While speculation about Pennsylvania drainage route origins and more specifically water and wind gap origins can be traced back to the early 1700s the questions still persist and the speculation continues[1] This inability of researchers to explain Pennsylvania drainage features, such as the large-scale Susquehanna River and Delaware River water gaps, is puzzling as similar features are found throughout the United States and in numerous other regions and are also unexplained or poorly explained. The geological sciences have made great advances since the early 1700s and can provide very precise explanations for many geologic features and events, yet Pennsylvania and most other drainage routes are as much a mystery today as they were 200 years ago. Pennsylvania drainage routes exist and they originated in some manner and it should be possible to determine how the drainage routes originated, yet there still is no accepted interpretation that explains Pennsylvania drainage route origins. Could the problem be that researchers investigating Pennsylvania drainage route origins have repeatedly been making the same flawed assumptions, which have prevented the discovery of solutions to the Pennsylvania drainage route puzzles?

19^th century Speculative Hypotheses and Assumptions

To test the hypothesis that flawed assumptions have been preventing researchers from solving Pennsylvania drainage route origin questions this essay seeks to identify and evaluate assumptions previous drainage route researchers have made. Up until the mid 1800s most researchers speculated that oceanic water movements of various types were responsible for eroding Pennsylvania and other drainage routes, including the spectacular water and wind gaps. These early speculative hypotheses probably had at least some roots in then prevalent religious beliefs, although these hypotheses were also supported by real geologic evidence reported by reputable geologists. This reputable evidence included descriptions of and explanations for diluvium, which today is interpreted to be glacial and lag gravel deposits. While many Pennsylvania areas are south of the recognized continental ice sheet margin diluvium is abundantly present just to the north in New York State. James Hall, in his 1838 report describes the diluvium as follows:

“The indications of diluvial action are every where perceived in the accumulations of gravel, sand and pebbles, with boulders of all dimensions strewed over the surface…In pursuing these investigations we are irresistibly led to the conclusion, that an oceanic current has flowed over the limestone terrace from the Genesee to the Niagara river, and indeed over the whole western part of our state…These materials all attest the action of violent currents, and not of a single and uniform current, but of opposite or conflicting ones. The presence, in the same locality, of boulders from the north, with those from the south, proves that opposite forces have prevailed either at the same or at different periods…. The extent of these diluvial formations, with the great numbers of erratic blocks, and the evidence of long continued wearing action on the limestone, prove that the force was not sudden and violent, but continued for an indefinite period. If, then, we admit the presence of an ocean covering the continent or a part of it, we should also admit the condition of an ocean as we find it at the present time.”[2]

Once the Louis Agassiz’s glacial hypothesis gained acceptance most diluvium could be explained in terms of one or more continental ice sheets that once covered northern North America and explanations involving “oceanic currents” disappeared from the mainstream geologic literature. Also contributing to rejection of the “oceanic current” hypotheses was the acceptance of Hutton’s principle of uniformitarianism, which among other things, used gradual processes to explain the development of river drainage systems and rejected the hypotheses of violent torrents. Playfair’s comments about rivers in his 1802 Illustrations of the Hutton Theory of the Earth include the following:

“If indeed a river consisted of a single stream, without branches, running in a straight valley, it might be supposed that some great concussion, or some powerful torrent, had opened at once the channel by which the waters are conducted to the ocean; but, when the usual form of a river is considered, the trunk divided into many branches, which rise at a great distance from one another, and these again subdivided into an infinity of smaller ramifications, it becomes strongly impressed upon the mind, that all these channels have been cut by the waters themselves; that they have been slowly dug out by the washing and erosion of the land; and that it is by the repeated touches of the same instrument, that this curious assemblage of lines have been engraved so deeply on the surface of the globe.”[3]

A close look at Playfair’s interpretation of the Hutton proposed uniformitarianism hypothesis suggests several significant, but seemingly reasonable assumptions are being made. First, is the assumption that water originating at numerous points found throughout a drainage basin was responsible for eroding the drainage basin. In terms of what we see today there is no question that the water flowing from a drainage basin originated at numerous points within that drainage basin. However, even though water flowing in modern drainage basins definitely is eroding the drainage basins and transporting the eroded sediment, it is possible the drainage basins were originally eroded under very different conditions than the conditions that exist today. For example, during the melting of a large continental ice sheet massive and prolonged melt water floods may have crossed the landscape and completely overwhelmed all existing drainage routes. Headward erosion of deep valleys from coast lines, or other deeper valleys, during such floods may have systematically captured the massive and prolonged melt water floods so as to systematically create the drainage basins that we see today. If so, the Playfair hypothesis may have led physical geographers and geologists down a dead end street.

Assumptions made by William Morris Davis

Without question William Morris Davis did more to shape late 19^th century and early 20^th century research endeavors related to Pennsylvania drainage routes than any other researcher. Davis (1850-1934) was born and raised in Philadelphia, although his professional career was spent at Harvard and after retirement in various southwest cities. Many of Davis’ ideas were introduced in his classic 1889 National Geographic Magazine article titled “The Rivers and Valleys of Pennsylvania”, subsequently reprinted in Geographical Essays[4]. Davis in that 1889 paper openly accepts the Playfair uniformitarianism interpretation with statements like the following:

“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…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 location of the streams in their present courses; to…trace the development of several river systems…from their ancient beginnings to the present time.”[5]

By making these statements Davis is ruling out the possibility of massive and prolonged glacial melt water floods that overwhelmed all pre-existing drainage routes so as to enable headward erosion of deep valleys while creating a completely new and different drainage system. Instead Davis assumed major river systems descended from late Paleozoic drainage routes through a gradual evolutionary process of captures and flow reversals. Based on stratigraphic studies Davis determined the original (late Paleozoic) master stream in present day Pennsylvania was what he called the northwest flowing Anthracite River, which flowed from highlands in the east. While this assumption may be valid for late Paleozoic time, the implication that modern drainage systems are descendants of such a late Paleozoic drainage system requires the assumption that nothing happened to interrupt Davis’s proposed gradual drainage evolution.

The very title of his Pennsylvania rivers and valleys paper implies that Davis is also focusing his research attention on Pennsylvania’s major rivers and their valleys and has made the unstated assumption that Pennsylvania drainage divides and first and second order valleys, perhaps due their complexity or numbers, have little or no research value. By ignoring Pennsylvania drainage divide areas and the state’s first and second order drainage basins Davis could not see and consider evidence that water at one time or another crossed all present day Pennsylvania drainage divides. Failure to see such evidence further supported Davis’ assumption that Pennsylvania drainage basins had evolved gradually over time and contributed to his failure to consider hypotheses that Pennsylvania’s drainage basins originated during massive and prolonged melt water floods that overwhelmed whatever drainage system existed previously.

Davis near the conclusion of his rivers and valleys of Pennsylvania paper makes what is perhaps his most telling assumptions. “If this theory of the history of our rivers is correct, it follows that any one river as it now exists is of so complicated an origin that its development cannot become a matter of general study and must unhappily remain only a subject for special investigation for some time to come…The history of the Susquehanna, the Juniata, or the Schuylkill, is too involved with complex changes, if not enshrouded in mystery, to become intelligible to any but advanced students…”[6] In other words, Davis is assuming Pennsylvania drainage history is so complex that only the most obvious and prominent drainage features can be studied and explained. By assuming Pennsylvania drainage basins did not develop in an identifiable sequence Davis and his followers (who made a similar assumption) did not look to see if Pennsylvania drainage basins had originated in an identifiable sequence and consequently found none.

Assumptions made by Davis’ followers

Davis’ 1889 paper stimulated additional research as others attempted to better explain Pennsylvania drainage route origins. Douglas Johnson, who had been one Davis’s graduate students at Harvard and who then went on to a distinguished career at Columbia University, in 1931 proposed regional superposition from a coastal plain cover, which had destroyed the pre-existing drainage, as a way to explain Pennsylvania’s numerous water and wind gaps[7]. This hypothesis assumed that during Cretaceous time much of eastern and perhaps even central Pennsylvania was below sea level and had been covered by marine sediments. According to the hypothesis the routes of Pennsylvania’s rivers were established when the region was uplifted to again be above sea level at a time when the Cretaceous marine rocks completely buried the underlying structures. Erosion subsequently completely removed the Cretaceous marine rocks as the river valleys eroded downward into the underlying bedrock. The hypothesis solved the problem of how rivers could cut the deep water and wind gaps seen today, but could not be proved because there is no evidence the Cretaceous marine cover ever existed.

In spite of the lack of evidence for Johnson’s Cretaceous marine cover Johnson’s hypothesis was for many years considered to be the most acceptable of many hypotheses that had been proposed to explain the Pennsylvania drainage routes. Shortly after the publication of Johnson’s hypothesis Ruedemann[8] proposed an early Susquehanna River had flowed in a southwest direction from headwaters north and east of the present day Mohawk-Hudson River valley on what he considered to have been the Permian surface (much as Davis had tried to reconstruct the Permian surface to establish his Anthracite River route). Ruedemann’s southwest flowing Susquehanna River had a long course to reach the sea and was susceptible to capture by younger southeast oriented streams with more direct routes to the sea. While stream capture evidence does support a hypothesis that southeast oriented drainage did capture southwest oriented drainage and that southwest oriented water once moved from north and east of the present day Mohawk-Hudson River valley into the present day Susquehanna drainage basin, Ruedemann’s hypothesis fails to explain much of the stream capture evidence and, like the Davis 1889 hypothesis, assumes present day drainage routes descended through gradual and uninterrupted evolutionary processes from the hypothesized Permian drainage routes.

Meyerhoff and Olmstead in 1938, like Davis and Ruedemann, started with the assumption that the Appalachian orogeny destroyed the previous drainage and that present day drainage routes are descendants, through evolutionary processes, of Permo-Triassic drainage routes established following the orogeny. Unlike Davis, however, they saw the divide between the Atlantic drainage and the western drainage to be west or northwest of the Folded Appalachians meaning the post orogeny drainage was to the southeast and despite numerous adjustments during the Mesozoic and Cenozoic, many present day southeast oriented drainage routes still follow routes established in the early Triassic. Adjustments made since the early Triassic included captures that occurred along lines of structural weakness (e.g. faults), which enabled streams and rivers to erode deep water gaps as their valleys eroded deeper and deeper into the underlying bedrock.

Strahler, who was studying under Johnson’s supervision and who subsequently filled Johnson’s Columbia University professorship following Johnson’s death, in 1945 critically examined all hypotheses advanced to that time for the origin of Appalachian drainage routes, limiting himself to problems and evidence found in the Folded Appalachians of Pennsylvania[9]. Contrary to statements made by Meyerhoff and Olmstead and by others Strahler was unable to find evidence of faulting at most water and wind gap locations. Further, Strahler found locations where there were obvious faults along ridges that streams had not utilized to erode through the rocks so as to create a water gap. As a result of his detailed analysis he was able to eliminate all hypotheses proposed up to that time except Johnson’s hypothesis of a widespread coastal plain cover. Strahler did not consider his findings to be proof that a coastal plain cover had existed, but it was the only explanation that he could not eliminate.

Marie Morisawa in a 1989 paper[10] reviews progress made by the above and other geomorphologists since publication of Davis’s 1889 Rivers and Valleys of Pennsylvania paper. She states, “Before and since Davis’ 1889 paper the explanation of the drainage of the Appalachians has intrigued geomorphologists. Those hypotheses advanced since Davis have relied primarily on various solutions he suggested. Despite the constriction on him by geologic paradigms of his day, Davis’ agile mind managed to conceive basic principles which affected future studies of geomorphology.” In other words, since Davis’ time most geomorphologists have used many of the same assumptions that Davis used and as a result are no closer to understanding the origin of Pennsylvania drainage routes today than Davis was more than 100 years ago. Perhaps the most telling assumption these researchers have made is stated by Morisawa, who says that the “present drainage is the result of many processes, still on-going.” In other words all previous researchers have assumed Pennsylvania drainage routes evolved over long periods of time in some type of uninterrupted process. And all previous researchers have assumed that Pennsylvania drainage routes are so complex and complicated that most details of their origins will never be determined.

Changing the assumptions

As seen in the above discussion William Morris Davis defined assumptions that he and future geomorphologists made while trying to decipher Pennsylvania drainage route histories. While different workers have used varying subsets of these assumptions none of these workers has developed an explanation that satisfactorily explains the evidence they worked with (and all workers limited themselves to major rivers, which represent only a small fraction of the evidence needing to be explained). This inability to satisfactorily explain Pennsylvania drainage history suggests some or all of Davis’ assumptions are flawed. Perhaps the present day drainage did not continuously evolve over long periods of time, but is the result of some as yet unrecognized event that completely altered the pre-existing drainage. Perhaps the evidence needing to be studied is not the major rivers, but the drainage divides and the first and second order streams that originate along or near those drainage divides. Perhaps the melting of continental ice sheets produced massive and prolonged melt water floods that overwhelmed whatever drainage Pennsylvania previously had and then eroded new sets of deep valleys and tributary valleys to create the drainage routes now seen in Pennsylvania. Entries on this website illustrate how such massive and prolonged melt water floods can explain numerous Philadelphia area drainage features that Davis and his followers claim are too complicated to explain. Perhaps the early geologists who used ocean currents to explain the origin of Pennsylvania water and wind gaps were not completely wrong. Perhaps continental ice sheets were like oceans on the land and that when they melted those frozen oceans were unlocked so water from those high level oceans flowed across Pennsylvania to reach the Atlantic Ocean and the Gulf of Mexico and in the process created Pennsylvania’s present day drainage systems with their water and wind gaps.

[1] Sevon, W.D., 1986, Susquehanna River water gaps: many years of speculations. Pennsylvania Geology. 17(3): pp. 4-7.

[2] Hall, James, 1838, Second Annual Report of the Fourth Geological District of New York: Assembly Document 200. pp. 307-308.

[3] Playfair, John, 1802, in Kirtley F. Mather and Shirley L. Mason, 1970, A Source Book in Geology, 1400-1900, Harvard University Press, Cambridge, MA. P. 132.

[4] Davis, W.M., 1889. The rivers and valleys of Pennsylvania. National Geographic Magazine, 1: 183-253: reprinted in Geographical Essays, Ginn and Company, 1909.

[5] Davis in Essays, p. 413

[6] Davis in Essays, p.482

[7] Johnson, D.W.,, 1931, Stream Sculpture on the Atlantic Slope. Columbia University Press, New York, N.Y., 142 pp.

[8] Ruedemann, R., 1932, Development of drainage of the Catskills. American Journal of Science, 5^th Series, 23:337-349.

[9] Strahler, A.N., 1945, Hypotheses of stream development in the Folded Appalachians of Pennsylvania. Geological Society of America Bulletin, 56: 45-88.

[10] Morisawa, M., 1989, Rivers and valleys of Pennsylvania, revisited. In: T.W. Gardner and W.D. Sevon (Editors), Appalachian Geomorphology. Geomorphology, 2:1-22.