third point concerns local specificity. On the scale of an urban site of this size, in the heart of an agricultural setting with many resources, it is possible to observe a certain degree of autonomy in the production of common ceramics connected both to domestic use and to the processing of foodstuffs. A salient example is the Brittle Wares, no doubt produced from kaolinite-type clays, which were produced either at Kellis or in its immediate surroundings. These constitute a distinctive and original group, finds of which are largely concentrated in the western part of Dakhla. Discovering the origins of the regional ceramic productions of the Great Oasis and establishing their connection to specific urban settlements remains a vital field of research, going well beyond the scope of monographic studies of this type. But such monographs are the indispensable foundation for such future investigations.
Pascale Ballet
Université de Paris Nanterre, UMR 7041, ArScAn – ESPRI
(translated by Roger Bagnall)
INTRODUCTION
Amheida is located in the western part of the Dakhla oasis, 3.5 km south of the medieval town of El-Qasr (Figure 1).1 Known in Hellenistic and Roman times as Trimithis, Amheida became a polis by 304 CE, as evidenced by a papyrus from Ismant el-Kharab/Kellis.2 Trimithis in fact served as the major administrative center of the western part of the oasis for the entire fourth century CE.3
The site dates back to as early as the Old Kingdom and was occupied as late as the late fourth century CE. The extension of the visible ruins is approximately 2.5 km north–south and 1 km east–west.4 Two brief exploratory seasons at the site (Amheida: Dakhleh Oasis Project site no. 33/390-L9-1) took place in 1979–1980 by the Dakhleh Oasis Project (DOP), directed by Anthony J. Mills.5 Among other features, the survey revealed several kilns for small-scale production of ceramics dated to the Roman period (Area 1), which have been cleaned and mapped.6 The systematic archaeological excavations at Amheida, which started in 2004,7 have been carried out by an international team under the sponsorship of Columbia University, and since the 2009 season by New York University’s Institute for the Study of the Ancient World. The project is directed by Roger S. Bagnall (ISAW); the archaeological director is Paola Davoli (University of Salento, Lecce); and the pottery study has been supervised by Pascale Ballet (University of Paris Nanterre).
Archaeologists have so far divided the site into 11 areas (Figure 2),8 four of which are under excavation. Area 1 is characterized both by private houses (so far, only House B2 was excavated in its entirety) and workshops that developed around one of the main streets (S1) of the settlement, and it presents highly diversified pottery, dating from the early Empire to the beginning of Late Antiquity.9 Area 2 is a residential area at the base of the hill that dominates the urban area, characterized by large private decorated houses and public buildings, in which ceramic materials are mainly dated to the fourth century CE. So far, four buildings have been excavated in Area 2: House B1 (Area 2.1); a school of rhetoric and Greek (B5 – Area 2.1); a Late Roman public bath (thermae) (B6 – Area 2.2); and a church (B7 – Area 2.3).10 Area 4, on top of the central hill of the site, shows remains of successive temples with different construction phases dating from the New Kingdom to the Roman Period and a deep stratigraphy that testifies to earlier occupations dating back to the Old Kingdom.11 Area 8 is a densely settled quarter at the northern limits of the city, in which the investigation of a domestic complex (B10) dated to the late third and beginning of the fourth century was undertaken in 2015.12
Most extant written remains from Amheida take the form of inscribed Greek ostraca, although Amheida also preserves hieroglyphic, hieratic, and demotic evidence.13 Among the Greek texts, a group of fourth century documents found in Area 2.1 stands out. It concerns a number of individuals, some of whom were members of the local city council. One of these men, Serenos, was a member of the municipal elite and a Trimithis city councillor, and is associated with the domestic structure designated as House B1 (Figure 3).14 This house, of which the walls are preserved as high as 2.5 m, originally had a square plan (15 x 15 m) and consisted of at least 11 rooms, with a staircase to reach a second floor or the roof-terrace. The two entrances are located on the west street (S3) and the east street (S2), and give access to vestibules (Rooms 7 and 12, respectively), each directly connected to a central space (Room 2), around which all the other rooms are arranged. The main room of the house was a large space (Room 1), entirely plastered and painted with subjects of Greek mythology and scenes depicting the family that owned the house. It is also the only room with a domed roof. Geometric and floral decorative motifs decorated the walls of three other rooms (11, 13, 14), painted with bright colors that are still well preserved.15 At a certain point after 350, the house was enlarged, incorporating the area occupied by a school building (B5) just north of B1. At the time of Serenos’ acquisition or incorporation of B5, some rooms were demolished and others changed their original purpose: e.g., Rooms 9 and 10 were transformed into a courtyard and Room 15 was converted into a storehouse. The house was probably abandoned around or soon after 365 CE.
The excavations below the occupation layers of B1, S2, and S3 also revealed the remains of previous features from the baths, such as a round laconicum (below courtyard 9) and a latrina (below Room 10), which were demolished down to floor level and partly covered by dumped material when the school building (B5) and House B1 were built (Figure 4).16 The change in use of the area occurred, according to ostraca, coins, and other artifacts, not long after the first quarter of the fourth century CE.17
This volume aims to provide a comprehensive study and classification of the ceramic assemblages recovered in House B1 and the two streets (S2 and S3) adjacent to it, as well as in the foundation deposits below them.18 Most of the ceramic fragments found in the area occupied by these features (Area 2.1) were produced in the Dakhla oasis, and more generally in the Great Oasis. Lesser in quantity are the imports from the Nile valley and the Mediterranean. Upon a detailed examination of the pottery, it became evident that the vast majority found above floors in B1, S2, and S3 can be securely dated to the second half of the fourth century CE, thus belonging to the last occupation phase of the area. In contrast, the ceramics accumulated as dumped material below the floors are pre-fourth century in date, and came from dumps in which different kinds of materials were mixed, with objects and sherds dating from predynastic times to the early fourth century CE.19
The main categories of vessels encountered during the analysis of the assemblages in the field are here arranged according to functional and typological criteria.20 Each type is dated by its occurrence in the well-dated stratigraphic contexts, thus making the presence of new types or the disappearance of existing ones readily noticeable. Therefore, comparison with the vessels found in the contemporaneous settlements in Dakhla, and Egypt more broadly, also helped to set the general chronological boundaries for each type, from the end of the first century BCE up to the fourth century CE. Ceramological analysis, along with the evidence of stratigraphy, ostraca, coins, and other objects, allows us to define with more precision the dating range of the various phases of construction and restoration that occurred in House B1.21
1. The medieval town of El-Qasr, dating to the Fatimid period but originating in late antiquity, reached its highest stage of development in the 18th–early 19th century, see Bagnall and Ruffini 2004: 143; Bagnall et al. 2015.
2. P.Kell. I G. 49.1–2 (ή Τριμιθιτών πόλις): Worp 1995. See also Wagner 1987: 190–2.
3. Bagnall and Ruffini 2004: 144; O.Trim. 1.1–2.
4. In general, the site is well preserved, although the strong wind erosion and high humidity (due to its proximity to cultivated areas and spring water) are a threat to the integrity of the buildings made of unbaked bricks. The overall extension of the ruins is still to be determined, obscured by mobile
