3.3. Contacts between the Persian Gulf region and India

Scholars have been uncertain about Egypto-Roman maritime trade contact with the long coastal tracts of Oman north of the Kuria Muria islands and with the thriving market-towns inside the Arabian-Persian Gulf (Casson 1989:281, Potts 1990:313,317, Salles 1993:493-523). Of course, as far as the land routes are concerned there were well established connections between the eastern Mediterranean and the old Arabian-Seleucid\Parthian trading centres about the Gulf such as Gerrha, Apologos, and Charax Spasinou (Raschke 1978:643-644; Sidebotham 1986 a: 171- 172) Prolific inscriptions from Palmyra refer to merchants and caravan leaders from this trading centre active in the Gulf region in Imperial Roman times (Raschke 1978:643). Beyond the Gulf they must have sought contacts with India

Characeneans apologus failaka a persia Parthians a bahrain & Chaldeans ed-dur thaj (gertha?) b (omana bidya SS ohara B mleiha B Parthians arabia Parthians mazirah is. (sarapis) Kushanas Indo-Parthians ss barbaricum Sakas +barygaza b india arabian sea a LEGEND Indian artefacts 6 Mediterranean artefacts Archaeological site Reference in texts NOTES Ed-Dur yeilds RPW, Indian coarse pottery, Indian base metal coin, Roman glassware, ceramics and coin. RPW also at Sohar, Iranian coast. Indian coarse pottery at Bahrain, Thaj?. Fig. 24. Early Historic Ports and Settlements in the Persian Gulf Region 150 50

151 along the old maritime highway whose functioning preceded the use of the EgyptIndia searoute Positive evidence for the Palmyra-India connection in the early centuries A.D. comes from epigraphical sources which record Palmyrenes sailing from the Gulf to the mouth of the Indus in the 2 nd century A.D. (Salles 1993 512,516). Emperor Trajan's wish to consolidate his conquest of Parthia and reach India shows that this area continued to be important for the Romans even after the initiation of the Egypt-India seatrade (Sidebotham 1986 a: 147,155-158). Excavations at the islands of Failaka and Bahrain and the coastal sites of Thaj, Ed-Dur, Mleiha and Sohar have yielded Roman artefacts such as the terra sigillata, pillar moulded glass bowl fragments and Imperial Roman coins (Fig. 3, 24). Was this material reposited as a consequence of overland trade with the Mediterranean or did the goods arrive by sea? The overland Mediterranean-Persian Gulf connection is interlinked with the issue of Roman maritime trade with this region. In a seminal paper on the evidence (or non-evidence) in the Periplus relating to Roman trade with the Oman\ArabianPersian Gulf, Salles (1993:493-523) offers an fresh view on this question. He takes the cue from the Periplus, which apparently ignores the Persian Gulf as an area of Roman seatrade. Salles finds the scant attention given to the Persian Gulf at sharp variance with the detailed description of harbours, sailing distances and maritime conditions provided for adjacent regions (e.g., the Gujarat coast) by the Periplus' author. In contrast to the meagre evidence in the Periplus for Roman trade with the Persian Gulf, Salles (1993:499-503) points to the widespread occurrence of a range of Roman export-commodity remains on the Oman\Persian Gulf coast. He proposes that much of the Mediterranean artefacts were re-exported to the Gulf after they had reached western India along the searoute from Egypt. In the words of Salles (1993:516) :- 'I would suggest that the western goods which were found in the archaeological sites of the Gulf had been carried along the nonRoman segment Apologos-Barygaza of the Indian Ocean trading routes, they had first reached the north India harbours on Roman ships and were then cargoed from Barygaza and Barbaricum to the Gulf by Arabo-Persian merchants and sailors' The probable routes of arrival of Roman goods in the Persian Gulf/Oman thus resolves into three possibilities: (1) Overland from Palmyra\eastern Mediterranean (2) Goods originating from Egypt brought north along the coast of Oman through the Gulf of Aden (3) Re-export from western Indian harbours.

152 ( The second probability, i.e., the circumnavigation of Arabia by Roman ships to reach the Persian Gulf ports is viewed with extreme doubt by Salles 1993:515). But then the Periplus ( section 36 ), does mention the sea run from Kane to Omana, the later a port-site placed by the Periplus inside the Persian Gulf. This suggests that the Oman peninsula was being circumnavigated regularly and the option of this route may have been used by Mediterranean sailors to enter the Persian Gulf (see also Heyerdahl 1982 for emulation of Bronze Age navigation about the Gulf). Furthermore, to study the pronouncements of the Periplus closely, we find that inspite of unclear information on the geomorphic features about the Straits of Hormuz (which Salles emphasises as a point in his argument) the sea guide does provide reliable nautical bearings upto Omana: 'Sailing through the mouth of the Gulf, after a six days course there is another market-town of Persia called Omana* (Periplus Maris Erythraei 36). It is this crucial bit of information on sailing distance which Potts (1990:306-310) uses to demonstrate the identification of the historical site of Ed-Dur located 120 km inside the Straits of Hormuz with Omana. Potts' identification is supported by Salles (1993:511; see also Haerinck et al. 1993:183). It is beyond Omana that the Periplus becomes vague, referring only to the far port of Apologus and expressing ignorance of the busy commerce on the islands of Bahrain and Failaka where Salles and his team have found ample archaeological evidence of Mediterranean contact (Salles 1993:493-523). The point sought to be made here is that there was nothing to stop Omana\Ed-Dur receiving direct Roman maritime commerce from Egypt/Red Sea. The traditional exports of this port and indeed of the Persian Gulf - pearls and dates were commodities lucrative enough to draw merchants by sea from the Mediterranean. Western demand for the commodities of this region is indicated by the copius evidence of brisk overland trade with Palmyra/eastern Mediterranean which continued to flourish even after the initiation of Romes oriental seatrade from the Red Sea. What role did Omana\Ed-Dur play in the context of Roman and Indian sea trade with the Persian Gulf region? The substantial area of the site of Ed-Dur (4 km along the lagoon of Umm al-Qaiwain with 1 km inland spread), together with ceramics and other artefacts of Roman, Indian, Characenian, south Arabian and Parthian origin found at the site indicate that Ed-Dur/Omana was a central emporium of the Persian Gulf (Haerinck et al. 1993:183-193). Now we are sure, drawing upon the Periplus, that the port of Omana had principal commercial connections with western

153 India, being engaged in the import of metals, timber and foodstuff (Periplus Maris Erythraei 36). At EdDur contacts with India are indicated by the discovery of decorated coarse storagejar fragments (Salles 1984: 246-247) Possibly these jars were used to carry Indian foodstuff to the arid Arabian region ( Salles 1993 513, discussion below) Other finds the Red Polished Ware and an Indian base metal coin metal coin (Haerinck et al. 1993.186-187, Salles 1993:501-502). are It is in the ships coming from Barygaza where Salles sees the possibility of Roman goods (presumably the bulk of them) arriving at the Gulf ports Kervran (1996:42) in her delineation of Indian connections with the port of Sohar also suscribes to Salles 're-export' theory, surmising with regard to the find of a Tiberian aurei hoard at Sohar that if no direct contact existed between Rome and the Persian Gulf, it is very possible to imagine this hoard was brought back from India to Sohar...' The hypothesis for indirect Roman seatrade with the Persian Gulf has to be examined within the broader archaeological context of the northern Arabian Sea. Maritime commerce in the Erythraean Sea was a complex affair and especially within the closed and integrated trading arena of the northern Arabian Sea, any number of ports may have dealt in any number of commodities. However, essentially in terms of the available data, doubts can be raised about any significant re-exports of Mediterranean goods to the Persian Gulf from western India. For instance, it has been noticed that the Gulf sites barely yield Mediterranean amphora of the Imperial Roman period. In particular there is an absence of the Dressel 2-4 variety, which as we have noticed, is found in good numbers along the main route of seatrade between Egypt and India. The amphorae largely found on the Gulf sites are a few Rhodian amphorae of the 3 rd-1 st century B.C. (Potts 1990:266,270; Salles 1993:502). The following opinion of the excavators of Bahrain can be applied to the scanty Mediterranean amphorae finds about the Gulf- 'No vessels (amphora) are mentioned in Bahrain or in Eastern Arabia and very few complete amphoras are published on the other sites of the Gulf and even in southern Mesopotamia." (Lombard and Salles 1984:117; bracket mine).' This situation is in direct contrast to India where Mediterranean amphorae finds are widespread (Chapter II ). Roman glassware finds also provide a point of contrast Glassware remains seem to be occurring in greater numbers in the Gulf sites than in western India. In the latter region, confirmed Mediterranean glass findings are relatively meagre, being confined to scattered fragments from Ter,

154 Roman Paithan, Nevasa, Paunar, Adam and Amreli (Chapter II). On the other hand, at the coastal site of Ed-Dur, pieces of broken Mediterranean glass are quite common surface finds' (Haerinck et al 1993:186) Furthermore, Ed-Dur has yielded a rich variety of complete glassware such as the pillar-moulded bowls, pear-shaped flasks, small jugs and beakers (Potts 1990:286, Haerinck et al. 1993:186). glassware has also been excavated at Bahrain (Herling and Salles 1993:175-177), Mleiha (Potts 1990:266), Bidya (Salles 1993:501) and Dhahran (Potts 1990:217) The greater density of Roman glasswares in the lower Gulf sites (especially at Ed-Dur) suggests their transportation overland down the old Tigris-Euphrates route, south to Gulf and then to India along the Apologos-Omana-Barygaza prominently mentioned in the Periplus (35-36). sea route So Roman glassware seems to have been an important commodity on the Silk routes reaching Central Asia from the eastern Mediterranean. Roman glassware is found at Taxila (Marshall 1975: vol. II, 685), Begram (Hackin 1939; Taddei in Cimino 1994:214-217) and further beyond on the Central Asian sites (Warmington 1928/1995:271-72, Raschke 1978:627). In all likelihood, the Roman glassware reaching Central Asia was sourced either from the Persian Gulf (via the Indus port of Barbarikon) or sought overland from the eastern Mediterranean.