With the painting of the newly installed shingles over the stone side of the Stone House, we have completed a milestone phase of the Stone House Restoration Project. As we explained in our last issue, the shingles have been painted with a red linseed oil paint mixed with copper naphthenate. In this article we will address the historical evidence for red paint on early-American wood shingle roofs and explain more about how copper naphthenate is a modern compromise to extend the life of the shingles themselves.

Historically, the red paint used on wood shingles was made of red iron oxide and linseed oil. Iron(III), also known as ferric oxide, is an inorganic compound with the formula Fe2O3. Basically, the historical paint was iron rust mixed with the oil from flax seeds. The paint we are using for our project is Ottosson Linseed Oil Paint, Iron Oxide Minimum color #LFRM. It is mixed with one part copper naphthenate to around three parts of the linseed oil paint. Copper naphthenate is a wood preservative that is the copper salt of naphthenic acid. It makes wood and other cellulosic materials inedible to insects and fungi. Historically, copper naphthenate was not available in the period to which we are restoring the house. Nevertheless, people of that time may have noticed that wood shingles abutting copper flashing did not deteriorate as fast as the other shingles on the same roof. The copper naphthenate will help to ensure a longer lifespan for this expensive reproduction roof.

Our consultant for this roof project is Mr. James Houston. He is retired from the Pennsylvania Historic and Museum Commission, where he was involved with roofing projects like our roof on the Stone House at historic sites such as the Daniel Boone Homestead and the Ephrata Cloister. Mr. Houston salvaged period examples of side-lap shingles from a structure in Frederick, Maryland that had surviving red oxide paint on them. (See photo below.) In 2008 Mr. Houston, along with his former colleague Mr. John N. Fugelso, published an article in the APT Bulletin: Journal of Preservation Technology titled “Fabricating and Installing Side-Lap Roof Shingles in Eastern Pennsylvania.” In that article they address the historical use of various methods that were employed to prolong the lives of different wood shingle roofs. They note that even George Washington had the wood shingle roof at Mount Vernon painted red. In coastal regions fish oil mixed with brick dust was applied to shingles to achieve a similar effect.

For those readers who live in our neighborhood, please make a point of coming to see the newly completed roof over the stone side of the Stone House. We hope to have the rest of the shingles delivered soon so that the roof over the log side can be finished by this time next year.

One of the greatest challenges we face in the business of historic preservation is controlling unwanted moisture in and around our structures. Water will find a way down. When it gets down and is trapped in between things made of wood, it can cause serious deterioration and rot. We have found evidence in the Stone House that there has been a history of serious water leaks where the roof over the stone side meets the south gable wall of the log addition. Sheet metal flashing has been the method used to prevent and repair leaks in this part of the house for over a hundred years. This has certainly been the case ever since the metal roof was installed over the stone side of the house. Before the advent of rolled metal roofs, wood shingles were tucked under weatherboards in places like this, where a roof met an exterior wall. The weatherboards were cut in relief to the profile of the shingles. This would leave a sawtooth pattern on the bottom edge of the weatherboards. We have an example of this kind of woodwork in the attic of the Steele & Bro. Store. (See image below.)

In this instance the weatherboards of a former exterior wall were left in place and were encapsulated under a new metal roofline that was installed above. Even though the old wood shingle roof and its supporting structure was demolished in the 1880s, its profile remains in the surviving weatherboards.

We are using this method of tucking our wood shingles under the bottom edge of relief-cut weatherboards on the south gable of the log addition. To minimize potential damage to the wood shingles over the stone side of the structure, we began by having our roofing contractor Frank Stroik, and his crew at Country Homestead, install a few courses of the side-lap wood shingles in that place where they abut the south gable wall of the log addition, leaving the remainder of the roof structure over the stone side covered with temporary tarps. This made it possible for Vintage Inc.’s crew to install the weatherboards above without requiring them to walk on any new wood shingles. After the abutting shingles were installed, they were painted with a red linseed oil paint mixed with copper naphthenate. There is historical evidence for red paint on wood shingle roofs, but the copper naphthenate is a modern compromise to extend the life of the shingles themselves. When the rest of the shingles are installed, they will also be painted with the same mixture.

The installation of the weatherboards over the shingles required careful measuring and the leveling of each board in relation to its vertical alignment. Every effort was made to ensure that the gap between the top of the shingles and the bottom of the weatherboards is about the width of a carpenter’s pencil. This allows both the shingles and the weatherboards to “breathe” and prevent water from getting trapped.

This method of weatherboard and shingle installation works because water does not run uphill. By sealing the end grain of the weatherboards and the shingles with paint, we are also significantly reducing the capillary action and absorption factor of the boards themselves. In our next installment we will unveil the finished roof over the stone side of the house. We have been looking forward to this for years.

We completed the work of fitting the gutter brackets to their respective places along the bottom edge of the roof sheathing boards. It is last thing we needed to do before installing the shingles on the stone side of the Stone House. This project was not as easy as we had thought it would be on the rear side of the building. This is mainly because the roof rafters and eaves on the back side of the building are very irregular. They have suffered the most over the nearly two hundred and sixty years of the structure’s lifetime. Additionally, when the ell addition was built on the rear of the stone side of the house in the early twentieth century, the workers removed the part of the soffit that was in the way. Everything that was missing had to be painstakingly restored while preserving the remaining building fabric. Consequently, the eaves and the sheathing boards above, to which the gutter brackets are attached on the rear side, are not completely straight or level. This is in contrast to the front of the stone side of the building. The work of attaching the gutter brackets to the front of the stone side was accomplished with the help of Dennis Clem, the proprietor of Cedar Creek Blacksmithing. Some of you may have seen the posts on our social media pages featuring photos of Mr. Clem while he was working.

Mr. Clem had to bow out from the work on the rear side due to scheduling constraints. As noted above, the work of fitting the brackets to the rear side has proven to be very time consuming. Our President Butch Fravel took over the project for the rear of the stone side. Mr. Fravel found that the irregularities of the rear side eaves and sheathing boards made the task very challenging. As the detail from the architectural drawing below illustrates, the angle of the bend below the shank that is attached to the sheathing board would theoretically remain constant while the amount of the drop below that bend would vary to insure the gradual slope of the gutter toward its downspout.

• 

Because of the irregularities in the placement of the sheathing boards, the angle of the bend, as well as the amount of drop below the bend, varied from one bracket to the next. The number of compound angles and the variables associated with them required approaching each bracket with new measurements. Some brackets had to be rebent and adjusted more than twice to get them in the right places. Mr. Fravel used a settling torch and traditional blacksmithing tools to accomplish the work. (See image below.)

• 

Once we were sure that all the gutter brackets were fitted properly to their respective places on the sheathing boards, we then took them to be powder coated on the 16^th of September. That work was done by Coatings USA LLC in Front Royal. This will ensure that the brackets will not rust and deteriorate as quickly as they would otherwise. After the brackets were powder coated, they were reinstalled in their respective places. Each bracket is uniquely marked with a lettering system that will be used to denote where they belong in their respective slots on the bottom sheathing boards. In our next instilment we will report on the installation of the shingles.

• 

On the 26th of April we received our first load of shingles for the Stone House roof. Many of our readers may recall the previous post on 5th of February 2019 titled “Long Biaxially Tapered Side-Lap Shingles.” As explained in that article, these shingles “are biaxially tapered, or wedge-shaped in profile, because they are split out of large red oak logs and then finished by shaving them smooth with a drawknife.” They are also free of sapwood and core heartwood. The only way to manufacture these shingles is by hand. They are not sawn into shape. They are split along the natural lines of the wood’s grain. This helps them not to warp or “cup” after they are installed and exposed to the elements. Because they are made by hand, they represent many hundreds of hours of hand-skilled labor. The skills and obsolete technologies used to manufacture these shingles are no longer commonly understood and practiced by those in the carpentry business, finding contractors who can make them can be a challenge. Fortunately, our consultants Doug Reed and James Houston were able to help us connect with a man named David Dauerty who has a great deal of experience making this type of shingle. Mr. Dauerty personally delivered the first load of our order for the Stone House on that day. (See photo below.)

Mr. Dauerty also delivered a punch bench that will be used in the installation process. This device ensures that the nail holes are correctly placed on the shingles. The pre-punched holes prevent the shingles themselves from splitting during their installation on the roof lath. (See photo below.)

This load of shingles will be enough to cover the stone side of the Stone House. Once we have the remaining part of the order from Mr. Dauerty, we will finish shingling the roof over the log addition of the house. The next steps we will be taking this summer will involve the necessary work of removing the old metal roof and installing the gutter brackets which will be mounted to the rafters and roof lath before the shingles go on above. This is an exciting time in the history of the Stone House, and we will share our progress on our social media pages, in our next newsletter, and on our website. Stay tuned!

Before the decision was made, we struggled with the evidence. The question seemed relatively simple: “Should we install gutters on the Stone House?” The earliest photograph we have of the log side of the house seems to show that there were no gutters installed on that side of the structure at that time. Nevertheless, that does not prove there were no gutters installed at an earlier time, nor does it preclude the possibility of there being gutters on other parts of the house at that time, or at an earlier time in the history of the structure.

Historical written sources and physical evidence of gutters on circa 1830 houses in western Virginia are hard to find. It is likely that those who had gutters on their homes in Newtown/Stephensburg during that period had installed wood gutter systems that employed boards fixed in a V-shape. Some wealthier homeowners may have had these V-shaped wood gutters lined inside with tinplate iron sheet metal. At first, only the wealthiest members of society could afford gutters made entirely of tinplate iron sheets. Terne sheet metal is a form of tinplate that is thin iron or steel coated with an alloy of lead and tin. The earliest half-round metal gutters were made of terne sheet. They were also rolled on both edges. (See the image below.) Our historic structure consultant Doug Reed has a collection of antique gutters and downspout segments that he has saved over the course of his career working on old buildings. We would like to thank him for sharing this collection with us and allowing us to take the photograph featured here.

The way gutters were held in place on buildings in the 1830 period included wrought iron brackets. Our President Linden Fravel had seen examples of period wrought iron gutter brackets during a visit to Old Salem in Winston-Salem, North Carolina. In turn, we contacted their Museum of Early Southern Decorative Arts to request photographs of examples from their collection. Among the examples they had was a gutter bracket designed to be mounted under the shingles and attached to the side of a roof rafter. (See images below.)

In the end we decided that in the interest of preserving original building fabric and protecting the restored parts of the structure, it would be best to install period-appropriate gutters on the Stone House. Those reproduction gutters have arrived. (See image below.) We are having gutter brackets (similar to the one pictured here) made by Williamsburg blacksmith Mark Sperry.

As we await the delivery of our 4,900 handmade shingles, there is good news and progress to report. In the course of the work to prepare for the installation of those shingles we are restoring features that will touch on or interface with that new roof. This includes the tops of the north and south gable ends with their rake boards, beaded weatherboards, and windows. The weatherboards we installed on these gable ends are copies of the original weatherboards we discovered under the roof of the circa 1869 shed addition behind the log side of the house. The windows are based on the evidence recovered by Doug Reed during his forensic investigation of those areas of the house. They are also based on other surviving historical examples of windows from the late eighteenth and early nineteenth centuries that have survived locally in this town and region. Nails and nail holes in original studs were the only clues we had to work with for the south gable window. With the north gable window we were fortunate to combine nail hole evidence with “ghost” lines on the original studs. These lines showed weathering and oxidation shadows where the original window frame and its trim used to be.

We are also restoring the soffits, fascia, and crowns at this time because of their relationship with the roof. One of the biggest challenges with the soffits and crowns is the way the building’s walls have settled and warped over the last two-hundred-plus years. While they were originally built plumb and square, they are now quite out of line in some places. This is particularly the case with the east wall of the stone side of the house. The evidence we have discovered implies that there was significant work done to this east wall of the stone side of the building during the just after first quarter of the nineteenth century. This is also the wall that we have had to construct buttresses alongside to bolster it until we can build a lean-to porch against it as a permanent solution. The bulges and lack of straightness across the top of this east wall are most visible where the crown and soffit meet the top of the stone. (See photo below.)

We also have restored our chimneys, and there is an interesting story about the main chimney top in the center of the house. Historically chimney tops needed maintenance every 30 to 50 years. Hot, dry smoke meeting cold, wet air produces acidic condensation that dissolves mortar. Combined with the freeze-thaw cycle, this moisture causes masonry to deteriorate. For this reason we were fairly certain that the brick top of the main chimney had been reworked since 1830. During our initial investigations we could not tell if that main chimney top had originally been completed in brick or stone when the roof over the log addition was raised in 1805. When our mason Edward Ashby and his men were removing the old brick chimney top, they discovered that the bricks transitioned to stone unevenly below the roofline up to the ridge. (See photo to the below.)

If it had been originally executed in brick, the masons would have simply started from a level stone surface. This uneven transition was a clear indication that the original top of the 1805 chimney was built in stone and that sometime later (probably during Henry Dinges’ 1869 renovations to the log side of the house) that stone top was replaced in brick. Please check out our posts on our social media pages for more photos of the work that has been completed.

The work on the Stone House roof restoration has begun. In preparation for the installation of the side-lap shingles, we are addressing the structural problems in the rafters, collar beams, and their support framing. The roof structure over the log side is in relatively good condition and has required comparatively minimal reinforcement. It is in the structure over the stone side of the house that we found the most serious problems. That timber-framed system is the oldest part of the house that has sustained the greatest damage from the elements and from man-made alterations over last 250+ years. This article will focus on the important work we are doing to save this original timber-framed roof structure over the stone side.
Before work could begin on the roof, we needed to address the structural integrity of the east wall of the stone side of the house. This east wall leans out too much at its top and away from the structure’s center of gravity. To temporarily solve this problem we have constructed a system of wooden buttresses along the outside of that east wall to support it and prevent any more movement. (See photo below.) Eventually we will be reconstructing a lean-to porch along two-thirds of that east stone wall. It will serve as a permanent buttress and prevent future structural problems.

With the east stone wall stabilized we then turned our attention to the roof structure over the stone side of the house. To make it possible to document the structural problems and to have access to the rafters, collar beams, and their supporting frames, we had to remove the remaining plaster and lath in the second floor over the stone side of the house. That plaster most likely had been installed during Henry Jackson’s period of ownership. This was a difficult but necessary step to take in our efforts to save the house. We already could see there was serious deterioration in sections of the east wall rafter tails and soffit framing. (See left photo below.)

With the plaster off we also found considerable decay and structural cracks in some of the joints of the rafters and collar beams. (See right photo above.) Two of the collar beams simply fell to the floor during the work on the structure. We also found that the rafters at the roof ridge were out of alignment in some places. This required the sistering of all the rafters and collar beams with new framing boards as well as cross bracing. (See the photo below.) The roof structure over the Stone House is now stabilized and ready for the future. We would like to thank the crew of Vintage Renovation & Construction, Inc. for their hard work and the careful execution of their talents as craftsmen.

We were hoping that by this time we would be able to report on the installation of our side-lap shingle roof. Unfortunately, the contractor who is making the 4,900 shingles by hand has run into problems with the wood supplied to him for the job. With shingle making the quality of the wood is critical. When warping and splitting of shingles happens, it leaves you with nothing but what our contractor calls “expensive firewood.” Our contractor is still working on getting the job done as soon as possible while we prepare the roof structure that will support them.

In the meantime there was one more discovery that we made in our forensic examination of the stone side of the structure that is worthy of our attention here. As our historic structure consultant, Mr. Doug Reed, removed plaster, wood trim, and framing from around the rear doorway in the east wall of the stone side of the house, he discovered clear signs that the opening in that wall had been altered from a window to a doorway. The evidence included the use of bricks to fill in gaps left where stones were removed as the opening was extended down to the floor level. (See photo below.)

 

Sandwiched in between the courses of infill bricks are wooden blocks that served as nailers for an earlier door frame to be mounted in this opening. These bricks and the wood nailer blocks have whitewash on them, as do the stones on the other side of this doorway opening. The whitewash is typical for the former exterior surfaces of the Stone House. As we have said previously, we have found whitewash on all the historical outer wall surfaces that were later covered or encapsulated. Because these whitewashed surfaces were protected from the elements they were preserved for us to study today.

This discovery of a window becoming a rear door means that the original design of the Stone House was not the way we had previously believed it to be. It did not have opposing front and rear doorways in the main room of the first floor. This was a bit odd. One of the most common features of Virginia’s Georgian vernacular architectural tradition is the use of ground-floor doorways that are situated in the front and rear of the structure. These doorways were typically positioned to line up, and make it possible to pass directly in a straight line through the house from one side to another. It is recognized by architectural historians that this allowed for better air circulation on hot summer days when a cross breeze through the building would be very welcome. The original rear window opposite the front door in the stone side of the Stone House no doubt performed a similar service, except with smaller volumes of air.

So the question this discovery raises is where was the back door in the original stone structure? The problem with this question is that its premise assumes that there had to be a back door in the original floor plan of the Stone House. Architectural historians who study late-medieval, and early-modern vernacular houses in Great Britain have found that small cottages of one or two rooms like the Stone House did not always have more than one entrance. In fact, the late Ronald W. Brunskill (1929-2015), one of the leading British academics who studied the historic vernacular architecture of his homeland, noted several examples of single-entry historic house types in his published works. Could it be possible that George Cabbage, the original builder of the Stone House, was following this custom that he knew from his homeland when he conceived of its floor plan? If so, it would not be the first time that a British immigrant to the American colonies built a house like they had previously known in their homeland.

Our historic structure consultant, Mr. Reed, has also discovered a blocked doorway behind the wall on the east side of the ground-floor fireplaces and chimney stack. This disused doorway once allowed passage between the stone side of the house and its log addition. Mr. Reed believes that also it originally could have been a doorway leading to the exterior of the stone side of the house before the log addition was constructed. While this theory remains controversial, it could answer the question about where a back door of the Stone House was originally located.

Ultimately, these discoveries do not have any bearing on our restoration of the Stone House to its 1830 appearance. The rear window on the stone side of the house had already been turned into a door by then, and the doorway to the east of the ground-floor fireplaces was also already blocked off by that time. Even so, these discoveries are interesting and worthy of our attention.

In our last post we began to explain how the nails discovered in the dado panels of the first floor south room of the stone side of the house are helping us know when those panels were installed. We also discussed how dating the installation of those panels relates to the construction of the rear bulkhead entrance to the cellar. In this second post on this subject, we will address the other nail evidence that we have discovered and analyze what it might mean. To do this we will focus on the nails that were made with machine-cut shanks and mechanically stamped heads during the long period from the 1810s to circa 1900. (See figure 1 below.)

It can be very difficult to determine the manufacturing date of nails like this because they all can superficially appear to be very much alike. Despite their similarities in appearance, there are certain characteristics that distinguish the early cut nails with machine-stamped heads from the ones made later in the period. To begin to recognize the earliest cut nails with machine-stamped heads from later ones, readers need to understand the difference between wrought iron and steel. Today almost all nails are made with steel. Steel is primarily an alloy of the elements iron and carbon. Prior to the invention of the Bessemer process in the 1850s, the manufacture of steel was relatively time consuming, labor intensive, and more expensive. Thus, steel was used sparingly where it was needed. Nails were not made of steel until the late 1870s.

Wrought iron is an alloy of elemental iron with very little carbon in it. Its manufacturing process was less complicated than the way steel was made in that period. It involved heating and hammering or hot-rolling smelted iron. The principal difference between wrought iron and steel for our purpose here is that wrought iron has siliceous slag inclusions that are layered in such a way as to form what resembles grain in wood. Steel does not. This grain-like characteristic in wrought iron runs lengthwise through a bar, and these siliceous slag inclusions tend to cause weaknesses in the iron’s molecular bond where they are located. Bars intended for nail manufacture were sent to rolling and slitting mills to be made into plates before they were cut into nails. At first the plates from which cut nails were made were rolled and cut so that the grain ran perpendicular to the nail’s shank. This caused the earliest cut nails to sometimes break in half (instead of bending) where the siliceous slag inclusions caused inherent weaknesses in the nail. By the early 1820s the nail-manufacturing technology had improved. The plates that the best cut nails were made from were rolled so that the iron grain ran parallel to the nail’s shank. This made it possible to bend or clinch a cut nail of this type. By the early 1840s nail manufacturers stopped making cut nails with the iron grain running perpendicular to the nail’s shank, as they were too easily broken.

Along with the nail we discussed in our last issue, we found these two types of cut nails in the dado panels. The ones that were used to nail the dado panel to the wall were the type with iron grain running perpendicular to the shank. As expected, the tips of these nails were all broken off due to the weaknesses in their molecular bonds around the siliceous slag inclusions. These nails could not bend or be clinched without breaking. (See image below.)

The other cut nails used in the dado panels are the type with their iron grain running parallel to their shank. In other words, these nails could bend without breaking. They were used to fasten the panel’s baseboard to the individual repurposed cabinet doors that were lined up side by side to make up the dado panel. They were driven through the baseboard and then clinched so their tips were bent over and driven into the wood. (See photo below.)

The presence of these three types of cut nails in this dado panel implies that the carpenter who hammered them in was working at a time when all three types of nails were still commonly in use. That time was around 1830, at the end of the period when the earliest cut nails were being replaced by the new cut nails that could be clinched like the nail pictured above.

In our earlier posts we addressed the dado panels in the first floor south room of the stone side of the house. The significance of these dado panels became apparent when we discovered that the one on the east wall had been in place before the exterior stone wall behind it was reworked to create the rear bulkhead opening to the cellar underneath. In other words, this meant that we would know when that rear bulkhead entrance had been created in that stone wall if we could also come to a more certain conclusion about the age of the dado panels inside. In this article we will address the dado panels in greater depth and explain the significance of the evidence that has come to light since our last issue of this newsletter.

To begin, what is a dado panel? Dictionary definitions of the term dado basically say that it is from the Latin word datum, and it can refer to the architectural section of a pedestal between its base and crown. It also refers to the lower part of an interior room wall that has been decorated differently than the upper part of that same wall. In this case, these lower wall decorations consist of panels. Sometimes called wainscot paneling, they are situated under the dado rail (or chair rail) and above the baseboards. The panels we are dealing with at the Stone House were made by cutting skillfully crafted cabinet doors down to the length needed for the height of the panels and then nailing them together between the dado rail and the baseboard. Some of the graffiti we discovered on the reverse side of one of those dado panels most likely dates to the time when they were still serving as cabinet doors. Because the panels in question were repurposed materials from an earlier period, we need to rely on a more certain method to determine the date when they were constructed and installed. One of the most reliable methods of dating architectural work is the study of the nails that hold together the materials in question. As noted above, this paneling is constructed with nails.

Nail manufacturing technologies have developed quite rapidly over the time that has elapsed since the first part of the Stone House was built in the 1760s. Beginning in the Iron Age and on through until about 1800, nails were made by heating the end of a long, thin iron rod and hammering that section into the shape of a nail. The nail was then broken off the rod and finished with a tool that allowed the blacksmith to hammer the head of the nail into shape. These all-hand-made nails are sometimes called rose head nails because their planished heads resemble petals on a rose. (See Figure 1.) By the 1790s new technology was being developed that made it possible to mechanically cut and shear off a section of a thin bar of iron to form the shank of the nail. This machine operation reduced the amount a labor required to make the nail, but the head was still hammered out by hand. (See Figure 2.) By the beginning of the second decade of the 1800s, new technology was being developed that made it possible to employ two machine operations to manufacture an iron nail completely without skilled hand labor. This type of nail had a cut shank and a machine-stamped head, and it was used into the first decade of the 1900s. (See Figure 3.) The round wire nail that is still in use today was introduced in the 1890s. These nails are entirely manufactured by machine operations and are made of mild steel instead of iron.

 

Figure 1 – Hand-forged nails like this were made until circa 1800.

 

Figure 2 – Nails with machine-cut shanks and hand-forged heads like this were made from the 1790s until the 1820s.

 

Figure 3 – Nails made with machine-cut shanks and mechanically stamped heads like this were made from the 1810s to circa 1900.(Illustrations above are by the author after the work of Thomas D. Vasser, University of Vermont.)

 

We can see that the nails used to construct the dado panels are cut nails. The first nail we pulled is like Figure 2 above and formerly held the dado panel to the east wall of the stone side of the house. (See photo below.) As already noted, nails like this were made from the 1790s until they were replaced by cut nails with machine-stamped heads in the 1820s. If the other nails we find in the dado panels date to this same period, we will have the answers to a number of our questions about the history of the Stone House. In our next post we will have more to report.