Season of Frustration

I have unpardonably left you, dear reader, in a long state of suspense, with our boat engine on fire and facing considerable uncertainty in how we could continue to work, stranded as we were on Paget Island...

The Bermuda Government, in the form of Kenneth Bartram and Kimberly Bean of YSR and Paget Island's own Captain Andre, saved the day (or rather the last two weeks) by giving us discounted charters of the Paget Island boat Sin or Swim to continue to take to and from Smith's Island and the St. George's Lab each day. Our ability to keep working with Bermudian volunteers became complicated, but many came out anyway - having become hooked on the thrill of finding artifacts, getting dirty, sweating profusely, and our usual fever-dream stream of consciousness working chatter even after they earned this year's unique field school shirt. The students enjoyed rotations through the lab working with Valerie in the final week.

With transportation sorted out, we focused on completing the mad scramble to get our planned excavations completed at Smallpox Bay and Oven Site by finishing off feature excavation and mapping at Smallpox Bay while taking all six units at Oven Site down to bedrock in the hopes that a nicely spaced line of post holes would be revealed.

Smallpox Bay

When the dust settled at Smallpox Bay, we frustratingly did not find a decisive end to the Big Building but rather post holes on the eastern edge of our easternmost units that indicate the structure keeps going. We now know that it is at least 42 feet long with the same regular post spacing at three-foot intervals.  Post holes in the eastern units are shallower than those previously excavated, which aligns with the sharp bedrock slope in this area: historically, the soil was thicker here, so Moore's workmen did not have to intrude far into bedrock to seat wall posts at the same 2-3-foot depth.  

Indeed, the final post hole identified in the southern wall line was the first one discovered to date that did not intrude on bedrock at all:  there was both sufficient depth of soil AND survival of intact stratigraphic layers below the 1970s plowzone to preserve a post hole cut and fill.

The cut was entirely situated in a prehistoric soil layer heavy with semidissolved limestone nodules and entirely devoid of artifacts. A heavy rain revealed this ancient was  unlike all Bermudian soils I have ever encountered: it was water impermeable. The excavated post hole became a puddle that would not drain. 

 Although seemingly mundane, this find profoundly reminds us that whatever post hole patterns we may find through excavations, they will represent an imperfect map of the town's buildings and layout. It is now clear that the ground surface in 1612 was higher in many areas than it is today, and that natural erosion forces, clearing and occupation during the various 19th-century epidemic periods, and especially high-impact mechanical clearing and plowing in the 1960s through 1980s has compromised our ability to fully discern the First Town and its various structures. Exact locations of early 17th-century artifacts not found within sealed feature contexts are also now known to be more suggestive than precise, given both horizontal and lateral plowing displacement.

Elsewhere on the north wall line, we uncovered a horizontal sheet of limestone rubble with quicklime concentrations that abutted a post hole - an intriguing find that may either represent an early floor of some sort or evidence of daub making and application, where quicklime and rubble construction material fell abutting the inside of the north wall and was packed down under the feet of workmen. It is located in a low-lying area relative to the western portion of the building, which may account for its survival; similar deposits occurred all along the wall's interior but have been swept away elsewhere.  Because the full contours of this sheet deposit were not defined, excavation will have to await a future season. Excavation of units to the north will also be necessary to determine if the sheet deposit extends to the wall's northern exterior, where daub was also applied.

We had planned to open a large area within the interior of Big Building but our small team size meant only two new meter-square units were excavated. These revealed a rough post hole that was never apparently finished, a large, not fully exposed feature in the room's center,  and the survival of a thin apparently undisturbed soil layer with numerous 17th century artifacts just above the bedrock surface. Although our focus has been on defining the walls and dimensions of this building first, excavation of the floor will also be important for revealing the building's use and, should post lines be discovered, interior partitioning that would problematize its hypothesized attribution as the town's church.

Excavations of three units in Locus C, far to the west of the stone ruin at the edge of the 1970s field, carried on to the bitter end, past quitting time on the last day. The cut and mortared building stones visible on the surface were found to be "floating in space" rather than protruding from a buried foundation; they were pushed there by plow or bulldozer from some other location but do not appear to be part of or related to the "GR" stone ruin.

 

The plow zone layer in the three units was remarkably thick, 40 to 50 cm, and contained a high concentration of artifacts spread across the 17th, 18th, and 19th centuries. 

The most notable find was a sherd of colono-ware, an unglazed, low-fired vernacular earthenware whose manufacture and use is usually associated with Africans, African-Americans, and Native Americans. Although this is the first example of colono-ware recovered on Smiths' Island, it has been found in 17th and 18th-century contexts at several sites in St George's. Intriguingly, it is highly unlikely that colono-ware was manufactured in Bermuda (given the rarity and poor quality of local clay), so recovered sherds reflect the deliberate importation of (most probably) African creole-made ceramics from Caribbean and southern coastal North American enslaved communities based in ports or plantations.

All in all under Ewan's excellent supervision, the SIAP team excavated 55 layers and 15 features in 14 meter-square units at Smallpox Bay this season, yielding approximately 6,000 artifacts. Artifact counts were lower in the Locus A Big Building area as one moves east, suggesting the core of the 1612 Town lies west of the "GR" stone ruin. High artifact yields in the Locus C units reflects both the thickness of the plow zone from whence they came and proximity to multiple occupations nearby, including an 18th-century presence that is largely absent in Locus A and B to the east.

Detailed databasing and analysis of the Smallpox Bay field records and artifacts will be ongoing throughout Fall 2026 at the University of Rochester.

 Oven Site

The quest to find the elusive Captain's House, overseen by Xander, also ended in frustration, through no fault of his own. The six meter-square units bracketing the 2017 trench with post holes proved heavy-going during excavation due to thick roots, a pronounced slope, thick stratigraphic layers, highly uneven bedrock surface, and ultimately, the lack of post holes where they were supposed to be! It did not help that we had a lot of rough weather in the last week, including some storms that formed out of the blue nearby on days predicted to be fine.

Unlike Smallpox Bay, Oven Site's stratigraphy is remarkably undisturbed. The artifacts recovered in these units all dated to the 17th and very early 18th centuries and were consistent with those found in the cistern and Oven Site kitchen - including several additional chert lithic flakes. Indeed, much of the assemblage was architectural material related variously to the burning of the Phase I kitchen, construction of the Phase II kitchen and water cistern, and repair and desctruction of the outbuilding surrounding the cistern and included large numbers of fully or partially fired quicklime, bricks, and several highly detailed daub fragments.

The soil-bedrock interface in the northern units was highly irregular, topped with a dense concentration of large limestone rubble cobbles. The bedrock surface in several areas that were initially thought to be features was blackened and heat altered, suggesting that this area had been excavated down to bedrock and used to support a makeshift limekiln - and then was naturally covered over as it became a yard surface in the later 17th century, perhaps lying between the cistern outbuilding and the main house somewhere to the west.

The post holes that were thought to be part of the main house ultimately did not align with any discernable structure and are quite puzzling. The close spacing of two well preserved and sharply detailed post holes in the 2017 trench were ultimately only associated with one additional, similar post hole a foot to the northwest and along a curving line.  These are clearly not part of a substantial structural wall and defy the ubiquitous forms of English linear architecture.  A short curved wall would be highly unusual and have no clear purpose. Alternatively, the three posts may have been successive rather than have existed at the same time, but if they relate to later repairs or reinforcement of a structure, why are they so distinct and diverge from the numerous examples of "double post holes" we have encountered at Smallpox Bay? 

Another unique feature also emerged by widening the original meter-wide trench. What appeared to be a small step with two "mini-caves" in 2017 we now see as a long linear apparently architectural feature with an unknown form and function.  

Close examination revealed that the higher "bedrock" south of the step, between it and the cistern, is actually poured cement/mortar. The "caves" actually seem to be footings for horizontal or low-angled beams or joists for a structure extending northward, with the cement poured while the timbers were in situ to secure them. A profile of the western units shows both the slope of the terrain, the step and post anchor, and sections of the post holes showing their cylindrical form and depth:

This annotated profile marks the horizontal strike of the step ridge (red), an hypothesized 3D projection of a joist or beam seated into the step (green), and the edge of the cistern (blue).

Example of a Cruck House, Stowe St,
Lichfield, GB

What could this feature be? It does not resemble anything yet found in Bermuda or early Chesapeake sites, other than a "cruck framed" building (a medieval A-frame-like structure used for barns and large halls), one of which was unearthed at Martin's Hundred in Virginia (1618-1622). The post footings at Oven Site are too small and too shallow to support even a small house, however.  

This puzzle prompts us to consider the many activites that must have occurred on early Bermudian farms and in the first decade, when newly arrived settlers experimented with various (and generally unsuited) pursuits to make money, replicate familiare European (and African and indigenous American) practices, and sustain themselves in Bermuda's different new environment.

The modest size of the anchored beams (apparently split cedar trunks of 6-8 inches in diameter) and spacing at one-foot centers suggests a framed platform that could either be covered with palm fronds or cedar shingles to form an open-sided lean-to or a bare frame for aerating and drying something, such as tobacco or fish.  17th-century frames for curing salted cod in Newfoundland could produce a similar archaeological signature and, indeed, are still used today in Iceland and Norway. Bermuda's humid climate and lack of fish species amenable to open-air curing makes this unlikely beyond the first decade of experimentation; because the hypothesized frame base aligns with the cistern, this would indicate that the cistern also dates to the 1610s (unlikely) or the frame feature and cistern are products of a more mature stage of Bermudian settlement.  

Frames for drying and curing tobacco persisted throughout the seventeenth century, but were usually situated inside tobacco barns under roofs to keep them dry.  

Detail from a 1667 French engraving of tobacco and manioc preparation in Jean Baptiste DuRertre, Histoire generale des Antilles habitees par les Francois II:419, John Carter Brown Library; SlaveryImages.org.

Oven site's 2026 excavations did not pin down the footprint of the Captain's House, but contributed greatly to understanding the farm layout as a whole. The six units excavated produced 27 contexts, including three features (one post hole, two declevities with lime burning attributes) and approximately 1,500 artifacts - all consistent with and often related to the construction, destruction, and steady use of the nearby kitchen and cistern. Two puzzling pieces of early Bermudian architecture have emerged that bear further study through future excavation. How far does the stepped lean-to frame structure extend to the west?  Are there features to the north (such as post holes in a line parallel to the stepped ridge) that can help us establish this structure's form and dimensions? Perhaps there are artifacts within the interior that might reveal its age and function.

The search for the Captain's House is becoming my "Moby Dick", at risk of developing into an obsession (although unlikely to lead to the death of my crew). We will resume the search next summer, hopefully with a much larger team and with more time to spend excavating. Irrational superstition leads me west, to the Big Yellow Tank that since 2010 has sat upon exactly the place I need to next dig. When we hurled it deep into the forest downhill in 2015, I thought it would no longer haunt my site logistics, but like a bad penny it keeps turning up...

Public tour of Oven Site in 2017. I am standing in the partly excavated cistern. The yellow tank bides its time patiently to the west.

Team Val at the Globe Museum Lab

While we raced to finish excavations at Oven Site and Smallpox Bay, Lab Director Val supervised the most complex set of operations to date, working with our most dedicated lab volunteers, Eleanor and Martin, as well as students rotating through this specialization and local volunteers. Besides the necessary task of washing, sorting, analyzing, and inventorying the new material coming in from the field daily, Val made a significant dent in SIAP's backlog of floatation samples collected in 2023 and 2024 at both sites and further re-inventoried site contexts from previous years to make them more consistent with our current recording and databasing methods. She too had to deal with rainy and hot, humid days when conditions in the musty cellar of a 1699 mansion and its adjoining alley entrance were less than ideal. 

The most exciting aspect of Val's work was in making new discoveries in old contexts through her expertise in faunal analysis. Besides differentiating mammals and fish from birds and reptiles, Val the presence of particular animals in specific contexts.  In one 18th-century feature, she found most of a rat skeleton (right), suggesting the little critter had been killed and tossed

into a shallow pit.  A re-evaluation of the Cave Site assemblage (dug 2014-2015) turned up the bones of a cane toad near the surface as well as details about the sheep (or goats), pigs, and cows found deeper down that apparently reflect the illicit feasting activites of unknown but very likely enslaved 18th-century occupants.

All too soon it was June 29, the season ended, and it was time to pack up, move out of our Paget Island home, and return to the Real World (or, in the case of supervisors and Andrew, Southampton).

Against my better judgement and because we were so short-staffed, the students left before our traditional "Fill in Friday" day of burying the season's open excavations with backfill to preserve them. Storms scuppered two scheduled fill days in early July. After bringing our workboat back to life (with skilled engine surgery), the Crystal Seas made one final mission to Smith's Island, where Xander, Ewan, Andrew, Larry Mills (going above and beyond any volunteer's duty), and I put Smallpox Bay and Oven Site to bed for another year and then fetched off all our remaining field equipment for long-term storage.

We then turned our attention to finishing up field and lab documentation, initial processing of 3D site models, artifact inventories of late-processed contexts, and writing up our initial interpretations.

We were blessed with being able to do this in the wonderfully genteel and historical setting of Cedar Hurst, an 18th century mansion in Southampton Parish as the guests of Ed Schultz, a long-time SIAP supporter, veteran shipwreck diver, and keen student of Bermudian and maritime history. We especially appreciated the anachronism of air conditioning after living in a state of nature on Paget Island.

Looking back from the end of August, the 2026 season seems like a long time ago, following personal travel in Norway (to get in touch with my roots/inner Viking) with Anna and extensive prep work for the Fall 2026 term, teaching a new graduate historiography seminar and becoming interim Director of the Digital Media Studies program. We got a respectable amount of work done with a very small group that had to contend with many rain days, high heat and humidity, injuries and sickness, no wifi, a broken boat, and unusuall swarms of mosquitos on both Smith's and Paget Islands.

SIAP will surely resume in summer 2027, hopefully with a larger group of mew students and the return of skilled lab and field veterans and dedicated Bermudian dirt/lab addicts  volunteers. 

I am enormously grateful to everyone who made the 2026 season a success, from Charlotte Andrews and the Bermuda National Trust for our lab, promotion, and public tour coordination to Chef Debbie at Somers Market, who kept us well fed and gave us the occasional Sweet Treat. We had the largest ever turnout for Bermudian volunteers, several of whom became regulars. The generosity of Edward Schultz, savvy seafaring of Andre Hollinsid, and support of Bermuda Government's YSR leaders kept us shuttling between our island home and workplaces. Smiths Islanders Samantha and Neil threw us a welcome wrap-up party and provided key medical guidance for a certain keen but injury-prone student. Ted Pagano and University of Rochester's Center for Education Abroad and Jacqui and Melissa at the History Department helped insure that the dig would happen in the first place. 

My biggest thanks goes to Andrew, Lexi, Rosa, and Rebecca for embracing the educational chaos of "immersive experiential learning" in a volatile state-of-nature classroom, and to Ewan, Val, and Xander for their professional expertise, student mentoring, new analytical insights and interpretations, and keeping me sane as we all kept everyone safe and the dig going.

A rare sighting of the Shannon Sloth in his
native environment, searching for students