Three areas were opened (AC; 52 sq m; Fig. 1), adjacent to Road 40 on the west, in places where concentrations of knapped flint artifacts were discovered in the survey and the trial trenches that preceded the excavation. The excavation was conducted within a grid of 2 × 2 m squares and yielded flint artifacts dated to the Middle Paleolithic period. The site extends east of Har Miyah, where a past survey identified several flint knapping sites, some dating to the Middle Paleolithic period (S-756/2017, S-832/2018). Near the excavation, on the slopes to Naal in and Naal Boqer, several sites dating to the end of the Middle Paleolithic and to the early Upper Paleolithic periods were discovered (Marks 1981; 1983; Barzilay and Boaretto 2016). In the vicinity of the excavation, high-quality brown, fine-grained Eocenic flint nodules were identified, which served as raw material for flint knapping.
 
Areas A, B. The areas are close-by, comprising together nine squares that were excavated near agricultural terraces dating from the Byzantine period. Most flint items were found in the surface layer (thickness 0.1 m; Figs. 2, 3). Below the surface layer lay a sterile layer of soil, which covered a light-colored, compacted layer of soil (0.2 m depth below the surface). Bedrock was exposed below the light-colored layer.
 
Area C. Four squares were excavated, yielding meager flint finds, most from the surface layer (thickness 510 cm); below this layer the bedrock was exposed.
 
The Flint Assemblage. Altogether 467 flaked flint items were recovered, including debitage, cores and tools (Table 1). Flakes are the most abundant artifact; in addition, the percentage of primary elements is high, and in this category as well flakes outnumber blades.
 
Table 1. The Flint Assemblage
Type
Area A
Area B
Area C
N
%
N
%
N
%
Primary flake
37
31.9
40
22.5
14
27.5
Primary blade
4
3.4
6
3.4
 
 
Flake
68
58.6
115
64.6
35
68.6
Blade
3
2.6
12
6.7
1
2.0
Bladelet
 
 
1
0.6
 
 
Core trimming elements
2
1.7
3
1.7
1
2.0
Overpass
2
1.7
1
0.6
 
 
Total knapping debitage
116
100
178
100
51
100
Chips
 
 
34
25.8
9
60.0
Chunks
72
100
98
74.2
6
40.0
Total debris
72
100
132
100
15
100
Cores
15
6.1
15
4.0
2
2.9
Tools
44
17.8
46
12.4
 
 
Knapping debitage
116
47.0
178
48.0
51
75.0
Debris
72
29.1
132
35.6
15
22.1
Total
247
100
371
100
68
100
 
Thirty-two cores (Table 2) of widely varying types were collected in all areas. The most common type is the single-platform core (43.7% of all cores; Fig. 4:1, 2), used mostly for producing flakes (31.3% of all cores). This pattern of flake production was observed also in the knapping debitage, indicating that flake production was the dominant industry. Various Levallois cores (15.6% of all cores; Fig. 4:3, 4) were also found.
 
Table 2. The core assemblage
Core type
Area A
Area B
Area C
Total
N
%
N
%
N
%
N
%
Single-platform core for flakes
4
26.7
5
33.3
1
50.0
10
31.3
Single-platform core for blades
1
6.7
1
6.7
 
 
2
6.3
Single-platform core
 
 
2
13.3
 
 
2
6.3
Levallois unidirectional convergent
1
6.7
 
 
 
 
1
3.1
Levallois unidirectional convergent preferential cores for points
1
6.7
1
6.7
 
 
2
6.3
Levallois centripetal core
1
6.7
 
 
 
 
1
3.1
Bidirectional flake core
1
6.7
2
13.3
 
 
3
9.4
Multiplatform core on two surfaces
2
13.3
1
6.7
1
50.0
4
12.4
Levallois bidirectional for blades
1
6.7
 
 
 
 
1
3.1
Core fragment
3
20.0
2
13.3
 
 
5
15.6
Undetermined
 
 
1
6.7
 
 
1
3.1
Total
15
100
15
100
2
100
32
100
 
Areas A and B yielded 90 tools (Table 3), with end scrapers, side scrapers and notches being the most abundant. The most common are simple end scrapers on flakes (31.6% of all end scrapers; Fig. 5:1) and converging end scrapers (21.1% of all end scrapers). The two main types among side scrapers are convex (44.4% of all side scrapers) and straight (38.8% of all side scrapers; Fig. 5:2). Converging side scrapers were also identified (11.1%; Fig. 5:3), and a single side scraper with only the proximal end being convex. The notched items (Fig. 5:4) did not show a standard characteristic for either the retouching technique or for the position of the notch. Tools produced on flakes comprise 44.4%; of these, a quarter for producing end scrapers. Tools produced on Levallois flakes comprise 18.9%, and these demonstrate no preference for specific tool types. More than half of the tools (52.2%) are covered with brown (‘desert varnish) or whitish patina.
 
Table 3. The Tool Assemblage
Type
Area A
Area B
Total
N
%
N
%
N
%
End scrapers
15
34.1
5
10.9
20
22.2
Side scrapers
7
15.9
11
23.9
18
20.0
Notches
5
11.4
14
30.4
19
21.1
Denticulates
4
9.1
5
10.9
9
10.0
Awls
2
4.5
2
4.3
4
4.4
Truncates
1
2.3
0
0.0
1
1.1
Double tools
3
6.8
2
4.3
5
5.6
Retouched flakes
5
11.4
3
6.5
8
8.9
Retouched blades
2
4.5
3
6.5
5
5.6
Handaxe
 
 
 
 
1
2.2
1
1.1
Total
44
100
46
100
90
100
 
A single handaxe found in Area A (Fig. 5:5) represents a Lower Paleolithic stone industry.
 
The flint assemblage dates mostly from the Middle Paleolithic period. Some of the items were produced using the Levallois technique (Hovers et al. 2008; Hovers 2009). All items exhibit medium to high wear, and most bear a brown patina (‘desert varnish’) characteristic of assemblages from this period. It seems that the items recovered in the excavation do not represent in situ activity but were rather washed down from a nearby hill.