Phase II and Phase III Archaeological Database and Inventory
18PR33 Warfield
Diagnostic Artifact Data
Features
Summary
The Warfield Site (18PR33) is an apparent Late Archaic and Woodland period base camp near Bowie in Prince George’s County. The site is situated on the Pleistocene terrace and floodplain of the Patuxent River just south of a first order tributary. An extensive freshwater swamp separates the site from the Patuxent River. The soil was at one time cultivated, but in more modern times has been partially wooded and also the site of a residential structure. Soils at the site are of the Collington-Wsit complex.
The Warfield site has long been known to collectors and to avocational archeologists. One of these early investigators, Carroll Thomas, first visited the site in 1935 and later described his findings in detail. Thomas noted three main occupation areas within the site, where artifacts tended to cluster into different types. Thomas reports recovering thousands of lithic artifacts as well as pottery sherds. On a map he created of the site, Thomas recorded particularly notable finds revealing the recovery of 1 whole bannerstone, 6 bannerstone fragments, 7 groundstone celts, two stone axes, 3 gorget fragments, a mortar, 3 pestles, an abrader, 34 drill fragments, at least 31 pottery sherds, and a clay pipe stem. Thomas also reports that other collectors were active at the site in the 1930s and 40s.
The site was first examined by professional archeologists in 1980. Researchers from the Maryland Geological Survey (MGS), Division of Archeology investigated the site as part of a Phase I reconnaissance of nearby Maryland Route 3 in both Prince George’s and Anne Arundel Counties. At the time, Route 3 was under consideration for possible realignment and/or reconstruction in order to facilitate the relatively heavy flow of traffic on the road at that time. Three test units (of undocumented size, but likely 1 X 1 m) and several shovel test pits (STPs) were excavated within the confines of the site as documented by Thomas. No artifacts were encountered in any of the STPs, however, the test units, which were situated in and near a vegetable and flower garden at the aforementioned residence did produce artifacts. The property owner reported recovering a significant amount of artifactual material from the plowzone in this area over the years. All test unit soils were excavated by trowel in arbitrary levels and screened through hardware mesh.
Artifacts recovered during the Phase I survey include a Piscataway point, 3 rhyolite flakes, 9 quartz flakes, 2 quartzite flakes, 22 lithic chunks (jasper, quartz, and quartzite), 5 pieces of shatter (rhyolite and quartz), a battered quartz cobble, and a Mockley body sherd. In addition, the property owner’s collection was examined and found to include 1 Lamoka point, 2 Vernon points, 3 Bare Island, 7 Broadspears, 4 Piscataway points, a Selby Bay point, 2 Potomac points, and 1 Potomac Creek cord-impressed sherd. Based on these findings, the MGS recommended Phase II testing be carried out at 18PR33.
A combination of subsurface testing techniques was utilized when MGS researchers returned to the site in late 1982 to conduct the Phase II research. The strategy involved systematic shovel test pitting followed by excavation of one-meter-square test units. STPs roughly 40 cm in diameter were excavated to sterile subsoil, and the soil was sifted through hardware mesh, with all cultural material retained. Detailed notes on standardized forms were taken on each pit as to soil color and texture, stratigraphy, and artifactual content. Given the very large size of 18PR33 (in excess of 5,000 m2), the shovel test pits were excavated on an interval of 10 meters. This strategy allowed for definition of artifactual concentrations and characterization of stratigraphy. The second phase of shovel test pitting consisted of additional STPs excavated between the 10 m intervals at 5 m intervals in areas of heavy artifact concentration.
Based on soil stratigraphy and artifact concentrations identified by the shovel test pits, the locations for 20 one-meter square test units were selected. The general procedure for excavation of the one-meter tests was to remove the plowzone and excavate subsequent strata by natural layers. When natural layers were thicker than 10 cm they were subdivided into 10 cm arbitrary levels to maintain vertical control. Exact provenience was recorded for diagnostic artifacts recovered in situ below the plowzone. All soils were sifted through hardware mesh screens.
It was evident, based on the shovel testing, that the site extends beyond the Phase II project limits as artifact counts did not appreciably diminish (an in many cases increased) toward project boundaries. Moreover, there was little in the way of discrete artifact clusters that could be identified. However, three loci of occupation were identified. An Early Woodland component is located on the Patuxent River floodplain. Later ceramics, especially Mockley, tended to occur along the terrace edge of the adjacent first order tributary. A southern slope area appears to have been primarily used by Archaic hunters where quartz, quartzite, and rhyolite debitage was plentiful and a number of Late Archaic points were found.
In general, the test units located in the southwestern portion of the site exhibited a mixed stratigraphy. Cultural material was present throughout the humus, Ap, and A2 horizons as well as slopewash layers. The dispersal of artifacts both horizontally and vertically through the A2 horizon suggested considerable reworking of soils in this area. In one unit, mending portions of a Bare Island point were recovered from 43 and 73 cm below datum respectively. A disturbed hearth feature associated with Mockley ceramics was encountered in this portion of the site as well, in a unit at the northwest edge of the site. Accokeek and Popes Creek ceramics were found in soil layers beneath the feature, suggesting that this area of the site may be vertically intact, if somewhat disturbed horizontally.
Test squares on the Pleistocene-age terrace in the central portion of the site had shallow profiles. A shallow plowzone or combination plowzone/slope wash was situated atop a shallow A2 sandy clay loam horizon, which graded into a B horizon of sandy clay. Some units exhibited more evidence of disturbance than others and in some areas the plowzone was overlain by an additional layer of fill (associated either with house or road construction).
Test squares on the floodplain to the northeast showed strong evidence of an undisturbed profile. Essentially, an old A horizon lies buried under 36-68 cm of recent floodplain deposits. Although testing was limited, this buried A horizon appeared to be unplowed. Underneath the buried A was a 12-30 cm thick sandy silt loam layer. Most of the cultural material came from the interface of this soil layer and the A horizon above. The materials appear to be Terminal Archaic to Early Woodland in age. This floodplain locale would today be situated in the highway median, capped by both highway fill and the aforementioned floodplain deposits.
Thirty-two diagnostic projectile points were recovered during the Phase II excavations at the site. The assemblage included 7 Bare Island points, 2 Brewerton side-notched points, 2 Brewerton eared-notched points, 1 Lamoka point, 1 Claggett point, 2 Vernons, 2 Normanskill points, 3 Piscataway points, 2 Susquehanna Broadspears, 2 Orient Fishtails, 3 Calvert points, 1 Jack’s Reef Pentagonal point, 1 Levanna, 2 Madison/Potomac points, and 1 elongate triangular form. The remainder of the flaked stone assemblage consisted of 3 chert scrapers, 14 quartz scrapers, 2 quartzite scrapers, 3 quartz drills, 24 other quartz bifaces, 16 other quartzite bifaces, 12 quartz cores, 1 quartzite core, 69 utilized/retouched flakes, 5,107 flakes, 298 chunks, and 611 pieces of shatter. The vast majority of the raw material in the assemblage is quartz and quartzite, followed by much smaller quantities of rhyolite, chert, jasper, sandstone, siltstone, argillite, ferric sandstone, chalcedony, silicified sandstone, and unknown materials. A total of 1,170 pieces of fire-cracked rock were encountered.
Other lithic artifacts recovered from the site include an atlatl weight (bannerstone) fragment, a flake of a groundstone implement, a polished slate piece, 3 fragments of ground sandstone, a steatite bowl fragment, and 6 hammerstones.
Two Hundred and Ninety-six ceramic sherds were identified during Phase II work at 19PR33 ranging from Early Woodland Marcey Creek sherds to Late Woodland Potomac Creek. The ceramic assemblage consisted of 1 Marcey Creek sherd, 3 Selden Island sherds, 55 Accokeek sherds (3 rims), 32 Popes Creek sherds, 117 Mockley sherds (3 rims), 28 Townsend series sherds (25 Townsend and 3 Rappahannock Incised), 50 Potomac Creek sherds (1 rim), and 10 sherds of an unidentified Late Woodland courseware (1 rim). The unidentified ware was a thick-walled ceramic with course quartz and sand temper and cord-marked exterior.
A summary of chronological indicators (projectile points and ceramics) at the Warfield site shows evidence for occupation from the Late Archaic through the Late Woodland (4000 BC-AD 1600). The site is characterized by a series of overlapping loci of occupation. Generally, projectile points were more scattered across the site than ceramics, and were therefore less useful in determining loci of occupation. The Archaic occupations do not appear to cluster in any one part of the site as Archaic artifacts are found in every area of the site. In the southern part of the site the presence of diagnostic lithic tools and scarcity of ceramics is taken as evidence for Archaic activities on the upper slopes away from the Patuxent floodplain and its nearby tributary. Early Woodland diagnostics are a bit more concentrated than Archaic. Based on Calvert points, steatite-tempered sherds, and Accokeek sherds, there are 3 distinct Early Woodland loci on the site: one on the floodplain and two on the Pleistocene terrace. Popes Creek ceramics, which date slightly later, are mostly heavily concentrated on the floodplain. Mockley ceramics (Middle Woodland) are concentrated on the northwest part of the site, although they occur on the southern slope as well. The distribution suggests a broad utilization of the area, and, in terms of total numbers of ceramics, also suggests more intensive activity. This could take the form of larger group sizes, occupations of greater seasonal duration, or more frequent visitation to the site. Late Woodland representation at the site includes a moderate number of Townsend ceramics and comparatively more Potomac Creek. Both of these types are concentrated to the north adjacent to the first-order tributary of the Patuxent. A Late Woodland coarseware is also present in smaller quantities in this area.
The Warfield site probably served primarily as an interior hunting site and possibly as a seasonal fishing station. Its cultural affiliations may have been quite diverse, as it seems to display similarities to the Piedmont Potomac, as well as to the estuarine and coastal Patuxent throughout the entire span of occupation.
The significance of the Warfield site should be considered in the context of its position in the Patuxent River Valley. It is one of the largest recorded sites this far upstream along the Patuxent River and has yielded abundant artifacts indicating occupation from the Late Archaic through the Late Woodland. In addition, it has one scarce and one previously unidentified ceramic type (i.e. Selden Island and Late Woodland courseware respectively). Because of its position in the upper freshwater portion of the Patuxent, this site could provide essential information concerning the extent of movement, trade, and communication across river drainages in the interior coastal plain as well as between the interior upland and the outer coastal plain. The site should also be considered an especially scarce archeological resource because many previously recorded sites in this section of the Patuxent Watershed have been destroyed in modern times due to extensive sand and gravel mining.
Both archeological work and examination of the site by a pedologist revealed that most of the site was disturbed by erosion or colluvial action. Two small areas of the site retain sufficient integrity to warrant further investigation: one on the lower floodplain and one on the Pleistocene terrace. On the floodplain, artifacts in the buried humus layer include a steatite bowl fragment, steatite and sand tempered ceramics, and a Calvert projectile point, which all date to the Terminal Archaic-Early Woodland. This lower floodplain portion of the Warfield site has excellent preservation and research potential because it is sealed in the floodplain and it represents an occupation from a short time span. Furthermore, road construction since the Phase II study involved primarily the import of soils to build up the road grade rather than excavation, further preserving the site.
Another area on the upper terrace has potential for some stratigraphic separation and possibly feature preservation in the A2 and subsoil beneath the buried plowzone. The size of the area is estimated to be approximately 400 square meters based on both test units and STPs. One of the units in this area revealed the scattered remains of a hearth 45 cm below surface. This feature occurred in a band of soils where colluvial activities have stratified cultural deposits. Distributional analysis indicated that this was the most intensively used portion of the site. In addition, Carroll Thomas mentioned the presence of “fire pits” on this section of the site. The Phase II investigations suggest that a Selby Bay component can probably be identified and isolated.
Based on the Phase II investigations, the Warfield site was recommended to be eligible to the National Register of Historic Places for its integrity, scarcity, and research potential. The site lies preserved in the median and in areas surrounding Maryland Route 3.
|
