The Search for Battery Halleck: Chapter 3
  • 3D Rendering of Shiloh Mound

    Southeast Archeological Center

    Cultural Resources National Park Service

SEAC: The Search for Battery Halleck




Spanish Hammock, the presumed site of Battery Halleck, is a well-defined finger of land, oriented east-west and overgrown in trees and underbrush, located in the tidal marsh in the north-central part of Big Tybee Island (Figures 1 95.6KB and 4 98.2KB). Situated approximately 100 meters south of U.S. 80, the hammock is on undeveloped land almost exactly one mile southeast of Fort Pulaski. That it was likely the location of Battery Halleck was discovered by Fort Pulaski National Monument staff rangers Talley Kirkland and Kent Cave in 1989, when they made a thorough comparison of Civil War period and contemporary USGS quad maps. Little change was evident in the local topography between 1862 and the present, facilitating the close relocation of the area where Battery Halleck was presumably situated. The only undisturbed, elevated ground in the area corresponding to the location of the battery on Civil War and modern maps was the elevated feature we have here named Spanish Hammock. 

In 1989 Cave and Kirkland contacted Edward L. Trout, the historian at Great Smoky Mountains National Park and formerly the historian at Fort Pulaski, to ask his opinion about their ideas on the location of Battery Halleck. Trout went to Fort Pulaski on May 22, 1989, and visited the Spanish Hammock area with the two rangers. In a subsequent memo to the Fort Pulaski superintendent dated May 31, 1989, he stated: 

I dropped by the fort last Monday (May 22), while on annual leave on Tybee Island, and was sorry to have missed meeting you. 

The purpose of my visit, aside from seeing old friends, was to honor Kent Cave's request that I look over what he and Talley [Kirkland] feel probably are the well-preserved remains of Battery Halleck. We walked the ground and discussed this interesting feature. Having been in Marine Corps artillery myself, it all looked and felt very familiar. We studied the three depressions on the little ridge, their relationship to each other, and their orientation toward the fort. The symmetry of the craters, as well as their symmetry of arrangement, all fit the gun- magazine-gun layout as described and delineated by General Gillmore and others. 

In efforts to rule out other possible explanations for this feature, we considered the notion that these may have been borrow pits. That did not make sense, as borrowing would have begun at the end of the ridge, not in the middle of it. We also considered that windthrown tree/root systems may have left the holes. However, it is unlikely that three trees of equal size, growing in an arc, would have blown over simultaneously. 

Back at the fort, we compared the enlarged Mylar overlay made from the Gillmore drawings to the modern chart for that vicinity. They are almost perfectly congruent. 

It all adds up on the ground. It all adds up on the maps. Only an archeological survey would tell for certain, but the evidence up to this point says that this feature is almost certainly Battery Halleck. It would be very exciting to see what Southeast Archeological Center could turn up, if they have time to do a survey. 

Given the rapid development I've seen on Spanish Hammock in recent years, this site may not long repose in its presently undisturbed condition. 

It was this visit and memo, in conjunction with the earlier map analysis conducted by rangers Cave and Kirkland, that prompted an inquiry for assistance to IASD in 1990, and that ultimately led to the fieldwork reported here. 

A detailed one foot (30 centimeters) contour map of the hammock area was produced during the 1990 field investigations. It provides a useful reference for the discussion of the immediate site area that follows (Figure 7 49.8KB). A modern, hardtop access road, Catalina Drive, cuts through the eastern end of the hammock and runs to a cluster of modern houses on Chimney Creek about three quarters of a mile to the south. These houses are on the edge of a rise marked Spanish Hammock on the 1978 USGS Tybee Island, Ga., S.C. 7.5' quadrangle sheet, and it is clear that it is from this feature that local residents have derived the name of the sand hammock explored in 1990. Most of the hammock, which is about 30 meters wide (N/S) and 200 meters long (E/W), is elevated 1 to 2 meters (3 to 6.5 feet) above the surrounding tidal marsh. At the time of the fieldwork, it was densely overgrown in live and scrub oak, saw palmetto, Spanish bayonet, and prickly pear vegetation. Approximately 75 meters west of the road, the hammock rises appreciably, to a height of 3 to 4 meters (10 to 14 feet) above the surrounding marsh. This elevated, central portion of the hammock extends for about 60 meters before it gradually drops back to the lower elevation (1 to 2 meters), feathering into the tidal marsh at the western end. 


On July 2, 1990, a limited archaeological reconnaissance of the Spanish Hammock area was conducted by Anderson assisted by Fort Pulaski personnel Cave (Chief Ranger) and Kirkland (Park Ranger). The reconnaissance, which lasted the entire day, found that the elevated central part of the hammock had been extensively disturbed at some undetermined time in the past. Three large, irregular depressions roughly 10 meters across and from 1 to 1.5 meters deep were found in the central and western portions of the hammock, in a roughly linear arrangement with each depression offset from the next one by a distance of about 22 meters (75 feet). The line of the depressions ran parallel and just back from the central line of the hammock, with the eastern and western depressions back against the marsh edge and the central depression near the crest of the hammock. The irregular ground surface, over and above the existence of the depressions themselves, indicated that extensive earth movement had occurred in this area. Parenthetically, the existence of a fourth depression, adjacent to the central feature (since then interpreted as the antechamber for the central powder magazine), was not recognized at this time due to the presence of dense underbrush in this area. This feature was discovered when extensive brush clearing occurred in conjunction with the mapping of the landform during the intensive field program in September. 

A metal detector survey was also made in and around the area of the depressions on July 2. The work was conducted for approximately three hours under the direction of Kirkland (Figure 8). A number of mid- to late-twentieth century artifacts were found, including 0.22 and 0.38 pistol, rifle, and shotgun shells; a 1964 nickel; and aluminum beverage ring tops. These collections, and all 1,154 artifacts found during the 1990 fieldwork, are described by provenience in Appendix 2. No earlier (i.e., pre- twentieth century or, specifically, Civil War period) artifacts were encountered, although the depth to which objects could be detected by the instrument was no more than about 10 centimeters (4 inches). Local residents had informed the park staff that the depressions themselves had seen use as blinds by hunters and as forts by local children. Tin and other metal debris, as well as asphalt shingle fragments, were widespread on the surface in and around the eastern and central depressions, and presumably represent crude roofing used for the toy forts or hunting blinds once present in the area. This metal debris greatly hindered the metal detection survey, but it likely also had the beneficial effect of frustrating looters interested in the area. 

From the visual inspection of the ground surface and the results of the metal detector survey, it was apparent that the upper deposits in the central portion of the hammock had seen extensive recent historic activity. The sides of the depressions appear to have slumped appreciably. A lightly worn trail ran along the northern margin of the three large depressions, and there is some suggestion that small amounts of fill have been added in this area to create a more level trail for bicycles or motor bikes, or possibly for easier pedestrian access along the spine/crest of the hammock. This recent filling is minor compared to indications of an earlier and much more extensive effort of earth movement encompassing virtually the entire 60-meter central section of the hammock, which the 1990 excavations indicate was associated with the excavation of the depressions and the creation of a parapet in front of them. 

A series of four small shovel tests were opened in the three identified depressions to check for subsurface deposits (Figure 8). One test each was opened in the eastern and westernmost depressions, while two were opened in the central depression. These units measured 30 by 30 centimeters on a side and were excavated by shovel, with all fill dry screened through ¼-inch mesh. The test opened in the easternmost depression (the presumed right battery) was placed in the exact center and went to a depth of 45 centimeters before terminating in a massive tree root. Two mid- twentieth-century artifacts-a 1955-D penny and a shotgun shell base-were found in the upper fill, as well as an appreciable amount of ash and charcoal from recent fires (Appendix 2). No other earlier historic materials were encountered. The depression itself measured about 20 (E/W) by 22 feet (N/S) (inside measurements). A small slot trench about 60 centimeters wide and of unknown age was cut through the east side of the depression. The south side of the depression abutting directly on the marsh (the presumed rear of the mortar emplacement) was lower than the north side, which faced toward Fort Pulaski. 

The two tests in the central depression (the presumed powder magazine) were located in the north central and central portions of the disturbance. This depression was found to be almost exactly 30 feet in diameter. Like the other two depressions, the south side was lower than the north. Both archaeological tests were taken to a depth of 90 centimeters, with all fill screened; no artifacts were found in either test. The northernmost test, opened on the interior slope of the depression, demonstrated that the fill in this area was all comparatively recent, and had apparently been placed there to create a more level trail along the crest of the hammock. The central test was taken to the water table, which was at 90 centimeters, with no evidence found for artifacts of any kind, or for obvious construction layers, burned materials, or artifacts. 

The single test in the western depression, the presumed left mortar, was also placed in the center of the bowl-shaped depression. Like the central depression, this feature was exactly 30 feet in diameter and, like the eastern depression, the south side opened directly onto the marsh. The test was taken to a depth of 80 centimeters, at which point water was reached. No evidence for artifacts or construction stages were encountered. A probe was used to examine the soil to a depth of a foot below the base of each unit, with no artifacts encountered. 

Although the limited archaeological testing conducted on July 2, 1990, did not yield conclusive evidence for Civil War utilization, given the small area examined and the extensive modern disturbance to the upper deposits, this finding was not altogether unexpected. Given the brief period over which Battery Halleck was constructed and in use, few Civil War period artifacts would, in fact, be expected, particularly since the various emplacements were set on wooden frames that were themselves soon removed. Previous metal detector survey by collectors, and the disturbance by local residents (i.e., hunters and children), furthermore, may have also removed obvious artifacts and other traces of Civil War occupation. Thus, the only evidence for Civil War period use of the area found during the initial reconnaissance was the existence and disposition of the earthworks themselves. 

On July 11, 1990, a memo describing the results of the fieldwork was submitted to the superintendent of Fort Pulaski. In this memo additional archaeological investigations were recommended to adequately map and test the hammock area for possible Civil War period remains. A week's worth of fieldwork for six people was proposed. Work would be directed by two IASD staff members, who would be assisted daily by at least four members of the park staff. This proposal was accepted, and plans for the fieldwork were initiated. 


Intensive archaeological testing activity was conducted at Spanish Hammock from September 17 through 21, 1990. The fieldwork was directed by IASD staff archaeologists David G. Anderson and John H. Jameson, who were assisted each day by a crew of from three to seven park personnel and a local volunteer, David Fuerst. The primary goal of the intensive testing program was to evaluate the possibility that the depressions and other disturbances found on the landform dated to the Civil War period, a finding that would reinforce the conclusion that it was the site of Battery Halleck. 

Mapping Activity 

Because the hammock was densely overgrown, the testing operations began with extensive vegetation clearing, specifically the removal of most of the dense underbrush, followed by detailed mapping operations. During the clearing, which was done using hand tools, a fourth depression, a probable antechamber or loading room, was found a few meters to the southwest of the central depression, which had been previously interpreted as the powder magazine. Once the landform was sufficiently clear to permit the effective shooting of lines-of-sight, it was mapped using a Lietz utility transit, a stadia rod, and a tape (Figure 9). A series of five temporary datums were established at about 30- to 40-meter intervals along the crest of the hammock. From these locations, distances to over 800 points (N=813) were shot, forming the basis for the site map (Figure 7 49.8KB). A permanent datum, marked by a piece of rebar, was established on the north side of the easternmost depression. This brush clearing and mapping occupied approximately one-third of the time devoted to the total field effort. 

The mapping documented the location and size of the four large depressions, and also found evidence to suggest that extensive earth movement had occurred in the central portion of the hammock. Appreciably higher elevations were found in this area, particularly in front of the depressions, suggesting an intentional effort had been made to raise the ground surface (Figure 7 49.8KB). Such earth-movement would be consistent with the historic accounts describing the construction of parapets in front of the mortars and magazine to protect the battery emplacements from hostile fire from the fort. The mapping also demonstrated that the central depression, interpreted to be the site of the powder magazine, was surrounded by appreciable quantities of earth, encompassing much more fill than was evident around the mortar depressions to either side. These deposits were particularly thick and well defined between this and the adjacent depression-the presumed antechamber-where fuse-loading would have occurred. Construction of a thickened earthen embankment around the powder magazine would protect nearby troops in the event of an accidental explosion of the stored materials. 

Detailed descriptions of how these features were likely constructed are given in the account by Lieutenant T.B. Brooks in Appendix 1. In brief, all were built on wooden platforms, surrounded by sand or earthen walls behind a high parapet, and, in the case of the powder magazine and antechamber rooms, roofed over with several feet of sand. In addition, covered shelters or splinter proofs for resting soldiers (the battery was operated around the clock by three reliefs) would have been located nearby. Appreciable quantities of both wood and iron, particularly iron fasteners, would have been needed to construct these features. Even though quickly erected and soon dismantled, some of these remains would be expected to have been left behind. 

Test Excavations: East Mortar Platform

Eleven one-meter test units, or TUs, were placed in three of the four depressions on the hammock during the testing program (Figure 10 55.7KB). Seven of these units were opened in the eastern depression, the presumed east or right mortar battery, in a north-south trench running from the marsh edge to the northern interior wall. Another two units, oriented east to west, were opened in the center of the presumed powder magazine. The final two units, also oriented east to west, were opened near the south wall of the presumed antechamber. No units were opened in the westernmost depression, the presumed left mortar platform. All of the fill was removed in 10-centimeter levels and dry screened through ¼-inch mesh (these units were combined into 30-centimeter levels in the field, reflecting upper disturbed and lower, less disturbed layers). An appreciable number of mid-twentieth-century artifacts were found in the upper levels of these units, together with a number of iron fragments and concretions much deeper in the deposits. These latter remains appear to derive from much earlier activity, possibly associated with the inferred Civil War period use of the hammock. 

The seven units in the eastern depression were opened to between 60 and 100 centimeters in depth (Figures 11,12, 13 54.8KB and 15). The northernmost four units (TUs 1, 5, 6, and 7) were placed across the center of the depression, while the remaining three to the south (TUs 2, 3, and 4) were placed through the slightly elevated rear wall to the edge of the tidal marsh (Figures 11 and 12). Extensive historic disturbance and large numbers of recent artifacts were found in the upper 20 to 30 centimeters of the five northern units that were opened largely or entirely within the depression. Very few artifacts, in contrast, were found in the two southernmost units on the opposite side of the sand rise at the depression/marsh margin. 

The upper 20 to 30 centimeters of the interior of the eastern depression had been filled comparatively recently, based on the large numbers of recent artifacts that extended through an old charcoal rich A-horizon located about 30 centimeters below the surface (Figure 13 54.8KB). This modern debris included glass, coins, bullets and shell casings, iron and asphalt roofing debris, aluminum pull-tops, and charcoal (N=953 artifacts; see Appendix 2). A few fragments of oyster shell and bird and cut large mammal bone were also found. These, given the several charcoal briquettes that were also found, appear to represent the remains of recent cookouts. The old A-horizon, which was a dark band about 20 centimeters thick, gives a bowl- like appearance to the bottom of the depression when it is examined in profile (Figures 13 54.8KB and 15). 

Below the dark A-horizon (recent debris) layer, distinct light gray merging to yellow sand layers were found that also conformed to the shape of the depression. These layers were thin at the center of the depression and much thicker at the front and back margins, where they formed much of the fill. Few artifacts, mostly modern materials in disturbances, were found in these layers, suggesting they were brought in as fill, perhaps to surround the central depression. Underlying these layers at depths from 60 to 90 centimeters was a strata of sand characterized by extensive mottled iron staining and heavily eroded concretions that appears to reflect an old disturbed surface. The nature of the deposits Battery Halleck was built on is stated in the historic accounts, where a Union officer observed that "the earth on which the Sherman and Burnside platforms rest, is a blue clay, or mud, mixed with sand. All the others are built on pure, fine quartz sand" (Brooks in Gillmore 1988:85 [1862]). 

While some of the iron staining in this lower level may be caused by tidally induced ground water fluctuations, the presence of occasional much larger, heavily weathered iron concretions in a few of the units, notably TUs 1 and 2 near the rear of the depression, suggests old historic disturbance. None of these concretions could, however, unequivocally be considered an artifact. The profile of the trench, which passed from the center of the depression through the back wall, demonstrates that the twentieth-century disturbance layers postdate the construction of the depression itself. Unfortunately, in the underlying levels, no artifacts were found in good enough context to give us the opportunity to date these deposits. 

In the construction of the eastern depression, earth was apparently removed from the central area and added to the northern face, or front of the hammock, adding to the height of the land surface there. Some earth may have been added to the rear as well, given the low rise in that area. The gray and yellow sand layers, conforming to the outline of the depression itself, were thickest near the margins. This suggests that fill was brought into the area to surround the depression and to build up the landform itself, perhaps to augment the height and thickness of the parapet. There are references in the historic accounts to the use of sandbags of fill in some battery areas; it is possible that the gray and yellow sand layers were formed in this fashion, although no soil distinctions indicative of individual bags or basket loads were observed. 

Test Excavations: The Powder Magazine

Two one-meter units were opened to a depth of 90 centimeters below the surface in the south- central part of the presumed powder magazine, in the deepest part of the depression (Figure 14). A few mid- to late-twentieth-century artifacts and one heavily weathered wooden beam fragment about 8 by 2 by 2 inches in size were found in the upper 30 centimeters of the deposits, together with one feature filled with charcoal briquettes that extended to a depth of 90 centimeters. These upper deposits were heavily disturbed and the presence of alternating sand and humus layers showed signs of one or more episodes of recent filling. 

Aside from a chunk of what appears to be melted styrofoam, probably from a disturbance, and a smaller wooden beam fragment like that found in the upper levels, no historic artifacts were found below 30 centimeters. Again, as in the eastern depression, heavily weathered iron concretions were fairly common. In this case, they were found from 30 to 60 centimeters in depth, and may be from older historic activity (Figure 16). Evidence for disturbance, probably caused by earth moving activity, was indicated from 30 to about 70 centimeters where the soils were mottled and appeared mixed. From 70 to about 90 centimeters in depth, appreciable iron mottling and staining were observed, again at least partially caused by ground water fluctuations, but also probably reflecting the weathering of iron on an old surface. The iron staining in this unit was much more pronounced than in the eastern depression, suggesting that more parent material (iron artifacts?) were originally present. 

Test Excavations: The Antechamber 

Two one-meter test units were also placed in the depression adjacent to and southwest of the powder magazine, in what was inferred to have been the antechamber or loading room (Figure 14). Placed on the back, or south side, of the depression near the marsh, these units were opened to a depth of 60 centimeters. The upper 40 centimeters or so of these deposits were highly disturbed (Figure 17). Under a thin recent humus layer was about a 30-centimeter- thick band of lighter gray sand that was, in places, mixed with, but more typically overlaying a darker humus zone about 20 centimeters thick that appears to be an old A-horizon/ ground surface. Below this was found a compact, undisturbed subsoil of mottled sand. Appreciable quantities of rusted iron, consisting mostly of small, heavily weathered fragments and concretions, were found in the levels above the old ground surface. While most of these fragments were so weathered they could not be considered artifacts with any assurance, two unequivocal concentrations of badly eroded and highly fragile iron plating were found resting on the old surface at a depth of 35 centimeters (Figure 18). While the age of this material could not be determined due to its poor condition, the extreme weathering suggests it had been in the ground for some time. 

Few recent historic artifacts were found in the antechamber units; no artifacts were found below 40 centimeters. The deposits in this feature are comparatively shallow and appear to have seen much less historic disturbance (i.e., in trash disposal or as a play or hunting area) than the other depressions. This is not surprising since the antechamber depression was extensively overgrown and not even recognized during the initial reconnaissance. The extensive weathering on the iron plating and the large amount of iron concretion-like debris in the depression's upper layers suggest that appreciable quantities of metal were used or at least discarded here. 

Chapter 4: Conclusions.

Chapter 4: Conclusions

Return to index or cover page