The American Geologist/Volume 18/Origin of the high terrace deposits of the Monongahela River
ORIGIN OF THE HIGH TERRACE DEPOSITS OF THE MONONGAHELA RIVER.
By I. C. White, Morgantown, W. Va.
At the Minneapolis meeting of the A. A. A. S., in 1883, the writer presented a paper before Section E in explanation of the terrace deposits along the Monongahela river, as well as those along the old and abandoned Teazes valley, which extends from the Great Kanawha at St. Albans, along the C. & O. R. R. to the Ohio river at Guyandotte.
In that paper the origin of these deposits was referred to the hypothetical glacial dam in the region of Cincinnati, evidence for the existence of which had just then been published by Prof. G. F. Wright, of Oberlin, Ohio.
Continued studies of the river between the Great Kanawha and the Monongahela have still led the writer to refer the terrace deposits of the latter river to a glacial dam, but not to the one which Prof. Wright believes existed at Cincinnati.
It is now pretty surely established, through the work of Carll, Spencer, Hice, Foshay, Chamberlin, Leverett, and others, that the Monongahela, lower Allegheny, and upper Ohio waters drained northward into the lake Erie basin in pre-glacial time. The great ice-field which covered northern Ohio and Pennsylvania, and descended southward nearly to the Ohio river at Rochester, or Beaver, Pa., would, of course, effectually stop the northward drainage of this pre-glacial river, and impound the accumulating water into a vast lake-like reservoir, until it filled up to the level of any divides that might lead the surplus water across them to other drainage channels.
If any such old outlets exist, they would furnish almost a demonstration of the reality of this supposed glacial lake, which I have named lake Monongahela. The writer has often called the attention of geologists to a remarkable old gap in the divide which now separates the Ohio and Monongahela waters, at the head of the latter stream, just west of Salem, on the Baltimore and Ohio Southwestern R. R., thirty-six miles west from Grafton. It is a valley 300 to 400 feet wide, filled with clay deposits, and separating Ten Mile creek, on the Monongahela side, from Middle Island creek waters of the Ohio drainage. The summit of this coll, one mile west of Salem, is 1,102 feet, A. T., while the bottom of the cut through it on the railroad or bed-rock under the clay-covered summit, is 1,081 feet, A. T.
At the head of Rock Camp creek, another tributary of Ten Mile creek, which heads several miles (8 to 10 miles) north of Salem, there is another gap through the divide at exactly 1,143 feet, A. T., as determined by an experimental R. R. survey. Also, Prof. T. M. Jackson informs me that there is a third gap through the same divide, one mile north from the one on the Baltimore & Ohio R. R., and through this the North West turnpike passes.
The general elevation of the divide which separates the waters of Ten Mile and Middle Island creeks is 1,400 feet to 1,600 feet, A. T. Here, then, are three gashes or wiers cut down 300 feet below the general level into valleys of considerable breadth from whose level summits the waters now go in opposite directions.
The escape of the impounded water of lake Monongahela would sufficiently account for these old channels or colls. Curiously enough, there is another outlet at this same level (1,100 A. T.) near the head waters of the West Fork of the Monongahela, since the wide level summit at Arnold's, on the West Virginia and Pittsburgh R. R., is 1,095 feet, A. T., and the railroad is cut five to six feet below the level stretch of land which makes the divide between the Monongahela and Little Kenawha waters. This 1,100 foot divide at three or four different outlets from lake Monongahela, could hardly be a coincidence merely, but must represent the old colls through which the surplus water of the imprisoned rivers escaped to lower levels during glacial times. We thus have at hand an explanation of the immense deposits of stratified silts, clays, boulder beds, and other trash found at the junction of streams all along the Monongahela river, and especially above the level of the upper slopes of the river gorge, beginning at one hundred to one hundred and fifty feet above the present river.
A large stream, Decker's creek, joins the Monongahela at Morgantown, coming in from the east after cutting through Chestnut ridge and draining a large mountain area east from the latter. At its junction with the Monongahela it has dumped large deposits of sand, boulders, and clay, which after much erosion still retain, in favorable localities, a depth of seventy feet. From this fact, and also because these deposits were first studied there, I have named them the Morgantown beds, confining the name to the deposits which rest on the pre- glacial rock-floor of the river. These beds, which are often as distinctly stratified as the under-lying Coal Measure rocks, can be found wherever any shelf of the ancient rock-floor of the river has been preserved, from the head of the Monongahela to Pittsburg, and on northwestward along the Ohio, and up the Beaver until they are met and submerged by the vast deposits of the terminal moraine. About one mile north from Morgantown, and near the Flats school-house, these beds contain beautifully preserved fossil plants imbedded in a bluish gray pottery clay of impalpable fineness. The plant bed lies at an elevation of 240 feet above the present river, or about 1,080 feet A. T.
The small collection that I made of these plants and deposited in the West Virginia University, several years ago, together with collections made at different times by Prof. S. B. Brown, of the University, and also some collected by Mr. Walter Hough, of the National Museum, were all sent to Dr. F. H. Knowlton, the accomplished paleobotanist of the U. S. National Museum, Washington, D. C., for identification. Under date of Sept. 17th, 1895, Dr. Knowlton sent me the following account of these fossils:
Report on a Collection of Fossil Plants from Morgantown, West Virginia.
By F. H. Knowlton, Ph. D.
Some years ago I was informed by Dr. Walter Hough of the U. S. National Museum of the existence of finely preserved fossil leaves in the vicinity of Morgantown, W. Va. A few small fragments were brought to the Museum by Dr. Hough, but it was not until 1894 that a collection of any magnitude was obtained. These proved to be of so much interest that the collection in the West Virginia University has also, by the kindness of S. B. Brown, professor of geology, been placed at my disposal. This material has all been made use of in the following examination.
The study of these plants has not been quite completed, but enough has been done to transmit the following presentation. A few well preserved examples appear to be new to science or at least not readily identifiable with known forms.
Thus far ten forms have been determined with much certainty as follows:
1. Equisetum arrense L.
A small fragment of a sterile branch.
This species is distributed from Virginia to California and northward to Greenland, and is also found in Europe. It is very abundant and attains perhaps its maximum development in New England.
2. Cyperus sp.
There are a number of stems that belong evidently to some cyperaceous plant, but they are too fragmentary to be determined.
3. Potamogeton robbinsii Oakes.
There are a great number of fragments of stems and leaves of this species, all so well preserved as to leave no doubt as to the correctness of their identification. Its present distribution is from New Brunswick to New Jersey, north of lake Superior and northward.
4. Liquidambar styracifolia L.
A number of fruiting heads and leaves are referred to this species. Its present distribution is from Connecticut and Illinois to Florida and Texas.
5. Platanus occidentalis L. Sycamore.
Several fruiting heads are referred to this species. It is found from Maine to Vermont, south.
6. Ulmus racemosa Thomas. The White Elm.
A well preserved leaf is referred to this species. It is now found from Ontario and Vermont to Missouri and Kentucky. It is nowhere very common, but is most abundant at the north.
7. Quercus falcata Mich.
Two finely preserved leaves of this species. It is found from Long Island to Florida.
8. Betula nigra L. Black Birch.
There are several leaves of this species all perfectly preserved. The tree is now distributed from Massachusetts to Florida.
9. Fagus ferruginea Ait. Beech.
A large number of well preserved leaves are referred without hesitation to this species. It is a common tree from Nova Scotia to Florida and west.
10. Castanea pumila Mill. Chinquapin. A single leaf only is referred to this form. It now grows commonly from Pennsylvania to Florida and west to Indiana and Texas.
The species thus far determined all belong to living species. Some of them enjoy a wide distribution and are still found growing in the region, while others are now only found to the north. One in particular, Potamogeton robbinsii, is confined to the north.
It seems probable from the evidence of the plants that they were pushed down from the north during the ice invasion and were entombed, while the species has retreated again to the colder area.
The occurrence of Potamogeton robbinsii in these beds is of special interest, since it practically demonstrates that there was during Glacial times a movement of water from the edge of the ice near Beaver, Pa., southward along the Mononguhela valley through the escape wiers just described, which brought with it this northern plant. It is possible that a systematic search would bring to light many other such northern forms, as well as throw much light upon the slight changes that have taken place in species since the Glacial epoch, because there can be little doubt that these plants were embedded in their present matrix during the Ice age The particular locality in which the plants occur is near the head waters of two little streams which rise against each other, and then flowing north, empty into the Monongahela, their mouths being two and one- half miles apart.
The broad level summit between the heads of the two streams is covered with the clay deposits up to 251 feet above the present river bed, and at one locality (Mr. Baker's well) they have a thickness of 65 feet. The surrounding hills are made up of the soft shales of the Barren or Elk River coal measures, and it is in just such a sheltered bayou back from the main channel of the river that we would expect to find such deposits in lake Monongahela.
A fine quality of clay for common blue stoneware, or crockery, is always found among these deposits where any considerable stream empties into the Monongahela from the west (soft rock areas), while sand and boulders predominate at the mouths of those streams draining from the east (mountain or sand-rock areas). The celebrated pottery clays of Geneva and Greensboro, Pa., just north from the West Virginia-Pennsylvania line are deposited opposite the mouth of Dunkard creek, a stream entering the Monongahela from the west, and draining quite a large area of Permian shales. Here the following structure is exposed at the clay diggings back of Geneva in descending to the Monongahela river:
Rounded boulders, sand, and river trash
Soil, boulders, etc., to top of clay bed
Reddish clay, fine, tough
Clay mixed with sand, and brown iron ore
Bluish gray clay, fine, tough
Bed-rock to level of Monongahela river
The elevation of low water here is 772 feet A. T. so that the rock floor of the old river is now 912 feet A. T.
The same kind of clay as that at Greensboro and Geneva, and which occurs at 150 feet above the river on the Millan farm, West Morgantown, was analyzed by Dr. De Roode, chemist of the U. S. Agricultural Experiment Station at Morgantown. W. Va., with the following results:
Oxides of iron
Water and loss
At Morgantown, sixteen miles above Geneva, the level of the old rock floor at the head of High street is 916 feet A. T. or 129 feet above river level, and the same at the University buildings, which are situated on top of this old river channel. But on the west side of the river, at Keek's hill, where the deposit is 70 feet thick, the level of the old floor is only 905 feet A. T. This difference between Morgantown and Geneva is to be explained because of the soft shale floor of the old channel at Morgantown, which was eroded deeper than the hard sandstone of the old channel at Geneva.
At Uffington, three and one-half miles above Morgantown, near the mouth of Booth's creek, we find the level of the old floor on the hard Mahoning sandstone under a large deposit of sand (brought out of the mountains from the east by Booth's creek), at 915 A. T., or 125 feet above the present bed. Fairmont and Palatine at the junction of the Tygart's valley and West Fork rivers, 26 miles above Morgantown, are situated on the wide plain excavated by the pre-glacial river, and as might be expected, the old rock floor is covered to a great depth with terrace deposits, consisting of sand, rounded boulders, etc., from the Tygart's valley, and pottery clays from the West Fork river. The elevation of the old rock-floor is 972.5 feet A. T., or 122 feet above the present river, while the top of these deposits extends up to 1,067 feet A. T.
The terrace material at Fairmont has been well exposed along Fairmont avenue, and its stratified condition is well shown. At the Smith & McKinney building the excavation showed the ditferent stratified layers of sand and clay dipping about 10^ to the west.
A layer of clay, about eight feet thick, at near 1,000 feet A. T., has long been mined for pottery clay, in the Fairmont region. The corresponding bed at Morgantown has an elevation of 945 feet A. T.
At Clarksburg, thirty miles above Fairmont, the West Fork river is joined by a large tributary, Elk creek, and the two streams had carved out a wide valley in pre-glacial time, whose rock-floor under the city has an elevation of 986 feet A. T., or about 70 feet above low water (916 feet A. T.) at the junction of Elk and the West Fork rivers. Here a great deposit of clays and quick-sand, twenty-five feet thick, covers all the level surfaces up to 1,020 feet A. T. There are some layers of rather coarse sand in the deposits, but the clays predominate, and they are found all along the West Fork river, and its tributaries from Fairmont to Clarksburg and on beyond to Weston, thirty miles south, where they extend to only about thirty feet (1,020 feet A. T.) above water level (990 feet A. T.).
At Grafton, on the Tygart's valley river, twenty-two miles above Fairmont, the elevation of the top of the railroad pier in the river is 997 feet and the terrace deposits extend only about twenty-five feet higher.
The principal town sites along the Monongahela river from Weston to Pittsburg, viz: Weston, Clarksburg, Grafton, Monongah, Fairmont, Palatine, Montana, Morgantown, Point Marion, Geneva, Greensboro, Rice's Landing, Fredericktown, Brownsville, Belle Vernon, Charleroi, Monongahela, Elizabeth, McKeesport, Braddock, Homestead, all have their upper and more level portions situated on the ancient floor of the pre-glacial valley. It is this wide and almost level area of deposits, stretching from Braddock to the Monongahela, across to the Allegheny river via Homestead, and East Liberty, which forms the principal site of Pittsburg itself, as well as Allegheny, beyond. Here a remnant of this ancient valley floor is appropriately named Monument hill, rising, as in does, like an island between the present mouth of the Allegheny river and its filled-up and more ancient channel just north of the island. The bed rock under the terrace deposits of Monument hill is now 190 feet above low water or 890 feet above tide, and the rock floor under the vast terrace deposits at Bellevue, still further down the Ohio river comes at the same elevation, while the summit of the same deposit there, as well as in Pittsburg and Allegheny, has nowhere been observed above 990 feet A. T., or 290 feet above low water in the present rivers.
This old elevated valley floor can be followed down the Ohio river to Rochester, and up the Beaver river past New Brighton, Beaver Falls, Rock Point, Wampum, Mahoningtown, and New Castle, beyond which, up the Shenango, it sinks out of sight under the drift-filled valleys, and at Sharon is submerged by 60 feet of cover, 40 feet below the present water level, or 780 feet A. T.
The exact course of the pre-glacial river from Sharon northwestward to the lake Erie basin has not yet been clearly delineated, but Mr. Leverett's studies leave little doubt that from Sharon the course was northwestward through the low drift- filled divide at Warren, O., (now only 900 feet A. T., and its rock floor probably 200 feet lower), and thence northward along the general course of Grand river.
Just how long lake Monongahela existed, and drained its surplus waters southward through the Middle Island and Little Kanawha gateways, cannot be estimated except by the deposits just described; but finally the barrier along the upper Ohio, (probably at the "narrows" below Moundsville, W. Va., as believed by Profs. Chamberlin and Leverett) gave way, and the level of lake Monongahela was speedily lowered by the rapid cutting away of the soft rocks along the present Ohio valley. Just how much of Ohio river history dates from this interglacial origin, it is yet impossible to determine, but the character of the topography along the river, everywhere between Rochester, Pa., and Cincinnati, O., (below which the writer has not studied it), would lead to the conclusion that all this portion of the river is new, and that the conclusions of Prof. Tight, of Denison University, Granville, O., are well maintained, viz: that in pre-glacial times the waters of the Muskingum, Scioto, and other rivers of Ohio, all flowed northward into the lake Erie system.
This being true, it would follow as a corollary that all of the West Virginia rivers which rise on the western slope of the Allegheny mountain-divide, including the Little and Big Kanawhas, and the Big Sandy, must have once gone northward in pre-glacial times and joined the Scioto, Muskingum, and other Ohio streams in their northward course, right athwart the present valley of the Ohio, but on a plane probably 200 to 300 feet above the present rock tloor of the Ohio. The search for these former high level valleys north from the Ohio, as bisected by it, is reserved for a future chapter, and it is possible that therein may be found an explanation of the old and abandoned high level Teazes valley, which does not admit of the explanation suggested by Prof. William M. Davis, in a recent number of "Science," viz: stream piracy of Coal river by the Great Kanawha, because the old valley in question is filled with transported boulders and gravel that could have come only from the crystalline rocks of the Blue Ridge in Virginia, and hence the Kanawha itself must once have flowed along Teazes valley, and been captured by another stream which led it northward to the newly cut Ohio valley, at Point Pleasant.
If the conclusions here inferred could be fully sustained by further study of the Ohio river system, it will be readily perceived that a pre-glacial drainage map of Pennsylvania. Ohio. West Virginia, and Kentucky would bear very little resemblance to a map of the present drainage, for if the Monongahela, Lower and Upper Allegheny once went northward into the lake Erie basin, there can be no doubt that the Upper Susquehanna also took this northward route to the sea, in pre-glacial times, and that much of its present course through Pennsylvania is new. The resultant modifications which such a change of direction in a great river system must necessarily bring to pre-existing drainage lines, are so varied and amazing even in contemplation, that their study must furnish many problems of surpassing interest to topographers of the modern school, so ably typified by Prof. Davis and his work.
he following table represents the descent of the ancient Monongahela river toward the lake region for a portion of its course between. Weston, W. Va., and Sharon, Pa., as well as the present water levels between the same points, and also the highest levels of the Morgantown beds, where known:
This table shows that while the present river falls 290 feet between Weston and Pittsburg, the floor of the ancient stream descended only 110 betwen the same points, or only about one-half foot per mile, which is practically that of a base-leveled valley, or one quite advanced in age, to say the least. The rapid descent of the modern river below the old rock floor of the ancient one gives rise to the steep slopes and gorge-like character of the Monongahela valley every where between Weston and Pittsburg, while the same is true of the lower courses of all the larger tributary streams, the descent in these being very rapid in the last few miles of their courses, while in their upper reaches the slope is gentle and the topography much older looking.
The smaller streams, which rise only five to ten miles back from the Monongahela, descend to it in the last mile or two of their courses by a series of rapids and cascades, while the still smaller brooks and rivulets have cut more notches in the bounding valley walls below the level of the ancient rock floor of the Monongahela, so that every feature of the topography tends to confirm the conclusion that the age of the river below the old base-leveled plain shown in the table, has been of comparative short duration, and must be measured by only a few thousand years at most.
The writer had intended to present a map with this paper, showing in a crude way the pre-glacial drainage of the areas herein described, but so much detailed work must yet be done before any approach to accuracy can be hoped for that the mapping is left for other and more skillful workers in this promising field. In this connection I would call the attention of students of the subject to the former northward course of the Little Beaver and the Slippery Rock.
The former once went northward along with the Monongahela drainage, but being dammed up by the northern ice it cut a new channel about fifteen miles long, southward into the Ohio river drainage, and this fully accounts for the wonderful change in topography along the lower portion of its course, as well as for the southward transportation of large granite boulders several miles beyond the limits of true glacial movement, thus giving rise to the phenomena called the "fringe" by Lewis and Wright ; for wherever a stream was thus impounded the blocks of ice floating southward across these temporary lakes from the terminus of the glacier would of course bear away and, melting, scatter over the surface many masses of imbedded rock.
The Slippery Rock now joins the Connoquenessing in a curious manner, by meeting it direct, and the combined stream turns off at right angles to enter the Beaver at Rock Point. In pre-glacial time the Slippery Rock left its present channel a short distance above Kennedy's upper mill, and following the present valley of Big run, turned northwestward near New Castle, and cutting squarely across the present gorge of the Neshannock (which is also of post-glacial origin) passed two and one-half miles north of New Castle and entered the ancient Monongahela near Harbor Bridge. All this is fully attested bv the wide drift buried valley, which can be followed from Slippery Rock clear through to the Shenango, and the course of this ancient stream also confirms the conclusion of Leverett that the pre-glacial drainage took the Shenango route via Sharon instead of the Mahoning, and that the lower portion of the latter stream is of recent origin. The Slippery Rock was turned southward at Kennedy's upper mill by the advancing ice-sheet which filled its pre-glacial valley, and on retreating left a dam of drift 100 feet high across its former course, while the wild gorge and raging stream between Kennedy's mill and the Connoquenessing, so different from the wide valley and sluggish current above the mill, though the outcropping rocks are the same, fully attest the recent origin of the twenty-mile cut between the mill and the Connoquenessing, near Wirtemburg.
- A "G. S. A." paper, read before Section E. of "A. A. A. S.," at the Buffalo meeting, August, 1896.