Pramāṇa · Search · About · API

History of Technology in India, Vol. I: From Antiquity to c. 1200 A.D. v1.c24.s6: Irrigation

History of Technology in India, Vol. I: From Antiquity to c. 1200 A.D. v1.c24.s6

Language
English

Irrigation

v1.c24.s6.1

Water is essential for the growth of crops. Rain water is the first and the foremost source of water supply. But, in many cases it is either insufficient or irregular or not available at the proper time. Rain water has to be supplemented by man-made devices of supplying water. Even in the case of other natural sources of water, such as rivers, streams, lakes and ponds, human effort is required to carry or direct water to the field. Thus, irrigation falls into two clear categories, natural and artificial.

v1.c24.s6.2

Scholars have made several, sometimes conflicting, assessment about the nature and volume of rainfall in the Harappan period, hence we cannot determine precisely the extent to which agricultural prosperity of the Harappans was due to this source of irrigation. In the absence of any definitive evidence about artificial irrigation the fact of the predominance of winter crops in the Harappan economy has been taken to be based on ‘flood irrigation farming’ using moisture and silt provided by river floods.

v1.c24.s6.3

The ruined Gabbarbands (lit. stone dams) in south Baluchistan placed by Lambrick between c.3000-1000 B.C., have been related by Hargreaves with the Nal culture. Though no traces of gabbarbands are reported from Harappan sites, it has been suggested that the Harappans also used such dams for storing water for irrigational purposes.

v1.c24.s6.4

At Lothal a trapezoid reservoir measuring on an average 214 × 36 meters has been excavated. It is surrounded by an embankment wall of kiln-fired bricks, the shorter arm on the north having a gap for the inlet of water with another gap in the south providing for spillway. Following S.R. Rao (pp. 70-73) the excavator, it is taken to be a dock, but Leshnik (vide Possehl 1979; AA 1970. pp. 911-22) and Junghans (Possehl 1979. p.210) argue for its having been an irrigational tank. It had a capacity of about 500 cubit meters with provision for refilling. In any case, this shows that the Harappans had a developed tank technology. The Great Bath at Mohenjodaro is an illustration of it, but we do not have any indication to confirm its use for irrigation. This holds good for wells that have been exposed at a number of habitational sites of the Harappan culture and are the earliest excavated stained wells having burnt brick lining.

v1.c24.s6.5

The wells at Harappan sites show a developed technology. Those at Allahdino (Baluchistan) with their small diameter, location at the top of the site, and the divided drains with movable sluice block suggest well water irrigation (Possehl, 1979, p. 210). Some of the water lifting devices associated with the Harappan wells can be determined. Sign of friction by the ropes on the sides imply drawing of water by a rope with pots attached to it. It has been suggested that the water-lifting device of shaduf was in use. The picture of a man using shaduf at a well is embedded on a piece of lime at Mohenjodaro. The ring-stones recovered from Mohenjodaro and Lothal are explained as counter-weights for the arm of a shaduf. Mackay suggests that the scored pottery from Mohenjodaro was attached to a wheel, which was managed by an endless rope, for drawing water for irrigation (Marshall 1931, I, p.318)54.

v1.c24.s6.6

The earliest archaeological evidence about an irrigational device is available from Inamgaon, a Chalcolithic site. It is in the form of an embankment and a channel belonging to the early Jorwe period (c. 1300 B.C.). It is 240 m. × 2.25 m. on the right bank of the Ghod river and is built of stone rubbles set in mud in its basal courses. Parallel to it is a channel running north-west measuring 118m × 4m and 3.5m deep. It was dug through hard ‘marrum’ down to rock and prevented much loss of seepage. The huge basaltic outcrop on the left bank of the river checked the flow of flood water. The channel served not only for flood diversion but also for storing water (54000 cubic feet) of which excess could be diverted by gravity flow into the fields to the west and in winter when the water level sunk pot irrigation was resorted to (Misra and Bellwood, 1985, p. 388).

v1.c24.s6.7

The Vedic literature testifies to the realisation of the importance of rain. The high position of Indra in the Vedic pantheon and the allegoric references to his killing Vṛtra and releasing the cows glorify the phenomenon of rains. In ancient texts we have numerous references to the benefits of rains, to the eager waiting of it, and to the joy and festivities on its coming.

v1.c24.s6.8

The Vedic literature refers to sacrifices being resorted to for bringing rains. Yāska (Nirukta, II.11) names a group of hymns in the Ṛgveda (X.98) as Varṣakāmasūkta or hymns recited at a sacrifice performed to bring about rain. The belief in the efficacy of sacrifice to cause rains is recorded in the Gītā (III.14) and the Manusmṛti (III. 76). R. Gangopadhyaya (p.112) goes too far in asserting that this belief had ‘a substantial basis’ as ‘famines were few and far between in the time of the Ṛgveda when Indra worship was greatly in vogue’.

v1.c24.s6.9

The natural phenomenon accompanying rains was carefully observed and analysed. The Purāṇas describe the formation of clouds and their different types with their characteristics (Vā.Pu I. 28-46; Br. Pu. II 22.23.50; CXXV 11-35, see also Sen S.C.). The Arthaśāstra shows an early development of the knowledge of clouds when it mentions (II.24.9) three types of clouds which continually rain for seven days, eighty which pour minute drops and sixty which appear with a sunshine.

v1.c24.s6.10

The process of the formation of clouds and of rainfall was described in terms of an embryo and delivery. The pregnancy lasts six months and a half. It begins in autumn-winter, but there was a wide difference of opinion about the actual period of the formation of the rain-embryo. According to some it occurred after the bright half of Kārttika, but the majority view placed it from the first day of the bright half of Mārgaśīrṣa (Bṛhatsaṃhitā XXI. 5.6). The period of the delivery of the rainfall was determined accordingly (ibid. 9-12). Varāhamihira records auspicious marks for the nourishment of the rain foetus, symptoms in different months favouring their health, and also indications of miscarriage (ibid., 13-28).

v1.c24.s6.11

Another aspect of study concerned planets and other connected phenomena for determining the nature, volume and timing of rains. This was closely associated with the development of astronomy to which significant contributions were made by the requirements of the Vedic cult of sacrifices and later by the Jainas and various schools of astronomy, including Greek and Roman. The application of astronomical knowledge to agricultural work was noticed by Megasthenes (Fragment I, p. 41; Majumdar p. 237) who says that the philosophers of India, when gathered together at the beginning of the year, forewarn the assembled multitudes about droughts and wet weather and also about propitious winds.

v1.c24.s6.12

Rain gauging was prevalent in India even before Pāṇini (III. 4.32) who provides for terms referring to measures of rainfall. The Arthaśāstra (II. 5.7) describes the rain-gauge as a basin with a mouth one aratni (= 24 aṅgulas) in width. According to the Bṛhatsaṃhitā (XXIII.2) the raingauge is to be one hasta or cubit (18 inches) in diameter. When filled to the brim it indicates 50 palas or 1 āḍhaka of rainfall. Parāśara (as quoted by Bhaṭṭotpala on Bṛh. S. XXI.32) mentions a raingauge 20 aṅgulas (15 inches) in depth and 8 aṅgulas (6 inches) in diameter. When filled to capacity, it indicates one āḍhaka rainfall. Parāśara equates four āḍhakas to one droṇa. He mentions another method of measuring rainfall. Water spread over one dhanuṣ (=4 cubits) of land indicates one droṇa of rainfall. But, in the Kṛṣi-Parāśara (verse 26) the measurement of one āḍhaka of rainfall differs radically. It says that an āḍhaka measure of water is a hundred yojanas wide and thirty yojanas deep.

v1.c24.s6.13

The Meghamālā defines droṇa as a continuous rain for seven nights together (Shastri,p.298). Varāhamihira in his Samāsasaṃhitā (quoted by Bhaṭṭotpala on Bṛh.S., XXIII.2) refers to the system of measurement of rainfall as prevalent in Magadha, which seems to have been recognised as the standard one.

v1.c24.s6.14

Kauṭilya (AS, II. 24.7-8) says that a forecast of rainfall can be made by observing the position, motion and pregnancy of Jupiter (Bṛhaspati), the rise, setting and motion of Venus (Śukra), and the natural or unnatural aspect of the sun.

v1.c24.s6.15

Ṛtuviñāna or climatology developed along with astronomy. The Bṛhatsaṃhitā, the most standard work of this type, has many chapters dealing with several aspects of the subject. Varāhamihira himself refers to some earlier authorities, but their writings are not available in their original form. Quotations from these reproduced by Bhaṭṭotpala in his commentary confirm their existence. The early medieval period witnessed the composition of several works on weather-forecasting. Of these only the Gurusaṃhitā has been critically edited. Some others, such as the Mayuracitraka and the Meghamālā are yet to be printed. These books contain weather-maxims similar to those current in many provincial languages. Though their scientific validity has not been investigated, Indian peasants depend on them for the rich practical wisdom stored in them. They present a strange mix-up of astronomical considerations and natural phenomena. Though the indications appear to be superstitious in some cases, they were evidently based on careful observation spread over a long period.

v1.c24.s6.16

The Kṛṣi-Parāśara shows how astronomical knowledge was used for forecasting rainfall in the interest of cultivation. One could have a knowledge of rainfall on the basis of the planets which happen to be the lord and minister of the year (verses 11-22). It lays down indications for determining the nature of rainfall in various months from Pauṣa to Śrāvaṇa on the basis of the planets and natural phenomena (verses 30-64). The movement of planets from one zodiac to another is also taken to indicate rainfall and draught (verses 71-79). The text mentions certain natural phenomena and the behaviour of human beings, birds, animals etc. as indicating immediate rainfall (verses 65-70).

v1.c24.s6.17

The importance of irrigation for agriculture was well realised. The Arthaśāstra (VII.14.21) refers to the irrigational works (setubandha) as the source of crops. Bṛhaspati (XIV. 23) mentions the suitability of a field for irrigation as one of the qualities conducive to a rich crop. The prosperity of a kingdom was often attributed to its irrigational work. (MB. II. 5.77; Rām II.100. 44-45; Kirata,I.17). Megasthenes correlates the fertility of the Indian soil with the fact that a greater part of it was under irrigation from the water of the rivers (McCrindle, Frag.I, XI; Majumdar, pp. 232, 252). The vogue for irrigational works is indicated by the terms *devamātṛka* and *adevamātṛka* or *nāḍīmātṛka* (Amara, II.1.12; MB II. 5.77, II.51.11) referring to the classification of regions on the basis of their depending on rains or using irrigational work.

v1.c24.s6.18

Varāhamihira in his *Bṛhatsaṃhitā* devotes a full chapter (No. 53) to the subject of water-divining. It is termed as *dakārgala*, of which the more explicit form is *udakārgala*, Varāhamihira admits that he presented in Āryā verses the *Dakārgala* composed by Sārasvata and narrates in Vṛtta verses another written by Manu (verse 99). Though both these texts are not available in their original form, quotations from them are reproduced by Bhaṭṭotpala.

v1.c24.s6.19

Varāhamihira speaks of eight underground arteries of water, four flowing in the four cardinal points and four in the intermediate directions with a ninth which shoots upward (verses 3-6). He takes the presence of certain trees and ant-hills in different directions of specified trees as indicating underneath water at a waterless place (*ambu-rahita*).

v1.c24.s6.20

He mentions the depth at which water will be available, the nature of the water, the direction in which it flows, the type of stones and soil to be met at different levels and fish, frogs, rats, snakes and tortoise to be noticed (verses 6-85). The depth at which water is indicated for a desert land (*marudeśa*) is not applicable to a *jāṅgala* land (with little water). In the case of *anupa* land (with much water), if certain specified trees are found in ant-hills, then water will be at a distance of three persons length (*puruṣas*), in *jāṅgala* land at a distance of five *puruṣas* and in *maru* land at a distance of seven *puruṣas* (verses 86-89). There are general indications of the presence of water in the form of colour and nature of soil, availability of vegetation on them, and the presence of anthills and insects without home (verses 90-96).

v1.c24.s6.21

On the basis of the *Dakārgala* of Manu, Varāhamihira says that where trees, shrubs and creepers have glossy and thick foliage and have milky juice, there an underneath water artery exists. He names certain objects, grasses and trees as indications of water underneath. Water is also found at the bottom of a hill which is placed on another hill (verses 100-2). Varāhamihira mentions the nature of the soil, its formation and colour as indicating that the water will be sweet or salty (verses 103-4). Water is far from the surface where certain trees grow and where trees, shrubs and creepers have sparse leaves and are dry (verse 105). A soil of the colour of sun, fire, ash, camel and ass is devoid of water. If the Karīra trees have red sprouts and have milky juice, water will be available under the rock (verse 106).

v1.c24.s6.22

Varāhamihira distinguishes rocks on the basis of their colour for the easy availability or remoteness of water (verses 107-10). For breaking rocks, which cannot be easily split, he records ways and means on the authority of Manu (verses 111-17).

v1.c24.s6.23

The subject of water-divining was not developed further by any later scholar. We have, however, the *Viśvavallabha* of Cakrapāṇi⁵⁵ which deals with the subject of location of underground water in chapter 1. It is clearly based on the Bṛhatsaṃhitā of Varāhamihira from which many verses have been borrowed. We do find some elaboration and new information about the indicators of underneath water. There is some change in the scheme of the chapter with verses 32, 33 and 34 presented as separate sections dealing with water respectively in arid areas, marshy lands, and mountainous country. In the absence of the texts of other authorities mentioned by him, we cannot determine the extent to which Cakrapāṇi added details from his own experience.

v1.c24.s6.24

Varāhamihira makes it clear that the treatment of the subject of dakargala was for the digging of wells (Bṛhatsaṃhitā 53. 77, 121- 123). He considers wells situated in the Āgneya (east-south), Nairṛta (south-west) and Vāyavya (west-north) directions to be inauspicious and recommends their placement in the other five directions (verses 97-98). Likewise, he recommends a vāpī (reservoir) which is east-west long and not south-north; the latter is destroyed by the water agitated by winds. In such a reservoir the flow of water is to be enclosed by strong wood, stones or baked bricks and the embankments are to be hardened by being crushed by elephants, horses and other animals (verse 118). At one place there was to be an outlet gate with a well made bed, the channel lined with stones and a panel, without aperture, fixed in a frame and covered by grit heaped against it (verse 120).

v1.c24.s6.25

The Arthaśāstra (II.1.20) mentions two types of setus (embankments), the sahodaka referring to tanks, wells etc., which have a natural supply of water, and the āhāryodaka, which store water in reservoirs. Discussing the comparative benefits from the different types of irrigational works, Kauṭilya remarks that one which is of perennial water is better than that which is fed with water drawn from other sources, and of works with perennial water that which can irrigate an extensive area is better (AS. VII. 12.4.5).

v1.c24.s6.26

Nārada (XI.18) classifies irrigational works into two: kheya (canal) or that which is dug into the soil in order to drain off excessive water, and bandhya (embankment) or that which is constructed to prevent water from flowing away. The Bṛhatkalpasūtra Bhāṣya (123a) (Motichand, p.165) records the regional peculiarities of irrigation devices. Certain regions (like Lāṭa) depended solely upon rains, others (like Sindhu) used the rivers, some others (like Dravida country) relied upon reservoirs, some (like Uttarāpatha) used wells, at some places (like Banāsa) seeds were sown after the floods had receded, while in others (like Kananadvīpa) paddy crop was cultivated on boats.

v1.c24.s6.27

Lakṣmīdhara lauds the merit of constructing a special type of reservoir called dvārībandha. It was made by damming a mountain spring and thus forming a high level reservoir useful for irrigation (Kṛt. pp. 292. 286 and Introd. pp. 116-17).

v1.c24.s6.28

Suyya, a minister under King Avantivarman of Kashmir, accomplished a remarkable work of engineering skill. He dammed the river Vitasta (Jhelum). He built stone embankments for seven yojanas and temporary stone dams at all threatened points and succeeded in shifting the junction of the river with the Sindhu to its existing position (Rāj, V. 84-121).

v1.c24.s6.29

The Vedic literature implies a clear knowledge of wells and water being drawn from them (RV. I.116.9). RV. X.101.5, as translated by Griffith, reads: “Arrange the buckets in their place; securely fasten on the straps. We will pour forth the well that hath a copious stream, fair-flowing well that never fails.” Verse 7 of this hymn (*RV.* X. 101.7) has been translated by S.A. Dange (p.11) more effectively than by Griffith: “Sprinkle out this well, that is drinkable for men, which is endowed with the wooden containers fitted with stone wheel and wherein the reservoir is provided with a top guard.” Dange, however, envisages an ordinary well and does not suggest that the buckets were joined to form a chain. Ghurye (p. 226) sees in the passage a reference to the modern Rahat, but does not use the name. This, in any case, could have been noria and not the Persian wheel, because of the absence of a reference to the operation of the device by the bulls and to the gearing system (Gopal 1980, pp. 153-154). The significance of another passage (*RV.* X. 102.11) has been not been precisely analysed. The term *kucakra* occurring in it is taken by Dange (p.11) to refer to a water-wheel, a device of circular jars. It is not unlikely that it signifies a wheel with earthen jars attached to it (Gopal, 1980, pp. 154-159).

v1.c24.s6.30

Most of the P.G. Ware sites are situated on the river banks. The location gave them the advantage of inundation of the fields and also provided the facility of overflow irrigation through shallow canals. At the P.G. Ware site of Jakhera (Etah, U.P) the exposed moat like water channel (Sahi, II, p. 104) was possibly used for irrigation also. In the *Ṛgveda* (V. 4.1,7) Ṛbhus are the craftsmen who led forth the rivers. Some passages suggest embankments obstructing rivers. Indra broke them to set the rivers free (*RV.* III. 3.7).

v1.c24.s6.31

Rivers were an important source of water for irrigating the fields. In a *Jātaka* story (V. 291) the Śākyas and the Koliyas are represented as being very keen in asserting their rights to use the waters of the Rohiṇī for irrigating their fields.

v1.c24.s6.32

Irrigational activity was promoted by the establishment of larger kingdoms and the increase in the sphere of state activities. The concept of state’s responsibility for irrigational work issued directly from the theory connecting failure of rains and the occurrence of natural calamities with some grave lapse on the part of the king and requiring him to provide help and redressal against calamities.

v1.c24.s6.33

The *Arthaśāstra* (II. 1.20-22, 39) requires the king to construct reservoir filled with water either perennial or drawn from some other sources. He is to provide sites, roads, timber and other necessary materials to people undertaking the cooperative construction of irrigational and other works. He is advised to maintain those constructed earlier and to provide for new ones. The actual implementation of this advice is confirmed by the fact that even in the outlying province of Gujarat Puṣyagupta, the governor of Candragupta, the Maurya king, constructed a reservoir 240 cubits long, just as many broad and 72 cubits deep, and later on it was restored and adorned with conduits. We have epigraphic evidence to trace the history of this tank from the times of Candragupta Maurya and Aśoka to Rudradāman, the Śaka Kṣatrapa, and Skandagupta and to note how it was carefully maintained on account of its usefulness.

v1.c24.s6.34

The testimony of Megasthenes about the control of river irrigation and the sluices is corroborated by the *Arthaśāstra* (III. 9.27-31, 38). It provides for ensuring that the water of the tank is let out only at its sluice-gate and not at any other place, the flow of water from the sluice-gate is not obstructed, and the natural flow of water from a higher to a lower tank is not stopped.

v1.c24.s6.35

The lexicons list special terms for different types of reservoirs (Abhi. R. 675-77: Vaij, p. 154). The Aparājitapṛccha (pp. 18-85) gives an elaborate classification of reservoirs: ten categories of wells, two of small wells (kūpikās), four of vāpīs, four of kuṇḍas and eighteen of tanks. The Mānasollāsa (II, p. 137, verse 1613) names a miśra tank as one which every year receives fresh water channels and is bound up with embankments.

v1.c24.s6.36

The construction of tanks is evidenced by archaeological excavation. Among the earliest belonging to the historical period are those from Rupar and Ujjain (IAR, 1954-55, p.9 ; 1957-58, p.36) which are dated roughly to the phase 600-200 B.C. They were large tanks. At Rupar the tank embankment is made of a brick wall and a brick-built inlet provides for the coming of rain-water.

v1.c24.s6.37

The tank at Ujjain is made of burnt bricks. The walls of a tank of Dharmarajika at Taxila are made of the rubble masonry with a plaster coating to prevent water from percolating (Marshall 1951, I, p. 247).

v1.c24.s6.38

At Sringaverpur (Allahabad) excavations have exposed a rectangular tank (IAR, 1978-79, p.58) made of bricks measuring 41.5 to 44 × 27.5 to 28.5 × 6 to 7 cm. It is assigned to the beginning of the Christian era. Its measured length is about 100m but its ends have not been determined. The sides are retained by three successive walls of baked bricks with battered back against the natural soil. The tank was fed by the water of the Ganga brought through a channel 2m wide. The mouth of the tank made of baked bricks was shaped like an arc of a circle. The bricks at the base of the channel were set in a stepped fashion so that the base and side walls were not eroded by the gushing water. The floor of the tank near the mouth was paved with unusually large bricks 64 × 12 cm.

v1.c24.s6.39

The tank was divided into two parts interconnected with a channel paved with large-sized bricks measuring 64 × 48 × 12 cm. Cascading over a flight of steps, water entered the inner and larger part of the tank. There was staircase of five steps, each having four courses of bricks, which connected the mud platform of the retaining wall with the brick platform projecting from the lowest retaining wall. The lowest retaining wall turns outwards at this point, which marks the beginning of the inner part of the tank. Another platform with 17 courses intact is seen on the other side of the channel. The inner part of the tank with south-west orientation is both bigger and deeper than the outer one. The length of this part of the tank remains to be ascertained.

v1.c24.s6.40

A number of inscriptions belonging to the early centuries of the Christian era refer to the pious work of digging tanks. Excavations also support it. In many cases this activity is connected with foreign people entering India from the north-west. Possibly the experience of irrigational work in Central Asia contributed to the promotion of tank-digging. Excavations show that in this period the practice of offering votive tanks had become popular. The Smṛti literature shows that the digging of wells and tanks was included in the list of works of high religious merit designated as iṣṭāpūrta.

v1.c24.s6.41

It was mostly in South India that huge tanks of great engineering skill were constructed. The Pallavas carried out a lavish programme of irrigational work throughout their kingdom. The technical skill attained in the construction of tanks at that time is clear from the fact that even now they serve the purpose for which they were made. The Tenneri tank near Kanchipuram, the Mahendrataṭāka in Mahendravadi in North Arcot, the Parameśvaraṭāka, the Vīrameghataṭāka in Uttaramerur, the Kaveripakkam tank and the tank at Mamandur are associated with Pallava kings: some of them are mentioned in the ancient epigraphs (ANRASI, 1903-4, pp. 203 - 4; IC, XII, p. 74; Gopal, 1989, pp. 283-286).

v1.c24.s6.42

The mathematical texts show the actual relevance of the functioning of the reservoirs. They refer to the regulation of fountains (nirjharas) for filling a vāpī, the time taken to fill it when a given number of channels (pranalis) are operated, and for water to flow out of it through a device (yantra) (Li. p.39, no.95; GSS. VII. 32-33; IV. 28-39).

v1.c24.s6.43

The Kāśyapīyakṛṣisūkti is the only known Sanskrit text which deals with practical details about irrigational facilities for fields. It emphasises the need for the construction of a water reservoir near a village or town. It is to be constructed near a hill or on a table-land with a big lake and, in the case of a plain on a firm-land, near a perennial spring. It is to be provided with a causeway and channels and sluices for filling and emptying it. It may be fed by a mountain rivulet, a big lake, a forest brook or a big river and should be provided with a suitable contrivance for the distribution of water. It is to be provided with a very strong and big culvert and several small channels branching out of it for the easy outflow of water.

v1.c24.s6.44

In the Rāmāyaṇa (II.80.1, 11-12) we find a reference to the yantrakas who made provisions for water. When the road from Ayodhyā to Śṛṅgaverapura was being constructed, they dug wells and tanks where there was scarcity of water. At places they built small dikes which later gained the shape of tanks. A very interesting confirmation of it is available from the excavations at Sringaverapur (near Allahabad) where a tank has been exposed.

v1.c24.s6.45

The Pali text Dhammapāda (verses 80,145) refers to nettikas who forced water to go where it would not go by itself. They were specialised builders of canals and aqueducts.

v1.c24.s6.46

The term odyamtrika occurring in the Nasik Cave inscription (Luders no.1, 1,137) is translated as hydraulic engineers. But the precise nature of the work done by them and the device they used are not known.

v1.c24.s6.47

Patañjali (Keilhorn, I. 11) vividly describes a well-digger, who, in excavating, is covered with dust particles, but the hardship suffered is more than compensated when he uses the water, because the resulting prosperity (abhyudaya) leads to welfare (yoga) of many type. But, here we do not get any details about the technological aspect of digging a well.

v1.c24.s6.48

The Kāśyapīyakṛṣisūkti recommends irrigation from ponds, tanks and, more particularly, from wells. After ascertaining the presence of water by the examination of the soil by an expert of the rules of water-divining and by observing the strata of the earth from the direction of the roots of trees a well or a pond is to be dug. At first a big pit is to be dug, oblong, square or circular in shape. Digging is to be continued till water appears. Then, for making the foundation strong the base is built in latticed masonry of burnt bricks and sometimes in stone-slabs. Bricks masonry is to be gradually raised over the base. The walls are to be made of bricks laid in lime mortar. Steps are also to be provided wherever necessary. The structure above ground is to be made of stone-slabs in which an opening is to be kept towards the east or west according to the situation of site.

v1.c24.s6.49

Archaeology reveals wells in different parts of India which belong to various periods. A comparative study of their technical details, shape, measurement including depth, lining of the walls, shape and size of bricks, and the device for drawing water and their flow to the fields, will throw welcome light on this aspect of irrigational technology. To exemplify we may refer to the well at Kumrahar datable to c. A.D. 300-450 (Altekar and Misra, p.31). It is brick-paved and has a diameter of 2.25 ft. There are 104 courses going to the subsoil water. For sealing it a special kind of bricks were used which were 10 inches long and 2 inches thick, the breadth being 7 inches at the inner edge and 9 inches at the outer one. The wells of the Gupta period exposed by the excavations at Bangarh (ANRASI, 1903-4, pp 4-9) are made of bricks which are broader on one side. The big well excavated at Rajghat has walls measuring 2.7 metres, the inner diameter being 2.35 metres. The inner surface is plastered with lime water (IAR, 1965-66, pp 54-55; 1966-67, p. 44).

v1.c24.s6.50

Large scale canal work in India is evidenced only from the early Muslim times. The absence of a regular term for a canal in Sanskrit would create the impression that canals were not much in vogue in ancient India (Laping 1978, pp. 4-5). The term praṇālī occurs in the Hathigumpha inscription of Kharavela. This is the earliest epigraphic reference to an aqueduct dug out (oghātita) by king Nanda (possibly Mahāpadma) three hundred years earlier. But praṇālī may refer to a narrow water course and not to a large canal (Sircar 1966, p. 256). In the Arthaśāstra (III. 8.21,23) also the term implies a channel, a drain, a small water-course, a trench for sanitation or rain-water.

v1.c24.s6.51

But, the presence of canals is inferred from passages referring to water reservoirs which could be maintained only through a continuous supply from a perennial source56. A more direct reference can be seen in khātapravṛttim (AŚ.III.9.35), one of the three types of irrigation works, which means dug out to channelise. Likewise, the irrigation work (setu) of the type made by digging (khanya), as mentioned by the Nāradasmṛti were for diverting water (toyapravartana) and imply canals. A Ṛgvedic passage (III.33.4) may imply canals carrying water of rivers. Parallel to it is the Agnipurāṇa (272.4) reference to arms of rivers made to run into tanks.

v1.c24.s6.52

But, archaeological testimony does not leave any doubt about the actual digging of canals in ancient India. The excavations at Kumrahar (Bihar) have revealed a canal assignable to the Mauryan period. It measures 45 ft. x0 ft. deep. It possibly ran between the Son and the Ganga (IAR. 1954-55, p.19). Megasthenes testifies that under Candragupta Maurya canals were a regular part of irrigational activity. It was on such a large scale that it required the appointment of a special class of officers who “superintend the rivers, measure the land, as is done in Egypt, and inspect the closed canals from which the water is distributed into the conduits” (McCrindle, Frag, XXXIV; Majumdar, p. 268).

v1.c24.s6.53

The excavations at Besnagar (Vidisa, M.P.) have revealed an old irrigation canal in use by 165 B.C. which could have been Mauryan or pre-Mauryan in origin. Its surviving walls measure 7' x 195'4" with a height of 5'6". The canal possibly extended upto the river Bes (ANRASI, 1914-15, pp. 69-71). It has been suggested that it was an inundation canal and served the purpose of storing water.

v1.c24.s6.54

The pioneering work in this respect was done by Karikāla I, a Cola King, belonging to the close of the first century A.D. He erected embankments to control the river Kaveri. For utilising the water of the river Karikāla cut a number of canals so that water could reach the fields (Aiyangar, PTS; pp. 360 ff; Aiyangar SK. pp. 93 ff; Sastri, pp. 26 ff; Dikshitar, pp. 71ff).

v1.c24.s6.55

The Kāśyapīyakṛṣisūkti says that in the absence of the facility of a water-reservoir a canal is to be provided. Such a canal is of four kinds, as it is fed from a river, a lake, a reservoir, or mountain slopes (i.e. a mountain rill). The feeding source of the canal is to be on a level higher than the region to be irrigated. An embrasure is to be made in it. In constructing the canal its bed is to be laid. The breadth of the canal is to be four, five, six, seven or ten hastas. Its height should be proportionate to its bed and its depth adjusted everywhere to the volume of its flow. Places with terrene faults, which cause the water to disappear either by absorption or desiccation or are covered with saltpetre, are to be avoided. At places two or three canals are to be dug according to requirement. The canals are to be made to fall gradually into a lake or pond. Where there is no such water- reservoir, the canal is to terminate into the fields, but, in such places, only small culverts are recommended.

v1.c24.s6.56

The Arthaśāstra (II.24.18) refers to a number of devices for lifting water for irrigation. These are hastaprāvartimam, skandhaprāvartimam, srotoyantra-pravartimam and nadīsarastāṭakakū-podghatam. The precise meaning of these expressions has not been determined. Kangle translates them as meaning ‘set in motion by the hand’, ‘set in motion by shoulders’, ‘set flowing in channels by a mechanism’ and ‘lifted from rivers, lakes, tanks and wells.’ It may be pointed out that prāvartimam here does not qualify a mechanism or device set in motion. It refers to the diversion of water or its being brought from one’s own irrigation work (svasetu). Kangle takes hastaprāvartimam as referring to the carrying of water in pitchers. Laping suggests that it may refer to any one of the devices - the swing - basket, the shadoof, dhenkli or tath, the picota, the spoon irrigation lever and bucket being emptied into higher basin. Skandhaprāvartimam is interpreted by Laping as referring to the raising of water by the help of draught animals, as in the case of mot. Meyer takes srotoyantra to refer to the Persian wheel. Laping suggests that it may refer to a more primitive water wheel, one turned with the feet, but prefers to take it to denote ‘some device for the regulation of the outflow, probably a sort weir or spillway’. Kangle’s rendering, though in effect brining out the nature of the irrigational method, does not emphasise the device referred to in the original expression. It is to be rendered better as ‘diverted (regulated or brought) by the device of sluices’. In this case the water flows to the fields through channels. In the fourth expression the term udghātam reminds us of oghāṭitam in the Hathigumpha inscription. Kangle’s rendering of it as ‘lifted from’ does not indicate the nature of the device. It may refer to the digging out of channels. This does not involve any complicated or expensive device for manipulating or regulating the flow of water from the original source of reservoirs. It meant the natural flow of water through a channel from a source (river, lake, tank or well) at a higher level to a field at a lower level⁵⁷.

v1.c24.s6.57

The Cullavagga (Nalanda ed., p.212, line 1) records the names of three water-lifting devices which were permitted to the monks by the Buddha: *tulam*, *karakaṭakam* and *cakkavaṭṭakam*. In drawing water many earthen pitchers (*bhajana*) are broken, hence the Buddha permitted the monks to use three kinds of vessels (*varaka*) made of iron, wood and leather. Buddhaghoṣa (fifth century) in his commentary *Samantapāsādikā* adds information about these devices. *Tulā* is the pulley for lifting water, resembling the balance of a merchant⁵⁸. *Karakataka* is the device for drawing water by hand or by yoking bulls and with the help of long straps or ropes. *Cakkavaṭṭaka* is the *arahaṭṭaghaṭiyantra*. The leather vessel attached to a *tulā* or a *karakaṭaka* was known as *cammakhaṇḍa*.

v1.c24.s6.58

Hemacandra gives seven *Desī* terms for the contrivance *tulāyantra* for raising water from a well. We cannot be sure if all these refer to the same appliance or signified several slightly differing devices (Gopal, 1989, pp. 289-90).

v1.c24.s6.59

Scholars often use the expression Persian wheel for a water- drawing device in ancient India (Sharma, 1967, p 41; Nath 1970, p. 84), though reservations have been expressed by a few (Needham, IV(2), pp. 361-62; Basham 3rd ed. p.194). Irfan Habib explains the device as being novia and holds that the Persian wheel proper, an innovation of the Arabs, was introduced in India in the 13th-14th centuries (PIHS, 1969, PP. 152-53).

v1.c24.s6.60

A panel from Mandor (near Jodhpur, Rajasthan), assigned to the eleventh century, depicts a Persian wheel (Gopal, 1980, pp. 114-68). It has a wheel with a chain of terracotta buckets. The wheel is presented in profile and the gearing mechanism would be on the other side. Hence it need not be a noria.

v1.c24.s6.61

A wheel device for drawing water from well is evidenced in the *Ṛgveda*, but its nature cannot be determined precisely. We have to allow margin for regional variations and for improvement and developments, even at a very slow pace. We find several terms, which, in some cases, were not used in their strict sense. *Araghaṭṭa* was the term for the full device. *Ghaṭiyantra* referred to the main part consisting of the wheel and the pitchers. The rotation of the cycle of pitchers is often referred to metaphorically in a number of texts in a philosophical context of the cycle of birth and death (*Kuv.* pp. 227, 277; *Upa. Pr.* pp.16, 418, 723; *Chana. U* comm. II. 15.5; *Br. Su.* II.2.19). The texts mention only those parts of the device which are relevant to their context, hence the negative argument of the absence of reference cannot be carried too far. The functioning of the device is also mentioned in the *Pañcatantra* stories (*Tan*, pp. 142-43; *Pañc.* pp. 231-44), writings of Bāṇa (*Hṣa.* II, p 42; *Kād.* pp. 85, 322), *Mṛcchakaṭikā* (X.59) and *Gathasaptasati* (V. 90). The earliest epigraphic reference to the device is traced in the Mandasore inscription of Yasodharman (A.D. 352). The *Upamitibhavaprapañcakathā* of Siddharṣi (A.D. 906) gives a detailed account of the various parts of the device and its functioning.

v1.c24.s6.62

The text refers to sluices (*nirvahani*), channels, fields, seeds and a sower and adds that sown seeds, irrigated (*sikta*) by the *araghaṭṭa*, fructify into crop. It speaks of reservoirs (*pāṇantika*) attached to the well for the purposes of irrigation (secanārtham). The device is associated with a well so deep that its bottom in not visible. The text makes a clear distinction between the araghaṭṭa, the kūpa and the ghaṭīyantra. The araghaṭṭa is described as fitted with all the materials or parts. The text refers to four cultivators (karṣaka) also called sarathins (drivers of bullocks). The sixteen bullocks are speedy and agitated with vigour. The text speaks of two big gourds (tumbadvaya) only, but the number is dictated by the requirements of Jain philosophy and implies many pitchers. The term araka (spoke of the wheel) and the reference to the device revolving constantly (satatabhrama) indicate that the device is not ordinary pura or purahaṭa.

v1.c24.s6.63

A critical analysis shows that the device in ancient sources was similar to noria in some cases, but in others it was nearer to the Persian wheel. The Upamitibhavaprapañcakathā the chain of pots is mentioned as mounted on the wheel and as quite separate from it; the pots are not attached to the rim of the wheel. There is no clear reference to a gear system in the device, but neither is there a reference to the rotation of the wheel through manual labour. The text implies that there were many other parts of the device (sarvasāmagrīsaṃyukta), besides those specifically mentioned. The references to the device revolving continuously and through its own action also suit a Persian wheel.

v1.c24.s6.64

According to D. Sharma, the Persian wheel was known before the beginning of the Christian era. R. Nath holds that it was used as early as the age of the Imperial Guptas. Metrodorus, a Hellenised Persian, who came to India in the fourth century, found that water wheels were regarded in India as a new invention (Needham, IV(2), p, 369). In the absence of specific details we cannot identify the device referred to by Metrodorus. The combined testimony of the Pañcatantra, Gāthāsaptaśatī, Mṛcchakaṭikā, Samantapāsādikā and the Mandasore inscription would imply that the device had been in use before the times of Hāla and Viṣṇuśarman. The reference in Cullavagga suggests a date several centuries before the Christian era.

v1.c24.s6.65

The use of the term cakkavaṭṭakam in the Cullavagga and the first occurrence of the name arahaṭṭaghaṭīyantra in the Samantapāsādikā suggest that a specific name for the device took centuries to establish. For the introduction of the araghaṭṭa device a date in the 2nd-3rd century may not be much beyond the mark. Amarasimha, the lexicographer, does not record technical terms associated with the device, suggesting that they had not come into regular usage. A critical analysis of these terms in the lexicons (Gopal, Aspects, pp 142-49) later than the Amarakośa shows how these terms gradually came into general acceptance.

v1.c24.s6.66

Epigraphic references suggest that the araghaṭṭas came into a larger use in the early medieval period (Gopal, Econ. p.290). The device involved expenses which were not within the means of any ordinary cultivator. The Upamitibhavaprapañcakathā mentions the landlord, his cultivators, supervisors and female labourers in connection with the device. The complicated device required specialists to operate them. The Tantrākhyāyikā (pp. 142-43) has a story relating to a person who operated an araghaṭṭa. Epigraphic evidence also confirms it (Akh. p. 146, v.12).

v1.c24.s6.67

The Samarāṅgaṇasūtradhāra (I,pp. 178-79, verses 109-14) describes four water-machines (vāriyantra) to bring water down (pāta), to raise it first and then to bring it down (ucchrayāsamapāta), to bring it down and then to raise it (pātasamocchrāya), and to raise it (ucchrāya). But all this was for the recreation of the rich class. It is highly improbable that they were commonly used for irrigation.

v1.c24.s6.68

The Kāśyapīyakṛṣisūkti (verses 167-69) gives an account of the devices for lifting water from a well and for using it for irrigating the fields. For raising water a place for fixing the machine is to be made on stone-slabs on the brink of the well, and for the outflow of water a small conduit is to be made on a hard surface near the edge of the well. Of the different kinds of machines for raising water the one drawn by bullocks yoked with strong chains is the best, that by the elephant with its trunk is the mediocre, and the one by human labour is the inferiormost. Water on lower levels in steadily raised to the surface of the wells etc. by the revolution of the water-raising machine. Water is then carried to the fields by means of small channels.

v1.c24.s6.69

Irrigation of fields through channels carrying water is evidenced by the Vedic literature. Khanitrima (vide Macdonell and Keith) in it refers to a water channel for irrigation. There are references to channels for irrigating fields (RV. X. 19.4), to the digging of channels for a forward course (ibid, X. 75), and to hymns sung when cutting a channel (AV. III. 13)59. Kulyā is the term for a channel for irrigation (Amara, I.10.34). The term occurs in the Vedic literature. Following Sāyaṇa modern scholars have taken it to refer to an artificial stream or water-course falling into a reservoir (vide Macdonell and Keith, I, p. 214). A stock example in Vedānta (Vedāntaparibhāṣā, p.23) speaks of the water of a reservoir issuing from sluices and entering the fields flowing through a kulyā. A passage in the Mahāvagga (VIII.12.1-2), as explained by Buddhaghoṣa, refers to fields in Magadha being irrigated by rectangular and curvilinear channels and ridges. The legend of Āruni, the pupil of Dhaumya, lying down to stop a breach in the water-course of a certain field (Mbh.I.³) indicates the nature of the irrigation channels.

v1.c24.s6.70

³ Frakfort reports a ploughed field in the excavations of the Harappan site of Shortughai (Eastern Bacteria) (see Lal and Gupta, p. 303).

v1.c24.s6.71

The system of using water for irrigation through channels is indicated in the Arthaśāstra (III.9.27) which mentions the flow of water from the reservoir or source (ādhāra), through the outflow or a channel (parivāha), to the field being watered (kedāra).

Linked texts (50)