Urban planning and transport in ancient and medieval Sri Lanka
by Kamalika Pieris - June 2007
Urban planning and transport were important subjects in ancient and medieval Sri Lanka. Urban centres, known as nagara, were in existence in the 5th century BC. Early Brahmi inscriptions refer to several of these nagara such Kasaba nagara and Chita nagara. The Dipavamsa says that in the time of Vijaya, Upatissa nagara had fine, well divided streets and apana sala. These cities cannot be located today.
However, the urban centres of the 3rd and 2nd century BC can be located. They include Kelaniya, Magama, Dighavapi, Seru nuvara and Panduwasnuwara. Using aerial surveys it has been possible to find the old city boundaries of Anuradhapura, Sigiriya, Mantota, Kaduruvela and Badalattali. Badalattali (now Paranagampitiya) was the seat of the ruler of Dhakkinadesa from the time of King Kumara Dhatusena (508-516). It was the control point for arrivals along the Deduru Oya. The boundary walls could not be traced in Dambadeniya or Gampola.
The capital city was located at a strategic point. Anuradhapura is on the highest elevation of the plain as are Sigiriya and Yapahuwa. The capital consisted of an inner and outer city. The inner city contained the palace complex and the administrative establishments. The palace complex consisted of the royal palace, the council chamber and a host of other buildings. Plans of the layout of Anuradhapura, Polonnaruwa, Sigiriya and Kotte can be seen in W. I. Siriweera's History of Sri Lanka (p. 109)
The outer city in Anuradhapura occupied an area of about eight square miles. The main monastic complexes were located in the north and south and the main reservoirs were situated in the west and eastern sides. The area beyond the bunds of these tanks had a substantial suburban population. The remains of roads, stone bridges and small monasteries are visible. In Polonnaruwa, in addition to the monastic complexes, a well laid out market complex with a main shopping street can also be seen. The area beyond the outer city had craft production centres and agricultural lands. I think that the inner city and outer cities can be seen best at Polonnaruwa.
The capital city was fortified with embankments, moats and gateways. Anuradhapura had a massive outer wall built of brick. This wall has been identified and traced. It seems to have been originally surrounded by a moat and at least four gateways set in the cardinal directions. The gateways in the north and south are still traceable. The distance from the north to the south gate was four gavuvas. At the eastern entrance, the wall is 10 feet wide and 18 feet deep extending at least four kilometres. It consisted of 23 million ancient bricks. The road near the right doorway when excavated was 57 feet wide. The embankment would have had a total length of about four kilometres. Aerial photos identify the rampart too.
In Sigiriya the citadel had three ramparts and two moats on the west and a single rampart and a moat in the east. The western and eastern precincts were laid out in a precise square module around north-south and east-west axes, which met at the centre of the palace area on the summit of the rock. Beyond the inner city there was an outer wall enclosing the outer city. Suburban settlements were beyond this. The royal monastery was at Pidurangala. There was a reservoir beyond the south eastern area. Sigiriya is considered as one of the best preserved examples of urban planning in a single phase construction in all South Asia.
Polonnaruva had three moats and fortified walls of receding height. The outer city had fourteen gates. The royal enclosure had a twelve feet wide wall with one entrance in the north and guardrooms on either side of the gateway. There was a soldier parade walk eight feet wide. Panduvasnuvara had a moat 50 feet wide and the main entrance at the eastern gate. A set of guardrooms flanked the gateway. The width of the outer wall was eight feet. This was reduced at the upper level to four feet. Two monasteries were located outside the city walls, to the north and south, with two reservoirs in the northeast and west.
Yapahuwa still has an inner and outer moat and wall. They have been traced and partially restored. The outer moat was a hundred feet wide while the inner wall was eighty feet. The base of the outer and inner walls was of stone and the upper section was of bricks. The inner walls had a five feet wide platform, which served as the parade walk for the soldiers. The guardrooms at each gate had flights of steps to reach the parade platform. The main gateway had a three tiered flight of steps. Elephants, horses, chariots and carts came through the side entrances facing east and west.
The western entrance still retains remnants of roadway flanked by shops on either side. The religious centre was placed east of the rock and outside the walls and moats. Reservoirs were situated north of the rock. The high ground to the south of the rock outside the walls was occupied by peasant settlers. Yapahuwa is similar to Sigiriya in layout. The gateways are also similar. At Kurunegala the remains indicate that walls were built only in places where the natural rock formation did not provide defence.
The major cities were also walled complexes. The Thupavamsa says that three moats surrounded Vijitapura. Mantota had a double moat. The boundary walls at Badalatthali are of interest, since they are an example of the oldest type of wall. They were made of well compacted earth and dressed stone. Magama also had walls made of mud and pebbles.
The capital city and major towns had planned layouts. Roland Silva said that the medieval village and towns were well designed and had aesthetic balance. Capital cities such as Mahagama, Dambadeniya, Yapahuwa and Kurunegala, and large market towns such as Mantota, Padaviya and Vahalkada were divided into quarters by streets. In Anuradhapura the inner city (citadel) was divided into four quarters by two main streets that intersected almost at the centre. There were separate quarters for the foreign traders and foreign emissaries. These had multi-storeyed houses. There were shopping areas with food centres. There was an underground system of terracotta and pipes for an efficient water and sewerage disposal system. There were also parks, ponds and baths fed by underground pipes. Mantota, a port city, was paved with wide streets by the side of which were buildings constructed of bricks, granite, and coral stone, with tiled roofs. The literature refers to two and three storied buildings and suggests that the streets were lighted in the night with oil lamps. There were commodity storage centres of mercantile guilds.
The capital cities and major towns had well planned streets with clear street lines. Anuradhapura had Mahaveli vithi and Raja vithi. Mangala vithi started at the southern gate near Thuparama and went eastwards and northwards. Magul maha vithi, the main highway, ran north to south. There was also Kevattta vithi near Thuparama. The streets were wide and level. They were paved with bricks and lined with walls built of bricks. According to Fa Hsien (411-413 AD) the streets and side streets were well kept. Mahagama had a Magul maha vithi, Rajamagul vithi and Tala veli vithi. Dambadeniya had Parivara vithi, Setthi vithi and Agampodi vithi. Smaller places also had named streets. There was Kumbal vidiya at Kumbalgama.
Urban planning had also to take into account the placement of monasteries. Kings did not construct religious buildings within the walled city, except to keep the Tooth and Bowl relics near the palace. The Vinaya rules stated a monastery should ideally be situated about 500 dunus or bow lengths from a village or town and be situated in an arama or park. In Anuradhapura the monasteries of the Mahavihara, Abhayagiri and Jetavana formed a ring of maha viharas, placed between the city walls on the one hand and the agricultural community with its reservoirs and fields on the other. An additional ring of forest monasteries were added in 7th and 8th centuries.
Roland Silva found that in urban areas, the religious edifices were placed in the centre. The plan was a devale and depository at the two ends of the main processional street with an agricultural ring of high and lowland around it. Embekke has a devale facing east with the depository of weapons placed at the eastern end of the processional street. The main street held the dwellings of the office bearers of the devale. Their paddy fields were behind. The threshing compound or kamatha was located at a high point in the fields. A wayside rest hall or ambalam was located on an outcrop of rock in the middle of the paddy field. This functioned as the community centre and sports pavilion as well. Lankatilaka, Gadaladeniya and Mahan Saman devale also followed this plan. At Ukgal aluthnuwara the central complex of a maha vidiya, devale and depository is surrounded by another circular street with houses facing the vidiya on either side. The fields were beyond.
"Drift to the southwest"
By speaking of a "drift to the southwest", the public has been encouraged to believe that the whole population took off with the king, each time he went to live in a new area. There was no such mass migration. For one thing, the places that eventually became capital cities were in use well before that. Polonnaruva was an important military post from early times. Then from the 6th century onwards, it was an additional residence for the king, particularly during the reigns of Silakala (518-31), Aggabodhi III (629-39), Aggabodhi IV (667-83) and Udaya I (797-801). The Kurunegala district, which has a climate and rainfall similar to the North Central Province seems to have been a centre of political power from early times. According to the Tammannava inscription, Vasabha's son was the provincial ruler in the Kurunegala district.
The abandoned capitals continued to flourish long after the ruler had gone to live elsewhere. Anuradhapura continued to be a centre of political activity during the Dambadeniya period. Dambadeniya was an important city in the 15th century when the capital was at Kotte. Buvanekabahu II who ruled from Kurunegala used Yapahuwa as a sub-capital. Places that were not capital cities also had settled populations. Kegalla, Colombo, Kalutara, Galle, Matara and Kandy had a settled population from the earliest times. Brahmi inscriptions dating from pre-Christian times have been found in these areas. Trincomalee was important from about 5th century. The western ports such as Colombo came into operation about 10th century.
There were roads in the island from the third century BC. I think that since there was much commercial, administrative and religious activity in Sri Lanka, good roads would have been essential. I think that there would have been roads from villages to towns, from production centres to markets and ports, from the central administration to provincial, district and village centres, and to places of religious worship.
By the medieval period, Sri Lanka had a well developed system of roads covering three quarters of the island. These roads met at nodal points. Siriweera in his History of Sri Lanka (p. 135) provides a map of the major routes. If we take Colombo as the starting point, one road extended over the full length of the southern coast and up the eastern coast beyond Ampara then turned and went to Mahiyangana. From there roads went to Anuradhapura, Polonnaruva, Trincomalee, Mantota and Kankesanturai. There was also an upward route from Colombo through Kurunegala, and Buddhagama to Anuradhapura.
Of all these roads, the most important was the road that connected the two major administrative divisions of the island, the southern region and the north central plain. Dutugemunu used this road, but except for Buttala, the other places named cannot be identified. In the Anuradhapura period, this road went from Anuradhapura to Polonnaruva and from there via Dastota ford along the right bank of the Mahaweli river to Mahiyangana, Yudaganawa, Buttala and Kataragama ending at Tissamaharama (Mahagama). This road was in use until the modern network of roads was constructed. Major Davy used it to go to Kataragama in 1815. H. Parker found the Buttala-Mahaweli stretch of this route still in use. Roland Silva found that the portion near Ritigala was used by villagers. It was known as Raja Weediya. It crossed a rivulet further up where remains of an ancient stone bridge were visible.
In the Polonnaruva period (12th century) there were two routes that linked the north and the south. One went along the eastern coast from Ambalantota, Tissamaharama, Cagama (west of Tirukkovil) and Mahiyangana to Polonnaruwa. The other went along the southern coast, from Ambalantota, Dondra, Valigama, Bentota, Totagamuwa, Kalutara, Nuvarakele (near Hettipola) Manikdena, and Nikavartiya to Anuradhapura
There were other roads as well. According to the Mahavamsa, there were roads leading out from Devinuwara, Kotte, Kelaniya and Gampola. Roads went from Anuradhapura to Mihintale and to Uruvela, which was near the mouth of the Kala oya. From Mihintale there were roads to Malvatu oya and Kasatota (modern Magantota). Trincomalee was linked to Anuradhapura and Polonnaruwa. One road went along Mihintale, Mahakandarawa, Pankulam and Ratmale.
In Ruhuna, roads went from Mahagama to Serunuwara and Dighavapi. There were routes into the central hills. There were three routes to Sri Pada. They were from Gilimale, in Kuruvita Korale, from Kehelgamuwa, near Ginigathhena (then known as Kadaligama) and from Uva. Parakramabahu I used a route along the Kala Oya to carry out campaigns in the central hills. There were alternative routes as well. Parakramabahu I used the Batalagoda, Laggala route to enter Dhakkinadesa and returned through Opanayake and Badalattali.
Within the Anuradhapura city there were was a route from Mahavihara to a Dhakkina vihara and another to Suddassana Padhanagaraya. Other places would also have had an internal network of roads. These roads met each other. The Visudsdhimaga sanniya refers to a person who did not know which route to take at a crossroads. The Mahavamsa refers to two and four roads meeting. The Saddharmalankaraya also refers to a crossroad. We are told that in Polonnaruva, in the time of Vijayabahu III, three and four roads met at crossroads.
There were three types of roads. There were footpaths and tracks. The Rasavahini refers to a path from the Tisa wewa bund. Even today, bunds of the smaller tanks function as the main footpaths. Secondly, there were "cart roads" or "gal maga", which were designed for carts. Thirdly there were wide, well built roads designed for heavy use. Elephants went on these roads, transporting goods. Items like the Galpotha at Polonnaruwa, which is 24 feet long and 4 feet wide would have come by road. The bunds of the larger tanks, and the major canals would also have functioned as roads. The Kalawewa bund is used as a road today.
The roads had "gavu kanu", which gave distances. Nissankamalla's gavu kanu are the best known. These were pillars inscribed with his name and carrying verses of royal advice. Nissankamalla set them at intervals on the Mahagama-Polonnaruwa-Anuradhapura road. They can be seen today at Buttala, Weligaththa, Vanduruppa, and Katugaha galge. One pillar is in Ratkinda temple in the Bintenne area. The inscriptions on some of these pillars have been published.
Parakramabahu I had a programme for planting a Bo tree and setting up a shrine at every gavuva (2 1/2miles). Roland Silva stated in 1984 that he had carried out a study of the roads at Kegalle district, where the old roads were traced entirely on the evidence of the ambalama and the Bo tree shrines within "meal distance" and "sight distance". Bo trees within "sight distance" can be seen easily on the ancient roadway from Bogoda near the Hali ela crossing point.
The major roads were built by the kings and their subordinates. Ilanaga constructed a road to Mahathupa. Mara vidiya at Dimbulagala was by Sundara devi, the queen of Vickramabahu I. Vijayabahu I improved the roads to Sri Pada and provided resting places for monks and pilgrims. Roads were enlarged to accommodate ox wagons and carriages drawn by horses. Vijayababu IV widened and levelled the road between Polonnaruva and Dambadeniya. Devapatiraja, minister to Parakramabahu II improved the road from Gampola to Botale.
Roland Silva says that these roads were mainly earth tracks cleared of vegetation, raised at the edges with possibly drains on either side. They were paved with hard granite to a reasonable distance from every bridge and crossing point. Records indicate that the road to Mahathupa, constructed by Ilanaga, was strengthened where it went near the tanks. Mara vidiya at Dimbulagala starting from Manampitiya had flattened slabs of stone. The Annekutti inscription at Mihintale suggests that there was a separate department for the construction and maintenance of the main roads. Sirimal Ranwella says the Kinihirikanda inscription shows that there were elephants that were used exclusively for roads.
There were three kinds of bridges according to the Visuddhi magga sannaya. There was the edanda, made of a single log of wood for small streams. Saddharmalankaraya mentions an edanda made of an erabadu log. There were foot bridges, made of planks over which four or five persons could go together abreast and large bridges to take carts across. The Pujavali refers to four bridges constructed on the road to Sri Pada, which supported elephants, horses and cattle. The Pujavali also mentions bridges that were 500, 200, 150 and 86 cubits long.
The army commander of Parakramabahu I built a long solid bridge, 20 or 30 cubits broad, across Kala Oya, passable by elephants, horses and carriages drawn by horses. Iron bands, iron nails and timber were used. Devapatiraja built a bridge of thirty five cubits in length at Botale and another of thirty cubits at Kanamadiri oya at Ulapane. He also built a bridge of 34 cubits in length at Ambagamuwa.
Most bridges were of wood. They were roofed so that the timber was protected. But important ones were of stone or iron. There are about 30 ancient stone bridges in existence. The best example is that at Mahakandarawa on the Anuradhapura-Trincomalee route. It was 80 feet long and 8 feet wide. It consisted of 14 sets of stone each 10 feet 16 inches wide and 12 inches high. There were 42 stone supports set on the Kana oya bed. There are seven bridges over Malvatu oya. The one in Anuradhapura consisted of stone columns standing in holes driven into the Malvatu oya bed. Not a single nail was used.
The ancient texts refer to bullock carts, horse carriages, and even elephant-drawn open carts. The king used elephants, horses, and palanquins for transport. The palanquin was known in Sri Lanka around the 4th century BC. The horses were imported from Sind. High officers too were carried in palanquins. The wealthy used chariots as well as palanquins. On festive occasions, these chariots were beautifully decorated. The Galpota inscription of the Polonnaruva period refers to "commander of vehicles," "chief of the department of vehicles" and "chief of elephants" but there is no further information on these designations.
The ordinary man walked or rode on a cart. Most of them walked, even when travelling long distances, carrying their own food. There were wayside shelters (ambalam) for them with special supplies of water at hand. The smallest ambalama consisted of a foundation of four beams with four posts at the corners and a thatched road. Some had compartments for those who had to spend the night there. Vijayabahu IV had built wayside shelters every half-yojana (approx 6 miles).
Merchandise was taken in carts drawn by bulls. The Medirigiriya and Mahakalattawa inscriptions show that these carts were on occasion driven by buffaloes. Oxen were also used as pack animals, grouped together in a tavalam. The Badulla inscriptions (10th century) refers to the use of pack animals for transporting commodities to towns. Each cart was drawn by four oxen, two on extra harness. Siriweera suggests that small quick sebu oxen would have been used. The Saddharmalankaraya speaks of a caravan of 60 carts drawn by 240 oxen and loaded with pillars of stone, crossing the ford at Kasatota (modern Magantota) on the Mahaveli. There were wayside stops "all over the country" for groups of tavalam. The Galapata vihara inscription of Parakramabahu II speaks of a caravan leader named Sattan and a stopover known as Tavalama close to Bentota.
Rivers were used for transport, but Siriweera says navigation was not possible in most parts of the Mahaveli or any other river. The Culavamsa refers to the anchoring of small ships in the Mahaveli river, which means that up to a point, it was navigable. The Visuddimagga sannaya talks of crossing the rivers in crafts. Boats, barges, rafts, and double canoes are mentioned in the medieval literature. Rafts and barges were made of bamboo or wood and boats were usually carved out of wooden logs. Timber was sent down rivers in the medieval period.
The writings of M.B. Ariyapala, Senake Bandaranayake, Nimal de Silva, P. A. T. Gunasinghe, I. Munasinghe, C. W. Nicholas, Roland Silva, W. I. Siriweera and V. Vitharana were used for this essay.