1879.]

[Stevenson.

Surface Geology of South-west Pennsylvania and adjacent portions of West Virginia and Maryland. By John J. Stevenson, Professor of Geology in the University of the City of New York.

(Read before the American Philosophical Society, August 15, 1879.)

INTRODUCTION.

I. Benches along Cheat River and vicinity-Along the Monongahela River-George's Creek-Redstone Creek-Along the National Road-Chestnut Ridge-Youghiogheny River—Westmoreland County, west from Chestnut Ridge-Ligonier Valley-Somerset County-East from the Alleghanies.

II. Resumé of the Facts-The Horizontal Benches-The River Terraces -Conclusions.

The region in which the observations recorded in this paper were made, embraces of Pennsylvania the counties of Fayette, Westmoreland, Washington and Greene, with parts of Alleghany, Beaver and Somerset ; of West Virginia, parts of Monongalia, Preston, Mineral and Hardy; and of Maryland, the extreme western portion. But the observations were made chiefly in the four counties of Pennsylvania first named and in Monongalia county of West Virginia.

The Alleghany Mountains of Pennsylvania form the eastern boundary of Somerset county. West from these in the same county are Negro Mountain and the Viaduct Axis, which unite with the Alleghanies near the northern line of Maryland to form Prof. W. B. Rogers' Tygart's Valley anticlinal in West Virginia. This great anticlinal is divided by Tygart's Valley into two monoclinal ridges, known as Cheat Mountain and Rich Mountain.

Laurel Ridge, the next west from Negro Mountain in Pennsylvania, is a bold mountain separating Westmoreland and Fayette counties from Cambria and Somerset; it extends into West Virginia and dies out before reaching the Great Kanawha river. Chestnut Ridge passes through Westmoreland and Fayette counties at twelve or fifteen miles further west, and is separated from Laurel Ridge by the Ligonier Valley, a canoe-shaped synclinal, which disappears southward not far beyond the southern line of Pennsylvania. Chestnut Ridge becomes insignificant soon after entering West Virginia, but in Pennsylvania it is almost as imposing as Laurel Ridge or the Alleghanies. West from Chestnut Ridge, only the Saltsburg anticlinal is strong enough to affect the topography. It forms Brush Ridge in Westmoreland and Fayette counties, but disappears very near the line of West Virginia.

The drainage west from the Alleghanies belongs to the Ohio river system and the smaller streams are tributary to the Conemaugh, the Youghiogheny, the Cheat, or the Monongahela. The first of these large streams flows into the Alleghany, while the second and third are tributaries to the Monongahela, which unites with the Alleghany at Pittsburgh to form the Ohio.

Stevenson.]

[Aug. 15,

The Alleghanies of Virginia are reached in Maryland. They are not the same with those of Pennsylvania, but belong to the series of anticlinals next east. In this region, the Potomac river, rising near the Alleghanies of Pennsylvania, breaks through the Alleghanies of Virginia and flows past Cumberland on its way to the Atlantic Ocean.

Within the whole area examined, no traces of glacial drift occur, except along the banks of the Ohio river, where such material appears on the terraces, having been brought down by the Beaver and Alleghany rivers from the northern counties.

Some notes respecting the river terraces of the Ohio and Monongahela rivers were given in my report on the Greene and Washington district of Pennsylvania (1875). The observations begun in 1875 were continued during 1876 and 1877, and the surface geology was studied as closely as was possible without interfering with the economic investigations, which were the main object of the survey. Some of the observations are recorded in my report on the Ligonier Valley of Fayette and Westmoreland counties (1877). Barometric measurements were made wherever a terrace or bench could be recognized, but, in very many cases, these observations proved to be worthless as no spirit-leveled line was within reach to be used as a base. The measurements given in this paper were verified by direct reference to leveled lines and by repetitions.

I.

BENCHES ALONG CHEAT RIVER AND ITS VICINITY.

Cheat and Monongahela rivers unite in Pennsylvania at about two miles north from the line of West Virginia. Their channel-ways diverge rapidly, so that the former stream issues from Chestnut ridge at little more than three or four miles south from the State line, while the latter breaks through the same ridge at fully thirty miles further south. The course of Cheat is rudely north-west, and that of the Monongahela rudely north and south.

Stewartstown Benches. In the peninsula between these rivers and west from Chestnut Ridge, a line of high knobs begins at about a mile west of south from the mouth of Cheat and continues in an irregularly south direction to Dorsey's knob beyond Morgantown in West Virginia. These hills are nearly alike in height and reach to about 600 feet above the rivers at their junction. They are rudely conical and their sides are terraced. On one of these hills is the village of Stewartstown, at nearly three miles from Cheat river. In descending from this place to either river, the following series of benches was seen :

Ice's Ferry. Descending from the State line to Ice's ferry, on Cheat river, at nearly nine miles above its mouth, one sees the first, second, third, sixth, seventh and eighth benches of the series just given, but the fourth and fifth have been so disguised by erosion that they cannot be recognized. Above the third, there are no transported fragments, and the higher benches to the sixth are covered by an irregular deposit of sand with more or less clay. But no clay was observed on the eighth bench. Ascending the opposite side of the river and following the road toward Morgantown, one finds rolled and polished stones, mostly of small size, occurring in great numbers up to the level of the third bench, where they suddenly disappear. The ninth bench is reached on this road at about three miles from the river near the church and school house. The mile-ground, a level stretch about a mile long and nearly the same distance from Morgantown, is on the eighth bench.

Standing here and casting the glass around the horizon, one finds this bench marked by flat-topped hills far beyond Decker's creek, which enters the Monongahela at Morgantown, while the same bench is distinctly continuous along the face of Chestnut Ridge southward to certainly beyond the gorge of Booth's creek, which enters the Monongahela river at somewhat more than 12 miles south from the State line. As ascertained from the records of wells bored or digged on the mile-ground, the deposit consists of coarse sand, sometimes containing a little blue clay, in all from 4 to 22 feet thick.

The seventh bench of the Stewartstown series is reached further along the road, where the valley of Decker's creek is first reached, and it is continuous thence around the hills to the Monongahela side. The third bench is well marked at Morgantown and, opposite that village, decayed shells of unio have been ploughed up at that level.

Line Ferry. Here, at three miles from the mouth of the river, the seventh, eighth and ninth benches of the Stewartstown series are distinct on the West Virginia side, but the lower ones have been removed or they have been disguised by erosion so as to be unrecognizable.

In descending from the north towards this ferry, one has difficulty in recognizing any of the benches, as the hill is abrupt and covered with forest; but rolled stones appear suddenly and in great numbers at 210 feet above the river bed, and increase thence in size as well as in number to the surface of the stream; while on the opposite side, transported fragments are numerous up to the level of the fifth bench. On both sides the higher benches are covered with sand, showing no polished fragments, but containing many small pieces of rock belonging to the immediate vicinity, which have been little affected by weathering and evidently have not been subjected to the action of running water.

BENCHES ALONG THE MONONGAHELA RIVER.

Between the State Line and George's Creek. Coming from the south to Point Marion, where the Cheat and Monongahela rivers unite, one finds all the Stewartstown benches distinct except the third, which has been de

Stevenson.]

[Aug. 15, stroyed by erosion on the east side of the Monongahela, though it is still shown on the west side. The higher benches are almost equally distinct on both sides of the river. Polished fragments occur in great numbers up to the line of the fourth bench and, even at that elevation, many of them are of enormous size.

Crossing the Ferry at Point Marion and taking the hill-road to New Geneva, at the mouth of George's creek, one reaches the third bench at half a mile from the river. It is covered by coarse sand mingled with clay, which has boulders freely distributed throughout its whole mass. Such fragments are numerous up to 280 feet from the water's edge, the line of the fifth bench. The eighth bench is reached at somewhat more than a mile from Cheat river, and at the first cross-road it bears a deposit of sand whose thickness has not been determined. A flat-topped hill near this place marks the level of the ninth bench.

The fifth bench is of interest here. It is reached on this road very soon beyond the divide separating George's creek from Cheat river. There it bears a valuable fire clay with pockets of excellent glass sand. The irregularity of the deposit is shown by sections obtained in two neighboring pits. The one shows :

5. Clay, white, good, but containing ferruginous streaks 7 feet.
6. Clay, ferruginous, seen..

10

66

Respecting the thickness of the ferruginous clay, No. 6 of the first section, nothing is known further than that at 10 feet from the top the tools used in boring became hopelessly fast. An attempt was made to bore through the sand, No. 5, of the second section, but the tools could not be pushed beyond 15 feet.

At a little distance beyond the pits, the third bench of the Stewartstown series is reached, but the fourth bench has been masked by crosion so that the wash from the thick deposit on the fifth covers the place of the fourth and becomes continuous with the deposit on the third. Throughout the deposits on both benches, transported fragments are found in vast numbers. A well digged on the third bench is 23 feet deep and does not reach the bottom of the sandy deposit. At a little way beyond, the bottom of the sand is reached and the thickness to the rocky shelf is shown to be 35 feet.

The black clay of the first section, commonly known as the "swamp clay," was reached in a boring here at 25 feet from the surface. In it, amid the numerous fragments of half-rotted wood, the cupule of an acorn was found along with what seemed to be berries of the black haw. The latter were thoroughly carbonized and crumbled rapidly on drying; but the acorn cup was still tough, and it was kept for some time as a curiosity. The wood in this clay is tough enough to snap in breaking and it has the peculiar tint characterizing the half-rotted wood so frequently seen in peat bogs.

A point of the third bench projects towards George's creek at a little north from the road. Unio shells, much decayed, are common here and are barely covered by the soil.

A rise in the road at half a mile nearer to New Geneva brings it again to the level of the fifth bench. There a small deposit of glass-sand was found, but it was soon worked out. It rested on an irregular deposit of clay and sand.

This fifth bench is continuous along the river hills from the State line to the mouth of George's creek, except where cut away by streams, but sometimes it is so defaced by erosion as to be recognized only with difficulty. It is handsomely preserved at Greensboro', on the west side of the river opposite New Geneva, where it shows vast numbers of rolled stones. Mingled clay and sand occur above it to 300 feet above water-level.

Between George's Creek and Redstone Creek. Below Greensboro' on the river hill, benches occur at 20, 180, 265 and 310 feet above low water. At the mouth of Whitely creek, in Greene county, the third bench of the Stewartstown series, at 180 feet, is very distinct up the creek to the village of Mapletown, and polished fragments are numerous all the way. At the mouth of Muddy creek, in the same county, the fifth bench, covered with rolled and polished stones, is handsomely shown, and along the creek it is quite perfect as far as Carmichaels, where the detrital coating is as thick as it is along the river.

On Pumpkin run, in the same county, this bench is shown with the same features. The measurement given in the Greene and Washington report* is erroneous and the bench is confounded there with one which was not seen in the Stewartstown series, but which seems to be intermediate between the fifth and the sixth, and to be persistent along several of the streams in Greene county at 30 feet above the fifth bench. At the mouth of Ten-mile creek the first, third and fifth of the Stewartstown series are shown, together with the supplemental one just mentioned; while at Frederick, in Washington county, or rather, at two miles back from the river, and near Frederick, is a still higher one, which is clearly the same with the eighth of the Stewartstown series, though the erroneous measurement given in the Greene and Washington county report would make it intermediate between that and the seventh,

* During part of the season of 1875 I used a barometer which proved to be quite bad. In this way came the erroneous measurements referred to.

PROC. AMER. PHILOS. SOC. XVIII. 104. 2L. PRINTED SEPT. 26, 1879.