In lieu of an abstract, here is a brief excerpt of the content:

sampling error. The 1930s excavators most certainly had a bias against nondiagnostic chipped stone, as shown by the 13:1 ratio of stone to sherds in the 1999 collection compared with a ratio of 0.01:1 in the 1933–1934 collection. Even tools were sometimes missed in the 1930s if they were small. The 1999 collection from the apron west of Mound A also contains historic items. None of these are diagnostically of Civil War origin. Some, such as the redware and whiteware sherds, and possibly the brick fragments, are likely to originate from the summer cottage used by Park Superintendent Delong Rice from 1914 to 1929 (Beditz 1980a:30; Smith 2004:125). Others no doubt come from foot traf¤c along the gravel path (once asphalt) leading to the stairs up the mound; this is clearly the origin of the asphalt listed in Table 7.4 and the limestone pebbles listed in Table 7.3. Several of the items listed in Table 7.4 are, ironically, of archaeological origin. The wood fragments are pieces of sawn and milled lumber, similar to the wooden grid stakes used by Roberts. Some of these pieces of wood were recovered from within the back¤lled trench, which establishes their origin securely. The plastic labeling tape bears embossed letters spelling “JERRY” and may have come off of one of Gerald Smith’s tools when he was installing a grid monument on the apron in 1976. Of course, the label could have fallen off a tourist’s possession, but that makes a less interesting story. The principal reason for excavating test units on the apron west of Mound A was to relocate the 1933–1934 trenches and thus be able to match their grid coordinates to the 1998 Welch grid. In this, the excavations were a quali¤ed success. The edges of the north-south trench were found, permitting the eastwest coordinates of the two grid systems to be matched within a few centimeters . However, the failure to delineate the intersection of the north-south trench with the east-west trench leaves room for a grid mismatch of 20 to 30 cm. The research design also mentioned the possibility of reexcavating a portion of the deep trench to expose the apron’s stratigraphic pro¤le. Pressures of time and questions about the advisability of using a backhoe to accomplish the reexcavation left this goal unful¤lled. The third locus of excavations by the SEAC archaeologists in 1999 was the “Dike.” This name appears on the 1933–1934 site map, attached to a linear ridge that is visible for 180 m or more (see Figure 7.8). There is no indication that Roberts excavated any part of this feature (his ¤eld notes contain several mentions of excavations on a dike, but those are references to the much broader, built-up ridge running west from Mound D). With the other earthworks at the site pretty well accounted for, this unusual feature has elicited puzzlement from every archaeologist who has seen it. The ridge is suf¤ciently close to the historic Browns Landing Road that a historic origin cannot be ruled out, but there is no evident reason for such a feature to have been 1999 SEAC Fieldwork 247 constructed in Historic times. The orientation and location of the ridge would not have made it useful at any point during the Civil War battle, so a Civil War origin seems unlikely. By default we are left with a prehistoric origin. To gain additional information about the feature, a 1-m-wide trench was excavated from the southern edge to the center, orthogonal to the ridge, at a point where the ridge is 50–60 cm high and particularly well de¤ned. A GPR transect across the ridge at this point showed a radar re®ection about 60 cm deep, presumably the surface under the ridge. This in itself is a signi¤cant observation, indicating that the ridge is a built-up feature rather than a ridge left high by the removal of soil to either side (as by the erosion documented in Chapter 5). On excavation, however, the ridge proved to be just as puzzling inside as it is on the outside. The ridge is composed of compact silt with astonishing quantities of sesquioxide (iron-manganese) concretions. In the upper 40 cm of the ridge ¤ll, the concretions ranged from pea-sized to marble-sized, but in the lower 20 cm they were as large as golf...

Share