Pramāṇa · Search · About · API

History of Technology in India, Vol. I: From Antiquity to c. 1200 A.D. v1.c32.s2: Irrigation Works

History of Technology in India, Vol. I: From Antiquity to c. 1200 A.D. v1.c32.s2

Language
English

Contents

Irrigation Works

v1.c32.s2.1

Epigraphy and literature refer to this systematized practice of irrigation works as *Pāthas-Śāstra*. In Sanskrit *Pāthas* stands for water; as Varuṇa, the Lord of the waters, had borne the name *Pāthas-pathi*, hydrology, which deals with water, was appositely termed *Pāthas-Śāstra* (Mac donell, 1924, p. 160). Although regular treatises on this subject are not presently available, mention of the construction of reservoirs, canals and sluices with day, date and year and the arrangements for the distribution of water therefrom can be gleamed from a large body of inscriptions, copper plate grants and to a certain extent from the literary works. Besides, the works on architecture, the *Viśvakarmā Vāstuśāstram*, also gives an account of the construction of reservoir and wells. It would, therefore, appear that a special treatise on the science of hydrology must have existed in ancient times.¹

v1.c32.s2.2

There were specialists in the construction and execution of irrigation works. The *Yajurveda* pays tributes with reference this: "Homage to him expert in constructing wells and water falls. Homage to him who knows how to construct canals and tanks (*Kulyāyaśca sarasyāyaśca*)" (*YV*. XVI. 37). The persons who possessed the knowledge of the science of water was known as *āpo adhyānvacāris* (*YV*. XX.22) in the Vedic period. Engineers specialised in the construction of irrigation works were called *Jala Sūtrada* (Hydraulic Engineers). We find references to them in the inscriptions mostly from South India. Similarly, there were water-diviners (*Kūpadarśakas*) in the south (*SII*, XII p.2) who were professional water-diviners eking out a living by private practice. It is believed that they might have been working as inspectors of wells in those days. The duties of the engineers were purely technical. They were perhaps employed in the department of waterworks (*vāri-gṛha-karaṇa*)². It was a separate body under which many subordinate department functioned. The hydraulic engineers did not confine their attention to the building of dams, tanks, canals etc. but the maintenance and repairs of irrigation works were also attended to. So there seemed to remain a sharp distinction between the hydraulic engineers and water-diviners. In the earliest traditions of Kashmir recorded by Kalhana in his *Rājatarangiṇī* (Stein, p. 200), the construction of irrigation canals and other waterworks by the genius of engineer-minister under Avantivarman increased the prosperity of the valley. He regulated the river Vitasta by systematic arrangements for the construction of irrigation channels. For these the water of various hill streams was utilised as well as that of the main river. "After examining the different classes of land he procured a supply of water for the villages, which thus were no (longer) dependent only on the rainfall. After watering all village (lands), he took from (each) village (some) soil, ascertained by (observing) the time it took to dry up, the period within which irrigation would be required (for each soil, respectively). "He (then) arranged (accordingly) on a permanent basis for the size and distribution of the water course for each village and by (using for irrigation) the Anulā and other streams, embellished all regions with an abundance of irrigated fields which were distinguished for excellent produce". In the south under the supervison of the Hoysala Ballala Kings the hydraulic engineers brought the waters of the Yagaci river which flowed through a valley ten miles away and divided by a range of hills from the Halebid valley by af channel to supply the capital with water and fill the neighbouring tanks (IA, 1, p.44). These two recorded instances show the role of the hydraulic engineers in the execution of irrigation was considered important in the past.

Dams

v1.c32.s2.t1.1

Dams are barriers, natural or artificial, to hold back flow of waters. They were in existence from time memorial. Such prehistoric dams, now called Gebr-band, are still to be found on many water-courses in the Western parts of the region. The pre- Aryan method of agriculture depended upon natural flood and flooding the lands on the banks of smaller rivers by means of seasonal dams without regular masonary to obtain the rich alluvial deposit of silt to be stirred by the barrow. These artificial barrier dams broken by Indra are mentioned in the Rgveda (2.15.8, vide Kosambi, 1975, pp. 74-75) in connection with the Vṛtra myth. Here the term Vṛtra stands for ‘obstacle’, ‘barrage’ and so on. Such artificial dams are mentioned in the Rgveda as ‘rodhas’. These mountainous dams had flood gates which are described as the soft belly of the mountain from which the waters flowed. The flood gates are mentioned as bila, ūdha or dūra (RV. 5.32.11, vide Shendge, 1990, p. 255).

v1.c32.s2.t1.2

The bridge-cum-dam are generally mentioned in Sanskrit literature as Setu. The word ‘aṇai’ in Tamil denotes a dam which we find mentioned in Tamil literature and inscriptions. Hemadri in the Dānakhaṇḍa of his Catur-varga-cintāmaṇi quotes from Devipurāṇa which says “The measurements of a superior type of Pālibandha are given as 200 cubits by the specialists in their treatises” (Satya Srava, 1953, p.5). The term Pālibandha possibly denotes a dam or embankment. Another term Vipāli mentioned in the Devipurāṇa must be smaller in dimension or subsidiary to Pāli. So the system of damming the rivers was practised for the purpose of irrigation.

v1.c32.s2.t1.3

The Kunāla Jātaka (prior to 4th cent. B.C.) mentions a dispute between the Sākiya and the Koliya tribes for the distribution of water from the dam across the river Hohini (Cowell, p. 19; Gupta, 1933). The earliest known reservoir with a dam is the famous ‘Sudarśana’ lake in Kathiawar (c. 300 B.C. - 457 A.D.) (EI, VIII, No. 6, p. 46). This lake was formed by a natural dam and furnished with well-provided conduits and drains. An important fact about this reservoir-cum-dam is that it continued to serve the Saurashtra region for about 400 years without any major repairs. The masonary work must have been very strong. However, the lake suffered heavy damage by a storm during the region of Rudradaman I in 150 A.D. The restoration work was carried out by his governor. After a gap of 300 years the lake again burst out during the reign of Gupta Skandagupta whose governor carried out the repairs and restoration of the embankment in 456 A.D.

v1.c32.s2.t1.4

In a late Tamil literary work, Cola-maṇḍala Śatakam (verse 28) we find a reference to the construction of a great dam known as Kal-Aṇai or the Grand Anaicut across the river Kaveri (Nilakantha Sastri, 1932, p. 36). This anicut was not made of earth but of brick and stone. These is no conclusive proof to ascribe the still existing ‘Grand Anicut’ to the early Cola king Karikalan. So the speculation is still rife as to the ruler who was responsible for this great work. Probably the greatest achievement of ancients in executing a large dam-cum-reservoir was the Bhojapur lake near Bhopal built in the 11th century by Bhoja Paramāra (died A.D. 1055-56), the King of Dhārā. This lake was created by carefully erected dams of moderate size and covering an extensive area of 250 square miles. They consisted of an “earthern central band faced on both sides, outer and inner, with immense blocks of stone laid one on the other without mortar, but fitting so truly as to be watertight, the two faces sloping inwards from the base. The lesser opening (barrage) was closed by a band 87 feet in height, and 300 feet thick at the base, or even more the greater (barrage), by one in places 40 feet high, and about 100 feet broad on the top; and, though the first mentioned band is now a complete wreck, the latter is intact and still continues to turn the river Kaliasot into Betwa, and from its top the old bed of the stream is recognisable” (IA, XVII, p. 350ff). The lesser but higher barrage was breached by Shah Hussain (Hoshang Shah) in the 15th century, and has never been restored.

Reservoirs

v1.c32.s2.t2.1

Reservoirs are constructed to impound water to irrigate large tracts of arable land in the absence of rivers and dependable channels leading off from them. Usually a reservoir is surrounded on three sides by earthern embankments and the fourth side is open to the catchment area from which water flows down to collect in the basin of the reservoir. The water reaches the fields by gravity flow, below the embankments through openings or sluice gates or valves. The reservoirs which are fed by channels from rivers are known as “system tanks”. These types of reservoirs were constructed mostly in the southern region to augment irrigation needs. The artificial reservoirs, usually referred to in Anglo-Indian jargon as ‘tanks’, were variously known as *śarasi* (Sanskrit), *ayam*, *eri*, *taṭāka*(Tamil), *samudra*, *sarovara* (Kannada) and so on. The reservoirs were named after the benefactors, royal dignitaries, rulers, dynasties, deities with the above mentioned names as suffixes. To indicate the variation in size and economic importance of tanks, *eri*, *samudra*, *sarovar* applied to very big water formations. Lesser formations, artificial or natural, were also designated by such names.

v1.c32.s2.t2.2

Much attention was paid to the proper location of tanks and dams. The *Mahābhārata* (Roy, Eng.Tr. II, p. 14) stresses that large tanks and lakes should be constructed at proper distances. Manu, the Law giver, advocates that dams, wells, tanks and *prasarvaṇas* should be built where boundaries meet (*MS*, VIII, 248; vide also Eng.Tr by Buhler). As their locations were vulnerable to attacks by the enemies the idea of locating them on the boundary lines was not advocated by Kauṭilya. He recommends that the tanks and dams should be constructed in the centre of a province or village (*AŚ*, III, 51, Shamasastry, 1961). Tiruvalluvar in his *Tirukkuṛal*; ‘Sacred Couplets’ (22.5; vide Eng.tr. Pope, 1886) compares wealth in the hands of the liberal minded people to the water in the Central tank and this confirms that tanks were located at the central place for all to benefit, such public tanks were known in Tamil Ūruṇi-kuḷam (drinking water pond for public).

v1.c32.s2.t2.3

The site selected for the construction of large reservoirs by the tank builders was in between two mountains; for the mountain slopes on either side served as the natural embankment. This sort of arrangement undoubtedly increased the storage capacity of the reservoir. This technique is, of course, possible where there is such a natural arrangement. The famous Sudarsana lake in Kathiawar was located round the mountain Girnar. It was “so well joined in construction as to rival the spur of a mountain” (EI, VIII, No. 6, p. 46). The Ciramegha taṭāka constructed in the first quarter of the 7th century A.D. by the Pallava Mahendravarman is by far the largest tank in the Cingleput district, Tamil Nadu. It is situated in between two hills which form the natural embankment for it. “The bund rests upon the bases of two hills and lets rise here and there in the centre of the reservoir making it the prettiest tank in the district” (Madras District Manuals: North Arcot, II, p. 305). As already mentioned, the Bhojapur lake near Bhopal was created by constructing two solid masonry dams across the watershed of a circle of hills. The Rāmappa lake constructed during the reign of the Kākatīya Ganapatideva at a distance of 44 miles north-east of Warangal (Andhra Pradesh) is an excellent testimony to the skill of the Kākatīyas in irrigation works of a high order. This magnificent lake is formed by a ring of hills on three sides with a colossal bund (embankment) on the northern side (Gopal Reddy, p. 62). The same technique was applied by them in constructing a tank at Buddhapuri (Andhra Pradesh) (CTI, p. 52).

v1.c32.s2.t2.4

After selecting the site either in a hilly tract or in plains, the bed of the reservoir in the basin system was strengthened with hard clay. Further, the entire area enclosed by a group of hills or earthern embankments was adjusted to form the bed of the reservoir. The elevation (kūrma) in the middle of the tank bed was totally avoided. Otherwise, the storage capacity of the reservoir would be reduced. The bed of the reservoir at the egress was formed of stony bed along a tortuous line, so that the water issued from it whirled round with great force, forming strong eddies (EI, XIV, p. 108, f.n. 8). It was provided with an inlet and outlet channels and the inlet channel was normally three yojanas distant from its source. In the basin system it was also the practice to carry the flood waters into the basin by canals. Main canals carried the water from a river. Distribution canals fed the water from the main canals into the basin. A dam across the portions of hills was constructed of a firm and compact stone wall. The stones from the quarry were straight and long to withstand pressure and strain. In the mountainous tract the question of an earthern embankment strengthening it with stone facing did not arise; whereas in the level plain it was necessary to throw up earthern banks all around the tank bund.

v1.c32.s2.t2.5

Taṭākas were dug and built either in the middle or the peripheral area of the village. Care was always taken that in the site chosen there were springs of good drinkable water. A well was usually dug in the middle of the tank with its length in the north south direction. The tank was also fitted with an inlet and outlet channels for filling and emptying it (Kalavali, 2:4; Nānmaṇikkaḍigai, 73:3). Big tanks attached to temples were provided with a number of deep sunk wells (urai-kiṇaru) and maṇḍapa at the centre which was known in Tamil as nīrāli maṇḍapa. Such tanks were generally provided with flights of steps, a covered verandāh or corridors or platform all round and ornamental gateways on all the four sides (VV. p. 51).

v1.c32.s2.t2.6

Although isolated tanks which were fed by the run off waters from their own catchment did exist, it was common to have tanks as part of an interconnected chain of tanks, wherein the surplus waters of one flowed down to the next tank down the line and so on. It would be of topical interest here to quote the rules laid down in the Arthaśāstra.

v1.c32.s2.t2.7

“The water of the lower tank, excavated later on, shall not irrigate the field (already) irrigated from the higher tank; and the natural flow of water from a higher to a lower tank shall not be stopped, unless the lower tank has ceased to be useful for three consecutive years” (AS. II. 9). The real significance of this injunction laid down by Kauṭilya is that there was such a system of carrying the surplus water of one tank to another in ancient times. We have two classic instances from the south. The Vīranārāyaṇa eri or Viranam tank (first half of the 10th century A.D.) is at the fag end of the Kaveri delta near Kattumannar Koil (South Arcot District) and 8 miles west of Gaṅgaikoṇḍasolapuram (Tirucirappalli district). The bund of the extant reservoir alone is 10 miles in length and has amean width of some 3 miles. “The circumference of the tank when it is full is nearly 25 miles” (GAAD, 1906, p. 132). The surplus water of the famous ‘Colaganam’ tank (now in ruined condition) constructed by the Cola Rajendra I in 1012-14 A.D. at Gaṅgaikoṇḍa-solapuram went to make up this tank through a canal. Another interesting instance comes from Karnataka. A Kannada inscription (IA, VI. No. 32, pp. 138 ff.) dated 1096-97 A.D. from Kāṭṭageri (Bādāmi Taluk, Katadgi district) refers to the existence of a ‘Larger Big tank’ during the rule of the Western Calukya ruler Vikramaditya VI. J.F. Fleet, who identified the place and the tank based on the inscription, attributes three reasons for the sustained water supply in that area after in situ inspection of the place. “The system of tanks at Kāṭṭageri was formerly a large one. First, there is the small tank, on the west of the Fort, and close under the walls ... Secondly, there is a larger tank, on a slightly higher level, which if repaired, would, when full, include the proceeding in itself, formed by an embankment running to the south from just below the south-east corner of the Fort. And thirdly, about 1/4 of a mile away to the south-east of the proceeding, and on a lower level there are the remains of a large embankment, now breached and quite useless, which when in repair, evidently formed a tank of considerable area”. From the above description, it is discernible that the method of draining the surplus water from the tank at the higher level to the tank at the lower level must have been successfully applied in this case. These two instances prove that the ancients had a sound knowledge of hydraulic engineering while planning and executing the irrigation works.

Sluices

v1.c32.s2.t3.1

Sluices are contrivances in the form of gates or valves through which the flow of water can be controlled or stopped. They are also used for surplus vents. For better flow of water from the reservoir, the sluices are usually installed at points of maximum depth and quite far from the embankments. The sluices were not unknown to the pre-Aryans who had barricaded the streams of seven rivers, Sapta-Sindhavah (RV. I. 51.4; VII 82.3). It is disappointing to note that the northern sources do not offer any tangible material in this connection. Udagabandha (Sanskrit), madagu, kalingu, tūmbu (Tamil), tubu (Kannada) and tumus (Telugu) are the terms generally met with in the inscriptions to denote a sluice (EI. III. No. 20; Silappadikāram, X. 109; XII. 4.1; ANRIE No. 257, 1952; No. 552, 1902, SII, III, p. 392).

v1.c32.s2.t3.2

In the Paripāḍal (verse, 20, ll.90-107; see also Balakrishna Mudaliyar, 1960, p.38), one of the Sangam anthologies, we come across a passage in which it is mentioned that the river Vaigai near Madurai was then a mighty river, and to control and regulate it when in freshes, several water-heads and large underground sluice shutters were constructed for the purpose. So it is not a matter of coincidence that we notice in the early and later Pāṇḍya records from the 8th century onwards a good number of references to the construction of sluices to river channels and small-sized reservoirs in and around Madurai.

v1.c32.s2.t3.3

References to the installation and renovation of sluice gates are more in South Indian lithic records pertaining to the Pāṇḍyas, the Pallavas and the Colas. It is equally interesting to note that the tanks constructed in one period were provided with sluices in another period. But they add a little to our knowledge of the types of sluices used in the regulation of water.

v1.c32.s2.t3.4

Depending on the size and water-spread of the reservoir the number of sluices were determined. The Tirayaneri in the Cingleput district constructed in the first half of the 8th century A.D. by some predecessor of the Pallava Nandivarman II was provided with 23 sluices and stone revetment (Sewell, 1882, p.188). The extant Viranam tank or Vīranārāyaṇppereri, a large reservoir, constructed by the Cola Parāntaka I (c. 950 A.D.) at the fag end of the Kaveri delta (South Arcot district, Tamil Nadu) has 74 sluices (GAAD, 1906, p. 132).

v1.c32.s2.t3.5

Similarly, the ‘Colagaṅgam’ reservoir constructed by the Cola Rajendra I was provided with several sluices and channels for the irrigation of a large area. During the period under review, the sluices were installed by the use of dressed granite slabs in place of laterite blocks; and some times the planks of palmyra trees joined together by means of ropes were also used as a make-shift arrangement (ANRIE, No. 417, 1912). A single evidence from Sirgāli (Tanjavur district, Tamil Nadu) records the exact measurement of a sluice (ANRIE, No. 499, 1918). It measured in modern conversion 35 inches in length and 8 inches in breadth. It is presumable that this sort of narrow sluice shutters were put to use to regulate the water-supply from the branch channels.

v1.c32.s2.t3.6

Head Sluice: The Head-sluice (Kūṛṛan-vay), as we understand from the Pallava records, is a sort of regulator. It is usually a strong masonary work at a point where the canal takes off from a river or tank; and its purpose here is to control the supply entering the canal. Muka-vāy, talai-vāy, vāyt-talai are the other terms in Tamil to be identified with the head-sluices (SII, I, 151; EI; V p. 49). Above the head-sluice was constructed a ‘head of water’ (uvantri or taluippelai) with a large mouth or opening which conveyed the excess flow of water entering the canal from the main sluice (SII, II. No. 93, p. 11; III, p. 530).

v1.c32.s2.t3.7

A remarkable piece of native engineering work, which is still in a fairly good condition supplying water to the town of Tiruchirappalli (Tamil Nadu) is the massive head-sluice (Vayttalai) of the Uyyakondan channel, which branches off from the Kaveri near the Veṭṭuvāyttalai Railway Station. This head sluice is made of a solid granite block.

v1.c32.s2.t3.8

On the bridge which crosses the head-sluice is placed a stone, which is stated to have originally formed part of one of the pillars of the sluice itself. This head-sluice was constructed about A.D. 1205-06 for conducting water to flow into the Uyyakondan Channel during the reign of the Cola Kulottunga III (ANRIE, No. 70, 1890; 1891 para 4; SII, IV, no. 394, p. 119). Another example of such a nature comes from the same district. This head-sluice, now popularly known as nāṭṭu-vāykkāl (native channel) or Periya-vāykkāl (big channel), is on the northern bank of the Kaveri near Musiri (Tiruchirappalli district). This head- sluice is spanned by a bridge across the channel Periya-vāykkāl. On one of the side walls of the sluice close to the bridge is an inscription of Rājarāja Cola III which records that in the fourth year of his reign (A.D. 1219), the head-sluice was built of stone at Musiri (ANRIE, No. 72, 1890; 1891, para 4; SII, IV. No. 396, p. 120).

v1.c32.s2.t3.9

The major part of the Pandya region in the South (Madurai and Rāmanāthapuram districts) depended on rain-fed servoirs for irrigation. They were carried out by the early Pāṇḍyas (7th and 8th centuries) and continued by the later Pāṇḍyas from the 9th century onwards. All the surviving reservoirs, mostly medium- sized, with stone sluices (Karttūmbu) of the period would indicate the technical skill involved in the construction of reservoir for irrigation. “The position and depth of the sluices demonstrate adequate knowledge in choosing the most suitable points and ascertaining the correct sill level” (Rajan, 1986, p. 159).

v1.c32.s2.t3.10

The following information regarding the working of the tūmbu sluices is based on in situ inspection of certain early sites by Rajan Gurukkal, since none of the tūmbu inscriptions is published (Fig.1). “The tūmbu was a single valve system operated without any devices for mechanical advantage. It consisted of two granite pillars installed in the reservoir on either side of the sluice-mouth. The pillars varied in height, according to the depth of the reservoirs. The distance between the two pillars was about one to one and a half meters. The pillars were connected to each other by a series of cross-slabs from the base of the sluice to the top. Each cross-slab had a hole in the centre with a diameter of about 5 to 8 inches. The base of the system was a rectangular enclosure built in granite on the bed of the reservoir. It had a rectangular opening of 1 ×2 feet. The stone above the opening had a long aperture and the stones on either side of the opening had slots, obviously for the movement of the shutter plank. It appears that, through the series of holes in the cross-slabs, a wooden rod lowered to the sluice-mouth. The edge of the rod must have been connected to a shutter-plank lowered through the long aperture so as to cover the opening of the sluice. The operation of the system was quite simple. The sluice-mouth could be opened by lifting the rod. The lifting of the rod was not difficult since a little opening would cause enough upward thrust from the flow of water for the shutter-plank to move easily. The sluice opened into an underground conduit which lead water to a well called etirakkiṇaru built outside the embankment. Unlike the ordinary wells, the etirakkiṇaru was constructed upwards Fig. 1. Tūmbu (sluice) system

v1.c32.s2.t3.11

Valve-rod

v1.c32.s2.t3.12

Shutter-plank

v1.c32.s2.t3.13

Sluice-mouth

v1.c32.s2.t3.14

Shutter-ball

v1.c32.s2.t3.15

Sluice-mouth

v1.c32.s2.t3.16

Fig. 2. Kumiḷi (sluice-pit) system (reproduced from Studies in History, Vol. 2, No. 2, 1986)

v1.c32.s2.t3.17

from the ground level. It had openings at its base into channels running in various dissections. The flow of water through the channels required no contrivance since it worked on gravitational force.

v1.c32.s2.t3.18

"The sluice-pit" (*kumiḷi*) was another system (Fig.2) for the regulation of the flow of water from the reservoir, used at required centres. In this case, the sluice-mouth was a pit, also enclosed and with a circular opening on the surface. The rod lowered through the cross-slabs in the system must have been connected to a global stone or wood that acted as the shutter. Here the difference was in the nature of the valve of the sluice. In the case of the *tūmbu*, water from the reservoir flowed into the sluice from the side and, in that of the *kumiḷi*, from above" (Rajan, 1986, pp. 160-161).

Canals

v1.c32.s2.t4.1

The canals system of irrigation probably had its beginnings in the use of artificial water courses by the Vedic people. The word *kulyā* (*RV*. III, 45.3; X.43.7) refers to artificial courses flowing into a river. The expressions *khanitrima* meaning ‘produced by digging’ as an epithet of *āpaḥ* clearly refers to artificial water channels used for irrigation. Such artificial canals were dug by using a tool called *Khanitra* (*RV*. VII. 49.2; *AV*. I. 6.4). Rituals were practised before leading the river water into the canals (*AV*. III. 13; I. 6.4; XIX. 2.2; Ramgopal, 1953, p. 133). Kautilya classifies the countries making use of the water channels in agriculture as *kulyāvāpaṇam* (*AS*. II. 24.116, Shamasastry (tr), p. 130). Patañjali in his *Mahābhāṣya* speaks of canals for irrigating paddy fields *śalyartham kulyāḥ* (Patañjali I.1.24; I. 82, see also Agrawala, 1953, p. 204).

v1.c32.s2.t4.2

In Sanskrit literature the expressions such as *nāla*, *nālikā*, *praṇāḷi* are used to designate different types of canals, drains and channels and these unmistakably point to the existence of artificial water-courses used in irrigation in North India. The canals which also helped to stop inundation by rivers are referred to as *Jalanirgamah* (drains) in the *Amarakośa* (9.7; 9.36).

v1.c32.s2.t4.3

A survey of South Indian epigraphical evidence shows that a number of canals and channels bearing different names would indicate their specific use. The reason is that in the South each agrarian unit had a network of channels to supply water in each field, leading either from the river or the village reservoir. Generally the canals were known as *vāykkāl* (Tamil) and *kāluve* (Telugu and Kannada) (*SII*. II. Nos. 4 & 5). In the Tamil speaking areas the channels were dug from every conceivable source, natural or otherwise. The channels dug to bring water from the rivers to feed the tanks in the vicinity were known as *ārruk-kāl* (*SII*. II. p. 352). In the days of the Pallavas, there were native channels (*nāṭṭukkāl*) dug from the river Palar for irrigating the adjacent fields (*SII*. I. p. 150).

v1.c32.s2.t4.4

The high level channels *meṭṭu-vāykkāl* or *mugaṭṭu-kāl* (*SII*. II. p. 435) were expressly dug to conduct water from the tanks to water the fields. The drinking water channels cut from the rivers and tanks were specifically known as *sennir-poduvinai* and for digging such channels there was a provision of free labour known as *sennir-veṭṭi* or *sennīr-amañji* (*EI*, XVIII, p. 124; XV, p. 72).

v1.c32.s2.t4.5

Branch channels from the main canals and tanks were variously named as *kuraṅgāru*, *kāl* and *kilai-kāl* (SII, II. p.352, III, p.106 & 154; *EI*, V. p.49). Similarly, the field channels flowing towards the fields in a village were either called *kai-yāru* or *ahai-vāykkāl*. Small irrigation channels issuing from the reservoir was known as *kaṇṇāru*. The famous Vairamegha *taṭāka* constructed by the Pallava Dantivarman (c. 795-845 A.D.) had three branch channels and as many as twenty five such *kaṇṇārus* taking water from this tank and watering the fields in and around Uttiramerūr (Chingleput district). This is attested by a number of Pallava and Cola records.

v1.c32.s2.t4.6

*Feeder Channel*

v1.c32.s2.t4.7

Feeder channels were dug from rivers to feed tanks which in turn were used as the main source of supply to fields by means of branch channels. Such a channel is mentioned in a record of the Pallava Parameśvaravarman who dug a feeder channel from the river Pālār to feed the Paramesvara *taṭāka* which was constructed by him at Kañcipuram (Cingleput district) in the 2nd half of the seventh century A.D. (*SII*. I. p. 154). From the Udayendiram grant of the Cola Parantaka I (920 A.D.) we learn that a feeder channel was constructed to the tank at Vinnamangalam (Ambur, North Arcot district) (*SII*. II. p. 389).

v1.c32.s2.t4.8

Among the many feeder channels dug in the extreme South India, the Uyyakoṇḍān Channel dug by Rājarāja Cola I (c. 985-1013 A.D.) is an important native irrigation work which is still serving the district of Tiruchirappalli. This channel branches off from the Kaveri near the Veṭṭuvāyttalai Railway station, supplies water to a good portion of the district and eventually falls into a large tank in the village Valavandān-koṭṭai, about 10 miles to the North-East of Tiruchirappalli. Robert Sewell remarks “there is no doubt that the channel is a very old work” (*ANRIE*, No. 72, 1890; *SII*.IV, No.396,p. 120; Sewell, I, p. 269).

v1.c32.s2.t4.9

*Inundation Canals*

v1.c32.s2.t4.10

The method of irrigating fields through inundation canals is indeed very ancient. Though the northern source materials are silent about this practice, but considering the geographical conditions and the river systems, the inundation canals dug from the rivers for irrigation purposes must surely have existed in many parts of north India. In the south, irrigation by means of inundation canals was quite common in the days of the Pallavas as evidenced by the records of the period. Shallow cuts are made through the river banks (*kulai*) into which the excess water flowed when the level of the water in the river rose higher in the floods which occur during the monsoon seasons. Usually the head of inundation canals are made on the exact or correct bank of the river from which they draw their supply. In the past, such spots were selected by experience only. In one of the records of Nandivarman Pallavamalla II we notice that the settlers of the newly constituted village were permitted to make inundation canals (*vellakkāl*) on the banks of the rivers Vegavati and Pālār and irrigate all the fields within the village. Though the use of such canals must have been extensively in vogue in other parts of India, but we hardly come across any references to such canals in the records (*SII*. II. p. 352).

Spring Channels

v1.c32.s2.t5.1

Spring channels (*ūrrukkāl* in Tamil and *ūttakālva* in Telugu) were also made use of for irrigating small parcels of arable lands and for collecting drinking water. They were made on river-beds, but the practice was not widespread because the non-availability of sufficient underground water during the hot seasons. Deep furrows were made on river-beds and water thus collected was channelised to the fields. The practice of collecting water from springs was quite prevalent during the days of the Pallavas whose records say that permission was granted to donees to dig *ūrrukkāl* from the beds of the rivers Palar and Vegavati (*SII*. I.p. 151). Even now such channels are found on the beds of the Pālār and other rivers in South India.

v1.c32.s2.t5.2

No tangible evidence is forthcoming from the above account regarding the materials and methods involved in the construction of canals. We shall now turn to archaeological evidences which throw much light on the ancient canals and their structural details. The interesting aspect is that each canal has its own distinctive features.

v1.c32.s2.t5.3

Besnagar

v1.c32.s2.t5.4

Excavations carried out by the Archaeological Survey of India in 1914-15 on the western bank of the river Bes at Besnagar, near Bhilsa (Gwalior), have brought to light the remains of an ancient irrigation canal (c. 300 B.C.). The masonry canal was 7 feet wide and 5 feet 6 inches high. The north-south limb of the canal was 185' 4", whereas the east-west limb could not be traced, since both the walls were broken off at the same point where the canal took a turn.

v1.c32.s2.t5.5

Most probably it extended as far as the Bes river which is hardly 2 furlongs away from the site. The masonry walls were plastered with lime mortar of a superior quality. D.R. Bhandarkar who supervised the excavation work opines that on chemical analysis it was found to rival that of the Romans. The use of such quality lime mortar is not only appropriate but essential also for such brick walls, which, otherwise, would have been easily destroyed by the percolation of water. The walls were deliberately sloped downward to counteract the pressure of water. Traces of a flight of steps found at the middle of the south wall of the canal suggested the various uses of the water that was flowing in the canal. That the canal served as a storage canal was indicated by the cross-wall joining the north-south walls at the end. By the application of suitable lifting devices water must have been raised from this canal to irrigate the surrounding area (Bhandarkar, 1914-15, pp. 69-71).

v1.c32.s2.t5.6

Kumrahar

v1.c32.s2.t5.7

A canal dating back to the Mauryan period was unearthed at Kumrahar, a village 3 miles to the east of Patna Railway Station. It was found along with the Mauryan wooden pillared hall and hence belonged to the same period. The canal was 43 ft. wide and 10 ft. deep at the centre. A.S. Altekar states that “this canal was ultimately connected with the Ganga and might have been used to waft, the huge pillars direct to the door of the hall from the chunar quarries where they were probably prepared” (Altekar and Mishra, 1959, pp. 25-30). Hence it is not possible to say whether the canal supplied water for irrigation of lands in the vicinity.

v1.c32.s2.t5.8

Ujjain

v1.c32.s2.t5.9

At Ujjain, a brick-lined channel 36 ft. wide at the top and 8' 6" at the base (bottom and foundation) was exhumed to a length of 150 ft. which ran south-west and north wards. The limited excavation did not reveal its connection with the river ‘Sipra’ which is quite close to the site. At the start of the canal there is a big pond like natural depression suggesting to be the feeding agency. It cannot be said with certainty whether this canal was meant for storage purpose or for irrigation (Banerjee, 1957-58, *IAR*, p. 38).

v1.c32.s2.t5.10

Kāveripūmpaṭṭiṇam

v1.c32.s2.t5.11

Kāveripūmpaṭṭiṇam or Pūmpuhār, the celebrated port city of the early Colas, is now an insignificant fisherman's hamlet on the eastern coast of Tamil Nadu when the river Kaveri joins the sea. Detailed excavation conducted by the A.S.I. at three places at Kāveripūmpaṭṭiṇam had brought to light, among other important monuments, an elegant structure - the water-reservoir at Vāṇagiri (*Fig.3*). Situated about a mile away from the sea, it was made of an earthern (24" × 16") bund with a facade built of bricks. It was beautifully designed with two graceful curves to receive the water and let into a pond. Probably it was a small feeding channel from the river Kaveri or its numerous off shoots. The paved inlet channel was covered by a corbelled arch (Raman, 1968, p. 239).

v1.c32.s2.t5.12

Fig.3. Outlet channel of a reservoir, Vāṇagiri, Kāveripūmpaṭṭiṇam (photo courtesy : ASI Southern circle)

v1.c32.s2.t5.13

Nāgārjunakoṇḍa

v1.c32.s2.t5.14

The ancient canal at Nāgārjunakoṇḍa, datable to the 3rd or 4th century A.D. when the Ikṣvāku dynasty ruled over the parts of Āndhradeśa with their capital at Vijayapuri in the picturesque valley is another example. “The exposed canal runs to a length of more than 800 yards flowing east-west with an average width of 50 feet and 6 feet depth. The embankments were constructed with a hard lime stone gravel with *Kaṅkar* and stones which were rammed to withstand the rush of water. They were gradually sloped to the canal bed which also contained similar flooring. In addition to this the southern side embankment was further strengthened by a random rubble wall of 4 feet wide” (Subrahamaniyam, *IAR*, 1958-59, p.8). The canal was built of large-sized bricks (12 × 12 × 3) and of extremely durable nature. Though the purpose of such a big canal as this is not clear, it can be reasonably surmised that it served as a water source for the dwellings in the vicinity, and other public use, in particular to the famous stadium, the Harati temple, great bath, etc. (Subrahmaniyam, *IAR*, 1954-55, pp.22-23).

Embankments

v1.c32.s2.t6.1

Embankments form an essential part in the construction of river banks, canals, and reservoirs. They are artificial wide wall of stone or earthern mounds which are built to keep the rivers, canals and tanks from overflowing the banks. They are essentially constructed to impound water and also to increase the storage capacity of reservoirs. The artificial embankments were used in the Vedic period.

v1.c32.s2.t6.2

They are mentioned in association with dams or barriers in the *Ṛgveda* as *Kṛtrimāṇi rodhāsi* (*RV*. II. 15.8). The embankments surrounding the rivers with whirls were known as *rodha cakraḥ* (*RV*. I. 190.7). They were, no doubt, earthern without regular masonry. So the idea of erecting embankments was current in those days and actually practised.

v1.c32.s2.t6.3

Kulai and *ārrukkulai* are the terms in Tamil used to denote the tank bund and river embankments respectively (ANRIE No. 292, 1908; *SII*.V.p. 603). The inscriptional evidence in Tamil clearly shows that the embankments of tanks were allowed to be raised within their limits to hold the maximum quantity of water. The Tiruvālaṅgadu plates of the Cola Rajendra I, for instance, the following specification: “the embankments of the tanks of this village shall be permitted to be raised within their (own) limits (to any suitable height) so as to hold the utmost quantity of water that may be let into those tanks” (*SII*. III, NO. 205. 11. 426 ff.; also ANRIE, No. 103, 1921).

v1.c32.s2.t6.4

We get a few references from the inscriptional sources about the material involved in the construction of embankments. While recording the restoration of the famous Sudarśana lake by Rudradaman I, the Junagadh inscription says that its embankments are stony, in breadth, length and height constructed without gaps as they are of stone (clay)" (*EI*. VIII. No. 6, p. 46). We do not know when the great embankment, over 100 feet thick at the base, finally crumbled. During the days of the Cola Karikāla the Kaveri was impetuous and of considerable magnitude. The Ceylon Chronicle *Mahāvaṃśa* says that he invaded the island of Ceylon (Sri Lanka), carried away about 12,000 captives and set them up to work on the construction of the banks which extended along the course of the river to a distance of about 100 miles from its mouth (Kanakasabai Pillai, 1904, pp. 8-9). This is well reflected in an eleventh century epigraph. “He (Karikāla), who curbed the pride of the subordinate, prevented the Kaveri, which by its extensive floods caused the earth to be deprived of its produce, by means of a bund formed of earth thrown in baskets carried in hand by (enemy) Kings" (TRV. III, pp. 154-155). This age-old and time-tested method of strengthening the river banks is still being adopted in rural area in South India. The excavation conducted by the Archaeological Survey of India since 1915 have clearly established that canal embankments were constructed with hard lime gravel mixed with stones which were rammed to withstand the rush of water. Bricks of unusually large size 24" ×12"×3" and of extremely durable nature were used in the construction of walls. Evidences show that in the later centuries the river banks or tank embankments were either raised out of earth or reinforced with stone revetment. In some cases granite stone blocks were used for the basement of embankments (Karthikeya Sharma, 1958-59, pp. 307 ff; Raman, 1968, p. 239).

v1.c32.s2.t6.5

The unabated technical skill of the tank builders in ancient India will be fully realised if one could make an in situ study of some of the tanks which still exists even to this day to afford irrigation facilities. A.F.Cox in North Arcot District Manual describes the Mahendra Taṭāka excavated by the Pallava Mahendravarman I in Mahendravāḍi (North Arcot District) thus: "The bunk (manmade embankment), is enormously high and might be restored to its original height, in which case a great extent of land could be brought under irrigation" (Cox, II. pp. 438 ff.). The tank at Kaveripakkam constructed by the Pallava Nandivarman III (c. 844-856 A.D.) is the largest tank in the same above mentioned district. Its extensiveness is exhibited by a "bund about four miles stretching from North to South" (Cox, II. pp. 438 ff). The bunk of the Viranam tank created by the Cola Parāntaka I is ten miles in length, and has mean width of above 3 miles. The large irrigation reservoir "Colagaṅgam" created by the Cola Rajendra I near the capital in the Tiruchirapalli district has an embankment of sixteen miles long running from north to south and provided with several sluices and channels for the irrigation of a large area (Nilakantha Sastri, 1955, p. 234).

v1.c32.s2.t6.6

Trees on the Banks of Tanks: The practice of planting trees on the taṅk bunds was in vogue in ancient times. The Sudarśana lake which was destroyed by a storm during the reign of Rudradaman in A.D. 150 was restored to its original glory by strengthening the dam and planting trees on all banks. In some places particularly in Āndhradeśa rows of trees called Kaṭṭava were planted on or by the side of the tank bund, evidently with a view to adding to its strength (CTI. No. 53; also Yazdani, 1964, p. 679). In Tamil country five fruit bearing trees, the Indian gall-nut, Nelli (Phyllanthus embalica or the embalika myrobalam), Tantri, Karañja-puṅga (Indian beach Pungania glabra) and Aśvatha (the Fig tree or Ficus religiosa) were generally planted all along the bunds of tanks. The principle which underlies the planting of such trees on the bunds of tanks is easily intelligible.

v1.c32.s2.t6.7

Trees arrest the soil erosion due to sudden or frequent floods and also hold up the water level which would fall without the banks. Hence the presence of trees near the tank was found very useful. Moreover, they contain some tannin in them. The tannin adds a slight sweet taste to the water and thus masks any excess of chlorine in drinking water. "It is also useful in cleaning the water of its impurities on the surface by carrying down to the bottom all dirt and clay that the water may contain, thus serving the same purpose as alum”. Hence these trees are generally planted near the bunds of drinking water tanks" (Venkatarama Iyer, 1916, p. 184).

Dykes

v1.c32.s2.t7.1

Dyke is a thick bank or low wall or embankment built to control water and prevent flooding. It is also a narrow passage or water- course dug to carry water away. Besides the use of canals and wells for irrigation, the construction of dykes (setu, which also stands for a dam) was known and used in ancient times. The importance of the construction of dykes is duly recognised by the Smṛti writers. Narada (Nārada smṛti, XI. 18) classifies the dykes according to their utility. Two main classes of dykes are mentioned in the Nārada Smṛti (XI. 20-2), namely Kheya which is dug into the soil in order to drain off excessive water and another called bandhya which is constructed as an embankment to prevent water from flowing away. The former serves the purpose of irrigation and the latter protects the fields from excessive water.

v1.c32.s2.t7.2

Both Yājñavalkya (Yājñavalkya Smṛti, II. 156) and Nārada (Nārada smṛti, XIV. 17) state that the erection of a setu made by the owner of one field in the neighbouring field should not be forbidden by owner of that field, if the loss of soil it causes (to the latter) is small as compared with the great benefit that it may offer. They also prescribe that one should make a dike or water-course on another’s land with the permission of that man or with the permission of the king as otherwise he cannot reap the benefit thereof. In South India such types of dykes might have been used for agricultural purposes; but we hardly come across any reference to them in inscriptions.

Wells

v1.c32.s2.t8.1

Well is a place where water comes from underground. Almost in all possible nook and corner of every village in India one finds the use of a well for irrigation and drinking purposes. Well irrigation no doubt extends the distribution network but slowly. However, it was common in those days, as it is today, to have a well dug in the midst of fields. Well water was utilised in the areas where the tank or canal irrigation was impossible or totally absent.

v1.c32.s2.t8.2

The discovery of a large number of wells within the housing complex at Mohenjo-daro points to the importance of well-water for domestic and ritualistic purpose. The occurrence of the word avata meaning a well repeatedly in the Ṛgveda (RV. I. 55.8; IV. 17.16; VIII. 49.6; X. 25.4) shows that the vedic farmers had known the method of digging wells. Such wells are described as unfailing (akṣita) and full of water (RV. X. 101.6).

v1.c32.s2.t8.3

Sometimes those wells appear to have been used for irrigation purposes, the water being led off into broad channels, sūrmi susirā (RV. VIII. 69.12; Macdonell & Keith, 1, pp. 39-40). Deep sunk wells whose bottom looks dark due to their depth are mentioned in the Jaiminīya-Brāhmaṇa (I. 292), a work related to Sāmaveda.

v1.c32.s2.t8.4

In the days of the Mauryan rulers, the farmers had to pay a revenue tax of 1/4 of the produce for using wells fitted with machines (AS. II. 24.117; see also Shama sastry, Eng. tr. 1967, p. 131). Pāṇini’s Aṣṭādhyāyī (IV. 2.73-74) mentions two types of wells: karkaṇḍu and sakandu. He also mentions wells on the banks of the river Vipāsā (Beas). Those dug on the right bank of the river were permanent wells, and those on the left bank were temporary ones subject to annual inundations.

v1.c32.s2.t8.5

In the early Tamil anthologies the expressions such as kūval, kulam, kiṇaru, ūruṇi, vāvi, keṇi and so on are mentioned to signify a well (Maṇimekalai. VII.76; Tirukkural, No. 523; Puraṉāṉūru, line, 132-133; Paṭṭiṉappālai, line, 244, see also Aiyangar 1928). These terms were evidently used to indicate variations in size and economic importance. One of the oldest regions in South India is the Koṅgu region (modern Salem and Coimbatore districts of Tamil Nadu). Although this region was watered by the Kaveri and its numerous tributaries, water scarcity was felt very much at all times. "The skill and energy with which the people of Kongu sunk very deep wells and brought out water by breaking hard rocks with their pickaxes drew the admiration of the ancient bards who celebrated them in their songs.

v1.c32.s2.t8.6

The wells supplemented to a great extent the limited water-supply of the people to convert arid tracts into arable land" (Subrahmanya Aiyar, K.V. II, p. 45). The Tamil epic, Manimekalai, mentions a well (endirak-kiṇaru) which was provided with filling and discharging pipes for conveying water (Maṇimekalai, XIX. 102, see also Aiyangar, 1928).

v1.c32.s2.t8.7

The lithic records from the north refer to two distinct types of wells - kūpa and vapi; the former is an ordinary well, while the latter appears to have been equipped with flights of steps and used for a tank of the smallest size. The South Indian records mention specific names to denote the wells the purpose for which they were put into utility - pānīya-kūpa (drinking water well), uruṇi-kulam (public water tanks), tirumañjana kulam (sacred tanks for use by the temples), puravai and kiṇaru (ordinary well), puṭṭai-k-kiṇaru (wells fitted with water-lifts) turavu (large wells for irrigation) and so on (EI. VIII. p. 145; SII, II. Nos. 4, 5, and 445; IPS. No. 156; SII. III. No 4; III. p. 102). The paṭṭiṉappālai (Chelliah, 1947, p. 48), one of the Tamil anthologies, records that in the outskirts of the city of Kāveripūmpaṭṭiṇam the inside of wells (urai-kiṇaru) were covered with baked clay instead of being built with stones. It is interesting to note from the Kannada inscriptions that the charity wells set apart for thirsty way-farers and cattle were specifically names as arabāvi (SII. XVII. No. 58. p. 68).

v1.c32.s2.t8.8

According to the Dharma Śāstras, the man-made receptacles are of four types kūpa, vāpi, puskarinī and taḍaga. The kūpa as a well is from five to fifty cubits in length (if rectangular) or in diameter (if circular). It has generally no flight of steps to reach the water. Vāpi is a well with a flight of steps on all four sides or on three or two sides or one side only and its mouth may be from 50 to 100 cubits; a puṣkariṇī is from 100 to 200 cubits in length or diameter and a taḍaga (a tank) is from 200 to 800 cubits. The Matsya Purāṇa states that a vāpi is equal to ten kūpas in merit (Kane, 1973, II, p. 893).

v1.c32.s2.t8.9

The Viśvakarma Vāstuśāstram (p. 51) clearly describes the vāpis. "Small tanks (vāpis) may be square, rectangular or circular in shape and their dimensions may be from 3 to 10 daṇḍas. It is enclosed by walls and has springs on all the four sides.

v1.c32.s2.t8.10

They are built down to a depth of about 10 cubits below the spring. They are also provided with bathing ghāṭs on all the four sides suitable for sporting in water. On one side of the tank shall be built a water-lifting device”.

v1.c32.s2.t8.11

Water-divination: If we are to take water divination as part of well irrigation, then we may say that water-divination is an important and independent science practised in ancient India to locate and avail the water sources from the subterranean region. The expression dakārgala means a branch of science dealing with water locked up or imprisoned in the bowls of the earth. Those who practised this science were known as kūpadarśakas. They leaked out a living by private practice for they paid taxes. It is believed that they might have been working as inspectors of wells in those days.

v1.c32.s2.t8.12

Varāhamihira’s Bṛhat Saṃhitā (Brh.S. pt.1, ch.LIV. pp.499-526 (Eng tr.) by V.Subrahmanya Sastri and M.R.Bhat, 1947) prescribes certain asterisms that are beneficial for sinking wells: They are Rohiṇī, Puṣyā, Maghā, the three uttaras (Uttaraphalgunī, Uttaraṣadhā and Uttarabhādrapadā), Hasta, Anurādhā, Dhaniṣṭhā and Satabhiṣaj. The author gives effects of sinking wells in the different quarters of a village or town. He says that wells sunk in quarters other than south-east, south-west, north-west, would lead to auspicious results.

First 76 of 197 passages. Read on

Linked texts (51)