Issuu

ANTIQUITIES SECTION SELECTED PAPERS

VOLUME III, NUMBERS 9-11

Department of Development Services Division of State History

December 1976

STATE OF UTAH Calvin L. Rampton, Governor

DEPARTMENT OF DEVELOPMENT SERVICES Milton L. Weilenmann, Executive Director

DIVISION OF STATE HISTORY Melvin T. Smith, Director

STATE BOARD OF HISTORY Mabel Oliver Helen Z. Papanikolas Howard C. Price, Jr. Elizabeth Skanchy Melvin T. Smith Richard O. Ulibarri

Milton C. Abrams, Chairman Juanita Brooks Dello G. Dayton Theron H. Luke Clyde L. Miller Elizabeth Montague

STATE ANTIQUITIES COMMITTEE Donald V. Hague Ray T. Matheny Melvin T. Smith Stanford S. Smith Chandler P. St. John Harold Tippetts

Milton C. Abrams, Chairman Clyde J. Benally Dello G. Dayton Evan DeBloois J. Eldon Dorman Richard E. Fike

EDITORIAL STAFF David B. Madsen, Editor Thomas J. Zeidler, Associate Editor Laurie M. George, Assistant Editor

EDITOR'S PREFACE

This volume is the third in a monograph series designed to examine and interpret the prehistoric cultures of Utah. Antiquities Section Selected Papers is specifically geared to Utah archeology, but includes papers from adjacent geographical areas and from ancillary disciplines relevant to the understanding of local archeological problems. The series has three goals: 1) to provide a vehicle for the publication of research carried out by the Antiquities Section; 2) to provide an outlet for archeological reports which do not have a general distribution (i.e., investigations done in conjunction with environmental impact statements); and 3) to allow publication of valuable manuscripts now on file and republication of articles now out of print and unavailable. Manuscripts from all sources, including state and federal agencies, educational institutions, and private individuals, will be accepted for examination and possible publication. Articles should be typed double spaced and should be accompanied by photo-ready line drawings and photographs. Submitted articles will be reviewed by the Antiquities Section staff or other qualified reviewers in the case of ancillary reports. Papers will be published on an irregular basis, depending on the number and quality of reports on file.

David B. Madsen December 1976

CUMULATIVE INDEX Antiquities Section Selected Papers

Number 1

Three Fremont Sites in Emery County Utah by David B. Madsen. Volume I, Page 1.

Number 2

Innocents Ridge and the San Rafael Fremont by Alan R. Schroedl and Patrick F. Hogan with an appendix by La Mar Lindsay. Volume I, Page 29.

Number 3

An Archeological Survey of the Northeast Portion of Arches National Park by Michael S. Berry. Volume I, Page 67.

Number 4

An Archeological Survey of the White River Area, Northeastern Utah by Michael S. Berry and Claudia F. Berry with illustrations by La Mar W. Lindsay. Volume II, Page 1.

Number 5

Man, Mammoth, and Lake Fluctuations in Utah by David B. Madsen, Donald R. Currey, James H. Madsen. Volume II, Page 43.

Number 6

Bulldozer Dune (42SL46) by David B. Madsen. Volume II, Page 59.

Number 7

Interstate Highway 1-70 Salvage Archeology by Curtis J. Wilson and Howard L. Smith with ceramic analyses by John Fritz and Christine Plimpton. Volume II, Page 67.

Number 8

Unusual or Enigmatic Stone Artifacts: Pots, Pipes, Points, and Pendants From Utah by La Mar W. Lindsay. Volume II, Page 104.

Number 9

Archeological Survey of The Bluff Bench/San Juan River and White Mesa Areas, San Juan County, Utah 1973-1974 by Richard E. Fike and La Mar W. Lindsay. Volume III, Page 1.

Number 10 Pint-Size Shelter by La Mar W. Lindsay and Christian K. Lund with appendices by La Mar W. Lindsay and Donald R. Currey. Volume III, Page 25. Number 11 Archeological Investigations In The Maze District Canyonlands National Park, Utah edited by William A. Lucius with contributions by Patrick Hogan, Leonard Losee, and William A. Lucius. Volume III, Page 75.

ARCHEOLOGICAL SURVEY OF THE BLUFF BENCH/SAN JUAN RIVER AND WHITE MESA AREAS, SAN JUAN COUNTY, UTAH 1973-1974 bv RICHARD E. FIKE

and

Bureau of Land Management Utah State Office

LA MAR W. LINDSAY Antiquities Section Division of State History

ANTIQUITIES SECTION SELECTED PAPERS Number 9

ARCHEOLOGICAL SURVEY OF THE BLUFF BENCH/SAN JUAN RIVER AND WHITE MESA AREAS, SAN JUAN COUNTY, UTAH 1973-1974 by RICHARD E. FIKE

and

Bureau of Land Management Utah State Office

LA MAR W. LINDSAY Antiquities Section Division of State History

ANTIQUITIES SECTION SELECTED PAPERS Number 9

PREFACE The archeological survey reported in the following paper is the result of an inventory conducted by the Bureau of Land Management in connection with a proposed land exchange between the federal government and the State of Utah. The purpose of this paper is primarily to report the sites identified in the inventory and secondarily to interpret spatial and temporal variations. The authors assume sole responsibility for the survey and interpretation; although, a few sites identified by the University of Utah and Brigham Young University are discussed. John W. Headley of BLM did the drafting.

ILLUSTRATIONS

TABLE OF CONTENTS Preface

Page 3

Abstract

Introduction

Location and Setting

Prehistoric Cultures of the Region Prehistoric Agriculturalists Mesa Verde/Kayenta Anasazi Proto-HistoricNumic and Navajo

11 11 12 12

Archeological Sites Archaic/Basketmaker II Basketmaker III Pueblo I/II Pueblo III Navajo ^ Discussion

13 14 14 15 15 15 16

Areal Comparisons

Appendix

References

Figure Page 1. Location and setting 6 2. Bluff Bench/ San Juan River area, eastern (upper) and western zones 8 3. White Mesa area 9 4. Legend: Archeological site types and cultural affiliations 9 5. San Juan River and lower Bluff Bench 10 6. Bluff Bench (looking east) 11 7. White Mesa with Abajo Mountains in background (looking north) 12 8. Distribution of prehistoric agriculturalists (from Marwitt 1970; Berry 1975) 13 9. Basketmaker III site (42Sa3813) on the lower Bluff Bench showing rock alignments of storage features and other cultural debris (upper) and closeup of vandalized slablined storage bin 15 10. Pueblo III site (42Sa3637) on White Mesa (upper) and closeup of rock alignment 16 11. Areal comparisons of component relative percentages 18

Fig. 1 Location and setting.

ABSTRACT

A survey conducted in San Juan County, Utah identified 67 archeological sites in the Bluff Bench/San Juan River area and 54 on White Mesa. The full range of Mesa Verde Anasazi development is represented. A few Navajo sites were also identified. Disparate settlement patterns are evident between the two areas despite the limitations of the survey and lack of internal site analysis. Site densities between the two areas differ; the relatively heavy Preceramic component on the Bluff Bench is virtually absent on the mesa. Site densities on the Bluff Bench/San Juan River show gradual decreases following Basketmaker HI times, however, they increase gradually after this period on White Mesa. These differences are attributed to improved soil and heavier precipitation on the mesa, and concomitant intensive dry farming. Percentages of Anasazi components for both areas differ somewhat from those of adjacent regions. These differences are attributed to changes in subsistence strategies in the context of spatio-temporal climatic variations as well as methodological factors such as investigator biases, sampling, differential interpretive models, and limited excavation and dating.

INTRODUCTION The purpose of an archeological survey conducted by the Bureau of Land Management in 1973-74 (Lindsay and Fike 1974) was to determine the presence and nature of archeological sites on the Bluff Bench/ San Juan River and White Mesa, San Juan County, Utah (Fig. 1). The survey provided additional information on Anasazi settlement patterns in southeastern Utah. The San Juan region has been the subject of considerable research for the past six decades. This no doubt has been prompted by the spectacular and numerous Pueblo III "ruins." Kidder (1924) was one of the first anthropologists to recognize that the region

appears to have been the breeding ground for many of " the basic traits of southwestern culture and the center of dissemination. A synopsis of previous research on Southwestern prehistory has not been included since this information is available elsewhere (e.g., Lipe and Matson 1973; Wormington 1951). This study comparatively employes recent research, much of which is unpublished. In this regard, we would like to acknowledge the cooperation of: Evan DeBloois of the U.S. Forest Service, William Lipe of the Museum of Northern Arizona, Ray Matheny of Brigham Young University, and Joseph Winter of San Jose State University.

. . . has yielded a great amount of material in regard to the oldest phases of human development in the southwest. More important still is the fact that the San Juan

LOCATION AND SETTING The Bluff Bench/ San Juan River area covers roughly 30 sq. mi. and is bordered by the San Juan River and the Navajo Indian Reservation on the south, Butler Wash on the west, Tank Mesa and Recapture Pocket on the north, and McCracken Mesa on the east. Cottonwood Wash and U.S. Highway 163 roughly bisect the area from north to south (Fig. 2). The White Mesa area, ca. 15 mi. north of the San Juan River, covers roughly 6 sq. mi., and is bordered by Corral Creek Canyon on the east, the Ute Indian Reservation on the south, and Cottonwood Wash on

the west. The northern boundary is ca. 1 mi. south of the confluence of Cottonwood and Westwater canyons. U.S. Highway 163 roughly bisects the area from northeast to southwest (Fig. 3). The terrain of the Bluff Bench/ San Juan River area is varied. In the eastern two-thirds, the broad river floodplain (elevation ca. 1340 m.) ends abruptly at a ca. 120 m. bluff on the north that runs roughly parallel to the river. To the west the floodplain is narrow and the bluff is either less pronounced or nonexistent.

By Jock HMdtoy

Fig. 2 Bluff Bench/ San Juan River area, eastern (upper) and western zones.

1/2 I Mile By Jack Headley

R.22 E, R. 23 E. JM3774 N !3775 ' M \ M3776

,3720 Q3_643

3642Q^ J37>7 3639 M ^ 3 6 2 3 a f m a 3638 M 3196 M M 0 MiJi'

13192 3,90 l M 3 1 9 1 M 3I94 3S30 M M|3189 M T3632 3 6 3 6 n

T3629

flD37"

363C7]

«^ "

Q 3778/ ^ 323t ™

3583^ 3585(7) 52Q3BSL 3B70#

MURPHY

SiM377< 364)03780 364! ^ M M363t-37J32 M3869 M3242/

W&m

324lH

32403

«7

'3696

,MJ3

, P3

^ X

SI99| ' ;A3763 3785Q#3784 / H 3 7 8 6 •' 3198 / H 3 7 8 7 f RESERVATION

Fig. 3 White Mesa area.

SITE

^ \ /

TYPES

s v&' y ^# jt *^

*S> *Y

* *

D O

a n & *

<$> V

CULTURAL

£>>

<£

G X

AFFILIATIONS

•

M J?

•

Fig. 4 Legend: archeological site types and cultural affiliations.

/y fj*

*S35*P.

Fig. 5 San Juan River and lower Bluff Bench

Bluff Sandstone of Jurassic age is the most prominent rock unit in the area. Shales of the Summerville Formation and the resistant Entrada Sandstone are exposed below the Bluff Sandstone along the western and southeastern perimeters. Exposed rock units overlying the Bluff Sandstone include, in ascending order: Salt Wash Sandstone, Recapture Shale, and Westwater Sandstone â€” all members of the Morrison Formation. Salt Wash Sandstone and Recapture Shale on Bluff Bench are generally covered by recent sand dune and other losse, surficial deposits. Terrace gravels of late Tertiary age cap several hills on the bench above the bluff, and also occur at lower elevations along the edge of the San Juan River (Craig Goodnight, BLM geologist, 1975 personal communication).

Sandstone and the Burro Canyon Formation. The Morrison Formation is exposed on the western and southern portions of the mesa in Cottonwood and Westwater canyons (Craig Goodnight, 1975 personal communication). The mesa is relatively stable with little evidence of erosion. In contrast, Cottonwood Canyon is highly dessicated. Several springs and seeps on the mesa and in Corral Creek Canyon to the east afford the only perennial sources of water. Rainfall varies from ca. 7 in. per yr. along the San Juan River to ca. 12 in. per yr. on White Mesa. Vegetation varies from the mesic species of the river floodplain, where Tamarix predominates (Fig. 5), to a variety of xerophytic types on the bench (Fig. 6). Representative species include: snakeweed {Gutierrezia sarothrae), greasewood {Sarcobatus vermiculatus), 4-wing saltbush {Atriplex canescens), soap-weed {Yucca spp.). Indian rice grass {Oryzopsis hymenoides), Mormon tea {Ephedra sp.), desert holly {Berberis fremontii), and prickly pear cactus {Opuntia spp.). Vegetation on the mesa is limited almost exclusively to sage {Artemisia tridentata) and grasses (GRAMINEAE). Juniper {Juniperus sp.) and pinyon pine {Pinus edulus) are restricted to a narrow band along the rims of the canyons. Many varieties of flora found along the San Juan River and on the bench are noticeably absent on White Mesa.

The San Juan River is the major perennial water source, although a number of springs and seeps are present in the major canyons. The bluff is cut by a number of washes including Recapture, Calf, and Spring canyons that flow roughly south to the river. To the west Cottonwood, Buck Creek, and Butler washes are the dominant drainages. The terrain of White Mesa is relatively uniform consisting of a number of low, rolling hills. Elevation varies from 1,580 to 1,670 meters. Eolian and alluvial deposits, which vary in depth to 5 m., overlie Dakota

PREHISTORIC CULTURES OF THE REGION

Five distinct populations have at various times occupied the San Juan region during the course of prehistory. These include Paleo-Indian "big-game" hunters, Archaic hunter-gatherers, Anasazi (Pueblo) agriculturalists, and the proto-historic Numic and Navajo. Evidence of limited Paleo-Indian occupations of the region contrasts with that of the locally abundant Anasazi. The duration of the Archaic (used here as a temporal designator) is open to question. A brief overview of prehistoric agriculturalists, and Numic and Navajo cultures, directly applicable to this survey is presented (modified from Berry 1975). PaleoIndian and Archaic occupations in the region are not reviewed here (see Berry 1975; Madsen and Berry 1975).

of southeastern Nevada and southwestern Utah (Aikens 1965; Jennings 1975; Sharrock 1966). "Fremont" is used collectively to refer to agricultural groups who inhabited Utah (exclusively) north of the Colorado River. Five variants are recognized: the Great Salt Lake and Parowan variants of the Great Basin, and the Uinta, Sevier, and San Rafael variants of the Colorado Plateau (Marwitt 1970). The origin(s) and taxonomic level accorded the Fremont are open to question. The long held notion of Archaic-Fremont continuity in the northeastern Great Basin (Aikens 1970) has been effectively challenged (Madsen and Berry 1975). Cultural continuity is not clearly established elsewhere. Current evidence suggests Fremont occupations to be no closer to the survey area than the Henry Mountains (C. Helm 1974; C. Berry, 1975 personal communication), the Book Cliffs (Wormington 1955), and possibly the La Sal Mountains (Hunt 1953). Mesa Verde/Kayenta Anasazi. The cultural sequence for the San Juan region, established at the Pecos Conference by A. V. Kidder and others (1927), applies equally to the Mesa Verde and Kayenta Anasazi. The

PREHISTORIC AGRICULTURALISTS The abundant evidence of agricultural groups contrasts sharply with that of earlier populations. Two major populations, Anasazi and Fremont, are recognized (Fig.8). The three Anasazi branches are: the Mesa Verde of the Four Corners area; the Kayenta of northern Arizona and south-central Utah (north and south of the Colorado River); and the Virgin Branch

•

*±^.-**&#.

mt*

fj£ Skk

Fig. 6 Bluff Bench (looking east).

*S£=S$

HftMiwrffn -

Fig. 7 White Mesa with Abajo Mountains in background (looking north).

framework is used here in modified form (after Jennings 1968): Latest:

Earliest:

consisting of subrectangular pit structures, and cave/ rockshelters (Dalley 1973; Wilson 1974; Lindsay et al. 1968). They are located in river valleys, canyons and on mesa tops (Lipe and Matson 1971; Matheny 1975). Cultigens include the addition of beans to corn and squash. Dated sites on Cedar Mesa (Dalley 1973; Wilson 1974) and Navajo Mountain (Lindsay et aL 1968) range from 1350 to 1250 B.P. Pueblo I is represented by the addition of neckbanded gray, red-on-black, and red-on-orange wares. The bow and arrow replaced the atlatl and was used with small corner-notched projectile points. Wattleand-daub surface structures were added to the subrectangular pit structures. They are typically located on low hillocks in river valleys and in open areas on plateaus (Matheny 1975). Dates from the "type" Site Number 13 at Alkalai Ridge cluster around 1200 B.P (Brew 1946). Pueblo II settlements are typically represented by circular, subterranean pit structures associated with one to four surface masonry rooms (Brew 1946). The pit structures are considered as "kivas" associated with ceremonial activities. This function is inferred from observations of similar structures of the historic Pueblo. Ceramics include the predominance of blackon-white and corrugated culinary vessels. The bow and arrow remained in use, and corner-notched points were replaced by side-notched points. Pueblo II marks a period of expansion, dispersed site distribution, and small settlements (Brew 1946; Jennings 1975). During late Pueblo II times, sites

Pueblo V 250 B.P. to Present Pueblo IV 650 to 250 B.P. Pueblo III 850 to 650 B.P. Pueblo II 1150 to 850 B.P. Pueblo I 1200 to 1050 B.P. Basketmaker III 1500 to 1200 B.P. Basketmaker II 1950 to 1450 B.P. Basketmaker I Pre â€” 1950 B.P.

Basketmaker I was originally conceived as a preagricultural, preceramic, hunter-gatherer "stage" in the Southwest. The stage is now considered a part of the Archaic. Basketmaker II is defined by the presence of corn and squash and the absence of ceramics. Other characteristics include slab-lined storage features, use of the atlatl, milling stones, distinctive sandals, nets, cordage, and basketry. Basketmaker II sites are commonly found in cave/rock shelters (Jennings 1975; Kidder and Guernsey 1919; Morris and Burgh 1954; Nusbaum 1922). Open sites with pit structures have been recorded from Cedar Mesa (Lipe and Matson 1973) and Castle Wash (Sharrock et al. 1963) on the Red Rock Plateau (Lipe 1970). Radiocarbon and dendrochronological dating of the open sites places occupation between 1650 and 1750 B.P. Basketmaker III is defined by the addition of ceramics, both plain gray and black-on-white, to the Basketmaker II inventory. Sites are reported as open,

became more nucleated with a corresponding increase in village size (e.g., Dalley 1973; Lindsay et al. 1968; Lister 1960; Sharrock, Day, and Dibble 1963; Wilson 1974). Pueblo II sites, located in river valleys and on mesa tops (Lipe and Matson 1971; Matheny 1975), have been found as far north as the town of Escalante in central Utah (Lister, Ambler, and Lister 1960) and Arches National Park (Berry 1975; Hunt 1953). Pueblo III settlements are typically large, nucleated, open and cave/rockshelter villages, often consisting of a number of kivas with associated masonry habitation and storage structures. Multistoried structures are common. Pueblo III settlements are located in river valleys and on mesa tops clustered around canyon head springs (Lipe and Matson 1971; Matheny 1975). The large centers of Mesa Verde, Canyon de Chelley, Chaco Canyon, and Betatakin exemplify the larger settlements. Although there are no settlements of this magnitude in Utah, small Pueblo III sites abound (Dalley 1973; Lipe and Matson 1971, 1973; Sharrock, Day, and Dibble 1963; Wilson 1974), indicating high population densities. By ca. 650 B.P. the region was abandoned for reasons not yet adequately explained. Proto-Historic Numic and Navajo. Initial dispersal of the Numic (Uto-Aztecan speakers), ca. 1000 B.P., from a southern California homeland has been postulated (Lamb 1958; Miller, Tanner, and Foley 1969). This was followed by their expansion into the Great Basin by 900 B.P. and the Great Salt Lake region by 650 B.P. (Madsen 1975). Although arrival on the Colorado Plateau remains undated, the presence of Paiute-Shoshoni ware (in archeological contexts near the La Sal Mountains) suggests they were present during late-prehistoric times (Hunt 1953). It is estimated that Navajo Athabascan speakers arrived in the Southwest sometime before 450 B.P. (Hester 1962). Linguistic evidence suggests possible origins in the northwestern Plains-Woodland area

Fig. 8 Distribution of prehistoric agriculturalists (from Marwitt 1970. Berry 1975).

(Sapir 1916 cited from Mandelbaum 1949). By the time the Spanish arrived the Navajo were well established. Major archeological site concentrations occur in such diverse areas as the upper San Juan River drainage, Chaco Canyon, Canyon de Chelley, the Hopi Buttes, and as far north as Glen Canyon (Hester 1962).

ARCHEOLOGICAL SITES survey was apparently confined to the rim above Cottonwood Wash. The survey conducted by Brigham Young University represents a roughly north-south transect along the Pinto-Abajo power line across White Mesa. The entire sample is no doubt biased by accessability and preconceived notions of settlement patterns, i.e., a majority of the sites in the Bluff Bench/ San Juan River area are documented along the river and washes. The few .transects across the bench indicate that a number of sites probably remain undocumented. The survey of White Mesa represents a more intensive, complete inventory.

Sixty-seven archeological sites were identified by the Bureau of Land Management in the Bluff Bench/ San Juan River area and fifty-four on White Mesa. Twelve sites had previously been documented on White Mesa by the University of Utah, and fifteen have since been reported by Brigham Young University (Berge 1975) (Figs. 2, 3, and 4). The Appendix gives a full listing of all recorded archeological sites. No formalized survey strategies were employed. The limitations of time and manpower dictated the following of various transects along the river, washes, and across the bench and mesa. The University of Utah 13

Archeological phenomena encountered include: habitation sites that are, for the most part, open structural sites (except for a few rockshelters along the San Juan River); temporary encampments or seasonal use sites; granaries; "towers"; and steps. Pictograph/petroglyphs, quarry sites, and chipping stations (or lithic scatters) are also represented. The full range of Anasazi development, as well as the Navajo, are represented. The assessment of cultural affiliation and development are based on the ceramic typology for the Mesa Verde region (Breternitz, Rohn, and Morris 1974; Madsen 1973), apparent structures, and site layout (Brew 1946; Dalley 1973; Wilson 1974). For explanations of site types and cultural affiliations see Figs. 2, 3, and 4. A number of multicomponent sites are represented. All are habitation sites that contain more than one level of Anasazi cultural development (Basketmaker III through Pueblo III). The assessment is based principly on the ceramic sequence. For example, the presence of Mancos Black-on-white (Pueblo II) and Mesa Verde Corrugated (Pueblo III) indicates the presence of both components. Also, multicomponent sites may have been identified by the presence of Pueblo III ceramics found on "typical" Pueblo II sites. Sites documented as Pueblo I/II on Cedar Mesa are not considered multicomponent because of the difficulty in separating the two as shown in excavated and dated sites on Cedar Mesa (Dalley 1973). In all cases the designation of sites as multicomponent is based on surface survey â€” without artifact collection â€” and is extremely tenuous. Site reports which, for instance, indicate Pueblo 11/III suggest that either affiliation could not be determined or the site is of a transitional nature. Matheny (1975 personal communication) cautions that clear differentiation of Basketmaker III from Pueblo I and late Pueblo II from early Pueblo III is impossible based on ceramics alone. Additional evidence, including dendrochronology, is required for accurate assessment. The data in this report represent the best possible estimates until further work is accomplished. The multicomponent category is included to indicate the range of possible developments at each site or to show that the assessment is open to question.

Preceramic Archaic and Basketmaker II developments are treated as one cultural unit. Distinguishing between the two is impossible without excavation and subsequent definition of the presence or absence of plant domestication. It is difficult to differentiate preceramic components from temporary encampments or camp sites of seasonal occupance of later developments. No Archaic/ BM II atlatl points (a principal preceramic diagnostic) were found in either survey area. The definition of preceramic sites is based entirely on the presence of large amounts of chipped and ground stone and fire-cracked rock in areas devoid of later, more sedentary Anasazi occupations, and are differentiated from the special use areas of the later developments. In addition, the majority are located in sand dune contexts, a preferred location for Archaic settlements in the Great Basin (Madsen 1976). Archaic/BM-II sites on the Bluff Bench (e.g., 42Sa 3213, 3223, 3226-3228) are located primarily in deflated sand dune areas along intermittent streams and washes. None were located immediately adjacent to the San Juan River. Sites documented on White Mesa (42Sa 3237, 3582, 3782) are located roughly in the center of the mesa. Because of their close proximity to later Anasazi sites (42Sa3240, 3641, 3780) the three are less likely preceramic than those reported from the Bluff Bench. A few temporary encampments and chipping stations on the bench (e.g., 42Sa3225, 3226, 3592) and mesa (e.g., 42Sa3192, 3611, 3783) of unknown cultural affiliation could be preceramic, however, the lack of diagnostics and the presence of a few unidentified ceramic sherds in close proximity precludes that assignation.

BASKETMAKER III Nineteen BM-III components are documented in the Bluff Bench/San Juan River area (25.7%) and 18 components on White Mesa (14.6%). BM-III is represented in 10 multicomponent sites on the bench (e.g., 42Sa3221, 3224, 3579) and 12 on the mesa (e.g., 42Sa 3208, 3238, 3242). The sites are documented as both temporary encampments and habitation sites on the bench (e.g., 42Sa3245, 3596, 3813) (Fig. 9) and mesa (e.g., 42Sa3193, 3199, 3870). Habitation sites are determined by the presence of slab-lined storage features (Fig. 9) and depressions suggestive of pit structures. BM-III Chapin and Lino Gray wares are the principle diagnostic ceramics (Breternitz, Rohn, and Morris 1974; Madsen 1973). The sites are distributed along the San Juan River, intermittent streams and washes on the bench, and along the rim above Cottonwood Wash on the mesa. The few temporary encampments and chipping stations, particularly on the bench, could be the product of BM-III occupations.

In the ensuing discussion the number of Anasazi components is arrived at through the tabulation of components from both single and multicomponent sites. The relative percentages are expressed vis-a-vis Archaic/ BM-II and known Anasazi components for each area.

ARCHAIC/BASKETMAKER II Twelve Archaic/BM-II sites are documented in the Bluff Bench/San Juan River area and three on White Mesa. These represent 16.2% and 2.4% of the Anasazi sites, respectively.

PUEBLO I/II

NAVAJO

Sixteen P-I components are documented in the Bluff Bench/San Juan River area (21.6%) and 20 components on White Mesa (16.3%). Fifteen P-II components are documented in the Bluff Bench/San Juan River area (20.3%) and 38 on White Mesa (30.9%). P-I is represented in 12 multicomponent sites on the bench (e.g., 42Sa3217, 3219, 3224) and 12 on the mesa (e.g., 42Sa3208, 3238, 3242). P-II is represented in 14 multicomponent sites on the bench (e.g., 42Sa3217, 3219, 3224) and 23 on the mesa (e.g., 42Sa3208, 3600, 3610). P-I sites are documented as both temporary encampments and habitation sites on the bench (e.g., 42Sa3220, 3230) and mesa (e.g., 42Sa3239, 3240). Similarly, P-II sites are documented on the bench (e.g., 42Sa3211, 3805) and mesa (e.g., 42Sa3240, 3777). In the absence of excavation, internal site analysis, and absolute dating no distinction is made between the two components (Figs. 2, 3 and 4). Dalley (1973) has shown that diagnostic P-I ceramics persist well into P-II times on Cedar Mesa. Without adequate dating, P-I designated sites may well be the later development. The majority of both P-I and P-II sites are recognized as habitation sites with depressions (pit structures) and surface masonry added to the latter. P-I Abajo Red-on-orange and La Plata Black-on-red and P-II Mancos Black-on-white and Corrugated are the principle diagnostic ceramics (Breternitz, Rohn, and Morris 1974; Madsen 1973). A few P-I/II sites are located along the San Juan River in the vicinity of Recapture Wash and along intermittent streams and washes on the Bluff Bench. However, the majority are distributed across White Mesa.

Eight Navajo sites are documented in the Bluff Bench/San Juan River area; none are on White Mesa. A majority of the sites consist of circular earthen rings ca. 7 m. dia., which may be the remains of "hogans." In some cases remnant timber is present.

PUEBLO HI Twelve P-III sites are documented in the Bluff Bench/San Juan River area (16.2%) and 44 on White Mesa (35.8%). P-III is represented in eight multicomponent sites on the bench (e.g., 42Sa3802, 3806) and 20 sites on the mesa (e.g., 42Sa3208, 3600, 3610). The majority are located along the San Juan River and across White Mesa. The greater number of sites are documented as habitation sites on the bench (e.g., 42Sa3803, 3804) and mesa (e.g., 42Sa3608, 3637) (Fig. 10). Habitation sites were determined by the presence of depressions (kivas) amid extensive surface masonry. These were associated with P-III Mesa Verde Black-on-white and Corrugated ceramics (Breternitz, Rohn, and Morris 1974; Madsen 1973). "Towers" on White Mesa (e.g., 42Sa3632, 3634) are probably P-III despite the absence of ceramics and are certainly no earlier than P-II (Winter 1974).

Fig. 9 Basketmaker III site (42Sa3813) on the lower Bluff Bench showing rock alignments of storage features and other cultural debris (upper) and closeup of vandalized slab-lined storage bin. 15

The structures were found in clusters of two or three, with four maximum. A few small "sweat-lodges" are associated with the hogans. Large amounts of refuse including metal and glass are present. The absence of pottery suggests the sites post-date the nineteenth century. The majority of sites are located in the vicinity of the heavy Anasazi occupation at Recapture Wash (e.g., 42Sa3220, 3868). Two pictograph/petroglyph sites (42Sa3218, 3589) that display horses are probably Navajo. Three rather enigmatic sites (42Sa3586-3588) on the bluff which overlook the river and consist of stone rings (ca. 3-4 m. dia.) lying on terrace gravels may also be Navajo (see Winter 1975).

DISCUSSION The Bluff Bench/San Juan River and White Mesa areas contrast in site densities and relative percentages of components documented in each area. The 67 sites in the 30 square mile area of the Bluff Bench constitute a density of 2.2 sites per square mile. The 54 sites on the 6 square mile area of White Mesa suggest a density of at least 9 sites per square mile. The contrast is further emphasized by the number of components computed from both single and multicomponent sites. Seventy-five components or 2.5 components per square mile, are computed for the bench/river area and 20.5 per square mile for the mesa. Relative percentages of components contrast over time between the two areas and in general differ inversely (Fig. 11). The relatively heavy Archaic/BM-II (16.2%) on the bench is virtually absent (2.4%) on the mesa although the later, heavier developments may have obscured the earlier, Both BM-III (24.7 and 14.6%) and P-I (21.6 and 16.3%) on the bench and mesa, respectively, are fairly evenly represented. Marked differences occur with P-II (20.3 and 30.9%), culminating with P-III (16.2 and 35.8%) on the bench and mesa. In the absence of rigorous analysis any attempt to sort cultural/environmental variables in contrasting the two areas is shrouded by speculation. Several climatic/ vegetational factors seem important when interpreted within the framework of changing Anasazi subsistence strategies and concommitant settlement patterns. Minimally, these factors are: 1. The great variety of flora immediately adjacent to and along the San Juan River, where both riverine and desert communities are represented, better supported limited populations partially dependent on food collecting. Access to several resource zones including the Glen Canyon (Jennings 1966) and associated Red Rock Plateau (Lipe 1970), Cedar Mesa (Camilli 1975; Lipe and Matson 1971), and Hovenweep (Winter 1974,1975) dictated settlement elsewhere. 2. The deep, sandy soils and limited precipitation (7 in. per yr.) on the bench/river probably limited

Fig. 10 Pueblo III site (42Sa3637) on White Mesa (upper) and closeup of rock alignment.

agriculture to floodwater farming not only for Basketmaker populations (Lipe and Matson 1971) but the later Pueblo as well. 3. The higher precipitation levels (12 in. per yr.) and the deep, well-developed soils of the mesa probably provided a suitable location for Pueblo populations more dependent on agriculture and able to engage in the more recent dry farming techniques (Lipe 1970; Lipe and Matson 1971). 4. Today's limited variety of flora on the mesa, unlike the abundant varieties along the San Juan, would not have met the needs of a population dependent upon gathering. 5. The different growing seasons â€” 145 frost free days at Blanding to the north and 160 days at Bluff to the south â€” would not be a factor. These differences, interpreted in the context of Anasazi developments, suggest that the Archaic BM-II cultures heavily occupied areas along the San Juan River and adjacent bench and relied on resources from riverine and adjacent eco-zones. Agriculture, introduced in Basketmaker II times, was probably confined to the floodplain along the San Juan and areas where the ground water is ample. Dry farming was not practiced until Basketmaker III times (Lipe and Matson 1971). The comparable percentages between the two areas during Basetmaker III and Pueblo I times reflect a general movement to higher resource zones where more effective moisture was available for dry farming. This indicates an increasingly greater reliance on agriculture by BM-III and P-I populations. The balance between collecting and agricultural production probably had shifted markedly by P-I times. The different percentages for the two areas during Pueblo II and III times probably reflects an increasing

reliance on agriculture and concomitant higher productivity of the mesa. During P-II times mesa tops were also farmed in the vicinity of Dark Canyon where a series of "check dams" have been reported (Brooks 1974), Cedar Mesa (Lindsay 1974; Lipe and Matson 1971), Hovenweep (Winter 1974, 1975), and Mesa Verde (Herold 1961). In contrast, dry farming was apparently not practiced on the Red Rock Plateau where precipitation falls below 10 in. per yr. (Lipe 1970). The gradual increase in the number of sites, particularly multicomponent, and high site densities through P-III times reflect the stability of a long term successful adaptation to the mesa by sizeable agricultural populations. Where lower precipitation occurred along the San Juan, agriculture was probably limited to areas of sizeable floodplain such as Recapture and Cottonwood washes where the few nucleated P-III settlements are reported. Large settlements also occur just west of the survey area at the mouth of Comb Wash. Elsewhere the floodplain is narrow and P-III sites are absent. In sum, the limited P-III occupation of the river probably reflects constraints imposed upon agriculture by the combination of low precipitation levels and lack of arable land. Throughout the region during P-III times, the main subsistence focus was floodwater farming of bottomlands (Lipe 1970; Herold 1961; Winter 1974, 1975). However, both dry and floodwater farming continued in shallow washes and deep alluvial soils (Camilli 1975; Lipe and Matson 1971; Winter 1974). The P-III settlement on White Mesa indicates that dry farming continued to be important despite the changes noted elsewhere in the Anasazi region.

AREAL COMPARISONS more comparable with the exception of the heavily represented P-III component. Settlement data from the northern (Dalley 1973; Wilson 1974) and southwestern portions of Cedar Mesa (Lipe and Matson various, 1975 personal communication) differ from each other and elsewhere in the region. To the south, extremely heavy BM-II percentages are reported followed by gradual decreases until the P-II expansion. To the north the BM-II component is almost entirely lacking, however, a BM-III component is heavily represented followed by P-II expansion. In both Cedar Mesa areas the P-I component is virtually absent. P-III representations are similar and roughly

Settlement patterns in the Bluff Bench/San Juan River and White Mesa areas differ somewhat from those of adjacent regions (Fig. 11). Changes in the settlement of Montezuma Canyon (Matheny 1962, 1975 personal communication) probably more closely represent the general Southwestern pattern: low percentages of Archaic/BM-II, followed by gradual increases, culminating in the P-II expansion. This is followed by nucleation, decrease in number, and corresponding increases in size of sites during P-III times. Similar patterns are reported from Hovenweep (Winter 1974, 1975) and North Cottonwood Canyon (Davis 1975). Patterns elsewhere deviate from these, although White Mesa is

times populations were evenly distributed throughout the region. The unexplained abandonment of Pueblo I populations on Cedar Mesa and the subsequent movement to the higher resource zones of Elk Ridge are an exception to this pattern (see Lipe and Matson for discussion 1975). Pueblo II is well represented everywhere with the exception of Elk Ridge where Pueblo II levels fall well below the earlier Pueblo I, and on the Bluff Bench which shows a slight decline. This expansion is probably the result of improved dry farming techniques and ameliorating climate allowing for more intensive occupation of mesa environs. Generally humid conditions, probably similar to the present, existed during P-II times (ca. 1050 to 850 B.P.) (Sears 1953). Pueblo III settlement follows a pattern of nucleation everywhere but on Cedar and White mesas where the P-II pattern of numerous small, semi-autonomous Pueblo villages extends into P-III times. This is explained in part by climatic deterioration beginning ca. 760 B.P. (Lancaster and Van Cleave 1954), and culminating in the "Great Drouth," 678 B.P. (Schulman 1947, 1954) which undoubtedly contributed to the demise of the Anasazi in the San Juan region. The adaptation of Pueblo III populations to higher resource zones (Cedar and White mesas) included both dry farming and supplimental hunting and gathering. This cushioned the debilitating effects of climatic flux and general overall deterioration felt elsewhere throughout the region. The persistence of Pueblo II settlement patterns into Pueblo III times on the westerly mesas may also be explained in terms of the diminution of Mesa Verdean influence. With the exception of a few rockshelters (e.g., Butler Wash) which remain small by Mesa Verde and Chaco standards, canyon bottom and rockshelter villages do not preclude the continued and intense occupation of the mesas. This lessening influence finds support in the appearance of Kayenta pottery styles on Cedar Mesa and elsewhere to the west during Pueblo III times (see Wilson 1974; Lipe 1970). Pueblo III mesa top settlements may have both adaptive and socio-political significance.

correspond with those on White Mesa. However, site densities differ, varying between 6 and 15 sites per sq. mi. (Camilli 1975) to 30 per sq. mi. (Lipe and Matson 1971) on Southern Cedar Mesa with 9 sites per sq. mi. on White Mesa. In the vicinity of Elk Ridge, located immediately north of Cedar Mesa, Archaic/BM-II sites are virtually absent and BM-III is only minimally represented (Evan DeBloois, 1975 personal communication). In contrast to Cedar Mesa, the majority of the Elk Ridge sites are P-I. The limited P-II component is roughly comparable in all areas. The decrease in the number of P-III sites on Elk Ridge compares with Montezuma and North Cottonwood canyons, Hovenweep, and the Bluff Bench/San Juan River area. Questions of causality underlying these regional differences are for the most part ignored because of a number of methodological problems (these by no means exhaust the avenues of critique): 1. Investigator biases and interests (compare Lipe and Matson 1971 with Dalley 1973, Matheny 1972, and Wilson 1974) (Fig. 11); 2. Thoroughness of surveys and sampling (compare Lipe and Matson 1971 with this study); 3. Interpretation of diagnostics and the classification of transitional sites (compare Matheny 1962 and Davis 1975 with this study); 4. Amount of excavation, dating, and internal site analysis (compare Dalley 1973 and Wilson 1974 with Winter 1974, 1975 and this study). Nowhere are sampling, interpretive models, and amount of excavation comparable. The great diversity of settings (riverine and canyon bottom to mesa top) and the variation in precipitation (ca. 5 in. per yr. to 20 in. per yr.) also provide some latitude for a variety of subsistence strategies and concomitant settlement patterns. Despite these factors, several points emerge from the comparative data. Preceramic Archaic/Basketmaker occupation of the region is heaviest on the Bluff Bench and southern Cedar Mesa. Drought conditions from ca. 1949 to 1450 B.P. (Schulman 1952) probably required utilization of all resource zones from riverine to pinyon/juniper on the mesa tops. During Basketmaker III/Pueblo I

APPENDIX BLUFF BENCH/SAN JUAN RIVER SITES Number 42 Sa 3211 42 Sa 3213 42 Sa 3217 42 Sa 3218 42 Sa 3219 42 Sa 3220 42 Sa 3221 42 Sa 3222 42 Sa 3223 42 Sa 3224 42 Sa 3225 42 Sa 3226 42 Sa 3227 42 Sa 3228 42 Sa 3229 42 Sa 3230 42 Sa 3231 42 Sa 3232 42 Sa 3233 42 Sa 3234 42 Sa 3235 42 Sa 3236 42 Sa 3244 42 Sa 3245 42 Sa 3246 42 Sa 3247 42 Sa 3248 42 Sa 3580 42 Sa 3581 42 Sa 3586 42 Sa 3587 42 Sa 3588 42 Sa 3589 42 Sa 3590 42 Sa 3591 42 Sa 3592 42 Sa 3596 42 Sa 3597 42 Sa 3598 42 Sa 3601 42 Sa 3602 42 Sa 3603 42 Sa 3605 42 Sa 3606 42 Sa 3607 42 Sa 3721 42 Sa 3722 42 Sa 3723 42 Sa 3797 42 Sa 3798

Type

Affiliation/ Development

Temporary Encampment Temporary Encampment Multicomponent Pictograph/ Petroglyph Multicomponent/ Steps Habitation Multicomponent Habitation Temporary Encampment Multicomponent Temporary Encampment Temporary Encampment Temporary Encampment Temporary Encampment Temporary Encampment Temporary Encampment Temporary Encampment Temporary Encampment Temporary Encampment Temporary Encampment Granary Temporary Encampment Temporary Encampment Temporary Encampment Temporary Encampment Temporary Encampment Temporary Encampment Temporary Encampment Granary Temporary Encampment Temporary Encampment Temporary Encampment Pictograph/ Petroglyph Habitation Quarry/Chipping Station Chipping Station Habitation Multicomponent Chipping Station Chipping Station Chipping Station Chipping Station Multicomponent Multicomponent Multicomponent Multicomponent Habitation Multicomponent Multicomponent Multicomponent

P-I, II Preceramic P-I, II, III Unknown (Navajo?) P-I, II, III, Navajo P-I, Navajo BM-III, P-l BM-III Preceramic BM-III, P-I, II, III Unknown Unknown Preceramic Preceramic BM-III P-I BM-III Preceramic Preceramic Preceramic Unknown Preceramic Preceramic BM-III Unknown Preceramic BM-III Preceramic Unknown Unknown (Navajo?) Unknown (Navajo?) Unknown (Navajo?) Unknown (Navajo?) BM-III Unknown Unknown BM-III BM-III, P-I Unknown Unknown Unknown Unknown BM-III, P-I, II BM-III, P-I, II BM-III, P-I, II, HI P-II, III BM-III P-II, III BM-III, P-I, II BM-III, P-I, II

Number

Type

Affiliation/Development

42 Sa 3627 42 Sa 3628 42 Sa 3629 42 Sa 3630

Habitation Tower Tower Multicomponent Tower Multicomponent Tower Habitation Tower Habitation Habitation/Tower Habitation Multicomponent Multicomponent Habitation Habitation Habitation Habitation Habitation Multicomponent Temporary Encampment Multicomponent Habitation

P-III P-III P-III P-II, III P-III P-II, III P-III P-III P-III P-III P-III P-III P-II, III P-II, III P-III P-III P-III P-III P-III BM-III, P-I BM-III BM-III, P-I, II P-II

42 Sa 3631 42 Sa 3632 42 Sa 3633 42 Sa 3634 42 Sa 3635 42 Sa 3636 42 Sa 3637 42 Sa 3638 42 Sa 3639 42 Sa 3640 42 Sa 3641 42 Sa 3642 42 Sa 3643 42 Sa 3799 42 Sa 3869 42 Sa 3870 42 Sa 3871 42 Sa 3872

EYED BY THE UNIVERSITY OF UTAH 42 Sa 3186 42 Sa 3187 42 Sa 3190 42 Sa 3191 42 Sa 3192 42 Sa 3193 42 Sa 3194 42 Sa 3195 42 Sa 3196 42 Sa 3198 42 Sa 3199 42 Sa 3200

Multicomponent Multicomponent Multicomponent Multicomponent Temporary Encampment Habitation Multicomponent Habitation Multicomponent Multicomponent Habitation Multicomponent

P-II, III P-I, II, III P-II, III P-II, III Unknown BM-III P-II, III Unknown BM-III, P-I BM-III, P-I BM-III BM-III, P-I, II

YED BY BRIGHAM YOUNG UNIVERSITY 42 Sa 3696 42 Sa 3774 42 Sa 3775 42 Sa 3776 42 Sa 3777 42 Sa 3778 42 Sa 3779 42 Sa 3780 42 Sa 3781 42 Sa 3782 42 Sa 3783 42 Sa 3784 42 Sa 3786 42 Sa 3787 42 Sa 3785

Habitation Multicomponent Multicomponent Multicomponent Temporary Encampment Habitation Multicomponent Habitation Habitation Temporary Encampment Chipping Station Temporary Encampment Habitation Habitation Temporary Encampment 21

P-II P-II, III BM-III, P-I, II P-II, III P-II P-I, II P-I, II, III P-I, II P-I, II Preceramic Unknown BM-III P-II P-II P-III

REFERENCES Aikens, C. Melvin 1965 Excavations in Southwest Utah. University of Utah Anthropological Papers, No. 76, Glen Canyon Series, No. 27. Salt Lake City.

Hester, James J. 1962 Early Navajo Migrations and Acculturation in the Southwest. Museum of New Mexico Papers in Anthropology, No. 6. Santa Fe. Hunt, Alice P. 1953 Archeological Survey of the La Sal Mountain Area, Utah. University of Utah Anthropological Papers, No. 14. Salt Lake City.

Berge, Dale L. 1975 Archeological Survey of the Pinto-Abajo Transmission Line, Southeastern Utah. A Special Report, Brigham Young University. Provo.

Jennings, Jesse D. 1966 Glen Canyon: A Summary. University of Utah Anthropological Papers, No. 81. Glen Canyon Series, No. 31. Salt Lake City.

Berry, Michael S. 1975 An Archeological Survey of the Northeast Portion of Arches National Park. Antiquities Section Selected Papers, Vol. I, No. 3. Salt Lake City.

1968

Breternitz, David A., A. H. Rohn, Jr., and E. A. Morris 1974 Prehistoric Ceramics of the Mesa Verde Region. Museum of Northern Arizona, Ceramic Series, No. 5. Flagstaff.

The Prehistory of North Hill, New York.

America.

McGraw-

1975 Prehistory of the Eastern Great Basin. University of Utah Anthropological Papers, in press. Salt Lake City. Kidder, Alfred V. 1924 An Introduction to the Study of Southwestern Archaeology. Papers of the Southwestern Expedition, No. 1. Phillips Academy, Andover.

Brew, John Otis 1946 Archaeology of Alkali Ridge, Southeastern Utah. Papers of the Peabody Museum of American Archeology and Ethnology, Harvard University, Vol. 21. Cambridge.

1927 The Southwestern Archaeological Conference. El Palacio, Vol. 23, No. 2. Santa Fe.

Brooks, Danny 1974 "Prehistoric Soil and Water Control in the American Southwest: A Case Study." Master's thesis, Northern Arizona University. Flagstaff.

Kidder, Alfred V. and Samuel J. Guernsey 1919 Archaeological Explorations in Northeastern Arizona. Bureau of American Ethnology, Bulletin 65. Washington.

Camilli, Eileen 1975 "Prehistoric Settlement Pattern on Cedar Mesa, Southeastern Utah." Master's thesis, Northern Arizona University. Flagstaff.

Lamb, Sidney M. 1958 Linguistic Prehistory in the Great Basin. International Journal of American Linguistics, Vol. 24, No. 2. Baltimore.

Dalley, Gardiner F., ed. 1973 Highway 95 Archaeology: Comb Wash to Grand Flat, Vol. I. A Special Report, Department of Anthropology, University of Utah. Salt Lake City.

Lancaster, James A., and Philip F. Van Cleave 1954 Excavations of Sun Point Pueblo. Archeological Excavations in Mesa Verde National Park, Colorado 1950. U.S. National Park Service Archeological Research Series, No. 2. Washington. Madsen, David B. 1975 Dating Paiute-Shoshoni Expansion in the Great Basin. American Antiquity, Vol. 40, No. 1. Washington. 1976 Bulldozer Dune (42SL46). Antiquities Section Selected Papers, Vol. II, No. 9. Salt Lake City.

Davis, Larry D. 1975 "An Archaeological Survey of North Cottonwood Canyon, San Juan County, Southeastern Utah." Master's thesis, Brigham Young University. Provo. Helm, Claudia J. 1974 Preliminary Report of an Archaeological Survey in Sevier, Emery and Garfield Counties, Eastern Utah. A Special Report, Department of Anthropology, University of Utah. Salt Lake City.

Madsen, David B. and Michael S. Berry 1975 A Reassessment of Northeastern Great Basin Prehistory. American Antiquity, Vol. 40, No. 4. Washington. Madsen, Rex 1973 Ceramics from the U-95 Sites. In Highway 95 Archeology: Comb Wash to Grand Flat, Vol. I, edited by Gardiner F. Dalley. A Special Report, Department of Anthropology, University of Utah. Salt Lake City.

Herold, Joyce 1961 Prehistoric Settlement and Physical Environment in the Mesa Verde Area. University of Utah Anthropological Papers, No. 53. Salt Lake City.

Lindsay, Alexander J., Jr., et al. 1968 Survey and Excavation North and East of Navajo Mountain, Utah, 1959-1962. Museum of Northern Arizona Bulletin, No. 42. Flagstaff; Lindsay, LaMar W. 1974 Preliminary Palynological Studies on Cedar Mesa. In Highway U-95 Archeology: Comb Wash to Grand Flat, Vol. II, edited by C. J. Wilson. A Special Report, Department of Anthropology, University of Utah. Salt Lake City. Lindsay, LaMar W. and Richard E. Fike 1974 Archaeological Survey of Bureau of Land Management Utah State Exchange Lands (U18819, U-18820) on the Bluff Bench and (U18821) on White Mesa, San Juan County, Utah. A Special Report, Bureau of Land Management. Salt Lake City.

Miller, Wick R., James Tanner, and Lawrence Foley 1969 A Lexicostatistic Study of Shoshone Dialects. Anthropological Linguistics, Vol. 13, No. 4. Bloomington. Morris, E. H., and R. F. Burgh 1954 Basketmaker II Sites Near Durango, Colorado. Carnegie Institution of Washington Publication, No. 604. Washington. Sapir, Edward 1916 Internal Linguistic Evidence Suggestive of the Northern Origin of the Navajo. American Anthropologist, Vol. 38. (Reprint 1949, Mandelbaum). Schulman, Edmund 1947 An 800-Year Douglas Fir at Mesa Verde National Park. Tree-Ring Bulletin, Vol. 14, No. 1. Tucson. 1952 Extension of San Juan Chronology to B.C. Times. Tree-Ring Bulletin, Vol. 18, Nos. 2 and 3. Tucson. 1954 Dendroclimatic Changes in Semiarid Regions. Tree-Ring Bulletin, Vol. 20, Nos. 3 and 4. Tuscon.

Lipe, William D. 1970 Anasazi Communities in the Red Rock Plateau, Southeastern Utah. In Reconstructing Prehistoric Pueblo Societies, edited by W. A. Longacre. School of American University Research of New Mexico. Albuquerque.

Sears, Paul B. 1953 Climate and Civilization. In Climatic Change, edited by Harlow Shapley. Harvard University, Cambridge.

Lipe, William D. and R. G. Matson 1971 Human Settlement and Resources in the Cedar Mesa Area, Southeast Utah. The Distribution of Prehistoric Population Aggregates. Proceedings of the Southwestern Anthropological Research Group Report, No. 1. Prescott. 1973 Progress Report on Year I of the Cedar Mesa Project. A Progress Report to the National Science Foundation. Flagstaff. 1975 Archaeology and Alluvium in the Grand Gulch-Cedar Mesa Area, Southern Utah. Four Corners Geological Society Guidebook, 8th Field Conference. Canyonlands.

Sharrock, Floyd W. 1966 An Archeological Survey of Canyonlands National Park. University of Utah Anthropological Papers, No. 83, Miscellaneaous Paper, No. 12. Salt Lake City. Sharrock, Floyd W., Kent C. Day, and David S. Dibble 1963 1961 Excavations, Glen Canyon Area. University of Utah Anthropological Papers, No. 63, Glen Canyon Series, No. 16. Salt Lake City. Wilson, Curtis J., ed. 1974 Highway 95 Archaeology: Comb Wash to Grand Flat, Vol. II. A Special Report, Department of Anthropology, University of Utah. Salt Lake City.

Lister, Robert H., J. R. Ambler, and F. C. Lister 1960 The Coombs Site, Part II. University of Utah Anthropological Papers, No. 41, Glen Canyon Series, No. 8. Salt Lake City. Mandelbaum, David G., ed. 1949 Selected Writings of Edward Sapir. versity of California, Los Angeles.

Winter, Joseph C. 1974 Hovenweep 1974. San Jose State University Archeological Report, No. 1. San Jose. 1975 Preliminary Report: Hovenweep 1975. A Special Report, Department of Anthropology, San Jose Stale University. San Jose.

Uni-

Marwitt, John P. 1970 Median Village and Fremont Culture Regional Variation. University of Utah Anthropological Papers, No. 95. Salt Lake City.

Wormington, H. Marie 1951 Prehistoric Indians of the Southwest. Denver Museum of Natural History, Popular Series, No. 7. Denver. 1955 A Reappraisal of the Fremont Culture. Proceedings of the Denver Museum of Natural History, No. 2. Denver.

Matheny, Ray T. 1962 "An Archaeological Survey of Upper Montezuma Canyon, San Juan County, Utah." Master's thesis, Brigham Young University. Provo. 1975 Cultural Developments in the Central Colorado Plateau. Archaeological Survey of the Pinto-Abajo Transmission Line, Southeastern Utah. A Special Report, Brigham Young University. Provo.

PINT-SIZE SHELTER by LA MAR W. LINDSAY and CHRISTIAN K. LUND

Antiquities Section Division of State History State of Utah with appendices by LA MAR W. LINDSAY and DONALD R. CURREY Department of Geography University of Utah

ANTIQUITIES SECTION SELECTED PAPERS Number 10

PINT-SIZE SHELTER by LA MAR W. LINDSAY and CHRISTIAN K. LUND

Antiquities Section Division of State History State of Utah with appendices by LA MAR W. LINDSAY and DONALD R. CURREY Department of Geography University of Utah

ANTIQUITIES SECTION SELECTED PAPERS Number 10

PREFACE Excavations at Pint-Size Shelter were conducted as part of a continuing program of contract salvage work conducted by the Antiquities Section, Division of State History. Field work was carried out from September 26 to October 10, 1975, by Michael S. Berry, La Mar W. Lindsay, and Christian K. Lund under an agreement with the Consolidation Coal Company. Donald R. Currey and James H. Madsen of the University of Utah provided interpretive assistance with, respectively, the geological sequence and the chipped stone analysis. Winston Gamble and John Whittaker assisted with the laboratory analysis. Laurie M. George, Suesan W. Taylor, and Amy R. Pringle aided in the preparation of the manuscript. David B. Madsen, State Archeologist September 1976

ILLUSTRATIONS

TABLE OF CONTENTS Preface

Page 27

Abstract

Introduction

Location and Setting Geology Climate Vegetation Fauna

31 31 31 32 33

Previous Research

Site Description Setting Excavation and Stratigraphy Cultural Features Summary and Discussion

35 35 35 36 37

Cultural Units

Dating

Artifacts Chipped Stone Ground Stone Ceramics Artifact Summary

40 40 51 54 55

Site Interpretation and Discussion

Figure Page 1. General site location 30 2. Aerial photograph of Castle Valley location 32 3. Pint-Size Shelter in Castle Valley 32 4. Cultural debris eroding out of the sand dune below the rockshelter 33 5. Plan and contour map 34 6. Site location and the limits of excavation 35 7. North-south profile of Pint-Size Shelter deposits 36 8. East-west profile of Pint-Size Shelter deposits 37 9. Site stratigraphy and Pit 1 38 10. Pit 1 outline and rodent disturbance 39 11. Pit 2 â€” Stratigraphic context 41 12. Pit2(closeup) 41 13. Fire hearth 41 14. Composite profile of cultural features and stratigraphy 42 15. Points 43 16. Drills 44 17. BifacesTypel, II 45 18. Bifaces Type III, IV 46 19. Bifaces Type V, VI, VII 48 20. UnifacesType I, II 49 21. UnifacesType III 50 22. Magnification: uniface and utilized flake 50 23. Hammerstones and cores 51 24. Slab metate; trough metate and mano 52 25. Manos 53 26. Ceramics 54 27. Schematic north-south profile of Pint-Size Shelter 61 28. North bedrock face of Pint-Size Shelter 63 29. Comparison of relative glaciation and depositional sequence at Pint-Size Shelter . . . . 64 30. Pollen diagram 69

Appendix I: Geomorphology of Pint-Size Shelter .. 59 Abstract 59 Introduction 59 Earlier Geomorphic Episodes 59 Late Quaternary Geomorphic Episodes 60 Evidence Provided by Epipetric Cortex Thicknesses 62 Geomorphic-Paleoclimatic Interpretation of Episodic Cover-Sand Accretion 65 Appendix II: Site Paleoecology: Palynology and Macrofossil Analysis Abstract Introduction Procedure Results The Pollen Record and Modern Vegetation . . . Temporal Variation Cultural Preferences Site Paleoecology

67 67 67 67 68 68 70 70 71

References

TABLES Table Page I. Distribution of Diagnostic Artifacts 38 II. Artifact provenience 55 III. Selected summary properties of sediments sampled from Pint-Size Shelter and nearby localities 60 IV. Summary of Holocene geomorphic episodes, at Pint-Size Shelter, with possible correlations 62 V. Thickness of CaCo 3 epipetric cortex on samples from bedrock face at north edge of Pint-Size Shelter 63

••V ;:V:i : l$:dvP : l : **i|i:l;l •

•

0 ''

Sft"

-—-^ p 'tjn,--:-:-

^^H

^JTA

s t\

s £ <*•

^ S^"

Fig. 1 Map: General site location.

^ % ^ wm#m

«S

ABSTRACT Pint-Size Shelter is a dual-component Archaic-Fremont site in Castle Valley, Utah. Radiocarbon determinations of 4520*210 and 3390*170 B.P. essentially bracket the Archaic deposit. The latter determination dates a Gypsum "atlatl" point. These dates closely correspond to Gypsum "phases" at Sudden and O'Malley Shelters. A 1790+100 B.P. determination, dates Emery gray pottery and the onset of Fremont occupance. Thickness variations in petrocalcic cortex (Currey, Appendix I) coincides with the stratigraphy. These data suggest a ca. 1500 yr. occupational hiatus separating the two cultural components and is similar to that of the northeastern Great Basin. Paleoecological data at Pint-Size suggest both Archaic hunter-gatherers and Fremont horticultural populations utilized the site for similar purposes.

INTRODUCTION Lake Base and Meridian. Archeological salvage excavation was required preparatory to coal mining operations, in fulfillment of a contract between the Consolidation Coal Company of Denver, Colorado, and the State Antiquities Section, Utah Division of State History.

The Pint-Size Shelter Site (42Em625) is a small, dual-component, Archaic and Fremont rockshelter on Christiansen Wash at elevation 1850 m., ca. 5.25 km. south of the town of Emery, Emery County, Utah (Fig. 1). The site is located in the SW'/4, SW'/4, Section 27, T22S, R6E (U.S.G.S. Mesa Butte 7.5 min. Quadrangle) at lat. 38Â°52'24" N, long. 111Â°14'22" W, Salt

LOCATION AND SETTING Pint-Size Shelter is located in Castle Valley on the western margin of the Colorado Plateau, ca. 8 km. west of the Coal Cliffs and ca. 32 km. north of the juncture of the San Rafael Swell and the Wasatch

Plateau (Fig. 2). The latter is the principle feature of the transitional zone between the Colorado Plateau and the Great Basin (Spieker 1949).

GEOLOGY

which crosses the San Rafael Swell, receives drainage from Quitchupah and Ivie creeks about three mi. south and four mi. southeast respectively of the rockshelter. The shelter is on a bench ca. 25 m. above and ca. 75 m. southeast of Christiansen Wash (a tributary of Quitchupah Creek).

Castle Valley (elevation ca. 1,800 m.) is a wide belt of relatively flat terrain located between the San Rafael Swell (elevation ca. 2,000 m.) and the Wasatch Plateau (elevation ca. 3,300 m.). The valley owes its existence to the relative softness of the Mancos Shale Formation which underlies it. The formation contains only one notable resistant, member, Ferron Sandstone, which is hard enough to form a low ridge on the valley's east side and contributes to the gradual westward sloping surface (Stokes and Cohenour 1956). The rockshelter is formed in Ferron Sandstone in the eastern portion of the valley (Figs. 3 and 4). The streams draining the region originate on the Wasatch Plateau and run southeastward en route to the Green and Colorado rivers. The Muddy River,

CLIMATE The climate of the valley is characterized as a midlatitude, dry climate with precipitation distributed equally throughout the year. Some precipitation is derived from intense convectional storms during the summer (Burnham 1950). These storms contribute significantly to advancing valley degradation. The weather station at Emery has recorded about 7 in. per year for the forty-year average (U.S. Weather 31

Aerial photograph of Castle Valley location Bureau Climatic Census 1957, 1965, 1961-70). Precipitation increases markedly on the Wasatch Plateau reaching 30 in. per year near the summit (Burnham 1950). The higher precipitation recorded at Salina, Utah, on the west side of the plateau, suggests that Castle Valley lies in a "rain-shadow" with precipitation mostly falling to the west. Temperature variance between the warmest and coldest months averages more than 130°F. The average frost-free period varies be-

tween 120 and 140 days. The average temperature during the growing season is 65° F. compared to an annual mean of 46° F. (U.S. Weather Bureau Climatic Census 1957, 1965, 1961-70). VEGETATION The site is located at the base of the pinyon-juniper belt in the Upper Sonoran Zone. Plant identifica-

5%&f*=^Si£ ^r Fig. 3

Pint-Size Shelter in Castle Valley (Wasatch Plateau in background). 32

•M-

tions made at the site indicate plant associations are within the Northern Desert Shrub Formation (Tidestrom 1925) and include Great Basin Sagebrush, Saltbush-Greasewood, and Pinyon-Juniper communities (Kuchler 1964). The latter community is sparsely represented on the site and moderately increases up the slope to the east. Plants identified on or near the site include pinyon, juniper, big sagebrush, low sagebrush, four-wing saltbush, big rabbitbrush, greasewood, and snakeweed. Serviceberry and squawbush are present in and near the rockshelter. Grasses, including sand dropseed and Indian rice grass, are present on numerous sand dunes in and around the site. Xerophytic species such as prickly-pear cactus and yucca were also observed. FAUNA The distribution of fauna varies from an abundance of game species and carnivores on the Wasatch Plateau and along the major water courses to only limited numbers in Castle Valley and on the San Rafael Swell to the east. The principle game species are mule deer, distributed throughout the region, and antelope on the San Rafael Swell. Elk are abundant on the plateau, and moose were reported during early historic times (Durrant 1952). Bison, currently maintained in the vicinity of the Henry Mountains about 50 mi. south, are known to have been in the area during prehistoric times (Lucius n.d.). Carnivores including mountain lion, bobcat, and lynx, are present on the Wasatch Plateau, while bears were reported during earlier times. Coyotes are present throughout the region. Other species include beaver,

Fig. 4 Cultural debris eroding out of the sand dune below the rockshelter. badger, porcupines, weasels, jack and cottontail rabbits, skunk, various species of bats, rats, mice, squirrels, and chipmunks (Durrant 1952).

PREVIOUS RESEARCH Creek (Shroedl and Hogan 1975), and three small sites along Ivie Creek (Wilson and Smith 1976). In addition, a number of the sites originally reported by Gunnerson (1957) have been revisited. Others have been identified from further surveys (Helm 1973; C. Berry 1974). Archaic components have been identified at Clyde's Cavern in the San Rafael Swell (Winter and Wylie 1974), Joe's Valley Alcove west of Huntington (DeBloois, 1975 personal communication), and Sudden Shelter in Ivie Creek Canyon (Jennings, Schroedl, and Homer n.d.). More recently salvage requirements have provided myriad data. An intensive survey of a transect through Castle Valley resulted in the identification of over 200 additional sites in Trough Hollow at the western edge of the Old Woman Plateau (Berge 1973). Several lesser surveys have identified sites along Quitchupah Creek and Christiansen Wash (Lindsay 1974; Berry 1975b).

Archeological research of the Castle Valley area (1929-30) was initiated by Noel Morss' (1931) and Henry Roberts' (in Morss) survey along the Fremont River and Muddy Creek. They named and identified the Fremont as a distinct cultural entity. Intensive survey and testing of archeological sites along Ivie Creek and elsewhere in Castle Valley was undertaken in 1954-55 by the Utah Archeological Survey (Gunnerson 1956, 1957, 1960, 1962). The 220 sites reported by Gunnerson (1957) provided a more comprehensive definition of Fremont and cultural variation on the northwestern Colorado Plateau. The survey and testing led to a more thorough program of excavation in which three large Fremont village sites were excavated: Old Woman and Poplar Knob sites (Taylor 1957) and the Snake Rock Site (Aikens 1967). A number of additional Fremont sites in Castle Valley have recently been excavated including three east of Huntington (Madsen 1975), one on Ferron 33

% o a

B u 3 O -t—1

c o o •a a cd

'Z 34

SITE DESCRIPTION trench profile for vertical control. Approximately 9 cu. m.. of fill, excavated with these controls, was passed through a '/4-in. mesh screen resulting in the collection of 1,444 provenienced artifacts.

SETTING The shelter is a heavily weathered L-shaped Ferron Sandstone outcrop (ca. 15 m. E-W by 50 m. N-S) facing south and west (Fig. 3). The shelter rises ca. 4 m. above semistable sand dunes which flow predominantly from west to east and contain the cultural deposits (Figs. 4 and 5). Although some chipping detritus was recovered from the surface below the west face, subsequent testing in that area revealed sterile sand deposits to a depth of ca. 2 meters. Testing below the south face where surface artifacts (chipped and ground stone and ceramics) were most abundant, revealed cultural deposits to ca. 1 meter. The surface below the south face slopes approximately 15Â° E-W and consists of unconsolidated sand and silt material primarily of eolian origin. The relatively shallow cover sand is exposed to wind and slope wash due to scant (ca. 50 percent) vegetational cover. The nearby NE to SW stream channel, cut to bedrock, is indicative of recent erosion produced by overgrazing. This has resulted in new alluvium deposits on the floodplain in the valley below (cf. Appendix I). Recent erosion has also contributed to deflation; consequently, artifacts were exposed and redeposited in the southwestern portion of the site where the deposits are shallowest. In all likelihood, cultural deposits along the west face of the outcropping have eroded out and were redeposited in the wash below. Only the bulkier artifacts were present overlying the sterile sand that remains. Below the site, Christiansen Wash has exposed about 2 m. of alluvium despite the abundant (ca. 75 percent) plant cover. Conditions of the wash during aboriginal times are unknown.

STRATIGRAPHIC SEQUENCE Removal of the trench deposits to bedrock, ca. 75 cm. below the surface, revealed four distinct strata (Figs. 7, 8, and 9). Stratum 1 is a discontinuous orange brown, weathered, colluvial sand deposit containing a high proportion of calcium carbonate and coarse gravel. The stratum directly overlies bedrock. No charcoal/ ash is present and although a few waste flakes and a mano are provenienced for the stratum (Table III), it is interpreted as a sterile, precultural deposit with the few artifacts derived from the overlying cultural fill. The stratum slopes from 0 to 15Â° to the south and was observed only in the southernmost 2 m. of the exploratory trench. Maximum thickness is about 20 centimeters. Absence of the stratum from the northern two-thirds of the trench is probably a precultural phenomenon resulting from an earlier position

EXCAVATION AND STRATIGRAPHY EXCAVATION PROCEDURES A complete recovery of surface artifacts was undertaken to add to those of the initial survey. Excavation of the site was initiated following the removal of vegetation from a ca. lO-sq.-m.-area below the southern part of the rockshelter. A north-south exploratory trench, 1 by 6 m., was directed toward the center of the southern portion in order to determine the presence and depth of cultural deposits (Figs. 5 and 6). A grid system was laid out in 2 m. intervals with a datum point established 2 m. beyond the southeast corner of the trench. The excavation area was expanded east and west of the exploratory trench by using the grid system for horizontal provenience control and the natural stratigraphy mapped from the

Fig. 6 Site location and the limits of excavation. 35

Rodent Disturbance

meters

Fig. 7 North-south profile of Pint-Size Shelter deposits.

of-the east-west channel presently located on bedrock about 5 m. to the south of the posited older course. Stratum 2 is a discontinuous lens containing heavily weathered stone slabs directly overlying bedrock. The stratum essentially replaces Stratum 1 where it is absent. It slopes from 0 to 15Â° north-south, and is visible in the northern two-thirds of the trench. The slabs, many charcoal/ash stained on the bottom side, are in a matrix of yellowish brown eolian sand and calcium carbonate. The matrix contains a modicum of charcoal, gray ash, and chipped stone detritus. The stone slabs were probably derived from bedrock through frost-wedging, although some may have been imported. This was followed by a natural deposition of eolian sand. Cultural debris subsequently filtered down from the overlying Stratum 3 to create the matrix containing the spall slabs. Stratum 3 is a continuous, dark brownish gray eolian sand deposit containing lenses of gray and whitish gray ash, charcoal, and chipped and ground stone artifacts. It slopes from 0 to 15Â° north-south and was observed overlying Stratum 2 throughout the excavation except in the southern 50 cm. of the exploratory trench where only Stratum 1 is present. The stratum has a maximum thickness of about 60 cm. where it serves as the fill for Pit 1. It is truncated by the erosion channel (presently 3 m. to the south) in the southern 50 cm. of the trench. The stratum contains a 2-m.-by-25-cm. spoil component (Fig. 8) derived from Stratum 2 in the construction of Pit 1. A lens of tabular stone slabs, similar to those of Stratum 2, is present in the northern two-thirds of the trench. The stone lens varies from the surface of the stratum to about 20 cm. below. A slab metate was taken from the stratum 35 cm. below the present surface. Stratum 4 is a continuous eolian deposit of orange brown sand consisting of charcoal, gray ash, abun-

dant chipped and ground stone, and ceramic artifacts. It overlies Stratum 3, and varies in thickness from 1 m., where it serves as the fill for Pit 2 at the base of the shelter, to 20 cm. in the vicinity of the hearth. The stratum terminates at the present and southernmost margin of the channel. (See Discussion in Excavation and Stratigraphy for possible subcomponents.) The stratum contains a 50-by-25-cm. spoil component (Fig. 8) derived from Strata 2, 3, and the upper half of 4 in the construction of Pit 2. Charcoal samples were obtained for radiocarbon dating from the bottom and top of Stratum 3 and from a fire hearth in Stratum 4. Soil samples were obtained from the four strata for soil and palynological analysis and one from Stratum 3 for analysis of plant macrofossils. In addition, stone samples were obtained for petrological analysis from the rockshelter, bedrock, and Strata 2 and 3.

CULTURAL FEATURES Only one cultural feature, a fire hearth, was uncovered in the exploratory trench. The level of origin is from within the basal portion of Stratum 4. The expansion of the northern third of the exploratory trench 1 m. to the west, 1 m. to the north, and 2 m. to the east exposed two additional features: Pits 1 and 2. Levels of origin are from the top of Stratum 2 and within the middle of Stratum 4 respectively. Pit 1, the earliest cultural feature, was observed in both the north and west profiles (Figs. 8, 9, and 10). Approximately half of the pit was excavated; however, the exact shape and dimensions of the pit cannot be estimated due to rodent disturbance. The level of origin is the top of Stratum 2. Depth of the pit is about 25 cm. below the top of the spoil component which overlies Stratum 2 at the edge of the pit. 36

immediate site vicinity. This occurred prior to the incision of present channels during a period which was more denuded than now (cf. Appendix I). Cover sand, silt, and entrenchment followed in which colluvial Strata 1 and 2 (in part) were deposited and subsequently removed, leaving isolated remnant patches. The weathering of spall from bedrock prior to site occupation constitutes a major property of Stratum 2. This occurred during the final phase of the precultural period and/or the onset of occupation. The relationship of Stratum 1 to Stratum 2 is not fully determinable because of rodent disturbance where superposition may have occurred (Fig. 7). The suggested sequence is based on the geological assessment that Stratum 1 consists of slightly coarser gravels and contained greater calcium carbonate properties. Also, the stratum was sterile except for a few unworked flakes which were probably obtained on or near the top, and hence were ultimately derived from Stratum 3. The spall, a principle component of Stratum 2, was no doubt derived prior to site occupation. The soil matrix of the stratum is a mixture of natural (precultural) and early (cultural) deposition. The latter infiltrated the natural deposit, percolating downward to create the soil matrix containing the stone slabs. If the cultural material within the stratum was not derived from Stratum 3, then Stratum 2 represents an earlier occupation. With the onset of Stratum 3 occupation, cultural and natural deposits were laid down simultaneously. Eolian deposition continued over the course of the two major site occupations, resulting in a more continuous sand and vegetational cover of the entire landscape. This is also suggested by the palynological evidence (cf. Appendix II). The homogeneous nature of Stratum 3, the earliest abundant cultural deposit, suggests the site was occupied without significant interruption for the duration of the stratum's depositional history. Pit 1 was constructed coincident with the depositional beginning. An occupational hiatus of considerable duration between Stratum 3 and Stratum 4 is posited. Although

Pit fill consists wholly of Stratum 3 with numerous tabular sandstone slabs that appeared to have been deliberately stacked. The fill contained greater amounts of charcoal and ash than observed elsewhere in the stratum. Ground and chipped stone artifacts are also present in the fill. Pit 2 was observed in the north and east profiles (Figs. 7, 8, 11, and 12) after the pit was partially excavated. Approximately three-fourths of the roughly circular (ca. 1.5 m. dia.) pit was excavated. The pit truncates the lower half of Stratum 4 about 20 cm. below the surface, all of Strata 3 and 2 to bedrock. Depth of the pit is about 80 cm. below its level of origin in the center of Stratum 4. A 20-cm. spoil component from the original construction was observed in the upper half of the stratum. Pit fill consists wholly of Stratum 4 and is marked by an abundance of stone slabs, ground and chipped stone artifacts, pottery, and other refuse. A trough metate was taken from the fdl ca. 2 m. below the present surface. The hearth, uncovered in the center of the exploratory trench, is located 3.75 m. south of the shelter and originates within the lower half of Stratum 4, about 10 cm. below the present surface (Figs. 7 and 13). The roughly circular (about 50 cm. dia.) hearth consists of a shallow depression lined with unworked, flat, tabular stone slabs. A charcoal sample was obtained for radiocarbon dating.

SUMMARY AND DISCUSSION Several problems were encountered in the course of excavation such as rodent disturbance, definition of stratigraphic subcomponents, and the levels of origin of.cultural features. These are addressed in the ensuing discussion of the natural and cultural depositional history of the site. An early precultural, pre-cover-sand deposition consisting of alluvial gravels has been identified in the

meters Fig. 8 East-west profile of Pint-Size Shelter deposits. 37

Site stratigraphy (north profile) and Pit 1 (lower left)

there is no evidence of an intervening sterile deposition between the two cultural components, the color and texture of the two strata are highly dissimilar. A major spall lens is variously present on and just below the surface of Stratum 3. This may indicate a lengthy period of weathering and frost-wedging of the sandstone outcrop above the site. However, the stone is wholly dissimilar to that encountered in Stratum 2 (assessed as a natural phenomenon), and the spall was probably a cultural import from flagstone exposed southeast of the site. Firm evidence of a stratigraphic break is derived from the variations in thickness of the petrocalcic cortex deposited on the face of the rockshelter. The cortex, formed by the leaching of calcium carbonates and subsequent percolation down through the strata adjacent to the rockshelter, varies in thickness from about 1 : 8 (Stratum 4 : Stratum 3) on the face of the shelter. This indicates that the cortex associated with the deposition of Stratum 3 is nearly eight times as thick as that associated with the overlying Stratum 4 (Appendix I). The radiocarbon dates obtained from the top of Stratum 3 and the hearth in the lower half of Stratum 4 corroborate the geological assessment (see Table II). Despite the lengthy occupational break, Stratum 4 appears to have been deposited in much the same manner as Stratum 3 â€” a continuation of the simultaneous deposition of eolian cover sand and cultural materials. Firm cultural evidence of any separation of Stratum 4 into lesser components is lacking despite attempts to provenience separately artifacts and pollen samples obtained from upper and lower levels (see Appendix II). The fire hearth and Pit 2 levels of origin

suggest the two features may have been constructed at separate times. The hearth appears to originate within the lower half (Fig. 7) and Pit 2 from the center of the deposit. However, regardless of the presence or absence of a relative sequence within the stratum, a minute amount of the component was deposited prior to the construction of the hearth. A thin postoccupational veneer consisting of loose, unconsolidated material is present on the surface of the site. This deposit was probably more abundant prior to recent denudation and consequent erosion. Much of the postoccupational debris and earlier cultural deposits, particularly in the southwestern portion of the site, probably have been displaced onto the floodplain below.

TABLE I Distribution of Provenienced Diagnostic Artifacts

ou O

CoU

2 NN

tate

ints

nts

Fig. 9

4Âť

o s

o e

ÂŤ

<u D. >* H

<u a. s*> H

*>> H

>> H

>-. H

>> H

Cultural Unit II

105

Cultural Unit I

ced

CULTURAL UNITS Unit II (Fremont) is the youngest cultural component. The unit consists of Stratum 4, Pit 2 and the hearth, both of which originate from within the stratum. Artifacts include a Type II, large, triangular, corner-notched point with expanding stem; two Type III, small, triangular points with rounded bases (diagnostic); a Type IV, small, triangular point with concave base; a whole trough metate, and three shaped (two-hand) manos. In addition, one Sevier Gray and 104 Emery Gray ceramic sherds were recovered from the unit. Other artifacts include a Type I point and an assortment of bifaces, unifaces, and waste flakes.

Two stratigraphic cultural units (I and II) are defined at Pint-Size Shelter (Fig. 14 and Table I). Unit I (Archaic) is the oldest cultural component at the site. The unit consists of Stratum 3 and Pit 1, the latter originating from the top of Stratum 2. Diagnostic artifacts include a Type I (Gypsum) projectile point, a slab metate, and five unshaped (onehand) manos. Two Gypsum points were found elsewhere; one from the fill of Pit 2 and one of unknown provenience. Nondiagnostic artifacts include two Type V triangular projectile point fragments (without bases), one drill, and various bifaces, unifaces, cores, and waste flakes. No ceramics are identified in the unit.

DATING the deposition of a Gypsum (Type I) projectile point recovered from the same locus. Sample RL-534 (1790*100 B.P.) was derived from charcoal obtained from the fire hearth which originates from within the lower half of Stratum 4. The sample dates Emery Gray pottery and the onset of Fremont (Unit II) occupation. (See Site Interpretation and Discussion.)

Three charcoal samples were collected â€” two from concentrations within Stratum 3 and one from the fire hearth in Stratum 4. Sample RL-535 (4520Âą210 B.P.) was derived from a small charcoal concentration at the base of Stratum 3 and dates the initial Archaic (Unit I) occupation of the site. Sample RL-536 (3390*170 B.P.) was derived from a small charcoal sample obtained from the upper portion of Stratum 3 and dates the terminus of Archaic (Unit I) occupation. The sample dates

Fig.

10 Pit 1 outline (west profile) and rodent disturbance (left side). 39

ARTIFACTS Artifact yield from Pint-Size Shelter is limited to nonperishable materials. Nearly two-thirds (1,444 of 2,380) are provenienced, although erosion on the southern and western portions of the site contributed extensively to an unprovenienced surface collection. The unprovenienced category also includes many artifacts from the exploratory trench. Provenienced diagnostic artifacts are clearly patterned temporally and spatially (Table II). As a result of the high degree of stratigraphic and feature controls, this pattern is an accurate reflection of the depositional sequence. Artifacts are classed technologically as chipped stone, ground stone, and ceramics. Chipped stone

comprises 93 percent of the artifact assemblage; ceramics 6 percent, and ground stone the remainder. Of the 2,220 pieces of chipped stone, 2,052 (ca. 92 percent) are classed as debitage. The remaining 168 pieces of chipped stone are classed morphologically as points, drills, bifaces, unifaces, cores, hammerstones, and utilized (?) flakes. These basic categories are subdivided according to configuration and flaking technique. Ground stone is descriptively classed according to the presence and extent of shaping. Ceramics are classified on the basis of tempering, inclusions, and design features.

CHIPPED STONE

of stone used. For example, if the bulb of force can be identified, artifacts can be described in terms of the striking platform, ventral and dorsal faces, and distal and proximal ends. Retouch, medial thinning, and collateral flaking are considered to be pressure flaking techniques. Thermal alteration, without microscopic analysis, can be identified only tentatively. Criteria such as variable luster in flake scars and gradients of color from the exterior through the interior (with both exterior faces identical in color), provide clues to possible heat treatment (Crabtree and Butler 1964). The 2,220 items of chipped stone are tabulated and discussed in the following order: Points, Drills, Bifaces, Unifaces, Cores, Hammerstones, Utilized (?) Flakes, and Debitage.

Artifact lithology is described in terms of Crabtree (1967) and Ford (1953). D. R. Currey and J. H. Madsen examined sample specimens and contributed to delineating broad categories for the rock types recovered from Pint-Size Shelter. The "flint-chert confusion" has been settled for purposes of this analysis by using the criteria of opacity and color. Light colored, transparent to semitranslucent cryptocrystalline specimens are called cherts; darker, opaque cryptocrystallines are called flints! Readily observable features are described in terms of opacity, texture, luster, type of fracture, presence or absence of structural inclusions, mineral banding, and possible thermal alteration. Artifacts recovered from the site separate into seven broad raw material classes: cherts, flints, siliceous limestones, quartzites, obsidian, jasper, basalt, and sandstone. Distinctive features of individual artifacts are described under the rubric "materials." Typing of raw materials must remain tentative until sources of raw materials are located and rock samples are subjected to appropriate laboratory analysis; Quarry sites for Pint-Size Shelter are probably variable. Sources may range from lag gravel deposits on Quitchupah Creek to nearby secondary deposits of siliceous materials, or perhaps to traded or imported preforms. The accurate identification of lithic materials is in part dependent upon identification of quarry sites. That level of accuracy is beyond the scope of this investigation. Although artifacts are ultimately typed morphologically and functionally, they are described â€” whenever possible â€” in terms of techniques used in forming them (cf. Crabtree 1967). It is assumed that technology is dependent upon the type and workability

POINTS The 22 points and point fragments from Pint-Size Shelter have been divided into four types. Points are defined generally as medially thinned lithic material having a hafting element and/or two sides forming an acute angle. Type I and II points are large and triangular in outline. Type III and IV points are smaller and triangular in plan. This sequence implies chronology (i.e., atlatl to arrowpoint) and generally corresponds to the depositional sequence. Type I No. of Specimens: 3 Description: The specimens are large, triangular points with straight to incurvate edges, asymmetrical in plan view, random-to-parallel flaking technique, corner-notched, and straight-stemmed. Two specimens are on curved flakes; one is lenticular in section. 40

Fig. 11 Pit 2 (north profile, right) â€” Stratigraphic context.

Figure 15 a: Material: translucent chert. Dimensions: length, 3.7 cm.; width, 2.0 cm.; thickness, 0.6 cm.; weight, 3.3 g. Provenience: unknown. Figure 15 b: Material: white, opaque chert. Dimensions: length, 3.8 cm.; width, 1.9 cm.; thickness, 0.5 cm.; weight, 3.3 g. Provenience: Stratum 4. Figure 15 c: Material: white, opaque chert. Dimensions: length, 4.0 cm.; width, 1.9 cm.; thickness, 0.5 cm.; weight, 3.3 g. Provenience: Stratum 3. Comparable Types: Jennings, Schroedl and Homer n.d., "Gypsum"; similar to Jennings 1957, W14; Schroedl and Hogan 1975, Type ID Figure 10, d-e; Gunnerson 1957, Figure 14, bb; similar to Aikens 1967, Figure 24, 1. Type II Fig. 12 Pit 2 (closeup).

No. of Specimens: 4 Description: The specimens are large, triangular points with straight edges, deep corner notches with expanding stems, and basal thinning. The specimens are lenticular in section with random pressure flaking and little edge retouch. Two specimens (Fig. 15,e and f) are fragmentary â€” one from unknown provenience, the other from Stratum 4. Figure 15 d: Materials: flint; gray vitreous, white inclusions (heat-treated). Dimensions: length, 3.4 cm.; width; 2.8 cm.; thickness, 0.5 cm.; weight, 3.0 g. Provenience: unknown. Figure 15 g: Materials: gray, waxy flint. Dimensions: length, 3.8 cm.; width, 1.9 cm.; thickness, 0.4 cm.; weight, 2.6 g. Provenience: unknown. Comparable Types: Jennings 1957,W18; Schroedl and Hogan 1975, Fig. 10,b and c ; Jennings, Schroedl and Homer n.d.; Aikens 1970, Elko Side-notched, similar to "s."

Fig. 13 Fire hearth (top view). 41

/^S»oiT^\

Stratum 4

Hearth

— ^ Stratum 3

Cultural

/?-

Unit

Stratum 3 Pit

Spoil

Cultural Unit 1

Stratum 2

Pit

Stratum 2

^^_

1 edrock

Fig. 14 Composite profile of cultural features and stratigraphy. Type III This category consists of small triangular points with corner notches.

Comparable Types: Similar to Gunnerson 1957, Figure 14; similar to Aikens 1970, Figure 18 e-i, Rose Spring Corner-notched.

Type III A

Type IV

No. of Specimens: 3 Description: The specimens are small, triangular points with straight to incurvate edges, lenticular to plano-convex in cross section; one specimen has short tangs extending toward base. Bases are slightly rounded and specimens exhibit little edge retouch. Figure 15 h: Materials: light gray, semitranslucent chert. Dimensions: length, 2.7 cm.; width, 1.3 cm.; thickness, 0.3 cm.; weight, 1.3 g. Provenience: Stratum 4. Figure 15 i: Materials: pink, translucent chert. Dimensions: length, 2.5 cm.; width, 1.5 cm.; thickness, 0.5 cm.; weight, 1.5 g. Provenience: unknown. Comparable Types: Similar to Gunnerson 1957, Figure 14 x; similar to Aikens 1970, Figure 19, Rose Spring Corner-notched; similar to Fowler et al. 1973, Figure 11, v-aa.

No. of Specimens: 1 Description: The specimen is a fragment; a small, deeply indented base, plano-convex in section. Figure 15 m: Materials: pink, translucent chert. Provenience: unknown. Comparable Types: Aikens 1967, Figure 24 g, unstemmed points. TypeV No. of Specimens: 8 Description: The specimens are points truncated at tip or base or both. Four specimens are triangular in outline with straight to incurvate edges; bases missing. Two are midsections with straight edges; one is a small tip fragment. Materials: cherts, flint, and jasper. Provenience: Stratum 3 (1); Stratum 4 (1); Unknown (6).

Type III B No. of Specimens: 3 Description: The specimens are small, elongated, triangular points, corner-notched with tangs extending toward base. Specimens are straight-stemmed with basal thinning. Specimen in Figure 15 1 is complete; the other two specimens are basal fragments. Two specimens (Fig. 15 k and 1) exhibit edge retouch. Figure 15 k: Materials: pink, translucent chert. Dimensions: fragmentary. Provenience: unknown. Figure 15 1: Materials: jasper; green, dull luster, brown inclusions. Dimensions: length, 4.0 cm.; width, 16 cm.; thickness, 0.4 cm.; weight, 2.0 g. Provenience: unknown. A fragmentary point of unknown provenience is not photographed. Materials: light gray, fragmentary chert. Dimensions: fragmentary. Provenience: unknown.

Discussion of Points Thirteen of the 22 points recovered from Pint-Size Shelter fall within the range of variation of dated types, but are limited as time markers. Only four diagnostic projectile points recovered from Pint-Size Shelter are provenienced: two Type I points (Gypsum) and two Type III A points (small, triangular corner-notched points). Type I (Gypsum) These points have a time range of nearly 4,000 years (ca. 4630 B.P. to ca. 950 B.P., Fowler, Madsen and Hattori 1973). Gypsum points are found in Fremont and Archaic contexts, but are far more 42

Fig. 15 Points Type I a-c; Type II d-g; Type III A h-j; Type III B k-1; Type IV m. (Scale in cm.) 43

numerous in Archaic deposits. There is the possibility that they were manufactured in the Archaic period and re-used in the Fremont period.

Type III A These small, triangular corner-notched points are presumed to be arrow points (Madsen and Berry 1975) and as such, are diagnostic of Fremont or Shoshonean cultures. Type III A points are not uniquely Fremont; however, they were recovered in stratigraphic association with Emery Gray ceramics.

Type II (Elko Corner-notched) and Type III B (Rose Spring Corner-notched) Elko Corner-notched points have little utility as time markers since they also are found in Archaic and Fremont contexts (Madsen and Berry 1975). All but one of the Type II points from Pint-Size Shelter are unprovenienced and could have eroded out of either cultural unit. The Rose Spring Corner-notched points are also unprovenienced. The assigned timerange corresponds closely with the Fremont Culture (ca. 1350 B.P. to 600 B.P.) (Hester 1970), although Aikens (1970) claims they appear at Hogup Cave as early as ca. 4450 B.P. Elko and Rose Spring Corner-notched points are morphologically similar and are generally distinguished on the basis of size: the larger points are typed as Elko Corner-notched and the smaller points as Rose Spring Corner-notched, with overlapping ranges of variation. Without the criteria of probable time-range (from stratigraphic association), probable function (e.g., atlatl or arrowhead), and a large provenienced collection, the area of overlap for the two types can be divided only arbitrarily. With these difficulties, neither type can be considered diagnostic.

DRILLS Drills are defined as bifacially worked material with bevelled points. Drill points (circular in section) may or may not have proximal haft elements. Six drills were recovered from Pint-Size Shelter. Two are possibly resharpened "points," but their technical lineage is indeterminate. Three points are truncated basally, and one is fractured longitudinally. Type I No. of Specimens: 2 Description: Complete drills. Figure 16 a: Description: The specimen is worked from a curved flake with a high dorsal medial ridge. One is shouldered with incurvate edges and has a point 0.2 cm. in diameter, extending 0.5 cm. to flaring blade. Material: obsidian. Dimensions: length, 2.9 cm.; width, 1.1 cm.; thickness, 0.5 cm.; weight, 1.3 g. Provenience: unknown. Figure 16 b: Description: The specimen is a small, corner-notched drill with a slightly expanding stem. It is lenticular in section and straight-based with a blunt point extending 0.6 cm. from distal end and curving sharply to maximum length. Blade edges are parallel. Material: dark red, translucent chert. Dimension: length, 1.9 cm.; width, 1.2 cm.; thickness, 0.3 cm.; weight, 0.5 g. Provenience: unknown. Type II No. of Specimens: 4 Description: Drill fragments range from lenticular to biconvex. Two specimens are truncated basally; one is amputated proximally and distally. One specimen (Fig. 16 c) shows edge retouch. The specimen in Figure 16 e is fractured longitudinally and is sharpened (2.2 cm.) dorsally along the blade. The specimen may be a sharpened biface fragment. The fourth drill fragment is lost. Materials: Figure 16 c: chert; Figure 16 d,e: silicified limestone. Provenience: Figure 16 e: Stratum 3; Figure 16 d: Stratum 4; and Figure 16 c: unknown. "Drills" is a functional category and has no utility as a cultural diagnostic. Drills have been recovered in stratigraphic association with Fremont and Archaic period cultures.

Fig. 16 Drills Type I a,b; Type II c-e. (Scale in cm.) 44

Fig. 17

Bifaces Type I a-e; Type II f-h. (Scale in cm.)

Type II

BIFACES

No. of Specimens: 3 Description: The specimens are lenticular in section with a random to parallel flaking showing some edge retouch. The specimens have straight to slightly incurvate edges and straight bases slightly rounded at corners. All specimens are fragmentary. Materials: Figure 17 f: semitranslucent chert; Figure 17 g: white, opaque chert; Figure 17 h: obsidian. Provenience: Stratum 3 (Figure 17 h) (1); Stratum 4 (1); Unknown (1). This category consists of leaf-shaped bifaces â€” lenticular to biconvex in section, and presumed leafshaped biface fragments (i.e., tips, midsections, and bases).

Bifaces are defined as lithic material that is worked on the ventral and dorsal faces. Bifaces may be limited to edge retouch along opposing faces of one edge or may have no edge retouch, but may merely exhibit ventral and dorsal primary flaking. The 74 bifaces and fragments recovered from Pint-Size Shelter have been separated into six types according to configuration and flaking techniques. Type I No. of Specimens: 11 Description: Specimens are lenticular to biconvex in section with parallel to incurvate edges. They exhibit random to parallel flaking and are medially thinned. Some specimens are retouched along edges (e.g., Fig. 17 a-e). Materials: Flint, semitranslucent chert, red jasper. Dimensions: Specimens range from ca. 6.0 cm. to 2.5 cm. in length; ca. 0.6 cm. to 0.3 cm. thick; and 1.2 to 2.4 cm. in width. Provenience: Stratum 3 (1); Stratum 4 (1); Unknown (9). Comparable Types: Jennings 1957, W42.

Type III A No. of Specimens: 2 Description: Specimens are leaf-shaped, lenticular to biconvex in cross section with random flaking. One specimen is retouched at the tip. Materials: Figure 18 a and b: gray green, semitranslucent chert; flake scars dull to vitreous; heat-treated. Dimensions: Lengths, 2.2 to 7.8 cm.; widths, 4.1 45

Fig. 18 Bifaces Type III A a,b; Type III B c; Type III C d,e; Type III D f,g; Type IV h,i. (Scale in cm.) 46

TypeV

to 4.7 cm.; thicknesses, 1.2 to 1.4 cm.; weights, 4.0 to 51.0 g. Provenience: Stratum 4 (1); Unknown (1). Comparable Types: Jennings 1957, W48; similar to Aikens 1967, Figure 27.

No. of Specimens: 4 Description: These bifaces are thick (at least 3.0 cm.) in section and are randomly flaked. One is oval in outline, one circular, one truncated with a curved edge, and one elongated. One specimen exhibits bifacial edge retouch. Weight range is 74.0 g. to 219 g. (e.g., Fig. 19 a and b). Materials: Siliceous limestone (3); light gray, semitranslucent chert (1). Provenience: Stratum 4 (4). Comparable Types: Type V may be functionally classed as choppers.

Type III B No. of Specimens: 1 Description: Biface midsection, biconvex in cross section. Materials: Figure 18 c: white opaque chert with gray interior; possibly heat-treated. Provenience: Unknown.

Type VI No. of Specimens: 13 Description: Type VI is made up of bifacially worked flakes. Three specimens have broad flake scars across two faces. The remaining seven are bifacially worked along one or two edges. Two fragments are possibly heat-treated (e.g., Fig. 19 c and d). Materials: Jasper; flint; chert; silicified limestone. Provenience: Stratum 2 (1); Stratum 4 (7); Unknown (5).

Type III C No. of Specimens: 8 Description: Specimens are large biface fragments truncated at one end and an acute angle at opposite end. Specimens are biconvex in section. Points are rounded to acute. Flaking techniques are generally random with some collateral flaking (e.g., Figure 18 d and e). Materials: Flint; semitranslucent chert; red jasper; white opaque chert. Provenience: Stratum 3 (1); Stratum 4 (1); Unknown (6).

Type VII No. of Specimens: 17 Description: Type VII is composed of various small biface fragments whose original dimensions cannot be surmised (e.g., Fig. 19 e and f)Materials: Flint, chert (various types); jasper. Provenience: Stratum 3 (2); Stratum 4 (10); Unknown (5).

Type III D No. of Specimens: 12 Description: Specimens are large, truncated bifaces with opposing ends square to rounded. Two fragments are medially thinned and nearly lenticular in section with some collateral flaking. The remaining nine fragments are biconvex with generally random flaking. No fragments exhibit edge retouch (e.g., Figure 18 f and g). Materials: Pink translucent chert with white inclusions; silicified limestone; three specimens are possibly heat-treated. Provenience: Stratum 3 (2); Stratum 4 (5); Unknown (5).

UNIFACES The 34 unifaces recovered from Pint-Size Shelter are separated into three types. The class of unifaces is distinguished as flakes with narrow. unidirectional flake scars along one or more edges. The worked flake may also exhibit broad flake scars across the same face. Type I

Type IV

No. of Specimens: 6 Description: Large flakes, plano-convex in section, with flake scars along one edge (e.g., Fig. 20 a and b). Materials: Limestone; silicified limestone; chert; flint. Provenience: Stratum 2 (1); Stratum 4 (1); Unknown (4). Comparable Types: Type I is usually referred to as side scrapers. Jennings 1957, W64, 65; Schroedl and Hogan 1975, Fig. 10 u; Berry 1975a, Fig. 10 f.

No. of Specimens: 4 Description: Type IV bifaces are relatively small, primary flaked, and thick in section. One specimen is triangular and one is oval in outline (e.g., Fig. 18 h and i). Materials: Flint; chert; jasper; three specimens are heat-treated. Provenience: Stratum 3 (1); Unknown (3). Comparable Types: Type IV specimens are commonly classed as preforms. 47

Fig. 19

Bifaces Type V a,b; Type VI c,d; Type VII e,f. (Scale in cm.)

Type II

Type IIB

Type II A consists of large flakes roughly triangular in outline and plano-convex in section. Specimens are worked along the edge opposite the apex (not necessarily the distal end). Type II B specimens generally correspond to the same criteria, yet are only 3.0 cm. to 5.5 cm. in length.

No. of Specimens: 3 Description: Specimens are small, vex flakes with small, unidirectional one end (e.g., Fig. 22 a). Materials: Flint (1); semitranslucent Provenience: Stratum 4 (2); Unknown

steeply biconflake scars on chert (2). (1).

Type III

Type IIA

Type III consists of small, unifacially worked flakes of various shapes. Type HI C is an elongated flake worked on three edges, and Type III B has graverlike promontories.

No. of Specimens: 8 Description: Specimens have high dorsal ridges and a series of deep parallel flake scars along distal end. Two specimens are worked on the proximal end and the bulb of force exhibits thinning scars, (e.g., Fig. 20 c-f). Materials: Silicified limestone (5); chert (1); jasper (2). Provenience: Stratum 2 (1); Stratum 3 (3); Unknown (4). Comparable Types: Functionally classed as end scrapers.

Type III A No. of Specimens: 13 Description: Specimens are amorphous in shape (weight range is 0.7 to 16.6 g.). All show undirectional edge retouch. Four specimens have sinuous edges. Four artifacts are worked along their dorsal faces. Two flakes are merely worked along one edge. Three flakes show broad, lateral flake scars on their 48

face (except on the striking platform) (e.g., Fig. 21 d). Materials: Light gray, semitranslucent chert. Dimensions: Length, 7.2 cm.; width, 2.5 cm.; thickness, 0.7 cm.; weight 15.8 g. Provenience: Stratum 4

ventral sides (i.e., ventral thinning to make an edge) (e.g., Fig. 21 a and b). Materials: Gray, semitranslucent chert; white, opaque chert; jasper; flint; silicified limestone. Provenience: Stratum 4 (8); Unknown (5).

UTILIZED (?) FLAKES

Type HI B

This class is weakly defined. Specimens have acute ragged edges. The edge flake scars are shallow and irregular. Whether the edge is a degenerate uniface, is merely used, or is fortuitous, is conjectural. The class is a convenient one for flakes with characteristics that resist clear definition (e.g., Fig. 22 b). No. of Specimens: 15 Provenience: Stratum 2 (3); Stratum 3 (1); Stratum 4 (5); Unknown (6).

No. of Specimens: 3 Description: Flakes have fortuitous promontories that have been worked unifacially into graverlike tips. One flake is thinned with a broad flake scar and worked unidirectionally along the dorsal face (e.g., Fig. 21 c). Materials: White opaque chert; gray semitranslucent chert; gray brown chert. Provenience: Stratum 3 (1); Unknown (2). Comparable Types: Commonly referred to as burins or gravers; Jennings 1957, W87.

CORES Type HI C

Cores are defined as multifaceted lithic material without acute worked edges. Cores are presumed to be sources of flakes to be knapped into artifacts. Twelve cores were recovered from Pint-size Shelter.

No. of Specimens: 1 Descriptions: Specimen is a long flake roughly triangular in cross section, flaring toward the distal end. The specimen is unifacially worked on the ventral

Fig. 20

Unifaces Type I a,b; Type II A c-f. 49

Comparable Types: Schroedl and Hogan 1975, (Fig. l l e , h ) . Type II No. of Specimens: 1 Description: Triangular in outline, multifaceted; two broad flake scars from acute angle. Possible used edge (chopper tool?) (e.g., Fig. 23 g). Material: Limestone. Provenience: Unknown.

HAMMERSTONES All five specimens are quartzite and have small, rough concavities that indicate repeated battering. Type I hammerstones are spherical-to-oval in shape; one is fragmentary (weights range from 59 to 590 g.). Type II consists of one specimen, oval in shape with a 1.5-cm.-dia. concavity at one end. It is circumscribed by a shallow (about 1.0 cm. deep) groove, midpoint between the ends. Length is 4.9 cm. and weight is 4.9 grams. All hammerstone specimens are from provenience unknown. Hammerstone, as the name suggests, is a functional category. Comparable Types: Schroedl and Hogan 1975, (Fig. 11 g). Fig. 21 Unifaces Type III A,a and b; Type III B,c; Type IIIC,d. (Scale in cm.)

DEBITAGE Approximately 90 percent, or 2,052 items, of the chipped stone is classed as waste flakes. Most flakes exhibit some work on the dorsal face and weigh from 0.59 to 620.0 grams. The flakes are amorphous in plan and variable in section representing all rock types (except obsidian) found among the artifacts. The two obsidian artifacts (a drill and a blade fragment) appear to have been finished. The absence of obsidian waste material suggests obsidian was imported as at least preforms or perhaps as finished pieces.

Type I No. of Specimens: 11 Description: Specimens are multifaced with one or more striking platforms. Eight specimens are roughly spherical in shape; two evidence cortex; three are roughly oval (e.g., Fig. 23 d,e,f)Material: Siliceous limestone; chert; flint. Dimensions: Weight, 88 to 247 g. Provenience: Stratum 2 (1); Stratum 3 (1); Stratum 4 (2); Unknown (7).

Fig. 22

Close-ups (Mag. 3X) of Utilized (?) Flake (left) and Uniface IIB (right). 50

Fig. 23

Hammerstones a-c; Cores: d-g. (Scale in cm.)

The debitage from Pint-Size Shelter generally reflects the stratigraphic distribution of other classed artifacts and is provenienced as follows: Stratum 1 (8); Stratum 2 (69); Stratum 3 (267); Stratum 4 (807); Unknown (901).

Slab Metate (Type I) This subrectangular slab of scaly sandstone (Fig. 24 a) is 36 cm. long, 28 cm. wide, and 6.0 cm. thick. One surface had been smoothed and deepened by pecking or grinding. Provenience: Stratum 3. Comparable Types: Marwitt (1970).

GROUND STONE

Trough Metates (Type II) The one complete specimen (Fig. 24 b) is trapezoidal in outline with rounded corners. The metate has a fairly deep grinding basin with one end open. The metate is sandstone, well-cemented with calcite. The specimen has no secondary depression and is not modified except for the grinding surface. Dimensions: Length, 41.0 cm.; width, 28.0 cm.; thickness, 12.0 cm.; depth of trough, ca. 6.0 cm. Three trough metate fragments were also recovered from Pint-Size Shelter. Provenience: Stratum 4 (metate and 2 fragments); Unknown (1 fragment).

Twenty-nine ground stone artifacts were recovered from Pint-Size Shelter. Ground stone is classed as metates or manos using descriptions and typology in Marwitt (1970). METATES The two complete metates recovered are classified according to the extent and shape of their grinding basins. Of three fragments recovered, two can be confidently identified as trough metates. The other has an extremely shallow basin. 51

Fig. 24

Metates Type I (above); Type II with mano (below). 52

Fig. 25

Manos Type I a,b; Type II c,d,e. (Scale in cm.) 53

MANOS Twenty-four specimens both complete and fragmentary are classed according to shape. Ten fragments could not be typed. Three of the remaining fourteen have been shaped by pecking. The rest are cobble manos ground on one or two surfaces. Type I No. of Specimens: 3 Description: Oval to subrectangular in plan, trapezoidal to subrectangular in section. Specimens have been pecked to form high-backed, loaf-shaped manos. One specimen is ground on three surfaces (e.g., Fig. 25 a and b). Material: Basalt; sandstone. Dimensions: Lengths, 13.0 cm. to 18 cm.; thickness, 7.0 cm. to 10.0 cm. Provenience: Stratum 4.

Type II No. of Specimens: 21 Description: Cobble manos vary from subrectangular to circular in plan and subrectangular to biconvex in section. Five of the manos and mano fragments (subrectangular in plan) are ground on opposing faces. Ten pieces exhibit one ground surface and are considered mano fragments (e.g., Fig. 25 c-e). Material: Vesicular basalt; well-cemented sandstone; friable red sandstone. Dimensions: The six circular manos are ground on one face. Lengths, 6.0 cm. to 9.0 cm.; thickness, 5.0 cm. to 6.0 cm. Provenience: Stratum 1 (1); Stratum 3 (5); Stratum 4 (6); Unknown (9).

CERAMICS One hundred thirty-one pottery sherds were recovered from Pint-Size Shelter. All ceramics are identified as Fremont pottery. One sherd recovered from the surface is defined as Sevier Gray (greater than 0.5 mm. black basaltic inclusions; R. Madsen 1972, 1976). The remaining 130 sherds are Emery Gray. Four of these have unsmoothed exterior coils. Emery Gray sherds are identified by small-grained gray basaltic temper (0.1 to 0.3 mm. inclusions). One rim sherd is from an approximately 50-cm. diameter olla. No bowl forms are evident. One sherd has an exterior drilled mending hole. Textures and surfaces indicate a minimum of four vessels. Three sherds have exterior fugitive red stain. Twenty-five Emery Gray sherds are from provenience unknown; 105 are from Stratum 4 (e.g., Fig. 26). Turner-Look is the type site for Emery Gray. Emery and Sevier Gray are usually distinguished by the area in which they are found, but difficulties are encountered as their areal distributions overlap. In these areas (e.g., Wasatch Plateau) Sevier and Emery Gray can be almost indistinguishable. They have the same type of tempering, and often have similar surface coloring and surface finish. Comparison of the ceramic distribution of Emery Gray to Sevier Gray prior to R. Madsen's redefinition in 1972 is problem-

Discussion of Ground Stone The ground stone at Pint-Size Shelter is limited to shaped and unshaped metates and manos. Distinctively Fremont artifacts such as spindle whorls or incised tablets are absent. In general, ground stone artifacts are of limited usefulness as time markers. Although shaped ground stone at Pint-Size Shelter may be Archaic (cf. Fowler, Madsen, and Hattori 1973), all such artifacts at the site were recovered in stratigraphic association with Emery Gray ceramics. The one complete trough metate was recovered with a shaped mano in situ. Little ground stone is associated with the Archaic level, yet what was recovered is unshaped. One of the most distinctive characteristics of the Desert Archaic is the use of "flat milling stones" (Jennings 1964). At Pint-Size Shelter one slab metate as well as numerous one hand cobble manos, were recovered from the Archaic level.

Fig. 26 Ceramics (a) unsmoothed Emery Gray; (b) Gray rimsherd; (c) smoothed Emery Gray with mending hole. (Scale in cm.) 54

atic. It is possible that the preponderance of Emery Gray at sites such as Snake Rock is even greater than reported. Ranges of temper sizes overlap and Emery Gray with large-grained basaltic inclusions may be reported as»Sevier Gray (D. Madsen, 1975 personal communication).

Provenience Unknown

Artifact

Provenience

ABL E II

E a «ha

"3 o H •

• * *

1 55 g £ *

£ S „ ? o « r„ ^ < w

Points Typel Type II Type III Type IV TypeV Total Drills Typel Type II Total Bifaces Typel Type II Type III A Type III B Type III C Type HID Type IV TypeV Type VI Type VII Total Unifaces Typel Type IIA Type IIB Type III A Type III B Type III C Total Cores Typel Type II Total Hammerstones Typel Type II Total Utilized Flakes Waste Flakes Metates

2 9

1 3

1 1

2 2 4

2 4 6

9 1 1 1 6 5 3

11 3 2 1 8 10 5 4 13 17 74

1 1 2 4 7 10 28 1

2 1 3

267

807

1 1

1 12

Typel Type II Total Manos Typel Type II Total Emery Gray Sevier Gray Total

6 15

3 4 6 1 8 22

1 3 5

2 8

5 5

1 1 1 1 1 5

1 1 1

5 5 36 4 4 1 5 2 16 7

6 8 3 13 3 1 34

11 1 12

1 6 7

4 1 5

9 9 25 1 li

3 21 24 130 1 131

105

ARTIFACT SUMMARY The variety of artifacts from Pint-Size Shelter is limited, yet there are adequate provenience controls of a sufficient number of technically different items to draw some conclusions. The cultural deposits at Pint-Size Shelter show a general increase in artifact yields from the lowest level through the highest. Stratum 1 yielded nothing other than waste flakes and one Type II mano; Stratum 2 yielded more chipped stone, but only two were tool-like items; Stratum 3 evidenced all classes of chipped stone and ground stone. The most recent level, Stratum 4, is the richest in all classes of artifacts and is the only level to contain ceramics. An attempt was made to distinguish an upper and lower component in Stratum 4. The limited sampling of the lower component yielded no artifacts, so the general increase seems to be interrupted here. An analysis of the ceramic distribution shows the pottery concentration slightly skewed to the north and east of the excavated parts of the midden. However, there is

901 2,052

3 6 9 105

The limitation of provenienced ceramics to Plain Emery Gray ware and the early radiocarbon date (1970* 100 B.P.) for the stratum suggests an early Fremont occupation. Plain Emery Gray is the dominant ceramic variety for the San Rafael Fremont, although surface manipulation, including pinched, punched, incised, appliqued, and corrugated varieties are common (Marwitt 1970). Painted and surface-manipulated pottery appears at Snake Rock, Innocents Ridge, and 1-70 sites; all have late occupation ranges (ca. 900 to 1300 B.P.). The Gooseberry Site in Salina Canyon (DeBloois 1976 personal communication) has only a few plain gray sherds, and is dated fairly early (ca. 1290*100 B.P.). The Crescent Ridge Site (Madsen 1975) is also limited to plain gray ware (Emery Variety, 32 of 33 sherds) and is dated at 1170*100 B.P. Windy Ridge Village (Madsen 1975) has two radiocarbon dates. Structure I contained predominantly Emery Gray and was dated fairly early (1260*120 B.P.). Structure II contained painted pottery (Ivie Creek Black-on-white) as well as corrugated and punctated Emery Gray, and was dated considerably later (980*110 B.P.). This evidence is not conclusive, but suggests a sequence from plain ceramics to decorated ware in the San Rafael Fremont artifact assemblage (see Wilson and Smith 1976). The Fremont occupation at Pint-Size Shelter is probably early in Fremont cultural development.

an insufficient number of sherds and the excavated area is too small to draw any conclusions. Chipped stone significantly increases in number and type in Stratum 4. The provenienced arrow points come from this stratum. The proportion of unifaces in Stratum 4 relative to Stratum 3 decreases. The larger unifaces tend to come from Stratum 3 and the smaller from Stratum 4. The lithology in Stratum 4 shows a decrease in the use of silicified limestones. Ground stone exhibits the clearest change through time. There is no shaped stone below Stratum 4. A

classically Archaic flat slab milling stone was recovered from Stratum 3. The most telling difference in artifact technology is, of course, the presence of ceramics in Stratum 4 and its absence in Stratum 3. The association of ceramics with chipped stone in Stratum 4 suggests that these artifacts differ technically and/ or functionally from the chipped stone in Stratum 3. However, with the wide area of morphological overlap, it is problematic to reconstruct these differences.

SITE INTERPRETATION AND DISCUSSION Two components, Archaic hunter-gatherers and Fremont agriculturalists, are recognized at Pint-Size Shelter. Temporal placement and artifact types compare favorably, for the most part, with established typological and chronological sequences. Although lacking the spectacularity of nearby larger sites, the shelter provides cultural ecological information that fills a void in interpretations of the seasonal round, ordinarily lacking at larger more permanent sites. The evidence of Archaic occupation at the shelter is based directly on the presence of a Type I (Gypsum) "atlatl" point in the basal, cultural deposit (Stratum 3). The temporal span of the occupation is bracketed by radiocarbon determinations (4520±210 B.P. to 3390±170 B.P.) which closely correspond to Gypsum points dated from 4670*140 B.P. to 3360* 85 B.P. at Sudden Shelter on Ivie Creek (Jennings, Schroedl, and Homer n.d.). The close correlation of dates from the two shelters seems to herald the initial appearance of Gypsum points in the Castle Valley region at a period corresponding with their appearance in the southern Great Basin (cf. Fowler, Madsen, and Hattori 1973). Pinto points are not present at Pint-Size and appear earlier in the Sudden Shelter sequence (7840*330 B.P. to 6310*240 B.P.). They are also dated at Joe's Valley Alcove to 6200*190 B.P. (DeBloois, 1976 personal communication). Other artifacts recovered from Stratum 3 are nondiagnostic although flat slab milling stones are commonly recovered from Archaic deposits (Jennings 1957, 1964). The only cultural feature is an uninformative pit, which originates from the top of Stratum 2 corresponding to the ortset of Archaic occupance. Pint-Size Shelter probably served Archaic populations in much the same manner as the Archaic occupation (ca. 3000 B.P.) at Clyde's Cavern (Winter and Wylie 1974). Both sites are located in very similar arid habitats. Pinyon-Juniper is sparse and Saltbush-Greasewood communities predominate at both sites. Coincident with the presence of ground stone, the pollen records and macrofossil analyses at both sites suggest high grass and Cheno-v4m usage (Lindsay n.d.[a]; Appendix II). The absence of faunal

material at Pint-Size and its paucity at Clyde's Cavern implies a high degree of dependence on wild plant food. The two sites differ in amounts of chipped stone, with a larger proportion recovered from PintSize Shelter. The absence of bone and other perishable material in the context of considerable chipped stone at the shelter is enigmatic, but may also be attributable to differential preservation. In sum, although chipped stone tools were manufactured, direct evidence of hunting is entirely lacking. Pint-Size Shelter and Clyde's Cavern contrast with Sudden Shelter and Joe's Valley Alcove since the latter are located in entirely difference ecozones where water, pinyon/juniper, and montane flora are abundant. Both yielded copious amounts of chipped and ground stone artifacts and a large variety of faunal material, suggesting an entirely different subsistence focus at the montane sites. The two types of Archaic occupation at the four sites — all located within a ca. 50 mi. radius — suggest a seasonal round similar, perhaps, to that of the protohistoric Numic of the Great Basin (cf. Steward 1938). Abundant wild game, including deer from the montane zone, were probably taken following a summer's fattening on browse. The fall animal hunt may have taken place in conjunction with an annual pinyon harvest. The valley provided grasses and Cheno-Ams during the early to midsummer months, and the montane zone provided additional floral varieties somewhat later (Coulam and Kreis n.d.; Lindsay n.d.[b]). Archaic populations would have necessarily followed the "horn-of-plenty" when and wherever provident. Clear evidence of occupational gaps at Pint-Size and Sudden Shelters (ca. 3400 to 1800 B.P.) essentially supports that posited for the eastern Great Basin (Madsen and Berry 1975). However, the evidence is less conclusive from Clyde's Cavern and Joe's Valley Alcove. Archaic occupation of Clyde's Cavern yielded a single radiocarbon date ca. 3000 B.P.; a determination ca. 1490 B.P. dates a possible Basketmaker II component (Berry 1975). There is some indication that the Archaic and Basketmaker components were separated by a sterile component. The intervening stratum 56

(2), an alluvial deposit, contained few artifacts (Wylie 1971). Deposition seems to conform to Neoglacial conditions identified in the northeastern Great Basin (Madsen and Berry 1975), and seems to coincide with a spall deposit separating Archaic and Fremont components at Joe's Valley Alcove. However, a radiocarbon date (3520*120 B.P.) has been obtained from a charcoal lens within the spall component and additional dates (2460*120 B.P. and 2410*130 B.P.) were obtained either within or just above the spall (DeBloois, 1975 personal communication). No artifacts were associated with these dates. In sum, evidence of an hiatus separating the Archaic and Fremont occupations at these sites is inconclusive. Pint-Size and Sudden Shelters provide the only direct evidence of an occupational hiatus on the northwestern Colorado Plateau. The Fremont occupation at Pint-Size Shelter as defined by pottery, is radiocarbon dated at 1790*100 B.P. The date is derived from a charcoal sample obtained from a fire hearth originating within the lower half of the cultural component (Stratum 4). This marks the earliest known Fremont occupation. An "early" 1550*95 B.P. date was derived from the Snake Rock Site (Aikens 1967), but was subsequently rejected by Marwitt (1970) in the context of his synthesis of the Fremont and the scant data then available on the San Rafael Fremont. Elsewhere in the region, plain gray pottery, predominantly Emery type (see Ceramics in Artifacts) was used to the exclusion of painted and corrugated varieties at Joe's Valley Alcove (1410*100 B.P.), the Gooseberry Site (1290*100 B.P.) (DeBloois, 1976 personal communication), Windy Ridge Village (1260*120 B.P), Cresent Ridge (1170*100 B.P.), and Power Pole Knoll (1040*130 B.P.) (Madsen 1975). If the limits of one standard deviation for the 1790*100 B.P. Pint-Size date are accepted, there is a 67 percent chance that the true age of the Fremont fire hearth falls between 1890 and 1690 B.P. If two standard deviations are taken, there is a 95 percent chance that the true age is between 1990 and 1590 B.P. Acceptance of two standard deviations brings this date into line with the 1550*95 B.P. date from Snake Rock and the 1410*100 B.P. date on plain gray ceramics from Joe's Valley Alcove. Marwitt's (1970) rejection of the date from Snake Rock is seriously questioned on the grounds of these corroborative dates.

Surface-manipulated and decorated pottery â€” entirely absent at Pint-Size Shelter â€” appear later (ca. 1250 to 950 B.P.) in the San Rafael region at Windy Ridge Village (Structure II) (Madsen 1975), Snake Rock (Aikens 1967), Innocents Ridge (Schroedl and Hogan 1975), Turner-Look (Wormington 1955), and the 1-70 sites (Wilson and Smith 1976). This coincides with the appearance of decorated pottery in the Parowan variant of the Great Basin (Marwitt 1970; Berry 1972, 1974). Sevier Gray pottery is absent from the Fremont cultural deposit at Pint-Size although one sherd was recovered from the surface. This pottery type is present in higher percentages at the late dated sites of the San Rafael region. The chipped and ground stone artifacts recovered from Pint-Size Shelter are not necessarily Fremont diagnostics. However, arrow points (Types HI A) appear no earlier than 1500 B.P. (Madsen and Berry 1975) and, on this basis, the small triangular (Type III A) points are clearly differentiated from those of the Archaic substratum. The trough metate recovered from the Fremont component at Pint-Size has counterparts in excavated Fremont sites elsewhere. Although one metate of this type has been recovered from Archaic deposits at O'Malley Shelter in southern Nevada (Fowler, Madsen, and Hattori 1973), they are absent elsewhere in Archaic contexts. The slab-lined fire hearth from which the 1790* 100 B.P. date was derived is diagnostic of the San Rafael variant. These have been identified elsewhere in pit structures at Windy Ridge Village and Power Pole Knoll (where they occur with prepared clay rims) (Madsen 1975), Innocents Ridge (Schroedl and Hogan 1975), and Turner-Look (Wormington 1955). The large pit which originates in Stratum 4 at PintSize is nondiagnostic. However, an "earth oven" with similar dimensions and fill occurs at the Fallen Woman Site (Wilson and Smith 1976). The two are similarly devoid of bone. The absence of Fremont occupational structures at Pint-Size Shelter contrasts with nearby Innocents Ridge and the 1-70 sites and suggests less permanent occupance. Despite the general dependence on cultigens by Fremont populations (e.g., Innocents Ridge), there is no apparent difference in the subsistence focus of the Fremont and the earlier Archaic populations at Pint-Size Shelter. Evidently the use of cultigens by the Fremont departed little from the seasonal round followed by the Archaic. The best guess is that PintSize Shelter principally served as a collecting site for both populations.

APPENDIX I LATE QUATERNARY GEOMORPHIC HISTORY OF PINT-SIZE SHELTER by Donald R. Currey Department of Geography University of Utah ABSTRACT Eolian sand is the predominant sediment in each of the four strata recognized at the Shelter. The sand probably underwent alluvial transport to the piedmont at the eastern edge of the Wasatch Plateau and eolian redistribution to the Shelter, mainly during relatively warm (interstadial) climatic intervals; cooler (stadial) intervals seem to have been times of sand stability. The basal strata (1 and 2) are undated, but are probably no younger than Altithermal. Radiocarbon dates indicate that strata 3 (Archaic) and 4 (Fremont) were deposited during earlier and later Neoglacial interstades, respectively. Radiocarbon evidence for a major nondepositional hiatus between Strata 3 and 4 is corroborated by differences in CaC03 epipetric cortexes associated with those strata.

INTRODUCTION Cretaceous age (Williams and Hackman 1971). Regional drainage is via the Muddy Creek basin, a tributary of the Dirty Devil and Colorado Rivers, and is generally southeastward (Hackman and Williams 1972). Local drainage is westward, into a southwesttrending reach of Christiansen Wash. Local relief within a few hundred meters of the site is approximately 40 meters. Local topography is dominated by discontinuously mantled sandstone knobs and swales that are marked by many bedrock ledges, ranging from a fraction of a meter to several meters high, and by several ephemeral watercourses. The sandstone is pale brown (Munsell scale: 10YR 7/3), moderately calcareous, medium-grained, moderately well sorted, and predominantly quartzose. In outcrop, the sandstone is friable to case-hardened, flaggy-to-massive, and where massive tends to be characterized by crossbedding, prominent high-angle joints, and cavernous weathering.

A series of late Quaternary geomorphic episodes is evidenced by surficial materials in the general vicinity of Pint-Size Shelter and within excavations dug by the Antiquities Section of the State of Utah. The following interpretations are based on observations and sampling conducted at the site by the writer on September 29, October 12, and November 22, 1975. Laboratory work was performed in the Gvosdetsky Laboratory for Environmental and Paleoenvironmental Studies at the University of Utah. In this project, I have benefited from helpful discussions with Michael S. Berry, La Mar W. Lindsay, and David B. Madsen. The site is near the west edge of a low interfluve, in a shadscale scrub/ big sagebrush/pinyon-juniper ecotone (Hackman 1973), at an altitude of approximately 1,850 meters. The interfluve is underlain by gently dipping bedrock strata belonging to the Ferron Sandstone Member of the Mancos Shale, of Upper

EARLIER GEOMORPHIC EPISODES WASATCH PLATEAU GRAVELS

REGIONAL DISSECTION

A discontinuous scattering of siliceous pebbles and cobbles occurs on exposed bedrock and on bedrock beneath younger cover sands at many places in the vicinity of the Shelter. The siliceous clasts are wellrounded and consist mainly of chert and quartzite, which are exotic with respect to local bedrock sources. It is highly probable that in late Tertiary or early Quaternary time the local landscape was part of a relatively continuous piedmont alluvial system, to which bed load was supplied by headwaters in the Wasatch Plateau, to the west and north.

Subsequently, probably in Quarternary time, the piedmont alluvial system underwent regional and local dissection. The local relief that is visible today began to emerge and interfluves became isolated from renewed inputs of siliceous gravels as Wasatch Plateau alluvium was increasingly confined to incised streams, leaving only dwindling lag gravels to persist into late Quaternary time on the interfluves.

LATE QUATERNARY GEOMORPHIC EPISODES In late Quaternary time, down to the present, the interfluve on which the Shelter is situated has been modified geomorphically by weathering, slope wash, eolian deposition, and eolian deflation. The sediments that are ascribable to geomorphic, or noncultural, stratigraphic processes at Pint-Size Shelter are mainly eolian deposits, although slope-wash deposits and bedrock rubble are present as well (see Table III). Several episodes of eolian deposition and nondeposition, probably no older than late Pleistocene and very

possibly no older than Holocene, are evident at the site. Eolian deposition produced sheetlike bodies of cover sand and, very locally, small dunes. Accretion of the cover sand seems to have been intermittent, with intervening episodes of relatively stable land surfaces during times of no net sand accretion. The sequence of probable geomorphic episodes represented by strata at Pint-Size Shelter (Fig. 27) is summarized in stratigraphic order in Table IV.

STRATA 1 AND 2 COVER-SAND ACCRETION and local slope wash, may predate Stratum 2, which is obviously pigmented with carbon from man-made fires and which seems in its sand-size fraction to be mainly eolian. Sandstone blocks and slabs of various sizes comprise approximately one-third by volume of both strata. Most of the larger sandstone fragments appear to have been spalled along bedding planes, and very

The basal strata, resting directly on the bedrock floor of the Shelter, are Strata 1 and 2. No radiocarbon date has been obtained from either stratum and even their relative age relationships are in some doubt. Stratum 1, which is light brown and in its sandsize fraction seems to be a mixture of eolian sand

TABLE III Selected Summary Properties of Sediment Samples from Pint-Size Shelter and Nearby Localities Percent by Weight" Sample Stratum 4, upper Stratum 4, lower Stratum 3 Stratum 2 Stratum 1 Knoll1 Dune, upper k Dune, lower'

Coarse sand" 9.6 8.6 16.2h 11.2 26.8' 9.0 7.2 10.8

Fine sand 77.8 78.8 70.0 79.0 68.4 80.4 82.8 71.6

Dust'

PH'

HCI Effervescence

12.6 12.6 13.8 9.8 4.8' 10.6 10.0 17.6

8.6 8.6 8.3 8.3 9.2 8.4 8.6 8.4

freely freely freely freely freely freely freely freely

Munsell Color" 10YR 10YR 10YR 10YR 10YR 10YR 10YR 10YR

5/3 5/2.5 4/1.5 4.5/2 6/3 6/4 6/3.5 7/2.5

Apparent Cultural Carbon very low low

high medium nil nil nil nil

Determined by dry sieving 50-g. samples. The samples included only relatively fine materials. However, Strata 1-4 all include sandstone blocks and slabs, orflagstones.An estimate of the coarser textural components in Strata 1-4 inclusive, based on a 420-point count in the sides of the excavation, is: sandstone blocks and slabs >10 cm. in longest dimension, 12 percent; smaller blocks and slabs, 22 percent; and fines, 66 percent. These proportions do not differ greatly from stratum to stratum. In each stratum the largest slabs average about 40 cm. long and 5 cm. thick. Thefinescomprising the 50-g. samples are composed predominantly of quartz grains and all samples appear to be largely wind-deposited. However, none of the samples shows evidence of long-distance wind transport, i.e., none shows exceptional rounding, frosting, or sorting. Includes coarse sand, very coarse sand, and granule sizes (0.5 to 4 mm.). No pebbles were observed except as noted in the sample of Stratum 3. c

Includes veryfine,fine,and medium sand sizes (0.0625 to 0.5 mm.). Includes silt and clay sizes (< 0.0625 mm.). e

1:1 mixture of sample and deionized water. 1.25 molar HCI. Air-dry sample.

Includes pebble-size sandstone fragments, possibly heat-shattered. Similar texturally to local bedrock. May contain much slope wash derived from cavernous weathering of sandstone at rear of Shelter. ^30-cm depth in cover sand on crest of knoll where dirt road approaches site, about 100 m. east of Shelter. 10- to 70-cm. depth in trenched sand dune, about 80 m. northeast of Shelter. 70- to 100-cm. depth in trenched dune. 60

Fig. 27

Schematic north-south profile of Pint-Size Shelter, showing relative positions of major stratigraphic units other strata at the Shelter, approximately one-third of the material comprising Stratum 3 consists of pieces of sandstone bedrock. Most of the larger pieces are very tabular, so much so that they do not appear to have been derived geomorphically from the nearest available bedrock, as in Strata 1 and 2, but instead appear to have been carried in selectively from scattered flagstone outcrops 10 to 30 m. beyond the southern perimeter of the Shelter. The age of Stratum 3 is closely bracketed by radiocarbon dates of 4520 B.P. and 3390 B.P., from near the base and top of the stratum respectively. By most Holocene paleoclimatic interpretations, that time interval is regarded as comprising an early-to-middle Neoglacial interstadial characterized by a slightly elevated upper timberline and by diminished mountain glaciers and pluvial lakes (e.g., Denton and Porter 1970; Denton and Karlen 1973; Currey and Madsen 1974).

possibly frost-heaved upward, from the subjacent bedrock floor. Based on indications provided by overlying strata, accretion of Strata 1 and 2 probably did not occur any later than the Altithermal climatic episode. For reasons that are discussed below, cover-sand accretion at Pint-Size Shelter is believed to have occurred during relatively warm (interstadial or interpluvial) climatic episodes. POST-STRATA 1 AND 2 COVER-SAND STABILITY Although the minimum age of Stratum 2 is unknown, it is very likely that accretion was followed by an interval of relatively little eolian deposition. Based on constraints provided by overlying strata, an interval of cover-sand stability subsequent to the accretion of Strata 1 and 2 probably did not occur any later than the early stade of Neoglaciation. For reasons that are discussed below, cover-sand stability at PintSize Shelter is believed to have occurred during relatively cool, stadial (little ice age or little pluvial) climatic episodes.

POST-STRATUM 3 COVER-SAND STABILITY Accretion of Stratum 3 was followed by a radiocarbon-bracketed period from about 3390 B.P. to about 1790 B.P. during which eolian deposition at the Shelter seems to have been scant or nonexistent. That time interval corresponds almost exactly to a widely recognized middle Neoglacial stadial characterized by a lowering of upper timberline and by a rebirth

STRATUM 3 COVER-SAND ACCRETION In its finer sizes, Stratum 3 is mainly eolian sand with admixed pebble-size sandstone fragments of possible fire-shattered origin. The stratum is darkly pigmented with carbon from man-made fires. As in 61

or expansion of mountain glaciers and pluvial lakes (ibid.).

POST-STRATUM 4 COVER-SAND STABILITY Although the time cannot be bracketed exactly, it would appear that much of the later portion of the present millennium has been a period of negligible cover-sand accretion at Pint-Size Shelter. If true, that interval would accord closely with the last major'Neoglacial stadial, i.e., with the Gannett Peak Stade of Neoglaciation in Utah and neighboring states (Richmond 1965).

STRATUM 4 COVER-SAND ACCRETION

A radiocarbon date indicates that accretion of Stratum 4, which in its finer sizes is essentially indistinguishable from sand in nearby dunes, began shortly before 1790 B.P. As in Stratum 3, Stratum 4 contains many very tabular sandstone slabs that appear to have been carried in selectively from behond the perimeter of the Shelter. The assemblage of Fremont artifacts recovered from the stratum suggests that accretion of Stratum 4 probably continued into the early portion of the present millennium. Therefore, the time span of Stratum 4 cover-sand accretion would seem to accord very closely with the last major Neoglacial interstadial interval (ibid.).

MODERN DENUDATION During the present century, eolian deflation and erosion by sheet wash and gully wash seem to have been making serious inroads into local cover sands, perhaps in large part because vegetation cover has been reduced through overgrazing. As a result, recently exposed bedrock is widespread in the vicinity of the Shelter. To the west, the flood plain of Christiansen Wash is aggrading due to an excessive load of sand from nearby interfluves.

EVIDENCE PROVIDED BY EPIPETRIC CORTEX THICKNESSES On the slightly overhanging bedrock face at the head of Pint-Size Shelter, the sandstone below the pre-excavation ground surface is coated with an epipetric cortex. The bedrock floor of the Shelter and the upper surfaces of sandstone blocks and slabs in the Shelter fill are likewise coated with the cortex. Composed of microcrystalline CaC0 3 and hard enough to be considered petrocalcic, the cortex clearly seems to be of pedogenic origin. The CaCO ? is derived from eolian parent material in proximity to the bedrock, and is not leached from within the bedrock itself. The contact between the cortex and the bedrock is abrupt and the CaC0 3 content of the calcareous sandstone bedrock adjacent to the contact seems not to be depleted. The transfer of CaC0 3 from eolian calcareous parent material to cortex might well be accomplished by C0 2 -charged capillary water.

One of the most striking features of the cortex is its variable thickness. At the urging of Michael S. Berry, a concerted effort has been made to examine cortex thicknesses for possible relationships to the stratigraphy of the Shelter. Cortex samples were collected from a prominent buttress on the bedrock face at the head of the Shelter (Fig. 28). Cortex measurements are summarized in Table V. In addition to thickness, internal layering, similar to that in speleothems, is observable. Both lines of evidence strongly suggest a twofold stratigraphy. Cortex adjacent to Stratum 3 fill has a mean thickness of 1.21 mm (n = 84, s.d. = 0.38 mm.) and is double-layered internally. Cortex adjacent to Stratum 4 fill has a mean thickness of 0.14 mm (n = 43, s.d. = 0.12 mm) and is single-layered. There is no cortex on above-ground surfaces. The change from thicker cortex adjacent to Stratum 3 and

TABLE IV Summary of Holocene Geomorphic Episodes at Pint-Size Shelter, with Possible Correlations Geomorphic Episodes

Possible Chronology'

Possible Equivalent Holocene Climatic Episodesb

Modern denudation Post-Stratum 4 sand stability Stratum 4 sand accretion Post-Stratum 3 sand stability Stratum 3 sand accretion Post-Strata 1 and 2 sand stability Strata 1 and 2 sand accretion

present to 100 (?) B.P. 100(?)to650(?)B.P. 650(?)toca.l790B.P. ca. 1790 to ca. 3390 B.P. ca. 3390 to ca. 4520 B.P. ca. 4520 to 5500 (?) B.P. 5500 (?) to 8000 (?) B.P.

Late Neoglacial stadial Middle/late Neoglacial interstadial Middle Neoglacial stadial Early/middle Neoglacial interstadial Early Neoglacial stadial Altithermal

Bracketing radiocarbon dates from Pint-Size Shelter are underlined. See also Fig. 29. 62

TABLE V Thicknessaof CaC0|3 Epipetric Cortex on Samplesbfrom Bedrock Face at North Edge of Pint-Size Shelter

Sample No. (See Fig. 28)

Adjacent Stratum

1 2 3 4 5 6 7 8 9

Stratum 3 Stratum 4 above ground Stratum 4 Stratum 3 Stratum 3 Stratum 4 above ground Stratum 3 Stratum 1&

Number of Fragment Edges Observed n

Mean Thickness (mm) X

Standard Deviation (mm) s.d.

37 9

1.22 0.14

0.37 0.16

1.55 0.13 0.00 1.02 1.60

0.34 0.11 0.00 0.29 0.20

15 34 many

32 8

* Measured using 20X total magnification and 0.04 mm. micrometer resolution. b Samples of cortex and underlying bedrock were removed as a unit with hammer and chisel. c Control sample to determine if lithologic differences in the bedrock could account for differences in cortex thickness. It is concluded that cortex thickness is probably independent of bedrock lithology, which is essentially homogenous among the samples examined. d

Cortex on upper surface of horizontal bedrock slab at base of Stratum 2.

50-i

, Bedrock

©.'

Recessed bedrock

cm. above

buttress

Recessed bedrock

datum

'Present Surface

©i

Stratum 4

\ Stratum 3 cm. below datum

©\

50 cm. east

Fig. 28

100

datum

Front elevation of bedrock face at head (north end) of Pint-Size Shelter, showing locations of epipetric cortex samples (see Table V). 63

'••.. Relative Glaciation (Denton and Porter, 1970, p. 107)

0.6

0.5

>. E E

•

ENg

C Date

E/MNg

M/LNg

LNg

0.4

to 4-

c o

0.3

< Z

1 l Ol3- 314 ^

B.P.

Fig. 29 Apparent rates of net accretion of Strata 3 and 4 at Pint-Size Shelter in relation to the average extent of mountain glaciers world-wide as a generalized record of global climatic changes during the late Holocene Neoglaciation (Denton and Porter 1970). ENg, MNg, and LNg are early, middle, and late Neoglacial stadials, or little ice ages, and E/MNg and M/LNg are intervening interstadials, or little interglacials. 64

thinner cortex adjacent to Stratum 4 is abrupt, and not a transition marked by gradual thinning. When compared with available radiocarbon dates from the Shelter fill, it is clear that cortex thickness is related to the age of adjacent unconsolidated sediments in a

manner that is direct, but not linear, i.e., although Stratum 3 sediments are roughly three times as old as Stratum 4 sediments, Stratum 3 cortex is approximately nine times as thick as Stratum 4 cortex.

GEOMORPHIC-PALEOCLIMATIC INTERPRETATION OF EPISODIC COVER-SAND ACCRETION be pointed out, however, that his sequence of geomorphic episodes is inferred on the basis of a minimum number of radiocarbon dates. The following is a possible paleoclimatic explanation of the dated geomorphic record.

Evidence examined thus far strongly suggests that tolian sedimentation was episodic, and not continuous, at Pint-Size Shelter during Holocene time (Fig. 29). Sand accretion seems to have occurred mainly during relatively warm periods. Accretion seems to have been greatly reduced during relatively cool periods. It must

STADIAL EPISODES

INTERSTADIAL EPISODES

Relatively cool, stadial episodes were probably associated with an expanded circumpolar vortex. Westerly flow and winter precipitation, much of it in the form of snow, would have been strengthened at the expense of the summer rainfall, Mexican monsoon, regime. Effective soil moisture would have been relatively high, plant cover enhanced, runoff moderated, and sediment yield minimized. Alluvial supply of sand for eolian redistribution to interfluves would have been greatly curtailed. Cover sands deposited previously would be inactive geomorphically, but would be relatively well covered by vegetation and relatively well endowed with soil moisture, leading to intensified pedogenic activity. Epipetric cortex deposition would probably have been more active during the pronounced, mid-Neoglacial stadial interval following Stratum 3 cover-sand accretion than during any time since. The present climate in the region of Pint-Size Shelter is probably intermediate between the regimes described above. The annual precipitation of slightly less than 20 cm. per year at the Shelter has a clear-cut maximum in late summer (Covington and Williams 1972), but not nearly as pronounced a summer maximum as that which occurs under the influence of the strong Mexican monsoon regime a few hundred kilometers to the southeast. Four cloudburst floods, which are notable for their alluvial fan building and which occur mainly in July and August, have occurred within a 10-km. radius of Pint-Size Shelter in historic times (Hackman and Williams 1972). Cloudburst floods, as a vehicle for bringing sand-rich debris into position for subsequent eolian redistribution onto interfluves, would doubtless be an even more potent geomorphic agent under a more intense summer rainfall regime. In short, Pint-Size Shelter seems to have been subject to episodic eolian sedimentation resulting from a delicately balanced chain of paleoclimatic-geomorphic cause and effect, thereby presenting an exceptionally detailed and sensitive Holocene stratigraphic record.

Relatively warm, interstadial episodes were probably associated with a contracted circumpolar vortex. Westerly flow and winter precipitation would have been relatively weak, replaced by strengthened Mexican monsoon circulation and accompanying summer rainfall. Precipitation effectiveness measured in terms of available soil moisture would have been relatively low, plant cover on xeric and sub-mesic sites relatively diminished, intense convectional precipitation events relatively frequent, and resulting sediment yield relatively high. Streams transporting large quantities of sediment, including much sand, from Wasatch Plateau sources to the west would have provided aggrading, easily deflated piedmont alluvial deposits as sources for eolian redistribution onto low interfluves nearby. Given a favorable climatic regime, the topography and lithology of the Wasatch Plateau slopes that drain in the general direction of Pint-Size Shelter are a combination that is well suited to providing high yields of sand-size sediment. Topographic relief on the east slope of the Plateau upstream from the vicinity of the Shelter averages 800 m. and ranges up to 1600 m. (Covington 1972). Relatively steep slopes and stream gradients are commonplace. The types of bedrock that underlie the Plateau slopes are sandstone, sandstone and shale, sandstone and conglomeratic sandstone, and shale containing significant percentages of sandstone (Williams 1972). Following removal from a Plateau source region of about 300 sq. km., sand-size sediment would tend to undergo alluviation in a lowgradient piedmont region of about 100 sq. km. between the Plateau and the Shelter. Walker Flat, Quitchupah Creek, and Christiansen Wash piedmont alluvium would be particularly well situated for subsequent eolian transport toward the Shelter by prevailing winds from the southwest.

APPENDIX II SITE PALEOECOLOGY: PALYNOLOGY AND MACROFOSSIL ANALYSIS by La Mar W.Lindsay ABSTRACT Thirteen samples were collected during the course of excavations at Pint-Size Shelter, a dual component Archaic/Fremont site in Emery County, Utah. The samples were obtained from soil deposits, metates, fire hearth, and a coprolite and subjected to palynological and macrofossil analysis. The results suggest a considerable temporal separation between the two cultural components. Despite the problems of pollen preservation and identification, the Archaic (ca. 4000 B.P.) is characterized by slightly less pine pollen production and a concomitantly higher CHENO-/lm production. This may indicate a slightly drier climate than that of today. During Fremont times vegetation was much the same as today. A variety of floral types was variously utilized during both periods including grasses, CUENO-Ams, greasewood, and yucca. The absence of bone in all deposits and cultivation in later levels suggests that Pint-Size Shelter apparently served similar purposes for both populations — seasonal gathering and lithic manufacture.

INTRODUCTION at Pint-Size Shelter. These data supplement existing paleoecological information for Castle Valley and San. Rafael Swell populations (Lindsay 1975, n.d.[a], n.d. [b]; Coulam and Kreis n.d.; Winter and Wylie 1974).

Soil samples were collected from archeological deposits and subjected to both extraction and flotation for palynological and macrofossil analysis. The samples yield information regarding the subsistence strategies employed by Archaic and Fremont populations

PROCEDURE All samples contained pollen that was heavily eroded and poorly preserved, making identification difficult. Samples collected from the contemporary surface and the fire hearth tended to be better preserved. Several samples were taken from metates by first brushing the grinding surfaces thoroughly in order to remove the surrounding soil matrix. This was followed by scrubbing and rinsing with deionized and filtered/ distilled water in order to sample sediments deposited during metate use. All but two of the samples in which counts were obtained (Nos. 2 and 5) reflect 200 grain counts. These are ordinarily sufficient to allow a relatively firm definition of economic floral types used by prehistoric populations (Martin 1963). However, poor preservation and limited sampling render any inferences somewhat tenuous.

Thirteen samples were obtained from site deposits and cultural features in the course of excavation. The samples were taken from a variety of phenomena including soil strata, pit fill, metates, fire hearth, and a coprolite. The samples cover the full span of Archaic (4520±210 to 3390*170 B.P.) and Fremont (1790±100 B.P.) occupations at the site. In addition, a contemporary surface sample was collected which reflects, to some degree, present floral distributions. All palynological samples were extracted following the procedures in general use for alluvial samples from the American Southwest and the Great Basin (Mehringer 1967). Extracted material was stained with basic fuchsin, mounted in glycerol, and counted under 600x magnification. Identifications were made using modern reference material made available by the University of Utah Herbarium. Pollen abundance and the state of preservation were highly variable, particularly in the earlier deposits.

The thirteen samples collected (Fig. 30) are described below. 1. The sample, obtained from precultural Stratum 1, the earliest deposit at the site, was collected from the east profile in the southern part of the exploratory trench. The sample was not used due to insufficient pollen, attributed to poor preservation â€” soil ph 9.2 (Appendix I). 2. The sample, obtained from Stratum 2, the earliest stratum containing cultural material, but derived from Stratum 3 above, was collected from the north profile in the center of the trench. 3. The sample, also obtained from Stratum 2, was collected from the north profile 50 cm. west of the center of the trench. Both samples (2 and 3) were collected from the matrix which contained the stone slabs present when the site was first occupied. The two samples (2 and 3) are combined providing a 300 grain count from the stratum. 4. The sample, obtained from Stratum 3, was collected from the north profile 50 cm. west of the center of the trench. Taken from the center of the deposit, the sample reflects some vegetational types present at the stratigraphic/temporal midpoint of Archaic occupation (ca. 4000 B.P). 5. The sample was obtained from a slab metate located wholly within Stratum 3. The metate, found lying face down, was taken from the west profile in the approximate center of the trench. 6. The sample, obtained from the lower component of Stratum 4, was collected from the north profile 50 cm. west of the center of the trench. The sample represents the earliest deposit resulting from Fremont occupation, the second and final occupation of

the site. 7. The sample was obtained from a fire hearth located approximately in the center of the trench. The hearth, which is associated with Fremont pottery and dated ca. 1800 B.P., originates from within the lower half of Stratum 4. 8. The sample, obtained from the upper half of Stratum 4, was collected from the north profile 50 cm. west of the center of the trench. The sample represents the final Fremont occupation at the site. 9. The sample was obtained from the center of the fill of Pit 2, located about 2 m. east of the center of the exploratory trench. The pit originates from the top of the lower component of Stratum 4. The upper component provides the fill of the pit. 10. The sample was obtained from a trough metate located in the center of the fill of Pit 2. 11. The sample was obtained from the trough of a metate fragment located within the fill of Pit 2. Both metate samples (Nos. 11 and 12) are artifacts of Fremont subsistence. 12. The sample was obtained from a coprolite located within the upper half of Stratum 4. In addition two samples, taken from deposits for macrofossil analysis are described below. 1. About 2 liters was obtained from Stratum 3 in the northern part of the exploratory trench. 2. About 500 milliliters was obtained from the trough metate found within the fill of Pit 2. The samples were processed by screening and flotation techniques as outlined essentially by Struever (1968) and used elsewhere on materials from Fremont and Anasazi contexts (Berry 1972; Wilson 1974).

RESULTS

the site vicinity, are anemophilous pollenated and are expected to be well-represented in the general pollen rain. Numerous shrubs such as sage {Artemisia tridentata A. spp.), snakeweed {Gutierrezia sarothrae), and other COMPOSITAE and forbs including saltbush {Atriplex canescens, A. spp.), greasewood {Sarcobatus vermiculatus), and rabbitbrush {Chrysothamnus nauseosus, C. spp.) have been identified in the vicinity of the site and elsewhere in Castle Valley (Schroedl and Hogan 1975, Appendix III). These types, mostly entomophilous pollenated, are also expected in the pollen record because of their proximity to the site. Grasses (GRAMINEAE) as well as a number of xerophytic species such as Mormon tea {Ephedra

THE POLLEN RECORD AND MODERN VEGETATION All pollen types identified from the archeological deposits are represented in floral varieties present today in Castle Valley (Schroedl and Hogan 1975, Appendix III) and on the Wasatch Plateau (Albee n.d.) The contemporary surface sample (No. 13) compares favorably with present day floral distributions. The conifers such as pinyon {Pinus edulus, P. monophylla), fir {Abies concolor, A. lasiocarpa), douglas fir {Pseudotsuga menziesii) and spruce {Picea engelmanni, P. pungens), and other types such as oak {Quercus gambelii, Q. turbinella) and aspen {Populus tremuloides) are present on the plateau. These types, in addition to juniper {Juniperus spp.) located in 68

sSc

I •_•_•_- I •

"-3 •a

a. ci

S 69

nevadensis), prickly-pear cactus {Opuntia spp.) and yucca {Yucca angustissima) have been identified on the site. The more mesic species such as lilies (LILIACEAE), sedges (CYPERACEAE), and willow {Salix sp.) are present in the wash below. The variety and relative percentages of floral types represented in the pollen record vary according to locational and distributional factors, modes of dispersal of the various types, climatic/pollen production factors, cultural preferences (utilization), and differential preservation of fossil pollen. These factors are considered in interpreting the record.

say n.d.[a], n.d.[b]; Martin and Sharrock 1964; Winter and Wylie 1974) and ethnohistorical evidence (Fowler and Fowler 1971; Kelley 1964; Steward 1938). In general the coprolite (Sample 12) should provide firm dietary information, however, ultimate origins may be questioned. The sample, which consisted of bone fragments and a variety of celluloid vegetal material, meets two of the three criteria for distinguishing human origins (Fry 1970). Coloration does not suggest that the sample was derived from a canine. The metates (Samples 5, 10, and 11) provide the more reliable economic information; the fire hearth (Sample 7) less so, but is included because of the better preservation of the sample. The high grass percentage from the slab metate (Sample 5) is consistent with interpretations of high grass utilization elsewhere during Archaic times (Fry 1970; Harper and Alder 1970; Kelso 1970; Lindsay n.d.[a]). High CHENO-/4w percentages vis-a-vis other floral types suggest that the chenopods were abundant and were also heavily utilized. This is further supported by the identification of goosefoot seeds obtained from the macrofossil analysis of Stratum 3. Grass pollen is also present on the metates (Samples 10 and 11) from the Fremont component, but in lesser amounts than from the Archaic. Grass is identified in use during Fremont times elsewhere, particularly at Clyde's Cavern (Lindsay n.d.[a]; Winter and Wylie 1974). The high CHENCMm percentages (compare with present day levels, Sample 13) suggest chenopods were harvested during the Fremont occupation. No macrofossils were obtained from Fremont deposits. Pollen percentages from the coprolite sample suggest human origins. The extremely high percentage of greasewood (24%) {Sarcobatus vermiculatus) suggests economic usage. Similar high percentages have been noted in coprolite samples from Fremont levels at Hogup Cave (Kelso 1970) and Clyde's Cavern (Lindsay n.d.[a]), and from a midden component at Innocents Ridge (Lindsay 1975). Greasewood is known to have provided a source for arrow shafts at Hogup Cave (Aikens 1970) and during ethnohistoric times (Steward 1938). The only known use of the plant for consumption is reported for the Cocopa of southern Arizona (Kelley n.d. in Kelso 1970). The coprolite sample also yielded evidence of lily (cf. Yucca spp.), which is reported in the ethnohistorical record (Kelly 1964). Other floral types including buffaloberry {Shepherdia canadensis), sage, Opuntia sp., Ephedra sp., and oak are variously present in the pollen record and are recorded as having been used during historic times (Fowler and Fowler 1971; Kelly 1964; Steward 1938). The absence of maize {Zea mays) pollen at PintSize Shelter probably indicates the site was not directly associated with cultivation. This contrasts with the identifications of Zea at nearby Innocents Ridge and Clyde's Cavern (Lindsay 1975; n.d.[a]). However, the absence of Zea may be attributed to preservational

TEMPORAL VARIATION The two cultural units, Archaic (ca. 4500 to 3400 B.P.) consisting of Stratum 3, and Fremont (ca. 1800 B.P.), Stratum 4, are separable on palynological grounds. 1. Marked differences in the preservation of fossil pollen is most striking between the two cultural units. In general, anemophilous types such as the conifers and grasses were less fragmented and distorted in the Fremont samples. Conversely, the higher percentages of CHENOPODIACEAE-.4ma/w2//*w.s (CHENOAms) and other pollen types generally more resistive to oxidation in the earlier strata, may be attributable in part to the lack of preservation of the more fragile types. 2. In general, the quality of the samples from the earlier deposits contained considerably less vegetal material, higher percentages of spores (not identified), and significantly greater amounts of siliceous material. If attributable to preservation these differences suggest a considerable time differential between the two cultural components. This is supported by the radiocarbon determinations and the geological assessment (see Appendix I). 3. If the differing percentages between the two cultural units are attributable to vegetation differences, rather than preservational or cultural factors, the limited sampling at Pint-Size Shelter suggests that vegetation during mid-Archaic times may have consisted of slightly less pine or at least less pine pollen production and concomitant high CHENO-/4m production. This may indicate a slightly drier climate than at present. The pollen record indicates that during Fremont times vegetation was not significantly different than today. Similar conclusions are drawn from the pollen records at Sudden Shelter (Lindsay n.d.[b]) and Joe's Valley Alcove (Schoenwetter 1974).

CULTURAL PREFERENCES A number of "economic" pollen types are defined from metates and other cultural phenomena. The types are interpreted in contexts of other archeological (Fry 1970; Harper and Alder 1970; Kelso 1970; Lind70

factors as suggested by the generally low percentages of other anemophilous pollen types at the site. The tentative identification of beeweed (cf. Cleome) from Fremont levels suggests that the plant was utilized.

Martin and Sharrock (1964) have demonstrated that beeweed was heavily utilized by the Anasazi in the Glen Canyon.

SITE PALEOECOLOGY Given the presence of chipped and ground stone and the absence of bone during Archaic occupance, Pint-Size Shelter probably served as a seasonal (summer) campsite for the collection of wild vegetal foods. The site was probably occupied only intermittently over the 1000-yr. Archaic span. The relatively large amount of chipped stone including finished atlatl points and debitage indicates that tools were manufactured at the site. Direct evidence of hunting activity is absent and is somewhat enigmatic considering the large amounts of faunal material (including antelope) recovered from nearby Sudden Shelter (Jennings, Schroedl, and Homer n.d.). During Fremont times, Pint-Size Shelter probably served a similar function â€” that of seasonal occupance in which subsistence activities were limited to collecting. Direct evidence of agriculture at the site is entirely lacking. The absence of bone compares somewhat to

the limited amounts recovered from nearby Innocents Ridge (Schroedl and Hogan 1975). However, the Innocents Ridge Site offers at least minimal evidence of agriculture in the forms of pollen and macrofossils and a more extensive, permanent settlement. The larger nearby Snake Rock Site (Aikens 1967) may have been the source for the seasonal Fremont occupance of Pint-Size Shelter. In sum, Pint-Size Shelter appears to have provided temporary occupation for small populations engaged in collecting and the manufacture of stone tools. Two separate and ordinarily diverse populations occupied the site: the Archaic who maintained a relatively high degree of dependence on collecting, and the Fremont for whom collecting was supplemental to agriculture. The evidence at Pint-Size Shelter suggests that collecting was a major feature of the San Rafael Fremont subsistence strategy in Castle Valley.

REFERENCES Aikens, C. Melvin 1967 Excavations at Snake Rock Village and the Bear River No. 2 Site. University of Utah Anthropological Papers, No. 87. Salt Lake City.

1974 The Evans Mound: Cultural Adaptation in Southwestern Utah. MA thesis, University of Utah. Salt Lake City.

1970 Hogup Cave. University of Utah Anthropological Papers, No. 93. Salt Lake City.

1975a An Archeological Survey of the Northeast Portion of Arches National Park. Antiquities Section Selected Papers, Vol. I, No. 3. Salt Lake City.

Albee, Beverly n.d. Plant Communities in the Area of Sudden Shelter. In Sudden Shelter, by Jennings, Schroedl, and Homer. University of Utah Anthropological Papers (in preparation). Salt Lake City.

1975b Archeological, Historical, and Paleontological Survey for Consolidation Coal Company in Emery County, Utah. A Special Report, Antiquities Section, Division of State History. Salt Lake City.

Berge, Dale L. 1973 An Archaeological Survey in the Castle Valley Area, Central Utah. Ms. on file, Brigham Young University. Provo.

Burnham, Robert C. 1950 The Climates of Utah. MA thesis, University of Utah. Salt Lake City. Coulam, Nancy J. and Peggy R. Kreis n.d. Paleoethnobotanical Analysis. In Sudden Shelter, by Jennings, Schroedl, and Homer. University of Utah Anthropological Papers (in preparation). Salt Lake City.

Berry, Claudia H. 1974 Preliminary Report of an Archaeological Survey in Sevier, Emery, and Garfield Counties, Eastern Utah. Ms. on file, Department of Anthropology, University of Utah. Salt Lake City.

Covington, Harry R. 1972 Topographic Relief Map of the Salina Quadrangle, Utah. U.S. Geological Survey Folio of the Salina Quadrangle, Utah, Map 1-591-C. Washington.

Berry, MichaelS. 1972 The Evans Site. A Special Report, Department of Anthropology, University of Utah. Salt Lake City. 71

Covington, Harry R. and Paul L. Williams 1972 Map Showing Normal Annual and Monthly Precipitation in the Salina Quadrangle, Utah. U.S. Geological Survey Folio of the Salina Quadrangle, Utah, Map 1-591-D. Washington. Crabtree, Don E. 1967 Notes on Experiments in Flintknapping: 3, The Flintkapper's Raw Materials. Tebiwa, Vol. 10, No. 1. Pocatello. Crabtree, Don E. and B. Robert Butler 1964 Notes on Experiments in Flintknapping: Heat Treatment of Silica Materials. Tebiwa, Vol. 7 No. 3. Pocatello.

Archaeological Excavations at Snake Rock, Central Utah. Ms. on file, Department of Anthropology, University of Utah. Salt Lake City.

1962

Unusual Artifacts from Castle Valley, Utah. University of Utah Anthropological Papers No 60. Salt Lake City.

Harper, Kimball T. and G. M. Alder 1970 The Macroscopic Plant Remains of the Deposits of Hogup Cave, Utah and Their Paleochmatic Implications. In Hogup Cave, by C. Melvin Aikens. University of Utah Anthropological Papers, No. 93. Salt Lake City.

Karlen Variations â€” Their Pattern Quaternary Research, Vol. 3, New York.

Helm, Claudia J. 1973 Preliminary Report of an Archaeological Survey in Eastern Utah. Ms. on file, Department of Anthropology, University of Utah. Salt Lake City. Hester, T. R. 1970 Chronological Ordering of Great Basin Prehistory. Contributions of the University of California Archeological Research Facility, No. 27. Berkeley.

Vol. 222,

Durrant, Stephen D. 1952 Mammals of Utah. University of Kansas Publications, Museum of Natural History, No. 6. Lawrence.

Jennings, Jesse D. 1957 Danger Cave. University of Utah Anthropological Papers, No. 27. Salt Lake City.

Ford,W.E.,ed. 1953 Dana's Textbook of Mineralogy. (15th Printing, original 1898) Wiley and Sons. New York.

1964 The Desert West. In Prehistoric Man in the New World, edited by J. D. Jennings and E. Norbeck. University of Chicago Press. Chicago.

Fowler, D. D. and C. S. Fowler 1971 Anthropology of the Numa. (John Wesley Powell's Manuscripts on the Numic Peoples of Western North America, 1868-1880.) Smithsonian Institution Press. Washington.

Jennings, Jesse D., Alan R. Schroedl and Richard N. Homer n.d. Sudden Shelter. University of Utah Anthropological Papers (in preparation). Salt Lake City. Kelly, Isabel T. 1964 Southern Paiute Ethnography. University of Utah Anthropological Papers, No. 69, Glen Canyon Series, No. 21. Salt Lake City.

Fowler, D. D., D. B. Madsen, and E. M. Hattori 1973 Prehistory of Southeastern Nevada. Desert Research Institute Publications in the Social Sciences, No. 6. Reno.

Kelly, William H. n.d. Unpublished research notes on file at the Arizona State Museum Library, University of Arizona. Tucson.

Fry, Gary F. 1970 Prehistoric Human Ecology in Utah: Based on the Analysis of Coprolites. Ph.D. dissertation, Department of Anthropology, University of Utah. Salt Lake City. Gunnerson, James H. 1956 A Fluted Point Site in Utah. American quity, Vol. 21, No. 4. Salt Lake City.

1960

Hackman, Robert J. and Paul L. Williams 1972 Map Showing Drainage Basins and Historic Cloudburst Floods in the Salina Quadrangle, Utah. U.S. Geological Survey Folio of the Salina Quadrangle, Utah, Map 1-591-N. Washington.

Dalley, Gardiner F. 1970 Palynology of the Evans Mound Deposits. In the Evans Site, by Michael S. Berry. A Special Report, Department of Anthropology, University of Utah. Salt Lake City.

Denton, George H. and Stephen C. Porter 1970 Neoglaciation. Scientific American, No. 6, pp. 100-110. New York.

An Archeological Survey of the Fremont Area. University of Utah Anthropological Papers No 28. Salt Lake City.

Hackman, Robert J. 1973 Vegetation Map of the Salina Quadrangle, Utah. U.S. Geological Survey Folio of the Salina Quadrangle, Utah, Map 1-591-P. Washington.

Currey, Donald R. and David B. Madsen 1974 Holocene Fluctuations of Great Salt Lake. Abstracts of the Third Biennial Meeting, American Quaternary Association, p. 74. University of Wisconsin. Madison.

Denton, George H. and Wibjorn 1973 Holocene Climatic and Possible Cause. No. 2, pp. 155-205.

1957

Kelso, Gerald 1970 Hogup Cave, Utah: Comparative Pollen Analysis of Human Coprolites and Cave Fill. In Hogup Cave, by C. M. Aikens. University of Utah Anthropological Papers, No. 93. Salt Lake City.

Anti-

Kuchler, A. W. 1964 Potential Natural Vegetation of the Conterminous United States. American Geographical Society Special Publication, No. 36. New York.

Mehringer, Peter J., Jr. 1967 Pollen Analysis of the Tule Springs Site, Nevada. Nevada State Museum Anthropological Papers, No. 13, Pt. 3. Carson City.

Lindsay, La Mar W. 1974 Report of a Preliminary Archaeological Survey of Coal Lease Lands (U-073039, U-073040, and U-073041), Sevier and Emery Counties, Utah. A Special Report, Bureau of Land Management. Salt Lake City.

Morss, Noel 1931 The Ancient Culture of the Fremont River in Utah. Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard University, Vol. 12, No. 3. Cambridge.

1975 Palynological Analysis and Paleoecology of Innocents Ridge. In Innocents Ridge and the San Rafael Fremont, by Alan R. Schroedl and Patrick F. Hogan. Antiquities Section Selected Papers, Vol. I, No. 2. Salt Lake City.

Richmond, Gerald M. 1965 Glaciation of the Rocky Mountains. In The Quaternary of the United States, edited by H. E. Wright, Jr. and David G. Frey. Princeton University Press. Princeton. Schoenwetter, James 1974 Palynological Records of Joe's Valley Alcove: A Multicomponent Site in Southeast Utah. A Special Report, Department of Anthropology, Arizona State University. Tempe.

n.d.[a] Pollen Analysis of Clyde's Cavern. Research data on file, Antiquities Section, Division of State History. Salt Lake City. n.d.[b] Pollen Analysis of Sudden Shelter Site Deposits. In Sudden Shelter, by Jennings, Schroedl, and Homer. University of Utah Anthropological Papers (in preparation). Salt Lake City.

Schroedl, Alan R. and Patrick F. Hogan 1975 Innocents Ridge and the San Rafael Fremont. Antiquities Section Selected Papers, Vol. I. No. 2. Salt Lake City.

Lucius, William A. and Jeffrey Colville n.d. Osteological Analysis of Faunal Remains. In Sudden Shelter, by Jennings, Schroedl, and Homer. University of Utah Anthropological Papers (in preparation). Salt Lake City.

Spieker, Edmund M. 1949 The Transition Between the Colorado Plateau and the Great Basin in Central Utah. Guidebook to the Geology of Utah, No. 4. Salt Lake City.

Madsen, David B. 1975 Three Fremont Sites in Emery County, Utah. Antiquities Section Selected Papers, Vol. I, No. 1. Salt Lake City.

Steward, Julian H. 1938 Basin-Plateau Aboriginal Sociopolitical Groups. Smithsonian Institution Bureau of American Ethnology, Bulletin No. 120. Washington.

Madsen, David B. and Michael S. Berry 1975 A Reassessment of Northeastern Great Basin Prehistory. American Antiquity, Vol. 40, No. 4. Washington.

Stokes, William L. and Robert E. Cohenour 1956 Geologic Atlas of Utah (Emery County). Utah Geological and Mineralogical Survey Bulletin, No. 52. Salt Lake City.

Madsen, Rex 1972 Ceramics. In The Evans Site, by Michael S. Berry. A Special Report, Department of Anthropology, University of Utah. Salt Lake City. 1976

Fremont Ceramics. Museum of Northern zona Ceramic Series, in press. Flagstaff.

Struever, Stuart 1968 Flotation Techniques for the Recovery of Smallscale Archaeological Remains. American Antiquity, Vol. 33, No. 3. Washington.

Ari-

Taylor, Dee C. 1957 Two Fremont Sites and Their Position in Southwestern Prehistory. University of Utah Anthropological Papers, No. 29. Salt Lake City.

Martin, PaulS. 1963 The Last 10,000 Years: A Fossil Pollen Record of the American Southwest. University of Arizona Press. Tucson.

Tidestrom, Ivar 1925 Flora of Utah and Nevada. Contributions the United States National Herbarium, 25. Smithsonian Institution. Washington.

Martin, Paul S. and Floyd W. Sharrock 1964 Pollen Analysis of Prehistoric Human Feces: A New Approach to Ethnobotany. American Antiquity, Vol. 30, No. 2, Pt. 1. Salt Lake City.

from Vol.

U. S. Weather Bureau 1957 Climatic Summary of the United States, Supplement for 1931 through 1952, Utah. Climatography of the United States, No. 11-37. U.S. Government Printing Office. Washington.

Marwitt, John P. 1970 Median Village and Fremont Culture Regional Variation. University of Utah Anthropological Papers, No. 95. Salt Lake City.

1961- Climatic Summary of the United States, 1961 1970 through 1970. Climatography of the United States. U.S. Government Printing Office. Washington.

Wilson, Curtis J. and Howard L. Smith 1976 Interstate Highway 1-70 Salvage Archeology. Antiquities Section Selected Papers, Vol. II. No. 7. Salt Lake City.

1965 Climatic Summary of the United States, Supplement for 1951 through 1960, Utah. Climatography of the United States. No. 86-37. U.S. Government Printing Office. Washington.

Winter, Joseph C. and Henry G. Wylie 1974 Paleoecology and Diet at Clyde's Cavern. American Antiquity, Vol. 39, No. 2, Pt. 1. Washington.

Williams, Paul L. 1972 Map Showing Types of Bedrock and Surficial Deposits, Salina Quadrangle, Utah. U.S. Geological Survey Folio of the Salina Quadrangle, Utah, Map 1-591-H. Washington.

Wormington, H. Marie 1955 A Reappraisal of the Fremont Cultures with a Summary of the Archeology of the Northern Periphery. Proceedings of the Denver Museum of Natural History, No. 1. Denver.

Williams, Paul L. and Robert J. Hackman 1971 Geology, Structure, and Uranium Deposits of the Salina Quadrangle, Utah. U.S. Geological Survey Miscellaneous Geologic Investigations, Map 1-591 (Sheet 1 of 2). Washington.

Wylie, Henry G. 1971 Report of Excavations at Clyde's Cavern (42Em 177), Emery County, Utah. A Special Report, Department of Anthropology, University of Utah. Salt Lake City.

Wilson, Curtis J. 1974 Highway U-95 Archaeology: Comb Wash to Grand Flat. A Special Report, Department of Anthropology, University of Utah. Salt Lake City.

ARCHEOLOGICAL INVESTIGATIONS IN THE MAZE DISTRICT CANYONLANDS NATIONAL PARK, UTAH

Edited by WILLIAM A. LUCIUS Contributors: PATRICK HOGAN LEONARD LOSEE WILLIAM A. LUCIUS Department of Anthropology University of Utah

ANTIQUITIES SECTION SELECTED PAPERS Number 11

ARCHEOLOGICAL INVESTIGATIONS IN THE MAZE DISTRICT CANYONLANDS NATIONAL PARK, UTAH

Edited by WILLIAM A. LUCIUS Contributors: PATRICK HOGAN LEONARD LOSEE WILLIAM A. LUCIUS Department of Anthropology University of Utah

ANTIQUITIES SECTION SELECTED PAPERS Number 11

TABLE OF CONTENTS Acknowledgments Introduction Previous Investigations

Page 78 79 79

Setting Geology and Geomorphology Water and Climate Vegetation Fauna

82 82 82 83 83

Site Classification Habitation Sites Storage Sites Rockshelter Camps Open Camps Chipping Stations Rock Art Sites

84 84 84 86 87 87 88

The Sites In Their Physical Settings The River Bottoms The Benchlands The Canyons The Grasslands

89 89 90 91 91

Artifacts Ceramics Chipped Stone Ground Stone Miscellaneous Artifacts

"2 92 92 93 94

Cultural Affiliation and Dating

Discussion Settlement Patterns and Subsistence Regional Cultural Relationships

95 95 96

Summary

References

ILLUSTRATIONS Figure 1. Map showing area of the Maze Survey 2. Bird Site (Harvest Scene) rock art panel of the Barrier Canyon style; located in Maze Canyon 3. Anasazi masonry granary located on Turks Head Bottom of the Green River 4. Upright slab and mud granary of unknown cultural affiliation located in the Doll House area 5. Typical slab-lined firepit with ashy soil, possibly of Archaic or Basketmaker affiliation 6. Previously unrecorded Barrier Canyon style rock art panel located in Range Canyon 7. Two story tower and adjacent structures located at Fort Bottom on the Green River

Page 80

84 85

87 88

ACKNOWLEDGMENTS The authors would like to gratefully acknowledge the continual help, hospitality, and advice provided prior to and during the course of the Maze District survey by District Ranger James Walters and members of his staff. Field work was not only facilitated by their help but also made more enjoyable. Also deserving grateful mention for his support, advice and editorial assistance, is Dr. Jesse D. Jennings, Professor of Anthropology at the University of Utah and principal investigator in the Maze project.

INTRODUCTION In accordance with agreements between the National Park Service and the University of Utah, an archeological survey was undertaken in selected portions of the Maze District, Canyonlands National Park. The Maze District incorporates approximately 325 square miles of rugged plateau parkland ,and canyons to the west of the gorges of the Green and Colorado rivers between Lake Powell and Dead Horse Point (Fig. 1). The following report is an edited version of two separate reports submitted to the National Park Service in 1975. The initial survey was performed and reported by Patrick Hogan, Leonard Losee, and James Dodge. A second survey and report was accomplished by Leonard Losee and William Lucius with Kendall Malstrom aiding in the Green River portion of the survey. This paper contains portions of both Maze survey reports combined to document the previously unknown archeological resources of the Maze District of Canyonlands National Park. The survey was conducted on an interrupted basis in the spring and late summer of 1975 by two threemember survey teams from the Department of Anthropology, University of Utah. Access to sites adjacent to the river or in the canyons below the major jumps was gained by boat. The bulk of the survey in the upper canyon regions was accomplished by backpacking into the canyons from their heads.

It was deemed advisable for reasons of preservation as well as logistics to limit collections to environmental sampling leaving all cultural material in situ, it was felt that field recording and identification of observed materials would be sufficient for survey analysis and interpretation. The field investigators, where possible, assigned cultural affiliation to sites relying on their familiarity with diagnostic and distinctive artifacts of those cultures known to have used the area (Sharrock 1966; Gunnerson 1969). Survey records typically consisted of a written description and location, a measured map of the site, an area sketch of the immediate environment, notes on artifact categories present, photographs, and, where applicable, botanical and soil specimens. Prior to this survey there had been no professional archeological study of the Maze District, although the existence of some pictograph panels was known. There was, therefore, uncertainty as to what would be encountered. Pre-survey research of work done in adjacent areas and a familiarization flight over the area itself, led to the prediction that sites would be scattered and there would be few, if any, permanent agricultural villages. The goal of the study, as stipulated, was an intensive survey in order to learn the probable cultural affinity and use of the area.

PREVIOUS INVESTIGATIONS mings was able to establish that the area had been occupied by a Puebloan group and did recognize the general settlement pattern which would be found to characterize the entire area. Between 1928 and 1931 a major expedition under the auspices of the Peabody Museum at Harvard University undertook the exploration and description of the^Upper Colorado drainage encompassing most of eastern Utah. Of specific interest in this treatise is the work of Henry Roberts in 1929 and 1930 (Gunnerson 1969). Roberts conducted surveys and limited excavations in the Salt Creek drainage, Ruin Park, Fort and Woodruff Bottoms on the Green River, and Barrier Canyon. In general, the results of his work indicated that the

The exploration of the Maze District can be viewed as a continuation of University of Utah concern with the prehistory of the Canyonlands region of Southeastern Utah. As a guide to previous investigations of Canyonlands prehistory by the University, as well as others, an annotated history of archeological research follows. The first professional work on record was a series of expeditions undertaken by Dr. Byron Cummings (1910) of the University of Utah between 1907 and 1909. Cummings described the ruins of White and Montezuma canyons, Fable Valley and the Beef BasinRuin Park area and Alkali Ridge. As Brew (1946) noted: "his two short generalizing papers contain little of use to students of the region." Nevertheless, Cum79

Emery Co. Wayne Co.

yq . V /

.-O

GLEN CANYON NATIONAL RECREATION AREA 1

Wayne Co. Garfield Co

Fig. 1 Map showing area of the Maze Survey.

Grand Co. San Juan Co.

^ 7 }

Plateau area for the National Park Service at the request of the Bureau of Reclamation (Baldwin 1949). The party visited 72 sites and noted the presence of hundreds of others. Occupation by both Basketmaker and Pueblo I — Pueblo II Anasazi was documented by ceramics, rock art, and other cultural details. Finally, a detailed study following Brew's (1946) work at Alkali Ridge was suggested. The first Park Service interest in the archeology of what is now Canyonlands National Park occurred in 1952 when Bates Wilson and Alice Hunt examined Horse Canyon and its tributaries in the Needles District (Hunt and Wilson 1952). The party located 44 sites and reported numerous well-preserved ruins in Cottonwood, Lavender, and Davis canyons as well. In a later report, Alice Hunt (Hunt 1952) noted the beginnings of vandalism and recommended further more intensive study and protection. In 1952, Jack R. Rudy conducted a survey of Beef Basin as part of the Statewide Archaeological Survey of the University of Utah. His report (Rudy 1952) noted the beginnings of systematic looting which he attributed to increasingly easy access to the area brought about by the construction of a new road and an unimproved airstrip. As the chances of protecting the sites seemed slim, Rudy returned in 1953 to excavate nine ruins, including seven surface pueblos, under salvage conditions (Rudy 1955). All sites dated between Pueblo II and Pueblo III periods, with a single exception thought to be a watchtower, were apparently occupied year-round.

areas were occupied primarily during Pueblo II — Pueblo III times by small groups of Mesa Verde Anasazi. The monumental architecture of the Mesa Verde area was not in evidence although several multiroom masonry structures were noted. The most spectacular find of the survey was the Great Gallery rock art panel in Barrier Canyon. Since the colossal, extremely detailed painted figures had no positive affiliation with the Anasazi, Roberts excavated the only major habitation site in Barrier Canyon in hopes of gaining more evidence of those artists. His notes, as interpreted by Gunnerson, indicate three occupation phases by Fremont and Pueblo I — Pueblo II Anasazi. Gunnerson's own interpretation of the data suggested that based on differential pottery distributions between the first and second occupations, the original habitation may have been by an earlier pre-ceramic people (Gunnerson 1969).Whichever interpretation proves valid, the site is important as one of the earliest stratigraphic excavations done in the area, and remains the only published site in Canyonlands with any indication of significant time depth. Between 1931 and 1933, Harvard University sponsored a second study in the general area of Canyonlands on Alkali Ridge. The Peabody Museum Southeastern Utah Expedition, under the direction of John Otis Brew, differed from previous work in the area in that it was conceived in order to answer a specific archeological question. Brew began excavations on Alkali Ridge "with the aim to define the culture of Pueblo II or to eliminate it from the Pecos Classification" (Brew 1946). Thirteen sites, excavated during the three field seasons, convincingly demonstrated the reality of the Pueblo II period. Moreover, the work provided the only comprehensive reported archeological study in southeastern Utah until the late 1950's when the University of Ut'.h and the Museum of Northern Arizona began a joint study in the Glen Canyon area. The scope and intensity of those studies permitted the archeological reconstruction of the culture history of the area and details of the lifeway of the aboriginal inhabitants of the Canyonlands region. Between 1945 and 1947, the Carnegie Museum of Pittsburgh sponsored three expeditions into the White Canyon, Fable Valley area. The expedition was variously under the field direction of J. LeRoy Kay, a geologist and paleontologist, and David W. Rial of the Carnegie Museum staff. The party made extensive surface collections and records for the Museum, but they were never analyzed nor was a report published. As part of the Glen Canyon project, Floyd Sharrock and Edward Keane reexamined collections and records from the Carnegie Museum and published descriptions of artifacts noting their provenience when known and redesignating the sites reported with the trinomial numbering system in use by the University of Utah (Sharrock and Keane 1962).

In August of 1955, James Gunnerson, while conducting surveys for the University of Utah, examined eight open camps and chipping stations on Bureau of Land Management land near Ekker's Cabin and French Springs in and near the present Maze District. (Not reported; data derived from University of Utah site records.) In 1958, Gunnerson conducted a survey of a proposed state park in the Dead Horse Point-Junction Butte area (Gunnerson 1958). A portion of that land is now the Island-in-the-Sky District. Sixteeen ruins were reported including surface masonry structures, open sites, rockshelters, and rock art sites. No work other than that of a protective nature was recommended. In 1959, Lloyd Pierson (Pierson 1959) of Arches National Monument, visited Beef Basin, Ruin Park, and the Needles to reconnoiter the potential of the area's cultural resources for recreational activities. The report, which concentrated on the accessibility, picturesque nature, and other generally aesthetic features of the ruins, stood in marked contrast to traditional site descriptions. In October 1963, C. Melvin Aikens, then with the University of Utah, visited the Needles District and in a letter report to Bates Wilson suggested a systematic mapping and collection of sites be begun by the rangers to offset increasing vandalism. For that same reason, some salvage excavations were suggested.

In October 1945, Gordon C. Baldwin conducted a brief survey of the Beef Basin, Fable Valley, and Dark ;i

In 1965, Floyd Sharrock conducted an extensive survey of the Needles and Island-in-the-Sky Districts under contract to the Park Service. His report (Sharrock 1966), in addition to providing a site inventory, made some preliminary statements of cultural affiliation and culture history, postulated migration routes into the Park and made some casual comments about settlement and subsistence patterns. John P. Marwitt and Henry J. Hall conducted rightof-way surveys for the proposed road from Squaw Flat to the Confluence Overlook in 1970 (Marwitt 1970). They located 15 sites, all chipping scatters. The report marked the first of several surveys initiated expressly to assess the impact of area development on the affected cultural resources. In 1973, a survey of road realignments in the Needles and Grandview Point areas by La Mar Lindsay and Rex Madsen noted lithic scatters along each route and recom-

mended testing prior to construction (Lindsay and Madsen 1973). Marvin Kay, a Park Service archeologist, was also involved in 1973 road surveys in the Maze District (Kay 1973). Kay located several special use camps, and reexamined sites in Horseshoe Canyon. Recommendations were made for surface collection and mapping of the sites as well as for the protection of the sites from damage due to increases in visitation brought about by the construction of the new roads. In 1974, rangers Brian McHugh and Richard Simmons conducted a two-month survey of the Needles District (not formally reported) to complete the inventory records of the area. In addition to the basic locational and descriptive documentation, those features of the sites that specifically related to the administration of cultural resources, such as the state of preservation and evidence of vandalism, were noted.

SETTING The stark grandeur of the Canyonlands Park area has prompted many popular and scholarly articles on the general topography of the region. Little, however, has been written beyond scenic description of the Maze District, Canyonland's most recent addition. Therefore, a short description of the structurally complex region is provided to establish the background for a discussion of aboriginal settlement and economic patterns.

Member of the Chinle Formation â€” rise approximately 300 ft. to a broad, open flatland. Those formations also form the sloping bases for remnants of the cliffforming Wingate in the many isolated buttes that dot the lower bench. Outcrops of the Lower Glen Canyon Group occur in canyons such as Horseshoe and Millard which cut the district's uplands. The domelike weathering of the Navajo Sandstone is commonly the rimrock present there. Sharrock (1966) has correctly pointed out the importance of the Canyonlands terrain in terms of the restrictions it places on "migration routes, population (site) density, and site type and locale . . ." The information from this survey tends to support those conclusions for the canyons and headlands of the Maze District.

GEOLOGY AND GEOMORPHOLOGY Physiographically, the Maze District falls within the Canyonlands Section of the Colorado Plateau Province (Hunt 1967). As the name implies, this region is characterized by numerous deep canyons dissecting the plateau surface. Principal among the canyons are the gorges of the Green and Colorado rivers which form the eastern boundary of the Maze District. Along most of the eastern boundary the Pennsylvanian marine sediments of the Hermosa and Rico Formations are exposed and form the cliffs of the river gorges. Rimming the formations and forming a bench about seven miles wide is the Cedar Mesa member of the Cutler Formation. "This platform is dissected by numerous canyons tributary to the Green and Colorado Rivers, and its surface is marked by buttes, rounded domes, and several tall slender spires . . ." (Baker 1946). That is the area of the Maze itself, the Doll House, and the Land of the Standing Rocks. Capping many of the 100 to 400 ft. red cliffs are the white sandstones of the Organ Rock Tongue and White Rim members of the Cutler Formation. From the inner fringes of the benchlands of the Cutler Formation, an irregular line of cliffs â€” formed by the Triassic Moenkopi Formation and Moss Back

WATER AND CLIMATE Apart from the Colorado and Green rivers, no perennial streams flow through the Maze District. Total annual precipitation is less than 10 inches, the majority falling during the winter and late summer. Because of a high percentage of exposed rock and poorly developed soil much of the rainwater and snow-melt runs off into the Colorado or Green River thereby providing little usable moisture. The low humidity and high temperatures produce ideal conditions for extreme evaporation and further aid in dessication of the area. Only the water absorbed by the permeable sandstones of the Moenkopi, Kayenta. Navajo and Cutler formations (Baker 1946) remains, discharging slowly from springs and seeps to provide a year-round usable water source. The springs, therefore, become a critical factor in man's presence within the Maze. Within the survey area, run-off is usually depleted 82

by late May and the bountiful flows of the Green and Colorado are usable only within their respective gorges and the mouths of the tributary canyons. Jumps (dry waterfalls) up to 300 ft. high block access from the canyons themselves. In late July and August torrential thunderstorms flashflood the canyons and provide usable water only where potholes trap some of the flow.

VEGETATION The nature of the vegetation within the Maze District is sensitive to the limited availability of water. Following Woodbury (1959), the flora of the Maze District can be divided into several general types based primarily on moisture requirements.

clay to rock talus. Generally dominant is the shadscale with joint-fir {Ephedra sp.) and grasses subdominant. Narrow leaf yucca and Utah juniper {Juniperus osteospermd) also occur, the latter mostly on the canyon bottoms and on wide shelves within the canyons. In those areas having shallow soil, the shrubs are dominant; where deeper alluvium is present there is a higher proportion of grasses. Of the many grass species present in the Maze, several, notably Indian rice grass {Oryzopsis hymenoides), produce seeds known to be attractive to Pueblo gathering groups (Jennings 1966). It is ricegrass that Schaafsma (1971) believes to be depicted at the Bird Site (better known as the Harvest Scene) in South Fork Canyon, along with two figures thought to be harvesters (Fig. 2). Certainly the presence of these grasses in the area could be a major reason for the occupation of the Maze District by aboriginal groups.

STREAMSIDE OR RIVER BANK VEGETATION The occurrence of this community within the Maze District is confined to the banks of the Colorado and Green rivers. Here the water table remains high and the roots of the plants are often underwater. The dominant shrub here is tamarisk {Tamarix pentandra), an exotic that has rapidly replaced the sandbar willow {Salix exigua) which now occurs only in isolated stands. The use of the willow by the aboriginal inhabitants of the river canyons is indicated by its use as framing for doors and hatches of granaries. Streamside vegetation is largely replicated around seeps of the larger canyons, and includes shadscale {Atriplex canescens), rabbitbrush {Chrysothamnus viscidiflorus), reed cane {Phragmities sp.), Fremont Cottonwood {Populus fremontii), and various grasses and herbs.

BLACKBRUSH SHRUB STEPPE This community has been described by Walter Loope (personal communication) for the benchlands, such as the Doll House and Maze Overlook, and the upland flats, such as Hans Flat and Flint Flat. As with the hillside flora, the blackbrush steppe vegetation is dependent solely upon precipitation for its moisture. As implied, blackbrush {Coleogyne ramosissima) is the dominant shrub with two species of joint-fir and shadscale as subdominants.

UPLANDS Present at higher elevations, such as the benches of Millard Canyon and Hans and Flint flats, are forests of pinyon {Pinus edulis) and juniper {Juniperus osteosperma). Extensive grasslands also occur at these higher elevations in areas with deep, eolian deposits.

TERRACE VEGETATION This group occurs on river terraces such as the Spanish Bottom and at the mouths of the benchland canyons where they enter the Green River. These plants are rarely flooded, but have a good moisture supply from river water carried through the soil by capillary action. Dominant species include squawbush {Rhus trilobata), greasewood {Sarcobatus vermiculatus), saltbush {Atriplex confertifolia) and arrowweed {Pluchea serices). Also occurring are Fremont cottonwood, cane reed, hedgehog cactus {Echinocereus aggregatus), narrow-leafed yucca {Yucca angustissima) and various bunchgrasses.

FAUNA The animal life of the Maze District as it relates to man appears, from informal observation, to include relatively few species. While various species of reptiles and small rodents are the most numerous animals of the area and were no doubt utilized for food by aboriginal groups, rabbits and large artiodactyls provide much more meat per individual and so were probably more significant to the effective environment of the prehistoric inhabitants. Mule deer {Odocoileus hemionus) are today the most common of the big game animals, followed in number by the pronghorn {Antilocapra americana), and the desert bighorn sheep {Ovis canadensis). The relative proportions of these species have apparently changed since aboriginal times, as archeological sites throughout the west show an abundance of mountain sheep bone, outnumbering deer by a ration of about 7 to 1 (Jennings 1966). Sheep, of course, may have been the preferred aboriginal prey.

HILLSIDE VEGETATION The plants of the hillside vegetation group are dependent directly on local precipitation to meet their moisture requirements. The group occurs, within the survey area, on the canyon floors and on the talus slopes. Those areas are dry most of the time and plant cover is sparse, often less than 20%. The communities of the group vary as local soil conditions range from sand dune to 83

, J

- -."-»{j''"^..

R i§: ?ijf S , • - 83

^SL

Fig. 2

Bird Site (Harvest Scene) rock art panel of the Barrier Canyon style; located in Maze Canyon.

Similarities in the amount of meat and in the nonfood raw materials provided by any of these three species makes the identification of the primary game animal

of the Maze area an unnecessarily fine distinction for the broad generalities of this report,

SITE CLASSIFICATION In the course of the survey within the Maze District, 248 sites were located and recorded. Following Sharrock (1966) they were grouped into habitation sites, storage structures, rockshelters, open camps, chipping stations, and pictograph/ petroglyph panels. In cases where a site fit into more than one category a hierarchy was established with occupation superceding special use. For example, a rockshelter with a chippage scatter and a rock art panel would be considered primarily as a habitation site with the other aspects of the site considered secondary. The divisions were made on the following criteria:

The canyon bottom has thick alluvial deposits and the presence of two granary sites nearby suggests that cultivation of crops was practical. The intensity of occupation at these sites is difficult to judge. The shallow deposits and paucity of artifacts suggest a short span of use. However, the sites are readily visible from the river and probably have been looted repeatedly. Two additional habitation sites located in the uplands were both near the Doll House, one of the few spots in the southern portion of the Maze District where access to the river was possible. Those two sites are, however, of uncertain cultural affiliation. They will be discussed further in the section on storage sites in the context of the unusual granaries they contain.

HABITATION SITES No sites were found in the Maze area that match Sharrock's characteristics of a habitation site as defined for the Salt Creek drainage: ". . . one or more structures obviously used for habitation, kivas, mealing bins, openair use areas, extensive pictograph/petroglyph panels, and artifacts for domestic use . . ." (1966). However, 11 sites on the Green River showed evidence of more than short-term occupation and were therefore placed in this separate category. Those eleven sites were of Anasazi cultural affiliation and reflect the essentially sedentary nature of that culture. One, a single tower structure, was essentially circular with walls of wet-laid, unshaped sandstone slabs standing on an exposed, relatively inaccessible ridge jutting from the mouth of Jasper Canyon.

STORAGE SITES Thirty-two sites with storage facilities were recorded during the survey. Of those, 21 were primary storage sites, while 11 had storage facilities in association with habitational remains. As expected, the large majority of the storage sites were along the rivers where good farmland was available. Storage structures noted during the survey were of four distinct types: typical Anasazi masonry granaries, bell-shaped capped pits, upright stone slab and mud granaries, and slab-lined floor cists. A majority of the 84

storage sites were of the first type, consisting of one or more masonry granaries (Fig. 3). All such sites occurred at river bottom locations. Generally the granaries consisted of semicircular, wet-laid, single-coursed walls, built against the back wall of a protective overhang or under a large boulder. Unworked sandstone blocks were used in the construction and workmanship overall was, with a few exceptions, rather crude. River mud was employed as mortar, occasionally strengthened with juniper bark or other vegetal fibers. Mortar often made up a considerable portion of the total wall volume. Commonly the roofs of the overhangs were incorporated into the structures but others had apparently been roofed with poles and mud. Doorways were small and designed to be sealed with a sandstone slab. A late Pueblo 11/ early Pueblo III Anasazi cultural affiliation was assigned to most riverine granary sites on the evidence of pottery types. A second type of storage unit found at one site was a bell-shaped pit lined with adobe plaster. Willow was used to frame a hatchway in the top of the dome and a square sandstone "manhole cover" was used to close the opening. A structure similar in design was described by Morss (1931) as occurring in Coleman Canyon in the Fremont River drainage. A third type of storage structure was found at five sites, four of them in the Doll House area, and the fifth on the river (Fig. 4). These structures were formed of large, upright sandstone slabs held in place and usually completely plastered with fiber-strengthened mud. Whole granaries of this type showed that poles were employed for reinforcement and roofing.

Cultural affiliation of these sites is uncertain. Similar structures were noted by Morss in the Fremont River drainage near Fruita, suggesting a Fremont origin (Morss 1931). However, no Fremont artifacts were found at the Doll House sites. A black-on-white Anasazi sherd was found downslope from one of the granaries but not in definite association with it. One habitation site had the basal remains of what appears to have been a granary of this type. Artifacts and rock art at that site suggested an Archaic or early Basketmaker affiliation. Also problematic is the location of these sites in an area apparently unsuitable for agriculture. Three possible explanations can be offered. First, the area may have had higher summer precipitation and therefore been more suitable for agriculture at the time of aboriginal occupation (Baerreis and Bryson 1965). Second, the structures may have been used for storage of pine nuts or wild grass seeds. Third, corn and other cultivated plant foods may have been carried up from Spanish Bottom for storage at the upland sites. It may be significant in this regard that all of the upland granary sites were located near the Doll House where access to Spanish Bottom is possible. All of the structures had been invaded by rodents and none of the original stored foods remained to help answer the question. A fourth type of storage facility was found at two sites, one in Range Canyon and the other near the Doll House. This type consists of slab-lined cists in the floors of two rockshelters. The cists averaged 50 cm. in diameter and about the same depth. Identical cists were found in a late Archaic/early Basketmaker con-

Fig. 3 Anasazi masonry granary located on Turks Head Bottom of the Green River.

text in nearby Cowboy Cave (Jennings 1975). The Range Canyon cist site produced Fremont sherds, but this could be due to re-use of the rockshelter. It is unknown what was stored in the cists; of the three cists at the two sites, two had been cleaned out completely while the third remained sealed with poles and mud and was left intact by the surveyors.

sion flakes (several showing use for cutting and scraping) jasper or quartzite hammerstones, jasper cores, blanks and preforms, broken knives and scrapers, occasionally one-hand manos and/or flat milling stones, and a rare projectile point fragment or a few sherds of pottery. One to three task-specific tool kits are thus in evidence. The hammerstones, cores, unfinished tools, and detritus indicate tool production. Judging from the apparent predominance of secondary flakes and two instances of caches of blanks, it appears likely that blanks were quarried elsewhere and brought to the camps for further reduction and/or finishing. The knives, scrapers, and projectile points as well as the utilized flakes imply the hunting and butchering of game and the preparation of hides. These activities were less in evidence than tool manufacture as far as total artifact numbers go. However, in view of the ease of recognition and desirability of the hunting tool kit as compared with the relatively nondescript nature of evidence of manufacture of tools, it seems feasible that hunting was a generally practiced activity which, for the most part, has escaped detection. This hypothesis can be supported only by systematic testing of sites, but evidence from undisturbed sites in Horse Canyon seems to bear out the suggestion. The milling of seed or grain was evidenced by the presence of ground stone. Rock art was found in or adjacent to twelve rockshelter camps. Styles present in those sites included Anasazi, Fremont, and Barrier Canyon. Cultural affiliation could be assigned to thirtyone rockshelter camps on the basis of pottery remains.

ROCKSHELTER CAMPS Overhangs and rockshelters showing evidence of brief or occasional occupation were classified as rock shelter camps. Seventy-three of this type were recorded during the survey. Typically, these sites had one or more fire areas and debitage indicating stone tool manufacture. Chipped and ground stone tools, most frequently shallow basin metates and small one-hand manos, were also common at these sites. Low, dry-laid stone structures were often found in sheltered camps with walls of two types: one was a complete enclosure of low walls (about 0.5 m.); the second had only two low walls with the front left open. Both types enclosed an average area of 9-12 sq. m. and were devoid of interior features. Judging from the artifact scatter, the structures were sleeping rooms while the open areas of the rockshelter served as kitchen and work room. Few finished tools were found in the alcove sites and those present were generally broken or worn out. A common assemblage included a scatter of percus-

Fig. 4

Upright slab and mud granary of unknown cultural affiliation located in the Doll H

ouse area.

Of those, 24 were of Anasazi affiliation and six were Fremont. Two sites, one at Fort Bottom and one in Horse Canyon, had both Anasazi and Fremont pottery. Rockshelter sites with pottery were found throughout the uplands as well as along the river. The majority of the rockshelter camps contained no pottery. On the basis of projectile point and ground stone types, as well as the lack of pottery, many of those sites were tentatively assigned to a pre-agricultural period.

••3?

«frM -, StiSL.X-

. 2,1, . - - • • • -1

S L'.-P- \ . j J B

. H^;

OPEN CAMPS Fig. 5 Typical slab-lined firepit with ashy soil, possibly of Archaic or Basketmaker affiliation.

Forty-one open sites showing evidence of brief occupation — possibly a few days or less — were classified as open camps. Often they were located in the lee of sand dunes with the cultural deposits in a blow-out. For this reason, the true nature of these sites is elusive because any structures or other features may be covered, or the depth of the deposit misjudged. The number of these sites is almost meaningless as their presence or absence on the surface changes from day to day with the shifting of the sands. Many of the recorded camp sites appeared to have been buried for an undetermined period before being recently exposed by erosion. A typical open camp consisted of a fire area surrounded by a sparse artifact scatter, usually simple chipping debris. A few open sites without evidence of fires were classified as open camps on the basis of the presence of pottery and/or ground stone in addition to the chipped stone characteristic of chipping stations. Erosion may easily account for the lack of camp fire evidence in such cases. Twenty-seven of the open camp sites, all in the upland areas, had one or more slab-lined fire basins. These basins, formed by a ring of upright sandstone slabs, ranged from .75 to 1.5 m. in diameter and contained ashy soil (Fig. 5). The large size of these basins and the frequent occurrence of two or more together, suggest that they served a special function, possibly for roasting some undetermined plant food. Open sites with slab-lined basins consistently lacked pottery. That fact plus the presence of Archaic type ground stone artifacts at a few of those sites, suggests that they date from the Archaic or early Basketmaker period. Kowta (1963) recorded an open camp with a large slab-lined basin in Capitol Reef National Park. That site also yielded only stone artifacts. Systematic excavation of a few of these sites might securely answer the questions of cultural affiliation and function. Nine slab-lined basin sites lacked any surface artifacts. They were assumed to be prehistoric on the basis of similarity to other such sites with artifacts, and the lack of associated historic remains. Other open camp sites were assigned to the Anasazi and Fremont cultures on the basis of pottery remains. These generally occurred in areas lacking suitable rockshelters.

CHIPPING STATIONS Throughout the area covered by the Maze survey, chert may be found weathering in large nodules from veins in the limestones of the Cutler Formation. The cryptocrystalline silicates vary in quality from a flawed, but serviceable brown chert to a bright red, high quality material. The abundance of cryptocrystalline material precludes the necessity for any concentrated quarrying activity. Rather, the procurement of lithic raw materials may be seen as an ever present cultural feature of the canyon region. Seventy-six sites were classified as chipping stations on the basis of the presence of finished stone tools, cores, blanks, hammerstones, and waste flakes, and a lack of evidence for activity other than tool manufacture. Many of these sites contained only waste flakes, but this may be due in part to the activities of souvenir collectors. Judging from the materials noted at the sites, the siliceous stone was reduced to a size practical for transport by the removal of cortex materials and transported from the site as irregular cores or blanks. The presence of caches of blanks at two alcove sites argues that in many cases, finishing phases of manufacture were done somewhere other than the collecting locality. Other small and diffuse chipping scatters were noted in large numbers, but were not specifically recorded. Commonly, these were comprised of small secondary flakes as might come from percussion retouch or finishing. Since these sites occurred most often near modern game trails at strategic points in the canyons it is felt that they are the work of hunting parties awaiting game or reworking their tools at a kill. Thirteen sites fall into Kay's proposed Quarry/ Primary Lithic Reduction Station category while the small scatters and the camps are classified as Secondary Manufacturing Stations (Kay 1973). However, the distinctions, at least in the Maze, are not fine. 87

-W;

Fig. 6

Previously unrecorded Barrier Canyon style rock art panel located in Range Canyon.

Such grouping describes only the modal activity and it must be emphasized that all construction phases took place to a greater or lesser degree at all sites with a lithic component. Diagnostic projectile points were rare at the chipping sites though the presence of Elko series points at a few suggests a possible Archaic/Basketmaker affiliation. Undoubtedly, some of the chipping stations were produced by the later Fremont and Anasazi inhabitants of the area.

ROCK ART SITES Rock art was recorded at 42 sites. Twenty-one of those were primary rock art sites while 21 were habitation, camp, and storage sites .with rock art present. The sites are evenly divided between petroglyphs (pecked) and pictographs (painted), but when separated on the basis of location — river vs. upland — a different picture emerges. Petroglyph sites outnumber pictograph sites along the Green River while the reverse is true in the uplands, where pictographs outnumber petroglyphs. Following Schaafsma's (1971) stylistic designations, three major styles were evident in the rock art sites — Anasazi, Fremont, and Barrier Canyon. The Anasazi style was present at ten sites, nine of them along the river and one in the uplands. Painted handprints, both negative and positive, similar to those reported by

Sharrock (1966) in the Needles District, were present at two Anasazi storage sites on Valentine Bottom and at an alcove site with storage structures in lower Horse Canyon. Pecked Anasazi motifs present at other sites include: burden-bearing human figures, a row of handholding dancing figures, animal tracks, deer, bighorn sheep, snakes, bullseyes, and a variety of geometric elements. Rock art of typical Fremont style was present at 11 sites, six of them along the river and five in the uplands. With one exception, the panels consisted of pecked or incised figures. Common elements at these sites were trapezoidal and triangular anthropomorphs (frequently horned) bighorn sheep, deer, and various abstract elements. One site, on a high ledge at Valentine Bottom, had a large panel of painted Fremont style anthropomorphs adjacent to a row of Anasazi granaries. Another site,, at Anderson Bottom, though apparently the work of a single artist, combined Fremont style elements — armless trapezoidal anthropomorphs and a well-executed shield-bearing figure with typical Anasazi style elements — animal tracks, and burden-bearing human figures. These two sites raise again the question of the authorship of Fremont rock art in the Canyonlands as discussed by Sharrock (1966) who recorded numerous Fremont style pictograph panels associated with Anasazi cultural remains in the Needles District. The Anderson Bottom panel may be evidence for the borrowing of Fremont rock art

panel in Horseshoe Canyon and the Bird Site (Harvest Scene) in the middle fork of the Maze Canyon (Fig. 2). Six of the 13 Barrier Canyon style panels occurred at camp and habitation sites. In all such cases pottery was lacking. Chipped stone and grinding implements, where present, suggesting a pre-agricultural origin for the sites. This would support Schaafsma's (1971) contention that the Barrier Canyon Style is the product of a pre-agricultural group, rather than of the Fremont as had previously been supposed. However, rock art/ artifact associations are always somewhat tenuous. Two panels in South Fork, although of uncertain stylistic affiliation, appear relatively recent and may be of Ute or Paiute origin. An interesting panel of boxybodied zoomorphs in Jasper Canyon are similar to some assigned by Turner (1963) to Glen Canyon style 5 and may be Archaic in origin. Seven more sites encountered during the survey are of less certain origin but could be assigned to either the Fremont and/or Barrier Canyon style.

motifs by Anasazi peoples as suggested by Sharrock. Unlike the Needles District situation, however, sherds and other evidence of both Fremont and Anasazi occupation were found on the Maze District side of the river. The canyonlands area was obviously a zone of Fremont/Anasazi interaction, but the nature and consequences of the contact remain unclear. The Barrier Canyon style was present at 13 upland pictograph sites. Motifs at these sites included anthropomorphs, zoomorphs — frequently at the sides of the anthropomorphs, plants, and raining clouds. Colors noted in the sites were red, purple, orange, and gray, with red by far the most common. The pigments appeared to consist of mineral materials — largely hematite — available in the area. One particularly noteworthy panel of this style was recorded in Range Canyon (Fig. 6). The panel consisted of lifesize human figures with varied torso decoration and halo-like headdresses. At the feet of eight of the figures were small upright animals, while a rainbow arched over others. The panel is reminiscent of the "Great Gallery"

THE SITES IN THEIR PHYSICAL SETTINGS and Colorado rivers (Woodbury 1959), but whether they were exploited by the prehistoric inhabitants of the area is unknown. Undoubtedly the major attraction of the river bottoms was the rich soil where such crops as corn, beans, and squash could be grown. Of the 55 sites recorded along the river, 45 could be securely attributed to Anasazi horticulturalists. Over half — 28 — were storage sites or habitation and rockshelter sites with storage facilities. Although a dozen named and unnamed bottoms were examined, the great majority of the river sites were located at four bottoms where conditions were particularly well suited for agriculture. Those were: Valentine Bottom (21 sites), Anderson Bottom (8 sites), Fort Bottom (5 sites), and Turks Head (4 sites). No surface evidence — field boundaries, irrigation ditches, etc. — was visible to indicate the location and extent of prehistoric fields. Flooding and, in at least two cases (Anderson and Fort Bottoms), historic farming have probably destroyed such features. Presumably, however, corn and squash and perhaps beans (corncobs and a squash button were noted) were planted near the river where groundwater level was high, and along stream channels emerging from side canyons where summer flash flood waters could be easily diverted. Though higher terraces were also examined, most of the river bottoms sites were found at the base of the first terrace above the floodplain in rockshelters and under large boulders. A notable exception to this

Due to the diversity of terrain found in the Maze District, it was deemed useful to discuss individually the various environments and the types of sites found in them.

THE RIVER BOTTOMS Along the stretch of river that forms the eastern boundary of the Maze District are more than a dozen bottoms and canyon mouth areas with land suitable for agriculture. Averaging about 3,900 ft. in elevation, these bottomlands are characterized by level, wellwatered areas along the river, surrounded by a series of increasingly higher and drier terraces backed by high cliffs. Three formations make up the bottomlands and terraces in the area surveyed; from north to south these are the Chinle, the Moenkopi, and the Cutler. The Chinle and Moenkopi are, for the most part, unsuitable for formation of rockshelters and it was in the Cutler Formation, particularly at the base of the White Rim Member, where the majority of the riverside sites were found. Climate along the river, as well as elsewhere in the district, is arid. Annual rainfall is under ten inches and it is only the presence of the river that allows for the comparatively rich habitat of the bottomlands. Fish, including channel catfish {Ictalurus spp.), squawfish {Ptychocheilus lucius), suckers {Castomus spp.), and chubs {Gita spp.) are common in the Green 89

Fig. 7 Two story tower and adjacent structures located at Fort Bottom on the Green River

rule was found at Fort Bottom where a group of structures including a circular two-story "tower" were perched atop a high butte (Fig. 7). The site commanded a view of a considerable area up and down the river and was accessible at only one spot. It appeared to be a defensible retreat or observation post, but may also have served some unknown ritual function. Pottery found at the river sites indicated that the great bulk of the occupation is traceable — as in the Needles District — to the Pueblo 11/early Pueblo III Anasazi expansion that occurred at about 1075 to 1150 A.D. (Sharrock 1966). The lack of large habitation sites and midden material also indicates that the majority of the occupation occurred within a short timespan — perhaps 50 years — and involved small groups, possibly inhabiting the sites only seasonally to plant and harvest their crops.

entirely upon precipitation and is available only in potholes in the exposed upper surface of the White Rim Sandstone, and then only for brief periods in the spring and following summer thunderstorms. Seeps and springs occur in the nearby canyons, but canyon walls are so abrupt that access is possible at only a few places. The benchlands are characterized by low semistabilized dunes and extensive areas of exposed sandstone. Vegetation is that described by Loope (personal communication) as "blackbrush steppe." As might be expected, sites were sparse in the benchland areas and of a very transient nature. Of 25 recorded, 20 were only chipping scatters while the remaining five were small open camps. The sites were concentrated along canyon rims where the uneven sandstone surface traps rainwater in shallow potholes. With the exception of Petes Mesa, the benchlands probably served more as access routes between more favorable areas than as resource centers. The north rim of Horse Canyon is particularly well suited as an access route, connecting the Green River with the pinyon-juniper covered uplands to the west. The few diagnostic artifacts — pottery was lacking — in the benchland sites suggested that an Archaic/ Basketmaker affiliation was likely for a large percentage of them.

THE BENCHLANDS Level to slightly sloping benchlands, averaging about 5,000 ft. in elevation, surround Horse Canyon and comprise Petes Mesa and the uplands west of the Standing Rocks around the heads of the Maze, Fins, and Range canyons. The environment in these areas is the most marginal in the district. Water is dependent 90

THE GRASSLANDS

THE CANYONS Canyons are the dominant landform in the Maze District, comprising about 75% of the total area. Although varying considerably in width and resulting environment, the canyons of the district are all characterized by abrupt, at times even slightly overhanging walls, numerous side canyons, sandy floors and, in most cases, a series of stepped benches. Canyon floors average about 4,800 ft. in elevation. Stream channels fill briefly with raging torrents following summer thunderstorms, but the only permanent water sources in the canyons are the infrequent springs and seeps. All else being basically equal, canyon width is the single most important factor in determining water availability and resulting vegetation types. Narrow canyon areas not only concentrate precipitation run-off more, but also retard evaporation by providing more shade than wide canyons. With the exception of Range Canyon, the canyons of the district contained relatively few sites. Those found, however, showed considerable variety, both in type and cultural affiliation. Rockshelter camps, many of them showing evidence of re-use, predominated. Chipping stations, open camps, and rock art in a variety of styles including Anasazi, Fremont, and Barrier Canyon were also documented. Habitation sites, as earlier defined, were lacking in the canyons and occupation throughout appeared to have been of a transient nature reflecting resource exploitation by various groups. Acquisition of lithic material was probably the major activity in South Fork, Jasper, Shot, and Water canyons where good quality jasper is found weathering out of extensive beds in the Cutler Formation. The presence of Anasazi camps on the ridge extending north from Chimney rock to Petes Mesa suggests an access route from the uplands into the canyons. Range Canyon had a relatively large number of sites (33) and two reasons can be suggested. First, lower Range Canyon has, as noted, a considerable area of grasslands offering Indian ricegrass and dropseed as well as the typical canyon plant resources and game; second, lying between the Wide ValleyErnies Country area and the pinyon covered uplands to the west, Range Canyon was probably on a major passageway. Although artifacts in the canyon sites indicated occupation by Anasazi and Fremont groups as well as Archaic/Basketmaker, no evidence of agriculture was noted anywhere in the canyons.

Deep, sandy soils occur at North Flat between Standing Rocks and the Dollhouse, Wide Valley, Main Flat, and portions of Ernies Country. These areas vary from level to gently rolling and average about 5,200 ft. in elevation. With the exception of a very few occasionally filled potholes, water is lacking and vegetation there is entirely dependent on the scanty precipitation. Nevertheless, these areas support a plant community that probably constituted one of the major attractions of the Maze District for prehistoric peoples of the area. The community consists largely of several species of sand-loving grasses, including dropseed {Sporobulus spp.), Indian ricegrass {Oryzopsis hymenoides), needlegrass {Stipa spp.) and grama grass {Bouteloua curtipendula). Also occurring are shadscale, snakeweed, and prickly pear, as well as an occasional juniper and barberry shrub. That the grasses were exploited prehistorically seems amply documented by the common occurrence of grinding stones in sites of the district. Sheep and cattle grazing was practiced until recently in the grassland areas but never on a large scale because of the scarcity of water. Apparently, little if any alteration of the local plant communities has resulted from those activities (Loope 1975, personal communication). Fauna of the grasslands include deer, antelope, rabbits, and rodents, but again, numbers are not high. Hunting was probably of secondary importance to plant food collection. Sites are relatively numerous in the grasslands, particularly in the area northeast of the Doll House. As would be expected, open sites including camps and chipping stations are most common. In the low cliffs surrounding much of the grasslands are many rockshelter camps and rock art panels. The latter are, with few exceptions, of the Barrier Canyon style. Storage structures occur at a few sites near the Doll House, including two habitation sites. As noted earlier, however, it seems unlikely that horticulture was practiced in the uplands. Artifactual evidence for both Fremont and Anasazi occupation of the grasslands was recorded, suggesting that wild plant food collection and hunting were important supplements to riverine agriculture for those groups. The majority of the grassland sites, however, were apparently of Archaic/Basketmaker affiliation and seasonal exploitation of wild grass seeds must have formed an integral part of the yearly hunting-collecting round. 91

In addition to their importance prehistorically as resource areas, the grasslands of the Maze District, by their very nature, must have served as important access routes between the river/canyon area and up-

land areas to the north and west. It seems likely that this fact accounts in part for the number as well as cultural variety of sites in the grasslands.

ARTIFACTS As mentioned, no collections of cultural materials were made during the course of the Maze area survey. It was deemed advisable that, since field observations could provide the basic information about cultural affinity and site utilization it would be of more value to leave the sites as untouched as possible for future intensive investigations. This approach was possible because of the relative remoteness of the survey area and lack of development of the Maze District. Standard procedure for in-field analysis involved the identification of any ceramic material, notation on the nature of any ground or chipped stone artifacts, and an approximation of the type and relative percentages of siliceous detritus. Assumptions based on this analysis were then incorporated along with environmental and locational clues into the site description and classification.

followed, as indicated by the preponderance of Mancos and McElmo Black-on-white pottery types, by increased Pueblo II and Pueblo III (900-1300 A.D.) use of both riverine and upland environments. Fremont ceramics were observed at 12 sites, four of which also contained Anasazi ceramics. All Fremont sherds were either Emery or Sevier Gray (7001200 A.D.) with field identification indicating Emery Gray as the most common type (Madsen 1973). One site contained six sherds of an anomalous vessel which could not be comfortably placed in either Fremont or Anasazi typologies and has no parallels in Shoshoni wares in University of Utah collections. No clay sources were discovered in any area of the survey and coupled with the general scarcity of pottery at all ceramic sites â€” only partially accounted for by visitor removal of portable artifacts â€” it is probable that vessels were manufactured outside the survey area and transported in by groups using the district.

In general, the sites recorded during the survey contained scanty artifact material. On the river sites, the lack of surface material is most likely due to years of collecting by river runners because the sites are readily visible and accessible from the rivers. Amateur collecting may also account for an absence of the more easily recognizable artifacts near areas of limited visitor use in the Maze proper, although our examination of many apparently undisturbed sites showed them to be only slightly more productive in terms of artifact yield. In general, therefore, the paucity of artifacts seems to be an actual feature of Maze district sites.

CHIPPED STONE In-field analysis of the chipped stone material was carried out with three questions in mind: (1) the function of the artifact as it relates to activities carried on at the site; (2) local procurement of raw materials; and (3) the cultural affinity of the site's occupants. Artifact function was judged on the basis of overall morphology and obvious use markings such as edge attrition. Locally available resources were determined by a comparison of material with the types noted during the study. Cultural affinity, in most cases could not be assigned as chipped stone assemblages of both the Anasazi and Fremont are extremely variable and the few artifacts encountered were utilitarian types with little typological distinction. Ignoring any possible cultural or temporal distinctions and considering the entire artifact inventory of the Maze area as a single assemblage, the general picture is of a homogeneous tool kit, fashioned from locally available cherts, consisting of probable knives, scrapers, an occasional projectile point, and the materials of their production: hammerstones, cores, blanks, preforms, and debitage. Some of the debitage

CERAMICS Pottery sherds, and in one case a restorable vessel, were observed at 58 sites. Ceramic sites yielded less than 20 sherds, each normally representing less than five vessels. Forty-five sites contained ceramics of Pueblo manufacture, which can be assigned to the Mesa Verde Anasazi. No recognizable Kayenta sherds were found. Three upland sites produced sherds of Pueblo I or early Pueblo II (800-1000 A.D.) affiliation with no later ceramics present. Cortez Black-on-white and Deadmans Black-on-red sherds (see Breternitz et al. 1974) at those sites suggests a sporadic early use of the upland regions which was 92

Cowboy Caves (Jennings 1975) unearthed a great number of milling stones in association with large amounts of chaff and seeds of various grasses and nongrass plants, notably Indian ricegrass, goosefoot, and amaranth. The gathering of wild seeds is further documented in the area by ethnographic accounts of the Southern Paiute (Kelly 1964) and the Hopi (Whiting 1939). (2) A second assumption that may be made on the presence of milling stones is that in many cases the parties that forayed into the Maze included both sexes. The concept of transient household groups moving through the area thus becomes more tenable. Ground stone, including metates, handstones, mortars, and axe grinding grooves, was observed at 46 sites with one isolated find of a complete shallow basin metate. Ground stone was notably absent from most riverine sites. Locally available sandstones were utilized with the exception of three manos from two sites â€” manos made of exotic stones which would have been carried into the area. The three loaf-shaped manos are of Fremont manufacture as evidenced by comparison with University of Utah collections and similar materials described by Wilson and Smith (1975). No identifiable Fremont metates or fragments were discovered. Five sites yielded distinctive Anasazi ground stone. Two sites were assigned Pueblo affiliation due to the presence of stone axe grinding grooves in conjunction with other diagnostic Anasazi artifacts. Three additional sites contained deep trough metates with one open end; metates which are characteristically Anasazi (Swannack 1969). No identifiable Anasazi corn manos or fragments were discovered. The remainder of the ground stone artifacts could not be initially identified as to either temporal or cultural affiliation. The majority of metate fragments found were from thin slab metates with shallow, pecked basins. Handstones, with the exception of those Fremont types described above, were small â€” 8 to 13 cm. in diameter â€” one-handed rectilinear-to-round stones with bifacial use. Analagous metates and manos were uncovered in excavations at Cowboy Cave, an Archaic/ Basketmaker site near the survey area (Jennings 1975). The similarity suggests probable Archaic/ Basketmaker affiliation for those sites yielding such ground stone remains. Further study is needed in order to clear up the status of shallow basin metates and round handstones. However, the presence of distinctive Fremont and Anasazi ground stone artifacts in the uplands, essentially a nonagricultural locality, suggests that shallow basin metates and round handstones are indicative of nonagricultural wild plant processing activity and may be considered diagnostic for Archaic/ Basketmaker cultural activity.

also served as casual cutting, scraping, and planing implements. The constancy of the artifact inventory argues further for a uniform usage of the Maze area environs. Chipped stone, including the debitage resulting from tool manufacture, was recorded at 193 sites and by far constituted the bulk of observed surface remains. Lithic waste and occasional tools in and around 78% of all recorded sites indicate the importance of chipped stone to aboriginal populations, and that prehistoric use of the area was in part due to its mineral resources. Four sources of cryptocrystalline tool material were recognized during the course of the survey. However, outcrops were merely noted and those listed do not exhaust the possible sources of raw materials. Outcrops occur in the deep canyons of the Maze proper. On the unnamed bottom south and across the Green River from Fort Bottom, a semitransparent mottled gray chert was found weathering out of the Moenkopi Formation. At Turks Head Bottom extensive reddishpink to white cherts litter the headlands and derive from the Cutler Formation. In. the Doll House area blood-red cherts were observed in limestone lenses of the Cutler Formation (Lohman 1974). Projectile points from 19 sites, as well as 10 isolated finds, were observed and tracings were made of the distinctive points. Recognizable points were predominately Elko-eared (ca. 6400-ca. 1250 B.C.) and Elko Corner-notched (ca. 6000 B.C.-ca. 1350 A.D.) as defined by Aikens (1970). The occurrence of those distinctive projectile points in the uplands of the Maze suggest use of the area by Archaic groups. A general impression imparted by those sites containing masses of debitage and broken preforms with a corresponding scarcity of tools and points was of site use primarily for reduction of tool materials to portable blanks and preforms. One important resource of the Maze district, and perhaps adjacent areas, would appear to have been mineral.

GROUND STONE The presence of ground stone at the sites is significant in two respects. (1) While the milling stone is part of both the Fremont and the Anasazi tool kits, its association with sites in an area where agriculture is not possible suggests the processing of wild plant foods, probably grasses. The so-called Bird Site or Harvest Scene, as Schaafsma points out, appears to document the gathering of grasses, particularly Indian ricegrass {Oryzopsis hymenoides). Ethnobotanical data and coprolite analysis from Clydes Cavern (Winter and Wylie 1973) reveals grass seeds and various nongrass plants, such as goosefoot {Chenopodium sp.) and amaranth {Amaranthus sp.), made up a high percentage of the diet of both Archaic and Anasazi/ Fremont period inhabitants. Excavations at nearby

Three mortars were observed during the course of the survey, but their occurrence at sites of varied cultural affiliations precludes assigning them to any specific culture. 93

MISCELLANEOUS ARTIFACTS

mains were identified as mule deer or bighorn sheep in very limited quantities. Vegetal remains, with the exception of a corncob and squash button were generally structural elements in granaries or of questionable antiquity.

In addition to the major classes of artifacts discussed, occasional finds of bone refuse, vegetal remains, and a drilled shell bead, were recorded. Bone re-

CULTURAL AFFILIATION AND DATING The nature of a surface survey of archeological materials lends itself, at best, to an approximation of cultural affinities and time scaling. The scarcity of cultural debris and absence of permanently occupied sites in the Maze makes statements about dating and affiliation of sites no more than an informed best guess, based on a few sherds of pottery and some assumptions about rock art styles. The 248 sites of the Maze survey presented difficult problems of interpretation in respect to their cultural affiliation. Primarily, lack of distinctive and diagnostic cultural remains at 138 sites, mostly chipping stations, meant that over 50% of all sites recorded could not be placed into any cultural category. One purpose of the survey, that of cultural inventory of all sites within the survey area, was thus partially prevented. The failure to establish the cultural identity of those sites indicates a need for post-survey research, possible in the form of test excavation. Thirty-seven sites contained possible Archaic/ Basketmaker components, affiliation being based on either ground stone or projectile points, with the possibility that a sizable portion of those nonidentifiable sites are also Archaic or early Basketmaker. Especially eligible are those open camps containing slab-lined firepits and associated debitage but never ceramics. All but one of the possible Archaic/Basketmaker component sites occurred in the Maze uplands. This almost complete lack of early remains on the Green River is problematic. Archaic sites are generally nondescript and as such would not attract visitors or promote removal of artifacts or site destruction. Sites located on areas subject to river action may have been lost, but generally it can be stated that pre-agricultural use of the Green River bottoms and benches was not evidenced by survey results. Two possible explanations are either that use of river resources did not promote establishment of camps or that river resources were not important to Archaic/Basketmaker populations. Opinions on the problem are not possible from survey data.

ary camps. That mineral resources were extensively exploited was evidenced by the volume of chipped stone debitage. The presence of slab-lined firepits with associated metates and manos in different ecozones suggests varied wild plant use. Deer, bighorn sheep, and rabbits are present in the survey area and may have been an attraction for early gatherers. The transitory nature of the recorded camp sites indicates resource exploitation by mobile hunting and gathering groups. No base camps were discovered in the survey area, suggesting that the Maze District was an important, but not central, activity area. The excavations of Cowboy Cave (Jennings 1975) and Sudden Shelter (University of Utah analysis in progress), located west of Canyonlands, suggest the Archaic cultures exploiting the Maze resources may have been based to the west. Fremont components were evident in 21 sites indicating minimal use of the Maze district by peoples of that culture. Fremont ceramics and ground stone, and in certain cases rock art styles and motifs, were considered diagnostic. The scarcity of Fremont remains and the strong Anasazi nature of the area suggest that the main transition zone between the two cultures may be found further to the west, possibly between the Maze District of Canyonlands and the Henry Mountains. Survey data indicate only sporadic use of the area by Fremont peoples. Twenty-six percent of the recorded sites contained Anasazi components. The presence of diagnostic ceramics as well as distinctive ground stone, architecture, and rock art aided in the identification of Pueblo occupation. Forty-one of the 65 Anasazi sites were located on the Green River, an area amenable to corn horticulture; the remaining sites were considered as bases for resource use of the uplands by essentially river valley farmers. Upland Anasazi sites generally occurred close to the rivers or along physiographic features which would allow for easy access to various upland areas. Desirable locations were characterized by two or more components and often even historic camp use; 14 sites showed prehistoric multi-component use. No large or long-lived habitations were found in any portion of the survey area. Even those Anasazi sites located on the river appeared to be seasonal farming

An Archaic mode of small group seasonal gathering in the uplands was indicated by dispersed tempor94

camps or, at best, short duration farmsteads. A temporal outline of the cultures responsible for the pre-history of the Maze District can be established. As no Paleo-Indian activity was indicated, Archaic populations were probably the first users of the district. However, the intensity and duration of Archaic hunting and gathering activities is unknown. It can be postulated that area use followed a similar time frame as did nearby Cowboy Cave, which shows discontinuous use for over 5,000 years from ca. 6800 to ca. 2000 B.P. (Jennings 1975). No Basketmaker III use of the area was observed and the total lack of recognizable Basketmaker III artifacts or sites invites explanation. It is suggested that either Basketmaker III sites of the Maze District are nonexistent or nonidentifiable. A gap of approximately 400 years is indicated between Archaic/Basketmaker III and Anasazi occupations. Only post-survey research can effectively probe the apparent hiatus. The initial Anasazi use of the area is established as occurring between 900 A.D. and 1000 A.D. by recovery of dated ceramic types. The early occasional use of the uplands was followed by intensive use of both upland and riverine environments by the Mesa

Verde Anasazi with abandonment apparently coinciding with the general abondonment of the San Juan cultural area by 1300 A.D. (Rohn 1971). Placement of the Fremont into the temporal outline is difficult as no artifact types of that culture are firmly dated or show recognizable variation through time. Thus, Fremont use could have occurred anywhere between the 700 to 1200 A.D. dates assigned to the San Rafael Fremont (Marwitt 1970). Fremont use could have coincided with Anasazi occupation but the interaction of the cultures, if any, could not be determined. Aboriginal use of the Maze District after Anasazi abandonment is not strongly indicated. Late utilization of the area by historic Ute and Paiute groups is only hinted by two apparently late petroglyph panels, 42Wn658 and 42Wn668. As pointed out by Sharrock (1966) Ute and Paiute " . . . leave few traces of their passage, and it is unlikely that [they] would be distinguishable even if found." No artifacts or sites of later aboriginal groups were discerned, indicating that prehistoric occupation and use of the Maze District was terminated by 1300 A.D.

DISCUSSION

SETTLEMENT PATTERN AND SUBSISTENCE Any accurate study of settlement pattern requires good control of cultural and temporal variables. The survey conducted within the Maze can make no claims to have completely isolated these key attributes. However, despite the broad technological differences between the agricultural and architectural Anasazi and Fremont on the one hand, and the earlier Archaic/ Basketmaker, on the other, the groups are comparable in the similarity of gathering activities â€” activities basically Archaic in nature. Following Steward's contention that similar cultures will elicit a similar response to similar environment (Steward 1955), we can hazard some general statements of exploitive strategies within the Maze that are applicable to all the groups thought to have utilized the region for gathering activities. Some allowances have to be made for the possibility of previous climatic conditions that might have made the area more favorable. However, barring long-term conditions drastically different from the present, the model should be valid. Three general resources have been discussed as probable objects of aboriginal exploitation: big game species, wild grasses, and lithic materials. Apart from

the spatial and seasonal restriction inherent in harvesting, further limitations are imposed by the uncompromising nature of the country. Principal among the latter appear to be the availability of water and restricted lanes of access and passage. Examination of site distribution reveals an overall restriction of sites to main drainage channels despite the presence of equal or better habitat in many of the tributary canyons. This patterning of sites along the arteries of passage and the absence of sites in the side canyons probably indicates the main canyons were the foci of subsistence activities. Surprisingly, many of the camp sites, some apparently occupied for at least a few weeks at a time, were located away from any present permanent water source. It seems, therefore, that occupations probably were seasonally timed to periods when water was more readily available; i. e., spring and early summer. This coincides with the ripening of many of the spring grasses and their harvest time as documented for the Hopi (Lipe 1970). Since game is available in the area throughout the year and jasper has no seasonality, it is probably the above combination of factors which governed the utilization of Maze resources. The canyon bottoms represent only a portion of the area utilized by the aboriginal inhabitants. The drainages represent only one of four topographic areas 95

within the district boundaries, and it is likely that each of the four played a unique role in patterns of use and occupation within the area. Furthermore, if previous climatic conditions were not radically different than those of the present, the Maze District could support agriculture nowhere but along the river gorges.

Unlike the Needles survey, the presence of the Fremont in the Maze was documented by Fremont ceramics at several sites. However, given that the area was subject to only occasional use, the problem of where the foraging groups may have originated becomes a concern. Surveys by Morss (1931) and Gunnerson (1957) place the Fremont as far east as the Colorado River in the drainages of the Dirty Devil, Fremont and Escalante rivers. To the north, work by Wormington (1955) and Gunnerson (1957) has located Fremont sites as far south as Thompson Wash near the present town of Green River. Unfortunately, little archeological data is available for the San Rafael and Green River deserts. However, judging from past finds in similar topography those regions were probably only occasionally visited by small groups. This transient use pattern seems to hold for most of the northern area as far as Turner-Look at the base of the Book Cliffs (Wormington 1955). It seems most likely that the site clusters described by Morss (1931) in the Pleasant Creek area and by Gunnerson (1957) at the base of the Henry Mountains near Hanksville will be found to be related to the Maze District. The disturbing fact that accompanies "guesstimates" of home base regions for the aboriginal groups using the Maze District canyons is that the distances to locations suitable for permanent settlements are so great. Our impression is that the broad block of marginal lands stretching from the west bank of the Colorado River to the high plateaus was an area of common, occasional use by all agricultural groups of the region. Therefore, the cultural boundaries are extremely hazy. In this area the Fremont and Anasazi would have met as hunter-gatherers. Such a buffer zone situation may aid in the explanation of why both cultures remained relatively unaffected by their contact.

REGIONAL CULTURAL RELATIONSHIPS The utilization of lands west of the Colorado River by the Mesa Verde Anasazi is contrary to the prevailing view in Southwestern literature that the Colorado marks the northern boundary of the Mesa Verde expansion. Such utilization is not unique to the Maze, however, as recent survey work in Arches National Park indicates (Berry 1975). That survey generated a picture of intermittent resource procurement essentially the same as that of the Maze. The information has led Berry (1975) to redefine the boundaries of the Mesa Verde cultural groups to the northwest bringing both Arches National Park and the Maze District within that culture area. At present, it is impossible to say whether the new boundary will encompass new localities of Anasazi occupation, or whether it merely bisects a buffer zone which was sporadically used by both the Anasazi and Fremont in the harvesting of wild resources. As for the canyons of the lower Maze District, it seems most probable that the Anasazi visitors originated in the Salt and Indian Creek drainages of the Needles District. Movement into that region during Pueblo II to Pueblo III times has been summarized by Sharrock as being ". . . primarily from the east by several routes but specifically via Indian Creek; from the south through Bobby's Hole, Ruin Park and Beef Basin into the grabens, thence eastward to the Salt Creek drainage system," (1966).

SUMMARY A survey of selected areas of the Maze District undertaken by the Department of Anthropology, University of Utah located and recorded 248 sites of archeological significance. The information derived from the sites suggests that the area surveyed was occupied intermittently by prehistoric groups apparently for the harvesting of wild grasses, hunting, and procurement of lithic raw materials. Horticulture was practiced only along the banks of the Green and Colorado rivers where sufficient water was perennially available. The evidence indicates an Archaic/Basketmaker II use followed by intensive Anasazi use of the district

between late Pueblo II and early Pueblo III. Fremont groups also utilized the region while protohistoric use by Ute and Piute parties is also a possibility. Research of literature on the surrounding area prompted speculation of a home base for the Anasazi in the Salt Creek drainage and for the Fremont in the Pleasant Creek or Henry Mountain regions. Five major communicative routes or channels were noted in the district. They were: the Labyrinthine Maze canyons, Horse Canyon and the benches north of it, the Standing Rocks-Doll House area, the Wide Valley-Main Flat area, and Range Canyon. 96

Survey of the Maze District raised almost as many questions as it answered, including: the presence of a previously undefined Archaic substratum; the cultural affiliation, age, and function of open camp sites with large slab-lined fire basins; the lack of Basketmaker III sites and artifacts; the cultural affiliation of certain

unusual upright slab and mud granaries and the nature of the material stored in them; the nature of Fremont/ Anasazi interaction in the area; and, finally, the age and cultural affiliation of the Barrier Canyon style pictograph sites. These problems indicate the need for further work in the district including limited excavation.

REFERENCES 1958 Archaeological Survey of the Proposed Dead Horse Point-Junction Butte Utah State Park. Unpublished manuscript on file Department of Anthropology, University of Utah. Salt Lake City.

Aikens, C. Melvin 1970 Hogup Cave. University of Utah Anthropological Papers, No. 93, p. 56. Salt Lake City. Baerreis, David A. and Reid A. Bryson 1965 Climatic Episodes and the Dating of the Mississippian Culture. The Wisconsin Archeologist, p. 215. Milwaukee.

1969 The Fremont Culture: A study in culture dynamics on the northern Anasazi frontier. Papers of the Peabody Museum of Archaeology and Ethnology, Harvard University, Vol. 59, No. 2, pp. 38-46, 6273. Cambridge.

Baker, Arthur A. 1946 Geology of the Green River Desert-Cataract Canyon Region, Emery, Wayne and Garfield Counties, Utah. United States Geological Survey Bulletin, No. 951, pp. 7, 11. United States Government Printing Office, Washington.

Hunt, Alice P. 1952 Recommendations for Additional Work in the Indian Creek Area, San Juan County, Utah. Unpublished manuscript on file (42Sa00 Vol. 1), Department of Anthropology, University of Utah. Salt Lake City.

Baldwin, Gordon C. 1949 Archaeological Survey in Southeastern Utah. Southwestern Journal of Anthropology, Vol. 5, No. 4, pp. 393-404. Albuquerque.

Hunt, Alice P. and Bates Wilson 1952 Archaeological Sites in the Horse Canyon Area, San Juan County, Utah. Unpublished report in survey files of the Department of Anthropology, University of Utah. Salt Lake City.

Berry, MichaelS. 1975 An Archeological Survey of the Northeast Portions of Arches National Park. Antiquities Section Selected Papers, Vol. I, No. 3. Salt Lake City.

Hunt, Charles B. 1967 Physiography of the United States. W.H. Freeman & Co., San Francisco.

Breternitz, David A., Arthur H. Rohn, Jr., and Elizabeth A. Morris 1974 Prehistoric Ceramics of the Mesa Verde Region. Museum of Northern Arizona Ceramic Series, No. 5. Flagstaff.

Jennings, Jesse D. 1966 Glen Canyon: A Summary. University of Utah Anthropological Papers, No. 81, pp. 21-22. Salt Lake City.

Brew, John Otis 1946 Archaeology of Alkali Ridge, Southeastern Utah. Papers of the Peabody Museum of Archaeology and Ethnology, Harvard University, Vol. 21, pp. 3, 23. Cambridge.

1975

Cummings, Byron 1910 The Ancient Inhabitants of the San Juan Valley. Bulletin of the University of Utah, Vol. 3, No. 3, Pt. 2. Salt Lake City.

Preliminary Report: Excavations at Cowboy Caves, June 3-July 26, 1975. Unpublished manuscript on file, Department of Anthropology, University of Utah. Salt Lake City.

Kay, Marvin 1973 Archaeological Road Surveys in Canyonlands and Capitol Reef National Parks and Adjacent Bureau of Land Management Areas, Wayne and Garfield Counties, Utah. Report prepared for the National Parks Service, Midwest Archaeological Center, p. 35. Lincoln, Nebraska.

Gunnerson, James H. 1957 An Archaeological Survey of the Fremont Area. University of Utah Anthropological Papers, No. 28, pp. 80-86. Salt Lake City. City. 97

Kelly, Isabel T. 1964 Southern Paiute Ethnology. University of Utah Anthropological Papers, No. 69, pp. 41-42. Salt Lake City.

Rudy, Jack R. 1952 Archaeological Survey Beef Basin-Ruin Park Region, San Juan County, Utah. Unpublished report on file, of the Department of Anthropology, University of Utah. Salt Lake City.

Kowta, Mary M. 1963 An Archaeological Survey of Capitol Reef National Monument. Unpublished report on file, Department of Anthropology, University of Utah, p. 11. Salt Lake City.

1955 Archaeological Excavations in Beef Basin, Utah. University of Utah Anthropological Papers No 20. Salt Lake City. Schaafsma, Polly 1971 The Rock Art of Utah: from the Donald Scott Collection. Papers of the Peabody Museum of Archaeology and Ethnology, Harvard University, Vol. 65, pp. 77, 128. Cambridge.

Lindsay, La Mar and Rex E. Madsen 1973 Report of Archaelogical Surveys of the Pipe Springs National Monument Water Supply System Project, Kiabab Indian Reservation, Mohave County, Arizona; lion National Park Sewer Extension Project, Washington County, Utah; Arches National Park Road and Sewage Disposal Area Projects, Grand County, Utah; and Canyonlands National Park Road Projects, Needles and Grandview Point Areas, San Juan County, Utah. Unpublished report on file, Department of Anthropology, University of Utah. Salt Lake City.

Sharrock, Floyd W. 1966 An Archaeological Survey of Canyonlands National Park. University of Utah Anthropological Papers, No. 83, pp. 51-84. Salt Lake City. Steward, Julian H. 1955 Theory of Culture Change. University of Illinois Press. Urbana.

Lipe, William D. 1970 Anasazi Communities in the Red Rock Plateau, Southeastern Utah. In Reconstructing Prehistoric Pueblo Societies, p. 100. Edited by William Longacre, University of New Mexico Press. Albuquerque.

Swannack, Jarvis D., Jr. 1969 Big Juniper House. National Park Service, Archeological Research Series. No. 7-C, p. 109. Washington. Turner, Christy G. II 1963 Petroglyphs of the Glen Canyon Region. Museum of Northern Arizona Bulletin No. 38. Flagstaff. Whiting, Alfred F. 1939 Ethnobotany of the Hopi. Museum of Northern Arizona Bulletin, No. 15, pp. 18-21. Flagstaff.

Lohman, S. W. 1974 The Geological Story of Canyonlands. United States Geological Survey Bulletin, No. 1327, U.S. Government Printing Office. Washington. Madsen, Rex 1973 Fremont Ceramic Types. Museum of Northern Arizona Ceramic Series. In Press. Flagstaff.

Wilson, Curtis J. and Howard L. Smith 1975 Interstate Highway 1-70 Salvage Archeology 1974. Antiquities Section Selected Papers, Vol. II, No. 7. Salt Lake City.

Marwitt, John P. 1970 Archeological Inspection of Proposed Road From Squaw Flat to Confluence Overlook, Canyonlands National Park, Utah. Memorandum report on file, Department of Anthropology, University of Utah. Salt Lake City. 1970

Winter, Joseph and H. G. Wylie 1974 Paleoecology and Diet at Clyde's Cavern. American Antiquity, Vol. 39, pp. 303-315. Washington.

Median Village and Fremont Culture Regional Variation. University of Utah Anthropological Papers, No. 95, p. 143. Salt Lake City.

Woodbury, Angus 1959 Ecological Studies of the Flora and Fauna in Glen Canyon. University of Utah Anthropological Papers, No. 40. Salt Lake City.

Morss, Noel 1931 The Ancient Culture of the Fremont River in Utah. Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard University, Vol. 7, No. 3, pp. 4-12, 15, 26. Cambridge.

Wormington, H. M. 1955 A Reappraisal of the Fremont Culture: with a summary of the archaeology of the Northern Periphery. Proceedings of the Denver Museum of Natural History, No. 1. Denver.

Pierson, Lloyd M. 1959 Archaeological Resources of the Beef Basin, Needles, Salt Creek Mesa Area. Unpublished manuscript on file, Department of Anthropology, University of Utah. Salt Lake City. Rohn, Arthur H. 1971 Mug House. National Park Service, Archeological Research Series, No. 7-D, pp. 26-27. Washington.