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 1. Egypt showing the location of the Dakhla Oasis.Figure 2. The Dakhla and Kharga Oases.
The subject of this book is an ancient city in a part of the Western Desert of Egypt that in those times was considered the western or "Inner" part of the Great Oasis and is today the Dakhla Oasis, administratively part of the New Valley Governorate (Fig. 1, Fig. 2). This city was called Trimithis in the Hellenistic and Roman periods; today the site is called Amheida. It is the subject of a field project that I (Roger Bagnall) have directed during the past fourteen years, with eleven years of excavation to date. We owe almost all of our knowledge of Trimithis to archaeology and to documents discovered through excavation, both at the site itself and from the work of the Dakhleh Oasis Project at Kellis, the archaeological site of Ismant el-Kharab, since 1986.1 As recently as 1987, the location of Trimithis was still debated, although the correct identification had been suggested already in the nineteenth century; it was known then only from a few papyri and one late antique bureaucratic compilation.2↩#p8
Figure 3. Left, the area of Amheida in 1967; right, the same area today.
There were multiple reasons for deciding to begin a field project at Amheida. Among them was a sense, shared with some other papyrologists—and eloquently articulated by Claudio Gallazzi in 1992 at the Copenhagen congress of papyrology—that our generation might be the last to enjoy significant opportunities to find ancient texts written in ink on papyrus, pottery, and wood in Egypt. Here, settlement expansion and the rising ground-water levels that are the result of the High Dam agricultural development and population growth more generally are rapidly destroying the dry conditions on the edges of the Nile valley and the Fayyum, as well as in the desert oases—conditions that preserved hundreds of thousands of papyri and ostraka until the twentieth century.3 Amheida itself, once largely surrounded by desert, is increasingly hemmed in by new cultivation (Fig. 3). Another motive was a more general assessment that there were few sustained high-quality projects in the archaeology of Graeco-Roman Egypt, especially linking texts to their archaeological contexts, and that more would be beneficial to the development of a more archaeologically grounded history of this society;4 a third motive was to furnish Columbia University, where I was then teaching, with a means for teaching archaeological fieldwork methods to its students.5↩#p9
Figure 4. The scarp bordering the oasis on the north.
The choice of site came more specifically from a combination of opportunity and my own research interests. Editing the wooden tablets of the Kellis Agricultural Account Book6 in the mid-1990s had brought me for the first time to the Dakhla Oasis. This region is located about 850 km by road from Cairo, and nearly 500 from the area of Luxor in the Nile valley; it is thus deep into the Western Desert. Like the other major oases of Egypt, Dakhla is a depression in the desert plateau, bounded by a high scarp on its north (Fig. 4). Amheida is in the northwest of this oasis and for most of antiquity was this area's dominant town, second overall only to the ancient and modern capital of the oasis—Mothis or Mut. The most important town of medieval, Ottoman, and modern northwestern Dakhla, the wonderful mud-brick settlement of El-Qasr, lies a few km away from Amheida and was, as we shall see, part of its territory in antiquity.↩#p10
We know quite a lot about this oasis by now, because it has been the object of regional survey for the last three and a half decades by the Dakhleh Oasis Project (DOP), directed by Anthony J. Mills, and because for more than twenty-five years an Australian team directed by Colin Hope has been excavating within the framework of the DOP at ancient Kellis.7 At Kellis have been found many papyri, ostraka, and wooden tablets, perhaps most famously its trove of Manichaean literary texts and family letters, but also a rich array of both documents and literary texts.8 It is from these documents that we get most of the information that we had about Trimithis before we began our excavations. For my own scholarly purposes, the fact that Kellis was abandoned around 400 CE, coupled with the fact that the focus of the excavations there was on areas belonging to the third- to fourth-century occupation of the village, means that its archaeological and documentary finds have been heavily concentrated in late antiquity, a period to which I was first drawn thirty-five years ago and continue to find of compelling interest.9 But it has something for almost every other taste and interest, too, as will become evident from the following chapters.↩#p11
The Great Oasis is a distinctive and remarkable part of Egypt. Because of its great distance from the Nile valley, it posed challenges for travel and transportation not otherwise present in a country bound together and granted cheap transport by a great river. In Chapter 1 we say more about the physical landscape and the geographical constraints imposed by location. Despite the difficulties of travel and transportation, the oasis was occupied already in prehistoric times when underground fossil water came to the surface in artesian springs—the mounds deposited by which are still visible today—even as the surroundings were turning from savanna into desert in the fourth millennium BCE. The oases have been thought to have led Egypt into the Neolithic revolution.10 Dakhla was explored and then occupied by the Egyptians of the valley under the Old Kingdom pharaohs, certainly already to some degree in the 4th Dynasty under Khufu, the builder of the Great Pyramid, then most extensively under the 6th Dynasty. There are even traces of the Early Dynastic Period to be found at Amheida, suggesting that there is much we do not yet know. The French Institute in Cairo began the exploration of the most important Old Kingdom site, ʿAin Asil not far from the village of Balat—where the governors had their seat and their tombs—near the other end of the oasis from Amheida, just before the start of the Dakhleh Oasis Project. Their work has produced important results: a palace, tombs, many striking objects, and written materials—Egyptian texts written, quite exceptionally, on clay tablets.11 Amheida and Mut also have Old Kingdom remains, and at ʿAin el-Gazzareen, a few km south of Amheida, a large Old Kingdom site has been partially excavated.12↩#p12
We know much less about the oasis, and even less about the other part of the Great Oasis—what is today the Kharga Oasis—for the next millennium-and-a-half, although there was certainly a New Kingdom temple at Mut, where the sequence is fairly complete, and (we now know) one at Amheida, to judge from surface finds of pottery and the discovery in 2014 of a stela of Seti II. Identifiable habitation sites, however, are not very numerous until they increase exponentially under the Romans. In late antiquity there was a moderate decline in numbers of sites, followed by a dramatic falling-off of the extent of habitation for quite a few centuries—although never any actual abandonment of the oasis, it must be added. But it is fair to say that at many sites we tend to find the last documents, or the last major horizon of documentation, to be the 360s, the period of the Kellis Agricultural Account Book, and Kellis itself seems to have been abandoned a few decades later. Other sites undoubtedly continue later, but little has been excavated so far from the last 250 years of Roman rule.13↩#p13
Figure 5. Plan of Amheida.
Amheida, with pottery of all periods from Neolithic to late Roman visible in surface survey, thus appeared potentially a diachronic key to the history of the oasis. Mut also presents a long time range, even longer than Amheida's, from early pharaonic into early Islamic, but only a tiny part of that site has survived the building of the modern city over it and the concomitant destruction of what little escaped that fate; the site was largely ruined even before the modern expansion of the city.14 Trimithis, in contrast, has never been reoccupied. In several areas—one of them very significant, as we shall see—groups in search of soil, baked bricks, and stones have dug pits of different sizes, but the site as a whole was not severely damaged by this activity. Natural causes have done far more harm, in fact, as is detailed in Chapter 1. Even so, the city mostly remains to be explored. We therefore have the possibility of studying a Graeco-Roman Egyptian town site at its full extent, something only rarely possible. That extent is very considerable, the Roman city stretching out on an uneven ground with the hill on which stood the temple of Thoth as its center (Fig. 5). Because some areas are still covered by sand dunes, we may not yet have a comprehensive picture of the full expanse of the ancient site. Including cemetery areas, it extends over an area of at least 2.5 × 1.5 km.↩#p14
We have a limited amount of information about Amheida's state in the nineteenth and twentieth centuries before the modern exploration of the oasis began with the Egyptian archaeologist Ahmed Fakhry and after him the Dakhleh Oasis Project.15 The travellers Archibald Edmonstone,16 Bernardino Drovetti,17 John Gardner Wilkinson,18 and Gerhard Rohlfs all visited the site in the nineteenth century. Wilkinson travelled to Dakhla in 1825 and provided the most interesting description of Amheida: "the extensive mounds of an ancient town with a sandstone gateway. The fragments of stone which lie scattered about appear to indicate the site of a temple, now destroyed. These mounds are about half a mile square, and below them to the E. is a spring called ʿAin el Keeád. They are also known as Lémhada. The only ruins now remaining are of crude brick; and from the state of their vaulted rooms, they appear to have been of Roman time." This description is remarkably accurate, containing the earliest mention of the modern name Amheida for the site. No other record exists of the sandstone gateway that Wilkinson saw, and it must have disappeared before Herbert Winlock visited Amheida in 1908.↩#p15
Drovetti was the first to remark on the brick pyramid, and Edmonstone noticed "a small remnant of a temple, and the fragment of a white marble statue. This last was apparently of Greek workmanship, and not without elegance, although so imperfect." Gerhard Rohlfs with two members of his team of German scientists visited the site in 1874. He identified the stone ruins as a fortress, not a temple. They also described "mountains of potsherds", fragments of stone vessels, small bronze objects and coins, all of which are still to be found on the surface of the site.19↩#p16
The Greek version of the name, Trimithis, has an Egyptian etymology, as is apparent from the Coptic equivalent Trimhite, "The Northern Storehouse."20 The reason why the town received this designation is unknown. Either the Greek or the Coptic name eventually gave rise to the Arabic version Lémhada, as recorded by Wilkinson in 1828, and the modern spellings al Amhâdeh (Winlock)21 and Amheida/Amhida (Dakhleh Oasis Project).22↩#p17
Ahmed Fakhry visited the site in 1963 and discovered a large funerary stela dating from the First Intermediate Period, which he took to Kharga.23 The systematic archaeological investigation of the site started in 1979, when the Dakhleh Oasis Project conducted its survey of that part of the oasis, with some limited test excavations.24 A workshop with pottery kilns was excavated,25 a corner of the central room in a painted villa (see §4.3 and 7.2), and several tombs in the cemetery, some of which had painted decoration (§4.2). In launching the excavation, we saw Amheida not only as the set of opportunities mentioned earlier but also as a chance to look at a number of historical issues of longstanding interest. None of these is limited to late antiquity, but all—at least in many views of Egypt in late antiquity—reach a kind of fulfillment in this period and are thus well approached through Amheida and other sites of its chronological horizon.↩#p18
One of these issues is economic growth in antiquity, for which the rapid development of the Dakhla Oasis in the early Roman period and its later virtual collapse raise all sorts of interesting questions relevant to contemporary debates about whether Roman economic growth was only extensive or intensive too—that is, whether it represented growth in average income per capita. The degree to which the development of remote areas was constrained by the cost of land transportation is an obvious factor here, but many other issues, like demand for products, technological change, and legal systems, come into play as well. Another area of interest is urbanism in Hellenistic and Roman Egypt, particularly the nature of diachronic change. Here in Dakhla, if anywhere, we should be able to see the Roman impact, the changes from classical to late antique, and a genuine decline, probably based on locally contingent factors. Because Trimithis at some point changed its status from a village to a city, it should be particularly interesting in this respect. Finally, the cultural mix of the oasis and its change over time are open to our gaze. An Egyptian village and religious center over the course of many pre-Roman centuries becomes a Roman city (polis). Kellis, which was evidently a new foundation in the Graeco-Roman period and never a city, offers us a handy basis for comparison and contrast close at home. How similar is Trimithis to Kellis? What distinctions between village and city and within the region would we find?↩#p19
For some of these questions we have found considerable evidence; for others, we anticipate finding it in future seasons. But the outcome of excavation is even less readily predictable than the weather, and there have been many surprises. In the following chapters, we have tried to describe what we have found so far. Like any dig, Amheida is a team effort, and this book is equally a product of a core team of authors.26↩#p20
2 Wagner 1987: 190–192, wavering between Ismant el-Kharab and Amheida. Wagner's book went to press before the Australian excavations at Ismant el-Kharab settled decisively that site's identity as Kellis. Lepsius 1874 is the first to associate Trimithis with the temple of Deir el-Hagar, even though this link could not be proven at the time.↩#footnote-0.2
4 For surveys of the archaeology of Hellenistic and Roman Egypt, see Bagnall 2001 and Bagnall and Davoli 2011.↩#footnote-0.4
5 From 2004 to 2012, the project included a field school for undergraduate students. It was terminated by order of the Supreme Council of Antiquities.↩#footnote-0.5
6 Bagnall 1997; referred to in this book as KAB.↩#footnote-0.6
8 A book edited by Professor Hope synthesizing the discoveries at Kellis is in preparation, to be published by Cambridge University Press. For this reason we have in the present volume avoided detailed discussion of these finds, but they have had a profound impact on our interpretation of our own finds at every stage. For these papyri, see the seven volumes of P.Kell.↩#footnote-0.8
9 See Bagnall 1993 for the broadest exposition of this interest.↩#footnote-0.9
11 The results of these excavations have been published in eleven volumes (to 2013) of Balat, in the series "Fouilles de l'Institut français d'archéologie orientale."↩#footnote-0.11
12 See Mills 2012 and Pettman 2012, with bibliography.↩#footnote-0.12
13 See Bowen 2012 for the church at Deir Abu Matta, where the pottery extends into the sixth century.↩#footnote-0.13
14 Excavations at Mut el-Kharab under the direction of Colin A. Hope have, however, been remarkably productive given the condition of the site. See, e.g., Hope and Pettman 2012 on the earliest chronological horizon there.↩#footnote-0.14
15 The following four paragraphs are the contribution of Olaf Kaper.↩#footnote-0.15
19 Rohlfs 1875: 129; Rohlfs actually identified the stone ruins as a fortress, not a temple. Cf. Kaper 2001: 240: "Rohlfs, Zittel und Ascherson machten auch einen Besuch zu steinernen Mauerresten in der Gegend zwischen El Qasr und Muschiya, die Rohlfs als "Rest einer ehemaligen, vielleicht römischen Befestigung" interpretierte. An der gleichen Stelle hat man "Berge von Topfscherben", Bruchstücke von Steingefäßen, kleinere Bronze-Gegenstände und Münzen gefunden. Es besteht kein Zweifel, daß dieser Ort die Stadtruine Amheida sei, der jetzt mit Sicherheit als die antike Stadt Trimithis identifiziert worden ist, wie schon von Lepsius vermutet worden war." Cf. Lepsius 1874.↩#footnote-0.19
20 Gardner, Alcock and Funk 1999: 276 [19].↩#footnote-0.20
24 Amheida was designated with a map reference as site no. 33/390-L9-1, and its cemetery to the south as site no. 33/390-K9-4; Mills 1980: 269-72.↩#footnote-0.24
25 Hope 1980: 307-11; Hope 1993: 123-27. See fig. 59 for some of these kilns.↩#footnote-0.25
26 Many more individuals have contributed to it through their work than only the authors of what follows. For a project directory and complete list of participants in the excavation from 2004 to 2015, see http://amheida.org/index.php?content=directory.↩#footnote-0.26
Introduction Figures
Figure 1. Egypt showing the location of the Dakhla Oasis.Figure 2. The Dakhla and Kharga Oases.Figure 3. Left, the area of Amheida in 1967; right, the same area today.Figure 4. The scarp bordering the oasis on the north.Figure 5. Plan of Amheida.
