An X-ray of a Dovetail Joint, & Letters and Comment on Puzzle Mallet
Excerpted from The Chronicle, Vol. 30, No. 2, 1977
by Christopher Tahk
While the skill of a joiner is evident in the quality of his cabinetwork, it may also he apparent in the tools he makes for his own use. In the Cabinetmaker’s Shop of the Farmers’ Museum, New York State Historical Association. Cooperstown, New York, is such a tool: a joiner’s mallet of ingenious construction demonstrating its maker’s knowledge of the subtleties of his craft.
As is apparent in the accompanying photograph (Figure 1). the exterior of the mallet suggests it has been constructed of only two pieces, each of a different hardwood. One piece serves as the handle and is securely held to the other that forms the head by a dovetail join. The cut of the dovetail is such as to prevent the head from flying off the handle when the tool is in use. This construction seems unexceptional until one considers how the mallet was assembled from the two component parts. Specifically, in view of the cut of the dovetail, how was the handle inserted into the head? Note that the handle and the head appear to be (and, in fact, are) each one piece and that neither has been cut and reglued.
Unfortunately for those who have been intrigued by the structure of this crafty join, the mallet and handle are not separable; even gentle persuasion would cause damage. Another approach, however, X-ray radiography, has succeeded in revealing enough of its intricacies to allow a good guess to be made as to its actual design.
The radiograph was taken in such a way as to give an interior view of the mallet head and mating handle parts as seen looking directly into one of the two striking faces. In the contact print of the radiograph (Figure 2), the mallet appears as it would in an ordinary photograph if it were made of a grayish glass instead of wood, the handle grip being too transparent to be visible. The interpretation of the radiograph, diagrammed in Figure 3, is that the head contains two wedges that hold the handle part of the dovetails firmly in place by friction and, possibly, with the assistance of some glue.
There are probably a number of similar cuts of the mallet head and handle, which would he in accord with the radiograph as just interpreted. One, supported by another radiograph, is depicted in Figures 4 and 5. Whether or not steam bending of the handle prongs would be a necessary part of the assembly step shown in Figure 5 is not clear.
The author wishes to express his appreciation to The Farmers’ Museum, Cooperstown, New York, for the opportunity to examine the mallet; the Cooperstown Graduate Program in the Conservation of Historic and Artistic Works for use of the X-ray equipment; and Minor Wine Thomas, Director of the New York State Historical Association, for reviewing the manuscript.
Dr. Christopher Tahk, Associate Professor of Conservation Science, Art Conservation Program, Cooperstown Graduate Programs ( Cooperstown New York), is a chemist with a long-standing interest in American clocks and woodworking tools whose appreciation of the latter was greatly increased through his experience in carrying out conservation work on the major part of the Garland Patch Tool Collection at Strawbery Banke, Portsmouth) New Hampshire.
Letters and Comment on Puzzle Mallets
Excerpted from The Chronicle Vol. 31, No. 1, March 1978
With all due respect to Dr. Tahk for his opening the can of worms regarding the intricacies of construction and assembly of the Farmers’ Museum mallet (Chronicle, 30, 2, June, 1977), I commend him for his interest in the beautiful and thought-provoking product of a master.
But I do not believe that the splines are wedge-shaped, as they appear to be. If they are, assembly would have necessitated that each spline – along its entire length – be “jumped” into its final position at the exact moment of maximum penetration; and this “jumping” is a mechanical impossibility.
I propose that the splines are conventional, parallel-sided, dovetail tenons. The face of the mallet opposite the insertion of the handle was wider than the face at the handle. The base of each dovetail mortice was parallel to its corresponding mallet-face. So the tenons were splayed from the centerline of the handle and its extension through the mallet-head. The degrees of the angles of divergence are impossible to state without knowledge of the exact dimensions of the head. Assembly of the tool very probably included steaming of the tenons. After assembly, each side-face of the head was dressed down, from “nothing” – where the handle entered – to a lesser-width-than-when-assembled of the face where the tenons emerged. This artifice of thinning the tenon toward its emerging encl would account for the wedge-shaped appearance on its face. As to the use of free wedges to seat the tenons, I believe that wedges would be unnecessary, and I am not convinced that they appear in the contact print of the radiograph.
Further detailing of construction and assembly would necessitate diagrams. etc. and would result in a lengthy article rather than a letter.
May the unknown joiner who made this tool look down on us mere mortals and enjoy the continuing appreciation of the evidence of his mastery of the trade.
— Tom C. Fuller
Mr. Fuller’s interesting suggestion as to the design of the Farmers’ Museum mallet is perfectly reasonable if only the external appearance of the tool is considered. There is no question that one could be assembled precisely as he has described. In fact, before I had an opportunity to study the piece in detail, an ingenious friend who happened to see it, offered exactly the same thoughts concerning its internal structure. It was something of a surprise. then. when a more detailed examination. including radiography, indicated its design was more complex. In the following. I will try to answer Mr. Fuller’s objections to my original proposal and attempt to show why, while the mallet’s construction may not be precisely as I have described. I believe it is more along those lines than the ones he has outlined here.
I certainly can appreciate that the print of the radiograph in the Chronicle does not provide convincing evidence for the presence of the wedges. Unfortunately, this copy has lost considerable contrast over that in the original and the faint but distinct white lines separating what I have interpreted as wedges from the handle prongs are all but invisible. One of these is just barely discernible in Figure 2 of the article and is located between the upper prong and the wedge below near the handle
(left) end of the mallet head.
There is a second bit of evidence for wedges. In dry weather, the mallet handle is slightly loose in the head and can be withdrawn about a quarter inch without undue violence. Further easy withdrawal is probably presented by the fact it would require inward compression of the outward-splayed handle prongs. Looking into the two small openings produced in this manner. one on each side of the head. it is possible to see, admittedly with difficulty, what I have illustrated in Figure 1. To be sure this view is not purely a product of my fancy. I have asked others for their account of what they see and all are in agreement with my version.
In Figure 1, one is looking almost directly into the crack formed in the mallet side. The mallet handle is on the left and its head to the right. The view is slightly from above and from the head end of the mallet. On the side illustrated, the wood of what I believe is a wedge is apparent behind that of the handle prong dovetail and is either diffuse porous or coniferous. In contrast, that of the handle is strongly ring porous. Also, the grain of the wedge runs parallel to that of the handle but at right angles to that of the head and so cannot he confused with the latter. The crack on the other mallet face shows the same view but is less convincing as the putative wedge also appears to be ring porous. Note that the front edge of the wedge is tapered, probably because it enters (or the maker thought it might) the region of the dovetailed section of the head as it holds the mallet handle prong in place there when the handle is fully inserted.
The principal objection Mr. Fuller raises to my idea of the mallet’s construction is that it represents a “mechanical impossibility” which requires “that each spline – along its entire length – to be ‘jumped’ into its final position at the exact moment of maximum penetration.” While it may appear that this is the case owing to my inadequate drawings that accompanied the article, this is definitely not true. Assuming the two outer handle prongs are flexible, whether naturally or from steaming, in the assembly step, each prong will slide smoothly into its mating mallet head dovetail, there being no mechanical discontinuity at any point in the process. I have attempted to illustrate this point in accompanying Figures 2-5 which show the assembly at different stages of the insertion process. the mallet head assumed to he transparent and handle parts becoming shaded as they enter the head. The important point here is that the handle prongs are free to move outward into the dovetailed regions of the mallet head during the insertion process. The extent of such outward motion of a prong into its mating dovetail mortice is restricted by the latter to an extent that is directly proportional to the extent to which the handle has been inserted into the head. For example, when the handle is halfway into the head, the prong can expand outwards halfway into its mating dovetail but not further.
An alternative way of looking at the problem is to realize that Mr. Fuller’s design and the one I have proposed do not differ significantly in mechanical principle. A prong of the handle of his mallet, in place in the mallet head, would appear as shown in Figure 6 before the mallet side was cut away to lessen the head bevel as he describes. (Note, the mallet head itself is not shown in the illustration.) The cutting away of the sides of the mallet head would remove an outer wedge-shaped piece of the prong (see Figure 7). If two further cuts were then made on the prong, another wedge of wood is freed from the inner prong surface. The prong (Figure 7) is now identical in shape to the one which would mate with the tapering dovetail mortice of the head illustrated in Figure 4 of my article. The wedge removed from the inner prong surface by the second two cuts is, except for the taper in the plane parallel to its base, identical to one of the wedges I believe present in the head of the Farmers’ Museum mallet.
Mr. Fuller, I am sure, believes his mallet is not a “mechanical impossibility” and that his handle can be smoothly inserted into its dovetail mortices in his mallet’s head. The only difference between his handle prong shaped by cutting and the one illustrated in Figure 7 is that, in his, the inner surface wedge of each prong is an integral part of it rather than having been cut away. Certainly, if a prong with an integral wedge can be slid smoothly into the dovetail, so could it be if the wedge were independent of the prong!
The only difference between the design of a mallet with a handle of the type illustrated in Figure 7 and the one illustrated in my article lies in the shape of the wedges. The mating head cross-section for the Figure 7 handle is shown in Figure 8. Notice that it is only a slight modification of the design I originally presented (Figure 4 of the article), differing only in the thickness of the dovetail mortice. Specifically, the only difference in the two designs is whether a wedge of tapered or rectangular cross-section is to be employed. Both designs would work equally well. Which design, if either is present in the Farmers’ Museum mallet may be a moot point, the evidence being locked securely within it.
“While we sit in a circle and suppose, The riddle sits in the middle and knows.” Robert Frost.
Editor’s Note: Roy Underhill shows how to build this mallet with a rising dovetail joint on Season 21, Episode 7, of The Woodwright’s Shop and in the April 2012 issue of Popular Woodworking Magazine (“It’s a Mystery”). I don’t know about the mallet shown above, but in Roy’s version, there are no wedges.