In the northern part of the excavated area, three elongated field walls built of piles of different-sized fieldstones, were uncovered (Fig. 3:1–3). All three walls were built adjacent to the bedrock on the western bank of the wadi, and extended eastwards, where they were cut by the seasonal south to north water flow. There was no evidence of the walls’ continuation on the eastern side of the wadi. The direction of the walls, across the stream, enabled the retention of runoff water and loess alluvium during seasonal flooding, thus slightly enriching the natural vegetation in the ensuing grazing season.
The northern field wall (Fig. 3:1) was built of fieldstones (0.3–0.5 m long) piled up along a rough line, and was preserved to a maximum height of 0.6 m, half of which lay beneath ground level, as it had been covered by alluvium over time. The middle field wall (Fig. 3:2) was similarly constructed, and it was preserved to a height of c. 1.4 m, again about half of it exposed beneath surface level. The southern wall (Fig. 3:3) was preserved to a height of c.1 m, built partially of large stones (c. 1 m long), and there may have been part of a building at its base that was destroyed by flooding. A granary formed out of a nearby rock ledge by constructing a wall of flat fieldstones beneath the ledge, was a recent Bedouin field storage installation (W150; Fig. 3: Section 1–1; Fig. 4).
 
A few buildings were unearthed south of the three walls (Fig. 2:4–8).
Building 4 consisted of two fieldstone walls built on the natural ground surface (W102, W114; c. 1 m thick) that together formed an elliptical structure and were probably both part of the same wall (Figs. 5, 6). The inner face of the walls was mostly built of upstanding stones, and partially of four courses of narrow stones laid horizontally (Fig. 7). A stump of an inner wall (W119) divided the internal space and may have defined a storage area. The bases of W114 and W119 were at a similar elevation (830.11 and 830.12 m asl), and this was probably the floor elevation. Upright stones incorporated in Wall 102, (one notably large c. 0.8 m high), and several flat stones, formed a raised corner above the height of the wall’s lower course (Fig. 5, shaded in gray in plan and section; Fig. 7). No finds were retrieved that could indicate the function of this distinctive feature. Inside the building, an archaeological accumulation, with no evidence of stratification, was found (L101; c. 0.5 m deep). The potsherds retrieved inside the building dated from the Nabatean, Roman, Byzantine and Early Islamic periods; twentieth-century CE refuse was found in the surface accumulation (0.4 m).
Buildings 5 and 6 were adjacent to each other (Figs. 5, 8), and each had a roughly circular or semi-circular wall whose southern part was preserved (W112, W116, c. 1 m thick). The height differences between the base of the walls indicate that Building 6 was probably constructed after Building 5, rendering obsolete part of the earlier wall. Alternatively, they may be partially enclosed installations, and the height differences may be due to the natural, slightly eastward-descending slope. In Building 5, no floor or habitation level was identified (L111) and no finds were recovered. Building 6 had a tamped earth habitation level (L106), c. 5 cm above the wall’s base. Immediately to its west, a small rectangular room with three walls (W104, W112, W118) was paved with flat stone slabs at the same elevation as the tamped earth floor. North of the small room, wall stumps attest to additional parts of the building that were not preserved. The few potsherds retrieved in the accumulation in Building 6 dated to the Byzantine period.
The area between Building 4 and Buildings 5 and 6 was covered with a c. 5–10 cm thick accumulation of soil containing ashy patches, a few potsherds, circular hearths not bounded with stones (L125, L127, L128), and many small burnt stones. This ashy layer overlay a layer of natural hard yellowish loess gravel, and also underlay Walls 104 and 118, located west of Building 6.
In the southern part of the excavation, two circular buildings were excavated in the wadi, at a slightly higher elevation than the other elements (Fig. 9).
Building 7 was circular (W131, c. 1.5 m thick; Fig. 10), and enclosed a uniform accumulation (L107) of light-colored soil that contained pottery from several periods: Early Bronze Age, Roman-Nabatean, Byzantine and Early Islamic periods. Immediately to its southwest, was a rectangular structure built of upstanding stones (W133) but devoid of finds, that was either a tomb or a storage installation.
Building 8 was a circular mound of stones (L109, W132; Fig. 10) that may have been a tumulus. No finds were retrieved in it.
South of Buildings 7 and 8, farther upstream, modern stone debris dumped in the past decade, has scarred the surface, making it impossible to identify any signs of ancient human activity.
The walls of all the buildings (4–8) were thick, at least two stone rows wide, built of fieldstones and founded on a thin layer of soil that covered the natural bedrock . Apart from Building 8, all the walls were constructed of several courses. Four of the five buildings had no definable floor and were built directly on top of the natural soil; Building 6 had a floor of tamped earth and flat stones. The stratigraphic relationship between the buildings was not established, apart from the layer of occupation debris including hearths, scattered ash and finds that was identified beneath the western part of Building 6 and that extended to the north. The pottery in the buildings, dating from several periods, provides no assistance in dating their construction and use.
The excavation, (including partial sieving) yielded 215 potsherds, of which 23% are small worn body fragments that could not be identified. Six percent, including a bowl (Fig. 11:1), are from the Early Bronze Age, and 8% from the Nabatean and Roman periods: Roman bowls (Fig. 11:2–6), a slipped goblet or small jar (Fig. 11:7) and a jug (Fig. 11:8); a Roman bowl (Fig. 11:9) and juglet (Fig. 11:10). Most of the pottery (60%) is Byzantine, comprising mainly cooking pots (Fig. 11:11–14). Seven potsherds (3%) are identified as Early Islamic, including a cooking pot (Fig. 12:1), jars (Fig. 12:2, 3), an oil lamp (Fig. 12:4) and a decorated body fragment (Fig. 12:5).
The pottery dating from six periods, including modern times, was retrieved inside and between the buildings, with no spatial or stratigraphic differentiation; it therefore only dates the periods during which the site was occupied.
The small quantity of pottery, including body fragments, retrieved from a 750 sq m excavation, may have one of two explanations. It may reflect the sporadic nature of the settlement comprising short periods of use during which little domestic refuse accumulated. Alternatively, it may be the result of taphonomic processes involving the dispersal and washing away of surface finds by the annual winter rains. The site may have been settled seasonally in the spring and summer, and deserted in the winter, whilst no steps were taken to protect the buildings from the floodwaters; the three cross walls provided no protection as they were built downstream from the buildings.
Additional finds, all undatable, include an orange stone bead (Fig. 12:6) found near the surface, a metal bracelet (Fig. 12:7), and an extremely worn coin. A round stone artifact with one worn flat surface and another face bearing strike marks (Fig. 12:8), probably a tool, was also recovered. A few animal bones belong to medium-sized mammals, possibly sheep/goat.
The flint assemblage consists of 244 items, originating as surface finds (Table 1). Most are debitage from a flake tool industry that used local flint available in the nearby streambeds. The majority of items are flakes (Fig. 13:1) and a few are blades (Fig. 13:2, 3); all the tools are ad hoc (Fig. 13:3–6), and the cores are flake cores and blade-and-flake cores (Fig. 13:7–9). 220 of the flint items are surface finds. The Mizpe Ramon site is located in a wadi with seasonal south to north water flow (also experienced on rainy days in the course of the excavation), and the spur to the south of the site contains flint outcrops that were used to manufacture flint tools during several periods. The flint items probably do not originate at the site itself, but from slightly further south, from a hill containing both flint outcrops and flint-knapping sites (Oron and Lavi 2018).
 
Table 1. Flint items
%
N
Type
16
38
Chunks
18
45
Chips
6
16
Primary items
38
92
Flakes
4
9
Blades
82
200
Total debitage
7
18
Cores
11
26
Tools
100
244
Total 
 
OSL dating was applied to two sediment samples from the excavation. The date retrieved for the sample beneath the eastern part of W102, was 1600 BP, whilst the sterile soil underlying the layer of human activity, supplied a date of 4900 BP. Both dates are consistent with the excavation data from the site, since the earliest evidence of human presence at the site reflected in the pottery finds, comes from the Early Bronze Age, and the other main period represented is the Byzantine period.
It is not possible to date the architectural units at the site since they contained a mixture of finds from several periods. The only relative stratigraphic consideration comes from Building 6, whose construction may postdate Building 5 and the habitation layer containing the hearths between it and Building 4. The evidence suggests that the buildings were used intermittently in several periods. Humans were first present at the site in the Early Bronze Age, and following a long settlement gap, some degree of settlement continued in the Nabatean, Roman, Byzantine and Early Islamic periods, as well as in the twentieth century CE.Evidence of a temporary presence at the site in recent generations was discovered on the surface, and seasonal grazing by Bedouin shepherds in the wadi is attested until the 1970s.
 
The finds from the excavation at Mizpe Ramon are compatible with temporary camp sites, seasonally resettled. The circular buildings, the paucity of material remains, the small diversion channels and the tumulus reflect the presence of pastoral nomads in the Negev highlands in the periods during which the site was inhabited: Roman-Nabatean, Byzantine and Early Islamic periods. The excavation consolidates the understandings from other surveys and excavations in the region, providing a record of human presence in Mizpe Ramon during the first millennium CE.