The Project Gutenberg EBook of The National Geographic Magazine, Vol. I., No. 3, July, 1889, by Various This eBook is for the use of anyone anywhere in the United States and most other parts of the world at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at www.gutenberg.org. If you are not located in the United States, you'll have to check the laws of the country where you are located before using this ebook. Title: The National Geographic Magazine, Vol. I., No. 3, July, 1889 Author: Various Release Date: November 4, 2015 [EBook #50383] Language: English Character set encoding: ISO-8859-1 *** START OF THIS PROJECT GUTENBERG EBOOK NATIONAL GEOGRAPHIC MAG., JULY 1889 *** Produced by Ron Swanson CONTENTS. The Rivers and Valleys of Pennsylvania: William Morris Davis (Illustrated by one map and twenty-five cuts.) Topographic Models: Cosmos Mindeleff (Illustrated by two plates.) National Geographic Society—Abstract of Minutes International Literary Contest to be held at Madrid, Spain July, 1889. PRESS OF TUTTLE, MOREHOUSE & TAYLOR, NEW HAVEN, CONN. THE NATIONAL GEOGRAPHIC MAGAZINE. Vol. I. 1889. No. 3. THE RIVERS AND VALLEYS OF PENNSYLVANIA.1 BY WILLIAM MORRIS DAVIS. "In Faltensystemen von sehr hohem Alter wurde die ursprüngliche Anordnung der Langenthäler durch das Ueberhandnehmen der transversalen Erosionsfurchen oft ganz und gar verwischt." LÖWL. Petermann's Mittheilungen, xxviii, 1882, 411. 1 The substance of this essay was presented to the Society in a lecture on February 8th, 1889, but since then it has been much expanded. CONTENTS. PART FIRST. Introductory. 1. Plan of work here proposed. 2. General description of the topography of Pennsylvania. 3. The drainage of Pennsylvania. 4. Previous studies of Appalachian drainage. PART SECOND. Outline of the geological history of the region. 5. Conditions of formation. 6. Former extension of strata to the southeast. 7. Cambro-Silurian and Permian deformations. 8. Perm-Triassic denudation. 9. Newark deposition. 10. Jurassic tilting. 11. Jura-Cretaceous denudation. 12. Tertiary elevation and denudation. 13. Later changes of level. 14. Illustrations of Pennsylvanian topography. PART THIRD. General conception of the history of a river. 15. The complete cycle of river life: youth, adolescence, maturity and old age. 16. Mutual adjustment of river courses. 17. Terminology of rivers changed by adjustment. 18. Examples of adjustments. 19. Revival of rivers by elevation and drowning by depression. 20. Opportunity for new adjustments with revival. 21. Antecedent and superimposed rivers. 22. Simple, compound, composite and complex rivers. PART FOURTH. The development of the rivers of Pennsylvania. 23. Means of distinguishing between antecedent and adjusted consequent rivers. 24. Postulates of the argument. 25. Constructional Permian topography and consequent drainage. 26. The Jura mountains homologous with the Permian Alleghanies. 27. Development and adjustment of the Permian drainage. 28. Lateral water-gaps near the apex of synclinal ridges. 29. Departure of the Juniata from the Juniata-Catawissa syncline. 30. Avoidance of the Broad Top basin by the Juniata headwaters. 31. Reversal of larger rivers to southeast courses. 32. Capture of the Anthracite headwaters by the growing Susquehanna. 33. Present outward drainage of the Anthracite basins. 34. Homologies of the Susquehanna and Juniata. 35. Superimposition of the Susquehanna on two synclinal ridges. 36. Evidence of superimposition in the Susquehanna tributaries. 37. Events of the Tertiary cycle. 38. Tertiary adjustment of the Juniata on the Medina anticlines. 39. Migration of the Atlantic-Ohio divide. 40. Other examples of adjustments. 41. Events of the Quaternary cycle. 42. Doubtful cases. 43. Complicated history of our actual rivers. 44. Provisional conclusions. PART FIRST. Introductory. 1. Plan of work here proposed.—No one now regards a river and its valley as ready-made features of the earth's surface. All are convinced that rivers have come to be what they are by slow processes of natural development, in which every peculiarity of river-course and valley-form has its appropriate cause. Being fully persuaded of the gradual and systematic evolution of topographic forms, it is now desired, in studying the rivers and valleys of Pennsylvania, to seek the causes of the location of the streams in their present courses; to go back if possible to the early date when central Pennsylvania was first raised above the sea and trace the development of the several river systems then implanted upon it from their ancient beginning to the present time. The existing topography and drainage system of the State will first be briefly described. We must next inquire into the geological structure of the region, follow at least in a general way the deformations and changes of attitude and altitude that it has suffered, and consider the amount of denudation that has been accomplished on its surface. We must at the same time bear in mind the natural history of rivers, their morphology and development; we must recognize the varying activities of a river in its youth and old age, the adjustments of its adolescence and maturity, and the revival of its decrepit powers when the land that it drains is elevated and it enters a new cycle of life. Finally we shall attempt to follow out the development of the rivers of Pennsylvania by applying the general principles of river history to the special case of Pennsylvania structure. 2 . General description of the topography of Pennsylvania.—The strongly marked topographic districts of Pennsylvania can hardly be better described than by quoting the account given over a century ago by Lewis Evans, of Philadelphia, in his "Analysis of a map of the middle British colonies in America" (1755), which is as valuable from its appreciative perception as it is interesting from its early date. The following paragraphs are selected from his early pages: "The land southwestward of Hudson's River is more regularly divided and into a greater number of stages than the other. The first object worthy of regard in this part is a rief or vein of rocks of the talky or isinglassy kind, some two or three or half a dozen miles broad; rising generally some small matter higher than the adjoining land; and extending from New York city southwesterly by the lower falls of Delaware, Schuylkill, Susquehanna, Gun-Powder, Patapsco, Potomack, Rapahannock, James river and Ronoak. This was the antient maritime boundary of America and forms a very regular curve. The land between this rief and the sea and from the Navesink hills southwest ... may be denominated the Lower Plains, and consists of soil washt down from above and sand accumulated from the ocean. Where these plains are not penetrated by rivers, they are a white sea-sand, about twenty feet deep and perfectly barren, as no mixture of soil helps to enrich them. But the borders of the rivers, which descend from the uplands, are rendered fertile by the soil washt down with the floods and mixt with the sands gathered from the sea. The substratum of sea-mud, shells and other foreign subjects is a perfect confirmation of this supposition. And hence it is that for 40 or 50 miles inland and all the way from the Navesinks to Cape Florida, all is a perfect barren where the wash from the uplands has not enriched the borders of the rivers; or some ponds and defiles have not furnished proper support for the growth of white cedars.... "From this rief of rocks, over which all the rivers fall, to that chain of broken hills, called the South mountain, there is the distance of 50, 60 or 70 miles of very uneven ground, rising sensibly as you advance further inland, and may be denominated the Upland. This consists of veins of different kinds of soil and substrata some scores of miles in length; and in some places overlaid with little ridges and chains of hills. The declivity of the whole gives great rapidity to the streams; and our violent gusts of rain have washt it all into gullies, and carried down the soil to enrich the borders of the rivers in the Lower Plains. These inequalities render half the country not easily capable of culture, and impoverishes it, where torn up by the plow, by daily washing away the richer mould that covers the surface. "The South mountain is not in ridges like the Endless mountains, but in small, broken, steep, stoney hills; nor does it run with so much regularity. In some places it gradually degenerates to nothing, not to appear again for some miles, and in others it spreads several miles in breadth. Between South mountain and the hither chain of the Endless mountains (often for distinction called the North mountain, and in some places the Kittatinni and Pequélin), there is a valley of pretty even good land, some 8, 10 or 20 miles wide, and is the most considerable quantity of valuable land that the English are possest of; and runs through New Jersey, Pensilvania, Mariland and Virginia. It has yet obtained no general name, but may properly enough be called Piemont, from its situation. Besides conveniences always attending good land, this valley is everywhere enriched with Limestone. "The Endless mountains, so called from a translation of the Indian name bearing that signification, come next in order. They are not confusedly scattered and in lofty peaks overtopping one another, but stretch in long uniform ridges scarce half a mile perpendicular in any place above the intermediate vallies. Their name is expressive of their extent, though no doubt not in a literal sense.... The mountains are almost all so many ridges with even tops and nearly of a height. To look from these hills into the lower lands is but, as it were, into an ocean of woods, swelled and deprest here and there by little inequalities, not to be distinguished one part from another any more than the waves of the real ocean. The uniformity of these mountains, though debarring us of an advantage in this respect, makes some amends in another. They are very regular in their courses, and confine the creeks and rivers that run between; and if we know where the gaps are that let through these streams, we are not at a loss to lay down their most considerable inflections.... "To the northwestward of the Endless mountains is a country of vast extent, and in a manner as high as the mountains themselves. To look at the abrupt termination of it, near the sea level, as is the case on the west side of Hudson's river below Albany, it looks as a vast high mountain; for the Kaats Kills, though of more lofty stature than any other mountains in these parts of America, are but the continuation of the Plains on the top, and the cliffs of them in the front they present towards Kinderhook. These Upper Plains are of extraordinary rich level land, and extend from the Mohocks river through the country of the Confederates.2 Their termination northward is at a little distance from Lake Ontario; but what it is westward is not known, for those most extensive plains of Ohio are part of them." 2 Referring to the league of Indian tribes, so-called. These several districts recognized by Evans may be summarized as the coastal plain, of nearly horizontal Cretaceous and later beds, just entering the southeastern corner of Pennsylvania; the marginal upland of contorted schists of disputed age; the South Mountain belt of ancient and much disturbed crystalline rocks, commonly called Archean; a space between these two traversed by the sandstone lowland of the Newark formation;3 the great Appalachian valley of crowded Cambrian limestones and slates; the region of the even-crested, linear Paleozoic ridges, bounded by Kittatinny or Blue mountain on the southeast and by Alleghany mountain on the northwest, this being the area with which we are here most concerned; and finally the Alleghany plateau, consisting of nearly horizontal Devonian and Carboniferous beds and embracing all the western part of the state. The whole region presents the most emphatic expression not only of its structure but also of the more recent cycles of development through which it has passed. Fig. 1 represents the stronger ridges and larger streams of the greater part of the central district: it is reproduced from the expressive Topographic Map of Pennsylvania (1871) by Lesley. The Susquehanna flows down the middle, receiving the West Branch from Lock Haven and Williamsport, the East Branch from Wilkes-Barre in the Wyoming basin, and the Juniata from the Broad Top region, south of Huntingdon. The Anthracite basins lie on the right, enclosed by zigzag ridges of Pocono and Pottsville sandstone; the Plateau, trenched by the West Branch of the Susquehanna is in the northwest. Medina sandstone forms most of the central ridges. 3 Russell has lately recommended the revival of this term, proposed many years ago by Redfield, as a non-committal name for the "New red sandstones" of our Atlantic slope, commonly called Triassic. FIG. 1. Part of Topographic Map of Pennsylvania, by J. P. Lesley (1871). 3. The drainage of Pennsylvania.—The greater part of the Alleghany plateau is drained westward into the Ohio, and with this we shall have little to do. The remainder of the plateau drainage reaches the Atlantic by two rivers, the Delaware and the Susquehanna, of which the latter is the more special object of our study. The North and West Branches of the Susquehanna rise in the plateau, which they traverse in deep valleys; thence they enter the district of the central ranges, where they unite and flow in broad lowlands among the even-crested ridges. The Juniata brings the drainage of the Broad Top region to the main stream just before their confluent current cuts across the marginal Blue Mountain. The rock-rimmed basins of the anthracite region are drained by small branches of the Susquehanna northward and westward, and by the Schuylkill and Lehigh to the south and east. The Delaware, which traverses the plateau between the Anthracite region and the Catskill Mountain front, together with the Lehigh, the Schuylkill, the little Swatara and the Susquehanna, cut the Blue Mountain by fine water-gaps, and cross the great limestone valley. The Lehigh then turns eastward and joins the Delaware, and the Swatara turns westward to the Susquehanna; but the Delaware, Schuylkill and Susquehanna all continue across South Mountain and the Newark belt, and into the low plateau of schists beyond. The Schuylkill unites with the Delaware near Philadelphia, just below the inner margin of the coastal plain; the Delaware and the Susquehanna continue in their deflected estuaries to the sea. All of these rivers and many of their side streams are at present sunk in small valleys of moderate depth and width, below the general surface of the lowlands, and are more or less complicated with terrace gravels. 4. Previous studies of Appalachian drainage.—There have been no special studies of the history of the rivers of Pennsylvania in the light of what is now known of river development. A few recent essays of rather general character as far as our rivers are concerned, may be mentioned. Peschel examined our rivers chiefly by means of general maps with little regard to the structure and complicated history of the region. He concluded that the several transverse rivers which break through the mountains, namely, the Delaware, Susquehanna and Potomac, are guided by fractures, anterior to the origin of the rivers.4 There does not seem to be sufficient evidence to support this obsolescent view, for most of the water-gaps are located independently of fractures; nor can Peschel's method of river study be trusted as leading to safe conclusions. 4 Physische Erdkunde, 1880, ii, 442. Tietze regards our transverse valleys as antecedent;5 but this was made only as a general suggestion, for his examination of the structure and development of the region is too brief to establish this and exclude other views. 5 Jahrbuch Geol. Reichsanstalt, xxviii, 1878, 600. Löwl questions the conclusion reached by Tietze and ascribes the transverse gaps to the backward or headwater erosion of external streams, a process which he has done much to bring into its present important position, and which for him replaces the persistence of antecedent streams of other authors.6 6 Pet. Mitth., 1882, 405; Ueber Thalbildung, Prag, 1884. A brief article7 that I wrote in comment on Löwl's first essay several years ago now seems to me insufficient in its method. It exaggerated the importance of antecedent streams; it took no sufficient account of the several cycles of erosion through which the region has certainly passed; and it neglected due consideration of the readjustment of initial immature stream courses during more advanced river-life. Since then, a few words in Löwl's essay have come to have more and more significance to me; he says that in mountain systems of very great age, the original arrangement of the longitudinal valleys often becomes entirely confused by means of their conquest by transverse erosion gaps. This suggestion has been so profitable to me that I have placed the original sentence at the beginning of this paper. Its thesis is the essential element of my present study. 7 Origin of Cross-valleys. Science, i, 1883, 325. Phillipson refers to the above-mentioned authors and gives a brief account of the arrangement of drainage areas within our Appalachians, but briefly dismisses the subject.8 His essay contains a serviceable bibliography. 8 Studien über Wasserscheiden. Leipsig, 1886, 149. If these several earlier essays have not reached any precise conclusion, it may perhaps be because the details of the geological structure and development of Pennsylvania have not been sufficiently examined. Indeed, unless the reader has already become familiar with the geological maps and reports of the Pennsylvania surveys and is somewhat acquainted with its geography, I shall hardly hope to make my case clear to him. The volumes that should be most carefully studied are, first, the always inspiring classic, "Coal and its Topography" (1856), by Lesley, in which the immediate relation of our topography to the underlying structure is so finely described; the Geological Map of Pennsylvania (1856), the result of the labors of the first survey of the state; and the Geological Atlas of Counties, Volume X of the second survey (1885). Besides these, the ponderous volumes of the final report of the first survey and numerous reports on separate counties by the second survey should be examined, as they contain many accounts of the topography although saying very little about its development. If, in addition to all this, the reader has seen the central district of the state and marvelled at its even-crested, straight and zigzag ridges, and walked through its narrow water-gaps into the enclosed coves that they drain, he may then still better follow the considerations here presented. PART SECOND. Outline of the geological history of the region. 5. Conditions of formation.—The region in which the Susquehanna and the neighboring rivers are now located is built in chief part of marine sediments derived in paleozoic time from a large land area to the southeast, whose northwest coast-line probably crossed Pennsylvania somewhere in the southeastern part of the state; doubtless varying its position, however, by many miles as the sea advanced and receded in accordance with the changes in the relative altitudes of the land and water surfaces, such as have been discussed by Newberry and Claypole. The sediments thus accumulated are of enormous thickness, measuring twenty or thirty thousand feet from their crystalline foundation to the uppermost layer now remaining. The whole mass is essentially conformable in the central part of the state. Some of the formations are resistent, and these have determined the position of our ridges; others are weaker and are chosen as the sites of valleys and lowlands. The first are the Oneida and Medina sandstones, which will be here generally referred to under the latter name alone, the Pocono sandstone and the Pottsville conglomerate; to these may be added the fundamental crystalline mass on which the whole series of bedded formations was deposited, and the basal sandstone that is generally associated with it. Wherever we now see these harder rocks, they rise above the surrounding lowland surface. On the other hand, the weaker beds are the Cambrian limestones (Trenton) and slates (Hudson River), all the Silurian except the Medina above named, the whole of the Devonian—in which however there are two hard beds of subordinate value, the Oriskany sandstone and a Chemung sandstone and conglomerate, that form low and broken ridges over the softer ground on either side of them—and the Carboniferous (Mauch Chunk) red shales and some of the weaker sandstones (Coal measures). 6. Former extension of strata to the southeast.—We are not much concerned with the conditions under which this great series of beds was formed; but, as will appear later, it is important for us to recognize that the present southeastern margin of the beds is not by any means their original margin in that direction. It is probable that the whole mass of deposits, with greater or less variations of thickness, extended at least twenty miles southeast of Blue Mountain, and that many of the beds extended much farther. The reason for this conclusion is a simple one. The several resistant beds above-mentioned consist of quartz sand and pebbles that cannot be derived from the underlying beds of limestones and shales; their only known source lay in the crystalline rocks of the paleozoic land to the southeast. South Mountain may possibly have made part of this paleozoic land; but it seems more probable that it was land only during the earlier Archean age, and that it was submerged and buried in Cambrian time and not again brought to the light of day until it had been crushed into many local anticlines9 whose crests were uncovered by Permian and later erosion. The occurrence of Cambrian limestone on either side of South Mountain, taken with its compound anticlinal structure, makes it likely that Medina time found this crystalline area entirely covered by the Cambrian beds; Medina sands must therefore have come from farther still to the southeast. A similar argument applies to the source of the Pocono and Pottsville beds. The measure of twenty miles as the former southeastern extension of the paleozoic formations therefore seems to be a moderate one for the average of the whole series; perhaps forty would be nearer the truth. 9 Lesley, as below. 7. Cambro-Silurian and Permian deformations.—This great series of once horizontal beds is now wonderfully distorted; but the distortions follow a general rule of trending northeast and southwest, and of diminishing in intensity from southeast to northwest. In the Hudson Valley, it is well known that a considerable disturbance occurred between Cambrian and Silurian time, for there the Medina lies unconformably on the Hudson River shales. It seems likely, for reasons that will be briefly given later on, that the same disturbance extended into Pennsylvania and farther southwest, but that it affected only the southeastern corner of the State; and that the unconformities in evidence of it, which are preserved in the Hudson Valley, are here lost by subsequent erosion. Waste of the ancient land and its Cambro-Silurian annex still continued and furnished vast beds of sandstone and sandy shales to the remaining marine area, until at last the subsiding Paleozoic basin was filled up and the coal marshes extended broadly across it. At this time we may picture the drainage of the southeastern land area wandering rather slowly across the great Carboniferous plains to the still submerged basin far to the west; a condition of things that is not imperfectly represented, although in a somewhat more advanced stage, by the existing drainage of the mountains of the Carolinas across the more modern coastal plain to the Atlantic. This condition was interrupted by the great Permian deformation that gave rise to the main ranges of the Appalachians in Pennsylvania, Virginia and Tennessee. The Permian name seems appropriate here, for while the deformation may have begun at an earlier date, and may have continued into Triassic time, its culmination seems to have been within Permian limits. It was characterized by a resistless force of compression, exerted in a southeast-northwest line, in obedience to which the whole series of Paleozoic beds, even twenty or more thousand feet in thickness, was crowded gradually into great and small folds, trending northeast and southwest. The subjacent Archean terrane doubtless shared more or less in the disturbance: for example, South Mountain is described by Lesley as "not one mountain, but a system of mountains separated by valleys. It is, geologically considered, a system of anticlinals with troughs between.... It appears that the South Mountain range ends eastward [in Cumberland and York Counties] in a hand with five [anticlinal] fingers."10 10 Proc. Amer. Phil. Soc., xiii, 1873, 6. It may be concluded with fair probability that the folds began to rise in the southeast, where they are crowded closest together, some of them having begun here while coal marshes were still forming farther west; and that the last folds to be begun were the fainter ones on the plateau, now seen in Negro mountain and Chestnut and Laurel ridges. In consequence of the inequalities in the force of compression or in the resistance of the yielding mass, the folds do not continue indefinitely with horizontal axes, but vary in height, rising or falling away in great variety. Several adjacent folds often follow some general control in this respect, their axes rising and falling together. It is to an unequal yielding of this kind that we owe the location of the Anthracite synclinal basins in eastern Pennsylvania, the Coal Measures being now worn away from the prolongation of the synclines, which rise in either direction. 8. Perm-Triassic denudation.—During and for a long time after this period of mountain growth, the destructive processes of erosion wasted the land and lowered its surface. An enormous amount of material was thus swept away and laid down in some unknown ocean bed. We shall speak of this as the Perm-Triassic period of erosion. A measure of its vast accomplishment is seen when we find that the Newark formation, which is generally correlated with Triassic or Jurassic time, lies unconformably on the eroded surface of Cambrian and Archean rocks in the southeastern part of the State, where we have concluded that the Paleozoic series once existed; where the strata must have risen in a great mountain mass as a result of the Appalachian deformations; and whence they must therefore have been denuded before the deposition of the Newark beds. Not only so; the moderate sinuosity of the southeastern or under boundary of the Newark formation indicates clearly enough that the surface on which that portion of the formation lies is one of no great relief or inequality; and such a surface can be carved out of an elevated land only after long continued denudation, by which topographic development is carried beyond the time of its greatest strength or maturity into the fainter expression of old age. This is a matter of some importance in our study of the development of the rivers of Pennsylvania; and it also constitutes a good part of the evidence already referred to as indicating that there must have been some earlier deformations of importance in the southeastern part of the State; for it is hardly conceivable that the great Paleozoic mass could have been so deeply worn off of the Newark belt between the making of the last of the coal beds and the first of the Newark. It seems more in accordance with the facts here recounted and with the teachings of geological history in general to suppose, as we have here, that something of the present deformation of the ancient rocks underlying the Newark beds was given at an early date, such as that of the Green Mountain growth; and that a certain amount of the erosion of the folded beds was thus made possible in middle Paleozoic time; then again at some later date, as Permian, a second period of mountain growth arrived, and further folding was effected, and after this came deeper erosion; thus dividing the destructive work that was done into several parts, instead of crowding it all into the post-Carboniferous time ordinarily assigned to it. It is indeed not impossible that an important share of what we have called the Permian deformation was, as above suggested, accomplished in the southeastern part of the State while the coal beds were yet forming in the west; many grains of sand in the sandstones of the Coal Measures may have had several temporary halts in other sandstone beds between the time of their first erosion from the Archean rocks and the much later time when they found the resting place that they now occupy.11 11 These considerations may have value in showing that the time in which the lateral crushing of the Appalachians was accomplished was not so brief as is stated by Reade in a recent article in the American Geologist, iii, 1889, 106. 9. Newark deposition.—After the great Paleozoic and Perm-Triassic erosions thus indicated, when the southeastern area of ancient mountains had been well worn down and the Permian folds of the central district had acquired a well developed drainage, there appeared an opportunity for local deposition in the slow depression of a northeast-southwest belt of the deeply wasted land, across the southeastern part of the State; and into this trough-like depression, the waste from the adjacent areas on either side was carried, building the Newark formation. This may be referred to as the Newark or Trias-Jurassic period of deposition. The volume of this formation is unknown, as its thickness and original area are still undetermined; but it is pretty surely of many thousand feet in vertical measure, and its original area may have been easily a fifth or a quarter in excess of its present area, if not larger yet. So great a local accumulation seems to indicate that while the belt of deposition was sinking, the adjacent areas were rising, in order to furnish a continual supply of material; the occurrence of heavy conglomerates along the margins of the Newark formation confirms this supposition, and the heavy breccias near Reading indicate the occurrence of a strong topography and a strong transporting agent to the northwest of this part of the Newark belt. It will be necessary, when the development of the ancestors of our present rivers is taken up, to consider the effects of the depression that determined the locus of Newark deposition and of the adjacent elevation that maintained a supply of material. 10. Jurassic tilting.—Newark deposition was stopped by a gradual reversal of the conditions that introduced it. The depression of the Newark belt was after a time reversed into elevation, accompanied by a peculiar tilting, and again the waste of the region was carried away to some unknown resting place. This disturbance, which may be regarded as a revival of the Permian activity, culminated in Jurassic, or at least in post-Newark time, and resulted in the production of the singular monoclinal attitude of the formation; and as far as I can correlate it with the accompanying change in the underlying structures, it involved there an over-pushing of the closed folds of the Archean and Paleozoic rocks. This is illustrated in figs. 2 and 3, in which the original and disturbed attitudes of the Newark and the underlying formations are roughly shown, the over-pushing of the fundamental folds causing the monoclinal and probably faulted structure in the overlying beds.12 If this be true, we might suspect that the unsymmetrical attitude of the Appalachian folds, noted by Rogers as a characteristic of the range, is a feature that was intensified if not originated in Jurassic and not in Permian time. 12 Amer. Journ. Science, xxxii, 1886, 342; and Seventh Ann. Rept. U. S. Geol. Survey, 1888, 486. FIG. 2. FIG. 3. It is not to be supposed that the Jurassic deformation was limited to the area of the Newark beds; it may have extended some way on either side; but it presumably faded out at no great distance, for it has not been detected in the history of the Atlantic and Mississippi regions remote from the Newark belt. In the district of the central folds of Pennsylvania, with which we are particularly concerned, this deformation was probably expressed in a further folding and over- pushing of the already partly folded beds, with rapidly decreasing effect to the northwest; and perhaps also by slip- faults, which at the surface of the ground nearly followed the bedding planes: but this is evidently hypothetical to a high degree. The essential point for our subsequent consideration is that the Jurassic deformation was probably accompanied by a moderate elevation, for it allowed the erosion of the Newark beds and of laterally adjacent areas as well. 11. Jura-Cretaceous denudation.—In consequence of this elevation, a new cycle of erosion was entered upon, which I shall call the Jura-Cretaceous cycle. It allowed the accomplishment of a vast work, which ended in the production of a general lowland of denudation, a wide area of faint relief, whose elevated remnants are now to be seen in the even ridge-crests that so strongly characterize the central district, as well as in certain other even uplands, now etched by the erosion of a later cycle of destructive work. I shall not here take space for the deliberate statement of the argument leading to this end, but its elements are as follows: the extraordinarily persistent accordance among the crest-line altitudes of many Medina and Carboniferous ridges in the central district; the generally corresponding elevation of the western plateau surface, itself a surface of erosion, but now trenched by relatively deep and narrow valleys; the generally uniform and consistent altitude of the uplands in the crystalline highlands of northern New Jersey and in the South Mountains of Pennsylvania; and the extension of the same general surface, descending slowly eastward, over the even crest-lines of the Newark trap ridges. Besides the evidence of less continental elevation thus deduced from the topography, it may be noted that a lower stand of the land in Cretaceous time than now is indicated by the erosion that the Cretaceous beds have suffered in consequence of the elevation that followed their deposition. The Cretaceous transgression in the western states doubtless bears on the problem also. Finally it may be fairly urged that it is more accordant with what is known about old mountains in general to suppose that their mass has stood at different attitudes with respect to base level during their long period of denudation than to suppose that they have held one attitude through all the time since their deformation. It is natural enough that the former maintenance of some lower altitude than the present should have expression in the form of the country, if not now extinguished by subsequent erosion. It is simply the reverse of this statement that leads us to the above-stated conclusion. We may be sure that the long maintained period of relative quiet was of great importance in allowing time for the mature adjustment of the rivers of the region, and hence due account must be taken of it in a later section. I say relative quiet, for there were certainly subordinate oscillations of greater or less value; McGee has detected records of one of these about the beginning of Cretaceous time, but its effects are not now known to be of geographic value; that is, they do not now manifest themselves in the form of the present surface of the land, but only in the manner of deposition and ancient erosion of certain deposits.13 Another subordinate oscillation in the sense of a moderate depression seems to have extended through middle and later Cretaceous time, resulting in an inland transgression of the sea and the deposit of the Cretaceous formation unconformably on the previous land surface for a considerable distance beyond the present margin of the formation.14 This is important as affecting our rivers. Although these oscillations were of considerable geological value, I do not think that for the present purposes they call for any primary division of the Jura-Cretaceous cycle; for as the result of this long period of denudation we find but a single record in the great lowland of erosion above described, a record of prime importance in the geographic development of our region, that will often be referred to. The surface of faint relief then completed may be called the Cretaceous baselevel lowland. It may be pictured as a low, undulating plain of wide extent, with a portion of its Atlantic margin submerged and covered over with a relatively thin marine deposit of sands, marls and clays. 13 Amer. Jour. Science, xxxv, 1888, 367, 448. 14 This statement is based on a study of the geographic evolution of northern New Jersey, in preparation for publication. 12. Tertiary elevation and denudation.—This broad lowland is a lowland no longer. It has been raised over the greater part of its area into a highland, with an elevation of from one to three thousand feet, sloping gently eastward and descending under the Atlantic level near the present margin of the Cretaceous formation. The elevation seems to have taken place early in Tertiary time, and will be referred to as of that date. Opportunity was then given for the revival of the previously exhausted forces of denudation, and as a consequence we now see the formerly even surface of the plain greatly roughened by the incision of deep valleys and the opening of broad lowlands on its softer rocks. Only the harder rocks retain indications of the even surface which once stretched continuously across the whole area. The best indication of the average altitude at which the mass stood through the greater part of post-Cretaceous time is to be found on the weak shales of the Newark formation in New Jersey and Pennsylvania, and on the weak Cambrian limestones of the great Kittatinny valley; for both of these areas have been actually almost baselevelled again in the Tertiary cycle. They will be referred to as the Tertiary baselevel lowlands; and the valleys corresponding to them, cut in the harder rocks, as well as the rolling lowlands between the ridges of the central district of Pennsylvania will be regarded as of the same date. Whatever variations of level occurred in this cycle of development do not seem to have left marks of importance on the inland surface, though they may have had greater significance near the coast. 13. Later changes of level.—Again at the close of Tertiary time, there was an elevation of moderate amount, and to this may be referred the trenches that are so distinctly cut across the Tertiary baselevel lowland by the larger rivers, as well as the lateral shallower channels of the smaller streams. This will be called the Quaternary cycle; and for the present no further mention of the oscillations known to have occurred in this division of time need be considered; the reader may find careful discussion of them in the paper by McGee, above referred to. It is proper that I should add that the suggestion of baselevelling both of the crest-lines and of the lowlands, that I have found so profitable in this and other work, is due largely to personal conference with Messrs. Gilbert and McGee of the Geological Survey; but it is not desired to make them in any way responsible for the statements here given. FIG. 4. FIG. 5. 14. Illustrations of Pennsylvanian topography.—A few sketches made during a recent recess-trip with several students through Pennsylvania may be introduced in this connection. The first, fig. 4, is a view from Jenny Jump mountain, on the northwestern side of the New Jersey highlands, looking northwest across the Kittatinny valley-lowland to Blue or Kittatinny mountain, where it is cut at the Delaware Water-gap. The extraordinarily level crest of the mountain preserves record of the Cretaceous baselevel lowland; since the elevation of this ancient lowland, its softer rocks have, as it were, been etched out, leaving the harder ones in relief; thus the present valley-lowland is to be explained. In consequence of the still later elevation of less amount, the Delaware has cut a trench in the present lowland, which is partly seen to the left in the sketch. Fig. 5 is a general view of the Lehigh plateau and cañon, looking south from Bald Mountain just above Penn Haven Junction. Blue mountain is the most distant crest, seen for a little space. The ridges near and above Mauch Chunk form the other outlines; all rising to an astonishingly even altitude, in spite of their great diversity of structure. Before the existing valleys were excavated, the upland surface must have been an even plain—the Cretaceous baselevel lowland elevated into a plateau. The valleys cut into the plateau during the Tertiary cycle are narrow here, because the rocks are mostly hard. The steep slopes of the cañon-like valley of the Lehigh and the even crests of the ridges manifestly belong to different cycles of development. Figs. 6 and 7 are gaps cut in Black Log and Shade mountain, by a small upper branch stream of the Juniata in southeastern Huntingdon county. The stream traverses a breached anticlinal of Medina sandstone, of which these mountains are the lateral members. A long narrow valley is opened on the axial Trenton limestone between the two. The gaps are not opposite to each other, and therefore in looking through either gap from the outer country the even crest of the further ridge is seen beyond the axial valley. The gap in Black Log mountain, fig. 6, is located on a small fracture, but in this respect it is unlike most of its fellows.15 The striking similarity of the two views illustrates the uniformity that so strongly characterizes the Medina ridges of the central district. Fig. 8 is in good part an ideal view, based on sketches on the upper Susquehanna, and designed to present a typical illustration of the more significant features of the region. It shows the even crest-lines of a high Medina or Pocono ridge in the background, retaining the form given to it in the Cretaceous cycle; the even lowlands in the foreground, opened on the weaker Siluro-Devonian rocks in the Tertiary cycle; and the uneven ridges in the middle distance marking the Oriskany and Chemung beds of intermediate hardness that have lost the Cretaceous level and yet have not been reduced to the Tertiary lowland. The Susquehanna flows distinctly below the lowland plain, and the small side streams run in narrow trenches of late Tertiary and Quaternary date. 15 Second Geol. Surv. Pa., Report T3, 19. FIG. 6. FIG. 7. FIG. 8. If this interpretation is accepted, and the Permian mountains are seen to have been once greatly reduced and at a later time worn out, while the ridges of to-day are merely the relief left by the etching of Tertiary valleys in a Cretaceous baselevelled lowland, then we may well conclude with Powell that "mountains cannot remain long as mountains; they are ephemeral topographic forms."16 16 Geol. Uinta Mountains, 1876, 196. PART THIRD. General conception of the history of a river. 15. The complete cycle of river life: youth, adolescence, maturity and old age.—The general outline of an ideal river's history may be now considered, preparatory to examining the special history of the rivers of Pennsylvania, as controlled by the geological events just narrated. Rivers are so long lived and survive with more or less modification so many changes in the attitude and even in the structure of the land, that the best way of entering on their discussion seems to be to examine the development of an ideal river of simple history, and from the general features thus discovered, it may then be possible to unravel the complex sequence of events that leads to the present condition of actual rivers of complicated history. A river that is established on a new land may be called an original river. It must at first be of the kind known as a consequent river, for it has no ancestor from which to be derived. Examples of simple original rivers may be seen in young plains, of which southern New Jersey furnishes a fair illustration. Examples of essentially original rivers may be seen also in regions of recent and rapid displacement, such as the Jura or the broken country of southern Idaho, where the directly consequent character of the drainage leads us to conclude that, if any rivers occupied these regions before their recent deformation, they were so completely extinguished by the newly made slopes that we see nothing of them now. Once established, an original river advances through its long life, manifesting certain peculiarities of youth, maturity and old age, by which its successive stages of growth may be recognized without much difficulty. For the sake of simplicity, let us suppose the land mass, on which an original river has begun its work, stands perfectly still after its first elevation or deformation, and so remains until the river has completed its task of carrying away all the mass of rocks that rise above its baselevel. This lapse of time will be called a cycle in the life of a river. A complete cycle is a long measure of time in regions of great elevation or of hard rocks; but whether or not any river ever passed through a single cycle of life without interruption we need not now inquire. Our purpose is only to learn what changes it would experience if it did thus develop steadily from infancy to old age without disturbance. In its infancy, the river drains its basin imperfectly; for it is then embarrassed by the original inequalities of the surface, and lakes collect in all the depressions. At such time, the ratio of evaporation to rainfall is relatively large, and the ratio of transported land waste to rainfall is small. The channels followed by the streams that compose the river as a whole are narrow and shallow, and their number is small compared to that which will be developed at a later stage. The divides by which the side-streams are separated are poorly marked, and in level countries are surfaces of considerable area and not lines at all. It is only in the later maturity of a system that the divides are reduced to lines by the consumption of the softer rocks on either side. The difference between constructional forms and those forms that are due to the action of denuding forces is in a general way so easily recognized, that immaturity and maturity of a drainage area can be readily discriminated. In the truly infantile drainage system of the Red River of the North, the inter-stream areas are so absolutely flat that water collects on them in wet weather, not having either original structural slope or subsequently developed denuded slope to lead it to the streams. On the almost equally young lava blocks of southern Oregon, the well-marked slopes are as yet hardly channeled by the flow of rain down them, and the depressions among the tilted blocks are still undrained, unfilled basins. As the river becomes adolescent, its channels are deepened and all the larger ones descend close to baselevel. If local contrasts of hardness allow a quick deepening of the down-stream part of the channel, while the part next up-stream resists erosion, a cascade or waterfall results; but like the lakes of earlier youth, it is evanescent, and endures but a small part of the whole cycle of growth; but the falls on the small headwater streams of a large river may last into its maturity, just as there are young twigs on the branches of a large tree. With the deepening of the channels, there comes an increase in the number of gulleys on the slopes of the channel; the gulleys grow into ravines and these into side valleys, joining their master streams at right angles (La Noë and Margerie). With their continued development, the maturity of the system is reached; it is marked by an almost complete acquisition of every part of the original constructional surface by erosion under the guidance of the streams, so that every drop of rain that falls finds a way prepared to lead it to a stream and then to the ocean, its goal. The lakes of initial imperfection have long since disappeared; the waterfalls of adolescence have been worn back, unless on the still young headwaters. With the increase of the number of side- streams, ramifying into all parts of the drainage basin, there is a proportionate increase in the surface of the valley slopes, and with this comes an increase in the rate of waste under atmospheric forces; hence it is at maturity that the river receives and carries the greatest load; indeed, the increase may be carried so far that the lower trunk-stream, of gentle slope in its early maturity, is unable to carry the load brought to it by the upper branches, and therefore resorts to the temporary expedient of laying it aside in a flood-plain. The level of the flood-plain is sometimes built up faster than the small side-streams of the lower course can fill their valleys, and hence they are converte...