Known and Suspected Prehistoric Sites in Flagler County, Florida

A list of known and suspected prehistoric sites located throughout Flagler County.

September 1977 collected information by Sisco Deen

8F146, Marineland Midden (located at Marineland)

South along the coast a distance of 2.4 miles from the Summer Haven midden is the town of Marineland, consisting mostly of the marine tourist attraction. Constructed in the 1940’s, and still undergoing expansion, Marineland sits on top of the largest aboriginal site in the study area. The location attracted the attention of both Douglass (1885) and Webb (1894) in their surveys of the Atlantic Coast; both of these investigators supplied unusu­ally detailed descriptions of the site. In 1941, a report was written by Vernon Lamme concerning excavations he made at the site, but no copy of the unpublished manuscript could be found. An apparent error in assignment was noted in plotting the recorded sites near Marineland. The earliest description, quoted at length below, is by Douglass (1885:75-76), under the name of Dupont’s Mound. From that reference, it was assigned the number 8F12 by Goggin. Such an impressive site was not overlooked by Webb (1894:697) who called it “large mound below Matanzas Inlet,” and illustrated its beach pro­file in a photograph (1894:Plates 83,84). From this source the site was assigned 8FL6 by Goggin. Based on these two descriptions, there is little doubt that the two sites are one and the same. 

Since the name Marineland is appropriate and since another, unrecorded midden was discovered during the present study at the location of Dupont’s homestead, it is suggested that the number 8Fl2 be closed. A new number, 8FIIO is assigned to the site at the Dupont homestead. 

Because it has been nearly completely destroyed in the past 30 years, both the Webb and Douglass descriptions are quoted in full. Additional information was collected during a brief visit as part of the present study. 

Douglass’s description is the earlier, but it is unlikely that the site changed significantly between the two visits: 

Most of the shell mounds on the coast are progressive in their construction and exhibit this very markedly when carefully examined. On the South Matanzas River, twenty-five miles south of St. Augustine, this characteristic is very beautifully shown in Dupont’s Mound. This Shell Mound extends from the sea coast westerly to the interior lagoon. Its length is about half a mile, and its width the same. On the sea coast it has been abraded by the surf, and some years since, its precipitous bank fifteen feet in height, afforded the best facility for examining its successive layers of shell debris. The black lines showing the hearths dark with burnt shell and mingled with fragmentary pottery, indicated successive stages of occupation. Mingled sparsely with the surrounding shell could be extracted bones of fish and fowl, of turtle, alligator and deer. From the stratification of the shell could be inferred the pre­ference of the natives for this or that shell fish. The conch, the clam and oyster, the minute donax, the large snail, were sometimes intermingled in the layers, but occasionally one or the other kind preponderated; the donax sometimes appearing in strata six inches in thickness. The oyster, however, was most abundant. The growth of this mound was evidently from East to West, and we could inspect at this portion of its bed its most ancient strata. Where the sea had abraded it, the perpendicular height, including three feet of super­incumbent soil, was twenty feet, and it was based upon a low outcropping of Coquina or shell conglomerate. The surface of the mound was rather rolling than level, stretching along in easy undulations covered for some hundred feet with a thick growth of the Spanish bayonet and scrub palmetto. This again gave place to a heavier growth of live oak and cedar, bay and wild orange; the surface became more irregular, out­lying moundlets of shell started out from the marsh, linked to the parent mound by a low shell ridge. The mound was now encroaching upon the area of the lagoon and still had reached but half its length. The surface generally dropped to a lower level and became more hillocky, the undulations were steeper and shorter, at length it resembled a giant honey~ comb. We were threading pits about fifteen feet in diameter, flanked and separated by ridges of from six to eight feet in height. Gaps in the ridges allowed passage between the several pits. It required but a small stretch of the imagination to imagine each wigwam nestled in its separate enclosure, its occu­pants lounging about the fire or crowding through the convenient pathways, while day by day the shell ridges grew in height and threatened to fall in and crowd out the tribe. At this point some calamity has swept them away forever, steadily, surely but slowly, vegetation closes over the scene, centuries pass, and the shells are buried far out of sight in a soil which nurtures the most massive trees under whose shade we pass out from the cluster of hillocks directly upon the bank of the lagoon (1885:75-76). 

Webb’s description, though perhaps less eloquent, is equally detailed: 

… disposition of remains gives a clue to the manner of formation of the mounds and is well shown in the large mound below Matanzas Inlet, which covers more than 30 acres (Pl. LXXXIV). The side facing the ocean is from 10 to 12 feet in depth, but has suffered from the encroachment of the sea to an extent which cannot be determined (Pl. LXXXIII). The highest part of the mound covers almost 2 acres, and back of this, extending to the Matanzas River, lies the remainder, disposed in circles of greater or less extent and covered with forest. These circles adjoin each other over a large part of the territory. They are from 4 to 8 feet across at the bottom. This was a dwelling place, and the daily refuse was thrown out on all sides, and so the circles of shells, bones, etc., gradually grew higher and higher, surrounding the rude dwelling like a wall. This wall would also serve for protection from the winds of winter and likewise as a pit for defense in case of attack. When this hol­low had become too deep I or the wall about it too high, it would be abandoned, and the owner, pitching his tent on the top of surrounding ridges, would use the hollow as a pit in which to throw refuse. The mound of which I am now speaking would appear to have been in some sort of a center of population for many miles around. A spring of water lies in the midst of it, and the waterway was kept open to the river. Smaller mounds are found scattered up and down the river for several miles in the vicinity (1894:697).

The Marineland Midden was visited briefly during the present study and found to bear little resemblance to its former condition. At the beach side where 15 feet of stratified midden once stood are now found the aquaria, pools and galleries of Marineland. Development extends across A1A towards the Matanzas River including tennis courts, a yatch basin, and administrative and research buildings. A small portion of the midden was found intact within an oak-palm hammock along the river, but even here most of the midden had been borrowed away leaving large Live Oaks pedestaled on top of 3 and 4 feet high islands of midden. A probable burial mound was located, and small rings of shell like those described above, but considerably shallower, could be recognized among the vegetation. On the basis of collection in the Yale Peabody Museum and the Charleston Museum, Goggin classified the site as St. Johns II. 

8F11O, Homestead Midden (1.5 miles south of Marineland)

Continuing south along the barrier beach we pass the mouth of Pellicer Creek. At least three sites are known along its banks, but as these are in a different environmental zone: they will be considered separately afterwards. About one mile further south, the Homestead Midden, previously unrecorded, was found during the reconnaissance survey. The site was recognized by its signature on-­false color infrared photography and noticed during an airplane flight, primarily because of its elevation and vegetation. Permission could not be obtained to survey the area on foot, however, the site was visited for confirmation. At the same location are found the Dupont cemetery and homestead. No buildings remain; however, a house symbol is shown on the Randolph plat of 1850. 

8F111, Washington Oaks Midden  (2 ½ miles south of Marineland)

Also unrecorded, this midden was found in the same way. It is located within Washington Oaks State Park about one mile below 8FllO, in an essentially identical environment. Late in the Second Spanish Period, J.M. Hernandez gave the 395 acre parcel on which this site is located to his daughter and son-in-law. The plantation Bella Vista was built soon afterwards; the original buildings are no longer standing. No cultural material was collected during the brief visit, but the apparently little disturbed midden was found to extend several hundred feet along the river. It was not deter­mined whether additional shell ridges exist along the shoreline either north or south. Such a pattern is likely. 

Seven aboriginal sites have been described in what might be termed a back bay-coastal situation. The pattern is well known for the east coast (Goggin,1952; Rouse, 1951) and is expressed south of the Palm Coast property in Vo1usia County beginning at the Tomoka Basin. No sites of this nature are recorded between the mouths of Long and Bulow Creeks, most likely because the estuaries which supplied abundant molluscs do not occur. The elevation respon­sible for the watershed boundary along the Lehigh grade seems to be a limiting factor.

When examined from an environmental perspective another pattern of site location is evident. It is represented by 2 recorded habi­tation sites on or near Pellicer Creek. Only one of these sites was visited; the other is recorded on the north side of the creek out­side the survey area. 

8F112, Wadsworth Midden (Mouth of Pellicer Creek)

At the north boundary of the Palm Coast property at the mouth of Pellicer Creek a large out-holding of more than 600 acres is owned by Lewis Wadsworth. An extensive shell midden was noted in this area on both sides of the property line, but since permission for access was not obtained, the northern extent and nature of the site are unknown. It is likely that habitation would extend around the point and into the mouth of Pellicer Creek, but as far as is now known, 8Fl12 appears to be oriented more towards the estuarine resources of Matanzas River than Pellicer Creek. with respect to exploitable species the site differs from the back bay-coastal sites only to the degree to which access across the Matanzas River was limited. Francis Pellicer owned and farmed a grant of 1,000 acres which includes the Wadsworth out-holding (Quinn, 1975). A plantation owned by Pomars is shown on an early map, and the property was later a hunting preserve, occupied, according to informants, by a Russian Princess (actually Angela Mills, wife Henry Mason Cutting, John Lorimer Wooten and Prince Boris Scherbatow – CSDeen). One piece of Spanish Olive Jar pottery was found on the Palm Coast portion of the site. 

8F13, Rhotan Midden (Pellicer Creek, North Bank – St Johns County)

Continuing up Pellicer Creek the next recorded site is the Rhotan Midden. The site was investigated by Douglass who provides what little we know about it: 

Immediately opposite this mound (Marineland Midden, 8F16), Pellicer’s Creek opens out on the western side of the lagoon, and following up its course about three miles we reach on the north bank, Rhotan Landing, where a shell bank and field indicate an early Indian village. Three-quarters of a mile westward from this landing is a sand mound known in my records as Rhotan Mound. The distance between this and Dupont’s Shell Mound (Marineland Midden, 8F16) is about five miles, about one and a half miles of which constitute the breadth of the Lagoon (1885:76). 

The site was not visited during the reconnaissance, but should fairly easy to find. 

A third group of sites which becomes apparent when locations are plotted consists of inland ceremonial mounds. Four are known in the area including the Rhotan Mound already mentioned. Since Douglass our only source, the quote is continued: 

8F14, Rhotan Mound (Pellicer Creek North Bank – St Johns County)

This is the first Indian Mound I had the oppor­tunity of exploring. Its form was that of a trun­cated cone, with unusually steep sides, rising at an angle of forty-five degrees. Its height above the trench at its base was twenty-two feet, and above the natural level of the old fields in which it stood, eighteen feet. The summit was a slightly convex plateau of thirty feet in diameter and the diameter of the mound at its base was eighty-five feet. It was composed of a fine cream-colored siliceous sand, and at the time I visited it, it was covered, both summit and sides, with a moderately thick growth of trees, some of them fifteen inches in diameter. The sand composing this mound had evidently been taken from the trench at its base, but it was now doubtless reduced in depth by the washings from above, during the lapse of centuries, and it melted so gradually and imperceptibly into the remote level of the plain that it was impossible to determine where the excavation terminated. The trench in its course about the mound was interrupted by three un-dug strips of the original level, equi­distant from each other. They were about six feet in breadth and were a very marked characteristic of the base. It might be inferred that the aboriginal constructors had a special reverence for this peculiar figure, attaching to it some religious or superstitious signification, but I imagine it was simply the design of the savage to reduce the labor of ascending with his load, when the increasing width of the trench com­pelled him to take it from a distant point. Gathering the soil in baskets from the widening excavation, he must descend into the trench, and again ascend the soft, fresh sand upon the steep slope of the mound, but by reserving this natural causeway he could avoid the extra descent and reach the slope at a point con­siderably above the base, and have a reduced proportion of the soft slope to encounter. The exploration of this mound was the work of several days, and not very fruitful in results. Burials, no doubt original ones, were found ten feet from the summit, but too decayed to be preserved or gathered. They comprised bones massed together with the skull on top, occupying a space about two feet by ten inches. The adjacent soil was slightly discolored by the red paint with which they had been coated, frequently only a few teeth were discernable. It was impossible to estimate the pro­bable number of burials–a matted mass of beads was exhumed at a depth of five feet, and, at a depth of four feet but quite apart from each other, were found three celts and two fine quartz pebbles, one grooved as a sinker, the other evidently at one time fitted with a handle and used as a skin polisher. I have singled out this mound rather as exhibiting the natural cause­ways designed to save labor. I rarely observed it in mounds of moderate height, say six to ten feet, but on larger mounds the idea was expressed in some similar way (1885:76-77). 

8SJ35, Pellicer Mound (Faver-Dykes State Park – St Johns County)

No information is recorded about this site except its approximate location and an earlier name: Yelvington Mound. Deed records or local informants may be useful in finding it. One long-time resident who was interviewed mentioned a mound within Faver-Dykes State Park, the exact location of which was known to park personnel. This could be 8SJ35, or might be the Rhotan Mound, 8F14.

8Fl13, Eatman Mound (Near Long Creek)

Situated on Palm Coast property in the vicinity of current construction, the Eatman mound was not recorded until the present study. Its location was revealed by a local resident. The site was visited and found to be heavily damaged. Bulldozer tracks led to the site from a recently cleared area where roads and lots were laid out. The vandalism was so recent that only a few small, grassy weeds had been able to grow up in the fresh, white sand. Since little weathering or erosion had yet occurred, the details of the mound’s destruction could be clearly recognized. 

A few small areas of undisturbed soil showed the mound to have been about three to four feet high and between 50 and 60 feet in diameter. Planted pines surrounded the site, but the mound itself had been apparently fairly clear. From the south­west corner the bulldozer began long, deep cuts across the diameter of the mound, each time choosing a different angle. Within a short time only a few isolated grassy spots, mostly hidden by dirt already pushed up, remained. Large piles of sand were spaced about the perimeter and cutting in the center had probably reached the original ground level. The bulldozer operator, or perhaps a later visitor, walked across the site a few times, picking up human bones and pottery, then leaving the collection in a pile in the center. Bulldozer tracks led away to the south. The mound, which had stood the tests of perhaps 2,000 years, was destroyed in probably less than 15 minutes. A surface collection consists of the following specimens: 

35        human bone fragments 
01        shell implement 
03        hematite nodules 
05        sand tempered plain sherds 
61        St. Johns Plain sherds (incl. 8 rims) 
02        Orange Plain? or semi-fiber-tempered? sherds 

The assemblage is not particularly diagnostic and a St. Johns I period assignment (100 B.C. to A.D. 800) is the most accurate that can be made. 

8F11, King’s Road Mound (2 miles north of Bulow State Park)

One last mound within the study area remains to be discussed, the King’s Road Mound, located about 2 miles north of Bulow Plantation. Douglass, again, is the original source:

A striking mound of somewhat similar construction is found near the southern boundary of St. Johns County, about three miles north of Bulow Forks, on the east side of the old “King’s Road.”  A spruce ridge, extending in a direction nearly northwest and southeast for several miles, here drops suddenly off, where it is intersected by a stream or branch  running eastwardly toward Bulow Creek, one of the northern affluents of Halifax River. The ridge, though a natural one, is slightly undulating, and the last knoll had a peculiar prominence which suggested the possi­bility of an artificial treatment. A further exploration disclosed the fact that it was a veritable Indian mound, formed on the last easy slope of the natural ridge, by cutting away the intervening ridge on the summit and side and heaping the earth upon the separated section of the ridge. The result was a mound eight feet high as viewed from the plain below. A thorough excavation revealed the faint but easily discernible line of the original level, about four feet above the lowest level of the connecting neck, which was narrowed to ten feet in width by exca­vation for earth from the slopes on either side. This ingenious construction produced an exceedingly imposing effect from a comparatively small outlay of labor. The only visible remnants of burials were human teeth and fragments of jaws. The relics were three celts and numerous shell beads (1885:80- 81). 

Douglass’ directions were detailed enough to locate the site on a quadrangle sheet, and when the site was visited his description was likewise found to be accurate. The mound is well known locally; at least three informants mentioned it. While not nearly so disturbed as the Eaton Mound, this site has been extensively dug in by collectors. No artifacts could be found on the surface, and with the exception of some small beads, no material is known to have been found by recent digging. 

8F15, Bon Terra Farm (6.5 miles south of Marineland)

One last site in the study area is recorded in the Florida Master Site File. This is 8F15, Bon Terra Farm. Although situated very near the coast, it was not included in the back bay- coastal group of sites because it may predate human occupation in the region. The site was originally discovered during canal dredg­ing in 1931 and brought to the attention of J.H. Connery at Rollins College. During excavations, Connery recovered vertebrate fossils, mostly undescribed, and a “chert arrow-head” in alleged association with a mammoth mandible. Bone needle pieces were found nearby (Connery, 1932:516).

More serious and extensive investigations were undertaken by E.B. Howard (1940), during which a rather large vertebrate fauna was identified. In 165 recovered specimens the following animals were represented: mastodon, camel, turtle, deer, tapir, alligator, beaver, mammoth, rattlesnake, giant sloth, horse, and stork. Two significant characteristics of the site were discovered: the bones were scattered rather than articulated, indicating secondary deposition, and no human remains or artifacts were found. Similarities between this site and the early man sites at Vero and Melbourne were noted by Howard. 

The question of human remains in association with extinct ver­tebrates at Bon Terra was considered again by Wilfred T. Neill (1953). In examining the site and the material recovered by Connery, Neill found the chert arrowhead to be in fact a stemmed scraper, an artifact type known to be much later than Late Pleisto­cene when the bones would have been deposited. Again, the fauna was compared with the “so-called Melbourne bed.” Neill notes that according to local residents, the site was underwater until the 1920’s when the Intracoastal Waterway was straightened. This altered the drainage pattern, draining the site which was then farmed. 

Similar late Pleistocene vertebrate localities have been noted around the coast of Florida. They exhibit a fauna which is suc­cessfully compared to Rancholabrean age fossils in the southwest .U.S. (Webb, 1974). Vero and Melbourne are the only two sites of this series to have yielded human remains. Since the faunal col­lections from Bon Terra and these two South Florida sites are so similar, early man in Flagler County is a possibility. 

Performed under contract for:
Pittsburg, Pennsylvania by
Tallahassee, Florida
James J. Miller & Robin Strassburger
September, 1977

Given to Claude Sisco Deen, Jr. by Edward P. Boyhen, 33 Cochise Court, Palm Coast, FL

Author: Sisco Deen