USING SCIENCE TO UNCOVER HISTORY:
A Geophysical Investigation of Wyatt Chapel Cemetery
Alison T. Henning, Ginger Burns, Richard Hoffman and Brian Jacoby
AbstractA Geophysical Investigation of Wyatt Chapel Cemetery
Alison T. Henning, Ginger Burns, Richard Hoffman and Brian Jacoby
Wyatt Chapel Cemetery is an abandoned cemetery located in Prairie View, Texas. Oral histories from local residents suggest that the cemetery originated as a slave burial ground in the mid-nineteenth century. The local community is interested in examining the cemetery in order to document the history of the area. In July 2007 and February 2008, participants in a graduate course at Rice University acquired global positioning system and ground-penetrating radar data at the cemetery. The soil at the field site is ideally suited for radar work and the subsurface image quality was excellent. Numerous anomalies were identified that are consistent with unmarked burials. Two of these anomalies were excavated and confirmed as burials. The stratigraphy consists of 3-6 feet of sand overlying a hard clay, and the boundary produces a very bright reflector. In the main clearing of the cemetery site, the sand-clay boundary deepens abruptly from 3 feet to approximately 5 feet. This anomaly was initially considered a possible man-made excavation, perhaps a mass burial site. While the stratigraphy does contain abrupt terminations, most depth changes occur gradually, suggesting formation by natural processes.
Background
Slave Cemeteries
Texas is not as often associated with plantation life as are other states in the deep south, such as Virginia and Georgia. But plantations did indeed flourish in Texas in the 1840s-1860s. Cotton was the main crop, made possible by slave labor. By 1860, there were over 158,000 slaves in Texas, comprising 31% of the Texas population. The Brazos River was surrounded by fertile agricultural land and many wealthy plantations were situated along its banks. Many of these plantations contained designated slave graveyards where slaves were buried before 1865. After emancipation, slaves were often buried in newly established church graveyards. Many of the early slave cemeteries have been lost and it is estimated that there are hundreds of abandoned and neglected cemeteries in the state of Texas dating from the plantation era. These cemeteries can contain valuable information on the cultural history of Texas, but most plantations are now gone and there are rarely any written records associated with these abandoned cemeteries.
Particularly striking documentation of slave burial practices come from oral histories of former slaves in Texas.
They just buried the dead most anywhere where there was a good place - under a tree, on a hill, and like that.These statements indicate that there were not always designated burial areas for slaves on plantations. Landscape features such as trees and hills may have served as the only markers for slave burials.
Sometimes they was buried in wooden boxes, and sometimes both white and black was buried just wrapped up. They wasn’t no places to make things [caskets] like now and no graveyards.
I remember when folks died, they built plank coffins and lined the inside with plain white cloth and the outside with black. They didn’t have funerals in them days - they just buried the folks in a burying ground we had.Some former slaves recall specific areas on the plantations set aside for slave burial grounds. Interestingly, these recollections also suggest the use of simple wooden caskets for the burials.
When a slave died, they made a box there on the place, and the folks go to the burying.
Grandma dies…and they puts her in the slave graveyard.
We had all our funerals at the graveyard. Everybody, chillun and all picked up a clod of dirt and throwed in on top the coffin to help fill the grave.
Liendo and Alta Vista Plantations
In 1822, Jared Ellison Groce of Georgia came to Texas with Stephen F. Austin as one of the “Original 300” settlers. Groce brought approximately 90 slaves with him to Texas and built the profitable Bernardo Plantation in what is now Waller County. After his death, his son Leonard W. Groce purchased “that part of the Liendo survey which lay north and west of Pond Creek”. In 1853, he built the Liendo Mansion on a small hill above Pond Creek, which flows into the Brazos River. Liendo was one of the wealthiest plantations in the state, the social center of the state, and employed over 300 slaves. Liendo is named for the original owner of the land grant, Justo Liendo.
After the Civil War, Liendo fell on hard times; without the slave labor, it was impossible to maintain such a large estate. Groce went bankrupt in 1868 and died penniless, like many southern planters during Reconstruction.
In 1848 or 1849, Groce’s cousin Jared Ellison Kirby moved to Texas from Georgia. By 1860, he owned 139 slaves and 8,000 acres on both banks of the Brazos River, including Alta Vista Plantation. The plantation house at Alta Vista was built between 1858 and 1861 and by this time the plantation employed 400 slaves. Years later, Kirby’s widow sold Alta Vista to the State of Texas and Prairie View A&M University was eventually established on the site.
Liendo and Alta Vista Plantations were home to hundreds of slaves between 1822 and 1865. However, there are no written records of a slave burial ground for either plantation. Oral history and a few old headstones suggest that Wyatt Chapel Cemetery, along the northern boundary of the Prairie View A&M University campus on what was formerly the Alta Vista Plantation, served as the slave burial ground for both plantations.
Wyatt Chapel Cemetery
Wyatt Chapel Cemetery is located along the northern boundary of the campus of Prairie View A&M University. The site is heavily vegetated with small brush, trees, and grass. Pond Creek flows around the cemetery to the west and north. Efforts to clear the area over the years have prevented the growth of large thick trees, but the dense low-lying brush and small trees have covered the area.
Students at Prairie View in the 1920s and 1930s reported the existence of a cemetery at the site, with fifty or more grave stones and metal markers . Oral history indicates that the cemetery was the burial place of slaves from the Alta Vista and Liendo plantations. Over the years, Prairie View A&M students and staff have continued to discover headstones in the dense brush. They also discovered a rusted metal fence that appears to enclose several burials.
In 1992, a historical marker was erected to recognize the significance of the site as the Wyatt Chapel Cemetery. The marker identifies the site as Wyatt Chapel Community Cemetery and states that
This cemetery is located on land that was originally part of Jared E. Kirby’s Alta Vista Plantation. According to oral tradition, the Kirby family set aside this land as a burial site for their slaves, as well as slaves from nearby Liendo Plantation, owned by Kirby’s cousin, Leonard Waller Groce.The historical marker also states that “the cemetery became associated with and named for Wyatt Chapel, a nearby African American church”. Wyatt Chapel Baptist Church was established on April 29, 1894, with Reverend George W. Wyatt serving as the first pastor. George Wyatt was born in Waller County in 1848, and family history suggests he was a state legislator from 1868 to 1900, while the Texas State Library Archives online suggest he served from 1883 to 1884. Little else is known about George Wyatt, including the date of his death. The fenced enclosure is thought to contain his grave, but there is no marker bearing his name.
One of George Wyatt’s daughters was Mattie Wyatt Wells. Prairie View resident Mrs. Ida Lou Wells Owens Pierce visited the cemetery in 1989 at age 82 and indicated that the grave of Mrs. Mattie Wyatt Wells within the wire enclosure was that of her mother. She also noted that a headstone bearing the name Caroline refers to her grandmother, a former slave who died in 1898. Caroline must have been her paternal grandmother, since her mother’s parents were George and Lizzie Wyatt. Her father was William Wells, but there is no additional information available about his family. She indicated that many slaves or former slaves were buried at the site.
Some confusion arises from the text of the historical marker that states “The oldest marked grave is that of Mattie (Wyatt) Wells (d. 1882), the daughter of a former slave.” If Mrs. Pierce was born in 1907, then her mother could not have died in 1882. Perhaps the marker should read “b. 1882”. To add to the confusion, there is also a marker for Mrs. Mamie Wyatt, with a death date of 3-17-62 (Figure 1). This was previously recorded as Mrs. Mattie Wyatt, date 8-17-82 (interpreted as 1882).
Most excavations of African American cemeteries have been of 19th century burials, mainly postemancipation. Wyatt Chapel Cemetery straddles the emancipation boundary, after which former slaves would have been freer to bury their dead how they wanted. If some of those buried at the cemetery were born into slavery, then the cemetery would represent an interesting chapter in post-emancipation history. There do not appear to be any burials from the pre-emancipation era.
Wyatt Chapel Cemetery is one of numerous abandoned African American cemeteries in Texas dating from just after the Civil War. For example, Evergreen Negro Cemetery in Houston contains burials from 1887-1950, approximately the same time range as Wyatt Chapel Cemetery (except for one burial in 1984, thought to be part of a family plot). Olivewood Cemetery, also in Houston, contains burials from 1867 (or earlier, some inscriptions are partially obscured) to 1961. Efforts are underway to restore both of these cemeteries, and initial efforts focus on mapping existing surface features. The clay-rich soil in the Houston area makes geophysical work difficult, but GPR did provide evidence of unmarked burials at Evergreen as well as the location of an old road through the middle of the cemetery.
Field Work
The Texas Historical Commission (THC) has outlined a preservation plan for historic cemeteries. One of the first recommended steps is to create a map of the cemetery grounds, including the location of landscape features and gravestones. The THC further recommends producing a description of gravestone material (concrete, marble, etc.), the name of deceased, vital dates, and the exact inscription on the gravestone. The THC also suggests recording vegetation and landscape features.
We utilized global positioning system (GPS) technology to record locations of key surface features, such as headstones, vegetation, and the creek. We used hand-held Garmin devices with average resolution in the heavily treed site of +/- 12 feet. These data were then loaded into a geographic information system (GIS) for interpretation. We used ArcMap 9.0 software and obtained georeference aerial photographs of the field site from 1930, 1956, and 2006. We also acquired ground-penetrating radar (GPR) data in order to obtain images of the subsurface. The
locations of the geophysical profiles were also recorded with the GPS units. Geophysical methods such as GPR are excellent tools for this type of project because they are non-invasive and non-destructive. In Texas in particular, GPR has been shown to be an effective tool for locating unmarked graves. We utilized a cart-mounted GSSI system with a 500 MHz antenna.
Geologic Setting
Prairie View is located in Waller County in southeastern Texas (Figure 2). Wyatt Chapel Cemetery is located north of the town on the eastern bank of Ponds Creek, approximately 7 miles east of the current Brazos River channel. The site is located on the Pleistocene Willis Formation, consisting of 1-3 million year old fluvial sands. Stratigraphically, the Willis Formation overlies the Lissie Formation of slightly older fluvial deposits, which in turn overlies the Tertiary (~20 million years) Fleming clay.
Several trenches were dug around the cemetery site in order to determine the stratigraphy. A layer of reddishbrown sand was found to overlie a hard gray clay. The sand ranges from 3 to 6 feet in depth across the area, with the depth generally changing gradually. We interpret the surface to consist of Plesitocene Willis sand overlying Tertiary Fleming clay.
Geophysical Survey
The use of ground-penetrating radar in archeology is well-established. GPR is a non-destructive method for studying the subsurface that can provide information on buried features without digging or trenching. GPR works by sending a tiny pulse of energy into a material and recording the strength and the time required for the return of any reflected signal. Excavation of cemeteries to answer archeological questions can be considered desecration. GPR provides a non-intrusive method for examination.
Ground penetrating radar is useful for locating graves by detecting anomalies in the soil where graves were dug or for the detecting of buried coffins or vaults. If a vault or coffin does not exist, GPR can be used to study disturbed soil or other remains of the burial. Remains of burials are easier to locate in sandy soils that do not contain tree roots or stones. GPR is operated above the ground surface, and produces a cross-sectional image of the ground.
GPR can produce ambiguous results and is often used in conjunction with other methods, such as magnetic surveys. Soil conditions and choice of antenna frequency are often responsible for this ambiguity. High clay content and water content can suppress the return radar signal. Unknown depth and dimensions of targets can make antenna choice difficult. However, given the right soil conditions and antennae, GPR can provide an accurate and detailed picture of the subsurface.
In July 2007, a group from Rice University’s Department of Earth Science conducted a preliminary GPR survey of Wyatt Chapel Cemetery. A GSSI cart-mounted 400 MHz system antenna was used for this survey. The dielectric constant was determined to be 17 based upon depth to a known target. Records were 80 ns in length, which corresponds to a depth of about 105 inches (8.75 feet).
With support from Prairie View A&M’s Department of Architecture, the group was able to acquire and interpret 59 GPR profiles and 100 global positioning system (GPS) positions in a period of two weeks (Figure 3). Numerous geophysical anomalies were identified and two of them were excavated with the help of an archeology team from Texas A&M University. One excavation resulted in the discovery of a metal plank thought to be a burial cover, and the second excavation revealed a grave shaft. The first anomaly that was excavated was offset about 5 feet from the headstone of Milo Wilson, Jr., which appeared to be in situ. We suggest that there may have been a delay between the burial and the arrival of the grave marker, which led to the marker being placed slightly off the actual burial location. The second anomaly that was excavated was located about 20 feet southeast of the first and was not associated with any surface marker. Most of the anomalies were located in the main clearing of the cemetery.
Survey Layout
In February 2008, additional GPR data were acquired at Wyatt Chapel Cemetery, along with magnetometer data. The main focus of this second data collection was to investigate a stratigraphic anomaly first identified on the 2007 data (Figure 4). In particular, we wanted to establish the boundaries of this anomaly and try to identify its cause as either natural or anthropogenic. Prairie View A&M students and the city of Prairie View had worked to clear the brush from the main clearing at the cemetery and removed much of the trash from the area in an effort to help facilitate the gathering of data by the GPR team.
The area of interest was identified based on the 2007 data. The main clearing that contains the excavated anomalies also shows a relatively deep stratigraphic anomaly. The real-time data display feature of the GPR system was used to locate the anomaly from the 2007 data and determine the best new site for data acquisition.
A 2-D grid of GPR profiles was acquired in the main clearing (Figure 5). A rectangular plot 24 feet by 60 feet was staked off and squared, and the perimeter was defined with string. The coordinates of each corner of the plot was obtained with a Garmin GPSMAP 60 handheld GPS unit, with accuracy of about 13 feet. The area was cleared of overhanging brush and limbs so that the GPR cart could be pulled smoothly across the site without losing contact with the ground. Stakes were placed along the perimeter line at increments of 2 feet and a string line was placed between opposing stakes to ensure that each run was straight. The equipment was pushed along the string line from the same starting line in each direction so that the lines were consistently North to South and East to West. A total of 44 lines were acquired in the main clearing, with 13 North-South lines (Lines 1-13) perpendicular to 31 East-West lines (Lines 14-44).
All the data was acquired in one day, so the field conditions were constant. The data were processed using GSSI’s RADAN software. Processing included adjusting time zero to the ground surface, converting the time sections to depth using a dielectric of 17, and applying a gain function. Data were interpreted in ArcGIS and Google Viewer.
Results
Aerial Photographs
We examined aerial photographs of the cemetery site from 1930, 1956 and 2006. The photos were loaded into ArcGIS and georeferenced. The 1930 photo clearly shows very little vegetation at the site, which is now very overgrown. It also shows Pond Creek very clearly, including a tightly curving meander loop located just northwest of the main clearing. It appears that the outer loop is now abandoned and the creek has moved south and east, toward the main clearing, based upon GPS positions taken along the present day creek. We note that this meanderloop would have been an easily recognizable landmark in 1930 and before, marking an area surrounded by water on three sides. The lack of vegetation would also have enabled easy access to the site. It appears that there were several clusters of vegetation north of the main clearing, which would have served as additional landmarks.
By 1956, there is more vegetation visible on the aerial photograph and the exact path of the creek is obscured, although the main clearing remains more sparsely vegetated than the surrounding area. By 2006, the entire area is densely vegetated and the ground is not visible on the aerial photo, with the exception of a small sinuous path that appears to coincide with the cluster of headstones. Some clearing had been done at this point by the city and university, so this may be an artifact of that clearing rather than representative of past use.
Headstones
Prairie View personnel had noted the existence of multiple headstones and efforts were made by students and staff to catalog these headstones. We used these data as a starting point for our reconnaissance. We located all the previously noted headstones, as well as some additional markers. Very few of the headstones appeared to be in situ, but their locations were recorded as found. All headstone locations were loaded into GIS, along with additional information such as inscriptions, in order to examine these data in the context of the aerial photos. A total of 33
headstones were catalogued, covering an area of approximately 0.3 acres. This is significantly less than the 5 acres thought to be included in the cemetery. If the area is extended southward toward the road, the area is about 4.6 acres.
Six of the markers are concrete headstones, two markers consist of concrete footstones only without associated headstones, three are US Army issue white stone, and seven consist of small rectangular metal plaques that are similar to modern day temporary grave markers. The remaining 15 markers recorded consist of stakes placed by earlier researchers to mark a burial. The three Army markers represent two veterans of World War I and one veteran of World War II.
Many of the headstones displayed illegible inscriptions, but several showed clear dates. The legible death dates ranged from 1922 (Albert Colling) to 1953 (Milo Wilson, Jr.). The temporary marker for Mrs. Mamie Wyatt is labeled 3/17/62, which we interpret to be 1962. However, previous records by Prairie View personnel interpreted the label to be 8/17/82, with a death date of 1882. The historical marker at the site also states that she died in 1882 and represents the earliest known burial at the cemetery. We were unable to corroborate this information. One marker reads “In memory of Caroline” and possibly “Died June 24, 1898”. The first two digits of the year are difficult to read, but 1898 is supported by Mrs. Pierce’s statements. Additionally, a previously recorded marker for Darrie Williams reads “Died Feb., 28, 1898”, but we were not able to locate this marker.
Geophysical Data
The GPR data were of very high quality and allowed a detailed interpretation of subsurface horizons. The depth of penetration was between 6 and 8 feet. The boundary between the sand and clay is evident as a bright reflector at 4-6 feet below the surface. Its appearance varies significantly on the GPR profiles acquired. GPR data acquired over anomalies first identified in 2007 showed the same anomalies, providing evidence of repeatability. Despite heavy rain the day before the field work was conducted, the surface was not wet and the quality of the GPR data was excellent, suggesting that the soil in this area drains efficiently.
Sharp breaks in the sand-clay boundary reflector are observed on some of the GPR profiles, but the majority of the data indicate a continuous horizon (Figure 6). Three-dimensional displays indicate excellent correlation between in-lines and cross-lines (Figure 7).
Discussion
The headstones in Wyatt Chapel Cemetery provided the first clue that the area was a burial ground. Many headstones are broken and not in situ. Inscriptions on headstones suggest the earliest burial to be either 1898 or 1922. As mentioned earlier, students at Prairie View in the 1920s and 1930s reported the existence of a cemetery at the site. If the earliest burial was in fact 1922, then the students would most likely have been aware of a new cemetery and not reported their “discovery” of an old cemetery. This suggests that the cemetery did in fact originate many years before. The partially legible death date of 1898 on Caroline’s marker is corroborated by her granddaughter
Mrs. Pierce.
Whether burials began in 1898 or 1922, the site is a likely place for a slave burial ground. The soil is very sandy and therefore would not have been suitable for agricultural purposes, which would have made the site a good candidate for a slave burial ground. It is bounded to the north and west by Pond Creek, making it easy to locate. And there are several large trees at the northern edge of the site, along the boundary of Pond Creek and near the fenced enclosure, that may also have served as location markers.
The geophysical data indicate that Wyatt Chapel Cemetery contains numerous unmarked graves. Excavations confirmed two geophysical anomalies to be burials. We also used geophysical data to investigate a subsurface anomaly to determine whether it had any cultural significance. We traced the extent of the sand-clay boundary where it suddenly deepens on the GPR data in an attempt to determine whether this deepening of the horizon was the result of natural processes or human activity. The southern limit of the deep horizon was identified, but dense vegetationprevented identifying the edges of the horizon in the other directions. The northern extent of the deeper horizon appears to coincide with the large trees that bound the main clearing at the site. This was initially interpreted as evidence that the deep horizon may have been man-made, since a paleochannel feature would have pre-dated the trees and should therefore not be affected by their placement. Tree roots appear as deep as 30 ns at the northern ends of the GPR profiles (see Figure 6), but the sand-clay boundary is significantly deeper than that (~60 ns) and therefore not likely to be affected by the roots.
The variability in the sand-clay boundary in the main clearing of Wyatt Chapel Cemetery has several possible interpretations. Anthropogenic causes for the change in this reflector are not entirely consistent with the data collected. The continuous horizon lies undisturbed for most of the bottom of the trench indicating a stream bed or a slump from subsidence. The geometry of the surface displayed in Figure 6 is consistent with a channel. Anthropogenic causes for this trench would not leave a continuous horizon; it would leave the horizon broken and disturbed. Therefore, the abrupt depth changes in the sand-clay boundary are most likely related to a paleochannel of Pond Creek rather than a large excavation or other man-made feature.
We were able to demonstrate the efficacy of utilizing 3-D grids of GPR lines at this site (see Figure 7), which should be followed up in future studies. 3-D grids of GPR lines can provide detailed pictures of the subsurface that can reveal additional information about burials at the site. We also recommend clearing a larger area to acquire more GPR data and map the boundaries of the cemetery.
No comments:
Post a Comment