the western part the inclination of the surface is uniform and gentle, and but few channels of any description are found. The rapid inclination of the surface to the southeast throughout nearly all Meade County and the southern part of Seward County gives so great a fall to the streams that their erosive action is more pronounced, and consequently the surface has been changed to greater depth and a corresponding rugged topography produced. The rainfall is so meager that few of the streams have water in them as much as a quarter of the year. Those which have worn their channels deep enough to come into contact with the general underflow water have springs and seeps in great abundance along them, and pools of living water throughout the entire year. This wearing down probably has occurred, however, since the main part of the erosion was done, and consequently has exerted but a limited influence on the general physiographic features.

SAND DUNES.

One of the interesting topographic features frequently observed is that produced by the sand hills or sand dunes. On the south side of the Arkansas, throughout its entire length in this area, a strip of country varying from 4 to as much as 15 or 18 miles in width is covered by loose sand which has been blown by the winds into the hills and hummocks so common in sandy countries. The area is exceedingly irregular in its southern boundary. In some places it is not more than 3 or 4 miles across, while in others the distance is much greater. Near the western side of the Garden quadrangle the sand hills reach southward from the river almost uninterruptedly to within Haskell County, a distance of from 18 to 20 miles. Immediately south of Garden the sand hills extend only about 7 or 8 miles, where a strip of country is reached on which there is but little sand. But farther east, through the eastern tier of townships in Haskell County and Finney County, another southern projection of the sand hills area reaches from 12 to 15 miles south of the river. Still farther east, in Gray County, south of Ingalls and Cimarron, the sands likewise extend from 12 to 18 miles south of the river, or to within 5 or 6 miles of Montezuma. East of this area again, throughout the remainder of the Dodge quadrangle, the sand hills area gradually contracts in width, so that immediately south of Dodge it is only 4 or 5 miles in width. The extent and location of the sand hills are well represented on the topographic sheets of the United States Geological Survey.

In most places, both north and south of the river, outside of the area of river sand hills, relatively little sand is found on the uplands. In the vicinity of Wilburn and Fowler, however, east of Crooked Creek, an area of sand hills exists, covering 50 square miles or more, which has a general appearance similar to that of the river sand hills area. The character of the sand is practically the same, and in most respects the conditions south of the Arkansas River are duplicated.

Again, in the southeast part of Meade County, in the broad Cimarron Valley to the north of Englewood, an area of from 12 to 18 miles in width is covered with sand dunes which are practically the same as those on the south side of the Arkansas. The whole face of the country here for many miles up and down the Cimarron River, except some irregularly shaped areas, is covered with the sand. The exceptions referred to are peculiar and interesting. The greater part of the valley occupied by Colonel Perry's ranch has but few sand hills. Here we have an area 4 or 5 miles across, lying between the sand hills proper and the Cimarron River, the general character of whose soil is that of the alluvial soil common to the flood plains of rivers. Farther down the river to the east the sand hills approach almost to the river bank.

To the north of Englewood, where the sand hills are best developed, a portion of which territory lies within the Meade quadrangle, the sand practically covers the whole face of the country. It is blown into hills and valleys, irregular in outline and position, with no apparent indication of the directions from which the principal winds came. Here and there the sand is still blowing, producing barefaced hills with no vegetable covering, showing that the movement is still in progress. The greater part of the surface, however, is well covered with vegetation, which implies a cessation of the sand movement. The principal sand dunes seem to be residual in character— masses of sand left behind after the finer parts have been carried away by wind and water. For a fuller treatment of this subject the reader is referred to a discussion of the Physical Properties of the Tertiary, by the writer, in Volume II, University Geological Survey of Kansas.

PECULIAR ARROYO EROSION.

A peculiar form of valley erosion is noticed throughout western Kansas, a form thus far unobserved elsewhere by the writer, and one upon which no literature seems to exist. The lesser tributaries to the principal drainage channels in their uppermost course frequently are quite void of water almost the entire year. As a result of this, buffalo grass or blue stem entirely covers the bottom of the arroyos. Such arroyos usually are from 50 to 200 feet in width, even to their very sources. The peculiar and characteristic feature of such arroyos is the shape of the bank at the margins. In almost all instances over the whole western part of Kansas such arroyos have a vertical wall at the outside part of the bank, varying in height from 2 to 3 feet to a minimum of only a few inches. The whole bottom of the arroyo is, as a rule, covered with grass, and seems to have no corrasion marks along it. But on the outer borders the arroyo is separated from the main upland plain by the vertical wall. This feature is sometimes noted to a limited extent along the little sink holes which are so abundant in this part of the country. Some of these depressions,

measuring no more than 10 or 15 feet across, have the buffalo grass growing all over them, and have their walls assuming this vertical character, in every respect similar to the walls of the arroyos.

These interesting features of the physiography of the Tertiary plains seem to have been caused by the underground creeping of the looser sands and clay which are not held together by the grass roots of the sod. As water is so rare in the arroyos, the conditions are as favorable for the growth of vegetation in the bottom of the arroyos themselves as on the uplands. When the rains come, the ground is softened probably more in the arroyos than elsewhere, and is thereby made more easily movable. The existence of the water is of so short a duration that the mechanical action of its flow is not sufficient to corrade the surface. But as the inclination is generally quite steep, and as the water softens the clays and sands, gravity will cause a slow, but constant creeping downstream of the material which is not held in place by the grass roots. In this way the effect is similar to that which would be produced were a blanket spread from bank to bank of the arroyo, a blanket which was not removed or carried away by the drainage, but which would allow the grounds beneath to become softened and creep downstream by the influence of gravity. The blanket in question would move vertically downward as the material beneath it was carried away, and the vertical walls at the outside of the blanket would be maintained, constantly growing higher as the materials from beneath were removed by the downward creeping.

GEOLOGY OF THE AREA.

The general geology of this part of the State is now fairly well known. Excepting a few small portions in the southeast, the whole area is covered with Tertiary sands, gravels, and clays. The lowermost formation is the Red Beds, which are exposed along a few of the bluff lines and in some of the lowest valleys in the extreme southeastern part of the Meade quadrangle. Farther to the east, in Clark County and beyond, the Comanche overlies the Red Beds. This formation thins westward, however, so that only a few feet of the black Comanche shales is found anywhere within the Meade quadrangle, and that in the extreme eastern part. It seems entirely to disappear westward, as no traces of it have been found. Above the Comanche lies the Dakota, a formation consisting largely of sandstone. It is not exposed at the surface anywhere within this area, but has been reached frequently by wells, and is known to exist both to the east and to the west, so we are sure it is spread over the whole of the northern and central parts. The Dakota is followed by the Benton, which is largely a limestone formation consisting of beds of limestone alternating with black shale. It is found exposed at the surface in a few of the arroyos of Crooked Creek and along the Saw