3 Nomad Migration in Central Asia KAZIM ABDULLAEV Institute of Archaeology, Samarkand THE ETHNIC AND CULTURAL IDENTITIES AND MIGRATION ROUTES of Central Asian nomadic tribes present one of the most disputed questions of those related to the Kushans. The tribes appear in literary sources preserved mainly in Chinese chronicles and in scarce references in the works of Greek and Latin authors. The migration of Central Asian nomads, particularly into Transoxiana can be divided into two categories. The long ‘trans-regional’ route is ascribable to the Yuezhi migration from the valley of Gansu, on the northern borders of China, to the territory north of the Oxus River (Amu Darya), while the migration of tribes like the Dahae, Sakaraules, Appasiakes, Parnes etc. can be classified as ‘local’ movements. Strabo (XI, VIII, 2) gives a description of the locations of nomad peoples. ‘The majority of Scythians, beginning with the Caspian Sea region, are called Dahae. The tribes living to the east of them are the Massagetes and Sakas; the others are commonly known as Scythians, although each tribe has its own proper name. All of them are nomads. Of these nomads, the best known are those who annexed Bactriana from the Greeks, namely the Asii, Passiani, Tochari and Sakaraules, who migrated from the other side of the Jaxartes near the regions of the Sakas and of the Sogdians occupied by Sakas’. The second part of this information is important, but complicated at the same time. This fragment provides information about the territory north of Bactria, if we take account of the fact that Strabo considered the River Oxus to be the border of the regions now in modern Kashka Darya, Surkhan Darya and southern Tajikistan. However, the last two territories are associated in scientific literature with northern Bactria. As experience of historic and archaeological investigations shows, however, the paramount goal in determining migration routes is the ethno-cultural identification of the remains of nomad people along their routes. Decisions within this task depend on the nature of the material currently available. The literary sources concerning this problem have a contradictory character and insufficiently reflect the course of historic events. Proceedings of the British Academy 133, 73–98. © The British Academy 2007. 74 Kazim Abdullaev The indicated geographic points have been the subject of discussion until now. The localization of ancient regions and cities remains hypothetical. The analysis of literary information without supplementary material (in this case archaeological) leads sometimes to conclusions far from the reality, such as those of Borovkova (1989). Zadneprovsky (1997a and b) justly criticizes Borovkova’s localization of the Yuezhi on the bank of the Syr Darya, but his own location of them in the Zerafshan valley is also disputable. Finds of Kushan coins in this area are extremely rare, but their abundance to the north of the Oxus shows that the Yuezhi settled there and created the core of the future Kushan kingdom. The events determining nomad migration in Central Asia are connected with the history of the northern and western borders of Han China in the second century BC. The Shiji (123), the Chinese ‘Book of History’ by Sima Qian, describes the conflict between the powerful confederation of the Xiongnu and the Da Yuezhi. The latter roamed between Dunhuang and the Qilian mountains (the Tian Shan of Gansu). In 176 BC, after suffering a crushing defeat by the Xiongnu and their leader Maodun, and later, Laoshang, the Da Yuezhi moved west, crossed the kingdom of Da Yuan and reached the banks of the Oxus. The Chinese ambassador and diplomat Zhang Qian met them there in about 129/128 BC. He found them already settled on the right bank of the Oxus. In this period they had not yet occupied southern Bactria (Da Xia), but already had power over it. Besides this, it can be determined from the ancient source, that the arrival of the Da Yuezhi on the banks of Oxus is impossible after this date. It is clear that they passed some time in the lands lying between their initial homeland and the banks of Oxus. Although the Shiji does not mention this, the gap is partially filled by the Han Shu (Chronicle of the Han Dynasty): the first and most important event which it relates is a Da Yuezhi attack on the Sakas. The second related event concerns Kunmo (king of the Wusun), who pursued the Yuezhi, pressing them to move west. Both events happened before the death of Shanyu, who reared Kunmo, that is to say before 160 BC, a date too early for the conquest of Bactria (Pulleyblank 1966; 1970). As for the Sakas, the same Han Shu chronicle (Han Shu, 96A, 5463) indicates that in the period when the Chinese established contact with Kashmir (Jibin) by a Suspended Crossing, they knew that the local rulers were Sakas, who had come from the North. They were also told that the king of the Sakas had been forced out of his homeland in the Pamirs by the Yuezhi. Crossing Da Yuan, the Yuezhi found themselves to the north of the Oxus. It is interesting that the Chinese sources do not mention the other, no less important, river of Central Asia, namely the Jaxartes (Syr Darya). In the opinion of some Chinese scholars, the Yuezhi tribes moved along the Yellow River and afterwards continued to migrate westwards to the Yili River valley (Lu Enguo, 2002). Probably, by moving in a westerly direction and crossing Da Yuan, the Yuezhi skirted the Altai and Turkestan ranges around their northern extremities and found their way through Usrushana into Sogdia (Abdullaev 2001). NOMAD MIGRATION IN CENTRAL ASIA 75 One of the most important sources of information on nomad migration in Central Asia is Justin’s Prologue to Pompeius Trogus (prologue to book XLII), which states that ‘the Asiani are kings of the Tochari and destroyed the Sacaraucae’ (Reges Tocharorum Asiani interitusque Sakaraucarum). It is possible to conclude from this extract that the Asiani and the Tochari were closely related tribes. What is more, it indicates that the ‘Asiani’ dominated the ‘Tochari’ (Reges Tocharorum Asiani). We can identify the ‘Asiani’ with the ‘Kushans’ (von Gutschmidt 1888; Haloun 1937; Bachhofer 1941; Daffina 1967), one of the leading tribes, which subsequently came to power and created a great empire. It is noteworthy that Justin says that the Tochari were ruled by the Asiani, while the Chinese sources identify them as the largest of the five Yuezhi principalities. In my opinion it is very possible that the name Da Yuezhi in the Chinese chronicles (for the early history of Kushans) was unknown in the West and therefore was not mentioned in the ancient western sources. Whereas the Chinese continued to call the tribes of Yuezhi by this name even after their migration, the Greek and Latin authors gave them different names. In this case the identification ‘Asioi-Tocharoi ⫽ Da Yuezhi; Pasianoi; Sakaraukai’ (Daffina 1967, 52) is disputable. The second part of Justin’s sentence is also important, for it gives information about the destruction of the Sacaraucae. I propose that ‘the Asiani kings of the Tochari’ opposed the Sacaraucae, and that this confrontation took place between two completely different ethnic groups. One can also recall that Strabo’s list of the main tribes who conquered Bactria from the Greeks included the Sacaraucae (Bailey 1985 and Daffina 1967 are among scholars who have discussed the variant forms of this name, Sakaraules and Sacaraucae, but only differing in the interpretation of the ethnonym itself). Thus the second part of the sentence concerns the war of Asiani and Tochari against the Sacaraucae. The latter were settled on the territory of Sogdia and Bactria in the period before the arrival of the Yuezhi. The Wusun problem and archaeological data After being defeated by the Xiongnu, the Yuezhi migrated westwards. The first to clash with them were the Wusun. According to the Chinese chronicles, the Wusun initially inhabited territory in eastern Turkestan prior to being defeated by Xiongnu in 176 BC. In about 160 BC, the Wusun moved to the area of Semirech’e following the same path as the Yuezhi. According to this version of events the locally found Wusun cultural remains must be dated no earlier than the second century BC. However, some investigators also attribute monuments of both earlier and later periods to the Wusun culture. They identify as Wusun most of the necropolis remains of the third century BC to fifth century AD in Semirech’e, Tian Shan, the valley of Talas, and at the foot of Karatau. This interpretation is not sufficiently substantiated as it is based on the absence of reliable chronology and a contradictory interpretation of literary sources. 76 Kazim Abdullaev The Chilpek, Buranin and Karakol necropolises were the first monuments of the ‘Wusun culture’ archaeologically excavated in the twentieth century by Gryaznov and Voevodsky. They date no earlier than the third century BC to the first century AD, on the basis of a study of the decoration, particularly of the gold rings and buckles. The presence of imported Chinese objects datable to the end of the third century BC has been used to establish a chronology. The sewn buckles were compared with objects of Sarmatian culture, as well as with material from the tumuli burials, kurgans of Shibe and Katanda in the Altai, the latter dated to the second century BC to first century AD (Voevodsky & Gryaznov 1938). However, later investigations showed that the Chilpek group of burials belong to a later date, while the Buranin and Karakol finds could be as early as the fourth century BC (Barkova 1978; 1979; 1980). The greatest contribution to the study of Wusun culture was made by Bernshtam, whose principle theory is based on the idea of the long continuity of the development of the Wusun culture and not just the migration of this tribe in the second century BC. Bernshtam introduced the term ‘Saka-Wusun’ to include cultural monuments of the preceding period as well as the previous epoch. Monuments such as the Semirech’e and Talas valley burials were attributed to this culture. They included ground-level tombs, with a mound erected over them, in which the deceased was placed in an extended position with its head oriented to the West. The necropolis of Berkarin is the archetypal monument of this kind. It includes two groups of burials belonging to two distinct periods: Saka (fourth to third century BC) and Wusun (second century BC to first century AD). However, it has a wider chronological framework for additional objects of a later period were also found here (Moshkova 1992, 27). Bernshtam (1952, 50) also identified the necropolises and partial burials of Alamyshik and Jergetal and the necropolis of Sokolovka and Jerges— dated by him to the second century BC to the first century AD —as monuments of Wusun and Yuezhi type. The dating of the Sokolovka necropolises was based on the analogies between the wood ceilings of its tumuli burials and the constructions of the Pazyryk burials. He dated the latter to the third century BC. However, according to the latest data, archaeologists are inclined to date the construction of Pazyryk — like the monuments of Shibe, Katanda and Berel — to the fifth to fourth century BC. Consequently, the necropolises of Sokolovka must be identified as monuments of the Saka period. As for the Jerges necropolis, the date of the material (pottery) found in it cannot yet be determined. The remains concentrated in the valley of the Ili River and excavated by Kushaev are associated with Wusun culture. Based on the peculiarities of the burial constructions and a study of the material, he identified three distinct phases: third to second century BC, first century BC and first century AD (Kushaev 1963). But these phases are only acceptable as a relative chronology, while the absolute chronology remains undetermined. In the Tian Shan mountains, the monuments of the Wusun period were explored by Kibirov (100 burials in 19 necropolis). Based on the information in literary sources and NOMAD MIGRATION IN CENTRAL ASIA 77 interpreting the studied remains as Wusun, the author links them with the mass migration of the Wusun population under pressure from the Juyans. The monuments have a wide chronological framework down to the fifth century AD (Kibirov 1979). From the archaeological data one can sum up the situation as follows: the nomad monuments are represented by several forms of burial construction which continued in use over a long period. Some of them precede the Wusun arrival in the second century BC and ‘present the remains of one people and belong to different periods of its historic development’ (Kozhomberdyev 1975). Amongst the great quantity of monuments of north-eastern Central Asia, one can only ascribe a certain number as the remains of the Wusun culture (Mandel’shtam 1983, 47–8). At the same time there are no reliable signs for determining this definitive culture. The graves of the Talas valley, Semirech’e and Tian Shan are more probably linked with the tribes of nomads, i.e. the cattle-breeder descendants of the Sakas. One of the widespread types of burials in the nomad culture of Central Asia is the so-called podboy, or tomb with an underground chamber and a side niche (alcove). The chronological frame for this type of burial extends from the mid-first millennium BC to the first half of the first millennium AD. The variations of certain details in the burial construction and the character of the contents are linked to differences in landscape and tribe. Podboy burials show such variations as the position of the chamber in relation to the entrance and the orientation of the deceased. Catacomb burials are distinguished by an entrance passage and the position of the catacomb itself: for example, in one type the distinctive feature is the continuation of the entrance passage and in another it is the perpendicular position of the catacomb in relation to the entrance passage. The question of ethnic attribution was posed by Bernshtam at the time of the discovery of the Kenkol necropolis. Analysing the material of the burials (the arrowhead and pottery forms and the presence of a silk textile dated first century BC to first century AD), he concluded that this type of burial was made by the Xiongnu. However, this attribution was criticized by Sorokin (1954 and 1956), whose analysis of the remains demonstrated the local character of the culture. He also proposed an alternative date for the Kenkol necropolis in the second to fourth century AD. Several investigators considered podboy and catacomb necropolises as evidence of a local culture of a population with certain imported elements (Moshkova 1992). There were attempts to distinguish certain groups of these burials as characteristic of different ethnicities. Zadneprovsky (1960, 137), for example, linked some of them with the Yuezhi migration, particularly the necropolis with podboy type tombs. However, this correlation is not supported by Chinese archaeologists such as Lu (2002), when excavating podboy tombs and similar burial constructions discovered in China (Daodunzi in Ningxia province in Northern China, Subeixi, Chawuhu III in Eastern Tian Shan, Hamadun in central part of Gansu province) as they identify the Chawuhu necropolis in particular as remains of the Xiongnu. 78 Kazim Abdullaev Podboy and catacomb necropolises are clearly present in Ferghana. Litvinsky (1972, 64–72) devoted some of his work to them. He traced elements of the Sarmatian culture in these burials. In his opinion, the catacomb burial construction forms arose in the Bronze Age and have parallels with the Saka kurgans in Besshatyr and Semirech’e, and the kurgans of the Lower Syr Darya (Chirik-Rabat). The burial monuments of Sogdia The greatest contribution to the study of the tumuli necropolises of the Samarkand and Bukhara regions has been made by Obel’chenko (1961). In the Samarkand area, he excavated the Agalyksay, Akjartepa, Mirankul and Sazagan necropolises. In the Bukhara region he investigated the necropolises of Kalkansay, Kyzyl Tepe, Kuyumazar, Lyavandak, Shakhrivayron, Khazara and Yangiyul (Fig. 1). Amongst the rich diversity of burial constructions, one can distinguish three main types: podboy, catacombs and graves. He associated the podboy burials with the Yuezhi migration, but the other two with the Sarmatian culture of the southern Urals and the lower Volga. However, the archaic appearance of a catacomb burial in the Kuyumazar necropolis (kurgan no. 3, near the Soinov mound) gave rise to some doubts about its strange character. The sub- Figure 1. The Map of the main Nomad monuments in Sogdia and Bactria. 1 Kalkansay; 2 Yangi Kurgancha; 3 Akjar; 4 Aksay; 5 Sazagan; 6 Mirankul; 7 Agalyk; 8 Kuyumazar; 9 Kiziltepa; 10 Shari Vayron; 11 Khazara; 12 Babashov; 13 Kokkum; 14 Aruktau; 15 Tulkhar; 16 Rabat-I, II, III; 17 Airtam. NOMAD MIGRATION IN CENTRAL ASIA 79 sequent discovery of the second millennium BC Zamanbaba necropolis in the Zerafshan valley confirmed that this form of burial was of local origin. The Zamanbaba necropolis included burials with an entrance passageway and a catacomb (Gulyamov et al. 1966, 119–29). The presence of such burials during the late Bronze Age (ninth to eighth century BC) has also been attested in southern Tajikistan. Obel’chenko’s conclusion that ‘the participation of the tribes of the Sarmatian world in the crushing defeat of the Graeco-Bactrian kingdom gives a somewhat different nuance to the Chinese Chronicles’ (Obel’chenko 1992, 230) is, however, disputable, as the theory of a Sarmatian invasion of Central Asia in the second to first century BC is not acceptable to other scholars (Mandel’shtam 1974; Zadneprovsky 1997b, 77; Abdullaev 1998a and b, 24–5). In the late 1980s several kurgans near the village of Orlat in the Samarkand region were excavated. The site became well known thanks to some remarkable bone plaques depicting various scenes of a battle, hunting and zoomorphic motifs. The catacomb burial construction type at Orlat, i.e. with the chamber at the axis of the entrance passage mainly oriented north-south, is similar to that at other Samarkand region necropolises, although a certain number have a west-east orientation. The burial finds include pottery and weapons: swords, daggers, arrowheads of a nomadic type, bone appliqués for bows etc. Pugachenkova (1989, 122–54) links the Orlat necropolis to the Kangju tribal group who held the territory to the west of Samarkand in the second to first century BC. In the same region of Samarkand (Koshrabat district), the excavation of the Sirlibaj Tepe kurgan revealed burials of various types and periods. In one of them, there was a catacomb located perpendicular to the axis of the entrance passage. The corpse was placed in an extended position, with the head oriented to the south, on a bed of twigs on the floor. A second burial was located in line with the axis of the entrance passage, which had five steps, with the remains of a coffin in the catacomb. A third burial, of podboy type, was discovered in the sub-burial construction. Its niche was located in the north-western wall of the grave. The head of the corpse was oriented north-east. Finally, a fourth burial was a grave with an entrance passage (the body was missing). The entrance passage and the grave are oriented on the same north-west to south-east axis. In the opinion of the excavators, this is the earliest of the excavated burials. It has analogies with the finds of the southern Urals (fourth century BC) and the necropolis of Tagisken in the eastern Arals (fifth to fourth century BC). The podboy burial is similar to that of the Orlat kurgan no. 9, dated first century BC to first century AD (Ivanitski & Inevatkina 1988). Burial monuments of the nomads of northern Bactria One of the most important studies of the nomads of Central Asia was the discovery and research undertaken principally by Mandel’shtam at the necropolis in southern 80 Kazim Abdullaev Tajikistan. His investigation of a great quantity of podboy type burials, mainly in Bishkent valley (Tulkhar, Aruktau and Kokkum (Fig. 1)) led him to the conclusion that the culture represented by these necropolises was that of the people who crushingly defeated the Graeco-Bactrian kingdom and ultimately established the Kushan Empire. The detailed study of the finds of these necropolises is given in the works of Mandel’shtam (1966, 1968 and 1975). Another nomad necropolis at Babashov in Turkmenia, also excavated by Mandel’shtam (1963), closely resembles the culture represented by the Bishkent valley burials. Nomad monuments survive to a lesser extent in Uzbekistan, because the development of land for cultivation included the area of foothills where necropolises were located, so that they have been destroyed. Such destruction has also occurred in more urban areas. Some idea of what has been lost is provided by the site of Airtam, 18 km to the east of modern Termez. Apart from the remains of monumental buildings decorated with stone reliefs, burials have been found here which can be linked to nomad culture. The construction and funerary finds of the Airtam burials are very similar to those of the Tulkhar necropolis (Turgunov 1973, 64–8). Two types of burials are found at the Airtam necropolis: a rectangular grave and a podboy (oriented north-south). The finds included a double-blade dagger (35 cm in length), three iron rings and an arrowhead. The pottery and other finds from the funerary complex show close analogies with the material from Tulkhar, Aruktau and elsewhere. The necropolis of Airtam can be dated to the second to first century BC. As mentioned above, the nomad monuments of Uzbekistan have been completely destroyed. One of them, the Rabat I necropolis in the Baysun district of Surkhan Darya region (Fig. 2), was situated within the site of Payon Kurgan, a fortress of the nomad Figure 2. Necropolis Rabat I. Modern and ancient tombs. NOMAD MIGRATION IN CENTRAL ASIA 81 migration period (Abdullaev 1999). This remarkable monument— part of a huge necropolis on the upper terrace of the Akjar River — extended from north to south for a distance of about 1 km. It was discovered by accident during preliminary digging for modern construction work. It was almost completely destroyed and the site is now covered by the vegetation and buildings of the modern village of Payon (which means ‘lower’ in Persian). Only ten tombs could be traced and in only one a skull with circular deformation was found. The archaeological material from the tombs— different weapons (arrowheads), a mirror and jewellery — is very similar to the finds from Tulkhar, Aruktau and other necropolises of the Yuezhi type. The Rabat II necropolis was discovered much later (Fig. 3). Evidently it was a continuation of Rabat I and was aligned in the same north-south direction. Here also, following the complete destruction of the surface remains, we traced and excavated ten tombs with some funerary material. The pottery collected on the surface (after the destruction of the site) belongs to the Kushan period and is dated first to second century AD (Fig. 4). The presence of piles of large stones suggests that the tombs were placed inside a stone enclosure which had been dismantled during the modern destruction of the site. The Rabat III necropolis is situated about 1 km south-west of the village of Rabat, at the foot of the Baysuntau mountain (Fig. 5). It covers an area of about 0.5–0.6 hectares. On the surface, the rectangular or oval stone enclosures of the tombs are clearly visible. Most of them appear to be oriented north-south, but the site has not yet been studied. The most representative excavated site of nomad type in Afghanistan is Tillya Tepe, where the tombs contained many remarkable objects in gold, silver, bronze, stone, and Figure 3. Necropolis Rabat II (first–second centuries AD). General view after destruction. 82 Kazim Abdullaev Figure 4. a (top): Plate with incised ornament from Rabat II. b (below left): Two handled pot (amphora shape) from Rabat II, first–second centuries AD. c (below right): Goblet covered with red slip from Rabat II, first century BC–first century AD. so on. For this complex, there are good chronological indicators— such as coins — which provide a relatively short time-frame of the first century BC to first century AD. But its ethnic identity is still disputed. The excavator, Sarianidi (Sarianidi 1985 and Sarianidi & Koshelenko 1982) considers it to be Yuezhi. The character of the finds is, however, notably mixed, for they include clear motifs belonging to the artistic culture of Pazyryk burials, as well as certain Chinese artefacts, and Saka elements that indicate a local tradition with strong Hellenistic influence. In any case, we have here an amalgamation of artefacts from Siberia to the Graeco-Roman world. NOMAD MIGRATION IN CENTRAL ASIA 83 Figure 5. Rabat III. General view of the Necropolis. Nomad city-sites An interesting aspect of nomad migration is the transition period from nomadism to a sedentary way of life, that is to say, the development of urban culture, the establishment of cities and the formation of a state, such as the Kushan state, for example. It should be noted that this process has not been recorded in the literary sources and is inadequately defined by the archaeological sites. The temporary nature of nomad dwellings allowed them to move easily over long or short distances in search of good pastures and camps. This usual process of migration, dictated by a way of life, is known to a certain extent from ethnological data. Of what does a nomad city comprise? This question has not yet been addressed by scientific literature or archaeological investigations. Chinese sources make special note of the capital of the Heavenly Empire, calling it jingshi, which means, according to Bichurin ‘mountain army’, that is ‘people’, because until the beginning of the Christian era, there was one military estate in China, and for the residence of the Emperor they chose the high banks of the Yellow River. The name jingshi applied only to the Chinese capital. For the capitals of other states they use the name du, meaning a residence (Bichurin 1950, II, 149, note no. 4; [editors: the etymology of jingshi proposed by Bichurin is more closely associated with later usages of the term, which would be better interpreted from ancient usage as the ‘king’s palace district’, as the character for jing shows a tall building and is thought to represent the king’s palace, rather than a hill; shi means a large group of 84 Kazim Abdullaev people, and hence the army, but in the context of jingshi the people are more likely to represent the people living around the king and hence the district around the king, Karlgren 1957, nos 559 and 755]. All three Chinese sources (Shiji, Han Shu, Hou Han Shu) mention another type of city (or capital), by adding the ending cheng— for example, Jianshi cheng or Lanshi cheng— meaning ‘surrounded by a wall’. The first to pay attention to this detail was Enoki, whose translation was used by Narain (1957, 129–30). Although the interior structure of the city remains uncertain, nevertheless, it is noteworthy that an important sign of a city was the wall surrounding it. This fact is especially appealing in that the cities mentioned in the sources of Han epoch belong to nomad states. As noted above, the term cheng is used for Jianshi and Lanshi. However, the same source, the Han Shu, when listing the capitals of the five principalities (xihou) formed after the settlement of numerous tribes in Central Asia, does not use the term cheng. If the sources record a concept of the city cheng as fortified, that is enclosed by a wall, then that can be interpreted as meaning a traditional, ancient Central Asian city complete with citadel as, for example, Dalverzin Tepe. It does not necessarily mean a type of nomad city. It is logical to expect the existence of a transitional form of city between the two types. The discovery of Kala-i Zakhoki Maron in the neighbourhood of the modern city of Karshi (the capital of the Kashka Darya region) is a clear example of just such an archaeological site (Fig. 6). The initial, most striking impression is the huge size of this site. Its first investigator, Kabanov (1977), described it as ‘one of the most extensive sites of the oasis— the ruins of an ancient fortification and citadel, located at the south-eastern extremity of Nakhsheb’. Kala-i Zakhoki Maron is square in form with sides 400 m in length. The site consists of three concentric terraces, ‘gradually rising to the centre’. The width of the exterior rampart is 30 m and the height 7 m. The exterior sides are steep and the interior ones more sloping. Kabanov (1977, 47) thought there were only two ramparts, but Masson (1973, 20–30) identified a third of greater dimensions, also square, with sides 1.5 km in length. Further archaeological investigations of the site confirmed the presence of the third rampart. The archaeological context suggests that Kala-i Zakhoki Maron was built in the second to first century BC. The remains of houses in the area adjoining the ramparts (wall I and II) are of a later period (Turebekov 1981, 9–10). A remarkable peculiarity of this site is the absence within the fortification — in the extensive area between the citadel and the city walls — of any remains of buildings. This peculiarity is noteworthy particularly if we take account of the huge expanse included by the third city wall. If we admit that it formed a cohesive part of the structure of the site —a proposal which is confirmed by the archaeological evidence— then the vast dimensions of Kala-i Zakhoki Maron (1.5 km2) mark it as the largest site in the region, surpassing even such examples as Afrasiab (Koshelenko 1985, 278). What is more, both the plan and structure of this site are unusual. The closest analogies to the plan of NOMAD MIGRATION IN CENTRAL ASIA 85 Figure 6. Kala-i Zakhoki Maron. Plan of the site, second–first centuries BC. Kala-i Zakhoki Maron are found at the sites of Shakhrivayron, in the Bukhara region, and Janbaskala in Khwarezm (Tolstov 1948, fig. 29a; Turebekov 1981, 9–10). Ancient literary sources, especially the Chinese chronicles, provide some information on the traditional cities of Central Asia, but there is nothing about the existence of the ephemeral settlements of migrating nomads, which must have appeared and disappeared at the flourishing agricultural oases on the endless expanse of the steppes. We can glean some idea of migration settlements from later sources, for example, the fragmentary preserved descriptions of European travellers who undertook long voyages to the courts of nomad kings, for instance, to Orda of Tataro-Mongol. So in the colourful description by Guillaume de Rubrouck (XIX, 213) of the Mongol king Batu, he gives the following information on the organization of the court: When I saw Batu’s court, I was surprised because his own houses appeared to be like a great city stretched out, and surrounded by people on all sides as far as three or four leagues. And, as among the Israelites each knowing where to set his tent around the Tabernacle, so they all knew which side of the court they should set their houses. So the court is called in their language orda, which means ‘middle’ [old Turkic orta ⫽ middle, centre], because it is always set in the middle of the people, but no one is positioned to its South, as it is the side onto which 86 Kazim Abdullaev the doors of the court open. But to right and left the people establish themselves as they wish, according to the lie of the land, so that they are not immediately in front or opposite the court. Ethnographic and archaeological investigations in Central Asia show that ‘a considerable portion of the nomad sites contain no sign of any permanent buildings. It indicates that part of these former cities were occupied by yurts (nomad tents) evidently set out in quarters like later and even modern cities’ (Viktorova 1980, 59). A similar type of city-site has been excavated at New Sarai on the Volga, where the house of a wealthy man with a paved yard has been unearthed (Fyodorov-Davydov et al. 1960, 71). In the area in front of it, yurts were found. It is interesting that even now in urban environments one can still see the combination of permanent dwellings and nomad tents, both in rural areas and in cities, for example, Ulan Bator (Schepetil’nikov 1960 and Maydar 1971). So this ancient tradition has been preserved through the millennia to remain a characteristic sign of the cities and settlements of former nomadic peoples. Even till the present day, the stock-breeding population in the mountain villages of Uzbekistan’s Surkhan Darya and Kashka Darya regions still place their yurts beside their modern houses. All these examples cited above provide substantial evidence for the origins of the nomad city and as a prototype we propose the archaeological site of Kala-i Zakhoki Maron. Very likely, the remains preserve the original plan of a single archaeological site, in spite of the evidence of later dwellings in certain parts of the city. According to the archaeological context, this city was built in the second to first century BC and can be chronologically linked directly to the migration of the nomad tribes mentioned above. So, one can suppose that in this case we have a typical nomad city with a citadel in the centre and divided by streets in earlier times into quarters filled with yurts (tents). The erection of the fortification wall demanded a colossal workforce and the means for building on such an enormous scale. These resources could have been supplied by using the local population as forced labour to erect the city wall. In all probability, the walls of captured cities were not always completely destroyed, but on the contrary, were reused by the nomads for their own defences. The archaeological evidence from the walls of Ai Khanum confirms this supposition (Leriche 1986, pl. 14). Here in the ‘lower city’, in trench no. 1, during excavation of the fortification wall, it was discovered that the wall had been repaired. It was strengthened with soil that had been taken from the moat which surrounded the wall. It is interesting to note that there was no connection between the re-fortified wall and dwelling complex. What is more, within the walls of the former city, burials of nomad type were discovered. Does this testify to the preservation and function of a fortification system after the destruction of the city? In this case, did it protect the nomads and their temporary camp within the destroyed city? I think, in this case, the answer can be yes. NOMAD MIGRATION IN CENTRAL ASIA 87 Images of nomads in figurative art When it is a question of nomad peoples, the task of cultural identification becomes complicated because of the absence of certain important elements that are characteristic, for example, of an urban population. The principal remains of the nomads are funerary edifices that have been given a detailed classification and typology in scientific literature. However, as the extensive research of Central Asian archaeologists has shown, the study of funerary monuments does not give sufficient information for the ethnic attribution of one or other tribe (Litvinsky 1972, 71–2). Many types of burials and funerary constructions co-exist chronologically. The archaeological material presented mainly by funerary utensils also sometimes has a universal appearance (for example, ornaments, utensils, armour, and so on). These funerary objects are termed ‘universal’ because they represent the common types and forms of the objects that constitute a part of the archaeological complexes of various ethnic groups but are found spread over a wide area in certain chronological frames, so that they are often used for the dating of archaeological and cultural complexes. For the clearer identification of nomad remains, it is necessary to use a combination of different sources of data from archaeology and other applied sciences. However, even such methods do not give an exhaustive answer to the various disputed questions of Kushan archaeology. Probably, in this instance, it is necessary to use other sources of information that could perhaps help reveal certain aspects of nomad culture. To such a category belong the objects of figurative art. They are, usually, part of a funerary complex and accompany the deceased as prestige objects. Their relative rarity in burial complexes belies their capacity for supplying useful information. The objects of art representing nomad images can be divided into two categories. The first belongs to funerary monuments; the second forms part of the interior decoration (relief or wall painting) of cult or secular buildings. The compositions mainly have a secular character: battle scenes, hunting or other related subjects provide information on the contacts between the steppe nomads and agricultural zones or oases. It seems that the nomad way of life of does not itself presume to have such a category of material culture as architecture, although certain elements of it, for example, burial constructions, demonstrate a good knowledge of building techniques (as can be seen in the tumuli burials of the Altai). On the other hand, the construction of transportable dwellings also demanded the experience of spatial decisions. In the absence of reliable sources, the interior design of nomad dwellings can only be reconstructed from modern ethnographic data. The investigation of certain burials that probably represent a model of a real nomad dwelling can also provide some information. The objects of art are sufficiently diverse and embrace different aspects of social life. They provide information on armour (and to some extent on military tactics), dress, hairstyle, details of domestic life, religion, mythology, anthropology, and so on. Identifying the find-spots of objects that closely resemble each other helps to trace the 88 Kazim Abdullaev routes of nomad migration and determine more precisely the areas of nomad settlement. So, for example, we can compare the representations on the bone plaques from the Orlat necropolis in the Samarkand region with the fragment of a bone plaque found in Kuyumazar necropolis in the Bukhara region (Figs 7 and 8). The similarities between them are not, evidently, accidental. Figure 7. Bone plate from Kuyumazar Necropolis. Bearded man. Figure 8. Orlat. Bone plate with representation of a battle scene. NOMAD MIGRATION IN CENTRAL ASIA 89 On the Kuyumazar bone plaque we have an abstract depiction of a male figure. His head is in profile; the body in frontal position. He has short hair and a young face without moustache but with a beard. The beard is primitively modelled in straight parallel lines. His dress is interesting, with a V-neck and ornamented with crosshatched lines to resemble the armour of a warrior. The armour on the Orlat plaques is more precisely depicted, but the representations of hairstyle and beard have close analogies with the Kuyumazar example. Certain figures in the Orlat plaque of a battle scene have pointed beards with or without a moustache. In the hunting scene, a figure in the lower section of the composition has a pointed beard, no moustache and short hair. Such a beard is characteristic of a series of art objects that appear to be from Sogdia. The knights in chain-mail armour have analogies in the Khalchayan reliefs depicting a battle of the Yuezhi against a Saka tribe (probably the Sakaraules). Apart from the chain-mail armour worn by the heavy cavalry of the enemies of the Yuezhi, the other characteristic sign of these warriors is long side-whiskers. This detail is typical of the portraits on early Sogdian coins representing a ruler on the obverse and an archer on the reverse. On the right bank of the Oxus, lies the remarkable site of Khalchayan, with its reliefs showing images of these nomads. These reliefs are one of the most famous examples of pre-Kushan art in Central Asia. Pugachenkova (1966; 1975) interpreted the main scene on the reliefs as a triumphal procession of the victorious tribes of the Yuezhi. She also identified the figures (based on a male head with moustache and diadem) with the Heraus clan. However, the numismatic finds, which suggest the succession of the Heliocles I imitations by the Soter Megas issues in Northern Bactria, did not include any coins of this ruler (Rtveladze & Pidaev 1981). Some figures and some fragments of the Khalchayan complex (the horseman shooting an arrow, the figure with captured armour, the horse’s hoof treading on armour etc.) illustrate a battle between the Da Yuezhi and other nomadic tribes, as proposed by Bernard (1987, 758–68) and developed by him with Abdullaev (Bernard & Abdullaev 1997; Abdullaev 1995a). There are some figures in the reliefs which are distinguished by their expressive character. Some of them were identified by Pugachenkova as satyrs and demonic creatures. Indeed these expressive figures with side-whiskers differ greatly from the tranquil and majestic faces and poses of the Yuezhi depictions. The Bactrian artist treated the images of his enemies in a grotesque manner, also characteristic of Chinese art (GrumGrjimaylo 1928). We think it is possible to identify all these grotesque personages with long side-whiskers as enemies of the Yuezhi and relate them to the Sakaraules mentioned above. Iconographically they are very close to the representations on the early coins with the archer on the reverse (Abdullaev 1995b, figs 7–8), which have mainly been found in the regions of Samarkand and Bukhara. We thus have distinguished two different and antagonist ethnic groups in the Khalchayan reliefs. They are represented in Khalchayan in a time of confrontation. Naturally these scenes of the Khalchayan figurative complex reflect a historical event which happened in the second half of the second century BC. In any case it was before the visit of the Chinese diplomat Zhang 90 Kazim Abdullaev Qian in 129–128 BC (this date is corrected from my previous erroneous dating of 123 BC, Abdullaev 1995a, 155). Have we any archaeological layers belonging to period of the Sakas, who occupied the northern lands of Bactria before the arrival of the Da Yuezhi? Clearly any layers of the Saka period must be situated immediately underneath the Yuezhi layers. These have been found at Dalverzin Tepe, one of the main Kushan sites in southern Uzbekistan (Belyaeva 1978). Two remarkable fragments of wall paintings, from a layer which preceded some substantial repair work after the destruction of a building, are particularly interesting (Belyaeva 1978). Indeed, the archaeological context attests here the presence of a population just before the Yuezhi arrival in northern Bactria. The polychrome painting fragment from Dalverzin Tepe shows a helmeted warrior (Abdullaev 1995a, 154–5) who resembles certain figures from Khalchayan. Iconographically, images of an armoured warrior can be linked to representations on the coins of Tanlismaidates, who was a local ruler. The majority of these coins have been found to the west of Balkh. We have thus examined some of the figures on the Khalchayan reliefs. Others— with a demonic expression and a very distinctive hairstyle with long side-whiskers — are paralleled in terracotta plastic art. We have at present several images resembling the figures of Khalchayan. One of them was found in a second to first century BC layer at the site of Old Termez (Fig. 9). It is a moulded yellow clay figure in high relief with a flat back, representing a male head with large ringlets and long whiskers, originally misidentified by the excavator as a female (Pidaev 1987, 89). There are vertical lines clearly identifiable as chain-mail armour on the neck. The second terracotta figurine represents a bearded male figure seated on a low conical chair. It was found at Kampyr Tepe (Surkhan Darya region) in layers of the second to first century BC (Fig. 10). The head was moulded while the hand was modelled in a primitive way. The figurine depicts one of the members of the ruling class of nomads, who in the second half of the second century BC conquered Bactria, shattering the once great Graeco-Bactrian kingdom (Abdullaev 1995b, 176–7). The third example is a hand-modelled figurine found at the site of Erkurgan (a major site near modern Karshi, in the Kashka Darya region to the south-south-west of Figure 9. Old Termez. Terracotta figurine with representation of a warrior, second–first centuries BC. NOMAD MIGRATION IN CENTRAL ASIA Figure 10. Kampyr Tepe. Terracotta figurine with representation of a warrior, second–first centuries BC. 91 Figure 11. Erkurgan. Terracotta figurine with representation of a warrior, second–first centuries BC. Samarkand) (Fig. 11). The representation is treated in a schematic manner, but there are some significant elements in its iconography. The terracotta armoured figurine has a high collar and long side-whiskers just like certain of the Khalchayan figures. It is dated stratigraphically to the second century BC. It is interesting to note that the archer figure on early Sogdian imitation coins has armour with a similar high collar (Mitchiner 1976, 5, 432, no. 652) Thus, both in the archaeological layers and in the figurative complexes, we can distinguish two different images of the northern Bactrian population. The Khalchayan reliefs show us the scene of a battle between the Da Yuezhi and Saka (Sakaraules, Sacaraucae). It is possible to attribute the Khalchayan reliefs to the first century BC archaeologically, but the historical events reflected in them belong to an earlier period, i.e. to around the middle of the second century BC. In general, one can say, that the Khalchayan complex represents a bridge between the Hellenistic art of Bactria and the Hellenized Kushan Art. In this connection, an engraved gem (intaglio) from Samarkand Museum depicting an image closely resembling the above personages is interesting. Its very early Sogdian inscription was first noted by Borisov (1963), who read it as ‘f-r-n’ (farn, parn). 92 Kazim Abdullaev Certain iconographic peculiarities of this image should be noted. The figure has a European appearance with a straight nose, a wedge-shaped beard, no moustache and short hair. The hairstyle resembles that on some figures from ancient Sogdia. A similar beard and no moustache can be seen on the bone plaques from Orlat, dated to the second to first century BC. The figures on these plaques we identify with the tribes surrounding Samarkand, namely the Sakaraules (Abdullaev 1995a and b). Early Sogdian coins with the name Hyrcodes and an archer on the reverse (dated 130 BC to first century BC by Mitchiner 1976, vol. 5, 436, type 669) have a certain similarity in the modelling of the face. The personage on these coins has the same beard and hairstyle, the head turned to the right. This is interesting, especially the coin cited by Zeymal’ (1983, pl. 31 no. 8). Another feature in common is dress, which is depicted with oblique lines to form a V-shaped neck. One can suppose that the figure on the gem represents the portrait of a local ruler, who appears to be ethnically close to the rulers depicted on early Sogdian coins, dated by Zeymal’ (1983, 269) to the second to first century BC. The gem from Samarkand Museum can be dated to the end of the second century BC to first century AD (Fig. 12). Conclusion Literary sources recount how the Yuezhi on their long journey from the valley of Gansu met different peoples. The first were the Wusun. After their defeat by the Wusun in the region of Semirech’e, the Yuezhi migrated further in a westerly direction. Passing through Da Yuan, they reached the region between the two rivers of Amu Darya and Syr Darya. Here, north of the Oxus River, they found another tribe already settled for some time, whom we can hypothetically identify as the Sakaraules. Figure 12. Samarkand Museum. Engraved gem (cornelian), first century BC–first–second centuries AD. Portrait of a governor. Figure 13. Early Sogdian coin. Portrait of a ruler, second–first centuries BC (?). NOMAD MIGRATION IN CENTRAL ASIA 93 This tribe is linked with the culture of the archaeological monuments of lower Syr Darya and is placed in eastern provinces of Khwarezm (Choresmia). Tolstov (1948, 243) and von Gutschmidt (1888, 58 and 70) identified them with the Kangju. In approximately the third to first half of the second century BC, these tribes migrated in a southern direction and settled around the Samarkand and Bukhara oases. In all probability, one can connect the representatives of the culture of the Orlat kurgans (tumuli) with the tribe of the Sakaraules, who dwelt in former times in the region of Kangju. If we follow this scheme further, the territory of the Sakaraules accords with the description in Strabo, who locates them as coming originally from the region beyond the Jaxartes river. Numismatically the Sakaraules can be associated with early Sogdian coins depicting a ruler in profile on the obverse and an archer on the reverse (Fig. 13). The other group that hypothetically can be related to this period or earlier is the early imitation of Antiochus I’s coins with a horse’s head on the reverse. Geographically both types are concentrated around Sogdia, including the Bukhara region. But chronologically, it seems very probable that the tribes around the Bukhara oasis had already gained their independence in the Seleucid period, while Sogdia and Bukhara itself became free from Greek power during the reign of Euthydemus I (Bopearachchi 1990). Numerous imitations of his coins testify to this. We have another imitation of Greek coinage, namely of Eucratides’ coins. Imitations of his obols are clearly local to the region of southern Tajikistan and partially in Uzbekistan. It is confirmed by Strabo’s information on the Parthian annexation of the Bactrian territories (i.e. the satrapies of Aspiona and Figure 14. Khalchayan clay sculpture, late second–first centuries BC. 94 Kazim Abdullaev Turivu). A more complicated imitation is the so-called ‘barbarized’ Heliocles coins. The finds of these imitations are mainly local to the modern Surkhan Darya region, south of Tajikistan and to the south of the Oxus river. The variety of imitation types supposes long continuity and circulation of these imitations in certain areas. Zeymal’ (1983, 110–28) gives a detailed typology of the Heliocles imitations. Amongst the 7 types of imitation proposed by him, the first two (I and II) hypothetically can be associated with Saka coinage. However, it is a proposition that needs to be argued in more detail. The problematic typology, chronology and attribution of the Heliocles imitations is beyond the scope of this paper. Very likely the Khalchayan reliefs demonstrate the historical event when the Yuezhi confronted the Sakas, probably the Sakaraules, and in defeating them, established their own power over the territory to the north of the Oxus (Figs 14–16). Later they subjugated the region to the south of the Oxus— the kingdom of Da Xia of the Chinese chronicles— and still later, they moved further into the Indian subcontinent. Figure 15. Khalchayan clay sculpture, late second–first centuries. Figure 16. Khalchayan clay sculpture, late second– first centuries BC. Representation of a warrior. NOMAD MIGRATION IN CENTRAL ASIA 95 So, according to the archaeological data, one can suppose that the nomad tribes gathered round the Bukhara oasis in its western and north-western parts were, in the majority, local. Similar tribes populated the east and north-eastern part of Margiana. It is very possible that those tribes moved to the Surkhan Darya and Amu Darya plains, and further that they were forced out by the Yuezhi arrival from Siberia and the Altai. There is evident iconographic similarity between certain Khalchayan figures and the portraits on the obverse of the Sogdian coins with the archer reverse. In particular, the long side-whiskers appear on both the coins and on the sculptural representations of Khalchayan (Abdullaev 1995a). References ASGE BSOAS CIAA CRAI IAN Turk SSR IMKU JRAS KSIIMK MIA RA TKAEE VDI ZDMG Arkheologicheskie Sbornik Gosudarstvennogo Ermitazha, Leningrad. Bulletin of the School of Oriental and African Studies, London. Circle of Inner Asian Art, London. Comptes rendus de l’Academie des Inscriptions et Belles-lettres, Paris. Izvestiya Akademy Nauk Turkmenskoy SSR, Ashgabat. Istoriya material’noy kul’tury Uzbekistana, Tashkent. Journal of the Royal Asiatic Society, London. Kratkie soobscheniya Instituta istory material’noy kul’tuey, Leningrad. Materialy issledovany po arkheology, Moscow, Leningrad. Rossiyskaya Arkheologia, Moscow. Trudy Kirgizskoy arkheologo-etnograficheskoy ekspeditsy, Moscow. Vestnik Drevnei Istory, Moscow. Zeitschrift der Deutschen Morgenlandischen Gesellschaft, Leipzig, Wiesbaden. ABDULLAEV, K. 1995a. ‘Nomadism in Central Asia. Archaeological Evidence, second–first century BC’. In A. Invernizzi (ed.), In the Land of the Gryphons. Papers on Central Asian Archaeology in Antiquity, Florence, 151–61. ABDULLAEV, K. 1995b. ‘Armour of Ancient Bactria’. In A. Invernizzi (ed.), In the Land of the Gryphons. Papers on Central Asian Archaeology in Antiquity, Florence, 163–80. ABDULLAEV, K. 1998a. ‘Nomady v iskusstve ellenisticheskoy Baktry’, VDI, 1998.1, 83–92. ABDULLAEV, K. 1998b. Sarmary ili Yuechzi ? (K voprosu o kul’turnoy identifikatsy nomadov i ih dvihzenia), Osh, 24–25. ABDULLAEV, K. 1999. ‘New Archaeological Discoveries in Nothern Bactria’, Circle of Inner Asian Art Newsletter, 10, 7–9. ABDULLAEV, K. 2001. ‘La localisation de la capitale des Yuezhi’. In P. Leriche et al. (eds), La Bactriane au carrefour des routes et des civilisations de l’Asie central (Actes du colloque de Termez 1997), Paris, 197–214. BACHHOFER, L. 1941. ‘On the Greeks and Sakas in India’, Journal of the American Oriental Society 61, 223–50. BAILEY, H.W. 1985. Khotanese texts, VII (Indo-Scythian Studies), Cambridge. BARKOVA, L. L. 1978. ‘Kurgan Shibe i voprosy ego datirovki’, ASGE 19, 37–44. BARKOVA, L. L. 1979. ‘Pogrebenie Koney v kurgane Shibe’, ASGE 20, 55–65. 96 Kazim Abdullaev BARKOVA, L. L. 1980. ‘Kurgan Shibe: Predmety material’noy kul’tury iz pograbal’noy kamery’, ASGE, 21, 23–31. BELYAEVA, T. 1978. ‘Ziloj dom bogatogo gorozanina’. In G. A. Pugachenkova, E. V. Rtveladze et al., Dal’verzintepa— kushansky gorod na juge Uzbekistana, 33–7. BERNARD, P. 1987. ‘Les nomades conquerants de l’Empire greco-bactrien. Reflection sur leur identite et ethnique et culturele’, CRAI, 758–68. BERNARD, P. & ABDULLAEV, K. 1997. ‘Nomady na granitse Bactry. K voprosu ob etnicheskoy identificatsy (Nomads on the border of Bactria —on the question of ethnic identification)’, RA, 1, 68–86. BERNSHTAM, A. N. 1952. ‘Istoriko-arkheologicheskie ocherki Tian Shanya i Pamiro-Alaya’, MIA, 26. BICHURIN, N. J. 1950. Sobranie svedeniy o narodah obitavshih v Sredney Azy v drevnei vremena, 2 volumes, Moscow-Leningrad. BOPEARACHCHI, O. 1990. ‘Podrazhaniya monetam Evtidema I i data nezavisimosti Sogdiany’, Kul’tura drevnego i srednevekovogo Samarkanda i istoricheskie svyazi Sogda, Tashkent, 30–1. BORISOV, A. Y. 1963. ‘Dva iranskih reznyh kamnya Samarkandskogo muzeya’, Epigraphika Vostoka, XV, 51–5. BOROVKOVA, L. A. 1989. ‘Zapad Tsentral’noy Azy vo II v. n.e.–VII v. n.e.’, Istoriko-geografichesky obzor po drevne-kitayskim istochnikam, Moscow. DAFFINA, P. 1967. L’immigrazione dei Saka nella Drangiana, Roma. FYODOROV-DAVYDOV, G. A., VAINER, I. S. & MUHAMADIEV, A. G. 1960. ‘Arkheologicheskie issledovania Tsarevskogo gorodischa (Novyj Saray) v 1959–1960 gg.’, Povolzh’e v srednie veka, Moscow. GRUM-GRJIMAYLO, G. E. 1928. Pochemu kitaytsy izobrajayut demonov ryjevolosymi, St Petersburg. GULYAMOV, Y. G., ISLAMOV, U. & ASKAROV, A. 1966. Pervobytnaya kul’tura i vozniknovenie oroshaemogo zemledeliya v nizov’yah Zarafshana, Tashkent. HALOUN, G. 1937. ‘Zur Ue-tsi-Frage’, ZDMG, 91, 243–318. IVANITSKI, I. D. & INEVATKINA O. N. 1988. ‘Raskopki kurgana Sirlibaytepe’, Istoria material’noy kul’tury Uzbekistana, 22, 44–59. KABANOV, S. K. 1977. Nakhsheb na rubezhe drevnosti i srednevekov’ya (III–VII vv.), Tashkent. KARLGREN, B. 1957. Grammata Serica Recensa, Stockholm. KIBIROV, A. K. 1979. ‘Arkheologicheskie raboty v Tsentral’nom Tyan Shane 1953–1955’, TKAEE 2, 107–23. KOSHELENKO, G. A. (ed.) 1985. Arkheologia SSSR—Drevneyshie gosudarstva Kavkaza i Sredney Azy, Moscow. KOZHOMBERDYEV, I. K. 1975. ‘Saki Kemen’Tyube’, Stranitsy istory i material’noy kul’tury Kirgizstana, Frunze, 174–8. KUSHAEV, G. A. 1963. ‘Kul’tura usuney pravoberezhia reki Ili (III do n.e.–III v. n.e.)’. In K. A. Akishev & G. A. Kushaev, Drevnyaya kul’tura sakov i usuney doliny reki Ili, Alma-Ata. LERICHE, P. 1986. Les remparts d’Ai Khanoum et monuments associés, Paris, LITVINSKY, B. A. 1972. Kurgany i kurumy Zapadnoy Fergany, Moscow. LU ENGUO 2002. ‘The Podboy Burials found in Xinjiang and the Remains of the Yuezhi’, CIAA, 15, 21–2. MALEINA, A. I. (translator) 1993. Plano Karpini i Giyom de Rubruk. Putishestviya v vostochnye strany, Almaty. NOMAD MIGRATION IN CENTRAL ASIA 97 MANDEL’SHTAM, A. M. 1963. ‘Nekotorye novye dannye opamyatnikah kochevogo naselenia yuzhnogo Turkmenistana v antichnuyu epohu’, IAN Turk SSR (Seria obschestvennyh nauk), 2, 126–37. MANDEL’SHTAM, A. M. 1966. Kochevniki na puti v Indiyu, Moscow. MANDEL’SHTAM, A. M. 1968. ‘Pamyatniki epohi bronzy v Yuzhnom Tadjikistane’, MIA, 145. MANDEL’SHTAM, A. M. 1974. ‘Proishozhdenie i rannyaya istoria kushan v svete arheologicheskih dannyh’. In Tsentral’naya Azia v Kushanskuyu epohu, 1, Moscow, 190–7. MANDEL’SHTAM, A. M. 1975. Pamyatniki kochevnikov kushanskogo vremeni v severnoy Baktry, Leningrad. MANDEL’SHTAM, A. M. 1983. ‘Zametki ob arheologicheskih pamyatnikax usuney’, Kul’tura i iskusstvo Kirgizy. Tezisy dokladov vsesoyuznoy nauchnoy konferentsy, Leningrad, 47–8. MASSON, M. E. 1973. Stolichnye goroda v oblasti nozov’ev Kashkadar’i s drevneyshih vremen, Tashkent. MAYDAR, D. 1971. Arkhitectura i gradostroitel’stvo Mongoly, Moscow. MITCHINER, M. 1976. Indo-Greek and Indo-Scythian Coinage, 9 volumes, London. MOSHKOVA, M. G. (ed.) 1992. Stepnaya polosa Aziatskoy chasti SSSR v skifosarmatskoe vremya, Arkheologia SSSR, Moscow. NARAIN, A. K. 1957. The Indo-Greeks, Oxford. OBEL’CHENKO, O. V. 1961. ‘Lyavandaksky mogil’nik’, IMKU, 2, 173–6. OBEL’CHENKO, O. V. 1992. Kul’tura antichnogo Sogda. Po arheologicheskim dannym VII v. do n.e.–VII v. n.e, Moscow. PIDAEV, Sh. 1987. ‘Stratigrafia gorodischa Starogo Termeza v svete novyh raskopok’. In Gorodskaya kul’tura Sogda i Baktry-Toharistana v antichnuyu i rannesrednevekovuyu epohi, Tashkent. PUGACHENKOVA, G. A. 1966. Khalchayan, Tashkent. PUGACHENKOVA G. A. 1975. Skul’ptura Khalchayana, Moscow. PUGACHENKOVA, G. A. 1989. Drevnosti Miankalya, Tashkent. PULLEYBLANK, E. G. 1966. ‘Chinese and Indo-Europeans’, JRAS, 9–39. PULLEYBLANK E. G. 1970. ‘The Wusun and Saka and Yueh-Chih Migration’, BSOAS, 33, 154–60. RTVELADZE E. & PIDAEV Sh. 1981. Katalogue monet iz Yujnogo Uzbekistana, Tashkent. SARIANIDI, V. I. 1985. Bactrian Gold— From the Excavations of the Tillya-tepe Necropolis in Northern Afghanistan, Leningrad. SARIANIDI, V. I. & KOSHELENKO, G. A. 1982. ‘Monety iz raskopok nekropoya, raspolozhennogo na gorodishche Tillya-tepe’. In Drevnyaya Indiya—Istoriko-kul’turnye svjazi, Moscow, 307–18. SCHEPETIL’NIKOV, N. M. 1960. Arkhitektura Mongoly. Moscow. SOROKIN, S. S. 1954. ‘Nekotorye voprosy proishozhdenia keramiki katakombnyh mogil Fergany’, Sovetskaya Arkheologiya, 20, 131–47. SOROKIN, S. S. 1956. ‘O datirovke i tolkovany Kenkol’skogo mogil’nika’, KSIIMK, 64, 3–14. TOLSTOV, S. P. 1948. Drevniy Khorezm, Moscow. TUREBEKOV, M. 1979. ‘Arkheologicheskoe izuchenie oboronitel’nyh sooruzheniy gorodischa Kalai-Zahoki-Maron’, IMKU, 15, 68–75. TUREBEKOV, M. 1981. Oboronitel’nye sooruzhenia drevnih gorodov Sogda (VII–VI vv. do n.e. –VII v. n.e. (unpublished dissertation), Moscow. TURGUNOV, B. A. 1973. ‘K izucheniyu Airtama’. In Iz istory antichnoy kul’tury Uzbekistana, Tashkent, 52–77. VIKTOROVA, L. L. 1980. Mongoly — Proishozhdenie naroda i istoki kul’tury, Moscow. 98 Kazim Abdullaev VOEVODSKY, M. V. & GRYAZNOV M. P. 1938. ‘U-sun’sky mogil’nik na territory Kirgizskoy SSR: K istory u-suney’, VDI, 3, 162–79. VON GUTSCHMIDT, A. 1888. Geschichte Irans, Berlin. ZADNEPROVSKY, Y. A. 1960. ‘Arheologicheskie raboty v yuzhnoy Kirgizy v 1957 g’, KSIIMK, 78. ZADNEPROVSKY, Y. A. 1997a. Review of Borovkova 1989, Drevnei nomady Tsentral’noy Azy, St. Petersburg, 65–73. ZADNEPROVSKY, Y. A. 1997b. ‘Puti migratsy Yuechzhey po novym arheologicheskim dannym’. In Drevnei nomady Tsentral’noy Azy, St. Petersburg, 74–9. ZEYMAL’, E. V. 1983. Drevnei Monety Tadzhikistan, Dushanbe.