Archeological survey of the Salt Fork of the Arkansas River Basin in north central Oklahoma

              A 2200.7 R434s no. 42
c. 3 ' 2001
ARCHEOLOGICAL SURVEY OF THE SALT FORK OF
THE ARKANSAS RIVER BASIN IN NORTH CENTRAL OKLAHOMA
BY
RICHARD R. DRASS
WITH A CONTRIBUTION BY
SUSAN C. VEHIK
The University of Oklahoma
Oklahoma Archeological Survey
Archeological Resource Survey Report No. 42
2001
Cover shows buried cultural deposits at GT31 and artifacts from various sites in the
survey area.
This publication, printed by the Oklahoma Geological Survey, is issued by the University
of Oklahoma and is authorized by the Director of the Oklahoma Archeological Survey.
One hundred copies of this publication were prepared at a cost of $532.00 to the
taxpayers of the State of Oklahoma.
ARCHEOLOGICAL SURVEY OF THE SALT FORK OF
THE ARKANSAS RIVER BASIN IN NORTH CENTRAL OKLAHOMA
RICHARD R. DRASS
WITH A CONTRIBUTION BY
SUSAN C. VEHIK
The University of Oklahoma
Oklahoma Archeological Survey
Archeological Resource Survey Report No. 42
2001
Oklahoma Archeological Survey
111 E. Chesapeake
Norman, Oklahoma 73019-0575
Copyright 2001
by
Oklahoma Archeological Survey
Norman, Oklahoma 73019
II
ACKNOWLEDGEMENTS
This project was possible through the assistance of many people especially the landowners in Grant
and Kay counties who gave us permission to survey on their land. The project was undertaken with
funding provided by the Oklahoma State Historic Preservation Office, the National Park Service,
and the Oklahoma Archeological Survey at the University of Oklahoma. Scott Brosowske assisted
with most survey work and some of the analysis. Students, Casey Carmichael, Jessie Ballenger, and
Robert Stokes, helped at various times during the fieldwork. A special thanks goes to three
volunteers. Charles Sanders, Mick Sullivan, and Mary Ann Drass, who helped with the survey of
several locations in the study area.
III
ABSTRACT
A pedestrian survey of portions of the Salt Fork of the Arkansas River basin in Grant and Kay
counties was undertaken in the winter and spring of 1999/2000. The project examined 4.1 square
miles (2622 acres) documenting 40 occupations at 36 sites. The survey included settings along Salt
Fork terraces, terraces along several tributaries, and upland settings near the river and smaller
streams. Research centered on the river basin in western Kay and eastern Grant counties. areas that
had very little previous archeological work. The project was designed to provide a preliminary
evaluation of the number and types of archeological resources in the central Salt Fork basin and
present recommendations for determining the National Register potential of the sites. Information
from the survey was not sufficient to determine the eligibility of any site for the National Register
and testing is recommended for many of the sites to determine their significance. Twenty-seven
prehistoric occupations have been identified and these range in age from Middle Archaic. Calf
Creek, to Late Prehistoric. Late Archaic/Woodland and Late Prehistoric camps or base camps appear
to be most common, but other sites may be deeply buried on stream terraces. Buried soils were noted
throughout the study area and one cultural deposit buried about 3 meters was dated to the Late
Prehistoric period. Prehistoric people of this area relied heavily on Florence-A cherts found in the
Arkansas River area to the east. The somewhat low density of artifacts at most sites and the
relatively high proportion of tools to debris may indicate that groups from the Arkansas River basin
were using this area seasonally or periodically rather than establishing long-term camps and villages.
This suggestion, however, needs further research to evaluate. Historic sites consist of 13 farmsteads
and 1 cemetery, most dating to the period after homesteading, after 1893, and many farmhouses were
probably abandoned by 1950. One historic site and a dugout may indicate earlier historic activity.
but the few recovered artifacts were not sufficient to accurately date the occupations. Many of the
farmsteads are destroyed or heavily disturbed by farming activities. In contrast, many of the
prehistoric sites appear to be in good condition and some may be buried with very little disturbance
from modern activities.
IV
TABLE OF CONTENTS
ACKNOWLEDGEMENTS iii
ABSTRACT iv
LIST OF FIGURES viii
LIST OF TABLES x
INTRODUCTION 1
Environmental Setting 3
Geology and Physiography 3
Soils 4
Climate 5
Flora and Fauna 5
Paleoenvironment 6
Cultural Setting 6
Paleoindian 7
Archaic 13
Plains Woodland or Early Ceramic 17
Late Prehistoric or Middle Ceramic 20
Proto historic or Late Ceramic 24
Historic 26
RESEARCH DESIGN AND METHODOLOGy 28
Methodology 30
SURVEY FINDINGS 35
Historic Sites in the Middle Salt Fork Basin 35
Prehistoric Sites in the Middle Salt Fork Basin 37
Village/Base Camps 42
Base Camps 44
Camps 44
Isolated Finds 45
Chronology and Site Distribution 45
Archaic/Woodland 45
Late Prehistoric 47
Summary 48
Lithic Material Use in the Middle Salt Fork Basin 48
Prehistoric Use of Florence-A Chert in the Salt Creek and Salt Fork Valleys,
by Susan C. Vehik 53
Introduction 53
v
Theoretical Considerations 56
Archaic Period 57
Woodland Period 57
Late Prehistoric Period 60
Florence-A Chert Usage in the Salt Creek Valley 64
Late Archaic Period 64
Woodland I Period 65
Woodland IIIIl1 Period 66
Late Prehistoric Period 66
Florence-A Chert Usage in the Salt Fork Valley 66
Archaic Period 67
Woodland IIIIII Period 67
Late Prehistoric Period 67
Discussion 68
Conclusions 70
Theoretical 70
Culture-Historical 71
General 71
Summary of Survey Findings 71
SUMMARY AND RECOMMENDATIONS 73
REFERENCES 78
APPENDIX A, DESCRIPTIONS OF SITES AND ISOLATED FINDS RECORDED
DURING THE SALT FORK OF THE ARKANSAS RIVER SURVEy 90
GT3-Porter 90
GT4-Hobbs 98
GT5-Spoon 99
GT8 102
GT9-Stalker. 105
GT31-Tebow Cut 108
GT32-Frank Kirby 112
GT33- Whitehead 114
GT34-Porter House 116
GT35-Kirby House 116
GT36-Mary Ann Smith 117
GT37-Tebow #2 118
GT38-Johnston Store 120
GT39-Friends Cemetery 120
GT40-Breckenridge 121
GT41-Zeloudek Historic 122
GT42-Joe Reynolds Historic 122
VI
GT43-Breckenridge Historic ]23
GT44-McGimmsey Historic 123
KA160 124
KA402- Tautfest 124
KA403-Pump House Site 125
KA404-Possum Site 125
KA405-Gary Wood 126
KA406-Walker 127
KA407-Wood Dugout 127
KA408-Wood #2 128
KA409-Wood Garden 128
KA410-Young#1 129
KA411-Young #2 131
KA412-Young#3 132
KA413-Young#4 132
KA414-Young #5 134
KA415- Young Historic 134
KA416-Stafford Historic 135
KA417-Fruit Historic 135
Isolated Finds 136
GTO/2-Mueggy IF#2 136
GTO/3-Kirby IF#1 136
GTO/4-Whitehead IF#2 137
GTO/5-Kirby IF#2 137
GTO/6-Hobbs IF#1 137
GTOI7-Dowell IF#1 137
GTO/8-Tebow IF#I 137
GTO/9-Knife IF#1 138
GTOIlO-Whitehead IF#I 138
GTOIlI-Mueggy IF#1 138
GTOll2-Mueggy IF#3 139
GTOll3-Breckenridge IF#1 139
GT01I4-Hobbs IF#2 140
GTO/15-Zeloudek IF#l 140
KAO/34-Stafford IF# 1 140
KAO/35-Woods IF#1 141
KAO/36-YoungIF#I 141
KAO/37-Young IF#2 141
KAO/38-Young IF#3 141
vii
LIST OF FIGURES
FIGURE 1: Map of the Project Area along the Salt Fork of the Arkansas River Basin ... 2
FIGURE 2: Salt Fork of the Arkansas River Basin in Northern Oklahoma 2
FIGURE 3: Plains Village Sites and Complexes 21
FIGURE 4: Surveyed Areas in the Salt Fork of the Arkansas River Basin 31
FIGURE 5: Salt Fork of the Arkansas River Research Area 54
FIGURE 6 Surface Geology of North-Central Oklahoma and South-Central Kansas ... 55
FIGURE 7: Artifacts Collected from GT3; a) Williams Point Base, b) Modified Flake Scraper,
c) End Scraper, and d) Arrow Point Fragment 91
FIGURE 8: Cordmarked Rim Sherds from Patterson Collection at 34GT3 94
FIGURE 9: Rim Profiles for Pottery from GT3A 95
FIGURE 10: Smoothed Rim Sherds and Repair Holes in Cordmarked Sherds from Patterson
Collection at GT3A 96
FIGURE 11: Point Fragment from GT4B 99
FIGURE 12: Artifacts Found at GT5, a) Fresno Point Fragment, b) End Scraper Fragment.
c) End Scraper, and d) Scraper/Knife 102
FIGURE 13: Part of Patterson Collection from GT8 104
FIGURE 14: Points Found at GT8, a) Fresno, b & c) Scallorn, and d) Calf Creek Barb
....................................................................................................... 104
FIGURE 15: Topographic Map ofGT9 and Grid Map of Excavations in Area B 106
FIGURE 16: GT9 Artifacts. Manos and Grooved Axe from the Reynolds Collection; Arrow
Points from Grid B excavations; and Sherds from Grid B.. . .. .. .. .. ... .. ... . ... .. .. . ... ... 107
FIGURE 17: Exposed Bank of Terrace at GT31 with Bone and Flakes at Base of Cut 109
FIGURE 18: Profile of Exposed Bank at GT31 110
VIII
FIGURE 19: Scrapers from GT31 112
FIGURE 20: Coring Truck at GT32 113
FIGURE 21: Artifacts from GT32 and GT33. a) Mano, GT32, b) Fresno Point Fragment,
GT32, and c) Knife Fragment, GT33 115
FIGURE 22: Looking Southeast at Terrace and GT36 118
FIGURE 23: Projectile Points and an End Scraper Fragment from GT36 119
FIGURE 24: Scraper from GT37 119
FIGURE 25: Biface and Drill Fragments Found at KA405 127
FIGURE 26: Artifacts from KA410 130
FIGURE 27: Scraper from KA411 131
FIGURE 28: Sherd from KA412 132
FIGURE 29: Morhiss-like Point from KA413 133
FIGURE 30: Beveled Knife Fragment and Bullet from GTO/9 138
FIGURE 31: Biface knife from GTOlll. 139
ix
x
LIST OF TABLES
TABLE 1: Previously Reported Sites in the Salt Fork Study Area 8
TABLE 2: Attributes of Historic Sites in the Salt Fork Survey Area 36
TABLE 3: Attributes of Prehistoric Sites in the Salt Fork Survey Area 38
TABLE 4: Chipped Stone Tools Found during the Salt Fork of the Arkansas River Survey
........................................................................................................ 39
TABLE 5: Temper and thickness for cordmarked and smoothed pottery from Salt Fork
sites 41
TABLE 6: Percentages of Lithic Material Types Found at Salt Fork Sites 50
TABLE 7: Counts of Lithic Materials for All Chipped Stone Found During the Salt Fork
Survey 51
TABLE 8: Chipped Stone Items from the Patterson Collection at 34GT3A 52
TABLE 9: Florence-A Chert and Attributes by Time Period for Salt Creek and Salt Fork
........................................................................................................ 65
TABLE 10: Recommendations for Sites Examined during the Salt Fork Survey 74
TABLE 11: Tool/Debitage Ratios for Salt Fork Sites 77
TABLE 12: Pottery from Roy Patterson's Collection at 34GT3A 93
TABLE 13: Chipped stone tools from the Patterson Collection at 34GT5 101
INTRODUCTION
This report summarizes the findings from the
1999/2000 archeological reconnaissance of
select portions of the Salt Fork of the
Arkansas River basin in Grant and Kay
counties. The project is part of the Oklahoma
Historic Preservation Office and National
Park Service efforts to increase our record of
cultural resources. The purpose of the survey
is to provide an initial evaluation of the types
of archeological resources and their
distribution in the Salt Fork basin in north
central Oklahoma. This information provides
an initial assessment of the National Register
eligibility of sites and will facilitate cultural
resource management decisions in the state.
The designated project area encompasses the
Salt Fork of the Arkansas River basin in Kay
and Grant counties (Figure 1). There are 147
recorded sites in the Salt Fork basin in these
counties. plus one site each in nearby Garfield
and Noble counties that are also within the
basin. Most of these sites, however, are found
in Kay County along tributaries of the Salt
Fork. Much of the rest of the Salt Fork basin
has never been intensively surveyed. A 1976
archeological survey included areas of the Salt
Fork in western Grant County (Ferring et al.
1976). but this work did not locate any sites
along this portion of the river. There has been
limited investigation of 34GT9, a Woodland
or Late Prehistoric camp near Pond Creek, but
there is no published data on this site.
Work in the Kaw Lake area of the Arkansas
River basin just east of the study area has
revealed sites ranging from Archaic to Plains
Village, but Woodland and Archaic period
camps are the most common. Prehistoric
quarries for Florence-A chert are found just
east and north of Kaw Lake. Farther east in
the Salt Creek valley investigations have
revealed some Late Archaic camps, many
Woodland base camps and villages, and Late
Prehistoric base camps (Vehik 1985b). Little
work has been done to the west in Alfalfa and
Woods counties and the prehistoric
occupation of the Salt Fork basin, and other
stream basins, is poorly documented in those
counties.
Based on the distribution of known sites in the
basin, the areas selected for field survey were
confined to a smaller section of the basin. The
project involved a sample survey, about four
square miles, in portions of the Salt Fork of
the Arkansas River basin roughly between
Tonkawa on the east and Nash near the
Alfalfa/Grant county line on the west (Figure
1). Emphasis was on terraces and uplands
bordering the Salt Fork and select tributaries
close to the river. These locations in Grant and
western Kay counties not only were poorly
documented in the archeological record, but
they also provided a cross section of settings
(e.g. first and second river and tributary
terraces, high river terraces, and ridges) to
identify prehistoric site distributions in this
region,
The survey location is also about midway
between the Great Salt Plains in Alfalfa
County and the chert resources of the Flint
Hills in eastern Kay and western Osage
counties (Figure 2). The selected Salt Fork
survey tracts permit evaluation of prehistoric
site distributions west of the better-studied
Arkansas River area, and, potentially, they
could provide evidence to determine if
different types of occupations occurred in
GRANT CO.
"'ash
Ir----- II
KAY CO.
543210
I W L;H
10 15
I
Figure 1: Map of the project area along the Salt Fork of the Arkansas River basin.
SCAlE Of MILES
Turkty C
"~
10 20 30 Miles
ec 4O~:l Kilometers
Figure 2: Salt Fork of the Arkansas River Basin in northern Oklahoma.
these western prairie settings. The survey also
could provide some information on the extent
of the use of Florence-A chert west of the
Flint Hills quarries and outcrops, a significant
resource for prehistoric groups in north central
Oklahoma. In general, the survey of the Salt
Fork of the Arkansas River basin in Grant and
Kay counties provides a systematic sample of
archeological resources and an initial
evaluation of prehistoric use of this north
central Oklahoma area.
2
ENVIRONMENTAL SETTING
Geology and Physiography
The Salt Fork of the Arkansas River flows
from Kansas south and east entering
Oklahoma in northern Woods County and
emptying into the Arkansas River in eastern
Kay County. This river drains 4755 square
miles (12,315 square kilometers, not including
the Chikaskia River and Bluff Creek) with
2549 square miles (6,602 square kilometers)
in Oklahoma (Oklahoma Water Resources
Board [OWRB] 1987). The Salt Fork cuts
through the red Permian shales and sandstones
of the Central Redbed Plains in Oklahoma,
but parts of Alfalfa and Grant counties have
large sand dune belts (the Western Sand Dune
Belts) along the river (Curtis and Ham 1972).
The most distinctive feature along the river is
the Great Salt Plains in Alfalfa County. This
natural concentration of salt was a major
resource for various Indian groups and early
settlers (see Ferring 1976). There are salt
deposits along the river in Grant County but
much less salt is found as you move east
toward the confluence of the Arkansas and
Salt Fork of the Arkansas rivers. The river,
however, is considered too saline today to use
as drinking water (Culver 1967:84). The Salt
Fork flows through gently rolling plains with
broad valleys. Most tributary streams dissect
the upland hills in relatively small valleys.
Principal tributaries of the Salt Fork of the
Arkansas River in the survey areas include
Pond, Deer, Boggy, Crooked, Coldwater, and
Wild Horse creeks. Of these only Coldwater
and Wild Horse creeks enter from the south
and they are both in western Grant County.
Wild Horse Creek has Quaternary deposits of
sand, gravel, and clay with both low terraces
and a high terrace. This creek has flooded
several times during the 20th century. Pond
~ I
Creek and Deer Creek are the principal
drainages on the north side of the Salt Fork in
the survey area. Pond Creek and its tributaries
drain much of central and northern Grant
County, and Deer Creek drains part of eastern
Grant and western Kay counties. Both streams
cut shallow valleys through gently rolling
uplands. Pond Creek extends for miles across
the higher terraces of the Salt Fork. In
contrast, Deer Creek flows through mostly
uplands before entering the Salt Fork terrace
system only a couple of miles from its
junction with the river.
The bedrock formations in the study area
consist of "flat-lying redbeds of clay shales,
siltstone and fine-grained sandstones of
Permian age" (OWRB 1987:3). Pleistocene
terraces and sand dunes cover much of the
bedrock near the river. The river currently
flows on alluvium that may be 10 to 50 feet
deep (Johnson 1980). In eastern Grant and
western Kay counties the sand dunes decrease
in frequency and appear to be more common
on the north side of the Salt Fork. The dunes
in western Grant County are typically
stabilized by grass and brush, but some are
active and sand from the flood plain is still
blown onto the dunes and terraces. The
uplands are gently rolling and cut by small
streams. In Grant County, most of the small
streams drain into the Salt Fork of the
Arkansas River or into Bluff Creek or the
Chikaskia River, which are also tributaries of
the Salt Fork. Larger tributaries are
predominantly north of the Salt Fork in Grant
County. Some of these have well developed
terraces, but streams in the sandy dune areas
in the western portions have smaller bottoms
that are often sand-clogged. Loamy and clayey
terraces are present in eastern Grant and Kay
counties, although sandy soils are frequently
present near the Salt Fork.
3
Stone suitable for prehistoric tool manufacture
is not common in the Salt Fork basin.
Sandstones and limestones that could be used
for grinding implements outcrop in some
upland setting, but often these settings are
some distance from the river and major
streams. Some cherts and quartzites suitable
for chipped stone tool manufacture are found
locally, occurring in the survey area within the
Pliocene Ogallala gravel outwash from the
Rocky Mountains. These Ogallala gravels
were created as a result of the uplifting of the
Rocky Mountains 60-65 million years ago,
which formed the Pliocene-aged plains of
southern Kansas, Oklahoma, and Texas from
clays, silts, sands, and gravels sloping off and
eroding out of the Rockies. Subsequent
erosional episodes and formation of major
river drainages have spread these gravels
southeast across Oklahoma (Banks 1984).
Many of the unidentified cherts found at
archeological sites. no doubt, are derived from
these transplanted gravels (Banks 1990). In
the surveyed areas, these unidentified cherts
and Ogallala quartzite gravels are not
abundant. and. generally, the cobbles are
small with most smaller than fist size. Sand
dunes near the river and larger streams may
cover Ogallala gravels, even on some higher
ridges and high terraces.
Soils
Soils in the study area have developed
primarily from weathered sandstones and
shales. or sandy, clayey, and loamy alluviums.
Soil associations reflect surface soils,
subsurface soils, and the unconsolidated, or
parent material, in which the soils formed.
The Grant and Kay County soil surveys use
different terminology, but the principal soil
associations in the surveyed areas can be
compared. The Grant County Soil Survey is
the more recent and terms will be used from
this source and compared with the Kay
County data. Primarily two soil associations.
McLain-Dale-Hawley (roughly equivalent to
Kaw-Brewer-Reinach-Lela in Kay County)
and Pond Creek-Bethany-Grant (equivalent to
Norge-Vanoss in Kay County), are
represented in the survey areas. The Kirkland-
Tabler association (= Kirkland-Tabler-
Bethany in Kay County) is common in the
uplands of the study area, but it is not
included in many surveyed settings. Almost
none of the sandy/loamy dune settings along
the Salt Fork in western Grant County were
included in the survey. These sandy soils are
found on the north side of the Salt Fork and
make up only about 10% of the soils in Grant
County.
The McLain-Dale-Hawley association and its
equivalent represent about 16-17% of the
Grant and Kay County soils (Culver 1967;
Williams et al. 1985). These soils are found on
the nearly level flood plain terraces of the Salt
Fork and major streams in the study area. The
soils have a loamy surface layer and clayey or
loamy subsoils. Terraces occasionally flood,
but these soils are highly productive for
farming (generally wheat and sorghums) and
as pastures. The majority of the areas
surveyed consist of soils in this association.
The Pond Creek-Bethany-Grant association
consists of well drained loamy soils on level
to gently sloping high terraces and gently to
steeply sloping uplands. These soils are
generally adjacent to the McLain-Dale-
Hawley association, but they are usually
higher above and farther from the streams.
This association makes up 16% of the soils in
Grant County, and its equivalent is about 13%
of the soils in Kay County (Culver 1967;
4
Williams et al. 1985). These soils are
productive for wheat cultivation in level areas
and they make good grass pastures in gently
sloping areas. The majority of upland settings
examined during the survey are composed of
soils from this association with a few areas
including Kirkland-Tabler association soils.
Soils of the latter association are found on
much of the broad upland plains in Grant and
western Kay counties. The association covers
a large area of the former upland prairies in
these counties, representing 33-38% of the
soils in the counties. These soils have a loamy
surface layer and clayey subsoils. Small, often
intermittent, streams drain the Kirkland-
Tabler association, and much of this
association is outside of surveyed areas.
Climate
The Salt Fork of the Arkansas River basin lies
between two climatic zones, the warm
summer temperate continental climatic zone
on the north and the subtropical humid
climatic zone to the south (OWSB 1987:4).
Mild winters and warm summers characterize
the area, but temperatures can change rapidly.
The average daily temperature in Grant
County is 59.8E F with winter temperatures
averaging 38E and 81E in the summer. Daily
temperatures can vary significantly with the
coldest recorded temperature of -9E in
January and the hottest recorded temperature
of 115E in July (Williams et al. 1985:2).
Summer temperatures exceeding 100E Fare
common. The growing season is about 195-
210 days with the last freeze in spring around
April 5 and the first freeze in fall around
October 24.
Prevailing winds are from the south with an
average as much as 14 miles per hour in the
spring (Williams et al. 1985:2). Annual
precipitation is 30-32 inches (76.2-81.3 em),
but the amount can vary widely from year to
year (records indicate a range of 22.3 to 37.2
inches [56.6-94.5 em] per year in Grant
County). The majority (70%) of the
precipitation falls during the growing season
between April and September (Williams et al.
1985:2). Thunderstorms (occasionally
containing tornadoes) occur frequently during
the summer, providing some locales
significant moisture. Rainfall patterns can be
spotty, however, and midsummer droughts are
normal with some periods of longer droughts
lasting several months. Evaporation is also a
problem in the dry, hot, and windy summers.
Snowfall averages 8 inches (20.3 em) and
usually it does not remain on the ground
longer than a few days.
Flora and Fauna
The native vegetation in the study area was
once a mixture of mid and tall grass prairies
with woodlands restricted to areas along the
Salt Fork and its tributaries. Much of the area
has been plowed or planted to pasture since
Euro-American settlement in 1893. Williams
et al. (1985: 51) estimates that 20% of Grant
County is rangeland where native vegetation
is used for grazing. Prairies are composed
predominantly of bluestem, Indian grass, and
switchgrass. Other common plants include
sideoats and blue gram a, dropseed, panicum,
leadplant, and sunflowers. A sand-sage
grassland is found on the dunes in western
Grant County. These dunes have a cover of
sand sage and midgrasses such as little
bluestem, sand bluestem, and hairy grama
(Risser 1974). The river valleys once
consisted of small mesic forests containing
cottonwood, American elm, green ash,
hackberry, black willow, blackjack oak, and
red mulberry along with other trees, shrubs,
and grasses.
5
The prairies and flood plain forests of Grant
and Kay counties support an abundance of
animals that can be hunted and used for food.
The forested areas support white-tailed deer,
wild turkey, cottontail rabbits, raccoons,
opossums, foxes, squirrels, quail, doves,
grouse, box turtles, and other animals. More
aquatic settings may have animals such as
otter, beaver, swamp rabbits, ducks, geese,
cranes, and other birds. The major fauna of the
grasslands would include bison before herds
were eliminated in the late 1800s. It is likely
that bison were also drawn to the bottomlands
for winter shelter and dependable water. Open
grasslands also offered grazing for deer and
antelope, hunting for hawks, and grains for
prairie chickens, quail, and doves. Jackrabbits,
ground squirrels, box turtles, and a number of
both poisonous and non-poisonous snakes can
be obtained in the prairies. Fish, mussels, and
various turtle can be found in the river and its
tributaries. In general prehistoric people
would have found an area adequate for
sustaining a foraging, mobile lifestyle. The
rich soils near the streams may also have
supported later village dwellers who
cultivated the sandy/loamy terraces along the
Salt Fork of the Arkansas River and its
tributaries. These areas are today cultivated
for wheat and other crops, although the
farmland now also extends onto the level to
rolling uplands.
Paleoenvironment
There is very little paleoenvironmental
information for the specific study area in north
central Oklahoma. The most detailed
information is to the east from the Cross
Timbers settings near the Caney River (Artz
and Reid 1984; Hall 1977a, b, & c; Vehik et
al. 1979). Hall (1977 a, b, & c) and Vehik
(l985c) provide summaries of general
paleocology over the past 2000 years. Mandel
(1987, 1992, 1994) has also studied Holocene
landscape formation in south central Kansas,
providing environmental data and information
on site stratigraphy and terrace formation in
the Walnut River basin and other areas. These
studies provide some evidence for changing
climatic and erosional/depositional sequences
in the area over the past 10,000 years. In
general, there would have been a fluctuation
in the location and composition of specific
biotic communities in the Southern Plains
related to changing conditions. The specifics
for the study area are not known and the
reader is referred to Vehik et al. 1979. Mandel
1987 and 1992, and Artz and Reid 1984 for
general information on past conditions in the
region,
CULTURAL SETTING
The north central area of the state contains
evidence of prehistoric occupation over the
past 10,000 years. from the Paleoindian
through the Late Prehistoric periods. The
earliest historic Wichita sites are also found in
this region. Ferdinandina or Deer Creek and
Bryson-Paddock were contacted by the French
who established trading relationships in the
1740s. These Wichita groups moved south to
the Red River in the late 1750s. Other groups
such as the Osage continued to use this area.
but it was not intensively occupied until the
area became part of the Cherokee Outlet and
was rented to cattle ranchers. The Ponca,
Tonkawa, and Otoe/Missouri were settled
west of the Arkansas River and the Outlet was
purchased in 1892 and opened for settlement
by a land run in 1893.
Most of the Salt Fork basin has never been
intensively surveyed for archeological sites.
6
Reported sites in the Salt Fork of the Arkansas
River basin range from the historic 101 Ranch
to Middle Archaic camps related to the Calf
Creek culture (Table 1). Surface collections
from 34GT6, in the study area, have revealed
evidence of a Calf Creek camp, possibly
dating back about SOOO years (Brooks 1995).
Much of the archeological work, however, has
been to the east in the Kaw Lake area of the
Arkansas River basin (see Galm 1979; Hartley
and Miller 1977; Rohrbaugh 1973, 1974; and
Vehik and Flynn 1982). Sites in this area
range from Archaic to Plains Village, but
Woodland and Archaic period camps are the
most common. Quarries for Florence-A chert
are found just east and north of Kaw Lake.
Farther east in the Salt Creek valley
investigations have revealed some Late
Archaic camps, many Woodland base camps
and villages, and Late Prehistoric base camps
(Vehik 1985b).
To the north in Kansas there has been
extensive work on the Walnut Creek area near
its junction with the Arkansas River (see
Hawley 1994). Most of this research has dealt
with Late Prehistoric and protohistoric
Wichita sites. Closer to the study area, there is
some information on sites along Bluff Creek,
a tributary of the Chikaskia River and
ultimately the Salt Fork of the Arkansas
River. The Bluff Creek complex is a Late
Prehistoric complex, but it is currently poorly
documented (Brown 1987c). To the west there
has been little research in northern Oklahoma
or southern Kansas. The Zimms complex, a
Late Prehistoric manifestation, is reported in
the northwest part of Oklahoma, but only one
site (Hedding, 34WDS) has been analyzed in
this area (Drass 1989). The Wilmore complex
is recorded in southwest Kansas and it has
recently been reexamined and redefined to a
Late Prehistoric variant, the Plains Border
variant (Bevitt 1999). Collections from two
Late Prehistoric sites in Woods County,
northwestern Oklahoma, have been examined
but not attributed to an archeological complex
(Drass 1999).
Oklahoma Archeological Survey site forms
indicate that many time periods are
represented at sites in the Salt Fork of the
Arkansas and Arkansas River basins. Isolated
projectile points and bones of mammoth,
mastodon, and extinct bison are evidence of
human occupation at least 10,000 years ago.
In addition, limited investigation of a few
upland and bottomland sites has revealed dart
points and early pottery that were used by
groups dating from over SOOO B.C. to A.D.
1600. Thus, the Salt Fork River valley and
surrounding uplands contain evidence of
almost continual use by people over the past
10,000 to 12,000 years. Based on the limited
information from excavations and surveys and
with data from sites in the region, we can
sketch a chronology and cultural-historic
framework for the region. A brief review of
the cultural historical sequence is discussed
below.
Paleoindian
There is currently a debate among
archeologists on the earliest inhabitants of the
Americas and their route of entry onto this
continent. The traditional view has been that
the earliest occupants were the Clovis people
who crossed over the land bridge from Asia
by about 12,000 years ago. Sites such as the
Meadowcroft Rockshelter in Pennsylvania,
however, have evidence of occupations dating
from 12,000 to possibly over 14,000 years
ago. The Monte Verde site in Chile has been
dated to about 11,000 to 10,SOO B.C.
7
Table 1: Previously Reported Sites in the Salt Fork Study Area.
Site # Time Period Site Type Recorder/Date Setting* Soil**
Ponca City SE
KA22 Archaic/Late Prehistoric Village? Hiatt 12-13-60 T V
KA330 Unid. Prehistoric/Historic Camp/Scatter Hughes-Jones 91 T V
KA333 Historic Farmstead Briscoe 6-16-91 T N
NB65 Historic Farmstead Briscoe 6-16-91 T LU
Ponca City (Salt Fork area only)
KA24 Woodland/Late Prehistoric Village? Slovacek 1-12-63 UP N-KD
KA27 Woodland/Late Prehistoric Village Slovacek 2-24-61 UP N-KD
KA45 Unid. Prehistoric Camp Slovacek 4-28-62 T KD-R-V
KA46 Woodland/Late Prehistoric Camp Slovacek 4-28-62, T Kaw
Vehik 1-12-90
KA47 Late Prehistoric Camp Slovacek 4-28-62, T Kaw
Vehik 1-12-90
KAI35 Archaic?/Unid. Prehistoric Camp Hiatt 3-4-66 T BR
KAI37 Late Prehistoric Village Wyckoff 2-11-67 T Kaw-
BR
KA321 Late Preh istoric/H istoric Village/- Sudbury 10-21-89 T V
Farmstead.
KA322 Unid. Prehistoric Camp Vehik 1-12-90 T/FP Kaw
KA323 Unid. Prehistoric Camp Vehik 1-12-90 T Kaw
KA324 Unid. Prehistoric Camp Vehik 1-12-90 HT N
KA331 Unid. Prehistoric Camp Hughes 2-23-91 T V
KA332 Historic Farmstead Briscoe 4-23-91 T V
KA334 Late Prehistoric Village Sullivan 9-15-91 T N
KA335 Late Prehistoric Camp Sullivan 11-23-91 T N
KA352 Late Prehistoric Village? Sullivan 5-16-91 T R-KD
KA353 Late Prehistoric? Camp Sullivan 5-17-91 HT N
Ponca City NW
KA31 Late Archaic/Woodland Camp Hiatt 9-8-61 T N-V
KA34 Archaic/Late Preh istoric? Camp? Hiatt 9-8-61 UP V
KAI38 Late Archaic Camp Wyckoff 2-1 1-61 UP N-KD
KAI39 Late Prehistoric Village Sudbury 9-8-67 HT V
KAI40 Unid. Prehistoric (Late Camp Sudbury 10-9-67 UP KD-Preh
istoric?) Kaw
KAI41 Late Prehistoric Camp? Sudbury 10-9-67 UP KD-T
KA222 Late Archaic/Woodland Camp/Scatter Wallis 3-12-80 UP N
KA223 Historic Farmstead Wallis 3-12-80 T N
KA224 Late Prehistoric Camp/House Wallis 3-12-80 UP N
KA225 Un id. Preh istoric/H istoric Camp/Scatter Wallis 3-13-80 UP N
KA226 Late Archaic/Woodland Camp Wallis 3-13-80 T N-Kaw
KA227 Unid. Prehistoric Camp Wallis 3-13-80 UP KD
KA228 Late Preh istoric Village? Wallis 3-13-80 UP T
KA304 Unid. Prehistoric Camp Hughes 8-31-85 T N
KA320 Historic Town ? T L
KA325 Late Archaic/Late Vi llage/Scatter Sudbury, Sullivan T V/N
8
Site # Time Period Site Type Recorder/Date Setting* Soil**
Prehistoric/H istoric
KA375 Historic Farmstead Briscoe 3-26-97 UP N
KA376 Historic Farmstead Briscoe 3-20-97 UP W
KA377 Historic Farmstead Briscoe 2-20-97 UP B
KA378 Historic Farmstead Briscoe 3-19-97 UP B
KA379 Historic Farmstead Briscoe 3-26-97 UP K-R
KA393 Historic Dump Briscoe 3-24-97 UP R
KA397 Historic Farmstead Northcutt 3-6-97 UP V
Newkirk SW (Duck Creek & Bois D' Arc Creek)
KA38 Woodland/Late Prehistoric Camp Hiatt 11-18-61, T N
Wallis 3-14-80
KAI46 Protoh istoric Camp Neal/Sudbury UP KD
KA211 Unid. Prehistoric Camp Wallis 2-19-80 UP KD
KA212 Historic House site Wallis 2-19-80 UP R-KD
KA213 Historic House site Wallis 2-20-80 UP N
KA217 Unid. Prehistoric/Historic House site Wallis 2-21-80 UP KD
KA219 Historic Farmstead Wallis 3-11-80 UP N
KA220 Early Archaic Camp Wallis 3-11-80 UP KD-R
KA221 Unid. Prehistoric/Historic Camp/Scatter Wallis 3-11-80 T/UP N-Kaw
KA229 Late Archaic/Woodland Camp Wallis 3-14-80 UP KD-R
KA230 Unid. Prehistoric/Historic Camp/Dump Wallis 3-14-80 UP KD-B
KA231 Archaic/Woodland Camp Wallis 3-14-80 UP N-Kaw
KA232 Unid. Prehistoric/Historic Camp/Farmstead Wall is 3-14-80 UP N
KA380 Historic Farmstead Briscoe 3-19-97 UP KD
KA381 Historic Farmstead Briscoe 3-19-97 UP Kaw
KA382 Historic Town Briscoe 3-20-97 UP T
KA383 Historic Farmstead Briscoe 3-20-97 UP B
KA384 Unid. Prehistoric/Historic Farmstead Briscoe 3-20-97 UP KD
KA385 Unid. Prehistoric/Historic Farmstead Briscoe 3-20-97 UP KD
KA386 Historic Farmstead Briscoe 3-20-97 UP T
Peckham (Duck Creek)
KA214 Unid. Prehistoric Camp Wallis 2-20-80 UP KD-R
KA215 Unid. Prehistoric Camp Wallis 2-20-80 UP N
KA216 Unid. Prehistoric Camp Wallis 2-21-80 UP KD-R
KA218 Historic House site Wallis 2-20-80 UP KD-R
KA233 Unid. Prehistoric/Historic Camp/Scatter Wallis 3-14-80 UP KD-R
KA387 Unid. Preh istoric/H istoric Farmstead Briscoe 3-20-97 UP Kaw
KA388 Historic Farmstead Briscoe 3-20-97 UP T
KA389 Historic Farmstead Briscoe 3-20-97 UP KD
KA390 Historic Farmstead Briscoe 3-20-99 UP KD
KA391 Unid. Prehistoric/Historic Camp/Farmstead Briscoe 3-20-99 T R-KD
KA392 Historic Farmstead Briscoe 3-20-99 UP R-KD
KA394 Unid. Prehistoric Camp Briscoe 3-29-97 T KD
KA395 Late Prehistoric/- Camp Briscoe 3-24-97 T/UP 0
Protohistoric
Braman
KA312 Archaic-Late Prehistoric Camp? Gettys 7-23-87 T BR
KA313 Woodland, Late Prehistoric Camp? Gettys 7-23-87 T RE
9
Site # Time Period Site Type Recorder/Date Setting* Soil**
KA3\4 Woodland, Late Prehistoric Camp? Gettys 7-23-87 T RE-BR
KA3l5 Archaic-Late Prehistoric Camp? Gettys 7-23-87 T BR
KA316 Unid. Prehistoric Camp? Gettys 7-23-87 T BR
KA348 Late Prehistoric Camp? Sullivan 3-1-92 T/UP N
KA349 Unid. Prehistoric Camp Sullivan 3-\-92 T/UP N
KA350 Archaic, Calf Creek? Camp Sullivan 2-27-92 T/UP N
KA366 Archaic-Late Prehistoric Village/burials Duncan 7-5-95 T Kaw
Blackwell
KA6 Unid. Prehistoric? Camp? Bareis 3-2-53 T? RE
KA326 Unid. Prehistoric/Historic Camp/Farmstead Vehik 8-22-90 T RE
KA341 Archaic Camp Sullivan 2-20-92 T RE
KA342 Unid. Prehistoric Camp Sullivan 2-20-92 T BR
KA343 Unid. Prehistoric Camp Sullivan 2-20-92 T RE
KA344 Archaic Camp Sullivan 2-20-92 T RE
KA345 Late Prehistoric Camp Sullivan 2-20-92 T RE
KA36\ Unid. Prehistoric Camp Vehik 12-2-92 T RE
Blackwell NW
KA336 Late Prehistoric Camp? Sullivan 11-23-91 T P
KA337 Unid. Prehistoric Camp Sullivan \\-23-91 UP/T P
KA338 Woodland, Late Prehistoric Village? Sullivan 1-3-92 T RE
KA339 Late Prehistoric Camp? Sullivan 1-9-92 T P
KA340 Archaic, Late Prehistoric Camp Sullivan 1-23-92 HT BR
KA358 Late Preh istoric Camp? Schermerhorn I I- T KD-R
3-92
GTI7 Late Prehistoric CampNillage Sullivan 12-24-91 T P
GT\8 Late Prehistoric CampNillage Sullivan 12-24-91 T D
GTI9 Unid. Prehistoric Camp Sullivan 12-24-91 T H
GT20 Archaic, Late Prehistoric Village? Sullivan 1-11-92 UP/HT PC
GT21 Archaic, Late Camp Sullivan 1-11-92 HT B
GT22 Archaic, Calf Creek Camp Sullivan 1-23-92 UP/T KD
GT25 Archaic, Calf Creek? Camp Sullivan 2-23-92 HT B
GT26 Late Prehistoric Camp Sullivan 2-27-92 T P
GT27 Late Preh istoric Camp Sullivan 2-27-92 T D
Kildare (Bitter Creek)
KAI Unid. Prehistoric Village? Bareis 2-15-55 T N
KA26 Unid. Prehistoric Camp Slovacek 3-1-61 T Kaw
KA28 Woodland/Late Prehistoric Camp Slovacek 3-6-61 T Kaw
KA43 Woodland/Late Prehistoric? Camp? Slovacek 3-14-61 T Kaw
KA44 Unid. Prehistoric Camp Slovacek 10-10- T Kaw
61
KA48 Archaic? Camp Slovacek 4-28-62 T KD-R-S
KA66 Unid. Prehistoric Camp Hiatt 3-\8-64 T? Kaw
KAI44 Unid. Prehistoric Camp Lopez 6-18-75 T N
KAl45 Unid. Prehistoric (L. Preh.?) Camp 6-18-75 UP N-Kaw
KA351 Late Prehistoric Camp Sullivan 5-16-92 T Kaw-
KD-R
Kremlin
GF52 Unid. Prehistoric Camp Briscoe 4-21-89 T R
10
I, ,
Site # Time Period Site Type Recorder/Date Setting* Soil**
Tonkawa
KA8 Unid. Prehistoric Camp? Bush 10-1-57 T V
KAI58 Unid. Prehistoric Camp Lopez 5-23-73 UP N-KD
Eddy
KAI59 Unid. Prehistoric Camp Lopez 12-11-75 UP 0
KA160@ Unid. Prehistoric Camp Lopez 12-11-75 T RE
KA311 Late? Archaic Camp Gettys 7-23-87 T RE-P
Marland
KA25 Late Prehistoric Camp Hiatt 2-4-61, HT V
Kraft 10-4-97
KA318 Historic 101 Ranch Briscoe 3-27-97 T y
KA328 Unid. Prehistoric Camp Vehik 8-24-90 T y
KA374 Historic 101 Ranch Briscoe 3-26-97 T RE
Nardin (Deer Creek)
KA303 10,000+ Tripp Elephant Wyckoff 8-2-85 T Kaw
KA308 Late Archaic/Woodland? Camp Gettys 7-23-87 T N
KA309 Unid. Prehistoric Camp Gettys 7-23-87 T N
KA310 Woodland/Late Prehistoric Camp? Gettys 7-23-87 T RE-P
GT7 Woodland/Late Prehistoric Village Patterson 12-6-75 UP M
GT30 Unid. Prehistoric Camp Bartlett 4-14-97 T N
Deer Creek (Deer Creek)
GTII Woodland/Late Prehistoric Camp? Wyckoff 10-28- T D
84
Lamont (Pond/Polecat Creeks) I
GTI Late Prehistoric Village Neal/Patterson 75 T PC
GT2 Woodland/Late Prehistoric Village Neal/Patterson 75 T D
GT3@ Wood land/Late Preh istoric Village Neal/Patterson 75 T D
GT4@ Woodland/Late Prehistoric Village Neal 75 T/UP D-PC
Lamont NW (Pond/Polecat Creeks)
GT5@ Late Prehistoric Village Neal/Patterson 75 T/UP N-PC
GT6 Calf Creek/Late Camp/Village Neal/Patterson 75, UP PC
Archaic/Late Preh istoric Brooks 1995
GT8@ Unid. Prehistoric Camp Patterson 10-19- HT N-PC
79
Pond Creek
GT9@ Woodland/Late Prehistoric Village Rose 10-80, 5-81 T PC
GTI3 Historic Chisholm Trail Briscoe 4-21-89 FP H
GTI4 Historic House site Briscoe 4-21-89 FP RE
GTI5 Historic House site Briscoe 4-21-89 T Y
GTI6 Historic Cemetery/Ranch Briscoe 3-1-91 T H
Medford
GTIO Historic House site Neel 12-4-80 UP KD
GTI2 Unid. Prehistoric Camp Briscoe 4-18-84 T KD
Medford NE
GT23 Late Prehistoric Camp Sullivan 1-30-92 T RE
GT24 Late Prehistoric Camp Sullivan 1-30-92 T PC
GT28 Late Archaic/Woodland Camp Sullivan 3-5-92 T D
GT29 Unid. Prehistoric Camp Sullivan 3-5-92 UP G
II
of this period. Kill sites predominate among
the early Paleoindian sites found on the
Southern Plains and tool assemblages reflect
this with the recovery of primarily butchering
tools and a few knapping or hide working
implements.
* T = terrace, HT = high terrace, FP = floodplain, and UP = upland.
** Soil Series: B = Bethany; BR = Brewer; D =Dale; G = Grainola; H = Hawley; KD = Kirkland; L = Lela; LU =
Lucien; M = McLain; N = Norge; 0 = Owens; P = Port; PC = Pond Creek; R = Renfrow; RE = Reinach; S =
Summit; T = Tabler; V = Vanoss; W = Waurika; Y = Yahola.
@ indicates sites revisited during the Salt Fork Survey.
Sites on creeks = Duck Creek (37); Chikaskia River (I); Bois D'Arc Creek (34); Deer Creek (8); Pond Creek (3);
Polecat Creek (3)
Total prehistoric sites = 121 (6 revisited during the survey)
Total historic sites = 32
Plus I mammoth site
47 Unidentified Prehistoric@ (40.9%)
26 Late Prehistoric@ (22.6%)
13 Woodland/Late Prehistoric@ (I 1.3%)
8 Archaic/Late Prehistoric@ (7.0%)
I Late Prehistoric/Protohistoric (0.9%)
I Protohistoric (0.9%)
8 Late Archaic/Woodland@ (7.0%)
6 Archaic@ (5.2%)
I Middle Archaic (Calf Creek)/Late Archaic/Late Prehistoric (0.9%)
3 Middle Archaic (Calf Creek) (2,6%)
I Early Archaic (0.9%)
Subtotal = I 15 sites (not including revisited sites)
indicating that groups had reached South
America at a very early period. In the
Southern Plains, there is evidence from the
Burnham site in northwestern Oklahoma that
people may have lived here as long ago as
26,000 years (Wyckoff 1989; Wyckoff and
Carter 1994). This age is still in question and
much more research remains before there is
good documententation of occupations older
than Clovis. At present, the earliest, well-documented
inhabitants of the Southern Plains
are the early Paleoindian Clovis and Folsom
groups who inhabited the area from about
11,500 to 10,000 years ago. Conditions were
cooler and wetter during this period and
nomadic bands hunted the large mammals of
the Late Pleistocene. Mammoth and large,
now extinct, forms of bison were hunted and
butchered. The large lanceolate spear points
used by these early hunters characterize sites
Clovis is the earliest complex in the region
with major sites at Blackwater Draw in New
Mexico, Lubbock Lake and Miami in Texas,
and Domebo in Caddo County, Oklahoma
(Sellards 1938; 1952; Wedel 1961; Leonhardy
1966; Holliday et al. 1983; Hester 1972).
These sites represent primarily mammoth kill
and/or processing locations. Fluted Clovis
points are characteristic of this complex and
are usually made of Alibates or Edwards chert
in this area of the Southern Plains. Earlier
people may have butchered a mammoth, the
Cooperton mammoth, excavated in Kiowa
12
County but no chipped stone tools were
recovered (Anderson 1975). No evidence of
Clovis occupation has been documented in the
study area although Hofman and Wyckoff
(1991) report two Clovis points from the
Arkansas River region in Osage County and
eleven are known from farther east on the
river in Tulsa County (Wyckoff and Rippey
1998: 15). A Clovis is also recorded from
Cowley County, Kansas, just north of Kay
County, Oklahoma (Brown and Brown
1987:9-17).
The Folsom complex developed about 11,000
years ago at a time when drying conditions
and possibly over hunting had eliminated the
mammoth. Large, now extinct, bison were still
common and Folsom people continued to
roam the plains hunting these bison. The
fluted Folsom point is distinctive of this
period, and studies of Folsom points found in
Oklahoma indicate a preference for the use of
Edwards chert in their manufacture (Hofman
1993). Two Folsom sites have been excavated
in northwestern Oklahoma. Both are in Harper
County some distance from the study area.
Bement (1999) has investigated a bison kill
containing Folsom points at the Cooper site,
and the Waugh site contains evidence of
Folsom kill and camp locations (Hill and
Hofman 1997). One Folsom point is reported
from Grant County in the study area (Hofman
1993). Another Folsom is documented in Kay
County from Kaw Lake east of the study area
(George 1978), and one is reported from the
Arkansas River in Tulsa County (Wyckoff and
Rippey 1998: 17).
About 10,000 to 8,500 years ago a variety of
late Paleoindian complexes marked by
distinctive unfluted, lanceolate point types
developed on the Southern Plains. Plainview,
. ,
Plano, Milnesand, and Portales are some of
the late Paleoindian complexes defined in this
area (see Thurmond 1991; Wyckoff 1992),
and Dalton groups occupied areas just to the
east (Wyckoff and Rippey 1998). These
people were hunters and gatherers, exploiting
diverse plants and animals. Bison may have
been the principal prey on the Southern Plains
(Wyckoff 1992). Tool assemblages include
grinding stones and a variety of knives and
scrapers. The most distinctive projectile points
include Plainview, Meserve, Milnesand,
Golondrina, Angostura, Agate Basin, Hell
Gap, Eden, Cody, and Scottsbluff on the
Plains (see Thurmond 1990) and Dalton
points in eastern Oklahoma and Kansas.
Artifacts and sites attributed to this period are
more numerous than during previous periods,
probably signifying an increase in populations
on the Southern Plains. No investigated sites
in the study area are attributable to the late
Paleoindian period, although a Plainview
point is reported from a private collection. Just
to the east in Tulsa County, Wyckoff and
Rippey (1998: 17-28) report finding Dalton
points and adzes, Agate Basin spear points,
Plainview and Cody points, and other early
projectile points from the Arkansas River.
Archaic
The Archaic period from roughly 6550 to 50
B.C. is associated with a time of drying and
warming conditions. The changing conditions
contributed to the extinction of the large
Pleistocene animals and people became more
dependent on hunting smaller game and
gathering plants. The Archaic is usually
divided into Early, Middle, and Late periods
associated with changes in technology and
probably subsistence. Few Archaic sites have
been investigated on the Southern Plains and
assemblages are often inferred based on
13
contemporary finds found in diverse settings.
Information is particularly scarce for the Early
and Middle Archaic (about 6550 to 2050
B.c.) on the Southern Plains, and it has been
suggested that warmer and dryer climatic
conditions during the Altithermal or
Hypsithermal may have resulted in the virtual
abandonment of some areas for several
millennia (Dillehay 1974; Stafford 1981;
Hughes 1991). Bison were apparently scarce
due to poor range conditions, and occupants of
the area may have subsisted on small animals
and plants. The best documented Early
Archaic site in the region is Gore Pit
(34CM131) in southwestern Oklahoma, and it
is dated to about 4,050 B.c. (Hamrnatt 1976).
This is an open camp with burned rock ovens,
shell middens, hearths, and burials near
Lawton, Oklahoma. Artifacts recovered here
include side-notched and corner-notched dart
points (Trinity, Ensor, Darl, Ellis, Frio, and
Meserve), grinding basins, scrapers, Clear
Fork gouges, and knives. Local Ogallala
quartzites seem to dominate the lithic
assemblage.
The Stigenwalt site (14LT351) in southeast
Kansas dates from 6190 to 7913 B.C. This is
a repeatedly occupied camp consisting of
burned rock features and a variety of lithic
tools and debris. Lanceolate, basally notched,
side-notched, and corner-notched dart points
have been recovered (Thies 1990). Other tools
include grinding basins and manos, drills,
knives, flake scrapers, hammerstones, bone
awls, and bird bone beads. Faunal remains
indicated exploitation of small mammals,
turtles, fish, frogs, and mussels. The groups
exploited locallithics but also obtained some
materials from the Flint Hills and the Ozark
Mountains.
Several Middle Archaic (roughly 5050-2050
B.C.) complexes are known for southern
Kansas and northern Oklahoma. An important
Middle Archaic complex is the Calf Creek
horizon identified across all of Oklahoma and
many areas of Texas, Arkansas, Kansas, and
Missouri (Wyckoff 1994; Neal and Drass
1998). This prehistoric manifestation is
marked by the presence of distinctive chipped
stone tools, principally projectile points. The
deep, basally notched Calf Creek point is
characteristic and may be accompanied by
Cossatot River points, practice notched pieces.
scrapers, knives, and other tools. Brooks
(1995) has identified a Calf Creek camp,
34GT6, within the Salt Fork survey area, and
Sullivan (1995) records two Grant County
sites along Bluff Creek where Calf Creek
points have been recovered. In addition,
numerous Calf Creek points have been found
at several sites in the Kaw Lake area just east
of the Salt Fork project area (Wyckoff 1995).
Excavations have been conducted at the Kubik
site revealing hearths, a burned rock oven,
bison and deer bone, Calf Creek points, and
lots of chipped stone debris (Neal and Drass
1998). This repeatedly occupied camp has
been radiocarbon dated to about 3360 to 3975
B.C. (calibrated ages). Camps seem to be
small and situated on terraces or uplands
overlooking rivers or smaller streams. The
Calf Creek groups in north central Oklahoma
appear to have relied heavily on Florence-A
chert for use in manufacturing their chipped
stone tools. Calf Creek points and sites are
rare west of the study area and in Kansas,
although Calf Creek materials are found at the
Coffey site in northeastern Kansas (Wyckoff
1995:179).
Other Middle Archaic complexes that occur
near the study area include the McKean,
14
Tom's Brook, and Grove complexes, and the
Chelsea phase. The McKean complex is
represented by a few sites in northern
Oklahoma that contain McKean, Duncan, and
Hanna points. Only a few of these points have
been found in Oklahoma, and there is very
little information on assemblages and
association of the finds with other Archaic
complexes (see Neal and Drass 1998:52). One
site with McKean complex points has been
identified in Kay County east of the study
area, but no dates are available for this
component.
The Tom's Brook and Grove complexes are
found in the Ozark Mountains area of eastern
Oklahoma, northwestern Arkansas, and
southwestern Missouri (Neal and Drass 1998).
These complexes are suggested to date
roughly from about 8000 to 4000 years ago.
They include small open camps along stream
terraces and rockshelters. Hearths and rock
features have been discovered at some camps.
Mixed deposits at many of these Archaic sites
have hindered definition of assemblages and
components. Variations in projectile point
morphology have been used to identifying
temporal changes during this period (Neal and
Drass 1998; Wyckoff 1984). Lanceolate and
lobed dart points such as Searcy, Rice Lobed,
and Jakie Stemmed are considered evidence of
early occupations dating some 7000-8000
years ago. These are followed by side-notched
points such as Big Sandy and straight-stemmed
points such as Johnson that were
used from about 6000-7000 years ago. Calf
Creek and McKean/Duncan/Hanna points
follow from 3000 or 4000 B.C. to about 1000
B.C. Other than the Calf Creek and McKean
complex points, the sequence has not been
defined west of the Ozarks in the study area.
Almost no information is available on Middle
Archaic sites west of the study area.
In southern Kansas, work on the EI Dorado
Reservoir resulted in definition of the Chelsea
phase, an Archaic manifestation that has been
dated to about 2050-2850 B.C. (Brown 1987b;
Grosser 1973). Base camps and small hunting
camps with large rock hearths or roasting pits
have been identified; many of these sites were
repeatedly occupied. These people hunted a
variety of game including bison, deer, small
mammals, birds, and fish. Grinding slabs and
manos indicate the processing of seeds or nuts
at these camps. Tool assemblages include
biface knives, scrapers, and predominantly
corner-notched dart points that resemble
Williams and Marcos types (Neal and Drass
1998:54). This phase is currently restricted to
the Walnut River Valley in the southern Flint
Hills area, but similar points are found
throughout northern Oklahoma and southern
Kansas. Reid (1984: 193) suggests that the
Lovelace site (34WN105) in the Copan
Reservoir area, northeastern Oklahoma, is
related to the Chelsea phase.
Sites and artifacts related to the period from
about 2050 to 50 B.C., the Late Archaic, are
better represented throughout the area
although information on complexes is not
much better than for the Middle Archaic
period. The climate appears to have become
moister some time after 2550 B.C. and hunters
and gathers may have moved back into the
prairies to hunt bison, which appear to become
more abundant. Temporary base camps
situated on stream or river terraces have been
reported from the Kaw Lake area and the
nearby Salt Creek drainage (Vehik 1985b).
Corner-notched, weakly barbed dart points
(Marcos or Ellis, Palmillas or Williams, and
Ensor) and contracting stem darts, primarily
15
Gary points, are found at many small camps in
Oklahoma. Other tools include grinding
stones, knives, scrapers, and a few bone tools,
primarily awls and flakers. Bison kill sites
dated between about 20 B.c. and A.D. 970 are
found in western Oklahoma, but are not, as
yet, documented in north central Oklahoma
(Bement and Buehler 1994; Hughes 1991;
Lintz et aI. 1991). To the east is the Lawrence
phase found in the Cherokee Prairie and Ozark
settings of northeast Oklahoma, southwest
Missouri, and western Arkansas (Wyckoff
1984). The Lawrence site (34NW6) is a semi-permanent
base camp and other large base
camps are found along major stream terraces
(Neal and Drass 1998:54). Many sites have
evidence of repeated occupation. Rock-lined
hearths and possible ovens have been
identified, as well as shallow pits and burials.
Corner-notched points such as Williams,
Afton, Marshall, Ellis-like, and Table Rock
are characteristic at these camps. Small
comer-notched points may indicate the use of
the bow and arrow at some of these sites.
Diverse tool kits are common and include
drills, scrapers, knives, flake tools, grinding
basin, manos, abraders, gorgets, bone awls
and flakers, and pendants. Subsistence
involved the hunting of deer and a variety of
small mammals, birds, fish, and turtles. The
grinding implements suggest plant processing.
Some domesticated crops have been reported
from rockshelters sites in southwestern
Missouri that were occupied at this time (Fritz
1997).
The EI Dorado phase, initially defined for the
Walnut Creek area in the southern Flint Hills
in Kansas, is now extended to other areas in
eastern Kansas (Brown 1987b:XII-13). The
phase is dated from about 1350-2050 B.C. and
is found in tall and mixed grass prairie
settings. Sites are predominantly large base
camps on river terraces. Features include
burned rock concentrations, hearths, shallow
pits, and structures indicated by the presence
of daub and post molds. Burials have also
been found. Dustin/Lamoka side-notched
points are characteristic at these sites, but
Table Rock and lanceolate points (Sedalia) are
also present. Other tools include bifacial
knives, drills, choppers, abraders,
hammerstones, celts, and grinding stones.
Faunal remains indicate exploitation of a
variety of animals including bison, deer,
antelope, small mammals, turtles, and
shellfish.
The Walnut phase appears to represent a
terminal Late Archaic occupation in southern
and eastern Kansas. The phase was originally
defined from the Snyder site in the southern
Flint Hills (Grosser 1970). Walnut phase is
dated from about 1200 B.C. to A.D. 1.
Investigated sites are seasonal camps.
probably hunting camps. Small corner-notched
points (Walnut Valley Comer
Notched) may represent early arrow points
(Brown 1987b). No pottery has been found,
but knives, bifaces, choppers, and large points
occur.
Several sites in the Copan Reservoir area, east
of the current survey zone, have been
attributed to the EI Dorado phase (Reid
1984: 193). Most of these appear to be
temporary camps. Similar temporary camps
with comer-notched dart points are attributed
to the Late Archaic period in the Kaw Lake
area just east of the study area (Rohrbaugh
1974). Similar sites are also present along Salt
Creek and one, 340S245, has been tested and
dated to about 160-430 B.C. (Buehler 1985).
This site has small pits and burned rock
16
features with contracting stem dart points and
ground stone. The site is not currently
associated with an identified complex. Many
of the Late Archaic sites in north central and
northeastern Oklahoma are found buried on
terraces near streams. Some of these are over
a meter below the current surface (see Beuhler
1985; Reid 1984). To the west of the study
area there is very little information on Late
Archaic sites, although site forms report
comer-notched dart points and other materials
that may indicate camps related to this period.
Plains Woodland or Early Ceramic
The Woodland period on the Southern Plains
is poorly documented and it is often difficult
to distinguish these sites from Late Archaic
sites. In western Oklahoma, sites dating
between about A.D. 1 and 900 are defined as
Plains Woodland. This period is marked by
the introduction of the bow and arrow and
pottery. Late Archaic tools such as corner-notched
and contracting stem dart points,
knives, and grinding basins continue to be
used by Woodland groups, but corner-notched
arrow points (Scallorn and others) along with
small amounts of pottery appear in
assemblages. The climate becomes wetter at
this time and populations on the plains may
have increased. After about A.D. 500, bison
hunting may have become less important and
Woodland groups hunted deer and smaller
game. Horticulture may have also been
adopted during this period, but evidence far
this adaptation is not presently documented in
this area.
Sites representing the Plains Woodland period
are more common near the study area than are
sites from earlier occupations. Research in the
Kaw Lake area of the Arkansas River basin
has provided many dated Woodland period
sites (see Vehik 1984). Work farther east in
the Cross Timbers of Osage and Washington
counties has also provided information on
sites of this period (Henry 1977, 1980; Reid
1984). Very little is known of sites to the
west, but several Woodland period complexes
have been defined in southern Kansas (see
Brown and Simmons 1987).
Three Woodland complexes have been
identified in southeastern Kansas and all of
these overlap in distribution. There is some
question as to whether these are all distinct
entities (Hofman and Brooks 1989; Thies
1990), but each is briefly summarized here.
The Butler phase is found in the Walnut River
area of the southern Flint Hills region in
Kansas, and it may represent the closest
complex to the study area. It has been dated
from about 1150-1450 years ago (Brown and
Simmons 1987). Sites are small hamlets or
homesteads containing one or two houses.
These were repeatedly occupied sites situated
along streams. Houses are small, oval-shaped,
made from bent poles and grass/twigs, and
about 5-6 meters in diameter (Brown and
Simmons 1987:XII-20). Shallow basin pits are
also present. Subsistence is based on hunting
bison, deer, antelope, and small game and the
gathering of wild plants. Scallorn-like serrated
arrow points and unnotched arrow points are
characteristic of this complex. Other tools
include bifacial knives, scrapers, modified
flakes, celts, manos, and bird bone beads.
Ceramics are present and typically include
two types. Vertically cordmarked jars
tempered with grit (caliche, sand, clay,
limestone, and feldspar) represent a local
Woodland style (Hofman and Brooks
1989:64). These vessels have straight rims,
flat lips, and conical bases. The second pottery
type has straight, tapered rims decorated with
17
diagonal, zoned dentate stamping. This type is
tempered with feldspar and grog. The second
ware is similar to Middle Woodland,
Hopewell-derived ceramics from the Kansas
City area and is considered evidence of
Hopewellian influence on local Woodland
groups (Grosser 1973; Hofman and Brooks
1989).
The second Kansas Woodland complex is the
Greenwood phase dating from 950 to 1550
years ago. Greenwood phase sites have been
identified from the southern Flint Hills east
through much of southeastern Kansas (Brown
and Simmons 1987:XIII-21). Sites include
large villages on terraces or floodplains of
rivers and small camps on smaller streams.
Villages contain long oval houses up to 19 m
long and 10m wide with a central hearth
(Blakeslee and Rohn 1982). Other features
include hearths, burned rock concentrations,
and shallow basin pits. Subsistence consists
principally of hunting deer, bison, and small
animals and gathering wild plants, although
some corn has been recovered from one site.
A variety of corner-notched, expanding stem,
and contracting stem dart points are found at
sites as are Scallorn arrow points. Other tools
resemble those found at Butler phase sites.
Ceramics include the Verdigris and
Greenwood pottery types (Brown and
Simmons 1987). Verdigris pots are conical-based,
thick-walled jars with vertical
cordmarks and are tempered with crushed
limestone. The vessels include two forms, one
with straight walls and one with slight
shoulders. The Greenwood type includes
cordmarked globular jars with straight or
slightly recurved rims and conical bases
(Brown and Simmons 1987:XIII-16). This
pottery is tempered with clay or shale. The
Greenwood phase may have developed into
the Late Prehistoric Pomona complex in
eastern Kansas (Brown and Simmons
1987:XIII-16).
Slightly farther east in southeastern Kansas is
the Cuesta phase, dating from about A.D. 700-
1000. This culture is characterized by a
variety of stamped, punctated, cordmarked,
and plain pottery types that strongly resemble
ceramics from Kansas City Hopewell
occupations to the northeast and Cooper focus
materials to the south in Oklahoma (Hofman
and Brooks 1989:64). The Cuesta phase is
considered to represent either Hopewellian
influence on a local Middle Woodland group
or migration of a Hopewellian culture into
southeastern Kansas (Brown and Simmons
1987:XIII-I0). Other artifacts from these sites
include expanding stem dart points. Gary
points, Scallorn arrow points. bifacial knives.
drills, scrapers, manos, grinding basins, celts,
gorgets, atlatl weights. and bone pins. awls.
fleshers, and beads. Settlements are either
large villages near major streams or
homesteads and small hamlets along small
streams. Houses are oval to circular measuring
11-15 m long and 8-12 m wide (Brown and
Simmons 1987:XIII-9). Basin-shaped pits.
hearths, and burials have been found at
villages. Subsistence is based on hunting (deer
and small game) and gathering. but corn and
sunflower indicate use of some cultivated
plants (Hofman and Brooks 1989:64).
There are three Woodland complexes
identified for northeastern Oklahoma. Vehik
(1984: 177) defines the Delaware A, Cooper.
and Delaware B foci in Neosho River area of
the Ozarks. These complexes are considerably
east of the study area, and they are not
reviewed here. However, it may be significant
that the Cooper focus is defined by
18
Hopewellian ceramics similar to Cuesta phase
materials in southeastern Kansas. Many other
artifacts appear to be similar in both the
Cooper focus and Cuesta phase assemblages.
Vehik (1984: 183) suggests that Cooper may
be a Hopewellian intrusion into northeastern
Oklahoma.
In north central Oklahoma there has been
considerable research at Woodland sites in the
tall grass prairie and Cross Timbers of Kay
and Osage counties. There are no defined
complexes, but Vehik (1984) has suggested
three groups or components for Woodland
sites. In the Arkansas River basin within the
Kaw Reservoir area there are numerous sites
that date from about 950 to 1850 years ago.
Other sites of this period have been recorded
from the Bird Creek drainage in eastern Osage
County. The earliest Woodland sites in these
areas are suggested to date between about
A.D. 100 to 300. (Vehik 1984). Sites in the
Kaw Reservoir area include small amounts of
Hopewellian-type pottery (Ozark Zoned,
Cowskin Dentate, and Cooper Zoned sherds).
Most other pottery is smooth surfaced but
cordmarked vessels are also represented.
Contracting stem (Gary type) dart points
predominate in the assemblages, but corner-notched
dart points and arrow points
(Scallorn) are also present (Hofman and
Brooks 1989:67). Sites include lithic
workshops, hunting camps (at open sites and
rock shelters in Osage County), and semi-permanent
camps marked by the presence of
daub, scattered post molds, and storage pits.
Burned rock mounds are also recorded in the
Caney River area of the Cross Timbers. Vehik
(1984: 188) suggests that these mounds were
in use through the Woodland period and may
be associated with hunting and kill processing
activities.
Around A.D. 300 to 800 pottery becomes
more frequent at Woodland sites, and it is
mostly smooth surfaced and tempered with
sand, limestone, or clay (Vehik 1984: 187).
This second Woodland component also has
predominantly comer-notched arrow points
(Scallorns) and Gary dart points. Knives,
drills, a variety of scrapers (thumbnail and
snub-nosed varieties increase in abundance at
this time), gravers, double-bitted axes, manos,
abraders, and stone beads occur at these sites.
Most sites appear to be hunting and/or plant
processing locales. Rockshelters occur in the
Cross Timbers, whereas in the Kaw Lake area,
open camps are found along tributaries of the
Arkansas River. There is no evidence for
extensive external contact with distant groups.
The third set of Woodland sites dates from
about A.D. 800-1100. Hofman and Brooks
(1989:67) suggest that some of these sites
represent transitional Plains Village
components. Projectile points are
predominantly arrow points such as Scallorn,
Washita, and Fresno. Pottery is variable but
most is smoothed and some has shell temper.
Other tools are beveled knives, snub-nosed
and thumbnail scrapers, gravers, hoes, manos
and grinding basins, abraders, and nutting
stones. Settlements include rockshelters and
burned rock mounds in the Cross Timbers that
were probably used as hunting camps and
processing locations. Open sites occur along
small streams and the Arkansas River valley,
but there is no evidence for permanent
occupations (Vehik 1984: 192). A burial
mound (340S99) is recorded for the Cross
Timbers area in Osage County and other
burned rock mounds appear to have been used
for some processing activity at this time.
Subsistence evidence is minimal, but hunting
deer, rabbits, and other small game, collection
19
of mussels, fishing, and gathering plants seem
to be the major activities.
In general, the Woodland adaptation in north
central Oklahoma indicates the establishment
of base camps and semi-permanent villages
along the Arkansas River and other major
streams early in the Woodland period (Vehik
1984: J 97). These settlements exhibit some
influence from Hopewell groups 111
northeastern Kansas or similar groups such as
Cooper focus or Cuesta phase people in
northeastern Oklahoma and southeastern
Kansas. The settlement pattern may have
changed toward a more mobile hunting and
gathering pattern during the late part of the
Woodland period, but more research is needed
to determine if late base camps/villages may
exist along river terraces or in other settings.
Subsistence during the period seems to have
emphasized the hunting of deer and a variety
of small game, fishing, and the gathering of
wild plant foods. Evidence of cultivated plants
is generally lacking during this period in north
central Oklahoma. Woodland sites in the
mixed grass prairies are not defined west of
the study area. but this probably reflects the
lack of archeological research rather than an
absence of Woodland occupation.
Late Prehistoric or Middle Ceramic
The Late Prehistoric period from about 500 to
1050 years ago is marked by significant
changes in technology. settlement, and
subsistence. Many Southern Plains villagers
appear to have developed from local
Woodland people. Changes that are evident by
about A.D. J 000 include intensification of
horticulture. the widespread use of subsurface
storage pits, expanded artifact inventories, and
the use of permanent houses and larger
settlements. Typical Late Prehistoric artifacts
include cordmarked and plain globular pots,
side- and unnotched arrow points. beveled
knives, numerous scrapers (particularly snub-nosed
end scrapers), bison bone horticultural
tools, and trade items from the Southwest or
Caddoan areas.
In general, more Plains Village sites have
been investigated in the Southern Plains than
sites representing most other periods. Several
archeological complexes have been defined
for the prairies of south central and
southwestern Kansas, but Plains Village sites
in northern Oklahoma have usually not been
assigned to complexes (Figure 3). Several
Oklahoma sites dating to this period have
been investigated and provide some
information on the people who Iived here. The
Kansas complexes are often based on
incomplete data and some of these are also
poorly defined. The Pomona variant
containing four phases is defined for a large
area of eastern Kansas (Brown 1987d:XIII-34-
42). The variant is found in the prairies
northeast of the study area. Pomona dates
from about 500-900 years ago. Sites consist of
one to three oval houses on stream terraces.
House shape is irregular with interior and
exterior bell-shaped, cylindrical, or basin pits.
Ceramics include smoothed and cord marked
bowls and globular jars that have flared or
excurvate rims (Brown 1987b:XIII-38). Some
rims are also collared and decorations are
restricted to rims and lips. Unnotched. side-notched,
and basally notched arrow points.
side and end scrapers, gravers. dri lis. abraders.
celts, pipes, manos, and grinding basins are
common on these sites (Brown 1987b:XIII-
38). Bone digging implements are rare.
Subsistence includes cultivation of maize.
squash, beans, and sunflowers, gathering wild
plants, and hunting deer, bison, elk, and a
20
· .....- ".~.~Jltelope Creek Phase '-"'- ~
Figure 3: Plains Village sites and complexes.
variety of small animals.
The Bluff Creek complex is defined directly
north of the current study area. Bluff Creek
sites occur in Harper and Sumner counties,
Kansas with the type-sites found along Bluff
Creek, which flows south into Oklahoma
eventually emptying into the Chikaskia River.
The complex was originally dated to around
AD. 1050 (Witty 1978), but more recent dates
indicate occupation extends to at least AD.
1250 (Huhnke 2000). Sites are small villages
encompassing a variety of house forms, oval,
square, and rectangular structures that were
made of wood posts plastered with daub.
Many of the houses have no interior hearths,
but cylindrical storage pits occur inside and
outside of structures. Chipped stone tools
resemble those found at most Plains Village
sites and include diamond-beveled knives, end
and side scrapers, drills, and retouched flake
tools. Arrow points are typically side-notched
and basally notched, but some Fresnos are
also found. Florence-A chert is the
IPaoli/Washita River Phases '_<: .::
predominant material used for chipped stone
tools. Bone tools include a variety of bison
implements such as tibia digging stick tips and
scapula hoes and cleavers. Pots are either
cordmarked or smoothed globular jars. The
pots are predominantly tempered with sand
and occasionally bone. Some cordmarked
vessels have collared rims similar to pottery
found at Central Plains villages. Decorations
are limited to tool-impressed lips (Thies
1989). Some shell tempered plain vessels
occur and are considered trade wares from the
south (Witty 1978:63). Subsistence is based
upon cultivation of maize and other plants and
hunting of, principally, bison. The complex
may have developed into the Great Bend
aspect.
The Pratt complex is northwest of the Bluff
Creek sites and it may have developed from
the Bluff Creek complex. Brown and
Simmons (1987:XVI-3) suggest that the Pratt
complex dates to around AD. 1400-1500, but
radiocarbon dates from one Pratt site extend
21
the age of this complex to as early as A.D.
1275 (Ranney 1994:85). Large villages are
found on ridge tops overlooking streams.
Houses are circular to oval grass structures
measuring 10 to more than 16 feet in
diameter. Central hearths are present and
storage pits may be present within or outside
of houses. Artifacts are similar to Bluff Creek
materials, but trade items such as Olivella
shell beads, Gulf marine shell, southwestern
pottery, turquoise, and obsidian are more
common at the Pratt sites. Lithic material use
also varies from Bluff Creek with Alibates
dominating chipped stone at one Pratt site
(Ranney 1994). Ceramics consist of
cordmarked jars with sand and/or bone
tempering. Vessels have straight to outward-flaring
rims and round or flat bottoms.
Subsistence is based on hunting bison and
deer, gathering wild plants, and small-scale
cultivation of maize (Brown and Simmons
1987:XVI-3). The Pratt complex is considered
ancestral to protohistoric Wichita complexes
in the Arkansas River basin of central and
south central Kansas
Farther west in Kansas is the Wilmore
complex found along tributaries of the Salt
Fork of the Arkansas River and the Cimarron
River in the southwest part of the state. The
three excavated sites used to define this
complex have considerable variability and the
complex has recently been combined with
sites and complexes in northwestern
Oklahoma to define a Plains Border variant
(Bevitt 1999). The variant is characterized by
small villages or hamlets containing houses
that are square to rectangular or oval and have
semi-subterranean floors. The principal
artifact characteristic is the pottery. Small to
medium jars typically have sand or bone
temper and have decorated rims on
cordmarked vessels (Bevitt 1999: 183). The
Wilmore complex sites vary in characteristics
but each has semi-subterranean structures.
Possible arbors or drying racks, pits, and
hearths are also present at these sites. An oval
pit structure was recorded at the Lundeen site
at the western extent of this complex, but
shallow, square structures are recorded for the
eastern sites (Bevitt 1999). Wilmore complex
ceramics are primarily cordmarked jars
although some pots have been smoothed. The
jars usually have constricted necks with out-flaring
rims, and rounded bases. Rims are
commonly decorated with pinched nodes.
fingernail punctations, or tool impressions.
Lips may also be incised or impressed with
lines or punctuations. There is a great deal of
variation in ceramics between the Wilmore
sites and Bevitt (J 999) suggests this reflects
temporal and functional variation in the
assemblages. Sand and bone are the most
common tempering agents. Chipped stone
tools resemble those from other Plains Village
sites. Raw material use varies between sites.
Flint Hills cherts. Smoky Hill jasper. Edwards
chert, and Alibates chert are represented to
varying degrees at the different sites. The
eastern, Comanche County. sites seem to
prefer to use Flint Hills cherts such as
Florence-A (Bevitt 1999: 177). Bison are
present in large amounts at all of the sites and
there is evidence that corn and other crops
were important. In general, the three Wilmore
complex sites seem to actually represent
different manifestations either different spatial
entities or variations from different periods of
occupation.
In northern Oklahoma, there have been
investigations at several Plains Village sites.
but research has not been extensive enough to
define complexes. Vehik (1985b:321-324)
22
reports 29 Late Prehistoric sites along Salt
Creek and Beaver and Little Beaver creeks in
eastern Kay and western Osage counties. Most
of these represent short-term occupations,
although sites near the Florence-A quarries
may be villages or base camps used to exploit
the chert for trade (Vehik 1990:133-137). The
initial work at Kaw Lake along the Arkansas
River did not reveal much evidence of Late
Prehistoric occupation before about A.D.
1450. However, lake erosion has since
exposed evidence of early Plains Village sites
on high terraces (Galm 1979; Vehik and Flynn
1982). The Uncas site (34KA 172) is the best
documented Plains Village occupation in the
area. The site represents a small hamlet or
village on a terrace above the Arkansas River.
Dates range from A.D. 1100-1400 but
occupation most likely occurred around A.D.
1325 (Vehik and Flynn 1982:7). Pottery and
house form distinguish this site from other
Plains Village sites to the north in Kansas and
to the south in Oklahoma. Houses are semi-subterranean
and square with rounded comers.
They have four center posts, a central hearth,
and some contain bell-shaped storage pits.
Pottery consists of globular vessels with
constricted necks, loop handles, and rounded
bottoms. These are predominantly plain
(Uncas Plain) or decorated (Coon Creek
Incised) vessels, although some cordmarked
sherds are present. Tempering is variable with
sand, bone, grit, or shell used in various pots.
Decorations consist of parallel, incised lines.
Vehik (1994) suggests some similarity with
ceramic types from the Central Plains
tradition, the Smoky Hill variant, and the
Lower Walnut focus of the Great Bend aspect.
Other artifacts resemble typical Plains Village
assemblages with side-notched arrow points
predominating in the assemblage (Galm
1979). Chipped stone debris is dominated by
j .
Florence-A with occasional pieces of Alibates
and other materials. These people hunted
bison, deer, and small game, but there is no
direct evidence of plant cultivation.
To the west of the study area there is, again, a
dearth of information on Late Prehistoric sites.
In the Oklahoma and Texas panhandles many
villages related to the Upper Canark variant
have been identified and excavated (see Drass
1998). These include the Zimms complex
found in northwestern Oklahoma, with the
Hedding site (34WD2) occurring in eastern
Woodward County. There is, however, some
debate as to whether Hedding should be
included with Zimms or separated into another
complex (Bevitt 1999; Brosowske 1998).
Hedding has an Antelope Creek phase house
with extended entryway and central channel.
Its pottery includes sand tempered Borger
Cord marked vessels and a plain ware, Wolf
Creek Plain. A few pieces are decorated with
incised lips or chevron designs on rims that
resemble Buried City complex pottery to the
west in the Texas Panhandle. Local, Day
Creek materials dominate Hedding lithics, but
about 10% of the chipped stone is made from
Florence-A.
Other Late Prehistoric sites 111 Woodward
County include the Loomis burial mound
(34WD 12) and the Traders Creek camp
(34WD5). Very little is known about the
Loomis site because few artifacts were
recovered. Two grit tempered cordmarked
sherds and a Fresno point suggest a Late
Prehistoric burial locale. The Traders Creek
site has had more extensive work, but there is
little published information. This appears to
be a base camp with pits and hearths present
(Buehler 1991). The assemblage includes
cordmarked pottery with sand and, rarely,
23
bone temper. Decorations are limited to
loop(?) handles, pinched rims, and tool
impressed lips or rims. Lithics appear to be
dominated by local, Day Creek chert, but
some Florence-A and Alibates are reported
(Buehler 1991). Radiocarbon dates indicate
occupation of Traders Creek around A.D.
1450.
Finally, two small hamlets or base camps,
Wilson (34WOI0) and Nelson (34W057), are
reported in eastern Woods County along Eagle
Chief Creek (Drass 1999). These Late
Prehistoric sites have assemblages consisting
of side-notched and unnotched arrow points,
end and side scrapers, and thin, cordmarked or
smoothed pottery. The ceramics are tempered
with bone and sand and contain a few rims
and lips decorated with fingernail
punctuations. Local materials dominate
chipped stone assemblages, but Flint Hills
cherts (Florence-A and Wreford) represent 26-
27% of the Late Prehistoric material (Drass
1999: 154). There is evidence of a house at the
Wilson site, but only test excavations have
been conducted.
Protohistoric or Late Ceramic
Around 500 years ago there are dramatic
changes in the cultures of the Southern Plains,
and these cultures continued to develop and
change after initial European contact in 1541.
Full historic documentation in this area occurs
after about A.D. 1750. Most of the Plains
Village complexes in the western half of
Oklahoma are considered ancestral to the
Wichita or other Plains Caddoan groups,
whereas southeastern Oklahoma is tied to the
Caddo. Apachean groups, also, migrated into
parts of western Oklahoma and Texas
sometime before A.D. 1550 (Wilcox 1981),
but Wichita and Caddo groups continued to
occupy central and eastern Oklahoma at the
time of historic contact. There are few early
historic accounts of native cultures in
Oklahoma, and the relationship between many
of the Late Prehistoric groups and the later
protohistoric and historic people is often
unclear. However, a number of archeological
complexes have been identified in the
Southern Plains for the period from about
A.D. 1450-1700.
The Wheeler phase (Drass and Baugh 1997) is
the primary proto historic manifestation in
western Oklahoma. There are fewer
protohistoric sites known in central Oklahoma
and none attributed to this complex in
northern Oklahoma. Wheeler phase groups.
however, had important contact with
protohistoric Wichita groups in north central
Oklahoma. Many of the western Oklahoma
sites have significant amounts of Florence-A
chert that were traded or brought out to the
area from sources in eastern Kay County or
southern Kansas.
The major Wheeler phase sites include
Edwards I (34BK2), Taylor (34GR8), Little
Deer (34CUI0), Duncan (34 WA2), and
Goodwin-Baker (34RMI4) In western
Oklahoma. The Wheeler phase, in generaL
includes villages with houses and arbors and
large encampments with fortifications (34BK2
and 34WA2). The fortified sites may have
served as seasonal rendezvous for communal
bison hunts and trade fairs. The fortifications
consist of large, circular ditches enclosing an
area about 50 meters in diameter. Earthen
ramparts were probably once present but are
not preserved at the Oklahoma sites. The
ditches are about 1 meter wide and 1 meter
deep. A similar fortification has been
documented at the Longest site (34JF 1), an
24
early historic Wichita site on the Red River
(Bell and Bastian 1967). Wheeler phase
subsistence patterns emphasize bison hunting.
Corn cultivation is suggested by the presence
of corn at some western sites, but its
importance in the economy has not been
established. Trade with Southwestern and
Caddo groups appears to have increased
during this period with many nonlocal pottery,
lithics, and shell items represented at Wheeler
phase sites. Radiocarbon, archeornagnetic, and
obsidian hydration dates place occupation
between about A.D. 1450 and 1650 or 1700
(Baugh 1986).
Artifacts characteristic of Wheeler phase sites
include a predominance of unnotched (Fresno)
points, some side-notched (Washita and
Harrell) points, and small amounts of a basally
notched point identified as Garza points. A
thin, black sandy pottery (Edwards Plain)
dominates the ceramic assemblages. These are
small vessels with round bases, constricted
necks. and everted or straight rims (Hofman
1984). Decorations are not common but
notched rims, fingernail punctates, incised
lines, and applique designs may be present. A
plain, clay, or grog tempered pottery (Little
Deer Plain) and small amounts of
Southwestern and Caddoan pottery are also
represented in these assemblages. Other
characteristic tools include diamond-beveled
knives, expanding base drills, ensiform pipe
reamers, and many, often large, end and side
scrapers. Lithic materials used in the
manufacture of the stone tools vary depending
upon location of the site, but usually include
large amounts of nonlocal cherts such as
Alibates, Florence-A, or Edwards. Bone and
shell artifacts are not abundant, but they
comprise a variety of decorative items as well
as tools such as bone awls and a few bison
bone digging stick tips or hoes.
Protohistoric sites in central and south central
Kansas are associated with the Great Bend
aspect that represents Wichita ancestors. The
Great Bend aspect has been divided into two
foci, the Little River and Lower Walnut foci.
The Little River focus is found in central
Kansas and the Lower Walnut division occurs
along the Walnut and Arkansas rivers in south
central Kansas. These groups may have
developed out of local groups such as the Pratt
or Bluff Creek complexes, or they may
represent other groups that moved north from
Oklahoma. The aspect dates roughly from
A.D. 1450 to the historic period in the early
18th century in this area. Both foci are at least
partly contemporaneous (Wedel 1959:586),
and sites were occupied when Coronado and
Ofiate traversed the area in 1541 and 1601.
These groups may have been moving south at
the time of Spanish contact and the Lower
Walnut focus may have persisted later than
the Little River focus (Hawley 1994:30). Two
sites (Deer Creek, 34KA46, and Bryson-
Paddock, 34KA 70) farther south along the
Arkansas River in Kay County, Oklahoma are
early 18th century Wichita sites that were
contacted by French traders (Wedel 1981).
Wichita groups subsequently moved south to
the Red River in south central Oklahoma and
northern Texas.
Great Bend aspect settlements consist of large
villages scattered along terraces and hillsides
near major streams. These were agricultural
groups with gardens of corn, beans, and
squash extending around the houses. Houses
are oval grass-covered structures with central
hearths (Brown 1987a). They are pole and
grass-thatch structures, which sometimes are
semi-subterranean. Arbors are also reported
25
(Lees et al. 1989). Bell-shaped pits are
common and may be very large. Some sites
have "council circles" consisting of shallow
ditches and low mounds (Wedel 1959).
Refuse mounds are common at many of the
sites. Artifacts from some of the Great Bend
sites include Euro-American items such as
iron axes and awls, copper, brass, or glass
beads, chain mail, and, later, gun flints. Great
Bend sites are also characterized by native
·long distance trade items such as
Southwestern pottery, obsidian, turquoise, and
pipestone pipes. Ceramics include Geneseo
Plain, Simple Stamped and Red Filmed
associated with Little River focus sites, and
Cowley Plain found at Lower Walnut focus
sites (Brown 1987a). The Geneseo wares are
typically sand tempered, but the Cowley
pottery is shell tempered. Vessels are bowls or
flat-bottomed jars with direct or flaring rims
and loop handles. Projectile points are
predominantly unnotched (Fresnos). Beveled
knives and end scrapers are common. Bison
bone tools are common and include scapula
hoes, squash knives, beamers, and fleshers.
Bone and shell beads and other ornaments are
also found. Other than gardening, bison
hunting was an important subsistence activity.
Special hunting parties are recorded for
historic Wichita groups from the fall to early
spring and these likely were important for the
protohistoric groups (Brown 1987a). A variety
of other game was also hunted and plant
collection was probably important.
Historic
The earliest accounts documenting indigenous
peoples in North America come from Spanish
records. In 1601, Ofiate traveled through what
is now western Oklahoma along the Canadian
River valley and northward into central
Kansas. Ofiate stopped at a Wichita village in
Kansas. By 1760, many of the Wichita
relocated to sites in south central Oklahoma
along the Red River as well as at a few
archaeologically recorded villages further
west and south. The move back south and
west was partially in response to Osage
aggression in the north.
In 1803, the United States purchased a great
tract of land west of the Mississippi in the
Louisiana Purchase. The purchase ended the
threat of Spanish reprisal for trespassing and
opened the Oklahoma area to a flood of
trappers and traders who capitalized on the fur
trade, which had already been established
further east and north by the French. Major
waterways connecting to the Arkansas River
and eventually to the Mississippi provided
easy access to the fur-rich resources of
Oklahoma, to trading tribes of the Plains. and
to the Santa Fe trade center.
Increasing settlement by Euro-Americans east
of the Mississippi River brought a demand for
the native residents of those areas to be
relocated so that settlers might have access to
their lands and by removal insure safety from
hostilities for European settlers. The Osage
ceded what is today northern Oklahoma to the
United States in 1825. In the 1828 Congress
created Indian Territory. an area set aside for
the relocation of various native groups from
the southeastern states. Cherokee started
moving into northeastern Oklahoma in 1828.
The Treaty of New Echota created the
Cherokee Outlet as an access to western
hunting areas (Morris et al. 1976). The study
area is within this Cherokee Outlet. but the
area was never occupied by the Cherokee. A
treaty in 1866 sold parts of the Outlet to the
Osage, Pawnee, Ponca, Kaws, Otos and
Missouris, and Nez Perces. The Nez Perces
26
moved back to Idaho in 1885 and their lands
were occupied by the Tonkawas (Morris et al.
1976). The Salt Fork of the Arkansas River
near its junction with the Arkansas River
includes part of the Ponca and Tonkawa
reservations. Areas farther west were
unoccupied by native groups. The same treaty
forbid permanent settlement of the Outlet by
Cherokees.
In the mid to late 1800s cattle from Texas
were driven north to Kansas for shipment east.
These trails crossed the Cherokee Outlet and
cattlemen often left their stock to graze in the
rich prairies in this area. Two trails crossed
the Salt Fork during this period. Before the
Civil War, the West Shawnee Trail was used
to bring cattle to Baxter Springs, Kansas. This
trail crossed the Salt Fork just west of its
junction with the Arkansas River. After the
Civil War, cattle were taken along the
Chisholm Trail, which crossed the Salt Fork
near Pond Creek in central Grant County.
Cattle became an important industry in the
Cherokee Outlet area in the 1870s. Ranchers
from Kansas and other areas began to use the
range illegally in the 1870s. The Cherokee
eventually began charging a grazing tax. In
1883, the Cherokee Strip Live Stock
Association consisting of a group of cattlemen
leased the Outlet for cattle grazing. The
cattlemen established large ranches along the
Salt Fork and in other areas until 1892 when
the Outlet was purchased by the government.
The Cherokee Outlet was opened for
settlement by land run on September 16, 1893
(Morris et al. 1976). Each homesteader could
claim a quarter section of land. The project
area was incorporated in Oklahoma Territory
and became Kay and Grant counties at
statehood in 1907.
27
RESEARCH DESIGN AND METHODOLOGY
This investigation involves the intensive
survey of select portions of the Salt Fork
of the Arkansas River basin in Grant and
Kay counties to identify prehistoric sites.
Previous research has involved very
little survey of the Salt Fork River basin.
One survey has included a small part in
far western Grant County and portions of
the river valley in Alfalfa County
(Ferring et al. 1976). This survey,
however, recorded no prehistoric sites
and only one historic dugout. Some
small surveys have been conducted
along tributaries in Kay County. Wallis
(1980) investigated Lost and Duck
creeks, tributaries of the Chikaskia
River, finding evidence of prehistoric
occupation from the Middle Archaic
period to the Late Prehistoric, plus some
late historic farmsteads. Most of the
prehistoric sites, however, are
considered to be Late Archaic or
Woodland camps. In addition,
avocational archeologists have reported
a variety of sites along Bois d'Arc
Creek, the Chikaskia River, and the
Arkansas River near its junction with the
Salt Fork (see Table 1). These sites
include Middle to Late Archaic camps,
Woodland camps and villages, and some
Late Prehistoric villages. Other research
includes extensive survey and
excavation in Kaw Lake northeast of the
study area and surveys and tests along
Salt Creek in western Osage County near
the Florence-A sources.
The current project is part of the
Oklahoma Historic Preservation survey
and planning process. Thus, one
objective of the survey is to record and
assess archeological resources for future
land management and preservation
planning. Given the relatively high
density of sites recorded in the eastern
end of the Salt Fork basin and in nearby
areas just east of the basin, we chose to
concentrate the efforts of this project on
the middle portion of the Salt Fork basin
in western Kay County and the eastern
half of Grant County. There are only 24
sites recorded in this portion of the Salt
Fork basin. These range from the
historic 101 Ranch to Middle Archaic
camps related to the Calf Creek culture.
There has been limited investigation of
34GT9, a Woodland or Late Prehistoric
camp, but there is no published data on
this site. The survey, thus, should permit
preliminary evaluation of prehistoric and
historic site distributions west of the
Arkansas River, and it may provide
evidence to determine if different types
of occupations occurred on the western
prairies. The survey should also provide
some information on the extent of use of
Florence-A chert in the prairies west of
the source area and quarries. In general,
the survey is designed to provide
information on the distribution and types
of prehistoric sites in the middle portion
of the Salt Fork of the Arkansas River
basin.
In terms of site distribution, previous
research in north central Oklahoma has
provided some comparative data on
settlements in the area just east of our
project. Research from the Kaw Lake
and Salt Creek projects indicates
occupation of that area from at least the
Middle Archaic period through the
historic period. Reported sites related to
the Archaic are typically small limited
activity camps scattered along terrace
settings, or upland lithic extraction
locales. Archaic occupations are not
28