Introduction

The site originally known as the “Baie des Saintes Wreck” is located in the middle of Saintes Bay (south of Guadeloupe, FWI) in 25 m of water in an area of sandy bottom. It was discovered by Claude Edouard in 1995 and excavated in the 1990s by a team of local divers without archaeological oversight.

The first archaeological assessment, conducted by Michel L’Hour and Jean-Luc Massy during a DRASSM project in 2001, dated the site prior to 1840 based on “Creil and Montereau” stamps on two ceramic artifacts: a saucer and an octagonal plate (L’Hour and Massy 2002). In addition, copper sheathing and cast-iron ballast were observed. Additional archival research was conducted and informed a hypothesis that the wreck represented the French schooner Anémone (Guibert 2013). The vessel was built in Bayonne, France, in 1823 and dispatched to the West Indies after its involvement in that year’s Spanish war. In Guadeloupe, the vessel was used as a tender ship, actively engaged in the struggle against the ongoing slave trade. Though the slave trade had been officially forbidden in the French colonies since 1817, the illegal trade continued until the 1830s. Anémone reportedly sank during a 1824 hurricane after sailing to Saintes Bay for protection (Jacob 1825; Lacour 1860:345). All of the crew and officers were killed during the sinking event.

Fieldwork and Methodology

Looted in the 1990s, the site had not been revisited until a doctoral dissertation proposed to further investigate and identify the site based on archival research (Guibert 2013). The first field campaigns in 2015 and 2016 relocated the site and determined its layout. Results confirmed the identification of the shipwreck and led to the development of a multiyear excavation program. The project goals focused on the shipbuilding and material culture of a naval schooner, the only known surviving example from a series of six ships built according to the Anémone design plan. The site represented an opportunity to compare the remains of the shipwreck with Jean Boudriot’s Jacinthe design plan (Boudriot 1989) and archival data including the ship’s plan from 1823 (Delamorinière 1823) (Fig. 1). The methodology included extensive excavation, hull structure and artifact identification and analysis, wood sampling and dendrochronology, and photogrammetry. The project team included specialists for each of those tasks.

Fig. 1
figure 1

Schooner Anémone plan type, 1823 (Delamorinière 1823).

Full size image

Field Campaigns: 2015–2019

2015–2016

The site was relocated in 2015 based on an estimate of its position in relation to known landmarks. The site was surveyed, and test trenches were excavated (measuring either 2 or 3 m2). A carronade found at the site supported the wreck’s potential identification as Anémone, which was reportedly equipped with two carronades of the same type (Boudriot 1989:26; Guibert 2016). The second campaign in 2016 yielded more precise information about the site’s layout. Two test trenches measuring 6 m2 were excavated: one in what was believed to be the forward area (Test Trench 1), the second in the assumed aft area (Test Trench 2) (neither area is shown on the site plan, Figure 2). After further analysis, it became evident that the trenches had exposed the starboard side of the hull (Guibert, Bigot, Lachèvre et al. 2017). Results of the material culture and ship-construction analyses during those first campaigns were consistent with the story of Anémone’s loss (Guibert 2016; Guibert, Bigot, Lachèvre et al. 2017) and helped to confirm the identification of the wreck as the French schooner Anémone.

Fig. 2
figure 2

General site plan of the Saintes Bay Wreck identified as Anémone. (Figure by Andrea Poletto, Hélène Botcazou, and Jean-Sébastien Guibert, 2019.)

Full size image

2017–2019

The field campaigns of 2017–2019 focused on excavating three areas of the starboard side of the hull: stern, bow, and amidships (Fig. 2). Test Trenches 1 and 2 were extended to 3 × 3 m units (9 m2), and additional units were opened to acquire a better view of the hull structure. Archaeologists documented the overall shipbuilding characteristics as well as particular structures and objects, such as mast steps, a bilge pump, a water drain, and ballast. A second goal was to integrate the material culture analysis with each hull-construction feature.

The 2017 campaign focused on the aft side of the hull. Two units (Units 3 and 4), each measuring 3 × 3 m, were opened and partly overlapped Test Trench 2. Unit 3 confirmed the identification of the stern area. The observed longitudinal hull components in this area include the keel, the stern curvature (the stern itself is missing), and the very degraded keelson. The planking is visible on both sides of the hull, but only a few ceiling planks are visible. The transverse hull components include 11 members, such as the filling frames and the station frames, placed according to a regular pattern (Guibert, Poletto, Bigot, Veyrat et al. 2017:39–49).

An object found close to Unit 3 and later identified as the rudder led to the excavation of Unit 4 nearby. The rudder, located 3 m away from the hull, measured 3.02 m long, 64 cm wide, and 18 cm thick. Its state of preservation was due to the copper sheathing. Two gudgeons were located in their proper place, but were broken, the upper part of the gear not protected by the sheathing was no longer extant. Such a hull component is rare to find preserved; its location and evidence of damage indicated that a violent shock occurred when the ship struck the seafloor (Guibert, Poletto, Bigot, Veyrat et al. 2017:50–58).

Drinking glasses, bottles, and ceramics found in the stern area are believed to be associated with the officers’ quarters. Among the bottles and glass, a plate of Montereau manufacture dating 1815–1825 indicates that the ship was a naval vessel. The plate’s decoration depicts the French defeat at Waterloo and the words of Cambrone: Le Général CAMBRONE sommé de se rendre répond\La Garde meurt et ne se rend pas! (General Cambrone, summoned to surrender, responds: The guard dies and does not surrender!). Other artifacts, such as personal weapons, were found and determined to be associated with the officers’ quarters (Guibert, Poletto, Bigot, Veyrat et al. 2017:68–70).

The 2018 campaign goal was to delineate the forward area of the ship to find the bow. Two 3 × 3 m units (Units 5 and 6) were opened and partly overlapped Test Trench 1 (Fig. 3). In Unit 5, excavations revealed that the bow had been destroyed, and the keel ends in this area. The foremast step has not been found. In Unit 6, the absence of large concretions and ballast permitted the team to dismantle part of the hull in order to analyze the underlying structural members and the water drain. Similar to the units excavated in 2017, this area of the hull contained the keel, outer-hull planking, and ceiling planking. The transversely oriented hull components are composed of 23 structural members following the same spacing pattern as those examined in 2017 (Guibert, Poletto, Bigot, Bianchimani et al. 2019:41–50).

Fig. 3
figure 3

Units 5 and 6 in 2018, showing the forward area under excavation. (Photo by Olivier Bianchimani, 2018.)

Full size image

Some of the artifacts examined in this area include a possible mast bar, usually used on deck for rigging, and a barrel pump, indicating a storage function. Parts of the units overlapped Test Trench 1 (2016), which explains the lack of artifacts in the area. Based on its presence in the various excavation units, the hull was sheathed with copper. A piece of copper sheathing was found in the bow area in 2019 and examined. It exhibits different draft marks than the others and represents the very end of the hull sheathing (Guibert, Botcazou et al. 2020:67–71).

The 2019 campaign focused on the central part of the hull remains, including the mainmast step and bilge pump. Three units were excavated, each measuring 3 × 3 m, forming a right angle. Units 7 and 8 were placed on the keel axis and Unit 9 in the upper part of the starboard remains. The shot locker was found close to the mainmast step, which was identified in spite of a large concretion covering the area. Although only the starboard side of the hull remains was preserved, the vessel was likely constructed with two bilge pumps. Due to the presence of pig-iron ballast and large concretions, examination of the lower-hull structure was limited. Instead, investigation focused on the keel and floor timbers in the bottom of the hull. Outer-hull planking and ceiling planks were examined in what remains of the upper part of the hull (Guibert, Botcazou et al. 2020:51–63).

Identified artifacts include several barrels, which indicated this area of the hull was used for storage, and pig-iron ballast. Artifacts associated with storage, but also with the ship’s surgeon, were found, including a demijohn, a lantern, ceramics, glass, and a syringe from the medicine chest.

Site Description

The hull remains are mainly preserved on the starboard side, with a total length of extant remains measuring ~17 m from stern to bow and ~4 m wide (Fig. 2). The only structural member that remains preserved on the port side of the hull is the garboard. Numerous planking and copper-sheathing nails were observed on the port side of the hull along the keel and in the sand adjacent to the hull. The presence of artifacts in the western part of the wreck supports the determination that the ship listed to the starboard side after settling on the seafloor. The wood has been preserved underneath the heavy materials (e.g., cannonball pit, pig-iron ballast, and carronades) that fell down or settled upon it. However, this material, while protecting the wood beneath it, created a challenge for archaeologists trying to examine and document the hull. The stern of the ship shows significant damage, likely encountered during the wrecking event, as shown by the keel’s state of preservation and the rudder with its broken hinges.

Hull Construction

The longitudinal structure of the wreck is composed of the keel, measuring 20 × 28 cm, and a false keel measuring 8 cm thick was observed in Unit 5. They are strongly fastened together by iron keel bolts, each measuring 60 cm in length. An upper false keel was discovered during the 2018 and 2019 field campaigns and had been observed at the stern of the wreck in 2017. This structural element measures 20 cm wide and 3–5 cm thick in Units 5 and 6, 14 cm thick in Unit 3, and is fastened to the upper surface of the keel. A segment of the keelson measuring 16 cm wide and 15 cm thick was observed in Units 3 and 6. Near the stern of the wreck, the rudder could be documented and measured due to the copper sheathing that originally covered the wood and preserved its original shape (Guibert, Bigot, Lachèvre 2017:50–58).

The transverse structural components are organized in a pattern of a predetermined coupling of two frames and two filling timbers. Each framing component is spaced 1 m apart. Chocks were used in the opening between the frame timbers to reinforce the framing components. The chocks were symmetrically grooved from one side of the master frame to the other and allow water to circulate laterally. The drain system is completed by limber holes carved in the floor timbers (Guibert, Poletto, Bigot, Veyrat et al. 2017:57). The master frame was not accessible due to the large concretions situated on top of the hull.

The planking on the port side of the ship is not well preserved, except for the first two planks. The garboard measures 22 cm wide and 5 cm thick, and is covered by copper sheathing. On the starboard side of the hull some planking at the waterline and toward the bottom of the hull was accessible for documentation in areas not covered by the pig-iron ballast and concretions. In 2018, 16 planks were documented in the bow of the wreck. They also measure 23 cm wide and 5 cm thick toward the lower hull, and measure 20 cm wide and 3.5 cm thick in the upper hull.

The ceiling planking is preserved in some areas, including 13 planks at the bow that measure 17–20 cm wide and 2.5 cm thick. Two of them are thicker (5 × 20 cm) and designed to reinforce the hull longitudinally (Guibert, Poletto, Bigot, Veyrat et al. 2017:43–44). The foremast step has not been identified, but the mainmast step was recorded during the 2019 field campaign. It is partially preserved on the starboard side, composed of three wooden elements on a transverse alignment. A fourth timber is placed on a longitudinal alignment. The frames are reinforced with a bigger chock between the floor timbers (Fig. 4) (Guibert, Botcazou et al. 2020:63).

Fig. 4
figure 4

Hypothetical reconstruction of the mast step and bilge pump. (Figure by Hélène Botcazou, 2020.)

Full size image

Wood analysis of the hull determined that 90% of the wood used for constructing the ship is oak, which correlates with the known usage of the French navy. Dendrochronological study concluded that the wood used in the hull’s construction dates largely from the period 1782–1820, with one sample dating 1822–1825 (Guibert, Botcazou et al. 2020:45).

In 2019, several meters of the keel were exposed for analysis of the hull’s copper sheathing. The sheathing is composed of rectangular sheets measuring 35 × 145 cm that overlay one another and are nailed at a 2.5–3.5 cm spacing. At the waterline, the copper sheathing is folded and attached to the hull planking with three lines of nails. In 2018, one sheet of copper included Roman numerals indicating the vessel’s draft marks (III––IIII––V) (Guibert, Botcazou et al. 2020:67–68).

Interior Features

The remains of a bilge pump were observed in the midships area of the wreck on the starboard side in Unit 8 (Fig. 4). An impression of the central tube, measuring 11 cm in diameter, was preserved inside a concretion located within a timber cribbing. This location also contains a cannonball pit, a barrel-storage area, and another area containing rigging elements and crew equipment. A bunker for storing objects associated with the galley was discovered near the bow, while the stern contained evidence of a living area apparently used by the ship’s master.

Site Interpretation

To date, the site is the sole example of a French naval schooner that has been extensively investigated and documented through archaeological analyses and archival research. As no other preserved examples of this type of vessel are known, archival research instead relied on Jean Boudriot’s (1989) La Jacinthe. The shipwreck is interpreted as a schooner built from the Anémone design plan, dated 12 February 1823 (Fig. 1), which was also armed with two 12 lb. carronades. Historical knowledge of the site is based on information about Anémone’s five sister ships (Boudriot 1989:26), including two archived design plans of Topaze and Émeraude dated 21 March 1823.

Anémone’s design plan, archival sources, the archaeological site plan, and a photogrammetry mosaic of the site were compared, and it was observed that the mosaic and the original design plan match perfectly, essentially confirming the identification of the site as Anémone. One observed difference was the slight collapse of the central part of the hull, undoubtedly due to the heavy pig-iron ballast and iron concretions. The differences between the original plan and the current condition of the wreck’s bow and stern can be attributed to site-formation processes.

As a trade ship, the schooner design originated in American and Caribbean shipyards. Its rigging, shape, and tonnage were adopted and adapted in European shipyards. From a functional point of view, Anémone was a military schooner commissioned as a customs ship in the French colonies. Material culture from the site dates to the beginning of the 19th century and supports interpretation of the vessel’s function as a naval schooner.

Material Culture: Preliminary Analysis

Analysis of the artifacts addressed three objectives: (1) to invalidate or confirm the identification of the wreck, (2) to define and describe the material culture associated with this type of vessel, and (3) to document the spatial organization aboard a French royal navy ship used as a customs ship in the Antilles.

During the 2015–2019 campaigns, a minimum number of 1,537 artifacts was recovered, representing 545 inventory numbers. The artifacts were recorded according to their location within the wreck (unit and location using the parameters x, y, and z) and classified by function. However, the site was heavily disturbed by looting and the anchoring of vessels. An additional 12 artifacts looted by an individual were studied by the team in 2016. A legal procedure by the French government to recover those artifacts is currently underway. Despite evidence of looting and other occasions of site disturbance, the team accomplished the three objectives enumerated above.

Material Culture Analysis

In the collection a distinction is made between features and objects related to the wreck’s hull construction and individual objects associated with the crew. The first category includes all artifacts related to the vessel’s operation, propulsion, and construction. Thus, 546 artifacts, corresponding to 35.52% of the collection, relate to this category and are subdivided into three subcategories: “tackle and rigging,” “architectural elements,” and “onboard equipment.” The second category, “crew-related artifacts,” is divided into six subcategories with 975 objects (63.96%) or fragments of objects. Finally, some artifacts in the collection (26 objects, or < 0.50%) cannot be identified or linked to a specific function.

Artifacts Related to Rigging and Hull Construction

Apparatus and rigging total 32 artifacts. This category includes a large number of copper-alloy pulley wheels, 62.5% of the total count. In 2019, many single or double wooden pulleys with remains of trunks were found, including the eight pulleys identified in the concretions in Unit 8.

Among the architectural elements, the group with the greatest number of examples is nails (Fig. 5d). This group is essentially composed of bordage or carvelle nails of different sizes used to attach the planking to the frames and represents 365 of the 546 artifacts in this functional group (66.85%). Shipboard equipment includes 92 artifacts, or 16.85% of this functional category, and is divided into seven subcategories.

Fig. 5
figure 5

Diagnostic artifacts from Anémone: (a) glass vial, (b) brass pan (1816–1822), (c) Louis XVIII 5-franc silver coin 1818, (d) bordage nail, (e) octagonal earthenware plate manufactured by Montereau (1815–1825), (f) Bordelaise-type bottle, (g) shell-edge plate (1815–1830), and (h) flint plate with pan and inscription: BRINGOL/PARIS/1812. (Figure by Franck Bigot, 2021.)

Full size image

Several draft marks were identified at the stern and the bow, and indicate the ship’s draft according to the load carried. Seven copper-alloy fragments, composed of six wires twisted in pairs and measuring 5 mm in diameter, are interpreted as part of a lightning protection device. A lead scupper collar and a ferrous scupper were observed on the site and represent the subcategory of “water drainage.” Indeed, a ship’s hull is never completely waterproof and requires valves to drain water from the hold. The scupper and scupper collar integrate this system of pipes and are responsible for the evacuation of water from the upper deck.

The excavation revealed 21 visible fragments of copper sheathing plates and 18 strips of lead of varying thicknesses and in various states of preservation. The plates are distributed throughout the lower portion of the wreck. The lead sheets are associated with repairs or caulking of the ship’s hull, with evidence of their attachment by square nails.

Personal Artifacts

The first category consists of artifacts related to artillery, small arms, and ammunition. Ammunition represents 68.06% of this group (minimum number of artifacts [MNA] = 669), which seems to correlate with the ship’s naval function. The artifacts documented on the site (e.g., 12 lb. carronade, pig-iron ballast, oven, kettles) support this identification. The category of “personal items” includes 31 objects. There are nine wooden knobs, circular in shape, and many are flat with one or five perforations cut into the back of the knob. Other examples are knobs with a perforated eyelet and convex face. Other artifacts include a single rectangular-shaped buckle of modest design constructed of metal that is unidentifiable due to poor preservation, and a watch rooster (an internal part of a watch). Surprisingly, no uniform buttons were observed. These artifacts reflect the fashion of the time and indicate the socioeconomic disparity within the crew.

Artifacts related to food consumption have an MNA of 193 and include plates and bowls of French and English earthenware, and numerous Rhenish stoneware and green-glazed earthenware pitchers (Saintonge), as well as fragments of French and English wine bottles (Fig. 5f). Finally, food was stored in large jars (biot) and drinks in numerous wooden barrels (MNA=8). A water pump, a barrel tap, and a demijohn complete the collection of artifacts associated with the transport of liquids.

Chronological Analysis

The material culture assemblage is essentially composed of objects dating from the first quarter of the 19th century. Indeed, this wreck is equipped with a galley stove known as “à la Kersaint” (PB2019-77). From the end of the 18th century, and especially during the first quarter of the 19th century, naval vessels were equipped with this type of stove (Colignon 1823).

The nails found to date represent various sizes. According to the typology proposed by McCarthy (2005:90), the nails appear to date to the 1820s. Similarly, the copper sheathing, of which large quantities were observed, dates to the first quarter of the 19th century.

The ceramics and glassware also date to this time period. For example, the shell-edge plate of English manufacture (Fig. 5g, PB2016-22) was very common during the period 1820–1850 (Métreau 2016:227). The French earthenware plates were manufactured by Montereau; one decorated example was produced between 1815 and 1825 (Fig. 5e, PB2017-76). For the glass artifacts, the temporal affiliation range is wider. Three different examples have been distinguished: a goblet (PB2017-65) decorated in a style that appeared around the 1800s and has been found in contexts from 1812 to 1814 (Jones and Smith 1985:41); a cylindrical flask (Fig. 5a, PB2019-86) with a thick, nearly flat base that appeared as early as the mid-18th century and was frequently used during the first quarter of the 19th century (Jones and Smith 1985:98); and the large, thick, black glass demijohn with a straight neck, cylindrical and wide body, and rounded shoulders (PB2019-77) that dates between 1810 and 1860 (Serra 2011:225).

Finally, the brass pans from the flintlock guns correspond to the year IX of the French Republican calendar (1800/1801) or to the 1816–1822 model (Fig. 5b, PB2017-53), while the flintlock side plate with pan bears the inscription: BRINGOL\PARIS\1812 (Fig. 5h, PB2016-12). In addition, the coins looted from the site date from the late 18th century to ca. 1818 (Fig. 5c, PB2016-01 to -08).

Spatial Analysis of the Material Culture

The study of Anémone’s material culture provides a glimpse into the functional spaces aboard the ship and thus highlights the hull’s division into three areas. The aft portion of the ship yielded numerous fragments of octagonal plates of fine earthenware, plain-stemmed wine glasses, cut-sided goblets, square bottles, and a table bell that was likely part of the table service for the officers (Guibert, Poletto, Bigot, Veyrat et al. 2017). Boudriot indicated that the officers’ quarters could accommodate a cavenette, or small cellar (Boudriot 2001:151).

Comparing this artifact assemblage to those in the forward and midship areas, a decrease in the value of consumption-related artifacts is evident. Access to fine ceramics differed, and therefore the crew used what was available on the ship. A considerable gap is evident in the ceramic sets, such as numerous fragments of poorer-quality bowls or pitchers of Saintonge manufacture (Guibert, Poletto, Bigot, Veyrat et al. 2017; Guibert, Poletto, Bigot, Bianchimani et al. 2019). Along with this common crockery, the personal objects, such as wooden buttons or turquoise-blue beads, found in these areas demonstrate a much lower socioeconomic status.

The material concentrated in the midships area of the wreck is characterized by a specific spatial organization. The account of Émeraude’s 1823 visit (Colignon 1823) indicated that this area of the ship was used for storing food and water. The presence of rigging objects, including wooden double pulleys that were necessary for maneuvering cargo, as well as cooperage materials in the hold, such as the eight barrels found in 2019, reveals the storage function of this part of the ship. The armament specifications of Topaze—a sister ship of Anémone (Jaffray 1823)—indicate that food and water were stored in “empty water barrels being stowed” and “were placed in the middle of the hull” (Jaffray 1823). A barrel tap and a barrel pump were found in the immediate vicinity of the wooden barrels and are clearly associated with the storage and filling of bottles to facilitate table service, as is the demijohn.

Other artifacts found in the midships area in 2019 are related to the ship’s surgeon, further illuminating daily shipboard life. This material consists of ointment jars, vials, weights, and syringes. Consequently, the artifacts on board, notably weaponry, everyday objects, and, more specifically, luxury objects related to consumption, provide evidence of the ship’s French affiliation. The analysis of the material culture as well as its spatial distribution across the site informed a better understanding of Anémone and its function as a naval customs ship.

Project Logistics

Project Team

Between 2015 and 2019, the Saintes Bay Wreck Project team consisted of approximately 50 people, most of them volunteers, including 34 professional divers and 6 students. Three people took part in every mission, four people took part in four missions (for which they are warmly thanked). The professional divers included 10 professional archaeologists or their equivalent (freelance archaeologists, master’s students, doctoral students, an assistant professor), 2 professional photographers, and a professional videographer.

Excavation and Artifact Inventory

A total area of 63 m2 of the site was excavated (of which 14 m2 overlapped) during 101 days of fieldwork, including 77 days of diving, for a total of 922 h of diving. A total of 557 inventory numbers were assigned over the five years of excavation.

Funding

The total project budget spanning five years amounts to 152,485 €, with 96,500 € from the Ouacabou Association and the Lesser Antilles Archaeology Association and 55,985 € from the University of the West Indies AIHP GEODE. This figure does not include voluntary work, in-kind donations, or time spent for project planning, scientific collaborations, and report writing.

Publications and Public Outreach

Five archaeological reports, three scientific articles, and ten scientific notices were published. In addition, the project was featured on the cover of issue 49 of Archéologie médiévale, published in 2019. In terms of public-outreach materials, 2 exhibitions were developed, 10 articles were published in the local and national press, and 2 news segments were aired on the Guadeloupe 1st and Martinique 1st television stations.

Conclusion

This overview of the Saintes Bay Wreck Project describes the excavation, documentation, and analyses completed over five years. The project was the first multiyear underwater project authorized by the French Ministry of Culture in the French West Indies. We hope it sheds light on underwater cultural heritage and will foster and inform other scientific research in the West Indies.