Himalayan Journal vol.07
The Himalayan Journal

Publication year:

Kenneth Mason
    (H.W. Tilman)
    (I. GENERAL r. finsterwalder)
  5. A VISIT TO NUN KUN, 1934
    (F. S. SMYTHE)
    (J. B. AUDEN)
    (A.P.F. Hamilton)
    (Captain F. KINGDON WARD)
    (P. C. DUNCAN)
    (Lieut.- Colonel Kenneth Mason)
  16. NOTES


A.P.F. Hamilton

To the Himalayan enthusiast the Siwaliks are often little more than a name, an 'insignificant range of foot-hills', to be crossed impatiently in answer to the call of the grander ranges beyond. On the other hand, modern geography is closely concerned with the economic values of geographical features, and to-day people take much more interest than formerly in their surroundings. It is hoped, therefore, that this paper may stimulate some further interest in the important, and perhaps notorious, part that this low range is playing in Punjab economics.

The damage caused by the great floods which periodically sweep down the Punjab rivers from the high mountains attract large headlines in the newspapers; and, though they may bring great havoc, they are, in the main, caused by unusual phenomena, such as the breaking down of glacier dams and cloudbursts, which are beyond the control of man. Yet nearer at hand, under the very eyes of the public, the Siwaliks and outer ranges of the Himalaya are being slowly converted into a source of destruction, the more dangerous owing to the insidious nature of its growth.

The object of this paper is to show how men, by misguided efforts to better their condition or under the stress of economic pressure, have contravened the laws of nature, thus meriting a punishment which only too often falls upon the heads of those innocent of the offence. The paper begins with a description of the role of mountains and forests as agents for the supply and distribution of water. A brief discussion of the geology of the Siwalik range follows, and the rest of the paper deals specifically with erosion in the Siwaliks of the Hoshiarpur district in the Punjab.

Mountains as a source of water-supply, and the role of mountain-forests.
Throughout the world mountain ranges play an important physical part in the water-supply of man. They act as barriers to the passage of winds laden with moisture received by evaporation from seas, from inland lakes, and from the land itself. The warm air currents, impinging against the mountain slopes, are deflected upwards; they are cooled, and they precipitate their moisture in the form of snow or rain.

In northern India, during the winter and early spring, depressions from the west penetrate far into the Himalayan barrier; heavy snow falls on the inner hills, but little rain is experienced on the outer slopes at this season. Snow melts during the late spring and early- summer, causing a gradual rise in the rivers and streams. At midsummer the rains of the South-west Monsoon, mainly from the Bay of Bengal branch, strike the mountain barrier, and bring rain to the successive ridges south of the Great Himalayan range, where the remaining moisture is given up. The rain is partly absorbed by the soil, but the greater part, as surface run-off, finds its way down through countless streams and rivulets to the main rivers, and thence down to the plains of India where it serves the many requirements of man.

Winter snowfall on bare alpine land melts and drains away slowly; but it does not require much imagination to realize what happens to rain-water falling on bare mountain-sides. The water dashes down the slopes with ever-increasing speed, carrying away the thin soil and discharging itself into the low-lying country with uncontrollable violence. Naked rock is left behind and the land receiving the discharge of such torrents is subject to constant floods. Fortunately for man, it is a provision of nature that certain regions of the globe are covered with forest, and mountainous country is one of them. Below the snow-line, the forest clothes the hill-sides down to the plains and controls the flow of the water from the catchment areas in the hills.

A forest consists of trees or scrub, growing together with bushes, herbage or grass and humus on a soil, of varying depth, derived from the disintegration of the underlying rock. The types of vegetation found in mountainous regions depend mainly on the amount of rainfall and elevation. They vary between wide limits, from dense fir forest to little more than grass, but all play their part as agents of protection. The chief roles of the forest are-

(1) reduction and control of surface run-off,

(2) conservation of moisture, and

(3) protection of the soil from erosion.

As regards the first of these, the leaf canopy retards the rain and breaks its force; the humus layer absorbs some moisture; and the roots of trees and shrubs penetrate the soil, make it more permeable, and increase its absorptive capacity. When the soil is saturated, surface run-off begins, but its force and speed are greatly reduced by the presence of the close-growing vegetation and humus. Steep ground and impermeable soil are conducive to the most rapid run-off and are most in need of protection from this point of view. When the protective covering is removed, for instance by heavy grazing and browsing, run-off is accelerated, and the banks of rivers and streams may be unable to hold the increased discharge. It must, however, be admitted that floods caused by exceptional meteorological conditions cannot be prevented by forests, though the destructive effects are certainly minimized.

Secondly, the forest, by retaining a covering soil in situ, by increasing the absorptive capacity of the soil, and by allowing water to sink slowly into cracks and crevices in the underlying rocks, enables the mountains to function as natural reservoirs. Long after the cessation of the rains the stored water gradually finds its way down to feed the springs and the water-table beyond the foot of the hills, while a higher level of stream-flow is maintained during the dry season. Briefly, therefore, a forest-covering tends to equalize the discharge of rivers and streams throughout the year and to increase the available water-supply; removal of this covering causes floods or desiccation, or both.

Thirdly, forests, by breaking the force of falling rain, particularly where heavy storms are frequent, by binding the soil with their roots, and by reducing the erosive action of run-off, prevent the erosion of hill-sides.

Erosion begins with the removal of the vegetable covering. There are two types of erosion, both of which are in evidence in any one locality but, according to the nature of the soil, one or the other is generally predominant. ' Gully' erosion originates, as its name implies, from the formation of small channels by water action; the channels deepen and cut back into the hill-side producing the well- known 'ravine' lands which are as common in bare undulating country as in more mountainous regions. 'Gully' erosion is most active in heavy, coherent soils, such as clays and marls, less so on sandy and friable soils, and least of all on stony ground derived from conglomerates.

The second type of erosion, 'sheet' erosion, implies the removal of soil particles by the flow of water over open surfaces rather than in channels. This form of erosion acts on all surfaces which are not protected by close-growing vegetation or a humus layer. Loose and friable soils, such as sandy soils and sand-rock, are the most liable to 'sheet' erosion. These coarse-textured soils maintain a smooth or rounded configuration under the action of water.

When slopes are steep and the soil and underlying rock of a friable nature, when the forest covering is scanty or absent and rainfall heavy, erosion may reach colossal proportions and the degradation of the hills is accelerated by the occurrence of landslips.

Erosion may be harmful in the following ways:

(1) The products of erosion are carried down by floods, and deposited over low-lying country which is thus rendered sterile. Desiccation, caused by the adverse hydrological effects described above, is increased by these deposits.

(2) Irrigation works and canals may be silted up, and the fields, which were formerly enriched by annual deposits of fertile soil derived from the protected slopes, now receive only a coarse, sandy deposit.

(3) Erosion increases flood height, for the detritus carried down by the water swells the stream volume to such an extent that the height of the water is raised far above what it would be if it were free from sediment. The effect of this may be gauged from figures calculated by a French expert who, for a certain mountain torrent, computed that 85,020 cubic yards of water brought down 221,052 cubic yards of detritus, or more than two and a half times the water volume.

(4) Silt deposition may entirely fill up a river-bed and so cause constant flooding; or at least the bed may be so raised that overtopping of the banks may be caused by comparatively light falls of rain.

Regarding the effect of forests on climate, this is a matter which has been in dispute for many years and evidence is contradictory. It is possible, but not probable, that forests have some effect on climate as a whole; but there is evidence, in the Terai of the United Provinces, which tends to prove that the clearing of the Sal forests during the period of British occupation has caused greater extremes in temperature; and there is little doubt that forests increase local precipitation both in abundance and frequency and that this effect is likely to be greater in hot, dry climates and in mountainous regions. Vegetation, by cooling the air both in and around it, also increases the condensation of moisture in the form of dew and clouds, reduces radiation, and prevents the evaporation of moisture from the soil by hot winds. The removal of forest covering, therefore, tends directly to increase desiccation.

Modern opinion is in general agreement that in countries consisting of hills and plains, the fertility of the plains, and with it the prosperity of the people, depends largely on a regulated flow of water from the hills, and that there is a direct relation between hill vegetation and floods. There are few mountain ranges outside the polar regions which, below the snow line, were not covered with vegetation before man began his work of destruction. Examples of the evils caused by disforestation can be quoted from earliest history.

Ancient Babylonia was once a fertile land watered by a vast irrigation system. Invaders may have destroyed the actual irrigation works; but the civilization of Babylon was more certainly destroyed by fire and axe in the hills to the north whence came the 'waters of Babylon'. Agents of destruction far greater than invaders were let loose upon the country, transforming it into the desert of which the greater part remains to the present day.

In France, up to the time of the Revolution, the forests were under the protection of powerful landlords. When they fell into the hands of the peasants, rapid destruction took place. In a short time floods from the denuded hills caused terrible havoc to villages and cultivation below, so that the French Government is, and has now been for many decades, spending huge sums on reforestation and torrent control.

In Ceylon the clearing of hill-sides for tea-planting and cultivation has led to erosion sufficiently serious to merit the attention of Government.

Examples could be multiplied, and few countries in the world have escaped the evil results of forest destruction, but the fertility of Egypt, unchanged throughout the ages, is due to the river Nile which has its source in the forest-clad mountains of Africa, beyond the reach of civilization.

Siwalik Geology and the Nature of Chos.

The Siwalik range skirts the Himalaya almost without a break from the Indus to the Brahmaputra, a distance of about 1,600 miles. Topographically the range belongs to the hills, geographically to the plains; for although its upheaval coincided with earth movements in the Himalaya, the latter were raised early in Tertiary times, while the Siwalik rocks are composed of sediments brought down from the rising ranges to the north. These freshwater sediments have been laid down along the front of the Himalaya ever since they started to rise. They fill what is known as the Indo-Gangetic 'deep' or depression, and the recent deposits still in process of formation along the foot of the hills are their direct successors.

The Siwalik range, therefore, is composed of rocks similar in nature to those underlying the plains of northern India and the range marks the northern limit of these freshwater deposits. It is remarkably uniform in character throughout, and though in some places it is pressed against the outer ranges of the Himalaya and not easily identified, it is more often separated from the latter by valleys, such as the 'dun' valleys in the United Provinces and the Kangra valley in the Punjab.

The most important lithological character of the rocks is their lack of consolidation by geological pressure, certainly in those strata which lie exposed. The beds are stratified, and consist of loose, pebbly conglomerates, soft-earths and barely coherent sand-rock. In the Hoshiarpur Siwaliks the sand-rock predominates and is exposed in massive beds, often several hundred feet thick. Narrow beds of clays, earths, and gravels are frequently associated with the sand-rock; wherever the strata are inclined by folding, these soft beds have been eroded and the sand-rock projects in a confused array of sharp points and steep ridges. Where the conglomerates occur they overlie the sand-rock, forming shallow beds of loose pebbles.

If the Siwaliks form only a small proportion of the catchment areas of the great Punjab rivers they are, owing to the nature of their rocks, liable to contribute a large quota of silt when the other factors favouring erosion are present. Moreover, particularly on the southern slopes, they border cultivation for the greater part of their length and are bound to have a considerable hydrological effect on the adjacent lands. East of the Jumna river, owing partly to sound forest policy and partly to a moister climate, the Siwaliks are well covered with vegetation; but in the Punjab all the conditions favourable for rapid run-off and erosion are present. Owing to a lower rainfall and to a type of monsoon weather characterized by heavy storms succeeded by bright intervals, the conservation of water in the hills and a steadying of the river discharges are most urgent.

In the Hoshiarpur district, in particular during the last eighty or ninety years, immense damage has been caused by the disforestation of the Siwaliks, and a brief history of how it originated is given later. The damage has mainly been caused by torrents, locally called chos, which sweep down from the southern slopes of these hills during the monsoon.

The true torrent is characterized by the steepness of its gradient and the violence and irregularity of its discharge. As a general rule the channels are dry except at times of rainfall, when sudden and short-lived floods develop as a result of uncontrolled run-off. When hill slopes are denuded of vegetation, torrents may result from severe 6 gully5 erosion, or streams, previously harmless, may be transformed into torrents. The amount of damage done may be very great, but it depends on a large number of considerations. Inside the hill tract velocity is maintained and the flood carries much material in suspension; even small boulders and stones may be transported, but at the point where the torrent debouches from the hills an easier gradient is encountered, velocity is reduced, and the heavier material is deposited in a characteristic fan-shaped mass called the detrital cone, or cone de dejection. The shape is due to the fact that the water is discharged along the top of a slowly rising convex bed formed by the detritus and is obliged to flow off in all directions, extending the 'fan' on all sides and increasing the radius of its damaging effects. In the earlier stages of its formation the convexity of the detrital cone is usually more marked, but as it spreads this feature becomes less evident, the velocity of the flood is reduced and the rate of extension of the cone is retarded, while floods in the country farther afield may become less violent. Thus if a curve of torrent activity could be drawn it would show that a peak is reached in the earlier stages; the curve would then gradually fall and finally run more or less level.

The behaviour of the flood and the damage it may do farther out must be considered. If from the beginning, the water finds, or can make, a definite channel which it can deepen, little damage to the surrounding country by flooding or silting may result, but the chances of its doing so depend mainly on the following factors:

(1) the lie of the land,

(2) the nature of the soil.

(3) the amount of sediment in suspension,

(4) the velocity of the current, and

(5) the final destination of the torrent.

The water naturally follows lines of greatest declivity and of least resistance; the lie of the land can either help the torrent to pursue a single channel or may force it to spread over the country. Firm soil, by providing stable banks, favours the first alternative, loose sandy soils favour the latter. The greater the amount of matter in suspension and the more sudden any reduction in velocity due to change of gradient at any point, the more silt is deposited; this either fills up any existing channel or prevents the current from making a new one. Apart from the silt factor, the faster the current the more easily it cuts out a channel for itself. When the torrent flows down a gradually decreasing gradient which ultimately leads to flat ground, it finds no exit, flow finally ceases, and the whole of the products of erosion are deposited either in the bed of the torrent which must gradually rise or, by inundation, over the surrounding country. For the most part the Hoshiarpur Siwaliks have a gradual slope from the edge of the plains to the crest, a rise of 600 feet to 800 feet in from three to five miles, and externally they appear unlikely to cause serious torrents. Owing, however, to the effect of water action on the soft rock, deep nullahs and ravines have been carved out of the hills, and the hill topography is characterized by steep, often precipitous, slopes and gorges. The main drainage channels are fed by numerous precipitous side-streams, and, on the upper slopes, by a network of converging feeders, which give to the channels catchment areas of a size and shape out of proportion to the external features of the range.

The chos and their main feeders have indeed been cut to such a level by erosion that one may leave the practically flat stretches of sand in the plains, and walk from two to four miles along the chos within the hill boundary without any perceptible change in the gradient. Levelling a cho of average size over a distance of two and a half miles showed an average gradient of less than i per cent. (50 feet in a mile). This peculiarity distinguishes chos from true torrents. At the point where the cho debouches from the hills the detrital cone is hardly noticeable, though most chos show a slight convexity at the centre of the sandy bed. At one time it was probably more marked, for after issuing from the confines of the hills, chos spread out to form expanses of sand which, in extreme cases, are nearly a mile wide.

The country subsequently traversed by the chos is best described as 'plains'; it is largely cultivated and the soil is sandy. The original drainage channels rapidly silted up when hill erosion and flooding started, with the result that the floods have been forced out over the country and numerous branches have been formed.

The branches of adjacent chos frequently unite forming a network of sandy beds which stretches for about twelve miles into the plains, and conditions shown in the accompanying map are typical of the country within a few miles of the hills. This network condition gradually becomes less marked with increasing distance, and the beds narrow down and finally die out at distances of from fifteen to twenty-five miles, as the crow flies, from the foot of the hills.

Without exception cho beds in the plains are broad rivers of sand with a very slight fall. Banks are often non-existent or, at the best, composed of unstable sand liable to be washed away by any flood, or of scarped cultivation which is being undercut by every flood. The various factors affecting torrent action and the harmful results of floods in general have already been discussed; and to show that these chos cannot but have a most destructive effect on the lands which they traverse, it is only necessary to add that conditions in their catchment areas are most favourable for very rapid run-off and erosion.

The question may be asked how it is that with so slight a gradient the chos have been able to acquire sufficient force to do so much damage. The rush of water across the plains has to be seen to be believed, and the explanation is to be found in the topography of the catchment areas and the barren nature of the slopes, both of which combine to discharge into the plains, suddenly and under great pressure, an irresistible flood of water and sand. The flood subsides as suddenly as it starts and the immense volume of water, much of which might have been stored in the hills and subsoil of the plains, is lost.

There is reliable evidence that about 100 years ago the chos ran between well-defined banks, and that in some places perennial streams, which were used for irrigation, issued from the hills. To-day, except for floods, they are dry throughout the year. No accurate measurements have been made, but observation shows that many beds are gradually rising. As a result, the chos frequently change their courses so that, apart from the fact that sand is a most unfavourable medium for constructing banks, the training of cho floods in their present state of violence becomes an engineering impossibility, except perhaps at a fabulous cost.

Such is the nature of the notorious Hoshiarpur chos. From the eighty miles' front of the southern slopes of the Siwaliks nearly a hundred such torrents discharge their floods into the plains; and enough has been said to show that no half-measures will be of any avail in controlling them.

History of the Hoshiarpur Chos.

During the Moghul raj the Siwaliks were in the hands of chieftains whose duty it was to protect the rich lands of the plains from invasion from the north. At that time the forests were strictly preserved in the interests of sport. Later, under Ranjit Singh's rule, the Rajput and other chiefs who were granted jagirs, appropriated all the land for their own use; the peasants, who previously had sufficient land to live on, were now compelled to seek a livelihood by grazing cattle and by cutting and selling wood from the forests. From this time the destruction of the forests began, but the inroads made were not very serious. In 1852 the first Land Settlement under British rule was made. As there were no records of land tenure and many of the Rajput chiefs had been dispossessed owing to their hostile attitude to British intervention, almost any one who claimed possession was granted it. The result was that the Siwalik forests became the property of countless irresponsible petty landlords.

Under British rule an era of prosperity and development began. Population increased, railways developed, cantonments were established, and the demand for firewood and charcoal was greatly extended. The improvident landlords wasted no time in getting what they could out of their newly acquired property; large areas of forest land were leased out for a few rupees to contractors who had a free hand to fell all the trees they liked. Cattle-herding, formerly a rather chancy source of livelihood, now became safe and profitable. Herds were grazed on the hills in increasing numbers, and goats, in particular, swarmed everywhere, eating down all shrubs and bushes and preventing any fresh growth from developing. In less than thirty years destruction was complete, and, to quote from a report of Mr. Baden-Powell, Conservator of Forests, made in 1879, 'So great has been the destruction that one may march for miles and miles with nothing in view but mud-coloured crests and rugged slopes, rarely dotted with greyish, browsed-down bushes. . .

For many years previous to this report the chos had been rapidly increasing in violence. Local officers had made it clear in reports that this was due to disforestation and that the damage could only be stopped by preventing the further destruction of the vegetation and by taking steps to reforest the hills. But it was not until 1902, when the worst of the damage had been done, that effect was given to legislation designed to stop the evil.

About this time exhaustive inquiries were made to estimate the damage which had been caused by the chos, and it is convenient to summarize the results here.

Loss of cultivation.

The area in acres under cho beds at three successive settlements in the Hoshiarpur district were as follows:

1852 1884 1897

48,206 80,057 945326

Thus losses increased by almost 100 per cent, in forty-five years. In addition about 26,000 acres had been partially damaged. In the Jullundur district 2,600 acres had been completely and 1,500 partially destroyed. Many houses were washed away and an uncalcu- lated area of grazing and forested land destroyed. The fertility of surrounding fields had been reduced by deposits of wind-borne sand.

Financial loss to villagers.

It was calculated that in the two districts, between the 1884 and 1897 settlements alone, the land value of the thousand odd villages affected had decreased by Rs. 20 lakhs as a result of cho action. The indebtedness of these villages amounted to Rs. 96 lakhs, a large proportion of which was directly due to loss of land.

Losses incurred by Government.

The annual loss in Land Revenue due to cho action must by now have reached a large figure and many grants of land in irrigation colonies have been made to those who had suffered most heavily.

From records it seems that cho activity reached a peak before 1897; figures collected in 1913 show that the rate of destruction of cultivated land had fallen off since 1897. The causes of the decline are-

(i) cho activity reached a peak before 1897;

(ii) on account of (i) the force of the floods has since been partly spent on land already devastated;

(iii) decrease in rainfall;

(iv) protection in the Siwaliks, to a very small extent.

Detailed examination of records and of the state of catchment areas and cho beds indicate that for some years chos have followed a more or less constant level of destruction, and this is likely to be maintained until improvement takes place in the catchment areas.

Official records show the area of land actually under cho bed at any one time; they do not give the large areas of sandy waste previously destroyed but not now lying under torrent influence. The total area destroyed since the beginning can only be estimated, and is probably not far short of 150,000 acres. Some reclamation has been effected, but in this connexion the following passages taken from official correspondence reveal the true situation:

'land reclaimed from chos is always of poor character . . . and eventually falls back, as a rule, to an unculturable condition' . . .


'the action of the chos tends to the constant diminution of the area of the good land, and every such diminution must, in a tract so thickly populated as that in question cause the prosperity of the people to decline'.

So far only the more obvious damage, that of the floods, has been considered; earlier in this paper it was stated that disforestation of hills may have other reactions which, though they may be less serious at first, are likely to be more widespread in effect in the long run.

The water-table in the Hoshiarpur and Jullundur districts has been slowly but steadily falling in past years. This was investigated and reported on in 1931 by Mr. G. de P. Cotter, sc.d., f.g.s., Superintendent, Geological Survey of India. He attributed the fall to-

(a) decrease of rainfall, perhaps partly caused by disforestation;

(b) increase of wells used for irrigation;

(c) drying up of the river Beyn, as a result of (a) and also probably through disforestation of the Siwaliks.

That the rainfall in the Hoshiarpur and Jullundur districts has declined in recent years is shown by the following figures:
Average annual rainfall from Inches Station
1862-82 36-70 Hoshiarpur district.
1883-90 Not available.
1891-1900 34'45 Average of Hoshiarpur and Jullundur.
1901-10 3I-58 ,,
1911-20 28-71 ,,
1921-30 26-61 ,,
This decline has undoubtedly affected the water-table, but what is the cause of the decline? Statistics supplied by the Director General of Observatories, India Meteorological Department, show that rainfall has not decreased in the Punjab generally during the past forty years. The possibility, therefore, that the decrease in these two districts may be due to local causes cannot be ignored. The regularity with which the decrease has taken place is significant, and disforestation may be a contributory factor.

The more intensive utilization of well-water must also affect the water-table, and if rainfall continues to decrease the drain on the subsoil water from this cause can only be aggravated. But there are, probably, other factors operating; for in a zone a few miles wide along the forefront of the Siwaliks, the fall in the water-level appears to be greatest; wells have not increased here, in fact many have dried up completely and in others the water is sinking fast. There can be little doubt that as a result of disforestation the flow of water from the Siwaliks by underground percolation into this zone has been seriously interrupted. Again, since the flow of water in the cho beds is confined to very short periods when the floods come down, little water is able to percolate into the subsoil. This factor might operate at a greater distance from the hills than that of underground percolation, and its effect would be felt more in the upper strata whence the wells draw their supply.

Mr. Cotter is not in favour of pumping; the water-bearing strata are not inexhaustible and the ultimate source of underground water is percolation from rainfall; rainfall cannot be increased at will, and the utilization of underground supplies is likely to extend. The presence of forest on the hills may or may not influence the amount of rainfall, but there can be little doubt that, when restored, it will eventually have the desired effect of increasing the availability of rainfall for utilization, by controlling its distribution from the Siwalik catchment area.

Nothing perhaps can be gained by attempting to assess the comparative effects of the various factors; the symptoms of desiccation are evident, and disforestation, the root cause, has been there for several decades.

Many attempts have been made to check the damage of the chos in the plains by constructing bunds; with a few minor exceptions, all these attempts have failed, as they are bound to; for, apart from the force of the water, there are no means of dealing with the accumulations of sand. The fact must be faced, and experts have for years pointed out, that reforestation alone can bring any measure of relief. Erosion and run-off can only be checked by replacing the vegetable covering.

The legislation enacted in 1902 provided some measure of protection, and goats were evicted; but by then there was nothing on the hills worth protecting, and the very scanty vegetation which has resulted is all that could be expected. Up to the present no provision has been made for systematic reforestation.

The first requirement is to stop the transportation of sand to the plains, the second is the formation of a forest soil under cover of a tree canopy. Owing to the geological formation erosion is mainly of the "sheet type'. Sand-rock disintegrates directly into the fine particles of which it is composed, rendering it peculiarly liable to removal by water and wind; for this reason the formation of soil is impossible except under the most favourable conditions. The presence of scattered trees and bushes has had little effect on surface erosion and none on soil formation. To build up a forest of trees and shrubs will lake a long time and it is essential to find a counter-erosive which can both rapidly and cheaply cover the huge area which has to be dlealt with. There is only one answer to this, it is grass. A close grown crop of grass has a considerable effect in reducing 'sheet' erosion. Sand-rock can support an excellent crop of grass which extends rapidly if closed to grazing. At present the large herds of cattle that roam over the hills effectively prevent grass from spreading; they destroy tree seedlings and at every step they break off and powder up the friable sand-rock, thus adding enormously to the amount of material to be transported by the floods. The first step is, therefore, to extend a system of closure over as wide an area as possible. Grass will not, however, by itself, solve the whole problem; it will not entirely prevent 'sheet' erosion, it has only a very moderate effect on run-off, it is useless as an agency in the formation of soil from sand- rock, and as a hydrological factor it is unimportant. In the end a vegetable covering of grass, bushes, and trees must be re-established from the plains to the watershed. The area is large and delay has increased the difficulties of reforestation; but the task is by no means impossible and, if carried out systematically, the work will not prove costly. The evil is the result of a mistaken policy; there is a debt owing to those who have suffered, and to those who will continue to suffer unless money is spent.

The chief obstacle to success is that the Siwaliks are the property of thousands of ignorant peasants who, although favoured by the gift of grazing-grounds far in excess of those ordinarily found in the Punjab plains, have, like most Indian villagers, consistently overgrazed them. The grass has become scanty and less nourishing, and the quality of the cattle has been sacrificed for quantity. In addition the villagers have not yet learnt that interference with their rights on the part of the Government will ultimately redound to their benefit. Other countries have had similar difficulties which they have faced by acquiring the land or buying up all rights. Apart from the heavy expenditure this would involve, there are other objections to such a policy here.

A scheme of reconstruction, working through the villagers themselves, and obtaining their co-operation, is what is required. They must realize that the Government, far from wishing to take away their rights, is prepared to act in the capacity of an agent, where protection is sought. The eviction of goats has extended the cattle- keeping habit-an improvement; now the villager must be taught to stall-feed his animals, feeding them on the grass which he will cut from the closures. There is no need to detail all the benefits that will result from this. Where closure was applied in 1902, the villagers are now making handsome profits from the sale of grass surplus to their requirements. Reclamation of devastated land must be encouraged so that there need be less dependence on cattle. Once the cho beds and slopes are forested there will be other sources of profit, firewood, bamboos, lac, and bhabar grass.

The landlords must be made to realize that there is more profit in fodder production than in grazing half-starved cattle, that their land can produce something more valuable than a few lopped trees and bushes. They will be enriched, and both the Government and their neighbours in the plains will be benefited. This may sound too optimistic, but there are hopeful signs, and if continuous and sympathetic encouragement is given to the willing, and pressure applied to the unwilling, the task is not an impossible one. Nature will take a part in reclothing the hills with vegetation, but artificial aid will be required. This is the work of experts and cannot be done by the villagers themselves. It will take long, and continuity of action must be assured by the adoption of a well-conceived plan.

Meanwhile, owing to increasing population, there is a demand for land on all sides. This can only be met by reclamation of the wasted land. Reclamation implies, first, the restoration of the land to a condition suitable for permanent cultivation and, second, the directional training of the chos. Success depends, ultimately, on the extent to which the rush of water and sand is reduced. Under present conditions eho training is possible only to a very limited extent. There are, however, thousands of acres of sterile and partially sterile land ready for reclamation. How is it that, in past years, the villagers have done so little to better themselves? The two main reasons are:

(1) Owing to loss of cultivation the people have tended to become more pastoral, and professional herdsmen have taken advantage of the increased opportunity for grazing, often at the expense of the real owners of the land.

(2) Reclamation fs a technical operation which requires organization and instruction; it cannot succeed under laissez-faire methods.

Reconstruction is again the key-note. Grazing must be reduced where necessary and replaced by cultivation and fodder crops. This implies closure and stall-feeding, a system of pasture management. The poor land must be manured; stall-feeding will help, and in the Jullundur district experiments with green manure have given good results. But huge areas have to be dealt with and nature's way is the best; the sandy wastes must lie under a crop of grass and trees for several years; it is believed that this is the only way that such soil can be rendered permanently cultivable. Other advantages are:

(1) Closure will indirectly improve the quality of the cattle, and the grass-crop, which can be improved and increased by sowing the most suitable species, will help to solve the grazing problem.

(2) The zamindars will derive an income from the sale of surplus grass and shisham trees, both of which grow readily in the sand.

(3) By leaving a few trees scattered over the fields and along the borders, the land, when brought under the plough, is protected from the drying effects of sun and wind.

This technique has been followed in many cases and has produced quite good soil, but too often good intentions have been spoilt by impatience or frustrated by the invasion of graziers.

All this the people can do themselves, but constant supervision and instruction are essential; some landlords who are unable to protect their property from the herdsmen have sought Government protection, and no doubt others will do the same. And as more land becomes fit for cultivation, the herdsmen, who are often tenants, will tend to give up their pastoral habits.

But reclamation by closure is going to do more than this; it will ultimately train the chos themselves. The surest way to form sandbanks is to fix the sand with a growth of vegetation. Reclamation and consolidation will proceed inwards from both sides as conditions improve in the catchment areas. Enterprising villagers have evolved a rough technique in cho training on a few minor branches. If this is improved upon and continuously and judiciously directed, the chos will gradually become canalized between tree-lined banks, the condition they were probably in before the trouble began.

I have dealt only with the Hoshiarpur Siwaliks. Only a very small portion of the Punjab foot-hills has been touched upon, but it is enough to demonstrate the urgency of a strong forest protection policy, in order to arrest an evil which is slowly but surely spreading throughout the length and breadth of the Punjab foot-hills.

Erosion in mixed clay and sand-rock strata

Erosion in mixed clay and sand-rock strata

Severe ‘gully’ erosion in soft earthy strata

Severe ‘gully’ erosion in soft earthy strata

Bare slopes of soft sand-rock: where run-off and 'sheet' erosion are greatest

Bare slopes of soft sand-rock: where run-off and 'sheet' erosion are greatest

Cho-wasted land: lining of well, 20 years ago surrounded by rich cultivation, now exposed by violent torrent action

Cho-wasted land: lining of well, 20 years ago surrounded by rich cultivation, now exposed by violent torrent action

Typical "Cho" Country

Typical "Cho" Country

Closure to grazing: dense grass on sand rock slope

Closure to grazing: dense grass on sand rock slope

A small cho, originally 60 yards wide, completely canalized by sand- banks fixed with vegetation

A small cho, originally 60 yards wide, completely canalized by sand- banks fixed with vegetation