Combined Caesarea Expeditions 

Areas TP and LL in the 2000 Excavation Season

By: Kenneth G. Holum, Jennifer A. Stabler, and Farland H. Stanley, Jr. 


The Combined Caesarea Expeditions conducted a six-week season in 2000, from June 5 to July 12. The project engaged a staff of 31, (1) more than 130 volunteers from the U.S., Canada, Israel, and several other countries, and for three weeks, a group of eight to ten Arab workmen from Jenin. Participating institutions were the Universities of Haifa, Maryland, Oklahoma, Kansas, and South Dakota, Concordia College, and Dominican University in Illinois. Directors were Kenneth G. Holum of the University of Maryland and Avner Raban of the University of Haifa. Financial support for the project came from the Joseph and Mary Keller Foundation, the Dorot Foundation (for volunteer scholarships), and the directors' respective universities. The volunteers were the backbone of the labor force, and their fees made up a good part of the field budget. Our gratitude to them is boundless.

On land, the excavators worked in about equal numbers on two projects at Caesarea. Area LL is the site of a large port-side horreum or warehouse founded (apparently) in King Herod's time, rebuilt in the 4th century, and replaced gradually after the Muslims conquered Caesarea in the 7th century, by a dwelling quarter that developed continuously until the 12th and 13th century Crusader period (fig. 1). Area TP is the Temple Platform, the site, successively, of King Herod's temple to Roma and Augustus, of an Early Christian Church, of a Muslim-period dwelling quarter, and then of Crusader Caesarea's Cathedral of St. Peter and associated monumental buildings (fig. 1). The report that follows presents discoveries during the 2000 season: For convenience, this report adopts the phasing system developed by our colleagues from Haifa University-although we believe that dividing the report in this manner often misrepresents the continuous and organic development of an urban site.

Area TP (the Temple Platform - fig. 2)

This area is located on the low, flat-topped ridge that rises about 13 m on the east of King Herod's main harbor basin and the Inner Harbor (area I). In the early 1960s Avraham Negev, excavating on behalf of the Hebrew University, Jerusalem, cleared the site of remains of the Bosnian village established within Caesarea's medieval walls in the 1880s and abandoned in 1940. Beneath former Bosnian houses Negev exposed ancient and medieval remains, including, on the southeast, the ruins of Crusader Caesarea's Gothic Cathedral of St. Peter. Walking across the site in 1998, K. G. Holum of the University of Maryland and Y. Tsafrir of the Hebrew University observed concentric octagonal foundations, as well as marble columns, capitals, revetments, and fragments of chancel screens, and estimated that they belonged to a previously unrecognized Early Christian church. To judge from its position-near the city's geographical center, above the Inner Harbor, at the highest point in Caesarea's generally flat terrain-this must have been one of Byzantine Caesarea's most impressive Christian monuments. Forming a partnership with A. Raban and Haifa University, Holum began exploring the site in 1989. Excavation proceeded in 1989-92 and 1995-2000, a total of ten seasons so far.

The area TP project proved even more illuminating than expected. In 1990-91 the project located remains of King Herod's temple to Roma and Augustus, mentioned by Josephus (Jewish War 1.414, Jewish Antiquities 15.339). It was thought at first that the Early Christian church replaced the temple ca. 500 C.E., but excavations since 1995 revealed an enigmatic building or complex called-for want of a proper identification-the "intermediate building," that replaced the temple on the same site and preceded the church. The church collapsed in the eighth century, perhaps in the earthquake of 749, to be replaced by a Muslim domestic occupation that resembled Muslim occupation in areas I, LL, and elsewhere inside Caesarea's medieval fortifications, and this in turn gave way to another monumental phase in the 11th or (more likely) the 12th century, represented by broad and deep foundations that probably represented large vaulted structures, eventually related to the Crusader cathedral on the Temple Platform's southeast flank built in the 12th century.

The strategy in the 2000 season was to address unanswered questions about all occupation phases in area TP except the earliest (Hellenistic). In the course of the summer, the excavators worked in a total of ten 10 x 10 m squares, most of them explored already in earlier seasons. Squares excavated were TP27 and 30 on the north side of the temple and church; TP9, 10, 23, and 25 generally in the northeast central part of the site; and TP7, 11, 28, and 31 in the southeast, south, central, and southwest part of the area TP. No structures or soil fills came to light that dated to the Hellenistic phase, when the town of Straton's Tower stood on the site. The 2000 excavations did, however, bring to light important new evidence for all other periods in Caesarea's ancient and medieval history.

Phase XV, Herod and Archelaus: 25 B.C.E.-6 C.E. (fig. 3)

In TP27 and TP30, between 4.75 and 9.9 m north of the temple stereobate, more fragments of the building's superstructure came to light where they had apparently fallen when the building was dismantled ca. 400 C.E. (fig. 4). The elevations of these fragments, ca. +10.69 to +11.48 m, presumably correspond roughly with the level of the temenos pavement that surrounded the temple, or, more likely, with the fill beneath the pavement-the pavement itself having been robbed out before the temple itself was pulled down. These fragments were three fragments of stucco with molded fluting that had covered one of the columns of the temple's north colonnade. Like the larger fragment uncovered in 1999, this stucco had separated from the column of kurkar segments, presumably when it collapsed northward. The column segments had been removed for reuse, but the unserviceable stucco remained where it lay (fig. 5). The other two fragments, which lay about one column length north of the stereobate, were a fragment of a kurkar architrave, and possibly a fragment of a Corinthian capital, likewise in kurkar. These fragments join a small but growing group of pieces clearly associated with the temple that have enabled Edna Amos and Anna Iamim of the Caesarea staff to propose a new reconstruction of the temple's original order.

The 2000 excavation also brought to light further evidence of the temple's stereobate itself, which is preserved, for the most part, only as leveling courses resting on bedrock. In TP10, for example, partly-exposed wall 10329 was two courses of kurkar stones, in the upper course measuring ca. 0.6 x 1.2 m, oriented east-west and laid in line with the well-preserved stereobate foundation in TP13 (fig. 6). On its north face one of the stones displayed remains of a boss reminiscent of the bosses in TP13. These courses, top elevation +8.95 m, were cemented to bedrock at +7.98 m by a layer of hard, dark gray ashy mortar ca. 4 cm thick. Also exposed in 2000, on the south side of the stereobate, was 28037, a matrix of kurkar stones set in dark gray mortar at +11.21 m that was presumably a leveling and binding layer for the foundation just above bedrock (fig. 7). Finally, excavators found the southwest corner of the Herodian stereobate in TP7. Here, a leveling course of kurkar blocks was discovered beneath foundation wall 7009 associated with the Byzantine church. It was bonded to the bedrock with the same type of gray mortar employed in TP10 and TP28 (fig. 8). The top elevation of the leveling stones was +10.6 m, and the natural level of the bedrock was +10.00 m, higher than to the north but apparently lower than in TP28 to the east. In TP7, as the bedrock sloped downward to the north, successively larger stones were employed to create a level upper surface for constructing the stereobate. Below the west face of the stereobate foundation the bedrock had been quarried straight down more than 3 m to +6.90 m. The south edge of the quarrying corresponded with the south edge of the temple stereobate. Perhaps the bedrock was quarried to receive a structure associated with the west approach to the temple.

Phase XIIb, Early Byzantine A: 340-420 C.E. 

Above the foundations of the temple stereobate, and surrounding it on all sides, the excavators have found no structures that post-dated the temple's construction and predated its apparent destruction in ca. 400 C.E. (2) Though no additional evidence appeared in 2000, it is thought that the destruction or dismantling represented by the collapse northward of a column and architrave in TP27 and TP30 (above) occurred near the end of the 4th century or the beginning of the 5th. This destruction entailed not only dismantling the temple's superstructure, and setting aside many of its stones for reuse elsewhere, but also thorough robbing of the stereobate right down to the leveling courses and mortar matrices cemented onto the bedrock. (3)

Shortly thereafter, by about 420 C.E., a new building or complex of buildings went up on the Temple Platform that occupied much of the space of the temple, especially its north half, but differed from it in orientation and extended beyond it (fig. 1). The function of this building or complex is utterly unknown, (4) so the archaeologists have dubbed it the "intermediate building" because of its chronological position midway between Herod's temple and the octagonal church. So far, evidence for it consists of 18 or 19 relatively uniform segments of "foundation." These are a light gray cement matrix containing kurkar and limestone chips, small cobbles, shell, potsherds, and flakes of carbon. The upper surfaces are virtually all smooth and display no traces of stones attached to them, or of floors abutting them. The widths of the segments vary from 0.35 to 0.7 m and their depths from 0.5 to 0.7 m, and all were set in a fill above the robbed out temple stereobate at ca. +11.50 m. If this was the approximate elevation of the as yet undiscovered "intermediate building" floors, a 1-2 m fill must have been brought in to level the area after robbing of the stereobate foundations.

During the 2000 season, a 6 or 7 additional segments of "intermediate building" foundation came to light, bringing the total to 18 or 19 segments. Newly uncovered were TP 09115 and 09135 (fig. 9), both continuations of 25056 excavated earlier; also 23208 and 28031; and related segments 10311 and 10351 (fig. 10), the first to extend the "intermediate building" on the north beyond the confines of the Herodian stereobate. In addition, part of mortar 31014, similar in construction and located at the southeast corner of the octagonal church, may have belonged to the "intermediate building." Apart from these segments of "intermediate building" foundation, three fill layers removed in TP7 may have dated from the same construction phase. These were yellow sand 7309, 7312, and 7314 containing 4th-5th century pottery and coins, located at +10.81 to +6.90 m in the void west of the temple foundation where the bedrock was quarried to receive the foundations of an unknown structure along the temple's west facade. If this association is correct, presumably the unknown structure had been robbed out when the rest of the temple was dismantled and the resulting void was filled in when the site was leveled to receive the "intermediate building."

As in earlier season, slight evidence appeared in 2000 that the "intermediate building," immediate predecessor of the octagonal church, was destroyed by fire later in the 5th century. In TP28, just above 28031, a suspected segment of "intermediate building" foundation, the excavators isolated 28028, a thin layer of dark gray soil containing ash, identified as a "burn" layer, that may represent burning of organic material in the building's superstructure or roof.

Phase XI, Middle Byzantine: 490-540 C.E. (fig. 11)

The octagonal church, original objective of excavations in area TP, dates to about 500 C.E., to judge from the rich numismatic evidence recovered in previous seasons. A glance at the building's plan reveals that it consisted of a) an outer octagonal wall and inner octagonal foundation for a colonnade with corner piers, and b) a series of rectangular rooms on foundations that form two concentric squares, within which the plan of the octagons was inscribed. Despite expressions of a contrary view, meticulous examination of the building's extant foundations confirms that a) and b) date to the same episode of construction ca. 500 C.E.

The 2000 season yielded substantial new evidence for the entire complex. In TP30 the excavators exposed more of foundation 30027, 80 cm wide, which supported the outer octagonal wall of the church, and of 30024 and 30066, which apparently supported walls flanking the church's north entrance (fig. 12). Wall 30022, 60 cm wide, part of the outer square that formed rectangular rooms surrounding the church, preserved evidence of 6.0 m wide opening aligned with the church's north-south centerline and the center of a threshold block still in position above foundation 30027. The opening in wall 30022 apparently accommodated double doors, represented in the stones by curved grooves for hinges on both sides (fig. 13). In TP10 wall 10352 represented the outer octagonal foundation, and in TP9, TP10, and TP27, wall 9140, 1.2 m wide, is the inner octagonal foundation that supported the stylobate (fig. 14). In TP27, at the northeast corner, this foundation preserved the imprints of two stones in the characteristic white mortar that cemented its upper courses that probably represent a corner pier corresponding to one that is better preserved at the northwest corner. Further west, ca. 1.5 m from the TP27 corner pier, the excavators found 23222 cemented to the top of 9140 (fig. 15). This was a kurkar block, square in plan, 70 x 70 cm, and 60 cm high; revetted with marble slabs, it probably supported a marble column base and column that formed part of the colonnade above the inner octagonal foundation. The team also found imprints in the white mortar of a corresponding block 2.4 m to the west that would have supported the west column of the pair supporting an arch above the north-south centerline, in line with the threshold preserved on wall 30027 to the north and the newly-discovered double doors further north above wall 30022. Thus, important new details emerged in the 2000 season about the superstructure and the interior arrangement of the church.

In TP7, on the southwest, the excavators exposed more of 7009, ca 60 cm wide, one of the foundation walls that represent rectangular rooms surrounding the octagonal church. Here it was laid, at +10.60 m, directly on top of the leveling course for the robbed Herodian stereobate. In TP31, on the southeast, the outer foundations of the same series of rectangular rooms was represented by wall 31001, likewise 60 cm wide, laid in a foundation trench, 31010 and 31012, that extended up to 30 cm from the wall and contained ceramics dating indiscriminately to the 4th-7th centuries (fig. 16). Similarly, in TP28 the team identified a foundation trench, 28034, extending 45 cm from inner octagonal foundation 2180 at +11.76 m down to the temple stereobate at +11.31 m, containing ceramics of the 5th-6th centuries. This foundation trench penetrated part of layers 28027, 28029, 28050, 28051, and 28054, a fill excavated between +12.50 and +11.36 m containing pottery from the 5th century. These were layers of imported fill that lay alongside and above mortar foundation 28031 and above the "burn" layer 28028. Like corresponding layers elsewhere, they represent a massive effort to raise the ground level of the site before construction of the octagonal church. That the foundation trench of the church foundation penetrated part of this fill suggests that destruction of the "intermediate building" and leveling the site began long before church construction commenced.

The 2000 season also recovered numerous fragments of the marble slab floor of the octagonal church, or at least of the mortar matrix in which the slabs were set or of its makeup. In TP27, for example, a fragment of marble floor, 27010, elevation +12.83 m, was lifted and the makeup beneath it explored. Beneath the gray mortar matrix, 27009, was a cobble and hamra layer, 27025, at +12.67 m, and beneath that a thin mortar layer, 27028, lenses of hamra, 27034, and a compact layer of crushed kurkar, 27029, at +12.45 m (fig. 17). The construction in TP28 was similar. Mortar matrix 28021 and 28043 preserved only a few fragments of marble at +12.83 m. This matrix rested above a thin layer of compact soil, 28023 and 28047 and a thick layer of hamra, clay, and cobbles 28026, at +12.51 m. As the excavators discerned them, the layers did not correspond exactly, but the overall picture was clear enough. Other fragments of marble church floor studied in 2000 were 10348, 23188, 23194, 23221, and 23224. As elsewhere, the elevations were consistent in the range of +12.73 to +12.93 m (fig. 18).

Finally, for the first time the team identified a fragment of the pavement that surrounded the octagonal church and its flanking rectangular rooms. In TP31, at the southeast corner, the excavators explored pavement 31002, consisting of large kurkar slabs laid at +12.40 m directly against foundation 31001 (fig. 16). Excavation beneath part of this pavement revealed several fill layers containing ceramic material dated indiscriminately to the 4th-7th centuries, but it is clear from the structural relationships that the pavement belonged to the occupation phase of the church.

Phase IX, Latest Byzantine: 590-640 C.E.

As understood last year, it appears that an annex was built beyond the north portal of the church about a century after its original construction. This is represented by the north extension wall 30066, by its parallel on the west, 30024, and by pavement 30031 between them, of which more was exposed this year (fig. 12).

Phase VI, Tulunid, Ikhshidid: 870-940 C.E.

In 2000, the excavators found no evidence for the collapse of the octagonal church, thought to have occurred in the 8th century, perhaps in the famous earthquake of 749 C.E. After collapse of the church, a period of abandonment and robbing presumably ensued, for which (again) we found no evidence in 2000. This was followed by the establishment of a domestic quarter on the Temple Platform that resembled that in areas I, LL, and elsewhere in the Old City of Caesarea. Subsequent occupation and archaeological activity destroyed much of the evidence for this phase, so it is represented mainly by subterranean structures: drains, sink pits (or "soak-aways," for disposal of waste water and runoff), cisterns, wells, and (eventually) grain storage bins.

The earliest such structure was a sink pit in TP7 built on the west side of wall 7009 and using it as its west side (fig. 19). Walls 7035 and 7303 formed the north and south sides of the sink pit, which originally had a vaulted roof and measured 2.65 x 4.5 m and was preserved to 4.3 m in depth. It contained two fill layers, 7304 and 7313, that indicate a process of infilling and abandonment early in the 9th century. Apart from this, only a suspected garbage pit in TP9 (9116) and a few disconnected surface fragments and layers in TP30 were associated with the Tulunid-Ikhshidid occupation phase.

Phase V, Early Fatimid: 940-1000 C.E.

During this period, Caesarea developed into a flourishing Muslim town, and the Temple Platform clearly participated in the upswing. An indication of prosperity was the construction of subterranean grain storage bins associated with large courtyard houses. Typically, the storage bins were laid against existing foundations and were built of crude outer walls of kurkar masonry lined with rectangular slabs that fit tightly together and were fixed to the outer wall by a thick layer of mortar. The floor was usually a sturdy tessellation of large limestone tesserae. Such bins were apparently thought to be safe from penetration of moisture or vermin and were, thus, suitable for storing grain. One such bin, later used as a cistern, came to light in TP9 (previously designated bin 4116, since its southeast corner penetrated into TP4). The bin's original dimensions were 2.5 m north-south by 5.6 m east-west. Later, it was subdivided and converted into a cistern (below). Apart from this bin, only disassociated walls and fills in TP30 were dated to this phase during the 2000 season.

Phase IV, Mid-Fatimid: 1000-1050 C.E.

In TP23, a pavement of kurkar slabs, 23116, was recorded at an elevation of +13.29 m. It measured 1.7 m east-west and 4.85 m north-south. Ceramic evidence from soil layer 23182 beneath it, dated this pavement to the 11th century. It may have represented the interior courtyard of the house associated with grain bin 4116. In TP30, well 30057 may have been dug during this period. It was square in plan and lined with kurkar stones (fig. 12). Its top - not well-head, however - came to light at +12.50 m, and the fill within was excavated but contained little pottery. Dating depended on fills and fragmentary structures surrounding the well, and is therefore unreliable.

Phase IIIb, Late Fatimid: 1050-1101 C.E.

In TP30, sink pit 30006 probably dated originally to this phase. This was a vaulted structure, 2.1 m north-south and 1.5 m east-west, built of dry-laid kurkar blocks (fig. 21). Inside the sink pit fill 30096, excavated from +11.17 to +10.17 m, contained little pottery but many animal bones and lenses of fish scales. This deposit ended on 30108, the crushed kurkar floor of the sink pit. Near the bottom on the southeast corner was an opening that allowed a connection, through a cutout in foundation wall 27012 of the octagonal church, with sink pit 27013 on the south side of the wall. Fill 30087, against the east side of the sink pit, contained much pottery dating to the 11th century, suggesting the dating. A smaller sink pit, or perhaps a drain channel, 30075, was built in the north part of TP30, ending at later wall 30043 on the north and apparently abutting 30022 in the south. Fill 30101 beneath it contained 11th century ceramics, indicating a date for the sink pit or channel.

Phase IIIb, Crusader, 1st Kingdom: 1101-1187 C.E.

During the Crusader period, general construction traditions continued at Caesarea with little change, but in area TP the fundamental occupation pattern changed from domestic architecture to monumental buildings associated with the Crusader cathedral on the southwest. The main building type associated with this transition was large rooms flanked by broad, deep foundations that presumably supported vaulted superstructures. Although there is still debate about the dating, this transition is probably represented in the 2000 discoveries by the large plastered room excavated in TP8 during the 1992 season and explore further in 2000 in TP11 (fig. 22 and 23). This room measured, overall, 6 m north-south by 9.65 m east-west. A thick layer of white plaster, poured over small cobbles, formed its original floor at ca. +12.83 m, and the plaster continued up the room's original walls. On the walls, incised herringbones indicated that another layer of plaster had been affixed that had fallen away without a trace. In TP11, wall 11004 was the west wall, 11116 the north wall, and 11132 the south wall. The team dug a probe in the northwest corner, between 11004 and 11116 in order to explore the relationship between the two walls, and hopefully, to secure additional dating evidence. In this probe, the plaster floor lay above a layer that contained 11th century pottery, but this does not settle the dating issue. It turned out that only the three upper courses of 11116 abutted 11004, while lower courses were set back, but there need be no doubt that the two walls were contemporary. A similar phenomenon came to light earlier in TP2 with abutting, apparently contemporary walls. A circular depression in the plaster floor to the east, 11116, may have contained a tabun or ceramic storage vessel (pithos) during this period.

Phase II, Crusader 2nd Kingdom: 13th century C.E.

Several fragmentary structures in TP30 apparently dated to this phase. Pavements 30026 and 30089, in the north part of the trench, lay on or near the modern surface at +12.95 and +12.80 m. Fill 30045 to the north of them contained a tumble of building stones, perhaps from adjacent wall 30005, and 13th century pottery. This fill lay against east-west wall 30043 to the north, which abutted 30005 on the east and wall 30050 on the west. The function of these walls and surfaces is unknown. Dating to the same phase were two drains. Fragmentary stone-lined drain 30063 ran north-south between walls 30005 and 30066, but its source and destination were not preserved. Ceramic pipe drain 30065 came from the south, ran over wall 30066, and apparently deposited its flow in stone-lined drain 30064, which led to pit fill 30068, perhaps part of a sink pit limited on the east by wall 30070 (fig. 12). Fill 30068 contained much pottery, including Crusader pieces, as well as eggshells, animal bones, a metal spur, a loom weight, and a zoomorphic object resembling a frog.

It is likely as well that the upper fill within sink pit 30006 dated to this period. This was 30090, containing 13th century ceramics as well as a large volume of fish scales. This represents the progressive infilling of the sink pit during the 13th century.

In TP11 the plastered room was provided with a new, higher floor, and a buttress was installed against wall 11004, presumably to support vaulting over the room and the street to the west. The new floor consisted of two layers of kurkar stones, 11121 and 11122 (fig. 24). The stones of the former were so irregular that they could not be considered a distinct phase of flooring but only makeup for a floor, while the stones of the higher floor, at +13.22 m, were more regular. Ceramics from between the stones and beneath them dated to the 13th century. Remnants of a compact mortar and soil layer, 11120, perhaps represented the actual surface of the higher floor. Buttress 11117 consisted of three courses of kurkar stretchers containing a rubble fill laid against 11004. Two courses lay below the new floor level and one survived above it.

In TP7 the later Crusader phase was represented by a large east-west wall, 7024, to the west of Byzantine wall 7009 (but continuing to the east of 7009, excavated in the 1991 and 1992 seasons) (fig. 25). Apparently, 7024 was the north foundation of a large east-west vaulted hall built adjacent to the north to the Crusader cathedral. In TP7 this foundation was 2 m wide and was constructed of kurkar blocks of various sizes laid carefully on the outer edges but loosely in between, contained in a rubble of kurkar stones and mortar. Mortar was also found between the courses. Six courses were excavated in 2000 at elevations from +12.40 down to +10.41 m. Part of 7024 extended over the earlier, presumably abandoned sink pit from phase VI (above). To avoid collapse into the pit, the builders added a buttress wall, 7032, against the north side of 7024, of which 12 courses were discovered, extending from +12.16 m down to +8.86 m. Further, a square buttress, 7024, was added on the east side of 7024.

On the north side of 7024 a fill layer, 7301, was dumped into the abandoned sink pit in order to create a level surface for the pavement of a courtyard to the north of the vaulted building represented by 7024. Portions of a large courtyard had been discovered in earlier seasons in TP1, TP12, and TP13. The top of this fill was at +11.29 m, and it contained pottery dating to the 13th century. Finally, a stone-lined pit, 7084, was sunk in the 13th century into the floor level of the vaulted building against the south side of foundation wall 7024. It contained much 13th century pottery and a section of pipe.

Structures from this phase were also studied during the 2000 season in TP9 and TP23, all of them, probably, related to the Crusader monumental phase evident in TP7 and TP11. Pavement 9006=23175, made of kurkar slabs at +13.60 m, extended 5.80 m east-west from the west side of north-south wall 9051. Ceramic evidence from fill 9108 beneath it dated to the 13th century. It abutted wall 9051, which thus dated to the same phase. North-south wall 9097, north-south wall 23105 to the west, and east-west wall 9095 on the north. Alternatively, it may have functioned as a buttress against wall 9074. In the same sector, subterranean grain bin 4116 from phase V was reduced in size by division wall 9136, provided with vault 9083 that permitted wall 9074 to rest on top of it, and was brought back into use as a cistern measuring 2.1 m north-south by 2.42 m east-west (fig. 26). Access to the cistern was by a square shaft embedded in the superstructure of wall 9074, which must therefore have been encased in a niche within the mass of the wall. Fill 9124 within the cistern contained pottery from the 13th century, presumably reflecting the progressive silting of the cistern during its period of use. Dating the cistern to the 13th century thus also helps to assign broad wall 9074 to the same phase.

Phase Ib, Mamluk: 14th-15th centuries C.E.

In TP11, wall 11116 was dismantled and its remaining upper stones were incorporated in a stone pavement that extended over TP8 and TP4 to the north. In TP11 this pavement was 11115 at +13.68 m, made of dry-laid stones robbed from earlier structures. The fill below, 11119, contained lenses of ash and lime as well as Mamluk pottery. The function of the extensive Mamluk pavement remains unknown. 

In TP30, a stone structure of unknown function, 30076, was laid against wall 30005. In design it consisted of three steps of kurkar stones, ascending to the south.

Phase Ia, Modern: Late 19th to Early 20th centuries

In TP11, soil layers 11114 and 11118 represented this phase, at +13.52 to +13.20 m. The higher fill overlay kurkar stone rubble, probably representing wastage as stones from dismantled structures were prepared for reuse in the Bosnian period. Two modern iron fence posts were embedded in the fill layers.

In TP30, three soil layers were diagnosed as modern. Locus 30042 was backfill from the 1999 excavation, and 30046 below it probably dated from the Negev excavation. It contained a section of rebar.

No excavation is planned for 2001. Instead, team members working in area TP will evaluate finds from ten seasons of excavation with a view to planning the final report volumes on the TP excavation. It is likely that further excavation will be required in subsequent seasons to clarify debated issues.

Area LL (Warehouse and Domestic Quarter)

This area is located to the north of the Inner Harbor basin and on the east side of the main harbor basin (fig. 1 and 27). It was first investigated in the 1970s by Lee Levine and Ehud Netzer of the Hebrew University, who designated it the Central Excavation Area. Levine and Netzer identified a Byzantine warehouse fronting on the Inner Harbor and a limestone-paved street running east-west along the north side of the warehouse. Various phases of Islamic and Crusader residences were also noted. Avner Raban and Robert Stieglitz, directing the Caesarea Ancient Harbour Excavation Project (CAHEP), returned to the area in 1986 and 1987 to examine a pier west of the warehouse complex. Several piers and walls were noted on the west and south sides of the warehouse that dated to the Herodian period (25 B.C.E. - 6 C.E.). A set of four Islamic storage bins dating to the late 10th or early 11th century was excavated in the southwestern part of the warehouse.

The Combined Caesarea Expeditions (CCE) began excavations in Area LL in 1996. The major goal of this ongoing effort is to study a residential or commercial district adjacent to the harbor dating from the Herodian period. Since the area was converted from a warehouse in the seventh century Byzantine period to a residential district in the early eighth century Islamic period, new information was expected on Byzantine-Islamic transition.

Trenches LL1, LL2, and LL3 were new trenches opened in 1996 along the south side of the Hebrew University excavations to follow features previously exposed. Trenches LL1 and LL2 were reopened in 1997 and Hebrew University's trench E/5 was reopened as CCE's trench LL4. Work resumed in trenches LL1 and LL4 in 1998 and Hebrew University's trench E/4 was reopened as CCE's trench LL5. In 1999, trenches LL1, LL4, and LL5 were continued and LL6 and LL7 were opened to the north of LL4 and LL5.

In the 2000 season, excavation continued in LL1 in the corridor portion of the large Byzantine warehouse. The main objective in this trench was to clear the Islamic and Crusader structures from the warehouse's corridor to expose Byzantine and any earlier remains in that area. A Byzantine floor surface for the corridor was identified and was sectioned in the northern part of the trench.

Work continued in trench LL7, where an Islamic residence was uncovered in the 1999 season. The main objective was to clarify the layout of the Islamic residence and to study urban planning in the Islamic phase in area LL. Once the Islamic phase was comprehended, the objective was to remove portions of the residences to reveal any earlier structures below. We were hoping to reveal buildings from the Byzantine and Roman periods to the north of a large stone paved street that runs through the area.

Trench LL6, in the northwestern portion of area LL, was continued from the 1999 season. This trench also contained remains of an Islamic residence that was first identified in the 1999 season. Excavations in 2000 aimed to delineate the features and layout of the Islamic residence and then to proceed downwards to earlier levels. 

Trenches LL8 and LL9 are located to the west and north of LL6. A fortification tower was found in the northwestern portion of LL6 in the 1999 season. Trenches LL8 and LL9 were opened in order to expose more of the tower and to clarify its phasing.

Phase XVI, Late Hellenistic: 100 - 25 B.C.E.

Several fill layers containing ceramics dating to the first century B.C.E. were excavated in LL6 and LL7 (6003, 6098, 6135, 6137, 6138, 7078, and 7086). The fills were found in the northern portions of the trenches, ranging in elevation from +5.89-+4.80 m. No structures were found in association with the fills.

Phase XV, Herod, Archelaus: 25 B.C.E. - 6 C.E.

Portions of the foundation walls of a Herodian warehouse, walls 1402 and 1411, were revealed in LL1 below Byzantine warehouse walls 1266 and 1217 (fig. 28). The Herodian walls were reused as the foundations for the Byzantine warehouse walls. The foundations consist of very large kurkar stones set on top of the natural bedrock.

Phase XII - X, Early to Late Byzantine: 340-590 C.E.

The Byzantine warehouse in LL1 was found to date to about 400 C.E. from previous excavations. In the northeast and northwest corners of the warehouse corridor, two large square structures (1256 and 1398) were exposed, consisting of kurkar stones tightly mortared together (fig. 29). Their function is uncertain, but they were added to the warehouse some time after 400, when the Byzantine renovations occurred. The upper portions of these features were dismantled at a later date when a new tessellated floor, 1400, was laid in the corridor. This floor consisted of white tesserae with an 8 cm wide border of red tesserae, varying in elevation from +3.60 to +3.65 m. A small section was cut through the floor to search for datable materials. The tesserae were set into a bed of mortar overlying a layer of hamra and cobbles, used as a leveling surface. One coin with the emperor Constantius II (351-61 C.E.) was found below the floor, along with ceramics dating to the 3rd century C.E. A firm date could not be assigned to this floor due to the dearth of datable material. The tessellated floor appears to have deteriorated in antiquity and was then covered with a plaster floor (1305, 1391, 1395, and 1406), varying in elevation from +3.76-+3.86 m. This plaster floor covered the square structures, 1256 and 1398 in the corners of the corridor. Numerous coins and coin blanks dating to the 5th and 6th centuries C.E. were recovered from the various floors above the tessellation.

Portions of an east-west wall and a tessellated floor that were identified in LL6 and LL8 possibly date to the Byzantine period. Wall 6114/8012 is located in the southern part of LL6 and LL8, bordering the north side of a wide limestone paved street associated with the Byzantine warehouse to the south (fig. 30). The wall is 53 cm wide and is built in a header-double stretcher pattern, characteristic of the Byzantine walls of the warehouse. North-south wall 8041 is 47 cm wide and bonds with wall (fig. 31). Walls 8012 and 8041 are plastered on the inside. Loci 8040 and 8045, which are also wall fragments, may also belong to the same phase as 8012 and 8041. These walls may define a room within another warehouse located on the northern side of the east-west limestone street. Further excavation is necessary to clarify the function of these walls.

A white mosaic floor, 6106, was identified to the north of east-west wall 6114 at +4.56 m (fig. 32). The mosaic abuts wall 6114, but was not excavated. Therefore, its phasing is uncertain. Another east-west wall, 6130, was exposed in the northwestern part of LL6. The wall is 1.2 m wide and was preserved to a height of +5.15 m. It may date to the Byzantine phase, but further excavation is necessary to provide a firmer date.

Phases IX, Latest Byzantine, and VIIIb, Transitional: 590 - 690 C.E.

The warehouse appears to have been abandoned shortly after the Muslim invasion in 640 C.E. Numerous storage jars were left abandoned inside the warehouse, where sand began to accumulate. Portions of this abandonment layer were excavated in 1998 as loci 1240 and 1242. A small section of the abandonment layer was left in the eastern warehouse room in LL1 below pedestaled features from the Islamic period (loci 1357, 1360, 1382, 1393, and 1397). Numerous coins and ceramic sherds dating to the sixth and seventh centuries, were recovered from this abandonment layer, further indicating that the warehouse was abandoned around the time of the Islamic conquest.

Phase VIIIa, Umayyad: 690 - 750 C.E.

Remains from the Umayyad phase were identified in LL8 and consisted of a plaster floor, 8038, associated with walls 8040 and 8045. The floor was found at an elevation of +4.80 m. The fills, 8038 and 8042, excavated below the floor contained ceramics dating to the 8th century C.E. This possibly suggests that an earlier Byzantine building was reused in the early Islamic period.

Phases VII-VI, Abbasid and Tulunid, Ikhshidid: 750 - 940 C.E.

Remains from the early Islamic phases were noted in LL6, LL7, LL8, and LL9. In LL7, a horseshoe-shaped sinkpit (locus 7105) was excavated from +5.89-+4.80 m. Ceramics from the fill inside the sinkpit dated to the 9th century C.E.

In LL6, a north-south wall (locus 6134) or dry-laid stretchers, is located in the north central part of the trench and abuts an earlier east-west wall, 6021. Five courses of wall 6134 were preserved to a height of +6.14 m. 

In LL8, walls 8015, 8017, 8024, 8025, and 8030 were built in this phase and formed a room in a residential unit (fig. 31). This structure was later converted into a storage bin, probably some time in the 10th century C.E. To the west, the remains of a plastered cistern (loci 8019, 8021, 8022, and 8023) measuring 3.20 m north-south by 2.28 m east-west, were exposed at +4.99 m. 

Phases V-IV, Early to Mid-Fatimid: 940 - 1050 C.E.

Remains from the Early-Mid Fatimid phases were noted in LL1, LL6, and LL8. Well 1309 was exposed in the 1999 season and was excavated this summer (fig. 28 and 33). The contents at the base of the well suggest that it was constructed in the tenth century and remained in use through the Crusader period. Fifteen courses of dry laid stones lining the inside of the well were excavated from +5.11-+1.49 m. The shaft was excavated another meter into the kurkar bedrock, reaching a depth of +0.41 m. The well was filled in the 12th century C.E. Cistern 1316 was also exposed in the 1999 season, but was not excavated until the 2000 season. Ceramics recovered from the fill around the outside of the cistern suggest a construction date in the 11th century. The fill inside of the cistern dates to the 13th century and indicates that this feature was used throughout the late Islamic and Crusader phases.

A sink pit, locus 1385, was partially excavated in the 1999 season and continued in the 2000 season. This feature was interpreted this summer as a well that was converted into a cistern. The fill excavated this summer inside of the sink pit dates to the 10th century. Two walls in the northern part of the corridor were completely excavated in the 2000 season. East-west wall 1224 with foundation 1346 bonds with wall north-south wall 1225 with foundation 1257. These walls were dated to the 11th century in the 2000 season and formed two rooms in the northern part of the corridor of the warehouse, after that structure was converted into a residence.

Cistern 6004 and the upper portions of well 6036 were probably constructed in this phase (fig. 34). A ceramic pipe runs from the cistern to the well, suggesting that the well was used to catch water overflowing from the cistern. The well was constructed at an earlier date, but was renovated in the tenth century to accommodate the overflow from the cistern. 

In LL8, the room bordered by walls 8015, 8017, 8024, and 8025 was converted into a storage bin. The imprints of the square stone liners are visible in the soft white mortar used to line the interior of the bin with a watertight seal.

Phase IIIb, Late Fatimid: 1050-1101 C.E.

Remains dating to the Late Fatimid phase were excavated in trenches LL1, LL6, and LL7 in the 2000 season. A stone passage, locus 1278 was exposed in the 1999 season and was completely excavated in the summer of 2000. The ceramics found in association with this passage indicate a date of construction in the 11th century C.E.

In trench LL6, a sinkpit bounded by walls 6005, 6115, 6050, and 6089, was exposed in the 1999 season and the interior excavated in 2000 (fig. 30). Ceramics recovered inside of this sinkpit indicate that this feature was in use from the 10th to 11th centuries and was dismantled some time in the 13th century. This sinkpit was originally connected to sinkpit 7043, which lies to the east. These two sinkpits were separated at a later date, when a larger residence was probably divided. 

Pavement 6022 was exposed in the 1999 season and was excavated in the summer of 2000. This pavement lies directly above cistern 6004 in the northern part of LL6. The fill found below the pavement contained ceramics dating up through the 13th century. The cistern was likely constructed in the 11th century and continued in use through the Crusader phase. A settling basin, 6071, also associated with the cistern, was excavated in the 2000 season and was dated to the 11th century based on the ceramics found below it. 

Features from the Late Islamic phase in LL7 include a possible surface at +5.57 m, located adjacent to wall 7064 in the west central part of the trench. Another hard packed surface, locus 7075, was also identified at +5.66 m. The soil fills below these surfaces contained pottery dating up to the 11th century. A possible settling basin, locus 7109, and staircase, 7107, were excavated in the south central part of the trench. The basin appears to post date the supposed staircase, as part of the fill below it covered a lower stair. The main feature in LL7 dating to the Late Fatimid period is a large courtyard, locus 7012, paved with white tesserae and reused flat marble fragments at an elevation of +6.09 m. An earlier paving, locus 7115, was found below the tessellated courtyard, but was not excavated this season.

Phases IIIa-II, Crusader: 12th - 13th Centuries C.E.

Remains from the Crusader phase were identified in trenches LL1 and LL7 in the 2000 season. The Islamic residence in LL1 was reused in the Crusader period. Sinkpit 1233 was reused in this phase and was probably constructed at an earlier date. It was not excavated to its base, so its date of construction remains uncertain. Most of the other Crusader features in the corridor of the warehouse were excavated in the 1999 season.

Remains from the Crusader phase were very fragmentary in LL7. Features dating to this phase include two drains, two walls, and one floor surface. A north-south drain, 7029, with capstones 7028, ran through the north central part of the trench at an elevation of +6.42 m (fig. 35). This drain ran over several Islamic period walls and canceled those features. Another east-west drain, locus 7009, was excavated in the east central part of the trench. This drain was located at an elevation of +6.37 m and is probably associated with drain 7029. The soil surrounding this feature contained ceramics dating to the 13th century.

A partial plaster floor, locus 7110, was excavated from +5.95-+5.83 m in the northern part of the trench. Another fragmentary kurkar surface was noted in the north central part of the +6.39 m. This floor appears to be associated with a sinkpit in the northeast corner of LL7. This sinkpit is composed of walls 7053, 7054, 7055, 7056, and 7097 (fig. 30). The fills excavated inside of the sinkpit contained ceramics from dated to the 13th century. 

Excavations in area LL in the 2000 season yielded material from the foundation of the city by King Herod in the first century B.C.E. up through the Crusader period. Many of the Crusader and Islamic features in LL1 were removed to expose Byzantine levels in the corridor level of a large warehouse complex that occupies the southern portion of area LL. Further excavation of the later features in LL1 will expose more of the earlier levels in future seasons. Some of the Islamic features that were dated to the Umayyad phase in the 1999 season were found to be later in date and should probably be placed in the 10th century Islamic phase. 

Excavation in trenches LL6 and LL7 clarified the phasing of some of the medieval features, which were found in fragmentary condition due to subsequent robbing, probably for the late Crusader fortification walls. Some earlier fill levels dating to the Herodian period were reached in the northern sections of LL6 and LL7, but no associated walls or floors have yet been found. Remains of several large walls running east-west through both LL6 and LL7 may belong to another Byzantine period warehouse. Excavation has not progressed far enough in these areas to more accurately define these features. Excavation of the medieval features will continue in future seasons to expose the remains of earlier structures in LL6 and LL7.

Two new trenches, LL8 and LL9 were opened in the 2000 season to identify the extent of a large fortification tower exposed in the 1999 season and initially phased to the Crusader period. The northern and southern walls of the tower were further defined and portions of the inside area was excavated. The remains found inside of the tower suggest that this structure was possibly built in earlier, in the 11th century Fatimid period rather than in the Crusader period. Future efforts will attempt clarify the phasing of this structure. 

Since summer 2001 will be devoted to study of remains from area TP, the next anticipated season in area LL, as is area TP, is 2002. Much work remains to clarify what has already been uncovered in this area, and there is promise of rich rewards to knowledge about all periods of Caesarea's history from continued exploration in area LL. 

1. Caesarea 2000 field staff on land were Ronni Toueg, stratigrapher; Farland H. Stanley, Jr. and Clayton M. Lehmann, area supervisors; Jennifer Stabler, Audrey Shaffer, Jerry Black, Clayton Lehmann, Marion Brew, and Tina Hill trench supervisors; Robert Hutchins, Heidi Racine, Stephen Murphy, Janet Davison, Stacey Poulos, Matthew Hill, and Brad LeMarr, assistant trench supervisors; Yael Arnon, registrar and ceramics; Laurie Brink, associate registrar; Barbara McCauley and Hagith Sivan, assistants to the registrar; Anna Iamim, chief surveyor; Peter Lampinen, numismatics; Edna Amos and Debra Taylor, architectural fragments; Tatiana Meltsen, drafting; D. Charles Smith, photographer; J. J. Gottlieb, darkroom; Jennifer Ramsay, palaeobotany; Carole Cope, palaeozoology; Janet Davison, Janet Blit, and Radha Dalal, administrators; Chad McDaniel, computers and database. 

2. Hence a series of layers found in TP30 that contained 1st and 2nd century ceramics probably represented fill imported at the time of destruction or shortly after. These were 30041, 30051, 30082, and 30109

3. The exception to this thorough robbing was in TP12 and TP13, in the NW, where the underlying bedrock sloped downward rather sharply. Here 4 stereobate courses survived above the bedrock, creating a level subsurface foundation at +10.70 m upon which foundations of the octagonal church were later set. Why these courses were preserved through the robbing episode ca. 400 C.E. remains a mystery. 

4. A priori one might wish to identify the "intermediate building" as a church, but nothing about its structure suggests monumental architecture or an ecclesiastical function-i.e. there is no indication of nave or bema.