Mersa/Wadi Gawasis, Mission 2007-2008 - Kathryn A. Bard and Rodolfo Fattovich et alii

Archaeology - K. A. Bard, C. Calcagno, R. Fattovich, I. Incordino, E. Mahfouz, T. Spurrier and C. Zazzaro

In 2007-08 the western and southern slopes of the fossil coral terrace continued to be investigated (Figure 14). Samples of cedar ship timbers were also examined to provide evidence for a dendrochronological dating of the site. [3]

Western Slope - K. A. Bard, C. Calcagno, E. Mahfouz, T. Spurrier and C. Zazzaro

In 2007-08 excavations were conducted on the top of the slope of the fossil coral terrace, in front of the entrance to Cave 6 (WG 32) where some large pieces of timber from seafaring ships were found and carefully studied in situ, because their removal could badly damage the wood. These timbers provided nautical archaeologists with important information about the construction technique of ancient pharaonic ships. Excavations outside Cave 6 also provided more fragments of clay sealings with the imprint of seals dating to the late 12th Dynasty.
Excavations were also resumed to the north of entry to Cave 3 (WG 33) where the remains of a mud-brick structure, hearths, inscribed stelae and exotic ceramics were found. A votive stela dating to year 21 or year 31 of the reign of Amenemhat the III was found there.
Two new excavation units were opened on the top of the slope, to the south of the entrance of Cave 6 (WG 55 and WG 56). In this area the opening to another cave (Cave 7) was found and partially excavated. A large quantity of wood debris from ship timbers and lithic artifacts were found in front of this entry (WG 55), suggesting the occurrence of an activity area there. A stelae with an offering inscription to Wsir Wedj-wer (Osiris of the sea) was excavated there.
To the south of the entrance to Cave 7, another opening to a rock-cut cavity was found, but was not completely excavated due to safety concerns. In front of this opening was an arrangement of three upright conglomerate stone slabs and a rock-cut wall (along the vertical wall of the fossil coral terrace), suggesting a ceremonial use of this area.

WG 33[4]
Excavations in WG 33 began on 2 January, 2008. This was where Chen Sian Lim and Giuseppe Lebro had excavated an area (7 m x 3 m) in 2006-07 (see Fattovich and Bard 2007). The principal discovery in 2006-07 in SU1 was a limestone stela of Senusret III, broken into two fragments and in a mediocre state of preservation. Broken niches for stelae were also recorded in the wall of the fossil coral terrace there, and an unfinished anchor was found in SU2.
The objective of excavations in WG 33 in 2008 was to open as much as possible the surface of this area in order to better understand the organization of this sector in front of the entrance to Caves 3-4. Excavations were conducted in a 7 m x 7 m unit. After removing the surface fill of the 2006-07 excavations, a small limestone stela (Stela 16, 16.0 cm x 14.0 cm x 5.0 cm) was found. The stela is in Abydos style and dates to year 31(?) of the reign of Amenenhat III; it belongs to the Guard of the Great Palace, Ameny.
SU1 consisted of a mix of sand, small pieces of wood, salt, pebbles, and fragments of limestone. At the base of SU1, a very damaged stela (Stela 20, 41.0 cm x 23.5 cm x 7.0 cm), covered with a thick crust of salt, and the lower part of another stela (Stela 17, 21.0 cm x 19 cm x 3.0 cm) were found.
After the removal of modern debris, a stratum (SU2) of sand mixed with pebbles, limestone fragments, and pieces of wood and potsherds was recorded. In this area were three hearths (Hearths 1, 2 and 3, from south to north; Figure 15). Hearth 1 was found in the southeastern sector of the unit and contained charcoal, pieces of burnt wood, and potsherds. Hearth 2, the largest of the three hearths, was roughly circular in shape and contained charcoal, pieces of wood from boxes, and sherds of imported pottery. Hearth 3 contained the remains of an ancient burnt box, pieces of wood, potsherds, charcoal and carbonized cereal. Between Hearth 3 and the wall of the coral terrace, a rectangular ship timber was found.
A layer of mud-brick (SU3) was encountered and WG 33 was enlarged to 10 m x 7 m. There was no evidence of a wall structure or foundations for a wall and the mud-brick layer appears to be some kind of (thin) platform construction, only partially preserved. The mud-bricks, which were locally made and tempered with sand, seem to have been laid in one course on top of a surface prepared with a thin layer of (mangrove) leaves and sometimes twigs of wood (mangrove?), below which is a stratum of windblown sand. In the northwest sector of WG 33, beneath the mud-brick construction, was a large sherd of a typical Middle Kingdom storage jar.
The layer of windblown sand (SU4) on which the mud-brick platform was constructed post-dated (and covered) the entrances to Caves 3-4. The purpose of the mud-brick platform is unknown.

WG 32/WG 53 [5]
Excavation unit WG 32 was first opened to the south of the entrance to Cave 2 in 2005–06. Investigations were resumed in a 6 m x 12 m area of the excavation unit in 2006-07. Relevant findings from this area included inscribed stelae, clay sealings, two ship timbers, ca. 40 complete or fragmentary wooden boxes and two boxes with a painted inscription, potsherds, and ca. 50 shallow bowls (complete or fragmentary). The entrance of a man-made cave (Cave 6) was also discovered in the fossil coral terrace, in the southern part of the excavation unit (see Bard and Fattovich 2007: 60-61).
Excavation of WG 32 was resumed in 2007-08 in order to investigate the earlier occupation levels, in an area 6 m x 4 m.
A new excavation unit, WG 53, was also opened southeast of WG 32 in order to investigate the extension of the different occupation levels toward the slope. Investigation in WG 53 was conducted in a 4 m x 4 m area.
The surface (SU1) in WG 53 consisted of a thin stratum characterized by windblown sand, a few potsherds and fossil coral or conglomerate rock fallen from the terrace. The bottom of this stratum included several scattered wooden box fragments and potsherds. The strata below the surface, in WG 53, were consistent with strata excavated during the 2006-07 excavation in WG 32 (Fattovich and Bard 2007: 18-19), although the quantity of materials and occupation evidence was reduced. This is typical of the strata at Mersa/Wadi Gawasis, where the archaeological deposits are more concentrated in the slope area near the terrace wall and at the cave entrances.
Noteworthy is a painted rimsherd dating to the Second Intermediate Period – Early New Kingdom (see Wallace-Jones, this report), recovered at the eastern end of the WG 53 unit on top of a stratum of windblown sand (SU10) which covered the wooden boxes in WG 32.
Excavation of the floor on which the wooden boxes lay continued in WG 32-53. A hearth (SU39) arranged within a large sherd from a jar was found on top of a stratum of windblown sand (SU25) dating to the end of the 12th Dynasty. This hearth was associated with other hearths excavated in 2006-07 (see Fattovich and Bard 2007: 18-19).
The lower layers identified in 2006-07 included a thick stratum of salt concretion (SU38) and a concentration of organic materials (SU33). During the 2007-08 excavations, the molded shapes of several large ship timbers, covering an area ca. 2.30 m x 1 m, were identified beneath the salt concretion of SU38, at the entrance to Cave 6, in squares B5 and C5. These included a steering oar blade (T72), similar in form and wood to the two blades recovered in 2004-05, but larger (see Calcagno and Zazzaro, Ship Components in this report; and Bard and Fattovich 2007: 150-153). Excavation was discontinued in this area until the timber conditions could be further assessed.
Excavation continued in squares A4, B4 and C4, in an almost sterile sand stratum (SU40); a few charcoal fragments, two sherds and salt crystals, were excavated. This stratum (SU40) covered an irregular stratum (SU42), ca. 10-40 cm thick, containing scattered cobbles and one badly preserved wood fragment, possibly a tenon.
A concentration of mud was recorded to the west of the Cave 6 entrance, covering a natural stratum of cobbles (SU41), 1.5 m thick. SU41 can be compared to strata of cobbles in Cave 2 and between Cave 2 and Cave 3 (see Bard and Fattovich 2007b: 63). It is also similar to the curved wall of the "ceremonial" structure found in WG 56: a natural stratum partially shaped and consolidated with mud plaster for structural purposes.
In WG 53, stratigraphic unit SU25 covered a stratum of cobbles (SU43) without cultural materials. This stratum seems to be associated with the timber accumulation in WG 32.
A test pit, 1 m x 1 m and ca. 1.70 m deep, was made in WG 32, A4, in order to investigate if earlier occupation phases existed. Two different strata were identified: a sandy stratum (SU44), ca. 85 cm thick, covering a more compacted stratum (SU45), ca. 85 cm thick, consisting of mud and pebbles.
Copper alloy artifacts excavated in WG 32 included a pin (ca. 8.7 cm in length) found in SU4 and what appears to be a blade fragment (ca. 8.2 cm x 0.2 cm x 0.2 cm) in SU25.
Excavation in WG 32 conducted during this field season allowed researchers to better investigate the three different occupation phases already identified by Andrea Manzo in 2006-07 (see Fattovich and Bard 2007: 18-19).
Below the first occupation level in WG 32/WG 53, excavation also revealed evidence of a lower layer of cobbles (SU42) similar to the floor of cobbles from one of the earlier occupation phases at the entrances of Cave 2 and 3 (units WG 24 and WG 16; see Bard and Fattovich 2007: 54-57, Figures 23 and 24). However, unlike the living floor found in the area outside Caves 2 and 3, there is no clear evidence of human occupation associated with the cobble layer in WG 32.
Consequently, at present, the first level of use attested in this area consists of a floor on which was found a concentration of organic materials and ship timbers which had accumulated at the entrance of Cave 6 and were associated with the two walls giving access to the cave. The westernmost wall, made with mud and cobbles, was probably built in order to separate Cave 5 from Cave 6.
The floor on which the boxes and hearths were found represents a successive phase dating to the reign of Amenemhat IV (ca. 1797-1790 BC), based on texts inscribed on two boxes (see Fattovich and Bard 2007: 47-48).
The later period of abandonment at the site during which windblown sand and leaves accumulated may date to the early New Kingdom, on the basis of the associated potsherd found in WG 53. This phase is characterized by very little archaeological evidence.
In summary, WG 32 has revealed only three phases of use, which implies a less intense level of human occupation compared to the sequence of seven phases identified in the area outside Caves 2 and 3. It suggests that Cave 6 was possibly carved and used later than the other caves. Further investigations outside Cave 5, to the west of Cave 6, will provide more information on the stratigraphic connections between the different areas excavated on the slope.

WG 55 [6]
WG 55 is located on the western slope of the fossil coral terrace south of/around the corner from the entrance to Cave 6. The unit was oriented to the northeast in order to follow the natural orientation of the terrace wall. Excavations consisted of squares C1, C2, C3, D1, D2, D3 and E1, E2, E3 (E squares excavated by C. Calcagno). Squares A1, A2, A3 and B1, B2, B3 were not excavated because they consist of a large build-up of sand which may be supporting a weakened area of the coral terrace between Caves 6 and 7. The excavation area of C2-C3-D2-D3 was 4 m x 4 m in area; squares C1-C2 were just under 2 m x 2 m because they were shortened by the presence of the terrace wall (see plan at end of report.

WG 55, C-D 1-2-3
Excavation unit WG 55 is located on the western slope of the fossil coral terrace south of the entrance to Cave 6. The major reason for beginning a unit in this area was because the edge of a very large stone with worked edges and small, carved square recesses was visible on the surface. SU1, which consisted primarily of eolian sand, covered the large stone, which was rounded on one edge and straight on the other. It has been proposed that the square recesses are quarry marks. At the end of the field season, once Cave 7 had been discovered, it appeared that this stone had been standing on end in front of the entrance, but had fallen over at some point. Along the bottom edges of the stone, mixed in with the sediments, was modern debris such as fragments of tea package labels, newspaper, plastic, styrofoam, etc. It is unknown how recent this debris is (5 or 50 years?), but this evidence indicates that the sediments and sand have moved around considerably in recent times. There was no debris directly under the rock, only along the sides indicating that it fell some time ago, but was buried more recently in sand. SU1 consisted of a layer of sand approximately 1 m deep. Corresponding to the bottom of the rock, at the end of SU1 and the beginning of SU2, there was a layer of sea grass which extended to the coral terrace wall.
Near the bottom of SU1 and once the large stone was fully exposed, a large hole was visible in the northwestern corner of square C1. At this point, it was unclear whether this was man-made, such as a cave entrance, or a continuation of a large fissure along the side of the coral terrace. For the time being, excavations in this corner were discontinued in case the sand was supporting the coral terrace. The hole was filled with bricks for support until further investigations could be done.
At the bottom of SU1 and the top of SU2, there was some sea grass and a red residue indicative of burning along the coral terrace wall. Also, the terrace wall began to be indented at this point, as if forming an alcove. In square C2/3, there were the remains of a hearth. It was decided to continue excavating in SU2 in squares D1-2 along the terrace wall where the burning, sea grass, and indentation were, to investigate whether there might be a cave entrance below. (In previous field seasons, evidence of burning and sea grass were found above the entrances to the other caves.) Along the terrace wall, there were clods of mud (it was uncertain if they were originally mud-bricks or natural deposits picked up in the wadi), which may have been used to seal or mark a cave entrance. SU2 was filled with wood debris, twigs, sticks and leaves, as well as an intact corner of a reed mat and a very well preserved piece of heavy cloth. Also excavated in SU2 were a large piece of folded linen and two hearths.
At the bottom of SU2, we decided to continue excavating down into SU3 in a smaller area of square D3 to better understand the stratigraphy and to see if there was a cave entrance. SU3 consisted of large cobbles in sand. There were occasional salt deposits throughout the top of SU3 on the cobbles. The cobble layer appears to slope downwards towards the wadi. By examining the baulk, the cobble layer appears to continue into the other squares of WG 55 and it corresponds to the cobble layer in WG 56. The cobbles are also present in the coral terrace wall (also note the sea grass and red burning from the bottom of SU1/top of SU2, which are visible at the top of the photo). Underneath the cobble layer was sterile sand with no cultural remains (SU4).
It was suspected that the cobbles in the terrace wall were actually man modified and built at the entrance to a cave because they resembled the modified and shaped cobble walls between Caves 5 and 6, and Caves 2 and 3. Also, there were mud deposits which could originally have been mud-bricks used to seal an entrance. Unfortunately, after much work with a pick-axe and hammer, the cobbles remained in place: they were a natural part of the coral terrace deposited geologically at some point long ago. There was no cave entrance. Continuing on, it was decided to excavate SU2 in the neighboring D3 square so that all units in WG 55 could be at the same stratigraphic level. In D3 SU2, the composition of the sediments was the same as in D1 SU2, thus the same SU number was used. Interesting finds in this square were a piece of an ostracon and a Middle Nubian rim sherd.
After finishing SU2 in D3 and reaching the top of the cobble layer (SU3), which continued from square D2, it was decided to move to C1 and begin systematically leveling the C 1-2-3 squares down to the cobble layer (SU3) as well. Located in C1 was the above mentioned hole (that was filled with bricks). Once C1 had been cleared of sand that had fallen in from the terrace slope, it was decided to investigate a small section of the top right corner because the hole appeared to be square, which clearly indicates that it was man-made, or at least altered, at some point. A limestone stela (Stela 23) was discovered in an upright position at the top of the hole. (For the inscription and more details, see Mahfouz, this report.) The stela had a layer of salt encrustation and its bottom right corner was resting on top of a large potsherd. The stela measured approximately 34 cm x 27 cm x 9 cm.
Upon the removal of Stela 23, it was revealed that its bottom left corner was sitting near another limestone stela, (Stela 24) which was lying horizontally in the cave entrance. Stela 24 measured approximately 52 cm x 33 cm x 10 cm and was quite large in comparison to Stela 23. The inscribed side was facing upward and was encrusted in a layer of salt.
The two stelae were discovered in SU8, which was medium grain loose sand that appears to have been deposited naturally once the cave was in a period if disuse. Also in this layer were many clay sealings (some with very clear details), pieces of inscribed boxes and ostraca. It is unclear how the stelae came to be in their positions and whether it was purposeful, or if they had fallen from somewhere else. At one time there may have been stelae niches carved in the coral terrace wall directly above the cave, but this wall had deteriorated over time no niches are now visible.
Because of the finding of important artifacts such as the stelae and clay sealings, it was decided to continue excavating in square C1 and to investigate approximately 1 m deep into the cave entrance, but not to excavate the full opening because of the fragile nature of the slope. The cave entrance (approximately 1.5-2 m long) was not fully opened because of the fragile nature of other man-made caves found in previous field seasons. All of the man-made caves need the attention of a specialist to assess the structural integrity before excavating their interiors.
Within the entrance to the cave, there was a large, salt encrusted timber in very deteriorated condition that continued into the unexcavated area beyond the entrance. It was so encrusted with salt that it first appeared to be a rock: the surface was very hard and uneven. This timber (T57) was left in situ because of its fragile nature and unknown length. The position of the timber would indicate that it fell from its intended position, or was thrown into the cave after a period of disuse because it was deposited in a layer of compact sand, and not in any particular position. In the past, timbers have been discovered parallel to cave walls as if neatly stacked there at the end of an expedition. It will not be possible to completely assess the position of T57 and its state of preservation, however, until it can be fully excavated in a future field season.
Approximately 130 cm from the top of the cave entrance in SU10 there was an almost intact small jar. The bottom of the jar was missing and the entire vessel was encrusted with salt. It measures approximately 15 cm tall and 8 cm wide at the shoulder. Directly under the small jar, was a possibly well preserved timber (T74). At this time, the stratigraphy changed to SU11 because the timber could have been used as a threshold for the cave and may indicate a living floor. (Note: SU11 in C1 appears to correspond to SU2 in the other squares.) This timber was lying parallel to the entrance along the bottom. It was left unexcavated and was covered to preserve its environment until a conservation plan can be implemented. At this time, the ceiling of the cave entryway began to crumble and excavations inside the cave were discontinued.
To the right side of the cave entrance large rocks of conglomerate stone had probably been placed together and may represent a small wall or border of some sort that has shifted somewhat over time. In Square C1, SU11 and continuing into C2, SU2 outside the cave entrance and perpendicular to it, was a large, well preserved timber (T75). It had a layer of salt and a bright green square of copper residue. It measures approximately 93 cm x 19 cm x 6 cm. It appears that the timber fell from another location as it was not in any perpendicular or parallel position to the cave entrance. T75 was found sloping downward toward the cave entrance, which is not the position one would expect if it had been used for a type of ramp (sloping downward toward the wadi). This will be better understood through excavation in a future field season.
Once the excavations of C1 were finished (down to the top of the cobble layer), it was decided to continue excavating in C2. Upon examining the next strata of C2, it was determined to be the same as SU2 in the D squares and the composition of the sediments was the same. In C1, there are many more stratigraphic units because the sediments fell differently into the cave entrance and directly outside of it. Associated with the end of T75 in C2 were a small wooden finial type of artifact and a wooden nail artifact approximately 15 cm in length. In C2, SU2 under the timber were the remains of a hearth, Feature 3 (SU12), approximately 35 cm x 30 cm, and a concentration of burnt seeds. This may indicate that there was more than one period of occupation in SU2 if there was a hearth, and later a timber placed/fallen on top, though this is not clearly visible by looking at the composition of the sediments alone. It is not possible to determine how many occupation periods there may have been because they were probably temporally close and the stratigraphy is not minute enough in this area.
Also of note, sea turtle bones were found in the strata of C2, SU2, stuck between the coral terrace cave entrance wall and conglomerate rocks. A few pieces had the preserved remains of skin on them (Figure 16). In C2, SU2 there was a small piece of papyrus (without writing) as well as exotic potsherds and a tiny Minoan sherd.
At this time, the excavation moved to square C3, SU3, which was done under the supervision of C. Zazzaro. C3, SU2 consisted of the same composition of sediments as SU2 in the other squares and a hearth (Hearth 4) was uncovered, Feature 4, SU13. The hearth had a large jar fragment on top of it.
The area outside of Cave 7 was a workshop area. The wood debris there is indicative of re-working ship timbers. The many clay sealings and wooden box pieces also indicate that this area was used for unpacking and packing goods, either upon first arrival to Mersa/Wadi Gawasis, or in re-packing the goods from Punt for transit to the Nile Valley. This is also supported by the fact that Cave 7 is near the work area in front of Cave 6, where the majority of clay sealings and wooden boxes were found in past field seasons. Even though WG 55, SU2 is not a very thick stratum, it appears to have multiple occupation periods, which are evident by multiple hearths, clay sealings, timbers and an abundance of wood debris. Different occupations are very difficult to understand by the strata alone because there is little to no change throughout the composition of the sediments, but the features and artifacts are easier to interpret.

WG 56 and WG 55 E1-2-3 [7]
Excavation unit WG 56 was opened east of WG 55 in order to investigate the rock shelter area of occupation all along the coral terrace wall.
The unit was oriented to the northeast in order to follow the natural orientation of the terrace wall. The excavated area, measuring 6 m x 6 m, was originally covered by a large deposit of sand (SU1) that sloped from the terrace wall toward the wadi bed with a height difference of ca. 2.30 m.
A stone structure composed of four boulders, associated with an opening carved into the wall terrace, and with a large curved wall cut into a natural stratum of cobbles, was discovered in WG 55-56 and was partially investigated (see below, Figure 17) [8].
The uppermost stratum (SU1) was composed of windblown sand and small pebbles coming from the deterioration of the conglomerate stone stratum of the terrace wall. This naturally accumulated deposit, ca. 40-20 cm in thickness, included sherds with a very eroded surface, suggesting they had been exposed for a long time on the upper terrace surface, and fell down together with the deterioration and collapse of parts of the terrace (see below, description of SU6). A few centimeters below the surface in the northeastern part of the unit, several modern hearths were found that had been made recently by workers to boil water. The westernmost area comprising WG 55 E1-2-3 also yielded significant modern rubbish and cloth debris. An inscribed stela (Stela 28, see Mahfouz, Epigraphy, this report) was discovered at the base of this stratum, in the southern limit of the unit.
The sand deposit (SU1) covered two large slabs of conglomerate stone (SU6) that had collapsed at one time from the coral terrace, and a structure composed of four large stone blocks (SU12, SU13, SU14, SU16). SU1 also completely covered the stratigraphic sequence of the terrace wall, including a stratum of consolidated sand (SU5) and a stratum of cobbles (SU4). The consolidated sand (SU5) is a natural stratum of wadi sediment cemented with carbonate (ca. 1.30 m thick), into which a man-made opening (SU15) was carved. This stratum was possibly intentionally modified all along the exposed surface.
The alcove-like opening in the terrace wall has two large horizontal fractures above it, and its ceiling lacks structural integrity. For this reason, excavations were conducted only in part of the sand deposit filling, in which two different accumulation levels were distinguished. The upper level (SU17) is ca. 1.0 m thick and consists of windblown sand that had fallen in through the crack at the top of the opening. The lower level (SU7) consists of windblown sand and sea grass. A small limestone slab (20 cm x 21 cm x 4.5 cm), possibly with an eroded inscription, was found at the top of SU7. A strip of painted linen (20 cm x 3 cm) with a red mark and hieratic inscription was recovered from the middle of this stratum. At the bottom of SU7 were clay sealings with seal impressions, a few fragments of wood, and, close to the stone structure, some large sherds of jars, shells, fish bones and some pieces of mud.
The opening was delimited southwest by the stone structure, which consists of three erected conglomerate stones arranged in a U-shape and a fourth conglomerate stone leaning against the southernmost stone. Another conglomerate stone seems to lie at the base of the structure and will be investigated in the next field season. The three principal blocks forming the structure are almost rectangular in shape and feature several small rectangular recesses on their faces or in the corners between two different faces [9]. The leaning stone has an irregular pentagonal shape and measures 67 cm x 63 cm x 22 cm. The stone structure was constructed by cutting into the natural cobble stratum (SU4). The two lateral stones are consolidated and supported with small coral blocks (ca. 10-15 cm in diameter) stacked in a pile on either side and mixed with some mud blobs and organic material. The remains of a wooden pole was found inserted in the southwestern consolidation mound, between the stone structure and the "cobble wall" (SU4); another pole was found in the corner between two stones.
The two large collapsed slabs of conglomerate stone (SU6) lying in front of this opening were broken up and removed, in order to continue excavation in the southwestern portion of the trench to better investigate the wall cut into the cobble stratum (SU4) connected to the stone structure and to the opening. The natural cobble stratum is part of the fossil coral terrace stratigraphic sequence and consists of cemented conglomerate. It was also found in WG 32, in WG 55, and in Cave 2, and had been artificially modified in several places. In WG 56 the cobble stratum was shaped into a large curved wall ca. 50 cm thick, which had been partially consolidated with mud plaster. Outside the wall structure was evidence related to construction of the structure, as well as evidence of human occupation.
A stratum (SU8) of mud and organic material (leaves, branches, rope, textile fragments and Papyrus 2) extending into WG 56, in the area around the external part of the structure, most likely represents deposits accumulated next to the retaining wall. This stratum corresponds in WG 55 E2-E3 to SU4; a fragment of an inscribed papyrus (Papyrus 1) and a Middle Nubian potsherd (see Wallace-Jones, Pottery in this report) were also found in this stratum. Thickness of these two strata is very irregular, varying from 10 cm to 30 cm. The windblown sand stratum, SU1, and the two large collapsed blocks (SU6), originally covered this stratum. A particularly thick, compact concentration of mangrove leaves (WG 55 SU 6; ≤ 12 cm), extended along part of the terrace wall within WG 55 E 1-2, continuing westward into WG 55C-D.
Below SU8 a stratum of compacted sand (SU11) associated with three hearths and with a great concentration of organic materials (wood fragments, ropes, bones) and potsherds (including dishes, bowls and small bottles), was found at the base level of the "cobble wall." A well preserved ceramic jar, missing only its neck, was found on top of this stratum and was associated with the stone structure. SU11 was not completely excavated.
Three hearths associated with this stratum were found in WG 56 near the "cobble wall." Hearth 1 (SU9), ca. 40 cm x 40 cm, was located close to the southern limit of the excavation unit. Hearth 2 (SU10), ca. 40 cm x 30 cm, was located close to the wall structure. Hearth 3 (SU2), ca. 85 cm x 40 cm and 10 cm deep, included one large, burnt potsherd and a flat granite stone. It was located at the eastern limit of the unit and the "cobble wall."
A test pit, 70 cm x 50 cm and 40 cm deep, was excavated at the eastern end of the unit in order to investigate other possible occupation phases. A layer of organic material (SU3) was found that seems to be related to SU11. No earlier occupation phases were identified.
The archaeological deposits associated with the stone structure discovered in WG 55-56 seems to be distinct from deposits that are usually excavated on the terrace slope, which typically include evidence of woodworking activities and/or remains of accumulated materials connected to the expeditions. The stone structure and the related mound delimited by the "cobble wall" can be compared to features excavated on the top of the terrace and interpreted as ceremonial structures (Bard and Fattovich 2007: 244-245). The associated types of ceramics (dishes, bowls and small bottles); the possibly votive, large jar; and the hearths found all along the structure enclosure suggest ritual activities performed in this area. The evidence of papyrus fragments also points to the administrative-ceremonial importance of this area.


Southern Slope - K. A. Bard, R. Fattovich, I. Incordino and T. Spurrier

In 2007-08 excavation in the so-called "harbor area" provided firm evidence of the use of this area in the Middle Kingdom. Two main phases of use of this area in the 12th Dynasty were identified. In the earlier phase, part of the area was used as storage for large jars. In the later phase most of the area was used as a camp with clear evidence of large hearths and many fish remains.
The following excavation units were delimited and partially excavated between the fosil coral terrace and the present bed of the wadi in the southwestern sector of Wadi Gawasis (the so-called "Harbor Area") in December 2007–January 2008: WG 45/46/47/48/49/50; WG 51; WG 52; WG 54; WG 57. Excavation was conducted by R. Fattovich, with the assistance of K. A. Bard (WG 51; WG 52), T. Spurrier (WG 54), and I. Incordino (WG 45/46/54/57).

WG 45/46/47/48/49/50
These excavation units occupied an area of 600 sq. m. at the base of the coral terrace and included the geological test pits T8-A1 and T8-A2/2006-07 (see Fattovich and Bard 2007). The excavation was aimed at 1) testing the hypothesis that the harbor was located in this area, and 2) identifying possible ancient shorelines (see Bard and Fattovich 2007: 77-90).
The surface in the investigated area sloped gently down from north to south, with a difference in level of about 0.9 m between the northern edge of WG 45/46/47 and the southern one of WG 48/49/50.
Two trenches (Trench 1; Trench 2) were excavated within this area. They included the following squares, 2 m x 2 m in area:
Trench 1: WG 45, E1 (with a northern extension WG 54, E5), WG 46, A1, A2, A3 (with a northern extension WG 57, A5), over an area of 24 sq m.
Trench 2: WG 46, E1, C2, D2, E2, C3, D3, E3, C4, D4, E4, D5, E5; WG 47 A1, B1, A2, B2, A3, B3, A4, B4, A5, B5; WG 49, D1, E1, D2, E2, D3, E3, D4, E4, D5, E5; WG 50, A1, A2, A3, A4, A5, over an area of 148 sq m.
Squares WG 46, D3 and WG 50, A4 corresponded to the 2006-07 geological test pits T8-A1 and T8-A2, respectively.

Trench 1: WG 45, E1; WG 54, E5; WG 46, A1, A2, A3; and WG 57, A5 [10]
Three stratigraphic units were distinguished in this trench (Figure 18).

The upper stratum (SU1), ca 1.4 m thick, consisted of a crust of salt, ca. 0.1-0.15 m thick, covering a deposit of thin layers of windblown sand alternating with thin layers of gravel from recent wadi activity, ca. 1.1 m thick. Finds included a few shells (1 Lambis sp., ca. 45 cm below the surface; 1 Trinacria sp., ca. 75 cm below the surface) and some potsherds, ca. 80-110 cm below the surface. At the bottom was another crust of salt, ca 0.1-0.15 m thick.
Below SU1 a stratum of dark compact wet sand (SU2), ca. 0.6 m thick, was recorded. Some hearths, the largest of which measured 20 cm in diameter, were found at a depth of 1.50 m from the surface in WG 54, E5 and WG 57, A5. Excavation in square WG 57, A5 was suspended at the level of the hearths (the top of SU2). In square WG 54, E5 some fish bones and shells, wooden sticks, charcoal, and a copper strip were associated with the hearths. At the bottom of SU2 small white algae (Acetabularia) were clearly visible on the sand, together with small shells and fish bones, suggesting a marine or brackish water environment.
At the bottom of the trench, ca. 1.8 m in depth, a layer of small pebbles and sand mixed with larger shells was found (SU3), suggesting with an ancient seashore. At the depth of 1.9 – 2.20 m a very hard coral rock was found in squares WG 54, E5; WG 45, E1; and WG 46, A1.

Trench 2: WG 46, E1, C2, D2, E2, C3, D3, E3, C4, D4, E4, D5, E5; WG 47, A1, B1, A2, B2, A3, B3, A4, B4, A5, B5; WG 49, D1, E1, D2, E2, D3, E3, D4, E4, D5, E5; WG 50, A1, A2, A3, A4, A5 [11]
Five strata were recorded in this trench.
The upper stratum (SU1) consisted of a light windblown sand alternating with thin layers of gravel from recent wadi activity, ca. 0.9-1.0 m thick. Most of SU1 was sterile, with a few (washed?) potsherds in the upper layers, ca 40 cm below the surface (WG 47, A2; WG 49, E1; WG 49, E2), 50 cm below the surface (WG 47, A2), 70 cm below the surface (WG 46 D4; WG 47 B1), 80 cm below the surface (WG 46, C3), and 90 cm below the surface (WG 46, C4; WG 47, A1).
A greater quantity of pottery occurred at the base of SU1, just above the salt crust sealing SU2, ca. 1.0 – 1.1 m in depth (WG 46 A3; WG 46 C3; WG 47 A2; WG 49 E3; WG 50, A2). One Middle Nubian potsherd was found at the SU1/SU2 Interface in WG 49, E2.
A number of hearths were found at the interface between SU1 and SU2 in WG 47, A1, A2, A3, B1, B2, B3, D4; WG 46, E1, E5; [WG 47, A3, B3: large hearth/FP 1]. Many fish bones were associated with a hearth in WG 46, E1.
Below SU1 a stratum of dark compacted sand (SU2), ca. 60-70 cm deep, was found. This stratigraphic unit was covered at the top (SU1/SU2 interface) with a salt crust, ca. 15-20 cm thick, over the whole trench. [12] SU2 was contained many ceramics, which had mostly accumulated at the interface between SU2 and SU3. An alignment of stones delimiting the concentration of pottery to the south was recorded at the SU2/SU3 interface in WG 47, A4, B4; WG 49, D3, E3; and WG 50, A3. Very few fragments of pottery were found to the south of this alignment in WG 49, D4, D5.
Below SU2 a stratum of dark compacted sand (SU3), ca. 40-50 cm deep, was recorded. No pottery was collected in SU3.
SU3 covered a stratum of red/brown sand with many red/orange streaks (SU4, ca. 70 cm deep), most likely the remains of ancient mangrove roots. This SU does not occur in WG 45, E1/WG 47, A1. One fragment of pottery associated with some shells was found in WG 47, A2.
Finally, a stratum of sand mixed with pebbles (SU5) from an ancient beach was found at the bottom of the sequence.

WG 51 [13]
Excavation Unit WG 51 was delimited at the mouth of a small wadi that drains from the fossil coral terrace into the Wadi Gawasis. Sediments from this small wadi had covered the whole "harbor" area in the southwestern sector of the site. The location of this EU was chosen to test the hypothesis that a straight alignment of bushes perpendicular to the wadi indicated the remains of an ancient feature (possibly a small dam/wall). Excavations in this area, however, did not uncover the remains of any structure.
A transect, 8 m x 4 m in area, was initially excavated across the alignment of bushes, including squares C2, C3, C4, C5, D2, D3, D4, D5. The transect was later extended to the northwest with the addition of square B2.
At the top of the stratigraphic sequence a stratum of light windblown sand alternated with layers of pebbles and gravel from wadi sediments (SU1), ca. 80 cm thick, was recorded. Artifacts occurred mainly at the interface between SU1 and SU2. They included pottery, including some large potsherds and fragments of a large jar, in WG 51, C3, D4; one squared block of basalt and some small pieces of copper in WG 51, C2; and hearths in WG 51, C4, C5.
In WG 51, D3, SU1 covered a stratum of gravel mixed with sand (SU2), ca. 20-25 cm thick. Below it was a stratum of sand and gravel (SU3), ca. 30-40 cm thick, with a few washed potsherds. Finally, a stratum of dark brown, wet and compacted sand (SU4) with evidence of ceramics was recorded at the bottom.

WG 52 [14]
This excavation unit was located on the top of the slope along the southwestern wall of the fossil coral terrace at an indentation of the coral wall. An 8 m x 4 m unit was initially opened and was later reduced to a transect, 6 m x 4 m in area, in the center of the unit. On the surface many pieces of wood (cedar?); a large piece of linen cloth, sewn in the center; and a small rope or string were collected. Only the upper stratum of eolian sand, covering the coral terrace, which slopes down towards the wadi to the southeast (SU1) was excavated in this unit. Much wood was found in SU1, but very few potsherds. One small stone bowl was also found in SU1.
Next to the terrace wall was a thick, irregular deposit of salt, ca. +50 cm thick, which had to be hammered through, as there was a clear vertical cut in the wall above and beneath the salt deposit. Many stones of collapsed fossil coral from above were mixed with the salt deposit, but no artifact or potsherd was found in the windblown sand in between.
A small natural cavity was excavated to the south of WG 52, with evidence of sea grass and windblown sand. Two birds' nests were found in this deposit, and ancient rope fragments that the birds brought there to use as material for their nests.

WG 54 [15]
This excavation unit, 6 m x 4 m in area, was opened on the western side of the southern slope of the fossil coral terrace, immediately to the north of WG 45, moving into the possible harbor area. Excavations were conducted in squares A2, A3, B2, B3, C2, C3.
The top layer of WG 54 consisted primarily of sand and debris deposited by wind from the top of the terrace and nearby areas. The composition of the layers was very similar, except for the salt layer in between SU1 and SU3, and it was difficult to distinguish sediment layers. Given that the salt layer was quite compact, it was decided to continue in only the eastern half of the unit in squares B2/B3. At approximately 1 meter below the surface was the coral terrace. On the coral terrace surface there were 2 (possibly 3) hearths and one in the baulk profile in the northeast corner. A ceramic bowl with large pieces of charcoal encrusted to it by salt was found in SU3. Also in SU1 was a vitrified piece of a potsherd with some copper attached to it, but it had probably been deposited there later and was not found in situ with the hearths. Most of the potsherds excavated were salt encrusted.
A trench, 4 m x 6 m in area, corresponding to squares A1, A2, B1, B2, C1, C2, was opened in the northwestern sector of the unit, and excavation was conducted down to the bedrock. Three stratigraphic units were identified in this trench.
A stratum (SU1) was excavated, ca. 90-95 cm thick, with layers of coarse large grain, colluvial sand and thin layers of gravel and pebbles from wadi sediments, mixed with debris from the coral terrace. This stratum overlaid the coral terrace to the west and included some hearths associated with potsherds in WG 54, C1 and WG 54, C2. A hearth was recorded along the edge of B1-C1 and at the northwest corner of B1-C1. Finds included pieces of charcoal and wood, copper, potsherds, bone, and a unique burned potsherd with vitrification on part of it and some copper melted to it. Hearths associated with large potsherds (in situ) and wood were also found.
A crust of salt occurred at the base of SU1. A hearth associated with ceramics, pieces of wood and leaves occurred at the interface between SU1 and SU2, over the crust of salt. Two hearths, ca. 1 m in diameter, and pottery were also included in the crust of salt.
A stratum of dark brown windblown sand and debris from the coral terrace (SU2), ca. 1.0 m thick, was beneath the crust of salt, with evidence of hearths at the top (SU1/SU2 interface). Finds included much wood debris, charcoal, and a few potsherds beneath the upper layer with hearths. The bottom of a bowl with charcoal cemented to it by the salt was also found. Potsherds were collected beneath the layer with charcoal.
A stratum of compacted sand with a granular texture mixed with small pebbles and a great quantity of sea shells (SU3) was found at the bottom of the excavation. According to Alfredo Carannante, this stratum represents an ancient beach.
Evidence of a small hearth with many fragments of burnt wood at the interface between SU2 and SU3 was excavated ca. 1.8-1.9 m beneath the present surface.

Pleases find attached .pdf files with the full report and the related figures