6. Summary

The archaeological review of ports/market-towns, maritime trade routes and trans-oceanic contact undertaken above corroborates and amplifies the picture emerging from ancient textual/epigraphical records (notably the Periplus) of seaborne commerce in the Red Sea-northern Indian Ocean during the early centuries of the Christian Era The 'pan-oceanic' view enables us to discern enduring patterns of contact and interaction between the Roman World on the one hand and India and Indian Ocean lands on the other. A salient fact which becomes clear in our study is that most of the harbours and coastal settlements mentioned in the Periplus and Geographia can be traced to archaeological sites. This fact particularly relates to the main ports of trade: Myos Hormos or Luekos Limen (Quseir, Egypt), Berenice (Berenice, Egypt), Adulis (Massawa, Ethiopia), Qana (Bir Ali, Yemen), Moscha (Khor Rori, Oman), Omana (Ed-Dur, U.A.E.), Barbarikon (Banbhore, Pakistan), Barygaza (Bharuch, Western

245 India), Sopara (Sopara, Western India), Kalliena (Elephanta-Kalyana, western India), Camara (Kaveripattinam, eastern India) and Podouke (Arikamedu, eastern India). Even the relatively less important ports and anchorages listed in the GraecoRoman sources now proliferate on the archaeological map of the Indian Ocean Mundus (Heis, Somalia), Opone (Ras Hafun, Somalia), Tabai (Chori Hordoi, Somalia), Dioscorida (Socotra, Yemen), Kammoni (Kamrej, western India), Mandagora (Kuda-Mandad, western India), Kantakosylla (Ghantasala, eastern India) and Sippera (Saupara, eastern India). A number of Early Historic port-sites on the Indian subcontinent which cannot be related to harbours listed in Graeco-Roman sources have, nevertheless, yielded material evidence of contact with the Mediterranean World and Indian Ocean lands. Among these sites, the important ones are Mandvi, Dwarka, Prabhas Patan and Nagara in Gujarat, Mantai and Godavaya in Sri Lanka and Dharanikota on the Andhra coast. Of course, many of these port-sites find mention in early Indian literary sources. Recent attempts to define and grade Early Historic harbours integrate both archaeological and geomorphic-oceanographic features in the descriptive framework. > We therefore find a 'holistic' framework being applied by Stiles (1994:103) in his review of East African harbours, by Flinder (1977:127-139) in his description of Jezirat Fara'un as Ezion-Geber and Somasiri (1991:84-90) in his discussion of ancient Sri Lankan ports. In this study also, a multiplicity of factors have been integrated to explain the nature and layout of harbours, particularly at Kamrej, Elephanta-Kalyana, Chaul, Kuda-Mandad, the Krishna Estuary and the Mouths of the Ganga. Our survey shows that Early Historic Indian ports involved in Indian Ocean trade can be located within three geomorphic environments: estuary, lagoon/backwater and creek. The major estuarine harbour-sites are Bharuch/Barygaza on Narmada, Kamrej/Kammoni on Tapi, Alagankulam on Vaigai, Kaveripattinam on Kaveri, Dharanikota on Krishna, and Tamluk/Tamralipti on Rupnarayan. Typically, all these sites are represented by extensive archaeological deposits and structures (wharfs, fortifications) compared to ports located on lagoons and creeks. Anchorages and ports on lagoons/backwaters are distributed on the Coromandel coast (Arikamedu, Karaikadu) while the most tradeports on creeks are located on the Konkan coast (Chaul/Semylla, KudaMandad/Mandagora, Janjira/Melizigara). Interestingly, there are no major Early Historic ports in India located on bays. However, all important trade ports in the Red Sea Quseir, Berenice, Jezirat Fara'un, Adulis - are situated on the shore of sheltered bays. In the Gulf of Aden, the port-site of Qana is located on a sheltered bay which -

246 Sedov (1996:11-12) calls 'probably the best landing place on the southern coast of Yemen.' The major ports of the Periplus on the Egypt-India route which have been identified with archaeological deposits are Quseir, Berenice, Qana, Khor Rori, Ed-Dur, Bharuch and Arikamedu. It is significant to note that all these interlinked harbours were either founded in the transition years from 1 st century B.C. to 1 st century A.D. or had their floruit during this period. The port of Quseir is dated to the Imperial Roman times (1 st-2 nd century A.D.) It has not revealed any Ptolemaic foundations so far (Sidebotham 1992 12-78; Dr. J-Y. Empereur, C.N.R.S., Alexandria: personal communication). Though Berenice was founded by the Ptolemies, the epigraphical evidence from the temple at the site and inscriptions on the desert route connecting Berenice with Coptos show that the port saw its most flourishing phase in the period of Tiberius Caesar (early 1 st century A.D.) (Meredith 1957:56-70; Sidebotham 1986 a: 52). At Qana, the excavators date the earliest levels of the port to the early 1 st century A.D. According to Sedov (1996: 23) 'It is most likely that the foundation of Qana as a port-city was directly connected with the establishment and expansion of the regular sea-trade between the Red Sea and Indian Subcontinent in early 1 st century A.D.' Besides Qana, the other Hadhramauti port of trade was Moscha, identified now with the coastal site of Khor Rori. Ceramics found at Khor Rori point to the 1 st century B.C. - 1 st century A.D. for the beginning of operations at Khor Rori (Yule and Kervran 1993:93-106). In the Persian Gulf region, the principal site from our point of view is Ed-Dur. Ed-Dur has been identified with the trade-port of Omana mentioned in the Periplus (see above). Though the periphery of Ed-Dur shows traces of Bronze Age occupation, the centre of the site appears to have been occupied principally between the first and the third or fourth centuries A.D.' (Potts 1990:274, also Haerinck et al. 1993:183-193). At the site of Bharuch (identified with Barygaza) limited excavations carried out by the ASI reveals occupation of the city going back to 1 st millennium B.C. We cannot be certain about the settlement size and level of activity at Bharuch on the basis of available evidence. However, as we have discuss in Chapter V, the major Early Historic sites in western India show escalation of industrial and building activity in 'Roman-contact' levels dated between the beginning of the 1 st century A.D. to 2 nd century A.D. Similarly, at the Indo-Roman trading station of Arikamedu, the most flourishing phase has been attributed to 'Roman-contact' levels by Begley (1983:471-472) Wheeler et al. (1946:24) date the beginning of Roman contact indicated by amphorae finds - to the end of the first century B.C. or beginning of the first century A.D., with an inclination towards the latter date." -

247 Sedov firmly The foundation or expansion of major port-sites of the Red Sea-Indian Ocean in the beginning of the Christian Era may not be mere coincidence associates the foundation of Qana with the demands and opportunities generated by the regularisation of the Egypt-India trade in the Early Roman period. It is not inconceivable that the very same opportunities led to similar emergence/expansion of ports and markets in other parts of the Indian Ocean. In this chapter we explored the 'extra-territorial' dimension of Indo-Roman interaction in far-flung areas of the Persian Gulf and Southeast Asia. In the last chapter we observed the sudden proliferation of Mediterranean artefacts to all parts of the northern Indian Ocean in the 1 st century A.D. We also discussed the appearance of 'syncetic' objects like imitation Roman coins and pseudo-terra sigillata pottery. These developments point to the 'overarching' nature of the Indo-Roman interaction. In this context, we repeat here the observation of Glover (1996: 131), 'The great expansion of Southeast Asian exchange which is evident in later prehistory is, I believe, closely connected with this Indo-Roman commerce and can be explained in part, at least, by a rising demand for exotic and prestigious items of consumption and adornment in the sophisticated urban civilizations of the Mediterranean basin, India and, of course, China The projection, though specific to Southeast Asia, can be applied in turn to other trade zones which came to be integrated through the dynamics of Indo-Roman exchange in the early centuries of the Christian Era.