A Cultural Resource Assessment: The Historic Terreplein at Castillo de San Marcos National Monument, St. Augustine, Florida


Material Information

A Cultural Resource Assessment: The Historic Terreplein at Castillo de San Marcos National Monument, St. Augustine, Florida
Series Title:
Castillo de San Marcos Terreplein Project
Physical Description:
Carl D. Halbirt
Publication Date:
Physical Location:
Folder: Terreplein report


General Note:
NOTE: Most images in the "Figures" section (pages 40-52) are missing

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
97-3003; 93-3000; 94-3000
Castillo de San Marcos
System ID:

Full Text


A Cultural Resources Assessment: The Historic Terreplein at
Castillo de San Marcos National Monument, St. Augustine, Florida


Carl D. Halbirt
City Archaeologist

Planning and Building Department P.O. Drawer 210 St. Augustine, Florida 32085-0210

August 2001

Table of Contents
List of Figures .................................. ................. iii
List of Tables ..................................................... iv
Introduction ....................................................... 1
Previous Archaeological Research ................................... 4
Excavation Results: The Stratigraphic Record ................................ 8
Test Unit 1 ............... ........... ...................... 8
Stratigraphy ............................................... 9
Coquina-stone Vault ..................................... 14
1939 Concrete Wall ........................................ 15
Test Units 2 and 3 ............................ .......... .. ......... 15
Stratigraphy ............................................ 16
Test Unit 4 .................................................. 17
Stratigraphy ........................................... 18
Test Unit 5 ............................... .................... 19
Stratigraphy ................................. ......... 21
Coquina-stone Vault ...................... ............... 26
Material Culture Inventory ............. ............................ 27
Conclusion .................. .......... ......................... 33
References Cited .............. ..................................... 37
Figures 1 through 13 ............................................... 40-52


List of Figures

Figure 1. Location of 1939 and recent excavations on the terreplein ................ 40

Figure 2. NPS illustration of the sequence of pours used to construct the terreplein ..... 41 Figure. 3. Plan and profile perspectives of Test Unit 1 .......................... 42

Figure 4. Photograph of tabby blocks at the base of terreplein deposits in Test Unit 1.... 43 Figure 5. Photograph of grooves in coquina-stone vault about Room 16 .............. 44

Figure 6. Profile of Room 16 showing relationship of capstone to coquina-stone vault .... 45 Figure 7. Photograph of grooves and cuts along edge of capstone above Room 16 ...... 46 Figure 8. Photograph of 1939 concrete wall trench above Room 16 ................. 47

Figure 9. Plan and profile perspectives of Test Units 2 and 3 above Room 7 ........... 48

Figure 10. Plan and profile perspectives of Test Unit 4 ............................... 49

Figure 11. Plan and profile perspectives of Test Unit 5 ........................ 50

Figure 12. Photograph of soil stratigraphy in Test Unit 5 .............................. 51

Figure 13. Photograph of tabby blocks in Stratum Q of Test Unit 5 ................. 52


List of Tables

Table 1. Summary of Stratigraphic Deposits in Test Unit 1 ........................ 10

Table 2. Summary of Stratigraphic Deposits in Test Unit 5 ......................... 20

Table 3. Artifact Types Recovered from Test Unit 1 According to
Stratigraphic Level ................................................ 28

Table 4. Faunal Remains Recovered from Test Unit 1 According to
Stratigraphic Level ............ ................................... 29

Table 5. Artifact Types Recovered from Test Unit 5 According to
Stratigraphic Level ............................................... 30

Table 6. Faunal Remains Recovered from Test Unit 5 According to
Stratigraphic Level .............. ................................. 31



As the City's oldest standing building and the nation's oldest stone fortress, Castillo de San Marcos National Monument is a significant landmark. Its history is inextricably linked to colonial conflicts that occurred along the Atlantic seaboard and to international politics that precipitated those events. The story of Castillo de San Marcos, its builders and occupants, is one of resilience and fortitude. Throughout much of its colonial history, the community of St. Augustine was a military outpost subject to the vagaries of an unreliable supply system known as the situado, an annual crown subsidy of funds and goods (Sluiter 1985; Bushnell 1981) and the constant threat of invasion. On five separate occasions the town was attacked and all or portions of it were destroyed: in 1577 by Native American Timucuans (Connor 1925:263; Manucy 1962:15); in 1586 by Sir Francis Drake (Bushnell 1983); in 1668 by the pirate Robert Searles; in 1702 by Governor James Moore (Crane 1956; Arnade 1959); and in 1740 by General James Oglethorpe (Arana and Manucy 1977). This list does not include the various enemy forays by both European and Native Americans that occurred around St. Augustine from the late 1600s until the end of the First Spanish period in 1763. The one thing that enabled the Spanish to maintain their sovereignty of La Florida, especially during the turbulent 18th century, was the Castillo. It was a symbol of Spanish power and provided sanctuary for the town's population during times of crisis.

Castillo de San Marcos (referred to herein as Castillo) was built between 1672 and 1695 at the site of a lime kiln (Arana n.d.a), a short distance north of the ninth wooden fort also referred to as San Marcos. This site was selected based on its strategic location, which "offer[ed] full opposition to any adversary, without resorting to the destruction of the town's houses in the name of military necessity" (Arana n.d.a: 9).1 Construction was a response to two primary events (Arana n.d.a): 1) the devastating 1668 pirate raid of Englishman Robert Searles, an activity sanctioned by the English government (Ritchie 1986), and his threat of return, and 2) the establishment in 1670 of the English settlement of Charles Towne (later Charleston), South

1 Eventually, the town would move to the north, a move that precipitated the destruction by the Spanish of those structures within the Castillo's "field of fire" during Colonel James Moore's 1702 siege of St, Augustine (Arnade 1959). The field of fire corresponds to a distance of about 750 feet, or the distance of "a musket shot" of the fort.

Carolina. Along with the unreliability of earlier Spanish wooden forts to withstand the harsh Florida climate,2 these two events prompted the Spanish Crown in 1671 to allocate 10,000 pesos for the construction of a more permanent symbol of Spanish sovereignity in what was then La Florida. This area extended from Florida into the coastal environments of Georgia to St Catherine's Island near Savannah, Georgia. Eventually, 138,375 pesos were spent in the construction of the masonry fort, which would take 23 years to complete. Subsequent remodeling and repair activities occurred throughout the 18th and 19th centuries, with the most notable modification occurring between the late 1730s and the late 1750s. A comprehensive review of the construction history of the fort can be found in Arana (n.d.a, n.d.b), Arana and Manucy (1977), Bearss and Paige (1983), and Sastre (2000) and need not be reiterated here.

This report focuses on one aspect of the Castillo's unique architecture that has witnessed repeated repair and remodeling efforts (Manucy 1939; Arana n.d.b; Bearss and Page 1983). That is, that portion of the modern terreplein, or gun deck, that occurs over the coquina-stone bomb vaults (i.e., rooms or casemates), which have been in existence since the mid-18th century. The terreplein is a complex architectural system that consists of "an earthen [tabby] deck upon which defensive cannon are operated... stone firing steps, sentry boxes, chimney penetrations, and a rainwater management system consisting of drainage ports, scuppers, and flashing" (Frazier, Copeland, and Arana 1986:93). It was built on top of fill deposits within the Castillo's four bastions and the coquina-stone bomb vaults that replaced the wooden framed, tabby rooms constructed in the late 17th century. These earlier rooms were not "bomb proof" (Manucy 1939:3) and were in a state of disrepair by the 18th century (Arana and Manucy 1977).

The modern terreplein has been a problem since its construction due to a continual "dripping-type" water leakage into the bomb vaults, which were the fort's living and storage quarters (Arana n.d.b; Bearss and Page 1983). The issue still exists even after resurfacing efforts by the National Park Service (NPS) in 1939 and 1960, using impermeable materials. The problem

2 Prior to the construction of Castillo de San Marcos, wooden forts had protected the colony of St. Augustine. Nine forts were built between 1565 and 1672 (Connor 1929), seven of which were constructed in the 16th century.

is that too much moisture is
being allowed into the coquina vault masonry. This water infiltration,
in conjunction with fluctuating temperature and stagnant air with
excessive'humidity, causes alternating drying/wetting cycles, which are detrimental to the masonry fabric: The coquina is weakened, the mortar is leached, and the plaster softens and spalls. In addition, the excessive
moisture permits vegetation growth (especially ferns), which
mechanically accelerates masonry deterioration (Frazier, Copeland, and
Arana 1986:99).

Leakage also is a major problem over the bastions where structural destabilization can occur as a result of the settling of fill deposits. This settling has resulted in cracks that have developed on top of the bastion, especially the San Pedro and San Pablo bastions. Some of these cracks have been documented as being about 12 feet deep (Manucy 1939:14).

To rectify this problem, the NPS in early 2001 began an extensive stabilization project of the entire terreplein surface. As part of this process, archaeological investigations were necessary to define and clarify the nature and extent of the fill and earthen surfaces that comprise the gun deck both on top of bomb vaults and on top of bastions. The purpose of the investigation is to develop a course of action to correct the leakage problem while creating as little impact as possible to the historic fabric of Castillo. The NPS divided the work into two projects. The City of St. Augustine Archaeology Program conducted investigations over the bomb vaults from November 1997 to March 1998. Southeast Archeological Center (SEAC), part of the NPS, conducted work for the bastion area (Lawson and Cornelison 2001). This report discusses the results of the City's investigation as well as its relationship to previous archaeological investigations that have occurred at Castillo, especially on the terreplein.

Previous Archaeological Investigations

Archaeological research has been conducted at Castillo for more than 60 years. Fischer (1976), Deagan (1980), and Lawson and Cornelison (2001) have summarized much of this research. These studies need not be reiterated herein except to say that some researchers have been critical of previous NPS archaeological efforts, especially in terms of objectives and methods (Deagan 1980; Kelly 1940).

Of the work conducted on the terreplein, the project that must be considered here is that reported by Albert Manucy (1939). Then a junior research technician at the fort, Manucy had the opportunity to investigate the terreplein before the establishment of the 1939 concrete gun deck. He investigated five test units over two casemates, or bomb vaults (Rooms 7 and 16). The objectives of the project were threefold:

1) to examine the construction of the original terreplein; 2) to examine
the construction of casemate arches; 3) to record historical evidence
revealed by uncovering the original terreplein (Manucy 1939:6)

Work proceeded after a "PWA crew" (Public Works Administration) removed the concrete tile, which had been placed over the historic terreplein in the 1880s or 1890s (Manucy 1939:5). Afterwards, PWA crews excavated five test units: three test pits and two test trenches (Figure 1). Excavation occurred under the direction of engineers and contractors, and Manucy recorded the results of these excavations. The fact that the 1939 report was the result of lay personnel untrained in archaeological field methods was one of the criticisms levied at Manucy's report by A. R. Kelly, chief regional archaeologist of the NPS (Kelly 1940). This and other criticisms were responded to by Manucy, who stated that while "Dr. Kelly's criticisms are justifiable, and to the point, it is difficult to apply a newly developed procedure to already going emergency work" (Manucy 1940: 1).

The first two test units recorded by Manucy were along the west terreplein surface over Room 7, historically a multipurpose room. One test unit (Unit 1) was excavated atop the arch,

whereas the other (Unit 2) was placed within the spandrel (i.e., the space between the arches of Rooms 7 and 8). Both units showed evidence of three floors, which consisted of "burnt oyster shell, lime, sand, andcoquina gravel [that] is remarkably hard" (Manucy 1939:8). The floors, labeled Layers A, B, and C, were considered by Manucy to have been formed by pouring "wet mortar," known as tapia, and then tamping the surface. Floors were separated by a "uniformly dark, greenish-brown stain" at the surface, which Manucy (1939:8) considered to be caused by algae formation in between the pouring of each floor. An anonymous note written on Manucy's manuscript, however, suggests that the algae represent "different levels at separate periods." Layer C, the lowest floor, was found to occur atop the coquina-stone arch, becoming thicker as it extended over the spandrel (i.e., from 3 to 12 inches). The spandrel had been filled initially with a 3-1/2-ft layer (Stratum E) of loosely packed coquina gravel, followed by an 18-inch layer (Stratum D) of hard-packed coquina chipping. One unexplained historical construction detail observed by Manucy was a

curb-like strip extending from one parapet wall to the other along and
over the center of the arch.. No similar feature was found elsewhere on
the terreplein. Apparently of red clay and lime, poured in place over
stratum B, this strip is about 4 inches square, and its top is at the level
of stratum A (Manucy 1939: 9)

The function of this "curb-like" feature will be discussed later, based on the results of the City investigation.

The other three test units were placed over Room 16, then known as the Fern Room
"because of its luxurious natural growth of southern maiden hair fern [that] festoon the walls and arch, watered by seepage of moisture from the top of the roof' (Manucy 1939:11). Historically, Room 16 had been the location of the King's treasury. Excavation involved a test pit (Test Unit 5) and two test trenches (Test Units 3 and 4). The pit was placed over the center of the arch, whereas the trenches were placed in the spandrel on either side of Room 16 (Figure 1). This strategy was used in response to an administrative decision to preserve the Fern Room, while ensuring that the moisture did not spread over adjacent rooms. The two trenches correspond to

the location of buried concrete barriers that would limit the movement of moisture.

Results from the two test trenches (Test Units 3 and 4) revealed a series of stratified deposits extending down 3.5 ft to "the old terreplein surface." The upper deposit was a tapia layer 5 inches thick that was followed by sand and shell, coquina gravel, coquina shell, and broken shell. These strata extend to both the curtain and courtyard walls, whose condition also-were noted. The curtain wall was stepped and more finished than the "rough character of the masonry of the courtyard wall" (Manucy 1939:12). In Test Unit 5, the arch of Room 16 was found to be 6 inches below the tapia surface.

Based on the results from the five test units, as well as observations from elsewhere on the gun deck, Manucy was able to make some general remarks concerning the composition of the terreplein that covered the casemates as well as the bastions. Concerning the terreplein over the casemates, Manucy observed that

most of the eastern (especially the section near the ramp) and
western terrepleins were found surfaced with tapia of considerable
hardness, but the north and south terrepleins were found to be
surfaced mainly with a loose mixture of sand and coquina
gravel... Occasional small holes in the terreplein were found
plugged with American period cannon balls (Manucy 1939:14).

Manucy (1939:14) attributed the inconsistent character of the terreplein over the casemates as a consequence of actions taken by Lt. Stephen Tuttle in 1833. Tuttle removed those portions of the tapia that were broken. Eventually, these locations were filled in with sand when the terreplein was resurfaced. Manucy's findings became the foundation for the NPS brochure, which illustrates the method in which the terreplein was constructed (Figure 2).

Manucy (1939:17) also documented the presence of 6-inch, hand-wrought, iron spikes
that had been driven into the tapia in many places, especially the western front. These spikes were regularly patterned, indicating that they supported tents or a small wooden building that had been placed atop the terreplain. Manucy attributed this activity as occurring during and after the Civil

War when Union troops were stationed in St. Augustine (Arana 1986:53-54). The other possibility is that these structures could be associated with the incarceration of indigenous Southern Plain Indians from 1875 to 1878. It was realized that housing the Native Americans in casemates would be injurious to their health. According to Lt. Richard H. Pratt, who was in charge of the Native American prisoners, "a musty sickly odor" permeated every casemate and each was dripping with water. Therefore, both tents and shed barracks were constructed on the terreplein (Bearss 1983:276-281). In 1886, Chiricahua Apaches were imprisoned at the Castillo and similar responses by the military were undertaken.

Two other archaeological investigations have occurred atop the terreplein prior to the present NPS-sponsored investigations. In 1988, a limited testing program was implemented by Bruce Piatek, of the Historic St. Augustine Preservation Board, to locate the historic firing step at the apex of the San Pedro bastion. SEAC archaeologists expanded Piatek's efforts later in 1998 (Lawson and Cornelison 2001). In 1991, Ken Wild (SEAC) monitored a series of 4-inch diameter, 3-ft deep bore test holes placed in the terreplein to evaluate the condition of the two modern concrete surfaces (Wild 1991). The core samples interpreted by Wild were not consistent with Manucy's findings. Rather, Wild concluded that there might be more variation in the stratigraphic deposits than originally thought. Moreover, Wild noted that the estrados (i.e., the top surfaces) of the casemates were closer to the present surface of the gun deck than that originally indicated by Manucy. This is especially true along the curtain or scarp wall because of the slope of the terreplein, which slopes downward from the courtyard wall to curtain wall.

To clarify the discrepancy between Manucy's findings and the work conducted by Wild, as well as to define the characteristics of the historic terreplein, NPS officials selected five locations to be investigated by the City. The purpose was to gather data to properly plan for the replacement of the existing concrete slab with a new waterproof roof and walking surface. Although NPS personnel realized that not all questions related to the discrepancies between Manucy and Wild would be answered, the five areas selected would "confirm some important conditions" that would allow for the design phase of the project to progress (Anonymous 1997).

Excavation Results: The Stratigraphic Record

The five test units excavated provide a unique glimpse into the construction of the modern terreplein (gun deck), which spans approximately 250 years of use. The information obtained during this investigation indicates that the terreplein deposits vary considerably in the types and degree of material used in its construction. This section addresses the various beds or stratigraphic deposits found in each test unit and how those deposits associate with the coquinastone casemates, which support the gun deck.

It is evident from this investigation that the quantity and composition of materials used in the construction of the terreplein differed. For example, 15 distinct strata comprise the terreplein surface between Rooms 15 and 16 (Test Unit 1). This contrasts with the 37 strata noted over Room 21 (Test Unit 5), and the 11 strata noted over Room I (Test Unit 4). Yet, stratigraphic deposits appear to be consistent between the courtyard wall and the curtain over each room-as exemplified by Test Units 2 and 3, which were above Room 7 and the findings of Albert Manucy over Room 16 (Manucy 1939:12). Clearly, historical repairs and/or resurfacing of the terreplein varied; however, attempts were made to maintain some consistency over each room.

The following describes the stratigraphic deposits that comprise the historical terreplein of Castillo de San Marcos. For Test Units 1 and 5, the strata are discussed in order of earliest (deepest) deposits to latest (upper) deposits given the complexity of the deposits. In the other three test units (Test Units 2, 3, and 4) the deposits are discussed from top to bottom.

Test Unit 1
Test Unit 1 was excavated to determine the characteristics of: 1) the 1939 wall
constructed in the middle of the spandrel that separates the coquina-stone vaults of Rooms 15 and 16, and 2) to evaluate the nature of the terreplein deposits above the Room 16. Initially, a 1-m by 3-m test unit was excavated, which straddled the 1939 wall, with the excavation proceeding according to the cultural deposits that form the terreplein. Later, with approval of park

Superintendent Gordon J. Wilson, the test unit was expanded an additional 4 m to the east (Figure 3a), which effectively covered the width of Room 16. Within this unit a total of 15 different strata or substrata were observed (Figure 3b; Table 1).

The earliest of these deposits (Stratum A) is a pale yellow (Munsell 2.5Y 8/4) coquina sand, which has been compacted. This sand represents a single filling episode that bridges the lower arches of Rooms 15 and 16.3 It extends from 0.97 m to more than 1.37 m below the established site datum (BESD), which is at the southwest corner of the courtyard wall. Toward the base of Stratum A is a layer of broken tabby blocks at approximately 1.27 m BESD (Figure 3b). The depth of these blocks was not determined due to the constricted area of excavation. The possibility exists that they may extend to where the vaults of Rooms 15 and 16 converge. The blocks may have constituted some type of buttressing system to help stabilize the vaults and/or terreplein. The top of these blocks forms a pavement-like surface (Figure 4), which Manucy thought to be the "old terreplein surface." The overlying coquina sand deposit (Stratum A) would have permeated the crevices of these intentionally placed blocks.

The tabby blocks are irregularly shaped. They measure more than 20 cm in length and width and vary from 10-15 cm thick. The origins of these blocks are unknown, but they are probably the remains of demolished structure(s)-possibly the original walls and/or roofs of the 17th-century tabby casemates (Arana and Manucy 1977:32).

Above the coquina sand fill is a gray, partially consolidated tabby zone that is composed of three distinct substrata labeled B-1, B-2, and B-3 (Figure 3b). These substrata separated easily from each other during excavation. This deposit is 22 cm thick; Zone B extends from approximately 75 cm to 97 cm BESD. The tabby surfaces varied in color from light gray (Munsell 2.5Y 7/2) to light brownish gray (Munsell 2.5Y 6.2) and contained various amounts

3 The lower elevation of this fill deposit was not determined given the presence of the 1939 wall and the constricting coquina-stone arches that effectively limited further excavation.

Table 1. Summary of Stratigraphic Deposits in Test Unit 1.

Strata Munsell Color Soil Characteristics

L 10YR 8/2 1960s concrete slab K O1YR 8/1 1939 concrete slab J 10YR 8/1 Concrete pavers I 2.5Y 7/2 Coquina sand fill H 2.5Y 7.3 Coarse coquina sand with shell and some humic material, also includes numerous tabby rubble fragments G 2.5Y 8/4 Coarse coquina sand with some coquina-stone F 2.5Y 7/2 Coarse coquina sand with tabby chunks and fragments. (Zone may have originally represented a tabby surface.) E 2.5Y 8/4 Coarse coquina sand, similar to Zone C2, surface is flat D 10YR 4.3 Brown fill zone composed of sandy silt with numerous tabby, charcoal, shell, sand concretion inclusions. (Soil possibly originates from glacis deposits outside fort) C2 2.5Y 8/4 Coarse coquina sand, more uniform, smaller shell fragments than CI C1 2.5Y 7/4 Coarse coquina sand, numerous shell fragments of varying size, compacted and possibly lightly cemented B3 2.5Y 6/2 Tabby floor with numerous tabby rubble, very carbonaceous and friable. (Floor may have become friable through use) B2 2.5Y 7/2 Tabby zone with numerous tabby chunks and rubble, much less charcoal than Zone B3. (B2 is more consolidated than B3 with tabby chunks cemented into matrix.) B1 2.5Y 6.2 Tabby floor with oyster shell fragments and some charcoal A 2.5Y 8/4 Compact coarse coquina sand/gravel fill TS ---- Tabby Stone Fragments

and sizes of tabby and coquina-stone rubble as well as charcoal and shell. Zone B is situated where the arch of the coquina-stone vault becomes less acute. Consequently, the three substrata in this zone might have formed the first floor surface to the terreplein--a fact that is supported by the numerous artifacts recovered from this stratum (see below).

Distinctions between the three substrata in Zone B were based primarily on compaction and quantity and types of inclusions. For substratum B-1, the deepest deposit, the tabby matrix was friable and contained mostly shell inclusions with some charcoal and tabby and coquina-stone rubble fragments.4 This differed from substratum B-2, which was more consolidated and contained numerous tabby and coquina-stone chunks that were better cemented into the tabby matrix. Substratum B-3, the upper tabby zone, was very carbonaceous and friable. Associated with the tabby matrix of B-3 were some large tabby blocks found over Room 15 (Figure 3b). No tabby blocks were documented in substratum B-3 for Room 16. In terms of preservation, B-2 was in a better condition than either B-1 or B-3. The poor condition of B-3 may be a result of an extended exposure to weather and/or use. This supposition is supported by wear marks consisting of grooves and depressions at a comparable elevation as the top of B-3 (i.e., 75 cm BESD) in the coquina-stone vault over Room 16 (Figure 5). The carbonaceous nature of B-3 could be associated with fires on the gun deck.

Following the gray tabby zone, the next deposit (Stratum C) consists of a coquina sand/ gravel layer that extends from 52 cm to 75 cm BESD. Stratum C is represented by two substrata (C-1 and C-2). Both are pale yellow (Munsell 2.5Y 7/4 and 2.5Y 8/4) and both appear to be partially cemented. The characteristics that distinguish C-1 and C-2 are compaction and particle size. Substratum C-1 was more compact with variable particle size, which included some chunks of coquina stone. Substratum C-2 had smaller particles, including some coquina-stone rubble and tended to be of a looser nature. Whether Stratum C constitutes a floor surface or fill deposit is

4 The terminology used for tabby and/or coquina-stone construction debris found with a stratum is based on particle size: block refers to fragments greater than 10 cm in length, chunks of tabby range from 5 cm to 10 cm in length, and rubble is less than 5 cm in length.

equivocal. The stratum does appear to have been cemented to some degree, but its compaction simply might be due to compression from above. Conversely, it is possible that the fill of Stratum C was used to elevate the terreplein to just below the vault's capstone. The top of Stratum C is 20 cm below the top of the capstone of Room 16 (Figure 3b).'

Atop the coquina stone was a thin (3-5 cm) brown (Munsell 10YR 4/3) band of sandy-silt labeled Stratum D, which contained numerous inclusions of tabby and coquina-stone rubble, sand concretions, shell, and charcoal. What this band of sandy-silt represents is uncertain. Clearly, it constitutes soil deposited onto the terreplein surface, which may have originated from the surrounding fort glacis or courtyard area. Stratum D essentially levels the terreplein to the top of the coquina-stone arch or the base of a capstone at the crest of the vault (Figure 3b). The undulating nature of Stratum D's surface could have been caused by foot and/or artillery traffic. The numerous artifacts found within the deposit suggest that they might represent incidental inclusions or be a by-product of occupation at this level on the terreplein.

Above Stratum D was another zone of coquina sand (Stratum E), which was similar to
Stratum C-2. Stratum E is pale yellow (Munsell 2.5Y 8/4), with numerous inclusions of coquinastone rubble and chunk fragments. The stratum was poorly consolidated with little-if anyevidence of the matrix having been cemented with tabby. Stratum E sloped slightly from east to west, with the thickness varying from 15 cm over Room 16 (32 to 47 cm BESD) to 11 cm over Room 15 (39 to 50 cm BESD). Of particular importance is the fact that the thickness of Stratum E over Room 16 is comparable to the thickness of the capstone (see below) over the same room (Figure 3b). It is not until Stratum E, which is only 28 cm below the existing gun deck, that we find evidence for the deposits within the spandrel to be at an elevation similar to the capstone atop the arch of Room 16. That the seven previous strata were lower than the capstone indicates that the coquina-stone vaults were obstacles in the movement of artillery and soldiers along the terreplein for part of its use-life--a fact substantiated by wear marks on the estrado of Room 16.

5 A similar pattern could not be discerned over Room 15 given the location of the test unit, which did not extend far enough west to determine association between the coqunina-stone vault and Stratum C.

The next three deposits (Strata F, G, and H) are similar to the coquina-sand matrix found in Stratum C and E. These three strata were not recognized during the excavation due to the use ofjackhammers used to remove the surface concrete over Test Unit 1. They were recognized, however, when the profile drawings were made (Figure 3b). These three deposits are either light gray (Munsell 2.5Y 7/2) or pale yellow (Munsell 2.5Y 8/4) and contain numerous inclusions of white tabby rubble and chunk fragments, especially in Stratum F. The combined thickness of these three strata varied from 14-20 cm, with the shallower deposits occurring over Room 16 and the deeper deposits over Room 15.

It is likely that these three strata correspond to Manucy's (1939) upper deposit that was approximately 5 inches thick, which might have replaced the broken and worn out compacted tapia surface mentioned by Lt. Stephen Tuttle in 1833. The possibility exists that Stratum F might have represented the compacted tapia floor surface due to the quantity of tabby rubble and chunks observed in the profile of the test unit. The tabby fragments were similar to the well-preserved tabby levels documented in Test Unit 4. Stratum F was subsequently broken, which resulted in a discontinuous layer of tabby that was covered with a coquina-sand fill.

The final four strata (I, J, K, and L) are associated 20th-century preservation activities, all of which had been removed by the jackhammer when the test unit was opened. The earliest is Stratum I: a fine-grained, white (Munsell 2.5y 7/3) sand that has a thin petroleum-rich lens along its base. Stratum I occurs only above Room 16 (Figure 3b) and is considered to be associated with the 1939 terreplein preservation efforts. The stratum is approximately 8 cm thick (10-18 cm BESD). It extends to the 1939 concrete wall construction trench, where it either ends or becomes a discontinuous lens, depending on which profile one examines. No evidence of Stratum I was found over Room 15.6 The other three strata are concrete. Stratum J consists of 5 cm thick (5-10 cm BESD) concrete pavers placed over the terreplein deposits of Room 16 and the

6 That no evidence of Stratum I was found above Room 15 can be attributed to preservation concerns in 1939. At that time, it was decided to preserve the "luxurious natural growth of southern maidenhair fern" within Room 16 (then known as the fern room) by not applying a waterproof concrete membrane over this one room (Manucy 1939:11). The remaining rooms, including Room 15, were sealed under a poured concrete slab, which in Test Unit
1 is Stratum K.

1939 concrete wall construction trench. Strata K and L are poured concrete slabs over Room 15, which have been reinforced with wire-gauge mesh. The combined thickness of Strata K and L is 17 cm (5-22 cm BESD). A thin (1-cm thick) Laykold membrane was applied over these concrete surfaces (Frazier and Copeland 1986).

Coquina-Stone Vault
Given the length of the excavation area for Test Unit 1, which was 7 m east to west, the architectural characteristics of the coquina-stone vault could be reconstructed. Figure 3b illustrates the articulation between the historical terreplein deposits to the upper elevations of the coquina-stone vaults of Rooms 15 and 16. The thickness of the vaults varies from 21.6 inches (55 cm) along the arch to 25.6 inches (65 cm) at the capstone, according to elevation readings for Room 16. Based on the projected configuration of the vault ceiling onto the piers for Room 16, the fill deposit (Stratum A), which includes the tabby blocks, extend to a depth of at least 5 ft (1.55 m), from 97-249 cm BESD. It is not until the vault's estrados (top of the arch) become less acute at a depth of 97 cm BESD that a potential floor surface begins (Zone B). This also is the point at which the estrados evidence scouring marks in the form of grooves and depressions (Figure 5). The movement of artillery might have caused these marks. Whether this pattern is repeated over Room 15 could not be determined with the existing database, although it is assumed--given the consistency of the historical terreplein deposits between the two rooms--that the characteristics of that encountered for Room 16 would be duplicated for Room 15.

Unlike previous interpretations (Manucy 1939; Arana and Manucy 1977; Frazier, Copeland, and Arana 1986), the top of the coquina-stone vault of Room 16 is not rounded. Instead, the arch gives way to a flat coquina-stone capstone that is embedded into the arch and rises 16 cm above the estrado. The capstone measures 4.47 m wide and is offset from the center of the vault (Figure 6). The capstone's depth along the east side was not determined in this investigation, given that authorization was not given to expand the test unit farther east. As such, the capstone's depth along the east side was not determined, although it can be speculated given the offset nature of this feature that it is deeper. Deep grooves are present along both edges of

the capstone (Figure 7), indicating that this rise also formed an obstacle in the movement of artillery.

1939 Concrete Wall
One of the principal goals in excavating Test Unit 1 was to determine how the
construction of the 1939 concrete wall impacted the historical terreplein deposits and what are the characteristics of that wall. To address these questions, Test Unit 1 bisected the wall enabling a complete profile to be obtained (Figure 3b). It is evident that the 1939 wall and associated construction trench extends approximately 1.15 m (10-127 cm BESD) into the historic terreplein deposit. The construction trench is tapered and varies from 50 cm at the top to 22 cm at its base. Fill deposit within the construction trench is coarse gray (Munsell 5Y 6\1) coquina sand. The trench stops at the top of the tabby blocks within Stratum A--a fact that is supported by the presence of some concrete that has extruded onto the top of some of the tabby blocks at an elevation of 1.26 cm BESD (Figure 8)

The concrete wall was constructed by building a wood form on the east side of the wall,
probably near the center of the trench and then using the historical deposits on the west side of the construction trench as the other buttress. As a result, the east face of the concrete wall is smooth, whereas the west face could not be excavated. The concrete solution had effectively permeated and cemented the historical terreplein deposits for a distance of between 2 cm at the top to 15 cm at the base. It is estimated that the concrete wall varied from 18 cm wide at the top to 28 cm wide at the base, with the discrepancy due to a 5-degree slope for the wall along the east face.

Test Units 2 and 3
Within a few days of completing the excavation of Test Unit 1, maintenance staff from Castillo de San Marcos opened two new test units over Room 7 that measured 1.5 m by 1 m in area. The units were opened using a jackhammer. The purpose of these units was to confirm Albert Manucy's (1939) observations at Room 7, as well as to establish the slope of the historical terreplein between the courtyard wall and the curtain as proposed by Wild (1991). The units were

placed at the crest of the coquina-stone vault above Room 7, with the aim of documenting the association between the historical terreplein deposits and the configuration of the coquina-stone vault. Unfortunately, the placement of these units did not take into consideration the possibility of a coquina capstone, similar to Room 16. As a consequence, only 18-32 cm of terreplein deposits were present, including the modern 20th-century concrete surfaces and these were removed during jackhammering. Thus, the archaeological investigation of Test Units 2 and 3 was limited to cleaning the units and documenting the deposits.

The stratigraphy of Test Units 2 and 3 was comparable. The same four strata were
identified in both units (Figure 9): two layers of poured concrete that were reinforced with wire mesh and then covered with a Laykold membrane;7 a lens of light gray (Munsell 2.5Y 7/4) sand; and a hard, white (Munsell 2.5Y 8/2), tabby-coquina surface. These strata overlay the coquina capstone that varied from 18-32 cm below the existing terreplein. In terms of specific elevational differences, the concrete strata ranged from 4-22 cm BESD in Test Unit 2 and from 16-32 cm BESD in Test Unit 3. The light gray sand was fully visible in Test Unit 2, occurring between 22-28 cm BESD, but only partially present in Test Unit 3 occurring as a thin band (32-33 cm BESD) along the east side of the test unit. The tabby stratum was found primarily in Test Unit 2, from 28-36 cm BESD, with only trace amounts along the east edge of Test Unit 3 at 33-35 cm BESD. The only cultural zone, which was level, was the capstone that occurred between 35-36 cm BESD in both test units.

Although the excavations in Test Units 2 and 3 did not confirm the findings of Albert
Manucy's investigations, given the placement of the test units, it did provide information relevant to understanding variation in the terreplein elevation between the courtyard wall and curtain. The terreplein surface was, on the average, 13 cm higher along the courtyard wall than at the curtain. Most of this difference is due to the presence of the tabby layer and the light gray sand found in

1 The two concrete strata documented in Test Units 2 to 5 correspond to Stratum K and L in Test Unit 1.

Test Unit 2, which account for 14 cm of the 33 cm of historical terreplein deposits documented. This is in comparison to the same strata, which constituted only 3 cm of the 18 cm of historical deposits, recorded in Test Unit 3. When the 9.2 m distance between Test Units 2 and 3 is considered, these historical deposits have a 1.6 degree slope from east (the courtyard wall) to west (the curtain). Only the capstone of the coquina-stone vault was level. This indicates that although the terreplein may have slope, as a means of controlling water drainage, the superstructure used to support the terreplein was plumb.

In terms of information related to characteristics of the coquina-stone vault, little can be
inferred with the extant data. The only information recovered concerns the capstone, in which the test units were placed directly atop. This information indicates that the capstone extends from the courtyard wall to the curtain and that the top was level. No attempts were made to determine the width of the capstone, although the orientation of both test units indicates that the capstone is more than 5 ft wide over Room 7.

Test Unit 4
The next area investigated was Test Unit 4 (Figure 10), an L-shaped excavation located over Room 1 near the courtyard wall. The unit was placed in the approximate location of a proposed elevator shaft, which will enable people with disabilities access onto the fort's gun deck. The intent of investigating this area was to define the characteristics of the historical terreplein between the current concrete slab and the coquina-stone vaults that would be impacted by the proposed elevator.

The L-shaped test unit proved to be within historical terreplein deposits, which extended more than 3 ft deep (see below), but not all of Test Unit 4 could be excavated. Upon removal of the two concrete surfaces, a very hard tabby-coquina layer was encountered. Because the jackhammer had loosened the initial 10-12 cm of this tabby, it was possible to use pick-axes to remove this layer. It was discovered, however, that a second hard tabby layer existed and this could not be excavated using pick-axes. The only way to excavate this second tabby-coquina

layer was by using a jackhammer, and it was decided that this could potentially create an adverse impact to the coquina-stone vault of Room 1. Therefore, a smaller 40 cm by 50 cm exploratory test pit was excavated in the northeast corner of the unit using chisels and sedge hammers (Figure 10a and 10b). The excavation of this small exploratory test unit took at total of 22 personhours--one of the most difficult units that was investigated at the fort. This exploratory test pit was taken down to 102-107 cm BESD at which point the coquina-stone vault of Room I was encountered. No further investigation of Test Unit 4 was undertaken after the excavation of the exploratory test pit.

As with the earlier test units investigated, the initial excavation of Test Unit 4 was done using a jackhammer to remove the 20th-century concrete deposits. The top elevation of the concrete surfaces varied from 2 cm BESD near the courtyard wall to 4 cm BESD at the south side of the test unit. The base also had a slight slope with the elevation varying from 20 cm BESD toward the courtyard wall to 24 cm BESD at the south end. Below these concrete slabs was a thin (1-2 cm) lens of light gray (Munsell 2.5Y 7/4) sand (Figure 10c), which was comparable to that found in Test Units 2 and 3. The sand also sloped from north to south with the elevation ranging from 21-25 cm BESD.

Following the removal of 20th-century deposits, a series of seven very hard tabby layers was documented (Figure 10b). These layers are similar in composition, consisting of a hard, white (Munsell 2.5Y 8/2), tabby-coquina mixture--the same mixture as found in the tabby level documented in Test Units 2 and 3.8 A minimal amount of oyster shell and charcoal inclusions were found within this tabby-coquina mixture. The tabby-coquina layers ranged in thickness from 8-13 cm and occur between 22-102 cm BESD at which point the coquina-stone vault was documented. A distinct boundary, which was smooth and flat and flaked-off during excavation, separates each of the seven tabby-coquina zones.

8 The tabby-coquina mixture documented on the terreplein is similar to some of the colonial road surface of historic St. George Street in St. Augustine (Halbirt 1997).

Based on the base elevation of Level 2, the one tabby-coquina layer that was completely excavated, the other six tabby-coquina layers (Levels 3 to 8) appear to be essentially level. The only surface that is sloping is the top of Level 2 (Figure 10c), which drops from 20 cm to 24 cm BESD-a 1.5 degree slope from north (the courtyard wall) to south (the curtain wall). This is the same slope as that found in Test Units 2 and 3. Had the test unit been placed atop the crest of the coquina-stone arch of Room 1, the bottom of Level 2 would be at the approximate elevation as the top of the coquina capstone found in Test Units 1, 2, and 3.

In terms of construction methods, the information from Test Unit 4 would suggest that it was not until after the terreplein deposits had reached the top of the coquina-stone capstone that the surface of the gun deck was sloped, thus allowing for water drainage. Prior to this elevation, the deposits appear to have been laid flat.

Test Unit 5
The last unit excavated was Test Unit 5, a 2 m by 1 m area located over Room 21. The focus of this test unit was to document the nature of the terreplein deposits near the fort's east curtain, as well as to verify historical documents and physical evidence relevant to the height of the coquina-stone vaults along the east terreplein. The top of the casemates along the east terreplein are estimated to be approximately 3 ft deeper than those along the north, west, and south sides of the fort.

In this unit, a total of 37 different deposits constitutes the terreplein (Figure 11 and 12; Table 2). These deposits extend more than 223 cm BESD. The bottom nine strata are at or below the top of the coquina vault. The remaining 28 strata are above the coquina-stone vault and vary from 15-148 cm BESD. Clearly, terreplein deposits along the east curtain are deeper and more complex than those found along any other section of the fort.

9 A parallel test unit (Test Unit 6) was to have been excavated near the courtyard wall to determine the similarity and elevational characteristics of deposits atop the coquina-stone vault of Room 21. Inclement weather necessitated the closing of Test Unit 6 within a few days after removal of the concrete slabs.

Table 2. Summary of Stratigraphic Deposits in Test Unit 5.

Zone Munsell Color Soil Characteristics

CC 1960s concrete slab BB 1939 concrete slab AA2 10YR 5/2 Tabby layer with crushed coquina shell AAl 10YR 5/2 Tabby layer with some conquina stone fragments Z 10YR 7/2 Tabby layer Y 10YR 8/6 Very consolidated finely crushed coquina X5 10YR 8/6 Crushed coquina shell X4 10YR 6/1 Tabby layer X3 10YR 5/1 Tabby layer with small coquina-stone and tabby rubble X2 10YR 6/1 Tabby layer with coquina-stone rubble X1 10YR 5/1 Tabby layer with charcoal flecks W 10YR 8/4 Crushed coquina shell V 10YR 5/2 Tabby layer with small chunks of coquina-stone and tabby U 10YR 8/4 Crushed coquina shell with some coquina-stone fragments T 10YR 4/2 Tabby layer with some coquina-stone S 10YR 8/6 Crushed coquina shell with coquina-stone fragments R 10YR 8/6 Crushed coquina shell with sand Q 10YR 7/3 Tabby with a coquina shell base, some tabby rubble P IOYR 7/2 Tabby deposit O 10YR 6/3 Large tabby chunks within an friable layer of tabby N 10YR 6/3 Tabby with some large chunks of tabby M 10YR 4/3 Sandy-silt earthen lens L 10YR 5/2 Tabby with tabby and shell fragments and crushed coquina shell K 10YR 6/3 Pocket of small tabby fragments and crushed coquina shell J 10YR 7/2 Crushed coquina shell and sand I 10YR 7/3 Crushed coquina shell and sand with some tabby fragments H 10YR 5/1 Tabby zone

Table 2. Summary of Stratigraphic Deposits in Test Unit 5 (continued).

Zone Munsell Color Soil Characteristics

G 10YR 5/2 Crushed coquina shell with tabby and coquina-stone fragments GG 10YR 5/1 Crushed coquina shell F 10YR 7/2 Crushed coquina shell and sand EE 10YR 6/2 A thin layer of tabby DD 10YR 3/1 A thin black band of silty-sand with organic debris E 10YR 6/3 Crushed coquina shell and sand D 10YR 7/2 Crushed coquina shell and sand C 10YR 6/2 Crushed coquina shall and sand B 10YR 8/2 Silty-clay deposit A 10YR 6/4 Crushed coquina shell and sand

The earliest stratum documented was light yellowish brown (Munsell 10YR 6/4) coquina sand referred to as Stratum A. This stratum does not occur as a level deposit (Figure 1 Ib), but was concave with the top elevation varying from 178-193 cm and 195 cm BESD, respectively. The deposit extended to more than 223 cm BESD.

Stratum A is considered to represent fill material used to bridge the spandrel between the coquina-stone vaults of Rooms 21 and 22, similar to Stratum A deposits found in Test Unit 1. The only difference is the absence of tabby blocks, which were found in Test Unit 1 but not in Test Unit 5. This difference simply might be due to the discontinuation of excavation efforts in Test Unit 5 before the tabby blocks were encountered. In Test Unit 1, the tabby blocks occurred approximately 30 cm below the top of Stratum A- depth that was not achieved in Test Unit 5. On the other hand, the difference may have been intentional, with different approaches being used to stabilize the base of the terreplein deposits between the coquina-stone vaults. That Stratum A in Test Unit 5 was concave could be due to tamping activities used to consolidate and compress the deposits, which provided a firm foundation.

Above the coquina sand deposit is a layer of compacted, white (Munsell 10YR 8/2) silty-clay referred to as Stratum B. This deposit also is concave (Figure 1 Ib), with the top varying in elevation from 174-188 cm BESD. The base of this deposit is between 193-195 cm BESD. The nature of Stratum B is uncertain, although it probably was intended as some type of barrier from moisture and/or erosion.

The next three strata are a series coquina-sand deposits of varying coarseness and color that are considered to represent fill deposits, which elevated the terreplein surface to an angle where the coquina-stone vault becomes less acute. Strata C and D occur between 175-188 cm BESD and level the depression observed in earlier deposits. Strata E is a pale brown (Munsell I 0YR 6/3), coarse coquina sand and includes more shell. It extends from 165-175 cm BESD with a slight rise toward the coquina-stone vault (Figure 1 Ib).

The terreplein's depositional history becomes more fragmented above Stratum E. The next 11 strata (Table 2) are discontinuous layers of fill deposits, identified only on one or two profile walls. For example, the earliest of these deposits (Stratum DD, EE, and GG) are found only on the south profile wall (Figure 12c), but not the east (Figure 12b). Stratum DD is a thin, 1-cm, black band of silty-sand with considerable organic debris. Stratum EE is a 4 cm band of light brownish gray (Munsell 10YR 6/2) tabby. Stratum GG is a pocket of gray (10YR 5/1) crushed coquina sand. Conversely, Strata L, M, and I are found only on the east profile. These three strata consist of 8-10 cm thick bands of coquina sand with some tabby rubble (Strata I), tabby lens with numerous inclusions of shell fragments and construction rubble (Stratum L), or a sandy-silt earthen lens (Stratum M). Other strata (e.g., Strata F and G) that consist primarily of coquina sand are found on both profiles. The elevation of this fragmented zone of 11 strata extends from 133-141 cm BESD at the top to a depth of 165 cm BESD at the base.

What this fragmented zone of 11 strata represents is probably an interval when the terreplein surface was undergoing numerous changes over a short period of time. These strata probably reflect the necessity of constructing a functional terreplein that extended above the casemates prior to the War of Jenkins' Ear and General James Oglethorpe's siege of St. Augustine in 1740. As such, whatever material was readily available was used. The result of this activity was the formation of a heterogeneous zone of different soil types.

It is not until Stratum N that the depositional sequence in Test Unit 5 becomes continuous again (i.e., found in all profiles). The stratum is the lower deposit to a series of three poorly preserved tabby bands (Strata N, O, and Q) that constitute the next 45-50 cm of terreplein deposits. Stratum N is a very friable, light brownish-gray (Munsell 10YR 6/3) tabby. The stratum extends from 118 cm BESD at the top to a depth of between 133-141 cm BESD at the base. Within the tabby was a series of irregularly shaped tabby blocks that averaged approximately 25 cm per side by 10 cm thick. These blocks appear to be randomly placed, as if used as filler in the tabby matrix. The tabby appears to contain some deposits that are of a sandy-silt composition. This resulted in the tabby matrix being very crumbly and poorly preserved.

Above Stratum N was another poorly preserved tabby lens, which is Stratum O. Stratum O had the same characteristics as Stratum N: it was a friable, light brownish gray (Munsell 10YR 6/3) tabby. A distinct boundary, which was flat and flaked off during excavation, separates the two strata. The only distinction between the two strata was the presence of large, horizontally laid tabby blocks in Stratum O. These blocks measured more than 30 cm in diameter by approximately 15 cm thick and formed a pavement-like surface (Figure 1 Ia; Figure 13). The stratum extended from 102-106 cm BESD at the top tol 18 cm BESD where the surface of Stratum N was encountered.

The upper layer in this sequence of tabby is Stratum Q, but it is not the next stratum
documented for Test Unit 5. That distinction goes to Stratum P-a pocket of light gray (Munsell O1YR 7/2) consolidated tabby found in the northeast corner of the test unit (Figure 1 Ib). This deposit is 13 cm thick and extends from 108-121 cm BESD. Unlike other tabby deposits that contain construction debris (e.g. tabby or coquina-stone blocks or chunks), Stratum P did not contain any of this material. This suggests that the deposit may represent an area where a batch of tabby was dumped onto the terreplein surface prior to the pouring of Stratum Q.

The third and final layer in this thick zone of tabby is Stratum Q-a very pale brown
(Munsell 10YR 7/3) tabby with a coquina shell base. This deposit does contain some tabby and coquina-stone blocks and chunks, especially along the eastern portion of the test unit, but not to the degree of Strata N and 0. Stratum Q is from 26-30 cm thick, extending from 90-95 cm BESD to 121 cm BESD. Unlike the other two tabby surfaces, Stratum Q is not continuous along the entire east face of Test Unit 5. In the southeast corner, the stratum has been interrupted by Stratum S-a pocket of yellow (Munsell 10YR 8/6) crushed coquina sand with some coquinastone chunks (Figure 1 Ib). This pocket appears to have filled a depression in the tabby surfacea possible consequence of construction or use--over which a thinner deposit of Stratum Q can be found. Stratum S is not found on the south profile wall of the test unit, indicating that the depression is to the east of Test Unit 5.

Above the thick tabby deposits that constitute Strata N, 0, and Q was a series of five
alternating floor and fill layers (Strata R, S, T, U, and V) of varying thicknesses (Figure 1 lb and 1 Ic), which represent a discrete zone. The first of these was a yellow (Munsell 10YR 8/6) layer of coquina shell and sand. Referred to as Stratum R, this deposit is about 8 cm thick, with the top ranging from 85-90 cm BESD and the base from 90-95 cm BESD. Most likely, this deposit represents fill used to elevate the terreplein. The layer is compacted, but there is no evidence of it having been cemented with tabby.

Stratum T is a thin band of dark grayish brown (Munsell 10YR 4/2) tabby with tabby and coquina-stone rubble. This stratum ranges from 81-84 cm BESD at the top and 84-88 cm at the base, for a thickness of 3-4 cm. The tabby does appear to have some sandy-loam incorporated into its matrix that may have contributed to its condition (i.e., a poorly preserved and friable surface).

Atop the thin, friable tabby surface was another layer of fill deposits. This layer, referred to as Stratum U, consists of very pale brown (Munsell 10YR 8/4) coquina shell and sand. The layer is approximately 12 cm thick, extending from 70-81 cm and 84 cm BESD. This deposit is similar to Stratum C1 and Stratum C2 in Test Unit 1, which were pale yellow coquina sand that appear to be partially cemented. The only difference is that Stratum U does not seems to have been cemented.

Stratum U was followed by Stratum V-a thin grayish brown (Munsell 10 YR 5/2) tabby with coquina-stone and tabby rubble. The deposit is between 3-4 cm thick, extending from 67-70 cm BESD. The tabby does appear to have some soil incorporated into its matrix and is similar to Stratum T, a poorly preserved and friable layer.

The next stratum was a thin layer of very pale brown (Munsell 10YR 8/4) coquina sand.
Unlike previous coquina sand fill deposits, which averaged 10 cm thick, Stratum W was only 2 cm thick, extending from 65-67 cm BESD. This fill deposit separates the previous unconsolidated fill

and floor layers from a mass of consolidated (i.e., hard packed) and fused tabby layers. As such, Stratum W may have served to form a base to even or level the terreplein.

Following Stratum W is a series of tabby-coquina surfaces: Strata X1 to X4. These surfaces varied in thickness from 2-5 cm and extended from 51-65 cm BESD. The primary difference between these strata is the inclusions within this gray (Munsell color 10YR 6/1 or 5/1) tabby zone: Stratum X1 had charcoal flecks throughout, Stratum X2 had coquina-stone inclusions, Stratum X3 had both tabby and coquina-stone inclusions, and Stratum X4 had no inclusions. Except for a thin (1 cm) lens of coquina sand (Stratum X5) between Strata X1 and X2 along the northeast portion of Test Unit 5 (Figure 1 Ib), these tabby layers were laid directly atop each other. The hardness of Strata X1 to X4 is similar to the strata documented in Test Unit 4; however, there are substantial differences in color, composition, and inclusions between the two locations.

The next stratum is a very consolidated layer of fine, yellow (Munsell 10YR 8/6) coquina sand that could only be removed by using a pick-axe. Stratum Y extends from 45-51 cm BESD, forming an essential flat or level surface-similar to the previous stratum. It is likely that Stratum Y represents a floor, given that the coquina sand appears to have been cemented, although it may simply represent a layer of compacted fill.

Above Stratum Y were two deposits of tabby that comprise the final extant historic floors of the terreplein along the east curtain. The lower of these is Stratum Z-a thin (1-2 cm), gray (Munsell 10YR 5/2) tabby layer with few inclusions. Stratum Z extends from 43- 45 cm BESD, forming an essentially level surface. Stratum AA is a second gray (Munsell 10YR 5/2) tabby deposit, which has been divided into two subzones based on the presence of particular inclusions: Stratum AA1 contains coquina-stone rubble and Stratum AA2 contains coquina sand. Stratum AA extends from 35-43 cm BESD. The two subzones probably represent different variations in batches of tabby used to build a floor. Stratum AA1 is found along the northern half of Test Unit
5 (Figure 1 Ib), whereas Stratum AA2 is found along the southern half (Figure 1 lb and 1 Ic).

The final sequence of deposits found in Test Unit 5 are associated with 20th-century
activities at the fort. The two concrete layers (Stratum BB and Stratum CC) are similar to those documented in the other four test units excavated. The combined thickness of these two layers is 16 cm, with the strata extending from 16-35 cm BESD. Above the concrete slabs was the thin (1 cm) Laykold membrane.

Coquina-Stone Vault
The excavation of Test Unit 5 exposed a considerable portion of the coquina-stone vault over Room 21. The top of the vault was 152 cm BESD, which is notably deeper than that recorded in Test Units 1, 2, and 3.10 This would support the physical evidence in the courtyard area, in which the interior windows along the east wall are much lower than those found along the north, west, and south walls. The difference between the top of the vault over Room 5 as opposed to Room 7 and Room 16 is approximately 120 cm. Whether a coquina capstone, like that found over Room 7 and Rooml6, was present over Room 5 could not be determined with the extant data, it is unlikely. Test Unit 5 was placed over or very close to the crest of the vault and no evidence of a curb-like feature was exposed (Figure 1 Ib).

Unlike the coquina-stone vault over Room 16, which had been part of the terreplein surface, there is no evidence that the coquina-stone vault over Room 5 was similarly used. The vault over Room 5 did not evidence any wear marks on the coquina stone. Moreover, the 11 stratigraphic deposits that occur toward the top of the vault above Room 5 (i.e., Strata F to M, DD, EE, and GG) are diverse and are hypothesized to be related to rapid filling activities. Whatever soil and/or construction material was readily available, that material was used to construct an operable terreplein, which may have been a response to the 1740 siege of St. Augustine by General James Oglethorpe of Georgia. The first definable floor surface is Stratum N. This floor surface is approximately 30 cm above the top of the coquina-stone vault. This is in contrast to the first floor surface recorded in Test Unit 1, which was 60 cm below the top of the coquina-stone vault.

0o The top of the coquina-stone vault over Room 1 was not determined given problems in excavating Test Unit 4. It is assumed that the vault is similar to those over Room 7 and Room 16.

Material Culture Inventory

The excavation of the five test units resulted in the recovery of 177 artifacts and 132 faunal bone fragments. All of the artifacts and bone were from either Test Unit 1 (Table 3, Table 4) or Test Unit 5 (Table 5, Table 6); none were retrieved from Test Units 2, 3 or 4. The artifact types are similar to those documented by Lawson and Cornelison (2001) for terreplein deposits at the San Pedro and San Pablo Bastions, which have been described and referenced in detail. The descriptions and references need not be reiterated here.

The presence of artifacts in Test Unit 1 and Test Unit 5 is probably because these two units were not disturbed historically. Test Units 2, 3, and 4 had been part of the resurfacing efforts along portions of the south, west, and north terreplein during the Second Spanish Period, which effectively had removed earlier deposits down to the vault's estrado (Arana n.d.b:42-43). Consequently, what is reflected in the projects' artifactual inventory are items that became part of the terreplein prior to the Second Spanish Period-a fact that is substantiated in the recovered ceramics and kaolin pipe stems. Datable European ceramic fragments are Puebla Blue-on-white, Puebla Poly, and Reyware; all of which date during the mid-18th century (Deagan 1987). The pipe stems had an aperture diameter of 5/64th of an inch (or 17 mm), which generally ranges from 1710 to 1750 (Hume 1991:298).

It is evident from Tables 3, 4, 5, and 6 that both test units essentially share the same
assortment of artifacts, although the proportions are not comparable in some instances. The two notable differences are the presence of pipe stems in Test Unit 1 (Table 3) and the large quantity of roofing tile fragments in Test Unit 5 (Table 5). These differences are probably the result of the period in which that side of the terreplein was constructed. The terreplein on the east side of the fort, where Test Unit 5 was located, was constructed in the late 1730s. This is in contrast to the terreplein on the north side of the fort, where Test Unit 1 was located, which was constructed in the 1750s. Pipe stems are present in St. Augustine from the 1600s onwards, but they do not become a common artifact type until after 1740 when commerce is firmly established between the

Table 3. Artifact Types Recovered from Test Unit 1 According to Stratigraphic Level.

Artifact Strata Type H F E D C B A Total Native American Ceramics
St. Johns Plain
St. Johns Decorated 1 7* 8 St. Johns UID 3* 3 San Marcos Plain - 1* 1 San Marcos Decorated 3* - 3 San Marcos UID 1 4* 5 Abo Discard -1 - 1 European Ceramics
Puebla Blue-on-white -1 1 2 Puebla Poly - 1 1 Reyware -1 1 Olive Jar 2 - - - 2 Unglazed CEW -1* 4* 5 Glazed CEW - - 1 1 Porcelain 1 - 1 Glassware
Goblet fragment -- 1 1 Bottle fragment 1 1 - 2 Metal Items
Iron Nails 6 3 1 5 15 Iron Spikes - - 1 2 3 UID Iron Objects -1 1 2 UID Iron fragments - - 2 2 Armaments
Gunflint (gray chert) -- - 1 1 Lead shot - 1 1 1 3 Lead sprue - - - Miscellaneous Items
Pipe Stems - 2 1 5 8 Worked Wood/Bone 1 - - 1 2 Tile fragments 1 1 1 3 6 Total 3 9 7 5 0 41 0 65
* Some of the artifacts in this class had tabby adhering to the surface, suggesting that the artifact may have been included within the tabby mixture instead of trash left on the terreplein surface.

Table 4. Faunal Remains from Test Unit 1 According to Stratigraphic Level

Taxon Strata H F E D C B A Total
Mammalia (mammals)
Artiodactyla - - 5 5 Pig (Sus scrofa) - 5 4 -9 Oxen (Bos taurus) 1 - 1 2 4 cf. Goat (Capra) - - UID mammal - 2 3 3 8 Osteichthys (bony fish)
Catfish (Ariopisfelis) - - - Sea Bass (Centroporistis sp.) - - - UID fish - 2 4 2 8 Chondrichthyes (sharks) - - 1 - 1 UID Bone 1 2 4 14 21 Total 1 1 4 15 5 30 0 56

Table 5. Artifact Types Recovered from Test Unit 5 According to Stratigraphic Level.

Artifact Strata Type AA-X W V T R Q O N L IG-A Total Native American Ceramics
St. Johns varieties - 1 1* 4 1 - - 7 San Marcos varieties - 9* 1 - 2 3 - 15 Abo Discard - 2 - 1* -3 European Ceramics
Pueblo Blue-on-white - 1 - - 1 Olive Jar - - - 1* - - 1
Unglazed CEW 1* 6* 1* 1* 9 Glazed CEW - - 1 1* 2 Glassware
Bottle fragment 1 - - - - 1 Metal Items
Nail fragments 2 1 3 3 4 -1 14 UID Iron objects 1 - - 3 - 4 UID Iron fragments 2 1 - 10 1 -- 1 15 Armaments
Lead sprue - - 1 Miscellaneous Items
Pipe stems - .- 0 Worked Wood/Bone - 1 1 Tile fragments 3* - 2 1* 4 12* 16* - 38 Total 9 1 10 4 3 18 20 14 20 8 2 3 112
Some of the artifacts in this class had tabby adhering to the surface, suggesting that the artifact may have been included within the tabby mixture instead of trash left on the terreplein surface.

Table 6. Faunal Remains from Test Unit 5 According to Stratigraphic Level

Taxon Strata AA-X W V T R 0 O N L J I G-A Total
Mammalia (mammals)
Artiodactyla 1 1 4 1 7 Pig (Sus scrofa) 1 1 Oxen (Bos taurus) - 1 2 1 2 6 cf. Goat 1 I 1 UID mammal - 4 - 1 8 - 1 1 15 Aves (bird) - - - 1 1Osteichthys (bony fish)
Catfish (Ariopisfelis) - - - I 1 Sea Bass (Centroporistis sp) 1 1 UID fish - - -- - 5 1 10 Chondrichthes (sharks) 1* 1 UID Bone 4 3 3 2 4 - 1 15 32

Total 5 8 3 4 23 1 1 27 4 76
* The shark remain was a fossilized tooth.

city and English settlements along the Atlantic seaboard. The presence of roofing tile primarily in Test Unit 5 could correspond to the state of the governor's quarters and arms room, which were in ruins by 1737, according to Royal Engineer Antonio de Arredondo (Arana and Manucy 1977:34). This material, or another public structure that was in a state of disrepair, could have supplied a durable fill material to be included into the tabby matrix. Of the 38 roofing tile fragments recovered from Test Unit 5, 87 percent came from Strata L, N, O, and Q (the tabby layers), and many of these fragments had tabby adhering to their surface.

The majority of artifacts and bone fragments are from the tabby floors within both units, which are considered to represent the original 18th-century terreplein surfaces. In Test Unit 1, 63 percent of the 65 artifact fragments and 54 percent of the 56 bone fragments are from Strata B, which is composed of a three friable tabby layers. In Test Unit 5, 65 percent of the 112 artifact fragments are from Strata L, N, 0, and Q, all of which are tabby layers with various types and sizes of inclusions including tabby blocks. The exception to this pattern is the bone count for Test Unit 5, in which only 37 percent of the 76 fragments recovered were from tabby deposits. The majority of bone fragments recovered in Test Unit 5 were from Strata I (Table 6), a layer of crushed coquina shell and sand with some tabby fragments that occurs below the tabby surfaces. This stratum is one of 11 that was probably hastily formed to ready this side of the terreplein for action against vessels entering the harbor.

While an unknown percentage of the cultural material (i.e., artifacts and animal bone) can be considered primary refuse (material discarded at its origin of use [Schiffer 1976:30]), not all of the artifacts are associated with this type of activity. Some of the items were either accidental or purposeful inclusions put into the tabby mixture prior to pouring. This is especially the case with many of the Native American ceramics, coarse eathernware (CEW), and tile fragments that had tabby solution adhering to the surface of the item (Tables 3 and 5). Other objects were incorporated into the fill used to elevate the terreplein surface (e.g., Strata C in Test Unit 1). Thus, any interpretations related to activities on the gun deck, as inferred from the artifactual or faunal assemblages, are speculative.


The five test units excavated provide a wealth of information for interpreting the stratigraphic history of the terreplein, as well as the composition of the coquina-stone vaults that supported these deposits. This information is critical in evaluating both Manucy's (1939) and Wild's (1991) conclusions regarding the composition of the historic terreplein, especially in light of NPS plans to stabilize this feature of Castillo.

Evident from this investigation is that the composition of the terreplein is diverse, with numerous strata overlaying a system of bomb vaults of varying elevations and configurations. This finding supports Wild's (1991) conclusion that the terreplein is composed of various stratigraphic deposits and that the estrados to the casemates were closer to the present surface than originally indicated by Manucy (1939). In defense of Manucy, however, Test Unit 5 of his study did reveal that the top of the casemate was only 6 inches below the tapia surface.

Most of the strata documented are unique to a particular area of the gun deck, although a few deposits are found in two or more test units. This suggests that the development of the terreplein was not uniform throughout its entire surface area, but fluctuated according to various criteria. These criteria would include, but are not limited to, technological abilities, material availability, time limitations, and period of construction. None of these criteria is mutually exclusive, however.

There is some repetition of deposits across the terreplein surface, however. This is
particularly evident for the 20th-century concrete layers and Laykold membrane, which are found in all test units. Other stratigraphic deposits that are found in two or more test units include: 1) the hard, white tabby strata found in Test Units 2, 3, and 4; 2) the presence of a coquina-sand fill (Strata A) at the base of the historic terreplein deposits in Test Unit 1 and Test Unit 5; and 3) the presence of a thick tabby zone with inclusions of large tabby blocks (Stratum B3 in Test Unit 1 and Strata N and O in Test Unit 5).

The most obvious explanation for these similarities is that the deposits were laid down at
approximately the same time period. As such, comparable materials would have been used in the terreplein's formation. This may have been the case for Test Units 2, 3, and 4 over Rooms 1 and 7, in which a hard, white tabby matrix was encountered. Historical documents from the Second Spanish Period indicate that portions of the terreplein over the west, south, and sections of the north curtains were resurfaced with tabby from the top of the coquina-stone vaults (i.e., the estrados) to the surface of the then gun deck. Based on the results from Test Unit 4, these deposits were laid down as a series of sequentially poured tabby lenses that measured from 8 cm to 13 cm thick. No soil deposits or cultural fill were found between these lenses, indicating that they were poured in quick succession. Moreover, the results from Test Unit 2 and Test Unit 3 indicate that these deposits extended above the coquina-stone vault, which includes the capstone. No mention is made of portions of the terreplein along the east curtain having been resurfaced during the Second Spanish Period (Arana n.d.a).

The explanation that the deposits were formed at approximately the same time period is not the case, however, for similar deposits that were documented over Room 5 and Room 16. A lacuna of at least 15 years separates the construction of the bomb vaults and associated terreplein along the east curtain as opposed to that of the north curtain. The similarities between these deposits are most likely related to two factors: 1) following known guidelines in the construction of fortifications, and 2) the availability of raw materials. The construction of Castillo de San Marcos is known to have followed the military specifications of its time (Arana and Manucy 1977; Arana n.d.a; Sastre 2000). The coarse coquina sand was probably the material of choice to initially elevate and level the area between the coquina-stone vaults. Likewise, the series of tabby surfaces were used to form what was hoped to be a waterproof surface. The stability of the friable tabby was reinforced by the inclusions of large tabby blocks that had been salvaged from demolished structures built as part of the original fort in the late 17th century.

The differences between Test Unit 1 and Test Unit 5 are of particular significance, for these show that although certain conventions were followed prior to the Second Spanish Period, the

modus operandi was one of eclecticism. Whatever materials were available at the time were used in the maintenance and repair of the terreplein. In some cases, the material could form a 10 cm to 15 cm thick deposit, whereas in other examples the deposit was only 2 cm to 5 cm thick. Eventually, three or more feet of floor and fill deposits were added to the original coquina-sand base used to elevate and level the spandrels between the coquina-stone vaults.

The coquina-stone vaults show some peculiar characteristics, documented in earlier
investigations of the gun deck, but not understood (Manucy 1939). This has resulted in some erroneous interpretations and illustrations of the Castillo's structural composition (Figure 2). In particular is the presence of a capstone that was part of the coquina-stone vault. This feature was found atop Room 7 and Room 16. Based on the extant information, three details distinguish capstones: 1) they are wide, measuring more than 4 meters; 2) they extend the distance between the courtyard wall and the curtain; and 3) they are not centered atop the coquina-stone vault, but are offset to the right (Figure 6). Just how the capstone articulates with the right side of the coquina-stone vault was not determined in this investigation. The presence of capstones is assumed to have been part of the later rebuilding efforts for the casemates along the north, west, and south curtains that were implemented between 1752 and 1756. Whether or not capstones were part of rebuilding efforts for the east curtain in the late 1730s was not determined, although based on the curvature of the exposed coquina-stone vault the capstone is thought to be absent.

The other aspect documented was the difference in elevation between the top of the coquinastone vault found in Test Unit 5 from that found in Test Units 1, 2, and 3. The vaults for rooms along the east curtain (e.g., Room 21) are approximately 1.2 m lower than the vaults along the north and west curtain (Room 7 and Room 16). This disparity reflects the different eras in which the casemates were rebuilt and the conditions under which they were built. Casemates along the east curtain were constructed during a period when hostilities between Spain and England were simmering, which culminated in the War of Jenkins' Ear and General James Oglethorpe's siege of St. Augustine in 1740. In contrast, the 1750s were a period of relative tranquility on the frontier, and, probably, more time could be allotted to constructing casemates according to the

specifications of 18th-century military fortifications.

This discrepancy in construction dates for the casemates also might account for the
differences in the stratigraphic location of both tile and tabby block fragments. These fragments were found in the lower tabby surfaces of Test Unit 5 (i.e., Stratum L, N, and 0) and were incorporated into the floor as part of the poured tabby solution. Conversely, in Test Unit 1, tile is essentially absent and the majority of tabby blocks were used to form a platform near the deepest portion of the spandrel over which additional layers of fill (Stratum A) were poured. For the east side of the terreplein, the rubble (including roofing tile) could have originated from the detached governor's quarters and arms room that were in ruins by 1737, according to Royal Engineer Antonio de Arredondo (Arana and Manucy 1977:34). The rubble material for the north side of the terreplein could have come from the 17th-century tabby vaults, which were demolished in favor of coquina-stone bomb vaults.

In conclusion, the investigations undertaken by the City of St. Augustine Archaeology
Program on that section of the terreplein over the coquina-stone bomb vaults has clarified the issues raised by Wild (1991). This clarification was essential before the NPS began stabilization of the terreplein, which would attempt to stem the problem of leakage into the casemates by removing a section of the extant terreplein and replacing it with a waterproof membrane.

This investigation reveals that although the historic terreplein deposits extend more than 3 ft deep, it is the elevation of the casemates that should determine the course of action. The capstones over Room 7 and Room 16 vary from 7-12 inches (18 to 32 cm) below the present surface, depending upon its location between the courtyard and curtain walls. Moreover, the coquina capstone along the curtain wall has been impregnated with concrete from the 1939 resurfacing. As such, any attempts to remove the 1939 concrete surface would have an adverse impact on the historic fabric of bomb vaults, especially along the curtain wall. The only deposit that can be removed without damaging the historic fabric of the coquina-stone bomb vaults is the 1960 concrete surface. It is recommended that the NPS limit its excavation efforts to this deposit.

References Cited

1997 Scope of Work Exploratory Excavations of Casemate Terreplein. Contract for
Archaeological Excavation, Castillo De San Marcos National Monument, Florida.
MS on file, City of St. Augustine, Planning and Building Department, St. Augustine.

Arana, Luis R.
n.d.a The Honor of the Spanish Arms, 1784-1821: Construction and Repairs of the
Castillo de San Marcos during the Second Spanish Period. MS on file, Castillo de
San Marcos National Monument, St. Augustine.

n.d.b The Endurance of Castillo De San Marcos 1668-1763: Construction and Repairs
during the First Spanish Period. Unfinished manuscript, Castillo de San Marcos,
National Monument, St. Augustine.

1986 Fort Marion in Civil War Times. El Escribano 23:48-65.

Arana, Luis R., and Albert Manucy
1977 The Building of the Castillo de San Marcos, Eastern National Park & Monument
Association for Castillo de San Marcos National Monument.

Arnade, Charles
1959 The Siege of St. Augustine in 1702. University of Florida Press.

Bearss, Edwin C., and John C. Paige
1983 Historic Structure Report for Castillo de San Marcos National Monument, St.
Johns County, Florida. Denver Service Center, NPS, U. S. Department of
Interior, Denver, Colorado.

Bushnell, Amy
1981 The King's Coffer, Proprietors of the Spanish Florida Treasury, 1565-1702.
University Presses of Florida, Gainesville.

1983 The Noble and Loyal City, 1565 1668. In The Oldest City, edited by Jean
Parker Waterbury, pp. 68-83. St. Augustine Historical Society.

Connor, Jeanette T.
1929 Nine Old Wooden Forts. Florida Historical Quarterly 4:103-111, 171-180.

Crane, Verne
1959 The Southern Frontierl670 to 1732. W. W. Horton, New York.

Deagan, Kathleen
1980 Excavations at the Castillo De San Marcos. MS on file, National Park Service
Southeast Archeological Center, Tallahassee, Florida.

1987 Artifacts of the Spanish Colonies of Florida and the Caribbean, 1500-1800,
Volume 1: Ceramics, Glassware and Beads. Smithsonian Institution Press,
Washington, D.C.

Fischer, George
1979 Draft Archaeological Research Prospectus, Castillo de San Marcos National
Monument. MS on file, Southeast Archeological Center, Tallahassee, Florida.

Frazier, C. Craig, Randall Copeland, and Luis Arana
1986 Historic Structure Report, Volume 2. Castillo De San Marcos National
Monument, Stabilize Fort Pkg 116. MS on file, Castillo de San Marcos National
Monument, St. Augustine.

Halbirt, Carl D.
1997 Of Earth, Tabby, Brick, and Asphalt: The Archaeology of St. Augustine's Historic
St. George Street. ElEscribano 34:70-97.

Hume, Ivor Noel
1991 A Guide to Artifacts of Colonial America. Vintage Books, Random House, N.Y.

Kelly, A. R.
1940 Review of Terreplein Construction Fort Marion National Monument.
Memorandum, Branch of Historic Sites Technical Comment, NPS, Washington,

Manucy, Albert
1939 Terreplein Construction Fort Marion National Monument: Notations on Its
Original Character. MS on file, Castillo De San Marcos National Monument, St.

1940 Memorandum for the Superintendent. MS on file, Castillo De San Marcos
National Monument, St. Augustine.

Sastre, Cecile-Marie
2000 Castillo de San Marcos National Monument, History of the Cultural Landscape for
the Cultural Landscape Report. MS on file, Castillo de San Marcos National
Monument, St. Augustine.

Sluiter, Engel
1985 The Florida Situado: Quantifying the First Eighty Years, 1571-1651. Research
Publications of the P. K. Yonge Library of Florida History, Number 1. University
of Florida, Gainesville.

Wild, Ken
1991 Trip Report on Investigations at Castillo de San Marcos National Monument and
Fort Matanzas 6/3-7/2/91. Southeast Archeological Center, National Park
Service, Tallahassee.


Figure 1. Location of 1939 and recent excavations on the terreplein.

mm. 1

Mancuy (1939)

city (2001)

1 Scale: 100 feet 1


Figure 2. NPS illustration of the sequence of pours used to construct the terreplein.


tj s h Roo



Figure. 3. Plan and profile perspectives of Test Unit 1.


Figure 4. Photograph of tabby blocks at the base of terreplein deposits in Test Unit 1.


Figure 5. Photograph of grooves in coquina-stone vault about Room 16.


Figure 6. Profile of Room 16 showing relationship of capstone to coquina-stone vault.


Figure 7. Photograph of grooves and cuts along edge of capstone above Room 16.


Figure 8. Photograph of 1939 concrete wall trench above Room 16.


Figure 9. Plan and profile perspectives of Test Units 2 and 3 above Room 7.


Figure 10. Plan and profile perspectives of Test Unit 4.


Figure 11. Plan and profile perspectives of Test Unit 5.


Figure 12. Photograph of soil stratigraphy in Test Unit 5.


Figure 13. Photograph of tabby blocks in Stratum Q of Test Unit 5.

rl l \b 138'-10" 3- }



.s: IO


OM-9 3 -0

110- -2"


SEP-15-97 MON 11:49 NPS-DSO-ARCHITECTURE FAX NO. 3039876608 P, 02/09

. .'. 42 u09

R 7" AL ,nc
Mfiic(6 WT PIT) "o "o '-"- e
2 5 T 9AU0 01 29 22PT3O244. hkso o m P
SEP-15-97 MON 2:52 PM 3039B76608 P. 2

eay1W'v: ~?toI:

r'~ ;~'# Z, ~;H,.,E
;~w ~-~44,E4

~ I 9c~~FA ; '~&DQ4 ~
~:8 l~ r" ii~l" I-'
Bi~r A7 Bi ;
A~ -o./ -4~ / t~~'w 'i _____I e ~ il~l~~"r rA-p4~A

-";: ~1' ;i ,
ai.II a~z~ ~ O, a .~b~r Aa" ~
-'" ~Q
4'p72 l& ~lli~1 DIAl~xi~ ICl ji~i l l
:i; p. ". ; ~-~.~.-:

KA6~ L~ .AI
4~.f4 ~-~ -i ~#oq.

; IIIIll

w-~ ~

~ ~ -~

I4~ / T~4)~Jt29S) ~A


~ ~



8~, ~

~ 4r~

4 t


/~~~~~ I/Cl &.l);'<4 o LA (-UJ/ #k fa- Art
.: ,u a:: ::Cr?> e I::: ea w b e/(./A,

'i Ii : : I I-::
e- V65 +P
-a we. 6

.w ('t d*r~o .74*a 1a

0, d q -f" 0 "
re h at A ? 7 sd

r d e' frrT

~~y)~2 w/tit'~Yjy/ ) yre4d w
o +/>* 7* 7- s /"T

0 yQ~c.,a -r e C o-s~ssuAJt/*w
71 ~:W -: :I yi'.~i~ ::i: )44:77

crA- A Ef b <,, +Y +' /t J

7- --se e-I 4M rI frf vA9 1 m ', o As del / ekA -fas y er
71, sk ar me Ii + arg' ct so < <.r -- e ;-r cle: wed .. De r Fare

I w- F ac swe T --r m fi





-Pm ~ ~

~FtJ a

oil -~ -~ --- (i~y~) ~
---------------~------,~ -- -~---~--~ / IA,

~ ~ p / ~ "a

. ..... ....... .~ ~o~,..P~ d~. -~~ CnnH ,

5 -r- s a -Ik s 7 ,

a .... > + ...... s > we.a 4 e--e J- --. -a aref ./ i;--- 7 4 ---- --a--s -t
...., .../. .

.. ....... .......I~..d ~as~s cI Lh~ c o.l c1 I
.t.. ... --. ...s ... .. .. ....

iAe. Au'A 79t- N e-e .

--r -u


+4- rFs ,- Y+

S0 r T A ~' 1 K)fr a 1~ i
l'n~vv~.J~% : )

S 4J I:r,- v q1 1

I j


1 .14

s ?

'1. ~V

Outline for Casa Terreplein Project

1. Introduction
a. Purpose of report
i. As the nation's only extant colonial period fort, the Castillo de Marcos is an invaluable cultural resource. A. As such, the documentation and preservations of the fort's characteristics are of parimount concern. B. To help understand the fort's resources..... ii. Use draft scope of work introduction (paraphrase) A. Four areas of concern to the NPS
(a) List these
b. NPS contracted with the City of St. Augustine to conduct archaeological
i. The City is one of a few municipalities in the US to have an ordinance
which mandates that archaeological investigation be implemented prior to any new ground-penetrating construction activities on either public or private properties
A. The intent of the ordinance (see Florida Anthropologist and La Punta report)
(a) The scope of the terreplein project is within the purview of the city's ordinance

2. Background
a. Existing interpretations of the Castillo's terreplein is based primarily on historical
documents with little corroborating archaeological information
i. This had to some potentially erroneous conclusions related to the
construction and development of the Castillo from the 1730s onward A. Examine Arana and Manucy (1977:44)-also NPS pamphlet b. Archaeological excavations at CASA are not new
i. Various projects have been conducted both with and outside the fort's confines within the NPS property (references) A. List some of the more significant project (Smith et. al; Deagan references) including recent bastion excavations ii. One of the earliest projects was Albert Manucy's archaeological
investigations of the terreplein (Figure-show location of test areas) A. Discuss Manucy's (1939) report and criticisms
(a) use tape transcripts
B. Manucy's report as basis for later interpretations
(a) 1986 draft HRS
iii. Subsequent archaeological investigations
A. In 1991 the NPS conducted a coring project on the terreplein
(a) no existing written report
(i) verbal information indicates (see draft scope) iv. Because plans are being made for resurfacing the fort's gun deck, it was

necessary to define the archaeological deposits that comprise the terreplein in accordance with section
A. The archaeological investigations undertaken by the City of St.
Augustine's Archaeology Program are intended to address some of the confusion that exists relevant to the terreplein's construction and composition.
(a) The investigations are not intended to be exhaustive, but to address specific archaeological questions that will allow for the design and replacement of the terreplein surface.

3. Task Definitions
a. In the draft scope of work, the NPS defined six tasks that need to be undertaken
by the City in the archaeological investigation of the terreplein surface.
i. Each task defined the location of a test unit and explain the purpose for that unit
b. Only five of the six tasks were completed.
i. Task 5, the excavation of a test unit over Room 21 near the courtyard wall, was not undertaken given the heavy rains received during the winter of 1997/1998. Moreover, it was decided that because additional emphasis had been given to Task 1 (Rooms 15 and 16) this would replace Task 5. ii. The five completed tasks and their purposes are as follows. Figure shows the location of the five test units A. Task 1 (Rooms 15 and 16-Test Unit 1) B. Task 2 (Room 1-Test Unit 4) C. Tasks 3 and 4 (Room 7-Test Units 2 and 3) D. Task 6 (Room 21-Test Unit 5) c. Excavation Methods and Data Recovery
i. How were test units excavated
A. After the concrete surfaces had been removed by personnel from the NPS, the excavation of each test unit proceeded by hand using shovels and trowels.
(a) When necessary chisels and sledge hammers were used to break through the hard tabby surfaces (in particular, those encountered in test unit 4 over Room 1)
(b) As it turned out, the excavation of each test unit was able to follow the terreplein's cultural stratigraphy with each level representing a different type of soil deposit
(i) Including the concrete slabs, the number of levels varied in a test unit from 3 to 31 B. All deposits from the terreplein were screened through quarter-inch wire mesh to retrieve any artifactual material that may exist
(a) In particular, attention was given to recovering broken pieces of pottery, animal bone, metal objects, brick and tile fragments, gun flints and lead shot, and items of personnel adornment and use

(b) Not collected where the numerous fragments of broken tabby and coquina stone use the construction of the fort

4. Excavation Results
a. The five test units excavated provided a unique glimpse into the construction of
the fort's terreplein (gun deck) surface that spans approximately 260 years of use.
i. The information that emerged from the excavation was that the fort's gun deck varies considerably in the types and degree of material used in its construction
A. Of

ii. exists
A. In 1991 the NPS core project (no existing written report but verbal information)

C. 1986 draft Hrs D. Discuss Manucy's report and criticisms ii.