Williams Fort was occupied from about 1811 to 1838 by Aaron and Tabitha Williams. American Resources Group tested and mitigated the site for White County Coal of Carmi, Illinois, in 2005 and 2006. The site principally functioned as a farmstead but was briefly fortified during the War of 1812 period and served as a rural tavern during the 1820s and 1830s.
Williams Fort is in Hawthorne Township within White County on a sandy ridge at the eastern margin of a historic prairie in the lower Wabash River valley of southeastern Illinois. This 1876 state atlas map contains superimposed data from multiple sources, most notably the extent of Big Prairie (in green) in 1809, and the locations of six forts estimated from a combination of old hand-drawn maps and descriptions in the 1883 county history. These civilian defenses were supported by an organized militia stationed in the prairie. Aaron and Tabitha Williams moved to Big Prairie in ca. 1811. A historical account of the New Madrid earthquake, the first of which occurred on December 16, 1811, ties the family to the area by then as one of their daughters was badly burned while cooking when the earthquake hit.
When we began the Phase II testing project at the Williams Fort site in 2005, we were not expecting to discover any evidence of a blockhouse or fort. In fact, previous investigations by another group suggested only the possibility of a tavern at the site. Unmistakable sections of the stockade wall trench began to emerge during the testing phase and was exposed in full during the mitigation phase of the project.
The stockade wall formed a roughly squared enclosure with an elongated southeastern corner. A squared bastion projected from the center of the north wall, and a gate was located directly opposite on the south wall. Baked clay and daub recovered along the stockade footprint would have been used to chink gaps between the posts. The slightly irregular stockade footprint meant that existing buildings had been incorporated into the enclosure.
The stockade wall appeared to contain four horizontal log buildings. Openings at the southwest and southeast corners were interpreted as the location of two blockhouses. The one at the southeast corner was the first or main blockhouse judging by the significant activity occurring to the southeast. Compacted subsoil at the proposed blockhouse locations suggests dirt-floored buildings. Two rectangular cabins—the earliest buildings at the site, joined by the bastion, were located at the two openings along the north wall.
No wells or cisterns were located within the enclosure meaning that little time was spent inside. While the absence of a water source within a fort was not unusual based on comparative data, it was certainly more precarious judging by accounts of settlers being attacked or killed while fetching water.
The height of the Williams’ stockade was extrapolated from known building techniques to average 11 feet with some deeper-set taller posts ranging from 12 to 14 feet.
The taller posts may have been used for special functions like the two associated with the gate accommodating a set of hinged swinging doors on the south wall. This was the only provenience where numerous modified antler tips and tines were found.
The hypothesized cabins on the north wall each contained shallow, rectangular subfloor pit cellars (Features 2 and 19) with very high artifact and faunal density. These were the two earliest features, and interestingly, they also produced most of the site’s gunflints and lead balls. Though there is no written record of Williams Fort having been attacked, three Kaskaskia iron arrow points recovered here and at a neighboring farmstead that was excavated at the same time offer strong evidence of conflict with the local Indians.
The site’s first well (Feature 16) was found under and just outside the stockade wall trench. This feature offers the best evidence for the construction of the fort after the site was already established. This well is thought to have collapsed during the 1811 earthquake and might be the reason it was filled and not included within the stockade.
Three activity areas were identified outside of the stockade: Area 1, in the center; Area 2, to the south; and Area 3, the site periphery to the north and east. Artifact assemblages from the features suggest that there was a progression of development from north to south at the site.
The most intensively used activity area outside the enclosure was Area 1, which is adjacent to the southeastern corner of the stockade and is associated with the farmstead and tavern. Area 2 is located about 20 m south of the fort. The types of features, artifacts, and density of faunal remains support food processing activities, such as smoking, curing, and butchering occurring here.
The well (Feature 42/50) in Area 2 is the site’s second well, though it was still built very early, probably right after the first well collapsed during the 1811 earthquake. Due to its location directly south of the gate, it would have been easily accessible from the fort when confined within the stockade. This well was located in the floor of a square cellar almost 5 feet below the surface. A large number of redware storage vessel fragments from this feature is indicative of a cold storage function on the upper “shelf” of the well. Post molds around the periphery of the feature may have been supports for a building, possibly a barn. Adjacent to the south side of the cellar and well was a midden dense with faunal remains.
Area 3 encompasses the eastern and northern peripheries of the site. The most significant aspect of this area was the identification of at least four fire pits, which are thought to represent campfires associated with the militia during the first decade of site occupation.
The largely vacant space within the stockade, in conjunction with the fact that published accounts by two early travelers in 1817 made no mention of a stockade, indicates the fortification had been dismantled early on, probably around 1815 after the potential threat of Indian attack had diminished with the close of the war.
More than 2,300 historic artifacts were collected, three-quarters of which were associated with food preparation, storage, and service vessels. The Feature 7 well directly east of the stockade was the site’s last well based on artifact analysis. The contents of this well also offer the best evidence for a rural tavern in the final years of occupation, specifically this cache of artifacts found in one massive heap in the well. Aaron and Tabitha Williams’ probate records also support the presence of a tavern based on the inventory of 31 bed sheets, two trumpets possibly used to herald arrivals and departures of stage and mail coaches, and a large number of tea and tablewares and serving vessels.
The closest redware potteries to the site would have been 8 miles northeast in New Harmony, Indiana, and 20 miles north in Albion and Wansborough, Illinois. One small redware fragment with a bright green copper glaze is French in origin. The stoneware was probably made out east since it was generally not being produced in Illinois before the 1830s. One stoneware fragment bears an incised exterior decoration painted in cobalt with an outline of a house or cabin—a design that was popular from the late 18th century to the 1840s.
Decorative motifs on tea and tablewares are primarily handpainted fineline floral soft pastel and earthtone varieties that were popular in the 1810s and 1820s.
This rich glassware assemblage is very rare for a rural frontier-period site in southern Illinois and provides the best evidence for a tavern dating to the late 1820s through the 1830s. Vessels in the collection include pattern-molded flasks, a large molded beehive bottle, large globular bottles, cylindrical bottles, tumblers, stemmed glassware, a decanter, a salt cellar, a pocket flask, a cruet or caster for condiments, and several wine or port bottles.
Among the personal and sewing artifacts was a British military button from an officer’s coat that dates to the late 18th to early 19th century. The silver coin is a cut Spanish Real 1 also known as a “piece of 8".
Based on the architectural assemblage, windows on the buildings at Williams Fort were covered with animal skins, fabric, or greased paper rather than glass, and log construction was utilized. Wells are thought to have been lined with wood. A relatively large sample of horseshoe nails and equestrian artifacts compared to other farmstead sites in the region may be indicative of a higher than average number of travelers or visitors to the site.
Williams Fort was abandoned by 1838 after Aaron Williams died. Although Tabitha lived for four more years, the 1840 census reported her living in her son’s household, probably nearby but not at the site.
The fortification was strictly civilian in nature but more extensive than was suggested by written accounts and produced important comparative data on an early 19th century fortified farmstead in southern Illinois. Although noted in the literature as a blockhouse, our excavations revealed a stockade wall with two log blockhouses at the southwest and southeast corners and a bastion flanked by two log cabins on the north wall. The Williams Fort stockade wall was taken down around 1815, and the residence and activity areas moved from the north towards the southeast and south.
Aaron and Tabitha Williams never applied for a tavern licence, and the only evidence for a tavern at the site was from the probate records and the excavation of the terminal-period well yielding a large primary deposit of drinking and food service vessels supporting a public establishment during the final decade of site occupation. This may have been a more formal expansion of activities already occurring at the site as settlers’ forts served as community centers and gathering places.