Main box stops

From previous posts, we have seen that there is a great deal happening within the aprons. Now, we need to make room for one last thing…

Proper right apron interior showing upper & lower racks with spring mechanism.

In this post, I described the process fitting the stop rods keeping the top from shooting off the back of the carcass when the works are set into motion. Something similar is needed to keep the main box from coming out the front. Two (fairly) sizable bolts, to minimize the force exerted on the stops, were used in the original. These need to be installed from within the side compartments of the main box such that:

  • they avoid contacting the lower racks,
  • the grooves in which they run must not protrude through the carcass front,
  • the main box is allowed to move to its fullest extent,
  • they avoid interfering with existing guides and drives located on the main box sides, and
  • they are installed above the main box floor.

Not too much to ask for, right?

Proper right compartment in the original main box. Circle shows stop bolt head (barely visible).

When Oeben created the original, he grooved the aprons just below the lower guides for the main box. This out-of-the-way location prevented anything else from interfering, allowing the bolts to slide unimpeded. So long as the bolts are installed behind the front legs, these channels won’t show through the carcass front. The other ends, naturally, aren’t a problem since the main box stops behind the rear apron.

Layout showing vertical alignment of main box and carcass.

This addresses the majority of the above criteria except for the “horizontal location”, or how far above the main box bottom are the bolts mounted? The challenge here lies in the limited available space between the lower racks, and the main box floor; less than 1/4 inch. Not much! Referring, once again, to the original, Oeben solved this by drilling a hole through the lower racks for the stop bolts to protrude. This has the added advantage of strengthening the bolts since overtime repeated collisions might loosen the bolts mounted in the wood alone.

Main box proper left side showing proposed stop bolt location.

As with the stop rods previously once these bolts are installed, they will be used to mark the location and extents of their grooves, and stop plates.

Postscript: having trouble visualizing things? Check out the first image in this post.

Cognacq-Jay Museum

In fall 2017, the Cognacq-Jay Museum was my first official appointment in Paris. I became aware of it through searches in woodworking discussion fora back in 2011. Their diminutive table is unlike the others in several regards…

The first thing one notices is the striking difference in marquetry motifs of this table versus other Oeben tables of similar function. Typically, the aprons are covered by parquetry, and the top in marquetry. This table, however, features an assortment of flowers connected by a circuitous vine on a background of yellow veneer throughout both the top and sides. Two curators have independently commented on its similarity to a known fabric pattern. Something else that is distinctive about this marquetry, I was told by Rose-Marie Mousseaux, the museum director, is the flowers appear to have been rendered life-sized. It is as if the marquetier laid the flowers directly on the veneer to cut them. The marquetry on the legs appears to be nearly symmetric. Viewed from either side as well as front, or back, the leaf and flower patterns are near-mirror images. Perhaps they were packet cut? Inside another visual anomaly is the veneer covering the interior of the compartment lids. This is typically tulipwood arranged in a reverse-diamond match (image below right of the Getty table). While the pattern is consistent, however, the choice of veneer species is not. It appears to be mahogany (below left image – Cognacq-Jay table).

Finally, this table has no keyholes! Without locks, or winding axles, this table can only be opened by manually pushing the top back.

First most obvious thought is that we are not seeing the table in its original state. Approximately 260 years have elapsed since it was built, and the differences could simply be due to changes and modifications from repairs. It could also be that this table is the work of someone other than Oeben. In his style…, so to speak. Although this is unlikely, the piece is not signed, and unfortunately, there is little existing provenance on it. Finally, perhaps this was a prototype constructed to test the concepts of a movable top and main box to refine the mechanisms [or a demonstration piece!]. This is a supposition. However, with a piece this complex a good maker would want to demonstrate the concept before showing it to prominent clients. It is reasonable that the racks and gears responsible for operation of the top and main box would be tried “small scale” first.

I appreciate the opportunity to view this piece first-hand, and want to thank Madam Mousseaux and her staff for hosting me on the museum’s off-day.

A note on my Oeben table examinations

Since before my first encounter with one of the two Oeben mechanical tables at the J. Paul Getty Museum in Los Angeles in 2012, I have been fascinated (obsessed some might say!) with this atypical furniture form. To learn more about them and their “works”, I arranged a series of examinations. In 2016, I got up close with the two other tables in North America: one at the Metropolitan Museum of Art, and the other at the National Gallery of Art. In October 2017, I visited three publicly displayed tables in Europe: the Cognacq-Jay Museum (Paris), the Rijksmuseum (Amsterdam), and the Victoria and Albert Museum (London). During these encounters, I made a series of notes which I will share in upcoming posts.

To this end, I want to reiterate that I am not a conservator, nor have I had any curatorial training. When I examine these pieces, I do it from the perspective of a woodworker trying (as best I can) to put himself in the shoes of the period craftsman. What I present are my interpretations. So, I am happy to discuss alternate viewpoints.

I’d like to thank the conservators, curators, and administrators who took time to arrange and accompany me on these visits.

A sticky situation

Apologies up front. This is going to get a bit technical…
Hopefully, this figure will help you understand how everything relates.

The parts of the table specified.

Now that I have latch plates, it’s time to start the process of fitting them into the prototype. It’s straightforward:

Underside of the top showing one of the racks. The chalk marks denote the travel extents.

Begin by locating the stop plates (small, hardened metal plates which impact the stop rods during opening) in the table top. Install the stop rods in the aprons, using them to define the grooves in which they will eventually be housed. The extents of the table when open, and closed can be determined by positioning the top (and main box) in each state. Since the two pieces are connected through the mainspring barrel drive gear, this also governs the plate locations limiting the main box in grooves nestled out-of-sight within the aprons. Bolts visible within the side compartments of the main box protrude through its walls, and travel in these channels. With the open, and closed positions of the main box and top defined, the main box latch plate can be marked and mortised. As with everything in this project, sequencing is critical!

The prototype table (stripped to almost its essentials!) showing the areas of consideration.

So, I begin by reassembling the prototype carcass, and installing the racks and guides. For the moment, let’s leave the mainspring barrels out to simplify maters, and just concentrate on getting the top to move smoothly. Problem is, it doesn’t. It binds. Still. This was the issue three years ago when I first installed them. They were each slightly stiff, and needed to be “opened” a bit by the machinist. Now, individually they slide smoothly, and freely, but they’re still binding when assembled…

Observations and thought follow…

Consider that when the original table was built around 1750, micrometers weren’t a standard tool in an ébéniste’s kit (they still aren’t). So, while the various parts and pieces fit tightly together, it should not require fractions of a millimeter alignment to get things working.

Check each guide by holding a straight edge to its back – it’s ever so slightly convex, bulging into the travel path. Could this be the hang up?

Just a slice of light between the rack and rear apron.

The gaps in the back panel through which the top racks slide when actuated feel crowded – but we can see light between them, and the rack when deployed. Remove a few shavings to ensure clearance…

Could it be due to wood movement? It’s possible, but…

Following about 40 minutes of discussion with Jim, my studio “neighbor”, I realize the upper racks have been installed just slightly askew. Rather than the distance between them being identical at each end of the rack, they are farther apart on one end than the other. So, if I install the top from the rear of the carcass into the guides, it fits albeit tightly. However, when I attempt to slide the top into the carcass from the front, the racks are too far apart to fit into the guides. To compensate, remove one rack, adjust its mortise, then reinstall it. This can be prevented in the future version by using a spacer stick or panel during assembly. But that’s not the only problem in this case.

Underside of top illustrating out-of-parallel racks (lines exaggerated for clarity).

Turns out, when the top racks were installed they were set against the guides, marked, and fastened. This left no space between the guide and the vertical part of the rack where it passes the guide on its way to the top. What I’m feeling is that friction is increasing as the top is slid into the guides – the further it goes, the more these rub against one another. Try spacing the guides further apart, about 1 mm – friction decreases!

By opening this gap 1 mm, the top slides freely!

Eventually, addressing all of these issues I’m able to get the top sliding smoothly through the guides. When the mainspring barrels are added, and the top is pulled into its closed position, it gently swooshes into its open position when released. Success (finally)!