Sanders, J. E., and Merguerian, Charles, 1994b, The glacial geology of New York City and vicinity, p. 93-200 in A. I. Benimoff, ed., The Geology of Staten Island, New York, Field guide and proceedings, The Geological Association of New Jersey, XI Annual Meeting, 296 p.
ABSTRACT
The fundamental question pertaining to the Pleistocene features of the
We here summarize our reasons for re-establishing Fuller's 4-glaciation
classification and for rejecting a latest-Pleistocene age of
We accept as Woodfordian only the youngest till deposited by a glacier that
flowed from NNE to SSW, down the
The next-older till, our III, was deposited by a glacier that flowed over the
Farther E on
In the subsurface at
The next-older glacial advance, our No. II, which also crossed the
We regard the newly exposed Gilbert-type delta foresets in Fuller's Manhasset
Formation as deposits of
Glacier III (and/or II; we cannot as yet distinguish between their erosive effects) sculpted the bedrock with prominent grooves trending NW-SE as found in Manhattan by Dr. L. D. Gale in 1828-29 (published in Mather, 1843), in Central Park by Hanley and Graff (1976) and in many parts of the New York City region by us (Merguerian and Sanders, 1988; 1991a, e; 1993a, c; 1994a; Sanders and Merguerian, 1991a, 1992). We ascribe the numerous effects of glacial flow oriented NW-SE, as reported from the erratics in the Brooklyn Botanical Garden (Gager, 1932) as shown on the Glacial Map of North America (Flint, 1945) and also as found by C. A. Kaye (1982) in Boston, MA, and vicinity to glacier(s) II and/or III.
Tills II and III, separated by a few meters of reddish-brown outwash of the
kind deposited in a proglacial lake, are exposed in eroding bluffs along the E
shore of the
The earliest glacier to leave its mark in
Glacier No. I deposited the gray-brown till that underlies red-brown till (No. II?) at Teller's Point, Westchester Co. and the gray-brown till at Target Rock, L.I., which contains "greenstone" indicator stones from the Maltby Lakes metavolcanics SW of New Haven [misidentified by Sirkin and Mills (1975) as Palisades Dolerite and thus assigned by them to a till that was deposited by a glacier that flowed NW-SE; the upper till here, deposited by a glacier that did flow from NW to SE, is gray and contains no Newark erratics but rather indicator stones, such as Inwood Marble and the Cortlandt Complex, derived from what we have called the "crystalline corridor" of southeastern NY and western CT].
At least two pre-Woodfordian glaciations are implied by the relationships at three localities on Staten Island: (1) Till, exposed in coastal cliffs eroded into a terminal-moraine ridge in southern Staten Island, containing but-minor quantities of decayed pebbles (only the "greenstones") and capped by a well-developed paleosol, yields provenance data which prove that at least one glacier flowed regionally across Staten Island from NW to SE (across the Hudson Valley, thus our Till II and/or III) not NNE to SSW, (down the Hudson Valley) with local diversions to the SE, the pattern inferred for the Woodfordian ice by R. D. Salisbury (1902) and accepted by many one-glacier advocates. (2) At the AKR Excavating Corp., much-decayed stones in outwash gravels that underlie comparatively fresh red-brown till and overlie Cretaceous sands imply a pre-Wisconsinan age (possibly a product of our Episode I). (3) Superposed glacial striae and -crescentic marks on the dolerite exposed at the Graniteville quarry are inferred products of two ice-flow directions: an older NW to SE (our II and/or III) cut by a younger NNE to SSW (our No. IV).
We infer that on
We have not yet turned up any absolute-age data that would settle the age assignments of our multiple-glacier interpretation and that would totally destroy the one-glacier-did-it-all view which we think is not correct. However, we think the case we have made for the pre-Woodfordian age for the Harbor Hill Moraine is very compelling. If we have correctly interpreted the subsurface relationships Rampino (1978 ms.) established at Jones Beach, then Long Island's two world-famous terminal moraines were not only not made by the fluctuating margin of the Woodfordian glacier, as has been universally believed for many years, but were made by two different glaciers whose appearance on Long Island was separated by an interglacial episode when the glacier retreated back into Canada and the sea rose nearly to its present level.
At present, the best hope for settling our chronological impasse is
amino-acid-racemization analysis of shells from the Wantagh Formation.
Our inference is that the
INTRODUCTION
This article has been written to provide regional geologic background about the
Pleistocene sediments in the
In this article we: (1) briefly summarize selected items from the large geologic literature about local Pleistocene stratigraphy and -glacial history; (2) present our credentials and background and reasons for studying New York City's glacial history and offer a new stratigraphic classification; (3) summarize our field results under the headings of (a) inferred directions of glacial flow based on eroded bedrock in the New York City region, (b) directions of glacial flow inferred from sediments (erratics and indicator stones and long axes of drumlins), (c) discussion of the ice-flow evidence, and (d) stratigraphic superposition of tills and Pleistocene sediments; and (4) a summary of the glacial geology of Staten Island with respect to our proposed stratigraphic classification.
SUMMARY OF SELECTED GEOLOGIC LITERATURE ABOUT LOCAL PLEISTOCENE STRATIGRAPHY AND -GEOLOGIC HISTORY
In our review of previous work, we make no attempt at complete coverage. Moreover, in this section we include only major articles about the stratigraphic relationships with emphasis on the number of glaciations and the glacial history particularly with respect to the ages of the terminal-moraine ridges. We defer our attempts to evaluate these previous results until after we have presented our own results. We include some previous results on orientations of striae and on indicator stones alongside our own work.
Work Done Prior to 1914: Concept of One Glacial Episode is Born
In this category we include published work by L. D. Gale (1828-29, published in Mather, 1843), by James Gates Percival (1842) in Connecticut; by Mather (1843), by Warren Upham (1879), by T. C. Chamberlin (1883); by R. D. Salisbury (Salisbury, 1902, 1908; Salisbury and others, 1902; Salisbury and Peet, 1894; Peet, 1904); and by J. B. Woodworth (1901). Because of the difficulty in locating many of these old references, we quote from some of them at length. (Fuller, 1914, p. 4-19 summarizes all of the pre-1914 literature on Long Island Pleistocene sediments.)
Figure GG-1. Physiographic sketch map of
The first systematic attempt to record directions of striae and grooves on the
bedrock in
Back then, most of the present-day streets had been laid out, but only a few buildings
existed north of what is now known as
As was common in his day, Gale supposed that the grooves and scratches had been made by water currents, perhaps assisted by icebergs. The presumed significance of water is implied in his use of the term "diluvial." L. D. Gale (1839, Geological Report of New-York; New-York island; published in Mather, 1843, p. 209-210):
"Diluvial grooves and scratches have been found in
every section of the island, from Sixteenth-street
on the south, to 200th-street on the north, (or to
the southern termination of the limestone;) and
from the banks of the
river on the east. The furrows generally are most
distinct where the rock has been recently uncovered,
and least where it has been long exposed to the
action of the elements. They have been found on
the highest rocks, and at the lowest tide-water
marks, being a difference of more than one hundred
feet perpendicular height. The furrows are always
most strongly marked on the northwestern slopes of
the hills, and least so on the southeastern. In
many instances they are very distinct on the
western and northwestern slopes, extending to the
highest point of the rock; but no traces are to be
seen on the eastern and southeastern slopes,
although both slopes are equally exposed.
"Direction of the furrows. Observations of
the diluvial furrows were made in between sixty and
seventy different places on the island. Taking
together the whole series of observations, the
general course of the current was from northwest to
southeast, or north forty-five degrees west, but
varied in the extremes from north twenty-five
degrees west to north forty-eight west, making a
difference of twenty-three degrees. Of the whole
series of observations, thirty-nine were north
forty-five degrees west, twelve varied from north
forty-five degrees west (seven being north thirty-
five degrees west), two were north forty-eight
degrees west, and a few scattering ones varying
from north thirty-five degrees west to north forty-
five degrees west.
"Abundance of the furrows. The furrows occur
most abundantly in the middle portions of the island,
between the city and the Harlem and Manhattanville
valley, somewhat less in the western, and least of
all in the eastern.
"Direction of the furrows in particular
neighborhoods. Half of all the places where the
furrows were noticed were in the middle portion of
the island, in the line of the Eighth avenue from
Sixtieth-street to 105th-street, where without
exception the direction is north forty-five degrees
west. About one fourth of all are on the west side,
and vary but little from north thirty-five degrees
west; and about one-eighth on the eastern side, where
the direction varies from north twenty-five degrees
west to north thirty-five degrees west. In
connection with this subject, I have examined the
surface of the greenstone on the neighboring shores
of New-Jersey (sic), and find their grooves and
scratches abundant, and their general direc-/ (p. 210):
tion is north forty-five degrees west. Hence it
appears, that the diluvial current which once swept
over this island from northwest to southeast, on
reaching the western shore, was deflected southward,
as by the action of some force at a right or some
other angle to its course; and that the same current,
before it reached the middle of the island, again
assumed a southeasterly direction, but was again
diverted southerly on approaching the eastern shore.
That some portion of the current was diverted
southerly on reaching the western shore of the
island, is evident, not only from the diluvial
furrows, but from the boulders of anthophyllite
found in large numbers in the lower part of the
Eighth avenue near Fifteenth-street, a distance of
two miles in a south-southwest direction from the
only locality whence they could have proceeded.
Again, the white limestone of Kingsbridge has been
distributed along the eastern shore of the
island, in a direction almost due south of the only
locality in the vicinity where it is found in place;
whereas had they been carried in the general
direction of the current, they would have been
deposited eastward in Westchester county, as before
stated.
"Magnitude of the furrows. The size of the
furrows varies in the same and different localities.
Sometimes they are the finest scratches, not more
than a line in diameter horizontally, and of the
smallest appreciable depth; from this they increase
to grooves four inches deep and eighteen inches in
horizontal diameter. In a few cases, they are furrows,
or rather troughs, more than two feet wide and six
or eight inches deep. A case of the latter kind
occurs on Eighth avenue, between Seventy-ninth and
Eight-first-streets; and one of the former on the west
side of the island, on the very banks of the Hudson,
five hundred yards north of Mr. John H. Howland's
country seat (near Ninety-seventh-street).
"Convenient places for examining the diluvial
furrows. The nearest places to the city for
examining the furrows are at the junction of Twenty-
second-street and First avenue, south of the Almshouse
yard; and again about half a mile northward at Kip's
bay, at the junction of First avenue and Thirty-fifth-
street. Both of these localities will soon be
destroyed by grading the streets. Some of the most
interesting localities have been made known by cutting
through Eighth avenue, from Bloomingdale road, at or
near Sixtieth-street, to Harlem and Manhattanville
valley at 105th-street; these locations are on both
sides of the avenue, and very conspicuous. Another,
equally interesting in many respects, is on the banks
of the Hudson west of the Bloomingdale road, about
six miles from the city, and about six hundred yards
northwest of Burnham's hotel. The interest excited
by this locality arises from the fact, that the
furrows ascend from beneath the lowest tide water,
up to an elevation of seventy feet in three hundred
or four hundred feet distance."
Gale's observations clearly suggest the effects of two contrasting flow directions, (a) nearly all the "diluvial scratches and furrows" indicating flow from the NW to the SE and (b) the displacement of indicator erratics (the anthophyllite-bearing rock and the white limestone) showing transport from the NNE to the SSW. Yet his interpretation of his data was that of a single event, which he expressed as "the diluvial current." Gale tried to show how the changes in flow of a single such current could account for both the regional trends of the scratches and furrows on the smoothed bedrock and the displaced indicator erratics. In this regard, Gale began a pattern that would be followed by most subsequent students of the "diluvial" deposits: trying to account for all the disparate observations: trying to force fit all data into a single transport event. But Gale's single transport event differed significantly from the one favored by later investigators. Gale concluded that his single "diluvial current" had flowed from NW to SE and he sought aberrations in this flow direction to account for the displacement from NNE to SSW of indicator erratics.
Independently of and nearly simultaneously with Gale's survey of Manhattan, that great genius of Connecticut geology, James Gates Percival, was mapping the geology of the state of Connecticut. Percival was well acquainted with both the bedrock and what we would now refer to as the Pleistocene deposits. As did Gale, Percival classified these as "Diluvium," or the "unstratified (sic) materials," and contrasted them with the Alluvium, "those arranged in strata." Percival (1842, p. 453-456) cited many examples of distinctive kinds of rocks that had been displaced from NW to SE:
"The greater part of the Diluvium was apparently
deposited by a general current, traversing the surface
from N. N. W. to S. S. E. This is satisfactorily
indicated both by the bowlders, scattered over the
surface, or imbedded in the diluvial earth, and by the
smaller fragments included in the latter, as well as
by its general character (sic)." (Percival, 1842, p. 453).
Percival emphasized that knowledge of the composition of the bedrock was absolutely essential for reconstructing the directions of the "diluvial currents:"
"In order to determine the direction of the
diluvial currents, a particular knowledge of the
local character (sic) of the rocks, as indicated in
the account already given of the different local
formations, is indispensable. Several of these local
formations are so peculiar in the character (sic) of
their rocks, that the latter cannot be mistaken, to
whatever distance they may have been transported.
These, by the distribution of their bowlders and
fragments, furnish conclusive evidence that the
more general (sic) direction of the diluvial current
was S. S. E." (Percival, 1842, p. 454)
Despite the numerous examples he cited that demonstrate transport from NW to SE, Percival reported that some rocks had been moved from NNE to SSW. As did Gale in Manhattan, Percival supposed that this transport to the SSW had resulted from local deflections of the general SSE-flowing diluvial current:
"Although the general direction of the diluvial
current was apparently S. S. E., yet in some instances,
from local obstructions, its course was deflected to a
S. S. W. direction. This is most distinctly obvious
along the Western border of the larger Secondary
formation, where blocks and fragments of the Trap and
Sandstone of that formation are accumulated, sometimes
quite abundantly, in such a direction from their apparent
source." (Percival, 1842, p. 457).
In contrast to both Gale and Percival, the single flow event most later workers invoked was from the NNE to the SSW. They called upon aberrations from this "main-flow" direction to explain the scratches and furrows that trend NW-SE.
Mather (1843) described the geology of Long Island emphasizing strata exposed in the north-shore cliffs (Figure GG-2). At Lloyd's Neck, a storm exposed dipping strata that had been
truncated and are overlain by horizontal strata. Mather was not able to interpret these strata as any modern geologist would do. After all, Mather's date of publication preceded general acceptance of the concept of Pleistocene continental glaciation and was 42 years before G. K. Gilbert (1885) presented his analysis of the topographic features of lake shores in which he proposed the terms topset, foreset, and bottomset as the three kinds of lacustrine deltaic strata formed along the shores of ancient Lake Bonneville, Utah (and 47 years before Gilbert's Lake Bonneville monograph appeared in 1890).
Mather used the name "Long Island Formation" informally for the sediments that he thought underlie most of the island; he assigned this formation to the Tertiary. Mather's term has been abandoned, but we are considering the feasibility of reviving it for the extensive suite of pre-Wisconsinan sediments deposited in a lake the occupied much of what is now Long Island. (We refer to it as Long Island Lake and discuss it in a following section.) Mather mentioned the two prominent curvilinear ridges now known to be terminal moraines, but he did not realize they were of glacial origin.
Warren Upham (1879) mapped and discussed Long Island's two terminal-moraine ridges and associated outwash plains (Figure GG-3). He inferred that each had been built at the margin of a separate glacier. (Upham recognized two tills throughout southern New England and on Long Island.) He noted that on Long Island, till is abundant W of Roslyn and generally absent to the E.
T. C. Chamberlin (1883; 1885; 1895a, b) laid the foundation for the stratigraphic classification of North American Pleistocene deposits. He described the conditions at the margin of the continental glacier in terms of lobes and inferred that the main axis of concentrated flow in eastern New York state had been down the Hudson Valley (1883, map following p. 346). He reviewed the previous work on the Long Island terminal moraines and outwash plains but differed with Upham's assignment of these to two separate glaciers. Based on the lack of differential erosion and dissection of the associated outwash plains, Chamberlin argued that both the terminal-moraine ridges and associated outwash plains were products of the latest glacial episode. Chamberlin's argument has been nearly universally accepted.
R. D. Salisbury and associates mapped the glacial geology of New Jersey and reported ice-flow indicators oriented not only NNE-SSW, as they expected, but also NW-SE, which they did not expect. Early in the studies of the glacial deposits, Salisbury concentrated on the flow indicators on the Palisades ridge. The position of the Palisades ridge was critical with respect to Chamberlin's view that a main axis of accelerated ice flow had been down the Hudson Valley. Chamberlin supposed that within the margins of an ice sheet are localized zones within which the ice flows faster than it does in adjacent areas. At such places of localized faster flow, he imagined that the ice-flow "stream-lines" would be crowded close together. On either side of such supposed zones of concentrated flow, the ice tends to spread out toward each side. Chamberlin had illustrated this concept by sketching a map of ice-flow indicators in the region surrounding Lake Michigan (Figure GG-4).
Figure GG-2. Mather's profile-sections of Long Island. Lloyd's Neck on the North shore cliffs is discussed in text. (Mather, 1843, Plate 4, fig. 16.)
Figure GG-3. Map of Long Island and vicinity showing the locations of the two terminal moraines and profile sections showing subsurface relationships. (Wolff, Sichko, and Leibling, 1987, fig. 14.3, p. 128.)
Salisbury accepted Chamberlin's flow model. But if the zone of fast ice flow had followed the Hudson Valley, as Chamberlin had supposed, then divergent flow across the Palisades ridge should have been from NE to SW. Salisbury and Peet went to considerable trouble to study the glacial geology of the Palisades ridge. After they found virtually all the ridge-crest striae indicating glacial flow from from NW to SE and none from the predicted NE to SW, Salisbury reconciled the situation by shifting the axis of the presumed accelerated flow westward and placing it in the Hackensack Valley. From this inferred zone of concentrated flow down the Hackensack Valley, Salisbury and assistants (1902) thought that the ice had flowed toward the SSE over the Palisades ridge and Manhattan, and toward the SSW over the crests of the Watchung Ridges in New Jersey (Figure GG-5). This was consistent with their findings that glacial-flow indicators over the Watchung ridges had been predominantly from the NNE to the SSW, the predicted flow direction for the Palisades ridge for a fast-flow axis located either in the Hudson Valley or the Hackensack Valley.
Salisbury admitted that the regional distribution of erratics of the distinctive Silurian Green Pond Conglomerate from northwestern New Jersey and the divergent orientations of the glacial grooves and -scratches constituted anomalies to this explanation of marginal-flow divergence within a single glacier. Salisbury acknowledged that another succession of events which could explain the distribution of erratics of Green Pond Conglomerate involved two glaciations, but he merely mentioned the possibility of two contrasting glaciers.
"No single Green Pond mountain conglomerate
bowlder has been found on the ridge. West of
Hackensack, such bowlders are found in abundance,
and this in spite of (sic) the fact that in New
Jersey the movement of the ice along the Green Pond
mountain range was to the southwest, approximately
parallel to the range itself. Glacial movement in
this direction could not have carried bowlders from
the New Jersey part of the the Green Pond mountain
formation to the Hackensack valley. It would seem
that the conglomerate ledges which furnished the
Hackensack valley bowlders must have lain somewhere
north of New Jersey, in the axis of the ice lobe, or
perhaps a little to the west of it, and that the
bowlders derived from this ledge were carried
southward in the direction of ice movement, and
finally out of the valley onto the highlands to the
west by the westerly-diverging currents, but that
they were not brought within the influence of easterly
diverging currents, and therefore were not carried
eastward upon the Palisades ridge. Another hypothesis
which would equally well explain the distribution of
the Green Pond mountain conglomerate bowlders, but for
which there is no demonstrative evidence at hand, is
that these bowlders were carried southeastward from
their parent ledges by an earlier ice movement, the
movement in the last epoch being to the southwest over
or along the Green Pond mountain formation. A good
deal may be said for this suggestion. The distribution of
these bowlders has not been studied beyond the State
of New Jersey" (Salisbury, 1894, p. 180).
All of Salisbury's work on the glacial deposits of New Jersey was based on the flow pattern shown in Figure GG-5. He published it repeatedly in various folios that the U. S. Geological Survey published in the region (New York City, Passaic, Franklin Furnace).
Figure GG-4. Sketch map of area west of Lake Michigan (mostly in Wisconsin, but including parts of Michigan and Illinois), showing concept of divergent flow from a narrow zone (centered above Green Bay, Wisconsin) of rapid flow within an ice sheet. (R. D. Salisbury, 1902, fig. 31.)
Figure GG-5. Map of northeastern New Jersey and southeastern New York showing inferred flow lines within the latest (and supposedly the only) Pleistocene glacier that reached the New York City region. Further explanation in text. (R. D. Salisbury, in Merrill and others, 1902, fig. 12, p. 13; also 1908, fig. 11; also, H. B. Kummel, 1933, fig. 13, p. 66.)
Woodworth (1901) reported on the Pleistocene geology of Queens County. About inferred directions of flow of the glacier(s), he wrote:
"Frontal moraines mark the position of the ice
front. The motion of the ice, at least near its
margin, will tend to be toward that front; hence,
since (sic) the moraine in this part of the island
trends to the south of west, forming a lobate line
across this region and that adjacent in New Jersey,
glacial striae in this part of the island should run
to the east of south. A number of ledges of gneiss
in Long Island City meet (sic) this requirement."
"The southeastward movement of the ice on this side
of the Hudson valley is further attested by the drift.
The moraine from Brooklyn as far east as Oyster Bay
contains trap boulders, the nearest known site of which
rocks is in the Palisade trap ridge on the west bank of
the Hudson river.
"Stratified red sands, also undoubtedly derived
from the area of Triassic red sandstones now found only
on the west bank of the Hudson, occur in a section by
the roadside from Corona to Astoria, being there overlain
by 8 or 9 feet of gray till... " (Woodworth, 1901, p. 652),
With respect to divergent flow directions, Woodworth wrote:
"This fanning of the ice sheet to the eastward
on the east side of the lower Hudson and to the
westward on the west side is consistent with the form
of the moraine across the mouth of the river. The
axis of the lobe thus indicated has been fixed by
Salisbury on the west side of the Palisade trap
ridge. (ftn. 1)
Footnote 1: "Salisbury, R. D. N. J. geol. sur. An. rep't state geol. for 1893. 1894, p. 161." (Woodworth, 1901, p. 653).
Woodworth discussed ancient glacial lakes that lay between the high parts of Long Island and/or a terminal-moraine ridge on the S and the glacier itself on the N and sandy/gravelly delta deposits built into such lakes. He included, as Plate 8, a photograph by Heinrich Ries of Gilbert-type delta foresets and -topsets taken in the large Port Washington sand pit during the early days of its active phase. In Figure GG-6 we have modified Woodworth's figure 9 (p. 658) to show a gap between the ice front and a terminal moraine on the S which could serve as a dam to hold in the water of a proglacial lake. Woodward's map and text clearly indicate that is what he had in mind even though in his figure 9, he showed other relationships.
Woodworth also discussed examples of older gravels assigned to the Columbia Formation (p. 624-637) that include an interstratified thin unit of boulder-bearing till (his "boulder clay bed" of p. 627). He described it as follows:
"The boulder clay bed. In many of the coastal
sections on the north shore an unstratified (sic)
mixture of pebbles, sand and clay in a bed varying
from 3 to 10 feet in thickness may be seen in a
position to indicated that it is interstratified
with these older gravels; but it is only in the
sand pits on Hempstead bay that a bed of this
character (sic) is fully revealed. About half
way up the bluff, or about 100 feet above the bay,
there is a bed of boulder clay from 2 to 3 feet
thick, traceable in all the pits open in 1900
south of Bar beach. The matrix of this bed is
an unctuous dark blue (sic) clay locally sandy
or gravelly. Scattered through it and sometimes
in close contact with each other (sic) are glaciated
boulders ofter over (sic) 1 foot in diameter and
numerous pebbles attesting the glacial origin of
the deposit. Several large boulders examined in
1901 by Dr. F. J. H. Merrill and the writer were
recognized by the first named as having been
transported in all probability from the Adirondacks.
Other small boulders carrying Silurian fossils
indicated their origin in the Hudson Valley north
of the Highlands. The longest journey made by these
materials appears to exceed 200 miles."
In summary, it is clear that by early in the twentieth century, T. C. Chamberlin and R. D. Salisbury had stamped on the glacial geology of the New York City their one-glacier view with flow deviation at the margin based on the behavior of inferred ice lobes. Lost in the shuffle were J. B. Woodworth's important contributions, especially that his older series of gravels and interbedded till form a foundation upon which the Harbor Hill moraine was deposited and with respect to his proof that the direction of flow of the glacier which had deposited the Harbor Hill Moraine had been from NNW to SSE and to his clear evidence for multiple glaciation based on the older unit of interstratifed till and gravels as seen in sand pits and north-shore coastal cliffs.
A note about language: Salisbury and Woodworth always used the term striae as the plural for glacial scratches, whereas T. C. Chamberlin adopted the plural of the attribute word "striation" and wrote "striations." In the interests of being gramatically correct, we accept Salisbury and Woodworth and reject the "striations" of T. C. Chamberlin and his legion of followers.
Figure GG-6. Schematic profile-section showing Gilbert-type deltas on N side of Lake Long Island, which formed in the lowland beteen the ice front on the N and the terminal-moraine ridge on the S. Two levels are shown: the lower at +40 feet and the upper at +80 feet (referenced to modern sea level). Highlands of Long Island, underlain by Cretaceous strata, could also serve as a dam for the lake on the S side. (Sanders and Merguerian, 1994, fig. 1, p. 103; adapted from Woodworth, 1901, fig. 9, p. 658.)
Fuller's (1914) Monographic Results: Four Glacial Episodes
The fundamental study of the stratigraphy of the glacial deposits in the New York metropolitan region is Fuller's (1914) monumental treatise on the geology of Long Island. Fuller found deposits that he interpreted as products of 4 glacial advances; between some of the glacial sediments, he found nonglacial strata. Table GG-1 shows the names- and stratigraphic relationships of Fuller's units. Notice his assignment of the Harbor Hill Moraine and the Ronkonkoma Moraine to the post-Illinoian (Early Wisconsinan); he inferred they are both younger than the Vineyard unconformity but assigned them to the Early Wisconsinan. (Later workers changed Fuller's age assignment from Early Wisconsinan to latest Wisconsinan, or Woodfordian).
Fuller inferred that during the Vineyard erosion interval, which he assigned to the Sangamonian, streams had eroded the deep north-facing valleys. Although Fuller did not discuss the relationship between glaciation and low sea level as contrasted with interglacial conditions and high sea level, he did realize that the extensive valley erosion he assigned to this interval required that base level be relatively low with respect to Long Island. He attributed the low base level to uplift of Long Island.
Fuller applied the name Manhasset formation (with upper Hempstead gravel member, middle Montauk till member, and lower Herod gravel member) to the "Columbia formation" of Woodworth. Because, as mentioned, Fuller inferred that the extensive network of north-flowing valleys had been cut into the Manhasset Formation, he classified this formation as being older than the Vineyard erosion interval. Fuller assigned the Manhasset to the Illinoian. Fuller drew many sketches of strata exposed in coastal cliffs. Among these, he illustrated examples of Manhasset Formation showing one-directional dips (Figures GG-7 and GG-8). We discuss these further in a following section.
Early Harbor Hill Terminal Moraine
Wisconsinan G Ronkonkama Terminal Moraine
Sangamonian I Vineyard Fm. (marine deposits and peat);
surface of erosional unconformity with relief of ca. 300 ft.
Hempstead Gravel Member
Manhasset "Ice-erosion" unconformity
Illinoian G Formation Montauk Till Member
"Ice-erosion" unconformity
Herod Gravel Member
Jacob Sand
Yarmouthian I Gardiners Clay
Kansan G Jameco Gravel
Aftonian I Unconformity surface of great erosion
Pre-Kansan G Mannetto Gravel
Table GG-1. Fuller's stratigraphic classification of the Pleistocene deposits of Long Island. G, glacial; I, interglacial. (Fuller, 1914, p. 20.)
The next-older unit in Fuller's classification is the Jacob Sand. Fuller took the name from Jacob Hill, "a high point on the north shore of Long Island, 8 miles northeast of Riverhead, near which the formation is well exposed" (Fuller, 1914, p. 107).
The Gardiners Clay grades upward into Jacob Sand which Fuller described as:
"In its most characteristic (sic) form the Jacob
sand consists of exceedingly fine sands, mainly quartz
flour, but with many grains of white mica and some of
dark-colored minerals. In color the sands commonly
range from a very light gray to yellowish (sic) and buff
tints, but where laminae of true clay are present they
may be stained reddish externally. They are everywhere
clearly stratified, although individual beds several
feet thick and appearing structureless to the eye are
encountered. When wet most of them are somewhat
plastic but lack the toughness of true clay; all are
decidedly gritty to the teeth and most of them to the
touch. Interbedded with the fine varieties of the
Jacob deposits are some more distinctly (sic) sandy
beds, usually buff or yellowish, and several feet
thick, in which particles of fairly fresh granitic
minerals can be recognized" (Fuller, 1914, p. 107).
Fuller (1914, p. 113) assigned the Jacob Sand to the Illinoian, but considered it to be transitional between the interglacial Yarmouthian Gardiners Clay below and overlying Illionian glacial materials of the Manhasset Formation.
"There is reason to believe that the change in
deposition, as was pointed out in the discussion of
the source of the material (p. 107), was caused by
the advent of glacial silts brought down from the
north during the advance of the Montauk ice, but
long before it invaded the region under discussion"
Fuller (1914, p. 113).
Fuller (1914, p. 92) gave the name Gardiners clay from "Gardiners Island, situated between the North and South flukes at the east end of Long Island, on which several clay beds with included sands are well exposed at a number of points."
"On western Long Island, where the formation
reaches its maximum development, the Gardiners clay
consists of irregular dark-colored beds alternating
with layers (sic) or lenses of sand and fine gravel
and attaining near Brooklyn an aggregate thickness
of 150 feet. In this region the clays, unlike those
in the localities farther east, grade downward through
glauconitic (sic) and locally fossiliferous sand into
the Jameco gravel, representing in fact transitional
deposits. The clays themselves consist of a very fine
silt, dark from the contained organic matter and carrying
more or less lignitized wood. The included sandy layers
are commonly from 5 to 10 feet thick and at some places
have yielded fossil remains (Fuller, 1914, p. 93).
"The great body of the Gardiners clay rests upon
the Jameco gravel, but along the borders of the Jameco
next to the Cretaceous land mass, especially along the
edges of the great depression in the vicinity of Jamaica
Bay, the clay laps up on the eroded surfaces of the
Cretaceous and Mannetto (fig. 57) or even upon the
metamorphic rocks (fig. 62), with sharp erosion (sic)
and overlap unconformities (sic) (Fuller, 1914, p. 94).
Figure GG-7. Fuller's sketch of exposure 1 mile west of Rocky Point, Montauk showing dipping diamictons and intercalated well-bedded strata. Based on what is exposed at Caumsett State Park, we infer that all these dipping layers are deltaic strata, the diamictons being products of subaqueous debris flows, not tills. (M. L. Fuller, 1914, fig. 156, p. 143.)
Figure GG-8. Fuller's sketch of exposure 0.5 mi S of Cullodan Point, Montauk, showing a dipping succession of diamictons and well-bedded strata and overlain by a horizontal till (f, at top; "Wisconsin till" in Fuller's caption. Other letter notations, after Fuller, are: a, "Montauk till member"; b, clayey sand; c, clay; d, gravel, and e, sand. (M. L. Fuller, 1914, fig. 157, p. 143.)
In all of the localities where the Gardiners Clay is visible at the surface, evidence of ice-thrust deformation is unmistakeable (Fuller, 1914, p. 96-102, figs. 65-86).
Fuller's two oldest units, the Jameco Gravel and Mannetto Gravel, are known mostly from wells. They form valley fills. The oldest valleys on Long Island, the pre-Mannetto valleys, were completely filled in by the Mannetto Gravel (Fuller, 1914, p. 44). The Jameco Valley cuts the Mannetto Gravel; it in turn has been filled in by the Jameco Gravel, which filled the valley and obliterated it as a landscape feature.
Fuller took the name Jameco Gravel "from the Jameco pumping station, near Jamaica South, 3 miles south of Jamaica, in western Long Island," where Veatch first recognized and named these deposits from deep wells.
"The Jameco gravel, although it has not been
definitely recognized at the surface at any point
on Long Island, has been encountered in a considerable
number of wells. In its type locality, in the area
extending from Jamaica Bay northward toward Whitestone,
it occupies a broad //(p.86) depression in the under-
lying rocks (either Cretaceous or Mannetto). It is
easily recognized in the wells in this locality because
of its striking dissimilarity to all other Pleistocene
beds (except the Montauk till member of the Manhasset
formation) and to the Cretaceous formations. The
difference between the Jameco gravel and the Mannetto
gravel is especially marked. Although the older beds
are prevailingly light-colored (sic) and composed
principally of quartz, the Jameco is generally a very
coarse dark-colored gravel containing a predominance of
granitic pebbles with a few streaks of black (sic) or
other dark sands or finer silts..."
"Where lithologic characteristics are not
determinative, the formation is recognized by
its position beneath the fossiliferous Gardiners
clay" (Fuller, 1914, p. 85-86).
"The Mannetto gravel was named from the Mannetto Hills (West Hills), on the crest of which just west of Melville some of the best exposures of this gravel on the island were found" (Fuller, 1914, p. 80).
Fuller described the Mannetto as follows:
It "consists of stratified (sic) and in some
places cross-bedded gravels composed mainly
of well-rounded pebbles of quartz from half
an inch to an inch in diameter mixed with
coarse yellowish quartz sand, but carrying
everywhere a few deeply weathered granitic
pebbles and scattered large bowlders of
crystalline rock, also deeply weathered or
disintegrated. It includes a few thin
intercalated beds of yellowish clay. The
granitic fragments can usually be crushed
by the finger or by a slight blow of a hammer,
and even the quartz is far more friable than
fresh fragments. The quartzose (sic) and stained
character (sic) of the gravels, the deep weathering
of the pebbles, and the complex flow and plunge (sic)
structure are the distinguishing features of
the formation " (Fuller, 1914, p.80).
We note a distinctive anomaly in the relationships Fuller described for the Jameco and contrasted with the Mannetto gravels. Both are prominent subsurface units in western Long Island where the upper unit, the Jameco, fills a major depression eroded in the Mannetto and is therefore the younger unit. Fuller reported that the Jameco is not known at the surface on Long Island. By contrast, the type locality of the Mannetto is on the surface in the Mannetto Hills.
The key feature that Fuller used to identify the Mannetto is the decayed granitic pebbles. Such decayed pebbles can originate in two ways. (1) They can be distinctive erratics, indicator stones, for example, of a region of pre-glacial decayed bedrock, such as is found in the northernmost 500 feet of the Garrison tunnel of New York City's Catksill aqueduct (Berkey and Rice, 1921, p. 101-103; Berkey and Fluhr, 1948); or (2) they can be the result of intensive in-situ chemical weathering after deposition and thus indicate great age.
Two indices of the first alternative are that (a) the pre-glacial (probably even pre-Late Cretaceous; Blank, 1978) decomposition of the feldspars was accompanied by the dissolution of quartz; and (b) other stones do not show comparably advanced states of decomposition.
The key feature of the second alternative is that all stones display the effects of advanced stages of decomposition. We have not checked the surface exposures that Fuller assigned to the Mannetto Gravel for dissolution of quartz, but have noticed a contrast in states of decomposition: only in the granitic rocks has the feldspar turned into clay. In other kinds of pebbles, only the effects of incipient decomposition are visible. Therefore, the decay of the feldspars should not be considered a badge of extremely ancient age (as we have previously supposed).
Given the validity of the first alternative, we raise the possibility that the Mannetto Gravel of the surface exposures may not be correlative with the subsurface unit Fuller assigned to the Mannetto Gravel in wells from western Long Island. This possibility would be highly likely if the degree of decomposition of the pebbles in the subsurface units assigned to the Mannetto is more or less uniform throughout in contrast to decay of granitic pebbles only in the surface exposures. Whatever is the outcome of this discussion about the Mannetto, the message is clear: decayed feldspars by themselves do not necessarily prove an early Pleistocene age.
Work Done After 1914: Most Return to the Single-glacier Hypothesis, But Multiple Glaciers Do Rear Their Ugly Heads Again
In the mid-1930's, W. L. S. Fleming (1935) and MacClintock and Richards (1936) published their analysis of the Pleistocene record. Both rejected several key age assignments in Fuller's stratigraphic classification. Fleming argued that the age of the Manhasset Formation is Wisconsinan; he invoked several Wisconsinan glacial advances to account for the Montauk Till and changed the age assignment of Long Island's famous terminal moraines from Early Wisconsinan (Fuller's interpretation) to Late Wisconsinan. MacClintock and Richards led the multitude back to the one-glacier view that Fuller had thought he had buried. Not only did they move Fuller's Manhasset Formation up into the Wisconsinan but also they shifted the Gardiners Clay from the Yarmouthian interglacial, where Fuller had placed it, into the Sangamonian (but MacClintock and Richards weaseled by allowing as how the age of this clay might be partly Sangamonian and partly Yarmouthian).
According to Sirkin:
"...the Gardiners Clay, was believed to represent
an Early Pleistocene interglacial (Fuller, 1914)
and was subsequently placed in the Sangamonian
Interglacial Stage (MacClintock and Richards, 1936).
In historical usage, a variety of fine-grained
sediments of both fresh water (sic) and marine
orgin have been called the Gardiners Clay. These
strata, which have been observed in surface exposures
and well sections, can vary considerably from the
original fossiliferous marine sediments of the type
section (Upson, 1968; Sirkin and Mills, 1975).
Gustavson (1976) has shown that certain so-called
Gardiners Clay units contain fossil faunas quite
unlike the fauna from the type section, while (sic)
Sirkin and Stuckenrath (1980) indicate that some
strata identified as Gardiners Clay could be of
Portwashingtonian age, particularly in the absence
of radiometric ages for either the original or the
presumably correlative units.
"The inclusion of such strata in the Woodfordian
moraines only show that they predate (sic) the
Woodfordian advance. As a surface deposit, the Manetto
Gravel, although well weathered, is probably Woodfordian
outwash (Sirkin, 1971), derived from deeply weathered
granite and granite gneiss in Connecticut. The Jameco
Gravel and the Gardiners Clay as recognized in well
section are undoubtedly post-Cretaceous and probably
represent Late Pleistocene deposits that are older
than the overlying glacial deposits" (Sirkin, 1982,
p. 38).
MacClintock and Richards experienced difficulty in recognizing Fuller's Montauk Till Member of the Manhasset Formation and did not accept Fuller's stream-dissection hypothesis for the origin of the north-facing valleys. According to Fuller, streams of post-Manhasset age eroded valleys into the Manhasset Formation. Because Fuller thought this interval of erosion correlated with the Sangamonian interglacial age, he assigned the pre-erosion Manhasset Formation to the Illinoian. By contrast, MacClintock and Richards argued that the valleys had not been stream eroded, but were left as depressions (somewhat analogous to giant kettles) because they had been occupied by tongues of glacial ice. While the ice tongue remained in what is now the valleys, thick bodies of outwash sand were aggraded in between. After the ice had melted, valleys appeared. Valleys having such an origin are not younger than the adjoining sands, but of the same age. We prefer Fuller's interpretation.
The theoretical background in support of the concept that one and the same continental ice sheet could display multiple flow directions was proposed at the time when the modern version of the Laurentide Ice Sheet was advocated (Flint, 1943). According to Flint, the Laurentide Ice Sheet began as one or more snowfields in the highlands of northeastern Canada. With continued additions of snow, an ice cap appeared and it began to spread southward and westward. The azimuth from the northeastern Canadian highlands to New York City is 195° or along a line from N15°E to S15°W. After this ice cap had become a full-fledged ice sheet and had attained something close to its full thickness, it is presumed to have itself become a factor in localizing where further snow would fall. Flint inferred that the ice sheet could divert the flow of moisture-bearing winds from the Gulf of Mexico and thus would have acted as a self-generating orographic source of precipitation. In other words, the ice sheet forced the air to rise and to be cooled and thus to drop its moisture. Enough snow is therefore thought to have been heaped up at various localities near the outer margin of the ice sheet and thus to have formed ice domes whose relief altered the direction of flow. Thus, the initial direction of regional rectilinear flow toward the SSW as a result of snow supply from northeastern Canada, could change locally to centers of quasi-radial flow under of the influence of the ice domes each of which could display divergent flow patterns, including zones of flow from NNW to SSE. During retreat, the above-described situation would be reversed. The factors responsible for radial ice-dome flow, including sectors from NNW toward the SSE, would cease to operate and those causing rectilinear flow toward the SSW to resume their former pre-eminence. The predicted pattern of flow for each advance of such an ice sheet, therefore, would involve three phases in the following order: (1) rectilinear flow from the NNE toward the SSW; (2) quasi-radial flow from the glacier-marginal ice domes, but locally from the NNW toward the SSE; and (3) rectilinear from the NNE toward the SSW.
Because he was convinced that if two glaciers had flowed over an area and both had extended their influence deep enough to polish and scratch the bedrock, then the younger glacier would tend to obliterate all traces of the older one, Flint opposed the multiple-glacier hypothesis. Accordingly, he argued that all the striae must have been made by only one glacier, the youngest one.
On the Glacial Geologic Map of North America, Flint (1945) mapped the two contrasting flow directions, one from the NNE to the SSW and the other from NW to SE. Figure GG-9 show examples from the upper midwest and from SE New England that demonstrate glacial flow from the NW to the SE.
The first post-Fuller challenge to the one-glacier-did-it-all school of thought came from one of the staunchest one-glacier partisans, Richard Foster Flint (1961). In south-central Connecticut, Flint found two tills in direct superposition. He gave the name Hamden Till to the upper till, whose flow indicators imply glacier movement from NNE to SSW. He proposed the name Lake Chamberlain Till for the lower till, whose flow indicators showed glacier movement from NNW to SSE.
Despite Flint's results, Sirkin (1968, 1971, 1977, 1982) has attached to his noteworthy paleobotanical contributions a strong adherence to the one-glacier interpretation. Because he disagreed with some of Fuller's correlations, Sirkin swept aside all Fuller's work. Because we think Wehmiller (in Ricketts, 1986) has destroyed one of the keystones of Sirkin's interpretation, namely the supposedly mid-Wisconsinan Portwashingtonian warm interval (the amino-acid-racemization results suggest an age of 200,000 yr for the shells that gave a radiocarbon age of about 40,000 yr BP), we accord Sirkin the same treatment that he applied to Fuller (1914) and for the same reasons. [For further discussion of the problem of contrasting ages between these two dating methods in specimens from Port Washington, Long Island, see Muller and Calkin (1993, p. 1841), and for the Pleistocene deposits at Sankaty Head, Nantucket Island, MA, see Oldale and others (1982).]
We think that Sirkin's Roslyn Till probably is the same till that Upham (1879) noticed in extreme western Long Island and also the "trap-rock"-bearing gray till at Corona mentioned by Woodworth (1901). Like Sirkin, we assign the Roslyn Till to the Woodfordian (our Till IV); unlike Sirkin, we interpret the Roslyn Till as being younger than the Harbor Hill Moraine (which we assign to our Till III and place in the Early Wisconsinan).
In his study of the Quaternary sediments in the Boston area, Kaye (1982) found deposits that he ascribed to several Wisconsinan glaciers. Kaye's inferred flow directions of these Boston glaciers are virtually identical to those that we infer for the ancient glaciers in the New York City region. To quote from Kaye's paper:
"The direction of ice flow in the Boston Basin
and adjoining uplands was studied by means of the
orientation of striations (sic) and grooves on the
bedrock surface, the orientation (sic) of the long
axes of drumlins, the direction of transport of
erratics in till, and the direction of thrusting and
overturning of bedding in glacially deformed drift.
These data range through 360 degrees in azimuth.
Analysis of this confusing message shows the existence
not of an ever-shifting ice current but of at least
four separate and distinct ice currents of different
ages. Three of these flowed fairly rectilinearly, but
one (the last) was multicomponent and marked by strong
lobation" (1982, p. 31).
Figure GG-9. Sketch maps showing other regions in the United States where glacier flow was from NW to SE.
A. Swarm of drumlins south of Charlevoix, MI. (Frank Leverett, and F. B. Taylor, 1915, p. 311; redrawn by L. D. Leet and Sheldon Judson, 1965, fig. 13-20, p. 188.)
B. Boulder trains in New England, all products of regional glacier flow from NW to SE. (J. W. Goldthwait, in R. F. Flint, 1945; redrawn by L. D. Leet and Sheldon Judson, 1965, fig. 13-22, p. 190.)
Kaye numbered these tills from I (oldest) to IV (youngest). Tills I, II, and III flowed from the NW to the SE, with means as follows:
Till Mean flow direction
III S31°E, +/- 02°
II S64°E, +/- 18°
I S23°E, +/- 01°
Recently, Sirkin has moderated his one-glacier viewpoint. In an open-pit mine at Sanford Hill, in the central Adirondacks, Muller, Sirkin, and Craft (1993) described two tills separated by 3.6 m of brown Tahawus lake- or pond clay containing wood fragments older than 55,000 radiocarbon years B. P. that were exposed in 1963 in the National Lead Company's (now NL Industries) open-pit mine. They assigned the Tahawus Clay to the Sangamonian; according to them, it contains "an interglacial pollen record, the first one identified in northeastern New York" Muller, Sirkin, and Craft, 1993, p. 163).
In their summary of the glacial events in New York State, Muller and Calkin (1993) did not recognize any pre-Wisconsinan tills in the New York City region. They wrote: "The pre-Wisconsinan record involves saprolith and till in the Adirondack Mountains, marine clay on Long Island, multiple tills at Fernbank, Otto, and Gowanda, and major drainage derangement of the Allegheny River" (Muller and Calkin, 1993, p. 1829).
After summarizing the amino-acid-racemization results, they continued:
"...it is difficult to escape the conclusion
that two temporally distinct stratigraphic units are
involved. Indeed, Stone and Borns (1986) propose
that the name Gardiners Clay be reserved for brown
marine clay and silt (sic) with interbedded sand and
gravel of probable Sangamonian to Eowisconsinan age
(Table 3). This action clarifies the age of the
Gardiners Clay, and implicitly acknowledges the
probability that marine clays of both Sangamonian
and pre-Sangamonian age are present" (Muller and
Calkin, 1993, p. 1830).
OUR WORK
We subdivide this section into two parts: (1) a narrative of our background and credentials for investigating the glacial deposits in the New York City region, and (2) a summary of our interpretation of the stratigraphic framework of the Pleistocene deposits in the New York City reigon.
Background and Credentials of Investigators
In 1949-50 Sanders (hereafter abbreviated JES) took the late Professor Richard Foster Flint's graduate course in Geomorphology and Glacial Geology and for 10 years (1954-1964) was a faculty colleague of Flint's at Yale University. In the late 1950's and early 1960's, JES mapped the bedrock geology of several 7.5-minute quadrangles in and around New Haven at the same time as Flint was mapping the surficial deposits. They undertook several joint field conferences during which Flint interested JES in looking for and recording the orientations of features on bedrock surfaces that are valuable in determining the direction of flow of a glacier. JES found this a logical extension of his interest in the features made in sediments that enable directions of paleocurrents to be inferred, a topic he began to study during 1954 while a postdoctoral fellow working in Europe with the late Professor Ph. H. Kuenen, of the University of Groningen, The Netherlands (Kuenen and Sanders, 1956).
In 1964 JES moved to New York. He continued to work closely with Flint as a co-author of the Longwell, Flint, and Sanders Physical Geology textbook published in 1968. In 1968, JES began teaching the introductory geology course in the Department of Geology at Barnard College, Columbia University. Guided by Ina Alterman, a City College graduate then a graduate teaching assistant in the Department of Geology, Columbia University, JES started to examine local features at Fort Tryon Park (Figure GG-10) and the Palisades, places where many generations of geology students had been taken on required half-day field trips. One of the first things that JES noticed was the prominent evidence at both localities of striae and grooves made by a glacier that had flowed from about N15°W to S15°E. In search of what the experts on Pleistocene geology had made of such evidence, JES read into the geologic literature on the Pleistocene geology of the New York City region. In the writings of H. B. Kümmel (1933) and of R. D. Salisbury (1902, 1908; Salisbury and others, 1902), JES found how Salisbury, one of America's foremost specialists on Pleistocene geology, had explained this evidence of glacial flow from NW to SE. (See Figures GG-4 and GG-5.) JES thought Salisbury's explanation, which had been accepted by Kümmel and many other geologists, was somewhat unusual, but initially found no reason to challenge it.