of a horse and rider on the whitish yellow plaster of the apse; the inscription “Allah” on a whitish yellow plaster surface (Fig. 30) at the bottom of a probe opened along the outer face of the southern wall, where robbers had dug a trench; and faint traces of Arabic script on one of the carved chancel screens (Fig. 31). These graffiti are most probably contemporaneous with the Kufic Arabic script of the eighth–ninth centuries CE documented in the church in a previous excavation (Hirschfeld 2006:25–27).
 
The rock-hewn tomb appears to be a family tomb from the second – mid-third centuries CE. This dating is based on its resemblance to the Roman-period Tomb at ‘Avedat. This dating corresponds with a sherd of a Roman-period round lamp found at the site by Hirschfeld (2006: Fig. 6:23). The Tomb expands our knowledge of the site, which until the current excavation was regarded as containing architecture and finds that are solely of the Byzantine period. The excavation of the church revealed evidence of damage caused by an earthquake and subsequent repairs and down-sizing in the mid-seventh century CE. The earthquake is most likely the same local event that destroyed the towns of ‘Avedat and Mezad Yeruham (Y. Baumgarten, pers. comm.) and caused damage at Shivta (North and South churches) and Nizzana (Nessana; North Church and south wall). Such an event explains the need for the revetments in the second phase. The finds from the church seem to indicate that it was abandoned before the beginning of Abbasid period.