Chris Schwarz's Blog

On Gaps and Dovetails and Winterthur

Years ago when touring Winterthur, I saw a lot of wacky Pennsylvanian dovetails on old chests. These joints had been wedged through their pins – a feature I had not seen in person before. While the museum personnel wouldn’t let me take photos, I did make a few sketches.

Whenever I cut dovetails, my mind drifts to these wedges. Were they decorative? Did they close unsightly gaps? How – exactly – were they driven in?

Yes, I know that sounds dumb. The answer is: With a percussive instrument. But is the slot through the entire pin, which means it was cut before assembly? Or is it at 45° like a corner spline, which could be cut after assembly?

I don’t have the answer, but I have been fooling around with the different methods.

This week we are building tool chests from “The Anarchist’s Tool Chest” at Kelly Mehler’s School of Woodworking, and a couple students cut some dovetails on the wrong side of their layout lines. This has created some gaps.

To fill them, I have been investigating these wedged dovetails that I saw at Winterthur and have been reading up on during the last couple years.

Because I have only my visual record to go on, I’ve been trying to replicate what I saw with the minimum amount of effort. So I have been kerfing my pins after assembly and have driven in wedges that are about 1/16” wide at the tip.

The result? The wedge does close up some gaps. It is more effective on closing some gaps than others. I’ll need to shoot a video to explain myself.

The good news is that we assembled five shells of the 11 tool chests today during only the second day of class. The rest will be assembled tomorrow before lunch. None of these five chests needed the wedges I described above. I might just have to wait and find some gappier dovetails to experiment on.

— Christopher Schwarz

19 thoughts on “On Gaps and Dovetails and Winterthur

  1. jacon4

    Yeah, i have been looking for the answer to this question for going on 30 years, ever since i got an 18th century painted blanket chest that had them. Even the drawers had wedged dovetails, it’s very obvious this was done on purpose and not to “fix” flaws/gaps in the chest. It’s Germanic construction no doubt however, no one that i know of can answer why german american cabinetmakers did this. I’d post pics but this blog does not have that feature?

  2. Kelley Blanton

    I have posted some pictures here: http://www.blogger.com/blogger.g?blogID=5965309449786983041#editor/target=post;postID=1757438391430368394

    I imagine it depends on where you live, but most any antique shop in the Shenandoah Valley where I am will one or more of these chests on display.

    I do not know how to make computer drawings and link to them, although that is a good idea a previous commenter requested. Perhaps someone who is SketchUp savvy (PW staff?) could help.

    As far a published information, I am not aware of any of any substance. That might be because everything that I have checked is of an English background, and this comes from an German tradition. Maybe someone who reads German might have seen something.

    The web has little more to offer. All the threads seem to start with someone like Chris S. saying wow look at this, what’s up, and the answer that shuts it down is that someone was trying to tighten up some loose joints. Hundreds and hundreds of beautifully made chests with each and every pin wedged suggests otherwise.

    I have yet to talk to a museum curator who has given much thought to this. When I showed one what I was talking about, she suggested writing an article for FWW. Well okay, this is an interesting woodworking joint for sure, but it has so much more to offer to the study of period furniture. I know that most curators are not from a building background, but a sustained shop practice is just as much a signature as a bellflower design. There is also interesting stuff here about gluing and clamping procedures, always difficult information to come up with about early shops.

    1. Kelley Blanton

      Thanks for the great pictures! I will post some others as well.

      Your point about a chisel stab to get the wedges inserted is a good one, and the lines on the side of the drawer are very interesting. I had not seen that before. If indeed they are traces from a chisel splitting the pin, then it suggests not a very deep split.

      I have two reasons for why I think a saw cut might have been made . First, the chests that I have seen have wedges that are neatly and systematically placed in the pins. There is some slight variation within the same chest, but it is so consistent that I thought only a saw cut could account for that. Splitting that many pins with a chisel you might expect would produce some unevenness due to grain runout.
      The second reason is that very often the wedge is so thick for the width of the pin, it seems to me there has to be some space for all that wedge to go. Say an 18 inch deep chest has 10 pins or more and each wedge is near an 1/8th, the you are looking at over an inch or more of wedge! The flared dovetail cuts account for some of this, but the amount of slope that I see is not very much.
      If a sawcut was used, then of course it would have been made before assembly, there being no way to saw across the entire width of the pin after knocking it together. The cut does not need to go all the way down, which might explain why there no sign of a sawcut at the bottom of apin. However you would expect to find one every now and then, and I have never found one. This would support the chisel opening theory

      At any rate,I have used saw cuts when making these joints. I wll try the chisel cut approach sometime and see what happens. I am probably just too nervous about splitting the tail board.
      Perhaps both approaches were used. It wouldn’t surprise me. Maybe someone will come up with a way of telling for sure.

  3. carguy460

    This is interesting…if you need “gappier” dovetails to try out, I’m a quality manufacturer of gappy dovetails. In fact I specialize in all types of gappy joints. I’m what you call a gap expert.

  4. TobyC

    I would like to see pictures of original examples, but I would also like to see detailed drawings. Or an article showing them being made and used.

    Toby

  5. Village Carpenter

    It’s my understanding that the wedged dovetail is a traditional German method of joinery. They purposely cut the joint loose, knowing they were going to wedge it. It makes for a very tight joint. You often see it used in painted chests where the joint would not be seen. I learned this from Bess Naylor at Olde Mill Cabinet Shoppe, who is an expert in 18th-century PA German furniture.

    1. Kelley Blanton

      I have seen a good number of unpainted chests, mostly walnut, with wedged dovetails. In a diffuse porous wood like walnut or poplar the wedges practically disappear since the end grain blends so well. I was at MESDA recently and had a chance to look closely at 3 walnut chests on display, all with their pins wedged. One had drawers in the bottom, and those joints were wedged as well!

      I think it is important that no one gets the wrong idea of the “loose fit” that is being discussed here.
      Perhaps loose is not even the right word, since it suggests a sloppy fit,as you might get by accidentally cutting on the wrong side of the layout lines. Or due to poor overall cutting technique.
      The loose fit is purposeful in just the same way a wedged through mortise and tenon joint is cut “loose”. The tenon fits the mortise snug at the shoulder end and at the outside the mortise has been made wider so when wedges are driven into the end of the tenon it spreads and can not pull out. This slack fit eases assembly but that is not the purpose of the flared mortise. No one would suggest making the mortise too wide overall and driving wedges into the end to tighten it up.

      Likewise,in wedged dovetails, the fit between the pins and tails is very snug at their base, and open at the top. When driven home the wedge locks the two together.

  6. Kelley Blanton

    Chris,

    I am pleased to see that you are interested in what I call wedged dovetails. They have been an interest of mine for a number of years, ever since I was lucky enough to help restore a pipe organ
    built in Pennsylvania by David Tannenburg, America’s first organ holder. The case was expertly made in a strong German tradition with a great number of dovetail corners . Every pin was neatly wedged. I felt that surely this was not some cabinetmaker covering up hundreds of miscuts, but rather an important trade practice. My apprentice training was mid-late 18th century English and I had never seen this before.

    Since then I have looked at dozens of chests and case pieces from the 18th and 19th centuries and
    have been struck by how many of them have their pins wedged. I would say for the typical Germanic background chest, painted or not, the vast majority have wedged dovetails. When asked about this, museum curators will say the wedges were for tightening up loose joints. Period woodworkers will say the same thing, even though the possibility of miscutting every single joint in a chest stretches the imagination. Nevermind the fact that the boards should be split wide open from so many wedges driven into them.

    I believe that there are several very good reasons for the wedged dovetail joint. The answer became clear to me when I thought about David Tannenberg and his dovetailed organ case. Now we are talking about pieces 8 and 12 feet long, almost 2 feet wide and near 2 inches thick. I don’t know about you, but I get more than a little anxious just gluing up a little dovetail box what with all the glue to spread, clamps with special shaped pads, square to check for, etc. etc. And that’s all with modern slow setting glues and a warm building. Hot hide glue and a chilly shop! Help!

    The secret to a wedged dovetail is purposefully cutting the tails out of square. In other words, not sawing at right angles to the face of the board, but tilting the saw first one way then the other so the socket is wider on the outside than on the inside. The pins are cut as usual, except that each one gets a sawcut right down the center . Now the joint can be assembled dry with no glue. It goes together with hardly any effort since the joint is slack everywhere except at the very bottom, where it can be quite tight.

    At this point the chest is completely together, no clamps , no glue and no banging on tight fitting dovetails. Glue is brushed into each opening , then a wedge driven in which spreads the pin tight against the dovetail socket and at the same draws the joint down against the shoulder, completely avoiding the need for clamps. Each corner can be done individually. In fact, every pin/socket can be glued individually. You could stop and have lunch part way through if you wanted to.

    The ease of assembly is a by product of the main feature of this joint. It solves the one drawback of a “regular” dovetail corner. Rather than relying on glue and a tight fit to be strong in the direction of assembly, the joint is now mechanically strong in both directions. You wonder if they used glue at all. II have looked closely at old chests made like this and while most of the time it was not possible to tell if there was glue or not, a few of them showed glue on their insides, where it was not necessary for it to be cleaned off.

    If you would like I can try to post some close up pictures of antique chests with this style of construction. Especially interesting are shots showing the out of square cuts on the dovetails, and how surprising large the wedges are.

    I believe this is a very important trade practice that has been misunderstood and therefore overlooked in the museum/period furniture world. Many questions immediately come to mind. Some are easily answered, some not.

    Perhaps this lengthy post will start a discussion that will shed more light on this intriguing woodworking joint.

  7. Buche

    Chris,
    I remember seeing an old chest (hopechest?) with what looked like wedged dovetails in north-eastern Germany. Unfotunately, I don’t have a picture, but at the time I thought it made the joint mechanically locked in both directions.

    Cheers,

    Lars

  8. bglenden

    My father has a couple of old(ish) (cut nails) small chests he bought in Nova Scotia. In one of them it looks like this was done as some kind of a repair (the wedges aren’t in only a few pins, sometimes at an angle), on the other it’s in every pin, exactly bisecting it with an identical shape and length etc.

    I’m also not 100% convinced they are wedged, since the “wedges” don’t seem particularly thicker at one end than the other.

    I have often wondered what’s up with them! Let me know if a picture would be useful (I happen to be visiting these days).

    Cheers,
    Brian

COMMENT