CHAPTER 3

EXCAVATIONS

EXCAVATIONS: 1885

The earliest recorded archaeological investigation of this section of the Little Tennessee River Valley was carried out in the late nineteenth century by the Bureau of Ethnology, Smithsonian Institution under the supervision of Cyrus Thomas (Thomas 1894). In 1885 John W. Emmert concluded the investigation of the region surrounding Tomotley (Emmert 1885). Emmert identified three mounds in an area Thomas states corresponds with the location of "Tommotley" shown on Timberlake's map (Thomas 1894:379-380). Although Emmert refers to only three mounds in his field notes and on his field map, Thomas shows four mounds in the report. Thomas' report seems to have been in error on this point.

Because Emmert's primary interest within this region of the valley was directed towards the testing of the "Toco" mound (Toqua, 40MR6) situated just upstream, only limited excavations were undertaken on the mounds located near Tomotley. The mounds designated 7 and 8 by Thomas consisted of extremely compact fill. Emmert's notes describe Mound 7 as being 8ft high and having a diameter of 58ft. A child's skeleton with associated shell beads was uncovered 2.5ft below the surface of this mound. According to Thomas four skeletons and one discoidal were recovered from Mound 9. Emmert's notes, however, reveal that this material was actually associated with Mound 8 (Thomas 1894:387-388 and Emmert 1885). No other material was recovered from these mounds.

Excavations: 1967

The next subsurface examination of this region occured in August, 1967 (Milligan 1969). In an attempt to locate the Cherokee village of Tomotley, testing was undertaken by Charles Faulkner of the University of Tennessee, under the auspices of the National Park Service and the Tellico Archaeological Project. The area tested was located on the second terrace of the Little Tennessee River at 35o34'20" North and 84o11'30" West. Surface reconnaissance of the area produced a small quantity of Cherokee sherds but because the field season was nearly over only limited subsurface testing was undertaken.

The testing consisted of the excavation of two testpits and two test trenches. Four features, designated Features 1-4, were discovered containing ceramics, lithics, Euro-American trade items, and faunal remains. All units were trowel sorted. Approximately two buckets of soil recovered from Feature 4 was waterscreened (Richard Polhemus personal communication 1981). All excavated units were backfilled at the conclusion of the testing. The artifacts recovered from the finescreen were placed with the trowel-sorted material and analyzed collectively for the 1969 Interim Report (Salo 1969 and Milligan 1969). In the report Milligan concluded that the best way to recover sufficient Overhill material needed to verify the actual location of Tomotley would be to expose subsurface features by stripping large portions of the plowzone away with heavy equipment.

Excavations: 1973-74

In 1973 and 1974 the Tellico Archaeological Project, under Principal Investigator Dr. Alfred Guthe, returned to this region to resume the search for Tomotley (Guthe and Bistline 1978). Permission to excavate was obtained from Francis McCammon, whose property was situated on the first terrace approximately 1400 feet north of the 1967 excavations at 35o34'40" North and 84o11'16" West. The field supervisor for these excavations was J. Worth Greene.

The grid system incorporated during these two field seasons was setup along a base line which ran perpendicular to the main axis of the river terrace and parallel to a local farm road. Grid north was 21 degrees west of magnetic north. A datum point was defined at the intersection of the baseline and the tree line located along the river's edge. Spatial coordinates were calibrated according to the relative position of each locus to the base line and the datum point. This system was denoted: (feet west of the datum point) L (feet to the left of the base line) and (feet west of the datum point) R (feet to the right of the base line) (Guthe and Bistline 1978).

During the 1973 season 11,400 square feet of plowzone were skim shoveled. In 1974 59,800 square feet were excavated using a self-loading scraper to strip the plowzone. The majority of the material recovered was associated with a late Dallas occupation (Mississippian III). However, stylistically the ceramic material compares quite well with that described as Lamar associated with Middle and Lower Cherokee (Egloff 1967 and King 1972:vii) and quantitatively does not appear to be similar to the ceramics recovered from the Dallas component at Toqua (40MR6). A more distinctive Cherokee component was represented by 32 refuse-containing pits and eight burials. A second area located approximately 1400 feet south of the main excavations was also examined. Several 24in wide backhoe trenches were placed in an area believed to correspond with the location of the townhouse shown on Timberlake's map. No cultural material was recovered from these plowzone tests and, at the insistence of the property owner, the work was terminated (Guthe and Bistline 1978:10-11).

Excavations: 1976

Because the 1973-74 field seasons failed to conclusively locate the major area in which Timberlake's Tomotley was located and because the early maps suggested that the village was actually located on the second terrace, field work was resumed in the spring of 1976. Under the field supervision of Dr. Gerald Schroedl, survey and testing was begun in April of that year in an effort to locate the most potentially productive areas for later excavation. This stage of the endeavor concentrated on the reconnaissance of three demarcated areas: T.V.A. property tracts 2233, 3409, and 3410.

Tract 2233, formerly owned by Mr. Wayne Curtis, was the first to be tested. Using a backhoe to remove the plowzone, 26 test pits, 10 by 15 feet in size, were placed on the second terrace. The initial three pits, located along the front edge of the terrace, exposed five refuse-containing pits (designated Features A-E) and one burial. Each of the pits was trowel sorted and then the fill was returned along with a beverage can labeled with the unit's designation in the event that this region was more extensively examined later. The burial was only inspected enough to verify that it was indeed a burial and then it was similarly refilled.

A major portion of all the test pits were placed along the front and back edges of the second terrace (T-2). One of these units exposed a portion of the 1967 excavations. One pit was placed on the T-1 and three were situated on the T-3. A midden, approximately one foot in depth, was located on the upstream portion of the T-2 and a few postmolds were observed along the front edge of the terrace. In general no other significant material or subsurface features were located by these tests.

Work was then shifted to Tract 3409. This area had been sowed to wheat by Mr. James Atchley. With Mr. Atchley's permission three 10 by 10ft units were hand excavated in the field. These units were designated Test Pits 1, 2, and 3. The reluctance of Mr. Atchley to have any further tests disturb his crops prevented any other work from being done. Surface reconnaissance located the site of an early to middle nineteenth century structure (40MR30) along the front edge of the T-2 adjacent to the property line separating Tract 2233 from 3409. At the time it was believed that this area may have been tested earlier by J.W. Greene although imprecise field notes from that field season prevented the confirmation of this.

The final area tested that spring was Tract 3410. This property was located to the south of Tract 3409 and included the area adjacent to Toqua Creek. It was believed that this area would produce significant finds since Timberlake's 1762 map shows outlying structures in this region and, because, borrowing along the T-2 of the creek in 1975 had exposed a Cherokee pit. Unfortunately, continued borrowing had, by the spring of 1976, removed all traces of the once exposed Cherokee material. Most of the material collected in 1976 was associated with Archaic and Woodland occupations.

Testing was then moved to the front edge of the second terrace, northeast of the borrowed area. The field in this location had been prepared for planting corn which facilitated surface reconnaissance. After securing permission to do limited subsurface testing five test pits were placed in areas where surface finds suggested artifact concentrations. A majority of the artifacts recovered were associated with Archaic, Woodland, and nineteenth century American occupations of the area. Test Pit 4 exposed four features which were designated Features A-D. Although recovered shell tempered ceramics implied a Cherokee origin the features were not totally excavated. Stakes were numbered and, along with a beverage can, were driven into the pits in a manner similar to that employed during the Tract 2233 tests. It was then learned that future archaeological work on this property would not be possible due to the nature of the stipulations of the agricultural lease. All test units were subsequently backfilled. Of the three land tracts only 2233 was available for extensive excavation so, work was shifted to this region in early June.

Beyond the testing described above the goals of the 1976 excavations were to:

  1. "determine the location, size, plan, and internal arrangement of the townhouse,
  2. specify the nature of domestic structures and their associated burials and features,
  3. determine the village plan,
  4. acquire sufficient artifact samples for defining a Cherokee cultural assemblage,
  5. obtain sufficient samples of faunal and botanical remains for specifying subsistence resources,
  6. and provide comparative data for continuing studies of Cherokee settlement patterns, mortuary practices, resource utilization, and acculturation." (Schroedl 1976)

Excavation techniques: 1976

The following field techniques were used to provide the data needed to address these goals. The plowzone was removed by a self-loading pan provided by T.V.A. and operated by Norman Kennedy (Figure 3.1). Several 50-60 feet long test strips approximately one foot in depth were placed at various locations on the T-2 with particular emphasis placed on those areas which produced material during the earlier tests. When evidence of an occupation was observed, surrounding areas were expanded. Six areas were eventually opened in this manner (Figures 3.2 and 3.3).

The exposed surfaces were then skim shoveled to clarify the location of all subsurface features. After the opening perimeter was recorded, one half of each non-burial pit was excavated to reveal the profile for recording and then the remaining half was removed. The length, width, and depth of each pit was recorded to the nearest tenth of a foot. Pit fill lying above all identified burial interments was removed but not screened since excavation experience at Chota (40MR2) indicated very little material would be recovered. Body fill was carefully excavated and waterscreened. All postmolds were cored using a soil auger to determine their depth. The depth along with the radius was recorded to the nearest tenth of a foot. Each unit was mapped onto plot sheets at a scale of 1:24 through the use of a plane table/alidade. All excavated fill was carried to a waterscreen unit by a Melroe bobcat and passed through half inch, quarter inch, and window mesh screens. The material retained on the window mesh was designated finescreen and bagged separately from that trapped by the larger screens. In an effort to decrease the large ceramic sample size, all shell tempered sherds were passed through a half inch screen with only those trapped being saved for examination. With most of the smaller sherds left at the field camp, the insignificant number of recovered finescreen sherds were not examined for this report. Selected portions of 17 units had floatation samples taken (Features 295, 300, 302, 305, 313, 325, 331, 342, 343, 347, 349, 350, 357, 393, 401, 402, and 412).

The only deviations from these procedures involved the excavation of the eight main support posts of the townhouse (Postmolds 5025-5032), the profiling of three square Anglo-American postmolds (Postmolds 5908, 5931, and 5933), the floatation of the total contents of the cob pits, waterscreening only one bobcat load of fill from Features 358-361, and not waterscreening any of Feature 401. The main support posts were excavated to record their profile. The cob pit fill was floated to recover the dense concentrations of botanical residues. The square postmolds were profiled in order to determine the nature of their construction and probable origin. Features 358-361 were seen as probable Woodland pits which, upon inspection, contained virtually no artifacts and therefore time was not expended screening all of the fill.

Schroedl, in consultation with Dr. Alfred Guthe, agreed to use the grid system employed by Greene in 1973-74 as well as make the feature, burial, postmold, and structure designations consecutive with those of the earlier excavations. The angular orientation of the areas with respect to one another was due to the fact that most of the units were panned prior to the selection of the grid system (recall that Greene's grid was aligned 21 degrees west of magnetic north).

Six areas were opened which, combined, covered 98,595 square feet (see Figure 1.2). Table 3.1 lists the size and contents of each excavated area. Maps showing the locations of refuse-containing pits (designated "features" for simplification), burials, and postmolds by area of excavation are presented in Figures 3.4-3.10.