We would like to express our deepest gratitude to Mr Pang-Chi Wang, director of ECO Taiwan Design, for his research sponsorship.
1The joint French-Austrian mission of the IFAO, the University of Strasbourg/UMR 7044 Archimède and the Austrian Archaeological Institute (ÖAI) worked at Kūm al-Ḫarāba al-Kabīr Ğirza, in the ruins of ancient Philadelphia, from 8 May to 4 June 2023.
2Our present and fifth campaign at Philadelphia was programmed for a transition of our investigation focus from the southern zone, i.e. south of Darb al-Ǧirza, to the northern zone (Fig. 1).
Fig. 1. From south to north, locations of S38, S16 and S17, in which we worked during this PhF2305 season (R.-L. Chang, based on images available on Google Earth).
© IFAO. 17124_2023_NDMCN_001
- 1 Chang 2019; Chang 2020.
3Contrary to the southern zone, which before our mission, had been briefly excavated by a German mission during one single winter season in 1908/9, the northern zone had never been scientifically explored prior to our research in 2015 in Sector 12 and in 2018 in Sectors 13, 15, 39 and 40.1
- 2 Marchand, Nannucci, Chang 2018; Chang et al. 2020; Chang et al. 2023.
4In the southern zone, through our 2nd season in 2018, 3rd season in 2019 and 4th season in 2022, we have been focusing on the urban planning, by exploring a quarter of pottery production and its surroundings.2 Besides the discovery of previously unsuspected structures of kilns, with the considerable number of ceramics they contained (S37 and S41), our mission also brought to light extensive ancient rubbish dumps in Sector 38, the investigation of which we completed this year. We analysed the stratification of this sector dating essentially from the 2nd century BCE, as well as its upstanding remains, fortunately still intact.
5At the same time, in the northern zone, work was started in Sector 17, an area characterised by a small kūm. This kūm actually occupies the highest place of the whole ancient settlement remains, and contains a cluster of upstanding structures. It offers both pertinent specimens of the town architecture and further evidence for the urban evolution with much better preserved elevations in comparison to the vestiges in the southern zone. Among the phases of occupation that we discovered, the most ancient one dated from the second half of the 3rd century BCE.
6In addition, research on the ancient water management was pursued by investigating a shaft opening on the ground in Sector 16. The upper part of this shaft, which looked like a well, had been exposed by recent plunderers. It turned out to be part of an underground cistern cut into the bedrock. This cistern also comprised a staircase, an arched entrance at the foot of the stairs opening into the shaft, as well as a system of tunnels used as reservoirs. Our preliminary ceramological dating attributed the whole structure to the Ptolemaic period.
7For this season, our work in S38 was aimed at completing the excavation of the extended trench (EW 7 × NS 28 m) already investigated during two campaigns in 2019 and 2022. We finished exploring the primary ancient dump and the secondary one, that had previously been identified in a trench of EW 7 × NS 13.5 m. After that, we successfully reached undisturbed stratified contexts, in this part of Philadelphia settlement almost entirely destroyed by the sebbakhin. The preserved stratification measured as high as 1.5 meters, and the earliest layers presumably predate the mid-2nd century BCE, according to our study of ceramic remains.
8The earliest phase uncovered consisted of a foundation wall made of one single course of tafla bricks, running NS for about 20 m. The foundation trench was shallow, and cut directly into the virgin soil. No relevant floor remained. No dating material was found in the trench fill. This wall was partially dismantled prior to constructions of the next phase.
9Immediately above this initial phase and directly upon the virgin soil, a building in mud bricks was constructed. This building had been partially revealed in the northern and southern parts of S38, but in a poor condition of preservation due to sebbakhin’s destruction. This season, we discovered the better-kept part of it: a courtyard divided into compartments by mud-brick walls, preserved as high as 50 cm and abutting the thick perimeter walls on the eastern and southern sides (Fig. 2 and the upper part of Fig. 3).
Fig. 2. In S38 (PhF2305), the plan of the remains of a building belonging to the second phase of occupation, showing a compartmentalised space, with underlying foundation walls belonging to the initial phase of occupation (S. Nannucci).
© IFAO. 17124_2023_NDMCN_002
Fig. 3. The upper part shows a compartmentalised courtyard in S38 (PhF2305), dating from before the mid-2nd century BCE, viewed from the south-west. The lower part shows, at the same location, layers of abandonment and destruction after the second phase of occupation, dating from the mid-2nd century BCE, viewed from the north (S. Nannucci).
© IFAO. 17124_2023_NDMPF_001
10These compartments may have been designed for storing crops. Demotic and Greek ostraca discovered mention wheat and barley (see below, § “Work on Artefacts”).
11This building was subsequently demolished. The most thorough removal of its mud bricks happened to the perimeter walls. The ceramics pertaining to this phase date from the second half of the 2nd century BCE (lower part of Fig. 3). Afterwards, the whole area, being levelled out, came to accommodate a dump consisting essentially of pottery sherds, among which many complete specimens were found. According to the ceramic evidence, this dump also dates from the second half of the 2nd century BCE. It was also characterised by silty sediment mixed up with ashes, and the botanic remains as well as other ecofacts were only rarely spotted therein.
12At a certain time after the formation of this ceramic dump, a mud-brick mural structure, oriented somewhat differently in comparison to the vestiges of other buildings in this part of the settlement, was built right upon the irregular surface of the accumulated potsherds. Its brickwork was heterogeneous, pointing to two construction phases. Even after the elevation of this structure, the dump continued to be used as such, and gradually buried at least the lower part of the upstanding structure. This observation suggests that the mural structure was erected in order to retain and stabilise the dump. But we cannot rule out the possibility that it was in fact the remains of a more extensive building (upper part of Fig. 4).
Fig. 4. The upper part shows, in S38 (PhF2305), a mud-brick mural structure built directly upon the primary dump dating from the second half of the 2nd century BCE, viewed from the south. The lower part shows the situation of S38 (PhF2305) at the beginning of this campaign, with the secondary dump still in place, viewed from the north (S. Nannucci).
© IFAO. 17124_2023_NDMPF_002
13During the first decades of the 20th century, the stratification of S38 was seriously destroyed by sebbakhin. Their aggression often reached the virgin soil. However, the superimposed mural structure, along with the thick primary ceramic dump under and around it, probably contributed to the halt of sebbakhin’s attack. In the end, a part of the original stratification of S38 was thus kept intact for the research we have accomplished. As for the secondary dump around the primary one, it was a combination of (1) what the sebbakhin managed to dig out of the primary dump and (2) potsherds that the sebbakhin threw in from neighbouring buildings that they destroyed and which could also be dated to the second half of the 2nd century (lower part of Fig. 4).
14Our work on the kūm in S17 combined surface cleaning and excavation. After the removal of the remains resulting from the sebbakhin’s destruction, three occupation phases were revealed over a surface measuring EW 16 × NS 17 m (Fig. 5).
Fig. 5. The excavated area of S17 (PhF2305), viewed from the south-west; the numbers of the stratification units of the 1st phase are in white, those of the 2nd phase in magenta and those of the 3rd phase in blue (S. Awad Mohamed).
© IFAO. 17124_2023_NDMPF_003
15The earliest phase, on the eastern side of our trench, consisted of the remains – from an identical building (S17.B01.I) – of two walls (US 11 and 12) and their associated floors (US 18 and US 40). The foundation trenches were cut directly into the virgin soil of cemented coarse white sand. The courtyard to the north of U12 and to the east of U11 was divided into compartments, presumably used as storage spaces. The dividing mud-brick walls still bore plaster in places (Fig. 6). On the western edge of our trench, we also uncovered two stretches of walls (U36 and U37) belonging to the earliest phase.
Fig. 6. The stratification units of the first phase of occupation in S17 (PhF2305), with the compartments of a courtyard, viewed from the south (S. Awad Mohamed).
© IFAO. 17124_2023_NDMPF_004
16The second phase of occupation was attested by two buildings. The building on the eastern side (S17.B01.II) was constructed after the structures of the first phase had been demolished. The foundations of the perimeter walls (US 6, 9, 22 and 23) were cut into the virgin soil. The associated floors (US 13 and 16) were much disturbed, and only few remains were found within the enclosed space. The second building (S17.B02.II) was located on the western edge of the trench and under the destruction remains left probably by sebbakhin. For this season, there was only enough time to uncover its eastern wall remains (US 33, 34 and 35), where a doorway installed on the N-S wall (US 35) also cut into an adjacent N-S wall (US 36) belonging to the first phase. Here, no evidence for an abandonment phase between the first and second phases could be found.
17The third phase of occupation was represented by two other buildings. The first one (S17.B01.III), situated near the summit of the kūm of S17, was revealed north and west of S17.B01.II. This part of the building, which stretched northwards, showed a series of compartments formed by eight walls (US 04, 05, 07, 24, 25, 26, 27 and 28), suggesting some kind of storage function. The second building (S17.B02.III) consisted of four walls supposedly forming a square (US 29, 30, 31 and 32). However, the southern part was completely destroyed by sebbakhin.
18In fact, a huge portion of the walls and floors from the second and third phases had been removed by sebbakhin. Neither their functions nor their dates could be properly evaluated. Further excavation and research investment are necessary to uncover more of these two phases and understand them better. Nevertheless, the first phase of occupation can already be confidently dated by the preliminary ceramological study to the second half of the 3rd century BCE. Amongst the destruction remains inside the aforementioned courtyard compartments of S17.B01.I, we found a complete small jar which should date from the second half of the 3rd century BCE (see below, § “Work on Artefacts”). More decisively, in the trench fill of wall U12, we found a saucer-shaped lamp ceramologically dated to the same period.
19After the removal of a recent dump c. 3 m north of the exposed shaft opening, as well as the topsoil, the highest steps of a staircase cut directly into the bedrock were brought to light. A trench measuring EW 6.44 × NS 9.78 m was thus installed. In a southward descent, it allowed the discovery of an arched and elongated doorway giving access to the shaft. The lowest part of the shaft further opened westwards and northwards into an identical gallery, which ran in a somewhat quadrangular way and closed on itself. The whole structure was an underground cistern (Fig. 7).
Fig. 7. In S16 (PhF2305), the shaft opening and the stairs of an underground cistern, viewed from the north (S. Awad Mohamed).
© IFAO. 17124_2023_NDMPF_005
20Along the stairs and down to the bottom of the shaft, two backfill layers could be distinguished. The first layer was left by recent robbers, and consisted essentially of yellow sand mixed with potsherds, fragments of stonework and modern materials such as textiles and plastic bags. It stretched through the aforementioned doorway into the shaft and the gallery, where it became damp due to the rain infiltration. Under this layer, there is another backfill composed of grey and yellow clay mixed with sand and numerous potsherds, as well as stonework pieces. The formation of this layer was probably due to treasure hunters or sebbakhin of former phase of occupation. It could also be associated with an ancient abandonment phase. We have kept this layer in one fourth of the underground gallery for our next campaign to further investigate. In other parts, the cistern has been emptied to its bottom, for proper photogrammetry and a better understanding of its structure.
21The whole preserved cistern was cut into the bedrock of cemented coarse-grained sand. It consisted of a shaft (top ∅ 0.71 m / bottom ∅ 0.97 m / depth 5.8 m) presumably for water inpouring, and a staircase of 11 steps (length c. 3.3 m / width c. 0.86 m) for descending and drawing water, as well as for cleaning and maintenance. After reaching the doorway (height 1.9 m), the staircase was extended by three more downward steps leading into the shaft. Below, on the lower wall surfaces, there were indentations which allowed a firm footing and a safe descent to the bottom. The gallery was composed of four interconnected tunnels (lengths: T01 EW 4.10, T02 NS 3.00, T03 NS 2.35, T04 EW 4.00 m). Tunnels 1 and 3 stretched out from the shaft bottom, westwards and northwards respectively. Tunnel 2 stretched northwards from the middle section of Tunnel 1. At their northern ends, Tunnels 2 and 3 joined Tunnel 4, which was oriented EW. The width of a tunnel varied between 0.5 and 1.3 m. Only Tunnel 1 was constructed with a flat ceiling, while Tunnels 2-4 were all barrel vaulted (Fig. 8).
Fig. 8. A view from the 3D modelling of the underground cistern in S16 (PhF2305) (M. Kačičnik).
© IFAO. 17124_2023_NDMCN_003
22The bottoms of all the four tunnels gradually sloped down to the shaft. Both the shaft and the gallery had originally been entirely plastered with hydraulic mortar, including their bottoms. In Tunnel 1, there was evidence for two layers of plaster in the ceiling. In Tunnels 2 and 4, as well as in the shaft, there were painted lines or carvings, presumably pertaining to the measurement of water level or volume. The initial ceramic survey dated the potsherds from all the layers to the Ptolemaic period exclusively. Our next campaign shall forward our ceramological study and further research the archaeological layers that we have decided to save for closer scrutiny.
23Compared to our 4th campaign in 2022, this time an even greater number of artefacts were recovered in S38, from the dumps and the underlying stratification. With supplementary finds from S17 and S16, in total 113 objects have been transferred to and registered at the Kum Ushim Storehouse, under 106 numbers. They are 43 ostraca and dipinti in Greek or Demotic, 33 ceramic vessels and accessories (including a small vase in the form of a Nubian head ; Fig. 9), 12 lamps, nine terracottas and various objects, nine potters’ stamps, one inscription on a vase, one wooden saucer and five coins.
Fig. 9. From S38 (PhF2305), a small vase in the form of a Nubian head discovered in the primary dump (M. Kačičnik).
© IFAO. nu_2023_08953
24On the other hand, a total of 116 objects have been deposited at the storehouse for future study, including 82 Greek or Demotic ostraca and dipinti (or possibly drawings), 26 ceramic items, terracottas and miscellaneous artefacts, as well as eight coins.
25There are plenty of remarkable objects. Here are three examples. A Greek dipinto from the secondary dump of S38 mentions a 51st year. This year must belong to the reign of Ptolemy VIII, that is 120/119 BCE, which is in accordance with the ceramological dating of the dump in question. In the abandonment/destruction layers of the Ptolemaic courtyard revealed in S38, we found a Greek ostracon (Fig. 10) recording 62 artabas of wheat for the account of a person named Petemin, as well as 50 artabas under the name of someone called Horos son of Keleesis. The total amounts to 112 artabas, i.e. about 4346.5 liters of wheat (see below the edition of this account of wheat).
Fig. 10. From S38 (PhF2305), a Greek ostracon dating from no later than the second half of the 2nd century BCE and recording considerable amounts of wheat, discovered in the abandonment/destruction layers associated with the second phase of occupation (M. Kačičnik).
© IFAO. nu_2023_09367
26This ostracon’s content not only supports the hypothesis that the courtyard was used for crops storage, but its handwriting style also suggests that despite the ceramological dating, it would not be too off the mark to attribute the text to the second half of the 3rd century BCE. In the northern zone, from the destruction remains of the first phase of occupation of S17, a small jar preserved in intactness can be dated to the second half of the 3rd century BCE. It presents a classic form of the Egyptian pottery repertory, and was used to store yellow and red ochre stones.
27The retrospective study of the ceramic material from the S38 secondary dump excavated during our 4th season in 2022 has been completed. Although it could no longer be put back into their occupation context which had been destroyed by sebbakhin, this material provides an informative inventory of household ceramic assemblage from the 2nd century BCE in Philadelphia. As for S17, all the ceramic remains from the first phase of occupation seem to date from the second half of the 3rd century BCE, while the layers of remains close to the surface (US 01, 02, 08 and 10), due to the sebbakhin, are chronologically heterogeneous, with dates varying between the 3rd century BCE and the 2nd century CE.
28Our restoration work encompassed all the types of the discovered artefacts. For example, out of several potsherds, a thin-walled goblet from the secondary dump of S38 and dated to the second half of the 2nd century BCE, has been almost entirely resuscitated. It is characteristically Egyptian and made from local chalky clay. It carries drawings and some Greek writings. A headless statuette in unbaked clay of a female figure, found in the destruction remains of the first phase of occupation dated possibly to the second half of 3rd century of S17, is also particularly worth mentioning. It has been duly consolidated in order to well preserve the delicate and evanescent curves of the robe and belt that the figure, reminiscent somehow of Tanagra figurines, is wearing. Both of these objects have been deposited at the Kum Ushim Storehouse to be further studied.
29To conclude, we give a preliminary edition of the Greek ostracon particularly described above, which by its person names, enticingly allows a swift glimpse into an early period of life at Philadelphia, perhaps even vaguely evoking the times depicted by the Zenon archive (see below note 4).
- Inventory number: Kūm Ūshim Register no 100 = PhF2305.038.240.001 (Fig. 10).
- W. 12.5 × H. 6 cm.
- On convex side.
- Philadelphia, Fayyum.
- Earlier than the second half of the 2nd century BCE according to the context of ceramic remains.
30Two considerable amounts of wheat are booked under the names of two persons, and summed up at the end of this short note. The names display an onomastic context pertaining to Egyptian cults. The account may be about additions to a storage, but this cannot yet be confirmed. The scribe could be a bookkeeper recording reception and even outflow of grains. The first two lines are written in larger letters than the third and last line, but the hand, writing brush and ink are the same. Perhaps the ostracon, utilised maybe as a tag for containers, was originally only intended for one entry.
Πετεμιν
πυ(ροῦ) ξβ
Ὧρος Κελεήσιος πυ(ροῦ) ν (γίνονται) ριβ
‘Petemin, of wheat, 62 (artabas); Hōros (son) of Keleēsis, of wheat, 50 (artabas), sum total 112 (artabas).’
- 3 Lüddeckens 1980, p. 310.
- 4 On the archive of Zenon son of Agreophon, dating from 263-229 BCE, see Clarysse, Vandorpe, Verreth(...)
1. Πετεμιν. There is no palaeographical clue to an abbreviated form, e.g. Πετέμιν(ις). The present name, without a Greek inflectional ending and thus no accentuation, is a transliteration of an Egyptian name meaning ‘he whom Min begot; TM NamVar 58915 (www.trismegistos.org/namvar/58915)’.3 The indeclinable spelling Πετεμιν is hitherto attested in the papyrological documentation four times, twice in ostraca from the Roman Eastern Desert and twice in the Ptolemaic Zenon archive: P.Cair.Zen. 4.59745, 85 (255-254 BCE); PSI 4.368, 1 (250-249 BCE).4
- 5 Lüddeckens 1980, p. 1016; Clarysse 2018, pp. 201–202. For prosopographical information about the p (...)
2. Κελεήσιος. Greek version of an Egyptian name meaning ‘the one who takes refuge by Isis’; see TM Nam 399 (www.trismegistos.org/name/399).5
- 6 For an alternative reading of the present siglum as πυρῶν, see Clarysse 2014, pp. 103–105.
- 7 For a form of the Roman siglum for πυρός, see Chang 2014, p. 67.
2, 3. πυ(ροῦ). This can be understood as πυ(ροῦ ἀρτάβαι).6 The monogram consists of the first two letters of πυρός, stylised and superimposed on each other. It is reminiscent and possibly the forerunner of the Roman siglum for the same word.7 It is used in texts from the Zenon archive, e.g. P.Cair.Zen. 4.59745 (already cited above l. 1 n.), l. 6, 9, 10, 12, 33, 40, 52, 57, 70, 72, 75, 81, 95, 98, 99; PSI 6.551 vo, 1 (mid-3rd century BCE).