Archeological Exploration of Patawomeke: The Indian Town Site (44St2) Ancestral to the One (44Stl) Visited in 1608 by Captain John Smith T. DALE STEWART SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY ? NUMBER 36 SERIES PUBLICATIONS OF THE SMITHSONIAN INSTITUTION Emphasis upon publication as a means of "diffusing knowledge" was expressed by the first Secretary of the Smithsonian. In his formal plan for the Institution, Joseph Henry outlined a program that included the following statement: "It is proposed to publish a series of reports, giving an account of the new discoveries in science, and of the changes made from year to year in all branches of knowledge." This theme of basic research has been adhered to through the years by thousands of titles issued in series publications under the Smithsonian imprint, commencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the following active series: Smithsonian Contributions to Anthropology Smithsonian Contributions to Astrophysics Smithsonian Contributions to Botany Smithsonian Contributions to the Earth Sciences Smithsonian Contributions to the Marine Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to Zoology Smithsonian Folklife Studies Smithsonian Studies in Air and Space Smithsonian Studies in History and Technology In these series, the Institution publishes small papers and full-scale monographs that report the research and collections of its various museums and bureaux or of professional colleagues in the world of science and scholarship. The publications are distributed by mailing lists to libraries, universities, and similar institutions throughout the world. Papers or monographs submitted for series publication are received by the Smithsonian Institution Press, subject to its own review for format and style, only through departments of the various Smithsonian museums or bureaux, where the manuscripts are given substantive review. Press requirements for manuscript and art preparation are outlined on the inside back cover. Robert McC. Adams Secretary Smithsonian Institution S M I T H S O N I A N C O N T R I B U T I O N S T O A N T H R O P O L O G Y ? N U M B E R 36 Archeological Exploration of Patawomeke: The Indian Town Site (44St2) Ancestral to the One (44Stl) Visited in 1608 by Captain John Smith T. Dale Stewart SMITHSONIAN INSTITUTION PRESS Washington, D.C. 1992 A B S T R A C T Stewart, T. Dale. Archeological Exploration of Patawomeke: The Indian Town Site (44St2) Ancestral to the One (44Stl) Visited in 1608 by Captain John Smith. Smithsonian Contributions to Anthropology, number 36, 96 pages, 55 figures, 23 tables, 1992.?Excavations by Judge William J. Graham and by T. Dale Stewart of die U.S. National Museum/Natural History from 1935 to 1940 at Potomac Neck in Stafford County, Virginia (site 44St2), produced evidence of a Late Woodland palisaded village (seven concentric palisade lines), enclosing an area of about 1.4 acres (0.56 ha). In the village were circular and elongate houses, plus storage pits, and three mass graves (ossuaries). The village proper dates from pre-Contact times. Two burial pits (one outside the palisade, and one that intruded into the outer palisade line) contained European trade items, indicating use of the site during post-Contact times, probably limited to a short time following die first European visit to the area by Captain John Smith in 1608. Extensive collections of cultural materials and data exemplify what Karl Schmitt in 1952 named the Potomac Creek Focus. Ceramics are predominantly Potomac Creek Cord-impressed types; projectile points are small triangles, usually of white quartz; a majority of the clay pipes are of the obtuse angle type; and there is a diversity of bone tools. Shell beads are numerous, especially in die ossuaries. Burials are predominantly secondary bundle burials, widi a few articulated flexed or extended, and even fewer cremations. Coupled widi evidence of long-term occupation of the site (e.g., six rebuildings of die palisade), there is evidence of evolving pottery types. The earliest ware equates closely widi Shepard Cord-marked type of die Piedmont Potomac valley, and diis evolves into the Potomac Creek Cord-impressed and Potomac Creek Plain wares typical of the early 17di century. OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial copies and is recorded in die Institution's annual report, Smithsonian Year. Library of Congress Cataloging-in-Publication Data Stewart, T.D. (Thomas Dale), 1901- Archeological exploration of Patawomeke: the Indian town site (44St2), ancestral to the one (44Stl) visited in 1608 by Captain John Smith / T. Dale Stewart p. cm.?(Smithsonian contributions to anthropology ; no. 36) Includes bibliographical references. 1. Patawomeke Site (Va.) 2. Woodland Indians?Antiquities. 3. Woodland Indians?Mortuary customs. 4. Stafford County (Va.)?Antiquities. I. Title. II. Title: Archaeological exploration of Patawomeke. III. Series GN1.S54 no. 36 [E99.W84] 975.5'26-^lc20 92-28885 CIP ? The paper used in this publication meets die minimum requirements of die American National Standard for Permanence of Paper for Printed Library Materials Z39.48?1984. Contents Page Preface vi Introduction 1 Part I: Field Work 2 The Geographical Setting 2 Manson's and Graham's Explorations (1935-1937) 4 Multiple Burial 5 First Ossuary 6 Second Ossuary 8 Individual Burial 10 Third Ossuary 10 Fourth Ossuary 10 Field Work by Museum Parties 11 The 1938 Field Season 11 The 1939 Field Season 13 Beginning Exploration of the Fifth Ossuary 17 The 1940 Field Season 22 Back to Exploring the Fifth Ossuary 23 Mapping the Excavations 28 Features of the Site in General 35 Postholes and Structures 35 Storage Pits 37 Narrow Ditches 37 Subsistence Base 37 Part II: Cultural Remains 39 Introduction 39 Aboriginal Artifacts 39 The Ceramic Complex 39 Potomac Creek Cord-impressed Ware 42 Potomac Creek Sand-tempered Ware 46 Minority Wares 46 Rappahannock Fabric-impressed 50 Keyser Cord-marked 50 Shepard Cord-marked 51 Pope's Creek Net-marked 54 Unclassified Sherds 54 Summary 54 Tobacco Pipes: Ceramic and Lithic 54 Other Ceramic Artifacts 63 Lithic Artifacts 64 Bone and Antler Artifacts 66 Shell Artifacts and Fossil Shells 67 Cordage and Textiles 67 European Artifacts: Multiple Burial 68 A. Finds by Carl Manson 68 B. Finds by Judge Graham 71 Finds in the First Ossuary 74 Finds Common to the First Ossuary and the Multiple Burial 76 in IV SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY Part III: Mortuary Complex and Anthropometrics 80 Part IV: Conclusions (A Look Back) 92 References 94 FIGURES Frontispiece A. 1937 aerial photograph of Potomac Neck viii B. 1981 aerial photograph of Potomac Neck ? 1. Map of the Tidewater Potomac River (Wilstach, 1932) 3 2. Three Maps of Potomac River, 1612, 1635, and modern 4 3. Location of Patawomeke on Smith's 1612 map 5 4. Three views of the 1935-1936 excavations by Judge Graham 6 5. Three views of Judge Graham's Ossuary 1 during excavation 7 6. Three views of Judge Graham's Ossuary 2 during excavation 9 7. Judge Graham's 35-foot squares 12 8. Sub-square numbering system employed by author 13 9. Contour map of site, July 1938 14 10. Judge Graham's trenches and finds, squares 5-8 15 11. Judge Graham's trenches and finds, squares 3, 4, 9, and 10 16 12. Judge Graham's trenches and finds, squares 1, 2, 11, and 11a 17 13. Three views of 1938 excavations, showing 5-foot squares and lines of postmolds in trenches 18 14. Three views of 1939 excavations, showing 10-foot wide trenches 19 15. North end of die 10-foot wide trench late in the 1939 season 20 16. View of site from creek-end of wire fence in September 1940 and part of contents of Ossuary 5, just before the close of the 1939 season 21 17. Two stages in die exposure of bones in the eastern end of Ossuary 5 22 18. Ossuary 5 23 19. Three sequential views of Ossuary 5 during 1940 excavation 24 20. Additional sequential views of Ossuary 5 during 1940 season 25 21. Two sequential views of last bones exposed in Ossuary 5 27 22. Outline of Ossuary 5 showing the positions of arrangements of the articu? lated skeletons 28 23. Composite map of excavations, showing trenches and tests by Judge Graham and by the U.S. National Museum parties 33 24. Final map of site, showing major features recorded 34 25. The only restorable vessels from the Potomac Creek site 45 26. Vertical sections of 10 rim sherds of Potomac Creek Cord-impressed type . . 46 27. Potsherds of Potomac Creek Cord-impressed ware 47 28. Rimsherds of Potomac Creek Cord-impressed type 48 29. Additional rimsherds, Potomac Creek Cord-impressed type 49 30. Design motifs of rim sherds, Potomac Creek ware 50 31. Six miniature vessels and ladle of Potomac Creek Sand-tempered ware . . . . 51 32. Eight sherds of Rappahannock Fabric-impressed ware 52 33. Three types of introduced wares: Keyser Cord-marked; Shepard Cord-marked; and Pope's Creek Net-marked 53 34. Twelve pipes from die Judge Graham collection from Patawomeke 59 35. Pipe bits showing the range of variation 60 36. Photomicrographs of latex casts of tiiree pipe bowl decorations 61 37. Designs from on six pipe fragments 62 38. Designs from five pipe stem fragments 62 39. Additional decorations from pipe fragments 63 NUMBER 36 40. Silver dram cup or wine taster recovered by Hanson from multiple burial in 1935 69 41. Schmitt's memory sketch of the "star-shaped ornament" from the multiple burial 70 42. Krauwinckel counters (or jetons) copied from Bernard's 1916 publication . . . 71 43. Rolled-out drawing of maker's mark on trade pipestem 72 44. Bone double-tooth comb from multiple burial 73 45. Two views of four metal buttons from multiple burial 74 46. Two strings of glass beads and the copper chain from multiple burial, and a shell/ bead strand, 1st ossuary 75 47. Two views of flushloop trade bells 78 48. Scissors from multiple burial, as found, in X-ray, and as reconstructed . . . . 79 49. Scissors from 1st ossuary, as found, in X-ray, and as reconstructed 79 50. Four views of a large adult male cranium from the multiple burial 80 51. Drawings of four adult male crania from Ossuary 1 82 52. Drawings of four adult male crania from Ossuary 1 83 53. Drawings of two adult female crania from Ossuary 1 84 54. Cut marks on five post-cranial bones, Ossuary 5 91 55. Three abnormal long bones, Ossuary 1 91 TABLES 1. List of skeletons in fifth ossuary with their individual characteristics 29 2. Ceramics from Museum excavations, 1938-1940 42 3. Rim sherds from Museum excavations 1938-1940 43 4. Body sherds from Museum excavations 1938-1940 44 5. 1988 re-analysis of pottery temper and surface treatments 54 6. 1988 re-analysis of ceramic decorative techniques 55 7. 1988 analysis of cordage shown on ceramics 55 8. 1988 analysis of rims and rim treatments 55 9. 1988 analysis of rim tempering vs. decoration and surface treatment 56 10. Frequency of pipe traits by collection (by Karl Schmitt) 57 11. 1988 re-analysis of clay pipes 58 12. Projectile point types, Stewart and Graham collections 65 13. Types of glass beads (Kidd and Kidd, 1970) 77 14. Measurements and indices of 46 male crania from Ossuary 2 85 15. Measurements and indices of 38 female crania from Ossuary 2 86 16. Measurements and indices of six crania from Ossuary 3 87 17. Measurements and indices of 15 male crania from Ossuary 5 87 18. Indices of 10 female crania from Ossuary 5 88 19. Measurements of faces of four crania from Ossuary 5 88 20. Age and sex distributions, Ossuary 5 88 21. Patiiologies and trauma, Ossuary 5 89 22. Abstracts of cranial measurements and indices by ossuary at Patawomeke . . . 89 23. Patawomeke cranial indices compared 90 VI SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY Preface Several things that culminated in this publication are worth explaining in advance of the body of the text, because otherwise the organization of this report will not be immediately apparent. As it is an archeological operation being reported, the first thing to note is that the field work involved underwent an evolutionary course between the years 1935 and 1940. This five-year period divides naturally into two major parts, marked by the death in November 1937 of Judge William J. Graham. Beginning in 1935, he headed a small group of local avocational archeologists, who thus became the first modern explorers of the site. The second part of the period began in 1938, when I took over the project for three seasons of excavations (1938-1940), aided by a succession of field parties working for the U.S. National Museum (now the National Museum of Natural History). Gradually, during the second period, it became apparent to me that I had not seen important elements of Judge Graham's collection from the site. I realized that it would only be from these specimens, together with the two places where they were found, that I would be able to learn the true age and identities of the site. Especially significant in this connection were a variety of European artifacts, found by Judge Graham and his helpers, that really connected the palisaded village site with the arrival of Captain John Smith in 1608. The work of filling in the details of this approach to the site's history became the objective of the long writing period following 1940. Unfortunately, preparation of this report of necessity had low priority over the years after 1942. From that year until 1961, I served as Curator of the Division of Physical Anthropology, and later as Director, of the National Museum of Natural History, and the demands of administrative work forced me to set aside the report. Upon returning to the Department of Anthropology in 1966, I had hoped to finish the report quickly, but unavoidable demands kept absorbing my time. Finally, late in 1978, seven years after retirement, I once more took up the report and slowly carried it nearer to completion. Final work began on the manuscript in 1982. Almost immediately, I came down with herpes zoster of my left trigeminal nerve and since then have had reduced vision. Although I still come to the office almost daily, my work output has slowed noticeably. As a key element in the overall report, I counted on using the excellent artifact analyses compiled by Karl Schmitt. His work, added to my own report as the project director and as a physical anthropologist, make up the bulk of this report. In the early 1980s, I had been working on the report for some 40 years, and I saw a need to finalize it and ready it for publication. I edited the manuscript for unnecessary or redundant details, and thereby reduced it to the essential archeological findings and what they revealed of the Patawomekes when first met by Europeans. Comments from colleagues pointed out some gaps in aspects of the report and also recommended that the analyses and descriptions be updated, especially with reference to more current researches in the Potomac valley. To this end, Ms. Christine Jirikowic (graduate student at The American University, Washington, D.C.) was hired to review the artifact collection, not only from the 1935-1940 work, but also collections previously made at the site during the past century. Her analysis and descriptive details are added to the report, where indicated, as are the original details by Schmitt. Other elements of the report have benefited from the work of Howard A. MacCord, Sr., who has been active in Virginia archeology since the 1930s. He had known Judge Graham and had had a long and close archeological relationship with Carl Manson. Many other people provided me with help in important ways during the field work and the manuscript preparation. Their contributions are acknowledged at various points in the text. NUMBER 36 The following report incorporates the archeological efforts made at site 44St2, plus a description and interpretation of the results of that work. Because the archeological work was done almost 50 years ago, the report of it is obviously "dated," with many gaps and possible errors, when viewed from today's perspective. However, the data are still valuable and I have been urged by many workers in this field to publish my account. Unfortunately, little if anything more can be added to this study by means of further excavations owing to the fact that much of the site has been subdivided and houses have been built on it. T Dale STEWART Smithsonian Institution Washington, D.C. V l l l SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY FRONTISPIECE A.?1937 aerial photograph of Potomac Neck (lower left) and Marlboro(ugh) Point (right), Stafford County, Va., from 5000 feet (1500 m), looking northwest. Accokeek Creek is on the left, Potomac Creek at the bottom right, and the Potomac River above and upper right. Indian Point, at the end of Potomac Neck, is out of sight at the bottom left. The reported site excavations were crossed by the fence row approaching Potomac Creek to the right of Indian Point. The site of Colonial Marlborough is at the bend to the right of the roadway running down the Neck parallel with the river. (U.S. Marine Corp photograph.) NUMBER 36 IX FRONTISPIECE B.?1981 aerial photograph of Potomac Neck and Marlboro(ugh) Point, Stafford County, Va., looking north: (1) Indian Point (44Stl; 77?18'12"W, 38?21'00"N); (2) site of the palisaded village (44St2; 77?17"56"W, 38?2r07"N); (3) the spring on the beach; (4) site of Colonial Marlborough; and (5) Marlboro(ugh) Point. (Photo courtesy of Va. Dept. of Transportation.) Archeological Exploration of Patawomeke, The Indian Town Site (44St2) Ancestral to the One (44Stl) Visited in 1608 by Captain John Smith T. Dale Stewart Introduction The fact that the fields on Potomac Neck, located on the west side of the great eastward bend in the Potomac River in what is now Stafford County, Virginia, offer much evidence of Indian occupancy has been known to students of American history back as far as Thomas Jefferson (1787). The first professional archeologist to visit the neck and collect artifacts from the fields there appears to have been William Henry Holmes. During the period 1888-1910, he was Archeologist and Chief of the Smithsonian's Bureau of American Ethnology, as well as Curator of Antiiropology in the National Museum of Natural History. His visit was not followed up until 1934 when, according to Manson and MacCord, Sr. (1985), Hugh Stabler and Richard G. Slattery, young amateur archeologists of the Washington area, found the site. Whether Carl Manson, another local amateur archeologist, accompanied them or found his way there independently is not recorded. However, in 1935, Manson told Judge William J. Graham, Presiding Judge of the U.S. Court of Customs and Patent Appeals in Washington, who also searched fields along the Potomac River for Indian sites, that he had uncovered on Potomac Neck some artifacts of European origin in association with human skeletons. Because previously the Judge's hobby had led him to enlist my help in dealing witii Indian skeletal remains he had found on the Maryland side of the river (Graham, 1935), soon thereafter (28 March 1936) the Judge invited me to visit Potomac Neck with him. On that visit, I noted that the Judge had two local laborers uncover disturbances below the T. Dale Stewart, Department of Anthropology, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560. plowline, mostly postmolds, and when these stains were seen to form rows, to follow them. Because the Judge was not keeping detailed records of the excavations and the exposed sub-surface features, I suggested to him that he employ the current archeological procedure of staking out squares along a nortii-south axis, so as to map the site and its features. This the Judge proceeded to do but, as I discovered later, the squares had the somewhat unusual scale of 35 feet (10.5 m) on a side and were identified by an inflexible numbering system. The Judge continued visiting the site at varying time intervals until 10 November 1937, when he suddenly died (Anonymous, 1939:310). In the spring of 1938, Mrs. Graham turned over to the Smithsonian Institution the Judge's archeological collection, field notes, and an incomplete manuscript, titled "Potowomeke." This material bears Acces? sion Number 147,346. About this time, Mr. John L. Pratt of Federicksburg, Virginia, a retired vice-president of General Motors, came to me and expressed the wish that I carry on Judge Graham's work. As a inducement, Mr. Pratt offered to arrange with the new owner of the site for permission for me to continue the digging there. Because no one else on the staff of the National Museum of Natural History Museum, Smithsonian Institution (also referred to herein as the "Museum," the "National Museum," the "U.S. National Museum"), where I worked was familiar with the site, or had expressed interest in working there, and even though I had no formal archeological training, the museum administrators allowed me to take on the project. The field work I conducted at the site from 1938 through SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY 1940 accounts for the broad data base of Part I of the following report. Part II deals with the artifacts and cultural data recovered during the site's exploration. This is largely an edited and updated version of the late Karl Schmitt's M.A. thesis, published in 1965. To this, I have added my analysis of the European artifacts recovered by Manson and Judge Graham. Other cultural aspects deal with community data and the subsistence base of the Patawomeke Indians at the time of Contact. Part III of the report deals with burial customs and variations and with measurements and observations. Part I: Field Work The Geographical Setting The great eastern bend in the Potomac River has long been regarded as the main geographical feature marking the location of the neck of land to the west containing evidence of Indian occupation. This conspicuous feature shares in most character? istics of the tidewater part of the river, e.g., navigability, brackish water, tidal changes, fringe marshes, and abundant aquatic food resources. The 1937 aerial photograph (Frontis? piece A) shows only a limited area because of the low altitude of the flight. This panoramic deficiency is made up for, of course, by Frontispiece B and the plenitude of modern, detailed maps, as in Figure 1. However, mapmaking in the Potomac region goes back to the earliest Colonial times. Figure 2 shows the great bend quite accurately; the 1612 map is by Captain John Smith (1612), created from notes and sketches he made during his 1608 explorations of the Chesapeake Bay region; the 1635 map is attributed to Lord Baltimore. Part of the Smitii map in Figure 2 is enlarged in Figure 3. Smith included names and locations of the Indian towns he visited (or saw without visiting), and many of his locations have been confirmed historically or archeologically. What most of the maps and air views show is Potomac Neck, a western extension of Marlborough (now Marlboro) Point. The Neck is bounded by the Potomac River on the east and southeast and by Accokeek Creek on the west, with Potomac Creek linking the two on the south side of the Neck. Confirming the overall fixing of Potomac Neck on early maps are the many other identifiable features along the Potomac, such as Port Tobacco River, Aquia Creek, the mouth of the Anacostia River, and the location of the Falls of the Potomac. The southern end of Potomac Neck, the main consideration here, rises from the water line of Potomac Creek by a narrow beach and steep banks to a height of from 10 to 25 feet (3 to 7.5 m). The banks, especially in the vicinity of Marlboro Point, reveal Eocene marl deposits, the Aquia Formation, which contain well-preserved fossil shells, especially of gastropods and pelecypods. Countless numbers of these shells occur in indurated beds and weather out along the beaches, usually exposed at low tides. Many of the fossil shells are found in the archeological deposit at Patawomeke. Away from the river and creek banks, the Neck is relatively flat, rising gradually to a maximum elevation of 50 feet (15 m) above sea level. At the Potomac Creek shore, the highest elevation is about 30 feet (9 m), and it is at this elevation that the Indian occupational debris is most concentrated. The soils there are sandy loams, containing light gravels and highly enriched by the accumulated humus, ashes, and minerals that mark long-term human habitational areas. Below the humic topsoils, there is a succession of clays, gravels, and marls overlying the Aquia fossil beds. As Frontispiece B shows, a fence row ran from northeast to southwest toward Potomac Creek, dividing a large cultivated area in the western half of the Neck. Over a considerable area on both sides of the fence and a road paralleling the fence, the soil is distinctly darker, and it is replete with Indian artifacts and animal bones. Holmes had noted this concentration of debris and had concluded that this was the site of the 1608 village of Patawomeke (44St41, Frontispiece B (2)) (Holmes, 1903). According to Manson and MacCord (1985:8) the evidence of human occcupation seen in this location represents the largest prehistoric-protohistoric Indian site in the tidewater part of the Potomac Valley. At the foot of the bank fronting along Potomac Creek, not far from the 1937 line of the dividing fence, there is a good, never-failing spring, as of 1988 enclosed in concrete. This may have been the primary source of potable water for the aboriginal inhabitants, although the presence of other riverbank springs should not be ruled out. Just east of the spring, a gently sloping ravine leads from the site terrace to the shore, thus providing easy access to the spring, as well as to beaches suitable for canoe-related activities. At the extreme southwestern tip of Potomac Neck, the land formation is known as Indian Point. This Point was similarly covered with Indian debris, although today the Point has been completely eroded or graded away. Erosion of the Point had been active for centuries, as was noted in the 1870s (Reynolds, 1881). Little of the Point remained in the late 1930s, and regrettably no archeological work was undertaken at that time. Later, in 1957, Manson tested the small remainder of die Point, and his finds will be published separately (see MacCord, 1991). The impression given by Mansonss finds (and others') is that cultural material there was similar to that found at the main palisaded village (44St2), although with some greater concen? tration of European-derived trade items. From this, it appears that the occupation at Indian Point (44SH) lasted longer into the Historic Period than did that at 44St2. This topic will be discussed at greater length below. NUMBER 36 fi*t?1* 1 ^ * " - * > / \ ^ .ARL INGTON ^ S M H ?, COUNTS V&. o? S# ?i? /J W / FIGURE 1.?Map of the Tidewater Potomac River, showing the main geographic features and the boundaries of bordering counties. (Wilstach, 1932, frontispiece.) SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY PRESENT FIGURE 2.?Three maps of the Lower Potomac River: Left, Captain John Smith, 1612. Center, Lord Baltimore, 1635. Right, author. Note the remarkable similarity of three maps brought to the same scale. The star between Aquia Creek and Potomac Creek (on all three maps) marks the location of Patawomeke. The circle and dot on the "present" (1970s) map marks the location of Washington, D.C.; arrow points to north. Like most tributaries of the Potomac River, Accakeek Creek and Potomac Creek have silted in to the extent that they are no longer the deep streams that in Colonial times could accommo? date sailing ships from England. Aquatic plants now cover much of the surface of the streams and spoil the fishing. The Potomac River flowing by the Neck is nearly three miles wide and is tidal. It is only mildly saline, thus accounting for the lack of oysters there. On the other hand, mussels thrive, and crabs are still trapped or caught nearby. Migrating ducks and geese are plentiful in season and in the Neck uplands, wild turkeys never disappeared. Other game animals found in historic times included deer, bear, beaver, and numerous species of smaller animals. Nowhere else on the Virginia side of the river is there a site of an Indian town as well known and showing as much promise for study as the large one on Potomac Neck. However, by a middle 1930s, more archeological attention had been, or was being, given to Indian sites on the Maryland side, from the Anacostia River southward to Port Tobacco. Prominent in this work were Judge Graham, Mrs. Alice L. L. Ferguson, and the ethnohistorian William B. Marye. Despite many years of collecting of Indian artifacts at sites in the Potomac valley, little was known of the culture history of the region's prehistoric inhabitants. Information as to the Indians of the early 1600s is fairly abundant, based on the writings of early explorers and settlers. However, these data had not been verified nor augmented by archeological researches. It was timely, there? fore, that such researches began at Patawomeke in 1935, while on the Maryland shore, similar work was begun at the Accokeek Site by Mrs. A.L.L. Ferguson (Stephenson et al., 1963). Manson's and Graham's Explorations (1935-1937) As explained in the Preface and "Introduction," when the Judge died in 1937, I took over (in 1938) the archeological exploration of the east side of the cultivated fields along the south end of Potomac Neck. My assumption of management at that time makes it necessary for me now to treat the Judge Graham period work (1935-1937) separately from that of the Museum parties (1938-1940) that followed. Accordingly, I put together in this section of Part I what I regard as the most important items from the record of the work left by the Judge. This is primarily his diary-form record of what he had seen and/or done on each visit to the site. Selected quotations from this record (not available to me until after it had been accessioned at the Museum), all somewhat edited to save space, constitute the indented texts below. The partial manuscript left by the Judge bears the title "Potowomeke," which is one of the numerous spellings of the site name that has survived. The more frequently used spelling (and the one I prefer) is the one used in the title of this report, "Patawomeke." The entries from the Judge's diary (Graham, NUMBER 36 uqliquena >N FIGURE 3.?Location of Patawomeke on Smith's 1612 map from somewhat simplified detail by Arber (1884). At the bend in the river just south of "Quiyough flu" (Aquia Creek), a house symbol near the name "Patawomeck" locates the residence of the chief. 1935) that follow have been grouped according to the main features found and worked on by him and his helpers, beginning with the first such major feature, the so-called "multiple burial." MULTIPLE BURIAL December 1, 1935. Went to Marlboro Point with Manson. He had been digging in corn field on site of Indian village of Potowmack on Potomac Creek and on Nov. 30 had found a burial with two skeletons and some artifacts?some large round shell beads, a small silver wine-taster, brass star-shaped ornament, two bone awls, etc. Arriving at site, which was about 165 feet [49.5 m] from spring on shore and about 50 feet [15 m] in from west side of corn field, found that a ditch about 2 feet [0.6 m] deep had been driven through cornfield northeasterly, ending in rounded pit where burials had been found. Remains were from 2 to 2V2 feet [0.6-0.75 m] deep. Removed balance of one skeleton partially dug out by Manson. Then found another, of a middle-aged man (most of teeth gone). Bones were in a heap, with skull uppermost, badly crushed. Then took out skeleton of infant, (which judging) from teeth (was) about four months old (front incisors about to erupt). This skeleton to right of male. Then found flexed skeleton of young female to south of man. Arms seemingly folded over chest. This woman had fine teeth; wisdom teeth uncut. On top of this skeleton, as if buried with or upon it, was skeleton of young child, somewhat larger than the other. Beads everywhere in large quantities (mostly small shell disks; some of red, blue and green glass). Also found about 40 small bells, like sleigh bells, evidently of copper or brass; two disked copper buttons; one metal counter punched for use as a pendant, one-half of a shell gorget ornamented on inside witii punched pattern; one roll of clay for pots; and one large decorated potsherd. All skeletons were found in a small compass, not more than five feet [1.5 m] square, but were individual burials. Charcoal frequent in small quantitites. Bodies encircled by postholes and near a kitchen midden on north. Saturday, Dec. 7. Cleared out hole and soon struck bones of an infant (had eight teeth in upper jaw) in south bank of trench at extreme bottom, (body lay) with head to SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY east, on side...legs extended. Many beads?disk and cylindrical shell and some longer red beads. Worked lower jaw of deer (with teeth in place) lay on child's breast. A smooth, slate pipe, perfect and blue in color, lay between collar bone and ribs. Also just below this was a small conch shell. (Space between) lower limbs literally filled with shell of various sizes, some larger than any observed previously. Under body was a comb of horn or tortoise shell (fine teeth on one side, coarse of the other) in poor state of preservation, near feet was copper chain in four parts... evidently worn as a necklace. The amount of shell beads surrounding this child was astonishing? perhaps a quart. There were also four round copper buttons. Back of this child was skeleton of adult, flexed, face down, with head to east, arms folded under body. A few disk beads (were scattered among the bones) ? This skeleton was slightly higher than the child, and had broken decorated potsherds above it and about one foot from surface (evidently broken by plows). Back of this, and just above it, was anotiier adult skeleton, head to the west, face down, legs and arms folded under the body. No artifacts with this adult. Just before quitting uncovered part of another burial (child) in south side of pit. December 10th. Some had been digging (during our absence). Skeleton in south (part of pit) had been removed.. . . Burial pit cleaned out is 6.3 feet [1.9 m] east and west and 6.3 lli (feet) [1.9 m] north and south, 2.2 feet [0.6 m] deep at deepest point. Later, when the Judge made a map of his explorations, he recorded the dimensions of the "grave pit" as 8 feet (2.4 m) long on the NE-SW axis, 6 feet (1.8 m) wide, and 2x/i feet (0.75 m) deep. The pit was at Line 5 on the north-south axis (Figure 9). From the account of the first excavation it appears that a shallow pit about 6 to 8 feet (1.8-2.4 m) in diameter was encountered (Figure 4, top) containing at least 10 bodies of adults and children. One skeleton may have been disarticulated and bundled, but the others, except perhaps the children, seem to have been articulated and flexed. Because Judge Graham favored a view that the bodies had been buried individually and presumably at different times, I have refrained from designat? ing this group burial as an ossuary. In retrospect, the large quantity of European objects clearly indicates a date of burial either in the last part of the town's occupancy or after its abandonment. The postholes mentioned by Judge Graham as surrounding the burial pit are not clear as to details or significance. When later the area surrounding the pit was explored more fully, many more postholes were found, often in rows. The arrangement of the holes was confusing to the excavators and hence there are entries in tiie Judge's diary referring to them FIGURE 4.?Three views of the 1935-1936 excavations by Judge Graham: Top, workman late in 1935 or early 1936 on edge of shallow pit where first group burials were found. Lower left, Ossuary 1 after reopening in September 1936. Lower right, Ossuary 2, probably in spring of 1936. (Graham collection photographs.) first as a stockade and then as a house (also sometimes referred to as a "death house"?the "quioccasin house" of the coastal Indians). In an effort to settle this point, the Judge undertook, at my suggestion, following my first visit to the site on 28 March 1936, to make a careful map of his findings. However, as will appear, he never carried this project far enough to reach a valid interpretation concerning the "multiple burial" pit and any adjoining or enclosing structure. FIRST OSSUARY The second important find, discovered on 21 March 1936, was a large and undoubted ossuary. On leaving the site that evening, Manson noted bones around the opening of a NUMBER 36 groundhog hole, about 100 feet northeast of the group burial pit. On examination, these bones proved to be human. Further excavation the following day revealed many skeletons. Work was then centered here and continued as opportunity permitted for over a year, i.e., until April of 1937 (Figure 4, lower left, and Figure 5). One characteristic of this ossuary was its segmentation? groups of skeletons were separated by layers of earth. From time to time the Judge recorded in his diary that he believed he had reached the end of the pit, only to follow this with a statement that further clearing of the earth fill had exposed another layer of bones. The Judge's descriptions of this fill are as follows (page numbers in brackets are from his diary): On south side of pit, removed large quantities of charcoal and several sherds of different pots. Here, I think, the filling came from old camp sites near here and thus contained these pieces of charcoal, etc. Noticed, also, in this part of grave, quantities of various colored earth, in quantities and bunches as if they had been dumped from baskets and had been procured from various sources. Thus, dark swamp soil was found in contact with clay, or light sandy soil, while over the top of the whole were several feet of solid clay of homogeneous texture and appearance. This was due to wash and cultivation, I assume. [Page 53] After clearing out dirt (in south end), found many bones in two layers as sometimes before, about 1 foot [0.3 m] apart. Whole south side of pit showed mixed dirt of all colors. [Page 95] At last, when all bones had been removed, the outline of the ossuary pit was revealed as oval, the long axis running NE-SW. It was 3772 feet (11 m) long, 15 feet (4.5 m) across at the widest point, and 5 feet (1.5 m) deep at the lowest point (to the south). By count of skulls, the ossuary population was placed at 181?men, women, and children of all ages. Although I urged the Judge to save certain parts of each skeleton for census purposes, he did not do this consistently. Therefore the figure given for the ossuary population can be regarded only as an approximation. The Judge's diary also gives some indication as to the mode of burial, both as regards the arrangements of the bones and the placement of any accompanying cultural objects. However, the Judge was more interested in finding artifacts than in observing signs of mortuary practices, and his methods of excavation did not favor such observations. In all fairness to him, it should be pointed out that he had to make small exposures in pits because of sporadic visits to the site and because of the threat of vandalism during his absences. Further, he interspersed digging in the ossuary pit with work in other areas of the site. Secondary burials, consisting of bundles of disarticulated bones, seem to have been encountered frequently, as Judge Graham's diary entries indicate: Found the long bones of the legs and arms extending north-northeasterly direction, overlying miscellaneous bones of the body.... Bodies in small compass usually occupying space about 18 inches [45 cm] or 2 feet [60 cm] in length at maximum. [Page 10] Long bones of arms and legs laid lengthwise, side by side and on top of each other, as if laid there. Looked as if bundled and secondary. [Page 14] These individual burials were placed in no apparent order, but appeared as if in pockets, with intervening layers of yellowish earth, mixed with darker surface earth.... In all cases all bones of the body are present and the bodies occupy but FIGURE 5.?Three views of Judge Graham's Ossuary 1 during excavation. Photograph at top taken in June 1936; others taken in September 1936. (Graham collection photographs.) a small space, perhaps 8 inches [20 cm] in width and 2 feet [60 cm] in length. [Page 15] These burials usually occupy a space of 8 or 10 inches [20 or 25 cm] thick and bones are all compactly folded together. [Page 17] Bodies all together, but not articulated. [Page 19] Bones were distictly bundled here, long bones laid in piles. [Page 38] A male adult lay with head on top of long bones of legs, with ribs and clavicles to one side, as if the body had been buried in a sitting posture, and the head had afterwards fallen over the ribs. Lower jaw was above and to north of skull, while some teeth were at knees. Legs extended to south, but could not find feet. [Pages 43-44] The head was in center, ribs and vertebrae behind head, pelvic bones under head and legs and arms in front of head, extending parallel to each other. [Page 56] Long bones of these latter two skeletons.. .laid in piles, side by side. [Page 60] These skeletons are secondary burials?arm and leg bones laid together in 8 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY piles, with skull lying near in each case. [Page 66] These bones all at 5 feet [1.5 m] level and all together in pocket, with long bones bundled. [Page 69] Burned bones, probably indicating secondary burial, are mentioned as being encountered at three different points in the ossuary. In one case, the burned bones were those of a left foot; in anotiier they were of an infant. Further evidence of secondary burial is supplied by observations on extraneous materials in some skulls. In three cases, the Judge notes sand inside the skull: Three skulls empty, and one partially filled with light colored, sandy soil, differing in quality and color from earth of pit. [Pages 10-11] Skulls empty, except for fine sand which partially fills them. [Page 20] ... skull crushed and filled with soft, fine, yellow sand, not common to that locality. This skull had either been filled with sand manually, or had been formerly buried in sandy place. Much like skulls at Port Tobacco. [Page 60] Another time On removing skull of one man, found 12 disk shell beads in interior, resting on facial bones. [Page 49] Articulated, extended skeletons were also encountered, perhaps in a minority of cases, as indicated by the following diary entries: Body extended lengthwise, north and south. [Page 14] All bodies were together with no evidence of bundling ? They lay with feet in opposite directions, but with skulls touching... while bodies may have been buried at the same time, they were evidently buried here orginially?that is, they were not secondary. [Pages 20-21] Man was on back and [T] was able to see bones of body all in natural position, legs and arms extended fully. [Pages 22-23] In contrast to the Judge's meager descriptions of the bones in the ossuary, he gives fairly full details about the cultural objects found therewith. From the enumeration of these objects another characteristic of this ossuary emerges, namely the presence of considerable amounts of copper and glass of European origin. Copper was found associated with at least 28 skeletons and glass with six. Other European objects found were a pair of scissors, two hand-wrought iron nails, a piece of scrap iron, and three pieces of glazed "tile." Of these only the scissors was found in association with a skeleton; the others were in the earth filling the pit. The copper was usually in form of tubular beads, and more than 60 are mentioned. These are made by rolling thin sheets of the metal and varied in length from 5 mm to 5 cm and in diameter from 2 to 5.5 mm. Other copper beads, seeming few in number, were in the form of solid spheres. The only size given for these is a diameter of 3.5 mm. In addition, there were four badly disintegrated thin copper plates, probably square and small; two small balls (brass?); a few small cone-shaped pendants(?); and a small copper band fitted to a pipe stem. Owing to the oxidation of copper and the spread of copper salts to neighboring objects, some materials and relationships are preserved. These are described by the Judge as follows: ... several?probably six or eight?copper beads, cylindrical, 6 x 22 mm, laid side by side. [Page 18] One man had a plate of copper (on what appeared to be cross-woven bark cloth), very thin.... [Page 19] All these beads (shell, copper, glass) had been strung, several sections being adherent, owing in some cases to the exudation of the copper? [Page 23] Also, near one knee found a single large barrel-shaped solid copper bead, 7 mm in length and of an equal width. This had a part of the original string in it, of two strands, twisted counter-clockwise upwardly. The threads, as was usual, were made with a clockwise twist upwardly, and were made of a grass-like fiber. [Page 42-43] ... found copper stains and then beads... 15 were counted, of three varieties.... These had been strung and were separated by small disk shell beads about 1 mm in thickness. This had evidently been a necklace.... [Page 46] This long bead (copper?32x7 mm) was made by wrapping the plate copper around a reed, which still remains in (place), preserved by the copper salts. A section of the cord is also present. This is about 1 mm in diameter and is the longest I have found. It is composed of six threads, as nearly as I can discover, which are twisted from left to right. The filaments are of plant fiber and are twisted from right to left. This thread is large enough to fill the diameter of the large copper bead and is brown in color, when dry. [Page 61] The glass was in the form of small, crudely spherical beads. The majority of these (19) were dark blue in color and ranged up to 4.5 mm in diameter. Other plain, colored beads about the same size were violet-blue (3), light blue (2) and green (45). Also, there were eight red beads with black centers and one striped bead (red, white, and blue). In addition to these European objects, a considerable amount of material of native manufacture was found in association with the skeletons and to a lesser extent in the earth fill. Shell beads, both disk and tubular, were the most common objects. There were also pipes, shell gorgets, and potsherds. The beads were often associated with children's skeletons. SECOND OSSUARY While the first ossuary was being excavated, Judge Graham had a laborer make narrow trenches following lines of postholes. One trench was carried westward from the site of the initial excavations and through the fence row separating the corn field from a brush-covered lot. Just west of the fence row, on 28 November 1936, a second ossuary was discovered. The bones here were in better condition than those in the first ossuary (Figure 4, lower right, and Figure 6). Hence the latter, being already largely explored, probably received diminishing attention thereafter. Certainly the second ossuary was finished in much less time that the first. It was roughly oval, the maximum dimensions being 38x 11 x4 feet (11x3.3x1.4 m), with the long axis NE-SW. It was broadest and deepest at the northeastern end. By count of the skulls, Ossuary 2 contained an estimated 287 individuals of both sexes and all ages. As with the other burials, it is possible to get some idea of the relationships of the bones and cultural objects from the Judge's diary. He repeatedly observed that the bones were more closely and continuously intermingled than in the first ossuary. Hence, NUMBER 36 fc-^p^*.5K %x^^l FIGURE 6.?Three views believed to show Judge Graham's Ossuary 2 in the process of excavation. Judge Graham in top view; part of an articulated skeleton in bottom view. (Graham collection photographs.) they were more difficult to remove when he approached them from the side, as was his custom. The supposition that this arrangement of bones was due to secondary burial is supported by the following quotations: Many of the long bones of these three burials were alongside and back of these skulls. [Page 71] Bones, including skulls, were intermingled, making it hard th disentangle them. . . . These bones were evidently a secondary burial and buried at the same time. [Page 72] Removed seven bodies.... These were bundled closely in center of pit.... Long bones mostly laid in rows with skulls on top and close together in most instances. [Page 78] Skulls very close together. [Page 80] Skulls were close together, on two occasions so close as to intermingle. [Page 85] All eight skulls were within space of two square feet [0.6 square m], and were superimposed, three touching each other. [Page 90] ... found a large number of skulls close together, five of them touching each other, interspersed with other bones. [Page 93] Removed nine bodies in east end, closely jammed in together? All bones interlaced, not bundled, and exceedingly hard to get out. [Page 96] Many skulls piled in middle touching each other. [Page 100] Long bones piled up on both east and west sides; for instance, on west the socket ends of five humeri protruded from wall all bundled with heads close together. At one place seven skulls were so close together they touched. [Page 111] Further observations of practices associated with reburial are as follows: One skull contained a considerable quantity of green marl. [Page 107] Found one skull burnt over a considerable area on left side?seemed to be an adult. In one place was almost burned through. [Page 111] Began to encounter burned bones. One pile was lying between and under several skeletons which were untouched by fire and in some places intermingled therewith. This pile was evidently there when other skulls were placed in, as the new skulls had caked the burned bones to a solid layer in places. Intermingled with these burned (bones) were burned beads.. .These burned remains covered an area of about 2 feet [0.6 m] in length. [Page 112] In contrast with the evidence of secondary burial are two statements indicating the presence of articulated skeletons: On top (of bone layer) was skeleton of man lying with face to east, on left side. Ribs were in place and arms raised in front of head?all bones there. Legs were in place, but seemed folded at knees, as all bones were above knees. Backbone in place under ribs. [Page 73] In southeast corner skeleton lying slanting upward, head at bottom. [Page 110] The curious statement regarding first of these skeletons, namely that the "legs were in place, but seemed to be folded at knees" suggests that the lower legs may have been folded unnaturally forward at the knees (cf. Stewart, 1941, and Ubelaker, 1974). The distinguishing characteristic of the second ossuary was not the arrangement of bones, but the types of artifacts accompanying the bones. These did not include anything of European origin. Also, copper was limited to three small pieces, which testing proved to be of native origin. A letter in the accession record to Judge Graham from J. A. Scherrer, Chemist, National Bureau of Standards, dated 12 July 1937, states, "The sample of copper you left at the chemical laboratory was analyzed and found to be pure metal. No lead, tin, antimony, bismuth, zinc, nor iron were found to be present." The origin of this sample does not appear, but in view of the date, it probably was one of three pieces found in the second ossuary. The absence of European objects was more than offset by the abundance and variety of artifacts of native manufacture, especially shell beads. At first, the Judge meticulously recorded die number and average size of each lot of shell beads, but later he resorted to general terms, such as "many" and "several," or gave the quantity in terms of ounces, and described the types as 10 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY disk, round, barrel-shaped, etc., qualified by the words small, medium, or large. In general, the beads ranged from minute disks (1-2 mm in diameter) to long cylinders (up to 4 inches (10 cm) long and about the diameter of a pencil) and large spheres (up to 1.4 cm in diameter). There was some evidence in places that they had been strung together, retaining their relative positions until excavated. It is noteworthy that Marginella shells prepared for stringing were common in this ossuary, but were not encountered in the first ossuary. As noted before, most commonly the beads were found with children's skeletons. Another observation, perhaps significant, was the presence of large beads, mostly barrel- shaped or spherical, scattered over the floor of the ossuary. In many cases they are reported to have been embedded in the earth floor under the bones, as if sprinkled there before burial. The quotations indicate that this arrangement was frequent: On bottom of grave found several more large barrel-shaped shell beads. These are almost universally on the bottom, half-buried in sand of bottom. [Page 78] Under the bones removed were a number of shell beads, embedded in sand bottom, averaging about 9 mm by 1.1 cm?27 in all were recovered. [Page 80] About 40 large shell beads on floor of pit on east side, distributed over distance of about 2 feet [0.6 m]. [Page 82] Found 34 round shell beads under remains in east end, in good condition. All on bottom, partly embedded. [Page 92] Part of these (45 large round shell beads) were under a male pelvic bone, coiled as if they had been in necklace. All beads on ground under bones. They were either placed there or fell through bones cm ground. Seem to be quite deeply embedded in ground, so that they must be undercut to get them out whole. [Page 99] In some places beads, both disk and marginella, are in windrows on bottom as if thrown there before bones were put in. [Pages 106-107] Quite a large number of large beads.... Pretty well scattered over bottom, of all sizes. [Page 108] Beads of all kinds found, many so scattered over skulls as to indicate they were thrown there,perhaps as an offering. [Page 111] Other objects encountered were potsherds, broken animal, bird, and fish bones, clay pipes (a few whole), quartz arrowheads, scrapers, and pieces of hematite. Some animal and bird bones had been shaped into implements. INDIVIDUAL BURIAL On 16 May 1937, while Judge Graham was excavating his second ossuary, his hired helper struck human bones in a small test pit about 110 feet (33 m) southwest of the second ossuary (probably in the northeast corner of the square later numbered 4L5). These were not fully exposed until 9 September. Then it was found that, instead of being another ossuary as expected, the bones were part of single individual (a young adult male). Not only was the skeleton incomplete, but some bones had been burned. A partial skeleton of a dog was found in association. Unfortunately, the record does not give the position and state of articulation of either the human or canine skeleton. Because no bones answering this description have been found in the museum collection, either they were not collected, or they were discarded after reaching the museum. This being the case, no further reference to them will be made. THIRD OSSUARY The day after the individual burial was disposed of (10 September), the Judge's assistants located, presumably through test pitting, another true ossuary about 77 feet (23 m) southwest of the second. Smaller than the others, it was estimated to contain 67 bodies?as usual of both sexes and various ages. Excavation was completed on 17 September. The outline was somewhat egg-shaped with the longer axis NW-SE and the greatest width at the northwest end. Dimensions were 18x11x372 feet (5.4x3.3 x i m ) . Like the second ossuary, no European objects were found with the bones. Little even in the way of native manufactured artifacts was encountered: a few small shell disk beads, a number of bone awls, some parts of worked deer antlers, and broken animal bones and pottery sherds. Almost all of the Judge's statements about the position of the bones refer to articulated skeletons. Thus: .. .bones, thus far, are not piled up and bundled as is usual. [Page 116] Bodies are not piled together, and on top of one another, but seem to be laid, side by side, irregularly, on bottom. [Page 117] This was skeleton of adult woman, all parts present. She was lying on right side.... [Page 117] Bodies as before. [Page 119] There are only two references that seem to refer to secondary burial: All parts of bodies do not seem to be present. Few hand and foot bones found. [Page 116] ... found remains of one youth, cremated, lying by itself, under other skeletons. [Page 118] FOURTH OSSUARY The last entry in the Judge's diary?that telling of the completion of the third ossuary?is dated September 17. There is no further record of work at the site until 30 October, when Manson, testing away (north) from Potomac Creek on the west side of, but parallel with the fence row, found a fourth ossuary. The distance along the fence row from the second ossuary was about 105 feet (31.5 m). Between 30 October and 1 December 1937, Manson spent six days clearing parts of this ossuary of an estimated 41 skeletons. From his brief notes, it appears that the bones were poorly preserved. No European objects and very few objects of Indian manufacture (eight sherds, small piece of undecorated pipestem, baked clay ball) were encountered. Some bones were found in positions of natural articulation, but as a rule the parts were disarticulated and bundled in pockets. Burned bones were not noted. Overall size of this ossuary was judged to be 23x13x3 feet (6.9x3.9x0.9 m). None of the NUMBER 36 11 bones from this ossuary were retained for accessioning or study at the Museum. Field Work by Museum Parties THE 1938 FIELD SEASON When I came to the site in June 1938, to take up where Judge Graham had left off, I found the fields overgrown with tall weeds and the Judge's grid stakes in a dilapidated state. This was not surprising, because during the two years the Judge had worked there, farming had been discontinued pending settle? ment of the estate of which the farm was a part. On 27 August 1937, the Judge was advised of the sale of the property and after that date made arrangements with the new owner to renew the lease. The transfer of the lease from the Judge to me had been still more recent. Thus, everytiiing conspired to make unlikely the resumption of cultivation of the site fields. Accordingly, I set about unhurriedly to renew the Judge's layout of squares, make a contour map, and record thereon all evidence visible of the previous digging. RE-ESTABLISHING THE GRID.?A search of the old corn field on the eastern half of the site (east of the fence row and the parallel farm road) revealed the Judge's stake number 10 still in place (Figure 7). Without looking farther, I placed five stakes at 35-foot (10.5 m) intervals (the interval used by the Judge) in a line running north from this point. Then 38 feet (11.4 m) beyond and slightly west of his number 7 stake, I found another of his original, but unnumbered, stakes still in place. On the assumption that these two original stakes represented the north-south axis adopted by the Judge, I shifted my stakes accordingly. Next, I was fortunate to discover in the western field (brush lot) two more original stakes bearing the numbers 5D and 5E. The line of these two stakes made a right angle with the north-south axis at the number 8 stake. However, careful measurement showed the 5D and 5E stakes to be about three feet (0.9 m) short of their true positions. Probably as Judge Graham worked away from the main axis of his grid, he accumulated an error in the positions of his other stakes. Anyone who has attempted to lay out a grid over a large area without adequate help and proper instrumentation will recog? nize the difficulty that confronted the Judge. Fortunately, from my viewpoint, the Judge's recording had not reached the point where such errors were unsettling or beyond correction. Aside from more accurately locating the stakes, the only change made in the Judge's system was in numbering the squares (Figure 7). Instead of following his practice of assigning numbers as required, I adopted a more flexible system, whereby all squares were identified by reference to the north-south axis numbers and by an indication of the distance, in terms of squares, right or left of this axis. Thus, 3R1 represents the first square to the right of the one in the axis row bearing the designation 3. For further clarification it should be understood that, whereas the Judge had identified each square by the number of the stake in the southwestern, or lower left-hand, corner (facing north), I adopted the more common practice of naming a square after the stake in the southeastern, or lower right hand, corner (again facing north). This means that I was able to retain only numbers 1 to 6 of the Judge's squares without a change of numbers. Beyond this, my number 7 designating the axis square number 7 was north of his and my number 6. His square number 7 I made my number 6R1. The rest of the changes should be easily understood from inspection of Figure 7. Aside from allowing greater flexibility in extending the area of exploration, the change in numbering the squares resulted in a shift of the axis line the distance of one square (350 (10.5 m) to the east. By this shift, the intersection of the axis line with the fence and hedgerow was at or near the periphery of the site. In view of the uncleared state of the western half of the site, this was an advantage in charting the areas the Judge had trenched. Obviously, a 35-foot square (10.5 m square) is too large for anything but a general reference. It became necessary, therefore, to have means of designating the 49 five-foot (1.5 m) squares contained therein. As Figure 8 shows, this was accomplished by assigning letters from A to G to the seven east-west rows beginning at the south and going north, and assigning numbers from 1 to 7 to the seven north-south rows beginning at the east and going west. Thus, the subsquare in the lower right-hand (southeast) corner of the 35-foot square (10.5 m square) is designated Al and that in the upper left-hand (northwest) corner is G7. THE CONTOUR MAP.?The next step after re-establishing the grid was to work out the contours of the site, so tiiat the elevation of each stake could be shown on a map. For this purpose, a datum point was selected on the Potomac Creek shore nearest the axis line and at the point of high tide on 16 June 1938 at 10:00 A.M. Using a Brunton compass, the elevation of the first stake was determined, and in turn each of the others. As the resulting topographic map (Figure 9) shows, the highest elevation, 29 feet (8.7 m), represents a slight ridge running diagonally northeast to southwest across the site. On either side of this ridge the area of occupation descends gradually, and probably by not more than four feet (1.2 m), except to the southeast. This means that the creek bank south of the site is around 25 feet (7.5 m) above sea level. RECORDING JUDGE GRAHAM'S EXCAVATIONS.?With the grid system re-created and the topographic map made, I next located on the latter map all evidence of previous excavations. Obviously, surface disturbance is a crude measure of the extent of previous subsurface exposure. Nevertheless, short of re-excavation, this approach was deemed the next best thing for future reference and as a check on the Judge's records. Also, it was desirable to see where those signs of excavation appear in relation to the re-established squares, as well as to the Judge's charts (Figures 10-12). The agreement is fairly good, but, at the same time, the comparison shows that the Judge did not use 12 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY 5E 5L5 4E 4L5 3E 3L5 2E 2L5 IE IL5 5D 5L4 4D 4L4 3D 3L4 2D 2L4 ID IL4 6C 6L3 5C 5L3 4C 4L3 3C 3L3 2C 2L3 IC IL3 6B 6L2 53 5L2 4B 4L2 3B 3L2 2B J 2L2j IB IL2 y ' 6A 6LI 5A j 5LI j 4A 411 I 3A f 3LI 2A 2LI IA ILI - ; ' ? > 1/ *7 / 6 5 4 3 2 1 y y 7 6RI 8 SRI 9 4RI 10 3RI II 2RI IIA -'mi 16 6R2 15 5R2 14 4R2 13 3R2 12 2R2s y y 21 7R3 20 6R3 19 5R3 18 4R3 17 3R3 s 22 7R4 23 6R4 / N i 1 FIGURE 7.?Judge Graham's 35-foot squares extended and re-numbered by author. The Judge's numbers (top ones) apply to the lower left-hand corners of the squares; the superseding numbers of the author (bottom ones) apply to the lower right-hand corners. The corners circled are the only ones found with stakes still in place on 1 June 1938. The heavily outlined areas are those for which the Judge left charts. great care in charting exposed features. This is not surprising in view of the way the digging by the workmen had been carried out (Figure 4, for example). In addition to this problem, there was also the one created by the Judge's own excavations. How could they be avoided in further explorations? As will appear, much subsequent trench? ing by the Museum parties had to be planned to avoid areas previously excavated. SAMPLING BY FIVE-FOOT SQUARES.?Up to this point in 1938, the survey work at the site had required five visits during June and July by me and one assistant, Robert N. Ladd of Washington. Digging was not begun until 8 September. When we returned to the site on that date, after almost two months absence, we found the place transformed. The hedgerow had been cut down, and the field on the east side of the fence plowed. The plowing had broken off or removed many of the stakes so recently and carefully relocated. After correcting the damage to the grid system, we began to delimit the area of Indian occupation by exploring the site from the north. With the help of two hired laborers, we elected to excavate the five-foot (1.5 m) square (numbered Gl) in the northeast comer of each 35-foot (10.5 m) section, beginning at a point judged to be well outside the site and working south along each row of stakes until significant soil disturbances were encountered. The results were completely negative in the first (RI) row, between stakes 8R1 and 4R1 (four squares). As we mapped the squares and features noted, we found it helpful to indicate on the field map the directions followed in exploratory trenching. For this reason, in the following discussions, we will use the word "arrow" and a number keyed to Figure 23 to indicate the path of our test squares or trenches. South of stake 4R1, the land slopes off rapidly to the southeast (to the ravine leading to the spring), and thus did not seem promising. The northeast square in section 6 was also found to be negative. South of this point, the row had been disturbed by the Judge's trenches and hence was not further tested at this time. Moving west of the road and clearing away the brush as we worked, we found the northeastern corner of squares 7L1, 7L2, and 6L2 to be negative (Figure 23, arrows 2-4). These rows were not pursued further at this time. Later, we were to discover that such broad sampling was sometimes straddling occupational features. However, being ignorant of NUMBER 36 13 NORTH m G7 F7 E7 D7 C7 B7 A7 6 6 F6 E6 D6 C8 B6 A6 6 5 F5 E5 D5 C5 B5 A5 G4 F4 E4 D4 C4 B4 A4 6 3 F3 E3 D3 C3 B3 A3 6 2 F2 E2 D2 C2 B2 A2 61 Fl E! Dl Cl Bl Al SOUTH FIGURE 8.?Sub-square number system employed by author. A 35-foot square is subdivided into 49 five-foot sub-squares, showing the numbering system used. this, we next moved to square 6L3 (Figure 23, arrow 5) and there encountered a wide ditch. Perhaps because we were intent on finding postholes, this ditch was not thoroughly checked until much later. The desired postholes (in an east-west row) and a pit were found in square 6L3 (Figure 13, bottom left). Because we were then under the impression that this row of postholes was the outermost feature of this part of the site, we followed it, first eastward into square 5L2 (Figure 23, arrow 6), where it was interrupted by Ossuary 2, and then westward in a circular line through squares 5L4,5L5,4L5,3L5, and 2L5 (Figure 23, arrow 7). At the latter point a maze of postholes was encountered. After much sampling, it was possible to discern in squares 1L4-1L3 (Figure 23, arrow 8) a curving line of postholes that seemed to continue the circle we were pursuing. In the course of further sampling (Figure 23, arrows 9 and 10) we noted other curving lines of postholes in the southwestern section of the site (particularly in squares 1L6 and 0L4) outside, but concentric with, the circle first exposed. The postholes in the outer circle(s) had a fresher look, the contained earth being softer and blacker. On the basis of this discovery, we returned to the northern section of the site and sampled outside the first circle (Figure 23, arrow 11). Again, we picked up (in square 7L2) an outer circle with fresh-looking postholes. However, because of the approaching end of the field season, we were not able to pursue this circle very far in either direction. We did determine, happily, that both circles were continuing east of the fence and road. Yet by this time, the eastern field had been planted in barley, and no further explorations could be carried out there. Finally, we defined the limits of Ossuary 3 (Figure 23, arrow 12). The appearance of the site on 8 November 1938, as seen from the foot of the hedgerow (now cleared) is shown in Figure 13,top. These details should make it clear that by sampling an area no larger that five feet square (25 square feet) (1.5 m square, 2.25 m2) at a time, I was able to get an idea as to the layout of the whole site in a relatively short time and with minimum effort by the laborers. As a result, by the end of the 1938 season, I could see that the outermost circle of postholes on the west side, when projected eastward, could account for the location of many of the features discovered by Manson and the Judge. I deduced on this basis that the diameter of the outer palisade was about 280 feet (84 m). This would make the enclosed area about 61,575 square feet, or roughly an acre and a half (5541m2, 0.6 ha). On the other hand, the diameter of the inner palisade, the postholes of which were less well defined than those of the outer, was only about 175 feet (52.5 m), enclosing an area of about 24,328 square feet (2189 m2), or just over one-half an acre (0.2 ha). The contrast in the two diameters and the appearance of their postholes (the outer ones looked fresher or more recent than the inner ones), gave me reason to think that the town had existed and had expanded over a lengthy, but uncertain, time span. THE 1939 FIELD SEASON To develop plans and make arrangements for the second field season, I visited tiie Potomac Creek site on 15 May 1939, accompanied by Mr. Ladd. By this time the landscape had undergone further change (Figure 14, top). Now the field on the west side of the site had been plowed and planted with grass. Also, a heavy wire fence had been constructed parallel with, but across the road from, the former fence-hedgerow. All this meant, of course, that more of our stakes had been removed. However, I learned that after 1 June when the barley in the eastern field was to be harvested, no further farming operations were anticipated until Fall. On the basis of this information, I decided to set up camp at the site on 15 June and to work intensively as long as funds permitted. The camp was established as we had planned and the field party included Robert Ladd, James E. Gillis, Jr., and myself. Again, two local laborers were employed to move topsoil. The first operation was to re-establish the layout of squares that had been disrupted by the plowing. The process was simplified by the discovery in some cases that the plow had cut off the stakes below the surface. By establishing the approxi? mate location of a stake, a little digging often would locate a broken stake still in place. 14 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY FIGURE 9.?Contour site map, July 1938, combining squares, contours, and areas of exploration by Manson and Judge Graham (stippled). Contour lines not marked represent less than 29 feet. Inset in upper left corner is based on frontispiece; multiple burial site is at Line 5 on the north-south axis. SAMPLING BY 10-FOOT SQUARES.?As experience had shown that sampling by five-foot (1.4 m) squares was leaving to chance the discovery of many features, I decided to change the procedure this year and dig 10-foot (3 m) wide trenches where they could be extended a considered distance. Also, because the eastern field had not been explored in the 1938 season, I decided to begin trenching there. Accordingly, a trench was begun in the northwestern corner of square 1L1 and carried eastward for the length of this square (Figure 23, arrow 13). Beyond this point was an area excavated by Judge Graham. Next, we started a like trench at the southwestern corner of square 2 and carried it northward along the LI line through four squares (Figure 23, arrow 14, and Figure 14, left). From this trench, like trenches were carried eastward and northward (Figure 23, arrows 15-18), avoiding as far as possible the earlier disturbed areas. The latter trenches in turn were joined by a trench of the same width along the western side of the RI line (Figure 23, arrow 19). Besides revealing the expected circular trends in the lines of postholes and "ditches," these excavations clearly delimited the eastern extent of the site. NUMBER 36 15 4 (so S \ l? o ? FIGURE 10.?Judge Graham's chart of his trenches and findings in squares 5-8. The grave pit in square 5 is where work began in 1935. This chart fits the top of the one in Figure 11. Next, further trenching of square 1L1 was undertaken to examine the deep deposit at the edge of the bluff overlooking Potomac Creek (Figure 23, arrow 22). To our surprise, the postholes and ditches extended to the edge of the bluff, under nearly four feet of soil. The rich, but uniform, cultural content of this deep topsoil suggested that the latter had accumulated subsquent to the beginning of cultivation there, probably in the mid-seventeenth century. (For another analysis of this deposit, see Manson and MacCord, 1985). The large exposure of subsoil resulting from the use of 10-foot (3 m) wide trenches permitted and encouraged more thorough investigation of features than had been possible previously. In this way, we discovered that the bottoms of some of the narrow ditches contained rows of shallow postholes. From this, we concluded that these narrow ditches had been dug to facilitate the setting of posts for structures. At certain times of the year, this would have been easier to do than driving posts from the ground surface or digging indivdual postholes. At this stage of the excavations, it seemed desirable to move to the west of the new fence and to carry a 10-foot (3 m) trench northward across the center of the site. Thus on 1 July, I located a starting point for a trench that eventually extended to square 9L3, a distance of 245 feet (73.5 m). The last 35 feet (10.5 m) were completely devoid of features. Up to that point signs of human occupancy varied in intensity, with some squares made up of pale tan soils, as opposed to the midden-laden black soils usually found (Figure 14, bottom right). However, the trench yielded two surprising features: a moat-like ditch in the 7L3 square, and a new ossuary, number 5, at the junction of squares 2L2, 2L3, 3L2, and 3L3. The moat-like ditch was found to run close to and parallel with at least one of the outer rows of postholes encircling the site (Figure 15, top). The black soil in the ditch was full of potsherds, animal bones, and other debris. After this feature was plotted, we realized that it probably had a circular course and that we had encountered it at several points further east, especially in work done by the Judge. Because of its relation to a row of palisade posts, I have likened it to a moat. This ditch 16 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY FIGURE 11.?Judge Graham's chart of his trenches and findings in his squares 3-4 and 9-10. This chart fits the top of the one in Figure 12. may have been a source of earth banked along a palisade to help support the posts. Then, perhaps during the last years of the site's occupancy, the ditch served as a convenient repository for trash. In any event, this ditch, with its distinctive greasyblack earth fill, extending sometimes well over a foot into the subsoil, proved to be one of the most productive features of the site for cultural objects. To continue with the excavations, the next step was to run trenches westward at three points from the north-south trench along the L3 line. These were (1) just south of line 8 as far as L4 (Figure 23, arrow 26), (2) just south of line 7 as far as L6 (Figure 23, arrow 28), and (3) just south of line 5 to slightly beyond L7. These three trenches defined the northwestern periphery of the site, gave a fuller picture of the plan already revealed by the initial sampling, and verified the predicted course of the moat-like ditch and its accompanying line of postholes. In running the last-mentioned trench through square 4L6, a puzzling feature was encountered. The moat-like ditch ap? peared at the expected place (just southeast of stake 5L7 in square 4L6) but the refuse therein was found to extend to an unusual depth?four feet (1.2 m). Included in the refuse was an NUMBER 36 17 FIGURE 12.?Judge Graham's chart of his trenches and findings in squares 1,2,11, and 11a. The excavations in square 1 reflect his interest in the deep deposit near the edge of the bluff. unusual amount of pottery sherds larger than usual. Because on the day it was found an early morning rain had made the soil sticky, the excavator, James Gillis, did not realize at first that the pottery type differed from that usually found in the ditch. When I inspected the work, I noticed that the pottery was not the usual cord-marked, grit-tempered ware but was an incised, shell-tempered type. We immediately checked the possible stratification of the location (Figure 15, bottom). Apparently there had been an earlier, deep pit at this point containing the shell-tempered ware. Subsequently, the moat-like ditch was dug into the upper level of the deep pit. The deeper pit obviously predated the village construction, and this consti? tuted the only clearcut instance of stratification found in the work done at the site. BEGINNING EXPLORATION OF THE FIFTH OSSUARY Exposure of the bones in the fifth ossuary (Figures 23, 24) began on 12 July 1939, in the last week of the second field season. Between that date and 23-24 July, when we broke camp, there were two weekends and two rainy days. Thus only about 5!/2 days could be devoted to this part of the project. In 18 SMrTHSONTAN CONTRIBUTIONS TO ANTHROPOLOGY FIGURE 13.?Three views of 1938 excavations, showing 5-foot squares and lines of postmolds in trenches: Top, view to northeast from creek-end of former fence row on 8 November 1938. Bottom left, postholes and shallow pit near stake 6L3 exposed in five-foot-wide trench early in 1938 season. Bottom right, postholes exposed near stake 3L6 later in same season. the first place, the surface extent of the ossuary pit was determined and a fence built around it to keep out the cattle then grazing in the western field. Rather than remove the bones from the exposed side of a vertical cut, as Judge Graham usually did (Figure 6), I decided to work down from above and expose the upper surfaces of the skeletons before removing them. Although this was time-consuming and required constant protection of the exposed bones, it permitted fuller observation of the skeletal arrangements. The long axis of the ossuary being east-west, the initial NUMBER 36 19 FIGURE 14.?Three views of 1939 excavations, showing 10-foot wide trenches: Top, view to north on 15 May 1939 from creek-end of a new wire fence opposite former fence row. Bottom left, a 10-foot-wide trench being continued northward along ling LI in June 1939. Bottom right, south portion of trench along L3 line in July 1939. Deeper excavation marks location of Ossuary 5; arrow points to site north. exposure was begun at the eastern end and carried westward five feet (1.5 m) over the full width of the pit. When thus exposed, some bones were seen to be arranged in bundles, whereas others were articulated. Both bundled and articulated skeletons had been laid in without apparent regard for direction, seemingly according to the dictates of space. In at least two cases, articulated skeletons were seen to have their lower legs bent unnaturally forward at the knee joints so that the feet were 20 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY FIGURE 15.?Top, north end of the 10-foot-wide trench along line L3 late in the 1939 season. Curving lines of postholes are beyond the outermost palisade and accompanying moat-like ditch. Bottom, deep pit underlying moat in northeast comer of square 4L6 and found to contain Rappahannock Fabric-impressed pottery (Townsend series). Arrows point to site north. against the pelvis and lower abdomen. For a possibly similar find by Judge Graham, see Diary, page 22. Shell beads of various types were the most common cultural objects found with the bones. In one case, a large number of different kinds of beads were found inside a skull forming part of a bundle. This is evidence of secondary burial, as Judge Graham had noted earlier (Diary, page 20). Other evidence of secondary burial was in the form of burned bones, a pile of which was found against the southeastern wall. Again, Judge Graham had noted this feature before (Diary, page 22). Upon removing the bones from this part of the pit, the total depth of the pit below the base of the topsoil was about three feet (0.9 m). Having exposed the bones from above, I soon discovered the need to number each skeleton as it was identified. The simplest method seemed to be to assign a number to each adult skull as exposed. This system needed modification only for fetal or infant skeletons, which almost always were placed on top of adult remains. As a result, the small bones tended in time to sift down through and mix with the larger adult bones. The numbering in such cases was to assign the infant bones the number of the nearest intact skull, usually adult, and to add an appropriate letter (a,b,c,...), depending on how many immature individuals were represented. The advantage of exposing bones from above rather than from the side was demonstrated by one of the first skeletons to be fully uncovered (Figure 16, bottom, Burial 3). The skeleton was one of those found articulated with the legs folded forward at the knees. Numbers 1 and 4 showed the same burial position, although less clearly. A clearer example of this leg arrangement was found at the end of the season along the wall of the pit in the northwestern exposure. Figure 16 (top) shows the lower half of this skeleton. (See also Stewart, 1940, fig. 88). As pointed out on page 22, the Judge had found an example of this burial type in Ossuary 2. Artifacts found with the skeletons are equally noteworthy. In removing the earth filling the skull of Burial 5, an adult female, four cylindrical beads, the largest 70 mm long, were found. Mention was made on page 9 of similar finds by the Judge in Ossuary 1. Beads of large size within a skull are certain evidence of secondary burial. The beads in the case of Burial 5 could have belonged to no. 5, or to one of the two children associated with her, nos. 5a and 5b. Marginella and small disc beads were found with these two children. An unusual artifact found was a three-foot (0.9 m) long piece of pine heartwood, extending east-west on top of the bones along the north side of Burial 4 (Figure 16). It seems reasonable that this piece of wood dates from the placement of bones in the ossuary. It had probably been used to carry one or more bundles or bodies to the pit and then had been discarded with the bones. It is also possible that the wood had been part of a structure associated with any ceremony or practice used in the burial activity. Whatever its meaning, such a find is unique for ossuaries along the lower Potomac River, insofar as I know. Besides articulated and bundled skeletons, the eastern end of the ossuary included disarticulated skeleton number 1, lying directly on a pile of burned bones (Figure 16). Judging from the absence of other signs of fire, the bones, had to have been brought from a cremation site located elsewhere. Several individuals, mostly adults, were represented. As noted above, burned human bones had been found in three of the four earlier ossuaries, as well as with the single burial found. When the 1939 field season ended on 24 July, only 10 skeletons had been completely removed. At that time, the NUMBER 36 21 FIGURE 16.?Two sequential stages (bottom to lop) in the exposure of bones in the eastern end of Ossuary 5. Beginning removal of bundled skeleton no. 1 (bottom) has revealed an underlying pile of burned bones. No. 4 is another bundle. Skeleton no. 3 has the lower legs flexed abnormally forward. The star and arrow in each view locates a pine log; large arrow points to site north. 22 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY FIGURE 17.?Bone exposure in Ossuary 5 after removal of the first skeletons encountered in 1939 in east end. Knee bones in upper right corner provide a means of gauging further progress. Unfortunately, not all of the skulls here yet bear numbers. Arrow points to site north. exposed layers of bones were covered with earth, and the excavation was levelled. THE 1940 FIELD SEASON The incompletely explored Ossuary 5 provided the main reason for our spending a third season at the Patawomeke site. With this in mind and for the purpose of planning the work, I visited the site on 10 May, accompanied by Karl Schmitt. Again the land had changed. Now the western field (the locus of Ossuary 5) was planted in corn (Figure 18, top). Thus we could not re-open the ossuary until the corn was cut around the end of August. For this reason, I decided to use the extra time to learn more about the shell-tempered pottery found in the deep pit in square 4L6. Having noted in Bushnell (1937) the occurrence of similar pottery at sites along the Rappahannock River, I decided to spend the last two weeks of August visiting some of his sites. On 20 August, the camp was again established at Patawo? meke. Our crew consisted of Karl Schmitt, Chandler Rowe, and the writer. One laborer was engaged to join us when we could begin work in the corn field. Using the camp as a base, we made daily trips to promising sites along the Rappahannock. Suffice it to say, we found further evidence of the pottery we sought. Bushnell's site of Nantaughtacund (sites Ex 3, 4, and 5) on the east side of Tobago Bay, yielded the most plentiful sample of this pottery type, (Bushnell, 1937, pi. 13). Other sites produced smaller amounts of the incised, shell-tempered wares, and a few produced pottery identical with that in the majority at Patawomeke. On 26 August, the farmer begain to cut com in the field where we wished to work. As in previous seasons, an occasional stake, although broken, often had the distal piece still in place. After replacing broken stakes, we were able to re-establish the grid and then re-open Ossuary 5. When we had cleared away our backfill and exposed the bones, we found everything as previously charted (Figure 18, bottom). Accord? ingly, from then until 30 September, our main effort was directed toward completing the excavation of this ossuary. Because at times it was inconvenient for three men to work in the pit, Schmitt and Rowe took turns working elsewhere on NUMBER 36 23 ? - ( * - ? ? "%'W* JU H u ^ j j n ^ ^ K 1 "?^S^T? ? ^^?^^^^^^^^l ijw^fllFnTTFl^^ i' i^JjBr ^ f HHH26K^ EaralfiJL J l 1 ,/:>< ' ' / . h ',JS .- , X - J Li < \! ) s ) ^ ?" i r FIGURE 20.?Continuation of sequential views (bottom to lop) of bones exposed in 1940 in Ossuary 5. It is beginning to appear (center) that also numbers 34 and 40 have their lower legs unnaturally flexed. Arrows point to site north. 26 SMITHSONIAN CONTRIBUTIONS TO ANTHRCl OLQfiY workers rigged a tarpaulin over the work area (Figure 18, top). Although this helped to some extent, it also proved to hinder our shovelling and interfered with photographic efforts. As each new area of bones was exposed, a sketch of the layout was made to show the direction of the long bones, any natural articulations, any aggregations into bundles, and the position of all skulls. In addition vertical photographs of the exposed bones were taken from an elevated tripod to provide additional documentation (Figure 18, top). Correlating a sketch with a vertical photograph was an innovation for us in the 1940 season. After the recording was done, the exposed bones were removed. Then the most accessible and obvious skeletal unit, often a bundle, was next addressed. As the removal progressed, notes were made about bone arrangements (including the direction the skull faced) and other details, such as the presence of associated artifacts. One of the first photographs taken (Figure 19, center) shows that when articulated burial 15 was removed, burial 16 was revealed as another articulated skelton with unnaturally folded leg bones. Also, Burial 15, when examined, showed a surprising find?inside the skull was a large nest of the mud dauber, Sceliphron caementarium (Drury), with cells still containing larval remains (Stewart, 1941, fig. 68). Because mud daubers do not build nests underground, this find proves that after death the remains of the individual, an adult female, were allowed to decompose above ground during at least one warm season, as a primary treatment before burial. With the completion of skeletonization, the mud dauber found the skull an attractive nesting place. The nest remained in place when the skeleton was brought to and deposited in the ossuary. This is the first such occurrence I know of such a find. Figure 19, top, taken on 2 September, shows that the bones exposed at the end of the 1939 season had been largely removed. To connect the two periods of work, Figure 18, bottom, shows an articulated skeleton with unnaturally folded legs. This had been fully exposed by 2 September, and the skeleton was given the number 27. At this point, too, about one-half of the ossuary had been exposed. Five days later, 7 September, the work had reached the stage shown in Figure 20, bottom. Comparing the numbers on exposed skulls in Figures 19, top, and 20, bottom, indicates that all but three (nos. 25,28, and 29) had been removed. My notes state that removing the bones was difficult, because the limits of individual skeletons were poorly defined, numerous child or infant remains were intermixed with adult remains, and additional adult skeletons were found articulated and with the unnatural leg bone arrangements. Also, three more mud dauber nest were encountered, one each in skulls nos. 19 and 29 and the third among the long bones on no. 19. This shows that the first such find (with no. 15) was not a fluke. More important by now was the first find of copper in Ossuary 5. The metal showed up in association with cylindrical shell beads among the remains of two children (nos. 26a and 26b) in the form of two badly corroded knife-blade-sj| pieces about an inch long and a half-inch wide. A hole in the end of one suggested tiiat they had been pendants. Between 7 and 15 September, when the picture in Figure 20, middle was taken, three rainy days and a weekend prevented much work. During the period, many visitors, especially from the Fredericksburg area, dropped in to see the work. Also, on 15 September, we hosted a visit by the Virginia Indian Relic j Collectors Club, which in early 1941 was reorganized as the i Archeological Society of Virginia. Neverthless, as Figure 20 shows, the area of exposure of bones had greatly increased. One burial shown (no. 40) is articulated, and certainly shows the unnaturally placed leg bones. Also in Figure 20, top, taken probably on September 24, the southwestward curvature of the pit wall shows that we were nearing the west end of the ossuary. This photograph shows that most skeletons seen in Figure 20, center, had been removed. One of these, no. 38, was found to have a mud daubers' nest in the antrum of the maxillary bone. Ultimately, burials numbered 57 and 59, shown in Figure 20, top, were also found to have mud dauber nests among their long bones. This brought to six the number of mud dauber nests found in this ossuary. The last bones in the ossuary were exposed by 27 September (Figure 21, bottom). This view is notable for showing the second pile of burned bones found in this ossuary. The position of the pile against the southwestern wall of the pit (to the left of skulls no. 71 and no. 72 in Figure 21, bottom) thus matches that of the first pile against the southeastern wall, shown in Figure 16. Among the bones between these two piles, only two small lots of burned bones were discovered, each representing a child, numbers 40b and 58c. Interestingly, the last skeleton removed from Ossuary 5 (no. 77) was articulated, with its lower legs unnaturally arranged (Figure 21, top). Altogether, eight skeletons certainly were arranged in this manner. This is the maximum number of such cases thus far observed in any ossuary along the Potomac River. In nine other cases of articulation of the upper half of the body, the position of the lower legs was uncertain. Figure 22 shows the random distribution of articulated skeletons in Ossuary 5. Figure 22 provides the further information that as many articulated skeletons were on their backs as were face down. Only two were lying on one side. Also, about as many had the cranial end of the axis to the east, southeast, or northeast as to the west, southwest, or northwest. On the other hand, when bundled and articulated skeletons are considered, the axial directions were: east-west?43.8%, north-south?16.4%, and diagonal (northeast-southwest or northwest-southeast)? 39.7%. The north-south arrangement occurred mostly at the ends of the ossuary. Thus, compass direction seems not to have been an overriding consideration in placing remains in the pit The distribution of children's remains in Figure 22 also appears to be random. However, the randomness in this NUMBER 36 27 FIGURE 21.?Two sequential views (bottom to lop) of last bones exposed in west end of Ossuary 5 in 1940. No. 77 is still another skeleton with abnormally flexed lower legs. Arrow points to site north. 28 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY FIGURE 22.?Outline of Ossuary 5 showing the positions of the articulated skeletons so far identified. Bundles of disarticulated bones filled remaining spaces. A number in a box indicates the likely adult with which one or more subadults were associated. instance is likely exaggerated because the numbers plotted represent the well-preserved skull with which the child remains were found associated. (See Table 1 for details on these and other burial positions and associations.) In one such instance, burial 56, a 4-5-year-old child had six younger children in association. Therefore, although some family relationship may be represented, Table 1 also gives my best appraisal of the individuals occupying Ossuary 5. The total number clearly identified is 135, of which 63 are clearly adult and 72 subadults. I judged 34 of the adults to be males and 27 to be females. In two cases I was uncertain of the sex. Agewise, the subadults show the usual high mortality around the time of birth, with a rapid decline in mortality ages thereafter. For details as to cranial measurements and indices, refer to Part III of this paper. In Part III, details of mortuary customs and occurrences will also be presented, using data from all five ossuaries. With this will be limited comparisons with other ossuaries studied in the Potomac Valley. Mapping the Excavations As should be evident now, the work on the Potomac Creek site led year by year from 1935 through 1940 to a growing appreciation of the site's size and its make-up as an Indian town. During the first three years of this period, Judge Graham assembled quite a bit of evidence about the extent of the eastern and southern parts of the site, as indicated by subsurface features exposed in his random excavations. This led me to produce a couple of formal maps when I took over from the Judge in 1938. Both of these maps have been reproduced and dealt witii early in the preceding pages. One is Figure 7, a beginning division of the site into a flexible system of 35-foot (10.5 m) squares or sections. This eased the task of plotting the excavations already done and others still to be undertaken. The second map is Figure 9, showing the topography of the site and the locations of the Judge's trenches. The next step was to fit into Figure 7 the areas excavated through 1937, and again those in 1938, 1939, and 1940. The latter three efforts were subsequently assembled into a composite map, Figure 23, showing the total areas explored, including those of the Judge. In addition, Figure 23 shows by means of arrows the directions in which the Museum parties progressed during each season of work. What Figure 23 does not show are the subsurface features exposed by the excavations. These details are reserved for the final site map, Figure 24, drawn up for me after 1940 by E.G. Cassedy, illustrator in the Bureau of American Ethnology. Because so much detail was involved, the finished map was large (approximately three feet (0.9 m) square), with the center of the innermost palisade area being located about where the NUMBER 36 29 TABLE 1.?List of skeletons in fifth ossuary with their individual characteristics. Age adult adult adult adult lV2yrs. newborn adult newborn newborn young adult young adult Sex F M M M ? 7 F 7 ? F F? Condition of skeleton Art.1 -KB) +(B) +(B) Bundle + - + + + + + + + Orientation of skeleton Long axis NE-SW NW-SE NE-SW E-W ? ? N-S ? ? E-W NE-SW Position of head 2 ? NW(R?) NE(R) W(Ba) ? ? ? ? ? E ? Anomalies and pathological changes - - Cultural associations3 adult adult M adult adult adult adult young adult adult 9yrs. F M M M F F? 7 F? adult adult adult adult 7-8 yrs. adult adult M M F F ? F F +? E-W +(B) + - + + + + + NE-SW E-W E-W N-S E-W E-W ? ? E E N E(L) E ? hypertrophic arthritis of spine; destructive lesion of right acetabulum; Polydactyly of right hand osteitis of tibia, fibula, and radius osteitis of many long bones advanced osteoporosis of parietal fetus adult newborn young adult adult newborn newborn newborn nr. '/2 yrs. 13-14 yrs. newborn 7 F 7 M M 7 7 7 7 7 7 ? - _ +(F) ? 7 7 ? +(S) 7 7 + + + 7 7 7 ? 7 ? NW-SE 7 NW-SE E-W ? ? 7 7 E-W ? ? SE(F) ? SE(F) w ? ? 7 ? E(L) ? ? upper median incisors hypo plastic - - E-W E(Ba) + + + + + ? + E-W ? NW-SE NW-S ? ? N-S ? ? NW ? ? ?(Ba) S osteitis of many long bones frontal bone slightly thickened osteoporosis of parietals scars on frontal in right supra? orbital region; osteitis of long- bones lower legs unnaturally flexed lower legs unnaturally flexed lower legs unnaturally flexed large cyl. beads inside skull; cut marks on humerus marginella and smallest size discoidal beads among bones on top of pile of burned bones and itself superfically burned large and small discoidal beads among bones and under skull position of legs uncertain few medium-size cylindrical beads smallest-size disc beads in skull cylindrical bone beads; medium- size discoidal beads; spherical clay bead; few shells (beads?) of the marine snail, Littorina irrorala Say few small cylindrical beads; mud dauber's nest in skull two cylindrical beads close to skull; lower legs unnaturally flexed position of legs uncertain few large discoidal beads; convex biconical clay bead mud dauber's nest in skull; an? other under femur beneath burned bones smallest-size discoidal beads; small cylindrical beads few medium-size cylindrical beads 30 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY TABLE 1.?Continued. Age newborn 6-8 mos. adult adult newborn adult young adult late fetus newborn 6-8 mos. adult young adult Sex 7 7 M M ? M M 7 7 7 F M Condition of skeleton Art.1 7 ? +(S) ?KB) +? +(F) 7 7 7 Bundle 7 ? + ? ? 7 + + Orientation of skeleton Long axis ? ? E-W N-S E-W? E-W NE-SW ? 7 7 NW-SE NW-SE Position of head 2 ? ? W(R) N(R) E W(R) NE(Ba) 7 7 ? SE ? Anomalies and pathological changes Cultural associations3 newborn newborn nr. 4 yrs. nr. 15 yrs. unusual groove in midline of frontal; ostieitis of longbones left elbow arthritic; old osteitis of longbones marked bowing of tibiae -KB) E-W E(B) adult 5 yrs. late fetus late fetus adult adult adult adult 2 yrs. newborn 13 yrs. newborn newborn M 7 ? 7 ? M M M ? 7 7 7 7 7 ? 7 - +(F) ? 7 ? + + 7 ? + 7 + + 7 + 7 7 NE-SW NE-SW 7 7 NW-S ? E-W NE-SW NW-SE ? NE-SW 7 ? NE(Ba) SW(L) ? ? NW(L) 7 W(Ba) SW(F) ? 7 NE 7 - probably some osteitis localized osteitis in distal third of right tibia anomalous first ribs (seventh cer? vical ribs?) cylindrical beads of various sizes up to 1 inch (25 mm); two blade-shaped pendants(?) of native of copper lower legs unnaturally flexed position of legs uncertain; me? dium-size cylindrical beads; small discoidal beads position of legs uncertain few small shell beads of various shapes; mud dauber's nest in skull; position of legs uncertain numerous cylindrical shell beads of various shapes up to 3A inch many small discoidal and a few cylindrical beads of various sizes in skull; cut marks on neck of condylar process of mandible few small discoidal beads few small discoidal beads string of beads, consisting of cylindrical elements with dis- coidals (2) serving as rondells string of beads, consisting of large discoidal elements, with small discoidals (1 or 2) serv? ing as rondells; right lower leg unnaturally flexed (whole left leg extended over trunk) many small and medium-size dis? coidal beads in skull mud dauber's nest in maxillary antrum many small discoidal beads in and around skull; also one or two worked mussel shells lower legs unnaturally flexed many minute discoidal beads among bones; shell gorget on top of skull burned NUMBER 36 31 TABLE 1.?Continued. Field no. 41c 42 43 44 45 46 47 Age l - lV2yrs . adult young adult 10-11 yrs. adult adult 14-15 yrs. Sex 7 F? F 7 7 F M? Condition of skeleton Art.1 7 - - - Bundle ? + + + + + + Orientation of skeleton Long axis 7 NE-SW E-W E-W E-W E-W E-W Position of head2 7 NE ? ? W(R) W(L) W(R) Anomalies and pathological changes - - - - Cultural associations3 many small discoidal and a few cylindrical beads - numerous discoidal beads numerous cylindrical beads of all sizes up to IV2 inches (37.5 newborn nr. 16 mos. 7 7 -? +(B) +? E-W NE-SW ? NE young adult newborn adult M 7 F? 7 7 adult M E-W ? 7 NW-SE W(Ba) 7 ? NW(L) newborn newborn 3-4 mos. adult adult adult adult 4-5 yrs. 7 7 7 M F? F F? ? ? ? ? +(F) _? _? + ? ? ? +? + +? ? ? ? E-W ? E-W ? E-W ? ? ? W(R) ? E(R) ? E(L?) l ' /2yrs. 3-4 mos. 3-4 mos. newborn newborn newborn adult adult lV2yrs. 2-3 mos. newborn newborn newborn adult ? ? ? ? ? ? F M? ? ? ? 7 ? M ? ? ? ? ? ? +(F) ? ? ? ? ? ? ? ? ? ? ? - + ? ? ? ? ? + ? ? ? ? ? ? E-W E-W ? ? ? ? ? N-S ? ? ? ? ? ? E(L) W ? ? ? ? ? N(R) mm); one specimen of the marine shell Oliva sayana rav one small specimen of the marine shell Busycon contrarium con- rod over lower legs; string of small discoidal shell and cop? per beads (alternating) at neck arms and legs articulated; cut marks on radii, ulnae and tibia portion of the maxillary point (unworked) of a common loon many small discoidal and a few small cylindrical beads in de? pendent part of skull lower Legs unnaturally flexed; few smallest-size discoidal beads spine articulated legs flexed (unnatural ?); many small discoidal and a few small cylindrical beads; a little cop? per (badly corroded); many marginella orginally attached to garment. String of large discoidal beads lower legs appeared to be disartic? ulated; mud dauber's nest under right tibia mud dauber's nest among bones burned 32 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY Age adult newborn newborn adult adult 1-2 mos. newborn newborn adult newborn nr. 16 yrs. adult TABLE 1.?Continued. Condition of skeleton Orientation of skeleton Sex Art. Bundle Long axis Position of head 2 Anomalies and pathological changes Cultural associations3 M? ? ? M F ? ? ? F 7 ? ? ? + (F) + NW-SE 7 7 7 7 NW-SE NW-SE ? ? ? NW-SE 7 SE(Ba) ? ? SE(V) C ? ? ? SE(R) metopic suture nr. 15 yrs. adult nr. 6 yrs. adult adult adult M? M ? M F F + + + + + + NW-SE E-W NE-SW E-W E-W NE-SW SE(Ba) W(R) SW(Ba) E(Ba) W(R) SW(L) tibiae markedly and symmetri? cally curved; especially in fibu? lae osteitis of tibiae and fibulae; two lower incisors are fused M? M N-S N-S S(Ba) nr. 2 yrs. nr. 20 yrs. newborn 3-4 mos. adult newborn newborn adult newborn newborn newborn adult adult newborn 1-2 mos. ? M? 7 ? F 7 ? M 7 ? ? M M? ? ? 7 + + + ? + ? ? 7 +(F) ? ? ? + + ? + ? ? ? ? ? ? N-S N-S N-S N-S E-W ? N-S ? ? ? ? NE-SW ? ? ? S(L) ? ? ? W ? S(V) ? ? ? ?(Ba) SW(R) 7 7 arthritic deformity of mandibular condyles unusually slender bones one of the adults in 71-75 group shows osteitis of the tibiae, filbulae, and a metatarsal in one of the infants the ends of some of the long bone diaphyses are expanded 'Articulated adult skeletons were noted as being face up (i.e., on the back = B), face down (i.e., on the face = F), or on one side (= S). Time did not permit the working out of this detail for children. ^ t was noted that the skull, in addition to being located at one end of the skeletal axis (or near the center of a bundle = C), rested on its right (= R) or left (= L) side, on its base (= Ba), on its vertex (= V), or on its face (= F). For the relation of the face of the skull to its respective bundle see the accompanying text. 3Except where noted beads are shell. vertebrae articulated Numerous medium-size discoidal and a few marginella beads position of legs uncertain traces of copper beads about neck; parts of the maxilla and mandible (unworked) of a com? mon loon a few bones articulated a few marginella beads within skull vertebrae articulated; few small bones within skull cut marks on neck of right femur and near distal end numerous marginella beads; nu? merous cylindrical beads up to 1 inch (25 mm) one of the adults in the group 71-75 shows many cut marks on the long bones position of legs uncertain position of legs uncertain few cylindrical beads position of legs uncertain a mussel shell (worked?) and some cylindrical beads (up to 1 inch (25 mm)) were found among bones lower legs unnaturally flexed numerous marginella beads NUMBER 36 33 FIGURE 23.?Composite map of excavations, showing trenches and test by Judge Graham and by the U.S. National Museum parties. Ossuaries are identified by Roman numerals; ? = multiple burial; grids are oriented N-S/E-W east-west axis line 3 crosses the north-south line L3. This intersection is close to the north edge of Ossuary 5, where the latter is crossed by Une 3 and well southeast of Ossuary 3. A disturbing feature that stands out in the mapmaking is the fact that little of Judge Graham's excavation records can be accurately fitted into the system of 3 5-foot (10.5 m) sections with which he covered his part of the site. For example, his squares 5 and 8 (Figure 10), subsequently renumbered 5 and 51, included features that were re-excavated by the Museum parties and hence appear on the part of the final map as derived 34 L8 L7 L6 L5 l_4 L3 L2 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY L1 0 R1 R2 R3 n 0 H 6 -. 5 n 4 y 3 H 0 * * * ? < FIGURE 24.?Final map of site, showing major features recorded. Ossuaries are identified by Roman numerals; <8> = multiple burial; grids are oriented N-S/E-W from the 1938-1940 excavations. When these features of the two maps, Figures 23 and 24, are superimposed, the squares do not quite match. Nevertheless, it seems better to have the main features correctly placed and to assume that the main disagreement is in the location of stakes. Thus, the estimated limits of the Judge's excavations have been adjusted accord? ingly. Proceeding in a similar fashion, the Judge's map of his squares 9 and 10 (Figure 11), now known as squares 4R1 and 3R1, shows a line of postholes running north-south across them. Our subsequent work revealed that the only line of postholes matching this one was to the east. When these two features are superimposed, no serious distortion of the Judge's other finding results, and the combination of findings is logical. NUMBER 36 35 A similar slight shift resulted in a satisfactory adjustment of the Judge's finds in his squares 1, 2, 11, and 11a (Figure 12), now designated as squares 1, 2, 2R1, and 1R1. The key to this adjustment was the long narrow ditch crossing his square 1, and because the Judge's digging there was unsystematic, it is hardly likely that all subsoil distrubances were seen and plotted. The large area included within the subsequently numbered squares 3L1,4L1, 5L1, and 5L2 remains to be considered. The Judge left several versions of his map of these squares that I have had to combine. Fortunately, his findings can be located on the final site map (Figure 24) with a fair degrees of accuracy because they connect with Ossuary 2, the position of which is the same on all maps. In digging these squares, the Judge was probably not following rows of postholes as his map implies; more likely he was encountering a scattering of postholes in every exposure of subsoil, and it would have been impossible for him to detect the real pattern without a larger exposure. The first four ossuaries, opened by Graham and Manson, can be located on the final site map (Figure 24). The exact location of Ossuary 1 was not verified beyond noting the position of the surface evidence. All of the others were checked. In addition, it was apparent that the Judge had dug test pits at random over the site. When one of these was encountered in our subsequent work, care was taken to ensure that any subsoil disturbances were not due to this cause. Features of the Site in General In addition to the ossuaries (five in all) and the individual and multiple grave, all described in previous pages, the work revealed numerous patterns of postholes, plus many which seemed random. Other features recorded were pits, basins, and ditches of various sizes. The final map, Figure 24, shows most of these features, and they are more fully described below. POSTHOLES AND STRUCTURES The number of postholes encountered was so great that no attempt was made to measure each one individually. Mostiy, they ranged from 3 to 5 inches (75-125 mm) in diameter, the bottoms were pointed, and they penetrated 6-8 inches (15-20 cm) into the subsoil. Experience soon taught each member of the work parties to recognize postholes and to dismiss imitations thereof due to roots, rodents, or other causes. As already noted, postholes of the innermost palisade, encountered mainly in the 1938 season, were indistinct because the soil in them was only a little darker that the surrounding subsoil and only a little less hard. In contrast, the postholes discovered later in the more peripheral locations were much more distinct and for the opposite reasons. During the 1939 season, more of the faded (old?) postholes were seen in the following squares: 4-D6, D7, E7 outside the trench 5-F6, G6 in line but not in trench 5-F7, G7 in semicircle Occasionally, a deep hole was encountered that did not appear distinguishable from the usual posthole. Because their diame? ters were too small to class them as storage pits, it is likely that they had held extra-large posts, or they had served as smudge-fire pits or caches. Evidence to confirm either such latter use was not recorded. The major patterns of postiioles make up a series of concentric palisade lines, with adjoining bastions. These patterns seem to equate with seven individual defense lines, probably no two of them in existence at the same time. If this interpretation is valid, the sequence indicates that the village defenses were rebuilt six times. Since posts set in most Virginia soils are exposed to termites and other borers, the life expectancy for such post structures is only about 10 to 15 years. The indicated sequence of seven palisades thus indicates a village occupancy of from 75 to 100 years approximately. This possibility will be discussed later in this paper. Since the entire concept of palisading a village implies a defense system, we lump these together as "fortifications." It is difficult to improve upon the descriptions provided by Karl Schmitt in his 1942 M.A. thesis, from which we quote the following (Schmitt, 1965:6). FORTIFICATION.?The site of Patawomeke was heavily fortified by a system involving oval stockades and a ditch or moat concentrically arranged. Stockades were of two types: (1) with posts set or driven into individual holes, and (2) with posts set into small trenches. In the first type, the posts were three to five inches [75-125 mm] in diameter and simply set or driven into the ground, rarely penetrating more than a foot into the subsoil. Of course there was a 6 to 12 inch [15-30 cm] layer of plough-disturbed soil overlying the subsoil, and this would have increased the depth of the placing of the posts. In the second type, posts of the same size were placed in small trenches between six inches [15 cm] and one foot [30 cm] in width and of the same range of depth. These trenches were predominantly U-shaped in vertical section, but were occasionally V-shaped. Postholes occurred in some of the trenches but were very shallow, rarely over four inches [10 cm] in depth. In both types of stockade, posts were generally regularly and widely spaced, the intervening distances between posts averag? ing between six and eight inches [15-20 cm]. Three stockades, represented by individual posts, were more consistent and strongly indicated than the others. The innermost one was of oval shape with the longest diameter measuring 175 feet [52.5 m]. Possible entrances were indicated on the north and southwestern side where the lines of post holes appeared to overlap for a distance of 20 to 30 feet [6-9 m]. On the west, the stockade pattern was interrupted by a bastion-like arrangement of postholes, rectangular with rounded corners in shape, and approximately 18 feet [5.4 m] long and 10 feet [3 m] 36 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY wide. There was some indication that the posthole pattern of this feature continued in two parallel tines some 50 feet [15 m] in length toward the center of the village. Unfortunately the termination of this arrangement was not located. The most logical explanation for this feature would be that it was a defensive bastion, although there is a strong possibility, also, that it was the remnant of a long, arbor-like house which was left jutting out of the inner stockade. The second concentric stockade, again marked by individual postholes, was appproximately 200 feet 60 m] on the longest diameter. The pattern of postholes was not as completely followed as in the previous case though it was located in all expected places, except two on the west where a definite break occurred. Outside the second stockade, at a distance of two to five feet [0.6-1.5 m] on the north and only a few inches on the south, was a ditch or moat. This feature quite consistently measured five feet [1.5 m] in width; in depth it ranged between one and two feet. In vertical section it tended to drop sharply on the inner side and to slope very gently to the bottom on the outer side. The eastern quarter of the village apparently lacked the ditch, though there is the possibility that Judge Graham's excavations obliter? ated the evidence in that sector. A short break of 10 feet [3 m] coincident with the break in the median strong stockade, was present on the western side, undoubtedly marking the entrance to the inhabited area. No evidence of heaping of dirt on either the inner or outer side of the ditch was encountered, though it is probable that such a feature, if it had existed, would have been erased by plough action. The fill within the ditch was consistently a homogeneous rich, black midden-like material. It was from this refuse that the great majority of the artifacts and almost the totality of pottery specimens was recovered. Some evidence of a slight fill by wash was encountered in the bottom of the ditch. This was of a very light color, closely approaching that of the subsoil, and rarely, if ever, was over an inch [2.5 cm] in thickness. The midden-like material gave the impression of having been deposited within a very short period. Sherds of the same pot were occasionally found lying 40 feet [12 m] apart and sherds of one restorable pot were recovered from diametrically opposed portions of the ditch. A third strong stockade pattern, again composed of individual postholes, was found outside the ditch, and was approximately 280 feet [84 m] on the longest diameter. The total extent of this feature is not known, but it was found in all sections of the site excavated except in one of the western trenches. Four bastion-like patterns were found along this outer stockade, one on the north and three on the east. The northern example consisted of double series of postholes, the outer conforming to a rectangular shape with rounded corners and dimensions of nineteen feet by 18 feet [5.7 m by 5.4 m], the inner conforming to a more semi-circular shape with a 15 foot [4.5 m] diameter. The bastions on the east overlapped and very likely represented replacements after decay. The median one consisted of a line of single holes, elliptical in shape with measurements of 15 feet by 25 feet [4.5 m by 7.5 m]. Just to the north and partially intersecting the median bastion was a rectangular arrangement of small trenches approximately 25 feet by 15 feet [7.5 m by 4.5 m], indicating the position of anotiier bastion. To the south and again intersecting the median bastion was another pattern consisting of a line of single postholes. This last feature was not completely excavated but appears to have been rectangular with dimensions of fifteen by twenty feet [4.5 m by 6 m]. Unlike the innermost stockade, the outer stockade was not interrupted by the bastions but was continuous past them. Between the innermost and median strong stockades were evidences of weaker elements, four or five showing in every lengthy excavation trench. These features were small narrow trenches containing postholes in the bottom although occasionally they were represented by lines of single postholes. The precise nature and arrangement of these patterns was not evident. They were not continuous and they varied in number in the different excavation trenches. It would seem, however, that they represent short fortification elements set between two major stockades, possibly to create a maze-like entrance to the inner village. A smaller fortification element was found in the interior of the village, encompassing two large ossuaries. This was a stockade set in a small trench and was circular with a diameter of approximately 100 feet [30 m]. In connection with this, the following quotation can be cited from Beverly (1705:149), who in speaking of the Virginia Indians says: Their fortifications consist only of a palisado of about ten or twelve foot [3 or 3.6 m] high, and when they would make themselves very safe they treble the pale. They often encompass their whole town; but for the most part only their Kings' Houses, and as many others as they judge sufficient to harbor all their people, when an enemy come against them. They never fail to secure with their palisado, all their religious reliques and remains of their Princes. Schmitt postulates at least two periods of building of stockades. He thinks that the inner and median palisades, strengthened by an inner maze, were in use together, and that later the outer palisade was built. This would render superflu? ous the moat-like ditch, thus inviting the occupants to throw their rubbish into it and thereby fill it. He also provides a relative temporal placement for the Historic period Multiple Burial pit, although he called it an "ossuary" in the following quote: "The large ossuary [sic] from which Judge Graham obtained the great bulk of the historic material interrupted the NUMBER 36 37 inner palisade pattern and is presumably later in time" (Schmitt, 1965:8). Accompanying the above descriptions of stockade elements by Schmitt, he provides a schematic map, his figure 3, which gives an overview of the palisade relationships. These are shown in greater detail in Figure 24, where site layout and the arrangements of structural and other features, as they were mapped in the field, are more accurately shown. COMMUNITY PLAN.?The palisades, whether single or multiple, enclosed a cluster of other structures. Among these were houses of two types, and possibly storage structures. As with the fortifications, it is appropriate to quote from Schmitt's (1965) descriptions, as follows: No definite house type was discovered on the site. Possibly the bastion-like swelling on the western side of the inner stockade represents such a feature. If this is the case, the structure of fifty or sixty feet [15 or 18 m] in length and twenty feet [6 m] in width is indicated. A partially excavated circle of post holes, 20 feet [6 m] in diameter, occurs in the very center of Patawomeke. This possibly is the pattern of one of the conical type of houses but could equally as well represent a circle of dance posts such as is pictured in Beverly (1705). As can be seen in Figure 24, several rows of postholes and/or shallow ditches form alignments, which may be portions of rectangular, arbor-like structures. The number of such wig? wams that can be identified in clusters totals only three. The circular pattern of postholes noted by Schmitt could easily be seen as the residence of an important person in the site, due to its prominent location, as well as to its shape, which differed from the long houses used by most of the inhabitants. If this circular pattern predates or postdates the site then it is possible that the center of the site was an open area. From the evidence at hand, this question cannot be resolved, altiiough future work at the site might produce the evidence needed. STORAGE PITS Storage of surplus foods and other materials would have been an important aspect of village life. Such items could have been stored in the rafters and along the walls of the houses, or special storage houses may have been built, using shallowly set posts, for which no evidence has survived plow destruction. Another storage method would involve pits of varying sizes, similar to the common potato-cellar of historic farmers. In the work done at this site, only a few pits were found, indicating that pit storage was not a widespread custom. Several pits are noted or are described in the Judge's diary, as follows: about 50 feet [15 m] north from Square 1, encountered a fire or pit hole, round and dropping below surface of ditch (top of subsoil?) for about one foot [30 cm]. Full of all sorts of waste material. Pit lined at bottom with quartzite and quartz rocks, quartizite predominant and mostly broken. Much charcoal and a s h e s . . . . Inter ior of pit soft and digs eas i ly . , .deer horns are those of adults...shell of water turtle.. .many land tortoise shells in entire condi t ion. . . s t ra t i f ied sandstone pieces (slabs?) approximately 1/2 inch [1.25 cm] thick doubtless used for cooking purposes... large fossil shells, one. . .8 inches [20 cm] in length.. .am of opinion these were used for d i shes . . .many screw-shaped fossil shells (Turritellas). . . large num? ber of dorsal spines of f ish. . . . [Page 48] This pit was oval, 5 feet [1.5 m] in length, 3'6" [1 m] in width and two feet [0.6 m] deep from surface. Contents... as follows: (partial list) shells, oyster mussel fossil garfish jaw bone bead (bored) currier (beamer) beaver mandible abrading stone worked antler 10 43 7 1 1 1 1 1 4 potsherds deer bones small clay dish hammerstone anvil stones quartz scrapers points and fragments shell disc ornament bone awls 322 194 1 4 2 7 1 7 2 [Pages 50-51] At south end, in Square 1, fire pits are very deep?4 feet [1.4 m]?with much evidence of long occupied camp site. Much charcoal and ashes. [Page 54] a post hole [sic] was found, about 3 feet [0.9 m] deep, 2 feet [0.6 m] in width, and full of ashes, charcoal, soapy deposits, many potsherds, five bone awls, one large deer horn, one broken clay pipe, a cut deer bone, some miscellaneous pieces of deer, bird, and turtle bone and shell. [Page 116] Pits such as those described above had probably been dug for storing materials below ground, and when emptied, they had been used for the disposal of refuse. Evidence for hearths in and around wigwams was not seen or recorded. Because such features were often on top of the ground, it is likely that they had been completely obliterated by plowing. Some fire-pits, though, may have been dug for use as smudge fires for tanning hides, and some of the larger postholes may have served this purpose. NARROW DITCHES Narrow ditches ranked second to postholes in frequency of occurrence. Besides being narrow (4-8 inches (10-20 cm) across as a rule), tiie ditches had fairly straight courses for considerable distances, e.g., Figure 14, bottom left. Their sides were steep and their bottoms V- or U-shaped, sometimes with shallow postholes at the base. There is thus some reason to believe that all may have served to support posts. A majority of the ditches paralleled in a rough way the circular palisades, as defined by rows of postholes and the moat-like ditch. The contents of the narrow ditches rarely included cultural debris. Judging from the alignments shown in Figure 24, the narrow ditches seem most likely to have supported the side or end walls of rectangular wigwams. Subsistence Base Although no vegetal food remains were noted in the excavations, the presence of smoking pipes hints strongly that tobacco was cultivated. No remains of agricultural products were found, although the ethnohistorical record is replete with references to the abundant supplies of corn to be had through trade at Patawomeke. No tools that may have been used in agricultural work were found, either. However, it can safely be deduced that agriculture was an important source of food for 38 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY the villagers. This was supplemented by wide-ranging hunting, fishing, oystering, and gathering. Gathering no doubt included nuts, berries, greens, and fruits in season, although no remains of these foods were found or recorded. Much charcoal was noted in refuse accumulations at the site, and no doubt many nut hulls and possibly other plant parts would have been found in the charcoal, had an effort been made to sort or fine-screen this material. Judge Graham notes one charred hickory nut shell in midden (Diary, page 33). Hunting for game probably produced most of the protein foods sought. Hunting techniques would have included solitary hunts, group "surrounds" including the occasional use of a fire-drive, and routine use of snares and traps. Most small game was no doubt taken through trapping, whereas the killing of a deer or other large game animal would have been an infrequent occurrence at best. The results of successful hunting are reflected in the refuse bone found in the midden, as well as in the bones used as tools or other manufactured products. The species identified areas follows: Deer Odocoileus virginana Raccoon Procyon lotor Beaver Castor canadensis Bob cat Lynx rufus Muskrat Ondatra zibethica Gray squirrel Sciurus carolinensis No doubt any animal encountered would have been captured or killed, because this pattern has been identified at most late prehistoric sites. Sampling size and the vagaries of preservation could account for the somewhat short list shown above. Birds of all sort would also have been taken, although only the following species were identified from the refuse bones and artifacts: Turkey Red Tailed hawk Meleagris gallopavo Buteo jamaicensis borealis The absence of migratory waterfowl is conspicuous, as is the now-extinct Passenger pigeon. Reptiles, being comparatively numerous and easy to catch, served as an additional source of proteins. Of those which live in the area of Patawomeke, only the turtle varieties have been identified in the village midden. The common box turtle (Terrapene Carolina) was the most abundant species recog? nized. In addition to the refuse bones and shells found in the midden, the carapace was frequently trimmed and cut to serve as cups, dippers, ladles, and so on. The other species found was not identified at the time, but was referred to as a commmon water turtle. The broad Potomac River and its numerous tributaries provide an inexhaustible supply of fish, Crustacea, and shellfish. Bones and shells of these animals were plentiful in the midden and in the various pits excavated. Oysters do not occur in the river immediately adjacent to the site, but they do occur plentifully about 10 miles (16 km) downriver in the area of Pope's Creek, Maryland and Mathias Point on the Virginia shore. A few hours in a dugout canoe would suffice to obtain these foods, and the numerous oyster valves found in the excavations attest to the exploitation of this food source. On the other hand, freshwater clams (mussels) occur at the site and were heavily exploited. Some of the mussel shells were also used as spoons or small knives, sometimes with saw-tooth edges filed into one edge of the shell. Crabs were caught in quantities, and fragments of their carapaces, especially the claws, occured in the midden, usually charred to some degree. Finned fish provided much food for the villagers, and their bones occur frequently in the midden. Species identified are Sturgeon Garfish Yellow perch Catfish Sucker Acipenser sp. Lepisosteus sp. Perca flavescens Ictalurus sp. Catostomus sp. Colonial records indicate that the sturgeons sometimes attained a length of nine feet and a weight of over 200 pounds, thus making this fish a major game animal and an important food source. Although not normally considered a game animal, the domestic dog (Canis familiaris) was available as an emergency food. No dog bones were recognized in the midden, although one skeleton of a dog was found in the pit with an individual human burial. The role played by dogs in the day-to-day life of late prehistoric Indians has not been defined. They may have assisted in hunting or tracking, and they may have been effective at giving the alarm when a stranger approached the village. In addition, they no doubt were effective scavengers, thus aiding in keeping the village habitable. The excavations conducted by Judge Graham and by the several Museum parties produced much data about the site, its occupants, and their uses of the environment. Important, too, are the data concerning the burial customs of the Patawomekes. These data have been covered in some detail in the foregoing Part I of this study. Some of the topics will again be treated later in this work. One result of the excavations not yet described is the accumulation of artifacts. These will be described and analyzed in the following Part II. NUMBER 36 39 Part II: Cultural Remains Introduction My original intent for this report was to edit and publish in Part II the portions of Karl Schmitt's 1942 M. A. thesis dealing with the Indian artifacts from Patawomeke. Accordingly, a year before Karl's tragic death on 8 August 1952, I asked him to revise this part of the thesis. He complied in his usual cooperative manner, leaving it to me to make any necessary or desirable editorial adjustment. In doing these I had many helpful suggestions from Margaret Blaker, then employed in the National Museum's Division of Archeology. Margaret's capabilities in this field were exemplified in her excellent report on the ceramics from the Townsend Site near Lewes, Delaware (Omwake and Stewart, 1963:14-39). As time went by, however, the thesis was published in full in the Archeological Society of Virginia's quarterly bulletin (Schmitt, 1965). There it has served as an interim source of data on Patawomeke and the archeological work done there. As a tangential action, Schmitt had used his data from Patawomeke and from the Maryland site of Moyaone to define what he named the Potomac Creek Focus (Schmitt, 1952). A present- day weakness in the thesis as published is the absence of references to work done in the region since 1952, especially in die definition and distributional studies of the ceramics. Similarly, Schmitt's classification of projectile points is no longer current, and these needed to be updated. Another factor worth mentioning is that Karl had not attempted to analyze the horizontal distribution of the ceramics within the site. Doing so might have helped in dating the several palisade lines and their related structures. Because many of the potsherds sent to him at Chicago have been discarded, it is highly unlikely that such a study can be made now. What seems likely is that most of the extant sherds are from the periphery of the site, e.g., from the moat-like trench, and therefore may date from the closing decades of the village occupation, when the decorative elements seem to have reached maximum elaboration. The illustrations for this revised portion of Schmitt's thesis are new, although they include many of the same specimens used originally. The new pictures were made when prints of the original (1942) illustrations could not be found. The earlier photographs carry the USNM negative number 34,260, and the new pictures are numbered 41,414. During the summer of 1988, when this report was being edited and prepared for publication, the extant artifact collection from Patawomeke was re-examined to (1) confirm Schmitt's findings; (2) provide insights developed through archeological work in the Virginia-Maryland area since 1940; and (3) include, when feasible, data from other, earlier collections in the National Museum that had not been available to Schmitt. The re-examining was done by Christine Jirikowic, then a graduate student at the American University, Washing? ton, D.C. She was hired for this task by the National Museum, and the work was done under the overall guidance of Douglas Ubelaker, with help from Howard A. MacCord, Sr. A retired archeologist, MacCord had done work at Patawomeke and at numerous other sites in the Potomac-Rappahannock River valleys. Further, he had known Judge Graham and had worked extensively with Carl Manson. To a great extent, the following descriptions of artifacts and the tabulations and relationships are based on and quote from Schmitt's excellent M.A. thesis work. Where new or amended data are at hand, these have been added as indicated and credited to Ms. Jirikowic or to H.A. MacCord, Sr. Artifacts were found throughout the excavated area, occurring in the plowed topsoil and in pits and ditches penetrating the subsoil. Adjacent to the bluff fronting on Potomac Creek, some materials were found in the accumulated midden, sometimes as deep as four feet (1.2 m). Because of the high concentration of shells, bones, and wood ashes in the soil, bone preservation was normally good. The artifacts collected in the 1935-1940 work consisted of two main groupings: those of aboriginal manufacture, usually ceramic, lithic, bone, or shell; and those of European manufacture, usually glass, copper, or iron. To follow the usual conventions of archeological publishing, the aboriginal arti? facts will be described and analyzed first, beginning with the ceramic. This will be followed by sections on the lithics, bone, etc., and finally by the artifacts obtained through trade with Europeans. Aboriginal Artifacts THE CERAMIC COMPLEX SCHMITT'S ANALYSIS.?Schmitt's study of the pottery from Patawomeke is based on 252 rim and 367 body sherds from the Graham collection and 532 rims and 3215 body sherds collected by the Museum parties. Six pottery types were differentiated on the basis of attributes relating to whole vessels. Descriptions of two types, Potomac Creek Cord- impressed and Potomac Creek Sand-tempered, are reasonably complete; those of the third type, Rappahannock Fabric- impressed, are based solely on the Potomac Creek collection, although other collections are now available. The fourth type, Pope's Creek Net-marked, is only a suggestion. This ware has been fully described by R.L. Stephenson in his report on the Accokeek Creek site in Maryland (1963:92-96). The remain? ing types have been described respectively in two publications, e.g., Keyser Cord-marked (Manson et al., 1944) and Shepard Cord-marked (Slattery and Woodward, 1992). For the two wares in the Potomac Creek series, it is well to define the attributes used by Schmitt (and confirmed by Jirikowic). The three attributes described are: Aplastic (temper- 40 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY ing), Surface treatment, and Decoration. Schmitt defines these in the following quote (Schmitt, 1965:9-10): Aplastic The following types of aplastic were distinguished: 1. "Coarse" gravel: a quartz gravel apparently obtained from the beach of Potomac Creek just below the site. Particles range from 1.0 mm to 7.0 mm in size and possibly larger. 2. "Fine" gravel: apparently a variation of the former but with grains running consistently smaller. The larger particles were crushed to give a uniform size of 1.0 mm and under. 3. Sand: a fine sand which was more evident to touch than to the eye. Possibly this was naturally inclusive in the clay used. 4. Shell: crushed and iridescent shell occurring in flakes up to 5.0 mm in size. 5. "Leached" shell: a crushed shell temper which has leached out on the surface (shell flakes and lumps were usually discernible in fresh breaks). The leaching is possibly due to a difference in firing conditions. 6. Mixed shell and gravel: a mixture of the coarser gravel with occasional flakes of shell, possibly of accidental inclusion. 7. Sandy micaceous: sand tempered but with numer? ous mica flakes, giving an iridescent appearance. Possibly the mica was naturally inclusive in the clay or sand used. Surface Treatment Four categories of surface treatment were noted: 1. Cord impressed: apparently applied with a paddle wrapped with twisted cords. In applying this to the surface, strokes were usually overlapped to give a continuous linear effect at right angles to the lip. 2. Smoothed: probably done with the worked sherds found on the site. Smoother surfaces ranged in appear? ance from irregular, lumpy ones to carefully treated ones. Many striations were visible. 3. Plain plaited fabric impressed: often called "bas? ket impressed." The warp consisted of large, closely spaced elements while the weft was made of much finer, but also closely spaced, elements. Sherds exhibiting this type of treatment were usually badly weathered. 4. Net impressed: produced by impressing a knotted net. Decoration Decoration was broken down into the following techniques and design elements: 1. Pseudo-cord impressed: produced by impressing a stick, a sharp edge (such as a Unio shell) or occasionally a group of cords about which a twisted cord or possibly a leather thong had been wrapped. All three techniques produce a very similar impression, and minute inspection is necessary to differentiate between them. Cord wrapped stick predominates. a) Short parallel elements. This design, as the name implies, consists of a series of short parallel elements of pseudo-cord impressions (five to twelve wrappings of the cord), which are usually placed on the rim obliquely to the lip. b) Simple. In this category the impressions over? lapped end to end to form continuous lines. Designs consist of bands parallel to the lip; hachuring between the lip and a band parallel to the lip; a series of joined hachured triangles; and intersecting lines producing rectangles and diamonds in the negative. c) Complicated. Impressions are again overlapped to form continuous lines but other impressions are superim? posed on these to bring out various designs. Depth of impression is also used to bring out the design motif. 2. Cord impressed: produced by impressing individ? ual cords. a) Short parallel elements. A series of short parallel impressions are usually placed obliquely to the lip. Occasionally knot impressions are visible at one end. b) Simple. Impressions are longer and form continu? ous bands parallel to the lip, hachured triangles, and hachured lines between the lip and a band parallel to the lip. 3. Incised: usually produced with a blunt instrument but occasionally by one with a sharp point. Crude horizontal bands and bold stroked elements can be differentiated from a technique in which execution is more careful and in which rows of parallel nested triangles and parallel lines accompanied by punctates are utilized. 4. Punctated: usually produced by some tool, such as a hollow reed, which leaves a raised portion in the center of the impression. This type of decoration is very rarely used alone; it often accompanies complicated pseudo- cord impressed. 5. Finger-nail impressed: usually a design of parallel lines of concentric impressions at right angles to the lip. JfRTKOWic's ANALYSIS.?During the summer of 1988, Ms. Christine Jirikowic re-examined the pottery collection at the National Museum from Patawomeke. She examined not only the collections made by Judge Graham and the Museum parties during the period 1935-1940, but she also compared them with collections made over many years by others, from William Henry Holmes up to Richard G. Slattery, who collected there around 1934-1935. Her observations (most of which parallel those of Karl Schmitt) are set forth below. NUMBER 36 41 Description of Selected Variables (by Christine Jirikowic) I. Surface Treatments vary as follows: a. Smooth?smoothed surface with or without previ? ous cord marks still visible. (Note: this category is named Potomac Creek Plain by Stephenson et al., 1963:119). b. Fine cord-marked?fine cords tightly wrapped on paddles; carefully applied; resulting impressions are closely spaced, mostly vertical without much superposition. c. Coarse cord-marked?thicker cords; loosely wrapped on paddle; tends to be more haphazardly applied; resulting impressions are more widely spaced, horizontal and diagonal, as well as vertical and superimposed. d. Fabric?impressed with loosely woven fabric with weft threads thinner and more closely spaced than warp. II. Decorative Techniques largely follow those of Schmitt (1965): a. Twisted cord?design applied directly witii single length of cord composed of two strands of single or multiple ply cords twisted together, either left over right (z-twist) or right over left (S-twist). b. Cord-wrapped?design applied with an implement (stick, wire, cord, paddle edge, etc.), which had been wrapped with twisted cords. The resulting impressions form series of short parallel lines. c. Hollow punctate?produced with hollow tool such as a reed, resulting circular impression has a raised portion in center. d. Solid punctate?produced with a solid tool. e. Single cord/knot impression?series of individually produced single impression made from either a knot or a small section of twisted cord, e.g., wrapped once around a finger tip. III. Decorative Motifs: Schmitt's three motif categories can be enlarged as follows (with definitions below and shown in Figure 30): Schmitt Jirikowic "Short parallel elements" Short diagonals Short perpendiculars Short horizontals 'Simple' 'Complicated' Horizontal lines Bound diagonals Infilled geometries Fine cord-wrapped combinations Simple combinations a. Horizontal lines?impressions (either simple twisted cord or a cord-wrapped implement) overlapped end to end to form continuous lines (as per Schmitt). Usually 3-5 parallel lines encircling the rim edge (Figure 30-1). b. Bound diagonals?encircling band of parallel lines diagonal to rim edge; bounded on top by rim edge and below by simple horizontal line (Figure 30-11). c. Short diagonals?short parallel lines, diagonal to the rim edge and without horizontal line boundary. Often less closely spaced and less carefully placed than bound diagonal designs (Figure 30-111). d. Short perpendiculars?same as short diagonals ex? cept lines are placed perpendicular to rim edge (Figure 30-IV). e. Short horizontals?groups of short parallel lines placed parallel to the rim edge (Figure 30-V). f. Infilled geometric figures?parallel and diagonal lines forming figures filled in with closely spaced parallel lines. The figures are most frequently triangular (Figure 30-VI). g. Fine cord combinations?usually multiple bands of different designs; often one closely spaced motif superimposed on another, e.g., diagonals over horizontals. Usually executed with a fine cord tightly wrapped around a small implement. Fre? quently includes a punctate "accent" in the design. Also frequently associated witii a thickened rim (Figure 30-VIII). h. Simple combinations?applied with a thicker ele? ment and not as closely spaced as above combina? tions. Includes multiple bands of different simple motifs, e.g., one band of horizontals witii one of diagonals, simple superimposed designs and sim? ple open geometric figure designs (Figure 30-VII). (This is admittedly a catchall category.) Discussion The results of my examination of ceramics from the Potomac Creek collection of Graham, Stewart, Slattery, and Holmes concurs with the general results Schmitt obtained from his examination of the original Graham and Stewart collections. The majority of sherds from these collections are consistent with Schmitt's descrip? tions of Potomac Creek Cord-impressed or Potomac Creek Sand-tempered. The ware appears to encompass a fairly distinct range of tempers from pure sand to mixtures of sand and gravel to predominantly coarse gravel. Similarly, the decorative motifs range from plain cord-marked to simple arrangements of parallel lines to intricate geomet? ric figures and superimposed designs. The finer, more intricate designs are applied with fine cord-wrapped implements and tend to be associated with finer tempers. 42 This association is not, however, entirely consistent. This fact suggests that the makers of this pottery were experimenting (within limits) with a variety of stylistic expressions and manufacturing techniques. Without spatial or temporal control, it is impossible to determine whether these experiments reflect trends over time, or if the variations in style and manufacture were associated with particular functional or social contexts. Future work at the site should seek such associations or correlations. Schmitt's description of the ware types represented in the Potomac Creek collection is consistent with results of my examination. I would, however, question (as others have before me) the usefulness of distinguishing the Potomac Creek Sand-tempered pottery as a ware separate from other Potomac Creek pottery on the basis of data from this collection alone. Without contextual data, there is no justification for separating into a discrete ware the extreme examples of the range of variability found within the general parameters of Potomac Creek ware. OTHER ANALYSES.?The several ceramic wares and their included types have been described (and in some instances named) by other researchers. A Potomac Creek "type" was first named by James B. Griffin in 1943, during his analysis of ceramics from the Keyser Farm Site in Page County, Virginia (Manson et al., 1944:411-413). In naming the type Griffin was continuing the category first named by Holmes in his monumental work on the pottery of the eastern United States (Holmes, 1903:155-156). Holmes had included what we now recognize as several wares in his compilations for the Potomac River area. These include Pope's Creek and Marcey Creek wares, which he thought to be contemporaneous with the pottery he found at Patawomeke. Holmes did point out the distinctiveness of the sherds at Patawomeke, and he com? mented on the regional differences which he noted and described. It remained for later scholars to divide these wares into temporally defined wares. The Accokeek Creek site at the mouth of Piscataway Creek in Prince Georges County, Maryland, across the river and upstream from Mount Vernon, provided the ceramic assem? blage which could be so divided. Karl Schmitt had examined part of the Accokeek Creek assemblage and had commented on its resemblance to that at Patawomeke (Schmitt, 1965:21). During the 1950s, the collection excavated in the 1930s by Alice L.L. Ferguson was studied by Robert L. Stephenson. Despite the lack of stratigraphic evidence, he sorted the ceramics into sequential time periods, described the various wares, and assigned names where these seemed to be needed. He described in great detail the Potomac Creek Cord-impressed type, and also to a lesser extent the smoothed variety, which he named Potomac Creek Plain (Stephenson et al., 1963:115- 120). His descriptions will not be repeated here, because we will use the data derived by Schmitt in his analysis of the Patawomeke collections. I have paraphased the data deter- SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY TABLE 2.?Ceramics from Museum excavations, 1938-1940. Type Potomac Creek Cord-impressed Potomac Creek Sand-tempered Rappahannock Fabric-impressed Keyser Cord-marked Shepard Cord-marked Popes Creek Net-marked Unclassified TOTAL Number Percent 3410 184 80 20 8 2 43 3747 91.0 4.9 2.1 0.5 0.2 trace 1.1 99.8 mined by Schmitt, but essentially the following descriptions follow those set forth in his 1942 M.A. thesis (Schmitt, 1965). Table 2 shows the numerical representation of the six types of ceramics obtained by the Museum field parties. In addition to the 80 Rappahannock Fabric-impressed sherds recovered from the general refuse and listed in the table, 142 sherds of this ware, including 56 belonging to a restorable vessel (Figure 25c), were recovered from the large pit underlying the moat-like ditch in square 4L6 (see Diary, page 47). Tables 3 and 4 present the analysis of sherds by surface treatment, decoration and temper. Because the two series, one recovered during the 1938-1939 seasons and the other during 1940, were of similar size and had a similar distribution of attributes, they would appear to represent true random samples and therefore have been pooled in making up the tables. Potomac Creek Cord-impressed Ware PASTE.?Temper ranges from a finely crushed quartz gravel with particles over 1 mm across and up to 7 mm in size. These are present in medium to excessive quantities. Occasionally a small amount of crushed shell is mixed with the other tempering materials, and a few sherds seem to be untempered. Texture of the paste is medium to medium coarse, and exterior hardness is from 2.0 to 3.0 on the Moh scale. Exterior colors range from buffs through browns to grayish black, with the latter predominating. Interiors show a similar variety of colors, and in addition, some sherds exhibit a lustrous dark gray color indicating intentional smudging. Blackening of exteriors through use over a fire is common. METHOD OF FABRICATION.?"Coil-breaks" and cord- roughened surface finish indicate that vessels were made with coiled strips of clay malleated together by use of a cord- wrapped paddle. SURFACE FINISH.?Apparently a cord-wrapped paddle was used to roughen the surface of most vessels. Some were left with the entire surface cord-roughened; the surface of others was partially smoothed over. Even on relatively carefully smoothed sherds, occasionally an isolated, unobliterated cord impression remains. The strokes of the wrapped paddle were usually applied in an overlapping manner, so that the linear NUMBER 36 43 TABLE 3.?Rim sherds from Museum excavations, 1938-1940. Surface treatment and decoration CORD-IMPRESSED Not decorated Decorated with: Pseudocord elements: Short parallel Simple Twisted cords: Short parallel Simple Fingernail Incised SMOOTH Not decorated Decorated with: Pseudocord elements: Short parallel Simple Twisted cords: Short parallel Simple Incised Punctated FABRIC-IMPRESSED TOTAL PERCENT Coarse gravel 169 10 9 64 78 2 - 32 2 15 7 20 5 - - 413 81.1 Fine gravel 3 1 1 1 3 2 8 1 4 24 4.7 Sand 9 - - - - - 39 1 1 4 1 1 56 11.0 Temper Micaceous Leached sand shell 1 - - 1 2 - - 3 2 2 7 0.4 1.4 Shell 4 - - - - 1 5 1.0 Shell and gravel 1 - 1 0.2 Steatite - - 1 1 0.2 Total 186 11 11 65 81 2 1 76 3 24 8 28 9 1 3 509 Percent 36.5 2.2 2.2 12.8 15.9 0.4 0.2 14.9 0.6 4.7 1.6 5.5 1.8 0.2 0.6 arrangement of impressions perpendicular to the lip was produced. The basal portion was often clumsily smoothed, whereas the rim area was frequently treated more carefully before application of decoration. Interiors were smoothed, and numerous striations from fingers or smoothing tools are evident. Bumpy exteriors are common because of protruding temper particles and indifferent smoothing. DECORATION.?Impressions made by cords, cord-wrapped sticks, cord-wrapped edges, and cord-wrapped strings represent the major decorative techniques. The result produced by the latter two treatments has been called "pseudo-cord," following the terminology of Griffin (Manson, et al., 1944:407). Most pseudo-cord impressions seem to have been produced with a sharp edge, such as that of a fresh water mussel shell, about which a twisted cord or sinew had been wrapped. Pseudo-cord decorations are subdivided (following Schmitt) into three categories: (1) short parallel elements; (2) simple; and (3) complicated. The first category consists of impressions exhibiting five to twelve wrappings of the cord, placed parallel to one another and usually obliquely to the lip. The second category?simple pseudo-cord?has impressions overlapping and end to end and thus forming continuous lines so that the following designs are produced: horizontal bands parallel to the lip; hachuring between the lip and a line parallel to the lip; series of joined, hachured triangles; and angular, zoned patterns producing rectangles and triangles in negative. The third category?complicated pseudo-cord?also has overlapping impressions forming continuous lines, but other impressions are superimposed on these to bring out triangular or rectangular designs by means of varying depths of the impressions. Rarely, complicated pseudo-cord is accompanied by punctations. Cord-impressed designs also occur in series of short, parallel elements and in simple designs (simulating those produced by simple pseudo-cord). The more complicated designs are associated with the finer gravel tempering (under 1 mm in size). Less frequently used techniques of decoration were punctat- ing, incising, fingernail incising, and impressing with knots. Designs were largely confined to the rim area of vessels, except that, rarely, pseudo-cord impressions extend a short distance into the interior of the vessel perpendicular to the lip (Figures 27-30). FORM.?Rims are straight to slightly flaring. Occasionally, added rim strips occur, producing a weak-collared effect, and this is most often associated with fine gravel temper and complicated pseudo-cord designs. Lips are rounded or flat? tened, with the latter predominating. Lips frequently carry cord and pseudo-cord impressions, and when the latter are deep, a scalloped effect is achieved. Vessel bodies are elongated and taper down to the base. Slightly constricted necks and weakly rounded shoulders are often present. Bases are subconoidal and 44 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY TABLE 4.?Body sherds from Museum excavations, 1938-1940. Surface treatment and decoration CORD-IMPRESSED Not decorated Decorated with: Pseudocord elements: Short parallel Simple Complicated Twisted cords: Short parallel Simple Fingernail Incised Punctated SMOOTH Not decorated Decorated with: Pseudocord elements: Short parallel Simple Complicated Twisted cords: Short parallel Simple Fingernail Incised FABRIC-IMPRESSED NET-IMPRESSED TOTAL PERCENT Coarse gravel 1862 4 7 3 23 1 2 1 662 16 1 2 18 1 2603 81.0 Fine gravel 149 1 159 16 6 4 - - 335 10.4 Sand 29 1 1 91 1 2 1 2 2 1 2 133 4.1 Temper Micaceous Leached sand shell 10 2 1 1 1 1 - - 14 0.5 8 10 51 71 2.2 Shell 16 23 - 2 - 41 1.3 Shell and gravel 12 - - 5 17 0.5 Feldspar 1 - - _ - - 1 trace Total 2081 5 9 1 4 24 1 2 1 949 1 34 8 2 24 2 11 53 2 3215 Percent 64.7 0.2 0.3 trace 0.1 0.7 trace trace trace 29.5 trace 1.1 0.2 trace 0.7 trace 0.3 1.7 trace rarely flat. Thicknesses rang from one-eighth to eleven- sixteenths of an inch (3.0 to 18 mm), with greatest thickening occuring in basal sherds. Orifices are oval to round. No appendages were noted (Figures 25a,b and 26). DISTRIBUTION.?Potomac Creek Cord-impressed (and Plain) wares appear to have a relatively localized distribution. Bushnell's illustrations (1935) indicate it at a site on the Rappahannock River just west of Fredericksburg, and also at the Forest Hall Site, Richards Ford Site, Jerry's Rats Site, and possibly the Skinkers Ford Site, all located above the Fall Line on the Rappahannock. Bushnell (1937) illustrated the ware as occurring below the falls at the mouth of Chingoteague Creek in King George County. Other types, of course, are figured in Bushnell's illustrations. Holmes (1903, pi. 140) illustrates Potomac Creek Cord-impressed pottery from the Anacostia area of Washington. The type occurs plentifully at the site of Moyaone (the Accokeek Creek Site) on the Maryland shore. Here the ware shows a higher incidence of fine gravel tempering, and a greater frequency of complicated pseudo-cord decorations (Stephenson et al., 1963:113-120 and plates 13-18). Six sherds from an ossuary on Piscataway Creek (Ferguson, 1940) are closely related to the Potomac Creek Cord-impressed ware, and there are indications that they represent a variety that might date to later in the 17th Century. Besides the Piscataway Creek examples, Schmitt collected similar sherds from sites at Hedge Neck near Broad Creek in Prince George County, Bullocks Neck on Mattawoman Creek, and Cedar Point Neck near Nanjemoy Creek, all on the Maryland shore. Other sites producing at least occasional Potomac Creek pottery are recorded by W. Clark (1980). Griffin (in Manson et al., 1944) describes Potomac Creek pottery as a minority ware at the Keyser Farm Site, Page County, Virginia. It has also been found at the Cabin Run Site near Front Royal in Warren County, Virginia (Gardner, 1986:82). Dorothy Cross (1941, vol. 1, plates 10a and 22b) pictures sherds remarkably close in design and decoration technique from the Indian Head and Salisbury sites in southern New Jersey. As a distinct type, Potomac Creek Cord-impressed is most frequent in the upper tidewater region of the Potomac valley and would not be expected to occur south of the Rappahannock nor north of the Susquehanna. To the west it occurs as a NUMBER 36 45 ^^SP^' : ' n " l \ ? ' ? ' FIGURE 25.?The only restorable vessels from the Potomac Creek site: a and b, Potomac Creek Cord-impressed ware (USNM Nos. 385,240a and -234); c, Rappahannock Fabric-impressed ware (USNM No. 385,240). 46 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY FIGURE 26.?Vertical sections of 10 rim sherds of Potomac Creek Cord- impressed type, oriented with exterior surfaces to right. minority ware in the Piedmont area and in the lower half of the Shenandoah valley. Distribution to the east is not known, although it is uncommon along the Patuxent River, as well as in the valleys of the Severn and Patapsco rivers. A closely related pottery type, Shepard Cord-marked, occurs at the Shepard site in Montgomery County, Maryland, and at other sites in the Piedomont area (Slattery and Woodward, 1992). Page Cord-marked, a type similar to Shepard Cord- marked but with a different temper (limestone) is a minority ware at the Keyser Site. Holmes (1903, pi. 134) illustrates sherds attributable to both these types from a site near Luray, Virginia. The cord roughening, pseudo-cord impressions, added rim strips producing collars, and bold-stroked elements and pseudo-cord impressions at the junction of the rim and collar are excellently shown. Holmes does not show a variation, occurring at the Shepard Site and represented by a few sherds at Patawomeke, which has crude horizontal incised bands on the collar, as well as bold-stroked elements. The broader relationships of Potomac Creek ware lie mainly to the north, where similar shapes and decoration techniques, but not actual designs, are found. The pottery of the Owasco Aspect of New York (Ritchie, 1944) and of the Lake Michigan Phase (McKern, 1931), the grit-tempered wares from Fort Ancient sites in the central Ohio Valley (Griffin, 1943), and of the Monongahela culture in southwestern Pennsylvania (But? ler, 1939; Mayer-Oakes, 1955) are in the same tradition. A related ware, found at late (proto-historic) sites along the Rappahannock River, has been named Camden Plain (Mac? Cord, 1969). The relationships apparent between the various wares will be discussed and evaluated in Part IV of this study. Potomac Creek Sand-tempered Ware FIGURE 31 PASTE.?Temper consists of a fine sand, which is more noticeable by touch than visually. It is possible, but not probable, that the sand was a natural inclusion in the potting clay. Tempering material is present in sparse to moderate amounts. Texture of the paste is medium to medium coarse. Exterior colors are buff to brown, with occasionally reddish and grayish black tones present. Hardness is between 2.0 and 3.0. METHOD OF FABRICATION.?Sherds with coil breaks indi? cate that a coiling technique was used, though some of the miniature vessels appear to have been molded by hand from clay lumps. Some sherds with cord-roughened finish indicate that larger vessels were paddle-malleated. SURFACE FINISH.?Smoothing predominates, although a minor number of sherds are cord-impressed. DECORATIONS.?Decorated sherds are rare. Those present exhibit techniques of impressing with twisted cords, pseudo? cord, or fingernails and incising. Designs consist of short parallel elements and horizontal bands and triangles made by cord and pseudo-cord impressions and are confined to the rim area. FORM.?Vessels of several shapes are observed, including small pots under six inches (15 cm) in height, with rounded or flattened bases and straight rims. Larger, wide-mouthed vessels are also present and indicate that this type overlaps with the Potomac Creek Cord-impressed type. The majority of the vessels are small, many being miniatures. DISTRIBUTION.?This type is little known but would be expected to have a distribution similar to that of the Potomac Creek Cord-impressed type. Stephenson (1963:120) designated this as "Moyaone" ware, but this name has been challenged. Minority Wares FIGURES 32, 33 In the collections from Patawomeke there are at least four other named ceramic types, plus some unclassifiable speci- NUMBER 36 47 km w FIGURE 27.?Potsherds of Potomac Creek Cord-impressed ware. All but sherd a are of Potomac Creek Cord-impressed type; all but c are rim sherds; a is unclassified. USNM numbers represented are 378,892; 385,212; 385,241; and 385,192. 48 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY 'Xv<- X ,,,'*" it ">??/f>ii.,hit,X X- XX /,? 1/7-X' FIGURE 28.?Rim sherds of Potomac Creek Cord-impressed type (a,d,e,f, and g bear USNM No. 378,692; b, 385,241; c, 385,217). NUMBER 36 49 FIGURE 29.?Rim sherds of Potomac Creek Cord-impressed type (a-g); unclassified (h); and Rappahannock Fabric impressed type (/). All are covered by USNM numbers 378,678, -692, -725; 385,180, -235B, -241. 50 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY X . V W w n h A L'1 \nes K- focund Difl<\or\aU \0|MOWO?tMttt| HQOttWfDMMtt CofA-urorVtA AC 4H&t&tt*x*v W\&X CM-A SS W S HI. S\\o^ D\ttftor\&\^ Tg, Shott PerptUirjol&tt V . Short Wwiytifals Ooooo 0000O QOODfe tftooo\ 0D0OC 00000 2E, \r\f\\U 6t0rA^\t V\yfe& (fkCiryo\ Ats*yvS W ^ r W ? ? - i ? A rxjm r- c L FIGURE 37.?Designs from 6 pipe fragments (after Stewart, 1954): d and e include base of bowls; otherwise all are stems and d is a tubular pipe. Figure 7>lb supplied the detail in Figure 36 (bottom). USNM numbers are 378,734 (e), -776 (a); -885 (b); 385,155 (c,d, and/). (Scale bars = 1 cm.) FIGURE 38.?Designs from 5 pipe fragments (after Stewart, 1954): a and e extend from bowls onto stems; b on bowl only; c and d on stems. A detail of c is in Figure 36 (middle). USNM numbers are 378,892 (d); 385,155 (all others). (Scale bars = 1 cm.) The 1988 re-examination of the Patawomeke collection of pipes produced the tabulations shown in Table 11. Further, Ms. Jirikowic made comments, as follows: The sample of 295 clay pipes and pipe fragments include a total of 259 specimens from the collections of Graham and Stewart, 33 specimens from Slattery's Potomac Creek collection (USNM #417,373), two specimens from Holmes' collection (USNM #135,992 & 135,993), and one from other Smithsonian Institution collections (USNM #196,344). The discrepancies between Schmitt's tabulation and my own are most likely due to my more liberal inference of type (tubular or obtuse) for fragments that Schmitt perhaps deemed too incomplete for accurate determina? tion. My own judgements?perhaps unwarranted?were based on the clearly different characteristics of the more complete specimens of each type. The tubular pipes tend to be more crudely crafted of a perceptibly sandier paste, having a thicker construction and no surface finish such as decorations or burnishing. The obtuse pipes on the other hand, tend to be more finely crafted of untempered paste, most likely burnished and more likely to deco? rated. The stems of the tubular pipes tend to taper to a NUMBER 36 63 w ^W yf^% muif^ ?%^>v FIGURE 39.-?Designs from 5 pipe stems (after Stewart, 1954): a extends onto stem; a detail of b is in Figure 36 (top). USNM numbers are 378,819 (a); -823 (d); -914 (e); 385,151 (c); -152 (b). (Scale bars = 1 cm.) round bit with a diameter smaller than that of the stem closer to the bowl. Stems of the obtuse pipes are either undistinguished from the stem or noticeably swollen or enlarged. By far, the most common technique used to decorate clay pipes was what Schmitt called a ?'roulette-like" technique. It is interesting to note that this technique has an effect similar to that of the cord-wrapped implements used to decorate pottery vessels. The decorative motifs used on the pipes also recall some of the more popular motifs of pottery decoration such as the simple encircling lines (horizontal lines), bound fields of diagonal lines, and infilled geometric shapes (see illustrations of ceramic design motifs). Noticeably absent on the pipes are the thickened rims of much of the pottery. This could, of course, be due to the smaller scale and limited size of the field on pipes?although other designs apparently were adapted to these limitations. Another noteworthy differ? ence between the decorative motifs of pipes and those of pottery is the use of representative figures on two (possibly three) pipes?the "salamander" and the "ser? pent." (See foregoing sections on pottery for more description of motif variables.) OTHER CERAMIC ARTIFACTS A variety of ceramic artifacts (other than pipe or container fragments or whole specimens) were found randomly in the midden and topsoil, with no concentration or reliable associa? tions. Those found are as follows. LADLES OR SPOONS.?Ladles or spoons with shallow bowls, two inches (50 mm) in diameter, with round handles two or three inches (50-75 mm) long and one-half inch (12.5 mm) in diameter. No whole specimen was found, but the fragments found indicate the overall shape and size. The paste in these artifacts is sandy, as a rule. BEADS.?Spherical pottery beads were fairly common, 12 coming from the general refuse and several from ossuaries. Most are slightly over a half-inch (12.5 mm) in diameter, although larger specimens nearly an inch (25 mm) in diameter occur. Perforation was made by molding around a smooth reed or stick. One specimen possesses concave depressions a half-inch (12.5 mm) wide and an eighth of an inch (3 mm) deep at the perforated ends. One cylindrical pottery bead recovered is three-fourths of an inch (19 mm) long and a half-inch (12.5 mm) in diameter. Two pipestem fragments had been smoothed at the ends, apparently intended for beads. These were in the collection made by Judge Graham. One was from a native-made pipe, the other from a fragment of a white English-made pipe. The exact provenience of these two items is unknown. WORKED POTSHERDS.?These are generally rectangular with the edges partially rounded by smoothing. Size ranges from two to three inches (50-75 mm). Fourteen were found in a small pile in the refuse filling the moat-like trench. Five others were recovered in the general digging; two of these were made on sherds of Rappahannock Incised ware. These worked sherds would have made excellent tools for scraping or smoothing pottery, and striations on much pottery indicate the use of such a tool. HUMAN EFFIGY FIGURINES.?Most interesting are three small pottery figurine heads, which are an inch (25 mm) or an inch-and-a-half (37.5 mm) long. One of two complete specimens has crudely punched eyes and mouth and scratched lines on the face. The other complete specimen lacks facial 64 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY features, but was identified in shape and, like the first, possesses notches for ears. The third specimen is fragmentary and has notches for ears and a raised line on the back, possibly representing gathered hair. It also has a hole in the base (neck), probably to permit it to be fastened to a body by use of a small peg. Unfortunately, the face of this specimen is damaged, and so its features cannot be determined. A similar ceramic head, found at the Kiser Site in Colonial Heights, Virginia, is shown in an article on "Effigy Heads in Virginia" (MacCord, 1966:73). The latter head is made of shell-tempered clay, whereas those from Patawomeke are made of slightly sandy or tempered clay. MINIATURE POTS.?Several small, fragmentary "toy" pots in the collection are made from slightly sandy paste. They are plain on exterior surfaces, and the interiors are roughly smooth. They seem to have been molded over the end of an adult's finger, possibly to make a toy for a child. Dimensions varied, but most are under two inches (50 mm) across and one inch (25 mm) high. POTTER'S WASTE.?Several roughly cylindrical lengths of fired clay seem to have been parts of coils used in pottery making. Apparently, some of the coils had not been used, but had been thrown into a fire, where they became fired. They seem to have had no use or purpose. The source (or sources) of the pot clay used in making pottery vessels, pipes, and the other ceramic artifacts has not been identified. At the site, clay forms part of the underlying geological formations, and usable clay could have been obtained from the eroding river bank. On the other hand, upland sources may have been sought and used. Thus far, no mineralogical studies have been conducted on the Patawomeke ceramics. LrTHic ARTIFACTS Artifacts of stone were numerous, both on the surface and in the excavations at Patawomeke. Because the site had long been known to local collectors, undoubtedly many hundreds of such artifacts had been taken away in years past. The stone tools are of two major categories, chipped and ground. Conforming to the customary descriptions and uses postulated for various artifacts, I will describe and discuss those found in the current work in the same fashion, chipped items first. PROJECTILE POINTS.?Recognizable projectile points of stone were fairly common. Those which could be classified in accord with recent categorizations are listed in Table 12. They total 191, which could be typed, plus another 25 that were broken or otherwise unclassifiable. As seen in the tabulation, the points can be assigned to all cultural periods from the Early Archaic up to the Late Woodland, probably extending into the early Historic Period. Materials from which the artifacts had been made seem to locally obtained, witii only few exceptions. Those are the few points made from rhyolite, chert, or Harper's Ferry quartzite. The latter stone outcrops near Harper's Ferry, West Virginia, although pebbles of it occur in the lower Potomac Valley gravels, as do the other exotic stones. The Harper's Ferry quartzite was called "breccia" by Judge Graham, Karl Schmitt, and Carl Manson in the various notes they made on the collections. Most of the projectile points had been made of the locally plentiful quartz and quartzite, as follows: Quartz Quartzite Other Totals Point categories Stemmed types Levanna Madison Clarksville Totals and percentages 146(73.0) The triangular points, presumably used on arrows, are similar in shape, but vary in length and width dimensions. Measurements taken of those in the collection revealed the following median dimensions: 37(25.8) 8(42.1) 57(81.4) 44(97.8) ( . ) 16(24.2) 3(15.8) 4(05.7) 23(11.5) 13(19.7) 8(42.1) 9(12.9) 1(02.2) 31(15.5) 66 19 70 45 200 Type Levanna Madison Clarksville Average length 30 mm 26 mm 20 mm Average width 27 mm 20 mm 16 mm CHOPPERS.?Large, crudely, or hastily made, these are fragments of cobbles, usually with much of the cortex still present. A rough edge is flaked on one side, and the opposing side is usually the rounded surface of the cobble, thereby providing a comfortable gripping area. Such tools probably saw use in shredding fibers, macerating foods, or as crude scrapers for inital work in hide-scraping. DRILLS.?Four specimens that seem to have been drill- perforators are in the Patawomeke collection. One is made of quartz, one of jasper, one of Harper's Ferry quartzite, and one (triangular) is of black chert. QUARTZ PEBBLE SCRAPERS.?The Patawomeke collections contain a total of 87 quartz scrapers, manufactured from quartz pebbles, ranging in size from 22 to 54 mm in diameter, split roughly in half. The sharp edges were then retouched, usually not entirely around the tool, but varying between 25% to 80% of the periphery. The resulting tool ranges from 7 to 22 mm thick and has a working edge with angles averaging 76? (range 56? to 92?). The steep angle indicates use under conditions where heavy pressure would be needed, as in woodworking, bone-working, shredding, or in scraping and softening a hide. Although this tool seems to have been much used at Patawomeke, it has not been reported in such numbers at other sites in the region. The absence of such chipped tools as end and side scrapers, burins, and knives is noteworthy. If these were not part of the Patawomeke tool kit, then other tools must have been made to serve those functions. It seems more likely that they were not seen or collected in the work done. Ground and/or polished tools and ornaments of stone were not plentiful, but they did show a wide range of types and purposes. The specimens collected in the 1935-1940 work, NUMBER 36 TABLE 12.?Projectile point types, Stewart and Graham collections. 65 Type name Palmer Kirk LeCroy St. Albans Kanawha Stanley Morrow Mountain Halifax Lamoka Brewerton Orient Fishtail Bare Island Lackawaxen Savannah River Susquehanna Broad Holmes Claggett Calvert Poplar Island Piscataway Fox Creek-Selby Bay Levanna Madison Clarksville TOTAL Number 1 2 3 2 3 1 2 2 9 1 2 9 1 1 1 1 3 2 2 6 4 19 70 45 191 Group percent 5.8 1.6 14.1 6.8 12.0 60.2 Comment Early Archaic " " tt rt Middle Archaic " Late Archaic " " " " ? ff rf ' Early Woodland M rt ft Middle Woodland Late Woodland " Broken and unclassifiable points (not counted in above): 25. Names are from MacCord and Hranicky, 1979, which refer the reader to original sources of the names. plus those already in the USNM collection, were examined by Ms. Jirikowic, who provided the following comments to augment those of Schmitt. CELTS.?According to Schmitt, no whole celt was found. Judge Graham recovered four fragments, and the Museum parties found one that was nearly complete. All were made of a greenish igneous stone. The nearly complete specimen is five inches [12.5 cm] long, two inches [50 mm] wide, and an inch and a quarter [31 mm] thick. It is triangular in shape, with curved sides and an oval cross-section. This form seems typical of the axe-like tool or weapon used by the Patawomekes. A complete celt and four additional fragments already in the USNM collection seem to be identical with those found in the excavations. The whole celt measured 100 mm in length, 556 mm in maximum width, and 29 mm in thickness. HAMMERSTONES.?Although almost any hand-held stone could have served as a hammer, some were recognizable as deliberately made or altered. In the excavated collection, plus four others in older USNM collections, the following observations were made. Of the 24, two are unpitted; 10 have a shallow pit on one side only; and 12 had a pit on each flat face. Seventeen have been pecked to form an oval, biscuit-like shape. Several show battering marks on the flattish surfaces, as if the stone had also seen use as an anvil, perhaps in bipolar flaking of lithics. ABRADERS? These are small, flat sections of sand? stone that exhibit wear by slight concave depressions and scratches. A total of six was noted in the excavated collection, plus two previously collected by the USNM from other sources. Usually these stones had not been altered or shaped, but were used as found. PESTLES.?One elongate pebble of a hard sandstone was found in the Graham collection, plus a fragment was in the older USNM collection from Patawomeke. These are generally unmodified stones, except the one or both ends have been through pounding or grinding, as with a pestle and mortar combination. The specimen from the Graham collection shows pitting on opposing faces of the stone, probably to permit a firmer grasp of the tool in use. No mortar was found in the collection. GROOVED AXE.?One fragment of polished green? stone showing part of a groove was a surface find at the site. Axes of this type are usually assigned to the tool kit typical of the Late Archaic Period and Early Woodland Period, and hence were not part of the Patawomeke tool kit. GORGETS.?Stone gorgets or pendant ornaments were made from thin pieces of mica schist and were generally crude. Two are irregular in outline and have a single, central perforation. One has notches, or "tally marks" on one edge. Also, related to these, although not perforated, are two mica schist objects. One with a concave side is roughly rectangular. The other is also roughly rectangular with rounded corners, but has two sets of five notches on opposing sides. [Ms. Jirikowic noted an additional specimen in the older USNM collection (Catalog No. 196,335). This is of a cherty material, was centrally perforated, and has notches on one side.] WORKED SOAPSTONE (Steatite).?In addition to four pipes or fragments, several other artifacts made from this stone were found. One is a fragment of what appears to have been a bead, probably made from a small segment of a pipestem. It carries USNM Accession No. 379025. Another item is a cone-shaped object, possibly a paint pot. This object measures 34 mm in diameter and is 15 mm deep. Its Accession number is 378977. One small fragment of a steatite pot found probably represents an occupation of the site during the Late Archaic Period or Early Woodland Period. 66 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY BONE AND ANTLER ARTIFACTS Bones, antlers, and their fragments were plentiful in the refuse found in the midden pits and ditches. Among the bones and antlers were many artifacts and worked pieces of each. Obviously, the bone industry of the Patawomeke Indians was an important component of their tool kit. The recognizable tools and other artifacts of bone or antler are listed and described below. These items had been studied by Karl Schmitt, and most of the following are from his M.A. thesis (Schmitt, 1965:16-20). Findings and observations of Ms. Jirikowic will follow. BEADS.?Many small cylindrical bird bone beads were found. Usually these were not polished, the cut ends not even being smoothed, but occasionally care had been taken in tiieir manufacture, and finely finished forms did occur. COMB.?One comb made of bone was found in the Multiple Burial pit, and this will be detailed below in conjunction with European-made items. ARROW POINTS.?Antler arrow points one-and-one-fourth inches (31 mm) to two inches (50 mm) in length were made from cut tines that were scraped down to a sharp point and hollowed out for insertion of a shaft. Eight were recovered, seven being simple conical, socketed points, the other having a perforation meeting the hollowed portion, possibly for inser? tion of a pin to help secure the point to the arrow shaft. Representing refuse from the manufacture of this type of point are the numerous antler sections from which the tines had been cut or broken. Another type of arrow point was made from the terminal phalanges of the deer. These had had the proximal end cut off, exposing the natural cavity of the bone, which served for insertion of the shaft, and were sharpened to a point. Three finished specimens were found, and two with only the end removed, probably representing points discarded in the process of manufacture. TURKEY CALLER.?An interesting hunting implement iden? tical to those reported by Speck (1928:357) was found at Patawomeke. These callers were made from the wing bone of the wild turkey, cut and polished. One complete specimen, three inches (75 mm) in length, and two broken ones, two inches (50 mm) long, were found. Speck says that they were used to lure wild turkeys close to the hunter. The hunter emitted "yelps" or "calls" by blowing into one end and holding the other cupped in his hands. FISH HOOKS.?Three broken blanks and two broken hooks are in the combined collection. The blanks were made from a flat section of bone, roughly retangular in shape, by cutting out tiie center portion. From this blank, by cutting opposite sides, two barbless hooks could be made. The broken blanks are between an inch-and-a-half (37.5 mm) and two inches (50 mm) long and would have been about double that when whole. The hooks are both one inch (25) long. Possibly the spatulate implement found (described below) represents a part of the process of hook making, a step before the center portion was removed. Another implement, a double-pointed "awl," very possibly was a gorget or part of a compound fish hook. The specimen is two-and-a-quarter inches (56 mm) long and was polished all over. AWLS.?These were made from many types of bone. Most common were simple splinters, usually worked only at one end but occasionally polished over the entire surface. Ninety-two specimens of this type are present in the combined collections; six of these are narrowed and rounded at the butt end, and one possesses rude notches, excellent features if the awls were hafted in antler handles. One large awl made from the metapodal bone of deer was recovered. Two awls are made from deer ulnas and two from the splint bone of the same animal. Eight are made from the tibio-tarsus of the wild turkey. All these types preserve a joint, possibly for facility in use. BEAMERS.?These were fairly common, 15 broken examples being present, and were made from the metapodal bone of the deer Apparently, the bone was cut, and not broken at random, because the finished tool was invariably quite symmetrical. ANTLER HANDLES.?These were made from sections of antler from which the tines had been removed. None had been finished at both ends but all had a portion of the cancellous bone worked out to form a hollowed end. This furnished an excellent haft for some such implement as a bone awl. Six specimens are between one inch (25 mm) and four-and-a-half inches (11 cm) in length, with a hollow between three- sixteenths of an inch (5 mm) and three-eighths of an inch (9 mm) in diameter. A more finished specimen is two-and-one- half inches (72 mm) by seven-eighths of an inch (22 mm) with a hollow large enough to haft a small celt. ANTLER PUNCHES.?These are large tines up to six inches (15 cm) in length, which have been smoothed and polished. Seven specimens that might have fulfilled the use of a punch were found. One had been worked to an efficient point by flattening opposite sides. ANTLER CYLINDERS.?Antler cylinders measure between one-half inch (12.5 mm) and three-and-one-fourth inches (81 mm) in length, the longest specimen having a knob-like swelling at the largest end. These objects are smoothed all over and apparently were made with great care. A tool undoubtedly serving the same purpose is triangular in cross-section and made of what appears to be the jaw of a large fish, possibly a gar. This implement is two-and-a-quarter inches (56 mm) long, and the sides measure one-half inch (12.5 mm) and three- eighths of an inch (9 mm). Much care was taken in the working and polishing of this object. These cylinders are usually called flakers, though quite possibly they served some other purpose. Cadzow illustrates a number of antler cylinders, some exhibiting enlarged, rounded ends, and suggests that they were used for weaving or as counters in games (Cadzow, 1936: 182-183). SPATULATE IMPLEMENT.?This is a flat, polished section of bone, three inches (75 mm) in length, with rude notches at one end. This possibly represents a step in the manufacture of fish NUMBER 36 67 hooks. BEAVER TOOTH CHISEL.?One specimen was found in which the tooth had been split longitudinally and some of the rough edges partially smoothed. The original, natural cutting edge was the working portion used by the Indians. NEEDLES.?No complete examples were found, but 10 fragments of various sizes were recovered. The largest of these was four-and-a-half inches (112 mm) in length and was incomplete at both ends. These implements were long, flat, curved, pointed at one end, and possibly had perforations at both ends. They seem to have been made from deer ribs. TURTLE-SHELL CUPS.?Turtle-shell cups were frequently made by smoothing down the sides and backbone of the common land tortoise (Terrapene Carolina). One whole specimen and seven large fragments were recovered. PERFORATED DEER PHALANGES.?Deer phalanges were perforated in a peculiar manner, apparently by holding a pointed tool on the bone and striking it. This produced a jagged, transverse perforation. Three specimens were found. NOTCHED CANINE TEETH.?Two canine teeth of a small canine animal, possibly a raccoon, were found. These have shallow notches around the root end, possibly for use as pendants. In her examination of the collection in 1988, Ms. Jirikowic "concur[s] with Schmitt in his description of artifact types and materials." Her count of the various types "varies slightly" as follows: Splinter awls Needles Spatulate tools Beads Fish hooks Pendants/teeth Other worked bone 99 7 4 6 5 4 3 Beamers Antler points Bone points Antler punches Antler cylinders Worked antler tines 18 14 2 8 5 83 "The collection includes two notched teeth; three shark teeth; two highly polished pieces of turtle shell and a third with two perforations; and ten additional pieces of worked turtle shell." The turtle shell with two perforations had been collected by Slattery and bears the USNM Accession number 417379. SHELL ARTIFACTS AND FOSSIL SHELLS Refuse shell occurred throughout the midden at Patawo? meke, with fair preservation. Shells of oyster, mussel, and two types of snails were recognized. The snails were both aquatic and terrestrial, the latter probably present because they scavenged on the Indian organic refuse. Artifacts made from these shelled animals were rare, with only the riverine mussel used as tools. The tools usually involved only one valve of the paired shell, and this served as a scraper or as a spoon. In a few instances, the working edge of the shell was serrated, thus providing an excellent cutting edge. No evidence was noted that such shell had been fitted with a handle. The serrations were spaced too widely to have been a tool for impressing the so called "rouletting" designs on ceramics. BEADS.?Quoting from Schmitt (1965): Thousands of shell beads were recovered from the ossuaries, and occasional specimens were found in village refuse. Disk beads made up the greater portion of the collection. The smaller disk beads averaged an eighth of an inch [3 mm] in diameter and a thirty-second of an inch [0.8 mm] in thickness, whereas larger forms measured one-half inch [12.5 mm] by one-sixteenth inch [1.5 mm]; the former were particularly numerous. Marginella beads, made by rubbing a hole in one side of the shell, were nearly as numerous as the small disk type. Cylindrical forms approximately five-sixteenths of an inch [8 mm] in length and an eighth of an inch [3 mm] in diameter were common; this is the type often called wampum. Spherical forms with flattened poles, averaging a half-inch [12 mm] in diameter, and larger curved cylindrical shapes an inch-and-a-half [37 mm] by a half-inch [12.5 mm] also occurred. The larger cylindrical specimens occasionally exhibited multiple perforations; in these, the regular longitudinal perforations were met by others from the side, resulting in four holes. Large, flattened, roughly rectangular forms with rounded edges and concave ends were also present. A peculiar hemispherical type with a concave depression on the flat side was found. A necklace consisting of beads of this form was found in situ (in Ossuary No. 2). The hemispherical forms were grouped by twos, back to back, with the couples separated by small disk roundelles. Spherical forms were also discovered in situ, separated by alternating disk-shaped roundelles. GORGETS.?From Schmitt (1965): Gorgets of both shell and stone were found. The Graham collection includes two complete shell gorgets and two fragments; the U.S. National Museum collection has one complete specimen. All were circular in form, two possessing a double perforation with a connecting groove, and one a double perforation with no groove. Two have a single perforation. Designs were present on three and consisted of triangles arranged around the periphery with apices pointing toward a central star or geometrical figure. FOSSILS.?Although not manufactured or otherwise altered, the many fossil shells found in the midden can be classed as artifacts, because they had been collected and brought to the site by the villagers. Specimens of Turritella mortoni, a spiral shelled animal, number over 100. On the eastern side of Marlborough Point, a mile from the village site, ledges of fossil bearing marl outcrop. Many of the shells have eroded from the marl and cover the beach. As most of the shells found in the excavations seem to have been waterworn, they probably came from that beach. Other fossil shells found were those of Ostrea compressirostra, an extinct form of oyster, also found in the marl. These shells are large and flat and may have been used as plates. From the same marl source, the site's inhabitants also collected fossil shark teeth. CORDAGE AND TEXTILES An important and useful element in the inventory of artifacts made and used by Indians is the usually missing category of cordage, textiles, and basketry. Only under exceptional circumstances are these materials preserved, and their contact with copper is one such circumstance. As the copper corrodes, the by-products (copper sulfate and/or carbonate) penetrate the otherwise perishable materials and protect them against decay bacteria and scavengers. In many of the copper beads found in the Multiple Burial and in Ossuary 1, numerous instances of such cordage preservation were noted by Judge Graham. In one instance, Judge Graham observed "what appeared to 68 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY be cross-woven bark cloth" on a sheet of copper. Because he did not have the cloth analyzed, the identification as "bark cloth" is questionable. In any event, the weaving seems to have been a simple "over and under" weave, and the material may have been a split cane. In another burial, he found part of a string, made from two strands twisted together counter? clockwise upwardly, providing a "Z-twist" cord element. Each strand had been twisted clockwise, producing an "S-twist" for the finer cord. In another bead of copper, he found a section of cord about one millimeter in thickness, made up of six threads twisted together from left to right (clockwise) producing an "S-twist." The individual filaments are a plant fiber and are twisted from right to left (counter-clockwise), producing a "Z-twist." See also a study by Carole P. Falk (1983). Additional information on cordage can be obtained from the cord impressions found as surface treatment or as designs on ceramic vessels and tobacco pipes. Many of the pottery sherds recovered in the excavations show corded surfaces, and many of the design elements consist of cord or knot impressions made into various patterns. Ms. Jirikowic examined many potsherds during her work with the collections during the summer of 1988. She was able to define the twist pattern of cordage on about 986 sherds out of 1076 examined. Her findings as to distributions of the two twist modes and their correlations with other ceramic attributes are shown in Table 7. Of the 986 sherds, 89.4% showed the Z-twist, and 10.6% had an S-twist. This preference for Z-twist cords was confirmed in an examination of potsherds from the 1983 excavations done by MacCord (1985). The sherds were examined by Dr. William C. Johnson of the University of Pittsburgh, whose report (n.d.) states (in part): To seek evidence of possible change in cordage twist patterns through time, 258 sherds of Potomac Creek Cord-impressed ware from nine levels excavated in Square D (Manson and MacCord, 1985) were examined, using latex impressions. As shown in the following, no change is apparent: 7.-twist Levels 1 2 3 4 5 6 7 8 9 Totals No. 48 26 24 22 37 21 17 22 22 239 % 87.3 92.9 88.9 100.0 94.8 100.0 94.4 91.7 91.7 92.6 S No. 1 2 3 2 1 2 2 19 twist % 12.7 07.1 11.1 05.1 05.6 08.3 08.3 07.4 Tola 55 28 27 22 39 21 18 24 24 258 Site Winslow (18Mo9) Gore (18Mo20) Frye (44Ld4) Z-twist No. % 26 86.7 9 81.8 11 84.6 S-twist No. 4 2 2 % 13.3 18.2 15.4 Totals 30 11 13 To compare the above with the possibly antecedent ceramics of the Montgomery Focus, potsherds from three Piedmont Potomac sites were similarly examined, yield? ing the following data: The similarities noted (despite the small samples from the Piedmont sites) support and do not refute or preclude a temporal tie between the tested ceramic types. More detailed comparisons and comment will appear in a forthcoming article. European Artifacts: Multiple Burial When Karl Schmitt became a member of the 1940 field party, the last to work at the Potomac Creek site, all of the European artifacts had already been found by Manson and Judge Graham. This may account for Schmitt's selection of the more numerous Indian artifacts as his thesis subject at the University of Chicago. This is not to say that he did not examine many of the European artifacts. As he had not been present as these items were excavated, he was therefore less familiar with them than he was with the artifacts of Indian origin. For this reason, I devoted much time and effort to this major category of artifacts from the work at Patawomeke. In the following pages, I will describe them, and, to the best of my ability, I will try to compare them with other finds in the region and to place the artifacts into their proper relationships. Because there were no European settlements within 100 miles of the Patawomeke village from which European-made artifacts could have come, it appears that all had come to the site via visits by explorers and traders during the early seventeenth century. Some items could have reached the villagers from sixteenth century visits to the Chesapeake Bay area by ships from Spain, France, or England, but there is no sound evidence to support such a provenience. After the 1607 settlement of Jamestown, explorers came to the Potomac River repeatedly, especially to barter trinkets for corn and skins. Most of the trade items found are ornamental, as opposed to utilitarian, items. Exceptions to this are scissors, nails, clay pipes, and glazed tiles. The lack of such items as knives, axes, guns, kettles, fish hooks, hoes, and chisels indicates an origin early in the Contact Period, roughly pre-A.D. 1620 A. FINDS BY CARL MANSON According to the Graham diary (page 11), the European objects found by Manson in Square 5R1 on 30 November 1935 consisted of a "small silver dish" and a brass "star-shaped ornament." Silver dish: This object, actually a dram cup or wine taster (Figure 40), was not acquired by Judge Graham and hence is not part of his collection that came to the National Museum. NUMBER 36 69 FIGURE 40.?Silver dram cup or wine taster recovered by Hanson from multiple burial in 1935: a, view from side; b, from below; c, from above. Bottom, enlargement of maker's mark barely visible in b. (Courtesy Carl Manson.) When photographed in the late 1930s, the bowl was deformed somewhat, the diameter of the rim varying from l15/i6 to 2!/6 inches (48 to 54 mm). The outside depth of the bowl is 5/s inch (16 mm); the maximum spread across the handles is 27/s inches (72 mm). Because the maker's mark (hallmark) impressed on the underside of the bowl (Figure 40, bottom) is clearly defined, the possibility of identifying the maker and hence the date of 70 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY >rAL tt.'s ?{Wi$