365bet娱乐

[Full title: Indian Subcontinent; 1: Lower Indus and Western India; (3): Harbours and Coastal Settlements; (3) Konkan Coast (Maharashtra); (5) Mandagora and other market-towns]

Let us begin with the 'market-town' of Mandagora which in the listing of the Periplus, is situated immediately after Semylla. The first to attempt an identification of Mandagora was Lassen (quoted in Fergusson and Burgess reprint 1969: 205). Lassen locates this settlement at Rajapuri, the modern anchorage at the very mouth of the Janjira creek (Fig. 32). Fergusson and Burgess (reprint 1969: 205) place Mandagora in the vicinity of the Early Historic Buddhist caves at Kuda at the head of the same creek where three coastal village sites 'seemed to be named Mandad or Mandar a name suggestive Manda(na)gora.' A visit by me to the Kuda caves in 1993 confirmed that the village on the shore immediately below the caves is officially still called Mandad (Gupta 1996:52-58; Pl. I.a). Schoff (1912/74:201) locates Mandagora south of the Janjira creek, specifically on the mouth of the Savitri river at Bankot. Huntingford (1980:114) suggests either Bankot or the modern town of Mandangarh a little south of the same river. Casson (1989:297) follows Schoff and Huntingford in identifying Mandagora with Bankot. Recently Patel (1991:102-112) after conducting a survey of the estuary of the Savitri identified Mandagora with Kolmandala village (trans-Savitri) facing Bankot. (Fig. 32 for all positions). After considered review of the above toponyms, this study takes the position that the identification of Mandagora with Kuda-Mandad originally proposed by Fergusson and Burgess is the right one (see also Gupta 1996: 52-58). Lassen's candidate for Mandagora, the anchorage of Rajapuri at the mouth of the Janjira creek, becomes unsafe and unapproachable for ships during the monsoon months. We know that it was with the monsoon winds that maritime traders traversed the Arabian Sea. On a visit to Rajapuri in the non-monsoonal month of May I noticed that the sea here was windy and rough. A natural anchorage for ships in ancient times would have been inside the creek which is sheltered by hills on all sides. Huntingford identification of Mandagora with Mandangarh, taluk headquarter in Ratnagiri District of Maharashtra located 10 km south of the river Savitri and 22 km from the sea, does not also seem tenable. As Patel (1991:107)

198 argues, Mandangarh does not lay claim to any literary or archaeological basis for its existence in Early Historic times. It is neither located on the river Savitri nor the sea for it to be a coastal trading-station. The Now to come to Patel's identification of Mandagora with two villages on the Savitri's mouth facing Bankot (Patel 1991 107) In support of his identification Patel reasons that (a) the two villages Bagmandala and Kolmandala are located right on the seashore and (b) these contain the ruins of a village called Mandan name of the ruined village - Mandan is presented as the toponym for Mandagora However, as Fergusson and Burgess have pointed out, there are three villages by the name of Mandad/Mandar near Kuda which are suggestive of Mandagora Patel also does not specify whether the remains of Mandad on the Savitri are of recent date or go back to Early Historic period. In contrast, the village Mandad at the foot of the Buddhist caves of Kuda has an Early Historic mound in its vicinity (discussion below). The modern anchorage of Kuda-Mandad at the head of the Janjira creek (Pls. I.a.b) offers, historically and geographically, a number of important reasons for it to be the trading-station of Mandagora. We have pointed out above the likelihood of Mandagora being about the Janjira creek, a possibility enhanced by the MandadMandagora correlation. More significant, the Mandad-Mandagora correlation acquires credibility because of (a) the natural and sheltered anchorage offered in the interior of the creek where Mandad is located (b) the existence of the seafacing early Buddhist caves at this site like at at the ports of Kalyana Kalliena and Chemulaka/Semylla and (c) the existence of an Early Historic occupational deposit on the seashore near Mandad village indicating the functioning of the anchorage in ancient times. M.N. Deshpande (personal communication) exploring the beach between the shore and the Buddhist caves found profuse pottery scatter on this tract S. Nagaraju of Andhra University exploring 'near village Mandad in Mangaon taluka (found) an ancient site yielding black and red and red polished wares ascribable to the early centuries of the Christian Era.' (Indian Archaeology - A Review Luders 1978-79:98-99). In company of Dr. V.Shinde (Deccan College, Pune) I revisited Kuda/Mandad in November 1996 In our prospection of the shore immediately below the caves a disturbed Early Historic deposit was located near Kuda village (Pl. I.b) A number of coarse ceramics of the Early Historic period were collected from the ancient occupation The area of deposit. estimated to be 150 x 100 m, was only 200 m from the shore. It seems a larger occupational area existed but has been ploughed by the villagers from Kuda. It is not clear whether Nagaraju reports the same occupational deposit. In any case, the Early

199 Historic site so close to the shore and in immediate vicinity of the Kuda caves probably represents the site of an ancient exchange-centre. This brings us to the evidence of the inscriptions dated to the early centuries A.D. in the Kuda cave group. Many of these epigraphs allude to a family of Mahabhojas (fuedatory ruling class) with the title Mandavas (Luders List nos 1037-1066, Jacobi 1878:253-257). Among others the inscriptions refer to donations made by people employed by the Mahabhoja Mandava Khandapalita (Luders List nos 1037, 1045) and gift of the Kumara Mandava, son? of the chief? of the Mandavas (Luders' List no. 1049). There is also mention of a certain Mandavi, the daughter of some Mahabhoja (Luders' List no. 1052). The Kuda inscriptions also mention donations by srestins or bankers and a sarthavaha or goods caravan leader (Jacobi 1878: 253-257). The presence of these people points to commercial activity at the head of the creek and land routes from the Deccan touching this ancient port. The inscriptions clearly reveal that the family or clan called Mandava politically controlled the Kuda area. It is entirely possible, keeping in mind the contextual data, that the title Mandava of the local ruling family or clan in the early centuries A.D. may have formed the basis of the name of their settlements. The modern village of Mandad (and other villages of the same name come upon by Fergusson and Burgess in its vicinity) below the Kuda caves could very well be recalling the Mandavas. In the same vein Mandagora, which is a Greek corruption of an Indian place-name and which has been associated by us with the toponym Mandad, may also be alluding to the settlement of the Mandavas at the head of the Janjira creek. In this context, the Early Historic occupation at the foot of the Kuda cave complex represents, in all probability, the remains of the settlement of the Mandavas. It is this settlement, together with the cave complex, to which the Hellenes must have attributed the name Mandagora. How can we account for the other coastal settlements lying below Chaul: Hippokoura, Baltipatna Palaepatmae and the Island of Melizigara. We have pointed out that there was little likelihood of any regular anchorage or harbour existing on the stretch of coast between Chaul and Janjira. We have also discussed the possibility of more than one Early Historic anchorage being located about the deep, sheltered creek of Janjira Let us first focus upon the Island of Melizigara, placed by the Periplus and by the Geographia immediately south of Semylla/Chaul. McCrindle (in Shastri 1927:388) identifies Melizigara with Jaigarh at the mouth of the Jaigad river 20 miles south of Dabol. Huntingford (1980:114) and Casson (1989:297) follow

200 McCrindle in suggesting the same identification Schoff also considers Rajapuri at the mouth of the Janjira creek. Furthermore Schoff (1912/74:201, 232) associates the great western Indian port called Sigerus - mentioned by Pliny - with Melizigara A crucial bit of information contained in the Geographia may indicate the location of Island of Melizigara. McCrindle (in Shastri 1927:57) informs that Ptolemy 'makes Miliegyris to be an island about 20 miles south of Semylla Interestingly, this happens to be the length of the stretch of coast between Chaul and Janjira. Going by Ptolemy's information, the Island of Melizigara would be located somewhere about the Janjira creek. The most important and strategically located island in this area is the fortified medieval harbour of Janjira island located at the very mouth of the creek (Fig. 32). Janjia (a Marathi corruption of the Arabic Zejira or island) was heavily fortified by the Sidis, a group of migrant Abyssinian conquerers in the 16 th century and held by them upto 1947 (Chauhan 1994-97-102) The long history of the island's occupation suggests its strategic and commercial importance. Two large fresh-water springs made Janjira island attractive for habitation and contributed to the impregnability of the fortress in medieval times Janjira island, which may have been Melizigara of the Periplus and Geographia, should be prospected for remains of Early Historic period. Let us also examine the Sigerus-Melizigara connection in the context of the situation of Janjira Pliny describes Sigerus as a port in western India to which ships would regularly voyage from Syagrus or the promontory of Ras Fartak (Fig. 23). According to Pliny (in Schoff 1912/74:232-233) the Syagrus-Sigerus journey presented a shorter route to India than the Syagrus-Patala (Ras Fartak-Indus Mouth) voyage. Subsequently, 'a still shorter cut was discovered by a merchant and the thirst for gain brought India even nearer to us' (Pliny in Schoff 1912/74.232) The nearest coast was obviously the Malabar for Pliny, informing about the port of Muziris, calls it 'the nearest mart in India' (Schoff 1912/74:233). It would seem from Pliny's records that the voyage to the Indian coast got progressively shorter as the Hellenes found ways to utilize the monsoon winds to take straighter routes across the Arabian Sea. If Patala/Indus estuary destination was the longest from Ras Fartak and Muziris/Kudangallur the shortest, than the median sailing distance represented by the Syagrus-Sigerus run would situate Sigerus somewhere midway, most likely on the Konkan coast. Schoff (1912/74.201) points out that the anchorage of Rajapuri - which lies right opposite Janjira island on the mainland is the only port on the Konkan where vessels still come to trade -

201 directly from Arabia. These factors do provide the context for the SigerusMelizigara equation but are still far from being conclusive. Beyond Melizygris the Geographia mentions the coastal settlement of Hippokoura. It may be noted that Ptolemy mentions another Hippokoura but places it inland, somewhere in the western-southern Deccan (McCrindle in Sastri 1927:44-45). The various locations suggested for the coastal Hippokoura are Ghodabandar or 'Horse-Port' on the Thane Creek (by B. Indraji following Lassen), Ghoregaon or 'Horse Village' near Chaul creek (by Campbell) and Rajapuri at the mouth of the Janjira creek (by A. Yule, for all ref. see McCrindle in Sastri 1927:44-45) Lassen, Indraji and Campbell are obviously influenced by the allusion of the Greek word for horse in Hippokoura in choosing Ghodabandar and Ghoregaon respectively. We must be aware that horses were regularly imported from Arabia by the Konkani ports right upto colonial times and the settlement of Hippokoura may refer to one such ancient trading-station on the Konkan dealing exclusively in horses. The identification of Hippokoura awaits more evidence. After Hippokoura, Ptolemy names the settlement of Baltipatna which evidently is Palaepatmae of the Periplus (McCrindle 1884:327, Schoff 1912/74:201). Following A. Yule, McCrindle (1884:327), Schoff (1912/74:201), Huntingford (1980:114) and Casson (1989:297) situate Palaepatmae Baltipatna at Dabhol on the river Anjawel or Vasisthi some 100 miles south of Bombay Campbell (in McCrindle-Sastri 1927:45) suggests Pali located upstream on the Nagotna river. Recently Patel (1991: 105, 107,108) has suggested two places; the modern port of. Dabhol and Pal/Palpattan near Mahad as probable locations of Palaepatmae. Evidently the choice of Dabhol by more than one scholar has been influenced by the premise that the coastal settlements mentioned in the Periplus Geographia were necessarily situated one after the other along a north-south axis Dabhol being an important modern port immediately south of river Savitri aids such an identification. However, as we have discussed, the geography of the Konkan coast facilitated emergence of a number of ports about single creeks and rivers. From the point of view of place name similarity the candidates for Palaepatmae are Pali on the Nagotna river and Palpattan on the Savitri river on the outskirts of Mahad. Of the two places, Palpattan on the Savitri is the toponym preferred in this study. Palpattan, located about a mile from modern Mahad, happens to be the site of a large early Buddhist cave-group inhabited in the early centuries of the Christian Era (Fergusson and Burgess reprint 1988:209). The caves at

202 Palpattan/Mahad directly overlook the river Savitri (personal observation) According to M.N. Deshpande (personal communication) Palpattan/Mahad was the site of the ancient emporium of Mahahatti The river Savitri is navigable from the sea upto Mahad (steam ferry indicated in Survey of India Maps). A survey along the Savitri in the vicinity of Mahad undetaken by me in the dry month of May 1993 showed the river full of water. Palaepatmae may have been the Hellenistic equivalent of Palpattan, a riverine trading station directly accessible from the Arabian Sea in all seasons Going by the list offered in the Periplus, the rest of the coastal settlements of Dachinabades are Byzantium, Togarum, Aurannoboas, Island of Sesecriennae, Island of Aegidii, Is. of Canitae, Chersonesus and White Island. Except for Chersonesus, the Greek word for 'peninsula' which can be equated geomorphologically with the Karwar Peninsula (see below), it is difficult to search for modern toponyms of the other ports on account of absence of ancient-modern place name similarity. Besides, the littoral tract between the River Savitri and Karwar has not been well explored for archaeological remains.