This document is part of the online version of the book Amheida II: A Late Romano-Egyptian House in the Dakhla Oasis / Amheida House B2 by Anna Lucille Boozer, which is available at http://dlib.nyu.edu/awdl/isaw/amheida-ii-house-b2/. It is published as part of the NYU Library's Ancient World Digital Library and in partnership with the Institute for the Study of the Ancient World (ISAW). Further information about ISAW's publication program is available on the ISAW website. Please note that while the base URI of this publication is stable, the exact content available at that address is likely to change over time.
We excavated the domestic remains of Amheida House B2 over the course of three seasons (2005–2007). We dedicated three additional seasons to finds processing (2008–2010). The majority of the material presented in this volume was excavated in February of 2005 and 2006, and we conducted additional research in the vicinity of B2 in February of 2007. Finds processing took place during excavations and continued as study seasons during January–February 2008, 2009, and 2010 (Figure 4.1).↩#p149
We began excavations with the three eastern rooms of B2 (rooms 1, 2, 3) in 2005, under the area supervision of Eugene Ball. Excavation proceeded by room since the walls of the rooms were visible from the surface. We employed the single context system, as outlined in the previous chapter, in our excavations. We excavated these rooms simultaneously, with one supervisor managing each room. The supervision of room 2 changed during the course of our excavations, leading to a re-evaluation of the contexts during excavation and during post-excavation. This re-evaluation is visible in the analysis of the stratigraphy in room 2. At the very end of the 2005 field season we began excavating room 4, removing only a small amount of windblown sand. We continued the room 4 excavations in 2006.↩#p150
The 2006 excavations focused on the western two-thirds of the house under the supervision of Anna Lucille Boozer. As with the easternmost rooms, the walls were visible from the surface so we proceeded by room and employed single context excavation. We began new excavations in rooms 5 and 6. One supervisor directed excavations in rooms 4 and 5, while another managed room 6. We continued with rooms 7 and 8 and 11 with one supervisor overseeing this work. Rooms 9 and 10 were excavated last, and were supervised by one individual.↩#p151
In 2007 we focused our excavations on areas external to B2, namely in the street (S1) and the courtyard (C2). In 2004, Eugene Ball conducted preliminary research in this area but did not excavate. Rather, his team conducted surface scraping on structure 1.1 and in Street S1 in front of structure 1.1 in order to gain a sense of the time horizon or sequence of occupations in this vicinity of the site. In 2007, we chose to excavate a trench in Street S1 immediately in front of B2 in order to determine the relationship between the structure and the street as well as to clarify phasing in B2. Likewise, we excavated a trench (C2A) in Courtyard C2 in order to clarify phasing in B2, the function of this external courtyard, and the relationship between B2 and the courtyard.1↩#p152
Figure 4.1: Plan of House B2, Courtyard C2, and Street S2 with excavated areas shaded.
This chapter presents the general stratigraphic analysis of B2 and its surroundings and also offers conjectures about the phasing of this structure. The following pages delineate the excavation of this structure in a manner that is intended to closely reflect the excavation process. I begin with a discussion of what was visible from the surface and then move through the excavation of each layer as we encountered it, followed by a description of the visible architectural elements. I summarize the excavation results for each room and then review the phasing for each area from earliest to latest phases: B2, the street north of B2 (S1), and the courtyard south of B2 (C2). SUQ quantities, single context plans, additional photographs and other information for each DSU can be found by consulting the open-access database.↩#p153
B2 Description
General View
The house designated as B2 is a building of moderate size, 11 x 11 m (121 m²) (Figure 4.1).↩#p154
The structure was badly eroded by the wind. The greatest preserved wall height (from the top of foundations to the top of the wall) is in the northwest extremity of the house, where it is 1.32 m, while the lowest preserved wall height, in the southeast corner, is 0.65 m.2 In some parts of the structure fifteen courses of bricks are preserved above foundation level, while the more eroded parts retained only two or three courses (Figures 4.2, 4.3).↩#p155
Figure 4.2: Section drawing along east-west axis of B2.
We excavated all of the rooms to the uppermost floor level and reached additional floor levels in four rooms (rooms 5, 7, 9). In rooms 1, 2, 3, and 6 the clean foundation sand was reached, approximately 7–10 cm below the lowest preserved floors. A 1.75 m deep test trench was dug in room 2 in order to determine if there was any prior occupation in this area of the site. No evidence of older buildings was found here, and the matrixes were found to be largely sterile. This situation is similar to that found at Kellis, where Houses 1 and 2 in Area A were built directly upon the clay surface of the site or upon deliberately laid earth deposits with no traces of earlier occupation.3↩#p156
Because the walls of the structure were clearly visible from the surface, we divided the structure into the visible rooms for excavation. The measurements for the dimensions of the rooms indicate the area enclosed by the walls and do not include wall width. I provide a total depth of excavation for each room as well as upper and lower elevations for each depositional stratigraphic unit (DSU). The fixed point from which the elevations were calculated has an elevation of 138.4859 m asl. Dimensions for each DSU are given if they are smaller than the room interiors.↩#p157
Figure 4.3: Section drawing along north-south axis of B2.
Room 1 (3.74 m east–west x 2.62 m north–south)
Room 1 is located in the northeastern corner of B2 and is entered from room 7. Four walls,4 already visible from the surface, enclosed room 1, which was filled with stratigraphic units of varying types and reliability for a depth of 0.94 m. We thus proceeded with the single context method and used these walls to determine the parameters of the room (Figures 4.4, 4.6).↩#p158
Figure 4.4: Harris Matrix, room 1.
DSU 15 filled the entire room. It consisted of clean windblown sand covered with a dense lens of pottery sherds. Within the sandy matrix there were moderate to high densities of ceramics, clinker (sometimes identified as slag on forms and in the database), and plaster in addition to low densities of wood, bone, and glass. As the artifact densities lessened we changed DSUs, because the lower levels of windblown sand were less contaminated than the upper levels of windblown sand. We also suspected that lower levels of windblown sand belonged to a phase of building collapse and deflation and wished to capture this phase in more detail.↩#p159
DSU 126 consisted of clean windblown sand below the surface sand (DSU 1). The DSU change was done because we wanted to distinguish objects that may have derived from roof and ceiling collapse, including objects that may have been stored originally on the roof. There were low densities of artifacts, including ceramics, glass, plaster, bone, and clinker. This DSU was mostly confined to the southeast corner of room 1, while ceiling and wall collapse (DSU 4) occupied the rest of the room. The erosion of this stratigraphic unit may represent a deflationary process, and therefore the objects are more likely than the surface objects to have belonged to B2 itself and may have derived from the roof.↩#p160
DSU 47 consisted of mud-brick collapse from both the ceiling and the walls of room 1. Both vault bricks and wall bricks were found in the collapse. The vault bricks were evinced by the presence of markings on their faces as well as the way in which they were mortared together.8 The wall bricks from this room had strongly sooted plaster facing that matched one of the standing walls (F4) and probably derived from its collapse. Two complete vessels, a storage pot (Inv. 7977) and a flagon (Inv. 7652), were found within this matrix. Both vessels had mud plaster adhering to their bases. It is uncertain if this plaster was used to plaster these vessels into anything, namely in the roof that once existed above this room, or a wall niche, or if this plaster was simply the adherence of melted adobe upon the vessel bases. Also found within this matrix were several 75%–90% reconstructable vessels as well as moderate numbers of sherds. The ceramics appear to date to both the third and, especially, the fourth centuries CE. It must be borne in mind that this assemblage may be quite mixed due to the collapse of the roof. In addition to ceramics, moderate to high levels of clinker were found, particularly in the eastern half of the room. This clinker may derive from the ceramic kilns north of House B2. Small fragments of plaster were recovered especially in the vicinities of the walls F3 and F4. Some of the plaster was burned, while other fragments are black and possibly even red. We did not find colored plaster adhering to any of the walls.↩#p161
We separated the articulated vault bricks that we found in order to understand how the vaults in this house were constructed. We found a concentric curved line pattern9 and also the use of a triangular mud-brick and two chinking sherds between the rectangular vault bricks. Both the wedge shaped bricks and the rectangular vault bricks were sampled (S17 & S19).↩#p162
The collapse appears to have eroded to the same elevation as the top of the walls (Figure 4.5). This situation was most evident in the southwest quadrant of room 1. This erosion pattern signifies that the strong north wind was responsible for much of the erosion of this house. This wind appears to have eroded the entire northeast section of Amheida quite significantly, and the pattern of the finds also informs us that the majority of the deflation process took place after the walls of the structure had already collapsed.↩#p163
DSU 1610 lay underneath the ceiling and wall collapse (DSU 4). DSU 16 was the first secure context we encountered in room 1. It consisted of yellow sand with high densities of mud-brick debris that probably represented deterioration from the wall and ceiling as well as the erosion of collapse (DSU 4). Also included within this matrix were moderate to high densities of artifacts, particularly ceramics and clinker. Glass, wood, fauna, and plaster were also represented. We found a number of objects concentrated in the northeast corner of this context, including four Greek ostraka11, some of which are accounts of wine paid to various individuals. Some of these individuals were named while others were simply designated by their occupation. A beveled bronze ring (Inv. 3452) was also found, although it was heavily corroded and fragmented.↩#p164
Figure 4.5: Wall and ceiling collapse (DSU 4) within room 1. The wind has had a strong impact on the taphonomy of this area of the site, evinced by the even wear of collapse and standing walls in House B2.
DSU 2112 was found beneath the mud-brick rubble layer (DSU 16) and the sand layer (DSU 12). This context was secure only in the portion underneath the mud-brick rubble layer (DSU 16). It consisted of a brown sandy matrix containing a sandy loam as well as gray ash and yellow sand patches. There were fewer mud-brick inclusions in DSU 21 than the DSUs that covered it (DSU 12, DSU 16). Artifact densities were moderate in the categories of ceramic, clinker, faience, quartzite, and plaster. Organics, such as wood fragments, fauna, and charcoal, were represented in moderate to high densities. There were also two fragments of burned textile that were too poorly preserved to identify. It is possible that some of this material derives from post-occupational usage.↩#p165
DSU 2213 was also found underneath the mud-brick debris (DSU 16), primarily in the center of the room, and was therefore secure. It consisted of a compacted grayish-brown matrix immediately on top of floor surface F11. We identified both a high ash content as well as a high density of organics, including charred seeds and perhaps straw. Artifact densities were moderate and mostly consisted of poorly preserved potsherds and clinker. It yielded one complete vessel, a partially blackened lid (Inv. 8749) and two mud stoppers (Inv. 3521, Inv. 3532).↩#p166
Figure 4.6: Multi-context plan of room 1 showing DSUs (encircled), FSUs (within rectangles), elevations (with triangles), and small finds (without symbols).
F1114 was a floor represented by a compacted gray surface with numerous ashy pockets and inclusions covered by the occupational debris (DSU 22) (Figure 4.6, 4.7). The floor was only preserved in the middle of the room, under the ceiling and wall collapse (DSU 4) and is a secure context. The floor eroded away in the rest of the room in an erosion pattern commonly found in adobe structures; the floors do not survive as well at points of junction with the walls. Moderate artifact densities typified this matrix, although two jar stoppers (Inv. 3521, Inv. 3532; see DSU 22 above) and two coins (Inv. 3730 and Inv. 3485) were found directly on top of, and slightly embedded within, this surface. The pressure of the collapse probably pressed these artifacts into the floor surface, and therefore we can interpret these objects as potential intrusions from DSU 22 into the floor surface. Other finds included potsherds, bone, wood, and clinker. There was also a single gold glass bead (Inv. 3066). The top of the floor layer consisted of a smoothed surface, underneath which we found compacted debris. The division between surface and debris was difficult to distinguish with certainty after we excavated beneath the top 5 cm of F11. Therefore, the floor was given a great depth of 17 cm. The floor thickness was difficult to determine, because it was so poorly preserved and lay above compacted layers that could also have served as floor layers but that were less well prepared than the one that we identified initially. A comparison of this floor with the other floors from this house is presented in the next chapter. This comparison is particularly valuable in this instance because we were able to distinguish floors and floor preparation layers excavated in the 2006 field season, because these floors were better preserved than those discovered in the 2005 field season.↩#p167
Figure 4.7: Photo of F11, room 1.
DSU 3115 was an ashy matrix that was located centrally in the room as a large patch immediately below the floor (F11). DSU 31 is secure and may be interpreted as a dump of material used to level the site surface prior to floor construction. The sherd sizes from this matrix were quite large, suggesting that it was probably used as fill to stabilize the floor. These sherds may date to the second to third centuries CE, signifying re-use of material from an earlier occupation. DSU 31 contained high densities of organics, both charred and unburned, such as seeds, charcoal, ash, and wood fragments. Clinker and dross (waste matter) were also present. The high density of ceramics included a nearly complete vessel (Inv. 9201). Two burned mud bricks and several extremely dense pockets of light gray ash were also associated with DSU 31. This DSU was terminated when we encountered either a compacted sandy matrix (DSU 23) or a matrix of sand with yellow sand and ash patches (DSU 21).↩#p168
DSU 2316 consisted of a compacted sandy matrix that was possibly the result of trampling and/or water on the area. There were low to moderate levels of artifacts, including ceramics, wood, fauna, clinker, and dross. There was one intact, lightly burned flagon (Inv. 9232). Some of the ceramics appeared to share characteristics with second to third century forms, which may be related to DSU 24, below. There were a few lenses of ash that contained burned olive and date pits, as well as wood and charcoal. The majority of the matrix consisted of compacted yellow sand in an uneven formation above the clean windblown sand (DSU 24) that was used as the surface upon which the walls of room 1 were built. DSU 23 had a weathered appearance and as such resembled DSU 20 in room 2 (see below). It was clear that the layer occurred beneath floor level (F11), as did DSU 20 in room 2. It seems likely that this sand and debris was used for packing and stabilizing beneath the floor level. This compacted weathered surface only occurred in the central part of the room, nearly never abutting the walls. This wear pattern is consistent with the floor and wall use wear found in other rooms within this house.↩#p169
DSU 2417 consisted of clean windblown sand that appeared to form the base upon which the foundations of the walls rested. The matrix was also directly underneath the compacted yellow sand (DSU 23), the brown sandy matrix (DSU 21), and windblown sand (DSU 12). This matrix was excavated for only a few centimeters to insure that it marked the last anthropic material encountered in excavation for room 1. Some small quantities of material occurred in the top of the layer, as was found in the counterparts to this context in rooms 2 and 3 (DSU 26, 30). These top centimeters contained ceramics, wood, fauna, clinker, and dross. The ceramics shared characteristics with second to third century forms. The layer was devoid of cultural material deeper into the matrix. In room 2 this layer was at least 1.35 m thick and contained almost no artifacts.↩#p170
F5118 is the 78 cm wide doorway from room 7 to room 1. The north side was preserved to a greater height than the south side. The lower parts of the doorway, including the threshold, were all quite poorly preserved, and the threshold appeared heavily damaged. The sides of the doorway were the unmodified north face of wall F3 and south face of wall F4, with no sign of any jambs. There were possible sockets for wood consisting of undercuts at the bases of the west side of both walls. Immediately east of this socket was a mud-brick threshold consisting of two visible courses of bricks lying so that the stretchers were oriented east–west. The plaster of walls coved onto the upper surface of this mud brick threshold. By coving, I mean that the plaster between the floor and the wall covers over the wall plaster and creates a curve rather than a 90˚ angle between the floor and the wall. The mud plaster was brown and had a high organic content. None of the bricks was sufficiently defined to measure.↩#p171
F119 is an east–west oriented mud-brick wall along the north part of the room. It was highly eroded, particularly in the east, and was preserved only 80 cm above the foundation bricks in the west and 48 cm above the foundation bricks in the east. The south face of F1 was covered in straw-tempered grayish-brown mud plaster, which concealed the bonding pattern for this wall. The top face of the wall exhibited the alternating channels construction method (Figure 4.8). This construction method involved laying bricks on one side of the channel lengthwise along their headers, while bricks on the other side of the channel were placed with their stretchers lengthwise. Plaster filled the space left between the bricks. One complete channel was filled with poured mud mortar and measured approximately 80 cm long and 9 cm wide.↩#p172
Figure 4.8: Alternate channels construction method, with a channel of poured plaster located on one side or the other of the mud bricks.
There were three patches of white plaster still adhering to the wall face. The first was approximately 120 cm from the corner to F4, the second was 134 cm from F4 and was 35 cm in length, and the third was 180 cm from F4 and was 33 cm long. The height of the plaster was approximately 10 cm. The plaster was approximately 2 mm thick. This plaster may indicate the remains of a gypsum plaster strip painted on the walls, a common attribute of Dakhla houses.20 There is a 23 cm long gouge with a hole 87 cm east of F4. The diameter of the hole was 13 cm and it appeared to extend through the entirety of F1. This hole may be from a tie beam or a functional addition to the wall.↩#p173
The wall proper rested on a foundation that protruded approximately 18 cm from the face of the wall into the room. The foundation had five courses of mud brick, of which the bottom course consisted of stretchers placed vertically. The second course consisted of stretchers placed horizontally. The third consisted of headers, the fourth stretchers and the fifth headers. The bonding pattern is not regular. The top course was absent from the eastern portion of the wall for the last 65 cm. This portion was filled with rubble and compacted mud until the wall proper began.↩#p174
F221 is a north–south oriented wall forming the east wall of room 1. It was highly eroded with only approximately two or three courses preserved above the floor level (F11). The bonding pattern visible below the plaster was English bond. The top face of the wall exhibited the alternating channels construction method. The channels were 9 cm wide and 250 cm long. A brick that measured 16 x 6 cm was embedded within the northern channel.↩#p175
The western face of the wall is covered in mud plaster below approximately 20 cm from the top of the preserved wall. There was a portion of preserved white plaster 105 cm south of the corner with wall F1. This white plaster appeared to be partly covered with mud plaster. The white plaster was approximately 2 mm thick. The plaster was 10 cm high and was vertically located at approximately 10 cm above floor level and 10 cm from the top of the wall. The foundations of this wall were not exposed.↩#p176
F322 is an east–west oriented wall serving as the south wall of room 1. The top courses had mostly eroded away, particularly in the east. The wall is L-shaped in plan view, and the top of the wall showed the alternating channels method of construction. These channels were 7 to 8 cm wide. The brick-bonding pattern was English bond and the wall proper rested on top of vertically placed headers, which formed the foundation of F3. The foundation bricks were built at a higher elevation than the stretchers of wall F5. At, and just above, floor level, there was a shallow gouge visible, which was probably created through typical usage of the room. Approximately 5 cm above the foundation bricks there was a row of four roughly circular holes. Three of them were situated in the center of the wall. The fourth one was set to the east, approximately 80 cm from F5. The diameters varied between 10 and 15 cm. These holes may indicate the use of tie beams in the wall construction.23↩#p177
There was mud plaster adhering to the north face of this wall, which was 1.5 to 1.7 cm thick. It was straw tempered and grayish-brown in color. A portion of the mud plaster was covered in soot approximately 136 cm from the west end of the wall. Just west of the doorway the sooted patch was 20 cm high and 2 cm wide. There was a large hole 191 cm from the west edge of the doorway. It was 20 cm wide and 14 cm high. The wall around this hole appeared to have been repaired with mud plaster and mud-brick fragments. The hole appeared to go through to the south face of the wall. A second hole24 was visible 20 cm above the top of the foundation, began at the top of the preserved wall, and was visible 110 cm west of the corner with F2. The foundations were not fully excavated on the north face.↩#p178
F425 is a north–south oriented mud-brick wall that forms the west boundary of room 1. It was highly eroded, and it was preserved to greater height at the north end than south end, with a maximum of seven courses and a minimum of six courses preserved. The brick-bonding pattern was English bond. The top face of F4 showed the alternating channels method of construction. The channel was 10–12 cm wide and 125 cm long.↩#p179
Only 60 cm of wall were preserved above floor level, and there was mud plaster on this portion of the wall, which showed clear signs of replastering. The surface mud plaster was yellowish brown and straw tempered and covered a grayish plaster and possibly also a white plaster. There were patches of white gypsum plaster visible 36 cm south of F1. These white gypsum plaster patches measured approximately 9 cm x 11 cm and were 10 cm above floor level. These patches of plaster are the only remains of what probably would have been a strip of gypsum plaster painted along this wall. Such strips of white plaster are common in the oasis.26 The wall showed signs of burning 100 cm south of the corner with wall F1. This burned patch was 57 cm wide and 54 cm high. Both the mud-bricks and plaster were discolored to black, gray and purple hues and they were very sooted. This sooting may be attributed to post-occupational ephemeral usage.↩#p180
103 cm from the corner with F1 there was a large hole 24 cm wide and 16 cm high. It appeared to extend through the extent of F4. The low point of the hole rested at floor level. There was coved plaster in this area right above the floor, perhaps indicating a floor layer that was later than F11 and that was not preserved. Alternatively, floor F11 could have coved up to meet the walls in this room, as was common in other rooms of this structure. There was another hole visible where F4 meets F1, resting on top of the foundation bricks of F1. The diameter was 10 cm and appeared to extend for the extent of F4. The tops of the foundation bricks were just visible and appeared to stand vertically on their stretchers. This foundation protruded 7 cm from the base of the wall into the room.↩#p181
Interpretation of Room 1
Room 1 is a room of moderate size in B2. It is located close to a functional area within room 7, namely a bread oven and food storage area, which may have influenced the ways in which the occupants used room 1, although it does not show clear signs of intense food preparation activities. The room was moderately well preserved to a height of approximately 48–80 cm above the floor level.↩#p182
The entrance to the room was located along its western side and was 78 cm wide. The remains of the fittings for hanging a door were evinced by the remains of the threshold (F51). It is unclear if this door had any locking mechanisms, as the upper portions of the walls were not preserved, but it is clear that the room could be closed off. The door probably swung into room 7 when opened.↩#p183
The room appears to have been covered with a barrel-vaulted roof, as was evident from the articulated vault collapse (DSU 4) and the distinguishing features of vault bricks. This vaulted roof was probably covered to support a flat roof space that was used for storage, since we found at least one storage jar (Inv. 7977) with mud plaster adhering to its base. It seems likely that this jar was plastered into the roof of the house.↩#p184
The walls of this room were both mud plastered and gypsum plastered. It is likely that the remains of white gypsum plaster would have once been a gypsum strip along the north end of this room, as can be found in many houses in Dakhla.27 Within DSU 4 we found indications that this room may have had decorative red plaster in some locations, which also finds parallels in the Dakhla Oasis.28 There appear to be multiple episodes of replastering, indicating that the room was renovated somewhat over time. In addition to replastering there were minor repairs to the walls over the course of occupation. The periodicity of replastering for this region is unknown at this time. Some cultures replaster annually, as can be found in both present-day Egypt and the American Southwest, but other groups only replaster when it is deemed a necessity. The nearby settlement at Bashendi offers us regional ethnographic comparanda in which individuals re-plaster annually. The wall foundations were exposed in room 1 and appear to have been constructed on the site surface without substantial preparations.29↩#p185
The floor for room 1 was of the standard mud type and was well preserved only in the middle of the room. There are signs that it was repaired and replastered during the occupational history of this structure. There were no distinguishing characteristics associated with the floor itself that indicate differential usage of space in this room.↩#p186
During its primary occupational period, this room appears to have served for multiple uses, although we must eliminate cooking as a primary activity, as we found no signs of in-situ burning within occupational levels. Following the abandonment of this structure, the room seems to have had ephemeral usage, since a burn mark was found above floor level, and there are some ashy depositions (e.g. DSU 22). This evidence also suggests something of the nature of the decay of this structure. Presumably windblown sand came into the room, raising the surface, and then a minor burning episode took place, perhaps by squatters or passers-by. The wind deflation process of the walls occurred subsequent to some minor sand accumulation and this episode.↩#p187
Room 2 (3.3 m east–west x 3.00 m north–south)
Room 2 was located along the eastern side of B2, between rooms 1 and 3. It was entered from room 7 and provided access to room 3. Four walls30 enclosed room 2, which contained stratigraphic units of varying types and reliability for a depth of 3.93 m, including a sondage. The stratigraphy of this room initially appeared to be more complex than the other rooms in this structure, but post-excavation analysis has not corroborated this complexity (Figures 4.9, 4.11).↩#p188
Figure 4.9: Harris Matrix, room 2.
DSU 231 consisted of windblown sand filling the entire room. 47.30 kg of iron-rich fabric A sherds were found, most of which were concentrated directly on the surface or in the upper 10 cm of the unit. These sherds present a heterogenous dating profile. Morphologically, the ceramics include storage jars, pots, and open forms such as plates and bowls. In addition to sherds, we also found moderate densities of glass, wood, clinker, faience, plaster, and glass slag. Some burned olive and date pits were also found. This deposition was quite shallow in the northwest corner, where the western door to room 7 (F52) was situated, and where it covered debris from wall and roof collapse below (DSU 6). On the other sides of the room, the unit was much thicker, as there was no collapse debris here. A complete bowl was found in the southwest corner of this room right next to the wall (Inv. 7978) at approximately 30 cm from the door to room 3. A faience amulet representing the Egyptian god Bes was also found (Inv. 3233). These objects cannot be connected directly to the B2 occupation, but may inform us about the vicinity around B2 in Area 1. An arbitrary division was made at a depth of approximately 50 cm in order to distinguish this surface sand from the windblown sand beneath the surface (DSU 5).↩#p189
DSU 5 (= DSU 9)32 consisted of windblown sand next to DSU 6. As in DSU 2, high densities of sherds, pieces of glass, bone, wood, and pits from olives and dates were found. Likewise, the ceramics present a heterogeous dating profile. Pieces of charcoal were found distributed unevenly across the matrix. It had the same characteristics as DSU 2, but towards the bottom of the unit, more brown sand was mixed with it. This DSU may equal DSU 9.↩#p190
DSU 9 (=DSU 5)33 consisted of yellow sand above DSU 15 in the southeast corner. This unit was given a different unit number because, in this corner, there were many large fabric group A sherds, although the ceramics do not present a different dating or morphological profile from those in DSU 2 or DSU 5. Moreover, it was somewhat browner than the yellow sand of DSU 5. Despite these minor differences, it is possible that DSU 5 resulted from the same depositional episode as DSU 9.↩#p191
DSU 634 was below windblown sand (DSU 2). It was composed of wall and roof collapse and consisted of many mud-brick fragments and only a few complete bricks. Some parts of vault bricks were amidst this rubble, although they were mostly incomplete. The small surviving collapse had eroded to a high degree and was found only in the northwest corner of the room. Within the collapse we found many fragments of iron-rich A group ceramics as well as A11, B, and P37, which may date to the late third century. We also found glass, wood, bones, clinker, and a few fragments of plaster.↩#p192
DSU 735 (Figure 4.10) was found below sand layers (DSU 5, 8, 9) and ash (DSU 10). It consisted of a brown loamy sand layer, which was somewhat harder than the windblown sand. It was quite shallow and was not found in the entire room. This context was found only in the western part of the room and two deposits were visible with diffuse edges, one in the central-west part of the room and one next to it. These deposits were located in and next to the threshold of the door to room 3 (F58), where it is the same type of deposit as DSU 18 in room 3. This unit may be interpreted as the remnants of a floor, perhaps mixed with occupational debris. That this deposit was situated just above or at the same height as the foundation bricks in the north and south walls may sustain the theory that this matrix was once a floor or occupational level. The deposit lacked the strength and hardness characteristic of true floors, but it was also found above the threshold and rested directly on the threshold bricks. Artifact densities were high and we found iron-rich A group and A4 ceramics in addition to bone, wood, charcoal, plaster, and olive and date pits. We also found a Greek ostrakon (Inv. 3414) that was awkwardly written and may have been a writing exercise.↩#p193
DSU 8 (=DSUs 11, 15, 17)36 was found below windblown sand (DSU 5). It consisted of a grayish brown sand layer in the northeast quadrant of the room and at a depth deeper than where the possible floor level (DSU 7) may be presumed to have been. Although there was a (slight) difference in color, it is very well possible that this DSU was of the same nature as DSUs 11, 15 and 17. DSU 8 and 15 may be the same deposit, in which case there may have been an arbitrary boundary between these two contexts or the contexts may have been mixed. For the larger part of the room, no sealing floor layers were found. There were large amounts of iron-rich fabric A group as well as smaller amounts of faience, bones, and clinker. The ceramics include a fairly large number of storage jars, a flanged bowl, cooking pot fragments, and a fire dog.↩#p194
DSU 11 (=DSU 8, 15, 17)37 was underneath the possible floor level (DSU 7), the wall/ceiling collapse (DSU 6), and the sand layer (DSU 8). It consisted of a gray brown layer of loamy sand with ash and organic material. It was probably the remains of occupational debris at floor level or a rubbish deposition to create a level surface for the floor. It was a quite small patch in the central-west part of the room. It was approximately at the same height as the top of the foundation bricks. It mainly consisted of sand patches mixed with loamy sand and clear ash pockets. Glass, clinker, and plaster were found here. The organics consisted of bone, seeds, and coprolites. The coprolites suggest a possible connection to Courtyard C2, which will be explored in more detail when the excavations of C2 are complete.↩#p195
DSU 15 (=DSU 8, 11, 17)38 consisted of yellow and gray sand in room two. The division between DSU 8 and DSU 15 was arbitrary. DSU 15 was probably a mixed layer. It is uncertain if this layer was contaminated or not. The layer occurred approximately at the level of foundation bricks. For the most part, no floor levels were found sealing it. DSU 15 covered a cleaner sand layer at the foundation level (DSU 26). As in the other layers in room 2, DSU 15 also contained organics. Among other materials it contained some bones, charred wood, and coprolites. Iron-rich fabric A group ceramics occurred in great densities, and A4 and P37 fabrics were also represented. Clinker was also represented.↩#p196
DSU 17 (=DSU 8, 11, 15)39 was a gray brown layer of loamy sand with ash and organic material. This matrix probably signified occupational debris at floor level or rubbish under floor level (DSU 7). It was located at approximately at the same depth as the top of the foundation bricks. DSU 17 was a quite small patch in the central west part of the room. It consisted mostly of sand mixed with patches of more loamy sand, and clear ash pockets could be defined. The organics consisted of bone, four date pits, an apricot pit, five olive pits, and a fragment of rope. Iron-rich fabric A and A4 ceramics were represented. The calcium-rich fabric B and P37 ceramics, glass, and clinker were represented in small quantities.↩#p197
Figure 4.10: Photo of DSU 7, room 2.
DSU 1040 consisted of an ash pocket. It was located very close to the debris (DSU 6) in the northwest corner. It is doubtful that this layer is reliable. Because the boundaries of this unit were not distinct, it is difficult to say whether this deposit cut windblown sand (DSU 5). We found large quantities of fabric groups A and B as well as A4. Clinker was also represented in small quantities, and there was a single apricot pit.↩#p198
Figure 4.11: Multi-context plan of room 2 showing DSUs (encircled), FSUs (within rectangles), elevations (with triangles), and small finds (without symbols).
DSU 2041 was located below sand layers (DSU 8, DSU 9). It consisted of brown to brownish-yellow sand with pockets of more loamy sand. It was found only in the center of the room. It had a weathered surface, which was probably due to contact with water and wind. As such, the top of the layer had a rugged appearance. This context had the same appearance as DSU 23 in room 3, and it appeared to be a compacted surface. The layer was 20 cm thick at its maximum extent and contained a dense concentration of ceramics, all of which seem to be datable to the Roman Period and possibly the third century. One patch was particularly rich in artifacts; it contained two complete pots (Inv. 7983, Inv. 8743), iron-rich A group and A4 sherds, and three ostraka written in Demotic.42 Among organic remains, four olive pits, one apricot pit, wood, and bones were found. Clinker, glass slag, and plaster were also represented. DSU 20 seemed to be an undisturbed context that occurred at foundation level. We did not find clearly defined floor layers sealing this floor preparation context, but we found it partly under a layer of compacted loamy sand (DSU 7) that occurred approximately at the level where one would expect a floor. DSU 20 covered the clean windblown sand (DSU 26), upon which the foundations of room 2 seem to have been set.↩#p199
DSU 2643 consisted of yellow to brown-yellow windblown sand. The sand was finely laminated and deposited in various orientations. The top of the layer was still contaminated with various materials in multiple categories, such as organics and some pottery. Although the difference is not easily quantifiable, one could discern that the top of this layer was somewhat disturbed. DSU 26 had a depth of approximately 1.35 m under foundation level and covers DSU 32, a more compacted yellowish brown to reddish sand with chalk particles. The house foundations were set on or in DSU 26, and therefore they rest on windblown sand.↩#p200
DSU 3244 was a 1 x 1 m sondage composed of strongly compacted sand that had to be broken by pick. In between the harder sand layers there were some medium hard lenses that did not have to be broken by a pick but which could be troweled. This DSU was excavated to a depth of approximately 1.75 m under the foundations wall courses of room 2 in an exploratory cut in the southeast corner of the room. The layer contained no diagnostic pottery fragments, but various over-fired dynastic sherds were found in these compacted sand layers. The pottery deriving from this DSU probably does not signify occupation at this particular location, but it may be interpreted as evidence for a much earlier occupation nearby. The pottery that was found was iron-rich fabric A4 and was characterized by a polished red slip. The body sherds were from a closed rounded/pointed based vessel. The over-fired technology suggests a dynastic occupation, athough the precise dating is not known.45 There was one unidentified ring/recessed base fragment that may derive from an open form. There was one body sherd of polished red ware with a slip that may be dynastic as well.↩#p201
F346 is an east–west oriented wall. It is the north wall of room 2 and the south wall of room 1. The top courses have mostly eroded away, particularly in the east. The south face of the wall showed traces of soot on top of mud plaster, which was largely eroded, but a few patches were visible. These patches showed reddish-brown plaster with straw tempering. The bricks were grayish-brown with straw inclusions and were highly eroded.↩#p202
F547 is a north–south oriented wall of which the top courses have mostly eroded away, and it was totally eroded to floor level in the south. The wall had a clear man-made gouge approximately 15 cm high at its thickest extent and 215 cm long. Two holes were visible in the gouge. They were set approximately in the center of the wall. Their diameters were approximately 12 cm and they extended to a depth of 29 cm (measured from the complete bricks at former floor level). The facing of mud plaster was highly eroded in the northeast corner of room 2. Elsewhere the organics in the plaster and bricks were well preserved. The brick-bonding pattern was visible above the clear foundation bricks and was English bond. Underneath the six courses of English bond, there was a row of foundation bricks that protruded from the wall by approximately 10 cm. Underneath this course, the stretchers were placed vertically and formed the bottom row of bricks. There were gaps of less than 0.5 cm to 1.5 cm between these bricks and no mortar was visible. The bricks were set on finely laminated windblown sand. The plan-view of the wall shows that the wall was built of both headers and stretchers using the alternating channels method of construction. The northern channel was towards the west and it measured 2 m long and approximately 9 cm wide. The southern channel was 8 cm wide and extended into F9.↩#p203
F5248 is the doorway from room 7 to room 2, consisting of the lower part of the sides of the doorway and the threshold. The vertical elements were in fair condition aside from a gouge from the socket for a wooden threshold (now lost). The upper course of the brick threshold was highly eroded. The doorway consisted of a north face of seven courses (wall F7) and a south face of five courses (wall F3), both of which were coated in mud plaster with high organic content. As with doorway F51, the lowest two courses of the west sides of the walls were undercut; presumably this was a space for a socket for a wooden threshold that was removed or eroded. Traces of wood remained in the top of and inside the socket. The mud-brick threshold is bonded with the sidewalls and consisted of two visible courses of mud brick in an English bond. The upper course was laid with north–south headers, the lower with north–south stretchers. The only remaining traces on the north and south sides were of the upper course. The north side of the doorway was preserved to six courses and the south to five courses. On the north side of the doorway with room two were the remains of a feature filling in the corner between the doorway and the north wall. This feature consisted of parts of two courses of brick (x=30, y=20, z=14) bonded into uppermost preserved courses of north wall. The lack of a lower course was probably due to an accident of preservation, as a gouge is visible on the bond with the doorway. The south face of this doorway and this feature had continuous mud plaster.↩#p204
F649 is an east–west oriented wall forming the south boundary to room 2 and the north boundary to room 3. It had only three to four courses remaining. Right above floor level there was an extensive gouge and other wear, which was due to heavy usage of the room. The gouge was repaired in at least one episode, possibly several times. Several pieces of baked brick and stone were inserted into the gouge and the gouge also showed a clear repair face where mud mortar was used. The gouge was once greater than 20 cm wide, and it runs for at least 220 cm before it reaches the threshold of the doorway between room 2 and room 3 (F58). The bricks were straw tempered and followed an English bond pattern and were set on one course of stretchers protruding 14 cm. The wall rested on top of foundations that had their headers placed vertically. The foundation bricks were not mortared together, and they ran all the way to the west where they met and bonded with wall F7. Here, however, they lay under a threshold (F58) for the doorway between room 2 and room 3.↩#p205
An erosion gouge, as seen on the north face of wall F6, continued into the doorway (F58). The west face of F6 (east wall of doorway) was filled with clinker used for repair purposes. The top of the wall exhibited the alternating channel method of construction. The channels were approximately 7 cm wide. The south channel was clearly visible and was approximately 180 cm long. The south face was covered with reddish-brown mud plaster. The foundation bricks on the south side of F6 consisted of one course of stretchers protruding approximately 3 cm from the wall proper. Underneath those stretchers was one course of vertically placed headers. The stretchers running into F7 bonded with the foundation level. The foundation bricks appeared to be set on clean windblown sand continuing to the bricks of F9 and F5, which seem to have been set in a foundation trench. This feature abutted walls F9 and F5.↩#p206
F5850 This feature serves as a doorway between room 2 and room 3 and is approximately 60 cm wide. It is located at the west end of the south wall (F6). The threshold was made up of two courses of bricks laid on top of the foundations for the walls. The bonding pattern of these bricks was the same as wall F6. Indeed, this threshold was built as part of wall F6 in a single episode. The threshold itself was approximately 60 cm wide next to wall F7. The inside of the eastern portion of the doorway was covered with reddish-brown, straw-tempered mud plaster. In the middle of the wall there was a threshold hole. At the level of the threshold of the doorway between rooms 2 and 3, there was a large hole where the threshold beam once existed.↩#p207
F751 is a wall that forms the west boundary of rooms 2 and 3. The wall has an English bond and the plan view of the wall shows that it was built using the alternating channels method. The channels varied in length from 90 cm to more than 180 cm and the width of channels was approximately 7 cm. The bottom of the wall in room 2 had a gouge approximately at floor level, probably due to erosion from heavy usage. The gouge was approximately 1 meter long, from the doorway inwards.↩#p208
The wall F7 bonded to wall F6 in the foundations. The east face of F7 was also the west face of the doorway between rooms 2 and 3. Above the foundation layer proper in room 2, a maximum of eight courses of brick were preserved. The foundation bricks consisted of one course of vertically placed headers without mortar. The foundation bricks were gray-brown in color and were set on windblown sand. This foundation pattern changed approximately 20 cm in before the doorway into room 3. There, two bricks were placed horizontally, a header above a stretcher. The wall, part of which forms the west wall of room 3, had the same construction and appearance as F7 in room 2, but the foundation bricks were placed differently and followed the pattern already started on the south side of room 2. These foundation bricks were formed by one course of stretchers protruding approximately 4–5 cm from the wall proper and underneath that, one course of headers placed vertically. The brick dimensions were the same as foundation bricks in the other part of the wall. The bottom course of the foundation bricks in wall F7 was set on a course of horizontally placed headers from the foundation of F8 in its most eastern extent. The foundations of F7 and F8 thus partly overlapped.↩#p209
F5252 is the doorway from room 7 to room 2, consisting of the lower part of the sides of the doorway and the threshold. This was a simple doorway consisting of the north face of F7 (5 courses) and the south face of F3 (6 courses), both coated in mud plaster with high organic content. The doorway was approximately 80 cm wide, and 50 cm of the vertical component remained. The vertical elements were in fair condition apart from a scar left from the removal or deterioration of wood that once fitted a socket in the threshold. The upper course of brick for the threshold was virtually lost. Both the north face that was in the doorway and the long east face of the wall were plastered with reddish brown straw and sherd-tempered mud. As with doorway F51, the lowest 2 courses of the west sides of the walls were undercut, presumably for the placement of a socket for a wooden threshold, now lost. Traces of wood remained in the top of and side of the socket. The mud-brick threshold consisted of two visible courses of mud brick. The upper course was laid with headers running north–south, and the lower course with stretchers running north–south in a truncated English bond pattern. Only traces on the north and south sides remained of the upper course. The door probably opened into room 7.↩#p210
Interpretation of Room 2
Room 2 was the middle room within the eastern section of B2. It provided access to room 3 and thus was the only room in B2 aside from room 7 to serve as both a room and a hallway. This room was less well preserved than room 1, particularly in the poor preservation of the floor, and therefore the reconstruction of the architecture and the activities of this room are less secure.↩#p211
The entrance from room 7 into room 2 was located along the west side of room 2 and was 80 cm wide. The entrance between room 7 and room 2 contained a door, as evinced by the threshold (F52). The door probably opened into room 7. The entrance between room 2 and room 3 was in the southwestern portion of room 2. The entrance between room 2 and room 3 also contained a door, as evinced by the threshold (F58). This entrance from room 2 to room 3 measured 60 cm wide. It is unclear if these doors locked, as the walls were not preserved to a sufficient height to determine if there was a locking mechanism.↩#p212
Room 2 was once covered with a barrel-vaulted roof, as evinced by the collapse found here (DSU 6). There is no clear evidence within the collapse to indicate ways in which the roof on top of this vault might have been used.↩#p213
The walls of this room were mud plastered and show signs of repair. There were no signs that either gypsum or painted plaster was used within this room. The wall foundations exposed suggest that they were built upon the site surface.53↩#p214
The floor construction, although minimally preserved, appears to be consistent with what we found preserved in other rooms. The sand site surface appears to have been wetted with water and trampled to harden it, followed by fill consisting of debris, and then a smoothed plaster surface. Unfortunately, only the preparation layers, not the floors, were preserved.↩#p215
This room reveals that there were no structures underneath B2 but that there was probably a dynastic settlement nearby, since the few sherds found below occupational levels appear to be dynastic. Coprolites found below floor level may indicate animal husbandry or stabling in the vicinity of B2 prior to its construction. Little can be said about the occupational usage of room 2, although the lack of functional devices in this room makes it clear that dedicated activities such as cooking did not take place extensively in this location.↩#p216
Room 3 (3.80 m east–west x 2.67 m north–south)
Room 3 is located in the southeast corner of B2. It was accessed by means of room 2 and was the only room in the structure that was not accessed by room 7, with the exception of storage areas. Four walls,54 already visible from the surface, enclosed room 3, which was filled with stratigraphic units of varying types and reliability for a depth of 1.36 m (Figures 4.12, 4.14).↩#p217
DSU 355 consisted of windblown sand on the entire surface within room 3. There was a dense layer of iron-rich fabric A pottery sherds in the top few centimeters. Other inclusions found were charcoal and plaster fragments, glass, wood, and animal bones as well as some organic remains such as olive and date pits. Collapse (DSU 13) quickly became visible within the windblown sand, which somewhat interpenetrated the collapse.↩#p218
DSU 1356 consisted of mud-brick collapse made of reddish brown wall and vaulted ceiling collapse. The ceiling collapse was discernible from the concentric pattern on the bricks consistent with vault brick construction.57 This context was very incomplete and had eroded to the level of surface sand and walls; it was also mixed with sand (DSU 3). It is significant that many of the bricks were very thin (approximately 4–5 cm thick), which was probably the result of wind erosion. The organics in the bricks were still extant, as were the many other organic inclusions in the collapse, such as olive and date pits. The inner surface of the vault was heavily sooted but not burned, suggesting that oil lamps may have been used in the room. Within the collapse were found iron-rich fabric A and A4 fragments as well as calcium-rich fabric B and P37. These ceramics, particularly as indicated by a lamp fragment, probably do not date to earlier than the beginning of the fourth century. In addition to ceramics we found glass, wood, bones, and plaster in low to moderate quantities. There was also one small 0.010 kg fragment of a textile,58 three beads,59 and an unfired clay stopper (Inv. 3517). Most of these objects derived from the lowest elevations in DSU 13 and therefore probably ought to belong to DSU 14.↩#p219
Figure 4.12: Harris Matrix, room 3.
DSU 1460 was a brownish-yellow sand layer located directly beneath collapse DSU 13 in the eastern half of the room and above a yellow-brown sand layer (DSU 29) in the western half of the room. It contained a high density of organic material, both charred and un-charred, such as wood, bones, and pits from olives, dates, and apricots. There were significant quantities of iron-rich fabric A and A4 as well as moderate quantities of calcium-rich fabric B and P37. These ceramics may date to the third and, particularly, the fourth century. Low densities of clinker and an unfired clay loom weight (Inv. 3534) were also found. This deposition may have extended below floor level (F10), which was not found in the eastern half of the room. It was not found in the southwest quadrant where the floor (F10) was still extant. This brown sandy layer was mixed with ash and silty debris and may represent occupational debris mixed with destroyed floor (F10) and windblown sand (DSU 3). It was directly on top of sand (DSU 29) that lay underneath the floor (F10) as well as floor preparation layers (DSUs 25 and 28). DSU 14, as a mixed context, is probably the result of the roof and wall collapse (DSU 13), which caused all of these layers to interpenetrate one another.↩#p220
DSU 1961 was a brown loamy sand layer in the southwest corner of room 3 above occupational debris (DSU 18) (Figure 4.13). This matrix was similar to the occupational debris under it (DSU 18), but this context had much higher concentrations of organic material. It was filled with a large concentration of rodent coprolites and a dense concentration of other organics. We took many samples of this matrix in order to discern the nature of these organics.62 Within this layer we also found three intact pots63 and a jar stopper (Inv. 3518). Calcium-rich fabric A group ceramics were represented in moderation, as were calcium-rich fabric B and P37. These ceramics share characteristics with third- and particularly fourth-century forms. Glass and two polishing stones (Inv. 3741, Inv. 3733) were also found. Among the organic remains, high densities of olive, date, and apricot pits were found, as well as a piece of wood with two small holes from domestic furniture (Inv. 3523).↩#p221
Figure 4.13: Photo of DSU 19, room 3.
DSU 1864 consisted of brown loamy sand with high densities of organic remains. It may represent occupational debris above floor F10. It was similar to DSU 7 in room 2. The deposition was visible in two places with diffuse boundaries: in the threshold of the door between rooms 2 and 3 and in the southwest quadrant above a floor level and below the brown loamy sand layer (DSU 19). DSU 18 was probably not itself a floor, but it was a layer that occurred at floor level. In some areas it lacked the strength and hardness characteristic of true floors, as for instance F10 (which lies directly beneath DSU 18). Three small, burned pieces of a textile were found lying flat on top of DSU 18 in the section in and near the doorway threshold.65 There were large quantities of iron-rich fabric A group and A4 as well as calcium-rich fabric A11, B, and P37. These ceramics share characteristics with third and particularly fourth century forms. Glass, wood, and bones were also represented. A bead (Inv. 3493), an apricot pit, and olive and date pits were also found.↩#p222
F1066 was a small patch of compacted gray brown mud floor along the west wall and in the southwest quadrant of room 3. This mud floor was thin and varied between 1 and 2 cm thick. The mud was tempered with straw and other organic material, such as twigs. Although this floor level was made of compacted mud it was not a very hard surface, but it was the best-preserved floor level found in room 3, and in some respects it was more distinct than the floor visible in room 1. Erosion may have played a large part in the clearing of anthropogenic levels in parts of the room. The part of the floor that was preserved best was covered by the largest part of the collapse and was therefore a secure context. The lack of collapse over other parts of the floor may have left it open to the elements and caused erosion. Another possibility is that the pressures of the vault may have destroyed the rest of the floor, as could be seen in DSU 14, which was quite mixed due to the vault collapse penetrating through multiple stratigraphic units. Clay stoppers (Inv. 3180, Inv. 3525) were found embedded into this context, likely due to pressure from the vault pushing DSU 18 into F10.↩#p223
DSU 2567 was a yellow-brown sand layer under occupational debris (DSU 18) in the southwest quadrant of the room and directly under the floor (F10). It may represent a preparation layer for floor F10. It contained numerous sherds and organic remains. The sherds derived from iron-rich fabric A group and calcium-rich fabric B and P37 ceramics. Faience, glass, and dross were also represented. Among the organics, wood, bones, and burned olive and date pits were found.↩#p224
DSU 2868 was a gray-brown weathered surface with some salt crystals and inclusions of pottery, charcoal and gypsum bits as well as organic remains. It was underneath the yellow brown sand layer (DSU 25) in the southwest quadrant of room 3 and on top of another yellow brown sand layer (DSU 29). Several patches of this weathered surface were found elsewhere in the room; these probably belong to DSU 28, but since they did not connect with the largest patch in the southwest corner, we could not be certain. The salt crystals and weathering resulted from water. The layer was at roughly the same level as foundation bricks along interior of south wall. DSU 28 likely represents a preliminary preparation layer for floors, as we found in rooms 1 and 2 (DSUs 23 and 20).69 Among the ceramics, iron-rich fabric A and A4 were represented in moderate to large quantities while calcium-rich fabric B and P37 were represented in low to moderate quantities. These ceramics appear to share characteristics of ceramics dating from the second to third centuries CE. We took samples of the matrix (Inv. 189 = S15).↩#p225
DSU 2970 was a yellow-brown sand layer under DSU 28 in the southwest quadrant of the room. It was under DSU 14 in the rest of the room. It was at the same elevation as the foundation bricks in the surrounding walls. A vault brick with concentric curved patterns was found molded into the surface (the same as those found in the collapse), which suggests that the collapse of the ceiling caused the mixing of layers found in DSU 14. Among the ceramics, the iron-rich fabric A and A4 groups were represented in large quantities, and the calcium-rich P37 was also represented. The ceramic material from this layer is characteristic of the second to third centuries, suggesting pre-occupational placement. Faience, glass, wood, bones, and clinker were found in low quantities. Two olive pits and one date pit were found. A bronze ring (Inv. 3484) was found in the northeast corner of the room and a small terracotta figurine fragment (Inv. 3519) was also found.↩#p226
DSU 3071 consisted of yellow sand mixed in with some ash, which was especially evident in the east corner of the room. There was mud brick debris in the top few centimeters, but the matrix became cleaner with depth. This context is similar to DSU 26 in room 2, which also contained a lens of debris that gave way to cleaner sand below. We excavated the layer to the bottom first course of the foundation bricks in the southeast corner of the room. A large pottery sherd, likely from a large cooking vessel, was found towards the bottom of the foundation layer. This sherd appears similar to fragments of a similar vessel found in the layer above (DSU 29) and may suggest a mixing of the layers in the upper centimeters of DSU 30. This context therefore consisted of contaminated windblown sand in the upper centimeters as well as clean windblown sand below. The ceramics included iron-rich fabric A and A4 as well as calcium-rich fabric P37. These ceramic materials indicate a second to third century date. Glass, wood and bones were also represented in low to moderate quantities, and organics, such as olive and date pits, were also found.↩#p227
Figure 4.14: Multi-context Plan of room 2 showing DSUs (encircled), FSUs (within rectangles), elevations (with triangles), and small finds (without symbols).
F672 is an east–west oriented wall with only 3–4 courses remaining. It forms the north boundary of room 3. The north face of the wall was described above as part of room 2.↩#p228
The top of the wall exhibited the alternating channels method of construction. There were alternating headers and stretchers with 7 cm wide channels in between. The south channel was complete and extended for 180 cm. The south face of the wall was covered with reddish-brown mud plaster, but the brick-laying pattern was still visible as English bond. Immediately above floor level there was an extensive gouge and other wear that indicates heavy usage of the room. The gouge was repaired once or several times, as indicated by several pieces of baked brick and stone that were inserted into the gouge during occupation. The wall showed a clear repair face where mud mortar was used to cover this stabilizing fill. Prior to its repair in antiquity, the gouge was once greater than 20 cm wide and extended at least 220 cm before it reached the threshold at an approximate depth of 137.20. The wall was set upon one course of stretchers protruding 14 cm from the wall proper. These foundation bricks were not mortared together. The foundation bricks ran all the way to the west, where they met, and were bonded with, wall F7. Here, however, they lay under a threshold for the doorway F58 between room 2 and room 3. The foundation bricks on the south side of F6 consisted of one course of stretchers protruding approximately 3 cm from the wall proper. Underneath those stretchers was one course of headers placed vertically. The foundation bricks appeared to be set on clean windblown sand continuing to the bricks of F9 and F5, which seem to have been set in a foundation trench. The stretchers of F6 extended into the wall F7 and bond F6 to F7. This feature abutted wall F9 and wall F5.↩#p229
F773 is a north–south oriented wall that forms the west side of rooms 2 and 3. Above the foundation level in room 2, there were a maximum of 8 courses of brick remaining. The bonding pattern was English bond and the alternating channels construction method was used. The channels varied in length from 90 cm to more than 180 cm. The width of channels was approximately 7 cm. The bottom of the wall had a 1 m long gouge located approximately at floor level extending from the doorway into the room. This gouge probably resulted from erosion.↩#p230
The wall had the same build up and appearance as the north face of the wall in room 2, but the foundation bricks were placed differently in order to follow a pattern already started in the south side of room 2. Here, the foundation bricks were formed by one course of stretchers protruding approximately 4–5 cm into the room and, underneath that, one course of headers placed vertically. The brick dimensions were the same as foundation bricks in other part of the wall. The bottom course of foundation bricks in wall F7 is set on a course of horizontally placed headers from the foundation of F8 in the most eastern end of the wall. The foundations of F7 and F8 thus partly overlapped. The foundation bricks consisted of one course of vertically placed headers without mortar. The foundation bricks were gray-brown in color and set on windblown sand. This foundation pattern changed approximately 20 cm before the doorway into room 3 (F58). There, two bricks were placed horizontally, a header above a stretcher. The east face of F7 was also the west face of the doorway between rooms 2 and 3 (F58). At the level of the threshold of the doorway between rooms 2 and 3, there was a large hole approximately 15 x 20 cm with a depth of 15 cm. This hole could have held a threshold beam. The threshold was approximately 60 cm wide next to F7. The wall F7 seems to be bonded to F6 at foundation level.↩#p231
F58 is the doorway between room 2 and room 3 and was described above as part of room 2. ↩#p232
F874 is an east–west oriented wall running forming the south boundary to room 3 and the south wall of the house. It was badly eroded, particularly to the east, where only three courses of bricks remained, as opposed to the western side, where seven courses were visible above the foundations. The bonding pattern was English bond. The top of the wall was very badly eroded, but it was clear that the alternating channel method was used. The mud plaster facing was approximately 1 cm thick, but visible only in the top west corner of the north face. Starting approximately 75 cm from the west corner of the wall (moving east), there were three holes in the wall with diameters of approximately 7–10 cm and a depth of 10–25 cm (at least). The third hole terminated at 115 cm from the west end of the wall. The foundations were made up of seven courses of brick with a total height of 60 cm. The top course was built with headers placed horizontally and which protrude 30 cm from the wall. The entire foundation was an English bond, but its protrusion of 30 cm distinguished it from the rest of the wall. The tops of the foundation bricks were particularly weathered, appeared melted, and had accumulations of salt particles. The foundation bricks of wall F7 rested on the top of the foundation bricks of wall F8.↩#p233
F975 is a north–south oriented wall that forms the east side of room 3. It extended five courses above the foundation bricks. It eroded to a greater extent to the south than to the north and it eroded to, or even below, floor level in the southeast corner. No mud plaster was currently visible on the wall, although a mud plaster facing would have been likely. The brick-bonding pattern was visible as English bond and the alternating channel method was used. The stretchers of the foundation layer protruded 4 cm into the room. These foundation bricks consisted of one course of stretchers, followed by vertically placed headers, placed next to each other with no mortar used. Underneath this course was another course of headers laid horizontally. All of the bricks were straw-tempered and grayish-brown in color, the bottom course being marginally grayer than the rest of the bricks.↩#p234
Interpretation of Room 3
Room 3 was of slightly smaller dimensions than rooms 1 and 2. Room 3 was the most private room in the house, because it was entered from room 2 rather than from room 7, which provided access to all other rooms in this house. The walls were not well preserved, particularly in the southeast, but the floor and many of the objects retained a moderate to high level of preservation.↩#p235
The entrance to room 3 was located in on the west side of the north wall and was 60 cm wide. This entrance had a door as evinced by the preserved threshold (F58). It is not possible to determine if this door could be locked, as the walls were not preserved to a sufficient height to discern a locking mechanism.↩#p236
Originally, room 3 was covered with a barrel-vaulted roof, as indicated by the collapse (DSU 13) found in this room. We did not find any artifacts that definitively appeared to have once resided on a flat roof on top of the barrel vault, although some of the surface pottery fragments may belong to this category of material. There were signs of soot on the inner surfaces of collapsed vault bricks, suggesting the use of oil lamps or other sources of smoke (DSU 13).↩#p237
The walls were mud plastered. There were no signs of any gypsum or painted plaster in this room. There were some signs of re-plastering, although the walls were not preserved to a sufficient height to definitively determine the plastering of these walls. The foundations of the walls appear consistent with other rooms in that they rested upon the site surface.↩#p238
One floor layer was preserved, and the room shows the same floor construction pattern found in the previous two rooms: mud plaster over a matrix of mud brick rubble and trash, which was placed directly upon the windblown sand.↩#p239
This room showed the most signs of erosion within this structure. The wind was particularly strong along the southeast corner of the structure as this portion of B2 was not protected by the large kiln structure to the northwest. It is also clear that the roof collapse served as the greatest form of protection for floor and occupational layers. Despite the poor preservation of the walls it appears that some of the objects were in situ, namely an assemblage of jars76 from DSU 9 and DSU 10 above the floor (F10). The occupational usage of this room is not evident, although the presence of rodent coprolites within the aforementioned jars suggests that some food was stored within the vicinity. Other functional objects found here may indicate that it was used for minor functional activities.↩#p240
Room 4 (2.73 m north–south x 0.71 m east–west)
Three walls,77 already visible from the surface, enclosed room 4, which was filled with stratigraphic units of varying types and reliability for a depth of 0.80 m (Figures 4.15, 4.16, 4.17).↩#p241
Figure 4.15: Photo, view into room 4 from room 5.
Room 4 is located along the middle of the western boundary wall of B2. The walls were clearly visible from the surface. The remains of a vault still in place were clear in the northern extent of the room. The visible wall features had the standard alternating channels construction, with some variation in the vicinity of the vault (F21) and the stairs (room 8).↩#p242
DSU 2778 was begun in 2005 but we did not complete the DSU. During the 2005 excavations a dense concentration of sherds was removed from the surface, which is standard for this area at Amheida as well as much of the rest of the site. In 2006 we continued to excavate windblown sand below where this dense concentration of sherds once existed. The matrix filled the entire room and consisted of windblown sand, ceramic fragments, clinker, faience fragments, and a few mud-brick fragments. As we proceeded we found that windblown sand continued in the northern portion of the room (DSU 35), but further south we began to find mud-brick debris (DSU 44). DSU designations were changed on the basis of position (DSU 35) and composition (DSU 44).↩#p243
Figure 4.16: Harris Matrix, room 4.
DSU 3579 consisted of windblown sand below the surface (DSU 27) in the north. This DSU was designated on the basis of position in order to indicate that it was less contaminated than the surface windblown sand. The context was almost entirely devoid of artifacts with the exception of a few fragments of ceramic and bone. We terminated this context when we encountered either vault collapse (DSU 44) or mud-brick debris mixed with sand (DSU 36).↩#p244
DSU 4480 consisted of mud-brick vault collapse underneath DSU 27 and DSU 35. This collapse clearly derived from a vault because of the type of bricks used. The vault bricks from DSU 44 differed somewhat from our other vault bricks: they were slightly curved and elongated with curved impressions on one face. We took architectural samples of two bricks (S2). Amid this collapse were found low densities of sherds that were probably used as chinking sherds to hold the vault bricks in place. Aside from a few bones and charred pits, glass fragments, and a single potsherd, no other inclusions were found in this matrix.↩#p245
DSU 3681 was found below both mud brick collapse (DSU 44) and windblown sand (DSU 35). It could be distinguished from DSU 44 by the presence of small mud brick debris mixed with sand rather than the large vault bricks that characterized DSU 44. There was a moderate to high density of small mud-brick fragments mixed with possible wind-blown sand. These fragments are probably from the deterioration and subsequent collapse of the vault ceiling over room 4, which produced DSU 44. Artifact densities for DSU 36 were low, with very low densities of ceramics, glass, bone, clinker, and plaster. The DSU was terminated when silty, occupational debris, DSU 45, was reached.↩#p246
DSU 4582 consisted of a silty brown matrix with moderate densities of organic inclusions, such as charred seeds. Low densities of fauna were also represented. Artifact densities were low in all categories except for charred fauna and seeds. There were very few diagnostic sherds from this context, which is singular given the high density of largely intact vessels associated with House B2. This DSU was directly on top of floor F15.↩#p247
F1583 is the floor of room 4. It consisted of a mud-brick pavement that was originally covered in mud plaster (Figure 4.17). F15 was in a good state of preservation. The northern 80 cm still had plaster adhering to the mud bricks, while only small patches of plaster remained in the southern extent. Both the bricks and the mud plaster were light grayish-brown in color and were tempered with straw. There was a 55 cm by 60 cm compressed portion of floor 20 cm north of wall F26. It is possible that this area represented wear from jumping onto the floor from a trap door. The northern portion of the floor is less worn, probably due to less traffic in this area.84↩#p248
Figure 4.17: Plan of room 4 floor (F15) showing FSUs (within rectangles) and elevations (with triangles).
F1485 is a wall forming the west boundary to room 4. The northern portion was preserved to a greater height and was in a better state of preservation overall than the southern portion. Because so little mud plaster remained on F14, the brick laying pattern and individual bricks were clearly visible. The bricks were yellowish-brown in color and laid in an English bond pattern. The bricks to the north were in a good state of preservation and better preserved than those to the south (see room 5 below). There was blackening 130–220 cm south of the corner of F14 with F30. It is likely that the blackening resulted from the use of oil lamps in room 4. The northernmost 40 cm of F14 had mud plaster overlapping from vault F21 on to the south face. This vault (F21) springs off F14 (and wall F37 to the east). The vault springing seemed to be at a higher elevation to the south than to the north on F14 (see vault F21). The foundations for F14 were not reached.↩#p249
F3786 is a north–south oriented wall that formed part of the east boundary of room 4. There were four courses of mud bricks in a good state of preservation where the wall met the vault. The level of preservation worsened closer to the floor level. There was no trace of any plaster facing adhering to the wall. The mud bricks were laid in an English bond pattern and the bricks had a high organic content. To the east, the vault spring and features obscured the thickness of the wall.↩#p250
F3887 is a north–south oriented wall that formed part of the east boundary of room 4, blocking the gap between the west ends of F26 and F37 and F28 that led to room 5. This wall abutted and was not bonded to any walls. There were eight courses of mud brick, the upper three of which corresponded to the lower part of the vault spring above F37. The preservation of the wall was fair, with some damage to the bricks and erosion of the mortar. The wall height was the same at both ends of the wall, with traces of mortar for the bedding of the course above at the north end of the wall. Only the upper three courses retained traces of gray-brown plaster. The wall was essentially only two brick-lengths long. The courses of the wall followed an English bond pattern.↩#p251
F2188 is the barrel-vaulted ceiling of room 4. It was in a good state of preservation for this structure. The springing of the vaults and the meeting of the vault springs were still intact. The vaults sprang from F14 to the west and F37 to the east. It seemed that the elevation of the springing increased to the south such that the distance between the ceiling F21 and the floor F15 increased to the south. The northern extremity of room 4 would therefore offer a shorter vertical space that the southern portion. The underside of the vault was completely covered in grayish-brown mud plaster. The cracks between the horizontal courses were clearly visible, and fourteen courses could be distinguished. The individual bricks within these courses were concealed. Some bricks from the vault collapse were found in DSU 44. They were curved rather than rectilinear vault bricks. This shape is so far not paralleled at Amheida, but it is likely that a different vaulting technique was used for narrower vaults than was used for vaults that would span a greater distance.89 No chinking sherds were visible in F21, but it is possible that they were obscured by plaster. In the northern 26 cm of F21 there were 16 cm wide cavities filled with small mud bricks and mud plaster that marked the termination of F21. This cavity was 14 cm wide for the southern 86 cm of the joining between F21 and the walls F14 and F37, showing how the gap caused by the vault springs was filled and anchored.↩#p252
Interpretation of Room 4
Room 4 was a small, narrow room of low height with a preserved maximum depth of 1.28m. It consisted of a mud brick floor, walls, and a vaulted ceiling. It was located immediately underneath the stairs (room 8). The vertical space of this room narrowed to the north, increasing the confined nature of this space.↩#p253
The entrance into the room is not clear because of the level of preservation of this room. Based upon comparanda in the Fayum, it may have been accessed by a trap door through the stairway above.90 There is a 55 cm by 60 cm compressed portion of floor 20 cm north of wall F26 that may be the area above which the trap door opened. The activity of jumping onto the floor in this area may have caused the compression and wear that is currently visible. Likewise, the northern portion of the floor is better preserved because of the lack of traffic in this area. No other distinguishing marks are visible. Despite the comparanda from the Fayum, however, it is possible that room 4 may have been accessed through room 5 rather than a trap door. The southern end of room 4 was not closed off with a wall and there were some wood fragments in this vicinity. The lack of a wall in this portion of the room might be explained by poor preservation, but given that all other walls in this house have been well preserved, albeit at a low height, it seems possible that a door could have connected these two rooms. If this was the case, the worn portion of F15 could be interpreted as wear from traveling through room 4 into room 10.↩#p254
A barrel vault, still partially in situ, covered this room (F21). The bricks used in the vault construction (F21) show the use of different vault bricks than those found in other rooms in House B2 and other parts of Amheida. The use of special bricks was not uncommon in Roman Egypt, as they were sometimes made for pitched vaults and other features.91↩#p255
The walls were mud plastered, although it appears that much of the mud plaster was not preserved. There were some signs of blackening along the mud plaster and mud bricks of the wall (F14), suggesting the use of oil lamps in the room. The wall foundations were not exposed, as the mud brick floor was preserved.↩#p256
The floor of this room was constructed out of mud brick, as was common in storage areas. This floor was fairly well preserved and retained some mud plaster on its surface, although this was largely worn away, probably during antiquity, especially in the southern portion of the room, which would have experienced higher traffic than the northern portion. We did not excavate through the floor, so the details of its construction are not certain.92↩#p257
Room 4 was a storage room. This room is singular for B2 since it lacked significant quantities of material culture and it is unclear precisely what was stored in the vicinity. It is likely that the occupants removed valuable objects from this room upon abandonment, as most other rooms in this structure contained high densities of objects.↩#p258
Room 5 (4.14 m north–south x 2.58 m east–west)
Four walls,93 already visible from the surface, enclosed room 5, which was filled with stratigraphic units of varying types and reliability for a depth of 1.275 m (Figures 4.18, 4.20).↩#p259
DSU 3394 consisted of windblown sand on the surface of room 5. This surface layer had a high density of ceramics, a few pieces of clinker, wood, bones, dross, and a low density of glass and faience. It contained large quantities of iron-rich fabric A group and moderate quantities of calcium-rich fabric A11 and P37. These ceramics share forms with Late Antique ceramics. There were a few isolated mud brick fragments, most likely from wall collapse due to the lack of distinguishing marks. This DSU was considered unreliable both because of its position on the surface and because of the presence of modern rubbish in the top 10–15 cm.↩#p260
DSU 4095 consisted of windblown sand below the surface. The change between DSU 33 and DSU 40 was based upon position. We wanted to make a distinction between the more contaminated surface sand and the less contaminated sand below. This distinction is particularly important in this house because we had many apparently in situ objects very close to the surface. This context appeared to be right above floor level and we found five objects in DSU 40, including 75%–100% reconstructable vessels96 as well as an ostrakon (Inv. 11049). We also found a moderate quantity of sherds. There were large quantities of iron-rich fabric A group and moderate quantities of iron-rich fabric A4. There were low to moderate quantities of calcium-rich fabric B and P37. Most of these artifacts were found in close association with F14 in the southwest quadrant of room 5. In particular, many vessels were wedged into a gap in wall F14 and probably served to repair this gap. The vessels found within F14 are designated as deriving from that feature, but a few outside of the feature may be associated with DSU 40, as they may have fallen into the matrix. In general, artifact densities were moderate to high for ceramics. These ceramics are broadly datable to Late Antiquity. We had low densities of clinker, bone, wood, and glass. There were low to moderate densities of mud brick debris within this context. This debris was largely associated with F14 in the southwest quadrant of room 5 and to a lesser degree in the center and near the doorway to room 7. DSU 40 was terminated when we encountered ash (DSU 41).↩#p261
Figure 4.18: Harris Matrix, room 5.
DSU 4197 was covered by DSU 40 and was located next to DSU 42. DSU 41 was an ashy matrix associated with burned sherds and objects associated with fire. DSU 41 consisted of what appeared to be destroyed floor layers mixed with ashy fill. The ash itself varied in color and consistency but not in sufficiently definable parameters to designate different contexts. The reliability was uncertain because it was not clear with which floor level the finds should be associated, and the context was not sealed by substantial collapse. The small finds associated with this context could be tentatively associated with floors on the basis of elevation. Glass, wood, bones, and clinker occurred in low densities, and we uncovered one date pit, five olive pits, eggshell fragments, and a textile fragment. The ceramics consisted of iron-rich fabric A group in high densities and calcium-rich fabric B group in low densities. We also found a complete lamp (Inv. 10155), a fire dog (Inv. 10165), a baggy jar base (Inv. 10163), a complete ceramic vessel (Inv. 10092) and a bowl used for cooking (Inv. 10094). These ceramics are datable to the third and fourth centuries. Sample S5 was taken of this matrix. The context was terminated in order to provide more control over provenancing finds, although it was quite similar to the underlying context DSU 67 (see below).↩#p262
DSU 4298 consisted of occupational debris: it was covered by DSU 40 and was next to DSU 41. DSU 42 consisted of a brown silty loam with moderate to high densities of organics. This context appeared to consist of occupational debris since it rested immediately on top of floor F40. Unlike DSU 41 to the south, DSU 42 was very thin and defined. It had a few scattered pockets of ash, but the isolation of ash concentrations to the southern half of room 5 suggests a differential use of space within the room. The ceramics consisted of iron-rich fabric A in moderate densities and low densities of calcium-rich fabric B and P37. These ceramics are datable to the third and fourth centuries. Clinker and plaster were found in low densities, as were organics such as wood, bones, and one burned olive pit. DSU 42 covered F40.↩#p263
F4099 was the first floor we encountered in room 5. The mud plaster was in a good state of preservation, although it was present only in the northeast quadrant. This visible portion of floor had a high density of organics. Unlike the floors beneath it (F41 and F47), F40 did not show signs of activities related to fire. This distinction might be the result of the position of the preserved portion of F40 rather than a change in the usage of room 1. For example, charred mud bricks associated with hearths were found in DSU 40 at an elevation that could be associated with either F40 or F41. The ashy matrix of DSU 40 and the charred bricks may indicate that the southwest quadrant of F40 was used for fire-related activities, while the heavily trafficked northeast quadrant of room 5 was free from such activities. F40 did not quite meet up with walls F55 and F16, being heavily worn in these areas.↩#p264
F41100 was a floor covered by DSU 41 (next to DSU 67). F41 was the second floor layer we encountered in room 5. Although F41 was highly compacted and had other signatures of being a floor (such as flatness, the presence of flat-lying sherds, etc), its close proximity to floor F40 above suggests that it might have been a preparation layer rather than a proper floor. Otherwise it is possible that F40 represents a minor repair to a heavily worn area of F41. F41 was in a moderately good state of preservation, although it was largely absent in the southern third of the room. The northeastern quadrant of F41 was obscured by the presence of floor F40. The visible portion of F41, running approximately east–west across room 5, appeared to have been associated with multiple activities related to fire. The intensity of these activities may have catalyzed the destruction of floor F41 in the south, where a high density of ash, DSU 41, was found. Likewise, two bricks stacked on top of each other in the manner of ephemeral hearths were found in this destroyed southwest portion. No other discernible marks were associated with F41.↩#p265
DSU 67101 consisted of ash with sand pockets. DSU 67 was similar to the DSU that covered it (DSU 41) in composition and inclusions, but it had a denser consistency. DSU 41 was terminated after we encountered F41, the second (middle) floor layer in room 5. We changed context designations so that we could have more control over determining the location of artifacts and with which floor layer they should be associated. The change in context was therefore based upon position and density of material. This matrix had a high density of ash, charcoal, and organic inclusions. The ceramics contained high densities of iron-rich fabric A group, low densities of calcium-rich fabric A11, B, and P37. The artifacts from this DSU are probably associated with the F41 floor occupation and included a carinated bowl (Inv. 10392), a nearly complete bowl (Inv. 10137), and a bread plate or lid with handle in many fragments. These ceramics are datable to the third and fourth centuries. Organic densities were high and included pits, bone, wood, worked wood (Inv. 11520), and a textile in three fragments (Inv. 11526). We had low to moderate densities of glass and clinker. Many of the artifacts from this assemblage were charred, which correlates with the high presence of ash, suggesting in situ burning. The context was terminated when either floor F47 or DSU 71, the fill of hearth F46, was reached.↩#p266
DSU 71102 filled F46, a hearth feature. This fill consisted of compacted loam with a high degree of compacted ash. Below the surface crust of mud-brick melt the context seemed to be secure, so a soil sample (S7) was taken. A complete, charred bowl (Inv. 10090) was found embedded within this context. Artifact densities for ceramics were moderate, and many of the sherds were charred. These fragments were iron-rich fabric A. Other categories include clinker and bone in low densities. There were low to moderate densities of organics including wood, charcoal, and seeds. DSU 71 covered F46.↩#p267
Figure 4.19: Photo of hearth (F46), room 5.
F46103 was a circular hearth feature associated with floor F47 (Figure 4.19). It was composed of mud bricks with mud plaster coating. Its base was a prepared mud plaster surface. It was in a very good state of preservation, being complete. The mud plaster was pink except on the top of the exterior ring and on the base of the feature. The feature was filled with ash and burned artifacts (DSU 71), indicating its active use as a hearth or, less likely, a brazier.↩#p268
F47104 was the bottom of three floors within room 5. Ash (DSU 67) and a hearth (F46) covered this floor. It was in a poor to moderate state of preservation. There seemed to have been a large number of activities related to fire, such as cooking, at this floor level; for example, hearth F46 was associated with this floor. These activities made F47 quite friable and easily broken. There were large portions of the floor missing in the southeast, southwest, and northwest corners of room 5. Despite its fragility on the edges, the floor remained very solid. Floor F47 was probably the original floor level of room 5, as was suggested by the presence of foundation bricks for wall F44 directly underneath it. No other distinguishing features were visible, and this floor was not fully exposed in the northeast corner.↩#p269
DSU 77105 consisted of compacted sand with no visible inclusions. DSU 67 covered it. It appeared to be the natural surface upon which House B2 was built. We did not excavate DSU 77.↩#p270
F13106 was an east–west oriented wall forming the south boundary to room 5. It was moderately well preserved and was preserved to a greater height in its western than its eastern end. The north face of the wall had yellowish-brown mud plaster adhering to the eastern 135 cm of the wall and for the top 32 cm only. The western 12 cm of the plaster were blackened, probably due to fire-related activities in the southwest quadrant of room 5. To the west, the individual bricks and the bonding pattern were visible. The walls were laid in an English bond pattern. The bricks were moderately preserved and were yellowish-brown in color. A few patches of red mud plaster were visible; these seem to have been used for repair, since it was a different color than the mud plaster to the east. There was a gouge 0–77 cm east of the corner with F14. It was 32–55 cm from the top of the wall and 8 cm deep. For 28 cm west of this gouge and at the same elevation the wall was very ashy and had numerous charcoal parts adhering to the bricks. There was a depression 0–18 cm east of the corner with F14. It was 16–32 cm from the top of the wall and was 9 cm deep. It is quite rounded due to use wear of an uncertain nature. The foundations were not exposed.↩#p271
F14107 is a north–south oriented wall forming the west boundary to room 5. It was in a poor to moderate state of preservation. The northern portion was preserved to a greater height and in a better state of preservation overall than the southern portion. Because so little mud plaster remained on F14, the brick laying pattern and individual bricks were clearly visible. The bricks are yellowish-brown in color, and their bonding pattern was English bond. The southern 360 cm of F14 had bricks in a moderate state of preservation as opposed to the northern portion in room 4, which was better preserved. There was a substantial gouge 0–353 cm north of the corner of F14 with F13. Its height varied from 0–20 cm above floor F47 to covering the entire vertical extent of F14. The depth of the gouge also varied greatly from 8 cm to possibly the entire width of the wall (not visible). At 160–220 cm from the corner with F13 a part of the rubble and sand filling the cavity was removed. This rubble consisted of largely intact ceramics and sherds. Most of this rubble was charred and highly burned. Given the proximity of hearths in this vicinity, fire probably catalyzed the fragile state of the wall here. Further south, F14 was abutted by F13, and reddish mud plaster was used in a coved shape between the two walls. The entire southern half of F14 had mild signs of soot or blackening, again, probably associated with the hearth or cooking activities. Directly opposite from F26 and F24, F14 had 6 cm deep indentations 0–19 cm from the top of the wall. The cause of these indentations was unclear, but they might be related to a possible door located here for entry between rooms 4 and 5.108 We did not reach the foundations for F14 during excavation.↩#p272
Figure 4.20: Multi-context plan of room 5 showing DSUs (encircled), FSUs (within rectangles), elevations (with triangles), and small finds (without symbols).
F24109 is an east–west oriented wall stub forming part of the north boundary to room 5. It was in a moderately good state of preservation, although it showed clear signs of wear and burning in its immediate vicinity. For example, there was a prominent burn spot on the wall 12–54 cm from the top of the wall and 40–90 cm from the southwest corner of F24. The bricks were set in an English bond pattern. The 7 lower courses were all set as stretchers, and the bricks of the upper course were all set as headers. The western corner was rounded by a combined action of occupational use and later weathering. The mortar between the bricks was dark yellow in color and made of a very high concentration of sand with almost no inclusions. The original mud plaster was, for the most part, missing, and only scanty traces of it were visible by some joints. It appeared to have been rich in fine lime inclusions, which created a light yellow color. The bricks themselves were beige in color. The wall was probably used to strengthen the corner of room 5 and room 4, if room 5 did indeed give access to the cellar. This hypothesis is substantiated by the way F24 abutted F26 running parallel to it. The wall appeared to be resting not on proper foundations but on the preparation layer of the floor in room 5 and thus represented a later addition to the room just after the initial occupation of B2.↩#p273
F55110 is a north–south oriented wall forming the east boundary of room 5. Wall F55 is in a low to moderate state of preservation. The wall retained a maximum of 7 courses above the foundations on in the northern portion of the wall, while the southern portion had barely 7 courses preserved. The remaining courses appeared to be stable. Most of the weathering was probably due to the wind. The bonding pattern for this wall was English bond, and the east face of the wall retained only very small patches of plaster. The southern portion of the wall face was better preserved than the northern portion. The northern portion had three gouges. The first gouge was 70–100 cm from the north boundary of F55 and was 0–13 cm above the bottom of the excavated area; it was 6 cm deep. The second gouge was 70–192 cm from the north boundary of F55, 5–12 cm above the bottom of the excavated area and 7 cm deep. The third gouge was 50–160 cm from the north boundary of F55, 28–39 cm above the bottom of the excavated area and 10 cm deep. It is likely that these gouges resulted from normal usage of room 5, which seems to have involved a lot of wear on the walls and floors. No other marks were visible on the wall. The tops of the foundation bricks were just barely visible in the southern 10 cm of the wall, standing vertically on their stretchers. Measurements of these foundations were not possible because they were not fully visible. The north face of the wall had a moderate quantity of plaster adhering to the face, and no marks were visible.↩#p274
F26111 is an east–west oriented wall forming part of the north boundary of room 5. F26 was in a moderately good state of preservation, with the western portion preserved to a greater height than the eastern portion. The western end of the wall was highly deteriorated, partly due to heavy usage during the occupation of B2. The south face was covered with mud plaster 83–160 cm west of the east corner with F26. At 160–264 cm west of the east corner of F26 the south face was obscured by wall stub F24, which may have been built to repair or buttress F26. The visible bricks in the south face were built in English bond. They were highly weathered from occupational use, wind, and perhaps some minor water damage. The plastered area had a thick, 2.5 cm application of plaster with a high density of organic inclusions. Behind this plaster there were signs of burning, which may have accelerated the wall deterioration and necessitated the addition of wall F24. Most of the visible bricks were yellowish-brown in color, but the bottom course had reddish bricks, which projected 10 cm into the room. These bricks were probably for repairing the wall base rather than being a true foundation course, which tends (in this area) to stand on its headers. Some wood was evident in the west face of F26 in the vicinity of the space between this wall and wall F14. This wood may indicate the remains of a doorway between these rooms, or it may have served as the remains of reinforcements placed in this wall.↩#p275
Interpretation of Room 5
Room 5 was accessed by means of room 7. The entrance into room 5 was somewhat narrow (60 cm) and elongated, removing it slightly from the central room of B2 (room 7). There was clearly no door located here, as there were no signs of a threshold or pivot for a door, and the flooring and walls were sufficiently preserved to evince these signatures had they been present. Room 5 was one of the larger rooms in B2, being smaller only than room 7. This room showed the most significant signs of burning, debris, and cooking activity among all of the rooms.↩#p276
There were no clear signs of roofing for this room, in contrast to the rooms covered by a barrel vault. It is possible that room 5 was once at least partially covered by a jarid roof. First, DSU 42 contained a high density of organic remains that may indicate the decayed remnants of a collapsed organic roof. Second, it was necessary to move from the stairway to the flat roofs over rooms 1, 2, 3, and 6. Room 5 provided the only clear means of access between the staircase (room 8) and these other rooms. There is no evidence for the appearance or extent of this possible roof.↩#p277
The walls of room 5 were mud plastered, although many of the walls show signs of wear and burning, which negatively impacted the preservation of the mud plaster. The walls appear to have been modified marginally during occupation. For example, wall stub F24 appears to have been added after the initial occupation of the structure. Other minor modifications included repairs to the walls, such as F14, which contained a large amount of debris used to shore up an area of heavy wear. These walls seem to have undergone more extreme usage than the walls in the other rooms of B2 and many of them, particularly F14, showed signs of fire in their vicinity. The signs of burning likely derived from active cooking in this vicinity, as the burn marks were quite large, particularly on F14. The foundations of the walls were partially visible on wall F55 and appear consistent with construction techniques elsewhere in B2.112↩#p278
The floor construction appears similar to that found in all other rooms in this structure with the exception of the storage rooms (rooms 4 and 10). These floors were mud plaster laid on top of mud brick debris and windblown sand. We had three floor levels in this room, suggesting intensive use of the area. All of the floors showed signatures of fire, burning, and ash. A hearth feature (F46) was on top of floor F47, and floor F41 had an informal hearth consisting of stacked mud bricks.↩#p279
In summation, room 5 contained a large number of signatures that may be associated with uses of fire, including burn marks on walls, substantial deposits of ash, both formal and informal hearths, fire dogs, and cooking implements. This room appeared to have been modified marginally during its phases of occupational usage, and all of these modifications appeared to be necessitated for stabilizing the room because of its intensive use for fire. This preponderance of data suggests that cooking took place here. In most households of Late Antiquity, cooking probably took place on portable braziers of bronze or pottery in the courtyard of the house. Materials related to cooking activity such as cooking pots, mortaria, frying pans, and metal implements such as knives are frequently found in domestic contexts and indicate such activities.113 In a Romano-Egyptian context we can understand the presence of hearths, firedogs, and cooking vessels to be the most significant indices of food preparation outside of bread ovens. The close association between these signatures suggests that this room was a significant locale for the preparation of food within B2.↩#p280
Room 6 (2.82 m north–south x 2.73 m east–west)
A south door in room 7 gave access to room 6. Room 6 is located along the southern end of B2, between rooms 3 and 5. Four walls,114 already visible from the surface, enclosed room 6, which was filled with stratigraphic units of varying types and reliability for a depth of 1.665 m (Figures 4.21, 4.23). We gridded room 6 below the surface to control ceramic analysis. Equalities are given between the DSUs for room 6, which resulted from the gridding, but were the result of the same deposition.↩#p281
Figure 4.21: Harris Matrix, room 6.
DSU 34115 consisted of windblown sand on the surface of room 6. The sand was mainly yellow in color with a conspicuous quantity of pottery sherds. The iron-rich fabric A is represented in high densities, with moderate densities of calcium-rich fabrics A11 and P37. We found one small fragment of faience, a few bones, glass, traces of decomposed wood, and rather large pieces of clinker. The DSU was terminated when the contaminated surface of sherds was complete. This change was largely based upon position to signify that the windblown sand below (DSU 37) was less exposed to surface contamination.↩#p282
DSU 37116 consisted of windblown sand below the surface. The sand in this layer corresponded, both in composition and texture, to that of DSU 34, the only difference being the absence of finds because DSU 34 had a high concentration of surface sherds. DSU 37 was distinguished from DSU 34, upon the basis of position and artifact density associated with the matrix. It was terminated when mud-brick collapse (DSU 38 = 43) was reached.↩#p283
DSU 38 (= 43)117 consisted of mud-brick collapse, most probably from parts of the room’s walls. Several bricks were still intact, while others were preserved in rather large (around 16 x 16 cm) pieces. Only a few vault bricks were visible in the collapse. The collapse was located mainly in the northeastern half of the room, with large voids filled with the sand from DSU 37 along wall F54 and in the northwestern corner. Among the bricks, roughly in the center of the room, five complete vessels were found.118 Three fragments of white, red, and colored plaster were visible, two still attached to the brick fragments and one detached. The iron-rich fabric A group occurred in moderate to high densities, while the calcium-rich fabric P37 was found in low–moderate densities. Glass, bones, and clinker occurred in low densities. The DSU was terminated when a silty, ash-laden context (DSU 47) was reached. We decided to grid room 6 at the termination of this DSU so that we could determine more precisely the location of objects that were not small finds. This strategy was employed since many of the sherds from room 6 could be meaningfully reconstructed.119↩#p284
DSU 43 (= DSU 38)120 consisted of mud-brick collapse. The division of the collapse into two different units was done along a north–south axis in the room, dividing it into equal parts. The collapse in this DSU seemed to derive from wall collapse, since no clear vault bricks, noticeable from the typical half-circle marks on one face, were present. The DSU was terminated when a silty ash-laden context (DSU 46) was encountered. Finds included moderate–high densities of iron-rich fabric A as well as bones and plaster, none of which was colored.↩#p285
DSU 46 (= DSU 47)121 consisted of occupational debris in the eastern half of room 6; it corresponded to DSU 47 in the western half of the room 6 north–south section. It was a secure context, since it was sealed by the collapse above. This matrix consisted of an ashy loam with moderate densities of organics. Ceramic densities, in particular, were high and consisted of iron-rich fabric A group. Also represented in more moderate densities were calcium-rich fabrics A11, B, and P37. Other categories of material include wood in moderate densities, while glass, bone, and clinker occurred in low densities. We also recovered some variety of desiccated fruit, perhaps a doum fruit. This DSU covered floor F16 and DSU 69; it probably represented the latest phase of occupation for this room.↩#p286
DSU 47 (= DSU 46)122 is a layer below collapse (DSU 43) in the western half of room 6 and therefore represents a secure context. A sample (S3) was therefore taken. This context was composed of brown sand mixed with organic material (straw) and very fine mud-brick fragments. It rested on top of floor F16 and below where the floor was eroded. Portions of this context might be interpreted as a mixture of destroyed floor and occupational debris, while other portions only represented occupational debris. The iron-rich fabric A was found in high densities, while the calcium-rich fabric P37 was found in low densities. Glass, clinker, and dross were found in low densities. We also found fruit fragments, gypsum, and charcoal. A coin (Inv. 11084) and the remains of a sewn plaits basket (Inv. 11554) were also found in this layer. Some pockets of ash were visible as a result of organic materials’ decomposition.↩#p287
F16123 was the top floor layer for room 6. It was in a poor to moderate state of preservation, only occurring in patches in the middle of room 6. The surface was moderately to highly compact and it appeared to have been built on top of a layer of organic-rich sandy loam (DSU 69) that was used to stabilize it. A sewn plaits basket (Inv. 11554) was found lying flat on this floor in the southwest quadrant. Seeds and other organics were also associated with this floor.↩#p288
DSU 69 (= 70)124 consisted of a loose, organic-rich sandy loam below the level of the floor (F16) in the east section of room 6. This matrix had moderate densities of clinker, wood, and glass. The ceramics included iron-rich fabric A in high densities as well as calcium-rich fabric P37 in moderate densities, and a complete bowl (Inv. 10397). A wooden lid of many fragments (Inv. 11071) was also found, although it should be considered as part of the architecture of the room, since it served as a lid to cover an under-floor storage area. Organic levels were high as a result of the preservation of this lid and included 127 pits, with nine date pits, 105 olive pits, and one peach pit. A small test trench in the northeastern corner of the room revealed that this context rested on clean sand (DSU 76). The context was not fully excavated because the western portion of the context was not excavated to this depth.↩#p289
DSU 70 (= 69)125 consisted of loose sandy loam with low to moderate densities of ceramics in the western half of room 6. It was below floor F16 and presumably covered floor F18, if that floor level is present in the west half of room 6, since DSU 69 covered F18 in the eastern half of the room. DSU 70 was not excavated.↩#p290
F18126 was the first, original floor layer for room 6. It was in a poor to moderate state of preservation. Only a very small patch of F18 was exposed. It was clear that this was the original floor level, because foundation bricks were visible in wall F55 just below the elevation of F18. F18 itself consisted of highly compacted, very light gray mud plaster. It was pocked with a large number of small holes from occupational usage and weathering. It had a moderate number of ceramic inclusions.↩#p291
Figure 4.22: Photo, Inv. 10383 and Inv. 10216 in situ in DSU 76, room 6.
DSU 76127 consisted of compacted clean sand with no natural inclusions. It was below the foundation level of walls F7, F12, and represented the surface upon which these walls were built. The iron-rich fabric A group was found in low to moderate densities and the calcium-rich A11, B and P37 fabrics were found in moderate quantities. Glass, wood, and clinker were found in low quantities. There were a large number of pits including 52 olive pits. Two complete storage jars (Inv. 10383 and Inv. 10216) were found embedded in the natural sand at the base of this DSU (Figure 4.22). The excavation of the DSU was terminated at the bottom of these jars. All of the ceramics are datable to the Late Roman and early Late Antique.↩#p292
Figure 4.23: Multi-context plan of room 6 showing DSUs (encircled), FSUs (within rectangles), elevations (with triangles), and small finds (without symbols).
F53128 was the doorway from room 7 to room 6, consisting of the lowest portions of the eastern and western sides, threshold and sill. Three and a half courses of mud brick were preserved on the west side, and the remains of mud plaster coating obscured its construction. The socket for the door pivot was found at the north end of the west side of the doorway. The east side consisted of two courses of bricks, with the uppermost course quite worn. It was laid in English bond, as was the rest of wall F12, which served as the east part of this doorway. The threshold was well preserved apart from one missing brick at the center of the uppermost course on the south side. The threshold consisted of wall courses continuing below the doorway rather than a separate construction element. There was a distinct sill with a width of the thickness of one brick (20 cm) that ran across the doorway at the far north end, and the mud floor (F17) of room 7 lipped over it.↩#p293
F12129 is an east–west oriented wall forming the north boundary of room 6. It was in a good state of preservation, with the western part preserved to a greater height than the eastern. The bricks were clearly visible and were grayish-brown in color. The bonding pattern was English bond, and the alternating channels construction method was used. The south face of the wall was weathered and seemed to have suffered from some minor water damage. There was no mud plaster visible. There was a light burn mark 60–75 cm east of wall F55 that was 8–14 cm above the foundation level. Some insignificant burning episode probably occurred in this vicinity, potentially after the structure was abandoned. Two courses of the foundation were visible here, and both had the bricks resting on their stretchers. The top course protruded into the room from the wall proper by 9 cm. The second course protruded by an additional 12 cm.↩#p294
F54130 is an east–west oriented wall forming the south boundary to room 6. It was in a poor state of preservation, with the western portion preserved to a greater height than the eastern portion. There was no mud plaster still adhering to the wall, so the bricks and the bonding pattern were fully visible as English bond. The bricks were yellowish-brown in color and were heavily weathered from occupational use. A large gouge was present on this north face. Its dimensions were: 55–203 cm east of west wall F55; 0–24 cm above the second course of bricks (from bottom); and 19 cm deep at the deepest point. The groove was probably the result of occupational use. A piece of wood was embedded in the wall 58–68 cm east of west wall F55 and 29–34 cm above the top of the second course of bricks (from bottom). It may have been used for attaching elements and furniture to the wall that could not be attached directly to the mud brick. The abutment between F54 and east wall F7 was separating, and a 6 cm wide gap between the walls was visible. The abutment between F54 and west wall F55 was still secure, and the plaster of F55 still coved over the joint between the two walls. The foundations of F54 were not reached.↩#p295
F55131 is a north–south oriented wall forming the west boundary of room 6 and the east boundary of room 5. Wall F55 was in a low to moderate state of preservation, with only 7 courses remaining above the foundations. The northern portion of the wall was preserved to a greater height than the southern portion. The remaining courses seemed fairly stable. Most of the weathering was probably due to the wind. The east face of the wall within room 6 had some mud plaster remaining, mostly in the southern portion, making it difficult to see the bricks in this vicinity, but it was clear that the bonding pattern for this wall was English bond. In the southern portion of the wall, the top course of the foundation bricks was visible. These foundations protruded from F55 into the room by approximately 10 cm. There was no visible bonding pattern to these foundations. A gouge was visible right above the foundations in the southern extremity. It was 130 cm long, 4.5–5 cm high and 2 cm deep.↩#p296
F7132 is a north–south oriented wall forming the east boundary to room 6. It was preserved to a visible height of four courses at its north end (see F52). The wall was laid in English bond. There were traces of pale brown mud plaster with high organic content visible along the entire length of the wall, except in the extreme north next to F52, where a plaster of similar consistency but pale red color extended in patches from the doorway to approximately 45 cm to the south along the wall. This plaster color change probably represents a separate plastering episode, which was also visible on wall F3. It may be interpreted as repair.↩#p297
Interpretation of Room 6
Room 6 was located between room 5 and room 3. This room shows signs of moderate wear on the walls and floors. It contained a large number of objects, many of which may have fallen from the presumed flat roof above the barrel-vaulted ceiling. It contained a trap door leading to a storage area.↩#p298
Room 6 was entered through a doorway in the north wall at its easternmost corner. This doorway was 58 cm wide. The doorway certainly contained a door, as evinced by the remains of a threshold, sill, and pivot (F53). It is uncertain if there was a locking mechanism for this door, as the walls were not preserved to sufficient height to determine the presence of a lock.↩#p299
There was very little collapse from the roof, but it is likely that this room was barrel vaulted and that a flat roof covered this barrel vault. The extent of wall collapse along with the presence of so many complete ceramics close to the surface may suggest that these ceramics derived from the roof and that the roof had parapets. There were bits of colored plaster within the collapse (DSU 43 = 38), which suggest that the ceiling and/or the tops of the walls were painted red and white.133↩#p300
The walls of this room were mud plastered, and there is also some evidence that it was painted. We found three fragments of red and white plaster.134 There were some fragments of blackened plaster, which was probably caused by smoke from oil lamps. The walls show moderately more damage than some of the other rooms of this structure.135 This damage may indicate that this room was used more intensively than these other rooms. The wall foundations were clearly visible, particularly in (F7) and appear consistent with construction techniques used in walls found in other rooms of the structure.136↩#p301
The floor construction was consistent with that found in all other rooms of this structure except for the storage rooms (rooms 4 and 10). It consisted of mud plaster overlying mud brick rubble and debris, which overlay windblown sand. There appear to have been multiple floors in this room. The floor contained an opening to an informal storage pit containing two jars that were embedded in clean sand. This opening was covered with a wooden lid (Inv. 11071).↩#p302
Room 6 was roofed with a barrel vault and had a flat roof above it that held ceramics. The room itself was used for storage, as evinced by a storage pit, as well as other activities. It appears to have been used extensively, given the amount of debris left behind as well as the wear upon the walls. There was lots of debris from fruit and other foods here as well (DSUs 69, 47, 46).↩#p303
Room 7 (6.20 m north–south x 2.65 m east–west)
Room 7 was entered from room 9 and gave access to six rooms.137 It is the largest, most central room in House B2. Nine walls,138 already visible from the surface, enclosed room 7, which was filled with stratigraphic units of varying types and reliability for a depth of 1.650 m (Figures 4.24, 4.26).↩#p304
DSU 39139 consisted of windblown sand on the surface of room 7. This surface layer had a moderate density of ceramics with few other inclusions. It covered clean windblown sand (DSU 49) or collapse (DSU 48). Ceramics included iron-rich fabric A in very high densities and moderate quantities of calcium-rich fabric A11, B, and P37. Faience, glass, bones, clinker, and plaster were represented in low to moderate quantities.↩#p305
DSU 48140 consisted of highly deteriorated wall collapse with a few fragments of vaulted ceiling collapse. This context covers DSU 49, a layer of windblown sand below the surface sand, suggesting that this sand blew in before substantial collapse took place. In the northern quadrant of the context there were small pockets of gray ash that were the early signs of an oven feature (F19) we were to find below. Additional signs of this oven included portions of a large vessel associated with the mud brick wall of F19.↩#p306
Figure 4.24: Harris Matrix, room 7.
This context had a large number of finds associated with the collapse, including two nearly complete vessels (Inv. 10347, Inv. 10384). There was also a clay tablet (Inv. 11048), which was found directly underneath the vessel Inv. 10347 in the collapse. This tablet was exceptionally well preserved and is one of only four found in Dakhla (and, indeed, in all of Roman Egypt). The context was secure because it was embedded in collapse and covered by Inv. 10347. There is some possibility that these objects derived from a niche that was once in the wall, because of the close contextual association, but such a conjecture is uncertain. The ceramics included iron-rich fabric A in very high quantities as well as low densities of calcium-rich fabric P37. Bones, clinker, and plaster were represented in moderate quantities, and one peach pit was found. This DSU covered windblown sand (DSU 49) and ashy debris in the vicinity of F19.↩#p307
DSU 49141 consisted of windblown sand below the surface. This context was found directly underneath windblown sand (DSU 39) in the northern 120 cm of the room as well as the southeast and southwest corners. DSU 49 was covered by wall collapse (DSU 48) in the approximate center of the room. Two objects from this context were found in close proximity to floor F17 and near walls and thus may represent occupational debris. These finds included a largely preserved lamp (Inv. 10011) and a weathered statue fragment of a naked male (Inv. 11509). We found iron-rich fabric A group in high densities as well as moderate densities of calcium-rich fabric A11, B, and P37. Glass, clinker, and plaster were found in low densities. Among organic remains, we found one olive pit, eight date pits, eggshell fragments, and extremely small quantities of charcoal. In the course of exposing this context, the oven feature (F19) became significantly more visible, and we ceased excavation when we found ash in its vicinity. Elsewhere the context was terminated at floor surface F17.↩#p308
DSU 61142 filled oven feature F19 and consisted of clean yellow sand mixed with ash and charcoal parts. Given its contents, this context may represent the interface between windblown sand (DSU 49) above and ash (DSU 63) below. Iron-rich fabric A was found in moderate densities, and calcium-rich fabric P37 was found in low densities. Clinker and plaster were also represented, and one burned date pit was recovered. These artifacts appeared to be quite mixed and not necessarily associated with the oven feature (F19), except for the sherds found that matched the ceramic body of F19. This context was terminated when a more singularly ashy matrix was encountered below.↩#p309
DSU 63143 filled oven feature F19 and consisted of ash and charcoal parts. It formed the bottom and most secure context within oven F19 in the northern extremity of room 7. Although DSU 63 represents the most secure context in F19, it appeared to have been contaminated, to some degree, by windblown sand, suggesting that it was either always exposed to the elements or was exposed to the elements for a considerable time span. There were very few finds within the ash. These finds included fragments of a bowl and fabric A group, a few pits, clinker, plaster, and small charcoal fragments. Very fine fragments of mud-brick were visible right on top of the base of the oven.↩#p310
F19144 is an oven feature constructed out of a large ceramic vessel and mud-brick exterior walls (Figure 6.3).145 It was in a good state of preservation, although the western portion lacked some elements. The entire interior of the feature was blackened from usage and had darkened circular marks running around the interior. At 0–13 cm above the base of the feature there were lines running laterally along the ceramic. There was a row of circular burn spots 10–20 cm above the bottom of the feature; each circle had a diameter of 10 cm. The second row was visible 21–33 cm above the bottom of the feature, and each of these circles had diameters of 12 cm. The circular marks were probably caused by the presence of the bread itself plastered up against the walls of the oven after heating it.↩#p311
There was an aperture from the interior of the feature to the exterior. It cut through both the ceramic and the mud-brick wall. It was 8–17 cm above the interior base of the feature and 46–80 cm west of the east wall of the feature and had a diameter of 10 cm. The circular aperture was the draft hole and the space for stoking the oven. It is located opposite to the prevailing wind from the north.↩#p312
Feature F19 seems to have been built after the original occupation of House B2, since the area where it abuts wall F4 to the east covered over white gypsum plaster. Because F19 was plastered into floor F19, it must have been added during the first floor level occupation. The floor space in front of the aperture was repaired over time, indicating intensive use of the oven feature. The relationship between this oven and the other floor layers146 was unclear because they were highly deteriorated in the vicinity of this oven, as can be seen in the description for DSU 64.↩#p313
DSU 64147 consisted of a brownish silty loam with numerous mud brick inclusions. It seemed to represent the interpenetration of floor layers, the rubble between layers, and occupational debris associated with floor layers. Artifact densities were moderate for this context, with categories of wood, bone, ceramics, glass, seeds, and clinker represented. Iron-rich fabric A group was represented in high densities, and we found calcium-rich fabric B and P37 in low to moderate densities. These ceramics were broadly datable to the second and third centuries, suggesting reuse of dumped material for floor construction. The footed base of a glass bowl was found in fragmentary condition. The context was terminated in its horizontal axis somewhat arbitrarily. We sectioned DSU 64 in order to see the destroyed floor layers that composed this deposition. We terminated this context on its vertical axis when we reached floor F49, which was fully present underneath this context.↩#p314
F17 (= DSU 75)148 was the top (fourth from bottom) floor layer for room 7 (Figure 4.25). It consisted of highly compacted light gray mud plaster. It was in a very good state of preservation, although it was largely absent in the vicinity of three walls.149 There was also a large gap in the floor 0–58 cm west of wall F7 and 34–169 cm north of doorway F53. Several mud bricks were visible in this vicinity. These mud bricks were probably used in as preparation for floor F17 to create a stable and even surface. This matrix can be found described more thoroughly below (DSU 75). At 50–75 cm north of F12 and 110–130 cm west of F7 there was a gray ash concentration with charcoal inclusions. There were no other discernible features or marks associated with floor F17. The northern part of F17 was removed and termed as DSU 75.↩#p315
Figure 4.25: Photo, multiple floor levels exposed and visible in room 7.
DSU 75 (= F17)150 consisted of very hard compacted mud. It was the same deposition as floor F17 but was given a different DSU number to designate the portion of the floor that we excavated in section as a stratigraphic unit. This context was highly compacted and had few inclusions, either organic or inorganic, although a few pockets of silty, organic-rich debris were present within the matrix. This debris was probably occupational debris associated with floor F45 below DSU 75 = F17 as well as debris used as fill to prepare floor F17. The DSU was terminated when floor F45 was reached. A bone hairpin was found in this context (Inv. 11320) along with high densities of iron-rich fabric A and moderate densities of calcium-rich fabric B and P37. The ceramics are datable to the second and third centuries, suggesting possible re-use of material from the vicinity for floor construction. Glass and clinker were also represented. We had high densities of organic remains from this context including wood, bones, and 73 pits (two almond pits, 55 olive pits, and 26 date pits). A soil sample (S8) of ash seeds and organics was taken of this matrix.↩#p316
F45151 was the second floor level (from the top) of room 7 (Figure 4.25). It was in a very good state of preservation and consisted of highly compacted grayish mud plaster with whiter patches. It had a high degree of inclusions, such as charcoal parts, ash pockets, pits (mostly olive), lime spots, ceramic fragments, and organic fibers. It had two patches of ash concentrations. The first patch was a gray ashy concentration 25–45 cm from the west wall F30 and 150–130 cm from the north wall F35. The second patch was a dark black and gray ashy concentration 90–130 cm west from east wall F3 and 275–310 cm south of north wall F32. Not all of F45 was exposed; presumably more was present underneath F17 to the south. F45 was missing from the vicinity of doorways F51 and F52 and had patches missing from the vicinity of F27 and F30. It abutted wall F3.↩#p317
F48152 was the third (from the top) floor level for room 7 (Figure 4.25). Due to its close proximity to floors F49 below and F45 above, F48 may have been a preparatory level for F45 rather than a proper floor layer. Alternatively, it might have been a repair episode. Very little of F48 was exposed, since it was revealed only through the sectioning of room 7 in order to clarify floor levels. The small exposed area appeared to be in good condition. F48 rested on large cobbles and debris that were used to stabilize and prepare a foundation for the new floor construction. There were no significant features or marks associated with this floor level.↩#p318
F49153 was the bottom (presumably original) floor layer of room 7 (Figure 4.25). It consisted of a highly compacted grayish mud plaster. It was in a good state of preservation, although portions of it were missing in the vicinity of four walls.154 The floor was extremely flat and compacted. There was one raised area of the floor that seemed to represent a repair episode. Its dimensions were 73 x 65 cm, and it was located immediately south of the oven feature (F19). This repair was probably necessitated by the use of F19 in the vicinity, since the oven’s draft hole, which was also used for stocking fuel, was located here. The repair patch also had a gray ash concentration in the vicinity of the hole into F19. The ash measured approximately 28 x 40 cm. The oven (F19) was associated with F49 since the plaster of F19 coved, or curved, to meet floor F49.↩#p319
F12155 is an east–west oriented wall forming the north boundary of room 6 and the south boundary to room 7. It was in a good state of preservation. The western portion was preserved to a greater height than the eastern end. The bricks were clearly visible and were grayish-brown in color. The bonding pattern was English bond and the alternating channels construction method was used. There were no substantial gouges or holes visible. A repair, or patching, seemed to have occurred in the 13 cm west of doorway F53. It was 0–14 cm above floor F17 and it may have extended below F17. This spot was probably where the reinforcement wood of the doorway was located, as evidenced by the socket in the east face (see description for doorway F53).↩#p320
Figure 4.26: Multi-context plan of room 7 showing DSUs (encircled), FSUs (within rectangles), elevations (with triangles), and small finds (without symbols).
F27156 is a north–south oriented wall forming part of the west boundary of room 7. It was in a good state of preservation, and the northern end was preserved to a greater height than the southern end. The bricks were highly visible and in a good state of preservation. They were grayish-brown. The bonding pattern was English bond, and the alternating channels method of construction was used. The wall face had almost no mud plaster adhering to it, and little of the mortar between the bricks remained. No other gouges or holes were visible. Both the southeast and northeast corners of F27 were slightly rounded from usage, most likely from frequent walking close to the wall in passage through room 7 between rooms 5 and 8. A socket was visible in the northeast corner. It was 14 cm south of the corner and 14 cm west. It is 0–13 cm above floor F45 and 11 cm deep. It was probably the socket for a door between room 7 and room 8. This doorway was not given a separate feature number because no threshold was visible. The foundations were not visible.↩#p321
F3157 is a short section of east–west oriented wall forming part of the west boundary to room 7. It was located between doorways to room 1 (F51) and room 2 (F52) and it also served as the boundary between rooms 1 and 2. The wall was L-shaped in plan view, and this western end represents the perpendicular short portion of the wall. The central part was well preserved, but the lower portions to the north and south ends were destroyed when the wooden thresholds of the doorways (F51, F52) were removed. A thick layer of mud plaster containing abundant organic material and some relatively large potsherds largely obscured this face of the wall. At the south end of the wall this plaster overlay a similar plaster that was pale red in color, traces of which could also be seen on the south side of doorway F52 (see F52 description). The brick-bonding pattern was English and the alternating channels method was used. These channels were 7 to 8 cm wide.↩#p322
F32158 is an east–west oriented wall forming the north boundary to room 7. It was in a good state of preservation with the western portion preserved to a greater height than the eastern portion. The south face of the wall was covered in a reddish-brown mud plaster that had organic, pebble, ceramic, and lime spot inclusions. This mud plaster obscured most of the individual mud bricks, but it was possible to determine that the bonding pattern was English bond. The bricks were, for the most part, yellowish-brown in color, although there were some reddish bricks as well. In the 144 cm of F32 that provided the north boundary to room 7 there was a gypsum plaster band over the mud plaster. It was 32–73 cm above floor F45 and thus extended to the top of the preserved height of F32. The gypsum plaster originally may have extended to floor level, because its bottom edge was rough from breakage and was not the original lower end of the plaster. This gypsum plaster, and the mud plaster below, cove at the corners between this wall and wall F4 and this wall with wall F36. This coving was approximately 4 cm thick, while the gypsum plaster was less than 0.5 cm thick. The gypsum plaster is somewhat blackened and grayed in the corners, likely due to the presence of the oven feature (F19) close by. In plan-view the wall clearly used the alternating channels method. The channel was 9 cm wide and varied greatly in length. The foundations of wall F32 were not visible in room 7, but the coving with floors F44 and F45 was evident. This coving protruded approximately 50 cm from the south face of F32.↩#p323
F35159 is an east–west oriented wall forming part of the north boundary to room 7. The wall was in a good state of preservation apart from the lowest course, which was undercut. Traces of mud plaster remained over the whole of the south face of the wall, with two areas of white plaster preserved, one on the upper west and one on the east side, which formed a rounded corner with wall F36. The brick-bonding pattern was obscured by mud plaster. The join where this wall abutted the adjacent north–south wall to the west (F34) was filled with gray-brown mud that was quite different in color from the wall plaster. The upper face of the wall was uneven and was apparently preserved to its original height. The wall was two brick-widths wide; the north row of bricks was covered in its east part by a channel lined with a thick layer of mud plaster (0.7–1.0 cm thick), with a layer of white gypsum plaster above (0.5–1.0 cm thick), forming an L-shaped channel, which turned northwards to the west of F36 and above F43. The channel ended one brick-length from the adjacent north–south wall to the west of F34 with the upper brick course offset slightly to the north to create a 7.5 cm lip that was coated in mud plaster. Both this plaster and that on the rest of the wall had high organic contents.↩#p324
F36160 is a north–south oriented wall forming a west boundary to room 7. It was in a good state of preservation. The entire east face of the wall was covered with mud plaster followed by a coat of white gypsum plaster that was less than 0.50 cm thick. This gypsum plaster was blackened in some areas, likely due to its close proximity to oven feature F19. The presence of plaster obscured both individual bricks and the bonding pattern for F36. There was an aperture in F36 that provided access into room 11, a storage feature. The aperture was 17–73 cm south of the corner with the north wall, F32 and extended 6–54 cm above floor F49. The aperture was worn through usage and followed the course of the vault ceiling, F43, for its upper extent. The foundations of F36 were not visible as its bricks extend beneath floor F19. The base mud-bricks extended by 5 cm in the southeast corner for one brick length, while the rest of the bottom bricks were flush with the wall face.↩#p325
F4161 is a north–south oriented wall forming part of the east boundary of room 7 and the west boundary of room 1. The south end of the wall is described as part of the doorway to room 1 (F51), and the east face is described with room 1. It was preserved to a greater height to the north than to the south. In plan view the upper course was laid as stretchers. The bonding pattern was obscured by mud plaster faced with gypsum plaster, which was preserved around the oven (F19). The mud plaster had a high organic component and became thicker towards the floor (F49). The mud plaster was missing in small patches at the south end of the wall, just outside the oven (F19). This plaster seemed to have been placed over an earlier layer of mud plaster of a similar composition but a slightly more orange-brown color.↩#p326
F51162 is the doorway from room 7 to room 1. The north side was preserved to a greater height than the south side. The lower parts of the doorway including the threshold were all quite poorly preserved, and the threshold appeared to be heavily damaged. The sides of the doorway consisted of the unmodified north face of wall F3 and south face of wall F4, with no sign of any jambs. The faces were mud plastered. There were possible sockets for wood consisting of undercuts at the bases of the west side of both walls. Immediately east of this socket was a mud-brick threshold consisting of two visible courses of bricks, lying so that stretchers were oriented east–west. The plaster of walls lipped or coved onto the upper surface of this mud-brick threshold. The mud plaster had high organic content. None of the bricks was sufficiently defined to measure.↩#p327
F52163 is the doorway from room 7 to room 2, consisting of the lower part of the sides of the doorway and the threshold. The vertical elements (jambs) were in fair condition, apart from a scar left from the removal or deterioration or removal of wood that once fit a socket in the threshold. The upper course of bricks for the threshold was virtually lost. This was a simple doorway consisting of the north face of F7 (5 courses) and the south face of F3 (6 courses), both coated in mud plaster with high organic content. As with doorway F51, the lowest 2 courses of the west sides of the walls were undercut, presumably for the placement of a socket for a wooden threshold, now lost. Traces of wood remained in the top of and side of the socket. The mud-brick threshold consisted of two visible courses of mud brick. The upper course was laid with headers running north–south, and the lower with stretchers running north–south in a truncated English bond pattern. Only traces on north and south sides remained of the upper course.↩#p328
F53164 is the doorway from room 7 to room 6, consisting of the lowest extent of the east and west sides, threshold and sill. Three and a half courses of mud brick were preserved on the west side, and the remains of mud plaster obscured the construction details. The socket cavity for the door pivot was visible at the north end of the west side of the doorway. The east side consisted of two courses of bricks, with the uppermost course being quite worn. The threshold was laid in an English bond pattern, as was the rest of wall F12. The threshold was well preserved apart from one missing brick at the center of the uppermost course on the south side. The threshold consisted of wall courses continuing below the doorway rather than a separate construction element. A distinct sill that was a brick thickness wide (20 cm) ran across the doorway at the far north end, and the mud floor (F17) of room 7 lipped over it.↩#p329
F7165 is a north–south oriented wall forming the east boundary to rooms 7 and 6, and the west boundary of room 3. It was preserved to a visible height of four courses at its north end (see doorway F52). The wall was laid in an English bond pattern, using the alternating channels construction method. There were traces of pale brown mud plaster with high organic content along the whole length of the wall, except in the extreme north next to doorway F52, where a plaster similar in consistency but pale red in color extended in patches from the doorway to approximately 45 cm to the south along the wall. This patch, which was also visible on F3, probably represented a separate plastering episode following use wear within and around the doorway.↩#p330
F30166 is an east–west oriented wall forming the west boundary of room 7. It was in a good state of preservation. The western portion of the wall was preserved to a greater height than the eastern portion. The bricks were in a poor–moderate state of preservation due to heavy use wear and were grayish-brown. The bonding pattern was English bond, and the alternating channels construction method was used. It has no substantial holes or gouges. The northeast and southeast corners were rounded from usage during frequent passage from room 9 into room 7 and from room 7 into room 8. There was likely a door between room 7 and room 8, as evinced by the remains of a socket in F30 (see room 8 below)↩#p331
Interpretation of Room 7
Room 7 was the largest room in this structure and served as a critical access point for the other rooms in the house. As such, it shows signs that it was used heavily but was also well maintained through multiple repair episodes. Some of this heavy usage consisted of food preparation, as shown by functional features in the north of the room (oven F19 and the storage room, room 11). Other activities included passage into other rooms as well as other ephemeral activities.↩#p332
Room 7 contained a total of six apertures leading to other rooms.167 Some of these entrances contained doors,168 while others consisted only of an opening (rooms 5, 9). The width of these doors varied between 58 and 68 cm. It is unclear if the entrances that contained doors were lockable, as the walls were not preserved to sufficient height to discern the presence or absence of locking mechanisms.↩#p333
There is no surviving evidence of a roof. If there was a roof, a light jarid reed roof is the most likely option. If there was a jarid roof it must have been removed at the time of abandonment or not long after, whether by human action or by wind erosion, as there were no organic remains suggesting its presence left behind, and all other organic remains within this room and this house were fairly well preserved. The presence of a bread oven (F19), located in its northernmost extent of the room, strongly suggests that this room was left open. Moreover, this room had windblown sand evident at most stratigraphic levels, suggesting that it was exposed more than the other rooms in this structure. Given the lack of secure archaeological data, the architectural parallels are particularly important for reconstructing this room.169↩#p334
The walls of this room were mud plastered, and the north wall of the room (F35) also contained a strip of gypsum plaster along it. The foundations of the walls of this room were not exposed, because their construction techniques were visible from other rooms (rooms, 1, 2, 3).170↩#p335
The floors were made of mud plaster, as was common in all rooms in this structure with the exception of the storage rooms (rooms 4, 10). The construction pattern of the floors was particularly well preserved in this room.171 This room experienced more floor repair episodes than any other room, an experience consistent with its function as a central access node for the entire house. This room was modified over time, since the bread oven (F19) appears to have been added shortly after the initial occupation of B2.↩#p336
Room 7 was a central point of access for other rooms in B2 and was used intensively for a variety of activities, largely food preparation. This room was kept in a good state of repair by the inhabitants as can be seen in the careful repair of the floor and the lack of debris left behind in most of the room.↩#p337
Room 8 (0.86 m north–south x 1.85 m east–west)
Figure 4.27: Multi-context plan of room 8 showing DSUs (encircled), FSUs (within rectangles), and elevations (with triangles).
Two walls (F28, F30), already visible from the surface, enclosed room 8, which was filled with stratigraphic units of varying types and reliability for a depth of 0.52 m. From the surface, room 8 appeared to be an elongated rectangular room opening up onto room 7. We defined the room on the basis of the visible walls and began our excavation with the top layer contained within these walls (Figures 4.27, 4.28, 4.29).↩#p338
Figure 4.28: Harris Matrix, room 8.
DSU 51172 consisted of yellow windblown sand on the surface of room 8. There was a high density of ceramics on the surface. We also recovered low densities of clinker, bone, and glass fragments. There were no natural inclusions in this matrix, and most of the artifacts were clustered in the top 1–3 cm of the context. We terminated this context when we reached a matrix of dispersed mud-brick collapse (DSU 53).↩#p339
DSU 53173 consisted of loosely articulated mud-brick debris amid a general matrix of clean yellow windblown sand. These mud bricks were located in the approximate middle of room 8. There were no vault bricks clearly identifiable from characteristic markings, so this collapse probably derived from walls. Artifact densities were low and consisted of ceramics only, most of which were from the iron-rich fabric A group.↩#p340
DSU 55174 consisted of clean windblown sand below both the surface windblown sand (DSU 51) and the loosely articulated collapse (DSU 53). This context was largely devoid of artifacts, although low densities of ceramics, faience, glass, bones, and clinker were present. Among the ceramics recovered we found mostly iron-rich fabric A group with a lower density of calcium-rich P37 fabric sherds. A few isolated mud brick fragments were also included in this matrix. It covered a silty, brown matrix (DSU 60).↩#p341
DSU 60175 consisted of a silty, organic-rich, brown matrix, which we interpreted to be occupational debris above floor level. The matrix had moderate densities of ceramics and other classes of artifacts such as glass, clinker, bone, and two date pits. The ceramics included a large percentage of iron-rich fabric A group and a significantly lower percentage of calcium-rich fabric B group. We took a sample of this matrix (S6), since it seemed to have moderate densities of organic remains deriving from a secure context below collapse. This context was not found for approximately 5 cm against the north and south walls (F28 and F30). We interpreted the lack of context in these areas as the typical wear of floor and walls at their points of meeting. Furthermore, the lack of collapse (DSU 53) in these areas likely accelerated deterioration. DSU 60 covered floor F20.↩#p342
Figure 4.29: Photo of rooms 7 and 8.
F20176 is the grayish-brown compacted mud floor of room 8, located beneath the occupational debris (DSU 60). The floor was located between the last step of the stairs (F29) and the floor (F17) in room 7. It was in a moderate state of preservation, being missing for 27 cm from wall F28 and 15 cm from wall F30. The straw tempering of the floor was still visible in the compacted mud plaster of the surface. There were also a large number of seeds and pits embedded in the surface of the floor. We did not excavate this floor.↩#p343
F29177 is the staircase of room 8.178 It was in a moderately good state of preservation, but only in its bottom part. Three full steps were preserved in the first flight and the mud plaster facing of the steps was still present.179 This plaster almost completely obscured the bricks below, but it seems probable that the steps were made of two consecutive headers and one stretcher. The plaster facing was reddish-brown in color and the bricks were reddish. Small fragments of pottery, seeds, pits, and straw originally mixed with the mud to strengthen it were visible in the mud plaster. The mud plaster clearly abutted walls F28 and F30, making it later than the two walls. The mud plaster also could be a re-facing of the steps in a later moment from the original construction of the stair. The first and second landings were on top of the vault (F21) that covered room 4. The third flight is not preserved, but it originally would have covered the small room 10.↩#p344
F28180 is an east–west oriented wall forming the south boundary to room 8. It was in a good state of preservation. The alternating channels construction method was used for this wall, but the bonding pattern was not visible. It was impossible to investigate the lower courses of the north face of this wall because the stairs (F29) abut this face, and the rest of the wall was visible only to floor level (F20). The bricks still visible were beige in color and the facing light beige. There were traces of straw temper still visible in the mud plaster facing. Since this plaster facing still covered the bricks, it was not possible to see the technique used in setting them. A brick protruded by 14 cm from the north face of F27, and 83 cm west of the east corner of the wall. The function of this protruding brick was unclear, as it was not a foundation brick. It may be related to stairs F29, but it is more likely that it reflects poor bricklaying techniques. The wall protruded 16 cm from the wall face 107 cm west of the east corner of F27 and then narrowed for the stairs (F29).↩#p345
F30181 is an east–west oriented wall forming the north boundary of room 8. It was in a good state of preservation, with the western portion of the wall preserved to a greater height than the eastern portion. The individual bricks and the construction method were clearly visible. The construction technique followed an English bonding pattern and the alternating channels construction method. The bricks themselves were grayish-brown and in a poor to moderate state of preservation due to heavy use wear. We did not excavate below floor level, and therefore the foundations were not visible. The south face of the wall was somewhat weathered and had many narrow and shallow grooves, likely due to heavy occupational usage of the hallway and staircase (F29). The stairs (F29) that abutted this wall partially obscured the visibility of this south face. There was a small hole 8–31 cm west of the southeast corner of the wall and flush with floor level (F20). This hole probably served as a socket for a door in the doorway between room 7 and room 8. A threshold was not visible, so we did not designate a separate feature for this doorway.↩#p346
Interpretation of Room 8
Room 8 served as the hallway and staircase that connected room 7 with the roof of the house (now missing). It is likely that the original stairway turned 90° to the south and had a short landing over room 4 and then may have turned 90° again towards the east, followed by another flight of stairs to give it sufficient height above floor level.↩#p347
Room 8 was entered through room 7, on the east side. This entrance probably contained a door, as suggested by the possible socket located in F30. The width of this entrance is 0.86 m. Based on comparative evidence rather than on archaeological data, it is possible that the landing over room 4 once contained a trap door to the storage area below.182 Likewise, there may have been a trap door over the room 10 storage area. These possibilities will be discussed in greater detail in chapter 6.↩#p348
The walls of this room were mud plastered. This plaster shows some signs of wear due to heavy usage as a passageway to the roof. The foundations of the walls were not exposed in this room.↩#p349
The floor consisted of mud plaster, and it was worn away at the junctions with the walls. Otherwise it was in good condition. A high density of organic remains was embedded in its surface. It was not excavated. The stairway consisted of mud brick surmounted by mud plaster, and only three steps remained.↩#p350
The central location of this staircase is a slight variation from comparanda, in which the stairway is located either in either the center or the corner of houses.183 Houses from Karanis and elsewhere have wood reinforcements in areas of heavy wear, such as stair risers, door edges, etc.184 Except for doorways, there is no evidence for such usage of wood in House B2, and the stairs are in a good state of preservation. The lack of wood on this staircase is probably due to a lack of wood in Dakhla. Indeed, visitors to the houses of Bashendi will note that there are no wood reinforcements on stairs, and these instead are replastered seasonally so as to maintain their form.↩#p351
The depositions, although largely lacking in artifacts, reveal a bit about the usage and history of this room. The presence of a large number of olive pits and other seeds embedded into the floor suggests that these goods may have been stored on the roof that once covered other rooms within the house. Indeed, some of the jars found embedded in roof collapse (rooms 1 and 6) may have derived from the roof and could have served as storage vessels for these items. The presence of windblown sand (DSU 55) immediately above the occupational debris (DSU 60) likely derives from sand blowing in from open areas surrounding the stairs. The collapse of surrounding walls then sealed this area before it was covered with yet more windblown sand.↩#p352
Room 9 (3.0 m north–south x 3.10 m east–west)
Room 9 served as the sole entrance into house B2. It forms a general L-Shape or dodged entrance, which is a common attribute of Egyptian houses on the edge of the desert. Five walls,185 already visible from the surface, enclosed room 9, which was filled with stratigraphic units of varying types and reliability for a depth of 0.755 m (Figures 4.30, 4.31).↩#p353
DSU 50186 consisted of windblown sand on the surface of room 9. It contained a few mud bricks related to wall and ceiling collapse (DSU 52) that were visible from the surface. The ceramics densities were high for DSU 50, as is typical of surface layers in this vicinity. Other artifact classes that were represented include fauna, glass, and clinker in low densities. The DSU was terminated when either the collapse (DSU 52) or clean, below-surface windblown sand (DSU 59) was reached.↩#p354
DSU 52187 consisted of mud-brick debris concentrated in a few areas within room 9. These mud-bricks appear to be from both ceiling and wall collapse, as can be seen from the characteristic half-circle markings visible on vault bricks. Artifact densities were low to moderate, with only ceramics, fauna, and clinker represented.↩#p355
DSU 59188 consisted of clean windblown sand below the surface windblown sand (DSU 50) and a small deposit of collapse (DSU 52). It contained low to moderate densities of mud-brick fragments and debris. Low to moderate densities of artifacts were associated with this context, including ceramics, clinker, faience, glass, fauna, iron, and dross. The DSU was probably contaminated, as evidenced by the presence of a modern stick of wood. It covered a compacted mud surface (DSU 73).↩#p356
Figure 4.30: Harris Matrix, room 9.
F39 (= DSU 73)189 is the top floor layer that we encountered in room 9. It consisted of compacted grayish-brown mud plaster with a moderate to high density of organic inclusions. The floor was in a moderately good state of preservation, although there were areas where the floor was missing in the vicinity of wall F30. This area was 0–43 cm north of wall F30 and 8–60 cm west of the east corner of F30. Floor F39 may have extended an additional 180 cm east–west and 200 cm north–south to the west of where it was visible. The western portion was not as compacted and uniform as the eastern portion. We sectioned F39 in order to examine the floor construction and to determine if there were previous floors. This section was designated as DSU 73 and was located in the western 180 x 200 cm. We removed DSU 73 and found that it was consistently of a lower compaction than one would expect for a floor. The presence of a rubble layer (DSU 74) below, however, suggested that there was preparation for a floor level in this area. F39 likewise seemed to rest on top of the rubble layer used as a preparation to stabilize and support the floor (DSU 74). There was a small area of mud plaster melt 0–36 cm north of wall F30 and 63–88 cm west of the east corner of F30. Its proximity to the area where the floor was missing suggests that F39 was highly weathered in the vicinity of F30.↩#p357
DSU 73 (= F 39)190 is a compacted mud surface. This mud surface equals F39, the top floor layer for room 9. We gave a separate DSU number to the same surface because we wanted to section it in order to determine if it was definitely a floor and if there were additional floor layers. We excavated the western portion of the floor. The DSU itself consisted of moderately compacted loam with a low density of mud-brick debris and inclusions. It had a low to moderate density of ceramics and other classes of artifacts including clinker, glass, and bone. The DSU covered numerous mud brick fragments (DSU 74).↩#p358
Figure 4.31: Multi-context plan of room 9 showing DSUs (encircled), FSUs (within rectangles), and elevations (with triangles).
DSU 74191 consisted of numerous small mud-brick fragments and a few isolated mud bricks. In addition to the brick inclusions there were moderate densities of artifacts, including ceramics, wood, fauna, glass, clinker, and seeds. The DSU was terminated when floor F44 was encountered. This matrix probably served as a preparation layer for floor F39 above it.↩#p359
F44192 is the bottom (and presumably original) floor level of room 9. It was in a good state of preservation. It consisted of highly compacted light gray mud plaster with moderate densities of ceramic inclusions. Some organics, lime spots, charcoal parts, and ash lenses were also visible. Although F44 was in a good state of preservation, portions of the floor were missing in the vicinity of south wall F30, west wall F31, and the northeast corner where F32 and F33 meet. At 50–59 cm north of south wall F30 and 80–96 cm east of west wall F31, there was a small patch of dark grayish-brown compacted ash. Of note were a few reddish patches in the northern half of the floor. These red patches may be interpreted as repairs to the floor during occupation. The floor appears to have been built directly on top of sand, as sand was visible in some patches near the walls where the floor was missing. We did not excavate below floor level and therefore cannot verify this assumption, but it is consistent with our understanding of construction methods in B2.193↩#p360
F31194 is a north–south oriented wall forming the west boundary of room 9. It was in a moderately good state of preservation, with the northern portion preserved to a greater height than the southern. The entire east face of F31 was covered with reddish-brown mud plaster with organic and ceramic inclusions. The individual bricks and the bonding pattern of F31 were not visible due to the presence of this plaster. The plan view of F31 reveals that the alternating channels construction method was used. The channels were approximately 54 x 10 cm in length and alternate from the east to the west third of the wall. There was a substantial gouge 0–148 cm north of the corner of this wall with wall F30. It was 0–42 cm above floor F44. It was very deep and appeared to undercut most of F31. Measurements of its depth were not possible because it was necessary to stabilize the wall with mud-bricks. At 65 cm from the corner with F30 there was a crack running down the entire east face. There was an additional large gouge to the north associated with doorway F50. There was a thin depression 64–84 cm north of the corner with F30. It was 100–113 cm above floor F44 and 3 cm deep. The cause of this depression may also be related to doorway F50. Coving between the wall and the floor was evident towards the north. It extended approximately 6 cm beyond the extent of the rest of the wall face. The foundations of F31 were not visible.↩#p361
F32195 is an east–west oriented wall forming part of the north boundary to room 9, the north boundary to room 11 and part of the north boundary to room 7. It was in a good state of preservation, with the western portion preserved to a greater height than the eastern portion. The south face of the wall was covered in a reddish-brown mud plaster that had organic, pebble, ceramic, and lime spot inclusions. This mud plaster obscured most of the individual mud bricks, but an English bonding pattern was discernible. In plan-view the wall clearly used the alternating channels method. The channel was 9 cm wide and varied greatly in length. The bricks were, for the most part, yellowish-brown in color, although there were some reddish bricks as well. The western extent of the wall was abutted by doorway F50. Both F32 and F50 appeared worn from the frequent usage of this doorway. There was a hole or socket 12–28 cm east of the junction between F32 and F34. It was 31–45 cm above where wall F43 meets F32 and was 9 cm deep. It may be related to the use of the space above wall F43. The foundations of wall F32 were not reached during excavation, but the coving with floors F44 and F45 was evident and protruded approximately 50 cm from the south face of F32 into the room.↩#p362
F33196 is a wall stub in room 9 that formed both an east and a north boundary for the room. It was in a good state of preservation, although no mud plaster remained. The northern extent was preserved to a greater height than the southern extent. The west face and the south face of the wall were the only visible faces of the wall. The individual bricks and bonding pattern for the wall were clearly visible. The west face of the wall was built in English bond in three segments abutting one another that are approximately 35 cm wide each. They were joined by 3 cm thick mud plaster. F33 was associated with floor level F39 and not with floor level F44, because it was built on top of rubble, DSU 74, which, in turn, rested on top of F44. Wall F33 does not belong to the original construction of House B2 but rather served as a later addition. The south face of F33 shows the English bond construction pattern and does not show segmented construction because the segments run east–west and not north–south. There are also signs of coving between F33 and F39, which protruded by 5 cm from the south face of F33 into the room.↩#p363
F34197 is an east–west oriented wall stub forming part of the north boundary to room 9 and the west boundary to room 11. It was in a good state of preservation with the northern portion preserved to a slightly greater height than the southern portion. The south face of the wall was almost entirely covered in grayish-brown mud plaster, so the bonding pattern was not visible. In plan-view it was clear that a variation on the alternating channels method of construction was used. The bricks visible in plan view were both yellowish-brown and red in color. The mortar was grayish-brown. The southwest corner of F34 was rounded, probably from the usage of room 9 prior to the addition of wall F33. Wall F34 seems to be associated with the initial occupation of B2, since it was associated with the first floor level, F44. Its foundations were not visible.↩#p364
F30198 is an east–west oriented wall forming the south boundary of room 9 and the north boundary of room 8. It was in a good state of preservation. The western portion of the wall was preserved to a greater height than the eastern portion. The individual bricks and the construction method were clearly visible, showing an English bond pattern and the alternating channels construction method. The bricks themselves were grayish brown and were in a poor to moderate state of preservation due to heavy use wear. The north face of the wall was highly worn from use. There were two holes in the wall that were probably due to use wear. The first hole was 3–15 cm east of wall F31, 13–20 cm from the top of the wall and 3 cm deep. The second hole was 40–53 cm from wall F31, 30–40 cm from the top of the wall, and 7 cm deep. The easternmost 155 cm of the wall was particularly weathered with many narrow, shallow, horizontal and vertical gouges, particularly 20–30 cm from the top of the wall. None of these is substantial, however, and they probably represent wear from frequent passage from room 9 into room 7.↩#p365
F50199 is a 1.47 m wide doorway in room 9 leading north to the street (Figure 4.32). It was in a moderately good state of preservation. It consisted of 2 steps formed by 5 limestone blocks used as steps up 12–15 cm to the street. The jambs on either side of the blocks were constructed out of mud brick. These jambs bonded to the walls on either side (F32 and F31). The blocks were of irregular dimensions and, although they were generally rectangular in shape, some of them were quite irregular. One of the blocks contained approximately 18 gouges. The dimensions of the clearest gouges were: (1) 7 x 1.5 cm; 1.5 cm deep; (2) 6 x 1 cm; less than 1 cm deep; (3) 4.5 x 0.5 cm; less than 1 cm deep. Most of the gouges ran north–south but a few were diagonal. The cause of these gouges is not certain, but it is possible that the block was used as an informal whetting stone. Abutting wall F32 was a mud-brick jamb that was so heavily weathered that it was not possible to see if there were sockets for the door. The northernmost 16 cm protruded west by 16 cm, creating a narrowing of the doorway aperture. The western portion of the doorway made use of F31, the west wall of room 9. Wall F31 was in a better state of preservation than the western portion of the doorway itself, potentially because of the removal of wood from the doorway in antiquity. Doorway F50 had a depressed area 9–292 cm above the lower western limestone block and 0–280 cm north of the block’s south boundary. It was 3 cm deep. It was probably caused by weathering and use. Doorway F50 was plastered into floor F44 with mud plaster in its bottom course of blocks. Its relationship with floor F39 was unclear, since F39 was not clearly articulated in this vicinity.↩#p366
Interpretation of Room 9
Room 9 contained a doorway to the street and an entrance into room 7. The doorway to the street was significant and had an aperture of 1.47 meters. The street door could be locked from the inside by a sliding bolt; a square hole through the wall (F32) to the east of the door serves as clear indication of this attribute. Such locks were standard in domestic structures. The doorway to the street was modified over time to compensate for the rising street level. The stairs belonging to the doorway (F50) were added after the initial occupation of the structure. The aperture leading to room 7 did not contain a door, as there was no threshold or pivot evident, and the architecture was sufficiently preserved to indicate one had it been present.↩#p367
Figure 4.32: Photo of stairway and door F50 in room 9.
The walls of room 9 were mud plastered and show signs of repair due to heavy usage. The walls were highly worn down at their bases, particularly the west wall (F31). The foundations were only partially exposed, and they appeared consistent with wall construction patterns found elsewhere in this structure.200↩#p368
The multiple floors consisted of mud plaster, showing signs of repair and renovation or re-plastering over time. The construction pattern appears consistent with that found in other rooms of the structure, as was clear from excavation of these floors. These floors and the associated depositions contained fewer objects than many of the others in B2, probably because it was a transitional space rather than an activity space.↩#p369
This room served as the only entrance into the structure. It is clear that the strong north wind affected the architecture, as can be seen both in the dodged entrance and in the addition of stairs to compensate for the rising street level.↩#p370
Room 10 (0.74 m north–south x 1.47 m east–west)
This room served as a storage room underneath the stairs (room 8) and therefore formed a counterpart to room 4. Little material culture remained with which to make determinations as to what was stored in this room. Four walls,201 already visible from the surface, enclosed room 10, which was filled with stratigraphic units of varying types and reliability for a depth of 0.790 m. We thus proceeded with the single context method and used these walls to determine the parameters of the room (Figures 4.33, 4.34).↩#p371
Figure 4.33: Harris Matrix, room 10.
DSU 54202 consisted of windblown sand on the surface of room 10. A few mud bricks related to wall or ceiling collapse quickly became visible beneath this surface matrix of wind-blown sand. DSU 54 was terminated at this loosely articulated layer of collapse (DSU 56). DSU 54 had moderate to high densities of ceramics of iron-rich fabric A group and low densities of calcium-rich fabric P37, as is typical in B2. At the end of DSU 54 it became largely devoid of artifacts of all classes.↩#p372
DSU 56203 consisted of loosely articulated mud-brick debris. This debris was included within a general matrix of yellow windblown sand. Artifact densities were low, with only a few sherds of iron-rich fabric A deriving from this context. No other inclusions were visible. Following an examination of the bricks from this context it was unclear if these bricks derive from the ceiling or the wall, because they were highly eroded and melted together. DSU 56 covered DSU 57.↩#p373
DSU 57204 consisted of mud-brick collapse from an undetermined location in room 10, but none of the bricks clearly resembled a vault brick. There were few inclusions in DSU 57, only large pockets of windblown sand. Artifact densities were low, with only a few potsherds deriving from the unit.↩#p374
DSU 58205 consisted of sand below the collapse (DSU 57). This sand had low to moderate artifact densities for all classes of material. The material categories included low densities of fabric A ceramics, clinker, and glass. Among the organics, bone, charcoal parts, and ten olive pits were recovered. We found relatively few other inclusions, and these included only small fragments of disarticulated mud brick debris. The DSU covered compacted mud (DSU 66) and mud-brick floor (F42).↩#p375
F42206 was the floor of room 10, preserved only under the north wall of the room (F28) and as a single line of bricks in two rows towards the south wall (F26). The central part of the floor was missing and instead was filled with loose rubble and compacted adobe melt (DSU 66). Under the north wall the single course of floor that was visible consisted of mud bricks laid on their headers, with their faces oriented east–west. They were laid to form a lip with the wall 0.8 cm wide. The upper surface was coated with a thick mud plaster that was very poorly preserved. The other area of the floor was laid with the bricks on their faces with the header east–west; the lower course was, as far we could determine from the small area visible, laid with headers oriented north–south, forming what would essentially be English bond if this floor had been a wall.↩#p376
DSU 66207 probably equals F42, but it was distinguished from it because DSU 66 represents the deteriorated portion of F42, which gave it a silty, sandy loam texture. This DSU was not excavated since we terminated the excavation of room 10 at the level of floor F42.↩#p377
Figure 4.34: Multi-context plan of room 10 showing DSUs (encircled), FSUs (within rectangles), and elevations (with triangles).
F27208 is a north–south oriented wall forming the east boundary to room 10. F27 is in a good state of preservation. The northern part is preserved to a greater height than the southern. The grayish-brown bricks were highly visible and in a good state of preservation. The bonding pattern is English bond, and the alternate channels method of construction was used. The west face has no mud plaster adhering to it, and little of the mortar between the bricks remained. No other gouges or holes are visible. ↩#p378
F26209 is an east–west oriented wall forming the south boundary of room 10. F26 was in a moderately good state of preservation, with the western portion preserved to a greater height than the eastern portion. The north face of the wall was almost entirely covered with mud plaster that had blackened, probably from the use of oil lamps in this vicinity. The western 47 cm of the wall was highly deteriorated. This deterioration extended 39 cm above the floor (F42) and 8 cm deep. This western portion of F26 also contained some wood, the absence of which may have accelerated post-occupational deterioration, but use wear during occupation seems to have been a major cause as well. A gouge extended across the entire north face of the wall at floor level. It had a depth of 4 cm and a height of 8 cm. It probably occurred during the use of B2. The foundations of F26 were not reached during excavation.↩#p379
F28210 is an east–west oriented wall forming the north boundary to room 10. It was in a good state of preservation, with the mud plaster facing still present on the south face. The facing was light beige in color, and there were traces of the straw mixed in with the mud. Because the plaster still covered the bricks, it was not possible to see the technique used in setting them. However the topmost course was visible, indicating that the alternating channels construction method was employed. The south face of the wall met with floor F42 and lacked plaster in the bottom 20–30 cm above the top of the foundations. Here it was clear that two or more layers of mud plaster were applied. The foundation bricks of F28 were partially visible. They consisted of mud bricks resting on their headers rather than on their stretchers. No additional courses of foundation were reached during excavation.↩#p380
F38211 is a north–south oriented wall forming the west boundary of room 10. This wall served as a block between the west ends of walls F26 and F37 and F28 and was not bonded to them. The preservation of the wall was fair, with some damage to the bricks and erosion of the mortar. The wall height was the same at both ends of the wall, with traces of mortar for the bedding of the course above at the north end of the wall. There were only traces of gray-brown plaster on the upper three courses. The wall was two brick-lengths long, courses alternating header-stretcher-header, then a course of stretchers. The upper three of the eight remaining courses corresponded to the lower part of the vault spring above F37.↩#p381
Interpretation of Room 10
Room 10 served as a storage room underneath the stairs (room 8) and therefore may be seen as a counterpart to room 4. Little material culture remained with which to suggest what was stored in this room.↩#p382
Because of the poor preservation of the upper elevations of the walls, it is unclear how room 10 was accessed. It is possible that it was once reached by means of a trap door through the stairs (room 8), but it is also possible that it was reached through room 4 and that this access was later blocked by wall F38. This suggestion appears possible, as the space between room 4 and room 10 was only blocked by an abutting wall stub F38, rather than a bonded wall integral to the house construction.↩#p383
The walls of room 10 were covered in mud plaster, although this mud plaster was often not well preserved. The mud plaster on the south wall (F26) was blackened, probably from the use of oil lamps in the vicinity. The foundations of the wall were visible and were consistent with the construction patterns found in other rooms.↩#p384
Like room 4, room 10 appears to have had a mud-brick floor, although it was in a very poor state of preservation. It appears to have consisted of only one layer of brick on top of rubble.↩#p385
No significant objects were associated with this room; whatever was present seems to have been removed upon the abandonment of this structure. The room was probably used for storage, since areas under the stairs were commonly used for this purpose.212↩#p386
Room 11 (0.71 m north–south x 0.79 m east–west)
This storage room is a sub-area within room 7 (Figure 4.1). The upper portion of this storage area seems to have had a functional role, but it was not sufficiently preserved for its purpose to be clear. Four walls,213 already visible from the surface, enclosed room 11, which was filled with stratigraphic units of varying types and reliability for a depth of 0.870 m (Figures 4.35, 4.36, 4.37).↩#p387
DSU 62214 consisted of windblown sand on the surface of room 11. This matrix had a moderate to high density of surface sherds of iron-rich fabric A group, as was common for the surface layers of B2. There were low densities of clinker, the only other class of artifacts represented in this context. The DSU covered mud-brick debris (DSU 65).↩#p388
DSU 65215 consisted of small fragments of mud-brick debris. This jumble of mud-brick, sand, and small quantities of artifacts appeared to have been a fill used to create a flat surface on top of the vault (F43) over a storage area. Artifact densities were low to moderate and contained low to moderate quantities of iron-rich fabric A group and calcium-rich fabric P37; a few small pieces of clinker were mixed into the mud-brick debris.↩#p389
Figure 4.35: Harris Matrix, room 11.
F43216 is the vaulted ceiling of room 11. It was in an excellent state of preservation and was complete. The entire top or plan-view of the vault was covered in grayish-brown mud plaster, which obscured the brick-laying pattern beneath. Cracks ran north–south along the feature, presumably marking the courses of brick. Mud brick and rubble in the south and the southeast corner cover the vault. This covering provided a stable space for a plastered channel that ran around the eastern and southern extremities above the vaulted ceiling. It was unclear how this channel was used or how this area should be reconstructed, given the poor state of preservation. The vault springing rests on top of the compacted surface inside room 11, rather than wall F32 or F35, which are used to support the vault spring laterally rather than vertically. The interior of the vaulted space was completely covered in mud plaster. Some organic material adhered to the bottom of the plaster here. It was probably used to help solidify the vault. The dimensions of the bricks in vault F43 were not visible.↩#p390
DSU 68217 consisted of sand under the vault (F43) in room 11. This windblown sand had moderate inclusions of mud-brick fragments. Artifact densities were low for all categories, including fabric A group and P37 ceramics, bone, glass, and clinker. There was one fragment of oasis red slipware. The matrix may be windblown sand and debris from outside of room 11 and therefore may not be reliable. The DSU was terminated when a compacted surface (DSU 72) was encountered.↩#p391
DSU 72218 consisted of compacted mud and served as an informal floor for room 11. The DSU itself was brown, but it has a moderate degree of ash pockets in its eastern quarter, most likely due to its close proximity to the oven feature (F19). DSU 72 was not excavated.↩#p392
F32219 is an east–west oriented wall forming part of the north boundary to room 9, the north boundary to room 11 and part of the north boundary to room 7. It was in a good state of preservation with the western portion preserved to a greater height than the eastern portion. The bonding pattern is English bond and the laying pattern followed the alternating channels method. The channel was 9 cm wide and varied greatly in length. The south face of the wall was covered in a reddish-brown mud plaster that had organic, pebble, ceramic, and lime spot inclusions. This mud plaster obscured most of the individual mud bricks, which were, for the most part, yellowish-brown in color, although there are some reddish bricks as well. The foundations of wall F32 were not visible, but the coving with floors F44 and F45 nearby was evident.↩#p393
F36220 is a north–south oriented wall forming the east boundary to the interior of room 11. It was in a good state of preservation. The entire east face of the wall was covered with mud plaster followed by a coat of white gypsum plaster that was less than 0.50 cm thick. This gypsum plaster was blackened in some areas, probably as a result of its close proximity to oven feature F19. The presence of plaster obscured both individual bricks and the bonding pattern for F36. There was a 56 cm wide aperture in F36 that provided access into room 11. It was 17–73 cm south of the corner with the north wall, F32, and 6–54 cm above floor F49. The aperture was worn through usage and followed the course of the vault ceiling (F43) in its upper parts. The foundations of F36 were not visible; its bricks extend below floor F19 and extend by 5 cm in the southeast corner for one brick length. The rest of the bottom bricks were flush with the wall face.↩#p394
Figure 4.36: Multi-context plan of room 11 showing FSUs (within rectangles), and elevations (with triangles).
F34221 is a north–south oriented wall stub forming the west boundary to room 11. It was in a good state of preservation, with the northern portion preserved to a slightly greater height than the southern portion. The south face of the wall was almost entirely covered in grayish-brown mud plaster, so that the bonding pattern was not visible in section-view on the wall faces. In plan-view it was clear that a variation on the alternate channels method of construction was used. The bricks visible in plan view were yellowish-brown and red in color. The mortar was grayish-brown. Wall F34 can be linked to the initial occupation of B2, since it is associated with the first floor level, F44. Its foundations are not visible.↩#p395
F35222 is an east–west oriented wall forming the south boundary to room 11. The wall was in a good state of preservation apart from the lowest course, which was undercut. Traces of mud plaster remained over the whole of the south face of the wall, with two areas of white plaster preserved, one on the upper west side and one on the east side, which formed a rounded corner with wall F36. The brick-bonding pattern was obscured by mud plaster. The join where this wall abutted the adjacent wall to the west (F34) was filled with gray-brown mud that was different from the wall plaster. The upper face of the wall was uneven and seems to have been preserved to its original height. The wall was two brick-widths wide, and the north row of bricks was covered in its east part by a channel. This channel was lined with a thick layer of mud plaster (0.7–1.0 cm thick), with a 0.5–1.0 cm thick white plaster. These plaster layers formed an L-shaped channel, which turned northwards to the west of F36 and above F43. The channel ended one brick-length from the adjacent wall to the west (F34), with the upper brick course offset slightly to the north to create a 7.5 cm lip that was coated in mud plaster. Both this plaster and that on the rest of the wall had high organic content.↩#p396
Interpretation of Room 11
Room 11 was a small storage area in the north part of room 7. It consisted of a vaulted, enclosed space with the space above the vault used for an unknown function. The storage area was accessed through a 56 cm wide aperture in the east of the feature. A door could not have closed this aperture, as no signs of any pivots were found.↩#p397
Figure 4.37: Photo of room 11.
The roof of this structure was found complete and consisted of a pitched vault over the open space. We did not excavate the vault itself so details of its construction are not known. The walls of the interior were mud plastered and contained large amounts of organic debris adhering to them. The walls of the exterior of the room were mud plastered and contained gypsum plaster in strips around the aperture as well as the top preserved courses of the feature. The foundations of the walls were not visible. The floor of this room consisted of compacted mud, consistent with that found in room 7, except that it was less compacted than the more formal floors of room 7.223↩#p398
Storage rooms of this type are common near oven features and can be found at comparable sites, such as Kellis. For example, House 1 at Kellis had two ovens, which had replaced earlier hearths, in close proximity to a storage bin.224↩#p399
The function of the upper portion of the feature is unclear. Rubble was placed above the barrel-vault to create a flat surface. A plastered channel rests on top of this rubble. This channel follows the south and east sides of this feature and probably would have continued along the north side. The rest of this feature is not preserved. I have not yet found a parallel that was sufficiently preserved so that its function was clear. A similar feature can be found at Ain Manawir in the Kharga Oasis. The greater context there was a temple precinct, but the feature is outside the sacred areas. The excavator did not know what its function was.225 It seems likely that this feature would have been used for the preparation of food or liquid. For example, the plastered area could have provided a surface upon which the bread loaves from the oven could cool. No significant artifacts were associated with this feature that could clarify its function.↩#p400
Remarks on the Phasing of B2
The textual and numismatic data from B2 were insufficient to support firm absolute datings for the phases of the house. Rather, at this time, most of the phases within the chronology of this house are relative within a more general absolute time.↩#p401
Phase 1: Construction and Original Use
Interior and exterior walls of were laid out. House B2 was built upon a cemented sand dune with scatters of Dynastic artifacts. The foundation bricks were laid directly upon clean sand. Likewise, the deep sondages in room 2 and room 6 that revealed clean, compacted yellow sand below the house walls. The construction episode occurred during the early–mid third century, to judge from the ceramics found underneath floor depositions.226↩#p402
This construction episode appears to have been part of a large-scale development in Area 1. The alternating channels method, for example, is often a sign of large-scale construction since it is more malleable than more rigid brick-laying patterns.227 It is also a cheaper method because it is possible to use fewer bricks in favor of mortar and rubble. This construction pattern can be found from at least the Middle Kingdom onward and is common in Fayum buildings.228 This construction method will be discussed in greater detail in chapter 5.↩#p403
The first phase of the preserved room floors also can be linked to the initial occupation phase. This supposition is based upon analyses of floors and occupational debris above floors. It is clear that room 5 was used as a food preparation area during the initial phase of occupation.↩#p404
Phase 2: Structural Alterations
Phase 2a: Probably quite early in the occupation of B2, a bread oven (F19) was added to room 7. Feature F19 seems to be built after the original occupation of B2, since the area where it abuts wall F4 to the east covers the white gypsum plaster on F4. It was added during the period of use of the first floor level, however, because it is plastered into floor F49, the original floor level for room 7.↩#p405
Phase 2b: Minor floor and wall repairs take place, particularly in rooms 5 and 7.↩#p406
Phase 2c: New floor levels are put into place over original floor levels. The stairs (F50) to the street (S1) are added.↩#p407
Phase 2c: A final floor layer (F17) is added to room 7. In room 5, repairs are made to the west wall (F14), a new floor layer is added (F41), and an informal hearth is put into place above the previous hearth (F46). A wall stub (F33) was added to room 9. This wall was associated with floor level F39 and not with floor F44 (the original floor), because it is built on top of rubble, DSU 74, which, in turn, rests on top of F44.↩#p408
Phase 3: Abandonment, Collapse, and Deflation
Phase 3a: B2 is abandoned. It may have witnessed some casual use as indicated by some signs of burning on a wall (F4) in room 1 that are above the floor surface. During this phase some windblown sand blew into some of the rooms.229↩#p409
Phase 3b: There was a phase of considerable collapse evinced by the presence of ceiling collapse in some rooms.230 Some wall collapse is also evident (room 7).↩#p410
Phase 3c: The structure is partially covered in windblown sand, but this sand cannot have been in substantial quantity, since this would have protected the structure from wind damage. Rather, a dramatic deflation process occurred that was caused by constant wind erosion, evinced by the smooth and consistent deterioration evident in the east–west section of the northernmost portion of B2 (Figure 4.2). This deflation clearly occurred after ceiling collapse, since the collapse in all of the rooms was flush with the preserved walls, having eroded at the same rate as the walls post-collapse. This erosion process was particularly evident in room 1, while room 3 was the room most damaged by erosion.↩#p411
Phase 4: Bedouin Use Present-day
The Bedouin from the village north of Amheida herd their animals across the easternmost end of Area 1. They do not seem to have disturbed this house, and there are no signs of significant intrusions in the layers. The herding of animals may have accelerated the erosion process, but most of the animal herding appears to take place in a consistent path much further east of the present structure. Indeed, the hill created by the archaeological deposits represents something of a barrier to herding in this area.↩#p412
Street S1 (Subarea 1.2)
Street S1 is located north of B2 and functioned as a major east–west oriented street into Amheida. We laid out a 6.10 m east–west x 2 m north–south oriented trench in this area, aligning it against the north wall of B2. Three surfaces were exposed (DSUs 6, 10, 13), all composed of compacted mud. The DSUs start from 0 because subarea 1.2 is a changed area (Figures 4.38, 4.39, 4.40, 4.41).↩#p413
DSU 0231 was the surface of the street area that was cleaned in 2004. This area was located immediately south of Area 1.1 and is east of B2. The area was not excavated, but the upper levels were cleared of ceramics and other finds. These finds included a high density of iron-rich fabric A group, and low densities of calcium-rich fabric B group. There were also some very small bone fragments, high densities of glass, and moderate densities of faience.↩#p414
DSU 1232 consisted of soft yellow windblown sand on the surface of Street S1 and within our exploratory cut. The DSU sloped down dramatically to the east. It is unclear if this slope was the result of deflation, archaeological deposits, or the natural variations in the site surface. This surface context contained high densities of iron-rich fabric A and A4 as well as moderate densities of calcium-rich fabric A11 and B. There were high densities of clinker, glass, and bone, and low densities of wood, mud-brick fragments, and charcoal parts. A fragment of a glass bracelet (Inv. 11606) was also found. We terminated the DSU when collapse (DSU 3) became visible. We also wanted to distinguish surface sand (DSU 1) from below-surface windblown sand (DSU 2). This context was not secure.↩#p415
DSU 2233 consisted of soft yellow windblown sand with a small charcoal patch that included small fragments of wood and charcoal. DSU 2 contained a small amount of disarticulated mud-brick debris and a smaller number of sherds than found in the surface layer (DSU 1), as would be expected. These ceramics consisted of higher densities of iron-rich fabric A group and lower densities of calcium-rich fabric A11 and B group. A few of the sherds were quite large. There were some bone fragments and some clinker. The mud brick most likely derived from the collapse matrix of DSU 3. DSU 2 was terminated when we reached the denser collapse mass of DSU 3 and the more compact brown sand of DSU 4. This context was mostly found in the western portion of the trench, but there was also a small, disarticulated part at the eastern end of the trench.↩#p416
DSU 3234 consisted of mud-brick collapse from walls and a ceiling. Bricks with semicircular marks evinced the ceiling collapse, while plain bricks signified the wall collapse. The collapse likely derived from the north wall of B2 and the roof that once existed over room 9. It is also possible that the street itself was roofed, although it seems unlikely given the street width and lack of substantial collapse debris.235 Some mud bricks showed signs of burning, and there were also signs of ash in a small patch, which may have resulted from later ephemeral secondary usage. The collapse on the east side of the trench was more “melted” and less distinct than on the west side. Over a dozen vault bricks were eventually found among the wall bricks, mostly close to the north wall of B2. Mixed in with the bricks was a considerable amount of small, broken, gravelly, brick debris. Many levels of bricks were superimposed. There were high densities of iron-rich fabric A and A4 group, and moderate densities of calcium-rich fabric A11 and B. Of note is that seven sherds were covered in a resinous, tar-like black substance. Glass, bones, and dross occurred in moderate densities, while clinker occurred in great densities. Much of this refuse probably originated from the kilns north of B2. There was one apricot pit. Two terracotta figurine fragments were found (Inv. 11628 and Inv. 11633). The bone, some of which showed signs of burning, appeared to be from a single large animal, and it is possible that it derived from the modern village north of the site. At this stage, however, it is not possible to confirm or deny the contamination of this context. Small objects were found just beneath the collapse layer, in DSU 5. DSU 3 was terminated when no more dense associations of mud brick were found and when DSU 5, a sandy matrix with many inclusions of small sherds and clinker, was found. A soil sample of coprolites was taken (S2).↩#p417
DSU 4236 consisted of medium yellowish-brown sand with few or no inclusions. When first exposed, it appeared in small patches, primarily in the southwest corner of the trench. It remained in patches in DSU 3, ultimately giving way to DSU 5, which had more inclusions and spread across the entire trench. This context had moderate densities of iron-rich fabric A and A4 group and moderate densities of calcium-rich fabric B group. Bones were present in moderate densities, and clinker was in low to moderate densities.↩#p418
Figure 4.38: Harris Matrix, S1.
DSU 5237 consisted of a small patch of soft brownish-yellow sand with high numbers of inclusions, including small fragments of clinker, charcoal, and ceramics. The clinker was small and crumbly, a possible indication of high traffic across this matrix. The matrix to the east held fewer inclusions than that to the west. The matrix was also sandier in the southern half than in the northern half, and was visible first. The DSU was quite mixed in content and somewhat irregular in section, with the northern portion much thicker than the southern portion. DSU terminated when highly compacted matrices of adobe (DSU 7 and DSU 9), a more compacted matrix with many inclusions (DSU 6), or ash (DSU 8), were encountered.↩#p419
Figure 4.39: Multi-context plan of Street S1 showing DSUs (encircled), FSUs (within rectangles), and elevations (with triangles).
The complete head of a terracotta figurine (Inv. 11627) was found at the very top of this layer, directly underneath the mud brick collapse. The presence of two fragments of a single figurine distributed between two contexts238 indicates a degree of interpenetration for these two contexts. Among the ceramics, we found a high density of iron-rich fabric A and A4 group and low to moderate densities of calcium-rich fabric A11 and B. There were two A-group over-fired body sherds that appeared to be melted; these probably derived from the kilns across the street. There was also one A36 Nile Valley imported body sherd, and a 75% complete bowl (Inv. 11734) was also preserved. Faience and glass occurred in low densities. Bone occurred in moderate densities and clinker occurred in high densities. A large number of ash pockets were found as well as a donkey hoof. A Greek ostrakon (Inv. 11626) mentioned donkey driving and wheat. Charcoal, mortar, eggshell, and olive pits were also found in low densities.↩#p420
DSU 6239 consisted of highly compacted adobe (either melted or in situ) with high densities of inclusions such as gravel, clinker, and ceramics. All of the inclusions were quite small, signifying heavy wear and traffic over this context. DSU 6 grew increasingly compact and more intensely yellow with depth. In the northeastern quadrant of the context there was a large ash pocket with a sherd scatter, consisting of large sherds. It was probably a secondary trash deposit. It may have been a built-up surface over a more pronounced street surface (DSU 14). Iron-rich fabric A group occurred in great densities and calcium-rich fabric B group in moderate densities. Glass and plaster occurred in low densities. Bone and clinker occurred in moderate densities. Mortar and one complete apricot pit were found. One poorly preserved ostrakon (Inv. 11712) was found here, which may be a tag from a jar, referring to a well.↩#p421
DSU 7240 consisted of highly compacted adobe with flat-lying sherds both on top and embedded in it. It appeared to be a surface of some kind that was probably intentionally built in order to keep refuse from Street S1 out of the house (B2), or it may have served as a bollard to protect the house walls. There is a parallel compacted adobe surface on the west side of the entrance to B2 (DSU 13). These contexts may have served as repairs to the wall surrounding the door to House B2 (F50), since both DSU 7 and 13 cove up over the north wall (F32) of the house. This context was not excavated.↩#p422
DSU 8241 consisted of a soft layer of fine gray ash with high densities of inclusions, ceramic scatters, and large sherds. The southeast corner maintained the yellow windblown sand characteristic of DSU 5. The presence of misshapen ceramic “wasters” (e.g., a handle collapsed onto a body) may indicate secondary waste deposits from the kilns north of Street S1. Similar wasters were found in DSU 5, which may indicate that the kilns were contemporaneous with the house during at least part of its occupation or, at the very least, with the street usage. These ceramics likely date to the third century. This context contained high inclusions of organic material, especially bone, and clinker. Plaster occurred in low densities, and one fragment was painted dark red and light green. Iron-rich fabric A and A4 group occurred in high densities. Calcium-rich fabric B group occurred in moderate densities. Wood and plaster occurred in low densities. Charcoal, several olive pits, and mortar were found. This context was terminated when either the compact surface (DSU 10) or another ashy DSU (11) was encountered. The change was made primarily on the basis of position in order to signify that deposits in DSU 8 are above DSU 10.↩#p423
DSU 9242 consisted of a silty brown matrix over highly compacted adobe. There were high densities of medium-sized potsherds, as well as moderate densities of clinker, glass, and bone, and low densities of organics such as seeds and charcoal. DSU 9 appears to be occupational debris from immediately above street surfaces (DSU 6, DSU 13). Iron-rich fabric A group occurred in high densities while calcium-rich fabric B group occurred in low densities. There were many salts adhering to the ceramics from this context, suggesting that there was some wetness in the area at some point. Bones occurred in high densities. A fragment of the body of a terracotta figurine was found here (Inv. 11915).↩#p424
DSU 10243 consisted of an extremely hard matrix of compacted trash. The surface broke apart in large chunks and had several plaster inclusions, especially in the northwest corner of the context. The context was extremely flat, except for the western portion, which slopes upwards slightly, and seems to have been a prepared street surface, perhaps the lowest (first) street level. As such, it is probably equal to DSU 14 to the west. The thickness of this context was difficult to ascertain, because it was present only in a small northeastern portion of the excavated area in the trench. The southernmost 50 cm appear to have been excavated earlier in a trench from a previous campaign that we have not been able to identify.244 We drew a section of the east sidewall of the Area 1.2 trench with this prior trench visible in the baulk (Figure 4.40). The ceramics included iron-rich fabrics A and B in moderate densities. Glass and bones occurred in low densities.↩#p425
DSU 11245 consisted of a soft gray ashy matrix, with numerous inclusions, including many sherds from iron-rich fabric A and A4 group. Calcium-rich fabric B occurred in moderate densities. Glass and iron occurred in low densities, and there was a considerable amount of clinker and bone fragments. There was some glass and a handful of baked brick fragments. This context was very similar to the gray ash above (DSU 8), but a distinction between the layers was made on the basis of position; DSU 8 is above a “surface” (DSU 10), while DSU 11 was below DSU 8. DSU 10 was never present in the southeast quadrant of the trench. We believed that this area of the trench was contaminated, and thus this context, unlike DSU 8, was probably contaminated.↩#p426
Figure 4.40: Street S1 section, showing trench cut, as well as the door (F50) into House B2.
DSU 12246 consisted of soft, loose yellow sand. It was uncovered in 2 sections, an east–west section just north of the north wall of B2 and a north–south section just west of the eastern border of Street S1. This context contained a ceramic scatter in the eastern edge of the east–west section, which yielded a vessel with two spouts (Inv. 11933) and a bread plate (Inv. 12216). The western and northern areas of the context were much cleaner, with very few inclusions. Another possible kiln waster was found in the context, which extends the active period of the kilns back in time, as it was once conjectured that the kilns were only a very late addition on this street.247 Our other ceramic finds included iron-rich fabric A group in very high densities and calcium-rich fabric A11 and B in moderate to high densities. Wood occurred in low densities, while limestone, clinker, and dross occurred in moderate densities.↩#p427
DSU 13248 consisted of a hard brown compacted mud matrix with high densities of inclusions, especially plaster. There were some quite large sherds from this matrix, but the densities of iron-rich fabric A group and calcium-rich fabric B group were low. Other inclusions were bone and charcoal in low to moderate densities. This context probably represented the last proper road surface in Street S1 and may also have served as a sort of trash barrier for the doorway to B2. It thus seems that this context is a counterpart to another compacted matrix (DSU 7). The context was terminated when a highly compacted surface was found (DSU 6). The southernmost portion of DSU 13 adhered to the north wall of B2 (F34), and this portion was not removed. The northernmost 60 cm of the context were not excavated.↩#p428
DSU 14249 consisted of compacted adobe. It was a highly level surface, with flat-lying sherds on top, suggesting it was probably a prepared street surface. Artifact densities for DSU 14 were low in all categories, although many categories were represented. Iron-rich fabric A group occurred in high densities and calcium-rich fabric A11 and B in low densities. Faience and glass occurred in low densities, while clinker occurred in moderate densities. Many of these artifacts had salts adhering to them, suggesting that there was wetness in this area. This situation is similar to that found in DSU 10, and it is possible that they resulted from the same depositional episode. DSU 14 was terminated when we reached soft yellow sand and occasional pockets of ash.↩#p429
DSU 15250 consisted of clean yellow sand. It was almost entirely devoid of artifacts, especially after the first 5–10 cm. There were low densities of ceramics from A group and A11. There was also one piece of unbaked A group ceramics, again possibly indicating active use of the kiln across the street. The sand was finely laminated, showing thicker levels to the north than to the south. This context was terminated when we reached what appeared to be the lowest elevation of the north wall to B2 (F34).↩#p430
F1251 was the first (top) street layer for Street S1. It was highly variable in elevation and thickness and probably represented a means of holding back trash from B2, since it was thickest near the doorway into B2. The surface was quite uneven and was covered with a dense lens of small potsherds, much like F3. DSU 5, located between the two raised portions of F1, probably also served as part of the street. There were two patches of F1 on either side of the door to B2. Each was approximately 76 x 200 cm with thicknesses of 10–20 cm. DSU 5 may be street level F1 also, but it had a different composition than F1.↩#p431
F2252 was the second (from top) street layer for Street S1. It was light yellowish-brown in color and was composed of compacted mud. It also contained high densities of plaster fragments. The thickness and elevation of F2 was highly variable, varying between 0.5 cm and 15 cm or more. The thickness increased to the west and may reflect an attempt to even out a naturally hilly topography. In the southern portion of the trench, F2 was missing from the southernmost 1 meter of the trench. The lack of F2 in this area may be due to a trench dug in this area by a prior expedition to this site (see the description for F3 below). We removed the northern portion of F2 in the part of the trench because it appeared to be a patchy lens in DSU 8. To the west, the feature was quite thick and was built directly on top of F1, with no rubble or sand in between. To the west, it was also greater in elevation. The white plaster fragments continued throughout F2.↩#p432
Figure 4.41: Photo of Street S1.
F3253 was the earliest street level for Street S1. Only a small portion was fully exposed in the easternmost part of the trench. It was dark gray in color and composed of compacted mud with no distinguishing features other than the high density of small potsherds adhering to its flat surface. A trench of 0.50 m appears to have been cut through F3 in the south. We do not know who cut this trench. Because of this missing 0.50 m we cannot definitively state that F3 abutted wall F32 of B2, although marks of wear on F32 suggests that this abutment is likely.↩#p433
F57 from B2254 is an east–west oriented wall forming the south boundary to the street just west of the house. Only the north face and the easternmost 75 cm are described here. It was in a moderately good state of preservation, although it was not preserved to a great height. It was abutted by wall F31 to the east. A large gap filled with mud, 10–15 cm wide, joined the two walls. It seems that the wall was repaired over time in this vicinity, probably because of wear from heavy usage of the street. At 18–50 cm from the east end of the wall there was a gouge 17 cm high and 4 cm deep, which probably also signifies street wear. There was no mud plaster remaining on the facing of the wall, and the bonding pattern was clearly visible as English bond. The foundations and their bonding pattern were not exposed.↩#p434
F32 from B2255 is an east–west oriented wall forming the south boundary to Street S1 and the north boundary to the house. The north face is well preserved, but not to a great height. The western portion is better preserved than the eastern. At 150–320 cm from the west end of F31 there is a long gouge 50–78 cm above the foundations that is at approximately the same elevation as F3, the bottom street level. F3 probably once abutted F32, but an unknown excavator removed it recently. At 184 cm from the west end of F32 there was a large crack approximately 1 cm wide. Other small cracks were also visible. Only a small quantity of mud plaster remained on the wall, and it was present only in the westernmost 50 cm of the wall in small patterns. The bonding pattern was clearly visible as English bond. The foundations were exposed; they protruded north of the wall face by 5 cm. The top course consisted of headers facing out and the bottom with stretchers facing out. Only two courses were visible. A foundation trench was visible in section view. It showed that the wall was built into windblown sand and that it was built at the same time as Street S1 became active (visible in F3 of S1).↩#p435
Interpretations of the Street S1 Exploratory Trench
Street S1 was the street outside B2. Multiple street surfaces were exposed, indicating intensive use and maintenance over time. Barriers were found around the entrance into B2. The last (top) surface (DSU 6) had elevated sections (DSU 7) on either side of the doorway into B2, indicating that they may have been used as barriers to hold back the refuse that filled the lower areas on top of this street surface.↩#p436
The entrance (F50) into B2 was visible to the south. The entrance was 1.47 m wide. It was not decorated in any manner.↩#p437
The walls were mud plastered, although this mud plaster was not well preserved. The foundations of the walls were clearly visible and were consistent with techniques commonly used in this structure. A foundation trench for the south wall of the street (F32) was found, indicating that this north wall of B2 was constructed contemporaneously with the Street S1. This evidence substantiates the hypothesis that Area 1 was laid out in a single phase of construction. The details of this construction are given in chapter 5.↩#p438
The street layers consisted of informally packed earth with high densities of refuse packed into the surface. There were multiple layers of these surfaces. It is possible that some minor works were conducted to even out the street surfaces but that these works were not nearly as formal as floor construction.256↩#p439
The refuse found in the street contained several noteworthy objects: several terracotta figurine fragments,257 donkey hoofs, and an ostrakon in Greek that dates to ca. 240–275 (Inv. 11626, O.Trim. 1.36). This ostrakon was found embedded in DSU 5, a secure context representative of the last phase of frequent usage. Moderate densities of ceramics, clinker, bone, and glass were also represented. Ash and high densities of objects were found between the third street level and clean yellow sand, among them several reconstructable ceramic vessels, including the cover of a bread oven. The quantities of clinker and wasters suggest that the kilns north of B2 were in use throughout the occupation of B2 and the usage of Street S1.↩#p440
Remarks on the Phasing of Street S1
Phase 1: Construction and Original Use. The street appears to have been laid out in a single large-scale construction effort, as evinced by the foundation trench for F32. Likewise, the alternating channels construction method used for the walls tends to indicate large-scale construction.258 The planning pattern of this area supports this suggestion, as the plan of this quarter is quite regular and the single buildings are very similar to each other in plan, dimensions, materials, and techniques.↩#p441
There were three different walking surfaces that raised the street level over time. Each of these surfaces abutted House B2. The fill between these surfaces consisted of waste and ash.↩#p442
Phase 2: Structural Alterations. DSU 7 added to keep street refuse out of B2; it can be associated with F1 and possibly F2.↩#p443
Phase 3: Abandonment, and Deflation The abandonment of the street is difficult to assess since it may have been in use beyond the occupation of B2, although this seems unlikely given that the stairs from B2 lead directly up to the top street level and all of the surface ceramics in this area of Amheida seem to belong to the late third and early fourth century CE. As with B2, the wind dramatically deflated the walls on either side of the street throughout Area 1, particularly towards the east.↩#p444
Phase 4: Bedouin Use Present-day The Bedouin from the village north of Amheida herd their animals across the easternmost end of Street S1, but they do not seem to disrupt the area.↩#p445
Courtyard C2 (Subarea 1.4)
Courtyard C2 was described initially as a large L-shaped open space south and east of B2. It was thought to be an exterior courtyard belonging to B2, since the entrance to it is located just east of B2, and B2 was the house in the closest proximity to C2. Alternatively, we thought that the courtyard could have been a communal or shared courtyard, since it is surrounded by houses on all sides and is the only obvious large open space within the immediate vicinity of Area 1. Subsequent excavations in C2, still in progress as this volume went to press, reveal that it was a rectangular shape. A house (B9), identical in size and similar in layout to B2, occupies the area east of B2.259 Courtyard C2 belongs to House B9 and not to B2. Even so, our test trench informs interpretations of B2. We chose to excavate a trench (C2A) in Courtyard C2 in order to determine the function of a space in the immediate vicinity of B2 as well as determine comparative phasing with the house and the street.↩#p446
The walls of Courtyard C2 were clearly delineated, with the exception of the southeast corner. We laid out a 6 m north–south x 2 m east–west oriented trench against the south wall of B2 and the west wall of Courtyard C2, since wind deflation has a lesser effect towards the west and north. This trench was later expanded by an additional 4.2 x 2.0 m to the south and 1.5 x 1.5 m to the east, against the south wall of Courtyard C2. We expanded the trench because additional features (F4 and F5) became visible in the course of excavating the 4.2 x 2.0 m trench (Figures 4.42, 4.43).↩#p447
DSU 1260 consisted of soft yellow windblown sand. The top centimeters contained high densities of ceramics in medium sherd sizes. The density of sherds decreased below the surface. The surface topography of this context was uneven and sloped down to east and south, showing a similar preservation to B2. There were high densities of iron-rich fabric A group and A4 group and moderate to high densities of calcium-rich fabric B group. There were low densities of faience, glass, bones, and plaster and high densities of clinker. We also had a large number of small finds, including an ostrakon in Greek referring to year 30 (AD 335/6 or 353/4) (Inv. 11773), a terracotta animal figurine (Inv. 11914), another terracotta figurine (Inv. 11916) and a largely complete lamp (Inv. 11903). We terminated the context when a silty, grayish-brown matrix (DSU 2) became visible.↩#p448
Figure 4.42: Harris Matrix, C2A.
DSU 2261 consisted of a grayish-brown sandy loam with high organic content. It was covered by surface sand (DSU 1) and also wall and ceiling collapse (DSU 8). The southernmost 400 cm of this context was secure because it was covered by wall and ceiling collapse (DSU 8). We took two soil samples, S1 and S5, because there are high densities of organics in this matrix, such as straw, wood, seeds, pits, and also donkey and camel dung. As with DSU 1, the topography was uneven, with the elevations higher to the northwest than the northeast. We found the iron-rich fabric A and A4 group in moderate to high densities, as well as moderate densities of calcium-rich fabric B. Glass and wood were found in low to moderate densities, and there were moderate densities of bone. We also found an ostrakon in Greek that is too worn to read (Inv. 11774). The northwest and southeast corners have small quantities of windblown sand mixed in with DSU 2, but not in sufficient quantities so as to justify a separate DSU.↩#p449
DSU 3262 consisted of a silty loam with high organic content. It did not have as many coprolites or dung as DSU 2, but both inclusions were present. Artifact densities were low to moderate in the categories of glass, wood, clinker, and bone. There were moderate to high densities of iron-rich fabric A and A4 group and moderate densities of calcium-rich fabric B and A11. There were a number of small finds from this context, including two ostraka in Greek, one of which is an account of wheat and barley (Inv. 11723), mentioned above, and the other was a well tag (Inv. 11717) dated to a year 2 that is unlikely to be later than 285/6. The other small finds included a complete miniature lamp (Inv. 11917), an incomplete small jar (Inv. 11902), and an incomplete globular flask (Inv. 11738). In this area, the DSU was terminated when the yellower, sandier matrix, DSU 4, was encountered. A small feature (F2) was exposed in the northeast corner of the trench.↩#p450
F2263 is an east–west oriented wall in Courtyard C2. It does not seem to follow the same orientation as wall F56, but since only two courses are preserved it is difficult to tell if this orientation is accurate. The bricks of the bottom course rest with their headers facing out, and the top course of bricks rests with stretchers facing out. No mortar was used, and there was no plaster facing. It was built on sand with no foundation visible. Only 73 cm oriented east–west were visible. It is probable that it was a small dividing wall. It was not well constructed.↩#p451
DSU 4264 consisted of a brownish-yellow layer of sand with some organic content. Like DSU 3, it did not have as many coprolites as DSU 2, but a significant number were present. This DSU also contained low to moderate densities of glass, bones, seeds, clinker, and some pottery fragments from iron-rich fabric A and A4 group as well as calcium-rich fabric A11 and B. There were a number of small finds, including two loom weights (Inv. 11882, Inv. 11880) and a fragment of a handmade figurine head of a woman (Inv. 11920). We also found the desiccated remains of doum fruit, which correlates with a clay tablet (Inv. 11048) containing an account of these fruits, which we found in B2. We terminated the context when a compacted mud surface (DSU 6) and a deposit of clean yellow sand (DSU 7) became visible. Because the context had a mixture of organic material and clean sand, it is possible that the DSU 4 was a mixture of several contexts (DSU 3, DSU 4, and DSU 7).↩#p452
DSU 5265 was a circular ephemeral hearth. It is gray and has a rough surface, and the context cut DSU 4. It contained high quantities of gray ash, as well as charcoal fragments and burned wood. Due to its high organic content, we took a soil sample S3. This context was terminated when no more ash was visible and mostly clean yellow sand (DSU 7) became visible. It is presumed that DSU 6 was destroyed in this area due to the use of this ephemeral hearth. This context was probably associated with DSU 6 = F1, but once the ash was removed only DSU 7 was visible. There were no objects within this matrix.↩#p453
F1 = DSU 6266 was a highly irregular and poorly preserved floor or surface layer that only remained in patches across the trench. It was very thin in parts and uncertain in some patches, so it may be an informal surface in Courtyard C2, rather than a prepared floor surface. It was underneath room fill (DSU 4) and an ephemeral hearth (DSU 5).↩#p454
DSU 6 = F1267 was a compacted yellowish-brown surface with moderate organic content, defined as a floor surface (F1). It was well preserved only in the northernmost corner of the trench. This corner was the only area where trash and sherds were visible under the floor surface. Several concentrations of ash could be associated with the surface of DSU 6, most notably DSU 5. We removed the majority of this context in order to understand the stratigraphy of the trench better. The best-preserved part of the floor was the northernmost portion, which was left in situ. We found clean sand (DSU 7) underneath the context and therefore terminated it.↩#p455
DSU 7268 consisted of soft, largely clean yellow sand with few artifacts. Most of the artifacts were found within the top 10 cm of the context. These objects might have been placed under the floor surface (DSU 6 = F1) as a preparation layer for the floor. Hard areas were present towards the southwest corner of the trench, but these areas were probably related to the floor surface (DSU 6 = F1), and for this reason DSU 7 was cut by DSU 6. This context contained high densities of iron-rich fabric A and A4 group. There were low to moderate densities of calcium-rich fabric B and A11 group ceramics, as well as charcoal parts, seed pits, wood, and dross. There were moderate densities of bones and clinker. We also found a few objects, such as a complete carinated cooking pot (Inv. 11765) and an unfired loom weight (Inv. 11881). The ceramics appear to date to the third century. We terminated the context when we reached below the foundations of the west wall.↩#p456
DSU 8269 was a wall and ceiling collapse consisting of mud-brick debris. The ceiling collapse could be recognized by concentric circle designs on the bricks, and the plain bricks made it clear that there was also wall collapse. This collapse may derive from the unexcavated structure to the west of Courtyard C2. Because of the small amount of concentric circle design bricks, it is presumed that DSU 8 mostly consisted of wall collapse. The mud brick fragments of DSU 6 must have interpenetrated DSU 2. This is clear because the brown sand of DSU 2 mixed with the organic, softer soil of DSU 3 in patches throughout the area. DSU 8 contained low densities of glass and a high density of plaster. Iron-rich fabric A and A4 groups occurred in low to moderate densities, while there were low densities of calcium-rich fabric A11 and B groups. We terminated the context when the brown organic matrix of DSU 2 became visible.↩#p457
DSU 9270 consisted of gray compacted ash with charcoal parts. It was similar in appearance to DSU 5, but it was not the same DSU, since 5 represented a singular deposition: an ephemeral hearth. By contrast, this context appears to have been a diffuse deposit of ash and it had less-defined boundaries. In addition, this context mixed slightly with an organic brown matrix (DSU 3). The iron-rich fabric A and A4 occurred in moderate densities, while calcium-rich fabric B occurred in low to moderate densities. Wood, charcoal, and bones were found in low to moderate densities, and we had one apricot pit. This context was terminated when the browner organic matrix of DSU 3 became visible or more defined ash became visible (DSU 10 in F3, DSU 11 in F4).↩#p458
DSU 10271 was the ashy fill of an oven platform (F3). It was dark gray and contained many charcoal fragments. It had low densities of artifacts in all categories, although low densities of iron-rich fabric A and A4 ceramics and organics were present. There were some loose mud brick and sand inclusions mixed in with this context. The bricks probably derived from F3 itself and collapsed into the ash. DSU 10 covered F3.↩#p459
DSU 11272 was the ashy fill of a bread oven (F4). It was dark and contained many charcoal fragments. It had low densities of artifacts but several large pieces of ceramics were recovered. Some of these ceramics appeared to derive from the circular ceramic interior of the bread oven. There were low densities of iron-rich fabric A and calcium-rich fabric B. Others were blackened and may have been used in the oven. DSU 11 covered F4.↩#p460
F3273 was a platform for the oven (F4) to the east (Figure 4.42). It was below room fill (DSU 9) and ash (DSU 10). It was poorly preserved, with only 1 course visible above clean sand below it. It was built into the corner of F56 and F5. It consisted of a small dividing wall, 99 cm long east–west, built to the north, and another 96 cm long north–south wall built to the west. Each wall was only 1 course high and 1 brick thick. The base of F3 consisted of blackened mud bricks in no clear bonding pattern. The northwest 20 x 32 cm was not preserved.↩#p461
Figure 4.43: C2, F3 (platform), and F4 (bread oven).
F4274 was a poorly preserved bread oven (Figure 4.43). Only one course was preserved for most of it, and only two courses were preserved on its east wall. Together with F3 to the west, F4 was built into the southwest corner of Courtyard C2. Including all of the walls, the feature itself consisted of a rectangle of mud-brick 138 x 102 cm. This feature was found below room fill (DSU 9) and was filled by oven fill (DSU 11). The diameter for the circular interior of the oven was approximately 85 cm, but it was very poorly preserved and present only in patches. The area between the circular interior and the rectangular mud-brick exterior was filled with ash and mud brick. The draft hole of the oven was visible in the east wall of the feature. It was 7.5 cm from the bottom of F4, as measured from the exterior. The top part of the draft hole was missing, so it was not possible to say how high up it extended. The north–south oriented draft hole was 38–63 cm north of F5 (the south boundary of the feature). The platform (F3) for oven F4 was preserved to the west. Together the two features measure 106 cm north–south and 225 cm east–west.↩#p462
F5275 is an east–west oriented wall forming the south boundary of Courtyard C2. It was in a poor state of preservation, with only three or possibly four courses preserved above the foundation level. It was better preserved in the west than in the east. Only 315 cm of the western portion of F5 was exposed. Only the north face of the wall was exposed during excavation, and it is this face of the wall that will be described at this time. F5 abuts wall F6 to the west. There is a slight gap between the two walls, from 3 to 5 cm, where the abutment occurs. The wall was built using the alternating channel construction method. No plaster remained adhering to the north face of the wall, so the bonding pattern was clearly visible as English bond, although it becomes irregular approximately 85 cm east of the abutment between walls F5 and F6. Much of F5 was obscured by the presence of an oven platform (F3) and oven (F4) that abut F5. The foundations of F5 were not reached.↩#p463
F13276 is an east–west oriented wall forming part of the north boundary for Courtyard C2. The south face was moderately well preserved and was preserved to a greater height to the west than to the east. No mud plaster facing was currently visible, so the bonding pattern is clearly visible as English bond. F13 abuts wall F56 and it was shaped in a step pattern around the foundations and an indention west of F56. This stepping pattern reflects the fact that F56 was built first and that F13 was constructed with the existing construction pattern of F56 in mind. The abutment between the two walls was fairly well constructed, and mud plaster was used to fill in any 2–3 cm gaps. The foundations of F13 were not visible.↩#p464
Figure 4.44: Multi-context plan of Courtyard C2 showing DSUs (encircled), FSUs (within rectangles), elevations (with triangles), and small finds (without symbols).
F56277 is the west wall of Courtyard C2. It was in a moderate state of preservation, but very little of the height of F56 was preserved. Small quantities of gray-brown mud plaster adhered to F56 across its east face. The bonding pattern was English bond, and the wall followed the alternating channels construction method. F56 exhibited a foundation construction pattern so far not found elsewhere in Area 1. It was 4 courses high in most places, but near wall F13 it had 6 courses. From the bottom to the top of the foundations the pattern consisted of: (1) headers out; (2) headers out; (3) headers placed vertically; (4) stretchers out; (5) headers out from the wall face. Starting at 6 courses from the top of foundations, the wall protrudes by 10 cm into the courtyard. The 3rd–6th (from bottom) courses for the foundations protruded an additional 8 cm and the 1st and 2nd courses of foundations (from bottom) protruded an additional 21 cm into the courtyard.↩#p465
Interpretations of Courtyard C2A Exploratory Cut
C2 was an open area exterior to B2 and entered from an unexcavated house east of B2 (B9). This area was only partially excavated in 2007. During excavation, we exposed a compacted surface. Debris associated with these surfaces indicates animal husbandry and food preparation. A bread oven (F4) and platform (F3) were found in the southernmost part of the trench (the southwestern corner of Area C2).↩#p466
The entrance to B2 was through the south wall of House B9. The courtyard and House B9 are currently being studied further to understand their relationship, as well as their relationships to House B2↩#p467
The walls of this area were mud plastered, although much of this mud plaster was not well preserved. One of the walls (F56) shows a different technique of laying the foundations than the other walls excavated in B2 and may indicate different construction phasing or usage of the wall in this vicinity.278↩#p468
There was not a proper floor in C2 but rather an informal compacted surface. It was very rough but also quite firm. It appeared much denser than the street surfaces but less so than the surfaces in B2.279↩#p469
C2 was a courtyard area external to House B2, but belonging to House B9, east of B2. It appears that animal husbandry took place here, as evinced by the large amounts of coprolites excavated. It is also clear that food preparation took place here, as evinced by the bread oven and platform excavated (F3 and F4). The large quantities of rodent coprolites associated with these features also indicate in situ use of grain or other food of interest to the rodents. These findings are consistent with common usage of exterior courtyard areas.↩#p470
The C2 depositions consisted essentially of sand overlying occupational depositions and one poorly preserved compacted mud surface. The occupational debris consisted of significant deposits of dung and other organics. The coprolites of donkeys, camels, and cows were represented in addition to apricot, olive, and date pits. In addition to these organic remains, an oven and an oven platform were found built into the southwest corner of the courtyard. They were in a poor state of preservation, with only the bottom course of bricks remaining. The northern 6 x 2 m of the trench contained low to moderate densities of material remains and included a complete bowl, a loom weight, and first half of the third century CE ostrakon (Inv. 11723 = O.Trim. 1.51). Object densities were greater in the southern 4.2 x 2.0 m trench addendum and included a miniature lamp (Inv. 11917), a clay figurine head (Inv. 11920), loom weights, and several complete vessels.↩#p471
Remarks on the Phasing of Courtyard C2A
Phase 1: Construction and Original Use. Courtyard C2 was laid out at the same time as B2 and S1, as indicated by the wall bonding patterns in the area.↩#p472
Phase 2: Abandonment and Deflation. The abandonment of the courtyard is difficult to assess, since it may well have been in use beyond the occupation of B2. As with B2, the wind dramatically deflated the walls of the courtyard.↩#p476
Phase 3: Bedouin Use Present–Day The Bedouin do not seem to disturb this area of the site, as they herd their animals across an area just east of Courtyard C2.↩#p477
Remarks on the Phasing of B2 and its Surroundings
The results of the 2005–2007 excavations in Area 1 indicate that the street (S1) was laid out contemporaneously with the house (B2) and the exterior courtyard (C2). The foundation trench for the north wall of B2280 shows this simultaneous construction clearly. Likewise the lack of any clear evidence for activity in the excavated area prior to the building of B2 and Street S1 suggests a compressed phase of construction and occupation in this area of the site.↩#p478
The dating evidence points to the late third century CE as the primary phase of occupation, destruction, and abandonment in the early fourth century. The ostraka suggest dates between prior to 250 CE through the end of the third and perhaps beginning of the fourth century. A fragment of textile (Inv. 11527), of a type that was produced only in the third century, provides additional datable material. The ceramic assemblage suggests a third-century occupation, perhaps extending into the early fourth century. A more thorough examination of the chronology will be presented in chapter 8.↩#p479
Notes
1 This courtyard is currently undergoing complete excavation by the University of Reading (2012, 2013) and currently by the City University of New York, directed by Anna Lucille Boozer. For preliminary reports, see Boozer 2012b, Boozer 2013c.
117 Upper elevation 137.926; Lower elevation 137.446. Dimensions: x=26 cm, y=190 cm.
118 These vessels included a bowl with white slip on the rim and red dots (Inv. 10010), a complete bowl (Inv. 10392) a complete footed bowl (Inv. 10166) a complete section of a bowl (Inv. 10010), and a complete lid (Inv. 10393). Inv. 10392, Inv. 10010, and Inv.10166 were located in the upper levels of DSU 38, while Inv. 10010 and Inv. 10393 were embedded in the collapse.
172 Upper elevation 138.256; Lower elevation 138.016. Dimensions: x=185 cm, y=86 cm.
173 Upper elevation 138.061; Lower elevation 138.031. Dimensions: x=50 cm, y=72 cm.
174 Upper elevation 138.011; Lower elevation 137.686. Dimensions: x=150 cm, y=84 cm.
175 Upper elevation 137.781; Lower elevation 137.636. Dimensions: x=105 cm, y=82 cm.
176 Upper elevation 137.736; Lower elevation N/A. Dimensions: x=63 cm, y=93 cm.
177 Upper elevation 138.341; Lower elevation 137.736. Dimensions: x=120 cm, y=76 cm.
178 The overall dimensions are x=120 cm, y=75 cm, z=60 cm.
179 The dimensions of these three steps are: higher step: tread=26 cm, riser=10 cm; middle step: tread=25 cm, riser=10 cm; lower step: tread=25 cm, riser=14cm.
180 Upper elevation 138.336; Lower elevation 137.181. Dimensions: x=162 cm, y=62 cm.
181 Upper elevation 138.186; Lower elevation 137.461. Dimensions: x=317 cm, y=61 cm.
239 Upper elevation 138.066; Lower elevation 137.796. Dimensions: x=127 cm, y=200 cm.
240 Upper elevation 138.126; Lower elevation N/A. Dimensions: x=97 cm, y=200 cm.
241 Upper elevation 137.966; Lower elevation 137.681. Dimensions: x=240 cm, y=200 cm.
242 Upper elevation 138.430; Lower elevation 138.076. Dimensions: x=130 cm, y=200 cm.
243 Upper elevation 137.771; Lower elevation 137.526. Dimensions: x=145 cm, y=231 cm.
244 The DOP did not excavate a trench in this area (Colin Hope pers.comm.). It is also unlikely that Lynn Meskell’s test pits were located in this area (James Conlon pers.comm.). It may be that the SCA did some preliminary test pits in this area about which we have no information.
245 Upper elevation 137.716; Lower elevation 137.506. Dimensions: x=50 cm, y=220 cm.
246 Upper elevation 137.536; Lower elevation 136.926. Dimensions: x=50 cm, y=220 cm.
