2010g: Late Byzantine Remains near Shiqmona

‘Atiqot 63, 2010 Late Byzantine Remains neaR shiqmona: a monasteRy, a CemeteRy and a WinepRess Raz KLetteR intRoduCtion Following plans to widen the Haifa–Tel Aviv highway opposite Tel Shiqmona, a salvage excavation was carried out from December 1999 to February 2000 (map ref. NIG 196/747, OIG 146/247).1 The excavation was carried out immediately east of the highway, in an area about 8 m wide and 300 m long, from Ha-Toren Street in the north to Zarfat Road in the south (Fig. 1). The northern part of the excavated area is an exposed rock escarpment, while the central and southern parts lie on the more gentle western slope of the Carmel mountain, and were covered by a municipal garden. This area was part of the Shiqmona cemetery excavated by Elgavish (1994). The excavation revealed inds from the late Byzantine period, including a monastery, a large winepress, a dozen rock-hewn tombs (robbed), and two large buildings (see Fig. 1). Late Byzantine-period remains had been discovered near Tel Shiqmona (Elgavish 1968; 1974; 1977; Dauphin 1998:665–667) and in salvage excavations between the tell and the Carmel Mountain (Hirschfeld 1998; 2006; ‘Ad and Torge, forthcoming). Elgavish (1994) published a summary of his digs, but has not yet published a inal report of the Byzantineperiod remains. Our most important ind was the re-discovery of a chapel with mosaic loors (Ovadiah and Ovadiah 1987:132, No. 221) that had been excavated in 1939–1940 by Makhouly on behalf of the British Department of Antiquities of Palestine (Makhouly 1944; cf. Kletter 2006b:46–51). Although it was covered and fenced, the fence disappeared and, when studied ifty years later, Peleg (1988:25) reported that “all the remains have since been destroyed”. The present excavation proves not only that most of the chapel survived, but that it was part of a much larger building (not entirely excavated), most probably a monastery. Together with the monasteries reported by Dothan (1954–1955) and ‘Ad (pers. comm.), there is valid evidence of a concentration of monasteries related to Shiqmona. This strengthens the view that Shiqmona was a city during the late Byzantine period and not a village, as dozens of late Byzantine monasteries were located in proximity to cities, such as Jerusalem, Bet She’an, and Bethlehem. While this phenomenon is noted here, it merits a separate study, such as those that have been prepared for desert monasteries by Hirschfeld (1992) and Dahari (2000). Following the excavation, the architecture was dismantled and the area was released for development. the monasteRy (Fig. 1; Plans 1, 2) The Chapel The chapel was discovered in 1939 by N. Makhouly, a supervisor on behalf of the British Mandate Department of Antiquities of Palestine. Makhouly reported a large mosaic loor near the Haifa–Tel Aviv highway, which was threatened by damage due to exposure. It was excavated by Makhouly in 1940 and surrounded by a fence for protection. Makhouly 148 196 400 196 375 196 350 196 300 196 325 Raz KLetteR T1 T2 747 750 T3 T4 T7 hig hw ay T7B T5 T6 T8 Area D1 –T el A viv 747 725 Cemetery Ha ifa Monastery Building B 747 700 Area D2 747 675 270 000 230 000 190 000 800 000 150 000 Building A Sea of Galilee 760 000 Haifa Shiqmona 720 000 747 650 Caesarea 680 000 Mediterranean Sea Tel Aviv Winepress 747 625 T12 0 20 m 640 000 Jerusalem 0 Dead Sea 40km Fig. 1. Location map. identiied the mosaic as part of a small chapel and published a short notice (Makhouly 1944). He left a written report with plans, photographs, and suggestions for restoration (today in the IAA archive, Mandatory File: Tell es-Samak). Piecemeal publication and references to the discovery followed. The mosaic loors were published in a monograph on mosaic pavements in Israel (Ovadiah and Ovadiah 1987:132, No. 221, Pls. CLVII–CLIX). The building was 149 Late Byzantine Remains neaR shiqmona W16 l Av iv hi Haif a – Te L217 W25 L204 W5 ghw ay L203 W7 L206 W27 W8 0 2 m Plan 1. Makhouly’s plan of the chapel with our wall and loci numbers (IAA archive, Mandatory File: Tell es-Samak). discussed in the Corpus of Byzantine Churches in the Holy Land (Ovadiah 1970:165, Site No. 165) and a schematic plan appeared in a supplement to this work (Ovadiah and Gomez de Silva 1984:162, No. 49[165]). Makhouly (1940; 1944) mentioned ribbed Byzantine pottery in association with the chapel, but none was published or kept (Peleg 1988:27, n. 5). Almost ifty years after the excavation, Peleg (1988) published the site. Peleg (1988:25) believed that “all the remains have since been destroyed”; fortunately, however, this was not the case. What happened was that the fence disappeared and the mosaics and walls were covered by a thin layer of earth. No remains of the chapel were visible when we began the excavation (Fig. 2), although most of the building, including the mosaics, remained 20–30 cm beneath the surface. This history demonstrates the temporary nature of inal reports, but as the architecture was dismantled after the present excavation and the mosaics were removed for restoration, this report functions as the inal publication of the chapel. It must be remembered, though, that the chapel was part of a larger building, remains of which may still exist farther east, beyond the limits of the excavation. The plan of the chapel was published by Ovadiah and Gomez de Silva (1984:162; cf. 150 Raz KLetteR L200 10.67 10.82 W 10.60 4 W 6 10.00 L201 10.82 L208 L209 10.39 10.27 Area D1 9.50 9.19 W16 9.18 L217 L402 9.47 W40 9.39 9.18 10.29 L219 L204 10.02 W5 W45 9.37 L403 Coffin W44 9.62 W25 9.58 9.42 L203 # W7 10.47 W9 W27 L206 10.18 9.96 10.02 W8 L207 L218 # 9.80 10.03 L226 10.12 # 10.26 W21 # 10.03 10.00 # L224 W17 W20 9.78 9.94 # W23 9.79 L225 9.91 10.26 10.06 L228 9.91 4 m W1 0 9.65 10.25 9.97 L212 Plan 2. The Byzantine monastery. 10.32 Late Byzantine Remains neaR shiqmona 151 Fig. 2. The area of the chapel at the beginning of the excavation, looking north (entrance to T6 is in the background). Peleg 1988:26, Fig. 1) and is shown in Fig. 1 with two nearby buildings (A, B; see below). Makhouly’s plan is presented in Plan 1. Almost all the chapel survived intact, except for the area west of Walls 5 and 27 that was damaged either by the widening of the Haifa–Tel Aviv highway in the 1960s, or by erosion, as the area sits on the edge of the highway’s drainage trough. Parts of W5 and some other reported indings were not found; perhaps they were removed by Makhouly or stolen soon after. Makhouly prepared plans for the reconstruction of the chapel, which are in the IAA Mandatory ile. He assumed the chapel was a small independent building, and therefore his reconstruction does not entirely match the newly gathered data. The chapel (Plan 2) comprises a large room with a mosaic loor (L206; c. 2.5 × 5.5 m; Fig. 3), and north of it is another room (L203) with remains of yet another mosaic (Fig. 4). We found the concrete bases of the Mandatory- period fence that surrounded the mosaics with only rusted stumps remaining from the iron poles. The stones of the apse wall that separated Room 203 from Room 204 on the west were robbed before Makhouly’s excavation, but the wall could be discerned by the semicircular contour of the western edge of the mosaic in L203. A hard gray plaster loor was found in L203 and L204 (Fig. 5). The loor reached W5, which served as the base of a marble chancel found by Makhouly together with one of the marble bases of a door in the center of W5. This base and the northern part of W5 are lost; however, we found many marble fragments in the chapel surroundings, some probably originating from this chancel. The mosaic in L206 was almost intact when found in 1939, except for an area near its southwestern corner (Ovadiah and Ovadiah 1987: Pl. CLVII:2). Yet, by the time it was excavated in 1999, some 15–20 percent of its 152 Raz KLetteR Fig. 3. The mosaic loor in the side aisle of the chapel (L206), looking south. Note the round concrete base of the Mandatory-period fence on W8 in upper part of photo. Fig. 4. Room L203 with mosaic in east, concrete loor in center, and W5 to west. area was lost. The causes of this deterioration are unclear, but it seems that it was damaged after it was exposed and before it was protected. It was not an act of professional antiquities robbers as the parts that are missing cut across motifs and left the most beautiful ones intact. The reasons were probably vandalism or decay during the short period of exposure. Only the covering of the mosaics by earth protected them from further decay. Late Byzantine Remains neaR shiqmona 153 Fig. 5. Excavation beneath (L217) the chapel loors, looking west, with the apse wall (W25) in the center and the base of W16 to the right. Fig. 6. The apse wall (W25), looking east, with W16 on the left. Both mosaics (L203, L206) were removed for restoration by an IAA team with the intention of replacing the large mosaic from L206 in an open-air museum on site. Once the mosaics were removed, we excavated beneath the loors. The mosaics were placed on an excellent base of white zifzif (crushed beachrock) mixed with organic olive pits, superposing a layer of small stones. Larger stones form the base of the apse wall (W25), between L203 and L204 (Fig. 6). Nothing of this wall survived above loor level, and it is not clear wheather Makhouly exposed its stones; his plan (Plan 1) seems to indicate 154 Raz KLetteR that he restored the apse wall according to the gap between the mosaic of L203 and the loor of L204. Peleg (1988:25–26, No. 1) also mentioned an “inscribed apse”. In our excavation, which reached bedrock, the bases of some of the chapel’s walls, such as W7, W16, and W27, were exposed. They are composed of large, well-dressed stones, aligned with the width of the wall, probably to strengthen it for support of the superstructure. The same building method was used in the foundations of Building A farther south (see below). The Finds.— There were not many inds in the chapel area. All the pottery above the covering layer of earth was Byzantine, but comprised only sherds that were close to the surface and remained close to there when the area was excavated and reilled by Makhouly. These sherds cannot be used to date the chapel and, therefore, are not presented here. There were sherds beneath the chapel loors in L217–219, but they were too small, and L219 was not sealed on the west. All these sherds were dated to the late Byzantine period (see Calderon, this volume), a date supported by the numismatic evidence (see Kool, this volume). One coin, dated to the twelfth–thirteenth centuries CE (see Kool, this volume: Cat. No. 15), was found above the loor in L204, but in a disturbed context. All the other coins are Byzantine, with a few dating to the fourth–ifth centuries, and most dating to the early–mid-sixth century. One coin, found in the loor bedding in L217, was dated until the mid-sixth century (see Kool, this volume: Cat. No. 13); this coin, however, does not give an exact date for the building of the chapel because it has a long date range and the context in which it was found was not completely secure, i.e., the area was disturbed during the dismantling of W25, which was ruined to a level beneath the loor of L217. A small, oval bronze amulet was found beneath the mosaic loor in L206 (B2225; Fig. 7). The stratigraphic position of the amulet is certain. As the mosaic loor (L206) is original, and there were no later building phases, the Fig. 7. Bronze amulet from under the mosaic in L206. amulet gives a terminus ante quem to both the mosaic loor and the building. Unfortunately, it is not inscribed, or it is too worn to show remains of letters. The shape and material of the amulet are identical to inscribed Samaritan and Greek amulets from Horbat ‘Eitayim near Nahariyya, dated by pottery to the sixth century CE (IAA No. 66.1637; Reich 1986; 1994; 2002; Pummer 1987:254, with further references therein). According to Pummer (1987:260–263), such amulets were probably manufactured by Samaritans starting in the third century CE, but were used not only by them, so the inding of one Samaritan amulet does not necessarily indicate the ethnic nature of a building or site. Furthermore, similar amulets bear Greek inscriptions (e.g., Shiqmona— Elgavish 1994:152, Fig. 139:3; Gush Halav— Makhouly 1939: Pls. 31:5,7; 32:h1, h2). The pottery found in the loci under the loors of the chapel cannot be dated within very narrow limits, but the sixth century seems to be the best dating (see Calderon, this volume). Based upon the pottery, coins, and amulet found beneath the loors, it seems most likely that the chapel and monastery were built in the irst half of the sixth century. This its the date suggested for the mosaics on stylistic grounds (late ifth-early sixth centuries CE; see Ovadiah and Ovadiah 1987:132, No. 221; Peleg 1988:27–30). The building functioned until the seventh century. Late Byzantine Remains neaR shiqmona Remains North of the Chapel Excavations north of the chapel showed that this area (L208 and 209) was an open space. The area was illed with stone debris and small marble fragments, presumably thrown there when the walls of the chapel were robbed for stones. There were two poorly built walls (W4 and W6) farther north in L200 and L201, but their direction does not it the chapel and they seem to be part of a later terrace (Fig. 8). Fig. 8. Terrace Walls 4 and 6, north of the chapel (L200 and L201). The Finds.— The many marble fragments found in L208 and L209 probably originated from the chapel. A few marble fragments were found in other loci. Many small fragments are lat with rounded ends, probably from a chancel screen (Fig. 9:1–3). All are white marble with black or dark gray veins. They vary in thickness from 6 to 14 mm. None bear any decoration. A few fragments (Fig. 9:4–7) are thick, probably from offering tables. For comparisons to the fragments in Fig. 9:4 and 9:5, see tables from a monastery west of Jericho (Netzer and Birger 1990:198, Fig. 10) and from Horbat Hesheq (Aviam 1990:359–360, Fig. 10). For comparison to the fragment in Fig. 9:6, which perhaps belongs to a round table, see Ma‘alē Adummim (Magen and Talgam 1990: Fig. 22) and Kh. ed-Deir (Piccirillo 2000:69–70). Finally, a broken decorated fragment of a capital (Fig. 9:8) was found on the surface near T2 (see below). Remains South of the Chapel South of L206, we discovered the continuation of walls and rooms that relate to the chapel, indicating that the chapel was not freestanding, but part of a larger building. It seems that a large courtyard (L218 and L226) lay south of L206, between Walls 8 and 21 (c. 5.4 m wide; the length was not fully exposed). A loor of plaster and small stones (at elevation 10.12–10.30 m) was found in the eastern part of this courtyard Fig. 9 No. Locus Basket 155 Description 1 217 2187 White marble, gray veins, 6 mm thick, end fragment 2 202 2117 White marble, gray veins, 14 mm thick 3 202 2161/21 White marble, gray veins, 11 mm thick, rounded end 4 216 2149/6 White marble, gray veins, fragment of a table 5 202 2061/1 White marble, gray veins, table? 6 5 1030 White marble, gray encrustation, 2 fragments once attached on the lower side 7 207 2085 White marble, marks of chisel at the base 8 2 2068 Broken, worn capital, white and gray marble 9 W22 2221 Basalt bowl, broken, gray, crudely made, incisions inside are modern 156 Raz KLetteR 1 3 2 4 5 6 7 9 8 0 10 Fig. 9. Marble and other stone fragments. (L207; Fig. 10). This loor did not continue west, perhaps because the area was damaged. It could have been the base of a mosaic loor, but no traces of such mosaic were found south of L206. In the southern wall of the courtyard (W21) part of an entrance was exposed. Wall 21 and its entrance were built with the same type of stones as the chapel walls. Wall 21 was ruined farther west, but its location could be traced due to patches of the plaster loor that survived to its south. North of W21, in the area with no loor, four stone fragments were found on the same level as the stone loor in Fig. 10. Stone loor in L207, south of the mosaic in L206. Note the Mandatory-period cement base at top left. Late Byzantine Remains neaR shiqmona 157 Fig. 12. Three marble slabs (B2227) found face down in L226, looking east; W21 is at back. Fig. 11. Marble slab (B2218) found face up on the stone loor in L226, near W21, looking southwest. L226. One fragment was found face up and the three others were found face down, in close proximity to each other, in the courtyard (Figs. 11, 12). These fragments probably originated from the building and were, perhaps, torn off and thrown away when it was robbed. South of W21 was a set of intersecting wall segments, creating rooms L224 and L225. The walls follow the same direction as those of the chapel, but only the bases of small- and mediumsized crude stones survived (Fig. 13). Traces of loors were found in L224 and L225. In L224, traces of burning were found, together with broken bricks, perhaps remains of a domestic installation. Many broken sherds, mostly jars, were recovered from this area, especially from L224. It is dificult to explain the narrow space between W21 and W23; it might indicate two building phases; however, the walls follow the same direction and are tied by Floor L226/207 to the chapel walls. Furthermore, there is no evidence of multiple phases (such as the raising of loors or closure of entrances) elsewhere in the building. Similar to the area north of the chapel (L208 and L209), there was no loor south of W17 (L228), and it seems that this was the southern wall of the building. There was a stone wall (W1) slightly farther south with two abutting walls from the east in an area that could not be further explored. The direction of W1 is similar to that of Walls 4 and 6 north of the chapel, but its function and relation to the chapel are not clear. The few pottery sherds found in relation to W1 (L212) were Byzantine in date. The Finds.— The four stone fragments from L226 (21.5 × 106.0 cm when aligned; Fig. 14) are 4.2–4.5 cm thick with at least one intact joint. There are three drilled holes (diam. 6 mm, 25 mm deep) in their upper edge, situated quite symmetrically at 18 cm from the left end of the aligned stone, 34.5 cm further to the right, and another 37.2 cm to the right, i.e., 16 cm from the right end. This indicates that the four fragments comprise the complete stone, and no parts are missing. The holes were 158 Raz KLetteR Fig. 13. Room in L225, looking east, below loor level; W21 is at top left. 0 10 Fig. 14. Four marble slabs from L226; the fragment on right was the only one found face up. probably used for attachment. The lower edge is more worn than the upper one, but it seems to be complete. Only the front is decorated; the back is rough, as it was hidden from view. The geometric decoration contains a row of incised triangles divided into ten smaller triangles, interspaced with six inverted and uncarved triangles. The small triangles are not identical; most are 5 cm high and 4 cm wide at the base, but the one to the very left is about 1 cm higher. The incised triangles are cut to a depth of 0.3–0.5 cm, and have a crude surface to enable painting. They were illed with a thick layer of red ochre. Samples of the ochre were collected and kept. The ochre was better preserved on the three fragments found face down, which were photographed without cleaning. The fragment found face up lost most of its ochre, and was washed before being photographed. Makhouly (1944:206) mentioned decorated fragments of marble screen slabs, but these probably stem from the chancel in W5 (cf. Horbat Hesheq, Aviam 1990:362–364). Makhouly found a marble cross that he restored as positioned on the edge of the roof. In the IAA Archive ile, Makhouly (1940) mentioned “a good number” of marble slabs which “bear decorative carvings in triangular shape” (cf. Peleg 1988:27). However, none was published and there are no drawings or photographs. I am Late Byzantine Remains neaR shiqmona 159 Fig. 15. Clay cofin (B4010) in Area D1, looking south; W40 is in the foreground, overlying a modern pipe. not aware of comparable marble fragments, but denticulated triangles are known from mosaics of the same period (Seligman and Abu Raya 2002:129, n. 5). Area D1 Shortly after the termination of the excavation, two water outlets had to be cut through the highway. The IAA was asked to supervise the work; hence, we opened two very small probes, Areas D1 (northwest of the chapel) and D2 (near Buildings A and B; see Fig. 1 and below). The excavation was conducted in haste and far from ideal conditions. In Area D1, W40 was exposed, identiied as the continuation of chapel W16; no trace of loors was found north of W40, as with W16, and the directions and heights of the loors in Area D1 and the adjacent L204 are itting (see below). Only very small segments of W44 and W45, perpendicular to W40, were uncovered. These walls, reconstructed in Plan 2 (that of W45 is very tentative), were much damaged by the highway, and neither was preserved to an elevation of more than 9.7 m. The Finds.— A surprising ind in Area D1 was a clay cofin tucked into the space between Walls 40, 44 and 45 (Figs. 15, 16). It is about Fig. 16. Clay cofin (B4010) after excavation. 0.5 × 2.0 m, with wide (c. 8 cm) horizontal rims on the long sides (part of the rim is seen in Fig. 15). The upper lip was found at an elevation 160 Raz KLetteR of 9.70–9.71 m and the bottom at 9.37 m. The cofin was open and the lid was missing. It contained scant bone fragments (unidentiied) and earth. A modern disused pipe traversed the square at 9.81 m. The cofin was probably found and robbed dozens of years ago when the pipe was placed, to judge by its rusty state. The cofin rim at 9.7 m was topped by a lid and/ or mosaic loor. Thus, the loor level would be at an elevation of c. 10 m, comparing well with the loor of L204 at c. 10.15 m. There is a 30 cm step between L204 and the adjacent L206 and 203 (at c. 10.45 m), so the smaller difference between L204 and the assumed loor in Area D1 is not problematic. As the cofin belongs to a well-known type, it was photographed, but not restored after cleaning (Fig. 17). Comparable clay cofins are deined as the ‘plain’ type, common during the second–fourth centuries in the western Galilee (Aviam and Stern 1997: Fig. 4). They appear also in Cyprus, along the Lebanese coast and in Cilicia; all were probably manufactured in the bay of Iskandrun area (Parks 2003:255–257). A corner-fragment of another clay cofin (not drawn) was found in L216 in Building B (see below). Three clay cofins were previously found at Shiqmona, in a robbed hewn tomb that had six kokhim and two arcosolia, when Ha-Toren Street was paved in the ‘En Ha-Yam neighborhood in 1965. This tomb was dated to the second century CE (Siegelman 1966:19). How can we explain a second–fourth century cofin, almost built within walls that are supposedly the continuation of a late Byzantine (sixth–seventh century), one-period building? Shiqmona was occupied in the Roman and 0 early Byzantine periods (Elgavish 1977; 1994). A rim fragment of a fourth-century, moldblown glass honeycomb beaker was found in Area D1 (see Gorin-Rosen, this volume). However, it is dificult to assume that the cofin related to an earlier building that continued to be used until the late Byzantine period as all the other buildings outside the tell roughly date to the sixth–seventh centuries. Furthermore, such clay cofins are related to burial caves, not buildings. Therefore, it seems more likely that the cofin was taken from an earlier burial and reused in the late Byzantine monastery. It was not a reliquarium, which are usually smaller stone cofins located in a chapel. Perhaps it was used to bury a dignitary or donor; such burials were common in sixth-century churches (Piccirillo 2000:61) and monasteries (Tsafrir 1984:269). Fragments of a similar clay cofin were found on the loor of a Byzantine church at Horbat Medav in the Galilee, but there the cofin was full of ashes, perhaps from reuse as a ireplace (Aviam 2002:205, Fig. 105). Not even one locus in Area D1 can be considered sealed. Small fragments of a mosaic and tesserae thereof were found in and around the cofin (Fig. 18). These were of very high quality and much smaller than the tesserae of the mosaic in L206. At least six colors were identiied: white, black, gray, red, green and yellow. A few fragments have a pattern including what may be an eye (Fig. 18: middle fragment in third row), indicating the use of igurative and not just geometric and loral patterns (unlike the mosaic from the chapel). Unfortunately, little survived from this mosaic. 20 Fig. 17. Clay cofin after cleaning. Late Byzantine Remains neaR shiqmona 0 161 2 Fig. 18. Mosaic fragments from L403. Discussion The chapel excavated by Makhouly occupied the northeastern part of a much larger building that measured at least 17.5 m on the north–south axis (between W16 and W17). We do not know the extent of its east–west axis, but it must have been more than 10 m long (W9 to W45). The function of the building is unclear. It could have been a large villa of the type found by Elgavish (1994:109–110, 116), plans of which have not yet been published. Its location on the outskirts of a town, the large chapel and the inds, as well as the existence of similar buildings nearby (see below), favor identifying it as a monastery. Other monasteries were identiied in the vicinity of Shiqmona. In her ‘inal’ report of the chapel found by Makhouly, Peleg (1988:25) writes: “Judging by its close proximity to the site of a monastery (excavated by Dothan in 1951 on behalf of the Israel Department of Antiquities and Museums), it may have been associated with this complex. It is, however, impossible to indicate the exact connection or relation between the chapel and the monastery exposed by Dothan”. The reason, not explicitly rendered by Peleg, is that Dothan did not mention an exact location. He wrote that the building was “a few hundred meters” southeast of Tel Shiqmona and under the existing highway, its western side was damaged by modern pipes, and it presumably extended farther east (Dothan 1954–1955:216). Dothan was aware of Makhouly’s chapel and his wording suggests that it is a different building (Dothan 1954–1955:222, n. 14). Fortunately, Avi-Yonah (1964:342; my translation from the Hebrew) mentioned a more precise location for Dothan’s excavation: “a few hundred meters further south [of Makhouly’s chapel], opposite the irst gate of the immigrant camp Sha‘ar Ha-‘Aliyah, were found [by Dothan] remains of a Christian monastery”. Dothan’s excavation, then, was 162 Raz KLetteR south of Makhouly’s, near the present day Zarfat Road and not connected to the building published here. In the survey map of Haifa (West; Olami, Ronen and Romano 2003:34, n. 7), the mosaic loor is mentioned as element “XI” with reference to Naveh 1958 (which is missing from the bibliography). In 1998, ‘Ad and Torge (forthcoming) excavated remains of a public building with traces of mosaic loors under the Haifa–Tel Aviv highway, west of our area. Perhaps this was part of another monastery, although only a small part could be excavated. Byzantine-period monasteries are wellknown from written sources. Dozens have been excavated in Palestine, and there is a wide variation of architectural forms and sizes (Tsafrir 1984:265–284; 1993:1–16). Perhaps the best known monasteries are those in the Judean Desert (Binns 1999; Hirschfeld 1990; 1992) and Sinai (Dahari 1994), as they are prominent buildings in an otherwise sparsely populated desert area. We do not know the precise function of our building. Monasteries ranged from small, one-space buildings, to huge complexes. The present one is somewhere in the middle. Its plan suggests a chapel (L203, L204, L206) in the northeast, a courtyard (L226, L207) in the middle, and auxiliary rooms in the south (L224, L225). BuiLdings a and B (Fig. 1; Plan 3) The area of these buildings was excavated in harsh winter conditions, and the remains were severely damaged by the road and its drainage trench (Fig. 19). Floors were not found, except one small section with no clear relation to any wall. The pottery is late Byzantine (see Calderon, this volume), but all the loci were 8 W1 9.70 Building B 9.40 9.77 10.45 10.20 10.01 W13 9.50 Area D2 9.58 9.23 9.99 9.72 L223 3 W4 9.86 10.55 9.90 2 W4 10.77 9.94 10.15 9.74 5 W1 10.15 9.77 9.63 # L222 10.06 9.75 14 W 10.14 L221 W11 L220 9.63 8.74 L227 9.10 2 W2 Building A 4 W2 10.13 9.94 0 W10 L401 1 W4 10.04 9.60 L215 4 m 10.35 Plan 3. Buildings A, B and Area D2. Late Byzantine Remains neaR shiqmona 163 Fig. 19. Buildings A and B, general view to the north. mixed or disturbed; therefore, no interpretation for the buildings can be offered, and their dating cannot be exact. Building A Building A is a very impressive structure, but only its southeastern part was found (Walls 14, 15, 22, 24). Its walls are built in the same method as the walls of the chapel, i.e., foundations of large stones, placed perpendicular to the welldressed stones of the upper courses. Walls 22 and 15 survived with three or four courses of large stones, and W24 reached an elevation of 10.35 m in the west; W14 survived to a lower height. In the southeastern corner, the walls were built on bedrock, which is higher here. The corner was completely robbed, but is evidenced by leveling marks made in the rock in preparation for building. The base of the corner rested on rock at an elevation of 9.1 m. The bases of the walls in the west and north were at an elevation of c. 8.7 m on a natural ill of red earth with many small rounded stones that may have been the natural accumulation of a small wadi descending from the Carmel. Wall 14 was about 9.5 m long; its northern side was much damaged. The northern limit of the building was not found, as it lies below the Haifa–Tel Aviv highway. The existing walls do not disclose the plan or nature of this building. There is one complete room bounded by Walls 14, 15, 22, and 24 (c. 2.1 × 5.2 m). The stone protruding inward (eastward) from W22 is not part of a wall, so L221 and L227 are parts of the same room. We did not ind loors in this room, but a patch of a lime loor east of W14 (L220; Fig. 20) is at an elevation of 10.06 m, similar to the height of the loors in L224–226 in the monastery. The Finds.— The pottery found in and near Building A is late Byzantine. One basalt bowl (Fig. 10:9) was found in W22, in a secondary location. It is of a type common to many periods (cf. a bowl from Tel Sumaq; Dar 1999: Fig. 298:29). Building B This building lies north of Building A and is mostly covered (and ruined) by the paved 164 Raz KLetteR Fig. 20. Section of a lime loor (L220) east of W14, looking east; W10 is on the left. highway. Only one wall was found (W18). Its direction is slightly different from that of the walls of Building A, and its base is at a slightly higher level. Its documented length is c. 4 m. In its center was an entrance. The stones near the southern corner of W18 are placed in a somewhat rectangular structure that seems to indicate a later addition. One line of stones (much disturbed) continues to the south where a semicircle of small stones reaches beneath terrace W13 at 9.86 m. The Finds.— Pottery sherds found inside Building B (west of W18) and in its vicinity are late Byzantine (see Calderon, this volume). Area D2 In Area D2, a small probe within the paved highway, only crude traces of the bases of Walls 41–43 were found. The loci were mixed and the relationship to Buildings A and B is uncertain. Terrace Walls East of buildings A and B there were wide terrace walls (W10, W11), clumsily built from small stones on the outside and a ill of earth and rubble on the inside. Wall 10 (see Fig. 20) had an extension to the north (W13), which was built of one row of large irregular boulders that we irst interpreted as a modern construction of the drainage trench of the highway. Walls 10 and 11 do not it the direction of Buildings A and B and are therefore later in date. There was stone debris around them (especially in L222) that included many well-dressed building stones, presumably robbed from the nearby buildings. A few rounded clay bricks of a type found in baths were found here (see Calderon, this volume: Fig. 6:67, 68). From L215 northward to W1 (see Fig. 1; Plan 2), no traces of buildings were found. Discussion Buildings A and B (especially A) could have been an extension of the buildings identiied by Uzi ‘Ad (pers. comm.) farther to the west. The buildings are dated to the late Byzantine period, and the pottery to the sixth–seventh centuries with very few possibly earlier fragments (see Calderon, this volume). The coins do not Late Byzantine Remains neaR shiqmona provide a more precise date: one coin of the fourth century (see Kool, this volume: Cat. No. 1) and one of the mid-sixth century (see Kool, this volume: Cat. No. 7) were found above the walls in L216 (above L220 and 222). Another coin from the mid-sixth century (see Kool, this volume: Cat. No. 12) was found in L223 in a mixed context. the CemeteRy (Fig. 1; Plans 4–8) A group of six tombs were hewn north of the monastery, where the rock is exposed; they were reported as early as 1863 by Guérin (1875, V:191), who mentioned that they were quite worn. The tombs share a general plan and date, with minor variations in detail. Meager remains of one more tomb (T12) were found farther south. Hewn stairs, leading eastward, climb the rock toward the hill of Kh. Tinani (T2; Fig. 21). The western end of the stairs is ruined, terminated by the road escarpment at a height of 12.12 m. The stairs reach an elevation of 14.77 m opposite the entrance to T3 (see below) and continue to rise. We mapped 16 irregular stairs 165 that average 50–70 cm deep, 10–20 cm high, and 130 cm wide. The stairs were exposed on the surface, and cleaning did not furnish any data regarding their date. The pottery found during cleaning is Byzantine, as is almost all the pottery in the excavation. As the stairs it the direction of T3 and T4 (see below), they are probably contemporary with the Byzantine burials. A more precise dating is impossible on the basis of the evidence at hand. These stairs were documented in the survey map of Haifa (West) (No. 1; see Olami, Ronen and Romano 2003:33, Fig. 22:7). Tomb 3 (Plan 4) Tomb 3, parallel to and immediately south of Stairway T2, was briely described in the survey map of Haifa (West) (Olami, Ronen and Romano 2003:28–30, Site 22, Fig. 22.2: Cave 97). It has a square standing pit and three arcosolia with single troughs. The tomb was robbed and the bones were brutally broken and mingled inside. No whole vessels were found, but many sherds were collected and partially restored. The entrance faces west with the threshold at 13.55 m asl (Fig. 22). The Fig. 21. Stairway of T2, with T3 to the right, looking east. 166 Raz KLetteR B2018 14.43 13.76 13.55 B2014 13.40 1 B2063 12.88 13.43 12.76 B2015 B2019 12.62 12.77 1 B2017 B2016 B2027 0 2 m 15.00 14.00 13.00 12.00 1-1 Plan 4. Tomb 3, plan and section. tomb had a drainage outlet toward the west, a common feature of the Shiqmona tombs (Elgavish 1994:148–149); it cut through a round winepress (diam. 1.2 m, height 0.6 m) with a small pit at the bottom (diam. 0.4 m, height 1.2 m). Such simple winepresses appear in many periods and sites (Frankel 1999). Since this winepress was cut by the drainage channel, it is earlier than the late Byzantine period, but a precise dating is not possible. Anthropological remains from T3 include fragments of crania, teeth, and post-cranial bones of at least three individuals (see Eshed, this volume). Tomb 4 (Plan 5) Tomb 4 (No. 95 in the survey map of Haifa [West]; Olami, Ronen and Romano 2003:28– 30, Fig. 22.2) lies south of T3. Tomb 4 was also robbed and broken bones and pottery sherds were left mainly in the standing pit. The plan Fig. 22. Entrance, drainage, and round winepress leading to T3, looking east. Late Byzantine Remains neaR shiqmona 167 Fig. 23. Standing pit (looded) of T4. Plan 5. Tomb 4, plan and section. is similar to that of T3, with three arcosolia of single troughs (Fig. 23). The entrance faces west with the threshold at 12.89 m (the robbers broke the entrance so the grave loods after rains). The dromos is wide and irregular with a drainage outlet toward the west (Fig. 24). Two stairs lead into the standing pit, whose loor is covered with a white ‘industrial’ mosaic. Fig. 24. Entrance and drainage outlet of T4, looking east. Anthropological remains from T4 include crania fragments, teeth and post-cranial bones of at least eight individuals (see Eshed, this volume). 168 Raz KLetteR Tomb 5 Tomb 5 (No. 96 in the survey map of Haifa [West]; Olami, Ronen and Romano 2003:31, Fig. 22.2) is similar to T3 and T4, and robbed as well. It is located between T4 and T6, but farther east and outside the area destined for development; therefore, it was not excavated. Tomb 6 (Plan 6) Tomb 6 is larger than Tombs 3–5, with four troughs in three arcosolia, as the rear arcosolium is divided into two troughs (Fig. 25). The entrance of the tomb faces southwest and the threshold is at 14.06 m asl. Two stairs descend into the standing pit. Part of the ceiling collapsed, and a natural crack now connects T6 with the lower T8 (see Plan 7). Tomb 6 is possibly Tomb 90, only briely mentioned in the survey Map of Haifa (West) (Olami, Ronen and Romano 2003: Fig. 22:2). Tomb 6 was robbed, leaving behind many broken bones (mainly in the standing pit), but few sherds and other inds. Anthropological remains from T6 include crania fragments, teeth, and post-cranial bones Fig. 25. The standing pit and rear arcosolium of T6, looking east. 15.09 1 B2046 B2052 14.01 13.91 B2007 B2010 13.74 14.12 B2048 1 B2047 14.31 0 14.04 B2029 B2039 B2038 B2051 2 m 16.00 15.00 14.00 13.00 1-1 Plan 6. Tomb 6, plan and section. B2050 169 Late Byzantine Remains neaR shiqmona of at least seven individuals (see Eshed, this volume). Tomb 7 (Plan 7) Tomb 7 (No. 94 in the survey map of Haifa [West]; Olami, Ronen and Romano 2003:32) was found at the present rock escarpment. Tomb 7a, just southwest of T7b, is a shallow depression in the rock. It may have been part of a burial, but no artifacts were found because the rock surface was exposed. Tomb 7b is an opening in the rock at elevation 13.43 m, clearly manmade. Only the opening survived; therefore, it could have been the entrance to a tomb whose hewing was not completed. Tombs 7c and 7d (Nos. 92 and 93 in the survey map of Haifa [West]; Olami, Ronen and Romano 2003:32), below Tombs 7a and 7b, are scant remains of tombs at the foot of the escarpment. Only the eastern edge of these T7c T7d 10.89 11.70 11.09 10.81 tombs survived; the rest was removed when the present-day road was paved, probably prior to 1966. Part of the rear arcosolium of Tomb 7c, divided into two troughs, survived, and the wall separating the two troughs is partially preserved. The beginning of a third trough in the south indicates that T7c originally had four troughs, similar to T6. The loor of the troughs is at 10.89 m asl. Of tomb 7d, farther to the south of the same escarpment, only a small part survived. It has a similar plan with an eastern arcosolium divided into two troughs. The standing pit is ruined by a modern electric pole. Tomb 8 (Plan 8) Tomb 8 (possibly No. 91 in the survey map of Haifa [West]; Olami, Ronen and Romano 2003: Fig. 22:2) is below T6 with an entrance facing southwest. The rock slopes sharply. The tomb lacks a large external dromos, and the threshold is at an elevation of 12.28 m (Fig. 26). Tomb 8 has three arcosolia with deep troughs surrounding a rectangular standing pit (Fig. 27). The rock was cracked, perhaps even before the tomb was hewn. Tomb 8 was probably excavated by an archaeologist, as almost no bones or broken sherds were found. 11.89 12.39 10.92 13.20 13.43 13.81 T7a 14.63 12.12 T7b T8 12.28 B2045 B2056 11.92 B2031 12.12 13.91 T6 B2026 B2011 1 14.06 0 12.22 11.54 B2035 B2044 B2040 B2057 4 m 0 2 m 13.00 15.00 12.00 11.00 1-1 Plan 7. Tomb 7, plan and section. 1-1 Plan 8. Tomb 8, plan and section. 1 170 Raz KLetteR Anthropological remains from T8 include crania fragments, teeth, and post-cranial bones of at least two individuals (see Eshed, this volume). Tomb 12 Tomb 12 is isolated from the other tombs and in the far south, between the winepress (see below) and Buildings A and B. The modern asphalt road runs about half a meter to the west, so what remained of the grave is located in the drainage channel of the road. Furthermore, the cement base of a high-voltage electricity pole ruined the standing pit. We do not know if the tomb was seen and documented when the present road was laid. Fig. 26. Entrance of T8, looking north. Only part of one, probably the eastern trough, was preserved at a height of 10.98 m asl. A shallow, smoothly hewn depression in the rock, it contained long bones heaped in secondary burial with a few skull fragments to the north. There was no pottery and the remains were near the surface. The original plan of T12 and its exact date cannot be established. Anthropological remains from T12 include crania fragments and post-cranial bones of at least two individuals (see Eshed, this volume). Discussion The basic characteristics of all the tombs are similar: a wide dromos leading to a rectangular opening with a heightened threshold. The entrances were blocked by stones (not round, as there is no evidence of channels in the rock), in which a round stone could roll (unlike Elgavish 1994:148–149, who did ind such channels). Presumably, the entrances were blocked by stone doors on pivots inside the grave (cf. Elgavish 1994:151, Fig. 135). Indeed, one socket stone was found inside the opening of T4. The doors were secured with bolts inserted into recesses in the sides of the entrances: in the case of T3, round, and in the cases of T6 and T8, rectangular. Two or three stairs lead Fig. 27. Three arcosolia and standing pit of T8, looking south. Late Byzantine Remains neaR shiqmona down to a deep square standing pit (about 2 × 2 m), surrounded by three arcosolia. Most of the arcosolia have single troughs with partially hewn and partially built walls. All the tombs were robbed or excavated in the past and their entrances were exposed (as were most of the graves excavated by Elgavish 1994:158). Anthropological remains from the tombs were delivered to the Ministry of Religious Affairs for reburial. The tombs are simple arcosolia tombs with two variations: three troughs in three arcosolia (Tombs 3 and 4), or four troughs in three arcosolia (Tombs 6, 7c, 7d and 8). The irst variation inds very close parallels in the tombs excavated by Elgavish (1994:149, Nos. 10–21), who excavated more than a dozen tombs immediately to the southeast of our area (now incorporated in a garden). With the exception of one Middle Bronze Age tomb, the tombs date to the Late Roman–Byzantine periods. Unfortunately, only preliminary reports have been published so far. Five tombs were ‘cleaned’ by the survey team of the survey map of Haifa (West) in 1966, who found them robbed or partially robbed (Olami, Ronen and Romano 2003:34). No report or inds were published from this ‘cleaning’. Simple arcosolia tombs are very common in the country, and appear in the north at Lohame Ha-Geta’ot and in the Carmel area (Tsafrir 1984:384; Foerster 1986; Avni 1997:38–39, Type 4.1). They were most common during the third–sixth centuries (Avigad 1971:190; Feorster 1986; Avni 1997:38–39). They have many variations and some are very elaborate. The existence of various types at Shiqmona better its urban centers than rural sites, which usually have a more limited variety of types (Avni 1997:39–40). The ethnic identity of the buried cannot be ascertained from the present data. The tombs were all robbed or damaged. Although the inds were not in situ, they still provide a general dating. Except for one early amphora and a few very late Medieval and modern sherds, almost all the pottery was of the 171 ifth/sixth–seventh centuries CE (see Calderon, this volume). the WinepRess (Plan 9) The area between T12 and the winepress was cleaned to bedrock (Fig. 28). The only ind on this rough rocky slope was one hewn, mostly ruined basin, perhaps of a winepress (L10). A large ‘public’ winepress was found south of T12 (Plan 9; Fig. 29), utilizing a moderate western slope of exposed rock with some crevices. The main elements include a large treading loor (L9), two installations (L6, L7) and three collecting vats (L4, L5, L8). The treading loor (L9) was covered with a white industrial mosaic, little of which survived, as the area was grazed for the preparation of a modern garden (Fig. 30). Mosaic patches were observed in and around three small depressions (e.g., Fig. 31) arranged in a line, running south Fig. 28. The cleared slope south of T12, looking south; L10 is at the bottom. The winepress is between the sculpture and the palm trees at the top of the photograph. 172 Raz KLetteR 11.12 11.00 L6 11.70 10.90 10.61 11.32 11.85 2 10.98 2 11.55 9.29 L5 9.38 0 11.24 1 10.65 8.88 11.48 11.02 11.52 7.84 9.53 L4 9.76 10.70 9.49 11.61 L8 10.89 11.49 11.26 11.28 11.09 L7 11.42 L9 11.03 11.35 11.34 11.23 11.29 11.20 0 2 m 12.00 11.00 10.00 L5 L4 9.00 1-1 Plan 9. Winepress, plan and sections. Garden Haifa–Tel Aviv highway 1 10.78 2-2 1 m Late Byzantine Remains neaR shiqmona 173 Fig. 29. The winepress, looking southwest. Fig. 30. The winepress with collecting vat L5 at bottom of photograph, looking south. Fig. 31. Traces of a mosaic in one of three depressions in the treading loor (L9) of the winepress. 174 Raz KLetteR from L4. The depressions may have held remains of must produced during treading, or the bases of storage jars. The elevation of the treading loor in the east, where the mosaic survives, is 11.49 m, while the bottoms of the depressions are at 11.28, 11.23 and 11.2 m. Farther west, the rock reaches 10.98–11.26 m asl, but here the mosaic is not preserved and the rock is damaged. The minimum length of the treading loor, from L6 to the most southern depression, is 12 m. Two installations were found in the winepress area. The irst (L6) is a large square stone (1.00 × 1.03 m, 0.3 m thick) that is secured in place with small stones and mortar on the outside and slightly protrudes above the treading loor (elevation at top 11.7 m). A mortise in the stone indicates a single ixed-screw press rather than a stone weight (Frankel 1999:140). The mortise (35 × 35 cm) is perforated and widens by 5–8 cm at the bottom of two adjacent sides to accommodate the screw (Plan 9: Section 2–2). The second installation (L7) is a large round depression in the treading loor (Fig. 32). The elevation of the treading loor is at 11.26–11.35 m, while the bottom of the lat depression is at 11.09 m asl. A line of stones and mortar on the sides of the round depression form a square depression (perhaps unintentionally), in the center of which is a deeper rectangular depression (at 11.03 m) measuring 25 × 40 cm. It seems incomplete, as it does not widen at the bottom to accommodate a screw. The three collecting vats (L4, L5 and L8) are similar in size and plan. Vat 4 is about 2.20 × 1.95 m with a maximum preserved depth of 1.35 m. Three stairs lead to its bottom from the southwestern corner. A small settling pit is in its northeastern corner. Vat 4 is partly hewn and partly built with thin plastered walls along the northern and eastern sides that cover the hewn rock. The walls do not reach the full height of the pit. The reason for the walls is evident in the east where the center of the wall is damaged. There, the rock has a natural issure that is sealed by the wall to prevent escape of the must from the vat. The reason for the damage in the Fig. 32. Round installation (L7) in the winepress. wall is also clear: the upper eastern side of this vat is hewn into the rock. This was noticed by robbers who believed it could lead them to the entrance of a tomb. They followed the wall of the vat downward and broke it in their search for the entrance. The loor of L4 is paved with a white mosaic. There is a round depression near the rim of the northern side of the vat, perhaps for placing a jar. Vat 5 is the largest (2.7 × 2.1 m, max. depth 1.35 m). Similar to Vat 4, it has stairs, a small settling pit, and plastered sides. Unlike Vat 4, it is not paved with a mosaic loor. The upper side of the vat was damaged by the modern drainage channel of the highway and remained to a level of 10.65 m in the west; originally, it may have been higher. There is a rounded depression (c. 0.12 m deep) near the northeastern corner, which could have been used for placing jars (cf. Winepress 12 at Tel Sumaq, Dar 1999:100– 102, Fig. 65). Vat 8 is south of Vat 5 and measures about 1.9 × 2.0 m. It has three stairs, a small settling pit, and a mosaic loor like Vat 4. Vats 4 and 5 were illed with dark gray earth, whereas Vat 8 was mainly illed with brown earth. While it is easy to describe the elements of the winepress, it is more dificult to understand its function as a whole. There are some shallow ‘channels’ in the rock, but they do not lead in Late Byzantine Remains neaR shiqmona expected directions (to the vats). The most obvious channel leads from near L7 westward and does not connect with Vat 8. Furthermore, it was illed with stones and mortar of the same kind used elsewhere in the winepress. Thus, it seems that the channels are natural issures that were illed and covered, rather than functional parts of the winepress. We found no holes or pipes connecting the three vats or the treading loor and the vats. Perhaps means of connection were situated higher in the vats and have not survived. Other obstacles to understanding the winepress are: the area to the east is outside the boundary of the excavation, and the area to the west was obliterated by the highway. Thus, we cannot know whether this was one winepress or a pair. The single, ixed-screw press was used in wine production throughout Palestine (except in the upper Galilee where a closed dove-tailed mortise was used; Frankel 1996:214, Fig. 5; 1999:140). Archaeological remains of such screws include a stone base, usually in the center of the treading loor (Frankel 1996:214). Mortises with a widening in two sides, such as in L7, were found in the Lower Galilee, Mt. Carmel and Sharon areas (Frankel, Avitsur and Ayalon 1994:75, Figs. 82, 84; Frankel 1999:141–144; Dar 1999:100–107). The large collecting vats of about 6 cu m are typical of the Byzantine period (Frankel 1999:140; for reconstruction of wine yields, cf. Dar 1999:107). The date of the Shiqmona winepress is based on pottery from Vats 4, 5 and 8, which give a terminus ante quem. Comparisons and a thorough discussion of the operation of such presses are offered by Frankel (1999:141–144, the ‘Ayalon press’ type with rectangular components). A restored winepress with a single ixed screw can be seen at the Eretz Israel Museum in Ramat Aviv (Chidiosan, Ayalon and Yosef 1987–1989). Three coins found in Vat 5 were identiied, dating to the sixth century CE; one coin dated to Justinian in the year 556/7 CE (see Kool, this volume: Cat. No. 8). This would indicate that the vats were illed with refuse slightly after 175 this date. In all the vats there was a considerable amount of pottery sherds, small stones, animal bones, and occasionally pieces of broken glass, metal, and shells, clearly refuse thrown into the vats after they ceased to be used to produce wine and became convenient dumping places. The pottery from the vats included mainly large vessels. Except for a few ifth-century vessels, the pottery is mostly from the sixth–seventh centuries (see Calderon, this volume). Cave t1 Cave T1 (c. 8 × 8 m) was used during recent times and illed with refuse (Fig. 33). Originally, it may have been a tomb, similar to the ones known farther northwest. Two second-century tombs were documented when Ha-Toren Street was paved in 1965 (Elgavish 1994:149). The entrance of T1 faces northwest and is protected by a thick twentieth-century cement and stone wall. The cave lies outside the area of development and, therefore, was not excavated. A modern stairway (clogged by vegetation) leads from the entrance of the cave to Ha-Toren Street. The adjacent section of the Haifa–Tel Aviv highway is named Hagana Street after the pre-state underground organization, but, contrary to a legend I heard from locals during the excavation, this was not the central Hagana command post in Haifa during the 1948 war (Eshel 1978). Fig. 33. Entrance to T1, looking north. 176 Raz KLetteR 1 0 1 3 2 4 0 2 Fig. 34. Bronze button (1) and mollusk shells (2–4). otheR Finds One bronze button, possibly from Napoleon’s campaigns, was found (Fig. 34:1). As the site is so close to the sea, mollusk shells were common (Fig. 34:2–4). The larger (Fig. 34:4) is comparable to a shell from Kh. ‘Eleq (Bar-Yosef Mayer 2000). Glass fragments, mostly of the late Byzantine period, include a considerable number of raw pieces of glass, as well as wasters and distorted fragments, indicating that glass was manufactured nearby (see Gorin-Rosen, this volume). ConCLusions The present excavation contributes to the understanding of Shiqmona during the Byzantine period. Hirschfeld (1998:20; 2006:140) questioned the deinition of Shiqmona as a city, and suggested it was a large village. This was part of a larger debate concerning deinitions of settlements during the Byzantine period, regarding the term ‘city’ in Jewish sources (Hirschfeld 1996:15, n. 39; cf. Safrai 1998). In the case of Shiqmona, Hirschfeld’s view was perhaps inluenced by the results of his salvage excavations. Three of the four areas he investigated between the Haifa– Tel Aviv highway and Tel Shiqmona produced meager remains. The more recent indings of large well-planned buildings, all thriving during the late Byzantine period, change this picture. It seems that the archaeological inds and written sources point to Shiqmona as being a city, not a village. While sources describe Shiqmona as a Jewish ‘settlement’ (Hirschfeld 2006:131, 140), we found evidence of a strong Christian presence in the form of monasteries and churches (notwithstanding the problem of deining ethnicity from archaeological records, see Kletter 2006a, with further references therein). Hirschfeld’s suggestion (2006:141) that “Byzantine Shiqmona was apparently settled by Jews” was based solely on the (few) written texts. However, the number of monasteries outside the city, and not just Elgavish’s chapel of which Hirschfeld (2006:141) was skeptical, supports Elgavish’s conclusion that “a large Christian community was established there”. Perhaps Shiqmona was ethnically mixed, conforming to the deinition of a city and not a village. Hirschfeld (2006:142, n. 3), himself, notes that “Christians resisted setting up monasteries or churches” in areas settled by Jews. It is ironic that although the greatest desire of Elgavish (1994) was to ind the Jewish remains of Shiqmona, he did not ind them. Thus, we cannot suspect that he ‘missed’ them on purpose. The monastery and Buildings A and B are located outside the city proper. The same is true for buildings in the nearby excavations conducted by ‘Ad and Torge (forthcoming), Late Byzantine Remains neaR shiqmona the monastery excavated by Dothan in 1951 (Dothan 1954–1955) and a monastery south of the tell (Elgavish 1994:22–23). In the present excavations we did not ind any signs of a fortiication wall. Furthermore, the location of the buildings so near the Carmel Mountain leaves no room for an assumed wall to the east. The architecture discovered in the area was not planned for military or defensive purposes. The outer walls of the buildings are similar in thickness to the inner walls, and towers or other military installations are completely lacking. It seems that the buildings were erected in a period of relative security. According to Hirschfeld’s excavations, at least part of the area between these buildings and the tell was open. The monasteries were active during the late Byzantine period; an exact date of construction cannot be established. Elgavish (1994:94) dated the erection of many buildings on the tell to the fourth century and believed that their end came in the seventh century. There is no clear evidence in our area for more than one phase or stratum, such as lifting of loors or changes in plan. The latest Byzantine-period coin from the present excavation that can be precisely dated was minted in 575–578 CE (see Kool, this volume: Cat. No. 9). The latest coins from other excavations at Shiqmona are from the 590s (see Kool, this volume). Perhaps the end of the buildings came with a violent Muslim conquest in 638 CE (thus Elgavish 1994:144–145), although churches and monasteries continued to exist. ‘Ad and Torge (forthcoming) suggested that Byzantine Shiqmona suffered a severe earthquake in the seventh century. However, archaeological evidence of an earthquake (cf. examples discussed by Fabian 1998:21–26; Mazor and Korjenkow 1999:265–282) was not documented in the present excavation. The sinking and slanting of walls in the area excavated by ‘Ad and Torge is more likely the result of an unstable ill of natural silt rather than an earthquake. It is also possible that Shiqmona, or at least the present area outside the tell, was deserted around 600 CE, and, thus, 177 came to an end unrelated to a conquest or an earthquake in the seventh century. The identiication of these buildings as a monastery deserves attention. Their plan and proximity to each other do not it farmsteads, which tend to be isolated from one another and near ields. Also, inds, such as marble fragments of chancels and screens, indicate a religious rather than agricultural function. It is not always easy to deine Byzantine-period monasteries by archaeological remains. Rural villas and farms of this period were often large and impressive and could include chapels for their residents (Safrai 1998:25–26; cf. Bar 2005:51). Members of a monastery community often worked in agriculture and industry, and monasteries share architectural characteristics with farms (Hirschfeld 1992:199). The archaeological deinition of monasteries is often based on remains, such as chapels, marble items, and inscriptions, or even circumstantial data, such as the component “Deir” in later Arabic place names (e.g., Avner 2000:25*, 50*; Seligman and Abu Raya 2002:137; Kloner 2003:51*; cf. Patrich 1995:108). Aviam (2002:216) deines monasteries in the western Galilee as square complexes of 500–2000 sq m that include rooms, olive oil presses, sometimes winepresses, and usually a church. Not all scholars agree with him. For typologies of Judean Desert monasteries, see Hirschfeld 1992:18–47. The concentration of several monasteries in such a small area near Shiqmona deserves attention. It was perhaps related to the area’s special status in Christian tradition. Here, the coastal plain narrows, the mountain almost kisses the sea, and the site sits right on the road to Haifa. It is very close to major Christian sites on the Carmel Mountain, such as the cave of Elijah (Guérin 1875:181–182, 185, 187–189). The Shiqmona monasteries may have served pilgrims traveling to and from the Carmel. Still, why build monasteries so close to a city with a Christian population that could hostel pilgrims and care for all their needs? 178 Raz KLetteR Remote Byzantine-period monasteries in the Sinai and Judean Deserts have captured the imagination and been studied in depth (Hirschfeld 1990; 1992; Magen and Talgam 1990; Tsafrir 1993; Dahari 1994; Patrich 1995:6; Dahari 2000:150–168; Bar 2005:50). However, monasteries are also commonly found within and near contemporary cities, such as Bet She’an, Ashqelon, Jerusalem and Bethlehem. There are dozens of monasteries around Jerusalem (see partial map in Seligman and Abu Raya 2002:127, Fig. 1). Hirschfeld (1992:3) acknowledged that “many monasteries were established in or around large cities or near villages”, but thought that “monasteries established in desert areas were the most numerous”. He counted 35 small, 12 mid-sized, and only six large monasteries in the entire Judean Desert (Hirschfeld 1992:79), while Kloner (2003:16*) listed 43 church or monastery sites immediately surrounding Jerusalem alone. (Hirschfeld included some of Kloner’s Jerusalem sites in his desert study, as there is no distinct geographic demarcation between the Jerusalem area and the Judean Desert). Kloner wrote, “surrounding the city [of Jerusalem] were neighborhoods or concentrations, and scores of monasteries and churches”. They were distributed in a belt of 1.0–1.5 km outside the city walls, especially at Mt. Scopus and north of the city. It could be explained that “within the walls no vacant land remained for construction, thus monasteries were built near the city,” but “remains of scores of monasteries, churches and chapels were documented” also in a 2–5 km belt around the city wall (Kloner 2003:51*). Kloner (2003:51*) concluded that “networks of monasteries also existed in areas 5–6 km away from the city”. Both Jerusalem and Bethlehem were cities surrounded by scores of monasteries in close proximity (Kloner 2003:51*). Seligman and Abu Raya (2002:137) explained such monasteries as “rural monasteries... subsisting on cultivation of their immediate surroundings” (cf. Hirschfeld 1992:199; Kloner 2003:51*; for rural monasteries, see Bar 2005), or road-stations located “along the routes from Jerusalem to the Judean Desert laura and coenobia and from Jerusalem to Bethlehem”. Some of these monasteries are no doubt related to roads, e.g., Kh. el-Latatin (Zelinger 1998:77–80; 1999), but it seems to me that such explanations do not it all the monasteries in and around cities. For example, and without negating the importance of agriculture and roads, many monasteries around Jerusalem are not related to any major or even secondary road. Moreover, the area around Jerusalem had an agricultural population, evidenced by the remains of Byzantine-period farms. The monasteries were additional components in this agricultural hinterland. Kloner (2003:51*) noticed that “remains of monastic structures located at a greater distance from the city are often smaller than those situated in the strip closer by”. If so, it seems that monasteries coexisted side by side with the city and not exclusive of one another. Monasteries close to cities must have had mutual relationships with them and, perhaps, were dependant on the city. Even desert monasteries received food and building materials from Jericho and bought land and gardens in Jericho for their use. They maintained hostels in Jerusalem that served persons other than monks as a source of income (Patrich 1995:132). At times of religious crisis, the desert monasteries were dependent on the Jerusalem clergy while many monks from monasteries around Jerusalem served in religious institutions in the city and enjoyed a comfortable way of life from the ifth century onward (Patrich 1995:4–8, 27, 305– 306). Even if desert monasteries “enjoyed the greatest enhanced status” (Hirschfeld 1992:3), their seclusion suited relatively few. Even at their zenith, Judean Desert monasteries were home to a total population of c. 2200 people, or 3000 including secluded hermits (Hirschfeld 1992:79). The ideology of remote seclusion itted Syrian monasticism in its early stages, but later Syrian monasteries were built close to cities and villages and their residents dealt Late Byzantine Remains neaR shiqmona in agriculture and industry (Patrich 1995:18– 19; cf. Bar 2005:57–59). In Egypt, too, work occupied an important ideological place in some monasteries, where monks performed all types of work. Such monasteries bought materials from villages or cities and in turn sold them products; monks often worked in villages as hired laborers (Patrich 1995:13–16). Thus, it seems that, at least in the late Byzantine period, most nuns and monks in the country stayed close to civilization. “Urban” and “rural” monasteries were far more important than desert monasteries. Perhaps Israeli conceptions about “conquering the wilderness” are inluenced by the ideology of the irst decades of the State of Israel, when newcomers were sent to settle in new peripheral towns and villages. Even with twentieth century technology, such settlements were modest successes and remained peripheral. “Settling the wilderness” (Hirschfeld 1992:236) was an ideology that was never feasible without outside support. appendix 1: List oF LoCi Locus Deinition Chapel and Buildings A and B (Plan 2) L200 Surface to rock, between T8 and chapel L201 Surface to heads of W4 and W6 179 L202 Surface to stone debris, north of chapel, above L209 and L210 L203 Surface to loor, room of chapel, east of W5 L204 Surface to loor, room of chapel, west of W5 L205 Surface to head of W8, above L206 and L207 L206 Head of W8 to mosaic loor, room of chapel, north of W8 L207 Head of W8 to cobbled loor L226, south of W8 L208 Stone debris to rock, beneath northern part of L202 L209 Stone debris to rock, beneath southern part of L202 L210 Surface, above L226 L211 Surface to loor, above L224 L212 Surface, south of chapel, area of W1 L213 Surface, south of W1 L214 Surface, south of L213 L215 Surface, north of W10 L216 Surface, above L220 and L222 L217 Beneath L203 until rock L218 Beneath L207 until rock L219 Beneath L204 until rock L220 Mortar loor height 10.06 m, west of W11 L221 Heads of W10 and W11 to sterile earth, west of W11 and around L220 L222 Heads of W10 and W11 to sterile earth, east of W11 L223 Building A, along road, north of W15 and west of W14 L224 Under loor in chapel, room bounded by W17 and W21, east of W20 L225 Under loor in chapel, room bounded by W17 and W21, west of W20 L226 Cobbled loor till sterile earth, north of W21, south of L218 180 Raz KLetteR appendix 1 (cont.) Locus Deinition L227 South of L221 L228 South of W17 L400 Surface locus of Area D1 L401 Area D2 L402 Near clay cofin L403 L403 Clay cofin Tomb Area (Plan 1) T1 Unexcavated cave, recently used T2 Stairs and nearby area, cleaned T3 Hewn tomb T4 Hewn tomb T5 Hewn tomb (not excavated) T6 Hewn tomb T7 Hewn tomb T8 Hewn tomb T12 Remains of tomb south of the winepress Winepress (Plan 11; Loci 4–9) and Other Loci L1 Probe L2 General area with surface cleaned to rock L3 Probe, southern edge of area L4 Collecting vat L5 Collecting vat L6 Press with mortise for single ixed-screw press L7 Press L8 Collecting vat L9 Treading/working loor L10 Part of mortared basin, north of the winepress L11 Round pit, modern note 1 The excavation was conducted on behalf of the Israel Antiquities Authority (Permit No. A-3172) by the author with the help of Iskander Jabur and Amin Abu-Hamid (area supervisors). Yoav Lerer supervised the excavation of Area D. Deep thanks are due to them and to Vered Eshed (anthropology), Uzi ‘Ad and Hagit Torge (data on an excavation at Shiqmona), Tsila Sagiv (photography), Israel Vatkin, Vadim Pirsky, and Avi Hagian (survey and plans), Shlomo Ya‘akov-Jam (administration), Michal Ben-Gal (pottery restoration), Marina Shuiskaya (pottery drawings), Ela Altmark (metal conservation), Donald Tzvi Ariel and Robert Kool (numismatics), Yael Gorin-Rosen (glass inds), and Moshe Sade (archaeozoology). Rivka Calderon studied the pottery, prepared its publication and helped in various other ways. I also wish to heartily thank Prof. Ronny Reich, for information concerning Samaritan amulets, Dr. Gerald Finkielsztejn, for sharing his knowledge of the area, and Dr. Gideon Avni, for information concerning Byzantine-period towns. Late Byzantine Remains neaR shiqmona 181 ReFeRenCes ‘Ad U. and Torge H. Forthcoming. Shiqmona. ‘Atiqot (Hebrew). Aviam M. 1990. Horvat Hesheq—A Unique Church in Upper Galilee: Preliminary Report. In G.C. Bottini, L. Di-Segni, and E. Alliata eds. Christian Archaeology in the Holy Land: New Discoveries; Essays in Honour of Virgilio C. Corbo, OFM. Jerusalem. Pp. 351–377. Aviam M. 2002. Five Ecclesiastical Sites in Western Upper Galilee. In Z. Gal ed. Eretz Zafon: Studies in Galilean Archaeology. Jerusalem. Pp. 165–218 (Hebrew; English summary, Pp. 184*–185*). Aviam M. and Stern E. 1997. Burial in Clay Sarcophagi in Galilee During the Roman Period. ‘Atiqot 33:151–162 (Hebrew; English summary, p. 19*). Avigad N. 1971. Beth She‘arim III: The Archaeological Excavations during 1953–1958. Jerusalem (Hebrew). Avi-Yonah M. 1964. Haifa and Its Surroundings during the Second Temple, Mishna and Talmud Periods. In Essays and Studies in the Knowledge of the Land (Massot u-Mekhkarim Beydiot Haaretz). Tel Aviv. Pp. 315–342 (Hebrew). Avner R. 2000. Deir Ghazali: A Byzantine Monastery Northeast of Jerusalem. ‘Atiqot 40:25*–52* (Hebrew; English summary, pp. 160–161). Avni G. 1997. The Necropolis of Jerusalem and Beth Govrin During the 4th–7th Centuries A.D. as a Model for Urban Cemeteries in Palestine in the Late Roman and Byzantine Periods. Ph.D. diss. The Hebrew University. Jerusalem (Hebrew). Bar D. 2005. Rural Monasticism as a Key Element in the Christianization of Byzantine Palestine. Harvard Theological Review 98:49–65. Bar-Yosef Mayer D.E. 2000. Appendix IV: A Shell from Horvat ‘Eleq. In Y. Hirschfeld ed. Ramat Hanadiv Excavations: Final Report of the 1984– 1998 Seasons. Jerusalem. P. 749. Binns J. 1999. The Concept of Sacred Space in the Monasteries of Byzantine Palestine. In T.A. Insoll ed. Case Studies in Archaeology and World Religion (BAR Int. S. 755). Oxford. Calderon R. This volume. Pottery from the Late Byzantine Site near Shiqmona. Chidiosan N., Ayalon E. and Yosef A. 1987–1989. Two Byzantine Winepresses in the Eretz Israel Museum, Tel Aviv. In R. Zeevi ed. Israel—People and Land V–VI. Tel Aviv. Pp. 23–36 (Hebrew). Dahari U. 1994. The Monasteries of Southern Sinai during the Early Byzantine Period: Their Development, Internal Organization and Economical Conditions. Ph.D. diss. The Hebrew University. Jerusalem (Hebrew). Dahari U. 2000. Monastic Settlements in South Sinai in the Byzantine Period: The Archaeological Remains (IAA Reports 9). Jerusalem. Dar S. 1999. Sumaqa: A Roman and Byzantine Jewish Village on Mount Carmel, Israel (BAR Int. S. 815). Oxford. Dauphin C. 1998. La Palestine Byzantine: peuplement et populations III: catalogue (BAR Int. S. 726). Oxford. Dothan M. 1954–1955. Excavations of a Monastery near Sha‘ar Ha-‘aliyah. Yediot 18:216–222 (Hebrew). Elgavish J. 1968. Archaeological Excavations at Shiqmona: Field Report No. 1; The Levels of the Persian Period. Seasons 1963–1965. Haifa (Hebrew). Elgavish J. 1974. Archaeological Excavations at Shiqmona: Field Report No. 2; The Level of the Hellenistic Period. Stratum H. Seasons 1963– 1970. Haifa (Hebrew). Elgavish J. 1977. Archaeological Excavations at Shiqmona: The Pottery of the Roman Period. Haifa (Hebrew). Elgavish J. 1994. Shiqmona on the Seacoast of Mount Carmel. Tel Aviv (Hebrew). Eshed V. This volume. Bones from the Late Byzantine Site near Shiqmona. Eshel Z. 1978. The Hagana Battles in Haifa. Tel Aviv (Hebrew). Fabian P. 1998. Evidence of Earthquake Destruction in the Archaeological Record—The Case of Ancient Avdat, Israel. Field Trips Guidebook (Israel Geological Society, Annual Meeting, Mitzpe Ramon 1998). Pp. 21e–26e. Foerster G. 1986. A Decorated Christian Burial Cave near Kibbutz Lohamey Ha-Geta’ot. In M. Yedaya ed. The Western Galilee Antiquities. Jerusalem. Pp. 416–431 (Hebrew). Frankel R. 1996. Oil Presses in Western Galilee and Judea—A Comparison. In D. Eitam and M. Heltzer eds. Olive Oil in Antiquity. Padova. Pp. 197–218. Frankel R. 1999. Wine and Oil Production in Antiquity in Israel and Other Mediterranean Countries. Bath. Frankel R., Avitsur S. and Ayalon E. eds. 1994. History and Technology of Olive Oil in the Holy Land. Arlington. Gorin-Rosen Y. This volume. Late Byzantine Glass from the Site near Shiqmona. Guérin V. 1875. Description géographique, historique et archéologique de la Palestine: Galilée I, II. Paris. 182 Raz KLetteR Hirschfeld Y. 1990. List of Byzantine Monasteries in the Judean Desert. In C.G. Bottini, L. Di-Segni, and R. Talgam eds. Christian Archaeology in the Holy Land: Essays in Honour of Virgilio C. Corbo, OFM. Jerusalem. Hirschfeld Y. 1992. The Judean Desert Monasteries in the Byzantine Period. New Haven. Hirschfeld Y. 1996. Changes in Settlement Patterns of the Jewish Rural Populace before and after the Rebellions against Rome. Cathedra 80:3–18 (Hebrew). Hirschfeld Y. 1998. Tel Shiqmona—1994. ESI 18:19–20. Hirschfeld Y. 2006. Excavation at Shiqmona—1994. ‘Atiqot 51:131–143. Kletter R. 2006a. Can a Proto-Israelite Please Stand Up? Notes on the Ethnicity of Iron Age Israel and Judah. In A. Maeir and P. De-Miroschedji eds. Archaeological and Historical Studies in Honor of Amihai Mazar. Winowa Lake. Pp. 573–586. Kletter R. 2006b. Just Past: The Making of Israeli Archaeology. London. Kloner A. 2003. Survey of Jerusalem: The Northwestern Sector; Introduction and Indices (Archaeological Survey of Israel). Jerusalem. Kool R. This volume. Late Byzantine Coins from the Site near Shiqmona. Magen Y. and Talgam R. 1990. The Monastery of Martyrius at Ma‘ale ’Adummim (Khirbet elMurassas) and Its Mosaics. In G.C. Bottini, L. DiSegni, and E. Alliata eds. Christian Archaeology in the Holy Land: New Discoveries; Essays in Honour of Virgilio C. Corbo, OFM. Jerusalem. Pp. 91–152. Makhouly 1939. Rock-cut Tombs at el-Jish. QDAP 8:45–50. Makhouly N. 1940. Mosaic Pavement near Tell es Samak. Report, 15/3/40. Mandatory Site Files, IAA Archives, Rockefeller Building, Jerusalem. Unpublished MS. Makhouly N. 1944. Tell es-Samak, Haifa SubDistrict. QDAP 10:206. Mazor E. and Korjenkow A.M. 1999. Earthquake Characteristic Reconstructed from Archaeological Damage Patterns: Shivta, the Negev Desert, Israel. Israel Journal of Earth Sciences 43/3-4:265–282. Netzer E. and Birger R. 1990. A Byzantine Monastery at Nuseib ‘Uweishira, West of Jericho. In G.C. Bottini, L. Di-Segni, and E. Alliata eds. Christian Archaeology in the Holy Land: New Discoveries; Essays in Honour of Virgilio C. Corbo, OFM. Jerusalem. Pp. 191–200. Olami Y., Ronen A. and Romano A. 2003. Map of Haifa (West) (Archaeological Survey of Israel). Jerusalem. Ovadiah A. 1970. Corpus of the Byzantine Churches in the Holy Land. Bonn. Ovadiah A. and Gomez de Silva C. 1984. Supplementum to the Corpus of the Byzantine Churches in the Holy Land. London. Ovadiah A. and Ovadiah R. 1987. Hellenistic, Roman and Early Byzantine Mosaic Pavements in Israel. Rome. Parks D. 2003. Clay Cofins in the Cyprus Museum. RDAC 2003:247–260. Patrich J. 1995. The Judean Desert Monasticism in the Byzantine Period: The Institutions of Sabas and his Disciples. Jerusalem (Hebrew). Peleg M. 1988. A Chapel with Mosaic Pavements near Tel Shiqmona (Tell es-Samak). IEJ 38: 25–30. Piccirillo M. 2000. Architecture and Liturgy in the Early Church. In Y. Yisraeli and D. Mevorach eds. The Cradle of Christianity (Israel Museum Catalogue No. 473). Jerusalem. Pp. 49–72 (Hebrew). Pummer R. 1987. Samaritan Amulets from the Roman-Byzantine Period and their Wearers. RB 94:251–263. Reich R. 1986. A Samaritan Amulet from Horvat Itaim near Nahariya. In M. Yedaya ed. The Western Galilee Antiquities. Jerusalem. Pp. 393– 396 (Hebrew). Reich R. 1994. Two Samaritan Rings from Gelilot (El-Jalil), Near Herzliyya. ‘Atiqot 25:135–138. Reich R. 2002. Samaritan Amulets from the Late Roman and Byzantine Period. In E. Stern and H. Eshel eds. The Samaritans. Jerusalem. Pp. 289– 309 (Hebrew). Safrai Z. 1998. Ancient Field Structures— The Villages in Eretz Israel during the Roman Period. Cathedra 89:7–40 (Hebrew). Seligman J. and Abu Raya R. 2002. A Byzantine ‘Monastery’ at Khirbet Umm Leisun, Jerusalem. ‘Atiqot 43:127–148. Siegelman A. 1966. Burial Caves at Shiqmona. HA 17:19–20 (Hebrew). Tsafrir Y. 1984. Eretz Israel from the Destruction of the Second Temple to the Muslim Conquest II: Archaeology and Art. Jerusalem. Tsafrir Y. 1993. The Development of Ecclesiastical Architecture in Palestine. In Y. Tsafrir ed. Ancient Churches Revealed. Jerusalem. Pp. 1–16. Zelinger Y. 1998. The Identiication of the JerusalemLydda (Diospolis) Roman Road on the Madaba Map and Khirbet el-Latatin. BAIAS 16:75–85. Zelinger Y. 1999. The Byzantine Road Stations around Jerusalem. M.A. thesis. Bar Ilan University. Ramat Gan (Hebrew).