An area (c. 50 sq m) was opened and two construction strata (I, II) were exposed, after the surface soil (thickness c. 0.2 m) was removed manually. Two phases were discerned in the earlier and main Stratum II; bell-shaped pits and a shaft that probably belonged to a cistern were hewn in the first phase (IIB), and changes to these cavities that mainly included the quarrying of short tunnels for the purpose of connecting them were made in the later phase
(IIA). The end of Stratum II is dated to the Early Roman period (first century CE). In Stratum I, when the underground complex was mostly covered over, a wall was built above it and into the Stratum II remains.
The strata are described below from the earliest to the latest.
 
Stratum II
The earliest remains consist of a habitation level on levelled bedrock (L104; Figs. 2–4) upon which potsherds that dated to the Late Hellenistic period were found. Several underground cavities, comprising three pits (L107–L109) and two shafts (L105, L106) connected by tunnels were hewn in the chalky limestone bedrock beneath that level.
The three carefully hewn pits were bell-shaped with a flat floor and sides that curved toward the bottom. Bell-shaped Pit 109 (max. diam. 2.55 m, max. depth 1.5 m) had a circular opening hewn in its ceiling (diam. 0.65 m). The opening was covered and blocked with a small square pool (L103; inside dimensions 1.55×1.60 m, depth 0.2 m; Fig. 5) that was built above it. The sides of the pool were a single course of dressed limestone and its flat floor was mostly bedrock hewn. The sides and floor of the pool were coated with a thick layer (2 cm) of gray plaster mixed with crushed potsherds, patches of which had survived. Near the northwestern corner of the pool was a round, hewn depression in the bedrock (diam. 0.25 m, depth 0.33 m; Fig. 6) that was apparently used to stand a jar in. A short tunnel (diam. 0.6 m, length 0.5 m) that led to an adjacent bell-shaped pit (L107; max. diam. 2.35 m, depth 1.38 m; Fig. 7) was hewn in the southern side of Pit 109, 0.3 m above the floor. Pit 107 was slightly smaller than Pit 109, and its flat floor was hewn at a slightly higher level. Two round adjacent openings were hewn in the northern side of Pit 109; one was 0.6 m above the floor in the east (diam. c. 0.6 m) and the other—0.8 m above the floor in the west (diam. c. 0.7 m). A small niche, perhaps meant for holding a lamp, was hewn between the two openings. A hewn tunnel was in the southern side of Pit 107; only its beginning was excavated. This tunnel and the eastern part of Pit 107 were partly blocked by a wall (W112; length 1.8 m, height 1.38 m; Fig. 8) that was built of five courses of dressed limestone and small fieldstones. In the western side of Pit 107, 0.5 m above the floor, was another very short tunnel (diam. 0.6 m, length 0.3 m) that led to a circular hewn shaft (L106; diam. 1.2–1.8 m, min. depth 2 m; Fig. 9). The shaft widened unevenly downward and the excavation did not reach its bottom. In the southern side of the shaft was a rock-cut tunnel that split off into two branches: the eastern one (diam. 0.95 m) was not excavated, and the western one (diam. 0.7 m) led to another pit (L108). Pit 108 was discovered filled with limestone that had collapsed inside from the ceiling; it was only partially excavated, but it was possible to see that it too was originally a bell-shaped pit similar in size to the other two pits. An opening of another tunnel (not excavated) was exposed 0.3 m above the floor in the southern side of Pit 108.
A circular shaft (L105) blocked by roughly hewn stones (L105A) was discovered c. 1.5 m east of the Pit 109 opening. The deep shaft was not excavated due to safety precautions, but it seems it led to a water cistern.
Pottery vessels were found in situ on the floor of Pits 107 and 109 and inside Pool 103. On the floor of Pit 109, which was covered with a thin layer of light colored soil (thickness c. 0.3 m), cooking pots, jars, jugs, juglets and a lamp were found. On the floor of Pit 107, which was covered with a slightly thicker layer of light colored soil (c. 0.4 m), was a similar assemblage of cooking pots, jars, jugs, juglets and two lamps, one of which was imported. The largest number of pottery vessels was discovered inside Pool 103, including cooking pots, jars, jugs, juglets, a mortarium, two lamps and a small imported bowl. Several fragments of cooking pots dating to the Early Roman period were found in Shaft 106, as well as potsherds and several vessels from other periods (see below). The pottery assemblage found in situ on the floors of the pits and in the pool date the final use of these installations to the Early Roman period. Apart from the pottery recovered from in the pits, very few artifacts were discovered, among them a circular limestone basin, two basalt fragments, a small marble fragment, a broken iron nail and a small glass fragment.
 
Two phases could be discerned in the underground complex, but due to its nature and the limited scope of the excavation, it is difficult to define them clearly. The division proposed here is based on the stratigraphic data, in light of the information known from other contemporary rock-hewn underground complexes throughout the Lower Galilee. It is our understanding that in the first phase (IIB), the bedrock floors were made level, for example Habitation Level 104, and the three bell-shaped pits (107–109) were hewn; the floors belonged to dwellings whose walls did not survive and the pits were used as underground household granaries for storing agricultural produce. It also seems that Shaft/Water Cistern 105 was hewn in this phase. In the second phase (IIA), the tunnels that connected the pits were hewn, Pool 103 was built, obstructing the original opening of Pit 109, Shaft 106, which may have served as an entrance to the complex, was hewn and Wall 112, which blocked part of Pit 109 was built. These changes converted the cavities into a single subterranean complex. 
 
Stratum I
After the pool and the underground complex filled with soil and for the most part were no longer visible on the surface, a wall (W114; length 1.4 m, width 1 m, height 1.9 m) was built right above the remains of the pool. The wall consisted of dressed limestone, and although only a small segment of it was exposed, it turns out to have blocked the western side of Cistern 105’s shaft. Since no potsherds were found near the wall, it cannot be dated. In the soil fill that accumulated in Shaft 106 several potsherds from the Middle–Late Roman period and a bowl from the Abbasid period were found, proving that the shaft was not completely blocked at the end of Phase II. It can be concluded from this that at least some activity had transpired in the area during these periods and presumably Wall 114 was built in one of them.
 
Pottery
Fragments of pottery vessels dating to the Late Hellenistic and Early Roman periods were recovered from the excavation, as well as a few potsherds from the Middle Roman, Late Roman–Early Byzantine, and Abbasid periods.
Late Hellenistic Period (Fig. 10). The scant amount of sherds found on the bedrock level (L104) date the beginning of its use to the end of the second–beginning of the first centuries BCE. These include an open bowl of buff-colored clay with a ledge rim (Fig. 10:1), a buff-colored mortarium with a ridged profile (Fig. 10:2), a cooking pot with a concave, bow-shaped neck (Fig. 10:3), a jar of pale red clay with a wide aperture, without a neck (Fig. 10:4), a buff-colored jar with a thickened rim characteristic of the locally produced jars of the Galilee (Fig. 10:5) and a jar with a long neck that slopes outward and a rim with a square profile, characteristic of the jars from Judea (Fig. 10:6).
Early Roman Period (Figs. 11–13). Complete vessels and broken ones from the Early Roman period were found in Pits 107 and 109 and Pool 103, and potsherds were found in Pit 108 and Shaft 106 (Table 1). One hundred thirty-seven vessels were counted; in counting the vessels and rims, an attempt was made to count only one rim per vessel.
 
Table 1. Number of Vessels According to Vessel Groups and Loci
Locus
Terra Sigillata Vessel
Mortarium
Cooking bowl
Open cooking pot
Closed cooking pot
Jug and juglet
Jar (T1.3)
Jar (T1.5)
Phoenician Jar
Lamp
Total
Pool 103
1 bowl
1
 
7
17
16
3
3
1
2
51
Shaft 106
 
 
 
 
5
 
 
 
 
 
5
Pit 107
1 lamp
 
1
12
20
5
1
8
1
1
50
Pit 108
 
 
1
 
1
1
 
 
 
 
3
Pit 109
 
 
 
7
11
3
3
3
 
1
28
Total
2
1
2
26
54
25
7
14
2
4
137
%
1.5
0.7
1.5
19
39.5
18.2
5.1
10.2
1.5
3
100%
 
The cooking vessels are c. 60% of the assemblage and c. 17% are storage vessels (jars). The mortarium, jugs and juglets, mostly found inside the pool, were also used in the preparation of food. Four locally produced lamps were found and two imported terra sigillata vessels consisting of a bowl and a lamp. A count of the pottery vessels clearly shows that the main activity in the underground complex and in the pool above it was the preparation of food. The lamps were necessary to illuminate the subterranean cavities.
The small imported bowl (Fig. 11:1) is relatively rare in Israel and belongs to a group of terra sigillata vessels that were produced in Cyprus. The bowl is classified as Type P37B (Hayes J.W. 1985. Sigillate Orientali. Pp. 1–96 in Atlante delle forme ceramiche II: Ceramica fine romana nel bacino mediterraneo [Enciclopedia dell'arte antica classica e orientale]. Rome. Pl. XX.14). A similar bowl was found in the Temple of Pan at Banias where it is dated between the years 50 BCE–50 CE (Berlin A.M. 1999. The Archaeology of Ritual: The Sanctuary of Pan at Paneas/Caesarea Philippi. BASOR 315: Fig. 8:3). The mortarium with sloping sides is made of rough buff colored clay containing many grains of sand (Fig. 11:2). Its form is characteristic of Roman-period assemblages and it is probably also imported from Cyprus.
All the cooking vessels are characteristic of the Galilee in the Early Roman period. They include Kefar Hananya Type 1B cooking bowls (Fig. 11:3), and open cooking pots that are similar to the Kefar Hananya Type 3A vessels (Fig. 11:4, 5). Most of the closed cooking pots resemble Kefar Hananya Type 4A (Fig. 11:6–11) and several of them are similar to Type 4B (Fig. 11:12). The clay of the cooking pots did not undergo petrographic analysis; however, it does seem different from the clay of the vessels produced at Kefar Hananya.
The jugs and juglets have a globular body (Fig. 11:16), with a candlestick-like rim (Fig. 11:13, 14), with a sloping neck and plain rim (Fig. 11:15) or with an everted neck and a ridge in its center (Fig. 11:17, 18).
The overwhelming majority of jars have a ridged barrel-like body, some with an everted rim (Fig. 12:1–4) and similar to the jars classified as Type T1.3 (Díez Fernández F. 1983. Ceramica Comun Romana de la Galilea. Jerusalem, pp. 135–137, 181–184), and some with a ridge at the bottom of the neck and a recess inside the rim (Fig. 12:5–8) of the kind Fernandez classified as Type T1.5. The two kinds of jars are the most common in the Lower Galilee in the Early Roman period; however, Type T1.3 dates from the first century BCE until the second third of the first century CE, while Type T1.5 dates from the second third of the first century to the beginning of the second century CE; consequently their use only overlapped during the second third of the first century CE. Only two Phoenician jars with thickened square rims (Fig. 12:9) were found; these are characteristic of the Phoenician coastal plain and not the Lower Galilee.
A triangular handle bearing a palm tree decoration (Fig. 13:1) belongs to a special terra sigillata lamp with a spiral nozzle. Its clay is slightly powdery and the lusterless slip alludes to a Tyrian or coastal provenance rather than a Cypriot one. The lamp dates between the end of the first century BCE and beginning of the second century CE (Heginbottom-Rosenthal R. 1995. Imported Hellenistic and Roman pottery. In E. Stern. Excavation At Dor, Final Report. Vol. I, Areas A and C: The Finds [Qedem Reports 3]. Jerusalem. Pp.183–288, especially p. 241, Fig 5.20:2). Two lamps with knifed pared wick-holes (Fig. 13:2, 3) are the most common lamps in the Galilee in the first century CE, and two more lamps of the kind referred to as ‘boot lamps’ (Fig. 13:4, 5) are well-known from sites in the Galilee in the first century CE, such as Nazareth and Yodefat.
The pottery assemblage is very typical of assemblages in Jewish settlements in the Lower Galilee, for example, Yodefat, Karm er-Ras (Qana), Zippori, Nazareth and Migdal. It includes cooking vessels that are known from Kefar Hananya, but it seems that at least some of them were manufactured at other workshops, probably closer to Kabul. The assemblage dates to the Early Roman period, specifically to the second third of the first century CE, c. 60–70 CE. The Phoenician jar in the assemblage reflects Kabul’s location along the border of the Phoenician coastal plain. The two imported terra sigillata vessels also reflect the proximity to ‘Akko and the Phoenician coastal plain, and perhaps also a high standard of living.
The Middle and Late Roman Period and the Abbasid Period (Fig. 14). Several fragments of pottery vessels dating to the Middle and Late Roman periods were found in the fill that accumulated in Shaft 106, such as an open Kefar Hananya Type 3B cooking pot (Fig. 14:1). A casserole or fry pan and a matching lid (Fig. 14:2, 3) are characteristic of the Late Roman and Early Byzantine periods.
The shaft also yielded a glazed bowl of white clay, slipped white and painted with green and yellow strips and brown dots (Fig. 14:4). A similar bowl found in Yoqne‘am is dated to the Abbasid period.
 
The main findings of the excavation are the two construction phases of Stratum II that date to the Late Hellenistic and Early Roman periods. In Phase IIA, bell-shaped pits used as granaries were hewn in floors, probably in the basement of dwellings. It seems that the house was already built in the Late Hellenistic period—the Hasmonean period in Judea—and it is possible the first residents were Jews who arrived there from Judea. In Phase IIB, small tunnels were hewn and other modifications were made to the rock-cut cavities, converting them into a single subterranean complex. The complex was probably camouflaged beneath the dwelling. Food preparation was the activity documented in the complex. The presence of the pottery vessels on the floors of the pits indicates that in the end, the residents were forced to leave in a hurry, with no possibility of taking all their belongings with them. It is our understanding that the underground granaries in the dwelling were adapted as a hiding refuge as part of the preparation by the Jews for the Great Revolt against the Romans in 66 CE.
There is other archaeological evidence of converting rock-cut pits to underground complexes in the Jewish villages of the Lower Galilee during the Early Roman period. It appears that these changes were done in preparation of the Great Revolt under the leadership of the Galilean commander Josephus. A clear reference to the conquest of the town of Kabul by the Romans at the beginning of the war in 66 CE can be found in Josephus, who wrote, "At the head of a portion of his troops, Cestius marched against a stronghold in Galilee called Zebulon, a frontier town between Jewish territory and Ptolemais. This he found deserted, as the population had fled to the mountains; but it was full of valuables of every kind, which he allowed the soldiers to loot. The town, as he realized, was very beautiful, the architecture being similar to that of Tyre, Sidon and Beirut, but he set fire to it nevertheless" (The Jewish War, 2, translated by G.A. Williamson [1959], p. 160). It seems that this description clearly corroborates the excavation findings and what we have here is evidence of the first campaign of the Roman army to suppress the Jewish uprising in the Galilee.