CHAPTER 5

ASSEMBLAGE ANALYSIS

The material recovered at 40MR5 can be sorted into six contextual sets: surface recovered material, material from test units, material from tested features (partially excavated), material from postmolds, material from refuse-containing pits, and material from identifiable burials. Four recovery techniques were used during the 1967 and 1976 field seasons: surface collecting, trowel sorting, waterscreening, and floatation. Pairing recovery contexts with recovery techniques produces three collection groups which internally represent statistically comparable samples: surface collections, tested units, and village excavations (i.e. material from Areas 1-6). This chapter discusses the artifact collections recovered from each of these samples in terms of their cultural significance. A total site inventory is provided in Appendix IV.

Surface Collections

Material was recovered in 1967 from Tract 2233 and in 1976 from Tracts 2233 and 3410. These collections produced samples of aboriginal ceramics (n=122), lithics (n=849), faunal remains (n=33), and Euro-American items (n=66). This sample represents the following percentages of the total number of artifacts of each class recovered from the site: Ceramics: 0.868; Lithics: 16.960; Faunal: 0.354; Euro-American: 0.640.

Surface collections represent, approximately, a 7-12 percent sample of the plowzone (Baden 1982b). Even allowing for the effect of such a low sample fraction, the second terrace deposits at 40MR5 seem to contain relatively little material. This observation is supported by the results of a 1979 surface reconnaissance of the non-excavated areas in the vicinity of the village excavations (Kimball 1980c), where a total of 32,400 ft.2 of exposed plowzone was collected. Only five sherds and 63 lithic artifacts were recovered. In contrast to similar collections made at other sites with major occupations (40MR3, 40MR20, 40MR23, 40MR40, and 40LD24), these totals are extremely low. It is not surprising that the search for Tomotley required the use of heavy equipment to expose subsurface features. As will be shown later, most of the artifacts from this site are to be found in a few, widely separated features. A surface collection from such a site would consist of dispersed, small clusters of material rather than heavy concentrations of artifacts.

Ceramics

The ceramic sample was composed of 113 shell, three limestone, two coarse quartz, one fine quartz, and three grit tempered sherds:

Shell 76 Plain 16 Checked stamped 2 Rectangular checked stamped 1 Rectilinear complicated stamped 3 Indeterminate stamped 1 S-twist cord marked 2 Loop handles 4 Residual 7 Destroyed 1 Detached plain lip

Limestone 2 Plain 1 Indeterminate cord marked

Coarse quartz 1 Plain 1 Checked stamped

Fine quartz 1 Rectilinear complicated stamped

Grit 2 Plain 1 Check stamped

The major components represented in this collection are Mississippian IV (Cherokee) and Mississippian I/II (Martin Farm/Hiwassee Island). With the exception of the two loop handles, the relative frequencies of these types generally correspond with the composition of the excavated units and are reflective of the ceramic producing components at the site.

Lithics

A total of 849 lithic artifacts were recovered from all surface contexts. The surface collection produced a majority of the amorphous cores (n=25), nodules of utilizable materials (n=28), unmodified/preforms (n=2), projectile point fragment/preform (n=7), Lower Kirk corner notched (n=5), Upper Kirk corner notched (n=1), Morrow Mountain I (n=2), Savannah River stemmed (n=1), fragments of probable Archaic points (n=6), and serrated Upper Kirk corner notched (n=29). Based on this sample, these artifacts are more likely distributed within the plowzone than in subsurface features. Of particular importance is the high density of Archaic period projectile points found on the surface rather than in pits. This contrasts with a high density of Mississippian points found in features. This suggests that formalized tools associated with the earliest occupations at the site came from refuse scatters throughout the plowzone rather than from subsurface features as is the case with the material from later components. When temporal parameters are defined on the basis of lithic time markers (classic types), early components should be expected to be over represented in surface collections taken on such second terrace deposits. Understanding the site structure of these early components requires careful spatial examination of the lithic contents of the whole plowzone and recognizing the distinctive differences in refuse disposal modes between early and late occupations (see Kimball 1981).

Euro-American items

A total of 66 Euro-American items were recovered from the surface. Diagnostic artifacts included one gun butt plate, one gunflint, one trigger plate, and 20 ceramic sherds. The ceramic material consisted of three transfer print earthenware sherds, three saltglazed stoneware sherds, three creamware sherds, six pearlware sherds, one stoneware sherd, one industrial non-glazed stoneware sherd, and two nonspecific glazed earthenware sherds. This material represents 55.6 percent of the total Euro-American ceramic artifacts recovered from the site. Based on the temporal associations of these sherds (see Appendix II), both Colonial and Federal Period occupations existed at the site. The Colonial material could be associated with the Cherokee village occupation and the Federal material could be associated with either the Nettlecarrier's, Sarah West's, or a subsequent American occupation.

Fauna

Only 33 faunal elements were recovered from surface contexts. Indeterminate mammal bone comprised 54.6 percent (n=18) of this sample. The remainder consisted of one cow/elk/bison ilium fragment, four white-tailed deer elements (one burnt phalange, one metatarsal, and two teeth), one second phalanx from a horse, one right mandible from an opossum, one indeterminate element from a turtle, two boxturtle elements, four indeterminate bivalves, and one Actinonaias ligamentina bivalve. The general density of the remains are consistent with the overall sample from all contexts of the site.

Tested Units

Tested contexts consist of excavations that either revealed no culturally produced features or contained features which were only trowel sorted or partially excavated. The specific contexts include: two trenches dug in 1967, three test pits dug on Tract 3409, two test pits dug on Tract 3410, Features A and C from Tract 3410, Feature D from Tract 3409, and Feature E from Tract 2233.

Ceramics

One hundred sherds were recovered from these units with 30 specimens coming from Trench 1 in 1967, one from the test pits on tract 3409, and 79 from the tested features. The sherds removed from the two trenches were all shell tempered Mississippian IV types with more stamped (n=18) than plain (n=10) motifs. Test Pit 2 produced one plain limestone tempered sherd (Mississippian I/II). The ceramic contents of the tested features indicate Mississippian IV affiliations for Features A (Tract 3410), C (Tract 3410), and D (Tract 3409). Feature E (Tract 2233) contained 19 plain limestone sherds and is probably associated with a Mississippian I/II component.

Lithics

A total of 103 lithic artifacts were removed from the tested contexts. Of the seven artifacts recovered from tested features none were diagnostic of any time period. The two test trenches produced only two flakes. Test Pit 1 (Tract 3409) contained 44 artifacts including two Lower Kirk end scrapers, one Upper Kirk end scraper, and one steatite pipe fragment. Tract 3409's third test pit contained ten lithic artifacts including one probable Archaic projectile point fragment. Tract 3410's second test pit held 34 artifacts of which ten were undiagnostic bifaces. In terms of the lithics, the tested units, like the surface collections, provide scattered evidence of Archaic occupations on the second terrace with no sizeable lithic concentrations from trowel sorted feature contexts.

Euro-American items

A single kaolin pipe bowl fragment contributed the only Euro-American item from the tested units. It was recovered from Feature C in Tract 3410. With the ceramic and lithic assemblage from this unit revealing no evidence of pre-Mississippian IV material, this unit is probably of Cherokee origin.

Fauna

A total of 53 faunal elements were recovered from these test units. Of the total, 39 were indeterminate mammal bones. The remaining elements consisted of one turtle hyoplastrom (1967 Trench 1); one Actinonaias ligamentina bivalve (Feature E, Tract 2233); and from Feature A (Tract 3410) one black bear element, four white-tailed deer elements, one horse tooth, four bivalves, and two Actinonais ligamentina bivalves. Based on the total unit assemblage of Feature A, it is probably associated with the Mississippian IV occupation.

Village Excavations

The 1976 excavations at 40MR5 stripped open six units, designated Areas 1, 2, 3, 4, 5, and 6 (see Chapter 3 and Figures 3.5-3.10). The purpose of this section is to analyze the assemblage of burial, refuse-containing pit, and postmold contents to define the village patterns at the site. All of the lithic, faunal, and Euro-American artifacts recovered from the finescreening were excluded from the examination in order to treat each artifact class and each recovery context in a statistically valid and comparable way. Except for the presumed Woodland features which were not totally screened (Features 358-361) and the floatation recovered material from the cob pits (Features 302-307, 347, 357, 402, and 412) all recovered contexts can be considered comparable and suitable for intrasite analysis.

Detailed intersite comparisons, although important, are somewhat difficult to make due to the absence of necessary assemblage information in many published reports. The absence of substantial reports on excavations of Lower and Valley Town Cherokee sites makes comparisons between different Cherokee subgroups extremely difficult. Where possible, intersite similarities and differences in the archaeological record are noted and compared with expectations based on the historic documents. By interfacing the two sources of information, a credible statement of historical and cultural significance is possible. Future information from other sites can then be added to this analysis process to produce a more complete picture of Cherokee cultural development.

Structures

Traditional Cherokee village patterns, as described by Timberlake (1927), DeBrahm (1971), and Bartram (1928,1853), consisted of paired winter and summer structures, one townhouse, and one summer pavilion. The winter dwellings were circular structures (or nearly so) with four central roof supports. they tended to be approximately 30ft in diameter (DeBrahm 1971:110) and constructed in such a manner (wattle and daub) as to efficiently retain heat. The summer structures were rectangular buildings approximately 12-14ft wide and 20-70ft long. Their outer walls, like the winter houses, were composed of a wicker-work mesh of canes or small trees intertwined around vertical posts and plastered with daub. When such structures are burned the daub becomes fired and is often preserved in the archaeological record. DeBrahm (1971:110) noted one small door in the winter structures and two opposing entrances in most summer buildings. He also noted that the winter dwellings were constructed and used by two or more families. Although such paired dwellings were observed well into the nineteenth century, these vertical post buildings appear to have been slowly replaced by single, horizontal log structures,beginning in the 1780's (Steiner and de Schweinitz 1928). The replacement process seems to be related to the level of acculturation and personal wealth of the inhabitants. Steiner and de Schweinitz describe considerable variability in house forms as late as 1799 (1928:479,485,490).

The excavations at Tomotley did not expose any circular winter structures. The most prominent habitation dwelling (n=8), as identified from field and laboratory observations of postmold alignments, were rectangular in configuration. Traditional winter structures were either obliterated by later disturbances (including the stripping of the plowzone during excavation) or were never built in the portion of the site exposed in 1976. The most interesting characteristic of the rectangular structures is their apparent clustering into two or three member groups similar to Bartram's description of Creek villages (1853). These clusters include:

Structures 16 and 17 in Area 1
Structures 24, 25, and 26 in Area 4
Structures 30 and 31 in Area 5.

To what extent this arrangement agrees with that found in the Lower and Valley Towns cannot be confidently confirmed due to the absence of suitable data on excavations from these areas. Limited information on excavations of the Lower Town of Toxaway and the colonial Ft. Moore (38AK4) suggests that such structures were used (Polhemus 1975). An engraved powder horn documenting the 1761 Grant expedition uses labeled images of rectangular structures to record the villages burned (Polhemus 1975 and King 1976). No other structural forms appear on the engraving.

This pattern differs greatly from that exposed at other excavated Cherokee sites in the lower Little Tennessee valley. At Chota/Tanasee (40MR2/62) 32 clearly identified structures (not including two superimposed townhouses) were uncovered. Of these, ten pairs of winter/summer structures, seven isolated circular patterns (probable winter structures), and four isolated rectangular structures were identified (Schroedl 1982:611). The 1978 excavations at Citico (40MR7) uncovered eleven recognizable structures. Of these, there were two pairs of winter/summer structures; one small, circular pattern (hothouse?); one square structure; and five rectangular dwellings. All but one of these (rectangular Structure 14 - Mississippian II) were attributed to the historic occupation. Structure 18 was described as being similar to those found at Tomotley (Chapman 1979). Structure 18 is also the only case outside of Tomotley where a possible second structure is located near enough to form a Tomotley-like cluster (the partially exposed Structure 19). The excavations at Toqua (40MR6) uncovered, in addition to the Dallas (Mississippian III) structures, one pair of winter/summer structures, two circular patterns, and two rectangular structures. One pair of winter/summer structures was identified in the 1979 excavations at Wear Bend (40LD107) (Chapman 1980).

Thus, the predominate pattern of dwelling structures associated with historic aboriginal sites along the Lower Little Tennessee River is the paired winter/summer configuration. The only other indication of a Tomotley cluster pattern was observed at Citico. The lack of substantive documentation on Lower, Middle, and Valley village patterns prevents more than a handwaving argument that the Tomotley pattern is indicative of the 1750-1760 refugee phenomenon and distinguishes the village from the other so-called Overhill Cherokee sites of the same time period.

The next most numerous structure form is the square pattern (n=7 clearly identified structures). Such a configuration is commonly associated with Mississippian II and III houses but, apparently, it is also associated with Middle Cherokee villages (Dickens 1970:224,1976:100). Of these, Structure 21 in Area 3 appears to be a classic Hiwassee Island (Mississippian II) house (Lewis and Kneberg 1949:48-79) based on the postmold pattern and associated materials recovered from Feature 325. Structure 34 (Areas 3 and 6) is a partially excavated, large, square structure (41.9X36.3ft) similar to the round cornered townhouses at Toqua (Schroedl 1978a) and Coweeta Creek in North Carolina (Dickens 1970:225). No diagnostic materials were recovered from this structure. It might be hypothesized that these large square structures are related to a Middle Town component among the Overhill dating to the 1760's retreat from Grant's expedition. All other square structures located at the site are assumed to be associated with the historic Mississippian IV component based on burial and feature contents or the lack of other diagnostics.

In Area 6 an octagonal townhouse (Structure 28, Figure 5.1) with eight main roof support posts was exposed which resembled the Creek townhouse described by Hawkins (1938:67-68) and the two superimposed townhouses at Chota. South of this structure stood an apparent single summer pavilion (Structure 29) common in Cherokee villages and like that uncovered at Chota. The only hearth identified during the excavations was located in the townhouse (Feature 293). A relatively clear area existed to the south of these structures suggesting the possible location of the ceremonial grounds (sometimes referred to as the chunky yard).

Construction consistencies should be expected within any village complex. The archaeological residue of such patterning has been recognized in the postmold alignments which serve as first approximations useful in distinguishing structures from apparently randomly scattered posts. Further patterning might be expected in terms of postmold radii and depths, if it is assumed construction components must fulfill certain prescribed criteria. An examination of these variables for Cherokee structures from Toqua, Chota, Tanasee, Mialoquo, and Tomotley was undertaken by Richard Anuskiewicz (1981) and used previously to define the structural postmold statistics in Table 4.5. An examination of the variability in radii from Tomotley postmolds suggested that this variable would be unsuitable for defining constructional patterning. The depths, however, seemed to fall consistently between 0.1 and 0.3ft (remember that approximately one foot had been removed). To test the premise that posts of such excavation depths would most likely be associated with structures and not found consistently elsewhere, a map was produced which contained only those posts with depths falling in that range (Figure 5.2). Another map was produced which showed those postmolds with depths between 0.4 and 0.6ft (Figure 5.3).

The resulting patterns suggest while some buildings, like Structure 24, maintain higher recognition integrity others, like Structures 21 and 32, virtually disappear due to differential selection in the mapping procedure. Based on these maps, there are four types of excavated structural patterns at Tomotley. The first consists of structures recognizable on both maps, whose outer walls are composed of essentially equally spaced shallow and deep posts (Structures 17, 24, and 28). The second group is composed of structures which have primarily shallow posts (Structures 18, 20, and 29). The third group's structures have mostly deep posts (Structures 19, 21, 32, and 34), with the fourth group composed of the remaining structures which tend to be unrecognizable because both shallow and deep posts are required to produce an identifiable pattern (Structures 22, 23, 25, 26, 27, 30, 31, and 33).

The inferences and observations derived from these types and their patterns include:

  1. When excavation techniques necessarily preclude the recognition of both shallow and deep postmolds, our ability to identify structures will be impaired.
  2. Most of the postmolds in the ceremonial area south of the townhouse are of the deep variety with the more shallow ones seemingly forming a rectangular boundary around it.
  3. If known structures under these situations can produce a pattern like that in Figure 5.2 where the buildings of the fourth type are reduced to unpatterned concentrations of postmolds, then it might be true that similar concentrations elsewhere on the map may represent previously unrecognized structures. Figure 5.2 shows two such possibilities: north of Feature 1 in Area 2 and around Feature 389 in Area 6.

In general, postmold patterns display useful characteristics and structures can be related to one another in a coherent way that recognizes constructional similarities and differences. Burial and refuse-containing pits can also be compared to identify patterned relationships. By quantifying thier contents in a simplified manner refuse patterning can be made apparent.

Feature and burial analysis

If the relative frequencies of the non-finescreen ceramic, lithic, faunal, and Euro-American material for each refuse-containing pit and burial are calculated in terms of the total sample (not in terms of each individual unit) it is possible to quantify the distribution of the artifacts across the site (see Figures 5.4-5.7). The ten highest density units for each material class are listed in Table 5.2. Pearson correlations between pit volume and percent composition of each material class for all 166 units, the 148 features, and the 18 burials are provided in Table 5.3. The size of the pit is not strongly correlated with any material class although there is a tendency for the high density units to be relatively large in comparison to the average pit size. The 27 (16.3 percent) high density units listed in Table 5.2 contribute 33.2 percent of the total excavated pit volume. It may have been more important to first fill in the larger pits with available refuse leaving the smaller units or those located further away from structures to fill in gradually. The results suggest differential distribution of material with most remains found in relatively few units.

High density loci tend to be associated with structure locations and, likewise, each structure tends to have at least one of these features near by. This suggests a selective disposal practice of throwing trash into specific pits. The contents of these pits suggest that the disposal generally resulted from isolated activities such as butchering one or two animals or throwing away a broken vessel. Caution should be exercised in any analytical examination of a few "representative" features where the assumption is that the contents reflect generalized behavior when in fact they may be the product of unique events. Only by examining the total variability of refuse disposal (hopefully based on a large sample) can general statements be made about behaviors and activities.

The general observations of Tomotley's distributional patterning based on these statistics are:

  1. High density units tend to be located near structures. The presence of apparently isolated high density features may suggest the location of an unobservable structure. Such a situation occurs in Area 6 near Feature 389. This loci corresponds with a postmold cluster shown in Figure 5.2 which suggested the location of a structure. When found near the edge of excavation units, these high density units may be indicative of unexposed structures.
  2. Most of the lithic material was recovered from units located along the front edge of the terrace in Area 3. This portion of the site is associated with the Woodland and early Mississippian components. The absence of significant amounts of lithic debris from Cherokee features away from the front edge of the terrace suggests lithic technology played a limited role (or no role) in colonial Cherokee culture.
  3. The faunal remains tended to be concentrated in Areas 2 and 4, although high density units can be found near structures in each of the excavated areas. Their contents generally are indicative of a single disposal activity and do not suggest long term repetitive use of the facilities for disposal.
  4. Euro-American artifacts occur most often in non-refuse association with burials. This is primarily due to the higher proportion of none-finescreen glass and metallic beads in burials (n=7976) than in all other contexts (features n=1044 and postmolds n=139). The highest density of Euro-American artifacts are found in Area 1, especially in Burials 81, 82, 83, and 84.

To describe the feature and burial assemblage those units producing more than five percent of the total non-finescreen remains from one or more of the material classes are examined in more detail (Features 4, 313, 325, 336, 383, 385, 389, 415 and Burials 82, 83, 86, 90). (All percentages and sample sizes are in terms of the total non-finescreen recovered material).

Area 1

Burial 82 - This infant or child burial was interred along the outside wall of Structure 17. This unit contained 6.1 percent of the non-finescreen Euro-American remains recovered from the excavations. Most artifacts consisted of beads, sleevelinks, and brooches which are categorized under the Personal Class of material. Tomotley Mississippian IV burials are characterized by the largest occurrence of personal items of any context. A total of 8080 personal items were recovered from all the burial contexts. This represents 97.60 percent of the Euro-American material recovered from burials (n=8279), 79.78 percent of the total recovered from all pits (burials and features, n=10128), and 87.37 percent of all the personal items (n=9248) recovered from pits and burials.

Burial 83 - This burial was a child located across from Burial 82 in Structure 17. It contained 23.5 percent of the Euro-American artifact total for pits and burials. Again, this was primarily because of the high incidence of personal items, especially beads and sleevelinks. The remaining two burials (81 and 84) in this structure contained fewer personal items and more activity and arms related materials suggesting they were adults (probably males) (cf. Schroedl 1982).

Because the children in this house were buried with necklaces and beadworked garments they have more grave goods based strictly on the total artifact count. The nature of grave goods and their constituent elements should be considered in terms of the purpose for quantifying archaeological units. In this case the goal was to locate the highest density units, regardless of the elemental constituents. Once this is accomplished more detailed observations, like those offered here, can be made discerning the differences and similarities between units. Equating one bead with one knife fragment may not always be an appropriate way to address the question of assemblage variability.

Area 2

Feature 4 - This unit is located along the south edge of the excavation unit. It contains 17.9 percent of the faunal remains from the site. Its ceramic composition primarily consists of shell tempered plain sherds (n=206). Only ten lithic artifacts were recovered, six of which were steatite pipe fragments. It also contained 42 Euro-American artifacts, including three gunflints. The unit's sizeable distance from Structure 18 suggests that it may be associated with an unexposed structure beyond the excavation limits.

The faunal remains showed evidence of butchering and burning. The majority of the identifiable material came from one or maybe two white-tailed deer which contributed 35.85 percent of the total deer remains recovered from features. This pit contained the partial remains of several other animals (black bear, turkey, various fish, etc.), but the deer elements predominate.

Feature 313 - This unit lies adjacent to Feature 4 and again suggests the existence of a structure beyond the limits of the excavation. It contained 6.6 percent of the ceramics and 16.3 percent of the faunal remains.

The motifs observed on the shell (n=735) and grit (n=172) tempered sherds included plain (n=358), check stamped (n=175), and other forms of stamping (n=189). This high incidence of stamping as opposed to plain surfaces is consistent with Middle (Egloff 1967 and Dickens 1976) and Lower (King 1972:vii) town ceramics. If these motifs do reflect outside influences then the use of shell tempering, in the absence of a functional explanation, may be a result of easier access to shell materials than quartz and other igneous minerals. Temper agents may be less indicative of cultural associations and more a product of procurement strategies.

The faunal remains consisted of a wide variety of species, including at least 16 kinds of molluscs and gastropods. The mammal elements were represented by black bear, white-tailed deer, pig, squirrel, chicken, and passenger pigeon. The MNI for each of these is one. Once again the refuse material is most likely the result of a few butchering and consumption episodes and not the endproduct of long term, repeated disposal activity.

Area 3

Feature 325 - This unit is a deep, bell-shaped pit located adjacent to the square, Hiwassee Island-style Structure 21. It is a high density ceramic (8.2 percent), lithic (34.8 percent), and faunal (12.1 percent) containing pit. Use of this unit in comparative studies (Boyd 1982 and Schroedl et al. 1982) confirms its Mississippian I/II designation on the basis of lithics and ceramics. It has an associated C14 date of 985 + 129 yr. B.P. (GX-7732), corrected to A.D. 976 + 129 (following Damon et al. 1974). A single silver earring was recovered from this feature. However, since no other colonial-dated material was found, it is assumed that the origin of this artifact was intrusive from a Mississippian IV context.

The ceramics consisted, primarily, of plain shell tempered sherds (n=930) with 13 loop and 5 strap handles. Of the 1118 total sherds, 1028 were shell tempered and 88 were limestone tempered. The lithics included large amounts of debitage and several (n=10) projectile points (three of which are associated with the Archaic component). Although the faunal remains contained a wide range of species, there is little evidence to suggest that elements from more than one or two individuals of each species were disposed of in the pit. Some of these represent small animals (like mice and rats) which probably fell into this deep, steep-sided pit and became trapped.

Feature 336 - This unit was found inside the perimeter of Structure 21 . It contained, with the exception of some botanical residues, only 261 lithic artifacts (6.8 percent of the total). The contents of this unit are very similar to those recovered from Feature 338 (3.4 percent of the total lithics). No significant diagnostics were recovered from the pit making temporal assignments impossible. Lithic refuse seems to occur along the edge of the second terrace more abundantly than elsewhere at the site. Therefore, the high density of lithics from these units may represent redeposition of material from the plowzone matrix. To better distinguish primary from secondary refuse under such conditions, the density of artifacts in the surrounding plowzone would need to be compared to that observed within adjacent pits. A primary disposal pattern would be supported if more material was concentrated in the pits than was recovered in the adjacent plowzone. Unfortunately, such a comparison is not possible from pan-exposed contexts. Lacking such data, it is assumed that the unit pre-dates the early Mississippian occupation based on its relationship to Structure 21.

Burial 86 - This unit is located inside the small, Structure 19. This burial contained 11.8 percent of the Euro-American artifacts and 2.6 percent of the lithic debris. The high density of lithics is correlated with its location in Area 3 and supports the idea that the lithics are associated with general plowzone accumulations since they seem to occur in this burial as fill. The Euro-American artifacts consisted totally of glass beads (n=1194). Based on the grave associations observed in Structure 17 it might be assumed that this individual was a child; however dental evidence suggests an adult male. The beads appeared to have been from a necklace.

Area 4

Burial 90 - Of all the units, this burial, located in Structure 24, had the highest incidence of Euro-American artifacts (28.1 percent of the total). This adult was buried with the largest number of personal items of all the burials. The inclusion of a mirror and a knife suggests that this individual was a male (mirrors were worn by men, Rothrock 1929). This individual was also buried with an unusual ground magnetite object placed under the head (see Figure 4.17). Three burial-like features (346, 349, and 350) were located at the northwest end of this structure. The wide range of residues found in each, especially botanical and faunal, suggests, however, that this portion of the house was used as a storage or processing area (cf. Steiner and de Schweinitz 1928:479).

Area 5

Feature 415 - this feature is located at the eastern end of Structure 31. Its contents represented 7.8 percent of the aboriginal ceramics and 4.0 percent of the lithics. The ceramics were primarily shell tempered plain sherds (n=926) with a few (n=10) limestone tempered sherds from an earlier occupation. The large quantity of lithics also suggests that this feature lies in a refuse laiden area from another occupation, possibly related to the burial area lying to the northwest.

Area 6

Feature 383 - Located along the edge of the excavation unit, this feature may be associated with either Structure 27 or an unexposed structure. Its contents include 14.1 percent of the ceramics, 2.6 percent of the lithics, and 2.0 percent of the Euro-American artifacts. Like Feature 313 in Area 2, this unit's ceramics include significant numbers of stamped sherds. The Euro-American artifacts consisted of several gun parts and several metal fragments. The contents were very similar to those of Feature 313.

Feature 385 - This feature was located adjacent to Structure 27 and contained 5.9 percent of the faunal remains. The identifiable faunal elements primarily consisted of the remains of one white-tailed deer, which constituted 11.1 percent of the total deer remains recovered from the site.

Feature 389 - This unit lies adjacent to the portion of the site hypothesized on the basis of postmold patterning as containing an unidentified structure. The pit contains 9.0 percent of the ceramics recovered from the site. All of the shell tempered indeterminate cord marked sherds recovered from features came from this unit. Such sherds are apparently associated with Middle Cherokee contexts (Dickens 1976).

Discussion

The primary component present at 40MR5 is associated with a Cherokee Mississippian IV occupation. An Archaic component was identified primarily within the plowzone deposits of the 32,000 year old second terrace of the Little Tennessee River. Some ceramic evidence of Woodland occupations were found along the edge of the terrace. Indistinguishable Mississippian I or II material recovered from Area 3 suggests an early Mississippian occupation. In addition, a late Woodland or early Mississippian burial area was located in Area 5. Its circular arrangement within the vicinity of burial mounds last observed in the nineteenth century suggests that it may be a remnant of this burial complex.

The Cherokee component, on the basis of artifact associations and historical documents, was related to the the village of Tomotley (ca. 1750-1776). Historic records suggest the presence of refugee Middle, Lower, and possibly Valley Cherokees in Overhill country during this time. It has been hypothesized that Tomotley represents one such village. The structural remains may be similar to those found in Middle and Lower towns and are definitely different from those observed at Chota or Toqua. Also the frequency of stamping on ceramic vessels is higher at Tomotley than at Chota or Toqua making this site more comparable to Mialoquo and portions of Citico (Baden 1982c). The incorporation of stamping on vessels is more characteristic of Middle and Lower Cherokee techniques than Overhill.

In conclusion, the goals of the 1976 excavations were successfully accomplished. The village pattern and townhouse location were defined and significant numbers of artifacts were recovered to identify the site's components. Future analysis and comparison of the Cherokee material recovered from North Carolina, South Carolina, and Georgia will need to be conducted before a more definitive statement on the characteristics of Lower, Middle, Valley, and Overhill Cherokee can be made. The first step involves recognizing that the historical evidence supports multiple, coeval cultural occupations along the lower Little Tennessee River. By providing detailed descriptions of the material recovered from these sites, signs of the differences in the archaeological record can be observed. Thus intersite comparisons can be made and a more concrete understanding of the Cherokee archaeological record can be realized.