Stone Village Project 2007

Anna Stevens and Wendy Dolling

Contents

Background
2007 Season
Excavation
Environmental material and artefacts
Discussion
References

This report outlines the results of the third season of investigations at the Stone Village, which ran from the 4th of November – 27th of December, 2007. The team comprised Anna Stevens, Wendy Dolling, Caroline Hubschmann, Richard Long and James Milner. The fieldwork was carried out with the assistance of SCA inspector Raghib Abdel Hamid Khalafalla and SCA staff at the Minia and Mallawi offices. The work was funded by the McDonald Fieldwork Fund (McDonald Institute for Archaeological Research, Cambridge University), the G. A. Wainwright Fund (Oxford University), the Michella Schiff Giorgini Foundation, and the Thomas Mulvey Fund (Cambridge University). We wish to thank the Egypt Exploration Society for the loan of a total station, and Professor Barry Kemp for his ongoing support of the work.

Background

The Stone Village Project focuses upon a group of Amarna-period remains located on a flat plateau in the low desert to the east of the Main City. The main feature at the site is a roughly rectangular scatter of limestone boulders lying loose amongst sand and low mounds of gravel, sherds and desert marl. This covers an area of some 65 x 80 metres, and we have termed it the ‘main site’. The boulders were used as building materials, mortared in place with marl-based plaster. Some of the stones lie in rough alignments but very little of what is visible at surface level remains in situ. The broader landscape of the Stone Village encompasses the adjacent surface and edge of the plateau and the desert floor, where there are various denuded stone structures, scatters of sherds and chipped stone, and pits and scoops. The latter seem to represent intrusive digging; some have scattered mud-brick nearby, suggesting the presence of as yet unidentified structures. It is possible that other elements of the site have been entirely buried by windblown sand, particularly in the flat terraces around the edges of the plateau, where scatters of potsherds are visible.

In previous seasons, fieldwork focussed upon recording the surface of the main site; namely, the hand-planning of a sample strip of the southern third of the site as a supplement to the photographic record. Small-scale excavation was also undertaken, in part to test the relationship between the surface and sub-surface features. In 2005, a 7 x 3 m trench was opened over the very northern edge of the main site (Trench 1). Excavation revealed a series of plastered/trampled surfaces, the lower parts of two stone walls and traces of at least two circular ovens. All were badly damaged, but belonged to at least two intercommunicating spaces. They were probably open to the sky or lightly roofed. An east-west wall passed through the centre of the trench, preserved only to its foundation course. To its north was a thick trampled surface covered by ash- and sherd-rich deposits that were probably in situ ancient rubbish. In 2006, the trench was extended northwards to expose the edge of this surface, which seems to delineate part of the northern edge of the site.

In 2006, a 5 x 4 m trench was opened in the south-central quadrant of the main site (Trench 2). This revealed the back two intercommunicating rooms of a building that extended beyond the northern trench edge. The building had again been damaged, but the walls still stood up to 60 cm high. All were formed of unworked limestone boulders joined with marl-mortar and faced also with marl-based plaster. A cruder stone wall of uncertain function divided the eastern room in two. The fill contained several pieces of roofing plaster, and also bricks. An extension, measuring 1 x 1 m, was cut from the south-east corner of the trench to establish the width of an east-west wall, [11489], that ran through the southern baulk. The southern face of the wall was defined, but its base was not reached. It was speculated that the two exposed spaces may have been the back two rooms of a narrow house, similar to those at the Workmen’s Village.

2007 season

Aims

Four broad aims guided the work of the 2007 season: to complete the sample strip of surface plan across the main site; to plan the three most prominent stone structures on the plateau (Structures I, II and III); to extend Trench 2 to the north, west and south in order to define fully the building exposed in 2006, and determine what bordered it; and to open a trench through a dense ashy deposit that runs through much of the east margin of the main site. The results of this work are summarised here. In a separate period of work at the Stone Village, in March 2008, Malcolm Williamson collected data for a three-dimensional scan of the site.

Progress and results

 

Survey

Approximately three weeks was devoted to the hand-planning of surface remains, which saw the completion of the survey tasks at the main site and at Structures I–III. Structures I and II lie to the south of the main site, in shallow valleys on the perimeter of the plateau. The valleys link into a broader sand-filled wadi that leads down into the main site, and which was probably part of the Amarna period landscape.

At surface level, Structure I presents as a rectangular building measuring some 12 x 7 m. Its walls still protrude slightly above the desert surface, particularly the western wall. A break in the eastern end of the north wall may be an entranceway. There is no evidence, at least from the surface remains, of substantial intrusive digging within this structure. Structure II, on the other hand, seems to have been quite heavily disturbed and it is more difficult to gain a basic sense of its layout. A scatter of limestone boulders, some in vague lines, along with pits and low moulds of sherds, gravel and marl, cover an area of around 17 x 12 m. An unusual feature at its northern end is an arrangement of small stones with a vaguely and irregularly grid-like layout, which appears to continue northwards down the slope beneath the sand filling the wadi. A slight question hangs over the date of this arrangement. Structures I and II are connected by one of the spans of Amarna period ‘roadway’ that ring the Stone Village, and the surface plan was extended to include this and other sections of the road network immediately adjacent to these structures. The position of Structures I and II appears to be a strategic one, perhaps connected with traffic coming into the site. This we might understand as largely crossing over the ‘roadways’ rather than coming around them (cf. Kemp 2008).

Structure III is a much smaller denuded stone structure on the plateau surface to the north of the main site. At surface level it survives as a c. 3 x 4 m rectangular arrangement of small unworked limestone boulders slightly embedded into the gravely plateau, with an open projection on its western side. Its function is also not immediately clear. It does not interact directly with any of the surviving roadways. Its position would be suitable perhaps as a watch-post, whilst its layout recalls that of some altars and offering tables. For now, slight uncertainty should also surround its date, although the presence of a similar, but much larger, arrangement of stones on the plateau behind the Workmen’s Village goes some way to confirming its antiquity.

Excavation

Excavation was conducted from 17th November to 13th December. The three excavation areas were opened simultaneously: Trench 2, forming the western and northern extension to the 2006 trench; Trench 2, southern extension; and Trench 3, in the ashy deposit along the western margin of the main site. All excavation was conducted by team members, with five local workmen (Waleed Mohammed Omar, Ahmed Mokhtar, Mahmoud Said Nasser el-Din, Mustafa Mohammed Abdel-Gabr and Gamal Abdel-Halim) engaged to removal and sieve the spoil. All deposits were sieved through a 4 mm sieve, and the three-dimensional position of artefacts (excluding pot sherds) identified during trowelling was recorded.

Trench 2

Trench 2 was extended 8 m to the north and 2.5 metres to the west to form a total area of 7.5 x 13 m, including the original 4 x 5 m of the 2006 trench.

At commencement of excavation a deposit of loose wind blown sand, (11708), was removed. This covered the majority of the trench with the exception of a small area along the north-west margin. Here, the surface deposit comprised mixed sand and marl rubble fill (11714), although the exact horizontal transition between these deposits was slightly blurred. Once the surface layer was removed distinct clusters of limestone boulders became apparent; in particular, a dense scatter running along the western margin of the trench ((11747) and (11751)) two clusters in the central area ((11724) and (11722)) and several large poorly defined clusters at the northern end of the trench ((11721), (11748) and (11749)). Surrounding these was a series of loose fill units consisting of sand, marl, silt and ash in varying concentrations. Excavation proceeded by removing each deposit across the entire excavation area. None of the deposits revealed at this stage were original occupational fill but were instead either structural collapse or destruction deposits. As excavation progressed it became apparent that different fill types were concentrated in different areas of the trench. Whilst there was a degree of blurring between fill units, particularly at the upper levels, three broad areas could be distinguished and are discussed separately below.

Southern trench area

A line of relatively clean sand, (11720), was evident early in the excavation running north-south along the approximate alignment of wall [11491]. This wall was exposed in 2006 and at that time was thought to potentially mark the western limit of the structure. With increasing depth of excavation there remained a distinction in the fill units to the east and west of (11720): the south-east and south-west areas.

South-east area

This area lies directly to the north of the 2006 trench. Following the clearance of surface sand, distinct deposits of displaced limestone boulders, marl rubble and sand were evident. On excavation some of these deposits were found to be structural collapse ((11719) and (11753)), whilst others were clearly the result of a deliberate digging-over of the site ((11725) and (11726)). As excavation progressed, the south-east area was found to be bordered on all sides by stone and marl-mortar walls: to the east by wall [11734], continuing the alignment of [11490] exposed during the 2006 season; in the west by a northward continuation of 2006 wall [11491]; in the south by a well preserved wall [11481], also exposed in 2006; and in the north by a previously unidentified wall, [11729]. In general the construction method for all of these walls is comparable, although their state of preservation varies. They are constructed of uncut limestone boulders, or sometimes smaller cobble-sized stones, which are bonded together with marl mortar. The mortar contains only a small number of inclusions: minor charcoal and gravel, occasional pebbles and small sherds, but no obvious straw or chaff. Where preserved, the facing plaster appears to be of the same simple marl mixture as the bonding mortar, smoothed over the wall face. Both the bonding mortar and the facing plaster vary slightly in colour from a distinct orange brown to a more greyish brown. Traces of gypsum plaster are preserved on walls [11734] and [11491]. Whether the absence of such plaster on other walls in this area of the trench reflects their original state or is due to erosion or wear is difficult to determine.

Two low lines of east-west orientated stones transect the southeast area; the southernmost is unit [11893] and the northernmost unit [11892]. These appear to be roughly laid walls, although it is not possible to determine if they originally continued to roof level. [11893] consists of uncut limestone boulders preserved to a maximum of three courses, although over much of its length only one layer of stones is preserved. Marl bonding mortar is only clearly preserved in the central part of the feature, but at the western end a line of marl mortar, [11951], with stone impressions on the southern side, is located directly to the north of [11893]. It is probable that the adjacent stones of [11893] were originally bonded to the mortar before they slumped down and to the south. At this western end of the wall the stones butt against the base of north-south wall [11491]. The eastern end of [11893] is set back into a slight hollow or indentation in the face of north-south wall [11734]. There is no bonding mortar between the two walls, raising the possibility that [11893] was a secondary addition. A large intrusive pit <11901> undercut the southern edge of [11893], and during excavation several stones fell from the wall into the pit. Originally, however, there was no break in [11893] here.

Wall [11893] is bordered on either side by a partially preserved floor. The relationship of the wall to the floor varies along its length. The eastern portion of the floor may have been truncated to allow for the installation of [11893], as the floor edge slopes roughly down and out towards the underlying gebel, onto which the stones of the wall rest directly. At a point approximately half way along [11893], a section of floor 11953 passes beneath the wall. The section of marl mortar [11951] with stone impressions at the western end of [11893] also rests on top of a remnant of floor, 11952. Whilst the floors adjacent to [11893] are extensively damaged, they are still the best preserved floor remnants within this area of the trench. At least three distinct layers are visible in the truncated edges of the floors here. The uppermost level (units 11952, 11953 and 12003) is greyish brown in colour and contains fine nodules of orange brown marl, fine-grained sand, minor gravel and very minor charcoal. The floor is slightly laminated in appearance when seen in the truncated section, with some irregular ashy lenses. On the northern side of the wall there seems to be two plastering phases, 12003 and 12004, that are not clearly apparent in the equivalent floor on the south side. Beneath portions of these floor levels, a distinct layer of compacted coarse- to medium-grained sand, (11999), is visible. In turn, there is a lower plastered floor level, 11976, 12000 and 12005, beneath this sand lens. It is not clear whether the sand represents a deliberate foundation for the upper floors or accumulated over the earlier floors naturally. The lowest floor comprises greyish brown silty sand with minor inclusions of gravel, small sherds, and moderate charcoal. It has a burnt appearance and, like the later floor levels, portions of this surface clearly pass beneath wall [11893]. There is one small patch of white gypsum plaster, 2mm thick and measuring 16 x 10 cm, on the uppermost floor surface, 12003. Therefore, at least two phases of activity may be evidenced here. Some of the preserved floor layers seem to predate the construction of the east-west wall [11893], but may be contemporary with the more substantial walls to the south and west, [11481] and [11491], as traces of plaster flooring butt against the base of these walls.

Architectural modification is also apparent with the second low stone wall, [11892], and the more substantial wall, [11729], to its north. [11892] is constructed from uncut limestone boulders with marl mortar plugs. The marl mortar is degraded in many places so that the original width of the wall was difficult to determine during excavation. The wall is comparable to [11893], although the mortar is in a better state of preservation. Its eastern end is preserved to the greatest height, of 3–4 irregular courses. Here, the wall rests directly on to the gebel surface and butts against the face of north-south wall [11734]. The facing plaster of [11734] can be clearly seen passing behind the end of [11892]. The western end of the wall is less well preserved, with the last c. 90 cm of its length surviving only to one course. It is possible that this western end marks a door or access point. The central portion of the wall is built directly over a thick layer of burnt ash and chaff that also continues to the north and passes beneath structures located there, including wall [11729].

[11729] appears to have at least two phases of construction. The eastern end is of a comparable construction technique to wall [11734], to which it is clearly bonded. Both these walls are constructed of limestone boulders bonded together with orange-brown marl mortar, and faced with the same mortar. At a distance of 1.2 m from the eastern end of wall [11729], a modification or addition to the structure is discernable. Whilst the eastern portion of the wall rests on top of natural sandy gebel, at the point of modification a thick layer of ash and chaff rich fill (11983) passes beneath the wall, elevating the base level of the western end of the wall. The exact transition point is a little unclear as a layer of slightly irregular marl mortar is smeared over the junction of the eastern wall and western alteration. From the initial transition point for a distance of 1 m westward, the wall face is rough and irregular, whilst the last 1.3 m is faced more smoothly with marl plaster. The plaster in both these sections of wall has inclusions of straw, which is absent in the plaster facing on the eastern end. The irregular area in the central part of the wall is potentially a blocked doorway.

The western end of wall [11729] appears to be truncated, and there are remnants of mortar extending westward from its base. This suggests that the wall may have continued westward for a distance of c. 1 m, at which point a floor level 11846 borders the remnant mortar. A short length of north-south wall [11862] extends from [11729] to the south, where it is bonded to wall [11892] with an application of mortar. [11862] is constructed of the standard uncut limestone boulders with a marl mortar. It is probably of the same building phase as the west end of wall [11729], as it is built over the same deposit of ash and chaff (11983). [11862] forms the western limit of a space bordered on three other sides by stone and mortar walls [11729], [11734] and [11892]. Apart from the possible blocked doorway in [11729], the only potential access point to this space seems to be through the southern end of wall [11862], which is preserved to lower height than its northern span. The floor in this room has been heavily disturbed. There are only two tiny patches of plaster floor preserved, both in the south-east corner, with the remaining floor cut through to the level of the gebel.

South-west area

Once the surface units in the western margin of the trench were excavated, it was found to be covered by a scatter of limestone boulders in a matrix of sand and fragmentary marl rubble, units (11747), (11750) and (11751). Running along the eastern side of this stone scatter were several irregular mounds of crumbled and fragmentary marl, (11765), (11793) and (11807), slightly displaced to the west of the line of north-south wall [11491]. The mounded marl seems to be collapsed structural material from the wall, potentially degraded brick or marl mortar. The removal of both the limestone scatter and a large mounded marl deposit in the south-west corner of the trench revealed several structures: an east-west wall running along the southern baulk, [11950]; the west face of north-south wall [11491]; an additional north-south wall with distinct mud brick impressions in its upper surface, [11781]; and an east-west wall [11754]. The basic building material of these walls is consistent with those elsewhere: unworked limestone boulders and orange-brown to grey-brown marl mortar. The walls create a separate architectural space, the western limit of which lies outside the excavation area. A narrow area between walls [11781] and [11491] is difficult to interpret as a functional space. It may have been accessed from the south – external to the building – as there are indications of a possible break in the southern wall [11950] at this point. Floor deposits were poorly preserved here and have probably been deliberately cut-through. A deposit of slightly silty sand with marl clumps in the south end of this space, (11791), had a level surface with several embedded potsherds, and may represent accumulated fill that became compacted through activity or as a result of weathering. It overlay a similar, but slightly less compact, deposit (11943), which in turn overlay gebel.

To the west of wall [11781] a small area of slightly silty sand, (11784), was preserved that seems to continue beneath the wall itself; a comparable layer of silty sand underlay the wall forming the northern limit of this space, [11754]. To the north of wall [11754], the removal of rubble fill exposed a second space that is bordered to the east by a continuation of wall [11491], and to the north by a roughly constructed stone and marl wall, [11974]. As in the southern space, this area has undergone a degree of deliberate destruction, and was largely covered by a deposit of dug-over floor. Beneath this deposit a remnant of a possible working surface is preserved as a smear of slightly silty sand with a few small embedded potsherds, (11966). At the north end of the space a small circular depression lined with mud is possibly a vessel emplacement, [11957]. There is also a remnant of burnt mud floor, 12032, that continues beneath wall [11974]. This area it is potentially an exterior space. In comparison to rubble deposits from the south-east area, only a minor amount of roofing material was recovered here.

Northern trench area

This area comprises the northernmost 5 m of the trench. Underlying the surface deposits were scatters of limestone boulders, (11748) and (11732). These covered much of the space, although they were less dense in the north-east and north-west corners of the trench; in general, stone rubble was less concentrated in this northern part of the trench than in the south. The excavation of the boulders and associated sandy fill exposed several units, the majority of which comprised sand with significant amounts of ash and charcoal, and variable quantities of marl and chaff. These were probably the remains of a surface that originally covered much of this northern area, and has been dug over. The stone clusters are most likely the result of a combination of deliberate destruction and a degree of structural collapse from surrounding walls.

In the north-east corner of the trench a small space emerged, bordered by limestone and mortar walls; on the west [11787], on the south [11777] and on the north [11959]. The limits of this structure are located beyond the area of excavation. A limestone block, [11960], of uncertain function is mortared to the base of walls [11959] and [11787]. Poorly preserved sections of greyish brown floor plaster, 11961, lip down from the base of [11787] and from the stone block [11960], but the remainder of the floor has been cut through to the level of the underlying sandy gebel.

South of this structure a second small space was revealed, bordered to the north by wall [11777], the east by [11789] and the south by a slightly curving wall [11772]. The west end of the curving wall has been destroyed, but there is a line of hardened gebel extending northwards here that may mark an original continuation of the wall. Relatively well preserved white gypsum plaster lips down from the northern wall face over the remnants of floor plaster. The stones in [11772] are all relatively small (less then 20 cm) and are bonded with a slightly friable mortar that has inclusions of charcoal and is greyish in colour. This wall bonds to the east wall [11789], which is constructed of comparable materials. The two walls clearly represent one phase of construction. [11789] also has traces of white gypsum plaster on its east face. At the northern end of this wall. traces of grey marl plaster overlie the gypsum layer; this over-plastering is at the base level of the wall and may relate to a phase of floor plastering. Wall [11789] appears to abut wall [11777], and so possibly represents a later construction phase. The mortar of [11777] is also more distinctly orange-brown in colour than the greyish brown mortar of [11789] and [11772]. Floors in the southern space have been largely destroyed, although a remnant of floor, 11981, lipping down from the walls in the south-east corner has a matrix of sand, ash and minor gravel inclusions. This floor level is over-plastered by white gypsum in a few patches, generally at the transition from floor to wall. Underlying this surface is a pale grey compacted surface 11982 with minor inclusions of gravel, sherds and charcoal: either a working surface or possibly a deliberately laid floor. This rests directly on the gebel. The southern wall [11772] and most of the length of [11789] are built over the thick ashy straw-rich deposit (11893) that extends over much of the north of the trench.

An ash rich surface, 11846, covers much of the remaining area across the northern margin of the trench. The visible upper horizon of this surface is brown to black in colour and comprises silty sand with moderate gravel and chaff/straw inclusions. It has a few embedded sherds and charcoal pieces, and some distinct burnt parches. The immediate surface has been broken through in a number of places, exposing the slightly friable matrix, which is up to 9 cm thick. In section 11846 has a slightly laminated appearance with some irregular blackened lenses, but no distinct plastering phases visible. It seems to be an exterior surface and is reminiscent of the surface exposed in Trench 1 in previous seasons. It overlies the ash- and charcoal-rich deposit (11983) discussed previously.

Excavation of rubble from the north-west corner of the trench exposed a particularly thick section of a similar ash- and charcoal-rich surface, 11805. Its eastern edge drops down sharply to the underlying gebel, and in section it appears slightly laminated with a matrix of slightly variable ash and silt bands. There are inclusions of small marl nodules, charcoal, gravel, minor ceramic and very minimal chaff. As with 11846, there are no distinct laid floor levels visible. Surface 11805 butts against the base of a wall extending eastwards from the trench baulk, [11717]. This is constructed of unworked limestone boulders bonded with an orange-brown marl. The northern wall face is poorly preserved, but the south is formed of smoothed mortar. The base of the wall is visible only on its southern side, where it overlies the sandy gebel. The eastern end of the wall is truncated. On its north side here, a line of greyish brown mortar [12033] with flecks of white, possibly gypsum, steps out marginally from the wall base and continues eastwards in a slightly curving line. It rests on surface 11805. On the south side of the wall, a section of slightly irregular marl mortar with at least one limestone boulder, [12035], bonds to the base of wall [11717] and curves to the south-east. A roughly circular space is created by these features, filled by an unexcavated deposit of semi-compacted ash-rich fill with floating limestone boulders (12034). This might mark the location of an oven. Although no fragments of ceramic oven liner were preserved in situ here, a number of fragments of such were recovered from loose fill in the north-west corner of the trench. A second possible oven emplacement is located immediately to the south-east, where a deep pit <11971> cuts down through surface 11846 into the underlying sandy gebel. Although these features cannot be identified definitely as oven emplacements, there are clear indications that the burning of wood or charcoal was carried out in this area in the frequency of ash and charcoal in the disturbed deposits and in the sections of preserved flooring. Some spans of gebel surface exposed across the north of the trench also have a slight pinkish hue that suggest exposure to heat, whilst preliminary analysis of the charcoal from the site shows that it is dominated by Nile acacia, which is known to be an excellent fuel (Gerisch 2007).

Trench 2 southern extension

The southern extension to Trench 2 was opened mainly to investigate the question of whether the main site had an enclosure wall like that at the Workmen’s Village. In the 2006 season, an east-west wall [11489] forming the southern limit to the exposed building, and projecting beyond it to the east and west, was revealed through the immediate southern baulk (Figure 2). At around 30 cm wide, it was no thicker than the rest of the walls of this building, and much narrower than the Workmen’s Village enclosure wall, which was around 80 cm wide (Peet and Woolley 1923: 53). This suggested that structures along the southern part of the main site shared a southern wall, but were not contained within a substantial enclosure wall. In order to check that [11489] did in fact mark the southern limit of the sub-surface structures, in 2007 an extension measuring 1.5 x 4 m was excavated directly southwards from the south-east corner of the 2006 excavation area. Excavation revealed a series of deposits, largely stone rubble and wind-blown sand, overlying a denser horizon of fill richer, especially, in sherds. The latter is probably undisturbed ‘occupation debris’ that spread out over the desert surface in the Amarna period. The main result, therefore, was of no indication of any further structures, or a laid surface, immediately south of wall [11489].

Trench 2 and its southern extension: architectural overview

Trench 2, including the southern extension, is now the largest exposure across the main site. It has revealed the remains a structure comprising two small back rooms (the easternmost with a transverse ‘partition’) fronted by a series of three narrow chambers that were probably accessed from the west. These appear to form a single structural unit, albeit one that has undergone at least one modification phase. The walls of the structure are built of unworked limestone boulders, most mortared and faced with marl-based plaster. Several preserve a coating of gypsum plaster, which may have extended over some of the floors. The building was certainly roofed, with many roofing fragments found in the fill.

The south wall of the building, [11489]/[11950], continues to both the east and west of the excavation area and seems to mark the southern edge of the main site. The north-west corner of the building has been damaged, but its entranceway was probably located here. Access to the southern part of the structure could only be achieved by passing through the narrow rectangular spaces created by the low stone walls [11892] and [11893]. Although no obvious doorways are visible, the western end of [11892] and the eastern end of [11893] are preserved as a single course of stones and as such are potential access points. Given the rough construction of these two walls, we cannot be sure that these were full-scale floor to ceiling walls and not partition walls or emplacements of some sort.

The south wall of the building, [11489]/[11950], continues to both the east and west of the excavation area and seems to mark the southern edge of the main site. The north-west corner of the building has been damaged, but its entranceway was probably located here. Access to the southern part of the structure could only be achieved by passing through the narrow rectangular spaces created by the low stone walls [11892] and [11893]. Although no obvious doorways are visible, the western end of [11892] and the eastern end of [11893] are preserved as a single course of stones and as such are potential access points. Given the rough construction of these two walls, we cannot be sure that these were full-scale floor to ceiling walls and not partition walls or emplacements of some sort.

It is difficult to know how the walls exposed along the immediate west margin of the trench relate to the main structure. It is conceivable that they represent a separate building or buildings, perhaps accessed through a break in wall [11950]. But it is also possible that the chambered-structure is part of a larger complex that extends to the west. The most likely access point is via the north-west corner of the former, a part of the building that is unfortunately badly damaged. Both the northern end of wall [11491] and the eastern end of [11974] appear to be truncated, but perhaps not heavily, and the possibility remains that there was an access route here between the main structure and the spaces to the west.

The area north of wall [11729] is an external space partially covered by a thick ash-rich surface. It is not clear if this belonged to any one building complex or was ‘public’ space. A curving wall [11772] in the north-east quadrant defines part of a small chamber, probably entered from the north, its inner walls also covered by gypsum. The corner of another ‘room’ is present in the immediate north-east corner. It continues beyond the excavation area, and is not accessible from it. Immediately south of wall [11772] is a band of open space that was possibly a thoroughfare, although a line of floor plaster, and plaster protrusions on the adjacent parts of walls [11772] and [11729], may mark the line of a wall that blocked access here. It is very likely that ovens were located somewhere within the northern part of the trench, perhaps adjacent to wall [11717] in the north-west corner. The north-central quadrant of the trench has been badly damaged, exposing a large area of gebel surface, and it is conceivable that further walls or features such as ovens were once located here.

Trench 3

The dense slightly mounded ash band that spreads, discontinuously, along the eastern margin of the main site was chosen as the location of Trench 3 for the possibility that it represented in situ rubbish from the occupation of the site. This turned out not to be the case, but the results of excavations here were still of considerable interest.

A 6 x 1.5 m trench was laid out in the south-east corner of the main site, encompassing the dense ash through most of its length, but a sandier deposit with limestone boulders in its north. The upper layers of fill comprised variously dense black ash, clean sand, and a mixture of the two. At a depth of around 40 cm, the top of a circular ceramic oven liner emerged, unit [11820]. Further clearance of the ashy fills revealed several curved lengths of marl, sometimes with stones, and circular depressions that marked the original locations of further ovens liners. In the south-east corner, the tops of two large storage vessels were exposed, [11928] and [11929], surrounded by a packing of marl mortar and stones. In order to provide more working space, a 1 x 4 m extension was set out along the northern end of the west baulk. The upper levels of ash and sand were removed until a uniform level was reached across the excavation area. The final stage of excavation was then to clear the fill from in and around the ovens and associated features. Across the northern margin of the trench, this exposed a horizon of sherds, gravel and marl nodules (12024) that may represent rubbish that had accumulated over the desert surface before the ovens were built.

At the end of excavation, at least six probable oven emplacements could be identified, although only [11820] retained the ceramic oven liner itself. The ovens were not of uniform size or shape. [11820] had a diameter of 60 cm, and in form and scale it recalls the ovens well represented elsewhere at Amarna, often in domestic contexts (e.g. Samuel 1999: pl. 3). Another oven of similar size may have been located in the north-west quadrant of the trench, represented only by a circular break in the ash that had built up around it. Small curved mortar edges nearby seem to mark the edges of smaller ovens, perhaps better understood as supports for cooking vessels, or similar. A curving line of mortar immediately to the south-east of [11820] indicated the location of another small ‘oven’. A circular patch of ash west of the storage vessels marked the location of an oven, entirely removed, with a diameter of 47 cm, and a break in the ash north of [11820] may indicate another oven of similar size here. At the close of excavation, much of the southern and central parts of the trench were covered by a flat horizon, of ash, chaff and charcoal that seems to be debris from the ovens that was being continually trampled into a surface. It provided access at least to the emplacements in the southern half of the trench, with the storage vessels set into it. Spans of flat burnt surfaces, perhaps plastered, were exposed on a similar horizon in the immediate north-west corner.

It is difficult to reconstruct the setting of the ovens. It is likely that they were exposed to the sky because of the smoke they generated, which the absence of roofing fragments in the fill seems to confirm. But it is not clear if they were contained in a separate enclosure, or attached to a particular building. The best candidate for a wall proper, as opposed to an oven retaining wall, was a curve of well-mortared stone in the immediate north-east corner; however, this could mark the location of yet another oven, located beyond the north baulk. It is almost certain that the ovens were primarily for the preparation of food, and not for high-temperature industries such as metal or faience production, since the trench yielded no obvious by-products of these manufacturing processes. The ceramic corpus also included pieces of ‘bread platters’ whilst animal bone, including some butchered and baked pieces, was found in the surrounding fill.

Environmental material and artefacts

Relatively well preserved botanical specimens were recovered from all excavation areas, with matrix samples collected this year only from deposits that were rich in plant material or were undisturbed. Processing of the plant remains is scheduled for late 2008. A reasonably high quantity of animal bone was also recovered from all trenches. Most was unburnt, but some pieces showed evidence of baking and/or butchery. Preliminary examination suggests that it was largely mammalian, and fish bone could also be distinguished on site. During the Spring 2008 field season, Professor Tony Legge processed the animal bone from the 2005–7 seasons at the Stone Village, continuing work begun by Phillipa Payne.

The 2007 season also expanded the range of objects from the site. Trench 2 yielded the first examples of mud seal impressions. One of these, an incomplete impression, has the added interest of providing the first royal name yet found at the site, that of Akhenaten (obj. 38103). Two others bear parts of emblematic impressions (obj. 38102, 38437). Jewellery items remain amongst the most commonly occurring artefact: faience beads, finger rings and pendants, including an unusual Taweret pendant formed in the round (obj. 38133), and an occasional glass bead. A small bronze ring bezel with an incised representation of Bes (obj. 38361) was found in Trench 3, in sherd-rich fill that may predate the construction of the kitchen. A fragment of a pottery mould for a probable petal-shaped pendant points to the likelihood of faience production at the site (obj. 38051). As in previous seasons, organic materials were well preserved, and many small pieces of worked wood, including a wooden ear stud (obj. 38120), textile, and one small piece of a possible leather sandal (obj. 38449) were retrieved. Other finds included parts of mud and gypsum jar-seals, a limestone figure of a monkey, fragments of pottery cobra figurines, a small collection of flaked stone, many re-used potsherds, and a small group of metal implements such as needles, and corroded metal nodules.

Discussion

We are now in a position to expand on some preliminary observations made in previous reports on the layout, function and history of the site. The hub of the Stone Village was a complex built of mortared limestone boulders and brick that covered an area of similar size to the settlement at the Workmen’s Village. It probably had a wall around at least part of its perimeter, which lent it a sense of containment; this is particularly evident in aerial views of the site. But this boundary was not as imposing as the retaining wall at the Workmen’s Village and was probably more permeable, with everyday refuse left to accumulate around the perimeter of the site. Although we are yet to identify any definite points of access into the main site, there were conceivably several. One can possibly be distinguished at surface level in the south-west corner of the site, and there is a possible break through excavated wall [11950].

Internally, the site also differed from the Workmen’s Village: the Stone Village is not another example of uniformly laid out residences separated by narrow alleys. The main site at the Stone Village appears to have contained much more open space than the walled settlement at the Workmen’s Village, whilst the structures excavated in Trench 2 have little in common with domestic buildings generally at Amarna. The exposed narrow chambers are vaguely reminiscent of storage or production chambers represented, for example, at the ‘magazines’ south of the Great Aten Temple (Pendlebury 1951, 106, pl. XVIII). There are also similarities here with a building excavated in 1980 at the Workmen’s Village (Building 350), which incorporated a group of narrow gypsum-plastered chambers (see Kemp 1983: 10–12; 1984). Their function is not clear, although they may have been connected with a group of nearby animal pens.

The use of gypsum wall plaster at the Stone Village must provide some clue as to the use of the exposed structures, but pinpointing its purpose is again difficult. This surface finish is found elsewhere at Amarna in such spaces as bathrooms, butchering yards, on quern emplacements and on ritual or cultic emplacements (Kemp 1986: 71–5). The full extent or range of its uses is not understood, but it seems often to have helped demarcate a clean space, in some cases with a purifying effect. The gypsum plaster that features in many of the internal spaces in Trench 2 at the Stone Village could suggest an attempt to separate clean indoor activities from the dirtier activities outside. The latter included the burning of wood/charcoal, in large part probably in cooking fires.

Whilst the issue of who occupied the site, and what activities they engaged in, remain problematic, food production was clearly a prominent activity. The ash band into which Trench 3 was sunk is visible along much of the eastern edge of the site and conceivably marks a large, but perhaps not continuous, kitchen complex. We can speculate that the quantity being produced was not all for on-site consumption, and that one role of the Stone Village was as a supply centre, perhaps providing food for workers based out in the desert, such as army personnel, tomb-builders or quarry-workers. That it was established and run as a state enterprise is almost certain, given its position on the fringes of the city and ongoing requirement of such supplies as water from the main city, although this is not to dismiss the possibility also of private enterprise at the site.

Otherwise, the object corpus suggests aspects of textile production, faience production and perhaps some stone-chipping were taking place, but not necessarily on a large scale. There is also quite a strong ‘domestic’ strand to the object corpus, where such items as fragmentary figurines of cobras, and bowls with traces of incense are represented: items well suited to everyday cult. A striking aspect of the jewellery corpus is the relative large quantity of images of ‘household’, and often female-associated, divinities such as Bes and Taweret. Over two-thirds of the pendants from the Stone Village represent such figures (although the sample size of 16 pendants is very small) compared with 13% from excavations in 2004 and 2005 in a housing area in the Main City (Excavation Grid 12). If the site was indeed a supply centre or similar, it may still have been a largely residential environment. Although we lack any building that yet conforms to a typical Amarna house design, it is probably reasonable for now to retain the idea that a core group of people actually lived at the site. The residential space may have been located in a part of the site not yet exposed, perhaps in an upper story, or simply took a form that differed from the typical ‘Amarna house’.

The question of how long the site was occupied is difficult to approach archaeologically. Assuming it was not recirculated, the mud sealing with the name of Akhenaten from this year’s work suggests that the site was occupied, and presumably established, during this king’s reign. It is a less complicated piece of dating evidence than a hieratic jar label mentioning a year date of Year 4 found in 2006 (obj. 37808; Stevens and Dolling 2007: 9). The site was occupied for long enough for considerable rubbish and ashy debris to build up, and for quite extensive internal modification to have taken place. Very few of the walls in Trench 2 are built over gebel. Many of those across the southern half of the trench, especially through the south-west quadrant, overlie a fill of ashy sand and sherds that seems to be occupation debris that gradually accumulated on the desert surface before these walls were built. This includes the southernmost wall [11489]/[11950], which has potential as a reasonably long stretch of perimeter wall. This could imply that there was quite substantial activity at the site before the major alignments were set out, and raises the possibility that both the appearance of the site and conceivably its function changed quite considerably over time.

Future work

The Stone Village Project will run for two more field seasons. A final season of survey and excavation will focus on the extramural areas and the broader landscape of the site, looking especially at the roles of Structures I and II, and a study season will be undertaken to complete the recording of excavated materials.

References

Gerisch, R. 2007. ‘Charcoal’ in B. J. Kemp, ‘Tell el-Amarna, 2006–7’, Journal of Egyptian Archaeology 93, 52–3.

Kemp, B. J. 1983. ‘Preliminary report on the El-‘Amarna expedition, 1981–2’, Journal of Egyptian Archaeology 69, 5–24.

Kemp, B. J. 1984. ‘Report on the 1983 excavations. The animal pens (Building 400)’, in B. J. Kemp (ed.), Amarna Reports I, London: Egypt Exploration Society, 40–59.

Kemp, B. J. 1986. ‘Report on the 1985 excavations. Building 540/541’, in B. J. Kemp (ed.), Amarna Reports III, London: Egypt Exploration Society, 60–79.

Kemp, B. J. 2008. ‘Amarna’s ancient roads’, Horizon: The Amarna Project and Amarna Trust Newsletter 3, 8–9.

Pendlebury, J. D. S. 1951. The City of Akhenaten III. The Central City and the Official Quarters, 2 vols, London: Egypt Exploration Society.

Samuel, D. 1999. ‘Bread making and social interactions at the Amarna Workmen’s Village, Egypt’, World Archaeology 31, 121–44.

Stevens, A. and W. Dolling, 2006. ‘The Stone Village: 2005–2006’, in B. J. Kemp, ‘Tell el-Amarna 2006’, Journal of Egyptian Archaeology 92, 23–7.

Stevens, A. and W. Dolling, 2007. ‘The Stone Village’, in B. J. Kemp, ‘Tell el-Amarna’, Journal of Egyptian Archaeology 93, 1–11.