Skip to main content

Full text of "Soil erosion"

See other formats

Historic, archived document 

Do not assume content reflects current 
scientific knowledge, policies, or practices 

Issued J iyi< 






Expert in charge, Soil Water Investigations. 




Issued June 8, 1911. 






Expert in charge, Soil Water Investigations. 




Burean of Soils. 

Milton Whitney, Chief of Bureau. 
Albert G. Rice. Chief Clerk. 


Feank K. Cameron, in charge of Physical and Chemical Investigations. 
Jay A. Bonsteel, in charge of Soil Survey. 
Oswald Schreinek, in charge of Fertility Investigations. 
W J McGee, in charge of Soil Water Investigations. 


U. S. Department of Agriculture, 

Bureau of Soils, 
Washington, D. O., May 3, 1910. 
Sir : I have the honor to transmit herewith the manuscript of an 
article on Soil Erosion prepared by Dr. TT J McGee, of this bureau, 
and recommend that it be published as Bulletin Xo. 71, of the Bureau 
of Soils. 

Very respectfully, Milton Whitney, 

Chief of Bureau. 
Hon. James Wilson, 

Secretary of Agriculture. 




Agricultural duty of water 7 

The duty of soil 14 

Natural work of water 16 

Work of water in agriculture 24 

Abnormal work of water in agriculture 28 

Remedies for soil erosion 32 

Principles of treatment 32 

Treatment of the soil 33 

1. Deep tillage 34 

2. Mulching .* 34 

3. Fertilizing 35 

' 4. Seasonable plowing 35 

5. Draining 36 

6. Dust-mulching 37 

Treatment of cover 40 

1. Tree planting 41 

2. Grassing 42 

3. Nurse cropping 43 

4. Cover cropping 43 

5. Eradication of weeds 44 

6. Rotation of crops 45 

Treatment of slopes 45 

1 . Contouring 46 

2. Terracing 48 

3. Vineyarding 54 

4. Retain-walling 54 

5. Annular forestation 54 

6. Grading 54 

Treatment of water supply 56 





Plate I. Relation between slope and cover 62 

II. Dependence of slope on cover 63 

III. Gullying in an abandoned field 64 

IV. Flattening of slopes with removal of cover 65 

V. Typical effect of deforestation 66 

VI. Typical gullying at brow of hill 67 

VII. Typical old-field gully 68 

VIII. Erosion of steeper slopes on removal of cover 69 

IX. Advanced old-field erosion 70 

X. Advanced hillside gullying 71 

XL Skinning of the land 72 

XII. Typical road in southwestern Mississippi 73 

XIII. Typical road of western Mississippi . 74 

XIV. Interdependence of slope, cover, and water supply 75 

XV. Typical arid land erosion 76 

XVI. Gullying due to subterranean flow 77 

XVII. Typical old-field erosion 78 

XVIII. Old-field devastation 79 

XIX. Old-field erosion pushing into a grove 80 

XX. Old-field erosion checked by natural growth 81 

XXI. Effect of linear cultivation on rolling land 82 

XXII. A contoured and partly terraced field S3 

XXIII. Contouring with parallel or concentric balks S4 

XXIV. Contouring with meandering balks 85 

XXV. Contouring with inadequate balks 86 

XXVI. An ill-designed balk 87 

XXVII. Ill-designed contouring 88 

XXVIII. A terraced field 89 

XXIX. A terraced park 90 

XXX. Erosion stayed by terracing in China 91 

XXXI. Contouring and terracing in China 92 

XXXII. Reclamation by terracing in China 93 

XXXIII. Terraced paddy fields in Ceylon 94 


Fig. 1. Natural water supply of mainland. United States IS 

2. Farm level 52 

3. Hand level 53 




The soil is of three parts, one solid, another fluid, and the third 
gaseous. The solid part is made up of both organic and inorganic 
matter in fragmentary or granular condition; the fluid part is a 
solution consisting of water carrying more or less organic and min- 
eral matter; the gaseous part consists of air (nitrogen and oxygen) 
mixed with aqueous vapor, carbon dioxid, and other gases. The 
solid part forms the body and the fluid part the circulatory medium 
of the soil on which plants grow and animals live ; the gaseous part 
permeates the body of the soil and passes through it in a manner 
which is sometimes likened to breath. 

As a whole the soil of a country forms a unit or entity comparable 
to, although less complete and distinct than, the flora or the fauna. 
It differs from these in that it is suborganic rather than organized, 
and in that it contains a larger proportion of mineral matter; it 
simulates organized systems in that it has its own modes of action 
and self-perpetuation, and also functions in accordance with its own 
special properties. Like the flora and fauna, too, it may be defined 
by special types and classified in general groups; the types compris- 
ing more or less variable unit areas corresponding somewhat with 
organic individuals. 

The internal action or functioning of the soil goes forward chiefly 
through the agency of its fluid part ; this may be specifically denomi- 
nated soil-fluicl, a while the solid and gaseous parts are conveniently 
designated soil-body and soil-gas. In the absence of soil-fluid the 
soil soon becomes inert or dead, losing its suborganic character: in 
the presence of this circulatory medium it is vitalized, and its reac- 
tions are largely connected with the growth and decay of organisms. 

a The fluid part of the soil has been made known largely through notable 
investigations by Prof. Milton Whitney, chief of the Bureau of Soils, and several 
collaborators. Bulletin 9, " Soil Moisture," by Milton Whitney and Ralph S. 
Hosiner (1897), and Bulletin 10, "The Mechanics of Soil Moisture."' by Lyman 
J. Briggs (1897), contain the more important results of the investigations. 



Thus the functioning of the soil is correlative with the functioning 
of the flora and fauna ; the processes taking place in the soil being 
interdependent with those taking place in the plants, somewhat as 
these processes are interdependent with those taking place in ani- 
mals. Much of the substance of plants is taken directly and that of 
animals indirectly from the soil, and the growth of soil goes forward 
largely through the return of substance from plants and animals in 
a more highly differentiated or richer form; while the chief source 
of vital energy in the soil is derived from the growth and decay of 
plants and animals. Although the interdependence extends to all the 
materials and powers of both soil and organisms, it operates chiefly 
through the peculiarly potent substance, water, of which large quan- 
tities pass into the plants and thence into animals; and the vital 
energy of organisms, like that of soil, is measured by the circulation 
of their fluid portions which consist chiefly of water. 

As the efficiency of plants and animals is measured, by the pro- 
duction of substance and energy, so the efficiency of soil may be 
measured by the yield of plant product ; and in both cases the yield 
varies with the vital energy attending circulation. In most animal 
genera the circulation is fairly uniform throughout life, and the 
greatest efficiency is attained during a (secularly lengthening) period 
of maturity following a period of more rapid growth. Among most 
kinds of plants the circulation varies widely with the season, and 
the efficiency is at its best during the period of most rapid growth of 
highly differentiated substance, which occurs about the time of most 
active circulation. In soil the circulation depends largely on climate 
and season, and the efficiency in general varies with circulation, 
though most types of soil are capable of storing or conserving 
water in ways increasing and prolonging their efficiency. Other 
things equal, the internal work or functioning of soil is determined 
by its capacity for conserving water and conveying it to growing 

The ratio of the solid and fluid parts of soil varies widely. When 
the soil is in good condition, i. e., when it contains its optimum of 
moisture, the fluid ranges from 4 to 45 per cent (even more in mucky 
and peaty soils) of the weight of the solid, according to texture; 
e. g., if the productive layer averages a foot in depth the two parts 
average about 2,000 tons per acre, while the fluid ranges from 80 to 

c An excellent statement of the most acceptable modern view of the soil as 
an entity functioning in a fairly definite way appears in " The dynamic view- 
point of soils," by Frank K. Cameron (Jour. Industrial and Engineering 
Chemistry, vol. 1, Xo. 12, 1909). In the light of those properties of soil-fluid 
and other waters pointed out herein the internal work of soil might perhaps 
better be considered kinetic, which would be but a consistent extension of Dr. 
Cameron's conclusions. 


perhaps 900 tons per aere a . The solid part is relatively stationary, 
but the fluid part is in constant movement (and change of state from 
liquid to vapor) throughout the growing season or period of soil 
efficiency; and the water required for the optimum moisture will not 
itself suffice to produce a crop, nor will it even permit any yield 
whatever from most types of soil, unless replaced as it is exhausted 
by the growing plants. If properly cultivated and watered, an 
acre-foot of soil averaging 2.000 tons retains efficiency for centuries; 
but to be even moderately productive of crop plants it must convey 
to these plants fully 1J acre- feet of water, or an amount equivalent 
to its own weight, during each growing season. The average re- 
quired for ample yield from various soils is about 4VJ acre-feet, or 
something over G.000 tons per year. So the nominal ratio of solid 
and fluid parts of the soil (i. e., the ratio reduced to the yearly basis) 

a The quantitative view of water, except in smaller measures, is so new to 
thought that familiar units are lacking. Municipal and domestic water supply 
is generally expressed in gallons, irrigation water in acre-feet, stream flow in 
second-feet (or more accurately seconds-feet), and water for certain uses in 
the variable and indefinite miner's inch. There is urgent need of a unit appli- 
cable to the quantities commonly used for water supply, irrigation, and various 
other purposes. Moderate familiarity with the metric system would render 
convenient as such a unit the stere (equivalent to the kiloliter or cubic meter, 
the virtual basis of the metric system for capacity or tridimensional measure), 
which roughly approximates — much as the liter approaches the quart — the cubic 
yard in quantity and the ton in weight of water, while the kilostere approxi- 
mates 1,000 tons and an acre-foot. The kilostere is especially convenient in 
discussing the water supply of the United States in that it permits expression 
of the leading values in round numbers not too large for ready comprehension — 
the mean rainfall of 215.000,000.000,000 cubic feet totaling 6.000.000,000 kilo- 
steres, and its main derivative fractions being expressible in sixths of this 
total. For the present cubic feet, acre-feet, and tons may be employed. 

A few of the equivalents involved in the use of customary units for the 
measurement of water (reckoned at maximum density) follow: 
1 liter=1.057 quarts=0.264 gallon=61.023 cubic inches=2.205 pounds. 
1 kiloliter=l stere=l cubic ineter=l,000 liters=264.1S gallons=35.314 cubic 

feet=2,204.62 pounds. 
1 kilostere=1.000 kiloliters=264,180 gallons=35.314.45 cubic feet=0.81O7 acre- 

foot=l,102.31 tons. 
1 gallon=3.7S5 liters=230.972 cubic inckes=0.1336 cubic foot =8.34 pounds. 
1 cubic foot=2S.317 liters=7.485 gallons=G2.42 pounds. 

1 acre-foot=^43.5G0 cubic feet=1.2335 kilosteres=326,047 gallons =1,359.6 tons. 
1 cubic mile=147.197,952.000 cubic feet=4J(;s.207 kilosteres=3,379,165 acre- 

feet=4,594,656.258 tons. 
1 pound=27.6S cubic inches=0.4543 liter=0.12 gallon. 
1 ton=2,000 pounds=32.04 cubic feet=907.19 liters=239.66 gallons. 

The mean rainfall of mainland United States may be expressed as — 
215,000,000,000,000 cubic feet=6,08S,159,380 kilosteres=4.!)3r>.<;70.s<>!) acre-feet= 

1,460.6 cubic miles=6,711,000.000.000 tons. 
215,000,000,000,000 cubic feet±G.000.O<X),0<>o kilosteres±5.000,000,000 acre-feet± 

1,500 cubic inilesi7,000,000,000 ) 000 tons. 


is about 1:3; but since the solid part lasts for years or centuries, while 
the fluid part passes away, the actual ratio is much higher. The 
fluid part (or circulatory medium) of the soil is not pure water, but 
carries both mineral salts and organic substances in solution ; and the 
relative value of the solid and fluid parts in plant growth probably 
corresponds fairly with the strength of the solution, or one to several 
hundred. Pending exact determinations, it may be assumed that 
the strength of the solutions forming the fluid part of the soil and 
the ultimate ratio of the solid and fluid parts required to maintain 
efficiency are about equal and as something like 1 : 1,000. 

The water within the soil may be or may not be efficient in circu- 
lation (or in soil functioning) according to its quantity in relation 
to the soil texture; for with the quantity its condition may be said 
to vary from (1) static to (2) dynamic; i. e.. it may be either inert 
or active. The full capacity of a given soil for water ranges with its 
texture or porosity from some 30 per cent to over 50 per cent of its 
volume. This is conveniently called the water of saturation; it 
completely fills the interstices among the soil grains, displacing the 
soil-gas, and ordinarily moves hydrostatic ally under the impulse of 
gravitation; it impedes or prevents normal functioning of the soil 
and remains in a virtually static condition until the excess is removed 
by drainage or otherwise. The water required to form soil-fluid (or 
to furnish the optimum soil moisture) ranges with the texture of the 
soil-body from, say, 10 per cent for sand to 40 per cent for fine clay. 
The quantity suffices to form a film surrounding each soil grain in 
such manner as to permit capillarity to act throughout the mass, yet 
leave space for air (or soil-gas) within the interstices. Through sur- 
face tension these films tend to flocculate the finer soil particles and 
promote physical and chemical action both within the soil grains and 
between the soil-gas and the soil-body; apparently the films are the 
chief means of interchange between inorganic soil matter and grow- 
ing or decaying organic matter, and, though subject to gravitation, 
the water forming them moves mainly through capillarit}^ under 
stresses acting dynamically in the normal functioning of the soil. 
Probably the energy of internal action within the soil-fluid increases 
with the thinning of the films — i. e., with the diminution of the 
water — from the point of subsaturation at which capillarity begins 
to the indefinite point at which capillary contact is interrupted and 
the moisture becomes hygroscopic, so that functioning is most vigor- 
ous in a moist but drying soil. 

° Slichter computed, the porosity of aggregations of spherical grains to range 
from 25.95 per cent to 47.64 per cent of the aggregate volume ( " The Motions of 
Underground Waters," Water Supply and Irrigation Papers of the United States 
Geological Survey, No. 67, 1902, p. 20) ; King computed the porosity of soils to 
range from 34.91 per cent in coarse sand to 52.94 per cent in finest clay 
("Physics of Agriculture," fourth edition, 1907, p. 124). 


While the aggregate quantity of soil-fluid varies widely with differ- 
ent soils of varying texture, the limiting points of subsaturation and 
interrupted capillarity vary in a measurably corresponding way, so 
that an approximate estimate may be made of the soil-fluid available 
for plant growth in average soil. The basis of the estimate may be 
the 4 feet of soil and subsoil throughout which capillarity has been 
observed to operate freely , a for while ordinary annual crop plants 
root within the first foot from the surface, the underlying: 3 feet of 
subsoil forms a reservoir whence they derive much of the moisture 
required for their growth; indeed, it is probable that under certain 
conditions the moisture at much greater depths becomes available. 
Now, the mean moisture of average soil when in good condition ap- 
proaches 25 per cent, while the mean moisture when plant growth 
ceases by reason of exhaustion of the soil-fluid is probably less than 
10 per cent, & and the difference measures the store of water additional 
to the current rainfall on which the plants may draw. This differ- 
ence (15 per cent of 4 feet, or 7.2 acre-inches=816 tons per acre) may 
be considered as the effective soil-fluid of average soil. 

Much of the fluid conveyed by soil to growing plants evaporates, 
this change in the state of the water produced by the power of the 
sun constituting the chief motor force attending growth ; so that while 
the crops of a generation may far exceed the quantity of solid soil 
yielding them, the same crops form but a fraction of the quantity 
of soil-fluid utilized. A fair to good crop from an acre-foot of fertile 
soil supplied with 4^ acre-feet of water may be put at a ton of grain 
and 3 tons of dry forage (including stover or stubble), or 1 tons in 
all, i. e., yg^o °f the weight of the season's water. The yield of grain 
alone is much less, as shown in the accompanying illustrative esti- 
mates in which the mean ratio is 4 ^W °f the water. The yield of 
pasturage, green forage, fruit, timber, tubers, et al., on the other 
hand, is much greater ; the average of all crops in good farming may 
be put at 6 tons per acre-year, i. e., ^J ¥ the weight of the first foot 
of soil (solid and liquid), or approximately one-thousandth of the 
weight of the water received and conserved in the soil and largely 
conveyed to the growing plants. This may be considered the duty of 
water in crop production. So the agricultural duty of water may be 
defined as the production of one-thousandth part of its weight in 
useful plant crop. 

°King found that capillary lifting of water through fine sand diminished 
from 2.37 pounds per day at 1 foot to 0.91 pound at 4 feet, the diminution being 
less with clay loam ("Principles and Conditions of the Movements of Ground 
Water," Nineteenth Ann. Kept.. U. S. Geological Survey. 1899, Pt. II, p. 85). 

6 The mean of King's determinations of soil moisture " when growth is brought 
to a standstill" ("Physics of Agriculture," op. cit., p. 125) was 10.93 per cent 
for clover and 8.92 per cent for maize. 


Illustrative estimates of yields of grain with varying water supply. 

















Sums. . 

18, 360 
a 4, 590 



1 : 2, 980 



1 : 4, 500 




12, 120 


Averages . . 




a Equals 9,180,000 pounds. 

With present knowledge, the coefficient is but a rough approxi- 
mation. Measurements are vague and experiences variable; soils 
differ both in composition and in the texture controlling circulation ; 
and the yield of succulent vegetables or of juicy fruits or fresh for- 
age may be several times that of grain, nuts, or dry forage, so that 
while the final product (in nutrients, textiles, etc.) may vary less 
than the bulk and weight of the immediate product, it will probably 
be found needful in time to work out coefficients for particular crops, 
just as it is now convenient to reckon yields per acre in different aver- 
ages for the several crops. Still, if scientific methods are to extend 
to the farm, no inexactness in the coefficient or variability with dif- 
ferent crops can remove or reduce the need for recognizing some defi- 
nite relation between the water passing from soils to plants and the 
crop produced by means of this circulation. Already civil engineers 
recognize a duty of water in the development of power, and irrigation 
engineers recognize a duty of water in ditches or flumes or headworks 
supplying a given depth of water over a given area ; a yet it is vastly 
more important to the country and to the world for the farmer (or 
agricultural engineer)' to recognize a duty of water in terms of that 

a Powell long ago recognized the necessity for determining " the amount of 
water which is needed to serve an acre of land " ( " The Irrigable Lands of the 
Arid Region," The Century Magazine, vol. 39, 1890, pp. 770-771), and described 
this service as the " duty " of water measurable in acre-feet ; and irrigators 
have frequently applied the term to the measure of water rather than of the 
service (or duty) performed by the water — a service susceptible of useful meas- 
urement only in terms of what the water does or performs, just as the duty of 
an engine is measured by its performance and not by boiler pressure. Terms 
for the measurement of the water no less than of the service performed are 
indeed requisite. Prof. Fortier, in judiciously discouraging excessive use 
of water in irrigation, says, " We find that the average duty of water over two- 
thirds of a million acres of land was recently shown to be 4| feet per acre" 
(Proceedings, Seventeenth National Irrigation Congress, Spokane, 1909, p. 274) ; 
but it is needful to distinguish between the quantity of water required by a 
given soil and the service which that water renders in crop production. 


production which furnishes food for man and forms the foundation 
for all human industries and institutions. 

The water consumption expressed in the coefficient is, of course, gross 
rather than net, i. e., reckoned from the practical conditions of the 
average farm, rather than from the technical conditions and meas- 
urements of plants singly or in small groups. Such net coefficients 
as those worked out by King, Storer, and others for special crops and 
seasons are undoubtedly more exact ; a but they would seem to be less 
applicable to the practical agriculture of the country at large with its 
incidental and generally unnoticed wastes, and with the growing 
need for balancing yearly use with aggregate current supply. 

Naturally, the coefficient for plant yield will not apply to general 
farm production, including crops of meat, eggs, wool, hides, etc.; 
for not only do animals drink many times their weight in water 
annually, but they consume indirectly in their feed the equivalent 
of that much larger quantity required for the growth of the vegetal 
tissue of which the feed consists. The human consumption is still 
larger: In illustrative estimate, a pound of bread is the equivalent 
of 2 tons of water used by the growing grain, and a pound of beef 
the equivalent of 15 to 30 tons of water consumed b} T the animal 
both directly and indirectly through feed; and the adult who eats 
200 pounds each of bread and meat in the course of a year consumes 
something like 1 ton of water in drink and the equivalents of 100 

° King, in Wisconsin, found the mean weight of water used during the grow- 
ing season by barley, oats, maize, clover, peas, and potatoes in producing a ton 
of dry matter ranged from 270.9 to 576.6, and averaged 446.3 tons ( " Physics 
of Agriculture," by F. H. King, 2d edition, Madison, 1901, p. 139). Storer, sum- 
marizing numerous experiments made by observers in different countries on 
various kinds of crops, says : " More than 300 pounds of water pass through a 
plant, and are transpired by its leaves for every pound of dry solid matter 
fixed or assimilated by the plant" ("Agriculture," by F. H. Storer, 7th edition, 
New York, 1906, vol. 1, p. 15). In Idaho McPherson measured the water used 
on an experimental farm in 1906. and obtained the following ratios of water to 
crop : Alfalfa, 432.78 to 1 ; beans, 152.9 to 1 ; beets, 90.7 to 1 ; carrots, 77.18 to 1 ; 
corn of four varieties, an average of 135.75 to 1 ; oats, 90.86 to 1 ; potatoes, 
46.28 to 1; and wheat, 66 to 1 ("Third Annual Report of Alex McPherson, 
Director of Experiment Stations," dated Twin Falls, Idaho, Apr. 4. 1908) ; 
the measurements being made only during the growing months without allow- 
ance for accumulated ground water or natural subirrigation, and on the as- 
sumption " that the amount of water evaporated from a water-free surface, as 
shown by the evaporating tanks, was equal to the evaporation from the soil, 
the seepage, and the amount actually used by the plants " — an assumption un- 
doubtedly rendering the figures too low. All these determinations relate 
primarily to the plants themselves and have little reference to the total water- 
supply on which agriculture in the last analysis depends; they especially fail 
to include that soil moisture on which fitting texture of the soil depends during 
the nongrowing as well as the growing seasons, and of which there is generally 
a considerable loss through surface evaporation. 


tons in bread and 4.000 tons in meat, or 4,401 tons in all — besides the 
use in ablution of from 100 pounds to 200 tons (12 to 48,000 gallons, 
or from a gill to some four barrels daily), according to habit of liv- 
ing. These figures correspond fairly with current experience of 
intensive agriculture in the arid region, in which water is measured 
more carefully than in humid lands ; here a 5-acre farm supplied with, 
say, 5 feet of water, suffices for a family of five, or an inhabitant per 
acre (cities balancing more barren tracts) ; and on this basis the 
5,000,000.000 acre-feet, constituting the total yearly water supply of 
mainland United States, would suffice for a population of about 
1,000.000.000. which at the current rate of increase will be reached 
in some three centuries, i. e., a future span equal to that passed since 
the Pilgrims landed on Plymouth Rock. 


The measure of the solid part of the soil required to produce given 
crops varies widely with Avater supply, soil type, mode of cultivation, 
crop type, and other conditions. In ordinal farming the common 
annuals are sustained chiefly by the topsoil lying within plow depth, 
i. e.. nominally 4 to 8 inches, actually 2J to 5 or 6 inches. Especially 
in dry seasons, corn and some other plants send their roots so much 
deeper as to impede the next plowing, while alfalfa and certain other 
plants adapted to arid conditions send a part of their roots down to 
water even manj^ feet below the surface. On the other hand, the 
roots of corn and some other crops in wet seasons and in irrigated 
fields ramify near the surface; with dust mulching in the subhumid 
and semiarid regions, the roots tend to spread throughout the friable 
soil immediately beneath the pulverized moisture-retaining la}^er at 
the surface ; and the most exuberant native flora in the United States, 
including the lofty firs and rank undergrowth forming a veritable 
jungle throughout western Washington, roots wholly within depths 
of 18 to 24 inches, the minor root tips turning upward and terminat- 
ing where the abundant duff or natural mulch passes into the rich 
humus at the very surface of the soil. In a word, the rooting of crop 
and other plants is controlled chiefly b}^ the distribution of soil- 
fluid, i. e., by the soil circulation: and this circulation is but a part 
of a general system which includes the movement throughout the 
subsoil below and the plants above no less than within the layer of 
soil whence the plants draw their moisture. The chief role of the 
subsoil, indeed, is that of a reservoir of soil-fluid for the supply of 
the topsoil by which the crop plants, or the native grasses and shrubs 
and trees, are chiefly or wholly sustained ; and when the water supply 
is adequate this topsoil is rarely over 12 inches in thickness. So it 
is convenient to reckon soil, no less than water, in terms of acre- feet. 


The apparent specific gravity of properly cultivated soil averaging 
about 1.5, a cubic foot weighs between 90 and 100 pounds and an 
acre-foot weighs some 2.000 tons. An annual yield of 6 tons of (dry) 
plant tissue from such an acre-foot (assuming its extensions to bal- 
ance the part not fully employed) would thus equal 3J3 of its weight ; 
or (the ash being barely 5 per cent of the plant produced), say ^ l 6T 
of the solid soil matter presumably consumable by plants in time. 
Whether a given acre-foot of soil would endure for 70 centuries : 
whether its substance would be consumed and carried away : whether 
it would be enriched to any given degree by the differentiation of 
mineral constituents attending organic growth and decay: whether 
the material derived from water and air would counterbalance in part 
or in whole the loss of substance carried off in the annual crops — these 
are questions not to be fully answered with present knowledge. The 
broad facts of geology teach that the organic and the inorganic have 
interacted through the suborganic soil during the ago. and that in 
a general sense the successive floras of the world have (despite the 
loss of soil matter through erosion) accumulated and enriched the 
world's soils : yet geology has not yet yielded quantitative coefficients. 
Current agricultural experience commonly suggests that soils are 
impoverished by continuous cropping, in connection with the often 
neglected but appallingly large losses through destructive erosion, 
unless productivity is maintained by special treatment ; and the im- 
poverishment is commonly ascribed to the exhaustion of certain earth 
salts regarded as plant food. The more extended studies and com- 
parisons of scientific agriculture, on the other hand, reveal a pro- 
gressive increase in productivity with continued cultivation in this 
and other countries a (i. e., a trend parallel with that indicated by 
geology), which may be due in part to improvement in crop plant-, 
in part to fertilizing, and in part to better methods, but is certainly 
inconsistent with any view of progressive impoverishment of prop- 
erly cultivated soil. In any event, the foot of suborganic soil re- 
quired to produce annual crops is constantly changing: it is at once 
product and producing agent ; and if it loses toward the surface 
through absorption by the growing plants, it gains below in equal or 
greater measure by oxidation and disintegration of inorganic Mibsoil 
and subjacent rocks, so that the acre-foot of one century is not iden- 
tical with that of the last and may be quite different from that of the 
preceding millenium. Yet. since the farm is not limited in depth, 
the acre-foot of soil may be considered permanent, whatever the 
progressive internal changes due to the internal work attending the 
soil-plant circulation and other processes: and it- average duty may 

The data are summarized by Professor Whitney in "'A Study of Crop Yields 
and Soil Composition in Relation to Soil Productivity," Ball. 57, Bureau of 
Soils, Oct.. 1909. 


be defined as the annual production of ¥ J y of its weight of useful 
plant crop. 

This coefficient is of less consequence than that defining the agri- 
cultural duty of water, chiefly because soil efficiency depends pri- 
marily on water supply, partly because of the variability of secondary 
conditions. Thus, yield varies with depth of tillage, and with com- 
position and texture of the soil; it varies also with the contained 
moisture of crops ranging from succulent vegetables or juicy fruits 
and bulky forage to grain, nuts, and fiber; though with increasing 
knowledge the coefficient may be standardized by reducing the meas- 
ured crop to a uniform condition of dryness. 

Pending a more careful reckoning of the variables, the coefficient 
is useful chiefly in fixing the idea that in rational agriculture, no less 
than in mechanics and the physical sciences, cause and effect may 
and should be constantly balanced by measuring or weighing in com- 
mensurable units. Intensive methods have already shown that mul- 
tiplying depth of tillage is more profitable than multiplying the 
acres tilled, and the time is at hand for measuring production in 
terms comparable with both soil and cultivation. So, habitual 
thought in terms of any coefficient based on commensurable quanti- 
ties, rather than on random ratios between incommensurable units of 
weight or volume and area, can not fail to be useful. 


The sole original source of fresh waters on the lands of the earth 
is rainfall, including snow. In mainland United States (i. e., ex- 
clusive of Alaska and the insular possessions) the mean annual rain- 
fall has been shown by the Weather Bureau to average about 30 
inches. The total quantity is some 215,000,00a,000,000 cubic feet, or 
5,000,000,000 acre-feet (6,000,000.000 kilosteres, or 7,000,000,000,000 
tons) ; it is equivalent in volume to 10 Mississippi rivers running con- 
stantly. J A secondary source of fresh water is the supply derived 
from former rainfall and stored chiefly in the form of ground water, 
partly in lakes and ponds and running streams. The water thus stored 
within the first hundred feet from the surface is estimated at 
2,000,000,000.000.000 cubic feet (1,400.000,000,000,000 in the ground and 
600,000,000,000,000 on the surface, including the American part of 
the Great Lakes) or some 46,000,000,000 acre-feet (or 57,000,000,000 
kilosteres) ; it is equivalent to nearly 10 years' rainfall. The ground 
water alone would fill a reservoir conterminous with the entire coun- 
try to a depth of some 17 feet. 

Neither rainfall nor ground water is uniformly distributed; the 
mean annual rainfall ranges from less than 5 (locally less than 2) 
inches to more than 100 inches, while the soil wetness measuring the 


quantity of ground water ranges from the air-dry condition to com- 
plete saturation, or from, say, 4 per cent upward; i. e., from desicca- 
tion to drowning. Considered with reference to natural water sup- 
ply, mainland United States may be divided conveniently into the 
three sections shown in figure l, a viz. (1) the Eastward States, or 
humid section, comprising the two-fifths of the total area lying gen- 
erally east of the ninety-fifth meridian; (2) the subhumid section, 
comprising the one-fifth contained in the Median States (Xorth 
Dakota, South Dakota, Nebraska, Kansas, Oklahoma, and Texas) ; 
and (3) the Westward States, or semiaricl section, comprising the 
two-fifths west of the Median States, and generally beyond the one 
hundred and third meridian. The mean rainfall over the humid 
section is about 48 inches yearl} 7 , 5 or not quite enough for full pro- 
ductivity from average soils (the optimum being some 60 inches), 
aggregating nearly 140,000,000,000,000 cubic feet annually; the 
ground water within the first hundred feet probably averaged over 25 
feet at the time of settlement, but has been reduced perhaps 5 feet 
through injudicious industrial development. The mean rainfall over 
the Median States approaches 30 inches, or 40,000,000,000,000 cubic feet 
yearly; it is supplemented by natural subirrigation from the moun- 
tainous country farther westward to an amount estimated at 12 
inches in central South Dakota, and probably averaging 3 to 5 inches; 
the ground-water reservoir is sustained by this subirrigation, and is 
estimated as equivalent to some 25 feet in the first hundred. In the 
semiarid section the rainfall ranges from less than 2 to over 100 and 
averages about 12 inches, aggregating less than 40,000.000.000,000 
cubic feet; the soil ranges from desiccation to saturation, and the 
ground water probably averages 5 or G feet in the first hundred. On 
the whole, the water supply of mainland United States is hardly half 
that required for full agricultural production, and the distribution is 

a Explanation of Fig. 1. — Shows mainland United States, i. e., exclusive of 
Alaska and the insular possessions, classified as (1) Eastward States or humid 
section, (2) Median States or subhumid section, and (3) Westward States or 
semiarid section; showing also precipitation by isohyetal lines ranging east of 
the one hundred and twentieth meridian from 10 inches to GO inches, with 10 
inch intervals, and west of that meridian from 20 inches to 100 inches, with 
40-inch intervals. 

h This and other figures for rainfall are somewhat in excess of the means 
computed from the observations at Weather Bureau stations. They have pur- 
posely been made high, because it seems preferable, in aiming for practically the 
first time to show that the rainfall over even the more humid parts of the 
country is below the agricultural optimum, to err if at all on the safe side, i. e., 
to give the farm the benefit of the doubt: accordingly it is assumed that the 
precipitation over mountainous areas where the Weather Bureau stations are 
fewest commonly exceeds that, over lowlands, where the stations are more gen- 
erally distributed. 

77266°— Bull. 71—11 2 



such that about 1.000.000 square miles may be regarded as fairly well 
watered, while another 1,000,000 are sufficiently watered for about half 
the normal productivity, and the remaining 1,000,000 square miles 

of the country are so meagerly watered as to be practically unproduc- 
tive under existing conditions. 

The possible crop yield over the entire country may be estimated 
at something less than half the possible production from half the 


area with the same aggregate quantity of water equably distributed, 
or equivalent to full production from an area of say 1,300.000 or 
1,400,000 square miles. 

Of the total rainfall, about one-third flows into the sea through 
rivers and smaller streams; this is the run-off. A smaller fraction 
either enters directly into chemical combinations (largely through 
vital action) or else penetrates deeply into the earth to saturate 
hypogeal rocks and find its way into the sea through slow percola- 
tion : this is the cut-off. The greater fraction remaining is returned to 
the air by evaporation (largely after circulation through organisms), 
forming the fly-off; it augments the vapor-content of the atmosphere, 
and is partly reprecipitated and reevaporated over and over again, 
thereby aiding in the equable distribution of the natural water 
supply, though its transitions have not been fully traced or its move- 
ments finally measured. 

The slight variations noticed from season to season have given rise 
to widespread popular beliefs that the rainfall of particular dis- 
tricts is increasing or decreasing; but meteorologists and geologists 
generally have favored the view of secular uniformity on the evi- 
dence of rainfall records for series of years. Of late the influence 
of forests and other vegetal cover on soil-plant circulation has re- 
ceived critical attention. 5 while the extent of deforested and artificially 
wooded areas has so far increased as to afford a basis for instructive 
comparisons of records, some of which seem consistent with the well- 
known facts that in passing from the humid section toward the 
semiarid section, or from any other humid district to an arid one, 
there is not only a decrease in precipitation, but the waterways change 
from comparatively steady streams into sand washes generally dry 
but flooded by tumultuous torrents for a few hours or days after each 
great storm: and that the transition is gradual and attended by 
progressive transformation in the character of the flora or vegetal 
cover. This accords with the elementary principles of geology, and 
also with a fundamental physical law affecting all climatic conditions 
whereby widening of extremes is accompanied by lowering of means, 

°The water resources of the country are set forth iu greater detail in the 
Report of the National Conservation Commission (60th Cong., 2d sess.. S. Doc. 
No. 676). 1909, vol. 1, pp. 39-49: and in the Annals of the American Academy of 
Political and Social Science. Vol. XXXIII. 1909, pp. ."21-534. The principles 
involved are summarized in " Outlines of Hydrology." Bulletin of the Geological 
Society of America, vol 19. 190S. pp. 199-220. A fuller discussion of the norma] 
relations between water supply and productivity is reserved for later publication. 

6 A recent and noteworthy contribution to knowledge of the subject has been 
made by Prof. L. C. Glenn. " Denudation and Erosion in the Southern Appa- 
lachian Region and the Monongahela Basin," U. S. Geological Survey Profes- 
sional Paper 72, 1911. 


and vice versa. a Accordingly, since the rainfall in the humid section 
(averaging barely 48 inches) is below the optimum for vegetal growth, 
any such diminution as is shown by the records can only be regarded 
as indicating progressive desiccation of both the air above and the soil 
below the natural vaporizer formed b} T the vegetal cover, and hence 
as marking a stage in transition from fair productivity toward 
desert conditions. Over the Great Plains, once open prairie but now 
partly wooded, a converse transition is frequently suggested ; but the 
records thus far disclose no perceptible increase in rainfall since 
observation began. 

In the state of nature existing before settlement of the country, 
the three fractions of the natural water supply (fly-off, run-off, and 
cut-off) were closely balanced not only among each other but with 
the natural cover formed by the flora — indeed neither the whole 
world nor the marvelous balance between the heavenly bodies of sun 
and planet and satelite reveals any more delicate harmony than that 
arising in the interrelations between the movements of the waters 
and the features of the lands. The inorganic interrelations are rela- 
tively simple; they form the foundation of dynamic geology. Lyell 
long ago, and Powell later, proved that nearly the entire earth face, 
down to the minutest lineaments, was sculptured during the ages by 
running water, and that the same water carried the waste into seas 
to build later formations eventually forming newer lands. Except 
in the glaciated region and a few volcanic districts and inclosed 
basins, the land forms of mainland United States were shaped pre- 
dominantly by running water ; and even in the exceptional regions the 
final configuration was generally given by the same agency. When 
the slopes are steep the water works rapidly and chiefly on the sur- 
face; when the inclination is gentle it works slowly and largely be- 
neath the surface; in the one case the rainfall runs off quickly, in 
the other case it either lies long for evaporation or sinks into the 
soil to reappear as springs and seepage. So the hills are cut down, 
the valleys are deepened more slowly, the plains are lowered more 
slowly still, and sometimes caverns and sinks are produced by sub- 
terranean flow. 

The general process is erosion; it involves (1) corrosion or weath- 
ering of the surface, (2) corrasion or scouring of channels, and (3) 
sapping or undermining of banks and strata, with (4) transportation 

a The empiric law was brought out iu a discussion of " Maximum Synchronous 
Glaciation," Proceedings American Association for the Advancement of Science 
for 1SS0 (vol. 29, 1881), pp. 469, 482^S5. Although it may be formally ex- 
plained as among the many cases of arithmetic means of geometric variables, 
the relations undoubtedly express innate properties of water as a temperature- 
controlling agency. 


or removal of the materials in (a) solution, (b) suspension, and (c) 

The rate of erosion varies in a geometric ratio with the declivity 
or slope of the surface, so that it is many times more rapid in moun- 
tainous and rolling regions than over smooth plains; it also varies 
geometrically with the size and swiftness of streams, so that by far 
its greater part attends storms^/the secular rate of erosion of the 
lands of the earth commonly accepted by geologists ranges from 
23 utt t° eoVo of a foot yearly, or a vertical foot in from 2,500 to 6,000 
years; the latest estimate for mainland United States, made by 
R. B. Dole and H. Stabler, of the United States Geological Survey, 
for the Xational Conservation Commission, is ¥T Vo-of a foot GoVoo 
of an inch) yearly — the amount of material removed being com- 
puted at 783,000,000 tons annually. 6 The several figures give some 
indication of the rate at which the earth face is sculptured into hill 
and dale, butte and gorge, knoll and swale, and all the varied forms 
of landscape and countryside. All have been carved out of older 
earth-forms during the ages by the removal of hundreds, thousands, 
and even ten-thousands of feet of rock above the present surface; 
and the work goes on and on. c 

a The several processes are described and the last-named is defined somewhat 
fully in '• Outlines of Hydrology," op. cit., p. 199, where it is suggested that 
the movement of the finer no less than the coarser sediment, and even of the 
water particles, is saltatory. 

6 Op. cit., vol. 2, p. 130. Although lower than earlier estimates, this deter- 
mination, resting as it does on extended observations and measurements, is 
undoubtedly the most trustworthy thus far made. Still it should be noted that 
since it is based on measures generally made under average conditions, and is 
hence an arithmetic mean of geometric variables, it is probably too low. Other 
factors corresponding, the erosion of a vertical foot in 6.000 years would give 
1,190.000,000 tons, and in 2,500 years 2,856,000.000 tons, of material removed 
yearly. Doubtless the discrepancy is due to the making of earlier measure- 
ments in flood, during which more sediment may be moved in a few hours than 
at ordinary stage during the rest of the year. In any event, the recent deter- 
mination may not be regarded as indicating progressive diminution in erosion, 
which is known through common observation to increase with settlement and 
injudicious cultivation. 

c Current knowledge concerning the development of land forms was gained 
chiefly through the work of the Geological Survey in conjunction with a num- 
ber of State surveys and several universities, notably Harvard, Wisconsin, and 
Chicago. The interpretation and definition of land-shaping processes ranks as 
the most important work of the federal survey, if not as America's greatest 
contribution to systematic knowledge; indeed, it marks an epoch in scientific 
progress. The data are distributed through hundreds of publications, are in- 
corporated in text books and taught in schools, and are rapidly entering into 
common knowledge. Perhaps the most convenient summary is that recently 
published by the Geological Survey as Professional Paper 60, "The Interpre- 
tation of Topographic Maps." by R. D. Salisbury and W. \V. Atwood. While 
the researches and publications form a firm foundation on which the art and 


Resting on the sure foundation of dynamic geology, the more in- 
tricate interrelations between land forms and natural or artificial 
cover become clear ; and here geology and agriculture meet. As the 
slopes of the land flatten under secular erosion vegetation appears 
and spreads in ways eloquent of the pertinacity and persistence of 
vitality, gradually forming soil; and thenceforward vegetal cover 
snd suborganic soil cooperate to check erosion, convert free run-off 
into fly-off and cut-off, and utilize the rainfall for the maintenance 
and increase of living things. With the advent of organisms and 
soil the physical laws of land sculpture are modified or masked, if 
not nearly nullified; for once a region is mantled with soil covered 
by a flora, mechanical reaction to external force is largely replaced 
by chemico-vital interaction responsive to the internal forces mani- 
fested in circulation and "growth. The efficiency of different plant 
forms (themselves reacting primarily to water supply and sec- 
ondarily to soil) varies widely; sage and greasewood and other arid 
plants barely hold ground against wind and storm ; the grasses form 
sward shielding the soil against rains and rivulets on gentle slopes 
though not on steep; while dense forests and close shrubbery break 
beating rains into mists and lead the trickles through a carpet of 
mulch into topsoil and on into subsoil over steeper no less than 
flatter slopes. It is in this final relation that the fullness of natural 
harmony attains perfection, and the organic and suborganic and 
inorganic are, as it were, attuned to a common measure of which the 
theme is the thrill and throb of life on the land — a harmony no less 
delicate and far worthier of appreciation than that of the cosmic 
spaces. And so practical is it that the trained eye of the geologist 
(or geomorphist, in stricter definition) sees at a glance whether the 
lines and contours of the fields before him were originally wooded 
or grassed or sheltered only by scattered shrubbery; for the slopes 
reveal not merely the work of the waters during the ages but the 
modification of that work first through the varying forms of plant 
and animal life and finally through the soils produced by the organic 

While a fuller analysis of the normal role of water in the func- 
tioning of soils remains for later publication, it is needful to note 
the primary relations between (1) moving water in its three forms, 
(2) the soil and subsoil accumulated during the ages, and (3) the 
vegetal cover; for the cover is the natural or artificial crop grown 

the science of the farm may forever rest secure, the leading applications have 
thus far run chiefly toward mines and mining; yet it is reasonable to forecast 
that as the applications are turned toward farms and farming they will be 
found still more useful, perhaps in even higher ratio, than that of agricultural 
production to mineral production, or as more than 4 : 1- 


from the soil through the work of the water. Stated in terms of 
active work (or function), the leading relations are: 

(1) It is the proximate function of water, as the chief agency of 
geologic process, to grade and shape the lands of the earth. 

(2) It is a function of plants (in cooperation with other organ- 
isms) to decompose and disintegrate the inorganic rocks of the earth, 
thereby producing soil with its complex mineral salts and organic 
compounds available for plant food. 

(3) It is the function of soil (in cooperation with organisms) to 
sustain plants, and especially to conserve water for their use and 
supply it as needed for their growth. 

(4) It is the mediate function of water to initiate and maintain 
the vitality and cumulative power of organisms by means of cir- 
culatory systems — systems at once extending through soil and plants, 
.involving change of state from liquid to vapor, and accompanying 
stresses (both tensile and compressive) sufficient to accelerate or 
retard, if not to cause or prevent, the reactions involved in organic 
growth and decay. 

(5) It is the ultimate function of water to regulate the tempera- 
ture of the earth by means of its own distribution and changes of 
state in a comprehensive circulatory system (at once secular and 
telluric) including its own movements and changes of state no less 
than the shapement of the land and the development of both organ- 
isms and the suborganic soil. 

Of these functions the second, third, and fourth are directly in- 
volved in the current work of water in the form of soil-fluid through- 
out habitable lands; the first and fifth are involved indirectly and 
more remotely. Now, in a state of nature the functions are coor- 
dinated, or adjusted one to another, in such manner that throughout 
lands lying at moderate height above the sea nearly all (in arid 
regions all) the water supplied by ordinary rainfall is used in the 
growth of plants and related organisms, in tempering the air by 
means of aqueous vapor, or in producing and enriching soil — or in 
two or all of these processes combined. With the normal balance 
found in a typical forest or on a well-grassed plain, the vegetal 
cover holds the storm water until it sinks into the duff or mulch 
and on into the soil and enters the soil-plant circulation; save in 
extraordinary storms the streams run clear because supplied rather 
by springs and seepage than by surface run-off; while the seepage 
water percolating through the topsoil and subsoil serves partly as 
a reserve against drought and partly as an excretory vehicle to carry 
off injurious by-products (such as growth toxics and the excessive 
alkalis of ill-drained areas) and sweep the worst of them through 
the streams into the sea. So the natural valley or drainage basin, 
from upland rim to lowland channel, is in proper sense a complete 

24 soil erosion. 

water system, involving supply, distribution, use, and sewerage. In 
the final balance toward which nature works, both gentle rains and 
ordinary storm waters are utilized in organic growth and in soil 
making, and only the slight excess required for sewerage flows sea- 
ward : each acre takes care of its own, both in water and in life ; 
and all the acres cooperate in using all the rainfall, and in the use 
transform much of it into vapor for the benefit of all. This ideal 
balance may seldom be attained: yet it is often approached — as it 
was in much of the Mississippi Valley before the relations were dis- 
turbed by settlement. 


The basis of agriculture in mainland United States may be consid- 
ered as the first foot of soil over an area of nearly 2.000.000 square 
miles, or about a billion and a quarter acres; it weighs fully 
2.500.000.000.000 tons; and it is capable, when adequately supplied 
with water, of yielding perhaps 7.500.000.000 tons yearly of plant 
product (including the field staples and also the woods, nuts, berries, 
pasturage, browse, mast, and mulch of woodlands), i. e.. many hun- 
dred times the current useful yield. Such a yield would suffice for 
a population averaging 500 per square mile over the 2.000.000 square 
miles of productive area, or 1,000.000,000 in the aggregate, or all the 
water supply could sustain at present standards. This soil and yield 
form the agricultural capital of the country, the primary source of 
food and apparel and industrial growth. 

The processes of agriculture supplement and extend the processes 
of nature. They progressively reconstruct the natural cover and 
modify the soil, and thereby initiate changes in the natural forms of 
the land which affect the natural balance established by the running 
and circulating waters. The processes are often injurious to the 
soil: though when conducted judiciously they may easily be made 

In primitive agriculture the natural balance of land forms and 
soil and cover is disturbed but slightly and slowly. Most of the 
American aboriginal tribes were agricultural in that they seeded and 
harvested sufficiently to supply a considerable part of their provender 
and apparel; yet they neither plowed the ground nor cultivated the 
crop in proper sense, merely planting local patches in a manner gen- 
erally simulating natural arrangement and leaving the production 
largely to natural processes — pursuing the methods of hacklxtu rather 
than ackerbau, as expressed by German students, or of hoe work 
rather than fieldwork. True, corn and tobacco and some other plants 
were so completely artificialized during the centuries of aboriginal 
cropping as to find their way into use in foreign countries soon after 
Columbus and Cortez came into the New World ; yet the chief process 


was selection of seed rather than cultivation of the soil or protection 
of the plants. So the aboriginal agriculture had little effect on the 
natural harmony. 

Certain customs of the hunting tribes were of greater effect, not- 
ably the burning of woodlands often ascribed (albeit erroneously) 
to a design of enriching the range for game by extending the prai- 
ries; as a matter of fact the aboriginal huntsman neither reasoned 
from cause to effect nor exercised the inventive faculty, since he had 
not reached the stage of development to which these powers pertain ; 
but with animistic faith instead he invoked the power of fire as a 
deific agency to aid in the chase, and in hunting orgies akin to the 
warpath obsession he slaughtered ruthlessly the game animals crazed 
by the flames — whereby the prairies were indeed extended, though 
the areas were generally reduced in game-bearing capacity. This 
aboriginal fire clearing was hardly less shortsighted and recklessly 
destructive than that of later agriculture at its worst, in that its 
motive was extravagant feasting for the day without thought for 
the morrow's famine ; over considerable tracts it indeed removed the 
natural cover, yet the burned areas were not extensive enough to 
affect materially the natural balance between the rainfall and the 
slopes and soil and cover of the country as a whole — for the pre- 
Columbian population of what is now mainland United States was 
not over 2,500,000 (perhaps not more than 1,250.000), or much less 
than one per square mile. 

In the effective agriculture following white settlement, the natural 
cover and surface were progressively artificialized. At the outset and 
for long after there was little effort to define or measure, much less 
to modify or improve, the delicate balance between those natural 
factors involved in shaping the surface and forming the cover; land- 
were chosen for clearing and breaking largely on the evidence of 
fertility seen in luxuriant growth, partly on account of smoothness 
or other qualities indicating suitability for plowing and seeding and 
harvesting; and since the smooth and fertile lands were relatively 
little affected by the artificial change, the natural balance was not 
seriously disturbed. So the liquid and solid parts of the soil re- 
mained about as before, the soil-plant circulation extended through 
the vascular systems of the artificially chosen plants much as it did 
through those selected by nature through survival of the fit, and the 
plow-stirred topsoil took partial place of the natural mulch as a 
sponge for retaining rainfall and reducing run-off. The chief imme- 
diate change was the forcing of a land naturally bearing a fine flora 
to yield still finer products — two heads of grain were grown where a 
blade of grass grew before, luscious fruits or pliable fibers were sub- 
stituted for bitter shrubs, and the profitless acres were made to teem 
with material for food and clothing ; and under irrigation a hundred 


heads of grain replaced the blade of buffalo-grass, and a hundred 
head of kine grazed where an antelope or two wandered before. 
Meantime unnoticed changes in the natural balance were advancing 
cumulatively, especially in the long-settled districts; and as the 
heavily wooded and more rolling lands were cleared, the disturbance 
became more and more manifest — and this in several ways : 

(1) With ordinary plowing and common crops the natural mulch 
soon disappeared and the humus diminished, so that the soil grew 
harder and poorer; and when storms came there was more surface 
run-off and less water soaked into the topsoil and on into the subsoil 

(2) With crops covering the surface imperfectly and for but part 
of the year, the soil was subjected to evaporation directly rather than 
through the growing plants ; and some of the soil-fluid, the very life- 
blood of the farm, was sacrificed. 

(3) When the meager and temporary cover of common crop plants 
replaced a luxuriant and perennial forest cover, the unbroken rain- 
drops smashed the soil into slime at the surface and tamped down 
the earth beneath; and as the natural sponginess diminished, more 
water ran off the surface, bearing a burden of sand and silt, whereby 
it was enabled to corrade rills, quickly widening into gullies. 

(4) When snows came they lay on bare surfaces (despite the slow 
bottom melting which normally absorbs snow sheets on ground to 
which they are naturally adjusted) until thawed by warm winds or 
rains, a when the thaw water went off in destructive freshets instead 
of soaking into and softening the thawing soil as in that state of 
nature wherein melting at the bottom is supplemented by the melt- 
ing through the reflected rays from trunks and shrubs and leaves. 6 

a A recent observation by Prof. J. T. Rotbrock, made in Cbester County. Pa.. 
during a tbaw following a beavy snowfall and bard freeze, is typical : " In tbe 
woods, wbere leaves covered tbe ground. I found tbat it was possible to tbrust 
an iron-sbod cane without difficulty to a deptb of 18 incbes into tbe eartb, 
unless it was stopped by a root or stone. * * * On tbe open ground, wbetber 
tbe snow still remained or bad drifted away, tbe resistance to tbe tbrust of tbe 
cane was solid, almost as if I bad struck a rock. To tbis tbere was but one 
exception, wbere tbere was an unusually dense covering of long grass. Under 
A matted surface of tbis kind I could still tbrust my cane into tbe ground of 
an open field." (American Forestry. Vol. XVI, 1910. pp. 349-350.) 

6 Tbe practical aspect of tbe diatbermancy of snow and ice. long ago demon- 
strated by Sir Jobn Tyndall (Heat as a Mode of Motion. Lecture IX. Mar. 20. 
1S62; edition of 1871, pp. 322-325), is too little understood. Ordinarily tbe sun 
is witbout appreciable effect in melting snow or ice, for all of tbe solar rays 
capable of affecting H 2 in tbe solid form are, as it were, sifted out by tbe 
H : in gaseous form (or aqueous vapor) diffused tbrougb tbe circumjacent 
atmospbere. In order to become effective tbe rays must first impinge on some 
otber substance, wbereby tbey are broken up and rendered capable of absorp- 
tion by water in its tbree forms. It is by reason of tbeir diatbermancy to 
solar rays tbat snow and ice seldom even soften at tbe surface in cold and dry 


(5) With the increase of run-off and fly-off. less of the water fur- 
nished by rains and snows sank into the soil, the natural drainage 
and sewerage effected by the ground water was impaired, and the 
accumulated reservoir of ground water was progressively reduced. 

The general result of these specific tendencies is well known. Most 
pioneer homesteads were located by springs, of which by far the 
greater part have failed ; in the next generation the households were 
supplied by shallow wells, of which most have gone dry unless 
greatly deepened ; well-remembered trout brooks have ceased to exist, 
while many tamarack swamps and hundreds of prairie sloughs have 
shrunk or disappeared; numberless bosky dells and shady reaches 
of clear river are gone, leaving in their stead freshet-swept gorges, 
running dry in summer : in certain districts old-field slopes are gullied, 
and even the new-cleared patches wash quickly, and often the soil 
soon turns hard and lumpy. It is estimated that throughout an 
area of some 500.000 square miles in eastern United States the natural 
level of water in the ground has been lowered 5 to 30 feet since the 
country was settled; and that the depletion of this natural reservoir 
can hardly be less than 700. cubic feet of water, 
equivalent to more than 3 years' rainfall for the entire country or 
fifteen years' rainfall over the area depleted; i. e., an immense capital 
has been squandered thoughtlessly. 

The artificial disturbance of the natural balance between cover, 
soil, and slope with given rainfall brings out clearly the normal role 
of water in agriculture, which is precisely parallel to that enacted in 
nature. The ordinary practice of the farm is measurably to reinstate 
the mechanical conditions of land sculpture existing before the ad- 
vent and extensions of organisms, under which conditions the run- 
ning water tends to erode the slopes, cut clown the hills, clog the 
valleys, and generally flatten the land surface to that natural 
hydraulic grade known in geology as the base-level of erosion.* At 

air, however strong the sunshine, though they quickly melt under a warm wind 
or rain (perhaps to refreeze and form a crust during the night) ; while the 
rays penetrate the mass to the dark earth below, which warms quickly and then 
by conduction and radiation thaws both snow and frozen soil. It is for the 
same reason that leaves and twigs lying on snow and ice soon sink deeply in 
pits, preserving their outlines: that tree trunks melt the snow about them until 
they rise from wells sometimes several feet deep : and that under natural forest 
conditions the heaviest snow seldom forms freshets, but thaws progressively 
at bottom and about the trunks and along fallen stuff in such manner as to 
soak into the mulch and hasten the melting of the frost in the ground below. 

a In dealing with the natural distribution of water it is necessary to recognize 
three natural levels or horizons. The first of these is the natural surface of 1 lie 
continent, including both the lands and the inland waters. The second is the 
'• base-level of erosion " recognized and defined by Fowell in the course of his 
classic work on the geology and geography of tbe Rocky Mountain region: it 
applies to each and all points at which the altitude is such that while water 


the same time this short-sighted interference with natural relations 
emphasizes nature's lesson, and indicates ways in which methods may 
be so modified as to keep the mechanical conditions and powers under 
complete control. 

Summarized in a preliminary way, the requisites are (1) to retain 
mulch and humus, partly to hold the waters of rains and snows, 
partly to temper and so increase the friability and sponginess of the 
soil; (2) to till deeply in order that the soil-body may be kept open 
to free circulation of the soil-fluid; (3) to select crops partly for the 
sake of affording cover for the longest practicable portion of the 
year, especially on steeper slopes; (l) to rotate crops partly for the 
sake of alternating long and short seasons of cover and so checking 
any erosion started by defective cover; (5) in the dearth of vegetal 
mulch, to maintain a dust mulch tending to check evaporation from 
the surface and thus to promote circulation through the crop plants ; 
and (6) on all steeper slopes to plow and plant on contours, thereby 
closing gullies and channels and retarding run-off. 

These and related devices suggested by the natural balance between 
cover, soil, and slope as connected with rainfall lie at the foundation 
of judicious agriculture. Properly developed and applied, they may 
be made to preserve and perfect organic control of the inorganic, to 
maintain the interadjustment of the solid and fluid parts of the soil, 
to obtain the full agricultural duty of water, and to prevent waste 
of the natural water supply with the attendant direct loss and indi- 
rect injury. The end of effort in the artificial applications agrees 
with that toward which nature works; when each acre retains and 
makes good use of its entire rainfall, then and then only is the adjust- 
ment perfect. 


Since it is the function of soil-fluid to circulate through the inter- 
stices of the soil-body and the vascular structures of plants, any water 

will flow along its natural lines leading seaward, the slopes are too low to give 
velocity sufficient for erosion (i. e., along the seashore it coincides with sea 
level and thence slopes upward inland at a rate varying with the size and dis- 
tribution of the rivers and minor lines of drainage) ; so that the base-level of 
erosion may be conceived as a surface within the earth, generally lying con- 
siderably below the natural surface, but conforming broadly with its irregu- 
larities. The third horizon may be denoted the ground-level of water; it is 
the upper level or surface of the natural ground water, sometimes called the 
water table, i. e., the level at which the earth is normally so saturated as to 
supply well water and springs or seepage. In clear streams or points supplied 
by seepage the three surfaces coincide; generally the ground-level of water 
lies somewhere between the natural surface and the base-level of erosion; 
although in arid regions the ground-level may be discontinuous or absent, 
while the base-level is determined by lines of flow leading not to the sea, but 
to loci of natural balance between storm flow, evaporation, and soil absorption. 


in the soil departs from its normal function and works abnormally 
(1) when it evaporates from the ground, whether or not it carries up 
earth salts to be precipitated at the surface; (2) when it evaporates 
through weeds or noxious plants; (3) when it accumulates in quanti- 
ties so excessive as to prevent circulation ; (4) when it drains quickly 
downward to depths beyond the reach of plants; and (5) when by 
reason of hardness of the ground and steepness of the slope it fails 
to permeate the soil-body and runs off at the surface. 

Of these abnormal tendencies the last is by far the most common. 
It is also the most serious, not only because of wider extent, but by rea- 
son of far-reaching consequences. The chief consequence is destruc- 
tive soil erosion, one of the gravest evils confronting the American 
farmer. It was shown by the National Conservation Commission on 
the basis of estimates received from 30,000 farmers, representing 
every county in mainland United States, that 16,597 square miles 
of farm land have been abandoned, and that 6,076 square miles, or 
0.2 per cent of the entire area, have been devastated by soil erosion. 
It was pointed out that — 

In districts liable to extensive soil erosion the abandonment of fields is dis- 
astrous; in some cases the old-field erosion not only removes the soil proper, 
but carries away the subsoil, and even the surficial deposits, exposing bare rocks 
or "intractable formations, over which soils naturally redevelop with extreme 
slowness, and can not be extended artificially except at large cost. The fact 
that over 6,076 square miles, or 3,888,640 acres, of our abandoned fields have 
been destroyed in this way is appalling. Not only would the area form nearly 
100,000 farms, capable of sustaining a population exceeding that of any one of 
our 12 least populous States, but each gully starts others in such manner as 
continually to extend the devastation. The evil should be remedied without 
delay. Communities and States should be awakened to the sacrifice of public 
interest through old-field erosion. First, in connection with abandoned fields, 
and progressively in cultivated fields, soil wash should be considered a public 
nuisance, and the holder of the land on which it is permitted to occur should 
be held liable for resulting damages to neighboring lands and streams. 

It was also shown (as noted above) that the rivers annually dis- 
charge into the sea over 70,000,000,000,000 cubic feet of water, carry- 
ing in solution and suspension no less than 783,000,000 tons of sedi- 
ment, of which a great and rapidly increasing part is derived from 
the erosion of farm lands; and the annual loss due to this cause was 
estimated at 7 to 10 per cent of the product of upland farms. 

Considered with special reference to the farm, destructive soil 
erosion comprises (a) comminution (including both disintegration 
and decomposition) of organic and inorganic material in excess of 
the rate required for production, (b) leaching or solution of such 
material beyond the norm of soil-plant circulation, and (c) washing 

° Report of the National Conservation Commission (60th Cong., 2d sess., 
S. Doc. No. 676), 1909, vol. 1, p. 79. 


of the material comminuted or leached (the three minor processes 
forming the corrosion of the general process defined on p. 20), to- 
gether with corrasion, in which firmer materials are scoured and 
crushed or dislodged by the impact of solid particles dropped or 
driven by moving water, and the transportation of all the materials 
down the slopes and toward the sea. The several processes are largely 
interdependent, and cooperate in a cumulative way ; clear water will 
not corrade, and the capacity of silt-laden water for corrasion (e. g.. 
in gullying) increases, other things equal, in a geometric ratio with 
the sediment carried; washing is of little effect without comminu- 
tion and leaching, and generally increases geometrically with the 
material supplied in these ways; while leaching and comminution 
doubtless interact, each aiding the other, and each aided, first, by the 
washing, which lays bare the surfaces of particles, and then by cor- 
rasion, which lays bare larger surfaces. 

The cumulative character of these processes is clear to any observer 
of old-field erosion during a storm. When the rain begins to fall the 
drops patter on the surface, breaking into spray at first absorbed by 
the air-dry earth ; as the surface moistens the spray gathers in a film 
surrounding and softening the external grains, and the continued 
patter disintegrates the softer particles and quickly converts the 
watery film into thin slime. Xow, if the surface is level or rough 
the slime lies stationary until the water soaks into the soil and the 
ground below : but if it is inclined and smooth the slime slips slowly 
down the slope, soon gathering in rivulets gaining swiftness as they 
run. Here the action changes according to the aridity or humidity 
of air and earth : in the arid region the water is absorbed or evapo- 
rated more rapidly than it gathers sediment, so that the rivulets are 
quickly dammed by their own debris and the flow is rediffused into 
a sheet-flood of slime, c while in humid lands the absorption and 
evaporation are so much less that the rivulets enlarge and gather 
larger granules to grind against the bottom and sides and cut rills 
of rapidly increasing width and depth. Then rill joins rill and the 
swifter flow and scour quickly cut gullies, and within a few minutes 
after the beginning of a brisk rain the surface may be flooded by a 
muddy slime gathering in gullies and carrying literal tons of soil 
matter for each acre of ground. The flow in slime film and rill and 
gulley varies geometrically with the slope ; when the slope is doubled 
the rate of flow is more than quadrupled, while the amount of soil 
matter carried, itself increasing geometrically with the velocity, may 

a The sheet-floods of arid regions and their work are discussed in " Sheet-flood 
Erosion," Bulletin of the Geological Society of America, vol. 8, 1807, pp. 87-112. 
The destructive erosion of humid regions is described in " The Lafayette For- 
mation." Twelfth Annual Report of the U. S. Geological Survey, 1S91, pp. 


be more than octupled. Where gullies meet on increasing slopes, 
falls are formed and the floods descend inches, or feet, or even yards 
in roaring, foaming, churning cataracts: each cataract scours a pool 
at its bottom, in which the whirling waters sap the sides and under- 
mine the rim, eating their way upstream and generally increasing 
the height of the fall ; here the waste is worse than on uniform slopes, 
for each fall marks a locus of retrogressive erosion a where the energy 
of the flood is concentrated, while the lateral sapping initiates slough- 
ing of the banks, and great masses of the soil and subsoil slide into 
the channels. So the waste, which may be trifling on level land, may 
on a steep hillside become enormous — a single storm may sweep away 
the entire soil accumulated through decades or even centuries, of 
organic interaction, and then dig deep ruts and channels far into the 
earth beneath. 

Largely because of the geometrically increasing efficiency of the 
water with slope, partly because of nature's device of protecting the 
surface with root-bound sward, the worst havoc is wrought where 
gullies and falls last from one storm to the next as ways for pro- 
gressive devastation. The careful observer, standing on an old-field 
slope already denuded of soil and rent by gullies, and watching the 
work of a single heavy rain, sees the storm water gather in rivulets 
guttering and roaring down the old channels and cataracts, sapping 
the banks at every bend, and so both deepening and widening the 
trenches and undermining any protective sward on either side, yet 
always pushing most rapidly upslope. Each gully forms an open 
line of attack, occupied by a growing body of rushing, crushing, 
rending, grinding, scouring, sediment-bearing water; each water 
body is a monster of two-score arms each ended in a hundred wrig- 
gling fingers clawing into the humus, under the bordering -ward. 
through the softening surface, slashing the soil into bits and separat- 
ing these into the sand and the soluble and solid semiorganic grains 
of which the soil bed consists. Finally, the debris is sorted and 

a Retrogressive ( " riicklaufige " or "retrogradern") erosion was recognizee! 
in Persia by Emil Tietze in 1S7T in its geologic aspects (" Bemerkungen fiber die 
Tektonic cles Albnrsgebirges in Persien," Jahrbuch der K. K. Geol. Reich., 
XXVI. Band. Wien, 1S7T, S. 375, et. seq. : ibid., XXXIII. Rand, 1882, S. 742: and 
elsewhere). In this country it was recognized in its relation to the current 
work of streams as part of a process of varigradation under which running 
water tends to depart from uniform inclination ("Pleistocene History of 
Northeastern Iowa," Eleventh Annual Report of the U. S. Geological Sur- 
vey, 1S91, p. 2G9). The first result of this tendency is the formation of falls; 
the second is the retreat or retrogression of the falls upstream, with the con- 
tinual sapping of the earth by the swirling waters of the pools which the falls 
form at their bases. The tendency appears in all streams, but is most pro- 
nounced in storm rills on steep slopes; it forms a phase of that autogenetic 
sculpture which gives form to the major part of the lands of the earth tibid., 
p. 245). 


scattered ; the coarse sand is spread over the near-by bottom land and 
the fine sand is dropped in the stream channel, the silt is dumped in 
the neighboring river, and the slime and soluble salts and organic 
matters are swept on toward the sea, muddying and befouling the 
waters on the way. ' If willing to risk engulfment in near-by gullies, 
the observer may watch a single storm destroy a tenth of the soil 
within a radius of 50 yards from where he stands; or as the storm 
passes he may trace the destruction of literal acres in a rolling 80- 
acre old field ; and he can hardly fail to see that each acre of upland 
devastation ruins from an eighth to a quarter of an acre of fertile 
bottom by the overwash of sand and silt. And where destructive 
erosion is once well started, each passing season sees the streams 
impaired hardly less than the soil; more water goes off in local 
freshets and less in springs and seepage, while the low waters are lost 
in the sand washes, and many old-time springs and smaller streams 
go dry; so that (even apart from any diminution in rainfall due to 
reduced local evaporation) every season of neglected soil erosion sees 
the section carried a stage nearer desert conditions. The facts are 
common knowledge in regions of old-field soil wash, the effects are 
abundantly recorded by the camera, and both processes and results 
are of the highest scientific and economic significance. 


When the soil is viewed as a suborganic structure exercising nor- 
mal functions connected with its own circulation, it is easy to see that 
destructive erosion and other abnormal processes affecting the soil are 
analogous to the diseases affecting animals and plants, and that, like 
most diseases, they may be counteracted by treatment tending to re- 
tain or restore normal conditions — and, as proverbially in other dis- 
orders, an ounce of prevention is better than a pound of cure. 

The practical treatment for soil erosion involves both prevention 
and remedy, and while the special means may vary widely with nat- 
ural and artificial conditions affecting the soil the best preventive 
and remedial measures are much alike. In every case the key to the 
treatment is the same — maintenance of normal balance between slope, 
cover, soil, and water supply. 

Of the four prime factors the slope is most considerable, largely 
because the movement of water on the surface varies in a geometric 
ratio with its declivity, and the chief modifications in treatment are 
those growing out of variability in slope. Xext in influence on dis- 
order and treatment comes the cover, largely because agriculture nec- 
essarily involves more or less complete transformation of the natural 
flora and fauna ; and special cases arise to demand treatment on 
account of varying degrees in this transformation. The soil is 


usually the least variable factor and the one most susceptible of con- 
trol; with it the general course of treatment begins and in most cases 
ends — for only over a minor part of the agricultural land of the 
country is it necessary to undertake special control of cover or of 
slope, much less of water supply, to prevent or remedy destructive 
erosion. Commonly the last factor requiring consideration is that of 
water supply, whether derived from direct rainfall or depending 
either on natural inflow or on irrigation; the variability in supply in 
the humid region is much less than the inevitable variations in slope 
and in cover, while the natural supply in arid regions affects erosion 
chiefly in connection with storms and floods. In every case the treat- 
ment has for its keynote the end toward which nature wrought dur- 
ing the ages of land sculpture and plant development, i. e., making 
each acre take care of its own rainfall. 


The first requisite is to keep the soil and subsoil in such condition 
that its range in moisture capacity between (a) air dryness and (b) 
saturation is large, for when this is done the ground will absorb 
much water without drowning, while the abundant water modules* 
in the interstices tend to remain in contact, and hence in free move- 
ment — i. e., in normal circulation. Commonly this condition is at- 
tained when the soil is granular and open-textured or friable to a 
considerable depth: but it is promoted (a) when the particles or grains 
are variable in size and texture, 5 (b) when the inorganic particles 

The natural particle of water is postulated as a module, or a group or 
aggregation of molecules large enough to pass readily with change of tem- 
perature through the three stages in which water exists — i. e., to form a spicule 
of ice, a droplet of water, or a unit of vapor, according to the rapidity and 
amplitude of its own molecular undulations. The water module is discussed 
in ''Outlines of Hydrology,*' op. cit., pp. 206-210. 

6 It is well known that the glacial soils of northern United States are much 
less subject to erosion, other things equal, than the residuary soils* found over 
the greater part of the country, and concurrently that the streams in the 
glaciated region are on the average much clearer than those beyond the glacial 
limits, despite the fact that the topography is youthful in the former case and 
mature in the latter. Of course, certain factors are seldom equal ; rains are 
more largely torrential in both the more southerly and the more arid portions 
of the country; the effect of winter freezing (which may be viewed as natures 
plowing) is greater in the northerly regions: grasses and some other forms of 
low cover flourish best on the glacial soils: and to these and related causes the 
difference in erosion of glacial and residuary soils has commonly been ascribed. 
Yet an important factor undoubtedly resides in the heterogeneity of the glacial 
soils: the particles composing them range in dimensions from finest rock flour 
(below 0.005 mm. in diameter, and hence classed as clay in the mechanical 
analyses) up to pebbles and cobbles or bowlders, while the materials arc no 
less variable than the dimensions. This heterogeneity on the whole diminishes 
the interstitial space occupied by the contained water, but serves to promote 
77206 3 — Bull. 71—11 3 


are intermixed with organic particles, (c) when the humus or organic 
particles are abundant, (d) when the chemical state of the soil-fluid 
tends to facilitate flocculation of particles, and (e) when other less 
important relations are favorable. In nature the condition is brought 
about spontaneously through the interaction of organic and inorganic 
forces; in advanced agriculture it may be brought about and main- 
tained by various devices, notably the following: 

(1) Deep tillage. — Inorganic particles lying in contact tend to co- 
here and finally to unite, in a way resembling the spontaneous indura- 
tion of rocks, whenever the interstices and so the capacity for soil-fluid 
are reduced. In nature this is counteracted by the action of humus, 
by the impenetration of roots, and also by the heave of freezing, and 
other factors; in agriculture it is best counteracted by plowing and 
other forms of tillage which mechanically separate the particles and 
throw them into new and less compact arrangements. Thus, when a 
well-tilled field is properly plowed to a depth of, say, 6 inches, the 
average surface is raised an inch or more — which means that the 
interstices among the particles average 15 per cent to 20 per cent 
larger or more numerous than before the plowing. In addition, the 
overturning of the surface carries the stubble, litter, mulch, and any 
other organic matter down to decompose and form a richer topsoil, 
the organic particles themselves adding materially to the moisture 

(2) Mulching. — Organic substances are commonly porous and 
capable of carrying considerable water; and, especially when pro- 
tected from sun and air and so kept moist, they undergo fermenta- 
tive and other changes liberating gases or solutions which react on 
contiguous mineral substances. In nature the surface is littered with 
leaves, twigs, bark, and other waste which holds storm water for a 
time and then sends it into the soil charged with organic and inor- 
ganic substance in solution, while the root systems are generally 
larger than those of crop plants and on their decay leave their abun- 
dant substance in the soil; in agriculture corresponding benefits may 
result from surface litter, whether accumulated naturally or applied 

both absorption and circulation; at the same time it seems to facilitate the 
impenetration of roots and the charging of the earth with organic matter. When 
glacial soil is exposed to an eroding rill the work of the water is constantly 
impeded by particles, too large or too dense for suspension or saltation, which 
obstruct the flow and occasion grounding of a part of the finer load; and the 
finer material thus tends to line the bottom and sides of the rivulet and in 
turn protect the coarser and intermediate particles. The residuary soils, on 
the other hand, are more homogeneous not only in texture, but especially in the 
size and material of the constituent particles ; and a rill once formed finds little 
obstruction to the successive removal of particles corresponding with those 
adjusted to the slope on which the flow begins, for as the rill grows in volume 
it increases in velocity, and hence in capacity for both carrying and corrading. 


purposely — for although the quantity is generally less, its efficiency 
is increased through the better admixture secured by tillage. 

(3) Fertilizing. — The physical effects of manures and other ferti- 
lizing materials simulate those of organic mulch and humus; their 
solid particles mix with those of the soil, increasing that hetero- 
geneity which in turn augments the capacity for moisture, while 
certain soluble materials yield solutions promoting flocculation or 
granulation, and thereby again increasing moisture capacity and 
aiding circulation. The effects of fertilizers are by no means uni- 
form — indeed as shown by Professor Whitney they are so uncertain 
as to elude prediction. In some cases the application induces changes 
narrowing the range between (a) the drought limit and (b) satura- 
tion, so that the soil is more prone to drown or bake after fertilizing 
than before; when obviously the effect is injurious in so far as the 
mechanical conditions of soil circulation are concerned, and so of 
diminished benefit at the best. 

(±) Seasonable plowing. — Although the condition varies with 
climate and other circumstances, it is quite frequently preferable to 
leave the cropped surface undisturbed throughout the ensuing autumn 
and winter. This is especially true of the middle latitudes, in which 
the ground alternately freezes and thaws and in which snow falls 
usually or occasionally, yet does not commonly lie long. Generally 
evaporation is less from the cropped surface than from plowed land, 
partly because the stubble and litter form more or less of protective 
mulch so that a larger share of the soil-fluid is retained. This re- 
tained moisture in turn facilitates freezing (for it is not the solid 
part of the soil but only the fluid part filling the interstices that 
congeals) with the consequent expansion of the soil, which tends to 
remain open-textured after thawing; and there is generally less 
evaporation of the ice itself from under the litter and stubble than 
from the surface exposed b} T plowing. "When snow falls the differ- 
ence is still greater ; usually the stubble impedes the movement of the 
wind so that the snow lies deeper, softer, and more uniform over the 
cropped field than over the plowed land, while the bleached surfaces 
of the stubble and mulch and weathered ground neither absorb nor 
dissipate heat so vigorously as the black soil newly turned up by 
the plow, so that on the average the snow lies longer on the tin- 
plowed surface than on the plowed land. Now, snow is highly bene- 
ficial to soil. The farmer who declares that a foot of snow is bet- 
ter for his field than a heavy top-dressing is not without reason, 
for not only is a foot of snow equivalent to fully an acre-inch of 

a Fertilizers for Cotton Soils, Bureau of Soils Bulletin G2, 1909. especially 
tables on p. 10. showing certain cases of decrease in production following appli- 
cations of fertilizer. Similar relations are still more strikingly shown for other 
crops in bulletins of the same series. 


water (or over a ton to the acre), but it equalizes temperature 
in the ground beneath, facilitates the few freezings which open 
the texture rather than the manifold freezings and thawings which 
break up the granules, and in the normal thaw produced by warm 
wind and sunshine it softens the soil beneath in advance of the 
melting so that most or all of the snow water is absorbed. The 
result of, say, 18 inches of frost in the ground, protected and 
finally removed by a foot of snow, is to raise the average level 
of the surface fully an inch (as may be measured on a deeply 
planted post or partly buried bowlder) and not only add over 10 
per cent to the interstices in the soil, but leave this space filled with 
water ready for circulation immediately on the sprouting of seeds 
and springing of plants; if not equivalent to given tons of certain 
fertilizer, it is at least partly equivalent to a deep and thorough 
plowing; and the subsequent actual plowing is not only rendered 
easier by the softness of the ground, but may be shallower than would 
otherwise be needful. In regions of winter rains it may sometimes 
be preferable to sacrifice the water of excessive evaporation from the 
newly plowed field in order to save loss through erosion; but gen- 
erally the crop litter can be made to serve the same end with less 
waste. In other cases the growing of winter wheat or other fall- 
sown crops may require fall plowing; but even in this case the sub- 
sequent harrowing (with rolling, if used) produces more or less 
dust mulch and so leaves the surface in much better condition for 
resisting excessive evaporation and snow melting than that of freshly 
plowed land. In an} f event, the advantages of spring plowing should 
be reckoned thoughtfully, for not only does it conform to the natural 
order of things which led to the development of land forms and soils 
and floras (in which the winter frost is as nature's tillage), but it 
affords what is in some cases the most effective means of counter- 
acting the destructive erosion of hilly fields through vernal thaws 
accompanied by spring rains. 

(5) Draining. — Paradoxically, the circulation of the soil-fluid is 
diminished by excessive wetness of the soil ; and this abnormality can 
best be counteracted by drainage, preferably by tile drains laid at con- 
siderable depth beneath the surface. When soil is saturated several 
consequences follow : In the first place, the water ceases to circulate 
except as the evaporation at the surface is balanced by capillarity 
immediately beneath ; again, the reactions of humus are modified and 
the normal activity of the functioning proper to the soil is thereby 
retarded; likewise, the air necessary for the functioning of the root 
tips and for the complete soil-plant circulation is excluded, and the 
impenetration of roots is obstructed ; at the same time slow fer- 
mentation is often engendered, and acids are generated and (in the 
absence of circulation) retained in such quantity as often to sour the 


soil ; furthermore, the granules cleflocculate until the suborganic mass 
settles like leavenless dough, becoming no less sodden than sour, and 
under its own weight tends to indurate after the manner of inor- 
ganic materials — so that finally on drying it bakes, and when turned 
up by the plow forms intractable clods wholly unsuitable for soil 
circulation and for the plant growth dependent thereon. Now, the 
absorptive capacity of baked soil, whether in mass or in clods, will 
not suffice for the taking up of an ordinary rain ; while some of the 
water is sucked in through the hygroscopic surface, the interstices are 
too small and irregular for capillarity, so that most of it is shed as 
from so much solid rock to flow off in corrading rills quickly growing 
into gullies. Commonly the condition of such soil can be alleviated 
by the application of chemical fertilizers, such as slack-lime or lime- 
stone powder on superacid tracts, which tend to induce reflocculation 
toward the normal granular condition ; but the only permanent rem- 
edy lies in systematic drainage — preferably underclrainage. After 
a drain is so laid as to carry off the excess of water from beneath, 
aeration and flocculation occur spontaneously and gradually extend 
laterally on either side of the drain, sometimes actually raising ridges 
1 to 3 inches high which steadily widen until they merge with those 
of neighboring drains — the uplift measuring the increased porosity 
of the soil, including, of course, its capacity for absorbing rain and 
maintaining normal circulation. 

(6) Dust mulching. — As commonly practiced, dust mulching (the 
essential feature of "dry farming") is designed to produce a super- 
ficial layer of finely divided soil matter so loose in texture as to inter- 
cept capillary movement and hence prevent evaporation of moisture 
from the damper soil beneath the surface; and in fact the method 
is most effective in districts of natural subirrigation, such as con- 
siderable parts of the Great Plains and certain valleys and piedmont 
slopes farther westward. Yet it should be noted that the dust layer 
may under favorable conditions extract aqueous vapor from the air 
and distill it into the cooler earth below, and thus render it available 
for soil-plant circulation; the process being analogous to that of dew 
formation, though the product may be imperceptibly small. In sub- 
humid or arid regions and at seasons suitable for vegetal growth, 
the strongest hold on the water appears to be that of the plant, the 
next that of the soil, and the weakest that of the air; so that in time 
of drought (with a moderate quantity of aqueous vapor present in 
the atmosphere) growing plants are actually able to extract water 
from soil nearly or quite air-dry, while the dry soil with its hygro- 
scopic dust skin absorbs water from the air until its aqueous content 
is considerably the greater. It can hardly be doubted that this ability 
of finely divided soil to seize and hold water from the adjacent air 
and convey it to plants is the chief physical foundation for that dust 


mulching pursued empirically for centuries in Arabia and Egypt 
and applied scientifically of late in " dry farming " in this country ; 
though it is equally indubitable that the decomposition of any organic 
matter containing cellulose and starch intermixed with the soil must 
yield some carbon clioxid and water, the former available for absorp- 
tion by the plants and the latter for use in the soil-plant circulation 
(much as seems to occur within the tissues of cactus and other desert 
growths during prolonged droughts). Xow, in well-balanced dust 
mulching, the fine surface layer overspreads the friable soil through 
which the rootlets ramify, both for mechanical support and for 
material sustenance ; the moisture of this soil, whencesoever derived,- 
is conserved by the noncapillary dust ; normal circulation proceeds 
through capillarity in the soil and through the capillarity and 
osmosis extending from root tips to plant tops, and is kept up 
largely by evaporation from the stomata of leaves and other struc- 
tures of the growing plants ; ° while in case of rain at any season the 

a Soil-plant circulation can hardly be elucidated save in the light of a virtual 
autonomy (or self -activity ) of water, well exemplifying that fundamental at- 
tribute of matter described by Spencer as the " instability of the homogeneous ; " 
for although so common H 2 is one of the least-understood substances within 
the purview of ordinary experience, and its properties (while partly known em- 
pirically) are seldom defined systematically or in such wise as to throw light 
on the circulatory mechanism. Pending fuller synthesis, it will suffice to note 
two properties of water which seem measurably reciprocal, and in their inter- 
action apparently control the aqueous circulation, not only through soil and 
plants, but generally throughout those portions of the planet known technically 
as atmosphere and hydrosphere and lithosphere. These properties may be de- 
noted (1) latency and (2) diffusivitij. (1) By far the greater part of the 
water of the earth exists in that liquid state which constitutes the hydro- 
sphere, and which may be deemed its normal or optimum form, not only be- 
cause it is the predominant one, but because it is that in which its several 
characteristics (including the " instability" of liquidity) are most typically dis- 
played. Viewed in the large aspect, liquid water vigorously resists change into 
the other forms except in automatically limited quantity; technically, it pos- 
sesses high specific and latent heat which tend to maintain that form ; prac- 
tically, it will not solidify without giving up " latent heat of water " in sufficient 
amount to lower a corresponding volume 143°, which thermal equivalent in 
nature tends to raise the temperature of contiguous air, and water, and earth, 
and thereby limit the freezing ; nor will it gasify without taking up " latent 
heat of steam" sufficient to raise a like volume 967°, and this thermal equiva- 
lent is in nature taken from adjacent air, or water, or earth, in a manner tend- 
ing to lower their temperature and thereby reduce or limit the vaporization; 
furthermore, while neither the ice nor the vapor will resume the liquid form 
until after taking in or giving out the corresponding thermal equivalents, in 
nature the ice tends to move toward lower altitudes and latitudes in which 
warmth abounds to replace that lost in solidifying, and the aqueous vapor nor- 
mally moves toward altitudes or latitudes In which coolth prevails and is 
ameliorated by its condensation — whereby the terrestrial water continually 
equilibrates the temperature of the earth and maintains it mainly within the 
temperature range of its own normal, or optimum, or at least predominating, 


thin dust layer is quickly saturated, and the spongy soil below ab- 
sorbs the excess of water and forestalls that surface run-off to which 
erosion is due. 

The several devices for controlling erosion through treatment of 
the soil may be held to stand for nothing more than those of good 
farming. They are, indeed, nothing more; but this is only another 
way of saying that under ordinary conditions of cover, slope, and 
water supply, good farming may be depended on absolutely to pre- 
vent destructive soil erosion. Even where the slopes are high, the 
cover inadequate, and the rain liable to come in devastating storms. 
good farming palliates where it fails to prevent the evil ; and whether 
used as palliative or preventive, the special merit of good farming 
lies in this, that under all ordinary economic conditions the remedy 
is not only effective, but pays more than it costs in the way of pro- 
gressively increasing returns for progressively decreasing expendi- 
ture. All the devices tend toward that intensive agriculture which 

form. In the minute aspect (such as that revealed in soil-fluid) the force often 
described as molecular attraction, and in some phases known as surface tension 
and the spheroidal condition, restrains disruption of small water bodies ; while 
the substance is so far inert chemically as to remain stable (in the liquid form) 
in strikingly pertinacious manner and degree. However interpreted, the tend- 
encies of water revealed in both the large and minor aspects to persist in or 
return to the liquid state are most conveniently generalized as a sort of inertia 
or molecular harmony ; and in all its aspects these tendencies take a leading 
role in conserving and distributing over the globe the heat received originally 
from the sun. (2) Conversely, H 2 diffuses vigorously with atmospheric and 
other gases, passing into vapor from both solid and liquid states, and absorbing 
heat with extraordinary avidity in so doing, until a thermal balance is ap- 
proached; it also passes no less vigorously in the liquid form over the surfaces 
and about the constituent particles of various substances in ways commonly 
described as hygroscopicity of these substances, and in some cases this diffusion 
develops (or is attended by) enormous stresses in film pressure, etc. ; and it sim- 
ilarly diffuses vigorously through and even penetrates certain organic struc- 
tures — as in the capillarity and osmosis involved in plant circulation — in such 
wise as at least to contribute materially to the vital action of the organisms. 
Thereby H 2 tends in so extraordinary degree to diffuse its own substance that 
in a state of nature it impenetrates virtually all other terrestrial substances and 
affects their properties in a manner on the whole supplementing its direct 
agency as a conservor and distributor of the energy of insolation — the details 
being far too many for present mention. So effective are the combined proper- 
ties of latency and diffusivity in the reciprocal relations manifested by water in 
contact with earth and air and organisms that they serve to render the surface 
of the land (with its infinitely complicated extensions into the soil below and 
the vital forms above) the theater of perhaps the most intense continuous ener- 
gizing on the planet — energizing due initially to the power derived from the sun 
yet rendered operative essentially through the inherent properties of water. 
Seen in detail, whether in nature or art, every phenomenon and each tendency 
is automatic, or a phase in a process passing from antecedent cause to conse- 
quent effect; while viewed in general the phenomena and tendencies and pro- 
cesses seem to fall into a perfect autonomy in which the innumerable sequences 


in every country grows more and more necessary with increasing- 
density of population, and which is the sole ultimate hope of every 
agricultural land. 


The chief functions of the natural or artificial Cover are (1) pro- 
tection of the soil, (a) by shielding it from direct rain beat, (b) by 
dissipating heavy rains into mists and trickles soaking easily into the 
surface, (c) by accumulating debris to form mulch, and (d) by shad- 
ing and sheltering the surface from sun and wind and so reducing 
evaporation except from the cover itself; (2) promotion of soil- 
plant circulation by (a) catching and retaining rain and snow water, 
(b) retarding or preventing surface run-off, (c) keeping the soil 
moist and open, and (d) allowing the root tips to ramify near the 
surface; and (3) enrichment of the soil by (a) amassing mulch, (b) 
facilitating impenetration and spread of rootlets, (c) decomposing 
inorganic substance by the liberation of carbonic acid in growth and 
decay, and (d) conserving humus and other organic matters. 

These functions are exercised in somewhat different ways and de- 
grees by different types of cover. In nature the leading types are 
those of forest, prairie, and desert ; or. described in detail, close 
forest, open forest, shrubbery, grass, and scrub (the last including 
tufted and scattered grasses, cactus, mesquite, sage, greasewoocl. creo- 
sote, rabbit brush, bunched chapparal. and other partial cover in the 
subhumid and semiarid regions). With that agricultural trans- 
formation which opens the way for soil erosion, the types of cover 
may be described as those of (a) close forest, (b) open forest gener- 
ally available as range, (c) pasture or meadow grass, (d) field crop, 
and (e) scrub sometimes available as range. The types are fixed 
primarily by water supply and secondarily by character of soil, and 
are affected incidentally or materially by slope, temperature, and 

of cause and effect are unified in the primary disposition of H 2 to persist 
chiefly in the liquid state, and to pass into other forms only in sufficient measure 
to maintain that normal or optimum. In more humid sections these inherent 
properties of water may seem of little immediate consequence ; but in subhnmid 
and semiarid districts, where air and soil and organisms seem joined in cease- 
less strife for water, the primary strife is that of the water itself for its own 
normal state on which even the plants are dependent — and any human effort to 
regulate the soil-plant circulation must be adjusted to this condition, just as 
the structures and functions of desert plants have been during the ages adapted 
by nature to the same condition. (The natural adaptations of desert floras and 
faunas and human populations to their environment, which have exercised a 
controlling influence on the development of civilization, are discussed some- 
what fully in " The Beginning of Agriculture," American Anthropologist, vol. 8, 
pp. 350-375, and " The Beginning of Zooculture," Ibid., vol. 10. pp. 215-230— both 
recently reprinted by Prof. William I. Thomas under the title, " Influence of a 
Desert Environment," in his " Source Book for Social Origins," pp. 55-73.) 


minor factors; within limits they are subject to artificial control 
through selection of plants and other means. 

Xow, cover can best be used to palliate or prevent soil erosion by 
making it conform to the natural functions of protecting the surface, 
promoting the soil-plant circulation, and enriching the soil, which 
functions are normal to all plant growth in greater or less degree, 
and in themselves express the fundamental laws of life on the land. 
Yet since crops are selected and cultivation is pursued primarily 
for production and only incidentally to prevent erosion and other 
waste, it is commonly necessary to sacrifice part of the natural func- 
tions and to modify others so that the whole will be redirected 
toward the end of human welfare rather than that of natural balance. 
The means are many and are increasing with every advance in agri- 
culture. Some of them are as follows : 

(1) Tree planting. — When gullying once begins (especially in 
rolling lands formerly wooded and hence too steep for stability in 
field or pasture) the erosion tends to increase cumulatively and 
devastatingly ; but usually it can be checked by planting locust or 
other hardy trees or shrubs adapted to the soil and climate. If the 
destruction is advanced and storms are frequent, it may be necessary 
to start the planting above dams of logs or brushwood or stakes, 
perhaps reenforced with straw or other litter affording a temporary 
mulch. With ample humus, an occasional patch or thicket may 
serve to protect the field; but where the topsoil is thin or poor and 
rests on intractable subsoil it may be necessary to extend the planting 
throughout the length of the gully and well into the undisturbed 
soil on either side. So, too, if the case is a light one (that is, if the 
gullying is slight), the treatment may be temporary only, and after 
a few years the shrubbery or timber may be cleared away and the new 
soil accumulated and held by the roots may be put under cultivation 
and kept available for permanent cropping; but if the case is grave it 
may be necessary to replace the forest cover and preserve it perma- 
nently — of course, taking out the mature timber for fencing or fuel. 
Trees and shrubbery planted in gullies tend to form belts on lines 
along which the drainage runs naturally or may be turned; they 
are less serviceable in controlling surface run-off than belts of trees 
or shrubbery extending across the natural lines of drainage, since 
such transverse belts tend to catch the storm flow and convert it 
into seepage and ground water. So, when the nature of the soil and 
the methods of cultivation are unfit and the fields reveal premonitory 
symptoms of erosion (i. e., if the humus disappears and the topsoil 
turns yellow or red, if the soil hardens and grows intractable, or if 
the yield runs light on knolls) , it is best to adopt vigorous preventive 
measures by setting timber belts transverse to the drainage line- 
before the gullying begins. When the larger ravines are occupied 


by spring-fed or seepage streams during most or all of the year 
these should be preserved, partly as source waters of rivers and partly 
as safeguards against destructive floods in case of exceptional storms; 
in nature streams are commonly protected by bordering shrubbery 
or timber belts; and in thorough agriculture they may be maintained 
and improved by continuing and extending the border growth, which 
serves to strain the sediment from such surplus waters as may escape 
the fields, and thereby maintain the clarity and diminish the erosive 
power of the stream water, while the cream of the soil carried by 
the surface run-off is at least saved to the farm (albeit lost to the 
field) : while the water-side woods may be made no less profitable 
than the cleared acres through crops of grapes, nuts, small fruits, and 
berries, in addition to timber for domestic use. 

The cases in which tree planting is indicated are endlessly variable 
and too numerous for listing: and they are constantly increasing, 
not only with extension of agriculture into previously unbroken 
tracts, but with that growth of intensive methods which best marks 
agricultural advancement. The method is indeed nothing more 
than the artificial application of nature's process, for under natural 
conditions (wherever soil and climate are favorable to tree growth) 
the gullies and scarps breaking the surface are soon invaded by pines, 
locusts, old-field plums, or other trees, tending to shelter the surface, 
hold the soil with their roots, and form mulch with their waste. In 
any event, the sufficiency of the treatment is shown by the results: 
if narrow timber belts either along or across the gullies and along 
the streams serve to stay the surface run-off and prevent scouring 
the remedy suffices: if they fail at first (other devices being in- 
applicable) the belts should be widened until they work well, even 
if it be necessary to plant the entire surface — a necessity only prov- 
ing that the land is unsuitable for farming and fit only for forest. 

(2) Grassing. — Over gently rolling plains, whether prairie or 
woodland originally, erosion is commonly preventable by proper 
treatment of the soil itself; yet, since cultivated soil is less resistant 
than sward, scouring may begin either in gullies or over broader 
surfaces and, proceeding cumulatively by retrogressive erosion, may 
require vigorous remedial treatment. While severe cases may demand 
tree planting, the seeding of the scoured surface with grass com- 
monly suffices. If the topsoil is already gone before treatment begins, 
it may be necessary to scatter litter or apply top-dressing to form 
mulch in which the seed may sprout, or if the land is valuable it 
may be sodded. Especially in the subhumid and semiarid regions 
the natural sward is frequently broken in consequence of overdose 
pasturage, particularly by sheep, partly because the animals like to 
stand level and walk on contours and thus wear paths into the sub- 
soil; such cases are best treated (when the value of the land suffices) 


by regrassing and withdrawing the stock until the sward is restored. 
The grass to be selected varies with climate, character of the soil, 
and other factors. In southeastern United States Bermuda grass 
and Johnson grass are especially effective by reason of that length 
and strength of root which render them objectionable in some situa- 
tions. In general the preferable grass is some introduced variety 
found specially applicable; for. under the natural tendency under- 
lying such plant invasions as those of the English daisy and Japan 
clover, well-adapted alien forms flourish greatly and perform their 
office quickly in the new environment. 

Whether in scattered spots or strips or over entire meads the 
grassing designed to control the water and save the soil should be 
done with a view to early returns in hay or pasturage : generally it 
should form part of a system of crop distribution and rotation 
adapted to the locality: and frequently the grass first sown may be 
replaced later by more profitable varieties. As with timber, so with 
grass, the suitability of the treatment is shown by its success. Some 
tracts are quite stable under sward, but not at all under plowing, 
and these should be grassed permanently and made profitable to 
their capacity through forage or grazing, while certain bottom lands 
adjacent to streams, either timber-bordered or not, are best kept 
in meadow to catch excessive wash from neighboring hills, prevent 
river pollution, and store in soil and subsoil the source waters for 
steady seepage into the permanent waterways; for in well-balanced 
farming the general benefit to the water supply of the country is 
equaled or exceeded by the special benefit of keeping the soil on the 
farm, coupled with the current profit of ample yield in a crop so 
richly responsive to moisture and fertility as hay. 

(3) Nurse cropping. — While nurse crops are commonly sown pri- 
marily to shield seeds and shoots from destructive alternation- in 
moisture, from withering winds, and perhaps from frost, they at 
the same time protect the soil, and frequently they perform their 
chief function in maintaining that normal soil-plant circulation 
required both for the greatest soil efficiency and for the largest con- 
tent of soil-fluid — the precise conditions which, other things equal. 
most directly counteract soil erosion. So the device is commendable 
in general, and is especially worthy of testing whenever slopes are 
steep enough to permit free run-off. 

(4) Cover cropping. — The effect of grassing in orchards or vine- 
yards, as in meadow or pasture, is to form a sward serving to hold 
the soil against storm wash, and at the same time to produce mulch 
and humus and maintain normal soil-plant circulation: and the same 
thing is true of grasses or legumes planted partly to protect the soil 
from desiccation during the period of growth, though designed also 
to make mulch when plowed under. Grain or cultivated growths in 


orchards, intended either to keep down weeds or to form by-crops, 
aid in maintaining the normal condition of the soil, thereby increas- 
ing its absorbtive capacity and checking run-off. with consequent 
erosion: accordingly they may serve a useful purpose both directly 
and indirectly. In special cases Bermuda grass and Johnson grass 
or other growths in corn and cotton fields serve a useful purpose in 
holding the soil on unstable slopes during the late summer and 
ensuing winter, and so may be worthy of encouragement despite their 
weed-like character and their obstruction of plow and cultivator, 
especially when they afford grazing for sheep or swine in the non- 
growing season. Cover cropping and nurse cropping alike imply 
both soil -fluid and plant food in such abundance that the supply may 
be divided without material loss to the main crop, a condition existing 
only where the water supply is abundant ; otherwise the useful pro- 
duction may be so far reduced as to counterbalance the benefit of 
diminished erosion. Especially in the subhumid and semiarid re- 
gions dust mulching should be considered an alternative and gen- 
erally preferable device. 

(5) Eradication of weeds. — The weed is the sign and symbol of 
slack farming, and still more of failure in that complete command 
over the materials and forces of nature forming the aim and the end 
of human effort — it reveals not only imperfect work, but imperfect 
thought. Typically, the weed is useless, in that it contributes noth- 
ing to human welfare: it is injurious, in that it consumes water and 
plant food which would otherwise enrich useful plants ; it is noxious, 
in that it chokes the innocently good by means of its own gross 
luxuriance and by means of the toxic substances left in the soil ; and 
it is malignant, in that through unwitting human help it has acquired 
a better constitution, giving greater pertinacity and stronger persist- 
ence than that possessed by the ordinary plant. So wild artichoke 
and ambrosia not merely overshadow the corn but rob it of suste- 
nance, foxtail steals plant food from the wheat, and in the subhumid 
region the wild sunflower draws the scant ground water away from 
the oats and the barley, leaving them to wither in half -headed Ayeak- 
ness. while the futile farming that permits onty the toughest Aveeds 
to surAuYe and yield seeds for a still hardier generation renders the 
tale of each year worse than the last. True, weeds, like useful plants, 
maintain soil-plant circulation and contribute humus, so that they 
tend to preserve the natural balance between soil and land form, 
albeit no more effectively than the useful plants the}^ displace: yet 
their presence almost invariably marks a lapse from those standards 
of thorough and thrifty treatment which best counteract erosion. In 
general the graA'est single cause of soil deA T astation has lain in the 
abandonment of fields to weeds, either indolently or in the vague 
expectation that shrubbery and forest would follow in natural course ; 


for while the old fields do reset slowly with pine and hardwoods, the 
gullying during the period of growth often removes from entire 
acres the humus accumulated during centuries. As a device for 
counteracting erosion, weeds are in every case — whether on the old 
field, over the producing acres, or in the orchard — unsatisfactory and 
unprofitable, if not a delusion and a snare. And whatever other de- 
vice be adopted, it should end, if it does not begin, with the eradi- 
cation of weeds — indeed with the elimination of everything not 
contributory to human welfare and subject to artificial control. 

(6) Rotation of crops. — Shaped during the ages under a natural 
cover of forest and sward, most of the soil of the country is more or 
less unstable after removal of this cover; and in retrogressive erosion 
the scouring initiated by any storm tends to increase cumulatively 
with each recurring storm from season to season and from decade to 
decade until the entire surface is reconstructed in harmony with the 
reduced cover of the farm — indeed the conspicuous feature of any 
eroded landscape is the flattening of the slope. "While any season's 
injury may be unavoidable, the cumulative tendency may often be 
stopped merely by changing the crop — e. g., if scouring begins while 
the field is in corn, its course may be interrupted by sowing small 
grain next season, and would almost certainly be checked by seeding 
to clover or alfalfa. Although in a broad way each field has its own 
coefficient or measure of erosivity (depending on soil type, texture 
and depth of the humus, capacity for soil-fluid, and slope of surface) , 
the coefficient varies with crops, each of which reacts on the soil in 
its own way not only as cover but in effect on circulation, and this 
for its own special part of the year. So in land subject to erosion, 
crop rotation may have an office additional to those commonly recog- 
nized in the way of counteracting the exhaustion of plant foods and 
the intensification of growth toxics; and with progress toward 
intensive cultivation the full uses of rotation should be combined and 
worked out in a complete system adapted to each locality. 

The devices for palliating or preventing erosion through control 
of cover are akin to those operating through treatment of soil in that 
they are merely those of good farming. They will not always suf- 
fice in themselves, but they will always be found compatible with 
other sound remedial measures; and at the same time they have the 
merit of paying as they go. 


Throughout inorganic nature the slope of the land is the passive 
resultant of forces operative during the ages; but in organic nature 
the slope enters into active relation with the suborganic soil and 
organic cover, and exercises a normal function — which is that of 


removing the excess of rainfall. Now the quantity constituting- 
" excess " varies widely with the efficiency of the cover and the capac- 
ity of the soil, together with the susceptibility of the subsoil and 
underlying rocks to seepage ; and accordingly the normal slopes of the 
land before settlement are variable within wide limits, ranging from 
imperceptible inclinations to grades approaching 100 per cent or 45° 
(i. e.. having a rise of 100 feet in each 100 feet in distance). In gen- 
eral the slopes of prairie land seldom exceed a 10 per cent grade (or 
10 units of rise for 100 of distance) . though they average a little too 
steep for stability under the plow, and hence are subject to destructive 
erosion unless protected by appropriate treatment of the soil and 
cover ; actual reconstruction of the slopes being seldom required save 
where the soil is exceptionally sandy or the subsoil exceptionally 
insusceptible to seepage. TToodlancl slopes, on the other hand, run 
much steeper, frequently reaching grades of 20 to 25 per cent : and 
although some bottom lands are wooded the greater part of the wood- 
lands of the country are rolling. This country includes probably 
500.000 or more square miles of originally wooded arable land, other- 
wise available fcr agriculture, in which the natural slopes are too 
steep for stability under the plow. Throughout this area destructive 
erosion is imminent so soon as the land is cleared and broken: and 
while some of the land lies fiat enough to be protected by treatment 
of soil and cover, several hundred thousand square miles of original 
woodland remain, together with a small part of the original prairie 
land, which can be protected and made permanently productive only 
by appropriate treatment of the slopes. While the modes of treat- 
ment required in particular cases vary with the texture and other 
conditions of the soil and with the character of the crops, the keynote 
is always the same — it is to regulate or prevent surface run-off. The 
chief devices are contouring (or contour cultivation) on moderate 
slopes and terracing on steeper slopes, both of which tend toward a 
gradual reconstruction of the slope into what may be considered the 
agricultural angle of stability; and when judiciously designed and 
carried out they may be made to attain this end without impairing 
production or involving prohibitive cost. These and related devices 
are as follows: 

(1) Contouring. — On level ground it is convenient to plow and 
lay out crop rows in straight lines; and on moderately rolling land 
it is commonly preferable to run furrows and rows in lines straight 
on the ground, even though they curve considerably up and down 
with the undulations of the surface. On steeply rolling or hilly 
ground, on the other hand, labor of men and draft animals is saved 
by laying out the furrows and rows on level lines curving on the 
ground to fit the conformation of hills and draws. Such curved 
lines correspond with the contours, or lines of equal altitude, on a 


topographic map: and cultivation along such lines on the farm may 
be denoted contouring. 

In addition to saving labor, contouring saves the soil. Straight 
furrows and rows running up and down the slope form ways for 
the surface run-off during storms which, even when not developed 
into gullies, aid in the wash of humus and other soil materials clown 
the slopes and away from the fields into neighboring ravines or 
streams: while the ridges of contouring obstruct the surface run- 
off and the intervening troughs hold the storm waters for seepage 
into the ground, thereby preventing instead of aiding erosion. Ex- 
cept on slopes so steep as to compel the use of sidehill plows or other 
means of throwing the soil in a single direction, contouring also tends 
to retain the soil substantially in place on the slopes: for as the plow 
moves on level lines the earth is thrown up or down the slope largely 
at the will of the operator, while in plowing uphill or downhill in 
linear cultivation the greater throw of the soil is always downward. 

In contouring, the first requisite is to lay off the land. With 
moderate slopes or with soil of such texture that the land approaches 
stability, a high degree of accuracv in leveling is not necessary, and 
the opening or " dead " furrows may be laid off by the eye : on more 
rolling ground or sandy soil greater accuracy in leveling is required, 
and the opening furrows should be located with the aid of a suitable 
instrument — what is of late made and known as a farm level, or an 
engineer's or Y level, a surveyor's transit, a theodolite, a hand level, 
or some improvised apparatus, perhaps utilizing a carpenter's level." 

° During the fifties Thomas Skinker, of St. Louis, County. Mo., recognizing 
the need and the practicability of preserving the soil of what, was afterwards 
by reason of its continued productivity well known as the " Skinker Farm '* 
(the tract occupied in part during 1903-1905 by the Universal Exposition), im- 
provised an apparatus, using a plumb line instead of a level. It comprised a 
simple framework of three narrow boards connected in the form of a flattened 
letter A. with the plumb line dropping froin the apex past an index mark at 
the center of the horizontal board: this apparatus was either "walked" over 
the field (the span between the feet being about a rod) and leveled by moving 
the forward foot up or down the slope until the plumb line fell over the index, 
or used by sighting along the horizontal piece: and the curved line so laid out 
was marked with sufficient frequency to guide the opening furrows. In the later 
sixties and early seventies the late Professor Hunnicut and Mr. II. TI. Parks 
(now of Grandfield, Okla.) began terracing their farms in Coweta County, 
Oa.. with the help of an improvised apparatus consisting of a high trestle, 
modeled somewhat after a carpenter's bench, supporting a carpenter's level: 
after bringing the beam of the trestle to a level by moving the legs, it was used 
for sighting points on the ground, with the help of a rod carried by a flagman : 
and when sufhVient points were located the opening furrow was cast through 
them. A serviceable device consists of a glass tube, say 30 inches long, with 
the ends turned up at right angles in the same line (which is easily done after 
softening the tube, preferably after filling with sand, by holding it in a flame 
for a few minutes), so as to form a greatly flattened U; this is nearly tilled 


The chief difficulty in the way of laying off the field on contours 
arises in the " lay of the land," i. e., in the endless variability of 
slope, which is commonly such that two or more opening furrows, 
each lying level, are not equidistant on the ground; so that the 
number of crop rows on one side of a knoll may be two or three 
times the number that may be accommodated on the steeper side. 
With moderately variable slopes the inequalities in width may be 
adjusted partly by varying the width between rows, partly by some 
sacrifice of accurate horizontality in the lines of cultivation; and 
on steeper slopes the methods may merge into those employed in 

In every case contouring should be adjusted to both soil and cover, 
and also to other methods of counteracting erosion. In most soil 
types deep plowing is especially necessary, in order to provide a 
surface sponge capable of holding the water supply not merely of an 
ordinary rain, but (so far as may be) of exceptional storms or series 
of rains; for with shallow tillage over intractable subsoilthe entire 
topsoil is liable to melt into mud and slough or slide down the slope 
bodily, not only impoverishing the field, but ruining adjacent lands 
and streams. In all types, unless conditions render it needful to 
move the soil upslope in tillage, it is sure to work downslope slowly ; 
while despite any effort to the contrary the circulating soil-fluid and 
the wash of humus always tend to move downward, with the result 
of relatively impoverishing the knolls and divides, though generally 
at a much slower rate than with linear cultivation — which natural 
tendencies may be counteracted by more thorough cultivation (mulch- 
ing, top-dressing, and fertilizing, no less than tilling) and more care- 
ful attention to cover at the higher levels. Commonly the devices 
should be directed to maintaining and increasing immediate pro- 
ductivity rather than maintaining the original slope; for. in the 
natural course of things, it is the fate of the rolling field to flatten 
from year to year by the progressive removal of material from up- 
land to lowland in such manner as to develop an artificial, rather 
than a natural slope — the agricultural angle of stability, in lieu of 
that normal to the native flora and land-form. 

(2) Terracing. — When slopes are too steep for stability of the 
soil under contouring, they may be reduced to practicable grades by 
systematic terracing. Where the land is of sufficient value to warrant 
the expense, as in cities and sometimes in suburbs, this can be done 

with water rising mid-height of the upright portions; and when held in the 
hand or rested on a staff the two water surfaces are easily brought in the line 
of vision, when they cut points level with the eye of the user, which can then 
be marked to locate the opening furrows. Many other devices have been im- 
provised, but except for emergency or merely experimental use, the farm level 
or hand level (obtainable from any instrument dealer) saves times and trouble. 


quickly by engineering methods; but where the land is of value 
solely or chiefly for current agricultural production the only prac- 
ticable process is one whereby nature works and man merely directs 
the work, i. e., one wherein the farmer or agricultural engineer does 
little more than lay out the terrace system in connection with cus- 
tomary agricultural operations, leaving the upbuilding and final shap- 
ing to the natural movement of the surface waters and soil-fluid and 
solid soil as they occur normally under these operations. In other 
words, man but starts while nature carries forward a progressive 
process of reconstructing the natural slopes. 

Terracing differs from contouring rather in degree than in kind — 
the leading principles are identical; the chief difference lies in this, 
that in terracing the opening furrows are permanent, forming balks 
or breaks which gradually rise into banks separated by belts of 
plowland. AVith medium slopes the balks are best started by opening 
or " dead " furrows (the first laid to throw the earth upslope and 
the second to throw downward), subsequently seeded with some 
sod-forming grass, while on still steeper slopes any native sward 
may be left on a somewhat wider belt to be either grassed or set 
with shrubbery; thereafter plowing should be on contour lines, 
preferably with a hillside plow, throwing the earth downslope; the 
crop rows, too, should run on contours, when all the normal soil 
movements will be downslope to the balk, of which the upper part 
will gradually grow by accretion while its base will be lowered and 
its slope thereby steepened by the successive plowings and other 
movements of the soil below. In time the balk will become a steeply 
sloping bank or glacis a standing at the angle of stability fixed by 
the strength of the sw r ard or the hold of the shrubbery. Ordinarily 
its height should not exceed 5 to 8 feet, and the width of the belt 
of plowland it sustains, which is to be estimated when the terraces 
are laid out, should be so adjusted to the original slopes as to take 
the agricultural angle of stability when the terraces assume final 
form. So each slope ultimately becomes a series of steps — a giant 
stairway in which the "risers" are steep banks of tough sward or 
strong shrubbery and the " treads " are flat-lying belts of fertile field. 
In ordinary or extensive farming and on moderate slopes the glacis 
should be kept in tractable grass available for forage when cut by 

a The English language lacks terms distinguishing the horizontal and inclined 
elements of terraces. In geology, as in ordinary usage, the term "terrace" is 
generally applied specifically to the essentially horizontal element in contra- 
distinction from the more or less nearly vertical face below, while in military 
nomenclature the term "glacis" (either French or, preferably, anglicized) is 
applied to the slope sustaining a parapet in fortifications and to the similarly 
sloping part of the armament of warships. It will be advantageous to adopt 
and apply the latter term to the terrace face, i. e., to the sloping element of 
the terrace. 

TT2G6 — Bull. 71—11 4 


a properly designed mower or for pasturage in the off-crop season, 
i. e., it should be treated as an integral part of the farm merely 
devoted to a special use and should not be permitted to become a 
nursery of weeds and brambles pushing ceaselessly into the adjacent 
plowland; and as in contouring the higher portions of the terraces 
should receive special care, with a view to keeping up current pro- 
ductivity. With intensive cultivation and on higher slopes requiring 
stronger support than that of sward, the banks may be held by 
perennial shrubbery, which should be made at least partly produc- 
tive — perhaps by a strong hedgerow at the top, with berry bushes or 
vines below; for not only will the yield contribute to the yearly 
profits, but the cropping will direct toward the banks that continu- 
ous attention necessary to insure them against falling into neglect 
and gradually impoverishing the farm. 

As in contouring, so in terracing, laying off the land is the critical 
step; and on slopes steep enough to demand this form of protection 
the process requires intelligence and skill — it forms, indeed, one of 
the most difficult lines of agricultural engineering. The plans for 
terracing the field or the entire farm should be worked out in ad- 
vance with reference to (a) original slopes, (b) texture of soil and 
earth or rock beneath as bearing on the agricultural angle of sta- 
bility, (c) nature of prospective crops and of sward or shrubbery 
for protecting the glacis, (d) nature of subsoil and depth of rock, 
and (e) relative dimensions of terrace and glacis required for sta- 
bility of the entire system. Here, too, the difficulty arising in the 
"lay of the land" (or variability of slope) enters; it is seldom 
practicable to lay off a large field into belts or zones of uniform 
height, since they would be too variable in width; generally it is 
better to plan for banks of varying height protecting terrace plots 
of convenient form and dimensions, so that the field is separated 
into a combination of curved forms (annuli, crescents, lunes, etc.) 
with the banks running out on gentler slopes but rising high or 
merging with others on steeper ground. In practice the combina- 
tions of form, slope, and material are too variable for description; 
they are fully developed only by actual experiment, and can be 
pictured only through projects already carried out. The combina- 
tions are, indeed, so various and so totally unlike those produced in 
rectilinear agriculture that they involve distinct standards, both 
practical and esthetic, and distinct habits of both thought and work. 

When the plans are perfected for laying off the field or farm, the 
lines of the prospective banks should be located with the aid of 
suitable apparatus. Unless detail surveys and plats are contem- 
plated, the best instrument is a Y level mounted on a tripod or Jacob 
staff and used with the help of a flagman and target rod. Usually 
the instrument can be set up on a spur affording a view of a con- 


siderable area ; after it is brought to a level the flagman sets his target 
at the height of the cross hairs, then moves off to a paced distance 
of 20 to 200 yards (according to the slope), when the level man 
waves him up or down the slope to a point at which the center line 
of the target is cut by the cross hairs, where he leaves a stake or other 
mark (switches or stalks easily plowed under are convenient), and 
then passes on to other stations at greater distances. The plowman 
follows, casting his opening furrow through the marks and along 
the same horizontal line as measured by the eye between, first throw- 
ing the earth upward and on his return throwing downward in such 
manner as to form a ridge. The initial line may be extended by 
resetting the level, and the other lines for additional banks at higher 
or lower levels are laid out in the same way. In ground not before 
broken and retaining some sward, the return furrow may well be far 
enough from the first to leave a belt of the original surface, when 
the further breaking should be by a hillside plow throwing down- 
ward, in order that the terrace may be somewhat reduced in slope at 
the first operation. In ground already under plow (new ground 
previously wooded is usually friable enough and rich enough in 
humus to resist erosion for at least a year or two) it may also be 
found preferable to leave a belt for the prospective glacis and seed it 
in grass or set it in shrubbery, as the steepness of the slope may 

In the absence of a Y level, or when a horizontal survey and plat 
are desired, a surveyor's transit is the best instrument for laying out 
the lines, while an architect's level will serve admirably ; though for 
all ordinary purposes a simple farm level will suffice — and especially 
if made with an attachment for determining grades, it is more con- 
venient than any other apparatus (fig. 2 a ). The least expensive 

° Explanation of Figure 2. — Shows approved type of farm level, seen from 
above in outline (in upper part of drawing), and in horizontal section (in 
center of drawing), with attachments and adjuncts. The essential part is a 
metal tube carrying a perforated eyepiece at one end and a cap supporting fine 
metallic cross hairs at the other end, which is open and covers a considerable 
field of vision about the intersection of the cross hairs. The tube is mounted 
on an accurately turned turret, with adjusting screws by means of which its 
horizontal axis can be kept parallel with the planes of the upper and lower 
rims of the turret. Within the turret and parallel to the tube there is also 
mounted a metal case containing a spirit-level vial; this case, too, being adjust- 
able by means of screws to parallelism with the planes of the turret, and so 
with the axis of the tube. The turret rests and rotates horizontally in an accu- 
rately turned seat in the uppermost of a pair of parallel planes, of which the 
lower is reduced to triangular form and provided with three leveling screws 
(as in modern field instruments of American make generally). The lower plate 
is fitted to a tripod head and supported in use by a tripod. In use the tripod 
with the parallel plates attached is first set and approximately leveled by the 
eye; then the turret with its tube is placed in its seat, and the final leveling is 
effected by means of the screws. A target rod is then set up in front of the 



instrument of sufficient accuracy for ordinary leveling is what is 
known as a hand level (fig. 3 & ), which may be used with an im- 

Ditching Sight iEn/aryerf) 

Target {fteduced) 


-Farm level. 

j3rovised target; with careful work its error need not exceed 3 to 6 
feet per mile, or a third of that on a 40-acre field — a probable error 

instrument, and the target is adjusted to its height by bringing its quadrants into 
coincidence with the cross hairs ; thereafter the target rod is carried to meas- 
ured or paced distances, and shifted up or down tbe slopes by the signals of 
the levelman until the cross hairs again coincide with the quadrants, when the 
ground at the foot of the rod is on the same level with that beneath the instru- 
ment. With the aid of a graduated extensible target rod, the instrument may 
be used for measuring heights or grades, just as with an engineers' level. 
While in its simple form designed only for leveling, the instrument can easily 
be adapted to surveying with sufficient accuracy for farm purposes, and also to 
laying out grades — e. g., for ditches or for terrace inclines designed to carry 
storm waters from steeper toward gentler slopes. It may be converted into a 
surveying instrument merely by inscribing an index on the turret in line witb 
the vertical axis of the tube, and graduating the rim of the upper parallel plate 
immediately outside of the turret seat ; and it is adapted to grading by replacing 
the eyepiece with a patented contrivance consisting essentially of a rotating 
disk with perforations adjusted to varying departures from horizontality in the 
line of sight through the tube (in the illustrative disk in the drawing tbe aper- 
tures are adjusted to grades of 1, 2, 3, 4, 5, 6, 8, 10. 14, and 20 inches in each 
hundred feet of distance). Imperfect adjustment of either the leveling tube 
or the spirit vial is easily detected and corrected by inverting and reversing 
the turret and setting the screws in a manner similar to that employed in the 
engineers' level ; if adapted to surveying, any error of adjustment is constant 
and negligible: and when adapted to grading, the error for each aperture is 
independent and constant and should be determined by accurate measurement 
and recorded as a constant correction. 

& Explanation of Ficure 3. — Shows longitudinal and transverse sections of 
an acceptable type of hand level. The essential part of the instrument is a 



far less than that of the plowing, and one negligible in comparison 
with slopes steep enough to require terracing. 

Terracing, unlike contouring, tends rather toward local recon- 
struction of slopes than general flattening of grades; and especially 
where the soil is merely a thin mantle over rock or where it is but a 
talus at the base of a steeper slope above, the amount of soil matter 
removable without impairing production is limited. In such cases 

Section C 0. 

Fig. 3.— Hand level. 

particularly the treatment of soil and cover should be carried for- 
ward in connection with the treatment of the slope; mulching, and 
in case of need fertilizing, should be used on the upper part of the 
terrace, and cover crops should enter freely into the scheme of rota- 
tion. In less favored countries actual soil is sometimes added, but 
ordinarily in the United States this can be made profitable only 
under special conditions of crop and market. 

metallic tube, closed at the ends with caps fitted with plain glass discs (that 
at the eye end small, and that at the outer end large enough to take in a con- 
siderable field of vision) ; to it is attached a metallic case carrying a spirit 
vial with a small bubble, the tube with its appendages and the level being 
permanently adjusted in the factory. The upper part of the tube (as placed 
when in use) is perforated, as is the lower part of the outer case carrying the 
spirit vial, and within the tube, immediately beneath the center of the vial, 
is an inner metallic case carrying a small mirror placed eccentrically, one edge 
marking the (vertically) middle line of the tube; tbis mirror is inclined 45° 
to the axis of the tube (and spirit vial), so that to the eye at the eyepiece it 
reflects the bubble moving apparently in a vertical direction. Midway between 
this inner case and the eyepiece the tube contains a ground glass semidia- 
phragm, so placed as to screen tbe mirror from the eye at the eyepiece and also 
cut off the left half of the field of vision (which is also cut off by tbe mirror) ; 
it is made lenticular to magnify somewhat the image of the bubble as seen in 
the mirror. A fine metallic cross hair (not shown in the drawing) is so placed 
in the aperture in the upper part of the tube as to be reflected from the mirror 
to an eye at the eyepiece in the exact horizontal axis of the tube, where it 
crosses the vertical axis defined by the edge of the mirror. In use the instru- 
ment is brought to the eye by the hand (perhaps steadied by a staff or other 
rest) and leveled tentatively; and when the horizontal cross hair cuts the cen- 
ter of the magnified image of the bubble, its intersection with the edge of the 
mirror marks a point in the field of vision on the same level with the eye of 
the user. In the best type of instrument the tube is made extensible by a 
telescope construction toward the eyepiece. 


(3) Vineyarding. — On slopes too steep for the plow vines may be 
set out on level lines raised partly by mulching and partly by the 
natural movement of the soil as aided by hand cultivation, and whole 
hillsides may thus be made productive. The principles are essen- 
tially similar to those involved in terracing, while the processes may 
easily be extended to other crops, especially berries and small fruits. 

(4) Retain- walling. — Where intensive cultivation prevails and 
proximity to market gives special value to high-grade products, the 
sward-protected glacis of ordinary terracing may be replaced by a 
retain wall of stone or brick or concrete, and this may be raised in 
a parapet completely protecting the soil of the terrace. The risk 
of failure through accident or otherwise, and also the cost attending 
this device, increase geometrically with the height of the wall, so 
that, other things equal, lower structures and narrower terraces are 
preferable. Retain walls should never be vertical, but should be 
built with a batter sufficient to counteract the cumulative effect of 
frost heave and of the hydrostatic head of the ground water com- 
bined, i. e., the angle of stability of the wall should be determined 
no less carefully than that of sward-bound glacis or plow-land soil. 

(5) Annular forestation. — Commonly terracing, like contouring, 
is best adapted to slopes of only moderate length and of steepness 
diminishing upward, i. e., to rolling country rather than to foothill 
and mountainous country; for on long slopes increasing in steep- 
ness upward into foothills or mountains cloud-bursts or cataclysmic 
thaws are liable to form floods so great as to sweep away all ordinary 
protective devices with a violence only the greater for the temporary 
retardation of the waters. Accordingly, long slopes should be 
guarded more effectively than is necessary on shorter slopes of like 
grade; and this may best be done by transverse belts of woodland, 
widening as the slope steepens and narrowing as it flattens, in which 
even the extraordinary storm-born flood or thaw freshet will lose 
its force and lag amid the trunks and undergrowth until absorbed 
by mulch and soil. Such forest belts should, of course, be designed 
for production no less than for protection; for so much of the native 
forests have been cut that silviculture is bound to become an adjunct 
to farming, and wood for home use and market no less a crop than 
grain or meat, milk or cotton. 

(6) Grading. — Where land is dear and labor relatively cheap (as 
in cities and some suburbs, or wherever the soil forms gardens rather 
than farms) and where capital abounds, it may be profitable to un- 
dertake complete reconstruction of the natural slopes by extensive 
grading operations. Such operations pertain rather to engineering 
than farming, and can seldom if ever be made to meet the prime 
requisite of the farm — that of paying their own way by increased 
production within a reasonable period. 


While the treatment of soil and cover and slope should be carried 
out jointly, the last differs from the others in that, whatever the 
methods employed, it is essentially a process of reclamation of lands 
not adapted to agriculture in the original or natural condition. The 
treatment of soil and cover is applicable to all lands, looks to the 
present, and should pay for itself annually through increased yield 
of each ensuing crop ; while the treatment of slope is applicable only 
where rendered necessary by unfavorable natural conditions, looks to 
the long future rather than the present, and can seldom be expected 
to do more than prevent impairment of current production, as the 
tax of labor is capitalized in growing value of the land — a normally 
increasing capital which can be made secure only by continuous and 
consistent carrying out of the treatment until the land is wholly 
reclaimed, i. e., re-formed in a manner adapted to agriculture. As a 
process of reclamation, the treatment of slopes by terracing or other- 
wise affects agricultural communities no less than individual farmers; 
for although an occasional farm may be so situated that it can be 
terraced without regard for neighbors, most farmsteads are so placed 
with respect to the natural slopes that the terracing of one affects 
those adjoining: and in any event the control or neglect of run-off on 
any farm in rolling country reacts on neighboring farms. Thus the 
interdependence of farms and farmers in broken country may ap- 
proach that arising in arid regions, where entire communities take their 
water supply from a single source ; and thereby opportunity is opened 
for associations of citizens and for action in the common interest by 
townships or counties or even States. Similarly, a way is opened 
for collective operations in acquiring and reclaiming lands too broken 
for reclamation by individual effort and capital, in a manner analo- 
gous to that in which both dry and wet lands are reclaimed by col- 
lective or state action in the common interest. 

While the reclamation of rolling lands by treatment of slope is 
essentially investment or capitalization, the best medium is not so 
much money, or its equivalent in labor, as intelligence — intelligence 
applied in so directing natural agencies and processes that they will 
of themselves re-form the land. Xow in all natural processes, time 
is a large and often controlling factor. Especially in contouring and 
terracing, the translocation of material must proceed slowly else the 
natural processes will become destructive rather than constructive. 
The proper rate can not be worked out in advance or expressed in 
equations; it is best measured in terms of current yield of crop: So 
long as (other things equal) the yield increases, the rate may be 
hastened; when it decreases, the rate is too rapid; and when it holds 
its own. the rate is about right. 

Whether the translocation of material is destructive or construc- 
tive, it always works cumulatively; in the one case it robs the soil 


and in the other enriches it. at rates increasing from season to sea- 
son, up to limits hardly to be foreseen, except, perhaps, in the light 
of object lessons in older countries ranking among the most im- 
pressive in the history of the world. The present wastes of Palestine 
and western China and parts of Greece and Spain where' once the 
soils were fruitful, are due to cumulative robbing, at first so slight as 
to pass unnoticed, but later so swift as to defy regulation by human 
power. On the other hand, the wall-held vineyards rising tier on 
tier up the steep hill slopes of southern Europe show what regulation 
will do when started betimes, while the terraced rice fields of eastern 
China and Japan and the Philippines are not only marvels of suc- 
cessful agriculture (despite the crudeness of the methods), but illus- 
trate forcibly the cumulative character of constructive processes. 
At first sight the terraced plantations, like the tiered vineyards, are 
commonly thought to be the outcome of elaborate engineering work, 
in which each feature was laid out with a technical skill and fore- 
sight exceeding those of modern science, though in truth the great 
systems are little more than products of natural growth. The primi- 
tive tiller simply adjusted himself to that which lay nearest at hand, 
and with little labor and less thinking leveled a tiny plot in order 
that his few plants, (often known individually and even named like 
the sheep and cattle of a small herd), might have standing room safe 
from flood; and as he broadened the narrow terrace he naturally 
built up his glacis, though always with the least possible labor of 
hands or effort of mind, yet generally in such wise that the next 
storm helped rather than hindered the enlargement of his plot. 
Where good fortune rather than skill favored, the natural translo- 
cation diminished cumulatively and gave each tiller a wider vital 
margin, so that communities grew and the terracing extended ; while 
if chance — the adverse fate so large in primitive faith — frowned, the 
first incipient terraces were swept away, the tiller and his family 
were impoverished, and the community was cut off. The power of 
invention had not yet arisen among human faculties when these 
terrace systems of the Orient began, and the people and the terraces 
merely grew up together : and now that the faculty of invention has 
grown common and the habit of reasoning from effect to cause and 
from cause to effect has become fixed, these terraced hillsides point 
the way toward a more effective reclamation of the rolling and 
broken lands forming a considerable part of this country. 


The several modes of treatment designed to remedy or prevent 
soil erosion are alike in principle — all operate through regulating 
the movement of water. The primary object is conservation of both 
solid and fluid parts of the soil through a balanced distribution of 


the water supply. The ideal distribution is attained when all the 
rainfall or melting snow is absorbed by the ground or its cover, leav- 
ing n3ne to run off over the surface of field or pasture; in which case 
the water so absorbed is retained in the soil and subsoil until utilized 
largely or wholly in the making of useful crops, while any excess 
either remains in the deeper subsoil and rocks as ground water or 
through seepage feeds the permanent streams. Although this ideal 
distribution can commonly be brought about by proper treatment, it 
frequently fails on account of (a) inequality of supply, (b) catas- 
trophic storms, (c) defective drainage, and ( d) thriftless farming. 
In these cases the evil may be palliated rather than prevented, and 
by collective rather than individual action. 

Inequality of supply (commonly due to variable rainfall) is a 
menace, especially where men fall into the besetting tendency to fol- 
low standards set below rather than above the mean ; for the efficiency 
of flowing water (including its erosive power) increases in an extra- 
ordinarily high geometric ratio with the slope and its own volume. 
It is for this reason that both the destructive and the constructive 
work of water is cumulative in so high degree, that gullying and 
retrogressive erosion generally proceed so rapidly when once started, 
that the single storm may work greater devastation than the ordinary 
storms of a decade or a generation, and that the quantity of sedi- 
ment carried by streams is nearly always underestimated — since the 
load carried during an exceptional day may exceed that carried dur- 
ing an ordinary year. It follows that the flood from an unusually 
heavy rainfall or sudden thaw is liable to break over the contoured 
crop rows or root-bound terraces or other devices for retaining run- 
off — and then to sweep and gully the surface so savagely as not only 
to leave the land in worse condition than before the treatment began, 
but to bury neighboring lands and dam nearby streams. So the 
plans for contouring or terracing, or for otherwise holding the water 
on the land, should be adjusted to the extreme rainfall rather than 
the mean: and until the artificial reconstruction is so perfected that 
the heaviest rain can be kept where it falls, means should be pro- 
vided for carrying off the excess with the least possible injury. To 
this end (and also to counteract soil-cap movement and settling) 
it is well to work the level crop rows and terrace banks somewhat 
upward on the steeper slopes thereby giving the furrows a slight in- 

a The velocity of flowing water varies in some geometric ratio. with the 
declivity of slope; the competence of running water (or its ability to move par- 
ticles of given size, varies as the sixth power of the velocity: the total capacity 
of streams for load varies as something like the seventh or eighth power of the 
declivity: while the efficiency, or total working power of the stream, is estimated 
to vary as the square of the capacity, or, say, as the eighth or ninth power of 
the declivity : " Outlines of Hydrology," op. cit., pp. 201-2. 


clination toward the gentler slopes where any surplus water will have 
greater chance of absorption, while the damage in case of outbreak 
will be diminished. In some cases this may be done within a single 
field or farm : but since boundaries and line fences seldom follow the 
configuration of the land, it will generally be found needful to carry 
the surplus water from farm to farm in a manner involving coopera- 
tion between neighboring owners and coordination of their systems 
of contours or terraces. 

In general the coordinated plans should extend from the divide or 
water parting to the stream in which the excess waters gather; and 
the aim should be gradually to fill all ravines or swales except those 
carrying channels into which the ground water seeps with such free- 
dom as to form permanent streams. In certain cases, depressions 
may be maintained in the lower grounds to catch any surplus storm 
waters and store them for stock (as in the tanques of Mexican 
rancheros) : but generally this is little more than a temporary make- 
shift, since the ponds soon silt up. while it is easy' not only to store 
an equal supply but keep it from contamination in the form of 
ground water. The capacity of ordinary ground for water is seldom 
appreciated; soils readily carry 10 to 40 per cent, and subsoils will 
generally take up from 20 to 30 per cent of their weight in water, and 
the capacity of the underhung rocks is often nearly as great. De- 
rived from the rainfall on the surface of the land, this ground water 
distributes itself according to the slope and the texture of the earth 
matter containing it; it is not limited by farm boundaries or re- 
stricted by civil lines, but is the common possession of the community 
or district containing it, and the primary source for wells and springs 
and streams, no less than an emergency supply for the soil in time of 
drought. Just as complete reconstruction of the surface in accord- 
ance with the interests of the community may be made to guard 
against excessive water supply, so may the reservoir of ground water 
be made to provide against the other extreme : and in both cases the 
interests are those of entire communities, and the usage should be 
regulated by community action. In law and custom, the soil belongs 
to an owner entitled to its full use. while the water pertains to the 
community, and each land owner is entitled to the usufruct rather 
than full use. 

While cloudbursts or other catastrophic storms are commonly 
thought equally liable to happen anywhere, their distribution is. in 
fact, measurably limited to certain meteorologic districts not yet 
fully defined, and in general the probability of their occurrence may 
be inferred from any local series of weather records covering a series 
of years — the longer the better. In other words, catastrophic storms 
and also excessive droughts are merely extreme manifestations of 
what maj^ be considered the normal inequalities of rainfall. In the 


broadest and most general terms, it may be said that all those in- 
equalities in mainland United States tend to culminate where vol- 
umes of vapor-laden air either meet similar volumes from other 
sources or encounter decided obstruction to free movement in strong 
geographic features, so that there is a zone of liability to cloudbursts 
approximately parallel to the Pacific coast, another stretching north- 
ward into the Great Plains east of the Rocky Mountains where the air 
drifts from Pacific and Gulf meet, and a third in the southern Appa- 
lachian region, where the continental air drift is liable to meet 
Atlantic currents. In these regions, and wherever else the weather 
records indicate exceptionally wide inequalities in precipitation, the 
modes of controlling the water should be made especially effective. 
and in case of doubt as to their sufficiency the land should be left in 
that primal condition whereby (in two of the belts at least) nature 
maintained the balance between variable rainfall, soil, slope, and 
streams before settlers came to disturb it. 

Generally throughout mainland United States, the soil can not 
be considered safe from erosion unless its texture, cover, and slope 
are such that it will absorb an inch or more of water within a few 
hours after the last wetting; in districts of highly unequal rainfall 
its capacity should be increased to three or more inches. Some cases 
may require provision for open ways to carry the surplus flow toward 
larger streams or rivers and on to the sea, but better results will 
generally follow such treatment of the soil throughout each district 
that the waters of even the catastrophic storm will be held in check 
and compelled to move so sluggishly as to sink into the soil and the 
earth below and thereby increase the reservoir of ground water. 
The possibility of retaining the water of even the catastrophic storm 
is well shown in the arid region, in which the floods tend not to 
gather into streams, but to spread over extensive surfaces in sheet- 
floods, which perhaps wreak destruction here and there, but generally 
soak into the ground to serve as soil-fluid for ensuing months or 
years and also supply the neighboring springs and valleys through 
steady seepage. Especially where the range in precipitation is wide 
the treatment designed to save the soil from erosion will also save 
the water; yet it will commonly involve the combined action of 
entire communities. 

"While rolling and broken lands generally rise high enough above 
the base-level of erosion to facilitate the distribution of the entire 
water supply by gravity, it sometimes happens that the ground-level 
of water lies too near the surface to give the seepage required both 
for normal soil circulation and for disposing of surplus waters. Es- 
pecially where it prevails over any considerable area, this condition 
can be remedied only by systems of artificial drainage, more or less 
like those required for the reclamation of swamp and overflow lands, 


and thereby involving community action, extending- at least to 
cooperation of owners and coordination of plans. 

Ordinarily the chief obstacle in the way of so controlling the 
natural supply as to remedy or prevent soil erosion lies in the neg- 
ligence of shiftless owners and the indifference of renters concerned 
with little beyond the season's crop; and where conditions demand 
broad plans for distributing the waters over considerable areas, this 
obstacle may become insurmountable save by community action. The 
first requisite is an intelligent and enlivened public sentiment; the 
second is proper organization under legal sanction. In several States 
levee districts, drainage districts, and irrigation districts have been 
organized and work well; and their example is Avorthy of considera- 
tion wherever soil erosion occurs or impends, more especially when 
extensive contouring or terracing is required. The guiding principle 
should be that recognized throughout American history : The greatest 
good for the greatest number — and that for the longest time. As 
recommended by the National Conservative Commission, " First in 
connection with abandoned fields, and progressively in cultivated 
fields, soil wash should be considered a public nuisance, and the 
holder of the land on which it is permitted to occur should be held 
liable for resulting damages to neighboring lands and streams." 
With growing knowledge of the relations between natural water 
supply and productivity of the soil, such a practice is bound to arise, 
and in time extend to the regulation of moving waters ; for since the 
water is paramount to the land in producing value it can work no 
eventual hardship on any for each to so use his own as not to injure 
others. At the same time the plans for avoiding waste and making 
the best use of waters in each community should be adjusted to the 
needs of other communities and, indeed, other States ; for the waters, 
unlike the land, are essentially mobile and transitory, and as sources 
of interstate rivers are subject to joint administration for commercial 
and other uses adjunct to those arising on the farm. 


Description of Plate I. 

Shows effect of removal of natural cover on moderately rolling land of sandy 
soil. The gentle slopes were stable under forest, although somewhat too steep 
for stability under sward. The sward serves to protect the surface except 
where broken by stock trails or roads, or by run-off concentrated in the natural 
lines of drainage, when gullying begins and extends rapidly under retrogressive 
erosion. (West of Grand Junction, Tenn.) 

Bui. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

Plate I. 

Description of Plate II. 

Hilly land, partly wooded, partly cleared, and partly cultivated ; illustrating 
the tendency toward flattening of slopes by destructive erosion attending re- 
moval of the natural forest cover. The entire surface was stable under forest, 
and even the steeper portions remain stable where this natural cover is re- 
tained ; but over the cleared ground erosion has started on the intermediate 
slopes, grown into gullying, and extended somewhat into gentle slopes — the 
field in the foreground lying flat and thus far deriving benefit rather than 
injury from the adjacent waste, though it is foredoomed to destruction as the 
wash increases. The eroded surface soil is devoid of humus or other water- 
holding material, and the subsoil is baked hard, so that most of the rainfall 
is lost as run-off, leaving the ground too dry for productivity. (Yancy County, 
N. C.) 

Bui. 71 , Bureau of Soils, U. S. Dept of Agriculture. 

Plate II. 

Description of Plate III. 

Near view of a typical gully in homogeneous soil and subsoil. The sward 
suffices to hold the comparatively smooth surface until gullying begins, when it 
forms a slight glacis or scarp defining the margin of destructive erosion. The 
main gully and its tributaries with each storm retrogress farther and farther 
into the sward ; on steeper slopes the topsoil is removed and the subsoil baked, 
so that most of the rainfall is lost as run-off, leaving the surface dry and 
sterile during the greater part of the year. In general, the erosion tends to 
flatten the slopes by translocation of material from higher to lower levels. The 
pebbles in the subsoil washed out during storms gather in depressions or pools 
in the course of the gullies, measurably checkiug corrasion at these points 
(thereby illustrating the tendency of heterogeneity in the soil to dimmish 
erosivity). The view shows also the spontaneous spread of plum, locust, sumac, 
pine, and other shrubby and woody growths in a manner tending to retard 
the surface wash and eventually produce a natural cover under which the 
slope will again become stable (Near Oxford, Ga.) 

71, Bureau of Soils, U. S Dept. of Agriculture. 

Plate III. 

Description of Plate IV. 

Typical landscape in rolling land of sandy soil, showing characteristic devel- 
opment of old-field gnllies and (toward the left) the clogging of channels with 
sand washed from the gullies. The moderate slopes were stable under the 
continuous forest cover; when cleared and put under cultivation the surface 
wash impoverished the soil: and when the fields were abandoned a sward 
formed, but remained too weak to prevent washing and gullying on steeper 
inclinations. The destructive erosion is extending rapidly into gentler slopes, 
and within a few years (as shown by neighboring examples) will remove the 
entire soil and fill the ravines, thereby flattening the entire surface to fit the 
reduced cover. The destruction here is increased by the fact that the surficial 
formation (including soil and subsoil) is a moderately tenacious loam, while 
the underlying formation is a less coherent sand. (Three miles south of 
Oxford, Miss.) 

Bui. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

Plate IV. 

Description of Plate V. 

A landscape of gently rolling and sandy land deforested within a few years. 
Under the forest cover the slopes were quite stable, and both trees and under- 
growth were luxuriant, while the surface was carpeted with duff ; but where- 
ever clearing has extended washing has begun, and about the heads of swales 
and brows of hills gullies have gone down to the local base-level and are retro- 
gressing up the slopes in such manner as to threaten the removal of the 
entire surface. Wherever conditions favor (i. e., wherever the slopes are mod- 
erate and fires have not spread) pine trees and plum thickets are starting, and 
both promise at least partial protection with further growth, the former yield- 
ing an abundant and peculiarly effective litter of fallen needles and the latter 
binding the base of the SAvard with a mat of roots. (Two and one-half miles 
south of Oxford, Miss.) 

Bui. 71, Bureau of Soils, U S. Dept. of Agriculture. 

Plate V. 

Description of Plate VI. 

Landscape illustrating the development of gullying on the steeper part of 
a relatively gentle slope, i. e., at the brow of a low hill. The entire surface was 
stable and rich in duff and humus under the original forest cover. After 
partial clearing the sandy soil was impoverised by washing and the field was 
then abandoned except as pasture; but the sward (rendered feeble by annual 
burning and excessive grazing) was too weak to hold the soil on slopes exceed- 
ing, say, 5 per cent, and allowed active gullying to begin in stock trails or 
storm runnels and extend into the flatter surface in such manner as to threaten 
entire removal of soil and subsoil down to base-level ; i. e. to a depth averaging 
more than 10 feet below the natural surface. A typical example of old-field 
erosion in northern Mississippi and neighboring States. (One mile southwest 
of Oxford, Miss.) 

Bui. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

Plate VI. 


Description of Plate VII. 

Near view toward left of preceding, showing forms of autogenetic sculpture 
produced by retrogressive erosion extending to base-level in such manner as to 
reduce the slope to fit the reduced cover. The sandy surface in the foreground 
is growing up slowly in sedge and plum. The surficial deposit is brown loam 
to a depth of 2 or 3 feet ; below lies stratified sand identified with the Lafayette 
formation. The scattered hardwood trees on the old surface are quite insuffi- 
cient to afford protection, though their duff and root-mats retard the wash 
so that (as shown by neighboring examples) they form peninsulas and islands 
of the old surface sometimes persisting for years. 

Bui. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

Plate VII. 

Description of Plate VIII. 

Development of erosion beginning on steeper slopes in rolling land of sandy 
soil. The entire surface was originally wooded, and was practically stable in 
the natural condition ; but on clearing and cultivation the erosion beginning 
on the steeper slopes extended into the upland, leaving only scattered patches 
of the old-field surface. Several acres of the hill in the right middleground 
have been carried away to a depth of over 50 feet, while the valley in the 
foreground has been so completely clogged with debris as to ruin the bottom- 
land farms. (One mile south of Oxford, Miss.) 

Bui. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

Plate VIII. 

Description of Plate IX. 

A gently rolling landscape, originally wooded and stable, in which the mod- 
erate slopes were rendered unstable by clearing and cultivation. When the 
field was abandoned to pasturage a sward formed and grew strong enough to 
hold the soil, except where invaded by gullies: but the gullies, beginning on 
the steeper slopes (as shown in VI and VII), have undermined this feeble 
cover, carrying away the soil and subsoil to a depth of over 15 feet and leaving 
a new surface gradually weathering flat and becoming stable through natural 
growth of grasses and shrubbery. (Six miles southeast of Yazoo City. Miss.) 

Bui. 71, Bureau of Soils, U. S. Dept. of Agricultur 

Plate IX. 

Description of Plate X. 

Gullies, started by roads descending brow of hill, extending into level fields. 
A typical illustration of destructive erosion along the line of low bluffs border- 
ing the river and the delta country in Mississippi and neighboring States. 
Originally these bluffs, like other parts of the surface, were well wooded, and 
the soil was practically stable ; but with settlement and the destruction of the 
luxuriant underbrush and cane and larger trees of the natural cover the soil 
broke easily along trails and roads, forming gullies pushing retrogressively 
into the uplands, leaving temporary tongues and islets of the old-field sur- 
face (one well shown in left middle ground), themselves gradually crumbling 
as their protecting trees and brambles die out on the sun-baked soil. The 
surficial deposit is brown loam, with stratified Lafayette loam below. (Near 
Rocky Springs, Miss.) 

Bui. 71, Bureau of Soils, U. S. Dept. of Agricultur 

Plate X. 




.,. v- 

^ -^P % ** ■ »" * . Jlv 



•—•£ ■>, 




■ '■■■'•m 

' -^^^B 

aC^Jf'V^ ' ' uM 

Kb'"' ^J^^PSRlHt 





• ; 







■ * # 




^liifTMUBi" r ' 


St.; T, 

. . 


Advanced Hillside Gullying. 

Description of Plate XI. 

Landscape showing complete removal of a surficial formation through the 
flattening of slopes due to removal of natural cover. Under the original forest 
cover the entire surface was stable : with clearing the gently rolling surface 
washed enough to impoverish the field, which was then abandoned to pastur- 
age — when gullying started by cattle trails and roads extended in such manner 
as to remove the soil and subsoil (of brown loam) to a depth averaging some 
8 feet. In this case (as frequently along the lower reaches of Big Black River) 
the lower limit of corrasion is determined not so much by the base-level as by 
the greater porosity of the underlying deposits (loess), permitting it to absorb 
so much of the run-off as to diminish the efficiency of the flow in a manner 
analogous to that observed on arid plains where sheet-flooding prevails. The 
final effect is a decided flattening of the original slopes. (Three miles west of 
Edwards, Miss.) 

Bui. 71 , Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XI. 





Description of Plate XII. 

The manner in which country roads are cut down under travel to such depths 
as sometimes to control storm run-off and initiate gullies invading adjacent 
fields. The surface here is gently rolling in general, running into steep bluffs 
toward larger streams; the soil and subsoil are of Memphis silt loam (i. e., 
loess) to a considerable depth. (One and one-half miles southeast of Natchez, 
Miss. ) 

Bui. 71 , Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XII 

Typical Road in Southwestern Mississippi 

Description of Plate XIII. 

Manner in which the loess generally stands in nearly vertical walls while 
the roadway is worn down to depths of many feet or yards, yet occasionally 
yields in such manner as to initiate devastating gullies. The entire surface 
was originally rolling but rendered stable by the luxuriant forest growth. The 
breaking of the surface was inaugurated primarily by clearing and cultivation. 
(Three miles southeast of Vicksburg, Miss.) 

Bui. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XIII. 

Descriptiox of Plate XIY. 

Typical valley bottom in the arid region, in which the natural balance ha? 
been disturbed. Wherever sheet-flooding prevails, the debris gathered chiefly 
from footslopes and buttes and sierras is washed into the valleys, whose depths 
are eventually lined by the finest part of the sediment. Within the storm and 
thaw waters accumulate, forming temporary lakes or "sinks;"' during drier 
months the water evaporates, leaving mud-flats or playas with surfaces some- 
times incrusted with mineral salts. While the playas seem level, their margins 
really rise gently and merge with the surface of the valley plain, bearing its 
characteristic flora growing in sparsely scattered tufts and colonies : and when 
the balance is disturbed by over-grazing of the slightly undulating valley 
surface or otherwise, the floods run off more quickly and corrade deposits pre- 
viously accumulated about the playa margin. The initial gullies commonly 
retrogress and erode the surface extensively, the debris moving farther dowu 
the valley, either into lower playas or into streams flowing to the sea. Even 
when the natural balance is disturbed beneficially by the bringing in of water 
for irrigation, the excess of water may initiate erosion similarly, unless the soil 
is so treated as to prevent. (Near Fallon. Nev.) 

il. 71 , Bureau of Soils. U. S. Dept. of Agriculture. 

Plate XIV. 


Description of Plate XV. 

Erosion invading a characteristically distributed arid-land flora where the 
accumulation of surficial deposits lining the valleys has been effected by the 
joint action of water and wind. Material translocated by wind has accumulated 
in such quantity as to raise the level of the valley somewhat above the plane of 
stability for water-laid deposits; and erosion initiated by a freshet has retro- 
gressed up the main and tributary channels for the storm waters, except where 
the deposits are held by the scant growth. (Near Fallon, Nev.) 

Bui. 7 1 , Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XV. 

Description of Plate XVI. 

Effect of an exceptional flood in deepening storm channels and initiating 
gullies in arid districts. After accumulating valley deposits by ordinary floods 
for years, Rio Santa Cruz suffered exceptional floods in 1908 and 1909 ; these 
first overspread the valley to great width and coated it with slime like the tra- 
ditional Nile mud, and then as the flow subsided corraded the main channel to a 
depth of several feet. During the height of the flood the stratified deposits were 
saturated, and with the lowering of the base-level by the new channel the val- 
ly was drained, not only over the surface in such manner as to open surface 
gullies, but sometimes subterraneously through the more pervious strata in 
such wise as to produce slumps or sinks; and eventually the surface and sub- 
terranean flow joined in irregular chasms of considerable depth. (Four 
miles southwest of Tuscon, Ariz.) 

Bui. 71, Bureau or Soils, U. S. Dept. of Agriculture. 

Plate XVI. 

Description of Plate XVII. 

Invasion of a gently rolling field by gullies starting on somewhat steeper 
slopes. The sward might be sufficient to protect the surface if the gullying 
were prevented, but does not resist undermining by the numberless rills in 
which the storm waters gather. Beneath the surficial brown loam forming a 
fertile soil lies the sandy loam of the Lafaj T ette formation, containing fer- 
ruginous nodules and crusts, baked so hard by the sun as to shed the storm 
waters, leaving little to soak into the ground: while the ferruginous concre- 
tions gathered in the channels retard corrasion and give starting points for 
sedge and other grasses and shrubs, whose spontaneous growth indicates ways 
in which the destruction might be checked. (Near Grand Junction, Tenn.) 

Bui. 71 , Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XVII. 

■ % mM<^- ^ 

Description of Plate XVIII. 

Rear view of Branch (State) Experiment Station, Holly Spring, Miss. 
Erosion is so far advanced as to threaten the entire homestead, including aged 
oaks, century-old cedars, and the residence of the superintendent. A character- 
istic landscape of one of the largest counties in Mississippi, in which the surface 
was originally wooded luxuriantly and so well watered that the ground-level 
of water approached the surface and supplied abundant springs. Through 
clearing and cultivation the sandy soil was rendered unstable, and it is estimated 
that a third of the area of the county has been devastated as shown, while the 
ground-level of water has been so far lowered that most of the springs have 
failed. The soil and subsoil are of sandy loam resting on stratified sands 
identified with the Lafayette formation ; the photograph was taken when a light 
snow in the gullies brought out the relief. The devastated hillside was sub- 
sequently reclaimed and made productive by a combination of hand and plow 
work in deep cultivation and contouring. 

!ul. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XVIII. 

Description of Plate XIX. 

Shows in detail part of a landscape originally stable under forest cover, but 
invaded by erosion on clearing and cultivation ; the destruction now pushing 
retrogressively into the grove maintained about the residence (of which a 
glimpse appears at the right), and threatening to remove the entire hilltop. 
Plum and other shrubs shown in the foreground are beginning to form a new 
cover for the freshly exposed surface. The soil and subsoil are brown loam 
to a depth of some 4 feet, resting on the Lafayette formation. (Three-fourths 
mile southwest of Raymond, Miss.) 

Bui. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XIX. 

Description of Plate XX. 

Another view of part of the landscape shown in Plate XIX, indicating the 
manner in which (provided burning be prevented) upland sedge, plum, pine, 
and locust spring up on surfaces exposed by erosion, even extending into the 
gullies and along the scarps, gradually checking the devastation by restoring 
that natural balance in which the cover protects the slope. The roots of the 
sedge form a sward and its stems collect litter, holding the storm waters ; the 
roots of the plum and locust form a mat supporting the sward, and especially 
during summer provide a low cover, while in winter the fallen foliage forms 
moisture-holding and soil-protecting duff ; the plum yields annual crops of 
fruit and the locust furnishes fence posts in three to five years ; tbe pine is 
particularly effective, since it shelters the surface at all seasons, while the 
fallen needles form a heavy carpet, retaining storm water and keeping the soil 
so soft and friable that tbe water is readily absorbed and the run-off thereby 

Bui. 71, Bureau of Soils, U. S. Dept. of Agricultur 

Plate XX. 

Description of Plate XXI. 

A common result of linear cultivation (i. e.. plowing and planting in straight 
lines) where the slope exceeds the agricultural angle of stability: storm rills 
form and. running both along and across the rows, ruin much of the crop and 
remove the richer portion of the soil. (Xear Holly Springs, Miss.) 

Bui. 71 , Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XXL 

■' P*tt? 


f& 4 



Descbiption of Plate XXII. 

A field in moderately rolling land of loamy soil, in which the slopes were 
stable under the natural forest cover but are unstable in the steeper portions 
under ordinary cultivation. The upper and flatter portion of the field is suffi- 
ciently protected by plowing and planting on contours ; the lower and steeper 
portions are partly terraced. The contours and terraces are laid out roughly 
with little regard to accurate leveling, or to maintaining maximum productivity 
on every part of the field, so that the photograph illustrates a common rather 
than a model treatment of slopes. ( South of Raleigh. N. C. ) 

71 , Bureau of Soils, U. S. Dept. of Agricuiture. 

Plate XXII. 

Description of Plate XXIII. 

Moderately rolling land, quite stable under forest, wholly unstable under 
linear cultivation, and fairly stable under contouring. Botb plow furrows and 
crop rows run on approximately level lines, while the slope is broken by balks 
not utilized as part of the farm but allowed to grow up naturally in weeds 
and brambles, so that the example can not be considered a model. The slopes 
throughout the field are so nearly uniform that the balks are approximately 
equidistant. (Near North Garden, Ya.) 

Bui. 71 , Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XXIII. 

Description of Plate XXIV. 

Gently rolling land, quite stable under forest, but liable to both surface wash 
and gullying under cultivation. The field in the middle ground is cultivated 
on contours adjusted to well-designed balks so curved as to conform approxi- 
mately to the "lay of the land*' and yet remain substantially equidistant; 
they are allowed to grow up naturally in grass and weeds, and hence to form 
waste land. (One-half mile northwest of Abbeville, S. C.) 

j|. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XXIV. 

Description of Plate XXV. 

A field of slope too steep for stability under linear cultivation (though stable 
under forest), in which the surface wash and gullying have been partially 
checked by contouring and balks. The balks have been left to natural growth 
rather than artificial selection and planting, and are hence only moderately 
effective — that in the middle ground (running out in both directions on gentler 
slopes) being too narrow and too imperfectly covered to retain completely the 
run-off and close the runnels, which threaten to grow into devastating gullies. 
(Eastern Mississippi.) 

ll, 71 , Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XXV. 

Description of Plate XXYI. 

Contouring supplemented by a balk laid out with insufficient regard for 
slopes varying with the " lay of the land." While the lines of cultivation are 
curved they are not level, so that the storm waters gather and cut rills, as in 
linear cultivation, running both along and across the furrows, thereby pro- 
moting rather than preventing run-off, and allowing the waters to gather in 
sufficient volume to overflow and cut through the balk and gully the adjacent 
plow land. (Eastern Mississippi.) 

Bui. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

.Plate XXVI. 

Description of Plate XXVII. 

Bather steeply rolling land, practically stable under forest cover but liable 
to surface wash and deep gullying after clearing. The cotton rows are curved 
in such manner as to save labor in cultivation and reduce the run-off, but are 
not carefully leveled and adjusted to the " lay of the land." so that gullying 
has started at several points. While the slopes are not so steep as to require 
terracing, the entire field might be protected by introducing, say, two carefully 
leveled balks, the first some 25 feet higher than the bottom of the ravine in 
the middle ground and the second some 25 feet higher still (the two averaging 
about 200 feet apart), and adjusting to them the intermediate furrows and 
rows. (Northern Louisiana.) 

Bui. 71, Bureau of Soils, U. S. Dept. of Agricultur 

Plate XXVII. 

Description of Plate XXVIII. 

View of a field in gently rolling land with sandy soil, stable under natural 
cover, but liable to surface wash and also to gullying (as shown in the fore- 
ground) under cultivation. The terracing comprises horizontal balks allowed 
to grow up naturally in grass and shrubbery in such manner that each forms 
a low glacis protecting the narrow terrace above ; it is experimental merely, 
and might be improved by planting alternate balks in permanent shrubbery or 
bushes yielding berries or other crops, and with the growth of these allowing 
each intermediate balk to merge gradually with the terrace surface and dis- 
appear. The gullying in the foreground, outside of the line fence, illustrates 
the interpendence of adjacent fields and the consequent necessity for coordi- 
nation of plans and cooperation between neighbors. (One-half mile northwest 
of Abbeville, S. C.) 

Bui. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XXVIII. 

Description of Plate XXIX. 

Illustrates protection of a park by breaking tbe slope into a series of simple 
terraces. Tbe region was stable under forest, but after clearing became sub- 
ject to devastating erosion tbrough retrogressive gullies, sometimes expand- 
ing into "•gulfs'* 100 or more feet deep, with precipitous walls: tbe soil is 
loess (Memphis silt loam), extending to a deptb of 20 to 50 feet, and grading 
into gravel or sand which in turn often overlies Port Hudson clays. Tbe ter- 
races average about a rod in width and some 5 feet in vertical beight : and on 
their level surfaces the storm waters (including the drainage from tbe glacis 
or the upland above) lie until absorbed into the pervious soil and subsoil, 
whereby both surface wash and gullying are practically prevented. The sta- 
bility of tbe glacis is due largely to the properties of loess; with other soil 
and subsoil tbe steeper slopes would require stronger sward or the root mats 
of shrubbery. (Southeastern outskirts of Natchez, Miss.) 

Bui. 71, Bureau of Soils, U. S. Dept. of Agncultur 

Plate XXIX. 

Description of Plate XXX. 

A typical Chinese landscape, in which the rolling surface was stable under 
the original forest cover, but on deforestation became subject to devastating 
erosion, the gullies extending retrogressively and the topsoil washing away, 
often laying bare the stony subsoil, as shown in the foreground. A part of the 
area has been reclaimed by terracing, as shown in the background, the terraces 
being held in place largely by walls of rubble, as shown more clearly in the 
middle ground. 

Bui. 71 , Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XXX. 

Description of Plate XXXI. 

A steeply rolling Chinese landscape largely devastated by erosion due chiefly 
to deforestation. Originally wooded, the surface was fairly stable, despite the 
steepness of the slopes, and was drained by a fairly uniform stream flowing at 
the bottom of a rocky gorge. As cutting of the timber thinned the natural 
cover the soil was progressively eroded until the steeper slopes grew sterile and 
the floods became destructive even on the gentler slopes : meanwhile the coarse 
debris washed from the hills accumulated in the gorge to form a broad sand 
wash, completely overflowed by destructive floods after storms and thaws, but 
running nearly or quite dry between, the meager flow of the dry season sink- 
ing largely or wholly into the sand, as in the arid districts of this country. 
Portions of the surface have been reclaimed, partly by terracing and partly 
by vineyarding on contours : for even on the relatively steep slopes and rocky 
subsoil these devices so far hold the rainfall as virtually to prevent run-off 
and consequent erosion. 

Bui. 71, Bureau of Soiis, U. S. Dept. of Agriculture. 

Plate XXXI. 

Description of Plate XXXII. 

A typical landscape in the loess region of China. Originally the rolling lands 
were wooded and the surface was fairly stable, bnt with deforestation gully- 
ing developed on a stupendous scale, the fine sediment overloading the great 
rivers and even coloring the waters of the " Yellow " Sea. Locally the devas- 
tation was checked and the land largely reclaimed by extensive terracing, the 
efficiency of which depends chiefly on the pervious character of the loess in con- 
junction with its capacity for standing in vertical wails so long as these are 
protected from running water. The terracing is supplemented by contouring 
and sometimes by vineyarding, and in some cases habitations are partly ex- 
cavated and partly bailt into the vertical walls. 

Bui. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XXXII. 

Description of Plate XXXIII. 

A liilly region originally wooded but now mainly deforested and hence un- 
stable in slope, despoiled of humus and topsoil, and subject to surface wash 
accompanied by gullying down to the underlying rock. The foot slopes have 
been reclaimed by terracing. The terraces are narrow and irregular, but 
closely adjusted to the " lay of the land ; " for the first requisite in growing the 
upland rice is so perfect horizontality of the surface that water will stand until 
evaporated or absorbed (the white terrace surfaces shown here and there are 
due to standing water). Each terrace is sustained by a glacis of rubble and 
puddled earth rising into a low parapet sufficient to retain the water of both 
irrigation and storms and prevent surface run-off. Such paddy fields repre- 
sent continuous cultivation for centuries, with full utilization of the entire 
water supply, including (in the best examples) the run-off from the higher 
slopes with its wealth of salts and other earth matter carried in solution and 

Bui. 71, Bureau of Soils, U. S. Dept. of Agriculture. 

Plate XXXIII. 

1 IrhI 


i ' h n 

Ik * *£ '/ 

I ■■¥» ■■ >^»- ••. 

'-••*■- ■■*»»■* •