A.-Absolute desert conditions brought about by erosion on a once-forested area in northern California. Every vestige of the topsoil and much of the subsoil are gone, and deep gullies have been cut into the soft basal rocks

B.-Intermediate stage of land destruction on Susquehanna soil having a stiff clay subsoil. The devastation has almost reached the stage at which any attempt at reclamation will be unprofitable C.-Properly terraced field in the piedmont region. Unprotected soil of this type depreciates rapidly because of both sheet erosion and gullying

with their supply of lime and fresh organic matter, some temporary increase of productivity necessarily will follow the recession of the water. The degree of this increase can not be estimated with much accuracy with the small amount of available data; but probably its tangible money value does not greatly exceed 75 cents or a dollar to the acre of cultivated land each year for several years following a flood. It must be remembered that the Mississippi delta lands, particularly the predominant "buckshot" soil (Sharkey clay), as well as many of the other alluvial soils of the country, are among the richest soils in the world.

THE DANGER OF AVERAGES

The effect of erosion is extremely variable from place to place, on varying soil and varying slope, with varying vegetative cover and method of land usage. Hence, the average depth of surface denudation that has been commonly computed from river discharges alone, means very little. It implies that the surface everywhere, on steep hillsides and flat prairies, on sand dunes, loam and clay, is being planed down at an equal rate. This is far from the truth. The estimate so often read that erosion is lowering the Mississippi Basin at the insignificant rate of 0.0028 inch annually, is not only too small as an average, but since erosion does not operate over large areas of varying soils according to any plan of averages, such a statement is dangerous both for its inaccuracy and complacency.

A most important thing to know about soil erosion is the rate of cutting away the topsoil, and after that the subsoil of the individual soil types, in those regions of the more vulnerable lands, such as the region of loessial soils, the region of the Susquehanna soils, the Knox, Marshall and related soil regions, the Cincinnati soil region, the Houston clay soil region, and the regions where Orangeburg, Decatur, Cecil, Dekalb, Reeves, Vernon, Putnam, Fairmount and numerous other soils are important.

The studies already made in connection with soil-survey work show that there are many types of erosion, due to many variants, that have to do with the process, chief of which are, (1) soil type, (2) degree of slope, (3) climate, (4) vegetative cover, and (5) method of usage. Some soils can be cropped with a fair degree of safety on slopes having a gradient up to about 20 per cent, such as some of the very porous gravelly soils of the chert ridges in the southern part of the Appalachian Valley. On the other hand, some soils can not be cultivated without steady decline due to erosion, even where the slope does not exceed 1 or 2 per cent. The Knox silt loam, for example, is such a soil. On this soil erosion goes on in all tilled fields where there is any slope whatever.

On some soils greatest erosinal damage is done by gullying; on most soils, however, greater wastage results from that slow type of erosion called sheet erosion. On the Cecil soils of the piedmont region deep, broad V-shaped gullies form and finger out rapidly, whereas on the Orangeburg soils, the sides of the ravines are more nearly perpendicular, and they extend by a process of caving, when the loose sand of the substratum is washed out, so that rapid widen

ing and head-on extension takes place. (Pl. 7, A.) On the Grenada soils of the "brown loam" belt, by reason of a compact subsoil layer peculiar to this group of soils, the washing extends more nearly equally in all directions, and rapidly invades broad areas of fine loessial land wherever the erosion has been neglected in its infancy. (Pl. 7, B.) On gravelly red land in northern California, where smelter fumes have annihilated the forests and destroyed almost every vestige of vegetation, extremely deep, narrow ravines have developed, which make travel over these areas difficult and even dangerous. (Pl. 8, A.) On soils like the Susquehanna, in which impervious heavy clay lies near the surface, the material of the cultivated soil is converted quickly into an approximate liquid condition during rains. This causes the surface substance to flow away rapidly. Following this skinning-off process, the exposed stiff clay is attacked by erosion and gradually cut to pieces by gullies that render the land absolutely unfit for further cultivation. (Pl. 8, B.) Numerous other variations of the manner by which soils erode could be given, but this will not be necessary for the purposes of this circular.

Although there is some erosion on most tilled and bare areas, and probably always will be, wherever water runs downhill, provided the soil is not frozen or protected by hard snow, the damage is greatest in the southern and central parts of the Temperate Zone and in the Tropics. So long as the ground is congealed freezing gives practically complete protection, save on those soils that "heave" badly. Slowly falling rains are everywhere much less destructive as an erosional agent than hard, beating rains. For example, no important effect of surface wash is observable on cultivated slopes of the Fairbanks silt loam, a wind-laid soil, in the Tanana Valley in northern Alaska, where the ground is frozen during eight or nine months and the light precipitation occurs almost entirely as drizzling rain and light showers.

By simple and well-known laws of mechanics the erosive power of flowing water increases enormously with increase of slope, but the destruction accomplished varies greatly with the soil type. Deep sandy soils, as a rule, do not wash severely, especially where the subsoil does not consist of impermeable clay or hardpan. However, some areas of sandy land, such as the Norfolk sand, do wash rather badly, and even gully on those slopes where there is impervious clay at a depth of 4 or 5 feet or less, as is true of areas in east Texas having a stiff subsoil like that of the Susquehanna clay.

RELATION TO FLOOD CONTROL

It is obvious that the erosional débris entering the streams adds to the volume of the water. It is equally obvious that those methods of soil conservation which have been found effective in slowing down or controlling soil erosion, chiefly terracing the land and the growing of trees, grass, and other soil-holding plants, are also methods which will cause more water to be retained in the surface soil and to be stored in the subsoil. Terracing of fields and the growing of trees, grasses, and shrubs on idle lands and areas too steep for cultivation, and upon soils that are highly susceptible to washing, as a combination of practices, will, it is believed, have considerable to do with flood reduction by decreasing the runoff and washoff from many

Soil conservation is somewhat synonymous with moisture conservation. Nothing will hold back all the water, of course, but enormous quantities can be held temporarily or stored for summer-crop use, especially in the subhumid regions. At the same time the rich topsoil can be conserved by these proved implements of soil and water conservation. Soil conservation, therefore, should be an important adjunct of any long-continued system of flood control. To those who have seen the water from heavy rains rushing down unprotected cultivated slopes and bare areas, surcharged with soil matter, and carrying even gravel, cobbles, and bowlders, it is not necessary to argue about the effective contribution widespread use of these soilconserving methods would make toward flood control as supplementary measures to protection with levees, spillways, and reservoirs.

Suspended material to the amount of 428,715,000 tons annually passes out of the mouth of the Mississippi River alone. This is but a part of the solid material that enters the river and its tributaries since much is left stranded somewhere along the pathway to the sea. In considering the relation of this water-transported erosional material to increased floods, it is necessary to take into account its full significance, along with that of a far greater amount stranded between the source of supply and the streams, in its relation to the increased amount of water flowing off the land areas which have contributed the material. So many tons of silt in the river stand, unmistakably, for so many denuded or partly denuded acres of sloping land somewhere upstream-land enabled by its denuded condition to contribute to the stream at a faster rate more of the rain that falls upon it.

In discussing the relation of forest and other forms of vegetative cover to run-off water and floods, it is frequently contended that, although the methods may have value, the time required for a forest. to grow up is too great for this means of assistance to have any important relation to flood problems requiring immediate attention. In this connection the fact should not be lost sight of that the roots of trees and of other plants begin to function as effective agents for holding soil against erosion very shortly after the seedling begins to grow. Greatest efficiency in this respect will come, of course, when the forest or other vegetative cover, as grass, bushes, and chaparal, has made sufficient growth to develop an absorptive, spongy cover of vegetable litter. The immediate effectiveness of grass in holding both soil and water has been conclusively shown by results of the erosional test referred to above. It is said by those familiar with early conditions in the Prairie States that before the extensive cultivation of the land the matted turf of the prairies, in many places, hung like canopies over the banks of streams that carried clear water throughout the year. With the breaking of the land this situation was changed. The streams are more frequently dry in summer and are more heavily laden with silt when the rains come.

The following relates to the effects of rains on sloping areas in Orange County, Calif., following removal of a bush growth by fire: "

During the Orange County Farm Bureau Forestry Tour on November 19th, a remarkable demonstration of the effectiveness of chaparral cover in conserving water by preventing destructive erosion was seen at the Harding reservoir.

Information furnished by C. F. Shaw, University of California, in a letter to the writer. Data obtained from Extension Service Report, December, 1927.

88854°-28-3

In October, 1926, the heavy chaparral cover on this watershed was almost entirely destroyed by fire, leaving the slopes unprotected. During November a heavy rain fell during a 24 hour period. Santiago Creek quickly became a turbulent mass of muddy water, containing over 60% solid matter washed from the burn. The Santa Ana River, which in the past had never had a peak flow of more than 8,000 second-feet during similar rains, showed approximately four times that at the height of the flood.

Harding reservoir was completely filled with rocks, silt and ashes from the burn, and a deposit of a half inch to an inch of this material was left over the entire bed of the Santa Ana River when the flood subsided. Other streams in the vicinity, where watersheds were untouched by the fire, showed scarcely any rise at all, and the water in all of them was clear throughout the storm. After many weeks of shoveling and washing, the capacity of Harding reservoir is less than one-fourth of its original volume.

LIMITED AMOUNT OF DATA AVAILABLE

In this country only a limited amount of information has been acquired concerning the rates of erosion on different soil types, the holding effect of terraces of different build or the possibility of reenforcing them with various stabilizers such, perhaps, as grass, shrubs, or vines, and the rate of alluvial deposition under varying conditions. Only three or four soil types of the many involved have had their susceptibility to erosion measured. It will be observed in reading this circular that little information other than estimates and observations have been given. This is because exceedingly little research work has been done on the subject. It is not known, for example, precisely what type of terrace or what degree of terrace slope is most applicable to the loessial soils of the Marshall, Memphis, and Knox series. It is known that some types of terrace have not given entirely satisfactory results on these peculiar friable soils of such exceedingly high silt content and such low content of clay to bind the silt. Possibly the Mangum terrace, if properly modified and given precisely the right slope, would effectively control erosion on these exceedingly vulnerable soils. Information greatly needed in connection with the problem of erosion should be made available through experimentation and research work as speedily as possible. If a particular type of terrace does not hold in one place and does hold elsewhere, the reason for the failure, as well as for its success, should be determined and the significance of the facts turned over to the farmers of the Nation in forms available for practical use.

As a Nation we are doing very little to abate the evil effects of erosion. Every one who knows anything about it admits the problem is a serious one, but few realize how very devastating is the wholesale operation of erosion. There is necessity for a tremendous national awakening to the need for action in bettering our agricultural practices in this connection, and the need is immediate. Terracing of sloping areas to prevent erosion has been carried on for a long time in the southeastern part of the United States. (Pl. 8, C.) Recently use of this method has extended across the Mississippi River and is being extensively and increasingly employed in Texas, Oklahoma, and Arkansas. The Federal land bank at Houston recently adopted the policy of requiring all vulnerable fields to be terraced before money is loaned on the land. The bank has employed an erosion expert, who, according to press dispatches, not only decides whether or not the property upon which a loan is asked needs ter