The holy town of Dwarka (Lat.22'15'N Long. 69'E) in Jamnagar district of Gujarat State is believed to be the legendary city of Dwarka built by Lord Sri Krishna between 3500 and 5000 years ago. According to the epic Mahabharata the city in due course was submerged by the sea'. The Harivansh, an appendix to the Mahabharata says that Dwarka was situated on the bank of the river Gomati where it joins the western sea. The submergence of Dwarka and the cause of submergence are of historical and oceanographic interest because of historic Dwarka is likely to throw light on the Dark Age of Indian history.
The seven-storeyed temple of Dwarkadhish or Ranchodji at Dwarka rises to a height of 43 m at the confluence of the river Gomati with the Arabian Sea. The inner temple is assign- able to the 13th century AD and the subsequent enlargement in the form of an assembly hall known, as Ladwa mandap and the awe-inspiring sikhara are datable to the 15th century. There are some early temples on the bank of the Gomati river.
Within the enclosure of the Dwarkadhish temple complex shrines dedicated to Pradyumnaji, Devakiji, Purushottamji, Kuseshvar Mahadev and also the Sharda Peeth established by Adi Shankra Charya in the 9th century can be seen. A hundred years ago large boats could reach the Gomati ghat but the stonewall built by the erstwhile Gyakwad State in 1890 narrowed its mouth. A sand bar has now stopped the discharge of Gomati waters into the sea. The temple of Sea God or Samudra narayana, also known as Varun Devta or Chakra Narayan, which is situated near the ancient harbor at the western extremity of, is an important landmark from the point of view of archeological investigations.
The occurrence of a distinct ceramic ware known as the Red Polished Ware in layer 8 and the Lustrous Red Ware in layers 10 and 10A clearly distinguishes the early historic deposit from the Past and Historic. Layer 9 is yellowish sand and layer 10 is grayish silt, while layer 10A is pinkish sand and layer 11 almost the same without any artifact. The stemmed bowl and a miniature bowl of averted rim found in Dwarka excavation are comparable in form, fabric and treatment to those from Rangpur III. The excavation in 1979-80 pushed back the antiquity of Dwarka from 15th century AD to 15th century 'BC and suggested the destruction of a coastal settlement by sea about 3500 years ago.
In all, eight settlements can be distinguished at Dwarka (IAR 1979-80, 22 ff). The first settlement made in the 15th century BC was submerged or washed away and so also the second one made in the 10th century BC. After a long gap the third settlement was made in the 1st century BC/AD as suggested by the Red Polished Ware and copper coins known as Kansha Varna It is during this phase that Temple I was built. The lime-plastered surface of its stone masonry still retaining a few lines of red ochre painting suggests that the plinth was open to view above the then working level. After the sea destroyed the first temple the second temple came to be built on the ruins of the first. When the second temple was also destroyed, the third temple dedicated to Vishnu or Vasudeva was built in the 9th century. It was perhaps in the 12th century a storm-wave blew away the roof leaving only the walls and plinth.
The fourth temple came into existence soon after. The present temple of Dwarkadhish is the fifth in the series. Temples I to V represent respectively settlements III to VII and the modern town is the eighth settlement at Dwarka.
The Third Expedition exposed structures SE of DWK-SN transit line at four locations 200 to 700 m seaward of Samudra Narayan temple by removing the vegetation and sediment.' It is observed that large well- dressed blocks of limestone (1.5 X 1 X 0.5 and 1 X 0.75 X 0.3m) used in construction have been partly disturbed but the plinths of some of the structures are in tact. For purposes of convenience the structural remains are designated 'A' through 'D'. Structure 'A' near buoys 12 - 13 is the nearest to Samudra Narayan and 'D' the farthest near buoys 26 - 27. Structure A has two arms. It is difficult to say whether these structural remains form part of a single large building. But the crescents plan and the recovery of two moonstones (candrasila) from the neighborhood suggest the existence of a temple, for candrasila is the first step at the entrance to a temple. One of them has a wedge-shaped slit on the margin for joining.
The Fourth Expedition in April 1986 found that the top stones of the masonry of the structures were disturbed by current and swell. In fact they lay under a thick mantle of sediment held by vegetation. The overburden was removed and trenches were dug on either side to expose at least a couple of courses of masonry. The Fourth Expedition, though of very short duration, could discover four 3-holed stone anchors of the type found in Syria and Cyprus and used during the 14th - 12th century BC. A large one is damaged but it was retrieved carefully. As the sea became very rough by the end of December the Expedition had to leave hurriedly. The experience of working in different months from October to May in four expeditions has shown that the proper season for exploration is from mid-November to mid-February even though in November and December heavy snow fall in Kashmir brings severe cold wind to Kachchh- Saurashtra and the sea becomes suddenly choppy. For instance, on 13th December 1986 the sea became so rough that waves 5 to 6 m high were breaking 200 m from shore, In the Third and Fourth Expeditions probing with an iron rod indicated locations of structures in the sea and vegetation was removed. The Fifth Expedition was highly successful in uncovering a massive bastion and fort walls besides three other structures -- all in situ . Hundreds of architectural members are lying loose in the seabed within 800 m of SN temple. Five structures in situ were uncovered after removing vegetation and sand overlying the structures with an airlift or air jet. In several places small trenches had to be dug along the margin of structures with an air jet to trace a few courses of the masonry.
There is always the possibility of dislodging stones from the masonry if the sediments sealing the structures or acting as buttress on sides are removed in excavation. To avoid damage to structures due to strong current and swell, lower courses were not exposed. What is significant in the case of structures exposed beyond 600 m seaward of SN temple is that the building blocks of prismatic shape in situ give at least a partial plan of the structures. For instance, among the structural remains of Period I (Past and Historic) in the lower levels the bastion, fort wall and the corner of a rectangular building could be identified. It was not possible to reach the foundation of these structures, but it appears from the scoured plinth of the bastion, the builders heaped up boulders in seabed to build a wall or pier in coursed rubble masonry. The bastion and fort w | 
