The White Court site, like 31DR14, faces Kitty Hawk Bay (Figure 3) and is exposed to the full force of prevailing northeasterly storms in the winter months. Unlike 31DR14, where vegetative cover extends down the dune slope to the top of the beach, the White Court site is experiencing extreme high bank erosion which has created a nearly vertical cut bank along its shoreward side (Figure 9a-b). It was probably the shell spilling down the bank from this site, rather than 31DR14, that Haag (1958: 40) observed as a "white scar" from Currituck peninsula across the sound. The amount of site loss from erosion is unknown, but Haag's estimate of 100 feet (30 meters) is probably conservative.

31DR33 lies along the northern slope of the dune ridge on a broad, shelf-like topographic feature created by the gradual incline from the 10 feet (at the edge of the cut bank) to 20 feet contour (Figure 3). While the exact area of the site was not determined, shell and cultural debris extended along the beach for a distance of 250 meters. This measurement probably exceeds the actual site area on the west; observation of intact humic and midden layers in the beach profile confirmed that the main site area lies east of the high point where the 20 feet contour is cut by erosion (Figure 9a). The surface of the modern dune of Fripp fine sand supports a mixed hardwood and pine forest and understory species (Page 25) typical of all such situations on the island. A cottage was under construction on the high point of the ridge along the shore in 1972 (Figure 9a); since then others have been built in the site area (Figure 8b). The site name is taken from the short street and cul-de-sac, Sir John White Court, which provide access to the area (Figure 2).

(a) (b) Figure 8. The Colleton Lane and White Court sites. Looking west along Kitty Hawk Bay Drive where it bisects 31DR34; the older humic zone can be seen near the top of the dune on the left (a). Aerial view (southwest) of the White Court site, 31DR33; the site extends from the sand spit at lower left to the cottage on the highest elevation in the center of the photo (b). (Page 26)

Surface collection was limited to the beach area in front of the site since the intact section lies buried below modern soil zones. The collected specimens (Table 3) are nearly equally divided between the earlier Mount Pleasant and the still partially intact Colington component, a percentage frequency more closely approximating Haag's collection from the Kitty Hawk Bay site (his C3) (Haag 1958: Table 1). It is probably that his 28 "grit tempered" and 67 "shell tempered" sherds came from 31DR33 rather than 31DR14, but he made no distinction between the two site areas, collecting the entire half-mile stretch of the north shore between the two sites as one unit.

Following the beach collection, temporary reference points were set on the beach and the top of the bank. These were used to control mapping of the beach profile, site topography, and a transit run back to 31DR14 in order to measure the exact distance between sites, trace the stratification along the bank, and confirm the negative evidence of sites along the beach between the two. Keying on the exposed stratification in the beach profile, four 2 x 2 meter squares (A-C, E) were laid out in a general east-west line to sample the midden extent observed in the profile. Subsequently, a 1 x 1 meter square, D, was opened on the southeast side of Square B to observe the size and overlapping of the lens-shaped midden deposits (Figure 3). Excavation and reclamation procedures were the same as those employed at 31DR14.

Stratification of the White Court site is the same as that at 31DR14, varying only in the depth of each stratum as a result of the particular processes of accretion and erosion at each site.

Zone 1: The gray-stained sand of the modern surface, averaging 10 centimeters in thickness, is the surface exposure of Fripp fine sand.

Zone II: The leached white to light gray Fripp fine sand of the modern dune. This culturally sterile stratum ranged from 10 to 50 centimeters in depth across the site, depending upon undulations in the underlying zone.

Zone III: This stratum of shell midden and dark ox,ganic soil was deposited in lens-shaped units rather (Page 27) than a comprehensive layer over the site. Maximum thickness of the midden was approximately 40 centimeters in Squares A and B, but its average was closer to 20 centimeters. This is the primary cultural deposit at the site.

Zone IV: The light gray humic sand of the former land surface, prior to site occupation ranged in thickness from around 10 to 30 centimeters with an average of 15 centimeters. Cultural material occurred only in the top few centimeters of this zone.

Zone V: The culturally sterile yellow sand of the older dune. Along the beach profile its observed thickness ranged from 2 to 6 meters, but maximum depth was not determined.

Representative examples of site stratification are shown in Figure 9b and Figure 10. A 10-meter-wide exposure of the Zone III shell and organic midden can be seen in the beach profile (Figure 9b), extending from the stake on the left to the pit profile (Feature 2) on the right. Other typical lens-shaped deposits are shown in the north profiles of Square A (Figure 10a), B (Figure 11b), and over- lapping midden units in the southern and southwestern profiles of Squares B and (Figure 10b). Breaks in the midden where it accumulated around tree trunks are seen in the middle of the profile in Figure 10a and to the right in Figure 11b; these features suggest that a forested dune similar to the modern situation existed at the time of initial site occupation. Figure 11a shows the top of Zone III exposed in Square E, and the gray ash and crushed shell fill of a pit (Feature 3) exposed along the north profile.

Three cultural features were excavated at White Court, along with a number of tree root intrusions. The cultural features were observed to originate in or at the bottom of Zone III and intrude into the underlying zones. Some of the tree roots were traceable from the top of the Zone III midden, but others originated in the top of Zone IV.

Feature 1 in Square A was a pit refilled with burned and charred oyster shell, scattered bits of charcoal and numerous very small unidentifiable fragments of bone. No artifacts occurred in this 40 centimeter diameter pit with a depth of 20 centimeters.

Feature 2 was salvaged from the beach profile (Figure 9b) where at least half of the pit had been lost by erosion. The outer fill of this pit, measuring 1 meter in width and 1.25 in (Page 28) depth was a black organic soil which enclosed an area of oyster shall mixed with organic soil. In this interior deposit were Colington and Mount Pleasant sherds. The feature was classified as a refilled pit of unknown function.

(a) (b) Figure 9. Eroded bank profiles at White Court. (a) View southwest showing the highest elevation at the west end of the site and a cottage under construction. Compare the exposed beach width with that in the 1979 aerial view in Figure 8b. (b) A shell midden deposit and Feature 2 in the bank profile. (Page 29)

(a) (b) Figure 10. Stratification at 31DR33, Squares A, B and D. North profile of Square A with tree root intrusion penetrating midden (a). Overlapping midden lenses in the southern and western profiles of Square B (foreground) and D (b). (Page 30)

(a) (b) Figure 11. Feature and profiles, Squares E and B. Top of Feature 3 exposed in Square E (a). Shell midden lens in north profile of Square B (b). (Page 31)

Feature 3 was an area of charred shell, ash, and bone observed at the top of Zone III in the north side of Square E (Figure 11a), The area measured 1.4 meters along the north profile and extended 80 centimeters into the square in a generally oval shape. The feature extended downward 20 centimeters to the bottom of 7, one III,, and was apparently a fire pit or hearth. One Colington fabric impressed sherd, 1 jasper spall (Table 3) and numerous fish bone fragments were recovered from the pit. All of the identifiable fish bones were white catfish (see Appendix B). Oyster shell from this feature was radiocarbon dated to A.D. 1230 + 65 (UGa-1087; B.P. 720 + 65), providing a date during the Colington phase occupation. A second radiocarbon sample, oyster shell from the lower level of Zone III in Square E, was analyzed as "modern"; the sample had apparently been contaminated either by modern intrusion of C14 on the site, or subsequently while stored in the lab.

The total number of artifacts from the surface and excavations was 186, distributed as shown in Table 3. As was the case at 31DR14, the frequency of artifacts of any class or type per excavated unit of midden is very low. Since no vertical differentiation of types or components was seen after analysis of the material from each arbitrary level in the midden zone, the specimens are listed only by square and zone or other provenience in Table 3. (Page 32)

Two ceramic series, Mount Pleasant and Colington, are represented in almost equal quantities, and a third, Cashie, by a few sherds. Ceramic sherds account for 135 of the 187 artifacts from the site.

The Mount Pleasant series of sand and small pebble tempered ,ware was most frequent in the surface collection from the beach, where 57 of the 63 sherds were found. The 6 sherds from the excavated units were randomly scattered throughout the later Colington component deposits (Table 3). Compared with Colington sherds from the beach, the Mount Pleasant specimens are much more eroded and water-polished, showing the typical rounded edges and smoothed surfaces resulting from long exposure in the surf. Mount Pleasant types, classified by surface finish, are: cord marked, 9 (Figure 12g), net impressed, 11; fabric impressed, 18; and residual, 25, the higher number of the latter resulting from surface erosion.

Colington series sherds occurred in reverse proportion to the Mount Pleasant distribution; 53 of the 69 specimens were recovered from their primary context in the Zone III midden, while only 16 specimens were picked up in the beach collection (Table 3). The surface specimens from the beach were less eroded than the earlier ware, although the shell temper had dissolved from most of them 9 (Figure 12c-d). All types of the shell tempered Colington series are represented at the site, including 42 fabric impressed 9 (Figure 12a), 3 simple stamped 9 (Figure 12b), 6 plain 9 (Figure 12c), and I incised 9 (Figure 12d). Sixteen other sherds were placed in the residual category because they were too small to permit determination of surface finish.

Four Cashie fabric impressed sherds, one of which is illustrated in Figure l2f, were recovered from Square B (Table 3), where they were associated with material of the Colington component. The low frequency of this ceramic type again suggests trade material from an inland source.

Fifty unretouched spalls, 2 of green slate and 48 of red, tan and gray jasper, were the total lithic collection from the site. All of these are small (.2 to 1.0 centimeter) and appear to be the by-product of both primary and secondary chipping procedures. Forty- seven of the spalls were recovered from the midden zone of Square E (Table 3). Jasper is the dominant material here as at 31DR14 and other coastal sites of the Colington phase. (Page 33)

One awl (Figure 12e), made from a deer ulna, was recovered in Square E. It is 7.6 centimeters in length and has a broken distal end.

Oyster shells (Crassostrea virginica) are the dominant food remain, occurring by the thousands and consisting of the bulk of the midden deposits. A sample of 43 shells from Squares A and B, representative of size range in this and the other sites on Colington Island, was measured to get some idea of the size of oyster being harvested. Only the lower valve half of each specimen was measured to obtain total length. The shells had a length range of 6.3 to 14.3 centimeters (2.4 to 5.6 inches) with an average of 9.6 centimeters (3.7 inches).

Quahog (Venus mercenaria) shells and miscellaneous fragments of other species were recovered, but their low frequency indicates relative unimportance as a food resource. The quahogs were more evenly distributed throughout the midden at 31DR33 than they were at 31DR14, however, and were sufficient in number to indicate intentional collecting.

Fish accounted for 94.2% of the usable meat (other than shellfish) at 31DR33, a percentage comparable to that at 31DR14. The species represented at White Court are, however, quite different in presence and percentage from those at Kitty Hawk Bay. A few spot, sea trout and Atlantic croaker, along with rays and/or sharks, were the only saltwater species present; the majority of the fish remains belonged to white catfish, a species whose normal habitat is the mouth area of fresh water rivers (Appendix B). Non-fish faunal remains include 2 deer, 1 raccoon and 2 rabbits among the mammalia, a duck and another (unidentified) bird, and snapping turtle, pond turtle, diamond-back terrapin and alligator from the reptile taxon (Table 13; Appendix B).

Seventeen charred fragments of hickory nut shells (Carya glabra-pignut hickory) were recovered in squares B and C. The shells had been broken prior to charring and were scattered through the midded in the two squares. Hickory trees probably were present in the forest type on the island contemporary with the Colington occupation, but the nuts could also have been brought as a storable food by the inhabitants from their base sites.

Exploitation of the aquatic environment was paramount at 31DR33 (Table 14), and terrestial resources are incidental to this pattern. In that respect, White Court is similar to 31DR14, but the evidence suggests a different seasonality for the two (Page 34) sites.

Figure 12. Artifacts from 31DR33. Colinton fabric impressed (a), simple stamped (b), plain (c) and incised (d). The shell temper has been leached from the specimens in c and d. A Cashie fabric impressed sherd (f) was probably obtained from inland trade. A typical Mount Pleasant cord marked sherd from the earlier component (g) and a bone awl (e) associated with the Colington component. (Specimens are half actual size). (Page 35)

Spring is the most likely season of occupation at White Court if the predominance of white catfish is considered. The relatively paucity of salt water fish and majority of catfish among the remains is best explained by an influx of fresh water into the Albemarle estuary. The influx was sufficient to support the latter species but brief enough not to seriously impair the level of salinity necessary for oyster beds. Since catfish remains were found throughout the site, Swift's (Appendix B) suggestion that storms were responsible for the influx is probably incorrect; the midden units were deposited over a longer period of time than the duration of a storm. Two other hypotheses can be offered for this pnenomenon. The first is a change in salinity of the northern sound area brought on by closing of inlets through the Outer Banks, a situation similar to present conditions when the level of salinity is too low for oyster growth north of Roanoke Island. Since the oyster beds were obviously productive at the same time catfish were being caught at White Court, this seems implausible. The second hypothesis is that the fresh water influx, and with it the catfish, was the result of increased fresh water flow from the rivers contributing to the sound system--the Roanoke, Chowan, Yeapin, North, Alligator, and others. While flow records for these streams are negligible, data obtained from the U. S. Department of Commerce Weather Bureau for the Roanoke River shows that in 15 out of 23 years between 1912 and 1959, maximum flow occurred between January and May. In seven of those years, the maximum flow occurred during March, April or May. If this pattern can be implied backward in time and compared with the season of maximum abundance of species represented in the 31DR33 faunal remains (Figure 17), then a spring occupation when the Albemarle system was being flushed with fresh water is most logical, An inlet open through the banks near Colington Island would tend to direct this flow toward Kitty Hawk Bay and the other waters around the island. Although evidence of hickory nut collection at 31DR33 could be interpreted as a fall activity, it is more likely that the occupants collected nuts still on the ground from the previous fall, or brought them to the island.

The White Court site was first occupied during the Middle Woodland period (300 B.C. - 800 A.D.) by people who pmduced the Mount Pleasant ceramics. Since 95% of the Mount Pleasant sherds were collected from the beach and exhibited evidence of long-term water polishing, it is apparent that the section of the site occupied by this component has long since eroded away. Erosion has also removed part of the suceeding Late Woodland period (800 - 1650 A.D.) Colington phase component, but a large part of the site occupied during this phase remains intact. One radiocarbon date, A.D. 1230 + 65, obtained from oyster shell in Feature 3, falls somewhere within the temporal range of the Colington component. A second sample apparently contaminated, produced a "modern" date. (Page 36)

White Court has the same intrasite pattern as the Kitty Hawk Bay site, shares the same stratigraphic sequence typical of the north end of the island, and has experienced similar erosional processes. In the latter situation, White Court is now being eroded at a rapid rate, perhaps because it occupies the northernmost, exposed point along the north shore, while the erosional processes at Kitty Hawk Bay appear to have abated somewhat.

The Colington component deposits at White Court are lens-shaped areas of shell, mostly oyster, and organic soil separated from each other by varying distances. The largest midden observed was approximately 10 meters across with a maximum central thickness of 25 centimeters. No evidence of structures was found within or between the midden deposits, and only three features, a fire pit and two refilled refuse pits, refer to cultural activities of a general nature. No artifacts were recovered from the old humus zone between the midden deposits, but a relatively small number of sherds and artifacts, faunal and botanical remains were mixed with the midden debris.

The pattern is one of small family groups with'a minimum of cultural equipment, occupying the site for brief intervals to pursue subsistence activities in the immediate vicinity. The primary activity was the collection of oysters, and perhaps clams (quahog) from beds immediately offshore from the site. Second in importance was fishing, but the main fish being caught is the fresh-to-brackish water white catfish along with a few of the smaller salt water species such as spot, croaker and sea trout. Hunting, collecting, and trapping of mammals (deer, racoon, rabbit), reptiles (turtles,, alligator), birds (duck), and forest products (hickory nuts) is a minor part of the subsistence pattern but does indicate a thorough knowledge of available resources in the various microenvironments. A spring occupation is suggested by the predominance of catfish, probably available during the period of maximum fresh water flushing of the Albemarle sound system. The site catchment area for all these resources is within a half-mile radius of the White Court site. (Page 37)