Four pieces of copper or brass were recovered from the Amity Site. While it is presumed that these are fragments of native sheet copper, such as was in use by the Coastal Algonquians at the time of contact (Quinn 1955:102), no chemical or spectrographic analysis has been performed to demonstrate this conclusively. The objects may be European-derived trade items. Only one item, a drilled copper pendant (Figure 30h), is definitely an ornament. It was recovered from Green�s Trench I and measures 22 mm long by 16 mm wide by 1 mm thick. A second copper object may be a pendant fragment (Figure 30i). It is roughly triangular, and a slight notch midway along its base may be the remnant of a suspension hole. It was recovered from the plowzone of square 190R192 and measures 20 mm by 13 mm by 1 mm. The other two pieces of copper, both rectangular, appear to be scrap. The larger measures (Figure 30k) 21 mm by 11 mm by 1.5 mm and was found in the plowzone of 185R200. The smaller fragment (Figure 30j) is only 10mm long by 4mm wide by 1 mm thick. It was recovered from Zone 2 of square 185R191. Subsoil in this square produced Colington, as well as Mount Pleasant pottery (Table B-4).

Aside from recent shot and bullets recovered from the plowzone (the field including the site is, among other things, a favored goose-hunting locale), only one object associated with firearms was recovered. That is a grey flint English gunflint recovered from Green�s Trench II (Figure 30e). The gunflint is 20 mm on a side and is 6 mm thick. The gunflint may be associated with the Postcontact period aboriginal occupation, or it may be associated with some of the later eighteenth and nineteenth century debris which occurs in the plowzone. As the arms carried by the Fort Raleigh colonists would have been matchlocks rather than flintlocks (Quinn 1865:26), it can be assumed confidently that the gunflint does not derive from the Fort Raleigh period.

The Amity Site produced a sparse assemblage of faunal remains (Table 8). This is not surprising, as acidic Coastal Plain soils rarely preserve bone in the absence of a sizable accumulation of shell. The vertebrate remains from the Amity site were analyzed by Annie Holm of the University of North Carolina Research Laboratories of Anthropology. Oyster shell was identified and weighed by the author. Only material from undisturbed contexts was analyzed.

Only 300 elements could be identified to class or more restrictive taxon. Four mammals are represented in the sample: white-tailed deer, black bear, rabbit and opossum. Also present but unidentifiable beyond the level of class were bird, fish, and turtle. Some of the turtle bone may be of the Florida cooter, a large lowland turtle that prefers quiet, well-vegetated waters (Martoff et al. 1980). In addition, about 1.5 kilograms of oyster shell were recovered.

Obviously, only the most tentative generalizations can be made on the basis of such a small sample. Little can be said about seasonality. On the basis of tooth eruption, Holm identified a deer mandible form Feature 8 as being that of a six to seven month old juvenile. Given a spring birth season, this indicates a fall time of death. The presence of turtle suggests a warm season exploitation, although turtles can be dug from their dens as they hibernate. Historically, late spring and early summer was the principal shellfishing season for the Virginia and North Carolina Algonquians (Waselkov 1987; Quinn 1955 :283), but Claassen (1986) has reported evidence of late fall to spring shellfishing during prehistoric times along the North Carolina coast. Overall the Amity site remains give some support to both a spring and a fall occupation of the site. However, a year-round occupation cannot be ruled out.

Calculation of the minimum number of individuals (MNI) represented by the identified specimens is made highly suspect by the small sample size. In fact except in the case of deer, a [Page 62] single individual of each species would be sufficient to account for all the bones recovered. The evidence of mandibular tooth eruption indicates that at least one adult and one juvenile deer are represented in the samples. Additionally, it should be noted that rabbit, turtle and fish were recovered from both Feature 8 and Feature 46. As these features are some 50 meters apart, it seems unlikely that the same individuals are represented in them.

A calculation of meat yield based on MNI would be inappropriate. With a small sample size the least frequently recovered species are overrepresented, since the first element identified always contributes to the MNI; whereas each additional element has a decreasing probability of adding to the MNI (Grayson 1984). At Amity, bear is represented by four elements which may very well derive from a single individual. However, it seems less likely that the 57 deer bones represent only one adult and one juvenile. A calculation of meat yield using these MNI figures and typical dressed weights (Smith 1975:34,173) would indicate the single bear to have contributed 95 kilograms of meat and the two deer only 59 kilograms. Hence, bear would be denoted as a significantly more important resource than deer. Given what is known of Native American hunting patterns (Swanton 1946; Smith 1975), it is safe to conclude that this is a fallacious conclusion.

With the data at hand, a more appropriate method for calculating meat yield is offered by the allometnc relationship between the weight of a bone and the weight of the flesh it supports. Using published data on skeletal allometry (Reitz and Scarry 1985; Reitz et al. 1987) it is possible to calculate meat yield directly from the weight of the recovered bone. Table 9, column 4, shows the result of applying the allometnc formula directly to the bone weights as measured by Holm. It can be seen that bear is relegated to a much less significant role as a food resource using this method. It is indicated as being only the third most important resource, providing only about one­fifth the meat as deer and less than one-half as much as oyster.

The utility of this calculation is lessened, however, by the fact that over 50% of the meat yield can be ascribed only to "unidentifiable mammal". Furthermore, over 33 grams of bone cannot enter the calculation at all, since it could not be identified to class thus precluding the selection of the appropriate allometnc formula. To mitigate these ill effects, a recalculation was [Page 63] made whereby the weight of the unidentified mammal bone was distributed among the identified mammal species based on the relative proportions of their bone weight, and the weight of the unidentified fragments was distributed proportionally among all vertebrates. The result of the recalculation of meat yield using the revised weights is shown in Table 8, column 6.

The general pattern of results is the same. Deer remains by far the most important prey species, but bear is now indicated to slightly surpass oyster in meat yield. Opossum now surpasses turtle in meat yield but remains a minor food item as does bird, fish, turtle and rabbit. In fact, nearly 90% of the total meat weight represented at Amity is provided by three species: deer, bear and oyster.

There is considerable uncertainty surrounding the meat yield calculation. There can be no escaping the fact that bone preservation at Amity was poor and the resulting sample very small. The total meat yield of the recovered faunal remains is only a paltry 1.5 kg. Oyster, in particular, with its dense and heavy shell is likely to be overrepresented due to preservation effects, whereas fish and birds with their light bones are likely underrepresented. Surely fish is not represented in nearly the numbers that the historical observers saw them taken (Quinn 1955), but neither is the Amity site located where weirs could be utilized.

Overall, the quantitative aspects of the faunal analysis are of suspect utility, but the general animal procurement strategy of considerable deer hunting augmented by the taking of other species and the collection of oysters seems perfectly reasonable given the site�s location. Although a larger sample would be most desirable, the site offers a tantalizing glimpse at coastal subsistence strategies at a non-littoral settlement.

Procedures: Three bodies of carbonized botanical remains make up the Amity archaeobotanical assemblage. First, 103 liters of soil from nine proveniences were water flotated using the "bucket method" (Watson 1976). A tea strainer with mesh size of 1.0 mm was used to capture the light fraction, while heavy fractions were collected by window screen. Second, a ten-centimeter level of midden from square 140R202 was waterscreened through window screen with mesh size of about 1.3 mm. Third, postmold fill from squares containing a portion of the house patterns was waterscreened by placing the fill of each postmold in a kitchen strainer (mesh size approximately 1.3 mm) and gently agitating it in a bucket of water. Artifactual material and potentially identifiable pieces of bone and charcoal were removed from the sieve and bagged separately. The remainder of the material was then returned to the bucket. After all postmolds from a square were sieved, the fill from the bucket was waterscreened through window screen.

At the laboratory all charcoal was identified using a binocular dissecting microscope. Identifications were made by the author and Kristen J. Gremillion. The flotation and waterscreening material was quantified in the standard UNC-CH fashion. Briefly stated (cf. Yarnell 1974 for more detailed discussion of methods) fragments greater than 2.0 millimeters in size were identified, and the identified taxa quantified by weight. The weight of material less than 2.0 millimeters in size was extrapolated from the relative proportions of the material greater than 2.0 millimeters in size. All seeds were quantified by count. Both the light fractions and the heavy fractions of the flotation samples were analyzed. Of the waterscreened material, only that from the southern half of square 140R202 was quantified. The remainder of the waterscreenings was examined microscopically and seen to mirror the quantified material. Results of the flotation and waterscreening are reported in Tables 9 and Table 10. No quantification by weight or seed count was undertaken with the postmold material, as the sample sizes are too small to make quantification meaningful. Instead, presence of various taxa of plant remains was noted. Results of the postmold analysis are reported in Table 11.

[Page 64] Results: A total of 98 grams of plant remains was recovered. This is about 34 grams of plant remains per 100 liters of soil which is reasonably successful recovery. Remains of food plants are much less well represented, occurring at a density of three grams per 100 liters of soil. The low density of food plant remains is a result of the exceptionally small amount of nutshell that was found. Nutshell density was only 0.6 grams of nutshell per 100 liters of soil, which is an extraordinarily sparse presence. The dearth of nutshell is not a reflection of a recovery problem, as small seeds are well represented. Seeds occur at a density of 77 seeds per 100 liter of soil if unidentified fragments are included, or at a density of 31 seeds per 100 liters if only identified seeds are counted. Either density is high.

Only 1.6 grams of hickory, walnut and acorn shell were recovered, but this represents 42% of the food plant remains. By weight 85% of the nutshell is hickory, 7% is walnut and 7% acorn. The three nuts vary dramatically in their respective proportions of edible meat, however. By weight hickory nuts are 35% edible meat (Watt and Merrill 1963), walnuts are 22% edible (Watt and Merrill 1963) and acorns are 62% edible (USDA 1984). Adjusting for the differences in edible proportion shows hickory to make up 76% of the nutmeat, acorn 20% and walnut only 3%. Ubiquity, expressed as the percentage of samples containing a plant remain, supports this rank order of nut resources. Hickory occurs in 55% of the float samples, acorn in 33% and walnut in 22%. Hickory occurs in 30% of the postmold samples and acorn in 20%, while walnut does not occur at all. Acorns are thus indicated to have been a more important foodstuff than walnuts. This is likely to be the case, as oaks are considerably more common than are walnut trees in the Outer Coastal Plain (Radford et al. 1968). Historical observers confirm that all three nuts were utilized by southern coastal Algonquians (Quinn 1955:351,354; Barbour 1986:151-152).

[Page 65] Remains of five crops were present in the samples: maize, bean, squash, maygrass and little barley. Maize was represented by kernels and cupules, the small structures of the cob from whence pairs of kernels arise. Maize was recovered from Amity in a density of 0.7 grams per 100 liters of fill. This is quite close to the median density of 0.8 grams per 100 liters reported for Late Prehistoric and Historic period Siouan sites from Piedmont North Carolina (Gremillion 1989). Maize comprises 52% of the food plant remains and occurs in 55% of the float samples and 44% of the postmold samples. By all criteria maize is indicated to be a staple. This is consistent with its known role in Native American culture during the Historic Period (cf. Quinn 1955:337-338; Barbour 1986:157-158). Maize remains were too fragmentary to allow any reconstruction of cob morphology, so no insight can be gained as to the varieties of maize consumed at the site.

One fragment of a bean was recovered from midden zone of the southeastern quadrant of 140R202. Beans are known from historical accounts to have been an important food (Quinn 1955:339; Barbour 1986:157), and are frequently recovered from sites of the Late Prehistoric and Historic periods. Rarely are they recovered in any quantity, however, so the small amount of bean recovered from Amity may not accurately reflect its importance as a food.

Likewise, only one fragment of squash rind was recovered, that from a postmold in square 178R204. Squash is known to have been an important food during historic times (Quinn 1955:340; Barbour 1986:158) and it is frequently recovered from archaeological sites in the eastern United States (Yamell and Black 1985). Its scant occurrence at Amity may be a result of small sample size.

[Page 67] Maygrass is a spring-ripening grass native to the Coastal Plain of the southeastern United States that was domesticated by Native Americans probably during the second millennium B.C. (Cowan 1978). Eight seeds (technically caryopses) were recovered from Amity, all from the midden zone of 138R202. Maygrass represents 9% of the identified seeds from Amity, which is a rather high contribution, but it is commonly recovered from archaeological sites dating to the last 2000 years (Asch and Asch 1985).

Little barley, like maygrass, is a spring-ripening grass. It is represented at Amity by 57 caryopses. It occurs in 55% of the flotation samples and 7% of the postmolds. It comprises 65% of the identified seeds. This is an extraordinarily large amount of little barley. (The number is conservative as the unidentified seed category is composed primarily of fragments that strongly resemble little barley but lack diagnostic characters). Little barley is not reported from archaeological sites as commonly as maygrass, but, like maygrass, it appears to have become an important resource in the Midwest by 2000 years ago (Asch and Asch 1985). It is apparently not native to the Southeast (Radford et al. 1968), so its presence at Amity is almost certainly the result of human husbandry.

Unfortunately, there is no historical account of Native American agriculture that includes a description of a plant that can be ascribed definitely to either maygrass or little barley. Hariot noted, however, "There is a kinde of reed which beareth a seed almost like unto our rie or wheat, and being boiled is good meate" (Quinn 1955:353). Smith, in his account of the Virginia Algonquians left the following vague description: Mattoume growth as our bents do in meddows. The seede is not much unlike to rie, though much smaller. This they use for a dainty bread buttered with deer suet" (Barbour 1986:153). As "bents" is an English term for a reedy or rush-like grass, it has been suggested that Mattoume may refer to Arundinaria gigantica, the common cane of the Eastern Woodland canebrakes (Bartour 1986:153 note 2). This seems unlikely, however. First, Arundinaria gigantica typically reproduces by rhizomatous sprouting and rarely sets seed (Radford et al. 1968). Second and more significantly, Arundinaria gigantica produces a seed seven to eight millimeters in length (Radford et al. 1968); whereas domesticated rye produces a seed of only five to six millimeters (Renfrew 1973:85). As Smith explicitly states Mattoume seed to be much smaller than rye, Arundinaria gigantica can be eliminated as a candidate, although it could perhaps be the "reed" observed by Hariot. I suggest a more likely candidate is one (or perhaps both) of the archaeologically common spring-ripening grasses, maygrass or little barley. Both have grains of about 2.5 to 3.5 millimeters long (Radford et al. 1968). As both prefer disturbed habitats such as old fields, they might well have been seen growing "in Meddows" around Native American habitations.

Three fleshy fruits are present in the Amity samples: grape, persimmon and sumac. Grape occurred in 22% of the flotation samples and one postmold. A total of five grape seeds were recovered. A single fragment of a persimmon seed was recovered from a postmold in square 172R206. Two sumac seeds were recovered from the midden zone of the southeastern quadrant of square 140R202. All three plants are commonly recovered archaeologically in the Southeast (Yamell and Black 1985). Grapes are available from late summer through the fall, while persimmons are available in the fall (Radford et al. 1968). Both, of course, produce edible berries. Sumac produces clusters of thin-fleshed drupes which have a tart taste and can be used to produce "Indian lemonade" (Medsgar 1966:213). In addition, they were used by Native Americans during the historic period as a hair colorant (Swanton 1946:527 quoting Bartram 1909) and as a smoking material. (Swanton 1946; Yarnell 1964).

Five plants identified from the Amity samples are of uncertain or doubtful economic importance. Two poke seeds were identified from postmolds. Poke can be harvested for greens when it is young, and its fruits, which lend the plant its sometimes name of "inkberry" (Medsgar 1966:142), are a likely source of dye or face paint. Bedstraw is represented at Amity by a single seed from a postmold. Bedstraw can be used to make a beverage (Hendrick 1919:285) and, as [Page 68] the name suggests, the plant can serve as bedding. Poke and bedstraw are so routinely recovered from southeastern archaeological sites, that I suspect they were of some economic importance. However, both are adventive weeds common in anthropogenic habitats, so the recovery of a few seeds could well be a result of inadvertent carbonization of commensal plants. Likewise, copperleaf and knotweed are vigorous weedy colonizers of disturbed soil. Tulip poplar, on the other hand, is a common component of the Mixed Hardwoods Flats forest that likely characterized the ridge on which the Amity Site occurs (Lynch and Peacock 1982). I suspect the tulip poplar, copperleaf and knotweed seeds were all introduced into the site as part of the natural seed rain, some of which made its way into fires to be carbonized.

Discussion: Although not a large sample, and collected with less than optimal equipment, the Amity archaeobotanical assemblage indicates that, if directly pursued, plant remains can be recovered from coastal sites even in the absence of refuse-filled pits. Historical accounts (Quinn 1955; Barbour 1986) make it clear that both agriculture and gathering of native plants were important subsistence activities of the southern Algonquians, but archaeological evidence of food plants has been slow to materialize in the coastal region. As yet, no report on a sizeable body of flotation-derived plant remains from a single component coastal site has been published. The little data presently available have indicated little more than the utilization of hickory, acorn and corn.

In North Carolina the 38 features at Uniflite site (31On33) produced about 67 g of carbonized plant remains greater than 1mm in size, but less than 0.2 g of plant food remains comprised of hickory nutshell, acorn shell and corn kernels (Glazier 1986). At the Flynt site (31On305) 35 features produced about 120 g of carbonized plant remains greater than 1mm in size, but less than 0.8 g of plant food remains. Hickory nutshell, acorn shell and corn kernels and cupules were the only food plant remains recovered (Glazier 1986). Both sites date to the Late Woodland period.

In Virginia 10 features from the multicomponent Plum Nelly (44Nb128) site (Potter 1982) yielded about 67 g of plant food comprised of about 61 g of hickory, 6 g of acorn shell and one specimen each of hackberry and grape seeds (Gary Crawford, personal communication 1980).

In South Carolina three Stallings Phase features at the Fish Haul site (38Bu805) produced 5g of plant food remains, entirely hickory nutshell (Trinkley 1986). At the Mattassee Lake sites in the inner Coastal Plain of South Carolina, about 476 liters of fill were floated from 39 features representing Archaic to Mississippian period occupations (Harris and Sheldon 1982). Hickory shell was the most abundant plant food remain, making up 46 g of the 549 g of carbonized plants recovered. Hickory was found in 90% of the tlotated features, acorn in 41% (0.2 g total), and corn was recovered from only a single pit. Fourteen taxa of plants, including the food plants chenopod, poke, blackberry black gum and grape, were represented by seeds (Harris and Sheldon 1982).

The Amity plant assemblage is unusual in the sparse occurrence of nutshell and in the abundance of spring-ripening grasses. This pattern suggests a late-spring deposit of the plant remains. This would be the time after the exhaustion of fall-ripening nut stores and during the period of availability of spring grasses. However, the grape and persimmon are fall-ripening fruits and were likely consumed then. The presence of a fall-killed deer (see p.61) reinforces the grape and persimmon in indicating a fall occupation. Absent from the Amity archaeobotanical assemblage are summer-ripening fruits such as blackberry and maypops, which are common archaeological finds elsewhere. This may be a result of small sample size, or perhaps the areas of the site with preserved midden were not used as garbage receptacles during the summer. Alternatively, the site may well have been abandoned during the summer, presumably for fishing and shellfishing stations at waterside. Certainly seasonal shifts in settlement location seem to [Page 69] have typified the coastal Algonquians dunng the early Historic Period (Barbour 1986:162; Feest 1978a).

The Amity site plowzone produced a wide array of recent material including square cut and wire nails, brick fragments, shoe sole fragments, modem lead shot and bullets, and unidentifiable metal scrap. None of this material is likely to be associated with the aboriginal occupation, nor does it suggest an early Euroamencan occupation. Rather much of it was probably introduced to the site from houses located along U.S. 264, some 200 meters to the west. In fact one resident of a nearby standing structure frequently chided the field crew that they were digging her family�s former trash dump. As much of the Euroamerican ceramics and glass fragments date to the early twentieth century, her statements may not have been entirely facetious. Also common in the plowzone were fragments of oyster shell. Some of this may have been plowed up from the Native American midden, but as oyster shell has until recently been used to lime the field for agricultural purposes, this cannot be known with certainty. For this reason all plowzone derived oyster shell has been considered a recent intrusion.

Euroamerlcan Ceramics: The Euroamerican ceramics from the Amity Site are predominantly nineteenth century in origin (Table 12). Whiteware (sensu lato including ironstone, hotel china and other semi-vitreous white-bodied wares) makes up 79% of the assemblage and post-dates 1820 (Majewski and O�Brien 1987). Stoneware makes up 14% of the assemblage. The Albany slipped, the gray salt-glazed, and the alkaline glazed pieces are all nineteenth or early twentieth century American manufactures (Greer 1981). The brown salt-glazed sherds may represent the English manufacture of 1690 to 1775 (Noel Hume 1970:114) or may be a later American manufacture. Three percent of the assemblage is pearlware, dating 1779 to 1820 (Noel Hume 1970:128-130). Seven fragments of porcelain, 4% of the assemblage, were recovered. [Page 70] Five are undecorated white and two are underglaze blue. As the latter are rather sloppily executed, they may be Canton porcelain of the 1800 to 1830 period (Noel Hume 1970:262; South 1978).

Significantly, the assemblage lacks any Euroamerican ceramic which must necessarily predate the Federalist Period. In particular, Amity produced no white salt-glazed stoneware, "Jackfield" ware, creamware, or delft, wares that are commonly found on coastal North Carolina Colonial Period sites. This would seem to indicate that the site was abandoned during this time period. If this is the case, then the kaolin pipe stems with small diameter bores are likely a part of the aboriginal occupation. This adds support to the applicability of the 1661 date to the Native American occupation.

Glass: In addition to the green bottle glass utilized by the Native American inhabitants of the Amity Site, an additional 239 glass fragments were recovered -- all from plowzone (Table 13). Fragments possessing screw tops, recent trademarks (e.g. beer and soft drink logos), or otherwise clearly recognizable as recent were not saved during the 1988-1989 season. Over 90% of the glass fragments are derived from bottles and 7% from window glass. The remaining 3% is comprised of three light bulb fragments, two pieces of automobile safety glass, one rim from a bud vase, and a handle fragment, probably from a pitcher.

Of the 216 bottle glass fragments recovered from the site only 14% are of green glass similar to that used to fashion the projectile points. As they are plowzone finds, their association with the Native American occupation cannot be demonstrated. Some is likely to derive from this occupation, however.

The remainder of the glass is of more recent origin. Two-thirds of the bottle fragments are of clear glass which postdates 1880 (Newman 1970). Ten of the clear fragments are base portions which bear the offset circular mold seam diagnostic of the Automatic Bottling Machine which came into use in 1903 (Newman 1970). Furthermore, three bases bear the outline diamond logo used by the Diamond Glass Company of Royersford, Pennsylvania after 1924, and one bears the outline diamond with interior "I" logo used by the Illinois Glass Company from 1916 to 1929. A fifth trademark consists of the outline diamond with interior "I" superimposed on an [Page 71] outline circle which was used by the Owens Illinois Glass Company from 1929 to 1954 (Toulouse 1971).

The amethyst glass is fairly recent as well, dating from 1880 to 1925 (Newman 1970). No trademarks were preserved among the amethyst fragments, but three amethyst pharmaceutical ("bitters") bottle necks show mold seams to with a quarter inch of the rim. This indicates manufacture by the semi-automatic bottling machine used from 1880 to 1913 (Newman 1970). The amber/brown fragments suggest recent beer bottles but are to small to be positively identified as such.