April 30, 2025. Dr. Chris Wilkins, Senior Conservator.

The general role of an archaeological conservator is to stop degradation and corrosion of artifacts as archaeologists unearth them. It is crucial for conservators to have an understanding of the degradation and corrosion process of various types of materials. In this sense, a conservator is also a materials scientist who focuses on historical materials.

At Jamestown Rediscovery, we’ve been focusing on glass objects recovered from our excavations. This focus is the result of conservation of recently-excavated glass from Jamestown, conservation of several glass bottles recovered from excavations at George Washington’s Mount Vernon, and the ongoing care and management of the Jamestown glass collections (see “In conservation, this will probably be the hardest thing you ever do.” | Historic Jamestowne; Fig. 1).

To conserve a glass artifact, a conservator needs to know the process of glass production, its general history, and glass degradation/decomposition. Additionally, the conservator and responsible party (e.g. curator, stakeholder, and/or owner) consult in the proper storage of glass based on its recovery and condition. There are different types of glasses, let us explore one.   

Glass Production

Alkali glasses are the most abundant glass type throughout history. The production of these glasses requires at least three components: silicon, alkali (sodium and/or potassium), and calcium. Glass is actually mostly silica, so why do we need these other elements?

Melting pure silica sand will produce a stable glass called fused silica glass. However, the melting point of silica is 1713°C (3115°F), a difficult temperature to reach in both kilns from the past and most kilns today.

An alkali fluxing agent is required to reduce the melting point to a more kiln-acceptable level of 800-1000°C (1472-1832°F) for softening or 1400-1600°C (2552-2912°F) for a true melt. The fluxing agent also lowers the viscosity of the glass melt, making it easier to work and form. Glasses made with fluxing agents, like sodium, are known as alkali glass. However, pure silica-sodium glass is unstable and will quickly dissolve in water or become crizzled (intact highly fractured glass) over time. The introduction of calcium oxide to the mixture counters this by stabilizing the glass, if included in appropriate amounts. Adding other materials to the glass mixture allows for manipulation of opacity and color. 

A Brief History of Glass

Pliny the Elder (d. A.D. 79 at Pompeii) recounts a story of the discovery of glass by natron traders landing on the shores near the colony of Ptolemais at the mouth of the River Belus in present-day Israel (Natural History XXXVI:192). Not finding any stones to support their cookware, they used natron blocks from their stores and, under the heat of the fire, a molten glass formed beneath. Reflecting on the components of a glass mentioned in the preceding paragraph, upon the sandy, presumably shelly, beaches, this mixture of sand (silica), natron (sodium), wood ash (sodium/potassium) and shell (calcium), could have produced an unintended glass (Gorem-Rosen 2000).

However, the manufacture of glass certainly predates Pliny, with evidence of it having occurred as early as 3500 BCE in the Near East (Pfeander 1996) if not earlier in other places like Mycenae, China, and North Tyrol. A clay tablet recovered from the library of the Assyrian King Ashurbanipal (696-626 BCE) provides a glass recipe and states that glass can be produced from ‘60 parts sand, 180 parts ashes of sea plants, 5 parts chalke’ equating to 25% sand (silica), 74% ash (sodium oxide and potassium oxide) and 2% chalk (calcium oxide) (tablet | British Museum). While the ingredients for glass (silica, sodium/potassium, calcium respectively) are good, the quantities provided would not produce a good quality glass.

Glass is a conservative technology with very little change over long periods. Analysis of clear and colored Roman glass from across the empire revealed compositions of largely silica (SiO₂), with some sodium, potassium and calcium oxides (Na₂O, K₂O and CaO), with additional inclusions for color, opacity and incidentals (Ganio et al. 2012 ). The composition of typical 17th-century English alkali glass is similar (Dungworth and Brain 2009). Totaling the alkalis reveals little significant difference between Roman and English glass despite a separation of 1600 years!

Glass Degradation

The degradation of glass is dependent on its composition and the environment. Modern fused silica glass is an extremely stable material. Most archaeological glass, however, is alkali glass. Alkali glass can become crizzled and could eventually fall apart if the amount of calcium originally added to the mixture during production is not appropriate for glass stabilization, although this is more common with historical glass than archaeological glass.

Archaeological glass is susceptible to two competing degradation processes involving water: dealkalization and network dissolution (Freestone 2001). Dealkalization simply means removing alkalis (sodium or potassium). Water will leach the alkali and some calcium from the glass structure replacing it with smaller hydrogen atoms from the water. The replacement of larger atoms with smaller atoms forces the contraction of the affected layers. This contraction causes the layers to partially separate from the pristine glass below and eventually flake off (Fig. 2). Known as ‘gel’ layers, they will expand and contract depending on the presence of moisture much like a true gel. Each layer, as seen by the naked eye, is actually several layers. Refraction of light passing through them can make them appear iridescent or avariation of dull opaque browns. Loss of these layers results in a loss of the original dimensionality of the object as well as any inscriptions and adornments connected to that surface.

Network dissolution is the removal of silica, the main component of glass. Ground water generally has a pH of 7. Dealkalization leaches sodium from the glass into the surrounding ground water causing the pH to increase.

If the pH goes above 9.5, network dissolution becomes dominant over dealkalization. For instance, if rainfall rinses away the water, the pH never becomes critical. If stagnant, the higher-pH water causes network dissolution, or the loss of silica, which results in pitted glass and/or general surface erosion and a generally weaker glass object. A glass object where significant network dissolution has occurred may appear intact in-situ but can fall apart upon simple handling.  

Glass Conservation

The condition of the glass determines the conservation that has to occur to save the object. If glass is wet, or ‘waterlogged’, and has gel layers due to dealkalization, it has to be control-dried to ensure the gel layers do not flake away from the object. Placing the glass into successive water baths with increasing concentrations of ethanol and letting the glass dry afterwards will result in less loss.

Glass recovered from Jamestown is not usually waterlogged but does have soil ‘adhered’ to it. Uncontrolled removal of the dirt by a conservator can pull away the gel layers or even larger fragments of glass revealing pristine glass located a few millimeters below the original surface. While ‘pristine glass’ may sound ideal, removal of the gel layers results in the loss of anything that was on the original glass surface, like inscriptions, as well as a loss in dimensionality. A dilute adhesive applied to the adhered soil and glass locks the gel layers into place and allows the conservator to control the removal of dirt from the surface. At Jamestown, we use a mixture of B-72 and acetone. Using a scalpel and/or solvent-impregnated cotton buds, mechanical cleaning occurs centimeter-by-centimeter across the surface to expose the glass and gel layers (Fig 3, also see Fig. 1). 

Once the glass is cleaned, application of a final consolidation layer in the form of a dilute adhesive produces a glassy finish while protecting the gel layers. Adhesive between the gel layers decreases light refraction, causing the colors on the surface to appear muted.

Glass Storage

The condition of the glass when excavated also determines the way it should be stored. Glass found in the desert or as a part of an historic collection is best placed in a dry environment. This kind of glass can survive in very low relative humidity. Glass recovered from archaeological excavations in wetter environments, like Jamestown, could still have molecule-thick layers of water in crevices and between gel layers, even if the glass appears dry. Placing this glass into a dry storage location could cause the gel layers to flake off as the water evaporates. A relative humidity of 45-50% (as found in your typical household) is ideal for archaeological glass. At a higher relative humidity, moisture could collect on the glass surface restarting the dealkalization process and further degrading the object.

Mechanical damage is an immediate concern when handling glass, archaeological or otherwise. Knocking a cannonball off a worktable will make a large sound, and at worse, break a toe. Dropping a glass object from the same height will likely cause it to shatter with potential loss of shards and possible cuts to the person trying to retrieve it quickly. Placing the glass into housings (e.g. bag or box) while in storage helps to protect the object from mechanical damage. If the box is dropped and the object breaks, the shards are contained with no loss and can be handled when time allows. A simple labeled box, with supports for the object inside, placed securely on the storage shelf and tracked through a database ensures a higher probability that the object will remain intact and available for the future. Boxes can be dismissed if specialized cabinets are used. These usually have multiple removable, adjustable shelves and, sometimes, a glass front but still serve to protect the contents within from mechanical damage (Fig. 4) and buffers changes in temperature and relative humidity.

In Conclusion

Next time you have a sip of iced tea or hot coffee from a glass, think of the 5000 years of glass history and innovation that went into it. Think of how we, as humans, thought to crush sand and shell, and combine it with ashes of wood and plants, and/or natural salt deposits to produce a material that can be transparent or opaque and any number of colors. A material that is both strong yet fragile, but where simple exposure to water coupled with an extent of time can pull it apart atom by atom. And think of us, preserving this glass against the will of physics and nature so that those in the future can learn about and see the beauty of the past.

References Cited:

Dungworth, D. and Brain, C. (2009).  Late 17th–Century Crystal Glass: An Analytical Investigation. Journal of Glass Studies 51

Freestone, I. (2001). Post-Depositional Changes in Archaeological Ceramics and Glass. Handbook of Archaeological Sciences. D. R. Brothwell and A. M. Pollard, John Wiley & Sons: 615-625

Ganio, M., S. Boyen, T. Fenn, R. Scott, S. Vanhoutte, D. Gimingo and P. Degryse (2012). Roman Glass Across the Empire: An Elemental and Isotopic Characterization. Journal of Analytical Atomic Spectrometry 5

Gorin-Rosen Y. (2000). The Ancient Glass Industry in Isreal, Summary of the Find and New Discoveries. In M.D. Nenna, La route du verre: Ateliers primaires at secondairs du second millénaire av. J.-C. au Moyen Âge (Travaux de la maison de l’Orient Méditeranéen 33). Lyon. Pp. 49-63

Pfaender, H.G. (1996). The History of Glass. Schott Guide to Glass. Dordrecht, Springer Netherlands: 1-15

February 27, 2025. Eleanor Robb, Archaeological Field Technician.

Every year, the field crew at Jamestown Rediscovery excavates three burials from the 1607 burial ground inside the fort, which contains approximately 36 graves from the first year of English settlement. For the past several years, the team has built a structure over this excavation area, not only to keep the human remains within safe from the elements, but also to keep them out of the view of the public during such sensitive excavations. Jamestown Rediscovery has also become increasingly selective about sharing photographs of these human remains on public platforms. This decision has been made partially out of respect for visitors who may find these photographs unsettling to view. Jamestown Rediscovery also strives to treat the human remains excavated here with the respect and dignity they deserve. After all, these were actual people who lived and died here. This means that we keep our burial excavations out of the public eye, both on site and online. As a public archaeology site, however, we strive to keep our visitors in the loop with everything we learn through our excavations. So, how can we present our findings during burial excavations in a respectful manner without sharing photographs? The answer lies in illustration. 

Archaeological illustration is used regularly at all stages of excavation, especially when recording the stratigraphy of the site. Conservation staff also regularly draw artifacts in order to better study them and make their elements easier to identify. The illustrations I made of the burials this year combine these efforts by representing and recording the excavation area, the remains we excavated, and the methods we use. The production of three illustrations for the burials at the 1607 burial ground began early in the excavation process. As a newer member of the crew, I did not participate hands-on with most of the excavation of human remains this year. Instead, I orbited between burials, supporting the team’s work while also closely observing the excavations and making live sketches of the human remains, of the archaeologists, and of the whole excavation area. These live sketches allowed me to take very detailed notes of the positioning and condition of the human remains without relying on a two-dimensional photograph and the limitations of its quality.

Click to enlarge images. Live sketch and notes on burials JR1237, JR1846, and JR1335, after the remains were exposed in situ.

After the excavations were complete, I consulted with Director of Archaeology Sean Romo and Senior Staff Archaeologist Mary Anna Hartley to decide what kinds of illustrations would be most useful and appropriate to produce. Three different types of images were chosen: two overhead illustrations of the whole burial site, one in watercolor and one in pen and ink, and one painting adapted from one of my live sketches which would show two of our archaeologists at work.  

Each illustration also required different kinds of preparation in order to achieve the desired effect. For the watercolor paintings, I used a test sheet to determine how to mix each color I needed to use. I painted a swatch on the sheet and labeled each with what paints I used to mix it. Additionally, I painted a small test of the sets of remains so that I could see how all the colors would work together in the final product, and made notes about my process during this stage. This “cheat sheet” was helpful in maintaining a consistent result in both paintings over the multiple days that they were in progress.

For my pen and ink illustration, I created another test sheet by drawing the outline of one of the burials four times over, and using eight different shading techniques across them. This enabled me to consult with other staff members, especially with Director of the Archaearium Jamie May, as to which method would be both the most visually appealing and convey the most information about the remains. Jamie’s insights were particularly valuable, as she is not only an archaeologist with decades of experience, but an expert archaeological illustrator as well.

(L) The large painting of all three burials in progress (R) Putting the final touches on the large pen and ink drawing

Finally, for both the painted and inked overhead compositions, I overlaid a grid across my reference image, drew the same grid on my paper, and transferred the necessary lines. Then I was able to use my “cheat sheets” to create the final illustrations, making adjustments where needed. For the painting depicting the excavations in-progress, I recreated my sketch on a larger scale, and added color to create a lifelike scene. 

Each piece serves a specific purpose. The watercolor painting of all three excavated individuals provides visual information comparable to what our final record photograph would provide. In full color, the remains are situated within the excavation area, including other archaeological features in the area as well as elements of the excavation structure. This painting strikes a balance between the accurate and the artistic. While all of the elements in view are accurately positioned, the painting is a step away from the minute detail of a photograph or a penned drawing, and it wouldn’t be possible or accurate to take measurements from a painting like this.

The pen and ink drawing gives a more technical view of the remains as archaeological finds. This drawing includes a 0.5 foot grid spacing in pencil, which is not only to scale with the remains represented, but is also oriented with the archaeological grid upon which all of our excavations are based. This means that this drawing is useful for more specific purposes, such as reports or presentations, and can be used to take quick or approximate measurements of the burials.

The working scene, another watercolor painting, is less accurate than the other two illustrations, because it is based off of a live sketch. Rather than conveying information about the remains themselves, this painting instead seeks to provide a look past the curtain into the way we excavate burials. Because our burial excavations are obscured from public view, it may be difficult for visitors to get a good sense of what goes on inside. This painting provides a window into our process, along with all of the preparation work and tools we use to get it done.

The results of this process were three illustrations which convey different types of detailed information for viewers who are interested in our burial excavations, but who may not want or need to look at photographs of the remains. The process of illustrating these burials benefits not only the public, but the artist — as a newer member of the team, illustrating these burials has allowed me the closest possible study of these remains without personally excavating them. To draw something, to faithfully reproduce it, is to know it. Through illustrating these three burials, not only do I know them better, but more visitors are able to become familiar with our burials and how we excavate them. Even so, these illustrations represent only a small portion of the work that went into our burial excavations this year. 

For more information regarding the construction of our burial structure, the methodology behind our excavations, and the forensic analysis and conclusions drawn from each of these burials, explore our Dig Updates from August, September, and October 2024, as well as our Dig Deeper video series on YouTube.

(L) The final watercolor illustration depicting burials JR1237, JR1847, and JR1335 in situ (M) The final pen and ink drawing depicting burials JR1237, JR1847, and JR1335 in situ (R) The final painted working scene, depicting Staff Archaeologist Caitlin Delmas and Senior Staff Archaeologist Mary Anna Hartley excavating burial JR1237 within the burial structure.

September 27, 2024. Sean Romo, Senior Staff Archaeologist, Mary Anna Hartley, Senior Staff Archaeologist, David Givens, (former) Director of Archaeology.

A 1608 map of Virginia, drawn by John Smith, depicts the region around the English settlement at James Fort. The Fort is drawn as a three-sided structure, with bulwarks at each corner, a cross indicating the location of the church, and a flag-shaped fortification, or outwork, to the north. If you look closely at James Fort, you’ll see a series of dots extending from the west bulwark: a road.

Many roads across the country have Indigenous roots. While waterways were the primary thoroughfares, Indigenous peoples across the continent also created paths across the landscape between hunting grounds, villages, and nations. When Europeans arrived in North America, they adopted these paths. The path depicted on the 1608 Zuñiga map leads directly to Werowocomoco, the seat of power for Wahunsenacawh—more commonly known as Chief Powhatan—who ruled over more than 30 tribal nations in the region. It’s likely that First Peoples used this path for hundreds of years before the Jamestown settlers arrived.

The English would co-opt this Indigenous route soon after arriving. Part of this existing path would become the “Great Road” (or “Greate” – spelling varies in the 17th century). For the Jamestown colonists, this was the earliest road established in English North America. It was the gateway to the Virginia’s interior and a fixture on the landscape throughout the 17th and 18th centuries. In later years, the Great Road would connect Jamestown with nearby English settlements, including Middle Plantation—modern-day Williamsburg. Archaeologically, roads and paths are key parts of understanding historic landscapes. The Great Road, as an Indigenous path and then an English highway, is the way thousands of people first experienced Jamestown Island, and it defined the landscape for both groups.

Archaeology of the Great Road:

Portions of the Great Road were uncovered during the 20th-century excavations led by National Park Service (NPS) archaeologist J.C. Harrington. Harrington and his team of Civilian Conservation Corps (CCC) excavators—made up of local Black men—uncovered major portions of Jamestown, including buildings, property lines, and burials. His team found parts of the Great Road in two locations: one area north of the Pitch & Tar Swamp, and another near the Tercentennial Monument. More recently, Jamestown Rediscovery has used excavation and ground-penetrating radar (GPR) to find additional portions of the Road, filling in some of the gaps. Unfortunately, nearly 1000 feet of the Great Road is still unaccounted for. Maps from the 17th and 18th centuries suggest its location, but the exact path of the Road—and where it crossed the Swamp—is unknown. Since the Road was a major part of both First Peoples and English landscapes at Jamestown, tracing its route is key to understanding settlement on the Island.

Less than a century since Harrington’s excavations, the Pitch & Tar Swamp has grown. Areas that were once dry land are now permanently underwater and several of Harrington’s archaeological trenches are completely inaccessible. This inundation prevents land-based—or terrestrial—archaeologists from uncovering features near the Swamp, blocking access to important pieces of our shared past. Maritime, or underwater, archaeologists might be able to investigate here, but Jamestown Rediscovery does not usually do that kind of work. Could an underwater team be brought in to help examine Jamestown’s flooded landscapes?

The Right Team:

To trace the Great Road, former JRF Director of Archaeology David Givens reached out to Elizabeth Moore and Brendan Burke, State Archaeologist and State Underwater Archaeologist, respectively, at the Virginia Department of Historic Resources (DHR). Elizabeth and Brendan recommended we bring in Southeastern Archaeological Research, LLC (SEARCH), one of the largest private archaeological firms in the United States. SEARCH has extensive experience in both terrestrial and maritime archaeology and their maritime team has led several internationally-recognized projects, including the searches for Clotilda and Endurance. SEARCH was eager to help, and sent an expert team of underwater archaeologists, led by John Albertson, to assist with the search for the Great Road.

The first step was to identify the likely path for the Great Road between where it had been found in the 1930s and 1940s. Using historic maps from the 17th and 18th centuries (including the ones above), archaeological findings from NPS digs and more recent Jamestown Rediscovery excavations, in-depth topographical analysis, and cutting-edge GPR surveys, the Jamestown team was able to pinpoint the most probable path of the Road through the Swamp.

The SEARCH begins:

After deciding on the likely path of the road, the SEARCH team descended on Jamestown in June 2023 to deploy their advanced equipment to help with the survey. Over two days plagued by summer thunderstorms, SEARCH, DHR, and Jamestown staff, including students of the 2023 Jamestown/University of Virginia Summer Field School, conducted a survey in the western section of the Pitch & Tar Swamp. First, the combined teams had to build a survey platform that would work in the swampy environment. A sub-bottom profiler was suspended between two kayaks and a data collection station was positioned on the makeshift deck. A sub-bottom profiler uses sonar (sound waves) to map inundated areas, not unlike a water-borne version of GPR. In theory, data collected by the sub-bottom profiler can show changes in density within the Earth, including areas of increased compaction typical of a road’s surface. Guided by GPS, the team pushed the scanning unit back and forth across the swamp, fighting thick brush, mosquitos, and deep mud. As they moved, the sub-bottom profiler recorded detailed information about the waterlogged environment.

The environmental challenges of deploying a sub-bottom profiler in the gas-filled swamp environment pushed the limits of the machine. Methane gas bubbles and the shallow depth of the swamp complicated the data collection process, but positive results were found in at least two locations! There, “hard returns”—evidence of a solid, compacted surface beneath the mud of the swamp—were visible in the sonar data. These spots are about where the team expected the Great Road, and might actually be the surviving road surface.

While these limited results aren’t 100% confirmation of the existence of the road, “they are worth troubleshooting and certainly warrant additional work in the swamp,” according to SEARCH lead John Albertson. This work shows that sub-bottom profiling in Jamestown’s flooded landscapes can work and provides additional circumstantial evidence for the Great Road’s location. Hopefully, Jamestown Rediscovery archaeologists can continue to work with SEARCH and DHR to follow up on these tantalizing results.

Want to learn more about the project? We’re releasing two new videos on our YouTube channel soon to share the ins-and-outs of how the survey came together. Make sure to subscribe so you don’t miss the latest updates.

Archaeologists recovered many swords and sword elements while excavating the Governor’s Well in the fall of 2023. However, the very first one to come out of the ground has proven to be one of the most intriguing finds from the feature. Five elements found together in situ are likely part of the same sword blade. Unfortunately, the hilt, grip, and pommel are missing, making the date and place of manufacture difficult to determine. Bricks discarded in the well were cemented to the iron blade by corrosion as it lay in the moist ground for 400 years.

As soon as excavators brought the sword into the lab from the field, it was dry-brushed and cataloged. Based on its measurements, it is likely a broadsword.

Conservators x-rayed the fragments, which revealed a maker’s mark on one side, which included letters, symbols, and possibly a number. Since this is the first marked sword blade from all of Rediscovery’s excavations, the reveal was exhilarating! The name “FERARA” appears one above the other within each of the blade’s fullers — along with an unknown symbol before the letter F and after the final A.

Next, the corrosion and concreted brick fragments were removed using air abrasion, and a second x-ray was taken to determine if the mark was easier to see. The marks before and after the name are somewhat more apparent but are still unidentified.

The blade was placed in a desalination bath (sodium hydroxide) for about 2 months to remove chloride salts that cause iron to corrode in lower relative humidity. In late March 2024, conservation of the two largest fragments of the blade were completed! Before conservation, the longest blade fragment was approximately 45cm long and 4cm wide and displayed minimal tapering from the tang to the broken end. Now that the conservation process is complete, we hope to see as many pieces of the blade as possible together to better understand the overall length. The conservation treatment that the sword has undergone will stabilize the artifact and preserve the mark. The sword is now housed in long-term storage, where it will be monitored for corrosion outbreaks. Perhaps one day it will go on display in the Archaearium museum.

The name Andrea Ferrara is one of the most ubiquitous marks found on 17th century swords. The widespread use of the name and variations in spellings, symbols, and other stylistic differences of the mark suggest that multiple individuals and workshops produced them. While a maker’s mark is generally intended to connect a product immediately to its maker, which can not only date the artifact but also provide a place of manufacture, the sword from the Governors Well is much more mysterious.

At some point, perhaps in the 17th century or even earlier, the mark of Andrea Ferrara began to be associated with a high quality weapon. This in turn led many English and European swordsmiths to use the name fraudulently on their products to enhance the perceived craftsmanship and value of their work. This explains the ubiquity of the mark. So who is this Andrea Ferrara?

The identity of Andrea Ferrara has been an enigma for centuries. A number of 19^th century antiquarians attempted to track down an individual with the name Andrea Ferrara, noticing the mark in historical documents and on extant swords. Sir Walter Scott, renowned Scottish author and arms collector writing in several historical accounts, diaries, and correspondence throughout his life, noted that Andrea Ferrara’s name was inscribed “on all the Scottish broadswords that are accounted of peculiar excellence.” (Scott, 1814)

The association between Ferrara and good quality Scottish basket-hilted broadswords became so strong that the swords themselves started being referred to as “Andrea Ferraras,” as in, “when he is armed with his pistols and a Ferrara, has good effect.” (Gilpin, 1792 p. 230). This association extended to literary works by prominent figures like Sir Arthur Conan Doyle (of Sherlock Holmes fame, writing c. 1889) and J.M. Barrie (of Peter Pan fame, writing c. 1890), who used the name to describe swords in their writings.

However, this early research was founded upon dubious facts, and Andrea Ferrara remained a mythological figure. According to early investigations, Ferrara was either Spanish or Italian. He supposedly not only exported swords from his home country but also traveled to Scotland to teach local metalsmiths his techniques. Depending on the account, Ferrara was invited to Scotland by King James III (reign: 1460-1488) (Gilpin, 1792), James IV (reign: 1488-1513), or James V (reign: 1513-1542) (Scott, 1814). In one account, Ferrara created a manufacturing operation in Scotland but was forced to flee the country after killing an apprentice who spied on his secretive sword making methods.  

It was not until the 1980s that arms researchers confirmed that a real Italian swordsmith named Andrea Ferrara existed and had a connection to the English market. Andrea Ferrara and his brother lived and worked in Belunno, near Venice, Italy. He was born sometime in the 1530s and died in about 1612 (Blair, 1998). On December 5th, 1578, the Ferrara brothers agreed to supply 600 swords to two English merchants monthly over ten years. Whether the massive order included marked swords or if the brothers made all 72,000 swords for the English market is unknown. If Ferrara made, marked, and fulfilled the late 16th century Englishmen’s order of at least some swords, it is conceivable that the sword found in the Governor’s Well was forged by him, purchased on the English market and brought to Virginia by an early settler who ultimately discarded their weapon at Jamestown. We may never know for sure. Although unique to Jamestown, this is not the only Ferrara-marked sword found in Virginia. Another one was found during excavations in Hampton, Virginia, at the site of Fort Charles, which was established by Lord De La Warr in 1610. Perhaps the two swords were owned by men who knew each other 400 years ago in Tidewater, Virginia.

sources

Blair, Claude. “New Light on Andrea Ferrara.” The Arms and Armor Society Newsletter, May, 1984.

Blair, Claude. The Crown Jewels. Vol II. 1998.

Ferrara, Andrea. Sword (Rapier) and Scabbard. c. 1575-1583, Philadelphia Museum of Art, Philadelphia.

Gilpin, William. Observations, Relative Chiefly to Picturesque Beauty, Made in the Year 1776, on Several Parts of Great Britain: Particularly the High-lands of Scotland …. United Kingdom. London, R. Blamire, 1792.

Mowbray, Stuart. British Military Swords. Volume 1: 1600-1660. The English Civil Wars and the Birth of the British Standing Army. A. Mowbray, 2013.

Pickup, David R. “Simon Mayne, Regicide, and Dinton Hall.” Records of Bucks. Volume 54. 2014.

Royal Armouries. “Weapons in Society conference – Day 2 – Keynote – Would the real Andrea Ferrara please stand up?” YouTube, June 2021, https://www.youtube.com/watch?v=67-VE9DUo9U.

Scott, Walter. Introductions, and Notes and Illustrations to the Novels, Tales, and Romances of the Author of Waverley. Volume 1. Edinburgh, R. Cadell, 1833.

Staples, Hamilton B. “Presentation of the Sword of Fitz-John Winthrop.” Proceedings of the American Antiquarian Society, October 1887.

Leather shoes are a type of artifact that is almost solely (a little pun there for you) recovered from waterlogged features like wells. The waterlogged, anaerobic environment of the lower layers of a well significantly slows bacterial degradation of leather. A total of 16 database records were made to capture fragmentary leather shoe elements excavated from the Governor’s Well feature. At least two, maybe three, leather shoes are represented by the fragments, including parts of shoe soles as well as upper elements of shoes. Also included is one latchet, or the leather strap that would have helped to hold a shoe onto a 17th-century foot!

Because the leather was wet when excavated, the fragments were kept wet as they came into to the Rediscovery’s lab. The first conservation task was to make a detailed record of each piece, what it looked like, its size, and any notable characteristics. The fragments were drawn and traced onto Mylar (thin transparent film) to record elements on the grain (skin exterior) and flesh (skin interior) sides including lace holes, thread holes, shoe nail holes, wear patterns, iron staining, large and small cracks, etc.

The first shoe is a matching insole (#141747) and outsole (#141743). The outsole is the part that faces the ground, and the insole is the part of the shoe that faces up towards the foot. Usually the two are separated by a midsole, but we haven’t identified one belonging to this specific shoe. The outsole and insole were excavated separately, but when laid on top of each other, the thread holes that stitched the two elements together align. The outsole, pictured below, even still had some of those threads in the holes!

Measurements of shoe leather are made in specific ways. Leather thickness is often given in ounces for shoe uppers, and irons for shoe soles. One ounce = 1/64″ and one iron = 1/48″. The term iron originates from the plates of iron that were used by European shoe-makers to measure the thickness of leather used for various parts of the shoe. The outsole of the first shoe is a 6-iron, meaning that the leather is 0.125″ thick. The insole is 8oz, also 0.125″ thick, same as the outsole. In total, it measures 18.93cm in length and 6.66cm in width at the ball of the foot. These measurements indicate the shoe fit a modern US child’s size 11-12. Whether it worn by a child or a petite woman, we may never know, but it shows shows the presence of women and children at the Fort around the time the well was filled with trash.

The second shoe is represented by an outsole, a midsole, and an extra heel fragment. When excavated, the outsole and midsole were held together by two visible iron nails near the toe. The heel fragment was found separately and matched the shoe because the thread and nail holes and heel curvature coordinated. A diagonal slice was made down the middle of heel fragment to extend and fit a added scrap of leather so that it fit the heel curvature and the rest of the shoe.

Larger than the first shoe, this outsole is 8-9 iron (0.167″ – 0.188″ thick), the midsole is 5 iron (0.104″) thick, and the additional heel lift is 8oz thick (0.125″). In total, the outsole measures 25.44cm in length and 8.86cm at the widest point, making it about a modern US men’s size 6-7.

Two iron shoe nails (or hobnails) holding the midsole and the outsole together are visible near the toe end of the outsole. This shoe was X-rayed by conservators to determine the number and location of shoe nails, and how they may be problematic in conserving and stabilizing the leather. The X-ray (below) showed more nails and nail elements are in the leather than visual inspection revealed! One nail has a clinched point, that is, it was bent over to ensure the shoe was securely attached. In addition, documentation revealed other interesting information about this shoe’s wearer. Near the ball of the foot is a “Dutchman,” a small leather fragment that was inserted in between the outsole and midsole when the shoe was made (see drawing above). This fragment indicates the shoe was made-to-order, and the additional 8-9 iron piece was likely placed to help correct a supination or overpronation of the wearer.

After recording and x-rays were complete, conservators cleaned the leather fragments with soft brushes and probes under water or with gently flowing water. As additional observations were made during the cleaning process, new aspects of the shoes (e.g. additional thread holes, wear patterns) were recorded on the Mylar tracings.

Next, the leather was immersed into sodium ethylenediaminetetraacetic acid to remove iron salts. Leather will shrink during the drying process, even if the shrinkage is undetectable. Iron salts, however, do not shrink. If they are present as the leather dries, the iron salts will cause the leather to tear. Iron salts left on the leather long-term could also cause chemical degradation to the leather over time.

Drying is the last major conservation step, conducted with either a freeze-dryer or through a slow-drying process. In the case of the shoe leather from the Governor’s Well, our conservation team decided to do a slow-drying process. To reduce shrinkage and distortion while drying, the leather was immersed in glycerol for several days, and then removed from immersion to slowly dry. The slow-drying process is ongoing.

Ideally, after conservation is complete, the leather shoes and other leather fragments will be dry and supple with little to no shrinkage, thereby allowing these precious objects to be handled and studied by researchers.

Just as they do today, shoes in the 17th century experienced a lot of wear and tear. They wore out or were outgrown, as evidenced by shoes recovered from Jamestown’s wells. By reflecting on artifacts as seemingly mundane as shoes, we can gain a deeper appreciation for daily life in early colonial America., and foster our connection to those who came before us.

Amanda Nedell recently completed an 8-week long internship with Jamestown’s Curatorial staff, focusing on a ceramic ware type called Spanish Coarseware. Most of the Spanish Coarseware vessels in the Jamestown collection are olive jars. Amanda recently graduated with her Bachelor’s degree in Anthropology/Archaeology from Mercyhurst University where she worked in the archaeology lab. To find out more about her project with us, continue reading.

Of the Spanish Coarseware ceramics excavated at Jamestown, most of the assemblage has been identified as olive jar fragments. These vessels are largely unglazed and have coarse fabric with a high rate of absorption. Despite their name, olive jars were known to be used to transport far more than just olives in brine. Water, wine, olive oil, vinegar, honey, chickpeas, various nuts, beans, pitch, gunpowder, and bullets are all known to have been stored in these large vessels (Goggin 1960:6; Avery 1997:190-195). The absorbent nature of the ceramic acted as a form of preservation; evaporation from the exterior of the vessel allowed the contents to remain cool (Deagan 2002:32). The primary use for olive jars was in commercial transportation, and this is likely how they came to be in the archaeological record at Jamestown. While no evidence of this exists at our site, olive jar fragments on other sites were used as architectural fill when constructing buildings and buried beneath dirt floors because the fragments acted as a natural dehumidifier (Avery 1997:103).

Olive jars were wheel-thrown in a few different shapes and sizes. At Jamestown, only two vessel shapes are found, globular (more round) and ‘carrot-shaped’, and both date to the late 16th to early 17th century. While most of the vessels found at Jamestown are incomplete as partially mended vessels or remain fragmentary as unmended sherds, at least 13 to 15 olive jar vessels are represented by over 1,000 individual sherds recovered to date.

Before we began this process, only 3 mended vessels and a few mended sections of vessels existed in the Jamestown collection, and most of the ceramic sherds were in long term storage in Jamestown’s lab facility. All of those fragments were pulled, assessed, and labeled before the mending could begin. The process of mending ceramics includes the use of both acetone and of 15-20% acryloid B-72 (w/v in acetone), an adhesive solution that also allows for disintegration and reversal if necessary. Prior to working on new mends, Associate Curator Emma Derry, Collections Assistant Lauren Stephens and I disassembled old mends to ensure the stability of the mended ceramic vessel. This was done by soaking acid-free cotton in pure acetone and wrapping it around the mended areas, allowing for the material used for the mend to dissolve.

The cross mending process is similar to putting a puzzle together, except we often don’t know exactly what the final result will look like. As we crossmend fragments, we group them together, creating a vessel or a “complete” puzzle. With olive jar fragments our research helped us to understand the possible shapes and sizes, but the color and coarseness of the fabric is quite similar which means that close analysis of the sherds is needed to mend them or put them with other fragments that closely resemble each other. We hoped that the crossmending process would help us to better understand how many vessels were present in the collection, and their shapes and sizes. This information then allows us to theorize on what they may have contained when they were shipped to Jamestown.

Different pieces are found all over the site and putting them together, even if the vessels aren’t complete, allows us to better understand different areas of the site and how the vessels were used. Two of the most complete vessels in the collection were both found in the Factory feature just outside the fort, a structure in use during the early fort period. The colonists were using fire for distilling and metallurgical purposes in this area, so they may have repurposed these olive jars to hold water for drinking, or to act as a 17th century fire extinguisher.

One of the vessels found in the Factory feature has been on display in the Archaearium museum since the museum’s opening in 2007. We temporarily removed the vessel from display for reassessment and cleaning. I was able to find one additional mend to this vessel, and began the process of disassembling and cleaning it.

I spent a large part of my internship reassembling and mending another of these two vessels. This olive jar is approximately 40 centimeters in length and, by my estimate, 60-70% complete. The vessel includes 60 sherds mended together, ranging in size from almost the entire top part of the vessel with the rim and shoulder, to sherds that are about the size of a thumbnail. It took around 5 weeks to completely find and mend this vessel. Following the mending process, I joined Collections Assistant Lauren Stephens in the photography process of the vessel with Senior Staff Archaeologist and Staff Photographer, Chuck Durfor.

Following the completion of the mending portion of the project, I had the privilege to use Jamestown’s Portable X-Ray Fluorescence, or pXRF, machine to analyze the chemical and elemental makeup of the glaze and fabric on various Spanish Coarseware vessels. We also compared several forms of unknown ceramic fragments to a known fragment of Spanish Coarseware to identify if the unknown sherds were, in fact that ware type. Based on its shape, one of the fragments that was previously classified as olive jar is now thought to be an orza, a Spanish Coarseware storage jar with a flat bottom. Our pXRF analysis of the orza fragments has confirmed that based on the composition of the clay, they are Spanish Coarseware!

After using pXRF on the various vessels, including the orza, olive jar fragments, and mercury jar fragments, the results were uploaded onto special software, called Artax, allowing me to read the wavelengths and identify which elements can be seen within the sherd. In addition to confirming that the orza was Spanish Coarseware, we were also able to test the glaze on different sherds. One small olive jar, previously thought to be covered on the interior surface with black pitch, was tested and inspected more closely and was found to actually have a dark green lead glazing (shown below). I completed this part of the project with the assistance of Collections Assistant Lauren Stephens and Senior Conservator Dr. Chris Wilkins.

This entire process has been incredibly rewarding, even though it was an extremely slow and tedious process, I was able to mend the majority of a vessel. I was also able to learn scientific analysis techniques on a sophisticated piece of machinery, an experience that would be hard to replicate elsewhere. It is my hope that the vessels I analyzed and mended will be used to educate the public and assist in the understanding and study of the Jamestown site overall for years to come.

references

Avery, George E. 1997. Pots as Packaging: the Spanish Olive Jar and Andalusian Transatlantic Commercial Activity, 16th-18th Centuries. Doctoral Dissertation, Department of Anthropology, University of Florida. University Microfilms International, Ann Arbor, MI.

Deagan, Kathleen 2002. Artifacts of the Spanish Colonies of Florida and the Caribbean Vol. 1: Ceramics, Glassware, and Bead, expanded and revised from 1987 edition. Smithsonian Institution Press, Washington, D.C.

Goggin, John M. 1960. The Spanish Olive Jar: An Introductory Study. In Papers in Carribbean Anthropology, Vol. 62, Irving Rouse, editor, pp. 3-48. Yale University Publications in Anthropology, New Haven, CT.

Peacock, Caroline 2023. From Olive This: A Characterization of the Spanish Olive Jar in mid-16th Century New Spain. Master’s Thesis, University of West Florida

Jackie Bucklew, Conservation Intern

In 2014 archaeologists Mary Anna Hartley and Danny Schmidt excavated fragments of a case bottle found in a fort period cellar feature of the northeast corner of James Fort’s 1608 addition. (Fig. 1) The case bottle had been partially mended by conservator Katie Cornelli, joining enough shards to establish the dimensions. After a few years in the collection, it was due to be reassessed. This included updating the description, noting the condition, taking photographs and measurements. Additionally, looking into the history of this settlement, some 17th century fort renovations had potential to spread shards of this vessel across the site. This prompted a search in the archive which produced additional shards that had not previously been associated with this bottle. At this point it was decided this bottle would be given further conservation treatment.

The condition of the glass showed it had undergone chemical degradation while in the ground. This process involves ground water molecules disrupting the silica-oxygen bonds by leeching the glass alkali stabilizers (i.e. sodium and potassium) used in glass making. This caused the formation of gel layers (glass delamination) and an uneven refractive index resulting in an iridescent appearance, and eventually in a loss of glass.

The goal was to prevent the continued delamination of the gel layers, which was achieved using 8% B-72 (w/v in acetone). B-72 is a synthetic acrylic resin used as a coating, but also as a consolidant, or an adhesive at higher concentrations. At 8% it’s quite runny, but it produces a thin layer that can be drawn into tiny crevices through capillary action, thereby consolidating the thin gel layers of delaminating glass. Once each of the total 200+ shards were coated, and had about 12 hours to set, it was time to start piecing it back together. Using a concentration of 20-40% (w/v in acetone), B-72 was applied by brush to both mending edges, thick enough to adhere but thin enough as to not cause a gap that could be problematic as the bottle is built up over time. After a few minutes holding the piece in place, the newly mended shards were supported in a sand box with cling film barrier and left to set overnight.

Adhering glass shards is more difficult than adhering ceramic sherds; the break is usually quite smooth and when joining, it seldom ‘grips’ like ceramics. With the smooth break edges the mends tend to drift if not perfectly supported, causing the joins to be close but ultimately needed adjustments. To do that without completely removing the mend, cotton wool is rolled into a cylinder shape, dipped in acetone and placed directly over the join. After a few minutes the mend is adjustable again. If needed, leaving the dampened cotton wool a bit longer allows the shard to be completely removed. When a shard is removed, a cotton bud and acetone are used to clean the residual adhesive off the break edge. Since B-72 was used both as consolidant and adhesive, extra care was taken to ensure the consolidant coating was not affected. If a shard showed signs of flaking after being adjusted or removed, a new layer of 8% B-72 was applied.

After joining all 132 shards, this bottle is easily the largest in the collection. It measures 37 cm tall, 14.5 cm wide and would have held approximately 193 fluid ounces or 1.5 gallons. For comparison, an average case bottle of this period holds approximately 45 fluid ounces and measures around 20 cm tall and 8 cm wide.

With the bottle as close to completion as our archaeologists could provide, it was apparent that it could not be stored in its assigned drawer standing upright. Generally, it’s best to keep objects in their originally intended orientation. When this is not possible supports and mounts are the next best option. A support of glass microspheres and B-72 was made allowing the bottle to be stored horizontally. To make this support, 3M glass microspheres were mixed with 40% B-72 (w/v in acetone) to produce a crumbly putty consistency. The bottle was gently pressed into the material to form a contoured support. Having the bottle fully supported reduces the risk of pressure points or damage to the individual shards and the bottle as a whole.

After the solvent evaporated, the support remains strong and lightweight. As for supporting the neck, a few layers of ethafoam were added to the base. The bottle will be housed with the rest of the glass collection in a temperature and humidity-controlled environment to help preserve these artifacts for as long as possible.

This project, though daunting, has ultimately been rewarding. When I began this project my mentor, Dr. Chris Wilkins told me, “in conservation, this will probably be the hardest thing you ever do.” Maybe trying to keep my expectations at bay, but also giving me an awareness of the estimated timeframe. Not only is the task of finding mends challenging and time consuming but also only being able to adhere 1-3 shards a day and waiting overnight for those pieces to set. A few times, this would be nearing completion, only to find a gap near the top, indicating a join was not perfect. This would require me to take down mends or make adjustments until the pieces could fit. Now with a renewed sense of patience, I will take this knowledge and experience onto my next conservation project.

Learn more about case bottles in the Jamestown collection.

Dendrochronology is the study of the size and spacing of tree rings within a wood sample to identify the potential date and origin of the tree from which it was harvested. This technique is extremely useful in archaeological research. For example, our knowledge that the 1587 Roanoke and early Jamestown settlements occurred during two of the most severe droughts in 800 hundred years was determined by Dennis Blanton’s study of bald cypress tree rings from the tidewater regions of northeastern North Carolina and southeastern Virginia.¹  Currently, the Jamestown Rediscovery (JR) staff hopes to apply dendrochronology to a few key wooden artifacts recovered from early context wells, e.g., the barrel staves found in the first well (c. 1608-1609) and the fruit wood handle of the Roman Lock pistol recovered from the second Jamestown well (c. 1610-1611).

To learn more about how samples are prepared for study and to share Jamestown’s resources with other important local historical researchers, I was pleased when 10-time National Scottish Fiddling Champion Dr. John Turner² asked if the JR staff could prepare high-resolution photos of an old violin to help identify its date and location of manufacture. As a longtime fan of John’s work (his annual Hogmanay performances in Colonial Williamsburg are not to be missed), I was delighted to have an opportunity to collaborate with this talented artist. The photos needed to have no glare, good focus, and detailed resolution of the tree ring structure (i.e., the grain lines of the violin must be full size, without pixelation, and at a resolution of at least 600 DPI).

David Rattray³ and Peter Ratcliffe⁴ provided additional information on electronic image preparation for dendrochronology (which will be very helpful for future Jamestown studies):

• High-resolution digital (scans and photographic) images have been very successful in dendrochronology studies.

• The use of digital images permits instant digital communication and rapid testing.

• With the increasing ease with which good quality, high-resolution digital images can be produced, it is much easier to commission a dendrochronology test. The costs of analyzing samples via the Internet have considerably reduced processing and travel expenses, which makes testing much more affordable.  

• When picture quality is high, (which reveal sharp tree-ring boundaries over the whole section to be tested), the accuracy of measurements is second to none.

• As a distinct advantage over microscopic analyses, digital files can be stored for re-assessment at a later date if required. 

• There is no set number of images required for testing, but sharpness and high resolution are required to collect accurate measurements from the rings. Instruments showing particularly tight growth rings may need more advanced photographic equipment than most ordinary cameras can provide. Most violin shops now have in-house photographic studios and can supply the best possible images. 

• The tree-ring measurements are gathered from a specially designed software module, where enlarged images are loaded. 

Using the same techniques as in artifact photography, images of the violin (see below) were collected using both full-frame and medium-format cameras. The photos were then transferred electronically to the UK for analysis. Given the fragility of our archaeological wood samples, learning that dendrochronology can be performed while the artifacts remain in the lab is very good news!

a summary of the results:

• The belly (spruce top) is made in two sections, joined down the middle.

• A total of 168 rings were measured on the bass side (left) and 155 rings on the treble side (right).

• The most significant cross-matches place the latest visible fully formed growth ring on the bass side at 1774 CE and at 1822 CE on the treble side. As the spring growth of the following year is just visible after the latest rings are measured, one year needs to be added to the dates, leading to an overall terminus post quem or earliest possible felling year of 1823.

• The results strongly suggest that the origin of the trees is the central Alps.

• Based on previous research, the most significant cross-matches for the bass side are a c. 1780’s Bavarian violin; a c. 1800 South German violin; an 1858 English guitar by Panormo; and an English violin of the school of John Lott, etc.

• For the treble side cross-matches include a Danish violin, 1848 by Thomas Jacobsen; a mid-19th c. Saxon violin; a c. 1890 French Mirecourt violin; an English guitar by the Roudhloff brothers; and several of other instruments.

• While no definite identification was achieved, one possibility is that the violin was made by John F. Lott (II), 1805-1871, who was known to make brilliant copies of 18th c. Italian instruments, and put facsimile labels in them. The label in this violin is a correct facsimile of an authentic Carlo Bergonzi label. (Carlo Bergonzi, who was born almost 40 years after Stradivari, completed several of the master maker’s instruments after his passing.⁵)

Finally, you might ask what was the charge for my work? A very high price indeed, once the violin is repaired to playing condition, John promised to play some period-appropriate music in the Historic Jamestown Church. I cannot wait!

1 The Lost Colony and Jamestown Droughts, D.W. Stahle et al, Science, Vol. 280, Issue 5363, 564-567 (1998)

2 http://www.fiddletree-music.com/johnturner.html

3 David Rattray is a violin maker and violin expert/curator of various collections in the United Kingdom. For more information please see –  http://www.davidrattrayviolins.co.uk and http://www.davidrattrayviolins.co.uk/publications.html

4 David Rattray HonARM, Strings Consultant RCS, Senior Fellow RAM, Ridge House, 22 Balwearie Gardens, Kirkcaldy, KY2 5LU. 

5 https://www.thestrad.com/lutherie/carlo-bergonzi-was-never-a-wealthy-violin-maker-but-he-still-used-the-best-quality-maple-ever-seen/13130.article

April 12, 2024. Janene Johnston, Associate Curator, and Dr. Chris Wilkins, Senior Conservator.

Archaeologists recovered a trove of interesting artifacts during excavations of the Governor’s well in the fall of 2023. Amongst these were several large iron objects including a billhook, three swords, a sawblade, and several pieces of armor. One particularly exciting find was a fairly intact rapier, comprised of the pommel, hilt, and a portion of the blade. In addition to this largely intact portion of the rapier, nine small fragments of the hilt were recovered throughout the well.

Though the rapier was heavily corroded, an edge of the original metal of the pommel was visible. The shape of the visible metal suggested the pommel was not solid, like most other sword pommels in the Jamestown Rediscovery artifact collection. The curatorial team suspected this could be one-of-a-kind for us, making the artifact even more intriguing. X-rays (below) taken by Jamestown’s conservators confirmed our suspicions that the pommel was hollow, and trefoliate (meaning resembles a trefoil) in silhouette.

Combined with the distinctive style of the quillons (the two arms forming the crossguard), it is likely that this is a Spanish rapier made during the mid-16th to mid-17th century. A similar Spanish rapier dated ca. 1575 is in the Victoria and Albert Museum collection.

We always anxiously await the end of the conservation process, as seeing the artifacts stabilized and expertly cleaned is an exciting reveal. Additionally, once conservation is complete for this rapier, we’d like to see if any of the additional hilt fragments recovered from the well can be reunited with the intact portions, and we would like to pursue further research on the artifact. However, the iron artifacts from the Governor’s Well are behaving quite differently than other iron excavated from the site, and the conservation process is taking a little longer than expected, including for this amazingly intact rapier.

Conserving the Well’s Iron:

Excavated iron from terrestrial Jamestown contexts generally looks like a bulbous mixture of sand, brick, and iron, cemented together with iron corrosion of varying shades of reds, oranges and browns. Exposed core metal appears gray to black, and the smell of the iron is generally the smell of the surrounding soils. Iron materials recovered below the waterline in well features at Jamestown typically have thinner corrosion crust and no distinctive smell. Iron artifacts recovered from waterlogged contexts like the well are initially kept in water to remove the potential of aggressive localized spot corrosion. However, due to the number of iron artifacts coming into the lab as the excavation was ongoing, collections staff worked together to start cleaning the artifacts as they were brought in, using toothbrushes and wooden tools to clear away excess mud and dirt.

As the artifacts recovered from the Governor’s Well were being cleaned, it was noted that the iron objects had a thin, black, almost oily coating and an indescribable industrial-like smell that was lacking in the well water itself. We took note of this, but proceeded with our normal procedures. After cleaning, the artifacts were dabbed dry in preparation for cataloging and x-radiography before they continued through the conservation process. During air abrasion, the next stage of conservation in which corrosion is removed, conservators noticed that the iron corrosion was more dense than usually encountered on iron from Jamestown. Removing this tough corrosion product means that the air abrasion process for each iron artifact took longer than normal.

Part of the conservation process for iron is the removal of chloride salts, which are a catalyst for continued corrosion. This process, called desalination, involves the immersion of the iron into a series of highly alkaline aqueous baths of sodium hydroxide, which stops the corrosion of iron and allows the salts to disperse into the surrounding liquid. After these baths, residual sodium hydroxide is washed out and the iron is carefully dried. During the desalination process of iron artifacts from the Governor’s Well, including the rapier, a moderately strong sulfur smell was apparent, and a new reddish-orange splotchy localized color on the surface was noted after drying. These unusual characteristics, combined with the strange smell and oil-like sheen of the iron as it was cleaned in the lab, were all indications of sulfate reducing bacteria (SRBs), a category of bacteria that requires an anoxic (lacking oxygen) environment to survive.

The Bacteria:

These phenomena could be caused by a number of factors that are unique to this particular well or possibly even from the prolonged effects of sea level rise that the island is facing. Though tests have proved inconclusive on the origin of these bacteria, they are producing hydrogen sulfide which is reacting with the iron. Characteristics such as the black biofilm, increased density of the corrosion crust, and sulfur blooms are all a direct result of the presence of these SRBs. Sodium hydroxide baths are known to remove chloride salts and are suggested within the conservation literature as an appropriate method for the removal of sulfide salts. Thus, our desalination process to remove chlorides is working because the smell of sulfur is very evident. As a result, we are extending the length of time an artifact remains in the sodium hydroxide baths until we can no longer detect the presence of sulfur and sulfur compounds in the aqueous solution.

The issue of sulfur and artifacts in anaerobic environments can be tricky depending on the material. For iron, it seems straightforward in the sense that it is readily detectable once you know what you are looking at. This also extends to other metals including aluminum, titanium, zinc, nickel, and copper alloy. Beyond metals, waterlogged wood can become partly infused with pyrite, a sulfur compound. Typically a concern for wood recovered from saltwater marine environments, evidence of pyrite infusion does not become visual until the formation of sulfur blooms years after it was conserved.

Moving Forward:

Having noted the strange characteristics of the iron recovered from the Governor’s Well, and having conducted the research on how to best move forward, Jamestown’s conservators are now armed with the information required to mitigate the damage before it becomes too severe for all material types within our collections. We are also prepared to deal with newly recovered artifacts in the future as they are excavated from anaerobic environments should they pose similar problems. The rapier and its hilt fragments are still undergoing their alkaline baths. Conservators will continue to carefully monitor it and other iron artifacts as they progress through the process. We will be sure to post an update to our social media platforms when the rapier is fully conserved.

March 20, 2024. Natalie Reid, Staff Archaeologist

In October of 2023, Jamestown Rediscovery archaeologists completed the excavation of an early 17th-century well located just north of James Fort. The team believes the well dates to ca. 1617 and was constructed for Deputy Governor Samuel Argall. Notably, the well matched another well located 200 feet to the south: both were brick-lined and the circular structure on both was constructed using whole bricks separated by small, hand-made triangular brick wedges. In total, eight of the 24 courses of brick had to be removed in order for archaeologists to be able to reach the bottom of the well. Many of these removed bricks displayed intriguing marks and deformities that we wanted to investigate more thoroughly. Luckily, Colonial Williamsburg’s experts in Historic Trades were only a few minutes away. In February, CW Masonry Trades Manager Josh Graml, Journeyman of Masonry Trades Kenneth Tappan, and Brickmaker’s Apprentice Madeleine Bolton visited Jamestown to examine bricks from the Governor’s Well.

Having made thousands of bricks just like the 400-year old examples in front of them, Josh, Kenneth, and Madeleine were able to reverse-engineer how the bricks had been altered, damaged, or deformed. Their lived experience gave them unique insight into the life of a brick before it enters the kiln, and how the specific timeline of molding, drying, and firing creates a chronology of markings on the brick itself. Here are a few examples:

There are striking differences in the surfaces of these two bricks. One is smoother (on the right) while the other is much rougher (on the left). The smooth face of the brick on the right indicates that it was laid on dry ground to dry. The rough face on the other brick indicates that it was laid on wet ground to dry.

Parallel lines appear on many of the bricks. The CW team theorized a potential cause: when transporting bricks to begin the pre-kiln drying phase, a still-malleable brick gets stacked on a brickmaker’s wheelbarrow called a hackbarrow. The brickmakers theorized that the lines on the bricks may have been made by the slats that make up the bottom of the hackbarrow.

Chopping whole bricks into triangular wedges to create a rounded structure such as a well or arch is a task often given to novice brickmakers, both 400 years ago and today. Josh, Kenneth, and Madeleine laughed while reminiscing on their own experiences doing just this!

While talking about their early careers in brickmaking, the brickmakers were also drawn to bricks from the Governor’s Well where perhaps a less experienced or hurried brickmaker flung the too-wet clay out of its mold, causing it to warp. The high number of warped bricks could potentially mean that the brickmakers were rushed for some reason or another.

Though it may seem like a small detail, the motivations of the brickmakers and how quickly they had to build the well could speak volumes about the priorities set by the colony’s leaders. For a structure like the Governor’s Well, constructed ca. 1617, these details become hugely important when considering how little was recorded about Jamestown during the 1610s.

Ultimately, these new insights provide tremendous aid in the continuing research of the Governor’s Well. Though the excavation of the well is finished, the analysis of the artifacts—including each and every brick—has only just begun. In learning more about the well’s bricks, Jamestown Rediscovery staff are able to piece together the story of the well, including hints of the political climate of the colony during its construction. Additionally, working with colleagues from Colonial Williamsburg enables the team to not only compare the well with other brick structures on the island, but also with brick structures across the entire Historic Triangle area and beyond.

Stay tuned as we continue to learn more about the incredible artifacts excavated from the Governor’s Well.