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.
The excavations at Kellis, where three churches have been discovered so far (see §5.3), have shown that Christianity was well established in the Dakhla Oasis in the first part of the fourth century. In light of Trimithis' size and particularly of its administrative, economic, and also cultural role there, the city seems likely in its last decades to have seen its urban landscape dotted with churches and other places of Christian cult and congregation, as is suggested by the existence of numerous churches at other sites in Dakhla and—even more profusely—in the neighboring Kharga Oasis. Any hesitancy in this assessment is largely owed to the fact that up to very recent years, no archaeological evidence of churches or any other type of Christian monuments had been detected at Amheida.↩#p185
During the 2009 conference of the Dakhleh Oasis Project in Lecce, Roger Bagnall and Raffaella Cribiore presented the limited then-available documentary evidence concerning the existence of a Christian community at Amheida.1 This material, which is discussed in more detail in the following section (§5.2), comes primarily from the temple area and strongly suggests Christianity's presence at Trimithis during the fourth century.↩#p186
Figure 93. Amheida: The church (B7) before excavation (view to the south).
In 2012, however, archaeological investigation was begun in Area 2.3, situated to the east of the public bath, B6, and roughly halfway between the north and south edges of the city. The precise location is the top of a mound that grants an unobstructed view (likely also in antiquity) over large portions of the city. The primary goal was to excavate a large building (designated B7) that had been mapped a few years earlier during the topographical survey. At that time, several features were noticed that, based on comparative evidence from other sites in Dakhla and Kharga, suggested a possible identification of this space as a church. In fact, most architectural features of B7 were already visible—although heavily eroded—at present-day ground level before we started excavating (Fig. 93). Only one cluster of partially collapsed walls was standing at some height above the ground, to the south of a large rectangular space labeled Room 1. Throughout the site we can observe the existence of buildings or rooms that stand at a considerable height, in apparent defiance of the destructive forces that have leveled most of the site. This phenomenon still awaits convincing explanation.↩#p187
Figure 94. Plan of church B7 at Amheida.
Room 1 is a large rectangular space (about 12 × 13.65 m), oriented toward the east (Fig. 94). It was part of a larger complex, which included a set of smaller spaces built along the south wall of Room 1, plus three roughly square rooms along the east wall. Further to the east, where the ground slopes down quite sharply nowadays, the topographical survey revealed extensive traces of a large rectangular space whose precise layout, function, and relationship to the rest of the complex are also yet to be determined.↩#p188
The three spaces along the east wall of Room 1 have almost completely disappeared, due to human intervention or (more likely) wind erosion. What is currently visible at ground level is the upper part—including sectors of mud-brick vaults, still in situ—of substructures once supporting those rooms, which were roughly at the same level as Room 1 and once opened onto it.2↩#p189
Figure 95. A passageway leading to church B7 (view to the east).
The main entrance to Room 1 was most probably somewhere along the west wall. Against this wall—and inside the room—are two square bases that define a space situated exactly along the room's main east–west axis. Topographical work revealed evidence of what might be a narrow passageway precisely in line with the west doorway and that main axis (Fig. 95). If the topographical relationship of this passageway with Room 1 is corroborated by future excavations, it may provide the physical link (or one of the links) connecting B7 with the urban fabric of Amheida to the west and therefore support the identification of the west doorway as the main access to the building. The two square bases have very deep foundations, suggesting that they might have supported heavy architectural elements, such as engaged semi-columns or pillars to frame a relatively monumental doorway.↩#p190
Two other doorways were found in the southeast corner of Room 1. One of them opened onto the southernmost of the three small spaces to the east, while the other led to a staircase ascending southward and, via the landing of the staircase, to a set of seven spaces to the south of Room 1. Some of these spaces, whose nature and functional relationship to the church are generally difficult to determine due to their poor condition and the lack of evidence, show clear signs of having been reused after they were abandoned. Among these signs are two clay and mud-brick stoves that were built in two of the rooms after the ceilings had collapsed—as evidence that we found at floor level indicates, and as we assumed from the fact that stoves were generally built in open spaces to allow the smoke to escape.↩#p191
Figure 96. Room 1 of the church showing a line of colonnades (view to the southeast).
Room 1 was divided into a central nave and side-aisles by two rows of mud-brick columns. The colonnades ran east–west, and their foundation walls were bonded at their east and west ends with north–south-oriented foundation walls, forming a rectangle. The two colonnades were seemingly joined at their west end with a return aisle,3 which created a sort of ambulatory along the inner perimeter of the room (Fig. 96). Our excavation revealed only the foundation walls of the colonnades and the square bases and lowest courses of a few columns, including two pillars of heart-shaped cross-section at the east end of the north and south colonnades. The foundations of an indistinct, roughly rectangular mud-brick feature are visible at the east end of the north colonnade. The location of this architectural element suggests that it may have been a pulpit. Indeed, this feature seems to find robust comparison at the fourth-century church of Shams ed-Din in Kharga, where a pulpit was built at exactly the same spot and in a shape that is not dissimilar.4↩#p192
A mud-brick bench, or mastaba, for the audience was found running for more than 5 m along the inner face of the church's south wall. It is likely that a mastaba like this ran also along the north wall of the room, although no traces are visible nowadays because of the poor preservation of that wall. The building had a floor of compacted clay, laid out in different phases, this too now poorly preserved. Even where the floor is more clearly discernible, we noted a significant difference between the elevation of the latest floor level in the north and south aisles and that of the nave of the church, which is about 10 cm higher.↩#p193
Figure 97. The stepped platform in Room 1 of the church (view to the southeast).
A well preserved rectangular platform abuts the inner face of Room 1's east wall. It was constructed centrally, in line with the west entrance and along the room's main axis (Fig. 97). Access to the platform was not frontal but via two flights of steps, one on each of its north and south sides. To the north and south ends of the platform are two rectangular mud-brick bases; these still bear evidence, though scanty, of the lowest course of engaged semi-columns, which framed a large opening (almost 2 m wide) onto a raised space to the east of Room 1. In fact, this platform provides conclusive evidence that this room was once architecturally linked with the central space to its east, now almost completely lost. This was presumably where the apse was, flanked by two side-chambers.↩#p194
The immediate fate of B7 after its abandonment is unknown. Once again, the entire structure shows, in its current state, noticeable signs of severe erosion, particularly along its west side. Architectural features are visible above the original floor level only in the eastern half, to a maximum height of about 0.75 m in the southeast corner. As mentioned earlier, on this mound only Room 7, to the south of Room 1, withstood erosion and human destruction, while most nearby walls were either destroyed or eroded to foundation level.↩#p195
Figure 98. A test trench dug along the north aisle of the church (view to the east).
The church had impressively deep foundations, reaching 3 m below the ancient floor level in the northwest corner of Room 1 (Fig. 98). B7 was built, at least in part, by first cutting through a leveling layer of broken mud bricks. Although it is yet to be determined whether this leveling layer lay on top of bedrock or other deposits—perhaps even the remains of earlier buildings—the church's foundations testify to the considerable effort that was put into the construction of a sizable, even monumental, edifice.↩#p196
Our investigation of the relatively limited stratigraphy inside Room 1 revealed some evidence for the collapse of walls, which seem to have once been covered, at least in large part, with white plaster. What is particularly worth mentioning in this context is a Greek inscription painted on a fragment of wall plaster that was found to the north of the stepped platform, although not in situ. Apart from several hard-to-read letters, the words ho theos (Greek for "God") were easily identified, thereby offering further evidence to support the assertion that the building was a church.↩#p197
Figure 99. Church (B7), large fragments of painted plaster from the ceiling.
Even though the walls seem to have been largely undecorated, at least in Room 1, the ceiling was another matter. We found ample evidence of a collapsed flat ceiling, including thousands of fragments of painted plaster. Many of these are still attached to a thick layer of mud that bears the imprints of palm ribs, pointing unmistakably to their original location as part of the ceiling. The larger pieces, as well as innumerable fragments, were collected and studied by Dorothea Schulz, with the goal of reconstructing the decorative program that once adorned the church's ceiling (Fig. 99). Both the larger patches and the fragments display a wide array of shapes, colors, and motifs. Among them is a purple band ornate with lines in different shades of yellow and a simple vegetal motif, which may have framed the decorated ceiling. The main design seems to consist of geometrical patterns including squares, triangles, and lozenges forming stars, as well as other interlocking shapes such as pentagons, hexagons, and octagons.↩#p198
The choice of these geometric forms and their elaborate combination tell us that the original plan was undoubtedly to create the effect of a coffer design, so intricate as to be very difficult to fully reconstruct. The aesthetic appreciation of the ceiling's decoration must have been considerably enhanced by the use of brightly colored shapes outlined by thick black lines and in dramatic contrast to the less ornate (if not—as mentioned above—entirely white) walls that once supported the roof.↩#p199
Comparative evidence for this type of ceiling decoration can be found in Dakhla, but in the context of domestic architecture. A Roman house dated to the second century CE, excavated by Colin Hope at Kellis, revealed an elaborate decorative project, including remains of a collapsed ceiling ornamentation of interlocking geometric shapes.5 These included rectangles, octagons, and lozenges forming eight-pointed stars (like those at Amheida), which, according to Helen Whitehouse, who studied the house's decoration, were meant to reproduce the perspective effects of a coffered ceiling.6 This seems to be the case at Amheida too. However, it is worth adding that the decorative program in the Kellis house included not only geometric shapes but also busts of divinities in octagonal frames, whereas no figural ornamentation was added to the church ceiling at Amheida.7↩#p200
In the neighboring oasis of Kharga there is evidence for the adoption of a very similar design, within a Christian funerary context and of a time closer to the church of Amheida. Inside one of the larger family tombs at the necropolis of Bagawat is a chapel (number 25) that has a vaulted apse ceiling, decorated at least in part with the same range of interlocking squares, lozenges, and octagons, although the choice of colors is different.8 This example along with the fragments from the ceiling of the Amheida church suggest that in Egypt a decorative style using geometric shapes to create the visual effects of coffering enjoyed a long life. It was certainly well known in the Roman period and continued throughout Late Antiquity, when it was adopted also within the context of Christian public and semi-public architecture.↩#p201
The funerary nature of the chapel at Bagawat is even more striking when one considers the eight human burials found in 2012 and 2013 inside B7. One skeleton was found in situ near the northwest corner of the church, although most of the burial pit had disappeared due to either human disturbance or erosion. All the other pits were intact. No funerary goods were found associated with the excavated tombs. The orientation of all the bodies was with the heads placed to the west (so that on rising the person would face east), which is quite standard for Christian burials, and compares with evidence from other Christian cemeteries in the oasis; we find it, for example, at Kellis, Deir Abu Matta, and at a site northeast of Muzawwaqa.9↩#p202
The physical anthropologists Tosha Dupras and Lana Williams analyzed some of the bodies and supplied a significant amount of data on the gender, age range, pathologies, and possible causes of death of the individuals who were buried in these tombs. Quite thought-provoking is the fact that people of both sexes and of all ages, including at least one adolescent, were buried inside the church, even in prominent spots near the sanctuary; this should lead us to think about how sacred space was used in funerary churches and—more specifically—about who its beneficiaries were, after death, within the ancient local community.↩#p203
Figure 100. The funerary crypt under the north pastophorion of the church (view to the southeast).
Three of the eight burials were found not inside or alongside the church's main body, but in what turned out to be one of the most exciting discoveries about this building. Excavations under the now-disappeared north side-room (presumably a pastophorion or sacristy) led to the discovery of an underground funerary crypt (Room 2 in the plan) (Fig. 100). It is a very well preserved space, with a mud floor and only the uppermost part of the vault no longer in place. Inside it are three sealed tombs with mud-brick superstructures. Through a doorway in the south wall, the crypt opened onto a so far unexcavated space below the apse that may have been part of the same crypt.↩#p204
It is beyond doubt, based on the rich body of available evidence, that B7 at Amheida was purposely built as a church for the local Christian community in the early fourth century. The church had features—a basilica with a central nave, side-aisles and west return aisle, a sanctuary accessible via a raised platform and flanked by service rooms—that were relatively standard in the Christian architecture of late antique Egypt. Other churches in the region of the Great Oasis, generally datable to the same period, present significant similarities to Amheida's. In particular, the Large East Church at Kellis shows a very similar layout—a central nave, two side-aisles, a west return aisle, and a set of rooms opening to the south of the building.10 Furthermore, the apse area was once accessed from the nave via a rectangular platform with steps built along its north and south sides, reflecting the arrangement adopted at Amheida. Another building that offers a significant number of similarities is the church of Shams ed-Din (Kharga), mentioned on p. 122. Here too, the layout reflects the basic architectural design of the church at Amheida, including a long rectangular platform placed in front of the apse, although at Shams ed-Din it was accessed frontally and not from the north and south sides, as at both Amheida and Kellis. The large space to the east of Amheida's church is also quite suggestive—although it is yet to be investigated—of a courtyard adjoining the Shams ed-Din church, which was identified by its French excavators as a kathesterion (or sitting space for visitors, equipped with benches around the walls).11 Clearly, Amheida's church fits with the architectural standards that were so widely adopted in the region of the Western Desert and throughout Upper Egypt in the fourth century.↩#p205
As regards the nature of the liturgical practices once carried out inside the church, the evidence is scanty at best. However, what is undeniable is that the presence of burials in the church and in the crypt suggests that B7 served, at least at some point (and probably from the very beginning, as the existence of the crypt itself would suggest), as a funerary church. Certainly, it remained in use after the deposition of the burials near the sanctuary, since the mud-brick coverings of the pits were sealed by a floor that extended throughout the east half of the nave and side-aisles. The association of a church with funerary practices is well attested at this time in Dakhla, for example at Deir Abu Matta, where Gillian Bowen discovered not long ago several burials inside the church.12 The West Church of Kellis may also have been associated with funerary practices, as the graves inside it, near the apse, and immediately to the east of the building, suggest.13↩#p206
The presence of the church in Amheida shows that the Christian community had the means to choose a prime location for the construction of such a building. We may suppose that B7 was one of several churches, considering the size and importance of the city, and another probable church stands at the northwest side of the cemetery. Indeed, B7 must have enjoyed a high degree of visibility as one of the city's most striking landmarks. This was the case elsewhere in Dakhla, for example at the rural settlement of ʿAin el-Gedida (§5.3). Here, a church was strategically situated at the center of the main mound, overshadowing the remains of an older Egyptian temple. Although Amheida and ʿAin el-Gedida are hardly comparable sites, with regard to size, monumentality, and administrative or economic significance, they effectively show how Christianity had become, by the early fourth century, a key element of the religious landscape even in this remote oasis of the ancient world.↩#p207
Even before we had reliably identified a church at Trimithis, there was adequate evidence for the presence of Christianity in the city, not least the titles of members of the clergy and a number of Christian personal names. Priests, presbyteroi, occur in two ostraka, and deacons in five. Christian names, as noted earlier, include Martyrios, Makarios, Paulos, Timotheos, Psenpnouthes, and also no doubt Moses, Ephrem, Jonah, and Joseph. Few of these are commonly found, however, and the direct evidence for Christianity from the ostraka would best be described as modest.↩#p208
One striking find from the temple hill is an ostrakon discovered in 2008 (O.Trim. 2.819) bearing a list of names including Jacob and Abraham, and headed by ho pater, "the father" (or "our father"). The other names, however, are not distinctively Christian, and fading or breakage has removed any numbers that may have been written on the right side of this ostrakon, if it was in fact, as we suppose, an account.↩#p209
A graffito of one Horigenes, son of Ioannes, on a stone is also probably Christian. Even though Horigenes is a theophoric name derived from Horos, it was also the name of the famous early-third-century Alexandrian theologian and scholar Origen. More importantly for our purposes, Ioannes is distinctively Christian.↩#p210
Figure 101. A graffito with an invocation of Ammon (infrared photo).
But most arresting of all is a stone block (inv. 3053) in the middle of which someone has written a Greek verse. This comes either from an altar (as Paola Davoli has suggested) or from the base of a statue (as Olaf Kaper thinks). The ink is quite faded in parts and demands persistent and prolonged autopsy—in natural-color photographs the text is almost illegible (Fig. 101). It reads: ↩#p211
ἀνθρώπων βιότοιο κυβερνήτης μέγας Ἄμμων [Great Ammon is the pilot of the life of men.]
This is a perfect hexameter line with an epic ring that calls to mind various literary reminiscences. It is in a way a kind of pastiche, reflecting the religious and cultural syncretism of Egypt.↩#p212
The concept of steering oneself through life occurs once in a letter of the Ptolemaic period. The metaphorical use of the verb "to pilot" and its noun becomes more frequent in literary attestations of the Roman period. In the second century, it is worth noting, Dio Chrysostom (in Or. 63.7.8) writes that "Tyche [the goddess of chance] governs a man's life," using the verb in a sense very similar to our line on the stone. What makes Dio's words notable is the fact that he is referring to a deity as the pilot of mortal life. In fact, man is usually presented as his own pilot, while life and its troubles are regarded as the waves of a tempest. Thus the Cappadocian Fathers, who frequently employ the phrase, encourage man to overcome the waves of misfortune they encounter, becoming their own pilots of their life; for example, Gregory of Nyssa, in his On Virginity, says that the good man, "like a good pilot with his boat, looks only up to heaven in guiding his life." Like Dio, however, John Chrysostom, in his On Genesis (In Genesim 53.118.16), regards God as the pilot: "We navigate through the sea of our present life, led by the great pilot, God."↩#p213
Ammon, the Greek form of the name of the Egyptian god Amun, was the dominant traditional god of the oases. The hellenized Ammon had his main sanctuary in the Siwa Oasis, which Alexander the Great famously visited in early 331; from this visit originated the claim that Alexander was Ammon's son. Ammon appears as the great god in the Historia Alexandri Magni (the "Alexander Romance").↩#p214
Although the concept that the gods, and one god in particular, give life to men is uncommon in Greek literature, it is a commonplace in Egypt from as early as the Old Kingdom. Amun was regularly seen as the source of life, but more interesting are the remarkably direct invocations of Amun as the pilot of life found in New Kingdom prayers discussed by Jan Assmann.15 One describes him as "Pilot who knows the water, rudder that does not lead astray." "You are Amun who comes to him, who calls unto him, the pilot who knows the water, the rudder that does not lead astray," goes another. "If a man's tongue is the boat's rudder, the Lord of All is its pilot," says yet a third. If our graffito of Horigenes is Greek in expression, the sentiment, by contrast, is deeply rooted in Egyptian religion. Although Amun was not the god to whom the temple of Amheida was dedicated, he was, as already noted, the principal god of the western oases. Olaf Kaper tells us that there is an image of Amun-Re of Hibis (Amenebis) on the gateway of ʿAin Birbiyeh, decorated under Augustus, which attributes to the god the following titles: "Amun-Re Lord of Hibis, the Great God, strong of might, King of the Gods, who gives [this verb is uncertain] the breath of life, who lets the constricted throat breathe, who causes all that exists to live."↩#p215
A good pagan hexameter line, then, Greek in expression but deeply rooted in Egyptian thought, written on a stone block in the temple of Thoth. Now, above this line are some rather faint traces of writing in a much smaller hand. We believe that at the top right it is possible to make out ete pnoute, which we take to be a Coptic gloss on the whole inscription, putting forward the view that it is God, pnoute, who is the governor of life, not Ammon. A bit of not entirely friendly religious dialogue in late antiquity, it would appear.↩#p216
These graffiti certainly indicate that by Amheida's last period of occupation, which on present evidence seems to be the last quarter of the fourth century, the temple was no longer in use as such, but was accessible to Christians who wished to leave a mark of their own religion on the structure and its contents. This is hardly surprising, but we have no means of telling at what date it became possible to do so. The excavations in the temple area also uncovered a considerable number of tags from jars of the same type found in the large fourth-century house designated B1. At least one of these is from the fourth century, namely O.Trim. 1.127, dated to a year 33, which is to be assigned to the reign of the emperor Constantius II and thus corresponds to the year 356/7. There are two other indications that activity on the hill was continuing at this time: the mentions of Psais the deacon, known from two other ostraka (O.Trim. 1.383), and of Nikokles and Philippos (O.Trim. 286). Both men are well known from the final occupation phase of House B1, during the 350s and 360s. We cannot, however, be certain from these ostraka whether the activity in question consisted simply of dumping trash on a partly abandoned hill, or whether it had something to do with habitation on the hill itself. Nor do the graffiti tell us the answer to this question, as their presence is consistent with either hypothesis.↩#p217
Figure 102. An ostrakon bearing an account headed "the father" (or "our father") (O.Trim. 2.819).
We might draw a slight hint from the list or account mentioned earlier, headed by "the father" (or "our father") (Fig. 102). A similar text was found in 2008 during Gillian Bowen's excavations of a church and perhaps a monastic site at Deir Abu Matta (situated between Mut and Amheida). The meaning of "father" in neither case is entirely clear, but we know of no similar texts of this period in which it refers to any secular office in local or Imperial government. If the reference is to leadership in some kind of religious community, this would certainly be consistent with the Deir Abu Matta excavators' hypothesis that that church was attached to a community of some kind, the adjacent building perhaps being a monastic keep.16↩#p218
Nothing has been found at Amheida so far to indicate with certainty the presence there of any monastic establishment. But given the present-day condition of the top of the temple hill, it is impossible to exclude any hypothesis—and this includes the notion that in the fourth century, interspersed with the contemporary brick buildings there may have been places for dumping debris from nearby parts of the city that were still fully active.↩#p219
5.3. Other Churches in Dakhla: Kellis, ʿAin el-Gedida, Deir Abu Matta, Deir el-Molouk, ʿAin es-Sabil
Nicola Aravecchia
More than a century ago Herbert Winlock, an Egyptologist affiliated with the Metropolitan Museum of Art, set off for the Dakhla Oasis on an adventurous trip.17 In his travel diary, published in 1936, he remarked on the astounding scarcity of Christian antiquities throughout the oasis, especially in comparison with the wealth of evidence from the nearby Kharga Oasis. For this state of affairs he blamed the relatively long distance of Dakhla from the Nile valley and its Roman garrisons, which exposed the oasis to the dangers of invasion and destruction by neighboring nomadic tribes—likely to have been the cause of most towns and villages of the region being abandoned during the late Roman period.18↩#p220
Winlock's remarks about the absence of substantial visible remains of Christian cemeteries and monuments—with the notable exception of the church of Deir Abu Matta—were undoubtedly justified at the beginning of the twentieth century. However, the picture has radically changed in the past few decades, thanks to intensive archaeological investigation carried out at several sites in Dakhla. We have already given some hint of the wealth of material for fourth-century Christianity in the oasis, and although it remains true that Kharga has more abundant Christian monuments and is relatively better known, there can be no doubt that Dakhla too proved to be a good location for thriving early Christian communities. The DOP survey recorded well over one hundred archaeological sites supplying evidence for human occupation in the oasis from about 300 to 700 CE.19 As Egypt was already in the mid-fourth century a profoundly Christianized country, it is likely that Christian communities were somehow linked to most or all of the "Byzantine" sites identified in Dakhla, but the mere fact that people from different ethnic, cultural, and religious backgrounds coexisted in Egypt in Late Antiquity prevents any easy generalization.↩#p221
It is, in particular, thanks to the work of Colin Hope and Gillian Bowen at Kellis that a large amount of evidence on early Christian architecture in Dakhla was first made available to the scholarly and general audience and became a point of reference for anyone interested in the origins of Egyptian Christianity.↩#p222
Figure 103. Plan of the West Church complex at Kellis (after Bowen 2002, 76, fig. 8).
Kellis (modern Ismant el-Kharab) is about 2.5 km southeast of the village of Ismant and roughly 11 km east of Mut, the capital of the oasis. In 1981, a DOP team surveyed the site, and large-scale excavations began in 1986. The DOP survey revealed substantial evidence of three churches at Kellis.20 One is located along the west edge of the village, and two are part of an extensive, multi-roomed complex at the south end of the settlement.21 The West Church, excavated in 1992–93, measures 15 by 7 m and is oriented to the east (Fig. 103). It consists of two rooms, one to the west, possibly functioning as a narthex, and one to the east, with a passageway placed centrally within the shared wall. Along the east wall is an apse with a raised floor, accessed up a step. The conch (semicircular apse) is flanked by engaged semi-columns and in front of it is a raised platform, reached from the west up another couple of steps. Two doorways, placed to the north and south of the apse, open onto small side-rooms. Low benches run along the walls of the two rooms forming the main body of the church, the only access to which is through a doorway in the south wall of the narthex. This opens onto a cluster of seven rooms, together with the church forming an architectural complex. The area covered by these spaces, the function of which is unclear, roughly equals that of the church.↩#p223
The only entrance to the complex is in the southwest corner; it opens onto a large rectangular room with benches, possibly functioning as an anteroom. Two Christian burials were found near the sanctuary, and others in its proximity outside the church. These discoveries led the excavators to identify the complex as funerary.22 The location of the two burials inside the West Church is very similar to that of some of the burials found inside the church of Trimithis. According to the numismatic evidence, the foundation of the West Church complex dates to around the mid-fourth century CE.23 The ostraka found in the buildings largely date to the third quarter of that century, and have links that can be established with similar material from the nearby site of ʿAin es-Sabil, mentioned below.↩#p224
Figure 104. Kellis, plan of the East Churches complex (after Bowen 2002, 66, fig. 2).
The two churches built on the southeast periphery of Kellis were, as has already been noted, once part of a rather large complex (Fig. 104). The so-called Small East Church stands near the southeast corner of its enclosure, built against the east wall. The church's overall dimensions are 10.5 × 9.5 m, and it is formed of two interconnected rectangular rooms oriented east–west. To the north is a large hall, originally barrel-vaulted, that was once accessible through a doorway in the middle of the north wall—bricked in at some point in antiquity—and through another door in the south half of the west wall. Only from this room could one enter the church to the south, via two doors—the larger one located in the middle of the wall separating the two rooms and the narrower one at the west end of the same wall. Gillian Bowen found ample evidence that the room had not been built originally as a church, and that its conversion into an ecclesiastical building had entailed several alterations.24 The most significant of these was the addition of a raised, tripartite sanctuary set against the east wall, with a central apse—delimited by two pilasters and richly decorated—and two side-rooms. According to ceramic and numismatic evidence, the Small East Church, which shares several significant similarities with the church of ʿAin el-Gedida, as we shall see, was in use during the first half of the fourth century.25↩#p225
Bowen argues that the Small East Church should be regarded as a domus ecclesiae, a domestic building converted by a Christian congregation in need of a place in which to gather and celebrate the Eucharist.26 In that case it would slightly predate the construction of the Large East Church, which was, by contrast, the result of careful planning and possibly served a rapidly growing Christian population. The Large East Church is a rectangular structure built against the complex's southeast enclosure wall (Fig. 104). Measuring 20 × 17 m and oriented eastward, it is in a fairly good state of preservation, and some of its walls stand to a considerable height. Access was originally via three doorways set in the western wall and connecting the church with the larger ecclesiastical complex.↩#p226
The main body of the church is divided by two rows of six columns into a central nave and two side-aisles. The bases of the two columns at the west end of both colonnades show that they originally had a trefoil shape. A west return aisle was created by adding an extra column between the north and south colonnades, against which is a stepped mud-brick platform. To the east, a transverse aisle with four columns completes the ambulatory, which runs along the four walls of the church and surrounds a central area paved with flagstones. There are benches against the north, west, and south walls. The north and south spaces between the columns were originally sealed with wooden screens, as was the northwest intercolumniation of the return aisle. The sanctuary consists of a raised apse centrally placed against the east wall, framed by two engaged pilasters and with a floor of triangular mud bricks. A rectangular bema (a platform used by clergy), accessed by two steps at its north and south ends, is situated in front of the apse and protrudes into the transverse aisle. The apse is flanked by two small service rooms, entered from the transverse aisle; the south room is also directly connected with the apse via two steps.↩#p227
Through the south aisle can be reached a set of four rooms to the south of the church. The function of three of these is unknown, but a staircase and two ovens were found in the westernmost room, which according to the excavators likely served as a kitchen for baking the bread used in the liturgy.27↩#p228
As will be evident, the Large East Church shares substantial typological similarities with Trimithis' church, described in §5.1. On the basis of the coins and potsherds found there, it is datable to the first half of the fourth century. The Trimithis church, as part of the redevelopment of Area 2, can be dated to the latter years of that half-century or the 360s.↩#p229
Further evidence for the intensive growth of Christianity in the oasis was gathered a few years ago with the discovery of the ecclesiastical complex of ʿAin el-Gedida, a site located a few km to the northwest of Kellis. This complex was identified and excavated, under my direction, between 2006 and 2008.28 Quite centrally situated on a large mound, it is easily accessible. Indeed, an extensive network of streets, comprising a main north–south axis plus smaller passageways and alleys, must have been quite effective in ancient times in shaping the movement of people around the hill and channeling them toward the church.↩#p230
Figure 105. Plan of the church at ʿAin el-Gedida.Figure 106. ʿAin el-Gedida, the church and the assembly hall (view to the west).
This complex has two large rectangular rooms and another five interconnected spaces to the north and northwest and is surrounded by several other structures, including a large kitchen, a domestic dump, a ceramic workshop, and passageways (Fig. 105). The church itself occupies the southernmost space. It is oriented to the east and measures about 11.35 × 3.65 m, and was barrel-vaulted. Mud-brick benches line the north, west, and south walls of the room; a semicircular apse with an L-shaped pastophorion is set against the east wall (Fig. 106). The church was originally accessible from the north through two doorways, one near the northwest corner, and via a large central passageway along the north side. These connected the nave with a vaulted rectangular space, likely functioning as an assembly hall, with benches built against the north, east, and south walls.↩#p231
Figure 107. The podium inside the church at ʿAin el-Gedida, with a bricked-in wall behind it (view to the north).
Against the south side of the central opening, inside the nave, is a stepped mud-brick podium accessible only from inside the church, that once enabled people in both rooms to see and hear the person—possibly a priest or a reader—standing on it. At some point, possibly due to a partial change of function of the room immediately to the north of the church, this passageway was closed, hindering access to the podium (Fig. 107). The assembly hall opens to the north onto a smaller and once barrel-vaulted rectangular anteroom. This space was used, at least in its latest occupation phase, for preparing food, as indicated by the discovery of a hearth and the imprints of several vessels on a platform and at floor level. A series of graffiti was found on the west and north walls, including two inscriptions—one in Greek and the other in Coptic—and some drawings of boats and a bird, among other subjects.↩#p232
One accesses the anteroom/kitchen from a long, roofless corridor running east–west to the north of the hall. This space ends on the east with a doorway that is the only entrance into the church complex from the outside. A narrow, vaulted passageway connects the anteroom with another, fairly large, space to the north. It was originally barrel-vaulted, and its outer walls form these buildings' northwest boundary. Recessed into the south wall of this room, by its southeast corner, is a mud-brick feature that might have been used as a cupboard. It is possible that the room was a storage space for the anteroom/kitchen to the south.↩#p233
As mentioned earlier, the church of ʿAin el-Gedida shares a considerable number of similarities with the Small East Church at Kellis—in particular, similar dimensions and an almost identical layout. We noticed differences, however, especially some dating from the later phases of alteration of the ʿAin el-Gedida church.29 Furthermore, judging by the available ceramic, numismatic, and documentary evidence, the early-fourth-century dating of the Small East Church matches that of the ʿAin el-Gedida complex.30↩#p234
There was ample evidence available to prove the existence of different construction phases within the complex of ʿAin el-Gedida. This includes the south wall of the church, irregularly laid out and clearly built at different times, and the whole semicircular apse with its L-shaped pastophorion, added later. There are also clear traces of foundation walls below floor level, belonging either to previous buildings or to earlier building phases of the church and the adjacent hall. Studying the relationships between each wall and those neighboring it revealed how the buildings had been significantly altered in the west and north sections, some rooms having been substantially changed and new ones added.↩#p235
The seating capacity of the church and of the assembly hall was more than seventy, or over eighty including the anteroom. This estimate does not take into account, of course, any individuals who might have stood in the church or the hall or those who accessed the complex merely to carry out more practical tasks in the other rooms. Considering not only the small-to-average size of the church and of the entire complex, but also the seemingly limited extent of the settlement—especially compared to nearby sites such as Kellis—this adds up to a considerable number of people.↩#p236
An intriguing question concerns the nature of the rooms that were altered to build the latest stage of the complex: namely, did they function as a church before their westward expansion and the addition of an apse to the sanctuary? In the first centuries of Christianity, groups gathered to worship in domestic buildings—the basilica form would be adopted around the time of Constantine. Indeed, the existence of such domus ecclesiae is attested archaeologically in the ancient world, including in Dakhla, as mentioned above in relation to the Small East Church at Kellis. Therefore, even though there is no evidence to support it, we cannot rule out the possibility that religious ceremonies were carried out in these spaces at ʿAin el-Gedida before their enlargement and/or the construction of the apse.↩#p237
In addition to the rich archaeological evidence from Kellis, ʿAin el-Gedida and Trimithis (as discussed in §5.1), other sites in the Dakhla Oasis testify to the existence of Christian communities throughout Late Antiquity. The 1977–1987 Dakhleh Oasis Project survey listed two churches whose substantial remains are still visible above ground level. One is at the site of Deir Abu Matta, about eight km southeast of ancient Trimithis. The area of visible archaeological remains, which had already been noted by H. E. Winlock in 1908,31 is fairly limited and is currently surrounded by patches of desert, cultivated fields, and a paved road. In 1980, DOP members surveyed the mound on which the church stands and dug some test trenches inside the basilica.32 An archaeological project involving the investigation of the church and its adjacent structures began in 2007, under the direction of Gillian Bowen.33↩#p238
Figure 108. Deir Abu Matta, external walls of the church (view to the west).
The church is the site's largest visible building. Oriented eastward, it is rectangular and measures 24 × 10.35 m. The mud-brick walls are nowadays the most impressive features, being over one meter thick and still standing several meters above ground level (Fig. 108). They were built in sections and originally supported a beamed roof, as suggested by the holes piercing the south wall. The church's interior is in very poor condition nowadays, but its layout is roughly discernible. According to a plan drawn by Peter Grossmann, the church was originally divided by two rows of six square pillars into a nave and two side-aisles, with an additional L-shaped pillar at the west end.34 A return aisle along the west side joined the two colonnades by means of two more square pillars, forming an ambulatory around the central nave. A bench runs along the north section of the west wall. Another bench, no longer visible, was once located against the south wall. Evidence of a fairly narrow door—possibly a secondary entrance—was detected toward the west end of the north wall. The sanctuary, in the shape of a triconch, is to be found inside the church along its east wall; it was once framed by two engaged pillars. To the sides of the lateral conches, against the northeast and southeast corners of the building, are L-shaped side-rooms (pastophoria).↩#p239
The test trenches dug along the north wall revealed numerous early Christian burials, although some of them, at least those excavated in more recent years, were found disturbed.↩#p240
Considerable evidence of different construction phases in the area of the church has been documented since 2008. Architectural features predating the basilica are visible to the north of it, possibly extending further south. Other walls—later than the church, according to the excavator—were found to the north and west. A wide, tower-like building was also excavated to the west of the basilica. It is possible that at least some of the structures found near the church were associated with a small monastic establishment, whose existence in Late Antiquity is suggested by the modern name of the site.↩#p241
According to the DOP team's survey, fifth-century coins, along with ceramics datable from the fifth to seventh century, were collected during the survey and the test excavation. The finds collected during the 2008–2009 excavations, which include coins, ceramics, and an ostrakon, were largely of the fourth/fifth century CE, although some of the ceramics may be sixth-century. This means it is possible that the church of Deir Abu Matta was built significantly earlier than previously thought.↩#p242
Figure 109. Deir el-Molouk, church plan (after Grossmann 2002, plan 181).
The other church mentioned in the survey is that of Deir el-Molouk, a few km northwest of Masara.35 It was built of mud bricks and had a cruciform shape (Fig. 109). A domed roof was at its center, and there was an entrance, the surveyors reported, along the poorly preserved north wall. The church was divided into nine square spaces by four central cruciform pillars. Against the east wall were three apses with small niches, and three more conches were located at the centers of the north, west, and south walls, emphasizing the building's cruciform shape. To the south of the church, and built against it, was a square room ending with a semicircular apse along its east side. This space, unconnected with the main building, was accessible through a narrow room outside the south apse. The south room, bearing traces of painted plaster, was possibly built shortly after the church itself and functioned as part of the same structure. The complex, measuring about 17.5 × 15.5 m (including both the church and the south room), shows evidence of architectural alterations, which include the addition of later walls near the southwest corner of the church and the entrance to the south room. There is insufficient evidence to determine whether the church originally stood in isolation or among other buildings. The dating of the church, too, is uncertain. However, the data gathered from the test trenching, and from examining the church's layout, suggest a significantly later construction period for the Deir el-Molouk church than for the others excavated or surveyed in Dakhla.↩#p243
In 2009, Kamel Bayoumi of Dakhla's Islamic and Coptic Inspectorate found another church, this one at the site of ʿAin es-Sabil, less than 1.5 km southwest of the archaeological site of Kellis. The church, which is oriented to the east, shows a basilical plan with a central nave and two side-aisles, defined by two rows of four mud-brick columns each, and a west return aisle. The entire church complex measures roughly 17 × 26 m, and the church proper about 9.4 × 10.6. The apse is rectangular and framed by two semi-columns. An arched niche is set into the sanctuary's north and south walls, which open onto side pastophoria through small doorways. The church at ʿAin es-Sabil has not yet been published and no archaeological data are available. From the typological viewpoint, the building appears to share some similarities with the church at Trimithis and the Large East Church at Kellis, both datable to the mid-fourth century. A similar chronology for the church of ʿAin es-Sabil is further supported by Rodney Ast and Roger Bagnall's analysis of documentary evidence that was found by the excavators in a complex adjoining the church.↩#p244
On the whole the archaeological and documentary evidence for the growth and flourishing of Christianity in Dakhla has, since the end of the twentieth century, become quite substantial. Excavations at Kellis, ʿAin el-Gedida, and Trimithis have revealed how Dakhla embraced Christianity, together with its artistic and architectural manifestations, from an early stage—since at least the early fourth century CE. They also tell us that Christianity was by then not just a phenomenon of the larger cities and towns, such as Trimithis, Mut, and Kellis, but a universal reality that encompassed the entire Egyptian society, including the small rural villages and even the most isolated hamlets of the oasis.↩#p245
