THE ARCHEOLOGY OF COASTAL NORTH CAROLINA
NOTE: Information in this section has been edited for the protection of archaeological sites. Please visit:
North Carolina Archaeological Resources Protection!!!
H 1, Cape Creek Site.
This site is an extensive midden consisting of a fairly uniform shell and bone midden. The total depth of this midden rarely exceeds one foot but it is quite evident that its extent was once much larger. The old humus zone upon which the midden accumulated is often one foot thick and strongly suggests a forest cover existed when the site was occupied. This old surface and the overlying midden are drifted over with sand that may vary one to five feet in depth. Throughout an area roughly five hundred feet square the midden is present except where winter storms have cut shallow ravines through sand midden and old humus. Local residents constantly attest that the land is greatly reduced all along the Sound side. This cutting back is said to be an every-year affair and not just the result of exceptional storms. The amount of the attrition cannot be very accurately fixed but it is quite evident that some has taken place. This may be judged from the fact that at every locality where a site once was (Page 28) or now is found, shells and potsherds are abundant in the water's edge. Where there is a long flat between two high points virtually nothing will be found in the water. The midden proper of H 1 is back from the Sound edge approximately one hundred feet, but a few shards were taken all along the stretch of about three hundred to four hundred feet. That these were washed into the Sound by wave action seems reasonable, but it is possible that the midden once extended farther Soundward.
Cape Creek site is the best midden found on the Outer Banks. It has been known for years and was visited by Porter and Harrington of the Park Service in 1938. The University of North Carolina Laboratory of Anthropology has material from this site collected by Coe. Despite the meager depth, the extensiveness of the site suggests a large population. All the surrounding dunes for fifteen hundred feet have shells and occasional sherds on their summits although at the site proper is the only place where any appreciable midden remains. These favorable facts led to a stratigraphic test of the site, the results of which are cited below.
[FIGURE 2 - Archeological Survey Map (Omitted) (Page 27)]
H 2, Indiantown Hill.
Indiantown Hill is a large dune. An old humus zone may be seen about four feet from the top of the hill and it is presumed that this zone was once of greater extent. A local informant who reported the site said at least three hundred feet has been removed from the site. Most sherds found were in the water's edge but shells and bones occurred in the humus zone and a few sherds were found at the base of the cut-bank face of the dune.
H 3, Lars' Hill Site.
(Page 29) Local residents estimate that about one hundred feet has been eroded away by waves. Fragments of pottery vessels and pipes were found at the base of the sand hill as well as on its sides and in the water. However, at no spot were extensive materials found.
H 4, Daily's Hill.
The artifactual remains are virtually all from the water. There is no promentory or hill here despite the name, and only a very small patch of an actual midden not more than one foot thick and about fifty feet in diameter is to be seen. In 1953 a skeleton was washed from this area but later reburied and subsequently could not be located. Potsherds were numerous along the shore and in the water; occasional finds of other artifacts came from the water. These occurrences extended for more than one thousand feet along the shore.
H 5, Sandy Bay Site.
(Page 30) No specific site or midden may be defined although sherds are localized in a few spots of not more than one hundred feet each along the shore. Many of the sherds are imbedded in the mucky shore but not all seem to be waterworn. It is probable that this site was always low now not more than two feet above sea level.
H 6, Duncan Point Site.
Scattered along a fairly-pronounced ridge of old partly-vegetated sand dunes are localized occurrences of sherds and pottery pipe fragments. Only on the higher dunes are cultural materials to be found. No vestige of a humus zone of midden could be seen. It is probable that the artifacts found are remains that have "drifted" down from sand layers removed by wind action. Winds are strongly evident here, modifying the growth of the covering vegetation.
H 7, Frisco Dune Site.
Archeologically, this site is not very impressive. It is a large blow-out in the dunes that is approximately five hundred feet east-west by three hundred feet north-south. In a few spots, especially on the north side of the depression, an old humus zone may be discerned. It may be surmised that this zone was once over the whole area now blown out and the potsherds found around the margin of the blow-out came from this (Page 31) zone. The many shells, mostly whelks and large bivalves, that floor the bottom of the blow-out hardly could be the remains of aboriginal eating. It is much more likely that this is beach debris, indicating the presence of the beach line here in former times. The overlying dune material thus has blown away and exposed the shell. The beach is now one-half mile distant to the South.
The archeological specimens recovered are not the sole reason for an interest in this site. In this vicinity a local resident recovered a number of coins and "a counter", all of which were sent or given to a "government man" some years ago (Quinn, loc. cit., p. 866). Because of the presence of "counters" in the archeological remains from Fort Raleigh, it was hoped that these objects might suggest something of the fate of the former colonists at Roanoke Island. Nothing new was discovered here although several visits were made. Glass ware, crockery, chinaware, and clay-pipe fragments indicated white residents here but a rather recent date--nineteenth century--seemed indicated.
About one mile from H 7 is another tremendous blow-out, H 8, about twenty feet deep and five hundred feet in diameter with shell covered floor. Many fragments of white clay pipes were found and a single stone pipe but none of these was of aboriginal origin. Many more evidences of recent white occupation were to be seen. However, an interesting feature here was the presence of numerous stumps and bark rings in a fairly heavy old humus zone. This is evidence of a rather heavy forest cover with a resulting floor of humus. It is concluded that dunes originating along the beach covered this wooded region and killed the trees. (Page 32) The aboriginal materials of H 7 came from this humus zone whereas the modern crockery and pipes seen to have been deposited more recently in the dune blow-outs.
H 9, Buxton Landing Site.
East of Buxton Landing, the hilltops generally have a sparse midden. If sherds are not found on the hill summits a few specimens will be seen on the slopes or at the bases of the hills, as though wave action was gradually reducing the face of all the Sound side here. Thus only heavier objects such as potsherds might remain.
The Buxton Landing materials may well be only outlying patches of H 1. Since it was hoped that the dune ridges so manifest across Hatteras would prove to be of quite different ages, it might be expected that the archeological remains here would aid in demonstrating this age differential. H 9 and H 1 thus would be correlated with different dunes and should contain different pottery remains. The two sites proved to be very little different in pottery content. It is probable that the dune ridges are far too old to have an age differential reflected in the archeological remains.
H 10, Original Little Kinnakeet Station.
The original site (Page 33) was on one of a series of low dunes that extended as a ridge from the present station to the sound. The site revealed no aboriginal materials but is recorded because of the wealth of widden materials of recent centuries. As historical studies are continued in the area, this site will be significant.
H 11, Hatteras Village Site.
A few sherds were found in the middle of the road where the level of the ground has been reduced by wind and vehicle erosion. The physical situation was such that some information might be forthcoming about possible sites on the beach side of the island. However, the sparse find did not reveal any such information.
H 12, Hatteras Coast Guard Station.
One large sherd and much shell at the water's edge indicated some aboriginal occupation, but this was so mixed with recent debris from Hatteras and the new station that nothing vital could be drawn from the site.
O 1, Springer's Point Site.
No suggestion of a wide-spread or large occupation of aboriginal peoples is found on Ocracoke. On the edge of the (Page 34) village at the head of land, a humus line about five feet above tide may be discerned in the wave-cut bank. In the water are found numerous shells that appear to be midden concentrate rather than recent debris. One potsherd was found in the humus zone but nothing else there or in the water. A more extensive midden may have existed at one time but nothing is left to suggest this. Local informants point to this spot as the source of arrowheads and other evidence of Indians.
Because of the belief that the Roanoke colonists may have come to Ocracoke, considerable attention seemed warranted for the area. Nevertheless, there is no archeological evidence that any people ever lived here in great numbers except directly at the modern village on Silver Lake. Nothing was found in the town area despite reports to the contrary.
For the purposes of this survey, all the Bank northward from Oregon Inlet to the Virginia line was called Bodie Island. This was purely for convenience and the area certainly has been several islands in the past. Not much evidence of Indian occupation may be found, but some interesting historical and geological information was gleaned from the several localities investigated.
B 1, Soundside Site.
B 1 contained no archeological remains of note, but it is an area that reveals something of the geologic processes active in the Banks. The cluster of houses locally known as Soundside lies southwest of Jockey Ridge, the highest dune along the coast. Directly west of the highest point of the ridge there is but one occupied house now (1955), but one (Page 35) vacant house and at least five sites of former houses may be discerned from piles of shell, broken glass, and other debris, as well as several live oak trees. The remaining resident said his house has been moved from the foot of Jockey Ridge to its present location. Thirty-five years ago the house had been built near the Sound shore and well away from the sand dune but the westward movement of the dune forced relocation of the house. Unfortunately, the exact spot where the house formerly stood is under the dune edge now but certainly the abandoned house and the other house sites give a clue as to the direction and rate of movement of the dunes. The vacant house is an interesting spectacle. It is two-storied yet sand has drifted entirely around the house, level with the second floor windows. However, wind eddies have kept the house largely free of sand. Only recently vacating, the occupants had aided the wind by shovelling sand away from the house edges. This house and the abandoned sites are about 800 feet from the Sound shore and all this "free" or "new" land has been formed by sand drifting into the Sound and extending the shore westward.
Jockey Ridge may be seen during the summer with a plume of sand spilling over the crest of the ridge toward the east. The prevailing summer winds from March to October are southwesterlies, but in winter the wind shifts to the east and northeast and these are stronger, steadier winds, hence, the dunes all tend to move soundward (westward).
Local collectors report sherds from this midden area but could produce none for inspection. One sherd was found on the Sound edge. No other Indian materials were forthcoming. A collection of chinaware, crockery, and glass fragments was made from the house middens that are gradually being covered by Jockey Ridge. The time of occupation can be (Page 36) fixed between 1875 to 1925 and historical research may profitably use collections of this nature. It seems obvious that future disclosures of modern chinaware of the eastern side of Jockey Ridge will give an accurate measure of the rate of its movement. (See B 3, below.)
B 2, Shellbank Point Site.
Shellbank Point is west of Kitty Hawk. A long, low, broken ridge extends from higher ground on the east to the Sound. Shell was abundant but other midden remains were scarce. A small collection of sherds was made. The suggested size of occupied area is about one hundred feet in diameter, hence no large village is represented.
B 3, Nags Head Site.
Southwest of the Nags Head are six small spots of midden. Each is less than ten feet in diameter and composed mostly of shell. When visited in 1954 only twenty seven sherds could be found but seventy five sherds were recovered when the site was visited in 1938 by Drs. Kelly and Porter. Another small collection of forty five sherds was made some years earlier by the National Park Service.
This site is significant because it may be an example of a former Sound-side site which has been (Page 37) uncovered now on the eastern flank of the dunes. It is virtually certain that shell has been hauled away from this site since 1938 for the six remnants are mere heaps of shells that appear to be in situ but were formerly a continuous stratum. Not more than three feet of shell is present in each hillock but sherds were found directly in those.
B 4, Martin Point Site.
Along the Sound shore a large midden area may be discerned. It is conservatively judged that for twenty five hundred feet along the shore shell and sherds suggest a fairly large occupation. Nothing was found more than one hundred feet inland but many sherds were recovered from the water.
B 5, Duck Dune Site.
In the vicinity of the small community of Duck the dunes are quite high, more than one hundred feet above tide. Much of the area is clean, shifting sand and an obvious movement of the dunes toward the Sound may be discerned. Several spots appear in this dune area that indicate the presence of former modern dwellings, but none had any aboriginal materials admixed with the modern debris. Nonetheless, there were several reports of the presence of aboriginal sites in the dunes.
(Page 38) Some years ago, between 1935 and 1940, a party picnicking in the Duck area found some modern materials, including two or three English coins which were sent to the Smithsonian institution. All the objects were identified and returned but none has been saved. However, Mr. David Stick, the Banks historian and researcher, has presented a number of projectile points accumulated over years from the Duck dunes. These specimens, plus a few found during the survey and test excavation, form the basis of a projectile point study from the Banks.
In the dunes area is one large depression or blow-out that is near one hundred feet deep. Grass grows on the moist bottom which is about fifteen to twenty five feet above sea level. An obvious numus line appears around the edge of the depression, a zone that outcrops about one hundred feet above tide. Often the more-compact old humus layer forms a shelf around the great hole. In and on the layer are found the old house sites with chinaware as well as indication of aboriginal remains. In 1955 a nearly complete skeleton of a deer was uncovered at this level; a projectile point was found less than ton feet away. The sherds of Indian pottery found here were always just down hill from the humus zone.
South of the same humus zone or, at least, a similar old humus zone appears. The patches that had been preserved by a dune cover and which now are being exposed by the wind yielded potsherds. The site was selected as a test area in order to recover the sherds in some stratigraphic order if possible. The results of this test are reported in a later section.
C 1, Beaseley Site.
(Page 39) This shell midden is much like others that have been reoccupied by later peoples. It is about one foot deep. Shells are abundant but sherds are scarce. The traffic of humans, wagons, and cars, plus later debris, often has masked the true aboriginal remains. All of the shells in the midden appear to be oysters, a similar condition in other middens on Colington Island. This fact argues for the presence of an ancient inlet through Currituck Banks somewhere near the north margin of Colington Island. No live oysters are found in these waters which are now too fresh.
C 2, Haygood Site.
An extensive midden stretches along the shore. Oyster shells are thick in the dark, sandy soil. The area of this occupation is a narrow ridge, four hundred feet wide by about fifteen hundred long. Scattered throughout the whole area are occasional potsherds.
C 3, Kitty Hawk Bay Site.
(Page 40) There are hills about fifteen or more feet in height that are capped with an oyster-shell midden. Generally, the shell is not more than a foot deep, but in a few spots it is impressively thick and when tumbled down the hillside makes a white scar. Sherds are rare but widespread in the shell. There are conflicting reports about how much the land has been cut back here, but all agree that there has been some regression. Most competent observers, such as Mr. Horace Dough, who has lived more than half a century in this vicinity, think that retreat has been on the order of one hundred feet.
C 4, Rush Point Site.
A layer of shells about one foot thick extends along the entire face of the sandy hill that has been cross sectioned by wave action. The shell layer is about twelve feet above sea level and contains little other than shell. However, wood charcoal, potsherds, and a hammerstone were plucked from the layer. Numerous sherds were found at the water's edge directly at the foot of the cut bank. As indicated above, all shells seemed to be oyster.
C 5, Stick Site.
(Page 41) At the north edge, the dune sand is quite high, probably exceeding thirty to forty feet in places. At varying heights through the dune sands there appears a thick lens of black humus. Nearly everywhere this humus is about one foot thick and contains no midden remains. A volume of several hundred cubic feet was sifted without finding anything. Occasionally this layer dips to lower levels and in each dip there is a shell filling of perhaps two feet, very densely packed, and consisting almost wholly of oyster. These spots are about ten feet above tide but the area of dark humus sifted was about twenty five feet above sea level. In the shell midden no bone nor stone artifacts were found but a few shards were recovered by sifting. In the water's edge shards and flints were recovered.
The great body of tidal waters called Albemarle Sound seems a most ideal locality for a heavy aboriginal population, but in only a few places does the expectation seem fulfilled. These spots were sufficiently remote from one another and so located that a reasonably good estimate of the archaeological content of the environs of Albemarle was obtained. Further, it was hoped that test excavations at two of the sites (A 1 and A 12) would establish definite relationships among several pottery wares.
(Page 42) The collections of the Laboratory of Anthropology, University of North Carolina have some materials from Tyrrell County. The pottery assemblage of this is dominated by the typical net-impressed (net paddled), sand-tempered sherds. In addition to the ubiquitous cord-marked and fabric-impressed types, steatite bowl fragments occurred. Broad, roughly chipped projectile points seem a characteristic of the Tyrrell materials. The materials are quite similar to those from both A 2 and A 12 as described below.
The following sites represent points at both ends of Albemarle. The north bank of the Sound probably has many more sites in the Pasquotank and Perquimans rivers areas. It was not possible to survey the area completely. The south bank of Albemarle gave much promise of occupation because of the high bluffs but nothing much was found. This is more or less as indicated by Lane and later visitors to the Sound.
A 1, Wright Memorial Bridge Site.
Scattered over a considerable area, is a cover of shells. This is an extensive midden that at one time probably covered the (area) above storm wave line. Local construction of a variety of types has altered the appearance and distribution of the site. When the approach for the bridge was constructed the major portion, and probably the thickest portion, of the site was used for fill. Skeletons were exposed during this operation. A persistent local story is that the burials were in a sitting position. Nothing was saved. Presently, a large area about one hundred yards square is preserved that now clothe the midden. In an (Page 43) undisturbed part of the midden a test excavation was made. A detailed report of this is below.
A 2, Pine Bluff Site.
For a distance of close to one thousand yards along the bank of (the) River is fairly high. For some four hundred feet of this distance sherds may be found directly along the crest of the bank. Evidently, a humus soil developed along this river bank and at about ten feet above the river sherds and other artifacts were recovered. Jetties have been constructed along the bank at this point in an effort to stop the erosion so evident. The winter storms from the west that do most damage.
A 3, Whaley Site.
This site is quite typical of many encountered in the Outer Banks area: the midden is concentrated along the edge of a terrace over-looking some body of water. The Whaley site is about three hundred feet long but not more than seventy five feet wide. The greatest depth hardly exceeds two feet. That some land (but how much is unknown) has been removed from the terrace edge is apparent, but whether the site was once quite extensive or even connected with A 1, which is quite probable, cannot be ascertained now.
A 4, Harbinger Site.
(Page 44) Along a rather large area--perhaps one-third a mile square--sherds and flint chips were discernible. No thick midden was evident.
A 5, Sampson Point Site.
On the Currituck Sound, a thin scattering of sherds and a few shells were found along the shore. In many places the wave-cut bank is five to six feet high, and about one foot at the top often is very dark humus material with shells and sherds included.
A 6, Waterlily Site.
Along the crest of this bank that a shell midden extends for about four thousand feet. Where discernible, the shell accumulation was about one foot thick. Sherds were scarce considering the great extent of shells.
A 7, Goose Site.
The extensive midden of broken shell. The midden, like A 6, of (Page 45) which it may be a continuation, is filled with broken shells that suggest a long period of plowing. The midden lies directly on the shore in an ideal location well above storm waves. Certainly a large population is suggested by sites A 6 and A 7 but it does not seen possible to definitely relate these middens with any historic tribes.
A 8, Ballance Site.
The site in a shallow midden of indeterminate extent. At the time of investigation the whole area was in pasture and only along the edge at the water line are sherds to be found readily. Projectile points and stone axes have been plowed up here, but no specimens or details are now available. Found here was a large bone slab, judged to be whale rib, about four inches wide, twelve inches long and one inch thick, that was much polished as though by wear.
A 9, Manns Harbor.
A thin midden appears that may be traced along the shore for another six hundred feet. This sparse sherd and shell evidence of aborigines is a less-than-one-foot-thick zone at the present land surface which is four to six feet above the level of the sound. Again, local residents aver that the Sound is encroaching upon the land. At this site numerous small stumps in the water suggest the earth has been recently removed from around them. Wave action in the shallow Sound must be pronounced (Page 46) during storms. It is indeed disappointing that no large collections were obtained here since A 9 must surely be the remains of the Indian village Dasemunkepeuc. This point will be considered below.
A 10, A 11 - south shore of Albemarle Sound.
The land is in many places about twenty five feet above the water. Ideal localities for Indian occupation are suggested but nothing except a few scattered sherds could be found. A 10 and A 11 are hardly to be recognized as sites: It seems reasonably certain that a more thorough survey would have revealed some small sites, but none of a large size seems present.
A 12, Bandon Site.
This fine midden is one of the largest in northeastern North Carolina. The results of test excavations are included in a later section. On the bank of the Albemarle Sound, the water is quite fresh at this point, and probably has been for centuries at least, since none of the shells of the tremendous lenses of the midden are oysters. The bluffs here are about twenty feet above the Sound and the midden is gradually tumbling into the water as the bluffs are eroded away. Presently the site extends approximately one thousand feet along the bank. Its width is difficult to ascertain. The greatest depth is judged to be less than ten feet. The test pit was slightly over six feet. It may be that the site was more extensive than the figures above suggest. Sherds have been found. Many projectile points have been found around two rather pronounced mounds of sand nearby. These latter are known to the local residents as Indian mounds but nothing short of excavation would prove this. They do have pine trees of considerable size but none that would.be more than one hundred years old.
Carolina Algonkian Project