Top Banner
23 e Topography of New Kingdom Avaris and Per-Ramesses Manfred Bietak and Irene Forstner-MÜller* § 1 Introduction Kenneth A. Kitchen, who has the most intimate knowledge of Ramesside inscriptions, has repeatedly drawn pictures and accounts of the major cities of Ramesses II, including, of course, Per-Ramesses. Into these imaginative reconstructions, as Kitchen modestly calls them, a most valuable amount of informa- tion drawn from the texts has been absorbed which has not lost its value over time, despite the long standing research of the Austrian Archaeological Institute and the Pelizäus-Museum Hildesheim in the region of Per-Ramesses in the last decades. We hope with this account, which is based both on surveys in the field and excavation results, to present some source of inspiration as a tribute to a great scholar for whom we have the highest admiration and to whom we owe so much. We need not repeat the long history of the search for Avaris and Per-Ramesses, since the identification with Tell el-Dabaa and Qantîr is now common knowledge in Egyptology. What remains to be done is to reveal more and more of the details of the topography of this once important residence of one of the greatest pharaohs of Egypt. On the surface practically nothing can be seen of the former splendour of this city. From the point of view of an archaeologist however, one can reveal the remains of a very fragmented picture and try to understand and interpret the scanty remains. § 2 e Primeval Landscape Located east of the easternmost Nile branch, ‘the Waters of Rea’, a navigable river channel flowing into the Mediterranean, the site of the later Per-Ramesses was situated at the beginning of the land A. H. Gardiner, ‘e Delta Residence of the Ramessides’, JEA 5 (98), 27–38, 79–200, 242–7; A. H. Gardiner, ‘Tanis and Pi-Raamesse: A Retractation’, JEA 9 (933), 22–8; M. Hamza, ‘Excavations of the Department of Antiquities at Qantîr (Faqus District)’, ASAE 30 (930), 3–68; W. C. Hayes, Glazed Tiles from a Palace of Ramesses II at Kantir (MMA Papers 3; New York, 937); P. Montet, Le drame d’Avaris (Paris, 94); L. Habachi, ‘Khataana–Qantîr: Importance’, ASAE 52 (954), 443–562; L. Habachi, Tell el-Dabaa, I: Tell el-Dabaa and Qantîr: e Site and its Connection with Avaris and Pi-Ramesse. Aus dem Nachlaß her- ausgegeben von E. M. Engel unter der Mitarbeit von P. Janosi und C. Mlinar, Redaktion E. Czerny (DGÖAW 23 = UZK 2; Vienna, 200), 65–95; J. van Seters, e Hyksos: A New Investigation (New Haven, 966); M. Bietak, Tell el-Dabaa, II: Der Fundort im Rahmen einer archäologisch-geographischen Untersuchung über das ägyptische Ostdelta, (DGÖAW 4 = UZK ; Vienna, 975), 23–32; M. Bietak, Avaris and Pi-Ramesse, Archaeological Exploration in the Eastern Nile Delta: Ninth Mortimer Wheeler Archaeological Lecture (Oxford, 979), 28–3; M. Bietak, ‘Ramessesstadt’, V (984), 28–46; E. Pusch, H. Becker and J. Fassbinder, ‘Wohnen und Leben oder: weitere Schritte zu einem Stadtplan der Ramsesstadt’, Ä&L 9 (999), 56. * e authors wish to acknowledge the assistance of Nikky Math for producing the illustrations, David Aston for reading the manuscript and Edgar Pusch for providing us with information on his research at Qantir.
28
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Bietak and Muller

23

The Topography of New Kingdom Avaris and Per-Ramesses

Manfred Bietak and Irene Forstner-MÜller*

§ 1 Introduction

Kenneth A. Kitchen, who has the most intimate knowledge of Ramesside inscriptions, has repeatedly drawn pictures and accounts of the major cities of Ramesses II, including, of course, Per-Ramesses. Into these imaginative reconstructions, as Kitchen modestly calls them, a most valuable amount of informa-tion drawn from the texts has been absorbed which has not lost its value over time, despite the long standing research of the Austrian Archaeological Institute and the Pelizäus-Museum Hildesheim in the region of Per-Ramesses in the last decades. We hope with this account, which is based both on surveys in the field and excavation results, to present some source of inspiration as a tribute to a great scholar for whom we have the highest admiration and to whom we owe so much.

We need not repeat the long history of the search for Avaris and Per-Ramesses,� since the identification with Tell el-Dabaa and Qantîr is now common knowledge in Egyptology. What remains to be done is to reveal more and more of the details of the topography of this once important residence of one of the greatest pharaohs of Egypt. On the surface practically nothing can be seen of the former splendour of this city. From the point of view of an archaeologist however, one can reveal the remains of a very fragmented picture and try to understand and interpret the scanty remains.

§ 2 The Primeval Landscape

Located east of the easternmost Nile branch, ‘the Waters of Rea’, a navigable river channel flowing into the Mediterranean, the site of the later Per-Ramesses was situated at the beginning of the land

� A. H. Gardiner, ‘The Delta Residence of the Ramessides’, JEA 5 (�9�8), �27–38, �79–200, 242–7�; A. H. Gardiner, ‘Tanis and Pi-Raamesse: A Retractation’, JEA �9 (�933), �22–8; M. Hamza, ‘Excavations of the Department of Antiquities at Qantîr (Faqus District)’, ASAE 30 (�930), 3�–68; W. C. Hayes, Glazed Tiles from a Palace of Ramesses II at Kantir (MMA Papers 3; New York, �937); P. Montet, Le drame d’Avaris (Paris, �94�); L. Habachi, ‘Khataana–Qantîr: Importance’, ASAE 52 (�954), 443–562; L. Habachi, Tell el-Dabaa, I: Tell el-Dabaa and Qantîr: The Site and its Connection with Avaris and Pi-Ramesse. Aus dem Nachlaß her-ausgegeben von E. M. Engel unter der Mitarbeit von P. Janosi und C. Mlinar, Redaktion E. Czerny (DGÖAW 23 = UZK 2; Vienna, 200�), 65–95; J. van Seters, The Hyksos: A New Investigation (New Haven, �966); M. Bietak, Tell el-Dabaa, II: Der Fundort im Rahmen einer archäologisch-geographischen Untersuchung über das ägyptische Ostdelta, (DGÖAW 4 = UZK �; Vienna, �975), 23–32; M. Bietak, Avaris and Pi-Ramesse, Archaeological Exploration in the Eastern Nile Delta: Ninth Mortimer Wheeler Archaeological Lecture (Oxford, �979), 28�–3; M. Bietak, ‘Ramessesstadt’, LÄ V (�984), �28–46; E. Pusch, H. Becker and J. Fassbinder, ‘Wohnen und Leben oder: weitere Schritte zu einem Stadtplan der Ramsesstadt’, Ä&L 9 (�999), �56.

* The authors wish to acknowledge the assistance of Nikky Math for producing the illustrations, David Aston for reading the manuscript and Edgar Pusch for providing us with information on his research at Qantir.

Page 2: Bietak and Muller

24

Manfred Bietak and Irene Forstner-MÜller

route along the northern Sinai to Palestine (fig. �). As such, the site was most favourably positioned for controlling traffic routes abroad.2 At the same time it oversaw the entrance into the eastern Nile Delta as the Horus road had to pass through a land bridge between the easternmost Nile branch and the natural Bahr el-Baqar drainage system which issued into huge overflow lakes and swamps protecting the approaches to the Delta. In the second millennium bc the northern coast of Egypt had a more southerly position and the site of later Per-Ramesses was still near the effects of the tides.3 This had the advantage that the channel was still navigable during the dry season in late spring and early summer, when the Nile was at its lowest. This is a prerogative for a seagoing harbour in a deltaic landscape.

The landscape of Per-Ramesses as it is now observable to the modern viewer is very different to the primeval landscape (fig. 2).4 Nowadays completely flat,5 it was then a system of turtlebacks—geziras—and eroded depressions in between them. Those extensive turtlebacks consisting of fine yellowish-whitish sand, easily recognizable during excavations, are the remains of older, partly early Holocene river sediments. They were not flooded during the inundation and are therefore ideal places for permanent settlement activity. The maximum flood levels during the Ramesside period were about 5 m above the modern sea level.6 In between those geziras in the lower parts of the Delta the Nile flood transported and deposited its mud and brought fertility to the whole area. These parts could only partly be used, mostly for agricultural purposes.

2 For the topography of the eastern Nile Delta, see Bietak, Tell el-Dabaa II, 47–��2.3 Brackwater fish such as dorados were found in the osteological material of Tell el-Dabaa, see J. Boessneck and A. von den

Driesch, Tell el-Dabaa, VII: Tiere und Historische Umwelt im Nordost-Delta im 2. Jahrtausend v. Chr. anhand von Knochenfunden der Ausgrabungen 1975–1986 (DGÖAW �� = UZK �0; Vienna, �992), 42–3.

4 Bietak, Tell el-Dabaa II, 49–��0; K. Butzer, Studien zum vor- und frühgeschichtlichen Landschaftswandel in der Sahara, III; Die Naturlandschaft Ägyptens während der Vorgeschichte und der Dynastischen Zeit (AAWLM Math-Nat �959/2; Mainz, �959); K. Butzer, ‘Delta’, LÄ I (�975), cols. �043–52.

5 The intensive agricultural use of the land and the sebakh-activity has destroyed most of the tells in the Nile Delta. At the end of the nineteenth century Griffith (F. Ll. Griffith, The Antiquities of Tell el Yahudîyeh, and Miscellaneous Work in Lower Egypt (MEEF 7; London, �890), 56–7) was able to walk on top of the tell between Khataana/Tell el-Dabaa and Qantîr which is a dis-tance of 2 km. In 2007 only a small part of this tell, north of the village of Tell el-Dabaa, is still visible.

6 J. Dorner, ‘Die Topographie von Piramesse-Vorbericht’, Ä&L 9 (�999), 78.

Fig. �. Historical Landscape of the Northeastern Nile Delta.

Page 3: Bietak and Muller

25

The Topography of New Kingdom Avaris and Per-Ramesses

Most of those topographic features are no longer visible due to agricultural levelling and modern irrigation engineering. The main components of the ancient topography, the Nile channels, the lakes and the artificial channels which cut through the ancient settlement and the turtlebacks were, however, made visible by the study of the sedimentation ridges caused by active Nile branches, and the depressions according to the contour maps of the Survey of Egypt7 and by investigating the nature of the sediments by

7 Bietak, Tell el-Dabaa II, 67.

Fig. 2. Primeval Landscape of Avaris/Per-Ramesses (adapted from Dorner, Ä&L 9, map �).

River Inundation Area

Page 4: Bietak and Muller

26

Manfred Bietak and Irene Forstner-MÜller

core drilling.8 The former is a good tool for assessing the water courses on a larger scale, the latter is suitable to probe water courses in a more restricted area but provides much greater precision. In recent years geo-magnetic surveying has also revealed the position of the Nile branches since they show where settlement abruptly stops along the Nile branches, channels and former lakes.9 The colours of such Nile branches show a darker hue than the surrounding land. Where modern housing does not cover ancient features, this most recent modern surveying tool shows the uppermost stratum of ancient settlements and buildings.

Thus the ancient Nile branches in the area of Avaris and Per-Ramesses can be assessed fairly accurately for the second millennium bc with the understanding that their date is provided by the settlements along their banks and by the sherds obtained from their sediments. The river branches during their normal water regime can be roughly estimated to have been 200 to 300 m wide. Josef Dorner has revealed the easternmost Nile branch west of the villages of Khataana and aEzbet Helmy in a south to north stretch of �.5 km (branch F1). North of aEzbet Helmy the river splits into two channels, forming islands after the point of bifurcation. Qantîr, the centre of Per-Ramesses, is situated at the southern part of a major Nile island. It is thus provided with protection and river transport at the same time. Remains of branch F1 were observed by Edouard Naville in �895 when it was still distinctly east of the Roman Tell Abu el-Filûs and it can still be recognized between aEzbet Rushdi el-Nimr and west of Tell Abu el-Shaf aei northwards beyond Qantîr, visible by the Bahr Faqûs drain and the borders of land reclamation towards the former riverbed.�0 It is very obvious that modern irrigation engineers placed the feeder channel Bahr Didamûn (called incorrectly Samaana channel by the Survey of Egypt) along the eastern levee of the branch F1, which was in antiquity artificially heightened by settlement activity.��

In addition to the Nile branches, Josef Dorner was also able to discover artificial channels by core drilling. K1 cuts south of the temple of Tell Abu el-Shaf aei and leads probably to a lake (S1) in front of the reconstructed temple of Amun-Rea-Harakhti-Atum (see below). The direction of the channel has to be revised most probably due to the different orientation of the temple of Tell Abu el-Shaf aei.

Another channel (K2) was observed along the direction of a modern channel leading from aEzbet Silmy northward, passing the Ramesside horse stables at area Q IV,�2 issuing, according to Dorner’s recon-struction, into a lake (S2) in the north of the town. Surely there were more channels within the northern part of Per-Ramesses, which remain to be discovered. One wonders if the depression associated with a sub-turtleback G2 southeast of aEzbet Silmy was not another channel.

There were also channels in the south, at the site of ancient Avaris, which was a component of Per-Ramesses. There is the fossil Nile branch F3 which meandered northwards, but which was already silted up in the Second Intermediate Period and covered with structures at that time. It is possible, however, that its northern course was still open in the form of a bay which was also fed by a canal from a basin S3

8 Preliminary report by Dorner, Ä&L 9, 77–83; final publication in preparation. The detailed reconstruction of the primeval landscape of the region of Tell el-Dabaa–Qantîr we owe to Josef Dorner’s survey activity published in J. Dorner, ‘Archäolo-gischer Survey in der Umgebung von Tell el-Dabaa’, ÖJH 54 (�983), 20–�; J. Dorner, ‘Die Rekonstruktion einer pharaonischen Flusslandschaft’, MAGW �23/�24 (�993/94), 400–5; J. Dorner, ‘Ergebnis der Geländeuntersuchungen zur Rekonstruktion der historischen Topographie von Auaris und Piramesse - ein Vorbericht’, in M. Bietak, J. Dorner, I. Hein and P. Janósi, ‘Neue Grabungsergebnisse aus Tell el-Dabaa und ‘Ezbet Helmi im östlichen Nildelta �989–�99�’, Ä&L 4 (�994), ��–�5; J. Dorner, ‘Bericht über die topographischen Untersuchungen im Gebiet von Avaris und Piramesse’, in M. Bietak et al., Tell el-Dabaa, VIII (in preparation).

9 E. Pusch, H. Becker and J. Fassbinder, ‘Wohnen und Leben oder: weitere Schritte zu einem Stadtplan der Ramsesstadt’, Ä&L 9 (�999); I. Forstner-Müller, W. Müller, C. Schweitzer and M. Weissl, ‘Preliminary Report on the Geophysical Survey at aEzbet Rushdi/ Tell el-Dabaa in Spring 2004’, Ä&L �4 (2005), �0�–9; I. Forstner-Müller and W. Müller, ‘Neueste Ergebnisse des Magnetometersurveys während der Frühjahrskampagne 2006 in Tell el-Dabaa/Qantîr’, Ä&L �6 (2007), 79–82.

�0 Personal surface exploration.�� Bietak, Tell el-Dabaa II, 8�.�2 E. Pusch, ‘Vorbericht über die Abschlußkampagne am Grabungsplatz Q IV �997’, Ä&L 9 (�999), Abb. 2.

Page 5: Bietak and Muller

27

The Topography of New Kingdom Avaris and Per-Ramesses

(see below) as seems increasingly probable from the results of the geophysical survey (fig. 2). South of gezira G5, southwest of Tell el-Dabaa, this fossil Nile branch F3 left a depression, which became in antiq-uity an overflow lake still to be seen in a reduced form today.

Still visible in surface survey is an ancient lake (S3), which gave Tell el-Dabaa its older name Tell el-Birka.�3 From a bay north of the Tuthmoside palace precinct (Area H) a channel (K3) leads towards this lake and another one (K4) towards the fossil branch F3. This could have been at least a drainage channel. It still remains to be assessed whether these channels were active in the second millennium bc. The drilling of Josef Dorner did not bring definitive results in this respect, but the features look too unusual to be natural. The deep slope between the geziras G5 and G6 (see below) makes such an assumption extremely likely.

Another channel (K5) surrounded aEzbet Machali and aEzbet Yanni from the south, separating this small turtleback from the bigger one in the south. According to recent survey work this seemed to have had an active connection to Nile branch F2.

The surface features and the results of core drilling show the position of the turtlebacks, the geziras, which remained outside the reach of the annual Nile flood and provided dry ground suitable for settle-ment.�4 They constitute the potential components of the town and also lead to its segmentation. The vicinity of huge geziras near the easternmost Nile branch contributes to the potential growth of a settle-ment into a large town. Such a situation was provided by nature in the area of Per-Ramesses, in addition to the utility of Nile branches as traffic routes.

The part of Qantîr on which the centre of Per-Ramesses developed seems to have been one huge gezira, as far as the drilling provided information. The gezira G1 provided an enormous space of at least 200 ha. It is intersected by the above-mentioned channels, which cut steeply into the sand mound. What is striking is that the southwestern segment of this gezira has a square shape, which is accentuated by the above-mentioned channel K2, which runs northwards. As the most important precincts of Per-Ram-esses were situated within this square, one gets the impression that this part of the gezira is the result of landscaping, enlarging an existing gezira to the south. All the geziras of Qantîr were settled during the New Kingdom, from the beginning of the Nineteenth Dynasty onwards. Pottery from the Eighteenth Dynasty�5 may originate from detached hamlets or from bricks produced during the enormous building activity of the Ramessides in the south at Avaris.

Gezira G2 on the southeast of Qantîr may have been separated by a channel, dredged through G1. As revealed by surface surveys and geomagnetic surveying, gezira G3 at aEzbet Yanni and aEzbet Machali displays the typical settlement structure of the Ramesside Period.�6 Gezira G4 at aEzbet Ziz has only been explored by surface survey and shows occupation of that time as well.�7

Geziras G5 and G6 represent the centre of Avaris with an occupation from the beginning of the Twelfth Dynasty till the end of the Hyksos Period. At the southern edge of G5 we have evidence of gardens from the Nineteenth Dynasty.�8 The problem is that, due to agricultural levelling, most of the Rames-side remains have been scraped away. The western part of G6 was occupied throughout the Eighteenth

�3 G. Maspero, ‘Notes sur quelques points de grammaire et de l’histoire’, ZÄS 23 (�885), ��–�3; Habachi, ASAE 52, 458; Habachi, Tell el-Dabaa I, 27; Bietak, Tell el-Dabaa II, 35 n. 65.

�4 Dorner, Ä&L 9, map �.�5 D. A. Aston, ‘Qantir/Piramesse-Nord – Pottery Report �988’, GM ��3 (�989), 7–32; D. A. Aston and E. Pusch, ‘The Pot-

tery from the Royal Horse Stud and its Stratigraphy: The Pelizäus-Museum Excavation at Qantir/Per-Ramesses, Sector Q IV’, Ä&L 9 (�999), 53–4.

�6 Forstner-Müller and Müller, Ä&L �6, fig. �.�7 Dorner, Ä&L 9, map �.�8 I. Hein and P. Jánosi, Tell el-Dabaa, XI: Areal A/V: Siedlungsrelikte der Späten Hyksoszeit (DGÖAW 25 = UZK 2�; Vienna,

2004), �87–8, Plan 2A.

Page 6: Bietak and Muller

28

Manfred Bietak and Irene Forstner-MÜller

Dynasty and shows activity in the Nineteenth Dynasty. The same is true for the precinct of the Seth temple on the eastern half of this gezira (see below).

Gezira G7 seems to be the continuation of G6 and was explored only in the area south of Khataana.�9 It contained settlement remains of the Second Intermediate Period until 500 m south of aEzzawîn.

In the east there is the enormous gezira G8 of Samaana, which extends more than 3 km in a north-east–southwest direction. Apart from a pommel of a dagger of Horemheb20 and a stone-lined well of Ramesses II, this region only has evidence of Ramesside cemeteries including slipper coffins from Gezirat el-Baghl, which have never officially been excavated. Scattered evidence of cemeteries are also reported further south at aEzbet Gayel between el-Samaana and el-Didamûn.2�

The area south of Ezzawin and the west bank of the Nile channel F1 have never been explored. At present it seems that only the Nile island of Qantîr (G1) was densely settled with an outskirt at G3 and a temple and other installations at Tell el-Dabaa and aEzbet Helmy (see below).

§ 3 From Avaris to Per-Ramesses

Per-Ramesses, the splendid residence of Ramesses II, was not a sudden creation but had a ‘pre-history’ connected to its very favourable topographic position, linking Egypt and the Near East (see above § �).

Avaris was a town of Near Easterners from the late Twelfth Dynasty onwards. It seems to have had a special status, and may have served the Egyptian crown particularly for trade expeditions to the Near East. Under such auspices it is understandable that Canaanite cults had been introduced to this place at that time, especially the cult of the northern Syrian storm god Baaal Zephon, as illustrated on a local cylinder seal (fig. 3).22 This god was identified with the Egyptian weather god Seth who became, with his Asiatic attributes, the ‘Lord of Avaris’ and more than 400 years later the ‘Father of the fathers’ of the Nineteenth Dynasty, a feature which probably originated from this place.23 In order to understand this continuity one has also to assess the local continuity of the town and its development as a residen-tial town, which started with Nehesy, a ruler of the Fourteenth Dynasty, the first king with the epithet ‘Beloved of Seth, the Lord of Avaris’.24

In the Hyksos Period Avaris became one of the biggest towns in the Near East and Egypt. It was accord-ing to the Kamose Stela a harbour town25 and the enormous quantities of imported pottery, especially the amphorae from Canaan and jugs from Cyprus are a proof of its first rate importance as a trading centre.

After the conquest of Avaris by Ahmose the town was first abandoned, except for the precinct of the temple of Seth where an unbroken activity till the Amarna Period can be observed.26 This temple seems never to have been abandoned. Along the banks of the Nile south of its bifurcation, at the village of aEzbet Helmy, enormous silo installations and a large magazine of the early Eighteenth Dynasty were erected.27

�9 Investigations J. Dorner, unpublished.20 M. Bietak, ‘Ein altägyptische Weingarten in einem Tempelbezirk’, AÖAW �22 (�985), 267–78, esp. 27� (fig. 7).2� Habachi, Tell el-Dabaa I, 37–8.22 E. Porada, ‘The Cylinder Seal from Tell el-Dabaa’, AJA 88 (�984), 485–8; M. Bietak, ‘Zur Herkunft des Seth von Auaris’,

Ä&L � (�990), �2–�3; C. Uehlinger, ‘Leviathan und die Schiffe in Ps �04, 25–6’, Biblische Notizen 7�/4 (�990), 493–526.23 Bietak, Ä&L �.24 J. von Beckerath, Untersuchungen zur politischen Geschichte der Zweiten Zwischenzeit in Ägypten (ÄF 23; Glückstadt, �965),

82–5; M. Bietak, ‘Zum Königreich des aA-zH-Ra NeHesi’, SAK �� (�984), 59–78.25 L. Habachi, The Second Stela of Kamose and his Struggle against the Hyksos Ruler and his Capital (ADAIK 8; Glückstadt,

�972), 37.26 Bietak, Ä&L �, �4. 27 M. Bietak, J. Dorner and P. Janosi, ‘Ausgrabungen in dem Palastbezirk von Avaris, Vorbericht Tell el-Dabaa/aEzbet Helmi

�993–2000’, Ä&L �� (200�), 59–67.

Page 7: Bietak and Muller

29

The Topography of New Kingdom Avaris and Per-Ramesses

Among the constructions was also a small palatial building with an assembly hall.28 Kerma house-hold pottery29 and flint arrow-tips of Kerma type30 document the presence of Nubian soldiers, particu-larly archers. Soon afterwards, still in the early Eighteenth Dynasty, the site was used as a camp and a burial place for soldiers.3� Osteological evidence again shows a substantial Nubian population component among them. Some burials seem to be the result of brutal executions.32

The pottery production with its specific blend of Middle Bronze Age and Egyptian typology con-tinued nearly unbroken into the time of the Tuthmosides, which indicates that the population of the Hyksos Period was still around, perhaps to some extent also used for military service.33 The same could be said of the scarab production and the weapon typology of the New Kingdom, which is, however, less specific for the site.34 This fact, together with the uninterrupted activity in the temple of Seth, explains the continuity of Canaanite cults on the site of former Avaris into the time of the New Kingdom.

In the early Tuthmoside Period, most probably during the joint reign of Hatshepsut and Tuthmosis III, a palatial compound of �3 acres was constructed on top of the cemeteries and the former Hyksos citadel.35 The palaces were built on high platforms, the biggest measured �60 × 87 m. Two of the palaces were furnished with Minoan wall paintings with Minoan court symbols, suggesting a close connection with the court of Knossos at that time.36 The sheer size of the biggest palace and the palatial precinct makes the residence of a member of the royal family, if not the king himself, likely. Of particular interest is building L beside the main palace G which lasted until the reign of Amenhotep II and gives the impression of high-level administration at this place.37 The evidence suggests that this was a major military stronghold, and the presence of the huge palatial compound raises the question as to whether this was the military and

28 Bietak, Dorner and Janosi, Ä&L ��, 65 (fig. 22). 29 I. Hein, ‘Kerma in Auaris’, in C.-B. Arnst, I. Hafemann and A. Lohwasser (eds), Begegnungen: Antike Kulturen im Niltal.

Festgabe für E. Endesfelder, K.-H. Priese, W.F. Reineke u. S. Wenig (Leipzig, 200�), �99–2�2.30 A. Tillmann, ‘Die Steinartefakte’, in M. Bietak and I. Hein (eds), Pharaonen und Fremde: Dynastien im Dunkel, Sonderausstel-

lung des Historischen Museums der Stadt Wien in Zusammenarbeit mit dem Ägyptologischen Institut der Universität Wien und dem Österreichischen Archäologischen Institut Kairo, Rathaus Wien, Volkshalle, 8. Sept.–23. Oktober 1994 (Vienna, �994), 257 (no. 348–9).

3� Bietak, Dorner and Janosi, Ä&L ��, 67–74.32 Bietak, Dorner and Janosi, Ä&L ��, 69–73.33 M. Bietak, ‘Where Did the Hyksos Come from and Where Did They Go?’, in W. V. Davies (ed.) The Second Intermediate

Period (13th–17th Dynasties): Current Research, Future Prospects (Leuven, in preparation).34 Bietak, in Davies (ed.), The Second Intermediate Period.35 Bietak, Dorner and Janosi, Ä&L ��, 74–�0�; M. Bietak, ‘The Tuthmoside Stronghold Peru-nefer’, EA 26 (2005), �3–�7; M.

Bietak, ‘Neue Paläste aus der �8. Dynastie’, in P. Jánosi (ed.) Structure and Significance, Bau und Bedeutung (Festschrift für Dieter Arnold) (Vienna, 2005), �3�–67; M. Bietak, ‘Egypt and the Aegean: Cultural Convergence in a Thutmoside Palace at Avaris’, in C. H. Roehrig, R. Dreyfus and C. A. Keller (eds), Hatshepsut: Exhibition catalogue of the Metropolitan Museum of Art (New York, 2006), �35–43; M. Bietak and I. Forstner-Müller, ‘Ausgrabungen im Palastbezirk von Avaris, Vorbericht Tell El-Dabaa/aEzbet Helmi Frühjahr 2003’, Ä&L �3 (2003), 39–50; M. Bietak and I. Forstner-Müller, ‘Ausgrabung eines Palastbezirkes der Tuthmosi-denzeit bei aEzbet Helmi/Tell el-Dabaa, Vorbericht für Herbst 2004 und Frühjahr 2005’, Ä&L �5 (2005), 65–�00.

36 M. Bietak, N. Marinatos and C. Palyvou, Taureador Scenes in Avaris and Knossos (Vienna, 2007).37 Bietak and Forstner-Müller, Ä&L �3, 39–50; Bietak and Forstner-Müller, Ä&L �5, 65–�00.

Fig. 3. Cylinder seal depicting Seth/Baaal Zephon from the Thirteenth Dynasty found in area F/I.

Page 8: Bietak and Muller

30

Manfred Bietak and Irene Forstner-MÜller

naval base of Peru-nefer (see also below the discussion of the harbours of Per-Ramesses)?38 We can also be fairly sure that Canaanite cults flourished in this town as the temple of Seth-Baaal continued to function, particularly as Canaanite cults are now well attested in a fortress of Amenhotep II in northern Sinai.39

After a period of abandonment the area was reoccupied late in the reign of Amenhotep III.40 In the district of the temple of Seth a scarab of this king may offer proof of continuity.4� Under Horemheb a huge fortress with a buttressed fortification wall of mudbrick was constructed towards the river upon the former palatial precinct (fig. 4).42 It is not yet clear if this fortification, which can be followed on the geo-magnetic survey for more than 400 m, is an enlargement towards the north of an already existing fortress south of the Tuthmoside palace precinct. Agricultural levelling in the last �00 years has, unfortunately, largely destroyed the inner structure of the Horemheb fortress. The northern fortification wall of this for-tress is parallel to the northwestern enclosure wall of the temple of Seth, which seems to have been rebuilt under this king,43 and it is probable that the fortification wall enclosed a huge area which included this temple in its midst. Parallel to the northern enclosure wall of Horemheb is a geophysical feature south-west of aEzbet Rushdi, which looks like an embankment at the northern edge of the water basin S3.

The fortress must have retained in its time the residential function from the Tuthmoside Period as we have found remains of beautiful alabaster and faience tiling within the fortress, just under the agricultural soil.44

In the late Eighteenth Dynasty, most probably under Horemheb,45 the site of Qantîr started to be used. The military workshops found in area Q I by the Hildesheim expedition seem to go back to this period.46

The Tuthmoside palaces and this enormous fortress of Horemheb are most important for our under-standing of the development of the town towards the Ramesside concept of a Delta residence such as Per-Ramesses. We have palaces indicating at least part-time royal presence and we have the temple of Seth, which was obviously destroyed in the Amarna Period and rebuilt in the restoration period under Tutankhamun and Horemheb.47 The latter king directed Egypt to a more active role in politics and warfare in the Near East and thought it necessary to fortify Egypt’s northeastern border and to rebuild the military base of the Tuthmosides in this region. Horemheb seems also to have constructed fortresses in North Sinai, such as Tell el-Borg.48

The duality of residence and temple of the town god was repeated under Seti I who continued to recon-struct the temple of Seth49 and built a new residence, this time in the hitherto largely untouched environment of southern Qantîr. The position is perhaps comparable to the North Riverside Palace of Akhenaten in the extreme north of Amarna. Seti I also continued to build fortresses along the land route to northern Sinai.50

38 Bietak, EA 26, �3–�7.39 J. K. Hoffmeier and K. A. Kitchen, ‘Reshep and Astarte in North Sinai: A Recently Discovered Stela from Tell el-Borg’,

Ä&L �7 (2007), �27–36.40 Aston, Ä&L ��, �94.4� Bietak, AÖAW �22, 276 (fig.�2).42 Bietak, Dorner and Janosi, Ä&L ��, �0�; Aston, Ä&L ��, �67–8�.43 Bietak, AÖAW �22, 267–78; Bietak, �990, �2–�3.44 Bietak, Dorner and Janosi, Ä&L ��, �0�, fig. 53.45 See the faience vessel with the cartouches of Horemheb, Bietak, Tell el-Dabaa II, 45; Habachi, Tell el-Dabaa I, 255, Tf. 49A.46 E. Pusch, ‘Metallverarbeitende Werkstätten der frühen Ramessidenzeit in Qantîr/Pi-Ramesse/Nord - Ein Zwischen-

bericht’, Ä&L � (�990), �06–8.47 Bietak, AÖAW �25, 267–78; Bietak, Ä&L �, �2–�3.48 J. K. Hoffmeier, personal communication.49 L. Habachi, ‘Sethos I’s Devotion to Seth and Avaris’, ZÄS �00 (�974), �00; Bietak, Ä&L �, ��.50 A. H. Gardiner, ‘The Ancient Military Road between Egypt and Palestine’, JEA 6 (�920), 99–��6; The Epigraphic Survey,

Reliefs and inscriptions at Karnak IV, The Battle Reliefs of King Sety I (OIP; Chicago, �986).

Page 9: Bietak and Muller

3�

The Topography of New Kingdom Avaris and Per-Ramesses

We may assume that the former site of Avaris remained in its function as a military, and most likely naval, base as in the Eighteenth Dynasty.

§ 4 The Temples of Per-Ramesses

His Majesty has built himself a Residence whose name is ‘Great of Victories’. It lies between Syria and Egypt, full of food and provisions. It follows the model of Upper-Egyptian Thebes, its duration like that of Memphis. The Sun arises in its horizon, and (even) sets within it.5�

Despite this definition, Ramesses II showed with the system of temples in his new capital that he wished to create a different religious world than in Thebes. Per-Ramesses was not intended to be a replica of the southern capital, as was attempted in the Twenty-first and Twenty-second Dynasties at Tanis, but was endowed with a religious concept in accord with the personality of the king. An antagonism to Thebes might be recognised, as the temple of Amun was not the centre of religious life. Ramesses II focused on

5� P. Anastasi II, �, 2, P. Anastasi IV, 6, 2: translation K. A. Kitchen, Pharaoh Triumphant: The Life and Times of Ramesses II, King of Egypt (Warminster, �982), ��9.

Fig. 4. Geophysical map of the Horemheb Fortress.

Page 10: Bietak and Muller

32

Manfred Bietak and Irene Forstner-MÜller

solar religion by creating a model combining the northern and the southern Heliopolis and including the local tradition, which is Seth. This is the key to understanding the sacred landscape of Per-Ramesses.

The major temple was dedicated to a group of state gods, focused on the solar circle of Heliopolis in the northern part of Egypt which seems to show, from the beginning of the history of Egypt, a kind of religious and perhaps political influence on the eastern part of the Delta.52 It was a temple for Amun-Rea-Harakhti-Atum. The traditional local god Seth was, despite his Asiatic guise, also connected to this circle by giving him the epithet ‘Son of Nut’.53 In addition, Ramesses gave the gods for whom other temples were installed a very personal accent, differing from their traditional character by adding epithets related to the king and the residence town such as ‘Amun of Ramesses of Per-Ramesses’, ‘Ptah of Ramesses of Per-Ramesses’, ‘Seth of Ramesses of Per-Ramesses’, etc. Many gods, like Amun, Rea, Ptah, Astarte, Anat, Wadjet and Horus, had their cult in temples in Per-Ramesses.54 Within this section, however, we shall focus on those who shed some light on topographical issues.

The main temple of the city, dedicated to Amun–Rea-Harakhti–Atum,55 was situated in the town centre (fig. 5).56 This temple is mentioned in contemporary documents: a hymn57 and the peace contract with the Hittites.58 Josef Dorner has convincingly reconstructed its site to the north of the major palace (see below).59 It could not have been situated south of the palace because Mahmud Hamza found there remains of faience workshops and wine magazines60 and the Hildesheim Museum excavated further in the south, in area Q I, chariot and military workshops.6� Indeed north of the site of the palace the German mission found quartzite blocks with monumental inscriptions.62 This would fit well with the previous observation of Eric Uphill that the most important parts of this building were made of quar-zite.63 Based on the fragments which could derive from this building, the length of this temple was esti-mated by Uphill to be about �80 m.64 Uphill also attempted to reconstruct the appearance of the front pylon65 with four colossal statues,66 two sitting, two standing. A canal found by Josef Dorner passed the front of the temple of Tell Abu Shafaei with the Nile branch F1 and continued into the lake S1 which he reconstructed as being at the front (west) of the Amun-Rea-Harakhti-Atum temple, which also provided a quay for the palace further south (see below).

52 H. Kees, Das Alte Ägypten (Berlin, �955), �05; W. Helck, Zur Verwaltung Ägyptens im Mittleren und Neuen Reich (PdÄ 3; Leiden, �958), �97–8.

53 P. Montet, Les nouvelles fouilles de Tanis, 1929–1932 (Paris, �933), pl. ix. 54 Bietak, Tell el-Dabaa II, n. 878; E. P. Uphill, The Temples of Per-Ramesses (Warminster, �984), 206–2�; recently H. Sourouz-

ian, ‘Seth fils de Nout et Seth d’Avaris dans la statuaire royale ramesside’, in E. Czerny, I. Hein, H. Hunger, D. Melman, and A. Schwab (eds), Timelines: Studies in Honour of Manfred Bietak (OLA �49; Leuven, 2006), �, 33�–54.

55 Uphill, The Temples of Per-Ramesses, 206–2��.56 Following Dorner’s convincing argumentation, ‘Zur Lage des Palastes und des Haupttempels der Ramsesstadt’, in

M. Bietak (ed.), Haus und Palast im Alten Ägypten (DGÖAW �4 = UZK �4; Vienna, �996); Dorner, Ä&L 9, 80.57 P. Anastasi II, IV.58 KRI II, 226–33; E. Edel, Der Vertrag zwischen Ramese II. von Ägypten und Hattusili III von Hatti (Berlin, �997).59 Dorner, in M. Bietak (ed.), Haus und Palast im Alten Ägypten; Dorner, Ä&L 9, 80. 60 Hamza, ASAE 30, 42.6� Pusch, GM ��2; Pusch, Ä&L �; Pusch in A. Eggebrecht (ed.), Pelizäus-Museum Hildesheim – Die Ägyptische Sammlung

(Zaberns Bildbände zur Archäologie �2; Mainz, �993).62 Dorner, Ä&L 9, 80.63 Uphill, The Temples of Per-Ramesses, �43 (T79/80), �55 (Q2), �94, 206–7.64 Uphill, The Temples of Per-Ramesses, 2��.65 Uphill, The Temples of Per-Ramesses, 206–��, pl. 2�;66 J.-J. Clère, ‘Nouveaux documents relatifs au culte des colosses de Ramsès II dans le Delta’, Kêmi �� (�950), 24–46.

Page 11: Bietak and Muller

33

The Topography of New Kingdom Avaris and Per-Ramesses

Ancient Egyptian towns were structured by their temples and by the processional roads connecting them. This may also have been the case with Per-Ramesses. Sources of the Ramesside Period (P. Anas-tasi II, �, 4–5 and P. Anastasi IV, 6, 4–5) give us some information on the topographic position of some temples of Per-Ramesses, but there is ambiguity if the directions in which the temples are mentioned are

Fig. 5. Main Palace at Per-Ramesses (adapted from Dorner, Ä&L 9, map 2).

Page 12: Bietak and Muller

34

Manfred Bietak and Irene Forstner-MÜller

aligned to the palace (bxn) or if they give positions within the town:

Its West is the house of Amun, its South the house of Seth, Astarte its East, Wadjet its North.

Those temples seem to be defined by the cardinal points and one might expect that they were situ-ated near the edges of the town and were therefore used to define its borders and limits. From those four temples only one can be considered to have been identified with certainty: the temple of Seth on the western edge of Tell el-Dabaa near the centre of former Avaris (fig. 6).67 Its enclosure wall was partly exca-vated by Naville and afterwards used as quarry. The Austrian Institute revealed the rest in test areas. It has a trapezoidal precinct of about 350 × 250 m, which must have topographical determinants. It is surely no coincidence that the northern external wall of the Seth temple is parallel to the northern enclosure wall of the fortress of Horemheb. The south wall, of more than 20 cubits thick, is precisely oriented east–west, as is the temple, which was largely destroyed by agriculture and stone robbers. Remains of a pylon and the foundations of a colossal statue were found by geophysical surveying.68 A smaller enclosure wall of the Eighteenth Dynasty was found beneath. The temple was surrounded, both inside and outside the enclo-sure, by a grove of trees whose pits were still found in a regular pattern. Inside were also vineyards.69

67 Habachi, ZÄS �00, �00; Uphill, The Temples of Per-Ramesses, 202, 238; Bietak, Ä&L �, �0–�4.68 Forstner-Müller and Müller, Ä&L �6 (2007), fig.�.69 Bietak, Ä&L �, ��–�2.

Fig. 6. Temple of Seth, with tree pits.

Page 13: Bietak and Muller

35

The Topography of New Kingdom Avaris and Per-Ramesses

Fig. 7. ‘400 Year Stela’ (photo D. Johannes).

Page 14: Bietak and Muller

36

Manfred Bietak and Irene Forstner-MÜller

The Seth temple in Avaris is the most likely original location of the ‘400 Year Stela’ (fig. 7).70 On this stela Seth is represented as a Near Eastern deity with a high crown with a long streamer, with horns and a half-Egyptian, half-exotic dress with crossed decorated bands and a kilt which reminds us of the dress of the Sea Peoples.7� Since Seth of Avaris is originally Baaal Zephon,72 the patron of seafarers, this kind of kilt could have been added to his symbolic regalia. Indeed the northwest wall of the temenos borders the presumed harbour area of Avaris. One may even envisage a side axis leading to the harbour area, which would explain the odd orientation of the eastern temenos wall, which is probably parallel to the axis of a northern gate. Nearby within the precinct an artificial lake was found, perhaps a sacred lake belonging to this temple.

From the representation of Baaal Zephon from the early Thirteenth Dynasty73 to the Seth on the ‘400 Year Stela’ is a long undocumented period to follow the development of Seth of Avaris. That we find this god, however, with Asiatic features already before Ramesses II may be deduced from the representation of Seth on the Stela of Seti I at Qadesh.74 On an Egyptian stela from Ugarit also dated to the beginning of the Nineteenth Dynasty we have the representation of Baaal Zephon in the same regalia as Seth on the ‘400 Year Stela’, with the high crown with the streamer, crossed bands and wAs-scepter.75 Part of the kilt and gown is not preserved. This stela proves again the close association of Seth of Avaris with Baaal Zephon.

The identification of this temple at Tell el-Dabaa was established by its position and size, by a lintel mentioning Seth ‘great of power’ and the names of Horemheb, which seem to replace the names of Tut-ankhamun.76

For the identification of the other three temples, mentioned in P. Anastasi II and IV the situation is more difficult. It was recently proposed that the temple at Tell Abu el-Shaf aei should be identified with the great Amun temple.77 At this site Shehata Adam discovered, in �952, a part of the base of a colossal limestone seated statue of Ramesses II, which can be reconstructed to a height of �5 m (fig. 8).78 Two further fragments and a column base were found nearby.79 These fragments certainly did not stand in isolation but in connection with a temple. Some years ago several large limestone blocks, which could belong to this temple, were unearthed in this area during drainage work.80

Recent geophysical survey by the Pelizäus-Museum Hildesheim brought this temple to light with more precision. It is orientated north–south opening to the south. It is in an oblique position to the river.8� For this unusual orientation a convincing explanation has to be found, because normally temples are oriented towards, or parallel, to the river.82 The explanation is, that the main reference point for this shrine is the temple of Amun-Rea-Harakhti-Atum to which it is aligned at an angle of 90°. There is

70 R. Stadelmann, ‘Vierhundertjahres Stele’, LÄ VI, �043; Bietak Ä&L �, frontispiece.7� Montet, Les nouvelles fouilles de Tanis, 1929–1932, 207.72 R. Stadelmann, Syrisch-palästinensische Gottheiten in Ägypten (Leiden, �967).73 On a seal cylinder Bietak, Ä&L �, �5, fig. 5. For another proof that Baaal is identified with Seth from the Middle Kingdom

onwards, see O. Goldwasser, ‘Canaanites Reading Hieroglyphs. Horus is Hathor? - The Invention of the Alphabet in Sinai’, A&L �6 (2007), �2�–60.

74 PM VII, 352, J. Pritchard, The Ancient Near East in Pictures (Princeton, �954), fig. 3�7; KRI I, 25.75 Louvre AO fig. �3.�76, see M. Yon, Ras Shamra-Ougarit, VI: Arts et Industries de la Pierre (Paris, �99�), 284–6, no. �.76 M. Bietak, ‘Kat. Nr. 393, Inschriftenstein des Haremhab’, in M. Bietak and I. Hein (eds), Pharaonen und Fremde, Dynas-

tien im Dunkeln (Vienna, �994) 282–3, nr. 393; Bietak, Ä&L �, ��–�2, fig. 2.77 Dorner, Ä&L 9, 80–�.78 S. Adam, ‘Recent Discoveries in the Eastern Delta (Dec. �950–May �955)’, ASAE 55 (�958), 3�8–24; Bietak, Tell el-Dabaa

II, pl. 43–5A.79 Adam, ASAE 55, 320.80 Dorner, Ä&L 9, 80.8� E. Pusch, personal communication.82 The reconstruction in Dorner, Ä&L 9, fig.2 shows the temple at a slightly different angle.

Page 15: Bietak and Muller

37

The Topography of New Kingdom Avaris and Per-Ramesses

Fig. 8. Feet of a colossal statue of Ramesses II at Tell Abu el-Shafaei.

Page 16: Bietak and Muller

38

Manfred Bietak and Irene Forstner-MÜller

no doubt that Abu el-Shaf aei was an important temple, however, the identification with the temple of Amun, mentioned in P. Anastasi II and IV, is highly unlikely since, firstly, the precinct is too small; the size of the Seth temple is three times as big, thus it would be odd that the temple for the most important state god of the New Kingdom would be smaller and, secondly, the orientation is north–south i.e. it is aligned 90° towards the entrance axis of the main temple of Amun–Rea-Harakhti–Atum, making it a temple subsidiary to a canonically east–west oriented major temple which is oriented according to the canonical orientation of main temples of that time east–west.

A much better candidate as the resident deity of this temple would be the Delta goddess Wadjet, whose shrine was 25 km to the north at Imet (Tell Nebeshe, eastern Buto) and who according to P. Anas-tasi II, �, 4–5 and P. Anastasi IV, 6, 4–5 had a house in the north of Per-Ramesses.83 Of course we have no proof for this identification except for the position in the north, but it is absolutely uncertain as to whether there were more temples further downstream.84 If we assume, however, that the four temples referred to were in the vicinity of the palace, then the identification of the temple deity as Wadjet would make sense. In this case the temple of Amun of Ramesses should be expected opposite the palace or the main temple of Amun–Rea-Harakhti–Atum, approximately at the site of the southern Qantîr bridge or more towards its west, as proposed by Kenneth Kitchen.85 No geophysical survey has yet taken place there. Such a position would rank the Amun temple nearly equally to the temple of Amun–Rea-Hara-khti–Atum and would explain why it was mentioned second in the Hittite peace treaty.86 It would align those two temples as exponents of the two major sun theologies (northern and southern Heliopolis) within the solar axis and would make understandable the above-cited definition ‘The Sun arises in its horizon, and (even) sets within it.’

If the idea that the passages in P. Anastasi II and IV reflected the situation of temples oriented and constructed around the palace, then a second temple of Seth, namely ‘Seth of Ramesses of Per-Ramesses’ has to be assumed southwest of the site of the palace and the temple of Amun–Rea-Harakhti–Atum, opposite to the temple of Tell Abu el-Shaf aei. Hourig Sourouzian already proposed this although on dif-ferent grounds; although there is no mention of Avaris nor of Seth of Avaris in the time of Ramesses II, despite a significant quantity of textual evidence, statuary mentioning Seth shows a strong association of this god to the Heliopolitan theology.87 This could be taken also as a physical proximity to the main temple of Per-Ramesses. Geophysical surveying or in situ monuments do not disclose any temple but one cannot rule out that a temple of more modest size was situated under aEzbet Yasergi or in the areas not yet surveyed to its south.

In connection with this system of temples one has to search for the temple of Ptah of Ramesses and should also assume that there was a processional axis from the major temple to the east towards the Astarte temple. In favour of such an assumption is the plan of a building found by the geophysical survey of the Pelizäus-Museum south of Qantîr, northeast of the royal horse stables excavated by Edgar Pusch at Q IV (fig. 9). This building, with a length of about �00 m within a precinct with numerous magazines, could be identified as either a temple or a palace.88 The room plan seems to comprise a first courtyard and

83 Bietak,Tell el-Dabaa II, 208–9.84 Indeed a lintel showing a scene of smiting enemies by Ramesses was found 400 m downstream within a precinct with en-

closure wall by E. van den Brink (Dorner, Ä&L 9, 8�), but nobody knows what kind of function this building had. For a temple defining the size of Per-Ramesses it is too small in any case.

85 Already K. A. Kitchen, Pharaoh Triumphant, �23 (fig. 42).86 KRI II, 226; Edel, Der Vertrag zwischen Ramese II. von Ägypten und Hattusili III von Hatti, �7.87 Sourouzian, in E. Czerny et.al (eds), Timelines: Studies in Honour of Manfred Bietak; Uphill, The Temples of Per-Ramesses, �90.88 Pusch, in Petschel and von Falck (eds.), Pharao siegt immer, 240 (fig. �).

Page 17: Bietak and Muller

39

The Topography of New Kingdom Avaris and Per-Ramesses

perhaps a second one, into which some installations were built. Then follows a portico with two rows of columns, a vestibule with two rows of columns, a central hall with six columns, flanked east and west by two square rooms with four columns each. To the south is a room which appears to be a vestibule, at the rear of which a square room with four columns, flanked by side rooms, may be reconstructed or, alterna-tively, a broad room, oriented east–west with two rows of columns. If a temple, it is probably aligned in relation to a major temple or ceremonial road, as main temples have an east-west orientation. Otherwise the temple would be oriented towards the south, to the river F2. We suggest that it could be a ‘House of Millions of Years’ of Ramesses, subsidiary to the main system of processional roads. A ‘House of Millions of Years of Ramesses in the House of Rea’ is mentioned in a title.89 Such a temple of a significant size could be expected in such a position, but no definite proof can be offered at the moment.

89 Habachi, Tell el-Dabaa I, 53.

Fig. 9. Hypothetical Temple of Millions of Years (after Pusch, in Petschel and von Falck (eds.), Pharao siegt immer, 240 (fig. �).

Page 18: Bietak and Muller

40

Manfred Bietak and Irene Forstner-MÜller

The position of the temple of Astarte is still unclear despite surveying and despite the existence of a block from her temple and a doorjamb of a sanctuary of Reshef, which was acquired by the Hildesheim mission.90 According to P. Anastasi II and IV one can assume that the temple stood in the east of Qantîr. Scarabs of Astarte and Seth-Baaal found at the royal stables in area Q IV in the southern part of Qantîr may indicate that the deities were still venerated by the local population.9�

We know of another temple of ‘Amun in the Harbour of Avaris’,92 which can only have been situated at the river in the southern part of Per-Ramesses, in Avaris.

§ 5 The Palace

The position of the major palace of Per-Ramesses can be assessed by several finds. In the south of Qantîr New Kingdom sherds were found within the mound at aEzbet Yasergi.93 This mound, which is today used as the local cemetery, is covered on its west by the village of aEzbet Yasergi and was most prob-ably once the core of an artificial platform, held together by walls the remains of which were found by the Pelizäus-Museum Hildesheim.94 The wall was constructed with its outer (plastered and whitewashed) face to the south. At the edge of the village, west of the cemetery, a huge plinth of red granite of consider-able size is still half in situ.95 It sank with the removal of sand for nearby building projects. The architec-ture of the palace is completely gone.

Southeast of the cemetery, a brickwork construction with regular spurs into the sand, which could still be seen until �968, can only be connected to the palace (fig. �0).96 It seems to have been a platform,

90 Pusch, in Petschel and von Falck (eds.), Pharao siegt immer, 24�.9� Pusch, in Petschel and von Falck (eds.), Pharao siegt immer, 262–3.92 B. Bruyère, Rapport sur les fouilles de Deir el-Medineh 1929 (FIFAO; Cairo, �930), 22; J. Yoyotte, ‘Religion de l’Egypte

Ancienne �969–�970 et �970–�97�’, Annuaire EPHE SSR 79 (�97�/72), �72.93 Dorner, Ä&L 9, 78–9.94 Dorner, Ä&L 9, 78–9. The position of the wall has not been published.95 Bietak, Tell el-Dabaa II, pl. xlii.a.96 Bietak, Tell el-Dab‘a II, pls xxxvi.a, xxxvii.

Fig. �0. Substructures of the Main Palace at Per-Ramesses.

Page 19: Bietak and Muller

4�

The Topography of New Kingdom Avaris and Per-Ramesses

which carried part of the palace on a lower level than the high sands of the cemetery area. This feature has unfortunately been removed and nothing now remains. As the area south of the cemetery between aEzbet Yasergi and aEzbet Silmy is now empty of archaeological features,97 we may expect an extensive garden and most likely a lake fed by a channel to the south of the palace.98

Mahmud Hamza failed to reveal any architectural features at the periphery of the palace but was able to identify the site as a palace by the discovery of numerous faience tiles with the protocols of kings, especially of Ramesses II. The repeated finds of magnificent tiles of Seti I, Ramesses II and other Rames-sides with inlays of calcite against the blue faïence, originally sold to different museums,99 stimulated the excavations of the Antiquities Service under Mahmud Hamza (fig. ��).�00 Changes of reigns seemed to have brought about changes in the tiles with the royal protocol. On such occasions the old tiles may have been buried in the sand mound.�0� The tiles were picked up by the local population and were sold to antiquities dealers, who sold them to museums. Mahmud Hamza reported the discovery of a factory for such tiles, which seemed to have been attached to the south of the palace.

Of particular importance is the collection in the Metropolitan Museum New York, whose curator W. C. Hayes succeeded in reconstructing throne podiums, steps, windows of appearances, including faience

97 Information kindly provided by E. Pusch.98 Concerning the channel, cf. E. Pusch ‘Vorbericht über die Abschlußkampagne am Grabungsplatz Q IV �997’, Ä&L 9

(�999), 2� (fig. 2, at the left edge).99 C. Boureux, Musée du Louvre, Departement des Antiquités Egyptiennes, Guide-catalogue sommaire (Paris, �932), II; Hayes,

Glazed Tiles from a Palace of Ramesses II; H. W. Müller, Werke altägyptischer und koptischer Kunst: Die Sammlung Wilhelm Esch, Duis-burg (München, �959); H. W. Müller, Die ägyptische Sammlung des Bayerischen Staates (München, �966); H. W. Müller, ‘Bemerkun-gen zu den Kacheln mit Inschriften aus Qantîr und zu den Rekonstruktionen gekachelter Palasttore’, MDAIK 37 (�98�), 339–57.

�00 Hamza, ASAE 30.�0� Dorner, Ä&L 9, 79.

Fig. ��. Faience Tiles from Qantîr (after Habachi, Tell el-Dabaa I, pls 5�A, 52A, 52D–F).

Page 20: Bietak and Muller

42

Manfred Bietak and Irene Forstner-MÜller

sculpture, etc.�02 In the Louvre a whole portal of Seti I could be assembled from the tiles.�03 Similar tiles were collected by the Cairo Museum.�04 The find position of the tiles indicates the approximate size of the platform of the palace and gives us a remote picture of the dazzling blue/yellow tiled interior decoration of the royal residence.

Dorner reconstructs the size of this palace as c.375 × �60 m,�05 which may be the maximum of what is possible. The side-by-side position of the king’s palace and the main temple is also known from the Tuthmoside Period at Karnak,�06 it is a feature known from the story of Apophis and Seqenenrea �07 and it is replicated in the so-called temple palaces at the ‘Houses of Millions of Years’ in Western Thebes.�08

Besides the access via lake S1, the palace seems to have had its main access from the south, where a channel can be seen distinctively to branch off from river F2 .

�09 This channel provides a connection be-tween the palace and the administrative quarter, according to Edgar Pusch.��0 Near this junction of the channel with the river is a recently excavated building, which can be considered as a reception pavilion of the king. It has also a podium structure with a lime plaster-coated façade. Nearby a fragment of a Hittite letter to the Egyptian crown was found.��� A short distance to the east the foundations of a small temple structure show up in the survey picture.��2 It is oriented north-south with the entrance towards the north.

According to the model of Tell el-Amarna one might expect more than one palace in such a splendid residence. One wonders if a large square precinct found by geophysical surveying in the northeast of the town, situated at the F2 channel of the Nile, could be such a palace.��3 Also palatial are elements from the time of Ramesses II, found east of the large palace under the horse stables of the end of the Nineteenth Dynasty with stucco decorations of the ceiling and gold dust on the floor.��4

§ 6 Mansions and Houses

The social hierarchy of the urban settlement of Per-Ramesses has been revealed in an exemplary fash-ion by the geophysical survey of the Pelizäus-Museum Hildesheim which will present its findings and evaluation in the near future. Their survey plots published thus far reveal a system of houses with streets,��5 which are influenced in their orientation by the river courses and channels which transverse the town (fig. �2). Whereas in some parts the system of plots is regular with an element of planning, in the eastern part of the town we already see an organic town developing with winding streets.��6 There are also free

�02 Hayes, Glazed Tiles from a Palace of Ramesses II; W. C. Hayes, The Scepter of Egypt (Cambridge ma, �959), II, 334–9.�03 Boureux, Musée du Louvre, Departement des Antiquités Egyptiennes, Guide-catalogue sommaire, II. �04 Habachi, Tell el-Dabaa I, 229–50, pls 5�–7.�05 Dorner, Ä&L 9, plan 2.�06 M. Gitton, ‘Le Palais de Karnak’, BIFAO 74 (�974), 69–7�. �07 A. H. Gardiner, Late Egyptian Stories (BAe I; Brussels, �932), 85–9.�08 R. Stadelmann, ‘Tempelpaläste und Erscheinungsfenster in den thebanischen Totentempeln’, MDAIK 29 (�973), 22�–42.�09 Pusch, Ä&L 9, 2� (fig. 2, below left).��0 Pusch, Ä&L 9, 2� (fig. 2, below left). ��� E. Pusch and S. Jakob, ‘Der Zipfel des diplomatischen Archivs Ramses’ II’, Ä&L �3 (2003), �43–53; Pusch, in Petschel and

von Falck (eds.), Pharao siegt immer, 259–60.��2 E. Pusch, H. Becker and J. Fassbinder, ‘Wohnen und Leben oder: Weitere Schritte zu einem Stadtplan der Ramsesstadt’,

Ä&L 9 (�999), �66–7 (figs 3–4, lower right).��3 Pusch, in S. Petschel and M. von Falck (eds), Pharao siegt immer, Krieg und Frieden im Alten Ägypten, 24�–2.��4 Pusch, Ä&L 9, �2�–33.��5 Pusch, Becker and Fassbinder, Ä&L 9, 2�, �48–9, �64–7.��6 Pusch, Becker and Fassbinder, Ä&L 9, �64–5.

Page 21: Bietak and Muller

43

The Topography of New Kingdom Avaris and Per-Ramesses

places of large size within the town and some few buildings differ in their orientation considerably from the general orientation. There is also social hierarchy visible. The more substantial house plots, represent-ing the upper class, can be expected in the part of the town which is west and southwest of the palace, especially along the banks of the river, whereas the smaller houses are to be found in the south-eastern part of the town.��7 The types of houses seem to be the same as those we have in Tell el-Amarna. A suburban area with small houses and narrow streets was also found at aEzbet Machali south of channel F2.��8

When digging the el-Samaana canal (now known as el-Didamûn canal), just west of the royal palace, blocks were found from over 25 doorways of at least four royal princes and high officials such as Royal Scribe and Steward of the House of Million Years of Ramesses II, Royal Deputy of Every Foreign Land, First Charioteer of the Great Stable of Usermaaarea-setepenrea, Treasurer of the King of Upper Egypt, three Favourites of the Good God, Fan-bearer to the Right of the King, Chief of the wab-Priests of Sa-khmet, Steward of the House of Ramesses-Meriamun, Kings’ Messenger to Every Foreign Country and Royal Scribe, Chief of Works in Every Monument, Overseer of the Cattle, Chamberlain of the Lord of the Two Lands, Royal Butler and King’s Herald, finally King’s Chief Herald, Chief Commander of the Army, and most likely also the house of the Vizier Paser (fig. �3). There can be no doubt that the highest

��7 Pusch, Becker and Fassbinder, Ä&L 9, �64–5.��8 Forstner-Müller and Müller Ä&L �6, fig. �.

Fig. �2. Geophysical map of an Elite Living Quarter in Per-Ramesses (after Pusch, Ä&L 9, Abb. 2).

Page 22: Bietak and Muller

44

Manfred Bietak and Irene Forstner-MÜller

ranks that needed daily access to the palace lived from the Qantîr Bridge southwards and west of aEzbet Yasergi. The façades of their houses seemed to have lined up at the area where the channel was excavated and their gardens opened up backwards with a wonderful view overlooking the river. One has to ask why the palace of the king did not open up towards the river. The demand to be beside the main temple and to have a processional road from there to another sacred building at the riverbank, probably the house of Amun, made the direct access to the river impossible. But, as the palace was on a high platform, one could overlook the river landscape anyway.

§ 7 Administrative, Industrial and Military Installations

Administrative quarters should be expected in the neighbourhood of the palace. Edgar Pusch identi-fied a building in the south of aEzbet Silmy as an administrative building, possibly even a ‘foreign office’ (fig. �4) because of its similarities with an architectural representation of an office in a tomb.��9 This large building does indeed look like a major administrative office. South of it is a huge courtyard with a central square building with four column bases. It is set against a separation wall, which splits the court into a northern and a southern half.

North of this administrative quarter the Hildesheim expedition has excavated stables for around 500 horses from the late Nineteenth and early Twentieth Dynasty (area Q IV) (fig. �5).�20 They were accom-modated within four units with a big hall for garaging chariots. These royal horse stables were situated along channel K2, which seems to have played a role in grooming and feeding the horses. Installations for training charioteers and quarters for metallurgical installations of a large scale for producing weapons were

��9 Pusch, Becker and Fassbinder, Ä&L 9, �46–52 (figs 6–9).�20 Pusch, GM ��2; Pusch, in Eggebrecht (ed.), Pelizäus-Museum Hildesheim – Die Ägyptische Sammlung; Pusch, Ä&L 9;

A. Herold, Streitwagentechnologie in der Ramses-Stadt: Bronze an Pferd und Wagen (FR 2; Mainz, �999); ‘„Von Pferdeställen und Wagenteilen“: Neuigkeiten über Pferd und Wagen aus der Delta-Residenz Ramses II’, Achse, Rad und Wagen 9 (200�), 4–�7; A. Herold, Streitwagentechnologie in der Ramses-Stadt: Knäufe, Knöpfe und Scheiben aus Stein (FR 3; Mainz, 2006).

Fig. �3. Door-lintel of the House of the Vizier Paser (after Habachi, Tell el-Dabaa I, pl. �8A).

Page 23: Bietak and Muller

45

The Topography of New Kingdom Avaris and Per-Ramesses

found south of the big palace, west of the stables at area Q I. The industrial quarter for arms production dates from the late Eighteenth to the early Nineteenth Dynasty.�2� The charioteer-installations on top of the former date from the early Nineteenth Dynasty to the middle of the reign of Ramesses II.�22 Remains of glass production were found and investigated by the mission of the Pelizäus-Museum Hildesheim in areas Q I, near the faience tile production of Mahmud Hamza, and at Q IV. The sites indicate that industry activities for metallurgy and glass production were concentrated south of the palatial area and the town in order that the prevailing northern winds would carry away the smoke clouds from the kilns.

Despite the fact that the terrain was widely levelled, surface pottery revealed that the enormous fortress of Horemheb seems to have been used until the Ramesside Period, most probably in its original military function. Remains of a huge limestone wall may originate from a temple of the Ramesside Period (§ 4 above).

There were surely more military installations at Per-Ramesses, but only excavation may determine the function of precincts which seem not to be living areas.

�2� Pusch, Ä&L �, 78. �22 Pusch, Ä&L �, 78.

Fig. �4. ‘Foreign Office’ (after Pusch, Ä&L 9, Abb. 2).

Page 24: Bietak and Muller

46

Manfred Bietak and Irene Forstner-MÜller

§ 8 The Harbours

Harbours in a river delta are not normally found on the coast itself, but nor are they a significant dis-tance from the coast, since ships, especially in a period of drought, need marine tides to reach and to leave the harbour safely (§ � above). Per-Ramesses was without doubt a harbour town for seagoing ships. It was a base for the Egyptian navy for its deployment in the Near East. It was ‘The marshalling place of thy chariotry. The mustering place of thy army, the mooring place of thy ship’s troops’.�23 This harbour had tradition. It was also already a major harbour of Egypt during the Hyksos Period. Kamose on his second stela reports hundreds of ships filled with the commodities of the Near East.�24 It seems that there was continuity from the Hyksos Period to the first half of the Eighteenth Dynasty to the Ramesside Period. If the military base with the huge palace compound of the Tuthmoside Period at the west of Avaris at aEzbet Helmy was Peru-nefer,�25 then the continuity of the combination of harbour function, the military base

�23 P. Anastasi III, 7, 5–6; translation R. Caminos, Late Egyptian Miscellanies (London, �954), �0�.�24 Habachi,The Second Stela of Kamose, 37, ll. �3–�5.�25 Bietak, EA, 26.

Fig. �5. Royal Horse Stud in area Q IV (after E. Pusch, ‘Tausret und Sethos II. in der Ramses-Stadt’, Ä&L 9 (�999), abb. 6).

Page 25: Bietak and Muller

47

The Topography of New Kingdom Avaris and Per-Ramesses

and the Canaanite cults till the Ramesside Period becomes understandable. The question is where now in Per-Ramesses may we envisage the main harbour?

We get some idea about the position of the harbour of Per-Ramesses from inscriptions on naos doors in the Pushkin Museum of Moscow from the Ramesside Period.�26 The inscription mentions a wab-priest of Amun, ‘great of victories’, at the harbour of Avaris. The epithet ‘great of victories’ is the epithet of Per-Ramesses. Therefore we learn that Avaris as a toponym was still known in the late Ramesside Period (Twentieth Dynasty) and it was the part of Per-Ramesses which was connected to its harbour.�27 We also learn that there was a shrine of Amun at this harbour.

If the harbour was at Avaris, the most suitable place was the basin east of the Temple of Seth (fig. 2). North and east of it is also the site where Horemheb had constructed his enormous fortress and it is reasonable to assume that it was situated near a harbour. At the eastern continuation of a bay we find a channel (K3) leading to a lake (S3), which would have offered protection to the ships. As we have men-tioned above (see fig. 2), its straight northwestern edge, which might originate from an embankment, is parallel to the northern fortification wall of Horemheb. The palaeogeographers Jean-Philippe Goiran and Hervé Tronchère (CNRS, University of Lyon) positiviely identified sediments from S3 in 2007.

Without doubt, Per-Ramesses had several harbours. We know of a ‘Mooring Place of Usermaaatrea-setep-en-Rea’,�28 which may have been the harbour at the royal palace, probably the lake L1 north of the palace at Qantîr. The mission of the Pelizäus-Museum Hildesheim discovered during its surveys a quay reinforced with limestone at the south-eastern edge of gezira G1, facing gezira G2.�29

§ 9 Cemeteries and Fringes of Per-Ramesses

Per-Ramesses, which was settled throughout the Nineteenth and Twentieth Dynasties and, to some extent, afterwards, must have accumulated an enormous number of tombs. However, one can only con-clude from small samples where those cemeteries had been situated.

In the area of the former fortress of Horemheb, at aEzbet Helmy, excavations have revealed a cemetery of the time of Ramesses II. Tombs of this cemetery were already noticed when the predecessor of the Didamûn canal had been dredged.�30 Finally, during excavations of the Austrian Institute at aEzbet Helmy, a part of a cemetery was found there. A large part was systematically excavated in 2005.�3� More than 80 tombs were found (fig. �6), mainly simple pit burials, some buried in slipper coffins,�32 two contained scarabs with the name of Ramesses II. The general orientation was east–west, some however north–south. Most bodies were buried in a supine position, although some were deposited on their stomachs because they were bundled up in a reed mat and it happened that the front or the back of the body were not recognisable at the time of interment. Remains of slipper coffins, perhaps part of the same cemetery, were found in secondary contexts in area F/I, several hundred metres towards the east.�33 Agricultural levelling largely denuded this part of the landscape. Most of the tombs were scraped away by motor ploughs and the coffin fragments were all that remained on a denuded surface.

�26 Bruyère, Rapport sur les fouilles de Deir el-Medineh 1929, 22; Yoyotte, Annuaire EPHE SSR 79, �72.�27 Bietak, Tell el-Dabaa II, �87–8.�28 Pap. Anastasi III, 2, 6–2, 9.�29 E. Pusch, personal communication.�30 Maspero, ZÄS 23, �2.�3� Bietak and Forstner-Müller, Ä&L �5, 95–8 (figs 27–9).�32 Bietak and Forstner Müller, Ä&L �5, 97; Dorner, Ä&L 9, 83, and another one found 2007.�33 Dorner, Ä&L 9, n. 53.

Page 26: Bietak and Muller

48

Manfred Bietak and Irene Forstner-MÜller

Fig. �6. Ramesside Cemetery in Area H/VI.

Page 27: Bietak and Muller

49

The Topography of New Kingdom Avaris and Per-Ramesses

A large and continuous necropolis can be expected at the western fringe of Gezira Samaana.�34 At Mahgu-bia pottery ushabtis turned up during ploughing and at Geziret el-Baghl large fragments of clay coffins.�35 At aEzbet Gayel half way to el-Didamûn, Labib Habachi found another cemetery with �� tombs,�36 and dated all of them to the Ramesside period. Only one tomb had architecture; it was built of mudbrick walls and reused limestone blocks, one of those blocks bearing the name of Ramesses III gives a terminus post quem for this tomb. Burials Nr. 4 and �0, in slipper coffins, date to the Ramesside Period as well. However, burial Nr. 2 and probably also Nrs. �, 5 can be clearly dated to the late Hyksos Period.�37 In area Q IV a tomb dating to the reign of Ramesses III was discovered by the German mission.�38

The hinterland of Per-Ramesses was famous for its fertility and agricultural products.�39 Wine was even delivered to the large temples at Thebes.�40 Such a vineyard was found to the southeast of the temple of Seth.�4� At the north-eastern outskirts of Avaris (area

A/V) the same system of small pits cut into layers of the abandoned town of the Hyksos Period was found and indicates that this area was not a settlement but a horticultural area.�42 Obviously land formerly used for settlement was now used for agriculture.�43 Archaeological excavations and prospections showed a chain of sandy hills in the south of the town centre.�44 Scarce remains like wells, tree pits and and destruc-tion pits attest settlement activity in the Ramesside period.

Another quarter may be localised at Samaana. Griffith found a heap of limestone and granite chips in the vicinity of the village.�45 Habachi discovered a well with an inscription of Ramesses II which was built of reused blocks from the Amarna period�46 and part of a lintel bearing the name of Paser, a vizier of Seti I and Ramesses II�47 which was reused in a modern house. Either the block of Paser was transferred

�34 Bietak, Tell el-Dab‘a II, 45–6; Dorner, Ä&L 9, 82.�35 Dorner, Ä&L 9, 82.�36 Habachi, Tell el-Dabaa I, 37–8.�37 See also Habachi, Tell el-Dabaa I, 254.�38 Pusch, GM ��2; A. Herold, ‘Ein Kindergrab im königlichen Marstall’, Ä&L 9 (�999), 85–�00; Pusch, Ä&L 9, 30.�39 P. Anastasi III, 2.�–3.9; P. Harris I, 8.2–7.�40 W. Spiegelberg, ‘Bemerkungen zu den historischen Amphoreninschriften des Ramesseums’, ZÄS 58 (�923), 26, 30.�4� Bietak, AÖAW �22, 278.�42 Hein and Jánosi, Tell el-Dab‘a XI, �87–8, Plan 2A.�43 See also Dorner, Ä&L 9, 82.�44 Dorner, Ä&L 9, 8�. Unfortunately there has not been any prospection in the north.�45 F. Ll. Griffith, ‘Gemaiyemi’, in W. M. F. Petrie, Tanis II: Nebeshe (Am) and Defenneh (Tahpanes) (MEEF 4; London,

�888), 45.�46 Habachi, ASAE 52, 48�–2.�47 Habachi, ASAE 52, 479–80; Habachi, Tell el-Dab‘a I, 38, �86–7 Kat. Nr. 52, pl. �8A.

Fig. �7. Slipper Coffin in the Ramesside Cemetery of Area H/VI.

Page 28: Bietak and Muller

50

Manfred Bietak and Irene Forstner-MÜller

from Qantîr�48 or comes from a building nearby.�49 In such a case a quarter for the elite like large villas or estates can be assumed. From the same area came a faience vase bearing the names of Horemheb and Mutnedjmet�50 and faience tiles from a cemetery.�5�

§ 10 Summary and Conclusions

The topography of Per-Ramesses is far from completely explored by geophysical surveying. We may guess that not even half of the city has been assessed. Large parts of the centre of Qantîr are covered by concrete buildings and cannot be recovered in the future. It is completely uncertain how far to the north the town extended. We are also not absolutely certain what features are waiting for us on the opposite bank of Nile branch F1. All that we know from dockets is that vineyards were situated there.�52 Per-Ram-esses proper was situated at the southern part of an enormous Nile-island, formed by the two branches F1 and F2 of the river and was intersected by channels and lakes. The geophysical survey shows a differenti-ated settlement pattern with villas of the upper class settling west and southwest of the palace facing one or the other branch of the river with their gardens. The more modest living quarters can be found in the east of the town, but even there larger houses break the rule. At the southern quarter of the island was the main palace on a high sand-filled platform, overlooking the country. The rooms, the dais of the throne and the window of appearances were encased in multicoloured faience tiles and sculptures. To the north of the palace must have been the major sanctuary of the town, the temple of Amun–Rea-Harakhti–Atum, which must have been connected within a system of processional roads with other important temples of the town, with the Wadjet temple in the north, with the Amun temple in the west and possibly with the Seth temple of Ramesses in the south, and by another road with the Astarte temple in the east.

On the southern part of this island were military quarters such as the royal horse stables, military workshops, workshops for glass and faience industries, barracks of charioteers and administrative quar-ters. The dense settlement system seems to have been maintained only at Qantîr. To the south of the Nile branch F2 the dense settlement system has only been observed at aEzbet Machali. For the rest, the land-scape towards the south seems to have been used mainly for horticulture and agriculture perhaps with scattered country mansions as far east as el-Samaana. An exception was the huge temple precinct of Seth, surrounded by tree groves and the harbour of Avaris east of it, in vicinity to the fortress of Horemheb at aEzbet Helmy.

This part of the town has to be seen in the tradition of a military base of the Tuthmoside Period, which was already endowed with a royal residence and which takes up traditions of the Hyksos residence as a harbour but also in respect to the continuity of Canaanite cults. This place seems to have been the major military harbour of Per-Ramesses. Other harbours can be expected near the palace and southeast of Qantîr island, where a possible quay of limestone has been found. The settlement was constructed on a huge sandy turtleback, elevated beyond the reach of the annual flood. The continuous settled ground of Per-Ramesses is only its northern part, which can be estimated to have covered a maximum of 5 square kilometres. Adding the temple of Seth and the Horemheb fortress one could estimate a maximum of 6 sq. km but not �0 or even 34 sq. km. The rest were gardens, agricultural fields and cemeteries, which were situated in the south of this splendid residence.

�48 Habachi, Tell el-Dab‘a I, 47.�49 Dorner, Ä&L 9, 8�.�50 Habachi, Tell el-Dab‘a I, 47, Kat.Nr. �52, pl. 49A.�5� Habachi, ASAE 52, 488.�52 Hamza, ASAE 30, 43–5.