Skip Navigation

Geologic Studies in Southwestern Kansas

Prev Page--Structural Geology || Next Page--Economic Resources


Physiographic Divisions

Adams, in 1903, classed most of southwestern Kansas as High Plains country, but showed the Smoky Hills upland as entering the extreme northeastern corner of the area, and the Red Hills upland as marking the border of the High Plains at the southeast. Fenneman (1931, pl. 1) maps the area as lying mainly in the central part of the High Plains region, but extending eastward into the Plains Border region. For purposes of local description, the area may be further subdivided. The general nature of these subdivisions has already been noted in the section on topography, and their origin has been considered in part in the sections on stratigraphy and on structure. Here it remains to summarize, systematize, and particularize. The following is an outline of the divisions proposed by me (fig. 19):

List of physiographic divisions
of the southwestern Kansas area
Upland areas:
Kearny area
Kalvesta area
Syracuse upland
Stanton area
Haskell area
Cimarron Bend area
Minneola area
Odee area
Intermediate and lowland areas:
Pawnee river drainage basin
Arkansas valley area
Scott-Finney depression
Finney sand plain
Cimarron valley area
Meade area
Red Hills
Ashland basin

The divisions are based on a detailed study of topography and drainage, together with a consideration of geologic history, so far as known. The broader topographic features involved are shown in the generalized contour map, plate 2, but for details the individual topographic sheets must be consulted. In some places transitions from one area to another are gradual rather than abrupt, and, in the absence of detailed knowledge of geologic history, boundaries must be drawn somewhat arbitrarily. Certain of the boundary lines are therefore to be regarded as provisional, and as subject to revision in the light of such added knowledge as future studies may provide.

Figure 19--Physiographic divisions of southwestern Kansas. 1A represents the Kearny area, 1B the Kalvesta area, 1C the Syracuse upland, 1D the Stanton area, 1E the Cimarron Bend area, 1F the Haskell area; 1G the Minneola area, 1H the Odee area, 2A the Arkansas valley area, 2B the Scott-Finney depression, 2C the Finney sand plain, 2D the Pawnee river drainage basin, 2E the Cimarron valley area, 2F the Meade area, 2G the Red Hills, and 2H the Ashland basin.

Physiographic divisions of southwestern Kansas.

Upland Areas

The upland areas of southwestern Kansas represent a deformed and dissected depositional surface, mainly on Tertiary formations but partly on Quaternary beds. The several areas differ from one another in amount and direction of topographic slope, in physiographic history, and in details of surface topography. Boundaries between them are controlled partly by stream erosion and partly by post-Ogallala deformation.

Kearny Area

The Kearny area lies in northern Hamilton County and northwestern Kearny County, and is named for the latter. It is not limited to the area considered in this report, but extends northward into Greeley and Wichita counties. At the east, it slopes toward and merges with the Scott-Finney depression. At the southeast and south, it slopes toward the Arkansas valley, although in Hamilton County the extent of this valleyward slope is uncertain, owing to lack of any detailed contour map. The area is believed to have been blocked out by post-Ogallala crustal warping, and was cut off from other areas to the south by subsequent stream erosion.

Kalvesta Area

This area is named for a small settlement in northeastern Finney County. It extends from the Scott-Finney depression at the west into an irregularly digitate set of uplands and divides between the several streams of the Pawnee river and Coon creek drainage systems at the east. Eastward it converges somewhat with the Arkansas valley, and its slope veers from east nearly to east-northeast, owing probably to warping of the original depositional surface. Its slope as a whole is gentler than that in the Kearny area, suggesting that the post-Ogallala tilting was progressively less toward the east.

Syracuse Upland

The Syracuse upland is a narrow highland paralleling the Arkansas valley in southern Hamilton County. It is named for the town of Syracuse just to the north. It was left in relief by stream erosion at the north and by down bending and faulting at the south. Eastward it converges gradually with adjoining areas.

Stanton Area

The broad Stanton area is named for the county that it encloses. It extends also into parts of Morton, Stevens, Grant, and Hamilton counties. The Bear creek drainage basin at the north constitutes the lowest part of the area. The southern portion is traversed by a few widely separated shallow valleys, which drain into Cimarron river. The average slope is east-northeast. The area is bounded on the south by the Cimarron valley and on the east by a dry tributary valley. It represents a slightly warped portion of the Ogallala surface.

Cimarron Bend Area

The Cimarron Bend area comprises a large area between the Cimarron valley and the Oklahoma state line. It is characterized by broadly undulatory topography, and by the complete absence of surface drainage from any except the marginal portions. There are numerous shallow basins and intervening swells, and many scattered sand-dune areas in various stages of development. Its Quaternary history is somewhat obscure, but probably involved more or less modification of the original Ogallala surface by slight warping, by possible stream deposition, by probable solution-and-collapse activity, and by wind action. The finding of a dip of 13° in volcanic ash in northeastern Seward County lends some support to the likelihood of solution and collapse.

Haskell Area

This area is named from Haskell County, which lies in its central portion. Its slopes contrast both in direction and in declivity with those of the Stanton area, being very gentle and dipping to the east-southeast. The area is bounded on the south by the Cimarron valley, on the east by the slopes of Crooked creek valley, and on the north by a moderate break in slope leading down to the Finney sand plain. Although its slopes are mostly regular, many minor irregularities in detail may be seen in the area of the Garden City and Lakin topographic sheets.

Minneola Area

The Minneola area, named for the town centrally located within it, comprises a sprawling upland in northern Clark County and southern Ford County, characterized by long, irregular projections to the southwest and west. It is separated from the Haskell area at its northwestern tip by a moderate break in slope, and from the Finney sand plain at the north by a similar topographic disconformity of uncertain significance (Garden City and Dodge quadrangles). It is bounded at the south by the Red Hills escarpment, and at the southwest by Crooked creek valley. The area is believed to represent a composite depositional slope on Odee, Kingsdown, and loess deposits, and possibly in part also on older formations. The slopes are essentially continuous, however, thus providing no basis for demarking transitions from one formation to the other. Examination of the topographic maps suggests that there was some broad arching of the depositional surface along an east-west axis in post-Kingsdown time.

Odee Area

The Odee area represents the southern extension of the southwestern projection of the Minneola area, severed from that area by the down cutting of Crooked creek. It seems to represent a depositional surface on the Odee formation, modified by the development of many solution-and-collapse depressions and by extensive dune building.

Intermediate and Lowland Areas

Pawnee River Drainage Basin

This district, comprising the valleys of Pawnee river, Buckner creek, and Sawlog creek, represents a deeply dissected belt in which encroaching streams are gradually destroying the upland. The more rapid progress of erosion here seems to be related to the facts that: (1) the main stream channels are lower than that of the Arkansas, thus providing steeper gradients for dissecting tributaries; (2) the impervious Cretaceous floor of the pervious Ogallala rises above stream grade, thus promoting seepage of and sapping by ground water emerging from the latter formation; (3) the configuration of the bedrock floor seems to be such as to place the ground-water divide close to the Arkansas valley, thus depriving that valley of any but the smallest proportion of subsurface runoff.

Arkansas Valley Area

The Arkansas valley area includes all land lying below the High Plains level and sloping toward or draining directly into Arkansas river. It comprises the present. floodplain, the terraces (described in later paragraphs), the sand-hill belt in part, and the bordering dissected slopes. The origin of the present course of the Arkansas valley is discussed under drainage history, in the next part of this report. The "perched" position of the valley (fig. 2B, C) with respect to the Cimarron valley at the south and the valleys of Pawnee and Smoky Hill rivers at the north is puzzling. The observed differences in elevation, amounting to more than 200 feet, cannot be explained as a result of Quaternary aggradation, for the valley fill is not known to exceed 50 feet in thickness. Conceivably the relatively greater depth to the impervious bedrock floor along the Arkansas was a factor (see Fenneman, 1931, pp. 23-25). Possibly the volume of sediment carried down from the Rocky Mountains was such as to leave the stream less energy for erosive work, but more probably crustal warping within the Kansas area played an important part in raising, or in preventing the lowering of the stream's gradient along this portion of its course. In any event, it seems probable that the anomalous condition in this area is closely related to events responsible for the Great Bend of the river just to the east. This, however, is a problem beyond the scope of the present paper, requiring for its solution a full knowledge of late Tertiary and Quaternary erosion and deposition within the Great Bend area.

Scott-Finney Depression

The asymmetric Scott-Finney depression represents a post-Ogallala downwarp along an axis trending roughly north-south. Numerous solution-and-collapse depressions have been superimposed on the floor of the depression. It slopes irregularly toward the Arkansas valley, but has no surface drainage into that valley.

Finney Sand Plain

The Finney sand plain comprises a broad, dune-covered area intermediate in level between the river valley and the plains upland of the Haskell area (pl. 2). It may be regarded as the outer part of the Arkansas valley, for it parallels the valley, and together they form a continuous area that lies below the High Plains level. Beginning in Kearny County, the sand plain attains maximum breadth in southern Finney County and northern Haskell County, and wedges out eastward in Gray County. Surface drainage is lacking.

The Finney sand plain is too broad and too irregular to be explained simply as a product of post-Ogallala stream planation. Rather it seems to have been depressed by crustal warping more or less continuous with that responsible for the Scott-Finney depression. It is crossed by certain elongate belts below the general level, which conceivably may represent ancient stream courses, possibly related to the Rexroad or Kingsdown formations (Garden City and Lakin quadrangles).

Cimarron Valley Area

The Cimarron valley area includes the narrow belt draining into Cimarron river west of Clark County. Its eastward extension is described as the Ashland basin. The cause for the bend in the river's course is discussed under drainage history.

Meade Area

The Meade area comprises the drainage basin of Crooked creek. Its anomalous trend is a result of structural control, and its diversified topography was developed by the interaction of faulting, downwarping, stream erosion and deposition, and the solutional work of underground waters. The tectonic movements involved were complex in character, and have not been fully analyzed.

Red Hills

The Red Hills represent a deeply dissected district in the Permian redbeds and overlying Cretaceous and Tertiary beds, in Clark County and eastern Meade County. The extent of the dissection, in contrast to that farther west along the Cimarron valley, seems to have been controlled by the following factors: (1) the grade of Cimarron river, which constitutes local base level, here lies at maximum depth below the plains level; (2) the relatively impervious sub-Ogallala beds rise well above stream grade, thus favoring the emergence of ground water; and (3) inferred downwarping is believed to have given rise to initial consequent streams of steep gradient, and to have lessened the amount of erosive work necessary to carve out the existing topography.

Ashland Basin

The Ashland basin, named for the principal town within its limits, is a broad lowland crossed by shallow stream valleys. It slopes gently toward Cimarron river. The origin of the basin is obviously related to that of the bordering Red Hills, but has involved considerable stream deposition in its later stages. The history has yet to be worked out in detail, however.

Drainage History

The drainage history of southwestern Kansas must be considered in relation to that of the central High Plains region as a whole. The stream courses in this region are all believed to be of post-Ogallala origin, and to have been initiated as consequent streams on the differentially upwarped depositional surface of the Ogallala formation (compare Fenneman, 1931, p. 23). These original courses were subsequently modified by piracy to form the existing drainage pattern.

Streams North of the Arkansas Valley

The drainage pattern in the area between the Platte and Arkansas valleys, although representing much more territory than is considered in detail in this report, must be considered as a unit. Many of the streams fail to cut through the Ogallala, thus indicating that their origin cannot antedate the deposition of that formation. As there must have been extensive and continuous lateral shifting of streams during Ogallala time, it is evident also that the stream courses cannot antedate latest Ogallala time. All the streams in this part of the region head in the plains, and none receive any runoff from the mountain area. All are essentially parallel to the regional slope, and together they form a semi-radial pattern, diverging from an apical zone in the western part of the Colorado Piedmont (fig. 9). These relations point to consequent 'Origin on the depositional surface, and the pattern itself suggests a broad, fan-shaped uplift of that surface, maximum along an east-west axis. As already pointed out, the physiographic relations of the Ogallala formation in eastern Colorado can be explained satisfactorily only on the basis of differential uplift of the depositional surface, the uplift being along lines inferred from the drainage pattern alone. Since the initiation of the drainage lines, there have been minor modifications. In the Colorado Piedmont, many of the streams within the radial pattern now flow on Cretaceous beds. It seems probable that these may have been superposed from a thin mantle of Ogallala, now stripped away. The sharp bend in Sand creek near the town of River Bend, in eastern Elbert county, where the stream turns abruptly from northeast to southeast, suggests piracy of a former northeastward-flowing stream by a stronger stream from the southeast.

Of the streams within the area of this report, only the unnamed southeastward-flowing streams in northern Kearny County and the Pawnee river system seem to form a part of the pattern described above. It is possible that Pawnee river once extended much farther west, and was continuous with one of the Kearny County streams, later to be dismembered by downwarping athwart its course, but it is equally possible, on the other hand, that all the concerned streams post-date the downwarp, and until the age of that feature is definitely known, uncertainties must remain.

Buckner creek and Sawlog creek depart from the general pattern in showing a northeasterly trend. This, in connection with the north-northwesterly trend of the contours associated, suggests consequent origin on a surface warped divergently from the regional slope, possibly at a later time.

Arkansas River

The Arkansas, alone of the streams in the area, heads in the Rocky Mountains. Its history is probably longer and more complex than that of the other streams, and must be considered in relation to the deposition of the Ogallala beds. Within the mountain area, physiographic studies by Powers (1935) indicate that the present course of the river dates back to pre-glacial time. Its Pliocene history is obscure, however, and much remains to be learned about the physiography of the Southern Rockies. Still there is no reason for doubting that the river's point of issuance from the mountains in Ogallala time was essentially the same as today.

To account for the river's course across the plains, the following hypotheses are presented: (1) stereotyping of a course held in latest Ogallala time; (2) headward extension from some point east of the Ogallala outcrop area; (3) consequent origin by following a trough or series of sags in the post-Ogallala uplift; (4) complex origin, involving combinations of the above, perhaps supplemented by subsequent diversions.

The first of these suggested explanations is readily seen to be inadequate to account for the course of the river as a whole. The late Ogallala streams were aggrading and probably branching in character, and there is a strong likelihood that their waters were dissipated before crossing the depositional area. It is very doubtful whether the course of a stream of this type would have been as straight as that of the present Arkansas. Furthermore, unless by sheer chance such a stream happened to coincide with the trough of the post-Ogallala uplift, its course could hardly have been maintained. This may have been the case in some stretches and perhaps the present river locally does follow segments of late Ogallala stream courses. This possibility seems to apply with less difficulty to the more easterly stretches, where post-Ogallala uplift was relatively mild. Even so, there is a probability that Quaternary crustal movements may have been sufficient to modify that portion of the stream's course, however acquired.

The second possibility, that of headward extension to establish the stream's course, is also untenable except possibly as a local factor applicable to minor irregularities in the trend of the valley. Any adequate explanation for the river's course must take account of the relatively large volume of water issuing from the mountain areas, and this must have been much greater during the glacial stages of the Pleistocene time than at present.

The hypothesis of consequent origin is more promising. The trend of post-Ogallala warping, as indicated on figure 9 and described previously, was such as to elevate areas north and south of the present Arkansas valley, while leaving that district as a broad sag or trough. It is obvious that any water issuing from the mountain area must have sought the lowest path across the plains, and that this would have been provided by the low place or places in the regional uplift. Either the river found its valley ready-made as a continuous trough, or developed it by the integration of a series of sags through filling and overflow. Subsequent dissection in the Colorado part of the plains region has been too deep and extensive to leave much evidence bearing on these details. In the Kansas portion, the evidence for a trough in the post-Ogallala upwarp is even less obvious, for the amount of uplift was less and the resultant relief lower. It seems probable, however, that the present course of the Arkansas represents its original course at least as far as central Kearny County. It is possible that beyond that point there have been shifts and diversions as a result of Quaternary warping. The northward bend in Finney County, for example, may have been caused by one or more diversions produced by continued subsidence along the southern part of the Scott-Finney depression.

The rough parallelism of the river valley and a bedrock depression in part of the western Kansas area is puzzling. Perhaps this parallelism is actually less close than appears on the bedrock contour map (fig. 8), which, because of inadequate data, is oversimplified. On the other hand, it is possible that the position of the river valley was controlled by downwarping along a zone of recurrent subsidence dating back to Ogallala or pre-Ogallala time.

Streams South of the Arkansas Valley

The stream pattern south of the Arkansas valley shows certain anomalies that must be accounted for in any hypothesis as to origin: (1) the parallelism of the northeasterly stretch of the Cimarron, its North Fork, and Bear creek; (2) the abrupt bend in the Cimarron from northeast to southeast; and (3) the abrupt bends in the upper stretches of Crooked creek and Bluff creek (pls. 1 and 2).

The subparallel streams at the west are essentially normal to the contour lines, suggesting that they were consequent on the lower flank of the Sierra Grande upwarp. If that is true, it is probable that, like Bear creek today, the western stretch of the Cimarron and its tributaries originally drained into the Arkansas valley. In that event, their present courses may readily be explained by piracy due to headward cutting of what is now the southeasterly stretch of the Cimarron. Such a stream would have had the advantage of a much shorter distance to trunk drainage lines, and thus a steeper grade. As a matter of fact, Bear creek could easily be diverted to the Cimarron drainage today by a trench about 20 feet deep and a few miles long.

The course of Bluff creek also seems best explained by piracy. The headward stretch of this stream aligns perfectly with a tributary of Rattlesnake creek (see Ashland topographic sheet and pl. 16), suggesting that it once formed a part of that drainage system. This ancestor of Bluff creek, however, must have originated later than the streams in the western part of the area, for it flowed on the Quaternary Kingsdown formation, and presumably was related genetically to the events of late Kingsdown time. The present sharp bend to the south is best explained as a result of piracy by a more vigorous stream of steeper gradient, cutting back from the Cimarron.

The history of Crooked creek is probably more complicated, for it shows more than one anomalous bend. The headward stretches, like those of Bluff creek, seem to be inherited from an ancestral stream draining into the Arkansas, but the course of this earlier stream is not ascertainable definitely. Haworth (1896, p. 371) suggested that it might have passed north of Minneola, thus implying that it may have been continuous with the upper stretches of Bluff creek. An alternative though less plausible interpretation is that it joined Mulberry creek. In any event, the sharp bend in the present stream is best explained as a result of disruption and beheading of the original stream by the Meade downwarp, the south-southwesterly stretch of the stream being consequent on the downwarp. Its development may have involved the tapping of a lake at the north by headward erosion of the stretch at the south. The southeastern bend in Crooked creek farther downstream suggests another case of piracy, but the cause is obscure. Conceivably it may have resulted from the integration of a series of sinks, such as are numerous in the upland strip of the adjoining Odee area to the south.

River Terraces

Arkansas Valley

Two recognizable terraces occur along the Arkansas valley in southwestern Kansas and it is possible that there are obscure remnants of one or more others at higher levels. The lower terrace, locally referred to as "second bottoms" (Darton, 1920, p. 3), lies about 5 to 8 feet above floodplain level. It is well displayed south of Syracuse, just northwest of Kendall, south of Holcomb, on the southwest side of Garden City, and south of Dodge City. This terrace is included in areas mapped as Laurel series by the Soil Survey (Coffey and Rice, 1912, pp. 77-82). The relation of the terrace to the floodplain suggests that it is of cut-and-fill origin, and of late glacial or post-glacial age.

Remnants of the higher terrace lie about 15 to 26 feet above the floodplain. It is possible that this terrace has some slope toward the river level to which it was graded, and that the interception of the terrace by the present floodplain at different points along this slope accounts in part for the differences in height. Available figures indicate that the average height of the terrace is about 20 feet, and suggest that the height may be slightly greater at the west than at the east. This terrace is underlain by sand and gravel having a maximum exposed thickness of about 20 feet. At Garden City and points west, the underlying Ogallala is exposed, but east of Garden City it generally is not, implying that the bottom of the terrace deposits passes under the floodplain level. The terrace is best exposed on the south side of the river, where, in many places, it forms a low bluff bordering the valley bottom. Elsewhere, however, it is masked by dune sand, or is marked by an inconspicuous transitional slope. The width of the terrace is uncertain, owing to the superimposed sand-hill topography. Probably it averages only a mile or two.

In Hamilton County there are gravel deposits that suggest terrace remnants at elevations of as much as 40 feet above floodplain level. Whether these represent a westward tilting of the 20-foot terrace or belong to a higher terrace unrecognized farther east remains to be ascertained.

The significance of the terraces in Kansas can be determined only through correlations with Colorado, where terrace remnants are more numerous and more prominent. West of the Royal Gorge, Powers (1935) described a series of seven terraces at heights of 20 to 390 feet above present river level. The lowest of the series is post-glacial, the next four fall within the glacial epoch, and the upper two are pre-glacial. East of the Royal Gorge, fewer terraces have been recognized, and correlations through the Gorge are uncertain. During reconnaissance observations, I found terrace remnants at various places in the Canon City quadrangle at elevations of 20, 70, 95, 120, 170, and 195 feet above river level, These represent at least three, probably four, and possibly five distinct terraces. In the Pueblo quadrangle, Gilbert (1897) recognized several terrace levels, but gave no details, and mapped them all together. I measured terraces, however, at heights of 20, 60 to 70, 80, 100, and 135 feet. In the Nepesta quadrangle, Fisher (1906) has mapped two terraces, the lower at about 40 feet, and the upper about 80 feet above the floodplain. Both are rock-cut terraces and both slope toward the river, a condition found also in other parts of the region. Farther east, in the Catlin quadrangle, Toepelman (1924a) has mapped three different terraces. The lowest is stated to have its base about 25 to 30 feet above the floodplain, the middle one about 50 feet higher, and the highest about 40 feet above the middle one. In the La Junta area, Patton (1924) recognized only two terraces. The elevation of the lower one was not stated, but may be inferred, from a comparison of his geologic map with the topographic map, to be 40 to 60 feet. The higher terrace was stated to be about 30 feet above the lower. Toepelman believes that these terraces correspond to his middle and upper levels. East of the La Junta area, no detailed physiographic studies have been made, but reconnaissance by me indicates existence of terrace remnants about 80 feet above floodplain level at Caddoa and near Granada, the latter place being about 15 miles from the Kansas line. Remnants of a low terrace at about 15 to 20 feet were found also.

Conclusions drawn from the foregoing data must be indefinite. Although terraces and terrace remnants are numerous, and occur at several different levels, it is obvious that they cannot be reliably correlated until more detailed field studies have been completed. There is a probability that certain of the terrace levels converge eastward, and that a single terrace level in Kansas may be equivalent to two or more of the Colorado terraces. Possibly the 80- foot terrace of the Granada area is represented by the questionable terrace remnants 40 feet above the floodplain in Hamilton County, or perhaps erosion has destroyed the topographic expression, or dune sand obscured it. The lowest terrace of the Colorado section, at 15 to 20 feet, seems to be persistent, and may correlate with either the 8-foot or the 20-foot terrace in Kansas. In any event, it seems probable that the axis of the Scott-Finney depression may have served as a hinge line in the intermittent uplift responsible for the cutting of the terraces, and that, unless this depression is of more recent origin than supposed, no remnants of high terraces will be found to the east. Thus, although the Pleistocene history of the river is recorded in the terraces, the record is not easily deciphered, and only by continuous correlations from Kansas to the area west of the Royal Gorge, where terraces have been related to glacial stages, is it to be expected that an adequate outline of the history can be prepared.

Cimarron Valley

The terraces of the Cimarron valley, especially the higher ones, are relatively inconspicuous, and in most places their presence is inferred from occurrence of gravel deposits rather than on the basis of topographic expression. Topographic forms produced by erosion in earlier cycles seem to have been destroyed or obscured by erosion in later cycles, so that remnants are few and discontinuous.

In Morton County, 5 miles north and 1 mile east of Elkhart, a deposit of coarse gravel uncovered in a gravel pit may represent a high-level terrace at 70 to 75 feet above the floodplain, almost masked by sand dunes. On the north side of the river 1 mile east of Rolla, another gravel pit is cut in the coarse alluvial cover of a 20-foot terrace.

In northwestern Stevens County, near the crossing of a county road, thin gravel deposits capping an undoubted lower terrace and a probable higher terrace were found at levels of 20 and 55 feet, respectively, above the Cimarron plain.

North of Liberal in Seward County, terrace gravels occur at heights of 40 to 50 feet above river level. Along U. S. highway 54, northeast of Liberal, a thick deposit of similar gravel lies 80 feet above stream level. A possible low terrace lies at 20 feet.

The correlation and dating of terrace remnants along the Cimarron valley must await much additional study. It is evident, however, that post-Ogallala erosion involved more than one cycle of downcutting.

Sand Dunes

General Features

The sand hills, or dunes, of southwestern Kansas vary greatly in relief, contour, soil cover, and stability. At many scattered points there is now vigorous drifting of free sand and in a very few places there are wide expanses of bare sand. By far the greater part of the area, however, is well covered with grass and shrubs (pl. 20A), and is thus protected from wind attack. The grass-covered dunes range from steep and irregular hillocks to broad, subdued swells. Transitions from the one to the other, and from dune to nondune areas, are abrupt in some places, but elsewhere are so gradual that it is difficult to map boundaries of dune areas. Some parts of the area are suited for cultivation but others are not.

Plate 20--Dune topography. A, Grass-covered sand hills south of Garden City. B, Active dunes about 4 miles west of Syracuse, September, 1939, looking west.

Two black and white photos of dune topography; top is grass-covered sand hills south of Garden City; bottom is active dunes about 4 miles west of Syracuse, September, 1939.

The maximum local relief of the sand-hill areas is about 70 feet, and the average relief is nearer half that figure. Individual dunes range in length from about 100 to 450 yards, but a few exceptional forms reach a length of 1 mile or more.

The dune areas are all alike in having no surface drainage. Local rainfall is absorbed or accumulates in the innumerable hollows, there to be dissipated by evaporation or seepage. Streams entering the dune belt from bordering slopes are almost all lost within a short distance.

On the ground, the dune areas show everywhere a repetition of the basic alternation of hills and hollows, but with endless variations in shape and size. In some places there are smooth, well-rounded forms, but in other places there are abruptly varied shapes. Little system can be discerned, and the topography as a whole seems best described as chaotic.

Available topographic sheets contribute little to a better picture of dune topography, and fail to give any adequate concept of dune morphology. On aerial photographs and mosaics, however, minute details of form are distinctly shown, and some pattern may be distinguished. Forms are forked, hooked, corded, pitted, furrowed, looped, reniform, amoeboid, and irregular. Patterns are reticulate, nucleate, and jumbled, they range from coarse to fine.

Only a preliminary account of the general features of dune history is essayed now, a more extended analysis of the dune forms, and comparison to other dune areas of the world, being reserved for a later paper. In succeeding paragraphs, modern wind work is described, then a general scheme of dune development is outlined, and finally the questions of sand source and wind direction are considered.

Modern Wind Action

Eolian action of today seems to be partly a continuation of activities long in progress under natural conditions, and partly a consequence of abnormal conditions more recently inaugurated by widespread farming and grazing. The former comprises the development of minor blowouts in older grass-covered dunes, and the continued modification of certain younger dunes not yet stabilized by vegetation. The latter includes a considerable range of phenomena, from minor sand drifting to the development of true dune forms. The results of natural versus unnatural conditions are generally distinguishable, sometimes strikingly so. Both are described below.

Sand Drifts Related to Cultivated Fields

Familiar sights in the "Dust Bowl" are the fence-line and hedgerow "dunes" (pl. 21A, B). These result from the surficial erosion of a thin layer of soil from a broad surface loosened by cultivation. Hedges, weed-choked fences, or other obstacles trap the drifting material and give rise to an essentially stationary type of dune. In some places, the buildings of farmsteads have caused sand accumulation on a considerably larger scale, with unfortunate consequences for the occupants (pl. 21C). "Dead air" spaces associated with trench-like road and railroad cuts provide another effective type of sand trap, leading commonly to the accumulation of sufficient sand to block or displace the line of travel. In Stevens County, the sand must frequently be plowed from the railroad tracks along many stretches to keep the lines open.

Plate 21--Artificial "dunes" resulting from wind erosion on cultivated fields: A, Hedgerow dune about 2 miles cast of Johnson, July, 1938, looking southeast, B, Fence-line dune 2 miles north of Feterita in Stevens County, August, 1937, looking west. C, Sand drift in farm yard about 4 miles north of Elkhart, August, 1937.

Three black and white photos of dunes; top is hedgerow dune about 2 miles cast of Johnson, July, 1938; middle is fence-line dune 2 miles north of Feterita in Stevens County, August, 1937; bottom is sand drift in farm yard about 4 miles north of Elkhart, August, 1937.

In some places, where no definite obstacle is encountered, drifting sand locally spreads over the ground as a thin sheet, overriding and destroying such crops as lie in its path (pl. 22A).

Plate 22--A, Drifting sand destroying crop in a cultivated field south of Feterita, in Stevens County, August, 1937. B, Spot blowout just south of Syracuse, view looking south. September, 1939. C, Areal blowout 11 miles north and 3.6 miles west of Liberal. view looking northeast, July, 1938. Note ancient dune bedding eroded in relief in foreground, and sand mound in background.

Three black and white photos of dunes; top is drifting sand destroying crop in a cultivated field south of Feterita, in Stevens County, August, 1937; middle is spot blowout just south of Syracuse, view looking south. September, 1939; bottom is areal blowout 11 miles north and 3.6 miles west of Liberal. view looking northeast, July, 1938.

None of the samples were subjected to disaggregating procedure, but some attrition probably took place during shaking.

Drifts of the types described are produced both from sandy soil and from silt-clay soil. The former in many places represents old dune sand, and, once started, drifts the more readily of the two. As soon as "blowing" is well under way, a surface of reduced resistance is developed, and much additional material is released by sand abrasion. The silt-clay soil is somewhat more stable. The drifts or dunes derived from it are made up partly of silt or clay pellets rather than of true sand. Such material, when soaked by rain water, tends to become more or less compacted and stabilized by the bonding action of the redistributed clay particles. Mechanical analyses of the material from two representative sand or dust drifts are tabulated in Table 4.

Table 4--Mechanical analyses of wind-drifted soil material, giving size of grains in millimeters.

Size Grade Sample number
1a 1b 2
1.0-0.5 1.06 0.0 0.05
0.5-0.25 12.1 12.5 8.2
0.25-0.125 18.6 40.0 36.0
0.125-0.062 45.4 30.5 28.7
< 0.062 22.9 17.0 27.1
Location of samples: (1) Hedge-row dune 2 miles east of Johnson (pl. 21); 1a, Partly compacted "dirt" from south side; 1b, Loose, mobile "dirt" from north side; (2) Fence-line dune along U. S. highway 50N, northeast of Garden City (surface loose).

The material in all samples is somewhat dark, owing to the presence of organic material. In the fractions between 0.5 and 0.062 mm from samples 1 band 2, microscopic examination showed that 10 to 40 percent of the material consists of pellets or granules of silt and clay, rather than of true sand grains. In sample 1a the percentage is smaller. This sample more nearly represents the true composition of the soil material, having been subjected to the action of rain water and weathering processes subsequent to eolian movement, thus leading to partial breakdown of the original clay pellets. Analyses of samples 1b and 2 resemble those for typical dune sand (Wentworth, 1932, pp. 12-19) in showing maxima in the 0.25- to 0.125-mm fraction, but differ in the larger proportion of finer material.

Sand Drifts Along Stream Channels

Along the floodplain of the Arkansas, no dunes or sand drifts were anywhere observed to be forming. Eolian action has been limited to the rippling of exposed channel sands. Along Cimarron river, however, some sand drifting is in progress, particularly in Morton and Stevens counties. North of Elkhart, considerable sand is being heaped up on the south bank of the river, and a minor amount has been drifted locally on the north bank. Farther east, in Clark County, there seems to be some minor drifting of sand on the north bank also. These phenomena are undoubtedly of very recent inception, for as pointed out in a later section of this report, the Cimarron channel was not such as to have provided a source of free sand when the area was first settled.


At many places in the sand-hill areas, grass-covered dunes are pitted or notched by small though active blowouts (pl. 22B). These lend a distinctive character to the landscape, and contribute to the scenic diversity of the dune areas. There is no reason for believing that this aspect of wind action has been limited to historic time, although it seems probable that it has been accentuated since the coming of white men.

Blowouts of a different and much larger type are found in many places where farming has been attempted in old, subdued dune-sand areas. On aerial photographs, these blowouts stand out as bare, white patches of elongate, rudely lobate, or irregular outline. They range from a few hundred feet to more than half a mile in length, and are broad in proportion. The soil may be eroded to a depth of several feet, thus involving more intense wind scour than that responsible for the ordinary fence-line dunes. Either an irregularly pitted surface or a shallow basin may be produced (pl. 22C). Additional sand is released continually by the abrasive action of wind-driven sand already present, and the denuded floor of the depression, where composed of semi-indurated material, commonly displays the miniature fluting, pitting, and studding characteristic of wind-etched surfaces (pl. 23A). The sand thus freed accumulates as an irregular sheet, as a series of drifts or humps (pl. 23B), or as a broad mound on the leeward side of the source area. The combined loci of erosion and of deposition, by reason of the broad area and comparatively low relief of the features involved, may be designated as an areal blowout, in contradistinction to the relatively deep and narrow form of the smaller blowouts.

Plate 23--A, Detail of wind-etched floor in blowout shown in plate 22C. B, Sand drifts invading field at north end of areal blowout of recent origin 6 miles south and 3.5 miles east of Kendall, September, 1939. C, Dune ridge of recent origin in same blowout, view looking west. Note abandoned windmill at left in background.

Three black and white photos of dunes; top is detail of wind-etched floor in blowout shown in plate 22C; middle is sand drifts invading field at north end of areal blowout of recent origin 6 miles south and 3.5 miles east of Kendall, September, 1939; bottom is dune ridge of recent origin in same blowout, view looking west.

Barchans and Transverse Dune Ridges

In a few of the larger areal blowouts, transverse dune ridges and barchan or sub-barchan forms have developed. These attain a maximum height of about 20 feet. They are especially well displayed at localities in southwestern Kearny County (sec. 27, T. 25 S., R. 38 W.) and in central Hamilton County (sec. 28, T. 24 S., R. 41 W.). At the former locality (pl. 23C) these dunes were reported to have been formed during the last few years. Two abandoned windmills near the center of the dune area suggest that the trampling of stock was responsible for breaking the sod cover and allowing wind action to start.

Comparison of the profiles of dune ridges and barchans of various sizes, and of the profile at points of different height along the same dune, indicates that the profile of the dune varies with its height. No steep leeward slope was found on dunes less than 3 feet high. Above that height, the relation of the lee slope to the gentler windward slope varies systematically. On low dunes, the rounded windward slope curves over the top and bends down the opposite side before intercepting the angle-of-repose slope. On dunes somewhat higher, the windward slope simply flattens out on top before meeting the lee slope. On dunes still higher, above 15 feet, the windward slope rises continuously and meets the lee slope at a sharp crest. This suggests that there is a gradual and continuous transition from a low, rounded sand heap to the typical sharp-crested dune form, thus according essentially with the observations of Oldham (1903) in India and of King (1918) in the Libyan desert.

The crest of the typical migratory dune is a puzzling feature. Although the reason for its perpetuation, once formed, is not difficult to understand, the cause for its inception is less clear. I suggest that the initiation of the steep leeward slope is due to a transition from laminar to turbulent flow of air currents, and that, for winds of a given velocity, which would be the maximum velocity at any particular locality, there is a critical dune height at which laminar or streamline flow breaks down on the forward slope of the dune, leading immediately to the dropping of such sand as may be in transit, at the angle of repose. Once initiated, the break in slope itself would probably induce turbulence in winds of less than maximum velocity, and thus perpetuate itself. This hypothesis is consistent with the observations of Whitfield (1939), who found that artificial reduction of the crest and lee slope of a dune, by means of a drag pole, allowed sand to be carried well beyond the dune and, in the absence of new additions of sand from the windward side, led to a marked lowering of dune height. This suggests the restoration of laminar flow. Obviously dune growth can take place only when the winds reaching the dune are already more or less "loaded" with sand. When the winds are fully loaded, dune growth should involve simple leeward extension without any corresponding shift of the windward slope, or perhaps with slight retrogressive banking of sand on that slope. When the approaching winds are underloaded with sand, moderate erosion may be expected on the windward side of the dune, followed by deposition of sand thus acquired plus sand already carried, on the leeward slope, leading to differential movement of the dune as a whole, more rapid at the front than at the rear, thus involving an increase in the breadth of the dune. Under either of these conditions of dune growth, it seems that the point of inflection in the profile gradually should shift toward the top of the dune, perhaps in relation to the stronger wind velocity encountered at greater height above the ground surface, where the retarding effect of frictional drag dies out. When the winds reaching the dune are carrying virtually no sand, maximum erosion on the windward slope is to be expected. If turbulence sets in at the crest or on the leeward slope, the sand thus won would be dropped, and the dune would advance uniformly, without lagging of the tail, but under conditions of laminar flow over the entire dune, there would be no abrupt dropping of sand, and the results reported by Whitfield would be in order. A more extended analysis of the aerodynamics of dune building, however, must be reserved for later papers.

In one locality, bare transverse dune ridges of much larger size and greater extent than those described above are found (pl. 27), and are believed to antedate the settlement of the area. These dunes are found in a belt about 1.5 miles wide and about 4 miles long on the south side of the Arkansas valley beginning just west of Syracuse. They constitute the largest expanse of free dune sand in the state of Kansas. They have the appearance of great waves of sand as much as 30 feet high, and nearly a mile long. Individual dune ridges trend approximately east-west, and range in plan from rectilinear through curvilinear to undulatory. Along the margins of the main belt, the sand ridges break down into smaller sub-barchan, lobate, and irregular forms. These dunes are being imperceptibly modified by every strong wind, and the sand is gradually being moved toward the river. They occur in an older dune belt, and seem to be of secondary origin, having been formed in much the same way as the smaller dune ridges of more recent origin, but on a far larger scale and under natural conditions. They are believed to represent the culmination of blowout action.

An area of somewhat similar dune ridges, but separated by strips of vegetation, and seemingly in process of stabilization, occurs in Kearny County a few miles southwest of Hartland.

The Sand-Dune Cycle

The wind action that has been described mostly represents only a minor and superficial modification of a much older dune topography. The history of this antecedent dune mass is complex, and may best be understood in terms of an ideal cycle of development (Smith, 1939). This developmental scheme is based partly on deductions from the internal structure of old dunes as revealed in various cuts, partly on inductive reasoning directed toward the linking of diverse forms in some orderly genetic sequence, and partly on analogies with current forms and processes, as no true homologues for certain of the original dune types are now definitely known to be in course of formation within the area. It must be emphasized that the wind work of today is insignificant in comparison with that at various times in the past, and that dune forms now being built under natural conditions are on a scale much smaller than that of the original dunes. It is probable that the major dune-building episodes of the past took place under climatic conditions very different from those of the present.

The dune cycle embraces two distinct phases, characterized by different processes: first an eolian, or active stage, and, second, an eluvial, or passive phase. During the eolian phase, the dune is built up. Stabilization by vegetation introduces the eluvial phase, throughout which the dune is protected from wind attack, and undergoes gradual wastage through weathering and creep. This second phase of the cycle, however, is subject to interruption through rejuvenation, whereby wind action is resumed and a new cycle inaugurated. In this second cycle, and in any subsequent cycles, the two phases of the initial cycle are repeated, though with variations in the stages attained and in the resultant morphology.

Eolian Phase

The initial dune cycle begins with wind scour on some bare sandy surface. This develops into a primary blowout, which is of the areal type. It consists of two parts: (1) a zone of sand removal, and (2) a zone of accumulation. Illustrative are the fence-line and hedge-row dunes and artificially induced areal blowouts already described. Topographic expression in the zone of removal depends on whether the source of sand is replenishable or nonreplenishable. The former condition exists only along stream channels or lake shores, and no lasting topographic expression is to be expected. Where the source is nonreplenishable, which is probably the more common case, either a wind-swept pavement or a wind-scoured hollow is produced, depending on the duration and intensity of wind attack, and on whether the locus of removal is stationary or migratory, confined or expansible.

The zone of accumulation is marginal to the sand source, and is commonly controlled by the vegetal mat, as witnessed both by observations on modern wind action and by analogous interpretation of rude casts of plant stems or roots in old, dissected dunes. Accumulation may begin as a sheet or drift of sand (pl. 23B), or may early take the form of a mound or ridge, according to the degree of resistance offered by the vegetation. In either case, a mound of some type generally develops sooner or later if accumulation proceeds. It may grow by the addition either of foreset, topset, or backset beds, or by combinations of these. denending on the equilibrium between the rate of plant growth and the rate of burial by sand. Steep foreset beds develop only when sand is swept in more rapidly than plant growth can keep pace, hence to meet with little resistance and be carried over the crest of the dune and be deposited on the leeward side at the angle of repose. Such bedding is very rare in the area under discussion, having been seen at only one place (pl. 26A). Backset bedding, the common type, is developed when plant growth keeps ahead of the influx of sand, thus to trap the sand and cause it to bank up, layer upon layer, on the windward side (pl. 26A). This type of bedding has a low angle of dip, and involves retrogressive growth of the dune, in the direction from which the wind is blowing. Topset beds may be laid down over either of the other types, and may be essentially continuous with the backset beds, differing only in their position at the top of the dune and in their essentially flat dip. Examination of numerous cuts through old dunes shows the backset type of bedding to be the prevalent one of the region. The original dunes grew upward and backward by accretion, under the continuous influence of vegetation, and were fixed in position from the beginning. Dunes so formed, or otherwise developed under the governing influence of vegetation, are here classed as phytogenic. [Note: The term phytogenic, signifying building through the agency of plant growth, was selected with the help of Professor Emeritus M. W. Sterling, of the Department of Greek at the University of Kansas.]

In ground plan, the primary blowout ridge is controlled by the shape of the source area and by the degree of differential erosion and accumulation during formation. Unmodified primary dune forms are extremely uncommon in the area studied, unfortunately, and their outline is largely a matter of conjecture. In one locality, however, south of Englewood in southwestern Clark County, presumably true primary forms occur (pl. 25). Their form ranges from roughly U-shaped to irregular, and probably corresponds to that of the so-called parabolic dunes of certain European writers. A progressive rather than retrogressive type of development is suggested, thus leaving some doubt as to whether they are representative of the still earlier generation of dunes of more widespread occurrence.

Plate 25--Aerial photograph of dune forms south of Englewood. The U-shaped to irregular ridges represent primary blowout forms in a stage of late youth to early maturity. They were formed by winds blowing from the south-southwest. They are extensively pitted by small secondary blowouts, many now active. The channel of Cimarron river crosses the bottom of the picture. The arrow points north and is about 0.5 mile long. Photograph from U. S. Agricultural Adjustment Administration, October, 1938.

Aerial photograph of dune forms south of Englewood. The U-shaped to irregular ridges represent primary blowout forms in a stage of late youth to early maturity.

Eluvial Phase

Primary blowout activity ceases and the eolian phase of the cycle comes to an end when plants in and around the dunes are enabled, through temporary climatic advantage, to spread sufficiently to cover completely and to stabilize the sand-swept area. This transition may be either abrupt and general or protracted and progressive. In the latter case, continued wind action on the blowout mound after the source area was stabilized may have an important effect on the final dune form.

After stabilization there begins the eluvial, or passive phase of the cycle, and the gradual degradation of the dune. Henceforth the main processes are soil building and soil creep. The incoherent sands are bonded together by silt and clay released through chemical weathering of silicates, and by organic material accumulating from the decay of vegetal matter. Some additional fine material may be added by dustfalls. The essential nature of these changes was recognized by Coffey and Rice (1912, pp. 51-52), who wrote that--

When the sand hills become stationary, weathering immediately begins to work changes in the character of the soil. A large proportion of the sand grains are feldspar and minerals other than quartz, and they break down readily and undergo chemical changes with comparative rapidity when exposed to weathering. While the original material varies in composition, the dune-shaped hills, which are now stationary, must owe their present loamy character to the decomposition of the sands once loose and incoherent. In some localities the hills have long been stationary and the dune-like contours have been modified by weathering and erosion.

As a result of mass creeping of sand and soil downslope, relief is decreased, the slopes are reduced, and the original blowout hollows and interdune depressions are gradually filled. The dune mass is lowered and spread out. Contours are rounded and simplified. These changes are essentially gradational and continuous throughout, but for convenience the eluvial phase may be somewhat arbitrarily divided into stages of youth, maturity, and old age.

During youth, a soil zone is formed and the steeper slopes are lowered. Youth is a transitional stage, somewhat precarious in character, and subject to easy reverses. Passage into maturity may be said to occur when the entire dune presents a smooth and regular profile. Breaks in slope are eliminated, angularities smoothed, and symmetry established. During maturity the major amount of degradation is effected, and the soil becomes thicker and more stable, sufficiently stable in some places to permit judicious cultivation. The slopes become less pervious, giving rain wash and even gullying greater opportunity to supplement the work of creep. The final transition to old age may be said to occur when the original form of the dune has become unrecognizable (pls, 26B, C and 24). Thereafter the same processes continue, but with waning vigor. The landscape is characterized by broad and gentle undulations, which, as time passes, grow fainter and fainter. In this stage, there is a likelihood that drainage integration may be effected-integration of inter-dune basins with one another, and with outside drainage.

Plate 24--Aerial photograph of area shown in plate 26B, C. The arrow points north and is approximately 0.5 mile long. Photograph from U. S. Agricultural Adjustment Administration, July, 1939.

Aerial photograph of area shown in plate 26B, C.

Plate 26--A, Exposure showing foreset, topset, and backset bedding in old dune along new railroad cut just west of Kismet. B, C, Old-age dune forms about 3 miles east of Fowler. See also, plate 24.

Three black and white photos; top is exposure showing foreset, topset, and backset bedding in old dune along new railroad cut just west of Kismet; lower two are old-age dune forms about 3 miles east of Fowler.

The stages outlined, although not necessarily of the first cycle, may be roughly correlated with the units mapped by the Soil Survey (Coffey and Rice, 1912). Youthful dunes, together with dunes still in the eolian stage, are mapped as Dune sand. Mature to old-age dune areas are included in the Richfield sandy loam, the Pratt sandy loam, and the Pratt loamy sand. The Richfield sandy loam, however, includes also some material that is not of eolian origin.

Interruption of the Cycle

The eluvial phase may be interrupted at any stage by rejuvenation, and the eolian phase of a new cycle initiated. Wind attack Ill;ay be renewed whenever the vegetal mantle is locally weakened or broken, as by drought, prairie fires, the trampling by herds of buffalo or cattle, lowering of water table incident to stream incision, cultivation, or other factors. As a result, secondary blowouts develop, and primary dune forms undergo dissection and reworking. These secondary blowouts are of three general types, all gradational into one another: the spot, the linear, and the areal types. The spot blowout is simply a pit or crater-like depression near the top of a preexisting dune mound or ridge (pl. 22B). It may grow into a linear blowout, which is a steep-sided, elongate, trough- or scoop-shaped depression, deep in proportion to breadth. The sand blown.from this excavation may be spread out as a fan or apron, or may be heaped in a secondary mound or ridge, depending on the slope of the older dune and on the resistance of the vegetation. Where individual blowouts of this type are closely spaced, they may converge laterally, even to the extent of entirely obliterating the original dune topography and giving rise to a new generation of dunes. The transverse dune ridges west of Syracuse (pl. 27) were probably formed in this way.

Plate 27--Aerial photograph of transverse dune ridges west of Syracuse. The sand is moving toward the river. Note contrast between present river channel and channel in abandoned meanders. The arrow points north and is about 0.5 mile long. Photograph from the U.S. Soil Conservation Service. August, 1936.

Aerial photograph of transverse dune ridges west of Syracuse. The sand is moving toward the river.

The areal blowout (pl. 22C), in contrast to the other two types, is similar to the primary blowout of the first cycle. It is very shallow in proportion to breadth, and is irregular in outline. It is formed most readily where the preexisting dune topography was of low relief. In rare instances, vegetation may be so completely overpowered "as to permit the rise of barchans or of transverse dune ridges, as in certain abandoned fields in the "Dust Bowl", such as those described here (pl. 23C) and those reported by Whitfield (1939). The end product of this type of blowout may be indistinguishable from that produced by the aggregate effect of many closely-spaced blowouts of the linear type, but the intermediate stages are very different.

Secondary blowouts, like their primary precursors, may be halted at any point by stabilization. When this takes place, the eluvial phase begins anew, and proceeds as before. Blowout scars are healed and topographic discordances incident to secondary wind sculpture are gradually smoothed. Eventually, barring interruptions, the markings of rejuvenation are obscured, and the end forms of the second cycle come to resemble those of the first cycle, differing only in the texture of the topographic pattern if at all.

Multi-Cycle Dune Topography

Given time enough, and freedom from external interference, the dune cycle may be repeated as many times as conditions permit. A multi-cycle topography, indeed, is characteristic of the area under consideration. The sequence of events in different localities may be different, however. New, first-cycle dunes may spring up in one place while earlier dunes undergo the vicissitudes of advanced age elsewhere, and the dunes of one locality may proceed uneventfully toward eluvial old age while neighboring localities undergo one or more episodes of rejuvenation. Examples of this are found in the dune belt south of Garden City (pl. 28). Heterogeneity seems to be the rule. Although local and sporadic wind action has been in progress much of the time, it is probable that the major periods of dune building were of general effect, were distinctly separated in time, and were related to climatic fluctuations of regional importance. No detailed analysis is attempted in this paper, however.

Plate 28--Multi-cycle dune topography south of Garden City. The recency and extent of rejuvenation increase from left to right. Note changes in the texture of the topography as a result of rejuvenation. The area at the left represents an old-age topography. The dark spots are low places where moisture accumulates and vegetation is denser. Photograph from U. S. Soil Conservation Service.

Multi-cycle dune topography south of Garden City. The recency and extent of rejuvenation increase from left to right.

In interpreting dune topography, it must be remembered that it is only the records of partial cycles that survive in the lineaments of the final landscape. Each complete cycle destroys all topographic vestiges of its predecessors. Stratigraphic indications of earlier cycles, however, may be preserved as soil zones or unconformities in the dune, and it is only from this type of evidence that the earlier history of many dune areas is likely to be worked out.

Thus through the continued interplay of wind, sand, and vegetation the dune complex of southwestern Kansas was evolved. So far as known, the dunes were essentially fixed in position in the beginning, were never of the truly desert type, and assumed their distinctive characteristics through the all-important role of vegetation.

Agronomic Implications of the Dune Cycle

As a dune area, progresses through the eluvial phase of the cycle, its soil cover becomes progressively thicker and more stable, and eventually may permit cultivation. This depends partly on climate, however, so that, in the eastern part of the area, it seems probable that the safe point for breaking of the sod comes earlier than in the western part, where rainfall is less, and the strength of the wind may be greater. In late youth and early maturity, the dune is obviously not yet ready for any type of cultivation. It may, however, be used for grazing, provided that care is taken to prevent overgrazing and excessive localized trampling, as around watering places. In advanced maturity or old age, according to the climatic belt in which the dune lies, the soil is sufficiently deep and sufficiently well bonded with silt, clay, and organic material, and slopes are sufficiently gentle, to allow farming if sufficient care is exercised. Any local spots where the topography has been set back in the cycle by rejuvenation are to be avoided, however, for they represent vulnerable points where "blowing" is easily started. Until the dune mass as a whole is well advanced in old age, more than ordinary care must be exercised in farming practices, for even in the early-old-age dune area of northeastern Meade County several large areal blowouts of recent origin, some having sharp fence-line boundaries, may be observed. During periods of drought, or at places of very low rainfall, as in eastern Colorado, it is probably unsafe to attempt cultivation of dune-sand areas, however far advanced in the cycle.

Hydrologic Implications of the Dune Cycle

The effectiveness of dune areas for ground-water intake and recharge varies with their position in the cycle. The more advanced their place in the cycle, the less pervious the soil, and the less readily is rain water absorbed. Thus in maturity and old age, infiltration is reduced, and the amount of surface runoff is increased. The latter, having no outlet to the exterior, accumulates in the interdune depressions, forming temporary ponds, from which a part of the water is lost by evaporation. The more advanced the eluvial development of the area as a whole, the broader and shallower are the depressions, and the greater the loss from this cause.

Source of the Dune Sand

The dune sands of southwestern Kansas seem to have been derived from sources very close to their present position. There is no reason for believing that truly migratory dunes of the desert type were ever important, for the steep leeward bedding associated with such dunes is extremely rare, Even where present, this type of bedding does not necessarily imply dune migration, but may simply represent headward growth of an essentially static dune. It might be argued, of course, that migratory forms were once common, but were completely destroyed by blowout activity during progressive fixation. Although it is true that the topographic form of the dune could thus be effaced, it seems unlikely that there would have been such complete reworking as to destroy internal structure, and consequently this possibility may be dismissed for want of any substantiating evidence.

It remains, therefore, to discover a near-by source for the dune sand. It has been assumed by some previous writers (Darton, 1916, p. 42; 1920, p. 3) that the present river floodplains constituted this source. Along the Arkansas valley, this is certainly not true, for no movement of sand from the channel toward the dune belt is to be observed. In fact, there are few if any dunes of any consequence either on the floodplain or on the lowest terrace, and the dune belt is far too wide to have been supplied from the present valley unless there was extensive movement of migratory dunes. The dune belt on the south side of the river overlies the 20-foot terrace along the strip nearest the river, and extends southward onto higher ground. It. is probable that the sand was derived at least in part from the terrace deposits, but it is uncertain whether dune building took place when the present terrace was still a part of the floodplain, or after down cutting to a lower level. Toward the southern part of the dune belt, there was probably a different source or sources of sand. At different places the sand may have been provided by older and higher terraces, as yet unrecognized, or by Rexroad or Kingsdown sands, or simply by denuded slopes cut in the Ogallala formation.

The source of the dune sand in the small, scattered areas north of the Arkansas valley is uncertain. It is unlikely that the Ogallala could shave contributed unless the hard calcareous beds at the top were first eroded away, so as to uncover the softer sand beds below.

The sand of the small dune belt of Stanton County was probably derived from Bear creek when that stream flowed at a higher level.

The dunes of the Cimarron Bend area present a problem. In the belt within the river valley at the west, the materials were probably derived from floodplain and terrace deposits. On the broad upland areas, however, the source is uncertain. Possibly a part of this area is covered with post-Ogallala fluviatile sands. Possibly solution-and-collapse basins affected the development of ponds at some time in the past, and led to sufficient reworking of surrounding Ogallala beds to release some sand to be picked up by the wind.

The high-level dune belt in eastern Seward County and southwestern Meade County may have had a source similar to that in the Cimarron Bend area, or may have been related to high-level deposits of Cimarron river. The dune belt in southern Meade County, however, seems to be superimposed on the depositional surface of the Odee formation, and probably derived its sand from that formation.

The dunes in northeastern Meade County are closely associated with Crooked creek valley. It is possible that the sand was derived, either wholly or in part, from the strand flats of a lake, which, it is believed, may once have occupied a part of the basin. At the one point where a dune is well exposed in cross-section, the dune sand overlies a soil zone on loess, and its bedding shows a low westerly dip, suggesting that the sand came from the west.

The sand of the dunes in southern Clark County was undoubtedly derived from fluviatile deposits of Cimarron river and Big Sandy creek. At least in part, the dunes were probably built when these streams were flowing at higher levels than at present.

Direction of Dune-Building Winds

Winds of the Present

The effective sand-moving winds of the present time, contrary to opinions previously stated, are predominantly southerly. Evidence for this, in fact, was mentioned by Haworth (1897b, p. 279), but seemingly was underrated. Along the Arkansas valley west of Syracuse, it is clearly seen that the steep leeward sides of the transverse dune ridges face north, and that sand is now being drifted toward the river. Similar evidence is presented by recent blowouts in many other parts of the area. In some places, however, a subordinate amount of sand has been moved in the opposite direction, as along Cimarron river north of Elkhart. Such sand movement is more or less irreversible in part, for the sand is generally trapped by vegetation, and is protected by that vegetation from return drifting by reversed winds. Thus persisting, this sand reveals the effects of northerly winds despite the greater strength of southerly winds.

Winds of the Past

The dune-building winds of the past, in the greater part of the area, were from a direction opposite to that of the winds now effective. The great extent of the dune belt on the south side of Arkansas river, in contrast to the absence of any but a few small patches on the north side of the valley, itself points to development by winds from the north. The testimony of dune bedding agrees, for where low-angle backset bedding is to be seen, it commonly shows a northerly dip (pl. 22C). It is evident that dune-building dates back to a time when wind movement was different from that of today, and it is suggested that the presence of a continental ice sheet during one or more of the Pleistocene glacial stages would have provided a ready cause for altered wind directions.

At exposures in northeastern Meade County and in eastern Seward County (pl. 26A), the apparent dip of backset bedding is westerly. This may indicate winds of intermediate direction during an intermediate interval of dune-building (compare Melton, 1938).

Along the Cimarron valley in Clark County, and along Beaver river in Oklahoma (Gould and Lonsdale, 1926, p. 13), the dune belt lies on the north side of the valley. This suggests either that the northerly winds were less effective at these latitudes, or that the dunes in these areas were formed more recently, after the prevailing wind system of today had been established. The early mature U-shaped dunes south of Englewood (pl. 25), for example, indicate a wind direction differing little from that of the present. These dunes may correspond to the intermediate series described by Melton (1938) from the southern High Plains. Correlation with his oldest series, however, is uncertain.

Sinks and Depressions

Sinks Developed within Historic Time

At least two sinks are known to have developed in the area within the last 70 years. The more recent of the two is situated in southern Hamilton County, about 11 miles south of Coolidge (NE corner sec. 22, T. 25 S., R. 43 W.). This sink is reported to have been formed in 1929 (Bass, 1931). Originally, it is said to have had a diameter of only about 60 feet, and lay just beside a county road. Subsequently its diameter has increased to about 200 feet, and it has engulfed the road, necessitating a slight detour. When visited by Bass in 1930, the sink was 40 to 50 feet deep, had overhanging walls, and contained a shallow pool at the bottom. Today, it is filled with water to a level within about 10 feet of the ground surface (pl. 29A), and the rim has been modified by slump and by short, steep gulleys. There is no visible overhang.

Plate 29--A, Sink hole 11 miles south of Coolidge, 1939. B, Solution-and-collapse crack across road about 6 miles north of Ashland, 1938. C, Characteristic features at the head of a draw, in northern Clark County.

Three black and white photos; top is sink hole 11 miles south of Coolidge, 1939; middle is solution-and-collapse crack across road about 6 miles north of Ashland, 1938; bottom is characteristic features at the head of a draw, in northern Clark County.

Bass (1931) postulated that this sink was formed by solution of the Greenhorn limestone. Landes (1931), however, presented evidence that the solution more probably took place in pre-Dakota salt or gypsum beds. Bass suggested that the sink represents renewal of movement in an older, broader sink, and I found, from examination of an aerial mosaic, that it is but one in a linear series of sinks, the relations of which are such as to indicate that it occurs along the line of a post-Ogallala fault.

The other sink of historic origin is situated on the east side of Crooked creek valley, less than 2 miles south of Meade (Johnson, 1901, pp. 706-710). It was formed in 1879, and was known locally as the "Salt Well", having filled with strongly saline water to a level about 14 feet below the ground surface. For a time it was used locally as a source of salt (St. John, 1887, p. 135). When described by Johnson in 1901, the Salt Well was a cup-shaped depression 150 to 200 feet in diameter, and about 35 feet deep to the level of standing water, which was 9 feet deep. The sink was surrounded by a broad zone of roughly concentric sod cracks. Further details may be found in Johnson's excellent photographs, map, and description. Today, the appearance of the sink is very different. The sides are deeply gullied, and the bottom has been filled, containing now only a shallow pool of stagnant rain water.

This sink is believed to have been formed by solution of underlying Permian salt beds. According to Johnson, a test well on the rim encountered bedrock at a depth of 292 feet, and gave strong indications of salt in the next 16 feet.

Other sinks may have formed within historic time, but have received less publicity. A phenomenon probably related is the development of deep cracks in the ground. Such a crack was formed across a county road near the edge of the upland about 6 miles north of Ashland, in 1938 (pl. 29B). The observed crack was as much as 8 feet deep and 2 feet wide. Considerable fill was necessary to make the road passable. Similar difficulties were encountered on the state highway through Big Basin, described below.

Big Basin and St. Jacob's Well

Big Basin and St. Jacob's Well, in western Clark County (secs. 24 and 25, T. 32 S., R. 25 W.) are perhaps the best known and most accessible of Kansas sinks. Both were figured by Johnson (1901, pls. 134, 135), and his photographs have been reprinted in other publications to illustrate solutional topography.

Big Basin (pl. 30) is a subcircular undrained basin about 1 mile in diameter and about 100 feet deep. It is crossed by U. S. highway 283. The rim on the east, south, and west sides is notched by small gullies, and at the north there are deeper and longer ravines, which probably represent beheaded segments of streams that crossed the site of the sink before subsidence took place. Except for minor undulations, the floor of the basin is essentially flat. At times it contains shallow, wet-weather ponds. In the sides of the basin, Permian, Cretaceous, and Tertiary rocks are exposed. Near the east side of the basin, a renewal of the downsinking is indicated by a small, steep-sided hole in the floor, seemingly of very recent origin. On the sides of this hole, moderately coarse gravels are exposed.

Plate 30--Aerial photograph of Big Basin and St. Jacob's Well, in western Clark County. Photograph from U. S. Agricultural Adjustment Administration, 1938.

Aerial photograph of Big Basin and St. Jacob's Well, in western Clark County.

Although Big Basin is undoubtedly of solution-and-collapse origin, it is uncertain whether it was formed by large-scale or by piecemeal downsinking. Its close association with smaller sinks at the southwest and at the east suggests the possibility of gradual coalescence of a cluster of smaller sinks, although its outline is less irregular than might be expected on that basis. In any event, the floor of the basin has undoubtedly been smoothed by the deposition of a veneer of fluvial sediments carried in by the streams from the north and by the gullies around the other sides.

The age of Big Basin is geologically not great, but undoubtedly ranges back several hundreds or even thousands of years. Some time must have been required for the streams at the north to have cut down from their original grade to the level of the basin floor, to which they are now graded.

St. Jacob's Well is a smaller sink, just east of Big Basin. The "well" is a deep pool of standing water. The sides of the sink are deeply gullied, and its age may be equal to that of Big Basin.

Other Depressions

Shallower and less striking basins than those described above are extremely common in many other parts of the area, and occur to some extent virtually everywhere in the upland areas. They are especially numerous, however, along the Scott-Finney depression, and in the Odee district (see Meade topographic sheet), where some are as much as 1 mile long, but are shallow and saucer-shaped in contrast to Big Basin. Some hold temporary ponds, and many have a marshy type of vegetation. These depressions are probably a result of subsidence due to solution of salt or gypsum beds in Permian or early Mesozoic formations, or possibly, in the case of the Scott-Finney depression, of calcareous beds in the Cretaceous. Some of the depressions may represent actual sink holes that reached the surface, and were later filled by slump and wash from bordering areas, but probably the greater number represent merely the sagging of yielding roof rocks over cavities that never reached the surface. They are probably more or less analogous to the subsidence area formed in Hutchinson as a result of extraction of rock salt at a depth of about 300 feet by the pumping method (Young, 1927).

It was suggested by Johnson (1901, p. 711) that "the innumerable upland basins, especially where the floor is Cretaceous to great depths, are clearly to be ascribed to grain-by-grain processes of readjustment and compacting, at work within the Tertiary only." Convincing proof of this process, however, is yet to be adduced.

Large Imbricate Blocks of Bluff Creek

About 0.5 mile south of the point where Bluff creek makes its sharp bend to the south, the stream channel shows a unique occurrence of large imbricate blocks (pl. 31A, B). Large, angular slabs of Cretaceous limestone, as much as 5 feet long, are stacked shingle fashion along the bottom and in the banks of the channel. Virtually all the slabs dip upstream, at angles as steep as 35°. The individual blocks are about 6 to 10 inches thick. Similar though less striking imbricate structure is found at various points for about 200 yards downstream, and for about 350 yards upstream, to the point where a limestone ledge across the stream forms a falls about 2.7 feet high. The rock in this ledge is similar to that of the blocks. The stream gradient here is about 40 feet per mile. It was not definitely ascertained whether the blocks are being moved by present floods, but it seems doubtful that more than minor readjustments are effected.

Plate 31--A, Imbricate blocks in channel of Bluff creek. B, Imbricate blocks in the bank along Bluff creek. C, Concrete slabs carried downstream by abnormal overflow waters after the breakup of the spillway of the dam at Meade County State Lake. The water moved from left to right. This illustrates the way in which the imbricate blocks were probably deposited.

Three black and white photos; top is imbricate blocks in channel of Bluff creek.; middle is imbricate blocks in the bank along Bluff creek.; bottom is concrete slabs carried downstream by abnormal overflow waters after the breakup of the spillway of the dam at Meade County State Lake.

Two possible explanations suggest themselves for the origin of the imbricate structure: (1) ice-jam action, and (2) torrential flood velocities. The former, although at first favored, was finally abandoned for lack of any supporting evidence. The likelihood of the latter was substantiated by the finding of concrete slabs in a comparable attitude a few tens of feet downstream from the dam at Meade County State Lake. The slabs were derived from the breakup of the spillway, and were carried downstream by overflow waters of abnormal volume and velocity. Although only a very few slabs were involved, their size was comparable to that of the imbricate blocks, and their final disposition was similar (pl. 31C).

Under the hypothesis of flood velocities, the blocks would have been loosened by undermining incident to falls-recession, and their imbrication would have resulted from short-distance transportation by floods having abnormally great velocities owing to local steepening of hydraulic gradient by the falls. This would involve transportation of individual blocks only a few feet or a few tens of feet at most, leaving them stranded at the point where velocity declined and carrying-power was reduced. As the point of heightened flood velocity receded with the falls, new crops of blocks would be stacked at points successively farther upstream, and older ones would undergo only minor jostling by undercutting.

At this point, it might be asked whether any imbricate blocks were actually found immediately in advance of the falls noted above. A few tens of feet below the falls, one small group was found embedded in the stream bank, on the outside of a slight curve in the channel. Under the lip of the falls there were a few flat slabs. Perhaps these were detached at a time of relatively moderate flow, and were awaiting imbrication by the next major flood. Although the display of imbrication here is less striking than might be expected, it is consistent with the hypothesis. A consideration of the processes involved indicates that the progress and continuity of imbrication would vary with the following factors: (1) size of available blocks, as governed by joints, thickness, and rate of detachment; (2) changes in the height of the falls during recession, owing to convergence or divergence of stream grade with the governing bed of hard rock; (3) chronologic spacing of minor and major floods; (4) lateral shifts in the line of swiftest current; (5) depth of channel scour and fill, in relation to opportunities for burial of imbricate blocks along some stretches.

A few hundred yards west of the bend in Bluff creek, an accumulation of somewhat smaller imbricate blocks was noted. No falls was found immediately upstream, but the channel was observed to be floored with bedrock in that direction, and a noticeable steepening of channel gradient was seen to be associated.

In summary, the giant imbricate blocks, for want of any better explanation, are attributed to the action of occasional major floods of locally accelerated velocities on slabs detached by falls recession. Plunge pool action may give the channel bottom an upward turn below the falls for a short distance, and the up drag of the flat blocks on this slope may result in the characteristic upstream dip and overlap. Possibly at one or more times in the past, climatic conditions were slightly different, and more favorable to torrential floods.

Minor Valley Forms

In many parts of the area, minor valleys have the form of blunt-headed, round-bottomed, steep-sided "draws". The bottom is generally grassy, and there is no distinct channel. The rim is sharp, and the sides may present a scalloped appearance. Sod cracks, somewhat crescentic in plan, are commonly found above the heads of these draws (pl. 29C) and in places there are definite depressions also (see Johnson, 1901, pl. 139). Little opportunity for study of these features was afforded during the field work on which this report is based, and for a consideration of their origin the reader is referred to discussions by Haworth (1897 a, pp. 18-21), Fenneman (1922, pp. 126-132), and Rubey (1928).

Historic Changes in Stream Channels

Arkansas River

St. John, in 1887 (pp. 133, 135), described Arkansas river as occupying--

a broad shallow valley comprising an immense area of level bottom-land, usually presenting two low benches, the narrower of which forms the present flood-plain along the margin of the stream, and which affords valuable meadows. The stream itself presents a very uniform appearance-a broad sandy bed threaded by shallow channels and bearing grassy islets, and confined within low earth-banks. In places the bars are composed of gravel; elsewhere treacherous quicksands prevail. The melting of the snow in the mountains about the sources of the Arkansas fill its banks brimful of turbid, sediment-laden water in early summer, when the volume of the stream is at is greatest .... None of these streams today have any timber, except a few scattering trees along the courses of some of the north-side affluents, in Hamilton County. Indeed, along the banks of the Arkansas only a slender belt of cottonwoods is seen in the same quarter.

According to Mead (1896), the stream was navigable during these times of flood, and--

as early as 1852 boats were built at Pueblo, Colorado, in which mountain traders and trappers, sometimes in parties of 15 or 20 in one boat, with their effects, floated down the swift current of the river to Arkansas.

Today the appearance of the channel is much the same as it was when St. John wrote, but the banks are different in being well wooded, supporting heavy stands of cottonwood and other trees. It is somewhat doubtful whether boats could now descend the river as reported by Mead.

Haworth, in 1897, noted that some filling of the channel had already taken place by that time, and stated that, in the preceding 15 years, the stream bed had been raised by as much as 15 feet (1897a, p. 28). It has not been ascertained whether filling has continued since that time, but some channel shrinkage does seem to be in progress. From west to east, the width of the present channel decreases by one-half or more. This seems to be at least partly a result of the recent encroachment of heavy stands of scrub cottonwood and other vegetation, for sandy strips of a recently wider channel may be seen beneath the undergrowth.

Physiographic evidence for changes in channel regime is suggested by the abandoned meanders west of Syracuse-a feature rare along the river in this section. The abandoned stretches of the channel seem to be considerably narrower than the present sandy, braided channel alongside, and suggest that the stream formerly flowed in a simple, open channel. The date of the change is not known (pl. 27).

Farther east, in the Wichita area, changes in the channel of the Arkansas were noted as early as 1896 by Mead, whose observations dated back to 1859. He reported the transition from a stream once flowing bank-full most of the time, to a sandy waste traversed by an insignificant thread of water.

The cause of the changes outlined above undoubtedly lies in the extensive diversion of river water for irrigation in eastern Colorado and in the western counties of Kansas. It was Mead who first noted the importance of this factor.

Cimarron River

Along Cimarron river, the changes have been of a different kind. In 1887, St John (p. 133) described that stream as--

a small brook only a few yards wide, and in places during a portion of the year its waters are lost in the sandy bed. In early summer its low banks are sometimes overflowed. The above-mentioned affluents mostly afford pools the year round, and like the larger stream they are subject to overflow from the heavy local rains that occur during the summer months. In the past these pools were the resort of herds that pastured the adjacent plains, and the Cimarron valley was, until recently, entirely occupied by stock ranches and thousands of cattle.

Johnson (1902, p. 663) later referred to the Cimarron as--

A notable example of such a valley floor, almost unvisited by runoff floods, yet with a perennial stream of constant volume looping intricately upon it. . . . The spring stream here merely occupies the valley; it has no part itself in valley making. It does not run full length. Though in some of its live sections it is a strong stream, unvarying in volume, there are other sections in which the bed is permanently dry to depths of 20 or 30 feet. . . . It happens that the Cimarron valley floor lies approximately at the ground-water level, though not precisely. At one point it may be a little above, at another a little below.
A few miles above Arkalon, in southwestern Kansas, there is a feeble "reappearance", as it may be termed, of the Cimarron, following a long dry section.

Haworth (1897, p. 63), writing of the Cimarron valley in Clark County and parts of Seward County, stated that the--

Cimarron river carries a large amount of water during a part of the year, and is rarely dry in this part of the state, as it is fed by springs.

Parker (1911, pp. 306-307) later wrote that--

from the old post office of Metcalf, Okla., to Point of Rocks, Kans., a distance of 25 miles, the channel of the Cimarron is often dry, but at Point of Rocks, Kans., the water comes to the surface at Wagon Bed Springs, a famous camp on the old Santa Fe trail, and the channel is usually full for a number of miles. It gradually sinks again before reaching Oklahoma a second time. . . . In Kansas, from Arkalon southward, the Cimarron river usually has water in it throughout the greater part of the year. The stream is subject to a June rise, which is caused by the melting of snows in the mountains at its head. . . . William Easton Hutchinson states that the Cimarron river is a constantly running stream throughout the entire width of Morton County, where it has a valley on one side or the other of the channel from one-half to three miles in width, on which an abundant crop of natural hay is cut. In Stevens County there is running water in Cimarron river at all seasons of the year. . . . In Grant County the river flows constantly and has a fine fertile valley on each side of the channel that is sometimes covered by floods.

Although this account contains certain minor inconsistencies, the general picture is clear.

Early residents of the area agree that the Cimarron originally had a narrow channel, clear water, and many fish and beaver. Today, only one short stretch of the river, in southwestern Haskell County and contiguous sections, retains any semblance of these characteristics (pl. 5B). Even this stretch has widened noticeably since I first visited it in 1937. Elsewhere the channel has widened enormously, and at the west its width has progressively increased to as much as 800 feet (pl. 5A). The stream bed has become a barren sandy waste, and there is little or no flow during much of the year. The fish and beaver have vanished. Much valuable meadow land, where hay was formerly cut, has been destroyed. From time to time it has been necessary to lengthen bridges to span the broadened channel. Whether these changes have involved any deepening or shallowing of the channel has not been ascertained.

Changes of similar character are reported to have occurred along many tributary streams. Once flowing in grassy channels through many pools, they have been converted to dry sand beds (pl. 32B).

At least in part, the changes in the Cimarron are reported to have been started by a severe flood in 1914. Antecedent causes undoubtedly lay in regional weakening or breaking of the sod cover by grazing and farming, leading to more rapid runoff, decreased groundwater recharge, and lessened protection of the soft Tertiary sediments. Long-range climatic fluctuations may have been an antecedent factor also.

Plate 32--Sheet-wash and gullying phenomena: A, Erosion in roadside drainage ditch, Meade County. B, Retreating gully head near the edge of the upland north of Ashland. The bluffs as formed by Ogallala "mortar beds." C, Sheet-wash in cultivated field after heavy downpour, Meade County.

Three black and white photos of sheet-wash and gullying phenomena; top is erosion in roadside drainage ditch, Meade County; middle is retreating gully head near the edge of the upland north of Ashland; bottom is sheet-wash in cultivated field after heavy downpour, Meade County.

Prev Page--Structural Geology || Next Page--Economic Resources

Kansas Geological Survey, Geology
Placed on web Feb. 8, 2017; originally published September 15, 1940.
Comments to
The URL for this page is