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.

University Microfilms International 300 North Zeeb Road Ann Arbor, Michigan 48106 USA St John's Road, Tyler's Green High Wycombe, Bucks, England HP10 8HR MASTERS THESIS 13-9436 CLARK, Wayne Edward THE APPLICATION OF REGIONAL RESEARCH DESIGNS TO CONTRACT ARCHEOLOGY: THE NORTHWEST TRANSPORTATION CORRIDOR ARCHEOLOGICAL SURVEY PROJECT, The American University, M.A., 1976 Anthropology, archaeology

Xerox University Microfilms, Ann Arbor. Michigan 48106

0 1976

WAYNE EDWARD CLARK

ALL RIGHTS RESERVED THE APPLICATION OF REGIONAL RESEARCH DESIGNS TO CONTRACT ARCHEOLOGY: THE NORTHWEST TRANSPORTATION CORRIDOR ARCHEOLOGICAL SURVEY PROJECT.

By

^ Wayne E. Clark.

Submitted to the

Faculty of the College of Arts and Sciences

of The American University

in Partial Fulfillment of

the Requirements of the Degree

of

Master of Arts

Anthropology

Signatures of Committee

Chairman Dean'^^the^ College

Date Date:

1976

The American University Washington, D. C.

THE AMERICAN UNIVERSITY LIERA]-;Y THE APPLICATION OF REGIONAL RESEARCH DESIGNS TO CONTRACT ARCHEOLOGY:

THE NORTHWEST TRANSPORTATION CORRIDOR ARCHEOLOGICAL SURVEY PROJECT

by

Wayne E. Clark

ABSTRACT

The applicability of incorporating recent methodological and

theoretical advances within the framework of conservation archeo­

logy is tested by using the three years of archeological investi­

gations of the Northwest Transportation Corridor as a case study.

The thesis places the archeological resources of the Gwynns Falls

Valley into a paleo-ecological perspective utilizing the present

habitat variables as a starting point for predicting past habitat

diversity; developes a diachronic synthesis of cultural develop­

ment for the Gwynns Falls Valley and surrounding Eastern Piedmont

province; and advances a culture history of the formulation, ex­

pansion and demise of the Potomac Creek complex during the Late

Woodland period. These major themes were formulated to place the

cultural resources of the Gwynns Falls Valley within the larger

perspective of the environmental and cultural variables of the

surrounding Chesapeake Bay region. The thesis demonstrates that

conservation archeology can utilize recent methodological and

theoretical advances in archeology. Preface

In 1973, the Maryland State Highway Administration requested the Maryland Geological Survey to conduct an archeological recon­ naissance and preliminary site examination of the Northwest Trans­ portation Corridor as an ancillary preliminary engineering service prior to determination of final route location. The Federal regula­ tion which applies to the investigation is the Federal-Aid Highway

Program Manual, Volume 7, Chapter 7, Section 4. The first phase of investigations in 1973 entailed a preliminary examination of the corridor and revealed six historic and six prehistoric archeological sites. Based on the results and recommendations of the 1973 survey

(Clark 1973), controlled surface collection of the prehistoric sites falling within the proposed Dolfield Road Interchange right-of-way were conducted in the summer of 1974 (Clark 1975a). Final phase excavation and additional controlled surface collection of the sites within the proposed Dolfield Road Interchange right-of-way was com­ pleted by May 2, 1975. Surveys of the Horsehead Branch, Red Run

Stream and Upper Gwynns Falls River Valleys were conducted in the summers of 1974 and 1975 to acquire settlement subsistence pattern data applicable to the thesis. The diachronic research design developed for the Gwynns Falls Valley and the synchronic research design developed for the Potomac Creek complex have been reformulated and refined. This thesis synthesizes the results of three years of archeological investigations of the Northwest Transportation Corridor.

During the three phases of the project and the preparation of the thesis, many individuals offered their assistance and encourage­ ment. First and foremost, I extend my appreciation to Tyler Bastian, Division of Archeology, Maryland Geological Survey. Mr. Bastian's interest and guidance throughout the progress of research have been instrumental in the completion of the various phases of the project and thesis. I wish to thank the public-minded Painters Mill Venture for allowing investigation of the prehistoric sites on its property and particularly to Howard Weinstein, the patient representative of the Painters Mill Venture. I thank all the amateur archeologists, high school students, and college students who contributed their time and enthusiasm. I especially wish to thank Betty Pearre of

Pikesville, Maryland, for her constant encouragement, cooperation and loan of her collection; Charles Hazard of Reisterstown, Maryland, for his assistance in excavation and loan of his collection; Louis

Phelps of Linthicum, Maryland, for his assistance in specimen pro­ cessing; and Virginia Clark of Adelphi, Maryland and Mary Kegley of

Richmond, Virginia for excellent editorial advice and assistance.

The constructive comments of Dr. Gary Hume and Dr. Charles McNett,

Jr., of American University were most helpful as were discussions with Russell Handsman of the University of Maryland and Howard

MacCord, Sr., of the Virginia State Library. The typing of the manuscript by Amy Palmer was deeply appreciated. I would like to thank the Maryland State Highway Administration and the Federal

Highway Administration, U.S. Department of Transportation for pro­ viding the impetus and funds for the project. Finally, the contri­ butions of my parents to the completion of the project are incal­ culable .

ii TABLE OF CONTENTS

Preface ...... 1

Table of Contents...... ill

List of Illustrations...... iv

List of T a b l e s ...... vii

Chapter

I. ARCHEOLOGICAL RESOURCE MANAGEMENT ...... 1

II. THE EASTERN PIEDMONT PHYSIOGRAPHIC PROVINCE ...... 9

III. SURVEY TECHNIQUES ...... 37

IV. ANALYTIC TECHNIQUES ...... 69

V. DIACHRONIC SYNTHESIS ...... 89

VI. THE POTOMAC CREEK COMPLEX...... 178

VII. SUMMARY AND CONCLUSIONS...... 227

Appendix A ...... 245

References cited ...... 265

iii LIST OF ILLUSTRATIONS

Figure 1: Physiographic provinces in Maryland...... 5

Figure 2: Major drainage basins in Maryland...... 9

Figure 3: Generalized geologic map of the Eastern Piedmont province...... 12

Figure 4: Vegetation map of the Eastern Piedmont province...... 13

Figure 5: Generalized soils map of the Eastern Piedmont province...... 15

Figure 6: Distribution of sites in the Gwynns Falls Valley...... 16

Figure 7: Dolfield Road Interchange: prehistoric sites 18 BA 106, 107, 112 ...... 42

Figure 8: View (facing southeast) from ridge of the 1975 investigations of the Gwynns Falls site, 18 BA 112 ...... 44

Figure 9: View (facing west) from Painters Mill Road of the 1975 investigations of the Painters Mill site, 18 BA 106 ...... 46

Figure 10: Excavated area, central section, of the Painters Mill site...... 49

Figure 11: Drawing of the profile of the west wall of square S42W48, 18 BA 106 ...... 51

Figure 12: Controlled surface collection of the Lee site.... 61

Figure 13: Spectographic read-out on lithic specimens from the Gwynns Falls Valley...... 76

Figure 14: Distribution of chronologically important sites.. 94

Figure 15: 1974-75 controlled surface collections of the Gwynns Falls site, 18 BA 112 ...... 144

Figure 16: 1975 excavations of 18 BA 112 ...... 145

Figure 17: 1974-75 controlled surface collections of the Painters Mill site, 18 BA 106 ...... 146

Figure 18: 1975 excavations of 18 BA 106 ...... 147

iv Figure 19: Distribution of cryptocrystalline artifacts...... 148

Figure 20: Distribution of feldspar rhyolite debitage 149

Figure 21: Distribution of black and plain rhyolite debitage...... 150

Figure 22: Distribution of vein quartz debitage...... 151

Figure 23: Distribution of opaque and cobble quartz debitage...... 152

Figure 24: Distribution of scrapping tools by lithic type...... 153

Figure 25: Distribution of cutting tools by lithic type... 154

Figure 26: Distribution of chopping tools, bannerstones, hammerstones, and fire cracked rock...... 155

Figure 27: Distribution of piercing implements by type.... 156

Figure 28: Distribution of Townsend comnlex vessels..... 157

Figure 29: Distribution of Moyaone Related and Moyaone Incised vessels...... 158

Figure 30; Distribution of Potomac Creek complex vessels...... 159

Figure 31: Lithic artifacts from 18 BA 106 (bifaces) 160

Figure 32: Lithic artifacts from 18 BA 1 0 6 ...... 162

Figure 33: Pottery from 18 BA 106 ...... 164

Figure 34: Lithic and ceramic artifacts from 18 BA 112 ... 166

Figure 35: Lithic artifacts from the Gwynns Falls Valley.. 168

Figure 36: Lithic artifacts from 18 BA 129 ...... 170

Figure 37: Bifaces from the Pearre collection...... 172

Figure 38: Bifaces from the Hazard collection...... 174

Figure 39: Lithic artifacts from the Hazard collection.... 176

Figure 40: Potomac Creek complex site distribution...... 192

Figure 41: The Potomac-Patuxent portion of Capt. John Smith's map...... 208 Figure 42: Flake angle and dimensional attribution definitions...... 248

Figure 43; Projectile point dimensional attribute definitions...... 248

Figure 44: Projectile point classification code...... 256

Figure 45: Functional attribute code...... 260

vi LIST OF TABLES

Table 1: Artifact category and lithic material percentage.. 63

Table 2: 18 BA 112 debitage data from 1974 controlled surface collection...... 249

Table 3: 18 BA 112 debitage data from 1973-1974 random survey...... 250

Table 4: 18 BA 106 debitage data from 1974 controlled surface collection...... 252

Table 5: Debitage data from 1973 random finds collection... 255

Table 6; Debitage data from 1974 random collection from 18 BA 112 ...... 255

Table 7: Piercing category (projectile point) attribute list...... 258

Table 8: Scraping category attribute list...... 258

Table 9: Cutting category attribute list...... 259

Table 10 Chopping category attribute list...... 259

Table 11 Wear pattern attribute list for 18 BA 106 ...... 261

Table 12 Wear pattern attribute list for 18 BA 112 ...... 262

Table 13 Summar, of 1973-75 artifact categories for 18 BA 106 ...... 263

Table 14: Summary of 1973-75 artifact categories for 18 BA 112 ...... 264

vii CHAPTER ONE

ARCHEOLOGICAL RESOURCE MANAGEMENT

Introduction

The increased demand for consultant services resulting from the passage of the National Environmental Policy Act of 1969 and the

Archeological Conservation Act of 1974 has presented archeologists who adhere to a deductive methodology with a dilemma. Under the deductive approach, specific hypotheses are developed based upon available knowledge and tested by planned, systematic investigations of the region or areas relevant to the problem. The deductive ethic calls for field investigations only when specifically required to

test relevant hypotheses. The site should be selected on the basis of the problem and not vice versa (Kl.ng 1971: 58). However, land- modification programs predetermine the area to be investigated

based upon external factors such as the boundaries of the Northwest

Transportation Corridor. Most support agencies and many archeol­

ogists currently conceive of the archeological input to an environ­

mental impact statement as an inductive, descriptive analysis of

the archeological resources within the impact area. The resultant

reports, when rarely published, seldom contain references to research

outside the immediate area and invariably fail to utilize or test

recent methodological or theoretical innovations.

To adequately satisfy the requirements of the federal legis­

lation, to insure that a part of our archeological heritage is

preserved for future generations, archeologists must increasingly

incorporate and operationalize the methodological and theoretical

advances of the past two decades into the framework of conservation archeology. The conservation ethic, if taken to a radical extreme, states that a moratorium on all archeological investigations (both surface and subsurface) of all non-directly threatened sites should be imposed until all previous research and collections are analyzed, interpreted and reported. Research would be strictly limited to boundaries established by land-modification forces, whether they be human or natural. A less radical viewpoint, adopted by this author, allows for the continuation of "pure", problem-oriented research if such research will contribute substantially to the clarification of problems relevant to determining the significance of sites within a given research region or regions. Excavation of sites purely as a teaching device should be discontinued. In reality, if the current rate of site depletion is not slowed, all research, regardless of the approach, will be of the conservationist nature.

Clearly the solution of the conflict between the conservation ethic and the deductive ethic as defined above lies in the incorpo­ ration of recent methodological and theoretical advances into the framework of conservation archeology. In the case of the Northwest

Transportation Corridor, Richard Ackroyd, Division Engineer of the

U.S. Department of Transportation, has specifically requested that

"where there are known artifacts or remains in the corridor, the

EIS should describe them in relation to their loss as a part of a larger resource base." To accomplish this request, multistage research and regional research designs are a necessity. The respon­

sibility for determining what constitutes an adequate survey or a

significant site rests primarily with archeologists, but only if we

accept and voice our concerns. Unfortunately, many state and federal agencies have failed to realize not only the need for conducting archeological investigations but also the type of investigation which should be conducted. Archeologists should attempt to educate

the various land-modification agencies and individuals of the

current crisis in American archeology and the necessity of develop­

ing a multitude of approaches to deal with the crisis. Detailed

studies in a number of states indicate that if the current rate of

land-modification continues for another 25 years most of our arche­

ological resources will have been destroyed (McGimsey 1974: 3).

Archeologists are responsible for generating large-scale,

long-range deductive research programs that focus on regions of

conservation concern and for structuring the conservation programs

so that they contribute to the ends of the overall research designs.

Such regional research designs, which could be developed by a con­

sortium of archeologists from area institutions and governmental

agencies, requires the cooperation, understanding and funding of

the land-modification agencies. The resultant reports, while pro­

viding the essential information required to assess the significance

and extent of the archeological resources of a project area, should

be oriented toward the development, testing and reformulation of

the hypothesis of the overall research design. Eventually the

results of various projects will allow the development of predictive

statements about the expected location and significance of the

archeological resources from the research region. Two regional

research designs have been developed by the author and are presented

in Chapters 5 and 6.

Realization of the goals of conservation programs necessitates

production of more than descriptive text. An assessment of the significance of the archeological resources directly and indirectly affected by land-modification forces requires the evaluation of the resources within the broader scale of a region; within the broader field of anthropology. Only by going beyond data collection and description to hypothesis formulation and testing, can a conservation program be of maximum benefit to the study of man and to the public.

General Research Design

Archeologists are increasingly striving to isolate the arche­ ological structure of extinct cultural systems. A cultural system is a set of constant or cyclically repetitive articulations between man's various levels of adaptation (social, technological, and idealogical) according to White (1959:8). The localities, facilities, and tools used by social groups to perform specific tasks result in the structuring of the archeological record. The procurement, manufacture, use, and loss of implements and facilities at different locations where groups of variable structures performed different tasks leave a partial record of the extinct society. By means of controlled surface collections and systematic excavations, the structure may be interpreted by studying the nature of the resultant population of artifacts and their spatial association (Binford

1973: 158). The isolation and definition of the content, the structure, and the range of a cultural system, together with its relationship to the biological and physical elements of the ecosystem, should be a primary research objective of the archeologist. As this objective may or may not be successfully accomplished under any given research design, several research designs applicable to dif­

ferent ecosystems and problems should be developed. The varying boundaries of different cultural systems require

development of both diachronic and synchronic research designs.

Diachronic studies develop models of change for cultural systems

over long periods of time and usually focus on specific problems.

Synchronic studies develop hypotheses correlating the sub-systems

of a single culture within a given time period into a structural

whole. Both approaches require an understanding of the physical

and biological as well as the cultural elements of an ecosystem. r ' APPAL^jAN PROVINCE PIEDMONT PROVINCE ^i'77 . /' "jKôWTHwt'fF" TRANSPORTATION ■v// .

V i/

PHYSIOGRAPHIC PROVINCES

AND THEIR DIVISIONS

IN MARYLAND I ' ' i & a

PROVINCE

Figure 1: Physiographic provinces in Maryland (Modified — from Vok.es 1961) .

Maryland's physical and biological elements can be divided

into five major physiographic provinces; Coastal Plain, Piedmont

Plateau, Blue Ridge, Ridge and Valley and Appalachian Plateau

(Figure 1). Each physiographic province offered different resources

from the others and although cultures frequently utilized the

resources of several provinces, synchroi.ic and diachronic research designs can be successfully developed for each province and the data correlated to formulate a synthesis of cultural development in the

Chesapeake Bay region. Although the biological element is generally divided into three major forest regions: the Oak-Pine forest, the

Oak-Chestnut forest, and the Mixed Mesophytic forest, each major forest region is composed of a mosaic of vegetation units offering a diversity of habitats (Paradise 1969: 4). Regional research designs for each physiographic province will eventually be developed as archeologists realize the need for incorporating the results of various conservation archeology investigations into a larger per­ spective. Cultures which spanned several physiographic provinces can only be understood by examining the total settlement-subsistence systems but elucidation of the whole first requires an understanding of the parts.

Boundaries for the cultural element in a particular area depend on the culture under consideration since cultural systems differ greatly in the limits of their adaptive range. For example,

the Late Woodland Potomac Creek complex utilized the Western Shore

section of the Oak-Pine Forest and the Piedmont section of the Oak-

Chestnut Forest while the Townsend complex remained primarily within

the Eastern Shore and Western Shore sections of the Oak-Pine Forest

region (Chapters 5 and 6). The Potomac Creek complex was riverine

oriented while the Townsend complex was marine oriented but both

utilized the resources of the Piedmont province (Chapters 5 and 6).

Archeologists have concentrated their investigations on the Coastal

Plain sections and the riverine microenvironments giving little

attention to the upper areas of the Piedmont section. The major riverine microenvironments of the Susquehanna,

Monocacy and Potomac Rivers have received varying degrees of arche­ ological research while the minor riverine environments of smaller rivers in the Piedmont section have received little or no attention

(for example, Witthoft and Farver 1971; McNett and Gardner 1971).

While deductive research is underway in the Potomac River basin

(Handsman and McNett 1974), the emphasis has been primarily on the

large floodplain sites with limitation of finds and other factors

preventing intensive surveys of the upland areas. The Western Shore

section of the Oak-Pine Forest region has received extensive inves­

tigation by amateur and professional archeologists (Holmes 1897;

Stearns 1943; Stephenson and Ferguson 1963; Wright 1973). This

emphasis on the comparatively rich archeological resources of the

major riverine and coastal habitats has resulted in research designs

which fail to formulate and test hypotheses concerning the prehistoric

utilization of the uplaid resources of the three major forest regions.

The Northwest Transportation Corridor provides an excellent

opportunity to sample the archeology of the Eastern Piedmont eco­

system (Chapter 2). The Corridor is entirely within the Piedmont

and follows the Gwynns Falls Valley for a distance of four miles

before leaving the floodplain and ascending into the upland areas.

The Corridor continues another six to seven miles through the uplands

to the eastern edge of the Patapsco River Valley where the corridor

ends at the Westminster Pike. Bisecting both riverine and upland

habitats of the Piedmont section and passing within a mile of the

unique biological microenvironment of the serpentine barrens of

Soldiers Delight, the Northwest Transportation Corridor provides an opportunity to formulate and test hypotheses relating to a diachronic research design developed for the Oak-Chestnut Forest region of Maryland (Chapter 5). The discovery of Late Woodland period Townsend and Potomac Creek complexes at the Painters Mill site stimulated the development of a regional research design for

the Potomac Creek complex (Chapter 6). The environmental setting,

field work and analysis of the data base from the Gwynns Falls

Valley is presented first (Chapters 2 to 4, Appendix A) and the

information then incorporated into the regional research designs

(Chapters 5 and 6). The limitations of the various methods and

theories developed and the potentials for future research summarize

the thesis (Chapter 7). CHAPTER TWO

THE EASTERN PIEDMONT PHYSIOGRAPHIC PROVINCE

The Eastern Piedmont region in Maryland is defined as the area west of the Coastal Plain (Fall Line), east of the Monocacy River drainage (Parr's Ridge), north of the Potomac River and southwest of the Elk River (Figures 1 and 2). The region ranges in elevation from 300 feet near the Fall Line up to 1100 feet towards Parr's

Ridge. Gently undulating uplands dissected by narrow steep-walled valleys characterize the area. Complete dissection by fluvial erosion in some places has resulted in a complex of hills and stream valleys. In areas underlain by Cockeysville Marble, wide alluvial valleys are formed (Brush 1975: 5). Draining directly into the numerous estuaries of the Chesapeake Bay, the region is postulated

to have been the hunting territory of various coastal cultures during some periods of prehistory.

APPALACHIAN PIEDMONT WESTERN % COASTAL

MARYLAND DRAINAGE BASINS mm I

Figure 2: Major drainage basins in Maryland (Modified from Vokes 1961). 10

The interrelationships between the biological and physical elements of the Eastern Piedmont region have been correlated and mapped by Grace Brush (1975). Brush's report serves as the primary reference for understanding present habitat variables. Before the affect on biota of changing climatic conditions during the Holocene can be inferred, the relationships between present climatic con­ ditions and the resultant biota must be elucidated.

Present Climate

The Eastern Piedmont region has a continental type of climate with well-defined seasons. The greatest period of precipitation is summer, with winter and autumn sharing the least amount. Droughts and excessive rainfalls are rare and precipitation is well distrib­ uted throughout the year. In the Gwynns Falls River Valley, the average annual amount of precipitation is 43.05 inches. The average

annual seasonal snowfall of 24 inches is distributed over a six-month

period with February as the peak month. The lightest winds occur

during the months with the greatest heat.

Located on the Patapsco River eight miles south of the Corridor,

Woodstock has a monthly mean temperature which varies from 33.8°F

in January to 75.1°F in July (U.S. Department of Commerce 1965). The

highest temperature was recorded at 103°F in July while the lowest

recording of -14°F occurred in February. At McDonogh School, the

average date of the first killing frost is October 27. As the aver­

age date of the last killing frost is April 13, the average growing

season is 197 days (Maryland Geological Survey 1929). 11

Present Vegetation Units

Although Shreve (1910) first noticed the relationship between forest units and soil units in Maryland, the close correlation between forest, geologic, topographic, and soil units was not clearly

defined and mapped until Brush (1975) conducted a detailed field

survey of the Eastern Piedmont region. This close correlation

suggests that available water is the most important environmental

factor controlling distribution of woody species in the Eastern

Piedmont region. A study of the relationship between forest units

and environmental variables on the Coastal Plain of Delaware indi­

cates that drainage is the prime factor in determining forest units

in the Coastal Plain province (Thomas et al. 1975).

In the Piedmont Province:

Blackjack oak and associated species occupy the arid areas within the region, the chestnut oak unit occurs consistently in the dry, well-drained areas, tulip poplar forests are present in areas of greater moisture, and sycamore, box elder, green ash, silver maple associated species occupy wet floodplains (Brush 1975: 39).

Geologic units of low water availability such as the serpentine

areas (Figures 3 and 4) are suitable for the growth of species with

low water requirements. The failure of some species to grow in

particular areas is not because they are out-competed, but simply

because the habitat is unsuitable. Each geologic-soil unit of the

environmental mosaic is occupied by those species which are best

suited because the unit contains (1) the requirements for the growth

of those particular species, (2) the pattern of available nutrients,

and (3) accessible water compatible with the physiological mechan­

isms of the species for obtaining these ingredients. In short. HCSIAVOII*••ETTtSOT'

LOCM w*vr*<

‘‘•nescuvoi*

lower Wissohicken schist ond Boulder gneiss BALTIMORE E3 upper Wissohickjn schiS» ond MorbufÇ schist

gronitic rocks

gobProlc rocks

Bolhmore gneiss 0 Trensect A A, Setters quorttiie ® Tronseci fi-B , ullrcmctic rocks © Tronseet C- C, diobose © Tronseet 0 * 0 , groveis

ES! New Oxford Fornetion

Cockeysville Wcrble > WASHINGTON □

Figure 3: Generalized geologic map of the Eastern Piedmont province (From Brush 1975: Figure 2). ittsfvciny.

Chestnut Ooi unit

|..^. j T ulip P oplor Unit

Syccmo^e-Boielder-Green Ash - ■ Silver Mople Unit

B lockioch Ook U nit

Shingle Ook Unit

Stands of Tulip Poplor B a ond Chestnut Oak

Stands 0* T ulip P oplor end Mimed Oaks exclusive of Chestnut Oak

Figure 4: Vegetation map of the Eastern Piedmont province (From Brush 1975: Figure 14). 14

similar species associations are consistently found on similar sites because the associations are environmentally controlled or determined

(Brush 1975; 41).

Discontinuities in forest type occur where the environmental change, is abrupt; as, for example, at the contact between the ser­ pentine rock and the Lower Wissahicken Schist, where the blackjack oak forest which occupies the serpentine is juxtaposed against a tulip poplar forest within a few meters. The contact between the chestnut oak forest occurring on the Setters formation and the tulip poplar forest on the Baltimore Gneiss is similarily abrupt (Figures

3 and 4). The contact between vegetation in floodplain alluvium and that of adjacent slopes is also abrupt (Brush 1975: 42-43). The geologic boundary between the Upper Wissahicken Schist and Lower

Wissahicken Schist does not result in abrupt vegetation change ; rather, the vegetation boundary coincides more precisely with the boundary between the Mt. Airy Soil association and the Manor-

Glenelg-Chester Soil association (Figures 3 to 5).

The abrupt changes between forest units was probably a factor

considered in the selection of the location of prehistoric base

camps. Sites may have been located in areas facilitating the

exploitation of two or more vegetation units within a short distance

of the camp. All of the sites occurring within the Red Run and

Horsehead Branch stream valleys are within ten minutes walk from

three vegetation units: the tulip poplar unit on the Lower

Wissahicken Schist, the sycamore-box elder-green ash-silver maple

unit on the Cockeysville Marble, and the chestnut oak unit on the

Setters Quartzite formation (Figure 6). The sites in the area I. Soils Formed from ConsoHdoted Scdimenlory Rock* WILES Penn-Reodmgfon-Lewisbçffv Assoc'Otion BALTIMORE 2 . Soils Formed from Igneous ond Melomorphic Rocks A. Soils Derived from Bosic Rocks

Neshommy MontollO Associolion

Cono#ingo-Aldmo'Chrome Associctfon

8. Soils Derived from Acid Rocks

Monor -Gieneig * Chester Associotion

Ml. A iry Gleneig - Lingonore AssOC OliOn

Boitimore* Hogerstown Associotion

3. Soils Formed from Unconsolidated Coostol Pioin Sediments ^WASHINGTON I I Bctfsville'Chilium«Groom Associotion

I I Chrisfiono'Sunnyside-Sossofros Associotion

4. Soils Formed from Alluvial Moteriols

\ ' ' Pope - Holsfon-Comus-Elk-BiDb Associotion

L n Figure 5: Generalized soils map of the Eastern Piedmont Province (From Brush 1975: Figure 3) 16

'^Bol20

FAILLS KSTOWN

r - ^ ♦♦♦ + t*+ ****’**\\♦ +♦♦ •n

. Î +% + 4 <+4 + ^4 4 + + + 4 4 . 12": + + 4 4 4 4 4^^4 4^4 4 Bo 127 'i Bo 107 4 - 1 T 4 ^ 4 4 4 +4 4444+4444. 4 a

HorMhMd Br. Bol2 9 ^

DISTRIBUTION OF SITES IN THE UPPER G WYNNS FALLS VALLEY rmuLTWAMAFIC ROCK-SERPENTINE ^OTT§^ PEEDDULTRAMAFIC ROCK-PYROXEMITE BRANCH I i mwER WISSAHICKEN SCHIST AND BOULDER GNEISS COCKEYVILLE MARBLE 2ZQSETTERS OUARTZITE 2 5 3 BALTIMORE GNEISS BFîÂg GABBROIC ROCK

Figure 6: Distribution of sites in the Gwynns Falls Valley. 17

depicted in Figure 6 àre within twenty to thirty minutes walking distance from the blackjack oak unit of the serpentine barrens of

Soldiers Delight. An understanding of the subsistence value of the forest units requires the correlation of data from Brush (1975) and

Thomas et al. (1975) to understand the biota of each vegetation unit.

Sycamore-Box Elder-Green Ash-Silver Maple Unit:

Green ash, box elder, sycamore and silver maple occur in order of decreasing dominance. Black walnut is a co-dominant species. Red maple, black cherry and ironwood occur in twenty percent of the sample areas (plots) surveyed by Brush (1975: 27). Pin oak and tulip poplar are infrequent species. Spicebush is the dominant shrub.

Forget-me-nots, lady's slippers and cattails are also reported

(Ramsey 1972: 7). The animal inhabitants range from insects, turtles, snakes and ducks (black), to deer, weasel, possum, muskrats, fox squirrel and formerly bear, mink, otter, turkey and beaver (Ramsey

1972: 1). Because of the greater natural browse in this unit, browse-oriented species such as deer find this an optimum feeding habitat. The superior browse and cover of this unit would be sought in the winter and may have served in the Gwynns Falls Valley as the route taken by deer herds to reach the rich browsing area of Soldiers

Delight (Figure 6).

Almost all floodplains of second order or larger streams are characterized by this unit which does not occur elsewhere in the

Piedmont. Headwater and first order streams are characterized by an upland type vegetation containing considerable numbers of red maple and none of the species characteristic of floodplain vegetation

(Brush 1975: 30). The Gwynns Falls, Red Run and Horsehead Branch 18

are second order streams while the small tributaries such as the one flowing past 18 BA 125 and 18 BA 127 are first order streams. The areas around most stream heads are wet, supporting the red maple vegetation discussed above. These differences may reflect differ­ ences in the hydrology in and on the alluvium of the larger streams although further research is needed to define the relationships.

The Gwynns Falls, Horsehead Branch and Red Run floodplains contain a sycamore-box elder-green ash-silver maple unit which is frequently associated with the broad floodplains underlain by the Cockeysville

Marble formation (Figure 6).

Tulip-Poplar Unit:

The dominant species in this association is tulip poplar with various oaks, including northern red, white, black, and scarlet the most common associated species. Red maple and dogwood occur in half of Brush's (1975: 25) sample plots with hickories and black gum in a third and beech in slightly over a quarter. The understory consists of Virginia creeper, spice-bush, southern arrowwood, and maple-leafed viburnum. Grape, azaleas and mountain laurel also occur. The percent of ground area occupied by trees is 0.28. Because of the heterogeneous nature of the mast-producing species in this forest unit, grouse, woodcock, deer, squirrel, fox, racoon, bear, elk, and turkey occur.

The more poorly-drained soils of the Lower Wissahicken Schist support the tulip poplar unit.

The tulip poplar unit occurs on almost all litholgies (sic) of the Piedmont except the serpentinite, gravel, and quartzite. The unit characterizes some of the bottomlands of the Upper Wissahicken Schist, but does not occur on the uplands where the soils are quartzitic (Brush 1975: 30-31). 19

Chestnut-Oak Unit:

Chestnut oak is the dominant species with the co-dominant species of black oak and white oak present. Black gum, northern red oak, and dogwood are also present. Dominant shrubs are blueberry, mountain laurel and azalea with Virginia creeper and raspberry frequently present. Percent of basal area to ground area is 0.25

(Brush 1975: 26). The unit is considerably more homogeneous than

the tulip poplar unit. All of the animal species noted for the tulip poplar unit are present but because of the larger amount of mast-

producing species, the unit is an optimum habitat for bear, squirrel

and turkey. Deer are seldom found in the unit during winter due to

lack of food and cover. The homogeneous nature of the unit is

unattractive to deer who range into this unit primarily during late

winter or early spring when they increase their home range (Thomas

et al. 1975: 14).

Occurring rarely on wetter areas, the chestnut-oak distribution

is restricted mainly to uplands and slopes of the Upper Schist, the

gabbro upland, sedimentary rock of the New Oxford formation, quartz­

ite of the Setters formation and gravel located near the fall line.

The unit is more closely correlated with the Mt. Airy-Glenelg-

Linganore Soil association than with the Upper Schist from which

the soil is derived.

Blackjack Oak Unit:

Blackjack oak is the dominant species with Virginia pine a

codominant species. White and post oak, black oak, northern red oak,

red maple and black gum are also present. The dominant shrubs are

various species of blueberry and greenbriers. Percent of basal area 20

to ground area is 0.21 (Brush 1975: 26). Cleaves, in his intense geologic study of the serpentine barrens of Soldiers Delight, characterizes the vegetation as follows:

Two major vegetative areas characterize the watershed, a xerophytic woodland, and a barrens. The barrens occur along ridge tops and ridge sides, have a very pebbly to bare-rock surface that is incompletely mantled by grasses, and are dotted by isolated scrub pine. Soil is lacking except as a filling between pebbles or in rock cracks or as a thin accumulation beneath clumps of grass. The barrens areas comprise 8.7 percent of the watershed. Woodland covers 78.9 percent of the watershed and consists of scrub pine and various species of oaks. Cedar is rarely present. Greenbrier, Virginia creeper, and poison ivy grow in profusion and form nearly inpenetrable thickets. The remainder of the watershed is pasture (6.7 percent) and right-of-way for an electric transmission line (5.7 percent) (Cleaves 1973: 54).

The serpentine areas may contain sundrops, tufted hair grass, black­ berry, meadow-sweet, chickweed, flame flowers and moss pink (Harsh- berger 1903: 342).

The diversified flora of the serpentine barrens supported a

large fauna. Spencer (1906: 143) states that the area abounded in

game and was a hunter's paradise. The interfluvial areas supported

a large rabbit population, the black oak acorns drew squirrels from

all the surrounding area and, every September, great clouds of

passenger pigeons used to descend upon it for acorns and sassafras

berries (Spencer 1906: 143). Wolves and wild deer were present up

until the middle of the late eighteen hundreds (Spencer 1906: 148).

Browse-oriented species such as deer and elk find this an optimum

feeding habitat. The stunted vegetation in the interfluvial area

provides excellent cover for bedding. The large variety of mast-

producing species in the interfluvial area also provides an optimum 21

habitat for such mast-oriented animals as bear, squirrels and turkey in addition to the passenger pigeon mentioned above.

The blackjack oak unit is restricted to the serpentine area of

Soldiers Delight and the Bare Hill barrens located on the Patapsco

River (Figures 4 and 6). The absence of a saprolite layer combined with the thin soils of the serpentine areas results in i dry environ­ ment supporting species which can survive under such environmental conditions. The stability of this unit and of other vegetation units described above will be elaborated upon in the ensuing discussion of the affects of cultural disruption of the units during the historic period. For elaboration on the relationship and composition of the vegetation units the reader is referred to Brush (1975). For an understanding of the carrying capacity of the units and of the habitat requirements of the fauna found in each unit, the reader is referred to Thomas et al. (1975).

Stability and Change of Vegetation Units

To what extent do the present forest units of the study area represent the forest units which existed at the beginning of the

Historic period? An answer to this question requires not only historic research but also an understanding of the stability and change of m o d e m forest units. We have already noted that species distribution is environmentally controlled, that permanent changes in the composition of the vegetation units are caused by permanent changes of the environmental variables which control species distribution.

Given relatively stable climatic conditions (such as the warm­ ing trend which has characterized the climate of the region since 22

A.D. 1600) there is little reason to expect changes in species composition due to changes in the environment caused by forest composition Itself.

Since the basic components of a mosaic unit are geologically and topographically controlled, the species occupying a site can be expected to have little influence on the basic composition of the mosaic.

The fact that the composition of the forest is uniform regardless of size within units indicates that at least within a time period of 100 years, no trend to replacement of species and change in composition is evident. Hence there is little reason to expect changes in species composition due to changes in the environment caused by forest occupation of itself. (Brush 1975: 43).

In forest units where open space becomes available through such processes as trees being blown over by the wind, the spaces are occupied by a seedling and small tree composition similar to that currently occupying the forest unit (Brush 1975: 44). When species of one vegetation unit occurs within another vegetation unit, they do so regardless of size of the tree. This indicates that the occurrence of species of one unit located within a larger unit reflects variability within the environmental unit rather than variability brought about by competition (Brush 1975: 40). Compe­ tition for sunlight may affect forest composition by limiting the number of individuals rather than species that reach maturity. The presence of forest does not appear to alter the environmental variables, thus resulting in stability of forest units given

stability of climatic and cultural factors.

Subjection of the environment to natural disturbance may be of

a temporary or permanent nature. Unless accompanied by other changes 23

in the environment, the devastation of forest fires, floods and windstorms is not likely to have a lasting effect on species dis­ tribution. Only those disturbances which permanently alter the basic characteristics of the environment will permanently alter the distribution of species.

Siltation of the Gwynns Falls floodplain during the Historic period may have altered the water characteristics of the soil suf­ ficiently to create a new habitat which was subsequently occupied by species suited to that habitat. Before the flood of 1868 and subsequent floods, the Gwynns Falls channel was of greater depth and of a rockier nature than it is today. The siltation from the floods combined with the stripping of the forest cover from the upland resulted in a decrease in the volume of water in Gwynns Falls as the channel silted and many of the smaller tributaries dried up due to a drop in the water table (Scharf 1971: 16). The decrease in moisture available to plants in the floodplain and upland areas would result

in the development of forest units composed of species requiring

less moisture such as the tulip poplar or chestnut-oak units. The

upland swamp areas of many smaller tributaries may have been wetter,

supporting species of the sycamore-box elder-green ash-silver maple

unit.

Disturbances which do not alter the structure of the habitat may result in no change or in a temporary adjustment. In such instances, given sufficient time, the forests can be expected to resume their original composition so long as the seed source is maintained. (Brush 1975: 45-46).

Disturbances induced by man will affect species distribution in a

manner analogous to natural disturbances. The duration of the 24

alteration will depend on whether or not the environmental variables have been altered.

For example, the hunting practices of the Potomac Creek complex

Indians resulted in repetitive firing of the forest of the Eastern

Piedmont region (Spelman 1613; Smith 1612; In Arber 1910). Vast barrens resembling the blackjack oak unit of the serpentine barrens

resulted (see Chapter 6 for a brief discussion of these hunting practices). Dr. Charles Carroll of Annapolis noted around 1753 that:

about thirty miles from the Navigable Water is a Range of barren dry Land without Timber about nine miles wide which keeps a course about North East and South West parallel with the mountains thro this province Virginia & Pennsilvania (In Marye 1955: 17).

Carroll's observation of the vast extent of the barrens confirmed an

earlier eye witness account written by Philemon Lloyd. Lloyd was

concerned that the "large Barrens, many miles over” isolated the

settlements along the fertile Monocacy River Valley from the eastern

coastal communities. He notes that "from the Heads of Patapsco,

Gunpowder & Bush Rivers, over to Monockawey, is a Vast Body of

Barrens; thb (sic) is, what is called so, because there is no wood

upon it; besides Vast Quantities of Rockey Barrens" (The Calvert

Papers, In Marye 1955: 16). An independent search of the land

records by Marye revealed that Carroll's observations that the

barrens were once nine miles wide is confirmed by the land records

(Marye 1955).

Marye's research indicated that the present area of Soldiers

Delight is a small portion of the "Patapsco Barrens" or the "Soldiers

Delight Hundred." The Hundred's eastern boundary line was the Old

Court Road, an old Indian path (Marye 1920), which extended from

Elkridge Landing (?) across Baltimore County to Joppa. Along the 25

area of the present Reisterstown Road, the Soldiers Delight Hundred met the Back River Hundred. The Northwest boundary was located between Westminster and Taneytown and north to the Pennsylvania line

(Spencer 1906: 144). This area was avoided by the early settlers because of its barren appearance.

Generally, the northwestern boundary was along the head streams of the Patapsco, the southeastern boundaries were near the

Susquehanna River, the head of Winter Run and the head of Western

Run (Marye 1955: 24-35). This area corresponds to the present dis­ tribution of the chestnut oak unit and the stands of tulip poplar and mixed oaks exclusive of chestnut oak (Figure 4). The relative dryness of this area may have been responsible for the development of the barren areas which may have resembled the blackjack oak unit of the serpentine barrens. This entire area was apparently a vast open barren in Lloyd's time but was already reverting to forest cover by the 1750s. As the presence of a saprolite mantle and the resultant abundance of soil was conducive to reforestation, once the

retardation factor of fire was discontinued, the non-serpentine areas

developed into mature forest except in those areas placed under

cultivation. In the serpentine areas the loss of soil resulting

from the firing would have lessened or eliminated the moisture avail­

able for plant development.

Emery Cleaves' detailed study of the serpentine barrens of

Soldiers Delight indicates that the environmental variables of this

geologic unit are very sensitive to disruption. Cleaves' study area

contained 56.7 hectares of gently rolling land drained by a single

perennial stream, supplemented by one large and a number of small 26

ephemeral streams. The serpentine at Soldiers Delight and other

serpentine areas lacks a saprolite mantle (Cleaves 1973: 55).

Saprolite is the weathered or altered layer of the earth's crust which provides the raw material for the soil formation process. The

lack of a saprolite mantle results in the thinness or absence of

soil in the serpentine areas.

In the barrens area, soil is generally absent. In the forested interfluvial areas test pits and road cuts indicate that the depth to rock is generally three feet or less. By way of contrast, vast areas of crystalline rock of the Eastern Piedmont of Mary­ land are mantled by saprolite. In felsic and mofic metamorphic rocks, saprolite thicknesses vary 20 to 40 feet or more beneath the uplands. (Cleaves 1973: 55-56).

Area underlain by saprolite reverted back to forest cover after burn­

ing by the Indians was discontinued but apparently was not so in the

serpentine areas.

The lack of soil and the nature of the rock of the serpentine

barrens have discouraged cultivation in these areas, although place

mining has disturbed the interfluvial area, probably resulting in

much loss of the soil. During the period of Cleaves' (1973: 56-57)

study, significant inorganic particulate load was not carried by the

stream nor was an accumulation of sand, silt or clay found on the

floodplain of the serpentine area streams. A large seasonal vari­

ation in base flow discharge and a high percentage of flood flow

relative to base flow is attributed to the absence of a saprolite

cover. In areas of thick mantles of saprolite, the saprolite acts

as a water storage reservoir (Cleaves 1973: 63). The absence of

water retention must have been noticed by the early agriculturalist

who described the barrens as dry land (Marye 1955: 17). The absence 27

of a thick soil layer and of alluvial floodplain deposits is attri­ buted to the weathering of the alumina-poor serpentine predominantly by solution, leaving little accumulation of secondary minerals.

"Consequently, areas of the Eastern Piedmont underlain by serpentine

are characteristicaly (sic) barren areas, with little or no soil

cover" (Cleaves 1973: 75). Soil development is further retarded by

lack of aluminum which, when combined with silica, forms clay minerals.

Large portions of the constituents of the primary rock go into

solution and are removed from the watershed (90 percent) (Cleaves \ 1973: 80). Deflation may remove soil Irom the watershed in the

barren areas.

As noted above, these geological conditions are repeated in

other serpentine areas along the eastern Piedmont. Similarities in

the vegetation units of the serpentine areas confirm the fact

(Harshberger 1903; Pennell 1910). The serpentine barrens of Maryland

and Pennsylvania are biological phenomenon reflecting geological

limiting factors which can be projected through time. Changing

climatic factors would alter the vegetation and man's influences

through the use of fire would also result in a change in the distri­

bution and composition of species but the underlying environmental

variables outlined above would ensure that this would always be the

driest unit in the Eastern Piedmont region.

While the serpentine barrens may have supported a thicker soil

mantle before the extensive burning of the protohistoric period, the

lost soil in this area resulting from deforestation would take a

long time to redevelop. Studies have never been undertaken on the

subject; however, geologists at the Maryland Geological Survey concur 28

that reforestation of the serpentine barren would be a very long process, taking from 100 to 200 years. Whether or not the serpentine barrens supported a chestnut oak forest before extensive firing is unknown. But if, as a result of the repeated firing, soil was lost from the barrens then the present stunted vegetation of the area can be partly explained. Again, whatever the prefiring characteristic of the barrens was before modification by the Indians, the unit would throughout prehistory be dryer than surrounding areas.

The effect of forest harvesting will affect species distribution if the environmental factors which control distribution of species are altered. Brush's (1975: 38) and Braun's (1950: 252) studies indicate that sprout forest continues the composition of the original stand thus allowing a consistency in composition over time. However, the construction of logging roads or the erosion of exposed areas may alter drainage patterns which in turn may provide a habitat for new species.

The effect of the chestnut blight did not result in the shift of species from one unit to another. The only effect of the blight was the demise of the species which was formerly most prominent in

the well-drained areas of the region (Maryland Geological Survey

1929: 395). Brush (1975: 24) indicates that the chestnut was contin­

uously distributed throughout the Eastern Piedmont region and has

been replaced by those species which occurred in proximity to it.

The data indicate that the present vegetation units represent

a continuity valid for at least 100 years and probably valid to

A.D. 1750 after the area had been reforested. The composition of

the forest units in prehistory can only be inferred. Apparently in 29

the area of the Gwynns Falls, constant firing by the Indians of the

Potomac Creek complex and possibly of the Townsend complex resulted in alteration of the vegetation units of the area. Unless this firing changed the environmental variables affecting species distri­ bution which may have been the case in the serpentine areas, then we expect that the present forest distribution continues back in time to a period of substantial climatic change.

Paleoenvironment

Projection of the present mosaic of vegetation units into the past becomes very difficult because paleoecological data are totally absent from the Eastern Piedmont region. The nearest sources of detailed paleontological and geological data are the Ridge and Valley province of Virginia, Maryland, and Pennsylvania (Carbone 1974) and the Coastal Plain province at the mouth of the Chesapeake Bay

(Whitehead 1967). Our understanding of the climate and vegetation changes from the Late Glacial period to the present in this area is in an adolescent stage of development.

While the composition of changing vegetation units responding to changing climatic conditions cannot be adequately projected, the existence of a vegetation mosaic throughout prehistory is suggested by the paleontological data. The environmental parameters which affect vegetation development today represent a continuum of para­ meters of the past. Thus the serpentine areas of the region were always the driest units in the Eastern Piedmont region. Increase in the moisture or decrease in the temperature during any particular period may have affected the moisture available and drainage patterns of the habitats, resulting in changes of the distribution of species. 30

Generally, a gradual rise in temperatures occurred from 10,000

B.C. to 3000 B.C. when a thermal maximum was reached. The tempera­ tures during the thermal maximum may have been four degrees Farenheit higher than the present. Temperatures steadily dropped until 500

B.C. and then began to gradually rise again until A.D. 1100 at which time another drop occurred which lasted until A.D. 1600 (Fitzhugh

1972).

As discussion about the cultural element of the Eastern Piedmont region will begin with the earliest period of occupation, analysis of the biological element of the region will begin with the Full

Glacial period. Recent excavations at the Meadowcroft Rockshelter near Avella, Pennsylvania, have yielded charcoal radio-carbon dated

to 14,250 B.C. (Adovasio et al. 1975). The charcoal was associated

with flakes, tools, and faunal remains, making this the oldest well

documented date of man in North America. Although evidence of man

in the Easteim Piedmont province definitely dates to around 8000

B.C., analysis will begin at 19,000 B.C. to provide a basis for

further research in case evidence of early man eventually is re­

covered in the area. The analysis will incorporate data on sea

level change as the formation of the Chesapeake Bay affected the

climate of the adjacent Eastern Piedmont region and the resources

available within the general area.

Stage I - After 19,000 B.C. but before 14,000 B.C.: A marked cool­

ing, lower evaporation rates, and a possible increase in precipita­

tion probably took place and corresponds with the Full Glacial

period and the Wisconsin Maximum. During this period sea level fell

rapidly, reaching a maximum depth of about -130 meters at the edge

of the continental shelf. When the continental glacier invaded the 31

headwaters of the Susquehanna River, the resultant increase in volume and velocity caused gravels to be deposited and alluviation to occur downstream (Butzer 1973: 123). Mammoth and mastodon remains found along the submerged continental shelf off of New

Jersey probably date to this period when the continental shelf was exposed (Emery and Edwards 1966). Megafauna in the Piedmont region may have been similar to the megafauna found in Saltville, Virginia.

The deposits in Saltville which were C-14 dated to 13,460 t 420 years

B.P. include the long-armed ground sloth, mastodon, woolly mammoth, caribou, moose, bison, woodland musk ox and musk ox. Other animals which have been reported from late Wisconsin times are the white tailed deer, the giant as well as the modern beaver, and two types of extinct peccaries (Carbone 1974: 94-95).

The climate of the Piedmont probably became more continental as the warming effect of the Bay was removed with the retreating waters of the interglacial sea. With the cold climate of the glacial period the faunas took on a distinctly boreal character. Toward the end of this period, vegetation in the Chesapeake Bay region was

composed of coniferous species mixed with birch and alder. In the

area of the lower Bay bridge crossing "the few hardwoods, deciduous

genera were represented by very small amounts of pollen, whereas the non-tree species may have been rather abundant locally" (Harrison

et al. 1965: 220). A pattern of spruce-pine, tundra and beech-birch-

hemlock units with a possible small number of oaks may have existed

in mosaic fashion in the Eastern Piedmont region.

Stage 2 - After 14,000 B.C. but before 9000 B.C.: During this

period spruce decreased in abundance as the pine forest reached a 32

maximum (Harrison et al. 1965: 220).

During the Late Glacial Period the spruce dominance in Virginia shifted to pine and in time the conifers gave way to beeches and birches. Hemlock gained in importance at this time also, while oak and hickory increase in percentage. (Carbone 1974: 89).

Similar changes probably occurred in the Eastern Piedmont region with tundra, if ever present, being replaced by spruce forest and with deciduous forest units. Possibly toward the end of this period the establishment of oak forest would have provided a greater resource base for the prehistoric occupants of the region. The late glacial faunas with distinctly boreal character probably dis­ appeared gradually between 9000 and 7000 B.C. (Carbone 1974: 95).

While sea levels during this period are difficult to project,

"initially sea level advanced rapidly across the coastal plain from

300 to 350 feet below the present sea level at rates greater than several feet sea level rise per century (Kraft 1971: 2131). Milliman and Emery (1968) place sea level at 81 feet below present at the end of this stage. Brackish, shallow-water environments may have resulted near the lower Bay Bridge crossing as a result of the transgressing

Pleistocene sea. Peat deposits from 82 and 85 feet below present received C-14 dates of 10,340 t 130 and 11,590 t 150 years B.P. respectively (Harrison et al. 1965: Table 2). The upper Bay Bridge crossing area would not have been affected by this early marine transgression except for a possible rising of the base level which would increase alluviation and filling of the paleochannel.

The beginning of the period was probably marked by cooling and lower evaporation rates followed around 10,000 B.C. by cool to temperate and moist conditions (Gardner 1974: 30). The end of the 33 period may have been the beginning of the overall drying trend of the late Glacial period (Gardner 1974: 30).

Stage 3 - After 9000 B.C. but before 7000 B.C.: A continuation of the warming trend and of the decrease in precipitation mark the end of the Late Glacial period. The following Boreal period probably may have been cooler and dryer (Gardner 1974: 31-32). Pollen profiles show a decrease in pine and an increase in oak at the lower Bay

Bridge crossing area (Harrison et al. 1965: 219). Megafauna, as discussed above, disappeared during this period.

A C-14 date of 8,135 - 160 years B.P. obtained from oyster shell

taken from the lower Bay Bridge crossing area indicates the estab­

lishment of shellfish resources in that area. The estuary environ­

ment would have extended up the Bay to the present upper Bay Bridge

crossing area.

A stage intermediate between fluvatile and estuarine may have

existed above the present upper Bay Bridge area, resulting in a

slightly smoothing effect on the local climate.

Stage 4 - After 7000 B.C. but before 3000 B.C.: Temperature

continued to rise until a thermal maximum was reached around

3000 B.C. at which time temperature was higher than the present.

Fluctuation in this general temperature rise and in the degree of

moisture also occurred (Gardner 1974: 32; Kerby 1964: 32). During

this period the pine maximum was replaced by an oak maximum with

modern vegetation units possibly becoming established toward the

end of this period.

The Chesapeake Bay developed its modern shape and oyster

middens may have developed north of the present upper Bay Bridge

crossing. Modern fauna was established. 34

Stage 5 - After 3000 B.C. but before 600 B.C.: This period is marked by a general cooling trend although periods of warm and dry climate may have occurred (Flint 1957: 220, 397). Species associations approaching those of modern semblance were probably established. In

the Great Lakes region, the forest composition achieved its modern

character during this period (Cleland 1966). The increase in mois­

ture and cooler temperatures would have favored the expansion of the

tulip poplar as well as the sycamore and other moisture-requiring

species.

Around 3,700 years before present, the sea level rise rate

decreased, rising a half-a-foot per century instead of the foot per

century which began approximately 8,000 years before the present

(Kraft 1971: 2131). The decreased rate of sea level rise would have

favored stabilization of habitats favorable for oyster growth.

Stage 6 - After 600 B.C. but before A.D. 1100: Rising temperatures

characterized this period. Slightly warmer and wetter conditions

than present began the period (Huntington 1945: 543). Toward the

end of the period, the climate began to cool and became drier. The

wetter conditions would have increased stream flow and increased

moisture which would in turn favor development of moisture-requiring

species. Sea level continued to rise causing continuing invasion of

the sub-estuaries with oyster colonies (Wright 1973).

Stage 7 - After A.D. 1100 but before A.D. 1600: The drier climate

of this period may have resulted in the establishment of larger

areas of forest species which require less moisture such as the

blackjack oak and chestnut oak units. The drier climate may have

contributed to the formation of the large barren areas which resulted 35

primarily from repeated firing of the Eastern Piedmont region (see pages 26-28).

The cool dry period of the 1500s was recorded by the early

Jamestown settlers and by the Spanish mission established in 1570 on the James River. The missionaries wrote in 1570 of famine and death which had caused suffering among the tidewater Indians of Virginia and which was probably caused by prolonged droughts andearly or late frost.

We find the land of Don Luis in quite another condition than expected, not because he was at fault in his description of it, but because Our Lord has chastised it with six years of famine and death, which has brought it about that there is much less population than usual...Since many have died and many also have moved to other regions to ease their hunger, there remains but few of the tribe, whose leaders say that they wish to die where their fathers have died, although they have not grain, and have not found wild fruit, which they are accustomed to eat. (Lewis and Loomie 1953).

This reference suggests that the climatic conditions of the sixteenth century may have resulted in major dislocations of the aborigines of the Chesapeake Bay region. That the possible drought reported in Virginia in 1570 was equally severe in the upper

Chesapeake Bay area is supported by the extent of the drought of the

1930s which covered the entire area (Hershfield 1971: 82-83).

During the severe drought year of 1930 streams in watersheds as

large as 18 square miles went entirely dry for several successive weeks and crop and livestock losses were tremendous (Hershfield

1971: 84). While such severe droughts are rare today, their occur­

rence during the cool and dry period from 1100 to A.D. 1600 may

have been more frequent, resulting in major disruption to the

horticulturists of the area. 36

The cold and drought conditions from 1500 to A.D. 1600 may have also resulted in the migration of the buffalo from the Mississippi

Valley and up the Great Valley into Maryland (Kerby 1967: 119).

However, bison remains have not been reported from archeological sites in Virginia or Maryland, although the presence of bison in both states was noted by early travelers and settlers (Kerby 1967 :

116).

Summary

The complexity of the modern mosaic of vegetation units has been reviewed and equated to climatic and culturally induced changes.

The sparsity of paleoecological research in the study area prevented

elaboration of the climatic periods and resultant vegetation units which affected man in this region. The framework developed will

serve as a basis for understanding the development of prehistoric

cultures in this region, specifically within the Gwynns Falls Valley.

Discussion will now turn to description of the data base upon which

explanation of the cultural element rests. CHAPTER THREE

SURVEY TECHNIQUES

Before the initiation of the Northwest Transportation Corridor archeological survey project in 1973, the archeological resources of the Corridor and the surrounding region were virtually unknown.

Published reports of previous investigations were non-existant and the few amateur archeologists active in the area had restricted their investigations to a limited number of sites (Chapter 1, page 7). The past two years of archeological investigations have increased our understanding of over 9000 years of cultural processes and prehistory within the study area.

The excellent historical research conducted by John McGrain

(1974) provides an encompassing account of the historical, archeo­ logical and cultural resources of the Corridor and surrounding region.

As this thesis emphasizes the prehistoric resources directly and

indirectly affected by the various alternate routes of the Northwest

Transportation Corridor, the reader interested in historic sites is

referred to McGrain's report. His report incorporates the results

of the 1973 survey (Clark 1973). Throughout the three phases of the

survey, historical artifacts were collected, analyzed and reported

in the resultant reports (Clark 1973; 1975a; 1975c).

Survey Procedures

The first fieldwork phase in any archeological survey is the

location of sites within the study region (Clark 1973). Once a

reliable and representative sample of the range of variation in sites

within the given region is determined, selection of sites for exca­

vation should be made on the basis of a general research design.

37 38

Selection can be aided by controlled surface collections and limited test excavations which define the spatial distribution of the stylistic and functional classes of artifacts and the nature of the subsurface matrix. Based on the interpretations of this data, excavations can be planned to yield controlled data on cultural features. Completion of this phase of excavation and analysis ful­ fills the "preliminary site examination" phase of investigation as defined in the "Federal Aid Highway Program Manual, Volume 7,

Chapter 7, Section 4." If the test excavations do not encounter a primary archeological context below the plowzone, power equipment should be used to expose a maximum area to facilitate the mapping and excavation of any cultural features found. This procedure should be undertaken if the final route location will directly impact the site. In other cases, alternative mitigation measures may be warranted by the nature of the archeological record and the extent of the previous research conducted at the various sites to be affected. Completion of this phase of investigation discharges the funding agency of their responsibility for providing for "salvage work" as defined in the above cited procedual manual.

The specific methods and techniques used in the archeological

survey of the Corridor were developed during the initial phase of

fieldwork and in the interim period before the second phase began.

The results of earlier phases of the field work were used in design­

ing the methods used in later phases (Binford 1973: 174).

PHASE I - INITIAL SURVEY AND RECONNAISSANCE: Carried out in July

and August of 1973, the preliminary fieldwork was designed to

intensely survey the accessible portions of the direct impact zone 39

of the Corridor as then defined by engineering maps provided by the

State Highway Administration (Alternates 1-4). As regional research designs for Maryland were lacking when research began in 1973, the initial surface survey and site location stages of investigations were conducted within an inductive framework as reflected by the resultant report (Clark 1973). The assumptions directing the initial survey were: (1) Large multi-component sites similar to those found on the Coastal Plain would be found in the Piedmont, floodplain habitat of the Gwynns Falls Valley; (2) Interior exploitive stations would be found along the major tributaries of Gwynns Falls and the

Patapsco River; and (3) Limited exploitive stations, bivouacs, and random finds of projectile points suggesting hunting activities would be found in the upland areas of the Corridor. These assumptions resulted in the concentration of fieldwork primarily to that section of the Corridor falling within the Gwynns Falls floodplain.

The survey attempted to locate as many historic and prehistoric sites as possible given the time and surface cover limitations. All cultivated areas between the Baltimore City line and Painters Mill

Road and all cultivated areas adjacent to streams from Painters

Mill Road to Route 30 were surveyed. The locations of all cultural items were plotted in relation to arbitrary points of reference. A controlled method of collecting was not employed.

In wooded areas, pasture or grass-covered fields, all areas of exposed surface were examined for artifacts. Test excavations, 3

feet square, were conducted in well-drained areas adjacent to open water resources and in all covered areas which produced surface

indications of cultural activity. The dense growth and the heavy 40

siltatlon in the floodplain prevented the adequate sampling of this area. Six prehistoric and six historic sites were located. The survey was not all encompassing; many sites may remain undetected beneath heavy ground cover or buried beneath layers of colluvium and

alluvium. To develop probability statements concerning the presence

of undetected sites, a regional research design was needed.

PHASE II - REGIONAL RESEARCH DESIGN DEVELOPMENT AND CONTROLLED SURFACE COLLECTION:

The preliminary survey was followed by an analysis of the data

and the writing of the preliminary report (Clark 1973). The report

recommended that controlled surface collection of the sites along

Painters Mill Road should be conducted. During the interim period

between the completion of the preliminary report and the resumption

of fieldwork in June of 1974, the results of the preliminary survey

were reviewed. The author then developed a series of synchronic

research hypotheses relating to the development, expansion and demise

of the Potomac Creek complex (Chapter 6). The hypotheses provided

explanations for the possible existence of trade between the Late

Woodland cultures of the Coastal Plain and Piedmont Plateau provinces

with each other and with cultures in the Appalachian province and

further west to the Ohio drainage; the displacement of the people of

the Sullivan Cove phase of the Townsend complex by peoples of the

Ferguson phase of the Potomac Creek complex; the settlement-subsist-

ence pattern of the Ferguson phase; and other matters relating to

the late prehistoric period. Chapters 5 and 6 will elaborate upon

the formulation, testing and revision of the hypotheses developed to

explain synchronic problems relating to the Potomac Creek complex in

the Eastern Piedmont and Coastal Plain provinces. The data from the 41

Painters Mill site was useful in testing hypotheses relating to the interior base camp sub-system of the Townsend and Potomac Creek settlement-subsistence systems and the influences of Piedmont cultures on Coastal cultures before the latter were displaced or assimulated. Many of the original hypotheses formulated in 1974

(Clark 1974b, c) have been rejected or revised, but some remain to be tested. The 1974 investigations concentrated on the prehistoric sites falling within the right-of-way of the proposed Dolfield Road

Interchange which were selected for an intense, two-stage program of controlled surface collection.

The initial stage of the second phase of field research involved the controlled collecting of the surfaces of the sites after the

1973 crops had been removed but before the 1974 cultivation began.

The second stage of investigations involved the controlled surface collection of the sites after the 1974 cultivation. The second collection increased the sample size and served as a control for the first collection. Unfortunately, unforeseen delays in the project resulted in a late start, June, after the fields had been cultivated and planted. While a controlled surface survey was not possible before the spring plowing, Tyler Bastian, with the assistance of several Boy Scouts of America, was able to make a random collection of the field. This survey revealed that the sites 18 BA 104, 105, and 106 which are located in the northern half of the field along

Painters Mill Road, are in reality part of one site which hereafter will be referred to as 18 BA 106, the Painters Mill site (Figure 7).

The limits of 18 BA 112, the Gwynns Falls site, were expanded to include the entire area of occupation in the southern half of the 42 above field. The northern and southern sites are separated by a low-lying swampy area which contains a slight scattering of cultural material (Figure 7). The failure to properly delineate the site boundaries in 1973 is attributed to the heavy ground cover of corn and grass encountered during the August survey. The benefits and the necessity of conducting follow-up investigations of sites located during the initial survey are clearly demonstrated by this example.

18 BA 107 /

Figure 7: Dolfield Road Interchange-prehistoric sites 18 BA 106, 107, 112. Modified from drawings prepared by Rummel, Klepper & Kahl, consultant engineers. 43

The selection of the size and location of the sampling units depends on the nature of the archeological record, the researcher's orientation, and the time and resource limitations of the project.

To a degree, the smaller the size of the sampling unit and the larger the area of the site surveyed, the greater the value of the results.

Ideally, the exact provenience of all cultural items relevant to the research hypothesis should be collected from the entire site surface.

This procedure is costly, time-consuming and theoretically unneces­ sary on large multi-component sites producing a mass of surface material (Ragir 1973). For small component sites like 18 BA 106 and 112, which produce a limited scattering of cultural items, the exact provenience approach to surface collection eliminates the need to develop random sampling procedures while providing the idea base for testing the given hypothesis. By plotting the exact proveniences of the cultural items and features, archeologists can determine what particular items and features are associated, although interpreta­ tion of these associations may fail to account for the specific cultural activities which produced the observed differential distri­ bution.

Following the 1974 cultivation, the Painters Mill and Gwynns

Falls sites were revisited after every rain until the soil was washed off of a sufficient quantity of artifacts. The entire surface area of the field at the beginning of the survey in June was exposed and well-washed.

The Gwynns Falls site (18 BA 112) was planted in a north-south direction for the entire length of the field, and the corn rows served as lanes for orienting the surface survey (Figure 8). Each 44

lane was followed with the location of all cultural items marked with a numbered stake. The artifacts were then placed in a bag and assigned a corresponding number. The artifacts were collected and bagged during the initial marking stage for fear of vandalism. A total of 78 artifacts were collected and marked for the entire site

(18 BA 112). Ten days of constant rain prevented the immediate mapping of the artifacts' locations. In the interim period, the corn grew two feet, obscuring the stake locations and necessitating the resurvey of the entire field. (Stakes numbered 43, 73, 75, and

76 could not be relocated.) Two persons were employed in the survey, one to hold the stadia rod and the other to locate the stakes while a third person, the author, plotted the artifact locations on the site map (Figure 15). Locations were plotted with a telescopic alidade and plane table.

A different procedure was employed for the surface collection of the Painters Mill site (18 BA 106). Following the corn rows, the investigators were instructed to mark each artifact with a tag pinned to a nail and to tie a survey ribbon to the top leaf of the corn stalk nearest the artifact. The artifacts were left in place. This method speeded up the survey process and reduced the loss of artifact locations. However, resurvey of the entire site by the author revealed that debitage, particularly of quartz, was overlooked by the initial survey. As quartz and serpentine lithics occur naturally on the site surface, serpentine was not collected and the marking of quartz cultural items may have been biased. As the collection of all

quartz lithics was beyond the capacity of the project's resources,

in-the-field evaluations of the quartz debitage were necessary. Had

all quartz lithics been collected, the quantitative data would surely 45 reflect the fact. Similarly, fire-cracked rock may have been over­ looked in the survey as the field-vein quartz fractures in planes similar to those resulting from heat treatment. River and streambed cobbles were absent from the surface of the sites. The rhyolite debitage, pottery, and artifacts were adequately marked by the sur­ veyors and represent a valid sample. The field was surveyed two additional times to insure against investigator bias by the various team members.

Figure 8: View (facing southeast) from ridge of the 1975 investi­ gations of the Gwynns Falls site, 18 BA 112. 46

After 18 BA 106 was resurveyed for the third and final time, the artifacts were plotted and collected. Each row was followed, with one person relocating the markers while a second person held the stadia rod. The author plotted the locations on the site map and assigned a number to the artifact (Figure 9). The number was then communicated to the field team who placed the artifact in a bag with the appropriate number. While this method appears to be redundant, had the location and plotting of artifacts been condensed into one survey by two persons, the sample would have had less validity for the reason stated above. Such a survey is feasible given an increased survey crew and a quality controller who could check the reliability of the sample as it is being collected: a procedure adopted during the 1975 investigations (Figure 9).

Figure 9: View (facing west) from Painters Mill Road of the 1975 investigations of the Painters Mill site, 18 BA 106. 47

PHASE III - CONTROLLED SURFACE COLLECTION AND EXCAVATION:

Following the 1974 survey, the various hypotheses developed during the previous winter were tested and revised. The interim report recommended excavation of the Painters Mill and the Gwynns

Falls sites to determine the relationships between the distribution of cultural items on the surface of the sites with the distribution of cultural features below the plowzone. Stratigraphie tests were also recommended (Clark 1975a).

Controlled Surface Collection; The final phase of investigation consisted of additional controlled surface collecting and preliminary excavation of the Painters Mill and the Gwynns Falls sites. The collecting technique developed at the Painters Mill site in 1974 was continued in 1975. The excellent response of approximately 40 volunteers combined with an efficient staff of five archeology students facilitated the adequate sampling of surface and subsurface cultural resources at the site.

The collector bias suspected of the 1974 investigations was confirmed by the 1975 data. In 1974, 35% of the collection from the

Painters Mill site (18 BA 106) consisted of quartz artifacts compared to a combined total of 53% quartz artifacts in 1975 (Table 1). The number of artifacts recovered also varied in response to field con­ ditions, collecting techniques and collector bias. In 1973, 37 artifacts were recovered compared to 146 artifacts recovered in 1974.

In 1975, under ideal collecting conditions and with adequate person­ nel, 477 cultural items were collected from 18 BA 106. The necessity of conducting multiphase research is demonstrated by the above data.

The Gwynns Falls site was not scheduled for a second controlled

surface survey because of the research priorities and the projected 48

time limitations. However, the sparsity of sub-surface features resulted in an extra day for surface survey. Given the time limita­ tion, a grid system was imposed which utilized 10-meter-square units in the areas yielding a clustering of artifacts in 1974 (Figure 15).

Twenty-meter-square and larger units were established in less pro­ ductive areas.

The eastern portion of both 18 BA 106 and BA 112 were cultivated in the fall of 1974 and were well washed by the time of the survey.

The surface of the sites was 100% exposed. Only the area cultivated in the fall of 1974 was subjected to control surface collecting in

1975. The uncultivated, western section of the field was in 80%

cover and was not systematically collected (Figures 15 and 17). The western section was randomly surveyed and cultural items were found

in the wash and at the base of the slopes (Figures 8 and 9).

Excavations: A base line was established from Bench Mark #5

and transected the sites in a southerly direction. The base line

ran the entire length of the cultivated field, bisecting both sites

(Figures 16 and 18). Two meter-square units designated by their

southwest corner were employed throughout the excavation. In squares

spatially isolated from other squares, the plowzone from the south­

west square-meter was passed through a 1/4-inch screen. For the

contiguous squares depicted in Figure 10, only two southwest square-

meters were screened. (The plowzone ranged in depth from 27 to 30

cm at 18 BA 106.) In seven of the squares excavated at 18 BA 106,

all the material not passing through the screen was saved and sub­

sequently wet-screened. For the remaining 12 squares at 18 BA 106

and the four squares at 18 BA 112, the excavator retained all 49

cultural material visible in the screen and discarded the residue. A comparison of the two techniques revealed a 40 to 50% loss of cul­ tural items resulting from the latter method. The high clay content of the soil, combined with the moist soil conditions contributed to the difficulty of recovering cultural items during screening.

66 I DBA I 06 POSTMOLD PROFILES

FEATURE I PROFILE

CLAY' BAND sTl t ROOTLETS LENSE

70

Op.M.3

EROSION

FEA. I îsfcr; ■ 72 p PM 9 ^PM e

20 CM FOR PROFIL! 74 NORTHWEST TRANSPORTATION I M FOR GRID CORRIDOR ARCHEOLOGICAL SURVEY. 1 9 7 3 EXCAVATION

76 50W54

Figure 10: Excavated area, central section, of the Painters Mill site. 50

The standard excavation forms developed by the Division of

Archeology, Maryland Geological Survey were employed. Upon removal of the plowzone, the floor of each square was drawn. Patterned soil discolorations were cross-sectioned and profiles of probable post molds and features were drawn. Excavation into the subsoil proceeded at 10 cm levels with all backdirt screened through a 1/4-inch mesh.

One feature and eight possible post molds were discovered at the

Painters Mill site (Figure 10). Soil samples were taken of the post molds and feature. An auger was utilized to take soil samples

from various parts of the subsurface matrix at 18 BA 106. The

samples indicated that the soil is highly acidic with a low phos­

phorous content. The repeated fertilization of the field has

distorted the soil chemistry of the site and prevented the use of

phosphorous levels to indicate concentrations of cultural activity.

Soil samples were also taken of the various levels exposed by

the strata cut excavated in square S42W48 (Figure 11). This cut

revealed a complex history of colluvial deposits but failed to

reveal any cultural material below the plowzone. A similar deep

cut was attempted at 18 BA 112 at square S302W2 but ground water

prevented completion of the excavation. A buried plowzone was

encountered but cultural features were absent.

Area Survey

As all but one of the sixteen proposed alternate routes directly

affects the Gwynns Falls and Painters Mill sites, the 1974 and 1975

field research focused on the acquisition of an adequate horizontal

and vertical sample of data from these sites to facilitate the

assessment of the extent and significance of the sites. But the 51

8W HW FiOURE II LINE LEVEL J

/PLOWZONE VERY DARK BROWN Sll*- ---

......

STRATIGRAPHY OF IXED z o n e : ORANOE-TAN H BANOrCUAY THE WEST WALL OF S42W4B, IBbAlOB DARK 3TAIN 0 10 20 CM

ROWN SAn P^IAIER.,',

DARK BROWN SAND

rCRAY SAND LAYER 100 :ORANOI9H, MOTTLED CLAl L A Y E R — NORTHWEST TRANSPORT ATI OH CORRIDOR MOTTLED ORAY.ORANBE ARCHEOLOOICAL lEO SURVEY; IS78 EXCAVATION______

MEDIUM BROWN BAND LAYER

140

ORANBE BAND LAYER • . A ' * Co p

Figure 11: Drawing of the profile of the west wall of square S42W48, 18 BA 106. significance of the sites within the direct impact zone could be evaluated only within the larger resource base of the Gwynns Falls

Valley. The 1973 survey located four sites in addition to the

Painters Mill and Gwynns Falls sites. Two of the four sites are located within the right-of-way of the proposed improvements in the

Northwest Transporation Corridor. While the Corridor transects a number of different geologic units (Figure 6), excessive ground 52

cover throughout the various microhabitats of these units often

frustrated the collection of a representative sample of archeological

data for each unit, necessitating the expansion of the boundaries of

the sampling area to include the entire Gwynns Falls drainage from

the Baltimore City line to the headwaters of the Gwynns Falls.

The Gwynns Falls Valley transects a number of different geologic

units. Each geologic unit may support several vegetation units which result primarily from differences in the availability of

moisture with sunlight, wind exposure and nutrient requirements of

secondary influence. But each geologic unit is dominated by a

particular vegetation unit and often only one vegetation unit is

found on certain geologic units. Diversified habitats result in

areas where several geologic units which are dominated by different

vegetation units are in close proximity to each other. If each

geologic unit supports several vegetation units, a complex diversity

of vegetation units within a relatively short distance may result.

While the composition and the extent of the various vegetation

units have changed over time as a result of culturally and climat­

ically induced changes, the mosaic vegetation effect resulting from

the different geologic units would have existed throughout prehistory.

As the various geologic units often support different vegetation

units, man's adaptation to these varying resources would have also

differed.

Areas containing several adjacent geologic units each of which

currently supports a different vegetation unit are hypothesized to

contain a greater density and diversity of prehistoric sites. Areas

containing a diversity of geologic units or only one geologic unit 53 which currently supports only a limited number of different vegeta­

tion units are predicted to contain a low density and diversity of prehistoric sites. The greater the number of microhabitats within each geologic unit, the greater the diversity and density of the

prehistoric sites. Thus the greater the diversity of vegetation

units between and within geologic units the greater the diversity

and density of the prehistoric archeological resources.

To test this hypothesis, surveys of all cultivated fields

adjacent to active water sources of the Gwynns Falls were conducted

over a three-year period. The contents of and the collection

techniques employed at the sites discovered during the survey will

be discussed separately for each geologic unit. The type of material

found will also be briefly summarized with more detailed analysis of

the assemblages presented in the diachronic synthesis in Chapter 5.

The Northwest Transportation Corridor begins within the Gabbroic

Rock formation; an area highly disturbed by urban development. When

surveyed in 1973, cultivated fields were non-existent; necessitating

surface survey of bulldozed areas and eroded surfaces with numerous

test squares excavated in wooded or grass covered areas adjacent to

water sources. The floodplains currently support a sycamore-boxelder-

green ash-silver maple unit while the adjacent uplands support a

tulip poplar unit (Figure 4). Extensive testing within the right-

of-way of the various proposed routes of the Northwest Transportation

Corridor located one prehistoric site (18 BA 100) and no random

finds.

Located on the south side of Gwynns Falls between the Baltimore

Beltway and relocated Old Court Road (Figure 6), 18 BA 100 is known 54

primarily as the historic Howard-McHenry Mill site (McGrain 1974) but limited prehistoric remains have been recovered. Three thirty- inch test squares excavated in 1973 produced two quartz flakes. The test squares revealed a plowzone approximately six inches deep over­ laying a yellow clayey subsoil which was excavated to a depth of 19 inches below the plowzone (four inches below the end of the historic occupation zone). The dense vegetation cover and the accumulation of historic midden prevented the delineation of the extent of the prehistoric remains.

The heavy vegetation and cement cover throughout this geologic unit hindered the sampling but the low density of sites probably

reflects the actual site density. The tulip poplar forest has a

low carrying capacity and contains a limited number of nut trees and

the dependent mast oriented fauna. This environment would not have

provided an abundant nut resource which was a staple in the prehis­

toric diets during the Archaic and Woodlands periods. The absence

of a well developed understory would have provided an unfavorable

environment for herd animals such as deer. While climatic changes

such as the warmer and moisture conditions of the Altithermal would

have changed the composition of the forest unit, this geologic

formation may have resulted in a lower carrying capacity than the

surrounding formations supporting other vegetation units. The low

density of sites in the other tulip poplar forest in the area supports

this idea.

The largest geologic unit in the Gwynns Falls Valley, the Lower

Wissahicken Schist formation, currently supports a tulip poplar

forest unit. Floodplains are rather narrow in the eastern part of 55

this unit while the western portion is characterized by well drained and gently sloping walls in the Upper Gwynns Falls Valley and U- shaped valleys for the smaller streams. Most of the land surfaces in this area are either in pasture or forested but examination of cultivated fields and other exposed land surfaces revealed several random finds of debitage, isolated bifaces, and one prehistoric site, 18 BA 128 (Figure 6).

Site 18 BA 128 was the only prehistoric site discovered in the

Upper Gwynns Falls Valley which is defined as the area north of

Reisterstown Road. The site extends a distance of 50 meters along the Western Maryland Railroad and from 20 to 30 meters inland from the railroad. Rhyolite and quartz debitage was observed scattered throughout the cultivated field when surface surveyed in 1975. A quartz, severely retouched, side notched point assignable to the

Late Archaic period was recovered (Figure 35).

Like the Gabbroic Rock formation, this formation has a low

density of archeological sites. While the floodplain of the Gwynns

Falls presently supports the sycamore-boxelder-green ash-silver maple unit, this formation is rather homogeneous. The low density

of sites supports the hypothesis that homogeneous forest areas in

the Eastern Piedmont province have a low density of archeological

sites. The Lower Wissahicken Schist formation was apparently

utilized for hunting as suggested by the stray projectile points

and occasional flakes recovered from this formation.

The multi-component Hazard site is close to the tulip poplar

and chestnut oak boundary which probably shifted over time in

response to changing moisture availability. This site is also at 56

the headwaters of the Worthington and Gwynns Falls Valleys where prehistoric trails may have converged and crossed over into the

Patapsco Valley.

The Baltimore Gneiss formation in the Gwynns Falls Valley is surrounded by the Setters Quartzite formation with the outer forma­ tion of this geologic complex consisting of Cockeysville Marble in contact with the Lower Wissahicken Schist formation (Figure 6).

Similar geologic formations occur to the north and result in similar topographic features and vegetation units (Figures 3-5). Although the serpentine barrens of Soldiers Delight may have increased the resources available in the Gwynns Falls Valley as opposed to other valleys, areas underlain by similar geologic units should exhibit similar site density and complexity. The Baltimore Gneiss geologic complex would have provided a range of habitats within close proxim­ ity to each other. On this basis alone, a higher density of arche­ ological sites was predicted; a prediction substantiated by the resultant survey data.

Like other geologic units supporting a tulip poplar vegetation unit, the Baltimore Gneiss formation contained a low density of archeological sites. Despite extensive surveys of the cultivated fields and subsurface testing along the Gwynns Falls and minor streams of this formation, only scattered artifacts were found. The

Pearre collection, obtained from the upland portions of this forma­

tion, suggest hunting and limited resource processing occurred in

this area (Chapter 5). Similar evidence was obtained from the upland

portions of the Gwynns Falls Valley. But the broad floodplains and

the steep valley walls which characterize this formation did not

produce any indications of semi-permanent or extended occupation 57

sites. This low density of archeological sites is repeated in other geologic units currently supporting a tulip poplar forest and suggest that areas currently supporting tulip poplar forest may have been of lower carrying capacity or desirability throughout prehistory as predicted.

The geologic units surrounding the Baltimore Gneiss formation have the highest density of archeological sites of any of the forma­ tions in the Gwynns Falls Valley. The Setters Quartzite formation surrounds the Baltimore Gneiss formation and currently supports a chestnut oak forest unit (Chapter 2; Figure 6). This unit is in

turn surrounded by the Cockeysville Marble formation which supports a sycamore-boxelder-green ash-silver maple forest.

Three sites were discovered on the Setters Quartzite formation and are situated along the top of the ridge overlooking the broad

floodplains underlain by the Cockeysville Marble formation. The

boundary between these geological formations is usually distinct,

consisting of a sharp rise from the flat floodplains and then a

leveling off at the top of the rise. This unit provided not only a

dry place for bedding in contrast to the often swampy floodplain

but also provided rich fauna and flora resources consisting of and

depending upon the more drought resistant forest species of the area.

All three sites discovered in this unit are situated in wooded

areas. Test squares revealed a shallow humus topsoil overlying a

yellow subsoil containing angular pieces of quartz bedrock. At all

three sites, the artifacts extended only a few inches below the

shallow humus topsoil. Various sections of this formation were under

cultivation during the survey and produced random finds of debitage 58

and projectile points, particularly around the stream heads which originate in the formation.

The excavation of three 30-inch test squares in 1973 revealed the location of 18 BA 101; situated on a ridge of Setters Quartzite overlooking the Gwynns Falls floodplain (Figure 6). One plain rhyolite and two quartz flakes were found at the juncture of the humus and subsoil. As the site was in woods, the extent of occupa­ tion could not be determined nor could the site be placed in a chronological position.

Site 18 BA 103 was situated in a similar topographic position but was closer to the Cockeysville Marble formation. Located on a knoll formed by the down-cutting of two gullies, the site extends

for a distance of 40 meters in an east-west direction and at least

15 meters from the edge of the ridge. The extent of the site was defined in 1973 by collecting rhyolite flakes from along the bridle

path which transects the site. Two test squares excavated near the

area of greatest surface concentration uncovered debitage to a depth

of ten inches below the present surface. As 87 percent of the

debitage was less than one centimeter in length and as the majority

of flakes have striking platforms which appear to have been detached

from bifacial surfaces, retouching or the final stages of tool

manufacture may have been a primary activity at the site. The site

was possibly utilized as a resting place or an overnight camp which

afforded the occupants time to resharpen their tools. Chronolog­

ically diagnostic artifacts were absent.

Located in the Red Run Valley, 18 BA 126 is also situated along

the edge of a ridge of Setters Quartzite adjacent to the Cockeysville 59

Marble formation (Figure 6). When surveyed in 1975, only one

rhyolite flake was observed on the surface of the site. The surface

is obscured by leaf cover and the site was not tested, but its

location at the mouth of a tributary of Red Run and in a similar

environmental setting as 18 BA 101 and 103 suggest that additional

artifacts are present.

As the Setters Quartzite formation is drier than the surrounding

formations, this area would have supported more drought resistant

tree species, such as nut trees, throughout most of prehistory. The

animals depending on the products of this forest unit, such as

squirrel, turkey and bear, may also have been an attraction to pre­

historic peoples. The dryness of this formation would have provided

suitable camp areas in contrast to the damp and often swampy flood­

plains underlain by Cockeysville Marble. The availability of floral

and faunal resources in the adjacent, broad floodplains underlain

by Cockeysville Marble would have enabled the exploitation of two

vegetation units with a minimum of base camp relocation. But all of

the sites discovered in this area are limited in size and apparently

represent camp sites of limited duration and involving primarily

maintenance tasks. The larger sites are found on the terraces and

colluvial fans of the Cockeysville Marble formation.

The Cockeysville Marble formation has the highest density and

diversity of archeological resources within the entire surveyed

portions of the Gwynns Falls Valley. As large sections of the Gwynns

Falls and Red Run Valleys as well as a small portion of the Horsehead

Branch Valley were cultivated; the larger number of sites reflects

in part the greater amount of exposed ground available for study. 60

Had these areas been in woods or pasture, many of the sites less than

30 meters in diameter probably would not have been located. The broad floodplains of the Red Run and Gwynns Falls Valleys would have provided, during the Woodland period, a dense vegetation cover favored by deer (Chapter 2). Two colluvial fans emerging from the

Setters Quartzite formation in the Gwynns Falls Valley were repeat­ edly occupied from the Early Archaic to the Late Woodland periods

(18 BA 106 and 112). As the survey techniques employed at the

Painters Mill and Gwynns Falls sites have already been discussed, further elaboration is not necessary. The collecting techniques employed at, and the nature of the archeological record at the remain­ ing archeological sites within the Cockeysville Marble formation will be presented.

Only one site, 18 BA 129, was found within the Cockeysville

Marble formation in the Horsehead Branch Valley (Figure 6). The eastern portion of the floodplain is very swampy, forested, and was not surveyed. The western portion was not nearly as broad but was swampy in the area of 18 BA 129. Other sites may exist in the vicinity, but as the remaining portion of Horsehead Branch was in pasture, testing was not conducted. Site 18 BA 129 is situated

along the ridge of a dissected terrace overlooking the swamp lands

of the Horsehead Branch (Figure 12). A base line was established

and all artifacts were collected in relation to the base line. Six

ten-foot squares were controlled surface collected to provide an

unbiased sample for comparison to other sites falling within the

Cockeysville Marble and adjacent formations. Diagnostic artifacts

falling outside the controlled collected squares were collected by THE LEE site: B Bo 129 Controlled surfoce collection Shoded area represent* percentage B of surface exposed Edge of field Boundary between excellent wosti and poor wosti surfoee 480 Swamp 481

20 40 60 eet

485 Poor

Excellent WI60SII0 •23

Poor WIOOS80 Excellent Wl3bS80 Datum 20 " •19

Horse tie od Brancti

Figure 12: Controlled surface collection of the Lee Site (18 BA 129). 62

the exact provenience method. The sample collected from the six squares revealed that 86 percent of the artifacts were assignable to the debitage category with rhyolite artifacts followed in popularity by quartz artifacts. The site was apparently occupied from the Early

Archaic to the Late Archaic periods (Chapter 5). The elevated position of the knoll adjacent to the swampy floodplain of the

Horsehead Branch was apparently an attractive camp spot for Indians throughout the Archaic period.

The northern portion of the Red Run Valley was under cultivation from 1973 through 1975 and revealed five prehistoric sites located along the edge of the first terrace above the floodplain. Although extensively surveyed, the floodplain did not produce any indications of occupation. Scattered finds of debitage and projectile points were found around the stream heads at the juncture of the Cockeysville

Marble and Setters Quartzite formations.

Controlled surface collected by the exact provenience method

in 1974 and 1975, site 18 BA 122 extends along the edge of the first

terrace for an east-west distance of 40 meters and a north-south

distance of 20 meters. Chronologically diagnostic artifacts were

absent although one cryptocrystalline flake knife may date to the

Early Archaic period and one expanding-base drill may date to the

Archaic period (Figure 35: P). Several cutting tools, a core, and

flakes of quartz and rhyolite were recovered but definable activity

areas were not present.

Extending 175 meters in an east-west direction along the edge

of the first terrace above the floodplain and from 30 to 50 meters

in a north-south direction inland, 18 BA 126 is the largest, purely 63

Archaic period site located to date within the Corridor. The site has produced the largest percentage (86) of quartz artifacts for any of the sites from which a relatively unbiased sample was obtained

(Table 1).

Except for a Soldiers Delight chalcedony core and flake which may date to the Early Archaic period, the chronologically diagnostic artifacts from this site are assignable to the Late Archaic period.

Projectile points of the Piedmont tradition (including one specimen of the Lackawaxen Contracting Stem type) were manufactured from both rhyolite and quartz. Two Dry Brook type points of the Susquehanna

Broadspear tradition, which dates to the end of the Late Archaic period, were also recovered. Artifact recovery techniques consisted of intense collecting of all cultural items falling within five

100-foot square units laid out over the site in an east-west direction.

SITE Percentages | No. NUMBER LITKIC MATERIAL CATEGORY CO H4H CO g 18 BA N N H g Ü i g g I 5 I O'I g I i I c/5 g c/5 H I to to

106 52 1 44 I ICO 84 2 16 100 662

112 9 + 87 + 100 92 - 8 100 483

122 37 - 63 - 100 89 2 11 100 35

125A 45 - 50 - 100 93 2 7 100 42

127 86 + 13 + 100 89 + 11 100 176

129 45 3 51 - 100 86 1 14 100 105 AVERAGE 46 I 51 + 100 1 11 100 1503

(+ = less than 1/2 percent)

Table 1: Artifact category and lithic material percentages. 64

The density and diversity of the 181 artifacts collected in 1975 which include scraping, cutting and piercing tools as well as an abundance of debitage (Table 1) indicates that this site was an interior exploitive station for various cultures during the Late

Archaic and possibly earlier periods. Heavy reliance on locally available lithic material is documented by the large percentage of debitage manufactured from vein quartz which occurs naturally in the

area. The presence of a small stream flowing along the western and

southern boundaries of the site was probably the primary feature which attracted people to this part of the topographically homoge­

neous Red Run floodplain.

Located along the edge of the first terrace of Red Run are three

separate clusters of prehistoric remains which suggest limited

occupation sites (Figure 6). Each cluster will be summarized

separately.

Site 18 BA 125A consists of a cluster of feldspar rhyolite,

gray chert and vein and cobble quartz debitage distributed over an

area 35 by 15 meters. One quartz and one rhyolite projectile point

diagnostic of the Piedmont tradition of the Late Archaic period and

the two gray chert flakes assignable to the Early Archaic period

were the only chronologically diagnostic artifacts recovered.

Site 18 BA 125B is located on a slight knoll and produced

rhyolite and quartz debitage as well as a Soldiers Delight chalcedony

flake side scraper which is assignable to the Early Archaic period.

The site extends over an area of approximately 35 by 30 meters.

Site 18 BA 125C is a very dispersed scatter of quartz and

rhyolite debitage and scraping and cutting tools as well as one 65

Soldiers Delight chalcedony flake side scraper. The scatter probably

represents resource processing activities, such as faunal processing, which may have resulted in the gradual accumulation of discarded

artifacts deposited throughout the Archaic period.

The three areas probably represent resource processing stations

and transient camps utilized primarily during the Archaic period.

The two sites in this formation located in the Gwynns Falls

Valley, 18 BA 106 and 112, apparently represented repeated occupation

of environmentally favorable locations over a long period of time by

a diversity of cultural groups. These sites are similar to 18 BA 127

and 18 BA 129. The actual size of any one occupation area may not

have been greater than the 30 meter area found at the smaller sites

along the first terrace of Red Run.

Summary

The field techniques and research objectives of the three phases

of fieldwork of the Northwest Transporation Corridor archeological

survey project have been discussed. The first phase of fieldwork

focused on determining what resources were present. The second and

third phases attempted to determine the extent and significance of

the Painters Mill and Gwynns Falls sites. The exact provenience and

grid system methods of controlled surface collecting were utilized.

A survey of the various geologic units in the Gwynns Falls Valley

was conducted to obtain a comparative sample of data essential for

evaluating the significance of sites within the various alternate

routes of the Northwest Expressway and Rapid Transit systems.

The original hypothesis that the greater the diversity of

vegetation units between and within geologic units the greater the 66

diversity and density of the prehistoric archeological resources has been supported by the available data. Where a number of geologic units supporting different vegetation units occur in close proximity, a high density of archeological sites are recorded. The Baltimore

Gneiss geologic complex (Baltimore Gneiss, Setters Quartzite,

Cockeysville Marble) currently supports three different vegetation units within close proximity to each other. This mosaic affect of vegetation facilitated the utilization of various habitats within a very short distance. The serpentine barrens of Soldiers Delight which support a black oak unit are also nearby and would have been a major attraction to the prehistoric hunter and nut gatherer.

Within the Baltimore Gneiss geologic complex, areas currently covered with a tulip poplar forest have a low density of archeolog­ ical sites while areas covered by chestnut oak have a medium density of sites. The highest density of sites fall within the Cockeysville

Marble formation which currently supports a sycamore-boxelder-green ash-silver maple forest unit. Other geologic formations in the

Eastern Piedmont province which support a tulip poplar forest also have a low density of archeological sites. The Lower Wissahicken

Schist formation occupies the largest portion of the Piedmont

section of the Gwynns Falls Valley and like other geologic units

supporting a tulip poplar forest has a low density of prehistoric

sites.

Within certain geologic formations, higher densities of sites

occur consistently in certain areas. In the Cockeysville Marble

formation, sites are located on the high ground adjacent to the

swampy floodplain. Sites are situated along the edge of the first 67

terrace above the floodplain and on colluvial fans which originate in the Setters Quartzite formation. Unique features such as colluvial fans or tributaries bisecting the homogeneous floodplains of the minor river valleys were a factor in the repeated selection of those sites for utilization as reflected in the higher density and diversity of the archeological remains.

Sites located on the Setters Quartzite formation occur on top of the ridges overlooking the swampy floodplains underlain by the

Cockeysville Marble formation. These areas offered dry places for temporary camps as well as mast producing tree species and the depen­ dent fauna. The various geologic units currently supporting tulip poplar forest have a low density of sites consisting primarily of random finds of debitage or functional tools around stream heads and a limited number of sites along the drier terraces away from the floodplain. Projectile points found scattered throughout the upland areas of the tulip poplar supporting geologic formations suggest the occurrence of primarily hunting activities.

Similar densities of archeological sites should be found in

the Jones Falls, Worthington and Green Springs Valleys which are underlain by similar geologic units. The two sites reported from

the Green Spring Valley are located on the Cockeysville Marble

formation near the Setters Quartzite formation (Division of Arche­

ology, Maryland Geological Survey, site survey file). A higher

density of sites may exist in the Gwynns Falls Valley because of

the presence of the serpentine barrens at the head of Red Run and

other tributaries of the Gwynns Falls.

While the vegetation units changed over time, offering pre­

historic people different habitats to exploit, the mosaic affect of 68

vegetation which is controlled by the underlying geologic formations persisted throughout prehistory. Therefore by studying the density of sites in geologic units currently supporting similar vegetation units, predictive statements on site density can be made; statements which are essential in developing measures to mitigate the adverse impacts of project development on known and unknown prehistoric sites. Future researchers conducting surveys of the Worthington,

Green Spring and Jones Falls Valleys can expect to find a high density of seasonal occupation sites along the first terrace above the floodplain and on whatever colluvial fans may be present. A medium density of sites should be present along the ridges of Setters

Quartzite adjacent to the Cockeysville Marble formation. The sur­ rounding Lower Wissahicken Schist should contain a low density of sites. CHAPTER 4: ANALYTIC TECHNIQUES

Throughout the investigation, the sampling techniques have been designed to minimize selective bias and to obtain a representative sample of the present population of artifacts at each site and from each ecological unit. The methodology changed during each successive stage of fieldwork as the analysis of previous data raised new

questions which required the obtainment of new data to provide

adequate answers. As the obtainment of a truly random sample for most archeological sites is more an "ideal to be striven for than

achieved" (Ragir 1973: 180), the selective bias which may have

affected the artifacts collected from each site has been elaborated

upon (Chapter 3 and Appendix A). The bias resulting from previous

amateur collecting of specific categories of artifacts cannot be

adequately determined although the nature of the distortion can be

predicted and will be discussed later in this chapter. After con­

sidering the post-depositional processes which have distorted the

original archeological record, the distribution of the artifact

classes and types should adequately reflect the location and nature

of past cultural activities, but only if the inferences concerning

the function and the chronological age of the artifacts under study

can be substantiated.

Analytical techniques were developed which insured that the

artifacts assigned to the same class and type shared similar attri­

butes (Rouse 1960: 320). The class of cutting as described below

contains attributes defined on the basis of published replication

experimentation conducted to determine the types of wear resulting

from use and manufacturing activities (Semenov 1973; Ahler 1971;

Tringham et al. 1974). The literature is primarily concerned with

69 70

experimentation on cryptocrystalline material. Although the need for analysis of wear patterns on conchoidally fracturing rocks such as quartz is recognized, the need has not been satisfied (Tringham et al. 1974: 178).

Whether identical wear patterns develop from identical use on different lithic material is a phenomenon which should be tested and not simply assumed. While detailed functional replication experi­ mentation was beyond the time, resources and scope of this report,

samples of Catoctin rhyolite from the Catoctin formation in Maryland

and of vein quartz from the Gwynns Falls Valley were utilized during

eleven man hours of flint-napping. The experiment was conducted to

familiarize the author with the fracturing properties of the lithic

material commonly utilized by the Indians in the Gwynns Falls Valley.

The differences resulting from the use of a soft versus a hard

hammer for detaching flakes was given particular attention. The

simple fact that a soft blow with a hard hammer often produced

identical flakes as a hard blow with a soft hammer became readily

apparent. For example, a lip could be formed by either technique.

As many flakes from the Gwynns Falls Valley appear to have been

formed during the thinning of bifaces, a number of bifacial edges

were manufactured and flakes removed from the edges to determine

what attributes would result. An acute angle between the ventral

surface and the striking platform, an acute flake angle, and a

number of negative flake scars on the ventral surface were the

primary attributes recognized for bifacially retouched flakes. On

rhyolite, flakes were produced during the manufacturing process

which could be mistaken as step flaking or edge crushing resulting 71

from a scraping activity. As a result of the replication experiments, greater care was given to determining whether particular wear patterns were the result of manufacturing as opposed to use activities.

As functional experimentation was not attempted, the functional attributes defined by Ahler (1971) have been adopted as a standard which can be referred to by the interested reader. Even if Ahler's experimental data cannot theoretically be applied to artifacts of a different material, the definitions of wear pattern attributes adequately defined and illustrated by Ahler (1971) are useful as a standard for identifying wear patterns which is readily accessible to the archeological community. The attributes comprising the basic unit of analysis are descriptions of a physical reality whose origin is a matter for debate but whose existence is a matter of fact.

Future researchers who have the benefit of replication experiments will be in a better position to infer the function reflective of

the attribute. But replication experiments can only increase the

acceptability of the inference, not prove that the wear on the

artifact was produced by the same processes which produced the wear

on the experimental specimen.

If archeologists are to explain culture process they must

proceed beyond mere data description, although to do so may decrease

the possibility of developing a "hard" science of material culture

(Dunnell n.d.). As analysis and explanation proceeds through levels,

each building upon the other, an error in identification, an invalid

inference in the lowest level of analysis will compound itself

throughout the remainder of the analysis. To decrease the quantity

of compounded errors, particular emphasis has been given to the 72

correct identification of the attributes as defined and consistently inferring similar function for similar combinations of attributes

(Appendix A ) . If the resulting interpretations are invalid, at least

they are consistently invalid and therefore can be corrected when

additional data become available.

Each of the 1882 lithic and 151 ceramic artifacts in the study

collection from the Gwynns Falls Valley were subjected to the same

analytical techniques. The attribute classes established have been

largely borrowed from previously published methodologically accept­

able terminologies (Appendix A). This procedure foregoes lengthy

definitions which are already better developed in the existing

literature. The development of new definitions is avoided when

adequate definitions already exist which can serve as standards.

Lithic Analysis

As 92 percent of the artifacts studied are manufactured from

lithic material and as faunal and floral remains are absent, lithic

material forms the primary base for interpretation of the past in

the study area. While functional artifacts have been traditionally

emphasized in many reports with only raw counts of debitage provided,

debitage often constitutes the majority artifact category from a

site or from a region. Of the 1882 artifacts studied from the Gwynns

Falls Valley, 87 percent were assigned to the category of debitage.

Elimination of all sites and amateur collections distorted by

selective bias resulted in an average of 89 percent debitage for six

sites with a range of from 84 to 93 percent (Table 1). In comparison,

the total lithic assemblage from the significant, deeply stratified

Meadowcroft Rock Shelter in Pennsylvania consisted roughly of 73

"...75% flaking detritus with the remaining 25% representing finished tools of one sort or another. Of this 25% most are projectile points followed in descending order by knives, scrapers, drills and other forms." (Adovasio et al. 1975: 18).

The high percentage of debitage obtained from the surface of the sites in the Gwynns Falls Valley can be attributed to a number of factors: (1) The volume of detritus resulting from the manufactur­ ing, use, and repair processes operative in the extinct cultural system; (2) The inability of the researcher to recognize wear or the absence of wear patterns resulting from a short use life, accumula­

tion of animal fats or post-depositional processes such as weathering or use of low magnification during analysis; and (3) The repetitive

collection of the surface of prehistoric sites by collectors who

selected for functional categories, particularly bifaces.

The nature of collector bias is suggested by a comparison of

the Hazard to the Pearre collection. Both collections were derived

from upland sites located around the head of small tributaries of

the Gwynns Falls. Both collectors shared an interest primarily in

projectile points followed by other functional artifacts and lastly

debitage exotic to the area or of unusual size or shape. Both

collections consisted of 12 percent debitage with the Hazard collec­

tion containing 52 percent projectile points in comparison to 53

percent in the Pearre collection. While part of the wide difference

between the upland collections and the floodplain collections may be

attributed to prehistoric cultural processes (i.e., hunting versus

more stable settlement in the floodplains) the largest percentage of

the discrepancy is attributed to selective bias. Perhaps a 75-25

percent ratio as suggested by the Meadowcroft data is a more accurate 74

percentage figure to be expected on uncollected hunting camp sites.

Similar selective bias was probably operative in the past at the sites in the Gwynns Falls Valley and one individual was reported to be collecting from sites 18 BA 106 and 112 (name unknown). Collec­ tion of artifacts in the Patapsco River Valley has been popular since at least the 1890's (Jones 1903; Stearns 1943).

Instead of obtaining a random sample from the population of debitage items and conducting detailed analysis on the sample and inferring the variability in the remainder of the assemblage, the author developed a basic attribute list. The basic attributes as defined below have been studied on each specimen. As analysis pro­

gressed the need for additional attributes emerged and the typology was redefined. In particular, greater attention was given to study­

ing the characteristics of the striking platform of flakes. As

analysis was conducted over a three-year period to produce informa­

tion for each subsequent report to the highway department (Clark

1973; 1975a, c), a final analysis was conducted in 1975 to differ­

entiate the various lithic classes not previously defined and to

study other attributes which became important for interpretation of

activity areas.

Segregation of activity areas at 18 BA 106 and 112 required the

analysis of lithic artifacts to determine if raw material types

could be used to isolate matching lithic items (artifacts derived

from the same core, bifaces, etc.). The quartz artifacts are

extremely difficult to match due to the homogeneous structure of the

material. Analysis was restricted to the study of the distribution

of opaque, cobble and vein quartz with further refinements not 75

attempted. Rhyolite exhibited greater heterogeneity, enabling matching of numerous artifacts. But a majority of the rhyolite artifacts was of the "plain" variety. The plain variety rhyolite was impossible to segregate on the basis of physical surface characteristics. The cryptocrystalline material was very heteroge­ neous, enabling adequate segregation of the assemblage. The results of the detailed ceramic and lithic analysis are incorporated into the diachronic and synchronic synthesis presented in Chapters 5 and

6.

Lithic Definition

To adequately interpret intrasite and intersite variability of the archeological resources of a region, the archeologist must study in detail the lithic materials from which the artifacts were manu­ factured. The lithic preferences of prehistoric cultures changed over time. Various cultures preferred specific types of material, but when a short-time need arose, locally available materials were often pressed into service. Thus vein quartz and Soldiers Delight

chalcedony, the locally available lithic materials of the study area, may have been frequently used to replace exhausted items of the tool

kit. Cultures which preferred fine grained quartz (opaque and cobble

types) may have also readily utilized vein quartz in the absence of

the more easily worked quartz types. Spectographic analysis was

conducted on eleven flakes to supplement the visual attributes of

the various lithic types and to determine if the visually established

classes are supported by the chemical elements.

Cryptocrystalline minerals

Soldiers Delight chalcedony (Figure 13: 8, 10-11): The 76

serpentine barrens of Soldiers Delight contain deposits of chalcedony minerals which occur in vein and cobble form. A crusty surface rang­ ing in thickness from two to four millimeters is found on most specimens. Colors cover the entire spectrum but white and reds predominate. A majority of the specimens observed are mottled in appearance. Impurities run throughout the material.

Spectographic analysis of three artifacts inferred to be manu­ factured from Soldiers Delight chalcedony did not reveal the same elements for the specimens. A mottled, white Soldiers Delight chalcedony flake contained the major elements of silicon, iron, nickel and chromium (Figure 13:8). The presence of chromium clearly documents the derivation of the material from the chromium-rich deposits located within the serpentine barrens. This material is shown in Figure 31: H. A flake from 18 BA 106 and a flake from

Figure 13: Spectographic read-out on lithic specimens from the Gwynns Falls Valley. 77

18 BA 127 are visually identical and inferred to be derived from the same core. The flake from 18 BA 106 (Figure 13: 10) has a trace of Ù aluminum, a minor amount of iron and a major percentage of silicon.

The flake from 18 BA 127 (Figure 13: 11) contains a minor amount of

iron and a major amount of silicon. Although the flakes were not

stiruck from the same core they probably were derived from the serpen­

tine barrens as similar samples were observed in situ at Soldiers

Delight. Other examples of this material are pictured in Figure 35:

E; Figure 31: H, J; Figure 35: I-K.

Brown jasper: A homogeneous and siliceous material, brown

jasper may be related to cobble jasper found in the Coastal Plain

province or may have come from jasper deposits in Pennsylvania. The

material is medium in texture, is free of impurities, and exhibits

excellent fracturing properties.

Point of Rocks jasper: This is similar to brown jasper in

texture and color but contains veins of highly siliceous material

and exhibits medium fracturing properties. While the material may

not come from the Point of Rocks outcrop, the material does resemble

samples taken from that outcrop. Flint Run jasper is similar to

Point of Rocks jasper but the material contains a higher silica

content, and the colors are richer in hue. A specimen of this

material is pictured in Figure 39: C.

Brown chert (Figure 13: 9): The dark brown color of this

mineral is caused by a high iron content. Brown chert is a homo­

geneous, fine-textured mineral with a very thin (1 mm), smooth

cortex. The material has good fracturing properties. Spectographic

analysis of one flake (Figure 13: 9) revealed a very low amount of

aluminum, a minor amount of silicon and a major percentage of iron. 78

Specimens of this material are pictured in Figure 31: A-C, I, and

Figure 37: A.

Gray chert (Figure 13: 9): A compact, smooth and highly siliceous material, gray chert ranges in color from gray to tan.

The colors are solid and impurities are absent. The material has excellent fracturing properties. While the sources are unknown, the material is similar to minerals common in the Appalachian Plateau province. Spectographic analysis of one flake (Figure 13: 5) revealed silicon to be a major element with a very low percentage of iron. Artifacts manufactured from this material are shown in

Figure 34: A, I.

Quartz minerals

Vein quartz: This type of mineral is found throughout the

Eastern Piedmont region. The material fractures poorly; the force diffusing rapidly and affected by various cleavage planes which occur throughout the material. The cleavage planes are opaque when held at certain angles to a light source. The material occurs in angular form with subangular edges although an occasional cobble of vein quartz is found on the surface of prehistoric sites. A direct source of this material in the Gwynns Falls Valley is the Setters Quartzite formation. See Figure 35: H, M-N for examples of cleavage planes in vein quartz.

Cobble quartz: The presence of a cortex is the primary diag­ nostic characteristics. But whereas vein quartz is generally white or tan in color, cobble quartz occurs in a variety of colors.

Cobble quartz has a finer texture and better fracturing properties

than vein quartz. Cleavage planes are absent. Specimens with a 79

cortex but exhibiting attributes of vein quartz are assigned to the cobble quartz category to denote that the material has been trans­ ported. Cobble quartz occurs in abundance in the Coastal Plain and rarely occurs in the stream channels of minor rivers in the Piedmont province.

Opaque Quartz (Figure 13: 4): A majority of samples exhibit structural properties similar to cobble quartz although a few exam­ ples contain cleavage planes. All specimens are white in color and exhibit excellent fracturing properties. The material is fine

textured. Spectographic analysis of a flake of opaque quartz revealed

a major percentage of silicon with a trace of iron (Figure 13: 4).

Metarhyolite minerals

Feldspar rhyolite: The distinguishing feature is the occurrence

of numerous phenocrysts of white feldspar and clear quartz. Mottled

surfaces with various shades of gray predominate although plain

surfaces also occur. Spectographic analysis of one flake of this

material (Figure 13: 3) revealed the major elements of silicon and

sodium with a minor percentage of iron and a trace of aluminum. The

material is found in boulder and bedrock form throughout the Catoctin

formation in Maryland and Pennsylvania. Boulders of this material

are intermixed throughout the Catoctin formation with boulders of

plain rhyolite (Ayers 1972: 28-33). Specimens of this material are

shown in Figure 31: P, V; Figure 32: Q.

Banded rhyolite: While this material may contain phenocryst,

the rhyolite is banded instead of mottled or plain. The banding

results from movement under intense pressure and heat. Specimens of

this type are readily assigned to specific varieties. Banded rhyolite 80

is probably found throughout the Catoctin formation. Examples of banded rhyolite are shown in Figure 31: Q; Figure 36: N.

Purple rhyolite: A red porphoritic metarhyolite, purple rhyolite is a moderate red to grayish-red or purple rock. Numerous phenocrysts of white feldspar and reddish-colored quartz are often found in the material. While two major deposits of this material have been separately mapped in the Catoctin formation near Cavetown and Jerusalem, Maryland (Jonas and Stone 1938), purple rhyolite is found in various places in the Catoctin formation (Ayers 1972: 28-29).

The Catoctin formation is a band of Precambian metamorphic rock types which extends from southern Pennsylvania through Maryland and into

Virginia. The major metarhyolite outcrops in Maryland occur in the area of the present day Catoctin National Park or in a southwest running deposit located north of Alternate Route 40. Purple rhyolite artifacts from the Cwynns Falls Valley are grouped with feldspar

rhyolite specimens for tabulation purposes.

Black rhyolite: Banding, mottling and phenocryst inclusions

are absent. The material is fine-grained and exhibits excellent

fracturing properties. Colors range from dark gray to black. Specto­

graphic analysis of two flakes revealed that both contained a major

percentage of silicon, with minor percentages of iron and sodium and

a trace of aluminum. Examples of this material are shown in

Figure 35: K, T, U; Figure 34: T.

Blue rhyolite: This material is grayish-blue or bluish-gray in

color. Excavated specimens are generally of darker shades but a

blue hue is still noticeable. The patinated specimens are non-

uniform in color, the results of varying substructural composition 81

of the material. White feldspar crystal inclusions vary in size between and within specimens. Inclusions of clear quartz crystals are present. Blue rhyolite, free of fractures or faults, is composed of fine grains which result in a smooth texture. Blue rhyolite occurs throughout the Catoctin formation (Ayers 1972: 30-32) and was extensively traded during the Selby Bay phase of the Middle Woodland period (Clark 1974d). Specimens of this material are rare in the

Gwynns Falls Valley and are tabulated with feldspar rhyolite speci­ mens for purposes of this report.

Plain rhyolite (Figure 13: 6-7): A catch-all category for rhyolite material which lacks any of the above distinguishing charac­

teristics. This category is composed mostly of light to medium gray specimens. Spectographic analysis of two flakes revealed a major percentage of silicon and sodium with minor amounts of iron and a very low amount of aluminum (Figure 13: 6-7). The presence of sodium

as a major instead of as a minor element indicates that plain

rhyolite may not simply be a weathered form of black rhyolite. The

homogeneity and the color of the material hinders the segregation

of varieties.

Plain rhyolite occurs throughout the Catoctin formation as

documented by the author's survey (8 man hours) of the Catoctin

formation in Maryland and by previous surveys by Ayers (1972: 27-33).

This material can be obtained from boulders abundantly found in the

Catoctin formation in Frederick County, Maryland or around Caledonia

State Park in Adams and Franklin Counties, Pennsylvania. The

material was probably quarried from the boulders instead of from

bedrock. Although quarry sites were not located during the author's 82

survey, attempts to reduce a large bedrock exposure with a large hammerstone proved fruitless. In contrast, a boulder of feldspar rhyolite 20 cm thick by 60 cm square was reduced into a pile of excellent flakes after 15 minutes of work. Many of the flakes struck from alternating sides of the flat boulder are identical to

flake blanks characteristic of the Selby Bay phase and possibly of

the Archaic period as well (Clark 1974d; Geasey 1974: Figure 4).

Because of the proximity of the Gwynns Falls Valley to the

Maryland sources, most of the rhyolite in the study area probably

came from the Catoctin formation in Maryland. The Catoctin formation

is located forty air miles from the sites in the Red Run-Gwynns Falls

Valleys. An Indian travelling on foot from the study area could

probably reach the nearest rhyolite sources in a minimum of three

days. Determination of the cultural processes resulting in the high

combined percentage (55 percent) of the rhyolite at the sites in the

Gwynns Falls Valley will require an understanding of the subsistence

settlement patterns of each culture utilizing the area; a task for

future researchers.

Functional Analysis of Utilized Lithic Items

The artifact analysis was conducted without reference to proveni­

ence data to prevent a biased interpretation of the function of arti­

facts known to be from the same proximity. During the measurement

of dimensional attributes, each artifact was examined for indications

of use. All samples which exhibited retouch or use-wear and all

examples which contained a potential functional edge were subjected

to examination under a binocular microscope. The analysis focused

on wear patterns which enabled the placement of specimens into one of 83

four categories: piercing, cutting, scraping, and chopping (defined below). The analysis is primarily limited to the active portion of the tool; the area in contact with the substance undergoing modifica­ tion.

Specimens lacking indications of use wear or intentional retouch were placed in an unutilized category (debitage). Each artifact exhibiting any visual indications of utilization was first examined under 7x with areas exhibiting wear or suspected of having wear subjected to study under 30x. The translucent properties of quartz artifacts restricted the latter analysis to the edge area in most instances although all facial areas of each specimen were examined under 7x.

The analytical techniques developed for the functional analysis

attempted to maximize the identification of functional attributes

during a minimum amount of time. If functional analysis is to become

an acceptable tool in conservation archeology, the time required to

conduct the analysis must be reduced while at the same time maintain­

ing quality control over the results. While ideally each artifact

should have been subjected to higher magnifications and should have

been coated to reduce reflection, such analytical procedures would

have passed the point of diminishing returns and delayed report

preparation past the deadlines set by the funding agency.

The terminology of functional attributes employed in this report

is derived directly from Ahler (1971: 38-39) with interpretations

based on Ahler's (1971), Semenov's (1973), and Tringham and others

(1974) experiments. The four lithic functional categories are

defined primarily on the basis of use-wear, with manufacturing and 84

stylistic attribute following in order of decreasing importance.

Although many tools are multi-functional, the primary function inferred for the artifact was used for purposes of tabulation and comparison (Appendix A).

The functional tools formally recognized represent the minimum number of utilized artifacts because such processes as the accumula­ tion of animal fats or short-term use prevents the formation of wear visible at this level of analysis (as documented by replication experimentation by Brose [1975]). Accumulation of fat on coarse­ grained lithics may be at a faster rate than on smooth-grained

lithics. Therefore, implements made of quartz or quartzite may

suffer greater loss of function due to fat adherence and may be used

for a shorter time than rhyolite or Soldiers Delight chalcedony. As

the use life of a quartz implement may be shorter due to fat adher­

ence, the wear patterns may be less extensive. Given the hardness

of quartz and quartzite, the wear patterns would be less likely to

develop than on rhyolite. These factors combined with the trans­

lucent quality of quartz which frustrates microscopic analysis of

untreated specimens suggest that many of the quartz artifacts not

assigned to a functional category may have been utilized.

Part of the explanation for the total assemblage ratio of 87

percent debitage to 13 percent utilized is that many of the functional

artifacts are not Identified at the given level of analysis. This

inference is supported by Brose's (1975: 88) research which when

utilizing similar analytical techniques on replication specimens as

used in this study on actual artifacts revealed that only 25 to 37

percent of the utilized artifacts exhibited wear. Other factors as 85

discussed above also contribute to the predominance of debitage in the assemblages from the Gwynns Falls Valley.

The following definitions have been used consistently throughout

the analysis.

PIERCING; This category is inclusive of the "projectile points" from the sites. Distal impact fractures, edge and facial smoothing of the distal tip, and the absence of facial wear along the remaining blade element are the primary attributes. Piercing implements are bifacial with edge angles ranging between 18-70° (mean = 48°). While

the blade-haft morphology may suggest a piercing function, inclusion

of the specimen in this category is based primarily on use-wear

attributes.

CUTTING: On bifacial implements, edge rounding and smoothing

usually extend along the entire blade edge. Surface wear may extend

along the entire blade element, depending on the degree of use and

penetration. Edge faceting and edge crushing may occur on specimens

used on hard surfaces. Flake cutting implements may exhibit similar

wear although to a lesser extent. Edge angles range between 25-70°

with a mean of 44°. Preverse and end shock fractures of bifaces as

defined by Crabtree (1972) are considered a diagnostic indicator of

breakage resulting from perpendicular force to the blade element; a

force resulting from most cutting activities.

SCRAPING: Use-wear may occur on flakes on unifacial and rarely

on bifacial Implements. Specimens exhibiting a plain surface

bisected by an angular surface frequently revealed wear indicative

of scraping. Edge rounding or smoothing may occur on the active

edge and extend onto the face of the angular surface. Edge faceting. 86

edge crushing, and step flaking indicate scraping of an unyielding surface. Edge crushing and step flaking may also reflect manufactur­ ing activities. Edge angles range between 40-125° with a mean of 77°.

CHOPPING: The active element contains bifacial edge crushing, step flaking and edge faceting. Edge rounding and smoothing may occur on the active element and extend to the face opposite the direction of force. Edge profiles are irregular due to the crushing but the plane view is almost always convex. Edge angles vary greatly on individual specimens due to use-wear but ranges between 25-75° with a mean of 55°. Hafted specimens have tapered elements.

DRILLING: Medium to heavy edge rounding occurs along the edges

of the active element or tip. Facial smoothing may also occur near

the tip. Transverse fractures also are diagnostic attributes. The

sample was too small to determine a range of edge angles.

Lithic Attribute Data

Definition of the attributes developed for analysis of debitage

and utilized artifacts is provided in Appendix A. Appendix A also

contains tabulations of the attributes for all of the lithic speci­

mens collected from site 18 BA 106, 112, 121 and all random finds

collected during the 1973 and 1974 field seasons. Functional descrip­

tions of the specimens collected during the 1973 and 1974 field

sessions are also provided. The plates of the diagnostic artifacts

collected during the 1973-75 field sessions are provided at the end

of Chapter 5.

This data can be utilized by the interested researcher in test­

ing some of the hypotheses advanced in Chapters 5 and 6 of this

report. Too frequently, archeologists concerned with establishing 87

research designs and understanding cultural processes do not provide data which can be used to retest the Inferences advanced (Taylor 1973:

31). The data supplied in Appendix A provide a representative sample of the material from the Gwynns Falls (18 BA 112) and the

Painters Mill (18 BA 106) sites. Similar tabulations and analysis were conducted for the material collected during the 1975 field season and is permanently stored (along with the field notes and artifacts) at the Division of Archeology, Maryland Geological Survey, Baltimore.

Ceramic Analysis

With the exception of a category for decorative motifs and a few minor adjustments for regional differences in attributes,

Bennett's (1974) laboratory manual was adopted as the basis for the ceramic analysis. A total of 41 attributes were established but only the most diagnostic attributes are described during the diachronic discussion presented in Chapter 5. The type descriptions utilized in this report have been previously defined elsewhere (Blaker 1963;

Stephenson and Ferguson 1963; Wright 1973; McNett and Gardner 1975).

Redefinition of the previously established types is not attempted due to the limited size and fragmentary nature of the sample collec­ tion from 18 BA 106.

One pottery fragment from 18 BA 112, one pipe bowl fragment and 149 pottery fragments from 18 BA 106 comprise the entire ceramic sample collected or observed from the Gwynns Falls Valley. Each of the 151 ceramic specimens was subjected to analysis. A binocular microscope was employed to identify temper orientation and type. A

Munsell soil color chart served as a reference for establishing general color categories similar to those established by Bennett 88

(1974: 103). A corner of each sherd was snapped to provide a fresh cross-section essential for determination of core position and color as well as temper density and orientation. The width of the cord impression or incised line, the number of parallel incisions or impressions per 10 centimeters, and the decorative motifs were also recorded.

The attribute list and code will not be provided but is avail­ able at the Division of Archeology, Maryland Geological Survey,

Baltimore. Instead, the analysis will focus on the isolation of vessels and discuss the spatial distribution of vessels from discern­ able phases. The fragmentary nature of the sherds limits the degree of reliability assignable to the vessel count. The following criteria,

in order of relative value, provided the framework for the division

of the sample into vessels and vessels into types; type of temper,

exterior and interior surface treatment, spatial distribution,

exterior and interior color, core position, core definition and core

color, vessel wall thickness, paste characteristic, and temper size

orientation.

The distribution of sherds inferred to be from the same vessels

documents the downslope displacement of artifacts and enables the

delineation of activity areas for the various components of the

Townsend and Potomac Creek complexes at the Painters Mill site

(18 BA 106). Examination of local collections over a period of five

years provided the basis for many of the inferences advanced on the

basis of ceramic analysis (Chapters 5 and 6). CHAPTER 5: DIACHRONIC SYNTHESIS

An adequately developed chronology is essential to the inter­ pretation of assemblages recovered from the surface of most prehis­ toric sites. A surface collection from a site, such as 18 BA 106, frequently represents the accumulated residues of a variety of cultures, conducting a multitude of activities, in different areas, over the passage of many millennia. Interpretations of cultural processes, of activity areas, of settlement-subsistence systems without first establishing the chronology of a region may result in a simplified and often erroneous elucidation of the complexities operative in the formation of the archeological record.

Even after the establishment of a regional chronological sequence, the majority of artifacts from the surface of a site may be unassign­ able to a given period; let alone a given tradition or phase. While lithic preferences may have changed over time and thus afford arche­ ologists a chronological control, the technological processes employed in the manufacture of lithic artifacts often span the entire prehistoric period: producing debitage which may be assigned to contemporaneous activities but which in actuality spans 5000 years of cultural activity. As greater attention is given to debitage

(which comprises the majority of artifacts from most assemblages) the controlled surface collections which may be difficult to inter­ pret today may become more meaningful in the future. As more arche­ ologists conduct and report on controlled surface collections, the

current problem of an absence of comparative data will decrease.

But until that time, establishment of a chronology depends on a

comparison of the available data to previous, established chronologies.

As projectile points are the primary functional class for which

89 90

chronological controls exist, this analysis will emphasize projectile point comparisons.

A chronology of cultural development in the Eastern Piedmont province has not been formulated. During the late '60s and early

'70s, archeologists from Catholic and American Universities focused on cultural development and adaptation in the Coastal Plain and

Piedmont province sections of the Potomac Valley. While excavations at stratified sites such as Swan Point, Ruppert Island, and Monocacy have clarified a segment of the confused chronology of the Potomac

Valley, chronological control over much of the Archaic period

assemblages still depends on regional comparisons to the firmly

established chronological sequences of Coe (1964), Broyles (1971),

and Kinsey (1972).

Three dissertations resulting from this research deal respec­

tively with the Archaic period, the Transitional period and the

Woodland period (McDowell 1972; Ayers 1972; Gluckman 1973). The

results of the Potomac River survey have been synthesized by

Dr. Charles W. McNett, Jr. in an unpublished book which clarifies

substantially the chronology of the Woodland period (McNett and

Gardner 1975). But the data accumulated over the years is insuffi­

cient for the development of a refined chronology for most of the

Middle Archaic period (ca 6000-4000 B.C.).

The diachronic synthesis presented in this chapter interprets

the various components represented at the known sites within the

Gwynns Falls Valley. These components will be related to other sites

within the Patapsco Valley and to sites in the surrounding region.

As the components in the Gwynns Falls Valley may represent sub­

systems of settlement-subsistance systems spanning the Western 91

Piedmont and Coastal Plain provinces, regional comparisons become essential to the understanding of the archeological record. While each period could be the subject of a dissertation, only the minimum of data required to interpret the archeological record from the

Gwynns Falls Valley is provided. For more detailed analysis, the reader will be referred to the relevant works.

The Betty Pearre collection from the upland area between

Reisterstown Road and the Gwynns Falls, the Bruce Chilcoat and

Charles Hazard collections from the divide between the Patapsco and

Worthington Valleys, the Northwest Expressway Corridor survey col­ lection from the Gwynns Falls and Patapsco Valleys and the relevant literature form the data base from which the following interpreta­ tions are formulated.

PALEO-INDIAN PERIOD (13,000-8000 B.C.)

When man first entered the Eastern Piedmont, the environment was probably quite different from the present. The Paleo-Indian hunter may have occupied a "tundra and broken spruce environment contemporaneously with caribou, mastodon, and moose-elk" (Fitzhugh

1972: 4). Toward the end of this period, the grassland and open spruce forest which supported these large herbivores were changing to a hemlock-northern hardwood type (Carbone 1974: 89). A diversity of forest units, corresponding in part to the moisture availability of the underlying geologic and soil units and to topographic features affecting exposal to wind, probably developed (see Haynes and Buell

1955 for a discussion of the physiological requirements of trees).

As the climate became progressively warmer, the resulting increased

diversity in forest units provided Early Man with a wider selection 92

of habitats and plant species while reducing the megafauna of the region.

The people of the Paleo-Indian period apparently subsisted primarily by hunting a variety of species over large territories.

The extent of the hunting territories may have been determined by

the distance to the nearest source(s) of cryptocrystalline deposits,

the rate of exhaustion of the tool kit, and the success of the hunt.

Collection of floral and of riverine resources may have supplemented

the basic meat diet; as suggested by the recovery of wild hawthorn

pits and fish bones from a hearth area dated to 8640 B.C. t 300

radiocarbon years at the Shawnee Minisink site in Pennsylvania

(McNett, Marshall, and McDowell 1975).

The decided cultural preference for cryptocrystalline materials

to the exclusion of most other lithic resources is one of the diag­

nostic traits of this period. While the Paleo-Indian groups may have

traveled over large areas, visiting a succession of quarries, the

evidence from the Thunderbird site in Virginia and the Shawnee

Minisink site (Figure 14) suggests periodic returns to a quarry base

camp upon exhaustion of the tool kit (Gardner 1975).

Locally available sources of cryptocrystalline materials, such

as Soldiers Delight chalcedony, were probably used to replace some

exhausted items of the tool kit. Soldiers Delight chalcedony may

have served as a source for replacing flake and core tools. Although

the proximal end of a biface cutting implement (Figure 36: I) was

found at 18 BA 129, the natural size of the material and the impuri­

ties in Soldiers Delight chalcedony may have frustrated the produc­

tion of large bifaces. Journeys to the serpentine barrens by early

man may have been primarily to exploit the mammals which fed on the 93

probable grasslands of the area; with the lithic resources a secon­ dary attraction. Artifacts of Soldiers Delight chalcedony have not been found outside the Gwynns Falls Valley, suggesting the secondary importance of this lithic source to early man.

The presence of artifacts of Point of Rocks jasper and Front

Royal jasper at several sites in the Corridor suggests that quarry base camps may have been located near outcrops of this material.

Investigations of the Point of Rocks jasper outcrops did not produce functional artifacts from this period although debitage was collected

(Geasey 1975: per. comm.). The absence of extensive cryptocrystal­ line resources in the Eastern Piedmont province combined with the evidence that poor quality Soldiers Delight chalcedony was utilized to replace exhausted items in the tool kit suggest that exploitation of the Eastern Piedmont province during this period was limited to hunting expeditions.

Sites from this period are predicted to occur in the floodplain of the Gwynns Falls on colluvial fans extending from the valley slopes and are probably buried (Gardner 1974; Coe 1964). The crypto­ crystalline material recovered from the sites in the Red Run and

Gwynn Falls Valleys could be assigned to either the Paleo-Indian or

Early Archaic periods. A gray chert side scraper and a Point of

Rocks jasper core end scraper from the Hazard's collection probably date to this period and suggest utilization of the upland areas as well as the floodplain (Figure 39: B, C). One Pennsylvania jasper biface in the Chilcoat collection exhibits excellent technological

control over the material; a characteristic of the Paleo-Indian

period. The specimen (ten centimeters in length) has a snapped base 94

DISTRIBUTION OF CHRONOLOGICALLY IMPORTANT SITES I. Shawnee - Minesink 12 .Thunderbird Z.Townsend Site l3.Hand 3.Conowingo 4. Pointers Mill 5.UMBC S.EIkridge 7. Stoney Run (18 An 27) 8.Bigg Ford 9. Monococy lO.Selden Island I J.Accokeek Pa. , "foel.VNJ

Deiowore Boy L

Atlantic

Ocean

Figure 14: Distribution of chronologically important sites. 95

and lacks fluting but may date to the Paleo-Indian period. The point was found on the Forester farm in the Green Spring Valley

(Chilcoat 1973: per. comm.).

EARLY ARCHAIC PERIOD (8000-6000 B.C.)

By this period, the megafauna of the Wisconsin glaciation had given way to the species found in the beginning of the historic period. The coniferous forest elements in the region were probably being replaced by deciduous forest units (pages 32-33). The marine transgression may have progressed to the area of the present upper

Bay Bridge crossing, resulting in a slightly smoothing effect on the local climates in the Coastal Plain province.

During the earlier phases of this period, the technology and the subsistence-settlement patterns developed during the preceeding period were continued. But with the decrease in megafauna and the increase in the diversity of floral resources, the subsistence activity may have increasingly shifted to the scheduling of floral resources. Man's familiarity with mast producing forest units may have begun when these units occupied only limited areas in the spruce-beech-pine dominated region. Knowledge of the utilization and preparation of deciduous forest products may have spread from the southeast along with the concept of corner notched projectile points.

Corner Notched Tradition: The major change, which defines the early phases of this period, is the introduction of a corner notched and serrated tradition of projectile point manufacture. The prefer­ ence for cryptocrystalline materials as well as the remaining

functional items of the tool kit continued in popularity. The 96

increasing utilization of locally available, non-cryptocrystalline minerals during this period probably reflects increased localized adaptation in response to the shifting climatic and resultant faunal and floral changes as discussed above. By the end of this tradition, corner notched projectile points are found in a variety of geographic settings and in quantities suggestive of increased population size.

The duration of this tradition is discussed in the Middle Archaic

section of this chapter.

Except for a possible Palmer Corner Notched-like rhyolite pro­

jectile point from the Pearre collection (Figure 37: B) the first

datable occupation of the Corridor is related to the Kirk Corner

Notched-type projectile points as defined by Coe (1964: 69) and

Broyles (1971: 65).

Gwynns Falls Component (18 BA 112): A projectile point of gray

chert resembling the basally ground, Kirk Corner Notched, Large

Variety, projectile type (Broyles 1971: 65) was found in square

S370W50 during the 1975 surface survey (Figure 34: A). This projec­

tile point style dates to around 6900 B.C. at the St. Albans site in

West Virginia (Broyles 1971: 65). Probably utilized during this

period, the cryptocrystalline artifacts cluster within a 30-meter

area around S310W50 (Figure 14). The screening of one square meter

of plowzone from S302W50 yielded four of the 20 cryptocrystalline

artifacts recovered from the site (Figure 15).

A minimum of eight and a maximum of ten different varieties of

cryptocrystalline minerals are represented. Point of Rocks jasper,

gray chert, and an undefined gray and white speckled flint were

recovered. One projectile point (described above), one unretouched 97 but utilized flake scraper, and a series of flakes suggesting at

least two instances of biface retouch, two instances of edge retouch

and two instances of primary flake production are represented. While

these cryptocrystalline artifacts may have been produced and utilized

during the Paleo-Indian period, they are assigned to the comer

notched tradition based on the presence of the Kirk-like projectile

point and the clustering of the assemblage. During this period the

site served as a transient camp where maintenance activities occurred.

Painters Mill Component (18 BA 106): Sixteen cryptocrystalline

artifacts were recovered (Figure 19). Soldiers Delight chalcedony,

brown jasper, brown chert. Point of Rocks jasper, and Flint Run

jasper artifacts are represented. Two brown chert artifacts found

40 meters apart, an end scraper (348) and a flake (449), were struck

in succession from the same core (Figure 31: B, C). A core, a flake

end scraper, and two retouch flakes of brown chert were also recovered

(Figure 31: A, I). A Kirk Corner Notched-like projectile point of

brown chert from the Pearre collection (Figure 37: A) and a similar

style point of the same material from a site in the Patapsco Valley

suggest that brown chert is a local diagnostic trait of the Corner

Notched tradition. Except for two cores and two retouch flakes of

white Soldiers Delight chalcedony (Figure 31: H) which cluster around

S50W70, the cryptocrystalline artifacts cluster loosely in the south-

central section of the site (Figure 19).

One flake of red Soldiers Delight chalcedony from this area is

similar to a core of the same material found at a site along Red Run

(18 BA 127). Given the heterogeneity of this material, the flake at

18 BA 106 may have been struck from the core found at 18 BA 127, 98 although spectographic analysis of both specimens produced differ­ ences in minor elements (see pages 64-65). Based on the above data, the Painters Mill site was utilized during this period as a possible processing station where scraping-related activities occurred and where tools were manufactured and retouched.

Red Run Valley components: Four sites located on the first terrace of the Red Run Valley have produced cryptocrystalline arti­ facts. Site 18 BA 122 yielded a gray chert knife/scraping implement.

A red Soldiers Delight chalcedony core found at 18 BA 127 and a flake of similar material were recovered from the same hundred-foot square.

Two gray chert edge retouch flakes from 18 BA 125A and one Soldiers

Delight chalcedony flake side scraper from 18 BA 125B suggest limited occupation during the time of the Corner Notched tradition

(Figure 35: E).

A Kirk Corner Notched-like point manufactured from black rhyo­ lite was found north of 18 BA 127, in the formerly swampy portion of an extinct stream headwater area. These upland spring-head swamps may have served the fauna of the area as wallowing spots as well as

feeding grounds and would have been an attraction to the people hunt­

ing these animals.

Horsehead Branch Valley component (Lee site; 18 BA 129):

Located on an eroded remnant of a former terrace, site 18 BA 129

produced three artifacts manufactured from Soldiers Delight chalced­

ony. A primary flake, a block core, and a proximal biface fragment

were recovered from an area 30 by 30 feet square (Figure 35: I-K;

Figure 12: 8, 11, 20). A fire-cracked rock of this material was

recovered from W70S100 (Figure 12). 99

The data from the Gwynns Falls Valley suggest that people utilizing cryptocrystalline materials were exploiting the resources of the serpentine barrens of Soldiers Delight, of the upland stream heads, and the floodplain habitats. Transient camps were located along the first terrace of the Red Run Valley and on colluvial fans in the Gwynns Falls Valley. A projectile point of black rhyolite suggests that this lithic material may have been utilized during the

Corner-Notched tradition; a hypothesis supported by the occurrence of rhyolite points of this tradition found in the surrounding region.

External Relations: Within the Patapsco Valley, Kirk Corner

Notched-like points manufactured from brown jasper, rhyolite, quartzite and quartz have been reported from several sites (Clark

1975b; Stearns 1949). This projectile point style extends from the southeast to as far north as Staten Island and is a pan-Eastern development. While the increased quantity of projectile points from this period may reflect greater exposure of sites from this period, another interpretation is that the Indians were already exploiting the varying resources of the different forest units which were devel­ oping. An increased utilization of locally available resources is also reflected by the growing acceptance of non-cryptocrystalline, locally obtainable lithics. Towards the close of the Corner Notched tradition, the cultures in the Eastern Piedmont and Coastal Plain cultures of the middle and lower Chesapeake Bay area still retained the flint-knapping technological abilities of their ancestors and successfully applied those abilities to non-cryptocrystalline materials. The technology of flint-knapping somewhat diminished in quality throughout the subsequent Bifurcate tradition. 100

Bifurcate Tradition:

"The bifurcate tradition (Broyles 1966) appears to originate from a Kirk horizon in the west Appalachian zone where it is, initially, a distinct regional development, spreading later into southern New England and the central Atlantic regions." (Fitzhugh 1972: 8)

The bifurcated-base projectile points from this tradition are

the first abundantly represented Early Archaic point types which occur in local collections from Maryland (Cresthull 1969). The preference for cryptocrystalline material disappeared in the Patapsco

Valley during this period although the practice of serration contin­ ued. Locally available resources such as quartz in the Coastal Plain or rhyolite in the Piedmont became the primary source of lithic materials in areas deficient of cryptocrystalline sources. In areas where cryptocrystalline minerals are locally abundant, this material

continued to be utilized by the local cultures (Cresthull 1969;

Broyles 1971; Gardner 19/1). Minor stylistic variation between the

contemporaneous phases of the tradition suggests that the cultures

continued to become increasingly localized. But contact with other

groups was still frequent enough to result in a pan-Eastern similar­

ity of general point styles. The bifurcate tradition dates from

around 6800 to 6200 B.C. in West Virginia (Broyles 1971) but may have

continued into the Middle Archaic period further north.

Gwynns Falls Component: One plain rhyolite projectile point

resembling the MacCorkle Stemmed type in base morphology was recovered

from the wash area of the site (Figure 34: B). The specimen may have

eroded from the upper edge of the terrace (Figure 15: 67). The

MacCorkle Stemmed point type dates to around 6800 B.C. in West

Virginia (Broyles 1971: 71). Association with other artifacts from 101 the site could not be determined although some of the rhyolite debitage may date to this period.

Painters Mill Component: One bifurcate base, serrated, cobble quartz projectile point was found at the head of the wash, west of

18 BA 106 (Figure 31: D; Figure 9). As characteristic of this point style, the blade element of this specimen was retouched. Due to its isolated location, the point could not be associated with other artifacts from the site although some of the quartz artifacts found on the surface of the site may date to this period.

Other Components: At the Lee site (18 BA 129), one plain rhyo­ lite, bifurcate base projectile point was found which resembles the

LeCroy Bifurcated Base point type (Broyles 1971: 64). The point was probably abandoned after receiving a distal impact fracture (Figure

36: A). Three quartz projectiles are represented in the Pearre collection. Two (Figure 37: D) resemble the LeCroy Bifurcated Base type which occurred in a level radio-carbon dated at 6300 B.C. t 100 years at the St. Albans site (Broyles 1971: 69). One resembles the

Kanawha Stemmed type radio-carbon dated at the St. Albans site at around 6210 B.C. (Figure 37: E). The three specimens exhibit trans­ verse fractures of the blade element, suggesting either breakage while in use for cutting or post-depositional damage.

External Relations: Two LeCroy-like and two Kanawha Stemmed- like projectile points were recovered from the lowest level of the

stratified UMBC site (Figure 14), which is located at the head of a

tributary of the Patapsco River (Clark and Miller 1975: 35). Quartz

debitage and three rhyolite flakes were also recovered from this

level (Vitelli and Zaruba 1975: 62). Bifurcated base points are 102

reported from several sites in the Coastal Plain portion of the

Patapsco River Valley (Clark 1975b; Stearns 1949). In the upper

Chesapeake Bay area, a variety of lithic types were utilized through­ out this tradition (Cresthull 1969). Transhumance probably developed during this period as suggested by the presence of bifurcate points in a variety of habitats. Bifurcate-type points do not really become abundant until the development of the LeCroy Bifurcated Base style.

MIDDLE ARCHAIC PERIOD (6000-4000 B.C.).

With the culmination of the gradual warming trend from approxi­ mately 5000 to 3000 B.C., a climate generally much warmer and drier than at present characterized the region (Altithermal). The warmer and drier conditions would have favored the expansion of plant species with low moisture requirements. The chestnut oak and black­ jack oak forest units, while perhaps of different composition, probably expanded at the expense of units similar to the tulip poplar and sycamore-box elder-green ash-silver maple units. As the black­ jack oak and chestnut oak units are mast producing forest, the expansion of these forest units may explain the occurrence of Archaic period remains in areas which today are of comparatively low carrying capacity. With the onset of cooler and moisture conditions, the

tulip poplar and other moisture requiring forest types may have

started to develop towards their present composition and extent;

reducing the size of the mast producing units and thus requiring

cultures to shift to greater dependence on other subsistence activ­

ities as well as shifting settlements to the remaining mast producing

forest for the seasonally scheduled gathering activities. This is

one explanation for the occurrence of Archaic material in a diversity 103

of present habitats and the subsequent abandonment of this dispersed form of settlement pattern.

The expansion of deciduous forest units was not limited to the

Eastern Piedmont but also extended into the Northeast. The progres­ sively northerly establishment and expansion of deciduous forest may have been a factor in the spread of traditions of this period from the south to the north (Ritchie and Funk 1971). The establishment of shellfish in the Chesapeake Bay and other coastal environments in the Middle Atlantic region may have been another ecological factor which influenced the spread of cultural traits from the southeast.

More important than the concept of particular projectile point styles, the concepts and techniques of survival in a changing environment would have been of greater value. Such techniques would be of little value before the resources appeared.

With the establishment of shellfish in the Chesapeake Bay, the local people probably did not have to invent an entirely new tech­ nology to exploit the resources; such a technology may have already existed farther to the south. The spread of the concepts and tech­ niques relating to shellfish exploitation and of adaptation to deciduous forest may explain the south to north time slope character­

istic of this period. Projectile point styles were but one of the many concepts exchanged. But projectile points are one of the few

artifacts remaining in the Gwynns Falls Valley which are useful in

documenting culture change over time. While the entire assemblage

for any tradition or complex should be studied, definition of assem­

blages on the basis of surface collection requires an extremely

large sample; a requirement not fulfilled by the material available

for study from the Gwynns Falls Valley. 104

The chronology of the Middle Archaic period in Maryland is poorly understood. The chronology in the Gwynns Falls Valley for this period is virtually unknown.

Generally, the influences during this period flow in a south to north direction and may reflect a change in the resource base as discussed above. The south to north time slope hypothesis is sup­ ported by radiocarbon dates and suggest that cultures in the Maryland area probably fall between the northern and southern dates.

In the Carolina Piedmont, the beginning of the Middle Archaic period may represent a continuation of the Kirk Stemmed points and associated assemblages (Coe 1964: 35). The possible development from

the Kirk Stemmed to the Stanly Stemmed type points may also reflect

influences from the Bifurcate tradition. In the Carolina Piedmont,

the first evidence of polished stone implements (atlatl weights) is

associated with Stanly Stemmed points (Coe 1964: 122). Evidence of

ground stone tools technology has been reported as early as 7700 B.C.

in the lowest levels of Graham Cave. At the Sheep Rock Shelter, a

ground celt fragment was found in a zone underlying a corner notched

point producing level dated to 6900 B.C. (Ritchie and Funk 1971).

Morrow Mountain type points and the associated assemblages are

believed by Coe (1964) to date not much before 4500 B.C. A Morrow

Mountain point from Zone D at the Stucks Bluff Rock Shelter in

Alabama was associated with a hearth radiocarbon dated at 4500 t 120

B.C. (DeJarnette et al. 1975: 116). This date compares favorably

with three dates from the Morrow Mountain level at Russell Cave

(4360 B.C. t 140; 4300 B.C. t 190, and 4030 B.C. ± 200) (Griffin

1974). Morrow Mountain points have been found in the lower zones 105

of the Tennessee Valley shell middens and have been found at the

Stanfield Worley Bluff Shelter in association with flexed burials

(DeJarnette et al. 1975: 116).

In Virginia, Morrow Mountain points were observed in oystermen collections photographed by the staff at the Virginia Research Center for Archeology. Reportedly dredged-up from oyster deposits in the lower Chesapeake Bay, the context of the specimens suggest utiliza­ tion of shellfish resources toward the close of the Middle Archaic period. Shellfish resources at the mouth of the Bay were present as early as 6000 B.C. and may have been utilized, although the sites of this period would have been long submerged. Widespread exploitation of oyster or shellfish resources by 4000 B.C. is documented by the above data and by the Tennessee data, by the reported radiocarbon dates (ca 4000 B.C.) of basal shell deposits in Kent County, Mary­ land (Wilke 1974: per. comm.), and by shell deposits in the Hudson

Valley (Brennan 1974). Brennan (1974: 89) reports the utilization of shellfish by 5000 B.C. in the Hudson Valley and argues for an

even earlier antiquity; the earlier sites being lost during the marine transgression.

North of Maryland at the Sheep Rock Shelter, a corner notched

point was found in a level dated to 5100 B.C. t 250 years (Michels

and Smith 1967: 863). This date for a c o mer notched point contrasts

with Bettye Broyles' chronological placement of the Comer Notched

tradition but may be the result of the confused stratigraphy at

Sheep Rock. While the Sheep Rock date for the C o mer Notched tradi­

tion is questionable, Herbert Krafts' excavations at Harry's Farm

site in New Jersey produced a Kirk-like point from a "hearth 106

containing charcoal that assayed to 5430 B.C. t 120 years (1-6133), and 5370 B.C. ± 120 years (1-6600)" (Kinsey 1975).

On Staten Island, a similar series of late dates have been reported but may date the bifurcated base points found in a multi- component level (Ritchie and Funk 1971). At the Hallowell site, a number of LeCroy Bifurcated Base and Kanawha Stemmed type points were found in a level associated with ovate or oblong choppers, a possible polished atlatl weight, fragmentary celts, and other func­

tional tool types. Charcoal samples collected from two hearths yielded age determinations of 5310 B.C. - 125 years (1-5512) and

6300 B.C. ± 140 years (1-5331) (Ritchie and Funk 1971). The latest

date may apply to the Stanly-like points while the earlier date may

apply to the Bifurcate tradition points found in this level.

In Maryland, the lowest level at the UMBC site produced LeCroy

Bifurcated Base and Kanawha Stemmed points made from quartz and rhyo­

lite. In the natural level sealing the Bifurcate tradition level

were seven Kirk Stemmed points (Clark and Miller 1975: 35-36). The

occurrence of Kirk Stemmed points above points of the Bifurcate

tradition suggests that the stemmed point style developed during the

Corner Notched tradition may have persisted throughout and after the

Bifurcate tradition. This hypothesis is supported by the presence

of Kirk Stemmed points in the Bifurcate tradition levels at the

St. Albans site. The "one culture-one point style" hypothesis

advanced by Broyles (1971: 48) may not be applicable to all time

periods; a cautionary note also advanced by Fitzhugh (1972: 3). Then

again, the Kirk Stemmed "point" may be a functional knife form which,

like so many other items of the Early Archaic tool kit, remained 107

basically unchanged as projectile point styles changed over time.

The Middle Archaic point styles of the Stanly and Morrow

Mountain types occur throughout the Coastal Plain and Piedmont province of Maryland but points of these types are absent from the

Gwynns Falls collections. While the Bifurcate tradition may endure

to the fifth millennium B.C., points of this tradition are also rare

in the Gwynns Falls Valley. The only points which may date to this

period resemble the Kirk Stemmed points from the UMBC site but they

lack serration and do not have the same flaking patterns on the blade

element (Clark and Miller 1975: 35-36). Collected by Charles Hazard,

the points may be the type following the Kirk type in the Gwynns

Falls Valley (Figure 38: A-C, J). While the Piedmont tradition

point styles are in the majority in the collections from the Gwynns

Falls Valley, the various styles are assignable primarily to the Late

Archaic period.

What happened to the Middle Archaic period in the Gwynns Falls?

While we know that the Middle Archaic point styles documented by Coe

for the Carolina Piedmont decrease toward the north, points and

other tools of this tradition are represented in sites only ten miles

from the Gwynns Falls Valley (Figure 14: 18 AN 27). Sites of this

period may be obscured by ground cover or remain buried under

alluvium, although sites with material from the Early Archaic period

have been discovered. Other point styles may have developed in the

Piedmont province of Maryland which have not been formally defined

or which have not been properly assigned to the Middle Archaic

period. Perhaps the Piedmont tradition dates to an earlier period

than documented for the Piedmont tradition in the Delaware Valley 108

(Kinsey 1972). The hypothesis that this area was not utilized during the Middle Archaic period is unacceptable. Clearly the gap in the archeological record stems from inadequate data or faulty interpre­ tation of the available data; a situation which will require the gathering of additional data before the problem can be resolved.

Perhaps renewed excavation to test for stratified deposits at the

Gwynns Falls and Painters Mill sites will provide the essential stratigraphie data necessary to clarify the confused chronology.

LATE ARCHAIC PERIOD (4000-1000 B.C.)

The general warm and dry conditions of the Altithermal continued until around 3000 B.C. but was gradually changing to a general cool­ ing trend which has continued with minor variation to the present.

The minor variations may have consisted primarily of periods of cooler and drier or warmer and drier conditions (Flint 1957). The general increase in moisture and cooler temperatures would have favored the expansion of moisture-favoring forest units such as the tulip-poplar unit at the expense of the forest units like the chest­ nut oak which is composed primarily of less moisture-requiring species (Figure 3-4). Modern forest units probably developed during this period although the horizontal distribution of the units may have differed from the present.

The cooling climatic conditions probably contributed to the decrease in the relative sea level rise rate from a foot to a half- a-foot per century beginning around 3700 years ago (Kraft 1971:

2131). This decreased rate of sea level rise would have initially

favored development of stable oyster resources and would have slowed

the rate of destruction of marine oriented prehistoric sites. 109

This period witnessed the development of a number of traditions in the east and an increase in cultural complexity. The apparent increase in cultural complexity may be related to our recovery techniques and factors of preservation as well as to real diversifi­ cation and elaboration of the cultures. The study collection from the Gwynns Falls contains assemblages of or similar to the Laurentian tradition, the Piedmont tradition, the Broadspear tradition and the

Orient Fishtail tradition (Kinsey 1972). Mortuary ceremonialism is first reported for this period and utilization of shellfish resources by coastal cultures is well documented (Brennan 1974; Ritchie 1969).

Evidence of seasonal settlement patterns is extensively reported for this period (Ritchie and Funk 1973). For a concise summary of the external relationships of these various traditions the reader is referred to Fitzhugh (1972). The complexity of the interrelationships between the traditions, phases, and components of this period prevents elaboration. For a description of the material culture and settle- ment-subsistence patterns of the various traditions see Kinsey (1972).

This analysis will define the presence and extent of the traditions represented but will refrain from the establishment of new complexes or phases because of the limited size of the sample.

Piedmont Tradition; This tradition developed in the southeast

during the Middle Archaic period and spread along the Atlantic coast

into Eastern Canada and may have reached as far north as Newfoundland

and Labrador by 4000 B.C. (Fitzhugh 1972: 9). Regional expressions

of the tradition are reported to the north (Kinsey 1972) and the

south (Coe 1964) of the Eastern Piedmont province and have been dis­

cussed above. During the Late Archaic period in the Carolina 110

Piedmont, the Guilford projectile point styles and associated assemblages were followed by the Savannah River complex which per­ sisted into the Early Woodland period north of Maryland. Intrusions of other cultures into the area formerly occupied by Piedmont tradi­ tion cultures resulted in a complexity of cultural development.

While influences from the Laurentian tradition are documented in

Maryland, the Piedmont tradition in Maryland continues throughout the Late Archaic.

McNett and Gardner (1975) have defined the Vernon, Holmes and

Piscataway complexes which date to the Late Archaic period. The

Vernon complex is characterized by projectile points of the Vernon type and the Piscataway complex by projectile points of the Piscata­ way type as defined by Stephenson and Ferguson (1963). Both the

Vernon and Holmes complexes may predate the Piscataway complex which has been assigned a date of approximately 2000 B.C. The Piscataway complex may extend back in time considerably earlier than 2000 B.C.

The Broadspear and Orient Fishtail traditions followed these complexes of the Piedmont tradition.

Painters Mill component: A rhyolite, square stemmed knife resembling the Savannah River projectile point type (Coe 1964: 44-45) was found in the northern wash area of the site (Figure 31: 0). The specimen exhibits heavy edge faceting and smoothing along one blade edge. Specimen No. 316 resembles both the Savannah River Stemmed and the Lackawaxen Expanded Stem projectile point types, and may

represent a regional manifestation of the Piedmont tradition (Coe

1964: 44; Kinsey 1972: 408-409). A projectile point, four edge

retouch flakes, and a snapped edge retouch flake are concentrated Ill

within a 30 meter area around S50W50 (Figure 20). A snapped-edge retouch flake (143) which exhibits edge faceting and a primary flake of this material was recovered from S130W50. Rhyolite artifacts are distributed fairly equally across the site with only the northeast portion of the site containing a low density of items (Figure 20-21).

Feldspar rhyolite loosely clusters in the north-central portion of

the site (Figure 20). As the majority of the projectile points from

the Piedmont tradition in the Gwynns Falls are manufactured from

rhyolite, part of the rhyolite debitage found on the surface of the

site probably dates to the occupation of the site during this tradi­

tion. The bannerstone manufactured from serpentine may also be

assignable to this tradition (Figure 32: V).

Three quartzite projectile points (Figure 31: K-M) are similar

in style to the Vernon type and may be assigned to the Vernon complex

(Stephenson and Ferguson 1963: Plate 24; McNett and Gardner 1975).

Two Piscataway points diagnostic of the Piedmont tradition were

recovered (Figure 31: DD-EE). McNett and Gardner (1975) define three

varieties of points for this complex; the points from the Painters

Mill site are assigned to the rounded stem and pointed stem varieties.

A third variety is similar to the pointed stem variety but made from

rhyolite instead of quartz. The rounded stem variety resembles

heavily reworked Guildford points. Occupation of the Painters Mill

site appears to have been limited during this tradition although

the sparsity of occupation is inferred primarily from the sparsity

of projectile points which constitute only five percent of the total

assemblage.

Gwynns Falls component: Two rhyolite projectile points diag­

nostic of the Piedmont tradition were found in the wash area west of 112

the 1975 area of cultivation (Figure 34: D, E). The square base specimen is similar to the Savannah River knife type (Coe 1964;

Figure 41) while the other specimen resembles the Lackawaxen Expand­ ing Stem type (Kinsey 1972: 408-409). The Lackawaxen-like point may predate Kinsey's type and may be of a yet-undefined complex of the

Middle Archaic period. The one quartzite Clagett type point from

the site (Figure 34: J) may be from the regional complex related to the Lackawaxen complex. If we compare Stephenson's and Ferguson's

(1963) Clagett and Bare Island types to Kinsey's (1972) Lackawaxen

Expanding Stem, Straight Stem and Contracting Stem types, the major differences are in the lithic preference.

One gray chert rounded base Piscataway point (Figure 34: I) and

a straight-stern rhyolite biface were recovered from the surface of

the site. The rhyolite biface is severely retouched and exhibits

edge faceting and rounding. Fifteen flakes of the same variety of

rhyolite cluster around S330W50 (Figure 15). Biface production and

maintenance, cutting and possibly scraping activities are represented.

One specimen recovered during the 1974 survey is also diagnostic of

this tradition (Figure 34: F), while the specimen shown in Figure 34:

G may represent either a severely retouched knife of this tradition

or of the latter Middle Woodland tradition (Table 12). Part of the

rhyolite debitage at the site was probably produced by peoples of the

Piedmont tradition.

The data suggest that the site served as an interior exploitive

camp and that occupation areas may be located towards the head of

the colluvial deposits; a hypothesis which will be tested during the

final salvage phase of investigations. 113

Red Run component (18 BA 127): While quartz bifaces predominate at this site (86 percent), a number of stemmed bifaces were collected.

Most of the bifaces were collected after the plates were prepared and are not illustrated. One feldspar rhyolite projectile point similar to the Lackawaxen Contracting Stem type was retouched at least once and may have been abandoned after receiving an end shock fracture.

A proximal fragment of a contracting stemmed point of rhyolite and a severely-retouched straight stemmed rhyolite knife (Figure 35: G) was also recovered. As rhyolite comprises only 13 percent of the total lithic material from the site (Table 1), occupation during this tradi­ tion is assumed to be limited. This interpretation is based on the assumption that only a portion of the rhyolite collected is assign­ able to the Piedmont tradition and that quartz material was not utilized extensively by the people producing the rhyolite bifaces.

Both assumptions are fairly weak and should be tested. If a con­

trolled surface collection of the site is conducted as recommended

for the salvage phase of investigations then the increased sample

size may enable testing of the interrelationships of the cultural

items.

Other sites in the Red Run Valley which produced stemmed arti­

facts are sites 18 BA 125A (Figure 35: A); 18 BA 125B (Figure 35: C)

and the area around a stream head north of 18 BA 127 (Figure 35: N).

One rhyolite pointed stem Piscataway type point was found at

18 BA 129 (Figure 36: E). While these stemmed styles are assigned

to the Piedmont tradition, the knives may have been produced during

other traditions or periods; the stemmed style being a functionally

practical knife form utilized in one type or another from at least

the Early Archaic to the Middle Woodland periods. 114

Pearre collection: The utilization of upland resources by cultures of the Piedmont tradition, as first suggested by the Red

Run data, is supported by the presence of Piedmont tradition point styles in the Pearre collection. Of the nine vein and cobble quartz bifaces utilized for cutting, piercing and scraping, only one has not suffered a fracture to either the distal portion of the blade or snapped haft element (Figure 37: R). Most of the bifaces appeared to be multi-purpose tools and were probably abandoned after breakage instead of loss during the hunt. The predominance of quartz arti­

facts (59 percent) in the assemblage, particularly of the cobble quartz variety, suggest that people hunting in this part of the valley were based primarily in the Coastal Plain province where cobble

quartz artifacts of the Piedmont tradition are abundantly represented

(Stearns 1943: 1949). But utilization of the area by Piedmont-based

groups is suggested by the presence of one Lackawaxen Expanding Stem­

like point (Figure 37: B), one Vernon-type point (Figure 37: N), and

three Lackawaxen Contracting Stem-like points (Figure 37: W, X). Two

generalized side notched points which only vaguely resemble the

Halifax types as defined by Coe (1964) are represented (Figure 37:

N, T). Halifax points identical to Coe's type have been found with

an associated assemblage at 18 AN 27 (Figure 14) and are reported

from other sites in the Patapsco Valley (Stearns 1949). This type

is dated to around 3500 B.C. in the Carolina Piedmont sequence and

may represent northern influences into the Carolina Piedmont (Coe

1964: 123).

Hazard collection (Hazard site): Piedmont tradition point and

knife forms predominate in the Hazard collection. Early projectile 115

point types of this tradition as discussed above may be represented

by the stemmed points in Figure 38: A-C, J. Points similar to the

Lackawaxen type predominate. Eleven Lackawaxen Expanding Stem

(Figure 38: F, M, R), three Lackawaxen Straight Stem (Figure 38: L)

and four Lackawaxen Contracting Stem type points (Figure 38: D) were

observed. The Chilcoat collection, which was primarily collected

from the Hazard site, contains numerous examples of the Lackawaxen

type points which are also primarily made out of rhyolite.

Whereas the Pearre collection consisted of 32 percent rhyolite,

the Hazard collection consisted of 55 percent rhyolite specimens.

This higher percentage of rhyolite artifacts may reflect the site's

location closer to the rhyolite outcrops of the Catoctin formation;

greater utilization of this area by people based in the Piedmont

province; and/or the presence of a path crossing over from the Gwynns

Falls and Worthington Valley drainage to the Patapsco drainage. The

last hypothesis is supported by the presence of projectile points

identical to the Lackawaxen type defined in the Delaware Valley

(Kinsey 1972).

The presence of bannerstone fragments, both finished and in

process (Figure 39: D-F), of a grooved axe fragment (Figure 39: A),

and of a diversity of tool types suggest that the site may have

served other than as a transient camp. The above items are assigned

to the Late Archaic period and may be associated with the stemmed

points of the Piedmont tradition as this is the predominate point

style found at the site.

External Relations: External relations have been discussed in

the introductory section. The data suggest that utilization of the 116

Piedmont province increased during the Piedmont tradition with a variety of habitats being exploited. Both marine oriented cultures from the Coastal Plain and riverine oriented cultures from the

Monocacy Valley may have established seasonal camps in this area.

Floodplain habitats, upland stream head, upland area, and the serpen­ tine barrens of Soldiers Delight may have been exploited. The last area was the probable source of the serpentine utilized for the manufacture of bannerstones ; the two broken bannerstone fragments suggesting that manufacture of bannerstones was one function occur­ ring in the valley. An intense survey of the serpentine barrens

should produce evidence of the roughing out of bannerstones. A possible path network is suggested by the projectile point types at

the Hazard site.

Laurentian Tradition; Unlike previous tendencies for cultural

concepts to spread from the south to the north, the Laurentian tradi­

tion apparently developed in the Midwest and spread eastward into the

culturally unified area of the Piedmont tradition. The earliest

evidence in the northeast for this tradition is at the Sylvan Lake

Rock Shelter which contained an Otter Creek type point bearing level

which was radiocarbon dated to around 4600 B.C. (Fitzhugh 1972: 8).

Fitzhugh believes that influences from this side-notched point pro­

ducing tradition spread into the mid-Atlantic States between 3000-

2500 B.C. Several complexes have been defined for the New York-Creat

Lakes area; the Vergennes being the earliest complex. The Vergennes

complex and subsequent complexes may not extend south of New York but

related complexes, yet to be defined, are represented in the Coastal

Plain province as far south as Virginia. 117

Ground slate semi-lunar ulus, a number of plummets and side notched points diagnostic of this period are represented in collec­ tions from several sites in tidewater Virginia (McGary 1975: 80).

Side notched points decrease in frequency and become particularly sparse the farther south in Virginia the collections are made (MacCord

1976: per. comm.). One semi-lunar knife and infrequent side notched points are reported from the estuary portion of the Patuxent River in

Maryland (Mayr 1975: per. comm.). Collections from the Coastal Plain and Piedmont province of Maryland frequently contain low percentages of side notched points (Clark and Miller 1975; Stearns 1949).

These assemblages suggest that Laurentian influences were affect­ ing coastal cultures in Maryland and possibly Virginia. The spread of this tradition may be related to the development of mortuary ceremo­ nialism in the northeast. But while the Laurentian tradition in the northeast apparently intruded at the expense of the local Piedmont cultures, in the Maryland area the Piedmont tradition continued to develop; eventually resulting in the assemblages defined as the

Broadspear tradition.

Painters Mill component: One argillite side notched point from the surface of the site may reflect influences from the Vergennes complex of the Laurentian tradition (Figure 31: W ) . Influences from the Brewerton complex of the Laurentian tradition may be reflected by the occurrence of a quartz projectile point similar to the Brewerton

Eared-Notched type points (Figure 31: S). This projectile point type dates to around 3200 B.C. in the Delaware Valley (Kinsey 1972: 407).

Two feldspar rhyolite specimens with rounded upper subelements of the haft and concave base elements may also be related to Laurentian 118

Influences (Figure 31: U). Two side notched points of vein quartz may date to the Late Archaic period and cluster in S50W80 (Figure 27).

The data suggest limited utilization of the site by cultures reflect­ ing influences from the Laurentian tradition.

Pearre collection: Three side notched points of varying styles are represented (Figure 37: H, L, 0). A quartz point has been retouched to the maximum extent and was probably discarded after becoming dull (Figure 37: C). Of the specimens represented, a rhyo­ lite specimen exhibits the greatest stylistic and technical similar­ ities to the Otter Creek type points (Figure 37: 0). Limited utili­ zation of the upland resources of the area is suggested.

Hazard collection: Three rhyolite specimens in the Hazard col­ lection are virtually identical to the Otter Creek type points as defined by Ritchie (1961: 40-41, 97). The three points were retouched at least once (Figure 38: N-P). A rhyolite and a quartzite point may reflect influences from the Laurentian tradition (Figure 38: Q). A

slate serrated gorget (?) from the Hazard site may also date to this

tradition (Figure 39: F). The similarities of these point types to

the northern types support the hypothesis that a path may have

crossed the divide between the Patapsco and the Worthington Valleys

during the Late Archaic period.

External relations: Regional comparisons have been discussed

in the introductory section. The rhyolite side notched point from

the Hazard collection is similar to Group 13 at the UMBC site while

the one quartzite point (Figure 37 : H) from the Pearre collection is

similar to the Group 11 points from the UMBC site (Clark and Miller

1975: Plate 5: S-250; Plate 6). The occurrence of one Group 11 point 119

in Level 3 at the UMBC site suggests that this point style may reflect the earliest influences of the Vergennes complex. An abun­ dance of Otter Creek-like points are represented at the Harmens site in the Patapsco River Valley (Figure 14) as well as other sites in the surrounding Coastal Plain and Piedmont province. Clarification of the interrelationships between the Piedmont and Laurentian tradi­

tions in this area will require excavation of stratified sites and

a more intense survey of surface collections.

Broadspear Tradition; This tradition may have developed in the

Potomac River Valley and spread northward. The tradition is defined

by the presence of steatite vessels. A preference for certain types

of rhyolites, and broad bladed point styles. A shift to greater

dependence on riverine resources is reflected by the sparsity of

artifacts from this period outside of as compared to within major

river valleys. Detailed analysis of this tradition for the Eastern

Piedmont province has been conducted by Ayers (1972) and McDowell

(1970). The tradition was originally defined by Witthoft (1953).

The reader is referred to these publications for elaboration of the

processes represented during this tradition.

Projectile points, steatite artifacts, and debitage diagnostic

of this tradition are not represented in the collections from the

Gwynns Falls floodplain. One rhyolite Perkiomen Broadspear-like

point is represented in the Pearre collection (Figure 37 : DD). The

poor control over the material indicated by the flaking patterns

suggest that people accustomed to working with jasper may have

utilized locally available materials to replace exhausted items in

their tool kits. This hypothesis is supported by data from the Holly 120

Branch site where a jasper Perkiomen Broadspear point was found near a quartz Perkiomen preform which may have been abandoned after an unsuccessful attempt to utilize local materials (Clark 1975b), The

Hazard collection contained a rhyolite point which is similar in style but not in dimension to the Susquehanna Broadspear type

(Figure 38: V). This point style may someday be recognized as an intermediate form between the Susquehanna Broadspear and the Dry

Brook type points.

Isolated pockets of steatite (talc-schist) occur in a geologic envelope of rock surrounding the serpentine barrens of Soldiers

Delight (Growl 1975: per. comm.). While these talc-schist deposits have not been surveyed, a brief reconnaissance of stream systems in the vicinity of the deposits failed to reveal artifacts manufactured

from talc-schist. Only one steatite artifact was recovered from the sites in the Gwynns Falls Valley (Figure 32: W). Apparently, the

talc-schist deposits of this area were not utilized although the

serpentine was extracted during the Late Archaic period and possibly

earlier for manufacture of gorgets and bannerstones. The shift to

riverine sites along the major river systems suggested by other

studies (Ayers 1972) is supported by the sparsity of sites of the

Susquehanna Broadspear tradition in the Gwynns Falls Valley. Only

two sites in the tidewater portion of the Patapsco Valley produced

diagnostic artifacts of this period (Clark 1970; 1975b).

Orient Fishtail Tradition: The study collections do not con­

tain specimens of steatite sherds which are diagnostic of this tradi­

tion. The development of steatite tempered pottery by participants

in this culture marks the end of the Late Archaic period. The Dry 121

Brook point type is the major diagnostic of this tradition although

Orient Fishtail-like points also occur.

Painters Mill component; Two rhyolite points resemble the Dry

Brook type (Figure 31: U-V) while one quartz specimen has a haft- blade morphology similar to both the Orient Fishtail and Dry Brook

types (Figure 31: T). Three points of the Orient Fishtail tradition

cluster in the west central portion of the site (Figure 24). Occupa­

tion was apparently limited to transient camps. Debitage associa­

tions could not be determined because of the absence of diagnostic

attributes for debitage of this period.

Gwynns Falls component: One quartzite Dry Brook type point

(Figure 34: K) was the only artifact which could be assigned to the

Piedmont tradition.

Red Run Valley components: One rhyolite pointed stem Piscataway

type point was found at 18 BA 129 (Figure 36: E). A poorly-manufac­

tured Orient-like point was also recovered (Figure 36: D). A classic

Orient Fishtail type point of black rhyolite was recovered from the

serpentine barrens of Soldiers Delight (Figure 35: T). Transient

camps are suggested. The Hazard collection contains three Dry Brook,

two Orient Fishtail, three Dry Brook-like and one Orient Fishtail­

like type points (Figures 38 and 39). A transient camp is again

suggested with rhyolite points predominating.

External relations: Orient Fishtail tradition type points are

represented in a number of collections from sites in the Coastal

Plain province of Maryland (Cresthull, Mayr, Clark). The data from

the Gwynns Falls Valley replicates observations from similar cultures

in New York State. Ritchie and Funk note the absence of soapstone 122

vessels from the back country sites of this tradition and state "it appears that the rather heavy and bulky pots were not carried to the fall-winter outpost by the necessarily small and highly mobile hunt­ ing parties" (Ritchie and Funk 1973: 346). That the hunting parties were small and highly mobile is supported by the sparsity of and limited diversity of functional tool classes. The known soapstone quarries in Maryland are near water sources which would have supported canoe traffic and facilitated movement of the vessels to permanent or semi-permanent camps (Wilkins 1962; Clark 1974a).

EARLY WOODLAND PERIOD (1000-500 B.C.)

The beginning of the Early Woodland period is defined by the first appearance of pottery in the area. The earliest pottery in the Piedmont province of Maryland occurs at the Monocacy site at the mouth of the Monocacy River (Figure 14). Marcey Creek-type pottery from the Monocacy site is associated with a radiocarbon date of 950

B.C. t 95 years (McNett and Gardner 1971). The Marcey Creek phase represents a continuation of the Orient Fishtail tradition but the substitution of steatite-tempered pottery for steatite vessels repre­ sented a major technological innovation. McNett argues for the con­ tinuation of the Dry Brook type point in the Piedmont section of the

Potomac River while in the Coastal Plain section the Vernon type point continues (McNett and Gardner 1975). Gluckman suggests that the Orient tradition continues into the Woodland period to perhaps

800 B.C. (1973). The Selden Island phase represents the introduction of conical base vessels and surface treatment of vessel exteriors.

But steatite temper continues and the distribution of pottery of this

type correlates closely with the distribution of pottery of the 123

Marcey Creek phase; suggesting a continuum of settlement-subsistence types. For a discussion of the Orient tradition influences in the

Gwynns Falls Valley the reader is referred to the previous discussion in this chapter (pages 119-120).

External Relations: Both McNett and Gluckman have expanded upon the settlement-subsistence patterns and site distribution data

for the Marcey Creek and Selden Island phases (McNett and Gardner

1975; Gluckman 1973). Components of the Marcey Creek and Selden

Island phases have been reported from numerous sites in the Coastal

Plain and along the major river systems of the Piedmont province

(McNett and Gardner 1971; Mayr 1975: per. comm.). But the minor

riverine drainages apparently decreased in importance during this

period. The occurrence of projectile points and the exclusion of

other functional categories of the assemblages, again suggest that

the Eastern Piedmont province may have been utilized by marcoband

units of a highly mobile nature. McNett's hypothesis that Selden

Island and Marcey Creek pottery is not found in the Lower Coastal

Plain province because of the absence of soapstone is incorrect as

recent articles and observation of amateur collections have demon­

strated the presence of soapstone at a number of Coastal sites

(Luckenbach, Allen and Holland 1975). Greater emphasis on seasonally

abundant resources with a resultant decreased demand for the faunal

resources of the Eastern Piedmont is one explanation for the sparsity

of remains in the Corridor for this and subsequent phases. Through­

out this period the tulip poplar and sycamore-box elder-green ash-

silver maple forest units probably continued to expand. This

expansion of low carrying capacity units may have been another 124

stimulus for the increased dependence on resources found in the major river valleys and the estuaries of the Coastal Plain.

Accokeek Creek Phase; Thin quartz or sand tempered, cord-marked pottery and square stemmed projectile points characterize this phase.

A quartz industry may also be diagnostic (Wright 1973). The phase may have begun as early as 750 B.C. (McNett and Gardner 1975), but the close of the phase is still a matter of debate. Wright places the close of the Accokeek phase to around A.D. 100 in the Coastal

Plain while McNett would have the phase ending around 500 B.C. in the Piedmont and Coastal Plain portion of the Potomac Valley. This confusion may be attributed to the development of the Popes Creek phase in the Potomac Valley but a continuation of the ceramic styles of the Accokeek type in the Coastal Plain. The Accokeek phase as defined by Wright (1973; 10-13) makes more sense if considered as a complex. The Accokeek complex would be composed of a number of phases which would incorporate the Popes Creek phase as defined by

McNett and Gardner (1975), the Accokeek Creek phase as defined by

Wright (1973) and several other phases which need to be developed by refined ceramic analysis and an increased sample size.

MIDDLE WOODLAND PERIOD (500 B.C.-A.D. 500)

This period has traditionally been divided into two phases.

The Popes Creek phase (500 B.C.-A.D. 1) represents a continuation of the grit-tempered, cord- and net-marked ceramic and square­ stemmed projectile point tradition. The settlement-subsistence pattern centered around resources of the zone between fresh and salt water (Bandsman and McNett 1974). A similar settlement-subsistence pattern was followed during the Mockley phase (A.D. 1-A.D. 600). 125

Shell-tempered pottery and a lithic industry of various exotic material is diagnostic of this phase. The reader is referred to

Handsman and McNett (1974) for a summary of the most recent concepts

concerning cultural development and adaptation during this period.

These recent concepts may prove to be a simplification of a compli­

cated interaction sphere operative during this period. As suggested

above, the Popes Creek phase may have developed out of the Accokeek

Creek phase in the Potomac Valley but similar developments are not

represented in the middle and upper Chesapeake Bay area.

Painters Mill Component: One side notched knife of the Mockley

phase was found but other artifacts of this period are not repre­

sented.

Gwynns Falls Component: The Calvert type projectile point from

this and the other sites described under the Accokeek phase may be

related to the Popes Creek phase. Three Mockley phase, stemmed

projectile point knives, two of rhyolite and one of argillite, are

present in the Hazard collection (Figure 39; P-Q). The Hazard site

would have served as a convenient resting spot for people traveling

to or from the blue rhyolite sources in Pennsylvania via the

Monocacy-Patapsco-Worthington Valleys (see page 80 for a definition

of blue rhyolite).

External relations; Trade networks with the Adena cultures in

the Ohio Valley developed around 300 B.C. and continued until

A.D. 250 (Wright 1973; Ford 1958; Thomas et al. 1974). Corn may

have been introduced during the Popes Creek phase as suggested by

the presence of corn pollen from the Dismal Swamp in a peat level

dated at 250 B.C. radiocarbon years (Whitehead 1965). 126

Painters Mill Component: Two fragments of cord-paddled, grit-

tempered sherds (Figure 33: A) were found which are similar to the

Accokeek Creek Cord-Marked type. The sherds might also represent an early Late Woodland variety of pottery from the Piedmont province.

Gwynns Fa. s Component: One cord-paddled, grit-tempered sherd

identical to tae two specimens discussed above was found. Except

for this sherd, all the sherds in the Gwynns Falls Valley came from

the Painters Mill site.

Other Components: Two quartz projectile points in the Pearre

collection (Figure 37: S), one quartz point from the Hazard collec­

tion (Figure 38: E), and one quartz point from 18 BA 129 (Figure 36:

F) were of the Calvert type; the diagnostic point style of this

phase. The Calvert projectile point type is probably associated

with both the Accokeek and the following Popes Creek phases as

defined by Wright (1973) and McNett and Gardner (1975).

External Relations: Settlement-subsistence patterns emphasizing

the exploitation of coastal and major riverine environments continued

throughout this period. The eastern Piedmont province served

primarily as a hunting territory for small and mobile groups of

people who left behind little more than an occasional projectile

point or knife. Either the women were not accompanying men during

the hunt or they were not bringing along their pottery. Of course,

this assumption is based on the premise that by this phase the manu­

facture of vessels, which may have been a cooperative venture when

steatite vessels were being produced or even during the beginning of

the manufacture of pottery, was now a task assigned primarily to

women. For a summary of thé Early Woodland period the reader is 127

referred to McNett and Gardner (1971). Only transient camps are expected to occur in the Gwynns Falls Valley for this period.

Refinement in the Accokeek as well as the Mockley pottery types

for this period may be possible in the future (Thurman and Barse

1974). This period has been elaborated upon by a number of recent

reports but due to the sparsity of Middle Woodland period remains in

the Gwynns Falls Valley, this data will not be synthesized. Instead,

the thesis will focus on the developments of the Late Woodland

period. Only by understanding the developments of the Late Woodland

period can we apply the direct historical approach to the problems

of the Middle Woodland period.

LATE WOODLAND PERIOD (A.D. 600-A.D. 1600)

While corn may have been introduced into the region during the

Early or Middle Woodland periods, horticultural activities may not

have become an important part of the subsistence pattern until the

Late Woodland period (Wright 1973). The extensive trade networks of

the Mockley period became dysfunctional by around A.D. 600. The

subsequent ceramic styles and settlement patterns in the Coastal

Plain suggest that the cultures became very localized, with examples

of horticultural, marine and interior sites occurring within the

same drainage system. Extensive trade networks did not begin to

form again until around A.D. 900.

Except for one corner-notched and one large triangular pro­

jectile point from 18 BA 106 (Figure 31: Y) and 18 BA 129, and

perhaps several large quartz triangular projectile points from the

Pearre collection, diagnostic artifacts from the period from A.D. 600

to 1100 are absent. Beginning around A.D. 900 the Monocacy, Potomac, 128 and Shenandoah River Valleys were occupied by people producing grit-tempered pottery and probably practicing corn horticulture.

Medium sized, concave-base, rhyolite, triangular projectile points of the earlier phases of this tradition are represented in the Pearre and Hazard collections and one was found on the surface of the

Painters Mill site (Figures 31: CC; 37: EE; 39: 0). Interior exploitative stations were not established in the Eastern Piedmont province until during the development of the Townsend complex.

Townsend Complex: This complex, defined by Lopez (1961) and

Blaker (1963), accounts for the majority of the shell-tempered vessels recovered from the Painters Mill site. The two phases of this complex, the Little Round Bay and the Sullivan Cove phases, are primarily differentiated on the basis of Townsend Incised and Town­ send Corded Horizontal pottery respectively (Wright 1973). Recent investigations of the Townsend pottery producing Slaughter Creek phase in Delaware has yielded radiocarbon dates for the Townsend

Incised pottery of A.D. 1085 t 75 and A.D. 1285 t 85, and for the

Townsend Corded Horizontal of A.D. 1360 i 60 (Griffith 1975: per. comm.). The former date is in agreement with the radiocarbon date of

A.D. 1105 - 125 for the related Shantok Incised pottery from Connec­ ticut (Salwen and Ottesen 1972). Excavation of numerous features of

the Slaughter Creek phase in Delaware indicates that the Townsend

Incised pottery is replaced by Townsend Corded Horizontal as the

complex developed (Griffith 1975: per. comm.). Therefore, the Little

Round Bay phase precedes the Sullivan Cove phase and not vice versa

as inferred by Wright (1973: 27). The reader is referred to Wright

(1973: 22-24) for a definition of the two phases under discussion. 129

The Painters Mill site is the only component of this complex found within the study area and will be described in detail.

Painters Mill Component: This component is defined primarily on the basis of the ceramic sample. The high acidity of the soil of the site has resulted in the complete leaching of the tempering agent of the pottery and may have also dissolved all evidence of faunal remains.

Ceramic Analysis: Of the 40 sherds recovered, four are so frag­ mentary that temper is the only observable attribute. The remaining

36 sherds represent a maximum of 24 vessels. The shell tempered vessels cluster in three areas: the northeastern section (3 vessels), the central section (13 vessels), and the southeastern section (8 vessels) of the site (Figure 28). Division of the ceramic sample into vessel groups sharing similar attributes will clarify the data.

Little Round Bay Phase: One vessel, defined on the basis of one neck (Figure 33: B) and three body sherds was found distributed over an area 20 by 30 meters in the southeastern section of the site

(Figure 28: 363, 366, 372, 403). The body sherds are fabric im­ pressed and wiped, while the neck sherd exhibits parallel line inci­

sions with a row of gashes incised parallel to the lower line. The

interior paste color is light gray and the paste is well-fired. A

core is lacking. Vessels exhibiting similar paste, temper, and

firing characteristics were recovered from S120W50 (353), S70W60

(212), S40W40 (261) and S170W50 (617). The data indicate that the

site may have been visited several times during the Little Round Bay

phase.

Sullivan Cove Phase: Two neck sherds from two different vessels

suggest occupation during this phase. One vessel (609) exhibits cord 130

wrapped stick, horizontal decorations of the neck and has a higher mica and sand content than any of the other shell-tempered sherds

(Figure 33: C)• The other sherd (266) is characteristic of the

Townsend Corded Horizontal type but the decoration, made with a

loosely twisted cord instead of with a cord-wrapped stick, is a

minority technique of this phase (Blaker 1963: 16). Both occur in

the northeastern section of the site (Figure 28).

Townsend Miscellaneous: Except for the three sherds dicussed

above, the majority of sherds are from the body of the vessel and

cannot be confidently assigned to a particular phase within the

Townsend complex. But sherds tempered with finely crushed shell and

with smoothed-over fabric or cord-marking on the exterior, may be

related to the Sullivan Cove phase (Wright 1973: 23). The Townsend

vessels in the northeastern section of the site (S80W50-S40W70) are

generally thicker (% = 6mm) than sherds from the southeastern and

southwestern sections of the site (% = 4mm). The northeast concen­

tration of sherds (S40W40) may be related to the Sullivan Cove

component (Figure 28).

Townsend Related: Three rim sherds from three different vessels

were recovered which exhibit decorative motifs identical to those

diagnostic of the Potomac Creek complex but the sherds are tempered

with shell instead of grit (Figure 38: D-F). Two rims have applied

lips while one has a straight lip. One is cord-wrapped wire-

impressed with the impression applied parallel to the straight lip;

another is impressed with a tightly wrapped cord in a plat motif

which was blurred by wiping; the third sherd is impressed with a

cord-wrapped stick in a plat motif. The leached tempering agent has 131

left flat and subangular spaces. Sherds with similar temper but wiped on both faces were recovered. Based on the temper similarities,

these body sherds are tentatively assigned to this group. The eleven

body and three rim sherds represent eight vessels (Figure 28).

The diagnostic motifs of this group fall within the general

corded decorative tradition which developed in the Piedmont during

the early Late Woodland period. The decorative motif is part of a

cultural tradition which extended from the upper Shenandoah to the

Susquehanna River during the period from approximately A.D. 900 to

1600. Whether the sample from the site represents influence from

the Piedmont cultures on the Coastal Plain cultures is a question

which will be discussed in Chapter 6. Based on typological analysis,

the ceramics of this group are assigned to a period following the

Sullivan Cove component of the site. The pottery from this group

clusters in the southwestern section of the site around S90W30 (3

vessels), the northwestern section of the site around S60W70 (2 ves­

sels) and the southcentral portion of the site (3 vessels) (Figure

27).

Lithic Industry: The overlapping distribution of ceramics from

the Townsend complex and the Potomac Creek complex prevents the

establishment of lithic association of debitage items (Figures 21,

22, 27).

Little Round Bay phase: Henry Wright (1973: 23) places humped-

back knives and small- to medium-size isosceles triangular points

with concave bases in this period. He does not note the lithic

material utilized but data from the Elkridge site suggest that cobble

quartz predominated. One humped-back knife of cobble quartz was

found at 18 BA 106 and occurs in the same area as a vessel of the 132

Townsend Incised type (Figure 27: 353). Projectile points of this phase are conspicuous in their absence. The site has been collected since 1969 by a local collector who states that mostly triangular quartz projectiles were found. This collection was not studied.

Sullivan Cove phase: In the Coastal Plain province quartz predominates with pebble jasper of secondary importance (Wright 1973:

23, 35).

"Also characteristic of this phase are small convex- edged end scrapers and an industry of small irregular cores of locally obtainable quartz and jasper pebbles from which small irregular points predominate on the surface of Sullivan Cove phase sites." (Wright 1973: 22-23).

A cobble quartz triangular point (14) of this description was found at 18 BA 106 (Figure 31: FF). While cobble quartz debitage and tools are present at the site and may be related to this phase, they may equally be assignable to the Potomac Creek complex. Peoples of both

complexes probably utilized the locally available vein quartz which may have been exposed in the area of greatest concentration (Figure

22).

Other Components: The Pearre collection contains four quartz

projectile points similar to those of the Little Round Bay phase

(Figure 37: AA-CC). One pebble jasper and three quartz (cobble)

projectile points from this collection are diagnostic of the Sullivan

Cove phase (Figure 37: EE-GG). The Hazard collection contains four

projectile points of the Sullivan Cove phase, all made from cobble

quartz (Figure 39: S-U). The remaining sites in the study collection

lack diagnostic artifacts of this complex.

External Relations: Regional relationships are discussed in the

introductory section of this complex (page 128) and will be elaborated 133

upon in Chapter 6. An interior exploitative station of the Sullivan

Cove phase is inferred from the data reported from a site on Soap­ stone Branch, a tributary of the Patapsco (see Figure 14, the Soap­ stone Branch site, and Stearns 1949: 4). Pottery similar to Townsend

Corded Horizontal in decorative motif and paste characteristic but tempered with quartz instead of shell is reported (Stearns 1949).

Similar pottery has been found at sites along the Western Shore section (18 AN 29) and the Eastern Shore section of the Coastal Plain province and may represent influences from the Piedmont on Coastal cultures. Interior exploitative stations of this complex, particu­ larly of the Sullivan Cove phase, are expected to occur in the

Eastern Piedmont region.

Potomac Creek Complex: In the sixteenth century, the riverine and coastal based cultures in Maryland were Increasingly subjected to dislocation, assimilation or death as a result of circumstances which may eventually be recognized as emanating from the indirect impact on local cultures of European contact with aboriginal cultures to the north and south and from the direct impact of drier and cooler climatic conditions during this century. Related to the riverine oriented Montgomery complex of the Piedmont province, the Potomac

Creek complex may represent the displacement of groups of people

from the Mason Island phase by people of the Luray phase (McNett and

Gardner 1975). The coastal oriented cultures were in turn displaced

to the more estuarine portions of the coastal rivers. The Townsend

component at the Elkridge site on the Patapsco River was displaced

by the Potomac Creek component at an early date: the Sullivan Cove

phase is present but poorly represented (Clark 1970). The interior 134

exploitative camps of the Townsend complex may have been utilized during the Ferguson phase of the Potomac Creek complex as documented by data from the Painters Mill site.

Ferguson Phase: The Ferguson phase is defined by Charles

McNett, Jr. (McNett and Gardner 1975). A clayey-textured, crushed

quartz tempered and cord impressed pottery type is the major diag­ nostic of this early phase of the Potomac Creek complex. Corded

horizontal and diagonal as well as decorative motifs which became

popular during the subsequent phases were being developed during

this phase (McNett and Gardner 1975). Small triangular projectile

points, often manufactured from flakes of opaque quartz, and split

cobble scrapers are associated with the permanent stockaded village

sites (Schmitt 1965). The concept of and the necessity for palisaded

villages developed during this phase which also witnessed the intro­

duction of the custom of mass or ossuary burials. While radiocarbon

dates are lacking for this phase, extrapolation from data on the

appearance of similar traits in other areas indicates that the

Ferguson phase developed in the Piedmont province around A.D. 1350

and was in the process of occupying the Coastal Plain by A.D. 1450.

The Patawomeke phase is assigned to the period from A.D. 1450 to the

arrival of John Smith in the upper Chesapeake Bay in A.D. 1608 (the

beginning of the historic period). The Indian Point phase represents

the period of direct contact with Europeans and spans the period from

1608 to the leaving of the last recognized group of Piscataways from

Conoy Island around 1711.

Painters Mill Component: The Painters Mill site is the first

reported interior exploitative station of the Ferguson phase. The 135

1974 interim report elaborated upon the cultural processes of this phase which were inferred from the spatial distribution of artifacts

and from a study of historic documents (Clark 1974c)• But synchronic

analysis requires diachronic control over the data. This analysis will focus on the diachronic aspect of the Painters Mill component while Chapter 6 will explore the culture history of the Potomac Creek

complex.

Ceramic Analysis: The 109 sherds diagnostic of the Potomac

Creek complex which were recovered between 1973 and 1975 represent a

minimum of 40 vessels. The total sample is distributed randomly

across the site, requiring refined typological analysis to differ­

entiate culturally significant clustering.

Potomac Creek Corded Horizontal: Similar to Townsend Corded

Horizontal sherds, the sherds are thick (r = 5.5-8mm, % = 7mm), wide-

cord-wrapped stick-impressed (r = 2-3.5mm, % = 3mm), interior and

exterior surface-wiped, and contain an average of 30 percent crushed

quartz filler (Figure 33: G-I). Two groups of three vessels each

cluster in the northwestern (S80W70 to S40W70) and the northeastern

(S60W30 to S30W30) portions of the site (Figure 30). The sherds are

identical to samples from the Ferguson component of the Elkridge

site (see Figure 14 and Clark 1970). The similarities between this

group and the Townsend Corded Horizontal group of ceramics of the

Sullivan Cove phase indicates that these cultures were interacting

with and influencing each other. The displaced groups of the Sulli­

van Cove phase continued into the Historic period along the Patuxent

River, but they probably abandoned the Piedmont before the Historic

period. 136

Potomac Creek Cord Impressed: Rim or neck sherds from eight vessels exhibiting motifs characteristic of this type were recovered during the surface surveys and excavation (Stephenson and Ferguson

1963). The decorative motifs of parallel and diagonal stick or paddle edge impressions continued, but the width of the stick or paddle and the size of the cord decreased (Figure 33: I-M). Plat decorations occur (Figure 33: N), and loosely-twisted cord-impressed rim sherds were recovered (Figure 33: 0-P). Body sherds representing eight additional vessels may also be related to this group. The

sherds are thinner (x = 5mm) and exhibit a higher percentage of

smudging. Sherds of this group loosely cluster in the southern portion of the site from S80 to 8140 with one sherd found at S170W50

(Figure 30). Several vessels are represented in the northern area

of the site, overlapping the distribution of Potomac Creek Corded

Horizontal sherds. While the ceramic groups may represent functional

types, comparable data from other sites suggest chronological differ­

ences. Accepting the latter interpretation as given, a shift in

residential units, from the northern to the southern area of the

site is inferred (Figure 30).

The assumed relationship between residence areas and the occur­

rence of pottery could not be tested as subsurface artifacts were not

encountered during the 1975 excavations. The postmolds and features

depicted in Figure 11 may be related to either the Townsend complex

or Potomac Creek complex although artifacts of the latter have a

higher frequency in the area. Three-fourths of the soil from Feature

1 was flotated but only wood charcoal was recovered. The postmold

pattern does not suggest to the author the outline of any inferable

structures. 137

Soil creep, frost action, rain wash, cultivation, and other post-depositional processes active during the Historic period have distorted the archeological record. Sherds from the same vessel and flakes resulting from the working of the same variety of lithic material, were found scattered over areas measuring 20 to 30 meters in an east-west direction (downslope) and 15 to 20 meters in north- south direction (Figure 29). The historic post-depositional processes active at the site have disturbed the midden of the Potomac Creek complex and distorted the relationship of the cultural items found therein. While all of this sounds very negative, clustering and association of cultural items are still significant but the post- depositional processes must always be considered before making inter­ pretive statements.

Moyaone Incised: One vessel, defined on the basis of eight sherds, was found scattered over a 20 by 15 meter area around S40W60

(Figure 33: R-S). The sherds are sand tempered and medium in texture.

Gashes may have extended around the neck while the square-lipped rim sherd exhibits tightly twisted cord impressions. The exterior sur­ face is highly weathered. This type, as defined by Stephenson and

Ferguson (1963), probably represents influences from or occupation of cultures from the Piedmont province (Montgomery complex).

Moyaone Related: Six vessels defined on the basis of seventeen sherds, exhibit paste and temper attributes identical to the Potomac

Creek Cord Impressed pottery from the site but the neck sherds are

incised with a variety of decorative motifs which are diagnostic of

the Montgomery complex or Blue Rock phase (Figure 33: T-BB). Body

sherds are wiped on the interior and exterior. Whether the sample

represents the presence of a component of the Shanks Ferry or the 138

Montgomery complex, the experimentation by women of the Potomac Creek complex with decorative ideals of neighboring cultures, or other hypotheses will be discussed in Chapter 6. Moyaone related vessels loosely cluster in all but the southeast section of the site (Figure

29). Sherds from one vessel in the northeast area of the site were scattered over a thirty by fifteen meter area. This area was inun­ dated during tropical storm Agnes (June 1972) which may have contrib­

uted to the distortion. The contruction of Painters Mill Road com­

bined with the silting of the Gwynns Falls floodplain has contributed

greatly to the flooding of the site which has probably been similarly

affected by other severe storms during the past hundred years. Flood­

ing during the prehistoric period was less likely as the area was in

forest cover, the river was not as silted, and Painters Mill Road was

not present to act as a dam to hold back flooding waters.

Lithic Industry: A number of hypotheses concerning the intra­

site settlement patterns of the Potomac Creek complex which were

developed in 1974 have been modified as a result of the 1975 analysis.

The spatial distribution of feldspar and black rhyolite indicates

that this material was not utilized during the Ferguson phase

although black rhyolite does cluster in areas with high concentrations

of Townsend complex pottery. This conclusion is supported by data

from other Ferguson phase sites (Stephenson and Ferguson 1963: 195;

McNett and Gardner 1975) and from the close correlation between

quartz lithics and pottery of this phase at 18 BA 106 as discussed

in the following paragraph. Except for the possible exception of

black rhyolite, the rhyolite artifacts found on the surface of the

site and recovered from the excavations are assigned to the Archaic 139

period although a small percentage may be related to the Mockley phase. A majority of the quartz artifacts probably date to the Late

Woodland period. Except for opaque quartz artifacts, the bulk of the quartz artifacts cannot be assigned to particular phases.

Opaque quartz artifacts overlap the distribution of Potomac

Creek Corded Horizontal pottery (Figure 23 and 30). Six out of the

forty-two opaque items occur outside of the northeast and northwest

concentrations. The direct correlation between ceramic type and

lithic type, the occurrence of opaque quartz at village sites of this phase (Stephenson and Ferguson 1963) and the presence of opaque

quartz, particularly projectile points of opaque quartz which are

diagnostic of the Potomac Creek complex, indicates that at least two

residential areas of the earlier occupations of the Ferguson phase

are represented. Activities within the residence areas included

projectile point replacement, tool manufacture, scraping of soft

material, cutting of soft material, smoking (i.e. pipes), and cooking

or/and food storage. Each residential area is associated with a

butchering area inferred from the tight clustering of scraping,

cutting and chopping tools of vein and cobble quartz (Figures 24-26).

In the northwest area, the artifacts cluster within a ten meter area

centering around S50W70 (Figures 23-26). This area produced a signi­

ficant clustering of quartz debitage as well as split cobble scrapers

and distal knife fragments.

The greatest concentration of vein quartz artifacts occurs from

S50 to S80 (Figure 22). Six vein quartz scrapers, two vein quartz

knives and two chopping tools occur within a ten meter area around

S60W50. The 1974 hypothesis that this was a hide preparation area

is supported by the 1975 data (Clark 1974c: 27). 140

Seven cutting implements, four scraping implements and one chop­ ping implement of vein and cobble quartz cluster due east of the inferred hide preparation area. This area may have been a primary butchering area (Figure 24-26).

The northeast butchering and hide preparation area discussed above are adjacent to but do not overlap with the inferred habitation area. The butchering and hide preparation areas may have continued in use during the occupation of the site by peoples producing the

Potomac Creek Cord Impressed pottery. While vein and cobble quartz debitage occurs in the southern part of the site, functional imple­ ments manufactured from vein or cobble quartz are absent with the

exception of one split cobble scraper (Figure 33: M). The butchering

area in the northeastern portion of the site may have continued in

use throughout the Potomac Creek phase and may have been utilized

during the Townsend complex. This area is close to the Gwynns Falls,

was possibly downwind from the habitation areas and was probably

close to the exposed vein quartz deposits in gulleys. The abundance

of vein quartz in a localized section of the site may be related to

the amount of vein quartz exposed on the surface during the pre­

historic period.

Other components: Other sites in the study area have not pro­

duced ceramics. The Pearre collection contains four quartz Potomac

Creek type projectile points. The Hazard collection contains two

quartz projectile points which may date to this complex.

External Relationships: Richard Stearns investigated two sites

in the Eastern Piedmont region which produced pottery diagnostic of

the Potomac Creek complex. A site (located in the Jones Falls River 141

Valley) produced at least one sherd of what appears to be Potomac

Creek Cord Impressed pottery (Stearns 1966: Figure 2). Sherds from a half-dozen vessels were found at a site at the head of Soapstone

Branch, a tributary of the Patapsco River (Figure 14). The sherds were similar to shell-tempered, incised and fabric-impressed vessels from the Coastal Plain but differed in that they were quartz-tempered and cord-decorated (Stearns 1943: 4, Plate V). The sherds probably are of the Potomac Creek Corded Horizontal type.

Interpretation: Interior exploitative camps of the Potomac

Creek complex, Ferguson phase, are found in the Eastern Piedmont region in the minor river valleys and at the head of tributaries of

the major river valleys. Latter decorative motifs of the Patawomeke phase are not represented which supports the hypothesis that the area was abandoned during the initial development of the Patawomeke phase probably as a result of pressures from the historic Susquehannock

groups. While Susquehannock artifacts were not found, pottery sug­

gesting either the presence of or influences from the Shenks Ferry

complex cultures was recovered from the Painters Mill site. Conclusion

Man has utilized the resources of the study area for at least nine thousand and perhaps as many as eleven thousand years. Through­

out prehistory, the floodplain and upland resources of the Eastern

Piedmont portion of the Gwynns Falls River Valley served primarily

as a hunting territory. The presence of artifacts made from Soldiers

Delight chalcedony suggests that the serpentine barrens of Soldiers

Delight were hunted from the earliest periods of human occupation.

Prehistoric sites were found along the first terrace in all

stream valleys underlain by the Cockeysville Marble formation. The

broad floodplains in these areas and the proximity to the serpentine

barrens as well as upland swamps must have been a major attraction

in the otherwise sloping and hilly valleys of the Eastern Piedmont

region. In areas not underlain by Cockeysville Marble, the valleys

are narrow with steeply sloping sides, depending on the underlying

formations. In these areas, sites are predicted to occur on the top

of the valley slopes (18 BA 101 and 18 BA 128). The upland areas

were hunted throughout prehistory, particularly around stream heads

where upland swamps formed. This resulted in a scattering of projec­

tile points lost or replaced during the hunt, and an occasional

scraping or cutting tool suggestive of initial processing of game.

Beginning around 2000 B.C. and established by 1000 B.C., a new

subsistence-settlement pattern apparently developed which resulted

in decreased use of the study area. The successful adaptation to

marine resources may have decreased the importance of hunting and the

necessity of extended hunting trips to the Eastern Piedmont region.

The introduction of horticulture during the Woodland period and

142 143 particularly of corn by no later than A.D. 500 may have further reduced the importance of this area.

This under-utilization of the resources of the region continued until around A.D. 1100, when components of the Townsend complex began establishing interior exploitative stations in the Piedmont (Painters

Mill site). These groups were replaced by groups of the Potomac

Creek complex who also established an interior camp at the Painters

Mill site. The Potomac Creek people were probably forced by the

Susquehannocks to abandon the area. The firing of the Eastern Pied­ mont which resulted in large barren areas noted during the Historic period (Marye 1955) may have begun during the Townsend complex occupation and continued during the Potomac Creek and Susquehannock complex occupations. The area continued to be utilized by Indians during the early Historic period with the last Indian occupation possibly dating to 1759 (McGrain 1974). The cultured history of the area during the Historic period is described by John McGrain (1974).

Chapter 6 will elaborate upon the development and demise of the

Potomac Creek complex. slseo 3 ^ 3?0 3,40 3?0 3?0 sp spo efO . 2 pO 2 7 0 2 Ç0 2 pO 2 ^ 0 2 ^) 2 fO 3j70 flWYNNS FALLS SITE: IS SA l i t jO METERS NOKTNWEST TRANSFORTATION SCALE CORRIDOR ARCHEOLOSICAL tURVCT - 7 0 IS T 4 .IS 7 S CONTROL SURFACE DATA

Tl. TO T» <•» » - ■ #f— ■*'— — Ü K ■* ------

V JBO •tt [■ k «ft RsM« r -*• J» • •» ;: J? ■M I* •H -5 0 .

JS JW

•10 • • *** «If

JSO IS

•• ••• Jll -20

•t

, •» JO

•4 •! •f .# * *î»

; ...... -,- ,.4-

PA INTERS MILL ROAD

Figure 15: 1974-75 controlled surface collections of the Gwynns Falls site 18 BA 112. NORTHWEST TRANSPORTATION CORRIDOR ARCHEOLOSICAL SURVET. IS SA I IE : IB7S EXCAVATION to « I CONTOUR INTERVALS «.TO

IE OF 1*78 CULTIVATION «.SO sssy J tS O

to

«.to

« .1 0

RÔXD

Figure 16: 1975 excavations of 18 BA 112.

Ln W ♦90

^ 8 0

SCALE «O S I OF JT» Cu lt iv a t io n ITI, *70 Sfr •IW .47% •44$ ^*4 çV*^^4V#7|! J" I f 1 . — TT *60 t f l . . 4 $ 4 ,4 7 # />10 .4 4 » 4 T # 4#&iii2_ 4## #ni Nl7#.l9»,t#4 ■.M.S J3— nr , 4 ' ,ai4 U 'l 1 ;#“ ]«■ — *60 so# ««V44C 47r »## *4lS 4 4 » 4»T#»rJV l '-' j» 1% % • ,zl *T Wot47 4 « 4 * « # | SIX’''Si -ff. a*7[ t» r ,A t4#.*t49 I«1 A ’f* Li' i . 4Ù- '■ ••a* '"hf- 7 •»•# *404 .»#! M4 # # '•> l a r *40 < 'r %#^ ,11» ' * %#0 U. ,*0 1 .4 *7 ,#» •••* m,' « « **44 ,.44» I ***' s#rs#7 - H X h - e |^ » 4 'SO sn 4io\>?rm M»# 4#Ty .4 ;*/ • ‘ L. w« •c. ■Æ aot, ML ** ***^ S»T4»> t4H.ll i'F - 'W ^01 n r - 111- •in 4 ## ■720 1‘S*‘ *M. PAINTERS MILL SITE: IB B* 10# »n & % # 4 ICOLLICT!•’•' .Cl lea NORTHWEST TRANSPORTATIONS I jZl_ CORRIDOR ARCHEOLOGICAL SURVEY I s iW - 10 I 1974 - 1979 CONTROLLED SURFACE. - Î . EXACT PROVIENCE DATA S • ■.N .4 1s t) 14b iA> 1^0 11*0 lob 90" 90 7b 90 80 40 "36 2 6 Ï T

Figure 17: 1974-75 controlled surface collections of the Painters Mill site 18 BA 106,

ON IM 7 0

□ m ,

im w n r 6 0

• T t « M•T4WS0

3 0

PAtNTERS MILL SITE IB BA I OB NORTHWEIT TRANSPORTATION CORRIDOR ARCHEOLOSICAL SURVET. IBT5 EXCAVATION BO * # CONTOUR INTERVALS

Figure 18: 1975 excavations of 18 BA 106. ÏOWCTER SC/J-E cocr ep tt75 ImjiWTjCM , - - @

J» ii. ;j33 Il)aovtso "■•'* .7» •;a*..4 .71..*p#

B .U . 9 2 <11 *60 'tl*420 ...24l.*t4« _':0f .»#« *ii” w #mo ;tOt .4»r I0« U2 *9<3 ^ I 4*^^^ \ 91/ •M9 "99.909.ac* 420 '4:f999 P« •JTT* 27| '138•|28 18291.2 I IVf 122.7> 124. lit. .424 , 129 1.411 39* .19 427 419 |4.vi.t34 30* \a? 2 i'*'* PAINTERS MILL SITE: 18 BA lOQ l u t / « C f iCeLklCT ION NORTHWEST TRANSPORTATIONS CORRIDOR ARCKEDLOCICAL SURVEY I 374- 1375 CONTROLLED SURFACE, OD.U.A EXACT PROVIENCE DATA EfO io

Figure 20: Distribution of feldspar rhyolite debitage (lines connect specimens of the same variety), \o ,iOMpeiiy lO tl OF «T8 ««yriWTiOt _ %tùUê,P»9 9l

» .isi.xplzo^tio

M i69'0 « > mM i il ” >77 .<*• dUtll nrm 4«ri4 •MSI 4 4 4 #1

. , , r % ' L • 4#4"0< .4*7 ,n 449 I *#pB4#7 »*F

70 W # # * m itMlt- .>n M .MV

PAINTERS MILL SITE: |« BA 10# NORTHWEST TRNtSPORTATIONS CORRIDOR ARCHEOLOGICAL SURVEY 1974 . 1*79 CONTROLLED SURFACE. EXACT PROVIENCE DATA

Figure 19: Distribution of cryptocrystalline artifacts; brown chert (O), Soldiers Delight chalcedony ( C 3 ) , Point of Rocks jasper (O)» brown jasper ( ^ ) , gray jasper , excavated material

7 %$^ ; connecting lines represent artifacts of same variety of material. 00 SCALE eoer e. i»ts «uiJiwTie!# _

»i»r •*»

r* m ;v.? B.U . s ?t5 *50 14711) I*. o» IVt *1 iii2 #« e#o ur> *28»

u«• III .*419 ••• 4ft/in«ao- •929U« ,3 7 0 7C%'" 301. !?4 ”^ ' “ ■• JCLlût l‘«î* PAINTERS WILL SITE: 18 BA lOQ I t 7 9 IUI7 ACt fCOkLlCT lOM NORTHWEST TRANSPORTATIONS CORRIDOR ARChEDLOGICAL SURVEY -I 1974- 1975 CONTROLLED SURFACE. EXACT PROVIENCE DATA

Figure 21: Distribution of black (d) and plain* ((O) rhyolite debitage (lines connect artifacts of same variety of material). *Plain rhyolite includes all rhyolite types except black and feldspar. o lOUITtB [50: OF HT8 eULJIWTIOK .

I •’=îhiw^. TTTTnr [TO 2^:4.4 I .021 412 (4)4 A74.I31 114 I • ll!**320 8.U. 5 J4ri4 € MCI )4S^«^ .1 I 4 ) 444**ii ■»•■ W <5> 49t /T I9»> I .410 MiJ J3?:..i»i.‘2 72 to 12*472J1» ro»lOl*

i t z •••* &

0 " ^ ® 'I5t I I C I PAINTERS WILL SITE: IS BA IOC ■ I ■ I V AC* iccLLtCT loa NORTHWEST TRANSPORTATION CORRIDOR ARCKEDLOCICAL SURVEY 1974- 1775 CONTROLLED SURFACE, EXACT PROVIENCE DATA

ë o • 5b

Figure 22: Distribution of vein quartz debitage (CD). Ln t o o r 0» 1* 7 : euLTivATiou .

»ITf •>•

*• try«71 >7t.i3i ni ©..• .«i‘.b‘.’.0)jT"..V 8.U.»

to ti *50 •il3*3 >44S4$ awf*a 444*,‘ II »». /84 UM.tt lO X * #8 ■■0- MC ~ 10140% .4 t7 tc# ,tix QQI , 441 I ItOMtr Jit JIO C at AA£L '42flt1 .*«•* 42T_4_1> ; ; r . -t >‘S** PAINTERS MILL SITE: 10 BA 106 NORTHWEST TRANSPORTATIONS CORRIDOR ARCHB3L0GICAL SURVEY 1974- 1975 CCNTROLLD SURFACE, C9.U,4 EXACT PROVIENCE DATA EfO Y )

Figure 23: Distribution of opaque (O) and cobble (C3) quartz debitage. Ln sa OMÎT ; ÎDOÎ OF lira _ ^ ri ,4 7 : *177 ni Trrrsr y 31 .Oil 206.7)0 """" .70 al) .tii.ap t 4 1 2 SI. ./)%0 4 # ) , I 4 7 # 11 8 , .*.4) 233 . *21 * 9 ^ » 4 # f 4TT.* X M . 81 l # 4 8.U.5 l ia : .. "G%K4"h:_"'..T, Wo»4iXJ8# âïoTT <*i 04fl4« MO» 4I4/.44I S ' l * * 4 4,4-" 1471*3 '41#tJii &c:Liioi *;**, ## 101 ,417 .11# 4 #4 1»0M#T •13) •:.I'«4I ,37* 41/4 *,“». I '««• ,*41» !iiê' ■”• vnîîî " 4 2 7 43 > •.4M,4)7 , i ) j <9 # I PAINTERS MILL SITE: IS BA lOQ pcsLLicra r i c c 10» NORTHWEST TRANSPORTATIONS CORRIDOR ARCKEDLOCICAL SUKVEY I 974 . 1975 CONTROLLED SURFACE EXACT PROVIENCE DATA

Figure 24: Distribution of scrapping tools by lithic type; rhyolite «>), Ln quartz C 3 ) , cryptocrystalline < C 3 ) . W roc* OF 1*75 yj-J'ï^nôM

,17? n # zrrrm H±.K_q% ,i9i .TO 48) .71 %f# •St» » # iM'UtOU.P 3I& lu/jil id 1, 4*7» .tl* &L *o .19 t4 J4ri4« MSI 4i4*U4B IW tTS iîs’j »4 a 44» Sf» 2/r Va‘.Va„„' '“ •■*' i24.‘..m ‘/.;gfl r r »04 .111 MC j'.» ni4 fit 1 K M IT , .11: Z i t , U S • 1*1 - I______it < 47fm e €T ©>0» 1ISS. a e tA a i u;

Figure 25; Distribution of cutting tools by lithic type; rhyolite (O). vein quartz (O) , cobble quartz (O), cryptocrystalline 5 Z D ) . Lnjs lourr SC.'iE f o e r OF 1*79 eULTlMT^Oÿ, _

"7 MI.;S3î #»TT n* -ri*5ci # 4*4

*T* II) 4 t & •**» 4TT * * i*5"):o.#»r »#( 8.U.S iM Ü FpU Ka . 4.1 j3# 120 'iii.in 2*4*7# W0» <•1 «5 0 ,347*141 MO# JW*I5 11*44 22^ *4. J9 1*0 V|»*I47 *1*3 i04.'..t*i‘.; nOr 24# #1*1. *f* 'lO l .ail ML llO'lf* "13 .50# .at# »«#*,*» 4?/j*S

42# 42T4)5 4 M .134 l‘cf * PAINTERS MILL SITE: 18 SA 100 icckLicr 10# NORTHWEST TRANSPORTATION CORRIDOR ARCHESLOGICAL SURVEY I S74 - 197S CONTROLLED SURFACE, (DD.U.A EXACT PROVIENCE DATA 0 ?b

Figure 26: Distribution of chopping tools ), bannerstones (C D ), hammerstone (CD), Ln fire creacked rock (O). Ln W

‘SÛ

’80 JOWETERB if»r- — — n# SC/J-E toer or itT» CuLTiwTiow _ 17^ *70 O. 01 fO» 1/ 9* r .. ’ " . . .S ll J77 •3M .44# "% •«« s *A6 a u .1.7 •«» 4S0 / ! M. j c m Tt# !3*l2l. .424I 42» • .Kfc *44» 400 14) . n » * -iîU2^ y > i __ !)CI 1D7 . 900 "HT U1 1 7 1 120 »04 PAINTERS WILL SITE: 18 EA 108 itut/ACt JT» {CCLkter lea NORTHWEST TRANSPORTATIONS i l î - CORRIDOR AROHEOLCCICA-SURVEY . 10 HSO* I 9V4 . 1775 CONTROLLED SURFACE, EXACT PROVIENCE DATA s OD. U,< 1 5 b 14b jÿo iÿo iib lob* s G" bo iQ ^0 • 5b 40 10 ’0

Figure 27: Distribution of piercing implements by type; triangular (A)> Piscataway ( ^ ) , Ln side notched (CD), Orient Fishtail tradition CD), corner notched 0 •), stemmed , ON |10UÏTERp SCALE eooi OF lore oultivatiok

*74 .aw J T T *58

• îOiF.eîp2P» "9!: «*&' 4#k7T# I 5) 2*4 I*S**J B.U.) 4 ,*9* ««< 3TCH’229 artJ«4 9T9 2 2 r 2*5 la r F /I 24*. .24 * .4 20 •£*> .2 3 » l'\V, ;u J f.V I*. *799 • « J I 3 3 r o . 10» > 7 * w9 ? 723'

"4VL.:'.:a.»o».»o* 4I//12- .,40 420 •42r3*2

n

I . r PAINTERS MILL SITE: 18 BA 106 II « t l liu . r «CK fO LLl.T 10. NORTHWEST TRANSPORTATIONS CORRIDOR ARCHESLOGICAL SURVEY 1974 - 1775 CONTROLLED SURFACE, EXACT PROVIENCE DATA 0 3 . U , 4 sro Ÿ0

Figure 28: Distribution of Townsend complex vessels (lines connect sherds from same vessel); Townsend Incised (dj , Townsend Corded Horizontal (O ), Townsend related ( C D ) , U1 Townsend undiagnostic (I l) , excavated material (isa,*, ♦). ~sj lOUETER P _ - n i l------EDO: or 1*73 tULTlVATiCIt _

au> au. 0 4 .1 9 2 992

!tO L22I 3ICU4. ./^20 Sifr , l«^as 299 * * ieJ**i20Ji*yjii/ 3*7.' 777 B .U . 9 — —1 ÎÏO ’ilj,!!.. 224*79 ______«*t *60 4S S39 4I4.,.44* 47 >?2 9 . rt24# i)3*) 579 2 J r 949.*?U 499. f.:\i -Ù 2 )9 ,A 244. .1 4 9 '41# ot*_ua 27**99 r 909 .3*1 M9. >2.*72 .0>, f *.94/*# COT. I t u m :. IO* ,0k 19b* ( ,119 9 4 " sao M f . *c#‘j .1.2 1'f % i? 3*0'4*7 "902 >?9.I }39^_ •*•» ■ fcai-j .1 ___ .311 247...it«i .2 9 9 * l# 9 ^C< J 9 0 399 ,,I7'44I II# •41 #33.909.909 I T lil* X* •2*«,286 :r,« r ” 27.U*.«2* 342 .424I 4:9 , 1:3 I1.43# '4*2 I 429* • .300 /*3 " *”’ 4:74)> '4\#.l)4

PAINTERS WILL SITE; 18 BA 108 NORTHWEST TRANSPORTATIONS CORRIDOR ARCHEDLOGICAL SURVEY 1 9 7 4 - 1975 CONTROLLED SURFACE, EXACT PROVIENCE DATA OD. U,4 ^ 0 Yo

Figure 29: Distribution of Moyaone related (dD) and Moyaone Incised (O) vessels (lines Ln connect sherds from same vessel). 00 IOm e t e r S c i Z Ë ^ C30I or t*7a cw>JiVAT20!t _

.IT ? • » • V 8.U.9

0 ; > ' ‘‘t “.L

SI J $ 9

74 »01, .550 ' ' II? 5C5_3Û2._I____t»3* *JB9LU na.^01 ÜHI PAINTERS MILL SITE; 18 BA 108 |c&\kCT 10* NORTHWEST TRANSPORTATIONS CORRIDOR ARCHEDLOGICAL SURVEY 1974- 1975 CONTROLLED SURFACE; EXACT PROVIENCE DATA FO • 5 0 go

Figure 30: Distribution of Potomac Creek complex vessels (lines connect sherds from same vessel); Potomac Creek Corded Horizontal (CO), Potomac Creek Cord Impressed flCZJ), Potomac Creek Ln miscellaneous - plain (^ ) , cord paddled I) , excavated material ( ^ ,fgi). VO 160

Figure 31: Lithic artifacts from 18 BA 106.

Proveni­ Lithic Letter ence Data Description Material

A 170 tabular core brown chert B 449 flake knife brown chert C 348 flake end and side scraper brown chert D 618* bifurcate base point vein quartz E-F 11, 21 generalized side notched points vein quartz G 581 ground base side notched point vein quartz H 198 polyhedral core S, D. chalcedony I S72W52 flake end scraper brown chert J 351 flake spurred end scraper S. D. chalcedony K-M 322, 57, 572 Vernon type points quartzite N, S 322** shallow side notched point vein quartz 0 169 Lackawaxen Contracting stem knife plain rhyolite P 316 Lackawaxen Expanding stem point feldspar rhyo­ lite Q ** side notched point and knife banded rhyolite R 320 Dry Brook orient? point & knife vein quartz T 473 Orient Fishtail point vein quartz U-V 511, 407 Dry Brook orient-like point feldspar rhyo­ lite w 44 Otter Creek point argillite X 317 expanding stem knife vein quartz Y-Z 61, 319 corner notched points opaque quartz AA 567 small stemmed point cobble quartz BB 75 Selby Bay knife plain rhyolite CC ** Mason Island triangular point feldspar rhyo­ lite DD-EE 495, 37 Piscataway points opaque, cobble quartz FF 14 Sullivan Cove point cobble quartz GG-KK 519**, Potomac Creek points opaque quartz 245**, 244

* Found in gulley north of site at base of Setters Quartzite formation ** General surface IftMM* ’I * ' * ' *1 *

At

T9I 162

Figure 32: Lithic Artifacts from 18 BA 106

Proveni­ Lithic Letter ence Data Description Material

A 553 unifacial knife and scraper plain rhyolite B, C 443, S32W60 ovoid base bifacial knives quartzite D 102 straight base bifacial knife vein quartz E 301 flake retouch knife vein quartz F,G, L 435, 26, 2 adze plain rhyolite H, I 109, 587 bifacial flake knives vein quartz J 582 bifacial knife and chopper vein quartz K ** scraper and spokeshave banded rhyolite M-0 374, 531, 263 keeled scraper vein quartz P 324 hafted bifacial scraper feldspar rhyo­ lite Q 80 unifacial flake scraper feldspar rhyo­ lite R, S 494, 541 convex edge unifacial scrapers vein quartz T 218 thumbnail scraper vein quartz U 156 distal knife fragment cobble quartz V 519 bannerstone fragment steatite w S42W64 bannerstone or pipe fragment soapstone X 214 spokeshave vein quartz Y-AA 163, 160, 95 adze vein quartz - 574 broken blank bannerstone steatite BB 107 flake knife opaque quartz CC 125 flake unifacial scraper opaque quartz DD 36 proximal knife fragment opaque quartz EE ** Little Round Bay point blank cobble quartz

** General surface 163

m

L ■ 164

Figure 33: Pottery from 18 BA 106

A: 198, Accokeek Cord Marked B: 403, Townsend Incised C: S42W52, Townsend Corded Horizontal D-F: 381, 490, 547, Townsend Related G-I: 545, S38W40, 486, Potomac Creek Corded Horizontal J-Q: 285, 90, **, 539, 333, 454, 473, 444, Potomac Creek Cord Impressed R, S: 206, S42W64, Moyoane Incised T-AA: 290, 476, 360, 232, 554, 433, 167, 368, Moyoane Related BB: 429, cord marked body sherd, Potomac Creek complex CC: 326, smoothed over cord-marked body sherd, Potomac Creek complex DD: S38W40, smoothed body sherd, Potomac Creek complex 165

# 8 6

« m

u *## AA BB

,7 18 19 llo ill llg 13 14 ll‘

9 IS IC . ONn.jii

I t I I I I I 1 1 t I I I I I 1 I I 1 I ( I I I ( f I I ' I f i ' I ' I ' I ! I I ' ^ ' I ' ' ' ’ ^ ^ ' 1 ' ^ I ’ ' ' ''I 166

Figure 34; Lithlc and ceramic artifacts from 18 BA 112

Proveni­ Lithic Letter ence Data Description Material

A S320W20* Kirk Corner Notched point gray chert B 67 MacCorkle Stemmed-like point plain rhyolite C S350W20 Lackawaxen Expanding Stem point feldspar rhyolite D, E ** Lackawaxen Straight Stem point feldspar and plain rhyolite F ** Lackawaxen Contracting Stem point plain rhyolite G, H S340W65, Selby Bay knife plain and banded S330W40 rhyolite I S290W80* Piscataway point gray chert J ** Clagett point quartzite K S370W65 Dry Brook point cobble quartz L 27 hafted knife vein quartz M S340W8* bifacial knife vein quartz N 43 side scraper cobble quartz

0 ** adze plain rhyolite P S370W50 end scraper and knife banded rhyolite Q, R S302W2, distal knife fragment feldspar rhyolite S325W40* S 14 knife and spokeshave plain rhyolite T 18 adze and knife black rhyolite U S330W65 flake side scraper and knife feldspar rhyolite V ** Accokeek Cord Marked pottery grit temper W 'k'k bifacial knife and scraper plain rhyolite X S295W10* Marrow Mountain I-like point base feldspar rhyolite Y 50 thumbnail end scraper feldspar rhyolite

* Exact provenience coordinates ** General surface 167 Figure 35: Lithic Artifacts from the Gwynns Falls Valley

Site Provenience Letter Number Data Description Lithic Material

A BA 125A 18* retouched stemmed knife feldspar rhyolite B BA 125C 5 flake end scraper knife plain rhyolite C BA 125A 24 contracting stem point vein quartz D BA 125B ** flake fragment vein quartz E BA 125B ** flake end and side scraper S. D. chalcedony F BA 127 ** Orient Fishtail-like point plain rhyolite G BA 127 ** retouched Lackawaxen Straight feldspar rhyolite Stem point H BA 127 ** flake scraper opaque quartz I BA 127 ** Kanahwa Stemmed-like point vein quartz J BA 127 WO-WlOO bifacial flake knife vein quartz K, L BA 127 W400-W500, flake side scraper and knife black and plain rhyolite W200-W300 M, N BA 127 W400-W500 examples of cleavage planes vein quartz

0 BA 127 W4G0-W500 straight back scraper vein quartz P BA 122 18 stemmed drill feldspar rhyolite Q BA 122 8 flake knife plain rhyolite R, S BA 122 5. 15 cores cobble quartz T *** ** Orient Fishtail point black rhyolite U Red Run Random Find Kirk Comer Notched point black rhyolite V, X BA 128 ** knife fragment and retouched point feldspar rhyolite and vein quartz w Red Run Random Find Lackawaxen Straight Stem point vein quartz

* Exact provenience map not provided for 18 BA 125A-125C, and 122. ** General surface *** From serpentine barrens of Soldiers Delight ON 0 0 169

0 I 170

Figure 36: Lithic Artifacts from 18 BA 129

Proveni­ Lithic Letter ence Data Description Material

A 2 1 * LeCroy Bifurcate Base point plain rhyolite

B 6 straight stem knife feldspar rhyolite C 17 straight stem point plain rhyolite D 5 Orient Fishtail-like point plain rhyolite E 19 Piscataway point feldspar rhyolite F 2 2 Calvert type point quartzite

G, H 2 . 1 generalize side notched points vein quartz 1 8 bifacial knife S. D. chalcedony

J, K 1 1 , 2 0 flakes S. D. chalcedony L 15 Susquehanna Broadspear point? vein quartz M 1 2 triangular blank vein quartz N W20S100 flake banded rhyolite

0 3 graver feldspar rhyolite P W20S100 triangular knife vein quartz Q, R 21, 9 bifacial flake knives plain rhyolite S 15 flake knife quartzite T W160S10C keeled end scraper vein quartz

See Figure 12 for exact provenience location. 171

•«IIIA B C D e

1 0 1 2 3 4 5 6 7

■ C M .

I NCHES

10 12 172

Figure 37: Bifaces from the Pearre collection.

Letter Description Lithic material

A, C Kirk Corner Notched-like point brown chert, plain rhyolite B Palmer Corner Notched-like point feldspar rhyolite D-E LeCroy Bifurcate-like points cobble and opaque quartz F Eshback or Eva-like points cobble quartz G, M Dry Brook Orient-points cobble quartz H, 0 Otter Creek-like points quartzite, banded rhyolite I Big Sandy-like point plain rhyolite J bifacial knife and burin feldspar rhyolite K knife and graver vein quartz L severely retouched Otter Creek Point clear quartz N, T Halifax-like? point cobble quartz P, U, untyped points plain rhyolite, quartz Q corner notched point feldspar rhyolite R, W, X Lackawaxen Contracting Stem points cobble quartz, feldspar and plain rhyolite S Calvert point cobble quartz V Lackawaxen Expanding Stem point plain rhyolite Z side notched point cobble quartz AA-CC Medium size triangular points cobble quartz, plain rhyolite DD Perkiomen Broadspear point plain rhyolite FF-GG small size triangular points black chert, cobble quartz 173

n □

V W X

1 0 1 2 3 4 9 6 7

10 1* 174

Figure 38: Bifaces from the Hazard collection.

Letter Description Lithic Material

A-D, F-K broad stemmed Piedmont tradition feldspar and plain points and knives rhyolite, vein quartz E Calvert point cobble quartz L Lackawaxen Strait Stem point feldspar rhyolite M, R Lackawaxen Expanding Stem points feldspar and plain rhyolite N-Q Otter Creek points feldspar and plain rhyolite S-T Orient Fishtail points feldspar rhyolite U Orient related point feldspar rhyolite V small Susquehanna Broadspear point feldspar rhyolite W Dry Brook point feldspar rhyolite 175

f l

M N

1 0 1 2 3 4 5 6 7

V w 10 12 I li 176

Figure 39: Lithic artifacts from the Hazard collection.

Letter Description Lithic Material

A full grooved axe fragment trap rock B side scraper gray chert? C core end scraper Point of Rocks jasper D preform bannerstone fragment serpentine E polished bannerstone fragment serpentine F notched bannerstone? fragment slate G, K expanding-base drill plain rhyolite and cobble quartz H-J end scrapers opaque quartz L Dry Brook-like point plain rhyolite N-M.R-T untyped points plain rhyolite and cobble quartz 0, S, U medium triangular points feldspar rhyolite and vein quartz P-Q Selby Bay knives argillite, blue rhyolite 177

4 4 CHAT’TER 6 : THE POTOMAC CREEK COMPLEX

Introduction

The chronological ordering of the data from the Gwynns Falls

Valley was a prerequisite to the elucidation of cultural processes and the evaluation of the significance of the archeological resources directly affected by the proposed development in the Northwest Trans­ portation Corridor. Attempts in 1974 to develop a synchronic analysis of the Late Woodland period complexes at the Painters Mill site was partly unsuccessful because the time depth of the assemblage was not

considered (Clark 1975a). Activity areas were assigned to the Potomac

Creek complex which in reality represented activities spanning

several periods. The increased sample size resulting from the 1975

controllv d surface collection combined with the subsequent diachronic

research enabled testing and substantial revision of the hypotheses

developed in 1974. The revised hypotheses were presented during the

diachronic analysis in Chapter 5 and will be elaborated upon in this

chapter. The diachronic analysis of the Townsend and Potomac Creek

complexes provides a framework from which the ensuing cultural

historical analysis of the Potomac Creek complex will proceed.

Culture history studies are a combination of the objectives of

synchronic and diachronic studies. As discussed earlier, diachronic

studies develop models of change of cultural systems over long

periods of time and usually focus on specific problems. Synchronic

studies develop hypotheses correlating the subsystems of a single

culture within a given time period into a structural whole. Culture

history studies attempt to explain the changes of a cultural system

throughout the history of the culture and focus on the changes of

178 179 the subsystems of a culture in response to environmental or cultural variation in the surrounding region. Culture history studies are a refined form of diachronic studies which is concerned more with problems of culture process than culture development.

The discovery of a few sherds of Potomac Creek series pottery at the Painters Mill site in 1973 raised two primary questions which oriented the direction of subsequent research at the Painters Mill site. What cultural processes resulted in the establishment of a

Potomac Creek camp so far north of the developmental center of the

Potomac Creek complex along the Potomac River? Does part of the assemblage at the Painters Mill site represent a hunting camp sub­ system of the Potomac Creek complex settlement-subsistence system?

The first requires the development of culture history hypotheses which explain the formation, expansion, and retreat of the Potomac

Creek complex during the late prehistoric and protohistoric periods.

The second question requires a synchronic analysis of the known village and upland sites to develop alternate hypotheses concerning changing subsistence-settlement systems. Both questions necessitate examination of historic documentation of Indian cultures in the

Chesapeake Bay region and of archeological and historic documentation of other Indian cultures in the Eastern Woodlands.

Clearly, this detailed analysis exceeds the time limitations imposed by the completion deadlines of most conservation archeology projects. But the core of the research design was developed during an independent study course in ethnohistory; the paper from which

served as the basis for directing the 1974 excavations at the Painters

Mill site (Clark 1974: b, c). The attempt at purely synchronic 180

analysis in 1974 was abandoned in 1975 because the synchronic analysis was applicable primarily to the Painters Mill site while the 1975 report to the highway department required a diachronic analysis applicable to all the sites within the Corridor. The final report incorporated the 1974 synchronic analysis in determining the significance of the Painters Mill site in relation to the surrounding region.

The culture history analysis of the Potomac Creek complex repre­ sents hypothesis formulation based on data examined since the author's initial analysis of a Potomac Creek component at the Elkridge site

(Clark 1970). Testing of the hypothesis must await an exhaustive analysis of available collections and controlled surface collections and excavations of other Potomac Creek villages and interior camps.

The data from the Painters Mill site are of value for formulating and testing only a few of the hypotheses advanced in the research design. The research design was developed to provide a framework for future research in the Coastal Plain and Eastern Piedmont regions.

Researchers involved in future conservation archeology projects in

the region can test and reformulate the relevant hypotheses if they

find components of the Townsend and Potomac Creek complexes during

their surveys. The same applies to the diachronic research design

developed in Chapter 5. The diachronic research design should be

applicable to the Eastern Piedmont province. The culture history

research design will find wide applicability by researchers conduct­

ing surveys from the Rappahannock River in Virginia to the Susque­

hanna River in Pennsylvania and from the Western Piedmont to the

Eastern Shore section of the Coastal Plain province. Researchers

who find Townsend or Potomac Creek complex material during the course 181

of their surveys will hopefully refine this research design and thus avoid the primary fault of many survey reports; the failure to relate the resources within the direct impact zone to the resources of the surrounding region and to contribute substantially to the study of man.

Methodology

A persistent debate in anthropology centers around whether our knowledge of the past is only as good as our knowledge of the present.

Ethnologists have argued that the perishable nature of the archeo­ logical record and of cultural processes prevents the recovery of cultural data from archeological remains. Lewis Binford (1968) believes that phenomenon, particularly those of a processual nature covering considerable periods of time, can be observed by the arche- ologist but not by the ethnologist. He argues that: "If we define

an hypothesis as the statement of a relationship between two or more variables, and if both variables are observable in the archeological

record, then the hypothesis formulated is testable." (1968: 268).

Archeologists concerned with explanation of cultural process depend

upon the knowledge currently available on the range of variability

in form, structure, and functioning of cultural systems. Enthnolog-

ical data from the point of view of the archeologist is background

information for explaining the similarities and differences observed

in the archeological record. The interpretations of the archeolog­

ical record between observed phenomenon and ethnographic or historic

phenomenon allow the development of postulates which must be sub­

stantiated by a number of testable hypotheses (Binford 1968: 269).

Under this conceptual framework, ethnographic data serve as resources 182

for testing hypotheses which seek to relate material and behavioral cultural phenomenon and serve as the basis for models of particular social relations which are postulated to have been the context for an observed archeological structure (Binford 1968: 270).

While some investigators argue that the archeological record must be studied in and of itself and that analogy will forever supply

explanations of only varying degrees of certainty (Dunne11 n.d.).

Binford emphasizes that we are still dependent upon the knowledge

currently available on the range of variability of cultural systems

for our interpretations of the past. Regardless of one's theoretical

or disciplinary orientations, inferences, models, or hypotheses are

formulated on the basis of an ethnocentric conception of reality

which is modified to varying degrees by intensive studies of past

and present cultural systems. The hypotheses and models formulated

and tested on the basis of the archeological record are in reality

influenced by the researchers' comprehension of past and present

cultural systems as reported in historic documents and the anthro­

pological literature.

Given that archeologists must study other cultural systems to

formulate explanations of the archeological record, the closer in

time and in space the ethnological or historical data is to the

archeological record, the greater the validity of the analogy between

the units under comparison. The greater the comparability between

analogous units, the greater the reliability of the derived infer­

ences. A basic criterion for analogy between an archeological

phenomenon and a historical and ethnological phenomenon is that the

cultures under comparison should be at similar levels of cultural 183

development and in similar ecosystems. If an archeological site can be demonstrated as being the site described in an historic account, then by tracing the ethnological and historic data of the living descendants of this culture, anthropologists and historians can account for the cultural bias of the early records and gain insight into the problems of acculturation. This approach of start­ ing from the known and working back to the unknown has been labeled the direct historical approach (Fenton 1952). Asking anthropolog­ ical questions of historical material, the approach has been success­ fully developed by historian and anthropologist alike (Wright 1968,

Spicer 1962).

While a number of disciplines may assist in the direct histor­ ical approach, only three are essential to the method; ethnology, history, and archeology. Ideally, two or more scholars from these disciplines should work collectively on the direct historical method with emphasis on their discipline. But when research is conducted

by only one researcher, disciplinary bias may result. This natural

disciplinary bias will often be reflected in his less than adequate

handling of other contributing disciplines. If such be the case of

this analysis, it will serve to emphasize the multi-disciplinary

nature of this method and stress the necessity of such multi-disci­

plinary projects as the Piscataway studies currently underway at

American University. This research design should be of assistance

to the students participating in these studies.

As only highly assimilated remnants of the indigenous cultures

of the Chesapeake Bay region remain, ethnological studies can con­

tribute little to the direct historical approach and are given minor 184 consideration. The indigenous cultures of the Eastern Piedmont province failed to survive to the beginning of the historic period.

With the exception of the Delaware, only assimilated remnants of the

Eastern Shore cultures remain (Parker 1936; Weslager 1943; Seabease

1969). The highly acculturated descendants of the Potomac Creek complex Indians have only recently become concerned with what remains of their cultural heritage (Scarupa 1976). To the south, the Rappa­ hannock (Speck et al. 1964), the Catawba (Speck 1946) and the Nottoway and Pamunkey (Binford 1967) have survived to the ethnological present but are similarly highly assimilated.

The ethnohistorian is confronted with an absence of historic documentation of the indigenous cultures of the Piedmont province

and the area north of the Patuxent River on the Western Shore of the

Chesapeake Bay. The historic aboriginal groups along the Potomac

River, who produced artifacts we recognize as the Indian Point phase

of the Potomac Creek complex, have been the concern of several his­

torical accounts of major importance to the understanding of previous

cultural development (Spelman 1613; White 1632). Historical accounts

of the Patuxent Indians (the people who produced Sullivan Cove phase

artifacts) are also of value (Smith 1612; Pory 1624). Historical

data concerning surrounding cultures will also be incorporated when

applicable. But the analysis will depend primarily on the archeo­

logical record for development of the majority of the hypotheses.

Formative Processes

Beginning around A.D. 900, a marked increase in the utilization

of the area extending from the Shenandoah to the Susquehanna River

Valleys is suggested by the appearance of several complexes. The 185

riverine oriented cultures which developed in the Piedmont province during this period formed the base from which the Potomac Creek culture evolved. The development of population aggregates in the major river valleys in the Piedmont may be attributed to the accept­ ance of additional cultigens, specifically beans, into the horticul­ tural complex. The cultigens of tobacco, com, beans, melons, pumpkins and squash were probably present by A.D. 900 although only corn and squash have been recovered (McNett and Gardner 1975).

Inadequate sampling techniques may be partly responsible for the absence of the remaining cultigens; all of the relevant sites were excavated prior to the development of wet screening and flotation techniques.

The cultures in the corridor extending down the Shenandoah, up the Potomac, along the Monocacy and throughout the Susquehanna

Valleys shared many related cultural processes and were probably in close communication via the various trail systems which survived into the historic period (Marye 1920: 250; Wallace 1971: 105). The trail systems would have served as the routes for the movement of marine shell, other commodities, and people from Coastal Plain cultures to the cultures in the interior. The Late Woodland period was a time of complex interaction between various cultures in the Piedmont and between culture in the Piedmont with cultures to both the east and

the west. The absence of palisades and the similarities in assem­ blages of different but related complexes strongly suggest that warfare was of minor importance during the period from A.D. 900 to

1300.

In an unpublished manuscript which synthesizes (for the first

time) cultural development in the Potomac River Valley, Charles 186

McNett has defined a number of complexes in the Piedmont province

(McNett and Gardner 1975). McNett's complexes and phases for the

Potomac River Valley and Wright's (1973) complexes and phases for the Middle Chesapeake region have been adopted as a basis for further elaboration. To understand the development of the Potomac Creek complex and early trade networks, the various complexes from the period A.D. 900 to 1300 will be summarized. The complexities of the cultural interactions during this period are only suggested by the available data. All of the complexes discussed require division into phases and elaboration of the subsistence-settlement patterns operative during this period; a task requiring renewed and refined investigations beyond the scope of the present analysis. The com­ plexes of concern during the Late Woodland period from A.D. 1300 to

1600 have already been defined (pages 127-141) and will be discussed within the larger perspective of the Chesapeake Bay region.

A ceramic series similar to and possibly derived from the

Albermarle Cord and Stony Creek Fabric-Impressed types characterize

the Catoctin Creek complex. Based on material excavated from the

stratified Monocacy site (Figure 14), McNett has differentiated the

Aquaduct phase (McNett and Gardner 1975). Cord and fabric-impressed

pottery tempered with crushed limestone or quartz (chert rarely) and

occasionally decorated with incised rims has been labeled Aquaduct

Cord and Aquaduct Fabric-Impressed pottery (McNett and Gardner 1975).

The Aquaduct phase also includes earth burial mounds which date

from A.D. 900 to 1300. Of particular significance is the Lewis Creek

Mound located in the Upper Shenandoah Valley. This mound produced

radiocarbon dates ranging from A.D. 1000 to 1300. Excavation of the 187

mound revealed a number of burial levels which yielded an abundance of marine shell beads, ceramics, pipes and other artifacts. The lowest levels produced a high percentage of marginalia beads which decreased in frequency in the upper layers; being replaced by beads of the "wampum" or tabular type (Manley 1963: 38). This change in popularity of bead types may be of significance in understanding the ossuaries of the Potomac Creek complex. The excavations of this highly disturbed mound in 1965 revealed 159 lower jaw fragments but from 200 to 400 individuals are estimated to have been buried in the mound (MacCord and Valliere 1965: 38). Calcined bones, charcoal lenses, and burned shell beads suggest cremation burials or some form of fire ceremony associated with the burial complex. A number of the flexed and bundle burials were covered with stone slabs. The abundance of shell pendants, olivella, marginella and columella beads and fossil shark teeth documents the extensive trade with coastal cultures along the Chesapeake Bay and Atlantic Coast where these shell species are obtainable. Shell beads of the same species have been recovered from other components of the Catoctin Creek complex; supporting the hypothesis that widespread trade networks were estab­ lished during the early part of the Late Woodland period. The incis­ ing techniques exhibited on some Aquaduct pottery specimens from the

Monocacy site are identical to the incising motifs of ceramics from the Little Round Bay phase of the Townsend complex. The potters of the Aquaduct phase were probably experimenting with the design motifs of the coastal cultures and in some instances, potters from the coastal cultures may have actually lived at the various Piedmont sites. Except for the component at the Monocacy site, the Aquaduct 188

phase sites are confined primarily to the Upper Shenandoah and James

River Valleys.

While the Clemson Island complex in the Middle and Upper

Susquehanna Valley is poorly understood (Heisey 1971) , this complex is contemporaneous with the Catoctin Creek complex and is apparently closely related. The ceramics of this complex are cord and fabric- impressed, tempered with quartz or chert, and decorated with simple designs executed with the edge of a cord-wrapped paddle, sharp stones or bone points (Jones 1931: 96; McNett and Gardner 1975). Most of the designs consist of horizontal lines. According to McNett and

Gardner (1975) , this pottery is almost identical to the Aquaduct

Cord and Fabric-Impressed types. The Clemson Island complex is named after the Clemson Mound which is similar to the Lewis Creek

Mound in several attributes. Although almost totally destroyed by plowing, the mound yielded remains of at least nineteen individuals.

The placement of stone slabs over the burials and the occasional cremation of individuals were practices shared by both the Clemson

Island and Lewis Creek mound builders. Cannibalism may have been practiced by the participants in the Clemson Island complex (Jones

1931: 95). Very little material culture was recovered from the

Clemson mound and marine shells were absent. Additional research of

the Clemson Island complex is needed as investigations in the Susque­

hanna Valley have focused primarily on the Shenks Ferry and Susquehan-

nock complexes (Heisey 1971).

A similar complex is found on the lower Shenandoah, the Potomac,

Piedmont and the Monocacy River Valleys but the lower Susquehanna

Valley was apparently sparsely occupied during the period from 189

A.D. 900 to 1200. The Montgomery complex was originally defined as

the Montgomery focus (Slattery, Tidwell, and Woodward 1966) and has

been refined and relabeled by McNett and Gardner (1975). The ceramics

of this complex consists of collared or plain rims decorated with

simple slash marks or cord-wrapped-stickdesigns. Quartz temper pre­

dominates in the lower Shenandoah Valley while granite temper pre­

dominates in the Potomac Piedmont. The Biggs Ford site in the middle

Monocacy Valley yielded grit tempered, cord-roughened pottery decor­

ated with rows and plats of cord-wrapped-stick impressions with

occasional exterior and interior punctations. The pottery from this

site combines features of the Shepard Cord-Marked type and the Clemson

Island types (Bastian 1974). Shell disc and drilled columella shell

beads were recovered from the graves. The graves were scattered

randomly across the area excavated.

The Biggs Ford and other sites of this complex have yielded

medium size triangular projectile points of rhyolite and quartz and

oblique angled pipes. A circular pattern of trash pits, the sparsity

of feature in the middle of most sites, and the documented house

patterns suggest a circular concentration of houses around a central

plaza. The Biggs Ford site was palisaded but the palisade may date

to the subsequent Luray phase occupation of the site. Most of the

Montgomery complex burials were individual interments which usually

lacked grave goods although marine shell beads and drilled oyster

shell pendents frequently were placed in the graves of infants. The

Winslow site along the banks of the Potomac River received dates on

charcoal which ranged from A.D. 825 f 75 and A.D. 1315 t 80 radio­

carbon years. The Mongomery complex was contemporaneous with the 190

Catoctin Creek and Clemson Island complexes, and was probably in contact with both.

McNett believes that the Montgomery complex was but little removed from the Potomac Creek complex. Witthoft also noted the similarities in ceramic types and the experimentation evident at the

Selden Island and related sites in the Potomac Piedmont (1963: 6 6 ).

Partial excavation of the Selden Island site may have revealed a

circular concentration of houses surrounding an open plaza (Witthoft

1963). However, Howard MacCord, who participated in part of the

excavations of the Selden Island site, states that not enough of the

site was excavated to ascertain the presence of a plaza or a house

although part of a palisade line was located (Kent and Smith 1972:

11). The graves at the Selden Island site contained native copper

beads, a variety of marine shell beads, exotic pipes in the styles

of the Ohio and Tennessee Valleys as well as other artifacts

(Witthoft 1963: 6 6 ). Contact with Mississipian and Ohioan groups is

also suggested by the subsequent development of ossuaries, cremations

and confederacies which characterize the subsequent Potomac Creek

complex.

Handsman and Hunter (1972) have plotted the distribution of

Historic period Indian sites along the Potomac River and the distri­

bution of sites producing pottery of the Potomac Creek complex

(Figure 40). This map not only shows the overlap of locations of

the two sets of data but also illustrates the concentration of

Potomac Creek sites in the Eastern Piedmont portion of the Potomac

River Valley. An examination of pottery obtained during a surface

collection of the Selden Island site in 1974 revealed sherds 191 identical in decorative technique to the Potomac Creek Corded

Horizontal type (Clark 1974a). Unfortunately, the excavated material from this site is currently in Florida and was not available for study. The Selden Island and related sites in the Eastern Piedmont portion of the Potomac River Valley were the locations of the evolu­ tion of the Montgomery complex into what has been labeled the

Ferguson phase of the Potomac Creek complex. This hypothesis will need additional testing but should prove to be valid. The cultures

in the lower Shenandoah and Monocacy River Valleys may or may not have evolved along similar lines. But the abandonment of the Pied­ mont province by the participants in the Potomac Creek complex may

have followed soon after the formation of the complex. The people

in the Eastern Piedmont portion of the Potomac moved downriver into

the upper portions of the estuarine Potomac River while the cultures

in the Monocacy River Valley may have moved into the upper Chesapeake

Bay and lower Susquehanna Valley. 192

POTOMAC CREEK COMPLEX SITE DISTRIBUTION MODIFIED FROM HANDSMAN AND H U NIER, 1972

Miles Monacacy River Potopsco

Ferguson end PotowomeMe Phosi - prehistoric Sites . Conowlngo 2 .Pointers Mill 3. Blond 4. Soapstone Broncti 5. Elkridge 6.Hormens 7. Obrecht 8. West Bonk - Forked Ck 9. Nottlngtiom 10. Selden Island Indlon Point Phot* Historic Sites 11. Nocotctitonke 12.Tessa motuck 13. Moyoons 14.Clnquaetuck 15. Pomacocack (Acco keek Creek ?) 16. Motougtiquomend 17. Nussomek IS. Nushemoulk IB.Potopaco 20. Secowocomoco 2 1. Monanouk 22. U tm m d (Yooconoco) 23. ZatavKoman 24. Patawomeke 2 5. Porno coco k 26. Touxenent 27. Unnumed 28. Nomossmoment 29. Assomeck

Figure 40: Potomac Creek Complex Site Distribution. 193

Development During the Ferguson Phase

The Ferguson phase is defined on the basis of the material and features uncovered by Alice L. L. Ferguson at the Accokeek Creek site

(page 134). The Ferguson phase pottery from this site is very similar to the late Montgomery complex and early Ferguson phase material found on sites in the Potomac Piedmont (McNett and Gardner

1975). The site may represent the actual relocation of Piedmont based peoples of the Potomac Creek complex following their abandon­ ment of the Piedmont province. The absence of radiocarbon dates for the Potomac Creek complex requires the extrapolation of dates from other sources to estimate the approximate movement of Potomac Creek cultural groups into the Coastal Plain province.

Excavations of the Accokeek Creek site revealed seven successive palisade lines which were inferred by Ferguson and Ferguson (1960:

11) to represent about 300 years of occupation before John Smith's visit in 1608 (A.D. 1300). A circular stockade 300 feet in diameter was discovered 150 feet south of the outer stockade of the major village area bringing the total number of stockade lines to eight

(Stephenson and Ferguson 1963: 55). At least one of the stockade

lines (c-d) consisted of posts which apparently took root after planting, suggesting that honey locust were utilized (Stephenson and

Ferguson 1963: 50). This was apparently the first stockade erected.

The occupants may have harvested the trees which would be the easiest

to cut but which would last a long time: locust. As the first

stockade would have depleted the available locust from the area, a

variety of species may have been harvested for the later stockades,

although transport of the desirable species via the Potomac River 194

may have facilitated selectivity of species. The tremendous energy required to transport over land the quantity of logs necessary to build a stockade may have been one factor considered in locating

the village along the bank of the river.

If the primary purpose of the palisade was defense, then at

some point following initial construction, the posts would have

decayed to such an extent that repairs to the palisade would no

longer suffice; necessitating the construction of a new stockade.

If chestnut posts, which have been reported to last from 50 to 80

years, were used for the palisade, then the palisaded village would

be between 400 and 640 years old; a much too early date for the

Potomac Creek complex. If the posts, five to eight inches in diameter,

represent the remains of honey locust of similar diameter, then we

can estimate the rate of decay and predict how often the stockade was

repaired. Personal observations of the decay of locust post planted

in sandy loam soil in 1949 near Baltimore, Maryland indicate that

locust posts have an average life span of from 20 to 25 years. While

Ferguson feels that the Accokeek Creek site was abandoned around

A.D. 1630, Thurman places the abandonment of the site to an earlier,

unspecified date (Thurman 1973: 40). If the eight stockade lines

representing rebuilding of the palisade every 20 to 25 years, and

assuming the site was abandoned around A.D. 1600, then the palisaded

village was occupied for 160 to 2 0 0 years which places the construc­

tion of the first palisade between A.D. 1400 to 1440.

This time range is based on many assumptions, but the dates

fall within the range for the earliest stockaded village reported

from Virginia; the Sullins site, which was radiocarbon dated to 195

A.D. 1425 - 95 (MacCord 1976: per. comm.). These dates are also in

accordance with the Piscataway Indians' traditional history as told

to the English in 1660 (Archives of Maryland, Proceedings of the

Council 1660: 403). This tradition states that the first "king" of

the Piscataway came from the Eastern Shore and was succeeded by

thirteen generations of rulers without interruption until the time

of Kittamaquund who died around A.D. 1640. If we accept the defini­

tion of a generation as the average difference in age between parent

and offspring, then by estimating the age at which the Indian was

considered to be an adult, we can guess the age of the development

of the Piscataway "confederacy." Ubelaker, in his study of the

Hurley ossuary at Nanjemoy Creek, noticed subadult-adult differences

among the miscellaneous bones (1973: 21). These differences

may reflect a subadult-adult dichotomy which were recognized cultural

institutions during life and represented after-death as well. The

rites of Huskenaw, which apparently occurred around the age of 15,

may represent the ceremony which demarked the obtainment of adult­

hood. This does not seem to be an unreasonable age for a generation,

particularly since the life expectancy at birth of the people of the

Potomac Creek complex was about 21 years (based on the data from the

Hurley ossuary). Subtracting 195 years (15 years x 13 generations)

from A.D. 1640 we get a date of A.D. 1445 for the formation of the

Piscataway "confederacy;" the approximate date of the construction

of the first palisade at Moyaone, the supposed center of the

Piscataway Confederacy.

According to the traditional history of the Piscataway, we can

infer from the passage quoted below that the Piscataways may have 196

occupied a substantially larger territory than the area occupied along the Potomac in historic times.

"...that long agoe there came a King from the Eastern Shorre who commanded over all the Indians now inhabiting within the bounds of this province (naming every town severally) and also the Patowmecks and Sasquehannaughs, whome for what he did as it were imbrace and cover them all they called Vitta- poingassinem..." (Archives of Maryland, Proceedings of the Council 1660; 403).

Given our current knowledge of the historic relationships between the Piscataway and Susquehannocks, (Ferguson and Ferguson

1960) it is highly unlikely if not impossible that people of the

Potomac Creek complex ever dominated the powerful and agressive

Susquehannocks. But the distribution of Potomac Creek Corded Hori­

zontal and Cord Impressed pottery suggests that the ancestors of the

Piscataway and related tribes along the Potomac River occupied the

upper Chesapeake Bay region throughout the Ferguson and Patawomeke phases until displaced southward towards the close of the Patawomeke

phase. Participation of the historic Patawomekes in the Piscataway

Confederacy before the domination of the southern shore of the

Potomac by the Powhatan Confederacy is alluded to in a letter of the

Governor and Council of Virginia to the Board of Trade;

"Have also revenged the treachery of the Piscaticons and their associates, the greatest of those parts, for cutting off Captain Spilman and Mr. Putis' pinnace and murdering great numbers of their ancient allies, the Patawomekes"(Calender of State Papers, Colonial; 1574-1660: 56, In Marye 1935: 186).

Except for the excavations at the Patawomeke site (Schmitt 1965),

very little is known about the early Potomac Creek complex sites

between the Potomac and Rappahannock Rivers. The Patawomeke site

may have been occupied from the close of the Ferguson phase, through 197

the Patawomeke phase and into the historic Indian Point phase.

Schmitt (1965: 11-12) does not provide sufficient information to substantiate this inference, although such data is available in the National Museum collection. His pottery descriptions and illus­ trations indicate the possible presence of Moyaone Incised pottery and Potomac Creek Corded Horizontal pottery but Potomac Creek Cord

Impressed pottery apparently predominates.

Turning to the postmold data, if we assume that the inner ditch represents a fence surrounding the ceremonial center of the site as suggested by Stephenson and Ferguson for a similar feature at the

Accokeek Creek site (1963: 53), then the village had from two to seven stockade lines. Schmitt believes the data suggests two con­ struction phases. Similar patterns of parallel concentric ditches have been reported from southwest Virginia and inferred to represent the excavation of soil which was placed on both sides of a palisade located between the ditches; the remains of which have been subse­ quently obliterated by cultivation. If this was the case at the

Patawomeke site, then five stockade lines are predicted to have existed through the life of the village. The five stockade lines may represent 100 to 125 years of rebuilding. As the site was abandoned sometime between A.D. 1634 to 1669 (Schmitt 1965: 5), we can infer that the site was first occupied sometime during the beginning of the sixteenth century; later than the initial occupation of the

Accokeek Creek site by people of the Potomac Creek complex. The data suggest that the areas south of the Potomac were occupied later

than the areas north of the Potomac River; an hypothesis first

advanced based on a study of the distribution of Moyaone and Potomac 198

Creek Corded Horizontal pottery and to be elaborated upon as the discussion proceeds.

Expansion During the Ferguson Phase

Early research on the distribution of Potomac Creek Cord

Impressed pottery revealed it did not extend very far south of the

Rappahannock or very far north of the Susquehanna Rivers (Manson,

MacCord, and Griffin 1943: 411). Current research has supported this hypothesis and clarified the east-west distribution as well. Except for the very early village sites in the Potomac River Valley and occasional hunting camps like the Painters Mill site, Potomac Creek complex sites do not extend west of the fall line and are concen­ trated along the upper portion of the estuarine rivers. This prefer­ ence for the less saline portions of the coastal rivers is attributed in part to the Piedmont, riverine base from which the Potomac Creek complex developed. Potomac Creek sites decrease toward the east but are reported along higher salinity portions of the coastal rivers and even in Kent County on the Eastern Shore of Maryland (Wilke 1974: per. comm.).

The widespread distribution of ceramics of the Potomac Creek

complex in the Coastal Plain cannot be explained as simply resulting

from an invasion of the Coastal Plain and displacement of the indig­

enous Townsend complex cultures. The shell trade, established at an

early date (A.D. 900), probably resulted in repeated interaction

between the Coastal and Piedmont adapted cultures. Experimentation

in similar ceramic manufacturing and decorative motifs in both

regions suggest the sharing of ideals as well as of stylistic or

technical innovations. That the Potomac Creek cultures developed as

the dominant force in the Coastal Plain may reflect interaction with 199

cultures in the Ohio and Mississippian drainages. The spread of the religious beliefs in conjunction with ossuary burials must have increased the political cohesiveness of related groups of the Potomac

Creek complex. Both displacement and assimilation of the Sullivan

Cove phase coastal cultures probably occurred; those groups desiring political and social automony moved outside of the areas environmen­

tally or politically desirable to the peoples of the Potomac Creek complex.

The data from the Accokeek Creek site suggest the establishment of a village by people directly removed from the Piedmont-based

Montgomery complex cultures in an area formerly occupied by peoples of the Townsend complex. Both the Potomac Creek Corded Horizontal

and Moyaone Incised types from the site are similar to the Montgomery

complex. The Moyaone Incised pottery type is probably contemporane­

ous if not slightly earlier than the Potomac Creek Corded Horizontal

pottery and is very similar to the Blue Rock phase of the Shenks

Ferry complex which is located along the lower and middle Susquehanna

River. Both Moyaone Incised and Shenks Ferry Incised share similar

collar forms, similar uncertain and wavering incising techniques,

and similar design motifs (Stephenson and Ferguson 1963: Plate 18: J,

Plate 19; Heisey 1971: Plates24-25). The Blue Rock phase is contem­

poraneous with the Montgomery complex. Based on the ceramic data,

the Moyaone Incised pottery at the Accoteek and other sites in the

Coastal Plain is inferred to represent the earliest establishment of

Piedmont derived groups or individuals in the Coastal Plain province

during the Late Woodland period. These early groups would have pro­

duced a minority of Moyaone Incised pottery and a majority of Potomac 200

Creek Corded Horizontal pottery. As the Potomac Creek complex developed, the incising technique was abandoned and the Corded Hori­ zontal decorations were refined and elaborated upon until the Potomac

Creek Cord Impressed pottery began to predominate. These ceramic styles represent a continuum of gradual change. Variation in temper and decorative techniques may someday prove of value in establishing synchronic differences of the allied groups of this culture.

Located near the tidal head of the Patapsco River, the Elkridge site contained both Ferguson and Patawomeke phase components and was probably a palisaded village site although the area of the palisade has probably been destroyed by gravel removal (Clark 1970: 52). The absence of Moyaone Incised pottery suggests that the site was occupied after the initial development of the Ferguson phase ceramic styles.

Potomac Creek Corded Horizontal pottery predominated but complicated design motifs and serrated lips characteristic of the Potomac Creek

Cord Impressed pottery are also represented. The sherds are quartz tempered and are identical in texture to the sherds of the Potomac

Creek Corded Horizontal type. While grit temper is a primary diag­ nostic attribute of Potomac Creek pottery, the kinds of temper may reflect variation between cultural groups as well as over time.

The largest percentage of sherds from the Elkridge site was of the Townsend Incised type with only four sherds of the Townsend

Corded Horizontal type pottery recovered (Clark 1970: Plate II; 12-

14, 16). The Townsend Corded Horizontal type pottery is inferred to represent the acceptance by various cultures in the Coastal Plain of the corded horizontal decorative motifs derived from the Montgomery and related complexes in the Piedmont province. The near absence of 201

this type of pottery at the Elkridge site suggests that the site was abandoned by the Sullivan Cove phase cultures before occupancy of the site by people of the Ferguson phase.

The Harmens site, located only a mile south of the Elkridge site on a tributary of the Patapsco River, yielded pottery of the Townsend

Incised, Townsend Corded Horizontal and Potomac Creek Corded Horizon­ tal types (Figure 40). Two vessels were recovered which are identi­ cal to the Townsend Corded Horizontal type in all attributes except temper. These sherds were crushed quartz temper instead of shell tempered and shared identical corded design motifs as one sherd from the Elkridge site (Clark 1970: Plate II: 12). This data contradicts the Elkridge site data; suggesting that certain groups of the

Sullivan Cove phase may have adopted ceramic manufacturing and deco­ rative motifs of the Ferguson phase. An examination of Norman Brice's collection from a site on the Severn River revealed vessels identical to the Potomac Creek Corded Horizontal and Cord Impressed types except that the vessels were tempered with shell instead of quartz

(Brice 1976: per. comm.). This hypothesis implies that the Ferguson phase cultures were dominant but may not have displaced all of the cultures of the area. Absorption of individuals from the local area into the new settlements probably occurred. Groups influenced by the

Potomac Creek complex but desiring autonomy would have been forced to migrate to other areas; an hypothesis to be tested by an examination of data from the Patuxent River Valley.

The Painters Mill site was apparently a hunting camp for

cultures of the Townsend and the Potomac Creek complexes. This is

the only known site of both complexes which has been controlled 202

surface collected and the only hunting camp component intensely analyzed. Previous discussions have traced the culture history of the Late Woodland period occupations at the site. The occurrence of sherds with identical decorative motifs as the Potomac Creek Corded

Horizontal type but with shell temper (Figure 33: D-F), suggests that the site was occupied by people of the Sullivan Cove phase who had adopted the decorative motifs of the Ferguson phase. The presence of

Moyaone Incised type pottery suggests early influences from cultures of the Montgomery complex and may represent actual occupation of the site by groups of the Montgomery complex. If the evolutionary sequence of the Potomac Creek pottery types from Moyaone to Corded

Horizontal to Cord Impressed is valid, then the site may represent the occupation of Sullivan Cove groups' hunting territories by

Montgomery complex groups. The Montgomery complex groups or families may have periodically returned to the same camps until their ceramic styles had evolved into the Potomac Creek Corded Horizontal type. An even more speculative hypothesis which requires further research is that some of the Sullivan Cove groups, influenced by the cultures of the Ferguson phase, developed identical ceramic techniques and pro­ jectile point types as the Potomac Creek complex. While the over­ lapping distribution of Townsend and Potomac Creek complexes supports this hypothesis, the distributional data may also reflect environ­ mental parameters such as the location of dry land, the availability of quartz detritus, the distribution of plant species or reoccupation of previously cleared areas (Figures 18-30).

On the Susquehanna River, the Conowingo site has produced ceramics of the Townsend Incised, Potomac Creek Corded Horizontal, 203

Potomac Creek Cord Impressed and Shenks Ferry Incised types (Stearns

1974; Plate 13; Cresthull 1976: per. comm.).

Two sherds of the Townsend Corded Horizontal type, only tempered with crushed quartz, are of interest in understanding the development of the Shenks Ferry cultures (Stearns 1943: Plate 13). Do these sherds offer slight evidence of the development of the Shenks Ferry complex from a Montgomery and Townsend complex base? Heisey has pointed out the close similarities between Shenks Ferry wares and wares of the Montgomery and Catoctin Creek complexes (1971: 55-56).

But he does not "suggest that Shenks Ferry ceramic traditions were derivative from any of these southern- and central-Piedraont wares."

(Heisey 1971: 56). But the rejection of the southern origin hypoth­ esis for Shenks Ferry is based primarily on the inconsistency in dates; the southern complexes were previously assigned to the ter­ minal part of the Late Woodland period. However, the Montgomery complex apparently preceded and was contemporaneous with the Blue

Rock phase of Shenks Ferry (McNett and Gardner 1975). The major source of the initial Shenks Ferry population in the lower Susque­ hanna Valley may have come from the Monocacy Valley; a poorly docu­ mented area from an archeological and historical perspective.

Researchers should be less hesitant to extend Shenks Ferry influences south of the Pennsylvania state line. The discovery of Funk Incised pottery at the Tuscarora Rock-Shelter in the Monocacy River Valley

(Geasey 1971: 8 ) and the presence of Shenks Ferry Incised-like sherds of the Accokeek Creek site suggests that these cultures were inter­ acting. The expansion of the Potomac Creek complex to the known area of the Shenks Ferry complex would have resulted in contact and 204

possible conflict between these groups. This might explain why the

Shenks Ferry cultures could not migrate south when pressured by the

Susquehannocks to move away from the rich floodplain of the Susque­ hanna River.

Displacement During the Sullivan Cove Phase

As the Ferguson phase expanded from the Piedmont into the

Potomac Coastal Plain and northward along the fall line to the

Susquehanna River, the indigenous coastal cultures of the Sullivan

Cove phase were either assimilated or displaced. This hypothesis is supported by the experimentation in ceramic manufacturing tech­ niques in some areas and the absence of experimentation and other shared concepts in other areas. The corded horizontal decorative techniques were apparently adopted by certain groups (clans?) of the

Townsend complex beginning approximately at the time of the initial expansion of the Ferguson phase (ca. A.D. 1300-1400). A study of the distribution of Townsend Corded Horizontal pottery in the Tide­ water reveals significant geographical limits (Blaker 1963: Figure

3). The absence of Townsend Corded Horizontal pottery in the upper

Potomac suggests that this ceramic type developed after the Coastal

Plain was occupied by Ferguson phase groups. An examination of the ceramics from the Farmington Landing (Clagett Farm) site revealed only a few sherds of Townsend Corded Horizontal pottery. This site, located on Piscataway Creek, produced an abundance of Townsend

Incised and Townsend Fabric Impressed pottery as well as Potomac

Creek Corded Horizontal ceramics of the Ferguson phase.

A similar hypothesis would explain the absence of Townsend

Corded Horizontal pottery on the sites along the upper Rappahannock. 205

This apparent absence may also reflect an inadequate sample or examination thereof. Only an examination of ceramics from the

Rappahannock sites will clarify the question of when the Rappa­ hannock Valley was occupied by Potomac Creek groups. McNett and

Gardner (1975) believe the Rappahannock was occupied by the forma­ tive groups of the Potomac Creek complex who had moved from the

Shenandoah Valley. But the ceramics from the Keyser Farm site in

the Shenandoah Valley are of the late Potomac Creek Cord Impressed type; suggesting that a small group of Patawomeke phase Indians moved in with the people producing Keyser Cord marked pottery

(MacCord 1976: per. comm.).

Townsend Corded Horizontal pottery is reported from the Severn

River Valley northward to the Patapsco River Valley and may extend

as far south as the West River and as far north as the Susquehanna,

but these areas are not as well documented. All of the sites in the

Delaware Bay and Choptank River areas reported by Blaker (1963)

have produced this type of pottery. An examination of collections

from the Patapsco River Valley, including the Painters Mill site,

revealed various decorative and tempering techniques suggestive of

a period of initial experimentation. But refined decorative tech­

niques similar to the pottery from the Townsend site was not observed

except at the Painters Mill site (Blaker 1963: Plate 8 A: 8 , 11-13).

The initial decorative motifs of the Townsend Corded Horizontal

pottery may have been parallel, oblique or vertical lines and com­

binations of the parallel with the vertical or oblique designs.

Through time, the width of the cord or cord-wrapped stick decreased

and the designs increased in complexity. This evolution in decora­

tive techniques parallels the evolution from Potomac Creek Corded 206

Horizontal to Potomac Creek Cord Impressed pottery. Eventually a

Townsend Cord Impressed pottery should be defined but an adequate sample prevented such a definition at this time. The ceramic

similarities suggest interaction between the Sullivan Cove and

Ferguson phase cultures and the gradual acceptance by the Sullivan

Cove groups in certain sections of the Coastal Plain of many of the

culture concepts of the Potomac Creek complex. Elucidation of the

other cultural concepts adopted must await detailed studies of

available collections.

Not all coastal cultures may have been influenced by the Potomac

Creek complex. The religious concepts associated with ossuary

burials and other cultural traits of the Potomac Creek complex

Indians may have proven unacceptable for certain conservative or

militant groups. Townsend Corded Horizontal pottery is not reported

south of Nottingham on the Patuxent River (Stearns 1943; 1965; Blaker

1963). But Townsend Incised ceramics are abundantly represented.

Inferring from the distribution of Potomac Creek complex sites that

the lower salinity portions of the tidal rivers were preferred by

the Potomac Creek complex groups, the absence of Townsend Corded

Horizontal or Potomac Creek pottery on the high salinity portion of

the Patuxent suggests that Townsend Incised pottery continued to be

manufactured up to and through the Historic period. The Patuxent

Valley may have been a refuge area for the displaced cultures of the

lower salinity portions of the Coastal Plain. Other groups in the

upper Bay area may have been pushed north, joining with the Shenks

Ferry cultures or east, joining with the Nanticoke or Delaware who

were culturally more similar than the Shenks Ferry. 207

Turning to the various sites along the Patuxent River, Stearns notes the occurrence of quartz tempered, cord impressed sherds but does not differentiate between the Accokeek and the Potomac Creek types (1943, 1951, 1965). At the Nottingham site near the head of the Patuxent subestuary (Figure 40), Stearns reports the occurrence of a few sherds of the Potomac Creek types. Mayr (1975; per. comm.) observed the excavation of a conical shaped pit at Nottingham which contained Potomac Creek Cord Impressed pottery. An examination of

Dennis Webb's surface collection and of the pottery on the surface of the site in 1974 revealed that except for a rare sherd of Potomac

Creek Cord Impressed pottery, the predominate ceramic types were

from the Townsend series. Howard MacCord (1976: per. comm.) exca­

vated a test square in 1950 which produced Rappahannock Fabric

Impressed pottery. Other surface collections contain samples of

Townsend Incised as well as Rappahannock Fabric Impressed pottery.

Based on a study of William Marye's personal notes on the

location of the historic village of Mattpament, Stearns associates

the site at Nottingham with the historic site of Mattpament (1951:

19). While the archeological assemblages could be representative of

the protohistoric period, the lack of observed or reported contact

material cautions against placement of this site in the Historic

period. The minority occurrence of Potomac Creek ceramics and the

predominance of Townsend series sherds tend to support the refuge

area hypothesis. The Potomac Creek pottery which does occur would

be expected in sites where there is contact between the peoples. The

predominance of Townsend series ceramics would be expected if this

area was occupied by remnants of the Tidewater culture for over 100 208

years after the abandonment of the other Tidewater sites. Stearns's failure to locate the sites reported by Smith (Figure 41) may reflect his preconceived bias that such sites would produce a quantity of

Potomac Creek ware. If this was a refuge area of Tidewater cultures as postulated, then future archeological investigations should find

Townsend series ceramics in association with historic material although an occasional Potomac Creek sherd is expected.

ÛLXâ0>LJten e83ri iWeiilffm I % J y ^ C : f : héuiiité^nil,, Ul$ïk rwîu ' ' % 'y. Y A .i.

’W ^ A t

r i-A \ X ' jcs f

Figure 41: The Potomac-Paiuxent Portion of John Smith's Map. 209

Smith noted in 1608 that the Indians along the Patuxent "inhabit together and not so dispersed as the rest." (In Arber 1910: 349).

The population of 200 warriors (600 individuals according to Semmes calculations [1937: 716])consisted of three distinct groups; Acquin-

tanacksuak, Pawtuxunt, and Mattapanient (Figure 41), During the missionary efforts of the Jesuit between 1630-50s, the King of the

Patuxent, Machaquomen, was reported to have great influence and

authority among the Indians (Campbell 1906: 299). Whether this meant he had the authority of an emperor such as the Piscataway

"Emperor" is unknown.

In 1621 John Pory reported that the Patuxent Indians were

employed by Opechancanough, the head of the Powhatan Confederacy,

to kill Thomas Salvage for injustices committed against the Pamaunkee

Indians (Smith, In Arber 1910: 568-9). Although this attempt failed,

the incident shows that the Patuxents were friendly with the Powhatan

Confederacy in 1621. The Patawomeke were enemies of the Pamunkey

(Piscataway) in 1622 (Handsman and Hunter 1972: 27). Therefore one

can conclude that the Piscataway and Patuxent represented separate

political entities at that time. In 1666, articles of peace between

the colonists and the Piscataway Confederacy listed the tribes which

accepted the Piscataway emperor as their spokesman. They included

the Anacostank, Doag, Mikikimonans, Mannacatick, Hangemaick, Porto-

backes, Socayo, Pamyoyo, and Choptico but not the Patuxent (Marye

1920: 185). References to the Patuxent Indians as members of the

Piscataway Confederacy are lacking from the primary sources. Why

were the southeastern members of the Piscataway Confederacy fleeing

the lower Potomac occupations for the more densely settled areas 210 when at the same time in 1634, the Patuxent groups were still occupy­ ing villages 42 miles to the north? These villages were abandoned around 1670 as a result of expansion of plantations throughout the upper Patuxent Valley and not necessarily because of pressure from the Susquehannock. While the Patuxents were attacked by the Susque­ hannocks they did not receive the devastating attention given to the groups of the Piscataway Confederacy.

If this was an enclave of the coastal groups, then why did the

Piscataway Confederacy not destroy them? The development of the

Powhatan Confederacy to the south and of the Susquehanna military power to the north must have changed the Potomac Creek position from one of offense to one of defense. As the Potomac Creek people were pressured from the north and south the northern outposts were aban­ doned and the population shifted to the south. A split in the confederacy may have occurred between the southern Potomac members and the northern members as a result of pressures from the Powhatan

Confederacy. The alliance of the two groups along the Potomac is alluded to in a letter to the Governor of Virginia as reported in

the Calender of State Papers, Colonial, 1574-1660 (Marye 1935: 186).

The consolidation of the tidewater people along the defensive posi­

tion of the Patuxent, the riverine orientation of the Potomac Creek

phase, and the growing pressures of new power centers in the Tide­ water all may have contributed to the survival of the coastal people,

if indeed this is the case. Mutual protection against a common

enemy, the Susquehannocks, may have forced these former enemies into

a defensive alliance. Perhaps the Potomac Creek complex people were

concerned primarily with cutting off direct trade between the coastal

cultures and interior cultures and were not concerned with simply 211

displacing the coastal groups to make room for themselves.

South of the Rappahannock, Townsend Corded Horizontal pottery is not reported (Blaker 1963). While influences from the Potomac

Creek complex in the form of ossuary burials are reported from the

Chickahominy River, Townsend Corded Horizontal pottery was not

reported (Barka 1976: per. comm.). However, Rappahannock Fabric

Impressed and Incised pottery is abundantly represented on Late

Woodland sites in the area (Evans 1955). The Chickahominy Series

as defined by Evans (1955) contains pottery of the Mockley, Swans

Point and Rappahannock or Townsend types (McNett and Gardner 1975).

The Townsend, Rappahannock and Chickahominy Fabric and Incised wares

are regional expressions of a closely related ceramic tradition of

the Late Woodland period. Contrary to Blakers (1963) observations

incised decorative techniques on Chickahominy pottery is widespread

in the lower Chesapeake Bay region but is not as abundant as in the

middle and upper Chesapeake Bay region. The decrease in incising

motifs toward the south may support the hypothesis that this tech­

nique originated in the northeast coastal area and diffused south

reaching the lower Chesapeake Bay very late in the protohistoric

period (see page 114 for a discussion on Shantok Cove Incised

pottery which dates from A.D. 700 to 1100).

Similarly, the concept of corded horizontal decorations was

not readily accepted in the lower Chesapeake Bay region although

occasional sherds of this type have been reported (Smith 1971).

Parallel incised lines may have increased in popularity during the

latter part of the protohistoric period (ca. A.D. 1525 to 1600),

although more detailed analysis is required to test this hypothesis 212

(Painter 1967: 106). Excavations at the site of the historic

Wyanoke village on the lower James River has yielded European trade items and has produced an abundance of Chickahominy Incised pottery identical in many attributes to the Rappahannock Incised pottery type (Gregory and Jarrell 1976). The information from this site and the data presented above suggests that shell tempered fabric impressed or incised pottery similar to the early Townsend types continued to be manufactured in the lower Patuxent and in most of the lower Chesapeake Bay area. Cultural groups in the upper Chesa­ peake Bay region were beginning to adopt the ceramic techniques of the Townsend Corded Horizontal type but were either displaced by or assimilated into the Ferguson phase cultures. On the Eastern Shore, cultures of the Townsend complex continued in contact with the

Potomac Creek complex groups and eventually adopted the decorative techniques of the Potomac Creek complex. Concepts, such as the religious beliefs associated with ossuaries may have also been adopted (see Feest [1973] for further insights into the burial customs of Indians in the Chesapeake Bay region).

The available data suggest that the Ferguson phase expanded from the Potomac River Valley to the south and to the north, both displacing and absorbing people of the coastal area. Sites along

the fall line were occupied first but eventually control extended

over the entire Western Shore north of the Potomac except for the

lower Patuxent where refuge groups of the coastal cultures of the

Sullivan Cove phase combined to form a formidable barrier in the

expansion of the Potomac Creek complex. Control may have extended

to the upper portions of the Eastern Shore in the area of Kent 213

County (Wilke 1974: per. comm.). Pressures from the Potomac Creek complex may have forced the palisading of the historic Tockwhogh at

the head of the Bay and the alliance of that group with the Susque­ hannocks. The development of the Susquehannock Confederacy to the north around A.D. 1550 and of the Powhatan Confederacy to the south around A.D. 1570 was to substantially lessen the territory controlled

by the Potomac Creek complex. The Potomac Creek complex was probably

the earliest confederacy in the Chesapeake Bay region and was sub­

sequently weakened by development of other confederacies which may

have developed in response to European influences on native cultures.

The Potomac Creek complex may represent the only truly native con­

federacy to develop in the Chesapeake Bay region and may have been

stimulated by the introduction of religious and political concepts

from the Great Lakes and Mississippi Valley cultures.

Development and Displacement During The Patawomeke Phase

The difference between the Ferguson and Patawomeke phases are

distinct on either end of the time spectrum but the boundaries

between the two are blurred by the gradual evolution of ceramic

technology as well as by other aspects of the cultures. Most sites

producing Potomac Creek Corded Horizontal pottery also contain

Potomac Creek Cord Impressed pottery. The decorative motifs of

Potomac Creek Cord Impressed pottery represents refinement and

variation of the corded horizontal motifs of the Ferguson phase.

Elaborate and well executed decorative motifs, serrated lips, and

excellent firing techniques characterize thelater ceramics of the

Patawomeke phase.

While stockades were constructed during both phases, McNett and

Gardner (1975) suggest that ossuary burials were adopted during the 214

Patawomeke phase. If the time between ossuary burials represented only three or four years (Ubelaker 1973) and if the location of the ossuary rotated from village to village within the same clan group, then given the number of known ossuaries in the Potomac Valley (19) and the absence of ossuaries in the Piedmont, we can infer that ossuaries may have been adopted during the Patawomeke phase. Yet

the early political cohesiveness as suggested by the domination of

the coastal cultures by the Potomac Creek complex cultures indicates

that the religious concepts and practices associated with ossuary burials may have been accepted during the developmental period of

the Ferguson phase.

Thurman (1973) suggests that ossuaries in the Potomac Valley

can be chronologically ordered on the basis of associated grave

goods. The sequence begins with no grave goods followed by a period

of tubular columella shell beads, then shell beads and native copper

beads and finally European copper and glass beads mixed with shell

beads. While the type of items found in the ossuaries are important

in chronological studies as also suggested by data from the Lewis

Creek Mound (Manley 1963: 38), the absence or presence of various

types of grave goods may indicate the status of the individuals

(Feest 1973: 3/. Placement of shell beads with infants was a prac­

tice developed during the Montgomery complex and continued by parti­

cipants in the Potomac Creek complex. As the Montgomery complex

components at the Selden Island and Biggs Ford sites produced

individual burials, the question arises, where did this custom

originate?

The presence of native copper at the Selden Island site and in 215

ossuaries of the Potomac Creek complex suggests contacts with the

Great Lakes region. Native copper is found in the Blue Ridge province of Virginia and Maryland such as in the rhyolite deposits of the Catoctin formation. Evidence of copper workshops has not yet been reported from either Maryland or Virginia. The similarities between the cultures in the Great Lakes region and the Potomac Creek complex suggest that ideas as well as copper was being traded.

Copper would have been a valuable commodity to exchange for the shells from the Coastal Plain province.

Between A.D. 900 to ’ ' two related populations developed in

Southern Ontario which shared the cultural traits of palisaded villages and the beginning of the ossuary burial complex (Wright

1968: 101). The absence of archeological evidence of wide trade relationships suggests that these populations developed in relative isolation and would not have influenced the Montgomery and related cultures in the Piedmont province. Around A.D. 1300 the military conquest of some magnitude occurring in southeastern Southern Ontario and New York strongly suggests the development of a confederacy of clan villages (Wright 1968: 102) which may have paralleled the political systems of the Piscataway Confederacy as described during the Colonial period (Ferguson and Ferguson 1960: 11-12). By around

A.D. 1350 large ossuary burials and an elaborate pipe complex were adopted by the northern cultures. An elaborated pipe complex was adopted by the builders of the Lewis Creek Mound before A.D. 1300

(Manley 1963; MacCord and Valliere 1965) and may represent Missis-

sippian influences. With the appearance of beans and squash in the

north ca. A.D. 1400, a self-sufficient horticultural base was 216 realized which markedly reduced the necessity of relying upon game to supplement the corn diet (Wright 1968: 103). Whether beans and squash were introduced at this late date or earlier in the southern

Piedmont is a question still unanswered. Population increases and cannibalism appear approximately at this time and regional trade networks are established in the north.

The adoption by Coastal and Piedmont cultures in the Potomac

Valley of the concepts of ossuaries, palisaded villages, confeder­ acies (?), and ceremonial cannibalism (?) suggest the development of an early trade network between the Ferguson phase cultures and the cultures of the Great Lakes region. During the later part of the

Patawomeke phase, contact with Mississippian cultures is suggested by shell pendants with incised weeping-eye motifs. The people of

the Potomac Creek complex may have acted as middlemen between the

coastal and interior cultures. Expansion to the north along the

fall line may have been stimulated by the desire to cut off direct

trade contacts between the interior and coastal groups.

These same trade networks may have been operative when the

fishing fleets and, later, the French fur trade directly influenced

cultures in the St. Lawrence Valley. From the beginning of the

sixteenth century, if not earlier, fishing fleets from England,

France, Spain and Portugal repeatedly visited the Grand Banks off

Newfoundland in the spring and hunted whales and walrus in the Gulf

of St. Lawrence (Biggar 1965: 18-37). While the size of the early

fleets is presently unknown, by 1578 over 300 vessels were reported

fishing off of the banks of the Atlantic slope (Hakluyt 1589).

Although fishing was the primary activity of these fishing fleets,

trade with the natives developed at an early date and increased in 217

regularity through time. Some of the European trade items of cloth, knives, charms, copper beads, and glass beads may have eventually passed through the trade networks well before the establishment of the Jamestown Colony in 1607. Thurman suggests that European trade items may have appeared during the Patawomeke phase as early or earlier than A.D. 1550 (Thurman 1973: 40).

In historic times, the Nacostians on the Anacostia River con­ trolled the trade relationships between the coastal tribes and

Europeans with the northern groups in Ohio. A reverse relationship may have been in effect during the protohistoric period with these

Indians controlling the flow of European trade goods from the north to the groups in the south (Ferguson and Ferguson 1960: 24-25).

Even before European trade goods entered the trade network, the

Potomac Creek complex people may have controlled the flow of shell beads and other marine resources from the Coastal Plain to the interior. A desire to cut off direct contact between the coastal and interior cultures may have been a primary motivation for expand­ ing from the Rappahannock to the Susquehanna Rivers along the fall line zone. By located village sites at the heads of all the coastal rivers in this area, the Potomac Creek cultures, in effect, monop­ olized the distribution networks of marine shell and, later,

European trade items in the middle and upper Chesapeake Bay region.

The coastal cultures either had to trade with the Shenks Ferry or

Potomac Creek cultures or try to sneak through to the interior; a practice which may have been both dangerous for the trade parties and for the villages where reprisals may have been focused.

Alternative hypotheses concerning the expansion of the Potomac 218

Creek complex may have some validity but the trade hypothesis is the most logical. The relocation from the west of Keyser Cordmarked pottery producing Luray phase cultures beginning as early as A.D.

1350 (Bastian 1974) may have forced the Ferguson phase groups to abandon the Piedmont and move eastward; occupying areas environ­ mentally similar to the Eastern Piedmont portion of the Potomac

River. The upper portion of the Chesapeake Bay (north of the

Patapsco River) is of low salinity and similar to the upper portions of the estuarine portion of the Potomac River. Perhaps the cooling and warming trend of the period resulted in periodic famine, requir­ ing heavy dependence on other non-horticultural portions of the sub­ sistence pattern (see page 35). The sites at the headwaters of the

coastal rivers such as the Elkridge site on the Patapsco (Clark 1970) or the Obrecht site on the Severn (Wright 1973: 24) were situated

in heterogeneous areas ideally located to exploit the annual fish

runs as well as in easy walking or canoeing distance to the interior

and coastal fauna and flora resources. The Potomac Creek complex may have expanded to the north to increase the size of the hunting

territory. The larger hunting territory may have served as a buffer

against famine induced by early or late frost or droughts. The

hunting territory may have been important in the seventeenth century

as furs were traded from the south to the north in exchange for rare

European trade items or other commodities. By 1610, only three years

after the establishment of Jamestown, the Potomac Creek complex

groups were already heavily involved in the fur trade. The communal

hunt involving 200 to 300 people may have been conducted primarily

to obtain large quantities of furs for the fur trade. The passage

referring to the hunting practices of the Potomac Creek groups as 219

observed by Spelmen, who lived briefly at Patawomeke, states that the Indians were primarily concerned with the deer skins, and not with provisions:

"But in that time when they goe a Huntinge ye weomen goes to a place apoynted before, to build houses for ther husbands to lie in attt night carienge matts with them to couer ther houses with all, and as the men goes, further a huntinge the weomen (goes before) follows to make houses, always carrienge ther mattes with them ther manner of ther Huntinge is thiss (wher) they meett sum 2 or 300 togither and hauinge ther bowes and arrows and every one with a fier (sti) sticke in ther hand they besett a great thickett round about, which dunn euery one sett fier on the ranke grass (and) which ye Deare seinge fleeth from ye fier, and the meenn cumminge in by litell and litle incloseth ther game in a narrow roome, so as with ther Bowes and arrowes they kill them at ther pleasuer takinge ther skinns which is the greatest thinge they desier, and sume flesh for ther provision" (Spelman, In Arber 1910: cvi, cvii).

Possibly, the spring communal hunting technique was employed in the fall when the demand for furs increased. Similar spring-fall communal hunts apparently developed among the seventeenth century

Delaware groups (Thurman 1974: 126). That the communal hunt, which employed the fire-ringing technique, was a cultural practice of considerable antiquity is suggested by the extensive barrens in the

Eastern Piedmont at the beginning of the Historic period (see pages

24-25). The increased use of this technique as the demand for furs

increased is inferred.

The middleman trade position of the Patawomeke phase cultures may have been one reason why the Susquehannocks focused their

attacks more on the Piscataways than on the cultures of the Patuxent

or Eastern Shore areas. Expansion of the Piscataway Confederacy

must have still been in progress when the Susquehannocks began their

expansion. While the members of the Potomac Creek culture were

powerful enough to depopulate large areas of the tidewater, they 220

were not powerful enough to retain the conquered land. The pattern

of withdrawal from the less populated areas to the Confederacy's

center which was observed by the first Maryland settlers in 1632

(White 1632) must have begun late in protohistoric period ca. A.D.

1570.

With the development of the Susquehanna conquest power around

A.D. 1550 (Witthoft 1963: 64) the indigenous Shenks Ferry cultures

of the lower Susquehanna Valley were attacked with the remnants

assimilated into the Susquehanna culture (Heisey and Witmer 1964).

With the beginning of the Susquehannocks war of domination of

surrounding territories, the Piscataway Confederacy's outpost to the

north was abandoned as indefensible (Clark 1970) . As the entire

area from Herring Bay to the mouth of the Susquehanna River may have

been depopulated by the Piscataway Confederacy, the Susquehannocks

were able to extend their control further south on the Western Shore

than on the Eastern Shore where the indigenous cultures may have

escaped the destructive powers of the Piscataway Confederacy. John

Smith noted that this entire area was depopulated by 1608 (In Arber

1910: 349). The tribes in the upper part of the Eastern Shore were

consolidated in palisaded villages and allied with the Susquehannock

who were forcing the abandonment of sites along the lower Potomac

(White 1632). As the Susquehannocks consolidated approximately

A.D. 1550 and had to first subject the Shenk Ferry's culture in the

Susquehanna Valley, then their expansion south of the Susquehanna is

estimated to have begun around A.D. 1570 with the Elkridge and

Painters Mill sites abandoned before A.D. 1600.

The Susquehannocks had lived in scattered hamlets near the 221

present Pennsylvania and New York border before A.D. 1550. After that date they formed single compact communities and turned their combined military might against their weaker, unconsolidated neigh­ bors and the Five Nations.

"By A.D. 1600 they had effectively penetrated throughout New Jersey, dominated the Upper Ohio Valley at least as far as the Scioto, and had occupied the Upper Potomac in the vicinity of Romney, West Virginia" (Witthoft 1963: 65).

Witthoft contends that the Susquehannocks were driven away from the early sources of European trade goods, and towards sources of beaver and other native commodities. Hunter (1963: 73) notes that

the Iroquois were weak during their developmental period, suggesting

that "the Susquehannock were in fact drawn, not driven to the lower

Susquehanna by a desire for better and safer trade opportunities."

When Smith visited the head of the Bay in 1608, the Tockwhoghs living

on the Eastern Shore had obtained European trade goods from the

Susquehannocks (Smith, In Arber 1910). Hunter suggests that these

goods may have come from the Dutch who, previous to the establish­

ment of a trade post on the Delaware in 1626, had traded with the

local populace. If this was the case then why did not the Tockwhogh

deal direct? Witthoft's suggestion that the Susquehannock were

acting as middlemen between the northern trade centers and the

sources of native furs is logical as the Dutch influence developed

after the Susquehannocks had become a powerful force in the area.

Once the Dutch trade began, the Susquehannocks continually attacked

the Five Nations trade parties going to and from the Dutch in New

York and the Delaware. The territories subjected by the Susquehan­

nocks either provided control of the areas around the trade center

or the areas of the resources; a similar control may have been 222

exercised by the Potomac Creek complex culture before their power was effectively broken by attacks from the Susquehannock and

Powhatan groups.

While some historians attribute the depopulation of the upper portion of the Potomac and Ohio drainage to the Iroquois, Hunter

(1969: 14) suspects that epidemics, the Susquehannocks, and later raiders all ravaged the region before the Iroquois. A large site of the Keyser Cordmarked or New River ceramics style of the proto­ historic period, the Welsh Run Site has scattered Susquehannock sherds intermixed in the fill of Keyser's pits.

"This site is therefore later than 1500 A.D. and was probably a stopping point in the Susquehannock march to the Keyser-style site at Romney, West Virginia" (Witthoft 1963: 66).

Apparently the Susquehannocks were partly responsible for the depopulation of the Piedmont and Alleghany provinces of Maryland.

The effect of the Piscataway Confederacy on the Piedmont cultures is unknown although the possibility is strong that any members of the Confederacy still living in the Piedmont were displaced south by the Susquehannocks. Refuge Patawomeke phase groups may have joined Luray phase groups in the Piedmont as suggested by the

Potomac Creek Cord Impressed pottery from the Keyser Farm site

(Manson, MacCord and Griffin 1943). The Luray phase groups were subsequently forced to abandon the Piedmont probably as pressures from the Susquehannocks increased. The Eastern Piedmont province and the Western Shore of the upper and middle Chesapeake Bay region had probably been depopulated by the time of the Jamestown settle­ ment in A.D. 1607. The Susquehannocks continued to attack the

Potomac Creek complex groups until the Susquehannocks were effectively 223

defeated by the Iroquois in 1674.

In 1678, long after the Piscataway had lost their control over

the Eastern Piedmont province to the Susquehannocks, who had then only recently been destroyed as a nation by the northern tribes in

1674, the Delaware Indians laid claim to the lands in the upper parts of Baltimore and Cecil Counties (Maryland Archives, Vol. XV: 175).

Were the Delaware basing their claim on the fact that the Eastern

Piedmont and upper Chesapeake Bay region was their hunting territor­

ies before the Susquehannocks and the Potomac Creek complex cultures

forceably removed their access? Place names in Baltimore County

include: "Delaware Falls," "Delaware Bottoms," and "Delaware

Hundred" (Marye 1920: 351).

"The name (Delaware Hundred) doubtless has some con­ nection with the fact that the south Branch of the Patapsco Falls seems to have been formerly known as the 'Western or Delaware Falls' of Patapsco" (Marye 1920: 351).

Marye cites numerous other references of the Delaware place

names and notes that other references could be cited. The Delaware

villages along the Delaware may have been surrounded by large terri­

tories where each clan of a community owned a particular bounded

section of the community's territory (Weslager 1972: 39). But

Weslager also notes:

"From a careful study of the land transfer records in the Delaware Valley, I have formed the opinion that in some instances a large hunting territory was used by an entire community of Delawares, the ownership rights vested communally in the people of that village. These hunting territories were not necessarily contigu­ ous to the village where the owners resided, and in many instances they were some distance away. This is con­ sistent with the Delaware's crest of seasonally leaving their village to hunt, each group of families pursuing game on their respective territories" (Weslager 1972: 39). 224

Thurman has raised several objections to the existence of hunting territories among the seventeenth-century Delaware (1974: 123-125).

Thurman indicates that family hunting territories may not have developed among the Delaware and the belief that such systems existed may simply represent misinterpretation of the historical record. But Thurman does note the practice of communal hunting drives during the spring and winter and cites examples of fire-ring­

ing hunts near the mouth of the Delaware Bay in December of 1631

(Thurman 1974: 124). The fire-ring hunts were probably responsible

for the formation of the large barren areas in the Eastern Piedmont

province as previously discussed.

The Susquehannock trail ran along the fall line portion of the

Eastern Piedmont province from the Nanticoke path along the Susque­

hanna River to the Potomac River. The trail crossed the Gunpowder

and Western Run and may have followed the Old Court Road which

crosses the Northwest Transportation Corridor (Marye 1920: 114, 117).

A local historian whose information is not always reliable reports

that the Susquehanna path came down through the Caves Valley

(Figure 6) into the Green Spring Valley, past Chattolunee, and

Stevenson, and on down to the Bay (Davidson 1967: 10). Another path

in the area of the Northwest Transportation Corridor was a trail

followed by black bears to their wallowing spot below the Hannah

More Academy at the Bear Wallowing Pond Branch. This path was

eventually incorporated into the Reisterstown Road but may have been

used by Indians who used animal paths for trails (Huttenhauer and

Helwig 1967: 57). If this is true, then the Indian path passed

within a mile at the Painters Mill site. The Susquehanna path was 225

used both by war parties and families going to their hunting territories (Marye 1920); a function which may represent a continuum since at least A.D. 900 and perhaps earlier.

The "Susquehannock" path may have been established by the

Delaware for trade and for travel to their hunting territory. The

Potomac Creek complex groups may have used the path for trade, travelling to hunting territories and attacking the Sullivan Cove phase (Delaware?) groups in the upper Chesapeake Bay region. The path may have served a similar function for the Susquehannocks.

Pressures from the Susquehannocks resulted in the abandonment of the upper Chesapeake Bay region by A.D. 1608 (the time of John Smith's travels). By the beginning of the Historic period, the Patawomeke phase people were reduced f.o groups along the Potomac and Rappahan­ nock Rivers and were abandoning the lower portions of the Potomac

River (White 1632).

This analysis has focused on the cultural historical processes operative in the region during the prehistoric and protohistoric

periods and will not discuss the historical period as this subject

has been treated elsewhere (Ferguson and Ferguson 1960; Handsman

and Hunter 1972; McMillan 1972; Smith 1971). The Potomac Creek

complex groups represent displaced Piedmont cultures with possible

assimilated remnants of coastal cultures mixed throughout the complex.

The historic groups along the Potomac also represent displaced

cultures from the areas to the north. The population density along

both the Patuxent and Potomac rivers in A.D. 1608 probably reflects

concentrations of people who removed to refuge areas for protection.

The development of the Powhatan Confederacy around A.D. 1570 further 226

reduced the power of the Potomac Creek complex cultures so that the extent and strength of the historic Piscataway Confederacy may have been but a small reflection of the power formerly held by these people during the prehistoric and protohistoric periods.

This analysis has considered only a portion of the available historical and archeological data base presently available to the researcher interested in more refined analysis. Many private collec­ tions, university collections, and collections in the National Museum contain information essential for the testing and reformulation of the various hypotheses advanced. Hypothesis formulation and testing is an ongoing process which requires constant modification or expan­ sion of the hypotheses as additional data become available. For example, Norman Brice of Annapolis, Maryland has excavated a site on the Severn River which produced pottery identical in all attri­ butes to Potomac Creek Cord Impressed pottery except that the vessels are tempered with shell instead of with quartz (Brice 1976: per. comm.). The pottery from this component should logically be defined as Townsend Cord Impressed pottery but such a type definition will not be developed until a larger sample from several components becomes available. The evolution of this potential pottery type from the Townsend Corded Horizontal type is highly probable and may represent continuing contacts between groups of the Patawomeke and

the Sullivan Cove phases. As future conservation archeology research increases the number of Potomac Creek and Townsend complex sites available for study, the Townsend Cord Impressed pottery and the

explanation of the parallel development between the two cultures may be further elucidated. CHAPTER 7: SUMMARY AND CONCLUSIONS

This thesis has provided the reader with a diversity of data; from detailed descriptions of individual artifacts to a synthesis of the Potomac Creek and related complexes. The underlying reason for examining the various topics is to demonstrate that contract or conservation archeology does not have to be purely mundane and des­ criptive and can be provocative and explanatory. As more of our limited professional personnel divert their research interests to partake in conservation archeology, and as more students become

involved in this field during their graduate training, the greater will be the necessity to incorporate theoretical and methodological

advances to insure that archeological research contributes to the

study of man. Archeologists, not the funding agency, are responsible

for the formulation of regional research designs and for the imple­

mentation of conservation archeology programs which contribute to

the discipline while at the same time providing the required informa­

tion for the funding agency.

Only by developing regional research designs can archeologists

adequately evaluate the significance and the meaning of the archeo­

logical resources in the area of conservation concern. The prehis­

toric archeological resources discovered within the Northwest Trans­

portation Corridor cannot be interpreted unless they are compared to

the ecological and archeological resources of the Gwynns Falls Valley,

of the Eastern Piedmont Province and of the surrounding Chesapeake

Bay region. As only a small percentage of the land surface of the

Gwynns Falls Valley was exposed by cultivation, the relationships

between archeological, geological, and ecological variables were

examined to determine if direct correlations of predictive value

227 228 were discernable (Chapter 3)• The investigations revealed a direct correlation between site density and geologic units which was valu­ able during the formulation of mitigation measures for the salvage phase of investigations.

The location of prehistoric and historic sites was the primary objective of the initial phase of investigations in 1973 (Chapter 3).

Once the nature of the archeological record was determined, subse­ quent research (preliminary site examination phase) was oriented toward the determination of the extent, the nature, and the signifi­ cance of the archeological sites discovered. As John McGrain's historical analysis adequately discusses the historical archeological resources of the Corridor and surrounding Gwynns Falls Valley, the

1974 and 1975 fieldwork focused on the prehistoric resources of the

Corridor and surrounding region. Only by considering the archeo­ logical resources of the Valley could the sites within the Corridor be appraised in regard to their significance. Had surveys been limited strictly to the right-of-way, predictive statements about site density and elucidation of the cultural processes would have been inadequate. This information was essential in determining the effect of the various subsequent alternate routes developed after

the initial three routes were surveyed in 1973.

As the exact provenience method of controlled surface collect­

ing has never been reported from the Middle Atlantic States region,

this method was employed at two sites (18 BA 106 and 112) to deter­ mine if such studies were useful in delimited activity areas, in

predicting the location of subsurface features, and in determining

if post-depositional processes have distorted the distribution of

the cultural items on the sites. Past control surface collections 229 were conducted primarily to delimit activity areas and based on this data, to predict the location of subsurface features (Binford et al.

1970).

While controlled surface collections are of value in such studies, the primary reason for conducting controlled surface collec­ tions should always be to obtain a representative sample of the total population of artifacts in the plowzone or on the undisturbed surface of the site. The contextual relationships of the artifacts in the plowzone may be all we have to understand the cultural processes operative at a majority of our Archaic period sites in the east.

Even on large Late Woodland period village sites where we know a multitude of subsurface features exist, controlled surface collec­

tions should be conducted to obtain a sample of the artifacts from

the areas between and within the features. Controlled surface col­

lections can obtain comparative data which would otherwise require

complete screening of the entire plowzone of a site; a procedure

which is not always feasible or practical.

The controlled surface collections of the Painters Mill and

Gwynns Falls sites revealed a complex accumulation of debris repre­

senting a variety of activities, in a number of different loci, over

a considerable period of time. The majority of the debitage items

could not be assigned to a specific time period, let alone a specific

tradition, complex or phase. The primary attribute useful for the

placement of debitage items into a chronological position was the

lithic material from which the items were manufactured. By comparing

the lithic material of the debitage to the lithic material of chrono­

logical diagnostic artifacts and comparing the clustering of artifacts

at the controlled collected sites, placement of the debitage into a 230 chronological framework was sometimes possible. While the clustering of certain assemblages from the two controlled collected sites was observed, a much larger sample of sites collected under similar methods is necessary before the benefits of controlled surface collections are fully realized.

The distribution of artifacts on the surface of the Painters

Mill site was useful in defining activity areas for the various phases of the Late Woodland period. Opaque quartz artifacts and

Potomac Creek complex pottery cluster together. If similar co­ occurences are reported from other Potomac Creek complex hunting camps, then we can begin to discuss intrasite activity areas for this complex; to do so on the basis of one site is dangerous. The distribution of sherds from the same vessel was particularly valuable in determining the post-depositional processes active at the site.

A 30-meter downslope and a 20-meter across slope displacement of artifacts has apparently occurred at the Painters Mill site (Figure

30). This distortion is particularly pronounced in the northwestern section of the site where two gulleys transect the site and may have contributed substantially to the movement of artifacts. Soil creep was observed at the site during the excavations in 1975 as the base­ line was displaced 8 centimeters downslope during a rainy, two-week period. Sheet wash and cultivation has probably also contributed to the distortion of the original context of the artifacts at the site.

As the Painters Mill site is located at the base of the steep slope of the Setters Quartzite formation on a gently sloping colluvial fan,

this distortion was expected. Distortion of the original context of artifacts at the flat Gwynns Falls site was due primarily to erosion 231

and transport with sheet wash and cultivation of less severity,

since cultivation results in equal displacement of artifacts; main­

taining the approximate relationships of the artifacts to each other.

Slope wash, soil creep and erosion may occur in different areas at

different rates and results in unpredictable distortions of the

archeological record. Without the exact provenience method of con­

trolled surface collecting, these distortions may not be discernable

and the delimitation of activity area of decreased validity.

The techniques developed to implement the exact provenience

method are enumerated in detail. The benefit and the necessity of

conducting more than one surface collection was revealed by the

results of the three surface collections from the Painters Mill site.

The 1973 collection was hindered by heavy ground cover and by an

absence of controlled sampling. The 1974 controlled surface collec­

tion from the Painters Mill site clarified the extent of the site

but was not adequate for interpreting the chronological complexity

of the site. The final controlled surface collection in 1975 was

conducted under ideal collecting conditions, with adequate staff

and with proper safeguards established to minimize collector bias.

This data enabled refinement of the spatial and temporal analysis.

The problems encountered during the three phases of investigations

are enumerated upon for the benefit of the reader who may wish to

utilize this relatively new method. While the primary purpose for

implementing this method was to obtain unbiased settlement-subsist-

ence data, another consideration was to demonstrate that conservation

archeology programs can be utilized to develop and refine new tech­

niques and methods in archeology. 232

The surface survey of available cultivated fields in the Gwynns

Falls Valley also revealed varying site densities which correlates closely with the underlying geologic units. The Cockeysville Marble formation contains a high density of prehistoric archeological resources while the adjacent Setters Quartzite formation contains a medium density of resources. The remaining geologic units in the

Eastern Piedmont section of the Gwynns Falls Valley contain a low density of prehistoric archeological resources. These resources from the Gwynns Falls Valley are of low density in comparison to the archeological resources in the major river valleys and in the Coastal

Plain province.

The direct correlation between resource density and geologic units may exist because of the variation in vegetation units associ­ ated with the different geologic units (Chapters 2 and 3). While the composition and the extent of the vegetation units changed over time in response to culturally and environmentally induced changes, the mosaic effect of vegetation resulting from different moisture and nutrient availability of the geologic units would have existed throughout prehistory. Areas (such as that underlain by the Baltimore

Gneiss geologic complex) which currently support a diversity of vege­ tation units within close proximity would have provided a similar diversified subsistence base in prehistory. The higher density of sites within the Baltimore Gneiss geologic complex is probably related to the diversity of resources which would have been of greater attrac­ tion to prehistoric peoples. By studying present vegetation zonation on the micro-level instead of on the macro-level, the correlation between prehistoric site density and environmental variables can be 233

clarified and predictive statements developed which may be valid for areas with similar environmental conditions.

As the Jones Falls, Green Spring and Worthington Valleys are situated in identical ecological-geological settings, similar archeo­ logical resource density to that found in the Gwynns Falls Valley is predicted. Indications of hunting and resource processing stations are predicted to exist around the upland stream heads. A high density of sites on the colluvial fans and the first terraces of valleys underlain by Cockeysville Marble is predicted as is a medium density of sites along ridges of Setters Quartzite. A lower frequency of

Paleo-Indian and Early Archaic sites is predicted when the crypto­ crystalline and stunted vegetation resources of the serpentine barrens of Soldiers Delight are not nearby as in the Red Run, Horsehead Branch and Gwynns Falls Valleys. Late Woodland period sites may be present as this entire area may have been a barren as a result of the communal hunting practices of the Indians of the Late Woodland period (Chapter

2). These predictions should be of less validity for minor river systems in the Eastern Piedmont province which are not underlain by

the Baltimore Gneiss geologic complex.

Future conservation archeology projects in the minor river sys­

tems of the Eastern Piedmont province will enable testing of these

hypotheses. Refined paleoecological interpretations will be necessary

to test the inferences concerning changes in local habitats during

the prehistoric period. Only by understanding the local habitat can

archeologists begin to interpret localized adaptation, particularly

since faunal and floral remains will be rare in the highly-acidic

soils of the Eastern Piedmont province. Until such studies are 234

conducted, interpretation of the past must depend on extrapolation from present habitats, regional pollen sequences and the lithic, ceramic, and feature remains of extinct cultural systems.

The artifact analysis was conducted to provide answers for specific questions, to provide a basic data base for the interested reader and to determine if new methodological advances could be in­ corporated into the time framework of conservation archeology pro­ jects. Replication experiments, functional analysis, lithic analysis and soil analysis were conducted (Chapter 4). The contamination of the soil by modern fertilizer and the low density of occupation at the Painters Mill and Gwynns Falls sites frustrated the attempt to delimit areas of human activity by studying the chemical elements in

the soils of the sites. Time prevented detailed replication experi­ ments but the limited studies conducted did familiarize the author with the working quality of the various lithic resources utilized by

the Indians in the Gwynns Falls Valley. Detailed technological and

functional replication experiments on rhyolite, quartz, and quartzite

are needed to fill the vacuum of knowledge about wear patterns and

fracturing properties of these commonly utilized lithic materials.

The functional analysis was conducted without benefit of such detailed

studies but was useful in assigning items to functional classes. Much

of the debitage may have been utilized but the nature of the quartz,

quartzite and rhyolite materials frequently frustrates recognition

of or development of wear patterns. The detailed debitage analysis

and resultant attribute list (Appendix A) were of limited utility as

the majority of the debitage items could not be assigned to chrono­

logical positions. Such detailed analysis may be justified for 235

stratified, in situ, material but reaches a point of diminishing returns when applied to surface collections. Greater attention should have been given to the striking platform and ventral surface attributes of the flakes which would have been of assistance in assigning functions to certain activity areas. The attribute list developed was thought to include the basic data which might be of concern to the interested reader. But such detailed analysis resulted in delays in completion of the various preliminary reports and did not contribute substantially to our knowledge of the cultural pro­ cesses of the region. While the functional analysis was of value and well worth the cost, the detailed debitage analysis was too time- consuming and of limited value. Such analysis is probably beyond the requirements of most conservation archeology programs and should be limited only to excavated in situ material.

The summary definitions of lithic material which incorporated the results of spectographic analysis was essential to the discussion; many of the lithic materials have not been defined. Identification of lithic materials by archeologists is still largely an art passed on from the experienced to the inexperienced with few systematic programs of identification offered. As a result, two archeologists rarely ever agree on the identifications on all the lithic materials from a region. Only by defining the chemical composition and the physical attributes of the lithic types or varieties can archeologists hope to control the comparisons of different lithic materials which are subsumed under the same labels. The lithic definitions were essential in interpreting the controlled surface collections from

the Gwynns Falls and Painters Mill sites, and should be of value when 236

other archeologists conduct conservation archeology investigations in the Piedmont and Coastal Plain provinces.

Diachronic Synthesis

The absence of in situ artifacts from the Gwynns Falls Valley necessitated the comparison of chronologically diagnostic artifacts to previously established sequences in other sections of the Carolina

Biotic province in order to establish a chronological sequence for the study area. While such "look-alike" comparisons of surface collections to excavated material should be applied with caution

(Kinsey 1971: 3) this was the only feasible method of analysis given the limited sample available for study. While point styles during the Paleo-Indian and Early Archaic periods are pan-eastern in distri­ bution, point styles beginning with the Middle Archaic period became increasingly diversified and may reflect decreasing interaction between members of neighboring groups (Michlovic 1976: 15). In the

Chesapeake Bay region interaction spheres developed from 300 B.C. to

A.D. 250 between Adena and Coastal Plain groups, from A.D. 100 to

A.D. 600 between cultures in the Middle Atlantic States region, and

from A.D. 900 to A.D. 1600 between cultures in the Coastal Plain with

Piedmont and Appalachian Plateau cultures. The trade networks opera­

tive during these periods may have been responsible for the shared

attributes of many of the cultures of the region. But during the

Middle and Late Archaic periods, interaction may have been a more

localized phenomenon.

The apparent absence of Middle Archaic projectile point styles

similar to those of the Coe's (1964) sequence to the south and Kinsey's

(1972) sequence to the northeast may represent sampling error, the 237

absence of occupation during this period, or the failure to recognize undefined local point types for this period. The latter hypothesis may be the correct one and can be tested by obtaining a larger sample size and by excavating stratified sites which may exist in the Gwynns

Falls Valley.

Throughout the various periods of occupation, the Gwynns Falls

Valley apparently served primarily as a hunting area. The sites pro­ ducing the greatest concentration of artifacts can be interpreted as environmentally favorable areas which were repeatedly selected for by the various cultural groups visiting the area. The Gwynns Falls,

Painters Mill, Lee and Red Run sites all represent the accumulation of numerous small hunting or flora processing camps' debris over a considerable period of time by a diversity of cultures. The smaller size sites such as 18 BA 121 and 18 BA 125B may reflect the actual size of single component sites (20 to 30 meters). Exploitation of the upland resources is suggested by the finding of an occasional projectile point between stream valleys and the presence of small processing camps around stream heads. These small processing camps usually consist of a scattering of debitage and an occasional pro­ jectile point, scraper or knife. Quarry related activities may have occurred in the serpentine barrens of Soldiers Delight where serpentine and Soldiers Delight chalcedony may have been obtained. Local vein quartz was utilized at all the sites in the floodplains. Variation in the percentage of lithic material types at each site may be of chronological significance although the high percentage of rhyolite artifacts at the Hazard site may simply reflect the closer proximity of that site to the sources of rhyolite. Such a pattern was not 238

discernable for the sites falling within the Cockeysville Marble formation.

The diachronic synthesis represents a first attempt at ordering the study collections from the Gwynns Falls Valley. Many sites failed to produce chronologically diagnostic artifacts. Previous collecting of these sites by avocational archeologists may have depleted the bifaces and may also account for the high percentage of debitage items at the sites. Archeologists should always assume that their

"newly discovered site" has already been heavily collected unless proven otherwise. Multiple collections from the various sites would increase the sample size and probably increase the number of compo­ nents represented at the sites. The greater the number of collections from a site, the greater the chronological complexity the site will tend to exhibit.

To assist the environmental analyst of the Maryland Highway

Administration, a summary of each site was included in the final report in addition to the diachronic synthesis which is elaborated upon in Chapter 5 (Clark 1975c).

This summary was incorporated into the area survey section of

Chapter 3. The summary of the sites should also be of value to the archeologists concerned with research problems other than those which

this author chose to focus upon. Site summaries are greatly appre­

ciated and often demanded by the environmental planners who have to

incorporate the data from the archeological report into the Environ­ mental Impact Statement. But site summaries should not be the only

data provided by the archeologist. The sites cannot be properly

evaluated unless they are placed into a regional framework. 239

Culture History Synthesis

A cultural historical analysis of cultural development in the

Coastal Plain and Piedmont province of the Chesapeake Bay region during the Late Woodland period has been presented to demonstrate the complexities of the archeological record and to assist in the place­ ment of the assemblage from the Painters Mill site into a regional perspective. The data from the Painters Mill site in the Northwest

Transportation Corridor was useful for only a small portion of the discussion but this is the value of developing a research design.

As archeologists conduct conservation archeology projects in the region, the data from Late Woodland components will be useful in refining the cultural historical hypotheses advanced in this chapter.

A cultural historical approach was developed because only by first understanding diachronic problems can archeologists begin to conduct processual analyses. If the sole purpose of this thesis was to discuss the Late Woodland period, then analysis of the cultural pro­ cesses of the Potomac Creek complex would have been developed. But the purpose of this thesis is to demonstrate that conservation arche­ ology can contribute to the advancement of the study of man in numerous different ways.

Placement of the cultural processes recorded in the archeological resources of the project area into the larger perspective of a region has resulted in the elucidation of problems which were not even known, let alone explained.

The culture history discussion of Chapter 6 and the diachronic discussion of Chapter 5 represented primarily hypothesis formulation

stimulated by questions raised specifically by data gathered during 240

the three years of the Northwest Transportation Corridor archeolog­ ical survey project. The research designs were generated because the data collected in 1973 and 1974 required those particular types of research designs to provide specific answers. The 1975 investi­ gation refined the earlier research designs, particularly concerning the development of the Potomac Creek complex. Many facets of the hypotheses advanced concerning the Late Woodland period could have been expanded upon but discussion was limited primarily to cultural historical problems.

The development of the Potomac Creek complex was traced from the formative processes in the Piedmont province, to the expansion in the

Coastal Plain and to the eventual contraction of the complex in re­ sponse to development of the Susquehannock and Powhatan confederacies.

All of the hypotheses represent, inferences with varying degrees of supporting data. The inferences need to be tested, refined, and modi­ fied or abandoned as additional data become available. However, many of the inferences generated on the basis of the data from the Elkridge,

Painters Mill and Accokeek Creek sites have proven valid as other

sites have been studied.

The cultural historical discussion has been presented to stimu­

late thought, to evoke critism, and to provide a foundation for

future research. Just as John Witthoft's (1963) original hypotheses

concerning the development of the "Potomac Creek Conquest State"

greatly facilitated the development of this synthesis, may this

synthesis inspire more refined analysis.

It was during the process of disproving many of Witthoft's

inferences that the present culture history synthesis was formulated.

May the inferences advanced in Chapter 6 be but a starting point for 241

explanation of the cultural processes operative during the Late

Woodland period in the Chesapeake Bay region.

Mitigation Philosophy

The discipline of archeology has undergone substantial trans­ formations within the past two decades and will probably continue at an accelerated rate to refine and reform the methodological and theo­ retical base of the present. The primary objective of any conserva­ tion program should be to insure that a representative portion of the archeological record is saved for future researchers who will be asking new questions and employing new techniques to obtain answers.

What constitutes a representative portion of the resource base for any one area can only be determined by considering the archeological resources within the larger perspective of a region. Once the signi­ ficance of the archeological resources has been assessed, the arche- ologist should first strive to minimize the effect of project devel­ opment on the resource base.

By providing input during the early planning stages of a project, significant archeological sites may be saved from destruction by alteration of the proposed improvements. The Painters Mill site was the most significant archeological site found within the Gwynns Falls

Valley because of the time depth of the assemblages and the presence of several Late Woodland components. Except for the "do nothing" and

the "widen Reisterstown Road" alternatives, all sixteen proposed alternate routes result in the complete or partial destruction of the

site. Even if the proposed transportation improvements were shifted

to conserve the site, unless the site was purchased and set aside as

a park, the site would surely be destroyed by commercial development. 242

While the sites along Red Run may be saved from destruction by shift­ ing the access roads in this area, the Gwynns Falls and Painters Mill sites could not be saved from indirect destruction because of the value of this commercial property once the transportation facilities were constructed. Even if the Northwest Expressway is not built,

this area will probably be destroyed by expansion of the existing

industrial park in the area.

If sites are saved from direct impact, archeologists should

strive to purchase or otherwise set aside archeological resources.

Zoning changes may be one means of preventing development. If only

a portion of the site is affected by project development, perhaps the

remaining portion can be protected by expanding the right-of-way to

include the site. But obtaining funds to purchase the site may re­

quire years of involved negotiations which few archeologists have the

resources or time to pursue. When faced with the inability to insure

the long-term preservation of the resources, it may be better to

salvage the site when funds are available than to save the site from

immediate destruction only to have the site destroyed by private

development. Conservation of the resources by excavation should be

undertaken only after all other efforts of conservation have failed

because, regardless of the rigor applied in recovering the resources,

no excavations can recover all the data available at any one site.

While archeologists cannot foresee the needs of archeologists

during the next century they should strive to meet the needs of the

present which means utilization of the latest methodological and

theoretical tools available. Schiffer's (1974) philosophy has been

adopted by this author and the investigations oriented toward ful­

filling the minimal conditions quoted below. 243

"When archeological investigations are conducted they must be in full accord with modern standards of archeological research. This entails the following minimal conditions:

a. Multistage research throughout project duration.

b. Explicit research designs.

c. Rigorous sampling designs.

d. Undertaken with multidisciplinary collaborators.

e. Experimentation, simulation, and other activities as needed to solve subsidiary problems in method, technique, and theory.

f. Rapid publication of results to a wide audience."

(Schiffer 1974).

Conclusions

The three years of investigations of the Northwest Transporta­ tion Corridor has attempted to meet the conditions quoted above; have striven to go beyond the purely descriptive text of the past; have pursued many avenues of thought and employed new techniques currently being tested in the discipline. This study provides but a foundation upon which other studies can expand. The results are applicable to areas well outside the boundaries of the Northwest Transportation

Corridor or the Gwynns Falls Valley. Future conservation archeolo­ gists investigating projects in the Eastern Piedmont province or in the area covered by the culture history synthesis of the Late Woodland period can contribute to the refinement and reformulation of the research designs developed in this thesis.

Following the final phase of investigation when the archeolog­ ical resources of the Gwynns Falls and Painters Mill sites will be studied for the last time, the Northwest Transportation Corridor transportation facilities will be constructed and the process of 244

development along this new transportation route began. Before the turn of the century, the Gwynns Falls Valley will, except for a thin strip of green open space, be a highly urbanized area with the Gwynns

Falls River little more than an open drain. Most of the archeological resources of the Valley will be destroyed by the year 2000. But by employing modern standards of archeological research, by developing regional research designs and studying the resources from the entire

Gwynns Falls Valley, the Gwynns Falls Valley will not be a blank in the archeological record. The final phases of investigation will result in refined hypotheses and although the sites may be destroyed, the history and knowledge they contain about the first inhabitants of this land will not be totally lost and may serve someday to pro­ vide vital data for the elucidation of the total subsistence-settle- ment pattern for various prehistoric cultures which formerly lived along the pleasant shores of the Chesapeake Bay. 245

APPENDIX A: LITHIC ATTRIBUTE DEFINITIONS AND DATA

Flake angle and flake measurement attributes are defined in

Figure 42. Projectile point measurements and stylistic attributes are defined in Figures 43 and 44. Functional attributes are defined in Figure 45. Tabulation of attribute data for all specimens col­ lected from 18 BA 106, 112, 122 and random finds during the 1973 and

1974 field seasons are provided in Tables 2 to 12. The total assem­ blages from 18 BA 106 and 18 BA 112 are tabulated in Tables 13 and

14. These data provide a representative sample of the artifacts from the Gwynns Falls and Painters Mill sites which may be of use to future researchers concerned with comparative studies of controlled surface-collected materials.

Lithic Attribute Data

Formal definition of the attributes will either be provided below or the reader will be referred to the original sources which have previously developed methodologically acceptable terminologies.

The following definitions have been extracted from Hume (1976) with only minor modifications influenced by the works of Crabtree

(1972) and Wilmsen (1970). The reader is referred to Hume (1976) for elaboration on the definitions and for definition of the dimen­ sional attributes utilized for functional edges. The projectile point terminology is based on Bebrich's (1971) definitions.

Illustrations of many of the functional artifacts discussed are provided at the end of Chapter 5 (Figures 31 to 39; pages 160 to 177). 246

I. Debitage - This is defined as residual lithic material resulting

from tool manufacture or use which was not subsequently utilized

or modified for use.

1. core - A natural nodule from which one or more flakes have

been removed. Flake cores are not recognized for the

purposes of this study (flakes struck from flakes).

2. core fragment (core f) - A piece of raw material which has

one or more negative flake scars that cannot be classified

as a flake fragment.

3. chips and chunks - Chips are lithic materials which lack

diagnostic characteristics of a flake or core. Chunks are

the same as chips only larger (thick). Both represent

shattered material although they are not always clearly

distinguishable from flake fragments.

4. flake - Any piece of stone removed from a larger mass by

application of force which results in the formation of a

striking platform, bulb of percussion, and complete distal

end.

5. flake fragment - Three types are possible. A bulbar flake

fragment (bulbar ff) retains the striking platform and the

bulb of percussion but lacks the distal segment. A medial

flake fragment (medial ff) may retain part of the bulb of

percussion as well as stress lines but lacks the distal seg­

ment and striking platform. A distal flake fragment (distal

ff) lacks a striking platform and bulb of percussion. A

bipolar flake fragment, which is split down the axis of per­

cussion, is placed in the bulbar flake fragment category. 247

II. Utilized - Debitage which is modified either intentionally for

use or unintentionally by use is given special consideration

(Table 2 to 6 and Appendix B).

1. position - Area of utilization in relation to bulb of per­

cussion and striking platform of flake. The area opposite

the striking platform is defined as the distal edge with the

edges defined as sides.

2. edge angle - Intersection of two faces in the area retouched

or utilized. In case of retouch, the angle is formed between

the retouch scars and the face from which they were struck.

A lens stand coordinate was used to measure the angles.

3. direction - Denotes whether the flakes are unifacial or

bifacial or unidirectional or multidirectional.

4. edge shape - The outline of the edge. The plane is the

outline of the edge when the object is seen in plane view.

The edge profile is the dorsal-ventral variation of the edge.

The profile can be straight (st), concave (cc), convex (cv),

or irregular (irrg) all of which are determined by laying

the edge against a flat surface.

5. edge length - Measured along a straight line between the two

maximum points of the functional area. 248

Figure 42: Flake angle and dimensional attributes definitions.

flake angle bulb of . striking proximal' percussion 6 p la tfo r m 1

m e d ia l

distal edge distal /

A-platforn width, B-platform depth C-flake length, D-flake width «Catalog No. 18 BA \Ob-ff 1<4.

Figure 43: Projectile point dimensional attribute definitions.

A-tlade length B-haft length C-distal-distal width D-nedial-medial width A blade E-proximal-proxlral width

proximal zone of Juncture medial zone of juncture distal zone of juncture ) haft

« From Bebrich (1967) 249

Table 2: 18 BA 112 Debitage data from 1974 controlled surface collection.

C a ta lo g i 'î a t e r ia l Subdivision U t i l ­ Pin t f o r n lAngie F la k e ( n n j H in g e d M un b e r iz e d Widlhl l«Dtn 1 L e n g th I V id 'tn l 2 q r t z - v ch u n k 2 3 15

Table 2: continued

C a to lo g l U t e r l a l Subdivision U t l l - Platform A n g lo Flak?!(mm) Kinged ■riinber ii-.-d 'JidthI Donth ierurthl Width

5 8 R h y l f la k e 9 2 6 0 1 5 1 3 5 9 rhyl bulbar ff 7 2 2 5 7 11 X 6 0 rhyl distal ff 1 0 1 0 6 2 r h y l c h ip 1 5 1 5 6 3 r h y l f la k e 11 4 3 5 1 6 2 1 64 rhyl flake 5 1 30 2 4 2 7 6 5 f l n t f la k e 11 2 11 1 7 6 6 f l n t d i s t a l f f 1 5 1 1 6 9 r h y l f la k e 5 1 1 7 1 7 ? 0 r h y l c h ip 2 5 1 8 7 7 1 r h y l c h ip 1 8 1 0 7 2 q r t z - v c h ip 1 3 9 7 3 r h y l f la k e 8 2 8 8 2 5 2 4 7 4 r h y l d i s t a l f f 1 2 1 5 7 5 q r t z - c c h ip 1 7 1 6 7 6 r h y l f la k e • 4 5 1 2 1 0 7 7 r h y l - f d i s t a l f f 1 5 1 1 7 8 r h y l f la k e \ 8 2 2 0 2 8 7 9 r h y l f la k e c ru s h e d 19 2 9 X

Table 3: 18 BA 112 debitage data from 1973-1974 random survey.

C a to lo g I - I a t e r ia l Subdivision U til­ P la tfo r m A n g le . F la k e 1 an) H lngec Num ber iz e d W idth lûepth L e n g th 1 W id th

8 0 * r h y l f la k e 1 0 4 1 8 31 4 8 6 0 r h y l c h ip 2 4 1 3 8 0 r h y i - f c h ip 17. 1 3 80 rhyl-f flake 7 4 35 1 5 1 9 80 rhyl flake 3 1 2 0 1 7 8 0 r h y l c h ip 9 8 8 0 r h y l b u lb a r f f 7 1 9 1 0 8 0 r h y l c h ip 8 8 8 0 ja s p e r f la k e * 4 2 1 0 9 8 0 q r t z - v c h ip 1 4 1 4 8 0 q r t z - v c h ip 2 0 1 7 8 0 qrtz-v chunk 2 0 1 8 8 2 * * r h y l f la k e 4 1 8 8 2 5 2 4 8 3 r h y l f la k e 8 3 1 5 1 5 • 8 4 r h y l c h ip 2 7 1 4 8 5 r h y l c h ip 12 1 0 8 6 r h y l c h ip 1 9 1 2 8 7 r l q r l f l a k s 9 2 3 5 2 6 3 2 8 8 r h y l - f d i s t a l f f 11 1 4 8 9 f l n t c h ip 1 0 7 9 0 r h y l f la k e 1 3 2 7 5 1 5 2 4

* 1973 test excavations ** 1973 surface collection 251

Table 3: continued

Catalog Material Subdivision Util­ Platform I Angle Flake (r.m) Hinged Number ized Width]Depth] lengthj Width

91 R h y l f la k e y 2 1 0 15 92 r h y l f la k e X 7 3 60 2 0 39 9 3 r h y l b u lb a r f f 5 2 .5 8 9 9 5 r l i y l f la k e 1 3 4 5 7 2 6 2 d 9 6 r h y l f la k e 4 1 9 10 9 7 r h y l f la k e 3 1 13 12 96 qrtz-c flake 7 2 .5 67 2 7 21 9 9 q r t z - c c h ip 14 14 100 rhyl flake 1 3 4 4 5 2 3 39 102 f l n t f la k e V 4 1 1 4 10 103 r h y l f la k e 4 1 1 4 2 0 1 03 r h y l f la k e 5 2 1 7 19 1 * q r t z - v c h u n k 4 0 2 3 J a s p e r f la k e 1 0 1 3 5 42 r h y l f la k e 5 1 1 6 17 r h y l b u lb a r f f c ru s h e d 1 6 18 r h y l f la k e 1 3 3 .5 2 7 33 r h y l f la k e 5 2 4 2 25 2 6 r h y l . c h u n k . 2 7 21 r h y l f la k e 6 5 7 4 1 8 2 9 r h y l - f f la k e 8 . 2 4 0 1 9 19 r h y l d i s t a l f f 1 7 15 r h y l f la k e 6 1 .5 2 4 27 r h y l d i s t a l f f 1 8 11 r h y l f la k e 1 2 3 34 2 5 2 0 r h y l b u lb a r f f 8 2 8 5 15 14 rhyl distal ff 14 15 r h y l c h ip 2 4 16 r h y l - f bulbar ff 6 1 15 13 r h y l - f c h ip 2 6 14 r h y l f la k e 4 1 7 0 17 18 r h y l fl.a k e 7 2 .5 7 0 15 14 r h y l c h ip 1 4 11 r h y l f la k e 1 5 12 13 r h y l c h ip 1 4 9 r h y l f la k e8 2 9 15 r h y l b u lb a r f f 5 1 1 3 14 r h y l f la k e 3 1 6 5 1 0 15 r h y l f la k e 5 1 8 8 r h y l c h ip 1 3 13 r h y l c h ip 1 4 11 • r h y l c h ip 1 5 10 r h y l f la k e 1 5 7 9 r h y l c h ip 1 0 10 q r t z f la k e 3 1 9 7

* 1974 random surface collection made by Tyler Bastian. 252

Table 4: 18 BA 106 Debitage data from 1974 controlled surface collection.

C a ta lo g Material Subdivision Util­ Plat f o m Angle Flak' (n.m) H.in.;ed Humber ized WidthlDeath Length!Width

1 rhyl flake missing 2 r i i y l f la k e X 23 22 4 r h y l c h ip 16 12 5 r h y l - b f la k e 4 2 20 9 12 X 6 qrtz-c flake 13 4 57 21 21 9 r h y l f la k e 8 1 68 14 13 10 J a s p e r f la k e 10 3 60 29 30 11 r h y l - f m e d ia l f f 11 23 12 r h y l - f f la k e 9 3 60 22 29 13 r h y l ^ c h ip 15 11 16 r h y l - f c h ip \ 16 8 18 r h y l - f m e d ia l f f 27 16 19 q r t z - v f la k e 10 4 73 16 17 X 20 r h y l - b c h ip 10 8 22 qrtz-v flake 17 10 79 39 25 25 rhyl-f bulbar, ff 11 2 36 18 27 27 qrtz-v chunk 25 21 28 q r t z - v c h ip 12 10 29 qrtz-v core f 39 .. 27 31 r h y l f la k e ' 5 2 58 11 14 34 7 f la k e 15 4 55 22 20 •34 q r t z - v c h ip 20 10 35 r h y l - f m e d ia l f f 31 22 36 q r t z - v c h ip 20 17 38 r h y l - f c h ip 9 7 39 r h y l - b f la k e 7 2 72 21 19 40 r h y l f la k e 3 1 78 21 25 41 r h y l d i s t a l f f 13 11 42 rhyl-b bulbar ff 6 2 42 14 15 43 r h y l - f f la k e 7 2 65 25 19 45 r h y l - f f la k e 10 2 54 16 21 X 46 r h y l f la k e 4 1 72 9 15 47 r h y l - f c h ip 11 10 4 8 qrtz-v chip natural rock 18 15 49 r l i y l - f f la k e 10 3 56 19 16 50 r h y l f la k e 3 1 88 8 8 X 52 r h y l - b f la k e 6 1 70 12 16 54 r h y l - f d i s t a l f f 12 12 55 q r t z - v c h ip natural rock 26 15 56 q r t z - v c h u n k natural rock 35 22 58 r h y l- b flake crushed 9 18 59 r h y l- b f la k e 7 1 10 8 60 r h y l - f f la k e 5 2.5 4 0 12 17 62 qrtz-v chunk natural rock 36 22 63 qrtz-v bulbar ff 11 3 12 18 64 r h y l- b b u lb a r f f 5 1 55 9 10 65 r y h l - b b u lb a r f f 4 1 65 9 10 66 r h y l f la k e 6 2 53 18 25 67 r h y l f la k e 9 3 57 14 19 68 qrtz-v chunk natural rock 27 20 7 0 r h y l c h ip 13 12 25.3

Table 4: continued

Catalog M aterial Subdivision U til­ P la tfo r m A n g le F la k e ( n o ) H in g e d Num ber iz e d W idthI Depth Length |W idth

7 3 r h y l - f f la k e c ru s h e d 1 6 2 1 7 4 r h y l f la k e 1 7 5 72 2 8 3 3 7 6 r h y l - f f la k e 4 1 6 4 1 7 1 6 7 7 r h y l f la k e 9 2 60 1 8 2 0 X 7 8 r h y l - b c h ip 1 3 9 7 9 r h y l - f b u lb a r f f 21 4 35 3 7 3 0 81 r h y l - b b u lb a r f f 8 1 8 1 3 8 2 r h y l f la k e 7 3 4 1 8 3 9 8 3 q r t z - v c h ip 2 7 1 3 8 4 r h y l f la k e 1 0 3 8 2 2 3 2 8 5 r h y l f la k e 3 1 5 8 1 2 1 0 X 8 7 qrtz-v bulbar ff ^. 6 2 1 8 1 9 8 9 q r t z - v f la k e 1 9 4 7 6 1 4 2 1 91 r h y l - f medial f f •’ 1 8 1 4 9 2 q r t z - v f la k e 6 3 7 6 1 5 2 1 9 3 q r t z - c f la k e 3 1 35 1 0 1 2 9 4 q r t z - c f la k e - 5 1 1 5 1 1 9 9 5 q r t z - v f la k e X 1 6 8 65 4 2 3 7 9 7 r h y l m e d ia l f f •. 1 6 1 7 9 3 q r t z - v c h ip 1 3 1 1 1 01 rhyl-f bulbar ff 1 3 3 2 2 1 3 1 6 •1 0 6 q r t z - v c h ip 1 2 8 1 0 7 q r t z - v c o r e f X 1 5 2 1 1 0 8 q r t z - c b u lb a r f f 2 3 . 10 2 5 2 8 3 5 1 0 9 q r t z - v b u lb a r f f X 30 12 3 6 4 5 1 1 0 q r t z - c c h ip 1 9 1 3 111 q r t z - v b u lb u r f f 6 2 6 0 2 0 1 9 1 1 2 q r t z - c c h ip 2 0 1 9 1 1 3 qrtz-v bulbar ff 11 87 3 7 3 8 1 1 4 q r t z - v c h ip natural rock 3 5 • 2 0 1 1 5 r h y l f la k e 2 1 52 1 0 1 5 X 1 1 6 r h y l - b f la k e 4 1 38 1 6 1 6 1 1 7 r h y l c h ip 1 3 7 1 1 8 q r t z - v c h ip natural rock — — 1 2 0 q ir t z - v c h ip 2 0 1 4 121 r h y l - b f la k e X 21 2 2 2 2 7 1 2 2 r h y l - f f la k e 2 1 11 2 1 1 2 5 q r t z - v c h u n k X 2 2 2 5 1 2 6 r h y l f la k e 1 7 2 100 1 3 1 3 X 1 2 7 q r t z - v c h ip 1 0 1 3 131 r h y l - b c h ip X 2 6 2 2 1 3 3 r h y l - f f la k e 8 4C8 2 2 1 0 1 3 6 q r t z - c f la k e 7 2 8 3 1 0 1 1 1 3 7 rhyl-f flake 5 2 4 2 1 6 2 2 1 3 8 r h y l f la k e 5 c ru s h e d • 1 9 4 1 1 3 9 q r t z - v c o re 3 8 3 8 1 4 0 r h y l f la k e 1 3 5 2 0 4 1 31 141 r h y l - b b u lb a r f f X 2 5 7 1 9 4 0 142 r h y l - f c h ip 9 1 3 1 4 3 r h y l - f f l a k s X 38 5 32 4 3 4 2 254

Table 4: continued

C a ta lo g M a t e r ia l Subdivision U t i l ­ P la tfo r m Angle Flake(mm) H ia g e d N im b c r iz e d Width Ijepth Length Width

144 r h y l f la k e 6 2 17 15 X 145 r h y l - 6 c h ip 9 8 146 r h y l d i s t a l f f 19 18 148 r h y l - b f la k e 12 3 77 17 26 r h y l - f f la k e X 14 4 46 41 45 157 r » iy l- b c h ip 13 23 158 qrtz-v c h ip 24 24 159 r h y l b u lb a r f f 11 3 11 22 164 qrtz-v b u h la r f f 10 63 48 4 0 165 r h y l - f c h ip 45 26 166 r h y l c h ip 12 8 168 r h y l - f d i s t a l f f 21 23 — — B A 104-1* r h y l f la k e 12 3 58 18 35 r h y l - f f la k e • 7 2 38 22 25 r h y l f la k e 12 2 47 18 18 r h y l - f f la k e 5 1 20 15 X r h y l f la k e •6 • 2 68 14 17 r h y l ‘ f la k e 11 • 3 48 13 14 r h y l f la k e 7 2 66 14 18 r h y l - f b u lb a r f f 3 5 11 12 r h y l b u lb a r f f 5 2 27 11 13 r h y l - f f la k e 5 5 12 10 r h y l f la k e 2 5 6 9 r h y l - f f la k e 6 1 58 10 9 r h y l f la k e 6 1 10 14 X r h y l b u lb a r f f 4 3 4 0 9 12 r h y l - f f la k e X 23 17 48 20 23 r h y l b u lb a r f f 20 27 flnt b u lb a r f f 10 1 14 13 r h y l - f c h ip 21 27 r h y l - f c h ip 18 18 r h y l c h ip 19 10 r h y l - f D e d ia l f f 23 22 r h y l d i s t a l f f 15 14 r h y l f la k e 6 1 8 10 r h y l - f d i s t a l f f 21 19 Ttyl f la k e 10 3 78 17 25 qrtz-c c h u n k * 45 26 qrtz-c f la k e 14 4 53 14 15 qrtz-c c h u n k 13 21 qrtz-c b u lb a r f f 11 1 59 13 14 qrtz-v ( 8 ) c h ip s qrtz-v (12j chunks

* Formerly site No. 18 BA 104, now included in 18 BA 106. 255

Table 5: Debitage data for 1973 random finds collection.

C a ta lo g M a t e r ia l S utdlviiiion Utll-I PI a".forr. A n g le Flalc.5 (mmj H in g e d H u n b er Ized 1 W idth D e p th Ler,gth Width

B A -5 r h y l f la k o X 9 2 1 1 8 27 B A -6 qrtz-c distal ff X 3 1 0 35 2 8 2 7 BA-11 qrtz-c chip 1 0 1 3 BA-11 rhyl-f b u lb a r f f X 35 1 0 4 0 21 38 BA-12* qrtz-c chip. 2 1 2 6 B A -12 qrtz-cchip 18 2 4 BA-12 qrtz-v f la k e 26 7 2 0 2 8 2 0 B A -1 2 qrtz-v flake 20 17 2 0 32 2 8 BA-13 qrtz-c f la k e 17 4 16 2 5 2 2 B A -1 5 r h y l - f f la k e . 3 1 - 5 8 . 9 B A -1 5 r h y l - f f la k e '13 3 45 17 2 2 E A -1 7 ** r h y l f la k e 9 2 30 2 7 BA-17 qrtz-c chip 25 15 BA-18 Rhyl-f bulbar ff 10 3 4 5 17 2 2 BA-19* rhyl flake 5 2 5 20 16 B A -1 9 r h y l c tru n k ' "X 53 2 6 B A -1 9 r h y l c h ip 23 2 5 BA-20 rhyl-f chip 22 11

* Falls within expanded boundary of 18 BA 106. ** Falls within expanded boundary of 18 BA 122.

Table 6 : Debitage data for 1974 random collection from 18 HA 112.

C a ta lo g M a t e r ia l Subdivision U t i l ­ Pint f o r a A n g le F la k e (mm) H in g e d dum ber iz e d W idthjDepth Length Width

1 r h y l m e d ia l f f 2 1 2 0 2 r » q r l f la k e 1 1 3 8 5 1 2 16 3 r l ^ l f la k e 1 2 5 4 8 50 2 0 5 q r t z - c c o re 4 4 4 0 7 q r t z - c c h ip 2 6 2 3 9 r h y l - f b u lb a r f f 1 4 3 4 8 21 22 11 r h y l - f b u lb a r f f 7 2 4 3 1 6 2 2 12 r h y l - f d i s t a l f f 2 7 21 13 r h y l - f b u lb a r f f 5 1 32 2 0 2 5 1 4 r h y l d ie t a x f f 1 5 St 1 6 q r t z - c f la k e 8 5 8 0 2 0 2 2 256

Figure 44: Projectile point classification code.

1 . BLAD2 OUTLirra 0-- indetcrninate 1 - Incirrvate 2 - triangular A

3 - ovate ) t \ 4 - parallel-ovate J/\ 5 - excurvate 5 6 - cxpandins-ovate 7 - cxcurvate-incurvate 8 - pentagonal 8

2,TRANSVE3^33 S2GTI0tIi observed at the midpoint of the blade 0 - in4terr.iinato 1 - plano-convex 2 - piano-triangular 3 - biplano 4 - biconvex 5 - bitriar-gular 6 - asymmetrical 7 - triangulc-convex

3,L0NGrrUDH,'AL SSGTIO.Vi outline of blade element 0 - Indeterminate 1 - biconvex 2 - concavo-convex 3 - plano-convex 4 - biplano

4 .BASS om -Lixs 0 - indeterm inate 1 - straight 2 - concave 3 - c o n v e x 4 - triangulo-convex(point) 5 - triangulo-convex(arc) 6 - parabolic 7 - trlangulo-concave(point j 8 - triangulo-concave(arc) 9 - recurvate-concave 257

Figure 44: Projectile point classification code (cont.).

5,juNcrruB3 A 0 - Indeterninate 1 - lateral-lateral 2 - lateral-basal 3 - lateral-coincidental 4 - lateral-base 5 - coincidental-basal 6 - basal-basal 7 - lateral-axial 8 - triangle 8

6 .AKt;LS OF TIE UPPSR SUBELEMEMT OF TH2 HAFT 0 - Indeterninate 1 - curvate concave 2 - expanding 3 - parallel 4 - contracting-concave 1 . . Ï \ r V V ' 5 - contracting-teonvex 1 2 6 - contracting

7.AKGL2 OF THE SHOULDER: the angle formed at the longitudinal axis by a lin e drawn along the shauld'to the longitudinal a x e s . I Z p 0 - indeterm inate 1 - a c u te 2 - right-angled 3 - obtuse

*Adapted from Bebrich (196?) and Kenti (1970) (1970) ' I I 258

Table 7: Piercing category (projectile point) attribute list.

S i t e C a ta , M a t e r i a l L e n g th W id th ■ S t y le Edge Re­ M u l t i - 183A tio . B la d e :r o x ,| M e d ia l [ D is t a l 1234567 A n g le touch Function

106 1 4 q r t z - c 2 7 2 2 27118 70 e 1 0 6 1 7 q r t z - v 11 1 8 1 5 16? 1 1 1 3 1 0 6 21 q r t z - v 1 4 1 0 1 7 1 5 1 6 3 6 1 1 1 1 3 4 0 X 106 3 7 q r t z - c 1 8 10 14 2 7 3 5 7 5 45 1 0 6 51 r h y l - f 343 4 2 c 106 5 7 q t z t 2 8 1 2 1 4 1 3 2 3 37325 2 3 4 8 X 1 0 6 61 q r t z - c 3 8 10 18 1 3 2 1 34335 2 2 4 0 X 1 0 6 1 2 9 r h y l - b 11 3 5 1 0 6 169 r h y l 6 5 1 6 1 7 1 7 2 3 27322 3 3 45 X c 1 0 6 1 r s * q r t z - v 1 6 1 2 2 0 1 7 2 3 21325 1 3 60 X 1 0 6 2 r s q r t z - c 1 7 A 6 2 4 3 2 8 26 1 0 6 3 r s q r t z - c 1 8 1 6 2 2 3 1 8 32 106 4ra rhyl-f 30 •2 3 2 4 1 2 8 5 9 X ■ c 1 0 6 5 t 3 r l ÿ l • 3 5 . 8 .1 4 1 3 1 6 34435 1 3 4 2 X 1 1 2 2 7 q r t z - v 6 ) ? 17 2 2 . 2 3 3 3 26315 1 3 4 0 e 1 1 2 3 8 r h y l 2 5 1 8 3 5 1 1 1 2 6 7 r h y l 4 4 7 16 2 3 1 ? . •- 2 4 2 6 3 8 1 1 3 47 1 1 2 101 r h y l - f ^ ' 2 6 3 4 0 X 1 12 1 r s r h y l 51 12 1 0 1 2 1 8 34312 4 3 3 5 c 1 1 2 2 r s r h y l 34 25 112 BA17 q t z t 5 1 1 5 2 1 1 8 2 2 25325 1 3 70 X

* Random survey, 1974.

Table 8 : Scraping category attribute list.

S i t e C a ta . M a t e r i a l s u b ­ D ir e c t io n Active Slerient Length liultl- 18BA MO. d i v i s i o n o f t e a r AngiejLengthjPlan|Proflie slad e 1 Hai't fu n c t io n

10.5 t o r h y l - f biface nulti 4 0 - 4 5 31 c v St 17 1 9 106 8 3 q r t z - v b i f a c e u n i 6 5 - 7 0 3 0 c v s t 106 103 q r t z - v b i f a c e u n i 6 0 - 7 5 27 c v i r r g 1 0 6 109 q r t z - v biface nulti 65-70 2 2 i r r g i r r g 1 0 6 125 q r t z - v ch u n k , u n i 70 2 1 i r r g i r r g c 1 0 6 131 r h y l f l a k e u n i 7 5 2 2 c e s t 1 0 6 154 r h y l - f f l a k e u n i 125 32 i r r g c c 106 16 2 q r t z - v b i f a c e u n i 7 5 - 8 0 3 5 c v i r r g e 1 0 6 1 6 3 q r t z - v f l a k s n u l t i 6 5 3 3 c v i r r g c h 1 0 6 BA19 r h y l ch u n k u n i 8 8 2 2 c e i r r g • u n i 90 4 8 e t i r r g " u n i 6 5 3 5 I r r g i r r g 1 1 2 1 4 r h y l - f f l a k e u n i 70 2 0 c c s t c h 1 1 2 1 6 r h y l c h ip u n i 4 0 2 0 S t c e 1 12 17 r h y l f l a k e u n i 4 0 17 c c s t 1 1 2 43 q i t z - c b i f a e e u n i 5 8 - 6 5 30 I r r g I r r s e 1 1 2 4 6 q r t z - c b i f a c e u n i 90 25 c v c v X c h 1 1 2 50 r h y l - f c o r e f u n i 70-75 2 9 c v s t 259

Table 9: Cutting category attribute list. s i t e C a ta . K a t e r l o l C u b - C o n d i­ A c t l v n t H a f t B ase i - a f t - l l u l t l - 18HA KO. d i v i s i o n t io n Krj;lc i i a n rl*oiiie1 t e n . i.e n . W ld . ed. F i n c t io u

et^2J 1 0 6 1 4 q r t z - c b i f a c o c o o p i^ ^ O s t 27 2 2 X P 1 0 6 3 6 q r t z - v b i f a c o p r o x . 5 5 CV s t 25 X 1 0 6 51 r h y l - f biface dist. 42 c v s t 7 P 1 0 6 7 5 r h y l b i f a c o p r o x . 5 5 25 25 X 1 0 6 1 0 2 q r t z - c b i f a c e p r o x . 5 0 s t s t 2 2 2 4 X 1 0 6 1 0 7 q r t z c core-f coap. 5 6 c v c v 15 106 121 r h y l - b f l a k e c o a p . 1 8 c v s t 15 1 0 6 1 2 5 q r t z - v chunk c o a p . 6 0 s t s t 12 • 1 0 6 1 2 9 r h y l b i f a c e p r o x . 30-^M) c v s t 12 1 1 X P 1 0 6 1 4 1 r h y l - b b u l h a r f f c o o p . 24 c v a t 15 1 0 6 1 4 3 r h y l - f f l a k e COOD. 2 5 c c s t 17 " 2 7 Irrg s t 4 6 1 0 6 1 5 6 q r t z - c b i f a c o d l t t * 70 c v s t • 1 0 6 1 6 2 q r t z - v b i f a c e c o a p . 4 0 - 6 0 i r r g l i r g 32 >8 1 0 6 1 6 9 r h y l bifaee coap. 45-50 s t s t 6 2 1 6 17 X P 106 4rs* rhyl-f B lf a c e c o a p . 50 c c s t 30 2 8 X P 1 0 6 6 n r h y l - f b l f a c e d i s t . 30 s t s t ' 7 . 1 1 2 9 r h y l - f f l a k e c o a p . 2 8 c v s t ' 1 2 1 1 2 1 8 r h y l f l a k e c o a p . 6 0 - 7 0 c v s t 1 2 ch 1 1 2 2 7 q r t z - v blface prox. 4 0 s t s t 6 3 ? 8 1 4 X P 1 1 2 43 q r t z - c blface coon. 58-65 i r r g s t 31 39 8 112 61 r h y l biface dist. 40-70 st irrg 1 1 2 1 r s r h y l b l f a c e p r o x 3 5 c v s t 51 12 2 1 X P 1 1 2 3 r s r h y l c o r e c o a p 40-50 irrg I r r g 71 45 112 4 r s r h y l - f blface dist. 4 2 c v s t 1 4 122 8 r h y l f l a k e conp 5 6 31

* Random survey, 1974.

(1 ) comp.= complete, prox.= proximal, medl.= medial, dist.=distal.

(2 ) st - straight, cv = convex, cc = concave, irrg - irregular.

Table 10: Chopping category attribute list.

S i t e te a ta . Material Sub­ D ir e c t io n Active Slenent 1 L a r .z th l l u l t i - 18BA « 0 . d i v i s i o n 3:' W ear Angle ll«nrtn FlanI Profile 1 ‘.ila c e 1 r d f t f u n c t i o n

1 0 6 2 r h y l f l a k e n u l t i 3 5 - 6 2 2 1 c v i r r g 1 1 10 1 0 6 2 6 r h y l b lf a c e * iin l 30-70 21 c v i r r g 1 9 2 5 1 0 6 9 5 qrtz-v flake n u l t i 70-75 2 6 i r r g I r r g u n h a fte d 106 160 qrtz-v flake n u l t i 2 6 c ru s h e d u n h a fte d 1 1 2 1 4 r h y l - f f l a k e u n i 4 0 - 6 0 3 3 c v c v 1 9 2 0 8 1 1 2 1 8 r h y l f l a k e n u l t i 60-70 30 c v i r r g 1 1 1 0 c 1 1 2 81 r h j l b l f a c e u n i 4 0 - 5 0 31 cv irrg 30 27 260

Figure 45: Functional attribute code.

1. Types of wear: P - polish S - smoothing ST - step flaking R - rounding F - faceting C - crushing

2. Nature of edge retouch: A - extensive retouch causing obliteration of wear patterns B - partial retouch, some wear still present C - utilized after retouch D - undiagnostic

3. Types of fracture: T - transverse I - inverse D - distal

4. Position of wear: The number refers to the functional edge under discussion. A specimen can have any number of functional edges.

5. Inferred function: B - burin C - cutting CH - chopping P - piercing C - scraping 261

Table 11: Wear pattern attribute list for 18 BA 106.

C a t a l o g F i g u r e Facial Edge Edge Frac­ I n f e r r e d F u n c t i o n N o . N o . Wear Wear Retouch ture FirstISecond 1 Third

2 3 2 : L 1 - S 1 - S T . 2 - R , CH 3 - S 1 4 3 1 : FF 1 - S - R - F 2 - A P C S 1 7 TP 2 1 3 1 : E 1 - 2 - A T) P 2 6 3 2 :G 1 - 2 - 1 - 2 - 0 - 3 , R-S 3 - R CH 3 6 3 2 : DO 1 - S C 3 7 3 1 -.e e 1 - R . 2 - R - S P 5 1 1 - S . 2 - G - R T-D C P 5 7 3 1 : L 1 - R 2 - A P 6 1 31:Y 1-2-R 1-2-B TP 7 5 3 1 : BB 1 - P 1 - 2 - R - F T C 8 0 1-P 1-R, 2-R-S S 8 8 1 - R - S S 9 5 3 2 : 2 1-C, 2-R-S CH 1 0 2 32:0 1-2-P 1-2-F-S C 1 0 3 1 - S 1 —F—S—C S 1 0 7 3 2 :B B 1 - S 1 - 2 - S C 1 0 9 3 2 : H 1 - 0 , 2 - R 2 - B TS 1 2 1 1 - P - S T , 2 - P c 1 2 5 32:CC 1-2-C-R, s C 3 - R . 4 - F 1 2 9 1-2-P 1-R.2-S I p 1 3 1 1 - P 1 - S . 2 - P S 1 4 1 1 - S - R , 2 - F c 1 4 3 1 - 2 - R c 1 5 4 1 - R 1 - R S 1 5 6 3 2 : 0 1-R-F-S,2-F c 1 6 0 3 2 : 2 1 - C , 2 - S CH 1 6 2 1 - C - S , 2 - S S C 1 6 3 32:Y 1-C,2-S-P CHS 1 6 9 3 1 : 0 1-2-P •1-2-R D C P B BA 19-1 32:K 3-P 1-ST-C-F-R, S 2 - S T - C . 3 - C 1 r s 31:S 1—2—C—S 1 - 2 - C P 2 r s 3 1 : J J 1 - 2 - P 1 - 2 - P DP 3 r s 3 1 :G G 1 - 2 - P IP 4 r s 3 1 :C C 1-F-S-R,2-S P C 5 r s 3 1 : Q 1 - 2 - S - F PC 6 r s 1 - P 1 - F , 2 - R - S C 262

Table 12: Wear pattern attribute list for 18 BA 112.

Catalog Figure Facial E d g e E d g e F r a c ­ Inferred Function N o . N o . W e a r W e a r Retouch ture First Second T h i r d

9 1 - R - F S C 1 4 3 4 : S 1 - S T , 2 - C S CH 1 6 1 - S - R S 17 1-S 1-SÏ-R S 18 34:T 1-S-P 1-ST.2-P-R C 2 7 3 4 : L 1-F,2-F-R P C 4 3 3 4 : N 1 - S C S 5 0 3 4 : Y 1 - S T - S S 6 1 1 - P 1 - 2 - F - R C 67 34:B 1-2-S I-B P 8 1 3 4 : 0 1-P 1—C-R—S, 2-S C CH 1 0 1 1 - 2 - P DP 1rs 34:F 1-S-F, 2—S-R TCP 2 r s T P 3 r s 3 4 :W 3 - P 1-R-F,2-ST, 3 - P C 4rs 1-F-R,2-R C 8 * 3 5 :Q 1 - P 1 - F - R C

«From 18 BA 122, 263

Table 13: Summary of 1973-75 artifact categories for 18 BA 106.

y CO M Cd N fH w f-i < I Is; I L P h H « t?: <: C f 5 to o < M M w I ! CATEGORY Hi ►P H n o f w I W M K a È3 •i M CP I*.1 6 5. . . , V'i F-« w 0 M a c II CP cH M > u I I ?n p, >-3 to c> CP a Kto 1.1 5 g Flake 1 0 2 42 1 6 3 147 3 2 3 1 1 1 1 3 6 6

Chip 7 6 18 2 0 3 0 1^14 Chunk 3h 4 3 41 Gore 4 1 2 7

Scraping 14 1 5 2 2 1 2 5

Cutting 16 6 2 3 5 3 2 Chopping 3 2 5 Piercing E, Archaic 1 i

M, Archaic 1 1 2

L, Archaic’ 2 8 3 3 1 1 7

E, Woodland 1 1

M, Woodland 1 1 L, Woodland 7 1 8 Undianostic 1 1 2 Grinding Stone 1 1 Hammerstone 1 1 Bannerstone 2 1 3 Fire Cracked ■1 Rock 6 1 6

1 1 2 1 2 6 6 2 Total 1 2 5 8 88 8 97 1 9 0 6 2 6 Ceramics Shell Temper UO Qi;artz Temper 93 Saad Temper 17

Total 1 5 0 264

Table 14: Summary of 1973-75 artifact categories for 18 BA 112.

N ►a CO N Ni E ' C) g H t ' œ < 5 g 3 z;i « 0 g od Z) c y % -=c Cf Of H tP M << u cf w ca N o s T0T.4L CATEGORY Z) E-. C.1 E-t CO Ï5 m cf to M H M < 5 5 w 0 K Z3, i-J S I > • 0 0 0 o ’ '-M 1 2 ;^ CO E-<

Flake 5 6 4 1 3 13 1 3 4 7

Chip 6 2 2 1 1 14 49 1 7 6

Chunk 8 1 7 Core 1 Scraping 2 2 6

Cutting 3 8 1 5 Chopping 2 2 Piercing E, Archaic 1 2 M. Archaic 2 3 L, Archaic 2 6 I'!, Woodland 1 1 Undianostic 3 5 Fire Gracke Rock 1 Total ( 2 5 1 2 7 2 5 75 3 3 6 3 1 7 1 483

Ceramics 1 Accokeek Coijd Ma-.-ked 1 Total 1 ______References Cited

Adovaslo, J. M., J. M. Gunn, J. Donahue, and R. Stuckenrath 1975 Excavations at Meadowcroft Rockshelter, 1973-1974: a progress report. Pennsylvania Archaeologist 45: 1-30.

Ahler, Stanley A. 1971 Projectile point form and function at Rodgers Shelter, Missouri. Missouri Archaeological Society Research Series 8 .

Arber, Edward 1910 Travels and works of Captain John Smith President of Virginia and Admiral of New England, 1580-1631, Vols. 1 and 2. Burt Franklin, New York.

Ayers, Harvard G. 1972 The archeology of the Susquehanna tradition in the Potomac Valley. Ph.D. dissertation. Catholic University of America University Microfilms, Ann Arbor.

Bastian, Tyler 1974 Preliminary notes on the Biggs Ford site, Frederick County, Maryland. Division of Archeology, Maryland Geological Survey. Typed ms.

Bebrich, Carl A. 1967 Lithic artifacts from Sheep Rock. In Archaeological Inves­ tigations of Sheep Rock Shelter, Vol. 2, edited by Joseph Michels and Ira F. Smith, pp. 193-221. The Pennsylvania State University, University Park.

Bennett, Ann M. 1974 Basic ceramic analysis. San Juan Valley Archeological Project Technical Series, Eastern New Mexico University

Contributions in Anthropology, Vol. 6 .

Biggar, H. P. 1965 The early trading companies of New France: a contribution to the history of America and discovery in North America. Argonaut Press, New York.

Binford, Lewis R. 1967 An ethnohistory of the Nottoway, Meherrin and Weanock Indians of Southeastern Virginia. Ethnohistory 14: 104-218.

1968 Methodological considerations of the archeological use of ethnographic data. In Man the Hunter, edited by Richard B. Lee and Irvin De Vore, pp. 268-272. Aldine, Chicago.

1973 A consideration of archeological research design. In Contemporary Archeology, edited by Mark P. Leone, pp. 158- 177. Southern Illinois University Press, London.

265 266

Binford, Lewis R., Sally R. Binford, Robert Whallon, and Margaret Ann Hardin 1970 Archaeology at Hatchery West. Society for American Archaeology, Memoir 24.

Blaker, Margaret C. 1963 The Townsend site near Lewis, Delaware. The Archeolog 25: 14-34.

Braun, E. Lucy 1950 Deciduous forests of eastern North America. Hafner Publish­ ing Company, New York.

Brennan, Louis A. 1974 The lower Hudson: a decade of shell middens. Archeology of Eastern North America 2: 81-93.

Brose, David S. 1975 Functional analysis of stone tools: a cautionary note on the role of animal fats. American Antiquity 40: 86-94.

Broyles, Bettye J. 1971 Second preliminary report: the St. Albans site, Kanawha County, West Virginia. West Virginia Geological and Eco­ nomic Survey Report of Archaeological Investigations No. 3.

Brush, Grace 1975 Forest ecology of the eastern piedmont Province of Maryland. Maryland Academy of Science, Baltimore.

Butzer, Karl W. 1973 Environment and archeology: an ecological approach to prehistory. Aldine, Chicago.

Campbell, B. U. 1906 Early Christian missions among the Indians of Maryland. Maryland Historical Magazine 1: 293-316.

Carbone, Victor A. 1974 The Paleo-environment of the Shenandoah Valley. In The Flint Run Paleo-Indian complex: a preliminary report 1971-73 seasons, edited by William M. Gardner. Catholic University of America, Archeology Laboratory, Department of Anthropology, Occasional Publications No. 1: 84-99.

Clark, Wayne E. 1970 The Elkridge site: initial excavations and interpretation. Maryland Archeology 6: 32-57.

1973 Archeological survey of the Northwest Transportation Corridor, Baltimore County, Maryland. Division of Arche­ ology, Maryland Geological Survey. Typed ms. 267

Clark, Wayne E. 1974a Archeological sites: prehistoric period. Governor's consulting committee report: October 1974 meeting. Maryland Historical Trust, Annapolis. Xeroxed.

1974b The application of the direct historical approach to historical and ecological problems in Maryland, part I. Department of Anthropology, The American University. Xeroxed.

1974c The application of the direct historical approach to historical and ecological problems in Maryland, part II. Department of Anthropology, The American University. Xeroxed.

1974d Cony-Selby Bay-Fox Creek "Projectile Points": a problem in classification. Paper presented at the Fifth Middle Atlantic Archeological Conference,Baltimore.

1975a Northwest Transportation Corridor archeological survey project: 1974 interim report, 18 BA 106 and 112. Division of Archeology, Maryland Geological Survey. Typed ms.

1975b Controlled surface investigations of an Archaic period hunting camp. Maryland Archeology 11: 1-18.

1975c The Northwest Transportation Corridor archeological survey project, 1973-1975: final report. Division of Archeology, Maryland Geological Survey. Typed ms.

Clark, Wayne E., and W. Dana Miller 1975 Projectile points. In Report on the excavations at UMBC site 18-BA-71, edited by Karen D. Vitelli. The Arche­ ological Society of Maryland, Miscellaneous Papers 10: 27-52.

Cleaves, Emery T. 1973 Chemical weathering and landforms in a portion of Baltimore County, Maryland. Unpublished Ph.D. dissertation. Depart­ ment of Geography, The John Hopkins University, Baltimore.

Cleland, Charles E. 1966 The prehistoric animal ecology and ethnozoology of the Upper Great Lakes region. University of Michigan, Museum of Anthropology, Anthropological Papers 29.

Coe, Joffre L. 1964 The formative cultures of the Carolina Piedmont. Trans­ actions of the American Philosophical Society 54.

Crabtree, Don E. 1972 An introduction to flintworking. Idaho State University Museum, Occasional Papers 28. 268

Cresthull, Paul 1969 Bifurcate base and other Early Archaic points in Northern Maryland. Maryland Archeology 5: 39-49.

Davidson, Isobel 1967 Real stories from Baltimore County. Tradition Press, Hatboro.

DeJarnette, David L., John A. Walthall and Steve B. Wimberly 1975 Archaeological investigations in the Buttahatchee River Valley II: excavations at Stucks Bluff Rock Shelter. Journal of Alabama Archaeology 21: 100-119.

Dunnell, Robert C. n.d. Archaeological potential of antrhopological and scientific models of function. In Archaeological Essays in Honor of Irving Rouse, edited by R. C. Dunnell and E. S. Hall, Jr., Chandler and Sharp, San Francisco (In press ms. 1974).

Emery, K. 0. and K. L. Edwards 1966 Archeological potential of the Atlantic Continental shelf. American Antiquity 31: 733-737.

Evans, Clifford 1955 A ceramic study of Virginia archeology. Smithsonian Institution Bureau of American Ethnology, Bulletin 160.

Feest, Christian F. 1973 Southeastern Algonquian burial customs: ethnohistorical evidence. In Proceedings of the Fourth Annual Middle Atlantic Archaeological Conference, edited by Ronald A. Thomas, pp. 1-16.

Fenton, William N. 1952 The training of historical ethnologists in America. American Anthropologist 54: 328-339.

Ferguson, Alice L. L. and Henry G. Ferguson 1960 The Piscataway Indians of southern Maryland. Alice Ferguson Foundation, Accokeek.

Fitzhugh, William 1972 The eastern Archaic: commentary and northern perspective. Pennsylvania Archaeologist 42: 1-19.

Flint, Richard F. 1957 Glacial and Pleistocene geology. John Wiley and Sons, Inc., New York.

Ford, T. Latimer 1958 Adena traits in Maryland. Eastern States Archeological Federation, Bulletin 17: 10-11. 269

Gardner, William M. 1972 Research proposal of the Flint Run complex: a study of Paleo-Indian cultural systems in the Shenandoah Valley of Virginia. Submitted to the National Science Foundation. Department of Anthropology, Catholic University of America. Xeroxed.

1974 The Flint Run Paleo-Indian complex: a preliminary report 1971-73 seasons. Catholic University of America, Arche­ ology Laboratory, Department of Anthropology, Occasional Publications No. 1.

1975 Eastern North American archeology: general summary. Paper presented to 1975 seminar class in Eastern North American Archeology, Catholic University of America. Mimeographed.

Geasey, Spencer 0. 1971 The Tuscarora Rock Shelter (Site 18 FR 9). Maryland Archeology 7: 1-16.

1974 Types of rhyolite cache blades found in Frederick County, Maryland. Maryland Archeology 10: 1-6.

Gluckman, Stephen J. 1973 Woodland in the Potomac basin: a model building approach to definition. Ph.D. dissertation. The Catholic University of America, Anthropological Studies 24.

Gregory, Leverette B. and Polly V. Jerrell 1976 The Powell's Creek site: a preliminary report. Virginia Research Center for Archaeology. Xeroxed.

Griffin, John W. 1974 Investigations in Russell Cave. National Park Service, Publications in Archeology 13.

Hakluyt, Richard 1589 The principal navigations voyages traffiques and discoveries of the English Nation, Vol. 8. Reprint 1968, University Microfilms, Ann Arbor.

Handsman, Russell G. and Charles E. Hunter 1972 Ethnohistory and anthropology: culture contact and conflict among the Piscataway. Department of Anthropology, The American University. Xeroxed.

Harrison, W., R. J. Mallow, Gene A. Ruemak, and J. Terasmae 1965 Possible Late Pleistocene uplift, Chesapeake Bay entrance. The Journal of Geology 73: 201-229.

Harshberger, J. W. 1903 The flora of the serpentine barrens of Southeast Pennsylvania. Science 18: 339-343. 270

Hayes, G. L. and Jesse H. Buell 1955 Trees also need water at the right time and place. The Yearbook of Agriculture. The United States Government Printing Office, Washington.

Heisey, Henry 1971 An interpretation of Shenks Ferry ceramics. Pennsylvania Archaeologist 41: 44-70.

Heisey, Henry W. and J. Paul Witner 1964 The Shenk's Ferry people. Pennsylvania Archaeologist 34: 1-34.

Hershfield, David H. 1971 The frequency of dry periods in Maryland. Chesapeake Science 12: 71-84.

Holmes, W. H. 1897 Stone implements of the Potomac Chesapeake Tidwater province. Fifteenth Annual Report of the Bureau of American Ethnology.

Hume, Gary W . 1976 The Ladizian: an industry of the Asian chopper-chopping tool complex in Iranian Baluchistan. Dorrance and Company, Philadelphia.

Hunter, William A. 1969 The historic role of the Susquehannocks. In Susquehannock M iscellany, edited by John Witthoft and W. Fred Kinsey, III, pp. 19-60. The Pennsylvania Historical and Museaum Com­ mission, Harrisburg.

Huntington, Ellsworth 1945 Mainsprings of civilization. John Wiley and Sons, Inc., New York.

Huttenhauer, Helen C., and C. Alfred Helwig 1967 Baltimore County in the state and nation. Board of Educa­ tion, Baltimore.

Jonas, Anna I. and G. W. Stone 1938 Geologic map of Frederick County. Maryland Geological Survey, Baltimore.

Jones, Robert W. 1931 The Clemson Mound. In Fifth Report of the Pennsylvania Historical Commission. Commonwealth of Pennsylvania.

Jones, Talbot D. 1903 Collection, catalog, and notes. Division of Archeology, Maryland Geological Survey. Xeroxed. 271

Kent, Barry and Ira F. Smith, III 1972 Community pattern studies. In Proceedings of the Third Middle Atlantic Archeological Conference, edited by Ronald A. Thomas, pp. 4-20.

Kerby, Maude Levey 1964 Climatic succession in the Western Hemisphere from the Third Interglacial. Quarterly Bulletin of the Archeological Society of Virginia 19: 26-35.

1967 Some notes on bison and elk east of the Mississippi River. Quarterly Bulletin of the Archeological Society of Virginia 21: 113-120.

King, Thomas F. 1971 A conflict of values in American archaeology. American Antiquity 36: 255-262.

Kinsey, W. Fred, III 1971 The Middle Atlantic culture province: a point of view. Pennsylvania Archaeologist 41: 1-8.

1972 Archeology in the Upper Delaware Valley: a study of the cultural chronology of the Tocks Island Reservoir. Pennsylvania Historical and Museum Commission Anthropo­ logical Series 2.

1975 Faucett and Byram sites: chronology and settlement in the Delaware Valley. Pennsylvania Archaeologist 45: 1-103.

Kraft, Herbert C. 1970 The Miller Field site, Warren County, New Jersey: part I. The Archaic and Transitional stages. Seton Hall University Museum Publications.

Kraft, John C. 1971 Sedimentary facies patterns and geologic history of a Holocene marine transgression. Geological Society of America, Bulletin 82: 2131-2158.

Lewis, Clifford M., S. J. and Albert J. Loomie, S. J. 1953 The Spanish Jesuit mission in Virginia, 1570-1572. The University of North Carolina Press, Chapel Hill.

Lopez, Julius 1961 Pottery from the Mispillion site, Sussex County, Delaware, and related types in surrounding areas. Pennsylvania Archaeologist 31: 1-38.

Luckenbach, A. H., R. 0. Allen, Jr., and C. G. Holland 1975 Movement of prehistoric soapstone in the James River basin. Quarterly Bulletin of the Archeological Society of Virginia 29: 182-202. 272

MacCord, Howard A., and Oliver D. Valliere 1965 The Lewis Creek Mound, Augusta County, Virginia (44 Au 20). Quarterly Bulletin of the Archeological Society of Virginia 20: 37-42.

Manley, P. C. 1963 Excavations of the Lewis Creek Mound, Augusta County, Virginia. Quarterly Bulletin of the Archeological Society of Virginia 18: 37-42.

Manson, Carl, Howard A. MacCord, and James B. Griffin 1943 The cultures of the Keyser Farm site. Papers of the Michigan Academy of Science,Arts, and Letters, Vol. 29:

Marye, William B. 1920 The Old Indian Road. Maryland Historical Magazine 15: 107- 124, 208-229, 345-395.

1935 Piscattaway. Maryland Historical Magazine 30: 183-240.

1955 The great Maryland barrens. Maryland Historical Magazine 50: 11-23, 120-142.

Maryland Geological Survey 1929 Baltimore County. The John Hopkins Press, Baltimore.

McCary, Ben 1975 Semi-lunar knives from tidewater Virginia. Quarterly Bulletin of the Archeological Society of Virginia 30: 80-83.

McDowell, Ellis E. 1970 Ruppert Island and Eastern United States cultural develop­ ment. M.A. thesis. Department of Anthropology, The American University, University Microfilms, Ann Arbor.

1972 The Archaic stage of the Potomac River Piedmont: a techno- ecological approach to archeological data. Ph.D. disserta­ tion, Department of Anthropology, The American University, University Microfilms, Ann Arbor.

McGimsey, Charles R., III 1974 These are the stewards of the past. University of Missouri, Columbia.

McGrain, John W. 1974 Industrial archaeological survey of the Northwest Trans­ portation Corridor Baltimore County, Maryland. Division of Archeology, Maryland Geological Survey. Typed ms.

McMillan, Barbara 1972 An archeological survey of St. Mary's County, Maryland. M.A. thesis. Department of Anthropology, The American University, University Microfilms, Ann Arbor. 273

McNett, Charles W., and William M. Gardner 1971 Early pottery in the Potomac. In Proceedings of the First Middle Atlantic Archaeological Conference, edited by Charles W. McNett, Jr., and William M. Gardner, pp. 42-52.

1975 Archeology of the lower and middle Potomac. Department of Anthropology, The American University. Xeroxed, draft ms.

McNett, Charles W., Jr., Sydne B. Marshall, and Ellis E. McDowell 1975 Second season of the Upper Delaware Valley Early Man Project. A report prepared for the National Geographic Society and the National Science Foundation. Department of Anthropology, The American University. Mimeographed.

Michels, Joseph, and Ira F. Smith 1967 Archaeological investigations of Sheep Rock Shelter, Huntingdon County. Pennsylvania, Vol. 1 and 2. The Penn­ sylvania State University, University Park.

Michlovic, Michael G. 1976 Social interaction and point types in the Eastern U. S. Pennsylvania Archaeologist 46: 13-16.

Milliman, John R., and K. 0. Emery 1968 Sea levels during the past 35,000 years. Science 162: 1121-1123.

Painter, Floyd 1967 Geometrically incised decorations on Great Neck ceramics, the Long Creek midden, part 3. The Chesopiean 5: 94-110.

Paradiso, E. Frances 1969 Mammals of Maryland. North American Fauna 66.

Parker, Arthur G. 1936 The Nanticoke. Pennsylvania Archaeologist 5: 83-90.

Pennell, R. 1910 Flora of the Conowingo barrens of Southeastern Pennsylvania. Proceedings of the Academy of Natural Science 62: 541-584.

Pory, John 1624 The observations of Master John Pory of Virginia, in his travels. In Historié of Virginia, New England, and the Summer Isles, edited by Captain John Smith. Reprinted in 1910. In Travels and Works of Captain John Smith, edited by Edward Arber, pp. 566-571. Burt Franklin, New York.

Ragir, Sonia 1973 A review of techniques for archaeological sampling. In Contemporary Archeology, edited by Mark P. Leone, pp. 178- 192. Southern Illinois University Press, London. 274

Ramsey, A. Ogden 1972 Ecology at McDonogh or whither Chinquapin Hallow? McDonogh. Spring Summer Catalog. McDonogh School, McDurogh, Maryland.

Ritchie, William A. 1961 A typology and nomenclature for New York projectile points. New York State Museum and Science Service, Bulletin 34.

1969 The archaeology of Martha's Vineyard: a framework for the prehistory of southern New England. A study in coastal ecology and adaptation. The Natural History Press, Garden City.

Ritchie, William A., and Robert E. Funk 1971 Evidence for Early Archaic occupations on Staten Island. Pennsylvania Archaeologist 41: 45-59.

1973 Aboriginal settlement patterns in the Northeast. New York State Museum and Science Service, Memoir 20.

Rouse, Irving 1960 The classification of artifacts in archaeology. American Antiquity 25: 313-325.

Salwen, Bert and Ann Ottesen 1972 Radiocarbon dates for a Windsor occupation at the Shantok Cove site. New London County, Connecticut. Man in the Northeast 3: 8-19.

Scarupa, Henry 1976 New identity for Maryland's Indians. The Sun Magazine, January 11 Issue: 6-9. The Baltimore Sun, Baltimore.

Scharf, J. Thomas 1971 History of Baltimore County and City. Regional Publishing Company, Baltimore.

Schiffer, Michael B. and John H. House 1974 The Cache River archeological Project, Draft 3. Xeroxed.

Schmitt, Karl 1965 Patawomeke, an historial Algonkian site. Quarterly Bulletin of the Archeological Society of Virginia 20: 1-36.

Seabease, Wilsie 1969 The Nanticoke and other Indians of Delaware. The Easton Publishing Company, Easton.

Semenov, S. A. 1973 Prehistoric technology. Harper & Row Publishers, Inc., New York. 275

Semraes, Raphial 1937 Captains and mariners of arly Maryland. The John Hopkins Press, Baltimore.

Shreve, Forrest, M. A. Chrysler, F. H. Blodgett and F. W. Besley 1910 The plant life of Maryland. Maryland Weather Service Special Publication 3.

Slattery, Richard G., William A. Tidwell, and Douglas R. Woodward 1966 The Montgomery Focus. Quarterly Bulletin of the Archeo­ logical Society of Virginia 21: 49-51.

Smith, Captain John 1612 A map of Virginia, with a description of its commodities, peoples, government, and religion, 1612. In Travels and Works of Captain John Smith, Vol. 2, edited by Edward Arber, pp. 41-174. Burt Franklin, New York.

Smith, Gerald P. 1971 Protohistoric sociopolitical organization of the Nottoway in the Chesapeake Bay-Carolina Sounds region. Unpublished Ph.D. dissertation. Department of Anthropology, University of Missouri.

Speck, Frank C. 1946 Catawba hunting, trapping, and fishing. Museum of the University of Pennsylvania-The Philadelphia Anthropological Society, Philadelphia.

Speck, Frank C., Royal B. Hasarick, and Edmund S. Carpenter 1964 Rappahannock taking devices: trapping, hunting and fishing. Museum of the University of Pennsylvania and the Philadel­ phia Anthropological Society, Number 1.

Spelman, Henry 1613 Relation of Virginia. In Travels and Works of Captain John Smith, Vol. 1, edited by Edward Arber, pp. ci-cxiv. Burt Franklin, New York.

Spencer, Edward 1906 Soldiers Delight Hundred in Baltimore County. Maryland Historical Magazine 1: 141-154.

Spicer, Edward H. 1962 Cycles of conquest, the impact of Spain, Mexico, and the United States on the Indians of the Southwest, 1533-1960. The University of Arizona Press, Tucson.

Stearns, Richard E. 1943 Some Indian village sites of tidewater Maryland. Natural History Society of Maryland, Proceedings No. 9.

1949 Some Indian village sites of the Lower Patapsco River. The Natural History Society of Maryland, Proceedings No. 10. 276

Stearns, Richard E. 1951 An Indian site survey of the Patuxent River, Maryland. Maryland Naturalist 21: 1-20.

1965 Indian village sites of the Patuxent River. The Journal of the Archeological Society of Maryland, Inc. 1: 40-49.

1966 Three aboriginal sites in the Pimlico area of Baltimore. The Journal of the Archeological Society of Maryland, Inc. 2: 70-73.

Stephenson, Robert L., and Alice L. L. Ferguson 1963 The Accokeek Creek site, a Middle Atlantic seabound culture sequence. University of Michigan, Museum of Anthropology, Anthropological Papers 20.

Taylor, Walter W. 1973 Old wine and new skins: a contemporary parable. In Contemporary Archaeology, edited by Mark P. Leone, pp. 28- 33. Southern Illinois University Press, London.

Thomas, Ronald A., Daniel R. Griffith, Cara L. Wise, Richard E. Artusy, Jr. 1974 A discussion of the lithics, ceramics and cultural ecology of the Fox Creek-Cony-Selby Bay paradigm as it applies to the Delmarva Peninsula. Report presented at the Fifth Middle Atlantic Archeological Conference, Baltimore.

1975 Environmental adaptation on Delaware's Coastal Plain. Archeology of Eastern North America 3: 35-89.

Thurman, Melburn 1973 A short paper on ossuaries. In Proceedings of the Fourth Annual Middle Atlantic Archaeological Conference, edited by Ronald A. Thomas, pp. 37-45.

1974 Delaware Social organization. In a Delaware Indian sym­ posium, edited by Herbert C. Kraft, pp. 111-134. The Pennsylvania Historical and Museum Commission, Anthropo­ logical Series 4.

Thurman, Melburn D., and William P. Barse 1974 Outline: Mockley and Mockley-like pottery in the Mid- Atlantic region. Paper presented at the Fifth Middle Atlantic Archeological Conference, Baltimore.

Tringham, Ruth, Glenn Cooper, George Odell, Barbara Voztek, and Anne Whitman 1974 Experimentation in the formation of edge damage: a new approach to lithic analysis. Journal of Field Archeology 1: 171-196. 277

Ubelaker, Douglas 1973 The Hurley ossuary at Nanjemoy Creek, In Proceedings of the Fourth Annual Middle Atlantic Archaeological Conference, edited by Ronald A. Thomas, pp. 17-36.

U.S. Department of Commerce: Environmental Science Service Adminis­ tration, Weather Bureau. 1965 Climatological data, Maryland and Delaware, annual summary 1965, Vol. 69, No. 13.

Vitelli, Karen D. and William P. Zaruba 1975 Other chipped and polished stone. In Report of the excava­ tion at UMBC site 18-BA-71, edited by Karen D. Vitelli. The Archeological Society of Maryland, Miscellaneous Papers 10: 62-84.

Vokes, Harold E. 1961 Geography and geology of Maryland. Department of Geology, Mines and Water Resources, Bulletin 19.

Wallace, Paul A. W. 1971 Indian paths of Pennsylvania. The Pennsylvania Historical and Museum Commission, Harrisburg.

Weslager, Clinton A. 1943 Delaware's forgotten folk: The story of the Moors and Nanticokes. University of Pennsylvania Press, Philadelphia.

1972 The Delaware Indians, a history. Rutgers University Press, New Brunswick.

White, Andrew 1632 A relation of Maryland. March of America Fascimile Series 22. University Microfilms, Ann Arbor.

White, Leslie A. 1959 The evolution of culture. McGraw Hill Book Company, New York.

Whitehead, Donald R. 1965 Prehistoric maize in Southeastern Virginia. Science 150: 881-883.

1967 Studies of full-glacial vegetation and climates in the Southeastern United States. In Quarternary paleo-ecology, edited by E. J. Cushing and H. W. Wright, pp. 237-248. Seventh Congress of International Association for Quarternary Research.

Wilkins, Elwood S., Jr. 1962 A preliminary report on the Harland Mill steatite quarry. Bulletin of the Archeological Society of Delaware 2: 1-12. 278

Wllm&en, Efeln N. 1970 Lithic analysis and cultural inferences: a Paleo-Indian case. Anthropological Papers of the University of Arizona. Number 16.

Witthoft, John 1953 Broad spearpoints and the Transitional period cultures. Pennsylvania Archeologist 23: 4-31.

1963 Section 8, general interpretations. In The Townsend site near Lewis, Delaware, edited by Gregory Omwake and T. Dale Stewart. The Archeolog 25: 59-68

Witthoft, John, and S. S. Farver 1971 Two Shenk's Ferry sites in Lebanon County, Pennsylvania. In Foundation of Pennsylvania Prehistory, edited by Barry C. Kent, Ira F. Smith, III, and Catherine McCann. The Pennsylvania Historical and Museum Commission Anthropo­ logical Series 1: 425-466.

Wright, J. V. 1968 The application of the direct historical approach to the Iroquois and the Ojibwa. Ethnohistory 15: 96-111.

Wright, Henry T. 1973 An archaeological sequence in the Middle Chesapeake region, Maryland. Maryland Geological Survey, Archeological Studies No. 1.