In 1991 and 1992, the Office of State Archaeology, NC Division of Archives and History, undertook extensive investigations at a medium-sized prehistoric shell midden and village site on the central coast of North Carolina (Mathis 1993, 1994). Overlooking Bogue Sound in Carteret Co., the site -- 31Cr218 -- was first identified in 1987 during a survey of the proposed Broad Reach Subdivision and Marina development project (Martin & Drucker 1987). The purpose of this paper is to offer a summary discussion and interpretation of one element of the site -- the Late Woodland period human burials. However, as a point of departure, I would like to provide a little background on the project and a few comments concerning the initial survey vis a vis the nature and extent of the site, and how we have traditionally dealt with this type of site in compliance archaeology situations. Those of you whom have heard this from me before can, I hope, bear with me one more time. My comments do have a direct bearing, however, on our understanding of mortuary practices along the coast, as I hope you will see.

The 1987 survey was conducted using a standard set of field techniques, including intensive surface inspection, shovel testing at 100 ft. intervals, and limited test excavations (Martin and Drucker 1987). The survey identified a "core" area of intact shell midden covering approximately 3 acres and a relatively low density scattering of shell and artifacts on the surface and in the plowzone over an area of an additional 10 or so acres. No apparent undisturbed cultural deposits were encountered in any of the subsurface tests outside of the midden "core", leading the archaeologists to conclude that the remainder of the site probably had been disturbed by plowing and otherwise not likely to yield significant data. As a consequence, recommendations for future testing and data recovery focused on and immediately surrounding the shell midden "core". This information subsequently was used by the State Historic Preservation Office for negotiating project plan modifications with the developers, whereby the "core" of the shell midden would be avoided by the direct impacts of the marina construction.

Between the 1987 survey and 1991, however, when the development plans began to take shape, we decided, based on a observations and experiences at other sites with shell middens, to "challenge" or test, if you will, the assumptions and results of the initial survey. Thus began the Broad Reach Project. At the time, the intent of the project was simple: to examine the relative efficacy of the standard survey techniques employed along the coast (and elsewhere, I might add). In other words, how accurate have our assumptions, deductions and recommendations been, based on limited testing and surface artifact distributions, even in cases of plowed fields?

With the cooperation of the developers, Broad Reach Investments, Inc., we removed the plowzone from two areas adjacent to but off the main body of intact shell midden. A total of just under 8.5 acres of plowzone was removed using a combination of road grader and paddle pan. A total of 696 features were recorded and mapped, not including the hundreds of postmolds and other features less than 20 cm. in diameter. Dozens of other features, primarily subtle humus stains, were initially observed but not recorded either due to time constraints or the fact that literally disappeared under blowing sands or the leaching effects of the sun.

The principal feature type identified was the shell-filled or predominantly shell-filled pit (over 350). I would like to note here that much of the shell scattered across the surface of the site, away from the shell midden, is a result of plow-truncation of these shell-filled pits. Other feature types included human burials, hearths, general refuse-filled pits, cobble tool caches, bipolar flaking and lithic workshop areas, and smudge pits. In addition, all or portions of 11 structure patterns were identified, including oval, round and rectangular structures. One pattern was of a rectangular structure measuring approximately 4.9 x 6.1 meters; another was a true longhouse approximately six meters wide and more than 17 meters long. The full length of the longhouse could not be determined due to time constraints and the fact that it extended into the area that was not stripped of plowzone. [See: Broad Reach Site Plan]

The point I want to make here is that during the 1987 survey, shovel testing was conducted across the entire area covered by our 1991-92 investigations, yet encountered no evidence of subsurface features. Had we not undertaken this "test", and had focused subsequent mitigation efforts on the area of intact midden, we would have acquired more information on the shell midden itself, but at the expense of the habitation and mortuary areas.

During the investigations, we excavated all or portions of 127 pit features. Of these, 15 contained human remains. The burials included two ossuaries, a primary burial containing two individuals, two primary flexed burials, six secondary bundle burials, and four burials containing only fragments of bone. The only burial forms not identified in the sample are extended burials and isolated cremations. In at least three of the four fragmentary burials, the bulk of the skeletal remains apparently had been removed in prehistoric times.

In addition to the definite burials, 20 or more features were excavated near known burials which contained no skeletal or substantial artifactual remains, at least some of which are now thought to have been temporary or "interim" burial pits based on their comparable sizes, shapes, depths and paucity of fill inclusions. That is to say, they appear to have been burials from which the skeletal remains were disintered by the prehistoric inhabitants. As I will discuss further below, it is quite possible that the remains were subsequently interred in an ossuary.

Spatially, all but three of the burials occur in clusters, referred to here as "Cemetery Areas". Cemetery Area 1 included one primary flexed (Burial 13), three bundle and one partially disintered burial (Burials 3, 4, 5 and 7), along with 12 potential "empty" burial pits. Cemetery Area 2 contained both of the ossuaries (Burials 1 and 6) and two burials containing only fragments of skeletal remains (Burials 9 and 11), both of which appear to have been deliberately disintered. Cemetery Area 3 consisted of three burials, two of which were secondary bundles (Burials 2 and 14); the third, Burial 15, is problematical due to disturbances but may have been another partially disintered burial. Many other burials may have existed around these areas but not observed, since identification usually required careful shovel-shaving of the quick-drying and blowing sands, a process which was not afforded more than a small portion of the total project area. Because of this, Cemetery Area 1 and Area 2 could actually be one in the same, since they were located only about 20 meters apart. It is for the same reason that I would say that the three apparently isolated burials may not actually be isolated; we just did not have time to go looking for others in the vicinity.

One of the isolated burials was a primary flexed interment (Burial 16), almost identical to Burial 13, found in Cemetery Area 1; another was a secondary bundle (Burial 10), similar to those also found in Cemetery 1. The third of the isolated burials (Burial 8) contained two individuals and a number of artifacts, including two Hanover type, clay tempered, pottery vessels, at least eight turtle shells, six of which were cupped and bundled together, along with a deer antler and canon bone, a beaver tooth and a conus shell. An adjusted radiocarbon of AD 365 +/- 100 places this burial solidly within the Middle Woodland period of the coastal region. The other burials appear to be associated with Late Woodland occupations based on both radiocarbon dates and associated artifacts (Table 1).

Although the absolute number of burials is small by comparison to sites in other regions, the Late Woodland burial assemblage at Broad Reach is the largest and most varied thus far excavated in the outer coastal or tidewater region of North Carolina. Based on the previous work in North Carolina and the Middle Atlantic region, and when cautiously framed within an ethnohistorical context, the Broad Reach assemblage may well represent a viable cross-section of a Late Woodland mortuary complex. As proposed here, this complex consists of a series of dynamic mortuary processes, whereby the remains of the deceased were physically moved through one or more treatment stages, prior to final interment. The treatment afforded a particular individual may have depended on their status in the community, but other factors may have been involved, such as the time and place of death and perhaps even the availability of next-of-kin for insuring the proper treatment of the remains.

To digress momentarily: In 1984, David Phelps, East Carolina University, reported the excavation of two Late Woodland period, Colington Phase burials at the Tillett Site, on Roanoke Island, located on the northern coast of North Carolina (Phelps 1984). The site is within the region occupied by Algonkian-speaking groups from at least AD 900 and into the historic period (Phelps 1983, 1984). The distinguishing feature of the Tillett site burials is that they were not ossuaries. One was a secondary bundle, the other a flexed inhumation. According to Phelps (1984:38), this was the first documented archaeological instance of non-ossuary burials associated with the Colington phase or "other contemporary phases in the Middle Atlantic subarea" (cf. Phelps 1983:40-42; Snow 1978). However, as Phelps suggested at the time, burial forms other than ossuaries were probably common and important elements in Colington phase mortuary practices. The paucity of single inhumations was, and remains today, more likely a reflection of the paucity of archaeological investigations than cultural reality.

While there are many similarities in the material cultures found in the northern and central coastal areas of the state, there remains considerable question regarding the actual cultural associations (cf. Phelps 1983, 1991; Loftfield 1990). However, three presumed hallmark traits of the Algonkian Colington phase -- shell-tempered ceramics, longhouses and ossuary burials -- are common to the central coast and, to a lesser extent, as far south as the Cape Fear River, near the North Carolina-South Carolina border (cf. Bogdan and Weaver 1989; Loftfield 1987, 1990; Mathis 1986). The most readily visible characteristic, at least at a superfical level of analysis, is the ceramics. Shell-tempered ware to the north of the Neuse River is generally called Colington. To the south, it is referred to as White Oak (Loftfield 1976) or Oak Island (South 1964). Fabric-impressing is the dominate surface treatment, with plain and occasionally burnished surfaces. Simple-stamping is common to the north but rare to absent south of the Neuse River. However, in general terms, there appears to be relatively little stylistic, technical or functional difference between the Colinton and White Oak/Oak Island wares, suggesting some measure of commonality and continuity. In virtually all other associated artifactual aspects, there is little difference between the archaeological assemblages of the sites in the northern and southern subareas of the North Carolina coast.

Longhouses have been found at at least three sites in the Carteret and Onslow counties area of the central coast, including the Broad Reach, Uniflite (31On33) and Permuda Island (31On196) sites (Loftfield and Jones n.d.). More recently, the corners of one and possibly a second square or retangular structure were identified at the Shelly Point Site (31Cr53), located just a mile southwest of Broad Reach (Reid 1994). Longhouse construction is generally considered an Algonkian trait in the outer coastal region and suggests and Algonkian association for at least a portion of the Broad Reach assemblage.

Unfortunately, the relatively small sample of ossuaries north and south of the Neuse River presents a more complicating than clarifying pattern. Of the seven documented ossuaries excavated south of the Neuse River, none fits a specific pattern in form or content; ranging significantly in form, size and content (cf. Loftfield 1990).

For instance, the Flynt Site (31On305) ossuary, located to the south of Broad Reach, contained the bundled and mixed remains of at least 150 individuals, tentatively identified as "Algonkian", although an "Iroquoian" affiliation cannot be ruled out given the apparent osteological similarities between the two groups (Bogdan and Weaver 1989). Nearby, the Jarretts Point site (31On309) ossuary contained the mixed remains of less than a dozen individuals but also was identified as probably "Algonkian" (Loftfield 1987). The Flynt Site ossuary was dated to ca. A.D. 1290-1402 (uncorrected, 1-sigma). A nearly identical date of A.D. 1297-1408 (1-sigma) was acquired for the Jarretts Point ossuary. The Burial 6 ossuary at Broad Reach, by contrast, contained the tightly bundled clustered remains of at least nine individuals. Burial 6 was dated to A.D. 1032-1247 (1-sigma, corrected). No biocultural identification has been attempted for the remains. At the Cold Morning site (31Nh28), located in New Hanover County, at the southern end of the region, was been identified as more likely associated with Siouan groups (Coe, et al. 1983), and was dated to A.D. 886-1024 (uncorrected, 1-sigma).

Based on overall archaeological characteristics, including the ossuary form of burial, the ceramics and other artifacts, I am reasonably comfortable in suggesting that there is some, albeit nominally documented, cultural relationship between the Late Woodland societies of the northern and central coastal subareas.

As suggested previously, ossuary burials have and continue to be the most common form of human interment excavated thus far along the outer coast. Based on the excavations, all in salvage contexts, there may not be a "typical" ossuary form, per se, although most tend to occur in roughly round pits ranging from 1.5 to 3 meters in diameter and contain skeletal remains from as few as six or seven individuals to as many as 150 or more (Phelps 1983; Mathis 1986, n.d.b.; Loftfield 1987, 1990; Ward ____; Coe, et al. 1983). Perhaps the best documented examples of Colington Phase ossuaries are from the excavations at the Baum Site (31CK9), located on the northern coast (Phelps 1980, 1983). The Baum site ossuaries contain both secondarily deposited articulated remains and fully disarticulated bundles, as well as randomly scattered bones. These features are generally characteristic to all of the dozen or so known ossuaries along the northern and central coast, with the exception of those found at Broad Reach.

While there are characteristics common to most of the known ossuaries, there also are some differences, particularly between sites, although multiple ossuaries have been excavated at only two sites - Baum and Broad Reach. Most of the known ossuaries contain all or the partial remains of from 20 to 60 individuals, although the Flynt Site ossuary contained portions of no less than 150 (Bogden and Weaver 1989). In contrast, one of the Broad Reach ossuaries, Burial 1, contained no apparent articulated elements and only portions of from four to six individuals, plus a cremation. An inclusive cremation is known from only one other ossuary, located on Knotts Island, on the north coast (Mathis n.d.b.). At least seven copper beads were randomly scattered among the bones in Burial 1, as if tossed onto the bones after placement in the pit.

The other Broad Reach ossuary, Burial 6, contained nine distinct partially articulated adult bundles with no mixing and no "scattered bone". However, one of the bundles also contained elements from two or more subadults (Liz Monahan, Wake Forest University, personal communication). In addition to the bundle segregation, Burial 6 differs from other recorded ossuaries in it's depth (over a meter, including the plowzone). Furthermore, over the top of the ossuary was a thick layer of shell, the vast majority of which was clam, even though the majority of the shell found across the site and in the nearby midden is oyster. In the pit fill, but primarily under and around the edges of the shell layer, were large pottery sherds and scattered charcoal or carbonized wood. At this point I am inclined to suggest that the sherds may have been used as shovels used in the preparation of the burial pit. At the base of the pit, beneath the bundles, was a thin but clearly visible organic stain. The fact that the stain covered the entire base and slightly up the sides of the pit indicate that it's origin was not from the bundles alone, but that the pit probably had been lined or covered by grasses, wood, mats, skins, or other organic materials. One additional unusual feature of the ossuary was the presence of deliberately interred artifacts with two of the bundles. One of the characteristics of previously excavated ossuaries is the general lack of burial offerings. One bundle, which was placed away from the others, was accompanied by a shell-tempered pottery vessel, a small perforated groundstone cup to which a string of cut shell disk beads had been attached, a turtle carapace, at least five clusters of marginella shell beads (possibly bracelets), and what has been tentatively identified as a bundled baby dog. Another bundle was accompanied by over 700 marginella beads which may have been sewn onto the original bundle wrapping.

How these differences between ossuaries within and between sites reflect or may be a reflection of socio-cultural variables, such as social status in the community, or simply a factor of regional variation on the general ossuary theme, remains to be determined. It is reasonably clear, however, that if we were to assign status on the basis of burial complexity and accouterment, the Burial 6 ossuary at Broad Reach meets all possible criteria as a high status interment. By contrast, the mixed and incomplete remains found in the smaller Burial 1 ossuary, could represent a lower status or commoner burial.

Social status may be reflected in the structure and skeletal content of individual ossuaries, with or without accompanying burial offerings. It is possible, however, that individuals of different status were ultimately interred in the same ossuary. In such cases, it may be that the stage of articulation vs. disarticulation, bundling vs. scattering, and degree of skeletal completeness within an ossuary is a reflection of the status of the individual, rather than that of the ossuary population as a whole. An additional consideration, however, and probably of greater significance, is the timing of death. How, where and when an individual died surely had a direct bearing on the completeness and condition of their skeletal remains when eventually placed in the ossuary. I will return to this point momentarily.

Unfortunately, while the 16th and 17th century ethnohistorical accounts of the English colonists of Virginia and North Carolina provide passing references to the treatment of the dead, including the so-called "commoners", none makes mention of ossuaries. For these we must look to the Great Lakes region and the Huron "Feast of the Dead", as reported by the Jesuit missionaries during the first half of the 17th century (Tooker 1964). The Huron ceremony was an elaborate affair lasting many days and often involved more than a single community. At selected times, estimated to range from every 8 to 12 years, the people of a community would bring together the remains of those who had died since the last "Feast of the Dead" for burial in one or more large, communal pits. Neighboring communities were often invited to participate in the Feast. If they so desired to have the host village serve as the final resting place for their deceased, they were welcomed and treated as if members of the host community, although they may choose to place their deceased in separate pits.

Of particular interest is the observation that upon notification of the impending ceremony, family members went to the cemeteries to remove the remains of their deceased (Tooker 1964:136). In instances where the flesh had dried or decayed, family members cleaned the bones of any remaining flesh and bundled them together in new cloths or skins. The recently deceased were "left as they were and covered with new robes" (Tucker 1964:136).

Following several days of preparation, and when all participating families were gathered together at the site of the ceremony, the "whole bodies" were placed in the pit. In the morning of the following day, the bundled bones were "thrown into the pit" and arranged with poles or paddles. When the pit was full or all bones interred, it "was covered with sand, poles and wooden stakes, thrown in without any order" (Tooker 1964:139). Over the next few days, nets full of corn were delivered to and thrown over the pit. [Recall, if you will, the layer of clam shell covering the Burial 6 ossuary at Broad Reach.]

Granted, the seventeenth century Huron are rather far afield from the coast of North Carolina and certainly not to be used as a source of direct ethnographic analogy. Among other things, the outer coastal and tidewater regions of North Carolina were dominated by Algonkian-speaking groups. In addition, the ossuaries found in the Great Lakes region, and even as far south as the Virginia coast, tend to be larger and contain the remains of more people than the Algonkian ossuaries of North Carolina (cf. Cleland ____; Trigger ____; Ubelaker 1974). This may be a reflection of greater population, greater significance of the ceremony among the Huron and Middle Atlantic Algonkians, or some other fundamentally different cultural or religious characteristic. Nonetheless, given the structural similarities (i.e., secondarily deposited bundles, articulated, partially articulated and scattered bone) it is reasonable to suggest that some of the same basic mortuary processes were involved in the creation of the ossuaries, regardless of size and skeletal population.

For example, in the case of the Huron, the bodies of the deceased were commonly placed on a scaffold shortly after death, where they would remain for some unspecified time (Heidenreich 1978:374). Eventually, however, the remains would be placed in an ossuary during a "Feast of the Dead." A similar treatment evidently was common among the Virginia and North Carolina Algonkians (Smith 1986b; Spelman 1910; Feest 1978; see also Feest 1978:245) and John White's 1586 drawing and description of a charnel house indicates that the remains of some individuals, identified as chiefs, received special attention prior to final interment (Hariot 1588). What happened to the remains after treatment in the charnal house is not reported, although one can probably assume they, too, were placed in an ossuary (cf. Phelps 1983).

Primary interment also was practiced among the Huron, although the Jesuit reports suggest that this was more often afforded individuals who had died during battle or had drowned (Heidenreich 1978:374). In these instances, a shrine was placed over the grave pending the next "Feast of the Dead", at which time the remains were disinterred for placement in an ossuary. Similarly, among the Nanticoke, in the Chesapeake region, primary inuhumations were commonly followed by secondary interment in an ossuary. In the Virginia area, Smith (___) reported that "ordinary burials" involved simple placement of the corpse in a "hole."

It is unclear from the ethnohistorical accounts whether the scaffold treatment was unique to higher status individuals or was afforded all members of the community, if they so desired or was possible, given the location and circumstances of death. John Smith does mention, and I paraphrase here, that commoners were simply placed in holes in the ground and left there to "rot" (Smith 1986:169). If so, this could explain the primary flexed burials. Does this mean that the single secondary bundle burials are of higher status individuals? I doubt it. The remains obviously had been stored elsewhere prior to interment, but probably not because they were of status individuals. Among other things, when considering the ethnohistorical documents we must recognize that by the time John Smith and others made their observations, significant changes had already taken place in the social fabric of the native Algonkian cultures (Phelps 1984; Potter 1993; Roundtree 1989). How these changes may have affected the mortuary processes can only be imagined at this point, yet cannot be overlooked. Furthermore, if local equivalents of the Huron "Feast of the Dead" occurred only at intervals of 8-12 years, or even as little as the 6-8 year interval suggested by Ubelaker (1974), and given the fact that there is no mention of them in the historical accounts, it seems quite likely that the Europeans were either deliberately excluded from or were otherwise not around to view the ceremony, or even that the ceremony itself had ceased to exist as a regular and viable element of the cultural system of the Virginia Algonkians. Given the numbers of ossuaries recovered in the Mid-Atlantic region which contain European trade items, however, the latter case does not appear likely.

In light of the above, I would suggest the following: Based on the limited sample form Broad Reach, when an individual died, the body was either placed directly in the ground (e.g., the flexed burials) or, more commonly, placed on a scaffold for a period of time. At some point in time, perhaps when additional room was needed on the scaffold, when the flesh had decayed sufficiently, or when the village was temporarily or permanently abandoned, the remains were bundled and secondarily interred in cemetery areas around the village or stored in relative's houses or a temple. In some instances, the remains may have been carried away to another settlement. If a "Feast of the Dead" was scheduled in the near future, however, the remains might be taken directly from the scaffold and prepared for placement in an ossuary. At the designated time for the local "Feast of the Dead", family members returned to the cemeteries (and/or scaffold area) to recover the remains. The "empty" pits and pits containing only portions of human remains in the cemeteries thus represent an intermediate stage in the treatment of the dead, with the final "anticipated" destination being an ossuary. Given the fact that most of the known ossuaries in the region contain isolated and fragmentary skeletal remains, within the context of this process it stands to reason that the rest of the remains were either lost "in transit" or left behind in the intermediate stage pits. In the Broad Reach case, analyses are currently underway to determine if the remains left behind in the individual pits match those found in the Burial 1 ossuary. That some of the individual pits still contain the more or less complete remains of an individual may be due to rather simple and very human factors, including but not limited to the abandonment of the village, death of the family members responsible for the final movement of the remains, or other even "misplacement" of the remains (i.e., forgetting where they were buried).

To a certain extent, the social and economic status of the individual may well be represented by the relative completeness of the remains ultimately placed in the ossuary. In effect, skeletal completeness relates to both the length of time since death and the intermediate form of treatment (in addition to the cause and location of death). If commoners were afforded only temporary tending on a scaffold, followed by placement in cemetery pits, then the likelihood exists that only a portion of their remains would ulimately make it into the ossuary, in addition to experiencing a relatively greater degree of deterioration. Most of the recorded ossuaries, including Burial 1 at Broad Reach, appear to fit this description to some extent.

In contrast to the individual burials and the Burial 1 ossuary, with it's fragmentary contents, the Burial 6 ossuary truly appears to be that of higher status individuals. The bundles appear to include all or the vast majority of the individual's remains, and when taken in tandem with the pit structure and the burial offerings, indicate far more careful and deliberate treatment of the deceased. In this instance, it appears conceivable that the remains were not previously stored in pits but tended to in a charnel house or comparable mortuary context prior to interment.

Although I am tempted to expand this model, as it is, to include the coastal ossuaries and funeral practices beyond the Broad Reach situation, too many unanswered questions remain, with too little data. Perhaps the most significant result of the Broad Reach Project thus far, however, is that the sheer size of the investigations have afforded us a much better opportunity to even think about such questions as mortuary processes. The fact that all but one of the burials recovered at Broad Reach were located well away from the shell midden, and even outside of the site boundaries initially defined for the site in 1987, suggests that we need to carefully consider how we approach these types of sites in the future, and how we might best structure our investigations to insure the protection, preservation and understanding of coastal prehistory.

Acknowledgements:

The Broad Reach Project was made possible by funding from Broad Reach Investments, Inc., the Friends of North Carolina Archaeology, Inc., the North Carolina Archaeological Society, the North Carolina Division of Archives and History, and the U.S. Department of the Interior, National Park Service. By late 1992, over 125 individuals had volunteered time on the fieldwork component of the project. To each I extend my most sincere appreciation. I also wish to thank Mssrs. Steve Stroud and Carlton Midyette, Broad Reach Investments, Inc., for their cooperation and interest in the project; Mr. W.F. Parker, W.F. Parker Construction Co., for the use of his heavy equipment and the crew of operators. At the Division of Archives and History, thanks are in order to virtually the entire staff of the Office of State Archaeology, not only for their assistance in the field and lab, but their patience and tolerance over the life of the project.

Originally presented at the 50th Annual Meeting of the Southeastern Archaeological Conference, Raleigh, NC, November 5, 1993.

Note: Since this paper was written, several changes have occurred in our understanding of the Late Woodland period of the central and southern coast, particularly with respect to the ceramic sequence. These changes, however, do not substantially modify the proposed model mortuary processes observed at Broad Reach. (MAM, 6/99)

References:

Coe, Joffre, H. Trawick Ward, Martha Graham, Liane Navy, S. Homes Hogue (Wilson) and Jack H. Wilson, Jr.

1982 Archaeological and Paleo-Osteological Investigations at the Cold Morning Site, New Hanover County, North Carolina. Interagency Archeological Services, National Park Service, Atlanta.

Bogdan, Georgieann and David S. Weaver

1989 Report on the Human Skeletal Remains form the Flynt Site (31 On 305). Report prepared for the North Carolina Office of State Archaeology, Raleigh.

Loftfield, Thomas C.

1987 Archaeological Data Recovery and Osteological Analysis at 31On309, Marine Corps Base, Camp Lejeune, North Carolina. National Park Service, Atlanta.

1990 Ossuary Interments and Algonquian Expansion on the North Carolina Coast. Southeastern Archaeology 9(2).

Martin, Debra K. and Lesley M. Drucker

1987 Archaeological Inventory Survey of Proposed Marina Construction Areas, Broad Reach Tract, Carteret County, North Carolina. Resource Studies Series 106. Carolina Archaeological Services, Columbia, SC.

Mathis, Mark A.

n.d.a. Field Notes on the 1985-87 Salvage Excavations at the Flynt Site (31On305), Onslow County, North Carolina. On file, North Carolina Office of State Archaeology, Raleigh.

n.d.b.n Field Notes on the 1989 Knotts Island Ossuary Excavation (31Ck24), Currituck County, North Carolina. On file, North Carolina Office of State Archaeology, Raleigh.

1986 The Flynt Site Ossuary (31On305) Excavation Summary. Paper on file, North Carolina Office of State Archaeology, Raleigh.

1993 Beyond the Shell of Shell Middens. Paper presented at the Conference on Shell Midden Archaeology, Georgetown, SC.

1994 Broad Reach: The Truth About What We've Missed. In Site Destruction in Georgia and the Carolina, edited by D.G. Anderson and V. Horak. Readings in Archeological Resource Protection Series No. 2, Interagency Archeological Services Division, National Park Service, Atlanta.

Phelps, David S.

1980 Carolina Algonquian Ossuaries. Paper presented at the 37th Annual Meeting of the Southeastern Archaeological Conference, New Orleans.

1983 Archaeology of the North Carolina Coast and Coastal Plain: Problems and Hypotheses. In The Prehistory of North Carolina: An Archaeological Symposium, edited by M.A. Mathis and J.J. Crow. North Carolina DIvision of Archives and History.

1984 Archaeology of the Tillett Site: The First Community at Wanchese, Roanoke Island. Archaeological Research Report No. 4, East Carolina University.

1991 Archaeology of the Southern Algonquians: The Carolina Sounds. Paper presented at The Southern Algonquians: A Symposium on the Native American Chiefdoms of the Middle Atlantic Tidewater, Greenville, North Carolina.

Potter, Stephen R.

1993 Commoners, Tribute and Chiefs: The Development of Algonquian Culture in the Potomac Valley. University Press of Virginia, Charlottesville.

Roundtree, Helen C.

1989 The Powhatan Indians of Virginia. University of Oklahoma Press, Norman.

Snow, Dean R.

1978 Late Prehistory of the East Coast. In Handbook of North American Indians, Volume 15 (Northeast), edited by B.G. Trigger, pp. 357-361. Smithsonian Institution.

Tooker, Elizabeth

1964 An Ethnography of the Huron Indians, 1615-1649. Smithsonian Institution Bureau of Ethnology Bulletin 190.

Ubelaker, Douglas H.

1974 Reconstruction of Demographic Profiles from Ossuary Skeletal Samples: A Case Study from the Tidewater Potomac. Smithsonian Contributions to Anthropology No. 18.

Source: "Mortuary Processes at the Broad Reach Site" by Mark A. Mathis, Office of State Archaeology,
Archaeology and Historic Preservation Section, NC Division of Archives and History 1993 (Modified slightly 1999).