The Urban Organization of Kolophon and Its Necropoleis: the Results Of

Home , Claros

The Urban Organization of Kolophon and Its Necropoleis: The Results of the 2011–2014 Surveys Author(s): Verena Gassner, Ulrike Muss, Benedikt Grammer, Martin Gretscher and Olivier Mariaud Source: Hesperia: The Journal of the American School of Classical Studies at Athens, Vol. 86, No. 1 (January-March 2017), pp. 43-81 Published by: The American School of Classical Studies at Athens Stable URL: http://www.jstor.org/stable/10.2972/hesperia.86.1.0043 Accessed: 17-02-2017 11:39 UTC

REFERENCES Linked references are available on JSTOR for this article: http://www.jstor.org/stable/10.2972/hesperia.86.1.0043?seq=1&cid=pdf- reference#references_tab_contents You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://about.jstor.org/terms

The American School of Classical Studies at Athens is collaborating with JSTOR to digitize, preserve and extend access to Hesperia: The Journal of the American School of Classical Studies at Athens

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms hesperia 86 (2017) THE URBAN Pages 43–81 ORGANIZATION OF KOLOPHON AND ITS NECROPOLEIS The Results of the 2011–2014 Surveys

ABSTRACT

After a long period of neglect, archaeological research in the ancient Ionian city of Kolophon in western Asia Minor was renewed from 2010 to 2014. As a starting point, a joint team of researchers from Turkey and Austria used the documentation from the 1922 and 1925 excavation campaigns led by the American School of Classical Studies at Athens and the Fogg Museum of Art at Harvard University. The new team explored the city using extensive surveys, geophysical exploration, and LiDAR scans. The results from this research give context to the American excavations, provide new information on the extent of the city and its necropolis areas, and create a preliminary model of the settlement patterns in the ancient city.

INTRODUCTION

Starting in 2010, Christine Özgan from the Mimar Sinan University at Istanbul has directed an archaeological survey at the ancient city of Kolophon in Ionia, in collaboration with Ulrike Muss and Verena Gassner from the Institute of Classical Archaeology at the University of Vienna.1 As this survey represents the first major research project at the site of Kolophon in almost 90 years, after an excavation campaign under Hetty

1. Our warmest thanks for this contribution, as well as Anton Bammer thank the anonymous reviewers of collaboration go to our colleague and (archaeologist), Erich Draganits (geolo- Hesperia for their suggestions. The friend Christine Özgan, who also gist), the students Carina Hasenzagl, survey project “New Investigations at helped us with all administrative ques- Andreas Hochstöger, Jasmin Scheifin- the Ancient Site of Kolophon in Ionia,” tions. We also thank the representatives ger, and Sarah Wanek; Torsten Riese directed by Ulrike Muss, is funded by of the Turkish Government, Umut and Sebastian Pfnorr (archaeologists- the Austrian Science Fund (FWF, Doğan (2010), Ayşe Akman (2011), geophysicists of Posselt Zickgraf P 24763-G 21). Olivier Mariaud’s Cansu Yilmaz (2012), Onur Yıldırım Prospections [PZP], in 2013). Luigi participation was partially funded (2013), and Seval Ekintan (2014). Vecchio (epigraphist, University of by the LUCHIE of the Université Thanks go to Mario Rathmanner of Salerno) joined our team in 2014. de Grenoble. The “Introduction” and Airborne Technologies for making the Kristina Klein provided valuable aid “Methodology” sections were written LiDAR scan possible. Participants of with the preparation of the plans and by Verena Gassner and Ulrike Muss. the campaigns were the authors of this photographs for publication. We also

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 44 verena gassner, ulrike muss, et al.

Burial ground

2 2

5 0

0 City wall 0 Riverbed Yaren Tepesi Modern street/path Modern building North East Necropolis Contour line (10/50 m) "Valley Cemetery"

Sivrikaya Tepesi Sırtlan Tepesi

Mound 1 "Cemetery B"

0 25

Kalabacık Tepesi Kuru Dere 2 0 0 0 0 Halil Ağa 2

200 250 Değirmendere

300 Northeast Area

350 1 5 0

2 0 0

Southwest Area 400 Kavakli Dere

5 Stoa 0 Yurtpinar Tepesi

4 50 Metroon Kuru Dere Acropolis Ambarkaya Tepesi 500

2 0 0

5 50

6 00 "Cemetery A/Cemetery Hill"

Southwest Necropolis Çamönü 650

South Necropolis 700 Taşyalak Tepesi

2 0 5 Southern Territory Burial Ground 35 0

0 500 m Kale Tepesi

Figure 1. Topographic plan of Goldman and Carl Blegen in 1922 and a short season in 1925, a large ancient Kolophon with burial 2 variety of research questions prompted our renewed interest. The main grounds and modern buildings. focus in 2010 was to study the ancient city’s extent, development, and B. Grammer spatial organization over time by using modern survey methods during the annual field campaigns. After an initial orientation within the modern topography, the Turkish team decided to concentrate its activities on the 2. For a summary of the excavation acropolis, while the Austrian team started research in the plain and the history, see Holland 1944. Smaller res- adjacent hills where the necropoleis are located (Fig. 1). cue excavations had been carried out in Before our new investigations started, the best known part of ancient the 1990s and early 2000s by the Archaeological Museum of Izmir, but Kolophon was its acropolis, as this area had been the main site of the have not been published, with the excavations carried out by the Americans in 1922 and 1925. Their results exception of some short remarks in 3 were published in detail by Leicester Holland. These excavations unearthed Şahin 2008. four major architectural features that are easily distinguishable on the plan 3. Holland 1944. Some of the struc- (see Fig. 3, below): the so-called Metroon on the eastern side of the acropolis tures mapped on the accompanying plates were not described in the article; hill comprises a stepped altar on a huge terrace and two small additional they were reserved for a separate pub- rooms to the south. The elongated central plateau of the acropolis is lication, which in the end was never dominated by a large, L-shaped stoa with a northern and western wing; each realized. Restudying these unpublished wing contains a number of differently sized rooms. To the immediate north structures is a part of the new Kolo- and especially to the west of this stoa, residential quarters were uncovered, phon project and will be presented with dwellings articulated around courtyards and along paved streets.4 within the framework of the final project monograph. Farther to the west, a larger bathing complex was found. Finally, parts of 4. For a discussion of this residential other, smaller structures of undetermined function were encountered by quarter, see Hoepfner and Osthues the Americans in trenches farther up the hill. 1999, pp. 280–291.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 45

When the Turkish team first examined the state of the ruins excavated by the Americans on the acropolis of Kolophon, in particular those in the area of the stoa, they found that these structures were generally in poor condition.5 In the following years their research concentrated in particular on the Metroon6 and then shifted to the documentation of several monuments scattered in the surrounding area that can be understood as rock-cut altars in the tradition of Phrygian monumental altars, probably also dedicated to Meter.7 The Austrian team started in 2010 to retrace the course of the Early Hellenistic city walls, originally documented by Schuchhardt and Wolters in 1886,8 which are—apart from the buildings on the acropolis—the only built structure of the city still visible.9 In 2011 we focused on a study of the modern village of Değirmendere. We also initiated studies of the archival material from the 1922 and 1925 excavations now held by the American School of Classical Studies in Athens.10 From 2012 to 2014, we explored the extent of the ancient city with surveys inside and outside the fortification walls. This work comprised extensive and intensive surveys, geophysical prospection, and documentation of microtopographical features, as well as interviews with field owners and tenants. In 2013 an airborne laser scan (ALS or LiDAR) was conducted to guide the survey efforts in the city and provide a secure topographical basis. This article offers a preliminary summary of the five seasons of field research conducted at Kolophon by the Austrian team concentrating in particular on the results of the survey inside the fortification walls and on the necropoleis.11

METHODOLOGY OF THE SURVEY

Initially, we were encouraged by the relative absence of modern construction in the territory surrounded by the Early Hellenistic city walls. Nevertheless, some problems arose during our work, necessitating a modification of our approach. Because our time frame was constrained by administrative concerns, our field season coincided with the agricultural use of some fields, and the temporary installation of solid greenhouses in others. A few other selected areas were inaccessible for a variety of reasons. The situation was further complicated by the intensive activities of illegal diggers (Fig. 2). Until now their work had continued largely unhindered and so they felt disturbed by our research activities. As a result our theoretical strategies had to be abandoned, making the selection of survey areas more random.12 In this context we were quite fortunate that the owners and tenants of the

5. Bruns-Özgan, Gassner, and Muss 8. Schuchhardt 1886. ment, nor our results of a survey in the 2011, pp. 214–218, figs. 23–26; Özgan 9. For the city walls, which are not territory of Kolophon (Özgan et al. et al. 2012, pp. 268–269, figs. 1, 5, 6. the focus of this project, see in particu- 2013). The work of the Turkish team 6. Bruns-Özgan, Gassner, and Muss lar Bruns-Özgan, Gassner, and Muss on the acropolis will be presented in a 2011, pp. 219–222, figs. 27–31; Özgan 2011, pp. 203–213, figs. 4–22. separate study. et al. 2012, pp. 269–271, figs. 3–5; Öz- 10. These studies are currently being 12. For a strategic choice of survey gan et al. 2013, pp. 196–197, figs. 3–7; carried out by Martin Gretscher and areas, see, e.g., Bintliff et al. 2007, Özgan et al. 2014, p. 98, figs. 1, 2. Ulrike Muss. Our warmest thanks go p. 44, with examples from Thespiai 7. Özgan et al. 2014, p. 98. For to Natalia Vogeikoff-Brogan for her and Haliartos. For the use of modern the topic in general, see Scherrer continuous support. fields as units, see Burgers 2012, 2000, p. 60; Bammer and Muss 2006; 11. In this preliminary report we do p. 16. Berndt-Ersöz 2006. not discuss a possible Geometric settle-

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 46 verena gassner, ulrike muss, et al.

Figure 2. Hole dug by looters in field NO07. Photo U. Muss fields in the northeast and southwest, which make up the central part of the Early Hellenistic zone of the city, allowed us access so that we could proceed with our work.13 Both extensive and intensive surveys took place within and outside the city walls.14 During the intensive survey, most of the fields were documented with line-walking by not more than five people, using simple hand-held GPS units. The intervals were between 5 and 10 m. On some occasions, we also laid out rows of quadrants, 5 × 5 m each, and one person collected the finds in a maximum time of 10 minutes. The relatively small amount of pottery found in most areas inside the city walls was striking.15 For the cultivated fields—at the time of the survey mostly stubble fields—one possible explanation for this may be the fact that they are plowed only to a depth of 10–15 cm, thus often not coming into contact with ancient contexts. We were also surprised by the juxtaposition between the mostly very small and badly preserved sherds from the surveys and the finds left by illegal diggers in situ at the margins of their “excavations.” Only here do we have better preserved and larger fragments of pottery, allowing the development of a more precise chronological spectrum. As all available modern maps are relatively imprecise, Schuchhardt and 13. We thank these field owners Wolters’s sketch from 1886 was still the most accurate plan of Kolophon and tenants, especially Osman Ceviz, when the project started in 2010.16 As a first step, using Google Earth Aydın Özkan, and the members of the and later orthophotos photographed during the LiDAR scan, we defined Karaoğlu family for their support and individual working areas by giving short codes to the fields in the supposed patience. area of the city, including the necropoleis and part of the surrounding 14. In this article we focus on the surveys carried out inside the fortifica- territory. These areas are basically oriented on the basis of the topography tion walls. For the methodology of and the modern situation of the territory. To the east of the acropolis a urban surveys, see in general Vermeulen broad plain sloping slightly from southwest to northeast is surrounded by et al. 2012. the Early Hellenistic city wall. Today the area is cut by a modern road in a 15. At this point in our research, north–south direction, running from the village of Ataköy via Çamönü to exact calculations on the frequency of pottery per m2/ha are not yet available. Değirmendere and from there to Menderes. This modern road dominates 16. Schuchhardt 1886. The plan the space and has been considered as a marker for the definition of the published in Holland 1944 likewise topographical areas. The area to the northeast of this modern road has been uses the Schuchhardt sketch as a basis.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 47

TN03 Kalabacık Tepe TNO50 City wall J TNO49 TN01 TNO62 TNO58 TNO56 City wall (reconstructed) TN02 DO04 TNO14 Riverbed DO06 TNO15 Field TN34 TNO57 TNO13 NO31 Prospection: Ground-penetrating radar TNO60 TNO07 TNO12 DO05 TNO06 Prospection: Magnetic DO23 K H TNO01 TNO05 TNO16 Contour line (20 m) NO32 NO30 TNO03 TNO02 G TNO08 SW41 TNO18 Halil Ağa NO29 TNO04 SW02 TNO09 SW03 NO16 TNO11 SW08 SW07 SW04 NO01 TNO10 TNO17 NO02 NO18 L NO08 Ambarkaya Tepe SW06 SW05 SW09 NO21 SW12 NO03 SW10 SW14 Değirmendere DS04 NO04 SW13 NO10 NO17 SW39 SW15 NO20 SW11 NO22 SW48 SW16 SW40 NO09 NO13 NO19 SW38 SW36 NO23 NO14 NO15 SW35 NO05 NO11 SW37 SW42 NO24 TSO74 SW17 NO12 NO25 SW47 NO06 F SW31 SW30 NO26 KabaklidereSW32 Stoa NO07 SW46 SW29 SW18 SW33 NO27 SW19 SW28 NO28 B Kurudere SW27 SW25 A01 SW20 TSO01 SW21 TS01 SW34 TS42 Metroon SW49 SW24 SW45 SW26 SW44 E TS02 TS41 SW22 A SW23 TS40 Acropolis D TSO03 SW43 TS03 TS08

TSW06 TS07 C TS06 0 200 m

Figure 3. Aerial photograph (2013) classifi ed as NO,17 and the part to the southwest as SW. Area SW covers of Kolophon, including plan of the slopes of the acropolis hill and is confi ned to the south and northwest American excavations, labeled fi elds, by sections of the city wall (Fig. 3). The part to the northeast incorporates and areas of geophysical prospection. the Kalabacık Tepesi in the north, the Ambarkaya Tepesi in the east, and B. Grammer; georeferencing of American excavations B. Zickgraf (PZP) is surrounded by section E–J of the city wall. The sectors outside the area of this region of the Early Hellenistic city have been marked with the identifi cation T (for territory). TS (territory south) covers the sector lying between the southeastern city wall and the modern village of Çamönü. In ancient times a road proceeding to Klaros and the sea passed through 17. NO stands for “Nordost” in this area. The village of Değirmendere was divided into four sections, German and corresponds to “northeast” 18 in English. For more information about designated as DO (Değirmendere East), DS (Değirmendere South), DW the terminology of the city wall sec- (Değirmendere West), and DN (Değirmendere North). tions, see Bruns-Özgan, Gassner, and As a great part of ancient Kolophon—especially the area of the Muss 2011, p. 200, fi g. 1. necropoleis—is densely overgrown with pine trees and maquis, survey from 18. DO = Değirmendere Ost, i.e., the air became an essential component of our study. A better understanding Değirmendere East. of the organization of the ancient city was made possible by an airborne 19. We thank Mario Rathmanner, Airborne Technologies, Wiener Neu- laser scan, carried out and processed by the Austrian company Airborne stadt (Austria), for his contribution to Technologies.19 This made it possible to create a reliable topographic map, our investigation. but in particular facilitated the exploration of the extent of the necropoleis,

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 48 verena gassner, ulrike muss, et al. as will be seen below. Exemplary geophysical prospection—both magnetic survey and ground penetrating radar (GPR)—allowed first insights into the urban organization of the city in the plain, in particular in the northeastern area.20 During the survey campaigns we collected between 5,000 and 6,000 items of pottery, including fragments of bricks and tiles that sometimes represented the majority of the finds.21 As study of these finds is still ongoing, we here refer only briefly to some of the difficulties the material presented for chronological classification. Their mostly fragmentary state of preservation is normal for material originating from surveys, but the original surface of most sherds was so badly damaged that their diagnostic quality was lost. In the case of supposed black-glazed pottery, for example, the gloss was not preserved. This fact also complicated the classification of painted pottery of the 8th–6th century b.c. In many fields, mostly fragments of tiles were found, which are difficult to date as detailed studies of tiles in Asia Minor are largely missing. While it seems probable that tiles with black or red paint date from the 7th to the 5th century b.c., the question remains open as to whether tiles without paint generally belong to the 4th century b.c., or can also be dated earlier.22

THE URBAN ORGANIZATION OF THE TOWN IN THE PLAIN

One of the principal aims of the current project was the exploration of the extent of the urban area of Kolophon in a diachronic way. As the city walls belong to the Early Hellenistic period, we expected that they included the city of the late 4th and the 3rd centuries b.c. Evidence from the acropolis showed that at the end of the 4th century b.c., part of the previous residential quarter was changed into a public area by the construction of a stoa and the Metroon, the only sanctuary ever discovered.23 The famous inscription from the Metroon (dated between 311 and 306 b.c.) tells us that the newly built fortification walls also included the ancient agora in the plain, thus indicating a construction date for the new city at the end of the 4th century b.c.24 In 1922, the American excavators opened a small number of trenches in this plain as well; they can be approximately located to the northeast of the acropolis hill. The main result here was the discovery of a major structure some 115 m long, which was probably a public building.25 Unfortunately, the topographical position of this structure was not precisely

20. The geophysical prospection survey, so only 200 items were collected. the chronological sequence see Bruns- was carried out by Sebastian Pfnorr 22. For a helpful discussion of this Özgan, Gassner, and Muss 2011, and Torsten Riese (PZP, Marburg, class and the chance to have a look at pp. 218–219, with previous bibliogra- Germany). Benno Zickgraf was the Archaic (and mostly painted) mate- phy. Verena Gassner, Ulrike Muss, and engaged in the processing of data, rial from Klaros, we are grateful to Martin Gretscher contributed to this which will be published in more detail Nuran Şahin. For a painted tile from section. at a later date. Ephesos, dated to the 5th century b.c., 24. Meritt 1935, p. 371; De La 21. Most of the material came out see Scherreret al. 2006, pp. 130–132, Genière 1994. in 2012 and 2014. The 2013 season was 146. 25. Holland 1944, p. 95. dedicated mainly to the geophysical 23. For a recent reconstruction of

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 49

Figure 4. Traces of walls and tile fragments in an illegally dug hole in field NO12.Photo U. Muss

recorded and therefore cannot be retraced. Two other trenches north of this structure also produced a number of more modest walls, belonging to at least two buildings of unclear function.26 We began surveying in the plain expecting to find part of the Early Hellenistic settlement, though it remained to be discovered how the settlement had grown and prospered. We were also interested in learning how the transfer of the Kolophonian population to Ephesos during the reign of Lysimachos influenced the development of the city.27 Similarly it was unclear if and in what way the development of the settlement had occurred in the plain between the Kabaklidere and Kurudere Rivers during the Archaic period. Further evaluation of the geophysical survey is still ongoing, but continues to be elusive as many factors have hindered the measurements. We began by surveying the fields located in the plain between the Kabaklidere and Kurudere Rivers in the northeastern quarter of the ancient city (NO03–NO05, NO09–NO10; see Fig. 3). Here the existence of houses was known from open trenches in these fields. Remains of walls were also visible in several trenches opened by illegal diggers both in the plains and in the forested area near the slopes of the surrounding hills (Fig. 4). The geophysical survey confirmed the existence of a coherent 26. This can be understood from sketches in the original excavation settlement pattern. In particular in field NO10 we found six parallel lines diaries, which are kept in the Archives reflecting walls oriented north-northeast to south-southwest (Fig. 5). The of the American School of Classical average distance between these walls is 21 m; in the case of the second and Studies at Athens. the third line from the east the distance is only 4.7 m, perhaps indicating 27. For the history of Kolophon the presence of a street or a broad corridor. The space between the first and a summary of ancient sources, and the second walls provides evidence for the inner organization of a see Rubinstein 2004, esp. pp. 1077– 1080. house. Two small rooms are visible in the northern part, accompanied by 28. For Kolophon, see Holland a corridor on the western side. In front of the two rooms to the east we 1944, pl. XI; Hoepfner and Osthues find more walls, perhaps indicating the area of aprostas and a courtyard. 1999, pp. 285–290, with figures. For Houses with similar ground plans have been excavated on the acropolis of Priene, see, e.g., Wiegand and Schrader Kolophon, and also in Late Classical and Hellenistic contexts at Priene.28 1904, pp. 285–297; Rumscheid 1998, pp. 140–149; Hoepfner and Osthues To the west of the sixth line there follows an area of about 60 m where 1999, pp. 338–351; Hellmann 2010, no anomalies were visible. It is possible that this free space is due to the p. 63, fig. 75. modern situation or it could be of archaeological significance. If the latter, it

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 50 verena gassner, ulrike muss, et al. could be interpreted as an open place. In field NO03 we found the remains of another insula/house, which shows the same width of 20 m. Also striking is a square structure with a length of about 40 m farther east of the area we investigated, which is visible in the geomagnetic prospection. Its large dimensions could point to an important public or sacred building, which probably belongs to a period preceding the houses. Both in NO03 and in the southeast of field NO05 the ground-penetrating radar also provided evidence for walls that could be reconstructed as smaller houses, which do not display the regular layout of the houses described above. As their orientation is different, it seems probable that they belong to another period of the settlement, though a different function cannot be excluded. The very modest dimensions of about 5 × 6 m as a maximum and the irregular building plan are reminiscent of Archaic houses like those at nearby Klazomenai.29 Small structures were also found in the second area chosen for geophysical prospection, via both magnetic survey and ground-penetrating radar. Fields SW07 and SW09–SW10 (Fig. 6) are situated to the south of Halil Ağa Tepesi, near field SW06, where we carried out our first intensive survey in 2010. Though we had been given access to SW06 in 2010, the owner denied access in 2013. This forced us to carry out the prospection on the fields to the west, where modern interference had to be expected. Consequently, the results from these fields were not very conclusive. There are a few structures, mainly in field SW10, which might belong to small houses, maybe of an earlier date, or which may reflect artisanal activities, but their orientation did not correspond to any of the other known orientations. The nearly rectangular structures in the southwestern part of the area remain enigmatic. They might represent archaeological phenomena like pits, but show a striking correspondence to lines interpreted as modern drains. Regarding the results in the northeastern fields, the existence of an overall street grid with insulae stands out. The width of a house can be approximately 21 m, but only in one case could the position of a street be assumed, so the overall organization of the quarter remains unclear. The inner organization could resemble Late Classical and Early Hellenistic houses like those of Priene.30 The fact that evidence for the inner walls of the houses was found only in the far east and in the west is intriguing. We could not decide if this meant that only these inner walls survived, or if it reflects the fact that the interiors of only some of the houses had been constructed, while for others only the outer walls were erected. This strange evidence perhaps can be connected with the abandonment of Kolophon under Lysimachos. During the 2014 survey, it was not possible to walk all the fields in this area, as we were not granted permission to work in the fields immediately connected to those in the northeastern area of the city. Therefore we explored the southwestern part of the territory immediately east of the Kabaklidere River, to ascertain if the settlement area also continued here.31 29. Ersoy 2007, e.g., pp. 156–159, These gently sloping terraces had been investigated by the Archaeological figs. 2, 3. 30. Hoepfner and Osthues 1999, Museum of Izmir in the late 1990s, in two areas where residential buildings pp. 341–345. See also references in were found, parts of which are still visible today (Figs. 7, 8). Most probably n. 28, above. the walls present two phases, the latter of which show large, finely worked 31. Muss et al. 2014b.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 51

NO10

NO03

NO04

NO09

NO05

0 40 m

Figure 5. Interpretation of GPR on the NO fields with likely archaeological structures (red). B. Grammer; PZP SW07

SW09

SW06

SW10

SW11

Figure 6. Interpretation of GPR on the SW fields with likely archaeological structures (red). B. Grammer; 0 40 m PZP

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 52 verena gassner, ulrike muss, et al.

Blocks/walls City wall Riverbed Fields Fields surveyed in 2014 SW42

SW31/1-4

SW31/5-6

SW31 SW31-8 SW31-7 Southwest Area survey fields

SW29

Kabaklidere

Museum of Izmir excavation SW33 SW33-4

SW33/1-3

Acropolis

SW49

Museum of Izmir excavation 0 100 m ashlar blocks at the corners, which seem to be spolia from older structures. Figure 7. SW area surveyed in 2014. Fields SW31 and SW33 revealed evidence for an ancient division of the B. Grammer area, recognizable by sporadic ashlar blocks with clearly worked edges in obvious alignment. The locations of all these features were recorded,32 and their correlation with the walls identified by the geophysical survey of 2013 in the northeastern part of the city showed that they corresponded to the orientation of these structures. This provides evidence for an overall grid- system organization of the urban area from at least SW31 to NO10, in a 32. The geodetic surveying was north-northeast–south-southwest direction. handled by Kamran Citak and his One of our crucial research aims was to determine whether the Archaic collaborator (Mira Harita, Selçuk, city extended as far as the northern part of the valley, where the modern Turkey).

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 53

Figure 8. Remnants of residential village of De irmendere is situated. We assumed that this area was part of buildings in field SW49, excavated ğ the Archaic settlement in the plain because the favorable environmental by the Museum of Izmir in the situation of the village provides good water supply during the whole year 1990s. Photo U. Muss and is oriented toward the fertile plain of Menderes. From 2011 forward, we documented stone and marble spolia as well as many fragments of ancient tiles built into the houses of the village. Older inhabitants confirmed that the building material was collected in the area of the village, where at least some of the houses were built before 1922. However, it could not be proven whether the area belonged to the ancient settlement; even our surveys of some empty spaces in the village did not reveal any diagnostic indications. As a result of our investigations in the plain, we know now of at least two periods of settlement, but a more detailed chronological framework still needs to be established. The typological classification of the houses led to the assumption that the small houses are evidence for the Archaic period, while the grid system with the insulae can be connected with contexts of the Early Hellenistic period. The direction of the street grid coincides with the orientation of sector G of the fortifications, belonging to the Early Hellenistic period (Fig. 9). When we georeferenced the stoa building on the acropolis, we discovered that it also followed this direction, so that a contemporaneous date of the city walls, the stoa, and the buildings detected in the plain seems most probable.33 The construction of the stoa can be connected with the reorganization of the town at the end of the 4th century b.c.34 Preliminary analyses of the diagnostic pottery demonstrated that the majority of these finds belonged to the Archaic period, while examples of 33. The georeferencing was done by the late 4th/early 3rd century b.c. are rare. Widely dispersed classes and Benno Zickgraf of PZP. types belonging to the Hellenistic period, such as moldmade bowls or 34. Holland 1944, p. 169. black-glazed echinus bowls, are totally absent, although they should have

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 54 verena gassner, ulrike muss, et al.

Riverbed City wall City wall part G City wall (reconstructed) Grid system Walls in the Northeast Area (Geophysical prospection/survey)

Walls/domestic buildings in Southwest Area (excavations Museum of Izmir/survey)

Acropolis, Stoa (American excavations 1922)

0 200 m been easily recognizable even if poorly preserved.35 Though it is too early Figure 9. Grid system of the Early to draw far-ranging conclusions, this might indicate that a large part of Hellenistic(?) period. B. Grammer; PZP the plain both in the northeast and the southwest was densely inhabited in the Archaic period, while the reorganization of the settlement in the 4th and the early 3rd centuries b.c. seems to have come to an abrupt end.

THE NORTHERN NECROPOLEIS

As is the case for the settlement, our knowledge of the burial areas of Kolophon is scanty because all the material from the early excavations led by Goldman was lost shortly after the excavation.36 Only one grave was published by Bridges in 1974,37 an LH IIIB–C tholos tomb, which had been partially robbed in antiquity. In 2011 a preliminary study of the archival material

35. For the frequent appearance of warmly thank James Wright for hav- pp. 8, 29, n. 38. these types in the survey materials of ing drawn his attention to this article 37. Bridges 1974. This documen- nearby Metropolis, see Meriç 2014. during his stay at Bryn Mawr in 2004. tation has been partially scanned and 36. For the archaeological, aca- On the brief mention of the necropo- is available on the internet site of the demic, and geopolitical context of leis of Kolophon, see Fowler 1922, American School (http://www.ascsa. Goldman’s excavations in Kolophon, esp. p. 259; Picard 1922, p. 540 (n. 2), edu.gr/index.php/archives/Excavation- see Davis 2003. Olivier Mariaud, who 547 (n. 2), 729; Goldman 1923; Lori- Records#Kolophon). authored this section, would like to mer 1950, p. 38; Hanfmann 1953,

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 55

concerning the Geometric tumuli was published.38 But the main body of the archives of the 1922 and 1925 campaigns remained useless without direct observations in the field. These records include the Northeast (“Geometric”) Necropolis, as well as the two others, the South (“Classical”) Necropolis, and the North Necropolis (Değirmendere Cemetery/“Cemetery B,” i.e., the area of the Late Bronze Age tholos; see Fig. 1). In 2010, some preliminary topographic identifications were made during the first survey season.39 In 2013 and 2014, these were confirmed, modified, and expanded with systematic comparisons between archival material, field observations, and GIS and LiDAR data. In addition to general funerary topography, we also collected and documented any funerary features that could help to clarify the chronology of the necropoleis, which was far from precise.40 Finally, we focused on the documentation of the most impressive features, two large burial mounds designated as Mounds I and II in the Blegen and Goldman archives. Sketches were drawn and photographs taken in order to compare them with the evidence in Blegen’s notebooks. In sum, our goals were to integrate the evidence from the American excavations in their actual geophysical context and, in general, to improve our knowledge of the burial places of Kolophon.41

The North Cemetery (Cemetery B or

DeĞirmendere Cemetery) The North Cemetery, identified as the “tholos necropolis” and also named Cemetery B or Değirmendere Cemetery in the Blegen and Goldman notebooks, was first located in the area of field DNO01–04.42 However, a closer observation of the topography led to the conclusion that, even though a funerary use of this field cannot be excluded, it did not correspond to the places excavated in 1922 and 1925, where Goldman uncovered the tholos. According to the notebooks and most of all to some photographs unknown in the 2010 season, it seems clear that the place where the tholos and a dozen other graves were found was situated in a flat area, bordered by a large hill on its north-northeast side (Fig. 10:a). This area corresponds exactly to the small plain bordered at the south-southeast by the hills of Halil Ağa and of Kalabacık Tepesi, while its north-northeast side is defined by the Yaren Tepesi (Figs. 1, 10:b).43 Though the precise locations of the many trenches excavated by Goldman and Cox have been obliterated by time, we were able to clarify their topographical context. Even though the total number of graves of the North Cemetery is not yet clear,44 because some of them surely date to the Late Bronze Age, the

38. Mariaud 2011. participated in this survey of the the images in Figure 10:a and 10:b 39. Bruns-Özgan, Gassner, and North Necropoleis: besides Mariaud, shows that the workers probably dug Muss 2011. Erich Draganits and Ö. Naz Bitrim in an area a little closer to the foot- 40. The collected pottery material (North Necropolis), Benedikt Gram- hills than where we stood when we could not be stored, but was photo- mer and Martin Gretscher (Northeast took our picture. graphed and sometimes drawn on the Necropolis). 44. The number is between 12 spot before being redeposited. The 42. Bruns-Özgan, Gassner, and and 15, but since the excavations are quantities were too low to warrant sta- Muss 2011, pp. 226–227. described in two different notebooks tistical analysis. We thus limited our 43. One can clearly distinguish in two different years (see Goldman work to a basic documentation and the sharp top of the hill, free of vege- 1922; Cox 1925), the risk of confusion diagnosis of the few uncovered sherds. tation, which remains similar to the is high. 41. Several members of the team 1925 picture. A close examination of

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 56 verena gassner, ulrike muss, et al.

Figure 10. North Necropolis, Değir- mendere Cemetery (a) with men in foreground searching for graves; (b) view of North Necropolis, looking north. (a) Photo courtesy American School of Classical Studies at Athens, Archives, ExcRec, Kolophon Archive, b photo C156; (b) photo O. Mariaud temporal context for the tholos must also date to that time period.45 The other graves however seem to be associated with a later context, probably Geometric, but they may also have a date as late as the 7th century b.c.46 The area is a vast alluvial plain, formerly crossed by the eponymous Değirmendere (literally “river of the mill”), which became channeled at some point in the modern era. Today, the location of the necropolis comprises a farm and its adjacent fields. The intense modern agricultural exploitation has heavily disturbed the archaeological evidence. During our survey, thick straw covered most of the area, while the narrow path coming

45. For instance, “Grave 1” of Cox’s with those of the Baitinger IA2 Group the funerary material have been lost, notebook contained a “stirrup jar.” (Baitinger 2001, p. 9, pl. I:8–14). It including the arrowhead itself, this dat- 46. This low dating of one of the should be considered though, that the ing cannot be confirmed. The only graves (grave 12) is based on the pres- absolute dating of these small bronze remaining evidence is a rough sketch ence of an arrow or javelin head in items is based on a broad time frame, plan of the arrowhead in Goldman’s the grave goods (M. Gretscher, pers. and any attempt at precision becomes notebook (1922, p. 73). comm.). Gretscher associates its shape hypothetical. Since all other objects in

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 57

Figure 11. (a) Worked stone in TNO31; (b) Holland 1922, p. 6, grave 15; (c) Cox 1925, p. 8, grave 1. (a) Photo O. Mariaud; (b, c) photos courtesy American School of Classical Studies at Athens, Archives, ExcRec, Kolophon Archive (source image modified) c down from the Yaren Tepesi and the small dried riverbed separating TNO30 from TNO11 were used as dumping grounds. 47. Worth mentioning here are Consequently, although undeniable evidence of ancient activities was at least two metallic slag, which indi- recorded,47 no evidence remained to clearly signify the presence of ancient cate the possibility of an undated metal tombs. This should not be surprising. Already in 1974, Bridges noted that workshop located in the area, at least there were few signs of the graves in that area.48 Forty years later, only a few one large ancient wall, many fragments haphazard stones located on the foothills of the Yaren Tepesi, close to the of tiles, and coarse wares. These dis- coveries will be detailed in the final abandoned house shown on Fig. 10:b (TN31–TN32), can be associated with publication. early material. One piece in particular (Fig. 11:a) shows stunning affinities 48. Bridges 1974, p. 266. with the cover slabs described by Holland and Cox (Fig. 11:b, c), which

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 58 verena gassner, ulrike muss, et al. belonged to one of the early grave sites. In fact, these slabs covered most of the tombs excavated by the American team in the North Necropolis. This covering technique, together with the apsidal shape of the pit, recalls some of the Protogeometric graves of Lefkandi.49 Such individual pit graves covered with large slabs and containing only a few grave goods are quite common at this early period in the Aegean area, including in Ionia and Caria. Besides the confirmation of the location of the excavation area in the field adjacent to the Yaren Tepesi, this inquiry demonstrates that very few elements could be associated with an ancient funerary use. Only the forthcoming publication of the data from the Goldman and Cox notebook, examined in their true topographic frame, will shed new light on the general funerary disposition of the North Cemetery, providing a long-awaited context for the tholos tomb.

The Northeast Necropolis (Geometric Necropolis or Valley Cemetery) Before 2010, examination of the American notebooks alone did not provide any secure location for the Geometric tumuli graveyard, otherwise known only from scanty mentions in reports and books. The topographical indications were limited and the analysis of the 15 photographs concerning this necropolis remained insufficient without a basic knowledge of the terrain. This explains the tentative nature of the southern location proposed in the preliminary account of the Geometric graves presented in 2007 and published in 2011.50 Also in 2011, Muss and Gassner suggested a new location for the necropolis along the northeast margins of the city (TNO61), which was immediately convincing.51 Not only did a road sign bearing the name Kurudere echo a name frequently used in the notebooks but, as we shall see, the form and size of the main monument of this necropolis corresponded almost exactly to the largest of the excavated tumuli, namely Mound I. Analysis of the LiDAR imagery from 2013 and the field observations recorded in 2013 and 2014 confirmed this identification. This then enabled us to concentrate our efforts on the documentation of some graves (Mounds I and II). Unlike the North Cemetery, in the Kurudere cemetery the marking of the graves—large tumuli—were still visible in the landscape, despite their poor state of preservation.

Mound I The first tumulus is a huge, irregular, but mainly conical structure that rests on the low slope of the eastern flank of the Yaren Tepesi. It is, therefore, lower on its south-southwest side than on its north-northeast side, and this profile has remained unchanged since it was first documented in the 1922 pictures (Fig. 12). But with a size of 6–7 m in height and 40 m in width, the actual tumulus is both lower and wider than it was a century ago, when it measured 9 m and 30 m, respectively. Aside from natural erosion, it seems clear that 49. See Popham and Lemos 1996, pp. 60, 80, pls. 18, 28. human disturbance—either scientific or illegal—has also contributed to a 50. Mariaud 2011. substantial subsidence of the structure. This heavy perturbation makes the 51. See Bruns-Özgan, Gassner, and convergence with Blegen’s description of 1922 even more difficult, if not Muss 2011, pp. 227–229.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 59

Figure 12. North trench, Mound I, Kurudere Necropolis. Photo courtesy American School of Classical Studies at Athens, Archives, ExcRec, Kolophon Archive, photo C181

Figure 13. Northeast Necropolis, Mound I, view of central pit looking east. Photo O. Mariaud

impossible. The mound is actually largely cut by seven main pits, above all a large central crater (9 × 9 × 4 m; see Figs. 13; 14:a, b:K); Figure 13 also depicts the southern berm, the conical accumulation of earth at the mound’s foot (the “cone of dejection”) resulting from erosion, and the bedrock reached by the illicit diggers. Besides this huge crater, six more holes of various sizes have disemboweled the tumulus (trenches 1–6; see Fig. 14:b). Most of them cannot be easily matched with those from Blegen’s excavations mentioned in his journal or visible on photographs because of our ignorance of the level of backfilling in 1922 and the intensity of illegal digging.52 The trenches of the inner tumulus do, however, allow some observations. 52. One exception could be trench 1 For instance, we note that the inner filling of the tumulus is composed north of the tumulus, which might correspond, by its size and positioning, to of alternative and irregular layers of orange-yellowish to very light brown the “North trench” excavated by Ble- earth, at least in the upper and middle parts of the trench. At the bottom, gen’s team (see Fig. 12). orange earth dominates. We can also consider the presence of some

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 60 verena gassner, ulrike muss, et al.

Figure 14. Northeast Necropolis: Relative height of terrain (a) ground plan of Mounds I and II; High : 0.5 (b) sketch of Mound I, both super- Low : -0.5 imposed on LiDAR visualization Contour line: 0.5 m (Local Relief Model/Hillshade). Drawings O. Mariaud and B. Grammer

Mound 2

Mound 1 Kurudere

Kolophon

0 40 m a

Transect 1 B 0 10 m

Transect 2 Bk-1

A Bk-2 K 1

Bk-4

4 Bk-3

B3-2 B5-1 B3-1 3 B3-3 2 B

A b

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 61

Figure 15. Northeast Necropolis smooth stones of average size (0.30 × 0.20 × 0.15 m; 0.20 × 0.15 × 0.10 m), Mound I, trench 5, profile B5-1, horizontally set in the profile and which seem to follow the same orientation detail of the stratigraphy. Photo as the surface (Fig. 15). According to Blegen, this “highly colored red and O. Mariaud yellow hard earth” was found around 0.80–1.00 m below the tumulus surface (in its 1922 state) and represents the remains of the mudbricks that should have constituted the main body of the mound.53 However, the irregular aspect in the stratigraphy of these layers, together with the presence of small stones, leads to the conclusion that this compact earth was not placed as mudbricks, but more likely represents a random distribution of a great quantity of earth, probably carried in individual baskets considering the small size of each colored area in the profile. The survey of Mound I revealed very few pottery sherds of good quality, undecorated coarse ware, with red-orange or orange clay. These sherds were not diagnostic, but correspond to Blegen’s description. Along with the decorated fine wares he regularly mentions some coarse ware or undecorated fine-ware pots with this type of clay (red-orange, with no mica), generally at some distance from the graves themselves. Finally, the presence in one of the berms, trench number 5 located right at the south-southeast of the central crater (berm B5-1), of a relatively high quantity of charcoal or carbonated residue should be noted (the largest 53. Blegen 1922, p. 150 (south pieces are 0.02 × 0.02 m). This residue appears as a layer in the middle trench), p. 172 (north trench). of the stratigraphy along a slightly horizontal orientation more or less 54. See, for instance, Blegen 1922, following the surface inclination, at a depth of approximately 0.60–0.70 m p. 174 (north trench). The possibility and for a distance of approximately 2.50–2.60 m. Again, this observation of these residues being the remains of calcinated roots is unlikely since these matches Blegen’s explanation that carbonated residues were noticed when residues are irregular in size and do not approaching the pyres of the primary cremations and were obviously linked form a continuous line. to them,54 although the residue we saw is not the remains of the pyre itself.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 62 verena gassner, ulrike muss, et al.

Mound II Less impressive than Mound I, Mound II was recorded at a location of 100 m north of the former.55 A regular heap of ca. 20–22 m diameter was first detected in the area by the LiDAR imaging and was then found on the ground (Fig. 14:a). Far less damaged than the first mound, the structure is also considerably lower. Its real height is difficult to estimate because of the field declivity, which in a few dozen meters goes from a steep foothill to the flat riverbed of the Kurudere. Set in the lowest lying area, we can postulate a height of 1–2 m, which is the same height recorded by the American excavators.56 One cannot help but notice the huge discrepancy in the height of the two tumuli, which is unlikely to be due to natural erosion or anthropic factors. Other probable tumuli, very close in size and form with Mound II, are visible in the vicinity, which makes the impressive dimension of Mound I even more intriguing. There is no doubt that Mound I, which is unique to this day in Kolophon and elsewhere in Ionia, must have been a remarkable landmark in the necropolis, and therefore the area that provides access to Kolophon from the northeast. No material of any sort was found on or around Mound II. Among the few surface features (numbered F1–9), some can be associated in one way or another with observations in the American notebooks. For instance, hidden under dense bushes, a large pile of stones of the “river boulder” type57 can be related to those of the grave E enclosure. They could have been gathered at a short distance from the grave.58 But it is mainly four large green schist stones (F9)59 that attract our attention for their resemblance to some cover stones of one grave (grave A) associated with the mound (Fig. 16). Unlike Mound I, Mound II displays enough affinities with the notebook descriptions that we 55. Blegen 1922, pp. 148–149. may cautiously superimpose our surface plan on the sketch plan from the 56. Blegen 1922, p. 183. 1922 excavations in order to provide a tentative reconstitution of the grave 57. Dimensions of stones: 0.10 × structure rigorously documented by our predecessors. 0.15 m; 0.30 × 0.25 m; 0.50 × 0.25 m. 58. See also Mariaud 2011, figs. 3, 5. Despite our strides with Mound II, we still face many challenges. It 59. Stone dimensions of F9 are the is our hope that the location and the general topography of the two main following: F9-1: 0.80 × 0.30 × 0.25 m; funerary areas in the northern outskirts of the city (excavated by Goldman F9-2: 0.75 × 0.50 × 0.20 m; F9-3: and her team) can be considered secure. Our documentation of Mounds I 0.60 × 0.50 × 0.20 m; F9-4: 0.65 × and II will allow a full publication of the data from 1922 and 1925, which 0.40 × 0.20 m. will, at last, secure the place of Kolophon in the burial customs of early 60. The interpretation of the LiDAR data was carried out by Benedikt Ionia between the Late Bronze Age and the Early Archaic period. This Grammer. In the field, the survey was will in turn lay the foundation for future excavations in the area. primarily conducted by Grammer and geoarchaeologist Erich Draganits, with Martin Gretscher, Olivier Mariaud, THE SOUTHERN AND SOUTHWESTERN and Seval Ekintan supporting the sur- NECROPOLEIS vey efforts on several occasions. All participants contributed greatly to the successful interpretation of the data and Many of the features of the Southern and Southwestern cemeteries were provided pleasant company during the elucidated with the help of the LiDAR survey. This survey was conducted entire field trip. Erich Draganits and through a high-resolution digital elevation model (DEM) generated by Michael Doneus offered valuable advice means of an airborne laser scan.60 The aim of the survey was to document concerning the methodological and technical aspects of the ALS interpre- all of the monuments visible on the surface in the hills surrounding the tation and the ground check. Grammer city, including the previously known burial ground of the South Necropolis thanks all the aforementioned for their (Figs. 1, 17). The airborne laser scan produces a highly detailed DEM assistance.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 63

F9-2 F9-1

F9-4

F9-3

Figure 16. (a) Northeast Necropolis, Mound II, slabs of schist; (b) grave A, Mound II, before opening. (a) Photo O. Mariaud; (b) photo courtesy American School of Classical Studies at Athens, Archives, ExcRec, Kolophon Archive, photo C191 b

with a resolution of 0.5 m.61 Vegetation can be virtually removed from the resulting DEM, making it a very practical and efficient method for discerning surface structures even in a densely wooded and overgrown area. As such it was an appropriate method for surveying the hills around Kolophon, where intensive fieldwalking is impractical due to the inaccessibility and scale of the terrain. The DEM was interpreted after various filters had been applied to the data, and possible archaeological 61. For the method in general, see features were marked for a ground check.62 In 2014, during a two-week Doneus et al. 2008; Bofinger and Hesse survey, a team of two to three people, using the LiDAR interpretation for 2011. 62. The methodology of the LiDAR orientation, documented about 150 archaeological features, among them interpretation will be published in more 95 burial mounds, 10 grave terraces, 18 graves, and 17 walls or structures detail in a later article. of unclear function (Fig. 17). The majority of features are located within

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 64 verena gassner, ulrike muss, et al.

0 200 m Metroon

Acropolis

Kolophon

TSW05-1

Southwest Necropolis

TSW13-1

TSW10-61 South Necropolis

TSW10-65

Kurudere

TSW10-6 TSW10-7

Kale Tepe Ridge

Cave Grave Burial mound Grave terrace Wall Structure City wall City wall (reconstructed) Riverbed Contour line (20 m) TSW10-52

Figure 17. South Necropolis, LiDAR shading, and archaeological features. B. Grammer

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 65

the South and Southwest Necropoleis on the hills of the Kale Tepe and Taşyalak Tepe of the Sivridağ mountain range. The analysis of the documentation and the find material is still in process, but a preliminary overview of the results can be offered here.

South Necropolis: Kale Tepe The campaign’s survey work focused on the previously known South Necropolis located on the Kale Tepe south of Kolophon. During previous extensive surveys of the area, around 30 burial mounds had been mapped, and two mounds that had been excavated by the Museum of Izmir in 1999 were later documented in an architectural survey in 2013.63 The results of the LiDAR interpretation showed that the extent of the burial ground had been underestimated and a complete survey of the area was deemed necessary. An easy-to-follow pathway leads from the southern plain of the city to this area, where the Museum of Izmir excavated several domestic buildings. As mentioned above, the Austrian team conducted a fieldwalking campaign in 2014 that went to the highest point of the city wall on the Kale Tepe, where a large structure is situated (TSW05-1). Curiously, the structure was not marked or mentioned during the first exploration of the city by Schuchhardt and Wolters at the end of the 19th century, even though they most likely visited this part of the city wall, as they marked a nearby grave.64 Closer examination showed TSW05-1 to consist mostly of a large heap of smaller rubble stones with a large pit in the middle, probably the result of illegal digging. The connection to the city wall could not be determined due to the badly preserved state of the structure and the thick vegetation. It remains unclear if the structure was a tower associated with the city wall that had been completely covered by refuse, or if it was possibly a burial mound that was incorporated into the defensive structure at a later point. From the location of this structure onward the pathway continues upslope on the eastern side of the Kale Tepe; beyond the intersection of a modern access road, the first burials of the South Necropolis become visible. The South Necropolis can be divided into three separate areas: a northwestern and southeastern part, as well as the ridge of the Kale Tepe.

South Necropolis: Northwestern Part The ridge of the Kale Tepe is characterized by a protruding stone ledge. An abrupt rise in slope separates the lower northwestern part of the burial ground from the higher Kale Tepe Ridge. The topographically prominent point after the steep rise of the hill is marked by TSW10-61 (see below). The northern area is located on both the eastern and western slopes of the Kale Tepe (Fig. 17). It is possible to move from one side of the hill to the other at certain points on the northern part of the hill. On the western slope no clear pathway is visible, but the terrain forms a relatively flat, natural terrace, which facilitates movement. The grave monuments consist of 13 burial mounds, six grave terraces, and 10 simple graves on both the western 63. Muss et al. 2014a; Grammer forthcoming. This section was written and eastern slopes of the Kale Tepe. The burial mounds are composed of by Benedikt Grammer. earth and smaller rubble stones, with sizes ranging between 3 m and 9 m 64. Schuchhardt 1886, p. 402. in diameter. The height was difficult to determine due to erosion, later

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 66 verena gassner, ulrike muss, et al. disturbances by looters, and the general slope of the hill, but usually it did not exceed about 1 m. The graves are built as rectangular terraces that were either freestanding or set against the slope of the hill, forming small grave districts of up to 6 m along one side. Simple grave markings consist of setting of smaller stones on top of a simple pit dug into the earth. The find of roof tiles next to illicit dig sites suggests that tiles might have been used for marking, covering, or defining the grave. It appears likely that additional graves with only a simple marking of stones are located in the area, but were not visible due to a thick layer of pine needles covering the entire ground and an overgrowth of bushes. All graves had visible signs of disturbance by thieves. The illicit excavations provided an unobstructed view of the mostly destroyed graves, which apparently were either simple pits reinforced or covered with stones, or partially cut into the rock and partially built of small stones. Sometimes large stone slabs were found in the vicinity, and one was still lying on top of a looted grave and partially covered it. In other cases no traces of constructed graves could be found. The size of the illegal trenches suggests that they contained inhumation burials, but since no traces of the burials could be found, the existence of cremation burials cannot be excluded. It seems very likely that the American excavations on “Cemetery Hill” in 1922 took place in this part of the South Necropolis. In Goldman’s notebook the location of Cemetery Hill is described as being to the east of the acropolis ridge and to the west of a tower in the city wall. The comparison of a photograph taken in 1922 showing Cemetery Hill and a contemporary photograph of Kale Tepe taken from the Metroon on the acropolis displays a close resemblance.65 A notation in the excavation notebooks, that a grave exists on a higher ridge to the east, indicates that the American excavations probably started at the lower western slope of Kale Tepe and then gradually moved upslope. Later during the excavation Goldman mentions a grave “overlooking Tratscha” (modern Çamönü), suggesting that at some point graves on the top ridge or eastern slope of the hill were excavated.66 The type of tombs described by the Americans include burial mounds of about 5 m in diameter, consisting of stone walls filled with rubble, and simple inhumation graves covered by stones. The pits were either lined with stones or cut into the bedrock and supported with smaller stones, where necessary. Both types were covered with stone slabs. Small finds in the graves included black-glaze pottery and an Early Hellenistic or Classical coin; these finds led to the designation of the necropolis as the “4th-century necropolis.”67 Although the sketches of the excavation could not be matched directly to any of the tombs found during the survey, the description fits the general types of tombs that were documented very well. Due to the lack of other datable evidence, these small finds from the American excavations remain our only lead to the proposed Classical or Early Hellenistic chronology of the burial ground. With the exception of roof-tile fragments found next to the illegal pits exposing the graves on the northern part of the area, no new finds were discovered in this area.

65. Muss et al. 2014b. 67. Goldman 1923, p. 68. 66. Goldman 1922, pp. 43, 53.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 67

South Necropolis: Southeastern Part The path which leads into the burial ground from the city continues along the slope of the Kale Tepe, but for a distance of about 180 m no graves were discernible either in the LiDAR or during the ground check until the path crosses the riverbed of the Kurudere (see Fig. 17, above). East of this riverbed a portion of the gently ascending hill is densely clustered with burial mounds. This area is encircled by the path, which continues eastward for 100 m but then turns south and finally west toward the ridge of the Kale Tepe, thus circumscribing the burial ground. Two tombs were found in a rather isolated position: a single burial mound was located to the west of the riverbed and 60 m south of the path, while tomb TSW10-65 (see p. 69, Fig. 19, below) was found at the northernmost edge of this part of the burial ground. The remaining 32 burial mounds, a single grave terrace, and one rock-cut grave are all found either directly adjacent to the path or at a distance of only a few meters from it. Judging from the way the path encircles the burial ground, the close connection to the burial mounds and the clear visibility of the path in the LiDAR data, it seems very likely that the path has been used since antiquity. Whether the same holds true for the northern part of the path within the urban area is uncertain, due to modern reinforcement with rubble from its origin in the plains up to the point where it reaches the city wall. It would, however, seem likely that the same path has connected the domestic quarters of the city in the plains to this burial ground since antiquity. If the burial ground was still in use after the construction of the city wall in Early Hellenistic times, a gateway of sufficient dimensions to accommodate funerary processions would have to be expected at the point where the path reaches the city wall, but no evidence for such a construction was found. This might be due to the badly preserved state of the wall and the large structure TSW05-1 located here, but could also indicate that the South Necropolis was no longer in use after the construction of the fortifications. The tombs found in the southeastern part of the South Necropolis were almost exclusively burial mounds with diameters between 6 and 10 m, but occasionally reaching up to 15 m. Similarly, their heights often exceeded those of the mounds of the northern part of the South Necropolis, where the largest mounds reach only a height of up to 4 m. Sometimes the tombs were grouped together in clusters of two or three, but are otherwise relatively evenly spread over the area of the southeastern part of the burial ground. The two burial mounds excavated by the Museum of Izmir (TSW10-6 and TSW10-7) and documented in detail in 2013 (Fig. 18) can be found in the southeastern part. These two excavations offer a closer look at the way mounds were constructed. The foundation of TSW10-7 consisted on a carefully built krepis (a load-bearing wall, as opposed to a simple circle of stones), while TSW10-6 was constructed of stacked-up irregular or roughly broken stones of various sizes. Both had a filling of mostly smaller stones and reddish earth. Within TSW10-7, three graves were visible, and they exhibited the most common type of construction: they were partially cut into the rock

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms rock-cut grave TSW10-6

TSW10-7

68 verena gassner, ulrike muss, et al.

Figure 18. South Necropolis: (a) Burial5 m mounds TSW10-6 and TSW10-7, ground plans; rock-cut grave rock-cutTSW10-6 grave TSW10-6 (b) burial mound TSW10-7; (c) burial mound TSW10-6. (a) Drawing J. Scheifinger, S. Wanek, and B. Grammer; (b, c) photos B. Grammer TSW10-7 TSW10-7

5 m 5 m a

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 69

Figure 19. South Necropolis, grave TSW10-65. Photo B. Grammer

and partially built of small stones. All were probably originally covered by stone slabs, as one such slab was still found lying on top of a grave. There were also several other slabs in the immediate vicinity. No sign of a grave could be found within mound TSW10-6. The construction of other mounds in the burial ground was not as readily discernable. Though these other mounds were covered by small, eroded stones, it was not clear whether they were laid down as the top layer of the fill, or discarded by looters. As far as we could tell, the other burial mounds in the area mirrored the construction of these two, although it appears the regularity of the mantle walls of TSW10-7 is rather exceptional. A peculiar variation of tomb construction was present in the grave already mentioned above (TSW10-65), which apparently took advantage of a natural cave several meters in depth (Fig. 19).68 A series of regular, small steps were cut into the rock around the opening of the cave, perhaps serving the purpose of either holding a platform, a cover for a grave, or both. Erosion and illicit digging have heavily damaged the area surrounding the grave, but several larger, regular blocks found to the north suggest the existence of a krepis forming a half-circle or circle around the cave, which may have been covered with earth forming a burial mound of ca. 10 m in diameter. During the 2014 survey no pottery fragments of any kind were found anywhere in the area other than a number of small modern pots probably connected to the extraction of tree resin. In 2013 during a thorough examination of refuse from the 1999 excavation a rim of a krater emerged. While it can probably be dated to the 5th/4th century b.c., it cannot be connected with certainty to any particular grave or activity and is of limited use for dating the structure. At the moment, given this very scarce dating evidence, this part of the South Necropolis can only be tentatively associated with the Classical period.

68. In Ionia the use of natural caves Melie (Kleiner, Hommel, and Müller- for burials is otherwise only rarely doc- Wiener 1967, p. 161) and Samos umented, but examples are known from (Tölle-Kastenbein 1976, pp. 102–103).

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 70 verena gassner, ulrike muss, et al.

Figure 20. South Necropolis, burial Southern Territory: Kale Tepe Ridge mound TSW10-61. Photo B. Grammer Following the path to the west one reaches the ridge of the Kale Tepe (Fig. 17). The ridge is defined to the west and east by two small structures built of rubble for a length of about 1 km, forming a 15–25 m wide plateau. The structures bordering the plateau show up clearly in the LiDAR data, but were found to be harder to trace on the ground due to a thick cover of pine needles and heavy erosion. These small walls do not measure more than about 0.5 m in height and consist of loosely stacked, mostly small stones. They do not seem to have served any practical function as a fortification or fence, but merely marked the path along the ridge. Due to time constraints and the length of the structures it was not possible to clean and document them in their entirety.69 Likewise the stratigraphic connection to other monuments situated on top of the ridge was not determined. Starting at the northernmost point of the ridge, we located six large burial mounds. The location at this outpost, with the terrain falling off steeply to all sides other than the southern one, gave the large burial mound TSW10-61, ca. 15 m in diameter, a particularly prominent and elevated position (Fig. 20). The mound was mostly destroyed by looters, the refuse of which covers the structure and surrounding area. On the northern side, parts of a wall were visible, perhaps forming the krepis of the mound. Other walls on the southwestern and southeastern sides might have been part of a confinement wall or small grave district which adjoined the mound. The addition of later graves to existing burial mounds by adjoining a krepis is a 70 configuration known from Samos and Larisa, but an intensive cleaning of 69. A similar demarcation of a hill’s this structure would be necessary to determine the stratigraphic relationship ridge with rubble walls was discovered between the walls and the mound. The mound itself was composed of in the LiDAR data on the ridge of the irregular rubble, which have been dispersed by looters over the entire area. Taşyalak Tepe to the west, but this was The large number of stones suggests that the mound once was considerably even harder to trace on the ground, so documentation was postponed due to higher than the ca. 1.5 m it measures in its current state. time constraints. About 30 m to the south of TSW10-61 another group of three large 70. Samos: Boehlau 1898, p. 33; burial mounds was found. Each of the mounds is situated on top of the Larisa: Meyer-Plath 1940, p. 109.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 71

Figure 21. Cave on the ridge of Kale Tepe. Photo B. Grammer

eastern ridge and the structures marking the plateau seem to be missing in this part, maybe indicating that they were dismantled and used as building material for the mounds. Due to the lack of a visible stratigraphic relationship between mounds and walls it was not possible to confirm their sequence. The mounds were constructed similarly to TSW10-61, but only reached diameters of 8–12 m. Two more mounds of similar proportions and building technique can be found another 250 m to the south. They are grouped closely together and occupy almost the entire width of the ridge. A singular, badly preserved wall of rubble could be seen protruding from the eastern burial mound, but a direct connection to the mound could not be determined (we speculate that it might be part of a dromos). The only finds at the mounds were large roof tiles lying on top of the refuse left by looters, which probably served as the original grave coverings. Following the ridge and the accompanying structures a very large cave eventually presents itself (Fig. 21). It is at least 8 m deep, but could be considerably deeper, as the entire depth of the cave is not discernible from above. Wall-like structures similar to those mentioned above and

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 72 verena gassner, ulrike muss, et al. accompanying a ca. 20 m wide path were also found to the northwest of this cave, but they do not seem to lead to any particular point of interest or archaeological feature. The arrangement of these marked pathways meeting at this point and delineating an open square of about 2,000 m2 suggests that the cave served some sort of function, perhaps as as a grave similar to TSW10-65, or possibly a sanctuary.71 However, no artifacts of any kind could be found in the immediate area, so any interpretation is purely hypothetical. An examination of the cave itself was not possible during our survey due to the lack of climbing tools. About 60 m further to the southwest and situated at the highest point of the crest, another small plateau is isolated from the ridge of the Kale Tepe. On the southern end of this plateau and protecting it from the steep slope to the south, the remains of a large, ca. 15 m long wall were found (TSW10-52; Fig. 22). It was carefully constructed out of mostly small stones and is badly preserved, but reaches a height of up to 1.5 m. Where the wall has collapsed, its refuse covers the section that is still standing and makes it hard to determine the actual width, but it probably exceeded 1 m. While a very large amount of rubble can be found in the vicinity on the slopes of the hill, the material visible is not sufficient to reconstruct any large stone structures. No stones showed any signs of decoration or workmanship, but it seems possible that the large wall served as a foundation for a mudbrick structure that was covered with roof tiles, large fragments of which can be found in great number on the plateau. Due to the good visibility of the surrounding area and the large size of the wall, the structure possibly served as some sort of fortification or watchtower. An interpretation of the structure as a possible sanctuary connected with the cave to the north also needs to be considered. While not clearly visible on the ground, the LiDAR data shows that the stones which define the Kale Tepe ridge extend up to the end of the plateau, indicating a connection between the marked pathway and structure TSW10-52 (Figs. 17, 23). The only finds were fragments of roof tiles and a single piece of lead, possibly carrying the inscription of a letter (∆) on one side.

71. A cave connected with a sanc- (Meriç 1982, pp. 38–40), Pergamon religious centers of the city, which is tuary is known in Ephesos and has (Nohlen and Radt 1978), and Aizanoi likewise only found in Athens. Addi- been associated, albeit not conclusively, (Naumann 1967). A general connec- tional and sometimes earlier sanctua- with the cult of Apollo (Karwiese 1995, tion of the cult with both isolated ries of Meter would therefore not be pp. 215–219), Demeter (Soykal- mountain sanctuaries and caves has surprising and the topographic situa- Alanyalı 2005), and Athena (Scherrer been noted before, based on both the tion of the cave in the hills would 2000, p. 66). An association with caves location of the sanctuaries and numer- be a highly suitable venue for such a was also proposed for the Pan-Ionian ous epitheta (Körte 1898, p. 94; Lam- cult. However, besides the lack of finds cult of Poseidon Helikonios in his brechts 1970, p. 138; Nohlen and at the cave, this seems so far to be function as an “earth-shaker” (Herda Radt 1978, p. 70). In Kolophon it is contradicted by the absence of rock- 2013, p. 427, n. 22). Another goddess yet uncertain if the Metroon on the cut altars, thrones, or niches, which closely connected to rocks of various acropolis had a predecessor before the are usually intrinsic for the cult of kinds is Kybele/Meter, whose sanctua- 4th century b.c.; a publication of the Meter and which can be found near ries are usually attested in the form of findings of the American excavations the Metroon on the acropolis (Özgan rock-cut altars or niches; in general, in 1922 and 1925 is in preparation by et al. 2013). A closer examination of see Naumann 1983; Graf 1985; Roller Martin Gretscher. The sanctuary on the the cave with suitable equipment 1999. Caves in particular served as acropolis is unusual in the way it was would be needed for a verification sanctuaries in nearby Metropolis embedded into one of the political and of this hypothesis.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 73

Figure 22. Kale Tepe plateau, with the rubble wall TSW10-52. Photo E. Draganits

Figure 23. Kale Tepe ridge, with wall TSW10-52, looking north from the plateau. Photo E. Draganits

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 74 verena gassner, ulrike muss, et al.

Southwest Necropolis Another previously unknown burial ground was explored on the eastern slope of the continuation of the acropolis ridge (Fig. 17). Despite their relatively close proximity, this burial ground was apparently unconnected to the South Necropolis. Instead, it is most easily reached from the ridge of the acropolis by means of an access road. No pathways between this burial ground and the South Necropolis could be found, while the pathways leading up to the burial grounds from the city and within the necropolis areas themselves were clearly visible within the LiDAR data. It is possible that an earlier connection between the two burial areas was destroyed with the construction of the modern access road. Since no traces of pathways can be seen at all in the LiDAR it appears more likely that the Southwestern Necropolis was accessed only from the direction of the acropolis. Within the Southwest Necropolis the predominant structures that were found and confirmed were 21 burial mounds, 3 graves marked by simple stone settings, and a single terrace. The mounds in the burial ground differed substantially from those within the South Necropolis; they were primarily heaps of earth of up to 8 m in diameter, mostly set against the sloping hill in the form of a half-circle. These mounds were partially covered and reinforced by rubble. Sometimes the setting of stones on the downslope part of the mound resembled a crudely built krepis wall, but in general the use of stones was limited. All burial mounds were looted, and sometimes the illegal digging in this area took the form of regular, large, and deep trenches, with dimensions up to 3.5 × 3.5 × 2 m, indicating the removal of very large objects by looters. Clay or stone sarcophagi are known to be used frequently in combination with burial mounds in ancient Ionia and could be expected in Kolophon as well.72 Given that not all sarcophagi were decorated and therefore most valuable as complete objects, illicit excavations should have produced some remains of destroyed or broken sarcophagi. No such fragments of any kind were found in the vicinity of any of the graves, even in trenches with evidence of recent looting. Unique within the burial ground was a single terrace (TSW13-1) measuring 8 × 8 m (Fig. 24). It is framed by badly preserved built walls, which included several large stone blocks of up to 1.2 m in length. Several heaps of smaller rubble stones and a larger stone lying on the terrace possibly marked the presence of graves.

OTHER ARCHAEOLOGICAL FEATURES IN THE TERRITORY OF KOLOPHON

In addition to the features documented within the burial grounds, several other monuments were discovered along the Kale Tepe and Taşyalak Tepe ridges in the south and in the hills surrounding the city to the north (Yaren Tepe) and northeast (Ambarkaya Tepe). Most of these can probably be 72. Philipp 1981, pp. 155–157; Mariaud 2007, pp. 58–66. Grammer interpreted as isolated funerary monuments, in particular burial mounds would like to thank one of the anony- and grave terraces. While very few artifacts were discovered in the burial mous reviewers of this article for high- grounds in the immediate vicinity of the city, the structures in the outlying lighting this issue.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 75

• 350.96 rock • 350.96 rock rock • 350.96 rock rock rock •348.06 •348.06 •348.06

grave marking? 351.14 gravegrave marking? marking? • 348.37 351.14351.14 • • • •350.15 • 348.37• 348.37 •350.15•350.15

• • 350.82 350.82• 350.82 349.88 348.51 •349.88• • 348.51 • •349.88 • 348.51

349.13•349.13• 0 0 3 m 3 m 349.13• 0 3 m

Figure 24. Southwest Necropolis, grave terrace TSW13-1, ground territory could be associated with Greek and Roman ceramics, the analysis plan. Drawing B. Grammer of which is still pending.

Southern Territory Burial Ground Of particular interest was an isolated group of two small burial mounds and at least two simple grave markings made of rubble, found even further south of the South Necropolis on the Kale Tepe. This limited cluster of burials is located about 1.7 km south of the city wall, but no clear pathways connect it to either the South or Southwest Necropolis. As usual, the graves were looted and fragments of roof tiles were found in the remaining refuse. A small plateau to the east of these graves (TSW11-1) was encircled with a low, heavily eroded wall of rubble stones measuring only 0.5 m high. No building structures could be found on the thickly overgrown plateau, but the dimensions of the area defined by the encirclement, 50 × 25 m, suggest that the entire plateau served some sort of function in antiquity. To the west of the small burial ground we discovered another plateau, on which traces of low rubble walls could be found (TSW11-6). Looters heavily disturbed the entire area. The location of these graves and structures at a considerable distance from the city and its official burial grounds, in a rather inaccessible part of the hills, could be an indication of settlement activity within the southern territory of Kolophon.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 76 verena gassner, ulrike muss, et al.

TSW10-7 4 5 6 7 12

Kolophon Old Smyrna 0 5 10 m

Figure 25. Burial mounds of Kolo- phon and Old Smyrna. B. Grammer, DATING AND REGIONAL CONTEXT OF THE after Miltner and Miltner 1932, p. 159, BURIAL MOUNDS IN KOLOPHON fig. 88

A prevalent feature of the Kolophon necropoleis is the similarity of the burial mounds to those appearing in other Ionian and Greek cities on the western coast of Asia Minor. Geometric burial mounds are so far only known from Kolophon’s Northeast Necropolis. An early, distinct burial mound tradition can also be found in Caria from the Protogeometric period onward,73 but a regular use of burial mounds in the western Anatolian region starts in the Archaic period.74 In almost all Ionian cities isolated burial mounds have been found, but currently burial grounds with larger clusters of mounds are known only in Kolophon, Klazomenai, and Smyrna. The occurrence of these burial mounds in northern Ionia has often 73. Radt 1970, pp. 217–235; been associated with a Lydian or Anatolian “influence” in Ionia, but the Carstens 2008, pp. 71–93; Henry 2009, Geometric mounds in Kolophon clearly predate the Lydian examples, and pp. 81–102. 74. Aeolid: Eckert 1998, §4.4.2. find close parallels in contemporary burial mounds in Athens, Halos, and Ionia: Mariaud 2007, pp. 260–265. 75 Naxos. Only a very few tombs in Ionian cities, like the Tomb of Tantalos Lycia: Hülden 2006, pp. 109–135. in Smyrna, show a definite typological resemblance to the Lydian tombs Lydia: Il 2009. with the use of a dromos, a grave chamber, and “phalloi” as a crest.76 Close 75. Mariaud 2011. parallels to the mounds from the South Necropolis of Kolophon are those 76. Miltner and Miltner 1932. On found in particular in Old Smyrna (Fig. 25), Klazomenai, and Larisa. There the question of Lydian/Anatolian and Greek burial mound traditions and the are at least 16 examples known from Klazomenai, 35 from Old Smyrna, meaning of burial mounds in a regional and a large, but unknown number from Larisa.77 The majority can be dated context, see the forthcoming publica- to the Archaic period, more specifically the 6th century b.c., but their use tion of a conference talk given by Bene- continues until at least Hellenistic times. No extensive documentation of dikt Grammer. these mounds exists, but a clear resemblance to the Kolophonian mounds 77. Old-Smyrna: Miltner and Milt- ner 1932, pp. 158–162; Akurgal 1950, can be seen in their size, construction, the existence of a krepis, their location pp. 79–82; 1983, pp. 58–59. Klazom- on sloping hills adjacent to the settlement, and the use of rock-cut and cist enai: Hürmüzlü 2004, p. 78; Aydın graves for the inhumation burials. In Larisa the occurrence of tiles in the 2006, p. 99. Larisa: Meyer-Plath 1940. construction of graves has also been noted, just as we have seen at some The exact number of burial mounds in of the burial mounds in Kolophon. In light of the regional trends and the Larisa is not reported. The total num- typological similarities, it seems probable that some of the many burial ber of graves discovered is around 100, several of which appear to be relatively mounds in Kolophon date to the Archaic period. A dating based purely low burial mounds marked by either on typology is severely limited, however, in particular with such long-lived circular or rectangular stone settings or and relatively simple grave forms. low walls.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 77

CONCLUSIONS

After many years, new archaeological research in the Ionian city of Kolophon in western Asia Minor was resumed in 2010. In a series of field campaigns the city and surrounding areas were explored through extensive surveys, geophysics, and LiDAR data. Our recent investigations made it possible to contextualize the results of the original American excavations, to gain new information on the extent of the city and its necropoleis, and to establish a model of the settlement’s development. Concentrating on the area of the city situated in the plain, geophysical survey revealed the existence of a regular street grid as well as the inner structure of single houses. Their inner organization resembles that of Late Classical and Early Hellenistic domestic architecture, while the orientation of both streets and houses closely matches that of the stoa on the acropolis and section “G” of the city walls in the plain. The geophysical data now supports the previously known epigraphical evidence for an overall reorganization of the city in the Early Hellenistic period.78 A pottery survey demonstrated that the majority of finds dated to the 6th and 5th centuries b.c. Examples from the 4th and early 3rd century b.c. were rare, and no ceramics were found from the later Hellenistic or Roman periods. The reorganization of the city at the end of the 4th century b.c. appears to have ended abruptly only a few years after it had begun. This evidence seems to correlate with the literary account of Lysimachos’s deportation of the inhabitants of Kolophon, which would have terminated major settlement activity in Kolophon. In some of the areas examined in the northeastern sector the GPR readings also indicated structures of approximately 5 × 6 m with various orientations. We suggest interpreting these structures as the remains of Archaic houses. By integrating data from our geophysical and pottery surveys, we can suggest a considerable residential settlement of the area at the foot of the acropolis hill during the Archaic and perhaps even the Classical periods, with a second, but short-lived floruit of this region of the city in the Early Hellenistic period. During the American excavations in 1922 and 1925 two necropolis areas to the north of Kolophon were explored, but were barely mentioned in Holland’s 1944 report. Combining archival data from the excavations and a careful observation of the territory we were able to determine the locations of these burial grounds. Remains of the North Necropolis (“Cemetery B” in the American field notebooks) were almost completely absent, but a close analysis of the excavation notebooks seems to indicate that this burial ground was used mainly during the Late Bronze Age79 and perhaps until the Geometric period, if the vague evidence for a Geometric period occupation is accepted. In the Northeast Necropolis (called the “Valley Cemetery” in 1922), two of the large tumuli (called Mounds I and II) excavated by the Americans could be retraced, but were found to be heavily disturbed. Both tumuli were carefully documented and these notations were subsequently 78. Meritt 1935. integrated with the original American descriptions, thus assuring that these 79. It contained the so-called tholos monuments will play a role in any future interpretation of the funerary rites tomb of Kolophon; see Bridges 1974. and topography of Kolophon.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 78 verena gassner, ulrike muss, et al.

The survey based on the LiDAR scan greatly facilitated the discovery of surface monuments in the urban area, particularly the burial grounds to the south of Kolophon. The vast majority of features documented were burial mounds and a smaller number of grave terraces. The rest most likely related to funerary activities, though their true functions remain unclear. Evidence for dating the tombs was meager, with a single diagnostic pottery find within the South Necropolis pointing to the Classical period. Further evidence for a Classical or Early Hellenistic date comes from finds discovered during the American excavations, most likely from the northern part of the South Necropolis. The Classical date for some of the tombs therefore seems very probable, but it is still unclear whether the burial grounds remained in use during the Classical period. Due to the very high number of mounds, typological comparisons, and the vast extent of the South and Southwest Necropoleis, it is possible that both also contained burial mounds from the Archaic period. This would then suggest that burial mounds had been part of the funerary tradition in practice since the Geometric period, and might have been used continuously, in different forms, through each stage of occupancy in Kolophon. Nevertheless, we do not have any direct material evidence to support this assumption, so any conclusions about chronology must remain preliminary. If the South and Southwest Necropoleis do not contain any Archaic tombs, the presence of another undiscovered burial ground needs to be considered. Possible locations include the fields west and east of the 80 city occupied by the modern villages of Değirmendere and Çamönü. Unfortunately, our consecutive survey efforts yielded no additional evidence, due to the modern agricultural and building activity in both areas. During the American excavations of Cemetery B in the plains to the northwest of Kolophon, at least one grave that can possibly be dated to the Archaic period was found, but the general occupation of the burial ground was sparse and some of the other tombs in the area can be dated to the Late 81 Bronze Age. Some of the isolated burial mounds mentioned before on 80. Bruns-Özgan, Gassner, and Yaren Tepe and Ambarkaya Tepe could also be part of larger burial grounds Muss 2011, pp. 226–229. with less overtly visible grave markers on the surface, which are hard to 81. Grammer, forthcoming. A more discern during fieldwalking. The hills of the Ambarkaya Tepe east of the detailed publication of these graves is city in particular show a high number of clearly visible pathways, as well being undertaken by Olivier Mariaud and Martin Gretscher within the scope as possible burial mounds and evidence of illegal digging in the LiDAR of the Kolophon project. See also the interpretation. The existence of another densely arranged burial ground contribution of Mariaud, pp. 54–62, in this area thus seems likely. above.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 79

REFERENCES

Akurgal, E. 1950. Bayraklı kazısı ön Bruns-Özgan, C., V. Gassner, and rapor (Zeitschrift der philoso- U. Muss. 2011. “Kolophon: Neue phischen Fakultät der Universität Untersuchungen zur Topographie Ankara 8.1), Ankara. der Stadt,” Anatolia Antiqua 19, ———. 1983. Alt-Smyrna 1: Wohn- pp. 199–239. schichten und Athenatempel, Ankara. Burgers, G.-J. 2012. “Urban Landscape Aydın, B. 2006. “A Survey of the Surveys on the Salento Isthmus, Clazomenaean Close Territory in Italy,” in Urban Landscape Survey Turkey-İzmir-Urla 1997,” in Studies in Italy and the Mediterranean, in Honor of Hayat Erkanal: Cultural ed. F. Vermeulen, G.-J. Burgers, Reflections, ed. A. Erkanal-Öktü, S. Keay, and C. Corsi, Oxford, Istanbul, pp. 95–107. pp. 13–22. Baitinger, H. 2001. Die Angriffswaffen Carstens, A. M. 2008. “Tombs of aus Olympia (OlForsch 29), Berlin. the Halikarnassos Peninsula: The Bammer, A., and U. Muss. 2006. “Ein Late Bronze and Early Iron Age,” Felsdenkmal auf dem Bülbüldağ in Halicarnassian Studies V, ed. von Ephesos,” Anatolia Antiqua 14, P. Pedersen, Odense, pp. 52–118. pp. 65–69. Cox, D. 1925. “Excavation Notebook, BCH, “Chronique” = “Chronique des Colophon 1925,” unpublished field fouilles et découvertes archéolo- notebook, American School of giques dans l’Orient hellénique Classical Studies at Athens. (nov. 1921–nov. 1922),” BCH 46, Davis, J. L. 2003. “A Foreign School of 1922, pp. 477–556. Archaeology and the Politics of Berndt-Ersöz, S. 2006. Phrygian Rock- Archaeological Practice: Anatolia, Cut Shrines: Structure, Function, 1922,” JMA 16, pp. 145–172. and Cult Practice (Culture and His- De La Genière, J. 1994. “Quelques tory of the Ancient Near East 25), réflexions à propos des murailles de Leiden. Colophon,” RÉA 96, pp. 137–140. Bintliff, J. L., P. Howard, A. M. Snod- Doneus, M., C. Briese, M. Fera, and grass, and O. T. P. K. Dickinson. M. Janner. 2008. “Archaeological 2007. Testing the Hinterland: The Prospection of Forested Areas Using Work of the Boeotia Survey (1989– Full-Waveform Airborne Laser 1991) in the Southern Approaches to Scanning,” JAS 35, pp. 882–893. the City of Thespiai (McDonald Eckert, A. 1998. “Ein Grab für Könige Institute Monographs), Cambridge. und Bürger: Studien zum monu- Blegen, C. W. 1922. “Excavation Note- mentalen Tumulusgrab als Mittel book, Colophon 1922,” unpublished der Selbstdarstellung mittelmeeri- field notebook, American School of scher Eliten vom 8. bis zum 6. Jahr- Classical Studies at Athens. hundert v. Chr.” (diss. Univ. of Boehlau, J. 1898. Aus ionischen und Hamburg). italischen Nekropolen: Ausgrabungen Ersoy, Y. 2007. “Notes on History and und Untersuchungen zur Geschichte Archaeology of Early Clazomenae,” der nachmykenischen griechischen in Frühes Ionien: Eine Bestandsauf- Kunst, Leipzig. nahme. Panionion-Symposion Güzel- Bofinger, J., and R. Hesse. 2011.” Der çamlı, 26. September–1. Oktober 1999 Einsatz von Airborne Laserscan- (Milesische Forschungen 5), ed. ning zur Entdeckung von archäolo- J. Cobet, V. von Graeve, W.-D. gischen Geländedenkmalen,” in Niemeier, and K. Zimmermann, Mit Hightech auf den Spuren der Mainz, pp. 149–178. Kelten (Archäologische Informa- Fowler, H. N., 1922. “The American tionen aus Baden-Württemberg 61), School of Classical Studies at ed. J. Bofinger and M. Merkl, Ess- Athens: The Excavations at Colo- lingen, pp. 70–89. phon,” Art and Archaeology 14, Bridges, R. A. 1974. “The Mycenaean pp. 256–260. Tholos Tomb at Kolophon,” Hespe- Goldman, H. 1922. “Excavation Note- ria 43, pp. 264–266. book, Colophon 1922” (unpublished

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms 80 verena gassner, ulrike muss, et al.

field notebook, American School of 20–21 October 2001, ed. A. Mou- Miltner, F., and H. Miltner. 1932. “Be- Classical Studies at Athens). staka, E. Skarlatidou, M.-C. Tzannes, richt über eine Voruntersuchung in ———. 1923. “Excavations of the Fogg and Y. Ersoy, Thessaloniki, pp. 77– Alt-Smyrna,” ÖJhBeibl 27, pp. 127– Museum at Colophon,” AJA 27, 95. 188. pp. 67–68. Il, Ö. 2009. “Tumuli Asiae Minoris: Muss, U., A. Bammer, E. Draganits, Graf, F. 1985. Nordionische Kulte: Reli- Untersuchungen zu den phrygi- V. Gassner, B. Grammer, M. Gret- gionsgeschichtliche und epigraphische schen und lydischen Tumulus- scher, and O. Mariaud. 2014a. “Col- Untersuchungen zu den Kulten von gräbern der Eisenzeit im zentralen ophon 2013,” Forum Archaeologiae Chios, Erythrai, Klazomenai, und und westlichen Kleinasien” (diss. 71/VI/2014. http://farch.net. Phokaia (Bibliotheca Helvetica Univ. of Heidelberg). Muss, U., V. Gassner, B. Grammer, and Romana 21), Rome. Karwiese, S. 1995. Gross ist die Artemis M. Gretscher. 2014b. “The Colophon Grammer, B. Forthcoming. “The Burial von Ephesos: Die Geschichte einer der Survey 2014,” Forum Archaeologiae Mounds of Colophon,” in Proceed- grossen Städte der Antike, Vienna. 73/XII/2014; http://farch.net. ings of the International Olympos of Kleiner, G., P. Hommel, and W. Müller- Naumann, F. 1983. Die Ikonographie der Mysia (Uludağ) Studies Workshop in Wiener. 1967. Panionion und Melie Kybele in der phrygischen und der Büyükorhan and Orhaneli Towns of (JdI-EH 23), Berlin. griechischen Kunst (IstMitt-BH 28), Bursa, 8th–11th May, 2014. Körte, A. 1898. “Kleinasiatische Stu- Tübingen. Hanfmann, G. M. A. 1953. “Ionia, dien III: Die phrygischen Felsdenk- Naumann, R. 1967. “Das Heiligtum Leader or Follower?,” HSCP 61, mäler,” AM 23, pp. 80–153. der Meter Steunene bei Aezani,” pp. 1–37. Lambrechts, P. 1970. “La quatrième IstMitt 17, pp. 218–247. Hellmann, M.-C. 2010. L’architecture campagne de fouilles à Pessinonte Nohlen, K., and W. Radt. 1978. Kapı- grecque 3: Habitat, urbanisme, et (Turquie),” TürkArkDerg 19, kaya: Ein Felsheiligtum bei Pergamon fortifications, Paris. pp. 133–142 (AvP 12), Berlin. Henry, O. 2009. Tombes de Carie: Archi- Lorimer, H. L. 1950. Homer and the Özgan, C., V. Gassner, U. Muss, tecture funéraire et culture carienne, Monuments, London. E. Draganits, and Ö. Teraman. VIe–IIe s. av. J.-C. (Archéologie et Mariaud, O. 2007. “Necroionia: Arché- 2014. “Kolophon Antik Kenti 2012 culture), Rennes. ologie, espace et société: Recherches Yılı Yüzey Araştırmaları,” Araştırma Herda, A. 2013. “Greek (and our) sur les nécropoles et les sociétés Sonuçları Toplantıları 31, pp. 97–109. Views on the Karians,” in Luwian d’Ionie à l’époque archaïque (700– Özgan, C., M. Özgen, K. Eren, A. Po- Identities: Culture, Language, and 500 av. J.-C.)” (diss. Univ. Bor- lat, and B. B. Aykanat. 2013. “Kolo- Religion between Anatolia and the deaux 3). phon Antik Kenti 2011 Yılı Yüzey Aegean (Culture and History of the ———. 2011. “The Geometric Graves Araştırmaları,” Araştırma Sonuçları Ancient Near East 64), ed. A. Mou- of Colophon after the Excavations Toplantıları 30, pp. 195–206. ton, I. Rutherford, and I. S. Yakubo- of H. Goldman, 1922: Reflexions on Özgan, C., M. Özgen, K. Eren, A. Po- vich, Leiden, pp. 421–506. the Burial Customs of Early Iron lat, and Ö. Çınar. 2012. “Kolophon Hoepfner, W., and E. W. Osthues Age Ionia,” in The “Dark Ages” Re- Antik Kenti 2010 Yılı Yüzey Araş- 1999. “Kolophon” in W. Hoepfner, visited: Acts of an International Sym- tırmaları,” Araştırma Sonuçları To- ed. Geschichte des Wohnens 1: 5000 v. posium in Memory of William D. E. plantıları 29, pp. 263–283. Chr.–500 n. Chr.: Vorgeschichte, Früh- Coulson, University of Thessaly, Volos, Philipp, H. 1981. “Archaische Gräber geschichte, Antike, Stuttgart. 14–17 June 2007, ed. A. Mazarakis in Ostionien,” IstMitt 31, pp. 149– Holland, L. B. 1922. “Sketchbook, Ainian, Volos, pp. 785–799. 166. Colophon 1922,” unpublished field Meriç, A. E. 2014. “A New Deposit Picard, C. 1922. Ephèse et Claros: Re- notebook, American School of from a ‘Banquet Hall’ of the Bath- cherches sur les sanctuaires et les cultes Classical Studies at Athens. Gymnasium in Metropolis,” ÖJh 83, de l’Ionie du Nord (BÉFAR 123), ———.1944. “Colophon,” Hesperia 13, pp. 11–41. Paris. pp. 91–171. Meriç, R. 1982. Metropolis in Ionien: Popham, M. R., and I. S. Lemos. 1996. Hülden, O. 2006. Gräber und Grabtypen Ergebnisse einer Survey-Unternehm- Lefkandi 3: The Toumba Cemetery: im Bergland von Yavu (Zentral- ung in den Jahren 1972–1975 (Bei- The Excavations of 1981, 1984, 1986, lykien): Studien zur antiken Grabkul- träge zur klassischen Philologie 142), and 1992–4 (BSA Suppl. 29), Oxford. tur in Lykien (Antiquitas 45), Bonn. Königstein. Radt, W. 1970. Siedlungen und Bauten Hürmüzlü, B. 2004. “Burial Grounds at Meritt, B. D. 1935. “Inscriptions of auf der Halbinsel von Halikarnassos Klazomenai: Geometric through Colophon,” AJP 56, pp. 358–397. unter besonderer Berücksichtigung der Hellenistic Periods,” in Klazomenai, Meyer-Plath, B. 1940. “Nekropolen,” in archaischen Epoche (IstMitt-BH 3), Teos, and Abdera: Metropoleis and Larisa am Hermos: Die Ergebnisse der Tübingen. Colony: Proceedings of the Interna- Ausgrabungen 1902–1934 1: Die Roller, L. 1999. In Search of God the tional Symposium Held at the Archae- Bauten, ed. J. Boehlau and K. Sche- Mother: The Cult of Anatolian Cybele, ological Museum of Abdera, Abdera fold, Berlin, pp. 109–112. Berkeley.

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms the urban organization of kolophon 81

Rubinstein, L. 2004. “Ionia,” in An Scherrer, P., ed. 2000. Ephesus: The New und Kore im sogenannten Felsspalt- Inventory of Archaic and Classical Guide, rev. ed., trans. L. Bier and tempel in Ephesos,” in Synergia: Poleis, ed. M. H. Hansen and T. H. G. M. Luxon, Istanbul. Festschrift für Friedrich Krinzinger, Nielsen, Oxford, pp. 1053–1107. Scherrer, P., E. Trinkl, S. Fabrizii- ed. B. Beck-Brandt, V. Gassner, and Rumscheid, F. 1998. Priene: Führer Reuer, and M. Scherrer. 2006. S. Ladstätter, Vienna, pp. 319–326. durch das “Pompeji Kleinasiens,” Die Tetragonos Agora in Ephesos: Tölle-Kastenbein, R. 1976. Herodot und Istanbul. Grabungsergebnisse von archaischer bis Samos, Bochum. Şahin, N. 2008. “Notion-Klaros- in byzantinische Zeit: Ein Überblick: Vermeulen, F., G.-J. Burgers, S. Keay, Kolophon Üçgeninde Myken Befunde und Funde klassischer Zeit and C. Corsi, eds. 2012. Urban Sorunu,” in Patronus: Coşkun (Ephesos 13.2), Vienna. Landscape Survey in Italy and the Özgünel’e 65. Yaş armaganı = Schuchhardt, C. 1886. “Kolophon, Mediterranean, Oxford. Festschrift für Coşkun Özgünel Notion, und Klaros,” AM 11, Wiegand, T., and H. Schrader. 1904. zum 65. Geburtstag, ed. C. Özgü- pp. 398–434. Priene: Ergebnisse der Ausgrabungen nel, E. Öztepe, and M. Kadioğlu, Soykal-Alanyalı, F. 2005. “Überle- und Untersuchungen in den Jahren Istanbul, pp. 329–338. gungen zu dem Kult von Demeter 1895–1898, Berlin.

Verena Gassner Martin Gretscher University of Vienna kleine alee 3 institut für klassische archäologie 07407 rudolstadt franz klein-gasse 1 germany 1190 vienna [email protected] austria [email protected] Olivier Mariaud Université Grenoble-Alpes Ulrike Muss laboratoire universitaire histoire University of Vienna cultures italie europe (luhcie) institut für klassische archäologie 1281 avenue centrale franz klein-gasse 1 38400 st-martin-d’hères 1190 vienna france austria [email protected] [email protected]

Benedikt Grammer University of Vienna institut für urgeschichte und historische archäologie franz klein-gasse 1 1190 vienna austria [email protected]

This content downloaded from 84.114.116.80 on Fri, 17 Feb 2017 11:39:01 UTC All use subject to http://about.jstor.org/terms