Phase II and Phase III Archaeological Database and Inventory
18CE29 Hollingsworth Farm
Diagnostic Artifact Data
Features
Summary
The Hollingsworth Farm Site (18CE29) consists of Late Archaic to Late Woodland base camp occupations and a prehistoric burial. It is located south of the city of Elkton in Cecil County in an area known as Historic Elk Landing. The site is situated partly in an agricultural field and partly in woods along the first terrace and adjoining ridgeline on the bank of one of the tributary creeks to the Elk River. It extends for approximately 1215 m along the shoreline and 180 m inland at its widest point. The southern portion of the site is located within the 2002 Maryland Historical Trust easement at Elk Landing. The site’s eastern and southern edges abut a marsh that borders the stream. Soil on the terrace is classified as Elsinboro silt loam (0-2% slope) and the remaining soil is Elsinboro silt loam (2-5% slope), moderately eroded.
Archival research indicated a rich history in the site area. During his trip in the area ca. 1608, John Smith gave the Elk River its name when he supposedly saw a herd of elk along the river’s edge. At that time, the Massawomekes, Susquehannocks, and Tockwoghs occupied what would become Cecil County. The Susquehannocks (or Susquehanna) lived in the area at the fork of the Elk River. The eastern branch of the river came to be called the Big Elk Creek and the western branch is known as the Little Elk Creek. In the mid-1600s, a Swedish and Finnish community developed in the fork of the Elk River. Recorded history of the site dates back to 1655 when the local Indian tribe, the Minquas (the Susquehanna), presented the Governor of New Sweden (now Wilmington, DE) with land they called “Chakakituqe” on the Elk River in exchange for the promise that the Swedes would establish a trading post there. The place name became “Sahakitko” (meaning “squeaky saw” in Finnish) among the Scandinavians. Trade between the Scandinavian settlers and local Susquehanna Indians was carried out between 1655 until the local aboriginal population disappeared around 1720. A Swedish or Finnish settlement never appeared to develop on the land, nor did the Dutch, who conquered the Swedes later that year, appear to lay claim to the land.
In 1672, a patent was issued to an Englishman named William Price for “Price’s Venture” on a 250 acre parcel located somewhere on the north side of Elk River. In 1678/9, a patent was granted to J. Browning and R. Nash for a 500 acre tract called “Successor” in the fork of the Elk River. In 1681, Nicholas Painter patented a 1400 acre tract called “Friendship” that was located on the west side of the northeast branch of Elk River (Big Elk Creek) adjoining Successor. Painter gave 3 Finns 50 acres each in exchange for building a mill on his other land. Sometime around 1693, a Swedish Indian trader named John Hanson Steelman acquired the south 200 acres of the Successor tract facing Little Elk Creek where he remained until ca. 1710. History indicates that Steelman established a trading post called Transtown on the parcel. It has been postulated that a log structure (demolished in 1917) that abutted the extant Stone House may have been his original trading post. However, a late 17th century construction date for the log building has been debated. An early 19th century date of construction seems more likely and it has been suggested that the trading post is to be found farther south on the property, possibly in the wooded area. However, ground disturbances around the house in the second half of the 19th century may have removed any archeological evidence associated with the earlier occupation. In any case, by the early 18th century Steelman had moved his trading post west following the migrating Indian population.
In 1711, most of the settlement called “Sahakitko” was absorbed into a 150 acre farm that was started by Henry Hollingsworth. The land was described as being bounded on the southeast by the Steelman plantation. After Henry’s death, the farm was expanded by his son Zebulon. Zebulon Hollingsworth added an additional 425 acres to the Hollingsworth farm between 1727 and 1752. On a 1742 deed, Zebulon was listed as “Innholder”. He died in 1763 and the Hollingsworth farm and lands were taken over by his son, Zebulon Jr.
During the 18th century, the area of Elk Landing was an important transportation hub for shipping food and other materials. A wharf was located on the east bank of Little Elk Creek, just southwest of the stone house (Elk Landing Wharf is recorded as site 18CE328). An advertisement from 1767 indicated that Zebulon Hollingsworth Jr. was involved in the shipping trade. “Good stores” for the reception of materials in transit and “houses of entertainment” were available at the Head of Elk, which was incorporated into the town of Elkton in 1787. These facilities may have been established on the 200 acres Zebulon Sr. purchased in 1742. It was postulated that the log structure was Zebulon Jr.’s upper storehouse. A stone addition was believed to have been built onto the log structure in 1775, although dendronological testing indicated that the stone addition could not have been built onto the log structure until after ca. 1783. Even so, very few late 18th century artifacts have been recovered from excavations in and around the stone house that would support a late 18th century date of construction. It was postulated that the log structure was converted to use as a kitchen when the stone addition was built. It is presumed that the brick Hollingsworth House was also built in the early 1780s. A map from 1857 shows the Stone House and the Hollingsworth House and two other structures that are labeled “wharf” (probably the wharf and associated warehouse). William Hollingsworth took over the land at the time of his father’s death in 1812. During the War of 1812, a small earthwork and a few cannons (some say only 1 cannon), known as Fort Hollingsworth, was located just southeast of the Stone House on the Hollingsworth Farm. William died in 1844 leaving the house to his wife. In 1848, the Hollingsworth House was gutted by fire and subsequently rebuilt in its present Gothic-Revival style. Descendants of the Hollingsworth family continued to occupy Elk Landing until January 2000 when the Town of Elkton and the Historic Elk Landing Foundation Inc. signed a 99-year lease for the restoration, management, and operation of the site as a living history museum. Historic Elk Landing is a ca. 44 acre area starting at the foot of Landing Lane and encompassing the area between the forks of the Elk River.
The Hollingsworth Farm prehistoric site was first recorded in 1970 by staff from the Maryland Geological Survey, Division of Archeology, with information from local informants and collectors. At that time, stemmed and side-notched Archaic points, soapstone bowl fragments, and thick prehistoric pottery sherds were noted at the site and in the informants’ collections. Presumably, the site boundaries were also established at that time, likely reflecting the extent of finds as discerned by information from the local informants. In 1980, the first systematic study of the site was conducted. A Phase II testing program was initiated ahead of construction of the Cecil County Correctional Facility. The study area was confined to the area of potential effect which encompassed the north central portion of the overall site. At the time of the archeological investigations, the field had been plowed in the past but was recently left fallow; a plow disturbed layer measuring up to 30 cm was noted in some parts of the project area. However, the majority of the proposed development area appeared to remain undisturbed. The effect of years of surface hunting by local collectors was also a consideration when examining the artifact assemblage. Field methods employed during the investigations included surface controlled survey and collection of 176 square surface units (each of which measured 100 m²), and the excavation of 63 shovel test pits (STPs) dug at 20 m intervals and 13 test units, mechanical stripping of five long, 2 meter wide trenches, and feature excavation. All artifacts were retained.
Based on the artifact distribution, it was noted that the southeastern and south central portions of the study area contained the most evidence for lithic related activity. Five possible cultural features were identified as a result of the subsurface testing. There were 2 small soil stains, a pebble and cobble filled depression, a prehistoric pit (Feature 3) containing several ceramic sherds, lithic debris and fire-cracked rock, a bowl-shaped and charcoal filled pit, and a possible hearth. The archeologists determined that the site represented a seasonal camp where subsistence –related activities were carried out.
A total of 2,386 prehistoric artifacts were collected during the 1980 Phase II investigations. There were 552 pieces of debitage, 63 cores and core fragments, 4 knives, 7 scrapers, 17 utilized flakes, 4 perforators, 6 preforms, 27 bifaces, and 6 uniface tools/modified flakes. Thirteen points and point fragments were collected including Brewerton eared and side-notched(minimum of 1 each), Bare Island, Rossville, and Madison-like (minimum of 1) diagnostic types along with unidentified Paleo-Indian points, Early Woodland points, and Middle-Late Woodland points. There were six other lithic items (4 hammerstones, 1 chopper, and 1 nutting/grinding stone). The only groundstone item recovered was a pestle. In all, 1,428 fire-cracked rock (FCR) fragments were retained. The lithic materials identified were jasper, flint, quartz, and quartz in decreasing order of frequency. A total of 252 ceramic sherds were also recovered. Diagnostic types included Wolfe Neck, Coulbourn, Mockley, and 1 Dames Quarter and 1 Marcey Creek as well as 225 unidentified Early-Middle Woodland types that were recovered from pit Feature 3. Except for the single Dames Quarter and Marcey Creek sherds that were identified, the site report indicated that all the remaining sherds could be classified as either Wolfe Neck, Coulbourn, or Mockley; therefore, a minimum estimate of 80 of each of the three types was recorded in the table above. The diagnostic lithics and ceramics suggest occupation from the Late Archaic through the Late Woodland period, with the most intensive use possibly in the Early-Middle Woodland period. In addition to the prehistoric materials, 261 historic artifacts were also found. In the original report they were simply cataloged as 114 sherds, 60 coal, and 87 ‘other’ items. The coal and ‘other’ were recorded as miscellaneous items in the table above and the sherds were recorded as ceramics in the table.
The final assessment of the Phase II archeological testing contained recommendations for mitigating the impact of the proposed construction on the site. It was decided that conservation of the resources in place was not feasible. Therefore, it was determined that all areas within the impact zone would be subjected to extensive grading and excavation with heavy machinery. The Memorandum of Agreement (MOA) between the County Commissioners and the Maryland Historical Trust provided that Phase III mitigative activities would consist of soil chemical analysis, subsoil exposure, and archeological excavation of cultural features. In addition, a second controlled surface collection was conducted using the same grid as in the Phase II survey as rainfall and wind erosion had exposed additional material. Phase III data recovery was conducted in the later part of 1980 and early 1981. Systematic soil samples were taken at 5 m intervals using an auger from the areas of potential effect. Soil phosphorous and pH tests were run on the samples. Following the mechanical removal of topsoil from the affected portion of the proposed correctional facility tract, all potential features observed in the subsoil were mapped. The topsoil was not screened. Trenches were then excavated in areas away from potential features in order to expose stratigraphic profiles and provide information on the possibility of buried cultural horizons. In all, 147 soil stains were identified in the mechanically stripped areas. Of these, 68 were archeologically investigated. Each feature was excavated in its entirety and all soils were screened and soil samples were taken from each feature for flotation.
The features were placed into 1 of 4 general categories. Feature Type A consisted of 7 possible prehistoric postholes, a determination based on the recovery of prehistoric material from the fill, and 2 possible historic post holes and molds from which no historic artifacts were recovered. Feature Type B consisted of 13 small prehistoric pits. Six of the pits had fire-reddened soil and 1 was full of FCR. One pit (Feature 147) was filled with pottery, FCR and lithic debris. Feature Type C consisted of 2 large prehistoric pits (Features 57 and 37), the fills of which contained flecks of charcoal and relatively large numbers of FCR, lithics, and pottery. Feature 57 also contained several small animal bone fragments. Feature 37 was a burial pit that had been excavated into a tree fall. The burial pit was cut into by a hearth feature (Feature 155). The grave was excavated to a depth of 55 cm below ground surface. It contained the remains of an adult female in a semi-flexed position. No artifacts were identified in direct association with the burial but Late Woodland ceramics were noted in the overlying stratum. Feature Type D consisted of 37 pits that were determined to be natural tree falls. The investigators hypothesized that these natural depressions were used as loci of activity, for example as trash pits, hence the presence of pottery and lithics within the falls.
The prehistoric assemblage collected during the data recovery total 2,900 artifacts. Counts for the flakes, cores and fire-cracked rock were taken from the artifact inventory in Appendix D of the original report and the remaining counts were taken from text/descriptions in the report. A total of 927 flakes were recorded of which it was noted that approximately 90% had evidence for utilization without modification. There were 199 cores and core fragments. Flake tools (n=36) included 16 preforms, 6 early stage bifaces, 3 choppers, 11 scrapers. The 84 projectile points and point fragments were separated into 9 groups based on morphology: Group 1 was comprised of 18 points that correspond to Bare Island/Lackawaxen; Group 2 was comprised of 8 points that correspond to Bare Island/Lackawaxen and Poplar Island; Group 3 consisted of 12 points that correspond to Levanna; Group 4 consisted of 17 unidentified points; Group 5 consisted of 2 Lamoka, 2 Brewerton, 1 Meadowood, 2 Normanskill, and 3 unidentified points; Group 6 consisted of 2 Otter Creek points; Group 7 consists of 2 unidentified points with concave bases; Group 8 consisted of 1 Perkiomen broadspear and 1 Jack’s Reef Pentagonal; Group 9 consisted of 13 unidentified point fragments. Groundstone items (n=6) included 2 fragments of slate pendants, a third sliver of ground slate recovered during flotation that was thought to be a portion of a pendant, a notched and ground green stone object of unknown function, a shaped grinding stone, and a mortar. Two fragments of soapstone vessels were retrieved. A total of 1,497 fire-cracked rock fragments were recorded. A significant number of prehistoric ceramic sherds (n=207) were retained. This included 16 Marcey Creek plain, 20 Dames Quarter, 31 Wolf Neck cord-marked, 9 Wolf Neck Net-impressed, 21 Hell Island, 11 Potomac Creek plain, 21 Rappahannock fabric-impressed (recorded as Townsend in the table above), and 78 unidentified sherds, many quartz and sand tempered, including 4 rim sherds. In the discussion section in the site report, the investigators mentioned that Accokeek pottery was present at the site; however, it was not listed in the detailed examination of pottery types in the previous chapter or in the Artifact Inventory. The Accokeek sherds may have been some of the unidentified quartz and sand tempered sherds only recognized after the artifact analysis. An estimated 20 animal bone fragments (unmodified faunal) were collected from pit Feature 57 and charred nut hull fragments were reportedly identified in many of the features. In addition to the prehistoric materials, 235 historic artifacts were recovered from the site. Unfortunately, no further information regarding specific classes or types was provided in the text; therefore, they were recorded as miscellaneous items in the table above. It appeared that majority of these artifacts were located in the southern half of the study area. These artifacts were almost certainly associated with the 18th to 20th century Elk Landing Site (18CE60).
An analysis of artifact distributions indicated that the majority of the lithic production occurred in the southeast quadrant of the study area (the area closest to Big Elk Creek and the edge of the terrace), where 47% of the flakes were recovered. This quadrant also contained 64% of the cores, 57% of the FCR, and 46% of the features. The Early Woodland ceramics were located primarily in features in the northern quadrants of the study area.
A 4.11 m (13.5’) core sample was collected from the floodplain-marsh located on the eastern side of the Hollingsworth Farm site. Results of the pollen analysis revealed that available foodstuffs in the area during the prehistoric period spanned a wide range of types, from nuts to herbs and roots. Preservation of pollen in the core was excellent and permitted an extensive identification of genera. The marsh environment further contributed to the preservation of grains. Wood and stem examples included within the sample were not analyzed but were preserved in glycerin. It was determined that the area would have been suitable for human habitation during the colder period of peat formation but that the variety of plant species would have been rather limited. Instead, ideal conditions developed during what the author referred to as Zone B when the presence of an abundance of oak, some hickory, walnut, and chestnut along with berry and herb-bearing plants would have helped to support a gathering economy. Presumably this occurred around the terminal Archaic/Early Woodland period (what the author refers to as the “Woodland Corner”).
In 1985, staff from the University of Delaware Center for Archaeological Research analyzed three assemblages of artifacts from the Hollingsworth Farm site and 2 other nearby sites (18CE30 and 18CE31) that belonged to a local collector. The collections had been carefully recorded, catalogued and provenienced by the collector. This research was part of a larger series of projects conducted in 1985 and 1986 that were initiated by the Maryland Historical Trust. The majority of those projects, including the analysis of the Hollingsworth Farm assemblage, were focused on investigations of the patterns of prehistoric lithic procurement and utilization in Cecil County, Maryland. It was the goal of the analysis to expand the archeological record of the Elk Landing area with the minimum outlay of limited research funds. The analysis involved a three-pronged approach. Using aerial photos and topographic maps general locations of artifact concentrations were identified. Next, each concentration locale was examined and verified during a pedestrian survey. Then, the catalogue of field specific artifact collections was cross-referenced with identified site locations to yield site specific assemblages. The 1980/81 excavated material described above derived from a limited area of the site (the area affected by the proposed construction); however, the material from the collection was obtained through less systematic means but from a more extensive area of the site. When comparing the 2 assemblages this fact should be kept in mind.
In all, 235 prehistoric artifacts were catalogued in the site collection. A total of 223 flaked stone artifacts were examined. The collection did not include a systematic sample of lithic debitage, prehistoric ceramics, or fire-cracked rock fragments. It did contain a significant sample of both damaged and unfinished bifacial tools. There were 64 utilized flake tools, 9 modified flake tools (5 knives, 2 drills, 1 scraper, and 1 chopper) and 20 early stage and 21 late stage bifaces. A total of 109 points were recorded including 4 Kirk (recorded as notch in the table above), 21 small stemmed, 43 large stemmed, 18 notched, 4 Orient Fishtail, 2 unspecified broadspears, 1 Rossville, 1 Fox Creek, 1 Jack’s Reef, 9 large triangles, and 5 small triangles. Despite the lack of lithic debitage in the collection, some assessment of on-site tool manufacturing was possible. Just over a third of the assemblage represented bifaces that were rejected during either early stage or late stage reduction. Cryptocrystalline chert, jasper, and chalcedony comprised 86% of the assemblage, with 34% being of primary quarry material (Cecil Black Flint, Newark Jasper, and Broad Run Chalcedony) and 52% being of secondary cobble material (quartz, quartzite, chert, and jasper). Ironstone was represented in 11% of the assemblage and rhyolite (1%) and argillite (2%) constituted the remaining material types. Groundstone artifacts (n=5) included a heavily worn celt, a bannerstone fragment, and 3 pendant fragments. The 5 other lithic items were all hammerstones. Evidence for continued occupation into the Contact period consisted of a large basal fragment of a heavy green wine bottle that had one edge flaked into a scraper, and a heavily used jasper gunflint , the characteristics of which indicated aboriginal manufacture. The glass scraper and gunflint were recorded as uncommon objects in the table above.
In 2000, the Historic Elk Landing Foundation signed a 99-year lease for the restoration, management, and operation of the site as a living history museum. Archeological survey and testing at Historic Elk Landing was commissioned in order to provide data that would aid in interpreting the historic and prehistoric past in a living history format. Historic Elk Landing is a ca. 44 acre area starting at the foot of Landing Lane and encompassing the land between the forks of the Elk River. The historic structures located on that property include two 18th century dwellings and several 19th to 20th century outbuildings. The total area inclusive of those structures (approximately 6 acres within the 44 acre property) has been designated as site 18CE60, which is located adjacent to the west side of site 18CE29.
Work at Historic Elk Landing was conducted in the early spring of 2002 and consisted of the excavation of 392 shovel test pits and 7 test units (each measuring 1 m²) in an approximately 22 acre area on the terrace above the floodplains of the Little Elk and Big Elk Creeks. Included in the survey was an area of approximately 5 acres located within the boundaries of site 18CE29 as currently delineated by the Maryland Historical Trust. The northern boundary of the 2002 survey roughly corresponded with the southern boundary of the 1980/81 survey discussed above. The project area for that earlier survey was not part of the Historic Elk Landing lease. An estimated 119 STPs, and no test units, were excavated inside the site boundaries. The original report on the 2002 testing did not differentiate between site and non-site STPs so the estimated number of STPs is based on the site map in the text. STPs were excavated at 10 m intervals in the western and southern portions of the project area and at 20 m intervals in a corn field located in the northeastern corner of the project area where artifact densities were noticeably low. In the section of site 18CE29 located within the project area boundaries, the southern third was subjected to 10 m interval testing and the northern two-thirds were subjected to 20 m interval testing. The prehistoric and historic artifacts that were recovered from STPs within the boundaries of 18CE29 have been included here and in the table above. Fire-cracked rock, oyster shell, coal, slag, and brick were not collected; however, they were counted before being discarded and are therefore included in the artifact totals in the table above.
A total of 174 prehistoric artifacts were recovered from within the boundaries of site 18CE29. There were 136 pieces of debitage (16 quartz, 50 jasper, 4 rhyolite, 38 chert, 5 quartzite, 3 argillite, 18 chalcedony, 2 unidentified material) and 1 jasper core fragment. There was 1 jasper biface tool, 4 utilized flakes (2 jasper, 1 quartz, 1 chalcedony) and 1 quartz Bare Island point which dates to the Late Archaic and Early Woodland periods. There were 3 other lithic items (1 hammerstone and 2 pebbles) and 1 groundstone item (a net sinker). Four ceramic sherds, 2 with sand and grit temper and 2 with sand and quartz temper were also found. A total of 23 fire-cracked rock fragments were counted. The majority of the prehistoric artifacts were recovered from the southern portion of the study area on the terrace edge overlooking Big Elk Creek. While smaller groupings of prehistoric artifacts were identified elsewhere in the project area, none were as dense as the concentrations along the edge of the terrace.
As noted above, the historic component from the site likely relates to the 18th-20th century site 18CE60. In all, 354 historic artifacts were recorded. There were 3 activity items (1 horse shoe nail, 1 fishing reel part, and 1 terra cotta flower pot sherd). There were 111 architectural items (35 brick fragments, 26 cut nails, 10 wire nails, 1 roofing nail, 5 unidentified nails, 26 shards of window glass, 3 pieces of flat glass, 1 spike, 1 nut and bolt, 1 window latch, 1 door bolt, and 1 tack). Only a single clothing item, a porcelain button, was recovered. Kitchen-related items made up the majority of the assemblage (n=199) with 62 pieces of bottle glass and 3 fragments of vessel glass, 110 ceramic sherds (1 Jackfield, 8 ironstone, 2 Rockingham, 24 creamware, 7 pearlware, 2 porcelain, 1 blue and gray stoneware, 1 Bristol glazed stoneware, 1 unidentified stoneware, 33 whiteware, 1 yellowware, and 29 redware), 3 bottle crown caps, 4 can fragments, 5 pieces of animal (1 burnt), and 12 oyster shell fragments. There were 39 miscellaneous items (10 pieces of unidentified metal, 2 pieces of plastic, 1 piece of Bakelite, 24 pieces of coal, and 2 pieces of slag). A concentration of historic artifacts was noted in a series of about 10 STPs located just north of the concentration of prehistoric materials.
The Hollingsworth Farm Site (18CE29) consists of Late Archaic to Late Woodland base camp occupations and a prehistoric burial. The most intensive period of occupation appears to have been between the Late Archaic and Early Woodland periods with a less significant presence by the Late Woodland/Contact period. The undisturbed portion of the site has been shown to contain evidence that would be pertinent to questions regarding prehistoric subsistence and settlement models. Given the extensive period of occupation at the site, and at Elk Landing more generally, any ground disturbing work should at the very least be monitored by a professional archeologist. Furthermore, it should be kept in mind that archeological testing has only been conducted within a relatively small portion of the site. As site 18CE29 is adjacent to site 18CE60, an 18th-20th century site, the prehistoric artifacts collected from 18CE60 probably relate to 18CE29 (see the synopsis for 18CE60 for artifact counts).
|
