Demilune Redux- defense of design
Bill Tindall, E.Tn.
There were criticisms of the demilune table design from persons that have a good eye for design. This post is an answer to the questions and criticism raised and an opportunity to discuss better ways of accomplishing the need.
First, form must follow function when building something to serve a purpose. In this case I had a request to build a table that would be located at an entry to my daughter’s house. Tables in this location provide a place to set stuff coming and going. The table was to have a drawer, and it had to fit between the entry door and another door located 3’ away. Additionally, the table must not restrict movement in the area of these two door passages.
A Demilune (half moon) design fit the requirement for the space without presenting corners that would obstruct passage in this area, while providing enough depth to be stable and useful. Demilunes were common to the Federal period. They usually had fancy skirts and sometimes had a satinwood panel inlayed in the skirt that could resemble a drawer front. I never saw one with a drawer, a near insurmountable engineering challenge to add to the design. Never-the-less I set out to design a Federal inspired Demilune. The drawer would be inspired by the satinwood panel found in Federal tables. Instead of satinwood, it would be birds eye maple rescued from the damaged back of my grandmothers china cabinet that was restored for this same daughter.
Demilunes are straightforward to make. The curved skirt is assembled in one way or another, veneered, and the front legs attached with bridle joints and the back legs with conventional mortise and tenon. Cut a drawer opening out of this skirt arrangement and there is little left of the skirt between the legs to hold the table together. The engineering solution was to make “rails” like in a conventional table between the legs and tie them into the curved skirt beyond the legs with mortise and tenon joinery. PIC 1
The space between the back legs and front legs is veneered with crotch walnut. The rails between the front legs are not veneered but were chosen to present an exposed plain rift grain on the front edge. Additionally, two filler pieces of rift walnut were added between the leg and drawer opening to cover the secondary wood of the skirt in this location. The visual result is shown in pictures 2 & 3. PIC 2 and 3.

Several objected to this result, namely figured walnut between the front and back legs and rift grain and a maple drawer between the front legs. Indeed viewed from the perspective of picture 3 there is a discontinuity at the front leg. But, nobody is ever going to view the table from this angle and distance to detect the mismatch of rift grain. It is all plain enough to become featureless at normal viewing, especially after finishing, Figure 4.
PIC 4
Personally, I think it works to have the light colored maple drawer between the font legs. The 100 year old maple looks as good as satinwood in this application. I will not further defend this decision.
I chose to use plain rift grain to frame the drawer opening. An alternate suggestion was to use crotch. Using crotch would have involved adding crotch veneer around the drawer opening, which would have left the edge of this veneer exposed where it would eventually chip. Using thick veneer would solve the chipping risk but then there would be a mismatch of skirt thickness at the bridle joint- thick veneer on the drawer side and conventional thickness on the back side. I suppose this issue could be worked out, if it was worth working out.
Figured wood can be overdone. To my eye highly figured walnut framing the figured drawer front would be too busy and distract from the drawer which I intended to be a focal point of the front. From a typical viewing angle and distance, the plain walnut framing for the drawer provides a featureless border to the drawer, just the featureless look I was after. The leg provides a clear separation of the figured skirt from what is going on with the front- drawer and framing.
Conclusion- I think the cock beaded maple drawer works in the design. While I agree with the grain mismatch objections when viewed straight on, up close and not finished, I believe the issue becomes unimportant as the finished table will be viewed in use. The potential problem was not worth the effort to figure out a more elegant solution......unless someone has an easier solution I didn’t discover.




