An Archaeological Overview and Assessment of the
Main Unit, Petersburg National Battlefield, Virginia

Chapter Six:  An Assessment of Archaeological and Historic Resources (Part A)

Project Origins

The Department of Anthropology at the University of Maryland, with the financial support of the National Park Service, undertook an inventory and assessment of the archaeological and related historical sites within the Main Unit of Petersburg National Battlefield during 1999.  The Main Unit is located within Prince George County, Virginia, east of Petersburg on the south side of the Appomattox River.  Petersburg is famous for the Civil War siege from June 1864 to April 1865 that led to the fall of Richmond and the surrender of the Confederate Army of Northern Virginia at Appomattox Court House.

The Main Unit contains extensive physical remains from the Civil War siege, including earthworks, battle sites, and encampment grounds.  Since the inner coastal plain of Virginia—the location of Petersburg—witnessed repeated occupation during the prehistoric and colonial periods, the Main Unit contains archaeological traces of many periods of Virginia's past, from early Native American hunter-gatherers through English colonists and African-American slaves to the Civil War and post-Civil War eras.

Prehistoric Evidence

As mentioned previously, a limited number of excavations have been conducted within the Main Unit, and those have focused on historic sites.  The only prehistoric site that has been positively identified is a Late Archaic Halifax component in association with small lithic assemblages at the Taylor house site on a ridge between Poor Creek and a tributary.  A prehistoric pottery sherd was recovered near Confederate Battery 5 on the bluff of the 100-foot terrace between Harrison Creek and an unnamed stream to the north.

A survey was undertaken by MAAR Associates within the adjoining grounds of Fort Lee in the 1980s (Opperman and Hanson 1985).  The site inventory from the portion of Fort Lee nearest the park (i.e., the portion on the Prince George quad) constituted Table 3.2 in this report, and the applicability of the survey data to an interpretation of prehistoric occupation of the Main Unit was discussed in Chapter Three.  Survey data indicate a predominance of sites related to the Late Archaic and Middle Woodland phases.  Mouer et al. (1985) suggested that Middle and Late Archaic sites are found locally in a wide range of locations, but particularly in uplands, higher terraces, and on upland spurs above stream confluences, an observation supported by Opperman (Opperman and Hanson 1985:5-2, 5-3).  The Taylor house site is positioned on a ridge above Poor Creek, a location likely to have Late Archaic occupation evidence. The database from Fort Lee suggests that other Late Archaic localities probably exist within the Main Unit.  A stratified random sample was discussed in Chapter Three.

Mouer et al. noted that Later Archaic/Transitional sites are, by contrast, often found along lower stream terraces.  Opperman (Opperman and Hanson 1985:5-2, 5-3) extended the latter observation to encompass Early and Middle Woodland sites.  It has been argued that Transitional and Middle Woodland phases reflect settlement systems with larger habitation sites (aggregation of "fusion" pattern) and small procurement/habitation (dispersal or "fission") sites (Mouer et al. 1985).  The possibility of such smaller sites from these later phases within the Main Unit clearly exists.  Opperman reinforced statements by Turner (1976) that the inner coastal plain of Virginia witnessed high population densities during the Woodland period, and further that the role of the interior sites (i.e., those not along the major rivers) in the overall settlement system was not fully understood (Opperman and Hanson 1985:3-6, 6-5).  Prehistoric sites within the Main Unit would therefore contribute to an understanding of the settlement system on the inner coastal plain.  A particular question in this regard would be the meaning of the different settlement patterns manifested during the Late Archaic and Early-Middle Woodland phases.

Historic Domestic Sites and Potential

The potential for seventeenth and early eighteenth-century domestic sites is a realistic one, although most seventeenth-century sites in Virginia are river-oriented loci at low elevations, i.e. generally less than 30 feet above sea level (Smolek, Pogue, and Clark 1984).  Definite evidence of c.1760-70 occupation has been found within the park along the higher elevations east of Petersburg and the Appomattox River.  The 1863 map of the Confederate defenses surrounding the town (O.R. Atlas 1863) indicates the precise locations of these late eighteenth-early nineteenth century plantations, most of which did not survive the Civil War siege: Jordan, Friend, Hare, and Taylor.  Two dwellings were constructed during the second quarter of the nineteenth century on the smaller Gibbons farm and on the lot owned by Griffith (Figure 6.1).  The historical framework of the Dunn house remains unclear at present.

Archaeological projects conducted by the National Park Service during the 1970s focused upon the plantations lying east of Petersburg that were destroyed in the initial fighting as the Union Army of the Potomac approached the town in June 1864.  Two of these sites were archaeologically investigated.  The Hare house was constructed in the latter half of the eighteenth century and was standing when the land tax records were initiated by the State of Virginia in 1782.  The house and grounds were overrun by Union forces early in the siege, and by July 1864 it would appear that the house had collapsed.  Limited excavations conducted in 1976 exposed portions of the brick foundations within the cellar of the dwelling (Blades and Cotter 1978).  A wealth of mid-nineteenth century material culture was also encountered, suggesting that the Hare house cellar may be a "time capsule" of the possessions of this wealthy planter during the summer of 1864.  Material evidence of activity patterns surrounding the house should remain substantially undisturbed since the site was not reoccupied after the Civil War.

The Taylor house had been constructed by Richard Taylor during the third quarter of the eighteenth century.  By the Civil War the plantation was owned by William Byrd Taylor.  This dwelling may have burned as early as June 18, 1864, since Union Army reports refer to the location as "The Chimneys" shortly afterwards.  Taylor did return to reoccupy his plantation after the war, but he chose to rebuild a smaller dwelling on the surviving foundations of the combination kitchen-laundry-slave quarters.  A dairy farm was maintained on the property during the early twentieth century until the site was acquired by the National Park Service.

The site was studied during a systematic excavation survey in the summer of 1978, a remote sensing survey in the fall of 1979, and, in the summer of 1981, limited testing of a large remote sensing anomaly which proved to be the cellar of the original Taylor dwelling (Bevan 1979, 1980; Bevan et al. 1984; Blades 1993).  The survey yielded considerable information concerning changing land use patterns surrounding the ante bellum dwelling, but evidence of these activities is to a certain extent obscured by the continued late nineteenth and twentieth-century occupation of the property.

The other historic plantations and farms remain unexplored from an archaeological standpoint and, together with the Hare and Taylor sites, represent varying levels of site preservation and research potential.  The landscape of large agricultural slave plantations that arose along the 100-foot terrace above the Appomattox River during the eighteenth century was not representative of the majority of smaller farms throughout Prince George County.  These smaller farms increasingly emerged in the nineteenth century as some of the older and larger plantations were subdivided upon the deaths of owners.  Two smaller farms emerged during the mid-nineteenth century on lands currently within the Main Unit.  The Gibbons farm dwelling, although less valuable than its eighteenth-century neighbors, was still more valuable that the median for the county.  The Griffith home was located on a small lot along the Jerusalem Plank Road leading to Blandford.  The Dunn house may represent a third smaller property, but its historical framework remains obscure at present.

The plantations and farms participated to varying degrees in the slaveholding economy oriented in the nineteenth century to the production of wheat and corn agricultural crops.  The Civil War and particularly the siege of Petersburg brought an abrupt halt to the forced labor system on which this agricultural production was based, and in most cases brought an end to the domestic occupation of these sites.

The date ranges of occupation, apparent later disturbances, and consequent archaeological research potential for the plantation and farm sites that have been identified within the Main Unit are summarized in Table 6.1.

The highest potential is manifested at three locations.  The Hare site reflects a slave plantation from the colonial eighteenth century through the first half of the nineteenth century.  The cellar contents are an indication of the material culture of a wealthy planter and since the house evidently did not burn in the summer of 1864, artifact preservation would most likely be enhanced.  The plantation was not reoccupied and thus the outbuildings and evidence of activity patterns across the site are probably relatively undisturbed.  A strong potential for interpreting African-American slave occupation, particularly in the vicinity of the kitchen/laundry outbuilding, exists.

The Gibbons farm site is especially interesting since it appears to have been occupied for less than a decade (1857-1864) prior to the destruction of the house during the siege.  Some evidence of a later roadway and a twentieth-century boxwood nursery are indicated.  The house site may provide something of a contrast to that of the Hare house.  The importance of spatial activity patterns and African-American occupation evidence on a smaller plantation or farm may contrast with patterns and evidence from older Hare plantation on the opposite side of Harrison Creek.

The quarter or probable overseer/slave quarter area of the Friend plantation on the low terrace below the 100 foot terrace bluff represents yet another perspective on the domestic environment.  The surrounding fields were the scenes of extensive agricultural activity in the nineteenth century and probably earlier, but the low terrace remains an area of importance for prehistoric and historic occupation potential.  The site is the only known area within the Main Unit associated with an overseer and/or slave dwellings that was physically separate from the dwelling of the plantation owner.  The associated activity patterns will therefore be distinct from those of the owner and as a consequence more clearly indicated because of the physical separation.  The historic occupation may date at least to the late eighteenth century since this area was a portion of "White Hill" estate by 1797 and evidently by 1782 when the land tax records commenced.

Three sites are considered to have moderate potential, although excavations at the Taylor site revealed that such "moderate" quality is still quite important, depending upon the questions under consideration.  The Taylor site has been investigated archaeologically, revealing two major temporal components:

It was possible to isolate temporally distinct spatial patterns and the pre-Civil War dwelling cellar appears to have intact stratigraphic deposits, although the structure that stood above it was destroyed by fire.  There is also historical and remote sensing evidence of Union artillery battery earthworks on the property.  Repeated occupations are interesting from a diachronic standpoint but have disrupted evidence from all previous periods, including the Late Archaic lithic scatter on the ridge between Poor Creek and a tributary.

The Griffith site is similar to the Gibbons site in having been occupied for a brief period between 1857 and 1864.  The lot was owned by Lydia Fells, a free African-American, at an earlier period when no building values were recorded in the annual land tax, which suggested that no buildings stood on the lot until 1857.  The later Griffith occupation would make an interesting contrast to the rural farms and plantations, since the lot size was suggestive of the properties in the neighboring town of Blandford.  The proximity of the site to Jerusalem Plank Road (modern Crater Road), the post-Civil War golf course, and park developments all combine to have a least a potential disruptive impact on the archaeological integrity of the property.

Archaeological excavations could certainly address the historical questions relating to basic dates of initial and final occupation of the Dunn house.  The farm site may prove to be a corollary in physical scale and economic scope to the Gibbons farm.  The proximity of World War I encampments from Camp Lee probably resulted in damage to the Dunn site and raises questions about the integrity of the associated archaeological deposits.

The Jordan and Friend plantations and dwellings had similar origins dating to the eighteenth century, but their histories changed during the Civil War siege when the Jordan plantation dwelling was destroyed.  The Friend house, although probably damaged, survived to serve as a Union headquarters and later as a farm residence.  Both sites became foci of NPS development activities during the twentieth century and these developments, coupled with the continued occupation of the Friend dwelling, have reduced but certainly not eradicated the research potential at these sites.

Late nineteenth and twentieth-century occupations are often regarded as "disturbances" of earlier sites and site patterns.  Such occupations within the Main Unit are reflected in evidence of World War I Camp Lee activities and CCC projects from the 1930s, in addition to evidence of early NPS development of the park.  A twentieth-century dairy farm was located on the Taylor house site, and occupation of the Friend site continued after the Civil War.  Private homes dating to the twentieth century along Hickory Hill Road were eventually purchased by the NPS.  It remains uncertain what if any contribution traditional archaeological excavation may make toward the understanding of these activities.  Historical documentation and landscape analyses represent other avenues of research that are probably more potentially fruitful in the short term.

Historic Military Sites and Potential

The Main Unit contains extensive evidence of Union and Confederate earthworks constructed in 1864-65 during the siege of Petersburg, in addition to earlier Confederate works of the Dimmock line which were intended to aid in the defense of the town.  Some excavations have taken place within and near the Crater, the depression created by the explosion of a Union mine filled with black powder beneath Confederate Elliott's Salient in July 1864 (Griffith and Wilson 1962; Wilson 1976).

During the summer of 1978 the National Park Service conducted excavations within the Confederate picket line earthworks that had been constructed in front of the Crater following the mine explosion (Blades 1981).  The excavations, undertaken by Brooke Blades and Mark Ohno, recovered few military artifacts, a reflection of the limited resources available to the Confederate defenders of Petersburg.  A pit containing a series of coal fire hearths was encountered.  Although advanced picket duty was obviously hazardous, exposure to the elements was a more persistent foe.  Confederate General Bushrod Johnson, who commanded the division that occupied the picket line and Crater area, wrote on December 31, 1864, that several men in his command had no shoes (O.R. Series I, Vol. 42, Part 1:925).  Desertions increased within the Confederate divisions in February 1865 and General Lee realized that he must oppose a growing Union Army of the Potomac with diminished numbers of poorly-supplied troops.

Bevan conducted geophysical investigations on the site of Union Fort Morton east of the Taylor house site in the fall of 1979.  He continued to examine this site during the 1980s and early 1990s, utilizing ground-penetrating radar, proton magnetometer, and conductivity meters that confirmed and amplified the initial results (Bevan 1992, 1993, 1996).  He has generated a map of subterranean features that compares most favorably with mid-nineteenth century plans of Fort Morton.  This extensive research demonstrated the utility of non-destructive remote sensing technologies in locating and mapping features such as buried cellars and Civil War fortifications.

Archaeological research on Civil War sites within the Main Unit resumed during the late 1990s.  Dr. David Orr and his NPS staff excavated a portion of the buried trenches from Fort Morton in 1998; they were assisted in this endeavor by students from the Virginia Military Institute.  During the summer of 1999, Gail Brown and Michael Wilkens, graduate students in the Department of Anthropology at the University of Maryland, conducted a limited excavation of a portion of the Union picket line opposite the Confederate lines near the Crater.

The Civil War earthworks and encampment sites are reflected by a large number of underground sites and standing remnants of works in varying states of erosion.  Major categories of earthworks and sites are discussed in the following paragraphs.

Dimmock line: The Confederate fortifications and discontinuous earthworks prepared in 1862 and 1863 to defend the town of Petersburg are reflected in the Main Unit by the remnants or sites of batteries 4-13, which were the first ones to fall to the advancing Union forces on June 15, 1864.  The conditions of these works vary since some were enlarged by Union troops and all have been subject to more than a century of erosion.  These works provide an interesting comparison with the subsequent siege line fortifications (Figure 6.2).

Harrison Creek line: A hastily-prepared and discontinuous line of earthworks was erected by Confederate troops on the night of June 15-16, 1864, west of Harrison Creek.  This line was abandoned two nights later as the Confederates retired to a line on the bluff west of Poor Creek.  A small remnant of the southern section of the earthworks is barely visible above ground; the northern section is no longer visible on the surface.  The limited period of occupancy may render the trenches amenable to investigations seeking the definition of activity patterns related to this brief Confederate occupation.  However, the density of Union troops in the area during the siege and subsequent agricultural activities have probably resulted in artifact deposition that has overprinted the limited ones left by the Confederate defenders.

Confederate siege line: The extensive works of the main Confederate siege line generally west of Poor Creek are composed of artillery fortifications and primary and secondary trenches, although most of the latter are no longer visible within the park.  An extensive mapping project to accurately record earthwork positions within the park geographic information system (GIS) should be a high priority.  The quality of the archaeological preservation within these works may be high to judge from data recovered during the 1978 excavations at the Confederate picket line near the Crater.

Union siege line and secondary works: The main Union siege line consisted of artillery forts and continuous—although at times no longer visible—connecting earthworks.  Various secondary works, isolated artillery lunettes, communication trenches, and apparent encampment areas lie between the main siege line and Harrison Creek.  Few elements of these secondary works are recorded on the Michler map of 1867.  All of these works should be recorded using global positioning system (GPS) technology and the resulted data should be entered in the park GIS system.  However, the challenge posed by tree falls in these wooded areas should be recognized.  The presence of a CCC camp within the park also raises the probability that many of the standing works were "restored" during this period with little or no documentation of such activities.