picture at 50mm from top frame The British Museum Expedition to Kom Firin

picture at 50mm from top frame The British Museum Expedition to Kom Firin Report on the 2009 season

The British Museum has been working at Kom Firin, a large settlement site in Beheira, since 2002 (fig. 1). Following initial topographic and geophysical surveys, research priorities were concentrated on the foundations of the Ramesside temple, a modest structure of mudbrick with limestone used for columns and doorways; this temple is now fully published. 1 Subsequent survey and excavations have revealed the architectural context for this monument, set at the heart of a mudbrick enclosure wall, covering an area of c.231m east-west, and c.200/218m north-south (fig. 2). The corners of this complex were provided with external towers, and an imposing gateway provided access through the north wall onto the temple axis. Excavations have concentrated on the interior north-eastern corner of this building, where a series of architectural phases have been identified, spanning the late Ramesside and Third Intermediate Periods (fig. 3). Areas with large round silos lie alongside suites of narrow storage magazines, with nearby rooms of a smaller scale, perhaps parts of dwellings. It is clear, from auger coring undertaken in 2007, that this Ramesside complex was founded on a Pleistocene sand turtleback. Furthermore, it seems Kom Firin was effectively an island during parts of antiquity (fig. 2). A major river channel formed the southern boundary of the site, but a secondary channel has been identified to the north; it was towards this waterway that the city faced during the Ramesside era. Inscriptions upon a doorjamb in the temple hint that the walled complex at Kom Firin was seen as defensive, symbolically if not practically. Such a role fits well with the historical context, as textual sources record a series of confrontations with Libyan groups along the western Delta in the 19 th and 20 th dynasties. Unfortunately, the New Kingdom toponym for Kom Firin has yet to be identified. More recent seasons have focused upon the later history of the settlement, namely the construction of two later temple temenoi, enclosing a much larger area (480x390m) than the New Kingdom complex (fig. 2). In addition, an area of 7 th -5 th century BC housing in the Citadel area of the site, west of the archaeological resthouse and magazine, was excavated in 2007 and 2008. The well-preserved lower parts of several houses, with oven emplacements and beehive-shaped silos, but also a pottery kiln, attested to a complex series of architectural renovations, modifications and re-use across a period of two centuries (figs. 4 5). The 2009 season 2 was largely devoted to the study of already excavated material, in particular the pottery from a long slot trench placed across the late enclosure walls (NA, figs.6 8), but also the ceramics from the Citadel (trenches CA, CB, CC, CD; fig. 5). In addition, finds were re-studied and collated, and an area of newly exposed Late Period enclosure wall was cleaned and recorded. 1 See N. Spencer, Kom Firin I: The Ramesside temple and the site survey (British Museum Research Publication 170, London, 2008). 2 The eighth season of fieldwork at Kom Firin (Beheira), took place between October 3rd and 23rd, 2009. The team consisted of Neal Spencer (director, archaeologist), Květa Smoláriková (ceramicist, all areas of site other than Citadel), Ross Thomas (ceramicist, Citadel) and Mat Dalton (illustrator). The research of Thomas and Smoláriková forms the basis for the discussion of ceramic assemblages within this Report. A debt of gratitude is of course due to the Supreme Council of Antiquities, for granting the concession, and for their support throughout. Thanks are due to the Secretary General, Dr. Zahi Hawass, and the General Director of Foreign and Egyptian Mission Affairs, Dr. Mohamed Ismail. Mustafa Rushdi, chief inspector in the Damanhur inspectorate, facilitated our work. Our site inspector was Nagwa Ezz el-din Mohamed, and we are also grateful to Ahmed Kamal Adham for his help and knowledge. As ever, Hisham el-eithy proved invaluable in facilitating our work..

3 Ceramics from trench NA The magnetometry survey of 2005 provided evidence for three near-parallel rectilinear features, reasonably interpreted as enclosure walls, which could be placed in a proposed layout of enclosures on the basis of other surviving segments (fig. 9). Trench NA (35x2.5m8) was laid across these three features in 2006, just east and south of STN 15, and revealed parts of three large walls (0440, 0417+0416, 0436; fig. 8) and intensive dumping activity to the north of wall 0436. The ceramics from the associated, stratified, deposits are important for understanding the dating of these features. Significantly higher concentrations of pottery were recovered in this trench than anywhere else at Kom Firin, and study of this material could not be resumed until 2009. The following discussion thus complements and updates material contained within the Report for the 2006 season, which also includes extensive plans and section drawings. 3 The foundation level of the southernmost wall (0440) was not reached, but the deposits north of the wall included 0429, notable for its large storage jars (C4122) of very gritty fabric, but also other forms in the same fabric: medium-sized jars, shallow bowls or plates with everted rims (C4123) and large amount of shallow bowls or plates rolled rims, but also a tray fragment with wet-smoothed surface (C4130). 4 A more imposing wall to the north (0417), measured at least 3.4m thick, and was further thickened with the addition of brickwork 0416. North of this wall 0416/0417, a key stratigraphic relationship was revealed which provided evidence for the dating of wall 0436, the northernmost of the three walls in this trench. A series of deposits, shelving off to the north, had built up against the wall 0416: some were windblown in character (0428), while others contained a mixture of sand, degraded bricks, charcoal fragments (most notably 0426, 0427). A construction cut (0425) had truncated these deposits, in preparation for the building of wall 0436, with Late Period and Ptolemaic sherds lying above the cut, suggesting the wall dates to the Ptolemaic era. Unfortunately the northern face of wall 0436 was not revealed due to the disturbances from dumping activity discussed below. Rather, dense pottery and rubble dumps were found here (including 0402, 0404, 0418, 0422, 0422A, 0432), sloping off at the limit of our trench. The ceramic evidence suggests a reverse chronological sequence, presumably caused by this material being redeposited through dumping outside the wall. The deepest layer reached (0432) consisted of compact sand mixed with mudbrick debris, and yielded parts of Graeco- Roman casseroles, shallow bowls and plates, numerous Late Period vases with moulded rims, fineware juglets (C4151), bowls with incurved rims, carinated bowls and medium-sized lids with a groove under the rim. The context above (0422A), was similar but included some Third Intermediate Period and Late Period jars and bowls alongside Hellenistic amphora handles and coarse plates made of Nile C (C4149). Above this, layer 0422 included a painted body sherd of a shouldered bottle (?) with thick cream slip and set of red horizontal bands on the exterior (4th 3rd centuries BC), a large amount of imported and Egyptian amphorae, Phoenician torpedo jars (end of the 4th and beginning of the 3rd centuries BC) and wide-mouthed jars and bowls 3 http://www.britishmuseum.org/pdf/kom%20firin%202006%20low%20res.pdf. 4 K. L. Wilson, Cities of the Delta, II: Mendes (ARCE Reports 5. Malibu, 1982), fig. 15/10; A.J. Spencer, Excavations at Tell el-balamun. 1991 1994 (London, 1996), fig. 61.

4 of the Late Period. Subsequent deposits shelve less noticeably to the north, though can still be characterised as dumping episodes, due to the concentration of large pottery vessels found therein. Deposit 0418 was characterised by fragments of basket-handled jars, 5 rims and handles of Phoenician jars, 6 and a significant amount of imported amphorae. In contrast, layer 0419 yielded large storage jars, but once again rims and body sherds of Phoenician amphorae with short thick necks (C4147, C4148). The three uppermost contexts at this elevated, northern, end of the trench contained an even higher concentration of ceramic material. Deposit 0404, stained a distinctive pinkorange by the degraded pottery, seems to have been deposited within a short time frame, perhaps partly waste from industrial activity. Ceramics included a 3rd century BC unguentarium, globular juglets manufactured of Nile silt (late Persian-early Hellenistic period), large bowls with modelled rims (4th century BC) and handmade miniature vessels, but also coarse, large stoppers (C490, C4091), imported amphorae (C4092, C4093) and Roman amphorae with massive, slightly rounded, toes (C4094), 7 storage jars, Late Period pot-stands (C4114) and a medium-sized beaker (C4115). Amphorae with buff-slipped surfaces, or slender with squat rims (2 nd /1 st centuries BC), were identified, alongside Samian examples (C4134), 8 and fragments from Cos, Rhodes and Mendes. A small layer (0406) above this in the north-western corner of the trench yielded the upper part of elegant black glazed flask from the Hellenistic period (?), a handle of a 4 th century BC painted goblet (C428, fig. 13) from Cyprus, and more globular juglets, but also storage jars with restricted or short neck and numerous toes of amphorae. A ring-base from a black glazed jug (C4141, fig. 11) 9 was recovered from 0406, but also amphorae and storage jars in significant quantities (C4144), among them several of a Samian provenance (C4145, fig. 11) from the late 4 th century BC. The latest significant deposits (0402, 0402A), lay beneath topsoil 0401. Deposit 0402 displayed clear tip lines, and yielded a fine pottery fragment of a red-figured vase (C426, early 4th century BC), stamped amphorae handles including two with the Rhodian Eponym Nikasagoras (198-161 BC), restricted storage jars of the Third Intermediate Period, but also shallow plates or bowls with inner rims, finer ware of later periods consisted of bowls with incurved rims, pilgrim flasks (C4106), dinoi with ledged rims (3 rd -2 nd centuries BC); large storage jars with slipped surface (C4117), jugs or cooking pots and lower parts of goblets and amphorae of late Ptolemaic date. These amphorae include Egyptian ware with squat 5 D.B. Redford, Excavations at Mendes, I: The Royal Necropolis (Culture & History of the Ancient Near East 20, Leiden and Boston, 2004), fig. Z/2, dated to the 5 th -4 th centuries BC. 6 Redford, Mendes I, for wide range of this type see plate H. 7 W.D. Coulson, Ancient Naukratis, II, Part I: The Survey at Naukratis (Oxbow Monograph 60, Oxford, 1996), 61, fig.30, second row, mainly type VW3/54, wide range of time from 1 st century BC to the 2 nd century AD. 8 Coulson, Naukratis II/1, 58, fig. 28/728, as at Naukratis, these are derived from a late 4 th century BC Samian type. 9 Coulson, Naukratis II/1, 26, fig. 11/type B1:503, Ptolemaic Period.

rim 10, Samian examples with a mushroom rim (C4108) and Coan 11 (C4086) and Mendean (C4087b) 12 types. 13 Finer pieces of black glazed or painted pot (C4082, C4083) were also recovered. Context 0402A yielded eroded fragments of large storage jars and amphorae of coarse fabric, a small offering bowl (C4125, fig. 10), and cooking pots and dinoi of different sizes (C4126) of the Roman era, but also Late Roman 7 amphorae with ribbed exteriors. Fragments of at least forty coarse Egyptian amphorae were also found in this context, and an imitation of a 3rd century BC Chian amphora. It is thus possible that 0432, 0422A, 0422 and 0418 represent a redepositing of material from the south, thus lying partly in reverse chronological sequence. Above these layers lie several ceramic rich layers that are dumped at much less of an angle, perhaps as wall 0436 was no longer a dominant feature of the landscape. These seem to follow a more predictable ceramic sequence: 4th-2nd century BC (0404, 0402) and 1st century BC-3rd century AD (0402A). Thus one could propose a wall of the early Ptolemaic era, gradually obscured by dumping of earlier and contemporary material from the south, before the build up of rubbish (again dumped in some concentration) above these tipped layers from the mid(?)-ptolemaic era onwards. The study undertaken in 2009 emphasise that the majority of ceramics from this area comprise cooking pots, casseroles, dinoi, bowls, jars, jugs, goblets, saucers, 14 some fine tableware items of black and red glazed types, and a large quantity quantity of amphorae. 15 Fragments of trays and pot-stands were also found. The majority of this material dates to the Ptolemaic period, with some earlier (presumably redeposited) and some Roman era ceramics. Later material is rare, some perhaps as late as the 7 th century AD. The most important result of this season s analyses is that wall 0436, the later of the two post-new Kingdom temenoi, can now be firmly dated to the early Ptolemaic era, rather than the end of the Late Period. 5 Ceramics from the Citadel Two seasons of excavations (2008, 2009) revealed parts of a dense housing area west of the modern antiquities resthouse, constantly modified and developed over a 200 year period. Later buildings, of the Hellenistic era, lie outside the excavation area, and were excavated by Shafik Farid between 1949 and 1951. A sondage by the Naukratis Project (American Research Center) supports this date. 16 10 A. Leonard, Ancient Naukratis: Excavations at a Greek Emporium in Egypt, Part 1: Kom Ge if (Annals of the American School of Oriental Research 54, Atlanta, 1997), 181, fig. 6.16/5. 11 Coulson, Naukratis II/1, 48, fig 23/826, 348, 3 rd century BC;Redford, Mendes I, fig. B/10. 12 Coulson, Naukratis II/1, 59, fig. 29/type JK2:751, examples which derive from Mendean amphora of the mid-5th century BC. 13 It is quite difficult, on the basis of visual examination, to satisfactorily determine which are true imports and which are locally produced imitations. 14 Berlin in Leonard, Naukratis I, figs. 6.1 6.68 with very wide repertoire of Hellenistic pottery. 15 Coulson, Naukratis II/1, 46 64 Figs. 22/31. 16 W.D.E. Coulson, A. Leonard Jr, A, Three Seasons of Excavations and Survey at Naukratis, JARCE 19 (1982), 79 80, fig.10.

6 Architectural and stratigraphic evidence indicated eight principal phases of architectural activity, and this study season provided the opportunity for a re-assessment of the ceramic evidence. The pottery from the Citadel encompasses material of Ramesside through the Roman period, but the vast majority of forms can be dated, through parallels, to the mideighth to mid-fourth century BC, placing the assemblage within the 25 th to 27 th dynasties. There appears to be a concentration of pottery distinctive of the sixth to fifth centuries BC that include the operation of a kiln in this area (fig. 14). Rare material from the surface deposits were of Ptolemaic and possibly Roman date, including red- slipped and black glazed fine-wares, but very few sherds can confidently be assigned to dates after the mid fourth century BC. Storage jars were found in large numbers, with plain or flaring rims (late Third Intermediate Period-Saite), while amphora included ribbed examples with a thickened rim, large neck-less examples with a bead rim (C3086, fig. 17), and Phoenician examples in Gaza silt or Levantine marl. A small number of fragments with a Chian or Samian provenance were recovered, and also a Thasian amphora handle stamped ΣorΚΩΝΙΔ, perhaps of 4 th 3 rd century BC date (C2013). Fragments of Cypriot jugs or bottles represent the other main form of import. The most ubiquitous form of pot recovered in the Citadel were the pigeon pots, deep jars with moulded rims and a circular aperture in the base, of the sixth and fifth centuries BC. The kiln (1619) found in trench CD was undoubtedly producing these vessels (fig. 14), with wasters being found (fig. 17), and broken examples used to re-line the exterior of the kiln wall (fig. 15). Other types of silt coarse wares, typically found in settlements of this date, include beer jars, vases with ring-bases, shallow bowls with cord decorations, bread trays, beakers and thin-walled flared cups and carinated bowls. A small number of black-glazed bowls form part of the later material found within the deposits. The vast majority of ceramics were of Egyptian manufacture, made of Nile silt or marl fabrics, with a local red-brown Nile silt being the most common, often with a grey-black core and coarse temper of rounded quartz, limestone, chaff, dung and mica. It was used for the pigeon pots produced in the kiln (1619), and for a range of table-wares, coarse-wares, jars and amphorae. Egyptian marl fabrics were soft, crumbly and filled with organic temper, used for dishes with moulded external rims and carinated bowls. Finer marl fabrics were used for bottles and fine-wares. Imported marl fabrics were found with the the Phoenician and Palestinian amphorae. Those typical of the Levantine coast were found, with at least three different variants (orange-pink marl, pale yellow coarse lime tempered marl and rare black [volcanic] sandy marl). A soft brown silty fabric was also used for Phoenician amphorae, most likely produced at Gaza. Mixed silt and marl fabrics were also used for Phoenician amphora copies and ribbed rim amphora that could be of Egyptian origin. Excavated features in the Citadel area appear to be mainly of fifth and sixth century BC date. This date is confirmed by the proposed provenance of the Palestinian and Phoenician amphorae. Given the numerous architectural phases identified within this 200-year timespan (figs.4 5), the area excavated clearly consists of a densely occupied urban zone, inhabited by persons with access (direct or indirect) to vessels imported from elsewhere in Egypt but also from the eastern Mediterranean, but relying largely on locally produced vessels. The area feature houses with cooking areas, but also large-scale cereal storage silos and a kiln installation, seemingly dedicated to producing just one form of vessel: the pigeon-pot.

7 Finds All small finds from the excavations around the New Kingdom complex, the Late Period enclosure walls and the Citadel were collated, and new photography completed. This included a visit to the SCA magazine at Marea, west of Alexandria, where registered material is now housed. 17 Cleaning and recording Recent activity related to agricultural areas had revealed a significant length of mudbrick wall (fig. 19). A 5x3m trench (NC) was laid out, orthogonal to the edge of the kom, in an attempt to ascertain the alignment of the ancient wall (fig. 20). The wall (1900) is preserved to six courses in height, and built of bricks in a variety of clays, including a sand-rich yellow clay and crumbly black clay (37 42x17 22x9 199cm; fig. 21). The outside (northern) face of the wall had been cut away by farmers, but the eroded inside edge was defined. The wall was founded on a series of shelving deposits (1905-1912) rich in clay, brick debris and sherds (fig. 22). These deposits may represent a leveling operation in advance of the construction of the enclosure wall. Furthermore, the presence of ceramics of 26 th dynasty date beneath the wall (notably externally-moulded rims from pigeon-pots, in 0903), and a broader range of Late Period material above it (0901), confirm our previous interpretation that this wall is of 7 th or 6 th century BC date. Surveying of the wall confirmed it lies on the line of the Late Period enclosure wall (fig. 2), other surveyed segments having been revealed in the Citadel area (1030), and in the north-eastern part of the site (0436, 0600). 18 It is rather striking that the surveying revealed the northern edge of the kom was now 18m further south than in 2002, illustrating the gradual destruction of the ancient remains. It is hoped that the Supreme Council will continue to permit further research at Kom Firin, with a season planned for autumn 2010, and a second monograph, Kom Firin II: The Ramesside complex and the post-new Kingdom city (British Museum Research Publications), due to appear in 2011. Dr. Neal Spencer (Director, Kom Firin project) Assistant Keeper (Curator), British Museum nspencer@britishmuseum.org March 10 th, 2010. 17 We are grateful to Ahmed Kamal Adham for his help in providing access to these objects. 18 See Spencer, Kom Firin I, 24 5, pls.70 1.

8 Figure 1.Map of the region around Kom Firin. The location and extent of archaeological sites is taken from Survey of Egypt/EGSA maps; the positions of the modern road, the Rosetta branch of the Nile, canals and modern towns (labelled in upper case) are from satellite imagery. The course of the other two branches (the Canopic and Western Nile branches) is indicative only, being based on satellite imagery and survey maps. These channels will have meandered over time, and further bifurcations are likely to have occurred at different periods.

9 Wall segment 1900 (trench NC) Figure 2. Topographic map of Kom Firin, showing reconstruction of ancient river channels (based on auger cores, AS 01-10) and the phases of enclosure walls.

10 Figure 3: View south over excavated area within the northeastern corner of the Ramesside enclosure (trenches EC, EE) at the conclusion of the 2008 season. Figure 4. View south over excavated area of Citadel (trench CB) at the conclusion of the 2008 season.

Figure 5: Composite plan of architectural phases within Citadel (trenches CA, CB, CC, CD). 11

12 Figure 6: View north along trench NA (2006), with wall 0417 in foreground. Figure 7: East-facing section, northern end of trench NA, showing deposits accumulated over wall 0436. For location of section, see figure 8.

Figure 8: Plan of enclosure wall segments revealed in trench NA. Southern part of trench at top, northern continuation below. 13

14 Figure 9: Magnetometry survey of eastern Kom Firin, with Ramesside temple and enclosure visible near bottom of image. Trench NA runs across the walls in the northern area. Survey conducted by the University of Akron (2004-2005), led by Ann Donkin. Figure 10: Offering bowl C4125 from deposit 0402A (trench NA). Figure 11: Ring-base from a black-glazed jug C4141 from deposit 0406 (trench NA).

15 Figure 12: Samian amphora toe C4145 from deposit 0406 (trench NA). Figure 13: Fragment of a Cypriot black-painted juglet C428 from deposit 0406 (trench NA). Figure 14: Kiln 1619 in trench CD, overlying sandy deposit 1647, with structure 1605 around it. Feature 1657 with broken pots (C3141-3156) against east side of kiln. View south.

16 Figure 15: Detail of vessels (C3141-3156) as part of feature 1657 against east edge of kiln 1619. Brick wall 1605 in background. View southwest (trench CD). Figure 16: Detail of stamp on imported amphora handle C2013, from deposit 1044 (trench CA). Figure 17: Waster from a pigeon-pot, found in deposit 1636 within kiln 1619 (trench CD).

Figure 18: Egyptian siltware amphora C3086, found within deposit 1608 in trench CD. 17

18 Figure 19: Cut into the northern edge of Kom Firin, revealing Late Period enclosure wall (segment 1900). Figure 20: Mudbrick enclosure wall segment 1900 after excavation (trench NC), with fields to at top.

19 Figure 21: Plan of trench NC, showing wall 1900. Figure 22: Elevation of wall segment 1900, within trench NC (for location see fig. 21).