Version classiqueVersion mobile

Water management in rural South India and Sri Lanka

Patrice Cohen
S. Janakarajan

Part I. History and traditional irrigation

1. Socio-Economic and Technological History of Irrigation in Tamil Nadu

Palaniappan Komathinayagam

Texte intégral


1A. Wittfogel Karl (1957) has been quoted in most of the literature on irrigation. His idea of "Hydraulic Civilizations" emphasized the existence of a strong central authority to manage large – a scale irrigation works. Evidence of democratically functioning traditional irrigation institutions in Tamil Nadu suggests that this was not always the case. Irrigation in Tamil Nadu has a long history. As the state of Tamil Nadu lies on the rain shadow area of the Western Ghats, the torrential rain of the south-west monsoon has little effect over Tamil Nadu. On the other hand, the north-east monsoon that is less dependable but is often accompanied by cyclones, pours heavily in short spells over Tamil Nadu. In there geographical conditions, irrigation played a key role in the development of agriculture in the state. There is evidence of well-developed irrigation systems in literature and in epigraphs and as well as in remnants of structures. Behind these existing indigenous systems of irrigation, there are thousands of years of tradition. Water management techniques thus had an opportunity to be perfected over time.

Figure 1.1. Kumili– Plug and socket sluice

Source: personal

Sources of History

2One of the most difficult problems faced by scientific investigation is that of elucidating the origin of things and the manner in which they have developed to the present stage. Written records throw light only on minute periods of history, while information regarding the preceding periods is indirect. Here, even myths, legends and folklore can help, because many are founded on fact. It is known from literature that the Tamils had established the first ‘Sangam’ (Society of Literates) in south Madurai around 10 000 BC. The city was submerged under the sea. The Tamils then moved north around 5 300 BC and established the second Sangam at Kapadapuram. This was also submerged around 2000 BC. Again they moved north-wards and established the third Sangam at the present Madurai in 1650 BC. It lasted till AD 200. Studies on oceanography point out the fact that due to rise in global temperature, polar snow started melting from 10 000 BC to 2 000 BC resulting in the rise in sea level and submersion of large land masses (Joseph Weinberg and Howard Parish). These scientific findings confirm the deluge that submerged ‘Lemuria’, a lost continent in the Indian Ocean. The Soviet scientists who studied under-ocean archaology confirmed the same. The most surprising part of this, is that a study of the world’s earliest civilizations reveals a whole series of riddles that can be solved only by using the hypotheses of Lemuria, a large land mass in the Indian Ocean that was in habited not just by Lemurs, and not even by Pithanthropos, but by human beings who had reached a high level of civilization (Alexander Kondratov 1974). Similarly the decline of Sangam is attributed to severe drought and famine. It is inferred by climatologists that the rainfall dwindled from AD 50 to AD 200.

3The literature belonging to the third Sangam (1650 BC to AD 200) represents the earliest records of the history of Tamil Nadu. The socio-economic conditions of Tamil Nadu in the first millennium BC are portrayed well in Sangam literature, compiled in the beginning of the Christian era. Another source is ‘Periplus of the Erythrean Sea’ (AD 81-96), which provides information about the trade between the Roman Empire and Tamil kingdoms. The most important development of this period is the spread of irrigated rice cultivation. The evidence appears epigraphically from AD 500 Some of the irrigation structures described in Sangam literature are still functioning and they are the standing monuments of indigenous irrigation technology.

Sangam Age

4During Stone Age agriculture, irrigation was possible only on deltaic terrains with loose soils that were easy to handle with stone implements. The world’s earliest civilizations developed in the valleys of the Nile, Euphrates, Tigris and Indus. Agriculture was possible in the uplands only after the discovery of iron. The Indus people did not use iron. But the Iron Age coincides with the Sangam age. An archaological discovery in the sepulchral urns on the banks of the Thamiravaruni River contained iron tools. Archeologists have estimated by carbon dating that they belong to 1 000 BC. The Sangam poems inform us as to the type of agriculture and irrigation technology of the period. Thiruvalluvar in his Tirukkural, written in the first century BC, emphasized the importance of rainwater. He says that the duty of a government is to know how to acquire resources and to collect, conserve and distribute them worthily. This applies to water resources also. There are a number of poems describing tank irrigation. One poet, Kudapulavianar, advises the king to build tanks wherever the topography permits in order to store rainwater. Those kings who build tanks will be remembered by the world. Another poem, written by Kariasan, lists the components of tank irrigation. He declares that only those who do the following would go to heaven:

  • Build a tank
  • Provide a surplus escape
  • Provide clear waterways
  • Provide tank feeder channels, field supply channels and drainage channels
  • Develope the command areas
  • Dig a common well.

5There are poems describing water lifting devices, sluices, gates and channels.

Development of Tank Irrigation

6Most of Tamil Nadu lies on the leeward side of the Western Ghats and therefore misses the intensive rainfall of the dependable south-west monsoon. But the north-east monsoon, which is less dependable, brings more rain over Tamil Nadu. However, the north-east monsoon is often accompanied by cyclones and pours heavily in short spells. This hydrological feature necessitates the creation of facilities to hold the rainwater of the north-east monsoon to be utilized at a later date. The layout, structure and construction of these small tanks reveal the ingenuity of our forefathers, who adopted a style quite fitting to the gradual fall of the contours and easily adaptable to the system of decentralized village administration they had. The north-east monsoon is more active in the coastal districts of Tamil Nadu, which also have the most tanks. Although there is mention of tanks in Sangam literature, the "Pallava" kings have the credit of creating more tanks between AD 500 and AD 900.

7During this period, there was a rapid development of tank irrigation in the northern parts of Tamil Nadu. Similarly, the Pandyan kings contributed to tank irrigation in the southern parts of Tamil Nadu. Epigraphs provide the names of a number of tanks which were in existence in the Pallava period, and many of these continue to function even today.


8The tanks had their off-take sluices in the deepest portion. Even now one can observe two stone pillars protruding from the water. These are the "plug and socket" type sluices called Kumili in Tamil. They are referred to as karkum/Lf in inscriptions. The protruding stone pillars are the identifying marks of the location of a sluice when the tank is full. A person has to swim to the pillers and go underwater to open or close the sluice. There will be two inlets for water. The one on top is the socket in which plugs will be used to open or close the sluice. The Top shaped plug provides an annular ring opening for partially lifting of the plug. This arrangement allows of a varying discharge with accordance to the water level in the tank. If the water level goes down to the socket level, the second opening, usually closed with clay or a wooden piece, will be opened. The arrangement is shown in Figure 1.1. In this type of arrangement, there will be minimum or no dead storage. In the present day, the tank sluices are formed in such a way that the barrel is limited to the size of the bund. This invariably leaves a dead storage. Other types of sluices, either Madai (with paddle shutters) or Madagu (with screw gearing shutters), were also used in place of Kumilis The arrangements are shown in Figure 1.2. The Pullikkan Madagu (tiger’s eye sluice) was common in the southern districts. This sluice has three circular holes on the vertical wall of the barrel. There is one hole at the top and two at the bottom. The arrangement of holes in this way resembles the eye of a tiger and hence this name. The discharge through an orifice (hole) will be equal to Cd 2gh (L x h).

Surplus Escape

9The Kalingu or surplus weir is the most important component of a tank. It was constructed with care and its protection and maintenance received the special attention of the villagers. One Kural says that one who lives an extravagant lifestyle will lose all his wealth, like a tank without surplus escape that loses all its water. The surplus escape is the safety valve of the tank. Usually, the escape was located on the extreme end of a tank. If a tank was formed across a stream, the weir will be higher than if located on the stream and will also be costlier. Providing dam stones for the surplus weir was a common practice followed for centuries until recently in Tamil Nadu. These stones were generally 2ft. x 2ft square and 2ft to 3ft long. They are built into the crest of the Kalingu. The top of the dam stones would be at the maximum water level (MWL) of the tank. The dam stones would be fixed at 1 to 2 ft intervals. The spaces between the dam stones were blocked up with clay and turf, with the result that water could be held up to MWL. When the water level rises over the top of the turf bund, it would be washed away rapidly and the surplus weir releases the floodwater quickly (Fig. 1.3).

Figure 1.2. Madai (Paddle) / Madagu (Screw gearing shutters) sluices

Source: personal

Chain of Tanks and Water Distribution

10These chains of tanks were planned in such a way that the tanks are smaller in size at head reaches and largest at the lower ends. Invariably the tanks in the region where the north-east monsoon is more acute are larger in size than the tanks in other regions. Modern irrigation management follows the rule that until the last tank in the chain is filled, the upper tanks should not be opened for irrigation. In the traditional tank system, this was ensured by keeping the last tank for either the king or the temple and, until it was full, no one started their activities in the upper tanks. To ensure that the last tank filled quickly, the people maintained the feeding channels in good condition.

The Grand Anicut

11The Grand Anicut built during the second century A. D., is an amazing engineering feat by the Tamils. This is a marvelous piece of hydraulic structure built across a mighty river in its sandy bed and serving to this date exceedently well. The science of building structures safely on a permeable foundation only developed in the last 200 years. Judging from the recorded data after the construction of the Krishnaraja Sagar and Mettur Dams across the Kaveri, heavy floods in the order of 1 86,000 cis. were discharged through this anicut with minimum or no damage. It is possible that much higher floods could have flowed over in the past when there were no other structures in the river. This old stone anicut fascinated Arthur Cotton, who called it the Grand Anicut and he had an innate desire to probe the foundations and see how this engineering feat had been achieved. He designed and constructed five scouring sluices on the eastern end of the anicut.

12While excavating for the foundation, Cotton was surprised to see that the Grand Anicut had stood all along on foundations laid with large boulders embedded in a layer of stiff clay underlying the sandy bed. Arthur Cotton wrote, in 1874, "It was from them, we learnt how to secure a foundation in loose sand of unmeasured depth. With this lesson about foundations we built bridges, weirs, aqueducts and every kind of hydraulic work. We are thus deeply indebted to the native engineers."

13Even before the construction of the Grand Anicut, rice cultivation had been practised in the Kaveri Delta. Sangam literature glorifies the Aaricultural productivity of the Kaveri Delta. The sheer necessity to save the crops in the delta and provide water for irrigation against all odds should have driven the ruler to embark on such a stupendous task with confidence. The technology with which it was built is so appropriate that this is the only earliest big structure of this kind, which continues to serve the purpose even today. Sir Arthur Cotton constructed the Dowleeswaram anicut across the Godhavari with the Grand Anicut as a model.

Figure 1.3. Tank Weir – Dam stones

Source: personal

Taming the Rivers

14The earliest form of river diversion works were known as Korambus. Casuarina or bamboo poles were driven into the bed of the river along a line between the banks. Cross poles were tied to these vertical poles. Nana! or elephant grass would be tied to the poles such that the arrangement acted like an obstruction to river flow. This raised the water level in the river and water started flow ing into a channel or inundated adjacent fields.

15Construction of anicuts or weirs on the bed of a river/stream raises the water level sufficiently and ensures gravity flow in the channels to irrigate lands which are at a higher elevation, and feeds tanks. The anicuts were built using boulders first, then undressed irregular cut stones and finally beautifully cut stones of different sizes and shapes. The boulders were held together by an ingenious method, which can be observed in the Kalingarayan anicut and at Jangamakatte in Mysore across the Kaveri. A hole was drilled in the boulder. Then an iron rod was inserted and molten lead was poured into the hole. This fixed the rods. Another boulder in which a hole had been drilled was placed in such a way that the rod fit into the hole. Thus two boulders were fastened together. This arrangement held numerous boulders across the river and formed the anicut. When cut stones were used, dovetail joints held the stones together. This prevented the stones from being wahed away by the water flow (Fig. 1.4).

16The anicut structures constructed during the early Pandya kingdoms, still serve the purpose. The anicuts across the Thamiravaruni River are classic examples of technical ingenuity. They are built on the riverbed at an inclined angle. They are either form an elongated "L" or are Horses-Shoe shaped. This shape was interpreted to mean that they were aligned to suit the rock outcrop. Whenever there was one off-take channel "L" shaped anicuts were built; when there were two channels off-take channels, the structure is in the form of horse shoe. There are two reasons for this. The first one is that this alignment will ensure that water flows into the channels even during low flows.

17The second important reason is that by increasng the length of the anicut, the affluxes during floods are contained within the river. In the Thamiravaruni system, the command area of each channel (contour canals) lies between the channel and river and the excess water drains back to the river (a method of reuse). Therefore, any increase in the water level of the river will submerge the cropped land adjoining (Fig. 1.5). Although at these anicut sites rock our crops are everywhere, they have followed this shape. This way of building anicuts diagonally on the riverbed can be observed in Amaravathi and Bhavani, which were built between the eighteen and thirteen centuries. When the British came, they built the first straight anicut across theThamiravaruni at Srivaikundam in 1858.

Figure 1.4. Joints in an anicut

Source: personal

Figure 1.5. Hydraulics of Duckbill Weir

Source: personal

18The Marudur anicut above this anicut was built by Pandya kings and has a length, which is more than three times the straight width of the river (Fig. 1.6). The historical flood records show that the height of water over the Srivaikundam anicut was enormous.

19Head of Discharge over Crest Year Marudur Srivaikundam 1877 1.02 m. 2.96 m. 1895 1.85 m. 3.66 m.1923 0.48 m. 1.93 m. This forced the British to build flood banks on both sides upstream of the Srivaikundam anicut for 3 km. This not only increased the cost but also created water logging and drainage problems in the ayacut behind the tanks. The other anicuts, however do not need any flood banks. The canals taking water from the anicuts were excavated as contour canals. The command areas were located between the canal and river or between the upper and lower canals. The drainage water from the command areas again enters the river as return-flows. Thus the water is used again many times. The Uyyakondan canal in the Kaveri near Tiruchirappalli excavated during the reign of Raja Raja Chola (A. D.985) and the Kannadian canal in Thamiravaruni was excavated during twelfth century; these are long contour canals running for more than 30 kms. The technology of anicut construction had been perfected and almost all rivers in the state have such anicuts, canals and tanks system.

Socio-economic History

20A notable feature in the early Tamil kingdoms was that there was little or probably no private property in land. All land was held nominally under the authority of the king with the "owners" holding merely the right of possession at the king’s pleasure. The produce of all arable land was divided into two main part: the Melvaram (upper share) belonged to the state and the Kilvaram (lower share) belonged to the local villagers. The majority of villages under the system called Vellanvagai (peasants’ custom) were common, containing a kinship community of peasants together with specialized and kinship groups of village artisans and village servants such as common irrigators.

21The village was the smallest administrative unit under the Pallavas and had a local self-governing institution called "village sabha", which consisted of all the male inhabitants of the area. They had complete control over the rural administration and acted on their own initiative independently of directives from kings. For the purpose of administration, the sabha created "committees." The management of tanks was entrusted to a committee called erivariyam. The committee appears to have changed its members annually. Kings, noblemen or the villagers for the purpose of maintenance of tanks made endowments in the form of gold or land. The erivariyam took up the responsibility of regular maintenance, failing which the members of the assembly were liable to fine and punishment. The leadership of the village was in the hands of farmers, who knew the importance of irrigation works and paid much attention to their maintenance.

Figure 1.6. Plan of Marudur Anicut

Source: personal

22After the decline of Tamil kingdoms, there was a series of invasions during the period 1300 to 1770 by the Delhi Sultanates, the Vijayanagar Empire and Maratha kings. They nominated viceroys to rule parts of Tamil Nadu. Each of these viceroys established dynasties and brought their own armies whose higher ranks were settled as non-cultivating aristocrats in the villages. Gradually, the villagers seem to have become progressively more "feudalized." The relatively democratic provincial assemblies of peasants disappeared. In most villages a distinctive class of non-cultivating land managers emerged who were in one way or other associated with the government which had usurped the region or been granted lands. This was the period when Dutch, French and British merchants entered and tried to gain control over the territories. During the latter part of 1700, there were a number of wars among the kingdoms, with the French and Dutch, on the one hand, and the British, on the other. The British East India Company progressively took hold of more areas. To pay for wars and for tribute extracted by bigger powers, up to 70 % of the net produce of paddy land was taken by the state, in 1768. Heavy revenue extraction and periodic destruction of irrigation works created hardship and famines.

The East India Company

23The East India Company made material changes in the village community structure. They dispossessed the village artisans and other village servants of their hereditary rights to land plots or to village produce. This eliminated important functionaries who maintained and managed irrigation. The colonial government adopted the village accountants and village policeman (Karnam and Thalaiyari) and paid them monthly cash wages. As the British insisted on cash payments of revenue, large proportions of crop had to be marketed. The ryotwari settlement, too, was of great interest. Under this settlement, the responsibility of commonly owned properties like eries vested with the government and therefore the maintenance responsibility fell to the revenue officials. The orientation of the revenue officials was of no help to the maintenance works. Therefore expertise was sought in this area and the civil engineers of the army were the only ones who could have been called experts.

24Thus a civil engineer was appointed as superintendent of tank repairs in the beginning of 1800, which was the first irrigation office created. In 1819, a whole department of civil engineers was created in the then Madras Presidency. Although the colonial government had to create a whole department for tank repairs, in a hundred years it never developed into anything more than a mere agency for financing and labour employment agency. As the revenues dependent on these tanks were considerable, the government was interested in the maintenance or restoration of the tanks to their former efficiency when deterioration occurred.

25The Famine Commission (1878-80) considered the important question of maintenance of tanks once they had been restored. The Commission recommended “Inspectors should be appointed in each district, whose duty should be to report annually to the Collector on every tank, besides special reports where necessary. These inspectors should be directly under the control of the Collectors, and their salaries should not be less than Rs. 150 per month.” “The repairs themselves”, they advised, “should be carried out by statute labour according to old custom called Kudimaramath.” The Commission submitted a draft bill for an irrigation act that was subsequently abandoned. Consequently, a tank superintendent was also not appointed, and maintenance was not carried out. As the traditional social responsibility of irrigation was eliminated, the government presumed that if an act were brought forward, the Kudimaramath could be revived.

26Hence the Government enacted the “Madras Compulsory Labour Act 1858”. Section 6 of the Act read like this:

27"Whenever by local custom any work for the purpose of irrigation or drainage, or connected therewith is usually executed by the joint labour of the village community, any person bound by such custom to contribute labour to such work, who neglects or refuses without reasonable cause to comply with a requisition for such customary aid made to him by the head of the village under the orders of the Thasildar or other superior revenue officer, shall be liable to pay a sum equal to twice the value of the labour which he is bound to contribute".

28The implementation of this act met with different situations in canal irrigation and tank irrigation. The efficiency of canal irrigation depends on the canals being periodically cleared of silt and on the observation of a fair standard of maintenance. The customary obligation to supply labour for this work was fully recognized. On the other hand, in the case of the tanks, the repairs could be postponed, and even if maintenance were neglected, the tank still had storage and served the purpose. Years of progressive deterioration would ultimately end in a major breach and then the government would have to intervene and restore the tank. This resulted in the cultivators losing all sense of their responsibility for the maintenance and upkeep of tanks, which custom had formerly imposed on them. Although the act elaborated the obligation of the raiyats, it never explained the obligation of the government. Hence the raiyats were not enthusiastic about Kudimaramath. The first Irrigation Commission (1901) observed that the maintenance of the tanks was not satisfactory. The irrigation commission made the following remarks:

29“We are ourselves reluctant to admit that so valuable an institution (Kudimaramath) is really dead and past restoration. We do not recommend its adoption for any repairs on a large scale. Government must, in its own interest, not only restore the tanks but also accept a liability for any important emergent works as well as for periodical general repairs, the avoidance of which would result in a great loss of revenue. But the frequency and cost of such emergent and periodical repairs will depend upon proper attention being paid in the intervals to the upkeep of works by those interested in them – a stitch in time that will save nine”.

30During the colonial period, the community organizations weakened. But still the cultivators probably kept them alive by their own effort since there was no other alternative available. After independence, canal irrigation was developed in a massive way, also in the tank-irrigated areas. Ground water exploitation got a lift in the form of heavily subsidized electric and diesel pumping and the availability of institutional finance. This resulted in financially better-off farmers favouring irrigation in the tank command. Their desertion further weakened the community organization. Thus only the poorest people who could not afford ground water equipment remained interested in tank irrigation.


31We have seen that the modern development initiated under British rule and accelerated by the successors contains many weaknesses. On the other hand, traditional irrigation and water management were not as backward as has been thought. A scientific fact is that continuous rice monoculture eventually leads to a salinity problem unless water management is followed. Rice monoculture in the Kaveri delta and Thamiravaruni valley has existed for more that 2000 years. It indicates the best water management followed in these parts. Detailed knowledge of traditional resource management methods may not only help in the better formulation of new development projects but can help in the gradual rectification process of the existing problems in the irrigation system, for the knowledge of water management handed down from generation to generation is extensive. Let us undertake detailed studies on these aspects to throw more light on what has been briefly considered in this paper.

32When the British came to India, they found that the pattern of agricultural development differed in western Europe and in Asia, and for them what they found in Asia appeared backward ‘and stagnant. For the Asians, the Europeans appeared inefficient. Sir Arthur Cotton wrote (1874):

33“There are multitudes of old native works in various parts of India. These are noble works and show both boldness and engineering talent. They have stood for hundreds of years. When I first arrived in India, the contempt with which the natives spoke of us on account of this neglect of material improvements was very striking; they used to say, we were a kind of civilized savage, wonderfully expert about fighting, but so inferior to their great men that we would not even keep in repair the works they had constructed, much less even imitate them in extending the system. All this was perfectly true, until lately”.

34The situation has not changed much even now. We are fortunate to inherit these tanks, which after thousands of years are still in good condition. Are we paying attention to their maintenance? Are we going to leave them in good condition for our future generations?



Alexander K., 1974. The Riddles of Three Oceans, Peoples Publishing House, Moscow.

Bishop R. Caldwell, 1887 (Reprint 1989). A History of Tinnelvelly, Asian Educational Services, New Delhi.

Dharma K. (ed.), 1982. The Cambridge Economic History of India 1757- 1970. Vol 1, Orient Longman, Hyderabad.

Gomathinayagam, P., 2000. History of Irrigation in Tamil Nodu {Tamil), Sneha Publishers.

Janakarajan, S., 1993. “In Search of Tanks – Some Hidden Facts”, Economic and Political Weekly, Vol. XXVII No 26, 26th June.

Ludden D., 1989. Peasant History of South India, Oxford University Press.

Meenakshi, C., 1956. Pallava Administration, Madras University Publication.

Mohanakrishnan, A., 1989. Selected Papers on Irrigation, Publication No.9. Irrigation Management Training Institute.

Muthukumaran, S., 1992. Some thoughts on the water resources of Tamil Nadu, PWD, Govt, of Tamil Nadu.

Nagasamy, R., 1987. “Maintenance of Tanks – Then and Now”, The Indian Express, Nov. 1987.

Pate, H. R., 1915. Gazetteer of Tinnevelli, The Government of Madras.

Proceedings of the Chingelpet District, Historical Seminar, 1978.

Sengupta N., 1985. “Irrigation: Traditional Vs Modern”, Economic and Political Weekly, Special Number, Nov. 1985.

The Tinneveli District Gazetter, 1915.

Wittfogel Karl, A., 1957. Oriental Despotism: A Comparative Study of Total Power, Yale University press.

Table des illustrations

Légende Figure 1.1. Kumili– Plug and socket sluice
Crédits Source: personal
Fichier image/jpeg, 37k
Légende Figure 1.2. Madai (Paddle) / Madagu (Screw gearing shutters) sluices
Crédits Source: personal
Fichier image/jpeg, 36k
Légende Figure 1.3. Tank Weir – Dam stones
Crédits Source: personal
Fichier image/jpeg, 58k
Légende Figure 1.4. Joints in an anicut
Crédits Source: personal
Fichier image/jpeg, 28k
Légende Figure 1.5. Hydraulics of Duckbill Weir
Crédits Source: personal
Fichier image/jpeg, 41k
Légende Figure 1.6. Plan of Marudur Anicut
Crédits Source: personal
Fichier image/jpeg, 51k


Water Resources Organization, PWD, Government of Tamilnadu, Chennai
graduated in Civil Engineering, with a postgraduate degree in Hydraulics and Water Resources Engineering and a Ph. D. in Economics. After graduation, he joined the Public Works Department of the Government of Tamil Nadu. He held progressively responsible positions in planning, design and construction of irrigation projects for fifteen years. Subsequently, he served as Assistant Professor in the Irrigation Management Training Institute, Thiruchirappalli for fifteen years before taking up the current assignment in the Centre for Water Resources, Anna University. His main interests are in water management, historical studies of irrigation and studies on conflicts over water resources. He has published a book on the History of irrigation in Tamil.

© Institut Français de Pondichéry, 2003

Conditions d’utilisation :


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search