The Excavations at Ismant al-Kharab: Volume II - The Christian Monuments of Kellis: The Churches and Cemeteries

The adoption of Christianity by the Egyptian populace was well underway by the late third century, but evidence for its presence in the archaeological record from the Nile valley is sparse. This is due, in part, to the loss of ancient settlement sites beneath modern cultivation. By comparison, Ismant al-Kharab, ancient Kellis, in Dakhleh Oasis, was abandoned at the end of the fourth century and many of its structures survive intact. The villagers, moreover, left behind a wealth of artefacts and documentation. By the late third century some had converted to Christianity and by the early fourth century three churches were built to accommodate their growing numbers. The churches afford an unparalleled window into three ecclesiastical complexes that served a single village. The Large East Church, moreover, is the earliest surviving example of a purpose-built basilica in Egypt known thus far. It provides a better understanding of the development of Egyptian church architecture and has forced a reappraisal of the dates of certain features that were previously attributed to the fifth century.

The community established three burial grounds: Kellis 2, with an estimated 3,500–4,000 graves, a funerary church and associated graveyard, and in a reused monumental mausoleum. Christian cemeteries are known throughout North Africa, Europe, and Britain, but in Egypt few are published in anything but a cursory manner. At Kellis, 800+ graves have been excavated; the earliest burials date to the late third century confirming the evidence of an early conversion by some villagers and its rapid expansion thereafter.

This volume provides the first detailed publication of the churches and Christian burial grounds. It incorporates a discussion of the spread of Christianity in Egypt’s Southern Oasis, drawing upon data from the rich textual documentation from the site. The material culture is presented in detail, especially the extensive collection of ceramics, glass, and coins.

• Language: English
• Publisher: Oxbow Books
• Release date: Feb 29, 2024
• ISBN: 9781789259643

Book preview

Introduction

Throughout the course of the survey of Dakhleh Oasis a variety of sites of different types was ascribed to the Late Roman period and identified as having been occupied or built by Christians. The date and religious affiliation were assigned largely on the data derived from the surface of the sites. Given the paucity of information on the early stages of Christianity in Egypt, and especially the chora, it was anticipated that Dakhleh would provide valuable new information. To this end, the site of Ismant al-Kharab, ancient Kellis, which preserved the remains of two temples and three churches, was selected. The site clearly spanned the transitional period when the villagers converted from the traditional religion to Christianity, and consequently, it had the potential to contribute significantly to our understanding of early Christianity and many other aspects connected with Roman-period Dakhleh. The survey and subsequent excavation soon revealed that occupation was confined to the first to fourth centuries of the common era, and possibly the very end of the Ptolemaic period. With the discovery of a wealth of documents relating to the activities of a thriving Christian community, it was imperative to explore the churches to deepen our understanding of that community. The exploration of Kellis 2 cemetery with its recognised Christian pattern of burials added another dimension. As it is excavated by the team of bioarchaeologists for the purpose of studying the human remains, the incorporation of its archaeology was crucial and has enabled a holistic approach to be taken when studying this early Christian community, which is unparalleled for fourth-century Egypt.

This volume is a comprehensive report of the results of excavations undertaken in three churches and three Christian burial grounds. In this Introduction, the Christian monuments are placed within the broader topographic context of the site; the excavation and recording strategy and the approach to presenting the ceramics are outlined.

The site was part of the concession to survey Dakhleh Oasis granted by the (then) Egyptian Antiquities Organization to the Dakhleh Oasis Project (DOP) in 1978 and was assigned the number 31/420-D6-1 in the DOP Archaeological Site Index (Mills 1982, 99–100). The site was surveyed between 1981 and 1983 during which time architect James Knudstad, assisted by Rosa Frey, determined its extent together with its general surface features which they began to map (Knudstad and Frey 1999, 189). Many buildings are visible with some walls standing up to 8 m in height. Six tests were undertaken, two of which were in the East Churches (Knudstad and Frey 1999, 189). In 1986, excavation began under the direction of Colin A. Hope (Hope 1988, 160). Hope divided the site into four areas for excavation purposes: Area A in the south-east is a fourth-century residential sector with a bathhouse and the East Churches; Area B, north of Area A, has large complexes, some multi-storey, rich with wall paintings all of which pre-date the fourth century. Area C in the north-east is contiguous with Area B; it includes domestic structures and workshops which span the first to third centuries. Area D, on the south-west, houses the temples, buildings of a formal nature (Hope 1988, 161–3) and the West Church Complex (Hope 2022, 9–14) (Figure 1).

In 2007, The Supreme Council of Antiquities deemed that directors of the sites within Dakhleh Oasis that fell under the aegis of the DOP apply independently for a concession to excavate. From that time onwards, the concession has been granted to Colin Hope, under the banner of Monash University. The concession to excavate Kellis 2 cemetery was awarded jointly to Tosha Dupras and Peter Sheldrick.

Area A Enclosure and the East Churches

The churches are all located on the periphery of the village (Figure 1). The main ecclesiastical area which houses the Small and Large East Churches is in the southernmost part of Area A (Figures 1, 2). The churches are situated at the easternmost end of a terrace that slopes down towards the south-east wadi (Knudstad and Frey 1999, 205; Bowen 2002a) (Figure 1). A major east–west street, some 8 m wide, separates this section of the village from the residences in the north. A bathhouse is located on the south-west of the street where it intersects with a north–south street that delineates Areas A and D (Figures 1, 2). To the east of the bathhouse is an extensive complex of buildings set within an enclosure wall, referred to here as the Area A Enclosure (Figures 1–3). The East Churches are within and adjacent to this enclosure, which is a pre-fourth-century development. There are no visible structures to the east of the churches or south of the enclosure. A well, which presumably served this part of the village, is located about 40 m south-east of the enclosure and beyond this is a row of collapsed tombs: the South Tomb Group (Figure 1). The dug canal is modern.

Figure 1 Plan of Kellis (drawing by J. Knudstad, J Dobrowolski and B. Rowney, complied by B. Parr).

Figure 2 Plan of Area A, showing Houses 1–5, the bathhouse, the Area A Enclosure and the East Churches (drawing by J. Knudstad, supplemented by B. Rowney and B. Parr).

The enclosure (Figure 3) is approximately 100 m east of the bathhouse; it comprises numerous rooms, courts of varying sizes, corridors, and at least one stairwell (Knudstad and Frey 1999, 205). The north and south walls are clearly defined and enclose a space of 47 by 60+ m; the east–west dimension has not been determined in its entirety, but Knudstad was able to trace the walls for about 60 m from its eastern extremity (Knudstad and Frey 1999, 205). The structures within the enclosure were altered considerably and could not be planned precisely because of the degree of preservation of the walls, which have survived well over 6 m in places, making it impossible to locate doorways below the windblown sand that filled the entire enclosure. Knudstad identified a major gateway in the east side of a bay in the south wall (Knudstad and Frey 1999, 205), but he could not find any other means of access.

Figure 3 Plan of Area A Enclosure, as determined by J. Knudstad in 1982.

Figure 4 Plan of Enclosures 1–4 showing the location of the West Church (drawing by J. Knudstad).

The south-eastern section of the enclosure was modified and extended for ecclesiastical use in the early fourth century. The Small East Church, as it has been designated, together with its associated rooms are located 10.5 m north of the south-east corner of the enclosure (Figure 3). The complex bears the location/area designation 31/420-D6-1/ A/8. Although excavation has only been carried out within the church itself, surface remains of the surrounding walls to the south and west indicate several building phases within this area. A series of walls >90 cm wide exposed by Knudstad are a third wider than the standard walls at the site which vary between 57–60 cm; the wider walls are set on a slightly oblique angle to the enclosure walls (Figure 3). They include a north-south wall 11.8 m west of the east enclosure wall, which forms the west wall of the church complex, and an east-west wall 4.5 m south of the church, which turns 90 degrees south 6.25 m from the east enclosure wall. A buttress bonded to the east enclosure wall on Knudstad’s plan appears to overlie the north section of the wider wall. As the area has not been excavated, it is not known whether the oblique north wall was reduced, and the enclosure wall built over it.

An east-west barrel-vaulted corridor extending from the centre of the east enclosure wall can be traced west for 12.5 m; it is 1.4 m wide at the west narrowing to 1.3 m in the east (Figure 3). The top of a doorway, 90 cm wide, that served as another point of access to/from the enclosure, was exposed at the east end of the corridor (Bowen 2003a, 153). A discrete block of rooms within the enclosure to the north of the corridor was modified to serve as the outer rooms of the Large East Church, which was built against the outer wall of Area A Enclosure to the north-east of the Small East Church. The Large East Church bears the location/area designation 31/420-D6-1/A/7.

The West Church Complex and its Surrounds

The third church, the West Church Complex, designated 31/420-D6-1/D/6, occupies the north-west corner of Enclosure 4 in Area D. The church is 42 m north of Enclosure 1 (Figures 1 and 4). It is one of four such enclosures in the western part of the site. Enclosure 1, the most southerly, houses the Main Temple, its dromos, and associated shrines and buildings; it was the earliest of the enclosures to be built (Knudstad and Frey 1999, 201; Hope 2022, 12). It occupies more than half of the area with the enclosures. Enclosure 2, north-east of Enclosure 1, has buildings of a formal nature which have not been explored. Enclosure 3 is an irregularly shaped area, built against the western section of the north wall of Enclosure 1 and the southern section of the west wall of Enclosure 2 (Knudstad and Frey 1999, 201; Hope 2022, 12); its north wall is buttressed and built on an oblique angle. No excavation has been undertaken in this area. Enclosure 4 is an expansion of the formal village enclosures (Knudstad and Frey 1999, 193; Hope 2022, 13) and was the last to be built; it shares common walls with Enclosure 3 and the northernmost part of the west wall of Enclosure 2. Two classical-style tombs, West Tombs 1 and 2, which pre-date the fourth century (Hope and McKenzie 1999, 61), stood on the open ground north of Enclosure 3 prior to the development of Enclosure 4, and possibly before Enclosure 3 itself was established. The tombs were brought within the confines of the village when the walls of Enclosure 4 were erected.

No buildings have been identified to the west of the church, and the only structure on the north is a stone-lined subterranean tomb chamber, West Tomb 3 (Hope 2004, 26), dug into the alluvial terrace about 25 m north of the enclosure. There is an earth dike running east-west in a wadi 20 m north-north-west of the church, and a well that Knudstad and Frey (1999, 201) state was dug in recent times, is in the south-west corner of Enclosure 3. A channel from the well passes through the west wall of the enclosure and continues into what Knudstad and Frey (1999, 201) describe as now derelict fields in the wadi beyond. This canal can be traced to within 40 m of the north-west corner of the church. The date that these water courses were cut is unknown. The other structures within Enclosure 4 are discussed in Section IVA.

The Cemeteries

A substantial cemetery comprising east-west oriented pit graves located to the north of the settlement was identified by Knudstad and tested by Edwin Brock and Peter Sheldrick in 1991 (Molto 2001, 83). It is designated 31/420-C5-2 (Mills 1999, 263). The cemetery is north-east of the North Tomb Group. A second cemetery of east-west pit graves was discovered within Enclosure 4 in 1999, it bears the designation 31/420-D6-1/D/7; two such graves were also found in the West Church (Hope 2003, 244; Bowen 2022b, 358–61). Excavation in North Tomb 1 during the 2000 and 2001 field seasons, revealed a series of intrusive east-west oriented pit graves that reflected Christian burial practices (Hope 2003, 264; Bowen 2022b, 362–3). North Tomb 1, designated 31/420-D6-1/NT/1, is the largest of the North Tomb Group which is situated on an open terrace northwest of the village (Knudstad and Frey 1999, 208).

The State of Preservation

The settlement was abandoned around the end of the fourth century and no large-scale occupation is attested after that time. During, or following its abandonment, the wooden fittings were removed. This included doors, their frames and lintels, screen walls, beams, and all portable timber items. The removal of wooden beams and lintels weakened the walls causing some collapse or large-scale cracks. Furthermore, the site is exposed to the prevailing winds that sweep down from the escarpment to the north and this has eroded the mud-brick architecture. The village was engulfed in windblown sand, which today is covered by a dense scatter of potsherds. Knudstad and Frey (1999, 192) explained this phenomenon, which they noted is prevalent on Roman-period sites, as the result of erosion of vaults and domes that are reinforced with chinking sherds. Kellis 2 cemetery and the north rooms of the West Church are the worst affected by erosion. This relates directly to their location as both areas are exposed to the winds. The cutting of a well to the south of the church has resulted in the formation of salt crystals on some of the floors in this structure.

The East Churches were less affected by erosion, being on the south of the settlement and on the downside of the terrace. The south-east corner of the Large East Church, however, was more exposed than the west and consequently that area of the church is less well preserved than the remainder.

Excavation and Recording

The Churches, Enclosure 4 Cemetery and North Tomb 1

A local workforce was employed, and an Antiquities inspector was assigned to oversee the work, as is required. That role was filled by an official from the Coptic and Islamic Inspectorate for the East Churches and the Pharaonic Inspectorate for the West Church Complex, Enclosure 4, North Tomb 1 and Kellis 2. The approach taken accorded with modern stratigraphic excavation techniques, and all artefacts were recorded. Given the paucity of information on early church architecture, it was decided that each of the churches identified during the survey would be investigated. The imposing nature of the North Tomb Group warranted the exploration of certain tombs and during the course of this, intrusive Christian burials were revealed. Once similar burials were discovered in Enclosure 4, it was targeted for full excavation. The importance of investigating Kellis 2 is axiomatic.

All buildings chosen for excavation are given a reference number comprising the DOP Archaeological Site Index site number followed by the Area designation and a numerical value based upon the excavation sequence for a given area. For instance, the Large East Church is designated 31/420-D6-1/A/7; A represents Area A, and 7 being the seventh structure/test to be excavated in that section of the site. The Small East Church, 31/420-D6-1/A/8, was the next structure in Area A to be excavated.

The buildings were visible on the surface enabling targeted excavation. The stratigraphy is simple. In most spaces the floor was covered by an accumulation of powdery material, which is usually covered with sand that had blown in after the buildings were abandoned. In some spaces mud-brick collapse from the roofs and walls was encountered at various depths and density. Each deposit was assigned a context number, removed, and recorded noting the composition of the matrix, the location and depth of the deposit and its relation to the other strata. As rooms and known features within a church such as nave and apse, were instantly recognisable, it was decided to name them accordingly, rather than to employ the term room, space, or locus. The same approach was adopted for walls, doors, cupboards, and other obvious architectural features. Moreover, it was deemed unnecessary to allocate context numbers to the latter.

Most of the excavation was undertaken before the wide-scale use of computers and digital cameras. Records of the excavations were kept in field notebooks and on context sheets. Plans and sections were drawn in the notebooks by the excavators, and in publication-quality by on-site architects. Site photographs were taken in both black and white and 35 mm slide format. Two 1 m scales, each divided into two 50 cm sections, were used for site shots. All material remains were placed into bags labelled by provenance and date of discovery and a full description of each was entered into sample books and retained for further study.

Recording followed the protocol adopted by the DOP for the survey. The finds were divided into two categories: objects and samples. Complete objects and those deemed to be of significance, such as broken figurines etc., were registered and designated numbers that included the site and structure reference followed by the object number. For example: the bronze figurine found in Room 1 of the Large East Church was the first object from the church to be registered; its registration number is 31/420-D6-1/A/7/ 1. Samples include broken objects, pieces of rope, glass sherds, fauna and flora, fragments of textile, basketry, shoes, pieces of worked wood, and minerals, amongst others. They were given similar identification numbers to the registered objects except that the item number was replaced by S followed by the final two digits of the year in which they were found, and a sequential sample number. A piece of leather from the Large East Church, for instance, has the sample number 31/420-D6-1/A/7 S96.93. Fallen architectural elements were described by the architects, and the wall paintings were conserved and described in situ; decorated plaster fragments were not assigned sample numbers but were kept according to their find spot for identification and to enable joins to be made. Much of the papyri was found in fragments; they were placed in humidifiers and framed by the conservators. As the papyrologists were able to join fragments, often from different deposits arranged in different frames, it became clear that allocating sample and registration numbers to each set of fragments was untenable. Instead, they were allocated P. (papyrus deposit) numbers; those published then received a P.Kellis Gr. or P.Kellis Copt. number, denoting Greek text for the former and Coptic for the latter. Potsherds are ubiquitous throughout the site and are found in almost all context types; the systems of recording are outlined below by Hope. Every effort has been made to record all finds, but the documentation is incomplete. We have been unable to access the on-site storage facilities for the last 10 years and consequently, small deposits of potsherds remain inaccessible, as do certain categories of samples. The silver and amethyst ring was surrendered before it could be cleaned and only a cast could be made.

Once the objects had been allocated a registration number they were conserved, where necessary, photographed, again in black and white film and 35 mm slides, and drawn by the illustrators; full descriptions were recorded on index cards. The objects were then studied by the appropriate specialists. Certain items were selected by the inspectors for registration within their system and were taken to the official magazine for storage. Access to the magazine was permitted, and with the advent of digital photography some objects were rephotographed. Those objects not registered by the authorities together with the samples were kept in the site or dig-house workrooms. Colour slides have now been converted into digital images and are stored online as are plans, sections, line drawings, and photographs. Records are now arranged in a series of databases.

Following the excavation of the Small East Church, the wall paintings were consolidated by the conservator, a wall of reused mud bricks constructed within 10 cm of the paintings and the intervening space filled with clean sand. The structure was then backfilled. The Large East Church has been backfilled with clean sand. The West Church Complex and Building A in Enclosure 4 have been engulfed by sand. All graves in Enclosure 4 and North Tomb 1 were backfilled following their recording; the human remains were placed in the hands of the bioarchaeologists.

Kellis 2 Cemetery

The bioarchaeologists devised the strategy and the recording system. Their aim is to understand the demographic profile of the oasis and consequently their focus is upon the recovery and examination of the human remains, and for this a concentrated area of the cemetery has been targeted for excavation (Molto 2001). The strategy is straightforward; excavate the grave to retrieve the human remains for study and backfill the grave. The prevailing winds soon engulfed the tomb enclosures. The archaeology of the individual graves and the enclosures was undertaken initially in field notebooks, and on context sheets from 2003. The following elements are recorded for each grave: the co-ordinates; the dimensions; any architectural features; whether the grave was intact or disturbed; the body alignment and position of arms and legs; body wrappings and inclusions; material remains such as grave goods, ceramics and flora. The plan of the grave or burial is sketched as is any sub- or superstructure. All graves are photographed, initially in black and white and 35 mm slides, and currently with digital photography. Plans of the tombs and sections of graves are drawn to publication standards. The master plan of the cemetery is added to following each field season and a typology of the graves has been developed.

Arrangement of the Volume

The current volume is arranged in sections; the first three focus on one church in order of its assumed foundation date. The individual sections stand as separate units. The first three include the excavation history, architecture and stratigraphy, followed by detailed accounts of the material remains and finally a discussion. This preserves the integrity of the individual monument and ensures the integration of the artefacts with the specific monument from which they were recovered. Categories of artefact are described and discussed by specialised contributors; these include in alphabetic order Colin Hope, for the ceramics, lamps and jar seals; Marie-Dominique Nenna for the glass, Anna Stevens for the terracottas; and Helen Whitehouse for the bronze statuette, acanthus leaf and sliver/amethyst ring. I have undertaken the study of the coins, wall paintings and miscellaneous items. C. S. (Rufus) Churcher identified the faunal remains, Peter Sheldrick the human remains, and Ursula Thanheiser the flora. Laurence Blondeaux has described her conservation of the wall paintings in the Small East Church, and Jaroslaw Dobrowolski’s assessment of the building sequence of the Large East Church is included, as well as Iain Gardener’s appraisal of the graffiti in the Small East Church, and Andrew Connor’s study of the jar-sealing inscriptions. I have summarised Klaas Worp’s translation of the ostraka. The results of the study of the human remains from the West Church and Enclosure 4 Cemetery were undertaken by J. Eldon Molto and Peter Sheldrick and those from North Tomb 1 by Tosha Dupras and Matthew Toscheri. The descriptions of the excavations are based upon my field notes, and those of Carla Marchini, who undertook the excavations in the North Pastophorium, and Joyce Haynes who excavated the South Pastophorium, both in the Large East Church.

Section IV has a different format with three sub-sections, each focused upon an individual cemetery. For the information on Kellis 2, I am reliant upon the excavation notes and context sheets complied by the various excavators, supplemented by Sandra Wheeler’s PhD dissertation, which includes information on the archaeology not available in the field notes. Photographs provided by Peter Sheldrick have been invaluable and have given me a greater understanding of some architectural features. I ceased collecting data in 2011 after 700+ graves had been excavated and access to the site became more difficult. Colin Hope has undertaken the study of the ceramics from the cemetery, Rosanne Livingstone examined the textile remains that she had access to, and Ursula Thanheiser identified the flora available to her.

The architectural appraisal of the churches was initially undertaken by James Knudstad; in 1997, Jaroslaw Dobrowolski examined the architecture of the Large East Church, followed in 1998 by Barry Rowney, who also drew plans and sections in the Small East Church and the West Church Complex. The site and object photographs were taken by Colin Hope; those of the painted glass and some of the wall paintings from the Large East Church by Carlo Rindi Nuzzolo. The various illustrators were supervised by Bruce Parr, who has also prepared the final drawings, compiled the data bases, and modified various plans, sections and line drawings.

The Presentation of the Ceramics

(Colin A. Hope)

The ceramics from the monuments published in this volume are organised by excavation areas and units, and then by find contexts, which are grouped into depositional/archaeological categories, namely fill, floor deposit, floor and sub-floor. Where multiple find contexts fall within a single category, such as fill in the Large East Church, then they are grouped and items from individual contexts are not distinguished. As the majority of the material derives from disturbed contexts, no attempt to estimate the minimum number of individual vessels has been undertaken and, on the whole, only one or a small number of examples of each type is illustrated from each category. Thus, the range of material is indicated but not the total quantity. However, the frequency of individual forms and the wares in which they were manufactured can be determined from the material given here. The study is based upon diagnostics (rims, bases, handles), although some body fragments of significance are included. Diagnostics with better contextual integrity are presented in their entirety. As a result of presenting the material by depositional categories and find contexts, no typology of forms within the major pottery fabric groups is presented. Several typologies of material from the same time frame as the ceramics within this volume and from sites within Dakhleh have been published recently (Dixneuf 2015; 2018; Caputo et alii 2020), and should be consulted by those seeking such typologies.

Material from each context or groups of contexts is divided according to the latest categorisation of material used in its manufacture. The Dakhleh Oasis Project Pottery Fabric/Ware Typology (Table 1) was commenced during the first season of the Dakhleh Oasis Project (DOP) survey of the oasis in 1978 and based upon an extensive collection of sherds from the surface of Mut al-Kharab. The fabrics were separated into two main groups, labelled A and B, distinguishing what, at the time, were identified as iron-rich fabrics from others that were primarily calcium-rich. The system, its chronological range, the diversity of fabrics and the precision in defining the characteristics of the material, have evolved considerably since then, and the process is indeed on-going. No attempt was made to distinguish or number fabrics by period of use, mainly with the aim of documenting the longevity of individual fabrics and to simplify the numbering system. The latter evolved incrementally as what were thought to be distinct fabrics were identified, and thus fabrics with adjacent numbers in each group may not resemble each other nor belong to the same time frame. In the four-and-a half decades of studying Dakhleh ceramics many changes have been made to the system and it is now possible to document the periods of use of many of the fabrics and indicated in Table 1. This has been made possible as a result of excavations at several sites, largely subsequent to the survey of Dakhleh, at which well-stratified deposits of ceramics have been found that can be dated relatively securely. For the historic periods, these range from the Old Kingdom onwards, though some phases are better understood than others. Of relevance to material in this volume, finds made across Kellis have enabled clarification of the evolution throughout the Roman period in Egypt, and publications of this material are referenced herein, though the majority remains as yet unpublished. Most derive from excavations within residential areas, the dating of which is discussed elsewhere (Hope 2015). The conclusions are supported by those at Amhida and ‘Ain al-Gedida (see references cited above), and to a lesser extent at ‘Ain Birbiyya, while Mut al-Kharab has enabled the characteristics of Ptolemaic period pottery in Dakhleh to be defined (Gill 2016). The nature of the main structures from which the ceramics published here derive, in association with other dating criteria, such as numismatics, render the date of the majority within the fourth century, and especially 330-380, certain. Extensive reference to parallels from other sites to date this material is therefore unnecessary.

The Dakhleh fabrics have been studied by several people who have also been responsible in some cases for its classification and recording: myself, Shirley Patten, who undertook a major investigation of Dakhleh ceramics from the Third Intermediate Period to Roman period (Patten 2000), Amanda Dunsmore and Andrew Jamieson. Of especial significance is the work of Mark Eccleston, which has refined the fabric typology and classification, and enabled a more detailed appreciation of the major groups. His work commenced with an evaluation of the Neutron Activation Analysis results of a large number of sherds from the early stages of the survey of Dakhleh determined by Professor R. Heiman of McMaster University (Eccleston 1997) and progressed to mineralogical examination at both the macro- and micro-level (Eccleston 1998; 2006), the latter study focusing especially on material from Kellis. His work confirmed the basic division between A and B fabrics, but showed that some fabrics initially assigned to the B group were in fact A fabrics. Fortunately, no fabrics assigned to group A belonged to B. Eccleston distinguished six main groups of interrelated fabrics with one, Group 2, divided into 2A and 2B. Groups 1–5 are ferruginous-rich fabrics and Group 6 is higher in calcium. He further indicated that Groups 2a/2b/3–5 are similar in many ways and distinct from Groups 1 and 6. His work also compared the Dakhleh fabrics with others from Kharga Oasis and the Nile valley, showing great similarity with the former and dissimilarity with the latter. It can be noted that in the publication of the material from the pottery workshops at ‘Ain Asil of the Old Kingdom to First Intermediate Period, it was proposed that the clays used for what is here labelled Group 1 derived from the Cretaceous basal Nubian strata and those designated 2a/2b derived from Upper Cretaceous strata (Ballet and Picon 1990, 82–3).

Amongst the material from Dakhleh there are sherds manufactured outside Egypt; these were not included in the later analytical programme and have been identified on the basis of visual characteristics. A few were, initially, incorporated into the DOP system, but are now referred to by other designations, and been removed from the system. Hence, all of those fabrics included with Groups 1–6 can be regarded as local to Dakhleh and/or Kharga; this is reinforced by comparison with pottery fabrics used by contemporary potters in Dakhleh.

Table 1 presents the grouping of the Dakhleh pottery fabrics as currently understood, indicating which fabric belongs to which group and the surface treatments identified with each. The combination of a specific fabric and specific surface treatment is used to produce the code used in recording the material. The table includes all of the fabrics of the historic periods; prehistoric material has been studied separately and published elsewhere (Warfe 2018). Eccelston’s work in its final stages focused upon Roman pottery, and his grouping is based upon material of those dates. In Table 1, I have assigned to his groups fabrics of other dates and those of the Roman period for which Eccleston lacked samples; they are given in parentheses. I have also indicated the known date range during which the fabrics have been documented; these are, obviously, open to revision in light of future study. The dates are derived from various studies of material from the DOP concession: Amy Pettman for the Old Kingdom (Pettman 2008; 2016); Seamus Scorgie for the Second Intermediate Period (Scorgie 2003); Richard Long for the New Kingdom (Long 2021) and Third Intermediate Period (Long 2007), Caroline Hubschmann for the Late Period (Hubschmann 2009), and James Gill for the Ptolemaic period (Gill 2016), plus a specific study of fabrics used in the manufacture of kegs and flasks by Hope and others (2000). I acknowledge the work of all in refining the Dakhleh fabric system. Naturally the work by colleagues elsewhere in the oasis at ‘Ain Asil, especially Pascale Ballet and Sylvie Marchand, and at Amhida and ‘Ain al-Gedida, especially Delphine Dixneuf and Clementina Caputo, is of major importance in confirming the viability of the system for use throughout Dakhleh and refining the chronological implications. While the DOP system is employed for the work at the latter two sites, at the former different designations are employed. In addition, the work of Pascale Ballet and Sylvie Marchand in Kharga enables comparison between the two oases and further refines the chronology.

The main pottery fabrics identified amongst the material published in this volume are described briefly below, based upon the detailed discussion in Eccleston 2006. They are also discussed in the works cited above on the ceramics from Amhida and ‘Ain al-Gedida. Some fabrics not attested amongst the current body of material but known from the same general time frame are also included. While only one fabric has been allocated to each of Groups 3–4, the designation group has been employed to allow for other fabrics to be added that might be identified in the future.

Brief Description of Dakhleh Pottery Fabrics

Ferruginous Fabrics: Groups 1–5

Group 1: Coarse, available widely throughout the oasis and in use before the historic period; from a sedimentary environment.

Group 2: Mudstone/claystone/shale fabrics with quartz; not really shale but platey, lath-like and pellet inclusions. Employ dense grey clays similar to those currently found at the base of the escarpment that are now used for water jars; more difficult to procure than clays used for Group 1. Shows the greatest diversity of all the groups.

Table 1 The Dakhleh Oasis Project fabric/ware typology.

Abbreviations:

OK = Old Kingdom, SIP = Second Intermediate Period, NK = New Kingdom, TIP = Third Intermediate Period, LP = Late Period, Ptol. = Ptolemaic, R = Roman, LR = Late Roman, // = parallel to, = equated with

Group 2a: Significant quantity of ‘shale’ inclusions.

Group 2b: Fine quartz inclusions as part of groundmass and less ‘shale’ than Group 2a, which has grey to green colour, though fired variable colours; contain feldspar and mica.

Group 3: Mudstone/claystone fabrics with significant quantity of vegetal temper. While no variations have been isolated for material of the Roman and Late Roman periods, a comparison with material from earlier periods, from the Old Kingdom onwards, may show several.

Group 4: Mudstone/claystone fabrics with sandstone; very rare.

Group 5: Fine mudstone/claystone fabrics with shale.

Coarse Quartz Marls: Group 6

Calcareous marl clay fabrics with light-weight, open texture, fired cream to greenish-grey, presenting a continuum in porosity and coarseness. Groundmass similar to Group 2b but with less quartz and probably of different origin in the Dakhleh Calcareous Silty Sediments.

Imports

The quantity of imported ceramics found within the Kellis churches is extremely small and does