Preliminary Examination of Archaeological Remains from Fort Johnson and Cantonment Davis

Abstract

In September 1814, Brevet Major Zachary Taylor established Fort Johnson on the eastern bluffs of the Mississippi to assert American control over the mouth of the Des Moines River and points northward. Completed in a matter of weeks, the fort and its works were razed in late October of the same year when the troops ran out of provisions and retreated downriver to the St. Louis area. The following October, a temporary winter camp called Cantonment Davis was established near the fort ruins as a staging ground for building Fort Edwards, a subsequent military trading post. This article presents highlights from the archaeological search for the fort/cantonment location and recent excavations undertaken at the site, with a special focus on the armaments-related assemblage.     

“Horned cattle and pack horses”: zooarchaeological legacy collections from the unauthorized (and unscreened) Spanish Fort

ABSTRACT

In 1689, the governor of La Florida ordered the construction of a fort near the Muscogee (Creek) ancestral community of Apalachicola, supplying it with a caravan of “horned cattle and pack horses.” The fort, referred to as “Spanish Fort,” was abandoned a year later. Archaeological investigations of the fort were carried out in 1960 without sieving, and a large collection of faunal remains was minimally studied. Despite the limitations of the recovery methods, a recent analysis of these zooarchaeological legacy collections provides insight into the provisioning strategies of the Spanish military 150 years after initial colonialism began. Spanish Fort was better provisioned than its predecessors, such as Fort San Juan, but with a limited range of domesticated livestock—only cattle and horses. The presidio may have traded with the Apalachicola community in order to diversify their diet, but butchering marks indicate that the presidio’s soldiers processed their own meat at the fort. Having learned hard lessons from earlier colonial expeditions, Spanish military colonialists minimized the outpost’s vulnerability by not relying heavily on the local Native American population, while building a transactional relationship with Apalachicola to ensure the community’s cooperation. The zooarchaeological materials from Spanish Fort also indicate that the fort was intentionally destroyed by fire, providing a glimpse of Spanish adaptive strategies as the mission of securing the inland Southeast from rival colonialism abruptly ended.     

Archaeological Site or Natural Marine Community? Excavation of a Submerged Shell Mound in Ninigret Pond,Rhode Island

ABSTRACT

Submerged shell midden sites and natural shell deposits can have similar characteristics and can be difficult to distinguish archaeologically. We excavated two test units from a large (at least 35 m×70 m) submerged shell mound in Fort Neck Cove in southern Rhode Island to assess whether it was natural or cultural in origin. This mound had been recognized as a potential archaeological feature as early as the 1970s. Excavation, radiocarbon dating, and subsequent laboratory analysis of excavated materials suggest that the mound was a natural oyster reef rather than a submerged archaeological site. No artifacts were found; there was no clear evidence for human modification of any shells; small species that would not have been targeted as food were present; and δ¹³C values of oyster shells from the mound were consistent with freshwater input into their growth environment, suggesting that they grew in an estuarine environment that did not exist prior to the inundation of the ponds. The stratigraphically oldest radiocarbon date we could obtain (430–190 cal BP, 2σ range), from 70 cm below the pond floor, placed deposition of shells at least 3,000 years after the inundation of the pond. The excavation methods that we used and the process of testing, irrespective of whether the feature is cultural, are valuable contributions to the methodological literature on submerged site archaeology and help provide insight for other researchers working to discern natural from cultural shell midden sites.     

Ritualized Middening Practices

Practice theory focuses attention on agency and the generative dimensions of sites and material culture in terms of social formation and reproduction. This approach is applied to the ritualization of midden deposits by Torres Strait Islanders of northeast Australia. Ritualization of middens was achieved by three depositional strategies of privileged differentiation: referencing of mortuary remains, discrimination of animal bones and mounding of deposits. In addition, ritualized middens were spatially separated into small residential mounds and large communal feasting mounds. Ritualized middening was part of a broader social process of maintaining the biographical status of midden materials as a dimension of community socialization, identify and cohesion.     

Archaeological Investigation of Military Sites on Inchkeith Island

Abstract

In August 2001, a programme of survey and trial excavation was carried out on the island of Inchkeith in the Firth of Forth in Scotland. The work was carried out as part of the first series of the BBC TV series Two Men in a Trench, and examined the WWII defences, together with some of the Victorian defences and a midden that probably dates to the 16th century.     

Establishing the Populations and Occupation Spans of Fort Johnson and Cantonment Davis,Hancock County,Illinois

Abstract

Fort Johnson and Cantonment Davis are two short-term War of 1812–era American military posts located above the east bank of the Mississippi River in Hancock County, Illinois, within what was then the western frontier of the United States. Because little has been written regarding the fort and the cantonment, most details of their occupations are poorly documented and occasionally contradictory. A review of the documents associated with the two military posts establishes that Fort Johnson was occupied by approximately 100 men for 42 days, while Cantonment Davis was occupied by as many as 825 military personnel for about six to eight months.     

Reconstructing the massacre at Fort William Henry,New York

Fort William Henry, in upstate New York, was the site of a legendary siege and massacre in 1757 during the French and Indian War. As part of the terms of surrender, the British garrison was to retreat with all their arms and possessions, thus denying the Indian allies of the French their spoils of war. Contemporaneous and fictionalized accounts of the resulting massacre have often been regarded as exaggerations of actual events. Five men buried in a mass grave within the fort, known as the crypt, however, were clearly victims of the massacre. These men were among the sick and wounded who were unable to make the 15 mile (24 km) journey to Fort Edward and were left in the care of the French. Four of the five men sustained pre-mortem leg trauma that would have resulted in their hospitalization and prevented them from walking. The other massive perimortem trauma on these remains vividly depicts the results of the massacre. Three of the five men were shot in the knee; two of these three were shot elsewhere as well. One man was decapitated. Both the front and back of all the bodies bear cut marks, probably from the use of both axes and long-bladed knives as weapons. The numerous gashes in the thoracic and pelvic regions indicate the men were mutilated. Our analysis of the remains from this mass grave confirms and enhances the historical accounts of the massacre at Fort William Henry. The skeletons of these five men provide gruesome testimony of the assault to which they were subjected.     

ANALYSIS OF EARLY-NINETEENTH-CENTURY MUSCOGEE CREEK FUR TRADE AT A UNITED STATES FACTORY STORE

Abstract

We conducted an analysis of economic transaction records from the Fort Wilkinson Factory Store in central Georgia in an attempt to understand trade behaviors by the Muscogee Creek people of the southeastern United States. Between February 4, 1804, and November 29, 1806, factory personnel recorded 2,168 trade transactions at Fort Wilkinson. During this period, 38,226.5 deerskins and skins, 482 hides and rawhides, and 569 furs entered the fort. In exchange for these items, the Creek Indians primarily received cloth, followed by manufactured items such as kettles, gun supplies, and consumable groceries. We analyze the month-by-month exchange trends and compare the trade-good assemblages to archaeological sites that represent Creek occupations that date to the same time period in Georgia and Alabama. Further, we assess the biases of the data and interpret the data in relation to existing historical and archaeological studies of the Muscogee Creek people.     

Fishing Booths and Fishing Strategies in Medieval Iceland: an Archaeofauna from the of Akurvík,North-West Iceland

Abstract

Excavations in 1990 in North-West Iceland documented a stratified series of small turf structures and associated midden deposits at the eroding beach at Akurvík which date from the 11th–13th to the 15th–16th centuries AD. The site reflects a long series of small discontinuous occupations, probably associated with seasonal fishing. The shell sand matrix had allowed excellent organic preservation and an archaeofauna of over 100,000 identifiable fragments was recovered. The collections are dominated by fish, mainly Atlantic cod, but substantial amounts of whale bone suggest extensive exploitation of strandings or active whaling. This paper briefly summarizes the excavation results, presents a zooarchaeological analysis of the two largest radiocarbon dated contexts, and places the Akurvík collections in the wider context of intra-Icelandic and inter-regional trade in preserved fish. Analysis of the Akurvík collection and comparison with other Icelandic collections from both inland and coastal sites dating from 9th to 19th centuries AD both reinforces evidence for an early, pre-Hanseatic internal Icelandic fish trade and supports historical documentation of Icelandic participation in the growing international fish trade of the late Middle Ages.     

Feeding the Crusades: Archaeobotany,Animal Husbandry and Livestock Alimentation on the Baltic Frontier

ABSTRACT

The integrated results of micromorphology, plant macrofossil, pollen, phytolith, and non-pollen palynomorph analyses represent an important study of two thirteenth-century Teutonic Order castles at Karksi (Livonia), and Elbl?g (Prussia). The research examines deposits that formed during the period of active crusading. At Karksi, the investigation of a midden and of the organic-rich sediment beneath allows the diachronic use of this area to be understood. Freshwater aquatic indicators are consistent with the occurrence of shallow stagnant water, as also suggested by a waterlaid pond sediment identified in thin-section. Coprophilous spore taxa suggest the use of the pond as a watering hole. Plant macrofossils from the midden represent a range of habitats, mostly from wet/damp areas, as well as pastures and meadows, and also woodlands. Fragments of millet are embedded within herbivore dung in thin-section showing the use of this grain as fodder. At Elbl?g, parasite ova may derive from animal feces as they also occur in the dung observed in thin-section, and a range of coprophilous fungal spore taxa were extracted. The results reveal information about the range of livestock that the Teutonic Knights kept, whereabouts within the castles the animals were stabled, and what fodder was used.     

Early Animal Management Strategies during the Neolithic of the Konya Plain,Central Anatolia: Integrating Micromorphological and Microfossil Evidence

ABSTRACT

This paper examines the contribution of microscopic multi-proxy approaches to the study of early husbandry practices and animal diet by integrated micromorphological, phytolith, and calcitic dung spherulite analyses of midden deposits at the three neighbouring Neolithic sites of Boncuklu (9th–8th millennium cal BC), P?narba?? (7th millennium cal BC) and Çatalhöyük (8th–6th millennium cal BC) in the Konya Plain, Central Turkey.

The results reveal considerable chronological and contextual variation in human-animal inter-relations in open areas between different communities and sites. At Boncuklu, middens display well-defined areas where phytoliths and substantial accumulations of omnivore faecal matter low in spherulite content have been identified. By contrast, open spaces at the Late Neolithic campsite of P?narba?? comprise large concentrations of herbivore dung material associated with neonatal ovicaprine remains from spring birthing. Here, the deposits represent repeated dung-burning events, and include high concentrations of dung spherulites and phytoliths from wild grasses, and leaves and culms of reeds that, we suggest here, derive from fodder and fuel sources. Late middens at Çatalhöyük are characterised by thick sequences derived from multiple fuel burning events and rich in ashes, charred plants, articulated phytoliths – mainly from the husk of cereals, as well as the leaves and stems of reeds and sedges – and omnivore/ruminant coprolites, the abundance of the latter declining markedly in the latest levels of occupation.

The application of this integrated microscopic approach to open areas has contributed to unravelling the complexity of formation processes at these sites, providing new insights into herding practices, diet, and the ecological diversity of Neolithic communities in Central Anatolia.     

Early medieval T⊘nsberg — dating problems and the results

According to the 13th century Icelandic Saga writer Snorre Sturlasson, there was a marketplace (in old Norwegian: Kaupstadir) in the T?nsberg area by the end of the 9th century. The origin of the medieval town of T?nsberg has therefore been much discussed in the light of this statement. Viking Age and early medieval settlements often lack material datable by archaeology alone. In T?nsberg radiocarbon dating has been used to distinguish between Viking Age deposits and medieval deposits, when stratigraphy alone does not state the differences. However, this dating method has produced various results and there is obvious need for a local calibration.     

The Archaeology of the Siege of Leith, 1560

Abstract

This paper summarises the findings of two evaluations carried out in public spaces in the Leith area, both of which were specifically designed to establish the survival or otherwise of remains relating to the siege of Leith in 1560. Between 6 and 10 November 2006 a trial trenching evaluation was carried out within Pilrig Park to the south of the French citadel to 'ground truth' geophysical anomalies identified during a previous phase of work (Banks et al. 2006) and also to establish the character of topographic features on the ground. The main aim of the project was to establish the presence or absence of remains related to the siege fort known as Somerset's Mount which from a map drawn in 1560 is known to have been located in the vicinity. The evaluation identified the remains of a WWII air raid shelter, the remains of a building associated with Pilrig House demolished sometime in the early to mid 19th century, and a possible ditch associated with the fort, into which several archaeological features including the foundation slot of a small building or other structure and a probable blacksmith's forge had been cut.

The second part of this report concerns the evaluation at Leith Links, around 1 km to the north east of Pilrig Park and to east of the French citadel. A week long evaluation similar to that carried out at Pilrig took place in April 2007, the aim being to locate further evidence for siege works. Once again, attention focussed on geophysical anomalies resulting from the previous survey. The Links are traditionally associated with the siege through the presence of two mounds, which at least since the 19th century have been identified as the platforms for siege guns directed against the eastern face of the French citadel. An area between the two mounds, which are scheduled ancient monuments, was trial trenched along with an area further to the north. Only the latter produced physical remains which seem to relate to the 1560 map. These took the form of an earthen bank and possibly related ditches which run along the same alignment as a feature shown on the early map, though there is some question as to its function.     

CHRONOLOGY,SITE FORMATION,AND THE WOODLAND-MISSISSIPPIAN TRANSITION AT BAYSHORE HOMES,FLORIDA

Bayshore Homes (8PI41) is a large mound and midden complex on Florida’s west-central Gulf Coast that was investigated originally by William Sears in the 1950s. From 1999 to 2009, the authors conducted survey, test excavations, and soil coring to address questions regarding site formation, chronology, and cultural affiliation. Radiocarbon dates and ceramic analysis indicate two separate occupations during the Woodland and Mississippi periods, cal. A.D. 140–565 and cal. A.D. 890–1390 (2 sigma). The earlier occupation is associated with the Manasota archaeological culture, sand-tempered plain pottery, burials in midden deposits, and interments in a sand mound accompanied by Weeden Island–related mortuary ceramics. The later occupation is associated with Weeden Island–related decorated and Pinellas Plain ceramics in midden deposits and represents the transition from terminal Weeden Island to the Englewood phase of Early Safety Harbor. A large burial mound and a platform mound are associated with this period of site use. Our results also indicate that the unusual ceramic sequence identified by Sears in the site’s large shoreline shell midden is the result of redeposition, which occurred sometime after cal. A.D. 1220. Possible explanations for the redeposition event include monumental mound construction or the elevation of the midden ridge to serve as a foundation for structures to protect them from rising sea levels or storm surges.     

Evaluating the Reliability of Fluoride Dating at Two Prehistoric Mound Sites in Mississippi

Abstract

Ion-selective electrode fluoride dating is used to address chronological problems at two mound sites located in Mississippi. At the Mississippian period (A.D. 900–1520) Lyon's Bluff site, 220K520, the fluoride content of deer and human bone is compared with radiocarbon dates and stratigraphic levels to evaluate the reliability of the fluoride dating method for understanding relative burial chronology. At a second site, Pocahontas Mound A (22HI500), fluoride content of deer bone is used to corroborate radiocarbon dates and to establish the proper chronology of levels associated with two distinct occupations (Archaic and Coles Creek/Plaquemine) identified in a large midden area. Fluoride analysis also is used to date a second midden area at this site. For Lyon's Bluff; fluoride dates did not correspond with accelerator mass spectrometry (AMS)-dated burials and midden strata. In contrast, fluoride dating proved very useful in determining chronological placement of midden levels and dating other areas at Pocahontas Mound A. Sample age, site location, soil, and bone preservation are considered as possible causes for the negative results at Lyon's Bluff.     

Eighteenth-century Copenhagen viewed from a refuse dump

Abstract

The excavation of a town refuse dump in Copenhagen has recovered a remarkably rich and varied assemblage of artefacts including ceramics, glass, animal bones, feathers, textiles, gloves, woollen socks, silk stockings and shoes. They were deposited c. 1750–65, but include many objects dating from the early years of the 18th century. Because the deposits were waterlogged, various unusual types of organic evidence were preserved. This paper gives an introduction to the site and its material, and outlines some of its research potential.     

Seventeenth-Century Glass Trade Beads from La Salle's Fort St. Louis and the Grand Village of the Kaskaskia

Abstract

Glass trade beads are described here from two seventeenth-century sites located in the upper Illinois Valley, La Salle's 1682–1691 Fort St. Louis and the nearby Grand Village of the Kaskaskia, destroyed during an Iroquois raid in 1680. Simple monochrome drawn beads characterize both bead assemblages and each contains significant percentages of very small (<2 mm) and small (2–4 mm) size beads. Dominant colors are blue, white, and black. Turquoise-blue beads were a staple of the French trade at this time in the Illinois Country, particularly in transactions involving La Salle and his successors based at Fort St. Louis. Comparative treatment provided as part of this research indicates that there are significant similarities between the bead assemblage from Fort St. Louis and the beads recovered from the 1686 wreck of La Salle's ship Belle in Matagorda Bay off the Texas coast.     

THE PROBLEMS WITH VISUAL IDENTIFICATION OF DOVER AND FORT PAYNE CHERT

Abstract

Similar characteristics including color, luster, texture, and mineral inclusions of both Dover and Fort Payne chert types coupled with the wide geologic and geographic distribution of these deposits compound interobserver errors associated with visual identification methods. Samples were obtained from outcrops of Dover from Kentucky and Tennessee and Fort Payne chert from Alabama, Georgia, Illinois, Kentucky, Mississippi, and Tennessee demonstrating their extensive geologic distribution as well as the great similarity in macroscopic traits. A logistic regression analysis upon the sample database demonstrates successful differentiation of the chert types. However, the Dover/Fort Payne case study provides a cautionary example of the potential pitfalls with sourcing studies solely reliant upon visual attributes to assign artifact provenance. In general, the use of visual chert identification should be a “first approximation” and it is argued here that large type databases systematically sampled from both primary and secondary sources highlight the possible wide range of macroscopic variation present.     

Joined-Up Archaeology at Old Scatness,Shetland: Thin Section Analysis of the Site and Hinterland

Abstract

Excavations on the multi-period settlement at Old Scatness, Shetland have uncovered a number of Iron Age structures with compacted, floor-like layers. Thin section analysis was undertaken in order to investigate and compare the characteristics of these layers. The investigation also draws on earlier analyses of the Iron Age agricultural soil around the settlement and the midden deposits that accumulated within the settlement, to create a 'joined-up' analysis which considers the way material from the settlement was used and then recycled as fertiliser for the fields. Peat was collected from the nearby uplands and was used for fuel and possibly also for flooring. It is suggested that organic-rich floors from the structures were periodically removed and the material was spread onto the fields as fertilisers. More organic-rich material may have been used selectively for fertiliser, while the less organic peat ash was allowed to accumulate in middens. Several of the structures may have functioned as byres, which suggests a prehistoric plaggen system.     

Further aspects of fur trade life depicted in the skeleton

Osteobiographies were reconstructed from the skeletal remains of four adults from Fort Edmonton, a 19th century trading post of the Canadian fur trade. Three males were Caucasoid and probably ethnic Scots, given the usual origin of fur traders in this region. The lone adult female in the sample was Mongoloid, either Indian or Métis, and likely the ‘country wife’ of a fur trader, since she was buried in the European tradition in the fort cemetery. The cause of death is not discernible from any of the skeletal remains and none of these individuals exhibit any evidence of chronic infectious disease, malnutrition or neoplasia. Trauma, arthritis and other indicators of physical stress do appear, however, and present an opportunity to expand our understanding of the effects of fur trade life on the skeleton. Viewed in the context of historical accounts of life at the fort in the early 19th century, stress markers on the skeletons of three males have led to the conclusion that they were voyageurs who engaged in trading trips by canoe or boat. Lesions of the capsule attachment area at the proximal tibio‐fibular articulation appear unilaterally in two males and may be associated with ‘mushing’ or driving a dog sled in winter. The musculoskeletal lesions on the one preserved female skeleton are consistent with the arduous domestic activities documented at the fort, which include milking cows, churning butter, stirring lye soap, and harvesting grain and root vegetables by hand. Since specific occupations or behaviours cannot be precisely determined from muscular attachment and other stress markers, these interpretations are made cautiously and only in the culture‐historical context of the skeletal sample. Copyright © 1999 John Wiley & Sons, Ltd.