Features of Interest as Archaeologists began a new excavation south of the Armoury.

About 5 weeks ago, at the outset of our spring excavation, we posted the picture above.  It was intended to help readers  see what we saw…areas that piqued our interest as we resumed exploration of a large pit discovered in 2012.   The dotted lines indicate differences in soil color.  The question marks identify those things we did not yet know, but hoped to learn before the end of April.  Now, with the “Sawpit Excavation” behind us, it’s time for some preliminary reporting on what we’ve learned.    

The image below is our “after” picture, illustrating what was found beneath all of those question marks.  As you can see, we called the “shapes” pretty well.  The “large pit” turned out to be smaller than expected:  12’ x 16’ instead of 12’ x 20’ (probing is not always an accurate indicator of a pit’s extent!).  We know that it has a rectangular shape, although you may notice that we left ¼ of the fill for future archaeologists to explore with their improved technology and different questions.    The pit measures about 3.5’ deep. 

Fully excavated features at the end of the project, 5 weeks later. What is it? Read on…

None of these characteristics is inconsistent with our original theory that this is a sawpit- a pit dug into the ground to enable pairs of sawyers (with a pit saw between them) to cut long plank.  Master Carpenter Garland Wood believes that 16’ is long enough to serve the purpose, and that the width of this pit would have accommodated at least 2 pairs of sawyers.  We hypothesize that, under pressure to the Armoury complex quickly, carpenters may have opted to prepare materials on-site.  

Recent excavation showing both pits in relation to the rest of the Armoury yard.

While this interpretation may still make sense, there are some unanswered archaeological questions.  If this is, indeed, a sawpit, where is the evidence for a framework?  At the very least we expected a trestle…represented in the ground by postholes… to support a platform above.  More significantly, where is evidence for a cover? We have learned through cruel experience that a single hard rain can spell collapse for a hole such as this, and yet the sides of the pit are straight and crisp indicating that they were never exposed to the elements.  Did later construction obliterate the posts we were looking for? Admittedly, our confidence in the sawpit interpretation waivered a bit during the course of this project. 

And then we found a second one. 

Just to the west of the first sawpit, the second pit looked (initially) to be a basin-shaped depression.  The upper layers produced large ceramic fragments, principally (and strangely) chamber pots, and below that,  quantities of brick rubble.  As excavation proceeded, the basin became a neat rectangular hole,  straight-sided, flat-bottomed, and with a drain cut through the center to channel rainwater. Unlike the first pit,  it exhibited the expected postholes—two on the east and two on the west—indicating that a trestle was supported over it.  While our 4 man sawpit might remain in question, there is no doubt that the small hole is a 2-man sawpit.  And (of course) there is guilt by association. 

A functioning sawpit.

At present we feel comfortable interpreting both of these features as sawpits, though there are still some details to work out.  We know that they were not there at the same time.  The smaller pit cuts into the larger one, making it more recent. The bottom layers of the larger pit are filled with trash from the Armoury: clinker, gunflints, and half-completed iron objects, suggesting that it had become a handy trash receptacle by the time the Armoury started to function.  That the smaller pit comes later is verified by the fact that there are very few “industrial” artifacts in it.  Instead, the fill consists mostly of household refuse.  It is possible that the brick rubble comprising its fill  represents the demolition of a house shown standing on the Frenchman’s Map (1782) just to the south.      

What else did we find? Although we are still a long way from having a clean and completed inventory of artifacts, we have formed some impressions of what was in the pit.  There were certainly lots of animal remains ….not just the butchered remains of Armoury meals, but the articulated skeletons of a cat, and what appear to be two ducks and three roosters.  Given that they were “whole”, it is unlikely that they were eaten.  Some readers remember that last year’s excavation produced 6 dog burials.  And so begins our next round of “question marks.” 

Here are some other artifacts that we stopped to photographed along the way…..

-Contributed by Meredith Poole, Staff Archaeologist.

Plan for the Reconstructed Armoury (Tin Shop shown farthest right).

Three-and-a-half short months ago our Historic Trades team raised the frame for the new Tin Shop.  Some of you may  have braved the cold and the pre-holiday rush to be present for that event.  It may seem hard to believe, but in the coming week construction on the Tin Shop draws to a close.  Once sealed with a coat of tar paint, the Tin Shop will join the Blacksmith Shop, the Kitchen, and the North Storage Building as part of the rapidly growing Armoury complex. 

This week’s gallery offers a look at recent activity on the site.  As you will see, the action is shared among the carpenters, joiners, and brick masons who have so ably reconstructed 4 Armoury buildings to date (3 remain!), the researchers who help to inform that work, and the many trades- and interpretive staff whose extraordinary skills and knowledge bring the Armoury to life.   This is truly a team effort!  

Image of the current excavation area, looking north.

Recently, the Roving Webcam has been covering early weeks of the 2013 archaeological season at the back (or south end) of the Armoury property.  This spring’s project picks up where archaeologists left off last fall, with the exploration of a large, rectangular pit feature that we are tentatively calling a sawpit.  Whether or not that identification holds water remains to be seen.    

Archaeological excavation is not always easy to decipher, and so for those of you wondering what has been taking place over the last three weeks, the following serves as a bit of explanation.  The image above (looking a little like a Rorschach test) shows the site as it appeared on Friday afternoon, before the late March snow and rain. (We hope that it will eventually dry out so that it looks this way again!).  You should be able to see a variety of colors…predominantly brown, but with some variation.  Below, the same image identifies modern “features” that have been removed in recent weeks as archaeologists have made their way down to earlier chapters in the site’s history.  Among the intrusions are 2 modern fencepost holes filled with cement, remnants of diagonal trenches dug in 1941 as excavators engaged in early efforts to find brick foundations, and two backfilled archaeological units dating to 1975.   

Modern features identified.

This week we will turn our attention to a few interesting forms that that have just appeared.  If your eyes are quite good, you may be able to use the unmarked image at the top of the page to identify the outline of the remaining “sawpit” fill extending northward from the 2012 sand backfill (hint: it’s easiest to see along the east side where it seems to be lined in black).  If not, we have dotted the edges (in the image below) of what we believe to be the pit’s extent.  Over the course of the next few weeks we will be excavating the fill, layer by layer, to see if the “sawpit” diagnosis seems plausible, or whether this hole proves more cellar-like in the final analysis.  

Areas of current interest.

Also intriguing is a rectangular area west of the sawpit which appears to be filled with brick rubble and artifacts.  In the image above it is marked with a “?” ….which sums up what we currently know about this feature.  It was a mystery in 2012 when we encountered the same “shelf-like” extension cut into the side of the pit (see white arrow, in the image below).  Perhaps the next few weeks will provide  answers.  As we get deeper… literally, and figuratively….into these questions, we will provide updates on what we’ve found, and what it tells us. 

2012 excavation showing the “shelf” cut into the side of the sawpit.

Meredith Poole, Staff Archaeologist.

There are two questions that seem to be on the top of readers’ minds today.  The first is “where is the roving webcam located?”  This morning we moved the roving webcam to the back of the Armoury property where it is focused on the new archaeological excavation (started on Monday), and on the brick masons laying the foundation for the Armoury’s workshop.  The graphic above should give you a sense of the camera’s position. 

The second question is about Eleanor, the Armoury cat, and how she is handling the changes (and the foul weather) all around her.  As you can see from the images below, Eleanor is faring quite nicely.  We should all experience stress in such comfort!

The longer we work at the Public Armoury, the broader the pool of experts from which we can draw! In this blog post we introduce a new author, Tim Sutphin, Director of Music, Military, and Evening Programs.  After the digging is done, and once the hammers are quiet, it is Tim who is responsible for introducing additional elements of interpretive programming to the Historic Trades activities already established on the site.  Interpretive programming uses the combined resources of historical research, archaeological research, trades research, and program and presentation skills.  Tim joins the discussion and shares his insights as program developer as we move into program planning and implementation phase.  

Preparing bread dough at the Armoury.

As we enter the last phase of construction at the Armoury site, our thoughts turn to programming, and how it will expand and evolve on this dynamic site.  There is already a great deal of activity at the Armoury: the blacksmiths have been working in the new shop for nearly a year now, and have given it an appropriate “lived in” look.  Foodways staff is bringing the kitchen to life by preparing simple working class-fare several days a week.  The hearty stews, and bread from the bread oven offer a great contrast to elaborate meals prepared in the Palace kitchen.  But there is more activity in store. With completion of the tinsmith shop anticipated for mid-April, we are interviewing applicants with hope of having an active tinsmith shop by mid-summer.  This will add a new trade to the Colonial Williamsburg Historic Trades program.

Making musket balls on the Armoury’s Opening Day.

Moving into spring and summer, guests will see increasing military activity with members of the Military Programming staff carrying out their work at the Armoury.  The Magazine’s armourer will be cleaning and repairing muskets on the site several days a week, and will be joined by other military interpreters casting lead bullets, doing maintenance work on artillery and, with the help of guests, transferring muskets between the Armoury and Magazine.

Throughout the summer and into the fall, carpenters will be constructing a workshop and a storage building on the south end of the lot.  The workshop will be a multi-purpose building, used seasonally, and for special projects.  One such project is the production of gun carriages, which involves the wheelwrights preparing the carriage’s wood components while, next door, the blacksmiths make iron hardware to assemble and mount the guns.  A gin will be built in the yard, enabling workmen to lift heavy artillery pieces on and off of the carriages.

We will include other military activity that we know was carried out at the Armoury, including leather work, some canvas work, and even the making of bone button blanks for uniforms.

Activity on the site will be increasing as we move into the spring season, with additional activity in the summer and fall.  Our anticipated grand opening will be in mid-November.  Stay tuned as programming evolves.

- Contributed by Tim Sutphin, Director, Music, Military and Evening Programs,  and Kenneth Schwarz, Blacksmith, Master of the Shop.

The kitchen’s clay floor soon after installation.

There is some activity in the Armoury kitchen this week that may attract viewer attention.  The clay floor is being pulled up, and a new clay floor is being applied.  In this blog post Matt Webster explains the process, why we’ve had to undertake this step, and what we’re learning about 18^th-century construction methods. 

The Armoury is one of those unique projects that you wait and hope for; it combines research, traditional materials, and construction techniques all in an effort to reconstruct a historically accurate site.  It is an opportunity to apply years of research and see how these buildings come together, and how the materials function.  Sometimes it works perfectly, but sometimes we have more to learn and need to adjust our approach.  The Armoury kitchen floor is such a case.

As discussed in an earlier blog post (“Installing the Kitchen Floor or ‘How Do We Know?’” -March 31^st 2011), portions of the Armoury kitchen floor were uncovered by archaeologists in 2010, allowing architectural historians to take and analyze samples.  From this we learned that t the mixture for the floor was 4 parts clay, 2 parts sand, ¾ parts quick lime, and 1/10 part brick dust. What we do not know is how the material was applied or if there was an additional finish application to the surface.  There simply was not enough evidence to help us answer these questions.

Now, two years after the floor was placed, we have seen how it has worn which has helped answer some of our questions.  When we first installed the clay floor, it was done in several sections.  This allowed us to make smaller batches of material and made finishing the surface easier.  Observing the wear on the floor, the junctions between the sections have deteriorated quickly.  The floor is deteriorating much faster in these areas, indicating that when the original floor was placed (sometime around 1760), it was done in one large application.  The large batch of material was dumped onto the floor and troweled as one uniform surface, not several independent sections.  We will be removing approximately three to four inches of the current floor and replace it using this technique.

The second problem we have seen is the floors inability to withstand hundreds of shuffling feet.  Granted, one year of visitation likely equals the amount of traffic the original floor saw in its entire lifetime, but we still need to solve the problem.  The floor samples collected by archaeologists are likely not the finished surface, but close to the bottom of the original floor, so we would not see evidence of a finish.  It is also unlikely that traces of any finish materials would have survived even if we had the upper layers.  Period recipes list primarily organic materials such as egg whites, blood, and linseed oil applications for finishing floors.  All of these materials would have deteriorated and any trace would have been lost long ago, even if the upper layers had survived.  We have decided to apply linseed oil as our finish surface, and are currently testing a panel to see how it will work.

It is simple to look at the written recipe or the results of the analysis and say we know how a period appropriate floor was made.  It is not until we actually try to replicate the process that we truly understand how it was done.  That is what makes this process so fascinating and a learning experience for all of us.

Contributed by Matt Webster, Director of Historic Architectural Resources

Update: February 25, 2013

Jason, Josh, and Kenneth are making rapid progress on the kitchen floor.  The old floor has been removed, new clay has been added, leveled, and smoothed, and soon it will be sealed with linseed oil finish.  What follows is a long period of drying…not the stuff of exciting webcam coverage.  The images below should help you to visualize the progress so far. 

Kenneth and Josh applying clay to the floor.

Josh smoothing the clay.

The new surface.

Left to right: The north storage building (here, not yet painted), main Armoury building, and kitchen. The storage building is painted with a (simulated) whitewash, the Armoury with white oil paint, and the kitchen with tar paint, reflecting differences in their periods of construction and status.

At the end of November, as the north storage building was receiving its final touches in view of the webcam, Kerry commented:  “Nice to see the paint on the new storage building. My question is:  did smaller structures not use beaded siding like those of larger ones?  Also there was much activity about the paint on the kitchen. Is the white correct for these small buildings? And what color will the tin shop be?”

Edward Chappell, the Shirley H. and Richard D. Roberts Director of Architectural History, supplies the following answer:

The varied finishes on the outside of buildings in the Armoury and Tinsmith shop complex reflect the varied dates and functions of the original buildings.  Hand-manufacture and importation of materials made oil paints expensive in the 18^th century.  Pine tar and limewashes were more economical choices.

Pine tar strengthened with iron-oxide red pigment was a common choice for riven (split) clapboards and shingles.  We have found it microscopically on the earliest frame parts of the Thomas Everard kitchen, applied before the walls were rebuilt with brick in Everard’s tenure.  The only painting contract to survive from early Williamsburg calls for tarring the roof of St. George Tucker’s kitchen with pine tar and iron-oxide pigment.  Applying this information to the Armoury property, we know that the tinsmith shop and the [Anderson] kitchen were built before the Revolutionary War and the rest of the Armoury complex.  They are the kind of ancillary buildings most likely to have been tarred.

The main Armoury building and workshop (not yet reconstructed) are perceived as being better-built during the Revolution than the smaller contemporary buildings like the storehouses.  Here there is a distinction is between higher and lower status, with oil paint on the principal buildings (the main Armoury building) and whitewash on the smaller and more cheaply built ones (the storehouse).

You may read more about the use and study of early paint in Williamsburg in the three first chapters of Architectural Finishes in the Built Environment, edited by Mary Jablonski and Catherine Matsen (London:  Archetype Publications, 2009).  The story of Williamsburg paint will be given more attention and broader context in The Chesapeake House, a book by Colonial Williamsburg architectural historians on early building in the region, to be published by University of North Carolina Press in March.

                                                                                     -Contributed by Edward Chappell, Director, Architectural and Archaeological Research

As blog followers recall, the Tinsmith Shop is part of the growing Public Armoury complex and, once completed, will allow guests to explore the Revolutionary War period activities of tinsmiths. 

But the Tinsmith shop was not always a tinsmith shop. 

Fig 1. Tenement in 1776 – A fence divides lot 17 from 18. The building in the background is the Anderson kitchen. Note the area behind the kitchen is probably fenced-in kitchen-garden.

Archaeological evidence indicates that the building was constructed sometime around 1760, long before the Public Armoury was a necessity.  It was, in all likelihood, constructed as a small, rather unprepossessing tenement, built right up to the fence-line on the east side of lot 17 (Fig.1).  Located on the side of a ravine, the tenement occupied a marginal (and not the most salubrious) location.  All evidence points to a low status building of cheap construction, with little by way of expensive treatments.  The close-up of the reconstruction (Figs.2 & 3) shows how the tenement may have appeared in 1776, before work began on the Public Armoury.  The building is a small one-room single-story tenement with very simple sliding shutter windows, and a basic set of steps leading up to the front door (Fig.2). 

Within the next three years the building underwent a significant transformation.  As Armoury construction began (in 1778) on neighboring lot 18, the fence separating the two lots was relocated, drawing the tenement into the complex for use as a Tinsmith Shop (Fig.4).  The virtual reconstruction reflects that change.  Larger windows were added to allow more sunlight (Fig.5).  The interior images show how work benches were placed under these expanded windows to maximize usable light (Figs.6 & 7).  The fireplace not only kept the workers warm in winter, but was used year-round to heat coals for the braziers used in the manufacturing process.  The shop’s tin products:  kettles, coffeepots, saucepans, and speaking trumpets, were made from sheet metal which was cut with shears, bent to shape, and the joints sealed with solder melted on a soldering iron which had been heated in a brazier.

Incorporation into the Armoury complex also meant greater security for the Tinsmith Shop.  It is thought that the whole Armoury complex was bounded by a “secure perimeter” protecting the valuable arms and materials within, and perhaps serving as a barrier to escape by the prisoners-of-war who worked there.  A diagonal fence was constructed, connecting the Tinsmith Shop to another former tenement on lot 17, and contributing to the Armoury complex’s secure core.  Archaeological evidence indicates that by 1779 steps on the west side of the Tinsmith shop had been removed, inhibiting access to the front door since this entrance was now outside the secure perimeter.  This process of controlling access to the site and buildings may help to explain why the Armoury shop was built between the kitchen and Tinsmith shop, rather than on the more open areas of the site to the south.

Removal of the steps then raises the question: how did the workers get into the Tinsmith shop?  With the original door closed, and the chimney located on the building’s south side, two options remain.  Either a new door was opened on the north side or an opening was cut on the east side, connecting the Tinsmith Shop directly to the Armoury building.  Both hypotheses have their merits, their drawbacks, and their supporters, and unfortunately no definitive archaeological evidence has survived for either possibility.  We have elected, therefore, to physically reconstruct the Tinsmith shop with a door on the north end–the configuration that allows our guests the easiest access from Duke of Gloucester Street, and our historic trades people the best traffic flow through the shop.  In the virtual model, however, we will show the alternative interpretation, with access through the Armoury.  That’s the beauty of the virtual world: we can visualize multiple hypotheses and the evidence supporting them! (click on images below to enlarge).

Contributed by Peter Inker, Digital History Center

Watch now.

Sills placed on the Tinsmith Shop foundation.

Webcam watchers will notice that they have a freshened perspective on the Armoury project.   The roving camera is now trained on the Tinsmith Shop where Historic Trades Carpenters, this morning, capped the foundation with sills and floor joists.  Frame-raising is scheduled for next Friday, December 21^st.  Beginning around 11 a.m., and wrapping up at around 4, Garland and crew will raise the west, south, and north walls, set the plates, and place the rafters…a full day’s work. 

If you happen to tune in between now and then, you may catch the carpenters building the east wall.  To accommodate the very tight fit between the Armoury and Tinsmith Shop, and to avoid damage to the chimney, this east wall will be built in place, before the rest of the frame is raised.   

So as you find breaks in your busy holiday schedules, please  join us either in-person or via the roving webcam to admire the progress being made at the Tinsmith Shop!

Early in November archaeologists began wrapping up work at the back of the Armoury lot.  It has been a long and challenging field season, culminating in a heavy soaking by hurricane Sandy (and earlier, by some of her lesser-known cousins).  The image below, a final photo taken on November 2^nd, after the final clean-up, shows some of the fruits of our labors.  It should also help you to visualize the site’s layout. 

Armoury Fenceline.

Curiosity about fences was the primary motivation for this summer’s project…not just where they stood and how to rebuild them accurately, but how the margins of the Armoury were treated.  Were they well-maintained or left untended?  Did workers cram activity onto every inch of available space, or did the Armoury’s busy core peter out toward the edges?  And with 40 Armoury workers employed on half an acre, how seriously was the property boundary regarded, anyway?  Is a fence really important during wartime?  And what was happening on the other side of the fence?  Was the grass really greener?

The archaeologists in the image (at left) are acting as props.  They are standing in postholes that once supported a fence at the western edge of the Armoury property.   To the right of this human “fence” is the Armoury lot.  To the left is Lot 17, the mid-18^th century ownership of which is still unclear.   You’ll notice, as well, that there are lots of “unmanned” holes. Over the course of the summer we encountered, excavated, and recorded many, many, (many) postholes representing numerous incarnations of a fence that marked the same boundary over hundreds of years.  Only one of those fences was standing between 1778 and 1780, when the Armoury was in operation.  Sorting out which is the “right” fence..the one that James Anderson might have claimed… has been a considerable puzzle, and one that we think we are close to solving.    

Perhaps the most curious aspect of this dutifully maintained fence is that it was virtually disregarded by those who worked at the Armoury.  This summer’s excavation showed that both sides of the fence were used for Armoury activities.  On the west side of the boundary, on land whose ownership in 1778 has not been clearly established, we discovered a large pit, now half-excavated, but tentatively identified as a sawpit.  We hypothesize that it was used by teams of sawyers to prepare plank for the Armoury’s construction…one sawyer above the pit on a platform, and another below, with a pit saw in play between them.  Once construction of the Armoury was complete, the pit was no longer needed and was filled.  Indeed, we found that the pit was filled soon after its abandonment with an assortment of the Armoury’s waste… clinker from the forges, French gunflints from the repair of weapons, and partially completed iron objects.  

Next summer we will return to complete the excavation of this feature, and to determine whether our working hypothesis about its function holds up to any additional information we might uncover.  In the meantime, we have filled the sawpit with sand to stave off any mid-winter collapse…and to provide us with a healthy warm-up activity for next spring!

Another “Armoury related” feature is a privy that was discussed in a recent blog post (as you may recall, it was the raspberry seeds in the fill that gave its function away!).  Like the pit feature, the privy, was located on the adjacent property.  And like the pit feature, it contained Armoury garbage.  So it seems clear that the Armoury was making use of space beyond what James Anderson owned. 

One final discovery that bears mentioning: the southern end of the Armoury lot was notable for the number of 18^th century dog burials it revealed.  The remains of seven dogs were uncovered along the Armoury fenceline…on both sides of the fence.  Three of these (and likely a 4^th that was not completely excavated) were found in discrete, purposely dug graves.  The remains of three other dogs were found in a single pit.  They ranged in age from less than a year, to mature, to “older-mature” (with evidence of arthritis, and significant tooth wear).  In our experience, it is uncommon to bury dogs in the 18^th century, so this unusual treatment.  The dogs have been removed from the site, and are now being studied in the Zooarchaeology Lab. 

Like many archaeological endeavors, the Armoury project has answered some questions, and raised others.  The end of the excavation hardly signals the end of the research, however.  Over the winter, artifacts recovered from the site will be washed, numbered, counted, measured, and evaluated.  We will use them to help us to affix dates to soil layers and to postholes, and to help us to identify the function of certain features.  Iron artifacts encrusted in rust (and there are many, given the nature of work on this site) will be x-rayed and conserved in an effort to better understand the nature and variety of work taking place at the Armoury.  Soil removed from pits of all sorts will be tested for plant remains and for chemical signatures that provide information about landscape and land use.  Individual field maps will be digitized and transferred to site maps where larger patterns may become evident.  And the dogs…all of those dogs….will be studied some more. 

In closing, and in the spirit of Thanksgiving, it seems appropriate to express our gratitude for the dedicated and determined archaeologists and interns who helped to excavate the Armoury site this year, often under adverse conditions: Lucie, Matt, Wes, Jeff, Andy, Meredith, Sarah, Loretta, Walt, Mike, Ron, and Dessa.  We are grateful for the opportunity to conduct this research, and for all of you whose interest and investment in Colonial Williamsburg allows us to move forward.  Most significantly, we are grateful to Mr. Forrest Mars for his commitment to the Armoury project.  Thank you….and Happy Thanksgiving!

Meredith Poole, Staff Archaeologist