Lost Art Press

Category: Roubo Translation

The Tool for Waves

Posted on by meghanlostartpress
Plate 314. Machine Appropriate for Making Flat Wavy Mouldings

The following is an excerpt from “With All the Precision Possible: Roubo on Furniture.” This book is the result of more than a decade of work by an international team that produced the first English translation of the 18th-century woodworking masterpiece: “l’art du Menuisier” by André-Jacob Roubo. This translation covers Roubo’s writing on woodworking tools, the workshop, joinery and building furniture.

In addition to the translated text and color images from the original, “With All the Precision Possible: Roubo on Furniture” also includes five contemporary essays on Roubo’s writing by craftsmen Christopher Schwarz, Don Williams, Michael Mascelli, Philippe Lafargue and Jonathan Thornton.

The excerpt below details a machine that had gone out general use even before Roubo wrote the original text. However, there is no denying that the illustrations and explanation of the device are captivating. The details on it inspired Jonathan Thornton to recreate one of these machines and write an essay on it for “With All the Precision Possible.” A portion of the essay will be the excerpt for next week.

Description of the Machine commonly called the tool for waves, and the way of making use of
it in different ways
The machine that I am going to describe is the largest and the most complicated of all the cabinetry tools, which once were much used. Now they are not used much, since they are only used for works of applied wood [moldings] and they have, so to speak, combined all their science to veneer the wood properly. However, since this tool is ingenious, and you cannot find it anywhere, I thought I must include it here, in order to save it for posterity, supposing that this work succeeds.3

The use of the wave-cutting Tool represented in Fig. 1 is for cutting onto the wood wave-mouldings, or patterns, precise intricate repetitive designs, whether flat, on the face or even in both directions at the same time.

It is composed of a box from 7 to 8 feet in length, by one foot wide and 9 to 10 thumbs in height, exterior outside measurements. This box is open on top and at the ends, such that the distance between the two sides is retained only by cross-pieces A and B, Figs. 1 & 2, placed at two ends of the box, where they are assembled by mortise and tenon. At about the middle of the height of the box is placed a plank C–D, Fig. 2, about 2–thumbs thick, called a sommier [or platform, mattress; in similar machines for printing lithographs this is called the couch or the cooch]. This, for more strength, should be fit together at the ends and braced from below. This plank, or sommier [platform], is held in a groove in the two sides of the box (which should not be less than one–and-a-half thumb in thickness) and serves to hold the mouldings to be wave-cut, as I will explain later, and which you can see in Fig. 2, which represents the machine viewed from above.

In the middle of the box is placed a square frame of about a foot in width, viewed from the side, and which extends from 9 to 10 thumbs above the box, to the sides of which it is attached with some screws, and in which it enters by tenon and a notch, as you can see in the evolution of this machine, represented in the following Plate [315] Figs. 5 & 6.

The width of this frame is determined by the width of the box, the sides of which the uprights of the latter are flush on the interior. It is in this frame that is placed a spring which presses on the toolholder [the cutterhead] E, Fig. 1. This spring is raised and lowered by means of the screw F, Figs. 1 & 2.

The whole machine is held on a base of a solid construction and widened [splayed] in the form of a trestle to give it a better footing. The height of this base should be from 2 feet 8 to 10 thumbs, so that is has about 3 feet in height from the axis of the crank handle G to the ground. This is the most comfortable height for the person who turns this crank handle to have all his strength, whether raising or lowering it.

There are in this machine two movements: one is horizontal, which is done by means of the handle G, Fig. 1, which by making the pinion turn placed in the interior of the box, moves the sommier A–B, Fig. 2, and consequently the work which is held above.

The other movement is vertical, downwards, and depends on the first. The rod, or wave guide/ channel H–H, Figs. 1 & 2 [Plate 315], which is held on the sommier, moves therefore with the latter, is raising the tool-holder F, Fig. 1, left, which then lowers immediately by itself, both by its own weight and by the pressure of the spring placed above. See Fig. 4, which represents a wave channel the size of the execution [ full-size/scale] Fig. 5, [which] represents a moulding completely wave-formed, according to the sinuosity of the wave channel in Fig. 4. Also see Fig. 3, which represents the cross-section of the tool-holder, which I will describe here later.

Fig. 6 represents a cutting blade viewed with different profiles, as large as the execution [ full-size/scale].

Figures 1 & 2 of this plate represent one of the transverse cross-sections of the machine, taken at the location of the pinions and the other the longitudinal cross-section of the same machine, so as to better understand the details of its construction and the mechanism of its operations.

Axis A–B, should be placed in the copper collars, a, b, so that they turn more easily. One should note at one of the sides of the box [is] a squared opening capable of letting pass pinions C–D, supposing that it is necessary to remove the axis outside. Pinions C–D, engaged in the toothed rack c, d, Fig. 1, and E, F, Fig. 2, which are embedded on the underside of the carriage [platform] G–G, same Figure, about 9–lines deep, and are held there by pegs, placed together in the sides of the latter, observing that the toothed racks are well positioned vis-a-vis the other, so that the two pinions C–D, Fig. 1, are contacted equally by the racks [platform] above. However, as it can happen that the teeth of the pinions are not well positioned vis-a-vis the other, one would do well, after having stopped/blocked one of the toothed racks, not to attach [secure] the other until after verifying that it fits well with its pinions, so as to be able to set it back or advance it as necessary.

Plate 315. The Development of the Machine Represented in the Preceding Plate

These racks can be made of iron or copper, which makes no difference for the machine, however it would be good that they be made of copper, given that the rubbing of two different metals is smoother and wears less than if the two pieces, that is to say, the rack and pinions, be of the same metal. [See Plate 314.]

The rods or wave conduits [channels or guide rails for the work piece] e, f, Fig. 1, and H, H, Figs. 2 & 6, should be also made of copper, and they should be bent at a right angle to have the ability of attaching them with screws on the carriage [platform] in which they are notched in all their thickness, as one can see in Figs. 1 & 6.

When you put these [wave] channels on the platform, you must pay the greatest attention that the guilloche [pattern] be not only fit well together, but also that they match at the same point of their contour with the contact of the tool-stand which bears on top of it, as you can see in Fig. 4. This represents the machine viewed from the end, and even better in Fig. 7, which shows the toolstand [tracing and cutting head] where you have removed the cheek [ fence] which holds the iron in place, as I will explain later.

The tool-stand is a frame I–L, M–N, Figs. 2 & 5, of about 2 feet in length, by a width equal to the interior of the box, less the necessary play to prevent any rubbing, which you avoid by diminishing the thickness of the uprights in the entire length, and reserving there some heels at the ends, so that the frame is held against the sides of the box, and cannot get out of place when you move it.

The frame of the tool-stand is attached at the sides of the box by means of two threaded bolts, represented in Fig. 3, half as large as executed here, where the extremity o ends in a cone, and bears on a copper collar embedded in the side of the box.

This screw is held in place in the frame by a nut placed in the middle of its thickness, normally. To prevent the movement of the frame so that it does not turn the screw, you put a counter-screw P outside, which you tighten against the frame, which prevents the screw from making any movement. See Figs. 3 & 5.

As it is sometimes found where it is necessary to lift the point of the movement of the toolstand, you pierce many holes in the copper collar attached to the side of the box, as I did in Fig. 2.

At the other end of the tool-stand, that is to say, where the cutting iron is secured/fitted, the cross-piece I, Fig. 2, should be very strong and assembled with a cover from above so as to present a uniform surface all along the length, which is the width of the tool-stand. Then you apply from above a piece of iron attached with some screws with countersunk heads, of a length equal to the width of the latter. And you make it overlap by about 5 or 6 lines at both ends, to make two frets that bear on the wave pattern [channels], and you make a notch in the middle of this piece of the size of the iron for positioning the cutting iron of the tool, as you can see in Fig. 7.

This iron is held in place by a cheek [ fence] (whether of iron or copper, either is equal), that you hold in place by means of two square-headed screws, g–g, Figs. 2, 4 & 5, where the nut is placed in the thickness of the cross-piece of the frame. See Fig. 3 of Plate 314, where I showed the cross-section of the tool-stand, with the contact I, the iron L, and the exterior cheek [ fence] M, which comes down as low as possible, that is to say, just to the bottom of the part the most hollowed of the latter.

The bottom of contact I [Plate 314] should be the thinnest possible (without however being a sharp edge), so that it follows well all the contours of the wavy pattern N–O. You must take great care that the point of contact for the fret be in the same direction as the iron cutting edge [both bevels are in the same direction], as I noted in this figure, so that the movement of the tool (which is made in describing an arc, where the center is found at the end of the frame) be less noticeable. I have partially remedied this by lengthening the point of the center of movement as much as has been possible.

The weight of the tool-stand should be almost sufficient to make the cutting iron bite into the surface of the wood workpiece. However, one must always put a spring there, both for augmenting the weight of the tool, supposing that it be necessary, and preventing it [ from] jumping around.

This spring h–i, Fig. 2, does not bear immediately on the tool-stand, but on a lever where its arms are loosely attached to the uprights of the movable frame of the box at m, Figs. 2 & 5.

The other end bears on the cross-piece of the tool-stand at n, which augments at the same time the strength and the elasticity of the spring, where the upper part is held below the small shelf O, Fig. 2, with screw P, where the nut is placed in the top of the frame Q. This screw serves, as I already said, to increase or diminish the pressure of the spring. The small shelf O through which passes the lower end of the screw, serves nothing but to hold it in place, and to press the heel o of the spring. As this small shelf is movable, you hold it from the opposite side of the screw with two pins, which you place across the uprights of the frame, as indicated by points p–p.

I made the head of the screw P in the form of a screw-eye, so that one cannot tighten it or loosen it by simply touching it, and so that you have need for a little pry bar or crank handle to do it. Those who approach the machine while it is adjusted cannot disturb anything there by simply touching it.

It is for this same reason that I prefer the screws with squared heads for closing the cheek [ fence] of the tool-stand, because a wrench is necessary to move these sorts of screws. You can eliminate their access from everyone’s hands, and consequently prevent anyone from changing anything on the tool.

As to the manner of using this machine, it is very simple. You begin by planing some wooden strips to the thickness of the profile that you have chosen, and the projection of the waves. This being done, you put in the tool-stand a smooth iron, which you adjust to the height equal to the projection of the moulding. You hold the strip on the platform, by means of little iron points placed [on the latter by equal distances from each other], and you make the machine move by turning the crank handle, which advances the platform forward. Consequently, the strip that is attached to the platform, after having passed many times under the smooth iron, is found to be wavy on its surface.

When the strip is thus finished, you remove the smooth iron, and you substitute the one that is shaped with a profile, and you begin the operation again, just until the iron is not cutting the wood any more, and consequently the moulding is perfectly finished.

You must take great care before running the moulding to verify that the wooden strip is placed truly parallel, which you know by making it pass the entire length under the blade that you hold elevated above. You should secure the strip on the platform only after having taken this precaution. You must also note that the pins that you place in the platform to hold the mouldings are positioned in the middle of their width [thickness of the moulding stock], and that they do not project enough to be able to meet the blade and cause any breakage, which you must take great care to avoid.

The blade of the waving tool is always placed perpendicular to the workpiece, which makes it scrape enough to cut, which cannot be otherwise, given that if you slant it in the normal way with moulding planes, it would scratch/drag on the wood as it comes against the grain, which happens at each undulation. What’s more, the blade thus slanted will no more be found in the same direction in all parts, which you must avoid as much as possible.

Since you can make many different blades, you must pay attention that they be all the same width, so that they completely fill the notch made in the piece which makes the cuts. You must also pay attention that they are all the same thickness, and that this thickness be considerable, to better resist the force of the wood in passing below.

The handle with which you move the waving tool can be placed to the right of the machine, as in Fig. 6, or to the left, as in Figs. 1 & 4, which makes no difference.

Each of these ways placing the handle has its advantages and disadvantages. If you place it to the right, which is the most natural way (since you made some effort pushing it), you cannot see the work well, behind which you position yourself. If on the contrary you place it to the left, you see the work clearly, but you are required to turn the handle in reverse. That is why, in order to eliminate these two inconveniences, I believe it would be better to position the two ends of the axis so that each one can receive a handle, like in Fig. 5, such that you can use it as you judge appropriate, whether on the right or on the left, or even from both sides at the same time.

3 It has not been possible to find a surviving wave-cutting Tool to make a good description of it. I have had only two iron blades, sold with other scrap metal which have nevertheless been very useful for fixing certain sizes, that I could not have known except for the description that Mr. Felibien made of this tool, which is otherwise very succinct, but imprecise, such that it could serve only to give me an idea of this machine, which I have then arranged in such a manner that it appeared to me the most likely. It has been greatly wished that those who have described this Machine in the Encyclopedia [of Diderot] had done something other than copy Mr. Felibien, instead of adding to the obscurity and inexactitude, as they have done. It would have been very useful to the public, and in particular to cabinetmakers, for whom they would have saved, or better said, presented one of their principal tools.

Meet the Author: Michele Pietryka-Pagán                                                    

Posted on by Kara Uhl

Editor’s Note: Michele Pietryka-Pagán is the French-to-English translator on the three-person team dedicated to bringing André-Jacob Roubo’s work to life. We have Michele, along with Don Williams and Philippe LaFargue, to thank for “With All Precision Possible: Roubo on Furniture” and “To Make as Perfectly as Possible: Roubo on Marquetry.”

These volumes are no longer in stock as we’re making room for new deluxe editions of each. The deluxe edition of “With All Precision Possible” will be for sale later this month and we plan to offer a deluxe edition of “To Make as Perfectly as Possible” soon.

Michele and Philippe have also completed the translations of more volumes of Roubo focusing on interior carpentry, garden carpentry and carriages. (You can read more about that on Don’s blog, here.)

Michele Pietryka-Pagán grew up in Vermont, the eldest of six children, born to native Vermonters.

“My parents were children of the depression, and so we grew up with a heavy dose of ‘Use it up, wear it out, make it do or do without,’” Michele says. “That’s a common Vermont philosophy. My parents were also educated, and they wanted all of us to be educated, too. There was always a subtext of do-it-yourself, and that included putting yourself through college, so we did.”

Michele’s dad was a mechanical engineer who liked to, and knew how to, fix most anything. In the early 1960s, Michele’s parents bought a 19th-century  house in Bennington, Vermont. It had no kitchen cabinets, so Michele’s dad drove to the lumberyard, bought lumber and taught himself how to make the base and upper cabinets. It was her first exposure to home renovation, helping her mom to wallpaper the old, horsehair plaster walls.

Michele’s mom was a teacher who stayed home to care for the family until Michele’s senior year of high school. When Michele was young, her mom taught her hand skills – sewing, embroidery and knitting.

“She taught me everything she could so maybe I would survive the next depression,” she says. “One of my earliest memories is getting a set of seven tea towels for Christmas one year, one for every day of the week, with a different motif to embroider on each one.”

As she grew up, Michele bought more and more complex patterns. By high school, she was able to make her own prom dress, and by the time she graduated from college, she made a friend’s wedding dress.

While Michele was growing up, her dad changed careers and became a high school industrial arts teacher and, later, a mechanical engineering professor at Vermont Technical College. Because of her dad’s position, Michele’s tuition at the University of Vermont was free. There she learned most of what she knows about textile science, in addition to perfecting her hand skills with fabric – turning 2D pieces of fabric into 3D garments.

“I had a real classical training in dressmaking and design,” she says.

Some of it she already knew – how to sew a straight seam and put in a zipper. She had whole semesters where she just studied tailoring or fabric draping. She spent two semesters studying textile science. She also learned how to make her own mannequin, which later came in handy when making mannequins for garments and costumes in museum exhibits. She graduated in 1973.

“Then, of course, the Bicentennial happened in 1976,” she says. “If you talk to a lot of museum folks of my generation today, we all got bitten really hard by the historic preservation movements that came about when the bicentennial celebration happened.”

Michele, John and Gracie, their terrier mix, today

In the mid-1980s, Michele earned her master’s degree in textile studies at the University of Connecticut. It was during this time that she met her husband, John Pagán, who was in the U.S. Naval Submarine Force. They married in 1984. Together they traveled up and down the East Coast, following John’s assignments both at sea, and at the Pentagon. In 1987 they moved to Washington, D.C., where they lived on and off for nearly 30 years.

Conservation & translation

While living in Washington, Michele and her husband hosted international students, researchers and writers. They had a particularly good experience with a young man from France, who stayed for a couple of summers. In part because of this, Michele began studying French in the evenings through a program with the U.S. Department of Agriculture.

Michele also studied textile conservation at the Smithsonian’s Museum Conservation Institute, and became one of their Research Associates. While there she helped a senior textile conservator with a small French translation project.

Don Williams, who was the senior furniture conservator at the Smithsonian’s Museum Conservation Institute, heard about her translation work. He had a couple of books about French carpentry written by an 18th-century woodworker named André Roubo that he wanted translated. He asked Michele if she’d be interested in volunteering.

“I naively said, ‘Sure! Why not?’,” she says. “For the next seven years, while most people were watching some sitcom on TV at night, I was sitting at my big dining room table surrounded by six or seven French-English dictionaries, a couple of them dating back to the 18th century,” she says.

Language changes over time. When Michele would get stuck trying to find an appropriate word in a 20th-century dictionary, she moved on to her 19th-century dictionaries, and then to her 18th-century dictionaries. She worked one sentence at a time: one paragraph, no matter how long it was, was always one sentence.

First, Michele would read the paragraph-long sentence and circle all the words she didn’t know. In the beginning, this ended up being about every third word because she’s not a woodworker nor a native French speaker.

“… So then I had to translate word by word, each word that I didn’t know,” she says. “I had to find the word in one of those dictionaries and then break up the paragraph into smaller sentences. That alone was a challenge because if I chopped up a paragraph into, say, three sentences, then I had to go back, after the translation, and see if the whole thing made any sense.”

Michele and Don, 2018.

With time, the work became much faster. Today, Michele can look at one of Roubo’s French paragraphs and typically type it into English, having to look up hardly anything.

“That’s how much I have improved over 18 years of doing this,” she says. “And now, of course, if there is a word that I don’t know, I just use Reverso. And the beauty of it is that it not only tells you what the word is, but it also puts it into context for you. So that’s really been lovely. But my French conversation still stinks!”

Michele and Philippe LaFargue, a native French speaker, work on the translations.

“My translated text then goes to Don, who adds contemporary information for today’s woodworkers,” she says. “Roubo was a master woodworker at the end of the 18th century in France. Some of that information translates to today, but not all of it. Don’s image of the project from the very beginning was to make this information as tangible and accessible as possible. Then the work goes to Philippe, who makes sure that my translation works with what Don is trying to say, for American woodworkers.”

Still learning – & teaching

Michel and John’s house in Dorset, Vermont.

John, Michele’s husband, retired in 2015. In May 2016, they bought and began restoring an 1825 farmhouse in Dorset, Vermont. It’s something Michele and John are well-accustomed to, having bought and restored four old townhouses while living on Capitol Hill. “All the homes on Capitol Hill are old, and they ALL needed a new furnace!”

Michele and a neighbor recently spent about five years researching 42 homes in their little village of East Dorset. In July 2025, the Vermont Division for Historic Preservation met and reviewed their application, calling for East Dorset to be named a historic district. It was approved.

“It was a long haul but definitely worth it,” she says.

More recently, Michele has been active with the Vermont Sampler Initiative.

“It’s a national search for 19th-century schoolgirl needlework samplers,” she says. “We’re trying to find them, document them, photograph them, analyze them and research the genealogy of all the girls who made these samplers and put them online. It gives me goosebumps. Nobody has ever done this before! Here we are, in the 21st century, and nobody has ever looked at a schoolgirl sampler, read her name, her birthdate, maybe her town all stitched there, and asked questions. Who were her parents? What kind of people were they? Did they have any role in making this country that we call the USA?”

Since November 2022, Michele and her team have found and documented more than 770 Vermont samplers. In 2025, in cooperation with the Vermont 250th Commemoration of the start of the American Revolution, Michele is coordinating a driving tour of 20 locations all over Vermont where visitors can stop and see exhibits of 19th-century schoolgirl samplers that all tie back to the American Revolution in some way.

In addition to research, for the past five summers Michele also served as a presenter at the Bennington Museum’s Summer Teachers Institute. There she teaches teachers seeking additional accreditation about how to use museum artifacts in their lesson plans.

“There’s nothing more gratifying than having an audience full of teachers, because when you’re teaching teachers, they are absorbing every single word you say,” she says.

Whether it’s translating, researching or teaching, Michele is all in. Case in point: She tore up part of her own meadow and planted flax, wanting to know more about how our ancestors planted, harvested, spun and wove it into linen. She brought the flax into her classes, along with different kinds of fiber for the teachers to observe under a microscope.

“This gives my life new meaning,” she says. “It’s a new chapter. I’m still really happy to be associated with Don and the Roubo project. What’s really special about working with Don is that he has so much respect for women: His wife and two daughters have raised him right!  Occasionally, Philippe will call from France – we have never actually met – but I can tell he’s a really great guy, too. So, this has been a truly wonderful project to be part of.”

The Roubo project is also giving back.

“My husband and I decided to use some of the royalty monies from the sales of the Roubo books to start an endowment at the Bennington Museum,” Michele says. “The endowment pays for one high school student per year to spend the summer working with the staff at the museum, for about eight weeks. We are into the fifth year of summer interns whom we have funded, and all we ask is that the student write us a little synopsis of what they did at the museum all summer. Since we don’t have children, this is our part of ‘touching the future,’ as Sally Ride, the astronaut, said.”

Despite living different lives, there’s commonality in Michele and Don’s work. In working on the translations, Michele says she was able to help Don better convey the antique processes and mindset for creating wooden furniture.

“Don and I are both conservators,” she says. “We both believe in historic preservation. We both believe in transmitting our cultural heritage from the past and making it accessible to today’s students. That’s why I enjoy making textile history from the past accessible to today’s teachers and their students. Don and I did the same thing in just two different specialties. We’re both educators. We’re both passing on information from the past to today’s and tomorrow’s students, teachers and historic preservationists.”

— Kara Gebhart Uhl

New Deluxe Edition of ‘Roubo on Furniture’

Posted on by fitz

Ever since we sold out of the original 12″ x 17″ deluxe editions of “With All Precision Possible: Roubo on Furniture,” Chris has been itching to do another luxurious book…if perhaps not _quite_ so extravagant (the almost-full-folio size of that edition was crazy expensive). Instead, we’re working on a new quarto-size edition (9″ x 12″), with gorgeous acid-free #100-pound interior paper, and plates printed in their full sepia glory. The signatures will, of course, be Smyth-sewn, and the end papers will be printed with images of a few of the plates.

The cloth cover for the boards will be Verona Blue Jay (a vibrant cobalt blue) with a gold foil stamp; the headbands will be blue and gold.

Then, we’re wrapping the book in a gloss laminate dust jacket (shown above)

While at $125 this new edition of “With All Precision Possible: Roubo on Furniture” won’t be inexpensive, I think when you see it you’ll agree it’s worth the price. It’s going to be resplendent – and built to stand the test of time, shop and kids (as well as look great on your bookshelf).

The new deluxe edition is at the printer now; it is due in our warehouse in March.

– Fitz

Clockwise from left: Proof copies of the dust jacket, cover die stamp and end papers (atop the interior pages).

p.s. Translating Roubo is the most expensive and sprawling project we’ve been involved in – and it is ongoing. Don Williams and his team are working now on the remainder of the volumes. If this deluxe edition of “With All Precision Possible” is well-received, we’ll offer suitably fabulous editions of the forthcoming volumes as well, and of “To Make as Perfectly as Possible: Roubo on Marquetry.”

Roubo on How to Saw Veneer

Posted on by fitz
Plate 278. The Way to Split Veneer Wood, and Its Explanation

The following is excerpted from “To Make as Perfectly as Possible: Roubo on Marquetry,” translated by By Donald C. Williams, Michele Pietryka-Pagán & Philippe Lafargue. It is the first English-language translation of the most important woodworking book of the 18th century.

While the title of this work implies that it is about marquetry alone, that is not the case. “To Make as Perfectly as Possible” covers a wide range of topics of interest to woodworkers who are interested in hand-tool woodworking or history.

In addition to veneer and marquetry, this volume contains sections on grinding, sharpening, staining, finishing, wood selection, a German workbench, clock-case construction, engraving and casting brasses.

But most of all, “To Make as Perfectly as Possible” provides a window into the woodworking world of the 18th century, a world that is both strangely familiar and foreign.

Roubo laments the decline of the craft in the 18th century. He decries the secrecy many masters employed to protect craft knowledge. He bemoans the cheapening of both goods and the taste of customers.

And he speaks to the reader as a woodworker who is talking to a fellow woodworker. Unlike many chroniclers of his time, Roubo was a journeyman joiner (later a master) who interviewed his fellow tradesmen to produce this stunning work. He engraved many of the plates himself. And he produced this work after many years of study.

As the wood that one uses for cabinetmaking is for the most part very expensive, because it costs roughly 10 sols up to 30 sols, and sometimes even one crown per pound, according to the different types of wood, we have great interest in using these woods sparingly; that is why instead of making furniture or other pieces of cabinetry in solid wood, we have tried to execute splitting [sawing] wood into laminates, or very thin sheets, that one applies on the furniture cases made of ordinary wood.

It is not the carpenter-cabinetmakers who split [saw] their wood, but the workers [sawyers] who do only this work, and who saw not only for the cabinetmakers, but also for the musical instrument makers, and generally all those who use thin wood. These workers or sawyers are paid by the pound, that is to say, according to the weight of the piece of wood that they use, including the waste-wood and sawdust, rendering the wood close to two-thirds more expensive, which makes a piece made in this manner very important.

Veneer wood is split [sawn] at about a thickness of 1 line at most [1/12″ to 1/14″]; when one wants to spare it, one makes from 10 to 11 leaves from a thumb-thickness [inch], which is worthless because even before the veneer is polished, it has left only a half-blade of thickness [1/24″ to 1/32″], which is then reduced almost to nothing when the piece is finished; it is absolutely necessary to avoid making veneers this thin, although that is used a lot at the present. When one wants to cut up a piece of wood to make a veneer, one begins by choosing the side of the log that allows for the easiest sawing, the goal being to orient the wood for the best advantage, and to yield the largest sheets of the veneer; then one puts the piece of wood in the vise, and with a standing saw [a saw to be used while standing, and a vise designed to facilitate that action], one saws it to a thickness that one judges appropriate (which I am going to explain, after having provided the description of the bench or vise with a standing saw, and of the saw appropriate for this task).

The saw appropriate for cutting wood from India, which we name also the saw with vise, Figs. 1 and 2 [to increase or decrease the tension on the blade] is a little bit similar to the saw for cutting used by the woodworking builders [often known in the modern era as a frame saw]. It is composed of two verticals and of two crosswise or crossbeam elements, of which the ends project out and are round ed, so that the two sawyers can hold the saw easily. The middle of these crosspieces is convex on the outside, in order to give them more strength, and that they not bend while one increases the tension on the saw blade.

The inside [interior] of the vise saw is from 15 to 18 thumbs [inches] wide [or approximately 9 inches on either side of the blade], is about 3 feet long, as measured from within the crosspieces or support piece. The blade of the saw has a 4–thumbs [inches] depth, at least, and is held at each end by a frame of iron, through which passes the crosspieces of the saw, or, better said, of its chassis. These frames of iron, represented by Figs. 4, 5, 8 and 9, are made of iron plate, and the largest possible, so that the saw cannot turn easily, and one tightens a nut to that above, for putting there a screw a b, Figs. 4 and 5, which serves to control the tension of the saw blade.

On the outside of the cross-members one insets a steel contact plate attached with some screws, which prevents the pressure of the screw of the frame to not ruin anything nor to make any holes. See Fig. 3.

The blade of the saw, as I just said, is 4 thumbs [inches] size at least, tapering barely toward the back [away from the teeth]. We do not put a set on these sorts of saws, because that would eat up the wood excessively with an unnecessarily wide kerf, and one takes great care that the teeth be perfectly straight on the horizontal, and that their teeth be also perfectly equal in height, so that they grab all equally, and that they do not chatter, resulting in uneven thickness of the wood, which is also to be feared, which ruins so many sheets of veneer. The teeth of these saws should be spaced equally, about 5 to 6 lines from one tooth to the next one at least, and should be positioned in such a way that the bottom [what we now call the tip] of each tooth is level with one another, because being so arranged, they are less subject to become dull, which would happen unfailingly if they were made ordinarily, as is seen that almost all wood from India is hard, and consequently causes more resistance to the teeth of the saw. See Figs. 6 and 7, which represent one part of the saw blade viewed from the front and side, half-size.

The standing saw vise, represented in Fig. 11, is one type of small bench, about 3 to 3.5 feet long, by 2 feet high, at the base of which one puts the vise, which serves to hold in place the piece that one wishes to saw.

In order for this vise to be solid [a stout twin-screw face vise], it is good that the brace [the jaw] A, Fig. 11, have about 6 thumbs [inches] thickness, as well as the top of the bench, in which the screws enter,
which to be good, should have at least 2.5 to 3 thumbs [inches] in thickness, and the threads be long enough so that when there is a piece of wood8 to 10 thumbs [inches] thickness placed in the vise, there remains at least enough length of the screw in the bench, as observed in this figure. As this bench is very short, and is subject to vibration by the movement of the saw, one loads stones on the bottom shelf to make it more solid; but I believe it would be better to make the legs of the bench long enough to be anchored to the floor of the shop, then one makes a hole in front of the bench to set in the piece of wood to be sawn in order to not extend upwards more than 3 feet above the top of the vise, locating it thus both for the comfort of the sawyers and for maximizing the yield of the piece being sawn. Not all the standing saw vises are part of an overall bench, such as the one represented here, in Figs. 10 and 11; this is why ordinary vises attached to a little bench are less solid than making them as I propose here.

When one wishes to saw with the vise, one begins by placing the piece to saw in the vise, of which the screws tighten with an iron lever, that one removes after being worked, so that it is not in the way; then, with an ordinary saw, one begins to mark all the lines to be sawn on the end of the workpiece, just up to 2 to 3 lines deep [3/16″], then one uses the frame saw, Fig. 1, which is guided horizontally by two men, observing the advantageous slight incline on the side of the tooth rake, and of the lifting up of the blade while pulling back, so as to relieve it, and that it not bind in the wood, or at least that the sawdust does not obstruct it. See Figs. 10 and 11, which represents a vise press upright, viewed in perspective, with the sawyers located as they should be.

When one saws with a vise, one begins with the outside edge of the log, so that the first sheets sawn bend away from the log and facilitate the passage of the saw, which could not be the case if one sawed in the middle; as one does when one saws large pieces of wood being used by carpenters or by ordinary woodworkers, given that the frame saw blade is very thin, and that it has no set. Sawyers at a vise do not lay out or mark a line on the side of the piece that they wish to saw; but after having begun on the end with an ordinary saw, they continue the rest by eye, which they do very well, for the most part; they are very sure to saw their veneers not only very straight, but still perfectly of equal thickness, as well. See Fig. 11, which represents the cut of the bench or upright vise saw, and a piece of wood sawn into sheets just up to the middle.

To finish what this looks like at the cutting of wood appropriate to the cabinetmaker I have represented in Fig. 12, a saw named the carving saw, which serves to cut up not only hard wood, whether wood with the grain or cross-grain, or standing wood, but also coral, ivory and mother-of-pearl. The framework of these sorts of saws is all iron, of which the upper branch is widened on the outside, so that one can adapt the blade and set it as one judges appropriate, which is done in the following manner.

After having pierced a hole in the blade of the saw, b, corresponding with that of the lower arm of the frame of the saw, you put this one [arm], and the one that is opposite, in a vise or other thing capable of bending them [squeezing them together], in a manner that they tend to meet one against the other, and tightens them as much as is judged appropriate, to give the saw all the tension necessary; then the blade of the saw, being stopped at point b, one makes it enter in the upper arm of the frame, and one traces the place for the hole at point a, which one pierces to place there a peg; this being done, one again bends the arms of the frame, just until it gives liberty to pass the peg below, and which serves to hold the peg in place, as one can see in this figure.

The blades for these sorts of saws are very thin, and one does not give them a set, so they have a very narrow kerf and lose less material, and they pass easily; one thins them on the back [away from the teeth], which one does with a file that one passes down the length just until they are thinned enough as one judges appropriate; then one rubs them with sand to remove the unevenness that the filing could have made; this operation is called “demaigrir” [thinning], a worker’s term.

‘Tables in General & the Different Types’

Posted on by fitz

The following is excerpted from “With All the Precision Possible: Roubo on Furniture,” by André-Jacob Roubo, translated by Donald C. Williams, Michele Pietryka-Pagán & Philippe Lafargue.

Representing a decade of work by an international team, this book is the first English translation of the 18th-century masterpiece: “l’art du Menuisier” by André-Jacob Roubo. This, our second volume, covers Roubo’s writing on woodworking tools, the workshop, joinery and building furniture.

In addition to the translated text and images from the original, “With All the Precision Possible: Roubo on Furniture” also includes five contemporary essays on Roubo’s writing by craftsmen Christopher Schwarz, Don Williams, Michael Mascelli, Philippe Lafargue and Jonathan Thornton.

“Roubo on Furniture” is filled with insights into working wood and building furniture that are difficult or impossible to find in both old and modern woodworking books. Unlike many woodworking writers of the 18th century Roubo was a traditionally trained and practicing joiner. He interviewed fellow craftsmen from other trades to gain a deep and nuanced view of their practices. He learned to draw, so almost all of the illustrations in this book came from his hand.

After Beds and Seats, Tables are the most ancient pieces of furniture, or at least the most useful. The number of Tables currently is considerable. There are Tables for the kitchen, Tables for eating, game Tables, Tables for writing, dressing Tables, night Tables, Bed stands, etc. which are composed of a top and of several legs, and which do not differ except in their size and the shape of their top or in their legs. That is why, before entering into any detail on the subject of these different Tables (which you can consider as being three different types, namely dining Tables, game Tables and writing Tables), I am going to address the different legs of these same Tables in general, so as not to repeat it when I come to their particular detail later.

The legs of Tables are of two types, namely those which are immobile, as in Figs. 1 & 2, and those that fold, like those in 3, 4 & 5. In the first case, the bases are composed of four uprights, of four cross-pieces [aprons] at the top and of four others [stretchers] lower down, as in Fig. 1, which is the most solid way to make legs for Tables. Sometimes you put there only two cross-pieces [stretchers] at the ends with a brace in the middle. Or even two stretchers at the ends and one on its rear side, such that there is one side free for providing access for the user’s legs, which is necessary for writing Tables and dressing Tables.

These sorts of legs are, as you can see, very solid. However, we often prefer those of the serpentine leg, represented in Fig. 2, which, although less solid than the first, have the advantage of being less heavily decorated and not to bother in any way those who are seated around it, whether for playing or writing; [this] is to be highly considered, especially when there is no need for much strength or they are not subject to changing place frequently. Because in the latter case, you would need legs like in Fig. 1, unless the Tables being very light, like little writing Tables, game Tables and others of this type.

The legs of folding Tables are of two sorts: namely those in x, whether in elevation, as in Fig. 3; whether in x in plan, like Fig. 4; and those of a folding frame, like Fig. 5.

In the first case, Fig. 3, these feet are composed of two frames assembled with a cap at the end, which would be about 2–and-a-half feet in length each, with a width equal to that of the table, less 2 to 3 thumbs, according to the greater or less width of the latter.

The width of the legs that I speak of should not be taken from outside of the uprights, but from the ends of the cross-pieces at the cap, at the end of one of which you make some dowels, a, b, which move in the hinges attached to the top of the table, which I will speak of next.

The frame that holds the dowels should be the narrowest so that in rounding off the latter, some shoulder remains in the mortise that receives the upright. One could not do this to the other frame, unless by moving it back a lot and consequently to reduce/narrow the frame on the interior as much as the exterior, and diminish at the same time the seating of the leg, of which it never has too much in the case being questioned here.

The two frames of the legs of the Tables that I just described are held together in the middle of their length by an iron pin which enters into each of the uprights at about the middle of their width, which requires that one not peg the wider frame after having placed the [iron] pins, which at 2 to 3 lines in diameter [will] suffice for giving all the firmness suitable.

I just said that you place the pins in the middle of the length of the frame. However, if you wish to give more spread to the leg, you could place them a little bit higher, which you do with no other change than to augment the length of the uprights a little. That is why when you make these sorts of legs, you [will] do very well to draw them in elevation in order to have the exact length of the uprights, the place of the hinges, Fig. 8, [& some racks] Fig. 9, which are attached under the table, as you can see in Fig. 7, which represents the leg folded under the table AB, which extends by about 5 to 6 thumbs at the end, at least normally.

The hinges, Fig. 8, (which Joiners improperly call pins), are made of beech, about a thumb’s thickness and from 5 to 6 thumbs in length; in the middle of which, and about 6 lines from the bottom, that is to say, from the straight edge, you drill a round hole a of about a thumb in diameter into which enters the pins of the crossbar of the leg. These hinges are attached under the table with some nails, which is the most normal way. However, it is much better to make them enter into a notch [that is] the thickness of their cheek in the underside of the table, [as] indicated by line b–c. This is not only more solid, but makes the top of the cross-piece of the frame support equally along the entire width of the table.

The racks represented in Fig. 9 are made of the same wood and of the same thickness as the hinges and are attached under the table with some nails, as with the latter. One is required to make some notches c–d, Fig. 3, into which enter the cheek of the rack. It would be good to make this enter into the notch in the tabletop of this same thickness, so that it attaches more firmly, and you are not obliged to make a notch in the crossbar of the leg frame, which conserves all its strength. However, as these notches serve to hold the leg in place, or at least to prevent it from varying, you can let the rack project by about 2 lines from the edge of the tabletop, [as] indicated by line d–e, which removes less of the strength of the crossbar and is sufficient to prevent the foot from varying. The racks normally have two notches, f & g, [NB: these elements are not present in the plate] to allow you to raise and lower the table as you judge appropriately, which you do by moving the cross-piece of the frame from one notch to the other, noting that the notch farther away is positioned such that the leg be at its normal height, which is for all dining Tables (where these legs are normally used) from 25 to 26 thumbs on the bottom of the table.

These sorts of legs are not used except for dining Tables of average size and are otherwise inconvenient and less solid, [with] their legs interfering with those who are placed around it. That is why one should prefer them in [an] X on the plan represented [as] in Fig. 4, which are the most solid, less awkward and less complicated, although constructed rather in the same manner, as you can see in this figure, of which inspection alone is sufficient.

The top of the uprights of this sort of leg should project past the crossbar by about 9 lines or 1 thumb, which is necessary to preserve the shoulder. This projection is necessary for entering into the notches that you put on the underside of the tabletop, so as to hold the leg in place. Sometimes you do not put a notch on the underside of the table, but you use some cleats into which enter the end of the uprights.

These sorts of table legs are very convenient for a dining table of a certain size because they do not interfere in any way with those who are seated around it and they take little space when folded, as you can see in Fig. 6, which represents this leg completely folded and viewed from above. This is preferred to all others for dining Tables of average size. What’s more, these feet are normally of a very simple construction and are consequently less costly, which is one more reason to prefer them.

Other folding legs are required, much more complicated than those that I just spoke of, but which are at the same time more solid. The leg represented in Fig. 5 is composed of six frame sections, or better said, of four – two on the side and two at the ends – which each break into two parts in the middle of their width. These frames are closed by pinned hinges on the inside on the frame and in the middle of the two outside. When you wish to fold them, you make them move toward the inside of each side, which make these fold thus, hardly 5 thumbs of thickness, as you can see in Fig. 10, which represents this folded foot held in place by a hook of iron ab, which you can remove when you wish to open it.

When this foot is open, you hold it in place with a flat iron hook, c–d, Fig. 5, which is placed behind the break in the middle. We also have the custom of placing there a movable brace, which is nothing other than a board of a length equal to that of the leg, and large enough so that it can hold the two uprights in the middle which enter into the notch in the ends of this spacer, which sometimes you make of assembled braces to make it lighter, like the campaign Tables represented in Fig. 6, Plate 251.

These sorts of legs are very solid and greatly in use for dining Tables of a medium size, of which the large projection over the leg [is] made such that it cannot harm those seated around the table.

There are serpentine legs, like Fig. 2, which fold in the same way as those that I just spoke about; that is to say, they fold in the middle of the cross-pieces of the ends, which instead of a tenon, have only a tongue-ending [like a wooden key], which enters into the serpentine leg on which they are tightened.

We also make a tongue at the fold in the middle of these crossbars and you note to make there a shoulder above and below so that they are more solid. These types of legs are frequently used, however they are less solid, no matter the care that you take when closing them. One should prefer the legs with a folding frame, Fig. 5, for large Tables or even that represented in Fig. 4 for small ones.

The size of the leg of Tables’ frames varies from 3 feet in length by 2–feet-3–thumbs in width up to 6 feet by 4–feet-6 thumbs by a height of 25 to 26 thumbs, which is general for all dining Tables. This cannot be otherwise since this height is determined by that of the person seated there, below the elbows of which the top of the Tables must be flush, at least for those of a normal size, which ordinarily is 26 to 27 thumbs in height from the top of the Tables. As to the size of the wood of these legs, 10 lines or 1 thumb thickness suffices, by one–thumb-and-a-half or 2 thumbs, and sometimes 2–thumbs-and-a-half for the width of the uprights, according to the size of the legs. Their crossbars should [proportionally] be a bit larger than the uprights, especially those that meet at the end of the latter so as to conserve the strength of the assemblage.

There you have in general the detail of all the different types of table legs in use for both dining Tables and Tables for games and writing, which, with some small changes, are always of the same form.