Lithic artifacts recovered at the Tillett site include projectile points, projectile point preforms, a drill, bifaces (blades) flake blades (or modified spalls), and abraders in addition to lithic raw material (modified and unmodified) and spalls from the production process. (Page 51)

Fifteen projectile points or fragments of points were found at the site. Two of these were collected from the surface of the exposed shell midden in Great Gut and the others, with one exception, were reclaimed from the excavations at the southern end of the site (-42 trench area and Squares A, B and -48). The exception was found in the midden of square 20R21. Seven points are whole, 6 are sufficiently complete to permit some measurements and observations, and two others are represented by distal tip fragments. All were made from red (2), gray (4) and tan (9) jasper. Jasper raw material and cores from the site suggest the reduction of pebbles or cobbles of varying quality to produce the points and other chipped stone artifacts. Both primary percussion and secondary flaking were utilized, but the points vary considerably in production quality and adherence to an "ideal" shape, as illustrated by those specimens in Figure 18a-j.

The thirteen complete or nearly complete points are classified as the Roanoke Triangular type, small variety, falling within the description by Coe (1964:110-111). The concept of a "small variety" is emphasized here, since the type as defined spans at least two sub-periods (Coe 1964:111), and the smaller points are generally associated with the Late Woodland period. In this instance, the provenience of all but the surface finds was either intact or disturbed midden deposits of the Colington component at the Tillett site.

Point shape is that of an isosceles triangle although two specimens approach the dimensions of equilateral triangles (Figure 18c). The points have a length range of 20 to 37 millimeters with an average of 28 millimeters, and a width range of 18 to 26 millimeters with an average of 22 millimeters. The average produce a width to length ratio of 1/1.3. Maximum thickness taken at mid-blade ranges from 4 to 6.5 millimeters with an average of 5 millimeters, and cross-section shape is generally lenticular although 3 or 4 more crudely flaked specimens are more nearly "diamond" shaped. Base shape is concave, ranging from very shallow (5 specimens) to deep (2 specimens). Two points (Figure 18d, j) had been resharpened, modifying the original shape of the blade edges which are usually straight but occasionally shallow concave.

The small Roanoke point type is generally associated with the Colington phase and other Late Woodland manifestations in the North Coastal region (phelps 1983:39), and has its precedent form in similar, but generally larger, triangular points in the Mount Pleasant phase.

Five preforms, or "blanks", for the Roanoke type projectile points were found in the primarily Colington component midden deposits (Page 52) in the -42 trench area and squares A and B. Two specimens are whole, two are basal sections without their tips, and one is a mid-section blade fragment. Four of the specimens are jasper (2 red and 2 tan), and the other is white quartz (Figure 18k). All exhibit relatively large flake scars along their edges and two retain the cortex of the original pebble or cobble on their faces (Figure 18l-m) . Basal shape is straight (2) or very slightly concave (2). The measurable specimens have a length range of 40 to 42 millimeters with an average of 41 millimeters, and a width range of 26 to 31 millimeters with an average of 28 millimeters, giving a width-to-length ratio of 1/1.5. Maximum thickness at mid-blade ranges from 6 to 16 millimeters with an average of 11 millimeters.

A jasper drill with an expanded, round base (Figure 18n) was found in the lower level of the Colington component midden in square -42L6. The specimen had lost its distal end and the remaining fragment has a length of 33 millimeters.

Five bifacial tools (or blades), all produced from red, tan or gray jasper, were recovered at the south end of the site; four were found in the Colington component midden of the -42 trench area and one was picked up from the surface of the site in the 1972 survey. Two are complete specimens and three are fragments, but all are relatively small, reflecting the limitations of available raw material of small cobbles and pebbles. Of the two whole specimens, one has an ovate-trianguloid shape (Figure 18o), maximum length of 35 millimeters and maximum width of 25 millimeters; the other is an irregular quadrilateral shape (Figure 18p) and is best classified as a flake, or spall, with edge retouch on three sides. Two of the fragments appear to be from triangular bifaces (Figure 18q-r), but a third is too small to indicate shape.

A tan quartz cobble hammerstone with pecked and battered edges and a central pit on one face (Figure 19a) was found in the Zone II shell midden of square -41.5R1, an extension of the -42 trench. The specimen is roughly trianguloid in shape with flat faces and measures 63 by 55 millimeters. The maximum thickness is 16 millimeters, and the central U-shaped pit is 13 millimeters in diameter and 3 millimeters deep. Except for the base, the edges of the cobble show extensive wear from striking, to the extent that several large flakes had been knocked off the face near the edge. (Page 53)

Figure 18. Roanoke projectile points (a-j), preforms (k-m), drill (n) and bifacial tools (o-r); all actual size. (Page 54)

Figure 19. Cobble hammerstones with pit (a), and calcareous sandstone abraders (b-d). All actual size. (Page 55)

The central pit has a relatively rough interior surface resulting from use either as a drill base or as an anvil or striking platform (Spears 1979), both of which have been suggested as functions for such pits. Hammerstones of this type are ubiquitous in North Coastal region sites.

The fragment of another quartz cobble hammer with battering wear on one end was found in square -48, level 2.

Two small fragments of polished grayish-green slate were found in square 20R21, level 1. One of these has a regular, straight edge which shows intentional shaping, and both probably came from a gorget, or similar artifact.

A fragment of a grinding stone showing abrasive polishing on both faces was collected from the surface of the exposed midden in Great Gut.

Abrading or polishing tools of calcareous sandstone were found in a number of the excavation units at the Tillett site. Three of the larger abraders are typical: A large, rectanguloid specimen has a smooth, concave abrasive-use surface on one side (Figure 19b), and measures 73 by 55 millimeters with a maximum thickness of 33 milli- meters; a second rectanguloid fragment (Figure 19c) with a concave abrasive surface on one side is 61 millimeters long, 41 millimeters wide and 10 millimeters thick; an oval, flat cobble (Figure 19d) from the eroding midden surface in Great Gut has a concave working surface on one side, measures 88 by 65 millimeters and has a thickness of 17 millimeters. In addition to these, three large and seven small fragments of other abraders were recovered. With the exception of the surface specimen noted above, all were found in the Colington component midden in the -42 trench area, and in squares A, 4L4 and 20R21. The calcareous sandstone is the same as that used for the abraders in the Colington phase sites on Colington Island (phelps 1981) and other sites in the north coast area.

The raw material from which the lithic artifacts, with the exception of the calcareous sandstone abraders, at the Tillett site were made is predominately jasper. Specimens representative of all stages of the lithic production process were reclaimed in the excavations, including unmodified cobbles and pebbles, cores and core fragments and spalls. The various categories and quantities of material are: (Page 56)

The majority of this material was found in midden levels associated with the Colington component; in those levels of squares and the two features (Feature 9 and Pit 1, square -48) which could be assigned primarily to the Mount Pleasant component, lithic material of any sort was negligible or absent.

It is obvious from the above quantities that the source of raw material was a deposit containing primarily pebble-sized stones and a few cobbles. Such sources are found in the older Pleistocene stream channels that trend perpendicular to the present coast and barrier island chain (Riggs and O'Connor 1974), and which are exposed along the beaches and shallow bottoms by various erosion processes. These older channels are remnants of rivers which carried heavier stream loads of pebbles and cobbles from Piedmont and Mountain source areas during the Pleistocene, and it is from these that the water-polished raw material was collected. The closest known exposed channel is at Nags Head on Bodie Island, a few miles north of and across Roanoke Sound from the Tillett site, but other such sources may have been available when the site was actively occupied. Almost all of the lithic raw material, production debris and artifacts are jasper, ranging in color from medium tan to deep red and gray. Other stone types are a very minor percentage of the collection (Table 5).

The higher frequency of pebbles to cobbles in the raw material source is reflected also in the production debris. The number of unretouched spalls with cortex remaining is slightly higher than those (Page 57) without cortex, indicating the majority of the artifacts were produced from pebble cores. Utilization of spalls (3) and modification of spalls by retouch (8) is relatively minor. The impression given by the data is that raw material was exploited from nearby sources during the occupation periods (seasonal) of the site, and only the immediately necessary tools were produced.

Figure 20. Shell (a-c) and bone (d-g) artifacts and ceramic smoking pipes (h-i). All actual size. (Page 58)

Six artifacts made from animal or bird bones are included in the Tillett site collection.

A bone object made from a mammalian (possibly deer) scapula was found in the Colington component midden of square -42L6. The object (Figure 20d) has a cut and polished top margin that slopes upward to an apex in a triangular, or reducing basal configuration. Although the fragment is too small to accurately convey the total original shape and function of the artifact, the existing section appears to be similar to two artifacts found as inclusive grave goods with Burial 9 in the Cashie component at the Jordan's Landing site, 31BR7 (phelps 1983:44). Both of these are scapulae, cut and polished into a "fan" shape with a distal point which expands to a maximum width at the base; the base then contracts to a trianguloid shape like that of the Tillett site specimen. The two artifacts from 31BR7 are presumed to be parts of a shaman's kit, as suggested by the range of artifacts included with them. A similar ideological function can be tentatively accepted for the Tillett site specimen.

Distal or point fragments of bone awls or perforators were recovered from square A, level 9, -48, level 5, and the top level of -42L6. The two specimens from square A were made from "cannon" bones of deer, and were found in a Mount Pleasant component midden context. The specimen from square -48 (Figure 20e) was produced from the long bone of a small mammal, and a tip fragment from square -42L6 is a modified bird bone (Figure 20f). Both of the latter specimens belong to the Colington component.

A fragment of deer antler from which the tip had been cut (Figure 20g) was found in the fill of Burial 2 and belongs to the Colington component of the site. (Page 59)

Although expectation of a higher quantity of shell artifacts might be expected in a coastal midden, only three were found at the Tillett site.

The fragment of a knobbed whelk shell (Figure 20a) from square -48, level 2, has a polished lip and the remaining margin of a hole roughly knocked out of the outer shell. The specimen is similar in form and species selection to whelks found in both Mount Pleasant and Colington phase contexts in coastal (Baum site, 31CK9) and inner estuarine or riverine sites (Shipyard Landing, 31BR1, and Liberty Hill, 31HF30). Since this specimen is lacking the distal end, it is not possible to determine its function and similar spire or body configurations can occur on shells used as hoes, net weights or in other functions. The specimens from the sites listed above have been tentatively classified as net weights because of the peculiar wear pattern on the foot of the shell. Painter (1981:31-33) has proposed a similar use for shells from the Currituck site, but other uses such as adzes or gouges (Sears 1982:86-87) and possible planing instruments (Trinkley 1980:213) have also been suggested.

A distal columella fragment from a large conch shell (Figure 20b) has extensive abrasive wear on its tip, and may have been part of a hoe or pick. The specimen came from square 20R21, level 3, in probable Colington component context.

In the Colington component midden of square -42L6, the modified section of outer shell of a large, knobbed whelk was found. The specimen (Figure 20c) has been cut and edge-ground into a general ovoid shape and retains part of the inner spire or whorl structure at one end, which would have facilitated its possible use as a "spoon" or ladle.

Food remains at the Tillett site included shellfish, the bones of fish, mammals and reptiles, crabs and charred botanical specimens. Shellfish were identified by the staff of the Archaeology Laboratory, East Carolina University, and the other materials were submitted to appropriate specialists (see Appendices C -- E). (Page 60)

The remains of shellfish represent the most numerous single class of food, occuring by the thousands in a classic shell midden deposit. Their massive quantity makes it rather obvious that exploitation of shellfish beds in the site area was probably the primary subsistence pursuit at the Tillett site. In fact, it is not unreasonable to assume that the original selection of this location was directly related to the proximity of oyster and clam beds. No shell count or estimates of edible meat weight were performed because of the degree of previous disturbance of the site, and also because the author is not convinced that a random shell column analysis produces any realistic statement about the number of people occupying the site at any given time. The rate of accumulation of shell is not constant over the site, deposits widely separated in time may be adjacent to each other on the same level, and momentary cultural preferences for particular foods can be misinterpreted as long-term trends. For example, a subjective observation of the number of oyster shells to quahog shells in the -42 trench and square 4L4 showed oyster dominant in the former and a nearly equal number of each in the latter. The equal number of clams and oysters can be interpreted as one family's preference during one seasonal visit to the site, as a change in general cultural food preferences, or as an environmental change through which one species increased while the other declined.

From each midden excavation unit a sample of shell species represented was taken and during the excavation estimates of relative numbers of species were obtained. The following list gives the six most prevalent species of shellfish in the order (1-6) of their estimated quantities from all excavations.

Oysters were by far the most numerous, and quahogs were well represented in most areas of the site. These two were probably the primary food species throughout the site sequence. The conchs, while nowhere as numerous as oysters and clams, were present in small numbers throughout the site. These probably served two purposes, first as food and then as a resource of shell raw material for producing beads and other artifacts from their columellae and outer shells. The cockle shells and a few other species occurred only in small numbers and, if eaten, did not add significantly to the food quantity.

Modern oyster beds in the general vicinity of the Tillett site are located around the mouth of Broad Creek, in Roanoke Sound south of Wanchese Harbor, and in Croatan Sound on the west side of the southern (Page 61) marshes (U. S. Army Engineer District, Wilmington 1977: Plate 9). It appears likely that a similar situation existed during the active occupation of the Tillett site.

Fish remains from the site (Appendix E) include some species commonly taken by sport and commercial fishermen today, and others that are not currently considered "food" fish. The most common species in the site were sheepshead, red drum, black drum, Atlantic croaker, longnose gar, white catfish and stingrays. All of these are estuarine or riverine species that invade brackish water and would have been present in the sounds around the island. Species rare in the site collection were sand tiger, hardhead catfish, flounder, jack crevalle, striped burrfish, seatrout and speckled perch. The large size of the red and black drum fish (1 to 3 feet) and some other species suggests use of large nets or harpoons. Fish weirs are more probable than nets, however. No noticeable difference was found between the fish species represented in the Mount Pleasant and Colington components. It is likely that fishing was secondary only to shellfish collecting in the subsistence systems of both components.

Faunal remains other than fish (Appendix D) in the Mount Pleasant component include deer, raccoon, rabbit, turkey, duck and various turtles. The number of individuals represented is relatively low, indicating only peripheral dependence on these species at the site. In the Colington component deer, rabbit, black bear, merganser, turkey, and freshwater, brackish and saltwater turtles. A domestic dog represented in the collection should not be included with the food remains.

Analysis of botanical specimens (Appendix C) identified wood charcoal as probable oak, hickory and red cedar, these samples representing wood burned in fires on the site. In addition, some seeds were probably from red cedar. Plant foods identified include acorns and hackberry in the Colington component, and hickory nuts in Mount Pleasant context. Seeds of cleavers and plantain from a Colington component feature may indicate collection of these plants for medicinal purposes.

The food remains from both components of the Tillett site suggest a continuity in subsistence patterns and environmental exploitation. Shellfish and fish were the dominant food resource and suggest the primary reason for establishing a site at this location. Both of these resources were available in the sounds near the site, and various turtles and ducks were also taken in the aquatic environment. The available evidence from the above species points to a late Spring/early Summer, and possibly early Fall occupation of the site. Land fauna from the upland terrestial environment (marsh fringe, maritime forest) were secondary to the aquatic resources, but deer, bears, raccoons, rabbits and turkeys were hunted in small numbers. Plant remains of acorn, hackberry and hickory presume an early Fall use of the site, and the collection of wild plant foods as a tertiary supplement to other resources. (Page 62)