Ground Penetrating Radar Survey at the Horton Plantation, Jekyll Island, Georgia

Ground Penetrating Radar (GPR) survey of portions of the Horton House archaeological site on Jekyll Island, Georgia were completed in March, 2002 by Rocquemore Radar Research, Box Springs, Georgia, as a sub-contract with Southern Research Historic Preservation Consultants, Ellerslie, Georgia. This pilot study was the first application of GPR technology on an archaeological setting on Jekyll Island and one of the few that have been conducted on Georgia's Barrier Islands. The results of this work were successful and indicate that this technique has useful application for archaeological sites in this environment. The three days that were spent conducting GPR fieldwork at the Horton House site allowed only a small fraction of the entire site to be examined. The survey examined 12 sample blocks within a portion of the Horton House site that extended from grid point 511-785 North and 310-473 East, or a distance on the North-South axis of 274 meters and 163 meters on the East-West axis. These sample blocks covered an area of approximately 732.5 m2. The majority coverage was placed around the Horton House ruin with lesser coverage at the DuBignon Cemetery and "the Brewery". The balance of the site is in heavily wooded conditions, which are not conducive for GPR coverage. The GPR radargram profiles revealed numerous anomalies across the site, particularly in the upper 60 cm of soil. At greater depths very few anomalies were identified and some of these may not be archaeological in origin. Some of the deeper anomalies, such as the pronounced anomaly that was observed at 1.3 meters depth in the northwest quadrant of GPR Block 4, are likely the result of geological or other environmental factors (possibly salt water intrusion in the water table). The upper stratum of the Horton House site (50 cm and shallower) contains abundant artifacts, as revealed from the various excavations. Many cultural features also are present in this zone, so much so, that the GPR mapping of these shallower depths yielded too many reflected signals. The density of large artifacts, artifact clusters, and features in the 0-50 cm soil zone, was not well suited for the isolation of individual features using GPR. The major discoveries of the Horton House GPR survey included: • ditch or trench feature, possibly part of a defensive palisade that surrounded the Horton House, which was located South of the ruin; • two burials within the DuBignon Cemetery, probably the two drowned sailors and, • large rectangular cellar, which was located about 10 meters east of the DuBignon Cemetery gate. These suspected features should be verified by future excavations. Other portions of the site, which are currently wooded, should be considered as potential areas for additional GPR survey should the ground surface on these areas become exposed. The angled linear anomaly in GPR Block 3 forms an interior angle of approximately 112 degrees. Such an angle may have been designed to allow for enfilading fire by gunmen defending against an attack. Most forts from this era incorporated these "salient angles" in their design, following military fortification theory of Vaughban and others (Lewis 1970). The suggestion of a palisade surrounding the Horton House seems logical, given that Major Horton established his plantation house on the hostile Georgia frontier, which was threatened by Spanish attack in the early part of the Georgia Trustee period. Fortified house sites were not uncommon in the colonial period. Recent excavations at the Galphin trading post at Silver Bluff in Aiken County, South Carolina, have revealed a substantial palisade surrounding. Galphin's house. That fortification dates to the 1740s and serves as a possible analog for the Horton House. Alternatively, this linear anomaly may represent a planting alignment in a formal garden, or possibly an undocumented utility trench. The two probable burials that were identified in the DuBignon Cemetery may be the same people who are commemorated with headstones. These stones indicate that the two men drowned on Jekyll Island. The headstones are located on the southern part of the enclosure and the suspected graves located by GPR are on the northern part. If these are one in the same, this suggests that the tombstones were erected sometime after the men were buried and were possible placed in the wrong location. Alternatively, the tombstones may actually be cenotaphs (markers placed to commemorate a dead person whose remains are not present) placed to commemorate two men, who were lost at sea and their bodies never recovered. This possibility could possibly be corroborated by a review of newspaper accounts of the tragedy, or from other historical records. If these are cenotaphs, then the two probable graves may be unrelated to the drowning victims. The fact that one of the tombstones has an accompanying footstone, however, supports the argument that these are not cenotaphs. The GPR survey did not identify any probable graves outside of the brick enclosure at the DuBignon Cemetery. The entire area around the enclosure was not examined, however, so the potential for additional graves cannot be completely eliminated. The initial conclusion, however, is that this cemetery was built exclusively for the DuBignon family, and was used sparingly in later years. The cemetery for the enslaved African-Americans, who were exploited by the DuBignons was likely located elsewhere.The suspected rectangular cellar, which was approximately 10 m east of the cemetery gate, measured approximately 7 m Northeast-Southwest by 5 m Northwest-Southeast. The age of this feature is not known, although it is suspected to be associated with the historic occupation. The orientation of this suspected cellar does not follow any cardinal direction. The DuBignon Cemetery enclosure shares this trait. In contrast, the Horton House was oriented along cardinal directions (Magnetic North). Generally, the Horton House site showed promise for the application of remote sensing geophysical techniques. These techniques are non-destructive and therefore may be a desired approach for long-term management of this State-owned archaeological site. Knowledge of what lies beneath the ground is useful information for managers who need to select areas for disturbance, such as utility lines, fences, parking lots, and posts. In the past few years GPR equipment has improved significantly and continued improvement is to be expected. In the near future the results of the present study may be vastly outmoded by new technology. Improved techniques and equipment may allow for the recognition of cultural features at very shallow depths. This would be an important advance for archaeology in Georgia since many sites are relatively shallow and features often extend only a few centimeters below the A-horizon. Such subtle differences were not recognizable with the equipment that was used in this study. The current configuration of the RAMAC 3XM radar unit and its cart did not allow for the attachment of a 1000 MHz antenna, which was unfortunate. A higher frequency antenna would provide more precision in mapping very shallow features. Other self-criticism of the present study concerns the machine settings that were used for the sampling frequency, number of samples, and number of stacks. The project would have probably benefited from greater experimentation with these variables, but time and budgetary constraints prevented this. A 500 MHz antenna was available for use in the survey but it was not employed since the 800 MHz antenna allowed for a better examination of the soil depths that were of most concern to the archaeologists. Hopefully, the present study will serve as a stepping stone in the continuing learning process and future projects can build upon it.