The geology of the oil regions of Warren, Venango, Clarion, and Butler Counties; including surveys of the Garland and Panama conglomerates in Warren and Crawford, and in Chautauqua Co., N. Y., descriptions of oil well rig and tools, and a discussion of the preglacial and postglacial drainage of the Lake Erie country

Front matter

SECOND GEOLOGICAL SURVEY OF PENNSYLVANIA: 18T5 to 1879.

.THE GEOLOGY

OIL

Table Of Contents

Pages .

Chapter 1. The topography and drainage of the outcrop of the Garland Conglomerate and Lake Erie, 1 to 10

Chapter 2. The Garland Conglomerate ; survey of its outcrop ; place in the series ; part of the Pottsville Conglomerate formation No. XII; underlying Sub- Garland sandstone, . 11 to 19

Chapter I.*

The topography and drainage of the country lying between the northern outcrop belt of the Garland conglomerate and the Lake Erie divide, in Warren and Crawford counties.

§ 1. A line drawn from Sngar Grove in Warren county to Jamestown in Mercer county represents approximately the northwestern margin of the outcropping Garland conglomerate as instrumentally traced from New York to Ohio in the second year of the survey, 1875.

§ 2. The bearing of this line is about S. 58° W., N. 58° E. being parallel with the average trend of the shore of Lake Erie and distant from it say 30 miles. (See map Plate I.)

§ 3. The outcrop of the Garland conglomerate, thus indicated, is, however, by no means an unbroken escarpment ;

* Report of the work done in 1875.

but a range of hilltops or outliers, separated by numerous shallow vales, through which streams flow towards the south; all of them being tributaries of the Ohio river ; and some of them having their head-springs within 8 or 10 miles of Lake Erie, on the crest of the "great divide" or water-shed, the gently sloping southern side of which is drained into the gulf of Mexico, while its steep northern slope is drained into the gulf of St. Lawrence.

§ 4. The original gaps in the ranges of conglomerate capped hills were gradually, in course of time, deepened and widened into valleys between elevated upland ridges ; and these ridges in their turn were worn down and cut/ through sideways, more or less, by the general rain- fall and smaller lateral streams, until now only an occasional detached outlier of the Garland conglomerate (or sandstone) remains on the crowns of the highest hills along the line above referred to.

§5. Pike's ridge, an outlier of this kind, the most northwesterly in the State of Pennsylvania, may be found about 2 miles from Wrightsville and 4 miles from Sugar Grove in Warren county, on the divide between Little Brokenstraw and StilhvUter creeks.

§ 6. To the north and west, the Garland conglomerate, with all the rocks above it in the series, has been removed by the general erosion of the country, which, deprived of its protection, has been carved into broad valleys and gracefully rounded hills and ridges of the softer and more homogeneous measures below it. The contrast is very striking to the sharp and rugged topography of all the country to the southeast, where most of the higher eminences are still protected from erosion by a considerable capping of the massive, current-bedded, micaceous sandrock deposits, underlying the Coal measures.

§ 7. At the Olean Rock City, in Cattaraugus county, N. Y. is found the next prominent exposure of the Garland conglomerate to the north and east of the one last mentioned. It bears N. 80° E. from Pike's ridge, distant 47 miles, and near the Pennsylvania State line, about 6 miles southwest of Olean.

TOPOGRAPHY AXD DRAINAGE. III. 3

§ 8. If now we start at the Olean Rock City and glance southwesterly along the line indicated on the map (Plate I) first to Sugar Grove, and thence to Jamestown, we notice that the principal streams cutting the line come in the following order : I. Tunangwant creek, flowing north into the Allegheny ; 2. the Allegheny river, flowing south ; 3. Conewango creek, delivering the waters of Chautauqua and Cassadaga lakes southward into the Allegheny ; 4. Little Brokenstraw creek ; 5. Blue Eye run ; 6. Big Brokenstraw creek ; 7. Oil creek, with two branches ; 8. French creek ; 9. the Conneaut outlet ; 10. Crooked creek, and 11, the Shenango river, all of them flowing south.

§ 9. More or less of the Garland conglomerate (or sandstone) remains in place along the line, on all of the ridges which intervene between these streams ; and in some instances (as in the vicinity of Meadville) spurs of hills capped with it extend still further north. The actual outcrop therefore, if followed along its windings, would make a very irregular line upon the map ; while several isolated patches exist to the north of the general strike-line completely cut off from the main body by lateral streams running nearly at right angles to the leading arteries of drainage.

Spring cresJc in Warren county, and Hyde (or Little Oil) creek, in Crawford county, are lateral streams of this character. The crown of the southerly water-shed of these two streams trends in nearly a due southwest course from Big Brokenstraw to Oil creek, and is virtually an unbroken ridge capped with the Garland conglomerate ; while over all the country to the northwest the rock lies only in detached masses on isolated knobs of the spur-ridges.

§ ] 0. Valley features. —To repeat what was said above, the configuration of the country and the character of the streams north of the main outcrop line of the Garland conglomerate are in striking contrast with those south of it. To the north, the erosion of ages, after accomplishing the complete destruction of the sandstone, has attacked the underlying shales, producing innumerable hills rounded to a graceful contour ; the valleys are broad, with bottom

4 III. JJKPOKT OF PROttKESS. JOIIX F. CARLL.

lands through which the sluggish streams sweep in curves and loops and flow almost entirely between banks composed of lacustrine or fluviatile re-worked Drift. Scarcely an exposure of bed-rock can be seen in all their meanders.

To the south, on the contrary, the hilltops protected by massive sandrock are angular and rugged ; the valleys are narrow ; the streams are hemmed in between high and precipitous bluffs ; and flow with considerable speed and directness through rocky channels, holding comparatively small accumulations of re-worked Drift.

§ 11. Water basins. — On a further examination of the hydrography of the region, the water-courses draining these trenched summit plateaux or basins arrange themselves naturally into four groups, the members of each group converging toward a common outlet, deeply eroded through the sandstones and shales at the south ; but the waters of all four groups eventually commingle in the Ohio river at the mouth of the Big Beaver, below Pittsburgh.

These groups of streams, 'or water-trees, are designated upon the map (Plate I) by the following names : Chautauqua.

Oil Creek.

Conneaut.

Pymatuning.

§ 12. The Chautauqua water basin. — All the drainage of the Chautauqua group, gathered from the numerous headwater branches of the Allegheny, from the Conewango and its lake feeders, and from the Big Brokenstraw and its affluents, flows down the Allegheny river in one body after passing Irvineton, which is situated near the centre of Warren county.

Near Thompson's station, six miles below Irvineton, there appears to be a slight anticlinal crossing the river ; but just where exactly located and of how much importance it would be difficult to determine without an extensive instrumental survey. Here the southerly rim of the Chautauqua water basin is cut by the Allegheny river valley.

Following the rim or dividing ridge aronnd, it may be seen on the map to, commence a little west of the head of

TOPOGRAPHY AND DRAINAGE. III. 5

Chautauqua lake ; thence, southward, passing near Corry in Erie county, to Thompson's station in Warren county ; to Kane in McKean county ; to Keating Summit in Potter county ; and sweeping around by the east of Coudersport, and the head of the Allegheny river, to the noted "Contitinental summit" near Raymond's Corners in Potter county.*

The outlet of this group of streams, at Tidioute, drains a larger area than that of all the other three groups combined.

§ 13. The Oil Creek water basin is comparatively small^including only Oil Creek and its branches. Its waters enter the deep cut through the lower carboniferous sandstones a short distance below Titusville, and after following a tortuous course for seventeen miles, along a narrow bed hemmed in by high walls on either side, commingle with those of the Allegheny river at Oil City.

§ 14. The Conneaut water basin includes French creek and its branches, some of them streams of considerable importance. It collects its water from the Lake Erie divide, the westerly rim of the Chautauqua basin, and the northerly and westerly sides of the Oil creek basin, and delivers it through French creek into the Allegheny river at Franklin.

§ 15. The Pymatuning water basin.— The area above Greenville drained by this group is inconsiderable in extent, but the physical features of the basin are worthy of note. It includes the Pymatuning swamp with its several small feeders and two outlets — the Shenango and Crooked creek. These outlets come together at Greenville in Mercer count}", and, re-inforced lower down by the Pymatuning from the west, the Neshannock from the east and the Mahoning from the west, from the Big Beaver river which enters the Ohio about 30 miles below Pittsburgh.

Pymatuning swamp lies only about 35 feet below Conneaut lake, and is separated form it by a low and narrow

* Where the rainfall on a fifty acre tract is carried off in one direction through the Geiiesee river into Lake Ontario and the St. Lawrence, in another through the Allegheny, &c., into the Gulf of Mexico, and in still another through Pine creek and the Susquehanna river into Chesapeake bay.

6 III. REPORT OF PROGRESS. JOITX F. CARLL.

divide.* At the point where the ol'd Beaver canal (taking its water-supply from the summit reservoir (Conneaut lake) passed through this divide ; but a short cut about 25 feet in depth, and entirely in drift, was required. The amount of Drift piled in here, and the configuration of the hills on either side, make it seem not at all improbable that the Pymatuning waters once flowed into the Conneaut basin.

Comparative levels of the Water Basins.

§ 16. To aid in obtaining a comprehensive idea of the relative levels of these groups of drainage channels, let us now imagine the Allegheny river at Thompson's station, where the conglomerate-capped hills rise abruptly between six and seven hundred feet above the stream, to be checked by an obstruction raised in the valley 200 feet above the present surface of the water. This dam, although slight as compared with the hilltops, would throw back the water of the Allegheny river to the north, and flood the wide valleys of all its principal tributary streams. It would raise the surface of Chautauqua lake about 31 feet above its present level, and Cassadaga lake 25 feet. It would cause slackwater to extend up the Allegheny and its branches beyond the New York State line ; flood the Conewango valley and a large part of the country between Warren and Chautauqua and Cassadaga lakes, and fill the trough of the Brokenstraw to a point above Garland.

Let us in like manner suppose Oil creek dammed to the

* To the southwest of Conneaut lake and swamp, just under the comb of the ridge, between 30 and 40 feet lower in level, is to be found the great Pymatuning swamp. This marsh is very extensive, as its western boundaries lie some distance within the State of Ohio. The river Shenango finds its source in this swamp, a river which runs in a southerly direction until it meets the Neshannock, in Lawrence county, at New Castle, and thus forms the Beaver river, which falls into the Ohio about thirty miles below Pittsburg. The area of the swamp, as near as could be ascertained by this cursory survey-, is nine thousand acres. I believe that all the water which flows from it contributes to the Shenango. A small bay on its eastern extremity, is cut off by the Erie extension canal, and forms a subsidiary reservoir for that work at Hartstown ; out of this end of the swamp flows Crooked creek, %vhich joins the Shenango about four miles above West Greenville — Surveys of the Pymatuning and Conneaut swamps, by Col. Jos. Worrall, C. E., Under the joint resolution of February 28, 1868.

TOPOGRAPHY AND DRAINAGE. III. 7

height of 200 feet where it enters the highlands just below Titusville. The side walls here rise about 400 feet above the stream. This would flood all the low country along Oil creek nearly up to Oil lake and form a sheet of water two or three miles in width in some places. The valley of East Oil creek would also be filled as far up as Colorado, and Caldwell creek as far as to Pleasant valley.

An obstruction of 100 feet on French creek between Franklin and the mouth of Sugar creek, would cover the Conneaut marsh, and the broad valley of French creek to Meadville, thus enlarging the boundaries of Conneaut lake and forming a sheet of water of no mean dimensions.

An obstruction 100 feet high in the Shenango river at West Greenville would form a lake with a large island in the center, extending up both branches of the Shenango, and embracing within its outlines the great Pymatuning swamp.

The surface of the Chautauqua basin were it filled with water as we have imagined it to be would lie at an elevation above the ocean of about 1330 feet ; the Oil creek basin, 1360 feet ; the Conneaut basin, 1080 feet ; the Pymatuning basin, 1070 feet.

§ 17. Into these four basins all the drainage between the Lake Erie divide and the almost continuous range of outcropping Garland conglomerate now centers ; and if they were flooded, as supposed, the inflowing streams from the north and the dividing ridges, would simply be feeders to these long, narrow, irregularly outlined lakes ; and the surplus accumulations of water from all this country, would pass out, as they pass now, only through the four gaps in the conglomerate first indicated.

§ 18. Going still further in this line of illustration, let us now compare these four imaginary sheets of water and note the probable effects of a change in their respective levels.

By increasing the obstruction in the Allegheny river at Thompson's station 10 feet, the Chautauqua basin would overflow to the north through Cassadaga lake into Canadaway creek and thence to Lake Erie.

By raising the Oil creek obstruction 10 feet, the Oil creek basin would overflow to the northwest, through the west branch of Oil creek into Muddy creek and thence into French creek.

By raising the French creek obstruction 10 feet, the Conneaut basin would oner/low to the west into the Pymatuning basin, thus converging all the waters from the Oil creek, Conneaut and Pymatuning basins into the Shenango river.

But if we should now raise the SJienango dam 30 feet, bringing it up to 1100 feet above tide, and build up the French creek obstruction to the same level, we should probably check the southern flow of water from all this area, and a new channel would be opened for it at the north, from Conneaut lake into Conneaut creek and thence into Lake Erie.

§ 19. We must bear in mind that the southern boundary of these basins is a highland capped with several hundred feet of the coarse sandy measures outcropingfrom beneath the coal fields of the south. At Thompson's gap the average top of the ridge may be put at about 1800 feet above tide ; at Oil creek about 1600 feet ; at French creek about 1400 feet ; and at the SJienango about 1200 feet.

§ 20. Thompson's gap. — We have seen that an obstruction in the Thompson's station cut, if raised to an elevation of 1340 feet above tide, would stop the passage of water at this point and open a new outlet for the drainage of the Chautauqua basin through Cassadaga lake into Lake Erie. By what agency, then, and when, was this gap (at Thompsons) opened ; since whatever it was that opened it must have commenced its work in the upper rocks at an elevation of 1800 feet at least — 460 feet above the place where the water now would have a free outlet to the north ?

§ 21. Oil creek gap. — So too on Oil creek, when the water is stopped, and thrown back until it finds an outlet to the north, by an obstruction 1370 feet above tide, 230 feet of the upper part of the chasm remains as a witness of the work of other agencies than the waters of the present basin.

§ 22. French creek gap. — The French creek gap also presents similar features. If its channel were filled to a sufficient height (1090 feet above tide) to cause the Conneaut basin to deliver to the west, there would still be an erosion

TOPOGRAPHY AXD DRAINAGE. III. 9

of 310 feet of the upper part of the old gap in the hills to be accounted for.

§ 23. Date of the formation of the gaps. — It appears therefore quite certain from the above presentation of the case that these southern outlets through this escarpment of massive conglomerates and sandstones, were opened long anterior to the time when the drainage of this area began to be influenced by its present topography, and before the waters were confined within the rims of the respective basins as we now find them. Otherwise the overflowing waters could not have cut through several hundred feet of sandstones and shales on the south, lying at a higher level than their surface ; but would have found a more ready outlet and a more rapid delivery through the softer measures of the north and west — the Chautauqua, through Cassadaga lake into Lake Erie — the Oil creek through Muddy creek into the Conneaut — the Conneaut into the Pymatuning — and the three latter (combined) into the low levels of the Shenango river country.

§ 24. The agency glacial. — The peculiar arrangement of these basins between the Lake Erie divide on the north and outcropping conglomerates on the south ; the positions of their several outlets, which singularly enough are cut down through the highest and hardest portions of their inclosing barriers — the highest reservoir having cut the deepest outlet ; and the immense amount of northern drift, accompanied by erratic bowlders, spread over nearly all this section— seem to point clearly to the inference that these features are not due to the action of aqueous and atmospheric erosion alone, but that the general outlines of the present topography were "carved out in the rough" prior to and during the glacial epoch, and afterwards modified and smoothed down by the gradually receding waters as the ice disappeared.

§ 25. No continental rock "barrier. — We can hardly imagine that the levels of the country between the Cincinnati anticlinal and the Chautauqua highlands were ever so adjusted as to hold a fixed body of water to the north at the high elevation of the Chautauqua basin ; and if they were so adjusted, then, when the first outlet opened to the south, down the dip of the strata, wherever that outlet might be, the erosion would be so rapid that only one waste weir would be required. This outlet, deepening, widening and lowering the surface of the water to the north, would be likely to leave in the basin behind (as the waters subsided by drainage) a system of streams all converging towards the common outlet. But instead of one basin of this description we have shown that there are four ; each one having an independent outlet, cut down deeply through ponderous rocks lying high above the average level of the country drained by it.

§ 26. JVo general submergence. — Neither are we prepared to accept the hypothesis of submergence below sea level to account for the superficial deposits and the facial peculiarities of this section. For in that case, when the land again emerged, and drainage commenced in any direction, we should have the same features as those above referred to repeated. Or, if an isolated basin remained unemptied, it would overflow and open an outlet for itself at the lowest point in its barrier, and not through sandstone cliffs rising hundreds of feet above its surface.

The orographic features of the district stand out boldly as witnesses of a fact which will hardly be questioned by any one after an examination of the country. They offer the best of evidence in themselves that they are not the results of the action of water alone, but of ice and water combined. In just what way the erosion was accomplished ; what its status was at the beginning of the glacial epoch ; and what the measure of its progress has been during and since that time, we can as yet only surmise.

Chapter Ii.*

The GARLAND (Olean\ or Sharon^) CONGLOMERATE and SUB-GARLAND (Sub-Olean-\- or Shenangoty SANDSTONE placed and traced.

a. Leveled Line of Survey.

§ 27. In selecting a route over which to run a spirit level line across northwestern Pennsylvania in 1875, for the purpose of connectedly tracing the several outcrops of Garland conglomerate, that they might thus be identified with cotemporaneous rocks in the State of New York on the one side and in the State of Ohio on the other, a due consideration of the amount of time that could be appropriated to the undertaking without prejudice to interests of equal importance in other parts of the oil district made the adoption of a route imperative which offered fewest obstacles to the speedy execution of the work.

In addition to the levels, courses and distances also were to be taken, in order that the line might be used as a base for future operations if occasion required. It was therefore desirable to follow some highway; and, under these circumstances, the choice naturally fell to the " State road" commencing at Busti, in Chautauqua county, N. Y. and running through Sugar Grove and West Spring creek in Warren county, Pa. ; Ricemlle, Meadvillc, Adamsville, and Jamestown in Crawford county to Kinney 's Corners in Trumbull county, Ohio.

§ 28. Although the selected line of survey does not keep constantly upon the outcrop of the Garland conglomerate it is an excellent one along which to study the geology of the

* Report of work done in 1875.

f Names assigned by Mr. Ashburner in his Reports (R, RR) on McKean, Cameron, Elk, and Forest counties.

{ Names assigned by Mr. White in his Reports (QQ, QQQ, QQQQ) on Lawrence, Mercer, Crawford, and Erie counties. ( 11 III. )

country, and it is as direct a line as any highway at our command between the points named.

§ 29. Messrs. F. A. Hatch and Arthur Hale, with C. A. Bodge as rodman, performed the work. The distances were measured with a ten foot odometer- wheel. Bearings were taken by transit- telescope and needle. Elevations were observed on the vertical circle of the transit instrument, and reduced to ocean level by reference to the data given in the profiles of the several railroads crossed.*

§ 30. The region traversed by this line of survey, rapidly executed for a special geological purpose, affords ample scope for the investigations of afield-party during an entire season, even if all the time were devoted to a study of its Drift phenomena alone. Its glacial grooves — its bowlders scattered at all elevations along the hillsides — its old water channels, some of them now filled to a height of 400 feet or even more — its drift-clays and lake beaches — each and all of these deserve close and careful study, and proffer the material data for a very interesting chapter in the geology of the State.

§ 31. But for the present, attention must be chiefly given to the more important task of tracing the Garland conglomerate (the Second Mountain sand of the oil wells) along its northerly outcrop across Pennsylvania ; identifying it with certain conglomerate strata outcropping in New York and Ohio ; and endeavoring to fix its horizon definitely in relation to the overlying coal measures and the underlying oil-bearing rocks.

The following chapters will deal chiefly with these topics, other matters being incidentally noticed as they were observed.

b. Place of the Garland Conglomerate in the Series.

§ 32. The term Garland conglomerate is used' in these pages as a convenient local geographical name for the low-

* As the survey along this line has not been amplified or extended in any direction, as it was originally intended to be, only a portion of their notes have been used in the preparation of the contour map and profiles (Plates 3 and 3B) herewith presented.

est member of the Carboniferous Conglomerate series in this part of the State.

It is apparently identical with the Olean conglomerate in. McKean county ; with the Sharon conglomerate in Mercer county ; with the Ohio conglomerate in Ohio ; and with the Second Mountain Sand of the oil wells.

§ 33. There has prevailed hitherto great confusion of ideas respecting the massive conglomeratic sand deposits underlying the Productive coal measures.

The Conglomerate (XII) at Pottsville in Schuylkill county (where it is more than 1000 feet thick) has been traced around the Anthracite coal basins and along the Allegheny mountain to the Maryland State line, without much difficulty ; and it has been recognized as a whole, throughout western Pennsylvania, wherever it appears upon the anticlinals, in gorges, and along the sides of valleys. But it is as thin in the west as it is thick in the east ; and the Mauch Chunk red shales (XI) which underlie it, and measure more than 3000 feet in thickness east of the Susquehanna river, and more than 1000 feet on the upper Juniata river, dwindle to 100 feet in Clinton county, and almost disappear on the Allegheny river and Beaver river waters. This brings almost into contact with the base of the Pottsville conglomerate (XII)' the top of the Pocono sandstone formation (X) in western Pennsylvania; and caused the heavy conglomeratic sandstone strata in the upper part of X to be confounded with the conglomeratic sandstones of XII. • The Garland conglomerate in Warren county, the Olean conglomerate in McKean county, the Panama conglomerate in Chatauqua county, N. Y. and the Salamanca conglomerate in Cattaraugus county, N. Y. have been looked upon not only as local exhibitions of one and the same deposit, but as the common representatives (in a thin form) of the whole Pottsville Conglomerate formation (XII) of eastern and middle Pennsylvania.

§ 34. But it will be seen in the progress of this report that both of these suppositions have been mistakes, due to superficial observations at far distant and necessarily unconnected points. More elaborate, connected, instrumental

14 III. REPORT OF PROGRESS. JOHN F. CAHLL.

surveys have placed clearly in light the true state of things, namely :

1. That the Panama and Salamanca conglomerates in the

State of J^ew York are different (and lower) formations from the Olean and Garland conglomerates in Pennsylvania.

2. That the Olean and Garland are two outcrops of one and the same conglomeritic sandstone deposit.

3. That this again is a different (and lower) formation from the Ilomewood conglomeritic sandstone of the Allegheny river and Beaver river valleys, denned as No. XII by the First Geological Survey of Pennsylvania in 183o- 1841, but which is only the upper part of No. XII.

§ 35. Between the Homewood sandstone of Butler county, (underlying the lowest bed of the Productive coal series, and therefore the representative of at least the upper part of the great conglomerate) and the Garland conglomerate exists a vertical interval of more than 200 feet.

In this interval lie the Mercer group of coal beds and the Sharon block coal bed — the variable coal beds of Crawford county to be mentioned further on, — the thin coal beds north of the Allegheny river in Venango and Warren counties — and the most northerly coal beds of McKean county.

These intermediate groups of coal beds attain their best development in Mercer county ; but they appear to be similar in physical character wherever they are found. They lie in "swamps," irregular in outline and uncertain in extent, and are often liable to "pinch out and terminate when least expected." They may be locally traced with considerable uniformity as to levels and position, but can seldom be relied upon as persistent over wide areas.

The interval in which these coal beds occur is occupied largely with sandy measures, and local layers of pebbles ; changing frequently from conglomerate to sandstone and from sandstone to slate and shale ; sometimes thin bedded, sometimes very massive ; one bed fading out and another of a different character coming in.

§ 36. The Pottsville conglomerate, (XII, the Serai conglomerate of Prof. Rogers,) the recognized base of the coal measures in eastern Pennsylvania, is 1000 feet thick at

GAUL AND CONGLOMERATE. III. 15

Pottsville, as said above, and cannot be looked upon as a solid homogeneous formation, but as a series, or group, of pebble-rock and sand-rock layers, separated by beds of coarse shale and slate, black slate, and coal. At the west end of the Pottsville basin it holds a number of small coal beds, and one—the Lykens valley bed — of great size and value.

It may very well be represented then in western Pennsylvania by the interval above described, extended to include the Homewood sandstone at the top, and the Garland conglomerate at the bottom, — an interval of full 300 feet in thickness, and with a character which closely resembles the description just given.