A good seat begins with a wide plank that is first shaped with adze and then scraped. Photo by Rich Frutchey.
“Windsor furniture is, I believe, the most characteristically American and the most historically significant furniture style to emerge from eighteenth-century America. It is a democratic style, one which appealed to and was used by all levels of society.”
-Charles Santore in The Windsor Style in America 1730-1830.
The enduring appeal of the Windsor chair is due to its comfort and striking form, and that it is at once both light in weight and very sturdy.
American Colonial chair makers selected their wood with care, finding the most suitable local species of the New World forest. The Windsor style continues to be very popular: it is adaptable to a variety of forms; the spaces between the narrow parts allow other features of the room to be seen; and Windsor chairs can be made by less expensive factory methods as well as by hand makers, like myself, using eighteenth century methods.
The Windsor style is characterized by both the legs and the back being socketed into a solid plank seat. One aspect of the great ingenuity in the design of the Windsor chair is in the use of three or more species of wood in each chair. Because different parts of the chair must withstand specific stresses and must fulfill other design requirements, the project begs for different species of wood for the various components. Each chair needs wood that is: large in dimension yet lightweight for the seat; smooth and hard for the turnings (legs, rungs, and arm posts); easily steam-bent and resilient for the back bows; and straight grained and strong for the slender back spindles. Based on my study of the chair-making tradition as well as my 28 years of experience in the trade, I have my own preferences for which wood species to use, while recognizing that other makers of good chairs might prefer different woods.
A comfortable seat
Even for a small person, a large chair seat is comfortable. Many antique chairs had a seat width of only 16 inches, but I use an 18-inch-wide plank for greater comfort. Basswood and tulip poplar (available farther south) are sometimes used for seats, but white pine is the most common choice. Pine grows wide, carves well, and is light in weight. Much of the weight in a Windsor chair is in the seat. Pine varies a lot in weight, the difference being in the amount of pitch in the wood. I’d like to say that I reject heavy, pitch-filled pine, but the reality is that I’m grateful for any planks wide enough and sufficiently clear that come my way from a few local sawyers who know what I want. Pine sometimes bleeds pitch, which can mess up a nice finish, so some makers prefer basswood for a painted chair or the rich color of butternut for a natural finish. Typically, these species aren’t found wide enough to yield a one-piece seat, so the chair maker glues up seat blanks out of narrower, more stable pieces.
I use planks with a full two-inch thickness, but wood that thick has a lot of stress in it, which often causes cracks near the center of the plank. For this reason, I don’t buy the center or “crown” plank from a log. Since a carefully made chair does not need to begin with a cracked seat, I must reject significant portions of some planks that have cracked during the drying process. Many chair makers expect to glue pieces together to make up their seat blanks, but I prefer the more traditional one-piece seat. However, with my accumulated rejected wood, I sometimes rip-saw sections to eliminate the cracks and join them to form glued-up seat blanks that may be stronger than the stressed wood in a one-piece seat.
Windsor chair making is one of the more complex projects of the “green woodworking” method, which refers to using fresh-cut wood rather than relying on dry dimension lumber. Except for the seats, I rive (or split out) all the other chair parts from fresh logs of white oak, hard maple, and ash. The riven wood cleaves along the continuous fibers of the wood grain. As I work, it’s easy to see vagaries in the grain, get a sense of the strength of the wood, and note the smoothness with which the wood cleaves. Pieces with curvy grain, knots, cracks, and other imperfections I condemn to the firewood pile. A really nice-looking chair has pieces of slight dimension, which makes it appear to be delicate. But when made of rived-out stock, it is strong, because the grain of the wood continues straight along the chair parts. In fact, a chair could be strong enough but have parts so thin that it lacks a pleasing visual presence. Factory-made chairs, on the other hand, need to be at least a little clunky to make up for the possibility of cross grain in the parts made from dimension lumber, which is sawn straight even though the fibers of the wood grain may curve. Your favorite (but broken) factory-made chair might well have failed where the wood grain crossed the rung, leg, or spindle because the piece was sawn straight from a curvy or twisted log.
I enjoy riving the logs with hammer and wedges and then going at these smaller pieces with mallet and froe. Then I make the stock into finished chair parts by using a drawknife and spokeshave, or hand-held turning chisels at the lathe. Such tooth and- claw methods put me into direct contact with the character of the wood in each log that I work up into chair parts. At the lathe, the thin shavings reveal the character of the wood: brittle wood will break when doubled over, but strong and supple wood could be tied into a knot.
Another advantage to splitting out the pieces is that if the wood is a little “dozy,” or weak, possibly from a diseased tree, it is apparent in the way that the wood cleaves. Dozy wood is another candidate for the firewood pile.
Choosing a tree in the forest before it’s cut into logs would certainly be a joy and allow me to better judge the health of the tree, and thus the health of the wood, but it is much less time consuming to buy a log from the local sawmill, where I usually can select a single log from a large assortment. It’s been an expensive education learning how to select a good log from the pile and judging the quality of the wood inside the log by the bark surface. No twisting grain, no knots or even lumps, please. Once the log is paid for, transported to my yard, and split open to reveal the character of the wood inside, it sure is difficult to give up on one that isn’t the best. For riving and shaving the wood by hand, I love creamy wood. And creamy, smooth-looking bark means creamy working wood inside. Even log inspectors buying for the mass market might pay more for “tight bark” maple than for logs with coarse bark because the wood inside is smoother.
The turnings: legs, rungs, and arm posts
Legs, rungs, and arm posts are turned on a lathe from riven wood. Turning these somewhat irregular pieces down to round is a little like running a jackhammer. I try to have on hand a sturdy young person trained to help with this task. I store the round stock until I do the finished turning, checking to make sure the wood doesn’t get moldy before it dries.
I use sugarmaple for turning the legs, rungs, and armposts even though my woodlot has many large, straight beech trees that perhaps should be harvested before beech bark disease kills them. It would be very satisfying, as well as cost-saving, to harvest the beech for turning stock instead of buying high-priced hard maple logs, but beech does not dry well when it is turned green and is not as hard as sugar maple. The joint where the rung goes into the chair leg requires a very hard wood to withstand the compressive racking force over generations of hard use that could cause a rung to pull out of the leg. The wood swells in summer’s humidity, causing the wood fibers around the joint to crush. These parts shrink in winter’s dryness, which might cause the rung to loosen from the leg.
It’s possible to “make do” with the softer red maple and white birch. You could compensate for this softness by compressing the tenon before assembly on the ends of the rungs and legs, which are socketed into other parts. You could abandon the use of hardmaple because it is costly and because it is impossible to verify what the grain is doing in hard maple without seeing a riven surface. Still, I prefer to stick with sugar maple, as it is the hardest suitable wood that is available to me, and it makes very crisp turnings. Maple seasons well by staying smooth and remaining stable in the selective drying process. I dry the ends of these parts, where they will be socketed into other parts, in hot sand before assembly to make sure that they will not shrink further, which would cause the joint to loosen.
When I have a chance to examine an antique chair, I like to see if there are riven surfaces on any of the turnings. If so, the chairmaker oriented these flat areas toward the floor or toward the inside of the chair, where they would be less visible. Turning rived out pieces down to round takes a lot of power, so the chairmakers working before the days of electric power sometimes left not quite enough spare material to make them completely round. These riven-surfaced (rather than sawed) flat spots are an indicator of early methods.
White ashmakes nice spindles for the chair back. It is very lightweight, strong, and stiff. Quite early on, chair makers learned that it was faster to turn spindles on a lathe rather than to carve them with drawknife and spokeshave. But turning long, whippy spindles at the lathe requires a lot of concentration and a little added thickness to make up for cross grain. Also, the strongest woods available for such delicate looking parts, such as oak, hickory, or ash, are ring-porous, so they don’t turn smoothly like birch or maple.
Riving the parts and carving the spindles offers an opportunity to get some work done outside the confines of the shop. I prefer to carve, rather than to turn, the spindles since I can do it under a shade tree in the summer, in my sunny shop porch in the winter, or at a craft fair where I can get some work done while entertaining the kids who love to watch the coiled shavings peel off. Straight, close-grained wood from old-growth trees may be best for casework and molding, but in my experience, I have found that healthy, fast-growing wood is the strongest and the most supple for spindles. The creamiest log of white ash that I ever worked had been growing very rapidly, showing only three growth rings to the inch. Ash is nice for spoke-shaving to a smooth finish and holds a painted finish reasonably well. At the sawmill where I buy logs, I look for a white ash log that, judging from the spacing of the annular rings visible on the ends, had some years of fast growth. I rive the ash into radial boards, then reject the heartwood because it is weak, and with many logs I must reject the last years of brittle, close-grained wood that forms in trees that have been stressed.
Reference books about antique chairs all refer to spindles of hickory, and it was sometimes used for steam-bent parts, too. This species, however, was the shagbark hickory mostly found farther south. The hog walnut or pignut hickory that grows in the more northern reaches of the Northeast forest lacks strength and often has curvy grain or pin knots. It is hard to paint, the pores open up to white lines when the paint dries, and the paint does not adhere well.
Some people consider red oak an alternative for spindles. Red oak is another species whose workability and strength vary considerably from one log to another. Red oak that is very lightweight and easy to cut is weak and should be rejected. The last time that I used red oak for spindles, I found it to be too squishy, coarse, and soft, and it did not finish smoothly. Sometimes the spindle failed when I tried to pound its oversized tenon end into the seat. Ash is generally stiffer, so it withstands the hard hammering into the seat sockets. White oak is very suitable but heavy, and I prefer to keep the chairs very light in weight.
There are many different configurations for the back of a Windsor chair. I usually make bow-back chairs, as they are the easiest to make and the most durable. I bend the bows using hot steam, as the lignin, which is the substance that binds the cells together, softens at this temperature, allowing the cells to slip on each other, and thus the wood to bend without breaking. There is tension on the outside of the bend that wants to pull the fibers apart, and on the inside of the bend, there is a compressive force that might cause the wood to buckle. Add to these destructive forces the fact that the bow is pierced completely through by all the spindles. Remember, these chairs are delicate-looking pieces with no extra dimension to them, so I use the very best wood that is available to me. White oak is my wood of choice for the bows. It is very strong, bends well, and has a finer grain than red oak or ash. White oak in this region is on the northern edge of its range and so is often stressed. The wood is scarce and tends to be of low quality. I have been fortunate to be able to buy logs from concentration yards or mills that draw from regions farther south.
The best wood for the task is quite critical for my peace of mind, since I guarantee my work for a lifetime – my own or the customer’s. Developing an eye for high-quality wood of the correct type for Windsor chairs takes some time and experience. But the rewards of learning to reject inferior wood have helped keep me in business making a quality chair for nearly three decades.
George Ainley has been making Windsor chairs for 28 years in his shop near Perkinsville, in southern Vermont. He offers chair making instruction at his workshop. Visit www.vermontwindsors.com.