This document is part of the online version of the book An Oasis City by Roger S. Bagnall, Nicola Aravecchia, Raffaella Cribiore, Paola Davoli, Olaf E. Kaper and Susanna McFadden, which is available at http://dlib.nyu.edu/awdl/isaw/oasis-city/. 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 specific content available online will be updated to include more links to supplementary material and related digital resources.
Figure 6. Satellite view of the Amheida and El-Qasr area (2000) (Base image from Ikonos)Figure 7. View of Amheida amid cultivation.
Lying in the western part of the Dakhla Oasis, in an oblong sandy area south of El-Qasr, Amheida is today completely surrounded by fields (Fig. 6). The oasis landscape is a patchwork formed by green land, laboriously reclaimed from the desert through irrigation, and sandy areas of various sizes, where water is not present and cultivation is thus impossible. Some of these barren areas are covered by sand dunes that encroach upon the green land, the roads, and the villages. The landscape is therefore variable according to the movements of the dunes and changes in the water supply. In this part of the Dakhla Oasis, the present pattern of the cultivated fields and the dunes clearly follows a north–south orientation, which has been produced by a stream of dunes that move down from the northern escarpment and flow toward the south at a rate of 7 m per year. So the cultivated fields are possible only between the dunes. The ancient landscape shared these characteristics, but its green areas were watered by different wells and thus located at least partly in different places (Fig. 7).↩#p21
The competing forces of the accumulation of sand and of erosion from winds and water due to climatic changes have also contributed to the formation and alteration of the oasis landscape. Human activities are responsible for substantial modifications as well, especially in the last thirty years or so, with the introduction of paved roads and of drilling machines that can reach deep aquifers. Intensive-agriculture policies, population growth, and changes in the way of life have led to a great consumption of water and to the consequent formation of wastewater lakes.↩#p22
The Dakhleh Oasis Project survey identified between 1977 and 1987 about five hundred sites of various historical periods and about four hundred of prehistoric eras.1 One of the major tasks of the DOP was to study the interaction between environmental changes and human activities in the region, which involved specialists in several fields collecting data and investigating both natural and cultural evidence and features. This complex regional and diachronic study, enriched by archaeological excavations led by different teams and by new surveys, is still under way.2↩#p23
Looking at the distribution of the features, settlements, and necropoleis of historical periods, we can gain some perspective on the changes in the distribution of the population, which almost certainly was determined by the availability of water. It seems evident that the extent of the ancient fields was different from that of the present green land, which is, taking into account the whole oasis, significantly greater.↩#p24
In this simple analysis, however, we should keep in mind that some ancient sites and features could have been invisible during the DOP survey because of the presence of the ever moving dunes or of cultivation. The regional survey should therefore, in principle, be repeated at intervals to discover new features or to confirm the collected data in areas with much wind-blown sand. The local landscape can change in just a few years, as is well demonstrated by recent villages and fields having been abandoned after the encroachment of the dunes, or by newly visible archaeological sites, as in the case of Area 11 at Amheida, which came to light only in 2013 after a dune had moved on southwards.↩#p25
Figure 8. Burials cut by a water channel.
The study of the Amheida area landscape started in 2006 with the documentation of the then present situation: the topographic survey of the visible buildings was accompanied by a survey of the main natural features, like the dunes, spring mounds, and yardangs, of other artificial features such as channels, and of areas covered by potsherds. The denuded geological strata together with the archaeological remains suggested deep changes over time in the ancient landscape, caused by deflation. In fact, some graves were found completely exposed and intersected by later channels running southeast of the pyramid among residual mounds and along the flat plane of the south cemetery. These graves are probably to be dated to the Old Kingdom, and the skeletons deposited at the bottom of the burials are now totally exposed and badly eroded (Fig. 8).↩#p26
Figure 9. Preserved barrel vaults.Figure 10. A sherd-covered site surface.
Harsh erosion is also evident on top of the temple hill, where it has been calculated that at least 1.5 m of stratigraphy have been lost. Erosion is also responsible for the disappearance of most of the upper parts of buildings: only a few are preserved above the present surface formed by wind-blown sand. The best-preserved buildings are those that were rapidly covered by sand; in contrast, those situated on top of natural elevations have been exposed for longer, so are eroded almost as far as floor level (like the church, B7, in Area 2.3 and the houses in Area 1). The buildings located in the lower areas were quickly invaded by clean sand up to the ceilings of the ground floors and eventually covered. This burial allowed for the preservation of ceilings, mostly barrel vaults, which are still in situ, as in the south quarters (Fig. 9). The natural erosion of whole buildings or of just their upper parts, human activity aiming to recover ancient building materials, and the accumulation of sand and dunes have produced a leveling of the city surface and a radically altered perception of its skyline and environment. The pottery sherds covering most of the site like a carpet (Fig. 10) are mostly derived from the erosion of the buildings, where sherds were employed as construction material in walls and ceilings and pots were stored on the roofs.↩#p27
Accumulation of sand and deflation are natural phenomena that occur continuously and are more severe in some periods, due to climatic changes.3 The geological and geomorphological study of the area carried out in 2010 by an international team is now focusing on residual yardangs so as to determine the major phases of erosion that occurred before the Roman period.4 In the present state of our knowledge, we can say that the settlement was built on an area periodically invaded by chains of dunes and exposed to erosion that was not uniform, as has happened in more recent times. To these phenomena we have to add dry phases in which the water table shrank, favoring rapid dune movement. The natural spring mounds needed to be reactivated several times by artificially deepening the wells, as is shown by the pottery found on and near those mounds. The west and north areas of Trimithis seem to have been sandier than the east, as is true today. Houses and living quarters moved according to the dunes, and thus the settlement was continuously having to deal with the local environment. The same would likely have happened to the farmland.↩#p28
It is the combination of these phenomena that caused the dis-appearance of some of the oldest settlements and features—of which only scattered potsherds remain—such as possible small Old Kingdom settlements and workshops west of Amheida. On the other hand, ruins like the Old Kingdom and Second Intermediate Period buildings on the central hill were not eroded, because they had already been buried below deep strata from human occupation.↩#p29
Figure 11. A Mut-formation mound of reddish shale.
The fossilized sand dunes on which part of Trimithis is built have been reached in several trenches in Area 2.1, and this level has been explored in depth in Street 2. At this point, the trench (3 m deep) reached a geological layer rich in organics, but could not reach the Mut formation (a reddish shale) readily visible on the surface in nearby places.5 Nonetheless, a small Mut formation mound is present south of Area 1 (Fig. 11), on one side of a round depression and near a series of kilns used for pottery production. It seems possible that the shale was artificially dug out from deeper layers and used for manufacturing pottery in the fourth century CE, as the materials of the local vessels suggest.↩#p30
Also, mounds of light-brown clay, employed for making mud bricks, were and still are present in the area. One such mound is preserved near the modern village of Bir Itnayn el-Arab and was used to make the mud bricks of the replica of the house of Serenos. Others can be seen in the fields between Amheida and El-Qasr. Considering the enormous quantity of mud bricks used in the city buildings and tombs, there must have been more such mounds around the settlement. Some, later covered by buildings, are in the area of the city and in the necropolis. Today the surface of those mounds is severely eroded, and very few features or buildings survive. The carpet of potsherds is the only evidence of the presence of ancient buildings there.↩#p31
Figure 12. A large spring mound south of Amheida.
The empty, barren area south of the city is characterized by mounds covering tombs and by spring mounds, one of which (Fig. 12) is of impressive dimensions. Potsherds included in the mud layers of the channel coming out of it suggest a long duration for the activity of the spring, possibly aided by some artificial reactivation. Given the deflation, it is extremely difficult to say when the low, flat areas were covered by soil and cultivation. However, a pigeon house was built in the fourth century, presumably near a farmhouse, at the north border of the cemetery. A topographical and pottery survey is planned for this area in the near future.↩#p32
The geological, geomorphological, and archaeological analysis of Amheida's present state of conservation has led to the following conclusions: Roman Trimithis was constructed on an uneven area characterized by fossilized dunes, soft dunes, mounds of clay and shale, and flat, lower lands, where most of the wells for the city water supply were probably situated.6 The area was already inhabited at least from the Old Kingdom onward (§2.1), but at present it is not possible to know whether it was only one settlement, how big it was, or whether it was continuously inhabited. Data collected by the Dakhleh Oasis Project survey and our observations at Amheida would suggest the presence of different, probably small, settlements and workshops located north and west of the Amheida area. The erosion and the accumulation of sand prevent a clear view of the older phases.↩#p33
The Roman-period settlement was built according to the gradients, favoring an orthogonal orientation of the streets and alleys, but also following the natural shape of the ground (see §3.1). It is not yet clear just how great was the extent of the Imperial-period city, but as far as we have been able to see in the excavated areas, some dwellings of the third century CE continued to be used in the fourth, while other buildings, like the public bath, were partly demolished and replaced. The stratigraphy found in all the excavated areas testifies to the absence of sand deposits between the different building periods; this probably means that the city was surrounded by fields—on the scale of today's cultivation or even greater.↩#p34
Figure 13. Standing walls on the slope of the temple hill.
Probably at the end of the fourth century or not long after, Trimithis was invaded by a field of sand dunes running from north to south.7 These moved further south in the following centuries and left behind a consistent deposit of sand that was trapped in the ancient buildings and streets. We know neither the shape nor the extent of the moving dunes, but their activity must have been instrumental in the variable preservation or erosion of the ancient buildings. Few walls are preserved for as much as a few meters above the present sand surface of the site, and it is hard to find an explanation without considering the possible changes in a landscape of moving dunes. Most of the buildings preserved up to the height of the roof vaults are covered by sand up to 138–9 m above sea level. And most of the features above that elevation, both natural and manmade, have suffered massive erosion. In some areas we find exceptions, as in the case of the still-standing walls (Fig. 13), but we can observe that, generally speaking, mud-brick buildings located above 139 m above sea level have been worn down to the foundation courses and, in some places—for instance, on the temple hill—even below that level. Only the very robust clay hills east and west of the site have partially survived to a considerable height. The accumulation of pottery on their surface is all that is left of the houses once standing on them. Most of what was lying below 138–9 m elevation has been preserved thanks to the massive accumulation of wind-blown sand that we still see.↩#p35
Sand dunes west and southwest of the site, all above 139 m elevation, still cover substantial portions of the ancient inhabited area. Taking them into account together with the hills covered with sherds and the areas in between, it is likely that the overall extent of the urban area of the Roman period was twice as large as what is visible on the surface today.↩#p36
The long life of the settlement, which was continuously rebuilt in the same place despite the constant changes in the landscape, is certainly due to its favorable position on the main road system but also to the abundance of wells and springs.↩#p37
1.2. The sacred landscape
Olaf E. Kaper
Cities normally depend on hinterlands. The city of Trimithis was no exception; it functioned within the agricultural region at the western end of the Dakhla Oasis, which was known as Sawahet, "The Back of the Oasis." This name is attested from about 1300 BCE onwards up to about 700 BCE.8 The prefix Sa-, meaning "back", is not known elsewhere in relation to an oasis, and Giddy9 warns against translating the term literally, because it may also be a feature of hieroglyphic group writing in the rendition of foreign place names. However, this instance shows no group writing, because it is followed by the regular spelling for the word "oasis", wahet.↩#p38
Figure 14. The Greater Dakhla Stela (Ashmolean Museum, Oxford; photo O. E. Kaper).
The earliest evidence for this toponym is from the New Kingdom, from the site of Amarna. A wine vessel from the palace of King Akhenaten was labeled as containing wine from "the vineyard of Sawahet."10 Apparently the city of Amheida formed part of a wine-growing region, and the name Sawahet referred to more than just the settlement. In the 20th Dynasty, the name is found again in Papyrus Turin 2074, verso, from Deir el-Medina. This papyrus contains a list of persons, four of whom are from Sawahet and nine from the town of Hibis.11 That the toponym could denote an entire region as well as a town is again confirmed by the text on the Greater Dakhla Stela (Fig. 14), which was found at Mut, in which specifically a "town of Sawahet" is mentioned.12 The region must have been sizable, because the stela mentions a part named "the west of Sawahet." It is also said to contain vineyards, which confirms the information already established for the time of Akhenaten.↩#p39
From the 26th Dynasty onwards, the toponym Sawahet occurs no more in the sources. Instead, the texts from at least the time of Amasis onwards use the toponym Setwah (written 4t-wAH or more rarely 4t-wHa), of which we may assume that the pronunciation was similar or even identical to that of the earlier written version of the name.13 What could be the reason to change the official name of an entire region? There are no parallels for this practice from other parts of Egypt. The fact that the pronunciation of the place name remained (more or less) the same may indicate that there had been some aversion against the earlier one, and perhaps "The Back of the Oasis" was considered pejorative.↩#p40
We may assume that the new name should be translated as "The Place of Endowment," and that it refers to a large endowment for the temple established in the time between the last occurrence of the earlier name Sawahet around 740 BCE and the first occurrence of the new name around 570 BCE. At present, there is no evidence for any official activity at the temple during the 25th Dynasty, and it is therefore possible that the building of a new and larger temple under Amasis coincided with the establishment of a large endowment during his reign.↩#p41
Figure 15. The stela of Seth from Amheida.
The Greater Dakhla Stela thus points at Amheida's close relations with the temple of Seth at Mut el-Kharab from an early date. Later in the Third Intermediate Period, the local governor during the reign of Takeloth III supplied the temple of Thoth with a small but regular endowment. In the reign of Piye, the same governor did the same for the temple of Seth. For both donations, the commemorative stelae erected have been preserved.14 The temple of Amasis was decorated with a large image of the god Seth of Mut, thereby demonstrating the importance of this god also for the town of Amheida. A large stela from the Ptolemaic or Roman period (Fig. 15) showing the same image was also found at the site. No references have been found to Hibis, or to other sites in the oases, but the amount of preserved text material is limited.↩#p42
For the Roman period, we know much more about the hinterland of the city of Trimithis. This is because of the number of monuments remaining from that time, such as the late-Roman fortress at the site of El-Qasr, the cemetery of El-Muzawwaqa, and the temple of Deir el-Hagar within the immediate catchment area of the city. The fortress, probably built in the late third century CE, was located three km north of Amheida,15 and there is no doubt that the city and its fortress were in close contact. The cemetery of El-Muzawwaqa is cut into a small rocky hill situated roughly four km northwest of the ancient city, and contains private burials in addition to burials of sacred animals of the local gods Amun-Re and Thoth. One of the priests of Thoth of the early Roman period, a man by the name of Petubastis, was buried in this hill in a chamber decorated with Egyptian scenes and hieroglyphic inscriptions, as well as a zodiac on its ceiling.↩#p43
Figure 16. Inside the temenos wall at Deir el-Hagar (photo O. E. Kaper).Figure 17. Visitors' graffiti at Deir el-Hagar (photo O. E. Kaper).Figure 18. Plan of the temple at Deir el-Hagar (James Knudstad/Dakhleh Oasis Project).Figure 19. A stibadium (banqueting bench) inside the temenos at Deir el-Hagar (photo O. E. Kaper).
The temple of Deir el-Hagar, about six km to the west of Amheida, is of great interest because there is no important settlement preserved around it that would justify the building's relatively large size.16 Only a small agricultural settlement lies south and east of the temple precinct. In addition, it has a temenos wall that has doorways at regular intervals (Fig. 16). No other temenos wall with a similar number of openings is known from Egyptian temples. Given these doorways, and given the large number of visitors' graffiti (Fig. 17) preserved on one of them, it seems that the building functioned specifically as a festival temple. The many doorways allowed a large crowd to gather in front of it and watch the spectacle of the god Amun-Re's procession (Fig. 18). One of the unique features of the courtyard is the several circular banqueting benches (stibadia) against the enclosure wall on either side of the processional route (Fig. 19), on which important guests could be waited upon while watching the proceedings.↩#p44
Figure 20. A relief at Deir el-Hagar depicting Thoth and Nehmet-Away (photo O. E. Kaper).
Among the graffiti on the Deir el-Hagar enclosure wall are a number of images of rams, a baboon, and an ibis. These refer to the sacred animals of the god of the temple, Amun-Re, and of Thoth of Amheida. In the relief decoration of the temple, Thoth likewise occupied a special place, even though the temple was dedicated to Amun and the other gods of Thebes, the religious center of Upper Egypt. In the first hall (called the pronaos) of the temple, the lower parts of the walls are usually reserved for its principal gods, and it is here that we expect to find Amun-Re and Mut of Deir el-Hagar, but Thoth and Nehmet-Away of Trimithis are also depicted (Fig. 20). Deir el-Hagar and Trimithis were closely associated not only because they were both located in the area named Setwah, but also because the population of Amheida came to attend the festival of Amun-Re at Deir el-Hagar. There they would find their town's god, Thoth, depicted in prominent positions on the walls, and some would leave votive offerings with his image, as happened during the reign of Vespasian.17↩#p45
Figure 21. A relief at Deir el-Hagar from the reign of Titus, portraying Amun-Re, Khonsu and Mut (photo O. E. Kaper).
The temple of Deir el-Hagar was probably first built as a small sanctuary in the reign of Nero, after which the decoration of the innermost room was finished in Vespasian's reign. Under Titus and Domitian, the temple was extended and decoration was added on a larger scale (Fig. 21). It is certainly significant that these works were executed at the same time that the larger temple of Thoth at Amheida was being built and decorated. At Trimithis, the builders made use of the stones from the temples of the Late Period, which were being demolished to make room for the new buildings. There is no evidence of a similar procedure having been followed at Deir el-Hagar. The two temples were built simultaneously and, arguably, decorated by the same artisans, as their workmanship is of similar style and quality.↩#p46
1.3. The economic landscape
Roger S. Bagnall
In the Introduction, we called attention to the distance of the Great Oasis from the Nile valley and to the apparently dramatic rise in the extent of settlement in Dakhla during the Roman period. The growth in accurate knowledge of the chronology of oasis pottery in the last two decades has led to the redating of some wares found in the DOP's survey of the oasis, which has led in turn to the identification of more Ptolemaic sites at places once thought to be only Roman.18 Consequently, the rise in habitation sites from Ptolemaic to Roman seems a bit less marked now than it did when the survey was carried out during 1977–1987, but it is still evident that growth in the Roman period was striking. The Egyptian oases are far from being the only parts of the Roman Near East to witness expansion to such unprecedented levels of occupation, often not matched even in modern times. Given the high cost of overland transport in antiquity—a theme familiar to anyone with even a passing acquaintance with ancient economic history—one might well wonder how such growth was possible. Why would anyone settle in so remote and seemingly forbidding a place as Dakhla? And how could people earn a living there, let alone become rich? Furthermore, why would rich people from outside the oasis invest there?↩#p47
The oases are constituted essentially of soil and water, plus some stone in outcroppings and the bordering scarp, mostly sandstone of moderate quality. Actually, that description sounds at first blush like much of the Nile valley, too. But in the valley it was relatively easy to supply anything not available locally, because transportation on the great river was cheap; for the same reason it was easy to export the agricultural produce of the valley's rich soil to places from which other goods were to be acquired. The oases lacked this cheap and easy transportation network. By ancient direct road, Dakhla is about 365 km from the valley at its closest point; by the less strenuous route allowing a stop in Kharga, it is still farther. What, then, was the economic landscape generated by the economy (described in §6.1) and embedded in the natural landscape described in §1.1?↩#p48
First, it required significant capital investment. Without investment in the infrastructure of wells and water-distribution systems, the oasis was just so much empty desert. Even if natural pressure was in most cases sufficient to bring the water to or near the surface, wells needed to be dug to the water-bearing strata. We do not know what the cost of digging a well and supplying it with control and distribution systems was, but one ostrakon from Trimithis (O.Trim. 1.19) suggests that the gross annual revenue from the use of a well's water was in the neighborhood of 50 gold solidi, or more than two-thirds of a pound of gold. The capital value may thus have been something like five to six pounds of gold, a very large sum and far beyond the reach of most Egyptians. As far as we can see, these wells were privately owned, and there is no evidence for the assertion sometimes encountered that the expansion of irrigation in the oasis was driven by central-government planning. Rather, the wells represent massive private investment. But because that investment was so massive, it must represent the activity of a relatively small number of rich individuals or families, not a broad-gauged movement. This fact will be important in thinking about the society of the oasis (§6.4).↩#p49
Second, the oasis landscape, as we noted in §1.1, is not one of continuous cultivation—a green expanse in the middle of the desert—but of plantings clustered around the wells. Sometimes these wells were, and still are, close to one another; but in other instances there are large stretches of desert between the islands of cultivation. This patchy landscape can still be seen today, and was much more prominent even a decade or so back, before the sprawl of deep wells dug in recent years. Wells and their settlements formed zones around the major population centers like Trimithis, Kellis, Mothis, and Mesobe. The economic landscape was thus one requiring plenty of local transportation to connect up the centers of production and dwelling. Not only were many long-distance camel caravans required to export the oasis products to the valley, but armies of donkeys, and their drivers, were needed to keep the extended territory of the oasis working as a unit. Indeed, the landscape was thick with this traffic of donkeys, which are ubiquitous in the papyri and ostraka.↩#p50
Around the wells was a complex array of cultivations. Because distance and transport costs made it uneconomical to provide the people of the oasis with staple crops from the valley, the oasis had to be self-supporting in basic foodstuffs like wheat, barley, legumes, and vegetables; to that list may be added ordinary wine. Although high-quality wine was often an item of interregional trade, everyday drink was too bulky and not valuable enough to make overland transport as far as Dakhla a reasonable proposition. The arable crops were grown in small basins watered via channels from the wells. We know that along with the more familiar food crops just mentioned the oasis fields grew millet and cotton in the summer, at a time when the Nile valley was under water. The landscape was thus seasonally variable, but in a different way from the valley. The arable plantings were probably inexpensive to maintain once the water-supply infrastructure was in place. Their value no doubt varied, cotton being the most valuable because exportable and, at least once spun into thread, high in value proportionate to bulk.↩#p51
Alongside these field crops, and perhaps in part interspersed with them, were large numbers of trees. It was here that the oases' distinct advantage really lay. In the valley, trees required elevated ground not reached by the Nile floodwaters; consequently, they demanded artificial irrigation by expensive water-lifting equipment. In the oases, by contrast, the water flowed 365 days a year and never flooded the trees' roots. Not only the vineyards but also the date palms and olive trees benefited from this geography, and these were the source of the high-value fruits exported for cash. They also, of course, are another form of investment, requiring some patience for eventual returns, as dates and olives, in particular, take years to produce their potential yields.↩#p52
Figure 22. A pigeon tower northwest of Amheida, exterior view.Figure 23. A pigeon tower east of Kellis, interior view.
One major drawback of not experiencing the Nile flood, however, was an acute need for fertilizer. In the valley, the arable lands received fresh silt from the flood each year, so that only the raised, artificially irrigated garden lands needed fertilization. In the oasis, all of the land did. For the arable areas, this must have come in large part from growing nitrogen-fixing leguminous crops destined for animal fodder—the "donkey gasoline" leguminous greenstuffs familiar to visitors to the oases today. For the vineyards and orchards, fertilizer meant dung: dung from the cattle tethered in the fields, human night-soil, and above all droppings from the millions of pigeons raised in the striking farm buildings that are a staple feature of ancient rural sites throughout the oases. These buildings (Fig. 22, Fig. 23) are largely of standard plan, with side-by-side vaulted storage rooms on the ground floor and pigeon quarters on the upper level. The pigeons both provided fertilizer and filled a need for dietary protein, just as they do today in rural Egypt. The pigeon houses constitute another element of necessary investment for the economic landscape of the oasis.↩#p53
Because the archaeology of these rural farmsteads is very inadequately explored so far, we know little about the facilities for processing produce that must have existed. The only such settlement to be even partly excavated so far is ʿAin el-Gedida (see §6.2). The ostraka reveal the presence of multiple wine-production facilities around the wells, the places designated in Greek as hydreuma and in Egyptian as pmoun ("the water of" followed by a name). The pressing floors and vats were located in these hydreumata, to minimize the distance that the bulky fresh grapes had to travel. None has yet been excavated in Dakhla, although such facilities have been excavated elsewhere.19 Similarly unknown to us are the olive-processing facilities; very few millstones from olive presses have been identified by survey in Dakhla so far—again, in contrast to other parts of the Roman world. But these too must have been situated around the hydreumata. Even date-pitting and -compression must have been part of the annual activity of the rural centers. And once again, wine-production installations and olive presses, which needed hydraulic (lime) plaster to protect the valuable liquids, were not cheap to create; they called for substantial volume to be economical. Large-scale pottery production was also required for the containers in which the high-value liquid products would be stored and shipped; not surprisingly, a pottery workshop was part of the production and storage facilities at ʿAin el-Gedida (§6.2).↩#p54
One other element of the production and distribution landscape of the oasis remains totally unknown so far: Where were the olives, olive oil, dates, cotton, and other export crops gathered and turned into caravan shipments? Were there marketplaces populated with brokers and consolidators? Were the transporters organized into companies or guilds, as seems to have been the case with the caravaneers in the Eastern Desert, based in family firms from the town of Coptos? So far, neither documents nor archaeology have given us a clue about this entire aspect of the economic landscape of the oasis. We are reminded once again of how much remains to be discovered.↩#p55
5 The trench was 1 × 1 m: the fossilized dune starts from an elevation of 135.97 m above sea level and stops at 133.5 m above sea level; the layer of sand and clay mixed with organics starts at 133.5 m. Small and badly eroded potsherds were present only on the surface of this ancient dune.↩#footnote-1.5
6 The head of one such well is visible at 137.5 m above sea level, southeast of Serenos' house.↩#footnote-1.6
7 These are still visible on satellite views.↩#footnote-1.7
