75 Attachment(s)
18th Century Dresser - or - The Jobs you Don't Want in the Early Joiner's Shoppe
This will be a long post.
Also: I apologize on behalf of myself and my countrymen, but I've completely reverted to imperial measurement in my woodworking. It just became too much to juggle the two systems so, while I will take to my grave that metric is better on a basic, fundamental level, I have to speak in imperial here.
Our apartment is pretty skimp on cabinet and closet space. We've got it sorted out to a livable point, but upgrading our cabinet-based storage was always on the cards. It wasn't so critical that we needed to go to the Labyrinth of Evil (IKEA), but it was high on the "honey do" list.
I wanted to make something that could be called "Early American", meaning using techniques and embellishments of the colonial American (18th Century) period. So I did a bit of research and learned that the majority of that work was done in Mahogany or Walnut, but I did discover that a fair amount was done in Eastern Maple (Acer saccharum), which is commonly called Rock Maple in AU and Sugar Maple in its native regions of the US. The most high end Maple furniture would've used what we today call Quilted or Tiger figured Maple.
I had some very nice, quarter-to-rift sawn QLD Maple (Flindersia brayelayana) which came from the Atherton Plateau region of QLD in the late 1980s, and which definitely fit the bill of having quilting/figuring, so I decided that was a winner, and would be a great pick for this build.
Attachment 424931Attachment 424932
The cabinet would be fairly large, at 42" high and around 32" wide (not including the overhang from the mouldings), but by no means a huge piece of furniture like a bookshelf or a twelve drawer, wide dresser.
The obvious first step would be to join the main carcase, but I wanted a bit of a morale boost before I jumped into dovetail la la land, so I decided I would start with the "fun" stuff, and cut the crown moulding. This brings me to the first job that you don't want to have in the Colonial American wood shop...
The Moulding Guy
In my mind, the moulding guy is some kid who is likely named Tobias or Martin or Jacob or some other Quaker version of what would be the 18th century equivalent of today's "Jared". Jared is around sixteen years old, and has just finished his apprenticeship. He spends his days walking the same eight foot lap with a 1" rebate plane, stopping every half hour or so to sharpen. He goes home with his clothes soaked with sweat and his hands stained from the beech tannins which leached out of the plane all day, along with the metal powder from sharpening the iron. The best part of Jared's day is when he occasionally gets to finally use a hollow or round to define the moulding profile he's working on, only to then return to his endless rebate plane shaving of sadness and despair. Jared dreams of marrying the shoppe owner's daughter, but we all know she's more interested in the dovetail guy...
In other words, cutting a big, solid, complex crown moulding by hand is some significant yakka...
I milled my 8ft stock to the square dimensions of the moulding profile (1.5" spring and 2.5" height), and sawed out the majority of the waste with two long cuts on the table saw (which Jared didn't have...). This offcut was reserved for the skirt moulding. This was a very serendipitous and appreciated savings of material.
Attachment 424933
The moulding profile was designed using Queen Anne profiles and basically selecting features I liked. It has a cyma recta which descends into two fillets, a large cove, a substantial bead, and then a small fillet and cove at the base. After the initial material removal, I set to work with the marking gauge, rebate plane, and moulding planes.
Attachment 424935Attachment 424936
This moulding was actually longer than my bench would clamp (Imagine that, Ian... It only took me a couple of years to max it out), so I made an extended length sticking board, which turned out to be invaluable for holding the moulding and also preventing it from bowing. These are the tools used for cutting the moulding:
Attachment 424937
The hollows and rounds were used to define the beads and coves, the shoulder plane was used for finish cuts on flat surfaces, the snipe bills and side rounds were used at or around the bead, and the scratch stock was used for definining the smallest cove and fillet at the bottom of the moulding. That was a day's work, and a sweaty one at that.
So the "fun" part was over, and now it was time to join the case. The top and sides would be QLD Maple, and the bottom, as well as all secondary material on the build, would be American White Ash (Fraxinus americana). It's not quite the secondary wood that Tassie Oak is, but it's pale in color, stable, nondescript, cheap-ish, and readily available to me in quartersawn stock, so it's my go-to. There is little, if any, period precedent for its use, but Chestnut was fairly common, and it looks similar, so...
Attachment 424938
The top is joined with half blind dovetails. This joinery will be covered by the crown moulding later. Some period examples show the tails visible on the top of the work. I see the draw, but it's not for me, so these would be hidden.
While most casework being done today is made with 3/4" (19mm) thick stock, period work is more commonly done in 7/8" (22mm) stock. I had the option of sawing off a veneer from each board and using 3/4" stock, but that's not really my thing, so this build was done with case sides and drawer fronts in 7/8" thickness.
This was a day's work from sharpening the saw to gluing up.
The next part was my least favorite part of the build: creating the drawer dividers and internal web structure.
This was actually started during the case-building process, as a couple of the sliding dovetail sockets required cutting before assembly, for reasons obvious in this photo:
Attachment 424939
After that, it was a matter of sawing them and fitting the drawer dividers. These were all around 2" wide and 7/8" thick. I learned a lot about cutting sliding dovetails, which had been a prior nemesis of mine, and I think I can handle them moving forward.
Attachment 424940Attachment 424942
They weren't all perfect, but they were all adequate, and I'm happy for them to be the only "show" joinery visible on this case when the drawers are closed.
I failed to document the building of the internal web structure, so a description will have to suffice.
A large square was used to align stopped dados with the drawer dividers along the back of the case, and 2" by 7/8" thick strips of Ash were installed. Then, strips of Ash were nailed into the sides of the dresser using cut nails made by the Tremont Nail Company (worth checking out).
The back would be what Charles Hayward calls a "traditional dresser back", meaning it was just boards nailed onto the web structure. I used 5/16" Ash for this job which was pre-drilled to allow for cut nails and was bevelled along the edges to create a "shiplap" join. This would prevent light being visible through the back, but would allow for movement and also provide some (fairly unnecessary) rigidity to the case.
Attachment 424943Attachment 424944Attachment 424945Attachment 424946
These are cut nails by Rivierre, which are available from Lie Nielsen (among others). I have nothing but good things to say about their quality and their holding power. I had to pull one because I made a mistake and it damn near required me to destroy the board to get it out.
Sidenote: Could I have used ply? Yes. I chose solid hardwood because I think it is stronger, better looking, and more period-specific. Given that this was resawn stock, it wasn't that expensive and, by the time I bought a sheet of 1/4" or 3/8" furniture grade ply it wasn't all that much more expensive.
At this stage, the main carcase was complete, and it received two liberal mop coats of blonde shellac on all internal parts
Attachment 424948
You'll notice that odd-man-out strip of ash along the top. That is done to provide further gluing surface and support to the crown moulding. It will not be visible.
Moving forward, the next step was to cut the skirt moulding. This was done with the offcut from the crown moulding. It took considerably less time:
Attachment 424949
A fairly simple steep cove, fillet, ovolo. The goal was to make it wide enough so that it, plus the bracket feet (see below) would stick out to almost the width of the crown. I drew this and carefully measured everything to make sure I didn't have an "oh no" moment at the very end.
So it was time to join the mouldings. For all the complex nonsense this build required, mitring these mouldings which I had spent so much time on was the most intimidating. I don't really trust the mitring attachments of any machine to be accurate... ever... so I was fully expecting (rightfully so) to have to adjust the mitres.
I wanted continuous grain, so I cut the front pieces first. These were simply mitred on each end and glued on.
Attachment 424950
The sides offer a unique set of problems with regards to wood movement, particularly in this setting, where I want the moulding to be flush with the top. You can't just glue it on, and you can't just screw it on. If you glue the whole thing, it will eventually either come off or tear the case asunder under tension, and if you just screw/nail it, then the mitre at the front will open up. SO, although there is no perfect fix, the best thing I found after a forums post, considerable digging, and even emailing some prominent period furniture makers, was as follows...
You glue the mitre and the front 20% or so of the moulding, up to a maximum of around three or four inches back from the mitre. Then, you cut slots on the inside of the case, and put screws through the slots into the moulding. This allows the sides of the case to move and the screw just moves with it. It's a fairly easy solution to execute, it's just one of those "knowing is half the battle" scenarios.
Attachment 424871 <<That's the inside of the case
The skirt moulding offered a different set of challenges. You still mitre the front piece and glue it on, but this moulding is small. It's only 1" tall, and the dovetail pins themselves are 7/8" tall. With that in mind, it's obvious that you don't have access to the inside of the case to screw this moulding on. This offers a simple but unfortunately solution, which is to glue the first 3" and the mitre, and to then nail the moulding on. I'd prefer not to put nails in the visible surfaces, particularly the embellished surfaces, of my work, but this was the right move. It was done with 7/8" brads from Tremont Nail Co.
Attachment 424872Attachment 424873
The holes were later filled with a shop-made putty of sanding dust, "titanium white" pigment, and white wood glue prior to polishing.
So with the mouldings attached, it was looking like something:
Attachment 424874
This brought me to a part of the build I had been dreading... The Ogee Bracket Feet.
I'm sure there are examples where it is done otherwise, but it is my understanding that a traditional, period ogee bracket foot would be joined with blind (aka secret mitred) dovetails. I'm not sure how this was done at the time, so all I can say is how I, personally, chose to do it. Note: Some photos are in pine. I did several practice runs on this before cutting it in my expensive and unobtainable (in the US) QLD Maple.
I start by cutting the moulding profile of the foot on a long length which can be used for several feet, and will allow for continuous grain matching. This is done with hollows and rounds after removing the waste on the table saw. (Note: The following description leaves out some critical minutiae and assumes knowledge of cutting blind or at least half blind dovetails by hand).
Attachment 424875
Next, what will be the sides of the feet are cut out of the board at right angles, ensuring to preserve grain continuity in the front two feet. Then, a rebate is cut into the end grain of either board behind the moulding.
Attachment 424876
The dovetails are laid out in the rebate and sawn, ensuring not to cut through the front of the moulded surface. After the saw work is done, it's chisel time. IMPORTANT: You MUST support the cove in the middle of the moulding while chiseling. I found out the hard way that this will split the work if not done properly. As such it is absolutely imperative you make a "negative" of your moulding profile that you can lay the work on while chopping:
Attachment 424879
After the joint is cut, you must carefully plane the mitre to fit. This should be done slowly and with the utmost care so as not to overshoot the join. A single plane stroke more is enough to open the mitre. I think it is wise to plane your mitres a half a degree or so shy of 45deg to ensure the tips of the mitre touch before the back. You'll end up with two pieces like this:
Attachment 424877
Which should join similar to, but far more cleanly than this (this was a practice attempt):
Attachment 424878
After the joints are cut and the mitres planed, use a template to bandsaw out your profile, glue up and you'll have something like:
Attachment 424880
The two rear legs do not require a blind dovetail, and are simply a half blind dovetail joined to the foot at the back.
The feet are screwed onto the cabinet using blocks glued to their back sides:
Attachment 424881 Attachment 424882<<These phillips screws were later replaced with slotted brass. This was done at the time to cut the threads, and because the brass hadn't arrived.
And after flipping it over...
Attachment 424883Attachment 424884Attachment 424885
Getting there...
At this stage it's obvious what's next... Drawers.
My drawer voids were roughtly 6", 7, 8 and 9" increasing in height from top to bottom. I decided to use five tails per drawer. I liked this because it allowed me to keep one of my dividers at the same setting for the pin sizes, and only increase the size of the tails. It would also be pleasing to the eye when looked at from above, because the perspective would create the illusion that all of the tails were around the same size. I also decided I would use the same dovetail profile on the fronts as well as the backs.
I won't bore you with drawer building tedium, as there is enough video and photographic evidence of guys cutting dovetails on the internet for a dozen lifetimes, but I would like to talk about joining these fronts...
These drawers would have lipped drawer fronts, meaning that the drawer would stop against the interior of the drawer front instead of sitting flush with the front of the case. The drawer fronts would sit proud, and would have a moulding cut around all four sides. This requires one of two approaches. You can either cut normal, flush drawers and then glue and screw on the fronts (the easy way), or you can cut rebates around the inside of the drawer fronts and cut the dovetails in the solid wood (the hard, and more traditional, high end way).
Of course, I did it the hard way. I wasn't going to all of those rebates by hand though, so that was a router job. I left enough material so that I could sneak up on a good fit with the shoulder plane. I was able to get all of the drawer fronts tight enough that they would stay put in the case without their actual drawers attached, but would also easily pull out.
Attachment 424886Attachment 424887Attachment 424888
After this, the dovetails were sawn by hand in the rebate. This is difficult and tedious, and requires considerable chisel work to fully clean the sockets out. Even marking them is a bit funny. I actually had to stop the presses and make a special awl to get in there and do it:
Attachment 424890Attachment 424891
And then it's fairly intuitive:
Attachment 424892Attachment 424893Attachment 424894Attachment 424895Attachment 424896
And then, with some luck...
Attachment 424897
You can't use a bench plane to flush the tails with the pins, so it has to be done with a shoulder plane. I found that after this, it helps to make circular strokes with a block plane across the grain to smooth it out and then scrape.
I was really happy with all of these joints on the drawers. I usually have a tail here or there, or maybe an entire joint, that is embarassing, but on this one they were all up to what I consider to be my "working standard".
The next step after the drawers was to cut the mouldings on the drawer fronts. Also done with a router using two bits:
Attachment 424898
I used a traditional roundover in a fillet. Typical period examples of this use something which is typically considerably more narrow, but I kind of liked the fillet a bit wider. This is the result:
Attachment 424899
And after a bit of drawer fitting and refitting, some sanding and fettling, and two mop coats of blonde shellac on all internal parts, the woodwork was done!
Attachment 424900
At this stage all that was left to do was French Polish it. There was nothing that could go wrong.
And with those famous last words, I give you the second Job Never to Have in Colonial American Joiner's Shoppe:
The Finishing Guy
The finishing guy is probably a first generation immigrant, possibly of Eastern European descent, with an abnormally and asymmetrically large right arm and who speaks little if any English. He seems nice enough, he works hard, but doesn't say much, and typically spends his twelve hour day vigorously polishing casework and tables. He smells like spirits, but not because he drinks. His fingertips are constantly a sickly shade of orange, and the only time anyone really notices him is during the occasional outburst of what can only be assumed to be a string of profanities in his native tongue, or when he eats whatever that awful-smelling fish is at lunch.
In other words... I had a bit of a debacle while applying shellac, and this piece continued to offer "learning opportunities" until the bitter end.
I have historically only used orange shellac. It has been appropriate for the projects, but I wanted this one to have a deep, dark, bold color, so I went with the more traditional garnet. Charles Hayward claims that you should always endeavor to use garnet whenever the project will allow. I can't speak with his level of experience, but I know that it was the right call on this project.
I have had lots of success French Polishing in the past, and up until literally the day I planned to bring it over to the house, everything was going well on this one. I used three mop coats, made several rubbers of varying sizes, and was able to tackle the mouldings and the feet with considerable ease and straight-forwardness. The sides were laborious, as it's harder on the arm, but I got them done. Approximately 20 "rubs" (I refuse to use the term "coats") per surface.
I saved the top for last, and gave it around 30 rubs. I wanted it to be real nice. And right as I was getting to the "spiriting off" phase, I noticed a blemish. My best guess is that I dripped something on it and then rubbed it too much, because it was down to almost bare wood. After maybe half an hour of fiddling with it, I'd made it considerably worse:
Attachment 424901
So I did what is often the best thing to do... Walk away.
After a second day, I had only made it worse. It wasn't until day four that I finally decided it was time to throw in the towel and sand back the entire top. This was on Thanksgiving day. This year I'm thankful for orbital sanders...
The second attempt went off without a hitch. The last things to do were put the hardware on, reattach the feet, and put clothes into it. I got the hardware from Whitechapel Ltd out of the UK. It was extremely expensive, at US$55 per pull, but I had done everything else to an extremely high standard thus far and wasn't going to skimp at the end. I chose these "Chippendale style" pulls because I thought they were appropriate, but were also understated compared to some of the more regal looking brassware of the time.
Attachment 424903Attachment 424904Attachment 424905
And then after screwing the feet back on...
Attachment 424906Attachment 424907Attachment 424908Attachment 424909Attachment 424910Attachment 424911Attachment 424912Attachment 424913Attachment 424914Attachment 424915
I even got one of these:
Attachment 424916
You'll all be proud to know those are Blundstones :cool:
All in all, this dragged out over a four month period. With that said, I was gone for 2.5 of those months, so it was, realistically, only a six week project of evenings, weekends, and the occasional day off. I, honestly, feel great about that. I share a shop space and I get a lot of commentary on how I'm able to work so fast even though I "do everything in the hardest possible way" and "spend so much time sharpening tools". I guess that's a compliment? I'm going to take it that way...
I'd say this is the best thing I've ever built. It's my favorite. Easily. I learned a lot from it without wasting a bunch of time or material, and it's an incredibly useful piece of furniture. And not bad to look at taboot...
I always like to comment on the wood I used, and wow... What can I say about QLD Maple? It planes like a straight grained wood even when figured, and even against the grain. It chops cleanly, sands quickly, and doesn't splinter. It's not too hard to be laborious to work, and not too soft to dent it in use. It polishes with ease to a high shine, and has tremendous depth of grain and consistent character and figuring. This is a truly world class cabinet wood. The discerning craftsman should never turn down an opportunity to try it out.
I've got work travel and then home for the holidays on the horizon, so this is the last piece of furniture I'll make in 2017. Nonetheless I can't complain too much about my productivity...
I made a king sized bed
Attachment 424917Attachment 424920
A sweet cabinet for my saws and moulding planes
Attachment 424921Attachment 424922
A coffee table for my neighbors
Attachment 424923Attachment 424924
And one for myself
Attachment 424925
Plus a handful of smaller "day or two" type jobs
Attachment 424926Attachment 424927Attachment 424929Attachment 424930
And a few other odds and ends that didn't make the "photographic evidence" cut. And that was with almost six months away from home. Not bad at all...
So yeah... it was a good one. I look forward to many more.
Cheers,
Luke
