TURNING

Checkered Bowls
Reinterpreting in wood the designs of the American Indian
by Irving Fischman

After turning bowls for several years, I have recently begun to explore the classic designs used since antiquity in pottery and basket making. I am now particularly interested in the pottery and basketry of the Indians of the Americas. Simple shapessuch as truncated cones or bellsare used to counterpoint intricate painted or woven patterns. I have tried to reinterpret this design approach into a different medium, wood. One bowl in particular has a simple bell or trumpet shape and a checkered pattern of teak and black walnut, woods that are richly contrasting. To make this bowl, familiarity with lathe work and a supply of clamps, both band and deep throated, are essential. Basically, the bowl is made of three layers of wood. A 15 or 16-inch square of one-inch walnut forms the top layer of the bowl and an 11-inch square of two-inch teak forms the bottom. In between is a checkered ring of teak and walnut one inch thick. Both the top and bottom pieces should be

planed or sanded flat and cut into a disk shape. To provide a means of attaching the faceplate, I glue a piece of 3/4-inch Baltic birch plywood directly to the bottom of the teak. No sheet of paper is used between them because the mass is so large that the paper might fail during turning. I use birch plywood because it is far stronger than either fir plywood or solid wood. The checkered ring in the center of the bowl is not an inlay as first observation of the completed bowl might suggest. It is a separate layer composed of solid truncated-wedge-shaped pieces glued together to form a ring. This technique is much easier than inlaying on the curved surface of a turned bowl and the pattern can be seen on both sides of the bowlin the manner of Indian baskets. The ring has twenty-four pieces, the smallest number that I felt would have a pleasing visual effect. Larger numbers of pieces are possible, but the accuracy of the angles of the pieces becomes correspondingly more critical.

Bowls in various stages of construction are shown in a flow sequence.

With 24 pieces the base angle of the wedge is 82.5 degrees or 7.5 degrees from the vertical. If the outside edge of each

wedge measures 1-7/8 inches, the ring will have a diameter of

approximately 14 inches (from the formula circumference equals pi times diameter; the circumference in this case is 24 times 1-7/8 inches).

identical pieces each of teak and walnut, flipping the board after each pass to get the wedge shape. If the angle has been set correctly, the 24 pieces should

form a perfect ring. This can be checked by clamping the pieces dry with the band clamp. Slight errors can be corrected
by sanding. Alternating the teak and walnut blocks, glue is applied to their edges and a belt clamp tightened around them. (A

The wedges are cut from identical pieces of one-inch teak

and walnut three inches wide by 27 inches long. In cutting the wedges on the table saw, the blade is kept vertical and the cross cut guide is set at 82.5 degrees. First, the end of the board is trimmed to this angle. Then the board is flipped over and the 1-7/8-inch base length is marked on the edge. This is used to set the distance between the fence and the blade. Because the crosscut guide is on the opposite side of the blade from the fence, using a fence for measuring each subsequent

tourniquet could be used instead.) I simultaneously clamp

the whole ring between two disks of 1/2 or 3/4-inch plywood, protected with wax paper, to assure that the ring is flat. I use yellow glue (aliphatic resin) throughout because it hardens

and is stronger than white glue. Plastic resin, resorcinol, or

casein are other possibilities. After allowing the glue to dry thoroughly, the ring is removed and touched with a sander or plane to make sure it is flat.

piece would cause the blade to bind. Therefore, we actually measure to a short hardwood block clamped at the front of the fence, which acts as a distance marker. Now we cut 12

It should be noted we are gluing end grain which is not

especially strong. However the ring will be glued between two

Approximate cross section of bowl showing placement of the layers. Bowls are turned by eye without templates.

solid layers and we have the advantage of not turning end grain. The entire turning block can now be assembled. The teak
base (with birch plywood block already attached), checkered ring, and walnut top are glued together and clamped using many clamps to assure good contact. Concentric alignment is important and either circles or crossed diameter lines drawn on the blocks will help. The block is allowed to sit one or two days to assure that the glue in the interior is dry. A faceplate is now attached to the birch plywood on the completed turning block. Because of its large size, the block

must be mounted securely on the outboard side of the lathe. I use a heavy duty lathe with four speeds600 and 1200 rpm for faceplate turning2400 and 3600 rpm for spindle turning. The handrest is mounted on a moveable tripod.

Truncated segments are cut with the cross-cut guide set at 82.5 degrees for 24-piece ring. Block clamped to rip fence safely eliminates need for measuring each piece.

With the lathe at its lowest speed, a heavy scraper is used to round the walnut top disk, both for balance and to determine the final dimension of the bowl. Then a gouge is employed to rough the outside shape. Starting at the bottom, material is removed from each of the three layers until there are no gaps

between the layers and the tool cuts solid material throughout its path. Keep in mind that the three layers will cut differentlythe easiest is the teak and the hardest the
walnut. Also, teak is notorious for dulling tools and frequent sharpening and honing will be necessary. Final outside shaping is accomplished with a scraper, taking very light cuts Any small tool can be used to apply the radius on the bottom edge. At this point I usually sand the outside of the bowl completely since I can apply as much pressure as I like to the still solid block. In other words, I don't tackle the inside until

the outside is completely finished. The inside of the bowl is tackled with the handrest facing
the top of the bowl. First I remove with a parting tool a central disk of walnut eight inches in diameter. This disk

Tourniquet clamp holds glued segments together while two plywood disks protected by waxed paper are clamped to hold
checkered ring flat between them.

Checkered ring is touched with sander to make sure it's flat; then it's glued between the two-inch teak base block and the one-inch walnut board that will form the top rim.

Outside of bowl is turned and finished first with lathe set at low, 600 rpms. Top walnut board gives the bowl stability during this stage. The handrest is on a moveable tripod.

matches the hole in the checkered ring and has not been removed until now to assure rigidity during both the gluing and the outside turning process. Because the parting tool is
not coming in from the side, be sure to widen the groove that it makes so that it doesn't get caught. When the cut is completed, the central disk is easily removed. Now I usually remove some material from the center of the bowl, so that the handrest can be moved in to act as guide for roughing out the inside. I prefer the scraper for this operation so there is no chance of the tool catching in the wood. The bowl takes shape very quickly now. The sides, completed first, are made parallel to the outside and between 1/4 and 3/l6-inches thick. Don't make the sides any

thinner or stability becomes a problem with these laminated bowls. The sides taper slightly at the rim and gradually increase in thickness at the base. A higher lathe speed should be used to finish the bottom. The bottom is left between 3/8 and 1/2 inches thick to give the bowl a solid feel. The interior and rim of the bowl are now sanded. Teak
sands very nicely, and I use only grits 60 to 120, wetting the surface occasionally to bring up the grain. The entire bowl can now be burnished with a clean rag if desired. The completed bowl is split from the Baltic birch backing and the bottom is hand planed or sanded flat. Finally, I prefer to give the bowl a rich oil finish, but a glossier finish can be tried. The finished bowl takes five hours to complete and about $10 in materials. The present design can readily be seen as a jumping off point for many variations. Contrasting veneers could be placed between the layers to form stripes in the

Thickness of bottom is measured and kept between 3/8 and 1/ 2 inches to give bowl a solid feel. Use a higher lathe speed

for bottom. Author sells the bowls for about $80.

finished bowl. Different numbers of pieces and different woods could create other patterns. However, I feel that the basic design approachusing a simple shape to compliment intricate patternsis essential to
a satisfactory finished product.

turned first parallel to the outside and no less than 1/4 to 3/16 inches thick in the interest of stability.

Inside of the bowl is now turned and completed. Sides are

Once outside is completed, a parting tool is used to cut the

central disk from the walnut top. Widen the groove so the tool
doesn't get caught.

When cut is completely through, centrifugal force holds disk in place until lathe is stopped. The disk should be made smaller

than the inside of checkered ring.

DEVICES.

Library Ladders
How the British had their steps and hid them too
by Alastair A. Stair

The image of a little girl converting the Castro sofa into a bed has been a familiar one to television viewers for years. Convertible furniture however, is not an invention of the twentieth century. The idea of creating household pieces to function in two or even three different ways has challenged furniture makers for centuries. The English maker in particular has continually demonstrated a special ingenuity for combination pieces since the seventeenth century. English technical skill has spawned all manner of technical devices that today delight the collector of English antiques. One specialized form that clearly illustrates this fact is the English library ladder. Library steps came into general use around 1750, and it is not a mere coincidence that this development was concurrent with the flowering of the Industrial Revolution. Various

technical advances made at this time were instrumental in both instigating the need for library ladders and in influencing the forms they were to take. Fostered by a favorable climate of conditions, the Industrial Revolution began in England toward the middle of the eighteenth century. One of its offspring was the quicker and less costly printing of books. A greatly increased production continued until 1798 when Earl Stanhope of London invented the iron press which made the work even more easy and rapid. As a result, more books were in circulation in the second half of the century than in the previous century, and more attention was devoted to the library and to furniture for it. Books were no longer locked away from view, and new prominence was given to the library breakfront bookcase that was often so tall that the upper shelves could not be easily

Convertible chair was commonly used in the Regency period. This chair is hinged at the seat rail so the back swings forward after a catch is released. The scimitar-shaped leg was quite popular, although other styles were also used. This chair fetches $1450 in the antique marketplace.

reached by standing on stools or chairs.

The ladders that were made to meet the need were often cleverly contrived to serve a dual purpose, with form and function uniting in most striking ways. Inspired by the new preoccupation with technical devices, English furniture
makers of the eighteenth century, and later in the early nineteenth century, employed simple mechanisms which

enabled chairs, stools, book stands and even tables to ingeniously convert into library ladders. Appearing along with straightforward step ladders and pole or "elephant"
ladders of all shapes and sizes (sometimes even spiraling, like

pulpits) these double-purpose pieces constituted a great variety from which the contemporary gentleman could choose. Playful, sometimes over-ambitious in its attempt to
combine beauty of line with practical function, yet often

quite elegant in design, the English antique library ladder has a unique appeal for the connoisseur.

One remarkable device of this kind is the library table that

folds out into a step ladder. Sheraton illustrates two designs for this form in his Drawing Book (1791-4) and according to the famous designer, such steps were apparently first made for King George III. Thereafter they became quite popular
because of their simplicity and ease of use. The steps can be put up in half a minute, and the whole can be taken down and enclosed within the table frame in about the same time.

The hinged top simply folds out and down to the floor to reveal a series of sturdy steps, with the uppermost step resting approximately five and a half feet from the ground. The inner
horse relieved by springs, unlocks, and becomes erect, along

An inlaid Sheraton library table was apparently first made for King George I I I ; convertible stools and benches (below) came in various designs. The cane-shaped handle

is held in a slot; everything else hinges. Because of their unusualness, devices like these don't stay long in the

marketplace and bring from $1800 to $2800.

with the multi-hinged handrail. The whole is supported by four strong, square legs. These ladders created marvelous architectural skeletons when open, and some display a very elaborate, often rhythmic handling of vertical, horizontal and diagonal effects. The visual aspect of the table when closed was considered as important as the function, and handsome woods were employed, often with decorative inlay and

elaborately inlaid case, has been attributed to Thomas Chippendale and dated c. 1770-1775. Over the years I have seen many examples of this type and they provide extra seating along with the additional advantage of a tool for saving space. It is easy for even a child to lift the upholstered seat and pull out the ladder or to carry it all the way down to

the floor where the seat can act as an upright support. A more rare article of cabinet work is the book rest,
pedestal or plinth, on which one side opens to produce a series of steps. Intended to support large, heavy books or folio volumes, these provide the maximum of convenience in the smallest amount of space. The enclosing cabinets were constructed of the finest mahoganies and endowed with carved effects. They were usually mounted on casters. As the draftsman's pen and the cabinet-maker's skill produced whimsical versions of ladders, much scope was provided for the contemporary metal worker and he was very adept in the art of forging the necessary hinges, springs, locks and metal mounts of all kinds. The metal worker of the Industrial Revolution enabled the designer and furniture

sometimes with painted ornament. Another special type of library ladder is the convertible chair, used commonly in the Regency period. In the decades of the late eighteenth and early nineteenth centuries,
archaeological excavations at Pompeii, Herculaneum, and in Egypt resulted in a lively interest in the furniture forms of ancient civilizations. One of the most popular chairs of this period possessed in-curving "scimitar" shaped front legs, wide arcing back uprights with reeded members curving in opposing directions in the manner of classical prototypes. This chair, with its graceful design, made a very attractive object when converted into a library ladder. A third type is the library stool that can be adapted into a ladder. Of necessity rectangular, and usually rather long, these were most often upholstered in leather and make a

craftsman to allow their imagination to run free in the pursuit of exciting, convenient and pleasing forms of ladders for use
in the library of the eighteenth century gentleman of taste and distinction.

quite handsome appearance. One such stool, with an

Trickiest of the lot is this book rest,

shown with one side folded down. It's been many years since the author last saw it work, but there's a hinge in the middle tread, a concealed pin hinge above the top tread, and what looks to be a hinge (or maybe a catch) where the small leg supporting the top tread meets the longer leg below it.

SHOWCASE

A Serving Tray
How to use up waste veneers and please the family as well
by Lionel Kay
It is always a pleasure to find a way to use up what might be called "waste" materials around the shop. My wife had asked for a new serving tray, so I was able to fill her wish with this project and have a lot of fun for myself at the same time. A classic pattern for laying veneers is known as the "X" design. It is a variation of the diamond pattern. I chose the "X" to use up a quantity of narrow mahogany veneer cutoffs left over from some previous project. The narrow strips are trued up with edges planed and parallel. Ends are cut at an angle of 55 degrees, though this is optional. Different effects will result with different angles. Strips are then arranged to form rhomboids, whose dimensions conform to the size of tray desired. The rhomboids (four are needed) are then cut and reassembled to form rectangles. These, in turn, are trimmed and taped together as one large surface ready for laying. For a diamond pattern, the right and left rectangles would be interchanged. I had a strip of inlay to attach as a border and then added four strips of satinwood as an additional margin. Corners mitered and the whole taped together, the top was ready. The tray core was a piece of 1/4-inch luan plywood. The bottom can be plain or fancy, as you wish. The top, core and bottom were laid at one time and were ready for finishing. The molding was cut from a piece of 3/4-inch pine. The bevel is approximately six degrees, though again, it is a matter of taste. Top and bottom were veneered with very straight-grained sepele. The inside edge was made from glued up avodire/walnut/avodire, and the outside edge was a thin strip of zebrano that was just lying around waiting for this job. The rabbet was cut for a snug fit with the tray surface and the work on the molding was done. I recommend finishing (according to your own preferred schedule) before assembly. I finished with an alcoholresistant urethane varnish. It is best to finish the molding in one or two long strips before cutting to size and mitering. This is similar to prefinished picture frame molding and assures accurate fit at corners. With finishing completed, the whole is assembled with glue and a single nail (counter-sunk and filled) in each mitered corner. No glue in the rabbet to allow for expansion and contraction!

Dashed lines in bottom drawing show how veneer is cut to make a rectangle from a rhomboid. Middle drawings show cross section of edge, including layering of veneers.

SHOWCASE

Stamp Box
What happened when the old relic box reached America
by Stanley D. Saperstein
The 17th Century stamp box is a little known antique in America, but widely known in Europe as a "relic box" to house prized religious relics. It makes a beautiful desk ornament. Because the relics were quite small, something was needed to house them, and small decorative boxes were carved out to serve the purpose. When the first Catholics arrived in Lord Baltimore's Maryland colony, the relic box came along, but in America there were no relics to be sold and the pretty little boxes were soon being made by local craftsmen to house different types of colonial stamps. The stamp boxes seemed to disappear after the Revolution and very few survive. The boxes varied as to size and could be square or rectangular. The most common sizes were three inches square, four inches square and four by five. The best woods for this are walnut, mahogany and cherry. Pine may be used, but it tends to break easily. This one is made of walnut and the bird is a caricature of a tufted titmouse. Owls and titmice were considered good luck and were common on stamp boxes. The box itself is made from one piece of wood (laminate if necessary) with the center scooped out. The easiest way to accomplish this is to honeycomb the center of the block with a drill. Remove the remaining wood with a gouge, and straighten the sides with a chisel. Carve the outside of the box before you take the center out to avoid cracking the sides. If you wish, leave the sides plain. The 3/8-inch thick lid is rabbeted to fit into the center block. To make the bird, use the pattern provided or preferably create your own any small bird will do, in a caricature style. Don't paint the bird; give it a natural finish and fasten it to the lid with glue or a screw. The inside of the box is covered with leather or felt.

All in One
Putting together favorite features
by Brian Considine

This piece is a synthesis of the features of many pieces that I have likedscalloped sides, a hutch top, library shelves with small drawers, etc. It's designed to be used either in a librarytype setting for books and the like, or hung over a sideboard in a dining room. The case is dovetailed together, top and bottom. But scallop the sides first and then cut dados for the shelves and the top facing piece. The molding is then mitered and nailed and glued on. Then the dividers are sized and dovetails are cut at either end and the case notched for them. The drawers have a 1/4-inch thumbnail on the face. They are dovetailed in the front and shouldered on the back. The bottom fits into the grooves in the sides and the front, and is nailed to the back.

Shaker Lap Desk

A challenging exercise in hand dovetailing
by Brian Considine

The Shaker lap desk may be considered the predecessor of today's briefcase. Like modern briefcases, lap desks were convenient for carrying papers. In addition, one can write on either the slanted top or on a special surface inside the desk. These desks are fine examples of the thought, care and skill the Shaker woodcrafters put into even the smallest piece. The variety and detail of the joints in this piece make it a challenging project for the modern craftsman. The general procedure for making these is to make and assemble the four sides and shelf, then the bottom and top, and finally the drawer. First make patterns of the large and small ends on pieces of manila cardboard and lay out and cut the dovetails' outline on these patterns. (In making dovetail joints, I like to make the tails first.) Next, size and plane the stock from the cutting list. Cut the pieces for the front, back and ends 1/8-inch over in width so that you can plane down the top edges when it's together. Cut the angle on the top of both the ends and leave a little over. With a marking gauge, scribe the depth lines of the dovetails on the end pieces. Now align the patterns along the bottom of the end pieces and transfer the dovetails with a scribe or awl. Then cut and clean them. Now position the four sides as they are to be and code the corners that are to be dovetailed together. Then one by one mark the lines for the

pins from the tails and proceed to cut them out. To help in the chiseling process, I like to make relief cuts with a bandsaw into the waste sections of the pins. But take care to place the piece on the saw table with the wide part of the pins up so that you don't cut into them by accident. Then fit the dovetails together, carefully paring down the pins where necessary. Size the shelf and lay out the grooves to hold it on the inside of the four sides. The bottom of the shelf should line up with the bottom edge of the small end. Cut the groove for the shelf with a router or with knife and chisels, but be careful not to go to the outside edge and cut through the dovetails. Then glue the sides together, placing glue blocks just beyond the dovetails so that they can close all the way. Clean any dried glue off with a chisel and smooth the corners with a block plane. Plane the bottom edges so that box sits flat. Plane the top edges so that they all lie on a plane. Size the bottom piece, round the edges and glue it on. Size the top. Glue on bread board ends (tongue should be 3/16-inch wide). Cut 1/8-inch molding to go around the mouth of the drawer. It should be rounded on the edge and protrude slightly. (The drawer will slide straight if you put extra molding behind the two sides of the opening. This should be done before gluing on the bottom of the desk.) Now size and fit the drawer front and proceed to dovetail

This lap desk was made of 200-year-old pine (on commission), although the author prefers to use cherry for the desks. Shakers often made dovetails as shown here, but to be more correct, dovetails should have been laid out to begin and end with half-pins, not half-tails.

Desk Parts
Large End Small End Back

Front Bottom
Shelf

Top

Drawer Parts
the drawer together. The front dovetails are of course half-blind. Dovetail the sides to the front first, put them together and slide them in the drawer opening. Cut the sides a little long so the drawer front will protrude at first. Then measure and cut off enough from the back end of the sides so the drawer front sits flush with the side of the case. You still have to size the drawer back and bottom and

Front
Molding Sides

Back
Bottom

dovetail the back to the sides. First cut a groove on the inside of the sides and front to hold the drawer bottom. Note that it is essential that the bottom edge of the groove be above the bottom half-pin. Otherwise when you pass the pieces through the table saw you will cut the dovetail. After you make these
grooves, size the drawer bottom to go in the grooves and bevel the edges. Then measure for the back piece. It will be as wide as the drawer front but its height will be the distance

from the top edge of the groove to the top of the sides. Dovetail this back to the sides by repeating the procedure for dovetailing the case. Finally, the drawer bottom should be slid in and nailed to the back piece from its underside. Once the drawer is completed, bevel the back edge of the
lid so it is flush with the back edge of the box and hinge it.

Green Bowls
Turn unseasoned wood, dry it, then turn again
by Alan Stirt

A big problem in bowl turning is obtaining thick, wide, dry wood. You might be able to get 4-1/2 or 5-inch thick mahogany or 4-inch teak from an importer. In the Northeast you might find some 3 or 4-inch maple, birch or cherry at local mills. These planks usually contain numerous checks and splits. If they are sound, they will be more expensive than thinner material. If you want to turn a number of bowls, such sources will be quite frustrating in terms of cost and available species. However, green (unseasoned) wood can readily be found and is often free. Even exotic woods are much cheaper when bought in the log. Working directly from the log gives you an opportunity to fit sizes and grain patterns to your own requirements, rather than accepting material that has been milled to a predetermined size. Green planks also offer advantages over dry wood. You can get larger sizes (the sawyer won't mind cutting extra-thick planks if he knows that he won't have to dry them), and the material will be in better condition. In rural areas, logging waste often containing the most figured wood sawmill slabs and storm-damaged trees are usually free or sold cheaply. Firewood piles yield nice chunks of local hardwoods. Small local mills usually are glad to cut logs to whatever dimensions you want. Here in northern Vermont, mills charge $40 to $50 per 1,000 board feet for milling logs that you bring them. If you buy a log from the mill and have it cut, the cost is 20 to 30 cents per board foot. If the log is in good condition, such material is virtually check-free. Even in cities, green wood can be had from local tree-removal services and highway departments. After you've found a supply of green wood, you have to dry it. One way is in planks or bowl-size blocks, but this is unlikely to produce perfect material. The easiest method is to turn the wood when it's green. Once the wood is in a bowl shape it dries much faster and with fewer defects than a solid chunk. You might start with a slab of lumber 4 or 6 inches thick, but if you turn the walls of the bowl down to an inch, it dries more like 4 / 4 stock. The analogy isn't exact because the grain orientation of the bowl isn't the same as that of milled lumber, but proper drying procedures minimize the differences. As the bowl dries it will warp and shrink, but once it is dry the walls are thick enough to be turned true again. As an example of green turning, I'll show how to get a dry bowl from a green log of lignum vitae about 9 inches in diameter. It had been drying for about two years, but it was still quite wet. Similar procedures can be used for most hardwood species, both native and exotic. First, cut about an inch off the end of the log to find check-

free wood. If the log has been in the sun, it may be necessary to cut a series of thin slices to reach sound material. In some hardwoods small center checks run the whole length of the log, but these will be removed when trimming the block for the lathe. Next cut off a cross section as long as the diameter of the log, and rip this piece along the grain through the center of the log. If there are any center checks, make this second cut parallel to them and the saw kerf will often obliterate them. It is important to make sure the center of the tree the pithdoes not end up in your bowl as it will almost certainly split. Note any other checks and defects and plan your cuts to eliminate them from the final shape. Next, flatten the outside of each slab. This will be the bottom of the bowl. The flat surface will make the block safe to cut on the band saw. On the lignum vitae I roughly flattened the bottom with a 1-1/2-inch carving gouge, but these cuts can be made with a chain saw or a band saw. To cut down vibration and make turnings easier, I taper the sides of the block. I used the gouge but the easiest way is to saw a tapered circle. My band saw just doesn't have the capacity to make this cut. The more you refine the shape with hand or power tools, the easier the initial turning will be. How far you go depends upon the size and species of your block of wood, the size and weight of your lathe, and your confidence and skill in using your tools. It's best to start with a balanced shape and discover how much unevenness you and your lathe can take. Even a small, out-of-balance piece can cause a lot of vibration. First I turn the back of the bowl, with the face that was at the center of the log attached to the faceplate. Use long screws to grip the wet wood since the bowl will be absorbing a

Bowls turned from green wood by the author. Largest, 15 inches across, is of quilted, broad-leaf maple. Others (clockwise) are from zebrawood, white ash and cherry burl.

number of hard knocks in getting it true. Even if you don't usually wear a face shield when turning, it's important to wear one now. In the early stages chips will fly in all directions and some of them will be rather large.

warping and checking; however, if the drying is too slow the

wood may succumb to fungus and decay. And the slower the bowls dry, the more storage space the turner needs.
One controlling factor is the coating on the bowl. If left
unsealed, the end grain will dry much faster than the rest. This can result in checking. Wax evens the drying rate and

Before turning on the lathe make sure the wood will not hit the ways or the tool rest. I start at a low speed and use a gouge, taking light cuts at first.
Don't try to decide the exact shape until all the rough spots are gone. Once the bowl is true, stop the lathe and carefully examine the wood. Note any defects which have to be removed, and interesting grain patterns to develop. The

slows the whole process. So far I have used only paste wax. I'm sure any sealer that would adhere to wet wood would

work to some extent. If I find that one layer of wax is not preventing checking I'll add more. The more layers of wax,
the slower the drying and, up to a point, the less the chance of checking. Each species of wood dries differently. In general, the higher the density of the wood, the longer it will take. But even within a single species the density can vary greatly.

shape and the grain can be made to work together to create

something more than just a bowl. On the lignum vitae bowl, I cut quite a bit off the bottom to ensure an interesting balance and pattern of heartwood and sapwood.

In shaping a bowl, I find the gouge to be the most efficient

and enjoyable tool. The wood cuts cleanly and thick, curly shavings usually fly from it. Lignum vitae is an exception, preferring to come off as chips. Some woods, particularly butternut, are so soft and stringy when wet that they are hard to cut with anything but a gouge. A scraper just pushes the

Sapwood will generally dry faster than heartwood and can cause extra distortion in bowls where both are present.
Among domestic hardwoods, cherry and apple check easily while elm, walnut and butternut are excellent; in general, fruitwoods are more susceptible to checking than nutwoods. Ash may check within minutes. This particular variety of lignum vitae proved to be very stable. Although I had to be very careful about checking, hardly any distortions occurred (by using many layers of wax and slow drying conditions I lost only one bowl out of 15

fibers around. To cut the straight foot, I use a 1/4-inch gouge with a slightly pointed nose.
When the contour of the bowl is done, flatten the bottom and make a pencil line to help reposition the faceplate. Before remounting the bowl, I drill down to 1 inch from the bottom using a 1/2 or 1-inch bit. This gauges the depth

completed ones). These bowls I turned from 1/2-inch to

3/4-inch thick. I dry most native wood bowls, turned 1-inch thick, for about three months. I dried the lignum vitae bowls from six to twelve months, according to size and thickness. You have great control over the drying environment, and the environment is crucial. Temperature, humidity and air

and makes the gouge work easier. The faceplate can now be mounted on the bottom, using shorter screws because the wood will be running true. If you align two of the screws with
the grain direction, the holes will probably remain in line during drying. Jot the screw size on the bowl for remounting later.

circulation are the important factors. In the winter I never

start bowls drying in a heated room. Usually I'll dry them in a spare room which stays around 45 or 50 degrees with moderate air circulation. After some weeks the exact time depending upon the experience with wood of this species and

First I clean up the front, taking light cuts with the gouge.
This can be a great help in reducing vibration, particularly if a chain saw was used to cut the log and the front is uneven. Now the bowl can be hollowed out. Because the wood is wet the tools stay cool and large amounts of wood can be removed before resharpening. I usually start at the center and work out

grain formation I move the bowls to a heated room.

A room which has good drying conditions during a period of high humidity can become an oven when the humidity drops sharply and stays down. Often the conditions can be
changed just by opening a door, for increased circulation and

toward the rim.

It's important to keep the thickness uniform throughout, so the bowl will dry evenly with less risk of checking. The
thickness is very important in determining drying time, and a

bowl turned down to 1/4 inch would dry very quickly with little chance of checking. However, it would distort more than a thicker bowl and when dry would be nearly impossible

faster drying, or closing it, to retard drying. If you want to be more scientific, you can outfit a room with temperature and humidity controls.

Once I found some 12 by 6-inch cherry bowls had checked

to turn truly round. For most native woods leave the walls and
bottom about an inch thick. I gauge the thickness with

calipers as the bowl nears completion, and examine it carefully for checks and knots. Checks present when the bowl is wet will get larger as it dries, and knots will often start
checks that spread through the wood. If you're satisfied with the condition of the wood, start the lathe and coat the bowl with a heavy layer of paste wax. I use Johnson's paste wax because it's cheap and I purchase it by the 12-pound case. Wax the bottom after removing the

during their first few days in my "normal" drying conditions. I dug out the checks with a gouge and rewaxed the bowls. Then I put them in my cellar which has high humidity. The bowls gradually dried without checking. However, they developed an unattractive blue-green stain from a fungus
which thrives on high humidity. I completed the drying in a

heated room and then finish-turned the bowls. The stain

went deep into the end grain and was visible after finishing. I later dried the cherry in conditions that represented a
compromise between my spare room and the cellar.
It pays to experiment with the facilities you have available;

faceplate.
It's a good idea to rough-turn in an uninterrupted sequence. If you have to stop before the bowl is hollow, wax the wood to keep it from drying. I have had unwaxed pieces start checking in minutes in a heated shop.

such experimentation should be a never-ending process. I have arrested checking by placing bowls in paper bags for a few weeks to choke off air circulation. Once you have an idea

Generally, the slower the drying the less risk of severe

of the principles involved there are endless ways to deal with problems.
To determine when the bowls are at equilibrium with the

relative humidity and temperature of the surrounding air, weigh them periodically. When they stop losing weight they are dry. Under average conditions, most native woods roughturned to a thickness of 1 inch will dry in about three months. I should mention an alternative to the drying procedures I

use. The green bowl can be soaked in a heated solution of

PEG (polyethylene glycol 1000) before drying. The chemical replaces the water in the cells and prevents them from shrinking. I experimented with PEG a few years ago and was not satisfied with the results. The slight differences in appearance and finishing qualities mentioned by PEG's proponents were real differences to me. Also, I was having success with natural drying and saw no need to continue with PEG. It can be useful, however, because with it you can turn bowls that include the pith of the tree. I know one professional turner who's satisfied with the results and I'm sure there are more. For further information, contact the Forest Products Laboratory and Crane Creek Company, both in Madison, Wisconsin. When your bowl is dry it can be finish-turned. First, plane the bottom flat. Before mounting, drill a hole to mark the finished depth. This will prevent turning through to the screws, which penetrate about 3 / 1 6 inch. I usually am able to

use the same screw holes as in the rough turning, and I use
the same length screw. Mount the bowl on the lathe and check to see that it clears the rest and the bed. I true the outside first, with the lathe at low speed. I usually use a gouge but I found light cuts with a small roundnose scraper ideal for the lignum vitae, which is very hard when dry. A larger tool might have taken too big a bite and forced the bowl off the screws. I finished off the outside shape with a skew scraper. At this point I usually sand the outside of the bowl. I turn most bowls relatively thin and when I am done hollowing, the walls vibrate. It's much easier to sand before hollowing, with little vibration. I start with 50 or 80 grit and work my way up to 220. I always wear a mask because the fine dust can be
quite harmful.

Now I clean up the rim of the bowl with a gouge. Next I get the inside rim true and work my way down to the bottom, using a gouge and scraper. I advise against using the scraper on the sides of deep bowls because it can really make a mess of end grain. When I'm satisfied with the contours and

thickness measuring with calipers I sand the inside of

the bowl using the same grit sequence as on the outside. The bowl can now be hand-sanded, if desired, to remove circular
scratches. Finish as you like.

The above procedures are only guidelines and can be adapted for almost any wood you'd care to turn. Exact methods of turning and drying should be worked out individually in one's own particular situation. I've had some
failures and will have more in the future, but I've had a high

rate of success. It is very satisfying to make a bowl when you control the whole process from log to finished form.

Before turning the back of a bowl cut from a green log (top photo), try to make it as round as possible. With the back

turned, the faceplate is then attached to the foot, and the bowl is rough-turned. Then the whole bowl is liberally coated with paste wax (bottom photo) to control drying. When dry, the bowl is remounted and finish-turned.

Queen Anne
Styling elements in table designs
by Franklin H. Gottshall

The Queen Anne style is generally the most popular of all the good English styles of the 18th century and is a good choice for craftsmen wishing to put together their own period design. The style's popularity is due to the fact that in the beginning it was distinguished for its clean lines, beautiful curved elements and restraint in the use of ornament. Queen Anne's short reign (1702-1714) was not distinguished for any personal influence she gave to the progress of fashion in her day, and so it must be assumed that the craftsmen themselves were largely responsible for the changes and improvements in furniture design during her reign. The happy result was that craftsmen, who understood both the practical possibilities as well as the limitations of their craft, were largely free of the domination by patrons whose wealth and position did not necessarily reflect good taste. Thus, at least in its early stages, the style was relatively free of the excesses in form, embellishment and elaboration so prevalent immediately preceding this style, and in those which followed. Cupboards, cabinets, chairs and tables became less elaborate and fussy, and were designed with a view to their function rather than to ostentation and display. Technical

improvements in both design and construction were made with pleasing results. The changes brought about by these factors, as well as an improvement of the economy in England during this period, made it possible for more people to share in the amenities which had previously been largely reserved for the privileged few. While the Queen Anne style came into being during the very beginning of the 18th century, its influence, once it was well established, continued well into the latter part of the century. Artists like William Hogarth greatly influenced design at this time, especially the employment of the reverse curve, both structurally and decoratively. Also sometimes called the cyma curve, it is used consistently and with good effect in Queen Anne style. On a portrait of himself, which now hangs in the National Gallery of London, Hogarth painted a palette on which appears a reverse curve with the caption "The line of beauty and grace." This aroused so much discussion that an explanation was demanded of him. He explained it by saying that "a beautiful curve by its serpentine, flamelike waving and winding simultaneously in different directions leads the eye in a pleasing manner from one end to the other." He sought to explain it further by saying that the principles involved were "fitness, variety, uniformity, simplicity, intricacy, and quantity all of which cooperate in the production of beauty, mutually correcting and restraining each other occasionally. In addition to this, he portrayed Queen Anne furniture in many of his paintings, which enjoyed wide distribution during the first half of the 18th century. The American colonies not only imported a great deal of furniture, once trade was well established, but they also made reproductions and adapted the styles to their own requirements. Fortunately, in the majority of cases, their adaptions reflected the simple, clean-cut lines and attributes by which we identify the style in America today.

Some styles of legs and feet found on Queen Anne furniture. Among the most widely used were the trifid (three-toed) webbed foot (second from left, also shown in cross section) and the pad foot (third from left). The ball-and-claw (second from right, below) later became a Chippendale hallmark. The Spanish foot is shown at right (above.).

Some Queen Anne Styling Elements

At present, good American Queen Anne furniture is more highly prized by American collectors of antiques than its

English counterparts. Among the reasons for this are that

skillful cabinetmakers like William Savery and others did
notable work in the style, basing their work on the early,

simple, clean-cut patterns imported from Europe. They used walnut rather consistently because it was available and plentiful in the areas where they worked. Maple, a wood not native

to the mother country, was also used; so was poplar as a secondary wood for drawer sides and like members. About 1720 and thereafter there was a gradual substitution from walnut to mahogany in England, but this change did
not take place in America to any great extent until a long time later, because mahogany was more expensive and no great improvement over the native walnut. One of the most appealing developments of the Queen Anne style was the small dressing table or "lowboy," sonamed to distinguish it from the "highboy," a similar piece with a chest of drawers on top. Lowboys are about table height, rarely exceeding 30 inches. (An antique purporting to be a lowboy which is much taller, or wider than the example shown, is probably a converted highboy and worth a lot less.) As the Queen Anne style metamorphosed into Chippendale, ornament became more and more elaborate, often featuring quarter columns and other refinements. But the best (and most highly prized) furniture of the Queen Anne style is characterized by the minimal decoration, simple outlines, beautifully formed curves and sound, sturdy

construction of the early period. The modern craftsman

would do well to adhere to these principles.

Bad Queen Anne

(taken from actual antiques)

Stacked Plywood
A fluid alternative to hardwood
by Ellen Swartz
Author with Surform Making plywood furniture and understanding that process as a possible vehicle of social consciousness are primary concerns of mine. The material I use is secondary, although many people are struck by it first, foremost and finally. In 1970 I was searching for material to use in my simple shop and "found" plywood. It was easily available, inexpensive, strong when laminated, and could be worked with the basic tools I had. The more I used it, the more it appealed to me. I use it so there is very little waste (political appeal); it isn't rare or precious so my prices can reflect that (social appeal); I can scrounge plywood at construction sites for smaller works (economic appeal); and the simple and direct methods of construction allow me to concentrate more on concepts than technique (personal appeal; i.e., knowing your strong points and limitations). All said, we seem suited for each other. The technique of making plywood by fastening thin strips of wood together and alternating the direction of the grain to give greater strength has been known since the time of the ancient Egyptians. At the Step Pyramid at Saqqara (2700 B.C.), plywood was found in which six 1/4-inch layers with alternating grain directions were fastened with wooden pegs. However, it was the demand for plywood for building ships and planes for the First and Second World Wars that brought about the product we know today. Until that time, the impermanency of casein, animal and vegetable starch glues retarded the development of plywood. The improved casein glues of the 1920s and 30s were used in airplane construction, but repeated cycles of wetting and drying caused eventual deterioration. The development of resin glues in 1935 made available an adhesive which is waterproof and immune to bacterial attack. It is unaffected by wet-dry cycles or intense heat and makes a permanent weld. In plywood glued with phenol resins, the glue lines are stronger than the wood. From an engineering standpoint, the advantages of plywood are impressive. It is extremely strong in comparison to its weight. It is durable, and it is permanently cured so there is no checking or splitting as with solid wood. There are many ways of putting plywood together to make furniture. For the past 40 years, molded plywood has been well developed by industrial furniture designers. Veneercovered plywood is widely used in panel and carcase construction and in cabinetmaking. But plywood can also be glued together in layers, building up to almost any form by laminating thin, cross-sectional slices. A stack of "side-view slices" can be face-glued together to block out the form of a chair that is then shaped and finished. Pieces of ply can be glued edge-to-face to make a right-angle joint, as long as the joining surface is large enough. If the area of contact is small I use screws or lag bolts for strength, as where the arms of a chair join the back. Sections can be hinged together. Laminated plywood lends itself especially to fluid, bending shapes because no joints are needed at the bends. As long as the material is thick enough for the stress it has to take, it will be extremely strong. Plywood isn't suitable for traditional joinery, although pinned finger joints, as at the back of a chair rocker, work

High chair, coffee table, side chair: plywood's horizons open wide when the designer's mind escapes from the 4x8 sheet.

well. They need to be rather large because there are no complete long-grain to long-grain glue surfaces. I make each section three layers of plywood thick and stagger the layers to
form the finger joint itself. It's very strong. To make a transition from horizontal to vertical, as in a table pedestal, I start by gluing the three vertical layers that form the column itself.

Then I measure up four or five

layers and glue a vertical lozenge-shaped piece to each side to form the transition, with its bottom edge square to the side
of the column. I usually shape

the column and the transition piece next, because there is still room to work, then I cut the base plate and screw it to the bottom of the column. Last, I cut the four or five layers that will fit between the transition piece and the base plate and glue them in place. If the measuring didn't work out and they're too thick, I plane off a layer of ply. If they're too thin,

Chair begins with four side-view slices, clamped together without

glue at the center.

I add a piece of veneer. Then I carve away the stair steps and shape the whole base. In terms of design, there are several ways I proceed. If an
idea is quite clear, I make a small, three-dimensional and

side-view sketch. Often I will ask the person I am working for whether he has any ideas to contribute, either verbally or through a drawing of his own. Other times I proceed with a
rather vague idea and no sketch. This approach often evolves forms I'm sure I wouldn't ordinarily think up. Once I determine the basic form I go directly to the plywood and draw up the first slice. This replaces the full-scale drawings many furniture makers use. The first and successive slices are cut out with a saber saw or band saw. The slices may be assembled from smaller pieces as long as the joints don't coincide from layer to layer. Usually I glue several slices at a time, sometimes as many as eight or ten. I use adjustable bar clamps and cee-clamps, placing one every 6 to 8 inches. I use
Titebond yellow glue, which dries quickly, but 24 hours is

When side has been shaped and four layers added to start the seat, gouge and Surform remove the stair-steps. A wedge-shaped section will join the two halves, with dowels for reinforcement because the seat is supported only at the front.

still not too long to wait if much stress is going to be put on the joints. For rough shaping I use a gouge and mallet, or a small electric chain saw. Then on larger pieces I use a high-speed

sanderan auto body grinder, reallyto take out the chain-

Sinuous bend of three

layers of plywood is made first; it is very strong and

Two small lozenges are glued and shaped for the

transition atop the base,

needs no reinforcement.

then the bottom plate is screwed on, and the halflayers of the base itself fit
in between. Similarly the

tabletop, except the surface piece hides the screws

into the upright section.

saw marks. But I use the Surform and rasps to give all the shapes, curves and transitions their final form. Next I file to

remove Surform marks. I fill checks, voids and splits in the

plywood with oak-colored plastic wood and sand. I usually start with 30 grit, then finish with 50 or 60 grit. Sometimes I sand as fine as 80, but rarely. Because of the striations, plywood doesn't have to be sanded as finely as solid wood; because plywood doesn't expand and contract like solid, the plastic wood won't fall out as long as a sealing finish is used. I generally apply six coats of urethane varnish, sanding the

second and fourth coats with worn 80-grit paper. The number
of coats and the sanding builds a smooth finish on the roughsanded ply. The major difficulty in working with plywood is the chipping out caused by working your tools across layers going in different directions. Abrasive tools are better than sharpedged ones for most operations because they cause less chipout. And the abrasive nature of glue lines quickly dulls a

sharp edge. Otherwise laminating plywood is a direct and

easy method and a lot of technical background is not necessary before beginning.

Rocking Chair with Message

Ellen Swartz made this rocking chair last year, when she was
invited to show at a gallery of contemporary craft. Before applying the varnish, she wrote a message on the back:

et, competitiveness pits people against people and people against nature. A reordering of what we value which would place respect, equality and unity above divisiveness, could be

In this society our responses to our work and to life in general are often conditioned by the cultural legacy of "rugged individualism," i.e. competitiveness. As a basic cultural ten-

a step toward re-establishing balance within society and with-

philosophers and doers, have a responsibility of evaluating their work and its relationship to the social, political and economic needs of society. If we remain on our ascribed pedesJewelry box is made of scrap. Cantilevered seat design of finished chair pushes material to its limits.
tals and only feel we need to relate to the world of aesthetics, we remain part of the problem, not part of the solution.
Many viewers stopped cold, wondering why she had ruined
the piece by writing all over it. But several inquired about the chair, and an art teacher eventually bought itas much for the sentiment as for the chair itself. Swartz usually writes on the furniture she makes for invitational shows, aiming the

in nature. I feel craftsmen/artists, as well as social thinkers,

messages at the audience such shows attract. "If I'm going to enter art shows," she says, "I want those people to listen to what I say, to react to it, even to say why did she ruin it."
Swartz, 31, is no raving revolutionary, but she believes American society needs some changes. She says that in traditional societies the artisan is an integral and necessary

part of the culture, making the functional things everybody

needs. In industrial society, mass production makes those

things, often not very well, and only the rich can afford craftsmanship. As a result, she contends, the work becomes
more sophisticated, elegant and technically elaboratebut also more hollow because its esthetic is "art for art's sake."

Thus she advocates using found materials and common construction plywood as a reaction against elegant, expensive
rosewood or walnut. Her shop is simple; her only machine is a band saw, although she is a trained cabinetmaker.

The point of all this is to integrate her work as a craftsman

with her life in the city. Her prices reflect the same ideas. When not bartering her work, she will build on commission
for the same hourly rate that the buyer earns.
"I'd like to see craftsmen consider other ways of working," she says. "I'll try to use my work for good things, not all self-

gain. Sure, I'm hooked into the system too, but it's a matter of degree. It's not to reject everything wholesale; it's small re-evaluations that may lead to bigger things." J.K.

Antiqued Pine Furniture

Distressing won't hide sloppy work
by B. D. Bittinger

"Probably it is the spirit of frank simplicity that gives this work (pine furniture) its fundamental appeal. It is on friendly terms with open fires, with wrought-iron hinges, with hewn beams and cornerposts. . . But 'ware lest you introduce a piece of mahogany to such company! The mahogany raises its eyebrows at favorite scratches and rounded edges of the pine, while the pine peeks out of the corners of its eyes at the painstaking satinwood inlay and wonders what it is all about."
The Pine Furniture of Early New England

Russell H. Kettell, 1929 Antiqued pine furniture has become increasingly popular in recent years. This style of furniture is characterized by thick (1-in. to 1-1/2-inch) table and case tops, and correspondingly sturdy carcase construction. It is constructed from knotty white pine. The antiqued and distressed finish is medium dark brown, with lighter brown highlights. The line of commercial antiqued pine furniture marketed under the Ethan Allen trademark is a good example of this furniture style. I view the style as a romanticized version of the pine furniture built by skilled joiners in rural America during the last half of the 18th centurynot rustic or common, but well-made country pine furniture. Antiqued pine furniture designs for use in present-day homes are necessarily adaptations, not authentic reproductions. After all, rural colonial families did not have king-size beds with "sleep sets" or stereo and TV cabinets. Freestanding desks were rare and all-drawer chests with large, plate-glass mirrors were unknown. They had as much use for coffee tables and bookcases as we have for dough boxes and flax wheels. Antiqued pine furniture designs are based more on feeling than on fact. Eastern white pine was used by early cabinetmakers and knotty eastern white pine can still be found. Western white pine serves as well and is available at most lumberyards. There

are several varieties of western pine, ranging in color from almost white to tan to pink. Kiln-dried knotty white pine is

made from 3/4-in. stock with a 1/2-in. thick panel is light in

generally available as #3 common 1x12 shelving in lengths of

up to 16 ft. Sugar pine is my choice when I can get it. It is usually tan in color with rust grain lines. Some boards also have a distinctive brown fleck marking. The knots are generally small and red-brown in color.

weight and relatively weak, it is satisfactory for this application because it will be amply reinforced by the back drawer dividers and facings. The frame is assembled with mortise and

tenon joints and the panel is retained in grooves in the stiles and muntins. The front edge stile should be 3/4 in. narrower than the back stile so that after the facing is applied the stiles
will be the same width.

Most furniture parts are less than 4 ft. long and 5 in. wide. In selecting material, I look at a 1x12 shelving board in terms
of the number of good pieces it may contain, not its overall

appearance. Some white pine boards are full of sap, heavy

and sticky, and are useless for making furniture. Thick (4/4, 6/4, 8/4) knotty white pine is not available at most lumberyards so I usually order from a dealer such as Educational Lumber Co. in Asheville, N.C. White pine is weak compared to most cabinet woods and
this must be considered when designing joints. Tenons in

pine, for example, should be as large as possible and somewhat longer than they would be in hardwood. Chair
turnings should be hardwood, but heavy pine seats and arms may be used in combination with the hardwood. Pine turnings for table legs should be heavy with simple, bold

After the end panels (frame-and-panel or solid) are cut to size, a series of 3/4-in. wide by 1/4-in. deep dadoes is laid out and cut to house the drawer dividers. Remember to cut a rabbet 3 / 8 in. wide by 1/4 in. deep on the back edges of the
carcase sides to accommodate the back panel of the case. Drawer divider units are made from 3/4-in. pine joined with mortise and tenon or half-lap joints. The width of the divider strips will vary according to the overall dimensions of

patterns. I prefer to glue up turning blocks from 3/4-in. stock. Small, firm knots in a turning block will usually cut and finish well.

Tools must be extremely sharp for cutting pine because it is so soft that the fibers tend to tear. I use a plywood-tooth saw blade on the radial arm saw and always crosscut with the good
side up. Even with a sharp, small-tooth blade the fibers may

the case. The strips at the ends of the divider frames, which run from front to back, should be about 1-1/2 in. wider than

the vertical facing strips of the front frame, so that they can
support and act as a bearing surface for the drawer sides.

break out on the bottom and leave a rough surface. Carving chisels must also be extra sharp. Let me emphasize that workmanship must be of the
highest quality. Antiquing and distressing will not cover or hide sloppy or careless work. Quite the opposite, antiquing will emphasize poor joints, hammer marks, clamp marks, and other evidence of careless workmanship.

The carcase is assembled with glueon the front edge only of solid end piecesand plug-covered, flat-head wood screws through the end panels. Plugs may be surface-cut round plugs, end-grain round plugs or square patch plugs. Finishing nails, set below the surface, may also be used for
case assembly. After setting the nail I use a modified nail set with a square tip (about 1/8 in. by 1/8 in.) to make a square set hole. The small round or square holes will blend with the overall distressed appearance. The facings and back panel will
be glued and fastened to the edges of the end panels and to the drawer dividers, to provide adequate strength whether the case is assembled with nails or screws.

Carcase construction
A carcase for a large chest of drawers includes most of the

particular problems of working with white pine. The large drawing shows the basic construction of such a case. What
follows are the working methods and finishing techniques I

have developed for achieving the style I like.

Carcase end pieces (using 3/4-in. stock) may be solid edgeglued pine or frame-and-panel assemblies. Doweling or
shaper-edge joining is not necessary with Titebond glue. It is important, however, to align the boards in the clamps, thus minimizing planing and sanding the finished panel. Wide

The front facing frames may be assembled as a unit with dowels, mortise and tenon joints or half-lap dovetails and then fastened to the carcase; or each strip may be individually attached with butt joints. I prefer the latter. In either case, the facing is fastened with glue and plug-covered screws or
finishing nails. If nails are used, they should be located at

random on the facing boards to avoid a regular pattern.

Plugged screws, however, should be placed in a symmetrical

shelving (12 in. or more) should be ripped into at least two

strips and the grain alternated before gluing, to reduce cupping. Excess glue squeezed out of the joints should be removed at once with a wet cloth. Glue will seal the wood and cause light spots when the stain is applied. If solid ends are used, special consideration must be given in assembly to avoid restraining the boards across their width,
else the case will be damaged by shrinkage or expansion. Raised-panel case ends are visually pleasing and solve the problems of expansion and contraction. Although a frame

pattern. Whether or not the facings were preassembled with mortise and tenon joints, 1/4-in. dowels may be set into the surface to simulate draw-bore locking pins. I usually make the base boards about 1-1/2 in. thick to balance the thick top overhang. They may be scroll cut or left
full width. The base is assembled with mitered corners and is glued and fastened to trie carcase. On some pieces, such as dower chests, the base boards may look better if they are
joined with through dovetails. The base mold is a modified stock molding. The bottom dust panel and back (lauan plywood) should

Bill Bittinger, 48, has been a woodworker for 20 years. Trained as an engineer, he is production superintendent at a tire cord factory. He lives in Shelbyville, Tenn.

not be attached until the drawer slides are installed and each drawer is accurately fitted to its opening. Case tops vary in thickness from 7 / 8 in. to 1-1/2 in.,

depending 00 the scale of the piece and the width of the

carcase facing strips. Tops are made from solid edge-glued pine. I use dowels on edge-glued thick pine to help level the pieces in the clamps. If bread-board strips are not used, the

opening. Knots should be at least 1 in. from each end. Cut the sides about 1/8 in. narrower than the front and

about 2 in. shorter than the case depth. Drawer sides should
be about 5 / 8 in. thick. Cut 1/4-in. by 3/8-in. bottom retaining grooves in the sides and front, and cut a mortise on the bottom edge of the back side of the front for the drawer slide part. Dovetail joints should be cut by hand. It is difficult to

end grain of the top should be carefully sanded or it will soak

up extra stain. Case tops are attached with screws from underside in oversize holes in the top mounting strips. Bread-board strips across each end of the top add to the appearance and strength of projects such as coffee tables and heavy trestle tables. Before joining the strips, I run either a 1/8-in. by 1/8-in. rabbet across the ends of the top or a quirk mold on the joining edge of the strip. Narrow strips may be attached with screws through oversize holes, to allow for expansion. The counterbored screw holes are covered with round or square plugs. The edging strip should be glued only at the center on a tabletop, or at the front edge on a case top. To attach unsupported strips to tabletops that are subjected to heavy loads, I use 2-in. wide stopped splines made of 1/2-in. plywood and two or more concealed tie bolts cut from 1/4-in. threaded rod. The spline is glued only at the center, to equalize misalignment caused by shrinkage or expansion of the top.

obtain sharp clear lines in soft pine with an ordinary marking

gauge. I use a sharp 3H pencil with a shop-made marking

gauge to lay out the dovetails. Dimension A is usually about

two-thirds the thickness of the drawer front and dimension B is 1/32 in. greater than the thickness of the side pieces. Side A is used as a marking gauge on the end of the front and on the matching side piece, to provide cutting lines for the length of the dovetails. Side B is used to mark the dovetail depth line (side thickness plus 1/32 in.) on the inside end of the front piece. I use another shop-made marking jig to lay out the halfblind dovetail on each drawer front end. The marker is used to draw the pins and the vertical-cut guide lines on the inside of the front.

Drawers
I don't think pine drawers should be lipped because a thin lip is fragile and a heavy lip is clumsy. I install drawers and doors with 1/8 in. of the edge exposed. When they are rounded by sanding, the chest has a softnot flatappearance. Cut the 3/4-in. drawer fronts for a snug press fit in the

openings, and index-mark each front to its corresponding

I start marking with a half-pin on each end, so that a tail will cover the bottom retaining groove in the side, and then fill in between by eye. Each end of the drawer front has the

same number of pins and sockets but they are not all identical
in size or spacing. Pins in pine should be only slightly smaller than the dovetail sockets. After sawing on the waste side of each pin line, I clamp the front to the bench to remove the waste with a very sharp chisel to within 1/16 in. of the guide lines. I trim the pins and the sockets to exact size with the front held vertically in a vise. Then I lay the side piece on the table saw and stand the front piece on top of it, using the saw fence to hold the front piece vertical. With the ends of the two pieces exactly aligned, I can use a sharp pencil to trace the

pin outlines onto the side piece. I cut the tails to size on the

band saw. There should be enough interference in the joint to require moderate pressure to assemble. I do not dry-fit dovetail joints because this compresses the pine and causes a
weaker joint. The dovetails are assembled with glue. The back piece is joined to the drawer sides with through dovetails. The bottom (lauan or pine plywood) is nailed to the lower edge of the back part with coated box nails. The

drawer slide is attached with brads and glue. Beveled glue

blocks keep the bottom from rattling. When the glue in the assembled drawer joints has set, I sand the protruding ends of the front with a belt sander. This

sanding, if carefully done, will leave about I/32-in. clearance

at both sides when the drawer is installed.
Install slides in the case and adjust them to center the

drawer fronts in their openings. Sand to round off the top

edge of the drawer front, to provide about I/32-in clearance.

A gap of up to 1/16 in. around the drawer is acceptable. Attach the 1/4-in. plywood back panel with glue and small coated box nails to the drawer frame edges and all around the Large crown molding may be assembled from several parts. The three-part composite molding shown below is made up

case edges for maximum strength.

Doors
Raised panel doors are included in some case designs such as a hutch base. Doors may be made from 1-in. or 3/4-in. pine stock. But conventional frame construction for a raised

from: 1) shop-made dentil backup piece, 2) modified commercial flat-crown molding, and 3) a nose-molded pine strip.
The dentil backup piece is machined to leave a 1/4-in. to 3/16-in. raised strip. The dentil is laid out and cut after the backup strip is beveled to fit the cabinet. The " t e e t h " should be laid out from the center to ensure symmetry, taking care to

panel door of 3/4-in. pine is structurally weak, and the door

does not feel comfortably heavy. To avoid weakness and add weight, I set the panel in a rabbet instead of a groove and add a back plate of plywood. First, I assemble the frames (rails,

locate one full tooth on each side of the bevel joint. The pieces of crown molding and the pine strips are mitered and
installed in turn over the backup strip.

stiles and muntins) with dowels and glue. Then I rout a

rabbet 3 / 8 in. deep by 1/2 in. wide on the inside back edges of the frame, squaring up the rabbet corners with a chisel. Then I cut the panels from 3/4-in. solid pine, allowing 1/8 in. for clearance on all four sides. I "raise" the panels by

cutting a bevel all around the front face, so that when the
panel is laid into the frame it will protrude by the merest 1/64 in. at the back. Finally I glue a piece of 3/16-in.,
1/4-in. or 5/16-in. pine or lauan plywood to the back of the

frame only, thereby pressing the panel into the rabbet and completely covering the door.

Some projects are enhanced by a closed dentil molding (a row of square depressions) along the top horizontal facing

strip. This decoration is particularly effective in bedroom sets

where it can be incorporated in chests, mirror frames, bedside tables, headboards and footboards. To make the closed dental facing strip, cut the facing about 1/2 in. wider and longer than finished size. Rip out a strip

from the facing board where the molding is to be located.

Start the dado layout in the center of the strip to ensure symmetry, and plan the cuts so that a raised tooth, at least full
width, will remain on the end of the strip at the facing joint.

Moldings One of the advantages of working with pine lumber is the wide range of commercial moldings. Before mitering and applying commercial flat molding, I glue a pine backing strip to the molding and then resaw to provide a larger glue surface. This is particularly important when attaching large cornice molding around the top of a cabinet. Resawing to 45 also helps in cutting miters on large moldings.

Cut and sand the dados, joint the glue edges and edge-glue

the strip back in place. Distressing

The piece should not be distressed until after complete assembly and coarse sanding. Experimentation is the only way

to determine the amount of distressing that will suit your taste. The procedure includes surface marking and removal of

material to create a worn appearance, and special finishing to make the marks look authentic. I do not like excessive surface marking. I usually make a few dents with the corner of a hammer claw and a few randomly spaced holes with an awl. The claw indentations are triangular

and the awl holes appear as small black dots when the
finishing glaze is applied. Distress marks are always randomly

spaced and are more numerous around the bottom and on the
top of a piece than on vertical surfaces. A tall bookcase, for example, would have very few indented distress marks above

the height of 4 ft.

Wear distressing should correspond to the imagined, as well as the actual, use of the piece. Stretchers and legs that are rubbed with shoe soles should be much more severely worn than areas touched by other parts of the body. Wear distressing may be done with planes, coarse sandpaper or even
rasps and files. In any case, the worn surfaces should be carefully sanded so that no tool marks remain.

After the piece is distressed, I sand with a high-speed orbital sander using 100-grit, 120-grit and 150-grit garnet paper. I complete the sanding by hand with 180-grit garnet paper wrapped around a felt pad. If scratches or other unplanned surface defects show up at this stage, they should be removed by going back to a coarser grit paper. Distressing

Tall chest: Designs are based on feeling more than on fact.

Tique deeptone antiquing glaze, made by C.H. Tripp Co.

But I suggest you make up sample blocks to determine your

preference. Rub or brush the glaze on the outside surfaces of the project and take care to fill all the distress marks with glaze. Wipe off when the glaze begins to appear dull. The glaze changes the color of the finish even though most of it is

does not camouflage sloppy and clumsy work or incomplete

finish sanding.

Finishing
I use Minwax Early American oil stain to antique pine furniture. Apply the first stain coat to all surfaces, inside and
outside, following the manufacturer's directions. After 24 hours, apply a second coat to the outside surfaces. This leaves drawer and case interiors lighter in color than the exposed
surfaces. At this point in the finishing schedule your beautiful piece of furniture will look very disappointingdull and splotchybut do not despair.

wiped off. Leave a film but not streaks. Wipe in the corners
and at surface intersections with a wadded cloth so that some of the glaze remains. If too much glaze is wiped off, you can recoat and start over. At this point the finish on your project will look very good and it will improve with the final steps. After the glaze has dried for 24 hours apply the second coat of varnish. Allow several days for drying and rub again with
used #320 emery cloth and 2-0 steel wool. The final step is to coat the entire piece with a good grade of paste wax. The wax should fill any nail set holes. Rub and polish.

I spray McCloskey Eggshell or semigloss Heirloom Finish for the seal coat and the final varnish coat. Glaze solvents do

not soften it and it has good rubbing (sanding) qualities. It

can also be brushed on. After the seal coat is dry (depending on the climate, this may take up to a week) sand all surfaces with wet/dry #320 paper. When the varnish is properly dry, sanding will form a white powder on the surface. There are a number of antiquing glazes on the market and

If you want an antiqued painted finish, substitute the paint of your choice for the stain and first coat of varnish and

proceed as described above. I like the clean, bright appearance of painted interiors on pieces such as hutch bases and dry sinks. Light blue paint goes well with the antiqued finish.
I usually use antiqued brass drop bail or white porcelain pulls and mortised antiqued brass hinges. Black-finished H or L hinges and hardware are also suitable for some pieces. These comments on design, construction and finishing also apply to small decorative pine projects such as spice cabinets, letter boxes, spoon racks, stools and picture frames.

James M. O'Neill, in his book Early American Furniture,

gives a formula for mixing an antiquing glaze. I use Tone 'n

Solar Kiln
Boards emerge bright, check-free
by Paul J. Bois
Curtis L.Johnson of Madison, Wis., has designed and built a sun-powered kiln for drying small quantities of furnituregrade lumber. His kiln is slower than a steam kiln (Fine Woodworking, Spring '77), but it is gentler. The boards emerge bright and clear, with virtually no checking. The kiln is an insulated, stud-framed wooden box that measures 10 ft. wide and 12 ft. long at floor level. It holds 800 bd. ft. of lumber in two stickered piles. The south wall slopes 40 from the vertical and is enclosed by storm windows of single-strength glass facing solar collector panels. Each collector consists of a flat box whose front is sheet metal painted flat black and whose back is a sheet of hardboard. An air space of 1-1/2 in. separates the glass from the collectors. The bottom of each collector is vented and the top is open. Johnson piles the lumber in equal stacks, well stickered,

Solar panels on Johnson's kiln are tilted to catch early morning and late afternoon rays. Small door gives access to rear of solar collector, large door to stickered lumber. leaving a central plenum 18 in. wide. Two overhead fans, spaced evenly atop the 8-ft. load, each deliver 1,200 cubic feet of air per minute into the plenum. Hinged baffles hanging from the fans ensure that all the air is driven down between the two stacks and out between the stickered boards. A thermostat turns on the fans when the interior temperature reaches 80 F. Two floor vents, each 4 in. by 10 in., admit outside air to the central plenum. The vents are screened against rodents and can be gradually closed as drying progresses, to keep the relative humidity as low as possible. The sun heats the kiln as high as 130 F in summer and 90 F in the Wisconsin winter. The temperature climbs as the moisture content of the lumber drops below 20%, because evaporation is slower and its cooling effect is less. To monitor the drying, Johnson cuts a sample from a representative board, weighs it, dries it in an oven and reweighs it. This allows him to calculate the initial moisture content, and he repeats the process throughout the run. Drying time depends on species, initial moisture content, thickness, season and latitude. Madison's latitude is 43 north. On July 15, 1975, Johnson loaded a green stack of 4/4 cherry at 60% MC and a stack of 4/4 white oak that had been air dried to 15% MC. Fifty-two days later, the cherry had dried to 6% MC and the oak to 7-1/2% MC. The following July 14, he loaded the kiln with 2x4 cottonwood studs at 130% MC and 4/4 black walnut at 85 % MC. After 47 days of excellent drying weather, moisture contents had reached an average of 10% and 12%, respectively. Johnson figures that about 80 summer days will dry most hardwood species from green to 8% MC or less. These figures assume 70 days of sunshine, or 400 hours of direct sunlight. Winter drying takes longer, from 150 to 200 days. A load of mixed green and air-dried stock will dry only at the rate of the green stock, but a full charge of lumber that has already been air dried to 15% MC will dry considerably faster.

This article is taken from Forest Products Utilization Technical Report No. 7, by Paul J. Bois, a wood drying specialist with the U. S. Forest Products Laboratory in Madison, Wis. The report is available from FPL.

All three Shaker round stands are made of cherry and finished in clear varnish. Stand at left (The Shaker Museum, Old Chatham, N. Y.) is subject of this article; center stand (Metropolitan Museum

of Art) has tapered Sheraton-style convex legs dovetailed into urnshaped pedestal; right, top of early 19th-century stand (The American Museum in Britain, Bath, England) is 16 in. in diameter.

Shaker Round Stand

Three legs are dovetailed to turned pedestal
by John Kassay

Round stand is a name given by Shakers to the pedestal, tripod candle stand. These stands were derived from the English Queen Anne and Sheraton styles. American cabinetmakers made them less ornate and Shaker cabinetmakers further simplified the stands in line and form, in accordance with their religious doctrines. The stands shown here were all made in the first half of the 19th century, when the Shaker sect was at its peak. They were used to furnish the retiring rooms of the Shaker's communal dwellings, which often housed 100 or more people. Today, these graceful stands make lovely accent tables. To duplicate the one shown, you need about 7 board feet of cherry. Construction of the stand should start with the legs (A in measured drawing). Draw a full-size pattern of the leg, including the dovetail pin, on a piece of thin cardboard and carefully cut it out. Mill the stock for the legs to the required 3/4-in. thickness and position the pattern on the wood with the grain of the leg in the longest direction, as in the front-view drawing. Trace the pattern three times and then separate the legs with rough cuts; make no attempt to cut to size. The ends of the blank that will form the dovetail and the bottom of the foot are now cut precisely to length and per-

John Kassay, 58, teaches cabinetmaking in the Department of Design and Industry at San Francisco State University.

pendicular to each other. Sandwich the legs together and firmly nail two scrap pieces of 1/4-in. plywood on these end surfaces. This assembly is now sawn to final shape and sanded as a single unit. Separate the legs, lay out the dovetail pins very precisely and cut the shoulders and sides with a dovetail saw. Of course they can be cut on the table saw, but that makes the operation less personal. Clean up with a chisel. To make the pedestal (B), turn a piece of cherry, 2 - 1 / 8 in. square and 20-3/4 in. long, to a 2-in. cylinder. Lay out the pedestal with its upper end at the tailstock. To make an exact reproduction of the original stand, mark pencil lines along the cylinder at 1-in. intervals to locate parting tool cuts at the diameters shown. The experienced turner, like the craftsman who made the original, will find it sufficient to part the smallest and largest diameters. Remove the excess wood with a gouge. Then the pedestal is rough and medium-sanded. A tenon, 1 in. in diameter and 11/16 in. long, is established on the upper end of the pedestal to hold the disc (C). On the original table, wood threads fasten the pedestal to the disc. If you have a 1-in. wood threading tap-and-die set, by all means use it. Otherwise, the tenon is simply glued into a hole in the disc. At the lower end of the pedestal, turn an absolutely straight cylinder, 1-7/8 in. in diameter and 3-1/2 in. long. This is where the legs will be dovetailed to the pedestal. Finish sanding the pedestal while it is in the lathe. Be sure to

dampen the wood to raise the grain, and sand along the pedestal with the lathe turned off in order to remove any circular scratches.
Choose the most attractive grain pattern on the pedestal and locate one leg here. The other legs will be located 120

right and left from this point (detail 2). Index the legs to the
pedestal with pairs of identification marks. Lines establishing the center of each leg arc drawn on the bottom of the pedestal

and extend along the sides. Then parallel lines 3 / 8 in. from
these center lines are drawn on each side. To enable the legs to seat properly against the pedestal, absolutely flat surfaces

must be cut within these outer lines. Cut a series of kerfs with a backsaw and pare away the excess wood with a sharp chisel.
Redraw the center lines on these flat surfaces, using the center

lines on the bottom as a guide. Draw lines 3 / 1 6 in. on each side of these center lines to indicate the thickness of the
dovetail pins, 3 / 8 i n . , at their shoulders. Hold each leg in

position against the bottom of the pedestal, and trace the outline of each dovetail pin (three lines for each) on the
bottom of the pedestal. These three lines, along with the two

parallel lines on the flats, determine the material to be removed to produce the dovetail sockets. With a brace and a
5 / 1 6 auger bit, bore a series of holes along the center line of

one leg. Count the number of brace revolutions to gauge the depth. The waste wood is removed with wood chisels.
Patience along with much trial fitting is now necessary. Make certain right from the start that the leg is being committed

perfectly straight; do not force the leg in place. These dovetail sockets may be produced using a router, a dovetail bit and a home-designed jig for supporting the router while the piece is still in the lathe (see p. 20). On the other hand, by the time
you have tooled up, you could be gluing up. After the legs

are fitted to the pedestal, they are disassembled and a spokeshave and abrasive paper are used to form the curves on the
upper and lower edges, as shown in section in detail 7. Plane stock for the circular disc (C) to the finished 3/4-in.

thickness. Mount it on a lathe faceplate and turn to the indicated diameter and chamfer. Finish-sand while the disc is on
the lathe and bore a 1-in. hole on center for the dowel end of

the pedestal. The disc may instead be bandsawn as described below. Bore and countersink four equally spaced holes for
flathead wood screws ( F ) , which fasten the disc to the top.

Fabricate the top (D) from three or four narrow boards that have compatible grain patterns. The circular shape and undercut chamfer can be turned outboard on a lathe, made with a hand router or a homemade circle-cutting jig on the band saw, or cut by hand with a bowsaw or a sabre saw.
Before assembly, the parts of the stand should be tough,

then medium-sanded. Make a jig of plywood to hold the legs and pedestal in an upright position while the glue hardens. Apply glue to both pins and sockets.
The disc is first screwed to the top with its grain 90 to the

grain of the top and then glued or threaded to the pedestal. Shakers often used a thin, sheet-iron plate (E) to reinforce
and cover the dovetail joint. If your dovetails have been well

made, the plate is not necessary. To finish the stand, raise the grain and fine-sand. Carefully
dust and apply a coat of clear furniture varnish. If a stain coat

is desired, it should be of a brown shade and is applied before the first varnish coat. Three or four varnish coats are necessary. The final coat is hand-rubbed with fine (4F) pumice or

rottenstone lubricated with lemon oil, and waxed.

The delicate interplay of glass and wood

by James Krenov

Showcase Cabinets

t is a puzzle to me why there are not more interesting showcase cabinets around. Certainly, living habits don't exclude this type of furniture. We do accumulate objects that are pleasing to behold and deserve a nice home of their own. Perhaps too many people have a preconceived, discouraging notion about showcase cabinets. Cabinetmakers may share such prejudices. Or the technical problems of doing doors with thin wood parts and closely fitted glass may discourage cabinetmakers. I suspect this is so. A way of getting past these problems is to use pretentious, special-effects glass and wild wood in all sorts of bubbly shapes. Interesting, although we may be missing opportunities by not taking advantage of the effects that simple glass set in pleasing facets can create. Work with glass and wood, if it is to succeed, demands great accuracy, patience and a way of conceiving and then doing a piece that is different from what some of us have been involved with. Someone says, "Showcase cabinets do not use enough wood!" This can be true. One is prompted (by mirrors and such) to forget, or at least to neglect, that this is in fact to be a cabinet, not an aquarium or a bar. At first, I, too, thought showcases were not truly cabinets. Then, because I liked the function of such pieces, I attempted to achieve some sort of balance between the wood and glass as related to the purpose of the piece. The real challenge is the pleasing interplay between you as craftsman and those who will use the showcase. You express something personalyour own version of a concept that is also a certain mood, while making something for someone to use and enjoy. Through usage the piece will achieve further expressions. Whatever mild interest I had in showcases from the outset has increased since then. It needn't always be so. Even a craftsman who tries to make such pieces with an open mind and a sense of the possibilities may conclude that it's not for him. Certainly, one of the things we should try to determine as craftsmen is the sort of work that is really for us. We are by our nature (the sum total of the traits we have or do not have) either finely tuned, meticulously inclined, or a bit of the opposite, the kind of people who do rough-cut, "unorganized" work. In the latter case this type of cabinetmaking is a frustration. Probably after trying we will then leave it. For those who discover something interesting here, I think such work leads to further discoveries and increased interest. Through one possibility you come to the next, and the next. That's the essential difference in our work between monotony or routine and this other thing, which really keeps it alive through the years. After all, I hope that some of us, in choosing our craft,

from resawn stock. Oil finish.

Showcase cabinet, Tasmanian blackwood (1977), 53 in. high, 30 in. wide (max.), 10 in. deep. The curved horizontal rails are laminated

are choosing a way to live and work and be happy doing it for a long, long time. Generally speaking, there are four basic solutions to showcase cabinets. Each poses problems and invites variations.
There is the flat, one-piece door or two flat doors in the same plane. The second solution is a V-shaped single door or two doors set at an angle to form a V. Another will be one or two

doors forming a convex curve. And last, a door or doors making a concave curve. It is a personal matter whichif anyhas a special appeal. The appeal will probably begin
with something visual: One likes the way a V appears, the way light plays on its glass, the idea of related angles and proportions. Or maybe one of the curved doors is more inviting.

Its softness, perhaps. At any rate, I believe the first thing that attracts us is visual, whether curve or flat or angled, glass and light as related to wood. After that, we must think in terms of
the work entailed in some particular solution we like. We probably should try from the beginning to choose something both possible and worth doing.

Besides the idea with the various demands we might first conceive, there are in showcase cabinets an enormous number
of details to be discovered; these we can play with and use. Some, of course, are directly anchored to the construction we decide on. Otherssuch as the various sensitive shadings we can use in connection with the doors, or the top and bottom

pieces, the thin strips with which we divide the glassare largely decorative. The center of this kind of work is aesthetic,
yes, but it is also the physical relation of wood to glass, the fits we need and how these relate to the various steps of the work. There is perhaps more organized method in making showcase cabinets than in any other kind of cabinetry.

Wall-hung cabinet of lemonwood (1966) has variation of the V-door: two smaller doors about 60 cm (23 in.) high.
ing the horizontal and vertical pans of a door in certain relations to one another, we can change the proportion of a given
shape or size without changing the size itself, and in so doing make a door appear to be wider or narrower; higher than it really is or not as high. There are options. Nor do the various

The sort and thickness of the glass is to be considered from the beginning. Once upon a time we had blown glass with a
greenish or brownish tint; it was alive, the real thing. This is now almost impossible to get. The modern imitations of antique glass do not appeal to me, so when I cannot obtain blown glass I use ordinary clear glass, about in. (just under

3 mm) thick. The average thickness of a door I make for a showcase cabinet that is, say, 16 in. by 30 in. or 24 in. by
30 in. is usually slightly under in. Simply put, flat doors are carefully chosen wood in pleasing proportions that belong to a well-balanced piece; they make a good first exercise. Curved doors are another and

parts of such a door need to be flush. There can be intentional yet slight differences in the thickness of some members, which will introduce little subtleties and
divide up the elements to give us proportions within

more involved matter. Usually one has to saw the laminates for the shaped parts and glue these up to the desired thickness on a mold. This is extra work, but it's worth it. I
wouldn't try to make curved doors cut from a thick plank of solid wood. First, it results in a lot of diagonal or wrong-way grain, which makes the cutting of rabbets or slots for glass dif-

proportions. These divisions are not for the sake of complexity, but because they are pleasant and give us variation
instead of a single impression, something for the eye

ficult and weakens the joints; and second, since it is very hard to predict exactly the visible pattern of wood on the pieces thus sawn, we're apt to end up with an imbalance. Once you have an idea for a cabinet and have considered the practical problems themselves, it is time to think also in terms of proportion. There are various possibilities with doors that may or may not be simple in their construction. By plac-

to play with. And it adds challenge in the work, since

such details are worthwhile
only when neatly done. With curved doors it is best to have the curved parts

Impractical Cabinetmaker, 1979 by Litton Educational

This article is excerpted from James Krenov's third book, The

themselvesthat is, the top and bottom pieces of the

doorextending out all the way. Thus the unbroken
Proportions: With the same outside measurements, a number of apparent variations are possible. Grain patterns will enhance or diminish the effect.

Publishing, Inc., to be published this fall by Van Nostrand Reinhold Company. His other books are A Cabinetmaker's
. Notebook and The Fine Art of Cabinetmaking

curve gives us a calm and complete sense of the shape

of the cabinet;

it is not

45

Thin wood layers glued to a curved form make the door rails. Start clamping at the center and work outward (top left). To make slots for the tenons, design a guide block for cutting slots that parallel the curve from a full-size drawing (top right). Using a spring clamp to tack the block to the wood, cut slots with a band saw or a table saw (bottom left, right).
"chopped off" by the vertical sides of the door. It is important to get this feeling of wholeness, to let the curve have its full intention. In such doors I find that having the sides, top and bottom flush on the outside as well as on the inside results in a calmness; it gives the mood that goes best with the soft curved intention of the piece. One should try to accentuate this feeling in the details of the various other parts. Think soft.. . . In its principal steps only, the work is apt to be as follows: 1) Concept of the piece, shape and size of door (or doors) related to whole. 2) If curves are part of the concept, mold for laminating. Saw and glue up laminates. Be generous with widths. 3) Vertical parts of door: Keep straight, true, and slightly thicker than laminated (horizontal) pieces. 4) Make slot and tenon joints as needed. 5) Dry-clamp, set up, then plan spacing of pins (vertical bars) to take various widths of glass. Replate these to position of shelves and whole cabinet as you'd like it. A bit of composing. . .take your time! 6) Lay out and make template (or templates) for routing rabbet in curved parts of doors. Do all this with care! Have scribe line along inside edge and stopcuts chiseled at ends. Rout a little at a time. 7) Assemble doors. Mark for rabbet in outside pieces and groove (or rabbet) in middle ones. Double-check margins as related to machining. Make stopcuts, scribe, and do all machine work, including bevel on inside of parts having groove. Before reassembling, plane the inward front edges on vertical door parts to a nicely rounded shape. 8) Assemble. See that rabbets or grooves meet as they should. Round edges along curved door fronts. 9) Make (shape) and fit vertical pins to hold separate pieces

later, they will be planed to the same curve (bottom). These small details make all the difference.

thicker than the laminated curves (top);

The vertical parts of the door are slightly

of glass. Cut them straight grain, and make lots of extras!

Think about glass fit.

10) Glue up doors. Prepare everything, study setup. Dryclamp first. (Neatness) 11) Clean all corners. Plane vertical parts of door to curve. Have plane iron razor sharp, fine set.
12) Do machinework for overlap fit of doors, angled to suit curve. Watch out! 13) Square doors. Check total width top and bottom when in right curvemake template or measuring stickand

proceed with work on cabinet case. (Lay out position of sides, how doors fit.) 14) Refit pins for glass, make rest of parts (to fit rabbets and
pins) that hold glass in place, drill, countersink. Fit to doors using pieces of plywood as thick as the glass, or the glass itself.) 15) Final fit of doors. Polish and finish all door parts

(handles?). Remove doors. Complete case, finish, make stand if needed. Go through all details before hanging
doors for keeps! The above list, or something like it, would be one of those

reminders I make for myself on a scrap of paper. Since the various procedures are closely interrelated, it will be difficult for me to give you an exactly parallel description to suit your projectwe should be together doing the work. As I have
been through it many times by now and am (almost) used to the zigzags, it is hard to foresee what your difficulties, if any,
might be. For the time being, you'll just have to ask and, I'm

sorry to say, try to answer, your own questions.

Choose the wood carefully. Relate the graphics and the color to the intention of the piece as a whole. The choice of wood can make or break not only a cabinet that is all wood, but also a showcase cabinet. Don't fool yourself into thinking that with these cabinets the choice of wood is any less impor-

tant than with others. Cut the various layers of laminate for
the door with care, keeping them in a visual relationship to the cabinet shapes. I make them 2 mm or 3 mm thick

ference between right and wrong in spacing these various

facets of glass is a matter of sensitivity, judgment and experience. It's an elusive thing. And neither I nor anyone else

(roughly

in.); on a door about 20 mm thick there are seven

can tell you when you've got it just right. You, yourself, have
to feel it. To "get it right," even in relation to ourselves, we should

or eight layers. No matter how cautiously we work, there will be deviations

along the way. Even at its best, our accuracy is not quite total.
The laminated door parts will emerge nearly, but not exactly, of an even thickness throughoutand they are curved. The

experiment with various curves and spacings. I tend to make

my curves more tense at the ends; that is, the door is slightly less curved along the middle and then tightens toward the

other pieces will need to be shaped (hollowed or rounded)

later on to form the whole smooth curve of the doors. Before doing this, however, we must make rabbets and perhaps

outsides. And after some experiment I have concluded that

the pins (the verticals dividing the glass) need to be closer together toward the outside of the door and farther apart at the

grooves in these vertical parts, and do so without going astray.

Our objective, you see, is to have all the details coincide at the corners; we don't want to chisel and chip and scrape try-

middle. How much and how little depends upon the curve
and, of course, how one feels about the shelves that are going to be in the cabinet. These make a horizontal division that is not obvious when you have only the door in front of you, so consider that there will be horizontal lines here and that they

ing to fit glass to wood. To provide a working margin here I

make the vertical parts of the door slightly thicker than the

laminated parts. This gives me the chance to make small adjustments before machining and to shape these pieces afterwards. Actually, the work is not as complicated as it may sound, though it is necessarily exacting. Once you get the
idea, you will discover a certain consistent logic about planning layout and methods. From then on, you can work

will "shorten" the door somewhatin the case of a curved

door, they will make it seem wider than it really is. I use thin strips of wood to simulate shelves and tape these onto the inside of the door between the horizontals. I can then move rhem up or down, observing how the "shelves" affect the door, until I arrive at something I believe will be fairly good.

without further accidentprovided you achieve the right

amount of concentration. Patience and care at every step are absolutely essential. If you are not set upon such an effort,

Someone says, "But I can use glass shelves, and then there
will be no negative effect." That is not always true, although sometimes glass shelves do work very well. Usually this is

perhaps it is better to avoid these cabinets altogether.

Having made the principal parts of our doorwhatever the shapeand cut the various joints that are needed, we fit

because the door and the rest of the cabinet do not need
definite horizontal lines to compensate for exaggerations or deficiencies of proportion.

them together dry. Now comes the most critical part, the rabbets in the top and bottom pieces. It is necessary to solve the
problem of rabbets in the horizontals before doing those in

The cabinet is somehow meant to have glass shelves. In

other cases, with a similar cabinet, shelves of glass may prove

the side pieces. If it is a curved door, we must make a very exact template of the curve with the various facets. But wait! To

to be all wrong; one will discover this when one changes from glass to wood here. This is a matter of experience, of judging
and observing, which also means experimenting.

do this we have to know the locations of the various sections

of glass; their widths and the thickness of the strips that will divide and hold them. This, in turn, should relate to the

Whether for glass or for wood shelves, I make two or three

holes for shelf pegs at each of the levels that seem well placed, thus allowing for a certain amount of leeway. Another eye

shape of the curve we have chosen, the size and proportion of

the door. We must now back up a bit, to the point where we conceived the idea for this piece and, with it, the main parts,

may find a variation in my choice, even better. Let's decide

on a spacing. And turn to thinking about making the faceted templates with which to rout the rabbets on the curved hori-

the relationship of the front to the rest of the cabinet. Right

or wrong, some details could hardly have been more than a guess. Now, deciding on how the vertical strips, and behind

them the shelves, will look, we do more than guess. The dif-

zontals of our doors. Plan this so the depth of the rabbets will be equal at the various corners. (The measurement of this depth is from the inside of the door.) Do take time now to get it as exact as possible. Later our rabbets in the sidepieces (ver-

front. If they simply followed the curve of the door,

The vertical strips (pins) that separate the pieces of glass are shaped to a tighter curve than the cabinet

they would appear flat, being less than

in. wide.

seems elongated, Right, with glass shelves.

18 cm (7 in.) deep, oil finish. The type of shelves, or lack of shelves, changes its effect: at left, with maple shelves, as it was made. Center, the same cabinet without shelves

Wall-hung showcase of Swedish maple, 88 cm (34

in.) high, 56 cm (22 in.) wide,

deals) will be made to coincide with these, and this early care will pay off then. Usually I cut the rabbets about three-quarters of the depth

(or thickness) of the door itself. In other words, there is rather little wood at certain parts (depending on the curve and how we divide it) of the front edge. The glass is up front in the
door, rather than being set back. I think this gives a sense of lightness; the door appears less thick and therefore less awkward. Besides, there is more room behind the glass for the fitted wood pieces that hold it in place.

we shift from one surface of the clamp block to the other, being careful not to change its relationship to marks that indicate the positions of pins and rabbet ends. We can do likewise with a pair of doors, or else make two

identical templates, which we handily use without shifting them on the clamp block: The lower left-hand one is also the upper right-hand one, provided they are really accurate! All
our planning is from the inside of the door. You will notice in the photos that the piece is clamped to the template so the router has access from what corresponds to the inside of the

Making one door only, we need a single template, which

door part being worked. We concentrate our attention and accuracy where it really countsto the way the glass will fit the door. If there are
small differences in the thickness of the wood at the front edge of the rabbetbetween the door front and the glassit does not really matter all that much; a very slight variation

here will bother no one. Granted, however, we do try to get each of our measurements as accurate as possible. Gradually, the relationship of details becomes clear; we coordinate the
steps and methods that are important, and we master them. We learn how to plan: where to allow in our measurements, and how much.

To do the rabbets neatly, we must prepare the various parts by marking off exactly where each cut is to be made. Mark (while clamped up) the starting and stopping points of each
cut, and chisel a notch there to prevent chipping out. Then, if the grain is at all difficult, it is safest to ensure the lower edge of the cuts with a very fine scribe. We do the more

Once the pin spacing is determined, convert the curve to a series of

below) is the same one used for laminating the stock. Rout with several small cuts rather than one large bite, to avoid chipping.

parts), is for a pair of doors. The clamp support block (seen from

the rabbets. This one, with both its halves alike (it could be in two

flat facets for the glass and make a router template with which to cut

complicated horizontal (shaped) pieces of the door first, then dry-clamp and mark accurately for the work on the sidepieces. Keep the margins as small as possible! With some doors,
where one side (toward the middle of the cabinet) is narrow, I machine a groove instead of a rabbet to meet the adjoining rabbets precisely. And here again the accuracy needed is rela-

tive: An allowance on the inside thickness of this frame part helps us to arrive at a common depth from which to set exactly the table saw or router. With the rabbets done, we can make the vertical strips or pins that will separate the panes of glass. From the beginning it is necessary to know the thickness of the glass we are to use. When we saw or rout the grooves to fit, there is one thing more to remember: The pieces of glass as they meet at these strips will be at a slight angle. The angle corresponds to the
various facets making up the door in its curve. Each groove,

therefore, needs to be a trifle wider than the thickness of the glass to allow for this slight angling. Take this into account,

but you should not make the groove sloppy and allow too much for the thickness of the glass. Without a proper fit, the glass is liable to rattle as a car passes outside. Some of you may prefer to cut the various pieces of glass to

Before routing, a small stopnotch is chiseled at the ends of each cut, and the edge of the cut
is scribed on the vertical parts (above). Where two doors come

fit. This isn't easy! Having tried, I now go to my favorite

glasmastare, as we call glaziers in Sweden, and have him do it for me. Before doing so I make a first assembly of the pins and the strips that are to hold them in place. Then, using
scraps of thin plywood veneer, I cut slip-ins that correspond
to the exact width of each pane of glass, all of which have a common length, namely the height of the door between rabbets. My glasmastare is very kind and patient; he cuts the

apt to be narrow and one can

together (left),

their parts are

saw or rout a slot instead of a

ford a better view.

rabbet. The frame is then beveled inward from the slot, to af-

glass extremely accurately. This, in turn, makes my work of final fitting much easier. If you plan to cut the glass yourself, try to do it as neatly as he does; it will pay off later on. I have tried clumsily, with my photos, to show the various

From left to right, top to bottom, account for the angle between the glass panels when routing grooves in the vertical pins, then cut the pins to shape.

Their tenons fit cutouts in the beveled strips that, fill the rabbets in the top and bottom rails. Clamp together and drill for the brads that hold everything in place, using thin pieces of plywood to maintain correct spacing. And now it's done. The glass should fit snugly, so it doesn't rattle.

steps in fitting the glass and the wood parts that hold it in place. There are beveled strips at the sides of the door and

shaped pieces in the rabbets, top and bottom, notched to fit the various pins between the sections of glass. Usually, when I make the laminated parts of the door, I make them wider
than need be and then bandsaw two or three thin layers off the curved shapes. These I later use to secure the glass. With such pieces there is no splash grain, and they are neat and

easy to work since they already have the shape of the door itselfI need only trim them to fit the rabbet and then notch
them for the pins. I do not use screws here, but prefer brads. With these it is easy to remove the glass if need be by carefully prying up the hold-in pieces. First I drill holes in the strips the size of an easy fit for the brads. Then I place the

glass in the door together with the various pieces of fitted wood to hold the glass. With a slightly smaller drill, now a very tight fit for the brad, I drill through the various parts at a
slight inward angle, making my hole deep enough for the whole brad. I countersink ever so slightly for the head itself. The brads are small and hardly noticeable, so they do not bother us as we view the cabinet. I take it for granted that we

have tapped and then set them without leaving any marks on the wood.
Before polishing and setting in the glass for keeps, I carefully fit the door with its hinges to the cabinet. To allow for the fact that it will sag a trifle from the weight of the glass, I

make the door fit a bit tightly upward. Then I polish the door and the various strips that will hold the glass in place. If there is to be a handle, I fasten this to the door. I finish everything
possible before I put in the clean glass and hang the door.
Trying to describe showcase cabinets, I find myself talking mostly about a door or doors and the process of making and

fore, that special attention to it and its problems is justified. When we can make a door on the level of our intentions for the rest of the cabinet, I believe the chances of success are very good indeed. Our showcase cabinet is more than a glassed door or two. We should be aware of this rather early, so that when we sketch or draw or otherwise plan the piece with its important front, we think of it in relation to the whole: the degree of detail and refinement, the proportions, the amount of the back that will be exposed, how the shelves will affect the proportions as well as how they will cast their shadows upon the back piece itself. Another point: Usually included in our idea of such a piece is the way it will be usedon a wall at a certain practical height, or perhaps with a stand that should be a pleasing part of it. Try to get a clear impression of these possibilities. When looking at the work during the various stages, imagine (or simulate) the final way the piece will want to rest. Come as near as you can to reality This ability to achieve a wellbalanced whole is dependent on observation (which becomes experience), and it is natural at first to be uncertain. When thinking of a stand, we may be tempted to imagine curves, shapesnot just a stand. All right, let your fancy go, but then allow for some other considerations. How will the stand serve its purpose, which is to support the case at the most pleasing height (with the doors open as well as closed)? How will the stand harmonize with the case? Harmony includes lines, and also volume; a stand has "weight," just as the case does. In reminding you of simplicity and harmony, I don't intend to be inhibiting; certainly a person with a mind for
fantasy and taste can combine flair and harmony skillfully. A last practical note: In showcases that have curves, the sides are apt to be set at an angle, which complicates the

fitting them. It seems a rather dull and one-note description.

Actually, a showcase cabinet is much more than a door or doors, or a glass front with a few objects showing through. Still, the door is usually the most difficult part. I feel, there-

joints needed in a strong and well-made stand. I usually make

these angled joints with spline tenons done neatly on the horizontal mortiser (that too-often overlooked machine). Take

time to lay out properly, get all the angles correct and fits snug. Dry-clamp, and then study the stand with the case together. . . .Return, and look again. When you decide to glue

work, discovering possibilities and satisfactions of their own. My attempts are limited; the illustrations here are meant only as a beginning. There is a great deal more to be done than I

up, have everything you need, including the specially fitted

blocks that go with getting such pieces together. I hope and believe some craftsmen w i l l go into this kind of

have even imagined. For those who try this path and then decide to abandon it, here's a consolation: One can learn something about oneself along the way.

rabbet. Note that two thin shims of cardboard are used to raise each piece slightly while drilling for the stop-nails. When glued and

The two outer pieces, mitered at the middle, form the rest of the

clamped, with the shims removed, the miter fits tightly together.

Above left, the vertical sidepiece and bottom rail for a V-shaped
paper against door surface, for a good grip.

glue. At right, the door is glued in only one direction at a time. The other half is merely set in place dry. Clamping blocks have sandFrom left to right, top to bottom, stock for the middle pin is only in. wide, so a hardwood stiffener (darker wood) is glued to the bevel is hand-planed. Slots for the glass are routed and the stiffener is removed. Both panes of glass are pushed into their slots in this center pin, then the whole assembly slides into the door frame from the back, to be held in place by wooden strips as before.
back of it. A neat slot is scribed and cut in the door rail, then the

door. Shading around spline tenon is a slight chamfer, to trap excess

Swedish-maple showcase (1965) has a single door set in a mild V, and

deep. Oil finish.

measures 86 cm (34 in.) high, 46 cm (18 in.) wide, 19 cm (

in.)

The horizontal rail for a V-shaped door is assembled from three pieces of wood. The inner piece (which forms the bottom of the rabbet that receives the glass) is bandsawn from a single piece of wood. This makes the basic shape without a complex joint.

Drop-Leaf and Gate-Leg Tables

Graceful proportions make all the difference
by Simon Watts

a bles with hinged leaves have been made in America for at least 300 years, although no surviving examples date back to before about 1700. The early tables were often used in taverns' and, when not in use, could be pushed back against the wall to make more floor space. The first drop-leaf tables had a square edge between the leaf and the topsimple but crude. The barrel of the iron hinge was left exposed and there was a substantial gap when the leaf was down. Later tables had a 45 angle cut in the lower edge of each board, with the hinge set in from the underside. At some point an unidentified genius invented the rule joint. This elegant detail allows the hinge to be concealed completely. Special steel table hinges are now made that have one leaf longer than the other. This longer leaf is attached to the table leaf; the shorter leaf is attached to the tabletop. I rout a short recess for the barrel of the hinge but there is no need to set the whole hinge into the wood. It's just extra work, and results in making the hinge more visible when the leaf is down. The mating edge profiles are best cut with a router, although originally molding planes were used. In my experience it is a mistake to make the joint too snug because crumbs and other debris tend to get jammed in it. The leaves of early tables were supported by slides or swinging brackets, but when larger tables with wider leaves were made, a swinging leg became necessary. These became known as "gate-leg" tables. For convenience I will call the first kind a drop-leaf and the second a gate-leg. The leaves of a drop-leaf table should not be narrower than about half the width of the top. I usually make the top 18 in. to 22 in. wide and the leaves 10 in. to 12 in. If the table is much narrower the legs get so close together that it might

overturn when used with only one leaf up. If the table plus leaves is much wider than about 44 in., it will look bulky unless it is made too long for the average dining room. A table 6 ft. long seats eight comfortably. The leaves do not have to be rectangular; they can be curved on the long side, with the slight disadvantage of reducing the space available at the end. I don't like oval-shaped drop-leaf tables because the curve crossing the rule joint makes part of the joint project in an unsightly way. Leaves can also drop from the ends of the table instead of from the side, although it is difficult to gain enough length for a whole place-setting without embracing a leg between one's knees. The table shown in the photos and drawing was made for a family with three small children. I suggested putting a radius on the corners but the parents decided that would detract from its appearance so the corners were left square. By now all three children have met a corner head-on, and one required stitches. I think a furniture maker is obliged to point out this kind of hazard, but should not insist on doing it his way. The choice of wood arid the boards selected for the leaves are important. This is because the drop leaves are not restrained, except along their top edges, and if they cup or twist there is no structure to prevent them from doing so. For this reason I use a very stable wood, such as mahogany, and try to use boards whose annual rings run nearly at right angles to the surface (vertical grain). The leaves have to look good when down and should match the top and each other when up. Ideally, one ought to make leaves and top out of boards cut from the same tree. Top and bottom surfaces of the leaves must be finished in exactly the same way. Otherwise, the side of the wood with less finish will pick up moisture or dry out more readily than the other, and the leaf will cup. EDITOR'S NOTE: For another version of the gate-leg table, see Fine Woodworking, Summer '76. For other ways of making tables that enlarge, see "Expanding Tables," Spring '77, "Gaming Tables," Fall '77, and "Designing for Dining," Winter '77. For Parsons tables, see Summer '78. ported by four hingesa pair near the ends, and a pair inboard of the slide supports.

The drop-leaf table shown in the plan drawing on the next page, has routed rule joints, (detail at right). Each leaf of the table is sup-

There must be enough knee room for people to sit at the ends of the table. My experience is that a top-to-apron overhang of between one-fifth and one-sixth the length of the
table accomplishes this and also looks right. The amount of taper of the legs is also important. Too much taper and the table looks nervousas if it were about to get up and walk away. Too little taper and the whole piece begins to look

look of these tables more than drooping leaves. I put a small blocking piece between the slide and the underside of the tabletop. This angles the slide up very slightly, and with a bit
of adjustment the leaf can be made to lie dead level. If the leaves do begin to droop due to wear, the blocking piece can easily be replaced with one slightly thicker. Some people make drop-leaf tables with a hinged bracket

clumsy. There are no rules except that what looks right generally is right. I always put a substantial chamfer around the bottom end of the legs where they meet the floor, especially if
the piece is too heavy to lift. Furniture always gets dragged around and the legs eventually wear down. If there is little or no chamfer the legs can splinter.

The most difficult detail on a drop-leaf table is designing the slides so they will support the leaves. Nothing spoils the

to support the leaf. This is not good practice because it strains the hinges and twists the apron. The strength of the table is in the joints between the aprons and the legs, so these must fit well and be properly proportioned. I use a haunched tenon joint and offset the mortises. The tenons should not meet inside the leg. If they do they seriously weaken the joint, which can then be split out by an accidental kick. The top is attached to the base with steel tabletop fasteners

SEE ERRATA AT END OF ARTICLE

and the support bracket. A brass pin in the slide stops its travel against the bracket when the leaf is down, and against the apron

Photograph of a drop-leaf table, taken from below, left, shows slides

right: leaf with curved edge, and leaves that are hung from the ends of the table.

when the leaf is up. Two variations of the drop-leaf table, center and

that fit in a groove or sawcut around the inside of the

apron. They can easily be

made in wood but the grain must run at right angles to the lip. The apron should never be screwed or glued directly to the top, as this would prevent movement

think this reduces the possibility of the leaf curling if it is exposed to the heat from a stove or radiator. If possible, choose vertical-grain stock.

The boards to be joined should not have any twist or wind

as this cannot be taken out by clamping. However, a bowed board can usually be straightened by putting it between two

aprons with metal clips or with

The top can be attached to the

straight pieces or by pairing it with a board bowed in the opposite direction. Mating edges can be planed by hand with a
long jointing plane or a power jointer. When using a power

The gate-leg or swinging along the inside of the apron. leg has many similarities. Since the open leaf is supported by a leg and not by a slide,

screwed to the tabletop and engages a saw kerf running

wooden buttons, Either device is

between the two caused by changes in humidity.

jointer, run the boards alternating their faces against the

fence. Any errors (due to the fence not being precisely at right angles to the bed) will cancel, not accumulate. It is good policy to plane the boards with a slight hollow (less than in.

over 72 in.) so that their ends are pressed together and there is
less chance of a joint opening under seasonal movement. On no account should there be any camber.

the middle section can be much narrower than the middle of

a drop-leaf. This way a table that is rather large when open

can be put against the wall when not in use, occupying very
little floor space. However, if you make the middle section less than about 12 in. wide, the table is liable to get knocked

I always glue up stock on edge vertically, holding the bottom board in a vise or standing it on sawhorses. If you try and do it on the flat the glue always runs down to the lower edge and you get a starved joint. I also dowel with -in. by 2-in.

over when both leaves are down. The maximum depth of leaf for a table of standard height (28 in.) is about 25 in. Any
deeper and the bottom edge gets too close to the floor and will be kicked. The shape of the top can be square, rectangu-

hardwood pins on 8-in. centers to keep the edges from sliding

on each other when clamping pressure is applied. Of course some use a spline for the same reason, but unless its groove is

lar, round or oval, but if the length is more than about 60 in.
a second gate is desirable, which interferes with seating. It

stopped near the ends, it will show.

Before gluing, the boards should be set up dry in final sequence. They should be placed on each other vertically and

also makes for a heavy leaf, awkward for one person to lift. The rule joint is the same but because the leaf is deeper the choice of wood is more crucial. If a single wide board cannot be found, a number of narrower boards will have to be
joined. Opinions vary, but my own feeling is that the heart

should sit fair without any rocking. Check one side with a
straightedge to be sure the surface is flat, then mark and drill the dowel positions. A quick way to close up the joints before

side and the sap side of adjacent pieces should alternate. I

Simon Watts makes furniture in Putney, Vt.

ventional leg-and-apron understructure, by adding a low rail between the legs where the gate can attach.
putting on clamps is to stand the whole assembly on the bench, pick up one end and then drop ithard. Then pick up the other end and drop it too. It makes an awful racket

Gate-leg table in open and dosed positions, as shown in the drawing on page 64. The gate-leg principle can also be applied to the more con-

points to the sides of the wood and square across. Now add pencil lines where the circle runs out on the faces and ends. The lines mark the point where no wood will be removed.

but is far quicker than winding up the clamps one by one.

The clamps should be alternated from front to back of the

assembly, and if the wood is not to be replaned, slip waxed
paper or plastic under them to avoid staining the wood. A clamp over every dowel is good practice, and always put one at each end. Bar clamps are rigid but pipe clamps, if threaded

The layout completed, I saw the chamfers and the bulk of the waste with the table saw set at 45, then finish the rounds with a rasp and sandpaper. Divide the width of the wood into
four or five fingers, saw in on the waste side of the lines, and chisel out the waste from both sides, just as for dovetails. You can get at the inner hollow on the ends with an in-cannel

at each end, can be made up with couplings to any length required. I like to keep some of each, but bar clamps over 6 ft. long are unwieldy. There should always be plenty of glue
squeezing out of the joints, but beware of too much pressureyou can squeeze out too much and starve the joint. The traditional wooden hinge can be mastered by anyone

gouge, but between the fingers the waste has to be removed with a chisel, bevel downward. Finally, fit the joint together and drill through both pieces at once for the pin. I usually use
a brass pin, peened over at both ends so it can't fall out. Make the end pieces of the table by cutting a board in half lengthwise; then bandsaw the waste and finally dowel and rejoin the two halves. A saber saw or a small frame saw could

with a bit of patience who is willing to make one or two practice sets. This example pivots through 180, but it can be stopped anywhere by varying the chamfer behind the knuckle
itself. Square up the stock, then lay out the joint by gauging the stock thickness all around the ends of both pieces. Draw in the diagonals on both edges, to locate the center of pivot,

make the cutouts without splitting the board, but the finishing tends to be tedious. The frames and gates can be left
square, chamfered or radiused. I prefer some rounding.

Any dry, reasonably stable hardwood can be used for these tables. I like walnut, cherry, teak and mahogany. I don't
think they look as good when made out of blond woods such

and with dividers scribe the circular shape. The circle crosses the diagonal at the bottom of the chamfer; extend these

as oak or maple, but that's just an opinion.

Cheap clamps
I don't like pipe clamps because the pipes bend under load and mar the work. I can't afford long bar clamps, so I make the 2x4-and-

folding-wedge clamps shown in the drawing below.

The materials bill for each clamp is an 8-ft. or 10-ft. utility stud crosscut in half, a foot of 1-in. hardwood dowel, and some scrap for cleats, blocks and wedges. Avoid twisted 2x4s, although a little bow doesn't matter. Make the wedges about 12 in. long, tapering about 1 in. over 9 in., for slow, firm squeeze. Keep the dowel holes about 6 in. apart and span intermediate widths by adding spacer blocks between cleat and wedges. I usually take the clamps apart and use half

of each as an assembly bed, then put them together right around the
work. Add blocks until the narrow end of both wedges just fits the space, then drive the wedges in together. Keep the joints in line by
putting hand-screws across them at the ends, and fine-tune with

smaller wedges driven between the clamps and the tabletop. Larry Green

ERRATA
I have just read Simon Watts' article, "Drop-Leaf and GateLeg Tables," in the Sept.'79 issue and am very impressed with his work. However, in studying the drawings of the drop-leaf table, there is an error in the measurement of the tabletop in the end view. If Watts generally makes his tops from 18 in. to 22 in. wide, and his leaves 10 to 12 in. wide, there is no way the width of the top and one leaf could possibly add up to 19 5/8 in. I was wondering if Watts has done or considered making the end of the table shorter and placing a drawer in the end, as the Shakers have done.. . .

Ralph E. Huss, Okemos, Mick,

Simon Watts replies: The top and one leaf should be 29 5 /8 in. Yes, I have put drawers in on occasion. With a drop-leaf table the overhang has to be kept small or the drawer is too far back under the tabletop. This can interfere with seating. Without leaves you can put the drawer anywhere you want but aprons deeper than about 3 in. can get in the way of people's knees. This means the drawer cannot be much deeper than 2 1/2 in.

A new book of measured drawings

by John Kassay

Shaker Blanket Chest

EDITOR'S NOTE: The best examples of Shaker furniture "were not the studied approach to design or a conscious effort to create masterpieces, but were rather expressions of utility, simplicity and perfection attributable to spiritual inspiration, moral responsiveness, dedication to a craft, and skill." This is how John Kassay, professor of industrial design at San Francisco State College, views the genius of Shaker designs. The following article is excerpted from Kassay's Book of Shaker Furniture (University of Massachusetts Press, Box 429, Amherst, Mass. 01004, 1980; $33). There are several other available books about Shaker furniture, most notably John Shea's The Amer-

ican Shakers and Their Furniture, Thomas Moser's How to Build Shaker Furniture, Robert Meader's Illustrated Guide to Shaker Furniture and Ejner Handberg's three-volume work, Shop Drawings of Shaker Furniture and Woodenware. With its concise introduction, its uncluttered format, superb illustrations and summary descriptions, Kassay's book is not intended for the scholar who's interested in social and religious history. Rather, the book is aimed squarely at the serious woodworker, who can scale his own shop plans directly from Kassay's drawings and purchase and dimension his stock from the accompanying bills of materials.

nscribed in burnt letters on the back of this one-drawer blanket chest of pine in original red paint is "April, 1837 Canaan." The chest is attributed to Brother Gilbert Avery (1775-1853), a member of the Upper Canaan family, which was a part of the New Lebanon [N.Y.] community. A plinth with dovetailed corners and convex cutouts raises the chest off the floor. Four corner blocks fastened to the inside corners lend added support. The sides and ends of the chest are held together with dovetails. A lidded till with a drawer beneath is at the inside left end. As an afterthought, a hole had to be cut in the chest bottom to allow air trapped behind the drawer to escape into the chest proper. The applied tongue-and-groove molding at the ends of the hinged top is typical of Shaker work. The key escutcheons are of bone. Collection of Mrs. Edward Deming Andrews.

Dovetail box, steambend handle

by John Kassay

Shaker Carrier

his not-so-difficult-looking project offers two challenges the hand-cut, through-dovetail corners and the sculptured, steam-bent bail (handle). Carrier is the Shaker name for a box fitted with a bail. Those carriers that exhibit pleasing form, fine construction, and quality craftsmanship were made for the Shakers' own use, whereas carriers made for sale in Shaker stores, though well crafted, look mass produced. With the exception of the manner in which the bail is fastened, this carrier is a fine example of one made for communal use. To make the carrier, thickness-plane enough pine (wood species is optional) to make the sides (A), ends (B) and bottom (C). All surfaces should be hand-scraped and sanded. Those surfaces that will be on the inside of the carrier should be finished surfaces and so marked. Now lay out the sides and ends and add in. to their widths and lengths, and cut accordingly. The extra length allows the ends of the dovetails to project minutely beyond the outside surfaces. After the sides and ends are assembled, these projections are planed or sanded off, resulting in a better appearing dovetail joint. The extra width is used for truing up the edges at the top and bottom of the carrier, again after assembly. Mark out and cut the bottom in. longer and wider than the overall length and width of the carrier sides and ends. Sand the inside surface and shape the upper edges as shown in the drawing. Nail the bottom in placea nice touch here would be to use -in. fine-cut headless brads (available from Woodcraft, 313 Montvale Ave., Woburn, Mass. 01888). Nailing the bottom onto the carrier sides may seem to contradict all we have been taught about wood movement, but it is the way the Shakers did itand they had central heating too. It has been suggested that the bottom ought to be let into a groove in the sides, like a frame-and-panel. However, I have rarely seen good results from altering a Shaker design. In this particular case, inletting the bottom would eliminate a characteristic Shaker form, the molding created by the protruding bottom, and it would greatly complicate the carrier's joinery. I think that when the bottom worked loose, the Shakers would just nail it on again. The bail is made of ash; red or white oak or hickory could be used instead. Mill straight-grained stock to overall thickness, width and length (detail 2), then steam it and bend it around a mold before tapering it to shape. Although it's difficult to shape the bail after bending, it's more frustrating to lose a pre-shaped bail during the bending process. The photo on the facing page shows my bending jig, with a back-strap made of four strips of 24-gauge galvanized sheet steel, spot-welded together at the center (FWW #8, Fall '77, p.40, and #30, Sept. '81, p. 84). This apparatus will bend John Kassay is the author of The Book of Shaker Furniture, available for $40 from University of Massachusetts Press,

Box 429, Amherst, Mass. 01004.

kiln-dried white oak that's been steamed for about two hours under low pressure (5 PSI to 10 PSI). If you use split-out green wood, the chance of a successful bend is greatly increased; you can probably substitute an ordinary band clamp for the steel back-up strap and end blocks. I leave the bent stock on the jig to set for a couple of days. When removed, it springs back just the right amount to fit the carrier. Now make a full-size pattern of half the length of the bail, trace it onto the bent wood and cut out the shape. With a block plane and a scraper blade, taper the bail in thickness from the center to the ends, as shown in the edge view, then spokeshave it to the cross-sections shown. Note that the undersurface is rounded, while the outer surface is left flat. Both ends of the bail are flat where they attach to the carrier ends, and chamfered on their outer corners. Fine-sand all the surfaces and ease any sharp corners, except those where the bail meets the carrier. Fasten the bail with four brass rivets and washers, two at each end; you could substitute countersunk flat-head woodscrews. The inside surface of the original carrier was protected with a wash coat of yellow milk-paint, while the outside was left natural. The bail was varnished.

Wedges hold bent stock against bending form while it cools and sets. Steel back-up strap with end blocks helps make the bend, but once bent, the strap can be tipped away from the stock, as shown.

Knockdown design for a dual-purpose project

by Tage Frid

Folding Stool With Tray

was asked to design a folding stool that would be light, take up little space when folded, and serve as the base for a tray. In addition, any parts broken during service would have to be easily replaceable. When the stool was finished it weighed 4 lb., and measured 1 in. folded. Nothing has broken yet, so I haven't had to take it apart, but I could if I wanted to and it would go back together good as new. I made the stool from ash. If I had used a weaker wood, I would have added to the thicknesses and widths for strength. The seat can be either leather or canvas. The one shown is canvas, with a single row of stitches to make a hem at the edges and a double row to hold the 3-in. overlap. The stretchers can be held to the legs with either T-nuts or barrel nuts and -in. stove bolts. The stool shown here has T-nuts, which leave the holes in the stretchers open. Barrel nuts would have filled the holes and looked like metal plugs. Where the stretchers butt against the legs there's a hidden dowel (or a steel pin) that keeps the stretchers from turning. A washer between the legs where they cross allows the stool, to fold easily, and double nuts are locked together so they don't have to be drawn too tight. If a tight single nut were used, the stool wouldn't fold. Washers under the bolt heads protect the wood. The legs are identical except for the angle on the footthe angle makes it a right leg or a left, to keep the dowel holes inside. Mill the leg blanks, square them and cut them to exact length. Set up the drill press with stops to locate the holes, and then drill them all. Notice that the holes for the dowels

or steel pins don't go throughmake these holes in. deep. After you've drilled the holes, taper the legs with a taper jig on the tablesaw or on the bandsaw. Cut a little wide so you can run the edges over the jointer to remove the saw marks, and then cut the foot angle. The stretchers are all the same length. I made the bottom stretchers in. wider than the top because people have a tendency to put their feet on them when they sit on a stool, but the stretchers could be all the same size. Mill them out and cut them to length, then use a stop on the drill press to make the holes for the T-nuts or barrel nuts. For T-nuts, make -in. holes; for barrel nuts, use -in. holes. Of course, regular nuts could be used if the others aren't available, but barrel nuts are easy to make. My students and I use them a lotthey make an attractive and strong joint if a piece has to be disassembled. They can be of -in. cold-rolled steel, aluminum, brass, or other rod stock. Cut the nuts to length, so they will be flush with the surface if you want them to show, or shorter if you want to use them in a blind hole. File and sand the ends, then drill and thread holes for the bolts. Use a V-block jig in the drill press to bore the hole. Remember to countersink these holes so the bolt will start easilywhen it comes time for assembly, you can wiggle the nut until you feel the bolt start to engage. If you use barrel nuts or regular nuts for the stool, use a -in. stove bolt, 2 in. long. For T-nuts, use a 2-in. long bolt. Tilt the drill-press table to 90 and clamp a jig to hold the stretchers while you drill the holes

The taper in the legs of this stool cuts down weight, leaves the wood where it's needed, and allows the stool to close up to a snug have to store this stool in the closet when you're not sitting on it make a tray that converts it into an occasional table or server, as shown above.

in. T-nuts in the stretchers allow disassembly. You don't

for the bolts and dowels. The stool is now ready to be assembled, but first chamfer all the edges with a router or a plane,
then sand and finish the pieces.

Don't make the tray before you have assembled the stool
and measured it to be sure that the tray will fit. This one is an ash frame with a panel of -in. walnut plywood in a

no trick to making the trayI cut the corners to 45, rubbed them together with hot glue, then strengthened them with a mock finger joint, which I learned from that wonderful book by Tage Frid.

groove. I didn't use solid wood for the panel because, to remain stable, it would have had to be in. thick, and that
would have made the tray too heavy to carry around. There's

Tage Frid is professor emeritus of furniture design at Rhode Island School of Design, and the author of Joinery:
Tools and Techniques and Shaping, Veneering, Finishing,

both of which are available from The Taunton Press.

Lots to learn from this 18th-century case study

text and drawings by Victor J. Taylor

Building a Secretaire-Bookcase

t's not often that you come across a piece of English antique furniture that can be dated precisely, but glued to one of the drawer linings of this handsome secretaire-bookcase is the following receipt: "B. Milward [the purchaser]. Jan 25. 1787. Bought of Mr. Evans, Broadmead, Bristol. Price 15.15." Today the piece stands in the Withdrawing Room of the Georgian House, Bristol, which is a real treasure store of late 18th-century household goods ranging from fine furniture and priceless paintings down to kitchen utensils. It is officially described as Hepplewhite style, but it seems to me that the date is too early for Hepplewhite, and that the piece is more likely late Adam. In drawing this complex piece, I was struck

by how instructive it can be of various features common to much simpler furniture. Rather than follow a strict (and probably oppressive) how-to-do-it formula, I have attempted to present the piece as a tour of period construction practices, with side-trips into alternatives for the present-day craftsman. As can be seen in the drawing on this page, the piece is composed of five sections: from the floor upwards these are the plinth, the cupboard (containing cutlery and linen drawers, and two butler's trays), the secretaire, the bookcase and the cornice. These sections were usually made as complete, separate units, then fitted together, although in this piece the bookcase and cornice are combined as one unit. Often the sections merely rested on each other so that they could easily be dismantled if they had to be movedindeed, quite often the main cupboard section had handles fitted to it to make lifting easier. Usually the weight of each section kept it in place, with various sorts of blocks and keys serving to keep things from shifting. Mahogany is used for all show-wood parts, with oak and pine for the hidden parts and groundwork, normal practice for the time. The mahogany is almost certainly Cuban, and the superb "Spanish Feather" veneer is virtually unobtainable these days. All veneer is laid on without benefit of counterveneer, which would be risky with today's central heating. In the following drawings, each part of the secretaire-bookcase is illustrated and its construction explained, beginning with the plinth and working upwards, which is not necessarily the order in which it would be built. All pieces are numbered to correspond to the listing in the bill of materials on p. 60. In each figure, there is a small diagram of the full cabinetthe shaded part of the diagram is shown exploded in the drawing. In drawing the piece, where it was impossible to see the joints, I have followed orthodox cabinetmaking practice. Doweling, incidentally, was a very common method in the old days. Craftsmen made their own dowels by trimming down a suitable piece of scrap wood, and then hammering it through a dowel plate, a piece of -in. thick metal in which holes of various sizes had been drilled -in., -in. and -in. were usual. Dowels were often shaped from offcuts from the parts they were intended to join, minimizing uneven shrinkage. Willow was also used; its stems could be made into dowels with hardly any trimming. At the time this piece was made, French polishing had not been invented (it did not become widespread in England until about 1820), so the piece was probably originally finished with linseed oil and wax, then French polished at a later date.

Victor Taylor, of Bath, England, spent many years in the furniture industry. He has written seven books, and was
editor of the British magazine Woodworker.

Figure 1: Plinth. The basic members

of this subassembly are the four corner posts (1), which are connected at the

front and back by rails (2 & 3), and at each end by a rail (4), The cupboard
rests on this base and is almost certainly keyed to it with blocks screwed beneath the cupboard bottom (5), though I couldn't see them.

The feet (8) are not weight-bearing,

but are merely glued as decoration

around the corner posts. The ear pieces
(9) are glued and doweled to the feet, then the moldings (10) are pinned and glued on. Brackets are glued and

screwed into each corner to strengthen the whole framework. Screws (handmade) were first introduced in the late
17th century and by 1720 were common. Nails and pins (brads), of course, have been used for centuries, and there is even a reference from 1343 on using

an adze to smooth "old timber" full of

nails. On the arris of the foot there is a

staff bead, whose profile makes any opening of the joint less obvious. A central frame rail (11) is mortised flush

into the front and back rails. Following the usual practice of the time, the main

carcase ends (7) are lap-dovetailed to

the cupboard bottom (5). There is a
filler strip (6) beneath the cupboard doors, and this is shown in section at A.

framing. The doors overlap the upright

ends of the carcase, therefore the carcase ends have to be stepped back by % in. below the point where the front secretaire separation rail (15) meets them.

Figure 2: Cupboard and drawer

The front and back drawer rails (12) are tenoned into the main carcase ends

(7), as are the top separation rails 15

and 38 (visible in figure 4, overleaf). Muntins (14), drawer bearers (13) and

a central bearer (16) connect these four rails. The two upper drawers are supported by this conventional framing, while the lower single drawers run on bearers (30) glued to the cabinet ends.
This ignores wood movement, but the bearers are still secure. The drawer construction is orthodox, with lapped dove-

tails on the fronts and through dovetails on the backs. The bottoms are solid wood, grooved into the sides and fronts
without being glued in, so that they can expand and contract. You could, of course, use plywood for the bottoms instead. The handles on the drawers are

solid brass and match those on the fall front; they are shown at A in figure 10.

Figure 3: Door construction. The

doors are hung by pairs of brass hinges.
As you can see at A, the frame is a straightforward mortise-and-tenon job with a rabbet for the panel, which lies

flush, glued and pinned. The rabbets on the rail run the full shoulder length,
while those on the stiles are stopped, as

at B. The tenons on rails 33 and 34 go

right through the stiles, and their ends

can be seen on the outside edges. On the closing edges, however, a thin strip (35A) has been glued to the edge to
mask them. Blind tenons would do just as well here, and the cover strips could thus be omitted. A thin astragal bead-

ing (35) is fixed as shown at G.

The corner brackets (36) appear to

have been glued behind the panel merely to add rigidity. The doors were constructed entirely of oak, with no veneer on the inside. At

the time the piece was made, veneers

were sawn and consequently were much
thicker than our present-day veneers. The central part of the veneered panel would have been laid in a press, while a

veneering hammer would have been employed to put down the border.
Workmen trimmed the edges of the veneer, after laying, with a cutting gauge (C), simply a marking gauge with a

small, sharp blade instead of the usual marking pin. A pair of dividers, with one point sharpened, was used to scribe
the corners, as at D, and the cut was finished off with a knife. Detail E shows the small ovolo bead-

ing being gluedalmost certainly, it was steamed first. Both this and beading (35) are a blond color and could be
birch, sycamore or holly. The left-hand door has brass bolts top and bottom,

and a false escutcheon that matches the lock on the other door.

Figure 4: Panel framing. This panel

(17) appears to be -in. thick pine, but I'd suggest birch plywood instead. It is

pinned and glued into

-in. by

-in.

rabbets worked on the edges of rails 15, 37 and 38. Although the frame lines up
at the front with the front edge of the carcase end (7), it falls in. short of the back edge, to leave room for the

back framing. The moldings (41) are glued and pinned on, corners mitered.

Figure 5: Secretaire fall and housing. The pigeon-hole section has a clever feature that I have not seen on other pieces from this period. It is contained

within a fall-front drawer, shown at right and at the bottom of the previous
page, that pulls out to provide knee

room for writing. The fall front (42) is made up of two pieces face-glued together to form a lip that fits into a notch cut into the side (43) when the fall is down (see B and D). The fall has three
hinges, and is fitted with two brass handles, shown at A in figure 10. Two mahogany lippings (47 & 48) mask the

drawer front's top edge and the exposed

pine edge of the drawer top.

The quadrant stay was made from solid brass, with a small fixing flange brazed on. Cut out a -in. channel in the drawer side for the stay to run in.

You might wish to install a lock in the

fall front and catches fitted into the sides

(see A and B). The fall front is veneered similarly to

the cupboard doors, with the addition of a black inlaid line running from top

to bottom in the center. Lippings are

applied after veneering so that the top edge of the veneer is protected. The ends of the fall are painted blacknot
an attractive featureand you may wish

to substitute another thin lipping.

Figure 6: Secretaire unit. This is a

real work of art, as all the parts are only in. thick, except for the drawer fronts,
which are in. thick (including the veneer). All parts are mahogany. With

such thin partitions, a practical joint is the interlocking joint shown at C. The main structure comprises the
bottom (50) and the two ends (49),

which can be butted together and glued to the "drawer" side and bottom

nected with interlocking joints. Construction of the drawers is shown at D. The veneer is enlivened by black and white stringing about in. wide, and the cupboard door is further embellished with an inlaid fan. Now we come to an intriguing item: the secret compartments (B). Frankly, they are rather obvious and clumsy compared to some I have seen, and you may wish to elaborare upon them. They are built in behind the two pilasters (see A). Once you have opened the door, the two inner walls can be pulled inward and taken out completely. I had to pry them out with the point of a penknife, but probably a leaf spring had originally been fitted behind the tray to help push it out.

(51, 52, 53, 54, 55 & 68) can be con-

(43 & 44).

The remaining partitions

Figure 7: Main carcase top and

bookcase door. The top of the main

carcase (69) is fixed to the carcase ends by means of blind tenons on the main

carcase end (7). Moldings (71) are attached to the top (shown in section at A) so that the removable bookcase section does not shift. The bookcase is
made up of the two ends (70), the bot-

tom (74), and the top. The ends extend up to include the cornice, and we shall
be dealing with the upper part, including the top (76), in figure 9. The bottom (74) is housed in a rab-

bet formed at the foot of the end. The joint is glued and then strengthened
with wood screws driven in from the outsidethe surrounding molding (71)

will conceal the screw heads. The shelf supports (70A) are glued and pinned to
the carcase endthese supports are made from a piece of in. stock which first has a small thumb molding worked

on its front edge, and is then sawn into separate strips. Note that their back
ends must stand in. away from the rear edge of the bookcase end to allow

for the fitting of the back frame. On the actual piece, the corners of the bookcase doors have through tenons, but
I have drawn blind tenons, on the assumption that you will prefer them.

Figure 8: Glazing bars. Once the

bookcase door frames are made, lay them on a flat board so that you can

pencil the centerlines for the glazing bars on it, following the pattern and measurements shown. Leave off the astragal headings until you have gotten the center bars fitted. Delicate joints such as these (AE) can be reinforced with
strips of linen soaked in glue. In the original piece, the glass is fixed in place with putty instead of the fixing

beads shown at F. I cannot recommend putty, as it has no resiliency, and consequently the glass will crack easily if the
wood swells or shrinksin fact, several of the panes have done so.

The last step is to hang the doors with three 2-in. hinges per door, and if
you wish, you can fix a closing bead on to the right-hand door to match the one

on the cupboard door. Door stops can be fitted beneath the top, where they
will be out of the way. As with the cupboard doors, the

right-hand door has a lock and an escutcheon, while the left-hand door has just an escutcheon plus a brass bolt at
top and bottom.

Figure 9: The cornice and top. The

cornice (75) consists of a piece of mahogany lap-mitered to the top of the bookcase end (70). It is rabbeted along

its lower edge to house the bookcase

top. Note the dado for the top in the bookcase end, as shown at A.

The carcase top (77) laps over the

back framing (shown in figure 10), and

it also laps over the bookcase ends (70)

and the cornice (75). In the original, it is screwed down all around, which does not allow for wood movement. The piece will look best if you reproduce the original moldings instead of substituting lumberyard patterns. The top molding (78) is quite

straightforward, but the one below it

(79), which comprises the dentil motif

with a cavetto beneath it, is not so easy.

Probably the best way to tackle it is to

run off the outline profile first on a

spindle shaper, and then use a router to

take out the slots for the dentils. Then

you will need to chop out the rounded end of each slot with a small scribing

(in-cannel) gouge.
Now for the bracket molding (80).

On many designs of the period this was

a straight run of molding with the

brackets joined together at the top. In

our model, however, they will need to

be sawn out separately with a fretsaw or

jigsaw, and the small pieces of beading glued on beneath them. These small

pieces were turned on a lathe as "split"

turningstwo small blocks were glued together with a sheet of paper between them and then turned; it was easy to

split them apart afterward. Lastly, we have to deal with the Grecian key motif (81), and the best way, again, is to use a router, squaring up
with a chisel.

Once the brackets are glued on, it

will be difficult to polish into all the nooks and crannies, so you can adopt

the method employed by the old-timers.

First they would have polished the cornice and the brackets as separate pieces, then they would lay the brackets on the cornice to scribe around them. When they removed the brackets, an outline was left and the polish was scraped away from this. Next they warmed up a metal plate (called a sticking board), so that the glue would not chill when it was spread on it. They would draw the backs of the brackets lightly across the sticking board so that each received a thin coat of glue, enabling them to be fixed with no fear of gummy crevices.

Figure 10: Back framing and han-

grooved in all around as shown at C

dles. Chances are that the framed panels in this piece were screwed into place,

not glued. This would have allowed

their removal, considerably lightening the piece if it had to be moved. It was

bear in mind that solid panels must not be pinned or glued in place but left loose to shrink or swellleave some space, too, in the groove.
If you make the groove in. wide

difficult to see how the back frame was

constructed, so I am giving details of

by

in. deep, you can then use it to

typical framings that you can use. Those on the original were of solid oak, although you may wish to use pine. In
any case, the frames are made up with conventional mortise and tenon joints.

accept the tenons as well as the panels. You may stop the grooves on the stiles

(84 & 85)

to avoid their running

through the top and bottom edges when

the frame is made up, although when

everything is in its place, finally, these

The fielded panels

(82 &

83)

are

edges won't show.

59

Bill of Materials
planing, etc., at the rate of about 1 in.
in length, in. to The dimensions given below are net, and you should allow extra for sawing, in. in width, and

in. in thickness. Where I have shown shoulder lengths you will need to add extra length for tenons. I have left the tenon dimensions mostly up to you, and you may, of course, use whatever joinery you prefer throughout the piece. I have not included parts for the secret compartments, as no doubt you will wish to design more ingenious (and less publicized) ones of your own. In measuring a complicated piece like this, one often finds that many of the parts were scribed from other parts or cut to fit, rather than laid out with a ruler. I found I had to adapt some of the measurements in order to get things to add up. Although I have made every effort to ensure accuracy, parts of the cabinet were inaccessiblecooperative as the folks at Georgian House were, no one was about to let me move it, let alone take it apart. I suggest that you temper haste with a bit of caution, and cut to fit as you go along.

SEE ERRATA AT END OF ARTICLE

ERRATA
I know that publishing a magazine of the scope and detail of Fine Woodworking is no easy task. Nonetheless, whoever proofread the bill of materials on p. 60 of your January 1983 issue, #38, was sleepy indeed. If only one of each part (save for 16 drawer sides) is required, then I want to be there to witness the assembly of the finished secretaire-bookcase! I am sure that the requisite number of parts can be puzzled out; even so, I hope you will publish a corrected list. Abram Loft, Rochester, N.Y.

EDITOR'S NOTE: Ouch, we goofed. Secretaire builders will need only one of some parts. For the rest of them, here is a list of part numbers followed by the correct number of pieces:
1-4; 4-2; 7-2; 8-8; 9-8; 10-3; 13-2; 14-2; 18-2; 20-4; 21-2; 22-4; 23-2; 24-2; 27-2; 28-2; 29-2; 30-4; 31-2; 32-4; 33-2; 342; 35A-2; 36-8; 37-2; 39-2; 41-3; 42-2; 43-2; 49-2; 51-2; 52-4; 53-2; 54-2; 55-2; 56-2; 57-8; 59-2; 60-8; 61-2; 62-16; 63-8; 64-8; 65-2; 66-2; 67-4; 68-2; 70-2; 71-3; 72-4; 73-4; 78-3; 79-

3; 80-3; 81-3; 84-2; 85-2; 86-2; 87-2. If you'd rather have a new
p. 60, drop us a note and we'll mail it to you.

A Southern Huntboard
by Carlyle Lynch

Cock bead is an elegant touch for doors and drawers

ith the Southern forests rich with game and the housewife eager for the results of the day's hunt," wrote Paul Burroughs in Southern Antiques, "the sport was engaged in by all classes. The hunting boards around which the owners of Southern plantations gathered before and after the hunt resemble the sideboard. They were often simple in design. . . As a general rule, they were taller than sideboards . . . and were used chiefly in halls, where members of the hunt could stand and partake of wine and food in the fashion of a buffet lunch.'' Besides serving as informal hall furniture, huntboards helped keep the muddy hunters off the chairs. This huntboard is adapted from one that I measured and drew in 1952 while it was on loan to the Museum of Fine Arts in Richmond. It's like most of those illustrated in Burroughs' book in that it has four legs instead of the six usually found on sideboards, and it's of a convenient size. Within reason, the piece can be made longer, deeper or taller without destroying its appearance. When I built the huntboard, I put doors on the two end compartments instead of the deep drawers of the original. The center compartment could be fitted with doors or with two drawers of differing depths. The edges of the doors and drawer fronts of the original were decorated with a plain, but elegant, molding called cock bead and I recommend retaining this detail. Though cock bead is defined as any beading that stands proud of the surface it is meant to decorate, it is best applied as a strip glued to the edges, as shown in the drawer detail in the drawing, rather than merely stuck on the front. Cock bead is common on drawer fronts and door edges of furniture of the Chippendale, Hepplewhite and Sheraton styles. I see no reason why it couldn't be used to good effect on more contemporary furniture. Because cock bead is an applied molding, it can be of a different wood than that of the drawer or door, giving the maker an opportunity to experiment with colors and textures. I made the cock bead of cherry, which contrasts subtly with the walnut used for the rest of the piece and with the holly inlaid in the doors and drawer fronts. Building the huntboard is straightforward. The carcase consists of two solid wood sides and a back mortised into the four tapered legs. Openings for the doors and drawers are

formed by rails attached to the front legs. Two solid wood partitions, mortised into the front stiles and nailed through the carcase back, divide the case into three compartments. Except for plywood doors veneered with walnut, I built with solid wood throughout. But you could substitute plywood for the drawer bottoms and the carcase bottom. Start by making the legs. They are rectangular in section, , as on the original. Lay out and cut the joints to join the sides, back, and front rails to the legs. Then cut and fit the stiles, the center rail, the drawer runners and the two partitions. Dry-clamp the carcase before gluing it up. Before assembly, groove the bottom front rail to accept the bottom; the bottom itself, though, can be fitted later. For added strength, the leg, back and rail tenons should be pinned after assembly. So they won't warp or swell, the doors should be made of -in. plywood veneered on both sides. Don't forget to allow for the thickness of the cock bead when sizing the doors. If you squeeze the leaves of the hinges a bit in a vise, you can mortise them entirely into the legs instead of into both leg and door edgethis makes a neater appearance. Drawer construction is conventional. I allow for the cock bead on the top and bottom edge of the drawer fronts by making the fronts narrower than their sides by an amount equal to twice the thickness of the cock bead. Or, you could glue up the drawer and cut down the drawer front after assembly. In either case, cock bead on the drawer sides is let into a -in. wide, -in. deep rabbet. The rabbet should be cut after assembly so that the rearmost edge of the bead will just touch the tapered ends of the dovetail pins. I make cock bead by ripping thin strips and then using a jack plane to remove the sawmarks and shape the small radius on the bead's front edge. Once made, the bead is simply mitered to length and then glued in place so that it projects about in. You'll have to cut a stopped miter where the wider bead along the top and bottom edges of the drawer fronts meets the narrower bead on the drawer sides.

Carlyle Lynch is a retired designer, cabinetmaker and teacher. He lives in Broadway, Va. More of his drawings are available from Garrett Wade or Woodcraft Supply.

Materials List

HUNTBOARD

A Small Highboy
by Carlyle Lynch

Plans for a Queen Anne charmer

Author with his highboy of cherry, adapted from the mahogany original.

he highboy is an imposing furniture form, too large for the spaces in which most of us live. But 18th-century cabinetmakers didn't always build grandly scaled furniture for stately halls. Shown here is a small, modestly proportioned highboy I found in the home of Mr. and Mrs. Richard P. Lewis in Augusta County, Va. Included in the drawings are a few adaptationssimpler moldings and a less arched front apron that accommodates one more drawer than the original. I built the piece to test these alterations. Here are the basic procedures; a bill of materials is given on p. 34. The legsBegin by squaring the pieces for the legs. The article that follows on p. 36 gives a method for shaping cabriole legs with a bandsaw and hand tools. Here's how the lathe can be used, in addition to the tablesaw, bandsaw and hand tools, to shape the foot and ankle of these cabriole legs: Make a pattern from the drawing on the facing page, and lay out the leg on the two inside faces, so that the apron, sides and back will all fit flush with the post block. Cut the mortises in the post blocks while the leg blanks are still square. To shape the legs, first draw diagonals on the leg ends to mark their centers, and punch a mark on each end in. off-center, as shown in the drawing below. Mount each blank in the lathe on its true centers, with the foot end at the tailstock, and turn the foot. Shape it to the top of the pad, but don't finish turning the pad yet, or you will lose the offset center you need to turn the ankles. Remount the blank on the two opposing offset centers and turn the ankles. To someone not used to making cabriole legs, the setup looks forbidding. Use slow speed, and take light cuts with a sharp gouge or round-nose chisel held tight and fed slow. You can turn and sand 2 in. to 3 in. of the ankle, and sand to the top of the foot, before remounting the blank on its true centers to turn and sand the pads. Take the blanks to the tablesaw, and with a smooth-cutting blade set for maximum height, cut the waste to form the post blocks. Set up a stop block to prevent going too far. In order to keep the post block flat on the table, cut two of the legs with the rip fence to the right of the blade, two with the fence to the left of it. Finish the cuts on the bandsaw and then rough out the rest of the leg. Bandsaw to the pattern line on one face, tape the scraps back in place, turn the leg 90 and saw again to the lines on the scraps. Final shaping is done with spokeshave, rasp and scraper. The lower caseMill out the apron, sides and back, then cut the tenons to fit the mortises in the leg. Cut the bottom edge of the sides to shape, but wait to scroll-cut the apron until a gentle fit of its tenons poses no danger of breaking it. Dove-

tail the top rail into the top of the front legs, and test-assemble the frame (drawing, above). Disassemble, and cut the mortises in the apron and the top rail for the drawer stiles, and in the stiles for the central drawer rail. Tenons are in. thick by 1 in. long, except the stile tenons, which are in. long. Use poplar, pine or other secondary wood for the partitions that mortise into the back edge of the stiles and into the case back. Nail three drawer runners and a kicker strip to each of these partitions. Now add the cock beading to the apron edge. Cut strips of mahogany in. thick, in. wide, and long enough to bend, around the curves with enough to spare for cutting the miter joints. Round one edge with a small plane. Use a small gouge to make a groove in a sanding block for smoothing the round. You can use this same block later, to sand the cock beading for the drawers. Bend the apron beading strips between pairs of plywood forms, shaped to accommodate clamps. Make the curves of the forms a little tighter than the apron radii shown, to allow for springback. Also, make sure the curves on the forms are smooth, as rough or flat places can show up in the bent strips. Boil the strips in a shallow pan of water for ten minutes or so and clamp them in the forms while hot. When they're dry, finish-sand the beading and the apron face, miter the strips, and attach them with glue and small brads. Then fit and attach the short, straight pieces of beading. Before gluing up the lower case, dry-assemble it to check

1. Mount blank on true centers and turn foot. 2. Mount blank

in. off-center and turn ankle. 3. Remount blank on true centers and turn pad.

for fit and squareness. Disassemble, scrape and finish-sand all

parts. Then glue the legs to the sides, clamping the two subassemblies together, if necessary, to make them lie flat. You can pin the tenons now, while these subassemblies are in clamps, or after the whole case is glued up. Drill -in. holes

receive the buttons that will hold down the case top (detail A, facing page). Mortise the sides to receive the drawer rails three front and three back. Cut the -in. long tenons on the railstwin tenons for the front, single tenons for the back. For the top and bottom rails, cut the half-sliding-dovetail slots in

(four in each side) into the post blocks and through the tenons, coat the inside of the holes with glue, and hammer in
-in. square pins. A small handsaw with set removed will trim off the pins that protrude, without marring the sur-

the top edge of the sides (detail B) and the dovetail mortises
in the bottom edge. Fit the corresponding dovetail tenons in the top and bottom rails. Test-assemble the case, then take it apart and plow the slots in the inside edges of the drawer rails to receive the tenons of the drawer runners (detail C). The runners have -in. long tenons that fit these grooves, and

rounding surface. Pare flush with a sharp chisel. While the sides are drying, glue up the front frame, then

glue the partitions between this and the back. When these are
dry, finish gluing up the lower case, clamping and checking for squareness, and pin the two apron tenons and the four tenons at the corners of the back. To shape the knee blocks, bandsaw the six blanks and glue

they will be left unglued; the distance between shoulders is

in. short to allow the sides of the case to shrink. The inside edge of the top front rail is also mortised to receive the loose tenon of the top-drawer kicker strip. This does not connect to the back rail, but like the drawer runners will be attached

each one to scrap wood with paper in the joint. Use the scrap to clamp in the vise while rough-carving the blocks. Match
the shape of When they're off the paper finish carving the blocks to the contour of the leg's knee. shaped, pry the blocks from the scrap, scrape and glue, and glue the blocks in place. Now the blocks to fair smoothly into the leg.

to the side with a single screw. Test-assemble all parts, then take the case apart and glue it up.
The case sides are plain. The top that overlaps them is now molded on three edges and fastened with wood or metal tabletop buttons. Nail the molding strip under the top.

The upper caseBegin the upper case by milling the stock

and cutting it to length. Along the back edge of each side, cut the -in. rabbet for the backboards, and cut the grooves to

The drawersAll the drawers are constructed alikedovetailed front and back, with the bottom slid into a groove in
the sides and the front, and secured with nails to the bottom edge of the drawer back, as shown in detail D. (For a

BILL OF MATERIALS

SMALL QUEEN ANNE HIGHBOY

detailed discussion of standard drawer construction, see

FWW #11, pp. 50-53.) The cock bead is

in. thick and in.

either in. or in. wide, depending on whether it goes on the top and bottom edges of the drawers or on the sides

(drawing, above). Either way, it should stand about

proud of the drawer face. Note that the drawer fronts should be at least in. shorter than their openings, to accommodate the beading, top and bottom, and to allow for possible swelling. When the drawers are glued up, but before the bottoms are slid in, rabbet the drawer sides for the -in. wide cock

beading with a fine-tooth circular saw, guiding the drawer against both the miter fence and the rip fence. Do all final
sanding of the fronts, and prepare the top and bottom beading. Cut these full-length and then miter the ends, actually only the front half of their width, to meet the narrower cock beading on the drawer sides. To miter, clamp the cock bead-

ing between a 45 angled block of wood and a backing

board. Using the wood block as a guide, cut down with a sharp chisel. Apply the top and bottom beading with glue and nails, then miter and apply the beading in the rabbets along the drawer sides.

Finishing touchesAttach the drawer stops to the drawer

rails in the upper case and to the drawer runners in the lower case. To ensure a close fit for the rabbeted moldings that provide transition between the upper and lower cases, position the moldings on the bottom of the upper case and miter them to length. Then, using a strap clamp, glue up the three mold-

ing pieces using the filler strip to complete the back of the
rectangle. Attach this assembly to the top of the lower case. To key the two cases, drill two -in. holes in the top of the filler strip about 4 in. from each side of the case. Use dowel centers in the holes to mark the position of corresponding holes in the back rail of the upper case when the upper case

is set in position on the lower case. Separate the cases, drill the holes, and then insert 1-in. long, -in. dowels for keys
(detail E, p. 35). Rabbet the white pine planks for shiplapping and lay them horizontally in the rabbets you've cut in the back edge of the

upper case sides. Space the planks in. apart, to allow for expansion, and secure them with small nails.

Carlyle Lynch, a designer, cabinetmaker and retired teacher, lives in Broadway, Va. His plans for a Southern huntboard appeared in FWW #39, and others of his drawings are available from Garrett Wade, Lee Valley

Tools Ltd., or Woodcraft Supply.

Boston Bomb Chest

by Lance Patterson

Bulging drawer fronts are all shaped at once

ore than 50 pieces of American bomb furniture made in the last half of the 18th century still exist. Surprisingly, all were built in or around Boston. The kettle-shaped bomb form (the term is derived from the French word for bulge) is characterized by the swelling of the lower half of the carcase ends and front, with the swell returning to a normalsize base. This shape is, I think, directly related to English pieces such as the Apthorp chest-on-chest, which was imported to Boston before 1758 and is now at that city's Museum of Fine Arts. Bomb was popular in England for only 10 to 12 years, but remained the vogue in Boston for nearly 60 years. In America, the carcase ends were always shaped from thick, solid planks of mahogany. In Europe, the ballooning case ends were most often coopered3-in. to 4-in. pieces of wood were sawn to shape, glued up, contoured and then veneered. Instead of veneering, the Americans worked with sol-

id wood. I think the magnificent grain patterns of this shaped mahogany are a major attribute of Boston furniture. The bomb form, I believe, also shows the enthusiasm that 18th-century cabinetmakers must have felt when wide, dear mahogany first became available to them. There was also an evolution in the treatment of the case's inside surfaces and, consequently, in the shape of the drawers. On the earlier pieces, the case ends are not hollowed out and the drawer sides are vertical. Some transition pieces have lipped drawer fronts, the lip following the curve of the case. The fully evolved form has hollowed-out ends and drawers with sides shaped to follow the ends. Some of the later pieces have serpentine drawer fronts. I will describe how I built a small bomb chest with four shaped drawers, ball-and-claw feet and a serpentine front. I didn't take step-by-step photos while building, so I'll have to

shows shape of serpentine front.

Patterson's bomb chest, based on an 18th-century design, has four dovetailed drawers and ball-and-claw feet. Side view, right,

Fig. 1: Patterson's plans for a Boston bomb chest

Detail A: Rail and drawer runner

Plan is based on four existing 18th-century bomb

chests. Three of the originals are at the Museum

of Fine Arts in Boston, the Rhode Island School of Design and the museum at Winterthur, and one

was privately acquired through Israel Sack & Sons.

rail height should equal the height of the next lower drawer. The wood for my chest was a 12/4 plank of South Ameri-

can mahogany, 12 ft. by 22 in., and a 4/4 mahogany board, 40 in. by 21 in., with secondary parts of poplar. For effective
use of grain, the symmetry of the ends and the continuity of the front are the most important considerations. I laid out the

ends book-fashion, with the bulge toward the wider annualring pattern (figure 2). Either face of the plank can be used as
the outside; both elliptical and hyperbolic annual-ring patterns are beautiful. I chose the bark side of the plank, producing a hyperbolic pattern at the bulge, as shown in the photo at left. De Rham's desk shows the characteristic ellipti-

cal pattern of the heart face, best seen in the photo on p. 56.
To avoid conflict between the long grain and the cross grain around the case, and to eliminate applied moldings, I departed from traditional construction. I used a thick mahog-

any bottom with the base molding cut into it. Thus the end

Vivid grain patterns are exposed when thick mahogany is shaped. Making the board's bark side convex yields a hyperbolic

base molding is end grain, but so is the molding on the top's

produces elliptical patterns, shown in the photo on p. 56.

figure, as in the author's chest, above; cutting into the other side

end, and there's a lot of end grain in the serpentine front

molding as well. I especially like end-grain molding.

To shape the chest, first rough-saw all the parts according to

illustrate some operations with photos of Jerry de Rham building his bomb desk at Boston's North Bennet Street the rough stock list on p. 53. Note that the final dimensions differ considerably in many cases, but the parts should be cut

School, where I teach. His version is of the basic bomb form: the front is not serpentine, but bulges to match the ends.
It's unclear how early cabinetmakers made the shaped

oversize to ensure that they can be shaped with the setups shown in figure 3. Next, rip the front pieces for the three
lower drawers at the angles shown on the side-elevation drawing in figure 1 so that they can be canted to provide the

drawers, but it probably was done by trial and error, then angle blocks and patterns were made for future reference.
There are graphic methods for figuring the angles, and mathematical methods are quick and accurate, too, as explained in

necessary thickness for the serpentine shape. Mark out the

rails from the centerline and bandsaw them to shape. I shaped the front as a unit, the method I recommend for

the box on p. 57. The same techniques can be applied when

designing anything with canted sides and ends, such as a cra-

any serpentine or oxbow casework. Mount the drawer fronts

and rails on the benchtop jig shown in figure 3A, made with two 2-in. wide supports cut to match the rail and drawer-

dle, dough box, or splayed-joint stool or table.

The first step in any project of this scope is to make fullsize orthographic drawings, primarily to facilitate making patterns for shaped parts. This also helps you work out joinery dimensions, and preview the actual size and look of the piece. In developing drawings, I like to gather information from all the sources I can find. I know of four original chests similar to mineone was privately acquired through Israel Sack & Sons, and the others are at the Museum of Fine Arts in Boston, the Rhode Island School of Design, and Winterthur. Measured drawings of the chest at RISD can be found in Masterpieces of Furniture by V.C. Salmonsky (Dover Publications). After building the chest, I revised my drawing by adding in. to the bottom drawer height. Because this drawer recedes from the common viewing angles, it appears narrower than it is. Usually I follow the rule of thumb that drawer height plus

front profile. I tack-glued the parts to each other at the ends and added two bar clamps for support during shaping. Using the full-size patterns, trace profile shapes on all four edges of the assembly. These lines, with the bandsawn rails, are your guides for the compound curves. With a large, shallow gouge, I first roughed out the concave areas and then the flat fields at the ends. Now spokeshave to the profile lines, using a square from the end surface to check the front. I used a bandsawn three-dimensional pattern, shown in figure 3C, to draw the line of the corner in to where the flat fields meet the serpentine shape. The rest of the front was shaped from this line. I did most of the gouge work across the grain, following up with spokeshaves, cutting from high to low in various directions. I sawed an inch off the handle of my No. 151 round-bottom spokeshave to reach all the concavities. The front should be symmetrical and free of lumps,

Most of the major parts were cut from a 12/4 South American mahogany plank, 12 ft. by 22 in. The case ends were laid out book-fashion with

the bulge toward the wider annual ring pattern.

Fig. 3: Shaping the curves

A: Contouring the front A benchtop jig positions all the drawer fronts and rails together so that they can be shaped as a unit The drawer fronts shown here are
in various stages of completion. The two supports are cut to conform to the slope of the backs of the drawer blanks, shown in Fig. 1,

and are notched to receive the rails.

about in. short of the profile lines traced on the edges. These parallel sawcuts allow the waste to be removed quickly with a wide chisel. The

B: Shaping the ends Carcase ends are rough-shaped by making tablesaw cuts every 2 in.
contours are smoothed with planes, spokeshave and scraper.

C: Matching the curves

allows you to draw a fairly accurate pencil line over the contoured surface to define the corner where the case ends meet the front.

Three-dimensional pattern, bandsawn to match chest outline in Fig. 1,

smoothly curved in all directions. The final scraping and adjusting of the field lines can't be done until the case and drawers have been assembled, yet before the cock beading is carved. The gouge work goes quickly and is fun. It's important to stay relaxed, and I try to keep a rhythm to my mallet

side of the half-dovetail with my shoulder plane because I

don't have a dovetail plane that will taper the angled side, as is usually done. Otherwise I used traditional methods of sawing, wasting with a chisel, and cleaning up with a router

blows. Hollow the backs of the drawer fronts individually, but leave enough flat area at both ends to pass each piece
over the tablesaw, to mark out the dovetails, and to check for any movement after shaping. If there is any winding or other movement, first plane it out of the back of the drawer front, thus making a reference surface for correcting edges and ends. Now put the front assembly aside, and turn to the case

plane. When drawn home, these joints establish the width of the case, and thus locate the mortises in the base. I made the case bottom from seven pieces, 36 in. by 3 in.
by 1 in., ripped from waste sections of the plank. The thickness lets you shape the base itself, rather than having to use applied moldings. The moldings on the top and base are cut on a spindle shaper, as are the rabbets in the ends. I grind my own

ends. Cut them to their finished width of 20 in., but leave the length and thickness rough. With the flat patterns, trace
the shape of the long-grain edges. Rough out the contours by making tablesaw cuts about every 2 in., as in figure 3B. I stopped the cuts about in. away from the profile line to

cutters. Nine tenons join each end to the base, allowing a lot of grain to run through to the molded edge. There is a short-grain problem inside the case ends between the tenons, so leave extra
wood here until after glue-up to prevent a crumbled edge. Rails on 18th-century Boston cases are typically 4 in. to 6 in. wide, with secondary wood often joined to the primary

allow for the inevitable movement: the top of my ends cupped outward, though I suspect that the bulge prevented significant cupping at the bottom. The sawcuts allow the
bulk of the waste to be chiseled away quickly, after which it's best to let the wood settle for a couple of days. Then plane the inside areas near the top and bottom edges flat and paral-

wood. I rhink that 3 in. to 4 in. is wide enough to keep the rails straight. Assemble the case dry, position the rails (aligning the center marks), and cut them to length. Slide the rails in on the stub tenons, scribe the stepped dovetails and cut the mortises for the runners in the rails (figure 1, detail A). The

lel, and retrace the pattern. Now scrub-plane across the grain
down to the line, and finish up with smooth plane and

case is ready to be glued up. To avoid friction between parallel edges, the dovetails joining the top and ends must be slid in individually before the rails and base are added.
Glue blocks or screws usually support the runners in bomb cases. To provide support while allowing movement, I used a vertical strip of poplar, notched to support each runner, at the back of the chest. Stub tenons hook these two

spokeshaves. Place the three-dimensional pattern over the

contour and trace the profile of the long edges. Bandsaw the ends to shape, and true up the top and bottom edges.

To join the case, first attach the ends to the top with housed
tapered dovetails. When fitting the joint, I tapered the square

vertical supports to the base, and screws through slots hold

them at the top. The runners are screwed to the case ends

through slots at the back.

tal boards with overlapping rabbets. The narrow boards are

screwed into the rabbets at the case ends, but the wide boards are free-floating and can move. The grain runs vertically in the feet and their ears. I bandsawed each foot with one ear as part of it and the other ear applied. This uses a little more wood, but it eliminates half the glue joints, and the grain match, of course, is perfect.

De Rham used oak supports

and fastened them both top and bottom with screws, as

shown in the photo at right.

I didn't glue the runner tenons in until later, after some

finish had been applied to the inside of the case.

The feet are stub-tenoned to the base, and I put a screw

through a sloppy hole in each ear into the base. The feet are carved as large as possible in the 12/4 pieces. (For an

Bevel dovetails are needed

for the lower three drawers,
since both their sides and

article on how to carve ball-and-claw feet, see FWW #10, pp. 58-59.) The side toes are angled slightly forward, but the
tips of the claw extend to the diagonals of the square blank. I

fronts are angled. As discussed in the box on the facing page, cut the drawer parts to length at the proper end and bevel angles, and
then before shaping the draw-

also keep the knuckles of the rear, side and front toes different distances from the floor, to avoid a box-like appearance. The bones of each toe get progressively longer as you go back

Notched vertical members support drawer runners at back.

from the nail, and the number of knuckles is anatomically

correct. The ball itself and the claws are smooth, but I left tool marks on the rest of the foot up to the ankle. The case

er sides, cut the dovetails. Next, to fit each drawer front into
the case, position it over its opening and scribe the curve of the case ends onto the dovetail pins from inside the case.

is designed so that the line formed where the flat field ends
and the serpentine shape begins runs across the base molding and around the transition as a miter corner, and ends at

Trim the pins to the scribe line to get the drawer front started in the opening and to check the fit. Then trim the rest of the
drawer front to the pins. The grooves for the drawer bottoms are cut on the shaper wherever possible; the rest is done by

the point representing the fetlock. I think knee carving suits

this design if the mahogany is highly figured. But if the wood is straight-grained, as mine was, then a less fancy style

hand. Assemble the drawers and plane the drawer sides down to the curve of the drawer-front ends. The drawer bottoms
are solid wood, with the grain running from side to side and

is better. Bomb chests deserve the best traditionally made hardware: I spent more money for the hardware than for the lumber. My thin cast brasses with separate posts are from Ball

three sides beveled from underneath to fit the grooves.

With the case on its back, block up the drawers flush with

and Ball, 436 W. Lincoln Highway, Exton, Pa. 19341. I

shaped a pine block to help bend the plates. To seat the posts
properly, I drilled post holes perpendicular to the tangent of

the rails. I established the line of the corner between the flat fields and the serpentine shape by running a compass along
the curve of the ends. Scrape the entire front to smooth all the contours, taking care to leave the flat fields at a uniform

the curve, then I adjusted the bails to fit the posts. I used #0
by -in. round-head brass screws to attach the keyhole escutcheons, so they are easily removable. All the locks needed

width that is crisp and clean. Then, after removing the drawers, use a scratch stock to make most of the cock beading on the rails and the case ends. The scratch stock will have to be adjusted for the middle rail, because of the angle of its face. The beading at the corners is carved. The back of the chest is made of wide and narrow horizon-

1 in. to the selvage. I used a slant-top desk lock on the lowest drawer and made strike plates for all the locks. To finish the chest, I gave it one very thin wash coat of orange shellac, to set up the grain for its final sanding, then
used boiled linseed oil. If applied in very thin, hand-rubbed coats, linseed displays the grain with depth, clarity and

warmth. I don't thin oil greatly with turpentine (never more

than 1 part turpentine to 20 parts oil), nor do I apply soaking coats. I don't think these methods significantly increase penetration, and I suspect that not all of the oil oxidizes, so you risk bleed-out problems. I do add a little Japan drier. I store my oil in a colored glass bottle placed in direct sunlightI

think this helps polymerization and drying. It is most important to apply the oil in the thinnest layers possible and to give it adequate time to oxidize between coats. Each coat should be rubbed hard to build up enough heat to force the oil into the pores and to level the surface. Carvings and moldings should be brushed vigorously to remove excess oil. Instead of

waxing, I prefer to build up the oil slowly to a high gloss.

Like most 18th-century furniture, each Boston bomb
piece is a complete design in itself, independent of its environment. It has character and warmth which are a joy to live with. Its shape continually invites you to run your hand over its curves, or even to tickle its carved feet.

Jerry de Rham, a student at North Bennet Street School, scrapes

his desk front smooth. Note the alternate traditional designthe front is not serpentine, but matches the curve of the sides.

Lance Patterson is a cabinetmaker and shop instructor at the North Bennet Street School in Boston, Mass.

How to make slope-sided boxes

Plans for slope-sided boxes, such as the drawers in the bomb chest on p. 52, aren't in the perspective we're used to. In a front-elevation drawing, the front measures less than its true height because it is tilted out of the plane of the drawing. Here is a method for reading tilted plans, laying the pieces out, and setting the tablesaw to cut the elusive angles involved. In this particular hopper, the front and the back could be cut with the same saw settings, but for clarity, let's consider just the front.

The ends of hopper pieces are cut at compound angles by tilting the sawblade and setting the miter gauge. You can use trigonometry to calculate the tangents of the angles, then from the tangents set a bevel gauge with which to set the saw, as follows.

1. The side elevation, above, shows the true cross section of the board that will be the front of the hopper. Use a bevel gauge to transfer the edge angle to the tablesaw, then rip the front to width.

4. As shown in the plan view, above, the cut has to be at 90 to the beveled edge of the workpiece, not at 90 to the face. To determine this tilt angle, place a carpenters' square flat against the beveled edge of the workpiece and crank the blade over until it lies tight against the arm of the square.

2. The end angle shown in the front elevation is distorted, but the drawing does show the true length of the top ana bottom edges. Measuring out from a perpendicular reference line, transfer the edge lengths to the workpiece. Then connect the edges to draw the true end-cut lines. This method also works for asymmetrical pieceswhere end-cut lines are at different angles.

5. Lower the blade to a safe height and cut to the line. If the piece is symmetrical, you can switch the miter gauge to the other side of the blade, turn the work over, and saw the other end. If asymmetrical, repeat steps 3 and 4. Repeat all five steps for the other sides.

3. With a straightedge, set the miter gauge so the end-cut line is parallel to the blade.

The problem of the hopper joint is traditionally solved by projection and measurement on full-size drawings. Lance Patterson derived the mathematics at right from the graphical method, then picture-framer Jim Cummins devised the no-math method shown above. For a photocopy of Patterson's mathematical proof, send a self-addressed stamped envelope to Hopper, c/o Fine Woodworking, Box 355, Newtown, Conn. 06470.

Chippendale Sofa
Templates for the basic frame, and some design options
by Norman L. Vandal

Any professional woodworker has to keep an eye on the market. Over the years I've earned a living making things that simply weren't available anywhere else, everything from period architectural components to period planes for restorers who wanted to stick their own moldings. Many of my furniture customers come to me because of the double jeopardy of buying antiques: originals are not only very high in price, they may also be in very poor condition. In January 1983, for instance, a Philadelphia camelback sofa sold at Christie's, New York, for the record price of $264,000, even though it had some serious problemsamputated leg bottoms had been pieced in, the rear legs had been cut off and refastened, and the stretchers had been replaced. Keeping all this in mind, I thought it a good idea to add a camelback sofa to my designs. I wanted to stay faithful to the lines

and solid joinery of the originals, so I studied Chippendale sofas in museums and period-furniture books. Surprisingly, my best source turned out to be copies of sell furniture stripped of its upholstery to ensure buyers that it's original, and many of the ads showed the entire frame and the joinery. I saw that period cabinetmakers varied the shape of the legs, front seat rail and crest rail without changing the shape of the basic frame much. I figured that I could do the same for my customers, and build a good frame to sell for a little over a thousand dollars, which compares favorably in price with factory "reproductions," and, in my opinion, greatly surpasses them in quality.

The Magazine Antiques. Dealers like to

Templates and variationsJoinery details are shown on the facing page. Sofas are not as difficult to make as they may

Adapting the basic templates yields this Philadelphia-style sofa with serpentine front.

look. Unlike upholstered chair seats, which are always trapezoidal, sofa seats are rectangular. Thus a measured drawing of the frame's end view shows many parts in true dimension. With these parts as a starting point, I worked out a reliable set of templates, shown throughout this article, for the angled parts. The templates take care of the tricky problems, ensuring that everything will go together and stand square. When building a sofa, you first make the end frames, which include the legs, end rails and side stretchers. Then you connect these with the seat rails, center and frame stretchers, and back frame, and finally you add the arms. In period sofas, there are variations in the arm roll and its supports, and I selected the system I felt worked best. The templates given here are for a New England style sofa with Marlborough legs, which can be blocked or left plain. Straight-leg sofas were the most numerous, exemplifying the Chinese influence in the Chippendale style. Yet the molded leg and the cabriole leg shown on p. 65 work just as well. For the Philadelphia look, as shown at left, the variations are simple: Marlborough legs, peaks on each side of the crest, and a serpentine front seat rail. You'll also find templates for the crest rail, vertical arm supports, and upholsterers' bar (the upright member underneath the arm at the backit gives the upholsterer a surface around which to pull and tack the material). Scale up the templates to full size, either by following the grid lines or by photo-enlarging them. I made the templates from heavy cardboard so I could cut them out and trace the parts directly from them. As you can see in the drawings, on my templates I've carefully laid out mortise and tenon dimensions and other useful information. You don't need templates for the front and rear seat railsjust mark them out

directly on the stock. (If you plan a serpentine front seat rail, of course, you'll have to work out a full-scale top-view template for the curve.) The center legs,
front and back, fall exactly in the middle

of the rails. As shown in the photo on p. 63, the front seat rail is one piece; the rear seat rail is two pieces, each tenoned

into the back center leg, which needs to be full length to support the center of the back. In addition to the low stretcher between the center legs, an upper stretcher prevents the frame itself from spreading.
This frame stretcher (which will be hidden by the upholstery) is tenoned off-center

into the seat rails, so as not to weaken the legs. Locate it toward the bottom of the rails, where it will not interfere with the setting of upholstery springs. Original sofas didn't have springs, just webbing,
but your upholsterer may suggest that the modern method is better. For more on upholstery and whether to agree with your

upholsterer or not, see p. 64.

The tenons on the side stretchers have only one shoulder, at the outside. I make

the center stretcher the same way, and offset the mortises in the center legs so the

stretcher will be centered.

MaterialsMost of the frame will be

hidden by upholstery, and period cabinetmakers knew this full well. Legs, which showed, were top-grade wood. Mahogany

predominated in Philadelphia; cherry was

peculiar to Connecticut. Walnut was used in high-style pieces from all areas, and you'll find that the finest sofas, with formal Marlborough legs or ball-and-claw feet, are always mahogany or walnut. But
secondary woods are another matter. I've seen seat frames made of maple and oak, and even chestnut in some New England

examples. The back frames are usually of a softer wood, sometimes pine, although yellow-poplar or basswood holds the

tacking better. If you can find it, soft maple is an excellent wood for the frame. Whatever you use, test some scraps, and
avoid any wood you can't easily drive a tack into, or one that won't hold it well. Arm rolls are always a soft wood, and pine or poplar is suitable. The vertical arm supports should be hardwood, but avoid

woods such as oak or ash because they

may split when the upholstery is tacked

on. Curly maple would be my first choice here (somebody once suggested plywood, but it doesn't hold upholstery tacks well).

ConstructionMany of my construction notes are shown on the templates,

but here are some additional hints. The tablesaw jigs shown in the photos
below will help when cutting the back

plates, and shape the legs and side stretchers. Then permanently assemble the end frames. The arm rolls, their support blocks and the vertical uprights will all be added later. Secure the tenons with pegs. Trial-assemble the end frames to the front and rear seat rails and the center parts. Do any fitting of the joints now, making sure that the tenon shoulders are square and the mortises true. This will ensure that the frame assembles squarely when you're gluing up. Muster all your speed and dexterity and glue up the frame. I don't install the medial stretcher yet, but measure it off the frame and slip it up into the other stretchers from below as one of the last steps in construction (it has

legs to shape. They ensure that the legs will match each other exactly and that the straight sections will be true. Begin construction with the end frames. Before test-assembling them, cut all the
joints shown on the end-rail and leg tem-

dovetails at each

stand square.

Tablesaw jigs position each rear leg exactly the same, ensuring that the frame will

Getting a frame upholstered

First, the bad news: a good upholstery job can cost as much as $1500, and you may have trouble finding a shop that will do it right. Now the good news: if the job is done rightand I'll tell you how to be sure that it isit will easily last 15 years.

by Bob McCarthy
tures of a piece. Damasks were popular on

Springs: Springs weren't used in upholstery until the mid 19th century, but they lend support critical to appearance and comfort. Well-tied springs should last for
years, webbing alone simply will not,

period sofas (a damask is a woven-pattern material, usually with floral motifs, whose
and dull surfaces). Period damasks were

design is accentuated by alternating glossy wool or silk. Many fine reproduction darnasks are available today, in wool, silk or
synthetic blends Another good fabric

working have both changed a lot since 1780 A true reproduction sofa frame would not have screws, modern glues or upholsterers' bars. There wouldn't be a machine mark anywhere, nor any trace of sandpaper. And a period upholstery job would have no springs or cushion, and would be stuffed with Spanish moss or horsehair. To most people, such a sofa would be very uncomfortable.

The period sofa: Upholstery and wood-

which is why I recommend springs even though they aren't authentic. For the seat,
I would insist on coil springs, hand-sewn

to the webbing and hand-tied together. For the back, I'd ask for Marshall spring units (light, muslin-covered springs), Padding: Instead of horsehair, cotton
batting commonly is used today Period materials are hard to acquire, will increase your costs, and won't show anyway Make certain that muslin is used to

fabric of a solid color embellished with a subtle embossed design. If you're fortunate, you may even find a decorator with
some leftover fabric (designers often buy excess material as insurance against running short of a particular dye lot). I've bought such bolt-ends for a quarter of

choice would be wool moreen, a heavy

and provide comfort at the same time. As you read on, keep the following basics
in mind: A Chippendale camelback sofa should be padded very tightly and never

Yet modern upholstery methods can recreate the period looktaut, crisp lines

hold all padding in place. Seats: If you want a traditional fabric, you should specify a tight seat, which means one with no cushion. This will look best,
and avoids the problem of a cushion that won't stay put, but of course it "wears faster. If you want a contemporary fabric, then a single thin cushion wouldn't look

their normal price. Have the upholsterer pad the rolls

slightly to accentuate crisp curves. Ideas

for piping, decorative brass tacking and

other traditional variations can be seen in antiques books and museums.

overstuffed. The back should not be padded too thickly or it will push the occupant forward. The seat should slope slightly from front to rear to hold a cushion, if used, in place. Do not allow staples anywhere; someday your sofa will be reupholstered, and staples are difficult to remove without breaking them, which leaves razor-sharp studs sticking up.

If all attempts at locating a qualified local upholsterer fail, do not despair. Learn-

bad. The cushion's box (the distance between the edge pipings) should be no more than 3 in. Cover the cushion on both sides so it can be flipped over. Zippers on
the back prevent you from flipping it four ways, but are hard to talk upholsterers out of. Use down filling if you can afford it.

Finding the right shop: Many shops

specialize in reupholstenng and are not qualified to tackle a bare frame, but any
large city probably has a shop that can do decorators, and keep looking.
Request a list of references from the

Fabric: A 6-ft. camelback sofa with a cushion requires 10 yards of 52-in. wide fabric. A material without a pattern can be "railroaded," that is, run horizontally thus saving some material.
Documentation for period fabrics can

nial Williamsburg, Box CH, Williamsburg, Va. 23187, Historic Charleston Reproductions, 105 Broad St., Charleston, S.C.

There are many books on the subject in libraries and bookstores. Few tools are requiredmostly patience. Sources: For traditional fabrics, try Colo-

ing upholstery is not all that difficult.

29401, Brunschwig & Fils, Inc., 410 East 62nd St., New York, N.Y 10021, Cowtan & Tout (chintzes), D&D Building, 979 Third Ave., New York, N.Y 10022, and
Stroheim & Romann, 155 East 56th St.,

your job right. Ask a nearby museum for recommendations, check with interior

be found in old advertisements and, sometimes from remnants uncovered during

reupholstery Period fabrics are readily available, and I've listed a few of the best

New York, N.Y 10022. For contemporary fabrics, contact Gretchen Bellinger Inc.,
330 East 59th St., New York, N.Y 10022, and Hasi Hester, 138 South Robertson

shop, then take the time to go and look at some of their work. A good shop will cooperate with you in making your sofa what it should bethey will allow you to specify materials and methods, and will put the agreement in writing.

be prepared to spend $20 to $50 a yard.

suppliers here. If you have a business letterhead, try to get wholesale prices. Still.

Blvd.. Los Angeles, Calif. 90048.

It's poor economy to save on fabric or its support, as these are the most obvious fea-

Bob McCarthy upholsters period pieces

and teaches adult-education courses on
the subject. He lives in Columbia, S.C.

end and a lap joint in the middle). Next make the back framethe specifics are

on many of the original frames have split.)

shown on the crest-rail template. I prefer a

single board for the crest rail, as on the originals, but you could glue it up. Size the materials, cut the joints, test-assemble, then glue the frame together. Peg the joints

With this much of the frame assembled, you can go on to the front arm supports, which consist of a curved vertical
upright and a support block. I first cut

saw with the rip fence as a guide, sawing them in length and thickness to fit. Then

I bandsaw the block's top curve.

Bandsaw the curved uprights and bevel the bottom edges so they set flat on the end rails. Glue and screw the uprights to

the support block's straight edges, either with the tablesaw leg jig or with the saw's

and cut the crest profile with a saber saw

Glue the back frame to the leg uprights, and, when dry, fasten it securely with #10 or #12 wood screws, driven

miter gauge. The block's joints into the

leg post and the end rail are critical: a tongue-and-groove at the front leg allows you to slide the block down until its tenon

the blocks, then fasten the assembled units to the frame with glue and pegs.
For the arm rolls, I always use clear pine, laminated from two pieces of 8/4

through the softer wood into the hardwood legs. (Period cabinetmakers usually used
clinched nails, and for this reason the legs

fits in the end-rail mortise. If these joints

are not precise, the arm will soon fail. I cut the tenon and the tongue on the table-

stock and one piece of 4/4 Period cabinetmakers used solid pine blocks, but
nowadays these are hard to get. The arms meet their supports at compound angles,

both front and back. It's best to cut these angles before shaping the arms. Trace the angles from the side-view and topview templates on the blanks, taking care
that the left and right arms will be mirror images, then cut the angles. You could set up a bandsaw for these cuts, but I find a fine-toothed handsaw easier. Test-fit the blanks, truing up their ends with a low-angle block plane if necessary With the blanks in place, trace the circles of the crest rail and the arm supports on their ends as far as you can reach with a pencil. Then remove the blanks and use the two circle templates to complete the end shapes. Bandsaw as much waste as possible, then carve the rolls to shape. I use a drawkmfe, spokeshave, carving tools and planes.

Variations on a theme
The Chippendale sofa frame I've designed the end-rail length, front-seat-rail shoulder

is a foundation that can accept many stylistic variations. For instance, I made the
classic Philadelphia-style sofa shown in

distance and stretchers to accommodate

the smaller post. If you'd like to try this

the photo on p. 60 with Marlborough

legs, a serpentine seat rail, and peaks on

yourself, you'll need to work out gains,

chamfers and other minor changes at the front corners so the parts fit neatly. The

each side of the crest. The sofa frame

shown in the photo on p. 63 is a simpler, New England design with molded legs.

arm-roll length also shortens by

in., but

this takes care of itself during truing-up. All the following variations are found

frames, fabric was tacked directly on the part of the frame it was covering. Most modern upholsterers prefer to pull their material through narrow openings in the frame and tack it down on the back side.
On our sofa, the lower back rail is higher than the seat rail, and provides such an opening there. The upholsterers' bar shown on p. 61 provides another opening at the junction of the sides and back frame. Although these bars aren't authentic, they add strength, and a frame with bars is easier to upholster. The bar fits into a gain in the arm, also shown on p. 61. Fair the edges of the bar to the shape of the arm and ease them so as not to strain the fabric. Then relieve all the other sharp milled edges of the frame with a file so the fabric will lie over them smoothly

Preparing for upholsteryOn period

Some of these modifications can be accomplished with very few changes in the basic templates. A serpentine front rail, for example, requires a curved template and affects the length of the two stretchers in the center. That's alleverything else can remain the same. Some variations call for more work. If you'd like to change the slope and splay of the arm roll, obviously you'll have to change the template for the vertical arm support as well as the length of the arm-roll templates and their end angles. The arm support block would probably be affected as well. In the leg designs shown below I'm recommending that you choose 1 -in. stock for the front molded leg. This allows you to reduce the size of the leg post above the carving to 1 in., the same size as the post on the Marlborough leg. On my sofas, I do it a little differently, because I like to keep the front and back legs the same overall width. I start with 1 -in. stock, reduce the post to 1 in., then vary

on original period pieces. Legs: Straight, square: straight, molded;

tapered, square; blocked-foot Marlborough,

plain or carved; ball-and-claw cabriole (no

stretchers) Some pieces have eight legs,

but six-legged sofas are more common. common; other lengths are options.
side of hump; varied in curve.

exposed and ornamented. A 6-ft. length is

Seat rails: Straight; serpentine curved;

Crest rail: Single-hump; peaked to each

Arm rolls: Straight, tapered, cylindrical;

serpentine; varied in slope and/or flare.

Stretchers: Plain; beaded-edged; reliefcarved or with open fretwork.

You'll note that I show no stretchers between the back legs, yet you might see them on many period sofas. Rear stretchers take great abuse from climbing children if the piece is placed near the center of the room. Also, they make it difficult to clean under the sofa. They're not needed structurally, and I prefer to leave them out. but the choice is up to you. N. V

The top edge of the crest rail should be rounded toward the front of the sofa. I scribe a line in. down the face, then round over the edge to this line with a drawknife and spokeshave. Don't bring the top back edge to a sharp point. After finishing the legs, I seal the entire frame with a coating of two parts boiled linseed oil and one part turpentine. This helps keep dimensional stability, and it

also improves the frame's appearance. One of my customers, upon receiving his completed frame, liked the look of it so much that he put off the upholstery job
for six months. People like that make the extra touches worthwhile.

Norm Vandal makes period furniture in Roxbury, Vt. He explained how he makes period molding planes in FWW #37 Black-and-white photos by the author

Building a three-cornered masterpiece

by Eugene Landon

Queen Anne Handkerchief Table

Landon ' s repro du ct i o n o f a r ar e 1 8 t h century handkerchief table is a study in pure Queen Anne lines.
n the middle of a very busy workday about four years ago, the phone rang. It was an elderly acquaintance who lived in a nearby town, calling to inquire whether I would be willing to repair a piece of furniture for her. "It's a handkerchief table," she said. I couldn't" leave what I was working on for a week or so, but agreed to go and look at the table when things slowed down a little. Secretly, I had my doubts about the piecethe handkerchief table is one of the rarest American furniture forms. Perhaps what she had was the larger version, usually called a breakfast table, or perhaps something else entirely. I mused about it for a moment, but almost as soon as I returned to what I had been doing, the table went completely from my mind. Two years later, out of the blue, I heard the same voice over the phone again:

"Mr. Landon, aren't you interested in my table?" The earlier conversation came back to me instantly, along with a considerable flush of embarrassment. "I'll be right over," I said. When I entered her home, I saw that she knew what she had all right. The little table took my breath away. Even though it was missing the leaf, the hinged leg and all its knee blocks, it had a presence that epitomized pure Queen Anne, before cabinetmakers under the Chippendale influence began to add shells and gingerbread. I have nothing against Chippendale, but while such decoration may sometimes enhance a piece, it may also serve to disguise basic flaws in design. The little table was so stylistically pure that any such shortcomings would have stood out immediately. As I walked around the piece, I gradually realized that it had no flaws at all; it was perfect. Much as I dislike extravagant claims, I believe that this handkerchief table is not only excellent Queen Anne, but that it is one of the finest pieces of furniture ever made in any time or place. You simply will not find a shaplier leg, nor one more perfectly proportioned to the rest of the table. The genius of the maker is evident everywhere: The four notched corners of the open top serve to restrain the eye's t r a v e l , y e t w h e n t h e leaf is down, the opened notches blend into a lovely curve. The back of the ankle is undercut just the right amount, the merest touch, to give the entire table poise and an irresistible uplifted energy. Even the chamfer on the front corner post strikes just the right balanceit defines the corner elegantly, yet is neither too sharp nor too weak. It turned out that the table had been made in Boston, circa 1740, and that it had been in my client's family ever since the day it was made. The owner agreed to let me make a copy for myself, and I was so

convinced of the rarity and authenticity of the find that I took photos of it to Israel Sack, the antiques experts, in New York City. Robert Sack told me that the firm, in 32 years of business, had handled only one or two similar tables. My client's was truly as rare as I had thought. I made my copy by following plans that I traced from the original. Figure 3 on p. 41, in fact, was adapted from a rubbing of the original table, which I made when I had the top off. The joinery has been added, and also the outline of the legs to show their orientation. I would advise you to redraw this top view full-size, and to add the joinery and wooden-hinge details as well. This step will immediately clarify the project and will also allow you to cut pieces to fit the sizes and angles on the drawing, rather than trying to measure them. My tablemy wife's table, as Jane would remind mestands with its folded leaf against a wall in our living room. This shows the decorative apron on both sides. The table could also go in a corner, with the leaf folded down in front, or it could stand next to an armchair or a sofa, being just about the right height to hold a reading lamp. Because my table stands with its 90 corner facing forward, I'm calling that leg the front one; the folded leaf and the hinged leg are at the back of the table. The table also has a left leg and a right leg, both at the rear. The legs all end up different, so it is important to know which one is which. The front leg is made like a regular Queen Anne leg, following the template shown in figure 1. Indeed, all the other

legs start the same way, but then each

must be modified. The rear legs, for example, are glued to the aprons, then their corner posts and the tops of their knees are reshaped to blend into the acute angle, as shown in the photo on p. 40. As also shown, the knee block of the left rear

Fig. 1: Leg template

Fig. 2: 18th-century

handkerchief table

leg must be modified so that the knee of

the hinged leg can nest behind it. These and other modifications will be considered as they arise.
The first step is to bandsaw all four legs to the curves shown on the template. Notice that the back of the foot is not under-

cient body for good looks. Center the bandsawn blanks on the lathe and turn the bottoms of the feet. Then just nick the back of the ankle to define the top of the
foot. Do not remove too much woodthe nick at the back of the ankle should be no more than in. deep, and your chisel

use rasps, files and scrapers, repeating each step on all the legs with one tool before moving on to the next. When the legs

are shaped and smoothed, set them aside.

The knee blocks will be sawn and added after glue-up, then shaped on the table. Cut the front aprons to length, but don't

cut at this timeto do so is a common

mistake that affects the visual balance of

should barely graze the top of the foot at

the front. These marks will be faired into a gentle curve as the leg is being shaped. I

bandsaw the curves yetthat step comes

just before glue-up. At the front end of both aprons, cut the tenons that will go

the leg and leaves the foot without suffi-

into the front leg. Then cut the mortises in

the front leg. There is a -in. shoulder at

the top of the tenonwhich leaves some

wood at the top of the corner post for

tice that this piece has no tenons). Then cut the hinged apron, leaving it about 8 in. overlong for the time being, and make the hinge. It is much like the card-table hinge

strength but there is no shoulder at the

bottom. I begin with an undersized bit in my drill press, being careful not to get too close to the lines, and not to go so deep that the brad-point will leave marks at the bottom of the finished mortise. (I don't like to leave any machine marks on my reproductions, even when they will be hidden.) I complete the mortise with chisels prior to mitering the front tenons on the aprons, which I cut a little short so they don't touch inside the mortises. As a final touch, I undercut the shoulders a little so

I described in FWW #47, but there are a

few noteworthy differences. It has a built-

in stop at 64 , and to make this work, the

hinge pin must be offset more than half

into it (see the photo below if this sounds confusing). You can use the leg template to determine the profile of the curve. Now glue the end blocks to the inner back apron, but don't glue on any legs yet. As you can see in figure 2, the hinged leg's corner post must first be half cut away so the leg can swing under the inner back

the board's thickness from the end. This

leaves some extra wood to bear against

apron. Work on the mortise-and-tenon

joint at the hinged leg until everything

the inner apron, as shown in the hinge details in figure 3. Final fitting of the stop is done by trial and error before the hinge
strip is glued to the inner back apron. As also shown in figures 2 and 3, there is an end block at each end of the hinged apron. One of these blocks is the fixed part of the wooden hinge, and the other can be made from the excess length of the hinged apron. It is by means of the end blocks that the back-apron assembly is tenoned to the legs. There is some careful fitting to be done before gluing up this assembly. First mortise the back legs and cut the tenons on the end blocks, then fit the hinge together. Next bandsaw the lefthand end block so the hinged leg can nest

fits, paring back the shoulders on the

hinged apron's tenon so that the leg ends

up in exactly the right nesting position. Finally, glue up the back-apron assembly, including the hinged leg but not the others. Now on to the a n g l e d mortise-andtenon, which is not nearly so difficult to make as it may look. First the tenon: I set a sliding bevel to the angle shown on the plans, then transfer it to the top and bottom edges of the apron blankI always mark such lines with a knife, since pencil lines are loo fat to be accurate. Then I simply bandsaw close to the lines and pare down to them with a chisel, as shown in

that the face of the apron will draw up

snug to the post.

Now we come to the interesting part,

the joinery at the back legs, which is

shown both in figure 3 and in the exploded drawing. The two front aprons are mortised to the back legs at 45. There are two back aprons, one of which carries the

hinged leg.
Cut the inner back apron to length (no-

the photo below.

The mortise is a little trickier, but not

A look at the table's back corner (left) shows how the knee

and post have been shaped to conform to the 45 angle. The on the facing page that both legs are oriented in the same

hinged back leg nests as shown at right; note in the drawing direction. Below, Landon demonstrates the setup for starting

the corner post in a V-block, and presses the leg against his forearm for extra control. Parders is shown at right below.

an angled mortise. He supports ing the tenon's angled shoul-

The first step is to bandsaw close to scribed layout lines. Then finish up with a chisel
and a 45 guide block.

really difficult. You can pre-drill most of the waste by supporting the corner post in a V-block as shown. Using the drilled

Fig. 3: Posts, aprons and hinge

holes as a guide, pare the mortise to full width with chisels. You can protect the very thin area at the inside corner from
splitting off by using the 45 guide block, just as when cutting the tenons. With the joinery cut, it is time to bandsaw the curves on the aprons. The pattern

is centered on the apron and extends as far as the tips of the knee blocks, as shown
in figure 4. After bandsawing the curves, remove the sawmarks with a rasp and chamfer the inner edges with a knife or c h i s e l . The original table shows rasp
Detail A: Fully open hinge

Detail B: Leaf hinge and beading

marks clearly, and the -in. wide chamfer ing is a series of very bold cuts.
Glue the table together upside down on a flat surface, and when it is dry, rub on the interior glue blocks. The front angle on the original table was 88 instead of
Detail C: Corner of top

90. I am not sure whether this was deliberate, so that the table would fit into a corner even if the room were slightly offsquare, or whether it was just one of those
things that happens. My table is also 88 , and if you choose to follow the plans exactly, yours will be, too.

Fig. 4: Apron detail

Now bandsaw the knee blocks and glue them onbut notice that the back knee
block on the left rear leg must be relieved, as was done with the end block, so that the hinged leg can nest inside it. Bandsaw

the relief cut before gluing on that knee block. Also notice that the forward-facing
knee blocks at the back legs are larger than the others so they can meet the posts at a 45 angle. They must also be cut from slightly thicker stock, but these differences will be obvious when it comes time

to make the blocks. Reshape the tops of the knees on the

back legs as well as the knee blocks to fair them back to the apron. Finish shaping the outsides of the back corner posts at

made mine with old planes, but any method will work. Pay particular attention to the location of the hinge pin, which determines how the leaf will align with the top in both the hanging and upright positions.

Brite pad. I smoothed the bottoms of the

feet by rubbing them with a brick, duplicating the moving around that the original must have experienced in its lifetime. For

the finish, I applied a home brew of green

walnut husks steeped for a month or so in water. This helps darken Brazilian mahogany so that it looks more like the Cuban variety used by 18th-century cabinetmakers. I sealed this with a brushed coat of thin

the same time. Next chamfer the outside corners of the front leg and the hinged leg, then go ahead and make the top.
I cut the top's molded edge with an ogee plane and a hollow plane, but if you

I attached the top with rubbed glue blocks and nails, the same method used
on the original. This allows no provision for seasonal wood movement, and you could fasten your top differently if you'd

don't have these you can begin by cutting a shallow rabbet and then finish up with files. The notches are cut with bandsaw and chisels.
There is a curious joint where the leaf
folds. It is not quite a rule joint, nor the

like. Some old pieces eventually split, and some did not. My table, in fact, has a nice small split already, which I welcome
as a sign of age. The original's top, ironically, is still fine after more than two hundred years. I don't like to think of it as faking, but

shellac. Five or six subsequent shellac coats were padded on, with some dry pigments mixed in to achieve a semitransparent patina. I took off the gloss with some 0000 steel wool, and everything came together at once. Suddenly there

tongue-and-groove that might have been found 50 years earlier, but rather a more

delicate nesting rabbet-and-bead that does

not conflict visually with the notched corners, whether the leaf is up or down. I

you could say that my table aged a little faster than the original. I added some wear marks where the original table had
them, then eased the edges with a Scotch-

were two old tables side by side, a gathering of the rarest of the rare.

Gene Landon restores antiques and makes

period reproductions in Montoursville, Pa.

Make a wooden-top 5-string

by Richard Starr

A Simple Banjo

never would have thought to design a banjo if several kids in my school shop hadn't wanted to make one (nothing seems too complicated to a 12-year old). Regular banjos have skin or plastic drum heads, but mine uses a wooden soundboard. It isn't as brilliant or quite as loud, but it has an appealing ker-chunky sound that is lovely for mountain-style clawhammer playing. This banjo's structure couldn't be simpler. There is no fancy joinery or bolts and no bent wood. The soundboard is glued to the rim of the banjo, eliminating the need for complex and expensive tensioning hardware. To make the banjo, a bandsaw is indispensable and a power jointer speeds the work, but you can, as do my students, manage with hand planes. Begin by drafting a full-size pattern for your instrument, both

top and side view. My students used dimensions from the book Foxfire 3, edited by Eliot Wigginton (published by Anchor Press/Doubleday, 245 Park Ave., New York, N.Y. 10167), but you can use the dimensions shown in figure 1, or copy an existing instrument. If you have such a model, be sure to note the following dimensions: length of the neck; width of the neck at the nut, fifth fret and where it joins the rim; height of the strings above the last fret and at the nut. Measure the positions of the frets to the nearest millimeter, using the nut end of the fingerboard as zero. You may choose to make a fretless banjoour design fits in well with the warm, primitive, fretless style of playing. After drafting the basic shape of your instrument, mark out the extension of the neck through the rim by drawing lines parallel

SEE ERRATA AT END OF ARTICLE

to the centerline, starting where the neck intersects the rim as

shown in figure 1. This defines the width of the whole neck piece and the shape of the rim halves. Trace the neck and the rim halves onto another piece of paper and cut them out as patterns to trace on the wood.
FrameSturdy native hardwoods like maple, ash, cherry and hickory were used by mountain instrument makers. I made my banjo of 8/4 cherry. For the neck, choose stock a couple of inches longer than the entire banjo and a tad wider than the section of the neck that extends into the rim. If your peg head is to be wider than that dimension, add the length of a second peg head to the overall length of the neck blank. Joint one face and
two edges of the stock. As you joint the second edge, bring the

er. This can also be done on a drum sander. The outside curve is best left rough until later. Hold one of the rim segments in place against the neck and
trace a line on the neck where the bevel meets the neck. Cut out the relief section of the neck as shown in figure 2. Bandsaw the neck and peg head to shape. Glue the rim segments to the neck as shown in the photo below. When the glue is dry, check to see that the top edge of the rim is exactly even with the face of the neck. If it's not, true it up with a hand plane. Trim the tailpiece extension of the neck flush with the rim. Plane the back surface of the rim and neck so that they're smooth
and flush with each otherI planed off about in.

SoundboardQuartersawn spruce, cedar or redwood gives a

stock down to its finished width. If you're widening the peg

head, cut the extra section off, rip it up the middle and glue the jointed edges of these two pieces to the sides of the neck. The two C-shaped segments that form the rim are made from stock the same thickness as the neck. Joint a face and an edge,
then trace the pattern with the ends of the C flush against the jointed edge. Bandsaw the C-shape and save the outer waste pieces to use later as gluing cauls.
Now bevel the inside edge of each rim segment. On the end of each segment, draw a line as shown in figure 2. This line establishes the 45 bevel. Use a compass to scribe a line around the inside of the rim as shown. Set the table of your bandsaw to

rich, resonant sound. An easy way to get a decent soundboard is to pick through a pile of spruce or cedar clapboards at your local
building-supply house. Look for annual rings that are perpendicular to the faces of the board. Cut two lengths of clapboard a couple of inches longer than the soundboard, joint the edges and glue them together to get the width you need. When the glue is dry, plane and sand the board a tad thicker than in. Trace the inside and outside edges of the rim on the bottom
side of the soundboard. The soundboard is round, except for a flat section where it butts up against the fingerboard. Be sure that

45 and carefully follow that line to guide your cut. Smooth the inner surfaces of the rim segments with a spokeshave and scrap-

the grain of the soundboard runs parallel to the neck. Bandsaw the curved shape in. outside the traced line, but cut right to the line on the flat section.

Cut a brace from spruce or pine as shown in figure 1 and glue

it across the underside of the soundboard. This cross-grain brace helps resist the downward pressure of the bridge and reduces

the chance of the soundboard splitting. Now you're ready to glue the soundboard to the rim. Hold the
banjo in a machine vise by the part of the neck that passes through the rim. Spread glue on the upper edge of the rim and set the soundboard in place. Be sure that the flat section of the disc is lined up where the neck joins the rim and that the soundboard overhangs the rim evenly all around. For a good glue job, it's important to apply gentle clamping

wood roughly the size of the fingerboard as a caul to distribute the clamping pressure. Now bandsaw the sides of the neck to match the shape of the fingerboard.

Carving the neckCarving the back of the neck is probably the

fussiest job in this project. The curve in cross section must be rounded almost to the top edge of the fingerboard, while the shape along the length is almost a straight line. At the same time,

pressure at every point on the rim. When the glue is dry, use the bandsaw to trim the soundboard flush with the rim.

FingerboardA fingerboard that contrasts in color with the neck of the banjo is appealing. Mountain folk used native woods: walnut, cherry or maple would work well. No part of the instrument receives more wear and tear, so very dense woods are bestthe finest banjos have ebony or rosewood fingerboards. Cut a piece of stock slightly wider than the neck and a couple of inches longer than its length. Joint one face and one edge, then thickness the piece to about in. Pencil a line up the center of the stock and lay out the shape of the fingerboard centered
on this line. If you want frets, mark the position of each fret along the jointed edge of the stock, then use a square to project each mark across the fingerboard. Fretwire has a T-shaped cross section and the shank of the T is jammed into a slot in the fingerboard. You'll need about 5 ft. of fretwire and a fretsaw, or a dovetail saw whose kerf gives a snug fit to your particular fretwire. If the saw cuts too wide, you can narrow the kerf by sliding a file lightly along the sides of the

the neck gets slightly wider and thicker from the nut to the rim and is faired gently into the peg head and heel. Rough out the shape with a spokeshave and refine it with a scraper or file. Sandpaper on a hard block works best for truing the surface lengthwise, A well-shaped neck is a musician's joy, so examine your work with your fingers, as well as your eyes. It's helpful to handle

a completed banjo to get an idea of how a neck should feel.

FrettingInject a small amount of white or yellow glue in the fret's kerf and tap in a length of fretwire using a lightweight, deadblow hammer. Be sure to support the neck with a block of wood directly under the fret you're installing. Using a pair of nipping pliers, trim the overhanging ends of the fretwire. File the
ends even with the fingerboard and round them slightly at the

top, so no sharp edges protrude. If you file the wrong way, you'll
lift the fretwire. To remove the file burrs, sand the edges of the fingerboard with 220-grit paper on a hard block. Run a long file lengthwise up and down the neck to level any high frets. Now you can smooth the outer edge of the rim with a spokeshave and sandpaper, and finish-sand the whole instrument.

saw, reducing the set of the saw's teeth. Don't make the kerf too
tight or the neck will bow when you hammer all the frets in.

Guide the saw against a block of wood clamped to the fingerboard, as shown in the photo below. This block can also serve as

Set-up Install tuning pegs according to the manufacturer's instructions, or make your own tapered friction pegs. To guide the fifth string over the fifth fret, we cut a simple notch in the fingerboard. You could also insert a small round-head wood screw between the fourth and fifth frets so that the head of the

a depth stop. Trim its height so that the back of the dovetail saw catches on its top edge when the cut is to depthabout in.

screw holds the string down tight on the fifth fret. Instead of a
tailpiece, we used five round-head brass brads driven into undersized holes at the tail end of the instrument. These brads secure the ends of the strings. Be sure that the heads of the brads stand about in. proud to catch the string's loop. Round the edge of the soundboard slightly where the strings bear on the corner. Fashion a nut and bridge from dense hardwood, and trim their height to give the proper action (the height of the strings at the nut and last fret). The nut glues against the peg head and the end of the fingerboard. File shallow notches for each string. String the banjo and position the bridge so that the 12th fret is midway between the nut and bridge. Adjust the bridge so that holding down each string at the 12th fret produces a tone one octave
higher than the open string. Move the bridge slightly closer to the nut if the octave is flat, further away if it is sharp. Don't glue the bridge to the soundboard. The tension of the strings will hold it in place. An oil finish, wet-sanded with 400-grit wet-or-dry paper, will

deeper than the shank of the fret. Practice cutting frets in scrap
before trying it on your fingerboard. After you've cut all the fret slots, saw the fingerboard to shape

and glue it to the neck. Be sure that the centerline of the fingerboard is true to the centerline of the neck and that the end butts
up against the flat section of the soundboard. Use a scrap of

give you a fine-looking and serviceable musical instrument. A

good book for beginners is How to Play the 5-String Banjo by

Pete Seeger (Oak Publications, Div. of Music Sales Corp., 799 Broadway, New York, N.Y. 10003). Good playing!

clamped to the fingerboard. Trim the block height so that the back of the dovetail saw catches on its top edge when the cut is about in. deeper than the shank of the fretwire.

To saw the fretwire slot, guide the saw against a block of wood

Richard Starr teaches woodworking at Richmond Middle School in Hanover, N.H., and is the author of the book Woodworking with Kids (The Taunton Press, 1982). Photos by the author. Banjo tuning pegs and strings are available from Stewart-MacDonald Mfg. Co., Box 900, Athens, Ohio 45701.

Erratum: There are two errors in the measurements for the banjo in FWW #53. The distance from the nut to the first fret should be 37mm instead of 32mm. The distance from the banjo ring to the peg head should be 20 1/8 , in. instead of 18 9/16 .

Frank Klausz created his ideal work area in a small space by pairing a hefty workbench with a utility table. The bench provides lots of clamping power and the table contains storage bins and drawers.

A Classic Bench
by Frank Klausz

Workstation's center is worth building right

f you are a serious woodworker who prefers handtools, one of your first investments should be a hefty, well-designed workbench. My joiner's workbench, shown in the photo above, is the heart of the ideal workstation. Based on a traditional design, my bench is outfitted with shoulder and tail vises and steel dogs that can clamp a workpiece in a variety of positions. And it's built solidly enough to be stable under any kind of sawing, planing, scraping, or pounding. Near my workbench is a wooden chest with my chisels and other handtools, all sharpened and ready to use. To make it easier to use the chest I built a small platform that raises the box

10 in. to 15 in. off the floor. If your bench is near a wall, you might prefer a wall-hung cabinet, as workmen in Europe often do. A 27-in.-high utility or helping table with a 40-in. by 60-in. work surface is located about 4 ft. behind the bench. This table, shown in figure 1 on the facing page, houses 12 plastic drawers (available from WW Grainger, Inc., 5959 W. Howard St., Chicago, Ill. 60648): small ones for dowel pins or screws, larger ones for chisels and other tools. Larger planes and portable power tools fit on its bottom shelf. Don't try to save steps by putting the drawers in your main workbenchif you clamp a large piece in the shoulder vise, you can't open the drawers. You could build drawers

Assemble the 27-in. high utility table with mortiseand-tenon joints. Locate shelves to fit standard plastic

drawers for small tools and odds and ends. Larger with the shoulder vise.

tools go on the bottom shelf. Make the 40-in. by 60-in.

plane long boards, right, Klausz uses a bench slave

top from particleboard and -in. maple plywood. To

that open from both sides of the bench, but putting them in the utility table is much handier. By arranging my workspace like this, I have plenty of room to

A good bench should be built of hardwood, heavy enough so that you can't move the bench with a stroke of your handplane. Hardwood is expensive, so I cut costs by buying green wood and

work comfortably and can easily step over to get a chisel or a handful of screws. Everything is at my fingertips. The workbench and table also work well together. I do all of my planing, sawing
and joint cutting on the bench, then assemble the pieces on the table. The table, being several inches lower than the bench, is perfect for holding a chair or a chest of drawers at a comfortable

drying it myself or scavenging rejects at local sawmills. The bench legs and base are cut from second- or third-class chunks of
red oak, white ash and beechany hardwood will do. It's not scrap, but it's not good enough for furniture. Though each workbench is a little different, depending on the material you have to work with, don't drastically alter the basic

work height. When I'm assembling on the table, I still have a clear workbench for trimming joints and other last minute touches. Apart from the knots in the base, my workbench looks pretty
much like any other traditional cabinetmaker's bench. Our ancestors invested more than 1,000 years in developing its design and they left very little for us to change. When I worked in Europe, I visited many different shops and the workbenches were always

dimensions shown on the drawings. You could go a little wider or longer without creating a monster, but scaling the bench down and using much thinner stock eliminates the weight essential to a good bench. The correct height of the bench is easy to
determine. Stand up, put your hands next to your pockets and your palms parallel to the floor. The distance between your

the same design and about the same size7 ft. by 3 ft.although the bench height was tailored to the height of the cabinetmaker who used it. Apart from little touches like the stops and oil dish
shown in figure 2, the only difference I found was that some

Fig. 1: Floor plan

craftsmen treat their benches with loving care and some don't. All the European cabinetmakers I visited used similar shoulder and tail vises to hold their work. The bench screw (available from Garrett Wade Co. in New York City and Woodcraft Supply Inc., Woburn, Mass.) on my shoulder vise gives it about a 7-in. capacity.
It can hold a short piece by itself or, working with a bench slave (see figure 2), hold a long piece in an efficient work position. The slave is a notched 1 -in. by 2-in. piece of hardwood tenoned into a cross-lapped base. A wooden block hanging from two wooden

ears connected with a dowel supports the work.

The tail vise can hold wood in the same manner as the shoulder
vise, but it's most often used with the bench dogs to lock pieces

down flat on the benchtop. I use traditional square metal dogs (I ordered mine from Garrett Wade). It's crazy to try to use dowels
for bench dogs. They might work if the dogs just kept the wood

from sliding on the benchtop, but they must also clamp the work
tightly against the top. Square dogs have slightly angled faces so
you can pinch the board between the jaws, then drive the dogs down to snug the piece against the top. A workpiece suspended in midair between the dogs will chatter when you work on it.

palms and the floor equals the bench's height. If you make the bench higher, you can't take advantage of your body weight when handplaning. Using your body weight, not just your arm
muscles, will give you hours of easy planing while the other guy is pushing and shoving.

bored in heavy bearers screwed to the underside of the bench

top. The weight of the top holds it on the base. Begin base construction by determining the height of your bench, as discussed above. I'm 6-ft. tall and my bench is 33-in.
high. Adjust the leg length, up or down, in the area between the stretchers and feet, then cut all the parts as shown in the plan. I cut the mortises with a hollow-chisel mortiser, but you could chop them by hand or mill them with a router. Drill a bolt hole through each stretcher mortise from the inside of the mortise.

Construction of the baseMy bench is supported by a sturdy

base: two heavy uprights joined by a pair of wide stretchers. The

pieces for the uprights are mortise and tenoned; the leg-to-topbrace joints are through-wedged; the others blind. So the bench can be broken down to be moved, the stretchers are fastened to
the legs with bolts and captured nuts. To position the top, bulletshaped dowel pins in the top braces of the uprights fit into holes

Insert the stretcher tenon and use the hole in the leg as a guide to bore into the end of the stretcher. Remove the stretcher and
deepen the hole to accept a 6-in. hex-head bolt. I rout a slot at

the end of the bolt hole to house the captured nut.

Assembling the topThe benchtop, with its tool tray and two
vises, is the most complicated part of the bench so you must

this avoids a lot of sawing and awkward cleaning up later. Since you want to reinforce the shoulder vise with a threaded rod

measure very carefully when making the parts. It consists of

2 -in. thick boards sandwiched between a thick front rail, which is mortised for the dog slots, and the tray and back rail assembly

through the top, as shown in the drawing, remember to bore a

-in. diameter hole through each component before assembly. You can take care of the splines and minor alignment problems

in back. Both ends are capped by heavy cleats. All pieces are
splined and glued. The vises themselves are constructed separately and then fitted to the top.

when you attach the vise. Glue the 2

in. pieces together with

1 -in. by -in. plywood splines, trim the assembly to size, then cut the grooves for the end cleats, as shown in the drawings.

For the 2 -in. stock, I used quartersawn maple, but you might want to jazz up your top by using several different woods. That's
OK if the different species are about the same density and will

Although the glued-up top is big and heavy, you can cut the
grooves by standing the top on end and passing it over your tablesaw's dado head. If this sounds too nerve-racking, use a router.

move with the seasons and wear at similar rates. Lay out the glue
joints so that the notch for the tail vise is created in gluing up

Always reference the top surfaces of the benchtop and cleats

against the fence or router base so the grooves will line up. Next,

Benchtop is positioned on the base by bulletshaped dowels, left. The base itself is lowgrade hardwood. Before assembling the top,

above, chisel out dadoed slots so L-shaped

dogs fit flush with top.

mill the front rail and the bench-dog slots in it. Note that the front rail and tail-vise face must be the same thickness so the
bench-dog slots line up. I cut the slots with a dado head on my radial-arm saw, then chisel the L-shaped notch for the dog's head by hand. After assembly, I glue a backing piece to the front rail to enclose the notches. Test the fit of each dog before you glue

the drawing is 1 in. in diameter by 13 in. long. The tail-vise screw is 1 in. by 17 in. Be sure to have the screw (and all other
hardware) before you build the vise. The tail vise has two partsa jaw assembly and guides fixed to the benchtop. The jaw assembly consists of a heavy jaw and face piece dovetailed together. The jaw houses the screw, the face

up. If they are too tight, it will be hard to trim the slots after the rail is glued to the top. I set the dogs into the bench at an 88 angle, nearly perpendicular to the surface. A greater angle might
increase the dog's down-clamping pressure, but you'd lose the ability to reverse the dogs and use them to pull something apartthe dogs would slide out of the angled slots. I use the dog's pulling ability in my restoration work. If I have to disas-

piece is the same thickness as the front benchtop rail and is likewise slotted for bench dogs. A guide rail, parallel to the jaw, is dovetailed to the face piece and a runner connects it and the jaw.
This assembly is further held together by two top caps, whose top surfaces will be flush with the benchtop. Two guide blocks bolted under the bench are notched for the runners that guide the jaw assembly. The vise-screw nut is housed in the end cleat.

semble a chair that's too fragile to withstand much hammering, for example, I reverse the dogs, fit the chair parts between the padded dogs, then crank the tail vise out until the joints separate.
This technique also works on other kinds of furniture. The tool tray is a piece of -in. plywood screwed to the underside of the 2 -in. top and housed in a groove in the back rail,

I cut the large dovetails on the bandsaw or with a bowsaw. The dovetails are very strong, beautiful and show craftsmanship. Finger joints would work, too. You can cut these on the tablesaw. The dog slots are cut using the same method as on the front rails.

Close off the open side of the slots by gluing on a piece of

-in. plywood after the jaw assembly is glued up.

which is in turn dovetailed to the end cleats. I glued two angled blocks in each end of the tray to make it easier to clean. The end cleats support the two vises. Six-inch by -in. hexhead bolts and captured nuts reinforce the splined glue joints.
The holes are not too long to bore with standard hand or power auger bits. Chisel or rout the blind notches for the nuts in the

To ensure proper alignment, bore the holes for both vise screws on a drill press before assembly. I first bored a 1 -in. dia. hole for the depth of the embedded nut, then, using the same
center point, bored a 1 -in. dia. hole through the piece for the screw. After boring the end cleat, I clamp the tail vise to the bench and use the drill bit to mark the center of the screw hole.

underside of the top. I leave the bolt heads exposed. That goodlooking hex head makes a handy little anvil for blunting nails so they won't split wood, or for tapping out hinges or other hardware. Before you glue on the cleats, however, make the vise
parts and assemble everything dry to make sure it works okay.

Unclamp, transfer the center point to the outside of the piece and bore a 1 -in. dia. hole. Make fine adjustments with a rasp.

After assembling the tail vise on the bench, I close it and, to

make sure its faces are parallel, saw through where the end of
the vise meets the bench with a sharp, fine-point backsaw, being careful to keep the saw between the two pieces. Then I glue topgrain cowhide to each face.

Design of visesI prefer 2-in. dia. wooden bench screws for vises, but they are so rare that most people use metal screws, even though they don't have as nice an action. Tailor your vise to fit the length of the screws you have. The shoulder-vise screw in

The shoulder vise is much more straightforward, but you may have a little trouble with the treaded rod running through the
top to reinforce the dovetail joining the end cleat and vise arm.

Flip-up bench stop is handy for crosscutting near tail vise, top left. Carved oil cup under

vise swings out when you need to lubricate plane sole or saw, left. Underside of bench near shoulder vise, above, shows hardwood bench stop and the track that guides vise jaw.
Since you drilled the top pieces before assembly, you should be able to clear the splines and any misalignments by running a bit
stain. Wax your bench regularly and resurface it every year. I believe lots of people, including customers, look at your bench as an indication of your craftsmanship. Besides, I am spending about 10 hours a day looking at and working at the thing, and it

on a 12-in. extension in from the front and back. Then bore the
vise block and arm separately before attaching the unit to the top. To finish the bench, level the top with a sharp jointer plane, checking by eye, straightedge or winding sticks, then sand with a
large vibrator-type finish sander. I put two coats of Waterlox (available at large building supply houses) on every wood sur-

should be beautiful.

Frank Klausz makes furniture and restores antiques in Pluckeand Wood Finishing, are available from The Taunton Press. For
min, N.J. Klausz's two videotape workshops, Dovetail a Drawer

face, then add several more coats to the top. Next rub on paste wax for a beautiful shine that will protect the top from glue or

more on building workbenches, see FWW #4.

Newport-Style Tall Clock

Tackling the tricky details
by Robert Effinger

Built with the aid of 20th-century technology, Effinger's stately

mahogany tall clock captures the graceful proportions and crisp carving of the 18th-century Rhode Island original. The dial face was hand painted by Judith W. Akey.

hen I moved to Maine in 1970, I left behind a career as a tool-and-die maker. Working with wood instead of metal, I managed to eke out a living selling my turned bowls and wooden novelties to tourists who drove through town in the summer. One day a local gentleman stopped in to ask me if I could make a tall clock. I'd never attempted anything that ambitious before but I took the job. Since then, I've turned out quite a few. Along the way I've developed some methods that make short work of the details; I'll explain several of these in this article. The clock shown is based on an 18th-century mahogany tall clock attributed to Newport, R.I., cabinetmaker John Goddard (1745-85). I scaled up the plan from a measured drawing in Wallace Nutting's book, Furniture Treasury: Vol. III (1933, MacMillan Publishing Co.). I'm not a period purist so my clock isn't built exactly like the Goddard original. I'll improve on the old construction methods if I can. For example, unlike many old clocks, mine are built to allow for seasonal wood movement in places where the old clocks might have nails, glue blocks and, more often than not, cracks. The most radical change I've made is in the supports for the seat boardthe horizontal board that supports the clockworks. On old clocks, the waist sides extended up into the hood and the seat board was nailed across them. My adjustable seatboard assembly slides up or down until the movement's at the correct height, then screws tight against the waist sides. The -in. plywood bottom of my clock is another break from tradition. Old clocks had a thick bottom that was often dovetailed to the base sides. This construction works fine until a weight cable breaks and the cast-iron weight wrecks the bottom, feet and sides of the clock. A falling weight will smash through my thin plywood bottom, without damaging the rest of the clock. Buy the movement and make the dial before you start cutting anything. The depth of the movement determines the depth of the case and the dial must be made to fit the hood or vice versa. It's easier to make your own dial than it is to redesign the Goddard hood around a store-bought dial. Some of the fancy old engraved dials were made from brass, but I cut mine from 16-gauge sheet steel and sent it out to be hand painted. The sources of supply on p. 78 lists a few of the many companies that sell movements. The movement I used in this particular clock is a cablewound, nine nested-bell movement (No. 213) from the Concord Clock Co., 96 Main St., Plaistow, N.H. 03865. Think of the clock case as three separate sections: the base,

Fig. 1: Setup for routing hood moldings

waist and hood. Figure 3 (p. 81) and figure 4 (fold-out section)

show how these sections are built and how they fit together. The
waist sides screw to the base while the hood just rests on the waist. The hood slides off the front to allow access to the works.

With a square, extend these lines across the width of the blank,

extend the line of the curve over the end of the blank. Fasten the
template to the mahogany with small screws making sure that the template marks line up with the lines drawn on the blank.

The

-in. pine back ties all three parts together, as shown in fig-

ure 4. In general, the waist must be about in. wider inside than the swing of the pendulum. Most old clock waists measure in.

Draw the molding profile on the ends of the blank as shown in

the drawing. You'll set your router bit against this profile. One-quarter-in.-thick wooden discs in increments of -in. in

across the outside and 7 in. to 8 in. from front to back. I increased the depth of my clock case because modern musical
movements are larger than the old ones.

diameter fit over a pin in the auxiliary drill-press table directly

underneath the bit. With the template side of the blank down on

I made the special one piece hinges for the hood door from
-in.-thick sheet brass. These hinges screw to the top and bottom of the door and pivot on -in. #2 woodscrews in the scroll board

the table, I select a disc that positions the bit where I want it
against the profile on the blank end, adjust the bit to the right height, then guide the template against the disc to make the cut, as shown in the photo, p. 77. One pass hogs the straight return

and hood molding. The waist door also requires special hinges
with an offset to match the -in.-thick lip on the hinge stile as shown in the detail, figure 4. Ball and Ball is the only company

moldings, another pass at the same setting cuts the curves. Next I
switch to a smaller disc to move the stock closer to the bit or a larger disc to move the stock away. The idea is to rout as close as

I've found that makes these hinges.

possible to the molding profile you've drawn on the end of the

blank. I do as much hogging as I can with a -in. straight bit then

The curved goose-neck, or swan-neck moldings at the top of

the hood are often the most intimidating part of a tall clock case.
In the old days they were shaped by carving and scraping, but I prefer to make them with a pin router. My method of pattern

I switch to smaller straight bits followed by whatever curved bit

gets closest to the line. After routing, I scrape and sand out any imperfections in the molding.

routing cuts both of the curved moldings and both of the return moldings that run along either side of the hood at the same time,
from the same piece of mahogany.

To make the moldings, I've converted my drill press into a pin

router (see FWW #37, pp. 26-27). My setup guarantees that the moldings will match up perfectly at the corner miters. Start with a mahogany blank in. thick, 10 in. wide and 31 in. long. Make a template by drawing the molding curves on a 6-in.-wide piece of -in. hardboard, as shown in figure 1 and

After routing, trace around the template on the back side of the blank. This line will become the cutting line for the top edge of the molding. Remove the template, set the tablesaw blade to 45 and cut the blank along the miter lines.
To mark for the rosette, score about in. deep with a -in.diameter hole saw on the back of the blank. This gives you a definite line to follow later on the bandsaw. Rip the return molding off the blank along the straight template line. Now, with the back side up, bandsaw along the curved template line that marks

bandsawing to shape. On this template, mark off the miter lines and the center lines for the rosettes. Place the template on the bottom of the mahogany blank and transfer the miter lines and rosette center lines to the blank.

the top edge of each goose-neck molding, including the radius marked by the hole saw. Flip the molding over. The cutting line for the bottom edge of the molding lies at the lowest point of

the radius, as shown in figure 1. If you run a pencil along the

bottom of this groove, it's easier to follow with the bandsaw.

the pediment moldings. The template rides against a wooden disc

ing profile drawn on the end of the blank (above). After a pass along the straight molding, the goose-neck molding gets a pass at location is scored with a hole saw (right), then the goose-neck is bandsawn free (far right).

With his drill press converted to a pin router, Effinger routs out

over a pin under the work. Bit height is adjusted against the mold-

the same setting (top right). After sawing the miter, the rosette

bandsawn from the blank. After sawing the top edge and the rosette, the blank is flipped over and the lower molding edge is
Fig. 2: Turning "quarter" columns

The moldings are now ready to glue to the scroll board. The smaller scroll-board arch moldings can be made using

the same technique, but I find it easier to mount a router on a

cobbled-up pivot to cut the semi-circular part and guide the hand-

held router against a straight edge to cut the straight sections.

You could also turn the semicircular molding on the lathe.

The quarter columns on the waist of old clocks were just that,
of a circle. To my eye, these look sort of flat. I thought that the effect would be more dramatic if the columns were just slightly

more than

of a circle. Here's the method I developed to turn a

"quarter" column that's really a 120 section of a circle. Make a fixture from two pieces of -in. scrap stock as long as

the column. Rip one piece 2 in. wide and one

Cut a

in. wide and

butt glue them to make an L-shaped fixture, as shown in figure 2.

-in.-square piece of mahogany for the column. Screw

this square blank into the L-shaped piece as shown. Make sure that your screws are recessed enough that you don't turn into

them later. Lay out the center on each end, remove the corners
on the tablesaw, if you prefer, and turn the column and the jig to shape. A new L-shaped jig must be made for each quarter column.

If your lathe has an indexing head, you can rig up a router box
and cut the flutes right on the lathe with a small veining bit in a

router (see FWW #37, p. 34 and #38, p. 40), but I have a different method. I have an old indexing jig that holds the column
between centers and allows me to slide it across the drill-press

table against a cutter chucked up in the drill press. My cutter is a

-in. Woodruff key seat cutter that I've ground to a radius as
shown (available unground from Manhattan Supply Co., Inc., 151

Sunnyside Blvd., Plainview, N.Y. 11803). A bronze sleeve over

the shaft acts as a bushing and limits the depth of cut.

There are lots of ways to make ogee bracket feet but I think
that my method is the easiest. I cut and glue up the joints while

Author cuts column flutes with a Woodruff key seat cutter

the stock is still square. By clamping the glued-up foot to a small

jig rests on drill-press table and slides by cutter. Sleeve on cutter

ground to a radius and chucked up in the drill press. Indexing

limits depth of cut.

wooden box for support, as shown in the photo at left, I cut the ogee curve on the bandsaw.
The front feet are joined with a splined miter. I cut the spline slot on the tablesaw with the blade at 45. The back feet are joined with half-blind dovetails. The rear section of the back feet is made from thinner stock and left flat to allow the clock to sit

closer to a wall.

The flame finials that crown the hood are turned from 3 in.square blocks about in. long. Turn the finial in the middle of
the stock leaving about 1 in. of waste on each end, as shown in the photo. For now, just turn the flame section to shapecarv-

ing comes later. On the bottom of the urn, mark off 24 divisions for reeding and stop fluting. If your lathe has an indexing head, you can mark and carve the finial between centers. I carve the
reeds with a V-tool working from larger diameter to smaller diameter. Tipping the tool to the left and right, I take off the sharp edges to round over the reed. About in. down from the top of the reeding I mark a line around the circumference and another line about in. from the first. This designates the lengths of the shallow flutes within the reeds. I carve these with a small gouge. On old clocks, the flames on the outside finials spiral in opposite directions. To lay out the flame spiral, I draw lines parallel to the finial axis that divide the circumference of the cylinder into six equal sections. Then I draw lines around the circumference, spaced in. apart, to form a grid. I connect the intersections with diagonal lines to form the spirals. Carve between the spiral lines with a small gouge. After the flame has been carved, cut the waste off and finish to a point. I sand the completed finial with a 220-grit flap sander chucked up in the drill press. One other detail worth specific mention is the shell carving on the waist door. Some of the old ones were glued onto the'door panel after carving. I like to make the door panel and shell from one board as thick as the combined thickness of the shell and panel. I set the thickness planer to remove in. and I stop the planer before the shell area goes through. The finished shell is about in. higher than the panel and overhangs each edge by in. I set the jointer for a -in. cut and joint the panel edge stopping when I get to the shell area.

shape on the bandsaw. Supporting the foot over a wooden box allows the ogee profile to be cut on the bandsaw (left). Finial is lathe turned, then flutes are and carved by hand. To lay out the flame, divide flame into six longitudinal sections and five latitudinal sections to form a grid (below). Connect points on the diagonal to form spiral lines. Pencil holder shown marks out axis lines.
marked out in indexing jig

up while square then cut to

Ogee bracket feet are glued

Robert Effinger makes period furniture in Fryeburg, Me.

Sources of supply
These firms sell tall-clock movements, clock supplies and hardware, except as noted. Selva-Borel, 347 T3th St., P.O. Box 796, Oakland, Calif. 94604. Mason & Sullivan Co., 586 Higgins Crowel Rd., West Yarmouth,
Mass. 02673. Turncraft Clock Imports Co., 7912 Olson Highway 55, Golden

Valley, Minn. 55427.

Klockit, P.O. Box 629, Highway H, North, Lake Geneva, Wisc. 53147.
Craft Products Co., 2200 Dean St., St. Charles, Ill. 60174.

Ball and Ball, 463 West Lincoln Hwy., Exton, Pa. 19341 (authentic reproductions of hood-door hinges, offset waist-door

hinges and clock hardware).

Judith W. Akey, 173 Harbourton Rd., Pennington, N.J. 08534

(hand paints clock dials). The Dial House, Rt. 7, Box 532, Dallas, Ga. 30132 (custom dials and hand painting).

Making a Wooden Clockworks

Part two: Getting things ticking
by Wayne Westphale

n part 1 we discussed the theory of how a clock works. Now is the time to make one. There are a variety of ways to cut gear teeth, methods that cover a broad range of accuracy, speed and expense. The method you choose will depend on your goal, your shop equipment, and your budget. As an example of how low-tech clockmaking can be, for my first clock I turned the arbors on a lathe setup that consisted of an electric drill (as a headstock) clamped to a 2x4 (the lathe bed). A piece of angle iron, drilled and tapped to carry a pointed bolt, became a tailstock. A chisel served as a lathe tool and a piece of scrap as a tool rest. My first tooth cutters were reground spade bits, as shown in the bottom right photo on the facing page. Needless to say, this was doing it the hard way. I've always tried to surpass each clock I've built with a better one, and along the way I've invested in some pretty sophisticated equipmentmachines more often found in a metalworking shop. These are not essential to building a good clock, but they allow me, as a matter of routine, to achieve repeatable accuracy with little fuss. Expect this clock to tax your ingenuity in getting the necessary precision from your own machines and tools. There are ways around every problem as long as you understand the features in a clock that are critical to its operation. Horologists don't speak of gears, but of wheels and pinions. Wheels, the large gears, have teeth; pinions, the small gears, have leaves. I cut teeth and leaves on two different machines, but the process is basically the sameI use a set of reground router bits to cut the gullets between the teeth. The preparatory step, laminating plywood gear blanks, was described in part 1. The photos on the facing page show some of the actual cutting, including my jig for bandsawing circles. To cut the wheel teeth, I mount a stack of gear blanks on a mandrel and mount the mandrel on an old metal lathe, which I also use to turn clock arbors. The tool-bit holder on the lathe's cross-slide and compound has been adapted to carry a router, with the router bit perpendicular to the lathe centerline. By cranking one of the lathe's control wheels, the router bit can be positioned closer to or farther from the work, then locked in position to give a cut at a set depth. By means of another control wheel the router can be moved precisely along the length of the work. The first step in cutting the teeth is to turn the lathe on at slow speed, then use an end mill or hinge-mortising bit in the router to trim the blanks to true round, sizing them to the correct diameter at the same time. This ensures that the arbor hole will be exactly centered. The lathe is then turned off, and the blanks are indexed by a pin and a shopmade plate. The router roughs out the gullets one

by one by traversing horizontally along the stack. I crank the router from the tailstock end up to the headstock end to cut a tooth gullet. Then I crank the router back to the tailstock end, turn the stack of wheel blanks to the next index location and repeat the process. (The escape wheel is a special case. It has three very critical surfaces on each tooth, and I make these as shown in figure 3, on p. 62.) To minimize chippingand maximize cutter lifeI make several passes, each with a different cutter. The first cutter, as shown in figure 1, has straight faces, is easy to sharpen, and has an included angle of about 32. It is a wasting cutter. I set it to about 80% of full depth. The next cutter profiles the tooth face and cuts to final depth. The last cutter eases over the tooth tip. The relief angle of this cutter is only 2 to 3the desired effect is to round over and burnish the tooth tip in one pass. Next, I lightly sand with 400 grit paper over a soft block to remove the burr left at the tips of the teeth. This produces a stack of identical chip-free wheels. The method suits itself both to small scale production or, if you are making just one clock, to making any identical wheels that may be in it (there are two identical pairs in my grandfather clock). Pinions are cut in a similar way on a milling machine, as shown in center photo. The same operation could be accomplished with a drill press fitted with a compound table (available from Sears for under $80 and from time to time in various bulk-mail catalogs for even less) and a properly contoured cutter. I profiled my cutters in a series of steps, as shown in figure 1. I began with a full-size drawing of each of the gear-tooth profiles. Figure 2, on p. 61, shows the exact profiles of the teeth and leaves in my grandfather clock. To achieve the necessary variety with the fewest number of cutters, I taper my cutters slightly at the tip, so that the tooth size, the width at the pitch circle, can be controlled by depthing the cutter as required. Pitch circle and other technical terms are explained in part 1. Arbors and bearingsI turn arbors in the metal latheit is fast and sure and will maintain 0.001-in. tolerances (exact sizes are shown in figure 2). I strive for a snug fit of wheel to arbor. A metal lathe is not absolutely necessary, though I would not recommend using dowels straight from the hardware store either. You'll find that commercial dowels are only approximately sized and only approximately round. I recommend a piece of tool steel or -in. drill rod be pressed into the arbor to serve as a pivot. Wood-on-wood is too inefficient at this point from the standpoint of friction as well as durability. The pivot must be accurately centered. If your lathe has a

At left, large gears are cut by a router mounted on the cross-slide

metal lathe, which the author bought used for $4,000. The blanks are indexed by the pin opposite the router bit. Above, Westphale cranks the router along a stack of six wheel blanks, backed up at each end by a hardboard

and compound of a South Bend

blank to prevent tearout.

Left, a milling machine is the metalworker's precision version of a drill press, equipped with a table

The stack of pinion blanks is indexed by a dividing head, which calculates angles by means of perforated plates and a gearbox. It takes forty turns of the crank handle to rotate the output shaft one full turn. Far left, an efficient circle-cutting jig: The board has a runner on the bottom that rides in the bandsaw's miter-gauge slot, and a number of axle holes to suit the various gear sizes.

on X and Y axes by hand cranks.

that can be moved horizontally

not be high-tech. Westphale shows two of his early gear cutters, reground spade bits, alongside the highly evolved ones he uses today.

the drawing at left. Tooling need

Cutters are reshaped as shown in

Fig. 2: Wheels, pinions and arbors

Fig. 3: Escapement data

Westphale makes the teeth on his escape wheels with a series of straight cuts, as shown above, then routs out the curved shape of the gullets using a template and guide bushing (far right). The escape-wheel blank (or a stack of blanks) is mounted on a mandrel through the arbor hole, and the mandrel is fixed to a divid-

ing head. The dividing head rotates and locks the wheel blank a fixed amount for each cut, ensuring even tooth spacing. Cutters are held in the chuck of a milling machine, the metalworking equivalent of a drill press. The milling machine adjusts precisely in three planes to locate the cutter relative to the work. The divid-

ing head is attatched to a sliding table, worked by hand cranks, that moves the work horizontally past the cutter and back again for the next cut. When routing the gullets, the work is indexed under the template by the dividing head. Spokes are routed the same way (photo, facing page) then rounded over on a router table.

hollow headstock you can drill the pivot holes as I do, with a bit in the tailstock. If not, I'd suggest clamping a piece of scrap to your drill-press table and drilling a hole the diameter of your arbor through the scrap just off the edge of the table. Maintain the setup but change the drill to a size a few thousandths smaller than your pivot material; I find that a #53 drill bit works well. Insert the arbor from the bottom and drill carefully into the end. As the arbors are different diameters on each end, at least two different setups will be required. Bearings, which I make from nylon rod, can be drilled with a similar setup. In this case, just drill part way through the scrap. For instance, if you use in.-dia. bearing stock, drill a -in. hole in. deep with a Forstner bit into the clamped scrap. Then drill a -in. hole all the way through. Cut your bearing stock into in.thick wafers. Insert the wafer, drill the appropriate size pivot hole, then push out the completed bearing from below.

Engagement testingTest wheel-and-pinion engagement patterns at various center distances. In a scrap of plywood, drill a hole for a pin that will represent the wheel arbor. Around it, draw a series of pinion-arbor holes, one at the nominal distance from center, the others at -in. increments from the ideal. Mount the pinion on a pin in various holes, revolve the gears, and note how the teeth mesh. Part 1 explains what to look for. Choose the distance that gives the smoothest action. There is some latitude, but many times, while working out the tooth profiles of the grandfather clock, I had to refine the contour of one cutter or the other, and sometimes both. You don't have to go with the exact tooth profiles and distances I've worked out, but they work well and I recommend that you try to match them. Once the teeth have been cut, the wheels can be lightened with any number of spoke configurations. Spoke shapes are limited only by what is practical and aesthetically pleasing. My

Fig. 4: Hands, weights and pendulum

The escape wheel nearing completion. The acrylic template remains stationary, with its far end clamped to a block on the workshop wall. To rout successive spoke holes, the work is turned by the dividing head, which has been set in position to hold the wheel horizontal. Spoke-hole patterns for some of the other wheels in the clock are also visible in the photo. spoke template is shown in the photo above. I use a router and guide bushing with a -in. veining bit. Some of my spoke patterns are a series of round holes of various sizes, which can be cut with a drill or a circle cutter as size dictates. Next the spokes can be rounded over on a router table, using a regular piloted roundover bit. After that, I seal the wood with a mixture of tung oil, polyurethane and mineral spirits, equal parts of each. I soak the wheel for a few minutes, then wipe off all the surplus. At the teeth surfaces, I use high-pressure air to blow away all external traces of the sealerall I want left is what has soaked into the wood. After drying, I repeat the sealing step a couple of times until there is enough finish on the wheel to be buffed and polished. The final step is to wax the surface and buff it, but take care not to wax the tooth surfacesthey must run dry. After finishing, the escape wheel gets a little extra treatment with 400 grit paper to polish the contact surfaces of the teeth. The wheels can now be balanced. Do this after the wheel and pinion are secured to the arbor (I use both glue and brass pins, driven at an angle). Wood density varies and sometimes wheels that you would reasonably expect to be balanced are not. To test them, I rest the pivots on the open jaws of a machinists' vise. The heavy side of the wheel will stop at the bottom. Rotate the wheel one-quarter turn and release. If the wheel stops in the same position, it needs balancing; if it doesn't turn, or the stopping position is random, it doesn't require balancing. Usually it will. Mark the light side of the back of the wheel near the perimeter and drill a hole about halfway through the wheel. Insert a small lead plug or piece of lead shot and test again. Add more weight if required until the stopping pattern becomes random, then use a small nailset or punch to expand the lead in the hole. You can plug if you wishI usually leave the hole open as evidence the wheel was balanced. Setting up The clockworks are supported by two outside plates. An intermediate plate carries the back end of the weight arbor. My template for the plates is shown in figure 2. Distances between pivots are critical, and should be adjusted in your clock according to how each wheel/pinion pair functions in the engagement testing described earlier. My clock case is an open frame that is 76 in. high, 18 in. wide, and 11 in. deep. The clock plates are attached to the frame's crosspieces with screws from beneath. A photo of my finished clock was shown in part 1, and you are welcome to copy my case design as closely as you care to, but feel equally free to design

Rout-a-clock
While editing Wayne Westphale's article, and watch him at work. He relies on precision metalworking equipment, which most woodworkers don't have, so I began trying to think of other ways to make

by Jim Cummins

I had occasion to visit him in Colorado

The index wheel turns on a in.-dia. pin that sticks up through a particleboard base. My index wheel is acrylic

photos. Here are the basics:

the router-template jig shown in the

division table that appears in FWW on

FWW #38, p. 12). For those who don't have access to the table, here's how it
the following factors: for 64 holes, 0.0491; for 48, 0.0654; for 40, 0.0785; and for 30, 0.1045. Work things out to as many decimal places as your calculator

Proven Shop Tips (or Methods of Work,

found in any shop.

a clockworks, using tools that might be

The various options seemed: to rig up

plastic (though it could be hardboard, etc.) with four concentric circles drilled

works: Measure the exact diameter of the circle in inches, then multiply by one of

a carriage on the drill press in imitation of Westphale's dividing head and milling
ing router, or to focus on the indexing setup. I chose the last course and devised

for 64, 48, 40, and 30 holes (various increments of these give all the necessary divisions in Westphale's design). the four index circles. I worked out the spacing for the holes by using the circle

machine, to rig up my lathe with a travel-

Mark what will be the centerpoint on the wheel, then use a compass to lay out

will allow. Set sharp dividers to the fig-

ure (you can measure with a machinists' rule in hundredths, or convert things to sixty-fourths as I did) and step off around

At left, routing an escape wheel on the indexing jig. The rough shape of the teeth square off the tops of the teeth. Minor

can be cut using the template as shown. A second template is then needed to

spacer for routing. The newspaper allows

the finished pinion to be split free.

tearout can be patched with epoxy and sawdust. Pinion blanks (above) are too small to be screwed to the index wheel, but they can be glued to an oversize

one of your own. Just be sure your design will accommodate the pendulum pivot as shown in figure 4.

You will probably have to make many adjustments to get the clock just right. Clocks don't really run at a steady rate, but speed
up and slow down minutely according to the weather and which

Mount the plate assembly on the clock frame and hook up the pendulum and weights, shown in figure 4. My standard weights are three pounds each, but you may find that your clock will run on less (mine do, hut I allow a 50% safety margin for customers).
The clock should be set level, and the escape lever must be adjusted so that it performs as shown in the tick-tock sequence in part 1. Let the clock run for a while, as a test to see whether it is

particular teeth are engaged at any one time. But these slight irregularities average out. My grandfather clock is accurate to a

few seconds per day. For the final bob adjustments you may have
to let it run a week or more before you can tell whether it is
gaining or losing time.

fast or slow, then adjust the pendulum bob a little to correct it. If the clock is running slow, shorten the pendulum, and vice versa. Keep a record of how often you make adjustments and of how
many turns of the adjustment nut you make each time.

If your clock has problems, a careful rereading of part 1 should allow you to understand what they are. Clocks are fascinating and magical, but they follow physical rules. The important

checkpoints are summed up here: The perimeter of wheels and pinions must be concentric with the arbor and the pivots must be

the circumference to show the location of the holes. If it doesn't come out exactly right, adjust the dividers minutely and try again until it does. When the hole locations are scribed,

drill the center hole. Next mount the wheel on its pin through the base, clamp the base to the drill-press table, and pivot the wheel around to drill the series of holes in each circle. I used a in. twist

guide bushing and a in. straight bit with at least a in.-long cutting flute. Shopping locally, I found that Black and $2.49 (I'm going to need at least that many more before I'm done). A router bit will give you a round bottom to the gullet, not as nice looking as Westphale's

drill (the plastic will ruin a brad-point). Next, fit your router with a in.

Decker bits were longer than Master Mechanic bits, so I bought a half-dozen at

wheel from beneath, with the screw holes where the spokes will eventually

square corners, but perfectly functional. Screw the wheel blank to the index

be. Use a spacer between (I used lauan

the small gears, the pinions, I had to make two stepped-down center pins for the jig, one for a in. arbor hole and

plywood scraps) so the router bit doesn't chew up the index wheel. For

another for a

wood in them for anchor screws, so I glued them to an oversize spacer with paper in the joint, as shown in the

in. arbor hole. The smallest pinions don't have enough

The teeth and spokes of this 48-tooth dial-train wheel were routed using the plywood template shown. To make the spoke template, the author bandsawed the spoke hole pattern, closed the entrance kerf by gluing a strip of veneer in it, then trued the shape with a rasp. Next step will be to round over the spokes with a piloted router bit.
shape. The template should be indexed by riding on the center pinthis ensures

needed to come down to the correct

just spent musing about the myriad little

photo on the facing page.

Make templates that will rout the shapes of the gullets shown on p. 61. To use the jig, rout a gullet, turn the wheel

that the final wheels will be the same diameters as the practice piecesand can be clamped as shown in the photo.

decisions to be made at each step. I remember that it took me the best part of an evening to realize that I couldn't rout In all, it's been as much of a challenge

a sharp bit and to rout against the rotation, a technique called climb cutting.
This helps prevent tear out. My lauan plywood templates took me several tries each before I was satisfied,

a notch and rout the next, continuing until done. One critical point is to keep

Spokes are routed similarly. Make a template for one spoke hole, then use the index wheel to rotate the wheel the The spoke edges can be rounded over using a piloted bit in a router table, if you have one (I just clamped my router upside down in a vise).

a tooth, but had to rout a gullet. Things like that.

as Westphale promised and I've enjoyed

correct amount to space them equally.

be ticking for at least a few more weekends, but the results so far are very promising, and I think I'm on the way.

the project thoroughly. My clock won't

but each practice run of a few teeth will show you exactly what modifications are

All this has taken me about five weekends so far, with a good part of the time

Jim Cummins, who putters weekends away at his frame shop in Woodstock,
N. Y., is an associate editor at FWW.

at dead center of the arbor. Pivots must be straight, not bent (set
the complete arbor, with wheel and pinion mounted, on the

open jaws of a vise, then rotate the shaft briskly to check for

wobble, warping, etc.) Allow -in. endplay between the arbors and the clock plates (even so, if the plates warp the arbors may
bind). Check for teeth jamming (bottoming or tips butting). Remove the escape lever assembly to check whether the rest of the
gear train spins freely. Test pinion leaves for uniform spacing with a micrometer. Never use oil. I've found that the most likely problem is eccentric wheels,

engaged teeth with small dots of masking tape. Then start the clock again. The next time it stops, look with suspicion at any taped pairs of teeth that are engaged as they were before. If it's not the teeth, the same test may pinpoint two gears that are slightly out-of-round, and that bind only when their long axes

are aligned. A little work with a file may be all that's needed to
put everything right.

Wayne Westphale designs and builds clocks at his shop, Contemporary Time, in Steamboat Springs, Colo. His grandfather clock

pinions or arbors. One diagnostic trick, which I hope you will never need to use, may pinpoint an intermittent fault. If your
clock regularly stops for no apparent reason, mark each pair of

took two years to develop, and is copyrighted. Westphale extends to individuals the right to make a copy of his clock for their own use, but not for commercial purposes.

Building a

Chest of Drawers
Joinery and design considerations

by Christian Becksvoort

nyone who has worked with wood knows that building a piece of furniture involves a series of decisions, from choice of wood, to design and construction details, to selecting the final finish. Even if you build identical pieces over a period of years, you'll probably find yourself making small changes each timeperfecting the proportions, or perhaps just exploring different joineries. In this article, I'll discuss the range of decisions and processes involved in designing and building a chest of drawers, as well as presenting some of the methods that I employ. One of the first things to consider in designing a chestand perhaps the most subjective part of designis the chest's overall appearance. You should style the piece to suit your tastes or to fit into the decor of the room the chest will inhabit. Since I live near one of the last two Shaker communities and have restored several original Shaker pieces, my designs show the strong influence of their simple, unadorned style. I prefer my furniture to By manipulating the design, construction, and materials, a chest of drawers can be built for any purpose or to fit into any decor. The trim lines of the author's cherrywood chest, left, display the usual austerity of the Shaker style. Inside the chest, below left, web frames support the drawers and add rigidity to the carcase. A dust panel fitted into the bottom frame keeps the chest clean inside. The finished back, below right, is built like a door and permanently fixed to the edges of the carcase, allowing the chest of drawers to be freestanding.

have clean, functional lines that are more at home in a wide

in the opening and bind. Wide cases can have two side-by-side

spectrum of settings than avant-garde or strictly traditional forms.

One nice thing about Shaker-based designs is that you can begin

with a basic carcase, then manipulate details like tops, moldings and bases for different effects. Moreover, Shaker construction
techniques are sound and durable, yet not so involved that a novice can't employ them.

drawers with a vertical divider between them. In casework, as in any furniture, the basic construction of the carcase affects both the look and the function of the piece. As is the case in designing anything, there are many paths leading to
the same point. Your choice of joinery for instance, will depend

on your level of skill or the amount of time you want to spend.

Material selection also plays a role in joinery methods. Although

I begin a chest design by considering what will be stored in it, since the number and size of drawers will determine the chest's overall size. Bulky items, like clothes, will require deep drawers, while smaller items will do well in a larger number of shallow drawers. Casepieces vary widely in their approach to drawer storage. Most chests, however, have drawers of graduated depth,

you may choose to use flat, dimensionally stable plywood, the following discussion of joinery will focus on solid-wood methods, which I prefer. Once its movement is understood and allowed for,

the advantages of solid-wood joinery and grain selection far

outweigh the drawbacks, in my opinion. All of the Shaker casework, and virtually all of the 19th-century furniture I've worked with, is constructed in one of two ways,
with solid-wood carcase sides or by solid-wood panels let into frames. I use both methods, but I'll focus here on what I call

with the deepest drawers at the bottom, the shallowest at the top. This serves to anchor the case visually and keeps the heaviest
objects in the lower drawers. Instead of a single shallow drawer

at the top of the case, Shaker cabinetmakers often used a pair of

drawers, separated by a vertical divider. Storage capacity and where the chest will befreestanding, against a wall, at the foot of a bedwill dictate the overall size of
the case and its dimensions. Proper proportions are important to

slab-sided construction. For more on frame and panel Shaker construction, see FWW #58. Figure 1 shows the basic anatomy of

a solid-wood chest of drawers. The construction is really quite

simple, and once you've decided on the design of your chest, you have only to select the joinery appropriate to the material and details you wish to employ.
Although the Post-Modern design movement has revived interest

the success of a chest design. If you aren't confident about what dimensions to use, measure a piece similar to the one you want to build or find a successful design in a book and go from there. One worthwhile option is to rely on the standard dimensions for

in the use of architectural details and moldings on furniture, I prefer relatively unadorned designs. Still, a simple carcase lends itself

furnishings in Architectural Graphic Standards by Ramsey and Sleeper (John Wiley & Sons, Inc., 605 Third Ave., New York,
N.Y. 10158). The volume costs more than $100, but most larger

to a variety of subtle details. To a large extent, these details will determine the kind of joinery you use to assemble the carcase. I prefer to join the top with dovetails, because they are strong

libraries have a copy. If you need the size of an average dresser, school chair, or even a tuba, it's all there in 750 invaluable pages. Another useful source for design standards is Human Factors

and decorative. But they aren't absolutely necessary and there

are several other options, as shown in figure 1A. The easiest,
perhaps, is the butt joint. When it's screwed and plugged or

Design Handbook by Wesley E. Woodson (McGraw-Hill, Inc.,

1221 Ave. of the Americas, New York, N.Y. 10020). When design standards don't suggest useful solutions, you'll

doweled, it's more than strong enough to properly join the case
together. The rabbet joint is slightly more work, and if planned

have to dimension the piece. Sketching on note pads is a good way to do this. If you are more comfortable with scale or even full-size drawings, by all means use them. They often help uncover tricky details you might otherwise overlook. As a general
rule, square shapes don't work well visually for cheststhey

correctly, it can be screwed from the side. A splined-miter joint, although strong, is tricky to cut if the panels are even slightly
warped. Another more contemporary-looking top joint is what I

call a bookcase top, where the carcase sides extend beyond the top, which is let into the sides with a dado, a rabbet and groove,
or if you're ambitious, a sliding dovetail.

appear monolithic and squat. However, a chest with a relatively

squarish frontthat is as wide as it is tallcan still look fine, as

long as its depth is less than about a third of its height. Rectangular shapes, for both the chest's front and side views, are far more
pleasing than squares. And rectangles based on the golden mean

are an excellent starting place. The golden mean is the oldest

and most widely understood proportioning method and can be

expressed as a simple ratio: 5 to 8. Thus, a golden rectangle is one whose short side measures 1.0 to the long side's 1.6. In instances where a chest's frontal dimensions aren't fixed by spacial constraints, it makes sense to build proportions around drawer depth. Generally, a drawer for clothes should be 16 in. to 20 in. deep and about 10 in. to 12 in. high. A deeper drawer has

two drawbacks: you have to take an awkward step back to open it

fully; and the carcase will have to be unpleasingly deep to accom-

Moldings can define edges and corners, crown or "set off" tops, break up monotonous surfaces, or cover up dowel and screw holes (see figure 1B). Further, even a simple cove molding can define the top of the case, without giving it the look of a period piece. A molding carried around the top edge of the carcase is another design option. Obviously, this introduces cross-grain constructionthe carcase side will shrink and swell, eventually loosening the molding. I've yet to see an old chest that hasn't succumbed to this problem. So if you choose to use a molding, you'll have to live with it. I glue the miter and about the first in. of the moldings, and nail the rest with brads or finish nails. The nails allow some wood movement. Another way of having a molded top without worrying about it falling off is to build an applied top. Attach the top anyway you choose, such as with blind

modate the drawers. Very shallow drawers, say less than 14 in., are
also a problem. It's too easy to accidentally pull them entirely out

dowels or tenons. In any case, there's no cross-grain problem

since the top moves in the same direction as the carcase sides. Deciding on the base presents another series of choices and

of the case, dumping the contents on the floor. Unless there's a

concrete reason to do so, such as having to store large sheets of

paper flat, I don't alter basic drawer dimensions much. A case can be as wide as it needs to be, but drawers themselves shouldn't be much wider than 40 in.; otherwise they tend to twist

figure 1C illustrates some options. A dovetailed case can sit directly on the floor, with or without a molding at the bottom
edge. This is neat, since you don't have to dust under the case, but it can look heavy and the chest won't sit flat on an uneven

Fig. 1: Chest-of-drawers design options

floor. To add visual lightness, you can raise the case a few inches off the floor using any of several methods. One easy-to-make

raised base is the plinth, which is simply a frame slightly smaller

than the footprint of the case bottom and 3 in. to 5 in. high. The plinth is a good contemporary treatment that would probably go

nicely with a bookcase top. Another way is to shape cutouts at

the front and/or sides of the chest. In this case, a web frame or solid bottom of the carcase will have to be joined to the sides

using one of the joints suitable for a bookcase top. Besides giving the chest some visual lightness, shaped feet also raise the bottom drawer so it won't drag on the floor upon opening. The

most involved base treatment I use is the traditional bracket

base. Bracket feet suffer from one major drawbackthe back feet

are bracketed only on one side, and pulling the chest across the
floor can easily break them off.

I start building a chest by cutting and gluing up panels for the

top and sides (and bottom, depending on the base). Once the

glued-up sides and top panels are dry, I rip them to width. To cut the carcase sides to length, I usually clamp them together and cut both at once on the radial-arm saw. These cuts must be absolutely square. Checking for squareness along each step of the
way is essential as minor errors early in the construction process

tend to accumulate and cause real headaches later.

Lay out the interior of the carcase by clamping the two sides together so that all the edges are flush, and mark the positions of

the drawers across the front edges. To figure vertical drawer spacing, first calculate the available drawer spacethe total inside measurement of the carcase minus reveal at the top, width of the

drawer rails and any applied moldings. Now divide this by the
number of drawers to get an average drawer height. If you want

the drawers to be evenly graduated from top to bottom, say in

1-in. increments, and you have an uneven number of drawers, figure the middle drawer will be the average height; each drawer

above it will be 1 in. smaller than the one below it; and each
drawer below it will be 1 in. larger than the one above it. If you have an even number of drawers, add in. to the average drawer

height for the drawer just below center and 1 in. to each drawer below it; subtract in. from the average for the drawer just above
center and 1 in. from every drawer above it.

Before assembling the basic carcase, you must decide how the
drawers will be supported and what they will ride on (figure

1D). Then, prepare the inside of the case accordingly. Very early or primitive chests often had solid-board dividers or bottoms between drawers. Later, light wooden web frames let into the carcase

were developed. These support each drawer and add stiffness to

the carcase without excess weight. Another all-wood method involves mounting guide strips inside the case that ride in
grooves cut into the drawer sides. This works well on banks of small drawers when you don't want to lose drawer height to all

those web frames. But the drawer sides must be

in. or thicker

to accept the groove. For fully extending drawers, such as file cabinets, you'll probably

want to use metal drawer glides. These come in many styles and
sizes (available from Grant Hardware Co., High Street, West Nyack,

N.Y. 10994-0600 or Accuride, 12311 S. Shoemaker Ave., Santa Fe Springs, Calif. 90670). They're strong and easy to mount, but most require -in. clearance between the carcase and drawer, so drawer fronts must be wider than the drawers to cover the gap.
If you choose to use web frames (the method I prefer), you'll need to cut the dadoes before assembly. You can use a dado

blade on the table or radial-arm saw, but if the sides have a slight

measuring diagonally from the top left to the bottom right corner
and vice versa: The readings should be identical. If not, shift or clamp the "longer" dimension until they are. Sight across all the

web frames to check for a warped case. This is seldom a problem, unless the top panel is severely warped.
To install the frames, turn the case back side up and slide each

frame into its pair of dadoes. Drive and set a few finishing nails
(6d if you're using -in. stock) at a steep angle through the frame into the case. Check the angle carefully, lest the nails

come out the other side of the frame or, worse yet, out the side of the case. Even though the frame grain runs perpendicular to the case side, wood movement is no problem, since the nails

allow for adequate movement. Purists who disdain the use of

warp and won't lay flat, cut them by hand or with a router to nails might use a sliding-dovetail joint instead. I have no problem with nails since they're found in almost all the traditional

ensure a consistent depth. For the router method, I made a jig

with two parallel bars, spaced the width of the router base and

furniture I've examined.

I won't go into detail about drawer construction here; for that,

about 30 in. long (see figure 2). The jig is clamped up square to the carcase side and a -in. bit ploughs a groove for the web frame at each location. The dado cuts can either be all the way across or stopped in. to in. shy of the front edge to allow the fitting of a drawer rail front strip, which I prefer.
Web frames can be four-piece assemblies, but on wider chests, a fifth center member adds stiffness to the carcase. Frames, as

I suggest you refer to FWW on Things to Make, pp. 20-24. However, I'd like to add a few of my own thoughts. First, I fit my drawer sides directly to the opening, making them about in. shorter than the case depth. I cut the two sides and front first, then run grooves in. to in. above the bottom edge to allow
room for half a dovetail below the groove. Then I plane in. from the bottom edge of the drawer front to keep it from dragging on the drawer rail. I lay out the dovetails on the sides with

well as other interior parts, can be made of a secondary wood,

such as poplar or pine. Measure the depth of the frame inside the

half dovetails, top and bottom, and as many full dovetails as

necessary in between. Avoid cutting one directly on the groove

carcase to the front edge, or to the stopped dado if you add the dovetailed primary-wood drawer rail across the front, as detailed
in figure 1D. I mortise and tenon the joints, then glue and pin

the frames for strength. Set the case on its back and cut, dovetail, and glue these into place. I prefer half-blind dovetails for a cleaner look on the side of the case, although through dovetails are acceptable and easier to cut. An easier option is to run
through-dadoes across the sides and use primary wood for the front members of the frames.

for the drawer bottomthat will leave a hole in the front. When the joints are cut, glued and assembled, measure the diagonals to check for square, After sanding, the bottom can be fitted. Make sure the grain runs side-to-side; if it runs front-to-back, any expansion of the bottom forces out the sides, weakens the joints and binds the drawer.
One trick sometimes found on old pieces is to let the bottom act as a drawer stop. The bottom extends in. to in. beyond the drawer, so it hits the case back. When the drawer is fitted, the bottom is trimmed until the front is flush. Seasonal moisture changes that cause the case sides to expand and contract will also make the drawer bottom expand and contract, and, in theory at least, the drawer fronts remain flush. Some makers install stops on the web frames that hit the inside top or bottom edges of each drawer, keeping it flush with the case front despite carcase expansion and contraction. A final word on finishing: I like to belt sand with 120-grit and 150-grit, then vibrate to 220-grit and 320-grit, followed by hand-sanding with 320-grit and 600-grit. On some surfaces, I finishscrape using my secret weapon: single-edge razor blades (available in boxes of 100 from most paint or hardware stores). Those who don't care for the ritual of sharpening scrapers will find these blades a real boon. They're tremendously sharp and stay that way long enough to do one or two sides. They can then be tossed out or saved as glue scrapers. Although I use an oil finish, the finish you use is not as important as how you finish. All exposed surfaces, top and bottom, inside and out, must be finished, because uneven moisture absorption causes uneven wood movement and warping. Many antiques were not finished inside, yet they were seldom built in the Northeast and shipped to the Southwest, nor were they exposed to the forced-air heating common in modern homes.

High-style period pieces that have web frames are often fitted
with dust panels, which are thin panels let into grooves in the web frames. The purpose of these dust panels is twofold: to keep

clothing from getting tangled in adjacent drawers and to provide a barrier against dust. Some makers still use these dust panels,
but I find them unnecessary. If I use a web frame, I do, however,

install a dust panel for the carcase bottom. Before the carcase is assembled, you need to decide on the kind of back it will have (figure IE). Traditional chest backs consist of individual boards (secondary wood is fine) joined together, let into a rabbet and nailed in place. A spline joint is perhaps the
best way to join the individual boards since it's easy to make and

locks the boards against misalignment in both planes. Under no

circumstances should the back boards be glued to each otherthe whole idea is to let the boards move. Each board is simply nailed

to the top, all web frames and bottom, and the nails are set. Freestanding pieces can have a frame-and-panel back built like a cabinet door out of primary wood and fit into the back rabbet. Panels can be flush, recessed or raised. A frame-and-panel back not
only looks much better, but gives the case tremendous strength against racking.

Some notes on assemblyAfter all the dadoes are in the right

places and of equal depth and web frames are identical in length, and square, the case should be dry clamped and tested for square. This is vital. If it's out of square, the drawers will
be a pain to fit. Now's the time to find out. Check for square by

Christian Becksvoort is a professional furnituremaker. He lives in New Gloucester, Maine.

Building a Stool
Compound angled joints on drill press and tablesaw
by Gary Rogowski

t was not in the sometimes-a-great-notion category that I decided to build a stool a few years ago. I needed something sturdy to sit on. "How hard could it be to knock out a stool?" I asked myself. My first attempt ended in a three-legged triumph of material over maker. It was astonishingly ugly and so precarious that you could sit on it only with great caution. It did hold a plant very nicely though. In the process of building my first stool, I learned a basic lesson. Effort, not luck, and planning, not good intentions, are required to successfully build a piece of furniture. This involves

a thoughtful approach to design, accurate drawings and careful construction. Gone is the innocent notion that one relaxing weekend of humming and puttering is enough to concoct a piece with style, grace and strength. So, I started over. Designing a chair or stool is a deceptive task, like setting up a model train. Kind thoughts blessed with the vision of an innocent invariably produce some degree of frustration. It only looks simple. You soon find the job involves more work than you expected. The process is a lot like designing other types of furniture in that it involves solving a series of problems, both aesthetically and

structurally. Stools do present unique design difficulties, however.

A stool's parts must strike a delicate balance between looks and weight. Stools look jaunty compared to chairs, are comparatively lighter and easier to move. Yet, looks can't come at the expense of strength. Thus, a stool is built with the strength of a timberframe house even though its airiness gives the impression it is built of matchsticks.
As if this didn't present enough of a challenge, recall that, by design, stools are meant to put you high off the floor or close to

it. Generally, the former design is more popular because there are more reasons in this world for sitting at workbenches, counters and bars than there are for sitting a few inches off the ground. Thus, stools are generally higher than chairs and narrower in front and side profile. This makes for a weight distribution
problem. Chairs are comparatively wider and more stable, so their legs can be perpendicular to or angled from the seat. Stools

need all the stability they can get, are more stable and look best

if their legs are splayed (i.e. they slope at a compound angle). A

stool with splayed legs distributes a person's weight over a greater area than one with legs perpendicular to the seat. Thus, a person

sitting on a stool with splayed legs must tip through a greater arc
before falling over. Splayed legs help, but don't entirely solve the stability problem. A stool's stability and attractiveness rely heavily on its

proportions. In my case, these were quickly arrived at through an empirical method. I wanted to sit comfortably at my workbench
and knew that chair height, about 17 in. or 18 in., would be too

low. I put a chair on top of bricks, placed phone books on its seat

and finally, sat myself on top of the phone books. I discovered

the correct seat height was about 25 in. With the stool's height

established, I went to the drawing board and made a series of drawings at in. scale, experimenting with various leg angles and spacing arrangements. I finally decided to locate the legs so they were 12 in. apart at the base and sloping 82. Given the stool's height, the slope of the legs looked just right. I used a

sliding protractor to copy the angle of the legs off the drawing.
Angled joineryIt soon occurred to me that although the sloping legs add to a stool's attractiveness and stability, they presented quite a challenge in joining together its partsboth in terms of rung-to-leg and leg-to-seat joinery. This was particularly troublesome in my case because, being influenced by James Krenov's work, I wanted to use exposed mortise-and-tenon joinery throughout. This meant the eight rung mortises would have to be cut to compensate for the slope of the legs, and the tenons on the ends of the rungs would need sloping shoulders. Furthermore, the compound angle does funny things to the geometry of the legs. The footprint seen in figure 3 shows an exaggerated view of what happens. Fortunately, the gentle curve I added to the legs had no effect on the joinery, because the legs curve only on their outside surfaces and were shaped after the joints were cut. Joining the seat to the legs seemed similarly tricky, but after some head scratching, this problem was easily solved by letting a lathe-turned dowel into the top of each leg (I explain how to do this below). A dowel joint eliminated the need to cut a tenon with compound-angled shoulders on the end of the leg. I started work on the stool by milling up my stock for the legs,

blade over 8, made a test cut and then checked it with the protractor. I repeated and tested the cuts until the saw cut a perfect 8 slope. To complete the compound angle, I set the miter gauge to 8, make another test cut and check as before. When I
can produce a perfect compound miter, I clamp a stop to the miter-gauge fence. Then, I cut one end, tip the miter gauge to

8 in the other direction, flip the leg over and cut the leg to finished length. Be sure to check that the top and bottom of the
leg are parallel after it is cut to length.

To bore the dowel hole in the end of each leg, I clamp the leg to a vertical sliding table bolted to the drill-press table. The sliding table moves toward or away from the drill-press column (see figure 2, above). To ensure the leg is plumb to the bit, I swivel the drill-press table around until the leg butts up to a long

brad-point bit chucked into the press, adjusting the drill-press

table so the leg lies along the bit's length. I then reposition the sliding table so a second face on the leg lies along the bit's length (referencing off a face perpendicular to the first face).

Once I'm sure the leg is plumb to the drill bit, I switch to a -in.-dia.
bit, center the leg's end under it and bore a -in.-deep hole. Next, I crosscut the rungs on the tablesaw with a plywood crosscutting jig with an 82 wedge tacked to its fence. With the wedge to the right of the blade and its narrow end pointed toward the" left, I cut one end of each rung. Then I positioned the stop block, flipped each rung over and cut it to length. Next, I

rungs and seat. Cutting the compound angles on the ends of the
legs seemed the trickiest job, so I started with them. I cut an accurate 8 compound angle on the ends of each leg on the table-

lowered the blade and repositioned the stop block to cut the
tenon shoulders. On each rung, I cut one shoulder, flipped the rung end for end and cut the shoulder on the opposite face. I then switched the wedge and stop block to the other side of

saw. With an extra-long fence on the miter gauge, I tilted the

double duty as a slot mortiser. Clamps un-

shape is checked with a template with a slot cut by the same end mill that bored the mortises. With addition of a shopmade sliding table, right, the author's drill press does

with a file to fit the mortise, above. Their

Tenons are cut square and then rounded

der the table hold stop blocks to set length

of mortises, which are first bored with a

brad-point bit, then finished with an end

right to angle the mortises, accounting for angle and locate stops is shown in place.

mill. The sliding table is sloped to the left or

the legs' slope. A test leg to establish table

the blade and repeated the procedure to cut the remaining two shoulders.

tipping the table until the rung laid flat against the drill bit.

Next, I roughed out each tenon cheek on the bandsaw to prevent the offcut from flying back at me and finished sawing the
tenon with a tenon jig on the tablesaw. Back at the bandsaw,

Without changing the drill-press alignment, I removed the drill bit and chucked a four-flute end mill into the drill press.

The end mill badly mauled a test leg at every speed I tried. The
remedy was to bore out the bulk of the mortises with a bradpoint bit. The mortises were cleaned up taking shallow passes with the end mill, running the drill press at 1,600 RPM. In boring with both the brad-point bit and the end mill, I prefer to stop in. or so shy of boring out the other side of the mortise. The leg's outside is shaped after the mortises are cut, so the remaining

each tenon was trimmed on its top and bottom edge to give it a shoulder on all four sides. Shoulders of two rungs must be pared, as described in figure 3, to compensate for the slight diamond shape of the legs. The tenons were rounded with a file to match the curve of the mortise and each was slotted on the bandsaw to receive a wedge. A -in. hole was bored at the bottom of the slot to prevent the wedge from splitting the tenon.

wood is cut away, leaving a clean opening. If you bore through

the other side, you will have to put a piece of scrap under the leg
to keep from cutting into your sliding table; you also risk tearing

Drill-press mortisingI moved on to cutting the leg mortises. I don't own a slot mortiser, but have the next best thinga horizontal sliding table for my drill press (see photo above, right). I set the table's angle for the sloping mortises using the angle on a rung end as a guide. I chucked a long drill bit into the drill press to serve as a positioning guide, stood the rung up on the table,

out the exit hole. The final shaping may not be able to remove the tearout if it's too severe. Mortises 1, 2, 3 and 4 are cut with the

table sloping to the right; mortises 5, 6, 7 and 8 are cut with the table sloping to the left, as shown above. Remember to mark the
table for two different sets of stops for boring the upper and lower rung mortises. When boring for mortises 3, 4, 7 and 8, I had to

Fitting rungs
I wish I had an extra 10 minutes for every time my mother told me not to tip back in my chair. I could retire. Mothers know a chair or stool's rungs are its weakest link;

by Jeremy Singley

tightly to each otherdouble jeopardy. This requires rung holes be bored in the leg at the correct angles and cut to an accurate length. If these requirements aren't met, the assembled stool or chair will have legs sloping at different angles. Fortunately, I've developed a bunch of techniques and jigs to make the job of fitting rungs to legs easier. For example, when I have just one or two stools to build, I bore the leg holes in the seat with a hand drill sighting along a sliding bevel gauge set to a leg axis line as a guide. I dry assemble the legs to the seat, then eyeball the alignment of the rung holes. The rungs are bored from inside the legs with an electric drill. If the room between the legs isn't enough to accommodate the drill and a full-length bit, I use a groundoff Powerbore, spade or twist bit. holes centered. I solve this problem with a marking tool or a try square and straightedge, shown in figure 7. I bore on the marks and measure for rung length. I check that the legs are the correct distance

to the woodworker, rungs are a pain in the neck. Not only do rungs have to fit tightly to the leg, in some cases they have to fit

ignations of leg positions: 1front left; 2front right; 3rear left; and 4rear right. This system assumes the stool is flipped over (the position it will be assembled in) so its left, front leg becomes the leg sticking up at the left rear. I use a color code to keep track of rungs. I crosscut boards and paint their ends a color to match rung length. No further coding is necessary, because turned rungs are not handeda top rung that fits on the right side of the stool would also fit the left side. Then I rip the boards into rung blanks.
you botch a rung hole? First, act innocent. Why bother the world with more bad news than it has already? Usually, turning a new leg is more work than the following alternatives. If the hole location is wrong and you plan to paint the stool, simply fill the hole with a plug and sand it flush. If the angle is wrong, turn a rung and then whittle the oversized tenon at an angle by eye to fit the hole (see figure 8, below). You can also glue a lathe-turned dowel into the hole and rebore. Since you have to bore the new hole exactly on the center

A final question: What do you do if

The tricky part here is getting the rung

of the dowel, position the plug with the tail center mark facing up and bore into the center mark with a brad-point bit.

Jeremy Singley makes chairs and stools in East Middlebury, Vt.

with a sliding tongue to measure the distance from the shoulder of one hole to the

between their ends, use an extension rule bottom of another, then add the depth of the second hole. If you find reading the ruler in this situation awkward, you can simply measure the depth of each hole and then measure the span between them. The rung stock is cut to this length, and its ends are shrunk to a snug fit with a heat

lamp just prior to turning and reshrunk just before the stool is assembled. The tenons swell from the moisture in the glue, locking them firmly in their holes. The eyeball rung-boring method works if you have only a few stools to build; if you have a large batch, it pays to set up a jig. It's crucial to keep the legs organized as
you bore. A box of 50 unmarked legs with

two rung holes in each makes a fine solution to the leisure-time problemyou can spend your day off sorting legs. To avoid this, before each leg leaves the jig, I mark its end with a number from one to four, corresponding to the four clockwise des-

March/April 1988

43

had to glue and clamp the stool in subassemblies. I glued up the

front and back pairs of legs and rungs (the legs perpendicular to

the long axis of the seat) and hammered the wedges into the

tenons. When the glue was dry, I inserted the remaining two rungs and then glued and clamped the two pairs together.
Seat shapingI was moving into the homestretch and started
work on the seat. To save time shaping, I wanted to saw out the curve in the seat, but my bandsaw wasn't large enough to accept

on the router table. Note that the template curves in two planes
test such a setup before trying an actual leg.

bandsaw and then cleaned up with a straight bit and template

The curves on the outside of the legs are roughed out on the

the seat blank turned up on edge. I solved this by making the

to accommodate the curve routed during the first pass. Always

seat from two narrower halves; I sawed the curve in each half then glued them together. I don't cut to the curved line, but
leave a slight amount of wood to allow for tearout when boring the leg holes. Save the curved offcuts, you'll need them later to
back up the seat when you bore the holes through it.

To mark for boring, I set dowel centers into the top of each leg, placed the seat top down on the bench and set the legs with

dowel centers against the seat's bottom. After checking that the
seat was positioned correctly relative to the leg, I tapped on top of each leg to mark the seat bottom and then returned to the
drill press. I use a hinged jig that bolts to the drill-press table to bore the

leg holes (see figure 4). This is simply two pieces of plywood connected by a piano or butt hinges. The jig is placed with the

hinge knuckle opposite the drill-press column and the jig's edge
parallel to the drill-press table. The top plate is wedged up from underneath until it's at an 8 slope, then clamped in place so the wedges can't move. Next, I tilt the table 8 to the left or right, using a rung, as before, to align the table relative to the bit. I rest

the seat top down on the table with the offcuts taped together underneath the seat. The brad-point bit enters the offcuts as it
exits the seat, reducing tearout on the top of the seat. The seat is parallel to the edge of the jig. Assembly is relatively easy compared to the rest of the project, but it takes considerable clamping force to bring the stool parts

together. I set concrete blocks on the shop floor and then put a piece of plywood that is slightly larger than the area covered by the stool's base on top of the blocks. I set the legs on top of the plywood and the seat on top of the legs, then rest cauls on the seat's long axis (the cauls are notched to allow the leg dowels to
project through the seat). I bring the seat and legs together with

shim the legs with a slip of paper under the rear, back edge to compensate for the compound slope of the legs.
With the setup fine-tuned, I cut the mortises and moved on to

shaping the legs. Working from a full-size drawing of the stool, I made Masonite templates shaped to the curve of the legs and seat. I transfered the marks from the Masonite leg template to one made of alder, which I curved in depth and plan view (see photo, above) to match the curve of the legs. I marked the legs out, rough sawed them on the bandsaw, then taped the template on top of one leg and finish shaped it on the router table with a
straight bit and ball-bearing pilot bearing against the template. I

bar clamps running from the plywood to the cauls. I have to flex the legs a few degrees to get them into the holes; this requires a fair amount of force. Once the legs fit in the holes, a generous amount of clamping pressure is required to bring the legs and seat together. Once the legs butt up to the seat, I take the clamps off, spread some glue on the wedges and bang them in place. After the glue has dried, I file down the seat dowels and spokeshave and sand the seat to its final curve. I file down any remaining projecting tenons. I lightly sand any areas that require it and then finish the stool with Watco or a similar oil. I prefer oil
finishes because stools are prone to being roughly handled, and oil finishes are easy to retouch. Through the years, I have made a number of variations of

then flipped the .leg over 90 and used the alder template to

guide the straight bit while cutting the second curve. As can be seen in the photo, the curve in the template's depth accommodates the first curve cut in the leg.
I sand the legs and rungs before assembly because it's impossible to do a quick and neat job after the legs are assembled. There is no way of simultaneously assembling all four legs and rungs on a stool single-handedly. Because I work by myself, I

these stools to suit the customer's needs. Their heights have ranged from 24 in. to 27 in. and with different rung heights, but I

haven't changed the basic design; neither have I changed my attitude toward building themanother relaxing, uncomplicated weekend project. I'll get started after brunch.

his Portland, Ore. shop.

Gary Rogowski builds stools and other intriguing projects in

44

Fine Woodworking

Beads and crown dress up a basic cabinet

by Tom Hagood

Building a Shaker-Style Wardrobe

've never been especially fond of period furniture; therefore, the commission I received to build a traditionally styled wardrobe cabinet was quite a challenge. The client did not specify the style, but the piece had to fit into an Early American bedroom dominated by a huge antique mahogany beda family heirloom. Because all my previous work had been with contemporary designs, I had some reservations about accepting the commission; a contemporary wardrobe would clearly be out of place in this bedroom. Thus, I set out to find a period style that would be traditional yet allow for some creative interpretation so I could incorporate my own design details. In this article, I'll tell you how I developed my design, worked out the details and built the cherry wardrobe cabinet pictured at right. Before I began designing the wardrobe, I searched through furniture reference books for a period piece to serve as a point of departure. In my search, I discovered a reference book called Chests, Cupboards, Desks and Other Pieces by William C. Ketchum, Jr. (published by Random House, Westminster, Md. 21157). Although primarily a book for furniture collectors, I find it a valuable resource for furniture designers: The book includes photographs and drawings of a dizzying number of styles, from English and French period pieces to Shaker designs to Wendell Castle stack-laminated work. The book even includes joinery details and dimensions for many of the pieces. Ketchum's book contains a photo of a simple Shaker wardrobe that seemed to fit my needs. The piece is traditional looking yet plain enough to benefit from the addition of some of my decorative details. I was already somewhat familiar with Shaker furniture and crafts, having at one time manufactured Shaker-style steambent oval boxes, and I like the Shakers' straightforward, functional approach to furnituremaking. There are also modern interpretations of the style that I like, such as the work of furnituremaker Thomas Moser of Maine. It was necessary to adapt the dimensions of the wardrobe to fit the client's room and accommodate the amount of clothing the cabinet would have to hold. To help visualize how the cabinet would relate to its surroundings, I made a perspective sketch of the room, including a view of the ceiling and other bedroom furnishings. I chose to make the wardrobe 78 in. tall and 42 in. wide to fit harmoniously with the client's large bed, and I made the wardrobe 22 in. deep to easily handle bulky winter clothes hung on a clothing pole inside. Pleased with the proportions of the cabinet, I went back to the drafting table to work out the small details, such as the cutout base beading around the door frames, the profile of the crown molding and the joinery to hold the cabinet together. Because solid wood would be used throughout, I had to design the cabinet to allow for expansion and contraction. In addition to frame-and-panel doors, this meant making a frame-and-panel back and a sliding joint for attaching the crown molding to the top of the carcase. The top, bottom and sides of die carcase would consist of edge-glued boards joined with dovetails at the top and pin tenons at the bottom. A face frame would be glued to the front of the carcase, and the back frame would be fitted into a rabbet in the carcase sides. I also shaped door, base and crown molding details to give the cabinet individuality.

The design of the author's cherry wardrobe cabinet, above, is based on a traditional Shaker piece that he modified, adding his

own details to the doors, base and crown molding.

By changing the position of the two movable rails that guide the router, the author works his way across the bottom of the carcase side, chopping mortises for pin tenons that will join the sides to the bottom of his wardrobe cabinet. guided by a homemade jig, as shown in the smaller photo above. The jig is a square template made up of two rails and two adjustable fences. A plunge router is set into the square, the bit is plunged and the router is then moved around inside the fenced area (router base bearing against the fences and rails) until the mortise is completed. The two adjustable fences, pinned into place with removable dowels, are repositioned for each of the seven mortises across the width of the two carcase sides. The routed mortises have rounded corners that must be squared up with a chisel. Next, I cut the multiple tenons on the ends of the carcase bottom. After marking these tenons by transferring lines from the mortises, I bandsawed away most of the waste, then pared the tenons for a tight fit into their mortises. As an alternative, you can use a router to rough-cut the shoulders, then pare them with a chisel. Either of these

Carcase constructionI began by building the basic carcase. After edge-gluing several narrow boards to make the sides, top and bottom, I cut the dovetails at the top of the case by hand, using a chisel, mallet and dovetail saw. You could also use a router and dovetail template. Next, I laid out and chopped the mortises at the bottom of the sides to accept the multiple tenons on the carcase bottom. These mortises were cut with a straight bit in a router

ing on p. 58. The shelf is notched so the edge of the shelf is flush with the front of the stile. Originally, I'd planned to let the base of my wardrobe rest on the floor, with a decorative molding at the bottom edge. However, I saw another base treatment I liked in the Ketchum book and adapted this base to fit my wardrobe. With this new base, the carcase is cut out to make four feet, with gentle convex curves coming up from the floor along the insides of the feeta pleasant visual effect.
Drawing in scale, I experimented with various curves until I came

up with one I liked. I scaled up this curve on graph paper to make a full-size plywood template, which I used to mark out the cabinet sides. After roughing out the curve with a sabersaw, I clamped the same template to the cabinet side and routed the final shapes with a piloted straight bit bearing against the template. Creating the same curved feet on the front of the wardrobe involved modifying the cabinet's otherwise straightforward face frame. I used a typical face frame, assembled with mortise-andtenon joints, on the front of the cabinet and a frame-and-panel

back. But the stiles weren't wide enough for the same curve I had
cut on the cabinet sides, so I glued in.-wide ears to the inside edges of the stiles, providing stock for the curved feet. I cut the bottom rail shorter than the top rail and tenoned it to the ear pieces instead of the stiles. If I'd made the bottom rail extra wide to accommodate the cutout, I would have created a wood-movement problem by joining a 6-in.-wide rail cross-grain to the face frame's stiles. Gluing up a cabinet of this size was quite a task in my small shop. The cabinet's size stretched every clamp in the shop to its

ness of each panel's edge by running the panel vertically by a straight bit. A fingerboard clamped to a spacer provides the pressure needed to stabilize the panel as it's fed. This fingerboard also prevents the climb-cutting bit from self-feeding the panel.

The author makes raised panels for the wardrobe doors on the router table. In the first step, shown above, he reduces the thick-

capacity. The carcase dovetails were tight enough that they didn't
need clamping, but the pin tenons needed to be pulled into their

mortises with pipe clamps. A 2x4 I had carefully bandsawn into a

bow shape (convex surface toward the cabinet) was used as a caul to distribute clamping pressure across the sides. After the carcase was dry, I glued on the already assembled face frame. Though plywood is a suitable and more-often-selected choice for the back of a large cabinet, I decided to stick to the more traditional solid-wood frame-and-panel back. This frame consists of three stiles and four rails mortised and tenoned together, dividing the back into four panels. The panels themselves are flat and only in. thick, the same thickness as the grooves in the frame. The entire glued-up back recesses into the rabbet cut in the back of the carcase earlier.

Hagood uses a homemade beading plane to detail the inner edges of the doorframe. The mahogany plane has two nonsymmetrical
bead-profile blades, one for cutting in each direction. A piece of

Building frame-and-panel doors After measuring the faceframe opening, I made a pair of door frames that fit snugly into the
opening. The frames are joined with haunched mortise-and-tenon

along the frame's inner edge for a straight cut.

quarter-round molding on the underside of the plane guides it

joints, and each member has a -in.-wide groove, cut with a dado blade on the tablesaw, to hold the panel. The bottom door

rails are wider than the top rails, to overcome visual foreshortening, which makes bottom rails look narrower than they are. I let

operations can be time-consuming, but I don't know a faster way to make pin tenons or an alternative joint that has the same integrity. As with most wardrobes, mine has a high shelf for storing clothing above the hanging garments. The -in.-thick shelf joins the

the stiles run long to prevent splitting during mortising, then

trimmed them to length after the doors were assembled.

I raised my panels with a router, homemade router table and

two different bits. The first operation was to reduce the thickness

inside of the carcase by sliding into a simple

in, by

in. dado

at the panel's edge with a

in. straight bit. The panel's edges were

plowed into the cabinet sides. The dado is the same depth as the rabbet cut along the back edges to accept the frame-and-panel back. I cut the dado with a in. straight bit in a router, using a straight board clamped across the cabinet side as a fence. Because die 4/4 shelf must span the width of the 42-in. cabinet, I supported the middle of the shelf with a lap joint on the in.-wide center stile. Dovetails at the top and bottom of the stile connect to the front edges of the carcase top and bottom, as shown in the draw-

routed with the panel on edge and held tight against the fence by a

fingerboard clamped directly above the bit, as shown in the top

photo on this page. To produce the cleanest cut possible, especially while shaping the panel's endgrain, I climb-cut the panels, feeding
them into the bit counter in the same direction it was spinning. If

you try this, make sure to take several shallow passes on each edge,
to prevent dangerous self-feeding, which can occur when climbcutting. The pressure of the fingerboard also prevents self-feeding,

and it shields your fingers from exposure to the whirling bit. After the panel edges were thicknessed, I cut a cove to finish the

panel raising. I used a

in. core-box bit in the router, shaping the

panels on the router table. I laid the stock flat on the table and shaped the cove in a few passes, raising the bit each time until its tip just contacted the flat in the panel edge cut from the previous operation. I then sanded the panels and glued up the door frames with the panels in place. The next step is to detail each door with a cock bead around the inner edge of the frame. I originally scratched the bead with an old scraper blade filed to the correct profile, but I was unhappy with the fuzzing and tearout. I chose instead to make my own beading plane, shown in the bottom photo on the facing page. The plane is designed to cut into the corners from two directions for work on an already assembled frame. (You can plane the bead in the frame edges prior to assembly, but you need to stop the beads

on the stiles where they meet the rails.) One of the plane's blades
is a standard cutter from a Stanley #55 Multiplane set. The other
cutter, a mirror image of the first, is made by modifying a straight

well as the cove-and-bead crown molding, provide the author's wardrobe with a distinguished degree of visual interest.
surfaces smooth, I cut the length of crown molding into three pieces and joined the mitered corners with splines and epoxy glue.

The beaded inner edges of the door frame and inlaid corners, as

plane blade. Both blades are held in place with removable wedges,
making it easy to lower one cutter and retract the other, depending on the direction of the cut. A strip of molding on the underside of the plane body acts as a fence to keep the cutter parallel to the frame's edge as it cuts the bead. Traditionally, the corners of a beaded frame were carved so the beads look mitered into one another. This is because the beading plane or scratch stock can't shape the bead all the way into a corner. Instead, I glued an end-grain dowel precisely into each corner, to serve as a return for the bead detail and to give the cabinet an
original touch, as shown in the photo above, right. I made a drilling

Crown molding has been traditionally nailed or glued on old

cabinets, but this can create problems. The molding, attached crossgrain on the cabinet sides, loosens over time because of expansion

and contraction of the cabinet's sides. To compensate, I attached

the assembled crown to the cabinet's side with a sliding dovetail joint, as described in "Cross-Grain Constructions," FWW #72, p. 80.

This allows the cabinet sides to move while holding the molding
securely to the cabinet. The crown is fastened to the cabinet sides with two 3-in.-long

guide by boring a hole through a scrap and gluing a tiny triangle to the bottom to reference against the frame's inside corner. Clamped in place, the guide established the precise location of the

male dovetail strips-one screwed to each cabinet side. These

strips slide in female dovetails cut into angled blocks glued to the back of the molding. After the male strips were in place, an assistant held the molding while I screwed the front molding on from the inside of the cabinet. Then, I applied hide glue to the female dovetail blocks and slid them onto the dovetail strips from the back. I held the components in alignment until the glue set. This procedure is not as difficult as it sounds, especially if you're good at cutting sliding dovetails. In addition to alleviating wood-movement problems, this method avoids difficult clamping procedures
and ensures that the molding is perfectly aligned. The whole

hole and kept the bit straight as I bored each

in. corner

hole with a hand drill. A tape flap stuck to the bit told me when the hole was deep enough. The dowels for the corners were made from cherry scraps, using a plug cutter in the drill press. Each dowel was glued in with epoxy, then trimmed and sanded flush with the surface of the door frame.

Crown molding-For this highly visible detail, I chose the fanciest piece of cherry I had. To shape the crown's cove profile on the
tablesaw, an improvised fence was diagonally clamped across the saw

table and a length of stock long enough for the entire crown was

passed over the blade, which is raised a little at a time. By varying

the angle of the fence and depth of cut, you can produce many different cove shapes. I experimented until I had a profile that mimicked the curve of the feet. After coving, I angled the molding's edges as detailed in the drawing on p. 58. The bead detail where the crown molding meets the cabinet is

crown is removable: This makes finishing the cabinet sides easier and lightens the heavy wardrobe, in case it needs to be moved. After final sanding and finishing (I used Watco oil to bring out

the cherry's color, but you can use any finish you like), I attached
the knobs, mortised in the locks on the doors and hung the doors
on the cabinet. Keeping with the Shaker style, I turned two small,

the same as the bead on the doors, adding visual detail and making any irregularity in the seam between the cabinet and molding less obvious. To hold the molding at the correct angle for beading,

plain cherry knobs. Each has a dowel turned on its back that is glued and wedged into a hole drilled through the door stile. I used standard in.-long butt hinges to hang the doors, three per

door, and mortised them into the door frame and face frame for a
clean fit with very little gap around the doors. Finally, I mounted a removable clothes pole inside the cabinet just below the shelf. My client is pleased with the wardrobe, and so am I. The adaptations I made to the design seem to mediate well between the room's ornate furnishings and the cabinet's simple Shaker origins. While the extreme austerity of original Shaker designs is not always completely satisfying to me, I enjoyed reinterpreting the Shaker

I first cut some small plywood triangles and hot-glued them to the
bench in a straight line. I then hot-glued the noncoved side of the crown molding to the triangles so the molding's bottom edge was facing up and level. The narrow edge gives the beading plane little support, so it is a bit tricky to get a straight bead. To stabilize the plane, try laying one hand on the molding alongside the plane as you walk down the length of the molding taking the cut. It'll take several passes to get the full bead profile, but in the end, this

wardrobe design. But, I'll always acknowledge the Shakers' basic

premise: "Keep it simple."

great-looking detail makes it all worthwhile. After scraping the

cove smooth with a curved scraper blade and sanding its show

Tom Hagood is a woodworker in Birmingham, Ala.

Starting from scratch with nuts-and-bolts joinery

Building a Pool Table

he distinctive crack of the break, brightly colored balls rolling across the expanse of green and the "plop" of the ball dropping into the pocket-these are the sights and sounds of satisfaction to pool afficionados. But, few of them have the satisfaction of sinking balls into pockets on a table they've built themselves. Building a pool table is somewhat intimidating, because of the size of the table and the weight it must support. In addition to the 400-lb. slate playing surface, a pool table must be able to support a person sitting on the edge while attempting a behindthe-back shot. On the other hand, aside from the compound angles, building a pool table doesn't involve anything that's beyond a competent weekend woodworker with the gumption to take on a large project. The table described here is constructed much like many highend commercial pool tables (see figure 1, above). Each section of the three-piece slate is glued to a wood liner and then screwed, but not glued, to the main table body, a box whose sides incline down and in at 15. A 2x8 leg support is screwed and glued be-

tween the sides at each end of the box. The four legs are attached to these supports with lag screws or hanger bolts. Bolted on top of the slate are six rails, which support the cushions and pockets. An apron, or blind as it's called on pool tables, surrounds the rails and covers the edge of the slate assembly. The choice of wood, the carving or other decoration on the table and the style of legs can be custom designed to suit an individual's tastes. But one thing is certain: The table must fit the slate, so buy the slate before you begin working. The slate comes from Italy or Spain, where it is cut to size and the pocket cutouts and holes are drilled for assembly hardware. The chart on the facing page lists common slate sizes, weights and playing areas. Unless you're buying 100 or more sets, you'll need to locate a local pool-table dealer in your area who is willing to sell you a single set. Don't buy slate that is not predrilled. You'll also need a set of #6 leathercovered pocket irons, type K-66 rubber cushions and facings, plastic sights for the "dots" in the rails and various hardware. After the table's built, you'll need cloth to cover the slate, balls, a triangle,

pockets, which are indicative of an antique-style table. The legs were purchased from a company that specializes in carving pool-table legs, and the relief-carved blinds are from a millwork-supply house. Once you've located the slates and accessories, there's nothing

This walnut pool table, built by the author, has sloping sides and leather-and-fringe

about building a pool table that's beyond the abilities of a competent weekend woodworker.

cues and a bridge. The retail price of the slate and other supplies

will come to about $800. Methods for covering the table and rails with the cloth are explained in the sidebar on p. 43.
Slate liner -The slate liner reinforces the brittle slate, anchors the staples holding the cloth and distributes the effects of the shims used to level the slate. You should use spruce or poplar 1x8 liners on the long sides of the table so there will be enough stock to accommodate the pocket cutouts. The rest of the liner can be from 1x4s. Be sure all the liner material is planed to the same thickness, or you'll be adjusting for the difference later. Butt the pieces together and glue the liner to the underside of the slate with Titebond glue or contact cement, as shown in the top, left photo on the next page. Hold the liner back about in. from the edges where the slates butt so they can be pushed tightly together. Cut out the pockets in the liner by running a sabersaw blade along the slate cutouts. The - or l-in.-dia. holes along the slate's edge are for the bolts that will be used to attach the rails. These holes should be drilled
Drawings: Kathleen Creston

through the liner, along with the smaller countersunk holes for

screws to hold the slate to the body.

Legs-There's no end to the design possibilities for pool-table

legs: They can be square, tapered, turned or carved. The only limitation is that you leave a way to bolt or screw them to the table. The legs on the table in the photo above were purchased from a dealer who specializes in duplicating intricate carvings (see sources of supply, p. 43). The cross section in figure 1 shows the basic construction of simple, square legs. An official pool table should measure in. from the bottom of the leg to the top of the slate. For the table described here, the legs should be about in. tall.

Body frame-The body sides are made from -in.-wide 6/4 flat and straight hardwood lumber. If you need to plane the pieces to flatten them, leave them as thick as possible. To determine the frame's dimensions, measure the slate's overall size and subtract 7 in. from each dimension, so the slate will overhang the body by in.

The -thick poplar slate liner is glued with Titebond glue and clamped to the underside of the slate. The slate's sides are lined with 1x8s to overlap the pocket cutouts, then 1x4 crosspieces are

To complete the body, the slate frame is screwed to the top edge of the body frame, and the crossmembers are installed to support the

butted between them and held back

that will butt together.

in. from the slate edges

joints in the slates. Glue blocks, screwed below the crossmembers,

add support. The slate frame stops short of the corners to allow pocket clearance, and a block ties the comer together.

all around. The incline of the sides is common, especially on antique tables. Joining these angled pieces, however, requires compound miters, so I initially crosscut the parts a couple inches longer than the final size to allow a little leeway for cutting these miters. I prefer to cut the compound miters with a sliding tablesaw. However, a radial-arm saw will do. The arm of the saw should be

swung 17 to the right of its normal, square position, and the blade
should be tilted to Because the blade may wander in bevel cuts, I cut each end twice: I first cut about in. past my mark, then slowly trim to the line. You can cut the compound angle on both ends of each piece without changing the saw-arm setting. First, place the board to the left of the blade, with its bottom edge against the fence and its outside face down on the saw table. After cutting this end, roll the board over and slide it to the right of the blade. Measure and mark the long, top edge of the board and cut to length. After the angles are cut, rip the edges of all four box pieces

at a 15 angle so the edges will be parallel to the floor when the

slope-sided box is assembled. If you're using a scalloped bottom

edge, shown in the photo on the previous page, saw the pattern now;
be sure to leave the first 10 in. from each end square for the legs.

Assembling the body frame -Many commercial pool-table builders join the parts with nails, which are set and hidden with wood filler or molding. I prefer the additional strength provided by screws; the wood plugs covering the screws also add a nice detail. I drill plug holes for the screws that secure the corners and leg supports at the locations shown in figure 1. Then, I drill pilot holes through for the screws. I sand to 120 grit with a belt or stroke sander and use a vibrating sander after assembly for finer sanding. I assemble the body frame on custom-made 20-in.-high sawhorses with 2x8x48-in. tops that can support the weight of the table and slate. The wide tops make it easy to assemble the sides, and the low height makes it easier to lift the heavy pieces of slate onto the

edge of the body, bandsaw the leg support back away from the scalloped portion of the end so the support doesn't show. Screw and glue the supports in place, running -in.-long screws through the predrilled holes in the body frame and into the leg supports. Glue wooden plugs in all the screw holes, and sand them flush. To drill for the lag screws or hanger bolts that attach the legs, make a plywood template the same size as the top of the legs and drill four evenly spaced -in.-dia holes. Place the template in the corners of the leg supports and drill through the template and the support. Use the same template when drilling the tops of the legs. Before I turn the table over, I round over the bottom, outside edge of the body and the four outside corners with a -in.-radius piloted router bit. With a helper, I turn the body over and reinforce the leg supports by gluing and screwing in hardwood glue blocks cut from scraps from the table sides (see figure 1). A frame to support the slate completes the table's body. It's made from 3-in.-wide 6/4 stock and should overhang the body frame by in. all around. Cut off the corners of the end pieces at 45 for pocket clearance, or let the parts come up short of the corner, as shown in the photo above, right. Screw and glue this frame to the top of the body. Install two crossmembers to support the joints in the slate. The crossmembers should be at least in. by in. (a standard 2x6) and cut to fit on edge from side to side. Glue them to the frame, and run a long screw into each of their ends from the outside of the frame. Screw and glue a block under the ends of each 2x6 crossmember for support. I usually also glue in side blocks to prevent twisting. With a sabersaw, cut out 6-in.-dia. arcs tangent with the body for the side pockets. Finally,

round over the inside edges on the top of the frame so it's easier to insert wedges between this frame and the slate liner when leveling the slate. Now, set the slates on the body and slide them

around until they're tightly together and centered on the frame. The railsThe rails, more than anything else, make a slate-top table
into a pool table. They can be made from a single piece of clear,

completed body. Set the sawhorses where the body will be assembled, and level them in place, shimming beneath their legs, if necessary, to ensure that the body can be assembled square and true. To assemble the four sides, I place the parts upside down on the sawhorses and brush a liberal amount of glue on the miters of two adjacent ends. Then, I carefully hold the corner in alignment and use a power screwdriver to run a #10 screw into one hole in the side and one in the end. I draw the corner tightly together with the rest of the screws, being careful not to strip out the end-grain holes. Repeat this process until the body is complete. The next step is to cut and install a kiln-dried 2x8 leg support at each end of the body. Crosscut the ends with opposite 15 bevels so the supports fit snugly side to side. Then rip one edge of each at 15 to fit against the ends of the body. If you scalloped the bottom

straight hardwood, but hardwood and softwood laminated rails,

shown in figure 2 on the facing page, reduce the possibility of

warping and provide softwood for stapling on the cloth.

The most critical rail dimension is the cushion height. According

to the Billiard Congress of America, the point of the cushion

should be in. from the bed of the slate on a standard table

with

-dia. balls. The balls tend to climb a lower rail and be thickness of the rail in conjunction with the 15 angle of its

forced down under a higher rail. This height is determined by the

face. If you vary the thickness of the rail, you must adjust the face

angle to provide the proper cushion height, as shown in figure 2.

To make the laminated rails, you'll need six

softwood boards (I prefer poplar because of its stability) and six

hardwood boards. Mill enough for one extra rail

angle in the rail's horizontal plane creates the correct-size opening

for the pockets. The angle in the vertical plane serves to extend

for tool setup and test cuts. Saw out a

rabbet

the cushion's nose around the ends of the cushions.

Cutting the pocket angles is much like cutting the opposite compound miters for the body: Cut one end, roll the board, move it to the other side of the blade and cut the other end. However, all the pocket-angle cuts should begin at the back of the feather-strip groove. So, cut the first rail to the right of the blade, with it's outside edge against the saw fence and its top side up so you can see where to begin the cut. The opposite cuts are made to the left of the saw blade with the top side of the rail face down. This makes it difficult to line the cut up with the feather-strip groove, so line up the first rail you cut with the other rails, bottom to bottom, to mark

in each piece of the softwood by making two cuts on the tablesaw.

Glue the hardwood segments into this rabbet, and when dry, plane
the hardwood flush with the softwood. Begin shaping the rail by ripping the 15 and 45 angles along the length of each rail's cushion face. On its top, cut a groove about back

from the 15 angled face. This groove is for the feather strip, which
secures the cloth to the top of the rail. On the bottom of the rail,

saw a shallow the softwood as you cut a

rabbet where the rail cloth will be -in.-deep dado for the blind into the

stapled. Finally, set the dado blade to just skim the back edge of hardwood. Mold the top edge of the hardwood with a router or

the location of the top-side-down cuts. The end rails get the 52
corner-pocket angle cut at both ends (in opposite directions), while

shaper to add a detail and break the sharp corner.

Use the chart on p. 39 to determine rail lengths for your playing area, and crosscut the rails to length. These lengths are figured for #6 pocket irons, so be sure that's what you use. Bore holes in

the four side rails have one of their ends cut at 52 and the others
cut with the saw arm reset to 15 for the side pockets. After the

angles are cut, use a router to extend the shallow rabbet on the
bottom of the rails, along the angled ends, for stapling the cloth.

the rail ends; 2 in. should be deep enough for the pocket "ears" (see
the left photo on p. 42). Center the hole back from the point

The cast pocket irons are anchored in the ear holes with
by 1-in. bolts run up through the rail. Drill l-in.-dia. holes, deep

created by the 15 and 45 angles of the angled face and hand drill, using the bit's long shaft to align the hole.

down

from the top of the rail. I drill the holes with a spade bit in an electric Assemble the rails and pockets on the slate to check your play-

enough to countersink the bolt heads and washers, directly below

the threaded holes in the pocket ears. Then, drill a hole

through to the ear hole. The bolt should screw into the ears when
the pocket is snug to the rail, and it should have enough play for

ing area. Be sure the pockets are tight in the ear holes, and check
the whole assembly for squareness. Tape pieces of cushion rubber

minor adjustments during final assembly.

onto rails across from each other, using double-faced tape, and
measure the distance between their "noses." The measurements

Attaching the rails-There are several different ways to attach the

rails to the table: lag screws, mortised nuts, threaded inserts and

should be your desired playing area. If your measurements are long,

you can trim the ends of the rails to leave the proper playing area. Now you can cut the pocket angles on the rail ends with the radial-arm saw. The angles and sizes for the pocket openings are shown in figure 3 on the next page, The ends of the rails that border the corner pockets are cut with the blade tilted to 10 and

threaded discs. Threaded discs, often referred to as metal rail

plates in billiard-supply catalogs, are used most often, but they may

be hard to find. I'll describe how to use them, but the same basic
procedures apply for the other methods as well.

To locate the threaded discs, assemble the rails and pockets

tightly together on the slate. Center the assembly by lining the

the arm set at 52. For the side pockets, the arm is reset to 15. The

Inserting the sightsThe last step in preparing the rails is locating

the sights used to line up bank shots. Figure 3, above, shows their
placement. Divide the playing area's width by four, or its length by eight, to get the distance between the three sights on each rail. The

end rail has one sight in the center and one, the calculated distance,
on each side of the center. Place them halfway between the featherstrip groove and the back of the rail. The center of the side rails falls in the middle of the side pockets, so measure from that point Sights come in various sizes, shapes and materials. The round,

plastic ones are the easiest to use. Just drill a hole the proper diameter and deep enough that the sight will stand just proud of the
rail. Then, brush in a little glue and drive them in with a hammer

Above, left: The pocket 'ear' is inserted into the hole in the rail's end, and a bolt is then run up from the bottom of the rail into the
photo above, right. The by 1-in. bolt anchors the pocket ear in the rail, and the rail-bolts screw into the threaded discs under the rails to secure the rails to the slate. The cloth is stapled and trimmed closely behind the pocket angle for a

and a wood block, to protect the sight. Once the glue is set, sand
the sights flush using 120 grit on a belt sander. The blinds-The blinds, or aprons, trim the rails and hide the edge of the slate assembly. They're about 4 in. wide and the same length as the rail they'll be attached to. If you bandsaw the ends

ear's threaded hole. The rails and pocket are upside down in the

tight fit between the rail end and the pocket leather.

with the curve shown in figure 2, they blend nicely into the pocket
fringe, but you'll have to cut them long to allow for this curve. The blinds fit into the dado in the bottom of the rail and are held in

side pockets up with the slate cutouts and measuring the rail assembly's overhang all around the slate. Eye the side rails to see that they are straight, and double-check the whole assembly

place with blocking, which is glued to the backside of the blind

and screwed to the underside of the slate liner. Glue two blocks to the back of each rail, making sure they won't be in the way when

for square. Then, reach under the slate liner, and with a pencil,
trace the holes in the slate onto the bottom of the six rails. Disassemble the rails, turn them over and mark the centers of the

tightening the rail bolts. Drill an angled pilot hole in each one to
screw the block to the liner. This makes for easy disassembly of the blinds when the rails are recovered. Assembling the table-After the table parts have been stained and finished and the cushions have been glued to the rails and covered with cloth (see the sidebar on the facing page), you can set up the table. Place the table body upside down on the floor of

holes. The side rails, as well as the end rails, should be interchangeable with their counterparts.
Drilling for the threaded discs is done with a fence clamped to the drill-press table set for the holes' front-to-back alignment. First, countersink for the discs by drilling a shallow hole the disc's diameter. Then, change to a bit and drill 1 in. deep in the center of each countersink for the shaft of the rail bolt. Use one of the discs, its flat side facing and parallel with the cushion side of

the rail, to locate the pilot holes for the discs' attachment screws.
Then, screw the discs in place (see right photo this page).

the billiard room. Bolt the legs in place and then turn the table over. Level the table by using a carpenter's level and placing plywood and plastic-laminate shims under the legs. Next, set the pieces of slate in place and screw them tightly to the body frame. A screwdriver bit in a hand brace simplifies this process. It's not a good

idea to screw down the center of the slates along the joints unless you need to pull down a bowed slate, which is very uncommon. Run your fingers across the four points where the slate joints

escape through this opening. Lastly, slip the blinds up into the dado in the rails, and screw through the glue blocks into the slate liner.

meet the edge of the table to detect any difference in height. If

there's a difference, loosen the screws in the lower slate and raise

Fine-tuning the tableTo check the table for final leveling, hold
a pool ball between your thumb and first finger about 12 in. above

it by inserting a playing card or folded paper shim between the

liner and body. Repeat this until the three slates are even at the

one of the rails. The line from your thumb to your finger should
be perpendicular to the front of the cushion rubber. Drop the ball

table's edge. If your body frame is straight and true, the slates
should be too. To double-check this, stretch a taut string from end

so it will strike the slate and the front of the cushion simultaneously. It should travel across the table at 90 to the cushion; if the table

to end, near the edge of the table, anchored to nails in the slate
liner. Slip a coin or poker chip beneath the string at each end and use another chip as a gauge to test for equal clearance along the length of the slate. Next, the joints are felt their whole length. If one piece is lower, drive a by 6-in. softwood wedge, tapering up to in., between the slate liner and crossmember. One person can tap the wedge in place while a helper feels for the moment when the two pieces of slate are even. When the three pieces of slate are as even as possible, putty the joints with "Durham's Rock Hard"

is the slightest bit out of level, it will roll toward the low point. As
an alternate technique, grab a cue and shoot a ball slowly the length of the table, about 6 in. from the side rail. As the ball slows and stops, it will veer toward the low point, if the table's out of level. The severity of the ball's drift, using either method, will help estimate the thickness of the shim needed under the leg toward the low point. Lift the table from the end so both legs are off the floor, and have your partner place the shims under the proper leg. If you pick up only the low corner, the torque is sure to break the puttied joints between the slates. Repeat this procedure around the

(available from local hardware stores) and a wide putty knife. Any
gouges, scratches or screw holes in the playing area should also be

table until the balls show no drift.

puttied and sanded smooth. After the putty has hardened, use 120grit sandpaper on a block to gently sand off any excess. Putty and sand again if necessary, then brush the table clean.
The bed of the table can now be covered with the cloth, as described in the sidebar below. Then, the rails and pockets are assembled on the covered bed, as shown in the photo (right) on the facing page. The rail bolts and washers are inserted through the slate liner and slate and finger-tightened in the threaded discs. Then the pocket-and-rail assembly is centered on the slate, and the end rail bolts are tightened with a socket wrench. Sight down the side rails and adjust them until they are straight, then tighten them down. To test their alignment, roll a pool ball down the side rails. The ball should pass the side pockets without catching a point of the cushion. Nail or screw the loose ends of the woven leather pockets to the underside of the slate liner, making sure that the balls will not

Paul Bowman is a freelance writer and builds pool tables on special order in Vancouver, British Columbia, Canada. For an article
on turning a pool cue, consult FWW #59, pp. 66-69.

Sources of supply
so find a local dealer or pool-table manufacturer who will sell you you'll need, including slates, from the following supplier:
the slates and accessories. If that fails, you can order everything
Tucker's Billiards, 3381 Ashley Phosphate Road, North Charleston, S.C. 29406

Wholesale billiard-supply houses usually won't sell to individuals,

For machine-carved legs and decorative blinds:

Adams Wood Products Inc., 974 Forest Drive, Dept. 1-3, Morristown, Term. 37814

Covering the rails and slate bed

Experience has convinced me that the best cloth for pool tables is a blend of 75% wool to 25% nylon, with a weight of 22 oz. per linear yard. A purist might prefer all wool,

by Eldridge Tucker
The cushion facings, made of laminated rubber and canvas, reinforce the cushion ends and cut down on the springiness of the cushion so the ball is not deflected too easily out of the pocket opening. Contact-cement them to the angled ends of the cushions and rails, flush with the beginning of the pocket
Trim the facing's other edges to the contour top and front edges of the facing with sandpaper, as shown in the top, left photo on the next page, so a corner won't tear the cloth. -in. by -in. softwood feather strips to secure the cloth to the rails. Lay a feather strip in the groove on each of the rails so it's flush at one end, and mark the other end for
angle and the top of the rail and cushion.

weight is well worth an extra couple dollars

around

but the nylon adds durability, and the heavy linear yards. Figure 4 at right

per yard. It comes 62 in. wide; you'll need

shows how to cut the cloth for an 8-ft. table with a playing area of 44 in. by 88 in. Adapt

this layout to the size of your slate.

of the rail with scissors or a knife. Break the

Attaching the cushions: After the finish

on the rails is dry, the cushions are glued to

should be flush with the top of the rail. If you use yellow glue, use masking tape to hold the rubber in place until the glue dries.
down against a board or workbench, align a knife blade with the pocket angles and slice through the rubber from base to nose to extend the angle. Dip the knife blade in water for a smoother cut.

the angled face with either contact cement or yellow glue. The top edge of the rubber

Covering the cushions: You'll need six

After the glue has set, trim the rubber to length. Hold the rail so the cushion nose is

length. Before removing the strip, make a

line near its center, extending onto the rail and cushion, as a reference for lining up the strip when the cloth is being secured, as
with a saw or chisel. Then, lay one of the

page. Remove the strip and cut it to length

shown in the top, right photo on the next

precut rail cloths so its edge lines up on the joint between the wood and the rubber. Place the feather strip on the cloth above the groove, align the centerlines and tap the center of the strip partway into the groove. Work toward one end, tapping the strip partway in and pulling a little tension on the

tinue to the side-pocket ends, but stop about 6 in. back from the corner-pocket ends.
While holding the feather strip down in the groove below the strip, as shown in the right

cloth in the direction you're working. Con-

groove, pull about 2 in. of cloth through the

photo this page, to create a "pucker" so you can stretch the cloth around the angled cushion. Drive the last 6 in. most of the way into the groove to secure the cloth. With a wood block, tap the length of the strip into the groove so it's just proud of the rail's surface. Trim the excess cloth by slicing with a knife against the cushion side of the protruding strip. Flip the cloth over the feather strip and cushion, and with a clothwrapped block, tap the strip flush with the rail's surface. Stretch the cloth tightly over the corner-pocket angle, and staple it to the rail tight behind the facing and on the rail's bottom near the edge. The photo (right) on p. 42 shows where to staple. On the sidepocket ends, fold and tuck the cloth like you would wrap a gift, with the open part of the fold on the lower edge of the cushion. Pull other ends. Starting at the middle of the rail, stretch the cloth tightly over the cushion and staple it every 2 in. in the shallow rabbet on the rail's bottom. Trim the excess along this

The cushion facing is contact-cemented to the angled rail end and cushion, then trimmed with scissors or a knife. The facing's sharp edge is eased with a sandpaper stretched tightly over the pocket angle.

The feather strip is driven into the groove to within 6 in. of the corner-pocket ends.

Here, about 2 in. of the cloth is pulled the comer-pocket angle.

block so it won't tear the cloth when

through the groove to create a wrinkle or 'pucker,' which helps stretch the cloth over

the cloth every 3 in. into the edge of the slate liner. Work toward the corner pockets, pulling the cloth slightly in the direction
you're working. Staple to within 2 in. of the

the cloth tight and staple like you did on the

pockets, to leave some play for stretching the cloth over the pocket cutouts. smoothly over the pocket cutouts, you'll need to make radial cuts in the cloth within cloth at the pocket and pull it down gently to reveal the outline of the pocket cutout. With a razor knife, cut in the center first and
then once or twice on either side to create

Trim the overhanging cloth around the table if it's long; if there's not too much excess, the blinds will hide it. Finally, cut circles out of the cloth with a knife for all the rail-bolt holes. If
you just make an X-cut, the cloth will get

caught in the bolt's threads. Now assemble the

rails and pockets, and bolt them in place on

Pocket cutouts: To stretch the cloth

rabbet and close to the staples on the ends.

slate bed. Start by lining the pocket cutouts with by 12-in. strips of cloth contact-cemented to the edges of the slate and liner. Brush the slate clean of any small particles left from sanding the putty, and lay the cloth down so it overhangs the slate evenly. Figure 5, below, shows the steps for stretching the cloth over the slate bed. Always begin at the center of a side and staple

Covering the slate bed: After the slate's joints have been puttied, you can cover the

the pocket areas, as shown below. Grasp the

the slate bed as described in the main article. All that's left is to place the center spot for locating the rack of balls for the break. I run

triangular tabs. Always cut toward the cloth's edge. Again starting at the center, pull down
firmly on the tabs and staple them in a neat

will be racked, and one between the middle sights on the end rails. Holding a level alongside the strings, I mark on the cloth the point where the two strings intersect and place

a string between the middle sights on two opposite side rails, at the end where the balls

the self-adhesive spot. Now you're ready.

Rack 'em up.

line near the slate liner's bottom edge. Trim the cloth as close to the staples as possible, to avoid a loose flap that a hand may catch when retrieving the balls from the pocket

Eldridge Tucker has been in the pool-table

business for 30 years. His shop, Tucker Billiards, is in North Charleston, S.C.

Fig. 1: Pencil-post bed

The Pencil-Post Bed

by Christian H. Becksvoort

Jigs for machining tapered octagons

he pencil-post bed is a classic form that has been in use for centuries, The high posts were originally designed to hold a canopy frame, or tester (pronounced teester), as shown in the photo at right. Before central heating, a canopy was standard equipment and was often accompanied by thick, quilted curtains on all four sides. Drawing the curtains at night created a room within a room designed to conserve warmth and provide privacy. In warm weather, the heavy curtains were replaced by fine lace netting, which offered protection against insects. Today, the testers are covered with either net or lace canopies or left bare, a decorative reminder of earlier times. I usually turn finials for the tops of the posts so the customer has the option of removing the tester entirely and capping the posts with the decorative turnings. Traditionally, bedding was supported by a rope mattress woven between the bed rails. Early on, the rope also held the bed parts together, but around 1750, builders switched to bed bolts, which could hold the bedstead securely together even when the rope mattress started to sag. The 6-in.-long bed bolts I use extend through the posts and thread into nuts that are trapped within the side rails and the head and foot rails. On conventional, low-post beds, the headboard and footboard are usually glued up and knock-down hardware is used for assembling the side rails. However, it would be difficult, if not impossible, to move an assembled headboard with tall pencil posts upstairs, through doors and around tight comers. Therefore, a pencil-post bed must be made to knock down completely into individual members: four posts, two side rails, a head rail and foot rail, a headboard and a six-piece canopy frame and/or four finials. Besides the eight bed bolts that secure the rails, the bed is held together by the mortises and tenons where the rails and headboard join the posts, and the lap joints on the tester. Before building this or any bed, check and recheck the actual box-spring dimensions, because they are anything but standard; few people today would want a bed that can accept only a woven rope mattress. I build pencil posts only in full- or queen-size, purely for reasons of proportion. The single-size seems too narrow and tall, while the king-size begins to resemble a cube. Pencil posts-The evolution of tall bedposts has been from massive square posts to thinner and tapered posts, then to even more delicate, eight-sided tapers, the shape we now associate with pencilpost beds (though most pencils today are six-sided). In addition, I prefer to taper both the upper portion of the post and the shorter, lower portion below the rails. I used to bevel the corners of the tapers with a handplane or spokeshave to create an octagon. This was time-consuming and presented the risk of tearing out the grain as well as the problem of shaping a smooth transition from

A pencil-post bed with traditional tester or canopy frame is shown above. The eight-sided, tapered posts are mortised to receive the four bed rails, which are held in place with bed bolts concealed behind the brass bolt covers. There are no glued joints, so the entire bed can be disassembled for ease of transport. the bevels to the square portion where the rails join the posts. To simplify this, I devised the two simple jigs shown in figure 3 on the next page. I use the jigs in conjunction with a tablesaw molding head, but they can be used with a shaper or a table-mounted router and a 45 cutter (you'll need 1 in. of cutting surface). The posts are first roughed out on the bandsaw, and the tapers are cleaned up on the jointer. Then, the corners of the tapered portions are beveled into octagons by running them over the molding head, using one jig for the upper portion and the other for the lower portion. Cutting the bevels by machine saves time and virtually eliminates tearout and, because of the cutter's arc, automatically leaves a smooth transition from the bevels to the square portion of the post. This results in a completely shaped post right off the ma-

Above, the author bevels the base of the post with the smaller jig. The larger jig pushes the post past the cutter; the smaller jig is pushed with the post; both jigs help to safely hold the post against the fence. The lines on the fence and the post indicate where to stop the cut.
chine, with nothing left to do but sand out the machining marks. To make the posts, I begin with 12/4 stock up to 12 in. wide by 7 ft.
or 8 ft. long. Since 1 don't have a jointer wide enough to true up

and the height of the 45 cutter in the molding head are set to expose

The photo at left shows the setup for beveling the upper portion of the post. An auxiliary fence
in. of cutting edge.

one face of these planks, I lay out the posts on the rough wood,
using a pattern cut from in.-thick Masonite. My pattern is 80 in.

tall, which, adding 3 in. for the finial, gives a total height of just
under 7 ft. The untapered portion of my posts, the section that will contain the mortises for the rails, is 8 in. long by in. sq. and begins 11 in. from the bottom of the post. The post tapers from this in.-sq. section to in. at the top and in. at the bottom (see
figure 1, p. 32). Because of this taper, you can save a fair amount of

After all four posts are laid out, bandsaw them apart, leaving them slightly oversize. Then, square up two adjacent sides of the in.-sq. by 8-in.-long section on the jointer, working to the pattern line on the side that's laid out. You only need to true up the area around the square sections, because you'll be bandsawing tapers on the rest of the length of the post. Then, bandsaw the top and bottom tapers on the posts, leaving the pattern lines to work to when cleaning up the shape on the jointer later. Crosscut the posts to length on the pattern lines, then lay out the pattern on the adjacent side of the posts that was previously trued on the jointer. Bandsaw
away the waste, then take the posts to the tablesaw and clean up the bandsawn sides of the in.-sq. sections by ripping them in.

wood by reversing the pattern as you lay out the posts side by side.

thick. Take the posts to the jointer and clean up the tapers to the pattern lines. One final pass on the jointer will plane off the sawmarks from the square sections and bring them to in.
To complete the shaping of the posts, you must bevel the corners of the tapers to make them eight-sided. The beveling jigs are

designed to cut bevels that increase in width as the post increases

in width, so that at any given point, all eight sides are equal in width, forming an equilateral octagon in cross section. Figure 2 on

the facing page shows my method for determining the dimensions

of the octagonal cross section at the top of the posts. On a piece of

paper, draw squares to represent the dimensions of the post's top, bottom and square section, and use this method to determine the
octagon's size at each of these points. You will have an octagon

with sides inscribed within the square representing the tops of your posts and an octagon with in. sides for the wide bottoms of the posts. Draw these octagons 011 the top and bottom of one of the posts, to use as a reference when setting the fence of the tablesaw and the height of the cutter. The -sq. portion of the post will give you an octagon with in. sides. This will be the maximum length of the cutting edge that should be exposed when the molding head's height is set. The jig for beveling the long, upper portion of the posts, shown in figure 3 on the facing page, consists of a cradle glued to a tapered shim that gradually lifts the post's length off the machine table, thereby yielding a tapered bevel cut. The shim is about 4 in. wide and 51 in. long, and tapers up to in. thick. It's glued under the 62-in.-long 1x4 portion of the cradle so about 2 in. of the shim's width overhangs the full length. This overhang will lift and support the post, although part of it will be ripped off with the first pass on the tablesaw. The lower portion of the cradle is bandsawn from a 32-in.-long 1x6, to conform to the shape of the lower 30 in. of the post and to wrap around the bottom end of the post, acting as a push stick. The two parts of the cradle are screwed together,

With the shaped post supported by a simple V-block, Becksvoort beltsands the machining marks from the tapered bevels of the octagon. A smooth, continuous motion and a light touch with the nose of the belt sander are requirements for sanding the curving

transition from bevel to square.

Push the jig and post, with its tapered side running along the fence as shown in the lower, left photo on the facing page, over
the spinning molding head. When the line on the post indicating

the top of the square section meets the line on the fence, pull the post and jig away from the fence. Repeat this procedure for the other three corners to form a tapered octagon on the top section

of the post. Bevel the bottom of the post in the same way with the
smaller jig, as shown in the lower, right photo on the facing page. Stop the cut when the line at the bottom of the square portion

lines up with the line on the fence. Repeat the cuts until the four
corners are beveled. Once all four posts are beveled top and bottom, you've completed the most difficult part of the bed.

with the shorter portion on top, which raises the push-stick part of
the cradle enough to contact the bottom of the post completely. The
smaller jig for beveling the bottom of the post is built on the same

I sand the machine marks from the bevels with a belt sander.
With experience, a light touch and a confident hand, I've gotten so I can use the front wheel or "nose" of the belt sander to sand the
curving transition without making it wavy (see the photo above). If

principle, except in this case, the post pushes the jig instead of the jig pushing the post. This variation makes it easier and safer to control the post for the short, lower portion of the post and vice versa for the
longer, upper portion. The shim for the smaller jig is 4 in. by 11 in.,

you don't feel comfortable using a belt sander for this, a cabinet scraper and a hand-sanding block will do the trick.

tapering from in. to zero. It's glued to a 13-in.-long 1x4, bandsawn to fit the post's taper and to wrap around the post's bottom. The setup I use, with a tablesaw molding head fitted with shopmade 45 bevel cutters, is shown in the lower, left photo on the

Rails, mortises and bed bolts Traditionally, the rails on pencilpost beds were nearly square ( in. by 3 in. was common), to
resist the inward pull of the rope mattress and the downward

facing page. The same basic procedures would apply to a router

table or shaper. Our earlier calculations told us that the sides of
the octagon at the largest portion of the post are in. wide.

weight of the bed's occupants. However, today's beds rely on a box spring that only bears down on the rails, so the more familiar, 1-in. or 2-in. by 6-in. board-on-edge serves nicely. I mill my rails
from straight 8/4 stock to a finished size of in. by 6 in. The 6-in.

So, we now set the cutter and an auxiliary fence to expose in. of the cutting edge so we can cut the tapered bevel in one pass. To double-check this, place the post, with the octagons drawn on

width hides all but 1 in. or so of a standard box spring. The mortises for the rails are in. and centered top to bottom and side to side on two adjacent faces of the square section

its ends, on the larger jig so that the tapered, top portion of the
post will run along the saw's fence, as shown in the photo. Raise the cutter so that at the peak of its arc it will bevel the corner right up to the side of the octagon. Slide the fence over to the side of the post and lock it in place, with the cutter centered on the octagon's side. With this setup, you can bevel all four upper corners on each post as well as all four lower corners. The jigs themselves compensate for the different tapers of the upper and lower portions of the

of the posts. This depth may seem shallow for bed-rail mortises, but cutting the mortises deeper would only weaken the posts. Besides, the bed bolts hold the tenons tightly in the mortises and help support the downward force on the rails, while the mortises and tenons provide alignment and prevent inward rotation of the rails. In the past, I've done my share of routing, drilling and
chiseling mortises, but I now have the luxury of a horizontal mor-

tising machine. However you cut your mortises, be sure they are

post. After setting the fence and the cutter height, use a square and
a felt pen to mark on the fence the location where the cutter goes into the table. This is the point where the cutter begins its cut. In addition, mark each post at 11 in. and 19 in. up from the bottom.
This designates where to stop the cuts, preserving the square section.

all the same length and the same distance from the bottom of the
posts. Use a square to lay out for the length of the mortises, transferring the lines around the corner of the adjacent faces of the post. Accuracy here will ensure a level mattress support.
After the mortises in the posts are complete, the two side rails

and the head and foot rails are crosscut to length to suit the size bed you're building. Double-check the box-spring dimensions, and don't forget to add 1 in. for the -in. tenon on each end of

the rails. Add an extra

in. to

in. so the box spring is easier

in. the rails are

to install or remove, and take into account the

set back from the inside corner of the posts. The

x1x5-in. ten-

ons are cut on the ends of all four rails, using whatever method is
comfortable for you: backsaw, tablesaw, router, or even bandsaw

or radial-arm saw. Because these joints will be assembled and disassembled, the tenons should slide into the mortises easily but without excess play. To ease assembly, sand or file a in. bevel on the tenon corners and around the perimeter of the mortises. After fitting the tenons to the mortises, you're ready to drill for the bed bolts. To locate the holes in the posts, I made a rectangular plug that fits into the mortise. On this plug, I marked the vertical

center of the mortise and then drilled two

in.-dia. holes

The nuts for the bed bolts are dropped into a 1-in.-dia. hole that intersects the bolt hole. Pie-shape pieces are bandsawn from a 1-in. dowel, chiseled to fit around the nut and glued in place to keep the nut aligned with the bolt hole. The nut is then trapped in place with a glued-in dowel plug.

through the plug: one in. above center for the head and foot rails; the other in. below center for the side rails. These holes accept a 6d finish nail, which is inserted through the appropriate hole and tapped into the mortise to locate the bolt holes. With the drill press, I drill a in.-dia. hole from the mortise through the post to locate the hole on the outside Of the post. Then, I turn the post over, and using a 1-in.-dia. Forstner bit, drill a in.-deep hole to countersink the bolt head. Finally, I change to a -in.-dia. twist bit and drill back through the in.-dia. hole to the inside of the mortise. I repeat the procedure for the other seven bolt holes.

The bolts will extend through the holes in the post and into
holes in the tenon ends of the rails. To align these holes, mount a
rail horizontally in a bench vise, find the mating post and mortise

for that particular rail end and slide the mortise onto the tenon.
Hold the post in place with one hand while you drill a in.-dia. hole through the post-bolt holes into the tenon ends. Unless you have an extra-long bit, you'll need to remove the post and extend the hole 4 in. deep into the rail to accommodate the 6-in.-long bed

bolts. Repeat this procedure for the remaining tenons. The rails
are now ready for nut holes.
To locate where the nut will be trapped in the rail, lay the rail

on the bench, with its inside face up, and slide a bolt partially into
the tenon hole. Align a straightedge with the centerline of the bolt shaft; with a pencil, mark this centerline on the rail. Measure in about 3 in. from the tenon shoulder, along the bolt's centerline,
and use an awl to locate the hole you'll drill for the nut. This

should fall about

in. from the end of the bolt, Traditionally, the

nut was dropped into a

x1x1 -in. slot chiseled into the rail. A

plug was then cut and glued into the hole to prevent the nut from falling out. These slots can be chopped by hand or cut with a in.wide hollow chisel or a plunge router and jig. I find it easier to
drill a 1 in.-deep hole with a 1-in.-dia. Forstner bit. Then, I

drop the nut into the hole and screw the bolt through. To hold the nut in place, I bandsaw a in. by -in. cross out of the end of a 1-in. dowel, saw off the four remaining pie-shape pieces and pare them with a chisel until they fit snugly around the bolt and nut, as shown in the photo this page. I glue them into place around the nut, leaving the bolt threaded through the nut until the glue sets, to ensure proper alignment. Then, I plug the hole with a 1-in.-dia. dowel.
Next, I mortise the inside faces of the two side rails to receive

the bed irons that support the box spring (see figure 1). You can

have a local blacksmith or machinist cut 1 -in.-long pieces from

-in.-thick 4-in. by 4-in. angle iron or bend x 1 x8-in. stock to 90. You can also purchase them from one of the sources listed at the end of this article. The irons should be drilled and countersunk for #12 flat-head screws. The mounting screws should be 1 in.
to 1 in. long. Each rail should have one iron 10 in. to 12 in. from

Toying with tradition

Just because the pencil post is a traditional form doesn't mean it has to look traditional. pencil-post theme.
Here are two variations that toy with the A few years ago, I made a "real" pencilpost bed for an architect. The photo (far
Photo: Kip Brundage

right) is the model I made from actual pencils. To make a full-size bed, the pencil posts had to be 11 times the actual pencil size. I usual pencil wood, is a bit too weak and difficult to obtain in 16/4 stock. The posts were
made four pencils in. by in. tall. I used cherry, because Port Orford cedar, the cut to six sides and sharpened with a block

plane; the grooves and flutes for the eraser

painted the posts to match a pencil's colors. The headboard is a section of ruler, to scale. A friend of mine, David Stenstrom of Portland, Me., built the pencil-post bed in the
photo (right) from maple, then had it sprayed to a high-gloss, candy-apple red. The holder were carved in, as was the lettering. I

stead in a stunning piece you can't take your

expected clash of style and color results ineyes off.

C.B.

Don't write-off the diminutive pencil-post bed (right) as just a small pun: Becksvoort made it as a model for a full-scale version. Woodworker David Stenstrom grew tired of his traditional cherry pencil-post bed, so he built another (left) and lacquered it bright red. The form is so simple and the lines so clean that this contemporary treatment doesn't seem at odds with the design of the piece.

each end and one in the middle. I mortise them into the inside

tester's corners are anchored with dowels in the four post tops.

face of the rails so the angle is flush with the bottom of the rail.
Headboard and testerThe bedstead is not complete until you've
made the headboard. You can choose from a variety of shapes, such as the profiles shown in figure 4 on the facing page, or you

Determine the size of the tester frame pieces by assembling the

bed and measuring from the top of the posts, outside to outside.

Make the lap joints in the side frames shallow, in. to in. deep, to help prevent sagging. Cut four crosspieces, one for each end and two middle pieces; lap their ends to fit the laps in the side pieces.

can design your own. I prefer a simple curved top with a semicircle
cut out of each end. My headboards are 14 in. to 18 in. wide, depend-

I use a doweling jig to drill the

in.-dia. by

in.-deep holes

into the tops of the posts for the dowels that secure the tester

ing on the bed's size. To determine the length and placement of the headboard, assemble the head rail and two posts. iMeasure up
about 15 in. (the combined thickness of box spring and mattress) from the bottom of the rail, and mark this point on both posts. This

frame. To aid in clamping the doweling jig to the tapered posts, I cut off a chunk from the thick ends of two of the strips bandsawn
from the post tops and use them as spacers. I chamfer both ends of in.-dia. by in.-long dowels and insert them in the post

is the bottom of the headboard. To determine the headboard's

length, measure between the posts where the upper tenons will be
and add in. for the two in.-long tenons that extend into

the posts. I glue up

in. stock to the desired width and bandsaw

tops. Drill in.-dia. holes through the corner laps of the tester frame to fit over the dowels in the tops of the posts. As an alternative to the tester, I also turn finials for the posts. Traditional forms include the ball, acorn, urn, tall urn and flame.
I've settled on a modified acorn oattern borrowed from a Shaker chair design. Drill a in.-dia., in.-deep hole in the bottom of

the headboard to shape. Make the four tenons that anchor the board to the posts about 2 in. wide and bevel them in back so the part of

the tenon that extends into the posts is about in. thick. Then, place the bottom of the headboard against the marks on the posts and locate the exact positions of the tenons (mortises-to-be) on the inside facets of the posts. Disassemble the head rail and drill, chisel or rout the mortises. The angle created by the posts' tapers is so minor over the 12-in. distance between tenons that it's easy to
compensate for by slightly angling the top and bottom of the mortises with a chisel. The fit should be loose but not sloppy, as this is

the finials to fit over the dowels in the post tops. The bed featured here is made from cherry with an oil finish, but it looks just as nice in American mahogany, walnut or figured maple. After finishing, the bed irons are attached and decorative bolt covers (usually six) are screwed over the bolt holes.

Me., and is a contributing editor to Fine Woodworking.

Christian H. Becksvoort builds custom furniture in New Gloucester,

not a glued joint. Reassemble the entire head unit to check for fit.
Now the posts, rails and headboard can be finish-sanded. If they were shaped and planed with no major tearouts, I begin with 120

Sources of supply
pressed and cast):
Bed bolts and nuts, wrenches, angle irons and bolt covers (both
Ball & Ball, 463 W Lincoln Hwy., Exton, PA 19341; (215) 363-7330. Horton Brasses, Box 120F, Cromwell, CT 06416; (203) 635-4400. Period Furniture Hardware, Box 314, Charles Street Station, Boston,
MA 02114; (617) 227-0758.

grit and sand through 400 or 600.

The tester is made from in.-thick stock, ripped to in. wide to match the width of the post tops and joined with unglued lap joints. If it was glued together, you'd have a large, flimsy, unwieldy frame to contend with when it was removed for transport or storage. The

A small table with an oriental flair

by Gary Rogowski

Building a Nightstand

nightstand shouldn't be so plain that it puts you to sleep. I designed the cherry nightstand in the photo below as part of a bedroom set, but I think the resulting table stands alone quite nicely. The cloud-rise pattern on its rails along with the rosewood pull on the drawer give the nightstand a look reminiscent of Chinese furniture. The basic design can be adapted for an end table, but keep in mind that if you change proportions in one area, other dimensions will need to change as well. Since it was designed as a nightstand, the first parameter for establishing dimensions was the 20-in. height of the bed's mattress, to which the stand would be a partner. I decided on a height of 23 in., and then proportioned the stand's width and depth to complement this height and allow enough space on top for a lamp, book and cup. The flaring legs add visual weight to the base of the nightstand, while the cloud-rise pattern on the rails softens this effect and lightens the load. The rails are inset from the legs to provide depth and shadow

play, and the overhanging top is beveled to give the entire piece definition and a certain oriental grace. The rosewood drawer pull is the final touch and gives the stand its own personal character. The nightstand is constructed following the same methods I use for larger cabinets or chests. The carcase frames are joined with haunched tenons in routed mortises and the panels that fill these frames are glued up from solid cherry. The bottom shelf and the nightstand's top, which are also solid wood, are mounted as shown in the drawing to allow for seasonal expansion and contraction. The drawer parts are all joined with sliding dovetails and the -in. plywood bottom is slid into grooves in the sides and front and screwed to the bottom of the drawer's back. Although simple in design, the nightstand is a challenge to build because of its joinery and details. Joining the carcaseThe nightstand requires about 4 bd. ft. of cherry for the legs and about 18 bd. ft. of cherry for everything else including the drawer parts. I mill the parts for the entire piece about in. oversize and let them sit for about a week in the shop, so they can warp or cup if they're going to, before final dimensioning. The -in. by -in. legs are then milled and crosscut to length, and all the mortises and panel grooves are routed in them before the outside face is tapered. Similarly, the rails are milled to size and the tenons tablesawn on their ends before the cloud-rise pattern is bandsawn. The top rails, drawer glide rails and side panels can be milled now as well. The panels are only in. thick, so I resaw them from full stock and sticker them so the interior and exterior moisture contents can equalize for a few days before I glue up the book-matched boards. The nightstand's -in. top and its bottom shelf can be glued up somewhat oversize at this time. I use a plunge router with a -in.-dia. bit to cut all the mortises, and then chisel the ends of the mortises square to accept the square tenons. You could round the tenons to fit the radius of the router bit instead, but on a joint this small I figure the more gluing surface the better. The mortises should be about -in. deeper than the tenons to ensure the tenons don't bottom out and to provide space for excess glue. When the mortises are complete, cut the tenons on the ends of the rails on the tablesaw. First, cut the shoulders with the rail flat on the table and the miter gauge set to 90. Then, make a simple tenoning jig to hold the rail vertically as you run it along the rip fence and trim the tenons to fit the mortises. Finally, bandsaw the tenon to create the haunch. I'm partial to haunched tenons because the haunch allows the tenon to run full width to resist twisting, while the main portion of the tenon can extend deeply into the leg or stile without weakening this member. The top rails above the drawer in the front and back of the

Although designed as a nightstand, this little cherry table would

serve just as well as an end table alongside a couch or chair in

any room in the house.

trace around the inside edge of the frame to define the panel's shape. Add in. around the entire perimeter of this line so the panel will fit into the groove, and then bandsaw the panels to size. I check the fit of the panels in the grooves and handplane the panels to fit snugly. Then, I finish-sand them and apply the first of three coats of a mixture of equal parts Danish oil and polyurethane varnish. I wait a day in between coats and finish both sides of the panels the same. Prefinishing the panels ensures that no
unfinished wood peeks free of the frame if the panel shrinks during an especially dry season. It's also much easier to wipe on and rub out the finish before assembly. After the third coat of finish on the panels is dry, you're ready to assemble the side frames. Use just a little glue on the tenons to prevent excess squeeze-out onto the legs or panels, and brush a more liberal coat of glue into the mortises, especially at the mouth

inside surfaces of the rail and leg, providing a consistent surface for the router base.
nightstand are dovetailed into the top of the legs. I cut these dovetail-shape mortises and tenons with a dovetail bit on my router table. The mortises are cut first, and then the fence is moved over
so it exposes only a small portion of the bit to cut the tenon on the

board is laid atop the shaped rail to fill the setback between the

With the side frame clamped to the workbench, Rogowski routs the panel groove with a bearing-guided slotting bit. A -in.-thick

of the joint. Don't forget to glue in the drawer glide rails. After the frame is glued up and clamped, check to see that it's flat and hasn't been twisted out of shape by clamping pressure. When the frame is dry, pin the panels in place with -in.-dia. dowels drilled from the inside of the top and bottom rails and centered in the middle keep the panels centered and prevent them from rattling too much if they shrink in dry weather. The front and back rails can now be glued into the side frames to form the carcase. In order to simplify the glue-up as much as possible, set aside the dovetailed top rails and back panel. It's not
necessary to glue these parts in place at this time because the panel can be slid in place and the rails glued into their mortises after the rest of the carcase is glued up. I get the clamps ready, one for each of the rails that I'm gluing in place, plus a couple extra to square up the clamped-up carcase if necessary. Because it's difficult to clamp to the tapered legs, bandsaw angled clamping blocks and tape them to the tapered surfaces so the clamps won't slip. Lay one of the side panels flat on the workbench with the mortises up, and glue the front and back rails that are below the drawer,

of the panel (see the drawing on the previous page). These dowels

ends of the top rails (see the top and center photos on the facing
page showing drawer construction). When all the mortises and tenons are cut, dry-assemble the frame to check for fit and to see that the tenon shoulders pull up tightly. Disassemble the parts and fine-tune the tenons if necessary. At this time you can also shape the rails. Because the same cloudrise pattern is repeated at the end of each rail throughout the piece, I make a full-size template from -in.-thick Masonite and use it to draw the shape on the rails. I bandsaw just outside this line, attach the template to each rail with double-stick tape and

trim to the template with a flush-trim bit on my router table. I stop just short of the end of the curves to prevent tearout and finish up with very light cuts with the router. Clean up the final shape and remove the router marks with a spokeshave or file and then sand the curves smooth. Next, dry-assemble the legs and the shaped side rails to form the side frames. However, leave out the drawer glide rails because
they'll be in the way as you rout the panel groove on the internal

the back bottom rail and the shelf support rails into their mortises. Then, with the other side panel laying on the bench with its mortises up, spread glue and fit all the tenons of the five rails into their mortises, and force the joints home with bar or pipe clamps. Then, place the nightstand upright on a reliably flat surface and, if all four legs don't make contact, damp diagonally across the front and back of the carcase frame to square them up. Test-fit the top rails into the dovetail mortises to see that the carcase hasn't been distorted out of
square in the horizontal plane by uneven clamping pressure. Now, you can insert the back panel and glue in the top rails.

edges of the frames. Clamp the frame to the workbench and use a
-in. slotting bit to cut a -in.-deep groove centered in the edge of the rails. The bearing of the slotting bit follows the curve of the rails but you'll need to place a -in.-thick board on the rails, as shown in the photo above, to bring the rail's surface flush with the legs. The panel is in. thick, so I reset the bit's depth and make two passes to cut a -in.-wide groove. Repeat this procedure to rout the groove on the back legs and rails for the small back

With the base of the stand assembled, the top of the stand and
the bottom shelf can be ripped and crosscut to size. I rout slots in the upper side rails with a -in. slotting bit for the tabletop fasteners that hold the solid top in place. The bit's depth must be set properly so the fasteners will hold without too much strain or

panel that covers the drawer opening. After the panel grooves are cut, disassemble the frames and draw the taper on the outside surface of each leg. Then, bandsaw close to the line and clean up the face on the jointer.

coats of my oil/varnish mix to the glued-up carcase frame and to the top and the bottom shelf. When the finish dries, I fasten the top and the bottom shelf in place.

slop, so I first cut a practice piece to check this setting. I bevel the edges of the top, as shown in the drawing, and then I apply three

Constructing the drawerOne of the nice things about using

sliding dovetails to join the drawer sides to the front, is that I can use the same setups to join the back of the drawer to the sides.
The secret to a good-fitting sliding dovetail is accurately milling

Making the panels and gluing up the carcase Now, turn your
attention to the panels. Glue up the side panels and cut them and the back panel at least 1 in. longer and wider than the inside dimensions of their respective frames. After the glue is dry, plane the panels in. thick. Then, lay the appropriate frame on its panel and

the parts. The drawer sides and back must all be the same thickness and their ends and edges must be square with the planed
faces. I use a -in.-thick front with -in.-thick sides and back. Rip

the drawer parts in. over their finished width so any chip out resulting from routing the dovetails can be jointed smooth. Crosscut the back a little long so you can cut it to fit after you've routed its
mortises in the sides. Before routing, carefully hand-sand all surfaces with a sanding block because surface-sanding the mortised parts after routing them will spoil the fit of the dovetails' shoulders. The top and center photos at right show the two setups for routing the mating parts. Before routing the dovetail mortises, I like to rout or dado -in.-wide slots centered where the mortises will be so the dovetail bit doesn't have to remove so much wood in a single pass.

Then I rout the mortises with the fence set for a

-in. setback. I hold

a scrap block in front of and behind the drawer part to reduce the

possibility of the bit tearing out the edge as it enters or exits the cut. After cutting the mortises, unclamp one end of the fence and move it into the spinning bit so it cuts into the fence. When only a small portion of the side of the bit is exposed, as shown in the center photo at right, reclamp the fence. I've found that I get the best fit by lowering the bit just a hair when I cut the tenons. However, to avoid a chance of lowering the bit too much and losing all
reference to the original height, you can use a thin paper shim under the ends of the workpieces as you cut the male portion of the dovetails. To set the fence for a snug-fitting dovetail, make a test pass on both sides of an off-cut from the drawer sides that has been sanded to the same thickness as the drawer parts. The test dovetail should slide into the mortise without excessive pressure, but without any slop. When you've got the fit just right, cut the tails on the front ends of the sides, and then slide them into the mortises in the drawer front and measure between the sides to get an exact measurement for the length of the back. Trim the back to length and rout the tails on both ends. Now, joint the edges of the

the scrap to the back edge. A -in.-deep groove, visible on the free end of the workpiece, is cut to remove most of the waste

drawer front. He holds a scrap block against the workpiece to prevent chip out and as he proceeds through the cut, he moves

In the photo above, the author cuts the dovetail mortise on the

from the mortises prior to a single pass with the dovetail bit. In the photo below, the dovetail tenon on a drawer side is being
cut. The fence has been moved to cover most of the bit, leaving

just enough of it exposed to result in the right size tail after making a pass on each side of the workpiece.

parts to remove sawmarks or chip out, and rout the grooves for the drawer glides in the sides. Finally, rout the bottom grooves in the sides and front, and rip the back piece narrow enough that the bottom will slide beneath it. I shellac my drawer parts on all sides, except the drawer face, before gluing them together, taking care not to get any finish into the joints. Shellac is a quick finish to apply and by adding a few drops of
your favorite scent, the drawer will smell much sweeter than an oilfinished drawer. The outer face of the drawer front is finished with the same oil and varnish mixture used on the rest of the nightstand.

When gluing up the drawer, spread glue only in the mortises to

avoid excessive squeeze-out. Begin by gluing the sides into the front one at a time with the drawer front standing right-side up near the edge of the bench, as shown in the bottom photo at right. Start the dovetail into the mortise and then press it home with a bar clamp spanning from beneath the bench to the top of the

drawer side. Keep the side moving once you start or it will freeze up and you'll need a great deal of force to break it free. Have a piece of -in. plywood handy to check the alignment of the bottom grooves. When both sides are in place, turn the drawer around and clamp the back into the side mortises. Then, slide the bottom into the side grooves and screw it to the bottom of the back. The two drawer glides are screwed to the drawer glide rail, as shown in the drawing on p. 89. The glides must be about in. wide to span the -in. setback between the rails and the drawer
sides, plus extend into the -in.-deep grooves in the drawer sides. To install the glides, predrill and countersink screw holes in them and then, with the glides in the side grooves, slide the drawer into the carcase. Pull the drawer out little by little while holding the

tises in the drawer front with a clamp, as shown above. Then, the drawer is reversed and the back is pressed into the side mortises.
see the completed piece from all directions and play with a number of options. I make models in cardboard or alder and paint them black, to provide contrast, and then try them on the drawer
face until I find a design that is complementary to the piece and easy to handle. The drawer pull for the nightstand is made of rosewood and screwed to the drawer face from inside the drawer with brass oval-head screws and finish washers.

The sliding dovetails on the drawer sides are pressed into the mor-

back end of the glide against the drawer glide rail from below, and screw the glides to the rails as you go. Then, wax the glides and
grooves for a smooth, sliding action. I always save designing my drawer pulls for last because I like to

Gary Rogowski designs and builds furniture in Portland, Ore.

Anderson Ranch Arts Center, Box 5598, Snowmass Village, Colo. 81615; (303) 923-3181.

This summer he will be teaching a course on stool building at

ith its modest proportions and uncluttered design, the chest-on-chest in the top photo at right can easily find a place in most of today's homes. Standing only 1 in. over 5 ft. tall, the chest has a friendly, unthreatening scale unlike many of its 7-ft. tall period counterparts that nearly graze contemporary ceilings and dominate an entire room. Judging by the clean lines, the good proportions and the shape of the bracket feet, this double chest was probably based on designs made by Thomas Elfe, whose work delighted the people of Charleston, S.C., during the mid-18th century. This chest was built to the plan on the opposite page 40 years ago by furnituremakers at Virginia Craftsmen, of Harrisonburg, Va., who have built reproduction furniture for many historical restorations. To avoid the problem of split sides, a common ailment of antique furniture, the designers at Virginia Craftsmen came up with a simple way to build carcase frames that allows the sides to expand and contract freely. To accomplish this, the rails that run side to side between the drawers are tenoned and glued into mortises in the sides, while the runners and center guides that run front to back within the case are tenoned, but not glued into the rails. These "dry" joints allow the sides to expand with increased humidity. In addition, the runners and guides are cut in. short to leave a -in. gap at each end, as shown in the bottom photo at right. These gaps provide space for the sides to shrink in drier conditions. Each runner is screwed to the center of the case side with a single 1 -in.-long, #8 flat-head screw to keep the runners from ever sagging and to resist any tendency for the sides to cup or bow. Building the base chestBegin construction by gluing up the base unit sides and cutting them to size as shown in the plan. Then, run a -in. by -in. rabbet on the inside rear edges for the backing. Next, make the front and back rails, drawer runners and center guides that make up the carcase framework. As you can see by the detail in the lower left on the plan, the top and bottom rails, both front and back, are dovetailed into the sides, while the other six rails that run between the drawers are mortised into the sides with double tenons. Bandsaw the dovetails on the ends of the top and bottom rails and mark the sockets from the rails onto the top and bottom edges of the sides. Saw and chisel out the sockets by hand. Cut the double tenons on the ends of the other rails with the

Pleasing proportions combine with a simple, unadorned design to give this chest-on-chest a friendly, informal feeling. In the 40 years since it was made, the Appalachian white pine has aged to a dark honey color. A peek inside one of the chests shows why the solid-wood sides have aged without splitting. The horizontal rails are mortised and glued into the sides but the joints between the drawer runners and rails are not glued. The runners are cut short, to give the sides room to shrink, and fastened at the center with a single screw.

dado blade on the tablesaw, and mark the case sides for the mortises
to receive these tenons. You can chisel the double mortises entirely by hand or speed up this process by drilling out most of the waste with a Forstner or brad-point bit on the drill press. You might notice that the drawer rails on the chest in the top photo on p. 76 are all

chests get a good deal of their rigidity from the back that is nailed

to the rabbets in the rear edges of the sides and to the top and
bottom rails. If you choose to use solid, vertical ship-lap for the back, as specified in the plan, this is another place you must allow for wood movement. The first board should be nailed only to the

dovetailed into the sides. This is a possible variation, but when drawing the plan, I chose to go with the double tenons because they are easier to make and quite adequate for this application. Next, cut the hardwood runners and guides to length. The runners and guides have -in.-long tenons on both ends that fit into
mortises on the inside edges of the rails. The detail in the lower left corner of the plan shows how the center guides lap over the top of the rails. Determine the length of these parts by measuring the actual width of the rails and sides, but don't forget that you want a -in. gap at each end to allow the case sides to contract. I cut the

rabbet in the edge of the side. The second board should overlap

the first, but should not be pushed up tight to it. Use

-in. spacers

to hold the boards apart and nail the second board to the top and bottom rails near the overlap so it holds the first board tightly to the back rails. Then, the third board is lapped over the second
board and nailed near this joint to hold it in place. By only nailing one edge of each board and leaving a -in. gap between boards, they are free to expand or contract. Continue in this way across

the back and then end with a narrow board, no more than about
3-in.-wide, and nail it to the rails where it laps the previous board and to the rabbet in the side.

cheeks on the tablesaw with a rip or combination blade first, and then change to the dado blade to remove the waste and establish the shoulders. It's a good idea to tenon the runners so their top surfaces will be about in. higher than the rail to ensure that the drawer sides slide on the runners and not on the front rails.
The -in.-wide by -in.-deep mortises on the inside ends of the rails that receive the runners can be made by cutting a stopped dado on the tablesaw. The mortises for the guides in the center of the rails can either be made with a router, drilled and chiseled square, or chiseled entirely by hand. After dry-assembling to inspect the joints for a good fit all around, take the framework apart

Making and installing the moldings and feetNow you'll need

to make the moldings that give a little style to the piece and provide transitions from one element to the next. There are three
shoe moldings: A quarter-round at the bottom of the base chest, and two ogees, one at the top of the base chest and one at the top of the upper chest. All the moldings are made from 2 -in. by -in. stock and you'll need about a 7-ft. length of each molding shape. First, shape the profile on the edges with a molding plane

and reassemble it with the rail tenons glued into the mortises in the chest sides. Clamp the carcase together across each rail and check
the case for squareness before placing it to dry on a flat surface. After removing the clamps, place the chest facedown on a padded bench and measure diagonally across the carcase to doublecheck that it's square before installing the backing boards. The

or shaper and then saw out the -in.-deep recess behind the profile. Miter the two base chest moldings to fit, and then glue and screw them to the rails and runners. The ogee molding for the top of the base is set in about in. from the chest's edges; since this molding retains the upper chest, it is the gauge for determining
the top unit's dimensions. Cross-grain splines are in order to reinforce these wide miters. You'll need to screw a -in.-thick filler

strip along the back rails, flush with the molding ends, to complete
the perimeter for each set of moldings. These strips fill the gap at the back of the chests and cover the ends of the backing boards. To make the shaped bracket feet, you'll need two x4 x28 boards. The boards are shaped to the ogee profile and then crosscut into six 8 -in.-long sections. Two of these sections are mitered together to form each front foot. The remaining two shaped sections are each dovetailed at a right angle to a x4 x8 flat piece to

form the rear feet. To make the concave shape, run the boards diagonally over the tablesaw blade, taking very light, repetitive cuts using the setup shown in figure 1 at right. Begin cutting with the blade exposed only about in. above the table and raise it about in. for each consecutive pass. Draw the desired profile on the end of one of the boards and as you approach the line, adjust the angle of
the fence to get the cove cut that you're after. When the cove is complete on both pieces, use a jack plane to make the top curve. Sand the sawmarks from the cove and smooth the top curve before cutting the boards into the six 8 -in. lengths that make up the feet. Now, take four of the pieces, and miter one end of each to make the front feet; cut the dovetails on the remaining two shaped pieces and the flat back pieces that go together to make the rear feet. After cutting the joints, bandsaw out the shapes and glue the pairs together, reinforcing the joints with stacks of corner blocks

to avoid the cross-grain problems you'd have with one vertical grain glue block. The feet are screwed to the bottom runners and rails of the chest through two x x5 blocks glued to the top inside edge of each foot. After the feet are in place, nail a -in.
plywood dust bottom to the bottom rails and runners. The base chest is now complete except for the drawers, which should be built all at once after both the top and bottom carcases are complete.

Building the upper chestThe dimensions for the upper chest

are derived from the frame formed by the base's top molding. Since it nestles into the lip of this molding, the top unit doesn't require any bottom molding of its own. The construction of the

upper chest is identical to the base, except for the absence of the bottom molding, the addition of the top board and some framing for the three small upper drawers. This framing consists of two vertical stiles between the small drawers that are mortised and glued into the two top front rails, and four wide hardwood runners mortised into the rails behind the vertical stiles. The lower runners
carry the drawers and have -in. by -in. strips glued to their surface to act as side guides. The two upper runners are called kickers and keep the drawers from tipping when pulled out. After the upper chest is assembled, the remaining ogee shoe mold is mitered and screwed to its top runners and rails, and the -in. spacer is screwed along the back rail as with the other moldings. Again, just as on the base, nail a -in. plywood dust bottom to the upper chest. Cut this dust bottom undersize and center it on

screwed to its bottom. To ensure accurate placement of these strips, insert one of the drawers into its opening until the front is against the face frame, and center the drawer side to side. Now, hold the guide strips alongside the center guide from below the drawer and drive three small brass screws down through the

drawer bottom to hold each strip in place. Remove the drawer from the chest, unscrew the guide strips and put a spot of glue on the guide strip near each screw hole before replacing the screws. When the guide strips are all in place on the upper chest drawers, you can screw the top board in place. Cut it to size to allow a 1 -in. overhang on both sides and the front, and run the same
ogee profile you used for the molding along the three overhanging edges. Fasten the top to the upper chest with eight 2-in.-long, #8 screws up through the top rails and runnersone near each corner and one 2 in. each way from the center of the runners. Countersink and slot the screw holes in the runners to let the top expand front to back. Brass pulls, escutcheons and drawer locks are available from Horton Brasses, Nooks Hill Road, Box 120F, Cromwell, Conn. 06416, or Ball and Ball, 463 W. Lincoln Highway, Exton, Pa.

the chest's bottom so it will fit within the frame formed by the ogee molding around the top of the base.

Drawers and hardwareBuild the drawers according to the detail on the plan. The sides are joined to the front with half-blind
dovetails with a -in. setback on the top and sides. The back is joined to the sides with through dovetails. During assembly, the -in. bottom is inserted into grooves that are in. up from the bottom edge of the sides, front and back. If you use solid wood instead of plywood for the bottom, be sure to allow room in the grooves for cross-grain expansion. After the drawers are glued up, saw and chisel a notch out of the

19341. However, if you are so inclined, the backing plates can be easily bandsawn from sheet brass (available from a hobby supply store) using a fine-tooth metal-cutting blade running at your normal woodworking speed. Unless you also want to venture into casting brass, you can buy the posts and bails from Horton Brasses
or Ball and Ball.

bottom of each back to fit over the center guide. Then, as a track for the guides, each drawer gets two -in.-thick hardwood strips

Carlyle Lynch was a Broadway, Va., designer, cabinetmaker and retired teacher. He passed away shortly after writing this article (see his obituary on p. 114). His drawings have appeared regularly in FWW and are available from Garrett Wade, Lee Valley

Tools and Woodcraft Supply.

Flush panels modify a Shaker design

by Christian Becksvoort

Building a Tinware Cupboard

his Shaker tinware cupboard is a fascinating piece because it breaks all rules of proportion. The golden rectangle, a proportional principle used since ancient Greek civilizations, maintains that a height-to-width ratio of 1.6-to-l is most harmonious, and other proportions create conflict. Yet this cupboard's size is what makes it one of the most versatile pieces I build. It is equally suitable in a hallway, bedroom, bath, kitchen or living roomanywhere space is at a premium-to store crystal, linens, clothes, camera equipment, papers, books, canned goods or even stereo components. My adaptation of the tinware cupboard is based on a Shaker original I saw at the Renwick Gallery of the Smithsonian Institution, Washington, D.C., in 1973. Built in New Lebanon, N.Y., the dark red painted pine piece was probably used in a kitchen or pantry. The first few cupboards I built were faithful reproductions, scaled from a photo. Since then I have made a few minor changes, such as lighter stock and movable shelves. I usually make the carcase from in.-thick edge glued cherry and use in.-thick stock for the face frame and doors. The top and sides are rabbeted to receive the back and are dovetailed together, while the fixed middle shelf and the bottom are dadoed into the sides. The bottom is also screwed to the sides through glue blocks. Visually and structurally, the most important changes were to replace both the slab back and raised-panel doors with frames and flat-flush panels, as shown here, which give the piece a distinctive look while accommodating the natural expansion and contraction of the solid-wood doors and back panels. The face frame and flat-flush back panel are mortised and tenoned, and then glued to the carcase, resulting in an extremely rigid piece. Because five of the shelves are adjustable, rather than fixed as in the original, the cabinet is adaptable to a variety of storage requirements. And, instead of exterior spinners and a lock, I use interior spinners operated by the doorknobs. While I usually prefer an oil finish on natural cherry pieces, I was intrigued by John McAlevey's article on ebonizing wood (FWW #76, pp. 47-49), a process that chemically dyes the stock, yet lets the grain show through. The Shakers often ebonized their chair frames, and so I thought why not an entire case? I experimented with one of my cupboards, making the carcase out of birch and leaving the interior natural to contrast with the dark ebonized exterior. When the doors are opened, the light, bright interior is a bit of a surprise, as you can see here. Carcase constructionTo build this cupboard, you will need about 60 bd. ft. of cherry or birch, most of which gets glued up into four slightly oversize panels. Two panels, will make the sides, and two more, will be cut into the top and bottom, the fixed middle shelf and five adjustable shelves. After planing these panels to in. thick, cut the individual pieces

The author's Shaker-inspired tinware cupboard is a functional piece that fits well in almost any room. The cupboard on the right is cherry with an oil finish. The other is birch, which Becksvoort ebonized on the outside and left natural on the inside. Adding to the beauty of each is a frame-and-panel back, shown on the cherry cupboard.

to the correct lengths and widths given in figure 1 below. On the

sides to fill in the gap left by the back-panel rabbet cut in the top,

tablesaw, rip a in.-wide by in.-deep rabbet along the inside back edges of the two sides and top to receive the back panel. To
dovetail the sides to the top, begin by laying out and cutting the

as shown in the detail in figure 1. Before assembling the top to the

sides, rout the dadoes for the bottom and the fixed middle shelf, guided by the dual-fence router jig shown in the top, left photo on

tails on the top. The tails at the front and back edges are half-tails,
and the tail at the back edge is positioned so the pin on the side

the next page. These dadoes can also be cut on the tablesaw guided by the miter gauge, but I find this operation unwieldy with long side panels. After sanding the interior surfaces, beginning with 120-grit paper and working up to 220-grit, I assemble the top to the sides. When clamping the dovetails, a spacer about a foot be-

does not interfere with the back-panel rabbet. Six more tails are
equally spaced between the front and back tails. Now, secure the
top to the sides with a 90 angle clamp, and lay out the pins by

scribing the sides with a knife. Because I'm used to cutting pins for dovetails in a vertical position, I clamp the sides to the stringer of

low the top keeps the sides from pulling together as the clamps
are tightened (see the photo at right on the next page).

my open stairway and stand on the stairs to cut the pins. You may
find it easier to lay the sides on sawhorses to cut the pins. When cutting away the waste for the pins, leave a stub at the back of the

While the glue dries, sand the middle shelf and bottom as before, working up to 220-grit sandpaper. Slide the middle shelf into the dadoes and toenail it from underneath with three 6d finish

Left: This dual-fence jig guides the router from both sides, minimizing any possibility of miscutting the dado, and batten clamped beside the jig aligns the panels.
it is long enough to rout both side panels at once. The

Right: To bold the sides in position for installing the top,

the author clamps them to the workbench and a sawhorse. A spacer wedged between the sides prevents the clamps from bowing the sides during glue-up.

Below: This shopmade drill guide automatically spaces

holes for movable shelf brackets and locates them in. from the carcase edge. A in. dowel on the drill bit controls the depth of cut.

nails per side. You could glue the shelf in place, but it isn't really necessary because the rear panel and face frame will prevent the
sides of the carcase from spreading. The bottom is assembled the

Making the face frame When made as a single mortised-and-tenoned unit, the face frame strengthens the carcase greatly and prevents the sides from spreading or compressing. I use the assembled carcase as a reference for measuring the stiles and rails and for laying out the mortises and tenons. This reduces measurement errors and ensures the face frame will fit properly. The carcase is easier to work on if laid flat across two sawhorses. Measure the diagonals to be sure the carcase is square, and sight from top to bottom, across a pair of winding sticks, to check flatness. Clamp from opposing corners to correct squareness, and shim as necessary to remove any twist. I cut the frame stiles and rails in. larger than the dimensions in figure 1 so I can plane the installed face frame to fit the carcase exactly. Place the stiles in position on the carcase, and lay the frame rails in position over the stiles. The center and bottom rails should be laid out so they cover only the bottom half of the middle shelf and the bottom. The exposed upper half of the shelf thickness acts as a doorstop. Markthie position of the rails on the stiles and then lay out the mortises, leaving at least in. to the ends of the stiles. I have a
horizontal boring machine for cutting the in.-wide by 2-in.-deep

same way, but for extra support, I prefer to add three glue blocks to each side and screw up into the shelf, as well as into the
sides (see figure 1). Using three separate blocks, rather than a sin-

gle cleat, allows the solid-wood sides and bottom to move as moisture conditions change. While you are working on this end of the carcase, lightly chamfer the bottom edges of the sides. This makes it easier to move the cupboard across the floor. Now that the case is fairly rigid, but before the back goes on, I drill the in.-dia. by in.-deep shelf-support holes using the shopmade drill guide shown in the bottom photo. I made several

of these in. by 4-in. hard-maple guides in different lengths for use in a variety of cabinet interiors. To make a guide, clamp a
fence to the drill-press table and bore a row of holes in. from

the edge of the drill guide at 1-in. intervals along the entire length. To use the guide, label one end "down" on both sides. Butt this
end against either the bottom or the middle shelf, and clamp the

guide so that the holes are

A section of

in. from the carcase edge. To get

mortises in the stiles, but you can also use a router or drill press, or
chop them out with a chisel and mallet. I cut my mortises first so that
I can test the adjustment of my tablesaw when crosscutting the ten-

the cleanest holes, I use a brad-point bit chucked in my hand drill.

in. dowel, drilled lengthwise and left on the bit,

serves as a positive stop to prevent drilling through the cabinet sides. Determine the length of the stop by subtracting in. ( in. for the
drill guide and in. for the depth of the shelf-support hole) from

ons. I set up the saw so the fence acts as a stop for the tenon, and I hold the stock against the miter gauge as I make multiple cuts on
each face of the rail to form the tenon.

the exposed length of the drill bit. Now, carefully insert the bit into the guide until it hits the cabinet side and drill until the dowel
stop bottoms out. Repeat this process for all eight rows of holes, sanding the "fuzz" around the holes with 220-grit paper.

After cutting the tenons, assemble the face frame and sand it as
before, working up to 220-grit. Dry-fit the frame onto the case so you can check for side-to-side fit, length and squareness. If the case sides are perfectly straight, you can joint the face frame's in. overhang

Hanging flush cabinet doors

Everyone develops their own methods and routines for hanging doors. But I like to keep things as simple as possible, with a minimum of measuring. The tinware cupboard described in the main article is a good project for practicing door hanging because the carcase can be laid flat on its back and gravity will hold the doors as you adjust and position them for a perfect fit. Although it's easiest to install doors with the case on its back, these techniques will also work on vertical installations. and the wood is dry because next summer's humidity will cause the stiles to swell. Leave the gaps a bit bigger all around if the cabinet is to be painted. On the other hand, if it is mid-summer and the moisture content of the wood is 11% to 12%, the gap can be quite small. Remember that the hinge-side gap will remain constant; so all the wood movement of the door frame will be apparent on the lock side. Also, I usually bevel the edge of the lock stile a few degrees for greater swing clearance. Hinge layout: Butt hinges are usually set at the intersections of the door rails and stiles: the top hinge even with the top of the panel and the bottom hinge even with the bottom of the panel. Setting the hinge as close as possible to the rail is not only structurally sound, but also aesthetically pleasing, forming a sight line between rails and hinges. Also, I set hinges with the barrel fully revealed, which provides clearance when the door is opened 180. To set the hinges, first place the door in a vise, hinge-side up. Using a square, locate one end of the hinge by transferring the rail-stile intersection to the edge of the stile. Turn the hinge upside down and press the barrel against the outside face of the door, as shown in the top photo, aligning the top or bottom with the mark. With a knife, scribe the stile carefully for all three sides of the hinge. Then set your marking gauge to the thickness of the leaf and scribe the outside face of the door between the knife marks. Rout or chisel the hinge mortise as marked. For hinges with tapered leaves, chisel upward to the back of the mortise, matching the bevel of the hinge leaf. Repeat for the other hinge. Mount the hinges with screws in just the center holes so you have the option of persuading the hinge into a better position later, should that be necessary. Now, carefully place the door in its opening with the desired gap left at the top and bottom. With a knife or razor, mark the tops and bottoms of the hinges where they touch the cabinet frame, as shown in the bottom photo, being careful not to shift the door. Remove the door and one of the hinges. Place the hinge barrel between the marks on the frame so that the barrel sits on the frame and the leaf hangs down along the inside frame edge. Then, scribe the hinge location with a knife and marking gauge and chisel out the waste as before. Test the fit by remounting the hinges with one screw per leaf, and make any necessary adjustments before installing the other screws. C.B.

slightly oversize and then saw or joint the lock stile to fit, taking the dimensions directly from the opening. Leave the door relatively tight in the opening; you can always shave off more later. To mark the door bottom for trimming, first place the door in its opening as tight to the bottom as possible, and then draw a line on the door, parallel with and as close to the frame rail as possible. Saw or joint to this line. The door should now fit on three sides. With the door placed back in its opening, mark the top for cutting by drawing a line a full in. below and parallel to the top frame rail. Saw or joint to this line and retest the fit. A flush door sitting in its opening is extremely difficult to remove. You can put a screw where the nob will go, predrill the knob hole to provide a fingertip grip or make a sheet-metal "door remover," as shown in figure 2 above, but don't try to pry the door out with a screwdriver or chisel. Selecting the appropriate hinge depends on the size of the piece and the weight and width of the door. Generally, the larger the piece, the bigger and heavier the door will be and the larger the hinge needs to be. Also, the wider the door, the greater the leverage it exerts on the hinges. I've found that in. or 2-in. hinges work on 90% of the case goods I build, but for larger pieces, like an armoire, I'll use in. or 3-in. hinges. For smaller projects, such as jewelry boxes, aesthetics become more important than strength, and so much smaller hinges can be used. The exact gap needed around the door is determined by the hinge type, the season of the year and the moisture content of the wood. As a very rough rule, I like to leave about in. plus on the two sides and the top, and a bit more on the bottom. Age, wear and gravity conspire to move doors down, never up. While the gap on the hinge side is predetermined by the hinge, it is not unchangeable. The gap on the lock side should be bigger if it is winter

Fitting the door: I make my doors

For best fit, hinge mortises must be accurately laid out. Here the author uses a knife
to trace the hinge outline to the door stile.

face frame. For accuracy, you must not

knife to transfer the hinge location to the

move the door when making these marks.

With the door positioned, Becksvoort uses a

July/August 1990

41

To install the back panel, turn the carcase facedown on the horses and fit the back snugly into its rabbets. The bottom rail on the back will be in. from the bottom of the carcase, but the stiles continue
to the floor to reinforce the fragile -in.-thick carcase sides where

rabbeted. Beveling the inside corners of the back panel makes it easier to slide into its rabbets, as well as provides space for excess glue.
Get the clamps and clamp sticks ready, apply the glue and clamp

everything togemer. The back should be clamped front to back, as

well as side to side, to fully seat it in the rabbets. If you don't have
enough clamps, 6d finishing nails help hold things together while the

glue dries. Countersink the nails deeply, and plug the holes by gluing in wooden pegs and chiseling them flush to the back panel.

frame rail, where clamps won't reach. Wedges under the feet of the

During glue-up, clamping strips are used to protect the face frame and the rabbet on the back edge. The large number of clamps ensures a good bond. The wedges and batten help secure the centersawhorse remove twist from the carcase.

The top molding shown in figure 1 is one of my favorites. While

the glue is drying, I shape the molding stock with a cutter in a

molding head on my tablesaw. Since you only need 46 in. of molding, however, it can be roughed and shaped easily with a block plane in 10 to 15 minutes. Start with a piece of in.-thick stock that is 4 in. to 6 in. wide by 50 in. long. Draw the profile on each end, mark off in. (the total thickness of the molding), clamp the

or you can wait until the face frame is attached to the carcase and

then plane the overhang for an exact fit. Before gluing the frame in place, get all your clamps ready; I use 20 or more, as well as protective strips for the face frame and the fragile rabbet along the back
edge. Then, spread the glue, position the face frame and clamp it to

stock in a vise and plane the stock to shape. When it looks just right, rip the in. molding on the tablesaw and finish-sand it.
Before applying the molding, I sand the carcase, starting with 150-grit paper in a belt sander and working up to 180-, 220- and 320-grit paper in a palm sander; finish up by hand-sanding with 320- and 600-grit. Now, miter the front molding and glue it in place. To keep the molding from slipping while clamped, I drive

the carcase, as shown above. Two

in.-thick boards, glued and

clamped to the back side of the center-frame rail and the top-frame

rail, as shown in figure 1, serve as the top doorstops. These stops extend the full width of the cupboard's interior, butt against the top
and the middle shelf, and extend in. below the frame rails, balancing the in. exposed edge of the middle shelf and bottom.

two small brads into the carcase at the molding location, snip off their heads and force the molding onto the pins. Glue the miters
and the first 2 in. of the side molding, and then nail the rest in

Building frames and panelsI like double-sided, flat-flush frames

and panels for the doors and back panel instead of the raised-panel

place with countersunk and filled 6d finish nails, which allow for seasonal movement of the carcase side. Now, sand the top, following
the same sequence as before, and touch up the molding as needed.

doors and the single-board back on the original cupboard. The

in.-thick doors and the in.-thick back are slightly oversize so they can be trimmed to fit. Mortise the stiles in. wide by 2 in. deep, as previously described, and then tenon the rails to fit Dado in.-wide by in.-deep stopped grooves on the stiles, and dado through

Be careful not to leave cross-grain scratches on the case sides.

The detailsI fit and hang the doors as described in the sidebar on the previous page, and then I install the door bumpers, knobs and spinners as shown in figure 1. Shopmade in.-dia. leather pads, glued to the lock side of the upper and lower doorstops,

grooves on the rails for the panels. By stopping the stile's groove, you eliminate the need to cut a haunched tenon on the rail without leaving a visible groove at the ends of the stiles. The in.-thick by in.wide panel tongues are cut with only two tablesaw setups. Although the shoulder cut can be vertical, I like a slight bevel, as shown in

make for quiet closings and convey quality construction. I turn the knobs from a block on the lathe after shaping the
knob tenon with a in. plug cutter on the drill press. Drill the

figure 1 on p. 39. The first setup cuts the face of the tongue. Set the
blade in. above the table and add a high auxiliary fence to support

knob holes with a in.-dia. bit, which will allow the knobs to turn easily in the doors, as shown in figure 1. The spinner, a scrap with a in.-dia. hole, is cut to a

the panel on edge. Adjust the fence to leave a in.-wide center section, which will form the tongue. Cut the panel by running it over the
blade vertically on all four edges, referencing for the first cut from

pointed oval shape so that it protrudes about

in. beyond the door

frame. The spinner's placement on the tenon is critical: too loose and it feels sloppy, but too tight and the knob won't turn. I carefully ham-

the face of the panel and for the second cut from the back of the panel. Then, tilt the blade 20 to 25 and set the fence in. from the
side of the blade farthest from the fence. With the panel fiat on the

mer the spinner into place until it feels just right, and then drill a pilot hole and drive a brad through the spinner and tenon to lock
them together. Cut off the excess tenon and install the in.-dia. dow-

table, cut the waste from the tongue, which now forms the beveled shoulder. Sand the tongue and bevel carefully to remove sawmarks.
Dry-fit the doors and back panel, making adjustments for a snug

but not-too-tight fit, and then glue and clamp the frames around the
floating panels. In keeping with Shaker tradition, I peg the doors'
mortise-and-tenon joints and the back panel. To save time and to free

els that stop the spinner in the open and closed positions. A final sanding and the finish of your choice completes the cupboard. I normally use Watco oil on my cherry cupboards, but I ebonized my birch cupboard following the process in McAlevey's article. I applied two coats of the steel wool/vinegar ebonizing
mixture, sanding lightly between coats. Some experimenting with

up clamps, I drill for the in. pegs right away, pound them through and then remove the clamps. Excess peg length is sawed off both
sides. Although you can plane and hand-sand the doors and back

ebonized scraps showed oil does not give very satisfactory results, but a sprayed lacquer finish brings out a rich ebony color and provides a satin sheen. Don't polish the lacquer. Remember, this is
a Shaker cupboard, not a grand piano.

panel to final dimensions, I rely on the wide belt sander at a local mill shop to save me hours of labor. 1 have the panels sanded to 180grit, which leaves barely perceptible cross-grain scratches on the
rails. I easily remove these scratches with a vibrating sander.

Christian H. Becksvoort builds custom furniture in New Gloucester, Maine, and is a Contributing Editor to FWW