Two weeks ago, the blog highlighted some of the more interesting artifacts recovered this season. Our focus this week is on several interesting features that have revealed more of the story of life during the 18th and 19th centuries in central Pennsylvania. According to my favorite archaeology text book by David Hurst Thomas and Robert Kelly, a feature is the nonportable evidence of technology; usually fire hearths, architectural elements, artifact clusters, garbage pits, soil stains, and so on. They are artifacts but they usually cannot be removed from the ground and can only be described. Fort Hunter was occupied by Europeans for nearly 300 years and contains a very large quantity of artifacts. Unfortunately, the majority of these cannot be dated unless found in features with other diagnostic objects – that is artifacts that can be chronologically placed to a specific and limited time period. Below we will review the more interesting and significant features encountered this season - some were datable and some were not.
(East wall of foundation)
Feature 22/55 is a rock foundation that we first encountered in 2009 but misidentified as a French drain. As we expanded the area north of the well, it became clear that this feature was a building foundation. It consists of a mixture of mostly rounded cobbles but also some dressed diabase. Many of the cobbles were large, 12” – 18” in diameter. In this part of the site, it is possible to identify the original ground surface present at the time of European contact and the foundation seems to be resting on this surface. At the same level as the foundation or just below it, we uncovered numerous fire-cracked-rocks which were part of a Native American hearth feature to the east. It does not appear that a builder’s trench was dug for this foundation but rather the foundation was simply placed on the ground surface. The east wall of this structure is at least 15 feet long. As we expanded the block north, we carefully excavated around and beneath the rocks but did not find any artifacts that would suggest a date for when the foundation was constructed. Since it rests on the original ground surface, it could be early.
(Northeast corner of foundation)
This year we reached the back corner and were hoping to solve this problem. The back or north wall makes a right angle turn to the west. Unfortunately, the back wall ended within two feet where it had been cut by a waste water ditch or where a large section of the bank eroded away during the hurricane Agnes flood of 1972. To our dismay, what appeared to be postmolds turned out to be rodent disturbances and the artifacts found in them could have dated anytime during the 19th century. In the future, we will investigate the area to the west of the erosion ditch and hopefully, uncover the west wall of the foundation.
(Feature 24/48 with Andrea standing on the iron pipe in the bottom leading to the well)
Another feature that attracted our attention this season was Feature 24/48. This was a large hole dug adjacent to the well but it also extends to the north for approximately ten feet. It was first identified in plan-view in 2009 as a large circular stain adjacent to the well. Our excavation eventually extended to a depth of 6.6 feet and an iron pipe was found at the bottom that apparently was part of a pumping system for the well. Based on historic photos from the 1890’s, there was a windmill approximately ten feet to the west and it is assumed that it pumped water out from the well. This feature was not completely excavated because part of it extended into the west wall. During last winter, the adjacent unit to the north slumped and exposed more of the feature. While troweling the wall this year, we recovered the U.S. Navy button dating to approximately 1809.
(Navy button probably part of the uniform worn by Thomas Gates McAllister, son of Archibald McAllister, who served in the U. S. Navy from 1805 to 1807.)
After the slump was cleared, approximately 15 inches of the feature was exposed in the floor and considering the date on the button, this required our attention. The button and several pieces of Middle Woodland pottery were the earliest artifacts found, however the majority were post 1850 in age and not particularly chronologically diagnostic. This portion of the feature was over seven feet in depth and revealed a mostly decayed log situated upright with an iron bar extending perpendicular through it. Most of the log had decayed but part was covered in creosote and therefore was preserved. Initially we speculated that the log was a wooden pipe that had been inserted into the ground. However, the iron bar suggested another scenario. Since the pipe connecting the windmill and the well are in the same excavation pit, we are now thinking that this log was part of this construction activity. In this scenario, the log was placed into the hand dug pit upright and the soil was filled in around it to secure it in-place. The log functioned as a “dead man” to which cables were attached to secure an adjacent structure; in this case most likely the windmill. We are reasonably sure that the windmill and pipe were part of the improvements made by the Boas family when they purchased the property in the 1870s. Therefore both structures probably date to that time. This does not help us much when interpreting the 18th century occupation but it does establish the construction chronology in this area of the site.
(Profile of Feature 24/48 illustrating “dead man” on the left)
Feature 77/90 is located at the back or north end of the icehouse. Originally, these were considered two separate features but after further excavation, they connected. The Feature 77 section is an 18” by 24” rectangular opening in the icehouse wall with a lining of motar on the bottom. The opening extends approximately two feet below the present ground surface and is approximately 6” below the current wooden floor of the icehouse. The opening does not extend inside the structure but is blocked by a dried laid brick wall. It is associated with a dark stain outside the wall that extended approximately 12’’below the plastered surface. Middle to late 19th century artifacts were recovered from both the opening in the wall and the dark stain area outside the wall.
(Opening in icehouse wall. Brick in back of the opening)
The Feature 90 section started out as a poorly defined stain extending north of the opening in the wall approximately 8 feet. It did not really take shape until the stain had been excavated to a depth of 8”. Rough cut wooden sides were exposed at this point with several long spikes protruding vertically from the sides. The spikes suggest that wood was attached to the top. It appears to have been a rectangular box extending from near the icehouse wall, north, possibly funneling water over the cliff that drops 36 feet to Fishing Creek. Depending on which records are accurate, the icehouse is either 15 or over 20 feet deep with stairs leading to the bottom. The records describe the ice as being stored on the bottom floor but, if the ice is stored on the bottom, why is there a drain at the top? Possibly the historic records are not accurate and the icehouse is not constructed deep into the ground and this feature acted as a drain for the melting ice. We have considered augering into the floor of the icehouse to determine its depth. However, augering inside a building would be very difficult since there is less than eight feet of roof clearance.
(Feature 90 with wood and nails visible on the left side)
The most productive feature we excavated this season in terms of 18th century artifacts was Feature 62. This was first identified in 2011. It appeared as a dark linear stain under the topsoil that produced a French gunflint and scratch blue salt glaze stoneware. This year, we recovered significantly more stoneware along with tin glaze earthenware. A preliminary analysis suggests these represent pieces of three or four vessels of scratch blue and at least two vessels of tin glaze earthenware. Unfortunately, these were found with 19th century artifacts. The stain was associated with a line of mostly dressed rocks but some rounded cobbles. It extended approximately 15 feet along the top of the slope leading down to the edge of the cliff overlooking Fishing Creek. The 18th century artifacts were generally confined to a six foot long section but, it seems to have been part of a dumping area that was used well into the 19th century. Several pieces of the stoneware and the earthenware have been mended and it will be interesting to see if there are any re-fits from other parts of the site. For example, the area around the bake oven produced a considerable quantity of similar stoneware and they also might cross-mend. Feature 62 produced more 18th century pottery than we have recovered in many years at Fort Hunter, unfortunately all of it was from a mixed context.
(the alignment of rocks to the left of the rock foundation in the center is Feature 62)
As is frequently the case, one of the most intriguing features was uncovered the last two weeks of the excavation. This is a circular rock foundation, 12 feet in diameter. Many of the rocks are large cobbles similar to Feature 22/55 but there is also a mixture of smaller cobbles and dressed pieces of diabase. However, in this case, there is a significant builder’s trench that is about 18 inches deep. The rocks do not appear to be aligned for a foundation but seem to have been disturbed. Possibly, the foundation was partially removed (robbed) to be used in some other structure. There also seems to be a small open ended rectangular structure on its north side. Artifacts are not common but most date to the 18th century. These include scratch blue stoneware, gunflint, musket ball and the near complete lock from a Brown Bess musket. Along with these early artifacts, the low density of artifacts also suggests that this is an early feature. Once Captain McAllister arrived in 1786, his artifacts became very common. Therefore, the low density of artifacts in this feature suggests a pre-McAllister structure. However, Feature 99 cuts through another feature that seems more recent, suggesting the foundation may not be as old as we think. The inside of the circle is disturbed but not to any depth so it is not a large well or cistern. We have discussed the possibly that it was one of the fort’s bastions or its powder magazine. However, a re-examination of historic records discovered that in 1798, there was an octagon shaped smokehouse in the back yard. It was elevated off the ground and smoke was produced via a stove on the outside of the structure – probably the rectangular structure to the north of the circle. However, the wooden smoke house was possibly constructed on top of a fort related structure. Clearly more excavation is required in and around the rectangular attachment and the features in the center of the rock circle. This feature will be a focus of our activity in the fall of 2015.
(Circular foundation, builders trench in profile and attached rectangular structure to the north)
Although, this season produced a relatively high frequency of 18th century artifacts, some of which may be fort related, with the exception of the hillside dump, all of these features could date to Archibald McAllister’s occupation rather than the French and Indian War fort. The features have also given us direction for excavation in the future. Practically all of these features are examples of the mystery and frustration involved in archaeological field projects. We spend considerable time in the excavation of features but their function and age frequently remain problematic.
Thomas, David Hurst and Robert Kelly
2007 Archaeology Down to Earth. Thomson Wadsworth Publishing, Australia.