In the shadow of the Wall. Hellenistic settlement in the Baysun and Kugitang piedmonts

In the shadow of the Wall. Hellenistic settlement in the Baysun and Kugitang piedmonts Ladislav Stančo ABSTRACT This article brings an up -to -date evaluation of the archaeological research in the core of the Bactro -Sogdian borderlands, i.e., in the vicinity of the Darband Wall, Baysun District, southern Uzbekistan, including the most recent results of the fieldwork of the Czech -Uzbek archaeological expedition. These are combined with the fruits of the efforts of other local and international teams busy in this region for the last twenty years in aspatiotemporal assessment. Building upon the lack of evidence, the author argues against the identification of the selected locations in the region as places where the events connected with the invasion of Alexander the Great took place. We also show that the area of the Baysun District including Darband was for the first time in history settled in the Seleucid / Greco -Bactrian period. The original function of the Darband Wall itself was most probably related to an event preceding the campaign of Antiochos III to Bactria and the presumed threat of nomads. KEY WORDS Hellenistic; Bactria; Darband Wall; Alexander the Great; Seleucid; Greco -Bactrian; settlement patterns. INTRODUCTION The contribution under the title above was originally presented at the conference eponymous for the current volume. In Autumn 2018, Idiscussed the state of research valid for that moment, Iknew however that the textual version of the talk should be quite different, since new data were popping up with enormous rapidness, only slowed down by the covid pandemic. Initially, Iplanned to publish atext resembling the final report of the project On the mountain of Oxyartes: detection of forts and refuges from the time of Alexander the Great in Central Asia. It would, however, require presenting such chronologically complex evidence that it would not fit the framework intended for these conference proceedings. Finally, Idecided to summarize the most recent research results and fresh evidence resulting in the present critical text, which aims to show where we are in the Bactro -Sogdian borderlands archaeology of the Hellenistic period right now, in 2021, and mayhap where to go and how to continue in the near future. The participants of the conference would not recognize the talk they heard in this paper; however, Ihope they will enjoy the reading all the same. We shall start with afew definitions of terms and delimitations of the topic. The wall in the title of this paper means not only the well -known man -made Darband Wall– afortification system founded in the 3rd c. BC in the vicinity of the modern village of the same name, but also the wall of the mountains overshadowing the valley of the Sherabad Darya, separating it from that of the Kashka Darya and its tributaries to the north -west and reaching at the highest places as much as 3000 m.a.s.l. The Baysun mountains as well as the range of Kugitang form the south -western spur of the larger Pamir -Alay system truly separating two remarkable historical regions– Bactria and Sogdiana. This mountainous type of landscape and its limSTUDIA HERCYNIA XXV/2, 64–95 65LADISLAV STANČO itations, offering only afew options to travel through the passes of 1500 m.a.s.l. and higher, played akey role in the historical processes and settlement dynamics of this area. Darband though, was as difficult to evade two thousand years ago as it is today for everyone travelling to Sogdiana from Bactria or vice versa,1 unless one wanted to add many amile and circumvent the mountains along the middle reaches of the Amu Darya. Thus, from the geographic point of view, this paper focuses on the area of the upper reaches of the Sherabad Darya and its tributaries for several reasons: 1) It was the area in which the research in the above -mentioned project was carried out, corresponding roughly with the present -day Baysun District. 2) At the same time, being interested in the Darband micro- -region and its settlement dynamics, we tried not to exceed awalking distance of ca 30–35 km from the Darband Wall, which was possible to cover within aone -day journey on foot and with transport animals. Both reasons exclude such sites as those situated in the Pashkhurt valley (Ghisht Tepa, Dabil Kurgan). These, in their turn, will be discussed in detail in the final publication of the Czech -Uzbek expedition projects in the Pashkhurt basin. By 2016, only afew archaeological sites within the upper reaches of the Sherabad Darya had yielded material dated to the ‘Hellenistic’ period. Note that when the term ‘Hellenistic’ is used throughout this paper, as meaning broadly the time span from the 320s BC (Alexander the Great’seastern campaign) to the mid-2nd c. BC (end of the Greek rule in southern Central Asia). It comprises the time of Alexander and the diodochoi, the Seleucid rule and ca. one hundred years of existence of the Greco -Bactrian kingdom until the rise of the Saka and Yuezhi tribes in southern Central Asia. It is far from being an ideal term for labelling this period, but still the handiest one which we have at our disposal. It is to some extent questionable whether the ensuing ‘transitional’ or Yuezhi period is to be included into the Hellenistic period, but Idecided to do so, assuming that at least culturally, if not politically, there is obvious discernible continuity. Our approach towards the study of the region and the period in question would be best defined through the idea contained in the title of the conference. Seen from Oxyartes’ Rock does not mean to look for the particular peak in the Baysun mountains where the nobleman and Roxane’sfather stood and contemplated the approaching Macedonian army; it means to be provided with alittle distant, but clear and complex view of what lies beneath, whatever the mountain and lookout point atop are called. In the call for papers of the Prague HCARN conference, we ambitiously aimed to ‘take alocal point of view and ask how local people experienced these turbulent developments, and how they coped with the strange [Greco- -Macedonian] newcomers.’ This is hard to achieve, however, having only limited written and archaeological sources at our disposal. All the more so it must not prevent us from trying. STATE OF RESEARCH The Hellenistic period research ranks among the most rapidly evolving sub -disciplines of archaeology and definitely among the most popular ones from the point of view of scholars focusing on southern Central Asia.2 In order not to repeat what has been published already, Iwill refer the reader to the overview of the past scholarship and state of research concerning the Hellenistic period in the Surkhan Darya province of Uzbekistan as awhole that had been 1 For atentative reconstruction of ancient roads and mobility in this region see Stančo– Pažout 2020. 2 So far, the best and most comprehensive overview of Central and South Asia in the Hellenistic period has been issued quite recently (Mairs ed. 2020). 66 STUDIA HERCYNIA XXV/2 put together by 2018 and that has been published only quite recently (Stančo 2020, 251–254). What is important for the present analysis is either what has happened since, i.e. in the field seasons 2018 and 2019, in the framework of the Czech -Uzbekistani archaeological expedition (which is an extensive surface survey in the Darband area, excavations of Iskandar Tepa, trial trenches at Daganajam, Kapchigay, Mirzali at Khojay Gor, an investigation of the Darband Wall itself, metal detector surveys at many sites, etc.), or has been published by the team of the Russian Academy of Sciences meanwhile. The latter is especially the case of aseries of publications related to the research at the fortress of Uzundara and its vicinity published by N. Dvurechenskaya and her team -members (Gorin– Dvurechenskaya 2018; Beľsh 2020; Dvurechenskaya, N.D. 2020; Dvurechenskaya, T.O. 2020; Dvurechenskaya, S.O. 2020). Moreover, the long expected final report on the Darband Wall excavations in the late 1990s is about to be published (Rapinetal. forthcoming). Thus, let us– after afew methodological remarks– start with an overview of the available data focusing on the principal characteristics of the settlements and other structures dating from the late 4th to the 2nd century BC paying special attention to the function of the respective sites. These are listed below alphabetically. METHODOLOGICAL ISSUES Most of the large body of data we work with now originates from the quite recent international field projects focused on the Hellenistic period in the Darband / Baysun area. Innovative as they might be in some respects, all of them suffer from methodological drawbacks and inconsistencies ranging from basic field work methods, through the documentation and publication, down to the most controversial one– the data interpretation. These have aserious impact not only on the individual research results on the one hand, but also on the data sets shared with other scholars for further research on the other. The common denominator of many of these issues, especially within the final interpretations, is an old scientific enemy: bias. The team of the Russian Academy of Sciences led by Nigora Dvurechenskaya focusing primarily on the long -term in -depth investigation of the Uzundara fortress based on regular archaeological excavations, managed to employ also asurface survey of the close vicinity of this site. As stated in the first publication on this topic, local informants– hunters– who had field experience from the excavations of the fortress itself, were employed in order to detect similar features in the neighbourhood (Beľsh 2020, 381–383). This method being as old as archaeology itself and usually very fruitful, led in this case to the detection of afew interesting architectural remains interpreted as fortifications of several types dated generally– without convincing evidence– to the Greco -Bactrian period and functionally linked to the nearby fortress. Problems related to the method are twofold: first, seeking specific features in the terrain may thus lead to omitting various other anthropogenic features having asimilar value. Amore appropriate approach would be, in my opinion, asystematic intensive surface survey documenting all relevant features present in aclearly defined area, and asubsequent analysis leading to afunctional interpretation of the individual features as well as of the entire system of which they may form apart; only then– built on solid chronological foundations– can we continue with ahistorical assessment. In the case of amore haphazard, targeted, survey, one would expect more consistent methods, too, to be not only employed, but also described in full detail and explained in the publication. Thus, we lack crucial information on the survey, such as whether it covered aspecific area of interest, or whether it was limited only to the published spots, whether there were some features surveyed and later discarded from the list and if so, for what (chronological / typological / other) reason. Finally, already having this set of data, for any further study it would be necessary to have precise spatial information (i.e., accurate coordinates) on the features in question. 67LADISLAV STANČO The present author is well aware of the fact that some of these critical remarks above apply also to the work of the Czech -Uzbekistani team. From the very beginning of the project On the mountain of Oxyartes: detection of forts and refuges from the time of Alexander the Great in Central Asia, we followed up on the previous experiences in the Sherabad Oasis (Stančo– Tušlová 2019) and especially in the Pashkhurt Basin (Stančo 2016; Stančoetal. 2017; Augustinováetal. 2017). Considering the fact that we are about to cover avery large research area (ca 4500 sq. km!) with atypically mountainous landscape, we built our methodology on an extensive rather than intensive surface survey. Thus, we surveyed 1) already known, but insufficiently studied archaeological sites in order to revise the published chronological information; 2) places topographically suitable for settlement or other past human activities (typically terraces of the river valleys, especially tepa -like mounds along the banks of piedmont water streams); and 3) courses of the presumed roads and paths crossing the region. As this list presents itself, the choice of the surveyed spots might also look rather haphazard. Even if our primary goals were related to the period of the second half of the 4th c.–second half of the 2nd c. BC, we systematically documented all anthropogenic features discovered. Anyway, not all previously known sites in the region were visited,3 only selected paths across the mountain ridges were walked, and the list of the surveyed riverbanks is far from complete by now. Some methods, moreover, such as ametal detector survey, were only tested at several sites in order to prove the effectivity and benefits of metal detectors for future research in similar conditions, while at the selected sites– e.g., the Darband Wall and Iskandar Tepa– they were employed in amuch more desirable systematic intensive way. The survey method based on the use of metal detectors is still questioned and sometimes strictly rejected by some archaeologists as aproblematic method damaging archaeological sites. In our opinion, when used responsibly and under the control of aprofessional archaeologist, ametal detector survey ranks among the most highly efficient non -destructive research methods. The majority of the surface survey is conducted on– and disturbs only– the topsoil (metal objects are basically picked up from the surface or from the thin subsurface layer slightly below, not deeper than 5–8 cm), at erosive slopes of archaeological sites (both tepas and natural mounds topped with asite), and below them hardly reaching archaeological contexts as such. It is worth mentioning that we do collect metal objects without discrimination, i.e., being made of iron, silver, gold, bronze / copper, and other alloys. Functionally, all sorts of items are represented, starting from sophisticated tools and weapons, and coins (even modern ones) to simple slag and iron rods. The precise coordinates and detailed documentation make the find assemblage apowerful spatiotemporal analytic tool for aproper assessment of the site dating and periodization. Last but not least, being well aware of the unfortunate development associated with the spread of illegal detector prospecting in Europe and elsewhere in the world over the last few decades, our further goal was also to save material from these sites from the inevitable interventions of this type in the near future here in Uzbekistan. The situation varies significantly from country to country but there is acommon denominator to all of them: an irreversible loss of an immense amount of archaeological data combined in some cases with an open war between archaeologists and ‘wild’ detectorists. To take an example from the Czech Republic, M. Čižmář (2006) drew attention to the looting of thousands of coins and other artefacts from alocal La Tène period (= Late Iron Age) site in 2002–2003 (cf. Čižmář 2006, 285–288 on the ethics of the issue). Some other nine contributions on this controversial 3 Some of the sites at higher altitudes in the mountain valleys to the north of Baysun listed in the appendix to Annaev– Annaev 2003 and Sverchkov 2005b were not visited by the Czech -Uzbek team, since the only material known so far dated from the High Middle Ages. 68 STUDIA HERCYNIA XXV/2 topic were published in the same volume of the periodical Archeologické rozhledy accelerating the discussion resulting in the prevailing opinion among the Czech archaeologists that we need to use metal detectors to protect archaeological heritage. To illustrate the dynamics over the last 15 years: in 2006 Čižmář spoke about 3000 metal detectors sold in the Czech Republic (Čižmář 2006, 286), while by now, there are already tens of thousands of these instruments distributed throughout the country. In Uzbekistan, we have the rare opportunity to protect the sites and the respective archaeological material by the use of preventive metal detector surveys, of course under the auspices of local authorities. In any case, aserious discussion on this topic comparable to the one that took place 15 years ago in the Czech Republic still awaits the professional archaeological community in Uzbekistan. For methodological remarks on the employment of the detectors in the archaeological surface survey see for instance Vích 2014 (note that the author specifically addresses the question of how to work with detectors in archaeology and not whether to do so, since he considers this problem to be solved). Athird remark goes to the excavators of Kurganzol that took place in two stages (2003– 2004 and 2008) and profited from 1) well -preserved archaeological contexts, which were not disturbed by any later use or inhabitation, and 2) very responsibly employed analyses of not only numerous artefacts, but also significant ecofacts– at least in the last phase of the excavations in the season 2008 –, putting emphasis on animal bones (Benecke 2013), plant macroremains (Neef 2013), as well as samples for both 14C and dendrochronology dating (Haussner– Boroffka 2013).4 As aresult, the final publication offers alively picture of the small early Hellenistic military base, or rather watch -post (Sverchkov 2013). Unlike Uzundara, metal objects are almost absent at the site according to the final report,5 since excavators intentionally refrained from using metal detectors during the work. As shown by our test metal detector sampling survey on the surface of the forecourt of the fort (three coins and some arrow heads of Hellenistic dating were detected; Stančoetal. 2019, 146; Stančoetal. forthcoming a), metal objects were not absent and might be– including coins– quite numerous even in undisturbed archaeological strata. These would serve well in discussions about the date of the fort foundation, life, and abandonment, currently based mostly on C14 (problematic for this time period representing the end of the so -called Hallstatt Plateau and not trusted by the excavators themselves), dendrochronological data, and ceramic typological analysis.6 Finally, the complete results and data of the detailed Darband Wall investigation, which took place more than twenty years ago (1996–2001) under Cl. Rapin and Sh. Rakhmanov, would be of the utmost importance, had they been properly and fully published. The same also goes for the abundant archaeological material that was brought to light during the excavation of Payon Kurgan (see below). In both cases, important archaeological material is briefly mentioned in the text of the report, and consequently used as an argument for dating, without being displayed in drawing or photography.7 Even if chronologically the necropolis of Rabat (1–3), which according to the excavators might be attributed to the Yuezhi tribes, lies beyond the scope of this paper, the scholarship 4 For acriticism of the results’ interpretation see Lyonnet– Fontugne 2021. 5 Altogether only four fragments of iron tools are published (Sverchkov 2013, 101, ris. 87:1–4, 102, ris. 88: 1–4). 6 For the latest assessment of the available data on the chronology of Kurganzol see Lyonnet– Fontugne 2021. 7 Analogically, we miss apublication of material from the sites of Maydan Kurgan and Dabil Kurgan in the Pashkhurt valley (out of the scope of this paper, but belonging also to the Kugitang piedmonts) advertised here and there as evidence for the Hellenistic dating of both these settlements. 69LADISLAV STANČO on the Hellenistic period in southern Uzbekistan would have greatly profited from adetailed publication of both the old and the more recent research at this site, where reportedly several dozen nomadic graves have been excavated (Abdullaev 2007, 80–81; Liang Yunetal. 2018; Liuetal. 2020). From the first stage of research there, the full publication would be appreciated, while in the case of the second, some other -than -Chinese overview article may help the international community to profit profoundly from the obviously inspiring research results obscured by the difficult Chinese language. DATA OVERVIEW ARCHAEOLOGICAL SITES In this overview, only the most significant sites are described in detail, general information on all relevant sites is given in asystematic way in Tab.1 below. The sites are sorted alphabetically (Fig. 1). Daganajam Tepa Daganajam Tepa (or Dakhna -e Jom Tepa; BA5_09; 38.139941, 67.041210) was discovered by way of visual observation of topographical anomalies in the village of the same name on 24 April 2018 by the Czech -Uzbek team and surveyed in detail on 1 and 5 May 2018 (Stančoetal. 2019, 146). Asix -day trial excavation led by J. Kysela took place in September 2019 (Stančoetal. forthcoming b). The site is situated on the right (i.e., west) bank of the Sherabad Darya some 230 m from the river itself, obviously out of reach of high -water level and floods in spring time (the slope below the site is quite steep offering additional protection). At the same time, an artificial water canal, which is diverted from the main course of the river stream ca 2.5 km to the north, runs just along the western outskirt of the site following the modern road. Water might have been brought to the site in exactly the same manner in Antiquity. The island thus created by the river and canal form afertile micro -oasis 9.7 km long, narrow in its northern part, but reaching as much as 1.8 km in width in the southern part. The total area of this micro -oasis calculated in the GIS reaches 820.2 ha, which makes it one of the largest irrigated areas in all of the Baysun foothills. The size of the site itself reaches 85 m (NNW–SSE) × 44 m (NE–SW) and its summit is elevated 4–5 m above the surrounding terrain. Its surface area including the erosive slopes measures 0.35 ha, more precisely 3492 m² (Stančoetal. 2019, 151). Neither the topography of the site nor the trial excavations indicate any remains of afortification so far (further research, however, may prove otherwise). This hillock has been reportedly used as achildren’sgraveyard in the recent past, and these activities heavily affected the upper strata of the mound’ssummit, which is not suitable for excavations of any kind. Altogether 35 pottery fragments that were collected during the 2018 survey were dated exclusively to the Hellenistic period (Stančoetal. 2019, fig. 6:11–13), the excavations only confirmed this dating. An intensive metal detector survey brought to light six Hellenistic coins belonging to the Seleucid king Antiochos I, as well as the Greco -Bactrian rulers Euthydemos I, and Demetrios I(Stančoetal. forthcoming a). Consequently, numismatic finds suggest that Daganajam Tepa was settled from ca. the second quarter of the 3rd c. to the first quarter of the 2nd c. BC.The primary function of the site is unclear, though its size and location points to arural settlement. Moreover, the site visually communicates very well with the rest of the micro -oasis, but not with the fortifications of Uzundara and the Darband Wall to the west, being separated by the ridge of Ak Tau (1120 m.a.s.l.). 70 STUDIA HERCYNIA XXV/2 Darband Wall The site (BA5_21–23; 38.208658, 66.973709) was mentioned for the first time by Parfyonov in 1930 (Rakhmanov– Rapin 1998, 5) and then found again in 1986 by E. V. Rtveladze as asupposed Kushan fortification (Rtveladze 1986), although the same author seems to ignore it in his later publication (Rtveladze 1990a). Later on, the French -Uzbek Darband archaeological detachment led by Sh. Rakhmanov and Cl. Rapin opened several trenches (sections of the wall and excavations of towers) in 1996–2001. They came to the conclusion Fig. 1: Map of the Bactro -Sogdian Borderlands in the second half of the 1st millennium BC, by author. 71LADISLAV STANČO that the wall was originally built in the Greco -Bactrian period, while in the Kushan period, it was repeatedly rebuilt and reinforced (Rakhmanov– Rapin 2003; Rapinetal. 2006).8 Afull report of these digs– in addition to several interim reports and short articles– has been finished just recently9 and shall appear in 2022 (Rapinetal. forthcoming). Again, only ahandful of (Greco -Bactrian) pottery is being made public in this new article, while the mass of the Yuezhi, Kushan, and also Medieval material that should dominate the finds in accordance with the dating is entirely omitted in the figures and only briefly mentioned in the text. The authors state here that the fortifications were built first during the reign of Euthydemos I(based upon the numismatic data from Uzundara– sic!), but taking into account the early dating of the majority of this king’scoins, the immediate threat of anomad invasion during the war between Euthydemos and Antiochos Iis no longer perceived as the one and only reason for the construction (Rapinetal. forthcoming). Finally, this new analysis admits that the function of the Wall might have been ‘politically symbolic rather than areal military’ one (Rapinetal. forthcoming). The paper does not provide adecisive argument to answer the question as to which direction the Wall was oriented towards, since the excavators date the towers projecting westwards10 to the Kushan period only. For the Greco -Bactrian period, no tower has been identified yet, thus the interpretation is based on the ‘niche that served as an inner tower accessible from the east’, which is, however, not visible in the relevant figure (Rapinetal. forthcoming, fig. 9). The Czech -Uzbekistani team returned to this site repeatedly within the framework of the project On the mountain of Oxyartes in 2018 (Stančoetal. 2019), 2019, and 2021 in order to gain fresh archaeological data in non -destructive ways. These included asurface topographical survey, ageophysical survey, pottery fragments collection (ca. 300 diagnostic fragments collected all over the wall’ssurface and erosive slopes below), an extensive systematic metal detector survey (629 metal objects, 364 out of these are coins of various historical periods;11 for details see subhead on coins below), and section cleaning and documenting. Consequently, alarge body of material has been obtained that is still being processed. Adedicated field report is in preparation. Basically, the wall stretches for ca 1000–1100 m in length from the steep slopes of Sarymas Mountain in the north across the valley towards the left bank of the Shurob Darya gorge in the south, where again the steep slopes of the Susiztag form the right bank. The wall itself, even though eroded, remained in avery good state of preservation with the exception of sections completely destroyed during the construction of an old road and anew motorway. 12 Thus the wall is nowadays divided into three sections separated by both natural and artificial depressions, the northern stretch being 220 m long, the central one 265 m long, and the southern one 335 m long.13 The latter in particular is heavily affected by recent interventions 8 Even after the results of the digs were published, some scholars stick with the original interpretation and Kushan period dating of the site (Sverchkov 2005b, 13). 9 Iam very grateful to Claude Rapin for allowing me to read his draft paper before publication. 10 Towers, typically square or oblong in their ground plan, used to be placed on the outside of the fortification structures in the ancient Greek defensive architecture providing additional defence especially to the closest vicinity and forefield of the wall. 11 It is remarkable that no coins were reported from the excavations in the late 1990s. The publication in press, however, presents two coins: an imitation of Heliocles I(supposed to belong to the turn of the 2nd and 1st c. BC), and adidrachm of Soter Megas (Rapinetal. forthcoming). 12 Rapin states that the wall lost 110 m, and 100 m respectively due to these construction works (Rapinetal. forthcoming). 13 The entries in Tab.1 respect this division, since we did so as well during the surface survey, thus we call the individual sections Darband Wall North, Darband Wall Centre, Darband Wall South. The 72 STUDIA HERCYNIA XXV/2 such as the construction works both of the roads and of arailway running at the bottom of the Shurob Say. According to our research, the core and the earliest phase of the wall resting on agypsum subsoil is formed by astone -made structure that was later reinforced by amud- -brick wall.14 As aresult, the wall completely blocked the passage through the valley except for the deep ravine of the Shurob Say. As to whether any measures preventing passage had been taken down there, we are not able to say, since the bottom of the valley left no traces of anthropic activity due to the railroad construction. Dogoba (Duoba) In 2003, aDemetrios Isilver coin was brought by local inhabitants of the village of Dogoba to the Termez Archaeological Museum. Members of the Czech -Uzbek archaeological expedition consequently visited the village seeking in vain for some more information about the context of the find. K. Abdullaev later published this coin– the first Hellenistic coin in the entire Darband / Baysun area– as found in the village of Munchak Tepa near Sairob (Abdullaev 2006, 108–109). Iskandar Tepa The site (ShD 243; 37.920690, 66.973576) was detected by the Czech -Uzbek team on the 12 th September 2017 (Stančoetal. 2017) and briefly surveyed and excavated on the 16th September 2017. Larger excavations took place in the next season in 2018 (September 14 th till October 7 th), while targeted excavations commenced after adetailed geophysical survey in September 2021. The site is located on the outskirts of the Loylagan valley, on the way from Ishtara to Loylagan, on the summit of anarrow promontory extending eastwards from an elevated ridge ca. 20 m above the bottom of aflat, narrow (ca. 500 m) and elongated valley, which itself stretches for ca. 5.4 km from north -west to south -east (Stančoetal. 2020). The surface area of the potentially irrigated land in this ‘Kulal Tepa micro -oasis’ was calculated as 246.7 ha. The site of Iskandar Tepa situated exactly in the centre of its length overlooks the entire micro -oasis. The core of the settlement site measures ca. 50×30 m only, while the entire summit that forms an elongated E -W oriented oval, measures ca. 260×40 m. The key evidence for the settlement activities is provided by the large number of storage jars cut into the bedrock. Repeatedly, ametal detector survey was employed during both excavations and surveys resulting in the accumulation of dozens of metal objects, including coins of Diodotos I, Euthydemos I, and DemetriosI. The dating based upon both coins and the pottery assemblage (Stančo– Martínez Ferreras– Kysela forthcoming; Stančoetal. forthcoming b) point, however, rather to the 2nd c. BC as the main period of occupation. Kapchigay Tepa The site (BA5_24; 38.216822, 67.031297) was surveyed repeatedly in the past by Rtveladze and Sverchkov and dated according to the collected pottery material to the Greco -Bactrian and High Medieval Period (Sverchkov 2005b, 13–14, no. 30, fig. 10:19–21). The ancient settlement was situated, along with the northernmost part of the modern Darband village from which it French -Uzbek team divided the Wall into four parts, the first two corresponding to our North (C1) and Centre (C2), while the southern part was further subdivided into two parts C3 (our South), and C4, which seems to be heavily affected by the construction of the railroad (Rapinetal. forthcoming, cf. esp. fig. 7). 14 The other stratigraphic section(s) of the Wall (French -Uzbek ‘trench 2 and 3’) apparently show adifferent situation, since the remains of the earliest double -faced structure is said to be built of mud -bricks filled with soil and stones (Rapinetal. forthcoming, fig. 11). 79LADISLAV STANČO Uzundara The site situated upon the eastern slope of the Susiztag ridge (38.146169, 66.947554) and named after the deep narrow gorge descending this way from the summit was discovered by Rtveladze in 1991 and taken to be aKushan fortification (Rtveladze 2002, 103–104). In 1997, Rakhmanov and Rapin recognized that this stone -built fortress with 13 defensive towers does not belong to the Kushan period, as previously thought, but predominantly to the Hellenistic one (Rakhmanov– Rapin 1998, 30). Intensive excavations of the Russian -Uzbekistani team led by N. Dvurechenskaya (RAS, Moscow) started only in 2013 and continue to the present day (Dvurechenskaya 2015; 2018; 2019; 2020a; 2020b; Rtveladze– Dvurechenskaya 2015; Dvurechenskaya– Gorin– Sheyko 2016). Uzundara yielded abundant archaeological material including alarge amount of Seleucid and Greco -Bactrian coins (see below), and became along with Kampyr Tepa and Kurganzol one of the reference sites of what is called northern Bactria. Its significance is yet to be fully understood and appreciated, especially after the whole body of material has been published. According to its excavators, the fort of Uzundara was founded not later than during the reign of the Seleucid king Antiochos I(Dvurechenskaya 2019, 159). N° Name Site code Modern place District LAT LONG ELEV (m.a.s.l.) Surface area Type Pottery (frgmts) Dating Published 23 Munchak Tepa 2 BA5-06 Kofrun B 38.084717 67.248264 823 950 S 45 21 G-B Stančo et al. 2019 24 No name MA_155 Khoja Dagyak D 38.353620 67.007403 1462 --- PS 21 ACH? Stančo et al. 2019 25 No name SA_050 Sarymas B 38.25421013 67.00874328 1894 --- U 1 G-B Stančo et al. 2019 26 No name tower 1 Susiztag B 38.197039 66.985102 1230 --- F 0 G-B? Beľsh 2020, 372, 373. 27 No name tower 2 Susiztag B 38.164738 66.948440 1780 --- F 11 G-B? Beľsh 2020, 373, 376, ris. 3 28 No name tower 3 Susiztag B 38.122909 66.952431 1490 --- F 0 G-B? Beľsh 2020, 372, 373. 29 No name tower 4 Susiztag B 38.112011 66.944986 1550 --- F 0 G-B? Beľsh 2020, 372, 373. 30 No name tower 5 Susiztag B 38.088142 66.932092 1854 --- F 0 G-B? Beľsh 2020, 372, 373. 31 No name SITE125 Panjob B 37.991670 66.870104 1035 --- PS ACH? Unpublished. 32 No name Sarymas B 38.256840 67.011926 1908 --- PS ACH Stančo et al. 2018, 138. 33 No name Toda B 38.204374 67.160610 1410 --- PS ACH? Berdimuradov et al. 2016, 111 34 Parch Sairob B 38.062699 66.952584 1140 --- F 0 G-B? Beľsh 2020, 378 35 Payon Kurgan BA5-18 Tuman Kurgan / Payon Kurgan B 38.159003 67.190708 1067 9100 S 44 G-B? YUE Stančo et al. 2019 36 Sanchil’ Sairob B 38.133911 22 66.955086 1468 --- F 0 G-B? Beľsh 2020, 375–378, ris. 4 37 steppe near Eriell Base 1 BA5-14 Kofrun B 38.043824 67.301230 785 --- PS 32 ACH Stančo et al. 2019 38 Sultan Kul’ DA5-25 Darband B 38.212498 67.017227 1037 --- U 0 [ACH] Stančo et al. 2019 39 Uzundara Susiztag B 38.146169 66.947554 1150 22000 F ALEX-SEL; G-B Dvurechenskaya 2015; 2018; 2019; Dvurechenskaya – Rtveladze 2015; Dvurechenskaya et al. 2016 40 Rabat Rabat B 38.170034 67.200936 1107 --- B YUEZHI Abdullaev 2001; 2002; 2007; Liu et al. 2020 80 STUDIA HERCYNIA XXV/2 Discarded sites Besides the sites listed above that may be with ahigh degree of probability dated to the Hellenistic period and interpreted as settlements and/or forts, there are several sites or rather larger areas claimed by various scholars to be part of the narratives of Alexander’shistorians (Grenet– Rapin 1998; Rapin 2013; 2018; Rtveladze 2002; Sverchkov 2013, 122–150). These are mainly the geological formations of Sarymas and Kapkagli Auzi (see below). Within the framework of the above -mentioned project, we aimed to verify these hypotheses, but finds relating somehow to pre -Kushan periods proved to be extremely rare there, almost non -existent. Analogically, there is no need to look for the ancient road in the Buzgala Khona gorge to the west of Darband (Rtveladze 2002, 111–114; Rtveladze 2019, 176–177), when there is an easier way leading around as the present -day motorway does (Stančo 2018, 141; Stančo– Pažout 2020, 9; Rapinetal. forthcoming). We do not intend to go into details here, since it is not an objective of this text and the literature on this topic is immense. From the archaeological point of view, however, there is not much to say regarding these places and their role in the events related to Alexander’scampaign, but the simple statement: presently, there is not enough evidence (Stančoetal. 2018; 2019). COIN FINDS When interpreting coin finds in relation to archaeological situations, we emphasize statistically meaningful data. In other words, every time we deal with an isolated find, we have to be very cautious about the historical implications resulting from it, whereas such abundant assemblages as the collection from Uzundara– or also from the Darband Wall –, allows for awhole range of analyses. In comparison with other historical regions of Central Asia, the Bactro -Sogdian borderlands present themselves with numerous and rapidly augmenting numismatic data. In the previous overview that was finished in 2018, Ipublished the general figure of 180 Hellenistic coins from known contexts from all over the Surkhan Darya province (Stančo 2020, 271–272, 268, tab.11.1).23 Based upon several preliminary publications, Uzundara alone had yielded 103 out of these pieces by then (Rtveladzeetal. 2014; Dvurechenskaya– Gorin– Sheyko 2016; 2017). Meanwhile, acomprehensive catalogue of all the finds from Uzundara has been published adding dozens of new finds (Gorin– Dvurechenskaya 2018). By now, there are 323 coins known to us from the Surkhan Darya province, and all of these 143 new finds (Tab.2) come from the Baysun and Kugitang piedmonts only! Uzundara, where thanks to the very thoroughly working RAS team dozens of new finds appear every field season and alarge quantity is just being prepared for publication,24 provided us with 69% of the coin finds in the area, while the second highest in yield comes from the Darband Wall (23%), where as many as 629 metal objects were detected in the 2019 field season only, 23 When Abdullaev published all the Hellenistic coins from Bactria and Sogdiana combined known at that time (just 20 pieces!), there was only one isolated find reported from the Baysun District– the above discussed obols of Demetrios Iwith aprovenance stated once as from Dogoba and once from Muchak Tepa (Abdullaev 2006, 108–109). 24 At the conference ‘The Hellenized East: traditions and innovations’ in Tashkent that took place on November 18, 2021, the following general figures were announced by A. Gorin, responsible for the Uzundara finds: the overall number of Hellenistic coins from Uzundara by now: 185 (including 25 new finds in 2021), the overall number of Euthydemos coins: 128 (including 2 silver ones), the overall number of Demetrios coins: 21 (one silver tetradrachm). The remaining 22 coins out of the unpublished Uzundara finds belong to various rulers. Iam very grateful to Alexey Gorin, who kindly provided me with his presentation and up -to -date numbers. 81LADISLAV STANČO out of which 364 were coins. So far 61 of these Darband Wall coins, combined with several finds from the 2018 season, were preliminarily interpreted as Hellenistic ones (closer examination might extend this assemblage by ca. 10 more pieces), but their proper attribution and publication is planned for 2022. Considerably smaller numbers come from the other recently surveyed sites: 9 from Daganajam (where almost no coins from other periods were detected), 5 from Mirzali Kurgan at Khojay Gor, 4 from Iskandar Tepa, 3 from Kurganzol, and the site of Kapchigay -fort yielded asingle coin. Note that without exception, these finds do not belong to trial excavation trenches, but originate exclusively from the surface metal detector surveys. The majority of finds from the foothills of Kugitang and Baysun are very well preserved, especially in comparison with those of the Surkhan Darya lowlands. Consequently, they are easy to read and classify.25 26 25 26 Tab.2: Overview of the Hellenistic coins found in the study area based on the published, but also so far unpublished recent finds.27 What do these coins and their spatial distribution tell us about the region in question? Afirst cautious observation shows that they were found predominantly at sites with primarily– and unquestionably– amilitary function, thus we can easily link them to the stronger presence of soldiers and mercenaries. The earliest,28 even pre -Seleucid coins of the so -called Alexander type (Gorin– Dvurechenskaya 2018, 43–57) come from two sites only: Uzundara and the 25 Figures in brackets show the number including unpublished finds from Uzundara. 26 The reported number of Uzundara finds valid by November 2021 is 185 according to the excavators, but the details of 7 pieces are not known. Fully published were 114 Uzundara finds detected in field seasons 2013–2017. Later on in seasons 2018–2019, some 46 coins were found, while in the last field season so far, 2021, another 25 coins were added. 27 The data in Tab.2 are based on Dvurechenskaya– Gorin– Sheyko 2016; 2017; Gorin– Dvurechenskaya 2018; Rtveladzeetal. 2014; Stančoetal. forthcoming a. 28 Rtveladze quoted an unpublished report of Parfyonov on acoin of Alexander the Great found at Abdulakhan rabat near Shurob– very close to the Darband Wall itself (Rtveladze 1990b, 137; Rtveladze 2002, 157), but Ishare the doubts expressed by Sverchkov concerning the accuracy of this classification (Sverchkov 2013, 136, note 103). RULER / SITE Alexander type Antiochos I Antiochos II Diodotos I and II Euthydemos I Demetrios I Antimachos Pantaleon Agathocles Euthydemos II Eucratides Heliocles OVERALL Daganjam 1 5 3 9 Darband Wall 1 2 6 31 12 1 1 2 4 1 61 Munchak Tepa or Dogoba 1 1 Iskandar Tepa 1 1 2 4 Kapchigay (fort) 1 1 Khojay Gur 3 2 5 Kurganzol 3 3 Uzundara 2 8 (9) 25 1 7 79 (128) 13 (21) (1) 1 (2) 3 (7) 114 (178) 26 OVERALL 3 9 (15) 1 (2?) 15 116 (165) 34 (42) 1 1 2 (3) 1 (2) 9 (13) 1 193 (263) 82 STUDIA HERCYNIA XXV/2 Darband Wall. The coins of Seleucid rulers with 16 or 17 known finds became quite numerous only lately, and are attested at four sites: Uzundara, the Darband Wall, Daganajam, and Kurganzol. Note that these find spots are situated very close to each other. Almost two thirds (61%) of the Hellenistic coins from the study area– including unpublished specimens from Uzundara– belong to Euthydemos I(165), while 16% of the finds were struck by Demetrios I(42), but these are more evenly distributed having been found at six sites in the region compared to five find spots of the Euthydemos coins. The third most numerous coins were those of Diodotoi (Iand II) with 15 finds, thus they are equal to the Seleucids in numbers. It is remarkable that Diodotos is the earliest mintage detected at the sites of the Kugitang piedmonts, i.e., in the southern part of the study region, while around Darband, coins of the first half of the 3rd c. are already well represented. Among the rulers after Demetrios, only Eucratides shows his presence in the Bactro -Sogdian borderlands with seven finds so far somewhat more markedly. Antimachos, Pantaleon, Agathocles, Euthydemos II, and Heliocles are generally rare and typically found at asingle site only. Uzundara and the Darband Wall are the only two sites where an almost uninterrupted sequence of Hellenistic rulers of the region was detected. It is worth mentioning that all these figures represent the current state of research, which is with all probability going to change very soon. For the time being, we leave aside the question concerning the occurrence of coins minted by local rulers after the fall of the Greco -Bactrian Empire and its effects. INTERPRETATION OF THE AVAILABLE DATA AND DISCUSSION While describing the historical development in general and the settlement processes in particular of the given region, we shall proceed chronologically to point out the principal features of the individual periods and the gradual changes. THE ACHAEMENID PERIOD For acorrect assessment of what happened after the campaign of Alexander the Great in the studied part of the Bactro -Sogdian borderlands, it is necessary to summarize what we know about the situation just before that (Pl. 3/1). 29 Larger and smaller sites of the Achaemenid period identified here and there in other parts of the Surkhan Darya Province (Mokroborodov 2015; Wu 2018; Stančo 2018; Havlík 2021) seem to be absent from the mountainous region around the ‘Iron Gate’. During the three years of the Czech -Uzbek team’ssurveys only very few fragments of pottery could have been very cautiously considered as dating to the Yaz III phase roughly corresponding to the Achaemenid period. These were found at Khoja Dagyak (Stančoetal. 2019, 145), in the steppe south of Kofrun (Stančoetal. 2019, 159), and in the Panjob valley (unpublished). In all these cases we find ourselves in rather peripheral areas, not directly in the Machay/Sherabad Darya valley itself. 30 Along the eastern edge of the Sarymas Mountain– above the Machay Darya gorge– only one isolated pottery fragment 29 An outline of the historical development of the Baysun District has been sketched by L. Sverchkov (2005c, 57), but only two sentences were devoted to the Achaemenid Period, while other scholars skipped this period entirely (Annaev– Annaev 2003, 9). 30 Rtveladze based his opinion about the course of the hypothetic old Achaemenid period route through the region on afew fragments of reportedly Achaemenid pottery he claimed to have found on afield close to the village of Old Ak Rabat, but never published (Rtveladze 2002, 110). 83LADISLAV STANČO was found dated with all probability to the period in question (Stančoetal. 2018, 138). Earlier, afew sherds were reported from asite near the village of Toda (Berdimuradovetal. 2016, 111), and the same dating was proposed also for the site of Kala -e Hissor at Sairob (Sverchkov 2005c, 59), though never confirmed by corresponding archaeological material.31 On the other hand, Sultan Kuľ, the site that was previously considered to be of Achaemenid dating (Sverchkov 2005b, 13), was searched by our team in vain and is perhaps not preserved any more (Stančoetal. 2019, 148). Ceramic material, adescription of which matches very well the finds from Khoja Dagyak mentioned above, was reported from the site of Mazar Tepa in the Munchak village, opening up the possibility to date the earliest settlements in this village to the 6th–4th c. BC (Sverchkov 2005b, 14). Some Achaemenid material, such as arrow heads and ceramics, was reported from the surface survey of the Uzundara fortress, but none of it from the excavated archaeological contexts (Dvurechenskaya 2019, 159). The last, though perhaps most important testimony comes from the site of Kurgan -iBolo (Old Baysun citadel), where Rtveladze recently claimed to have found Achaemenid pottery (Rtveladze 2018, 131), which was unfortunately not published and is unknown to previous authors (Sverchkov 2005b, 14, Nr. 36). To sum up, for now, we do not have any confirmed Achaemenid period settlement in the vicinity of Darband and Baysun,32 and the earliest site known so far seems to be the small fort of Kurganzol (Sverchkov 2013, 122), which has been dated to the time of Alexander’scampaign (Sverchkov 2013, 113–116). However, the absence of larger settlements combined with rather haphazard pottery finds here and there throughout the region, may indicate specific land use and subsistence strategies in this region: the piedmonts and even the mountain ridges might have been used as summer (or generally seasonal) pastures for the people otherwise living in the Sherabad lowlands and the Bandikhan micro -oasis. Only asystematic intensive surface survey of the region and / or excavations of the supposedly later sites may bring some fresh data and help to resolve the riddle of this seemingly blank space on the map of the Achaemenid greater satrapy of Bactria. FROM ALEXANDER THE GREAT TO SELEUCUS I(CA. 330–205 BC) Having shown the absence of substantial Achaemenid period archaeological material, we shall continue with the evidence for the ensuing period of Alexander the Great and the following two decades or so of political turmoil in the Upper Satrapies (Pl. 3/2). As shown above, the current state of research does not allow for confirmation of the hypotheses linking the area of Darband and especially the mountains and the so -called cuestas33 in its vicinity directly 31 The surface finds– both pottery and metal objects– point exclusively to the High Medieval period (Stančoetal. 2019, 155–157). In summer 2021, trial excavations were conducted by Beľsh and Bazarov at this site confirming the Medieval dating only (personal communication, September 2021). 32 It is worth mentioning that we also encountered asimilar situation in the Pashkhurt Basin: both an extensive and intensive survey revealed only very scarce evidence of the Yaz III / Achaemenid period material (Stančo 2016, 81 and 83; Augustinováetal. 2017, 126). For an overview of the research and isolated finds see also Mokroborodov 2021. 33 Cuesta (Sp.: slope) means ahill or aridge with agentle slope on one side, and asteep slope on the other. Rtveladze is thus wrong in classifying Susiztag, for instance, as acuesta (Rtveladze 2002, 101). His description would rather match another formation in geology called amesa (Sp.: table), characterized by its flat top and very steep slopes– escarpments, if it were not for the fact that the mesas usually rise significantly from aplain. Kapkagli Auzi, or on amuch bigger scale the nearby mountain of Ketman Chapty towering over Darband and Toda, would fit much better the definition of acuesta. The term was later incorrectly used also by other scholars (Sverchkov 2013, 136). 84 STUDIA HERCYNIA XXV/2 to the events described in the Greek and Roman written sources and related to the particular episodes of Alexander the Great’scampaign. In fact, our current project started among other targets with an ambition to confirm or to refute such hypotheses promoted especially by Cl. Rapin (2013; 2018) and E. V. Rtveladze (2002), thus asubstantial part of our field work capacities in seasons 2017 and 2018 was deployed to fulfil this task.34 Even if we had relevant archaeological material from the mountains such as Sarymas and Kapkagli Auzy– which is not the case –, we would not be able to prove that this hypothetic material belongs exactly to the narrow period in question. The exact itineraries of Alexander’sarmies, as well as the location of individual places of refuge remain unknown,35 despite the long -lasting efforts of renowned scholars to prove otherwise (Grenet– Rapin 1998; Rapin 2013; 2018). In my opinion, these hypotheses should not be taken for granted and repeated as such (Martinez -Sève 2020, 84–86) without further scrutiny. Unlike the mountain ridge of Susiztag, where not only the fort of Uzundara, but also several other fortification structures were discovered (Beľsh 2020), neither the mountain of Sarymas (the alleged Arimazes’ rock: Rapin 2018, 289–290), nor the elevated plateau of Kapkagli Auzy (the presumed Sisimithres’ rock: Rapin 2018, 292) yielded relevant archaeological material, be it the remains of afortification, or chronologically fitting small finds (Stančoetal. 2017, 139–141; Stančoetal. 2018, 170), except for afew pottery sherds dating generally to Antiquity (but definitely not to the Late Achaemenid or Early Hellenistic period) found at one isolated spot at Kapkagli Auzy (Stančoetal. 2018, 146, 160).36 At these sites, asampling surface survey with ametal detector was also employed with rather poor results compared to the other investigated sites. Despite it, we suggest asystematic metal detector survey as apossible means of future research. Ido not intend here to reject the very idea of these mountains’ identification with the places mentioned in the Greek written sources for good, the current state of research simply does not allow us to accept it. Moreover, even if we accept the possibility that they were used as temporary places of refuge (which is the function described by Curtius and Arrian), we cannot– Iassume– call them ‘fortresses’ (Martinez -Sève 2020, 84–86) of the borderland chieftains implying certain intentionally built structures and apermanence of occupation. Coins, otherwise avery eloquent source of information on Bactrian history, remain almost silent about this early period: so far, we do not have asingle find of aSophytes coin from the borderlands. The only pre -Seleucid coins known so far are the so -called Alexander type coins (minted after the death of Alexander and before the beginning of Seleucus I’srule in Central Asia at various places in the Mediterranean): two of them were found in 2016 at Uzundara, and asingle coin of this type was detected in 2019 at the Darband Wall. It is probably unnecessary to emphasize that such few and isolated finds cannot be used as arguments for any activity at the site at the time of their minting or shortly afterwards. Significantly more plausible is the assumption that they remained in circulation deep into the Seleucid period. 34 Note, however, that for us it was more important to find an answer to the question of whether alarger group of people might have lived up the Baysun mountain ridges for alonger time period, than to find aproper name of the place that persisted in the Greek written sources. 35 In this point Iagree with Sverchkov in his opinion that ‘the further route [beyond the Oxus] of the Greek army, except for the general direction to the north -northeast, cannot be reconstructed in detail’ (Sverchkov 2013, 122). 36 It is worth mentioning that we have the same doubts about interpreting the mountain of Kyz Kurgan near the village of Shina in the upper Surkhan Darya region as the place of Chorienes’ Rock (Rapin 2013, 75; Rapin 2018, 288–292; Martinez -Sève 2020, 84–86) for the very same reasons– the absolute lack of archaeological evidence (Stančoetal. 2018, 147, 170). Our task as scholars is to doubt and dispel constantly repeated assertions with no real basis. 85LADISLAV STANČO Not only the identification of the ‘rocks’ of local noblemen, but also that of six forts allegedly founded by Alexander during his 328 BC campaign, have been attempted. For Sverchkov, two of these forts could be identified as Kurganzol and Payon Kurgan (Sverchkov 2013, 133–134). The rather small fort of Kurganzol (ca 700 m² of inner space) with only nine rooms in three separate humble buildings does not offer any comparison with the known Greek fortification architecture. On the contrary– round ground -plans are well attested in Achaemenid Central Asia as Sverchkov admits (Sverchkov 2013, 17). Thus, even if one accepts the foundation date of the fort of 328 BC established by the dendrochronological method, there is no reason to connect the construction of the fort with the Greek / Macedonian army. Regardless of the authority giving orders, Kurganzol– unlike the stone -made Uzundara– was built by local people employing traditional concepts (circular ground -plan) and construction techniques (mud -bricks, timber). As amatter of fact, the Kurganzol fort– regardless of its later function– might have originally been built by locals to support the advance of Alexander’sarmy, or just as well to prevent it, or at least to slow it down. One could hypothesize about the material culture typical for the local inhabitants vs. that of the western newcomers, but unfortunately, by now we lack such aclear distinction and only the presence of pottery shapes of presumably western origin (without alink to the earlier local tradition) leads us to the assumption that the contexts in question relate to Greeks and Macedonians. As for the second out of the six alleged forts built by Alexander the Great, Payon Kurgan might have played the role of Alexander’sstronghold only hypothetically: among the published evidence, there is neither any information about layers dated to the late 4th century BC, nor any hints at fortification. Were it that easy to prove, one can find enough generally Hellenistic sites in the neighbourhood of the Darband Wall to choose six of them for this purpose. The absence of Late Achaemenid (and do we recognize it for sure?) material at these sites prevents us from such an interpretation. To sum up, we can assume that Alexander’sarmy marched through the Darband area, perhaps even more than once, but as for the precise identification of the places mentioned in the written sources, we must remain very cautious. Concerning the last decades of the 4th century BC, the only site that functioned in the vicinity of the Iron Gates was the small mud -brick fort of Kurganzol. Considering its tiny size and peripheral location, it hardly played any important role in the policy of the local rulers prior to the accession of the Seleucids. THE SELEUCID PERIOD (CA. 305–250 BC) Even if it is not possible to agree with Sverchkov’sstatement that ‘around the beginning- -middle of the 3rd century BC, Baysun Margania had been deserted again and the defence of the northern territories became vital again only at the end of the 3rd century BC, when on the borders of Greco -Bactria there appeared anew, much more serious enemy’, (Sverchkov 2005c, 60) the data currently available do not allow us to form detailed conclusions about the first half of the 3rd century (Pl. 3/2). According to the early material and especially coins (Tab.2), only some forts were in use at that time: Uzundara and the Darband Wall in its early form were most probably established simultaneously at that time, more precisely during the reign of Antiochos I(cf. Stančo 2020, 262), while Kurganzol continued its existence, having been founded two generations earlier. Among the rural settlements, only that of Daganajam might have been settled in this early phase. From the other sites, there is not enough evidence to support Seleucid dating. In case this time period was really characterized by the efforts of Antiochos Ito bring colonists from the Mediterranean and to found new cities in the lowlands of Bactria and Margiana (see below), it did not affect the Bactro -Sogdian borderlands. 86 STUDIA HERCYNIA XXV/2 Now, why should we expect that the territory was already settled and fortified in the time of Antiochos I? First, it was indeed ageneral Seleucid policy to bring many asettler group deep into Central Asia to found new cities (Strootman 2020, 17–20). Those would naturally need protection, not necessarily from an outside threat, but definitely from local ‘barbarians’, nomads and semi -nomads roaming this area at will. The most natural policy would have been the establishment of both military garrisons and rural settlements nearby. Second, the growing number of Antiochos Icoins from the Darband micro -region seems to exclude the otherwise acceptable explanation that these were used for along period of time and served well also the Greco -Bactrian community. The finds from Uzundara (9 pieces), Kurganzol (3), the Darband Wall (2), and Daganajam (1) may be already taken as evidence attesting to the presence of Seleucid military garrisons within the territories– not on the edge!– of the empire. The simple result of this reasoning is that afortification does not mean aborder of astate. Cl. Rapin arrived at avery similar conclusion (although speaking of the Greco -Bactrian period) assuming that the Darband fortification system including Uzundara served to ensure ‘economic and military control over the main communication route between Bactra and Samarkand’ (Rapinetal. forthcoming). THE GRECO ‑BACTRIAN PERIOD (CA. 250–140 BC) The situation in the researched region changed dramatically at this time (Pl. 3/3) when we observe the establishment– or in some cases maintenance– not only of fortresses, but also of rural settlements: the system of central sites of the micro -oases seems to have developed in the second half of the 3rd century BC.Thus, we may speak of Machay Kurgan and its position in the upper Machay Darya valley, Daganajam in the upper Sherabad Darya valley, Munchak Tepa 1 and 2 in Kofrun, Kapchigay in Darband, and for the 2nd–1st century BC also about Payon Kurgan, Iskandar Tepa, and Mirzali Kurgan at Khojay Gor.37 To further complicate the matter, we must admit that the primary function of some of these settlement sites remains unclear. Without their detailed investigation we cannot exclude that these were basically also fortified watch posts or some other sort of small military installations. It is their central position, typically in the very centre of an elongated piedmont river valley, allowing for the control of asubstantial stretch of the water stream and surrounding micro -oasis. Amilitary function was considered by scholars in the case of Munchak Tepa 1 (Mokroborodov s.d.), Kapchigay– lower settlement (Sverchkov 2005c, 60; Martinez -Sève 2017, 289),38 Iskandar Tepa (Stančoetal. 2017, 132), and Payon Kurgan (Sverchkov 2013, 133–134), even though without fundamental evidence of the matter. The spatial relations of the central sites and the micro- -oases themselves including an analysis of the capacity to sustain not only the oasis population itself, but also military garrisons in the neighbourhood, deserve special attention, which is beyond the scope of this contribution. 37 Whether to see the sites in the Kichik Ura Darya valley, such as Bilibayli Kurgan (Tab.1), in asimilar way remains to be clarified. For now, we consider them apart of the post -Hellenistic settlement changes. If, though, there were confirmed settlement sites similar to Daganajam, Munchak in Kofrun etc., it would be much more difficult to hold the position in the scholarly debate explaining the Darband area as aborder between two individual political entities. 38 Sverchkov (2005b, 13–14), whom Martinez -Sève quotes as her source, does not speak about afort, he just mentions the function of the site, which according to him was blocking the entrance to the Machay gorge. 87LADISLAV STANČO There are several breaking points in the history of Bactria in the Greco -Bactrian period that might, but also might not be visible in the archaeological record here in the Bactro -Sogdian borderlands. First, the declaration of Bactrian independence by the local Seleucid satrap Diodotos from his distant sovereign (be it Antiochos II or Seleucus II) taking place in the 250s or 240s BC.39 The second such event was the successful rebellion of Euthydemos Ibased maybe in Sogdiana (Lerner 1996; see also Stančo 2020, 225–266) against King Diodotos, second of this name, in the 220s BC. 40 The third would be the war between Euthydemos and AntiochosIII claiming back the Seleucid Central Asian dependencies. The first event seems to have left no traces in this region, which makes perfect sense, since Diodotos Iwas apparently able to keep both Bactria and Sogdiana including its capital Maracanda for the initial part of his reign (Rapinetal. forthcoming; Lyonnet 2020, 323) withdrawing later southwards to Kashkadarya as attested by the dating of Erkurgan between 250 and 150 BC (Sulejmanov 2000, 163–166; Lyonnet 2020, 322–321). Even if Diodotid coins are rare in the borderlands, the number of their finds increases steadily especially at the military bases of Uzundara (7 sp.), the Darband Wall (6 sp.), and Kapchigay– slope (1 sp.). Bearing in mind Euthydemos’ industrious attempts to define himself against the Diodotids from the very beginning by his own extensive coin issues (thus hypothetically not letting Diodotid coins circulate), we may cautiously assume that these coin finds really mirror the reality of the 240s and 230s. It means that Diodotoi most probably held the fortifications (or natural strategically located places fortified alittle later) in the borderlands controlling at the same time regions to the southeast and northwest of them (Lyonnet 2020, 83–85). In this point the new evidence seems to contradict what Isuggested in 2018, i.e., that the construction of the fortification system ensued after the departure of Antiochos III to India (Stančo 2020, 262). This leads us to the rebellion of Euthydemus and his role in the shaping of the borderlands. Iam not going to repeat my arguments concerning the interpretation of Uzundara and other sites that have been published recently (Stančo 2020, 261–266). After all, Iam not the only one who expressed some doubts about the function of Uzundara in the system of borderland fortifications (Rapinetal. forthcoming). One of my key arguments is related to the location of the fort at aplace better suited to watch the upper Sherabad Darya valley and to protect this place from aprospective enemy coming from the east (or passing by in the eastern foothills of Susiztag and Kugitang). The most recent discoveries of various fortification elements in the vicinity of Uzundara and Sairob (Beľsh 2020), if they really belong to the period in question, offer asimple explanation: they might very well have prevented not only the armed enemy, but also– or more so– the ordinary travellers, merchants, local semi -nomad pastoralists etc., from travelling in the foothills of Susiztag, forcing them to wander along the Sherabad River itself via Munchak Tepa (village), and Daganajam towards Kapchigay and further to the north. What the reason was for such arestrictive precaution is not known. At present, however, the precise dating of these defensive structures is open to debate, as is the original intention of their builder. One of the crucial issues repeatedly addressed by scholars dealing with the Greco -Bactrian policy and especially with the conflict between Euthydemos Iand the Seleucid king Antiochos I concerns the threat of nomadic incursions from ‘the north’. Now, what did Polybius say exactly? It is the notorious– reportedly Euthydemos’– sentence, stating that ‘…if [Antiochos] did 39 For adiscussion about the later date of the Bactrian secession from the Seleucid Kingdom see Jakobsson 2020. 40 Or against the hypothetic third Diodotid king Antiochos Nikator (Jakobsson 2020, 500–503). 88 STUDIA HERCYNIA XXV/2 not yield to [his, i.e., Euthydemos’] demand, neither of them would be safe: seeing that great hordes of Nomads were close at hand, who were adanger to both; and that if they admitted them into the country, it would certainly be utterly barbarised’ (Polybius XI, 34). Thus, if we accept the principal idea of anomad threat, there is no hint whatsoever as to the direction or aplace, from which these nomads were supposed to come. Here Iprefer to agree with the opinion expressed by Martinez -Sève that the nomads were rather an integral part of Central Asian societies than amere external threat (Martinez -Sève 2017, 288–289; Martinez -Sève 2020, 90). It is indeed avery good starting point for the discussion. On the other hand, new waves of nomads regularly migrating from the Eurasian steppe belt was areality and we can only surmise that not only sedentary populations in Bactria, but also their (semi)nomadic neighbours were alarmed by foreign nomads, especially when these arrived in large numbers. Based only on Polybius’ quote, we cannot in any case assume that these nomads came from the north. Even less probable seems to be the Bactrians’ fear of their Sogdian neighbours, who consisted of asimilar mix of asedentary and semi -nomadic population, after the latter gained their independence in the mid-3rd (?) century BC (Lyonnet 2020, 324). In any case, it cannot be accepted that the findings of the fortifications on the Susiztag ridge could put an end to the discussion on the border location between Bactria and Sogdiana (Dvurechenskaya 2019, 159). The remnants of afortification system– if it is truly acomplete system– need to be viewed much more comprehensively and approached without conclusions about their purpose and functioning made in advance. Aborder is too complex aconcept to be reduced to asingle cluster of small forts. In this way, advocates of the border being situated on the Amu Darya– and Ido not belong among these either– might take the presence of the whole chain of fortified points there in both Greco -Bactrian, and in the Kushan period as an argument in favour of their hypothesis. For example, isolated round towers like those located at Susiztag could have been constructed in aone -off military campaign or in an isolated conflict, and were not necessarily part of along -term strategy to protect the northern (or other) border of astate.41 For linear fortifications, it is necessary to look for suitable analogies in the Greek environment– based on the assumption that the structures were actually built by the Greeks and Macedonians (which was most likely not the case of Kurganzol). The excavations of Uzundara enriched our knowledge of everyday life in the Bactro- -Sogdian borderlands also through the rich finds of animal bone remains (Dvurechenskaya, S.O. 2020). More than 5000 analysed samples show, when compared with three other already published sample sets (from Kurganzol, Kampyr Tepa, and Kyzyl Tepa), astrikingly different picture, especially avery high proportion of donkey and dog, but an unusually low proportion of cattle (Fig. 2). Asignificant representation of transport animals, moreover growing over time, as well as their attested severe physical exertion and bone injuries, has been explained logically as related to their role in ensuring the supply of food for the garrison before fulfilling their other task: becoming apart of the garrison’sdiet themselves (Dvurechenskaya, S.O. 2020, 390). We can compare this specific anomaly with the enormous proportion of horse at the site of Kampyrtepa, which has been discussed elsewhere (Stančo 2020, 270–271). The representation of sheep and goat, on the other hand, is very similar to that of Kurganzol and Kyzyl Tepa, oscillating in the sample sets around 70%, and also the percentage of camel bones from these three sites, even if quite low, is equally rare, oscillating around 0.2%. The fact that except for deer, wild animals are very rarely encountered in the fort (Dvurechenskaya, S.O. 2020, 387), clearly demonstrates the garrison soldiers’ dependence on domesticated animals kept for meat production. 41 For apossible interpretation of isolated towers see Müthetal. 2016, 9. 95LADISLAV STANČO Sverchkov, L. 2005b: Archaeological sites of Baysun District. Review of Baysun Scientific Expedition. Труды Байсунской научной экспедииции 2. Tashkent, 10–20. Sverchkov, L. 2005c: Baysun. Trial historical reconstruction. History and traditional culture of Baysun. Review of Baysun Scientific Expedition. Труды Байсунской научной экспедииции 2, 56–65. Sverchkov, L.M. 2008: The Kurganzol Fortress (on the History of Central Asia in the Hellenistic Era). Ancient Civilizations from Scythia to Siberia 14, 123–191. Sverchkov 2013 = Сверчков Л.М.: Курганзол. Крепость Александра на юге Узбекистана. Tashkent. Sulejmanov 2000 = Сулейманов Р.Х.: Древний Нахшаб. Samarkand– Tashkent. Vích, D. 2014: Příspěvek kmetodice detektorové prospekce varcheologii. [Ein Beitrag zur Methodik der Metalldetektorprospektion in der Archäologie]. Archeologie východních Čech 7, 152–172. Wu, X. 2018: Exploiting the Virgin Land. Kyzyltepa and the effects of the Achaemenid Persian Empire on its Central Asian frontier. In: Lhuillier– Boroffka eds. 2018, 189– 214. Wuetal. 2015 = Wu, X.– Miller, N.F.– Crabtree, P.: Agro -Pastoral Strategies and Food Production on the Achaemenid Frontier in Central Asia. ACase Study of Kyzyltepa in Southern Uzbekistan. Iran 53, 93–117. Ladislav Stančo Institute of Classical Archaeology Faculty of Arts, Charles University Celetná 20, CZ-110 00 Prague 1 [email protected] 158 STUDIA HERCYNIA XXV/2 Pl. 3/1: Map of the Darband Wall area with spots with reported finds of the Achaemenid period. Map by L. Stančo. Basemap: Esri. 159PLATES Pl. 3/2: Map of the Darband Wall area with sites belonging to the late 4th– early 3rd c. BC (Alexander to Seleucus I). Map by L. Stančo. Basemap: Esri. 160 STUDIA HERCYNIA XXV/2 Pl. 3/3: Map of the Darband Wall area in the Greco -Bactrian period, small dots indicate the sites with uncertain dating. Map by L. Stančo. Basemap: Esri. 161PLATES Pl. 3/4: Map of the Darband Wall area in the Yuezhi period. Map by L. Stančo. Basemap: Esri.