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.
The Western Desert comprises two-thirds of the land within the current boundaries of Egypt. The only refuges from the hostile environment are the five major oases—Siwa, Bahariya, Farafra, Dakhla, and Kharga. In Arabic, Dakhla means the Inner Oasis and Kharga means the Outer Oasis (that is with respect to the desert, rather than the Nile). In Roman antiquity, these two oases were grouped together as the Oasis Major (“the Great(er) Oasis”). Dakhla itself is approximately 800 km southwest of Cairo and about 300 km west of Luxor (Figures 1.1, 1.2).↩#p70
Geology and Geomorphology
The Dakhla Oasis is a depression that extends more than 80 km from east to west and approximately 30 km from north to south. It rests along the southern edge of a limestone and sandstone plateau and the steep escarpment of Paleocene limestone defines the northern boundary of Dakhla (Figure 1.2).1 This scarp is elevated 300–400 m above the oasis floor.2 The floor of the Dakhla basin is approximately 100 m above sea level, though it varies in elevation across the depression.↩#p71
This oasis occurred as a result of geological changes occurring since the Early Cretaceous.3 Wind was probably the primary agent for excavating the oasis depressions, although tectonic activity also may have contributed to their formation. All of the oasis depressions occur in areas where hard rocks overlay soft ones; once the hard overlying rock was eroded, the softer rock was hollowed out quickly.4↩#p72
Underground water passes along permeable sandstone beds, confined between impermeable clays above and below.5 These beds are folded and faulted in places, while others have been uplifted or removed to other areas.6 Natural fissures or artificial borings allow water to reach the surface through artesian pressure. A conveying stratum of Nubian sandstone covered with impermeable materials confines the flow of water into the oases.7 Nubian sandstone covers vast areas of the desert in Dakhla’s vicinity.8↩#p73
Hydrology
The presence of oases in the Western Desert makes sedentary life possible in this arid region. The water that forms the oases derives from the Artesian-water sandstone underlying the oasis and the entire Western Desert. The basal stratigraphic unit in Dakhla is the Taref Sandstone Member of the Nubian Formation, the same unit that forms the floor of the Western Desert to the south. This unit contains one of the largest ground water reservoirs of the world. The thickness of this water-bearing sandstone column seems to range from 200 m in the Kurayim area to 1030 m in the Dakhla Oasis.9↩#p74
Water has always been a key concern in the oasis landscape, as can be detected by the emphasis placed upon wells throughout history. Writing in the late Roman Period, Olympiodorus noted that the inhabitants of the oases were renowned well diggers.10 As a mark of their significance, wells have always been the chief measure of wealth and importance in the oasis.11 Even today, well ownership is a complex and important issue. It is possible to own collective shares and individual shares of wells. Equally possible are individuals who own water but no land or land but no water.12 There has been much debate in recent years about how long this water is expected to last and if it is renewable. There is general agreement that the Saharan aquifer rate of recharge is substantially lower than the rate of withdrawal at present.↩#p75
Even within the Great Oasis, there are significant differences between Dakhla and Kharga in their sources of water. Most notably, parts of Kharga’s ancient water supply came from qanats. Qanats are water management systems designed to bring water to human settlements and agricultural fields in hot, arid climates. Sloping tunnels bring water from water-bearing strata in the hills into plains. They are equipped with a series of vertical shafts, which provide access and ventilation. The advantage of qanats is that they enable water to be distributed in hot climates without losing too much water to evaporation and seepage.13 Extensive evidence of these systems has been found in the Kharga Oasis, but few traces have been found to date in the Dakhla Oasis.14 These differences signal a divergence in the geomorphology between these oases as well as distinct approaches to water management among inhabitants.↩#p76
Although water reservoirs continue to support a sizeable sedentary population in the Western Desert, uncontrolled usage and methods of irrigation, dating back to antiquity, that emphasize flooding small basins cause severe problems with salinization, a problem present already in ancient times. Salinization has been responsible for vast tracts of wasteland, once irrigated but now abandoned, as well as salt marshes created by unrestrained water spilling.15 Today, considerable salt deposits can be found in the vicinity of Mut.16 Salinization is not only an issue for cultivation but has come to be an issue for the preservation of antiquities in the region as well, particularly of sandstone temples, such as ‘Ain Birbiyeh.17 The salt deposits seem to be long-present inclusions within the depositions of the western desert. Water and evaporation create these salinization issues.↩#p77
Climate
Climatic conditions in the Western Desert are severe. The extreme temperatures make daily life and fieldwork difficult. Winter temperatures range from as low as 0º–2ºC just before sunrise to 20º–25ºC by midday, and in the summer maximum temperatures reach 40ºC for extended periods and for stretches exceed this level.18 Heat provides a significant hindrance to many activities and even the human consciousness.19 Given the palpable influence of heat on the body and mind, it should be understood that it would have been a considerable issue in antiquity and most certainly influenced architectural choices and the organization of daily activities.↩#p78
The sand-laden northern winds were an obstacle in antiquity as well as today. The wind shaves down crops, fills in houses, and deposits dunes over paved roads. Herodotus tells the story of Cambyses’ army, which supposedly was engulfed by these winds during a campaign across the desert to attack the Oracle at Siwa.20 This account suggests that the fear of these winds was widespread across the Mediterranean basin. In antiquity, it was common to build wind and refuse barriers outside of houses to protect them from windblown sand.21 Today, the winds also serve as an obstacle to fieldwork, which is only possible in the winter months (October-March), and even then formidable sandstorms are not uncommon. From an archaeological standpoint, the strong oasis winds can be seen in the deflation of mud brick structures, with various consequences for the structure of depositional units, as well as oasis site topography.↩#p79
Agricultural History
Dakhla contains a green area of c. 410 km², although there is not a continuum of fertile areas within this quantity. Cultivation areas occupy scattered spaces in the lower land surface, stretching from Teneida in the east approximately 100 km to the west.22 Structures occupy high ground that is not suitable for cultivation within this space.↩#p80
There is evidence of continuous human activity in Dakhla extending from the Lower Paleolithic (c. 400,000 BCE) to the present day.23 Despite the distance between Dakhla and the Nile Valley, Dakhla was known as a fertile land throughout Dynastic Egypt. In particular, Dakhla was known particularly for prestige agricultural goods, such as olives, dates, and wine, until the Roman Period.24↩#p81
Currently, the Kellis Agricultural Account Book (KAB), taken together with archaeobotanical finds at Kellis, is our best source for understanding the agriculture of Roman Dakhla.25 From this account we have a good sense of the field and fruit crops available in Dakhla during the Roman Period. The predominant wheat is triticum aestivum, a bread wheat common in the Roman Mediterranean and found in Egypt from Ptolemaic rule onward. Durum wheat (triticum durum) and emmer wheat (triticum dicoccum) are not mentioned frequently in documentary sources.26 Other common crops include hay (trifolium alexandrium), arakia, fenugreek (trigonella foenum-graecum), sesame (sesamum indicum), tiphagion (an unidentified product), turnips (brassica rapa ssp. rapa), vetch (vicia), dried figs (ficus carica), dates (phoenix dactylifera) in multiple forms, doum (hyphanae thebaica), and olives (olea europaea).27 Cotton (Gossypium sp.) was grown in Roman Dakhla as this region provided a particularly favorable locale for the crop. Cotton is a summer crop that requires irrigation and is not suitable for the inundation and basin cultivation found along the Nile Valley. The oases provided the possibility of year-round irrigation from wells, which was more suitable for cotton cultivation.28 Wine is attested from the oases in KAB as well.29 Crops that also appear in Roman Dakhla include barley (Hordeum vulgare) and safflower (Carthamus tinctorius).30↩#p82
Due to the cost of transporting goods from Dakhla to the Nile Valley, it is likely that prestige agricultural goods, such as olives, dates, wine, and cotton, were the primary goods used for trade with that region.31 Bagnall has argued that bulk agricultural goods may have been traded more locally between the Nile Valley and the Small Oasis, suggesting that a more localized oasis economy focused on agricultural products was also possible.32↩#p83
Access
In antiquity there were two general means of access between Dakhla and the Nile Valley. First, one could travel via the Kharga Oasis and the routes offered between ancient Hibis and the Nile Valley. Second, it was possible to avoid Kharga by moving directly across the Libyan Plateau (Figure 1.2).33↩#p84
Traveling via the Kharga Oasis to the east, individuals used two routes most frequently to access the Nile Valley from Dakhla in antiquity: the Darb el-Ghubari and the ‘Ayn Amur Road. Winlock made the trip from Kharga to Dakhla in 1908 by following the Darb el-Ghubari route on camel.34 This route, fairly easy and level but somewhat longer than the ‘Ayn Amur road, departs from Hibis and follows along the southern extremity of the east–west oriented ridge that runs between the oases, ending at Teneida in the Dakhla Oasis. The route is clearly designated by cairns and ceramic scatters as well as signs of human activity in the form of petroglyphs, graffiti, and lithics.↩#p85
On his return to Kharga, Winlock followed the ‘Ayn Amur Road, which is located north of the Darb el-Ghubari.35 It departs from Teneida, traverses ‘Ayn Amur, and then connects up with the Darb el-Ghubari before the latter reaches Hibis. The ‘Ayn Amur route is 10 km shorter than the Darb el-Ghubari and runs across the plateau of the ridge between the two oases; it is a much more difficult route to follow than the Darb el-Ghubari because it is necessary to ascend and descend the escarpment. In his travel documents, Harding-King describes the descent into Dakhla in some detail:↩#p86
The sand had drifted up against the cliff we had to climb down, and once on the bank of sand the camels, by walking diagonally down the slope, were able to reach the bottom without difficulty. But at the top of the sandbank, the rocks…overhung to form a sort of cornice, so narrow that the baggage had to be lifted temporarily up from the camels’ backs to enable them to pass through the passage.36
This description provides but one example of the extreme difficulties that one had to endure on the ‘Ayn Amur Road when traveling by traditional methods.↩#p87
Second, naturally formed passes in Dakhla’s escarpment enable through-routes directly from Dakhla to the Nile Valley. The northern cliff barrier to the oasis is broken into a number of promontories by clear indentations. The most important of these passes are located to the north and east of el-Qasr, to the northeast of Balat and to the east of Teneida. The beds, which create a steep escarpment elsewhere, structure a gradual slope up to the plateau at these indentations. This slope allows the only practicable passes from the oasis floor up to the limestone plateau.37↩#p88
The most significant pass, located some 15 km east of the town of al-Qasr, unites with the Darb el-Tawīl route that leads north out of the oasis further east. There are a number of alternate routes within the pass that merge together in the same general direction.38 North of the escarpment, the floor of the depression of Farafra descends almost imperceptibly for some 150 km until one can perceive the steep cliffs on the east, west, and north of the Farafra Oasis. The total length from the apex of the Farafra depression in the north to the escarpment of Dakhla to the south is about 200 km.39 The old caravan route between Farafra and Dakhla covered these 200 km. By camel, it would have taken up to four days to reach Qasr al-Farafra from the town of al-Qasr in Dakhla, which is located just a few kilometers from Amheida. The ancient caravan route passes by Bir Dikkar and continues at the edge of the sand dunes until it descends the escarpment of Dakhla above al-Qasr. This route is still in use today as a 300 km paved road. It starts from Mut, passes by al-Qasr and Abu Munqar, and then goes on to Farafra.40↩#p89
Cultural History of Roman Dakhla
The archaeological evidence for human activity in Roman Dakhla is abundant and indicates that the oasis is an ideal location for studying daily life in Roman Egypt. The oasis has been known to the western world since Sir Archibald Edmonstone visited in 1819, but Dakhla was not intensively explored until the late twentieth century.41 The DOP discovered over 200 Roman Period sites during its survey, carried out from 1977 to 1987, basing its interpretation upon ceramic evidence.42 This data allowed for preliminary dating and identifications of sites, eventually leading to excavation at several sites. I provide a brief overview of some of Dakhla’s excavated sites below to indicate the range in settlement, mortuary, and religious sites that can be found in this oasis.43↩#p90
The site of Kellis (Ismant el-Kharab) provides significant archaeological evidence of domestic life in Roman Dakhla.44 The DOP survey of 1981–82 found substantial Roman ruins on this site. Excavation at Kellis began in 1986. Although they have not yet been fully published, the Kellis houses provide the best local comparanda to Amheida House B2. The Main Temple, located in the western part of the site, is the earliest structure known at Kellis and may have been constructed in the first century CE. It was dedicated to the protective deity, Tutu (in Greek, Tithoes), a local god in Dakhla. In addition to the temple and houses, three fourth century churches were excavated, along with a necropolis and a bath house.45 These urban remains, in addition to substantial mortuary remains, provide us with the most complete and holistic exploration of a single Romano-Egyptian town in Dakhla to date. A more thorough discussion of the Kellis houses can be found in the previous chapter.↩#p91
Ain el-Gedida, which underwent excavation by the SCA and Nicola Aravecchia, offers a site with a dense concentration of structures as well as a church.46 This settlement is significantly smaller than Kellis and Amheida and offers a non-urban perspective on daily life in Roman Dakhla. Full publication of this site is expected in the near future.↩#p92
The site of Mut el-Kharab, the ancient and modern capital of the oasis, is in highly ruinous condition. It consists of isolated islands of preserved archaeology distributed within the sprawl of present-day Mut. Despite the challenges that this urban archaeology offers, the DOP has excavated a considerable portion of the temple at Mut. Temple relief blocks and ceramics reveal that occupation extended back to at least the Persian Period and also attest to the high levels of wealth and status in the oasis.47↩#p93
Deir el Haggar, Dakhla’s signature tourist-accessible monument from the Roman Period, appears to have been the most important temple in the western portion of Dakhla, beyond Amheida.48 The temple of ‘Ain Birbiyeh, to the east of Kellis, has also been excavated and is undergoing conservation due to the destructive effects of salinization to the sandstone.49 The active temple building during the Roman Period attests to the continued significance of local cults during this time.50↩#p94
Qârat El-Muzawwaqqa appears to have been the primary elite burial site in the vicinity of Amheida. Research here has yielded several important painted tombs, which attest to the complicated ethnic identity affiliations among oasites in the Roman Period. In particular, the tombs of Petosiris and Petubastis show a complicated intertwining of Egyptian, Greek, and Roman elements suggestive of a multicultural local environment.51↩#p95
Documentary evidence also sheds light on the communities who inhabited Dakhla during the Roman period. I have mentioned already KAB as an important source for the agriculture of Roman Dakhla. We are fortunate also to have the Notitia Dignitatum for this area of the Empire, which gives us some clues about the distribution of forces in the Dakhla Oasis.52 Moreover, Kellis has produced numerous documents in the forms of papyri, ostraka, and an early book.53 Despite recent advances in our understanding of Roman Dakhla, we still do not have a good sense of where the expanded population came from in the Roman Period or why they came. Immigration seems to be the most likely explanation of the population boom, although increased economic potential may have raised the health and longevity of individuals in the region. Although the origins of the Roman Period population are uncertain, it is clear from the visible increase of local wealth in the Roman Period that economic reasons were a primary cause of the population increase.54↩#p96
Recent research at Amheida brings additional, significant evidence of daily life in Roman Dakhla to our current suite of data on this important oasis.↩#p97
Amheida
Amheida is a substantial ancient settlement located in the northwest part of the Dakhla Oasis (Figure 2.1).55 This multi-phase site reached its greatest extent under Roman rule (1st C CE–4th C CE). Its earliest known name was Set-wah (“resting place”) and it was known as Trimithis during the Roman Period. Amheida is the largest surviving ancient settlement within the Dakhla.↩#p98
Amheida was one of the most important towns in the Dakhla Oasis during the Roman and Byzantine centuries. Documentary sources indicate that Trimithis gained the legal status of a city by the fourth century and was regarded on the same level as other significant cities in the Oasis Magna.56 The diversity and magnitude of archaeological remains at Amheida attest to this significance. The substantial above-ground remains and surface pottery scattered across the urban center and cemeteries extend at least 2.5 km north–south and 2 km east–west. These remains represent dates ranging from Pharaonic to Late Antique periods, and the surrounding landscape contains evidence of prehistoric lithic scatters, Old Kingdom ceramics, and several cemeteries. Despite a long occupational history, Roman and Late Antique ruins dominate the visible site surface today. Like the greater Dakhla Oasis, Amheida appears to have reached its apex during the Roman Period.↩#p99
Figure 2.1: Map of the Dakhla Oasis (B. Bazzani, based on ASTER GDEM data, a product of METI and NASA).
Location of Site
The area of Amheida is bordered on the north by the stark limestone escarpment that dominates the entire oasis (Figure 2.2). To the west is the necropolis of Qârat el-Muzawwaqqa and the first century Roman temple of Deir el-Haggar. These monuments are rimmed in by Gebel Edmonstone followed by a barren desert replete with massive dunes leading out to the Great Sand Sea. South of the town, the land drops gradually over a minor escarpment and there is a dense band of present-day agricultural villages. Cultivation surrounds Amheida today.↩#p100
A small Bedouin village lies north of the site and modern agricultural fields extend to the east.57 There are minor springs in these vicinities. Other landmarks in the vicinity of Amheida include el-Qasr, a medieval Islamic and Ottoman town. El-Qasr likely was occupied already during the last period of Amheida’s existence in the form of the Roman fort of the Ala I Quadorum, and it may be the location to which Amheida’s inhabitants moved after they abandoned the city.58 There are signs that structures at Amheida, particularly the temple, were dismantled in early modern times to supply stone for construction at el-Qasr.59↩#p101
Figure 2.2: View of Amheida (to north).
History of the Amheida Project
Outsiders have known about Amheida since European explorers first ventured to Egypt’s Western Desert in 1819.60 The critical watershed for archaeological exploration in Dakhla came with the work of Ahmed Fakhry from 1936 until his death in 1973. Fakhry was astonished to find so many antiquities in Dakhla since early explorers had found so little there, compared with Kharga.61 Fakhry carried out a number of important excavations in Dakhla between 1968 and 1973, including at Ain Aseel, Qila el-Dabba, Balat, Qârat el-Muzawwaqqa , and Deir el-Haggar.62 Descriptions of Fakhry’s research have been published both by Fakhry himself and others who summarized his work, following his untimely death in 1973.63↩#p102
Since its European discovery, both specialists and non-specialists visited Amheida. The site came into prominence in 1979 when the DOP discovered Classical wall paintings on the site during its survey of the entire Dakhla Oasis.64 Modern archaeological fieldwork began in 2000, directed by Roger Bagnall.65 It was first under sponsorship from Columbia University; primary sponsorship passed in 2008 to New York University, in continued partnership with Columbia University. In 2012, the University of Reading became a partner in the project; the City University of New York assumed this role when Boozer joined the faculty of that institution. The first systematic excavations, begun in 2004, focused primarily on the house containing wall paintings discovered in 1979 (B1). These excavations later expanded to include additional domestic structures (B2), the temple mound, a church (B7) and conservation of a pyramid and of a monumental mud brick tomb.↩#p103
Topography of Amheida
Survey continues to define the extent and form of Amheida’s urban fabric, although preliminary conjectures regarding the layout can be made at this time (Figure 1.3).66 Amheida contains a diverse range of structures, as might be expected for a major regional center. The houses seem to be single—or, in some cases, possibly two—story mud brick structures with mainly barrel-vaulted roofs and some palm reed and mud flat roofs. Industrial areas are distributed among the domestic structures, but primarily along what we currently understand to be the edges of the city. A temple hill on the west side, around which the Roman settlement wraps, and mortuary structures along the southern side are also clearly evident among the surface remains. Agricultural fields occupied the low-lying surrounding landscape. Governmental, administrative, or so-called “public buildings” have not been identified securely yet, apart from the church (B7). Even so, it seems that the area just east of the temple mound may have been a focal point of the site, as surface remains indicate a number of large, elaborate structures that may have served civic functions. Moreover, the houses excavated closest to this area appear to be of particularly high status.↩#p104
There are two major streets currently identifiable at Amheida. First, a broad east–west oriented road (S1) provided access into the city from the east. It leads from the industrial and domestic area on the northeastern edge of the site towards an area north of the temple mound, but it turns sharply off axis and terminates before reaching the temple mound. Its width is 6.82 m, which corresponds generally with evidence of other Eastern Mediterranean cities.67 It appears that the road and the structures along its eastern end may have been built during a single phase of construction, as argued later in this report.↩#p105
The second major street (S2) is a north–south oriented road that extends from the mortuary area in the southeastern portion of the city to the north. This road appears to have had structures built into it over time, which made it less effective as a major conduit across the city. This phenomenon of narrowing streets fits a general pattern in the evolution from Classical to Byzantine cities seen at other sites in the Eastern Mediterranean. Scholars commonly interpret this encroachment of private structures upon street networks as reflecting the gradual decline of the central municipal authority.68↩#p106
In addition to these major roads there are some clear streets and alleys. One runs roughly parallel to the major north–south road and is located between this road and the temple mound. Some additional east–west aligned streets also occur. The project has not yet identified any streets connecting the eastern settlement area with the temple mound, nor has any trace of the temple dromos been found.↩#p107
There are several areas that indicate localized planning. For example, the northeastern part of the site appears to have been laid out in a single phase, and the structures in this area share similar plans. An area located in what appears to be the center of the Late Antique site also contains houses that, from the surface, appear to share similar layouts. Likewise, in an area just south of the temple, a series of roads delineate the eastern boundary of a housing block where at least one significant, decorated house is identifiable on the basis of surface architecture.↩#p108
Domestic Structures
Domestic structures have been a key focus of this project since its inception, and this project will contribute substantially to our understanding of Romano-Egyptian domestic life.↩#p109
Figure 2.3: Plan of House B1.
A fourth century house (B1) provides us with a wealthy dwelling that contained Greek mythological wall paintings in a central, domed room and wall paper motif wall paintings in a number of side rooms (Figure 2.3). These figural scenes are unique within Roman Egypt at this time and also inform us that occupants of the oasis were more comfortable with Roman Mediterranean cultural norms than might be expected for a peripheral Egyptian region. Measuring 225 m², this house was largely square in plan-view and had a clustered plan of access between rooms. Most of the rooms were covered in barrel vaults, although one room contained a dome, potentially with an oculus. There were flat roofs constructed out of palm reeds and beams with a mud plastering over at least two rooms.69↩#p110
Minimal material culture was recovered from this structure. Damp conditions did not preserve soft organic materials at B1 as well as is found at other sites within the oasis, such as Kellis. The other materials that survived include a substantial corpus of ostraka, which provide information about the owners of this structure, dependency relationships within the region, and small-scale economic information. These ostraka also provided a glimpse into the political framework of Trimithis since they repeatedly mention a man named Serenos, who is presumed to have owned B1 at some stage of its occupational history, and who can be inferred to have been a city councilor.70 Ceramics are generally of high quality with respect to the local assemblage. Additional artifacts point towards higher status occupants. Preliminary analyses suggest that the occupants consumed foods common within the broader Roman Empire and sustained an existence looking more towards the Mediterranean than one might have expected, given the peripheral location of the site.71↩#p111
The vicinity of the house has also been explored. Children and youths appear to have been educated in three rooms along the north end of the house. Red dipinti on the walls indicate that these pupils were educated in Classical Greek traditions and particularly rhetorical verse composition.72 South of the house lies another, unexcavated house with an identical overall footprint and a similar layout within.73 Sondages beneath B1 indicate that it is located on top of parts of a Roman bathhouse that possibly dates to the third century CE and was constructed out of mud and baked brick. Only a fraction of this structure has been excavated thus far, due to the overlying structures. Excavations have recently begun on an additional structure (B6), located northwest of this house. This structure contains approximately ten rooms and a central, columned room. These preliminary results may suggest Roman Mediterranean influences on the architecture, but more excavation is required to justify any conjectures.↩#p112
Figure 2.4: Aerial photo of House B2, facing west.
House B2, the subject of this volume, was also excavated in its entirety (Figures 1.5, 2.4). This house is located in the northeastern part of the site in an industrial and domestic quarter (Area 1). The house measures 121 m². Like the larger house (B1), this structure is constructed out of mud brick, demonstrates a clustered plan of access over its square plan, and employed barrel-vaulted roofs. The vicinity around this house has been explored as well. We excavated a trench in an exterior courtyard (C2A) as well as a trench in the street (S1) adjacent to the house (B2). To these excavations we can add the preliminary survey by the DOP, during which they cleaned and mapped the surface of a structure across the street from B2, revealing a structure that was provided with kilns for small-scale ceramics production.74 More will be said about all of these components of B2 in the following chapters.↩#p113
Temple
Amheida’s temple was dedicated to the Egyptian god Thoth, god of writing and wisdom (Figure 2.5). The temple itself, although completely dismantled by stone-robbing in antiquity (and again in more recent times) when looters and sebbakhin (diggers of decomposed organic matter) also dug deep pits, has provided significant information on the occupational history and religious life within the Dakhla Oasis. Excavations of the numerous looter and sebbakh (decomposed organic matter used for fertilizer) pits dating to the Ottoman and later periods that dot the temple mound have indicated that the site was occupied since at least the Old Kingdom. The material recovered from these excavations suggests that there may have been a settlement located on what became the temple mound in later years. New Kingdom remains include an ostrakon with a school exercise.75↩#p114
Figure 2.5: Area 4, the temple mound.
The epigraphic material from the temple has provided a particularly important collection of cartouches with Pharaonic and imperial names. The recovered cartouches provide us with some of the construction history of the temple and associated chapels, although interpretations may change as they are based on a highly fragmentary data set. Minor building took place under the Theban 23rd Dynasty (King Pedubast circa 800 BCE), and the early 26th Dynasty (Nekau II (610–595 BCE) and Psamtek II (595–589 BCE)). Major construction seems to have started under Amasis/Ahmose II (569–526 BCE), and the Persian ruler Darius I (522–486 BCE), both of whom contributed a new chapel with vaulted ceiling. The Roman Emperor Titus (79–81 CE) built a new Roman Period temple on the site. Domitian (81–96 CE) both demolished older structures and contributed a new larger sanctuary, which was extended (or at least further decorated) under later emperors, as yet anonymous.76↩#p115
The epigraphic material also provides us with some intriguing clues about local oasite traditions. In the Third Intermediate Period, Dakhla was ruled by an Egyptianized Libyan tribe, called the Shamain.77 This epigraphic material suggests that these late dynasties incorporated the oasis regional structure much more thoroughly into their regional ruling than had been attested previously. Moreover, the Persian Period finds have provided some of the most substantial indications of a Persian presence in this distant oasis.↩#p116
The material culture recovered from the temple mound includes fragmentary glimpses into local cultic practices and the occupational history of Amheida. For example, ceramic coffins containing un-mummified birds were recovered from the temple mound. Such deposits were common from the Late Period through the Roman Period and served as votive offerings. The remains included a range of birds (raptors, ibis) as well as other animal fragments.78 The ibis was long associated with Thoth, the principal deity at Amheida, while the raptors should be associated with other gods venerated in the oasis such as Horus or Re-Horakhte. Offerings to Osiris were also found in the form of forty-five bronze statuettes, fragments, and pendants of the god. In addition to these finds, ceramics dating back to the Second Intermediate Period have been recovered, including large quantities of bread molds.↩#p117
Chronology of Amheida
The chronology of human activity at Amheida is poorly understood due to limited excavation thus far. The recovered temple reliefs and archaeological material from the temple area suggest that there was occupation extending back to at least the Old Kingdom in this area of the site. Predynastic lithic scatters in the vicinity of Amheida extend the chronology of anthropogenic activity back even further. At this time, surface remains suggest that the fourth century was the last major period of activity represented at Amheida. One must be wary working with surface remains, however, as they conceal a deeper history that has been exposed only in keyhole insights into deeper stratigraphy. It is entirely possible that Amheida was a large, wealthy settlement for a considerable period of time prior to the fourth century, as may be gleaned from the temple reliefs. The present work is concerned primarily with the mid-third to early fourth centuries CE. This compressed time span entails that the archaeological evidence itself varies little in terms of chronological markers for each stratum discussed—at least from what we currently understand for the material culture. Future studies may make a more refined material-based chronology possible, which would be most welcome.↩#p118
5 Ball 1927b:106. For more on the geology, see Murray 1952. On the Artesian-water sandstone, see Schild and Wendorf 1977. On the phreatic layers beneath the Western Desert, see Ball 1927, Ball 1927b.
14 On the qanats at Ayn Manawir in Kharga, see Grimal 1995:527-574, fig 11. Bagnall and Rathbone 2004 state that no qanats have been found in Dakhla. Recently, the SCA reported on evidence of possible qanats in Dakhla (Youssef 2012). This evidence is promising, but there are no publications on the Dakhla qanats to date. The examples found and reported on by the SCA are significantly shorter than those found in Khargeh and run in parallel to one another. These potential qanats require further study before more conclusions can be drawn.
26 On these wheat crops, see Thanheiser and Bagnall 1997:36.
27 On hay, see ibid.:38. On arakia, see ibid.:39. On fenugreek, see ibid.:39. On sesame, see ibid.:39. On tiphagion, see ibid.:40. On turnips, see ibid.:40. On vetch, see ibid.:41. On figs, see ibid.:41. On dates, see ibid.:42-43. On doum, see ibid.:43. On olives, see ibid.:43-44.
28 Ibid.:39-40. On cotton in the Small Oasis, see Bagnall 2008b. Cotton was grown in Roman Egypt by at least the second century CE (Wild 1997:8, Wild and Wild 2007, Wild, Wild and Clapham 2008, Thomas 2007:156 and note 57).
41 For a historiography of early archaeological research in Dakhla, see Boozer 2013a.
42 See papers in Churcher and Mills 1995. The dating of sites was based upon ceramic seriations that are now outdated. Now that the regional assemblage is better understood, more Ptolemaic material has been recognized among the initial survey assemblage than at the time of collection. Even so, the Roman Period is represented in greater numbers than any other until the present day.
43 The publications on Dakhla are too numerous to indicate here. Please consult the following bibliography for a comprehensive selection of sources on Dakhla http://www.amheida.org/inc/pdf/amheida_bibliography.pdf.
44 See above for coordinates and an introduction to the site.
45 Bowen 2000, Bowen 2002, Bowen 2003a, Bowen 2003b. On additional Christian remains in Dakhla, see Bowen 2008, Bowen 2009, Aravecchia forthcoming, Bayumi 1998.
46 For preliminary reports on these excavations, see http://www.amheida.org/index.php?content=reports.
52 For a comparison between the Notitia Dignitatum description of fortresses and the material correlates of these descriptions in the Oasis Magna, see Boozer 2013b. For the correlation between the Notitia Dignitatum and Egypt’s geography more broadly, see Worp 1994.
53 For example, see Worp and Hope 2006, Worp 1995, Worp and Rijksbaron 1987. On the Manichaean community at Kellis, see Gardner 1993, Gardner 1996, Gardner 1997a, Gardner 1997b.
54 On migration to and mobility in Dakhla and the Western Desert more generally, see Boozer 2011, Boozer 2015d.
55 The coordinates of Amheida are 25° 40’ 122’’ N, 28° 52’ 502’’ E.
57 The villagers still herd their cattle across the site today, and many of them work with the current excavation team.
58 On the discovery of this fort, see Jobbins 2006.
59 Many of the structures at Amheida show signs that the architectural features containing wood were removed at the time of abandonment (see Chapter 5). Olaf Kaper has identified matches between blocks found at el-Qasr and blocks recovered from the New York University excavations at Amheida (pers. com.). The Dakhleh Oasis Project conducts research at el-Qasr today.
60 On the history of archaeological research in Dakhla, see Boozer 2013a.
62 Ibid.:220-221. For a description of Fakhry’s findings, see Osing et al. 1982. See also Mills 1985.
63 Fakhry 1973:218-221, n.5. Dieter Arnold and Jürgen Osing visited Dakhla in March 1978 in order to publish a volume on Fakhry’s research results in Dakhla as well as continue studies that Fakhry considered to be important, see Osing et al. 1982. During this time, Osing also took notes at several temples in the Southern Oasis and published a series of articles on them, see Osing 1985, Osing 1978, Osing 1985a, Osing 1985b, Osing 1986a, Osing 1986b, Osing 1986c, Osing 1990.
65 See Chapter 3 for the Amheida Project staffing.
66 Congedo and De Santis 2006. See additional reports at www.Amheida.org. For conjectures on the late Roman urban development of Trimithis, see Boozer 2014.
