a hellenistic early roman shipwreck assemblage off ashkelon

21
A Hellenistic/Early Roman Shipwreck Assemblage off Ashkelon, Israel E. Galili Israel Antiquities Authority, POB 180 Atlit, 30300, Israel, [email protected] V. Sussman Institute of Archaeology, Bar-Ilan University, Ramat-Gan, Israel G. Stiebel Institute of Archaeology, Hebrew University, Jerusalem, Israel, [email protected] B. Rosen Israel Antiquities Authority, POB 180 Atlit, 30300, Israel Underwater surveys at Ashkelon during 1998 exposed a Hellenistic/Roman shipwreck. The ship, c.15–25 m long, sank in the 1st or 2nd century BC, and its remains include iron anchors, lead sheathing, copper nails, assorted metal artefacts, and shipwright’s unused nails. Bronze vessels found—oil-lamp, shovel and ladles—could have functioned in cult rites. Weights and balance-scale parts demonstrate commercial activity. Fishing-net sinkers indicate involvement in fishing. This paper raises important points regarding the risk to shipwrecks on the Israeli coast from environmental and human interference, and suggests that Ashkelon never had a built-up port. © 2009 The Authors Key words: trumpet, balance, weights, iron anchors, symbolic ritual, shipwright. T he Mediterranean coast of Israel is char- acterized by shallow sandy shores, few natural shelters, and frequent seasonal storms (UK Hydrographic Office, 1976). Vessels caught in a storm drifted ashore and wrecked in the surf-zone, and heavy objects were buried in the sand. Such buried artefacts were protected from destruction and salvage, and survived as intact assemblages (Frost, 1962; Muckelroy, 1975; Ward et al., 1999; Galili et al., 2002a). In the last 50 years changing coastal sedimentation patterns, caused by natural processes combined with human activities (sand quarrying and the con- struction of marine structures), have resulted in shortage of sand and exposure of archaeological sites on the shallow sea-bottom. These phenom- ena, in combination with intensified underwater archaeological exploration, have resulted in the discovery of numerous wreck-sites (Galili et al., 2002a). The construction of the Ashkelon marina and the electricity power-plant south of the city (Fig. 1), caused a significant reduction in the volume of unconsolidated sediments, mainly sand, in the coastal zone. This resulted in the exposure of wide areas of the sea-floor, revealing numerous archaeological assemblages (Galili et al., 1993; Galili and Sharvit, 1994). After expo- sure, the archaeological artefacts were threatened by marine erosion and treasure-hunting. This situation necessitated the initiation of rescue surveys to ensure documentation and retrieval of archaeological material. Underwater rescue surveys at Ashkelon were carried out by the Israel The International Journal of Nautical Archaeology (2010) 39.1: 125–145 doi: 10.1111/j.1095-9270.2009.00249.x © 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society. Published by Blackwell Publishing Ltd. 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

Upload: meztli-hernandez

Post on 01-Dec-2015

45 views

Category:

Documents


0 download

TRANSCRIPT

A Hellenistic/Early Roman Shipwreck Assemblage offAshkelon, Israel

E. GaliliIsrael Antiquities Authority, POB 180 Atlit, 30300, Israel, [email protected]

V. SussmanInstitute of Archaeology, Bar-Ilan University, Ramat-Gan, Israel

G. StiebelInstitute of Archaeology, Hebrew University, Jerusalem, Israel, [email protected]

B. RosenIsrael Antiquities Authority, POB 180 Atlit, 30300, Israel

Underwater surveys at Ashkelon during 1998 exposed a Hellenistic/Roman shipwreck. The ship, c.15–25 m long, sank in the 1stor 2nd century BC, and its remains include iron anchors, lead sheathing, copper nails, assorted metal artefacts, and shipwright’sunused nails. Bronze vessels found—oil-lamp, shovel and ladles—could have functioned in cult rites. Weights and balance-scaleparts demonstrate commercial activity. Fishing-net sinkers indicate involvement in fishing. This paper raises important pointsregarding the risk to shipwrecks on the Israeli coast from environmental and human interference, and suggests that Ashkelonnever had a built-up port.

© 2009 The Authors

Key words: trumpet, balance, weights, iron anchors, symbolic ritual, shipwright.

The Mediterranean coast of Israel is char-acterized by shallow sandy shores, fewnatural shelters, and frequent seasonal

storms (UK Hydrographic Office, 1976). Vesselscaught in a storm drifted ashore and wrecked inthe surf-zone, and heavy objects were buried inthe sand. Such buried artefacts were protectedfrom destruction and salvage, and survived asintact assemblages (Frost, 1962; Muckelroy, 1975;Ward et al., 1999; Galili et al., 2002a). In the last50 years changing coastal sedimentation patterns,caused by natural processes combined withhuman activities (sand quarrying and the con-struction of marine structures), have resulted inshortage of sand and exposure of archaeologicalsites on the shallow sea-bottom. These phenom-ena, in combination with intensified underwater

archaeological exploration, have resulted in thediscovery of numerous wreck-sites (Galili et al.,2002a).

The construction of the Ashkelon marina andthe electricity power-plant south of the city(Fig. 1), caused a significant reduction in thevolume of unconsolidated sediments, mainlysand, in the coastal zone. This resulted in theexposure of wide areas of the sea-floor, revealingnumerous archaeological assemblages (Galiliet al., 1993; Galili and Sharvit, 1994). After expo-sure, the archaeological artefacts were threatenedby marine erosion and treasure-hunting. Thissituation necessitated the initiation of rescuesurveys to ensure documentation and retrievalof archaeological material. Underwater rescuesurveys at Ashkelon were carried out by the Israel

The International Journal of Nautical Archaeology (2010) 39.1: 125–145doi: 10.1111/j.1095-9270.2009.00249.x

© 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society.Published by Blackwell Publishing Ltd. 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

Antiquities Authority (IAA) between 1994 and2004. During the winter storms of 1998 a ship-wreck assemblage from the late-1st–2nd centuryBC was detected and excavated by the IAA. Thewooden hull had disintegrated completely, andonly heavy items survived. Ship-related itemsincluded iron anchors with stocks, lead sheathingand copper-alloy nails. Other copper-alloy findsincluded an oil-lamp, an incense-shovel, a caul-dron, a decorated ladle, a trumpet and a set ofbalance-weights. These artefacts are describedand discussed below, listed by a cataloguenumbers and diving reports when available.

The siteThe site is situated off the northern suburbs ofAshkelon (Fig. 1) between two municipal

beaches. The coastline in the area is straight andsandy with no shelter or mooring facilities forseagoing vessels. The beach and the shallow seasadjacent to it are covered by sand underlain bykurkar (aeolianite sandstone) and beach-rock for-mations. Due to coastal erosion these rocks areoften exposed by the shifting sand. The artefactswere scattered on the exposed sea-floor at a depthof 3–4 m, 80–100 m west off the coastline. Most ofthe heavy artefacts were concentrated in a smallarea, 15 ¥ 15 m, probably the location at whichthe vessel was wrecked. The rest of the artefactswere scattered, fan-like, east and south-east of thisconcentration, toward the beach, over an area ofc.30 ¥ 30 m.

The anchorsFour iron anchors were located on the wreck-site—a pair of 2-armed anchors with their stocksnot inserted, and two 1-armed anchors with theirstocks in place. Only three were retrieved. Thepair of iron anchors (nos 1 and 2) with their stocksbeside them, were lying concreted together but inopposite alignments on the western edge of thewreck-site. The two stocks were lying beside them,parallel to the anchor shanks (Figs 2–4) (Table 1).The larger anchor (no. 1) was broken at the top.The shank has a rectangular cross-section nearthe arms, narrowing to a polygonal cross-sectiontowards the central part. The upper 35 cm of theshank is flattened to a rectangular cross-section.The upper 10 cm is corroded but the concretionshowed that the vanished top section had a hole inwhich an iron ring was inserted. Most of the ellip-tical ring (23 ¥ 34 cm) survived. It was formedfrom a circular iron rod, 2 cm in diameter. Theshank was thickened into a box-like shape c.40 cmbelow its top, between the flattened upper areaand the polygonal shank. Within this was a rect-angular hole into which the stock could beinserted. The two arms have a rectangular cross-section, their upper 15 cm being flattened, chisel-like. An extension of the shank, below the arms,

Figure 1. Location map.

Table 1. Basic measurements of the 2-armed anchors

Anchor no. Type Weight LengthDistance between

points of armsLength ofeach arm

Lengthof stock

Thicknessof stock

1 2-armed c.50 kg 190 cm (est.originally c.2 m)

103 cm 70 cm 158 cm 6 cm

2 2-armed 40 kg 143 cm 65 cm 45 cm 110 cm 5 cm

NAUTICAL ARCHAEOLOGY, 39.1

126 © 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society

had been drilled, and inserted into this hole was asalvaging-ring, 12 cm in diameter, made from around iron rod of 2 cm diameter. The slightly-curved stock is rectangular in cross-section, withtwo round holes in the centre. One was pluggedpermanently by a protruding rectangular flat pin(4 ¥ 4 cm) which was riveted into the stock byhammering it on the side opposite the pin. Theprotruding pin was hollowed and a ring (4 cmdiameter), with an S-shaped chain link attached toit, was inserted in this hole (Fig. 5). A lost pin,which was probably hanging on this chain, couldhave been inserted in the second hole. The whole

fixture was designed to lock the movable stockfirmly in place (Fig. 6). As far as we know such amethod of joining the stock to the shank in antiq-uity has not been previously reported.

The smaller anchor is complete (Fig. 7)(Table 1). The shank has a rectangular cross-section in the lower section near the arms, and aheptangular cross-section in the upper part. Theupper 20 cm of the shank is flattened to a rectan-gular cross-section. It is hollowed at its top withan iron ring inserted in the hole. Most of theelliptical ring (170 ¥ 125 mm) survives, formedfrom a circular iron rod 15 mm in diameter. Theshank is thickened into a box-shape c.20 cmbelow its top, between the flattened upper areaand the polygonal shank, creating a rectangularhole into which the stock could be inserted. Thetwo arms have a rectangular cross-section andtheir upper 10 cm is flattened, chisel-like. Anextension of the shank, below the arms, wasdrilled and a salvaging-ring was probably insertedinto this hole. The slightly-curved stock is rectan-gular in cross-section, with two round holes in thecentre. One of the holes was plugged permanentlyby a protruding round, flat pin (5 cm in diameter)which was riveted to the stock by hammering it onthe opposite side. The protruding pin was hol-lowed to insert a ring with chain and another pin

Figure 2. Divers checking the iron anchors in situ (top): no1 proximal, no 2 distal (bottom). (S. Ben-Yehuda)

Figure 3. The iron anchors displayed in the National Mari-time Museum in Haifa. (E. Galili)

E. GALILI ET AL.: A HELLENISTIC/EARLY ROMAN SHIPWRECK ASSEMBLAGE OFF ASHKELON, ISRAEL

© 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society 127

(Fig. 6) similar to anchor no. 1, though neitherring, chain nor pin survived.

Two 1-armed anchors were discovered, withtheir stocks inserted, c.30 m south-east of themain artefact concentration (Galili and Sharvit,1998; Galili, et al., 2001). They were about 10 mapart and only one of them, no. 3, was recovered(Fig. 8). This anchor, which has lost muchmaterial through corrosion, weighs c.3.3 kg. The

Figure 4. Iron anchor no. 1. (S. Ben-Yehuda)

Figure 5. Detail of stock of anchor no. 1. (E. Galili)

Figure 6. Reconstruction of the 2-armed iron anchors.(S. Ben-Yehuda)

Figure 7. Iron anchor no. 2. (E. Galili)

NAUTICAL ARCHAEOLOGY, 39.1

128 © 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society

shank, rectangular in cross-section in the lowerpart, tapering to a square section in the thinnerupper part, and the rectangular cross-section arm,were forged together forming one piece. A roundhole in the upper part of the shank retains theremnants of an iron ring, 11 cm in diameter.Below the ring, an elongated stock, with a rectan-gular section, is attached to the shank by a roundrivet. The stock was placed asymmetrically so thatits mass faced the single arm. In the lower part ofthe shank there is a hole in which a salvage-ropewas once attached, perhaps by a ring.

DiscussionOne-armed anchors from antiquity are relativelyrare. They were first mentioned by Kapitän (1971)who published a 1-armed wooden anchor fromSicily. Examples from Israel have been discussedrecently (Galili and Rosen, 2007). One almost-intact wooden example was recovered from theMa’agan Michael shipwreck (Rosloff, 1991).Assembly pieces and stocks of such anchors werediscovered in Atlit and Yavne Yam (Galili andRosen, 2007). A 1-armed iron anchor similar tothe ones from Ashkelon was recovered south ofApollonia (Galili and Rosen, 2007). Its originalweight was c.10–15 kg, its shank and stock were

rectangular in section and attached asymmetri-cally by a round pin with the mass facing the arm.Possibly the stock was detachable. In the upperpart of the shank were traces of a hole holding aniron ring. One-armed anchors were mentioned inEgyptian papyri, so their presence in the Romanand Byzantine periods was known prior to theirdiscovery by underwater archaeologists. Casson(1971) suggested that they were manually placedon the shore, substituting for a mooring-stake or-stone. The most significant feature of this anchoris that the stock is joined to the side of the shank.This type of anchor was used continuously fromantiquity to the present—modern version arecommon on yachts.

Two-armed iron anchors have apparently beenused, in conjunction with wooden anchors, sinceat least the 3rd century BC (Haldane, 1984: 10).Ancient 2-armed iron anchors were dated andclassified by Kapitän (1984), and his classificationhas been used by several subsequent scholars (forexample Bockius, 2000). His early iron anchortype A, in which the arms on both sides of theshank are in the shape of an arrowhead, resemblesearlier and contemporary wooden anchors(Kapitän, 1984), and is dated to the 3rd to 1stcenturies BC. One example from the 1st century

Figure 8. One-armed anchor no. 3. (S. BenYehuda)

E. GALILI ET AL.: A HELLENISTIC/EARLY ROMAN SHIPWRECK ASSEMBLAGE OFF ASHKELON, ISRAEL

© 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society 129

BC appears in a mosaic from Pompeii (Ucelli,1950: 246). Kapitän’s type-B anchors are gener-ally later than type A. Two type-B anchors wererecovered off Neve Yam, Israel near a Hellenisticshipwreck-site dated to the late-2nd to early-1stcentury BC (Galili and Sharvit, 1999b). A type-Banchor was recovered with the Lake Nemi ship,dated to the 1st century AD (Speziale, 1931). Itseems that both types could be used simulta-neously, as depicted in a graffito from Beit Jibrin,Israel (Bridley, 1919). The pair of 2-armed ironanchors from Ashkelon can be classified by theirshape as an intermediate type between A and B.Thus dating them to the 1st to 2nd centuries BCseems to be justified.

The copper-alloy nailsBoth used and unused nails, all made of a copperalloy, were recovered from the wreck-site(Table 2). A concentration of 53 unused, straight,nails (nos 101–153), all of round cross-section,was recovered from an area of less than 2 m2

(Fig. 9). Additionally 14 (nos 160–174) bent/twisted/cut nails, bearing hammer-marks on theirheads, were recovered from a relatively wide area,encompassing the whole wreck-site (Fig. 10). Allthe nails have their heads, and no stem fragmentswere recovered, unlike other underwater wreck-sites. However, four of the recovered used nails

were damaged: three (nos 162, 168, 170) weremissing a small portion (1–3 cm) of the point, andone (no. 172) was missing more than a half thestem. The nails are described below according totheir properties, shapes, forms and apparent pat-terns of use, based on a typology established for alarge nail assemblage recovered from a Romanshipwreck off the north Carmel coast, usinghead-shape and stem cross-section as classifyingattributes (Galili et al., forthcoming).

All the nails have round cross-sections. Theunused nails have a stem diameter (just below thehead) ranging from 5–6 mm among the threeshortest, and 7.2–7.3 mm among the three

Table 2. Properties of the used nails

Cat.no.

Use type(fig. 10)

Totallengthin mm

Max. headdiameter in

mm

Max. stemdiameter in

mmWeightin gm Remarks

160 C 300 18 10 56 almost whole, bent at 174 mmhead to stem joint straight

161 D 270 18 8 44 almost whole, head to stem jointstraight

162 D 250 19 7.5 56 possibly up to 1 cm lost, headto stem joint straight

163 D 240 15 8 31 head to stem joint conical164 D 220 13 7 31 head to stem joint conical165 A+C 230 17 9 44 head to stem joint straight166 C 235 16 8 38 bent at 15 cm167 C 200 16 9 39 Small part of tip missing, head

to stem joint straight168 C 175 18 9 43 about 1–2 cm of tip missing,

bent at 15 cm169 A 165 19 8.5 41 head to stem joint conical170 A 145 15 7.5 33 broken, head to stem joint conical171 D 125 16 7.2 18.5 almost whole172 D 105 19 7.9 32 broken part of long, massive nail173 D 90 15 6 13.5 whole short, perhaps re-used

Figure 9. Unused nails nos 101–53, scale in cm. (E. Galili)

NAUTICAL ARCHAEOLOGY, 39.1

130 © 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society

longest. Generally the longest nails are the thick-est, but there are exceptions, which may have beencaused by deviation in production, or later bycorrosion. The diameters of the stems of the usednails vary between 8–9 mm in the three longest,and 7.5–10.5 mm in the shortest. There is no cor-relation between length and diameter of stem; forexample, the thickest used nail (no. 169, 16.2 mm)is of medium length relative to other used nails.

The lengths of the unused straight nails varyfrom 132 to 227 mm, and they do not cluster intogroups of roughly similar lengths. The lengthsof the used nails, however, vary between 90and 299 mm and they can be divided to threegroups: long nails (240–299 mm), nine, of whichtwo were broken; medium (125–164 mm), three,one broken; and one short nail, 89 mm long. Onenail (no. 172) is broken, with only about halfsurviving.

Two basic nail shapes (according to the typol-ogy of Galili et al., in press) were identified, G andI. In type G the head-shape is a flat cone or domewith a relatively thick rim (2–4 mm) circumscribedby a lip pointing downwards. The stem joins thebottom of the head at a straight angle. In type I thehead is a flat dome, its circumference ending in asharp tip, with no apparent widened lip. The stemjoins the head either with an arch-shaped widenedarea, or straight, as in type G. A few nails belong toan intermediate type between G and I. Most of theunused nails (44 of 53) belong to type G, six to typeI and three are undetermined. Most of the usednails (13 of 14) have type I heads.

The head-diameter of the three shortest unusednails varies between 13.5 and 15 mm, and that ofthe three longest 17.5–18 mm. Generally thereis a gradual increase of head diameter as lengthincreases, but there are several exceptions. Appar-ently the manufacturer of the nails did not pay asmuch attention to this dimensional relationship ashe did to the head-shape or the stem-diameter.

The head-diameter of the used nails variesbetween 15.5 and 19 mm. One nail (no. 164)had a considerably narrower head-diameter of13.2 mm, but this may be partly due to corrosion,as it belongs to the long nails group. There seemsto be no clear association between the head-diameter and length among the used nails, as allthree length-groups had approximately the samerange of head-diameters.

In general, in both used and unused nails, thereis a positive correlation between the length andthe weight. However there are exceptions. Theweights of the used nails vary between 13.5 g (no.173, 90 mm long) and 56 g (no. 162, 250 mmlong). The longest unused nail weighs 37.2 g (no.101, 226 mm long), the shortest 12.4 g (no. 153,133 mm long). The weight of the unused nailsvaries between 43 g (no. 11, 206 mm long) and12.4 g (no. 153, the shortest at 133 mm).

The unused nails have no signs of workingdamage or attrition on the heads, and no twists orbends in their stems, which were unusually straightfor nails recovered from shipwrecks. The used nailsdemonstrate several use-patterns: most have dis-tinct hammering marks on the head; five weredeliberately bent at about 90°, obviously to jointwo timbers together, probably the planking toframes (Fig. 11), but no clear clenching by doublebending was observed. In these five nails thedistance between the head and the bend rangesbetween 150 and 165 mm, approximately thethickness of the joined timbers, the planking andthe frame; three nails seem to have been bent byforces generated during the disintegration of thehull (Fig. 11). The distance from the underside ofthe head to this bend varies between 45 and 58 mm.One of these three also showed a deliberate con-struction bend. Six used nails, four long and two

Figure 10. Used nails, scale in cm. (E. Galili)

Figure 11. Reconstruction of nail deformation duringwrecking event. (S. Ben -Yehuda and E. Galili)

E. GALILI ET AL.: A HELLENISTIC/EARLY ROMAN SHIPWRECK ASSEMBLAGE OFF ASHKELON, ISRAEL

© 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society 131

short, have undefined twists and bends. In one nail(no. 172) with a type G head (unlike the rest of theworked nails with type I heads) most of the stemand point is missing. It has hammering marks onthe head and cutting marks on the stem, near thebreak, which seem to indicate that it was deliber-ately broken prior to the wrecking.

DiscussionThe unused nails may represent a cache of newnails belonging to the ship’s carpenter; less prob-ably they could represent trade goods. The usednails may be the remains of hull fastenings. Theyare spread over a relatively large area withoutassociation with any other pieces of scrap metal,so are unlikely to represent a collection of scrapmetal such as has been occasionally foundon shipwreck sites. Generally, in Israel, copper-alloy nails with round cross-section originatefrom Hellenistic/Early Roman shipwrecks, as forexample the Hellenistic shipwreck at Megadim(Galili et al., forthcoming).

Lead-sheathing fragmentsThe wreck-site yielded five broken, twisted andtorn fragments of lead sheathing (nos 4–8) (200–250 ¥ 100–170 ¥ 2 mm) (Fig. 12). They haveseveral perforations, both round (3 mm diameter)and square (3 ¥ 3 mm), arranged in a randompattern, marking the location of the nails whichwere once inserted in them to attach them to thehull. The imprints of the nail-heads are visible(Fig. 12 bottom). Lead sheathing was a thin layerof lead (average thickness 1.25 mm) fixed to theouter surface of wooden-hulled ships, from thekeel up to the first or second strake above thewaterline; the pieces overlapped each other, andwere fixed with copper tacks with large heads,enabling water to flow smoothly over the hull.This sheathing enhanced the ship’s structuralintegrity and stability, and protected the timbersfrom wood-borers. It was used from about themid-4th century BC to about the 2nd century AD(Kahanov, 1999: 219). The pieces of lead sheath-ing from the site could be fragments of completehull sheathing, or patches used to seal leaks and toprotect the timbers from biological deterioration.

Lead fishing-net sinkersSixteen lead artefacts definable as fishing-gearsinkers were recovered from the site (nos 9–24),and are classified according to shape, mode of

manufacture and to some extent functional pro-perties (Galili et al., 2002b). One sinker was aplano-convex net-ring (type L1.3.1), ten weresmall folded rectangular sinkers (type L2.3) andthere were two large sinkers in the form of a platebent to form a tube (type L2.2). Three of thefolded rectangular sinkers (type L2.3) have pro-truding decorations (diving reports 37/96/53;37/96/37). All three have a protruding line just onthe rectangle rim. One (92 ¥ 34 ¥ 2–23 mm) weigh-ing 46 gm bears inscriptions and two crossed linesin the shape of a long X bordered by two parallellines (Fig. 13). On one narrow side there are 3 or4 unclear letters in one line, parallel to the end. Onthe other narrow side there are about six unclearletters arranged in lines, parallel to the edge of thelong side.

The second decorated sinker (open 92 ¥ 31 ¥1.5 mm) weighs 46 g and is decorated with aRoman iron anchor, with two bent arms, thestock inserted, and a tying ring on the top. Theanchor is hanging inside a gable-roofed structure,its walls indicated by two pairs of parallel lines,standing on a foundation marked by crowded loz-enges. The third rectangular folded sinker (open

Figure 12. Lead-sheathing fragments bearing nail-holes(top), impression of a nail-head (bottom. (E. Galili)

NAUTICAL ARCHAEOLOGY, 39.1

132 © 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society

74 ¥ 44 ¥ 1.5 mm), weighing 36 gm, is bordered bya frame formed by two lines on each of the longsides and single lines on the short sides. Inside thepanel enclosed by the square frame is a 2-armedRoman anchor. On the upper part of the anchorshank is a lateral line and above it a round pro-trusion, probably the anchor-ring (Fig. 14). Theten small folded rectangular fishing-gear sinkersof type L2.3 probably came from a lead line or thefoot-rope of a cast net. The two large sinkers, type

L2.2, as well as the three decorated sinkers (typeL2.3) may have belonged to a gill or trammel net.The ring-shaped sinker could have belonged to abeach seine. The sinkers could have belonged toelements of fishing-gear habitually maintained onboard, as in numerous Roman merchantmen(Galili et al., 2002b; Galili et al., forthcoming).There is also the possibility that some of them areintrusive artefacts from other shipwrecks or fromfishing activities in the area. Another, less prob-able, explanation is that they represent scrap leadcollected for recycling.

Lead sounding-weightOne sounding-weight (no. 25) was recoveredabout 20 m south-east of the main concentrationof heavy metal artefacts. It is of type 1 (see Galiliet al., 2009), weighs 4.3 kg and is 125 mm high,with a maximum diameter of 85 mm (Fig. 15).The internal diameter of the straight lug hole is20 mm. It is shaped like a tall truncated cone withalmost straight sides. The suspension-lug was castwith the body and the head-angle is 10°. The twoupper corners of the rectangular lug are roundedand on the two long opposing sides it is separatedfrom the body by well-defined, almost right-angled, shoulders, while the narrow sides continuewith the body. In the bottom there is a shallow(6 mm deep), plain, concave tallow-cup type B.On the side are traces of an inscription madewith a cold chisel after casting. Several letters,arranged in an approximately straight line paral-lel to the bottom of the weight, can be discerned,but have been partly erased by erosion andimpact-indentations. Numerous wear-marks arespread over the sides, bottom and tallow-cup.

Figure 13. Lead fishing-net sinker decorated with unidenti-fied inscriptions. (photo E. Galili; drawing S. Ben-Yehuda)

Figure 14. Lead fishing-net sinker decorated with image ofiron anchor. (E. Galili)

Figure 15. The sounding-weight, no. 25. (E. Galili)

E. GALILI ET AL.: A HELLENISTIC/EARLY ROMAN SHIPWRECK ASSEMBLAGE OFF ASHKELON, ISRAEL

© 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society 133

DiscussionA typology of Israeli sounding-weights was estab-lished based on a collection of 66 examples (Galiliet al., 2009). The single example recovered fromthe Ashkelon shipwreck has certain characteris-tics found in type 1 (Late Roman/Byzantine).However, as it can not be considered as a true type1 example, the typology cannot be used to date it.Perhaps it is an early version of this type, or it maybe an intrusion, and not from the wreck beingdiscussed.

Bronze weights with a lead coreEight bronze weights (nos 300–307) with an innerlead core were recovered from the wreck-site.They range in weight from c.26 kg to c.0.9 kg(Table 1). No. 300, the heaviest, is in the form ofa rectangular box, c.18 ¥ 17 ¥ 10 cm (Fig. 16 top),with the top and the base extending slightly away.The lip of the base forms a sloping ledge sur-rounding the body, while the top is decorated byan extended roof-like cover, decorated with an

oval-topped ridge forming an internal rectangle,inside which a semi-oval loop (circular in cross-section), cast with the object, holds a suspension-ring. It is possible that the ring was also formedwhen the object was cast, and was subsequentlyseparated from the attaching loop and workedinto its final shape. The walls of the box arec.8 mm thick, and filled with a lead core, coveredby a thin plate tightly pressed into the base.

All the other weights (nos 301–307) (Fig. 16bottom) have a cylindrical or slightly cone-shapedbody with a mushroom-shaped top. Below the topall have a circular line, circumscribing the objectand forming a very slight ledge. Five of the sevenhave a similar line on the top, close to the mush-room head. All have had a suspension-ringattached to the top centre by a metal strip forminga loop. In the two heaviest (nos 301, 302) thesuspension-rings and the holding-strip survived.In no. 303 only the strip survived. In the otherfour, only traces of the strip can be discerned.Inside the outer wall (3–11 mm thick) is lead,inserted from the bottom, which was originallytightly sealed by a cover. The complete bottomcover remained in weights nos 301 and 303. Inweight no. 302 only part of the cover remainedand a copper-alloy nail or rivet was inserted in thebottom, probably to hold the bottom cover.Weight no. 306 has five copper nails, or rivets,stuck in the bottom of the lead core which mayhave served to hold the cover. The bottom cover ismissing from all the other weights.

The cylindrical weights range from c.4.6 kg to0.9 kg, in proportion to their size. Two pairs aresimilar in weight and dimensions. Nos 302 and303 weigh 4.367 and 4.266 kg respectively, sug-gesting that when undamaged (303 is missing itshanging-ring) they had the same mass. The otherpair, nos 304 and 305, weigh 2.643 and 2.446 kgrespectively. No. 305 is heavily eroded andmissing a large part of the bronze envelope, butprobably originally weighed the same as its pair.

Bronze balance-scalesThree bronze hanging devices (nos 26–28) wererecovered from the wreck-site. Two (nos. 27 and28) are similar, omega-shaped bucket-handlesweighing c.200 gm each (Fig. 17 top). They wereformed of a round-sectioned copper rod, taperedtowards both ends, and have openings of 24 and23 cm respectively. The third (no. 26, weighing498 gm) is composed of two connected parts,a free-hanging ring (73 mm in diameter) and

Figure 16. Rectangular weight, no 300 (top); cylindricalweights, nos 301–307 (bottom). (E. Galili)

NAUTICAL ARCHAEOLOGY, 39.1

134 © 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society

a crescent-shaped bar loosely joined together(Fig. 17 bottom). This bar has a tube on the apex,both formed by one casting, with a round hole inthe top of this tube. The free-hanging ring wasinserted into the tube. The ring (outside diameter74 mm) is bent, tightly closed, but not welded atthe joint. It is made of a copper-alloy bar, round-sectioned, 9 mm in diameter. The bar forming thecrescent has a triangular section. The flat outerflank of the crescent forms the base of the triangle,while the inner, knife-like, edge is formed by thetriangle top. Two parallel grooves across the knifeedge can be seen in the inner side of the crescentapex, opposite the tube holding the hanging ring.Two rivets are inserted through this implement,about 4 cm from the tips. Judging by distributionof the three artefacts on the wreck-site and thesimilarity in materials, dimensions and generalcharacter, they seems to be closely associated.

It is proposed that the balance-weight set is tobe associated with these artefacts, which formedparts of a massive balance of which the beamand the carrying-frame would have been madeof wood. The two omega-shaped objects were

connected to the tips of this wooden balancebeam. The balance-weights were placed on oneside of the beam, and the object to be weighedsuspended from the other. Baskets or similar per-ishable objects could have been used to hold bothweights and the weighed merchandise. A similarbalance is depicted on a North African mosaicshowing the unloading of a beached ship (Casson,1971: fig. 191) (Fig. 18), and on a 1st-centuryRoman tomb (Kleiner, 1992: 109). The crescent-shaped bar formed the pivot of the balance. It wasattached to the vanished wooden beam by therivets seen on its tips. The two parallel grooves onthe inner side of the apex could mark the placewhere this implement was positioned on the oppo-site edge which was fixed to the lost woodenframe.

A set of bronze vesselsAmong the artefacts recovered were four thatcould have been household kitchen vessels, orutensils for measuring trade goods, or intendedfor some special cult purpose. They were recov-ered from a relatively small area (3 ¥ 3 m) inthe centre of the site. The set (Figs 19 and 20)includes an oil-lamp, two ladles, and a shovel.Similar assemblages have been found on varioussites from several periods, including the Persian/Hellenistic assemblage from Shechem whichincluded a lamp and a strainer (Stern, 1980: figs.5:2, 5; pl. XIV:C and XV).

Oil-lampThe oil-lamp (IAA diving report 37/96/46/2)(Fig. 19) is constructed of a round flat body and

Figure 17. Bronze balance-scale hanging devices: nos 28(top) and 26. (E. Galili)

Figure 18. Large balance-scales in use. (E. Galili, afterCasson, 1971: fig. 191)

E. GALILI ET AL.: A HELLENISTIC/EARLY ROMAN SHIPWRECK ASSEMBLAGE OFF ASHKELON, ISRAEL

© 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society 135

an elongated nozzle, terminating in a round(16 mm diameter) wick-hole; a horizontal, slightlycurved handle is attached at the opposite side tothe nozzle. The two ends of this handle, made of around rod (7 mm diameter), are attached underthe shoulder, turning upwards and projectingslightly above the body. The lamp weighs 373 gmand is 166 mm long including the handle, 93 mmwide and 37 mm high, and stands on a ring-base54 mm in diameter. The sides of the lamp arecurved, ending in a sharp angle with the shoul-ders, which are narrow and flat, forming a com-plete circle. A wide circular filling-hole, 46 mm indiameter, surrounded by a low rim, 5 mm high, isplaced in the middle of the shoulder. This rim hasan inner, narrow (4 mm wide) flange on which amissing cover once sat, attached to the double-pierced hinge resting on the rim of the hole. Thelamp was cast as one piece using the lost-waxprocess.

Such metal artefacts are hard to date, becausethere are few precise parallels. Bronze vesselsremained in use for a long time, relative toceramic ones, as they were more durable andcostly. To date such lamps one has to rely onartefacts in the associated assemblage, such ascoins, or C14 dating of organic objects. If theseare not available, dating can rely on mass-produced ceramic oil-lamps which are moreclosely dated, helped by artistic, stylistic elementsin the design. The nozzle, narrow and elongated,flattened on three sides and curved at the bottom,continues Hellenistic traditions. The U-shaped

Figure 19. Bronze lamp, no 29 (top and middle), andbronze shovel, no. 32 (bottom). (E. Galili)

Figure 20. Bronze ladle no 31 decorated with a duck’s head (top); detail of the ladle pot (middle left); bronze ladle no. 30decorated with three duck’s heads (bottom); two duck’s heads facing in opposite directions (middle right). (E. Galili)

NAUTICAL ARCHAEOLOGY, 39.1

136 © 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society

handle has a long history both on metal andpottery lamps, and on other open vessels like largebowls and kraters from the Greek Archaic to theRoman period. Such handles were used in lampsat least since the 4th century BC and up to theEarly Roman period (Bruneau, 1965: Delos,pl.1.11, 4th–3rd century BC, and 42, 3rd–2ndcentury BC. Agora type 33–36). One of the crite-ria separating Hellenistic metal oil-lamps fromRoman or later examples is the height of thehandle projecting above the receptacle (Bailey,1972).

Cooking vessels with horizontal handles bentupwards are known from Ephesus (Zabehlicky-Scheffengger and Schneider, 2005: fig. 1). Suchhandles are used for hanging the vessel when notin use. However, the circular flat narrow shoul-ders also appear in Hellenistic oil-lamps. Thelamp may therefore be dated to the Late Hellenis-tic or Early Roman periods. Metal oil-lampscovered with a lid, which are less liable to break-age than ceramic ones, are useful on board ship.At least five bronze oil-lamps from Hellenistic,Roman and Byzantine periods have been recov-ered from shipwreck sites off the Israeli coast(Galili et al., forthcoming). Bronze lamps are rareon land sites, relative to ceramic ones, as brokenone were recycled.

Incense-shovelThe rectangular shovel (IAA diving report 37/96/46/3) (length 157 mm) (Fig. 19) weighs 193 gm ofwhich the concretion amounts to c.2–3%. It wascast as one piece by the lost-wax process. Theoverall dimensions are 158 ¥ 61 ¥ 21 mm. Theblade (61 mm ¥ 75 mm) has a flat base and asocketed handle. The edge around three sideswidens only slightly towards the 5-mm-wide rim.The grooved rim is folded at the rear into twopointed stretched corners, as if it was a bent plate.At the front the shovel base forms a slightlycurved, sharp opening to ease loading. Thehollow, cone-shaped, ridged handle is attached tothe shovel by a flat trapezoidal protrusion. Thehandle could have been extended by inserting awooden or metal extension into the hollowhandle. The artisan casting the shovel formed anintentional split on the bottom of the hollowhandle to prevent cracking when inserting awooden extension. The handle widens towardsthe hollow end, resembling a fluted column deco-rated by annuli. It terminates, on the top, in adouble frieze on which the plain flat section, thecapital, holds the shovel, and the two folded side

projections represent the acroteria. The base ofthe column is decorated, crossed by a wide bar.

Such tools have a long history, used in thehousehold and on cult occasions. In the Jewishcult they are known and described as the Mahta.This artefact is depicted on Byzantine mosaicfloors and on oil-lamps, placed beside the SevenBranched Menorah of the Second Temple Period.Shovels vary in size. Some are very elaborate anddecorated in varying styles. Several artistically-fashioned shovels were found in association withdecorated metal jugs and bowls in the Judeandesert caves (Yadin, 1963: 48–58). The shovel wasmade to hold small quantities of materials such asspices or incense. There are no indications of fireon this example.

Like the associated lamp, it is hard to date. Thestyle of handle could be influenced by Greekarchitecture. A shovel from Cyprus (Mitten, 1965:pl. II) with a handle in the form of a Corinthiancolumn, standing on four legs, was dated to theRoman period. A similar shovel, but larger, with avery long twisted handle and a ring for hanging(Karageorgious, 1987: 719, fig. 186) was dated tothe Cypro-Archaic-Cypro-Geometric 1 (1050–475BC). A number of similar-shaped iron shovelswith elongated handles were found in Israel: onenear the altar of the Temple at Tel Dan (Biran,1994: 181), another in the debris of the HellenisticPeriod Temple excavated at Beersheba (Aharoni,1975: 163–5, pl. 36:1) which was probably fromthe time of Yanaeus (2nd–1st century BC) andanother shovel, with elongated handle, was foundin a tomb at Jerusalem of the Second TemplePeriod (1st century BC to 1st century AD)(Mazar, 1982: 45, fig. 2:17). Finds from a burialtomb at Kafr Kama include a 10-cm-long bronzeshovel (Ben-Nachum, 2007: 106, fig. 4:3). Glassvessels in this tomb were dated by Gorin-Rosed tothe late-2nd–3rd century AD. An iron shovel (130¥ 90 mm) identical to the one discussed herewas discovered near Tel Abu Shusha (Siegelman,1988: fig. 53). Pottery incense-shovels were foundat Sepphoris and their function is discussed byMeyers (2006: figs 5–13). Because the shovel wasfound with the oil-lamp it is suggested that it maybe dated to the Late Hellenistic or Early Romanperiods.

Bronze ladleThis artefact (IAA diving report 37/96/46/5)(Fig. 20) consists of a bowl/cup (36 mm high ¥43 mm diameter) joined to a long handle.It is 540 mm long and weighs 188 gm of which

E. GALILI ET AL.: A HELLENISTIC/EARLY ROMAN SHIPWRECK ASSEMBLAGE OFF ASHKELON, ISRAEL

© 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society 137

concretion amounts to c.5 gm. Most of the cup’sbottom is missing as well as part of its side, theloss amounting to c.10–15 gm. The cup, having acurved base, was made in one piece and designedto draw about 30 ml of liquid from a deep jar oramphora. The handle is wide at the base with arectangular cross-section; it narrows to a square,tapering toward the end, which is shaped in aschematic swan’s head, and curved so it could behung up when not in use. Several parallels havebeen found, and the volume of their bowls seemsto be similar, though the length of their handlesvaries. Roman parallels of no specific date werecited by Guhl and Koner (1994: 449b). Twosimilar ladles come from Sarepta-Sarafend on thePhoenician coast. The materials associated withone ladle were dated to before 50 BC (Pritchard,1988: fig. 26). Both are wider and shallower thanour example. A similar ladle was reported fromCyprus (Flourentzos, 1978: fig.1:2; pl. 18:5) intombs ranging from Late Archaic to the Romanperiod.

Bronze ladle (simpulum)This implement (IAA diving report 37/96/37/12),weighing 391 gm, was made of two main parts, asmall spherical pot (75 mm high ¥ 94 mm diam-eter; 200 ml) to which an elongated handle wasadded (Fig. 20). The ball-shaped pot has a wide,short neck with flaring thin rim. The long handlecould have been intended to keep the user awayfrom heat, when holding over a fire, or to facili-tate ceremonial pouring. This vessel was designedfor transferring liquids from shallow containers.The handle, cast as a single unit, is divided intofour parts, a near part holding the pot, and threeparts forming the handle. The part hugging thepot is made of two wires extending left and rightfrom the near end of the second part, joining onthe side of the pot opposite to the handle byfolding the two ends together. The second part isa round rod, decorated with three sets of annularrings and a duck with its beak pointing toward thepot; it gets slightly wider where it joins a thirdpart, which is in the form of a flat bar extendingfrom the round rod. It is joined at the far end byanother duck’s head, with its beak away fromthe pot, to the fourth part, which is a thin rodbent into a hanging loop. Its tip ends in a thirddecorative duck’s head facing the pot (Fig. 20).Round handles with flat sections appear onwine-strainers such as one described by Pearson(1992:51) and Robinson (1941: pl. L). Comstockand Vermeule (1971: 606) described such a

strainer as Etruscan and dated it to the 5thcentury BC or later. Bronze artefacts similar tothe Ashkelon simpulum are displayed in the RoyalOntario Museum (cat. no. 117) (Hayes, 1984).

DiscussionThe bronze vessels recovered from the wreckcould have been used as kitchenware or measur-ing devices by traders, but also for other purposes.The character of these artefacts did not changemuch over time, including the use of decorativeduck’s or swan’s heads on handles, possiblymaking a link between liquids and water-fowl.Dating these vessels is difficult. A patera with awavy handle made of a rod (Flourentzos, 1978:fig.1, pl.IX:1–3, 5) is similar to ours but slightlywider (diameter 5.5 cm) and its handle is 285 mmlong. Ladles ending in a swan’s head are commonin Cyprus (kyathos) ranging from Late Cypro-Archaic to the Roman period (Robinson, 1941:194–7, pls L613–22, LI;). The swan’s head as anornament developed in the Orient during the 2ndmillennium BC, and remained a favourite into theRoman period. Naturally with hand-made objectsthe heads differ slightly from one ladle to theother; technically it was easy to turn the square orrounded rod/wire into a decorative element (byhammering?).

A decorated ladle is listed among the Hellenis-tic Period finds from Tel Anafa (Weinberg, 1971:106, pl. 19A). It has a duck’s-head handle and thebottom is decorated like an East Greek (Megar-ian) bowl. Among a large number of quality wareswas a strainer with a horizontal handle, dated100–75 BC. The shape of the body has parallels inHerodian 1st-century-AD pottery vessels (Bar-Nathan, 2002: 102–13, pl. 11.139). The origin ofthese vessels is difficult to determine, and theycould have been obtained in various locations.The workshop producing the lamp could havebeen in Egypt, as similar lamps are found in Alex-andria (Abdou Daoud, 1998). Similar ladles areknown from all over the Mediterranean. Besidethe mundane use of drawing liquids, the presenceof a set of such metal vessels on the shipwreckmay have another explanation, such as religiouscult activity. The shovel is a common Easterntool. Each of the individual bronze items mighthave had an everyday use, but together theysuggest that they were used as a group, perhaps insome ceremony. Finds from shipwrecks along theIsraeli coast show that ritual, ceremonial andapotropaic activities were commonly conductedaboard Greco-Roman watercraft. The possibility

NAUTICAL ARCHAEOLOGY, 39.1

138 © 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society

that this assemblage was associated with suchactivity should not be discounted.

Bronze trumpetA straight copper-alloy trumpet was recovered(IAA Diving Report 37/96/46/4) (Figs 21–22). Itwas found in two pieces, total length c.900 mm.After examination and reconstruction it seemsthat the instrument was composed of two parts: acast, solid copper-alloy mouthpiece which was notremovable, attached to an elongated tube 86 cmlong, ending in a bell-shaped cone, the end ofwhich is now missing. The minimum diameter ofthis tube, adjacent to the mouthpiece, is 17.5 mm,

and the maximum, at the end of the bell, is58.5 mm. The body was made of a long andnarrow band of metal (c.56 mm wide) which wasformed into a tube and joined by soldering. Theband was bent over three solid copper-alloy tubes,c.130 mm long and 3 mm thick, which acted asthe skeleton of the instrument’s body. The mouth-piece is a cast of copper alloy, 58.5 mm diameterand 37.5 mm long, shaped in a wide cup whichfitted over the lips of the blower (Fig. 21). Thecentral hole is crowned by a raised circularpattern. The mouthpiece gradually narrows to thedimensions of the body.

Compared with the complete bell of a bucinafrom Zsámbék (Hungary) (1280 mm+) (Stiebel,2007: pl. II.3A, 2–3), and judging by the shape ofthe bell-end of the Ashkelon bucina, one mayestimate its overall length as c.1000–1100 mm. Asimilar instrument made of ivory, c.1570 mmlong, is known from the Boston Museum ofArt (Comotti, 1989: 73). The mouthpiece of theAshkelon bucina is utterly different from thedetachable mouthpieces associated with a cornudescribed by Stiebel (2007: II, R.3.1). Icono-graphic evidence indicates that trumpets wereused in various Greek and Roman public celebra-tions, amusements, spectacles, and ceremoniesboth religious and profane, as well as in variousmilitary uses. An early ceremonial image from

Figure 21. The mouthpiece of the copper-alloy trumpet.(E. Galili)

Figure 22. Bronze trumpet (top) (Maritime Museum Haifa, reproduced with permission); demonstration of the use of atrumpet using modern replica (bottom). (E. Galili)

E. GALILI ET AL.: A HELLENISTIC/EARLY ROMAN SHIPWRECK ASSEMBLAGE OFF ASHKELON, ISRAEL

© 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society 139

Hellenistic Maresha demonstrates a trumpetbeing used in a religious parade (Kloner, 1984:141). A much later trumpet used in a religiousceremony appears in a Roman mosaic fromPiazza Armerina, Sicily, from the 4th century AD(Stiebel, 2007: pl.II.3B.1). In Roman art itappears to be used to signal for military manoeu-vres during training and campaigns, as shown onTrajan’s column (Stiebel, 2007: pl.II.3B.2). InIsrael a trumpet in battle appears on a Romansarcophagus from Ashkelon from the 3rd centuryAD (Avi-Yonah, 1976, fig 17).

Finding this instrument in a shipwreck assem-blage may indicate that it could have had a nau-tical function. The Salamis naval battle (480 BC),one of the most important in history, was startedby a trumpet call according to Aeschylus (525–465 BC) in The Persians, staged within livingmemory of the battle. Thus already during the 5thcentury BC trumpets were used to convey mes-sages at sea. This practice may be behind a depic-tion of the goddess Nike standing on a ship’sprow and holding a trumpet during a celebrationof a victory appearing on a 3rd-century-BC Mace-donian drachm of Demetrios I Poliorcetes (Basch,1987: 341, 343, figs 727–8). In pre-modern andmodern ships, sounds made by various windinstruments (trumpets, fog-horns, whistles) andbells were and are used to convey messages likewarning or manoeuvring in poor visibility, fogor darkness (Tryckare, 1973: 59). The Ashkelonbucina could have had a nautical function, or mayhave been carried for trade.

DiscussionMaritime AshkelonAncient Ashkelon is situated on the coast of theJudean plain, a productive and rich agriculturalhinterland, and adjacent to the Via Maris, theroute along the coast of ‘Philistia’ connectingEgypt with points north. Ashkelon also served asa terminus of the ‘perfume road’, connecting theeast with the Mediterranean coast. This locationgave it commercial and strategic advantageswhich, combined with the geopolitical situation,brought about its development as an importanturban trading centre, beginning in the MiddleBronze Age (20th–17th century BC) up to theCrusader period (13th century AD). A diversity ofimported goods was recovered in archaeologicalexcavations at Tel Ashkelon, and there is histori-cal evidence for trade in oil, wine and otheragricultural products, which were distributed all

over the Mediterranean (Stager, 1993; Devor-jetski, 2001:121–7; Stager, 2008). This evidencepoints to widespread maritime ties between Ash-kelon and the large trading centres of the Medi-terranean basin. Port facilities were needed inorder to maintain such extensive maritime trade.

Evidence for the role of Ashkelon as an urbantrading centre, crossroads and port can be foundin historical records. In the ‘Pseudo-Scylax’ sea-man’s guide from the 4th century BC, whichdescribes ports, cities and prominent coastalpoints in Phoenicia and Israel, Ashkelon appearsas a coastal city, but a port is not mentioned(Stern, 1974: 19). William, archbishop of Tyre,who visited Ashkelon after its conquest by theCrusaders in 1153, wrote: ‘Ascalon derives noadvantage from being situated on the seacoast,for it offers no port or safe harbour for ships. Ithas a mere sandy beach and the violent windsmake the sea around the city exceedingly choppyso that, unless the sea be calm, those who comethere are very suspicious of it’ (Brundage, 1962,126–36). The Arab historians Ibn Shaddâd andAbu al-Fidâ, citing an earlier source, stated thatAshkelon did not have a harbour in which shipscould anchor (Sharon, 1995: 65).

Victor Guerin, who surveyed the ruins of Ash-kelon in 1854–1863, echoed William’s description:‘The pattern of the Ashkelon coast is not at allsuited for giving shelter to ships, therefore Ash-kelon never had a port or anchorage that couldprovide safe haven for ships, but only a dangeroussandy beach’. He suggested that the medieval portof Ashkelon was in the south-west part of the city,but noted that this area was open towards thewest and therefore not safe, and that it was ananchorage, not a built port, and in any case not aworthy anchorage (Ben-Amram, 1984, vol. 2:100,109–10). Despite intensive surveys and excava-tions carried out in Ashkelon, so far no remains ofa built-up port have ever been found. No histori-cal description suggests that the city had a port,while some categorically deny its existence. Thecoastline in the area is straight, sandy, and lacksbays and islets which could provide shelter forseagoing ships during winter storms, or duringstrong westerly and northerly summer winds. Thenearest temporary shelters for ships (and not inthe winter) are Tel Ridan anchorage, 40 km to thesouth (Raban and Galili, 1985) and Yavneh-Yamanchorage, 35 km to the north (Galili and Sharvit,1991; 1996). Thus, Ashkelon is situated at thecentre of a 75-km-long coastal strip that lackshavens for ships during storms, and certainly does

NAUTICAL ARCHAEOLOGY, 39.1

140 © 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society

not enable safe loading and unloading of goods ina stormy sea.

The morphology of the coastline and shallowcontinental shelf off Ashkelon does not providethe city with any advantage over other coastalsections between Gaza and Yavneh-Yam. Appar-ently, therefore, the reasons for choosing Ash-kelon as a base for merchant ships stemmedmainly from considerations such as connectionswith the agricultural hinterland, accessibility toinland trade-routes, and the geopolitical situa-tion, and not because of marine environmentalfactors. The fact that Ashkelon served as a port,actively in use from the Late Bronze Age (~1500BC) to Crusader times (1200 AD), indicates theimportance of these social, economic, and politi-cal reasons. Based on the morphology of theshallow continental shelf, the physical character-istics of the coast, and the underwater archaeo-logical remains, it seems that the town never had abuilt-up port. It seems that in antiquity, shipsarriving at Ashkelon anchored in the open sea,hundreds of metres offshore, using kurkar sand-banks and underwater rocks to ensure anchorhold (Galili and Sharvit, 1996; 1999a; Galili andSharvit, 2000; Galili et al., 2000; Galili, et al.,2001). The danger to shipping on this coastis demonstrated by the fate of the vessel underdiscussion.

Site-formation and post-depositional processesThe coast of Ashkelon is inhospitable, and stormshave trapped ships along the coast over the mil-lennia, crushing them in the breaker-zone.Judging by the distribution of artefacts on thesea-bottom, site-formation and post-depositionprocesses may be proposed. The wreckage-sitediscussed represents the remains of a local orforeign craft, probably driven there by a storm,while sailing along the coast, or anchoring somedistance offshore. During the wrecking event thevessel drifted ashore, grounded and broke up inthe surf-zone, at a depth that was then 1–2 m. Thedepth now is c.3–4 m because of recent sanderosion (Galili et al., 1988). Analysis of the com-position of scores of wrecks along the Israeli coasthas demonstrated that the contents of a shipwreckin the surf-zone are generally separated into threemain groups by the action of the sea. People andlivestock will drift ashore. Post-depositional pro-cesses on wreck-sites acted as an ‘extracting filter’,discriminating between light objects and heavy,usually metallic, artefacts. Light objects such asloose hull-timbers, rigging, and floatable cargo

disintegrated or drifted ashore to be salvaged.Heavy metallic or stone objects will sink into thesediment during the storm or soon thereafter,accumulating on the substratum under the sand.Clay amphoras will either break, drift ashore orroll on the shallow sea-bottom and graduallymove away from the wreck-site. Usually onlyfragments of amphoras are found and in mostcases they will be well worn by the surf. Therefore,only heavy objects will remain at wreck-sites inthe surf-zone. It thus seems that the cargo wascomposed mainly of materials which did notremain on site.

In addition, post-depositional processes,including fishing activity, may have added someintrusive artefacts to the site. In the last century,sand quarrying and the construction of break-waters and quays along the Israeli coast hasinterfered with the movement of unconsolidatedsediment, creating a shortage of sand and chang-ing the patterns of coastal sedimentation. Wideareas of sea-bottom have been uncovered, andhundreds of sites, including that presented here,have been exposed and discovered.

The shipThe nails and the anchors may provide someindication as to the size of the ship. Judging bythe thickness of the planks (4–5 cm) and the sizeof the iron anchors (c.1.5–2 m long), it seemsthat it was a medium-sized vessel. According tothe nails, we estimate that the ship was some15–25 m long. The pair of iron anchors, foundwith their detached stocks beside them, werereserve anchors, while the small 1-armed anchorcould have been used for fishing boats or life-boats, or as an auxiliary anchor. The ship wasprobably constructed by the ‘shell-first’ method,using mortises and tenons; ships constructed bythe skeleton-first method appeared later (Casson,1971:14–16, 201–14). No obvious remains ofballast were recovered. Possibly the ship usedlocal stones, kurkar or limestone pebbles, whichcannot be distinguished from indigenous stones.It is also possible that the cargo was heavyenough that the ship did not need an appreciableamount of ballast.

Ritual activityBeside the mundane use of drawing liquids foreveryday use, the presence of the set of bronzevessels may have a religious-cult explanation. Thesea was always considered as a gateway to newhorizons and a bridge between cultures. However,

E. GALILI ET AL.: A HELLENISTIC/EARLY ROMAN SHIPWRECK ASSEMBLAGE OFF ASHKELON, ISRAEL

© 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society 141

it can be dangerous, hostile and frightening.Ancient mariners used a variety of symbolic, apo-tropaic and cult means such as idols, inscriptions,and decorations to ensure safe sailing, good luck,safe anchoring and the avoidance of storms,disasters and evil spirits. For example, a relief(Casson, 1971: figs. 142, 146) shows a scene of aRoman merchant ship celebrating its safe returnto the port of Ostia, depicting a person (themaster?) standing in front of a table/altar onwhich a large bowl is placed; he is sacrificing andthanking the Gods for the safe return. The set ofbronze vessels (oil-lamp, shovel, ladles) couldhave functioned in such a shipboard ceremony.

Commercial activitySeveral finds, especially the set of weights and thebalance-scale parts, may indicate that the ship wasengaged in commercial activities. The two deco-rated ladles, the shovel, and the oil-lamp, whichare not typical mariner’s possessions, could becult items, but could also be the property of theshipowner or a wealthy passenger, or be tradegoods. The ship probably carried a main cargo oforganic materials or stored in pottery containerswhich drifted away from the wreckage-site asdescribed above.

Uses of the fishing gearThe fishing-gear finds could indicate that the shipwas engaged in fishing or carried a smaller fishing-boat. The fishing-net sinkers fit the characteristicsof cast-net, beach-seine, gill-net and trammel-netsinkers (Galili et al., 2002b). Fishing gear foundon a given underwater site may have derived fromfishing activities or from a wrecked ship thatdeposited the remnants of the gear it carried.Ancient Mediterranean shipwrecks often containfishing-gear sinkers and fish-hooks (Frost, 1991:355–410; Parker, 1992: 330, 356–40). Apparentlyin the past (as well as nowadays) watercraft of allsorts commonly carried fishing gear to augmentthe diet of crew and passengers and help thempass the time. Ancient or modern fishermen mayhave deposited remnants of fishing gear underwater, due to loss or to a shipwreck. Organic partsof the fishing gear decayed, only the metal andstone elements survived (Frost, 1991: 355–410).These remains may intermix with those of intru-sive gear. In the present case, there are remains ofother shipwrecks reported in the vicinity of thesite (Galili et al., 2001) and some of the fishinggear recovered, especially lead sinkers, may beintrusive.

Uses of the sounding-weightThe sounding-weight belongs to Oleson’s class4A (2000). It is an essential tool for fishermenand navigators, especially on dark nights orduring heavy fog (Galili et al., 2009). It can beused to identify both the depth and the nature ofthe sea-bed, by the material sticking to thetallow-cup. When approaching shore at nightor in low visibility the sounding-lead was anessential navigational aid. The weight of thesounding-lead (4300 gm) is identical to theweight of one of the balance-weights recoveredfrom this shipwreck, suggesting that it was delib-erately made to a standard weight. Type-1sounding-weights are generally attributed to alater period, between the 3rd and 7th centuryAD. Given the early date of the shipwreckassemblage, the presence of this artefact in thissite may be explained by its being either an earlyexample or a later intrusion.

The shipwright’s kitAmong the most important crewmen aboardseagoing vessels were the carpenters. Carpenter’stools such as right-angles, plumb-bobs, ham-mers, chisels, saws, and consumables such aspegs, treenails and metal nails, which must haveaccompanied these tools, have been recoveredfrom shipwrecks. An early group of shipwright’stools was recovered in Israel from a 4th-century-BC shipwreck (Stieglitz, 2006). Anothersuch kit from the Roman period was discoveredoff the Carmel coast (Galili et al., forthcoming).The copper-alloy nails from the Ashkelon wreck-site were divided into two sets: used nails, fromthe wrecked ship, and unused nails kept onboard for maintenance, as a part of the ship car-penter’s kit. Similar clusters of unused nails havebeen recovered from other shipwrecks, such asthe Lake Nemi wreck (Ucelli, 1950: 158). InIsrael, scores of unused nails were recoveredfrom a Hellenistic shipwreck in Apolloniaanchorage (Galili et al., 2009), and a Byzantineshipwreck south of Haifa, which contained addi-tional metal artefacts including a bronze right-angle, probably remains of a ship carpenter’s orshipwright’s kit (unpublished diving reports no.A-3817/2003/71/1). The lead fragments recoveredfrom the Ashkelon shipwreck could representadditional remnants of such kit. Whole rolledsheets of lead, as well as numerous fragments,many torn and twisted, have been recoveredfrom Israeli shipwrecks and harbours and dated,where possible, to the Hellenistic, Roman and

NAUTICAL ARCHAEOLOGY, 39.1

142 © 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society

late Roman periods (Kahanov, 1999; Rosen andGalili, 2007; Galili et al., forthcoming). Some ofthe smaller pieces, with nail holes regularlyarranged along the edges, were clearly patches.The torn lead fragments recovered in the presentcase could have come from hull sheathing, butthey could also have been scraps to be used aspatches.

The set of bronze weightsBalances and steelyards have been recoveredfrom several Mediterranean Greek, Roman andByzantine shipwrecks (Parker, 1992: 30, 107,205, 250, 323, 444). Obviously they were used onboard for a variety of purposes. These couldhave included the weighing, and thus pricing, oftransported foods, arranging cargo according toits weight, and weighing caught fish. Therefore itwas natural for a merchantman to carry abalance or steelyard aboard. The recoveredmetal parts of what could have been a woodenbalance, and the set of weights, were probablypart of a single kit intended for weighing mer-chandise. In order to identify the system towhich these weights belonged, they werearranged systematically according to theirmodern mass. Assuming that the previously-identified pairs of weights each had the sameintended mass as each other, there are six differ-ent mass units (Table 3). The internal relationsof the set indicate that five pieces belong to oneset, while one seems not to relate to the samesystem. Of the two best-known local and moreor less contemporary weight-systems in antiquityexamined, the Roman libra and the Greek AtticPtolemaic Hellenistic Greek (PHG) system, theRoman libra fitted the less well. In the PHGsystem there are at least three published stan-

dards (Lendering, 2008). In one, the ‘Solonian’system, the heaviest unit, the talent, was equal to60 minai (27.47 kg). In the ‘coin’ standard onetalent was equal to 60 minae (25.86 kg). In analternative system the talent was equal to21.45 kg. and consequently the mina was357.5 gm. The heaviest weight in the Ashkelonset seems to fit best one talent in the PHG coinstandard, which can be divided to 60 minae of431 gm. Comparing this system with our setshows that five of the six weights fit well. Onlyartefact no. 301 does not fit this system, but doesequal exactly ten minae in the PHG Soloniansystem. Both systems were in use during theHellenistic period (Lendering, 2008). Possibly anodd weight (e.g. 1 Mina) was originally part ofthe set. Such a missing odd weight would haveconsiderably expanded the range of weighingpossibilities.

Dating the shipwreckGiven that no coins or organic materialsenabling radiocarbon dating were recoveredfrom the site, one may try dating the wreckageby the typology of the artefacts using finds fromdatable archaeological sites. The bronze ladles,however, were in use for a long time (2ndcentury BC to 1st century AD). The oil-lamp istypical of the early Roman to Hellenistic periods(2nd century BC to 1st century AD). Ironanchors with two arms were in use from the 3rdcentury BC to the 3rd century AD. The balance-weights are of a Hellenistic standard used forcoins during the 1st and the 2nd centuries BC. Ittherefore seems that the ship was of Hellenisticor Early Roman origin, and was probablywrecked in the 1st or 2nd century BC.

Table 3. Dimensions and properties of the bronze weights, and various possible weight units

Weightno.

Weightin gm

Hellenistic Mina(coin system)

431 gm(accurate)

Hellenistic Minafor coins431 gm

(rounded)

Hellenistic Mina(Solonian system,

accurate)457.8 gm

Hellenistic Mina(Solonian system,

rounded)457.8 gm

HellenisticTallant

25.86 kg

Romanlibra

323 gm

300 25,900 60.301 60 56.771 56.8 1 79.969301 4,620 10.719 10.7 10.091 10 14.215302 4,367 10.132 10 9.539 9.5 1/6 13.437303304 2,643 6.132 6 5.773 5.8 1/10 8.132305306 1,714 3.977 4 3.743 3.7 1/15 5.274307 901 2.090 2 1.968 1.9 1/30 2.772

E. GALILI ET AL.: A HELLENISTIC/EARLY ROMAN SHIPWRECK ASSEMBLAGE OFF ASHKELON, ISRAEL

© 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society 143

AcknowledgmentsWe wish to thank the Israeli Antiquities Authority under which the surveys and excavations were conducted; C. Pulak for hisremarks concerning the anchors; Dr P. Martin for editing the manuscript; the diver Y. Ayalon, who discovered the site; thearchaeologist J. Sharvit who participated in the surveys; the divers D. Moskovitch, H. Sali, and A. Ya’aqobovitch; Ms S. BenYehuda, B. Galili, and R. Galili who helped in preparing the manuscript and the drawings; and photographers Ms T. Sagiv andJ. Galili.

ReferencesAbdou Daoud, D., 1998, Evidence for the production of bronze in Alexandria, Bulletin de Correspondence Hellenique Supple-

ment 33, 115–24.Aharoni, Y., 1975, Excavations at Tel Beer-Sheba, Tel Aviv 2, 147–68.Avi-Yonah, M., 1976, Recently discovered Sarcophagi from Ashkelon, ’Atiqot 11, 72–6, pl. XVII.Bailey, D. M., 1972, Greek and Roman Pottery Lamps (2nd edn). London.Bar-Nathan, R., 2002, The Pottery; Lamps, in Ashmonean and Herodian Palaces at Jericho, Final Report of the 1973–1987

Excavations, vol. III. Jerusalem.Basch, L., 1987, Le musée imaginaire de la marine antique. Athens.Ben-Amram, H. (transl.), 1984, Guerin V. Geographical Description of Erez’Israel, vol. 2 Judea. Jerusalem (Hebrew).Ben-Nachum, H., 2007, Tombs of the Roman Period and Building Remains of the Byzantine and Early Islamic Periods at Kafr

Kama, ’Atiqot 56, 105–18 (Hebrew).Biran, A., 1994, Biblical Dan. Jerusalem.Bridley, H. H., 1919, A Graffito of a Ship at Beit Jibrin, Palestine Exploration Fund Quarterly, April, 76–8.Brundage, J., 1962, The Crusades: a Documented History. Milwaukee WI.Bruneau, P., 1965, Exploration Archeologique de Delos, 26 Les lamps. Paris.Bockius, R., 2000, Ein Romischer Stockanker aus Trajans Donaukanal biem Eisernen Tor, Serbien, Archaologisches Korre-

spondenzblatt 30, 97–116.Casson, L., 1971, Ships and Seamanship in the Ancient World. Princeton.Comotti, G., 1989, Music in Greek and Roman Culture. Baltimore and London.Comstock, M. and Vermeule, C., 1971, Greek and Roman Bronzes in the Museum of Fine Arts. Boston.Devorjetski, E., 2001, The economy activity and special agricultural products of Ashkelon from the Hellenistic to the Byzantine

periods, in A. Sasson, Z. Safrai, and N. Sagiv (eds), Book of Ashkelon, 119–34. Ramat Gan (Hebrew).Flourentzos, P., 1978, A Group of bronze utensils from a private collection in Nicosia, Report of the Department of Antiquities

Cyprus 1978, 22–124.Frost, H., 1962, Submarine Archaeology and Mediterranean wreck formation, Mariners Mirror 48, 82–9.Frost, H., 1991, Anchors Sacred and Profane Ugarit-Ras Shamra. 1986: The Stone Anchors Revised and Compared (Ras

Shamar-OugaritVI). Paris.Galili, E. and Rosen, B., 2007, One Armed Anchors from Israel, Archaeologia Maritima Mediterranea 3, 99–114.Galili, E. and Sharvit, J., 1991, Yavneh-Yam Anchorage, Finds from the Underwater Survey, in M. Fisher (ed.), Yavneh-Yam

and it’s surroundings, 111–21. Jerusalem (Hebrew).Galili, E. and Sharvit, J., 1994, Classification of Underwater Archaeological Sites along the Mediterranean Coast of Israel:

Finds from Underwater and Coastal Archaeological research, in C. Angelova (ed.), Actes du Symposium International ThraciaPontica V, 269–96. Sozopol.

Galili, E. and Sharvit, J., 1996, Ashkelon North—Underwater and Coastal Survey, Hadashot Arkheologiyot 106, 155.Galili, E. and Sharvit, J., 1998, Ancient coastal installations and the tectonic stability of the Israeli coast in historical times, in

I. S. Stewart and C. Vita-Finzi (eds), Coastal Tectonics, Geological Society Special Pubn 146, 147–63. London.Galili, E. and Sharvit, J., 1999a, Underwater surveys off Tel Ashkelon—evidence for shifts in the coastline and the issue of the

location of the ancient port, in B. Galil and Y. Mart (eds), The Mediterranean Continental margin of Israel, Symposium, IsraelOceanographic & Limnological Research, The National Institute of Oceanography, 12–15. Haifa (Hebrew).

Galili, E. and Sharvit, J., 1999b, Underwater Survey in the Mediterranean sea 1992–1996, Surveys and Excavations in Israel 19,96–101.

Galili, E. and Sharvit, J. 2000, Tel Ashkelon, Hadashot Arkeologiyot 111, 111–14, Surveys and Excavations in Israel 111, 83*–85*(Hebrew/English).

Galili, E., Dahari, U., and Sharvit, J., 1993, Underwater Survey and Rescue Excavations off the Israeli Coast, IJNA 21, 61–77.Galili, E., Raban, A., and Sharvit, J., 2002a, Forty Years of Marine Archaeology in Israel, in H. Tzalas (ed.), Tropis VII,

Proceedings of 7th International Symposium on Ship Construction in Antiquity, 927–61. Pylos.Galili, E., Rosen, B., and Sharvit, J., 2002b, Fishing-gear sinkers recovered from an underwater wreckage site, off the Carmel

coast, Israel, IJNA 31.2, 182–201.Galili, E., Rosen, B., and Sharvit, J., forthcoming, Artefact Assemblage Recovered from a Roman Shipwreck off the Carmel

Coast, Israel, Atiqot.Galili, E., Rosen, B., and Zviely C., 2009, Ancient sounding-weights and navigation along the Mediterranean coast of Israel,

IJNA, DOI: 10.1111/j.1095–9270.2008.00218.x.Galili, E., Sharvit, J., and Dahari, U., 2000, Ashkelon, Underwater Survey, Hadashot Arkheologiyot 111, 110–11, Surveys and

Excavations in Israel 111, 82*–83* (Hebrew/English).

NAUTICAL ARCHAEOLOGY, 39.1

144 © 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society

Galili, E., Sharvit, J., and Dahari, U., 2001, Ashkelon and the sea in light of the coastal and underwater archaeological finds,in Z. Safrai, N. Sagiv, and A. Sasson (eds), Book of Ashkelon, 11–38. Ramat Gan (Israel) (Hebrew).

Galili, E., Syon, D., and Finkielsztejn, G., forthcoming, A hoard of Ptolemaic Bronze coins and weights from a wreck off AtlitIsrael, Atiqot.

Galili, E., Weinstein-Evron, M., and Ronen, A., 1988, Holocene Sea-Level Changes based on Submerged Archaeological Sitesoff the Northern Carmel Coast in Israel, Quaternary Research 29, 36–42.

Guhl, E. and Koner, W., 1994, The Romans—their life and customs. London.Haldane, D. D., 1984, The Wooden Anchor, unpublished MA dissertation, Texas A & M University.Hayes, J. W., 1984, Greek, Roman and Related Metal ware in the Royal Ontario Museum; A Catalogue. OntarioKahanov, Y., 1999, Some Aspects of Lead Sheathing in Ancient Ship Construction, in H. Tazalas (ed.), International Sympo-

sium on Ship Construction in Antiquity Tropis V, 219. Athens.Kapitän, G., 1971, Greco-Roman anchors and the evidence for the one-armed wooden anchor in Antiquity, in D. J. Blackman

(ed.), Marine Archaeology, 383–95. Bristol.Kapitän, G., 1984, Ancient Anchors—technology and classification. IJNA 13, 33–44.Karageorgious, V., 1987, Chronique des Fouliies et Decouvertes Archéologique a Chypre en 1986, Bulletin de Correspondence

Hellenique Supplement 2, 663–733.Kleiner, D. E. E., 1992, Roman Sculpture. London.Kloner, A., 1984, Mistarei Maresha, in I. Zaharoni (ed.), Derech Ertetz, 140–41. Tel-Aviv (Hebrew).Lendering, J., 2008, Livius web site, http://www.livius.org/w/weights/weights.htmlMazar, A., 1982, A Burial Cave on French Hill, Atiqot 8, 41–5.Meyers, E. M., 2006, The Ceramic Incense-shovels from Sepphoris: Another View, in A. M. Maeir and P. de Miroschedji (eds),

I Will Speak the Ridles of Ancient Times. Archaeological and Historical Studies in Honor of Amihai Mazar on the occasion ofhis Sixtieth Birthday vol.2, 865–78. Winona Lake IN.

Mitten, D. G., 1965, Two new bronze objects in the McDaniel Collection: an Etruscan strainer and a Roman incense-shovel,Harvard Studies in Classical Philology 69, 165–7.

Muckelroy, K., 1975, A systematic approach to the investigation of scattered wreck sites, IJNA 4, 173–90.Oleson, J. P., 2000, Ancient Sounding-Weights: a Contribution to the History of Mediterranean Navigation, Journal of Roman

Archaeology 13, 294–310.Parker, A. J., 1992, Ancient Shipwrecks of the Mediterranean and Roman Provinces. BAR Int. Ser. 580, Oxford.Pearson, A., 1992, Ancient Greece. London.Pritchard, L. B., 1988, Sarepta IV; The objects from Areas II,X. The University Museum of the University of Pennsylvania

Excavations at Sarafand. Beyrouth.Raban, A. and Galili, E., 1985, Recent maritime archaeological research in Israel—a Preliminary Report, IJNA 14, 321–56.Robinson, D. M., 1941, Excavations at Olynthus, Part X; Metal and Minor miscellaneous finds. Baltimore.Rosen, B. and Galili, E., 2007, Lead Use on Roman Ships and its Environmental Effect, IJNA 36, 300–307.Rosloff, J., 1991, A one-armed anchor of c. 400 BCE from the Ma’agan Michael vessel. A preliminary report, IJNA 20, 223–226.Sharon, M., 1995, A new Fatimid inscription from Ashkelon and its historical setting, Atiqot 26, 61–86.Siegelman, A., 1988, An Herodian Tomb near Tell Abu-Shusha, in B. Mazar (ed.), Geva; Archaeological discoveries at Tell

Abu-Shusha, Mishnar Ha-’Emeq, 11–42. Jerusalem.Speziale, G. C., 1931, The Roman Anchors found at Nemi, Mariners Mirror 17, 300–320.Stager, L., 1993, Ashkelon, The New Encyclopedia for Archaeological Excavations in Israel 1, 103–12.Stager, L., 2008, Ashkelon, The New Encyclopedia for Archaeological Excavations in Israel 5, 1577–86.Stern, E., 1980, Achaemenian Tombs from Shechem, Levant 12, 90–111.Stern, M., 1974, Greek and Latin Authors on Jews and Judaism, I–III. Jerusalem.Stiebel, G. D., 2007, Armis et litteris, The Military Equipment of Early Roman Palestine in Light of the Archaeological and

Historical Sources, unpublished PhD thesis, Hebrew University, Jerusalem.Stieglitz, R. R., 2006, Classical Greek Measures and the Builder’s instruments from the Ma’agan Mikhael Shipwreck, American

Journal of Archaeology 110, 195–203.Tryckare, T., 1973, The Lore of Ships. Gothenburg.Ucelli, G., 1950, Le Navi di Nemi. Rome.UK Hydrographic Office, 1976, Mediterranean pilot, vol. 5: coasts of Libya, Egypt, Israel, Lebanon and Syria, South coasts of

Greek islands from Kriti to Rodhos and Turkey with the island of Cyprus. Taunton.Ward, I. A. K., Larcombe, P., and Veth, P., 1999, A new process-based model for wreck-site formation, Journal of Archaeo-

logical Science 26, 561–70.Weinberg, S. S., 1971, Tel Anafa: The Hellenistic Town, Israel Exploration Journal 21, 86–109.Yadin, Y., 1963, Judean Desert Studies: The Finds from the Bar Kokhba Period in the Cave of Letters. Metal Utensils. Jerusalem.Zabehlicky-Scheffengger, S. and Schneider, G., 2005, Ephesian cooking vessels of the Augustan period, in M. B. Brice and E.

M. Vaag (eds), Trade Relations in the Eastern Mediterranean from Late Hellenistic Period to the Late Antiquity: The CeramicEvidence. 63–75. Copenhagen.

E. GALILI ET AL.: A HELLENISTIC/EARLY ROMAN SHIPWRECK ASSEMBLAGE OFF ASHKELON, ISRAEL

© 2009 The Authors. Journal Compilation © 2009 The Nautical Archaeology Society 145