Four squares (c. 100 sq m) were opened on the southern fringes of Tel ʽAfula, c. 100 m south of the southern peak and the fortress situated on its summit. A building dating to the Byzantine period was exposed; one of its rooms was decorated with a colorful mosaic floor. The building was apparently constructed in the fourth or early fifth century CE, and was used, with several changes made to it, until the sixth or beginning of the seventh century CE. The structure was partly damaged by pits, dug in the Ottoman period. Pottery sherds from the Early Bronze Age I, the Intermediate Bronze Age, and the Persian and Hellenistic periods were found, scattered randomly throughout the excavation area, with no possibility of attributing them to any architectural remains.
 
The Early Periods. Fragments of pottery vessels from various periods were found scattered on the surface and in fills throughout the excavation area without any possibility of ascribing them to any building remains or even to discrete layers of fill. These included an EB I jar (Fig. 2:1); a rim of a Greek amphora dating to the Persian period, from the island of Miletus (Fig. 2:2); a base of a mortarium (Fig. 2:3), also from the Persian period; a jar (Fig. 2:4); a spindle juglet (Fig. 2:5); and two Rhodian amphora handles bearing traces of worn stamp impressions, one rectangular and the other round (not drawn) from the Hellenistic period.
 
Byzantine Period. Part of a large residential building (c. 60 sq m) with an extensive open area to its east, was exposed. Remains of three construction phases were discerned.
Phase 1: The phase in which the building was erected.
The building had several paved rooms that probably surrounded an inner courtyard (Figs. 3–6). The southern part of a large paved room (more than 7 m long, width 1.5 m) was discovered in the northern part of the area. Two walls survived: an eastern wall (W112, length 2.5 m, width 0.6 m) and a southern wall (W140, length 5 m, width 0.4 m), built of a single row of large fieldstones with a small-fieldstone fill. Wall 112 was built of large fieldstones; the fieldstones in its eastern face, which was the visible exterior of the building, were dressed and straightened, while their unworked side was hidden deep in the wall. The western face of the wall was completed with small stones. Wall 140, which was slightly narrower, was apparently an internal partition wall inside the building. A large stone pavement (L138, L122 and L137 from east to west) abutted the walls. These three sections of pavement were probably part of the floor bedding of a single room that was later severed by the walls of Phase 3. A basalt loaf-shaped fragment of an upper grinding stone was found in secondary use between the paving stones (Fig. 7:4).
Another room (L139), carefully paved with adjoining flat stones, some of which were dressed, was exposed south of W140. One of the paving stones was an Olynthus millstone in secondary use, whose flat bottom side was placed facing up (Fig. 5, in center; Fig. 7:5). The southern part of the pavement was cut by a large Ottoman-period pit (L125). The pavement abutted three of the room’s surviving walls (W140 to the north; W128 to the west—length 2 m, width 0.6 m; W126 to the east—length 2 m, width 0.6 m). Wall 128 was built similarly to W112, but here the dressed stone surface of the wall faced west. It seems that the wall continued at least 5 m to the south and was connected to another wall (W131), exposed at the southern end of the excavation. Eastern Wall 126 was not as well preserved as the others, but a single dressed stone survived in the eastern side, next to which were small and medium-sized fieldstones. Apparently, this method of construction was similar to that of W112, with an even outward-oriented face. The southern room was therefore 3.8 m long. Its width was unknown, but judging by the dimensions of the mosaic added to the room in Phase 2, it may have been c. 3 m wide.
No other architectural remains were discovered in the area east of W112 and W126, and this area was apparently the eastern end of the building. No other floors or walls were exposed from this phase to the west of W128, and apparently, this area too was an open area. However, since the room in the northern part of the building continued west beyond the line of W128 and delineated the open area to the north, the area was probably within the building itself, possibly an inner courtyard surrounded by rooms.
The dating of the building’s construction is uncertain since the excavation did not descend below the floors. A fragment of a jug from the Early Roman period (Fig. 8:1; first century CE) was found in the foundation trench of W112 (L120). The fill to the east (L118), which was cut by the foundation trench, contained fragments of pottery vessels from later periods, including a cooking pot (Fig. 8:3) and a juglet (Fig. 8:4) from the Middle Roman period (first–second centuries CE) and cooking pots from the Late Roman period (third–fourth centuries CE; Fig. 8:5, 6). These potsherds provide a possible early date for the construction of the building, and indicate that it was not erected prior to the fourth century CE. Another vessel from the Early Roman period (first century CE), a spherical juglet with a flat base and a thin neck (Fig. 8:2), was found in secondary deposition, lying at the bottom of a pit from Phase 3.
 
Phase 2: The building was probably renovated during this phase, and a colorful mosaic floor (L129; Figs. 4, 9) installed above the stone pavement in the southern room. The other rooms remained unchanged. A relatively large section of the mosaic (c. 1 × 2 m) that abutted W126 survived in the eastern part of the room. Another smaller section was preserved in the northwestern corner, which abutted W140 and W128 (Fig. 10). The mosaic floor was composed of white, black, yellow and red tesserae (1 × 1 cm) and was installed above two foundations layers—an upper layer of dark gray plaster (thickness c. 2–3 cm) and an underlying layer of scattered tesserae in hard gray plaster (thickness c. 1.5 cm). The bottom foundation layer was laid directly upon the stone pavement of Phase 1 (L139). The tesserae scattered in the mosaic bedding from this phase may be evidence of an earlier mosaic floor from Phase 1 situated elsewhere in the structure.
The decoration in the surviving mosaic section is composed of a panel with a double guilloche strip of intertwined ribbons alongside a white carpet decorated with tiny flowers and two rows of geometrical shapes made of black tesserae—mainly squares and symbols resembling the letter X (Fig. 11). It seems that the inner portion of the mosaic carpet was divided into square or rectangular panels, each decorated with a different geometric pattern. The small flowers appearing between the two main panels that were preserved may be part of the panel adorned with squares or a background decoration of the carpet between the panels (S. Miller, pers. comm.; Fig. 12). Such mosaics are common mainly in the fifth and sixth centuries CE, but the poor preservation of this specimen does not allow for an accurate dating based on the motifs.
 
That being the case, Phases 1 and 2 were dated to the fourth and fifth centuries CE. Three coins were found on the pavement of the northern room: two, the first a coin of Constantius II, dating to 351–361 CE (IAA 158010) and the other a Late Roman bronze coin dating to the fourth–fifth centuries CE (IAA 158011), were found between the paving stones in its eastern part (L138), sealed beneath the floor of Phase III (L121). Another coin of Constantius II (IAA 158012) was found on the westernmost part of the pavement (L137)—a section that was not sealed by the floors of Phase 3.
The construction of the mosaic floor in Phase 2 probably dates to the fifth century CE. This date seems to be corroborated by a fragment of a neck-less Byzantine cooking pot with an everted rim (Fig. 13:1) dating to the fourth–fifth century CE, found on Floor 138, sealed beneath the Phase 3 floor. A curved metal tool, possibly a hoe blade, was also discovered on this pavement (Fig. 14). 
 
Phase 3: Several internal modifications were made to the structure (Figs. 4, 15). A new partition wall (W108; length 5 m, width 0.6 m) was built above W140, separating the room with the mosaic from the paved room to its north. The wall was built of a combination of large ashlars, probably in secondary use, and medium-sized fieldstones. A concave basalt stone (Fig. 7:1) was found in secondary use between the stones; it was probably a bottom grinding stone, worn from use. This wall was wider than overlying W140. Its southern side was built above the edge of the mosaic and slightly covered the pavement, but did not damage the mosaic, and therefore, it seems that the floor continued to be used during this phase as well.
Two new walls were built in the northern room, dividing it into three separate spaces. The two walls severed the pavements of Phase 1 and abutted W108 to the north. One wall (W107; length 1.8 m, width 0.6 m), built of a single row of large stones, was constructed above Pavements 138 and 122 in the eastern part of the room, and another wall (W127; length 1.8 m, width 0.6 m), in the western part of the room, severed and separated Pavements 122 and 137. Wall 127, built of small fieldstones, abutted the western end of W108 and was a continuation northward of the line of the building’s western wall (W128), c. 10 cm to its west.
A new stone pavement (L121) was installed in the northeastern cell above Pavement 138, used in Phases 1 and 2. The new pavement was built of small and medium-sized fieldstones and abutted the three walls of the room. A fragment of a basalt bowl in secondary use (Fig. 7:2) was found between the paving stones. An intact clay lamp dated to the fifth century CE was found on the pavement (Figs. 13:7; 16, 17; Hadad 2002:37–50, Type 19).
Two additional walls, in the western part of the building, divided the open area into three separate spaces: a wall consisting of a single row of large stones (W141, preserved length 1 m, width 0.6 m) in the northern part of the area and a wall in the southern part, built of small fieldstones (W135; length 2.5 m, width 0.6 m). Wall 141 was constructed above the pavement from Phases 1 and 2 (L137), while the southern wall cut the western wall of the building from Phases 1 and 2 (W131). Bag-shaped jars dating to the Late Byzantine period (sixth–early seventh centuries CE), including both cylindrical and southern jars, common mainly on the Carmel coast and in the Sharon (Fig. 13:4–6), were found in the layers of fill (L133, L134, L136) abutting these walls.
Stone collapse exposed in the northern part of the building, both in the courtyard area east of W112 (L117) and on the stone pavement (L109, an accumulation above Pavement 137, not in the plan), are probably evidence of the building’s destruction. Additional pottery vessels, mainly jars, kraters and cooking pots from the Byzantine period (sixth-beginning of the seventh centuries CE; Fig. 13:2) were found in the fill above the pavements and among the collapse. A glass bottle with a cylindrical neck ending in a rounded upright rim dating to the end of the Late Roman–Byzantine periods was also found in the collapse. Another bottle rim and several fragments of glass vessels dating to the same periods were also found in the fills in the open area west of the building: an everted rim of a delicate bowl made of yellowish-green glass with a turquoise trail wound horizontally below the rim; a small fragment of a ring base belonging to a wine goblet; and two small beads (not drawn). Other artifacts include part of a Canaanean sickle blade (in secondary deposition, not drawn) and a fragment of a basalt bowl (Fig. 7:3). 
 
No architectural remains were found in the entire area east of the building, and it seems that the area was open during all the periods, possibly an external courtyard connected to the building. In the center of the area was a surface made of small fieldstones (L119) on which were sherds from the Roman and Byzantine periods, such as cooking pots (Fig. 13:3), as well as two Canaanean sickle blades (abraded and in secondary deposition, not drawn) and a relatively large quantity of bones (Marom, below). It seems that this was not a habitation level but rather the bottom of a large refuse pit. A few small fragments of basalt grinding stones (not drawn) were found among the stones. It is possible that some of the activities related to the ongoing operation of the building, such as grinding flour and cooking, were carried out in this area, as evidenced by the relatively large number of basalt bowls and grinding vessels that were discovered in all three phases of its use (Fig. 7).
 
The results of the excavation do not allow us to determine with certainty how the building was destroyed—whether intentional or it was destroyed after it was abandoned. It is clear, however, that the site itself was abandoned at the end of the Byzantine period and was not re-inhabited. Among the bones found at the site was a single pig bone and one bone with horizontal cut marks that are characteristic of a non-Jewish butchering method (Marom, below), which alludes to a gentile population at the site during this period.
 
Ottoman Period. This area remained open after the building was destroyed. No signs of a settlement in the Early Islamic, Crusader or Late Islamic periods were revealed in the excavation. The only evidence of any activity in later periods is a large pit that severed the floors of the building (L125) in which Gaza jar potsherds were found that dated the digging of the pit to the Ottoman period (Fig. 13:8, 9).
 
Archaeozoology
Nimrod Marom
 
The excavations in the Byzantine building on the fringes of Tel ʽAfula yielded a small assemblage of animal bones that were identified per biological species and skeleton part (Davis 1992: Table 1), and were measured to the extent possible (Driesch 1976: Table 2).
A total of 50 bones were recorded, some collected from the surface (N=19) and others, from fills that probably dated to the Late Roman period (N=11). The remaining bones were found in assemblages from the Byzantine period (N=20) and reflect debris that was discarded in the pit (N=8) and accumulations in other fills (N=12). The number of bones identified from each period is very small and insufficient to substantiate a quantitative analysis of the data.
Several remains of a donkey (N=5), cattle (N=9) and sheep/goat (N=5) were found on the surface; of the latter, three bones were identified as belonging to goats. The donkey and the cattle are mainly represented by teeth and hoof bones (N=12)—parts yielding little meat, discarded as butchering waste in the area when that part of the site was on the outskirts of the settlement. The bones of the sheep/goat are represented by parts of a humerus and a scapula (N=4), which are rich in meat. All the bones gathered from the surface are of adult individuals.
The bones from the Roman-period fill represent the lower body parts of cattle (N=8), including one bone (metacarpus) of an individual less than two years old. Two of the bones were identified with serious pathologies, indicating the animals were used for work such as carrying loads and plowing. Of the three bones of the sheep/goat, two are humeri of individuals older than one year and one is the jaw of a two-year-old individual identified as belonging to a sheep. Two of the cattle bones bore signs they were cut with a knife, which, during the Roman period, is characteristic of butchering in a rural environment; in larger population centers, it was customary to dismember the animals with a cleaver. On many of the bones (N=4) from the Roman fills there was evidence they had been chewed on by dogs, and three bones were broken when they were fresh, which was probably deliberate to remove the marrow.
The layers of Byzantine fill were rich in taxonomic terms, and included cattle bones, particularly those of the lower limbs and the head (N=7). Two of the bones (including one from the pit) displayed marks caused by dogs chewing on them, which is indicative of tertiary deposition: after consumption, the bones were discarded in a place accessible to the dogs, and only later were they placed in the pit and in the fill. Knife marks were observed on one of the digits of the cattle, which were caused when flaying the animal's skin. In addition, a horizontal cut mark was found on the ventral side of the axis vertebra. This was apparently caused when slaughtering the animal and is evidence of a distinctly non-Jewish method of slaughter and of the ethnicity of the inhabitants in this part of the site. Apart from one bone that belonged to a young individual, all the bones in the assemblage were of adult animals. Most sheep/goat bones (of which two goats and one sheep were identified) represent lower limbs and heads of adult individuals (N=5). One of the bones bears cut marks of a knife, and two (including one from the pit) were chewed on by dogs. Here, too, we can relate this to tertiary deposition of the assemblage’s bones in these layers of fill. In addition to the bones of sheep/goat and cattle, several pig bones (a jaw bone of a two-year-old) and the bones of a rooster, a goose and a horse (one digit) were found in Byzantine contexts.
 
In conclusion, it appears that the faunal assemblage from the Roman and Byzantine periods consisted mainly of slaughtering waste of adult cattle and sheep/goat (mostly goat), which were dismembered by an inexperienced butcher. This waste was stratified in tertiary depositions, most likely some distance from where the animals were eaten. The composition of the assemblage, and especially the predominance of cattle accompanied by goats, attests to intensive agricultural activity in the vicinity whence the animals were brought to the site (Redding 1984), but it is important to remember that due to the nature of their deposition, there is no connection between the bones and the consumption of the food by the inhabitants of the building where they were found. The presence of a bone belonging to a young pig in a pit from the Byzantine period probably indicates a non-Jewish population at the site at the time.
 
Table 1. The number of bones identified by species in the assemblage, according to period
Taxon
Stratigraphic Context
Surface Level
Roman
Byzantine
Cattle
 9
 8
 9
Sheep/goat
 5
 3
 7
Pig
 
 
 1
Horse/donkey
 5
 
 1
Rooster
 
 
 1
Goose
 
 
 1
Total
19
11
20
 
 
Table 2. Measurements (in mm)
Taxon
Period
Bone
Measurements
Cattle
Byzantine
M3
L M3=38.7
Cattle
Late Roman
Astragal
Bd=36.4; GL1=58.5; D1=32.7
Cattle
Late Roman
Phalanx I
Bd=27.7; Bd=26.6; G1pe=55.7; SD=23.4
Cattle
Byzantine
Phalanx I
Bd=25.2; G1pe=55; SD=22.5
Cattle
Late Roman
Phalanx II
Bp=26.5; Bd=22.3
Cattle
Late Roman
Phalanx II
Bp=22.7; Bd=20.2
Horse
Byzantine
Phalanx I
Bd=54.6; Bd=45.7; BFd=43.9; GL=84.1; SD=32.4
Sheep
Byzantine
Metacarpus
WCM=13.1; DVM=18; DEM=12.2; BFd=27.7; WCL=12.8; DVL=17.6; DEL=11.6
Goat
Byzantine
Astragal
D1=17.4; GL1=32.4; Bd=19.6
Sheep/goat
Byzantine
Metatarsus
BFd=28.7
Rooster/hen
Byzantine
Tibiotarsus
Bd=12.4
Goose
Byzantine
Coracoid
GLm=55.5
 
The excavation on the southern fringes of Tel ʽAfula revealed part of a large residential building with stone pavements and a mosaic floor, indicating that its inhabitants were probably members of the upper class. The results of the excavation show that the structure was built in the fourth century CE and continued to be used throughout the Byzantine period, when internal changes were made, such as the installation of the mosaic floor and the addition of interior room partitions. The building was abandoned in the early seventh century CE and was not inhabited again.
Previous excavations conducted next to the current excavation area exposed part of a large residential quarter from the Byzantine period that included intersecting streets aligned north–south and east–west and buildings, both residential and workshops constructed alongside them (Fig. 1: Permit Nos. A-6515 and A-6716). Some of the rooms of these buildings had floors covered with paving stones and mosaic pavements, but so far only white mosaic floors have been exposed. According to Dalali-Amos (pers. comm.), this residential quarter was established at the beginning of the Byzantine period and existed continuously until it was abandoned at the end of the period (early seventh century CE). The building exposed in the current excavation, likewise Byzantine in date and oriented north–south, was probably part of that residential quarter. Its presence there indicates that residents of the upper class also lived in the quarter. Based on the small assemblage of bones, it is reasonable to suggest that non-Jewish residents resided in the building.