Life and Death at the Florence Stockade, American Civil War, Prisoner of War Camp, South Carolina

Abstract

Extensive excavations of a guard camp at the site of Florence Stockade, South Carolina, in advance of development revealed extensive traces of Prisoner of War (PoW) activity despite an occupation of only from September 1864 to March 1865. Evidence from features, particularly pits, and from artifacts, can be combined with documentary sources to reveal many aspects of PoW life within this crowded and unsanitary camp. The excavations demonstrate that even such transitory sites can yield much information to illuminate the PoW experience.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

No official provisions existed for exchanging PoWs until July 22, 1862, when the Dix-Hill Cartel was signed. The agreement spelled out the mechanism for PoW exchange whereby those of equal rank could be exchanged on a one-to-one basis, while officers could be exchanged for specific numbers of enlisted men depending on their rank. Prisoners could be paroled under that system until they could be officially exchanged and rejoin their units. The cartel of exchange was ended by the US Secretary of War in May, 1863. The breakdown of the cartel is attributed to different causes by different researchers. Lincoln had been opposed to it because he felt that the signed agreement gave the Confederate government a degree of legitimacy that he was loathe to confer. Ultimately, the policy stated by Jefferson Davis to treat black PoWs as runaway slaves and to charge their commanding officers with leading a slave insurrection was a major cause, as was Abraham Lincoln’s desire to deny the South much needed replacements for their decimated armies (Martinez 2004; Sanders 2005; Speer 1997).

Whatever the reason for the breakdown, both sides were faced with dealing with large numbers of PoWs after May, 1863. According to Martinez (2004:ix), up to 210,000 Union and 220,000 Confederate troops were captured after the collapse of the Dix-Hill Cartel. That total does not count civilians arrested by both sides during the war.

Many types of facilities were used to house PoWs during the Civil War. Existing prisons, forts, camps, and industrial buildings were used, and new camps were constructed. Open air stockades were used by the Confederacy to supplement other types of prisons because they were cheap and easy to build. Camp Sumter, better known as Andersonville Prison (see Chap. 2), was built in early 1864 and was the first and most notorious of the open air stockade prisons (Sanders 2005; Speer 1997). The Florence Stockade was yet another expedient answer to the overwhelming question of how to deal with PoWs during the Civil War.

The Florence Stockade was built to house Union enlisted PoWs who had primarily been shipped east from Andersonville after the fall of Atlanta in early September, 1864. The PoWs were first sent to temporary quarters in Charleston, which were quickly filled beyond capacity and plagued by outbreaks of smallpox, yellow fever, and other diseases. Major Frederick F. Warley was ordered by the commandant at Charleston to construct a prison at Florence, where three railroads crossed. Warley assembled a work force of 1,000 slaves from surrounding plantations and began construction. The first PoWs were shipped to Florence on September 15, 1864, well before the stockade could be finished and before an adequate guard force could be assembled (King 1974:35–36). Ezra Hoyt Ripple was among the first 1,500 PoWs sent to Florence, and he indicated that the first group volunteered to go to Florence from Charleston. The PoWs were initially held in a cornfield surrounded by a small number of guards (Snell 1996:62); mass escapes were attempted by the PoWs while they were held there, but all were eventually recaptured and they were all moved into the unfinished stockade on September 18 (King 1974:36).

The stockade encompassed approximately 23½ acres, with palisades that were 1,400 feet long by 725 feet wide. The palisades were made of heavy, undressed timbers that extended 3–4 feet into the ground and projected approximately 12 feet high above ground. A ditch was dug 5 feet deep and 7 feet wide, with the excavated dirt thrown against the palisade to form a walkway for guards that extended within 3 feet of the top of the palisade. A “deadline” was placed 10–12 feet inside the palisade, and the guards were instructed to kill any PoW who crossed this. The palisade included a stream, Pye Branch, that ran through the prison to provide water for PoWs on the upstream (north) side and to drain sinks (privies) placed on the downstream (south) side. Approximately 6 acres of the camp around Pye Branch was swamp, and the stream separated the main living area of the camp to the east, from the hospital and other facilities to the west. There were 50 guard posts established around the stockade, with 29 picket posts, manned only at night, placed approximately 20 yards outside the stockade to guard against tunneling. Platforms were erected in each corner of the stockade for artillery that could rake the camp in case of an uprising or attempted mass escape (OR II, Vol. VII 1899:1097–1099; Snell 1996:62).

Images of the camp were prepared as watercolor prints by Sneden (Bryan et al. 2001:235–236) based on sketches he made while a PoW at Florence. His overall image of the camp appears to be grossly inaccurate, which places his other images in doubt. What is probably a more accurate map of the stockade was drawn by Sergeant-Major Robert H. Kellogg (1868:318) of the 16th Connecticut Infantry, showing the perimeter of the stockade and its interior layout (Fig. 3.1).

Fig. 3.1

Plan of the Florence Stockade (Kellogg 1868:318)

The camp hospital was located in the northwest corner of the enclosure and consisted of five sheds used to house patients at the time of Kellogg’s (1868:318) map. Andrews (2004:94), based on a visit to the stockade on October 19, 1865, noted that the hospital complex included seven log buildings that each measured 40 by 20 feet. He indicated that the buildings had been partially burned when the stockade was abandoned. Two intersecting streets were present in the section west of Pye Branch, and one of the streets extended across a causeway over the stream and into the eastern part of the camp that housed the PoWs. Kellogg remarked on the similarity of the layout at Florence to that at Andersonville, an observation shared by other PoWs who left journals. No shelter was provided for the PoWs save the hospital (Cook, 1996:62; Goss 2001:217; Kellogg 1868:317–319; Miller 1900:21).

The fortifications built to protect the camp from attack were described as being on “two sides of the stockade” (OR II, Vol. VII 1899:1097–1100). Earthworks are still visible near the east end of the stockade. Those earthworks are slightly north of the northeast corner, and probably represent defensive trenches or extensive rifle pits. What may be faint earthworks are present in the same relative position on the west side of the camp, and run parallel to an existing dirt road which may have been the original road between the stockade and the cemetery to the north.

Guard camps were located on both sides of Pye Branch. Very little historical information has been found about the guard camps, and much less is known about the guards than the PoWs. According to Kellogg (1868:237), the slaves who built the stockade and outer defenses lived in a camp that was “a few rods from the north side of the stockade.” It is likely that the guard camps were also located north of the stockade, given the fact that Pye Branch was probably heavily polluted by the time it flowed through the stockade to the south.

Sydney Andrews (2004:93) toured the South after the war and wrote articles for the Chicago Tribune and the Boston Advertiser. He visited the abandoned Florence prison on October 19, 1865, and noted that the stockade, the PoWs’ huts, outer defenses, and the guard camps were all still standing at that time. He observed that there were earthworks standing “twenty rods” outside the stockade to both the east and west, and that the earthworks were fronted by rifle pits. He placed the main entrance at the northwestern corner of the stockade, and said that there were two guard camps made up of “log houses” near the main entrance and beyond the northeastern corner of the stockade. An additional stockade entrance was located in the southeastern corner of the stockade.

Cemeteries were established north of the stockade for burial of the Union dead. The main cemetery was located along the road from Florence to Georgetown, and is preserved today as part of the Florence National Cemetery. Additional cemeteries, containing from 10 to 400 graves, were placed mainly between the stockade and the northern burial area, though their exact locations are no longer known (Congressional 1868:982–985; Rusling 1866). The total number of dead at Florence cannot be absolutely determined. Power (1991:16) has estimated the death toll among the PoWs to be 2,800, and that number seems reasonable considering the numbers held and mortality rates that are partially known.

The camp was also supported by a number of facilities. It is known there was a sutler present, and there was an early hospital placed outside the stockade. There are references to a headquarters that was probably located to the west. Storage facilities for rations were probably also present. Sadly, no map or images of the overall stockade complex have survived.

A report on the Florence Stockade, submitted to Lieutenant-General Hardee by Lieutenant-Colonel W.D. Pickett on October 12, 1864, provides insights into the status of the prison at that time. He reported that the stockade was approximately a week from completion, but housed 12,362 PoWs. That number included 860 men who were sick in the hospital and 20 men who had been paroled, but excluded the 807 men who had taken an oath to the Confederacy and enlisted in the Confederate Army. The guard force consisted of five battalions of the South Carolina State Reserves with about 1,200 effectives and additional regular troops that brought the total number of guards to about 1,600. The health of the PoWs was noted as “generally poor,” and blankets and clothing were in short supply. The death rate among the PoWs at that time was between 20 and 50 per day. The PoWs lacked shelters and cooking utensils, and their rations were issued raw (OR II VII 1899:972–974).

By late December, 1864, concern of the Confederate officials shifted to the advancing Union forces under the command of Major General William T. Sherman. As early as December 23, 1864, plans were discussed to move the PoWs back to Andersonville, but logistical problems and then events prevented that from being done. The flurry of correspondence concerning the need to remove the PoWs from Florence continued into February, 1865. The Assistant Secretary of War, J. A. Campbell, finally ordered the removal of the PoWs to North Carolina on February 13, 1865, and the PoWs who were able to travel were transported to Goldsborough, North Carolina, via the North Carolina Railroad on February 15, 1865. There were 7,187 PoWs at Florence when the evacuation began. Approximately 700 PoWs were left at Florence as they were too sick to travel. The PoWs were paroled at Goldsborough and sent on to the Union lines at Wilmington. The last of the PoWs did not leave Florence until early March (OR II VII 1899:1262, 1270–1271, 1286, 1302–1304, 1219–1221; OR II VIII 1899:13, 96, 127, 161, 172, 181, 191, 210–213, 218, 224, 225, 234, 238–239, 244, 449–454).

Numerous diaries and letters have survived that chronicle the life and death of Union PoWs at the Florence Stockade (Cook 1996; Elliott 2002; Fosdick 1887; Goss 2001; Hoster n.d.; Kellogg 1868; McElroy 2003; Miller 1900; Moore 1972; Newton 1896; Snell 1996; Stewart 1999). In contrast, only a few useful guard accounts survived from Florence. The best information available from the Confederate guards is contained in a series of articles written by Second Lieutenant Thomas J. Eccles, of Company D of Gill’s (3rd) Battalion of the South Carolina State Reserves for The Yorkville Enquirer under the name “E” (Eccles 1864–1865). He chronicled life at the camp from a decidedly southern viewpoint, but provided at least some information useful for this discussion. Rev. N.J. Holmes (1920) wrote a brief account of his experiences as a young guard at Florence that was published in 1920, and Walter D. Woods (1947), who was among the first Confederate soldiers assigned to Florence, also published an account of the events at the stockade. Information from the guards is supplemented by official Confederate reports, and the few records of the camp that have survived.

The day-to-day operation of the stockade would have been impossible without the cooperation and aid of at least some of the PoWs who filled a number of jobs within the stockade. The 200 members of the Police Club patrolled the inside of the camp, stood guard at posts in the swamp, the main street, the gates, and the hospital. They also regulated commerce within the camp and maintained general order. They were responsible for seeing to the welfare of PoWs who could not care for themselves. Other jobs filled by PoWs included working on the firewood and burial detail, serving as clerks and hospital staff, and even serving as camp musicians. Prisoners received extra rations for their work that, in the case of the Police Club, even included rations of beef that were denied to the PoWs at large. The available records reflect that 500–600 extra rations were dispensed on a daily basis in the camp, which may reflect the size of the prison workforce. The sergeants who were elected to distribute rations to detachments of PoWs organized into 1,000s and then into 100s also received extra rations. The members of the Police Club probably benefited from this arrangement more than the other classes of workers, and were held in the lowest regard by their fellow PoWs (Goss 2001:221–222; Hostern.d.; Kellogg 1868:340; Snell 1996:103–105).

Both PoWs and guards had to build their own shelters at Florence. The PoWs made do with whatever they could scavenge within the camp, including trees and stumps left over from clearing the stockade. A few brought tents with them, but most built their “shebangs” by digging a hole the size and shape of their blankets or tents, using forked sticks to support their blankets or tents as a crude roof (Hoster,n.d.:11–112, 131; Kellogg 1868; Snell 1996:66). Shelter provided to the guards never amounted to more than a few tents, and most apparently fashioned crude cabins built of logs and organized into “villages” to the east and west of the stockade. The roofs of the guard cabins were covered with wooden slabs, blankets, tent halves, or whatever else was available (Eccles 1864–1865). An important difference between the guards and PoWs when building shelters was that the guards had free access to the timber in the forests around the camp, while PoWs had to use whatever they could find in the stockade to build their shelters. The guards were also able to dig wells to secure safe drinking water.

The types of foods available to the PoWs were probably not very different from the basic rations provided to the guards, although the latter probably drew larger amounts. The guards had more meat and in greater variety than the PoWs, and probably had access to greens and other foodstuffs that the PoWs could not gather on their own. Meat was very scarce in the Stockade, and Hoster, as a member of the Police Club for a period of time, had beef for meals over 14 days while beef was rarely available to other PoWs. Goss (2001:240) states that beef was only issued three times to the general population within the Stockade.

It is impossible to reconstruct the quantity of food issued to the individual PoWs at Florence Stockade, as no detailed records concerning rations have survived. Hoster (n.d.:113, 128, 129), who recorded what and how much he ate each day in his diary, recorded that he received 1 quart of beans, ¾ pint of molasses, 2 “sanitary cups” of meal, 1 “sanitary cup” of hominy, and a tablespoon of salt as extra rations from the Police Club on October 15, 1864. After leaving the Police Club he drew 1½ pints of meal plus an unspecified amount of molasses as regular rations on January 18, 1865 and 1½ pint of meal and a gill (a quarter of a pint) of beans on January 25, 1865. It appears that the amount of any one foodstuff varied with the strength of supply, and that mixed rations in small amounts were given to the PoWs at any given time.

The diet of the Confederate guards is less clear. Archaeological excavations of a guard camp at Florence indicated that they had access to much more meat than has been described for the PoWs (Avery and Garrow 2008). The relative diet of the guards can be inferred through discussions of health issues. It is known that the PoWs suffered from a broad range of ailments, and that the death rate was extremely high. Malnutrition and scurvy, which resulted directly from inadequate and insufficient diets, were major killers among the PoWs. Inadequate shelter and a contaminated water supply contributed to a number of fatal illnesses. In contrast, the general health of the guards was relatively good. Eccles (1864–1865) indicated that “fever, measles, and mumps” were the major ailments recorded among the guards, although there were a few deaths from typhoid fever. At least some of the guards lacked adequate clothing and even shoes, but their diet was clearly sufficient to meet their needs and few guards are known to have died at Florence.

The guard force assembled at Florence was a ragtag group of reserves composed of boys too young to serve in the regular army or men who were too old to post to regular units. A report on Company A of the 3rd Battalion of the South Carolina Reserves filed by Captain M.W. Coleman of the 4th Battalion of the South Carolina Reserves on December 31, 1864 stated that the discipline of the company was “good,” their instruction was “fair,” their military appearance was “ordinary,” their arms were “inferior,” their accouterments were “none received,” and their clothing was “private” (Friends of the Florence Stockade 2006:11).

An “invoice of ordnance and ordnance stores” turned over by Captain H.S. Ingraham, Assistant Chief Ordnance Officer, to be sent to Florence provides one of the few insights into military items provided at Florence for use by the guards. The items on the list included one 6 pounder iron Napoleon cannon and a second 6 pounder gun; two 6 pounder carriages and limbers; four rammers and sponges; one gunner’s gimlet and four handspikes; one gunner’s level and four priming wires; four landyards and two tube patches; four thumbstalls and two vent patches; two field worms and two vent covers, 22 spherical case shot, fixed; 112 6 pounder canister, fixed; 300 muskets and bayonets, caliber 0.69; 12,000 musket buck and ball cartridges, 0.69 caliber; 100 friction primers; 40 paper fuses; and 35 packing boxes (Florence Military Records 1864–1865).

It appears that the reserves were not issued military accouterments, and had to supply much of their own gear. They arrived at Florence with inadequate clothing and some even lacked shoes. They were probably armed with a patchwork of obsolete military weapons and private arms. A monthly report prepared by Captain John C. Rutherford on November 5, 1864, indicated that the duty roster consisted of six commissioned and 17 noncommissioned officers, with 336 privates. Those on duty manned the sentry posts around the prison and served on other assignments inside and outside of the prison. The 300 0.69 caliber muskets issued to the camp on September, 1864, were nearly sufficient to arm the sentries that were on duty at any given time. That supply was probably supplemented with additional issues of weapons and private weapons brought by the guards because of the ever-present threat of Union raids.

The traditional southern view of the way PoWs were treated in the Confederate prisons differs radically from that expressed by the former PoWs. There is ample evidence, as reflected in the official records, that at least some Confederate officials worked hard to improve the lives of those incarcerated at Florence, but there is also ample evidence of cruelty and neglect on the part of the guards and Confederate officials. At the same time, the Union policy of suspending PoW exchanges led to a high mortality rate and a great deal of misery for their troops held in southern prisons.

Archaeology in the Camp

In 2005, the US Department of Veterans Affairs planned to expand the Florence National Cemetery in Florence, South Carolina. The proposed 10-acre expansion area, located south of the existing cemetery, included a portion of Site 38FL2, the Florence Stockade. Although the project area was north of the stockade itself, it was believed that part of the support system for the prison was located between it and the cemetery. Phase II archaeological testing (Grunden and Holland 2005) was conducted on the expansion area prior to construction, although ground clearing had already taken place. Testing revealed the presence of numerous Civil War-period features, but did not reveal the function of the area.

During the spring and summer of 2006, MACTEC Engineering and Consulting conducted a Phase III archaeological data recovery on the project area (Avery and Garrow 2008). The research design called for the stripping of the plow zone from the 9-acre portion of the project area within the boundaries of Site 38FL2, the recordation of all identified features and the excavation of 150 of them. The small tract adjoining the southern edge of the existing cemetery that was determined to be outside the site boundary was also monitored while construction was taking place. The types of features present and their locations revealed that the area had been part of a camp of the Confederate guards.

Features

In total, 521 features were recorded, including the 149 recorded during testing. During the main excavations 179 features were investigated, although some of these were determined to be trees or other non-cultural disturbances (Fig. 3.2). The excavated features were assigned to one of ten general categories based on size and shape in both plan and profile. Feature types included structures, trenches, privies, slit trenches, wells, pits, posts, trees, other disturbances, and prehistoric pits. While some categories provided more specific typological or functional information, such as privies or posts, others such as structures, trenches, and pits required further definition and interpretation.

Fig. 3.2

Feature location map

Mapping of the structural features at Florence revealed that the camp was arranged in company streets, but did not apparently strictly follow military rules. The location of the structures strongly influenced the positioning of the other features. Military regulations dictated how a camp was to be arranged, but these rules were often changed based on terrain, the number of men encamped and the duration of the occupation. For the Civil War soldier in camp, shelter took many forms, ranging from a shelter half to a log cabin, depending on the season, the tactical situation and length of deployment. The guards at Florence were far from the front and were there for a relatively long period of time which extended into the winter.

Excavated features revealed that several structural types were constructed. Three curved, shallow trenches probably marked the location of Sibley tents. Measuring approximately 18 feet in diameter, Sibley tents were conical in shape with a round base. They were often placed on top of a short wall of vertical logs that were placed in a trench. Sibley tents could accommodate at least 12 men but they were expensive and heavy, so they were relegated to use by rear echelon troops through most of the war (Nelson 2006; Whitehorne 2006). Each of these trenches had an elongated protrusion perpendicular to the main trench that probably marks an entryway.

More permanent structures included semi-subterranean huts and possibly small log cabins. With sawn lumber scarce, logs were the basic building material for these structures. Semi-subterranean huts consisted of a hole excavated 3–4 feet into the ground with short log walls on the surface. A roof, usually consisting of combined shelter halves or pine boughs, was placed on poles across the top of the wall. The dirt walls and floor were clad in boards if available. Chimneys were typically placed at a gable end of the hut and were constructed with bricks, mud-coated barrels or boxes, or sticks and clay (Nelson 2006; Whitehorne 2006). Feature 223 was a clear example of a semi-subterranean hut (Fig. 3.3). Measuring 10 feet by 10 feet and extending to a depth of 38 cm below the truncated ground surface, this hut was one of the largest structures recorded and was the deepest. Burned wood located on the floor of the structure, including fragments of bark and possible boards, may indicate that the walls or floor were covered as described above. The hearth of this hut was extremely well preserved.

Fig. 3.3

Pre-excavation view of Feature 223

Similar features were excavated but were generally much shallower. These may represent cabins constructed above a shallow excavation or may simply indicate that more of the feature was lost to plowing. One large example, Feature 540, may have been used as a guard house based on its location north of the main residential area and its large size. This structure measured approximately 10 by 15 feet and had a small pit cellar near its center.

The most common features recorded at Florence were pits, which ranged widely in shape, size and profile. Specific functions were determined for some, but the purpose for most of them was unclear. Many were probably excavated specifically for the disposal of refuse, while others may have served as sources of fill or other unknown functions. One pit, Feature 215, appears to have been excavated specifically for the disposal of trash as it was basin-shaped and contained discrete layers of artifact-rich fill. Two others appear to have been used as a source for clay based on their shape and depth. Feature 217 was excavated well into the hard, red clay subsoil and was bell-shaped in profile. Wooden planks were located on the base of the pit, which may have been used as a work surface. Feature 425 lacked the bell shape, but was dug well into the clay. Clay was widely used in the construction of stick or barrel chimneys and was not readily available in this area of the camp. An easy solution would have been to dig through the soft sand to the clay that lay below.

Clues to the function of another pit were provided by a historic photograph. Feature 286 was a long narrow pit with larger, oval ends. The base of one end was baked hard and a concentration of ash and charcoal was noted within the rest of the pit. The feature was presumed to have been used for cooking, but what form was unclear. A historical image of a soldier with the 153rd New York cooking on an oval stove indicates that the feature was likely used for that purpose.

Artifacts

In addition to providing a wealth of information on the material culture of the Confederate soldiers stationed at Florence, it was hoped that the 5,828 artifacts recovered would provide some information on the function of the features excavated. The distribution of the different artifact groups among the features was examined using South’s (1977) functional groups. Although South’s classification was designed for domestic assemblages it was deemed appropriate for this project due to the site’s overall domestic function within a military framework. The average number of artifacts from each group was calculated for each feature type to provide the basis for the analysis of distributions. The locations of certain individual artifact types were also examined in order to assist with the interpretation of feature types and the location of activity areas within the site.

Artifacts of the Activities group were most frequently located in pits. This is probably due to the recovery of a large number of metal fragments as a result of the discard of sheet tin items in refuse pits. The relatively high frequency of these materials recovered from the structures can be explained in the same way.

The Architectural group was more evenly distributed among the various types of features, but the majority of the material was recovered from the wells, with most of the remainder recovered from pits, privies and slit trenches, and the structures. Most of the architectural artifacts consisted of nails and brick fragments that were common across the site. The presence of these materials in the wells and pits is primarily from the dumping of refuse, although the base of at least one well was probably lined with a wooden crate or box held together with nails. Likewise, one of the privies was apparently lined with a wooden crate. The bricks and nails recovered from the houses may represent primary deposits derived from efforts to improve the structures with board walls and brick hearths.

The majority of the Arms group artifacts were recovered from houses and pits. Those recovered from the houses consisted primarily of ammunition components, such as percussion caps and bullets—small items that were easily lost. While ammunition was recovered from pits, canteen parts (Fig. 3.4) and cartridge box parts contributed to the assemblage from the pits, where these items were intentionally dumped.

Fig. 3.4

Canteen recovered from Feature 502

Exactly one half of the Clothing Group artifacts were recovered from structures, although this figure is primarily due to the relatively large number of buttons and button fragments directly associated with the burial in Feature 95. If the buttons from Feature 95 are omitted, the majority of the Clothing Group would have been recovered from pits, followed by privies and slit trenches. It would be expected to find these items in the houses as this would have been where the maintenance of clothing took place, and buttons were certainly easy to lose if dropped. Those found in pits suggest that they were intentionally disposed of, while those in privies and slit trenches may have been lost while unfastening and fastening garments.

All feature types contained Kitchen group artifacts but they were most commonly recovered from pits. Most of these materials were fragmentary glass containers and ceramic vessels that were probably thrown into the pits after they were broken elsewhere. However, a few pits appear to have been directly associated with the preparation of food. The high frequency seen in posthole fill was unexpected and consisted almost exclusively of container glass.

Artifacts from the Personal group were relatively evenly divided between privies and slit trenches, houses, and pits, although they were recovered from all feature types. The majority of the personal materials were located in privies or slit trenches and were probably lost from pockets or disposed of intentionally, while artifacts recovered from the houses were more likely lost. Broken tines from hard rubber combs were the most commonly recovered personal artifact and were primarily located in pits and houses.

Only a small number of tobacco pipe fragments were recovered, with the vast majority located in pits. No intact or complete specimens were recovered, indicating that they were broken elsewhere then disposed of in the pits. The next most frequent location for them was in houses, which more probably represents their location of use and presumably breakage.

Artifact Patterning

The distributions of some specific artifact types were selected in order to determine if patterns were evident that might provide information on the camp. This approach was of little analytical value using artifacts such as nails or container glass that were distributed widely across the site. Therefore, smaller assemblages and those with specific functions were examined.

Although a very small number of window glass sherds were recovered, their location is informative. All the sherds were found in structures, with 12 of the 13 fragments coming from Features 223 and 540 which may indicate that these structures were built with glazed windows. It was common for officers on extended duty, such as winter quarters, to place windows in their cabins or huts, often with window frames scavenged from other buildings (Nelson 2006; Whitehorne 2006). It was much less common among enlisted men. Feature 540 was large in plan but shallow, which may indicate that it was a fairly substantial cabin or guardhouse built primarily above the ground surface. As described above, Feature 223 was the largest of the subterranean huts and apparently had board walls. These two features apparently represent substantial structures and may very well have included glazed windows.

Only two calibers of bullets were recovered besides small buckshot. The 0.54 caliber bullets were probably used in either Mississippi/Palmetto rifles or Lorenz rifles. The 0.69 caliber balls could have been fired by a wide variety of older weapons, but were probably used in Model 1842 muskets (Coates and Thomas 1990). This disparity in weaponry may indicate the presence of different units of infantry, although a single reserve unit might be issued a mixture of weapons based on availability, especially late in the war. The distribution of the different calibers further suggests that at least two different units were present in the portion of the camp investigated (Fig. 3.5). The 0.54 caliber ammunition was concentrated in the northern portion of the site, specifically in and around Block A. This type was recovered only from structures and the large pits encountered in this block. One 0.54 caliber minie ball was recovered from Feature 540 within the northern perimeter of the site as well.

Fig. 3.5

Distribution of small arms ammunition. Light shading: features with 0.54 caliber ammunition; Dark shading: features with 0.69 caliber ammunition

The 0.69 caliber ammunition was much more widely dispersed across the site than the 0.54 caliber bullets. Ammunition in this caliber was recovered from as far south as Feature 109 and as far north as Feature 212. It was also recovered from a wider variety of features including structures, pits, and a privy. This may indicate that this caliber was more commonly in use, at least in this portion of the camp. It is interesting to note that no features contained both 0.54 and 0.69 caliber ammunition, but two 0.69 caliber balls were recovered from Feature 212, which is located in Block A immediately adjacent to two features that contained 0.54 caliber bullets. The presence of differing ammunition calibers within the same block may indicate that different units occupied this area over the time the camp was occupied.

The distribution of Kitchen group artifacts was more difficult to analyze because of their sheer numbers. For example, no attempt was made to pattern the container glass as it was recovered from every type of feature in every area of the site. Ceramics, however, provided a better opportunity for analysis. The distributions of refined wares and utilitarian stoneware were plotted separately. Refined ceramics were widespread, but were concentrated in the northern area of the site, including Block A. All of the wells contained small amounts of refined ware, including a blue transfer printed plate, while ten sherds from another blue transfer ware plate were recovered from Feature 239, a slit trench associated with a possible Sibley tent in Block E. The concentration of these materials in the structures and associated pits in the Block A area may mark a different status between the soldiers who lived on this block as opposed to the others. What this difference might be is unclear; these soldiers possibly held a higher rank than those to the south. Conversely, it could mean that these soldiers were relatively new recruits that arrived from home carrying their private dinnerware.

Stoneware was much more common than refined wares and was more widely distributed. However, it was concentrated in the northern portion of the camp that includes Block A. It was recovered from Features 223 and 95, but only as single sherds. Two of the wells, Features 518 and 502, both produced stoneware and tended to yield larger sherds and more complete vessels, such as the nearly complete jug and jar recovered from Feature 502. The largest number of sherds was recovered from Feature 376, a pit in Block B that produced 162 sherds from a single vessel. Likewise, the 22 sherds recovered from Feature 425 represented two vessels that were deposited in two discrete areas of the feature. What the distribution of stoneware implies is unclear, although it probably simply indicates that more of this material was in use on the northern end of the site than elsewhere.

Kitchenware, those items used to store, prepare, and consume food, was widely scattered from as far south as Feature 485 and north to Feature 217. Utensils were recovered from Features 212, 215, 223, 239, and 248. Most of these were forks or spoons, but a folding corkscrew was located in Feature 223. Two fragments of a kettle or dutch oven were recovered from Feature 217 while a portion of an iron spider skillet was found in Feature 518. The remaining kitchenware consisted of a nearly complete tin can and fragments of another. The actual number of tin cans and other food containers should probably be much higher as a very large number of tin fragments were recovered that could not be identified as to form or function.

Conclusions

The spatial information gathered during the archaeological work in the guard’s camp at the Florence Stockade has provided some insight as to its layout and who might have been living in that area. This camp was located just west of Pye Branch, while another was located to the east of the creek. Defensive earthworks were located to the west and north of the camp, as it was anticipated that any Federal attack would come from those directions. Army regulations of the day called for a camp to face the direction of a perceived threat, with the enlisted men to the front and the officers to the rear (United States War Department 1861). As no indication of habitations were located east of the wells during the archaeological research, it is possible that the portion of the camp excavated was inhabited by officers, assuming that some effort was made to lay the camp out according to regulations.

The distribution of the artifact groups among the various feature types provided valuable information for defining general activity areas across the camp and produced some intriguing evidence as to which units were living within the project area. However, it was less successful in determining feature function. The main difficulty was the lack of obvious primary depositional deposits within the features. While some primary deposits were certainly encountered, they were often impossible to separate from the secondary refuse dumped into the feature later. This is not a major concern with the more obvious features such as the structures, privies, or wells, where the morphology of the feature itself is usually sufficient to determine its function. But in features such as the pits where the main function of the feature generally cannot be determined by shape alone, the lack of primary deposits makes any interpretation much more problematic. Many of the pits, for example, may well have been intended as receptacles for secondary debris but this cannot be assumed. In other cases, the artifacts can be misleading, as with several post holes that were encountered that contained complete liquor bottles and animal bone.

The excavations conducted in the campground of the Confederate guards at the Florence Stockade provided a unique opportunity to examine the day-to-day life of rear echelon soldiers during the latter days of the Civil War. The short period of occupation and the relatively small number of men who inhabited the campground have provided a discrete sample of documentary, spatial and material data that is being analyzed in great detail. While much has been written by and about the Union PoWs who suffered and died within the prison walls, the Confederate guards have remained conspicuously silent through history. Excavations within their camp have shed some light on them and the conditions under which they served, and this research can serve as a starting point for more consideration of PoW guards in the future.