Site Discovery and Testing
The city of Boone, North Carolina has recently experienced a frightening rate of commercial and residential development due to the growth of Appalachian State University and the local tourism industry. As a consequence, most significant prehistoric and historic archaeological sites, including prehistoric villages and camps, and possibly Benjamin Howard's cabin site (where Daniel Boone is alleged to have resided when he was in the area), have been destroyed.
The editor had been alerted to the possible existence of a few prehistoric sites remaining in the Boone area while conducting test excavations at the Winkler's Meadow Road site (31WT259) in preparation for construction of an apartment complex. That site, which has since been destroyed, exhibited disturbed contexts but yielded artifacts dating to the Early Archaic, Middle Archaic, Late Archaic, and Early Woodland periods (Kimball and Whyte 1992).
During the summer of 1993, Appalachian State University Anthropology students Elaine Minton and Chris Schwarz, under the direction of the editor, conducted a land modification survey of parts of the city of Boone, North Carolina. The purpose of the survey was to locate areas of ground which had not been severely impacted by historic period land modification and which, therefore, may contain preserved remnants of prehistoric archaeological sites.
The survey was conducted by obtaining permission from property owners to auger areas of open ground (lawns, parks, vacant lots) using a one-inch-diameter soil sampling tool. The auger tool was used in areas which looked like they may contain relatively undisturbed ground. Cores obtained were inspected for evidence of disturbances by cultivation or artificial grading or filling.
In the process of conducting that survey, a prehistoric archaeological site was discovered on a commercial lot owned by Mr. Gwyn Hayes just behind the Cardinal Motel near the intersection of Rt. 321 and Deerfield Road (Figure 1) . A few quartz, chalcedony, and chert flakes were observed on the ground in the heavily eroded western portion of the lot where a house once stood. Augering revealed that only the western-central portion of the lot, where stumps of large trees remain intact, exhibited a relatively undisturbed soil profile.
Mr. Hayes was contacted in early November, 1993 concerning the presence and potential importance of this site and generously consented to an archaeological investigation by the editor and a few volunteers. The work was conducted on Saturday, November 20th, and included the excavation of nine 40X40cm test pits spaced at 5 m intervals along cardinal axes, and two contiguous 1X1m excavation squares (Whyte 1994).
Prehistoric artifacts were concentrated in the western portion of the site, were recovered only from the A and B soil horizons, and included burned rocks, lithic debitage of a variety of materials, cores, and a few stone tools. The recovery
Figure 1. Location of Gwyn Hayes Site (31WT308), Watauga County, NC
of one basal fragment of a Savannah River stemmed knife indicates that the site was at least occupied during the Late Archaic period (ca. 5,000-3,000 B.P.). A large rhyolite core/scraper (what local collectors refer to as a "turtle back") was also recovered which exhibits extreme patination and resembles ones found in Paleoindian contexts.
In the spring of 1994 a report of preliminary investigations of the site was submitted to the North Carolina Division of Archives and History (Whyte 1994). In May, 1994, the Appalachian State University Archaeological Field School, under the direction of the editor, returned to the Hayes site for further investigation. By that time the property had been purchased by Howell Associates Architects of Boone for construction of a multistory office building and associated parking areas. The field school investigations resulted in the excavation of three distinct areas of the site separated by clusters of large tree stumps (Figures 2 and 3).
In all, 28 1X1m units were excavated yielding thousands of lithic artifacts, remains of hearths and associated wood charcoal, and minor amounts of historic (mostly 20th century) debris. Temporally diagnostic spear points and a spear-thrower weight recovered tentatively date the site's many occupations to the Early Archaic through Early Woodland periods with a heavier manifestation of the Middle Archaic (Morrow Mountain I) period. Additional excavations in the fall of 1994 and spring of 1995 (eight additional 1X1m units) have confirmed these findings (Figure 2) .
During the fall of 1994, the analysis of materials recovered by the summer excavations was begun as part of the education of students enrolled in the Archaeological Laboratory Methods course taught by the editor. The analysis was completed in the spring of 1995 by students enrolled in Archaeological Site Reporting, also taught by the editor. Several students enrolled in the Field School and therefore involved in the excavation have participated in each of these courses, thus being involved in all aspects of the investigation.
The following report was written by students enrolled in Archaeological Site Reporting (Anthropology 4750) at Appalachian State University. It details archaeological investigations of the Hayes site up to and including the fall, 1994 excavations and in appropriate instances, makes reference to certain findings of the spring, 1995 excavations. Details of the latter, however, will be provided in a follow-up report.
Environmental Setting
The Gwyn Hayes site (31WT308) is situated at 3,140 ft. above mean sea level on a bedrock high area which formerly overlooked the confluences of Winkler Creek and the Middle and East Forks of the South Fork of the New River (Figure 1) . This landform has a slope of 0-2 o which is, for the most part, west-facing, although the larger terrace system of which it is a part faces more to the northeast. The landform is situated immediately west of where the Middle Fork emerges from a narrow gorge adjacent to Yarnall Knob in the Blue Ridge Mountains (Figure 1) . Winkler's Creek and the East and Middle Forks merge to form the South Fork of the New River only 1 km to the north-northwest (Figure 1).
Figure 3. Excavation of the Gwyn Hayes site (31WT308), viewing southeast.
Throughout the past 10,000 years, as humans made regular use of the upland regions of western North Carolina, the site area was probably characterized by deciduous forests of chestnut and oak with perhaps open meadows along larger floodplains. Recent house, business, and road construction have left the Hayes lot an island of relatively preserved ground which has generally remained a residential yard area during this century. Heavy disturbance to the eastern and western fringes of the site has resulted from house construction/demolition and construction vehicle traffic. Only along the western central axis of the lot, where large tree stumps guard the ground, does the prehistoric component of the site appear to remain intact (Figure 4).
The soil on the site is classified as Perkinsville loam, undulating phase and consists of a dark brown loam A horizon from 0-6 cm, a medium grayish brown silty loam B horizon from 6-13 cm, and a yellowish brown silty clay loam C horizon from 13-20 cm, and a yellowish-brown saprolite of undetermined depth. Quartz and phyllitic quartzite pebbles are abundant throughout the profile. The A, B, and C horizons are included within approximately 20 cm of pre-Holocene colluvium which blankets the landform.
Other Prehistoric Sites in the Area
Several prehistoric sites have been located near the headwaters of the South Fork of the New River. Most of these are known from chance discovery and surface collections in highly disturbed areas. Only one other site within the Boone City limits, the Winkler's Meadow Road site (31WT259) has experienced any professional excavation (Kimball and Whyte 1992).
Prehistoric sites identified in the area range in age from the Paleoindian period (ca. 12,000 B.P.) through the Mississippian period (ca. A.D. 1500). These include small temporary special-use or resource extraction sites, temporary and seasonal camp sites, and possibly small villages (Purrington 1983).
The Winkler's Meadow Road site (31WT259) was located approximately 1.5 km to the west on a first terrace of the south bank of Winkler's Creek (Kimball and Whyte 1992). This site was partly excavated by Appalachian State University in spring, 1992 under contract for the Mental Health Association in North Carolina. These and subsequent salvage excavations by the Appalachian State University archaeological field school revealed a multicomponent, temporary camp site containing artifacts ranging in age from possibly the Early Archaic (ca. 10,000 B.P.) through Early Woodland (ca. 2700 B.P.) period. Most artifacts, however, dated to the Middle and Late Archaic periods (ca. 8000-3000 B.P.).
Figure 4. Southeastward view of the Gwyn Hayes site (31WT308) area showing large tree stumps.
Recent History of the Gwyn Hayes Site
Deed research for the Gwyn Hayes site was done at the Watauga County Courthouse in Boone, NC. Altogether, thirteen deeds were found pertaining to the property. Due to the 1873 burning of the Watauga County Courthouse, the first deed on file dates to February 23, 1906. At that time, the lot was part of a larger tract and the description of the land is as follows: "Beginning on a spruce pine at the road, running south nine degrees west with the road twenty-eight poles to a maple; thence south fifty-four poles to a stake at turnpike road; thence south forty-seven poles to a spruce pine at the lower side of the turnpike near the creek; thence west forty-seven poles to a chestnut, J.W. Farthing's and W.W.D. Edmisten's corner; thence north 124 poles with J.W. Farthing's line to a stake; thence east sixty poles to the beginning, containing forty-seven acres, more or less" (Book 7, Page 517 of Watauga County Deeds Records). According to this deed, a C.A. Greene purchased the land from R.P. Edmisten and wife for nine hundred dollars.
The first deed to mention any type of structure or building on the land is dated June 25, 1945. This deed indicates that the previous owners of the land, a Mr. H.M. Lewis and wife Delpha, had built some kind of structure. According to Deed Book 58, Page 609, the location of the "old dwelling" was, "...one hundred and fifty feet to the beginning, being lots as aforesaid upon which is located the old Lewis dwellings and buildings". Also a plat map of 1945 divides the land into four plots, thirty-six, thirty-seven, forty-six and forty-seven (Appendix A). This map illustrates a house, apparently the old Lewis dwelling, in the southern half of lots forty-six and forty-seven.
Until June 5, 1979 the land had changed ownership but with no evident modification of the land. On June 5, 1979 Cardinal Associates purchased the land from a Ms. Cora Lee Wilson. Between June 5, 1979 and November 3, 1986, a motel was developed on a section of the southern lots (46 and 47). The remaining undeveloped lots were subdivided into two lots now being 0918 and 1973. These two plots of land were then purchased by Mr. Gwyn Olin Hayes in 1986. The property has since been purchased by Howell Associates Architects for commercial development.
Goals of the Investigation
Given that the lot on which the site is located was recently purchased for commercial development, archaeological investigations of the site were initiated to salvage material remains and preserve knowledge of the site's archaeological contexts. These salvage excavations were designed to answer questions about the site's age and function and the role it may have played in prehistoric human adaptations in the southern Appalachian region. Nevertheless, additional questions may be addressed of the resulting data as the prehistory of northwestern North Carolina is explored further.
CHAPTER II
ARCHAEOLOGICAL FIELD METHODS
The western-central portion of the lot was gridded into three blocks of contiguous 1X1 m squares with reference to the grid datum established for the 1993 exploratory excavations (Figure 2). The grid was oriented with magnetic north and surface and base excavation elevations were recorded using a transit. The 1994 grid was not aligned precisely with the 1993 grid. As a consequence, units 10N9E and 10N10E of the 1993 excavations are located at 9N10E and 9N11E according to the 1994 grid (Figure 2). Five units were excavated in Block A, six units were excavated in Block B, and 17 units were excavated in Block C. In addition, portions of two units in Block B adjoining the 1993 excavation units were excavated (Figure 2). An additional four units were excavated in the Fall of 1994 and eight more were excavated in the spring of 1995 but not included in this report (analysis pending).
Each square was excavated by removing the sod, subdividing the square into 50 cm quadrants, and excavating each quadrant in 10 cm levels according to a line level anchored to the grid pin of the highest corner of the unit (Figure 3) . Diagonally oriented corner balks of 10X10 cm were left in the corner of each unit to preserve grid and surface references. Soil was removed by trowel and dry-sieved through 1/4-inch (6.5 mm) mesh. Artifacts and burned rocks were point-plotted on field forms and collected in provenience-labelled bags. Larger (5 cm+) pieces of burned rock were pedestalled in the event that they proved to be parts of intact features.
Most units were excavated to the base of Level 2 (20 cm) which, in most areas, corresponded to the middle or base of the C horizon. In Block C where excavation was more concentrated, a 5 cm level (Level 3) was excavated in the northeast quadrant of each unit for wet-screening through 1/16-inch (0.5 mm) mesh for the recovery of microdebitage.
Features were excavated according to their nature and content. Feature 1, a historic post mold and post conveniently bisected by the east wall of Unit 7N20E, was excavated by removing the half exposed in the floor of the unit and screening the soil through 1/4-inch mesh to recover artifacts. Feature 2, a prehistoric hearth located in Units 9N19E and 10N19E was excavated by disassembling the burned rocks in three 10 cm levels, mapping and photographing the configuration at the base of each. Soil collected from Levels 1 and 2 within the rock cluster was dry-screened through 1/4-inch mesh, while soil collected from within the lower reaches of the hearth (Level 3) was returned to the laboratory for flotation.
Feature 3, a historic structure foundation, was excavated by means of exposing, photographing, and mapping the feature and collecting a sample of foundation rocks and mortar for laboratory analysis.
Feature 4, a historic hearth or burned area was deemed insignificant and therefore simply exposed and mapped.
When the 1994 field season closed, the corner balks remaining in Block C were excavated in 10 cm levels and all units were backfilled and resodded.
CHAPTER III
RESULTS OF THE FIELDWORK
Contextual and Material Preservation
The Gwyn Hayes Site (31WT308) has experienced many types of natural and cultural formation processes over the past ca. 10,500 years of recurrent site visitation. These processes have adversely affected the context of the site and the preservation of cultural materials. Formation processes, both natural and cultural, affect the size, composition, number, preservation, and location of artifacts and features through a variety of disturbances. Most of these processes are somewhat regular in their causes and effects, and thus leave recognizable signatures in the archaeological record. Necessarily, before discussing analyses and interpretations of artifacts and features, as well as their vertical and horizontal distributions, a discussion of the evidence for formation processes and their effects on the site will be discussed.
Contextual Preservation
Evidence of postdepositional processes which have affected the Hayes site contexts include disturbed soil profiles, horizontal and vertical dispersion of artifacts and burned rocks, worm and insect activity in the soil, roots, krotovina, historic period constructions, and vertical clustering of objects at the B/C horizon interface.
Cultural Formation Processes.
The site is situated on a bedrock high area which remains stable and has received little or no soil deposition since its initial human occupation. Initial auger tests at the site in 1993 discovered that the only area retaining an intact soil profile was located in the center of the lot. The destruction of the surrounding area was due largely to historic construction of buildings and roads. Houses were located immediately to the west, north, and east of the excavated area, and a motel stands just to the south. The foundation walls of an outbuilding (Feature 3) were uncovered in the northeastern corner of the excavations. The eastern portion was also disturbed by large trucks during construction of the Watauga County Medical Center, just to the northeast (Gwyn Hayes, personal communication 1993).
The small "island" of intact soil horizons near the center of the lot was probably protected during historic times by four large trees whose stumps still remain on the site (Figure 4). They may have protected that area from historic cultivation (no evidence of cultivation was observed) and construction.
Prehistoric cultural processes may also have had an effect on the condition of the site's contexts and artifacts. Although these processes are more difficult to detect, there is a good chance, for example, that hearth stones and artifacts were reused, recycled, trampled, and dispersed as a result of intermittent prehistoric occupation of the site.
Natural Formation Processes.
The area near the center of the lot exhibited a well-developed soil profile (Figure 5) . However, there was clear evidence of the mixing of soil between horizons primarily as a result of bioturbation. Worms, ant colonies, and other burrowing insects were frequently encountered and recorded during excavation. They all have a profound impact on the archaeological record. Earthworms move through the soil by either pushing it aside or ingesting it and excreting it behind or on the surface. This process slowly but effectively churns soil and causes vertical movement of artifacts and ecofacts (Wood and Johnson 1978). This may account for the fact that burned rocks and other artifacts were vertically concentrated at the B/C horizon interface. Ants and other insects, although perhaps less prolific than worms, have a similar effect on artifacts and features. While never directly encountered, the presence of burrowing animals was evident by larger krotovina observed.
Floralturbation is equally effective in churning soil, and therefore disrupting artifacts and features. Roots open up spaces in the soil, moving artifacts and features as they grow. Roots and root casts were encountered in nearly every square and within features. Tree falls may also have affected the site's contexts over the past several millennia (Wood and Johnson 1978), although no evidence of them was observed in the area excavated.
Cryoturbation (freezing/thawing) is a disturbance process to be considered on sites in the high elevations of the southern Appalachians. The area's wet climate and temperature extremes are conducive to disruptive freeze-thaw cycles which cause vertical movements (primarily upward) of soil and objects near the surface (Wood and Johnson 1978).
The natural formation processes mentioned above are probably most accountable for the fact that artifacts are vertically distributed from the surface to a depth of at least 30 cm yet concentrated at the B/C horizon interface (approximately 10 cm). Figure 6 shows the vertical distribution of artifacts by size and illustrates that larger objects are relatively more abundant near the surface of the ground than are smaller objects. This indicates that the vertical spread of artifacts is a result of postdepositional processes such as bioturbation.
The apparent concentration of objects at the B/C horizon interface indicates that objects moving downward as a result of bioturbation met with resistance when encountering the more compact, clay-rich C horizon where bioturbation was minimized. Generally, only smaller objects have continued to settle deeper into the C horizon (Figure 6).
Although various postdepositional formation processes have clearly affected the site's contexts and contents, certain features and artifact clusters remained relatively intact. One prehistoric hearth (Feature 2), for example, was only minimally disturbed at the top levels; the carefully assembled cobbles of the hearth's base and associated carbonized wood remained virtually undisturbed. Likewise, a dense concentration of vein quartz debitage centered in Unit 7N19E appeared to remain relatively undisrupted.
Material Preservation
The only materials recovered from the site include historic artifacts, historic vertebrate remains (bones of domesticated fauna), lithic artifacts, burned rocks, and charred plant remains. Animal remains and unburned plant remains, undoubtedly deposited on the site in prehistory, had evidently decomposed. The
acidic soils and moist climate of the region account for the lack of prehistoric archaeofaunal remains and unburned prehistoric archaeobotanical remains on the site. Soil acidity appears to have differentially affected the preservation of stone artifacts on the site as well. Many of the rhyolite flakes recovered, for example, are considerably patinated.
No prehistoric ceramic artifacts were recovered by the excavations, although one projectile point probably dating to the Early Woodland period was found. It is possible that prehistoric ceramic artifacts such as pottery sherds were part of the depositional assemblage but have not survived the deleterious processes which have evidently also affected the preservation of organic remains and the site's contexts.
Features
Four relatively intact cultural features were discovered by the excavations (Figure 7) . These features, discussed in detail below, include the remains of two prehistoric hearths (Features 2, 5), one historic hearth/burned area (Feature 4), one historic post mold (Feature 1), and one historic outbuilding foundation (Feature 3). In addition, several amorphous clusters of burned rocks recognized as prehistoric hearth remnants were discovered but not labeled as discreet features. These may have been disturbed by natural or cultural agencies as discussed above.
Artifacts and Ecofacts
Artifacts were recovered from all horizontal and vertical units of excavation. Those recovered include lithic debitage, cores, and tools, burned rocks, and historic (mostly 20th- Century ) household and architectural debris. Prehistoric artifacts recovered range in age between 10,500 and 3,000 B.P. Ecofacts observed and sometimes recovered include historic and prehistoric wood charcoal and two recently deposited bones of domestic fauna
Prehistoric artifacts were found immediately beneath the sod and extending to at least 25 cm but were most concentrated within the B horizon (approximately 10 cm). Several horizontal concentrations observed are discussed below. Historic period artifacts were also recovered from throughout the profile but were most prevalent between the surface and 10 cm deep.
CHAPTER IV
METHODS OF ANALYSIS
Lithic Artifacts
Lithic artifacts recovered by hand and 1/4-inch mesh screening from the Gwyn Hayes Site (31WT308) were washed, dried, and rebagged according to their site provenience. They were then separated into the following categories for varying levels of analysis: (1) vein quartz debitage; (2) nonlocal material debitage; (3) stone tools and cores; and (4) burned rocks (Feature 2 only). Prehistoric lithic artifacts recovered by 1/16-inch mesh wet-screening (Level 3) were separated from recovered sediment and rebagged for quantification. Data forms were designed for the the analysis of each category.
Vein Quartz Debitage
Most of the artifacts recovered from the Gwyn Hayes site are flakes and flake fragments of locally derived vein quartz. Countless researches have recognized the difficulty of studying vein quartz artifacts. Because of the heterogeneous crystalline structure of the rock, intentionally imparted fracture in the making of stone tools yields seemingly random structure to the debitage and seldom complete flakes. It is therefore difficult to identify the various structures (striking platforms, bulbs of percussion, etc.) of flakes useful in reconstructing methods of reduction and the kinds of tools manufactured. Given these obstacles, vein quartz debitage was analyzed according to only three criteria: size (in 0.5 cm increments), cortex type (absent, residual, water-tumbled), and quantity per provenience.
Non-Local Material Debitage
Debitage of non-local materials recovered from the site include flakes and shatter of Knox chert, chalcedony and jasper from the Shady Dolomite formation, Erwin formation quartzite, crystalline quartz (possibly obtained locally), rhyolite and green metasiltstone from the Carolina Slate Belt formations, and rhyolite from the Mt. Rogers group.
These artifacts were analyzed according to five criteria: lithic raw material, debitage type (bifacial thinning flake, core-derived flake, bipolar flake, indeterminate flake fragment,shatter); size (in 0.5 cm increments), cortex type (absent, residual, water-tumbled), and quantity per provenience. Bifacial thinning flakes were distinguished by the presence of an acutely angled, multifaceted, and ground platform, a flattened bulb of percussion, and multiple dorsal scars. Core-derived flakes were distinguished by the presence of a flat, less acutely angled platform, a pronounced bulb of percussion, and parallel dorsal scars. Bipolar flakes were distinguished by the lack of a platform and bulb of percussion, crushing at either or both ends, and (frequently) tightly spaced concentric compression rings. It is recognized that bifacial thinning, core reduction, and bipolar flaking occasionally yield similar byproducts and that flakes were occasionally misidentified in this analysis.
Fine-Screen Sample Analysis
Artifacts separated from the fine-screened material recovered from Level 3 of the northeast quadrant of each unit in Block C include primarily lithic debitage. Specimens larger than 1/4 inch were separated and analyzed within the appropriate groupings above. Microdebitage (less than 1/4 inch) was analyzed according to lithic raw material and quantity.
Burned Rock Analysis
Burned rocks, some of which are fire-reddened and some of which are fire-cracked, were recovered from all areas of the site, but only those associated with Feature 2, because it was relatively undisturbed, were recovered and analyzed. These were analyzed according to rock type, cortex, weight (to the nearest gram), and quantity. Most specimens were either metawacke, phyllite/phyllite metaquartzite, or metabasalt which occur in the local Grandfather Mountain Formation.
Lithic Tool Analysis
Stone tools recovered from the site include bifacial tools (projectile points, knives, unfinished bifacial tools), flake tools (utilized flakes, scrapers, other flake tools), a cobble tools (chipped cobble), and ground stone tools (one spear-thrower weight and one indeterminate tool fragment). Although cores did not necessarily serve a function other than as sources of flakes and tools, they were analyzed as tools.
Projectile points and other bifaces were analyzed according to lithic raw material, portion, source blank (core flake, cobble, etc.), length, width, and thickness (to the nearest 0.01 mm). Cores were analyzed according to lithic raw material, maximum dimension (nearest 0.1 mm), and weight (nearest 0.1 g). Flake tools were analyzed according to lithic raw material, source blank (core flake, bifacial thinning flake, etc.), maximum dimension (to the nearest 0.01 mm), form of utilization/retouch (unifacial, bifacial, grinding), and location of utilization/retouch (lateral dorsal, distal ventral, etc.).
Historic Artifact Analysis
Historic artifacts, since they are relatively few and mostly of very recent age were simply washed, sorted, labeled as to their function (window glass, bottle glass, wire nail, etc.) and tabulated. They generally include the kitchen and architectural debris expected of a 20th-century house lot.
Radiocarbon Sample Analysis
Wood charcoal was collected by hand (larger pieces) and by flotation from the base of Feature 2. The flotation sample was processed in the laboratory using tap water, a 5-gal plastic bucket, and 0.5 mm mesh screen. The soil was poured into the water-filled bucket and the contents were then stirred to promote the suspension of lighter particles. The 0.5 mm mesh sieve was passed through the water until repeated passes failed to yield additional suspended material. The heavy fraction was then wet-screened through 1/16-inch mesh to collect artifacts and ecofacts which had not become suspended.
After the two samples were dry, artifacts and charred plant remains were separated using forceps. Artifacts were then analyzed according to the criteria discussed above. Charred plant remains (approximately 6 g) recovered from the flotation sample were sent to Beta Analytic, Inc. for a radiocarbon assay. Hand-collected pieces of wood charcoal were cross-sectioned and examined at 20X with a binocular microscope for wood species identification.
CHAPTER V
RESULTS OF THE INVESTIGATION
Features
Four relatively intact cultural features were discovered by the excavations (Figure 7). These include the remains of two prehistoric hearths (Features 2, 5), one historic hearth/burned area (Feature 4), one historic post mold (Feature 1), and one historic outbuilding foundation (Feature 3). In addition, several amorphous clusters of burned rocks recognized as prehistoric hearth remnants were discovered but not labeled as discreet features.
Feature 1
Located in the eastern edge of square 7N20E, Feature 1 appeared to have been a historic post mold (Figure 8) . The uncovered portion of the feature extended 10 cm west from the east wall and measured 20 cm in diameter. It was revealed during the excavation of level two, 17 cm below the surface, and extended to a depth of 51 centimeters below the surface. The only artifacts recovered from the soil of Feature 1 were a rhyolite flake and the proximal end of a wooden post recovered from the base of the feature.
Feature 2
Located in the northern half of square 9N19E and the southern half of 10N19E, Feature 2 is a relatively undisturbed hearth (Figure 9) . Level one revealed a scatter of burned rocks in both squares covering an area of approximately 47 cm. Level two revealed a circular arrangement of rocks with a depressed center and measured 55 cm in diameter. The center of the feature at this level included a relatively high concentration of oak ( Quercus sp.) wood charcoal. The circular arrangement of rocks continued to a depth of 25 cm. The rock materials varied somewhat in composition from top to bottom (Table 1). Soil was collected from among the rocks between 20 and 25 cm below surface and floted in the laboratory for the recovery of charcoal. Artifacts were recovered from within and near the hearth in each level, the majority being vein quartz and rhyolite flakes. A rhyolite Morrow Mountain I projectile point was found 18 cm below the surface in close proximity to the feature, suggesting at least a Middle Archaic period use of the feature.
The height of the feature (approximately 25 cm) coupled with the vertical variation in rocks used in its construction, suggest that the hearth may have been rebuilt several times for reuse. In effect, this "site furniture" was "pulled up out of its matrix for reuse" (Binford 1979:264) by subsequent prehistoric visitors to the site.
Figure 9. Plan view of feature 2, Gwyn Hayes Site (31WT308).
|
