Reconstruction of Demographic Profiles from Ossuary Skeletal Samples A CASE STUDY FROM THE TIDEWATER POTOMAC Douglas H. Ubelaker SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 SERIAL PUBLICATIONS OF THE SMITHSONIAN INSTITUTION The emphasis upon publications as a means of diffusing knowledge was expressed by the first Secretary of the Smithsonian Institution. In his formal plan for the Insti- tution, Joseph Henry articulated 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 keynote of basic research has been adhered to over the years in the issuance of thousands of titles in serial publications under the Smithsonian imprint, com- mencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the following active series: Smithsonian Annals of Flight Smithsonian Contributions to Anthropology Smithsonian Contributions to Astrophysics Smithsonian Contributions to Botany Smithsonian Contributions to the Earth Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to Zoology Smithsonian Studies in History and Technology In these series, the Institution publishes original articles and monographs dealing with the research and collections of its several museums and offices and of profes- sional colleagues at other institutions of learning. These papers report newly acquired facts, synoptic interpretations of data, or original theory in specialized fields. These publications are distributed by mailing lists to libraries, laboratories, and other in- terested institutions and specialists throughout the world. Individual copies may be obtained from the Smithsonian Institution Press as long as stocks are available. S. DILLON RIPLEY Secretary Smithsonian Institution SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY . NUMBER 18 Reconstruction of Demographic Profiles from Ossuary Skeletal Samples A CASE STUDY FROM THE TIDEWATER POTOMAC Douglas H. Ubelaker SMITHSONIAN INSTITUTION PRESS CITY OF WASHINGTON 1974 ABSTRACT Ubelaker, Douglas H. Reconstruction of Demographic Profiles from Ossuary Skeletal Samples: A Case Study from the Tidewater Potomac. Smithsonian Contributions to Anthropology, number 18, 79 pages, 27 figures, 45 tables, 1974.—The excavation and analysis of two Late Woodland ossuaries from the Juhle site (18CH89) in southern Maryland are described in detail. The report includes a discussion of archeological features of the ossuaries, but emphasizes the reconstruction of population profiles derived from the analysis of the recovered skeletal samples. Ethnohistorical and arche- ological sources are consulted to suggest that ossuaries contain nearly all individuals who died in the contributing populations during culturally prescribed numbers of years and, consequently, offer somewhat unique opportunities for demographic analysis. Several methods are employed to estimate the chronological age at death of indi- viduals in both ossuaries. Subadult ages are derived from the formation and eruption of the teeth and from the maximum length of the femora. Adult ages are calculated from examinations of the symphyseal faces of the pubes and the degree of microscopic cortical remodeling in the femora. The latter method involved the preparation of 151 ground thin sections taken from the anterior cortices of the right femora, and it repre- sents the first application of Kerley's relatively new method (1965) to a large arche- ological population. The resulting death curves are compared and the methods evaluated. Data from the most reliable of these age-determinative methods are used to calculate curves of mortality and survivorship, life tables, and crude mortality rates for the populations represented. Population estimates are attempted by utilizing the crude mortality rates (calcu- lated from the life tables), the length of time represented by each ossuary (calculated from archeological data), and the total numbers of individuals in the ossuaries. The resulting population size estimates are considered against both archeological and ethnohistorical data to suggest the nature of the sociopolitical unit serviced by the ossuaries. Finally, both local and regional population-size estimates are compared with those estimated by others using different types of data. OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Smithsonian Year. SI PRESS NUMBER 4958. Library of Congress Cataloging in Publication Data Ubelaker, Douglas H. Reconstruction of demographic profiles from ossuary skeletal samples. (Smithsonian contributions to anthropology, no. 18) Bibliography: p. 1. Indians of North America—Maryland—Anthropometry. 2. Juhle site, Md. I. Title. II. GN1.S54 no. 18 [E78.M3] 573'.6'09701 73-16117 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402 - Price $2.25 (paper cover) Stock Number 4700-00287 Preface In August 1953, Mr. Bernwald Juhle and his son Hans-Bodo of Nanjemoy, Charles County, Maryland, discovered a concentration of human bone while they were connecting a water pipe to the family home on Friendship Farm. The Juhle's keen interest in local history and concern over the potential importance of the find led them to cover the remains carefully to prevent further dis- turbance and to call the Smithsonian Institution. At the Smithsonian, the call was directed to T. Dale Stewart, physical anthropologist in the Department of Anthropology of the National Museum of Natural History, who responded with a visit to the Juhle farm. His preliminary examination of the exposed bones revealed that an apparent ossuary had been found. Stewart returned to the ossuary with Marshall T. Newman of the Smithsonian Institution on 29 August 1953 and began the excava- tion. Unfortunately, the water pipe installed by the Juhles bisected the ossuary, dividing it into an eastern two-thirds section and a western one-third section. Since the water pipe could not be removed, Stewart and Newman were obliged to dig around it. Beginning with the eastern section, they worked as steadily as weather would permit and by 1 November had succeeded in removing all the skeletal material from the eastern two-thirds. They then closed the excavation for the season. Stewart was unable to return to the excavation until 29 August 1955. At that time, he removed the topsoil from the western third and continued the 1953 procedure of exposing, recording, mapping, and removing the skeletal material. On 29 September Stewart completed the excavation, backfilled the pit and returned to Washington with the complete ossuary skeletal collection. In the spring of 1971, eighteen years after the discovery of Ossuary I, Fredrich Berthold (Pete) Juhle was digging a post hole for a planned fence line on his mother's farm, when his auger struck bone. Recalling the excavation of the first ossuary during his childhood, he realized that the bones were probably human. Thereupon, he carefully covered the exposed bones with plastic, refilled the hole and finished the fence, without installing a post at that point. Soon thereafter, his mother noti- fied T. Dale Stewart at the Smithsonian of the discovery. At that time, I had just joined the staff at the Department of Anthropology, National Museum of Natural History, and accepted Stewart's cordial invitation to join him in investigating the discovery. On 10 May 1971, Stewart and I visited the Juhle Farm, confirmed "Pete'' Juhle's diagnosis that the bones were human, and began removing the topsoil to disclose the extent of the bone deposit. Since we were aided by numerous volunteers from Washington, D.C., and a grant from the Smithsonian Institution's Hrdlicka Fund, progress continued virtually uninterrupted until 1 June. At that time I was forced to leave the excavation to direct an eight-week project in South Dakota sponsored by the National Geographic Society and the Smithsonian Institution. During my absence, T. Dale Stewart assisted by David Ruzek of Washington, D.C., continued the excavation. Accompanied by James Yost of the University of Kansas Medical Center, I rejoined Stewart and Ruzek at the ossuary on 9 August 1971. By 16 August 1971, we suc- ceeded in removing the bones from all but the northwest quarter of the ossuary. Again aided by volunteers from Washington, D.C., and financial support from the Smithsonian Institution's Hrdlicka Fund, Stewart and I returned to the ossuary excavation on 16 May 1972. By 9 June 1972 all of the bones from this final quarter of Ossuary II had been removed and the pit refilled. As is usually the case with a project of this magnitude, many individuals offered their assistance and encouragement. First and foremost, I extend my appreciation to T. Dale Stewart, Physical Anthropologist Emeritus at the Smithsonian Institution. Dr. Stewart provided the initial encourage- ment to undertake this research problem, generously placed his original, unpublished data at my disposal, enthusiastically followed the progress of my research, and read many versions of the manu- script. His continued interest and guidance have been instrumental in my completion of this mono- graph and I am grateful for his generosity. Excavation of Nanjemoy Ossuaries: left, Ossuary I (1953). From top to bottom: Mrs. Lisa Juhle; husband, Bernwald; son, Hans-Bodo; niece, Hilda Karlsson; son, Fredrich Berthold; dog, Hero; T. Dale Stewart, right, Ossuary II (1971). Standing from left to right: Mrs. Marty Juhle; Fredrich Berthold (Pete) Juhle; Mrs. Lisa Juhle. Foreground: author. Numerous other colleagues at the Smithsonian have contributed substantially to the development of my ideas and general research design. In particular I wish to thank Clifford Evans, Chairman, Department of Anthropology, and Betty J. Meggers, Research Associate, for spending many Saturday afternoon coffee breaks listening to my interpretations of research results, as well as for reading and offering suggestions on various drafts of the manuscript. Waldo and Mildred Wedel also read the final draft and offered many helpful suggestions. Many others at the Smithsonian have been helpful. In particular I wish to thank J. Lawrence Angel for his advice on techniques of demographic recon- struction; Christian Feest and William Merrill for many conversations on the ethnography of the Maryland and Virginia Indians; Donald J. Ortner for making his research histology laboratory avail- able and for his advice on thin-section preparation; Joseph P. E. Morrison for identifying the shell, George E. Phebus and Clifford Evans for identifying the pottery, George R. Lewis for preparing the illustrations, Victor Krantz and the Smithsonian Division of Photographic Services for preparing the photographs, Neil Roth for his assistance in the statistical computations, Joan Horn for excellent editorial advice and assistance in leading me from manuscript to final published monograph without too many pitfalls. Several individuals outside of the Smithsonian Institution have contributed significantly to this research. I especially wish to thank B. Miles Gilbert of the University of Missouri for his identification of the faunal material, Clark S. Larsen of Kansas State University, Pamela L. Horn of Springfield, Virginia, and Nicki Horton of Washington, D.C., for their assistance in specimen processing, and Marnie Briggs of Washington, D.C., for her substantially underpaid assistance in thin-section prepara- tion and osteological analysis. David C. Ruzek of Austin, Minnesota, David Frayer of the University of Michigan, Richard Stewart of Silver Spring, Maryland, and Katchie McQuilkin of Washington, D.C., all contributed many hours of unpaid assistance during excavation. Dr. Carlyle S. Smith and Dr. Michael Crawford of the University of Kansas both read a final draft of the manuscript and offered many helpful suggestions. Finally I would like to thank Mrs. Lisa Juhle of Nanjemoy, Maryland, not only for allowing us to excavate on her property and disrupt her summer routine, but also for continually offering us her cold lemonade, delicious desserts, and gracious hospitality on those hot summer afternoons. It was Mrs. Juhle's immediate report of the discovery of an undisturbed ossuary that made this entire investigation possible. Consequently, this monograph is dedicated to Mrs. Lisa Juhle and her concern for local history. DOUGLAS H. UBELAKER Smithsonian Institution Washington, D.C. 30 June 1973 vn Contents METHODS OF POPULATION RECONSTRUCTION Ethnohistorical Methods Tribe-by-Tribe Inventories Death Rates Population Projection Sociopolitical Reconstruction. Depopulation Rates Archeological Approaches Site Surface Area House and Mound Frequency Settlement Patterns Ecological Resource Potential Physical Anthropological Approaches Application in Old World Application in New World . Research Model of Present Study... EVIDENCE OF OSSUARY BURIAL FROM EASTERN NORTH AMERICA Ethnography of Ossuary Practice The Huron Southeastern United States Mid-Atlantic Area of the United States Archeological Evidence THE OSSUARIES FROM NANJEMOY CREEK Ossuary I ... Ossuary II: Excavation Procedure Ossuary II: Contents Animal Bone Shell Pottery Human Remains Total Number of Individuals. Ossuary I Ossuary II Variability in Bone Representation Spatial Distribution Features within Ossuary II. Summary DETERMINATION OF SEX AND AGE AT DEATH Determination of Sex Determination of Age of Subadults Length of Long Bones. Epiphyseal Union Dental Eruption and Calcification Determination of Age of Adults Pubic Symphysis Internal Bone Remodeling Specimen Preparation Osteon Counts Page 1 2 2 2 2 3 3 4 4 4 4 5 5 5 6 7 8 8 8 10 10 11 15 15 15 19 22 26 27 27 31 32 33 33 37 40 41 41 42 42 43 44 51 52 53 54 56 IX Page RECONSTRUCTION OF DEMOGRAPHIC PROFILES AND POPULATION SIZE 59 Mortality Curves 59 Survivorship Curves 60 Life Tables 60 Crude Mortality Rates 65 Population Reconstruction 65 Regional Population Estimates 68 LITERATURE CITED 71 Tables 1. Estimates of New World population size in 1492 . 1 2. Distribution of bone weight within Ossuary II 22 3. Spatial distribution of bones of turkey {Meleagria gallopavo) within Ossuary II 22 4. Spatial distribution of bones of white-tailed deer {Odocoileus virginianus) within Ossuary II 26 5. Frequencies of ceramic types in ossuaries and habitation area 27 6. Frequencies of articulated adult bones in Ossuary II 31 7. Spatial distribution of burned bone within Ossuary II 31 8. Frequencies of skull positions in Ossuary II 31 9. Frequencies of face orientations in Ossuary II 31 10. Counts of complete and fragmentary adult long limb bones from Ossuary II 32 11. Number of individuals represented by each type of adult bone in Ossuary I 32 12. Number of individuals represented by each type of subadult bone in Ossuary I. ... 33 13. Number of individuals represented by each type of subadult bone in Ossuary II. ... 33 14. Number of individuals represented by each type of adult bone in Ossuary II 34 15. Order of representation of adults in Ossuary I as indicated by bone types 35 16. Order of representation of subadults in Ossuary I as indicated by bone types 36 17. Order of representation of adults in Ossuary II as indicated by bone types 36 18. Order of representation of subadults in Ossuary II as indicated by bone types 37 19. Partial inventory of bones from Fairty Ossuary, Ontario. Canada 37 20. Chi square values for distributional associations between young and old adults within Ossuary II 38 21. Chi square values for distributional associations between adult males and females within Ossuary II 38 22. Chi square values for distributional associations between adult and subadult bones within Ossuary II 38 23. Chi square values for distributional associations between the adult femora and other adult bones within Ossuary II 39 24. Chi square values for distributional associations between the subadult clavicles and other subadult bones within Ossuary II 39 25. Sex representation of adults in Ossuary 1 42 26. Sex representation of adults in Ossuary II 42 27. Maximum diaphyseal length of subadult femora from Ossuary 1 44 28. Maximum diaphyseal length of subadult femora from Ossuary II 44 29. Age distribution of subadult femora from Ossuaries I and II as determined from max- imum diaphyseal length and epiphyseal union 45 30. Age distribution of subadults in Ossuary I as determined from crown-root calcification and tooth eruption. 46 31. Age distribution of subadults in Ossuary II as determined from crown-root calcification and tooth eruption 47 32. Age distribution of adults as determined from pubes in Ossuary 1 52 33. Age distribution of adults as determined from pubes in Ossuary II 53 34. Percentages of remodeled bone and corresponding age estimates for 66 right femora from Ossuary 1 56 Page 35. Percentages of remodeled bone and corresponding age estimates for 84 right femora from Ossuary II 57 36. Age distribution of adults in Ossuaries I and II as estimated from femoral microscopic remodeling 57 37. Age distribution of subadults and adults from Ossuaries I and II as estimated from femoral length and epiphyseal union (0-20 years) and femoral microscopic remodeling (20+ years) 59 38. Life table reconstructed from Ossuary I age distribution 62 39. Life table reconstructed from Ossuary II age distribution 63 40. Life expectancy at birth in 15 world-wide populations 64 41. Correlation of ages at death with known annual death rates 65 42. Death rate estimates from eight skeletal populations 66 43. A comparison of possible time intervals (T) represented by Ossuary II, assuming different rates of decomposition (t), as calculated by the formula T= .42 t 66 44. A comparison of possible population sizes assuming different time intervals, as calculated from Ossuary II by the formula P= 67 M T 45. A comparison of possible population sizes assuming different time intervals, as calculated De° from Ossuary II by the formula P=KH 67 Figures Excavation of Nanjemoy ossuaries vi 1. Geographic distribution of reported ossuaries in the region of the Chesapeake and Delaware bays 12 2. Location of Juhle site near Nanjemoy Creek, Charles County, Maryland 16 3. Nanjemoy Creek, as seen from Juhle site, looking southwest 17 4. Detailed map of the Juhle site, Charles County, Maryland 18 5. Excavation of Ossuary I 19 6. The 0.6-meter square system utilized in the 1971 excavation of Ossuary II 20 7. Division of Ossuary II into northeast and southwest halves, looking northwest 20 8. Excavators working in square DL1, Ossuary II 21 9. Excavator lying on a board-support to work in square GR1, Ossuary II, looking south- west 22 10. Bone concentrations in Ossuary II 23 11. Bone concentrations in Ossuary II with extended skeleton 24 12. Northeastern extremity of Ossuary II 25 13. Profiles of Ossuary II 26 14. Articulated adult skeletons on the floor of Ossuary II, looking northwest 28 15. Articulated adult skeletons on the floor of Ossuary II ... 29 16. Examples of partial articulation and bundles in Ossuary II 30 17. Comparison of femoral head diameters from Ossuary I with those of Ossuary II 43 18. Comparison of age distributions from Ossuary I 46 19. Comparison of age distributions from Ossuary I.. . 47 20. Comparison of age distributions from Ossuary II 48 21. Comparison of age distributions from Ossuary II 49 22. Comparison of age distributions from Ossuary I 50 23. Comparison of age distributions from Ossuary II 51 24. Cross-section of human cortical bone 54 25. Comparison of adult mortality curves from Ossuaries I and II 58 26. Comparison of mortality curves calculated from Ossuaries I and II 60 27. Comparison of survivorship curves calculated from Ossuaries I and II 61 XI Methods of Population Reconstruction An examination of the literature dealing with the cen- tury-old problem of New World pre-contact population size reveals a tremendous range of estimates (Table 1). At the lower end of this range are hemispheric estimates of 8,400,- 000 by Kroeber (1939:166) and 13,385,000 by Rosenblat (1945:92) and at the upper end is Dobyns' figure of 90,043,000 (1966a: 415). Such widely fluctuating estimates naturally include contradictions. Thus, while Las Casas (in MacNutt, 1909:317) in 1541 believed that 15,000,000 In- dians died in the West Indies during the 40 year period after 1500, Rosenblat (1954:102) calculated that in 1492 only 13,385,000 Indians were alive in the entire western hemis- phere. Whereas Kroeber's estimate works out to a population density of only one person per 5 square kilometers (0.2 per km2), Borah (1962b: 179) suggested a density of 12 persons per 5 square kilometers (2.4 per km2). As the dates of the cited estimates indicate, the trend is now away from the earlier conservatism in estimation, but even the modern es- timates clearly retain a considerable amount of subjective interpretation. The current emphasis in American archeology on prob- lems of prehistoric settlement patterns, cultural ecology, and aboriginal demography has shifted attention to this old prob- Douglas II. Ubelaker, Department of Anthropology, National Mu- seum of Natural History, Smithsonian Institution, Washington, D.C. 20560. lem of prehistoric population size and accentuated the need for new and more accurate methods of estimation. Accurate evaluations of population size, density, and structure have become increasingly essential to the interpretation of pop- ulation-environmental relations and to the attempt to reach an understanding of prehistoric cultural process. While the attention recently given to new attempts to estimate popula- tion size attests to the need for better estimates, the need is hardly fulfilled. As Macgowan and Hester (1965:5) state, "it is anybody's guess" just which estimates are the most ac- curate in the absence of more refined technique. The wide range of population estimates reflects to a large extent the variety of methodologies employed. Each of the subdisciplines of anthropology has utilized different data and methods to reconstruct population size. In ethnohistory, scholars have turned to early descriptions of warrior counts, tax rolls, mission records, depopulation rates, and social organization. Archeologists have examined village patterns, house numbers and sizes, refuse mound composition, and pottery distributions. Physical anthropologists have recon- structed death rates and estimated population size from skeletal populations. In short, specialists have consulted a wide variety of data to produce estimates for large geo- graphic areas through two general methods: (1) by totaling estimates for each group within an area, and (2) by project- ing density estimates calculated from a single well-studied TABLE 1.- —Estimates of New World population size in 1492* Area Sapper {1924) Kroeber (1939) Rosenblat (1945) Steward (1949) Dobyns (1966) 2, 000, 000- 3, 500, 000 12,000,000-15,000,000 3, 000, 000- 4, 000, 000 5, 000, 000- 6, 000, 000 900, 000 3, 300, 000 200,000 »* 1,000,000 4, 500, 000 300, 000 800,000 1,000,880 4, 500, 000 225, 000 736,000 9 800 000 30, 000, 000 443,000 10, 800, 000 22, 000, 000-28, 500, 000 - 6, 600, 000 6,461,880 51,043,000 12,000,000-15,000,000 3, 000, 000- 5, 000, 000 3, 000, 000 1,000,000 4, 750, 000 2, 035, 000 6, 131,000 2, 898, 000 30, 000, 000 9, 000, 000 15,000,000-20,000,000 - 6, 785, 000 9, 029, 000 39, 000, 000 37, 000, 000-48, 500, 000 8, 400, 000 13,385,000 15,490,880 90, 043, 000 ♦Modified from Steward (1949:656). •♦Included in South American and Mexican Estimates. 525-413 O - 74 - 2 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 8 group within the area. Of course, each method has its own advantages and disadvantages. Perhaps the best way to evaluate the merits of individual methods is to examine them in detail, beginning with those that rely upon ethnohistorical data. Ethnohistorical Methods Although all ethnohistorical methods of calculating popu- lation size and structure depend upon early historical records for their basic data, they vary considerably in the type of historic data used, the manner in which population statistics are derived, and the level of population estimate attempted. In general, as already noted above, scholars have attempted to reach totals for large geographic areas by tabulating esti- mates for each constituent tribal group, or by projecting density calculations from smaller, well-studied areas. For small geographic areas, estimates have been derived from death rates, mission records, tax assessments, sociopolitical organization, and depopulation rates. The following compre- hensive, but not all-inclusive discussion examines some of the most frequently quoted ethnohistorical attempts at popula- tion estimation. Early estimates (prior to Mooney, 1928) are not examined in detail since they are of questionable value and are seldom quoted (for example, Schoolcraft, 1851: Lefroy, 1853). TRIBE-BY-TRIBE INVENTORIES Perhaps the most frequently quoted population estimates in North America are those of James Mooney (1910). Hav- ing been requested to write on the topic "population" for the "Handbook of American Indians North of Mexico," Mooney undertook an intensive review of the evidence for American Indian population size at the time of European contact. His interest and scholarship carried him far beyond the require- ments of the "Handbook" article and into an exhaustive study of early ethnographic accounts. He used a "tribe-by- tribe" accumulative approach, taking estimates from all available sources, including original census rolls. Unfortu- nately, by the time of his death in 1921 Mooney had not completed his proposed exhaustive treatise on aboriginal population size and its decline after European contact. The only published records of this work are to be found in the brief article in the "Handbook" and in his (Mooney, 1928) posthumously published monograph entitled "The Aborig- inal Population of America North of Mexico." Although Mooney did not explain in detail his methods for computing each tribal estimate, he used a direct ethnohistorical ap- proach for all areas except California, where he relied upon estimates by C. Hart Merriam (1905), calculated from mis- sion to nonmission population ratios. In 1939, Kroeber published his monumental Cultural and Natural Areas of Native North America. A major thrust of this work involves a calculation of population density at the time of contact for each of the cultural areas he defined. All of the population estimates utilized in this calculation were Mooney's except for the California estimate, where Kroeber used his own data. Although Kroeber felt Mooney's esti- mates were slightly high for some areas, he (1939:134) noted that "all in all, however, Mooney's estimates and computations have clearly been made on the basis of wide reading, conscientiousness, and experienced judgment. Until some new equally systematic, and detailed survey is made, it seems best to accept his figures in toto rather than to patch them here and there." Kroeber (1939:131) substituted his own figures on the California population over Merriam's, feeling that "my total is arrived at through a tribe-by-tribe addition or 'dead-reckoning' method, like all Mooney's other figures; whereas Merriam uses a mission to nonmission area multiplication ratio for the state as a whole." Rosenblat (1935, 1945, 1954) employed a similar ap- proach to Mooney's and Kroeber's in estimating 13,385,000 Indians in the western hemisphere at the time of contact (1492). Even though Rosenblat's estimates for North Amer- ica are higher than those of Mooney and Kroeber, they would still have to be considered conservative. DEATH RATES Although the "dead-reckoned" estimates of Mooney, Kroeber, and Rosenblat are the most frequently quoted, numerous other attempts have been made using different types of ethnohistorical data. Fray Bartolome de Las Casas (in MacNutt, 1909:317) wrote in 1541, "We give as a real and true reckoning, that in the said forty years [1500-1540], more than twelve million persons, men, and women, and children, have perished unjustly and through tyranny, by the infernal deeds and tyranny of the Christians; and I truly believe, nor think I am deceived, that it is more than fif- teen." Las Casas estimated that in 40 years, gold-seeking Christians killed nearly twice as many Indians as Kroeber believed existed in the entire New World. Of course, as Kunstadter (1966) has argued, these statistics would not be overly surprising even assuming a normal death rate for the population. A death rate of 40 per 1000 would indicate that over one and one-half times as many people would die dur- ing a 40 year period as would be alive at the beginning of the period. During the Spanish Conquest, the death rate would have been much higher. However, since Las Casas' mortality estimates were undoubtedly exaggerated; they have not been used extensively in modern population recon- struction. POPULATION PROJECTION MacLeod (1928:15-16) expressed some dissatisfaction with the estimates of Mooney and Kroeber and produced estimates of his own for the Maryland, Virginia, and Dela- ware area. Although MacLeod apparently utilized virtually the same sources as Mooney, he derived consistently higher METHODS OF POPULATION RECONSTRUCTION estimates. After producing what he felt was a reliable esti- mate for the mid-Atlantic area, he computed the popula- tion density and then projected his figures toward an esti- mate of three million for North America north of Mexico and 15 million for Central and South America. Many other investigators have attempted similar ethno- graphic-reconstruction projection methods for calculating population size (Aschman, 1959; Means, 1931, Meigs, 1935; Merriam, 1905; and Sauer, 1935, 1966). Merriam (1905) relied upon early mission records prior to 1834, which pro- vided relatively accurate accounts for the strip of California south of San Francisco that was heavily missionized. Mer- riam calculated the total from die missions, added a factor of 25 percent to allow for unbaptized Indians, and then multiplied by five for an estimate of the total population of California. Sauer (1935) estimated the aboriginal population size of northwestern Mexico from a careful examination of early Jesuit records. According to Sauer (1935:2), "total popula- tion was rarely taken and population figures usually must be arrived at by converting warriors, families, baptisms, or other items of the minister's record." The utilization of various types of data provided Sauer with several crosschecks on his population estimates, which certainly increased their relia- bility. Although Kroeber (1939:177) accepted Sauer's work as "revolutionizing," he apparently rejected his results: ". . it is difficult to meet Sauer's citations of seventeenth-century figures. . . ." Aschman (1959), Borah (1962a, 1962b), Borah and Cook (1963), Cook (1937, 1939, 1940, 1943, 1946a, 1946b, 1955a, 1955b, 1956, 1958), Cook and Borah (1957), Cook and Simpson (1948), and Meigs (1935) also attempted population reconstruction using early ethnographic docu- ments. Meigs (1935) turned to early mission accounts for population estimation in lower California, while Aschman (1959:147) used what he calls the "additive method" to compute population counts from individual mission roles in each of the six Baja California territories. In a similar man- ner, Borah, Cook, and Simpson at various times considered population estimates from a multitude of ethnographic sources in Mexico and California. Borah and Cook (1963) utilized tribute data recorded as early tax assessments. They broke the assessed tribute of each Province into terms of annual value in mantles and fanegas of grain and beans. The number of assessed families was calculated and multiplied by a factor of 4.5 to estimate the number of persons in the population. In short, their work (Borah and Cook, 1963:5) represents "an attempt to use pre-Conquest fiscal material for estimates of the pre-Conquest population of Central Mexico, as we applied fiscal and other administrative mate- rial (that is, Spanish tribute assessment, counts, and parish or missionary reports) for our estimates of post-Conquest population." SOCIOPOLITICAL RECONSTRUCTION Means (1931) reconstructed population size for the Inca Empire through an innovative analysis of Inca sociopolitical organization. Utilizing Spanish descriptions of Indian ac- counts of pre-Conquest Inca political hierarchy, Means was able to reconstruct the entire Inca political organization and subsequently derive an overall population estimate from the relatively ordered individual political units. Although Baudin (1961) and Mason (1957) have suggested that the Inca system was actually less well-defined, Means still can be credited with an admirable attempt to reconstruct popula- tion size. Of course, this approach has little applicability out- side of Peru. DEPOPULATION RATES All of the foregoing studies examined the earliest Euro- pean records to produce three types of population estimates: (1) estimates for specific groups in small geographic areas (Sauer, Meigs, Aschmann, Brau, Cook et al., Means), (2) estimates for large geographic areas by a "tribe-by-tribe" additive approach (Mooney, Rosenblat, Kroeber), or (3) estimates for large geographic areas by a projection of den- sity calculations from small areas (Merriam, MacLeod). A final group of scholars calculated aboriginal population size directly by considering the rate at which population numbers were being diminished as a consequence of European con- tact. Of these, the studies of Dobyns (1962, 1963a, 1963b, 1966a, 1966b), Kubler (1942, 1946, 1948) and Rivet (1924) are the most noteworthy. Rivet (1924) began his population estimate procedure with a comparison of the number of living Indians to the estimated number of pre-contact Indians in North America. He then assumed that the same "rate of depopulation" was generally applicable throughout the western hemisphere and calculated the aboriginal number from counts of the living. The method has been criticized by Kroeber (1939:160) as incorporating faulty logic and by Dobyns (1966a: 399) as relying upon incorrect counts of both living Indians and aboriginal population (Mooney's estimates). Kubler (1942, 1946, 1948) compiled population estimates from census records, reports, tax rolls, etc. for Mexico. Since Kubler's data were based on historic documents, he con- fined his study to careful examination of population shifts during historic times only and avoided the temptation to project his rates into prehistory. Dobyns (1962, 1963a, 1963b) also began with the rate of depopulation during historic times, but then extended his depopulation rates into prehistory. In his monumental (1966a) study, he very carefully examined the rate of de- population in different parts of the western hemisphere, pro- ducing a general, standard depopualtion ratio of 20 to one, i.e., 20 aboriginal Indians to one at the nadir or lowest point of population decline. The application of this ratio SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 to the entire western hemisphere produced an estimate of 90,043,000 (2.1 per km2). The Dobyns ratio has been criti- cized by scholars (Bennett, 1966; Bernal, 1966; Blasi, 1966; Deneven, 1966; Forbis, 1966; Kehoe and Kehoe, 1966; Kundstadter, 1966), who generally distrust his calculation of the nadir population and emphasize the variability of depop- ulation rates within the western hemisphere. The above review of ethnohistorical approaches to popu- lation estimation is by no means complete. It suffices how- ever, to document both the diversity of methodology and the similarity of assumptions employed in ethnohistorical reconstruction. Archeological Approaches Many Americanist anthropologists have criticized ethno- historical methods of estimating population size, claiming they attribute excessive objectivity to early observers. Others have claimed that even if ethnohistorical estimates are ac- curate, they do not represent the prehistoric populations and there is no reliable way to determine the difference. Conse- quently, many scholars have turned to prehistoric archeol- ogical data to produce hopefully more reliable population estimates. Archeologists have estimated population size from site surface area, frequencies of mounds and houses, settle- ment patterns in specific areas, and regional ecological potentials. Each of these approaches produces population estimates or density statistics and each has its own assump- tions and limitations. SITE SURFACE AREA Cook and Treganza (1950) suggested a correlation be- tween population size and total surface area of the site. Their study produced a general formula for estimating popula- tion size: log population = constantX log area (in square meters). This basic formula was then supplemented with data on area, volume, density, and mass from many sites. According to Cook and Treganza (1950:231), "For the estimate of the number of inhabitants living on a prehistoric site, there are no concrete data whatever. We have to de- pend upon our knowledge of habitation methods, family number, and many purely subjective criteria. Nevertheless, there is no good reason to suppose that guesses of this sort are seriously in error when they are supported by a careful study of a specific site and an adequate background in gen- eral ethnology." HOUSE AND MOUND FREQUENCY Investigations of house and mound frequencies at indi- vidual sites have produced several local population esti- mates (Ascher, 1959; Haviland, 1965, 1966; Nelson, 1909). Differences between individual estimates stem from several sources of variability such as the assessment of the number of houses occupied at any one time, the length of time of village occupation, and the number of persons per house. Haviland's recent work (1970) has demonstrated that a figure of five persons per house at the Mayan site of Tikal is reasonably correct, whereas the length of occupation and number of houses in use at one time remain questionable and highly variable from site to site. In the Mayan area, Brain- erd (1956) concluded that only one house in eight had been occupied, Morley (1947) felt they had all been occupied, while Ricketson and Ricketson (1937) put the figure at one in four. In the Plains, by contrast, Wood (1967) estimated 10 individuals per Mandan earthlodge, with nine out of 10 lodges being occupied simultaneously at the Huff site in South Dakota. SETTLEMENT PATTERNS Numerous recent studies have concentrated on the settle- ment patterns and site frequencies within regional areas. Bullard (1960), Dumond (1972), Griffin (1956), Heizer and Baumhoff (1956), Rouse (1952), Sanders (1953, 1956, 1960, 1963), Termer (1951), Welte (1966), and Willey, et al. (1965) represent a few of the excellent studies of site densities. They all supply a wealth of information about their respective areas, but are subject to the same general limita- tions as estimates from individual sites. These settlement studies consider the number and size of sites within a given area, but usually employ broad periods of time. For example, Coe and Flannery (1967) discuss population trends in the Ocos area of south coastal Guatemala. Their approach involves tabulating the total number of site components rep- resenting each temporal period. The resulting table demon- strates a population decline from Middle to Late Formative, a rapid increase to Late Classic, and then another dramatic decline in the Post-Classic period. This type of presentation is very difficult to interpret since the time intervals are not equal, and there are no data on the number of components occupied contemporaneously. Their evidence of maximum population size during the Late Classic could indicate simply a longer time span or that the populations were more mobile. In the absence of detailed ceramic seriation studies or at- tribute analyses, the question of simultaneous component occupation remains hypothetical. Schwartz (1956) introduced a method of studying surface pottery collections from 104 sites identified as Cohonina in Arizona. He used the plain (San Francisco-Mountain-gray) pottery as an index to delineate the affiliations of the sites and the painted pottery to provide the time scale. Schwartz (1956:28) then plotted the number of sites occupied for each 25 year period, assuming that "the relative increase or decrease in habitation units bears a direct relationship to the rise and fall of population without the necessity for taking into consideration the exact numbers of peoples. The simi- larity in size of the sites throughout the time span makes this METHODS OF POPULATION RECONSTRUCTION type of comparison possible. The analysis, then, is based on settlement density and its fluctuations." His results show a steady increase in population growth from A.D. 600 to 900, leveling off from A.D. 900 to 1100, a rapid decline between A.D. 1100 and 1200, and a disappearance of Cohonina sites after A.D. 1200. Schwartz' evidence for population decline beginning A.D. 1100 corroborates similar data to the north (Hall, 1942) and to the east (Colton, 1936, 1939). ECOLOGICAL RESOURCE POTENTIAL H.P. Thompson (1966) adopted an ecological approach to the problem of population size. He began with an exam- ination of the aboriginal Chipewyan Indian utilization of barren-ground caribou and calculated the upper and lower limits of the Chipewyan population that could be supported by the caribou. The technique has been criticized by Cook (1966) and others, but remains an innovative attempt at population reconstruction. Others have considered an ecological approach by calcu- lating the resource potential of a geographic area (Allan, 1965; Sapper, 1924; J.E.S. Thompson, 1951). For exam- ple, Cowgill (1962:283) studied swidden farming in the Maya lowlands, concluding that the area could support 100 to 200 persons per square mile (63-126 per km2). Havi- land's archeological data suggest an even higher density (Haviland, 1966), but even if Cowgill's data are correct, it by no means follows that the population density actually reached its maximum potential. Cowgill merely estimated what population numbers could have lived in the area. Although the archeological approaches discussed above consult a variety of different data, they all focus on popu- lation structure within small geographic areas, and avoid estimates for the entire hemisphere. Collectively, these studies document the diversity of population size and den- sity within the western hemisphere and stress the importance of developing accurate estimates for small areas. Physical Anthropological Approaches Population estimates from skeletal studies are potentially the most accurate of all. The number of skeletons found in cemeteries is not dependent upon the number of houses oc- cupied in the village, the number of individuals per house, the accuracy of ethnohistorical accounts, or the resource potential of the area. Unfortunately, however, adequate samples are hard to come by and physical anthropological techniques have their own unique limitations and require- ments. Problems and prerequisites in determining accurate demographic information from skeletal populations have been summarized recently by a number of authors (Angel, 1969a, 1969b; Acsadi and Nemeskeri, 1970; Genoves, 1963; Howells, 1960; Swedlund and Armelagos, 1969; Vallois, 1960). In general, studies from skeletal populations demand: (1) a knowledge of the completeness of the sample; (2) information about the archeological associations of the skele- tons; (3) a determination of the length of time the sample represents; (4) an adequate assessment of sex and age at death, and (5) a proper selection of demographic method- ology. If all of the above requirements are met, then skeletal populations may offer the most accurate data of all (Hausler, 1966; Stloukal, 1966). However, as Dobyns (1966b) and many others have emphasized, these conditions rarely are met, especially in the New World. APPLICATION IN OLD WORLD Acsadi and Nemeskeri (1970) recently summarized the development of paleodemographic research in the Old World. They traced the first analyses to early epitaph studies by Beloch (1886), Motta (1891), and Harkness (1896). Beloch and Harkness then influenced the statistician Pear- son (1901-1902) who used sex and age data from an Egyp- tian mummy series to reconstruct life expectancy and mortality. Following Todd's studies of skeletal age changes in the 1920s and 1930s, numerous investigators began shifting their attention to problems in paleodemography. Epitaph studies continued through the work of Gomme (1933) for 4th and 5th century B.C Athens, Richardson (1933) for ancient Greece as a whole, Valaoras (1938) for life expectancy in 4th century B.C. Greece, Wilcox (1938) for length of life in the Roman Empire, Hombert and Preaux (1945) for popula- tions in Greco-Roman Egypt, Etienne (1955) for 1 to 300 A.D. Bordeaux, France, and Szilagyi (1959) for Roman-era Hun- gary. Using Todd's (1920, 1921, 1929, 1930a, 1930b, 1931) published age standards, numerous authors studied demo- graphy from skeletal populations in the Old World. Nougier (1949, 1950, 1954, 1959) utilized the data of others to trace the development of prehistoric populations, especially those in France. Franz and Winkler (1936) studied the demo- graphy of the Bronze Age in Lower Austria. Vallois (1937) reconstructed length of life from Paleolithic and Mesolithic skeletons. Weidenreich (1939) extended demographic anal- ysis to earlier fossil populations. A number of others utilized skeletal samples in regional demographic studies, e.g., Riquet (1953), Giot (1951) Gerhardt (1953), Ferembach (1960, 1961, 1962), Fuste (1954, 1955), Broste and Jorgensen (1956), Cunha (1956), Gejvall (1960), Ivanicek (1951), Kurth (1955), and Stloukal (1962). In addition Acsadi and Nemeskeri (1957a, 1957b, 1958, 1959, 1960, 1970) and Angel (1947, 1948, 1951a, 1951b, 1953, 1954, 1957, 1959, 1968, 1969a, 1969b) have published a series of articles that both develop theory and methodology, as well as apply methodology to excavated skeletal samples. The large number of studies that successfully deal with demography of skeletal populations in the Old World re- flects to a great extent the accurate documentation that exists SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 8 for many Old World collections. Due to the greater antiquity of written records in the Old World, they are often availa- ble to supplement information derived from archeological features of the skeletons themselves. Such sources reduce the variables of an analysis, strengthen the sample, and ul- timately assure the accuracy of the investigator's conclusions. APPLICATION IN NEW WORLD In the New World, demographic studies of prehistoric skeletal populations have remained largely theoretical (e.g., Weiss, 1973). Perhaps for a lack of adequate skeletal sam- ples, Americanist physical anthropologists have been reluc- tant to attempt demographic reconstruction. As Dobyns (1966a), Howells (1960), and Vallois (1960) have indi- cated, numerous cultural factors (known and/or unknown) seriously limit the extent to which New World samples can be utilized. The tremendous New World cultural variation in mortuary practices continually raises the possibility of in- completeness in skeletal samples, due to such factors as selective cremation, trophy taking, death away from the village, and differential status burial. Since in most cases the scope of the mortuary complex is unknown, the assump- tion cannot be made that the skeletal sample is complete. In spite of the limitations, a few New World anthropolo- gists have attempted demographic reconstruction from skeletal samples. The most widely quoted American demo- graphic analysis is that of Hooton (1930) from Pecos Pueblo material. Hooton considered data from his skeletal analysis to refine what he felt were inadequate population estimates derived from ethnohistorical sources. The attempt was of limited success because he could obtain so few constants. He was forced to assume that the excavated sample was rep- resentative of the entire cemetery population and that Kid- der's (in Hooton, 1930:333) estimate—approximately 15 to 20 percent of the cemetery had been excavated—was cor- rect. Hooten then had to depend upon ethnohistorical estimates for the length of time the cemetery was in use. Howells (1960) reworked Hooton's data, producing an estimate of population size directly from the number of skeletons. Utilizing new archeological information from Kidder (1958) and assuming a death rate of 30 per thou- sand per annum, Howells produced what is probably a more realistic estimate; but still there is no assurance that the assumed death rate and time interval are accurate. Snow (1948) attempted a demographic analysis of the large Archaic skeletal population from Indian Knoll, Ken- tucky. The analysis simply involved estimating age for each skeleton and plotting the ages to form a mortality curve for the population. Unfortunately, the skeletal sample may not have been complete (Howells, 1960:169) and Snow's orig- inal aging criterion for adults was rather unreliable. Snow determined adult age merely by examining cranial suture closure, which has been demonstrated since then to produce quite variable results (Singer, 1953; McKern and Stewart, 1957). Johnston and Snow (1961) revised the original age estimates using more and supposedly improved criteria, producing new death curves and new estimations of life expectancy. However, Stewart (1962) effectively showed that their reassessment has little meaning, since again thev used unreliable aging criteria (dental attrition, for example) and did not reexamine the entire skeletal series. Blakely and Walker (1968) presented mortality profiles constructed from 479 Middle Mississippian skeletons ex- cavated from Dickson Mound in Fulton County, Illinois. Unfortunately, the accuracy of their age determinations cannot be evaluated since they were produced using methods derived from unpublished sources. In addition, whereas they attempt to explain the death frequencies in each age cate- gory through a discussion of age-specific disease, they do not consider how their statistics might be affected by in- adequate sampling. In general however, their mortality curves reflect a high infant mortality rate with an average age at death of about 23 years. In a later article, Blakely (1971) compares mortality profiles reconstructed from Archaic, Middle Woodland, and Middle Mississippian skeletal populations, showing that the average age at death increases from 27 years during the Archaic to 30 years during the Middle Woodland; but then drops to only 24 years during the Middle Mississippian period. He provides an excellent discussion of what disease or cultural factors could influence these statistics and cor- rectly cautions that the differences could reflect only dif- ferences in mortuary practices. Bass, Evans, and Jantz (1971) attempted to derive in- formation on mortality and population size from skeletal material of the Arikara at a historic site in South Dakota. After working with assumed death rates, total numbers of excavated skeletons, and archeological and ethnohistorical estimates of population size, they concluded that the ex- cavated skeletal sample is grossly incomplete (Bass, Evans, and Jantz, 1971: 160-161). They discussed age distributions of deaths, but again the demographic considerations are of limited value since such a large percentage of the skeletons are missing. In short, the applications of quantitative methods in physical anthropological demographic reconstruction in the New World have been limited severely by uncontrollable variables, principally sample inadequacy. According to Howells (1960:159), "Control of these sources of error is good only by accident, as perhaps in the case of an Iroquois ossuary, known to represent a limited number of years." Only rarely have the unique ossuary samples mentioned by Howells been considered as potential sources of demo- graphic information. Churcher and Kenyon (1960) made such an attempt using Iroquois ossuary material. Unfortu- nately most of their ossuary sample was initially destroyed by power equipment and then badly looted by local col- METHODS OF POPULATION RECONSTRUCTION lectors. The report is of limited value because of the in- completeness of the sample, their use of assumed death rates, and their reliance upon suture closure and dental attrition as age indicators. Still, the article emphasizes the importance of ossuary samples in studies of prehistoric demography. Several other scholars have studied ossuary burials, but have stopped short of demographic reconstruction. Ander- son (1964) analyzed an Iroquois ossuary near Toronto, but confined his demographic study to estimation of age at death. Since he did not state his criteria for determining adult age, it is difficult to utilize his data. Kidd (1953) excavated an ossuary in Ontario and presented a convincing argument that it corresponded to the skeletal deposit de- scribed in great detail by the Jesuit missionary, Jean de Brebeuf, in 1636. However, in spite of the tremendous opportunity to study demography from a documented os- suary, Kidd confined his study to archeological features of the site itself. In the Plains, Kivett (1953), Strong (1935), and others excavated apparent ossuary burial sites, but confined their analyses to the archeological features, providing only esti- mates of the number of individuals represented and their age and sex distributions. The Plains ossuary complex lacks die ethnohistorical documentation of the Huron, but still offers a valuable skeletal sample for advanced analysis. Research Model of Present Study Clearly, if advanced techniques of physical anthropology can contribute to our understanding of prehistoric popula- tion size and demographic structure, they must concentrate on small regional populations and must utilize the best skeletal samples available for analysis. Unfortunately most skeletal populations recovered from New World prehistoric cemeteries are too incomplete to qualify for detailed demo- graphic analysis. However, if archeological or ethnohistorical information indicates the extent of the completeness of the sample, then skeletal samples do have potential for such analysis. In particular, the rare ossuary samples offer ex- ceptional opportunities to reconstruct accurate demographic profiles and even to derive reliable size estimates for popu- lations of the small geographic areas represented by the ossuaries. Although ossuaries seldom have been considered as sources of demographic information, they probably have the greatest potential of any prehistoric New World skeletal sample. This study utilizes two prehistoric Late Woodland ossu- aries from the Juhle site in southern Maryland to recon- struct aboriginal population profiles. The following analysis first considers the nature of the skeletal samples, including a discussion of the ethnography of general ossuary burial, local history of the Juhle site, excavation approaches, cultural-archeological affinites, and cultural information de- rived from the excavation and analysis. Several methods are then utilized to establish age at death, including a modifi- cation of Kerley's (1965, 1969, 1970) new method involv- ing an assessment of the degree of microscopic internal bone remodeling. Demographic data in the form of mortality curves, survivorship curves, life tables, and crude mortality rates are reconstructed and discussed. Archeological informa- tion is utilized to establish the time interval represented by the ossuary deposits. Finally, population size estimates are derived from the crude mortality rates (calculated from the life tables), time interval, and total numbers of individuals in the ossuaries. Evidence of Ossuary Burial from Eastern North America Ossuary burial practice may be described generally as the collective, secondary deposit of skeletal material represent- ing individuals initially stored elsewhere. In the past, the term "ossuary" has been applied rather loosely to almost any mul- tiple burial, either primary or secondary (Yarrow, 1880; Bushnell, 1920; Weslager, 1942). In this study, the term is restricted to those secondary deposits that probably repre- sent the periodic redisposal of individuals, which took place after a culturally prescribed number of years. Ethnography of Ossuary Practice THE HURON Ethnographically, ossuary burial practices are best known from Iroquoian groups in Canada and the Great Lakes re- gion. The Huron in particular were visited during the 17th century by a number of explorers and missionaries, some of whom recorded lengthy accounts of their customs. The most thorough description of Huron ossuary burial customs is found in the letters of Jean be Brebeuf, a Jesuit missionary from the mission of St. Joseph, who observed the burial ceremony at the Huron village Ihonatiria on 16 July 1636 (Thwaites, 1896-1901, X:279-305). Since his lengthy ac- count has been reproduced elsewhere (Kidd, 1953:372-375) and frequently referred to by other authors, it will only be summarized here. De Brebeuf relates that every ten to twelve years, the Huron journeyed to their temporary burial areas and carefully removed the skeletons, with each family being responsible for its own deceased relatives. The decayed re- mains were then cleaned, wrapped in fresh robes, and brought to the site chosen for reburial. The Huron ex- cavated a deep, circular, ossuary pit, lined it with beaver robes, and then erected a wooden platform around it. At the time of deposit, the skeletal remains were grouped ac- cording to the villages and families to which they belonged, wrapped tightly in robes and blankets, and hung from the platform. Copper kettles and necklaces of shell beads were added as mortuary offerings. Later, after extensive feasting, the Huron emptied the skeletons into the pit, saving the robes and blankets they were wrapped in. After the bones were arranged by men with poles, they were covered with robes, mats and bark, and the remainder of the pit was filled with "sand, rods, and stakes of wood which were thrown in promiscuously." While Jean de Brebeuf provides the most detailed discus- sion of ossuary burial practice, the earliest description was probably by the explorer, Samuel de Champlain. Champ- Iain's account (Biggar, 1929:160-163) generally comple- ments that of de Brebeuf, but adds the fact that the primary deposit was either in "a cabin covered with tree-bark" erected above ground on four posts, or in the ground with the cabin erected over the grave. He also stresses the regular scheduling of reburial and believed the ceremony took place every eight or ten years. A third detailed description of Huron ossuary burial is provided in the journals of Father Gabriel Sagard, a lay brother of the Franciscans who was in Canada from June 1623 to Autumn 1624. Since the concise and informative Sagard account rarely is discussed, it is reproduced in full (Wrong, 1939: 211-212). Every ten years, or thereabouts, our savages and other sedentary tribes hold the great festival or ceremony of the dead in one of their towns or villages, according as it has been decided and ordered by a general council of all the people of the district; for the bones of the dead are entombed separately only for a time. The other neighbour- ing tribes are notified in order that those persons who have chosen that town to be the burying-place of their relations' bones may bring them thither, and others who wish to come out of respect may honour the festival with their presence. For all are made welcome and feasted during the days that the ceremony lasts, and nothing is to be seen there except kettles on the fire, and continual feasting and dancing, and this brings an immense number of people who flock in from all sides. The women who have to bring the bones of their relatives go to the cemeteries for them, and if the flesh is not entirely destroyed they clean it off and take away the bones. These they wash and wrap up in fine new beaver-skins, and with glass beads and wampum neck- laces, which the relations and friends contribute and bring, saying, "Here, this is what I am giving for the bones of my father, my mother, my uncle, cousin or other relative." A"nd putting them into a new bag they carry them on their backs, and also adorn the top of the bag with many little ornaments, with necklaces, bracelets, and other decorations. Then the skins, tomahawks, kettles, and other articles of value in their estimation, as well as plenty of provisions, are also carried to the place appointed, and when all are assembled there they put the provisions together to be used for the feasts, which are a great expense to them, and then hang up decently in the lodges of their hosts all their bags and skins to await the day on which every- thing must be buried in the earth. OSSUARY BURIAL FROM EASTERN NORTH AMERICA The grave is dug outside the town, very large and deep, capable of containing all the bones, furniture, and skins offered for the dead. A high scaffolding is erected along the edge, to which all the bags containing bones are carried; then the grave is draped throughout, both the bottom and the sides, with new beaver skins and robes; then they lay in it a bed of tomahawks, next kettles, beads, necklaces, and bracelets of wampum, and other things given by the relations and friends. When this has been done the chiefs, from the top of the scaffold, empty and turn out all the bones from the bags into the grave upon the goods, and they cover them again with other new skins, then with tree-bark, and after that they put back the earth on top, and big pieces of wood. To mark their respect for the place they sink wooden posts into the ground all round the grave, and put a covering over it, which lasts as long as it can. Then they have a feast again, and take leave of one another, and return to the places whence they came, with great joy and satisfaction at having provided ihe souls of their relatives and friends with something that day to plunder and wherewith to become rich in the other life. These ethnographic accounts clearly elucidate the import- ance of the ossuary burial practice to the Huron people and the significance they attached to carrying out details of the ceremony correctly. Although the early ethnographers disagree whether the length of time between ossuary deposits was 8 years (Champlain) or 10 to 12 years (de Brebeuf), there is little doubt that it was a fixed, reoccurring event that was integrated into their belief system and religiously adhered to. According to Bressani (Thwaites, 1896-1901, XXXIX: 29), the Huron feast of the dead was "the most sacred and solemn ceremony that they had. ..." The Jesuit Jerome Lalemant (Thwaites, 1896-1901, XXIII: 31) adds in 1642: "If there be anything in the world that is Sacred among the Hurons, it is their law of Burial. Their care in this matter greatly exceeds anything that is done in France. They are singularly lavish in proportion to their means, and despoil themselves to clothe their Dead and to preserve care- fully the bones of their Relatives, in order that they may repose after their death in the same spot." In situations where fire threatened both their homes and the scaffolds holding their dead, the Huron "did not feel troubled at incurring an irreparable loss, that they might save the bones of their departed before extinguishing the fire in their own cabins" (Bressani, in Thwaites, 1896-1901, XXXIX:31). Clearly, the Huron valued their dead highly and made every attempt to preserve each individual for ossuary burial. At the time of ossuary deposit, the Huron collected the remains of everyone who had died since the last ceremony. Each family assumed the responsibility for gathering the remains of their deceased relatives and preparing the bones for reburial. To insure the completeness of the skeletal material, the Huron even included those individuals who had just died. According to Jean de Brebeuf (Thwaites, 1896-1901, X:283), "The flesh of some is quite gone, and there is only parchment on their bones; in other cases, the bodies look as if they had been dried and smoked, and show scarcely any signs of putrefaction; and in still other cases they are still swarming with worms." In 1636, de Brebeuf (Thwaites, 1896-1901, X:285) describes how the Huron treated the body of an old man who had died the autumn before. This swollen corpse had only begun to decay during the last month, on the occasion of the first heat of Spring; the worms were swarm- ing all over it, and the corruption that oozed out of it gave forth an almost intolerable stench; and yet they hid the courage to take away the robe in which it was enveloped, cleaned it as well as they could, taking the matter off by handfuls and put the body into a fresh mat and robe, and all this without showing any horror at the corruption. All of these accounts document both the complexity and social significance of the ossuary practice to the Huron peo- ple. The ceremonies involved were not just an expensive method of disposal, but rather an important functional ele- ment of their culture that combined their religious belief concerning life after death with the communal gathering which reinforced social relationships and added cohesion to the community. Of course, in spite of the social significance of the ceremony, the missionaries looked upon it with dis- gust. Eventually they voiced their disapproval, especially for those Huron people who became Christians. The follow- ing paragraph from Biard's account of 1611 (Thwaites 1896-1901, 1:169) reflects the cultural conflict involved and the usual Christian response: I shall here relate another act of the same Sieur de Potrincourt, which has been of great benefit to all these heathen. A christian savage had died, and (as a mark of his constancy) he had sent word here to the settlement during his sickness, that he desired our prayers. After his death the other Savages prepared to bury him in their way; they are accustomed to take everything that belongs to the deceased, skins, bows, utensils, wigwams, etc., and burn them all, howling and shouting certain cries, sorceries, and invocations to the evil spirit. M. de Potrincourt firmly resolved to oppose these ceremonies. So he armed all his men, and going to the Savages in force, by this means obtained what he asked, namely, that the body should be given to the Patriarch, and so the burial took place according to christian customs. This act, inasmuch as it could not be prevented by the Savages, was and still is, greatly praised by them. Gradually, the Huron custom of ossuary burial was re- placed by Christian mortuary practices. However, the mod- ern Iroquois still placate their dead with semi-annual feasts reminiscent of the feast of the dead. Fenton and Kurath (1951) have shown that the modern ceremony is descended directly from the former practice and even retains many of the original terms. The practice of ossuary burial among the Iroquois pro- duced many ossuary pits over the years. However, because of the abundance of artifacts contained, the skeletal deposits have been looted heavily by local collectors. Although An- derson (1964) documents at least 216 ossuary sites in the Province of Ontario alone, Kidd (1952:73) notes that few remain undisturbed. Kidd (1952:73) estimates mat some of the looted ossuaries contained as many as 1000 individ- uals. Smaller ossuaries have been professionally excavated and reported by Anderson (1964), Churcher and Kenyon 10 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 (1960), Kidd (1953), and Ridley (1961). All of the exca- vations provided evidence consistent with the ethnographic accounts. In particular, Kidd (1953) believed that the one he excavated was that observed in use in 1636 by Jean de Brebeuf. The pit was bowl-shaped and about 24 feet in diameter. Many of the bones were scattered, some were partially articulated, and two were completely articulated and lying on the floor of the ossuary. A ring of 9-inch post- holes surrounded the pit and probably represents the plat- form described by de Brebeuf. Artifacts included shell beads and fragments of copper kettles. All of these features cor- roborate de Brebeuf's observation. SOUTHEASTERN UNITED STATES Burial customs in the Southeast include ceremonies similar to those described for the Huron, but with some regional differences. According to Romans (1775:88), the Choctaw also practiced "bone cleaning" of the deceased. However, the bones were not cleaned by relatives, but by a group of specialists who traveled through the Choctaw nation. These mobile Choctaw morticians were described by Romans (1775:88) as "a certain set of venerable old Gentlemen who wear very long nails as a distinguishing badge on the thumb, fore and middle finger of each hand." Their long nails not only symbolized their trade, but allowed them to remove the decaying flesh with ease. The removed flesh was then burned and the bones were painted, placed in a chest, and deposited in a "bone-house." Other accounts report the practice of bone-cleaning to have been common among other Southeastern groups. The method of final bone dis- posal seems to have varied considerably, however, with some groups apparently leaving the bones in the temples permanently. MID-ATLANTIC AREA OF THE UNITED STATES Mortuary customs in the mid-Atlantic area are best known from the writing of Captain John Smith and Thomas Harlot, and the drawings df John White. John Smith (Arber, 1910) recorded the manners and customs of the Virginia Indians as he observed them during the years immediately following the founding of Jamestown in 1607. Smith was especially fascinated with their manner of treating deceased leaders and wrote (Arber, 1910:75) : Their bodies are first bowelled, then dryed vpon hurdles till they bee verie dry, and so about the most of their iointes [joints] and necke they hang bracelets or chaines of copper, pearle, and such like, as they vse to weare: their inwards they stuffe with copper beads and couer with a skin, hatchets, and such trash. Then lappe they them very carefully in white skins, and so rowle them in mats for their winding sheetes. And in the Tombe, which is an arch made of mats, they lay them orderly. What remaineth of this kinde of wealth their kings haue, they set at their feet in baskets. These Temples and bodies are kept by their Priests. Thomas Hariot (1590) adds another description of how the southern Virginia Indians disposed of their deceased leaders. According to him, they opened the body to remove the entrails and cleaned the flesh from the bones, leaving ligaments intact. They then packed the body with leather until it resumed its original shape. The bodies of all leaders were placed side by side on a 9- or 10-foot scaffold or death house. The famous drawing by John White that accom- panies the Hariot text visually illustrates the description. The drawing shows nine deceased chiefs stretched out on a scaffold in the death house, with a fire burning at the entrance. Unfortunately, none of the accounts indicate what eventually happens to the deceased leaders after their stay in the death house. Early accounts of the treatment of deceased common people are less revealing. According to Smith (Arber, 1910:75), "For their ordinary burials, they digge a deep hole in the earth with sharpe stakes; and the corpes being lapped in skins and mats with their iewels, they lay them vpon sticks in the ground, and so couer them with earth. . . ." Smith's description is difficult to interpret in that his reference to "a deep hole" suggests an ossuary type burial, but his use of the word "corpes" implies primary burial. In an earlier account, Smith wrote (Arber, 1884:22) in 1608, that "their Kings they burie betwixt two mattes within their houses, with all his beads, iewels, hatchets, and copper: the other in graves like ours . . . ," suggesting that the com- mon people were buried individually in the ground. Spelman (1609-1610, in Arber, 1884: ex) in his Realtion of Virginia adds a contradictory account of common burial. He relates that after death, the body was wrapped in a mat and deposited on a scaffold, 3 to 4 meters (3 to 4 yds) above the ground. After the flesh decomposed, the Indians wrapped the remaining bones together in a new mat and hung them in "their howses, wher they continew whille ther house falleth and then they are buried in the ruinges of ye house. . . ." In 1676 Glover (1676:24-25) adds that the Virginia Indians burned the bodies of the dead and placed the ashes in mats near their relatives' dwellings. Clearly the variations in these ethnographic accounts in- dicate either variability in Alqonquian mortuary practice or inaccurate observation and reporting. Few scholars doubt that the drawing by White and accounts by Smith and Hariot accurately describe the treatment of deceased lead- ers. However, there is no agreement on methods of treating deceased common people. In the Maryland and Delaware area, direct, reliable, ethnographic accounts are not available. A considerable amount of indirect ethnographic evidence, however, sug- gests that the practice of bone cleaning and ossuary burial did occur. The missionary John Heckewelder (1819:92) observed in 1776 that the Nanticokes moving north out of Maryland through Pennsylvania carried with them the cleaned bones of their ancestors. OSSUARY BURIAL FROM EASTERN NORTH AMERICA 11 These Nanticokes had the singular custom of removing the bones of their deceased friends from the burial place to a place of deposit in the country they dwell in. In earlier times, they were known to go from the Eastern shore of Maryland, even when the bodies were in a putrid state, so that they had to take off the flesh and scrape the bones clean, before they could carry them along. I well remem- ber having seen them between the years 1750 and 1760, loaded with such bones, which being fresh, caused a disagreeable stence, as they passed through the town of Bethlehem. Heckewelder's observations were confirmed by his missionary supervisor, David Zeisberger in 1779 (Hulbert and Schwarze, 1910) and later by Brinton (1885). Evidence of bone cleaning among the Conoy of south- western Maryland can be found in the Proceedings of the Council of Maryland, 1676-1678 (Archives of Maryland, 1896:185). On 19 August 1678, "Nicotaghsen" (the Piscattaway or Conoy "emperor") and other Indian repre- sentatives were assembled at Lord Baltimore's Council to discuss business matters. Noticing that many of the Indian representatives were missing, the council inquired about the small attendance when such important business was to be discussed. After a pause, a Conoy leader replied that "most of their great men were very busie in gathering together their dead bones. . . ." Another passage in the Maryland Archives suggests that the Nanticokes at Assateague saved the bones of their leaders. The following account was recorded for 6 May 1686 (Archives of Maryland, 1887:480) : The King of Assateague complaines that severall of the Inglish (viz) Mr. William Browne, Edward Hammond, William Bowen, John Fossett, Henry Bishop &c were come and seated among them in the very Towne where they live—but particularly he complaineth against Edward Hamond for that whereas it is a custom among them upon the death of an Indian King to save his bones and make a case with skinns wherein they inclose the bones and fill it up with Ronoke, and other their riches, he the said Hamond about a month since had upon the like occasion of one of their kings dyeing stolen away the skinns and roanoke from the place where he was layd, which one Epimore a greate man of Assateague did see at the sayd Hammond's house and very well know to be the same, and alsoe one Manassen an Indian that lives with said Hammond did see him bring them home. Bozman (1837, 1:173-174) mentions a boyhood recollec- tion that the Choptank Indians kept the remains of their dead chiefs and leaders in a death house called the "Quioc- cason House." The term "Quioccason House" is used com- monly by Algonquian Indians to refer to their religious temples. In 1792, Dr. William Vans Murry of Cambridge, Eastern Shore of Maryland, submitted to Thomas Jefferson a vocab- ulary taken from the Nanticoke Indians of Maryland. The letter of explanation to Jefferson contains the following note (Speck, 1927:41) : "Wynicaco—the last king crowned of the Nanticoke tribe, he died at past 80 years since. His body was preserved and very formally kept in a Quacasun— house—chio-ca-son house, 70 years dead." Marye (1936:43-45) indicates that several references to the term "Quioccason" or modifications of it can be found in early land records in the Maryland area. Marye located two entries in a proprietary rent-roll of Somerset County, Maryland, describing a tract of land on the south side of the Nanticoke River. The first entry refers to the land as "Quiakeson Neck." The second entry describes the land as being by a swamp near "Indian Quiankeson houses." Marye believed that "Quiakeson Neck" was named because of its location near the death houses. Marye (1936) cites another occurrence of the word in a land record of 15 August 1761 in Dorchester County, Mary- land. The boundary of the tract was defined at one point by a group of trees standing in "cuiackason Swamp." In a later article, Marye (1937) reported a tract of land in Worcester County, Maryland, described in 1762 as being located on "Quaacotion House Point." He located the site near the Indian town of Askiminakonson, which was occu- pied by the Pocomokes. Adams (1890) and Harrington (1921) discussed the "Skeleton Dance" practiced by the Wolf Clan of the Dela- wares, who traced their origin to the Nanticoke. In the ceremony, the Delaware stripped the flesh from the bones of the deceased and buried it. The bones were then dried for 12 days, wrapped in white buckskin, and taken to the cere- monial dance. After the dance, the bones were interred collectively. The Nanticoke who were relocated in Kansas prior to 1875 placed the bodies on a 5 to 6 foot scaffold, immediately after death. After the flesh decayed, the bones were included in the "Ghost Dance" ceremony and later buried (Speck, 1937:148-149). Weslager (1942:144) feels both this prac- tice and the Delaware Skeleton Dance represent survivals of the earlier practices of bone scraping, temporary scaffold deposit, and final interment in ossuaries. The above ethnographic gleanings collectively constitute strong evidence for the practice of temporary placement on scaffolds or in death houses, systematic bone cleaning, and final secondary interment in ossuaries among the aboriginal inhabitants of the Maryland and Virginia areas. The ac- counts agree that this general mortuary practice applied to the aboriginal leaders. However, there may have been some regional variance in the treatment of the common people, especially in Virginia, since the accounts of Smith, Spelman, and Glover describe other patterns. Archeological Evidence The best documentation for ossuary burial in the mid- Atlantic region consists of the archeological discoveries of some 34 secondary skeletal deposits from sites distributed through Virginia, Maryland, and southern Delaware (Fig- ure 1). Although a few of these probably do not represent ossuary burial as defined herein, many possess archeological features nearly identical to those of Huron osssuaries. 12 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 ATLANTIC OCEAN 0 5 10 15 20 MILES L_l I I I 0 5 10 15 20 25 30 KM. I I—I 1 1—I I FIGURE 1.—Geographic distribution of reported ossuaries in the region of the Chesapeake and Delaware bays. OSSUARY BURIAL FROM EASTERN NORTH AMERICA 13 Perhaps the first skeletal deposit described as an ossuary in the mid-Atlantic area was found at site 7S-H-15 in 1808 about 2 kilometers (1 mi) from Laurel, Delaware. Appar- ently, some workers removing earth from a stream bank to repair a mill dam uncovered a concentration of human bones. According to Huffington (1838:16), several wagon loads of bones were removed and a considerable number were left in situ. Several old men at that time recalled that the Indians had buried the bones just before they departed from Delaware. Their recollection that the Indians brought "bones" for burial and not bodies suggests secondary, os- suary burial. However, Huffington (1838:16) relates that die skeletons "were laid side by side and each bone in its proper place. Among them were several frames which must have belonged to men of large growth. One in particular was said to have measured seven feet in length." With the con- flicting evidence and in the absence of any contemporary records, it is impossible to determine whether the Laurel discovery was an ossuary or even a secondary burial. In 1883, workmen uncovered 16 skeletons at site 18D042 while constructing a jail foundation in Cambridge, Mary- land. The skeletons were supposedly arranged in a circle, with no associated artifacts (MacLeod, 1928:208). The term "ossuary" has been applied to this skeletal sample by Weslager (1942:145) but again it does not fit the ossuary definition used here. Henry C. Mercer reported the exploration of an Indian ossuary on the Choptank River, Dorchester County, Mary- land (18D030) in 1897. The Choptank ossuary was 3 kilo- meters (2 mi) north of Cambridge, Dorchester County, Maryland, on the left bank of the Choptank estuary. The deposit had been previously disturbed by amateurs seeking crania, but still contained scattered human bones in an ir- regular pit, 7.62 meters (25 ft) in length by 6.10 meters (20 ft) in width. The bone deposit was 0.46 meters (1.5 ft) thick. A second deposit measuring 2.1 meters (7 ft) long, 2.4 meters (8 ft) wide, and 0.2 meters (8 in) thick was lo- cated above the larger concentration and separated from it by 0.46 meters (1.5 ft) of sand. At least 100 individuals were represented, with no bones in articulation and no arti- facts associated. What may be a third ossuary was uncovered at this site in the late 1930s close to Mercer's original excavation. The feature was completely destroyed by local collectors, who reportedly removed an unknown number of skeletons with a tremendous amount of accompanying grave goods, includ- ing over 100 complete gorgets, over 100 copper beads, effigy pipes, and a large quantity and variety of worked stone (Weslager, 1942:146). Although this discovery may repre- sent an ossuary, the artifacts suggest it more probably repre- sents an earlier deposit, perhaps Adena. Fowke (1894) reported two possible ossuaries from Al- leghany County, Virginia. The 2.4-meter (8 ft) pits each contained the disarticulated remains of approximately 25 individuals of all ages. Graham (1935:33) mentions a third ossuary found by Fowke in Orange County, Virginia, but Fowke's (1894:24-25) description indicates the discovery was just a mound containing some secondary skeletal material. Wigglesworth (1933) reported excavating 15 articulated skeletons from a cliff face just south of Rehoboth Beach, Delaware (7S-G-3). The skeletons (14 adults and 1 child) were apparently completely articulated and communally de- posited in a pit 2.8 meters (9.2 ft) long by 2.0 meters (6.7 ft) wide. Although Weslager (1942:145) includes the site in his discussion of ossuaries, it probably represents a group deposit of individuals, who for unknown reasons died at approxi- mately the same time. Davidson (1935:91-96) recorded the excavation of five individuals at Slaughter Creek (7S-C-7), 10 miles south of Milford, Delaware. One individual was represented by a bundle in a separate pit, one by an articulated flexed skeleton in a separate pit, and the last three by disarticulated bones in a common pit. Davidson (1935:91-96) interpreted the variety of burial forms present at the site as indicating that the material was transitional between the earlier practice of individual burial and the later practice of ossuary deposit. Weslager (1942:145) mentions that an ossuary was en- countered by a state highway crew near Killens' Mill Pond (7K-E—3) east of Felton, Delaware. Many skeletons were apparently removed, but no details were recorded. Weslager (1942:145) claimed to discover an ossuary on the Vincent Farm (18D041) near the junction of Whitehall Creek and the Choptank River, several miles east of Cam- bridge, Maryland. Actually, the site had been visited in 1936 by T. Dale Stewart (1940b: 363) and even earlier by local collectors. Stewart salvaged some of the material and verified the ossuary as authentic, but unfortunately could glean little further information. In 1935, William J. Graham, Presiding Judge of the United States Court of Customs and Patent Appeals, Wash- ington, D.C., privately published his analysis of five ossuaries from site 18CH95 near Port Tobacco, Maryland. Aided by T. Dale Stewart, Graham personally excavated four of these ossuaries. The fifth had apparently washed out of a bank and had been looted heavily before Graham discovered it. All of the ossuaries contained secondary, disarticulated skele- tal material, and two produced artifacts of European manu- facture. Graham (1935) estimated the four undisturbed os- suaries to contain 10, 25, 50, and 100 individuals. Judge Graham continued his excavation of ossuaries in 1937, this time moving to the site of Patawomeke (44ST1) on Potomac Creek. The habitation site is located on the west bank of the Potomac River, near what is now Marlboro Point, Stafford County, Virginia, and supposedly had close to 1000 inhabitants at the time Smith visited the site in 1608. Graham's excavations revealed three ossuaries, one of which contained artifacts of European manufacture. A fourth os- suary was discovered later by an associate of Graham's and a fifth was added by T. Dale Stewart in 1939 (Stewart, 14 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 8 1940a, 1941). Stewart's ossuary contained 135 individuals and many shell beads. The first four ossuaries were esti- mated by Graham (1935) to contain 181, 287, 67, and 41 individuals. Possibly the first ossuary excavated at Patawomeke was reported by E. R. Reynolds (1883). In 1869, a group of amateurs from the Shenandoah Valley excavated 12 skele- tons, buried together at a depth of six feet. Although Rey- nolds (1883) described the site as being at the junction of Accotink Creek and the Potomac, Stewart (1958a) showed that Reynolds actually meant "Accakeek" and not "Acco- tink." Consequently, the discovery was probably close to the site of Patawomeke. However, E. R. Reynolds' (1883) ref- erence to articulated skeletons indicates the feature was probably not an ossuary, as defined in this report. In 1937, T. Dale Stewart and Waldo R. Wedel of the Smithsonian Institution excavated two ossuaries in Ana- costia, just outside and southeast of Washington, D.C. Ap- parently, the ossuaries were associated with the village of Nacotchtanke visited by Smith in 1608, who observed "80 able men" at that time. The bones were uncovered by a power shovel when nearby Boiling Field was being extended. Between 63 and 70 individuals were present in each pit, represented by disarticulated bones grouped in bundles. Several shell beads and pottery sherds were recovered. Stewart (1940b) investigated another ossuary in 1939, this time at 44KQ10 along the York River in Virginia. The bones had been discovered several years earlier eroding out of the north bank of the river, between 6.4 and 8.1 kilo- meters (4 to 5 mi) south of the present town of West Point. By the time Stewart was notified and could investigate, over half of the ossuary had washed away. Bones could be found along the beach for a considerable distance. The remaining bones were stacked in bundles, some partly articulated. Six potsherds represented the only cultural material. During the years 1936-1937, Alice L. Ferguson, an amateur archeologist from southern Maryland, conducted an excavation at the site of Moyaone (18PR8) near the mouth of Piscataway Creek, in Prince George's County, Maryland, just across the Potomac from Mount Vernon (Ferguson, 1937; Stephenson and Ferguson, 1963). In the process of exploring the village area, she uncovered four os- suaries. Stewart again had an opportunity to examine, but not to excavate the skeletal material. Because of the man- ner in which the bones were recovered, exact statements as to the number of individuals, etc., are not possible. Accord- ing to Stewart (per. comm.), Ferguson's archeological ap- proach involved digging a deep trench around the deposit, and then inviting friends for an afternoon of bone collecting from the trench wall. A few selected bones were saved for Stewart to examine. In general the ossuaries presented the same type of associations found at Patawomeke. The remains were mostly disarticulated and partially arranged in bundles. The first two ossuaries contained an estimated 288 skulls, the third 248 skulls and the fourth 618 skulls. Later a fifth os- suary was found at the nearby site of Piscataway that con- tained 254 skulls, as well as a multitude of trade material. Specimen accession records at the Division of Physical An- thropology, National Museum of Natural History, indicate that a possible sixth ossuary was discovered in the Moyaone- Piscataway area as early as 1908. The records indicate that 99 human bone specimens were accessioned in 1908 as a result of excavations by A. Hrdlicka, J. D. McGuire, and J. H. Reams at the mouth of Piscataway Creek. By the time Weslager published his synthetizing article on ossuaries on the Delmarva peninsula in 1942, 32 suggested ossuaries had been discovered in the mid-Atlantic area. Of these, 21 probably represent true ossuaries (2-Choptank River, 1-Vincent Farm, 5-Port Tobacco, 5-Patawomeke, 2-Anacostia, 1-York River, and 5-Piscataway Creek), six were definitely not ossuaries (1-Orange County, 1-Cam- bridge Jail, 1-Rehoboth, 1-Slaughter Creek, 1-Choptank River, and 1-Patawomeke), and five (1-Laurel, 2-Alleghany County, 1-Killens Mill Pond, and 1-Piscataway Creek) re- main unknown due to the manner in which they were excavated. However, the 21 true ossuaries clearly substantiate the ethnographic implication that the practice of ossuary burial was important in the mid-Atlantic region, just as it was among the Huron to the north. The finding of both par- tially articulated and scattered bones of many individuals representing all age groups strongly suggests that, like the Huron, the people of the prehistoric mid-Atlantic region buried all of their deceased in ossuary graves. The presence of calcined bones in ossuaries at Anacostia, Choptank, Pata- womeke, and Accokeek indicates that if different mortuary practices were employed (i.e. cremations), the skeletal re- mains still were included in the ossuary collective deposits. The ethnographic-archeological evidence warrants the as- sumption that the large ossuary skeletal samples in the mid-Atlantic region represent nearly complete collections of aboriginal deaths for the time periods and populations the ossuaries served. Of course, the possible omission of the leaders, accidental deaths of individuals whose remains could not be recovered, and occasional loss of individuals from the primary deposit, theoretically prevent the sample from being 100 percent representative. Still, such losses must have been minimal if present at all, making ossuary skeletal collections far more complete than any other archeological skeletal sample in North America. If properly excavated, such relatively complete samples offer unique opportunities to reconstruct population profiles for the formerly living populations. The last two ossuaries discovered in the mid-Atlantic area, those from the Juhle site (18CH89), Charles County, Maryland, support the evidence sum- marized above for a Huron-like ossuary practice in the Tidewater Potomac. In the remainder of this report, these two ossuaries will be discussed in detail, and the skeletal material from them will be utilized to reconstruct demo- graphic profiles for the formerly living populations. The Ossuaries from Nanjemoy Creek By the late 1940s, students of mid-Atlantic archeology had learned a great deal about aboriginal ossuary burial. Enough ossuaries had been discovered to indicate that the practice was geographically distributed throughout eastern Virginia, southern Maryland, and southern Delaware (Figure 1), and chronologically confined to the late Woodland and early Historic periods (Davidson, 1935; Weslager, 1942). Ossu- aries were known to contain large numbers of both articu- lated and disarticulated skeletons, some scattered throughout the pit and others arranged in bundles. Some ossuaries con- tained large amounts of artifacts (aboriginal and European manufactured) and some included small amounts of burned bone. Although much had been learned, many questions re- mained concerning details of the mortuary custom, the physical type, and the demographic structure of the popula- tion. Most of the ossuaries had been excavated by amateurs, with only occasional professional supervision. Those that had received professional attention (Anacostia and York River, 44KQ10, for example) had been disturbed partially prior to excavation. Since the skeletal samples were incomplete, they were of limited use to physical anthropologists. Clearly, there existed a need for the professional excavation of an undisturbed ossuary and the analysis of a complete skeletal sample. As discussed in the preface, this need was fulfilled with the discovery and excavation of two ossuaries from the Juhle site in southern Maryland in 1953 and 1971. The Juhle farm (named Friendship Farm after an early English Ship) is located approximately 50 miles (80 kilo- meters) south of the present District of Columbia on the north bank of Nanjemoy Creek, a small tributary of the Potomac River. The ossuaries found by the Juhles and the probable associated habitation site (18CH89) are on a 33.5- meter (110 ft) bluff overlooking Nanjemoy Creek (Figures 2, 3). Ossuary I lies 3 meters (10 ft) north of the north kitchen porch of the present Juhle residence and one meter east of the circular, unpaved drive that connects the Juhle farmyard to the county road 91.0 meters (100 yds) to the north (Figure 4). Ossuary II lies about 30.5 meters (100 ft) northwest of Ossuary I. The site falls within the region occupied by the Algonquian-speaking Conoy or Pistcataway in early historic times, and possibly represents the Late Woodland pre-contact ancestors of these people. Although contact between the Potomac River Indians and Europeans began shortly after the founding of Jamestown in 1607, the Nanjemoy area was not colonized until the mid-17th cen- tury as evidenced by the ruins of one of Charles County's old homesteads, dating from ca. 1660, and located less than one mile north of Ossuary I. Ossuary I Stewart's 1953-1955 excavation of Ossuary I was directed principally toward recovering the complete skeletal sample in a manner that would facilitate the later assemblage of complete individuals. Feeling that the bones of each indi- vidual would be scattered over a small area, he constructed a 1.5 meter (5 ft) grid system over the ossuary and kept the bones from each square separated from each other. Unfor- tunately, few complete skeletons could be assembled, but his excavation did document many important features of ossuary burial. He found that the bones were typically dis- articulated and scattered throughout the pit, although some articulation and a few distinct bundles were noted (Figure 5). At least seven individuals were almost completely articu- lated and were concentrated on the floor. Numerous exam- ples of partial articulation were observed to be scattered throughout the bone layer. At least 12 distinct bundles were recognized, usually at the periphery of the bone concentra- tion. The bundles contained both articulated and disarticu- lated skeletal parts, frequently representing more than one individual. Ninety-four skulls were recognized and numbered by Stewart, although the presence of numerous infant cranial fragments associated with the numbered skulls indicated the count would be much higher (my own analysis indicates a total of at least 131 individuals). The few artifacts found with the skeletons were all aboriginal and suggestive of a Late Woodland occupation, just prior to European contact. Ossuary II: Excavation Procedure Excavation of Ossuary II was designed to remove the skele- tal material as rapidly as possible without sacrificing archeo- logical data. Initially, this involved removing the 25.4-centi- meter (10 in) layer of over-burden to disclose the pit outline. The outline was easily distinguishable since the soil outside of it was much lighter in color, more compact, and lacked the small particles of charcoal and shell that permeated the pitfi.ll. The pit enclosed an area 5.2 meters (17 ft) long by 2.1 meters (7ft) wide. Our excavation of the bones themselves utilized a 0.6- meter (2 ft) grid system superimposed over the bone mass 15 16 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 I MILE I KILOMETER FIGURE 2.—Location of Juhle site (18CH89) near Nanjemoy Creek, Charles County, Maryland. THE OSSUARIES FROM NANJEMOY CREEK 17 FIGURE 3.—Nanjemoy Creek, as seen from Juhle site (18CH89), looking southwest. (Figure 6). This unit size was chosen over a 1.5-meter (5 ft) system because it was large enough to include a sizeable amount of bone, yet small enough to allow us to study dis- tribution variation within the pit, and possibly to facilitate association of bones of the same individual scattered over a localized area. Since preliminary testing indicated that the bone concentration was approximately 1.8 meters (6 ft) wide, three rows of 0.6-meter (2 ft) squares were laid out parallel to the long axis of the bone concentration. By chance, the barbed wire fence being erected by the Juhles at the time the bones were encountered bisected the ossuary longi- tudinally and divided it into southwest and northeast halves (Figure 7). At Stewart's suggestion, we labeled the three longitudinal rows according to their positions relative to the fence. We assigned the letter "L" (left) to the row southwest of the fence (i.e., to the left as one faces away from Ossuary I) ; "C" (center) to the row directly under the fence; and "R" (right) to the row northeast of the fence (i.e., to the right as one faces away from Ossuary I). The 0.6-meter (2 ft) squares in each of the three longitudinal rows were lettered consecutively from A to H, beginning from the south- east end. Thus, the first squares of the three rows bore the designations AL1, AC, and AR1 (Figure 6). This arrange- ment proved easy to use and to remember. Although small amounts of bones extended beyond the left and right limits of the squares, and our original intention was to give them the number "2" (AL2, AR2, etc.), in actual practice we included these bones with those in the squares number "1". The number "1" was retained (AR1 as opposed to just AR) since our field notes originally were recorded in this manner. Our excavation procedure involved exposing the bones in situ; describing, mapping, and photographing them in the context of the grid system, and then removing them with as little damage as possible. Initially we planned to preserve the original shape of the trench, documenting the positions of the bones relative to the trench walls. Leaving the trench wall intact, however, forced the excavator into the some- what uncomfortable position of leaning down into the pit to work (Figure 8a, b). The situation was improved by the construcdon of a board support system which enabled the excavator to lie above the bone concentration while working (Figures 8c, 9). However, best results were obtained when 525-413 O - 74 - 3 18 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 0 10 20 30 40 50 FT i i l 1 1 1 i 1 1 ■ 0 5 10 15 M OSSUARY I (CONTOUR INTERVAL 2 FT J BUSH Q ;o TREE A = DATUM 3 TOOL SHED '"///,.. '''ill'-.. '"//""■•.„ ''''////,..., DATUM 2, 'I!'"' nu ^'V 4 -i A I «" It"' lN\#' ///,,,. , ,nii ,l" "' \SjS? ...««# '"' '"" , |l »«'" „ll,#»"" , 'I MIU" TREE nlM1'' ...>\\v0. TREE FIGURE 4.—Detailed map of the Juhle site (18CH89), Charles County, Maryland. the area outside of the ossuary was removed, enabling the excavator to lie at the same level as the bone concentration. To minimize damage from exposure, we confined our exca- vation to approximately one-third of the ossuary at one time, beginning with the row of squares AL1 through HL1 (Fig- ure 10a, b). When that material had been exposed, recorded, and removed, we shifted our attention to the row of squares AR1-ER1 (Figure 10c) and then ended the 1971 season with the center row of squares AC-EC (Figure 11). The final six squares in the northwest corner (FC-HC and FR- HR) were excavated simultaneously during the 1972 season (Figure 12). Because of the interlocking nature of the bone concentration, some material in the left and right rows could not be removed until the center row was excavated. These log-jams of bones constantly complicated the excavation. In addition, the 0.6-meter (2 ft) grid system only could be applied very generally since a high percentage of the large bones extended across the arbitrary unit boundaries. In such cases, a bone was assigned to the square containing over 50 percent of its length. Occasional bones appeared to be divided equally between two squares and these were as- signed both square numbers (AL1-BL1; AL1-AC; etc.). In spite of these complications, analysis later demonstrated that the 0.6-meter (2 ft) grid system provided a satisfactory approach to ossuary excavation. After the bones were removed, they were divided into four categories: (1) complete individuals, (2) partially articulated bones, (3) bundles, and (4) scattered, disarticu- lated remains. The few complete individuals were each as- signed a number and their positions in the grid were recorded. Bones that were still partially articulated (hands, feet, etc.) were assigned separate "partial articulation" num- bers and their positions in the grid were also recorded. Bundles of both scattered and partially articulated bones were assigned "bundle numbers" and their grid positions were plotted. Finally, scattered, disarticulated bones were assigned individual square numbers. This four-fold classifi- cation system allowed the recording of several culturally sig- nificant bone categories without the loss of information on spatial distribution. Aside from general observation, no at- THE OSSUARIES FROM NANJEMOY CREEK 19 FIGURE 5.—Excavation of Ossuary I: a, 1953—upper level of eastern section, looking northeast; b, 1953—lower level of eastern section, looking northeast (note articulated skeletons and bundles); c, 1955—western section, looking northeast; d, 1955—western section, looking east. tempt was made to distinguish upper from lower layers in the bone concentration, because no natural layering could be detected and long bones often were observed to extend from top to bottom. Ossuary II: Contents In general, our excavation revealed that the archeological features of Ossuary II were very similar to those of Ossuary I and other Late Woodland ossuaries in the area. The bones had been placed in a trenchlike pit with rounded ends (Fig- ure 13). The trench was shallow (maximum depth 0.5 meters (1.8 ft) at the northeast end and gradually tapered to a maximum depth of about 0.9 meters (2.9 ft) at the southwest end). The overall maximum pit dimensions were 5.0 meters (16.5 ft) by 2.2 meters (7.3 ft) with the long axis oriented northeast-southwest. A 20.3 cm (8 in) to 25.4 20 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 FIGURE 6.—The 0.6-meter (2 ft) square system utilized in the 1971 excavation of Ossuary II. cm (10 in) upper layer of dark soil marking the plow zone extended uniformly over the area. Below the plow zone, bones were encountered at depths ranging from 30.5 centi- meters (12 in) in square AL1 to 43.2 centimeters (17 in) in square CL1. No evidence of a Huron-like platform was found on the periphery of the pit; however, a small intersecting pit was discovered in the western corner. The circular inter- secting pit (designated feature I) was 0.8 meters (2.7 ft) in diameter and 0.5 meters (1.8 ft) in depth. It contained several small charcoal fragments and quartz flakes, several broken shells of oyster (Crassostrea virginica), a distal tibia of a white-tailed deer (Odocoileus virginianus), 10 frag- ments of softshell turtle (Trionyx spinifer), and two frag- ments from the three-toed box turtle (Terrapene Carolina). Intersecting pit outlines failed to reveal whether the probable trash pit was later, contemporary, or earlier than the ossuary. One of the oyster shells displayed a serrated edge composed of 13 notches that extended around the lip of the shell from the hinge. The notches are 1 to 2 mm in depth and 1 to 4 mm apart. Stephenson (in Stephenson and Ferguson, 1963:163) described 226 serrated mussel shells from the nearby Accokeek Creek site and suggested they may have served as scrapers or cutting tools. The bone concentration as a whole varied in thickness from 0.5 meters (1.8 ft) in square AC to only 7.6 cm (3 in) in square HO Table 2 demonstrates the distribution of bone weight per square. These figures represent the total weight (in kilograms) of the cleaned bones by squares and provide a clear indication of the overall distribution of bone within the pit. Although the southeast-center area was the deepest part of the pit, square DLI produced the most skeletal ma- terial. This is because the bones were piled higher on the FIGURE 7.—Division of Ossuary II (by barbed wire fence) into northeast and southwest halves, looking northwest. THE OSSUARIES FROM NANJEMOY CREEK 21 ^;3*Si#>v4^i> f*fy v«!?***-,*-» '':' / tl4* FIGURE 8.—Excavators working in square DL1, Ossuary II: a, looking southeast; b, looking northwest; c, looking northwest. 22 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 8 •.--•■« - * ■ :-A \ FIGURE 9.—Excavator lying on a board-support to work in square GR1, Ossuary II, looking southwest. sides. Of the total of 233.900 kilograms of bone in the pit, 90.915 kilograms (39 percent) were on the left side, 84.954 kilograms (36 percent) were on the right side and only 58.031 kilograms (25 percent) were in the center. Along the northwest-southeast axis, the maximum bone weight oc- curred in the "D" squares with the minimum in the "H" TABLE 2.- —Distribution of bt me xieight (in km) within Ossuary // % total % total Square Weight weight Square Weight weight AL1 12. 066 05. 16 FR1 11. 588 04. 95 BL1 14. 069 06. 01 GR1 8. 649 03. 70 CL1 7. 860 03. 36 HR1 2.048 00. 88 DL1 19.675 08.41 L 90. 915 38. 86 ELI 15. 572 06. 66 C 58.031 24. 81 FL1 10. 741 04. 59 R 84. 954 36. 32 GL1 8.475 03. 62 A 19. 319 08. 26 HL1 2.457 01. 05 B 31. 521 13. 47 AC 682 00.29 C 33. 223 14. 20 BC 3. 730 01. 59 D 41. 772 17. 86 CC 6. 671 02. 85 E 39. 202 16. 76 DC 8. 107 03. 47 F 39. 825 17. 02 EC 13. 936 05. 96 G 24. 302 10. 39 FC 17. 496 07. 48 H 4. 736 02. 03 GC 7. 178 03.07 A-B 50. 840 21. 74 HC 231 00. 10 C-D 74. 995 32. 06 AR1 6. 571 02. 81 E-F 79. 027 33. 79 BR1 13. 722 05. 87 G-H 29. 038 12.42 CR1 18. 692 07. 99 A-D 125. 835 53. 80 DR1 13. 990 05. 98 E-H 108. 065 46 20 ER1 9. 694 04. 14 Total 233. 900 100 00 squares. The weight distribution through the longitudinal axis largely reflects the greater depth of the southeast part of the trench. Whereas 125.835 kilograms (54 percent) came from the southeast half of the pit, only 108.065 kilograms (46 percent) were recovered from the northwest half. Since the bone concentration only partially extended into the "A" squares of the southeast end, only 19.319 kilograms (8 per- cent) of bone were recovered from those squares. In addition to human remains, Ossuary II included shell (worked and unworked), pottery sherds, projectile points, and animal bone. With the possible exception of shell beads, the cultural material appeared to be scattered haphazardly throughout the pit fill. Shell beads were found either inside crania or in concentrations that did suggest possible inten- tional placement. These beads may either have been placed in the pit in connection with the ritual of ossuary burial or were with the remains originally. ANIMAL BONE Animal bones, all fragmentary, represent seven species: white-tailed deer (Odocoileus virginianus), squirrel (Sciu- rus sp.?), fox (Vulpes sp.? or Urocyon sp.?), meadow jumping mouse (Zapus hudsonius), turkey (Meleagria gallopavo), muskrat (Ondatra zibethicus), and pine vole (Pitymys pinetorum). Most of the bones were either from white-tailed deer (over 32 fragments) or from turkey (eight fragments). Tables 3 and 4, which list the bones of these two species and their square distributions within Ossuary II, show clearly that the bones were randomly distributed throughout the pit. The other bones consisted of a right radius of a squirrel (Sciurus sp.?) from square ER1, a left ulna of a fox (Vulpes sp.? or Urocyon sp.?) from CL1, a right femur of a meadow jumping mouse (Zapus hudsonius) from DL1, and the following bones of a muskrat (Ondatra zibethicus) : vertebrae from AL1, right innominate from DL1, left femur from EC, the distal end of a right tibia from DR1 and a left mandible from the plow zone. These faunal remains probably do not represent grave offerings, but rather occupational refuse accidentally included. In addition, much TABLE 3.—Spatial distribution of bones of turkey (Meleagria gallopavo) within Ossuary II Bone Square Right innominate Femoral shaft Vertebra Left femur Vertebra Tarso-metarsus Tarso-metarsus DL1 DL1 DL1 FL1 FL1 DR1-ER1 DR1 THE OSSUARIES FROM NANJEMOY CREEK 23 FIGURE 10.—Bone concentrations in Ossuary II: a, squares AL1-HL1, looking southeast (note partially exposed articulated skeleton in wall); b, squares AL1-HL1, looking northwest; c, squares AR1-ER1, looking southwest. V v ''^ i * 'J*e THE OSSUARIES FROM NANJEMOY CREEK 25 FIGURE 12.—Northeastern extremity of Ossuary II: a, pit outline with bone concentration shown in profile, looking northeast; b, same section after soil was removed, looking northeast (note iso- lated bundles in the periphery of the bone con- centration) ; c, looking northwest. FIGURE 11.—Ossuary II: a, extended articulated skeleton in squares AC-DC, looking northeast; b, bone concentration in squares AC- EC, after removal of extended skeleton, looking north. 26 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 CR CC CLI FIGURE 13.—Profiles of Ossuary II: a, longitudinal profile through the center of ossuary; b, cross-section profile through the center of the "C" squares. TABLE 4.—Spatial distribution of bones of white-tailed deer (Odocoileus virginianus) within Ossuary II Bone Square of the skeleton of a pine vole (Pitymys pinetorum) was found inside a skull from square DC. The vole probably represents a secondary intruder into the ossuary. SHELL Shells representing seven species were found in Ossuary II: oyster: Crassostrea virginica; land snails: Anguispira alternata, Mesodon thyroidus, Triodopsis allolabris, Trio- do psis juxtidens; and seacoast land snail: Busycon carica. Oyster was the most plentiful, and examples were evenly scattered through all squares. In contrast, only 16 fragments of freshwater mussel (Elliptio complanata) were identified from seven squares. Like the faunal remains described above, the oyster and mussel shells probably do not represent of- ferings, but rather habitation refuse accidently mixed in with the pit fill. Land snails total 153 and represent four species. A single specimen of Triodopsis juxtidens was recovered from inside a skull in DLL Square GL1 produced one specimen of Triodopsis allolabris. Seven examples of Mesodon thyroidus were found within squares AL1, FL1, GL1, and AC. Two of these were inside crania and one appeared to be a recent specimen. The remaining 144 land snails, identified as An- guispira alternata, were found throughout the pit. Many of these snails were encountered inside crania, with a total of AL1 AL1 BL1 CLI DL1 DL1 DL1 ELI FL1 FL1 GL1 GL1 GL1 AC AC BC DC DC EC FC FC GC CR1 CR1 DR1 DR1 ER1 Inside skull Inside skull Inside skull Plow zone Plow zone Left first phalanx Temporal Right first phalanx Left astragulus Left central fourth tarsal Several long bone shafts Right first phalanx Several long bone shafts Left scapula Right rib Left distal femur Metacarpal shaft Left third phalanx Several long bone shafts Tarsal Right humerus Left zygomatic Left rib Several long bone shafts Several long bone shafts Right radius Left zygomatic Left astragulus Long bone shaft Right proximal tibia Right acetabulum of innominate Long bone shafts Left calcaneus Left distal tibia Lift distal radius Right humerus Left proximal femur THE OSSUARIES FROM NANJEMOY CREEK 27 65 originating from a single skull between AL1 and BL1. Initially, it appeared that the shells had been intentionally placed inside the skull. Later examination by Dr. Joseph Morrison, Curator of Mollusks, Department of Invertebrate Zoology, National Museum of Natural History, suggested that while they could have been placed in the skulls, mey more probably entered either by crawling up the scaffolds during the summer months or by tunneling down to die buried bones in late autumn. The presence of the fresh Mesodon thyroidus from AL1 favors a cold weather tunnel- ing interpretation. The remainder of the shell sample from Ossuary II con- sists of 107 large and 111 small shell beads. Ninety-one of the large beads are barrel-shaped, drilled from both ends, and vary in width from 6-11 mm and in length from 6-16 mm. A concentration of the large beads was noted in the middle of the ossuary and is borne out by the following square frequencies: four in BC, 30 in CC, 13 in DC, 12 in EC, 7 in FC, 2 in CR1, 21 in DR1, and 2 in GR1. Thirty-two were found inside crania, 27 in a skull in CC, and 5 in 2 skulls in DR1. The remainder of the large shell beads (16) were elon- gated, cylinders ranging in width from 5-7 mm and in length from 13-20 mm. These also were concentrated in the mid- dle of the center and right rows, with 1 in BC, 5 in CC, 3 in DC, 1 in CR1, and 6 in DR1. Just one bead was found inside a skull (DR1). According to Dr. Joseph Morrison, all of the large beads probably were cut from the columella of Busycon carica, a sea-coast land snail that occurs on the Maryland eastern shore. The concentration of the beads in one area within the pit and the occurrence of many within skulls strongly suggest that they were either placed in the ossuary at the time of burial or perhaps were placed with the body on the scaffold and then brought along when the bones were transferred to the pit. All of the 111 small beads were found around the arti- culated bones of a left foot associated with a bundle in GR1. They consistently measure 3 mm in width by 2 mm in length. Stephenson (in Stephenson and Ferguson, 1963:163-164) reported 986 of these beads from the Accokeek Creek site and stated they were made from mussel shell. POTTERY The 126 pottery sherds recovered from Ossuary II were scattered rather uniformly throughout the pit and probably were accidently added to the skeletal deposit when the pit was filled. According to Clifford Evans and George R. Phe- bus, Department of Anthropology, National Museum of Natural History, the sherds may be classified into the follow- ing type frequencies: Potomac Creek, 105; Moyoane, 10; Rappahannock, 6; and unknown, 5. These wares and pottery types have been described by Evans (1955) and Stephen- son and Ferguson (1963) and will not be discussed in detail here. As Table 5 indicates, the type frequencies are very similar to those from Ossuary I, and to a sample collected in the adjacent occupation area. All three of these samples demonstrate a high frequency of Potomac Creek and mar- kedly low frequencies of other, probably earlier, wares. TABLE 5.—Frequencies of ceramic types in ossuaries and habitation area Ceramic type Ossuary I Ossuary II Habitation area No. % No. % No. % Potomac Creek Moyoane 97 87.39 105 83.33 657 91.25 3 2. 70 10 7. 94 48 6 67 Rappahannock 3 2. 70 6 4.76 12 1.67 7 6.31 0 0 0 0 Stony Creek Unknown 0 0 0 0 3 0.42 1 0.90 5 3.97 0 0 Total 111 100.00 126 100.00 720 100.00 Other artifacts recovered in the excavation of Ossuary II include two bowl and two stem fragments from angular, un- derrated native-manufactured clay pipes; two small, trian- gular, quartz projectile points, and numerous quartz frag- ments. The projectile points strongly resemble the small, thin, triangular Potomac Points described by Stephenson and Ferguson (1963), that were produced by the pressure- flaking of white quartz. Collectively the cultural material suggests that both os- suaries and the nearby occupation area represent a 16th- century Late Woodand occupation, just before European contact in that area. The strong similarity between artifact assemblages from the two ossuaries suggests that aldiough one may be older than the other, the difference in age be- tween them is very slight. Since standards for recognizing subtle, temporal changes in artifacts are not yet available in the mid-Atlantic region, it cannot at this time be deter- mined from the cultural associations which of the ossuaries is the older. The associated artifacts indicate, however, that both of the ossuaries were pre-European, and that they represent approximately the same time period. HUMAN REMAINS The excavation of Ossuary II produced a total of 233.900 kilograms of human bone. Most of the material was well preserved and probably only a few of the smaller, infant bones were decomposed and lost after interment in the pit. All of the soil was sifted through a one-quarter inch screen, 28 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 so it is doubtful that any important fragments were missed during excavation. As stated earlier, four categories of human skeletal material were found in the pit: (1) com- pletely articulated skeletons; (2) scattered, partially artic- ulated skeletal parts; (3) bundles of both articulated and disarticulated bones; and (4) scattered, disarticulated bones. There were three completely articulated adult skeletons, apparently representing individuals who died shortly before the ceremony. One a young female, was lying on her back on top of the bone pile (Figure 11a, b) with legs extended. The other two were on the floor of the pit in CR1 and DR1 (Figure 14). One of these (35-40 year old male) was lying on his back (Figure 15a) with his legs very tightly flexed. The other (30-35 year old female) was lying face down with her femora extended (Figure 156, c) and her lower leg bones unnaturally flexed underneath the femora. Since the knee articulations were preserved, it appears that the lower legs had been placed deliberately in the above posi- tion, presumably to conserve space, as Stewart (1941:70) has suggested. Many partially articulated bones were found scattered throughout the bone concentration (Figures 15a", 16a). Such remains undoubtedly represent individuals who were incom- pletely decomposed at the time of ossuary deposit. These individuals had been dead longer than the three completely articulated individuals, but not as long as those represented by completely disarticulated bones. Table 6 presents a general listing of frequencies of partial articulation representing dif- ferent parts of the adult body. The table reveals that the greatest number of individuals (23) was represented by foot bones, followed closely by the tibia and fibula (20), and the thoracic vertebrae (20). To a large extent, the distribution in Table 6 reflects the relative strength of muscle, and ligament attachments and their resistance to decomposition. Appar- ently, decomposition produces separation first at the major joints such as the shoulder, elbow, wrist, hip, and knee. Separation next occurs at the joints between the sacrum and pelvis, bones of the hand, lower leg and foot, radius and ulna, sacrum and fifth lumbar vertebrae, skull and first cervical vertebrae, the lumbar segments, first and second cervical ver- tebrae, skull and mandible, and the third to seventh cervical segments. The thoracic vertebrae, tibia and fibula, and bones of the feet are the last to become disarticulated. Only 14 & FIGURE 14.—Articulated adult skeletons on the floor of Ossuary II, looking northwest. FIGURE 15.—Ossuary II: a, nearly completely articulated 35 to 40-year-old male from squares CR1-DR1, looking northwest; b, articulated 30 to 35- year-old female from squares CR1—DR1, looking northwest; c, closer view of articulated female (note unnatural position of the lower legs) ; d, eight articulated thoracic vertebrae from square GR1. 30 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER II FIGURE 16.—Ossuary II: a, two examples of partial articulation (three articulated thoracic vertebrae and five articulated left metatarsals) ; b, isolated ^ f - bundle in squares GR1-HR1. examples of partially articulated subadult bones were re- corded. Of these, 12 involved vertebrae, one carpal bones, and one a foot. The fewer examples of partially articulated subadult bones reflect both a faster subadult decomposition rate and the greater difficulty in detecting subadult articula- tion during excavation. Many of the partially articulated and wholly disarticulated bones probably were arranged originally in bundles. Unfor- tunately, formerly distinct bone clusters tended to blend together in the bone mass, and bundles could be distinguished only in the shallow northwest end (Figures 12b, c, 166). In this area, five bundles were located on the periphery of the bone mass from squares ER1, FR1, GR1, HR1, GC, and HO All contained bodi articulated and disarticulated bones, and all but one were composed of parts of several individuals, both adult and subadult. The exception was a bundle from GC that contained the relatively complete remains of an 18 or 19 year old female. Possible cremations were represented by 416 fragments of burned human bone. Of 320 identifiable fragments, 82 were from crania, 5 from mandibles, 2 from the scapulae, 4 from vertebrae, and 227 from long bones. Only adult bone was recognized but the presence of subadults could not be ruled out. The occurrence of two charred adult right mandibular condyles suggests that at least two individuals were rep- resented. As Table 7 shows, burned bones were found in 14 squares of Ossuary II, but were most frequent in EC and FC in the center of the pit. Within these squares, they were con- centrated on top of the bone pile. None of the surrounding bones had been scorched, indicating that the burning had occurred elsewhere before the remains were brought to the pit for burial. The evidence of exposure to fire varied, some THE OSSUARIES FROM NANJEMOY CREEK 31 fragments being completely calcined, some bleached and others barely scorched. Van Vark's (1970) laboratory studies suggest the calcined (white) fragments had been fired at a temperature over 800° C (1472° F) and the others at lower temperatures. TABLE 6.—Frequencies of articulated adult bones in Ossuary II One hundred and forty-one skulls of all ages were found more or less intact and numbered during the excavation of Ossuary II. Contrary to Ferguson's observations at Accokeek Creek, the skulls appeared to be neither segregated from the rest of the bone mass nor placed in any particular section of TABLE 8.—Frequencies of skull positions in Ossuary II Minimum no. of individuals Number of occurrences Articulated bones 23 20 20 13 12 11 11 9 8 7 6 4 3 46 40 59 (238 vert.) 21 (62 vert.) 12 11 19 (5 vert.) 9 88 14 12 7 3 2 2 Bones of foot Tibia-fibula Thoracic vertebra 3-7 cervical vertebra Skull-mandible 1-2 cervical vertebra Lumbar vertebra Occipital—1st cervical vertebra. Sacrum-lumbar vertebra Radius-ulna Tibia, fibula-bones of foot.... Bones of hand Sacrum-pelvis Femur-innominate Radius, ulna-bones of hand. . . Humerus-scapula Humerus-radius, ulna Femur-tibia, fibula Femur-patella Tibia, fibula-patella TABLE 7.— Spatial distribution of burned bone within Ossuary II Mandibular Square ■ and cranial Postcranial Undetermined Total No. % No. % No. % No. % AL1 - - 5 2 - - 5 1 AC — — — — _ _ : - AR1 - BL1 - - 1 1 - - i - BC — — — " : - BR1 CLI - - - - - - - - CC — — 4 1 _ _ 4 — CR1 1 DL1 - - 2 1 19 20 21 5 DC - - - - - - - - DR1 - - 1 1 1 1 2 1 ELI 6 7 16 7 8 8 30 7 EC 3 3 37 3 16 1 15 16 55 3 13 ER1 1 FL1 - - 1 1 - - 1 - FC 65 75 151 3 65 1 46 48 262 3 63 FR1 1 GL1 1 1 - - 2 2 3 1 GC 8 4 9 5 3 6 1 2 3 2 3 2 14 12 3 GR1 3 HL1 - - - - - - - - HC — — _ _ _ _ _ - HR1 - Total 87 100 233 100 96 100 416 100 Left side Base Face Ver- tex Right side ? Total Number 38 27 36 26 12 8 26 18 21 15 8 6 141 Percent 100 ? = Orientation uncertain due to excessive fragmentation of crania. TABLE 9.—Frequencies of face orientations in Ossuary II N NW W SW S SE E NE p Face down Total Number.... Percent... . 22 16 20 14 18 13 10 24 7 17 10 20 8 7 14 6 1 5 2 1 141 100 .'^Orientation uncertain due to excessive fragmentation of crania. the pit. In addition nearly all skull orientations were rep- resented (Tables 8, 9) with no particular position or direc- tion predominating. The presence of numerous smaller adult and subadult bones inside many skulls suggests that the skulls may have been utilized as containers to transport bones from the primary deposit to the ossuary. Total Number of Individuals At least 131 individuals are represented in Ossuary I and 188 in Ossuary II. Since ossuaries are secondary deposits of largely disarticulated bones, these totals were obtained by carefully listing the frequencies of each type of bone from both ossuaries and then comparing the minimum numbers of individuals represented. All complete and fragmentary bones were examined to make this count as accurate as possible. From Ossuary II, the counts of proximal and distal ends of complete and fragmentary long limb bones were used to indicate the number of adult individuals represented (Table 10). The number and size of isolated shaft fragments were examined also, but only in the fibula did they affect the mini- mum number of individuals represented. Likewise in Os- suary I (Table 11) the minimum number of adult individ- uals represented was determined from counts of proximal and distal ends for all long bones except the fibula where 54 left and 53 right adult fibulae were represented by proximal and distal ends, whereas the shaft fragments indicated that at least 61 adult individuals were actually present. 32 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 TABLE 10.—Counts of complete and fragmentary adult long limb bones from Ossuary II Complete L R Proximal Distal Minimum no. individuals represented Bone L R L R L R Humerus 68 66 64 78 75 44 70 63 62 73 78 48 1 13 24 4 6 13 12 14 24 10 5 14 19 14 13 3 6 30 14 17 15 8 15 33 87 84 80 80 Ulna 88 86 82 83 Tibia 81 93 Fibula 74 81 OSSUARY I At the time my analysis began, the skeletal material from Ossuary I already had been segregated into major bone groups. This preliminary sorting facilitated the collective examination that revealed the minimum number of adults (Table 11) and subadults (Table 12) represented by each kind of bone. In general, the "adult category" includes those individuals over 18 years at death, while those 18 and younger were classified as subadults. Since many of the smaller bones of the skeleton attain morphological maturity prior to 18 years, there may have been a tendency to include some smaller subadult bones in the adult category. However, this number is probably minimal since there were few indi- TABLE 11.—Number of individuals represented by each type of adult bone in Ossuary I (numbers in parentheses indicate actual number of bones) Bone Left Right Bone Left Right LONG BONES: Humerus Radius Ulna Femur Tibia Fibula IRREGULAR BONES: Clavicle Scapula Temporal Maxilla Mandible Gladiolus Manubrium Innominate Patella VERTEBRAE: Cervical: 1 2 3-7 Thoracic: 1-9 10 11 12 10-11 10-11-12 Lumbar Sacrum HAND BONES: Carpals: Navicular Lunate Triquetral Pisiform Greater Multangular. Lesser Multangular.. 68 64 64 65 65 60 66 68 68 69 54* 53* 60 54 64 57 55 56 50 50 62 30 29 63 62 61 51 46 63 51(255) 50(446) 28 36 37 37 32 50(248) 56 54 33 36 29 26 25 10(19) 17 22 25 19 22 HAND BONES—Continued Carpals—Continued Capitate Hamate Metacarpals: 1 2 3 4 5 Phalanges: Proximal 1-5 Middle Distal 1-5 FOOT BONES: Tarsals: Calcaneus Talus Cuboid Navicular Cuneiforms: 1 2 3 Metatarsals : 1 2 3 4 5 Phalanges: Proximal: 1 2-5 Middle Distal: 1 2-5 34 33 30 32 48 39 51 51 50 57 34 43 53 48(472) 31(243) 16(155) 44 60 60 61 58 57 54 58 54 52 52 43 38 54 42 50 53 52 54 49 47 55 47 62 42(83) 27(214) 6(41) 28(56) 3(19) 57 *Shaft fragments indicate at least 61 adults present THE OSSUARIES FROM NANJEMOY CREEK 33 TABLE 12.—Number of individuals represented by each type of subadult bone in Ossuary I (numbers in parentheses indicate actual number of bones) Bone Left Right 55 33 51 51 49 38 30 62 33 48 47 32 18 17 8 17(68) 5(28) 5(9) 55 32 Ulna 45 58 Tibia 50 Fibula Clavicle 26 36 61 Maxilla 37 44 48 39 Pubis 24 Patella 16 10 evident in Tables 15 (Ossuary I, adults), 16 (Ossuary I, subadults), 17 (Ossuary II, adults), and 18 (Ossuary II, subadults). These tables rank the bones in the order of the number of individuals they represent. Thus, according to Table 15, 69 adult individuals are represented in Ossuary I by tibiae, 68 by femora, 55 by fourth metatarsals, etc. These tables also show the percentages of represented and unrepre- sented individuals by each bone. According to Table 15, the patellae of 54 adults were recovered from Ossuary I (78 percent of the total number of all of the adults represented therein), and the patellae of 15 adults (22 percent) are missing. The differential representation of bones in Ossuaries I and II could reflect such factors as (1) loss of bone prior to secondary burial, (2) intentional cultural selection at the time of ossuary deposit, (3) differential decomposition in the ground, (4) accidental loss during excavation, or (5) accidental loss after excavation. Decomposition in the ground may account for the loss of some of the smaller infant bones and adult phalanges, ribs, and occasionally vertebrae. During excavation of Ossuary II, several of the infant bones crumbled upon removal. Although an attempt viduals in either ossuary between ages 15 and 20 years. The adult right tibia (69) and the subadult left temporal (62) were most numerous and indicate at least 131 individuals were present in Ossuary I. OSSUARY II To compute bone totals for Ossuary II, the contents of each square were analyzed separately and then totaled. This approach was taken because (1) following excavation and processing, the bones of each square were cataloged and stored together, and (2) bone frequencies from each square were needed to study spatial distribution features within the ossuary. Totals from all squares in Ossuary II are pre- sented in Table 13 (subadults) and Table 14 (adults). The minimum number of individuals in Ossuary II (188) was determined from the left subadult femora (89) and the right adult mandibles (99) which had the highest fre- quencies. Variability in Bone Representation This "bone-by-bone" inventory not only revealed the number of individuals in each ossuary, but also showed the great variability in numbers of the different bones. In Os- suary I, 69 adults were represented by tibiae, but only 3 by the second through fifth distal foot phalanges. Whereas 62 subadults were represented by temporals, only 5 were rep- resented by patellae. In Ossuary II, 99 adults were rep- resented by adult right mandibles, but only 9 by coccygeal vertebrae. Subadult variation ranged from 89 (left femora) to 1 (carpals). This variance in bone representation is clearly TABLE 13.—Number of individuals represented by each type of subadult bone in Ossuary II (numbers in parentheses indicate actual number of bones) Bone Left Right Humerus 71 68 Radius 47 45 Ulna 58 54 Femur 89 82 Tibia 67 75 Fibula 35 35 Clavicle 47 49 Scapula 59 52 Temporal 81 84 Maxilla 49 49 Mandible 52 49 Gladiolus 10 Manubrium 13 71 58 Ischium 43 48 Pubis 41 22 Patella 0 1 Rib 19 (494) Vertebrae* 29 (775) 10 Sacrum Coccyx 1 (4) Calcaneus 18 25 Talus 10 (20) Other Tarsals 3 (28) Carpals 1 (10) Metatarsals or Metacarpals 16 (295) 6 (271) *Cervicals, thoracics, lumbars. 525-413 O - 74 - 4 34 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 TABLE 14.—Number of individuals represented by each type of adult bone in Ossuary II (numbers in parentheses indicate actual number of bones) Bone Left Right Bone Left Right LONG BONES Humerus Radius Ulna Femur Tibia Fibula Clavicle Scapula Temporal Maxilla IRREGULAR BONES Mandible Gladiolus Manubrium Innominate Patella Rib VERTEBRAE Cervical: 1 2 3-7 Thoracic: 1-9 10 11 12 10 or 11 11 or 12 10, 11 or 12 Lumbar Sacrum Coccyx HAND BONES Carpals: Navicular Lunate. . . . ._± Triquetral Pisiform Greater Multangular. Lesser Multangular.. 87 84 80 80 88 86 82 85 81 93 74 81 70 72 85 92 92 91 86 81 98 99 56 58 84 86 73 78 72(1727) 98 97 76(389) 72(645) 63 80 59 37 11 77(385) 84 9(33) 59 63 54 46 41 24 18(35) 45 55 37 38 HAND BONES—Continued Carpals—Continued Capitate Hamate Metacarpals: 1 2 3 4 5 Phalanges: Proximal: 1 2-5 Middle Distal: 1 2-5 FOOT BONES Tarsals: Calcaneus Talus Cuboid Navicular Cuneiforms: 1 2 3 Metatarsals: 1 2 3 4 5 Phalanges: Proximal: 1 2-5 Middle Distal: 1 2-5 69 58 51 55 66 68 74 65 74 77 56 54 69 74 53(106) 66(524) 44(348) 29(57) 23(181) 86 86 93 90 70 73 69 80 78 73 73 68 73 75 82 74 72 63 77 85 77 74 81 73 60(125) 46(385) 19(118) 45(89) 9(55) was made to record bones lost in this manner, it is possible that a few decomposed bones were overlooked. Stewart's field notes and personal comments indicate that some small decomposed bones may have been lost from Ossuary I as well. It is doubtful that any bones were lost by vandalism or by accident while excavation was in progress. Since both were located on private property within 30 meters (100 ft) of the Juhle residence, vandalism was discouraged, and probably would have been detected had it occurred. Dur- ing the excavation of Ossuary II, all backdirt was sifted through a 5 mm (*4-in) screen that made the loss of even small infant bones unlikely. The material from Ossuary II was cleaned, restored, cataloged, and analyzed immedi- ately after excavation, either by myself or by Smithsonian volunteers directly under my supervision. Similar efforts were made to safeguard the Ossuary I sample. During die 17 years between the excavation and analysis, the material had been isolated from the regularly studied collections and stored on the fourth rotunda floor of the Natural History Building of the Smithsonian Institution. Consequently there is little reason to believe that any of that material had been lost over the years. It follows from the foregoing tiiat most of die missing bones were lost prior to the time of ossuary burial. Appar- ently the Indians did not transfer all of the bones from the THE OSSUARIES FROM NANJEMOY CREEK 35 TABLE 15.—Order of representation of adults in Ossuary I as indicated by bone types primary deposit to the ossuary. Instead, they selected those tiiat best represented the individual. Thus, in Ossuary I (Table 15), the greatest number of adults is represented by long bones. This is surprising since one would expect the Indians to associate maxillae and mandibles more closely with the individuals known in life. Yet 9 percent of the mandibles and 28 percent of die maxillae are missing. Of course, it remains anyone's guess what happened to them, but it is unlikely that the skull and mandible were left lying around the scaffold or death house. More probably a few of the skulls and mandibles were taken elsewhere for memorial purposes. The rest of the adult bone frequencies presented in Table 15 generally reflect what one would expect, namely, the occurrence of larger and more conspicious bones in greater abundance. Twelve percent of fibulae are missing, reflecting to a large extent the tendency of the fibulae to break easily. Once broken, the fragments could have been misplaced. The adult second cervical vertebrae are surpris- ingly well represented, with only 9 percent missing. A high representation of second cervicals would be expected if many of them were still attached to the base of the skull when the bones were transferred. However, the fact that 19 per- cent of the skulls (temporals) and 33 percent of the first cervical vertebrae are missing seems to discourage that in- terpretation. Foot bones are better represented than hand bones. (62 individuals represented by fifth metatarsals, 57 by third metacarpals), probably reflecting the greater tendency of the ligaments of the foot to resist decomposition and allow the bones to remain articulated. Bone representation of subadults in Ossuary I is shown in Table 16. In general, the inventory shows a high frequency of the larger bones and a poor representation of smaller bones, suggesting that the latter often may have been lost from the original scaffolds or dropped during the transfer to the ossuary. The subadults were represented most fre- quently by temporals, followed closely by long bones. Since the temporal is part of the skull and highly resistant to decomposition, it is not surprising that it occurs with such a high frequency. Although adult representation in Ossuary II (Table 17) is similar to that in Ossuary I, there are some striking differ- ences. In both ossuaries, more individuals are represented by larger, easily recognized bones than by the smaller, in- conspicuous bones. However, the frequency of occurrence is different. In Ossuary I, the maximum counts come from the tibiae and femora, with 9 percent of the mandibles missing, whereas in Ossuary II the maximum counts come from the mandible, with 14 percent of the femora missing. Second and third in frequency in Ossuary II, are the first and second cervical vertebrae. There are more tali than ulnae and humerii, and more individuals represented by third meta- tarsals than by femora, fibulae, or radii. For some reasons, 6 percent of the tibiae, 11 percent of the ulnae, 12 percent of the humeri, 14 percent of the femora, 18 percent of the Bone Tibia Femora Humerus Radius Ulna Scapula Mandible 2nd cervical vertebrae Fifth metatarsal Innominate Talus Fibula Clavicle Calcaneus Foot navicular Third metacarpal Cuboid Temporal Thoracic 10-12 Sacrum Fourth metatarsal Second metatarsal Patella Fifth metacarpal First metacarpal First cuneiform Second metacarpal 3-7 cervical vertebrae Maxilla 1-9 thoracic vertebrae Lumbar vertebrae Third metatarsal First metacarpal 1-5 prox. hand phalanges.. . First cervical vertebrae Fourth metacarpal Second cuneiform First prox. foot phalange. . . Hand navicular Capitate 1—4 middle hand phalanges. Hamate Gladiolus (sternum) Manubrium (sternum) Lunate First dist. foot phalange 2-5 prox. foot phalanges.... Greater multangular Triquetral Lesser multangular 1-5 distal hand phalanges. . Pisiform Middle foot phalange 2-5 distal foot phalanges. . . Absent Represented % % No. No. 69 100 0 0 68 99 1 1 68 99 1 1 65 94 4 6 65 94 4 6 64 93 5 7 63 91 6 9 63 91 6 9 62 90 7 10 62 90 7 10 61 88 8 12 61 88 8 12 60 87 9 13 60 87 9 13 58 84 11 16 57 83 12 17 57 83 12 17 56 81 13 19 56 81 13 19 56 81 13 19 55 80 14 20 54 78 15 22 54 78 15 22 53 77 16 23 53 77 16 23 52 75 17 25 51 74 18 26 51 74 18 26 50 72 19 28 50 72 19 28 50 72 19 28 49 71 20 29 48 70 21 30 48 70 21 30 46 67 23 33 43 62 26 38 43 62 26 38 42 61 27 39 36 52 33 48 34 49 35 51 31 45 38 55 30 43 39 57 30 43 39 57 29 42 40 58 29 42 40 58 28 41 41 59 27 39 42 61 25 36 44 64 25 36 44 64 22 32 47 68 16 23 53 77 10 14 59 86 6 9 63 91 3 4 66 96 36 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 TABLE 16.—Order of representation of subadults in Ossuary I as indicated by bone types TABLE 17.—Order of representation of adults in Ossuary II as indicated by bone types Bone Represented Absent No. % No. % 62 58 55 51 48 50 48 39 37 33 28 24 17 17 10 5 5 100 94 89 83 77 80 77 63 60 53 45 39 27 27 16 8 8 0 4 7 11 14 12 14 23 25 29 34 38 45 45 52 57 57 0 6 Humerus 11 Ulna 17 23 Tibia 20 Mandible 23 Ischium 37 Maxilla 40 Radius 47 Clavicle 55 Pubis 61 Fibula 73 Calcaneus 73 Talus 84 Patella 92 92 fibulae, and 19 percent of the radii were lost prior to ossuary burial. It is doubtful that such large bones could have been accidentally lost or left in the scaffolds. More probably, those bones were selected for disposal elsewhere. Once again, the predominance of foot bones (93 individuals represented by the talus) over hand bones (74 individuals represented by 5th metacarpals) probably reflects the tendency for the former to remain articulated (see Table 6). The high fre- quency of first and second cervical vertebrae is again difficult to explain. One would expect that some of these vertebrae might still have been attached to the skull at the time of ossuary burial. This would explain why more individuals are represented by the first and second cervicals than by the rest of the cervical vertebrae (76 percent), but not why there are more first and second vertebrae than bones of the skull (temporals 93 percent, maxillae 87 percent). The representation of subadults in Ossuary II (Table 18) is very similar to that in Ossuary I (Table 16). Here too, the long bones, temporals, and other larger bones occur most frequently. In Ossuary I, however, the most subadults are represented by temporals and 6 percent of the femora are missing, whereas in Ossuary II the maximum count comes from the femora with 6 percent of the temporals missing. These differences in the skeletal inventories from the two ossuaries not only reveal the type of skeletal material selected for ossuary burial, but also demonstrate the fallacy of relying upon counts of a single skeletal part for reconstruction of the number of individuals in an ossuary burial. An adult count based on femora would be 99 percent correct in Ossuary I, but would underestimate the total in Ossuary II by 14 per- cent. Estimates based solely on the number of skulls are Bone Represented Abs No. ent No. % % Mandible 99 98 97 93 93 92 92 88 87 86 86 86 85 85 84 84 82 81 81 80 80 78 78 77 77 77 76 75 74 74 73 73 72 72 72 72 69 68 66 63 60 58 56 56 55 55 54 53 46 45 44 41 38 29 23 19 18 9 9 100 99 98 94 94 93 93 89 88 87 87 87 86 86 85 85 83 82 82 81 81 79 79 78 78 78 77 76 75 75 74 74 73 73 73 73 70 69 67 64 61 59 57 57 56 56 55 54 46 45 44 41 38 29 23 19 18 9 9 0 1 2 6 6 7 7 11 12 13 13 13 14 14 15 15 17 18 18 19 19 21 21 22 22 22 23 24 25 25 26 26 27 27 27 27 30 31 33 36 39 41 43 43 44 44 45 46 53 54 55 58 61 70 76 80 81 90 90 0 1 st cervical vertebrae 1 2nd cervical vertebrae 2 Tibia 6 Talus 6 Scapula 7 Temporal 7 Ulna 1 1 Humerus 12 Innominate 13 Maxilla 13 Calcaneus 13 3rd metatarsal 14 Femur 14 10—11—12 thoracic vertebra 15 Sacrum 15 1st metatarsal 17 5th metatarsal 18 Fibula 18 Radius 19 Foot navicular 19 Patella 21 1st cuneiform 21 4th metatarsal 22 3rd metatarsal 22 Lumbar vertebrae 22 3-7 cervical vertebrae 23 3rd cuneiform 24 5th metacarpal 25 2nd metacarpal 25 26 2nd cuneiform 26 2nd metatarsal 27 1-9 thoracic vertebrae 27 Clavicle 27 Rib 27 Capitate 30 1 st metacarpal 31 Prox. 2-5 hand phalanges 33 Hand navicular 36 1st prox. foot phalange 39 Manubrium 41 Gladiolus 43 4th metacarpal 43 Greater multangular 44 Hamate 44 Lunate 45 1st prox. hand phalange 46 2-5 distal foot phalanges 54 1st distal foot phalange 55 Middle hand phalanges 56 Triquetral 59 Lesser multangular 62 1st distal hand phalange 71 2-5 distal hand phalanges 77 Middle foot phalanges 81 Pisifoi m 82 Coccygeal vertebrae 91 2-5 distal foot phalanges 91 THE OSSUARIES FROM NANJEMOY CREEK 37 TABLE 18.—Order of representation of subadults in Ossuary II as indicated by bone types TABLE 19.—Partial inventory of bones from Fairty Ossuary, Ontario, Canada (after Anderson, 1964:28) Represented Absent % % No. No. Bone 89 84 75 72 71 60 58 52 49 49 49 47 41 35 29 25 24 16 13 10 10 6 3 3 1 1 100 94 84 81 80 67 65 58 55 55 55 53 46 39 33 28 27 18 15 11 11 7 3 3 1 1 0 5 14 17 18 29 31 37 40 40 40 42 48 54 60 65 66 73 76 79 79 83 86 86 0 6 16 19 20 33 35 42 45 45 45 48 54 61 67 72 73 82 85 89 89 93 97 97 99 Femur Temporal Tibia Ilium Humerus Scapula Ulna Mandible Maxilla Ischium Clavicle Radius Pubis Fibula Vertebrae Calcaneus Ribs Metatarsals-carpals Manubrium Gladiolus Talus Phalanges Other tarsals Sacrum Carpals Patella also misleading. Although 94 skulls were recorded during the excavation of Ossuary I and 141 during the excavation of Ossuary II, the actual number of individuals in the pits are 131 and 188, respectively. The differential representation of bones in ossuary de- posits has been discussed by others. Anderson (1964) noted in his study of an Iroquois ossuary that a wide range existed in the minimum number of individuals represented by each bone (Table 19). His largest adult count came from the talus, followed by the temporal, ulna, and humerus, while his largest subadult count came from the humerus, followed by the femur, pelvis, ulna, and radius. No data were given on the frequencies of mandibles and maxillae. Churcher and Kenyon (1960:256) relied solely upon long-bone counts for their reconstruction of the number of individuals in two Iroquoian ossuaries from Toronto, since long bones were well preserved and could easily be identified as left or right. Although their analysis would be more meaningful if their bone inventory was complete, they do show that long bone frequencies varied from 213 individuals represented by femora to only 64 represented by radii. Unfortunately, complete skeletal samples were not saved from ossuary excavations in the mid-Atlantic region prior to 1950. Stewart encouraged Ferguson, Graham, and others to Bone Number oj items cataloged Minimum number individuals of Mature Immature Total 931 507 74*7 1,541 3,238 1,460 182 404 908 907 1,591 951 1, 169 1,753 1, 176 484 981 850 769 675 346 274 287 258 221 218 182 122 214 232 295 196 307 198 223 250 207 177 354 281 144 47 93 49 49 75 ■? 46 129 135 217 151 152 156 172 ? 147 83 40 69 490 Atlas 321 380 307 270 293 1824- 168 343 367 512 347 Ulna 459 354 395 Patella 2504- 354 260 Talus 394 350 save certain bones for his analysis, but he had no assurance that the samples were complete. Still, he determined that more individuals were represented by the skull and mandible than by long bones and judged that of all individual bones, the temporal would "furnish the best check on the number of individuals" (Stewart, 1940c: 15). Spatial Distribution Features within Ossuary II Past students of mid-Atlantic ossuary burial practices generally have suggested that skeletal material is randomly scattered within ossuaries with no evidence for order or structure. Referring to her excavation at Accokeek Creek, Ferguson (in Stephenson and Ferguson, 1963:68) stated that "the ossuaries were very similar. Most of the bones were flung into the pit with no regard for order or sequence." In his discussion of the Port Tobacco ossuaries, Graham (1935:17) similarly reported that "there was apparently no particular order used in placing the bodies, or skeletons in the pit." Other discussion by Ferguson and Graham suggests that, except for the presence of bundles, articulated indi- viduals, and small pockets of cremated bone, no order exists within ossuary burials. The use of the arbitrary system of 0.6-meter (2 ft) squares in the excavation of Ossuary II allowed the "random- scattered" pattern to be examined objectively. The contents of each square were analyzed independently. The total num- ber of each type of adult and subadult bone, and the sex and 38 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 age of adults as estimated from the skull and pubis were listed for each square. Adult age was divided into two gen- eral categories: those over 30 years and those between 18 and 30 years. Variables such as age, sex, and variety of bone in each square could then be tested against each other to determine if the spatial distribution within Ossuary II was truly random. A multivariate canonical analysis was used to examine a number of the variables simultaneously and the univariate chi square statistic was applied to test indi- vidual variables against each other. A combination of these two approaches revealed which variables were probably randomly distributed and which demonstrated spatial differ- ences of possible cultural significance. According to the multivariate canonical analysis, there were no significant distributional differences among adults in age (under 30 years vs. over 30 years), sex (male vs. female), or type of bone (skull vs. pubis). All distribution variance among these variables can be accounted for by chance alone as Ferguson and Graham had concluded. Significant differ- ences were discovered, however, when the distribution of adult bones was compared to that of subadult bones. These differences are perhaps best expressed by the individual chi square values. Chi square values were calculated for each pair of vari- es (0-E)2 ables tested according to the formula X2 = ^ , where 0 equals the observed frequency and E equals the expected frequency. Table 20 presents chi square values for differences between the distribution of young and old adults. Chi square values of 2.28 for the pubis and 2.48 for the skull support the results of the multivariate analysis indicating that the differences are probably due to chance. Table 21 presents the data on distributional differences between adult males and females. Again, chi square values suggest that the differences are probably due to chance alone and agree with the results obtained through the multivariate analysis. Table 22 presents chi square values for subadults vs. adult distributional differences in 13 different bones. The tem- porals, clavicles, radii, femora, and scapulae show no sig- nificant differences between subadults and adults. Significant differences, however, were recorded from the tibiae, humeri, pubes, ribs, tarsals, carpals, and vertebrae. Values for the vertebrae, and tarsals and carpals (168 and 68) indicate the odds are less than one in 1000 that the distributional differ- ences could have resulted from chance alone. These data TABLE 20.—Chi square values for distributional associations between young and old adults within Ossuary II Variables tested Chi Degrees Proba- square of bility value freedom 2.28 2.48 Pubis age 18-29 vs. pubis age 30 plus. Skull age 18-29 vs. skull age 30 plus. . .•10-. 20 . 10-. 20 TABLE 21.—Chi square values for distributional associations between adult males and females within Ossuary II Chi Degrees Proba- square of bility value freedom Variables tested . 05-. 10 . 20-. 30 . 10-. 20 3.80 1.31 2.28 Male pubis vs. female pubis Male skull vs. female skull. Adult skull vs. adult pubis. suggest that the major bones of adults and subadults were placed in the pit in the same manner, perhaps at the same time. Miscellaneous bones, however, such as ribs, carpals, and tarsals appear to have been handled differently, sug- gesting that they may have been segregated originally. More insight into the specifics of the bone groupings may be gained by examining associations within the adult and subadult groups. In Table 23, chi square values are presented for distributional differences between the adult femora and all other adult bones. The femur was chosen as a standard since it appeared to be the most overall randomly distributed bone, based on the distribution of weight in each unit. As the table indicates, most of the major bones of the skeleton have distributions highly correlated with that of the femora; only the phalanges and carpals have significantly different distributions at the .05 level of significance. Perhaps the most revealing aspect of the table is the order of the bone associations. The skull, mandibles, and first cervical ver- tebrae demonstrate the highest correlation with the randomly scattered femora, followed closely by other major long bones, innominates, scapulae, tali, etc. The bones showing the most disparate distributional patterns are small bones such as the carpals, phalanges, patellae, etc. The order of bone distributional associations closely approximates the order of total bone representation in the ossuary as a whole (Table 17). Since the chi square statistic operates independently of TABLE 22.—Chi square values for distributional associations between adult and subadult bones within Ossuary II Bones Chi square value Degrees of freedom Probability Vertebrae 167.63 68.02 12.07 6. 79 6.76 6.28 4.93 3.95 1.62 1.53 0.75 0.50 0. 13 0. 00 -0. 001 Tarsals and carpals Ribs . 00 - . 001 .00 - .001 Pubis .001- .01 .001- .01 Tibia .01 - .02 Ulna .02 - . 05 Scapula . 02 - . 05 Fibula .20 - . 30 Femur . 20 - . 30 Radius 30 - . 50 Clavicle . 30 - . 50 Temporal. 70 - 80 THE OSSUARIES FROM NANJEMOY CREEK 39 TABLE 23.—Chi square values for distributional associations between the adult femora and other adult bones within Ossuary II Femora vs.— Chi square Degrees of Probability value freedom Carpals 7.84 1 7. 75 ] 3.31 1 3. 14 1 3.09 1 2.81 1 2.75 1 2.52 1 2.43 1 2.05 1.85 1 1.74 1. 74 1.51 1.44 .89 .89 . 77 .73 .73 .57 49 .49 .40 .39 .33 .29 .26 0. 001-0. 01 Foot phalanges .001- .01 Patellae . 05 - . 10 Cervicals . 05 - . 10 Tarsals . 05 - . 10 Hand phalanges . 05 - . 10 Metacarpals . 05 - . 10 Thoracics . 10 - .20 Metatarsals . 10 - . 20 Radius .10 - . 20 Lumbars . 10 - . 20 Coccyx . 10 - . 20 Sternum .10 - .20 Ribs .20 - . 30 Clavicle .20 - . 30 Calcaneus . 30 - . 50 Fibula I . 30 - . 50 Pubis . 30 - . 50 Sacrum I . 30 - . 50 Humerus... . I . 30 - . 50 Talus 1 . 30 - . 50 I . 30 - . 50 Ulna I . 30 - . 50 I . 50 - . 70 Ilium I . 50 - . 70 I .50-70 Mandible I .50-70 Whole skull 1 .50-70 sample size, the data suggest a general correlation of bone representation with bone distribution that has to be ex- plained culturally. Similar bone distribution correlations exist within the sub- adult group. Table 24 gives the ranking of chi square values for distributional associations between the subadult clavicles and other subadult bones. The subadult clavicles were chosen for comparison since they demonstrated a close distributional correlation with the adult clavicles. Like the adult data, the subadult data indicate a strong distributional association among the major bones of the skeleton and a poor association between the clavicles and such miscellaneous bones as tarsals, carpals, metatarsals, metacarpals, vertebrae, and phalanges. Once again, the order of the distributional correlation generally approximates the relative frequency of representation in the ossuary. The data presented above indicate that three spatially distinct, generalized groups can be recognized within the bone concentration in Ossuary II: (1) major bones of both sub- adults and adults; (2) miscellaneous smaller bones of sub- adults, and (3) miscellaneous smaller bones of adults. Al- though these groups are neither spatially isolated nor mutu- ally exclusive, they do present distinct distribution patterns that appear to have cultural significance. TABLE 24.—Chi square values for distributional associations between the subadult clavicles and other subadult bones within Ossuary II Clavicles vs.— Chi square value Degrees of freedom Probability Tarsals and carpals Metatarsals and metacarpals 31.30 25.43 22. 71 6.52 5. 70 5.26 4.69 3.92 2.99 2.60 2.43 2. 18 2. 10 2.04 1. 35 . 54 .25 . 12 . 12 1 0. 000-. . 000- .000-. .01 - .01 - .02 - .02 - .02 - .05 - . 10 - . 10 - . 10 - . 10 - 10 - .20 - .30 - .50 - .70 - . 70 - 001 001 Vertebrae 001 Phalanges (hand and foot) . 02 Pubis . 02 Ulna . 05 Fibula . 05 Humerus . 05 Tibia . 10 Ribs . 20 Maxilla . 20 Femur .20 . 20 Ilium . 20 Scapula . 30 Sternum Temporal .50 7n Radius . 80 Mandible 80 These patterns could have been produced by the follow- ing procedure. The major bones of adults and subadults may have been mixed together at the time of ossuary burial. Since the miscellaneous, smaller bones of both adults and subadults display distribution patterns different from both the major bones and from each other, they must have been segregated prior to ossuary burial. Perhaps die villagers traveled to the place of primary burial (scaffolds or death houses) and gathered together the bones of their dead for reburial. At this time, they may have grouped together the major bones of adults and subadults and transferred them to the ossuary. Then, they may have cleaned out the scaf- folds to prepare them for future use and during this process gathered much of the skeletal material originally missed. These bones may then have been added to the ossuary, giving them a different distribution within the pit from the bones already there. If placement on the scaffolds was originally segregated by age, then these collections of mis- cellaneous bones might retain that segregation in the ossuary. Bodies placed in the scaffold areas may have been segregated according to whether they had gone through "huskanaw." According to Beverley (1705, ch. 6:41) males in the Vir- ginia area went through this rite of passage from boyhood to adulthood at about age 15. Since this pre- and post- huskanaw dichotomy was maintained in many cultural in- stitutions during life there is no reason to believe that it could not have been maintained after death as well. 40 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 Summary In summary, the ossuaries from 14CH89 appear to repre- sent the relatively complete remains of nearly all the indi- viduals who died in the population contributing to the os- suaries during two culturally prescribed periods of time during the Late Woodland. Associated cultural remains in- dicate that the two ossuaries were nearly contiguous in time and probably date from the 16th century A.D., just before European contact in the area. The ossuaries are very similar to those described for the Huron, the only major differences being the absence of large numbers of mortuary offerings and of a platform. Distributional features of skeletal material within Ossuary II suggest some segregation of bones prior to ossuary deposit, perhaps reflecting either the differential selection of bones for ossuary burial or an initial segregation in the scaffold area or death houses, or both. Although not all bones were equally represented in the ossuaries, the presence of burned bones and partially articulated skeletons suggests an attempt was made to include remains of all individuals who had died since the last burial ceremony, regardless of the manner of initial mortuary treatment or the stage of preservation and decomposition of the body. The ossuaries from 14CH89 represent nearly complete cross- sectional skeletal samples and offer a unique opportunity to examine population profiles for the populations represented. Determination of Sex and Age at Death Demographic reconstruction from skeletal populations de- mands an accurate assessment of sex and age at death. Be- cause the reliability of interferences about longevity and life expectancy is directly related to the accuracy of age assess- ments, it is vital to select those criteria tiiat produce the most accurate estimates under the circumstances. Several methods are here employed to generate the basic information needed for demographic reconstrutcion. The results of each method are then evaluated to identify the most accurate age and sex estimates for the overall populations. DETERMINATION OF SEX Estimation of sex is not attempted for subadults in the ossuary remains because none of the previous studies (e.g., E. L. Reynolds, 1945, 1947; A. Thompson, 1899; Boucher, 1955, 1957) provides sufficiently accurate methods for esti- mating subadult sex from fragmentary skeletal material. One of the most encouraging efforts is that of Hunt and Gleiser (1955), who based their sex criterion on the as- sumption that sexual dimorphism is more clearly marked in skeletal maturation than in dental eruption. By comparing the degree of skeletal maturation with the stage of dental eruption, they were able to identify sex with an accuracy of 73 percent at 2 years, 76 percent at five years, and 81 percent at eight years. Bailit and Hunt (1964) estimated subadult sex by observing the stage of canine eruption relative to premolar eruption. Their predictions were only 58 percent accurate and they concluded that sexual dimorphism in eruption sequence is not sufficiently marked to allow ac- curate sex estimates to be made. Garn, Lewis, and Kerewsky (1964) reached the same conclusion about tooth size. Of all methods of determining subadult sex, only that of Hunt and Gleiser (1955) for older children approaches the accuracy necessary for meaningful demographic analysis. Unfortu- nately this method has little application in an ossuary situa- tion, where subadults are represented by isolated bone fragments and both skeletal and dental information are not available for the same individual. Sex determination of adult skeletal material is of course much more reliable, with the pelvis providing the best data. Sex differences in the pelvis have been studied by many in- vestigators and their conclusions are available in several good summary articles (Krogman, 1962; Stewart, 1968). In general, female pelves are distinguished by the occur- rence of a quadrangular-shaped pubis body, broader sub- pubic arch, markedly everted ischio-pubic ramus, low sym- physis, small obturator foramen, small acetabulum, wide and shallow sciatic notch, lower and more flaring ilium, smaller sacro-iliac articulation, high frequency of a deep pre-auricu- lar sulcus, and general gracility. In addition, Phenice (1969) has developed a method that utilizes three areas of the pubis: the ventral arc, subpubic concavity, and medial as- pect of the ischio-pubic ramus, to estimate sex with 96 percent accuracy. Although metric methods are available for the pelvis (Washburn's 1948 ischium-pubic index for ex- ample), none is more accurate than the Phenice nonmetric method. General sex differences in the skull also are well defined and have been summarized by a number of authors (Krog- man, 1962; Stewart, 1968). Utilizing a subjective assess- ment of cranial morphology alone, however, an experienced investigator can determine sex correctly in only 80—90 per- cent of the cases (Krogman, 1962; Stewart, 1968). Slightly better accuracy has been reported for more objective meth- ods using discriminant functions (Giles and Elliot, 1963; Giles, 1964, 1970a,b). So far as other bones are concerned, sexual differences are well documented but not as helpful as the foregoing. In par- ticular, femoral head diameter is regarded as a fair sex in- dicator (Dwight, 1904; Maltby, 1917-1918; Parsons, 1913- 1914, 1914-1915). Femoral head diameters of 43 mm or smaller are usually female and those 46 mm or larger are male. The overlap, however, between male and female dia- meter-ranges and their variation among different popula- tions reduce the accuracy of identifications. The application of the above criteria to the skeletal re- mains from the Juhle ossuaries generally shows that males and females are about equally represented in Ossuary I, but females probably outnumber males in Ossuary II. Table 25 summarizes the sex distribution in Ossuary I as determined from the skull, pubis, ischium, and ilium. Each bone shows approximately an equal representation of males and females. 41 42 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 8 TABLE 25.—Sex representation of adults in Ossuary I TABLE 26.—Sex representation of adults in Ossuary II Bone Male Female No. % Total Percent of No. % adults Skull 31 23 30 32 55 51 54 52 25 22 26 29 45 49 46 48 56 45 56 61 81 Pubis 65 81 Ilium. . 88 Although slightly more males than females appear to be rep- resented, the fact that between 12 to 35 percent of the total adult sample are missing from each bone category suggests that the slight differences have little meaning. The figures serve mainly to document that both sexes occur in approxi- mately equal numbers. In contrast, skeletal evidence from Ossuary II suggests that more females may be present than males. Table 26 sum- marizes the sex representation in Ossuary II as deduced from the skull, pubis, and ilium. All three criteria suggest more females are present than males; however, only 61 and 82 percent of the total adult ossuary population are represented by pubes and ilia, respectively. If the missing 39 and 18 per- cent are predominantly males, then the representation of males and females would be approximately equal. Only 10 percent of the crania are missing; however, if as many as 20 percent of the crania present have been incorrectly estimated to be females, then again equal sex representation would be indicated. More insight into male-female representation in the two ossuaries may be gained by examining femoral head diam- eters. Figure 17 shows the frequency distribution of head diameters of left femora from Ossuary I, and right femora from Ossuary II, after the raw data have been grouped into 2-mm units and converted to percentages of the total. The distribution of head diameters for Ossuary I closely fits the model of equal male-female representation indicated by other criteria. An equal representation would suggest those femoral heads 43 mm and smaller are predominantly female, Bone Male Female No. % Total Percent of No. % adults Skull 36 24 29 40 40 36 53 36 52 60 60 64 89 60 81 90 Pubis 61 Ilium 82 those 46 mm and larger are predominantly male, with the two ranges strongly overlapping between 44 mm and 45 mm. This distribution generally agrees with the variation docu- mented for other populations (Krogman, 1962). The distribution of femoral head diameters from Ossuary II is somewhat different from that of Ossuary I. The shape of the frequency polygon in Figure 17 suggests a higher per- centage of smaller head diameters in Ossuary II than in Ossuary I. Of course, whereas only 2 percent of the adult femora are missing in Ossuary I, 18 percent are absent from Ossuary II. However, even if most of the missing femora are males, not all of the difference exhibited in Figure 17 can be accounted for. The evidence suggests either that males of Ossuary II had generally smaller femoral head diameters than males of Ossuary I, or that more females than males are represented in Ossuary II. The latter explanation is much more probable, since a higher frequency of females is sug- gested independently by data from the crania, ilia, and pubes. A higher frequency of females in Ossuary II would indicate either a higher female death rate or that some of the deceased males were not included in the ossuary. The missing males could have been leaders that were deposited perma- nently in death houses or, more probably, males that either died away from their village, or for some reason were treated differently after death. Only the greater frequencies of female-appearing skulls, pubes, ilia, and femoral heads can be positively documented. Explanations of these frequencies must remain speculations until further evidence can be obtained. Whatever the explanation it is doubtful that the possible slight shortage of males in Ossuary II significantly affects the demographic calculations. DETERMINATION OF AGE OF SUBADULTS Subadult age at death may be determined from any one or a combination of the following criteria: long-bone size, epiphyseal union or nonunion, dental calcification, or dental eruption. When possible, it is best to examine all of them. Since complete skeletons usually cannot be assembled from ossuaries, however, techniques for aging must be employed independently. The present analysis estimates subadult age by using three criteria: long-bone growth and maturation, dental eruption, and dental calcification. The resulting dis- tributions of ages at death are then compared and evaluated. Length of Long Bones Few good standards exist for estimating age from individ- ual bone length. Stewart (1968) published data on the size of Eskimo femora from birth to 18 years, developing regression lines to express the correlation between femoral length and chronological age estimated from tooth eruption. His analysis is confined to a small sample of femora, but represents the only cross-sectional growth study of aboriginal American skeletal populations that records growth from birth to 18 DETERMINATION OF SEX AND AGE AT DEATH 43 OSSUARY I n=67 OSSUARY U. n = 87 UJ 21 TAG z ia UJ o cr UJ Q. 15 37-38 39-40 41-42 43-44 45-46 47-48 49-50 FEMORAL HEAD DIAMETER (MM) 51-52 FIGURE 17.—Comparison of femoral head diameters from Ossuary I with those from Ossuary II. years. Johnston (1962) published data on the growth of long bones in archaic skeletal populations from Indian Knoll, Kentucky, but his study only extends from birth to 5.5 years. Numerous cross-sectional growth studies have been made on the living (Anderson and Green, 1948; Maresh, 1955). They are of limited value for skeletal analyses of archeologi- cal populations for two reasons: (1) they are based on living populations which may have different growth patterns from those of earlier populations, and (2) they utilize radiographic measurements instead of measurements of dry bone. Conse- quently, Stewart's (1968) Eskimo data would appear to be the most appropriate for estimating subadult age from iso- lated femora. Epiphyseal Union For later subadult years (after about 17 years), standards are available for estimating age from epiphyseal union. Through their study of 417 Korean War dead, McKern and Stewart (1957) documented age of epiphyseal union for young American males. Their study presents excellent data on the variability of epiphyseal closure, but lacks information about the early ages of closure since the Army sample did not extend below the age of 17. Stevenson (1924) and Krogman (1939, 1955) have pub- lished data on the age of epiphyseal union in both sexes. Both investigators examined dissecting-room skeletal samples which were weighted toward older individuals and repre- sented predominantly the lower economic class from regional areas. Krogman (1939) presented ranges or "central tend- encies" for the postnatal union of ossification centers. How- ever, Stewart (1934) and McKern and Stewart (1957) emphasize that more variability occurs than Krogman and Stevenson indicated. Tables 27 and 28 present the total maximum lengths of the subadult femora recovered from Ossuaries I and II. The "bone numbers" of Ossuary I are field numbers recorded on individual bones and refer to their locations relative to nearby skulls. All measurements were taken on bones that lack proximal and distal epiphyses, using a sliding caliper or osteometric board (depending on the size of the bone) and were recorded to the closest millimeter. Total length was estimated for fragmentary femora by comparing diem to unbroken bones of about the same size. In addition to those femora listed from Ossuary II, 9 left and 6 right occurred that displayed epiphyseal fusion at the greater and lesser trochanters and femoral heads, but nonunion of the distal 44 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 8 TABLE 27.—Maximum diaphyseal length (in mm) of subadult TABLE 28.—Maximum diaphyseal length (in mm) of subadult femora from Ossuary I, arranged in increasing order of length femora from Ossuary II, arranged in increasing order of length Bone no. Length Bone no. Length Bone no. Length Square Length Square Length Square Length Square Length RIGHT FEMORA RIGHT FEMORA 51 67 126-112 106 42 193 DC 53 GC 16 FLl 131 BRl 241 37 67 15 108 43 193 DR1 60 FLl 78 EC 135 DR1 248 103A 68 39 117 9-10 209 CRl 63 CRl 78 AL1- ■BL1 135 BL1-CL1 276 102-125 70 NB128 118 37 209 DR1 64 CRl 79 FR1 150 FC 287 W-117-119 71 42 119 119 225 DLl 65 EC 80 GL1 165 DR1 287 9-10 72 26 120 41 252 DR1 65 FLl 80 CLI 175 BR1-CR1 300 W-117-119 72 52-54 122 52 258 GL1 67 ER 82 FLl 175 ARl 313 115 74 54 125 42 270 D 1 68 AL1 85 FLl 175 BRl 320 9-10 75 42 130 1 275 FC 68 AL1- BL1 86 BRl 178 EC 330 W-117-119 75 101-120 130 3 277 BL1 68 DC 87 GC 181 DR1-ER1 330 p 75 N189 140 NE126 280 FC 72 ERl 90 FR1 190 FLl 332 102-125 77 39 140 118 280 BR1-CR1 72 DR1- -ERl 100 FC 200 DR1 342 W-117-119 79 5 145 44 310 BL1 73 FC 100 DLl 202 GL1 364 4-6 80 54 148 19 315 DR1 73 FC 100 FLl 205 BRl 387 122-124 83 42 150 4-6 323 EC 73 DLl 103 CRl 212 DLl 390 103 84 26 161 23 330 AL1-BL1 74 EC 105 EC 213 DC 412 19 95 4-6 175 B4-112-113 331 DLl 75 AL1 105 CRl 221 CRl 413 41 100 104 182 14 333 ERl 75 EC 120 GR1 222 ELI 423 30 101 464- 187 W102 39 378 380 FLl 76 ARl 122 GC 236 GR1 429 LEFT FEMORA LEFT FEMORA NE126 46 26 97 126-112 200 DC 53 DC 77 BR1- CRl 125 EC 208 ? 50 19 98 4-6 200 DR1 55 CRl 80 FL1 129 FLl 213 ? 62 9-15 106 9-10 211 BL1 57 FLl 80 DL1- ■ELI 131 EC 220 102-125 68 101-120 113 49 235 CRl 60 EC 80 DR1 132 GR1 223 102-125 68 52-54 114 41 256 CRl 62 DC 82 AL1- BL1 137 CRl 223 9-10 73 42 115 52 261 DR1 63 FLl 82 CR1 140 BRl 234 W-117-119 75 42 119 2-17 275 DR1 64 FLl 83 EC 140 DR1 239 9-10 75 26 120 118 278 DR1 65 AL1 85 GC 144 BL1 267 NE126 75 3, 100 143 NE126 281 FR1 65 BL1 87 DR1- -BRl 150 FC 287 26 76 42 144 2-17 310 GR1 66 FC 89 EC 150 DR1 288 W-117-119 77 117-8-9 146 44 313 GL1 67 EC 90 CRl 170 BRl-CRl 298 118-119 81 54 147 B+112-113 320 FLl 67 ER 90 DR 171 BRl 308 NE126 81 5 152 4-6 324 EC 70 CRl 93 DLl 175 BRl 327 42 86 126-112 180 23 330 FLl 71 DR1 96 CLI 175 FR1 331 9-10 92 464- 181 4-5, 14 345 EC 72 EC 96 FLl 175 DC 333 9-15 92 104 183 39 380 AL1-BL1 73 DLl 98 FLl 178 GL1 362 126-112 95 1 185 W-102 385 BR1-CR1 75 DR1 100 GR1 182 DLl 395 ERl FLl 75 75 EC ARl 100 122 DLl CRl 183 190 BRl CRl 405 416 epiphyses. Krogman (1962) suggests that the femoral distal GC 77 FLl 122 BC 202 GR1 426 females and 16 to 18 years in males, although McKern and Stewart (1957) indicate the male range extends up to 20 years. Table 29 compares femora from Ossuary I with those of Ossuary II after they have been grouped into 5-year age categories based on maximum femoral length and the stages of epiphyseal union. Age categories for femoral lengths are based on Stewart's (1968) Eskimo data, and the ages of epiphyseal union on data presented by Krogman (1962) and McKern and Stewart (1957). Actually all ages are de- rived from maximum lengths except the nine left and six right femora from Ossuary II already mentioned. All of these were aged between 15.0-19.9 years. As Table 28 re- veals, the right femora from Ossuary I and the left femora from Ossuary II produced the maximum femoral counts, and consequently provide the most reliable indicators of sub- adult age. Most of the age differences between left and right femora occur in the 0 to 4.9 year category, probably since smaller bones of that category could easily be lost prior to ossuary burial. Dental Eruption and Calcification Most investigators agree that dental eruption and calcifi- cation probably constitute the most accurate indicators of subadult chronological age. Garn, Lewis, and Polacheck DETERMINATION OF SEX AND AGE AT DEATH 45 TABLE 29.—Age distribution (in years) of subadult femora from Ossuaries I and II as determined from maximum diaphyseal length l and epiphyseal union 0-4 9 5.0- 9. 9 10. 0-14. 9 15.0- -19.9 Total Side No. % No. % No. % No. % OSSUARY I Left 30 59 13 25 6 12 2 4 51 Right 36 62 14 24 6 10 OSSUARY II 2 4 58 Left 56 63 12 13 7 8 14 16 89 Right 47 57 13 16 9 11 13 16 82 'Using Stewart's (1968) generalized growth curve of the Eskimo femora: < 178 mm = H,9 years; 179-263 mm=5.0-9.9 years; 264-357 mm= 10.0-14.9 years; >358 mm= 15.0-19.9 years. (1959), and Lewis and Garn (1960) have demonstrated that dental development is more closely associated with chronological age than is skeletal development. Dental de- velopment has been shown to have a strong genetic com- ponent (Glasstone, 1938, 1963, 1964) and is little influenced by nutritional or other environmental factors (Paynter and Grainger, 1961, 1962). Although specific diseases, such as hypopituitarism and syphilis can modify dental development (Bauer, 1944), most major diseases minimally affect teeth even though skeletal parts may be greatly altered (Nis- wander and Sujaku, 1965). Endocrine disorders and other maturational malfunctions affect teeth only one-fourth as much as the skeleton (Garn, Lewis, and Blizzard, 1965). Overall, general dental development appears to provide the closest correlation with subadult chronological age. In the past, physical anthropologists have utilized exten- sively an assessment of the degree of dental eruption and calcification to determine age at death of subadults. For the actual age correlations, most investigators have relied heavily upon a chart compiled by Schour and Massler (1941). The chart draws upon the observations of Logan and Kronfeld (1933), and Kronfeld (1935) on the teeth of 30 infants and children under 15 years of age (Kraus, 1959). It has been criticized by Garn, Lewis, and Polacheck (1959), Miles (1963), and recently Sundick (1972), since it is based on a small sample and many of the subjects examined suffered from illnesses that may have affected the time of tooth eruption. In addition, the chart combines data on eruption and calcification. Garn, Lewis, and Polacheck (1959), and Lewis and Garn (1960) have demonstrated that tooth for- mation and calcification have a higher correlation with age than does tooth erupdon. Tooth calcification is less affected by premature loss of deciduous and permanent teeth, crowd- ing, and dietary factors, than is the timing of tooth eruption (Fanning, 1961, 1962a, 1962b; Niswander and Sujaku, 1960, 1964). In place of the Schour and Massler (1941) chart, Hunt and Gleiser (1955) recommend the use of data provided by Robinow, Richards, and Anderson (1942) and by Meredith (1946) for the age of eruption of the deciduous dendtion and by Hurme (1948) for the age of permanent eruption. These studies soundly document eruption for North American Whites. Studies by Steggerda and Hill (1942), Hurme (1948), Garn and Moorrees (1951), Dahlberg and Mene- gaz-Bock (1958), and Hrdlicka (1908) all suggest that erup- tion of at least the posterior teeth occurs slightly earlier in Indian and other non-White groups. Unfortunately the stud- ies do not agree on the actual eruption ages for various teeth, and it is not clear from the discussion whether the variance is due to population genetic variation, dietary-nutritional differences, or methodological discrepancies. Of all studies of dental calcifications, those of Moorrees, Fanning, and Hunt (1963a, 1963b) supply standards most suitable for aging immature dentitions from archeological samples. They present two standards of tooth formation; one for three deciduous teeth, and one for ten permanent teeth. Their study is based on a large sample (n = 236) and provides both mean age and standard deviadon values for exactly defined stages of development. Although they pub- lished separate standards for males and females, their results can be pooled for assessment of archeological specimens where sex cannot be determined with confidence. All subadult dentitions from both ossuaries were aged first by a comparison with the Schour and Massler (1941) dental eruption chart and then by the calicification stand- ards of Moorrees, Fanning, and Hunt (1963a, 1963b). This approach not only provided an accurate assessment of sub- adult age at death, but permitted comparison of the stand- ards as well. Table 30 presents the results of that assessment for Ossuary I. The age categories in Table 30 represent convenient categories to judge dental eruption from the Schour and Massler chart. Numerical age values determined from the calification standard could easily be placed into those categories for comparison. As Table 30 reveals, the eruption standard gave consistently higher age estimates than the calcification standard. The differences between the two aging methods are presented more clearly in Figure 18 for the right maxillae and mandibles, and Figure 19 for the left maxillae and mandibles. In these figures, the number in each category has been multiplied by a normalizing factor so that sample size (n = 62) is the same for each curve. The curves were normalized to a sample size of 62 since that is the minimum number of subadults represented in Ossuary I. The actual sample sizes are keyed to each particular curve in the figures. Both figures also demonstrate that ages calculated from dental eruption are consistently higher than those calculated from dental calcification. Consequently, the calcification standard produces more age estimates for the earlier years and fewer for the later years in both the maxilla and mandible. A comparison of the two figures 46 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 TABLE 30.—Age distribution in years of subadults in Ossuary I as determined from crown-root calcification and tooth eruption % % % Tooth group No. 0-1.9 % 2.0-5.9 No. 6.0-11.9 No. 12.0-17.9 No. 18.0-19.0 No. Total Right mandible Left mandible. Right maxilla.. Left maxilla. . . Right mandible Left mandible. Right maxilla.. Left maxilla. . . CALCIFICATION 18 41 15 34 8 18 3 22 46 15 31 9 19 2 13 40 8 24 10 30 2 15 41 9 24 11 30 ERUPTION 2 16 36 14 32 10 23 3 20 42 14 29 11 23 2 10 30 10 30 10 30 2 11 30 11 30 12 32 2 0 44 0 48 0 33 0 37 2 44 2 48 3 33 3 37 RIGHT MAXILLA n= 33 RIGHT MANDIBLE n = 44 RIGHT MAXILLA n=33 RIGHT MANDIBLE n = 44 DENTAL CALCIFICATION GO UJ _l Q. < x CC UJ CD 28 27 26 25 24 23 22 21 20 - 19 - 18 - 17 - 16 - 15 14 13 12 II 10 9 8 7 6 5 4 3 2 I I- B- I 6-11 J L 12-17 18-19 AGE IN YEARS FIGURE 18.—Comparison of age distributions determined fom dental eruption and dental calci- fication in right mandibles and maxillae from Ossuary I. All normalized to n=62. DETERMINATION OF SEX AND AGE AT DEATH 47 TABLE 31.—Age distribution in years of subadults in Ossuary II as determined from crown-root calcification and tooth eruption % % % Tooth group No. 0-1.9 % 2.0-5.9 No. 6.0-11.9 No. 12.0-17.9 No. 18.0-19.0 No. Total Right mandible Left mandible. . Right maxilla. . Left maxilla. . . Right mandible Left mandible. Right maxilla. . Left maxilla. . . ( CALCIFICATION 24 45 15 28 6 12 8 15 0 0 53 27 49 14 25 7 13 7 13 0 0 55 15 29 18 35 9 17 10 19 0 0 52 17 33 15 29 9 17 ERUPTION 11 21 0 0 52 17 32 21 40 7 13 8 15 0 0 53 22 40 18 33 9 16 6 11 0 0 55 14 27 19 37 8 15 10 19 1 2 52 15 29 16 31 9 17 11 21 1 2 52 DENTAL ERUPTI DENTAL CALCI MANDI 28 r- 27 - 26 25 - 24 \- 23 22 21 20 19 18 17 16 15 14 13 12 II 10 9 8 7 6 5 4 3 2 I e-1 J L 12-17 18-19 AGE IN YEARS FIGURE 19.—Comparison of age distributions determined from dental eruption and dental calcifi- cation in left mandibles and maxillae from Ossuary I. All normalized to n=62. 48 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER I8 shows few differences between the left and right sides, and confirms the differences described above. In both figures, estimates from the mandible are lower than those from the maxilla. Differences between the maxillae and mandibles are largely a product of sampling, since on both the left and right sides, 11 more mandibles were present than maxillae. Furthermore, it is apparent from the shape of the curve that most of the missing maxillae were from individuals in the birth to 1.9-year and 2.0 to 5.9-year categories. These data suggest that the eruption standard of Schour and Massler (1941) produces consistently higher age esti- mates than the calcification standard of Moorrees, Fanning, and Hunt (1963a, 1963b). Although the calcification stand- ard probably is the more accurate, the actual ages at death may be even slightly earlier for this non-White archeological population. Of all curves presented in Figures 18 and 19, that obtained from the calcification of teeth from the left mandible is probably the most reliable since it is based on the more accurate standard (calcification) and largest sample (n = 48). Table 31 compares calcification and eruption estimates from Ossuary II. As in Ossuary I, calcification age deter- minations are generally lower than those obtained from dental eruption. Figures 20 and 21 compare ages at death as determined by eruption and calcification from the maxilla and mandible, left and right sides. Again, there is close agreement between age assessments from the two sides. As expected, the two main causes of age discrepancies are the bones selected for examination (maxilla or mandible) and the age standard utilized. The standards show close agree- ment on the maxillary teeth, with the calcification standard producing only slightly younger estimates. In contrast, the aging standards produced widely divergent results when applied to the left and right mandibular teeth. Estimates 46 44 42 40 38 36 34 in 32 UJ _i 30 0. 2 28 < X 26 UJ 24 u. 22 o 20 CC UJ 18 CD 2 16 3 14 12 10 8 6 4 2 _ 1 1 1 1 1 1 1 -^ „ -* \ \^ ^^ X. \ ^x- •**""• X. \ X "*^\> - • • * %*S^ X *«N\ 1 1 1 1 1 1 X *X>s 1 1 1 1 1 1 1 1 —— RIGHT MANDIBLE n=53 RIGHT MAXILLA n = 52 RIGHT MAXILLA n = 52 RIGHT MANDIBLE n = 53 } ENTAL CALCIFICATION DENTAL ERUPTION AGE IN YEARS FIGURE 20.—Comparison of age distributions determined from dental eruption and dental calcifi- cation in right mandibles and maxillae from Ossuary II. All normalized to n=89. 12-17 DETERMINATION OF SEX AND AGE AT DEATH 49 from the maxilla are higher than those from the mandible, but the discrepancy is not as large as in Ossuary I. Of these curves, that of the left mandible aged by the calcification standard is probably the most reliable since it is based on the largest sample (n = 55) and the more accurate standard (Moorrees, Fanning, and Hunt, 1963a, 1963b). The above discussion not only provides a basis for the accurate assessment of subadult age for the Juhle ossuaries, but also documents the great variation in results obtained from different methods and the hazards inherent in the arbitrary selection of a single criterion. In Ossuary I, a death curve based upon dental eruption of the maxillae would induce quite different demographic considerations from a curve based on mandibular tooth formation. A "students t-test" for differences between the mean ages estimated by the Schour and Massler chart and the mean ages derived from the standards of Moorrees, Fanning, and Hunt reveals a significant difference at the .02 level. These results not only demonstrate the error involved in arbitrary age esti- mation, but stress the need for investigators to utilize the best standards available and to specify which standards and which teeth were employed in the analysis. Figures 22 and 23 compare age distributions calculated from the most representative femoral length data and the most representative dental eruption and calcification data. The age categories were selected to allow meaningful com- parisons with dental eruption. In Ossuary I, all three curves are in close agreement, although the calcification curve in- dicates slightly more younger individuals are present than are suggested by the other two curves. In Ossuary II, the curves for femoral length and tooth calcification agree closely, but the curve based on dental eruption differs mark- n=55 n = 52 n = 52 LEFT MANDIBLE LEFT MAXILLA LEFT MAXILLA DENTAL CALCIFICATION DENTAL ERUPTION LEFT MANDIBLE n=55 44 42 40 38 36 34 -* ?? o 20 CC UJ 18 CO 2 16 z> z 14 12 10 8 6 4 2 6-il 12-17 18-19 525-413 O - 74 - 5 AGE IN YEARS FIGURE 21.—Comparison of age distributions determined from dental eruption and dental calcifi- cation in left mandibles and maxillae from Ossuary II. All normalized to n=89. 50 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 CO UJ _l Q. 2 < x UJ CC UJ CO 2 3 z 28 — ^k 27 26 25 - 24 _ >v ' '. x. 23 o«. \ 22 x^. \ 21 \._ >< 20 - X* >. 19 - 18 - 17 - ^s^^k.. 16 - * • . 15 ^^^^ ""'*•. 14 >^V 13 ^^X^X '*■ 12 ^v^^XJ- II n=58 X. 1 0 * m X. 9 (DENTAL ERUPTION) >v '•.\^ 8 LEFT MANDIBLE n=48 (DENTAL CALCIFICATION) Xv \ ^x. 7 x/\ X. 6 \ ', \ 5 x^* > \ 4 ^^/ . \ 3 2 1 1 1 1 1 1 2-5 6-11 12-17 18-19 AGE IN YEARS FIGURE 22.—Comparison of age distribution determined from (1) maximum length and epiphyseal union of right femora, (2) dental eruption in left mandibles, and (3) dental calcifi- cation in left mandibles from Ossuary I. All normalized to n=62. edly. Of the three curves, certainly those based on femoral length and tooth calcification are the most accurate. Any evaluation of the relative accuracy of the femoral and dental calcification curves must consider both the ac- curacy of the standards and the relative size of the samples. Stewart's study of Eskimo femora was based on ages deter- mined from dental eruption. Consequently, the data not only incorporate the error of correlation between femoral length and stage of dental eruption, but the additional error of the original age assessment from dental eruption. Even though these two independent sources of error may not be additive, they do indicate that age estimates based on femoral length are less reliable than those obtained directly from the dentitions. Even though standards of aging femora may be less accurate, however, the femora are much better repre- sented than mandibles in the ossuaries. In Ossuary I, only four subadults are not represented by femora, while in Ossuary II, the minimum subadult count comes from the femora. In contrast, 14 left mandibles are missing from Ossuary I, and 36 right mandibles are missing from Ossuary II. Although the close agreement of femoral and dental curves suggests that the loss of mandibles was approximately equal for each age category, it appears best to utilize the femoral data for further demographic computations since femoral representation in the ossuaries approaches 100 percent. DETERMINATION OF SEX AND AGE AT DEATH 51 LEFT FEMORAL LENGTH n=89 RIGHT MANDIBLE n = 53 (DENTAL ERUPTION) RIGHT MANDIBLE n = 53 (DENTAL CALCIFICATION) 38 CO 36 UJ -I 34 Q. 2 32 < X 30 UJ 28 u. ?6 o 24 (T UJ 22 CD 2 20 ") z 18 16 14 12 10 8 6 4 2 B- l 6-11 AGE IN YEARS FIGURE 23.—Comparison of age distributions determined from (1) maximum length and epiphyseal union of left femora, (2) dental eruption in right mandibles, and (3) dental calcifi- cation in right mandibles from Ossuary II. All normalized to n=89. DETERMINATION OF AGE OF ADULTS Adult age at death may be determined in a number of ways. As with subadults, it is best to consider as many criteria as possible from different parts of the skeleton to obtain an overall assessment of skeletal maturity and chrono- logical age. Since few complete skeletons can be assembled from ossuary deposits, however, aging methods applicable to each bone category must be considered independently. In the past, physical anthropologists relied heavily upon the degree of cranial suture closure and dental attrition for age determination. Age progression in the closure of cranial sutures was described in detail by Todd and Lyon (1924- 1925), who noted that endocranial closure progresses more regularly than ectocranial closure, yet the progression is not regular enough for aging purposes. This has been emphasized by Singer (1953) and verified by McKern and Stewart (1957) for North American males. Today suture closure is used in determining age only as a last resort. The degree of dental attrition is used occasionally for age determination, largely because teeth are frequently all that remains of some poorly preserved skeletons. Hrdlicka (1952) related general age categories with stages of dental attrition in American Indians, but Miles (1963) and Moor- rees (1957) have demonstrated that the correlation of age with attrition varies considerably within and between popu- lations. Although estimates may be improved by considering the diet of the population and the rate of attrition within each dentition, the method still can provide only very general age estimates. Kerley (1970) has discussed the ossification of rib cartil- age at the costo-chrondral junction as a general aging cri- terion. With increasing age, the sternal ends of the ribs gradually transform from a billowed appearance to an ir- regular, often ragged appearance. With advancing age, the ossification extends further out into the cartilage, eventually culminating in complete ossification of the costal cartilage. McKern and Stewart (1957) documented the variability of this feature, showing that the age changes are too erratic to yield reliable age estimates. Stewart (1958b) described osteophytic outgrowths of ver- tebral centra as general age indicators. Again, individual 52 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 variability is so great that the observation can only be ap- plied very generally. Well-developed osteophytes seldom ap- pear at many sites prior to age 40. Other age changes in bone that yield only very general age determinations are parietal porosity (Todd and Lyon, 1924— 1925; Kerley, 1970), lipping of diarthrodial joint margins (Johnson, 1962), and gross resorption of cancellous tissue within long bones (Hansen, 1953-1954; Schranz, 1959). At best, these methods allow age determinations to be made confidently, only within a decade of the real age. Conse- quently, they do not provide the accuracy needed for demo- graphic analysis. In 1950, Gustafson published a method of determining age at death from seven features of dental microstructure linked with age changes: cementum apposition, attrition, perio- dontosis, secondary dentin apposition, root resorption, trans- parency of the root, and closure of the root orifice. When considered together, these features yield an age estimate con- taining a standard error of only 3.6 years. The standard curve was determined from a series of 41 teeth taken from individuals of known ages that ranged from 11 to 69 years. Numerous modifications have been made on the Gustaf- son method since its formulation in 1950. Nalbandian (1959) and Nalbandian and Sognnaes (1960) reexamined the age correlations and established similar regression coef- ficients from Swedish and United States Caucasian samples. Dechaume, Derobert, and Reyen (1960) found that experi- ence was necessary to obtain reliable results, and eventually duplicated Gustafson's findings. All investigators agree that Gustafson's method is accurate when employed by an ex- perienced histologist with the proper sectioning equipment. Although Gustafson's method is the most reliable of those discussed so far, it was not employed in this study for two reasons: first, many teeth could not be associated with their mandibles or maxillae and second, it was not desirable to destroy adult teeth until they had been studied thoroughly. For these reasons, and because of the lack of reliability of the other methods mentioned, determinations of adult age were made only from the articular surfaces of the pubic symphyses and the anterior cortices of the femora. Pubic Symphysis In 1920, Todd defined 10 phases of age-related pubic symphyseal metamorphosis from his examination of 306 adult skeletons of known age (predominately old) from the Todd Collection. A year later (1921) Todd stated that Negro and White pubes showed essentially the same age changes. He noted also that the stages of change in females are basically the same as in males, but proceed more slowly. McKern and Stewart (1957) provided a new system for aging the male pubis as a result of an analysis of 450 youth- ful skeletons recovered in "Operation Glory" from the Korean conflict. Their system involves the examination of three "components" of the pubic symphysis: the dorsal pla- teau, ventral rampart, and symphyseal rim. This has the advantages of (1) being based on a well-defined population of one sex, (2) being easy to use, and (3) allowing different aspects of the symphyseal face to be assessed independently. Gilbert's (1971) study of female pubes now makes the method available for both sexes. Both studies indicate that symphyseal age changes are more variable than Todd's work had indicated. However, the McKern and Stewart's (1957) method still represents a highly accurate, objective method of age determination, especially since plastic models of each stage of symphyseal change are now available for both sexes. These three-dimensional plastic models allow objective com- parisons to be made and thereby reduce subjective error. The application of the McKern and Stewart (1957) and Gilbert (1971) aging methods to adult pubes from Ossuary I and II produced the distributions of ages recorded in Tables 32 and 33. More male than female pubes were pres- ent in Ossuary I, while more female than male pubes oc- curred in Ossuary II. In both ossuaries, more ages could be TABLE 32.—Age distribution (in years) of adults as determined from pubes in Ossuary I 40-44 30-34 No. % % Side 20-24 % No. 25-29 % No. 35-39 % No. % No. 45-49 No. No. 50+ % Total Average no. age MALE Right Left Right Left Total Right Total Left.. 2 10 3 15 7 35 5 25 1 5 1 5 1 5 20 34 3 12 4 15 10 38 6 23 3 FEMALE 12 0 0 0 0 26 33 0 0 1 6 7 44 3 19 3 19 2 12 0 0 16 36 0 0 0 0 5 46 3 27 2 18 1 9 0 0 11 37 SEXES COMBINED 2 6 4 11 14 39 8 22 4 11 3 8 1 3 36 35 3 8 4 11 15 41 9 24 5 14 1 3 0 0 37 34 DETERMINATION OF SEX AND AGE AT DEATH TABLE 33.—Age distribution (in years) of adults as determined from pubes in Ossuary II 53 40-44 20-24 30-34 No. No. No. % Side 25-29 % No. % 35-39 % No. % No. 50+ 45-49 % No. % Total Average no. age MALE Right Left Right Left Total Right Total Left. . 0 0 7 33 9 43 4 19 1 5 0 0 0 0 21 32 4 16 6 24 11 44 2 8 2 FEMALE 8 0 0 0 0 25 30 4 13 8 27 10 33 7 23 0 0 0 0 1 3 30 31 4 11 9 26 13 37 9 26 0 0 0 0 0 0 35 31 SEXES COMBINED 4 8 15 29 19 37 11 22 1 2 0 0 1 2 51 31 8 13 15 25 24 40 11 18 2 3 0 0 0 0 60 31 determined from left pubes than rights. Age estimates from the two sides are in general agreement, with minor discrep- ancies reflecting both the error of the aging method and the unequal sample sizes. The average age at death was higher for Ossuary I than Ossuary II, slightly higher for females than males in Ossuary I, but equal for both sexes in Ossuary II. Internal Bone Remodeling In 1965, Ellis R. Kerley published a method of determining age from the microscopic examination of cortical bone that advanced skeletal aging methodology to a new level of ac- curacy. His method is based on the fact that during the years of rapid long bone growth (0 to 17) most of the diaphyseal cortex consists of circumferential lamellar bone deposited by the periosteum. Beginning about the fourth year and con- tinuing throughout the life of the individual, this lamellar bone constantly is being remodeled. The remodeling process involves osteoclasts destroying the lamellar bone to form re- sorption spaces. These spaces are then gradually filled in by material generated by osteoblasts to produce Haversian sys- tems or osteons. Occasionally, the osteoclasts not only destroy unremodeled bone but osteons as well, leaving behind only fragments of the formerly complete osteons (Figure 24). The effect of this process increases the number of new osteons and osteon fragments with increasing age at the expense of the original circumferential lamellar bone (for details of this process see Enlow, 1963; Jowsey, 1960). Utilizing cross- sections through the midshaft, these structures can be objec- tively examined in four, precisely defined, circular, micro- scopic fields in the outer third of the long bone cortex. Age determinations are made from counts of these structures using regression formulae developed from Kerley's study of 126 undecalcified ground thin-sections of femora, tibae, and fibulae, taken from individuals of documented age, sex, and clinical history. The regression formulae contain standard errors as low as 9.39 years (femur), 6.69 years (tibia), and 5.27 years (fibula). In recent years, two important modifications of the Ker- ley aging method have appeared in the literature. Using thin sections from the mandibles, femora, and tibiae of 59 cada- vers of known age, sex, and medical history, Singh and Gun- berg (1970) examined in two circular fields the total num- ber of osteons, the average number of concentric lamellae per osteon, and the average diameter of the Haversian canals. From this data, they worked out regression equations with standard errors as low as 2.55 years. Unfortunately, the Singh and Gunberg (1970) study was based on a sample of individuals aged 39 to 87 years and clearly cannot be applied to ossuary material where evidence from the pubes suggests most adults died prior to age 40. In 1969, Ahlqvist and Damsten pointed out that (1) Kerley's reliance upon several structures for age assessment overly complicated the method while adding little accuracy; (2) that his use of circular visual fields necessitates specimen movement to distinguish structures on the borders of the field; and (3) that one of the visual fields he selected for examination falls directly upon the linea aspera, an area that demonstrates more non-age related osteon variation than other areas. In their modified version of Kerley's method, Ahlqvist and Damsten consider only the total per- centage of remodeled bone (osteons and osteon fragments) from four square fields placed just inside the outer surface of the bone. They spaced the fields around the circumfer- ence of the bone to fall between Kerley's fields and not on the linea aspera. Using a 100-square ruled ocular micro- meter, inserted into the eyepiece so that one side (10 squares) of the grid measured one millimeter at the level of the section, they counted the number of squares more than half-filled with either osteons or osteon-fragments, and rep- sented the result as a percentage. Finally, by applying this method to 20 unstained, ground sections from midshafts of 54 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY f*N* ' a>.jl » VV >-|HBBBS- FIGURE 24.—Cross-section of human cortical bone showing (a) osteons, (b) unremodeled circum- ferential lamellar bone, and (c) osteon fragments. femora of known age at death, they developed a regression expressing the relation between age and the percentage of remodeled bone. Their linear regression formula contains a standard error of 6.71 years; Of the Kerley, and Ahlqvist and Damsten studies, Ker- ley's offers the lower standard error, draws upon the larger sample, and would appear to be the better of the two methods. However, many sections prepared from the pos- sibly 400 year-old ossuary specimens lack the clarity and resolution of sections prepared from fresh anatomical speci- mens. Although it is possible to readily distinguish remodeled bone from circumferential lamellar bone, often it is difficult to distinguish whole osteons from osteon fragments. Largely because the Ahlqvist and Damsten modification does not re- quire the latter distinction to be made, it was chosen over Kerley's original method for use in this study. The femora were selected for sectioning since they occur- red in great frequency in the ossuaries and femora formed the basis for Ahlqvist and Damsten's (1970) study. The sample was comprised of 151 right femora, 66 from Os- suary I, and 85 from Ossuary II. Right femora were given preference because in both ossuaries they outnumbered the lefts. In each bone two horizontal cuts about 15 mm apart were made three-quarters of the way through the bone just below the anterior midshaft. The remaining quarter which includes the linea aspera was not removed since it was not to be used for the age determination and leaving it intact preserved the original length of the bone for future studies. A standard 10 inch hand saw was used for cutting and a Foredom Model 73 dental drill with a No. 8 round bit was used to disconnect the section from the long bone shaft. SPECIMEN PREPARATION Both Ahlqvist and Damsten (1969), and Kerley (1965) utilized ground, undecalcified, unstained thin sections for their age determinations, although the former mentioned that unstained decalcified sections would work as well for fresh bone. Ahlqvist and Damsten recommend cutting the decalcified bone in sections approximately 25 microns thick and grinding the undecalcified sections to a thickness of 30-50 microns. DETERMINATION OF SEX AND AGE AT DEATH 55 Unlike fresh bone, archeological bone often has lost mineral constituents due to prolonged exposure to soil chemi- cals and has been invaded by organic and inorganic sub- stances from the surrounding soil. These changes limit the quality of ground sections and necessitate the introduction of new techniques to reveal their details. Experimentation with bone samples from Ossuary I using several decalcifying solutions for different lengths of time, showed that de- calcified stained thin sections could not be used success- fully, because they lack enough remaining bone mineral to allow the structures necessary for aging purposes to be dis- tinguished. Accurate age determinations could be made from unstained ground thin sections, but even with these, two major problems were encountered. First, after a dry arche- ological bone section was mounted on a thick slide, it would absorb water during the preparation process, expand, and crack the nonabsorbant glass slide. Initially, the problem was solved by using specially cut, plastic slides which would bend and not break as the bone section expanded. Unfortunately, several of the plastic slides bent to the point where they would not adhere to the vacuum-supported chuck face of the cutting saw. The problem was eliminated only by infiltrating the bone specimen with an epoxy resin, which prevented the absorption of water and the resultant bone expansion and thus saved the slide from destruction. (The epoxy-resin used is Araldite AY105 with Hardener 935F. These products can be obtained from Chemical and Engineering Company, Inc., 221 Brooke Street, Media, Pennsylvania, 19063. Equal weights of the araldite and hardener were mixed thoroughly on a hot plate until the originally milky mixture became transparent. For infiltration, 5 grams of the araldite mixture were dissolved in 150 ml. of Toluene.) The second major problem involved the organic and in- organic contaminants mentioned above. Their permeation of the important periosteal third of the cortex obscured the microstructures used in age determination. In some cases, the extraneous material was removed by gently teasing the section surface with a dissection needle under a dissecting microscope. Unfortunately, the araldite, needed to keep the bone from expanding, also firmly held the extraneous mate- rial in the bone matrix. Eventually better results were ob- tained by placing all blocks in an ultrasonic cleaner and infiltrating them with a mild oxidizing agent prior to the araldite infiltration. This somewhat time-consuming modifi- cation allowed satisfactory age determinations to be made. In short, the procedure initially involved the cleaning of the bone block after its removal from the femoral shaft. The block was then infiltrated with araldite and mounted on a glass slide. At that time, a slab of bone with smooth parallel surfaces had to be removed from the block and remounted on a second glass slide for final grinding. This step was neces- sary since the final thin section had to be of even thickness and consequently be prepared from a bone specimen with parallel surfaces. The parallel-sided bone slab was removed by making two cuts through the original block. The first cut near the outer surface of the bone created one outer smooth surface that was parallel to the glass slide. A second cut, closer to the slide and also parallel, freed the bone slab from the original block. Since the bone slab had two parallel surfaces, it could then be mounted on the final slide and be evenly ground down to a finished specimen. Specifically, the entire process was as follows: The block was first washed with soap and water in the ultrasonic cleaner for one hour, thoroughly rinsed, and then washed again in distilled water in the ultrasonic cleaner for 30 minutes. This procedure removed much of the soil from both the medullary and outer surfaces and even some loose material from within the cortex. The block was then infiltrated with a mild oxidizing agent to remove as much as possible of the organic and inorganic material that had permeated the outer cortex. Several oxidizing agents were used successfully, but the most effective was a 20 percent by volume solution of household chlorox. After being left overnight under vacuum in a bell jar, the block was washed again in the ultrasonic cleaner for 20 minutes to remove any residues of the oxidizing reaction. Following cleansing, the bone block was prepared for sectioning by being infiltrated with araldite to limit expan- sion during the cutting process. Approximately 20 blocks at a time were placed in the toluene-araldite solution under vacuum in the bell jar. About 10 grams of araldite were added each day until the solution became quite thick. The specimens were left under vacuum for a total of five or six days, or until the araldite solution began to gel. At that time, the specimens were removed from the vacuum, cleaned of any adhering araldite and dried overnight in an oven heated to 60° C. When dry, each block was mounted on a ihick glass slide with the araldite-hardener mixture as the mounting medium. The medium was applied to one flat surface of the block and the latter was placed on a clean glass slide. Slight pres- sure was applied to the block to remove excess air bubbles. The slide and mounted bone block were then dried at room temperature overnight or until the mounting media had hardened. At that time it was necessary to remove the parallel-sided bone slab from the block for reasons discussed earlier. After the mounting media was thoroughly hardened, the slide with mounted block was held by water vacuum to the chuckface of an Ingram thin section saw. An initial cut was made on the outer edge of the bone (across the long axis) that re- moved the rough outer edge and created a smooth surface that was parallel to the glass slide. A second cut was then made approximately 3 mm closer to the slide than the first cut, which freed a bone slab from the original rough mount. The free slab had smooth, parallel sides and, of course, still represented a cross-section of the femur. One side of the free slab was then machine-polished. The free slab was moved 56 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 8 slowly against the direction of the high speed wheel for 90 seconds, rotated 180° and polished for an additional 90 seconds or until deep scratches and cuts were removed from the bone surface. The second polishing wheel was used to remove all fine scratches and irregularities from the bone surface. With the wheel spinning, the slab slowly was moved against the direction of the wheel for 30 seconds, rotated 180° and again polished for 30 seconds. The laboratory was equipped with an A.B. Buehler polishing apparatus (No. 48-1512) with two polishing wheels; (1) a high speed (550 rpm) wheel covered with an A.B. Texmet cloth (No. 40- 7668) and (2) a low speed (168 rpm) wheel covered with an A.B. Microcloth lap (No. 40-7218). The first wheel was coated with A.B. Metadi diamond polishing compound (3 microns) and a mixture of equal parts A.B. mineral spirits (No. 40-8140) and A.B. polishing oil (No. 40-8142). The second wheel had a thin coat of water and Buehler A.B. Bamma micropolish added to the cloth. The polishing was followed by a thorough washing with a grease-removing detergent, a 30-second rinsing in ether- chloroform to remove any remaining grease, and then a bathing in an oxidizing solution for 15 minutes in the ultra- sonic cleaner. The oxidizing agent was used again to insure removal of all extraneous material from near the bone sur- face, and was followed by a final 15 minute wash in distilled water. The bone slab was then dried overnight in an oven at 60° C. The dry slab was mounted to a clean slide of uni- form thickness, again using the araldite-hardener mixture as the mounting medium. The mounting medium, slide, and bone slab had been uniformly heated to 200° C to obtain uniform expansion, and thus to avoid slide breakage. Follow- ing this, the mount was placed in a "C" clamp between two flat sections of 3 mm cork. With the clamp tightened, the media was allowed to set overnight or until it hardened. Once the mount was firmly fixed, the slide was attached to the chuckface of the cut-off saw. A final cut was made ap- proximately 500 microns from the slide surface. The remain- ing mounted specimen was then attached to an Ingram thin-section grinder and ground down to a thickness of about 80 microns. The polishing procedures outlined earlier were repeated to remove all scratches from the bone surface. After a detergent wash, and a rinse in distilled water, the section was dried, and covered with a cover glass, mounted with Permount or Canada Basalm. If the finished thin section still displayed an excessive amount of foreign material, the cover slip was removed with xylene, and the surface of the specimen was brushed with the oxidizing agent. OSTEON COUNTS Having produced satisfactory thin sections using the above procedure, all observations were made in the manner defined by Ahlqvist and Damsten (1969). An ocular square- ruled reticule was inserted into the left eyepiece of a Leitz Ortholux microscope. A square reticule containing 100 squares was especially prepared by Bunton Instrument Com- pany, Washington, D.C., to meet the specifications of this project. One side of the reticule (10 small squares) measured exactly 1 mm at the level of the section. Four visual fields were examined along the periosteal surface of the bone sec- tions. In each field, the small squares more than half-filled with either osteons or osteon fragments were counted. The numbers for the four fields were averaged and the age at death computed by the linear regression formula y=0.991x — 4.96±6.71, where y equals age at death and x equals the percentage of osteons and osteon fragments in the four fields. An analysis of the data is presented in Table 34 for femora from Ossuary I and in Table 35 for femora from Ossuary II. Both tables present the average number of microscopic grid squares more than half-filled with osteons or osteon fragments (percentage remodeled) and the cor- responding estimated age at death for each specimen. Table TABLE 34.—Percentages of remodeled bone and corresponding age estimates for 66 right femora from Ossuary I Specimen Percent- Esti- Specimen Percent- Esti- number* age re- mated number* age re- mated modeled age modeled age 46+ 42 36.7 12-15 41 35. 7 39B 31 25.8 E. 109 47 41.6 113 36 30.8 42 32 26.8 27 34 28.8 111-112 41 35. 7 Bt 112-113 44 38.7 114 51 45.6 2-17B 29 23.8 110 43 37.7 8 42 36. 7 102 28 22.8 30 51 45.6 W119 27 21.8 NR128(A) 43 37.7 39C 47 41.6 2-17C 27 21.8 112-113 38 32.7 20-28A 50 44.6 38B 35 29. 7 61B 60 54.5 126-112B 54 48.6 34 54 48.6 103 52 46.6 464-A 38 32.7 128D 31 25.8 18A 38 32. 7 121 54 48.6 39A 42 36. 7 38A 38 32. 7 111? 45 39.6 7 51 45.6 20-38 31 25.8 101-120 50 44.6 18B 36 30. 7 108 39 33. 7 464-D 48 42.6 SE Cor. 48 42.6 2-17A 36 30.8 40-56-57 41 35. 7 61A 33 27.7 TR 42 36. 7 464-C 47 41.6 24 31 25. 8 10-13B 38 32. 7 45 40 34. 7 10-13A 32 26.8 29 36 30. 7 37 42 38.6 126-112A 36 30.7 46+B 31 25.8 2-17D 37 31. 7 20-38B 37 31. 7 NR61 37 31. 7 105 31 25.8 10-13C 48 42. 6 50-60? 33 27.7 128C 31 25. 8 115 56 50.5 W117-118 53 47.6 23A 45 39.6 15 36 30. 7 23B 33 27. 7 128B 30 24.8 *Number refers to location of nearby crania. DETERMINATION OF SEX AND AGE AT DEATH 57 TABLE 35.—Percentages of remodeled bone and corresponding age TABLE 36.—Age distribution (in years) of adults in Ossuaries I estimates for 84 right femora from Ossuary II and II as estimated from femoral microscopic remodeling Specimen number* Percentage Estimated remodeled age Specimen Percentage Estimated number* remodeled age 35 38 44 28 30 60 47 48 54 48 44 45 56 54 28 39 54 48 41 37 36 34 27 38 48 41 30 56 62 43 42 37 49 42 51 62 46 71 32 36 40 44 AL1-1 AL1-2 AL1-3 AL1-4 BL1-1 BL1-3 BL1-4 CL1-1 CLI-2 DL1-1 DLl-2 DL1-3 DL1-4 DL1-5 EL1-1 EL 1-2 EL1-3 EL1-4 EL 1-5 EL1-6 EL1-7 FL1-1 FLl-3 GL1-1 GL1-2 GL1-3 GL1-5 HL1-1 AR1-1 AR1-2 ARl-3 BR1-1 BR1-2 BRl-3 BR 1-4 BRl-5 BR1-6 CR1-1 CRl-2 CR1-3 DR1-1 29. 5 33.0 38.6 22.5 24.8 54.0 41.9 42. 48. 42. 38. 40.0 50.0 48.6 22.3 33.4 48.6 43.5 35. 7 31.2 30.2 28.5 21.6 32.2 42. 9 35. 1 24.8 50.5 56.5 37.9 36. 7 32.0 43.4 36.2 45. 1 56.0 40. 9 65. 7 26.8 31.0 34.2 38.9 66 65 55 56 55 34 34 55 37 29 45 38 45 44 34 45 40 27 47 44 33 39 33 38 60 37 46 44 40 46 58 48 57 43 39 52 38 44 36 37 27 56 60.0 55.2 50.0 50.5 50.0 28. 7 28.2 49. 1 31.7 24.0 39.4 33.0 39. 1 38.6 29.0 39. 1 34. 7 22.0 41.4 38.6 27.3 33. 9 27. 7 33.0 54.8 31.2 40.6 38.3 34.4 40. 1 52.3 43. 1 51.0 37.2 33.4 46. 1 ER1-1 ERl-2 FR1-1 FR1-2 FR1-3 FR1-4 CC-1 CC-2 DC-1 DC-2 DC-3 DC-4 EC-1 EC-2 EC-3 EC-4 EC-5 EC-6 EC-7 FC-1 FC-2 FC-3 FC-4 FC-5 FC-6 FC-7 FC-8 FC-9 AL-BL-1 AL-BL-2 BL-CL-1 BL-CL-2 DL-EL-1 DL-EL-2 DR-ER-1 DR-ER-2 32. 38. 28 SI. 21. DR-ER-3 B-l-1 B-2-1 B-2-2 B-4-1 B-5-1 50.0 *Number refers to either the grid square or distinct bundles (B) with which the femora were associated. 36 compares the data from the two ossuaries when the age estimates are grouped into 5-year intervals. Age estimates from the femoral osteon counts are mar- kedly different from those determined from the symphyseal face of the pubis. Figure 25 demonstrates this fact by means of separate adult mortality curves calculated from each os- suary, using the data yielded by each method. The curves express the percentages of age determinations in 5-year age categories that range from 20.0 years to 69.9 years. Al- Age group Ossuary / Ossuary // No. % No. % 20-25 5 14 16 13 9 7 2 0 0 0 66 8 21 24 20 14 11 3 0 0 0 8 10 18 17 11 9 7 3 0 1 84 10 26-30 12 31-35 21 36^0 20 41-45 13 46-50 11 51-55 8 56-60 4 61-65 o 66-70 1 Totals though all curves demonstrate an increase in the percentage of deaths between 30 and 35 years and a decline in the percentage of deaths after age 35, they differ in the relative frequencies for each age category. The pubis indicates nearly 20 percent more individuals were dying in the 30.0-34.9 age category than suggested by femoral osteons. After age 35, the pubes indicate a rapid decline in the percentage of deaths with no one living after age 50 in Ossuary II or after age 60 in Ossuary I. In contrast, femoral osteon counts indicate a slow decline in numbers of deaths after age 40 with some individuals living to nearly 60 years in Os- suary I and to 70 years in Ossuary II. Whereas the pubis suggests an average adult age at death of 34 years for Os- suary I and of 31 years for Ossuary II, femoral osteon counts indicate average adult ages of about 36 and 38 years for Ossuaries I and II, respectively. These differences are great enough that they would produce quite different demographic profiles and could lead to contrasting interpretations. The differences between the death curves described above result largely from two factors: differences in the accuracy of the aging methods, and differences in sample size of the bones selected for determining age. The second factor is probably the larger contributor to the discrepancy, but the first does warrant some discussion. Pubis metamorphosis is probably the less accurate of the two methods of aging used in this study. A major problem in determining age from the symphyseal face of the pubis involves distinguishing a ventral rampart that is building up in the early part of the fourth decade from one that is breaking down in the sixth. The problem is greater in females where scars of parturition may further alter the appearance of the symphyseal face. Consequently, it is possible that some of the fragmentary pubes judged to be in the 30 to 35 year interval were actually much older. The femoral osteon method is relatively straight-forward and presents few interpretational problems as long as rea- sonably clear histological sections are available for examina- 58 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 SYMPHYSEAL FACE OF PUBIS FEMORAL REMODELING OSSUARY H. n=60 OSSUARY I OSSUARY n n=84 OSSUARY I n=66 35-39 40-44 AGE IN YEARS FIGURE 25.—Comparison of adult mortality curves from Ossuaries I and II as determined from the appearance of the symphyseal faces of the pubes and the degree of microscopic cortical remodeling in the femora. tion. As discussed earlier, the presence of extraneous substances in the bone cortex presented time-consuming methodological problems in specimen preparation. However, additional sections for each specimen were prepared and subjected to more thorough cleaning until accurate evalua- tions could be made. Both aging methods assume that the standards developed from modern populations are applicable to prehistoric popu- lations from different cultural and environmental areas. Todd (1921) found that little difference exists between Whites and Negroes in the rate of symphyseal metamorphosis. Probably, therefore, if the prehistoric population represented in the ossuaries is different in this respect, the difference is minimal. The possibility is greater that the osteon remodel- ing rates are different between populations, since remodeling is strongly influenced by physical activity. If the Indians rep- resented in the ossuaries were more active than the popula- tions represented in Ahlqvist's and Damsten's sample, and had faster osteon turnover rates, then the regression formula might overestimate the age at death. However, Ahlqvist (pers. comm.) has indicated that the samples used in the Ahlqvist and Damsten study were taken from routine autop- sies, and probably represent predominantly lower economic class individuals, who may have approached the Indians in physical activity. Even though such factors as alcoholism, arteriosclerosis, and cardiovascular disease may affect osteon turnover rates (Ortner, 1970), there is no indication that these diseases were more prevalent in the Ahlqvist and Damsten sample than in the Indians from the ossuaries. In the absence of any further such adverse evidence, the femoral osteon method appears to be the more accurate method for the present purpose. An additional justification for relying upon age estimates calculated from femoral osteon counts involves the relative representation of femora and pubes in the ossuaries. Whereas age estimates from the pubes represent only 54 percent and 61 percent of adults in Ossuaries I and II, respectively, those from the femora represent 96 percent and 85 percent, re- spectively. Consequently, most of the difference between death curves can be explained by the absence of older pubes from the ossuary samples. This explanation is logical since the older, fragile, osteoporotic pubes would probably be damaged easier than the younger ones. Reconstruction of Demographic Profiles and Population Size The ethnohistorical and archeological information pre- sented earlier suggests that the ossuaries contain nearly a complete representation of all individuals who died in the contributing populations during a culturally prescribed num- ber of years. Although some individuals who died away from their villages or whose bones were lost prior to ossuary burial may not be represented, the skeletal samples still offer a relatively unique opportunity to reconstruct demographic profiles for the formerly living populations. The procedure used in these reconstructions assumes that (1) the skeletal samples are complete, (2) the ages at death can be accu- rately determined, and (3) the size of the living populations and their death rates remained constant during the time interval represented by the ossuaries. The completeness of the sample and the accuracy of the age determinations have already been discussed and validated. In regards to the third assumption, there is little doubt that fluctuations in population size and mortality occurred. However, since only a few years are represented by each ossuary, the demographic effect of these fluctuations would have been negligible. In this chapter, the ages at death of the individuals in the two ossuaries are used to construct mortality curves, sur- vivorship curves and life tables for the populations repre- sented. These statistics offer important demographic data on longevity, age-specific mortality, and life expectancy at all ages. Crude mortality rates (i.e., numbers of individuals dying annually for every 1000 of the living population) are calculated from the life tables. Archeological data then pro- vide clues as to the length of time represented by each ossuary. Finally, all of these data are considered in estimat- ing the total population size. The terms Population I and II are introduced in this chapter to refer to the living popu- lations represented by each ossuary. Table 37 shows the frequency of deaths by 5-year age intervals in the two ossuaries. As discussed in the last chapter, age estimates below 20 years were determined from femoral length and stages of epiphyseal union, while ages 20 years and over were calculated from femoral microscopic cortical remodeling. Only 5 and 8 percent of the total individuals in Ossuaries I and II, respectively, are not represented by femora. TABLE 37.—Age distribution of subadults and adults from Os- suaries I and II as estimated from femoral length and epiphyseal union (0-20 years) and femoral microscopic remodeling (20 +■ years) Ossuary I Ossuary II Age interval % % No. No. 0.0-4.9 36 29.03 56 32.37 5. 0- 9. 9 14 11.29 12 6.94 10.0-14.9 6 4.84 7 4.05 15.0-19.9 2 1.61 14 8.09 20. 0-24. 9 5 4.03 8 4.62 25. 0-29. 9 14 11.29 10 5.78 30. 0-34. 9 16 12.90 18 10.40 35. 0-39. 9 13 10.48 17 9.83 40. 0-44. 9 9 7.26 11 6.36 45. 0-49. 9 7 5.65 9 5.20 50.0-54.9 2 1.61 7 4.05 55. 0-59. 9 0 0 0 124 0 0 0 3 0 1 173 1.73 60. 0-64. 9 0 65. 0-69. 9 0.58 Mortality Curves Figure 26 converts the percentages contained in Table 37 to mortality curves for the living populations represented by the ossuaries. The curves for the two populations are gen- erally similar. They both show the highest frequency of deaths in the first five years of life, a dramatic decline through childhood into adolescence, a steady increase during early adulthood to a maximum adult death frequency be- tween 30 and 35 years, followed by a steady decline after age 35. Perhaps the most significant difference between the two curves is in the period of adolescence. The curve for Popula- tion I reached a low of 2 percent between 15 and 20 years, whereas that for Population II is 8 percent, an actual in- crease of 4 percent over the previous interval. Although there are only two individuals in Ossuary I between ages 15 and 20, Ossuary II contains 14 in that interval. Several of the 15 to 20 year-olds in Ossuary II had femora with small head 59 60 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 UJ o < z UJ o CC UJ Q_ 40 r- 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 OSSUARY I OSSUARY H 0-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70 AGE IN YEARS FIGURE 26.—Comparison of mortality curves calculated from Ossuaries I and II. diameters suggesting they probably were females, whose deaths perhaps reflect problems resulting from childbirth. The remaining differences between the two curves are largely products of greater longevity in Population II. More individuals were dying in Population I during the early adult years 20 to 35. Death frequencies were the same between ages 35 and 45, but after age 45, more individuals were dying in Population II than in Population I. No one lived past age 55 in Population I, whereas three percent of Popu- lation II exceeded that age, with one individual living to nearly 66 years. Survivorship Curves Differences in mortality between populations are more clearly revealed through survivorship curves (Figure 27). A curve of survivorship is merely the reverse of the mortality curve and plots the percentage of the original population surviving after each five-year interval. According to Figure 27 for example, 71 percent of all individuals born in Popu- lation I were still alive after five years, while only 68 percent of Population II were living. Survivors were approximately equal by ages 10 and 15 years. Then the higher mortality of adolescent females discussed earlier for Population II, created a higher survivorship for Population I by ages 20 and 25. The higher death rate between ages 25 and 30 in Population I restored an equality of survivors by age 30. After age 30, greater longevity in Population II was reflected by a proportionately greater number of survivors in each successive age category, until 70 years when there were no survivors remaining in either population. In short, the curve of survivorship accentuates the differences between the two populations, documenting that Population II exhibited a higher infant and adolescent mortality, but greater adult longevity. Life Tables Perhaps the most informative of all statistical presenta- tions of demographic data is the life table. The procedure of life table construction has been summarized by several in- DEMOGRAPHIC PROFILES AND POPULATION SIZE 61 UJ < z UJ o rr UJ 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 OSSUARY I OSSUARY n 50 55 15 20 45 10 25 30 35 40 AGE IN YEARS FIGURE 27.—Comparison of survivorship curves calculated from Ossuaries I and II. 60 65 70 vestigators (Swedlund and Armelagos, 1969; Acsadi and Nemeskeri, 1970) and will not be discussed in detail here. Essentially, it involves computing a number of attributes that characterize the demographic structure of the living population and that can be compared with data taken from both living and skeletal populations. The life table contains calculations for x, the age interval; Dx, the total number of deaths occurring in age interval x; dx, the percentage of total deaths occurring in age interval x; lx, the number of survivors of age interval x; qx, the probability of dying at age interval x; Lx, the total number of years lived between age interval x and the following age interval; Tx, the total number of years lived after a lifetime of all individuals who reach the age interval x; and e°x, the life expectancy of an individual of age x. Two types of life tables are usually employed by demog- raphers studying modern populations: generation life tables and current life tables (Dublin and Spiegleman, 1941). Al- though both types are organized in the manner previously described, they differ in the type of population sample they draw upon. Generation life tables are constructed from data obtained by following a generation of people from birth until death, recording age at death for each individual in the population. Obviously, such tables must be recorded in retro- spect, using well-documented death records. Usually the records necessary for generation life table construction are not available and demographers must con- struct current life tables instead. These tables are con- structed from age distributions within living populations, and assume that the mortality rates suggested by the popula- tion composition will remain constant throughout the lives of the individuals in the population examined. Of course, in modern human populations, this assumption is not valid. For example, changing medical practices markedly affect mortality rates over periods of 50 to 70 years. In fact, Dublin and Spiegelman (1941) have shown that the difference between current and generation life tables can be an effec- tive measure of improvement in public health. Although 62 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 some demographers (Fisher, 1923; Dublin and Lotka, 1963) have attempted to modify current life tables to compensate for this effect, their tables remain only approximations for the populations described. The basic concepts underlying current life table recon- struction have been applied by biologists and ecologists to the study of demography in wild animal populations. Non- human populations in natural environments display rela- tively stable mortality as long as the environment remains generally unchanged. Using various methods to determine age at death, biologists have constructed life tables, survivor- ship curves, and other demographic statistics for Dall moun- tain sheep (Murie, 1944), rotifers (Edmondson, 1945), herring gulls (H. Marshall, 1947), robins, blackbirds, song thrushes, starlings, and lapwings (Lack, 1943), barnacles (Hatton, 1938), rabbits (Green and Evans, 1940a, 1940b, 1940c), sunfish (Ricker, 1945), lizards (Tinkle, 1961, 1967; Tinkle, Wilbur, Tilley, 1970), mosquitos (Crovello and Hacker, 1972), and many other species (Deevey, 1947). All of these studies demand adequate samples and accurate age determinations and assume that demographic features re- main relatively stable throughout the lifespan of the population. Angel (1947), and Swedlund and Armelagos (1969) have suggested that the approach to life table construction utilized by biologists can be applied successfully to prehistoric human populations as well. Since most archeological populations did not experience the rapid nutritional and medical ad- vances of modern populations, they approach the demo- graphic stability documented for nonhuman populations in natural environments. The similarity of aging techniques available to both the biologist and physical anthropologist and the homogeneity of archeological populations further validate the use of the current life table in prehistoric demo- graphic reconstruction. It should be noted that Angel (1969a) now avoids using the life table for demographic reconstruction, feeling that population instability and the loss of infants during the period of cemetery use might intro- duce error. However, population instability should affect all demographic data, not just the life table, and should be negligible for the short periods of time represented by ossuaries. The life tables for Populations I and II presented in Tables 38 and 39 were computed as follows: the symbol *'x'' represents the total span of years contained in each age interval. Although modern life tables utilize age intervals of single years, a five-year age interval is used in this study since aging criteria for adults contain approximately a five- year standard error. A smaller interval would be misleading due to aging errors and a larger interval would limit the information gained from the analysis. The total number of deaths (Dx) for each age interval is weighted so that the total deaths in the life tables equal the total number of individuals in the ossuaries. The weighting assumes that the missing femora are evenly distributed among all age intervals in the group considered. Although there is no assurance that the missing femora are evenly distributed, the assumption must be made since a complete representation is needed. Adults and subadults were weighted separately to equal the minimum adult and sub- adult representation as determined from individual bone counts. Weighting factors of 1.07, 1.05, and 1.18 were ap- plied to the femoral counts of Ossuary I subadults, Ossuary I adults, and Ossuary II adults, respectively. A weighting factor is not needed for Ossuary II subadults since the mini- mum count actually came from the femora. The number of survivors at each age interval (lx) is calculated by subtracting the percentage of deaths (dx) in the age interval from the number of survivors entering that interval, beginning with an initial number of 100 TABLE 38.- —Life table reconstructed from Ossuary / age distribution Age interval No. of deaths % of deaths Survivors Probability of death Total no. yrs. lived between x and x + 5 Total no. yrs. lived after lifetime Life expectancy (*) (Dx) (dx) (lx) (fO (Lx) (Tx) (e°x) 0 0 0 100. 00 .0000 426. 125 2105. 600 21.06 5 39 29.55 70.45 .2955 323. 850 1679. 475 23.84 10 15 11.36 60.09 . 1612 284. 075 1335.625 22.60 15 6 4.55 54.54 .0770 268. 900 1071. 550 19.65 20 2 1.52 53.02 .0279 255. 625 802. 650 15. 14 25 5 3. 79 49.23 .0715 217.750 547. 025 11. 11 30 15 11.36 37.87 .0308 157. 150 329. 275 8.69 35 17 12.88 24.99 .3401 98. 425 172. 125 6.89 40 14 10.61 14.39 .4246 51.075 73. 700 5. 13 45 11 8.33 6.05 .5793 18. 875 22. 625 3.74 50 6 4.55 1.50 .7521 3.750 5.750 °. 50 55 2 1.52 0.00 1.0000 0.000 0.000 0.00 60 0 0 0.00 0. 0000 0.000 0.000 0.00 DEMOGRAPHIC PROFILES AND POPULATION SIZE TABLE 39.—Life table reconstructed from Ossuary II age distribution 63 Age interval No. of deaths % of deaths Survivors Probability of death Total no.yrs. lived between x and x + 5 Total no.yrs. lived after lifetime Life expectancy (*) (Dx) (dx) (lx) (qx) (Lx) (Tx) (e°x) 0 0 0 100. 00 .0000 425. 525 2297. 900 22.98 5 56 29.79 70.21 .2979 335.100 1872. 375 26.67 10 12 6.38 63.83 .0909 309. 850 1537.275 24.08 15 7 3.72 60. 11 .0583 281. 925 1227. 425 20.42 20 14 7.45 52.66 . 1239 251.325 945. 500 17.95 25 9 4.79 47.87 .0910 223. 400 694. 175 14.50 50 12 6.38 41.49 . 1333 179. 525 470. 775 11.35 35 21 11. 17 30.32 .2692 125.000 291. 250 9.61 40 20 10.64 19.68 .3509 81. 125 166. 250 8.45 45 13 6.91 12.77 .3511 49. 225 85. 125 6.67 50 11 5.85 6.92 .4581 23. 950 35. 900 5. 19 55 8 4.26 2.66 .6156 7.975 11.950 4.49 60 4 2. 13 0.53 .8008 2.650 3.975 7.50 65 0 0.00 0.53 .0000 1.325 1.325 2.50 70 1 0.53 0.00 1.0000 0.000 0.000 0.00 75 0 0.00 0.00 .0000 0.000 0.000 0.00 survivors. Survivorship values in the life table directly reflect mortality and correspond to the survivorship curves discussed earlier. The probability of death (qx) for each age interval is computed by dividing the percentage of deaths in the age interval by the number of survivors of that interval. The figure represents the probability of individual death in each age category and provides a valuable index of age-specific mortality. Calculations of life expectancy values (e°x) necessitate initial determinations of (1) the total number of years (Lx) lived between age interval x and the following interval, and (2) the total number of years (Tx) lived by all survivors of age interval x. The value of Lx is determined from the formula T 5(lx-f-lo) Lx=—r- where lo is the number of survivors of the age interval following interval x. The factor of 5 is introduced, since the life table is abridged to a 5-year age interval. The value of Tx is derived from the Lx value by the formula w-l Tx=2 Lx. X The value expresses the total number of years that can be lived by the survivors of each interval. The Tx value has no demographic significance in itself, but is used to calculate life expectancy by the formula Tx lx ' e*u=- Consequently life expectancy at birth is computed by the formula o_To e°-lo"' A comparison of life tables for the two populations re- veals important demographic information. The percentage of deaths (dx) and the number of survivors (lx) have already been discussed. A comparison of qx for each popu- lation reveals differences in age-specific mortality. The probability of dying in Population 1 is greater than in Population II for all ages except 15-20 and 20-25 years- In both of these age intervals, the probability of dying was greater in Population II, reflecting the high incidence of adolescent and young adult deaths expressed in the mor- tality curve. Life expectancy for Population II was greater than for Population I in all age intervals. A newborn entering Pop- ulation I could expect to live only 21 years. However, if the child was one of the 71 percent that survived the first five years, it could expect to live another 24 years. After age 10, life expectancy diminished with age until by age 50 only 1.5 percent of the original population was still alive and their life expectancy was a mere 2.5 years. Similarly, the average newborn entering Population II could expect to live about 23 years, an 8 percent increase over life expectancy at birth in Population I. If the child in Population II was among the 70 percent living to age five, it could expect to live an average of 27 more years. Life expectancy declined steadily thereafter until age 65, when the surviving 0.53 percent of the original population could expect to live only 2.5 more years. In Population I, the age intervals with the lowest 64 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 probabilities of death are 15-19 and 20-24 years. In Popu- lation II, they are 10-15 and 5-10 years, reflecting the higher mortality of adolescents and young adults in the latter population. Table 40 lists life expectancy figures at birth for several populations. While the figures for archeological populations are practically all that are available, they have little com- parative value, due to sample inadequacies and differences in methods of age determination. For example, the high life expectancy for Pecos Pueblo was calculated by Goldstein (1953) on ages published by Hooton (1930), and originally determined by Todd (1927:494), who admits that "the very young children are represented by such scattered remnants that, as usual, they must be discarded." In the absence of the probable high number of infant skeletons, Goldstein's calculation of life expectancy at 42.9 years has little meaning. Stewart (1962) has shown that the data on the Archaic Indian Knoll population from Kentucky (see Table 40) are of limited value since only part of the sample was considered and questionable criteria were used to determine age. The data from Nubia (Swedlund and Armelagos, 1969; Table 40) are derived from 328 skeletons excavated from a Meinarti cemetery. Although the archeologists intensively excavated the major portion of the cemetery, part of it had eroded away prior to excavation. In addition, most of the skeletons excavated had been disturbed, leaving the overall completeness of the sample somewhat questionable. Swed- lund and Armelagos estimated the ages at death from a combination of stages of dental eruption and changes in the pubic symphyses. The remainder of the comparisons in Table 40 suggest that life expectancy at Nanjemoy was very close to that of Ancient Greece, but substantially lower than that of 11—16th century England, 19th century United States Whites, 20th century United States Negroes and modern populations of India and England. In addition to the foregoing Churcher and Kenyon (1960) studied two Huron ossuaries which unfortunately had been disturbed prior to excavation, first by construction equipment and then by local looters. Their analysis revealed that of 90 recorded skeletal ages, only five were under 10 years, while 29 were between ages 16 and 20. Such an unusual preadoles- cent death rate probably means that a considerable number of preadolescent skeletons were lost from the sample prior to analysis. Anderson's (1964) analysis of another Iroquois ossuary indicates that his sample was probably complete. His selec- tion of age intervals and his uncertain criteria for determining adult age, however, do not permit close comparison. He found that 28 percent of his sample was between birth and 12 years, 14 percent was between 13 and 16 years and 58 percent was "adults" over 17 years. By comparison, 36 to 41 percent of the individuals from the Nanjemoy ossuaries was between birth and 10 years, 4 to 5 percent was between 10 and 15 years and 56 to 60 percent was over 15 years of age. This crude comparison generally shows that the mortality rate in the Nanjemoy populations was higher in infancy, but lower in adolescence than that in the Huron population. Unquestionably, data on life expectancy and age specific mortality provide an evaluation of overall population fitness and health status. However, until reliable aging techniques are applied rigorously to complete samples of carefully exca- vated skeletons, sound comparisons of earlier populations will not be possible. Of all skeletal samples previously re- ported for aboriginal North American Indian populations, only the Huron ossuary described by Anderson (1964) offers a sample suitable for meaningful demographic comparison. As we have seen, Anderson's Huron population displayed a lower infant mortality and greater adult longevity than the TABLE 40.—Life expectancy at birth in 15 world-wide populations, arranged in order of increasing magnitude Population Dates Life expectancy (years) Source Indian Knoll, Kentucky Nubia, Egypt Nanjemoy, Ossuary I Nanjemoy, Ossuary II Ancient Greeks Texas Indians European ruling families U.S. Caucasian U.S. Negro English India (Females) India (Males) Pecos Pueblo England and Wales (Males) England and Wales (Females) 3000 B.C 18.6 Johnston and Snov A.D. 1050-1600 19.2 Swedlund and Ar A.D. 1500-1600 20.9 A.D. 1500-1600 22.9 670 B.C.-A.D. 600 23.0 Goldstein, 1953: 4 A.D. 850-1700 30.5 Goldstein, 1953: 4 A.D. 1480-1579 33. 7 Goldstein, 1953: 4 A.D.1800 30. 0-35. 0 Goldstein, 1953: 4 A.D. 1900 33.8 Goldstein, 1953: 4 A.D. 1000-1100 35.3 Goldstein, 1953: 4 A.D. 1951-1960 40.6 Klebba, 1971: 2 A.D. 1951-1960 41.9 Klebba, 1971: 2 A.D. 800-1700 42.9 Goldstein, 1953: 4 A.D. 1965-1967 68.7 Klebba, 1971: 2 A.D. 1965-1967 74.9 Klebba, 1971: 2 DEMOGRAPHIC PROFILES AND POPULATION SIZE 65 Nanjemoy populations, but his choice of age intervals elimi- nates more detailed life table comparisons. Life expectancy in both the Huron and Nanjemoy popu- lations was markedly lower than among the 16th century European ruling families (Table 40). Sigerist (1965) has suggested however, that life expectancy at birth was as low as 20 to 25 years among European general populations during the 15th century. If his calculations are correct, they would indicate that overall life expectancy and mortality may have been approximately the same for the general populations in both Europe and North America during the 16th century. Recent medical advances of course, have led to markedly higher life expectancy in many countries, and even to a cur- rent peak of over 70 years in the United States. Most of this recent nearly 50-year increase in life expectancy can be ac- counted for by the remarkable reduction in infant mortality, following the introduction of antibiotics and other medical aids. Crude Mortality Rates Another statistic that not only provides information about mortality and longevity, but that can also lead to the recon- struction of population size is the crude mortality rate. This rate is usually expressed as the number of individuals dying per thousand per year, or less commonly as the number per hundred per year. The figure is a direct reflection of overall life expectancy and can present evidence for population decline, equilibrium, or expansion when considered in rela- tion to the birth rate. In the past, crude mortality rates have been derived by comparing age distributions within skeletal samples with age distributions documented for populations with known mortality rates. For example, Hooton (1920) compared the distributions of deaths determined from his Madisonville, Ohio skeletal material with documented deaths for a number of European populations (Table 41). His skeletal sample closely matched the Switzerland sample for the years 1873— 1877 which had a recorded death rate of 2.38 per hundred per year. Hooton then extended the death rate at Madison- ville to 3.0 per hundred to allow for infants that may have been overlooked during the excavation. Churcher and Kenyon (1960) used Hooton's 1920 table for a comparison of their skeletal sample from the Tabor Hill ossuaries. The distribution of ages at death from their ossuary material matched that recorded for France during the period 1866-1877, when the documented death rate was 2.46 per hundred. Allowing for a few missing infants, they rounded the figure to 3.0 per 100 and calculated the size of the population contributing to the ossuary. Assuming a constant rate of deaths, the crude mor- tality rate can be calculated directly from the life table by the formula M=i e° where M is the crude mortality rate and e° is life expect- ancy at birth (Acsadi and Nemeskeri, 1970:67). The ap- plication of this formula to the ossuary data produces a crude death rate of 47.8 per thousand for Population I and 43.5 per thousand for Population II. Table 42 com- pares these death rates with those calculated from other skeletal populations. The table shows that Nanjemoy death rates were lower than those of the protohistoric Arikara populations, the archaic Indian Knoll sample, and Angel's Middle Bronze age populations from Lerna, Greece, but higher than Hooton's Madisonville, Ohio, sample and Churcher and Kenyon's Huron population. Again the comparisons are of limited value since much of the variation probably reflects differences in the com- pleteness of the samples. Certainly the low death rate offered for the Tabor Hill ossuaries is influenced by the loss of infants. Consequently, perhaps the only really meaningful comparisons are those between the two ossu- aries from Nanjemoy Creek. Population Reconstruction The calculation of the crude death rate not only permits comparison of mortality rates, but also provides data that can be utilized to reconstruct population size. Since the mortality rate equals the number of individuals dying per thousand per annum, a knowledge of the total number TABLE 41.—Correlation of ages at death with known annual death rates (after Hooton, 1920:21) Place Per od Age 0-10 % 10-20 % 20+ % Period Aver, annual death rate Italy 1872-77 1866-77 1860-70 1875-77 1871-77 1865-77 1865-70 1870-74 1873-77 52 32 44 52 52 52 51 62. 36. 37 28 23 43 61 38 86 33 94 4.22 4.25 4.56 3.51 2.22 4.05 4.37 4. 13 3.72 43.41 63.47 51.21 44.06 45. 17 43.57 43. 77 33.54 59.33 1865-78 1865-77 1865-78 1865-78 1865-78 1865-78 1865-70 1865-75 1870-78 2 2. 2. 2. 3. 3. 3. 3. 2. 99 France 46 England 20 Prussia 72 Bavaria 09 Austria 18 Spain 12 Russia 67 Switzerland 38 525-413 O - 74 66 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 8 TABLE 42.—Death rate estimates from eight skeletal populations Site Location Death rate Madisonville Ohio 30 Tabor Hill Ontario 30 Nanjemoy, Ossuary II Maryland 44 Nanjemoy, Ossuary I Maryland 48 Sully South Dakota 54 Lerna Greece 56 Indian Knoll Kentucky 59 Leavenworth South Dakota 63 of dead in the ossuary and the length of time represented by the ossuary can lead to an accurate determination of the size of the contributing population. This calculation may be represented by the formula « 1000 N MT where P is population size, N is the number of individuals in the ossuary, M is the mortality rate expressed in the number of deaths per annum per 1000 of the population, and T is the number of years represented by each ossuary. The values of all of these figures except T have been determined for the ossuaries. The length of time represented by a single mid-Atlantic ossuary can be estimated from the different degrees of bone articulation, observed during excavation. A discussion of the three stages of articulation that were observed during the excavation of Ossuary II is given in detail on pp. 28-30: (1) completely disarticulated bones; (2) partially-articu- lated skeletal parts; and (3) completely-articulated skele- tons. Undoubtably, the degree of articulation directly re- flects the length of time the skeleton remained in the primary repository between the time of death and time of ossuary burial. Consequently, the ratio of individuals represented by disarticulated bones to individuals represented by par- tially articulated skeletal parts should increase as the length of time between ossuary deposits increases. Of course, the ratio of partially articulated individuals to completely articulated individuals would remain constant as long as the death rate and decomposition rate remains the same. Table 6 summarizes the number of adults represented by partially articulated skeletal parts in Ossuary II. At least 23 adults had articulated foot bones, followed by 20 adults with articulated tibiae and fibulae. These data indicate that approximately 20 percent of the 99 adults in the ossuary had died soon enough before the ossuary burial that their lower leg bones were still held together by ligaments. Con- sequently, if both the death rate and decomposition rate remained constant, then the length of time represented by the partially articulated bones would be approximately 20 percent of the length of time represented by the entire os- suary. This relationship can be expressed by the formula T = .42 t, where T is the length of time (in years) repre- sented by the entire ossuary and t is the length of time (in months) represented by the partially articulated lower leg bones. Table 43 presents a series of ossuary time intervals as calculated from different rates of decomposition using this formula. The table shows that if total decomposition of all soft parts was accomplished within one month after death, then the entire ossuary would represent less than a six month period of accumulation. At the other extreme, if two years were required for the decomposition of soft parts, then the ossuary would represent about a 10-year accumulation. Unfortunately, there are few data on the rate of body tissue decomposition, especially for a prehistoric scaffold situation. T. K. Marshall (1968) notes that decomposition rates can be quite variable depending upon such factors as climate, protection of clothing, and accessibility to scav- engers, rodents, and insects. Marshall claims that in a very dry environment, a body may become partially mummified and remain in that state for years. In contrast, a body left on the ground surface in damp climates may be reduced to bones in several weeks. Both T. D. Stewart and J. L. Angel (pers. comm.) have observed medical-legal cases where a body left exposed on the ground was reduced to bone in less than one month, especially when it was accessible to scavenging birds and mammals. Bodies placed by the mid- Atlantic Indians in death houses or scaffolds may have been protected from many scavengers, but bacteria, maggots, and perhaps birds would insure that decomposition and deflesh- ing would proceed rapidly. Thus, even allowing for deaths during the winter months, it is doubtful that lower-leg bone TABLE 43.—A comparison of possible time intervals (T) repre- sented by Ossuarv II, assuming different rates of decomposition (t), as calculated by the formula T = .42 t Decomposition rate "t" (possible no. of months required for lower leg bones to become disarticu- lated) Time interval "T" (corresponding no. of years represented by total accumulation) 1 4 2 .8 3 1. 3 4 I. 7 5 2. 1 6 2.5 7 2. 9 8 3.4 9 3.8 10 4. 2 11 4.6 12 5.0 18 7.6 24 10.0 DEMOGRAPHIC PROFILES AND POPULATION SIZE 67 articulation could have been maintained longer than eight months after death. This figure implies that the entire os- suary represents approximately a 3-year accumulation. Of course, this is only an approximation and the exact intervals cannot be determined until more data on decomposition rates are available for the mid-Atlantic area. However, the actual interval was certainly less than the 10-year interval documented for the Huron of Canada. The 20 individuals in Ossuary II with articulated lower legs would have to reflect a 2-year accumulation of the dead for the entire ossuary to represent a Huron-like interval of 10 years. It seems unlikely that the bones of the feet and lower legs could remain articulated after two years of exposure on a scaffold or in a death house above ground. More probably, articulation was maintained only about 8 months, suggesting the time interval of only 3 years. Although this suggested 3-year interval is considerably less than that documented for the Huron, it is compatible with intervals suggested for Algonquian groups in the Great Lakes region. In 1709, Raudot (1940: 368) mentioned that the Algonquian feast of the dead was held every 3 years. In contrast, Hickerson (1960: 88) believed that the event was annual with seven or eight distinct groups participating. He interpreted the accounts of Perrot (Blair, 1911: 88), Radisson (Scull, 1943: 199), and Beschefer (Thwaites, 1896- 1901, LXII: 201) to indicate that the seven or eight Algonquian groups communally buried their dead, and rotated the place of ossuary burial so that a particular group would host the ceremony once every 7 or 8 years. Table 44 gives a series of population estimates for each of the two Nanjemoy ossuaries, calculated from different theoretical time intervals by the formula 1000 N MT Those from Ossuary I range from 5481, assuming a time interval of 6 months, to 183, assuming a time interval of TABLE 44.—A comparison of possible population sizes assuming different time intervals, as calculated from Ossuary II by the , „ 1000 N juimuiu x ^ Time interval (T) Population size (P) Ossuary I Ossuary II . 5 5481 8644 1.0 2741 4322 2.0 1370 2161 3.0 914 1441 4.0 685 1080 5.0 548 864 6.0 460 720 7.0 392 617 8.0 343 540 9.0 305 480 10.0 274 432 15.0 183 288 De° TABLE 45.—A comparison of possible population sizes assuming different time intervals, as calculated from Ossuary II by the formula P=K-f Time interval (T) Population size (P) Ossuary I Ossuary II 0.5 5518 8641 1.0 2759 4320 2.0 1380 2160 3.0 920 1440 4.0 690 1080 5.0 552 865 6.0 460 721 7.0 397 618 8.0 346 541 9.0 307 481 10.0 277 433 15.0 185 290 15 years; whereas those for Ossuary II range from 8644, for a 6-tnonth interval to 288 for a 15-year interval. The time interval of about 3-years suggested by the partial- articulation data from Ossuary II indicates a size of 914 for Population I and 1441 for Population II. Population size can be estimated also by the formula P=K+^p where P is the average population size, D is the total number of dead, eg is life expectancy at birth, T is the time interval, and K is a correction factor equal to 10 percent of the value of T (Acsadi and Nemeskeri, 1970: 65). The application of this formula to the ossuary material produced the population estimates in Table 45 for different values of T. A comparison of Table 44 with Table 45 shows that population estimates calculated from the two formulae are nearly the same. It appears unlikely that the entire population contributing to either of the Nanjemoy ossuaries could have occupied the habitation site adjacent thereto. Preliminary testing by T. D. Stewart revealed that cultural debris (pottery, shell, animal bone, and lithic material) was concentrated within an area of approximately 106 square meters (1000 ft2), extending 50 meters (150 ft) west from the Juhle residence (Figure 4). If all of the population contributing to Ossuary II resided in that area, a population density of over eleven persons per square meter (1.2 per ft2) would be suggested. Obviously, such overcrowding was unlikely, and hence much of the pop- ulation must have lived elsewhere. It is possible that the occupation area is actually much larger than revealed by Stewart's testing; however, only exhaustive excavation can determine the actual extent. At this time, it appears more probable that several settlements in addition to the village at 18CH89 buried their dead in the ossuaries. John Smith's map of 1612 (Arber, 1884:384-385) lists 28 villages in the area occupied historically by the Conoy. 68 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 8 This area consists of the part of southwestern Maryland, south of the present city of Washington, D.C., that is bor- dered on die west by the Potomac River and on the east by the Chesapeake Bay (Figure 1). On Smith's map, 5 of these 28 villages (each hereafter referred to as a "chief's village") are accompanied by a symbol which indicates they contained the house of a chief, in addition to the houses of common people. The remaining 23 villages (each hereafter referred to as a "common village") lack such symbols and therefore contained only houses of common people. The map suggests a ratio of approximately one chief for five to six villages. Consequently, if we assume that the Nanjemoy ossuaries represent the communal dead of the five or six villages gov- erned by a single chief, then each village would contain an average of about 235 people. Such population estimates compare favorably with Smith's counts of 20 to 100 warriors for the villages in that area. The occurrences of ossuaries associated with many of the villages in the area may reflect a Huron-like practice of rotating the site of ossuary burial from village to village. Regardless of whether a number of villages were repre- sented by the ossuaries, a comparison of the population estimates derived from Ossuaries I and II suggests the exist- ence of temporal demographic oscillations. Unfortunately, the associated cultural artifacts are not sufficiently diagnostic to establish which of the ossuaries is the older. The ce- ramics indicate only that both ossuaries date from just prior to European contact. Whatever the temporal ordering, the population represented by Ossuary II was 58 percent larger than that of Ossuary I. Also, it exhibited a greater life ex- pectancy at birth, a greater longevity, a lower over-all mor- tality rate, and a higher adolescent mortality rate. If Ossuary I preceded Ossuary II, then the data indicate a rapidly expanding population with increases in longevity, adolescent mortality, life expectancy at birth and decreases in over-all mortality. If Ossuary II is older, then the evidence would suggest the opposite situation, namely, a rapidly declining population, with decreases in longevity, adolescent mortality and life expectancy, and an increase in over-all mortality. Be this as it may, Charles MacNett of American University (pers. comm.) indicates in his archeological study of Late Woodland settlement patterns that prior to 17th-century European contact aboriginal mid-Atlantic populations were expanding rapidly in response to improved agricultural tech- niques and more efficient food procurement systems. Of course, with the advent of cultural conflict and the introduc- tion of new diseases following initial European contact, ab- original populations decreased markedly in numbers. Since the Nanjemoy ossuaries lack European-manufactured trade materials, however, they probably represent pre-contact pop- ulations. I think it is reasonable to believe therefore, that they were enjoying the population expansion and increased longevity indicated when Ossuary I is assumed to be earlier than Ossuary II. Of course, the size difference between ossuaries also could reflect a change in the size of the social- political unit contributing to the ossuaries. Regional Population Estimates The question may now be asked: What bearing do these local population estimates have on the population estimates of the total tidewater Potomac area at the time of Euro- pean contact? Most of the frequently quoted figures for the latter area can be traced to James Mooney. Although he used figures on the tidewater Potomac area as part of the basis of his total North American Indian population estimate (Mooney, 1910), his more detailed account was not pub- lished until after his death (Mooney, 1928). In the preface, John R. Swanton cautioned that the figures had not been arranged in final tabulation at the time of Mooney's death, but he expressed confidence in their accuracy because "it is known that, in some cases, he [Mooney] carried his inves- tigations back to the original census rolls" (Mooney, 1928:2). Kroeber (1939) likewise showed confidence in Mooney's estimates by using them almost without alteration for his own purpose. Mooney estimated the total aboriginal population of Vir- ginia to be 15,100, but that of Maryland and Delaware to be only 4700, with the Conoy or Piscataway of southwestern Maryland contributing a mere 2000. Although the exact methods employed by Mooney to produce these figures have never been published, Mook (1944) detected a proportional relationship between Smith's Virginia warrior counts and Mooney's total Virginia population estimates. This sug- gested to him that Mooney's Virginia estimates might have been produced by multiplying Smith's warrior counts for each village by a factor of about 4.5 to account for the number of nonwarriors in the total population. The factor of 4.5 was derived by Mook by dividing Mooney's total population estimates by Smith's total warrior counts. Following Mooney's death, all of his unpublished manu- scripts, notes, and miscellaneous data became a part of the Bureau of American Ethnology Archives and recently were transferred to the National Anthropological Archives in the National Museum of Natural History. Among these records are several scraps of paper which contain Mooney's original population calculations for Maryland. These documents demonstrate that Mooney's estimates for the Conoy were derived by considering John Smith's warrior counts for vil- lages within the area occupied by the Conoy historically. The following warrior counts per village were listed by Mooney: Cecomocomoco, 40; Potapaco, 20; Pamacocack, 60; Moyaons, 100; and Nacotchtank, 80. He then added 200 more warriors from the villages Mattpament, Pawtuxnet, and Acquintanactuck to obtain a total of 500 warriors. All of these figures agree with Smith's estimates. The warrior counts were then multiplied by 4 producing a total popula- tion of 2000 for the Conoy. Mooney's original figures show DEMOGRAPHIC PROFILES AND POPULATION SIZE 69 an addition of 400 more individuals, apparently to account for numerous other villages shown on Smith's map, but not mentioned in his discussion of warrior counts. Subsequently, Mooney must have ignored these other villages, since his pub- lished estimate for the Conoy is only 2000. If Mooney's estimates are incorrect, then the error prob- ably originates from two sources: (1) his use of the 1:4 ratio of warriors to the entire population, or (2) his failure to include estimates of all of the villages on Smith's map. Mooney's notes do not disclose the source of his 1:4 ratio. It approaches, however, the ratio implied by John Smith (in Arber, 1884:65), who wrote, "the land is not populous, for the men be fewe; their far greater number is of women and children. Within 60 miles of James Towne there are about some 5000 people, but of able men fit for their warres scarse 1500." Smith's estimate of 1500 warriors in a popula- tion of 5000 represents a ratio of 3 to 10. Others have of- fered such ratios as 1:10 (McKenney, in Graham, 1935:4), 3:10 (Jefferson, 1787:139), and 1:5 (Proud, 1797-1798, 11:297). The most accurate determination of the ratio of warriors to overall population can be made directly from the skeletons. The life tables reconstructed from the Nanjemoy ossuaries specify the number of survivors entering each age category. The number of warriors would equal approximately one-half of the number of survivors between ages 15 and 40. Age 15 would approximate the lower age limit since at about that age, boys underwent "huskanaw," the rite of passage from boyhood to manhood. According to Beverly (1705: ch. 6:41) at that time they "unlive their former lives, and commence Men, by forgetting that they ever have been Boys." The upper age limit of 40 indicates the approximate age at which most men would terminate active warrior activities, although the age could have been even lower and certainly varied with each individual. An acceptance of these age boundaries, would place the percentage of potential "warriors" at ap- proximately 24 percent of Population I and 23 percent of Population II. The ratio of warriors to overall population would then be 1:4.12 for Ossuary I and 1:4.38 for Ossuary II, suggesting that Mooney's ratios were the most precise of all. The substitution of either of the calculated values of 4.12 or 4.38 for Mooney's factor of 4.0 would not significantly alter his overall population estimate. Although Mooney's 1:4 ratio appears to be nearly correct, his population estimate for the Conoy is overly conservative because he did not include all of the villages in the area. Mooney's estimates were derived directly from John Smith's warrior counts for only eight villages, although Smith's map of 1612 (Arber, 1884:384-385) shows 20 additional villages in that same area. Since none of these 20 additional villages was a "chief's village," Mooney may have felt that they were too small to add significantly to the overall estimate. Smith lists 60 warriors for the "common village" of Pamacocack, however, whereas as few as 40 warriors were recorded for the "chief's village" of Cecomocomoco. Clearly therefore, all of the villages on the map need to be considered. Warrior estimates are available for only two of the 23 "common villages," Pamacocack with 60 and Potapaco with 20. This suggests an average of 40 warriors per "common village." The application of Mooney's own factor of 4 pro- duces an average population size of 160 for the "common villages" and indicates that the 20 "common villages" not included in Mooney's estimate represent about 3200 people. Consequently, the inclusion of these villages raises Mooney's estimate to 8400. A comparable crude regional population estimate can be generated from the ossuary data. If the length of time between ossuary deposits was approximately 3 years, as sug- gested by the partial-articulation data, then a population of about 1440 would have contributed to Ossuary II. If the hypothesis discussed earlier is correct, Ossuary II represents the collective dead from five or six villages governed by a single "chief." If these political units were approximately of equal size, then the presence on Smith's map of five villages with "chiefs" would suggest a total of prehistoric Conoy population of 7200. In conclusion, the ossuary analysis and John Smith's map both suggest that Mooney's population estimate of 2000 for the Conoy was much too conservative. The actual number must have been at least 7000 and perhaps much higher. A total population size of 8400 (suggested by Smith's map) implies a population density of over 1.2 persons per square kilometer. Kroeber, using Mooney's population estimates, suggested a density of 0.13 person per square kilometer. Not only is Kroeber's population estimate for the Conoy (2700) too conservative, however, but he considers the area inhabited by the Conoy to be 20,100 km2 My calculations indicate that the area was closer to 6750 km2. The higher population density of 1.2 persons per square kilometer is substantially higher than that calculated by Kroeber for either the east coast (.07 person per km2) or for all of North America north of Mexico (.05 person per km2), but below the figure offered by Dobyns (1966a) for the entire western hemisphere (2.1 persons per km2). Although such figures do not necessarily mean that Mooney's tribal estimates for all of North America were too conservative, they do indi- cate the need for further review of his estimates, through similar detailed studies in local areas. Within the mid-Atlantic area, future ossuary excavations should aim to discover (1) whether the length of time be- tween ossuaries was constant in all areas; (2) whether the population size differences documented in this study repre- sent population expansion or merely changes in the organi- zation of the social unit contributing to the ossuary; (3) 70 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 8 the extent that European contact affected the demographic formation, such investigations should fill many gaps in our profiles of the aboriginal population; and (4) whether the knowledge about prehistoric mortuary practices, social sys- bone distributions observed within Ossuary II were unique terns, and demographic-environmental adaptation in the or represent a cultural pattern. Given this additional in- mid-Atlantic region. Literature Cited Acsadi, Gy, and T. Nemeskeri 1957a. Palaodemographische Probleme am Beispiel des Friimit- telalterlichen Graberfeldes von Halimba—Cseres. Homo, 8:133-148. 1957b. Contributions a la Reconstruction de la Population de Veszprem, X° et XIe Siecles. Annates Historico-Natu- rales Musei Nationalis Hungarici, 8:435-467. 1958. La Population de la Transdanubie Nord—Est, X° et XP Siecles. Annates Historico-Naturales Musei Nationalis Hungarici, 50:359-415. 1959. Analyses des Trouvailles Anthropologiques du Cimetiere de Kerpuszta (XF Siecle) sous L'aspect de L'age. Acta Archaeologica Academiae Scientiarum Hungaricae, 11:419-455. Budapest. I960. Le Population de Szekesfehervar X" et XIe Siecles. An- nates Historico-Naturales Musei Nationalis Hungarici, 52:481-495. 1970. History of Human Life Span and Mortality. Budapest: Akademiai Kiado. Adams, R. G. 1890. Notes on the Delaware Indians: Report on Indians Taxed and Not Taxed. Washington: U.S. Census. Ahlqvist, J., and O. Damsten 1969. A Modification of Kerley's Method for the Microscopic Determination of Age in Human Bones. Journal of Foren- sic Sciences, 14:205-212. Allan, W. 1965. The African Husbandman. New York: Barnes and Noble. Anderson, J. E. 1964. The People of Fairty. National Museum of Canada Bul- letin, 193. Ottawa. Anderson, M., and W. T. Green 1948. Lengths of the Femur and Tibia. American Journal of Diseases of Children, 75:279—290. Angel, J. L. 1947. The Length of Life in Ancient Greece. Journal of Geron- tology, 2:18-24. 1948. Health and the Course of Civilization. Interne, 15-18. New York. 1951a. Population Size and Microevolution in Greece. Cold Spring Harbor Symposium on Quantitative Biology, 15:343-351. Long Island Biological Association. 1951b. Troy: The Human Remains. Supplementary Monograph I. Princeton: Princeton University Press for University of Cincinnati. 1953. The Human Remains from Khirokitia. Appendix 2 in P. Dikaios, editor, Khirokitia. London: Oxford University Press. 1954. Human Biology: Health and History in Greece from First Settlement until Now. Yearbook of the American Philosophical Society, 98: 168-174. 1957. Human Biological Changes in Ancient Greece. In Year- book of the American Philosophical Society, 98:266-270. 1959. Early Helladic Skulls from Aghios Kosmas. Appendix in Aghios Kosmas. G. E. Mylonas, Princeton: Princeton University Press. 1968. Ecological Aspects of Paleodemography. Pages 263-270 in D. R. Brothwell, editor, The Skeletal Biology of Earlier Human Populations. Symposium for the Study of Human Biology VIII. London: Oxford University Press. 1969a. The Bases of Paleodemography. American Journal of Physical Anthropology, 30(3) : 427-437. 1969b. Paleodemography and Evolution. American Journal of Physical Anthropology, 31(3): 343-353. Arber, E., editor 1884. Capt. John Smith, Works, 1608-1631. The English Scholar's Library, 16. Birmingham. 1910. Travels and Works of Captain John Smith, 1580-1631. Edinburgh: J. Grant. Archives of Maryland 1887. Proceedings of the Council of Maryland, 1667-1687/8. Baltimore: Maryland Historical Society. 1896. Proceedings of the Council of Maryland, 1676-1678. Baltimore: Maryland Historical Society. Ascher, R. 1959. A Prehistoric Population Estimate using Midden Analysis and Two Population Models. Southwestern Journal of Anthropology, 15:168-178. Aschman, H. 1959. The Central Desert of Baja, California: Demography and Ecology. lbero-Americana, 42. Berkeley. Bailit, H., and E. E. Hunt, Jr. 1964. The Sexing of Children's Skeletons from Teeth Alone and its Genetic Implications. American Journal of Physi- cal Anthropology, 22(2) : 171-173. Bass, W. M., D. R. Evans, and R. L. Jantz 1971. The Leavenworth Site Cemetery: Archaeology and Phys- ical Anthropology. University of Kansas Publications in Anthropology, 2. Lawrence. Baudin, L. 1961. A Socialist Empire: The Incas of Peru. Translated by Katherine Woods, edited by A. Goddard. Princeton: D. Van Nostrand Co. [Originally published 1928 under the title L'Empire socialiste des Inka. Paris.] Bauer, W. H. 1944. Tooth Buds and Jaws in Patients with Congenital Syphilis: Correlation Between Distribution of Treponema Pallidum and Tissue Reaction. American Journal of Pathology, 20:297-314. Beloch, K.J. 1886. Die Bevolkerung der Griechischromischen Welt. Leipzig. Bennett, J. W. 1966. Comment on Estimating Aboriginal American Population. Current Anthropology, 7(4) : 425-426. Bernal, I. 1966. Comment on Estimating Aboriginal American Population. Current Anthropology, 7(4) :426. 71 72 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 8 Beverly, R. 1705. The History and Present State of Virginia. London. Biggar, H. P. 1929. The Works of Samuel de Champlain. Volume 3. Toronto: The Champlain Society. Blair, E. H. 1911. The Indian Tribes of the Upper Mississippi Valley and Region of the Great Lakes as Described by Nicolas Perrot, French Commandant in the Northwest. 2 volumes. Cleve- land: The Arthur H. Clark Co. Blakely, R. L. 1971. Comparison of the Mortality Profiles of Archaic, Middle Woodland, and Middle Mississippian Skeletal Populations. American Journal of Physical Anthropology, 34(1): 43-53. Blakely, R. L., and P. L. Walker 1968. Mortality Profile of the Middle Mississippian Population of Dickson Mound, Fulton County, Illinois. Proceedings of the Indiana Academy of Science, 1967, 77:102-108. Blasi, O. 1966. Comment on Estimating Aboriginal American Population. Current Anthropology, 7(4) : 426-427. Borah, W. 1962a. Population Decline and the Social and Institutional Changes of New Spain in the Middle Decades of the Sixteenth Century. Internationalom Amerikanisten- Kongresses, 34:172-178. 1962b. America Como Modelo ? El Impacto Demografico de la Expansion Europea Sobre el Mundo no Europeo. Cuadernos Americanos, 6:176-185. [Reprinted in English in Actasy Memorias, XXXV Congreso Internacional de Americanistas, Mexico, 1962, 3: 379-387.] Borah, W., and S. F. Cook 1963. The Aboriginal Population of Central Mexico on the Eve of the Spanish Conquest. Ibero-Americana, 45. Berkeley. Boucher, B. F. 1955. Sex Differences in the Foetal Sciatic Notch. Journal of Forensic Medicine, 2:51-54. 1957. Sex Differences in the Foetal Pelvis. American Journal of Physical Anthropology, 15(4) -.581-600. Bozman, J. L. 1837. The History of Maryland. Volume 1. Baltimore: James Lucas and E. K. Deaver. Brainerd, G. S. 1956. The Ancient Maya. Palo Alto: Stanford University Press. Brinton, D. G. 1885. The Lenape and Their Legends. Philadelphia: Brinton's Library of Aboriginal American Literature, 5. Broste, K., and J. B. Jorgensen 1956. Prehistoric Man in Denmark: A Study in Physical An- thropology: I, Stone and Bronze Ages. Copenhagen: E. Munksgaard. Bullard, W. R. Jr. 1960. Maya Settlement Pattern in Northeastern Peten, Guate- mala. American Antiquity, 25:355-372. Bushnell, D. I 1920. Native Cemeteries and Forms of Burial East of the Missis- sippi. Bureau of American Ethnology Bulletin, 71. Wash- ington, D.C. Churcher, C. S., and W. A. Kenyon 1960. The Tabor Hill Ossuaries: A Study in Iroquois Demog- raphy. Human Biology, 32: 249-273. Coe, M. D., and K. V. Flannery 1967. Early Cultures and Human Ecology in South Coastal Guatemala. Smithsonian Contributions to Anthropology, 3. Washington, D.C. Colton, H. S. 1936. The Rise and Fall of the Prehistoric Population of North- ern Arizona. Science, 84 (2181) : 337-343. 1939. Prehistoric Culture Units and Their Relationships in Northern Arizona. Museum of Northern Arizona Bulletin, 22. Flagstaff. Cook, S. F. 1937. The Extent and Significance of Disease Among the Indians of Baja California, 1697-1773. Ibero-Americana, 12. Berkeley. 1939. Smallpox in Spanish and Mexican California, 1770-1845. Bulletin of the History of Medicine, 7 :153-191. 1940. Population Trends Among the California Mission Indians. Ibero-Americana, 17. Berkeley. 1943. The Conflict Between the California Indians and the White Civilization: 1, The Indian Versus the Spanish Mission. Ibero-Americana, 21. Berkeley. 1946a. A Reconsideration of Shellmounds with Respect to Popu- lation and Nutrition. American Antiquity, 12:51-53. 1946b. Human Sacrifice and Warfare as Factors in the Demog- raphy of Precolonial Mexico. Human Biology, 18:81-102. 1955a. The Epidemic of 1830-1833 in California and Oregon. University of California Publications in American Archae- ology and Ethnology, 43:303-326. Berkeley. 1955b. The Aboriginal Population of the San Joaquin Valley, California. Anthropological Records, 16(2) :31—78. 1956. The Aboriginal Population of the North Coast of Cali- fornia. Anthropological Records, 16(3) :81—129. 1958. Santa Maria Ixcatlan Habitat, Population, Subsistence. Ibero-Americana, 41. Berkeley. 1966. Comments on Estimating Aboriginal American Popula- tion. Current Anthropology, 7(4) :427-429. Cook, S. F., and W. Borah 1957. The Rate of Population Change in Central Mexico, 1550-1570. Hispanic American Historical Review, 37:463-470. Cook, S. F., and L. B. Simpson 1948. The Population of Central Mexico in the Sixteenth Cen- tury. Ibero-Americana, 31. Berkeley. 1950. The Quantitative Investigation of Indian Mounds with Special Reference to the Relation of the Physical Com- ponents to the Probable Material Culture. University of California Publications in American Archaeology and Ethnology, 40:223-262. Berkeley. Cowgill, U. M. 1962. An Agricultural Study of the Southern Maya Lowlands. American Anthropologist, 64:273-286. Crovello, T. J., and C. S. Hacker 1972. Evolutionary Strategies in Life Table Characteristics Among Feral and Urban Strains of Aedes aegypti (L). Evolution, 26(2): 185-196. Cunha, A. X. 1956. Contribuicao para a Antropologia dos povos de Cultura Campaniforme em Portugal. Estudios Anthropologia Por- tugal, 123-137. Dahlberg, A. A., and R. M. Menegaz-Bock 1958. Emergence of the Permanent Teeth in Pima Indian Chil- dren: A Critical Analysis of Method and an Estimate of Population Parameters. Journal of Dental Research, 37:1123-1140. LITERATURE CITED 73 Davidson, D. S. 1935. Burial Customs in the Delmarva Peninsula and the Ques- tion of their Chronology. American Antiquity, 1(1): 84— 97. Dechaume, M., L. Derobert, and J. Payen 1960. De la valeur de la determination de 1'age par l'examen des dents en coupes minces. Annates de Medecine Legale, 40:165-167. Paris. Deevey, E. S. 1947. Life Tables for Natural Populations of Animals. Quarterly Review of Biology, 22(4) : 283-314. Baltimore. Denevan, W. M. 1966. Comment on Estimating Aboriginal American Popula- tion. Current Anthropology, 7(4) :429. Dobyns, H. F. 1962. Pioneering Christians Among the Perishing Indians of Tucson. Lima: Editorial Estudios Andinos. 1963a. Indian Extinction in the Middle Santa Cruz River Valley, Arizona. New Mexico Historical Review, 38:163—181. 1963b. An Outline of Andean Epidemic History to 1720. Bulle- tin of the History of Medicine, 37:493-515. 1966a. Estimating Aboriginal American Population: An Ap- praisal of Techniques with a New Hemispheric Estimate. Current Anthropology, 7(4) : 395-416. 1966b. Reply to Comments on Estimating Aboriginal Population Numbers. Current Anthropology, 7(4) : 440-444. Dublin, L. I., and A. J. Lotka 1963. Length of Life: A Study of the Life Table. New York: Ronald Press. Dublin, L. I., and M. Spiegelman 1941. Current Versus Generation Life Tables. Human Biology, 13:439-458. Dumond, D. E. 1972. Demographic Aspects of the Classic Period in Puebla- Tlaxcala. Southwestern Journal of Anthropology, 28(2) : 101-130. Dwight, T. 1904. The Size of the Articular Surfaces of the Long Bones as Characteristic of Sex: An Anthropological Study. American Journal of Anatomy, 4:19—31. Edmondson, W. T. 1945. Ecological Studies of Sessile Rotatoria, Part II: Dynamics of Populations and Social Structures. Ecological Mono- graphs, 15:141-172. Durham. Enlow, D. H. 1963. Principles of Bone Remodeling: An Account of Post-natal Growth and Remodeling Processes in Long Bones and the Mandible. Springfield: Charles C. Thomas. Etienne, R. 1955. A propos de la demographie de Bordeaux aux trois premiers siecles de notre ere. Revue Historique de Bor- deaux et du Department de la Gironde, 4:189-200. Evans, C. 1955. A Ceramic Study of Virginia Archaeology. Bureau of American Ethnology Bulletin, 160. Washington, D.C. Fanning, E. A. 1961. A Longitudinal Study of Tooth Formation and Root Resorption. New Zealand Dental Journal, 57:292-217. 1962a. Effect of Extraction of Deciduous Molars on the Forma- tion and Eruption of Their Successors. Angle Orthodon- tist, 32:44-53. 1962b. The Relationship of Dental Caries and Root Resorption of Deciduous Molars. Archives of Oral Biology, 7:595- 601. Fenton, W., and G. Kurath 1951. The Feast of the Dead or Ghost Dance at Six Nations. Reserve, Canada. Pages 139-165 In Symposium on local Diversity in Iroquois Culture. Bureau of American Ethnology Bulletin, 149. Washington, D.C. Ferembach, D. 1960. Les hommes du mesolithique d'Afrique du Nort et le probleme des isolats. Sociedade Portuguesa de Ciencias Naturais, 8 (2a). Lisbon. 1961. Quelques Facteurs de Formation et de Developpement d'un Isolat. Population, 16:71-90. 1962. La Necropole Epipaleolithique de Taforalt: Etude des Squelettes Humains. Centre National de la Recherche Scientifique et de la Mission Universitaire et Culturelle Franchise au Maroc, oriental. Ferguson, A. L. L. 1937. Burial Area in Moyaone. Journal of the Washington Acad- emy of Sciences, 27(1):261—267. Fisher, I. 1923. Report on National Vitality. National Conservation Com- mission, Washington, D.C. Forbis, R. G. 1966. Comment on Estimating Aboriginal American Population. Current Anthropology, 7 (4) :431. Fowke, G. 1894. Archeological Investigations in James and Potomac Valleys. Bureau of American Ethnology Bulletin, 23 Washington, D.C. Franz, L., and W. Wingler 1936. Die Sterblichkeit in der fruhen Bronze zeit Niederoster- reichs. Zeitschrift fur Rassenkunde, 4:157-163. Fuste, M. 1954. La Duracion de la Vida en la Poblacion Espanola Desde de la Prehistoria Hasta Nuestros Dias. Trabajo del Inst. Bernardino de Sahagun de Anthropologia y Ethnologia, 14:81-104. Barcelona. 1955. La Duracion de la Vide en las Poblaciones Humanas del Levante Espanol Durante el Periodo Neo-eneolitico. Pages 325-333, in D. Francisco Pardillo Vaguer, Homenaje Postumo. Barcelona. Garn, S. M., A. B. Lewis, and R. M. Blizzard 1965. Endocrine Factors in Dental Development. Journal of Dental Research, 43: 243-258. Garn, S. M., A. B. Lewis, and R. S. Kerewsky 1964. Sex Differences in Tooth Size. Volume 12: 125-126 in S. T. Genoves, editor, Yearbook of Physical Anthropology. Cordoba, Mexico: Instituto Nacional de Anthropoliga Historia. Garn, S. M., A1. B. Lewis, and D. L. Polacheck 1959. Variability of Tooth Formation. Journal of Dental Re- search, 38:135. Garn, S. M., and C. F. A. Moorees 1951. Stature, Body Build and Tooth Emergence in Aleutian Aleut Children. Child Development, 22:261-270. Gejvall, N. G. 1960. Westerhus: Mediaeval Population and Church in the Light of Skeletal Remains. Lund: H. Ohlssons Boktr. Genoves, S. 1963. Estimation of Age and Mortality. Pages 353-364 in D. R. Broth well, editor, Science and Archaelogy. New York: Basic Books, Inc. Gerhardt, K. 1953. Die Glockenbeckerlcute in Mittel- und Westdeutschland. Stuttgart: E. Schweizerbart. 74 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 Gilbert, B. M. 1971. A Method for Aging the Female Os Pubis. PhD thesis, Department of Anthropology, University of Kansas, Law- rence, Kansas. Giles, E. 1964. Sex Determination by Discriminant Function Analysis of the Mandible. American Journal of Physical Anthropol- ogy, 22:129-135. 1970a. Sexing Crania by Discriminant Function Analysis: Effects of Age and Number of Variables. Proceedings, VIHth International Congress of Anthropological and Ethno- logical Sciences, 1:59-61. Tokyo. 1970b. Discriminant Function Sexing of the Human Skeleton. Pages 99-109 in T. D. Stewart, editor, Personal Identifi- cation in Mass Disasters. Washington, D.C. Giles, E., and O. Elliott 1963. Sex Determination by Discriminant Function Analysis of Crania. American Journal of Physical Anthropology, 21:53-68. Giot, P. R. 1951. Armoricains et Bretons, etude Anthropologiques. Travaux de I'lnstitut d'Anthropologic Generate de la Faculte des Sciences de Rennes, 1. Glasstone, S. 1938. A Comparative Study of the Development in vivo and in vitro of Rat and Rabbit Molars. Proceedings of the Royal Society of London, series B., 126:315-330. 1963. Regulative Changes in Tooth Germs Grown in Tissue Culture. Conference on Genetic Aspects of Oral Struc- tures, Denver, Colorado, 1961. Journal of Dental Re- search, 42(6):1364-1368. 1964. Cultivation of Mouse Tooth Germs in a Chemically De- fined Protein-free Medium. Archives of Oral Biology, 9:27-30. Glover, T. 1676. An Account of Virginia, Its Situation, Temperature, Pro- ductions, Inhabitants, and Their Manner of Planting and Ordering Tobacco. Oxford: B. H. Blackwell. Goldstein, M. S. 1953. Some Vital Statistics Based on Skeletal Material. Human Biology, 25:3-12. Gomme, A. W. 1933. The Population of Athens in the Fifth and Fourth Cen- turies B.C. Oxford: Basil Blackwood. Graham, W. J. 1935. The Indians of Port Tobacco River, Maryland, and Their Burial Places. Washington: Private printing. Green, R. G., and C. A. Evans 1940a. Studies on a Population Cycle of Snowshoe Hares on the Lake Alexander Area, I: Gross Annual Censuses, 1932— 1939. Journal of Wildlife Management, 4:220-238. 1940b. II. Mortality According to Age Groups and Seasons. Journal of Wildlife Management, 4:267-278. 1940c. III. Effect of Reproduction and Mortality of Young Hares on the Cycle. Journal of Wildlife Management, 4:347— 358. Griffin, J. B. 1956. Prehistoric Settlement Patterns in the Northern Mississippi Valley and the Upper Great Lakes. Viking Fund Pub- lications in Anthropology, 23:63-71. Gustafson, G. 1950. Age Determinations on Teeth. Journal of the American Dental Association, 41:45-54. Hall, E. T. Jr. 1942. Archeological Survey of the Walhalla Glades. Museum of Northern Arizona Bulletin, 20. Flagstaff. Hansen, G. 1953-1954. Die Altersbestimmung am Proximale Humerus- und Femurende in Rahmen der Identifizierung menschlicheu Skelettreste. Wissenschaftliche Zeitschrift der Humboldt- Universitat zu Berlin. Mathematish- Naturwissenschaft- liche Reihe, 3(1): 1-73. Hariot, T. 1590. Narrative of the First English Plantation of Virginia. London. Harkness, A. G. 1896. Age at Marriage and at Death in the Roman Empire. Transactions of the American Philological Association, 27:35-72. Boston. Harrington, M. R. 1921. Religion and Ceremonies of the Lenape. Indian Notes and Monographs, 19. New York: Museum of the Ameri- can Indian. Hatton, H. 1938. Essais de bionomie explicative sur quelques especes inter- cotidales d'algues et d'animaux. Annates Institut Oceano- graphique, 17:241-348. Paris. Hausler, A. 1966. Comment on Estimating Aboriginal American Population. Current Anthropology, 7(4) :433. Havilland, W. A. 1965. Prehistoric Settlement at Tikal, Guatemala. Expedition, 7:15-23. 1966. Maya Settlement Patterns: A Critical Review. Middle American Research Institute, Tulane University Publica- tions, 26:21-47. 1970. Family Size, Prehistoric Population Estimates and the Ancient Maya. American Antiquity, 37(1) : 135-139. Heckewelder, J. 1819. An account of the History, Manners and Customs of the Indian Nations Who Once Inhabited Pennsylvania and the Neighboring States. Philadelphia. [Reprinted in Memoires of the Historical Society of Pennsylvania, 12. Philadelphia, 1876.] Heizer, R. F., and M. A. Baumhoff 1956. California Settlement Patterns. Pages 32-44 in G. W. Wiley, editor, Prehistoric Settlement Patterns in the New World. Viking Fund Publications in Anthropology, 23. New York. Hickerson, H. 1960. The Feast of the Dead among the Seventeenth Century Algonkians of the Upper Great Lakes. American An- thropologist, 6(1):81—107. Hombert, M. and C. Preaux 1945. Note sur la duree de la vie dans l'Egypte Greco-Romaine. Chronique d'Egypt, 20:139-146. Hooten, E. A. 1920. Indian Village Site and Cemetery near Madisonville,, Ohio. Papers of the Peabody Museum, American Archaeology and Ethnology, 8. Cambridge. 1930. The Indians of Pecos Pueblo: A Study of their Skeletal Remains. New Haven: Yale University Press. Howells, W. W. 1960. Estimating Population Numbers through Archaeological and Skeletal Remains. Pages 158-185 in R. Heizer and and S. Cook, editors, The Application of Quantitative Methods in Archeology. Viking Fund Publications in Anthropology, 28. Chicago: Quadrangle Books. LITERATURE CITED 75 Hrdlicka, A. 1908. Physiological and Medical Observations among the Indi- ans of Southwestern U.S. and Northern Mexico. Bureau of American Ethnology Bulletin, 34. Washington, D.C. 1952. Practical Anthropometry. 4th edition, T. D. Stewart, editor. Philadelphia: The Wistar Institute of Anatomy and Biology. Huffiington, W. 1838. The Delaware Register. Dover. Hulbert, A. B., and W. N. Schwarze 1910. Zeisberger's History of Northern American Indians. Ohio Archaeological and Historical Quarterly, 19 Hunt, E. E. Jr., and I. Gleiser. 1955. The Estimation of Age and Sex of Preadolescent Children from Bones and Teeth. American Journal of Physical Anthropology, 13(3) :479-487. Hurme, V. O. 1948. Standards of Variation in the Eruption of the First Six Permanent Teeth. Child Development, 19:213-231. Ivanicek, F. 1951. Staroslavenska Nekropola u Ptunu. Rezultati antropolo- skish Istrazivanja, Ljubljana. Slovenska Akademija Znanosti in Umetnosti, Razredza Zgodouino in druzbene vede Dela, 5. Jefferson, T. 1787. Notes on the State of Virginia. London. Johnson, L. C. 1962. Joint Remodeling as the Basis for Osteoarthritis. Journal of the American Veterinary Medicine Association, 141: 1237-1241. Johnston, F. E. 1962. Growth of the Long Bones of Infants and Young Children at Indian Knoll. American Journal of Physical Anthro- pology, 20(3):249-254. Johnston, F. E., and C. E. Snow 1961. The Reassessment of the Age and Sex of the Indian Knoll Skeletal Population: Demographic and Methodological Aspects. American Journal of Physical Anthropology, 19(3):237-244. Jowsey, J. 1960. Age Changes in Human Bone. Clinical Orthopedics, 17:210-17. Kehoe, T. F., and A. B. Kehoe 1966. Comment on Estimating Aboriginal American Popula- tion. Current Anthropology, 7(4) :434-435. Kerley, E. R. 1965. The Microscopic Determination of Age in Human Bone. American Journal of Physical Anthropology, 23(2) :149- 163. 1969. Age Determination of Bone Fragments. Journal of Fo- rensic Sciences, 14(1) : 59—67. 1970. Estimation of Skeletal Age after about Age 30. Pages 57- 70 in T. D. Stewart editor, Personal Identification in Mass Disasters. Washington, D.C: Smithsonian Institu- tion. Kidd, K. E. 1952. Sixty Years of Ontario Archeology. Pages 71-82 in J. R. Griffin, Archaeology of Eastern United States. Chicago: University of Chicago Press. 1953. The Excavation and Historical Identification of a Huron Ossuary. American Antiquity, 18:359-379. Kidder, A. V. 1958. Pecos, New Mexico: Archaeological Notes. Papers of the R. S. Peabody Foundation for Archaeology, 5. Cambridge. Kivett, M. F. 1953. The Woodruff Ossuary: A Prehistoric Burial Site in Phil- lips County, Kansas. Bureau of American Ethnology Bul- letin, 154:109-142. Washington, D.C. Klebba, A. J. 1971. Leading Components of Upturn in Mortality for Men. Vital and Health Statistics, series 20, number 11. Wash- ington, D.C: U.S. Department of Health, Education, and Welfare. Kraus, B. S. 1959. Calcification of the Human Deciduous Teeth. Journal of the American Dental Association, 59:1128. Kroeber, A. L. 1939. Cultural and Natural Areas of Native North America. University of California Publications in American Archae- ology and Ethnology, 38. Berkeley. Krogman, W. M. 1939. A Guide to the Identification of Human Skeletal Mate- rial. FBI Law Enforcement Bulletin, 8(8) : 1-29. 1955. The Skeleton in Forensic Medicine. Graduate Medicine, 17 (2 and 3). 1962. The Human Skeleton in Forensic Medicine. Springfield, Illinois: Charles C. Thomas. Kronfeld, R. 1935. Development and Calcification of the Human Deciduous and Permanent Dentition. The Bur, 35:18. Kubler, G. 1942. Population Movements in Mexico, 1520-1600. Hispanic American Historical Review, 22:606-643. 1946. The Quechua in the Colonial World. Volume 2:331-410 in Julian H. Steward, editor, Handbook of South Ameri- can Indians. Bureau of American Ethnology Bulletin, 143. Washington, D.C. 1948. Mexican Architecture of the Sixteenth Century. New Haven: Yale University Press. Kunstadter, P. 1966. Comments on Estimating Aboriginal American Popula- tion. Current Anthropology, 7(4) :436-437. Kurth, G. 1955. Vorbericht Uber Anthropologische Beo Bachtungen Bei der Jericho-Grabung. Homo, 4:145-156. Lack, D. 1943. The Age of Some More British Birds. British Birds, 36: 193-197, 214-221. Lefroy, J. H. 1853. On the Probable Number of the Native Indian Popula- tion of British America. 1832-33. Toronto Canadian Journal, 1:1832-1833. Toronto. Lewis, A. D., and S. M. Garn 1960. The Relationship between Tooth Formation and Other Maturational Factors. Angle Orthodontist, 30:70-77. Logan, W. H. C, and R. Kronfeld 1933. Development of the Human Jaws and Surrounding Struc- tures from Birth to Age of Fifteen Years. Journal of the American Dental Association, 20:379. Macgowan, K, and J. A. Hester, Jr. 1965. Early Man in the New World. New York: Doubleday and Co. Inc. MacLeod, W. C 1928. The American Indian Frontier. New York: Alfred A. Knopf. MacNutt, F. A., editor 1909. Bartholomew de las Casas: His Life, His Apostolate and His Writings. New York: A. H. Clarke. 76 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 18 Maltby, J. R. D. 1917—1918. Some Indices and Measurements of the Modern Femur. Journal of Anatomy, 52 : 363-382. Maresh, M. M. 1955. Linear Growth of Long Bones of Extremities from In- fancy through Adolescence. American Journal of Diseases of Children, 89:725-742. Marshall, H. 1947. Longevity of the American Herring Gull. Auk, 64:188- 198. Marshall, T. K. 1968. Changes After Death. Chapter 8 in F. E. Camps, editor, Gradwohl's Legal Medicine. Second edition. Bristol: John Wright & Sons, Ltd. Marye, W. B. 1936. Former Indian Sites in Maryland as Located by Early Colonial Records. American Antiquity, 2:40—46. 1937. Burial Methods in Maryland and Adjacent States. Ameri- can Antiquity, 2:209-214. Mason, J. A. 1957. The Ancient Civilizations of the Andes. Baltimore: Penguin Books. McKern, T. W., and T. D. Stewart 1957. Skeletal Age Changes in Young American Males. Tech- nical Report EP—45, Quartermaster Research and Devel- opment Center, Environmental Protection Research Divi- sion, Natick. Means, P. A. 1931. Ancient Civilizations of the Andes. New York: Charles C. Scribner's, Sons. Meigs, P. 1935. The Dominican Mission Frontier of Lower California. University of California Publications in Geography, 7. Berkeley. Mercer, H. C. 1897. Exploration of an Indian Ossuary on the Choptank River, Dorchester County, Maryland: Researches upon the Antiquity of Man in the Delaware Valley and the Eastern U.S. University of Pennsylvania Publication Series in Philology, Literature, and Archaeology, 6:87—98. Meredith, A. V. 1946. Order and Age of Eruption for the Deciduous Dentition. Journal of Dental Research, 25:43—66. Merriam, C. H. 1905. The Indian Population of California. American Anthro- pologist, 7:594-606. Miles, A. E. W. 1963. The Dentition in the Assessment of Individual Age. Pages 191-209 in D. R. Brothwell, editor, Dental Anthropology. New York. Mook, M. 1944. The Aboriginal Population of Tidewater Virginia. Amer- ican Anthropologist, 46(2) : 193-208. Mooney, J. 1910. Population. Pages 286-287 in F. W. Hodge II, editor, Handbook of American Indians North of Mexico. Bureau of American Ethnology Bulletin, 30. Washington, D.C. 1928. The Aboriginal Population of America North of Mexico. Smithsonian Miscellaneous Collections, 80(7): 1—40. Washington, D.C. Moorrees, C. F. A. 1957. The Aleut Dentition: A Correlative Study of Dental Char- acteristics in an Eskimoid People. Cambridge. Moorrees, C. F. A., E. A. Fanning, and E. E. Hunt, Jr. 1963a. Formation and Resorption of Three Deciduous Teeth in Children. American Journal of Physical Anthropology, 21:205-213. 1963b. Age Variation of Formation Stages for Ten Permanent Teeth. Journal of Dental Research, 42:1490-1502. Morley, S. G. 1947. The Ancient Maya. Palo Alto: Stanford University Press. Motta, E. 1891. Morti in Milano del 1452 al 1552. Archivo Storico Lom- bardo, 18:241-286. Milan. Murie, A. 1944. The Wolves of Mount McKinley. Fauna of the National Parks of the U.S., Fauna Series 5. Washington, D.C: U.S. Department of Interior, National Park Service. Nalbandian, J. 1959. Age Changes in Human Teeth. Journal of Dental Re- search, 38:681-682. Nalbandian, J., and R. F. Sognnaes 1960. Structural Age Changes in Human Teeth. In Nathan W. Shock, editor, Aging. American Association for the Ad- vancement of Science, 65:367-382. Washington, D.C. Nelson, N. C 1909. Shellmounds of the San Francisco Bay Region. University of California Publications in American Archaeology and Ethnology, 7(4) : 309-356. Berkeley. Niswander, J. D., and C Sujaku 1960. Dental Eruption, Stature, and Weight of Hiroshima Chil- dren. Journal of Dental Research, 39:959—963. 1964. Dental Caries, Retained Primary Tooth Fragments and the Eruption of Permanent Teeth. Journal of Dentistry for Children, 31:139-145. 1965. Permanent Tooth Eruption in Children with Major Physical Defect and Disease. Journal of Dentistry for Children, 32: 266-268. Nougier, L. R. 1949. Densite humaine et population au Neolithique. Societe Prehistorique Francaise Bulletin, 46:126-127. Paris. 1950. Le peuplement prehistorique, ses Stapes entre Seine-et- Loire. Societe Prehistorique Franqaise Bulletin, 47: 210. Paris. 1954. Essai sur le peuplement prehistoriques de la France. Population, 9:241-274. 1959. Geographie humaine prehistorique. Paris. Ortner, D. J. 1970. The Effects of Aging and Disease on the Micromorphol- ogy of Human Compact Bone. PhD thesis, Department of Anthropolgy, The University of Kansas, Lawrence. Parsons, F. G. 1913-1914. The Characters of the English Thigh Bone, Part I. Journal of Anatomy and Physiology, 48:238-267. 1914-1915. The Characters of the English Thigh Bone, Part II: The Difficulty of Sexing. Journal of Anatomy and Physiology, 49:335-361. Paynter, K. J., and R. M. Grainger 1961. In^uence of Nutrition and Genetics on Morphology and and Caries Susceptibility. Journal of the American Medi- cal Association, 177:306-309. 1962. Relationships of Morphology and Size of Teeth to Caries. International Dental Journal, 12: 147-160. Pearson, K. 1901-1902. On the Change in Expectation of Life in Man during a Period of Circa 2000 Years. Biometrika, 1:261— 264. LITERATURE CITED 77 Phenice, T. W. 1969. A Newly Developed Visual Method of Sexing the Os Pubis. American Journal of Physical Anthropology, 30 (2):297-302. Proud, R. 1797—1798. The History of Pennsylvania. Philadelphia. Raudot, A. D. 1940. Memoir Concerning the Different Indian Nations of North America. Appendix to the Indians of the Western Great Lakes, 1615-1760. Edited by W. Vernon Kinietz. Occasional Contributions from the Museum of Anthro- pology of the University of Michigan, 10. Ann Arbor. Reynolds, E. L. 1945. The Bony Pelvic Girdle in Early Infancy. American Jour- nal of Physical Anthropology, 3(4) : 321-354. 1947. The Bony Pelvis in Prepuberal Childhood. American Jour- nal of Physical Anthropology, 5(2) : 165-200. Reynolds, E. R. 1883. Ossuary at Accotink, Va. Pages 92-94 in Abstract of Transactions of the Anthropological Society of Washing- ton, D.C. (1881). Smithsonian Miscellaneous Collections, 25. Washington, D.C. Richardson, B. E. 1933. Old Age among the Ancient Greeks. . Johns Hopkins University Studies in Archaeology, 16. Baltimore: The Johns Hopkins Press. Ricker, W. E. 1945. Natural Mortality among Indiana Blue-gill Sunfish. Ecology, 26:111-121. Ricketson, O. G., Jr., and E. B. Ricketson 1937. Uaxactun, Guatemala, Group E, 1926-1931. Carnegie Institution of Washington, 477. Washington, D.C. Press of W. F. Roberts, Co. Ridley, F. 1961. Archaeology of the Neutral Indians. Ontario: Etobicoke Historical Society. Riquet, R. 1953. La population des Grottes de Baye. Bulletins et Memoires Societe d'Anthropologic de Paris, 4:45—67. Rivet, P. 1924. Langues americaines. Pages 597-712 in A. Meillet and Marcel Cohen, editors, Les langues du monde. Paris: Collection Linguistique, Societe de Linguistique. Robinow, M., T. W. Richards, and M. Anderson 1942. The Eruption of Deciduous Teeth. Growth, 6:127-133. Roman, B. 1775. Natural History of East and West Florida. New York. Rosenblat, A. 1935. El desarrollo de la poblacion indigena de Americana. Tierra Firme, 1(1) : 115—133. 1945. La poblacion indigena de America desde 1492 hasta la actualidad. Cuadernos de la Serie Stirps Quaestionis. Buenos Aires: Institution Cultural Espanola. 1954. La poblacion indigena y el Mestizaje en America, I. Buenos Aires: Editorial Nova, Biblioteca Americanista. Rouse, I. 1952. Porto Rican Prehistory. The New York Academy of Sciences Scientific Survey of Porto Rico and the Virgin Islands, 18(4) : 463-518. Sanders, W. 1953. The Anthrogeography of Central Veracruz. Pages 27-78 in Bernal and Hurtado, editors, Huastecos, Totonacos y Sus Vecinos. Mexico: Sociedad Mexicana de Antropo- logia. 1956. The Central Mexican Symbiotic Region in Prehistoric Settlement Patterns in the New World. G. Willey, editor. Viking Fund Publications in Anthropology 23:115-127. 1960. Prehistoric Ceramics and Settlement Patterns in Quintana Rao, Mexico. Carnegie Institution of Washington Publi- cations 606 [Contributions to American Anthropology and History 60.] 1963. Cultural Ecology of the Maya Lowlands, II. Estudios de Cultura Maya, 3:203-241. Sapper, K. 1924. Die Zohl und die Vlksdichte der Indianischer Bevolker- ung in Amerika von de Conguista und in der Gegenwart. Pages 95-104 in Proceedings of the 21st International Congress of Americanists, I. Leiden: The Hague. Sauer, C 1935. Aboriginal Population of Northwestern Mexico. Ibero- Americana, 10. Berkeley. 1966. The Early Spanish Main. Berkeley. Schoolcraft, H. R. 1851. Historical and Statistical Information Respecting the History, Condition, and Prospects of the Indian Tribes of the United States. 6 volumes. Philadelphia: Lippin- cott, Grambo and Co. Schour, I., and M. Massler 1941. The Development of the Human Dentition. Journal of the American Dental Association, 28:1153-1160. Schranz, D. 1959. Age Determination from the Internal Structure of the Humerus. American Journal of Physical Anthropology, 17(4):273-278. Schwartz, D. W. 1956. Demographic Changes in the Early Periods of Cohonina Pre-history. Viking Fund Publications in Anthropology, 23:26-31. Scull, G. D., editor 1943. Voyages of Peter Esprit Radisson, Being an Account of His Travels and Experiences among the North American Indians, From 1652 to 1684. New York: The Prince Society. Sigerist, H. E. 1965. Civilization and Disease. Chicago: The University of Chicago Press. Singer, R. 1953. Estimation of Age from Cranial Suture Closure: A Report on Its Unreliability. Journal of Forensic Medicine, 1(1) : 52-59. Singh, I. J. and D. L. Gunberg. 1970. Estimation of Age at Death in Human Males from Quan- titative Histology of Bone Fragments. American Journal of Physical Anthropolgy, 33(3): 373-382. Snow, C. E. 1948. Indian Knoll Skeletons of Site Oh2, Ohio County, Ken- tucky. University of Kentucky Reports in Anthropology, 4(3/2):371-554. Speck, F. G. 1927. The Nanticoke and Conoy Indians. Wilmington: Histor- ical Society of Delaware. 1937. Oklahoma Delaware Ceremonies, Feasts and Dances. Memoirs of the American Philosophical Society, 7. Phila- phia. Steggerda, M., and T. J. Hill 1942. Eruption Time of Teeth among Whites, Negroes and Indians. American Journal of Orthodontics and Oral Surgery, 28(b): 361-370. 78 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 1 8 Stephenson, R. L., and A. L. L. Ferguson 1963. The Accokeek Creek Site. A Middle Atlantic Seaboard Cultural Sequence. Anthropological Papers Museum of Anthropology University of Michigan, 20. Ann Arbor. Stevenson, P. H. 1924. Age Order of Epiphyseal Union in Man. American Jour- nal of Physical Anthropology, 7(1) :53-93. Steward, J. H. 1949. The Native Population of South America. Volume 5(3) : 655-668 of Handbook of South American Indians. Bu- reau of American Ethnology Bulletin, 143. Washington, D.C. Stewart, T. D. 1934. Sequence of Epiphyseal Union, Third Molar Eruption and Suture Closure in Eskimos and American Indians. American Journal of Physical Anthropology, 19:433—52. 1940a. Further Excavations of the Indian Village Site of Pata- womeke (Potomac) Explorations and Fieldwork of the Smithsonian Institution in 1939, pages 79—82. 1940b. The Finding of an Indian Ossuary on the York River in Virginia. Journal of the Washington Academy of Sci- ences, 30(8):356-364. 1940c. A Report on the Skeletal Remains from the Piscataway Creek Ossuary. American Antiquity, 6(1): 4— 18. 1958a. Further Notes on the Melton Mask. Quarterly Bulletin of the Archeological Society of Virginia, 13(2). 1958b. The Rate of Development of Vertebral Osteoarthritis in American Whites and its Significance in Skeletal Age Identification. The Leech, 28(3,4,5) : 144-151. 1962. Comments on the Reassessment of the Indian Knoll Skele- tons. American Journal of Physical Anthoropology, 20(2) : 143-148. 1968. Identification by the Skeletal Structures. Chapter II in F. E. Camps, editor, Gradwohl's Legal Medicine. Second edition. Bristol: John Wright & Sons, Ltd. Stewart T. D., and W. R. Wedel 1937. The Finding of Two Ossuaries on the Site of the Indian Village of Nacotchtanke (Anacostia). Journal of the Washington Academy of Sciences, 27(5) :213—219. Stloukal, M. 1962. Heidnische Elemente de Bevolkerung des grossmahrischen Mikulcice auf Grund der Befund in Begrabnisstaten. Homo, 13:145-152. 1966. Comment on Estimating Aboriginal American Population. Current Anthropology, 7(4) :438-439. Strong, W. D. 1935. An Introduction to Nebraska Archaeology. Smithsonian Miscellaneous Collections, 93(10). Washington, D.C. Sundick, R. I. 1972. A New Method for Assigning Ages to Immature Skele- tons. American Journal of Physical Anthropology, 37(3) : 452. Swedlund, A. C, and G. J. Armelagos 1969. Une recherche en paleo-demographie: la Nubie Souda- naise. Annates Economies Societes Civilizations Extrait, 6:1287-1298. Paris: Librairie Armand Colin. Szilagyi, J. 1959. A datok Az Atlagos Elettartam Kerdeseihez Aquincumban es Pannonia mas Reszeiben. Antik Tanulmanyok, 6:31— 80, 221-243. Termer, F. 1951. The Density of Population in the Southern and Northern Maya Empires as an Archaeological and Geographical Problem. Pages 101-107 in Sol Tax, editor, The Civiliza- tion of Ancient America. Selected Papers of the 29th International Congress of Americanists. Chicago: Uni- versity of Chicago Press. Thompson, A. 1899. Sexual Differences of the Foetal Pelvis. Journal of Anat- omy and Physiology, 33: 359—380. Thompson, H. P. 1966. Estimating Aboriginal American Population: A Tech- nique Using Anthropological and Biological Data. Cur- rent Anthropology, 7(1) :417-424. Thompson, J. E. S. 1951. The Itza of Tayasal, Peten. Pages 389-400 in Homenaje al Doctor Afonso Caso. Mexico City. Thwaites, R. G. 1896-1901. The Jesuit Relations and Allied Documents: Travels and Explorations of the Jesuit Missionaries in New France, 1610-1791. 73 volumes. Cleveland: The Burrows Bros. Company. Tinkle, D. W. 1961. Geographic Variation in Reproduction, Size, Sex Ratio and Maturity of Sternothaerus odorata (Testudinata: Chelydridae). Ecology, 42:68-76. 1967. The Life and Demography of the Side-Blotched Lizard, Uta stansburiana. University of Michigan Miscellaneous Publication, 132. Ann Arbor. Tinkle, D. W., H. M. Wilbur, and S. G. Tilley 1970. Evolutionary Strategies in Lizard Reproduction. Evolu- tion, 24:55-74. Todd, T. W. 1920. Age Changes in the Pubic Bone, Part I. American Jour- nal of Physical Anthropology, 3:285—334. 1921. Age Changes in the Pubic Bone, Parts II—IV. American Journal of Physical Anthropology, 4:1-70. 1927. Skeletal Records of Mortality. Scientific Monthly, 24: 481-496. 1929. Entrenched Negro Physical Features. Human Biology, 1:57-69. 1930a. Age Changes in the Pubic Bone, VIII: Roentgenographic Differentiation. American Journal of Physical Anthro- pology, 14(2):255-271. 1930b. The Roentgenographic Appraisement of Skeletal Differ- entiation. Child Development, 1:298-310. 1931. Differential Skeletal Maturation in Relation to Sex, Race, Variability and Disease. Child Development, 2:49-65. Todd, T. W., and D. W. Lyon 1924. Endocranial Suture Closure, Its Progress and Age Rela- tionship, Part I: Adult Males of White Stock. American Journal of Physical Anthropology, 7(3) : 325-384. 1925. Cranial Suture Closure, Its Progress and Age Relation- ship, Part II: Ectocranial Closure in Adult Males of White Stock. American Journal of Physical Anthropology, 8(1):23-71. Valaoras, V. G. 1938. The Expectation of Life in Ancient Greece. Pratika de I'Academie d'Athenes, 13. Vallois, H. V. 1937. La duree de la vie chez l'homme fossile. Anthropologic, 47:499-532. 1960. Vital Statistics in Prehistoric Population as Determined from Archaeological Data. Pages 186-222 in R. Heizer and S. Cook, editors, The Application of Quantitative Methods in Archaeology. Chicago: Quadrangle Books. Van Vark, G. N. 1970. Some Statistical Procedures for the Investigation of Pre- historic Human Material. Groningen: Rijksuniversiteit te Groningen. LITERATURE CITED 79 Washburn, S. L. 1948. Sex Differences in the Pubic Bone. American Journal of Physical Anthropology, 6:199-208. Weidenreich, F. 1939. The Duration of Life of Fossil Man in China and the Pathological Lesions Found in His Skeleton. Chinese Medical Journal, 55:34-44. Weiss, K. M. 1973. Demographic Models for Anthropology. Memoirs of the Society for American Archaeology, no. 27 [American An- tiquity, 38(2)]. Welte, C R. 1966. Mapa de las localidades del Valle de Oaxaca. Oaxaca Oficina de Estudios de Humanidad del Valle de Oaxaca. Weslager, C. A. 1942. Ossuaries on the Delmarva Peninsula and Exotic Influ- ences in the Coastal Aspect of the Woodland Period. American Antiquity, 8(2) : 141—151. Wigglesworth, J. 1933. Excavations at Rehoboth. Bulletin of the Archaeological Society of Delaware, 3 :2—6. Wilcox. 1938. Willey, 1965 Wood, 1967. Wrong, 1939. Yarrow 1880. W. F. The Length of Life in the Early Roman Empire. Volume 3:14-22 in Congres International de la Population. Paris: Herman. G. R., W. R. Bullard, Jr., J. B. Glass, and J. C Gifford Prehistoric Maya Settlement in the Belize Valley. Papers of the Peabody Museum of Archaeology and Ethnology, 54. Cambridge. W. R. An Interpretation of Mandan Culture History. Bureau of American Ethnology Bulletin, 198. Washington, D.C. G. W., editor The Long Journey to the Country of the Hurons by Gabriel Sagard. Toronto: The Champlain Society. , H. C. Introduction to the Study of Mortuary Customs among the North American Indians. Washington, D.C: Smith- sonian Institution Bureau of American Ethnology. U. S. GOVERNMENT PRINTING OFFICE : 1974 O - 525-413 Publication in Smithsonian Contributions to Anthropology Manuscripts for serial publications are accepted by the Smithsonian Institution Press, subject to substantive review, only through departments of the various Smithsonian museums. Non-Smith- sonian authors should address inquiries to the appropriate department. If submission is invited, the following format requirements of the Press will govern the preparation of copy. (An instruction sheet for the preparation of illustrations is available from the Press on request.) Copy must be typewritten, double-spaced, on one side of standard white bond paper, with IY2" top left margins, submitted in ribbon copy with a carbon or duplicate, and accompanied by the original artwork. Duplicate copies of all material, including illustrations, should be retained by the author. There may be several paragraphs to a page, but each page should begin with a new paragraph. Number consecutively all pages, including title page, abstract, text, literature cited, legends, and tables. The minimum length is 30 pages of typescript and illustrations. The title should be complete and clear for easy indexing by abstracting services. Include an abstract as an introductory part of the text. Identify the author on the first page of text with an unnumbered footnote that includes his professional mailing address. An index, if required, may be supplied by the author when he returns page proof. For references within the text, use the author-date system: "(Jones 1910)" or "Jones (1910)." If the reference is expanded, abbreviate the data: "Jones (1910:122, pi. 20: fig. 1)." Simple tabulations in the text (e.g., columns of data) may carry headings or not, but they should not contain rules. Formal tables must be submitted as pages separate from the text, and each table, no matter how large, should be pasted up as a single sheet of copy. Use the metric system instead of (or in addition to) the English system. Illustrations (line drawings, maps, photographs, shaded drawings) can be intermixed through- out the printed text. They will be termed Figures and should be numbered consecutively; however, if a group of figures is treated as a single figure, the individual components should be indicated by lowercase italic letters on the illustration, in the legend, and in text references: "Figure 96." If illustrations (usually tone photographs) are printed separately from the text as full pages on a different stock of paper, they will be termed Plates, and individual components should be lettered (Plate 9b) but may be numbered (Plate 9: figure 2). Never combine the numbering system of text illustrations with that of plate illustrations. Submit all legends on pages separate from the text and not attached to the artwork. In the bibliography (usually called "Literature Cited"), spell out book, journal, and article titles, using initial caps with all words except minor terms such as "and, of, the." (For capitaliza- tion of titles in foreign languages, follow the national practice of each language.) Underscore (for italics) book and journal titles. Use the colon-parentheses system for volume, number, and page citations: "10(2) :5-9." Spell out such words as "figures" and "plates" (or "pages" when used alone). For free copies of his own paper, a Smithsonian author should indicate his requirements on "Form 36" (submitted to the Press with the manuscript). A non-Smithsonian author will receive 50 free copies; order forms for quantities above this amount with instructions for payment will be supplied when page proof is forwarded.