唐代帝陵

                                  
　　After visiting Bamiyan in AD 632, Xuanzang proceeded southwards to Kapisha, the large valley ituated to the north of Kabul. The pilgrim reports that the king was aKshatriya by caste and a Buddhist, and had been able to subjugate the neighbouring countries. The Buddhist samgha was flourishing, since there were 100 convents and some 6000 monks, mainly Mahayanists (Beal 1884: 55; Watters 1904-5: I,123). However, Xuangzang says that there were also some tens of deva temples(that is, Brahrnanical temples), and ' 1000 or so of heretics', among which therewere 'naked ascetics, and others who cover themselves with ashes, and some whomake chaplets of bones, which they wear as crowns on their heads' (Beal 1884:55). Clearly, Shivaite ascetics are alluded to, partly identifiable with the Pashu-patas, partly with the Kapalikas, and partly unidentified,②
　　In the early 1970s a Shivaite temple was discovered at Tapa Skandar, a large mound in the Kapisha plain, which yielded an inscribed marble stela representing Shiva and Parvati (Kuwayama 1976: 381- 83, figs. 14- 15). Ithyfallic Shiva isrepresented with his vehicle, the bull, and his consort Parvati, resting against his left shoulder. This sculpture, which belongs ro the first phase of the temple, is dat-ed, on paleographical grounds, to the eighth century AD (Kuwayama 1999: 72).
　　Tapa Skandar is likely to be one of the deva temples cited by Xuanzang.
　　Farther north, in the plain once known as Bactria, the site of Dilberjin Tepe yieldeda painting with the same subject as that of Tapa Skandar, that is, Shiva seated on bull and Parvati, with a large figure of a devotee depicted on the fore (Kruglikova1974: 45, fig. 30; pl. 5). StratigraphieaUy, the painting can be attributed to the last phase of the temple in the ancient town, originally established in the early cen-turies BC, and can be dated to the seventh- eighth century, despite a much earlierdate suggested by the Soviet archaeologists (ibidem: 46).
　　The penetration of Shivaism into north- western India and Western Central Asiawas a well- established fact at the time of Xuangzang's journey. At Nagarahara(present - day Jalalabad), the Chinese pilgrim observed that, although most peoplewere Buddhists, the stupas were desolate and ruined (Beal 1884: 91). This was due to the fact that Gandhara had been cut off from the main routes connecting Central Asia and India (Kuwayama 1999: 37- 41), a thing that caused the col-lapse of trade, on which the Buddhist community traditionally lived. The increasing pressure of Hinduism, largely based on an agrarian economy, was made easier by these developments, and may have also favoured them.In India, Shivaism had grown as the most fearful movement opposing Buddhism. Itfound royal support and a strong social basis in South India, from where it moved northwards further strengthening the control that the Brahmanical order had al-ready imposed over ,society in many of the central and northern regions of the Sub-continent (Verardi 1996:233 -34). At the time of Xuanzang's journey, the situa-tion was not yet as bad as it was to become some time later. King Harshavardhana,ruling northern India from Kanauj, still acted as a Buddhist cakravartin, although
　　much pressed by the Shivaites. He could secure for some time patronage and free-dom to the Buddhists before his defeat by the new Hindu kingdoms of the Deccan.Only after his death in AD 647 Brahmanical power grew unrestrained, with the ex-ception of the north - eastern regions, where Buddhism was to resist for several centuries to come. We lack detailed information on the actual mechanics through
　　which Buddhism and the social sectors of the population that supported it were de-feated, but violence played a great role (Verardi 1996). Shivaism, which was orga-nized on a military basis, acted as an avant- garde of Brahmanicat reviva[ism, andwas so successful as to replace Buddhism in non- Indian regions also.In present- day Afghanistan Buddhist artistic production had always been either inschist (especially in the early centuries AD), or in clay and plaster of Paris. Ac-cording .to a long established tradition, which went back to the Greeks who had set-tled in Central Asia at the time of Alexander the Great,③ this was by far the com-monest technique used (the same largely employed in Xinjiang). ④ But the Brah-manieal icons were in marble, thus representing a clean break from the tradition: anew order was being imposed which was easily recognizable also from the point of view of the cult images.
　　Buddhism survived and even flourished in some areas after the journey of Xu-angzang. The best- known example is the Buddhist kingdom of Bamiyan.The large statues, recently destroyed, had been built one generation before Xuanzang'svisit (Kuwayama 1987: 725; Klimburg- Salter 1989: 90 - 91). The kingdom was able to control the trade activities from western to central Asia for some time more.Buddhism continued to prevail there throughout the eighth and even the ninth cen-tury after the arrival of the Muslim invaders. A small Buddhist principality flour-ished in the mountains of the Hindukush between Bamiyan and Kapisha: the sanc-tuary of Fondukistan was rebuilt there after AD 689 (G? bl 1967, II: 314). Theiconographic programme included a few images of the Buddha, dewas and the stat-ues of the local princes who were the patrons of the reconstructed sanctuary(Hackin et al. 1959: pls. 189 - 194). An upper coat of reddish clay, subsequentlypainted, was given ro the clay images. Stylistically, the production of Fondukistandepends on Indian models, as is also clear from the well - known painting of a deva or bodhisattva (Hackin 1959: pl. 196).
　　The large Buddhist monastery of Tapa Sardar near Ghazni was also entirely recon-structed in about the same period. The clay images, which show the same final coatof reddish clay on which the paint was given at the end of the image moulding pro-cess, are very close to those from Fondukisran (Taddei and Verardi 1978: 131-33), although the new artistic output may have begun earlier (Taddei 1999: 392-93), Images still depended on Indian stylistic models, although Chinese Buddhistart of the second half of the seventh and of the early eighth century had not passed unnoticed (see, as an example, the Buddha head TS 613 from chapel 17 in Taddeiand Verardi 1978: fig, 181), Actually, stylistic similarities with the early eighth-century Tang production of Tianlongshan have been noted, and the existence of a'common international style' inclusive of Tang China and the regions to the west and south of the Pamir has been suggested (Rhie 1988:29 - 33). Rhie maintainsthat the date of AD 720 'can be suggested as a terminal one for the Fondukistan and Tapa Sardar images' (ibidem).
　　However, a further accentuation of Chinese stylistic traits seems to be recognizable at Tapa Sardar after that date. Both sculptures and paintings appear increasingly dependent on Chinese models, and facial characteristics such as the ,small mouth,the elongated eyes and the arched eyebrows ridge become increasingly common (TS1465, fig. 1; cf also Taddei 1968: figs. 52-53; Taddei and Verardi 1978: figs.240, 252). As to painting, some of the Buddha and Bodhisattva images painted onthe walls of some monastery rooms, as for instance TS 1790 from the still unpub-lished room 52 to the south- west of the main stupa (fig. 2) The adherence to Tang stylistic peculiarities acquires now a previously unknown precision. This fulldependence on Tang art requires an explanation, which is to be found in the politi-cal and religious events that took place in India and China between the end of the seventh and the eighth-ninth century. They were of a different order, but taken together conjured up to create a very particular situation.
　　As is known, in the seventh century Tang China had successfully recovered the eastern Central Asian regions, where the four garrisons were established in 658 and, again, in 692 (Chavannes 1903: 113, note 2). In this period Buddhism reached its greatest success in China, which continued even after the fall of the Zhou dynasty in 705. Many new monasteries were built, and the already existing Buddhist establishments such as Maijishan and Longrnen were further enriched.Yet, the passage from a fundamentally Indian inspiration, which had lasted for at least two centuries, to a consciously unified, 'national' style, is evident. It can be considered the result of the blending of completely assimilated Indian models and arevitalized Chinese tradition. Indian influence can still be detected in the bodies'postures, but facial features and the rendering of clothes acquire a distinct Chinese aspect. Eighth- century Tang images usually show full rounded faces, almond-shaped eyes, arched eyebrows and a small fleshing mouth. This can be observed in the late seventh and eighth - century Buddhist production of Central Asia. As an example, the deities painted on the south wall of Cave 103 ar Dunhuang, dated to705-780 (Dunhuang wenwu yanjiuso 1987: pl. 152), can be recalled. They are characteristic of the Tang style, and show a certain resemblance with the images from Tapa Sardar, especially as regards the hair - do (see TS 1800 from room 52).Proceeding from East to West, the murals at Bezeldik (Turfan), dating to the eighth- ninth century, show a definite Chinese influenee both in style and iconog-raphy. The row of Buddhas seated on lotus flowers from the pedestal of the mainimage in Temple 19 (MIK III 8382; cf. H? rtel et al. 1982: pl. 85) are signifi-cant under this aspect. The clay images and the paintings from Sengim, again in Turfan, dated to the same period, also find reference in Tang art. A mural frag-ment from Temple 10 (MIK III 6761; cf. ibidem: pi.86) represent a deity whose chubby face, almond- shaped eyes and small, fleshing mouth are immediate-ly reminiscent of the contemporary Chinese production, as are the clothes and thehair- do.
　　The examples from Kucha are even more significant. Ernst Waldschmidt's classifi-cation of Central Asian styles cannot be easily accepted nowadays. The idea that Kucha was under 'western' influence in contrast with Turfan, ⑤ should be revised,at least partly, if we consider that Tang influence is well evident in eighth- century Qumtura, and that the Chinese style was probably introduced in the region even earlier, after that Kueha was occupied by the Chinese army in 648 (Chavannes1903: 266 - 67). In a fragment of painting from the Cave at the Bend representing a scene with the Buddha preaching (MIK III 9020; cf. H? riel et al. 1982: pl.58), the Buddha's face, his clothes, and the flaming halo look ' easterly', and the style of the murals in the Cave of the Apsaras (eighth century), in the Cave of the Nirvana and in Temple 12 (eighth - ninth centuries) is definitely Tang. A fewfragments among the paintings from Duldur Aqur show roundish forms and faces from where the Indian- style smile has disappeared. They are of particular impor-tance because they are very close, in style and technique, to the production of theKaiyuan era (713 - 741).⑥
　　When the four garrisons were re- established in 692, Qarashahr was replaced bySuiye (Forte 1994: 46), corresponding to Aq Beshim in present - day Kirghizistan(Clauson 1961 ). The Soviet archaeologists brought to light two Buddhist temples and a monastery there (Staviskij 1998: 111 - 25). The largest temple was excavat-ed in the early 1950s by L. P. Kyzlasov, who dated it to the late seventh century or early eighth (besides Kyzlasov 1959, cf. Hambis 1962 and Higuchi 1997:132 -46), and has been first identified with the Dayunsi cited in Du You's Dong dian,193. 5275, by Clauson (1961: 8). Forte (1994) has shown that this temple,contrary to the previously held opinions, and independently from the archaeological evidence, was not built in 748, but half a century earlier, at the time of EmpressWu between 690 and 705, The temple was probably run by Chinese monks for sev-eral decades (ibid: 54).
　　The majority of the iconographic and decorative materials found in the larger AqBeshim temple keep, in general, a distinct Indian look, as the Buddha images rep-resented on bronze plaques (Kyzlasov 1959: 206, fig. 38). Moulded clay plaquesand other parts of larger compositions,⑦ such as the leaves, share common charac-teristics with the moulded materials from the contemporary Buddhist sites of west-em Central Asia, including Tapa Sardar (Taddei and Verardi 1978: 58- 64).
　　When Chinese stylistic traits are observable, as in a fragmented Buddha head (Ky-zlasov 1959: 1999, fig. 33), they are. not particularly accentuated.
　　A head of Bodhisattva or deua found in the second Aq Beshim temple, excavated by L.p. Zjablin between 1955 and 1958 (Zjablin 1961) is, instead, very heavily de-pendent on Chinese models (cf. Staviskij 1999: 121, fig. 88; Higuchi 1997: 125,fig. 6. 7). The date of the second temple is somewhat uncertain (cf. Staviskij1999: 121 - 22), but such a head cannot be earlier than the first half of the eighthcentury, and in fact Rhie (1988: 39) has compared it to the Bodhisattva faces of Cave XVII at Tianlongshan, associable to the Tianbao era (742 - 55). We know that the Dayunsi was still standing in AD 750 after the attack of Wang Zhengjian,the military commissioner of Beiting, who destroyed the town (cf. Forte 1994),and it is probable that the monastery was still in existence when the Arabs invadedthe region, and that it survived for some time even after the Muslim conquest. This must be true also for the second temple of Ak Beshim. Normally the Muslims did not destroy the Buddhist monasteries, which emptied and were slowly abandoned.There is another Buddhist monument at Aq Beshim, excavated in 1940 by A. N.Bernshtam and interpreted as a monastery, which is dated as late as the ninth -tenth century (Stavisky 1993/94: 124- 25).
　　At Krasnaja Reeha (another site of the Chu Valley) there is an indisputable evi-dence of a direct Chinese presence in the first of the Buddhist temples, which was built at the end of the seventh or in the eighth century (Stavisky 1993/94: 120).A granite votive stela showing a preaching Buddha flanked by two Bodhisattvas and with two dvarapalas on the sides is a Chinese product of the eighth century (ibi-dem: pl. VI). ⑧Similar images of heavenly kings can be found in Longmen (see,
　　for example, the statues of the Wanfo Cave and of the Fengnan Gave, both dated to the late seventh century; Longmen Wenwu Baoguansuo and Beijing Daxue Kaouguxi 1992: figs. 78, 118, 141), and in Dunhuang (Dunhuang Wenwu Yan-jiusuo 1987: figs. 127, 128, 149) as well as in many other Buddhist sites of Chi-na. Only the availability of Chinese models and sculptors, or of priests carrying im-ages from afar can explain its presence. Even ritual objects may show a complete dependence from Chinese models. The pitcher bearing a moulded Bodhisattva head applied to the handle found at Sukuluk (again in the Chu Valley) makes us think to the possibility that Chinese monks were not only running the Dayunsi, but were ai-so in charge of other monasteries of the Valley.
　　Chinese artistic influence is also present in other sites of western Central Asia dated to the seventh- eighth century, among which Kuva in ancient Ferghana (northern Uzbekistan) can be recalled (Bulatova 1972). The colossal Buddha head from stupa1 at Merv (see it in Pugacenkova and Usmanova 1995: fig. 16) in ancient Mar-giana (Turkmenistan) is attributable to the same period.⑨ These sites may also betaken as participating to the ' common international style' influenced by Tang art that we have mentioned above. ⑩ One of the best known site of this epoch is Adzina Tepa in southern Tajikistan (Litvinskij and Zejmal' 1971). We remark the pres-ence there of a clay Buddha head with arched eyebrows, elongated eyes, and thetypical small, fleshing mouth (ibidem.: pls. 55 - 56), and especially of a BOd-hisattva head from Corridor XXVII (ibidem.:pls. 46- 47), a quite exceptional piece, closely dependent on Chinese art. Such a seulpure as this marks a strong change in the 
　　stylistic production of the site, as TS 210 (Taddei 1968: figs: 52-53), TS 1465 (fig. 1 ) and other pieces as well mark a new trend within TapaSardar's late phase. A Bodhisattva head from Cave G at Bamiyan kept in the MuseeGuimet in Paris can also be considered as sharing these characteristics (Hackin1933: fig. 82).
　　What we have seen so far is not sufficient to explain why Tang influence expanded to the south as far as Ghazni on the Iranian cultural borderland, becoming stronger and stronger in the course of time. Ghazni (or, for that matter, Barniyan and Adzi-na Tepa) was not politically subject to Tang China, and Chinese stylistic models could not spread easily or, even leas, be imposed. Clearly, the Buddhist community of Tapa Sardar opted for the Tang stylistic solution intentionally, and continued tofollow it for reasons that are independent from Tang religious policy and cannot be referred to the general acceptance of a new stylistic trend.
　　These reasons must be looked for in India, or, better to say, in the relations be-tween peripheral Buddhist establishments such as Tapa Sardar and Indian Bud-dhism. In the eighth and ninth century, Buddhism was under the final attack fromthe various Brahmanical movements, and was annihilated even in the Northwest,where once it had been so strong. A Vaisnava temple built at the top of a hill domi-nating the once Buddhist settlement of Saidu Sharif and dated to ca. 700 AD has re-cently come to light in Swat in northern Pakistan (CaUieri: forth.; Filigenzi:forth. ). Buddhism, already cornered in peripheral, mountain areas, was to beeradicated when Hindu dynasties were established in Kashmir, Gandhara and Kab-ul.
　　Far- away sites such as Tapa Sardar were safe from the onslaughts of the ortho-dox, but were cut off, at the same time, from the surviving Buddhist monasteries of India, the majority of which were located in Bengal. This region was not only along way from the Iranian borderland, but was difficult to reach because of the hos-tile presence of Hindu kingdoms in the vast territory stretching from Kabul to the middle Ganges valley. A number of marble statues of Hindu gods, indicating astrong Brahmanical presence, characterize eighth- century Kabul and the Gardez region,(11) bordering the territory of Ghazni, i. e., ancient Zabulistan to the east.The only viable option for such cornered Buddhist communities as that of Tapa Sar-dar wanting to keep the faith alive was to look northwards, towards the Buddhistsites of the Northeast, and the only possible models were those propounded by Tang China, which, as we have seen above, had established its artistic hegemony evenbeyond the boundaries of Xinjiang since about 700.
　　If the earlier production of the late sanctuary of Tapa Sardar is roughly contempo-rary with that of Fondukistan, and can thus be attributed in part to the first half of the eighth century, the process of strong sinicization that can be subsequently ob-served had to start sometime before 750, and cannot have continued after 840. This period is the turning point of Brahmanical policy in India, and as to China we knowthat Buddhism was suppressed there between 842 and 845 - a thing that would have prevented, by itself, any further influence of Tang Buddhist art westwards and, what is more, any further possibility for the extant Buddhist communities of western Central Asia, already isolated from Indian Buddhism, to survive. TheMuslim conquest of Central Asia, firmly established by that period, further con-tributed to the end of late western Central Asian Buddhism. The Buddhist laityconverted to Islam, and the monks had to move to the few countries, like Tibet,where the religion of dharma was still alive. Both in western and eastern Central A-sia, and to a large extent also in China, an entirely new story had begun.
　　References
　　Beal, S. (1888) Si - yu- ki. Buddhist Records of the Western Worlds, trans-lated from the Chinese of Hiuen Tsiang (A. D. 629). London.
　　Belenitsky, A. (1968), Central Asia. Geneva.
　　Bernard, P. (1973) Fouilles di Ai Khanoum (Memoires de la Delegation Archeologique Fran? aise en Afghanistan 21), 2 vols. Paris.
　　Bulatova, V. A. (1972) Drevnjaja Kuua (Akademija Nauk Uzbekskoj SSR, In-stitut Arxeologii). Taskent.
　　Bussagli, M. ( 1963 ), La peinture de l'Asia Centrale. Geneve. Callieri, P.(1996) 'Hephtalites in Margiana? New Evidence from the Buddhist Relics in Merv', in LaPersia e l'Asia Centrale da Alessandro al X secolo (Atti del Convegni Lincei 127), pp. 391-400. Roma.
　　Callieri P. (forth.), ' Excavations of the IslAO Italian Archaeological Mission in Pakistan at Bir- kot - ghwandai, Swat: the Sacred Building on the Citadel', in-South Asian Archaeology 2001. Paris.
　　Chavannes, E. (1903), Documents sur les Tou -kie ( Turcs ) Occidentaux. St.Petersburg.
　　Clauson, G. ( 1961 ) 'Ak Beshim - Suyab', in Journal of the Royal Oriental Soci-ety of Great Britain and Ireland, pp. 1 - 13. London.
　　Dunhuang wenwu yanjiusuo ed. (1987), Zhongguo shiku. Dunhuang Mogaoku.Beijing.
　　Filigenzi, A. (forth.), ' Stone and Stucco Sculptures from the Sacred Building of Bir- kot - ghwandai, Swat, Pakistan', in South Asian Archaeology 2001. Paris.Forte, A. (1994) 'An Ancient Chinese Monastery Excavated in Kirgiziya', inCentral Asiatic Journal 38, pp. 41 - 57. Wiesbaden.
　　Gies, J. e Cohen, M. eds. (1995), Serinde , Terre de Bouddha. Dix siecles d'art sur la Route de la Soie. Paris.
　　Gies, J. ed. (1996), The Arts of Central Asia. The Pelliot Collection in them Music Guimet. London.
　　Cobl, R. (1967) Dokumente zur Geschichte der iranischen Hunnen , 4 vols. Wies-baden.
　　Guenee. P., in collaboration with F. Tissot and P. Callieri (1998) Bibliographie analytique des ouvrages parus sur l'art du Gandhara entre 1950 et 1993(Memoires de l'Academie des Inscriptions et Belles- Lettres, n. s. 16). Paris.Hackin, J. in coll. with J. Carl (1933) Nouvelles recherches archeologiquesBamiyan (Memoires de la delegation Archeologique Fran? aisc en Afghanistan 3).Paris
　　Hackin J. (1959) 'Le monastere bouddhique de Fondukistan' , in Hackin J. etal., Diuerses recherches archeologiques en Afghanistan (Memoires de la Delegation Archeologique Fran? aise en Afghanistan 8), pp. 49 - 58. Paris.
　　Hambis L. et al. (1977), L'Asie Centrale. Histoire et civilisation. Paris. Ham-bis, L. (1962) ' Ak- Besim et ses sanctuaires', in Comptes rendus de l'Academie des Inscriptions et Belles - Lettres , pp. 124 - 37. Paris.
　　Hartel H. er al. (1982), Along the Ancient Silk Routes. Central Asian Art from the West Berlin State Museum. New York.
　　Higuchi, T. (1997), Chuo ajia hokubu no bukkyou iseki no kenkyu ( Stucties on Buddhist sites of Northern Central Asia ) (Silk Roadology 4). Kamakura.
　　Kak, R. C. (1923) Handbook of the Archaeological and Numismatic Sections of the Sri Pratap singh Museum, Srinagar. Calcutta.
　　Kak, R. C. (1933) Ancient Monuments of Kashmir. London.
　　Klimburg - Salter D. ( 1989 ) The Kingdom of Barniyan (Istituto Universitario Orientale, Series Major 5). Naples- Rome.
　　Kruglikova, I. T. (1974) Dilberdzin (raskopki 1970 -1972 gg. ), Part I.MoskvaKuwayama, Sh. (1976) 'Turki Sahis and Relevant Brahrnanical Sculptures in Afghanistan', in East and West 26, pp. 375-407. Rome.
　　Kuwayama, Sh. (1987) ' Literary Evidence for Dating the Colossi in Bamiyan', in G. Gnoli and L. Laneiotti eds, Orientalia Josephi Tucci Dicata (Serie OrientaleRoma), 3 vols., pp. 703-27. Rome.
　　Kuwayama, Sh. (1999) 'Historical Notes on Kapisi and Kabul in the Sixth-Eighth Centuries' , in Zinbun. Annals of the Institute for Research in Humani-ties, 34, pp. 25 - 77. Kyoto.
　　Kyzlasov, L. R. (1959) Arxeologiceskie issledovanija na gorodisce Ak -Besim u 1953 -- 1954 gg. (Trudy Kirgizkoj arxeologo- etnograficeskoj ekspedicij Akademija NAUK SSSR 2). Moskva.
　　Le Coq, A von (1933) Die buddhistiche Sp? tantike Mittelasiens. VII. Neue Bildwerke III (ed. E. Waldschmidt). Berlin.
　　Litvinskij B. A. and T. I. Zeimal' (1971) Adzina -Tepa. Arxitektura,Zivopis', Skul'ptura. Moskva. Longmen Wenwu Baoguansuo and Beijing DaxueKaoguxi eds (1992), Zhongguo shiku. Longmen shiku 2. Beijin.
　　Pugacenkova, G. A. and Z. I. Usmanova (1995)'Buddhist Monuments inMerv', in A. Invernizzi ed. , In the Lands of the Gryphons. Papers on Central Asian Archaeology in Antiquity, pp. 51 - 81. Firenze.
　　Rhie, M. M. ('1988) Interelationships between the Buddhist Art of China and the Art of India and Central Asia from 618 -755 A. D., Supplement no. 54 to Annali dell'Istituto Orientale di Napoli 48, 1. Napoli.
　　Staviskij, B. Ja. (1998) Cud'by Buddizma v Srednej Azii po Dannym Arxeologii (Kul'tura Narodov Voskoka, Rossijskaja Akademija Nauk, Otdeleniie Istorii ).
　　Moskua
　　'Stavisky, B. J. (1993/94), ' The Fate of Buddhism in Middle Asia in the Light of Archaeological Data', in Silk Road Art and Archaeology 3, pp. 113 -142. Kamakura.
　　Taddei, M. (1968) 'Tapa Sardar. First Preliminary Report', in East and West 18, pp. 109-24. Rome.                                          
　　Taddei M. (1999) Chronologtcal Problems Connected with Buddhist Unbaked - Clay Sculpture from Afghanistan and Surrounding Areas', in M. Alramand E.
　　Klimburg- Salter eds., Coins Art, Chronology. Essays on the pre- Islamic History of the Indo- Iranian Borderlands ( Osterreichische Akademie der Wissenschaften, Phil-Hist. Kl., Denkschrift, 280. Band), pp. 391-97. Wieh.
　　Taddei M. and G. Verardi (1978) ' Tapa Sardar Second Preliminary Report'in East and West 28, pp. 33 - 135. Rome.
　　Tarzi, Z. (1986) ' La technique du modelage en argile en Asie Centrale et au Nord - Ouest de l'Inde sous les Kouchans : la continuite malgre les ruptures', in Ktema. Civilisations de l'Orient, de la Grece e de Rome antiques 11, pp. 57 -93.
　　Strasbourg .
　　Verardi G. (1996) ' Religions, Rituals, and the Heaviness of Indian History',Annali dell'Istltuto Orientale di Napoli 56, pp. 215 -53. Napoli.
　　Verardi, G. '(1974) ' Osseruazioni sulla coroplastica di epoca Kusana nel Nord - Ovest e in Afghanistan in relazione al materiale di Tapa Sardar', in Annalidell' Istituto Orientale di Napoli 43, 1983, pp. 479-502. Napoli.
　　Watters, T. (1904 - 5) On Yuan Chwang' s Travels in India, 629 - 645 A.D., 2 uols. London.
　　Zjablin L. P. (1961), Vtoroj buddijskij xram Ak-Bescincgo gorodisca. Frunze.
　　———————————————
　　①To G. Verardi are due the pages concerning westem Central Asia and Afghanistan, while those regard-ing Tang China and Xinjiang are due to C. Visconti.
　　②According both to Beal (1884. 55, nora 197) and Watters (1904- 5. I, 123), the naked ascetics are to be identified with the Digambaras, a thing that, however, is very unlikely. The Jains never left India proper.
　　③The earliest examples known are, in fact, those found at Al Khanum on the Ainu Darya (Bernard 1973. 189- 93).
　　④On the clay techniques for the production of images in Afghanistan see Verardi (1974) and Tarzi(1986).
　　⑤ According to Waldschmidt, the western and eastern styles were distinguished also on the basis of the different painting technique and the colours employed (Cf. Le Coq 1933: 24- 31).
　　⑥ The reader is further referred to the fundamental study by Rhie (1988).
　　⑦ Compare the beaded plaques and the foliages from fempIe 1 in Aq Beshim (Kyzlasov 1959: 182, fig.25, 5-6, 9-10; 184: fig. 26, 7-8; 200: fig. 33, 1-3; etc.) with the analogous material at Tapa Sardar(Taddei and Verardi 1978: 58-66, 121-124).
　　⑧A chinese inscnption found on a pedestal at Ap Beshim testifies that Chinese statues were also donated to one of the Aq Beshlm temples (Higuchl 1997:148 - 50).
　　⑨The dating of stupa 1 and of the colossal clay Buddha head in Merv has been subject to a variety of opin-ions and revisions through time (cf. Callieri 1996 and id. in Guenee 1998;163-72), but the stylistic evidencesuggests a later date than those proposed so far.
　　⑩ The late production of Ushkur in Kashmir is also affected by this ' international' style of Tang origin(cf Kak 1923: 12, fig. Bc2; 1933: pl. 58a)
　　(11)Sh. Kuwayama has devoted his attention to this material on several ocasions. Cf. especially Kuwayama(1976, 1999:66 ff.)

唐墓壁画国际学术研讨会论文集/陕西历史博物馆编.--西安：三秦出版社，2006.10