Search for Fort Daniel

Submitted by Jim D’Angelo and Shannon E. Coffey ( and

The Gwinnett Archaeological Research Society (GARS) has been conducting an archaeological investigation at the traditional site of one of Georgia’s earliest frontier forts, Fort Daniel, in Hog Mountain (Hamilton Mill), Gwinnett County, Georgia.

New study of old documents shows that Fort Daniel was not original to the site, but rather constructed in late 1813 to replace an earlier fort dating to at least 1799 and perhaps earlier, when we know there was a militia at Hog Mountain. The traditional site of the second fort, and probably the first fort as well, is located on a 4-acre parcel of private land that is currently for sale and may be subject to commercial development. With the owner’s permission, GARS has been working at the site to establish the existence of archaeological remains dating from this period, and to determine what those remains represent.

The research design for this investigation included several phases, which have been carried out in succession: clearing the project area of underbrush, saplings and dead trees; laying out a 200×160-foot grid over the approximately 0.8-acre area to be investigated (Figure 1); shovel testing to characterize the soils at various points on the grid; creating a local 1-foot interval topographic map and mapping all surface features; carrying out an intensive metal detection (MD) survey of the upper approximately 1-foot of soil (Figure 2), followed by a ground penetrating radar (GPR) survey (Figure 3); and, based on the results of those surveys, carrying out limited archaeological excavations including mechanical stripping and hand-excavated test units (Figure 4). All excavated soil from two 20×20-foot units (a total of approximately 1200 cubic feet) is being screened (Figure 5), and soils from stripping in four other units has been replaced (and seeded) without screening. All recovered artifacts have been provenienced and registered in the field before being sent on for cleaning, stabilization, and curation. The results of this investigation along with historic research that is being conducted concurrently will then be published in a technical report as soon as possible.

Partnering with GARS in this ambitious project were members of the Gwinnett Historical Society; TRC (Norcross), whose Lab Director, Tommy Garrow, is cleaning and stabilizing metal artifacts; local members of a metal detecting club who ably assisted with the MD survey (and came away with a new appreciation of archaeological context); the Student Association of Archaeological Sciences (SAAS), a student-run organization loosely associated with The University of Georgia that conducted the GPR survey; the Gwinnett County Parks and Recreation Division, which provided machinery and an operator for clearing of push piles and mechanical scraping; and the Gwinnett County GIS Department, which has provided a Digital Elevation Map and rectified satellite image of the site and vicinity, and has offered continued support for developing a project GIS.

The project has yielded a large number of wrought nails of various sizes as well as some early cut nails of various sizes (Figure 6). Several ‚“buck and ball” buckshot balls, two period brass-plated buttons, and a variety of other metal objects have been recovered. There is also some historic pottery dating from the late eighteenth century through the nineteenth century, and thick green (‚“black”) wine/ale bottle glass.

As expected, given the landform and availability of water nearby, this is a mixed component site. Quartz lithics seem to be ubiquitous across the site and include cores, secondary and tertiary flakes, one Wade-like projectile point, and a white glass bead. According to J. Brain’s schema, this is a Type IA2 glass bead, which might be dated to 1725 based on one excavated example. Brain has no bead data from Georgia. [GARS had previously recovered a similar Type 1A3 blue trade bead from the Creekside Rock Shelter on the Elisha Winn property that, based on several known excavations, is given temporal range of 1650-1833 with a mean date of 1726.]

Ground-truthing of the GPR survey, which was limited to about 25 percent of the site due to time constraints, has so far turned up one linear feature and a possible hearth feature as well as several ‚“post holes” that turned out to be tree or rodent holes. The linear feature exhibits charcoal and hardened burned clay several inches deep. These features will be excavated in the near future. Unfortunately what was thought to be stockade wall trench turned out to be a deep vein of quartz. However, selective stripping in an area not covered by the GPR survey has intersected a north-south, 20-inch-wide trench that may represent the east stockade wall. This also will be further investigated in the near future.

A PowerPoint presentation on ‚“The Search for Fort Daniel,” including results of the GPR survey, will be made at the October meeting of SGA. A poster-board and artifact display will also be presented by members of GARS.