My internet service was out for a while and I wasn’t able to respond to the comments to Chris’ reading of “A Visitor Comes to Covington” or to the backstory of the book. Thank you for the many very kind comments.
I wasn’t sure how the book would be received. In the letter sent with the book my suggestion was to put it on a high shelf in the library, push it well to the back and put something heavy on it. Alternatively, it could be buried in the basement. Fortunately, the Stick Chair Badge Approval & Distribution Committee (Chris and Megan) liked the book and I heard there was a bit of teary-eyedness when each had read the book. I didn’t intend to make anyone cry but have to confess I got a bit of moisture around my eyes when Chris read the book.
Below is a photo of my last cat, Bunky Beanie Bronzini. He was a big and solid 15-pounder capable of herding me towards the kitchen when he thought I might be headed in the wrong direction. If he had lived another few years he would gained another name or two.
The following is excerpted from Peter Follansbee’s “Joiner’s Work.”
If you like green woodworking, “Joiner’s Work” is doctoral thesis on processing furniture-shaped chunks of lumber from the tree using and axe, froe, hatchet and brake. If you are into carving, Peter dives into deep detail on how he festoons his pieces with carvings that appear complex but are remarkably straightforward. And if you love casework, “Joiner’s Work” is a lesson on the topic that you won’t find in many places. Peter’s approach to the work, which is based on examining original pieces and endless shop experimentation, is a liberating and honest foil to the world of micrometers and precision routing.
The book features six projects, starting with a simple box with a hinged lid. Peter then shows how to add a drawer to the box, then a slanted lid for writing. He then plunges into the world of joined chests and their many variations, including those with a paneled lid and those with drawers below. And he finishes up with a fantastic little bookstand.
Construction of these projects is covered in exquisite detail in both the text and hundreds of step photos. Peter assumes you know almost nothing of 17th-century joinery, and so he walks you through the joints and carving as if it were your first day on the job. Plus he offers ideas for historical finishes.
The primary material for joiners’ work is oak (I use white oak [Quercus alba] and red oak [Quercus rubra] interchangeably) that has been riven, or split, from the log.
This results in boards whose face is the radial plane, the most dimensionally stable surface possible. The stock is initially worked fresh from the log, a state we call “green.” After initial planing, boards are selectively dried some, then re-worked – the decoration and joinery are cut once the surface is dry enough to take a good finish, yet the interior of the stock retains some of that moisture, making it easier to cut than air- or kiln-dried stock.
When studying period pieces, I see that case pieces – boxes, chests, cupboards and related items – often display a mixture of riven and sawn stock. Usually the mill-sawn stock is the secondary wood, and in New England pieces this is often Eastern white pine (Pinus strobus). Other softwoods might appear as well. I’ve noted Atlantic white cedar (Chamaecyparis thyoides) riven into chest bottoms and drawer bottoms. It also appears as applied mouldings.
I’ve seen yellow pine (probably Pinus rigida or Pinus resinosa) in furniture from Connecticut and western Massachusetts. You can make the whole piece from riven oak, and many period works are just that, but when making case pieces, I prefer to use white pine for at least the bottoms, and often the lids, too. This is just a way to save some oak for future joinery work.
While I am quite content using oak for most any piece of furniture, you might want some variety in your work. Black walnut (Juglans nigra) and ash (various Fraxinus species) are perhaps the most commonly found non-oak timbers I have seen in period joinery. I have used them and some other woods, too. I’ve seen Spanish cedar (Cedrela odorata) in some 17th-century joiners’ work from Boston, and other timbers besides. And recently, I’ve been using some Alaska yellow cedar (Cupressus nootkatensis) for carved boxes. Try what you have. I’d say the single-most important feature is straight grain. If nothing else, keep that in mind. Recently a friend gave me several sections of riven black walnut. They were short lengths, but just right for a joined stool. My previous experiences with walnut were a mixed bag. I often joke about being a personal monoculture; I have used oak so much for so long that I had no reference point for most other timbers. But it wasn’t until I used the riven walnut that I could truly appreciate this wood.
I used to think “it’s not for me” – but that was because I came at it the wrong way. I knew that for oak, riven green wood was the ideal. Why I thought it would be different for walnut is just one of those mental blunders that hit us all from time to time. I was judging this wood based on cuts that I would never deal with in oak. My first walnut project was a wainscot chair, made from kiln-dried wood, complete with grain run-out and all manner of un-straight fibers. Next up was a board chest, this time in air-dried wide boards. Nice, but expensive. It worked so much better than the previous project, but the chest had few details, and I really didn’t get much out of it.
Then it all clicked perfectly with the riven green wood. A nice straight-grained walnut log is a dream to work, easy to plane and to chisel – and the axe work in walnut makes you feel like Superman. Turning it is especially satisfying. When it came to carving, I had to lighten up my mallet blows. When I used my usual approach, I blasted the walnut to smithereens. Less mallet work, more hand-work and it cut fine. Live and learn. Now I have to take back all the horrible things I’ve said about walnut. Riven, straight-grained, green, free – that’s the way I like my walnut.
Around southern New England where I live, the ash tree most commonly found is white ash (Fraxinus americana). I have used this wood for furniture almost from my very first projects. When I made ladderback and Windsor chairs, ash was a favorite riving wood for parts. It is lightweight, strong and splits well – an all-around great hardwood. It has a short shelf-life in the log, though; it can go bad pretty quickly if you don’t get to it. Seventeenth-century turned chairs were often ash, but some joined work was, too. You work it just as you would oak. I have found it’s somewhat stringy, and when very fresh it can tear out badly under the plane. But if you work it green then let it sit a while, it will work very well afterward. The color is pretty bland, and there’s no figure on the radial face as in oak. So carve it up then paint it, and no one’s the wiser. Nice wood.
The Way I Work Some woodworkers keep a stockpile of lumber on hand, and draw from their stacks as they begin a new project. Others buy enough lumber (with some extra) for each piece they are planning to build, often working from a list that includes all the pieces in a given project and their rough dimensions. I work in a different way. I start with a log and split out almost every piece of oak that I use at the bench. Starting with the log is a lot of work, but it’s fun. This approach usually results in me starting several projects at once, then I leapfrog back and forth among them. Sometimes I have quite a few pieces underway, usually limited by shop space. As I write this, I have a joined chest, a chest with drawers, a chest of drawers, two wainscot chairs, a long table and several joint stools in progress. I just finished a carved box in the midst of all these.
It’s not an attention-span problem. It’s what happens when I open the log. I have an idea of what I want from a given log, but the tree has ideas of its own. I might be preparing stock for a joined and carved chest, but end up with more narrow framing parts than I need. These get shuffled over to the next project. Panels in a chest range from 7″-12″ wide, usually only about 14″-16″ long. Some logs offer wide panels that are longer than that, but are still too short to get two lengths out of them. These are ideal for carved boxes whose long front and back boards are usually around 20″-24″ long. It goes on and on.
Selecting the Timber If you are going to split your stock, it starts with the log. Finding the right log isn’t half the battle, but it’s a good chunk of it. Ideally, you’ll work with an oak that grew dead-straight, is clear, or free of knots, bumps and branches, and is large enough in diameter to yield wide boards along its radial plane. With all those criteria met, you can get even more demanding and look for an oak log that meets all these demands and is also slow-growing. In some places, you can find this timber at sawmills and log merchants, and sometimes even firewood suppliers will have short sections that fit the profile.
If you’re skipping the log-splitting aspect of this work, you can come close to this material if you look for straight-grained quartersawn boards. When using sawn stock, I prefer air-dried to kiln-dried. To me, the kiln-dried stuff seems stiffer and less cooperative. Air-dried is harder to find, but it’s out there. Flat-sawn boards can be used in joinery; they certainly are prevalent in surviving English work of the 17th century. They require some careful planning and much more effort in working, but you can make joined work with flatsawn stock – it just won’t be as much fun as working with riven stuff. When using riven stuff, you’ll be working the wood green, or unseasoned. If you have flat-sawn stock, it needs to be seasoned. The general rule of thumb is one year of air drying per inch of thickness. When I use sawn oak, I like to air dry it outside, then bring it in the shop for a few months before breaking it down to rough-size pieces.
Oak Most of the oaks are ring-porous hardwoods, meaning their growth rings are comprised of two different types of cells. The material that grows in the spring is generally more porous than that from summer growth. This results in a distinct division between the spring or “earlywood” and the summer or “latewood.” Also visible on the end-grain view are the ends of the medullary rays. These are cells that radiate outward from the log’s center, or pith, toward the bark. Ring-porous hardwoods, with only a couple of exceptions, split predictably both in line with the growth rings and perpendicular to them, along the medullary rays.
The end grain of a large oak shows two distinct colors in the wood. The outer inch or so (just below the bark) is the sapwood. This material is the part of the tree that conducts the sap. It is more prone to decay and infestation, and is considered weaker than the darker-colored “heartwood.” In oaks, the greater proportion of the tree’s diameter is taken up with the heartwood. A young sapling is all sapwood, but over time as the tree grows, there is a transition in which the inner layers of sapwood undergo a chemical change into heartwood. For our purposes, the principal difference between sapwood and heartwood is the greater resistance to decay that heartwood exhibits. In woods such as ash or maple, most of the usable stock is sapwood. In the oaks, the sapwood is generally discarded, and the heartwood is the stuff of choice.
When splitting out oak stock for furniture work, the radial face is the best one to use for a number of reasons. The primary benefit is the dimensional stability of boards oriented this way. In a straight-grained example, there is little shrinkage across this radial face, thus little to no distortion, either. Another feature of this radial plane is the ease of working it – the wood cuts more easily on this face than on the adjacent tangential face. So the carved work is always done on the radial face. Mouldings are often cut in this plane, also. When using riven oak, the stock is usually oriented so the radial faces compose the front of the piece.
The material nearest the center of the log is called juvenile wood. Formed when the tree was a small sapling, this wood is often very fibrous and can include twisted grain. There is no clear distinction between where the juvenile wood leaves off and the workable heartwood picks up. Each log is different. If the innermost fibers appear straight-grained, you can try to keep as much of that stock as possible. One place where you might push your luck with the juvenile wood is when you are trying to split out the widest panel stock you can get. (I’ll cover that in detail when we get to splitting and working panels.)
The following articles on lapped and double-lapped dovetails (aka “half-blind” and “blind” dovetails) are excerpted from Volume 3 of “The Woodworker: The Charles H. Hayward Years.”
The third book in our “The Woodworker: The Charles H. Hayward Years” series covers all types of woodwork joints, including how to design them, cut them and fix them when things go awry.
It’s difficult to overstate the importance of the book “Woodwork Joints” by Charles H. Hayward (1898-1998), which was first published in 1950 then reprinted many times and in several different editions of varying quality.
The compact 168-page book is beautifully illustrated by Hayward and contains the kind of spare prose that made him the best woodworking author of the 20th century. Like a good woodworking joint, Hayward’s text contains nothing superfluous and lacks nothing important to the task at hand.
In addition to Hayward’s take on joinery, this volume also contains the perspective of other British writers of the day that Hayward published in The Woodworker, including J. Maynard, Robert Wearing, K.J.S. Walker and C.A. Hewett.
A WOODWORKER SUPPLEMENT – LAPPED DOVETAIL JOINT
This joint is used chiefly for carcase construction and has the advantage of being entirely concealed at one side. To enable the joint to be started easily the inner corners of the tails can be chiselled off. The ports should not be joined together too often before gluing as this tends to loosen the joint.
Square the ends, and with a cutting-gauge, gauge dovetails and pins. The lap should be between one quarter and one third thickness of wood. For the pins set gauge to work from inside face. Mark both pieces. Re-set to thickness of tails and mark depth of pins.
Mark out dovetail with template or sliding bevel.
Using a fine saw cut dovetails. Do not let saw run past gauge line. Some find it easier to fix wood so that cut is vertical.
Clamp work to bench and make a sloping cut at gauge line. Chop down 1/16 in. from line, and remove chips by chopping in from the end. When about half way pare down to gauge line Reverse wood and repeat process.
To mark pins fix wood in vice and, with dovetails in correct position, scribe round as shown.
When sawing the pins hold saw to the waste side of marks. Inset shows part of waste cut away with saw.
With chisel slightly in front of gauge line chop down, and remove waste by chopping with the grain from the end. A 1/8 in. paring chisel is used for the corners and a bevel edge chisel for final paring.
Editor’s note: The image at the top of this blog post corresponds with the numbers above.
HOW TO CUT THE LAPPED DOVETAIL
The great advantage of this joint is that it is entirely concealed on one surface and at the same time is extremely strong. This gives it a special value for carcase work since the joint does not show at the sides. The top and bottom surfaces where the dovetails are exposed are out of sight Drawer fronts are also lap-dovetailed (though the spacing is rather different), and here there is the added advantage that the wedge formation of the joint resists the pull which is the chief strain the drawer has to withstand.
It makes no difference whether the dovetails or the pins are cut first; it is mostly a matter of personal preference, though choice may be determined by other considerations. For instance, the top and bottom may have to be glued up to make the width, and it would then likely be convenient to cut the pins in the ends whilst the joints are setting.
Marking out. Trimming the wood to size is the first procedure. The ends in which the pins are cut are obvious; they are the finished size of the carcase as shown in Fig. 1. It is clear that the top and bottom must be short of the over-all width by the combined thickness of the two laps in the ends. This lap size has therefore to be decided straightway. In Fig. 1 the required over-all width is 18 ins. Assuming that the lap is to be 1/8 in. it is clear that the top and bottom will have to finish 17-3/4 ins. long.
Use the cutting gauge to mark the extent of the joint as shown in Fig. 2. Set the gauge to work from the inside of the ends, the required lap projecting beyond, and mark both sides of top and bottom as well as the edges of the ends (see A). In this way the pins are bound to be the same size as the dovetails. Since the top and bottom sink their full thickness into the ends, the gauge is now re-set the thickness of these and the inner surface of the ends marked as at B, Fig. 2.
Dovetail positions. The tradesman usually roughly pencils in the dovetail positions and then saws straightway. Practice enables him to cut the true slope without exact marking out. The inexperienced man should either make the simple marker given in Fig. 3, or set an adjustable bevel to the slope (5/8 in. in 3 ins.). The spacing for normal small work is given at A; B is suitable for a wide carcase in which the small end dovetails resist any tendency for the ends to twist away at the corners. Drawer front dovetails are given at C.
Sawing the joints. Fig. 4 shows the dovetails being sawn. If the wood is 3/4 in. or more thick it is advisable to cut each joint individually, but in thinner wood two or more pieces can be fixed together in the vice and sawn together as shown. Most men saw straight across dead square and it is necessary to mark out first with the trysquare. Some prefer to make a slight taper fit and then the saw is taken at a very slight angle as at A—not more than the thickness of the pencil line. If this is done it is obviously important that all face or outer sides are to the front. Otherwise, instead of showing a close joint there will be a gap. A, Fig. 4, is given in exaggeration for clearness.
Marking the pins. Before the waste is chopped away the pins are marked out. Fix the end in the vice, and place the top upon it in the relative position it is to occupy. The inner end can be supported upon a block of wood of suitable thickness. Hold the top firmly down with the left hand, and, placing the saw in each kerf in turn, draw it backwards as in Fig. 5, so leaving a mark which corresponds exactly with the dovetail.
Sawing. When sawing the pins be careful to place the saw on the waste side of the mark as at A, Fig. 5. Don’t overdo it, but just leave in the marks. This will ensure tight joints without forcing. Judgment on this point is probably the most important point in the whole procedure, and it is something which comes with experience. The thickness of the saw has to be taken into account, far less allowance being necessary for a fine saw than a coarse one. Beginners are strongly advised to cut an experimental joint using the same saw that will eventually be used for the actual joint, and try it together. They will learn more in this way than reading a dozen articles.
Of course, the saw can only be taken down diagonally as shown in Fig. 6, and here again the inexperienced man should mark down with the trysquare first. Part of the waste can be sawn away as also shown in Fig. 6. It all helps to lessen the chopping-out with the chisel.
Another way of reducing the waste is to bore it partly away as in Fig. 7. Except for the thickest wood it is seldom practicable to use the centre or twist bit because the centre point is liable to emerge through the lap, but the Forstner bit is ideal as shown in Fig. 7. It has scarcely any centre point.
Chopping. The procedure is given in Fig. 8. Clamp down the work on a flat, spare piece of wood to avoid bruising, placing it over a solid part of the bench such as a leg. At each socket make a small sloping groove against the gauge line, and then chop downwards about 1/16 in. short of the line. Ease away the waste at the end with the grain, and repeat the process. Finally put the chisel right on the gauge line and cut down.
For clearing out the corners a bevelled-edge chisel will prove invaluable because it will work close in. Many men keep an old stubby bevelled edge chisel specially for the purpose. A, in Fig. 8, shows how the waste is gradually removed. Note that if the chisel were placed on the gauge line at the outset it would be forced in beyond the latter because of its wedge formation.
The dovetails are chopped in the same way, but of course, the chisel is used from both sides. The waste at the corners can be sawn away. When the pins are cut first the marking of the dovetails is done with an awl as shown in Fig. 9. Once again the saw must be used on the waste side.
To enable the joint to be started easily the inner eDges of the dovetails can be pared off. This will permit the joint to be partly put together to see that it fits. It should not be driven right home until it is glued as this is liable to loosen it. Place a piece of waste wood over the joint when knocking it home and strike this. Otherwise the wood may split owing to the local pressure, and in any case it may be bruised.
THE DOUBLE-LAP DOVETAIL
THIS IS SLIGHTLY more complicated than the single-lap joint, but is simpler than the mitre dovetail in that there is no mitre to bother about. In fact, the experienced man can glue up the two parts straightway without first assembling them dry. This is risky in the case of the mitre dovetail because almost inevitably a certain amount of trimming at the mitre is unavoidable.
The joint can be in either of the forms given in Figs. 1 and 2, and to an extent it depends upon the stresses to which the joint is most liable to be exposed, because the dovetail shape resists the The joint can be in either of the forms given in Figs. 1 and 2, and to an extent it depends upon the stresses to which the joint is most liable to be exposed, because the dovetail shape resists the pull more in one direction than the other. The more usual joint is that in Fig. 1, in which form it is often used for cabinet carcases. At the top the joint is entirely hidden, and at the side shows only as a thin line of end grain.
Preparation It is immaterial whether the dovetails or pins are cut first in the joint in Fig. 1, but in Fig. 2 the pins should be cut first as otherwise it is difficult to mark the one piece from the other. Whichever method is followed, the first essential is to square up the wood to the finished size, remembering to allow for the lap when calculating the length of the parts. This scarcely arises when just an isolated joint is cut, but in the case of, say, a carcase the sizes are obviously important. The rule to remember is that the piece with the projecting lap is always trimmed to the finished size, whereas that with the flush lap is less in length by the thickness of the lap on the other piece.
Marking out Assume that the joint in Fig. 1 is to be cut. The first stage in marking out is that of gauging as shown in Fig. 3, because until the thickness of the projecting lap is decided it is impossible to trim the wood to the finished length. Set the cutting gauge to the dovetail thickness (Z), which is the thickness of the wood less the thickness of the lap. Mark the end of the piece with the dovetail, the gauge fence working against the inner surface of the wood, and also the inner surface of that with the pins. Decide on the thickness of the flush lap (it is usually the same as the other), and mark the inner face of the dovetailed piece, and the end of that with the pins (see Y). Lastly set the gauge to the thickness of the wood with the pins, and mark the inner face of the dovetailed piece (X).
Cutting the dovetails The outer mark on the inner face and the end mark give the extent of the rebate to be worked on the dovetailed piece. Saw down across the grain and remove the waste by chopping with the chisel at the end. It may be necessary to trim the rebate afterwards with the shoulder plane. Note the sloping channel chiselled on the waste side of the line as shown inset in Fig. 4, to provide a convenient path for the saw. Mark out the dovetails and saw down as far as it is practicable to cut as in Fig. 5, taking care not to let the saw touch the projecting lap as any such marking would show badly later.
The rest of the work must be done by chopping with the chisel as in Fig. 6, and for this a bevelled-edge chisel is desirable to enable it to reach into the corners. Follow the usual practice of chopping down first short of the gauge line (see top arrow), and do not bring the chisel right on to the line until the majority of the waste has been removed. The latter is done by chopping in from the end (lower arrow). Since the saw cannot reach right into the corners owing to the projecting lap, it is necessary to cut down on the line of the dovetail, but only light taps with the mallet should be given as otherwise the wood may split. Clean out the corners and make sure that the bottom is level.
Marking the pins Transferring the marks to the pins is the next job. The one piece is held in the vice and the other placed upon it as in Fig. 7. A block of wood to the rear will make sure that the wood is level and is steady. Press firmly down to make sure that there is no movement, and pass a marking awl along each side of every dovetail.
The process of cutting the pins is given in Fig. 9. The saw is held immediately to the waste side of the mark, and is taken down as far as the diagonal. Part of the waste can be sawn as at (2), but finally it is chopped away as described last month. The stages are shown in Fig. 9, but it will be realised that each socket is not completed individually before the next is sawn. Rather, all sawing is done, then all chopping.
Finally the inner corners of the dovetail are lightly chiselled away to ease the assembling of the joint as in Fig. 8. Be sure to do this after the marks have been transferred. If the parts are lightly started together it will be obvious whether they fit or not. It is better to avoid knocking completely together dry before gluing.
As I write this, there is still room in Will Myers’ June 26-30 Build a Moravian Workbench class at the Pine Croft School of Woodworking – and that surprises the heck out of me. We get asked all the time if we’ll teach a workbench class, and the answer is a firm no. We don’t have the room or equipment for such shenanigans here – Pine Croft does (plus Will is a lot younger and stronger than we are, and funnier!). So if you want to build a solid bench – one that breaks down for travel – and have a good time doing it, sign up now!
And tomorrow at 10 a.m. (Saturday, Apr. 15) new sessions go on sale, including my October 14-15 Dovetailed Shaker Tray class…but I’d take the Oct. 7-8 class from the guy who first taught me dovetails: Kelly Mehler. He’s teaching a Dovetailed Candle Box – the very class I took with him two decades ago!
p.s. For those who don’t already know, Berea College bought Kelly Mehler’s School of Woodworking, so the classes are in that purpose-built shop, which is nestled into the foothills of the Appalachians, just south of Lexington, Ky., and close to Shaker Village at Pleasant Hill. In other words, it’s in a gorgeous location, and there’s plenty to do and see in the area, in addition to the classes.