‘Akko, Underwater Excavation

In 2009–2012, an excavation of ‘Akko’s southern seawall (Sharvit, Planer and Buxton 2013; Sharvit and Planer 2014; Fig. 2) uncovered previously unknown remains of the Hellenistic harbor quay, among other finds. The current excavation opened eight offshore probes (1–8; Fig. 3) near the Hellenistic quay, with the aim of identifying the side of the quay and determining its height and plan. Remains of the quay were found in Probes 1 and 4–8. The probes were dug with the aid of square metal excavation frames (2 × 2 m; Fig. 4), to prevent sand from sliding into the probes and ensure that the excavation remained undisturbed.

Probe 1. The probe was located 25 m south of the Ottoman wall of ‘Akko. In the beginning of the excavation, sandy soil (L200) was uncovered at 0.8 m bsl (below sea level). At 1.05 m bsl, a layer of stones, pebbles and loamy sand (L201) was revealed. The paving of the ancient quay (L202), composed of roughly dressed building stones (0.2 × 0.4 × 0.8 m), was found at 1.45 m bsl. The foundation of the quay was uncovered at 1.86 m bsl; it was built of small and medium-sized dressed kurkar stones (L203), similarly to the foundation discovered on land beneath the seawall (Sharvit and Planer 2014). Beneath this layer, at 2 m bsl, lay loamy soil rich in shells and pottery (L204). The natural bedrock (L205) was encountered at 2.57 m bsl.

Probe 2. The probe was located 40 m south of the Ottoman wall. A surface layer (L200) uncovered at 1.3 m bsl contained mostly modern refuse. At 1.8 m bsl, a layer of small stones, pebbles and non-local ballast stones (L201) was revealed. At 2.1 m bsl, a layer of loamy silt (L205) typical of the harbor contained abundant Byzantine pottery, including a CRS Form 7 bowl (Fig. 5:1) dating from the mid-seventh century CE (Hayes 1972:377–379, Fig. 81:1–7); a Late Roman Amphora 5 bag-shaped jar (Fig. 5:2; Reynolds 2005:573–575, Pl. 19: Fig. 140–148); a Late Roman Amphora 4 Gaza jar (Fig. 5:3) dating from the sixth–seventh centuries CE (Reynolds 2005:575, Pl. 20: Fig. 153:4); and an imported amphora (Fig. 5:4; Bouchenino 2010, Fig. 4:18). Most of the pottery is worn and partially coated with marine encrustation and shells. The bedrock (L206) was encountered beneath this layer, at 2.47 m bsl. Coral colonies had formed on the bedrock of the site. These are cushion corals belonging to the Cladocora species (Fig. 6), from an order of stony corals made up of polyps—small cylindrical columns that cluster together; they grow at a rate of 2–6 mm per year. The polyps of the corals found in the excavation were 8–12 cm long, and had therefore been growing for 30–60 years (Fishelson 1996). Radiocarbon analysis of the corals dates them to 335–94 BCE (Pietraszek 2018).

Probe 3. The probe was located c. 30 m south of the Ottoman wall. A surface layer of modern sand (L100) was uncovered at 0.8 m bsl. At a depth of 1.4 m bsl was a layer of dressed building stones (L101; stone dimensions 0.2 × 0.4 × 0.6 m), which were arranged in no particular order and were probably not part of the quay. After removing the stones, at 1.6 m bsl, a layer of sand, pebbles and broken shells (L102) was revealed. Beneath this layer, at 2 m bsl, was a layer of loamy silt (L103), rich in potsherds dating from the Roman–Byzantine periods. The bedrock (L104) was encountered at 2.4 m bsl.

Probe 4. A surface layer (L200) uncovered at 0.8 m bsl and composed of sand, small stones and modern refuse contained lead musket balls. At 1.2 m bsl, two large dressed kurkar quay stones were revealed (L201; 0.2 × 0.6 × 1.2 m). The stones were placed on the same level and abutted each other, similarly to the paving stones of the quay. The foundation of the quay (L202), built of large leveled kurkar stones, was discovered at 1.4–2.0 m bsl.

Probe 5. A surface layer (L300) uncovered at 0.8 m bsl was composed of sand and modern refuse. At a depth of 1.2 m bsl, flat kurkar quay stones were uncovered (L301; c. 0.5 × 0.6 × 1.2 m).

Probe 6. A surface layer (L400) uncovered at 0.8 m bsl was composed of sand and modern refuse. Beneath this layer, at 1.4–1.6 m bsl, was the quay’s paving (L401, L402), made up of rectangular dressed kurkar stones (c. 0.2 × 0.6 × 1.2 m). At the southern edge of the probe, beneath the quay’s paving and at 1.6–2.1 m bsl, lay the foundation of the quay (L403), which was built of roughly dressed kurkar stones (thickness c. 0.3 m).

 

Probe 7. A surface layer of sand and modern refuse (L500) was uncovered at 0.8 m bsl. At 1.4 m bsl was the edge of the quay (L501; Fig. 7), built like a retaining wall, with rectangular dressed kurkar stones (c. 0.4 × 0.6 × 1.2 m) arranged as headers and stretchers; it was preserved to a height of three courses. The upper part of the wall abutted the surface of the quay’s paving, described in Probe 6 (above). To the south of the edge of the quay, at 1.5 m bsl, was a layer of small and medium-sized stones (L502) containing potsherds, sand and shells that may have been collapsed rubble. The third course of the quay was revealed at a depth of 2.2. m bsl. It was built of headers, consisting of large rectangular kurkar stones (0.5 × 0.6 × 1.2 m) laid on top of the natural bedrock (L504). Harbor silt (L503) covering the bedrock was rich in shells and pottery, dated mostly to the Hellenistic period. The pottery included a bowl (Fig. 8:1) with a pointed out-turned rim and a conical body, dated to the second century BCE (Rotroff 1997: No. 730); a bowl (Fig. 8:2) with an outward-slanting rim and a rounded body, dated to the Hellenistic period (Handberg and Peterson 2010: Pl. 121:Dc-311); a globular cooking pot (Fig. 8:3) with an upright slightly out-turned rim, a straight neck and two handles between the rim and the shoulder, dated to the second century BCE (Berlin 2015:636; Fig. 6.1.8:7); an unguentarium (Fig. 8:4), with a thickened button base and horizontal grooves placed together in the center of the body, parallel examples of which are dated to the third and second centuries BCE (Anderson-Stojanović 1987:109; Saraçoğlu 2011:23, Fig. 3:U25); and a fragment of a mold-made oil lamp (Fig. 8:5) depicting the figure of Hercules, a parallel example of which is dated to the late first century CE (Rheinland-Pfalz collection 1970–1999).

Probe 8. A surface layer (L600) uncovered at 0.8 m bsl was composed of sand and modern refuse. The quay (L601) was revealed at 1.8 m bsl. It built of headers and stretchers, with three courses of rectangular dressed kurkar stones (c. 0.40 × 0.65 × 1.20 m). A layer of harbor silt at 2 m bsl contained potsherds dating from the Hellenistic–Byzantine periods; at this stage of the research, the pottery was left in situ. The natural bedrock was encountered at 2.4 m bsl.

As part of a monitoring program to check the condition of the seawall foundations, a previously undiscovered section of the quay’s paving (average stone dimensions 0.2 × 0.2 × 1.0 m; Fig. 1) was identified at a depth of 1.2 m bsl. This survey was conducted approximately three years after the completion of restoration work on the wall and the removal of the underwater rampart. It followed several severe storms that had washed away some of the sand and sediment which remained after the rampart was dismantled from the seabed.

During the three underwater excavation seasons, the quay of ‘Akko’s Hellenistic harbor was uncovered for a length of 7.4 m. Probes 7 and 8 revealed the southern extremity of the quay. The quay’s stones were apparently laid on a course of roughly dressed kurkar stones, which were placed on a layer of loamy silt containing pottery. The side of the quay faced the open sea and was therefore built as a massive retaining wall, using headers and stretchers in the Phoenician style of construction to prevent the stones from moving. The foundation course of the quay wall was laid directly on top of the natural bedrock. Probe 4 revealed the quay at a depth of 1.2 m bsl, whereas in Probe 4 it was 1.4 m bsl; the slope of the quay between the two probes is c. 2.7%.

By identifying the side of the quay, it was possible to measure its width as far as the southern seawall (at least 25 m). Since the Ottoman wall was built on top of the quay, it was impossible to measure its entire original width. The harbor silt with its abundant Hellenistic–Roman pottery, found in excavations of the southern seawall (Constantine 2019) and in the probes, shows that this is the largest and most impressive Hellenistic harbor so far discovered along Israel’s coastline, and one of the largest in the ancient world. During the Hellenistic and Roman periods, ‘Akko’s harbor was designed as a port for naval and maritime trading fleets. The absence of archaeological layers from later periods above the quay shows that the harbor quays were abandoned, and shipping was probably moved to the southwest, to the area of the Island of Flies and the current location of ‘Akko’s harbor. The colony of cushion corals found in Probe 2 shows that the area was free of sediment prior to 94 BCE and the water was clear. After this, some extreme event occurred that quickly decimated and buried the corals, possibly an alteration in the river course of Nahal Na‘aman. This event may have also been the reason behind the harbor’s relocation.