The restoration of the Fish Vault in Memory Hill Cemetery

Submitted by Stephen Hammack, Susan J. Harrington, Matthew Williamson, and Hugh T. Harrington (

The Fish Vault has been famous in Memory Hill Cemetery, Milledgeville’s city cemetery, as the place where poor Mr. Fish, despondent over the loss of his wife, had shut himself into the vault and killed himself while sitting in a rocking chair. Visitors to the vault are routinely told to knock at the door and ask “Mr. Fish! What are you doing in there?” With the person encouraged to place an ear to the crack in the wall in order to hear Mr. Fish’s answer, the response, of course, is “nothing.”

The vault is a partly subterranean, 1840s tomb made of handmade bricks. This style of structure is rare in central Georgia. It is large, approximately 14′ 3″ by 12′, and has a gable-style brick roof. Through the years it had suffered vandalism, which had caused the city to brick up the door in the 1960s. Secondhand reports from that time suggested that there were no visible remains. Now the deteriorating condition of the roof suggested that it was probably leaking. The front wall had been bumped by passing vehicles and had settled to the point where it was separating from the rest of the vault. Without intervention, the front wall would have eventually fallen, exposing the interior of the vault.

The Friends of Baldwin County Cemeteries, Inc., a 501(c)(3) nonprofit public charity and historical society, decided to undertake the vault’s restoration. Susan Harrington, chairperson of the Friends, acted as project coordinator. Since there was still the possibility that there may be remains and opening the vault may constitute “disturbing a grave,” the Friends consulted with the superior court judge, probate court judge and the city attorney. It was determined that the only Georgia law that seemed to fit was the Abandoned Cemeteries and Burial Grounds Act, Official Code of Georgia 36-72. Under that law, descendants were to be determined and contacted for permission, a permit from the city was required, and an archaeologist should be present for the vault’s opening. Stephen Hammack agreed to serve as archaeologist.

After genealogy was performed, descendants contacted, and permissions given, the city issued the permit. This process took approximately 8 months. In the meantime, a leading brick mason was contracted to do the brick masonry. The work began September 8, 2008, one day after the permit was received by the Friends. Security was a concern, and the city immediately erected a large fence around the vault as the brick mason began his work.

The Friends notified descendants in Moultrie, Macon, and Atlanta, who came to Milledgeville to watch as the vault was opened. The vault was opened on the hot and rainy afternoon of September 9, 2008, and since no human remains were expected to have survived, it was assumed that the archaeologist’s responsibility was to be on hand to ceremonially give the go-ahead so the brick mason could begin his repairs.


View of Burial 3 within the vault.

But the vault was not empty! Upon peering through the just-opened door, it was clear that there were multiple sets of human remains, all but one of which were visible from the steps. Taking two steps down, Hammack found himself surrounded on 3 sides of the vault by stone benches. Remains and casket material covered the floor and benches to his left and the wall opposite. On his right was another burial enclosed in a cast iron coffin immediately recognized as a Fisk Metallic Burial Case.

It was quickly decided that all the remains had to be removed before nightfall, to prevent any potential looting or vandalism, and the back-up plan was put into effect. The plan was that any remains would be placed in the care of a local funeral home for safe-keeping until re-interment. Because there were three obvious burials in the crypt, though one contained more than one set of remains, the decision was made to extensively photograph everything in situ and to number the sets of remains Burial 1, Burial 2, and Burial 3.

Burial 1 was on the left (south) wall and was that of an adult. The most obvious thing about this burial was that in addition to the remnants of coffin wood, there was an enormous amount of charcoal intermixed with the bones. Burial 2 was on the back (west) wall, was also an adult, and appeared to have contained lesser amounts of charcoal. However, it appeared that some of the bones had been disturbed at some point in the past and replaced in piles. These piles, however, also included the skulls and some bones of what appeared to be one or two children. Burial 3 was enclosed in the Fisk burial case and was not visible for inspection, though it was clear that the metal coffin itself had once been inside a wooden coffin.

The “methodology” for removing the remains was as follows. Since daylight was limited, the only logical choice of action was simply to remove the bones and large pieces of coffin wood and charcoal and place them in individual body bags. Burial 1 and Burial 2 were each placed in separate body bags and placed in the back of the funeral home van. Burial 3, which was noted to have a corroded layer near its bottom, was slowly and cautiously removed from the vault and placed on thick wooden boards, which were to provide support in case the iron coffin’s bottom gave way. The coffin and boards were then placed on dollies and rolled to the van and, with much effort, the coffin was placed inside. Several breaks were required over the course of the afternoon because of the rising humidity and temperature, which rendered the situation in the vault nearly unbearable for the archaeologist. Finally, just before dusk and with the assistance of a flashlight, the work was completed. The following day, the Friends contacted Dr. Matthew Williamson, an anthropologist at Georgia Southern University, who came to acquire the bones for identification in his lab.

A few days later at the funeral home, with the family’s permission, it was decided to remove the iron plate covering the glass face plate, or viewing port, of the Fisk Metallic Burial Case in order to assess the condition of the remains inside. The Burial Case was patented by Almond Fisk of New York City in 1848 and was designed to be air tight and filled with gas to prevent the decomposition of the body. The upper and lower halves fit tightly together and are sealed with a glue-like cement and fastened by screws. The Burial Case would have been ideal for Sarah Fish, who died in 1856 in Gordon Springs, Whitfield County, located in the northwest corner of Georgia, to be transported back to central Georgia for burial.

The body had not been preserved by the supposedly air tight coffin, and the skeletal remains were found to be in an advanced state of decay, evidently due to penetration into the coffin of water from the damp vault. It is interesting to note that a fair amount of fabric from what appeared to have been a striped dress was still intact and visible around the waist.

After washing and analyzing the artifacts, Hammack noted a variety of different sizes of coffin nails and other types of extant artifacts, such as wood and Prosser (China) buttons. Some curious findings were three rocks of granitic or gneissic composition that were found with Burial 1. It is not known if they were placed on top of the coffin or included inside with the body, but they were the only lithic artifacts discovered within the vault and may have been memorials placed purposefully with Burial 1. The presence of large amounts of charcoal with Burial 1, and the smaller amount included with Burial 2, were also noted. Research eventually pointed to the usage of charcoal as an odor-deterrent when bodies were shipped long distances, and this method was evidently also utilized for the same reason in vaults, especially where future burials were planned. It appears that in such burials the coffins were lined with charcoal, sometimes even to the point of surrounding the corpse with it. Other instances are recorded of charcoal actually being placed into the abdomen of the corpse.

Williamson determined that two adult males and three children (labeled Burials 2a, 2b, and 2c) of predominant European ancestry are represented by the Burial 1 and 2 remains. Dark brown staining was present on all the bones, which was consistent with other historic burials perhaps caused by tannins that have leached out of the coffin wood and were then absorbed by the bones. In addition, a small amount of dried adipocere was present at various locations on the skeleton, indicating that the body may have come into contact with water relatively soon after burial. Adult dental health was fairly poor based on the presence of several carious lesions and antemortem tooth loss while the juvenile teeth were in pretty good condition. In general, there were no lesions from infectious disease, metabolic disorders such as chronic anemia, nor any significant osteoarthritis present on the adults. According to local legend, Mr. Fish shot himself in the head while sitting in a rocking chair. From his analysis, Williamson found that there was no skeletal evidence of a gunshot wound.

Unfortunately, positive identification of people from historic cemetery contexts is generally impossible unless they are buried in separate grave shafts with an associated nameplate or grave marker. With this in mind, the skeletal remains have been tentatively identified as those of William Fish who died in 1843 (Burial 1), his youngest son Horace Virgil Fish who died in 1845 at the age of about 5 (Burial 2b), two infants whose names are unknown, and a 45–55 year old male whose identity is unknown. Sarah Harvard Fish, the wife of William Fish, died in 1856 and was buried in the Fisk metallic burial case (Burial 3).

Because of the concerns for security and the need to close the vault in conformance to the brick mason’s availability, all remains had to be returned to the vault and sealed in on October 1. This date provided Hammack and Williamson less than two and a half weeks to perform their analyses.


The Fish Vault after restoration.

The archaeological, anthropological, and historical preservation experience at the Fish Family Vault was, to say the least, a memorable one. It proved to be very interesting from beginning to end, and the best part was that after the restoration was complete and the contents of the vault subjected to study, all remains, casket materials and other items were reinterred with dignity in the vault where they rightfully belonged. Despite the tight timetable, this is a great example of how a project of this kind, with cooperation of many experts, is supposed to work. Preservation-minded Georgians everywhere should be proud of this endeavor!