Dovetails have always been joinery's hallmark of craftmanship. It's an old joint to be sure... it pre-dates written history, and we have functional examples of dovetails that go back to the earliest dynasties of ancient Egypt and China; well before the European dark ages. Although we now use dovetail design style as a way to date and distinguish time periods for furniture and construction, it's no wonder that the concept of dovetail joinery has survived for thousands of years. When things like joint strength, durability, joint longevity, space allowance, and beauty are taken into consideratioin, it is far and away the best way to join two perpendicular pieces of wood, and has been for a long time.
I find it somewhat amusing how time has changed the perception of putting them on display... Today, dovetails are shown off as a mark of quality, indeed, for their own asthetic beauty. Some dovetail methods, methods I might add, that took exceptional skills to perform, were born to hide them all together.
|
|
|
|
Dovetails come in a few different "flavors" so to speak. Some are still very common while others are no longer in vogue. They are all variations of a theme, each with benifits and detriments. The two biggies are English and French dovetails.
One of those styles that no longer sees a lot of favor is the pin and cove dovetail. It's the second one from the left up above. Around 1850, when America (and Europe) began to mechanize, master woodworkers were falling from grace rather quickly; there simply weren't enough of them to keep up with the demand for furniture. An American invention, and North American manufacturing's answer, was the pin and cove, a mechanized method which could be cut more quickly from a pattern by someone with a lower level of skill. Not actually dovetails, pin and cove only lasted about 25 years or so... roughly 1870 to 1895. By then, "router" improvements and assembly line interchangeable drawer parts were cranking out picture perfect machined dovetails, and were firmly entrenched in the furniture industry. The whole Arts and Crafts movement was about the displaced master woodworkers taking back their craft from the machine. Still, pin and cove has a certain historical "charm" if you will, and it's a great way to date some late Victorian and Eastlake furniture.
English dovetails...
French dovetails, also called sliding dovetails, are comparitively quick and simple.
They've long been a staple for shelves in bookcases and cabinets, joining rails
and stiles to table legs, etc. In fact, it's odd that they're called
sliding dovetails in almost every situation they're used in, except drawer
construction. When used for drawers, that's the only time I can think
of when they become French dovetails. Anyway, when used for drawer
construction, the socket is cut from the bottom. Also, the
socket stops short,
and isn't cut all the way through. The pin can be either straight or
tapered, but is usually tapered to make it a "locking" joint. Regardless,
with the socket stopping short, it
conceals the joint's construction
except when viewed from the bottom. French dovetails are often seen on
drawers with curved (think serpentine) fronts because of the nightmare it would
be to cut and fit traditional English dovetails with those curves.
French dovetails have a couple of problems. First, the sockets can't be on the extreme edge of the mating pieces. Once the glue failed, the drawer side would simply fall off. It can't be inset slightly... the first time the drawer hung and got yanked, it would fail and the side would simply fall off. It has to be inset far enough for the drawer to maintain mechenical integrity. The problem with this is the loss of space. Assume a drawer that's 16" long and 4" deep. Inset both sides by 1" for the drawer to maintain enough mechanical strength to survive for a few years, and you've lost (16" x 4" x 2 sides)/144 or approximately .89 cubic feet. Give that up in your freezer; no ice tray... or ice maker! And, of course, the second problem is that they're French. I don't know that they've ever produced anything I admire, except the Stature of Liberty and they exported that.
I plan on trying my hand at the sliding (not the French) dovetails in the near future on a couple of bookcases. I'll let y'all know how they come out.
Pins or tails first?
Dovetailing is a method of joinery that is said to be self squaring. Stock needs to be adequately, not heroically, milled. Once the stock is flat and square, if dovetails are cut correctly (read perfectly), it is, indeed, self squareing. However it's rare that the pins and tails are "perfect," so some very small degree of umph is typically employeed to square the final assembly. Keep in mind that, unless the pins and tails are horribly deformed, we're talking about fractions of a single degree here.
What follows is a faithful representation re-printed from Manufacturer and Builder. Manufacturer and Builder was a building and construction trade magazine that ran from 1869 through 1894. This particular article was from Volume I, Issue I, 1869. The original article layout was a three column format, and can be viewed here This is part of a collaborative effort called The Nineteenth Century in Print: The Making of America in Books and Periodicals , between Cornell University Library and the Library of Congress.
The strongest and most permanent joint made in carpentry and cabinet-making,
where pieces of wood are fastened together at right angles, is the dovetail.
When made in some of its most approved and perfected forms, it is equal, in
neatness and artistic finish to the mitre joint; The mitre is a comparatively
neater method of joining wood at an angle, but the dovetail possesses the
greatest possible strength. It is generally employed in articles made of thin
materials, such, for instance, as drawers, boxes, chests, etc. If we
examine a dove-tailed box, we observe that it consists of six pieces, or sides,
four of which are interlaced, or dovetailed together at the corners, forming a
rigid framework. This is shown in Fig. 1. When properly put together,
the joint formed by the two pieces A and B is strengthened and braced by those
of C and D. The rigidity of the box is of course still further increased
by attaching the bottom, which may be done by means of screws, or glue and
screws may be both employed. There is an advantage in the employment
of the latter, for if the dovetails should be somewhat loosely fitted, or the
glue lose its adhesive power, the screws will prevent any disposition of the
four sides to rack. When we wish to have the lid of considerable depth,
and fastened together by dovetails, the box is made sufficiently deep to form
both box and cover. The top and bottom pieces are then added. The
six pieces so joined have the appearance of a rectangular block or cube.
A saw is employed to separate the cover, or lid, from the bottom portion.
This method has the advantage of saving some labor in dovetailing. It
insures the exact agreement in size and form of box and cover. It would
be somewhat difficult to effect this were they separately made. This would
especially be the case in making such articles as desks and writing cases.
In all boxes where the bottom and lid are made together, the line of division
is marked on the four exterior sides. One of the dovetail pins is placed
on that line. This pin should be about twice as wide as the others.
When divided in forming the cover, either part is then of the size of the others.
In order to have the cover and box show a mitre joint, the dovetail and joint
pin are made to a mitre; whereas, if the pin were left square, or made as usual,
the box when cut open would show the rectangular lines of the pin and dovetail.
The top and bottom of the box may be fitted in various ways. They may
be glued, or otherwise fastened, on the square edges of its sides. They
may be rebated; or, to cause a more finished appearance, they may be both
rebated and mitred.
In Fig. 2, the common dove-tail joint is shown, in which the dovetails and
pins are seen upon both sides; When the gr~Un of the wood of the parts joined
together runs in the same direction, such parts will expand and contract equally,
should they be equally moist and dry. This expansion or contraction can then
take place without injury to the work. In order to effect this, it is advisable
to niake the work from the same board, or from pieces of boards of The
Manufacturer and Builder. The same quality of timber, and having come as near
as may be from the same relative position in the logs from which they were sawn.
This can be very readily ascertained by an examination of the ends of the pieces,
the curvature or configuration of the grain furnishing the desired information.
If the pieces of work have the grain running in contrary directions, as seen
in Fig. 3, it is evident that the portion D would entirely prevent the
expansion of C, and the restraint would cause injury to the joint. This would
be the inevitable result in case of much expansion or shrinkage. In order to
explain this more fully, let Fig. 4 represent the surface of a board, its edge
being indicated by a b, and the end by c d. It is evident that no contraction
can occur in the direction of the line a b, which is that of the grain of the
wood. Along this line the wood-fibres remain quite rigId, the shrinkage taking
place in the direction of the line c d. The flexible fibres are in this direction,
and more or less disposed to become curved from exposure to the atmosphere.
The four marginal lines of the board are not likely to change very materially
in respect to each other, and if made parallel and square at first will remain
so.
Fig. 5 represents the ordinary dovetail joint. Fig. 6 shows the parts before
they are put together.
The pins are shown at P, and the dovetails at C. The pins and dovetails
are commonly made of about the same size, as this gives the strongest kind of
attachment; but in many forms of cabinet work the dovetails are made on the
front, or more exposed portion, and the pins are cut about one fourth less in
size than the dovetails, in order that as little as possible of the end wood
may be seen. In laying out dovetails, the sides and ends of the boards
are first marked across with a gauge or square, to indicate the inside measure
of the box or drawer, and also the bottom of the pins and dovetails.
The portions beyond the lines are left a little longer than ultimately required.
In laying out the pins very little care is required. It is a common
practice to mark them off at random. Care must be taken to leave the
outside pins nearly twice as large as the inside ones. The dovetails
are then marked from the pins; they thus become exact counterparts of each
other. The piece upon which the pins are made is laid upon the work bench,
and that upon which the dovetails are to be made is held vertically upon it
in the exact position that the two pieces are to occupy. A fine point
or pencil is then used to mark the outline of each pin. This being done,
a fine saw may be used for both pin and and dove-tail, taking care that it
follows the outlines and does not cut below the gauge mark that limits the
interior of the work. By not cutting into the outline marks, both pins
and dovetails will be a trifle too large, and in driving them together they
will compress each other a little, and produce a close and accurate joint.
After cutting with a saw to the depth required for the pins and dovetails, a
sharp chisel is used to remove the wood that remains between the portions that
are to be retained. A ready method of doing this is to lay the several
pieces one upon another, similar to a flight of stairs, and cut half way through
them successively, and then turn them over, arrange them as before, and cut
the other side, which operation removes the pieces. This method enables
a mechanic to work with celerity, and also permits him to see what he is doing.
The chisel should be held so that the cut will extend inward at the base of
the pins and dovetails, meeting at the centre of the board. This will
insure a tight fitting joint when the work is put liogether.
It is advisable not to make the pins and dovetails with any considerable degree of bevel, or the connection will be deficient in strength. When pine or soft, brittle wood is employed, the parts should be slightly beveled; but more bevel should be allowed to hard wood, such as oak or mahogany. In making the outlines of the work upon some hard woods, mahogany for instance, it is difficult to observe the lines; but if the surface of the wood be rubbed with chalk before marking, the lines will be rendered conspicuous. In gluing the dovetails of a box, as seen in Fig. 1, it is best that the pins on the side A be first sparingly glued, and held by some sort of fixture in a vertical position, with the glued end uppermost. The end B, which is to be attached, is held horizontally, and forced down upon A. This is conveniently done by blows of a mallet or hammer, which, to prevent injury to the work, are received upon a piece of waste wood placed directly over the pins. When the pins come flush with the upper face of the board B, the piece of waste wood is placed beside them, so as to allow the pins to rise above the surface. It is essential that the pins and dovetails should exceed the thickness of the board, so that when the superfluous length is planed off, a good, clear joint is apparent. The second side, C, is then glued, similar to A, and held in the same manner, and B is forced down upon it. The three sides of the box are thus united. The pins upon the other ends of A and C are then glued; the end B is placed face downward upon the bench or floor, resting upon two smooth strips of wood, placed close beside the projecting pins. Care should be taken to have it rest firmly. The remaining side, D, is then driven down upon it, and the four sides of the box united. The superfluous glue, forced out by the crowding together of the parts, is then scraped off, while yet soft, with a chisel; but, where moisture will not affect the work, the remaining portions of the glue may be washed off with a sponge, wet with hot water from th~ glue kettle. Where it can be safely applied, this is the best method, as every vestige of the glue is removed from the surface of the work.
To ascertain if the box has become rhomboidal, or, as it is technically termed, out of square, it can be tested with the fitting square, and it should be immediately corrected before the glue becomes set, or dry. This correction can be made by pressure applied to the longer diagonal. A ready means of ascertaining any deviation from a square is to find the exact distance that should exist between diagonals, and then to cut a thin strip of wood of just this length, and so sharpen the ends that they will be more acute than the angles of the box. It will exactly fill the space between the diagonal lines if the angles are correct that is, if they are exactly ninety degreecs.
In some kinds of dovetails the pins are more or less concealed. Fig. 7
shows a form called the lap dovetail, which is generally employed for the
fronts of drawers.
Fig. 8 shows the manner of attachment. The pins are on the front of the
drawer, and are first made. Gauge lines are marked to denote how far the
pins shall extend inward. It. will be observed that neither pins nor
dovetails extend quite through the wood in which they are respectively formed;
hence the concealment of their end grain. It is, however, immaterial in
which piece the pins or dovetails are made. The depth of the latter is
generally about one eighth of an inch less than the thickness of the piece
containing them, and the pins are of a corresponding length. If the ends
of the pins were made even with the end of the wood,.they would reach the bottom
of the dovetails, and an open corner would be the result. The pins are
generally cut back one eighth of an inch from the front end; otherwise the
amount of the dovetails would be shallower tban the thickness of the piece in
which they are made. In drawers, the sides are about half the thickness
of the front; the dovetails are, therefore, cut quite through them, the same
as in the ordinary dovetail, shown in Fig. 5. If the front part of the
drawer is to be covered with veneer, the pins are cut quite through with a saw,
the kerfs being entirely concealed when the veneer is applied.
Fig. 9 shows the mitre dove-tail joint, and Fig. 10 represents its construction.
This is the neatest and the most difficult to make. The pieces must be
cut to the external dimensions of the work, and then rebated out square at each
end; and after the pins and dovetails have been formed, the square rebates are
cut into a mitre joint, a rebate plane being used for the purpose. When
properly finished, neither the pins nor their mode of concealment can be seen,
and the work has the appearance of a plain mitre joint. As in the mitre,
this rebate should have an angle of forty-five degrees.
In Fig. 11 is seen another form of finish sometimes given to the lap dovetail,
shown in Fig. 7. When the two pieces are put together, they resemble the
form seen in Fig. 12, and the final finish is given by simply rounding the
corner, removing the angular projections, until it assumes the appearance seen
in Fig. 11. Boxes are often made by rebating one side or piece, the end
of the other being received into this rebate, and then glued and nailed.
The successive operations present the appearances shown in the cuts. It
is quite a quick and cheap method of joining, and, when nailed from both angles,
is, next in strength to the dovetail.
Fig. 13 is a simple and strong fastening, somewhat resembling the exterior
appearance of the mitre dovetail joint. It is termed the mitre and keyed
joint. Small boxes, and similar light articles that do not require much
strength, and where cheapness of mannfacture is desirable, are thus made.
The extremities of the end and side pieces are mitred or planed to an angle of
forty-five degrees, and attached by glue or suitable nails. A few upward
and downward cuts are then made with a saw, and pieces of veneer or thin wood,
covered with glue, drawn into them and allowed to dry. These pieces are
then cut off flush with the sides of the work. If properly done, the ends
and sides will form aiigles of ninety degrees with each other, or, in other
words, the box will he exactly square.
As you can see, not much has changed with dovetail joints in roughly 140 years since this article was written. Indeed, not much has changed with dovetails since they were discovered on 5000 year old mummy tombs in Egypt. Of course, the process has become mechanized with jigs like the Leigh, Keller, or Porter Cable, and a router, but the joint is essentially the same.
