Irrigation
Irrigation works, the object of which is the artificial application of water to the land for purposes of agriculture as a means of supplementing the natural rainfall, although almost unknown in Northern Europe, have existed in India, as in most other Eastern countries, from time immemorial. This is a natural result of the conditions of climate. India contains large tracts, such as the deserts of Sind and the South-western Punjab, which are practically rainless, and in which cultivation without irrigation is impossible. There are others, such as the Deccan plateau, where cultivation is exceedingly precarious, owing to the great irregularity of the rainfall and the long intervals during which crops may be exposed to the fierce heat of the sun, and to dry and scorching winds. In such regions there may be an almost complete failure of crops in a year of short or badly distributed rainfall, and a succession of two or more unfavourable seasons might result in famine. In other more favoured tracts, such as the Himālayan submontane districts, a serious failure of the rainfall is seldom or never known, yet here also irrigation works, which are usually of a simple kind such as can be constructed by the people themselves, are of the greatest value in improving the out-turn of the crops. Lastly, there are districts which ordinarily receive so copious a rainfall that rice is almost the only crop grown; but for this water is required at certain critical periods when there may be a break in the rains, and a full harvest can only be secured by irrigation1. In general it may be said that the security of the harvest depends on the existence of some form of irrigation in almost all parts of India, except in tracts such as Eastern Bengal, Assam, or Lower Burma, where the average rainfall does not fall short of 70 inches per annum. As might be expected under these conditions, the simpler forms of irrigation works have been limited in number and capacity only by the resources and engineering skill of the people, by the insecurity of tenure, and by the interruptions caused by invasion or internal dissension. Some of these works were constructed by the former rulers of the country; but it was not until after the establishment of British rule that the larger and more important works, many of which irrigate annually several hundred thousands of acres, were undertaken. As an instance of the magnitude of some of these later works may be mentioned the Chenāb Canal, which irrigates nearly 2,000,000 acres, or about two-fifths of the whole cultivable area of Egypt, with an ordinary discharge of 11,000 cubic feet per second, or about six times that of the Thames at Teddington. ‘No similar works in other countries approach in magnitude the irrigation works of India, and no public works of nobler utility have ever been undertaken in the world1.’
Irrigation works may be conveniently divided into three great types : namely, lift, storage, and river works, which are represented by wells, tanks or reservoirs, and canals. In lift-irrigation the water is raised from a lower level to that which will command the area to be irrigated, the raising being effected either by manual labour or by animal or mechanical power. The source of supply is usually the subsoil water into which wells have been sunk, but lift appliances are often erected on the banks of rivers or pools from which water is raised to the lands to be irrigated. Storage works are reservoirs, formed by the construction of dams across drainage lines, for the purpose of storing the supply which passes down after every heavy fall of rain for subsequent use during long breaks in the rains or in seasons of drought. River works consist essentially of canals, drawing their supplies from rivers which are in continuous flow during the whole or the greater portion of the year. In most cases they include a weir, which is constructed across the bed of the river immediately below the off-take of the canal, for the purpose of holding up water to the full supply level of the canal during low stages in the river. The canals are usually graded with a much lower bed-fall than that of the river from which they take off, and, by a suitable alignment, can often be carried out of the river valley on to high land, or even on to the watershed separating the catchment of the river from that of the drainage line lying next to it. The great canals in Northern India, and in the Orissa and Madras deltas, command nearly the whole of the lands lying between the main rivers in the former and the branches or arms of the deltas in the latter. In the Peninsula proper, conditions are less favourable, and the formation of the country prevents canals being taken far out of the river valleys. It must be remembered, too, that some large rivers, such as the Narbadā and the Tāpti, are throughout useless for irrigation, by reason of their deep and often rocky channels.
Although these three types are in themselves as distinct as the sources of supply on which they depend, they sometimes overlap. Lifting appliances are often required to raise water from canals to ground on a higher level, and the cultivation under some tanks and canals can only be matured by the use of supplemental wells. Large storage works are incomplete without a network of distributing channels; and canals taking off from rivers in which the supply is uncertain and intermittent require storage works which may be constructed either at some distance up-stream from the head-works, in the valley of the main river or of a tributary, or as 'tail tanks' at suitable points on the line of the canal itself. But although in particular cases all three classes may be found in combination, each is so distinct from the others that it deserves separate consideration.
An important point in connexion with pure lift-irrigation works is that they have been constructed by private enterprise on the part of the owners or cultivators of the soil. The works are all of a petty character, and irrigate an area which may be as low as one acre and seldom exceeds twenty. Such works cannot conveniently be owned or maintained by the state, which can do no more than encourage their construction by agricultural loans and by liberal rules in the matter of land-revenue assessment. But although these works are small and have cost the state nothing, their importance in the aggregate is very great. Taking the average area annually irrigated from works of all classes in British India at 44,000,000 acres, it is found that nearly 13,000,000 acres, or about 30 per cent., are irrigated from wells. Well-cultivation is moreover of the highest order; and it may be safely estimated that the produce annually raised from pure lift-irrigation is, in value, considerably more than a third of the out-turn from irrigation works of all kinds. The most favourable conditions for well-irrigation are found in the alluvial plains of Northern India, the subsoil of which contains an inexhaustible supply of water. Out of the 13,000,000 acres thus irrigated more than 9,500,000 are in the United Provinces and the Punjab, and about 2,750,000 in the Madras and Bombay Presidencies. In other Provinces well-cultivation is inconsiderable, being confined to garden or other special crops.
When the spring level is near the natural surface and the soil is fairly stiff, a well may consist of nothing more than a hole in the ground over which the appropriate lifting appliances are erected. These wells, which usually fall in or are abandoned after one or two years' working, are called 'temporary' wells, to distinguish them from 'permanent' or masonry-lined wells, which, if solidly built, may last for an indefinite period. In some parts there is an intermediate type, in which the well is lined with a wooden cylinder or with brushwood or wicker-work. In the United Provinces less than one-fourth of the total number of wells are permanent; and both here and in the Punjab the number of temporary wells fluctuates greatly from year to year according to the season, being greatest in years of drought. The cost of a well varies from a few rupees for a temporary, to over Rs. 1,000 for a deep and permanent structure, but the average cost of a permanent well is between Rs. 300 and Rs. 600 according to size and depth. In the Punjab, where the wells are mainly permanent, the average area irrigated per well is 12 acres; but in the United Provinces, where so large a proportion of the wells are temporary, irrigating only 1 or 2 acres, the average area per well is not more than 4 acres. In Southern India the average area is between 2 and 3 acres; but here well-irrigation is, as a rule, applied only to the most valuable class of crops, which require frequent waterings.
The wells now in use throughout India, and the lifting appliances employed, do not differ materially from those in use in former times. The water is almost invariably raised either by manual labour or by bullocks. For the former the favourite contrivance is the well-known picottah, in which the water is drawn up by means of a bucket suspended by a rope to the end of a long pole, which is pivoted near the other end to a post so as to revolve through a vertical plane. The shorter end is weighted, so as nearly to counterbalance the weight of the longer arm and the full bucket. The bucket is lowered by the man in charge, who draws the long end of the lever by a pull at the rope, or in some cases, as in Southern India, by running up and down the pole. This contrivance is seldom used for lifts of more than 15 feet, and for very small lifts (of 4 or 5 feet) other simple appliances are used. For lifts of over 15 feet bullock-power is almost invariably employed, and is often used even for lower lifts. The most common contrivance, which is to be found all over India except in some parts of the Punjab, is the mot, which consists of a large leather bag, holding from 30 to 40 gallons of water, fastened to one end of a rope which passes over a pulley overhanging the well. When the bag has been lowered the other end of the rope is attached to a pair of bullocks, who then walk down a ramp of a length approximately equal to the depth of the well. When the bullocks arrive at the end of the ramp the bag has been drawn up to the top of the well, and its contents are emptied into a trough or sump, generally by a man who stands by, but sometimes by a self-acting mechanical arrangement. The rope is then detached from the bullocks, the bag is lowered, and the bullocks return without a load to the top of the ramp, when the operation is repeated. In some parts the rope is not detached and the bullocks walk backwards up the ramp, as the empty bag descends. The alternative to the mot is the Persian wheel, which is used extensively in the Punjab. This consists of an endless chain of waterpots passing over a vertical wheel erected over the top of the well, and rotated by means of rough wooden gearing which is worked by a pair of bullocks walking round a circular track. When bullock-power is used water is sometimes lifted for purposes of irrigation as much as 60 feet, but ordinarily the lift seldom exceeds 40. The use of any form of purely mechanical motive power is practically unknown in India. A few isolated attempts have been recently made by private owners to use oil engines, and a project is under consideration to irrigate about 50,000 acres from the Kistna river in Madras by similar means. But even if these should prove successful the practice is not likely to be extensively followed by small cultivators.
Wells play such an important part in the agricultural system of India, and are of such great protective value, that the extension of this form of irrigation must be regarded as highly desirable. There are, however, obvious limits to such extension, even in tracts where conditions of soil, climate, and spring level are favourable. The first cost of a well is only one of the factors to be considered, for its subsequent working requires a considerable amount of capital and very careful husbandry. The extension of well-cultivation must, therefore, always be gradual, and although it may be effectively encouraged it cannot be forced. The encouragement hitherto afforded by the state for the construction of wells and other works of agricultural improvement has been of two kinds—a system of agricultural advances known as takāvi, and a permanent or temporary exemption of the lands depending on the well from any enhancement of land revenue as a consequence of the improvements. Takāvi advances are made freely to approved applicants, the general rate of interest being 6¼ per cent., except in Madras and Bombay, where it is only 5 per cent. Recovery is made by instalments extending over periods varying from seven to thirty years. During the ten years ending 1900–1, as much as 625 lakhs was thus advanced by Government, 348 lakhs as agricultural loans, and 277 lakhs for the specific purpose of land improvement. In the Madras and Bombay Presidencies, ryots who construct wells or other works of agricultural improvement are exempted from any future enhancement of land-revenue assessment on the ground of the increase in the agricultural value of their land due to the improvements. In other Provinces the exemption lasts for specific periods, the term being generally long enough for the owner to recoup the capital outlay incurred in the construction of the work.
It is for many reasons very difficult to compare the area irrigated from wells at the present day (1904) with that irrigated as recently as twenty-five years ago. The Irrigation Commission of 1901–3 reported that in the four Provinces in which well-irrigation is most extensively practised—namely, the United Provinces, the Punjab, Bombay, and Madras—the increase in the number of permanent wells during the decade ending 1900 amounted to 11, 25, 33, and 44 per cent. respectively; but it seems certain that the increase in the total area irrigated from wells, in the Northern Provinces at any rate, has not been in any way proportional to these figures. This may be due partly to the imperfection of the earlier statistics, partly to the substitution of permanent for temporary wells, and partly to a reduction in the areas irrigated from old wells in tracts where canal-irrigation has been introduced. The Famine Commission of 1880 estimated the area irrigated from wells in 1876–7 at 5,794,000 acres in the United Provinces, and 3,176,000 acres in the Punjab. The Irrigation Commission of 1901–3 found that during the years 1900–2 the area irrigated from wells in the former Province varied from 4,544,000 to 6,122,000 acres, and in the latter from 3,018,000 to 4,155,000 acres. The Famine Commission of 1880 gave no details of well-irrigation in Bombay; but stated that the number of private wells in ryotwāri lands in Madras was believed to be about 375,000, and, on the assumption that the average area to each well would be about 5 acres, estimated the total area served at 1,800,000 acres. This assumption was probably much too high, for the Irrigation Commission of 1901–3 reported that the number of permanent wells in 1900–1 (excluding mere supplemental wells used on tankirrigated lands) amounted to 522,300, irrigating about 1,680,000 acres, or not more than 3·2 acres to each well. In this Presidency there appears to have been a material increase not only in the number of wells, but also in the well irrigated area.
Tanks or storage works are of all sizes, ranging from the great lakes formed by the construction of high dams across the beds of large but irregularly flowing rivers, such as Lakes Fife and Whiting in the Bombay Deccan, and the Periyār lake in Travancore, in which are impounded from 4,000,000,000 to 6,500,000,000 cubic feet of water, to the small village tanks to be found throughout Southern India, many of which irrigate less than 10 acres. Any general extension of large works of this description is of course limited by the cost of construction and the loss of supply by evaporation. The idea of forming reservoirs in the beds or valleys of streams which are in flow only for a few weeks or months in the year, must have presented itself to the rulers of the country or to the occupiers of the soil at a very early stage in Indian civilization, at any rate in those parts of the country where the conditions were most favourable; and in such tracts tanks of great antiquity bear witness to the readiness of former rulers to encourage the construction of works of this kind. Some of these works, such as the Chembrambākam and Cumbum tanks in Madras, are of great size, holding from 3,000,000,000 to 4,000,000,000 cubic feet, and with water-spreads of over nine square miles. There are remains of other works of even greater size, such as the Madag tank in the Dhārwār District of Bombay, which have failed or been abandoned. The inscriptions on two large tanks in the Chingleput District of Madras, which still irrigate from 2,000 to 4,000 acres, are said to be more than 1,100 years old. In Upper Burma there are many tanks constructed centuries ago, of which that at Meiktila is the most important. The celebrated lake at Udaipur (Rājputāna), which is said to be the largest in India, is not utilized for irrigation; but of the many old tanks in Rājputāna and Central India some are still so used, although the majority have been abandoned, or no longer hold water. In the Native States of Mysore and Hyderābād, tank-irrigation has been very highly developed. The area irrigated from large tanks is, however, small in proportion to that dependent on smaller works, most of which were probably constructed originally by the owners or occupiers of the lands protected, with but little assistance from their rulers. The most notable example of these is to be found in the smaller irrigation works in Madras. It is estimated that there are no less than 40,000 of these works in the ryotwāri tracts of that Presidency, of which less than 3,500 irrigate more than 200 acres, while many serve less than 10 acres. The former are cared for by the Public Works Department ; more than 31,000 Minor works are in charge of revenue officers ; and about 5,000 are small private works. The average area irrigated from the works controlled by Government amounts to more than 3,000,000 acres, or about as much as the entire area irrigated from all the great canals and larger irrigation works in the Province. These petty works include a certain number of small canals, but are mainly tanks or storage works. Tanks abound also in the zamīndāri estates of this Presidency, in which it is estimated that about 2,500,000 acres are irrigated, of which 2,000,000 are dependent on storage works, and the balance on wells and river channels. No approach to these figures is made in other Provinces ; but, excluding well areas, nearly two-thirds of the irrigation in the Gujarāt, Deccan, and Carnatic Districts of Bombay depends upon small storage works, and there are 50,000 small private tanks in the Central Provinces, which are said to be capable of irrigating from 150,000 to 650,000 acres, according to the season, being almost the only form of irrigation in that Province, as the area under wells is there inconsiderable. Tank-irrigation is practically unknown in the Punjab and in Sind, but is to be found in some form or other in all other Provinces, including Burma. The jhīls or natural depressions scattered at intervals over the flat alluvial plains of the United Provinces, most of which are privately owned, irrigate about 2,000,000 acres.
Storage works consist eventually of four parts. The dam, which may vary in height from 15 to 160 feet, may be constructed of earth, concrete, or masonry, but earthwork is seldom relied on for greater heights than 80 feet. When concrete or masonry is used, sound rock foundations are essential, and many sites otherwise eligible must be rejected when this cannot be found. The submerged area may, when full, have a water surface of many square miles, and here again sites otherwise eligible must be rejected on account of the large areas of cultivable land which would be thrown out of cultivation or the high compensation that would have to be paid. The outlet and distributing channels, through which water is drawn off from the reservoir and carried to the lands to be irrigated, often form a very expensive item. Lastly, there is the waste-weir or escape, through which flood-waters must be passed at times of abnormally heavy rainfall in such a way as not to imperil the safety of the work. The total cost of a storage work varies enormously according to the site. The two largest works executed by Government, the Nīra Canal and the Periyār Project, have each cost about Rs. 1,300 per 1,000,000 cubic feet in a full lake. Although these and many other large storage works have been constructed by the British Government, tank-irrigation is essentially a product of native rule, and almost the whole of the tankirrigated area is from old native works, the maintenance of which is now undertaken in whole or in part by the state. These works are mainly confined to the crystalline and sandstone formations, which occupy the whole of the Indian Peninsula between the Deccan trap area and the narrow strip of alluvium on the coast. The formation of the country in this area favours the construction of tanks, which form the principal means of irrigation. The area comprises the greater part of the Madras Presidency, nearly the whole of Mysore, half of Hyderābād, nearly two-thirds of the Central Provinces, the Baghelkhand States of Central India, the eastern half of Rājputāna, and portions of the United Provinces and of the Orissa and Chotā Nāgpur Divisions in Bengal.
It has already been stated that wells are constructed and maintained solely by private enterprise. This is also true of the smaller forms of tank, the petty works which serve a single holding or irrigate little more than 10 acres; but the largest works, which impound several thousand million cubic feet of water and submerge many square miles of country, could only have been constructed by the rulers of the country, and have always been maintained as state irrigation works. Between these extremes are the moderately sized tanks, capable of irrigating from 10 to 300 or 400 acres annually. Many of them appear to have been originally constructed by private or communal enterprise, but the state is so interested in the land revenue dependent on them that it has gradually undertaken their control or management and contributes largely to their improvement and up-keep. Thus it happens that in the ryotwāri tracts of Madras and Bombay all but the smallest tanks are controlled by Government. In the Central Provinces, on the other hand, the Government has no concern in the management of the mālguzār's tanks; and in the United Provinces, where the greater part of the irrigation recorded as from tanks is effected from the jhīls above mentioned, it maintains only the large tanks which have always been recognized as state works. In Ajmer-Merwāra and in Upper Burma most of the tanks are kept up by Government.
The average area recorded as irrigated from tanks in British India is about 8,000,000 acres, but the supply to many of the works is very precarious, and is apt to fail in years of extreme drought. This figure does not include the large areas over which rain-water is held up for some time after every fall by means of field embankments, such as are to be found under different names in Bengal, Bundelkhand, the Central Provinces, Gujarāt, and the Bombay Deccan. The agricultural value of these temporary storage works is considerable, but the areas protected by them are not usually recorded or regarded as tank-irrigation.
The third and most important class of irrigation works, viz. river works or canals, are generally of such magnitude, and affect so many different interests, that they could only have been constructed, and can only be efficiently maintained, by the state. There are, however, a certain number of petty works of this class which have been constructed and are still maintained by private enterprise, with occasional assistance in some Districts from Local funds. In the Punjab nearly 1,000,000 acres are irrigated annually from works of this class, of which more than 100,000 acres are dependent on inundation canals made by the people, but now under the control of the District Boards. Nearly 100,000 acres are irrigated from canals of the same kind belonging to private landowners, and about 800,000 acres receive irrigation from hill streams, or from jhīls from which water is diverted by means of simple works and small channels that have been constructed by communal effort. In the United Provinces about 750,000 acres are irrigated from private canals, of which the most important are the works constructed in the Bastī and Gorakhpur Districts by two English landowners, which irrigate areas of 40,000 and 15,000 acres of rice and rabi crops within their respective estates. In the permanently settled Districts of Bengal, and especially in Bihār, there are a great number of small canals, called pains, by which water is diverted from the innumerable streams which traverse the country and is conducted directly into the rice-fields or stored in small reservoirs (ahars) which have been formed by field embankments, for use during breaks in the rains. The total area protected in Bengal by means of these pains and ahars is estimated at nearly 5,000,000 acres. In Madras about 336,000 acres are irrigated annually from private canals, including the spring channels which are made in the beds of streams during the dry season, of which nearly nine-tenths are in zamīndāri estates. The Minor works under Government control in this Province also include many small river channels. In the Bombay Deccan and in Gujarāt there are many works of the same kind, most of which are under Government management. It is difficult to state with accuracy the areas irrigated throughout India from these small canals, as in the Provincial returns some are included under the head 'Canals,' others under 'Tanks' (which are often associated with them), and others, as in Bengal and the United Provinces, under 'Other sources'; but the total area protected by private canals is probably between 7,000,000 and 8,000,000 acres. Many of these works are, however, apt to fail in years of prolonged or severe drought.
We now proceed to consider the great canals which have been constructed and are maintained by the state, most of which extend over very wide areas, irrigate many thousand acres, and comprise hundreds or thousands of miles of distributary channels. These works may be divided into two classes, perennial and inundation canals. Strictly speaking, a perennial canal is one which has an assured supply of water all the year round; but this generally involves the construction of a weir or 'anicut' across the bed of the river, to hold up the low supplies and divert the whole, or as much as may be needed, into the canal. This is not, indeed, always necessary, for the head-works of the Swāt River Canal in Peshāwar District (Frontier Province) are so favourably situated that a perennial supply can be obtained without the assistance of a weir; which is also considered unnecessary for the Tribenī Canal now under construction in the Champāran District of Bengal. On the other hand, the river supply may fail in a river across which a weir has been constructed. As a general rule, however, a canal which takes off from above a properly constructed weir is assured of a fairly continuous supply during the season when water is required, and it is at any rate possible to utilize every drop of water that comes down, however low the state of the river. The term 'perennial' is therefore generally applied to a canal which can draw off a constant supply without the assistance of a weir, or which is provided with a weir which will render the supply to the canal dependent only on the quantity of water in the river and not on its surface level; while if the supply is not sufficiently continuous for the purposes required, it can generally be rendered so by the construction of storage works. Inundation canals are simple channels taking off from the banks of a river at a level which is generally considerably higher than the ordinary low-water level of the stream. Water flows into these cuts during the flood season, and they remain in flow until the river falls below the level of their beds, which is generally 2 or 3 feet higher at the end than at the beginning of the season, owing to silt deposits, which have to be cleared during the low-river season. The most important inundation canals are to be found in the Punjab and Sind, i. e. in the valley of the Indus and its tributaries. They are in flow, generally, from the beginning of May until the end of September, although the supplies fluctuate greatly during this period with the surface levels in the river. Cultivation, which is practically impossible without irrigation in the rainless tracts they serve, has adapted itself to these conditions. The average area irrigated from the inundation canals under Government amounts to about 3,500,000 acres, of which 2,500,000 are in Sind. In addition, about 1,000,000 acres are irrigated by the Bahāwalpur State Canals and by the District and private canals in the Punjab, 75 per cent. of this area being in the Native State of Bahāwalpur.
Although the present system of large irrigation works has been almost entirely constructed by the British Government, examples are not wanting of old river works undertaken by former rulers of the country. Irrigation was carried out on a large scale in the Cauvery delta, the supply being maintained by means of a weir, known as the Grand Anicut, which is said to have been constructed over 1,500 years ago. This is the work which Sir Arthur Cotton strengthened and improved in 1835–6, and the success of which encouraged him to propose the great Godāvari works which he afterwards constructed. Most of the weirs on the Tungabhadra, above which the present Tungabhadra channels draw their supplies, were constructed by the Hindu monarch, Krishna Rāya, at the beginning of the sixteenth century. In Northern India the Muhammadans appear to have made frequent attempts to utilize the water of the river Jumna. In the middle of the fourteenth century Fīroz Shāh, Tughlak, constructed a canal on the right bank of the river, 150 miles in length, for the irrigation of his domains in Hissār. This canal, which had silted up, was reopened in Akbar's reign, and a branch was made to Delhi in the reign of Shāh Jahān ; but during the decline of the Mughal dynasty both canals gradually silted up. A canal was also made on the eastern bank of the Jumna at the beginning of the eighteenth century, but appears to have likewise been abandoned, although it is said to have carried water to below Sahāranpur. These canals were subsequently cleared and reopened by the British Government, have since been re-aligned, extended, and improved, and are now first-class irrigation works known as the Western and Eastern Jumna Canals. In the Punjab the small Hasli Canal had been made by former rulers, for the purpose of carrying water from the river Rāvi to Lahore and Amritsar, on very much the same alignment as the present Bāri Doāb Canal. The simple system of irrigation by inundation canals appears to have been practised from time immemorial in Sind and the Punjab ; and many of the existing canals of this class in Multān, Muzaffargarh, and Dera Ghāzi Khān were brought to a fair state of efficiency under the more energetic Pathān and Sikh governors of these Districts, and were in operation at the time of annexation, but have since been considerably developed.
Although the earliest results attained by direct Government agency on the Cauvery, Godāvari, and Jumna works were sufficiently encouraging, the East India Company was reluctant to commit itself to any extensive scheme of state irrigation works. Funds for such works as were sanctioned were provided out of revenue, and the policy of a productive public works loan had not then been inaugurated. It was at first thought that the construction of large irrigation works, like that of railways, could be most advantageously entrusted to private enterprise. In 1857 the Court of Directors asked for proposals for new irrigation works in Madras, to be constructed by this means. The Madras Government, under the advice of Sir Arthur Cotton, proposed a great scheme, known as the Tungabhadra project, for the irrigation of large portions of the Bellary, Kurnool, Cuddapah, and Nellore Districts ; and the Madras Irrigation Company was eventually formed for its prosecution, with a Government guarantee. Operations were limited at first to the construction of that part of the general scheme known as the Kurnool-Cuddapah Canal, which was completed by the company ; but the capital cost had greatly exceeded the original estimates, the financial prospects of the project were very unfavourable, and the company was in such difficulties that the Government was compelled to take over the work in 1882.
In pursuance of the same policy the East India Irrigation and Canal Company was formed in 1860 for the purpose of carrying out the Orissa Canal Project, also proposed by Sir Arthur Cotton. Here, again, it was found that the capital cost of the works had been greatly under-estimated, while a very exaggerated estimate had been formed of the revenue which would be derivable from the work. The company was unable to raise the necessary funds for completing the undertaking, and in 1869 Government purchased the works and the company ceased to exist.
The selection of these two projects for execution by private companies was unfortunate, for experience has since shown that neither of them is ever likely to prove a directly remunerative undertaking. But it was soon recognized that, even in more favourable circumstances, the construction and control of irrigation works of this class could not conveniently be entrusted to private enterprise; and, under the viceroyalty of Lord Lawrence, a new policy was initiated under which irrigation works that gave a fair promise of proving directly remunerative were to be constructed by the state from loan funds, as 'productive' public works. From that time the construction and extension of works of this class by means of borrowed money has been actively and systematically proceeded with, the total capital outlay recorded to the end of 1902–3 on works classed as 'productive' amounting to about 38 crores of rupees. State expenditure has not, however, been confined to works which were considered likely to prove remunerative at the time they were undertaken, although loan funds have hitherto been applied only to works of this class. In accordance with the recommendation of the Famine Commission of 1880, several large works which gave little promise of proving directly remunerative have been undertaken on the ground of their value as a protection against famine. The capital cost of these undertakings, which are called 'protective' works, has been met from the Famine Insurance Grant1, the total expenditure thereon recorded to the end of 1902–3 amounting to more than 2 1⁄2 crores of rupees. These two classes of works, productive and protective together, make up what are known in irrigation accounts as Major works. They are works which have been wholly constructed or reconstructed by the British Government, and for which regular capital, revenue, and interest accounts are kept, which exhibit the financial results attained as fully as is usual in other commercial undertakings.
It should be added that Major works do not consist entirely of river works or canals. The irrigation works in the Punjab, Sind, the United Provinces, and Bengal are all of this type; but those in Madras and the Bombay Deccan include some purely storage works, and some canals, such as the Periyār Project and the Nīra Canal, with which large storage works are combined.
Considerable expenditure has also been incurred by the British Government on works which do not fall under the Major category. Reference has already been made to those old or indigenous irrigation works the maintenance of which, as Minor works, is undertaken by Government. The more important of these have, however, been not only maintained, but extended and improved, at the cost of the state. The cost of such improvements has been recorded as capital expenditure, although the funds have been provided from general revenues. There are also a few new works, generally of small size, the cost of which has been met in the same way. Both kinds of works are classified as 'Minor works for which capital accounts are kept.' The total capital expenditure on works of this class to the end of 1902–3 amounted to 329 lakhs of rupees, while the area irrigated from them is nearly 2,000,000 acres. The aggregate capital outlay to the end of 1902–3 on Major and Minor works together amounted to nearly 43 crores, and the area irrigated from them in that year was about 14,000,000 acres. The annual value of the crops raised by the works for which capital accounts are kept is estimated at 40 crores, or about 88 per cent. of the capital outlay thereon.
Details for each Major work, and, as a whole, for the Minor works for which capital accounts are kept, are given in the table on pp. 331, 332 for each Province. The figures are given for the year 1902–3, and may be regarded as fairly normal in the present stage of development of the irrigation system.
It will be seen that, although many individual works, especially those in Bengal and Bombay, show an insignificant return on the capital outlay, the Major works as a whole have proved a very profitable investment, the net revenue yielding a return of nearly 7 per cent., while the Minor works have been hardly less profitable. A brief description will now be given of the principal state irrigation works in each Province.
No Province presents greater facilities for irrigation than the Punjab with its snow-fed rivers and level plains. The
| Province. | Name of Work. | Capital outlay to end of 1902-3, in lakhs of rupees. | Area irrigated during 1902-3, in acres. | Percentage of net revenue for 1902-3 to capital outlay. |
|---|---|---|---|---|
| Punjab . . | Major Works. | |||
| Western Jumna Canal (including Patiāla section) . . | 1,84 | 595,229 | 7·7 | |
| Bāri Doāb Canal . | 1,96 | 893,862 | 12·9 | |
| Sirhind Canal (including Native States) . | 3,87 | 1,112,425 | 4·4 | |
| Upper Sutlej (including Lower Sohāg and Pāra Canal) . | 17 | 233,887 | 6·3 | |
| Chenāb Canal . | 2,75 | 1,829,169 | 21·3 | |
| Sidhnai Canal . | 13 | 60,791 | 3·9 | |
| Jhelum Canal . | 1,15 | 138,911 | – 1·0 | |
| Total Major Works 3 Minor Works | 11,87 13 | 4,864,274 235,087 | 9·7 8·0 | |
| Total . . . | 12,00 | 5,099,361 | .. | |
| North-West Frontier Province. | Swāt River Canal (Major Work) . | 42 | 173,772 | 10·6 |
| 1 Minor Work . . | 5 | .. | 23·0 | |
| Total . . . | 47 | 173,772 | .. | |
| Sind . . | Major Works. | |||
| Desert Canal . . | 26 | 187,061 | 3·7 | |
| Unar Wah Canal . | 6 | 64,350 | 11·6 | |
| Begāri Canal . . | 17 | 225,612 | 15·8 | |
| Eastern Nāra Works . | 64 | 256,938 | 6·8 | |
| Jāmrao Canal . . | 83 | 269,358 | 3·8 | |
| Dad Canal . . | 21 | 60,256 | – 1·2 | |
| Mahi Wah Project . | 10 | 8,426 | – 2·0 | |
| Total Major Works 8 Minor Works | 2,27 42 | 1,072,001 794,530 | 5·0 20·5 | |
| Total . . . | 2,69 | 1,866,531 | .. | |
| Bombay(Deccan and Gujarāt). | Major Works. | |||
| Hāthmathi Canal . | 5 | 287 | – 1·2 | |
| Lower Pānjhra River Works . . | 5 | 2,503 | 1·5 | |
| Kādva River Works . | 8 | 2,775 | 0·2 | |
| Lākh Canal . . | 4 | 1,206 | – 0·3 | |
| Muthā Canals . . | 68 | 8,101 | 2·4 | |
| Ekrūk Tank . . | 13 | 2,030 | 0·5 | |
| Kistna Canal . . | 8 | 5,963 | 3·7 | |
| Mhasvād Tank . | 21 | 4,075 | 0·02 | |
| Nīra Canal . . | 57 | 35,102 | 3·1 | |
| Shetphal Tank . | 7 | 216 | 0·02 | |
| Total Major Works 26 Minor Works | 1,96 76 | 62,258 38,237 | 1·9 0·4 | |
| Total . . . | 2,72 | 100,495 | .. | |
| Madras . . | Major Works. | |||
| Godāvari Delta System | 1,34 | 810,634 | 18.4 | |
| Kistna Delta System . | 1,44 | 627,850 | 15.8 | |
| Penner River Canals . | 63 | 163,641 | 5.0 | |
| Cauvery Delta System | 31 | 982,356 | 28.5 | |
| Srīvaikuntam Anicut System . . . | 15 | 44,526 | 6.1 | |
| Kurnool-Cuddapah Canal . . . | 2,17 | 60,154 | 0.5 | |
| Barūr Tank . . | 4 | 5,325 | 1.4 | |
| Periyār Project . . | 90 | 154,068 | 3.6 | |
| Rushikulya Project . | 49 | 92,399 | 0.6 | |
| Total Major Works | 7,47 | 2,940,953 | 8.7 | |
| 28 Minor Works | 1,05 | 584,081 | 7.0 | |
| Total . . . | 8,52 | 3,525,034 | .. | |
| Bengal . . | Major Works. | |||
| Orissa Project . . | 2,65 | 224,779 | 0.05 | |
| Midnapore Canal . | 85 | 87,464 | 0.8 | |
| Son Project . . | 2,67 | 483,567 | 3.3 | |
| Total Major Works | 6,17 | 795,810 | 1.6 | |
| 1 Minor Work | 7 | .. | -0.4 | |
| Total . . . | 6,24 | 795,810 | .. | |
| United Pro- vinces. | Major Works. | |||
| Ganges Canal . . | 3,04 | 871,862 | 10.5 | |
| Lower Ganges Canal (including Fatehpur Branch) . . | 3,98 | 849,061 | 3.8 | |
| Agra Canal . . | 99 | 236,721 | 6.2 | |
| Eastern Jumna Canal | 43 | 284,079 | 25.0 | |
| Betwā Canal . . | 45 | 64,457 | -0.5 | |
| Total Major Works | 8,89 | 2,306,180 | 7.0 | |
| 4 Minor Works | 31 | 137,375 | 2.8 | |
| Total . . . | 9,20 | 2,443,555 | .. | |
| Ajmer-Merwāra | Minor Works in 3 sub- divisions . . | 31 | 26,199 | 1.2 |
| Baluchistān . | 2 Minor Works . . | 17 | 2,939 | -0.6 |
| Burma . . | Mandalay Canal (Major Work) . | 47 | 7,223 | -0.2 |
| Grand Total Major Works . . . | 39,52 | 12,222,471 | 6.9 | |
| Grand Total Minor Works (with Capital Accounts) . . | 3,27 | 1,818,448 | 6.0 | |
| Grand Total Major and Minor Works . | 42,79 | 14,040,919 | .. |
restoration of Fīroz Shāh's Canal on the west bank of the Jumna was commenced during the administration of the Marquis of Hastings (1814-23), and the canal was gradually brought into such a state of efficiency that it was capable of irrigating over 500,000 acres in the year 1870. By reason, however, of its original faulty alignment, this canal, although a very profitable work, caused great injury to the country by water-logging, and could not be extended owing to the insecurity of the supply. In 1870 it was decided that the canal should be wholly remodelled; and the greater part of it has since been re-aligned and reconstructed, while new branches have been made, so that in the famine year 1897-8 it irrigated 764,000 acres.
The Bāri Doāb Canal, which takes off from the right bank of the Rāvi at the point where it debouches from the hills, and irrigates large portions of the Gurdāspur, Amritsar, and Lahore Districts, was undertaken almost immediately after the annexation of the Province, and was opened for irrigation in 1860-1, but it has since been greatly enlarged and extended. The Sirhind Canal, which taps the Sutlej at Rūpar, as it emerges from the Siwālik hills, was originally proposed by Sir William Baker in 1841, but the first estimate was only sanctioned in 1870, and the canal was opened by Lord Ripon in 1882. More than one-third of the cost of this canal has been contributed by the three Phulkiān States—Patiāla, Nābha, and Jīnd, which are entitled to a corresponding share of the supply. The canal irrigates large areas in these States, and also in the Farīdkot State and in the British Districts of Ludhiāna and Ferozepore. The Swāt River Canal, in the Peshāwar District of what is now the Frontier Province, is said to have been suggested by Sir Henry Lawrence soon after the annexation of the Punjab, but was first definitely proposed by Sir Henry Durand in December, 1870, and was opened for irrigation in February, 1885.
The Sidhnai, Lower Sohāg and Pāra, and Chenāb Canals were commenced in 1884. The first of these is a small canal for the irrigation of a portion of Multān District, which takes off from a weir constructed on the Rāvi at a short distance above its confluence with the Chenāb. The Lower Sohāg and Pāra is an inundation canal from the right bank of the Sutlej, and has no weir; it irrigates a portion of Montgomery District, and has recently been grouped with the Upper Sutlej inundation canal, formerly a Minor work. The Chenāb Canal was also originally designed as a small inundation canal, and opened as such in 1887, but in 1889 it was decided to convert it into a perennial canal of the first magnitude. A weir and head-works across the Chenāb. river at Khanki, about 8 miles below Wazīrābād, were completed in 1892. The canal has since been enlarged and extended so as to command the greater portion of the Rechna Doāb, the tract lying between the Chenāb and Rāvi rivers in the Gujrānwāla, Jhang, and Montgomery Districts. Great part of the area commanded by the Sidhnai, Lower Sohāg and Pāra, and Chenāb Canals was uninhabited crown waste, so that the success of the works depended on the introduction of settlers or colonists who would take up and cultivate the land. The first efforts were made on the two smaller canals with encouraging results, and the experience gained was of the utmost value in subsequently carrying out the great scheme for the colonization of the Rechna Doāb. The magnitude of this scheme and the success which has attended it are shown by the fact that up to the end of 1902–3 the area which had been sold or leased to cultivators amounted to 1,750,000 acres; while the population of the colonized tract, which in 1891 was practically non-existent, was returned at 792,000 in the Census of 1901. The colonists were drawn from the agricultural classes in the congested districts of the Province, and are now a thriving and prosperous peasantry. The total area irrigated by the canal in 1902–3 amounted to 1,829,000 acres. The Jhelum Canal was opened in October, 1901. It was commenced in November, 1898, and is at present estimated to cost 144 lakhs and to irrigate annually more than 600,000 acres. It takes off from the right bank of the Jhelum near Rasūl, the site of the battle-field of Chiliānwāla, and will irrigate a portion of Shāhpur District. Here also there is a large area of crown waste into which colonists must be introduced; but there is no reason to doubt that the canal will prove a successful and highly remunerative work.
The Minor Punjab works entered in the table include two systems of inundation canals, the Shāhpur and the Dera Ghāzi Khān Canals, for each of which a capital account has been opened1. There are, moreover, three other large systems of old inundation canals—the Lower Sutlej, the Lower Chenāb, and the Muzaffargarh Canals—which were constructed under native rulers, but are now maintained as Minor works by Government, although capital accounts are not kept for them, and which together irrigated 815,690 acres in 1902–3. The total area irrigated annually from state works in the Punjab and the Frontier Province is about 6,000,000 acres.
Although so much has been done to extend irrigation in the Punjab, the field is by no means exhausted. Apart from all extensions of existing works and many small schemes that may be adopted, two large tracts remain to be brought under irrigation. One of them is known as the Bāri Doāb, including nearly 1,600,000 acres in the Montgomery and Multān Districts, of which about half is crown waste. A detailed project for the irrigation of this tract, by means of a canal taking off from the right bank of the Sutlej immediately below the confluence of the Beās, has been prepared, but has been set aside in favour of an alternative proposal for irrigating the same tract by means of new canals from the Chenāb river, the cold-weather supply of which would be augmented by a feeder from the Jhelum. This canal will command and protect a much larger area than the Sutlej project, and will have the advantage of leaving the supply in the Sutlej available for further extensions of irrigation on the left bank, in British or Bahāwalpur territory. The second tract is the great Sind Sāgar Doāb, which comprises an area of over 5,000,000 acres lying to the south of the Salt Range and bounded by the Indus river on the west, and by the Jhelum and Chenāb on the east. The whole of this country could be commanded by a great canal taking off from the Indus at Kālābāgh ; but the work would be very costly, and a great part of the area is so covered with sand-hills that it would be unsuitable for irrigation. It is possible that this scheme, or a modification of it, may be carried out eventually, but it cannot be considered as urgent or as a very promising financial investment. Almost all the Districts of the Punjab contain stretches of inferior soils, impregnated with salt or reh ; and although it is beyond doubt that some extension, or further deterioration, of these areas has been the result of excessive water-supply, there is also evidence that such lands are not always unsuitable for irrigation, and that they have in some cases been improved or reclaimed by it. Moreover, these stretches are not so numerous or extensive as to render it unprofitable to carry canals through them for the improvement of better soils around or beyond them.
Agriculture in Sind, which is an almost rainless tract, is wholly dependent on artificial irrigation, and population is to be found only where the means of irrigation have been provided. Here, as in the Punjab, irrigation has completely altered the face of the country, and has converted barren wastes into fertile fields of wheat and rice. The works in Sind consist almost entirely of inundation canals from the Indus. The Desert, Unar Wah, and Begāri Canals all take off from the right bank, above Sukkur, and have been practically constructed by the British Government. The Eastern Nāra is a branch from the Indus, which takes off from the left bank, immediately above Sukkur, and discharges into the Rann of Cutch. This branch has been deepened at the head and brought under control, so that it provides a perennial supply to the canals which take off from it. Among these is the recently constructed Jāmrao Canal, which takes off from the Nāra at the lower boundary of the Khairpur State and commands a large area of waste land. Colonists are being introduced on the system which has been so successfully followed in the Punjab, and the prospects are promising. The canal was opened in November, 1899, and irrigated more than 269,000 acres in 1902–3. In addition to these Major works, Sind contains several large inundation canals, which are Minor works with capital accounts. Among these the largest and most important is the Fuleli Canal in Hyderābād District, which is navigable and generally in flow all the year round, and is capable of irrigating over 400,000 acres. There are also, as in the Punjab, a number of smaller canals, which are maintained by Government as Minor works, for which no capital accounts are kept, and which together irrigate about 900,000 acres.
In addition to the works mentioned, four small productive works are under construction on the left side of the Indus, and many extensions of existing canals have been proposed. It is estimated that when all these have been completed the total area annually irrigated, which may be taken as averaging about 2,700,000 acres, will be increased by about 20 per cent. The irrigated area fluctuates greatly according to the state of the river, the difference between a very favourable and an unfavourable season being as much as 800,000 acres. In spite of these fluctuations the canals never wholly fail, and are remarkably cheap and profitable works. It has been suggested that the area under irrigation might be greatly extended, and the efficiency of many of the existing canals increased, by constructing a weir across the Indus at Sukkur, and that this will be rendered necessary sooner or later by the continued abstraction of the water in the Upper Indus and its tributaries by the new Punjab canals. The cost and engineering difficulties of such a work would be very great, and no opinion can be pronounced on its feasibility until it has been more fully considered; but a great deal can be done in the meanwhile to improve and extend existing canals and the system of distribution.
It will be seen from the table already given that, although nearly every irrigation work in Sind is highly remunerative, those in the Bombay Deccan and Gujarāt are financially failures. Many of them do not pay their working expenses, and the best does not yield a return of much more than 3½ per cent. on its capital cost. The main reason for this is the great cost of storage works and the precarious and unreliable nature of the water supply. A second cause is the inconstancy of the demand in districts where in ordinary years excellent crops can be grown without the aid of irrigation, which is taken regularly only for the more valuable class of crops. For these high rates are paid; but the areas are small, and the total area irrigated by all the works shown in the table, on which a capital expenditure of 272 lakhs had been incurred up to March, 1903, does not on the average exceed 100,000 acres. The Major works and the Minor works with capital accounts are all of the same class and may be considered together. Of the whole thirty-six only two small works are situated in Gujarāt, the remainder being scattered over the Deccan Districts. The most important are the Muthā and the Nīra Canals, in connexion with which two fine Ghāt-fed storage works, known as Lake Fife (Kharakvāsla) and Lake Whiting, have been constructed, which are capable of impounding 3,833,000,000 and 5,312,000,000 cubic feet respectively. The area irrigated from the Muthā Canal may appear remarkably small as compared with its capital cost; but one of the main objects of the work is the supply of water to the cantonment and municipality of Poona and to Kirkee, and the whole cost of the waterworks has been debited against the project. The supply available for irrigation is therefore but a portion of the whole; and it is taken mainly for the highest class of sugar-cane, which is grown in the vicinity of Poona, and which requires frequent and heavy waterings and pays very high rates. The Nīra Canal has been constructed solely as a protective irrigation work, and, although hitherto not directly remunerative, it proved of the greatest value during the severe famines of 1896–7 and 1900–1.
In Bombay, as in several other Provinces, the area irrigated from petty works which are maintained by Government as Minor works for which capital accounts are not kept, exceeds that irrigated from the Major and Minor works on which capital expenditure has been incurred. There are 387 of these petty works in Gujarāt, which irrigate about 32,000 acres, of which more than two-thirds is in Surat District. The Deccan and Carnatic Districts contain 1,019 of these works, which irrigate on an average about 111,000 acres.
Irrigation works in Bombay (excluding Sind) having proved unremunerative, no new works of any importance have been sanctioned in recent years. But the Deccan Districts are so liable to severe famine that there is no part of India in which the protection which irrigation works can afford is more urgently required. The rainfall in the Western Ghāts provides an unfailing supply of water; but the cost of storing and utilizing it would be so great, and the demand for irrigation in ordinary years is so small, that there is little prospect of any work which can be constructed proving directly remunerative. Nevertheless, it is mainly by the multiplication of works of the Nīra Canal type that substantial protection from famine can be obtained, and the value of this may justify the state in incurring the permanent liabilities which the construction of such works must entail.
The most important among the nine Major works of the Madras Presidency are the three deltaic systems of the Godāvari, the Kistna, and the Cauvery, which together irrigate more than 2,400,000 acres. In these deltas the conditions are as favourable for irrigation as in the tracts served by the great perennial canals of the Punjab. It has already been stated that irrigation had been extensively developed in the Cauvery delta under native rule, but that the works were added to, strengthened, and improved by Sir Arthur Cotton in 1835–6. The irrigation is here effected not from artificial canals, but from the numerous natural channels into which the river divides after entering the delta. The works consist therefore almost entirely of weirs, escapes, and regulators, constructed for the purpose of controlling the distribution of the supply. After these works had been completed, Sir Arthur Cotton turned his attention to the Godāvari and constructed the canal system on that river between 1847 and 1850. He subsequently planned, but did not himself carry out, the Kistna project. The remaining Major works are of a wholly different character from these three deltaic systems, as they depend more or less on storage works or are carried out in more difficult country. The Kurnool-Cuddapah Canal, as already explained, is part of the great Tungabhadra scheme originally projected by Sir Arthur Cotton, the construction of which was entrusted to the Madras Irrigation Company. This work has for various reasons never proved a financial success, and even at the present time the gross revenue does not much more than cover the working expenses. The Periyār Project is one of the most interesting works in the Presidency. It consists of a large storage work in Travancore territory, on the western side of the Ghāts, formed by the construction of a concrete dam 156 feet in height across the Periyār river, the natural course of which is towards the Arabian Sea. The impounded water is, however, diverted by means of a long tunnel into one of the tributaries of the Vaigai river, which discharges into the Bay of Bengal below Madura. Many old irrigation works exist on this river, and new works have been constructed for the full utilization of the supplementary supply. The work was opened in 1896; and although the net revenue as yet only covers the charges for interest, it may be expected to improve before long, when irrigation is more fully developed. The only other Major works in Madras which have not proved remunerative are the small Barūr tank in Salem District, and the Rushikulya Project in Ganjām; but the latter was originally constructed as a protective work without any expectation that it would be directly remunerative.
The twenty-eight Minor works on which capital outlay has been incurred do not call for special remark. They consist either of anicut systems or storage works, or of a combination of these; and they may, on the whole, be regarded as directly remunerative, for the net revenue attributable to them in 1902–3 was equivalent to a return of 7 per cent. on the capital cost. But, as has already been shown, half of the total area in Madras irrigated from works under Government control is dependent on 35,000 petty works which are maintained by the state as works for which no capital accounts are kept. The revenue derived from these small works exceeds 80 lakhs a year, and about 20 lakhs is devoted annually to their maintenance or up-keep.
There is no Province in which irrigation is more appreciated or has been longer established than Madras. Many of the works are of great antiquity, and they are to be found in large numbers in almost every District. Except in the deltas, the conditions are less favourable than in the Punjab or Sind for large canals unaided by storage; but, on the other hand, the country is better adapted for the small storage works which form such an important feature in the agricultural economy of the Presidency. There is still a wide scope for the extension of the irrigated area, but few extensions are possible without storage. The supplies of even the great Kistna and Cauvery rivers are often insufficient for the irrigation at present dependent on them, or for the second crop which might be grown, and proposals have been made to form large storage works on both these rivers. From a protective point of view the most important undertaking which can be proposed is the completion of the Tungabhadra Project on the scale originally suggested by Sir Arthur Cotton. This work would protect very large areas in the Districts of Bellary, Anantapur, Kurnool, Cuddapah, and Nellore, all of which are specially liable to severe distress and famine in years of drought. But storage works of enormous capacity would be required on the Tungabhadra and elsewhere in order to obtain a reliable supply for the whole area that could be commanded, while the canal channels themselves would be very costly owing to the nature of the soil and the configuration of the country. As much of the area that would be served consists of black cotton soil, upon which in ordinary years ‘dry’ crops can be raised, the demand for water would be much less constant than in the rice-growing tracts; and such a work is not, therefore, likely to prove directly remunerative. There is no doubt, however, that an unfailing supply of water is available, and that, if it can be stored and utilized at an expense which will not be prohibitive, a very large area in these districts could be securely protected from famine. The question is now (1905) being investigated in detail. Numerous other works of a less ambitious character are possible in almost all districts, but practically all will involve the construction of storage works. Such works are likely, however, to be less costly and unremunerative than in the Bombay Deccan, and it seems probable that many schemes can be proposed which will not involve a greater permanent burden on the finances of the state than would be justified by their protective value.
State irrigation works have never hitherto proved remunerative in Bengal. The ordinary rainfall is so great that water is not in constant or regular demand, although there may be a great rush for it during long breaks in the rains, or at critical periods of the rice crop. The three Major works which have been constructed are all unremunerative. The Orissa Canals, to which reference has already been made, do not pay their working expenses, and the Son Canals in Southern Bihār, which are the most successful, do not on the average yield more than 3 per cent. on their capital cost. A small canal, the Tribenī, is now under construction in the north of Champāran District as a protective work. There is scope for extension of irrigation in Bengal, especially in the Bihār Districts, which are densely populated and the most liable to famine; but until higher rates for water can be obtained from the people, such works are never likely to be remunerative. In Eastern Bengal the rainfall is so abundant that there is no demand for irrigation.
It has already been shown how the canal constructed by Fīroz Shāh on the west bank of the Jumna was reopened in the early part of the nineteenth century, when it was within the Province of Agra. The small canal on the eastern bank was reopened and extended soon afterwards, and now irrigates portions of the Districts of Sahāranpur, Muzaffarnagar, and Meerut. The knowledge and experience gained on these works were subsequently applied by Sir Proby Cautley in the design and construction of the Ganges Canal, which was the first original irrigation work of any magnitude constructed in Northern India, and is even now second to none in the boldness of its conception, and to few in its utility and financial success. It was the pioneer of all the large canals subsequently constructed in the United Provinces and the Punjab. The canal, which was opened in 1854, takes off from the right bank of the Ganges immediately below the sacred ghāts at Hardwār, and after being carried over the Solāni river by the celebrated Solāni aqueduct, irrigates portions of nine Districts in the Jumna-Ganges Doāb. Seven Districts in the lower part of the Doāb receive irrigation from the Lower Ganges Canal, which also takes off from the right bank of the Ganges at Naraura, about 130 miles below Hardwār, and was first opened for irrigation in 1878. The Agra Canal, opened in 1874, is a smaller work which takes off from the right bank of the Jumna at Okhla, 11 miles below Delhi, and irrigates portions of the Districts of Gurgaon, Muttra, and Agra. All these canals have proved remunerative; but this cannot be said of the remaining Major work, the Betwā Canal, which takes off from the left bank of the Betwā, a tributary of the Jumna, about 12 miles north of Jhānsi, and irrigates portions of the Jālaun and Hamīrpur Districts in Bundelkhand. This canal, which was opened in 1885, has not a perennial supply, and the demand for irrigation during the period that the river is in flow is very slack, except in dry seasons, so that it does not pay its working expenses. A certain amount of storage (2,700,000,000 cubic feet) has been provided above the weir, where the river is impounded to a height of 56 feet above the lowest point of its bed ; but this is insufficient. The canal was sanctioned as a protective work and was not expected to be remunerative, but it proved of great value during the severe famine of 1896-7, when it irrigated more than 87,000 acres.
The four Minor works for which capital accounts are kept include three small systems of canals—the Dūn, the Rohilkhand, and the Bijnor, which are all profitable works ; and the Bundelkhand lakes, which irrigate 3,000 acres and just pay their working expenses. There are no other Minor works.
It must be added that the canals in the United Provinces have not always conferred unmixed benefit upon the land which they command. In many places their introduction led to a gradual but steady rise in the level of the subsoil water, and resulted eventually in the water-logging of the soil, the increase of malaria, and the further deterioration of ūsar or reh-covered tracts. To remedy these evils, and to prevent further injury, it was found necessary to incur a large outlay on the re-alignment of some of the older canals, and on the construction of a large system of drainage channels, the aggregate length of which now exceeds 3,000 miles. These measures have proved successful.
The field for the extension of canal irrigation in the United Provinces is not large. A project for a great canal, to take off from the right bank of the Sārdā river for the irrigation of the Oudh Districts, was prepared in detail more than thirty years ago, but was not sanctioned owing to the doubts expressed as to the necessity for such a work and the opposition of the owners of the lands which would be affected. The project has been several times revived and modified, but has found few advocates among landowners or revenue officers, while it is strongly opposed by many. A smaller scheme has been suggested which would protect the only portions of this tract which have in the past been liable to famine ; but there are doubts and difficulties attending this also, and it will not, in any case, be taken up at present. Additional protection from famine is most urgently required in the districts which lie to the south of the river Jumna ; but canals in this tract would have to take off from tributaries of the Jumna, the supplies in which, as in the case of the Betwā, can be relied on only for a few months in the year. The most important scheme recently sanctioned as a protective work is a canal from the river Ken, to protect a large portion of Bāndā District which suffered severely in the famine of 1896–7 ; but however valuable works of this kind may be in this tract in seasons of drought, there appears to be little hope of their proving directly remunerative.
The irrigation works in Ajmer-Merwāra consist entirely of tanks, which were either made or restored about 1850, under Colonels Hall and Dixon. There are no natural rivers on which reliance can be placed for irrigation, and the rainfall is so irregular, and often so inadequate, that wells are apt to run dry. In such tracts tanks are of the greatest value, not only on account of the irrigation directly effected from them, but as a means of maintaining the spring level in the wells in their vicinity. Tanks have been constructed in almost every available catchment basin, and every effort has been made to utilize to the utmost the scanty rainfall, so that there is little room for the extension of irrigation in this small territory. In favourable years the tanks are capable of irrigating nearly 40,000 acres, but in a dry year not more than one-third of this can be protected. In ordinary years the works yield a net revenue equivalent to about 3 per cent. on their capital cost, but in dry years working expenses are sometimes not covered. Apart from the tanks the only means of irrigation is from wells, the area under which varies from 44,000 to 108,000 acres according to the season.
Two small works have been constructed by the British Government in the Quetta-Pishīn District of Baluchistān, which irrigate between them from 4,000 to 8,000 acres, according to the season, but at present they do not quite pay their working expenses. Other small works have been proposed recently, but all permanent supplies of water are now so fully utilized that there is little scope for new state irrigation works. Private irrigation works, utilizing the waters of perennial streams and to a certain extent the flood-waters of hill torrents, are numerous; but the distinctive private irrigation works of Baluchistān are the kārez, or underground tunnels, which tap the subsoil supply of water on high ground and lead it on to the surface of lower country. The people display much skill in the construction of these works, for which the country is well adapted.
Irrigation is not required in Lower Burma, where the rainfall is so regular and abundant that rice, the staple crop of the country, can be grown without it. Up to 1901 there were no Major irrigation works in operation in Upper Burma; but two projects—the Mandalay and the Shwebo Canals—were then under construction. The former was opened in 1902–3 and irrigated upwards of 7,000 acres in that year; the latter was opened in 1906. These two works are estimated to cost 95 lakhs, and to irrigate together about 230,000 acres, and there is reason to believe that both will prove remunerative. Another scheme, known as the Mon Canals, of about the same scope as each of these works, has also been sanctioned, and is now (1905) under construction. There are about 200 Minor works, for which capital accounts are not kept, but which are now administered and maintained by the state. Most of these existed before the annexation, but many had fallen into disuse. About 50 lakhs has since been spent on restoring or repairing them, and in 1902–3 they irrigated about 400,000 acres.
Private irrigation works are also numerous in Upper Burma, especially in the Magwe District, or on the tributaries of the Irrawaddy. The area irrigated by these is estimated at 475,000 acres, of which 300,000 acres are from small canals, and the balance from tanks and other sources.
There are at present no state irrigation works in the Central Provinces, although, as already shown, certain rice-growing districts possess no less than 50,000 small private tanks, capable of protecting from 150,000 to 650,000 acres according to the season. The necessity for protective irrigation does not appear to have been recognized until recently, as the pressure of population was light. Moreover, until 1896 a complete failure of the rains was hardly known, and in many parts of the Province the cultivation is said to have suffered more frequently and severely from excessive rainfall than from drought. Under such conditions the people are never likely to pay as much for irrigation as it will cost to provide it, but the Province has suffered so severely from famine during the last five or six years that the construction of suitable state irrigation works is contemplated as a protective measure. In Berār there are practically no irrigation works of any kind, state or private, except wells. Here, too, the necessity for irrigation works has not often been felt, but the Province was seriously affected by the failure of the later rains in 1896, and severe famine occurred in 1899. Storage works may be possible in the hilly tracts in the north and south known as the Melghāt and Bālāghāt, but tanks in the latter tract would not be secure from failure in years of severe drought.
Irrigation works are not required in Assam, where the rainfall is heavy and unfailing; nor in any part of Coorg, except in a narrow strip on the eastern border, in which failure of the rains occasionally occurs and where a few petty works have been constructed.
The Irrigation Commission of 1901–3 estimated the total area ordinarily irrigated in British India from all sources as below, the total cultivated area being taken at 226,000,000 acres :—
| Source of Irrigation. | State works. | Private works. | Total. |
|---|---|---|---|
| Acres. | Acres. | Acres. | |
| Wells | ... | 12,895,000 | 12,895,000 |
| Canals | 15,644,000 | 1,235,000 | 16,879,000 |
| Tanks | 2,944,000 | 5,194,000 | 8,138,000 |
| Other sources | ... | 6,186,000 * | 6,186,000 |
| Grand Total | 18,588,000 | 25,510,000 | 44,098,000 |
It should be noted that the figures given in this table do not include the large areas directly watered by river floods, nor the many thousand acres which, though not directly irrigated in the ordinary sense, are artificially saturated by rain-water held up within field embankments.
It is difficult to determine the increase of area under private works since 1878. As regards wells, the question has been already referred to, and it may be doubted if there has been a greater increase under private works of other kinds than under wells. With regard to state works there is no difficulty, and it may be safely said that between 1878 and 1903 the area under irrigation increased from 10,500,000 to 18,500,000 acres; of this increase nearly 7,000,000 acres are due to Major works.
The financial results of state irrigation works are discussed in Volume IV, chapter vi (Finance).
The distribution of the total irrigated area among Provinces is shown in the table on the next page.
The actual area for each Province in any one year may differ considerably from that shown in the penultimate column. In a very dry year tanks are liable to fail and the area irrigated by them falls very short. On the other hand there is an unusually strong demand for water in such years, and the greatest inducement to economy in its use, so that maximum areas are often then irrigated by perennial canals or from wells in which the supply is maintained.
| Province. | Population (1901). | Average area annually cultivated. | Area ordinarily irrigated from all sources. | Percentage of irrigation on area cultivated. |
|---|---|---|---|---|
| Acres. | Acres. | |||
| Punjab and North-West Frontier Province | 22,357,000 | 28,207,000 | 10,430,000 | 37·0 |
| Sind | 3,211,000 | 3,323,000 | 2,923,000 | 88·0 |
| Bombay proper | 15,305,000 | 24,327,000 | 1,077,000 | 4·4 |
| Madras | 38,209,000 | 36,574,000 | 10,532,000 | 28·8 |
| Bengal | 74,745,000 | 63,664,000 | 6,349,000 | 10·0 |
| United Provinces | 47,692,000 | 41,086,000 | 11,055,000 | 26·9 |
| Ajmer-Merwāra | 477,000 | 388,000 | 142,000 | 36·6 |
| Baluchistān | 308,000 | (?) | 5,000 | .. |
| Upper Burma | 3,847,000 | 4,666,000 | 828,000 | 17·7 |
| Central Provinces | 9,877,000 | 16,814,000 | 700,000 | 4·2 |
| Berār | 2,754,000 | 6,820,000 | 56,000 | 0·8 |
| Coorg | 181,000 | 195,000 | 1,000 | 0·5 |
| Total | 218,963,000 | 226,064,000 | 44,098,000 | 19·5 |
A few remarks may be made about irrigation in Native States. The Irrigation Commission of 1901–3 obtained statistics from all the more important States in which irrigation is practised, excepting those in Burma and Baluchistān. These statistics covered an area of 438,000 square miles, the population in which amounts to 51,326,000. The average area cultivated is estimated at 71,076,000 acres, of which 7,763,000 acres, or 10·9 per cent., are irrigated in an ordinary year. After making some allowance for areas not included in the returns, the Commission estimated the total area annually irrigated in the Indian Empire, excluding the Native States in Burma and Baluchistān for which no particulars can be given, at 53,000,000 acres, of which 19,000,000 are irrigated from canals, 16,000,000 from wells, 10,000,000 from tanks, and 8,000,000 from other sources.
Among individual States the first place may be given to Mysore, where there are said to be about 39,000 tanks, or four to every three square miles. Almost every valley contains a chain of such tanks, the first overflowing into the second, and so on until the terminal tank is filled. The largest existing tank has a capacity of 3,118,000,000 cubic feet ; but there are only ten tanks with a capacity of over 270,000,000 cubic feet, and not more than 2,300 are capable of irrigating 80 acres or more. The average area irrigated annually by all tanks is about 540,000 acres. There are also about 1,000 miles of canals or river channels, which irrigate about 100,000 acres. The well area is about 70,000 acres. The State is now constructing two very large storage works, the Bora Kanave and Māri Kanave reservoirs, which are to have storage capacities of 2,354,000,000 and 30,000,000,000 cubic feet respectively. The latter work, which is being constructed on the Vedāvati or Hagari river, a tributary of the Tungabhadra, in Chitaldroog District, will have a masonry dam 142 feet high, and a waterspread when the lake is full of over 40 square miles. It will then have a capacity very nearly equal to that of the Nile reservoir at Assuān, but the lake will only fill during years of extraordinary rainfall, or perhaps not more than once in thirty years. In a year of average rainfall the tank will not impound more than 10,000,000,000 cubic feet, but it has been proved more economical to build the dam to the height proposed than to provide the necessary escape at a lower level.
In the Hyderābād State an area of about 773,000 acres is said to be irrigated in years of normal rainfall, but in a dry year, such as 1899–1900, this area is reduced to less than half. Most of this irrigation is from tanks, of which there are said to be 18,000, but the capacity of the largest of these does not exceed 300,000,000 cubic feet. Some of these tanks are of great antiquity, but the majority of them were until recently breached and ineffective. A great deal has, however, been done to repair them since 1893: more than 7,000 have been restored, while many projects have been prepared for the restoration of others or the construction of new works. There are also some small canals which take off from the left bank of the Tungabhadra river, and irrigate nearly 4,000 acres, and a new work of this kind has been recently sanctioned which is estimated to irrigate 10,000 acres.
In the Native States under the Madras Government, 625,000 acres, or more than 57 per cent. of the average area cultivated, are said to be irrigated. In Bombay the area irrigated in Native States, including Baroda, in years of normal rainfall is estimated at 1,147,000 acres, but this is less than 5 per cent. of the area cultivated. In the Rājputāna Agency, where the need for irrigation is very great, about 1,172,000 acres, or 18 per cent. of the area cultivated, are irrigated from wells and tanks in years of good rainfall, but the irrigation is very precarious. The States best protected are Jaipur, Bharatpur, Shāhpura, Kishangarh, Alwar, and Kotah. The Jaipur State has done a great deal during the last thirty years, under the advice of Colonel Sir Swinton Jacob, in constructing irrigation tanks and in advancing money for the construction of wells; and in Bharatpur a sum of 10 lakhs has been spent since 1897, principally in constructing embankments for impounding and distributing spill waters from the Bāngangā and other rivers, which has resulted in an increase of 50,000 acres in the irrigated area. There is scope for extension of irrigation in some of the Rājputāna States, but the rainfall is so uncertain that sites for reliable storage works are not easy to find, and the question is complicated by territorial difficulties. Rājputāna is, however, in such urgent need of protection that investigations are now (1905) being made with the object of drawing up projects for utilizing to the best advantage all available sources of water-supply. In the Central India States irrigation is not very extensively practised, and the total area irrigated from all sources does not amount to 6 per cent. of the average area cultivated; but many tracts are liable to severe drought. In the Native States of the Punjab nearly 2,000,000 acres are irrigated in ordinary years, of which about 750,000 acres represent irrigation from the Sirhind and Western Jumna Canals in the States of Patiāla, Nābha, Jīnd, Farīdkot, and Kalsia, while an equal area is irrigated in the Bahāwalpur State by means of the Bahāwalpur inundation canals which take off from the left bank of the Sutlej.
The revenue receipts from Government irrigation works are derived almost entirely from the charges made for water or for water advantages; and these charges depend throughout India not on the volume of water supplied, the amount of which may vary largely according to the character of the season and other considerations, but on the kind of crops cultivated and the areas actually or ordinarily irrigated. In former days, when land revenue was taken in kind, the state's share of the produce increased with the introduction of irrigation; and it was in anticipation of such an increase that rulers or farmers of land revenue constructed, or contributed towards the cost of constructing, irrigation works.
When, under British rule, the system of cash payments succeeded, and assessments were made with reference to average produce, lands which were assured of irrigation were naturally assessed at higher rates than similar lands which did not enjoy this advantage. The difference between this 'wet' rate and the ordinary 'dry' rate leviable on unirrigated land of the same class represented the true revenue earned by or due to the works. In Madras almost all the new works included a considerable area of old irrigation, dependent on works which were superseded by or incorporated in the new works. In the case of lands newly brought under irrigation, power was taken to impose a water-rate which might be levied on the whole area for which water had been applied (or in some cases on the areas actually irrigated) in addition to the 'dry' rate; but in course of time the two rates were consolidated at revision of settlement, the consolidated rate representing the land revenue assessed on the land entitled to irrigation, in consideration, among other things, of its water advantages. This system is still followed throughout the Madras Presidency, in Sind, on a number of old irrigation works in Bombay, and in the Burma Districts which have undergone settlement. It has many advantages where, as in Southern India, over 90 per cent. of the irrigated area is under rice cultivation and water is delivered to the same fields year after year, and when the practice is in accordance with immemorial usage. In Northern India, however, where the new perennial canals were carried into tracts where irrigation had been before unknown, and water was applied to a number of different kinds of crops, and to different fields in different seasons (for the area actually irrigated in a single year is often not more than a third or fourth of the area under command), it was found that this system was inapplicable, the demand from particular lands being less constant, and the irrigation much more widely diffused.
In the Punjab and the United Provinces, and also in Bengal (where the land is under permanent settlement), the charge for water is distinct from and independent of the land-revenue assessment, and is levied in the form of a water-rate on the occupier. The occupier's rate varies according to the crop grown, and in the first two Provinces is chargeable only in respect of fields for which water has actually been taken. In Bengal it has been found convenient to introduce a system of long water leases (five or seven years) for blocks of land comprising many holdings, and each occupier is charged according to the area of his holding within the block, whether he has actually taken water or not. The occupier's rate is subject to periodical revision, and remissions are freely given in case of failure of crops, even when this is not due to the irregularity or insufficiency of the supply. It has been found, however, that the occupier's rate does not prevent a rise in rents due to water advantages. In the Punjab and the United Provinces, where the land revenue has not been permanently settled, the Government can take a share of this increase at revisions of settlement; and under the Northern India Canal and Drainage Act, 1873 (which does not apply to Bengal), it may, during the currency of a settlement, impose, in addition to the occupier's rate, an 'owner's rate' on lands which have been brought under irrigation after the settlement was made, provided that such rate does not exceed half the increase in the rental value of the land due to irrigation. This rate, like the occupier's rate, is leviable only on lands which have actually received irrigation during the year.
On the inundation canals in the Punjab which were taken over at annexation, the land revenue was until recently fixed with reference to water advantages, in accordance with ancient custom, and no other charges were made for irrigation. But in accordance with the same custom the people were also required to carry out, or to contribute to the cost of, the annual canal clearances. On the Multān and Muzaffargarh Canals this obligation was enforced by means of the now moribund chher system, under which each owner was required to send to the canal clearances a number of labourers proportionate to the extent of his holding which had been irrigated during the previous season. On the Dera Ghāzi Khān Canals a fixed 'clearance-rate' equal to half the estimated average cost of the clearances was levied rateably on all owners. The supply in inundation canals is liable to such fluctuation that a fixed land revenue was open to many objections, and the people no longer desired the continuance of the chher system. At the last resettlements of these districts it was therefore decided to impose an occupier's rate in addition to an assessment at 'dry' rates, Government undertaking the clearance by hired or contract labour in the ordinary way. The change is being introduced gradually, and the practice of chher clearances (which has sometimes been incorrectly regarded as forced labour) is still followed on some of the canals, but will in a few years cease to exist.
Occupiers' rates are also levied on all the Major, and on several Minor, works in the Bombay Deccan. In Sind, where almost all cultivated land is irrigated, the charge for water forms part of the land-revenue assessment; but the latter varies with the method of irrigation applied (flow or lift), with the area under cultivation, and, to a certain extent, according to the crop sown. Nine-tenths of the revenue assessed on irrigated lands is regarded as irrigation revenue, for which the canals are entitled to credit.
The charges for irrigation, whether taken in the form of enhanced land revenue or of occupiers’ and owners’ rates, vary very much, depending on the kind of crop, the quantity of water ordinarily required for it and the time when it is required, the quality of the soil, the intensity or the constancy of the demand, and the value of irrigation in increasing the out-turn. In the immediate vicinity of Poona a rate of Rs. 50 per acre is paid for sugar-cane. This is a quite exceptional rate, and obtains only over a limited area. On other parts of the Muthā Canal the rate varies from Rs. 40 to Rs. 12, and on other canals in the Bombay Deccan from Rs. 25 to Rs. 10 per acre. In Madras the maximum rate for sugar-cane is Rs. 10, and in the Punjab it does not exceed Rs. 8-8. The rate charged for rice varies in Madras from Rs. 5 to Rs. 2, and in Bengal from Rs. 2-8 to R. 1-8 per acre. In both these Provinces irrigation is practically confined to rice; in the Punjab, where this crop is not extensively grown, the rate varies from Rs. 7 to Rs. 3-4 per acre. The ordinary rate in the Punjab for wheat, which is the principal crop, varies from Rs. 4-4 to Rs. 3-12, and for fodder crops from Rs. 3 to Rs. 2-8, per acre. The average rate realized on Major works from irrigation of all kinds is about Rs. 3-8 per acre, the Provincial averages being R. 1-9 in Sind and Bengal; Rs. 3-4 in the Punjab; and Rs. 4-8 in Madras, the United Provinces, and the Bombay Deccan. The working expenses per acre vary from 8 annas in Sind to Rs. 2-8 in the Bombay Deccan, the average for all Provinces being about R. 1-1. The charges for irrigation may be taken as varying from 10 to 12 per cent. of the value of the crop, except in Bengal and the Bombay Deccan, where the average is little more than 6 per cent.
During the famines that occurred in nearly every Province of British India, and in most of the Native States, between 1896 and 1901, the protective value of irrigation was very plainly shown, in spite of failures of some of the smaller storage works; and the question naturally arose whether this protection might not be extended to those tracts in which the most severe distress had occurred, and in which means of irrigation had not been provided or were inadequate. In 1901 the Government of India referred this question to a Commission which was directed to report (inter alia) on the extent to which irrigation had been provided by works constructed by the state, and on the results—productive, protective, and financial—which had been attained; to determine the scope which existed for further extensions of state irrigation works, and the probable net cost of carrying them out; and to consider the extent to which local capacities for irrigation had already been utilized by private individuals, and the manner in which it might be possible to stimulate the extension of private irrigation works. Sir Colin Scott Moncrieff was appointed President of the Commission, which completed its report (from which many of the statements and figures in this chapter have been quoted) in April, 1903.
The Commission reported that the Major irrigation works constructed by the state have been, on the whole, so profitable that the surplus revenue, after meeting all interest charges on the capital outlay, now exceeds a crore of rupees a year; but they found that the field for the construction of new works of any magnitude on which the net revenue would exceed the interest charges was very limited, being restricted to the Punjab, Sind, and parts of Madras, and that the tracts in which most of such works could be constructed were not liable to famine. They recommended that such works of this class as were feasible should be constructed as soon as possible, as they would not only be profitable investments, but would have a protective value in increasing the food supplies of the country. They recognized that irrigation works could not possibly prove directly remunerative in the areas where protective irrigation was most urgently required, such as the Deccan Districts of Bombay and Madras, the Central Provinces, and Bundelkhand; but they were of opinion that such works might be undertaken in these tracts as would tend to reduce the cost and mitigate the intensity of future famines, although they would not pay much more than their working expenses. For the protection of the Bombay Deccan they recommended the construction of storage works, like Lakes Fife and Whiting, in the Western Ghāts, where the rainfall has never failed even in the driest years. For Madras they recommended a detailed investigation of the old Tungabhadra project, or some modification thereof, and of a scheme for a storage work on the Kistna, on the understanding that it was not essential that either scheme should fulfil the conditions of a productive work, provided that it would afford, at a reasonable cost, reliable protection from famine to the districts which are now most liable to it. They also proposed that Government should undertake the construction of protective works in the rice-growing districts of the Central Provinces, and recommended the Ken Canal project in Bundelkhand and the examination of the rivers which traverse that territory for sites for storage works. The investigations which have been undertaken in the Rājputāna States were also started on the recommendation of the Commission, who proposed that they should be extended to the Central India and Kāthiāwār States. The Commission further sketched out a rough programme of new Major works to be constructed in different parts of India, which would cost not less than 44 crores of rupees, and would result in an increase of 6,500,000 acres in the irrigated area. It was estimated that the construction of these works would impose a permanent yearly burden of nearly 74 lakhs on the state, through the excess of interest charges on capital cost over the net revenue due to the works. Against this would have to be set, however, the reduction in the cost of future famines resulting from the construction of the works, which, the Commission thought, could not be estimated as likely to average less than 31 lakhs per annum. The balance of 43 lakhs would represent the net annual cost of the works to the state, or the price to be paid for the protection from famine which the works would afford, and for all other indirect advantages that might be attributed to them. The Commission also called attention to the fact that the area actually protected by private irrigation was greater than that which is dependent on all the works constructed or maintained by the state ; and considered that a great deal might be done to encourage the extension of such works by a more liberal and elastic system of Government loans, by grants-in-aid in famine tracts, and by other means. The Commission were, however, of opinion that, although much could be done by irrigation to restrict the area and to mitigate the cost and distress of famine, India could not be entirely protected from famine by means of irrigation alone. They estimated also that, if all the works which they proposed could be constructed, they would not intercept for purposes of irrigation more than 2¼ per cent. of the volume of water which now, unused, flows onward to the sea. The practical application of these proposals is under consideration ; and the annual allotment from revenue for expenditure on irrigation projects, which stood at one crore between 1900–1 and 1903–4, has now (1904–5) been raised to 1¼ crores.
The relative merits of railways and irrigation works as a means of preventing famine have often been the subject of discussion, but the claims which may be made on behalf of either are not conflicting, for the means of production and distribution must be regarded as mutually supplementary. Railways do not, like canals, increase the food-supply of the country, but they render possible the transfer of surplus supplies to parts which require them. They are the only means of ensuring a food-supply in times of famine to many tracts where irrigation is impossible or could only be supplied at a prohibitive cost. Again, the protective value of many of the most important irrigation works, such as those which have converted into unfailing granaries the unpopulated and uncultivated wastes of the Punjab and Sind, depends on the existence of railways by which the surplus produce may be readily and cheaply distributed. It may be that the construction of a particular railway might well have been held in abeyance in favour of an irrigation work, or vice versa ; but improvement in the means of production and of distribution must nevertheless proceed concurrently. The smallness in the difference of prices in all parts of the country during the famines which have occurred since 1896 is the best justification of the policy of railway extension as an alleviation of famine. Irrigation works can do much, but they cannot by themselves give as complete or universal protection as is required.
Navigation
Inland navigation and irrigation are closely connected, for each depends for its fullest development in India on the construction of canals. If the natural waterways of the country are to be utilized to the utmost for the purpose of communications, it is as necessary to connect them by navigable channels as it is to tap their waters by canals for purposes of irrigation. At first sight it may appear that the same channels might be made to serve both purposes, and that when the great cost of constructing a large irrigation canal has been undertaken, there should be no hesitation in incurring the extra expenditure required to render it navigable. Even if the revenue to be anticipated from navigation dues should be insufficient to yield an adequate return on the additional capital outlay involved, it may be admitted that unremunerative expenditure of this kind would be justified if there were a certainty that the canal would be extensively used for navigation, that a new line of communication would be established, and that a material saving would be effected in the cost of transport. Apart, however, from all questions of cost, it has been found that the exigencies of irrigation and navigation are not always compatible, and that traffic is not attracted to a navigable canal which does not pass through large cities or important trade centres, or which is not in uninterrupted connexion either with the seaboard or with the waterways which form the most convenient outlet for the produce of the tract which the canal traverses. Hence it happens that irrigation canals are not always suitable for navigation, and, on the other hand, that many canals have been constructed for purposes of navigation only and do not irrigate a single acre. It will be convenient to consider in the first place the system of irrigation canals which are also navigable.
The two most important navigable systems of irrigation works are the Godāvari and Kistna Canals in the Madras Presidency. The total length of the main canals and branches of these systems is 506 and 372 miles respectively, of which 493 and 332 miles are navigable. There were strong reasons for applying these canals to navigation. During many months in the year they carry away all, or nearly all, the river supply, and so cut off the upper waters of the rivers from the seaboard. They traverse flat and fully cultivated deltas, in which there are no great falls to be overcome, and which are ill provided with roads or other means of communication, while the people on the coast are accustomed to the use of boats. The navigable canals which have been made radiate from the head-works at Dowlaishweram and Bezwāda towards the sea, which connects their lower ends, while the head of each system is connected with the other, with the upper waters of the Godāvari and Kistna, with the Buckingham Canal, and with the railway system of the country. The total number of boats entering the two systems annually is about 80,000, with a tonnage estimated as varying from 700,000 to 800,000 tons, and there is also a small amount of raft traffic. The facilities for navigation are undoubtedly a great boon to the cultivators and tend materially to reduce the cost of transport; but although great use is made of them, it cannot be said that navigation even here is directly remunerative. The navigation receipts, which have been materially affected by the opening of direct rail communication along the east coast, average only 1½ lakhs per annum, and do not as a rule exceed the working expenses, so that there is little direct return on the expenditure incurred in rendering the canals navigable. On the other hand, the assessment on the irrigated lands would no doubt have to be considerably lower if the people did not enjoy their present advantages of cheap transport.
The only other irrigation work in Madras which has been adapted for navigation is the Kurnool-Cuddapah Canal, but here the conditions are very unfavourable. The canal is navigable throughout its entire length of 190 miles, in which there are no less than forty locks. It has, however, no water connexion with the seaboard, although it is connected with two railways, at its 109th mile and at its tail (Nandyāl and Cuddapah). The traffic is practically nil, although no navigation dues are levied, while the navigation charges exceed Rs. 10,000 a year. The advantages of navigation would probably have been more appreciated if the complete Tungabhadra project as originally conceived had been carried out, if a navigable connexion with the seaboard through Nellore District had been provided, and if large storage works had been constructed on the Tungabhadra so as to ensure at all seasons an adequate supply for purposes of navigation. However this may be, there is little prospect of navigation ever being successfully developed on the existing canal.
The three Major irrigation works in Bengal have all been adapted for navigation. The Orissa Canals comprise 280 miles of main canals and branches, of which 205 are navigable, as is the whole length of the Midnapore Canals (72 miles). It was originally proposed to connect these two systems by a high-level canal, which would afford a continuous water communication between Cuttack and Midnapore and thence to Calcutta; but this was only carried as far as Bhadrakh on the Salandī river. In the absence of such a connecting link, the boats on the Orissa Canals proceeded to Calcutta via Chāndbāli and the Bay of Bengal. But in order to provide inland communication for small craft, the Orissa Coast and Hijilī Canals, which are purely navigation canals, were subsequently constructed; and there is now free and uninterrupted water communication at all seasons between Cuttack at the head of the Orissa delta and Geonkhāli, near the confluence of the Rūpnārāyan and Hooghly rivers a few miles above Diamond Harbour. The construction of these two navigation canals has not resulted in such an increase in traffic on the Orissa system as was anticipated. During the year 1902–3 the combined navigation receipts of the three canals were less than Rs. 1,40,000, and the net revenue did not amount to Rs. 23,000, although the capital expenditure incurred on account of navigation alone cannot be estimated at less than a crore of rupees. Even less favourable results may be anticipated in future, as the traffic is affected by the recently constructed extensions of the Bengal-Nāgpur Railway. The Midnapore Canal has been more successful as a navigation work, but here also the traffic is likely to be affected by the railway.
The three main canals of the Son system, with an aggregate length of 218 miles, are all navigable and tail into the Ganges at Buxar, Arrah, and Dinapore. The traffic has always been very moderate, and the tolls do not cover the working expenses due to navigation. During the last two years the receipts have suffered serious diminution in consequence of the opening of the Mughalsarai-Gayā Railway; in 1902–3 they did not amount to Rs. 19,000, which is about a quarter of the average for the three years ending March, 1899.
In the United Provinces the main lines of the Upper and Lower Ganges Canals (275 miles) and of the Agra Canal (100 miles) are navigable throughout. The Ganges systems are connected with the Ganges river at Cawnpore, and the Agra Canal has a channel leading into the Jumna river near Agra, while it is connected at its head, by means of the Okhla navigation channel, with the Western Jumna Canal in the Punjab1. It might have been expected, therefore, that a considerable through traffic to Calcutta would be developed, but as a matter of fact the traffic is purely local and of inconsiderable volume. The average number of boats plying monthly on the canals in the United Provinces is less than 250, and the traffic receipts do not average more than Rs. 15,000 per annum, which is about half the amount of navigation working expenses.
In the Punjab the only navigable canals are portions of the Western Jumna and Sirhind systems. The former is navigable from its head to Delhi, and is then, as already stated, connected with the Agra Canal and the Jumna. A portion of the Hānsi branch is also navigable, the total length of navigable channels being 207 miles. The Sirhind Canal is navigable for 180 miles, or from its head at Rūpar to Ferozepore, where the canal connects with the river Sutlej, and there is a continuous waterway onwards to Karāchi. The boat traffic on these canals is so insignificant that the details are not recorded, and the boat-tolls on both together amount to less than Rs. 5,000 per annum. There is, however, a considerable raft traffic, particularly on the Western Jumna Canal, where the rafting dues average about Rs. 40,000 per annum. The rafts consist principally of timber, sleepers, scantling, and bamboos, which are floated down from the hills to the canal heads, and are then passed into the canals. These rafts used to pass down the rivers before the canals were constructed, but the abstraction of the whole of the cold-weather supply has rendered the canal route necessary. The effect of constructing the canals has therefore been to divert rather than to create traffic.
Inundation canals are not adapted to navigation, with the single exception of the great Fuleli Canal in Sind, which enjoys an almost perennial supply and is so situated that it has been possible to render it navigable at a very moderate cost. The Eastern Nāra, which takes off from the left bank of the Indus above Sukkur, is also navigable all the year round.
Besides the state irrigation works which are also navigable canals, Madras and Bengal possess other canals which have been constructed solely for purposes of navigation. In the former the most important work is the Buckingham Canal, which is connected with one of the branches of the Kistna system and then, proceeding almost due south, skirts the Coromandel coast for a distance of 262 miles, and passes through Madras city. It is a tidal canal, and costly locks have been required at each of the numerous river-crossings, so that the capital cost of the work amounts to nearly 90 lakhs of rupees. There is a moderate traffic on this canal, on which more than 200,000 tons of goods are carried annually; but this also has been affected by the extension of rail communication along the east coast, and the navigation receipts do not fully cover the working expenses. Nevertheless the work must be regarded as one of considerable public utility. There is also a navigation coast canal in Tanjore District, known as the Vedāranniyam Canal, and a small canal connected with the Chilka Lake in Ganjām District. Madras possesses no other navigable canals, but has many miles of natural waterways formed by river backwaters, especially on the west coast, and on these there is a very considerable traffic. Mention may also be made of the works on the first rocky barrier on the Godāvari, at Dummagudem, 120 miles above the head of the Godāvari Canals at Dowlaishweram, which render that river navigable for boats to the foot of the second barrier, a distance of 188 miles.
Reference has already been made to the Orissa Coast and Hijilī Tidal Canals in Bengal. There are two other important works in the Province, the Calcutta and Eastern Canals, and the Nadiā Rivers. The former consist of short lengths of locked canals and open channels which have been made for the purpose of connecting many of the natural river-channels in the deltas of the Ganges and the Brahmaputra, so as to provide a continuous interior line of communication between Calcutta and the Sundarbans, Eastern Bengal, and Assam, for the benefit of steamers, flats, and small craft constructed for inland waters. These works are of great importance, for they comprise a length of 735 miles of navigable canals and river-channels. There is a great deal of traffic, the average tonnage of cargo and passenger-boats amounting to over 1,000,000 tons a year. The works are more nearly self-supporting than any other navigation works in India. The capital expenditure has been about 68 lakhs, the average annual receipts amount to 4¼ lakhs, and the net revenue exceeds 1½ lakhs. The Nadiā rivers comprise the Bhāgīrathi, the Jalangī, and the Mātābhānga, which constitute the upper waters of the Hooghly. These rivers have not been canalized, nor has any capital expenditure been incurred on navigation works. Steamers can navigate them from July to October, but in very dry years they are not open even to small craft throughout the remaining months. Some expenditure is incurred annually in keeping open the channels, which have an aggregate length of 472 miles. The aggregate capacity of the boats using them amounts to about 500,000 tons per annum, and the tolls to 1 lakh, the average working expenses being about Rs. 1,20,000.
In Burma, which abounds in natural waterways, there are only two important navigable canals, the Pegu-Sittang and the Sittang-Kyaikto. These are maintained by Government, but no capital expenditure on them has been recorded. The receipts on the former amount to about Rs. 1,40,000 per annum, but do not cover the working expenses. On the latter no revenue is realized, but about half a lakh is spent annually on maintenance. Some expenditure is also incurred on the conservancy of navigable rivers, so as to facilitate navigation.
From the foregoing description of the navigable canals in British India it will be seen that not one of them, whether intended primarily for irrigation or constructed solely as a navigation work, is directly remunerative, and that in many cases the navigation receipts do not cover the working expenses. The extension of navigation works must therefore depend for its justification on the benefits which the country will derive from a reduction in the cost of transport. This reduction will be real and material only in those tracts where the public can be trusted to avail themselves freely of the advantages offered by water communications. There is no Province in which facilities for navigation are so likely to be appreciated as in Eastern Bengal, with its dense population, its thriving industries, and its innumerable river channels. There is great scope for further extensions of inland navigation in this tract; and if the works will not prove directly remunerative, they are not likely to impose a heavy permanent charge on the state, while they will be of great value in increasing the wealth of the community. Navigation advantages have also been fairly appreciated in the deltas of Orissa and Madras, and have certainly proved of great value to the cultivators. Elsewhere, as on the Kurnool-Cuddapah and Son Canals, or the great perennial canals in Northern India, navigation may be regarded as a failure. Not only have the works required in connexion with it been very costly, but the people have evinced no desire to avail themselves of the means of navigation afforded. There is, moreover, another objection to the combination of navigation with irrigation on perennial canals which deserves notice. Navigation is not only costly, but cannot be maintained during seasons of short supply without detriment to the efficiency of the canal as an irrigating work. In general it may be said that, outside the deltaic tracts in Bengal, Orissa, Madras, and Sind, navigable canals will never be of much use or value as a means of inland communication.
We have hitherto referred only to navigation works by which the natural waterways of the country have been artificially connected. It remains to consider briefly the value of these waterways themselves as a means of inland communication. Some of the large rivers of India, such as the Narbadā and the Tāpti, are unfortunately, by reason of their rocky beds and swift floods, practically useless for navigation except at their mouths. The Indus, the Ganges, and the Brahmaputra are navigable by steamers all the year round, or for the greater part of the year, for hundreds of miles above their mouths, or above the heads of the navigable canals traversing their deltas. The Indus is thus constantly navigable as high as Dera Ismail Khān in the North-West Frontier Province, 800 miles inland and 250 miles beyond the Sind border. Its tributaries, the Chenāb and the Sutlej, are open to small craft all the year round, but there is little traffic above Chiniot on the former and Ferozepore on the latter. The Fuleli Canal and the Eastern Nāra may be regarded as navigable branches of the Indus. The Ganges is navigable as high as Cawnpore, and steamers also pass up the Gogra as far as Fyzābād. Steam navigation on the Indus and the Upper Ganges has, however, been greatly reduced by the development of rail traffic, though on the Bengal portion of the Ganges it is still active. The Brahmaputra is navigable by steamers as high as Dibrugarh, and there is steam navigation on its tributary the Surmā as far inland as Sylhet and Cāchār. The Hooghly is navigable all the year round up to Nadiā, and steamers can pass up the Nadiā rivers above this point from July to October. The great rivers on the east coast of the Peninsula—the Mahānadī, the Godāvari, and the Kistna—are all navigable for some distance above the heads of their deltas, but the traffic on them is not very considerable. Many tributaries or branches of these rivers are also navigable during the monsoon months, and several are open to small craft all the year round. There are, in effect, all round the coast innumerable small rivers, creeks, and backwaters affording facilities for water transport which are fully utilized by small native craft; but outside the zone of such operations inland navigation is practically confined to the deltas and to the valleys of the great rivers which form the natural waterways of the country.
As to Burma, there is no Province in which the natural waterways afford greater facilities for inland navigation, or in which it is more extensively practised. In the Arakan Division the Mayū and Kaladan rivers are navigable by steamers during a great part of the year for distances of 50 and 90 miles respectively above Akyab. Many of the other rivers which fall into the Bay of Bengal are similarly navigable for some miles above the seaboard, and native craft can of course proceed much farther inland. Farther east the great Irrawaddy river, which traverses nearly the whole length of Upper and Lower Burma (excluding Tenasserim), is navigable by steamers at all seasons of the year as high as Bhamo, or more than 500 miles from its mouth, and steam launches and country boats can proceed much higher, or for some distance to the north of Myitkyinā. The numerous deltaic channels which fall into the Bay of Bengal from the Irrawaddy form waterways connected with the main river and with the sea, which are for the most part navigable throughout the year, while higher up many of the tributaries of this river are navigable for some distance above their confluences. Of these the two most important are the Chindwin and Myitnge rivers, the former of which is navigable by steamers during the monsoon as high as Homalin in the twenty-fifth parallel of latitude, and by smaller craft as far as Kyaukse. Farther east again are the Sittang and the Salween rivers, both of which discharge into the Gulf of Martaban. The former of these is navigable by small steamers during the monsoon as far as Toungoo, and for small craft as far as Pyinmanā in Upper Burma. All these rivers have numerous navigable affluents, and branches or mouths which spread fanlike over their deltas and are in navigable communication with each other. There are also, all round the Burma coast, innumerable creeks and backwaters open to steamers or smaller craft.
The tolls charged for navigation vary very much on different works, and cannot be easily compared. In the United Provinces and the Punjab monthly or quarterly charges are made according to the size of the boat, which is then free of the canal during the period for which the pass is given. On the Fuleli Canal in Sind the same system is adopted. In Bengal the charge depends on the maundage, which is taken roughly as half the displacement measured in cubic feet. There are fixed rates per 100 maunds for different reaches, which vary with the length and other circumstances. In Madras the Bengal system of distance-tolls on all vessels plying between certain stations was originally in force; but has been replaced by a system of annual and short-term (6 weeks) licences, for which moderate fees are charged, and which leave the holders free to navigate all parts of the connected waterway systems. In Burma navigation tolls are levied only on the Pegu-Sittang Canal, where they are charged on the carrying capacity of the boats expressed in baskets of paddy (‘unhusked rice’), and on the Shwetachaung Canal in Mandalay District, which is navigable for about 13 miles. Conservancy or registration fees are sometimes charged on vessels navigating the large rivers. On the Indus the charge is at the rate of Rs. 4 per annum per maund of registered burden for native craft, while steamers which pay no port dues are charged three times this amount.
The comparative merits of railways and navigable canals as means of communication have sometimes been discussed. The principal argument in favour of the latter is that the cost of haulage or transport is less. On the other hand, canal routes are more devious; cross-communications or connexions between different systems are more difficult; feeder canals cannot be taken into new areas of supply with the same facility as feeder railways; the time occupied in transport on canals is much greater; and, lastly, there are many tracts in which canal navigation cannot be maintained, even at great expense, without the diversion and absorption of a large volume of water which might be more advantageously used for irrigation. The comparison of these advantages and disadvantages is, however, only of academical interest. The commercial, the administrative, the military, and even the agricultural needs of India could never have been met by means of navigable canals alone; and the real question is not of the comparative merits of canals and railways, but whether in particular cases navigation canals may not suffice for immediate requirements, or may not serve as useful supplements to railway facilities already provided. Individual cases of this kind must be considered on their merits, but it is noteworthy that there is hardly a single tract in which navigable canals have been made by Government in which a railway has not been subsequently constructed. There may be now, or eventually, room for both, but the fact that the construction of a railway causes a very serious diminution in the volume of previous canal traffic indicates that the lower cost of haulage on canals cannot in many cases be set against the other advantages which may be claimed for railways. It may be said that during the severest stress of recent famines railways have sometimes failed for a time to cope fully with the situation, but this has been generally due to the want of a reserve of rolling stock, and navigation canals would be equally liable to failure from insufficiency of cargo boats. It is, indeed, difficult to conceive any practicable system of navigable canals by which the exigencies of famine-distressed districts could have been met as effectively as by the existing network of railways. Lastly, it may be said that, whatever the comparative capital cost of navigable canals and railways may be, the former have never proved directly remunerative, whereas the latter yield on the whole a net revenue which is more than sufficient to cover all the interest charges on their capital cost. Much can, no doubt, still be done for the improvement of communications in the deltas of Eastern Bengal and similar areas by improving the open waterways of the country and connecting them by navigable cuts, but outside these tracts the field for the extension of inland navigation cannot be regarded as either wide or promising.
T. Higham.
Bibliography
A. Deakin.—Irrigated India (1893). R. B. Buckley.—Irrigation Works in India and Egypt (1893). J. and R. Strachey.—Finances and Public Works of India (1882). Report of the Irrigation Commission, 1903. Reports of the Famine Commissions of 1878–80, 1898, and 1901. Annual Irrigation Reviews of Government of India and Local Governments. ‘Completion Reports’ are also published, under the authority of the Government of India, for each Major irrigation work when completed.
Notes
- 1. For a full account of the rainfall conditions, see Vol. I, chap. iii, ‘Meteorology.’
- 1. Sir J. Strachey’s India (1903), p. 222.
- 1. See Vol. IV, chap. vi, Finance.
- 1. Since 1902–3 the Dera Ghāzi Khān Canals have been transferred to the ‘Major’ head.
- *. The area shown under ‘Other Sources’ includes a large area (5,000,000 acres) irrigated in Bengal from private canals, and from water held up in natural depressions and in shallow artificial tanks.
- 1. The Agra Canal was closed to navigation in June, 1904.