And lo, on the third day of September, 2010, it was finished. A new roof, all walls reshingled, all windows repaired. A couple of weeks of painting and puttering followed, and finally, the scaffolds came down on September 20 to dramatically reveal Comstock House about as it appeared when the Oates' family moved in during Easter week, 1905. The old gal done cleaned up good.

There's still exterior work to be done - the porch must be shingled, about half of the trim still awaits attention - but mostly, the exterior is now restored. Better yet, there were no injuries, and the project came in under budget and early, despite a couple of serious problems with materials.

At right is a short video that reviews the work of the last two summers, along with a glimpse of how the house appeared when we first saw it in 2006.

Visitors often are disappointed that we don't have a ghost story about Comstock House, so here you go:

The old dark house was wrapped in a blanket of fog that August evening, the sky moonless. Although it was near the peak of summer there was a seeping chill. In the distance, a beast howled.

TOP: Start of the roof tearoff. Notice the tarp-covered frames to protect the plants
MIDDLE: A rain of shingles
BOTTOM: A tarp was hung over the scaffold to control dust and help conceal that 11 windows were removed
CLICK on any image to enlarge

Then suddenly from the attic, footsteps. First a heavy step, then in stumbling fashion, another clumsy tread, then another. Then another. Moving closer.

"Someone is upstairs!" I exclaimed, bolting from my chair. Up the steps into the dark I flew, a flashlight in one hand, a claw hammer in the other. My heart pounded as I ripped open the attic door, expecting the worst.

My flashlight's beam probed every corner as I sought that lurker in the dark, yet no skulking invader awaited. I was the only person in the attic. Living person. Clearly: Proof of a ghostly encounter. The footsteps were of the murderous tycoon James Oates, or any of the better souls who died in this very place.

Or perhaps... were the lumbering footsteps actually lumber? Could the sounds merely be a cascade of creaks caused by temperature and stress fluctuations in the century-old rafters, floor joists, and beams, after about two tons of old shingles were removed from the roof? Was the howling beast possibly our neighbor's lovable beagle? Was the homeowner in the attic with a hammer easily spooked? Could a blog update offer a more belabored intro? Who can say for sure?

The tearoff of the eastern roof was certainly dramatic; almost all of the living space in Comstock House is on that side, and soon after the demolition began, a new draft pulled through the entire house, even when the attic door was closed. Soon after came a faint smell of something very old (think of Dickens' dismal Miss Haversham) and by evening, a thin layer of talcum-like dust was everywhere. I sneezed for two days, when I wasn't charging into the attic with a hammer.

The mostly chill days of August were the busiest yet for the Comstock House restoration project. Most mornings found Filipe Garcia, sometimes supported by another worker from Ridgeline Roofing, methodically laying new shingles over the old sheathing boards, which were already dotted with nails from (at least) two previous roofs. Jim Scotchler continued shingling the white cedar sidewall. A dozen other contractors came and went, the driveway always blocked by trucks.

The groaning house and blanket of microdust aside, the most noteworthy event of the month was the discovery that the house originally had a bizarre gutter system, as discussed at length in a separate post on the Restora Obscura blog. Superior Seamless Gutter of Santa Rosa also came by and fabricated a 58' long gutter for the front of the house.

The greatest challenges were restoring the windows and frames, which suffer far more weather on the east than any other face of the house. But the experienced gained from a year of breath-holding work sessions during the 2007 restoration of an attic window now paid a reward. This summer there were eleven distressed windows removed at one point, all moving between the basement workshop and outdoor tables for various stages of repair.

LEFT: Filipe Garcia, who shingled the entire roof
MIDDLE: Jim Scotchler, finishing a section on the western face
RIGHT: Fabricating the long copper gutter for the front

All of the attic windows and most of the ones on the second floor had at least one lower joint that had separated. The top window at left below only needed glue and a clamp to bring the mortise and tendon back together; the window beneath was in far worse shape, with a large section of the stile rotted away. To restore it, I made a form on the window side and the bottom using popsicle sticks held in place with push-pins. (If I ever write a book on the Comstock House restoration, a chapter will be called, "Toothpicks, Popsicle Sticks, and Tiny Dental Tools.") The photo in the center shows the area filled with epoxy as viewed from the bottom of the window.

The final image below shows the same window in the process of glazing, using an unorthodox technique of my own. First I crudely pack the window with fresh, warmed putty - DAP 33 only, please. It should have the consistency of room-temperature butter; the only goal is to ensure the putty has good adhesion to the glass and wood. After it has dried for two or three days the putty will be firmed up similar to cold cream cheese, and you can make the 45° cut with a utility knife. Half (or more) of the putty will be sliced off and thrown away. Dip the end of a flat edge glaziers tool in linseed oil and drag it over the putty to achieve a smooth line. The linseed oil also helps the putty tighten when it dries. Allow it to sit in the sun for at least a week before painting. This technique is time-consuming, but overall less labor intensive than trying to finish glazing in one session.

Five weeks into it, the project already feels like ballet dancing while rounds of live ammo fly overhead.

Our schedule called for almost everything to be finished in 100 days, a time table that was tight but allowed for ten days of leeway. But the wet spring pushed the starting date back to the last week of June, nearly a month later than anticipated. Worse, that cold, rainy weather also delayed our contractors' work for their other customers, making everyone now very busy bees. The result is that any glitch can easily cascade into a freeway pile-up. Everything almost came to a halt when we discovered (A) there wasn't enough of a custom piece of trim, which (B) the painter had to paint before (C) the carpenter attached it to the house which had to be done before (D) the roofers could roof. If the good folks at Redwood Lumber Company hadn't been able to hustle through additional moulding on short notice, we easily could have lost a project week as the delay rippled through everyone's schedules. A couple of problems like that and the roofing project might be pushed dangerously close to the usual start of the autumn rainy season.

(RIGHT: Full house on the scaffolds, with five contractors)

Although it seems that there's always more repair work needed than expected, the job is mostly on time, here at the halfway mark. Filipe Garcia of Ridgeline Roofing finished the western side of the roof on August 6, just short of six weeks after the new rain gutters were installed (to be discussed in the next update). Jeff Patton of Patton Painting has continued the faux painting he did for the trim on the upper floors of the north face. Jim Scotchler has shingled all of the west gable, as well as finishing the last remaining sidewall. Except for the porches and upper floors of the east face, all of the Eastern White Cedar sidewall shingling is complete.

The main challenge of this phase was solving the mystery of the bargeboard, which is a subject of its own essay over at the "Restora Obscura" blog. Not far behind was grappling with the windows. We found the cathedral windows were in pretty good shape (thank heavens), but all other windows on the upper west face were verging of collapse. All had at least one mortise and tendon joint open (detail of the little shed window seen at right), the redwood frames in advanced decay from years of exposure to the elements, and the glass barely held in place by the remains of crumbling putty. Not to worry, though; we had already wrestled with a badly failing window as our first restoration project back in 2007, so we knew the techniques and materials needed to fix them like (almost) new.

Not so clear-cut were the problems of the leaded glass windows in the master bathroom. Every attempt at repair only served to reveal something worse. The glass in both windows was sagging and bowing. Candice resoldered the many broken joints in the H Came, which firmed up the window structure - but found she was unable to repair the broken reinforcing stained glass rebar in one of the windows. Then while cleaning out the old putty to reglaze, we discovered that the crucial U Came (the "U" shaped strip of lead that both holds together the assembled glass structure and is joined to the wood frame by glazing) was mostly gone, the material both crushed by an earlier repair attempt and oxidized into nothingness where the glazing putty had fallen out. We were left with no choice but a complete rebuild. This was another surprise that could have delayed part of the project, but were fortunate that Bob Tutone, president of SGO Designer Glass & Doors in Santa Rosa was game to personally handle our job. (This company really knows artisan glass; check out their portfolio of doors and ceiling panels and window overlays.) The outcome of his craftsmanship is a pair of windows that are probably in their best shape since FDR was in the White House.

LEFT: Candice resolders a broken joint on the leaded glass window
MIDDLE: Closeup of the smashed U Came, which must attach the window to the frame
RIGHT: Closeup of the window showing the warping

A final project in these busy weeks was small but rewarding. Despite the chaos of demolition and reconstruction, the summer of 2010 has brought family visitors, and the main guest bedroom lacked a single window with a wire screen for ventilation. The screen that had been on one of the windows was disintegrating (once I swatted a fly and watched as a large patch of wire mesh crumbled into dust) and the wood frame had the same peeling paint and rot as other exterior wood. All of that frame - including the half-round trim around the screen - was redwood, and probably custom built for the house in 1905.

Our research suggested that lacquered bronze mesh was the material of choice for window screens in that era, and that's what we used for the rescreening. We were happily surprised when we found a bit of the old screen still in original condition under the half-round trim, and sure enough, it was exactly the same kind of bronze screen as we are using. In another little way, Comstock House is again beginning to look as it did when new.

After a year's planning, the summer of 2010 is bringing a new roof to Comstock House. This posting is first in a biweekly series to show the state of progress.

The photograph below shows the status as of July 15. The old shingles on the west side have been completely removed, and new copper gutters are in place, protected with plywood against dropping accidents. Felipe Garcia of Ridgeline Roofing has completed shingling of the mansard portion of the northwest roof, and the peak of the roof is loaded with bundles of shingles to complete the west side. To the right are the towers of shingles that filled the driveway. The tallest pallete holds the Eastern White Cedar that will be used to complete the sidewall; all other shingles are Western Red Cedar for the roof.

LEFT: Sunlight blazes into the attic, sans shingles
RIGHT: Marshall Scaffolding built two sturdy towers, enabling us to reach the peak of the gable and step onto the top of the roof

LEFT: Master shingler Jim Scotchler completes shingling of the western gable. The half-round window was removed for repair
RIGHT: Contractor David Jessen applies sealant between the trim and nail board

Architect Brainerd Jones' directions were clear: don't paint any of the wood on Comstock House, inside or out. The interior was mostly spared the paintbrush (although one "expert" tried hard to convince the Comstocks that everything would look much nicer with a coat of white paint), but every inch of exterior wood has two, three coats, much of it now peeling away.

Something has to be done to protect the outside redwood from sun and rain, of course, and Jones' instructions were that instead of paint, it was to be treated regularly with a linseed oil/turpentine formula. Oiling your house a couple of times a year may be unusual, but it's not an onerous chore - at least for the first story, where windows and trim can be easily reached. Working on the upper stories, however, requires scaffolding or a dangerously high ladder, so it's difficult to imagine that the oil treatment was continued for very long, and thus some painting was unavoidable. But was there a paint color that could really match the look of oiled wood?

Our first actual restoration project was rebuilding an attic window, and I began experimenting with paints, settling on a reddish-brown color that was popular in the Arts & Crafts period, and just might pass for "redwood" if you squinted real hard. Real. Hard. I continued to dabble for months (we still have gallons worth of brown-ish paint in small cans, each with slightly different mixes of red or orange), and finally did a study of oil and turpentine on actual century-old redwood. It became apparent that it was a mistake to seek a single "redwood color;" the solution was adding very dark grain patterns over a red-brown color base. In other words, faux painting the wood too high to reach.

Although it's doubtful that the exterior of Comstock House was ever faux painted before, Brainerd Jones certainly knew about the technique; faux painting became quite popular in America during the two decades before WWI. It was used to fake the look of stone or fine wood (James Wyatt Oates had a set of four redwood bookcases faux painted to look like golden oak, which can be seen here in the background) and from 1895 onwards it was somewhat of a DIY fad thanks to Glidden, which had an advertising edge over other lacquer companies by selling its "Jap-A-Lac" varnish as part of a kit for homeowners to make splintery-wood floors more attractive as well as smooth and durable by using their graining kit.

(RIGHT: 1909 ad from McClure's Magazine - click on any image here to enlarge)

Probably anyone with a home built before 1930 would benefit from knowing a little about faux painting. With a few simple tools and just a little practice, the results can be quite rewarding. A couple of books about the technique from that period can be downloaded through the Comstock House library, but without question the definitive instructional book is "Professional Painted Finishes." I might even place it among the top 10 books I've found for old house restoration.

To finally get that century-old redwood color perfect along with the faux grain, we consulted with Sonoma County artist Patti Zimmer. Above is a short instructional video of Patti teaching Jeff Patton (Patton Painting/Santa Rosa) how to use the graining tool. Note that this is exactly the same tool shown in the ad from a century earlier.

Patti was able to precisely match the look of our old redwood. At left is our window sill test platform, with the actual oiled redwood on the outside (farthest left) and the faux painted sill closest to the camera. There is a varnish top coat on the paint to preserve it and give it the same semi-gloss appearance of freshly-oiled old redwood. On windows, the rails, stiles and mullions will be stripped of old paint and oiled, while the muntins and window putty will be fauxed.

Oxide Yellow (Y3)
Black (S1)
White (W1)
Oxide Red (R3)
The formula for the base color of "Comstock Red" is shown at right. It is Benjamin Moore exterior Aura paint flat, deep base four tint. Note that these details only apply to Aura exterior, and differ from even the brand's sister interior grade. The graining color is lamp black and raw umber mixed by eye in an acrylic glaze base. The top coat is a 50-50 mix of Acryvar Clear Satin 494 and Acryvar Clear Matte 492, which seemed to match the sheen of oiled redwood.

The photos below show Jeff Patton's remarkable work in faux painting a pair of the windows on the north face. Like the very best faux work, it deserves highest praise because it is never apparent. Only when we point it out and explain that it is painted do visitors notice, and always they marvel at the surprise of it.

The front porch certainly had seen happier days by the time we moved into Comstock House. Part of the floor was rotted and needed urgent replacement, the stone lions had crushed the stair side platforms (original 1905 redwood, too), almost all treads were split, and the entire step assembly was pitching forward away from the house. And, as we found out, that wasn't the bad news.

Our original restoration plan for the stairs was to simply replace the treads and risers, believing the supporting structure was still sound. The treads felt firm, except for a little sponge on the split boards. But once we looked closer, it became apparent that the steps could have collapsed at any moment. The side walls had no foundation at all and were simply sitting on dirt, with one of those walls haphazardly braced by a pair of 1x6 planks nailed to the back (a beer can was also found in the dirt next to this fine workmanship). While the base of the stringers rested unbolted on three sturdy concrete piers, the stringers themselves were rotting because they were unsealed raw pine. Except for a few puny nails connecting the top of the stringers to a floor joist - which itself was crumbling with dry rot - the entire structure was free-standing. Had any of the stringers failed, the whole mess would have lurched and collapsed to one side or the other, like a puff on a house of cards.

LEFT: What happens when 100 lbs. of stone meets 100 year old redwood
RIGHT: No foundation under the steps, except for the front piers

LEFT: A foundation was formed around the existing piers
RIGHT: Pressure-treated stringers are bolted into the new foundation

ABOVE: Framing the new side walls

Aside from a debate about period-proper handrails, construction was uneventful. Architect Mark Parry created a foundation plan where new concrete was poured around the old footings (which are two feet deep). A sill bolted into the new concrete provides support to the pressure-treated stringers, which are placed between the piers. The stairs now feel sturdy and should last for decades.

There's still some work to be done. The cedar steps need another coat of paint, and the caps on the side walls will have to be refinished (again), a project that will come up in a later post.

The back stairs were a lesser headache. Those steps were still firm, and might have lasted another ten years or more; the immediate concerns were instead the railings, which were becoming more wobbly each year on both sets of stairs. We already had made emergency repairs to prevent collapse of the kitchen railing when we moved into the house in 2006, and since then I had used consolidant to fill holes and major cracks.

None of that construction dated back to origins of the house; both entrances were apparently rebuilt at the same time, c1960. Some of the details matched the blueprints, other parts not so much. The railing on both landings retained architect Brainerd Jones' hallmark "Union Jack X," but the house side of it was attached directly to the building, not to a half post, as indicated. The caps on the backyard posts were square and flat, not rounded; the handrail was likewise a flat 2x4, where the blueprints suggest a curve. Yet the kitchen posts had arched tops, and that handrail seemed to follow the drawings with what was called at the time a "toad's back" (AKA toadback, toad-back) curve to the top, making it inviting to hold. Why were they different? Were they rebuilt by different carpenters at different times? Small mysteries.

LEFT: The backyard steps; notice the flat platforms atop the posts and the white patches of consolidant on the railings
RIGHT: Detail of kitchen railing

Although both stringers and decks were still in reasonable shape, it was decided to replace everything at once, using modern materials such as pressure-treated wood for the supporting structures. Moving the stringers adjacent to the house closer to the center also allowed a gap between the treads and the wall to prevent water from pooling on the steps.

Contractor David Jessen performed his usual magic in restoring the design shown in the blueprints. Once more the landing rail terminates into a half-post attached to the house; the tops of the posts are again curved on all four sides. At our request he placed the newel post on the second step instead of on the ground, as shown in the drawing. We believe this is how it was originally built, providing a "swing around" that's also found on the interior staircases.

None of material from those stairs was worth salvage, but we were able to refashion original redwood left over from the rebuild of the front porch balustrade in a few parts. The half-posts on both back porches are old wood, as is the kitchen porch top rail and the backyard porch newel post.

LEFT: The backyard steps before finishing; notice the old redwood newel post and half-post at the top
RIGHT: The kitchen steps after oil treatment

In theory, the north side of Comstock House should have been easier to shingle than the south. Yes, more shingles would be required because there's more square footage to cover; there's no chimney or porch cut-out to break up the wall on this side. But that also would mean, in theory, that the project should go faster because there would be less time-consuming work on chimney flashing and trimming all those end shingles into custom shapes to fit. It looked like a more straightforward job overall. In theory.

(RIGHT: Jim Scotchler unwraps the third of three shingle palettes that have been used to date. Each palette contains 13 squares, and each square is intended to cover 100 square feet. Actual coverage is less due to defects - shingles too narrow or uneven - and because we are strictly using the old standards where there is only 4½ inches of shingle exposed)

Before any work could begin, we had to evict our squatters. Honeybees had lived in a portion of the north wall for as long as forty years, and getting rid of them required tearing open the sheathing. We moved our scaffolding from the south side of the house to the corner with the bees, and made a date with beekeeper and "structural extraction" expert Coral Pawka to round 'em up and move 'em out. The tale of this adventure is told in a previous post (don't miss the video).

The other major obstacle was reaching the peak and upper section of the northwest corner. There's a three-foot bump-out for the kitchen, making it impossible to erect the usual scaffolding up against the house. This corner also has the electrical service, so there were also safety concerns about having a scaffold near the lines. We contacted Marshall Scaffolding in Santa Rosa and arranged for them to set it up, and their solution was ingenious. A traditional scaffold remained on the east side and on the west, sections were built up both from the ground and from the roof over the kitchen. Planks bridged over the twin scaffolds. A photo can be seen at the end of this post; every part of the structure was absolutely rock-solid, even at the very top. My highest recommendation goes to these good folks. But like the removal of the bees, this was essentially a separate construction project from the work at hand.

We have some concerns that the bees may try to return to the location of their old home in the spring, even though all the honeycomb is gone (we hope), their former residence between wall studs is now packed with insulation, and every possible entrance way is sealed (we hope). We likewise expect that the new shingles will deter the woodpecker that did so much damage over the windows in the servant's bedroom and attic.

The real estate agent told us that a servant had died in that bedroom (VERY doubtful) and hinted that some thought it haunted. Then a few weeks after we moved in I heard a knocking coming from there; assuming it was a plumber or someone working on the house, I checked outside. No one around. The same thing happened shortly thereafter, and confess that I was now mildly spooked. Finally I caught our "ghost" in the act; instead of the furious rat-a-tat-tat I associate with Woody Woodpeckers, this fellow makes a few pecks that sound like knocking or hammering, then peeps into the hole to see if there's anything for eats. Before the reshingling I kept a big aluminum skillet in the attic, and whammed it against the wall whenever I heard the "ghost," which seemed to exorcise it for a couple of days.

Older Posts