The University of South Florida Excavations at Sepphoris took place from June 12 to July 14, 1995 under the direction of Professor James F. Strange. Senior Staff included Thomas R.W. Longstaff, Associate Director, C. Thomas McCollough, Assistant Director and Field Archaeologist, Arlene Fradkin, Zooarcheologist, Richard Kremer, Architect, Mary Lynn Jones, Registrar, Joan Keller, Glass Historian and Glass Registrar, Alysia Fischer, Glass Technologist, Ronald Levy, Surveyor, Winfred McGee, Tools, and Carolyn Strange, Administrator. Area Supervisors included David Brown, Alysia Fischer, Jennifer Gibson, Alan Hix, Lynn Penley, Barbara Pilcher, Jason Rech, Sharon Sammons, John Serrage, Lora Sorkin, James R. Strange, Katherine Strange, and Elizabeth Strange.
The excavations for the 1995 season focused on the three areas of the peristyle building (a schematic drawing of which appears at the top of this page) in anticipation of completing the stratigraphical and architectural history of this massive structure. The first area was the northwest corner, where we anticipated exposing the intersection of the west and the south walls and the sidewalk/street running along the perimeter of the walls (Field V, Areas 96, 97, 98, & 99). The second area was the northeast corner of the building where we had previously exposed portions of a bath complex which appeared to be installed in the Byzantine I period (Field V, Areas 104, 105, & 106). This complex interrupted the paved Roman street running along the east perimeter of the building and appears to overlay and incorporate portions of the peristyle building. The third area was interior space along the east and south walls of the building (Field V, Areas 81, 91, 100, 101, 102, & 103). In this area we anticipated two results. The first related to the Arab and Byzantine reuse of the space along east and south perimeter of the building. Previous excavations had exposed extensive small industrial installations dating to the Byzantine and Arab periods that overlay and at points seem to displace the earlier building. The second was to gain a clearer picture of the configuration of space inside this part of the building. One of the problems to be solved is the architectural solution to the elevation differential between the west (264.00) and the east (261.00) sides of the structure. The excavation of the areas over the interior space could provide some clues to solve this architectural puzzle.
The excavation of the first area (northwest corner) proved very fruitful. The architecture that we had anticipated was exposed and the stratigraphical record recovered was consistent with earlier patterns. Modern fill (as deep as 1-2 m. in areas farthest to the east [V.98]) roadbed construction and in two cases modern Arab village housing (Field V, Areas 97 & 98) were excavated to expose the ancient materials. As in the past, the excavation of this side of the building yielded little of the Arab I/II occupation. Byzantine occupation was in evidence in the robbing of the west wall (V.96), the resurfacing of the street along the west perimeter of the building (V.96), and robbing of interior walls in V.98. The Late Roman/Middle Roman periods were exposed in two areas, V.97 and V.96. In V.96, the continuation of the outer west wall was recovered with plaster on the interior face. As in previous cases of exposing this plaster, we again noted the multilayers, beginning at least with the phase of the building when the mosaics were installed (3rd-4th century). A small fragment of this plaster showed signs of colored fresco. The soil loci from the bedding of the paving stones from the street and sidewalk yielded Early Roman pottery. It was clear from the excavations that these surfaces, while disturbed in later periods (e.g., pavers removed), nonetheless remained in use at least through the Byzantine II period.
The excavation of V.97 exposed a polychrome
mosaic floor and standing plaster that defined an interior room (approximately
5 x 5 m.) located on the north perimeter of the building. This mirrors what
we found along the south perimeter, and we therefore project a series of these
small rooms (shops?) along the entire north wall in the Middle Roman/Late Roman
phase.
The areas to the north and east of V.97 yielded remnants of the Early Roman phase of the building. Area V.99, the area immediately to the north, exposed bedrock which had been cut and the inside face plastered. The alignment with previously exposed fragments of the north and west outer walls suggested we had found the foundation for the northwest corner of the building. Early Roman ceramics around the bedrock foundation gave us further stratigraphical evidence to support an Early Roman foundation for the building. This area also exposed stratigraphic evidence of an Early Roman street (lime-packed surface) running along the north wall. This seems to be the continuation of the street we had exposed in V.6 and would establish a grid at the intersection of the Early Roman street we recovered beneath the paved road running along the east wall of the building.
Area V.98, the area immediately to the east,
had its Middle Roman/Late Roman phase destroyed (the mosaic floor of V.97 appeared
in the west balk) exposing an Early Roman plaster surface, foundations for Early
Roman interior and outer walls and drain, and street surface. The street appears
to be the continuation of the street exposed in the balks of Areas V.99 and
V.6. This is the street running along the north perimeter of the building which
would have formed an intersection with the Early Roman street exposed beneath
the Middle Roman monumental paved road running along the east perimeter. We
have suggested this is indicative of an intentional grid plan used in the construction
of the Herodian city. The Early Roman plaster floor is mirrored in Area V.53.
This may indicate that these corridor areas normally used plaster floors, and
more broadly it may suggest general use of plaster floors inside the building
in the Early Roman period prior to the 3rd C. installation of mosaic floors.
The excavation of the northeast corner also proved productive, especially in terms of the Arab and Byzantine occupation and reconfiguration of this portion of the building. The excavations in this area of the building in 1993 and 1994 had revealed that later occupants of the space had interrupted the north-south road and destroyed a portion of the building in order to erect a bath complex. Areas 104, 105 and 106 were opened to reveal more fully the bath and its stratigraphical and architectural relationship to the building. In short we found that the space had three phases of occupation related to the building of the bath. In the late 5th and early 6th centuries, at the time that larger public buildings were being constructed elsewhere at Sepphoris, paving stones from the Roman road and building stones from the peristyle building were torn from their original location and used in the construction of a bath. To this point, we have recovered only the hypocaust pillars and fragments of a floor from the original Byzantine I construction.
In the second phase, the Arab I phase, the bath appears to have suffered some destruction (perhaps owing to the Persian invasion) and rebuilding. We recovered several architectural remnants of this phase of occupation. In area 105, a shallow pool paved with serpentine stone was built into the Byzantine bath complex. The function of the pool is unclear but it may relate to ritual foot washing. In four areas (Field V, Areas 101, 102, 103, & 104) we exposed a portion of a polychrome mosaic floor that appears to have been constructed in the Arab I phase covering drains and other architectural elements of the Byzantine bath. Areas 104 and 106 also produced evidence of Arab I reconstruction of the floor of the hypocaust. A reconstruction of the architectural and historical events that led to the construction of the bath in this part of the building is yet to be established. After the destructive events of the mid-fourth century, the building was evidently in significant disrepair, and perhaps it was the easy access to building materials (robbed from the peristyle structure) that helped determine the location. It may also be that access to an ancient water source or to the water system of Sepphoris determined the location. This too remains speculative since a water source is no longer visible and the layout of the water system within the city is yet to be established.
The areas opened inside the outer walls of the building along the south and east perimeter, produced further evidence of Byzantine and Arab reuse of building space and materials for industrial and domestic occupation. Plaster work surfaces and reused stone defined small rooms from the Arab I, II and Byzantine phases and surfaced in V.81 and V.100. Scatted remains of glass manufacturing continued to appear and, given the as yet unpublished evidence of temporary glass furnaces of the same period recovered elsewhere, it now seems even more likely that the east perimeter of the building housed a glass-making industry from the Byzantine period through the Arab occupation.
The soil excavated below the Arab and Byzantine surfaces in V.81 yielded a barrel- vaulted cistern that appears to have been plastered and constructed in the Early Roman phase of occupation and continued in use through the Arab II phase. A large wall of bossed, Herodian stones running east-west and parallel to the cistern may form one side of an original entrance to building from the street.
The further excavation of V.91 (opened in 1994) exposed foundations for architectural elements along the south outer wall. Foundations for cobbled surfaces made up to the south outer wall were exposed and dated to the Early Roman period. At least a shadow of the foundation of the south wall was also exposed along with foundation for later (Byzantine I) walls that overlay and at points interrupted the earlier structures.
In sum, as noted at the outset, in terms of architecture, the season was especially productive. The northwest corner of the Early Roman building has been firmly established and the streets and sidewalks along the west and north have been exposed and phased to the Early Roman period. Our projection of small rooms along the perimeters of the building have been confirmed and show the anticipated phasing from Early Roman foundation to Middle Roman/Late Roman elaboration (e.g., mosaic floors) and Byz/Arab reuse as industrial and domestic space. The bath complex in the northeast corner was shown to have gone through three phases of construction and use beginning with Byzantine I and ending with Arab II.
(Submitted by James F. Strange, Thomas R. W. Longstaff, and C. Thomas McCollough)
Return
to the Archaeology Web Page