A total of 60 flotation samples from two seasons of work at the Clampitt site have been processed, but only 16 samples have been sufficiently analyzed to merit inclusion in this report. These 16 samples were recovered during the 1991 season and represent 15 contexts, most of which are pit features. Feature 8 and its associated post molds, numbers 7 and 8, are part of the inner stockade wall. The 16 flotation samples ranged in size from 12.5 to 21.0 liters and totaled 288.25 liters of soil.
Because separation of carbonized remains was imperfect, both heavy and light fractions were treated identically in analysis. 1 After drying, the fractions were gently sieved through a stack of geological screens (2mm, lmm and .5mm) to make sorting easier. All plant fragments greater than 2mm were completely sorted under a light microscope. Fractions smaller than 2mm were scanned under a light microscope at 10x to 45x magnification. All carbonized botanical materials except wood ( In an attempt to overcome the separation problem for samples collected during the 1992 field season half of the samples were randomly selected to be dried prior to flotation in hopes that drying would inaease the buoyancy of carbonized remains. Although an extremely preliminary count reveals that slightly more remains turned up in the light fractions of the dried samples, the difference is not statistically significant.) Charcoal from these smaller fractions were removed and identified (when possible) but the materials were not otherwise sorted.
It is not currently a standard practice in archaeo-botanical analysis to remove fragments of carbonized maize kernels, cupules or nutshell from size fractions smaller than 2mm. Such remains were removed and identified from the Clampitt samples, because this was my first experience with archaeo-botanical analysis and I wanted to be entirely sure that I would not overlook any remains. Unfortunately, this means that the results presented here are not fully comparable with other, more standard analyses. The maize, cupule and nutshell counts are artificially high, as are the counts of unidentifiable remains. I suspeet that this inflation is especially true of the hickory counts, since hickory is so dense that even very small fragments of it would tend to survive in the soil rather than being ground into dust as small fragments of acorn or maize would. Weight data, when available, should be better comparable with other sites, since weights will minimize skewing by the presence of small fragments.
In this analysis, as is usual for open sites in the eastern woodlands (Minnis 1981:147), only carbonized remains are considered ancient. All others are considered modern contaminants. Because soil acts as a seed bank for wild plants, uncarbonized seeds are commonly found to a depth of 20cm and occur up to a depth of lm (Keepax 1977 in Minnis 1981:148). Because Clampitt is a relatively shallow site, exclusion of fresh seeds seems especially prudent. In fact, large numbers of uncarbonized seeds were observed in flotation samples, but they rarely overlapped with the carbonized taxa, goosefoot and grasses being the main exceptions.
Seeds and other non-wood plant parts were identified to the lowest possible taxonomic level using standard reference works (e.g., Martin and Barkley 1961, Schopmeyer 1974). When possible, identifications were confirmed by comparison to modern carbonized and uncarbonized specimens in the collections at the Glenn A. Black Laboratory and Margaret Scarry's collections at the Kentucky Anthropological Research Facility. Dr. Scarry graciously checked all identifications except those from features 17 and 30.
Second, botanical remains encompass both actual foods and the by-products of processing those foods. Nutshell, for example, is a by- product of nut eating whereas maize kernels represent a plant part that was actually eaten. Therefore, in considering the ratio of hickory to acorn consumption on a site, the analyst cannot simply calculate a ratio of hickory-shell-to-acorn-shell but must also take into account the proportion of shell to nutmeat for each of those taxa (as well as the probable preservation bias towards dense, heavy hickory shell).
Ten domesticated bean fragments have so far been found at Clampitt. Since beans were probably not parched in preparation for storage as maize was, beans are less likely to appear in the archaeo-botanical record. Therefore, the fact that only ten bean fragments were found, as opposed to 850 or so maize fragments, cannot be taken as an absolute indicator of their relative importance in the Clampitt diet. Certainly, beans were important enough to warrant the effort involved in cultivation. But the dietary significance of beans, which provide critical amino acids missing in maize, will have to await further research.
Contrary to an earlier report (Bush 1991), no squash remains have been found so far at Clampitt. What had been initially identified as squash rind has been conclusively identified by Indiana University mycologist Michael Tansey as a fungus, specifically a stromatic pyrenomycae (pers. comm. 1993). According to Dr. Tansey, this fungus is common in south central Indiana and tends to grow on wood and grasses. It could easily have entered the archaeological record at Clarnpitt if wood to which it was attached was burned as firewood.
1. In an attempt to overcome the seperation problem for samples collected during the 1992 field season, half of the samples were randomly selected to be dried prior to flotation in hopes that drying would increase the buoyancy of carbonized remeains. Although an extremely preliminary count reveals that slightly more remains turned up in the light fractions of the dried samples, the difference is not statistically significant.