INFORMATION TO USERS

This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted.

The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction.

1. The sign or “target" for pages apparently lacking from the document photographed is “Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity.

2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame.

3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary/sectioning is continued again — beginning below the first row and continuing on until complete.

4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced.

5. PLEASE NOTE: Some pages may have indistinct print. Filmed as received.

Xerox University Microfilms 300 North Zeeb Road Ann Arbor, Michigan 48106 76-3576 THOMPSON, Woodrow Burr, 1946- THE QUATERNARY GEOLOGY OF THE DANBURY-NEW MILFORD AREA, CONNECTICUT. The Ohio State University, Ph.D., 1975 Geology

Xerox University Microfilms , Ann Arbor, Michigan 48106 THE QUATERNARY GEOLOGY OF THE

DANBURY-NEW MILFORD AREA,

CONNECTICUT

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

By

Woodrow Burr Thompson, B.A., M.S

******

The Ohio State University

19 75

Reading Committee: Approved By

Sidney E. White Richard P. Goldthwait Charles E. Corbato iser

Department of Geology and Mineralogy Frontispiece. Southward aerial view of Housatonic Gorge (center) and junction of Still River with Housatonic River. Photo by J. A. Pawloski. PREFACE

This report is based on field work done by the author

during the summers of 19 70 through 19 74. The work was funded

by the United States Geological Survey during the first

three summers and by the Connecticut State Geological and

Natural History Survey during the final summer. It was part

of a co-operative mapping program involving the state and

federal geological surveys. The geologic maps and much of

the other information in the present study will be published by the U.S.G.S. as part of the Geologic Quadrangle Map

series.

The author gratefully acknowledges the U.S.G.S. for

providing base maps and other essential materials for the

preparation of this report. Fred Pessl, Jr. and

William Langer of the U.S.G.S. visited the author in the

field and expressed helpful opinions concerning mapping

problems. Robert Melvin of the Water Resources Division

of the U.S.G.S. supplied bedrock contour and subsurface data

that were important in interpreting the Quaternary

stratigraphy of the study area.

The writer is especially indebted to Dr. Sidney E. White

for his help as faculty adviser and for reviewing this

iii manuscript. Editorial comments were also contributed by

Dr. Richard P. Goldthwait and Dr. Charles E. Corbato.

Dr. White and Dr. Goldthwait visited the study area and ex­ amined the Quaternary deposits that are discussed here.

Where possible, the geographic and geologic features mentioned in this report are located with respect to places that are readily identifiable on Plate I. However, it is occasionally necessary to use a grid system to refer the reader to a certain part of the map. The Danbury and

New Milford quadrangles (Plate I) are divided into ninths, each of which covers 2.5 minutes of latitude and longitude.

These divisions are indicated on Plate I by printed lines on the map border and by crosses that designate their corner pointsc The author has assigned letter symbols to the divisions in the following manner:

Danbury quadrangle New Milford quadrangle

NW NC NE NW NCNE

WC C EC wc CEC

SW SC SE sw SC SE

Localities that may be hard to find on Plate I are referred to one of the above divisions. This grid system is also used to help locate the sources of samples that are listed in Appendix B.

iv All grain-size measurements in this report are in metric units. Metric equivalents are given in parentheses for most other measurements except topographic map elevations.

Preference is commonly given to English measurements because of their usage in previous reports — particularly in pub­ lished well and test hole logs.

v VITA

December 10, 1946 Born - Plymouth, New Hampshire

1968...... B.A. Geology, Dartmouth College, Hanover, New Hampsire

1969-1971...... Teaching Assistant, Department of Geology, The University of Vermont, Burlington, Vermont

1971-1975...... Teaching Associate, Department of Geology and Mineralogy, The Ohio State University, Columbus, Ohio

PUBLICATIONS

Stone,.B.D., and Thompson, W.B., 19 71, Late Wisconsin . glacial readvance deposits on Isle La Motte, Vermont (abs.): Geol. Soc. America Abstracts with Programs, .v. 3, p. 51.

Thompson, W.B., 1971, Rodingite near Thetford Mines, Quebec: Mineralogical Record, v. 2, p. 45-46, 48.

Thompson, W.B., 1973, Theories on the origin of pegmatite — a summary: Earth Science, v. 26, p. 77-78.

Thompson, W.B., 1974, The Palermo Mine, New Hampshire: Mineralogical Record, v. 5, p. 2 74-279.

FIELDS OF STUDY

Major Field: Geology

Studies in Glacial Geology. Professor Richard P. Goldthwait

Studies in Geomorphology. Professor Sidney E, White

vi TABLE OP CONTENTS

Page

PREFACE...... iii

VITA...... vi

LIST OF TABLES...... ix

LIST OF ILLUSTRATIONS...... x

Chapter

I. INTRODUCTION...... 1

Purpose and scope...... 1 Means of investigation...... 1

II. DESCRIPTION OF THE STUDY AREA ...... 5

Geography...... 5 Geomorphology...... 7 Bedrock geology...... 10 Surficial geology...... 13 Previous work...... 16

III. TILL DEPOSITS...... 19

Introduction...... 19 Upper till...... 19 Lower till...... 34 Ages of the upper and lower tills...... 41

IV. GLACIAL-MELTWATER DEPOSITS...... '...... 52

Basis for delineation and correlation of map units...... 52 Meltwater-stream deposits...... 57 Glacial Lake Danbury deposits...... 6 3 Saugatuck River stage deposits...... 64 Pond Brook stage deposits...... 68 Pumpkin Hill stage deposits...... 73 Housatonic River stage deposits...... 75 Glacial Lake Kenosia deposits...... 79 Glacial Lake Candlewood deposits...... 80 vii V. LATE-GLACIAL AND POSTGLACIAL DEPOSITS...... 81

Eolian deposits...... 81 Stream-terrace deposits...... 84 Flood-plain deposits...... 85 Swamp deposits...... 85 Colluvium...... 87

VI. GLACIAL HISTORY...... 89

Pre-Wisconsinan events...... 89 Early Wisconsinan glaciation...... 91 Late Wisconsinan glaciation...... 92 Recession of the Late Wisconsinan glacier...... 96 Deposition of stratified drift during deglaciation...... 9 7

VII. LATE-GLACIAL AND POSTGLACIAL HISTORY...... 113

Late-glacial events...... 113 Postglacial events...... 118

VIII. SUMMARY...... 123

APPENDIX

A . LOGS OF WELLS AND TEST HOLES PLOTTED ON PLATE I ...... '...... 127

B. SAMPLE LOCATIONS...... ■...... 130

C. X-RAY DIFFRACTION DATA FOR CLAY FRACTIONS OF TILL SAMPLES...... 132

REFERENCES CITED...... 135

viii LIST OF TABLES

Table Page 1 Grain-Size Analyses of Upper Till Samples...... 25

2 Gasometric Determinations of Calcite and Dolomite Weight Percentages in Till Samples and Till Clasts...... 33

3 Grain-Size Analyses of Lower Till Samples...... 37 LIST OF ILLUSTRATIONS

Frontispiece Page

Southward aerial view of Housatonic Gorge and junction of Still River with Housatonic River...... i

Figure 1 Location of study area...... 6

2 Outcrop of rottenstone marble between U. S. Route 7 and Grays Bridge Road...... 12

3 Ablation facies of upper till at New Fairfield two-till locality...... 22

4 Grain-size distributions of till samples...... 24

5 X-ray diffraction pattern of clay fraction of upper till sample NW-9...... 2 8

6 Upper till on east side of Great Brook valley... 32

7 Vertical pit face in weakly oxidized lower till...... 36

8 X-ray diffraction pattern of clay fraction of lower till sample NW-12...... 40

9 Rose diagram showing long-axis azimuths of stones in upper till at New Fairfield two-till locality...... 45

10 Rose diagram showing long-axis azimuths of stones in lower till at New Fairfield two-till locality...... 46

11 Vertical pit face showing contact between upper till and lower till, east of New Fairfield 4 8

x Figure Pa9e 12 Gravel outwash overlying rhythmically- bedded lake-bottom sediments west of y. S. Route 7 at Boardman Bridge...... 62

13 Delta from Saugatuck River stage of glacial Lake Danbury, Germantown district of Danbury...... 66

14 Kame delta from Pond Brook stage of glacial Lake Danbury, west of Brookfield village...... 70

15 Foreset sand beds in kame delta west of Brookfield village...... 70

16 Collapsed boulder gravel in kame on west side of Still River valley...... 72

17 Flat-bedded glaciolacustrine silt overlying folded and faulted sand beds deposited in an ice-contact environment...... 72

18 Northward aerial view of Housatonic Gorge..... 74

19 Delta at Lanesville from Pumpkin Hill stage of glacial Lake Danbury...... 74

20 Collapsed gravel beds in kame terrace on west side of Housatonic River valley...... 76

21 Eolian sand and silt overlying topset beds of the Lanesville glacial lake delta...... 82

22 Rose diagram showing azimuths of bedrock striations and drumlins 9 3

2 3 Roche moutonnee developed on Inwood Marble northwest of Brookfield village...... 95

2 4 Maximum extent of glacial Lake Danbury during the Saugatuck River stage...... 99

25 Maximum extent of glacial Lake Danbury during the Pond Brook stage...... 10 3

26 Extent of glacial Lake Danbury during the Pumpkin Hill stage ...... 106

xi Figure Page 2 7. Maximum extent of glacial Lake Danbury during the Housatonic River stage...... 10 8

2 8 Kame complex in site of former river channel on west side of Still River valley...... 117

29 Weathering pits on Inwood Marble outcrop on east side of U. S. Route 7 ...... 121

Plate

I Surficial geologic map of the Danbury and New Milford quadrangles , Connecticut...... pocket

II Profiles of glaciofluvial, stream-terrace, and flood-plain deposits in the Danbury and New Milford quadrangles...... pocket

III Map of bedrock outcrops, thin drift, and bedrock contours in the Danbury and New Milford quadrangles...... pocket

IV Explanation to accompany Plates I, II, and III..pocket CHAPTER I

INTRODUCTION

Purpose and scope

The purpose of this report is to describe and interpret the Quaternary geology of the Danbury-New Milford area in western Connecticut. The author mapped the glacial and post­ glacial deposits of this region, and they are described in detail in the first part of this study. The Quaternary history of the study area is then reconstructed and cor­ related with other historical studies from Connecticut and neighboring states. Soils were given only brief attention in this investigation because they have been examined in the area by. Gonick and others (1970).

Means of investigation

Much of this report is based on field mapping of the surficial geology of the Danbury and New Milford 7.5-minute quadrangles. The author examined borrow pits and building excavations in till and water-laid sediments. This aspect of the field work was facilitated by rapid urbanization of the study area. Stream-bank exposures provided additional information, but in many cases it was necessary to identify surficial deposits by means of shovel or auger holes. Subsurface data were obtained from well and test hole logs, most of which have been published by the Connecticut

Water Resources Commission (Melvin, 19 70). This information was useful for determining thicknesses of map units and their stratigraphic relationships. Some of the most signif­ icant borehole sites are plotted on the surficial geologic map of the study area (Plate I). Appendix A contains graphic logs for these holes.

The author used map and aerial photograph interpretation to supplement the field work and subsurface data. Topo­ graphic maps and low-altitude aerial photographs provided a regional perspective for evaluating ground observations and locating features that were not readily visible in the field.- Surficial deposits wbre classified into chronologic mapping units, which were then plotted on scale-stable mylar copies of the Danbury and New Milford quadrangles. These maps are reproduced on Plate I, and the geologic symbols are explained on Plate IV. A bedrock map was available for the

Danbury quadrangle (Clarke, 19 58) and was used to determine possible sources of till and outwash. Robert L. Melvin of the United States Geological Survey provided the author with unpublished bedrock contour maps for lowland areas in both quadrangles. Stream drainage changes and effects of glacial erosion were inferred from these bedrock contours, shown

(with slight modification^ on Plate III. Plate III also shows individual bedrock outcrops and areas of numerous closely spaced outcrops.

Samples of glacial till and eolian sand were collected from the study area and processed in the Quaternary

Laboratory of the Department of Geology and Mineralogy at

The Ohio State University. Weight percentages of granules,

sand, silt, and clay were determined for all samples. The sand fractions were separated by sieving. Sediment that passed a No. 2 30 sieve was then separated into silt and clay

fractions by the hydrometer method. In accordance with the

United States Department of Agriculture (19 51) , Grim (1968) , and Carroll (1970), a 2-micron size limit was selected as the boundary between silt and clay. This limit approximates the natural boundary between clay minerals and minerals that occur in coarser size fractions.

While the clay fractions from the tills were still in aqueous suspension, they were prepared for identification by x-ray diffraction. A vacuum pump apparatus was used to

form clay coatings on porous ceramic plates. Three plates were prepared for each sample. Two of them were then air- dried, while the third was kept moistened with ethylene

glycol. All three plates were x-rayed, and the air-dried plates were x-rayed a second time after being heated to

450° or 550°C for two hours. This treatment modifies certain

clay mineral crystal structures and enables distinction of x-ray diffraction peaks that would otherwise be coincident

(Carroll, 19 70).

Till samples that contained carbonate minerals were analyzed for calcite and dolomite percentages. This was accomplished by gasometric determination in a Chittick apparatus of the type described by Dreimanis (1962). The author made six simultaneous determinations on each sample and averaged the results to get representative values. In order to avoid polygranular marble fragments, only the part of the till matrix that passed a No. 60 sieve was used in the Chittick apparatus. This size fraction was prepared and run in the manner described by Dreimanis. CHAPTER II

DESCRIPTION OF THE STUDY AREA

Geography

The study area for this report includes the Danbury and New Milford 7.5-minute quadrangles in western Connecticut

(Figure 1). The New Milford quadrangle adjoins the north edge of the Danbury quadrangle, and together they encompass

112 square miles (292 km ). This area ranges in latitude from 41°22I30"N to 41o37'30"N, and in longitude from

73o22'30"W to 73°30'W. The Danbury quadrangle lies almost entirely within Fairfield County, while the New Milford quad­ rangle includes parts of Fairfield and Litchfield Counties.

The study area is heavily settled, with the highest population density in the cities of Danbury and New Milford.

The climate of the Danbury-New Milford area is humid and continental. Average daily temperatures vary from a mean of about 25°F (-4°C) in January and February to 70°F

(21°C) in July. Annual precipitation averages about 46 inches

(117 cm), and it is usually heaviest in June and July

(Gonick and others, 1970). Wind directions are highly vari­ able, but they are most commonly from the northwest or southwest.

5 6

Ne Iford

W a te rb u ry Co

Donbyiry

B rid g e p o r t

m iles

XXECX3Q kilom eters

Figure 1. Location of study area. 7

The natural vegetation of the study area (on uncleared land) is a mixed forest that contains both deciduous and evergreen trees. Differences in vegetation are not suffi­ ciently great to have caused major variation in soil types

(Gonick and others, 1970). However, it is common for some plants to grow preferentially on certain soils, which in turn have developed on specific parent materials. Juniper and pine trees, for example, thrive on sandy, well-drained glacial kames.

Ge omo rph o1ogy

The Danbury and New Milford quadrangles lie within the New England Upland part of the New England physiographic province (Thornbury, 1965) . The topography of the area is very hilly. Land elevations range from 150 feet along the

Housatonic River (in an area that is now artificially flooded) to over 1200 feet in the southwest corner of the New Milford quadrangle. The terrain is knobby and irregular where con­ trolled by bedrock outcrops, but many hills are smoothed by thick accumulations of till. Glaciation has caused a strong northwest-southeast topographic lineation, especially in the Danbury quadrangle. This is the combined result of bedrock erosion and drumlin orientation. Slopes are commonly steepest on the southeast sides of bedrock hills, where

glacial plucking has occurred.

The Housatonic River is the largest river in western

Connecticut and one of the most important streams in.the study area. It flows southeast across the full length of the New Milford quadrangle and the northeast corner of the

Danbury quadrangle. Its valley is generally straight and « closely confined by the till and bedrock upland. While many of its tributaries are subsequent streams that follow struc­ tural or lithologic weaknesses in the bedrock, the Housatonic

River has cut across ridges of resistant schist and gneiss,

Two miles (3,2 km) south of New Milford, for example, the river passes through a deep bedrock gorge (henceforth called the "Housatonic Gorge"). In this same area, tributary streams follow the paths of least resistance and enter the river from the north or south. Previous geomorphic interpre­ tations of the Housatonic River's course are summarized later in this chapter.

The Still River, a tributary of the Housatonic River, is the other major stream in the study area. Most of its drainage basin lies within the Danbury and New Milford quadrangles. It originates from several brooks, ponds, and reservoirs within 3 miles (5 km) of the New York border.

The lowest point on the divide between the Still River basin and the Croton River basin to the west is on the state border, at an elevation of about 475 feet. After entering the southwest corner of the Danbury quadrangle, the Still

River flows east through the city of Danbury, and then north along a meandering path to the Housatonic River. It joins the latter stream just past Lanesville and north of the Housatonic Gorge. The gradient of the Still River averages

11 feet per mile over its 2 3-mile (37 km) course from the village of Mill Plain (Brewster quadrangle) to the

Housatonic River. At Brookfield village, however, the flat topography of the valley floor is interrupted by a bedrock high (NE; Plate III). Here the Still River flows on bedrock and loses over 30 feet of elevation in 0,5 mile (0.8 km).

A gap occurs at an elevation of about 380 feet on the divide between the Still River and Housatonic River basins.

It is located in the east-central part of the Danbury quad­ rangle, Another gap at about 410 feet occurs 2.5 miles

(4 km) south of the quadrangle, on the divide between the

Still River and Saugatuck River basins. It will be seen that

these two gaps played an important role in controlling melt- water drainage during deglaciation of the study area.

The Rocky River (C) was formerly an important tributary of the Housatonic River. It flowed south from near the village of Sherman (WC), doubled back around Vaughns Neck

(SC-NC), and flowed north to the Housatonic River, The

Rocky River was artifically dammed near its mouth, and much of its drainage basin was flooded to form Lake Candlewood.

This lake is drained through an aqueduct at its north end to generate hydroelectric power. A short segment of the

Rocky River still exists and joins the Housatonic River

1 mile (1.6 km) northwest of New Mi, 1 ford. 1° Most of the Danbury-New Milford region is well-drained by the many small streams that are tributary to the Still

River and Housatonic River, but swamps and other poorly drained areas are common. Small upland swamps occupy de­ pressions in-the bedrock surface. Other swamps are the re­ sult of drainage derangements caused by deposits of till and water-laid sediments. The latter variety includes the largest swamps in the study area, such as the one in the Limekiln

Brook valley (SE).

Bedrock geology

The bedrock of the Danbury-New Milford area includes igneous and metamorphic formations of the Precambrian and

Paleozoic Eras. Clarke (1958) mapped the bedrock geology of the Danbury quadrangle, and he distinguished a "Precam­ brian Highlands" region in the western third of the quad­ rangle. This area is underlain by metamorphosed, hetero­ geneous igneous and sedimentary rocks, which have been intruded by the Danbury Augen Granite. The remainder of the quadrangle is underlain by Paleozoic igneous, metasedimentary, and metavolcanic rocks. The metamorphic formations include the Fordham Gneiss, Inwood Marble, Manhattan Formation, and

Hartland Formation, These units have been intruded by the

Brookfield Plutonic Series and a "younger granite" (Clarke,

195 8), Hall (196 8) assigned a Precambrian age to the

Fordham Gneiss in the vicinity of White Plains, New York,

The Inwood Marble unconformably overlies the Fordham Gneiss 11 in this area, and Hall gave the age of the marble as

Cambrian to Lower Ordovician. The ages of the other

Phanerozoic formations recognized by Clarke are uncertain.

The rock formations of the Danbury quadrangle extend northward along strike into the New Milford quadrangle. The bedrock geology of the latter quadrangle was being mapped at the time of this investigation, so a published report was not yet available. The present author recognized two of

Clarke's formations -- the Inwood Marble and Danbury Augen

Granite — in the course of mapping the locations of glacial boulders in the New Milford quadrangle. Other boulders could only be designated in terms of their compositions (Plate IV).

Most rock formations in the study area have fresh, solid outcrops due to scouring and plucking by Pleistocene glaciers. However, the Inwood Marble is an important ex­ ception. The Inwood Marble occurs throughout the length of the Danbury and New Milford quadrangles. It underlies the valleys of Sympaug Brook and Limekiln Brook east of Danbury.

From Danbury northward, the marble belt underlies the entire

Still River valley, its topographic extension north of

Lanesville, and much of the Housatonic, East Aspetuck, and

Great Brook valleys (NC-NE). Many outcrops of the Inwood

Marble have a weathered zone in which the rock has disinte­ grated to such a degree that it is easily excavated with a shovel (Figure 2). The rottenstone grades into solid rock at variable depths. Test hole data indicate that the weathered Figure 2. Outcrop of rottenstone marble between U.S. Route 7 and Grays Bridge Road (EC). 13 zone is commonly over 10 feet (3 m) thick (Melvin, 19 70), and that it is as much as 100 feet (30 m) thick in at least one place (R. L. Melvin, 1970, personal communication). The dif­ ferences in thickness are partly the result of uneven glacial erosion of preglacially weathered bedrock. Unweathered mar­ ble outcrops probably owe much of their durability to their mineralogic composition. The Inwood Marble has several facies containing highly variable percentages of calcite, dolomite, and silicates (Moore, 19 35; Clarke, 1958; Hall, 196 8;

Priicha and others, 196 8).

Deeply disintegrated outcrops of other rock types occur in the study area, but surface exposures are rare. However, rotten schist, gneiss, and granite are frequently encountered in test holes drilled in valley areas (Melvin, 1970). These rottenstone occurrences are buried under glacial-meltwater deposits and till. They may have been partially protected from glacial erosion by their low topographic positions.

Surficial geology

Till deposits from Pleistocene glaciation are the stratigraphically lowest and oldest surficial materials in the Danbury-New Milford area. Two principal tills are present here — a compact, clay-rich, and deeply oxidized, lower till, and a friable, coarser-grained, and generally unoxidized upper till. The lower till is found in nearly all drumlins and other streamlined hills, where its thickness is generally greater than 40 feet (12 m). The upper till has 14 an average thickness of about 10 feet (3 m) . With the ex­ ception of drumlins, till deposits are thickest on hillside areas. They become thinner at high elevations, where bedrock outcrops are common; and they are believed to be thinner on valley floors. Measurement of till thickness on valley bottoms is hindered by the uncertainties of interpreting well and test hole logs.

Glacial-meltwater deposits are abundant in several stream valleys in the Danbury and New Milford quadrangles.

Most of them occur in the Still River and Housatonic River valleys and a few of their tributaries. Large deposits probably exist in the Lake Candlewood valley, but only a few scattered patches of sand and gravel escaped flooding by the lake. Meltwater deposits are uncommon in the upland areas.

The only large upland outwash accumulations lie between the drum-lins in the southeast part of the Danbury quadrangle.

Both glaciofluvial and glaciolacustrine sediments are common in the study area. Generally fine-grained lake-bottom deposits were laid down in the lowland that extends from

Bethel to New Milford and includes parts of the Still River,

Housatonic River, and East Aspetuck River valleys. Sediments of glaciolacustrine origin are also known to underlie Lake

Candlewood (Harvey, 1920). Many of the glaciofluvial sand and gravel deposits were graded to the glacial lakes that occupied these valleys. Other fluvial deposits were graded 15 to stream valleys or temporary base levels controlled by the distribution of stagnant ice.

A thin layer of eolian sand and silt commonly overlies both till and glacial-meltwater deposits in the Danbury and

New Milford quadrangles. It is a few inches to a foot thick in most places. It does not occur on flood-plain alluvium, nor has it been observed on stream terraces. Therefore, this eolian mantle is believed to have formed prior to the estab­ lishment of a vegetative cover in early postglacial time.

Stream terraces occur along the Still River and

Housatonic River, as well as along a few of the lesser streams.

These terraces are erosional surfaces that were cut in glacial- meltwater deposits, but they are commonly capped by alluvium.

More than one terrace level may be distinguished locally a- long the major rivers. However, they usually cannot be traced very far, so they are shown as a single map unit on

Plate I.

Recent flood-plain deposits occur along all of the rivers and some of the brooks in the study area. Flood-plain sediments are dark colored because of a high content of organic material. Although erosion is the dominant process along upland brooks, small deposits of coarse gravel occur along the lower parts of their courses. Only a few such deposits are large enough to be shown on Plate I. Flood- plain alluvium is one of the youngest stratigraphic units 16 in the region (along with swamp deposits) and is still being deposited.

Other recent surficial materials in the Danbury-New

Milford region are swamp deposits and colluvium. The swamp alluvium is a mucky, heterogeneous mixture of sediment and organic debris. It is currently accumulating in poorly drained areas. Colluvial deposits include talus at the bot­ tom of cliffs and a surface layer of till-derived material on steep hillsides. Both types of colluvium are the result of late-glacial and postglacial mechanical weathering and mass movement processes.

Previous work

All previous geomorphic studies of the Danbury-New

Milford area were concerned at least partly with the regional drainage history. The Housatonic Gorge, the underfit appear­ ance of the Still River, and the low gaps on drainage divides near Danbury have all caused speculation about possible drainage changes during or before glaciation. Hobbs (1901),

Harvey (1920), and Hokans (1952) claimed that the ancestral

Housatonic River originated as a consequent stream on a

Cretaceous peneplain or coastal plain surface. Uplift of this gently sloping plain supposedly occurred at the end of the

Cretaceous Period and/or in Tertiary time.

To account for the width of the Still River valley,

Hobbs claimed that the Housatonic River formerly occupied it and flowed across either the Still-Croton or Still-Saugatuck 18

Housatonic Gorge. Hokans claimed that a marine estuary ex­ tended far up the Housatonic River valley during deglaciation, and the drainage of its waters was another agent that was re­ sponsible for cutting the gorge,

Thompson (19 71) studied the drainage and glacial history of the Still River valley and its northward topographic ex­ tension. He evaluated the ideas of previous workers in light of more detailed surficial geologic mapping and newly availa­ ble bedrock contour and subsurface data. He concluded that the preglacial Still River and Housatonic River followed their present courses. The anomalous width of the Still

River valley was attributed to relatively easy fluvial and glacial erosion of the Inwood Marble. Thompson discredited

Hokans' marine estuary theory, which had rested in part on an assumed postglacial land uplift of 15 feet per mile and the uneven rebound of crustal blocks along hinge lines. In addition, the stages of glacial Lake Danbury were redefined and renamed.

It was previously mentioned that Clarke (19 5 8) has mapped the bedrock geology of the Danbury quadrangle. The

Connecticut Water Resources Commission has published a re­ port on the hydrogeology of the study area (Cervione and others, 19 72) and a bulletin containing detailed well and test hole logs (Melvin, 1970). CHAPTER III

TILL DEPOSITS

Introduction

There are two principal tills in the Danbury and

New Milford quadrangles. They differ with respect to color, texture, structure, and degree of oxidation. They are here called the "upper till" and "lower till" because of their stratigraphic relationship to each other. Two similar tills have been discovered at many other localities in southern and central New England, and their stratigraphic positions are always the same where they occur together. However, type localities and widely used names have not been estab­ lished for these tills.

Upper till

The upper till is a thin, patchy unit in areas of rough, bedrock-controlled upland topography. It may occur on the flanks of drumlins, but it is probably not the major con­ stituent of any of them. Water-laid deposits generally cover the upper till in the bottoms of large stream valleys, but the till is often exposed in excavations along valley walls.

The average thickness of the upper till is about 10 feet

(3 m). Apparent thicknesses of over 20 feet (6 m) have been 19 20 observed, but they are rare. The till is absent over many upland ledges, or so thin that it extends no deeper than the bottom of the soil profile.

The moist color of the upper till varies from light olive brown through olive to light olive gray

(2.5Y 5/4 - 5Y 5/3 - 5Y 6/2 on the Munsell Soil Color Charts

(Munsell Color Company, 1954)). The till is mostly fresh and unoxidized. Oxidation has occurred at a few localities, but it has seldom penetrated deeper than 3 feet (1 m) (not counting the usual soil thickness).

The upper till is sandy, structureless, and very fria­ ble. It can usually be excavated with ease by construction equipment, so borrow pits are common in this till. Stones of pebble to boulder size are abundant and typically form

10 to 25 percent of the till. Stone lithologies strongly reflect the local rock types, but the variability of the bed­ rock and absence of outcrops in the vicinity of many till pits often prohibit a more precise correlation. Clasts of sedimentary rock are absent in the upper till, and definite erratics are uncommon. Some well-rounded quartzites and low- grade metamorphic rocks were probably derived from areas far to the north. The till stones are mostly solid and un­ weathered, but many of them have surfaces that appear freshly broken. Only a few are striated, though it should be noted that the bedrock of the study area is coarse grained and not readily striated. Silt coatings occur on the upper and lower 21 sides of many of the stones. These coatings consist of fine till-matrix sediment that has been redistributed by ground­ water movement.

It is possible locally to distinguish basal and ablation facies in the upper till. The basal facies was deposited by accretion beneath a moving glacier. It may be very com­ pact, and it commonly contains inclusions of the lower till.

At localities where the two tills are in superposition, slabs of lower till appear to have been ripped up (presumably while frozen) and included in the lower part of the upper till. The tills are so thoroughly mixed in some places that it is hard to distinguish them. The author observed a simi­ lar structural relationship between basal upper till and schistose bedrock at a pit on the east side of the Still River valley, between Vail Road and Stony Hill Road (EC). Slabs of schist along the till-bedrock contact were lodged as they were being wedged up into a compact mixture of the upper and lower tills. Thin lenses of washed sediment may also occur in the basal facies of the upper till.

The ablation facies of the upper till is-loose and coarse grained. Large boulders are usually present, as well as lenses of washed sand, silt, and gravel (Figure 3).

Meltwater has locally winnowed out the fine fraction of the till, leaving pods of openwork gravel. Inclusions of lower till are scarce or absent in the ablation facies. The above evidence leads to the conclusion that this facies was 22

Figure 3. Ablation facies of upper till at New Fairfield two-till locality (NW). Note openwork gravel (above shovel) and beds of washed sand (left of shovel). 23

deposited in an englacial or superglacial environment where

the effects of meltwater were great.

The grain-sized distributions of upper till samples

from the Danbury and New Milford quadrangles are shown in

Figure 4 and Table 1. It is apparent from these data that

the till contains about 55 to 80 percent sand, 15 to 40 per­

cent silt, and 2 to 15 percent clay of less than 2-micron

size. Samples NW-6 and C-7 are unusually fine grained. The

upper till ordinarily contains over 60 percent sand and less

than 10 percent clay (not counting granules and larger parti­

cles) . It is possible that the two anomalous samples have been contaminated by finely intermixed lower till. This ex­ planation seems plausible because their colors are similar

to those of oxidized lower till, and sample NW-6 was collected

from the southeast flank of a lower-till drumlin. However,

the two samples lack jointing and iron-oxide staining, at

least one of which is almost always present in the lower till

of the study area.

The author examined the mineralogy of five sand size-

fractions from each till sample (very coarse to very fine sand

on the Wentworth Scale). Rock fragments are common in the

very coarse and coarse sand fractions, but the finer fractions

are almost entirely composed of monominerallie grains.

Quartz and feldspar are the dominant minerals in the upper

till, except in carbonate-rich members that were derived in

large part from the Inwood Marble. Quartz and feldspar CLAY 25

TABLE 1

GRAIN-SIZE ANALYSES OF UPPER TILL SAMPLES3-

Weight Percent Weight Percent of Sample > 2 mm Sand - Silt - Clay Number Description (of total sample) Sand Silt Clay

NE-1 Upper till 22.9 79.5 16.2 4.3

NE-2 Carbonate- rich _b __b upper till 2.7 71.0

NE-3 Partly leached , upper till 4.4 70.7 27.4 1.9

NE-4 Partly leached upper till 3.0 71.1 26.1 2.8

NE-5 Leached oxidized upper till 4.6 61.8 32.0 6.2

NW-6 Upper till 9.9 56.9 29 .6 13.5

C-7 Upper till 9 .2 55.7 37.7 6.6

C- 8 Upper till 5.6 63.7 29 .2 7.1

NW-9 Upper till 16.3 80.1 17.6 2.3

NW-10 C-horizon soil in upper till 15.1 68.4 29 .6 2.0

NW-11 B-horizon soil in upper till 12.0 49.0 40 .0 11.0

aExclusive of clasts >20 mm in maximum diameter. ^Undetermined, but essentially the same as sample NE-3. 26

together constitute at least 70 or 75 percent of the sand

v component of the till. The ratio of quartz to feldspar

generally increases with finer grain size. Both of these minerals appear very fresh. Quartz grains are lustrous and

angular, while feldspar grains are commonly subangular to

subrounded.

Ferromagnesian minerals (biotite, hornblende, and

augite) compose variable percentages of the sand fractions

in the upper till. Monominerallic ferromagnesian grains are nearly absent in the very coarse sand, but they become pro­

gressively more numerous with decreasing grain size. A maximum abundance of 10 to 20 percent occurs in the fine to very fine sand fractions. This observation may simply re­

flect the fact that ferromagnesian minerals are finer grained

than most quartz or feldspar in the bedrock of the Danbury-

New Milford area.

The hornblende and augite grains in the upper till

appear fresh and lustrous. Biotite is the only major con­

stituent of the sand component that is appreciably weathered.

Weathered muscovite is present in some samples; but it is

subordinate to biotite. The biotite flakes have lost some

of their luster and elasticity, and they appear to have been

altered to vermiculite. Some biotite is partly altered to

limonite, but this degree of weathering is pronounced only

in certain micaceous rock fragments and in tiny flakes that

are intergrown with garnet. Weathered biotite is the only 27 apparent source of the weak iron-oxide staining that occurs on 5 to 10 percent of the quartz and feldspar grains.

Almandine garnet is a ubiquitous minor constituent of the upper till in the Danbury and New Milford quadrangles.

It is present in amounts ranging from under 1 percent to about 5 percent. Although whole garnet crystals occur in the coarse to very coarse sand fractions, the mineral is most abundant in the medium sand. Garnet percentages diminish to their minimum values in the very coarse and very fine sand fractions. Therefore, it is possible that the mineral grains in the upper till tended not to be comminuted below the size of medium sand. The presence of whole crystals of magnetite, pyrite, and rutile in the fine to very fine sand fractions supports this theory. However, the terminal grain size varies among different minerals according to their hardness, cleav­ age, and other factors (Dreimanis and Vagners, 1971).

The silt fraction of the upper till was given a cursory examination. It is dominantly a mixture of fresh quartz and lesser amounts of feldspar. Ferromagnesian min­ erals are less common than in the very fine sand fraction, and they are mostly biotite flakes.

Clay-size minerals in the upper till were identified by x-ray diffraction. Figure 5 shows a typical pattern of diffraction peaks for the untreated till clay, and Appendix C gives the basal d-spacings derived from both treated and untreated samples. The results of glycolation and heat M- 0 do s=o 1 ►d i o H* 8 M M 'f

ON O O -P- 5=0 l-i co CN > 0 ■£" i Co 1 CD < tf=o & ffi H * hi

O H* O c+ d CD M O o O CD

O

Figure 5. X-ray diffraction pattern of clay fraction of upper till sample NW-9 from New Fairfield two-till locality.

to 00 29 treatment show that illite and vermiculite are the principal clay minerals in all samples. The illite peaks indicate basal d-spacings of approximately 10.1 8 (001), 5.01 8 (002), and 3.34 8 (003). The d-spacings derived from the vermiculite peaks are 14.1 A (001), 7.19 A (002), 4.79 8 (003), and

3.57 8 (004) . These values compare favorably with those given for illite and vermiculite by Grim (196 8) and Carroll

(19 70). "Illite" is a poorly defined term, but it is used here to mean any "clay-mineral micas of both dioctahedral and trioctahedral types and of muscovite and biotite crystal­ lizations" (Grim, 196 8, p. 42). Since biotite is the prin­ cipal mica in the bedrock of the study area, the illite is probably derived from it by slight weathering and hydration.

Similarly, the vermiculite is believed to be biotite that has been hydrated (and perhaps otherwise altered) to a greater degree than the illite.

A low x-ray diffraction peak persisted at the

14-8ngstrom position for some samples, even after they had been heated to 550°C. According to the method used to identify the clay minerals, this observation suggests the presence of a relatively small amount of chlorite. However,

Grim (196 8) claimed that vermiculite is subject to rapid rehydration unless it is heated to 700°C. Therefore, it is possible that the vermiculite in the samples was not com­ pletely collapsed to the 10-8ngstrom spacing when they were x-rayed. The variable disappearance of both the 14-8ngstrom 30 and higher order peaks at 450° and 550°C likewise indicates that dehydration of vermiculite is not always permanent or complete at 450°C. It is far more effective at 550°C.

Quartz may be present in the clay fraction of the upper till, but its principal diffraction peak is masked by the

(00 3) peak of illite. Several samples contain mixed-layer clays of uncertain identity, as indicated by peaks in the

11.6 to 12.5 8 range. Dolomite is present in samples of carbonate-rich till from a pit in the Great Brook valley (NE).

It is destroyed by calcination on heating to 550°C. X-ray diffraction patterns indicate that clay-size microcline, albite, and tremolite are also present in the Great Brook samples. However, only tremolite was identified visually in the sand fractions.

The carbonate-rich variety of the upper till is worthy of special attention here. It is restricted to those parts of the Danbury and New Milford quadrangles that are above or downglacier from the Inwood Marble belt, and it differs in several respects from the typical upper till of southern

New England. The carbonate-rich till contains much calcite and dolomite, both as marble fragments and individual grains.

Most of the marble clasts are composed of sugary, equigranular dolomite marble (with lesser calcite). Except for certain boulders, they are usually so weathered that they can be crumbled by hand. Some of the clasts must have been pre- glacially rotted because they were crushed and smeared out 31 into long stringers at the time of till deposition.

An oxidized zone occurs in the upper few feet of the carbonate-rich till in some places. This zone is pro­ nounced at the till pit east of Great Brook (Figure 6). In the oxidized till at this locality, boulders of schist and gneiss have been so thoroughly weathered that only shadowy outlines remain. Using a Chittick apparatus, it was learned that the oxidized till has been leached to such a degree that it contains less than 1 percent total carbonate (Table 2, sample NE-5). Partial leaching has occurred below the oxi­ dized zone in the section shown in Figure 6. Elsewhere in the same pit, partial leaching has occurred where there is no oxidation at all (Table 2, samples NE-2, 3, 4). In the latter instance the dolomite content of the till matrix is relatively constant with depth (36.4 to 41.1 percent), while the calcite content increases from 1.5 percent near the surface of the till to 6.5 percent at a depth of 48-50 inches (127 cm).

Much of the leached calcite is probably derived from the till matrix, but it was suspected that some of it comes from the rottenstone marble clasts. The author sampled marble clasts of similar size and lithology from both the unleached and partly leached unoxidized till. The clasts from the un­ leached till had an average calcite content of 18.3 percent, with the remainder being almost entirely dolomite. Those from the partly leached zone averaged 9.4 percent calcite

(Table 2, samples NE-19, 20). Assuming that sampling averaged 32

Figure 6. Upper till on east side of Great Brook valley (NE). Pit face shows 4 feet of weathered colluvium (below ground surface) and 7 feet of oxidized till (above shovel) overlying fresh till with rotten- stone marble boulders. 33

TABLE 2

GASOMETRIC DETERMINATIONS OF CALCITE AND DOLOMITE WEIGHT PERCENTAGES IN TILL SAMPLES AND TILL CLASTS

Sample Percent Percent Total Calcite Number Calcite Dolomite Carbonate Dolomite

NE-2 6.5 39 .6 49 .5 .17

NE-3 2.6 41.1 47.1 .06

NE-4 1.5 36 .4 40 .8 .04

NE-5 0.1 0.5 0.7 .19

SC-16 2.9 10.6 14.4 .28

NE-19 18.3 71.0 95.2 .26

NE-20 9.4 86 .3 102 .9a .11

aThis anomalous value probably resulted from heating of gas in Chittick apparatus due to rapid dissolution of marble. 34

out compositional variations, it is concluded that both the marble fragments and the till matrix have experienced post­ glacial leaching of calcite. Significant removal of dolomite has occurred only in the oxidized zone of the till.

Lower till

The lower till composes nearly all the drumlins in the

Danbury and New Milford quadrangles, and it underlies most

areas of smooth topography in upland regions. Its thickness

is usually over 40 feet (12 m) and locally as much as

160 feet (49 m) (J. K. Adams, U.S.G.S., unpublished data).

The moist Munsell color of the lower till ranges from olive brown through olive to dark olive gray

(2.5Y 4/4 - 5Y 4/3 - 5Y 3/2). Thus its colors are similar

in chroma to the upper till, but they are darker. The lower

till is oxidized in nearly every exposure in the study area.

One exposure of unoxidized, carbonate-rich lower till was discovered on the east side of Danbury, near the intersection of the Still River and Newtown Road (SC). The fresh till at this locality was olive gray, and it was overlain by about

10 feet (3m) of olive-brown oxidized lower till.

The lower till is very compact and difficult to excavate.

It has a blocky structure that results from sub-horizontal and sub-vertical jointing. The sub-horizontal joints are usually closely spaced and parallel to the land surface.

Dark-brown iron-oxide staining coats the joint surfaces at 35 most localities. Both the jointing and the staining become weaker at depth (Figure 7) .

The stone content (greater than 1 centimeter) of the lower till in the Danbury-New Milford area varies from about

3 percent to over 25 percent, with the usual range being

5 to 10 percent. The stones are crystalline rocks, and most of them were derived locally. Stones of cobble to boulder size are scarcer than in the upper till. The majority of the stones are fresh and solid, except for some very micaceous schists and a few other lithologies that are rotten through­ out. Most of the stones do not appear freshly broken, and few are striated. Many of them are coated by the same iron- oxide staining that occurs on the till joints; and they may be completely covered by a glossy, adhesive film of silt and clay.

No definite ablation facies of the lower till has been discovered in the Danbury-New Milford area or elsewhere in southern New England. Most of the lower till probably had a subglacial origin. Its compact, fine-grained texture indi­ cates that it did not form by ablation of stagnant ice. It is possible that an ablation facies existed at one time and was removed by a later ice advance.

Table 3 and Figure 4 show the grain-size distributions of lower till samples from the Danbury and New Milford quad­

rangles. These data show that the lower till contains about

40 to 65 percent sand, 25 to 45 percent silt, and 10 to 36

Figure 7. Vertical pit face in weakly oxidized lower till, about 15 feet below ground surface, south of Newtown Road and 0.1 mile east of Still River bridge (SC). Note widely spaced sub-horizontal and sub-vertical joints (upper right). 37

TABLE 3

GRAIN-SIZE ANALYSES OF LOWER TILL SAMPLES9-

Weight Percent Weight Percent of Sample > 2 mm Sand - Silt - Clay Number Description (of total sample) Sand Silt Cla\

NW-12 Oxidized lower till 9.9 56 .3 31.7 12.0

WC-13 Oxidized lower till 5.3 52.9 33.3 13.8

WC--14 C-horizon soil in lower till 8.8 54.6 34.4 11.0

WC-15 B-horizon soil in lower till 2.9 61.0 29 .5 9.5

SC-16 Carbonate- rich , unoxidized lower till 4.8 40 .4 44.6 15.0

EC-17 Oxidized lower till 5.6 64.2 25.3 10 .5

NE-18 Oxidized lower till 4.4 56.4 31.9 11.7

Exclusive of clasts >20 mm in maximum diameter. 38

15 percent clay of less than 2-micron size. It has higher clay and lower sand percentages than the upper till. The locality represented by sample EC-17 has unusually sandy lower till with small pods of washed sand and gravel. This till deposit may be La remnant of the ablation facies.

The sand fraction mineralogy of the lower till was examined. It is very similar in most respects to the miner­ alogy of the upper till. Quartz and feldspar are the domi­ nant minerals, and they are generally fresh, Ferromagnesian minerals increase in abundance from about 1 percent in the very coarse sand fraction to 10 to 15 percent in the fine to very fine sands. The amphiboles and pyroxenes appear fresh, but the biotite flakes are more weathered than in the upper till. However, even the biotite appeared fairly fresh in the unoxidized, dolomitic lower till from Danbury. A few grains of fresh chalcopyrite were also noted in the till matrix at this one locality.

Minor constituents of the sand component of the lower till include garnet, magnetite, and traces of sphene, epidote, and unidentified minerals. Garnet is slightly more common than in the upper till, with a maximum abundance of 5 to

10 percent in the medium sand fraction. Otherwise, the minor mineral contents of the two tills are identical.

Iron-oxide staining has affected up to 20 percent of the grains in certain sand fractions of the lower till, but the staining is not appreciably greater than in the upper till. Nor is it more pronounced in the silt fraction, which is a mixture of quartz,feldspar, and weathered biotite. Thus the difference in color between the tills is not obvious when they have been disaggregated. It is believed that some of the color difference arises from the clay fractions. Clays

from the heavily oxidized tills and till soils are slightly darker and/or browner than clays from their fresh parent

tills. This difference may be due to the presence of amor­ phous iron oxides, which are not recorded on the x-ray dif­

fraction patterns. However, the overall variation in color between and. within the two tills is not the result of any

single cause. It is due to the interaction of oxidation,

local variations in mineralogy, grain-size distribution, and moisture content.

The clay mineralogy of the two tills is also very

similar. Figure 8 shows a typical x-ray diffraction pattern

for untreated lower till clay, and Appendix C gives the basal

d~spacings for all samples that were x-rayed. Laboratory

treatment revealed that illite and vermiculite are the prin­

cipal clay minerals in the lower till. The illite diffrac­

tion peaks were generally higher than for the upper till, but

this relationship was not strong or consistent enough for

gneralizations about the abundance of illite. Dolomite

occurs in the clay fraction of the carbonate-rich lower till

from Danbury, and diffraction patterns indicate the presence

of mixed-layer clays in several samples. Minor amounts of O n

-O

ro >o

H* O >o

H' <4- M O O P

Figure 8. X-ray diffraction pattern of clay fraction of lower till sample NW-12 from New Fairfield two-till locality.

o 43 eastward continuations of the till sheets described in this report.

It is apparent from the above-mentioned studies that a two-till association is common across southern New England.

The colors and textures of both tills vary from one locality to another, depending on the local bedrock and weathering phenomena. However, the similarities outweight the differences among occurrences of either till. The upper till at each locality is typically loose, fresh or weakly oxidized, structureless, and relatively coarse grained. The lower till is compact, deeply oxidized, jointed, and fine grained. Thus it is believed that the two tills are correlative throughout southern New England. Similar tills have also been described from adjacent parts of New York (Connally and Sirkin, 1973) and New Hampshire (Koteff and Stone, 1971). Drake (1971) presented cogent evidence that the "hard till" and "soft till" in central New Hampshire are basal and ablation facies from a single glaciation. However, both of Drake's tills may be equivalent in age to the upper till described here.

Although most workers in Massachusetts and Connecticut assigned different ages to the upper and lower tills, a so- called "two-till controversy" existed through the late

1960's. One group argued that the upper and lower tills might simply be ablation and basal deposits from the most recent, Late Wisconsinan glaciation. Their opinion was based on the textural differences between the tills and the apparent 44 absence of interglacial soils or an ablation facies of the

lower till. The other group claimed that the upper till is

Late Wisconsinan and the lower till is older — Early

Wisconsinan or possibly Illinoian. It is now generally

agreed that the two tills are the products of different glaciations that were separated by an interval of at least

interstadial magnitude. Pessl and Schafer (196 8) summarized

the evidence in favor of this theory. Their strongest argu­ ment was based on field observations: the deep jointing,

oxidation, and iron-oxide staining of the lower till, and

the contact-zone relationships at two-till localities. The apparent absence of buried soils or an ablation facies of the

lower till is still a problem. The oxidation zone in the lower till may be the remnant of a thick C-horizon, and the upper part of the weathering profile (including ablation till) may have been removed during the following glaciation.

There is only one well-exposed multiple-till locality

in the Danbury-New Milford area. It is a small pit on the hill just west of Disbrow Pond and east of the village of

New Fairfield (NW). This occurrence was examined in search

of evidence that would bear upon the ages of the tills. The

New Fairfield tills resemble their counterparts in the rest

of the study area with respect to the properties that have

already been described. Figures 9 and 10 show their fabrics, which were obtained by measuring the long-axis azimuths of

stones on a goniometer of the type described by Karlstrom 45

0 2 1______I______I no, of stones

S

Figure 9. Rose diagram showing long-axis azimuths of 40 stones in upper till at New Fairfield two-till locality (NW). Azimuths were measured in di­ rection of plunge. (19 52). It was difficult to obtain even a 40-stone sample

from the upper till because most stones were not sufficiently elongated (with a:c axial ratios less than 0.5). The relia­ bility of both fabrics is suspect because of their maxima in the east-northeast to west-southwest direction. This may re­

flect a reorientation of the stones by downslope movement to the northwest soon after the tills were deposited. Work by Pessl (19 71) showed that more reliable lower-till fabrics in western Connecticut indicate an ice advance from the northwest.

Other lines of evidence provide better support for multiple glaciation at the New Fairfield site. The contact between the two tills is sharp, and slabs of lower till have been included in the basal part of the upper till (Figure 11).

There is also a difference in provenance. Stone counts re­ vealed that 30 percent of the upper-till stones are exactly

the same rock type that outcrops in the pit area, and many others were probably derived from nearby. The lower till

contains only 10 percent of the underlying rock type, and

the 1 0 0 -stone sample included several low-grade metamorphic

rocks of possible Taconic Mountains origin that were not re­

corded from the upper till. The latter observation is com­ patible with an ice advance that came more from the west

than the one that deposited the upper till. Pessl (1971) noted that the lower-till fabric mode usually lies west of

the upper-till mode in areas where the tills1 are superposed. 48

Figure 11. Vertical pit face showing contact (marked by 6 -foot tape measure) between upper till and lower till, east of New Fairfield (NW). Note inclusion of lower till (about 1.5 feet above center of tape). 49

Flint (1961) reached the same conclusion for the Lake

Chamberlin and Hamden tills.

If it is granted that the upper and lower tills are of different ages, their absolute ages remain to be established.

The upper till was deposited by the Late Wisconsinan glaciation that covered New England. This conclusion is based partly on its stratigraphic position and freshness. Suitable materi­ al for radiocarbon dating has not been found within the upper till, but limiting dates have been obtained from strat- igraphically older and younger deposits at several localities in the Northeast. The first maximum age was obtained from a Farmdalian (Middle Wisconsinan) peat deposit in alluvium along the Delaware River valley, near Trenton, New Jersey

(Connally and Sirkin, 19 73). Sirkin and others (19 70) dated the peat at 26,80011000 yrs B.P. The deposit contains a mix­ ture of coniferous and deciduous arboreal pollen, which was interpreted as indicating the onset of a glacial climate.

Sirkin (19 74, personal communication) has since discovered a younger Farmdalian shell and peat deposit in a fluvial sand unit at Port Washington on Long Island. It is located between the Montauk(?) and Roslyn Tills in the Harbor Hill

Moraine. These tills are believed to be Altonian and

Woodfordian, respectively. Radiocarbon dates from the organic material are >43,000 - 22,000 yrs B.P.; and the associated pollen (pine, hickory, oak) indicates an interstadial climate.

Sirkin's dates show that the upper till in western 50

Connecticut is probably no older than 22,000 years. This conclusion is consistent with Schafer and Hartshorn's (19 65) statement that the last major ice advance in New England occurred about 20.,000 years ago.

A minimum age limit for the upper till in the study area can be inferred from Connally and Sirkin's (1970) study of the upper Wallkill Valley in southeastern New York. Using a radiocarbon date and inferred sedimentation rate for

New Hampton Bog No. 1, they assigned an age of 15,000 yrs B.P. to the basal bog sediment. Connally and Sirkin (1970, 1973) then gave the same general age for the nearby Wallkill Moraine.

This moraine formed during a recessional stillstand of the

Late Wisconsinan ice sheet a few miles north of the bog. The bog, in turn, is just north of the Pellets Island Moraine.

Considering the locations of these two moraines, one of them is probably correlative with the deglaciation of the Danbury-

New Milford area. An exact correlation has not been deter­ mined because the moraines cannot be traced eastward into

Connecticut. However, the 15,000 yrs B.P. date is believed to be a good approximation of the minimum age of the upper till in the study area.

The age of the lower till is still uncertain. It was previously mentioned that several workers have considered it to be an Early Wisconsinan deposit. The lower till has not yielded any absolute ages, though; and no one has proven that it is not pre-Wisconsinan. A Nebraskan or Kansan age 51 is unlikely because the till is not much more weathered than the upper till. The present author favors an Early Wisconsinan age for the same reason. One would expect greater differ­ ences in stone weathering, matrix alteration, and clay min­ eralogy if the lower till were Illinoian. If the upper till correlates with Sirkin's Roslyn Till on western Long Island, it is possible that the lower till is equivalent to the

Montauk(?) Till in the same area (L. A. Sirkin, 1974, personal communication). It has been noted that the Montauk(?) Till underlies Farmdalian sediments and is believed by Sirkin to be Altonian (Early Wisconsinan). More work must be done to relate the Connecticut and Long Island stratigraphies with certainty. CHAPTER IV

GLACIAL-MELTWATER DEPOSITS

Basis for delineation and correlation of map units

Glaciofluvial and glaciolacustrine deposits are common

in the Danbury and New Milford quadrangles. They occur principally in the Still River and Housatonic River valleys

and along some of their tributaries. No meltwater deposits

are known to exist beneath either of the tills. Pit ex­ posures show that the fresh outwash overlies bedrock or one

of the two tills (usually the upper till). Therefore, it

is considered to have formed entirely during the final de­

glaciation of the study area in Late Wisconsinan time.

The glaciofluvial deposits were graded to local base-

level controls. They were either graded to stable base levels

such as valley floors and topographic cols, or they were

controlled by ephemeral levels such as proglacial lakes and

stagnant-ice masses. The glaciofluvial deposits in the

study area include eskers, crevasse fillings, kames, kame

terraces, and valley trains. Kames and kame terraces are

especially common because many of the meltwater deposits were

laid down in contact with residual ice masses. Keeping in

mind the influence of the wasting ice, the kames can be

52 53 studied in much the same way as stream deposits. Many of them contain uncollapsed sedimentary structures that reveal the direction of meltwater flow. Topographic profiles of kame and other outwash surfaces can be used to determine both the direction and gradient of flow (Plate II).

Glaciolacustrine deposits abound throughout the Still

River lowland, its northward topographic extension into the

Housatonic River valley, and the Lake Candlewood valley.

The deposits include lake-bottom sediments and the associated deltas that were built into the lakes. The glacial lakes formed during deglaciation of the study area as meltwater was trapped between the ice margin and topographic barriers to the south and east. Glacial Lake Candlewood occupied the same basin as modern Lake Candlewood. It drained southward into glacial Lake Danbury. In the- course of four stages, glacial Lake Danbury occupied the entire lowland that has developed on the Inwood Marble from Bethel to New Milford.

Glacial Lake Kenosia was a small water body that existed in the southwest corner of the Danbury quadrangle. No definite shore line features have been found in these glacial lake basins, but there are local breaks in slope that may have resulted from wave erosion.

End moraines are absent in the Danbury-New Milford area, but it is possible to draw conclusions about the style of deglaciation from a study of the meltwater deposits. There are two general ways in which the ice could have retreated 54

from the study area: by regional stagnation and downwastage,

or by the gradual withdrawal of an active ice sheet that was fringed by a stagnant zone*. The latter alternative is

favored on the strength of field evidence. The meltwater

deposits have topographic and internal characteristics that

enable them to be grouped into morphologic sequences as

defined by Koteff (1974).

Jahns (19 41, 1953) was first to apply the sequence

concept to the deglaciation of southern New England. He

concluded from his field work in Massachusetts that the glacial-meltwater deposits occurred in contemporaneous

groups, or "sequences". Jahns' sequences are associations of fluvial ice-contact sand and gravel deposits that were built from successive marginal positions of an active ice sheet during its northward retreat. Each sequence was graded

to a particular base level, and ice retreat caused deposition of new sequences by opening lower paths of escape for the meltwater. Jahns1 (19 41) measurements of the lengths of sequences parallel to the direction of ice recession showed

that the stagnant-ice zone had an average width of 3 miles

(4.8 km). \ Other workers in Massachusetts accepted and applied the concept of stagnation^zone retreat, including Currier

(1941a), Jahns and Willard (1942), and Hansen (1956). Koteff

(19 74) introduced the term "morphologic sequence" to stress that an individual sequence is a group of essentially 55 contemporaneous landforms. He described eight types of sequence deposits. Most of them are classified according to whether they are composed of glaciofluvial and/or glacio- lacustrine sediments, the spatial order in which these en­ vironments occur, and whether or not the upstream end of the sequence was in contact with the stagnant-ice zone. For example, a "fluvial-lacustrine ice-contact sequence" was formed in part by meltwater streams whose gradients were low enough for fluvial deposition to begin within or immediately beyond the glacial stagnant zone. The stream sediments can be traced to the point where the meltwater entered a lake and lacustrine deposition occurred.

Koteff's definition of a morphologic sequence is broader than the meaning assigned by Jahns. It includes valley trains and glaciomarine deposits as well as sediments that were deposited in actual contact with dead ice. Koteff also demonstrated that most of the sediment in sequence deposits was probably derived from shear planes at the con­ tact between active ice and the marginal stagnant zone.

The principal use of the morphologic sequence concept lies in tracing successive positions of a retreating ice margin, especially in the absence of recessional moraines.

Ice-contact sequences are good for this purpose because their proximal ends were built at the ice margin and usually have a topographic expression. Fine examples of these "heads of outwash" are seen in the southeast part of the Danbury quadrangle, near Interstate Route 84 (EC-SE). Plate I shows that scarps mark the upstream ends of several ice-contact sequences in this area. Map unit Qsr1 is a very well-defined morphologic sequence, especially in profile (Plate II). Where possible, lines have been drawn on Plate I to show parts of the distal edge of the Late Wisconsinan stagnant-ice margin during its recession across the study area. These ice-front positions have been shown only where they can be inferred on the basis of actual meltwater deposits. It is not possible

to determine the exact configuration of the ice margin across the full width of the quadrangles. The heads of outwash are not laterally continuous, and several correlations are possible between map units in different areas.

The morphologic sequence deposits in the Danbury-

New Milford area indicate that the last ice sheet withdrew by stagnation-zone retreat to the north or northwest. The distribution of glacial Lake Danbury deposits is best ex­ plained by several drops in lake level as the ice pulled back from Bethel to New Milford and lower spillways opened

up. Deltas and glaciofluvial current structures indicate meltwater flow to the south or southeast. This general flow

direction is also indicated by topographic profiles (Plate II),

changes in outwash textures, and pebble counts from gravel pits. The esker system that fed the delta at Lanesville

(C-SC) is over 1.5 miles (2.4 km) long. The lengths of this

and other glaciofluvial units indicate that the glacial 57 stagnant zone had an average width of at least 2 miles

(3.2 k m ) .

In summary, Koteff's model of a retreating ice front is accepted for the Danbury and New Milford quadrangles on the basis of field evidence. Meltwater deposits have accordingly been mapped as a series of chronologic units, which are shown on Plates I and II. Letter symbols were assigned to these units in accordance with U.S.G.S. conven­ tions. Some map units (e.g. Qgp^ 2 3 ) are whole sequences as defined by Koteff. Others, such as the glacial Lake

Kenosia deposits (Qk and Qkb), are fluvial-lacustrine se­ quences that have been subdivided in order to differentiate lake-bottom sediments from fluvial deposits that were graded to glacial lakes. On the other hand, units Qsr and Qpb2 include several poorly defined sequences that were graded to a common lake level. The individual map units are described in the remainder of this chapter.

Meltwater-stream deposits

The map units described in this section include all glaciofluvial deposits in the Danbury and New Milford quad­ rangles that were graded to non-lacustrine base levels. In­ cluded here are both fluvial ice-contact and fluvial non-ice' contact sequences (Koteff, 19 74). These deposits consist of sand, gravel, and minor silt. Many of them are kames or kame terraces that exhibit ice-contact slopes. Borrow pits 58 commonly expose slumped beds of coarse, poorly sorted sedi­ ment. Meltwater-stream deposits are described here in order of decreasing age, starting in the southern part of the

Danbury quadrangle.

Unit Qgb is located in the southeast corner of the

Danbury quadrangle. It is an ice-contact gravel and sand deposit with knobby topography and collapsed bedding. The glacial streams that formed this unit were probably controlled by a bedrock threshold near the northern border of the Bethel quadrangle. This inferred base-level control is 0.4 miles

(0.6 km) south of Berry School, at an elevation of about

485 feet. A prominent meltwater channel lies just north of unit Qgb. It was a feeder channel for at least some of the. meltwater that deposited the unit. Additional meltwater and sediment probably came from stagnant ice in the vicinity of King Lake.

Unit Qgs consists of scattered kames, kame terraces, and ice-channel fillings near the southern border of the

Danbury quadrangle (SC-SE). The unit is composed of ice- contact sand, gravel, and silt; and it is very bouldery and poorly bedded in places. Ice-channel fillings are prominent in the swampy basin northeast of Bethel. The Qgs deposits were probably graded to the lowest col on the drainage divide between the Still River and Saugatuck River basins. This control point is a narrow gap located 2.6 miles (4.2 km) south of the quadrangle border. It is at an elevation of 59 about 410 feet and will henceforth be called the "Saugatuck

Divide". Plate II shows topographic profiles of the highest surfaces on unit Qgs.

Three successive outwash units occur in the vicinity of the Pond Brook valley. Pond Brook is a small tributary of the Housatonic River that heads in the east-central part of the Danbury quadrangle. The glacial stream units in this area are Qgp^, Q9P2 f an<^ Q9 P 3 * Each is composed of sand, gravel, and silt deposited by meltwater that flowed southeast­ ward into the Pond Brook valley and thence to the Housatonic

River. The first two units were deposited directly into the head of the Pond Brook valley, near the east border of the quadrangle. Their elevation difference is due to the thinning and retreat of the stagnant-ice margin. Unit Qgp3 was graded to the lowest point on the divide between Pond Brook and the

Still River. This base-level control has an elevation of approximately 380 feet. It will subsequently be referred to as the "Pond Brook Divide". All three units were deposited between masses of dead ice and the till upland. Slabs of flowtill are locally interbedded with unit Qgpj, and kettle holes are present on unit Qgp2 (Plate I). The ice margin positions shown for the Qgp units on Plate I are inferred from the topography of ice-contact slopes and the locations of ice-channel fillings. The latter presumably formed within the glacial stagnant zone (Koteff, 19 74). 60

Unit Qgt occurs along the lower part of Town Farm Brook in the east-central part of the New Milford quadrangle. It consists of ice-contact sand and gravel that was deposited along the walls of Clatter Valley. There is no definite base-level control for this unit. It was probably graded to the till ridge that projects northward into the lower end of

Clatter Valley (northwest of Wolf Pit Mountain).

Unit Qgg is also composed of ice-contact sand and gravel.

It is a series of kame terraces that were deposited along both sides of the Great Brook valley, northeast of New Milford

(NE-EC). There are few pit exposures in this unit, but ex­ cellent kame terrace morphology is seen in the vicinity of

Center Cemetery. There is no apparent base level for the

Qgg deposits. Their highest elevations lie on an even depo- sitional gradient that slopes southward, but the kame terraces end abruptly at New Milford. Beyond their termination there is no suitably high surface that could have been a base-level control for the meltwater streams. There is only the

Housatonic River valley and its lower-level meltwater deposits.

It is possible that unit Qgg was graded to an ice or drift barrier at the Housatonic Gorge, but there are no correlative deposits between New Milford and the gorge. For lack of positive evidence, it is believed that the meltwater streams were controlled by a temporary threshold on stagnant ice near

New Milford. Unit Qgsh is located in the vicinity of Sherman Church, in the northwest part of the New Milford quadrangle. It is composed of gravel and sand in an area of knobby, bedrock- controlled topography. The internal structure of the unit is not exposed, but its external morphology suggests that it was deposited among stagnant-ice masses that occupied the

Morrissey Brook valley. The meltwater flowed south through the Sherman area along the Rocky River valley (since flooded by Lake Candlewood).

Unit Qgm is a small gravel deposit in the northwest corner of the New Milford quadrangle. It occurs as a kame terrace on the east side of the Morrissey Brook valley. This unit was deposited by meltwater that flowed northeast along the stagnant-ice margin and into the Housatonic River valley.

Unit Qgnm is a dissected valley train (NW-NC). It consists of gravel and sand that was deposited in the

Housatonic River valley when the ice margin had retreated into the Kent quadrangle. The unit is exposed in the pit complex at the north edge of the New Milford quadrangle and on the west side of U.S. Route 7 at Boardman Bridge. An excavation at the latter locality shows about 12 feet (3.7 m) of Qgnm outwash overlying very fine-grained glaciolacustrine

sediments (Figure 12). The log from nearby well NMI 15

reveals a similar stratigraphy (Appendix A). The principal bedding in unit Qgnm is nearly horizontal, and cross-bedding

dips generally south. The unit lies at an intermediate level 62

Figure 12. Gravel outwash (unit Qgnm) overlying rhythmically- bedded lake-bottom sediments (unit Qhrb) west of U. S. Route 7 at Boardman Bridge (NC). between higher kame terraces and lower stream-terrace or flood-plain deposits. Plate XI shows a longitudinal profile of the Qgnm surface. It is apparent that its gradient is likewise intermediate between older and younger deposits.

The variation in slope among these adjacent map units is the result of differences in depositional gradients combined with the effect of postglacial uplift. Glacial lake deltas and shore line features are too scarce and poorly distributed to enable an estimate of the amount of postglacial uplift in the study area. Jahns and Willard (19 42) found that it has been 4.2 ft/mi in the Connecticut Valley of Massachusetts.

Their figure is accepted for the study area, but it may be slightly high for the latitude of the Danbury-New Milford region.

Unit Qg includes scattered sand and gravel deposits of uncertain chronologic position. They were built by meltwater streams, but their correlation with the other map units (Plate IV) is uncertain.

Glacial Lake Danbury deposits

Following the convention of recent U.S.G.S. quadrangle reports, glacial-lake deposits here include both lake-bottom sediments and all deposits of meltwater streams that entered the lakes. Most of the map units distinguished here are subdivisions of fluvial-lacustrine ice-contact sequences.

The largest lake was glacial Lake Danbury. This was a proglacial lake that was impounded between the retreating 64

Late Wisconsinan ice margin and the highlands to the south.

Sand, gravel, and silt were deposited by meltwater streams that entered the lake; and sand, silt, and clay accumulated on the lake bottom. Glacial Lake Danbury had four stages, during which it covered most of the Still River lowland and part of the Housatonic River valley. The lake levels were successively lower from south to north because ice recession opened lower outlets to the north. Thompson (19 71) named the stages according to the outlet channels or river valleys into which the lake spilled. They were the Saugatuck River stage, Pond Brook stage, Pumpkin Hill stage, and Housatonic

River stage.

Saugatuck River stage deposits

The Saugatuck Divide was the spillway during the

Saugatuck River stage of glacial Lake Danbury. Units Qsr,

Qsr^, and Qsr2 are composed of glaciofluvial sediments that were graded to the lake during this stage. They were de­ posited in an ice-contact environment by meltwater streams that entered glacial Lake Danbury from the north, east, and west. Profiles of these units are shown on Plate II.

Units Qsr^ and Qsr2 were built from successive ice margin positions in the east-central part of the Danbury quadrangle.

The north ends of both units.are topographically expressed heads of outwash. Unit Qsr-^ exhibits a general gradation from boulder gravel in the north to sand and cobble gravel in the south. The isolated Qsr^ deposit near Plumtrees School 65

(SE) may be a kame delta that formed as the ice melted out from the Limekiln Brook valley. The crest of this hill lies on the same gradient as the main body of Qsr^ outwash

(Plate II).

Unit Qsr includes the remainder of the deposits that were graded to the Saugatuck River stage lake level. The majority are kames or kame terraces, and they are composed mostly of sand and gravel. Typical ice-contact features such as poor sorting and collapsed bedding are present in the Qsr deposits. One can infer local ice margin positions that were associated with the deposition of this unit (Plate I), but they cannot be correlated definitely with the positions shown for units Qsr-^ and Qs^. At least one glacial lake delta is included in unit Qsr. It is located in the German­ town district of Danbury {SCj Figure 13) . The contact between its topset and foreset beds is at an altitude of about 440 feet. This was presumably the local surface al­ titude of glacial Lake Danbury during the Saugatuck River stage. Other probable Qsr deltas are indicated on Plate I, though their internal structure is not exposed or it has been obscured. The hill at Wooster Cemetery (SC) is believed to be a kame delta that was built by meltwater streams flowing south from the vicinity of Lake Candlewood. Plate II shows that the Qsr deposits in this area lie on a uniform gradient that intersects the level of glacial Lake Danbury at the delta(?). There is another probable delta just north of 66

Figure 13. Delta from Saugatuck River stage of glacial Lake Danbury, Germantown district of Danbury (SC). Foreset beds dip to east. 67

Spring Street in downtown Danbury (SW). It may have formed by an influx of sediment from the ice that occupied the

Mill Plain Swamp basin to the southwest.

Unit Qsrb was deposited on the bottom of glacial Lake

Danbury during the Saugatuck River stage. It occurs at the

surface in lowland areas around the city of Danbury, except where covered by stream-terrace or flood-plain alluvium.

The unit is commonly more than 50.feet (15 m) thick, and it exceeds 100 feet (30 m) in places. Subsurface information

about the lake sediments is provided by many well and test hole logs (Melvin, 1970). Plate I shows some of the most

significant borehole sites, and their logs are given in

Appendix A. Rhythmic bedding is present in some of the Qsrb

deposits, and it may be of seasonal origin.

The Qsrb deposits east of Danbury occur at two topo­

graphic levels. These two surfaces are most pronounced south­ east of Newtown Road, in the Limekiln Brook valley (SC-SE).

Well-bedded glaciolacustrine sand forms a plateau north of the sewage treatment plant at an elevation of about 370 feet.

The adjacent and lower lake-bed has an average elevation of

300 feet, and it is underlain by silt and clay (test hole

DY 75). The same relationship exists at other nearby locali­

ties, and it indicates two closely related episodes of lake- bottom sedimentation. Older sand deposits were laid down

among masses of stagnant ice early in the deglacial history

of a particular locality. Current ripples and cross-bedding 68 in these sands are evidence of a swift current. The younger silt and clay beds were then deposited in a lower, quiet water environment when the lake was more fully developed.

Although the two lacustrine facies are distinguished by hachured contacts on Plate I, they are not assigned to dif­ ferent map units because of their time-transgressive nature.

The silt and clay facies near Bethel (well BT 29) could easily be the same age as a sand facies farther north.

Pond Brook stage deposits

Units Qpb-^ and Qpb2 are glaciofluvial deposits from the Pond Brook stage of glacial Lake Danbury. During this stage the lake level was controlled by the Pond Brook Divide.

Unit Qpb^ is located near the south end of Lake Candlewood

(C). An esker in the Beaver Brook valley marks the probable location of the glacial stagnant zone when this unit was deposited, but isolated masses of dead ice persisted beyond the main ice margin. Lenses of flowtill were interbedded with sand and gravel in the abandoned pit complex east of Great

Plain School. The upper part of the section at this locality

(now destroyed) may have been deltaic. Gravel beds with a topset-foreset relationship were exposed in the lobate hill just north of Interstate Route 84 and west of Beaver Brook.

The altitude of the topset-foreset contact was about 370 feet.

The Qpb^ deposits are shown in profile on Plate II. The scarp that is seen on this unit is an erosibnal feature. It 69 probably resulted from the later drainage of glacial

Lake Candlewood across the Qpb^ deposits.

Unit Qpb2 occurs principally along the west side of

the Still River valley, near the village of Brookfield.

It is a heterogeneous group of kames, kame deltas, and ice-

channel fillings. The origin of these deposits is problem­ atical because they seem too low to have been graded to the Pond Brook stage lake level. The only lower outlet from the valley is at the junction of the Still River with the

Housatonic River. A northward meltwater drainage to this point could not have developed. The ice margin blocked the

Still River valley north of Brookfield and built a south- facing delta across its full width at Lanesville (SC). The

Qpb2 deposits do not show any topographic expression of meltwater flow direction, but some of them do have undisturbed bedding that indicates unequivocal southward flow. Some of the coarse, ice-contact gravels may have been deposited as ice-channel fillings in the glacial stagnant zone. An ex­ ample is the north-south ridge in the pit complex west of

Brookfield (NC-NE).

Other Qpb2 deposits are small sand hills that formed as kame deltas. One such deposit is the partially excavated hill just north of the power line and west of Brookfield (Figure

14). These deltas have clearly defined, uncollapsed bedding that reveals the paleocurrent directions (Figure 15). The tops of the deltas have been excavated, but they never were Figure 14. Kame delta from Pond Brook stage of glacial Lake Danbury, west of Brookfield village. Foreset beds dip to south.

gg&sjg&gj

Figure 15. Foreset sand beds in kame delta west of Brookfield village. Cross-bedding indicates flow to south. 71

as high as the Pond Brook stage lake level. It is likely

that ice recession was so rapid that they were not built up

to grade with the lake surface.

The remainder of the Qpb2 kames are sand and gravel

deposits that are typically slumped and contorted. Bedding

may be completely absent, and there are masses of unsorted

cobble and boulder gravel (Figure 16). These sediments may have been dumped into local meltwater pools on the stagnant

ice. Alternately, the ice margin may have retreated so

rapidly that they were not built up to grade. In either case,

an unstable glacial-stream environment was soon followed by

deposition of undisturbed lake-bottom sediments. Exposures west of Brookfield showed that flat glaciolacustrine silt beds were laid down above greatly deformed sand beds. The

latter in turn had been deposited on coarse glacial-stream sediments (Figure 17).

Unit Qpbb consists of the lake-bottom deposits from

the Pond Brook stage of glacial Lake Danbury. They are partly concealed by more recent deposits along the Still River, but their occurrence on the valley floor is known from stream- bank exposures and subsurface data (e.g. test hole BD 8 ).

The Qpbb lake sediments have an average thickness of less than

50 feet (15 m) in the center of the Still River valley. The hillside deposits along the east wall of the valley (SE-NE) are probably just a thin veneer over the underlying till.

The unit is also thin and patchy at Brookfield, where most of 72

Figure 16. Collapsed boulder gravel (unit Qpb2 ) in kame on west side of Still River valley, northwest of Gallows Hill (SC).

Figure 17. Flat-bedded glaciolacustrine silt overlying folded and faulted sand beds deposited in an ice-contact environment. West-facing pit exposure is located in unit Qpb2 , west of Brookfield village (HE). 73

the relief is due to the outcrops of Inwood Marble

(Plate III) .

Pumpkin Hill stage deposits

Unit Qph occupies part of the river lowland between

Lanesville and New Milford. It includes a variety of glacio-

fluvial deposits that were formed during the Pumpkin Hill

stage of glacial Lake Danbury. The lake level during this

stage was controlled by a spillway across the north end of

Pumpkin Hill (on the southwest side of the Housatonic Gorge).

A gap on the ridge crest clearly marks the location of the

spillway (Figure 18). Its elevation is about 295 feet, which corresponds to the elevation of a major delta at Lanesville.

The Lanesville delta is the best indicator of the lake

level during the Pumpkin Hill stage. It trends northeast

across the Still River valley. The contact between the top-

set and foreset beds is exposed in several places at altitudes

of about 290 feet (Figure 19). The log from test hole NMI 1

(Appendix A) shows that at least 55 feet (17 m) of bottom-

set beds (of unit Qphb) were deposited in front of the delta.

A network of discontinuous eskers or crevasse fillings ex­

tends northward from the delta for nearly 2 miles (3.2 km).

They were deposited by the meltwater streams that flowed

through the glacial stagnant zone and fed sediment to the

delta. The north side of the delta is a steep ice-contact

face that is now breached by a small tributary of the Still

River. Figure 18. Northward aerial view of Housatonic Gorge. Gap on hill west of gorge was spillway for glacial Lake Danbury during Pumpkin Hill stage. Photo by J. A. Pawloski.

Figure 19. Delta at Lanesville from Pumpkin Hill stage of glacial Lake Danbury. Foreset beds dip to south. Topset-foreset contact is visible near top of pit face. 75

The remainder of the Qph deposits consist of sand and gravel that overlies a bedrock high north of Lanesville

(C-EC). None of this material is appreciably higher than

300 feet, so it may have been graded to a lower level than the Lanesville delta. The ice that, blocked the north end of the Housatonic Gorge may have melted enough so that the meltwater from glacial Lake Danbury could enter the gorge at a slightly lower elevation late in the Pumpkin Hill stage.

The Pumpkin Hill stage lake-bottom deposits (unit

Qphb) are of limited extent. The single ice margin position for this stage was very close to the associated spillway, and sedimentation occurred only in the vicinity of Lanesville.

The sand deposits farther south (unit Qpbb) are too coarse to have been transported beyond the finer deltaic sediment at

Lanesville. This textural change is the basis on which the dashed Qpbb-Qphb contact has been drawn.

Housatonic River stage deposits

Unit Qhr was graded to glacial Lake Danbury during the

Housatonic River stage. The lake drained directly through the Housatonic Gorge during this final period of its existence.

The Qhr deposits along the Housatonic River north of Wannuppee

Islands are a series of discontinuous kame terraces. They are composed mostly of sand and gravel, and they contain typi­ cal ice-contact features such as unsorted boulder gravel pods and collapsed bedding (Figure 20). Lenses of flowtill are chaotically mixed with unit Qhr in the pit near the north Figure 77 border of the New Milford quadrangle. Plate II shows that the topography of the unit in this part of the valley is also characteristic of ice-contact deposits.

The appearance of unit Qhr abruptly changes to the southeast of the constriction in the Housatonic Valley that is just upstream from Wannuppee Islands. The change in topography is evident on Plate II. The surface of the unit becomes relatively even and continuous, and its gradient steepens slightly. The sediments are also finer-grained and lack the ice-contact structures that occur farther north.

Pit exposures and subsurface data (Appendix A) show that these deposits overlie thick lake-bottom sediments that are part of unit Qhrb. Pits east of Ferriss Pond also revealed spec­ tacular fluvial dunes and other structures that were indi­ cative of a strong southward current at the time of deposi­ tion of the Qhr sands. Thompson (1971) concluded from this evidence that there must have been a transition from a glaciolacustrine to a glaciofluvial environment in the area between Wannuppee Islands and the Housatonic Gorge. The historical implications of this observation are discussed in Chapter VI.

The Housatonic River stage lake-bottom deposits are well-exposed only on the west side of the valley between

Lanesville and New Milford. Elsewhere they are overlain by younger outwash and stream-terrace or flood-plain alluvium.

Glaciolacustrine sediments are exposed in a scarp between units Qgnm and Qst near the west end of Boardman Bridge

(Figure 12), and they were encountered in well NMI 15.

Therefore, standing water extended at least as far north as

Boardman Bridge, but the lake sediments in this area are

probably younger than the part of unit Qhrb that lies south

of Wannuppee Islands. It has been shown that stagnant ice

still occupied the valley at Boardman Bridge when the Qhr kame

terraces were built. These kames are believed to be con­

temporaneous with the southern part of unit Qhr, which in

turn overlies part of unit Qhrb. However, the lake-bottom

sediments at Boardman Bridge must be younger than the adjacent

Qhr-kame terraces.

Rhythmic bedding is present locally in the very fine

sand, silt, and clay of unit Qhrb. The thickness of the unit

commonly exceeds 25 feet (7.5 m), and it is more than 100 feet

(30 m) in places. Subsurface data (Melvin, 1970) show that

the thickest lake-bottom deposits are located just northwest

of the junction of the Housatonic River with the Aspetuck

River (C-NC).

A problem arises from the fact that the average ele­

vation of the buried lake-bottom sediments just north of

the Housatonic Gorge is 19 5 feet. On the other hand, the bedrock elevation at the north end of the gorge is only

180 feet (Plate II). It is doubtful that glacial Lake Danbury

extended south of the gorge because there is no other likely

dam site farther down the river. Thus the spillway at the 79 gorge must have been at least 15 feet higher than the present rock floor. There may have been this amount of postglacial bedrock erosion, but the rock at the gorge is resistant schist and gneiss. Alternately, the gorge may have been blocked by a plug of drift and/or ice during deglaciation of the study area. The total thickness of such a plug is uncertain. It is possible that unit Qhr is correlative with outwash deposits south of the gorge, but recent flooding has obscured most fluvial deposits in the latter area. The drop in elevation through the gorge hinders efforts to correlate units on either side by means of topographic profiles.

Glacial Lake Kenosia deposits

Glacial Lake Kenosia occupied the Mill Plain Swamp basin in the southwest corner of the Danbury quadrangle. It drained eastward across the till upland and into glacial

Lake Danbury. Harvey (1920) realized that a glacial lake existed in this area, and she called it "Lake Kanosha"

(an old spelling of "Kenosia"). However, she thought that it drained in a westward direction.

Unit Qk is composed of sand, gravel, and silt that was graded to glacial Lake Kenosia. These sediments were deposited by meltwater streams from ice that lay to the west.

The kame terrace surface at St. Peter's Cemetery can be traced westward into the Brewster quadrangle as a series of succes­

sively higher kames. Sediment textures range from sand and pebble gravel at the cemetery to cobble gravel in the 80

Brewster quadrangle. The average thickness of these Qk deposits is greater than 50 feet (15 m). The elevation of the lake surface was no less than 470 feet (the altitude of the highest lake-bottom sediments). Therefore, the outlet channel (in till) east of Danbury Fairgrounds has been eroded to at least 30 feet lower than its original altitude.

Unit Qkb was deposited on the bottom of glacial

Lake Kenosia. It consists of silt, sand, and clay. The thickness of the unit is generally greater than 40 feet

(12 m) and locally as much as 100 feet (30 m ) .

Glacial Lake Candlewood deposits

A third glacial lake is believed to have existed in the Lake Candlewood valley. Present day Lake Candlewood is artifically impounded. It was created by damming the north- flowing Rocky River. Harvey (1920) realized that a proglacial lake occupied the Rocky River basin as it was deglaciated.

She described well logs from the Rocky River valley that the present author interprets as recording lake-bottom sediments.

Although the area is now flooded, patches of sand and gravel occur as islands and peninsulas that stand as much as 15 feet

(4.5 m) above lake level. Most of these deposits were probably graded to glacial Lake Candlewood. They are designated as unit Qc on Plate I, The lake spillway was at the lowest point on the drainage divide between the Rocky River and the Still

River basins. This point was near the south end of the modern lake, and it was at an elevation of about 430 feet. CHAPTER V

LATE-GLACIAL AND POSTGLACIAL DEPOSITS

Eolian deposits

A thin layer of windblown sand and silt overlies the glacial deposits in the study area. It is not shown on

Plate I because of its widespread patchy distribution. The eolian mantle is yellowish brown to yellow and usually lacks visible stratification. The thickness of the unit varies from a few inches in upland regions to 6 feet (1.8 m) on the Lanesville delta (Figure 21). The thickest eolian cover formed on glacial-meltwater deposits in valley areas, where its accumulation was favored by partial shelter from strong winds and the presence of easily eroded outwash. Some of the eolian mantle was probably derived from till — especially the loose upper till. Ventifacts are sometimes found where sections of eolian sand are exposed in borrow pits.

A representative sample of eolian mantle was collected from the Lanesville delta locality. It was sieved and found to contain 74 percent sand, 25 percent silt, and 1 percent clay. Seventy-three percent of the total sample was medium to very fine sand. This is a coarser and more poorly sorted grain-size distribution than Smith and Fraser (1935) reported

81 82

Figure 21. Eolian sand and silt (dark layer) overlying topset beds of the Lanesville glacial lake delta.

i 83 for loess samples from the Boston, Massachusetts, area. They noted strong maxima (56 to 83 percent) in the .062 to .031 mm range. However, Denny (1936) discovered that eolian "brown loam" from southern Connecticut contains only 12 to 50 percent particles that are finer than .0 74 mm. These regional dif­ ferences are not surprising when one considers the variability of source areas and wind conditions. Another eolian sand sample was collected from the surface of unit Qgs northwest of Bethel (SC), and it was both coarser and more poorly sorted than the Lanesville sample. It contained 74 percent sand, 20 percent silt, and 6 percent clay. Sixty-five percent of the sample was medium to very fine sand.

The sand fraction of the Lanesville sample was composed mostly of quartz, feldspar, and mica. These minerals were commonly rounded and frosted from the combined effects of wind and water transport. Small amounts (S 2 percent) of horn­ blende, garnet, magnetite, and sphene were also noted in the sample. Denny (19 36) similarly identified a variety of heavy minerals in the eolian sand of southern Connecticut.

The eolian sand and silt was deposited in late-glacial time before vegetation was sufficiently dense to curtail wind erosion. The wind-blown sediment layer is a surface deposit; so it is now part of the weathering profile.

Denny (19 36) pointed out that the eolian mantle coincides en­ tirely or in pa, it with the B-horizon. Thus soil development is responsible for the oxidation and brownish color of the 84 eolian sediment. Frost heaving has also disturbed it and mixed the wind-blown sand with near-surface glacial deposits.

Stream-terrace deposits

Stream-terrace alluvium (unit Qst) occurs principally along the Housatonic River and Still River. Terrace surfaces have been cut in glacial-lake and meltwater-stream deposits.

These surfaces are entirely erosional in a few places; but most are capped by contemporaneous sand, gravel, and silt derived from erosion of glacial deposits. Terraces also occur along a few of the smaller streams, such as the east and west branches of the Aspetuck River (NC-NE).

The stream-terrace surfaces usually lie 10 to 30 feet

(3—9 m) above modern flood-plain levels. The thickness of unit Qst is 4 to 8 feet (1.2-2.4 m) along the Still River between Danbury and Brookfield. The terrace alluvium over­ lies finer grained lake-bottom sediments of unit Qpbb in this area. It also has cross-bedding and current ripples that indicate a northward flow direction. Unit Qst is 10 to 20 feet (3-6 m) thick along the Housatonic River, with an average thickness of about 15 feet (4.5 m). Subsurface data

(Melvin, 19 70) show that it overlies deposits from the

Housatonic River stage of glacial Lake Danbury (unit Qhrb).

At least two terrace surfaces occur along the Still

River and Housatonic River. They are shown on Plate I, and and the Housatonic River levels are indicated on Plate II.

The age relationships between the terraces in one valley and 85 those in the other valley are not known. Consequently, all

Qst deposits have been mapped as a single chronologic unit; and hachured contacts have been used where it is possible to distinguish adjacent terrace levels.

Flood-plain deposits

Unit Qal is recent alluvium deposited on flood plains by modern streams. The flood-plain sediment is ordinarily dark colored because it contains much organic material. The unit is thickest along the Housatonic River, where it consists of 5 to 20 feet (1.5-6 m) or more of sand, silt, and gravel.

The Still River flood-plain deposits are 5 to 10 feet

(1.5-3 m) thick, and they are mostly sand and silt. Coarse gravel occurs along steep upland brooks.

Subsurface data (Melvin, 19 70) show that unit Qal over­ lies lake-bottom deposits of glacial Lake Danbury in parts of the Still River and Housatonic River valleys. The surface gradient of the unit is gentle, especially in contrast to some of the outwash units (Plate II).

Swamp deposits

Unit Qs includes peat, silt, and sand accumulated in swampy, poorly drained areas. It is one of the youngest surficial units in the Danbury-New Milford area, and it is still forming at the present time. The thickness of the unit is generally less than 15 feet (4.5 m) but is locally much greater. One test hole log (Melvin, 19 70) records 61 feet 86

(18.5 m) of "peat" southeast of Danbury Airport (in the

Bethel quadrangle).

The largest swamps in the study area developed on the former beds of glacial Lake Danbury and glacial Lake Kenosia.

Mill Plain Swamp (SW) and the swamps along Sympaug Brook and

Limekiln Brook (SC-SE) are prominent examples. Deposits of unit Qs are also common in parts of the upland where drainage is impeded by till deposits and irregular bedrock topography.

Two major upland swamps are located in the mountainous region between Sherman and New Milford (WC-C). Some swamps of the latter variety have formed in large kettle holes in till.

This was the probable origin of Bound Swamp, near the east border of the Danbury quadrangle (EC). Such swamps are favorable sites in which to conduct future pollen studies or look for material for radiocarbon dating.

Two fossiliferous bog marls are also included in unit Qs.

One of them is located along Dibbles Brook, just southeast of the cemetery on Walnut Hill Road (SE). Pawloski (196 7) originally described this locality, and Thompson (19 71) com­ mented on its fresh-water mollusks. A temporary excavation in the bog revealed about 15 feet (4.5 m) of soil, muck, and peat that overlay 15 feet (4.5 m) of calcareous bog sediment

(marl) (Pawloski, 1972, personal communication). Pawloski collected wood and bone samples from the marl, but they are probably too contaminated by carbonate minerals for radio­ carbon dating. Shells and the remains of plants and insects 87 are abundant in spoil piles from the excavation, but the mollusk species have little climatic significance (Thompson,

1971).

Cooper (1930) described a similar marl bog from the

New Milford area. The exact locality is unknown, and the bog may have been destroyed. It once was somewhere in the Park

Lane district of New Milford (NE).

Coriuvium

There are two types of colluvium in the study area.

One variety consists of taluses at the bottom of cliffs.

Taluses are most common in the steep, mountainous area around

Lake Candlewood. Large blocks of rock have piled up locally to such a degree that the land is rendered unsuitable for building or agriculture (though summer cottages are in fact built in such areas). The talus blocks are covered with un­ disturbed moss, trees, and other vegetation. They are probably not accumulating as rapidly as in late-glacial time, when frost action would have been greater.

The other type of colluvium is a thin mantle of dis­ turbed material on steep hillsides. It is very poorly sorted, and it was derived from the underlying glacial deposits (with a contribution from bedrock in places). Gravity and frost heaving are the principal agents that formed this colluvium.

It is typically less than 5 feet (1.5 m) thick, and much of it has been weathered during soil formation. Figure 6 shows 88 weathered colluvium overlying till on a steep hillside.

Downslope movement may still be occurring, but it is probably slower than in late-glacial time. CHAPTER VI

GLACIAL HISTORY

Pre-Wisconsinan events

The geography of the Danbury-New Milford area at the start of the Pleistocene Epoch was probably similar to its modern aspect. Glacial scouring removed a lot of bedrock along the Inwood Marble belt, but the Still River and

Housatonic River did not experience major drainage changes.

Thompson (19 71) presented evidence that the Housatonic River did not occupy the Still River valley in preglacial time.

The bedrock contours (Plate III) show that the floor of the

Still River valley has been scoured into a series of rises and depressions. However, the Croton Divide and Saugatuck

Divide are still at relatively high elevations. A minimum of 300 feet (92 m) of glacial overdeepening between

New Milford and Danbury would have been required to form the modern topography if the preglacial Housatonic River had flowed southward through the Still River valley and into the Saugatuck River or Croton River basins.

The pre-Wisconsinan regolith of the study area contained much rottenstone that started to form during or before the

Tertiary Period. Schafer (196 8) discovered as much as

89 25 feet (8 m) of rotten granite, schist, and gneiss in the nearby Waterbury, Connecticut, area. He noted that the over- lying glacial deposits are usually unweathered, and concluded that the bedrock disintegration predates the ice advance that deposited the lower till. In the Danbury and New

Milford quadrangles, remnants of the deep preglacial weather­ ing profile occur principally in the Inwood Marble; but there are a few disintegrated outcrops of other crystalline rocks. Rottenstone is also encountered in wells and test holes. It may be common on valley floors, where it is con­ cealed by glacial deposits.

The study area probably does not contain any glacial or interglacial sediments of pre-Wisconsinan age (assuming that the lower till is Wisconsinan). Any deposits laid down during this interval were removed by later ice advances. It is possible that buried remnants of pre-Wisconsinan deposits exist in the deep valleys, but the abundance of rotten marble in the upper till suggests that valleys along the marble belt were scoured to bedrock by the last glaciation.

Pre-Wisconsinan ice advances did occur in southern

New England, and they left deposits in coastal areas. Kaye

(1961) described four "drifts" from Boston, Massachusetts, the oldest of which was believed to be Nebraskan or Kansan.

Kay (196 4) also re-examined the Pleistocene stratigraphy of

Martha's Vineyard. He described a series of tills that sup­ posedly represent all four Pleistocene glacial stages. The 91 older tills are more deeply weathered and contain higher percentages of coastal plain sedimentary rocks than the younger tills. The extent to which early Pleistocene glaci­ ations affected or covered the Danbury-New Milford area is not known.

Early Wisconsinan glaciation

It is the consensus of those who have worked in southern

New England that the lower till was deposited by an Early

Wisconsinan glaciation (Flint, 19 71). The present author agrees on the basis of evidence presented in Chapter III.

Pessl's (19 71) till fabric study showed that the ice that deposited the lower till in western Connecticut advanced to­ ward the south-southeast. Flint (1961) determined the same ice advance direction for the Lake Chamberlin till, which may be correlative with the lower till. The Early Wisconsinan glacier deposited1 a thick layer of lower till in the Danbury and New Milford quadrangles. Many drumlins were built, and part of the regional topography was smoothed by till ac­ cumulation.

The recession of the Early Wisconsinan ice sheet un­ doubtedly resulted in the deposition of ablation till and outwash. With the exception of a small amount of possible ablation till, these materials are not exposed in the study area. Buried remnants may exist in valleys, but they would be difficult to identify from well and test hole records. 92 Elsewhere, Early Wisconsinan glacial ablation deposits were removed by the more recent ice advance.

Interglacial weathering in Middle Wisconsinan time caused the development of deep oxidation and iron-oxide staining in the lower till. The oxidized zone appears to be part of a C-horizon that was included in the Middle Wisconsinan weathering profile. All of the weathered zone except part of the C-horizon was removed by later ice erosion.

Late Wisconsinan glaciation

The Late Wisconsinan ice sheet advanced across New

England at about 20,000 yrs B.P. (Schafer and Hartshorn,

1965; Borns, 1973). Sirkin's radiocarbon dates on

"Farmdalian" sediments (Chapter III) indicate that the ice sheet had not deposited the Roslyn Till on western Long Island by 22,000 yrs B.P.

Figure 22 shows the directions of bedrock striations that resulted from Late Wisconsinan glaciation of the study area. The striations indicate an average ice-movement di­ rection of 160°. This direction is close to the average azimuth of 155° for lower-till drumlin axes (Figure 22).

Perhaps the Early and Late Wisconsinan glaciers moved in essentially the same direction, or the drumlins may have been slightly re-oriented by the Late Wisconsinan ice. It is believed that the drumlins were at least trimmed by the last glaciation. There is no upper till on their crests, and the no. of striations o 2 i— i i

N

S

0 8 Uau-i nl no. of drumlins

Figure 22. Rose diagram showing azimuths of 19 bedrock striations (upper half of diagram) and 45 drumlins (lower half). Data grouped in 10° classes. 94 upper till on the flanks of some drumlins has much included lower till.

The Late Wisconsinan glacier eroded much bedrock, expecially in the Still River valley. However, a prominent bedrock high was left at Brookfield (Plate III), and the ice sculptured roches moutonnees on the Inwood Marble in this area (Figure 23). The marble was resistant to erosion in places because of its compositional variability.

Marble and other rock types were incorporated into the upper till. Large boulders of resistant rock such as the

Danbury Augen Granite were able to survive glacial transport.

Local concentrations of these boulders dot the hillsides throughout the study area (Plate I). The boulders were plucked from the steep lee sides of bedrock hills and came to rest on stoss slopes a short distance downglacier.

The basal facies of the upper till could have been . deposited during both the advance and retreat of the Late

Wisconsinan glacier. However, a recent study in Glacier Bay,

Alaska (Mickelson, 19 73) proved that basal till deposition may occur only during the recession of an ice sheet. The ablation facies in the Danbury-New Milford area definitely formed during the melting of the ice. The upper till was deposited in both valleys and upland areas. Buried upper till is inferred to exist on valley floors on the basis of textural descriptions in Melvin's (19 70) subsurface data. 95

Figure 23. Roche moutonnee developed on Inwood Marble northwest of Brookfield village, on south side of North Mountain Road (NC). 96

Recession of the Late Wisconsinan glacier

The time of final ice retreat has not been determined for the Danbury-New Milford region. However, it can be ap­ proximated from radiocarbon dates that were obtained from nearby localities. Sirkin's (1967) bog study showed that the deglaciation of western Long Island had begun by 17,000 yrs

B.P. Davis (1969) determined the sedimentation rate at

Rogers Lake in southern Connecticut, and she concluded that

the ice was gone from the Connecticut coastline by 14,300 yrs

B.P. The study area was presumably deglaciated prior to the

Middletown readvance in the Connecticut Valley, which

Connally and Sirkin (1973) correlated with the Rosendale re­ advance in the Wallkill Valley of southeastern New York

(20 miles (32 km) north of the Wailkill Moraine). On the basis of Leopold's (1956) pollen study in central Connecticut,

Flint (1956) claimed that the Middletown readvance occurred not long before 13,000 yrs B.P. This readvance could actual­ ly be as old as 15,000 to 16,000 yrs B.P. according to

Borns (19 73). Connally and Sirkin (19 73) favor an inter­ mediate age of 14,000 to 14,800 yrs B.P.

It was noted in Chapter III that the upper till in the

study area is no younger than the Pellets Island or Wallkill

Moraines in New York. This conclusion is based on an east­ ward extrapolation of the margin of the Hudson-Champlain

Lobe from the positions of these moraines. Therefore, the

final deglaciation of the Danbury-New Milford region is 97 tentatively dated at 15,000 to 15,500 yrs B.P. by correlation with the Wallkill Valley chronology (Connally and Sirkin,

1970, 1973).

Deposition of stratified drift during deglaciation

The stagnant-ice margin had an irregular shape as it retreated from the study area, but its general trend was from west-southwest to east-northeast (perpendicular to the ice- movement direction). The southeast corner of the Danbury quad­ rangle was deglaciated first. Glacial streams issued from the meltwater channel west of King Lake and deposited unit

Qgb. The unit was graded to a bedrock-floored gap on a hill

0.1 mile (0.16 km) south of the quadrangle border. However, the meltwater must have ultimately flowed southward across the Saugatuck Divide (Bethel quadrangle) and into the

Saugatuck River. Ice margin positions associated with the deposition of unit Qgb and other units are shown on Plate I.

Further ice retreat resulted in the deposition of glacial-stream deposits (unit Qgs) in the vicinity of Limekiln

Brook and Sympaug Brook (SC-SE). Unit Qgs was probably graded directly to the Saugatuck Divide. It lies on a pro­ file line — along with kames in the Bethel quadrangle — that intercepts the divide. However, the slope of this line is less than the gradient of other meltwater deposits in the same area (Plate II). It is possible that open water formed north of the divide and was the base-level control for the

Qgs deposits. If this is true, they should be grouped with 98 the earliest glacial Lake Danbury sediments. A detailed study of the meltwater deposits in the Bethel quadrangle is needed to solve this problem. In either case, much stagnant ice remained between the main ice margin and the Saugatuck

Divide when unit Qgs was deposited. The sites of residual ice masses are now occupied by younger glaciolacustrine sediments and swamps.

The Saugatuck River stage of glacial Lake Danbury commenced when the ice margin was still in the southern part of the Danbury quadrangle. The dead ice disappeared from the area north of the Saugatuck Divide, and standing water was ponded in the vicinity of Bethel. The divide was the spill­ way for the lake during the Saugatuck River stage. Figure 24 shows the maximum extent of glacial Lake Danbury during this stage. Early deposits from the Saugatuck River stage (part of unit Qsr) occur in the town of Bethel (SE). They were graded to the lake when it had not yet expanded into the

Danbury quadrangle.

Unit Qsr-^ and associated lake-bottom deposits (part of unit Qsrb) were formed as the ice melted from the Limekiln

Brook and Dibbles Brook valleys (SE). Most of the Qsr^ out- wash was deposited between drumlins. It is probably contem­ porary with the kame delta (?) southwest of Plumtrees School.

Meanwhile, unit Qsrb was deposited at successively lower levels on the lake bottom as the Limekiln Brook valley became free of ice. A meltwater stream deposit (unit Qgp^) was i Figure 24, Maximum extent of glacial Lake Danbury during the Saugatuck River stage. Arrow indicates spillway at Saugatuck Divide. Barbed line and hachures represent approximate position of stagnant-ice margin and boundary of river lowland, respectively. 100 built at the same time as unit Qsr-^ when the ice margin reached the position shown for these units on Plate I. The

Qgp^ sediments were laid down in the head of the Pond Brook valley (EC). They were graded to a meltwater drainage level in the Housatonic River valley to the east.

Some of the undifferentiated Saugatuck River stage deposits in Danbury (unit Qsr) may also be correlative with unit Qsr]_. These sediments were washed into glacial Lake

Danbury from poorly defined ice margin positions in various parts of the city. Lake-bottom sediments of unit Qsrb were also laid down in this area. Some of the Qsr deposits came from a tongue of ice that lay to the west in the Mill Plain

Swamp basin (SW).

A younger Saugatuck River stage deposit (unit Qsr2) was formed when ice recession in the east-central part of the

Danbury quadrangle opened up a drainage path to the southwest along Stony Hill Brook (EC-SE). Meltwater was now able to follow this lower and more direct route to glacial Lake

Danbury. Thinning of ice in the Pond Brook valley resulted in the deposition of unit Qgp2 at the same time. The Pond

Brook outwash continued to be carried eastward into the

Housatonic River valley. Lake-bottom sediments (unit Qsrb) likewise continued to accumulate during Qsr2 time.

The Qsr deposits near the south end of Lake Candlewood are probably contemporary with (or slightly younger than) unit Qsr2 . The stagnant-ice margin stood in the Danbury Bay 101 area (C-WC), and meltwater streams built the deltas at

Germantown and Wooster Cemetery (SC). The Qsr kame terraces are evidence that much residual ice existed between the main ice margin and glacial Lake Danbury. Cobble and boulder gravel was deposited adjacent to the ice front, while sand and gravel were laid down farther south. The prominent meltwater channel west of Beaver Brook Mountain (C) also formed at this time. A meltwater stream truncated the drumlin east of Stadley Rough Road and deposited mounds of coarse gravel (of unit Qsr) to the south. Ice must have continued to block the Still River valley east of Beaver Brook Mountain throughout the Saugatuck River stage. Otherwise, glacial

Lake Danbury would have escaped across the Pond Brook Divide and fallen to its next level.

Glacial Lake Kenosia originated when the ice margin re­ treated westward from the Mill Plain Swamp basin (SW). This event followed the deposition of the Qsr sediments east of

Mill Plain Swamp, so it is probably contemporary with the late Saugatuck River stage of glacial Lake Danbury. The water from glacial Lake Kenosia flowed eastward’ across till and outwash to glacial Lake Danbury. Only lake-bottom sedi­ ments (unit Qkb) were deposited at first. Then the glacio- fluvial deposits (unit Qk) were graded to the lake as the ice margin withdrew into the Brewster quadrangle. There was thick sediment accumulation in glacial Lake Kenosia -- unit

Qkb is as much as 140 feet (43 m) thick on the south side of 102

Danbury Airport (in the Bethel quadrangle). The lake nearly filled with silt, sand, and clay. It was destroyed eventually when its spillway was eroded by the outflow of water to the east. Deposition of glacial Lake Kenosia sediments ceased when the ice front had retreated far enough west for the meltwater to escape southward into the Croton River system in New York.

A third glacial-stream unit (Qgp3) was laid down in the east-central part of the Danbury quadrangle following the deposition of units Qgp3 an<3 Qs r 2 • T*ie stagnant-ice margin was in the East Brook valley (southeast of Lake Candlewood), and the meltwater streams that deposited unit QgP3 were con­ trolled by the Pond Brook Divide. Glaciofluvial sediments were deposited on the west side of the Still River valley by meltwater that flowed eastward across the divide.

The ice in the Still River valley then thinned to the point where glacial Lake Danbury escaped across the Pond Brook

Divide. The lake level dropped to an elevation of about

380 feet, and the Pond Brook stage commenced at this time

(Figure 25). The lake water drained across the' divide and flowed down the Pond Brook valley to the Housatonic River.

The ice margin was still close to the south end of the Rocky

River (Lake Candlewood) valley during the earliest part of the Pond Brook stage. Meltwater streams in the Beaver Brook area graded the sand and gravel of unit Qpb-^ to glacial Figure 25. Maximum extent of glacial Lake Danbury during the Pond Brook stage. Arrow indicates spillway at Pond Brook Divide. Barbed line and hachures rep­ resent approximate position of stagnant-ice margin and boundary of river lowland, respectively. 104

Lake Danbury. Fine-grained sediments (unit Qpbb) were de­ posited simultaneously on the lake bottom.

Further ice retreat uncovered the Rocky River valley, and glacial Lake Candlewood was established. The surface elevation of this lake was about 4 30 feet. Its water level was controlled by a spillway on the divide (now flooded) between the Still River and Rocky River basins. Drainage of the lake probably carved the scarp in unit Qpb^ northwest of

Great Plain School (C). Sand and gravel (unit Qc) was graded to glacial Lake Candlewood as the ice front receded north­ ward into the New Milford quadrangle. Lake-bottom deposits must have formed, but they are concealed by the modern lake.

Pond Brook stage deposits (units Qpb2 and Qpbb) were laid down in the Still River valley between Brookfield and

Lanesville at about the same time that unit Qc was being de­ posited in glacial Lake Candlewood. Ice recession in the

Still River valley was rapid during the Pond Brook stage.

None of the Qpb2 deltas were built up to lake level, and the outwash stratigraphy suggests a quick transition between glaciofluvial and glaciolacustrine environments (Chapter IV).

Ice retreat in both the Rocky River and Still River valleys was facilitated by calving of icebergs into the proglacial lakes. Large ice-rafted boulders are encountered in the fine­ grained lake-bottom sediments (unit Qpbb) between Danbury and Lanesville. 105

When the stagnant-ice margin reached Lanesville (SC), the water in glacial Lake Danbury began to flow across the col at the north end of Pumpkin Hill (SE) and spilled into the

Housatonic Gorge. This event lowered the lake level to an elevation of about 29 5 feet, and it marked the beginning of the Pumpkin Hill stage of glacial Lake Danbury (Figure 26).

The large delta at Lanesville was built during this stage in the lake's history. While the stagnant-ice margin lay against the north side of the delta, the active-ice margin withdrew northward almost to New Milford. The intervening stagnant zone was initially the site of esker formation. Subglacial or englacial streams carried sand and gravel (part of unit

Qph) from the active-ice source to the Lanesville delta.

Further melting of the stagnant zone exposed the bedrock high north of Lanesville, and the remainder of unit Qph was deposited subaerially. The glacial meltwater also deposited a thick group of bottomset beds (unit Qphb) upon entering the lake at the Lanesville delta.

An unusual characteristic of the Pumpkin Hill stage of glacial Lake Danbury was the proximity of the ice margin to the spillway. A tongue of stagnant ice must have blocked the north end of the Housatonic Gorge during this stage. Other­ wise, the meltwater would have flowed directly into the north end of the gorge at a lower level than the Pumpkin Hill spill­ way. Some of the Qph sediments may in fact have been depos­ ited during a transition to this later drainage path. Figure 26. Extent of glacial Lake Danbury during the Pumpkin Hill stage. Arrow indicates spillway across Pumpkin Hill. Barbed line and hachures represent approximate position of stagnant-ice margin and boundary of river lowland, respec­ tively. 107

Glacial Lake Danbury did not extend south of Brookfield during the Pumpkin Hill Stage. The postglacial Still River

had begun to develop in Danbury during the Pond Brook stage,

and it was now able to flow northward across the newly ex­

posed meltwater deposits between Danbury and Brookfield. The

valley was poorly drained at first, and water may have been

ponded in places. However, the Still River began to erode

the earlier glacial deposits as it flowed on what are now

stream-terrace surfaces.

Further deposition of glacial-stream sediments (unit

Qgt) occurred contemporaneously with the Pumpkin Hill stage

of glacial Lake Danbury. The stagnant-ice margin was near

the north end of Clatter Valley (EC), but masses of dead ice

probably lingered in the center of the valley. Sand and

gravel was deposited on either side of the residual ice as

meltwater flowed southward into the Housatonic River. Unit

Qgt was graded to a hillside threshold northwest of Wolf

Pit Mountain.

Glacial Lake Danbury dropped to its lowest level when

the stagnant ice melted back from the north end of the

Housatonic Gorge. This was the beginning of the Housatonic

River stage, during which the lake drained directly through

the gorge (Figure 27). The disappearance of ice from the

Housatonic Valley opened up a basin in which lake-bottom

sediments (unit Qhrb) were deposited. The Qhrb deposits in

the lowland east of Fort Mountain (C) are higher than those MILFORD

lda n b u r y

Figure 27. Maximum extent of glacial Lake Danbury during the Housatonic River stage. Arrow indicates spill­ way at Housatonic Gorge. Stagnant-ice margin was north of map area. Hachures represent boundary of river lowland. 109 in the main Housatonic Valley to the east. They may have been laid down in a separate water body that was contained be­ tween the ice front and the Pumpkin Hill stage deposits

(Qph) north of Lanesville.

As the stagnant-ice margin retreated from New Milford, standing water of the Housatonic River stage expanded from the Housatonic Gorge to the vicinity of Wannuppee Islands.

The constriction in the valley west of Wannuppee Islands was choked with stagnant ice, and it dammed the glaciofluvial sediments (unit Qhr) that were carried toward the lake. This damming effect may explain the change in slope of unit Qhr on Plate II. The lake basin eventually filled with sediment, and there was a transition from glaciolacustrine to glacio- fluvial environments. Sand and gravel outwash of unit Qhr prograded across the lake-bottom sediments as this transition occurred. The high-level part of unit Qhr along the Aspectuck

River branches (NC-NE) was also deposited at this time.

A second body of standing water developed in the

Housatonic River valley when the ice disappeared from the area northwest of Wannuppee Islands (NC). The distribution of glaciolacustrine sediments proves that it extended at least as far north as Boardman School. This water body is con­ sidered to have been the final phase of glacial Lake Danbury.

However, the part of the lake that had existed southeast of

Wannuppee Islands was filled with sediment. These deposits in the New Milford area (units Qhr and the underlying Qhrb) 110 dammed the meltwater that now existed in the vicinity of

Boardman Bridge. Downcutting of this drift barrier subse­ quently lowered the base level for the Aspetuck River branches, and the lower level of unit Qhr formed in their valleys. The latter surface lies at elevations of 260 to 300 feet. It is partly depositional and partly the result of terracing of earlier Qhr deposits.

Additional glacial-stream deposits (unit Qgg) were

formed during the early part of the Housatonic River stage of glacial Lake Danbury. The ice margin stood in a series of closely spaced positions as the glacier thinned and retreated

from the northeast corner of the New Milford quadrangle.

These positions are now marked by a set of meltwater channels on the till upland near Hickory Haven (NE). Meltwater flowed down the steep upper end of the Great Brook valley and began to drop its sediment load where the gradient was sufficiently low. Unit Qgg was deposited as a series of kame terraces be­ tween the till upland and residual ice in the center of the valley. The unit was probably graded to a temporary base- level control on the ice itself (Chapter IV). The meltwater then emptied into glacial Lake Danbury in the vicinity of

New Milford.

Sedimentation probably continued in glacial Lake

Candlewood during the Pumpkin Hill and early Housatonic

River stages of glacial Lake Danbury. Harvey's (1920) well

logs indicate the presence of lake-bottom deposits in the Ill lower part of the Rocky River valley (C), but they are now hidden by the northeast arm of Lake Candlewood. Glacial

Lake Candlewood emptied into the Housatonic River valley when the ice disappeared from the area between Candlewood Mountain and Guarding Mountain (C). This happened sometime during the Housatonic River stage. The postglacial Rocky River was established following the disappearance of glacial Lake

Candlewood.

Unit Qgsh also formed during the Housatonic River stage.

It was deposited when the ice margin was north of Sherman

Church (iJW) . Meltwater streams built unit Qgsh as they flowed southward into the Rocky River valley. The valley floor was the base-level control for these streams.

Ice retreat from north of Sherman Church resulted in the deposition of unit Qgm. The ice margin was now in the northwest corner of the New Milford quadrangle. Meltwater flowed northeast along the ice front and deposited unit Qgm as a kame terrace in the Morrissey Brook valley. The glacial stream entered the Housatonic River valley in the Kent quad­ rangle, where stagnant ice was probably its base-level con­ trol .

Unit Qgnm is the youngest meltwater deposit in the study area. It was deposited when the ice margin was in the

Kent quadrangle and little or no stagnant ice remained in the New Milford area. Valley train outwash was laid down on the bottom of the Housatonic River valley at least as far 112 south as the mouth of the Rocky River. Unit Qgnm prograded across glaciolacustrine sediments that had been deposited in the last remnant of glacial Lake Danbury. Terracing of

Housatonic River stage deposits (unit Qhr) probably began farther downstream as the postglacial Housatonic River cut into the Qhr outwash in the area between Wannuppee Islands and the Housatonic Gorge. CHAPTER VII

LATE-GLACIAL AND POSTGLACIAL HISTORY

Late-glacial events

Deglaciation of the Danbury-New Milford area probably was complete by around 15,000 yrs B.P. — the estimated age of the Wallkill Moraine in southeastern New York (Connally and Sirkin, 1973). Disappearance of the Late Wisconsinan ice sheet was accompanied and followed by widespread depo­ sition of eolian sand and silt. The thickest eolian mantle formed on meltwater deposits in valley areas. The prevailing wind direction in late-glacial time is not known. The thick eolian sand on the Lanesville delta occurs on the east side of a high ridge. This may be due to a wind shadow effect, indicating that the wind came from the west. Hartshorn

(1962) reported that late-glacial winds in Massachusetts came from directions varying from north-northeast to west-northwest.

Wind erosion was curtailed within a few decades by the northward migration of vegetation. Pollen studies in southern

New England and nearby parts of New York and New Jersey have shown that tundra vegetation quickly moved into ice- free areas. Sirkin (1967) discovered late-glacial pollen zones that formed in southern Long Island bogs while the

113 114 glacier was still present on the north side of the island.

Leopold (1956) first discovered a late-glacial, non- arboreal pollen zone in southern Connecticut and named it

"zone T". This zone is dominated by herbaceous pollen, in­ dicating a tundra climate. Later reports on the pollen stratigraphy of Connecticut bogs and lakes were published by Deevey (1958) , Beetham and Niering (1961) , and Davis

(1969). Connally and Sirkin (1970) described a pollen se­ quence from the Wallkill Valley in southeastern New York, and

Sirkin and Minard (19 72) examined a similar occurrence in northwestern New Jersey. All of these workers encountered the zone-T pollen assemblage, and they considered it to be representative of a tundra or park-tundra environment.

Davis (1969) studied the pollen in cores from Rogers Lake,

5 miles (8 km) north of Long Island sound. She determined the sedimentation rate, from radiocarbon dates and concluded that a tundra environment existed in southern Connecticut during the period 14,300 - 12,150 yrs B.P.

One might expect to find periglacial features of late- glacial age in the study area. They have been reported from elsewhere in southern New England by Denny (19 36) , Birman

(1952), and Schafer (196 8). These investigators discovered involution zones, ice-wedge structures, and clastic dikes

(ice-vein replacements) developed in glacial deposits and weathered bedrock. No such features have been discovered yet in the Danbury and New Milford quadrangles, but there 115 are certain deposits that probably formed mainly in a peri- glacial environment. One of these is the talus that occurs at the bottom of cliffs. The taluses are partly covered by vegetation, and they are not enlarging noticeably at the present time. The same is true of colluvium on hillsides.

It is likely that both of these deposits formed more rapidly in the periglacial environment.

An unusual weathering phenomenon in the Inwood Marble may also be of periglacial origin. Many outcrops of rotten- stone marble have "silted joints". These are thin, dark-brown discolored zones that appear to follow a fracture pattern in the rock. Figure 2 shows a good example located between

U. S. Route 7 and Grays Bridge Road (EC). The "silted joints" at this locality are not due to compositional variation in the marble, though they are commonly parallel to the rock foliation. One would expect a discontinuity in the bedrock if they were clastic dikes, but this is not the case. The carbonate grains in the discolored zones are coated by a thin iron-oxide stain and very fine sand to silt-size mineral grains of extraneous origin. It was learned from Chittick analyses that this contaminated rottenstone contains 82 per­ cent total carbonate, of which 27 percent is calcite. Clean rottenstone from the same outcrop contains 99 percent total carbonate and 50 percent calcite. Thus the marble along the discolored zones is dirtier and more leached than the sur­ rounding rottenstone. The depth to which these zones extend 116

is not known. They dip at various angles and can be traced downward as far as it is possible to dig (usually 1 or 2

feet).

Formation of the "silted joints" probably began with the development of fractures in the Inwood Marble when it was still solid. Ground-water circulation along these joints disintegrated the carbonate grains to a greater degree than in the adjacent marble, but without the development of an open space. Repeated freezing and thawing of ground water in late-glacial and postglacial time enabled sediment from the overlying glaciofluvial unit to migrate downward into the weathered marble seams. The extent to which this process was late-glacial versus postglacial has not been determined.

Stream-terrace formation began immediately after ice retreat from the Danbury-New Milford area. The Still River re-established itself as glacial Lake Danbury withdrew from the valley. The river followed its preglacial course except near the village of Brookfield. Kame deposits of unit Qpb blocked the preglacial channel west of Brookfield (Figure 28).

They forced the Still River to the east side of' the valley, where it now flows on bedrock. The river gradient was slightly steeper than at present because significant postglacial up­ lift had not occurred. The hills at Brookfield may have dammed the Still River at first, but it eventually cut into the glacial deposits to the south and left the terrace sur­

faces between Brookfield and Danbury. Other terraces began 117

Figure 28. Kame complex (unit Qpb^) in site of former river channel on west side of Still River valley, southwest of Brookfield village (NE). Edge of kettle hole is visible in foreground. Note preferential growth of juniper trees on kame surface. 118

to form in late-glacial time along the Housatonic River,

East Aspetuck River, and West Aspetuck River following the shoaling of glacial Lake Danbury.

Postglacial events

As the postglacial climate warmed in southern New

England, tundra grassland was successively followed by open spruce forest (recorded by pollen zone A), pine forest

(zone B), and oak forest (zone C). This vegetation sequence has been demonstrated by all of the previously cited pollen studies. Hence, there is no reason to doubt that it also occurred in the Danbury-New Milford area. Davis (1969) gave the following limits for the pollen zones at Rogers Lake:

zone A 12,150 - 9,100 yrs B.P.

zone B 9,100 - 7,900 yrs B.P.

zone C 7,900 - present

The most recent surficial deposits in the study area are swamp sediments and flood-plain alluvium, which are still accumulating* Pollen studies indicate that swamp deposits began to form soon after deglaciation (Sirkin, 1967). Flood- plain deposits have likewise formed over much of postglacial time, but they postdate any adjacent stream terraces. Recent flood records for the Still River and other streams in the study area have been compiled by the U. S. Geological Survey

(Cervione and others, 19 72).

Holocene weathering processes have acted on all surfi­ cial deposits to form soil profiles. Most soils in Litchfield 119

County {which encompasses ahout one-third of the study area) belong to the Spodosol or Inceptisol orders according to the classification used by the U. S. Department of Agriculture

(Gonick and others, 1970). The soils that have formed in

glacial deposits are Spodosols. Most of them are brown pod-

zolic soils with well developed A-, B- and C-horizons. The

Inceptisols are the immature soils that are now developing in

the younger stream-terrace deposits and flood-plain alluvium.

Variations in parent material and relief (as well as time) have produced the wide variety of soil series and phases

described by Gonick and others.

Examination of a few Holocene soil samples from the

upper and lower tills revealed that postglacial weathering has affected the mineralogy of the tills in similar ways.

Biotite flakes in both B-horizons are weathered to a greater

degree than in the parent tills, and many of the flakes are

iron-oxide stained. Many of the feldspar grains have a

chalky luster and are very friable. Other minerals, such as hornblende and magnetite, have experienced little weathering

in postglacial time. Soil horizons in the two tills are also

similar with respect to clay mineralogy (Appendix C). X-ray

diffraction patterns show that the amount of vermiculite in

the B-horizons has increased at the expense of illite relative

to the C-horizons and parent tills.

Minor postglacial weathering has occurred on some Inwood

Marble' outcrops. Ground-water activity has formed pits on 120 ledge surfaces (Figure 29). These pits ,grow by disintegration of the marble rather than by total solution. The weathering process involves a combination of mechanical breakdown along grain boundaries and preferential solution of the marble's calcite component. Thus a small amount of the rottenstone marble in the study area has formed in this manner during postglacial time.

Indian artifacts and camp-site remains are evidence of prehistoric human habitation in the Danbury-New Milford area.

Dozens of archaeological sites have recently been discovered in the valleys of the Still River, Housatonic River, and other streams. Artifacts are found in shallow excavations on flood plains, stream terraces, and kames. A major Indian settlement existed on the north end of Pumpkin Hill, near the mouth of the Still River. The oldest radiocarbon date on this site is 3665 ± 120 yrs B.P. (J. A. Pawloski, 1972, personal communication).

Human activity did not significantly affect the geology of the study area until the twentieth century. Recent urban expansion in Danbury has resulted in modification of drainage patterns. Highway and building construction are causing much flood plain and swamp area to be covered by pavement.

The resultant loss of infiltration area is believed to be increasing the frequency and intensity of floods along the

Still River. The build-up of artificial fill is also re­ ducing the area into which flood waters can spread, which 121

Figure 29. Weathering pits on Inwood Marble outcrop on east side of U. S. Route 7, 0.7 mile south of junction with Route 13 3 (EC). 122 causes a further increase in flood magnitude.

Urbanization has spurred the rapid exploitation of sand and gravel resources in the study area. Borrow pits are an aid to surficial geologic mapping because they expose the stratigraphy of unconsolidated deposits. However, the re­ moval of these deposits destroys valuable indicators of the sequence of deglaciation. Whenever possible, it is important to examine morphologic sequence deposits before or during their excavation. CHAPTER VIII

SUMMARY

The study area for this report includes the Danbury and New Milford 7.5-minute quadrangles in western Connecticut.

It is a hilly region with many outcrops of Precambrian and

Paleozoic crystalline bedrock. The glaciated bedrock sur­ face is overlain by unconsolidated Quaternary deposits. These include till, glacial-meltwater deposits, wind-blown sediment, stream-terrace and flood-plain alluvium, swamp deposits, and colluvium. The surficial deposits were mapped and studied to help reconstruct the Quaternary history of the Danbury-

New Milford area.

There are two tills in the study area — the "upper till" and "lower till". The upper till is generally unoxi­ dized. It is light olive brown to light olive gray, sandy, and friable. It is also very stony and locally carbonate- rich. Inclusions of lower till and lenses of washed sediment are present in places. Sand and silt-size mineral grains are fresh (with the exception of biotite) and appear to have been derived from local rodk types. The clay minerals in the up­ per till are mostly illite and vermiculite.

123 124

The lower till (which forms the many drumlins in the study area) is olive brown to dark olive gray and deeply oxidized. It is silty, compact, and less stony than the upper till. Sub-horizontal and sub-vertical joints are com­ mon, and dark-brown iron-oxide staining coats the surfaces of both joints and stones. The lower till is similar to the upper till with respect to mineralogic composition. They contain the same clay minerals, and both tills have a car­ bonate-rich member. The most noticeable difference is the more intense weathering of biotite in the oxidized lower till.

The tills of the study area are believed to be cor­ relative with two tills that occur at many other localities in southern New England. Field evidence indicates that these tills are of different ages. It is known from radiocarbon dates from older and younger deposits (outside the study area) that a Late Wisconsinan glaciation deposited the upper till.

The age of the lower till is uncertain. It is probably an

Early Wisconsinan deposit because it is not much more weath­ ered than the upper till.

The glacier that deposited the upper till retreated from the Danbury-New Milford region about 15,000 years ago. De­ glaciation was accompanied by deposition of glaciofluvial and glaciolacustrine sediments in many valleys. These de­ posits were mapped as a series of chronologic units, shown on Plate I. Each unit contains meltwater deposits related 125 by age and depositional environment. The distribution of glacial lake and stream deposits indicates that the ice front retreated to the north. The active ice was fringed by a con­ tinuous stagnant zone, beyond which scattered masses of dead ice persisted locally. Karnes, eskers, and other glacial- stream deposits were laid down within or adjacent to the

stagnant ice. Non-ice-contact glaciofluvial deposits are

relatively uncommon in the Danbury and New Milford quadrangles.

Glacial lakes were prominent in the deglacial history of the study area. The largest was glacial Lake Danbury.

This water body occupied most of the Still River valley and part of the Housatonic River valley during its four stages.

Other glacial lakes existed in the Lake Candlewood valley and

the southwest corner of the Danbury quadrangle. These lakes were local base levels for many of the meltwater streams.

Following deglaciation of the study area, a thin cover

of eolian sand and silt was deposited until wind erosion was

curtailed by vegetation. Permafrost conditions may have

existed in late-glacial time and promoted the formation of

taluses and colluvium. Deposition of swamp sediments also began when the study area became ice-free. Rivers developed

in general conformity to their^preglacial drainage patterns

and started to cut terraces in the glacial-rneltwater deposits.

Pollen studies in southern New England indicate that

a tundra or park-tundra environment existed in the Danbury-

New Milford area in late-glacial time. Tundra grassland was 126 successively followed by spruce, pine, and oak forests during postglacial time. Soil profiles are forming in the Quaternary surficial deposits, and flood-plain and swamp sediments are still accumulating. Man has inhabited the study area for over 4,000 years, but human activity did not significantly modify the geologic environment until the twentieth century. APPENDIX A

LOGS OF WELLS AND TEST HOLES PLOTTED OH PLATE I

BT BT BT DY DY DY DY DY DY DY DY DY BD 2 7 2 9 38 35 37 69 75 7 5 79 83 90 118 8 TH TH TH TH TH TH TH TH la n d _ W W w W W surface V 7 7 2

EXPLANATION w well • • TH test hole • • X x X *0‘ • « • X X • ■ • • » fill • i I « I R • • * • • soil, muck, • • • peat » • oo oo o o o o gravel x * • AAA • • • sand and 80' A A « • gravel AAA A A fine to very AAA XXX coarse sand A A AAA xxx very fine A A sand silt and clay 120 ' A A A A A till XXX • • X X cored "bedrock t • a X XX • •

* • • 127 identification • • r uncertain |« « » 160' AAA A A R refusal RK bedrock

200 /L- APPENDIX A (cant.)

BD BD BD NMI N M t N M I NMI NMI NMI NMI NMI NMI NMI 9 16 20 1 2 6 7 13 14 15 16 16 17 TH W TH TH T H W T H TH W w TH W W la n d o o o o o ft • • WZL • 0 surf oce o o o o o ft • « o o a o o 0 0 0 o o_ • * o o OOP * • • OOP

4 0 ' -

• • « 4 • » • • O o o • • o o • • oO o£> o • • e O O 01 80' - RK

xxx x X 120 ' - XXX X X XXX X X XXX X X XXX X X XXX X. X 160' - XXX X- X XXX XJL 128

200 '1— APPENDIX A (cont.)

NMI NMI NMI NMI NMI 19 21 31 46 49 TH TH TH TH TH land 7 T 7 • » surfoce • • •

4 0 '

x x x XXXx x

80 '

A A

120 ' AAA A A AAA A A A A A XXX X X ix x x

160'>— 129 131

APPENDIX B (cont.)

NW-9 Upper till, pit on N side of hill, 0.1 mi W of Disbrow Pond and just S of Ball Pond Brook.

NW-10 Soil from C-horizon in upper till, 2.3 ft below surface, same pit as NW-9.

NW-11 Soil from B-horizon in upper till, 1.5 ft below surface, same pit as NW-9.

NW-12 Lower till, oxidized, same pit as NW-9.

WC-13 Lower till, oxidized, pit on W side of Padanaram Brook, 0.1 mi W of junction of Padanaram Road with Pembroke Road.

WC-14 Soil from C-horizon in lower till, 1.3 ft below surface, same pit as WC-13.

WC-15 Soil from B-horizon in lower till, 0.6 ft below surface, same pit as WC-13.

SC-16 Lower till, unoxidized, carbonate-rich, about 25 ft below surface, building excavation on S side of Newtown Road, 0.1 mi E of Still River bridge.

EC-17 Lower till, oxidized, shallow excavation on S side of Interstate Route 84, 0.4 mi E of Danbury-Bethel town line.

NE-18 Lower till, oxidized, ditch on N side of road, 0.3 mi S of Hickory Haven.

NE-19 Inwood Marble clasts, unleached zone of upper till, same pit as NE-2„

NE-20 Inwood Marble clasts, partly leached zone of upper till, same pit as NE-2. APPENDIX C

X-RAY DIFFRACTION DATA FOR CLAY FRACTIONS OF TILL SAMPLES

Table gives sets of basal d-spacings derived from x-ray dif­ fraction patterns for each sample. Data for glycolated sam­ ples are omitted because they are same as for unheated sam­ ples .

Sample Number Average d-spacings (Sngstroms)

NE-1 unheated 14.6 10.2 7.16 3.56 3.35

450°C 9.97 5 .02 3.34

550°C 10 .0 5.04 3.33

NE-3 unheated 14.6 10.2 8.56 7.19 5 .00 4 .80 3.59 3.34 3.25 3.19 3.12 2.88

450°C 14.2 9.98 8.42 .7.13 4.97 4.79 3.58 3.32 3.19 2.88

550°C 13 .9 10.1 8.50 5.03 3.35 3.25 3.20 3.12

NE-4 unheated 14.5 10.0 8.44 7.12 ■ 5.00 4.78 3.57 3.34 3.25 3.18 3.12 2.88

450°C 14.2 10.0 8.42 4.96 3.57 3.32 3.24 3.18 3.12 2.88

550°C 13.9 10.1 8.46 5.03 3.35 3.25 3.19 3.11

132 133

APPENDIX C (cont.)

NE-5 unheated 14.5 12.3 9.98 8.40 7.13 4.97 4.75 3.56 3. 32 3.25 3.18 3.12 2 .85

450°C 13.6 9.93 8.42 4.99 3.32 3.24 3.18 3.11

550°C 13.9 10.1 5 .02 3.35 3.25 3.20 3.12

C-7 unheated 14.5 11.7 10 .2 7.12 5.00 4.73 3.55 3.35 3.20

450°C 14.5 10.0 7.13 4.99 3.33

550°C 13.9 9.98 5 .02 3.35

C-8 unheated 14.6 12.4 10.1 7.15 5.02 3.55 3.35

450°C 14.0 9.98 7.13 5.01 3. 34

550°C 14.0 10.0 5.03 3.35

NW-9 unheated 14.5 11.9 10 .2 7.21 5.01 4.87 3.58 3.55 3.36

450°C 14.5 10.0 7.13 5 .00 3.33

550°C 14.0 10.0 5 .04 3.32

NW-10 unheated 14.2 11.9 10 .1 7.13 4.98 4.83 3.53 3.36

450°C 14.0 10.0 4.98 3.32

550°C ,13.8 9.98 5.01 3.31

NW-11 unheated 14.2 10.1 7.12 4. 84 3.54 3.34

450°C 10 . 3 4.98 3.34

550°C 9 .98 5.02 3.34 134

APPENDIX C (cont.)

NW-12 unheated 14.7 10.2 7.25 5.03 3.59 3 .37

450°C 10.0 5.01 3.32

550°C 9.98 5.03 3.32

WC-13 unheated 14.6 10.2 7.22 5.02 3.57 3.37

450°C 9.98 5.00 3.33

550°C 10.0 5.04 3.33

WC-14 unheated 14.0 11.9 9.98 7.08 4.98 4.71 3.53 3.32

450°C 14.0 9.9 8 4.98 3.32

550°C 13.8 9.98 5.02 3.35

WC-15 unheated 14.0 12.1 9.98 7.09 4.99 4.72 3.53 3.32

450°C 13.8 9.98 7.13 4.98 3.34

550°C 13.8 9.93 5 .02 3.34

SC-16 unheated 14.7 10.2 7.18 5.03 4.75 3.59 3.55 3.37 2.89

450°C 14.1 10.0 7.13 5 .01 3.33

550°C 13.9 10.1 5.03 3.32

EC-17 unheated 14.4 10.1 8.50 7.13 4.99 4.74 3.57 3.34 3.24 3.19

450°C 10.0 4.99 3.33 3.20

550°C 9.98 5.03 3.34 3.19

NE-18 unheated 14.6 11.9 10.2 7.19 5.02 3.59 3.35

450°C 14.4 10 .1 7.13 4.99 3.33

550°C 14.1 10.0 5.04 3. 33 REFERENCES CITED

Beetham, Nellie, and Niering, W. A., 1961, A pollen diagram from southeastern Connecticut: Am. Jour. Sci., v. 259, p. 69-75.

Birman, J. H., 1952, Pleistocene clastic dikes in weathered granite-gneiss, Rhode Island: Am. Jour. Sci., v. 250, p. 721-734.

Borns, H. W. , Jr., 1973, Late Wisconsin fluctuations of the Laurentide ice sheet in southern and eastern New England, in The Wisconsinan Stage: Geol. Soc. America Mem. 136, p7 37-45.

Carroll, Dorothy, 19 70, Clay minerals: a guide to their x-ray identification: Geol. Soc. America Spec. Paper 126, 80 p .

Cervione, M.A., Jr., Mazzaferro, D. L., and Melvin, R. L., 1972, Upper Housatonic River basin, pt. 6 o£ Water resources inventory of Connecticut: Connecticut Water Resources Bull. 21, 84 p.

Clarke, J. W . , 1958, The bedrock geology of the Danbury quadrangle: Connecticut Geol. and Nat. History Survey Quad. Rept. 7, 47 p.

Connally, G. G., and Sirkin, L. A., 1970, Late glacial history of the upper Wallkill Valley, New York: Geol. Soc. America Bull., v. 81, p. 3297-3306.

------1973, Wisconsinan history of the Hudsori-Champlain Lobe, in The Wisconsinan Stage: Geol. Soc. America Mem. 136, p. 47-69.

Cooper,.G. A., 1930, Fossil fauna of the marl deposits in the vicinity of New Milford, iri Flint, R. F., The glacial geology of Connecticut: Connecticut Geol. and Nat. History Survey Bull. 47, p. 238-259.

135 136

Currier, L. W. , 1941a, Disappearance of the last ice sheet in Massachusetts by stagnation zone retreat (abs.): Geol. Soc. America Bull., v. 52, p. 1895.

------1941b, Tills of eastern Massachusetts (abs.): Geol. Soc. America Bull., v. 52, p. 1895-1896.

Davis, M. B., 1969, Climatic changes in southern Connecticut recorded by pollen deposition at Rogers Lake: Ecology, v. 50, p. 409-422.

Deevey, E. S., Jr., 1958, Radiocarbon-dated pollen sequences in eastern North America: Veroff. Geobot. Inst. Riibel, v. 34, p. 30-37.

Denny, C. S., 19 36, Periglacial phenomena in southern Connecticut: Am. Jour. Sci., 5th ser., v. 32, p. 322-342.

Drake, L. D., 1971, Evidence for ablation and basal till in east-central New Hampshire, in Till — a symposium: Columbus, Ohio State Univ. Press, p. 73-91.

Dreimanis, Aleksis, 1962, Quantitative gasometric determination of calcite and dolomite by using Chittick apparatus: Jour. Sed. Petrology, v. 32, p. 520-529.

Dreimanis, Aleksis, and Vagners, U. J., 1971, Bimodal distri­ bution of rock and mineral fragments in basal tills, in Till — a symposium: Columbus, Ohio State Univ. Press, p. 237-250.

Flint, R. F., 1956, New radiocarbon dates and late-Pleistocene stratigraphy: Am. Jour. Sci., v. 254, p. 265-287.

------1961, Two tills in southern Connecticut: Geol. Soc. America Bull., v. 72, p. 16 87-1692.

------19 71, Glacial and Quaternary geology: New York, McGraw-Hill, 892 p.

Gonick, W. N., Shearin, A. E., and Hill, D. E., 19 70, Soil survey of Litchfield County, Connecticut: Washington, U. S. Govt. Printing Office, 105 p.

Grim, R. E., 1968, Clay mineralogy: New York, McGraw-Hill, 596 p.

Hall, L. M., 1968, Times of origin and deformation of bedrock in the Manhattan Prong, in Studies of Appalachian geology: northern and maritime: New York, John Wiley, p. 117-12 7. 137

Hansen, W. R., 1956, Geology and mineral resources of the Hudson and Maynard quadrangles, Massachusetts: U. S. Geol. Survey Bull. 10 38, 104 p.

Hartshorn, J. H., 1962, Late-'glacial eolian activity in Massachusetts (abs.): Geol. Soc. America Spec. Paper 68, p. 19 4.

Harvey, R. S., 1920, Drainage modifications and glaciation in the Danbury region, Connecticut: Connecticut Geol. and Nat. History Survey Bull. 30, 59 p.

Hobbs, W. H . , 1901, Still Rivers of western Connecticut: Geol. Soc. America Bull., v. 13, p. 17-26.

Hokans, D. H., 1952, Glacial water levels in the Housatonic- Naugatuck river valleys of western Connecticut: Clark Univ., unpub. thesis.

Jahns, R. H., 1941, Outwash chronology in northeastern Massachusetts (abs.): Geol. Soc. America Bull., v. 52, p. 1910.

------195 3, Surficial geology of the Ayer quadrangle, Massachusetts: U. S. Geol. Survey Geol. Quad. Map GQ-21.

Jahns., R. H., and Willard, M. E. , 1942, Late Pleistocene and Recent deposits in the Connecticut Valley, Massachusetts: Am. Jour. Sci., v. 240, p. 161-191, 265-287.

Karlstrom, T. N. V., 1952, Improved equipment and techniques for orientation studies of large particles in sediments: Jour. Geology, v. 60, p. 489-493.

Kaye, C. A., 1961, Pleistocene stratigraphy of Boston, Massachusetts: U. S. Geol. Survey Prof. Paper 424-B, p. B73-B76.

------196 4, Outline of Pleistocene geology of Martha's Vineyard, Massachusetts: U. S. Geol. Survey Prof. Paper 501-C, p. C134-C139.

Koteff, Carl, 1974, The morphologic sequence concept and deglaciation of southern New England, iri Glacial geomorphology: Binghamton, State Univ. of New York, p. 121-144.

Koteff, Carl, and Stone, B. D., 1971, Surficial geology of the Merrimack River valley between Manchester and Nashua, New Hampshire, Trip B-4 iri Guidebook for field trips in central New Hampshire and contiguous areas — New England Intercollegiate Geol. Conf., 63rd Ann. Mtg., Concord, N. H., 1971, p. 100-104. 138

Leopold, E. B., 1956, Two late-glacial deposits in southern Connecticut: U. S. Natl, Acad. Sci. Proc., v. 42, p. 863-867.

Melvin, R. L., 1970, Hydrogeologic data for the upper Housatonic River basin, Connecticut: Connecticut Water Resources Bull. 22, 33 p.

Mickelson, D. M. , 1973, Nature and rate of basal till deposition in a stagnating ice mass, Burroughs Glacier, Alaska: Arctic and Alpine Research, v. 5, p. 17-2 7.

Moore, F. II., 1935, Marbles and limestones of Connecticut: Connecticut Geol. and Nat. History Survey Bull. 56, 56 p.

Moss, J. H., 1943, Two tills in the Concord quadrangle, Massachusetts (abs.): Geol. Soc. America Bull., v. 54, p. 1826.

Munsell Color Company, Inc., 1954, Munsell soil color charts: Baltimore, Munsell Color Co., Inc.

Pawloski, J. A., 1967, Pleistocene invertebrate fossils from Bethel, Connecticut: Rocks and Minerals, v. 42, p. 664-66 8.

Pessl, Fred, Jr., 1966, A two-till locality in northeastern Connecticut: U. S. Geol. Survey Prof. Paper 550-D, p. D89-D93.

------19 71, Till fabrics and till stratigraphy in western Connecticut, iri Till — a symposium: Columbus, Ohio State Univ. Press, p. 92-105.

Pessl, Fred, Jr., and Schafer, J. P., 1968, Two-till problem in Naugatuck-Torrington area, western Connecticut, Trip B-l in Guidebook for field trips in Connecticut — New England Intercollegiate Geol. Conf., 60th Ann. Mtg., New Haven, Conn., 1968: Connecticut Geol. and Nat. History Survey Guidebook 2, 25 p.

Prucha, J. J., Scotford, D. M., and Sneider, R. M., 1968, Bedrock geology of parts of Putnam and Westchester Counties, New York, and Fairfield County, Connecticut: New York State Mus. and Sci. Service Map and Chart Ser., no. 11, 26 p.

Schafer, J. P., and Hartshorn, J. H., 1965, The Quaternary of New England, in The Quaternary of the United States: Princeton, Princeton Univ. Press, p. 113-12 8. 139

Sirkin, L. A., 1967, Late-Pleistocene pollen stratigraphy of western Long Island and eastern Staten Island, New York, in Quaternary paleoecology: New Haven, Yale Univ. Press, p. 219-274.

Sirkin, L. A., Owens, J. P., Minard, J. P., and Rubin, Meyer, 19 70, Palynology of some Upper Quaternary peat samples from the New Jersey coastal plain: U. S. Geol. Survey Prof. Paper 700-D, p. D77-D87.

Sirkin, L. A., and Minard, J. P., 1972, Late Pleistocene glaciation and pollen stratigraphy in northwestern New Jersey: U. S. Geol. Survey Prof. Paper 800-D, p. D51-D56.

Smith, H. T. U., and Fraser, H. J., 19 35, Loess in the vicinity of Boston, Massachusetts: Am. Jour. Sci., 5th ser., v. 30, p . 16-32 .

Thompson, W. B., 1971, The drainage and glacial history of the Still River valley, southwestern Connecticut: Univ. of Vermont, unpub. thesis.

Thornbury, W. D., 1965, Regional geomorphology of the United States: New York, John Wiley, 609 p.

U. S. Department of Agriculture, 1951, Soil survey manual: U. S. Dept. Agriculture Handb. 18, 50 3 p.

White, S. E., 1947, Two tills and the development of glacial drainage in the vicinity of Stafford Springs, Connecticut: Am. Jour. Sci., v. 245, p. 754-778. PLATE I

SURFICIAL GEOLOGIC MAP OF THE DANBURY AND NEW MILFORD QUADRANGLES, CONNECTICUT 3 S ill i9S» /*v e v j . s a h o b x o h Annexou 3230"

handles M*" t"1 Qal Cedar Island tq

Oak 0 I

Rock island

SPILLW AY ELEV 439

\ SPILLW429 AY ELEV

If.:**, r s?;n h

* 4 9 s

:-y. . #A. / >V' ,.

m S M . m * ■ v*VT' V V . »'! Q i m l Mf/IULWl a w V- X m Ay- * * >rs\* ••> ■"'"- R ® ' ! ^

4 M

Candle*------va Orchards Qc

Arrow head pf AN D LEW 0 0 D ILL WAV ELEV 4 2 9

Pocono^ -- -"PT^ 3 N A t Oflk'S I N M O A M 3 N )

bUALt I:Z4U0u 1 o ROAD CLASSIFICATION ooe 2ooo 3C0C1 4000 5000 6000 'OOC Fit' Heavy-duty Ligfst-duty «s Medium-duti' Unimproved dm = r CONTOUR INTERVAL 10 FEET interstate Route U S. Route State Route DATUM IS MEAN SEA LEVEL CONNECTICUT

QUADRANGLE LOCATION DANBURY, CONN. ''PLATE 1

SURFIGIAL GEOLOGIC MAP OF AND NEW MILFOBD QUAD CONNECTICUT

JOggsiX.™ n • 'tl, w 53S<5

&

as'

P5

•TSS ♦, '*s>sss>

LO 3 8 U 13Z9 .Thistle I

Candle' SPILLWAY ELEY 430

Candlei

i Cedar Island

Point Hedden

•i l l w a y e l 420

Turners

f bpear 'a Point f ipsar 'a Point

ivu ^ i aiKiicvv'.dd Lake CIu|)

% CandievvoodiMm I:. Ist e «

(’andleww Shon

Pinel Island

'andleww Orchard

Arrowheai Pine island

Orchardi

Arrowhea ’V Pt Jollywylt L A K§ B C ANDLEWOOD ?ark EUEV 429

T ^ Mb,

'"Driftwood

27 ' 30 "

ICandiewood fPines

4590

EattLake Reservoir fPILLWAYEU *9> e: *£

««! £os* Lo&e Reservoir WLLWAY £L%V w t:

el

«*bj 25'

m «82 200 000 F E E f 7 3 * 3 0 ' jp Mapped, edited, and published by the Geological Survey SCALE 1:24000 in cooperation with Connecticut Highway Department * .. »»H ! '.000 0 !000 0000 : 3000 4000 Gontroi by USGS, USC&GS, and Connecticut Geodetic Survey I Topography by photogrammetric methods from aerial photographs taken 1950. Field checked 1953. Revised from aerial •\ N photographs taken 1962. Field checked 1963 * CONTOUR INTERVAL 10 FEET ) ;}i8M‘Vs • DATUM IS MEAN SEA LEVEL Polyconic projection. 1927 North American datum 10,000-foot grid based on Connecticut coordinate system I . 1000-meter Universal Transverse Mercator grid ticks, rone 18. shown in blue OTH GRtO AND 1963 MAGNETIC NORTH declination AT center OF sheet PROFIL] DEP(

Maximum altil A-A*, B-B*, areas of rc reconstruci ized deposj Dashed blue glacial Lai postglacial Saugatuck B

UJ h-QC if> 5 0 0 '- | Qsr Q sr Qsr Qsr

3 5 0 VERTICAL EXAGGERATION X20

O < O

>- 5 5 0 ' Z Qsr.

5 0 0 '-

4 5 0 '-

3 5 0 '-

3 0 0 ' ■ PLATE II

PROFILES OF GLAGIOFLUVIAL, STREAM-TERRACE, AND FLOOD-PLAIN DEPOSITS IN THE DANBURY AND NEW MILFORD QUADRANGLES

Maximum altitudes of deposits are projected onto the planes o f sections A-A* t B-B*, G-G', and D-Df, from both sides of each section line. Small areas of recently modified topography in Danbury quadrangle have been reconstructed from map edition of 1953* Solid blue lines indicate ideal­ ized depositional gradients for related meltwater-stream deposits. Dashed blue lines show lowest possible levels for indicated stages of glacial Lake Danbury, based on present-day spillway elevations, A postglacial tilt of feet per mile is assumed for lake levels, Saugatuck River stage lake level is shown at maximum extent.

r—500‘ 450'

Qgs Q s r t/> Qsr Qsr 400'

350'

350 300'

Qsr.

Qgs N sections ne. Small ave been icate ideal- sits. tages of s. A vels.

CO r 4 5 0

OL

CO - 4 0 0 Q p b 3 5 0 ' - -3 5 0

3 0 0 ■300'.

on CD r 5 50

-500

Qgs

•Qsr. Q s r

■300 U\ CA BEND IN CO CA CO o CA o SECTION O ca o o o o O O o

ro

U3

BEND in SECTION

STONY HILL ROAD b o a r d m a n BRIDGE BEND IN SECTION

BOARDMAN BRIDGE STONY HILL ROAD

c-+

BEND IN SECTION U. S. ROUTE 7

BEND IN SECTION

Cl­

io

LIMEKILN BROOK LIMEKILN LIMEKILN BROOK co In in CO

- 3 0 0 SECTION BEND BEND IN

U.S. ROUTE U.S. 7 ROUTE GORGE " GORGE "HOUS ATONIC"HOUS SNJ D bedrock AND BEDROCK CONTOURS IN THE DANBURY AND NEW HILFGRD mDRAKLfiS

Bedrock contours modified after Robert !. Kelvin.

* MAP SHOWING BE AND BEDROCK CO)

Bedrock MAP SHOWING BEDROCK OUTCROPS, AREAS OF CLOSELY SPACED OUTCROPS, AND BEDROCK CONTOURS IN THE DANBURY AND NEW MILFORD QUADRANGLES

Bedrock contours modified after Robert L. Melvin

r> JED OUTCROPS, ) QUADRANGLES n

x:

&

A N^ew Milford ^New Milford "irrTnTTTTrmiT < %

^rriTrfTmmTTTTT> \ . ) Brookfield i a

// CD / /

Brookfield . • ^

TT

•I

V

in

450 & 3 5 0 in

0J PLATE‘IV

EXPLANATION ?0 ACCOMPANY PLATES I, II, AND III

CORRELATION OF MAP UNITS

Qt Q t

EXPLANATION OF MAP UNITS AND SYMBOLS

ALLUVIUM — Sand, silt, and gravel deposited on flood plains Qal by modern streams'^

SWAMP DEPOSITS — Peat, silt, and sand in poorly drained areas, Qs

STREAM-TERRACE DEPOSITS — Sand, gravel, and silt on terraces Qst cut into glacial-lake or glacial-stream deposits. Farmed in part during late-glacial time, UNDIFFERENTIATED DEPOSITS — Sand, gravel, and silt of uncertain Qgu age and origin. May be glacial or postglacial,

GLACIAL-STREAM DEPOSITS — Gravel and sand deposited by meltwater Qgnm streams in the Housatonic River valley north of New Milford*

GLACIAL-STREAM DEPOSITS — Gravel deposited by meltwater streams Qgm in the Morrissey Brook valley,

GLACIAL-STREAM DEPOSITS — Gravel and sand deposited by meltwater Qgsh streams that flowed south through the Sherman area into the Rocky River (Lake Candlewood) valley, GLACIAL LAKE CANDLEWOOD DEPOSITS — Sand and gravel deposited in Qc or graded to glacial Lake Candlewood, Concealed in large part by modern Lake Candlewood, GLACIAL LAKE KENOSIA DEPOSITS Qk Qk: sand, gravel, and silt deposited by meltwater streams that entered lake from the west, Qkb: silt, sand, and clay deposited on lake bottom. Q k b

GLACIAL LAKE DANBURY DEPOSITS Qhr Sand, gravel, and silt laid down by meltwater streams that entered lake, and lake-bottom deposits of sand, silt, and clay, Qbrb Qhr: Housatonic River stage deposits, Qhrb: Housatonic River stage lake-bottom deposits.

Qph Qph: Pumpkin Hill stage deposits, Qphb: Pumpkin Hill stage lake-bottom deposits,

Qphb Qpb^ and Qpb^: Pond Brook stage deposits, Qpb^ is older, Qpbb: Pond Brook stage lake-bottom deposits,

Qpb, Qsr, Qsr,, and QsrSaugatuck River stage deposits. Where numbered, Qsr^ lis older, • Qpb. Qsrb: Saugatuck River stage lake-bottom deposits.

Qpbb Qpbb: Pond Brook stage lake-bottom deposits. '

Qpb^ Qsr, Qsr., and Qsrg! Saugatuck River stage deposits. Where numbered, Qsr^ is older, Qpb, Qsrbs Saugatuck River stage lake-bottom deposits.

Qpbb

Qsr,

Qsr.

Qsr

Qsrb

- GLACIAL-STREAM DEPOSITS — Sand and gravel deposited by meltwater Qgg streams in the Great Brook valley, northeast of New 'Milford,

GLACIAL-STREAM DEPOSITS — Sand arid gravel deposited by meltwater Qgt streams in the Town Farm Brook valley, southeast of New Milford.

GLACIAL-STREAM DEPOSITS — Sand, gravel, and silt deposited by Qgp, meltwater streams that flowed southeast into the Pond Brook valley (Newtown quadrangle). Qgp. is oldest. Qgp,

QgPt

GLACIAL-STREAM DEPOSITS — Sand, gravel, and silt deposited by Qgs meltwater streams that flowed south into the Saugatuck River valley (Bethel quadrangle), GLACIAL-STREAM DEPOSITS — Gravel and sand deposited by meltwater Qgb streams east of Bethel,

UNCQRRELATED GLACIAL-STREAM DEPOSITS — Sand and gravel. Qg Chronologic position not assigned.

TILL — Poorly sorted rock debris deposited by glacial ice. Qt Includes both upper till and lower till. Pit exposures identified by letter symbols: t, upper till; T, lower till. Subscripts: a, ablation facies; b, basal facies; m, locality where lower till is intermixed with upper till, BEDROCK EXPOSURES — Ruled pattern indicates areas of closely spaced outcrops where surficial deposits are thin. Mapped in part from aerial photographs,

ARTIFICIAL FILL af af: obtained from till, sand and gravel, or bedrock, aft: dump material. Contains considerable trash. aft

Contact — dashed where approximate!v lonated. Qt Includes both upper till and lower till• Pit .exposures identified by letter symbols; t, upper till; Tf lower till. Subscripts; a, ablation facies; b, basal facies; m, locality where lower till is intermixed with upper till. BEDROCK EXPOSURES — Ruled pattern indicates areas of closely spaced 'outcrops where surficial deposits are thin. Mapped iri part from aerial photographs, • ARTIFICIAL FILL af af; obtained from 'till, sand and gravel, or bedrock, aft; dump material. Contains considerable trash. aft

Contact — dashed where approximately located.

Scarp separating adjacent surfaces within same map unit — Hachures on downslope side. Dashed where approximately located.

•1 5 0 - Bedrock contour — Shows approximate altitude of bedrock surface. Contour interval is 50 feet. Datum is mean sea level. Contours omitted in upland areas,

Drumlin — Symbol shows direction of long axis. Composed mostly or entirely of till.

22 £ Glacial striation Point of observation at tip of arrow, - .

Glacial boulder — Glacially transported boulder with maximum diameter greater than h feet; A, amphibolite and hornblende gneiss; B, "Brookfield Plutonic Series"; BGf biotite gneiss; D, "Danbury Augen Granite"; G, "gneissic granite and trondh- jemite"; GMS, garnet-mica schist; GR, granite, undifferentiated, white to pink, locally foliated; I, Inwood Marble; Q, quartzite; S, schist, undifferentiated; TM, tremolite; Y, "younger granite". Quoted names are those of Clarke (1958).

Area of numerous large till boulders.

Ice margin position —- Shows inferred position of stagnant-ice margin during deposition of adjacent meltwater deposits. Dashed where correlation between positions is probable.

Dip direction of delta foreset beds — Number is altitude in 44-0, feet of topset-foreset contact (where known).

Dip direction of fluvial cross-bedding.

Crest of esker or other ice-Channel filling,

Meltwater channel — Arrow indicates direction of stream flow.

Glacial lake spillway and/or local base level for glacial streams — Q K r -vAAr* Letter symbols indicate map units graded to lake level controlled by spillway or to local base level.

kt Morphology — Letter symbols indicate examples of glaciofluvial land forms; k, kame; kt, kame terrace; kd, kame delta. Contour interval is 50 feet. Datum is mean sea level, Contours omitted in upland areas.

Rrumlin— Symbol shows direction of long axis, Composed mostly or entirely of till.

. Glacial striation Point of observation at tip of arrow.

Glacial boulder — Glacially transported boulder with maximum -H D diameter greater than ^ feet: A, amphibolite and hornblende gneiss; B, "Brookfield Plutonic Series"; BG, biotite gneiss; D, "Danbury Augen Granite"; G, "gneissic granite and trondh- jemite"; GMS, garnet-mica schist; GR, granite, undifferentiated, white to pink, locally foliated; I, Inwood Marble; Q, quartzite; S, schist, undifferentiated; TM, tremolite; Y, "younger granite". Quoted names are those of Clarke (1958)*

Area of numerous large till boulders.

Ice margin position — Shows inferred position of stagnant-ice margin during deposition of adjacent meltwater deposits, - Dashed where correlation between positions is probable.

Dip direction of delta foreset beds — Number is altitude in 4±9> feet of topset-foreset contact (where known).

-- > Dip direction of fluvial cross-bedding.

Crest of esker or other ice-Channel filling,

Meltwater channel — Arrow indicates direction of stream flow.

Glacial lake spillway and/or local base level for glacial streams — Qhr Letter symbols indicate map units graded to lake, level controlled by spillway or to local base level.

kt Morphology — Letter symbols indicate examples of glaciofluvial land forms: k, kame; kt, kame terrace; kd, kame delta.

Till or sand and gravel pit — Extent of large pits shown.by i 6 5 s hachures. Letter symbols indicate materials in decreasing 'X' active order of abundance: cl, clay; st, silt; s, sand; ps, pebbly sand; p, pebble gravel; c, cobble gravel; b, boulder gravel; 6 P f, flowtill. Superposed symbols indicate superposition of X ^ t inactive materials. Numbers are thicknesses in feet.

Quarry active X inactive Location of logged well or test hole, See Appendix A for logs, -f I? Numbers assigned by Melvin (.1970), ■ well + 98 test hole Glacial boulder — Glacially transported boulder with maximum + D diameter greater than h feet: A, amphibolite and hornblende gneiss; B, "Brookfield Plutonic Series"; BG, biotite gneiss; D, "Danbury Augen Granite"; G, "gneissic granite and trondh- jemite"; GMS, garnet-mica schist; GR, granite,undifferentiated, white to pink, loyally foliated; I, Inwood Marble; Q, quartzite; S, schist, undifferentiated; TM, tremolite; Y, "younger granite". Quoted names are those of Clarke (1958)*

Area of numerous large till boulders,'

Ice margin position — Shows inferred position of stagnant-ice margin during deposition of adjacent meltwater deposits. Dashed where correlation between positions is probable.

Dip direction of delta foreset beds — Number is altitude in 44-0. feet of topset-foreset contact (where known).

Dip direction of fluvial cross-bedding.

Crest of esker or other ice-Channel filling,

Meltwater channel — Arrow indicates direction of stream flow.

Glacial lake spillway and/or local base level for glacial streams — Q hr — v/lAr^ 1 Letter symbols indicate map units graded to lake level controlled by spillway or to local base level.

Morphology — Letter symbols indicate examples of glaciofluvial kt land forms: k, kame; kt, kame terrace; kd, kame delta.

Till or sand and gravel pit — Extent of large pits shown.by s hachures. Letter symbols indicate materials in decreasing ' X - active order of abundance: cl, clay; st, silt; s, sand; ps, pebbly sand; p, pebble gravel; c, cobble gravel; b, boulder gravel; 6 P f, flowtill* Superppsed symbols indicate superposition of X ~ g f inactive materials. Numbers are thicknesses in feet.

Quarry active ■fc inactive Location of logged well or test hole. See Appendix A for logs. Numbers assigned by Melvin (.1970) • well + $ test hole