west by a very sharp descent from the divide between itself and the Little Missouri. From the top of this divide one sees, extending far to the eastward, an oval, ampitheatre-like valley whose general level is from 100 to 150 feet below the rim of buttes in every side. This whole basin-like area was apparently the watershed for Government creek. Near the center of the valley a lone butte lifted its summit to a level with the rim surrounding the valley. From this point one gets a view of the valley and the highland to the east, north, south and west. The low, round, dumpy, red-topped clay buttes of the valley surround the massive butte in the center and continue on to the east, a tumbling sea of irregular forms, gradually rising in elevation until they merge with the plains several miles to the eastward. On the north, one long, continuous ridge bounds the valley. Far to the south west, Sentinel and Square buttes are still visible, along with another very large butte in the north west. this valley, depressed so distinctly below its surroundings, we evidently have erosion forms of a later cycle imposed upon the remnants of an older and higher base level. Not the least striking feature of this valley is the great extent of burned clay. Most of the low buttes are red-topped and often capped with sandstone, which has assumed a jointed, columnar, structure; or, in some cases, it is fused into a conglomerate mass, due to the effect of heat arising from the burning coal beds beneath.

The Effect of Burning Lignite. The agent that effected this widespread change in the clays and the sandstone is still at work on the area. At the west side of the valley a ten-foot coal seam outcrops at the head of a draw, and a mile further to the east is found burning. The effect of this burning is not only to lower the level over the area burned by the depth of the coal seam, but to loosen and rearrange all the material lying above the seam, by causing it to cave in, and also by fusing a large part of the more fusible material into scoriaceous masses of large size. These jagged, angular masses resist weathering and erosion, and remain in many parts of the area to add greatly to the roughness. The loosening of the material also makes possible an underground seepage along the level of the burned coal, and a further cracking and loosening of the surface from this cause. An acre or more at the edge of the burning coal was thus depressed below its surroundings. The buttes themselves at this point were being rapidly undermined by the burning of the lig

nite. The great extent of burned clay throughout this whole region leads one to believe that the bad land erosion is aided in no small degree by these fires in the lignite. The burning of these seams of lignite goes on but slowly. The seam above referred to has been burning in this particular place since 1882, and yet but little more than an acre has been consumed. The coal burns slowly for the reason that the process goes on at a greater or less depth below the surface, and consequently in a somewhat smothered condition. Other beds were seen burning under conditions which indicate that but little progress has been made by the fire for several years. The valley above referred to contains from sixty to eighty square miles. If the fires which have burned the larger part of the coal at the present level in the area were burning in only a few points at the same time, the period necessary to cover the valley would certainly be several thousand years. But this period would be short compared with the time that these hills have been subject to the slow process of rain and erosion. Relatively, therefore, the effect of burning coal in reducing the area to the bad land conditions would be a rapid process.

FACTORS IN THE FORMATION OF THE BAD LANDS

The more important factors in the formation of the bad lands and the steep slopes of the valleys are the following:

The conditions under which these factors work are as follows: 1. Arid climate.

2.

3.

4.

1.

Materials of rather uniformly fine texture.
Horizontally bedded materials.

The presence of irregular concretionary forms.

Streams.-Erosion in the bad lands is associated directly with stream action, as in any other region. Steep slopes can be formed only where ravines and gulches have developed. Slopes and run-off are inseparable. No depressions occur anywhere, except sloughs which are to be associated with the drift, that cannot be traced to the action of the streams. Streams are necessary agents in the removal of material that is washed down

the slopes, and serve to direct the course of the run-off. The first work on all groups of bad lands, then, began at the valley of the master stream, and the future of every gulch can be inferred from the forms seen along the ravines and creeks. Just as surely as the torrential rill is the parent of the creek, so the gulch formed by an afternoon shower is the ancestor of miles of impassible bad lands.

The river is at work all the time, the larger creeks part of the time and the small creeks occasionally, in transporting materials. Heavy showers temporarily overload the streams, but they accomplish their task by continuing to flow long after the showers have passed. Thus the valleys do not fill up and most of the debris of one shower is carried off before the next one occurs. The melting of spring snows supplies much water, but so gradually that the streams during this season are probably receiving less material than they can carry, and load up with materials deposited at other seasons.

2. Torrential Rainfall and the Streams.-Meteorogical observations show that not only does most of the rainfall of this region come during the months of May, June and July, but that the characteristic rains are thunder storms; these furnish much rain on small areas during short periods of time. Several times during the trip down the river, five or six distant thunderstorm centers were counted, hanging over restricted areas of the valley. The waters from these showers get rapidly into the stream, as was shown again and again by a raise in the river following immediately on the passing of the storm. Two series of showers occurred on the afternoons of the 21st and 22nd of July, both too far up the stream to give much rain at the point where the party was encamped; but as a result of these two storms the river rose about five feet, filling its channel from side to side with a deep, brown, muddy current that hurried swiftly by, carrying brushwood and all sorts of debris picked up from the flats. Thus it is that most of, the run-off, and the work accomplished by it, is concentrated into a very few hours. The steep slopes along the creeks are affected practically only while the storm lasts; the creeks carry much water for a few hours, and the river is swollen for sevaral days after the storm has passed. The river is kept at a more regular flow than the creeks, because showers are more uniformly distributed along its valley.