The following is excerpted from “With All the Precision Possible: Roubo on Furniture Making,” by André-Jacob Roubo, translated by By Donald C. Williams, Michele Pietryka-Pagán & Philippe Lafargue.
For the fashioning of raw timbers, we understand the manner of splitting [hewing or sawing] and squaring them, which is done in different ways, according to the nature, the quality and the thickness of the wood. They are sawn by the mills, or even by hand, by workers called long sawyers or simply sawyers. I will not speak here of the harvesting of the woods in the forests. I will be content to say only that they are sawn and cut in sizes and lengths relative to our different needs, and that wood thus prepared is called wood samples.
They are found abundantly in all types and all qualities possible in the inventories of the wood Merchants, which they ordinarily cut for themselves and for their clients, and are transported to their storage lots or shops in Paris. [Suppose you go to a lumberyard and in the lumber section you will find 2×4, 2×8, 2×6 etc. This is what is called bois díechantillons. In this case it is oak, pine, walnut etc. of various types but standard dimensions.]
Cut or squared-up woods take different names according to their sizes, and according to the place which they occupy in the body of the tree: We call them dosses [slabs], countre dosses, swinging doors, framework, chevrons and finally planks and battens.
Slabs are the first cuts removed from a log in order to square it up, after having removed the bark, like those on sides g–g, Fig. 5 and 6.
When the diameter of the tree is considerable, and you fear that the first cut slab will become too thick, you make a double-cut, which is called a contre-dosse; that is to say, that it is between the first cut dosse [slab] and cut line and the heartwood [wood between the pith and the sapwood] like those on sides h–h, Fig. 6. When the wood is beautiful, the contre-dosse are very soft, being very close to the edges of the tree [closer to the sapwood layer]. They do not have any sapwood except at their extremities, instead of the first cuts, dosse, that have sapwood all over their convex areas. The thickness of the contre-dosse is not precise. It varies from 2 to 4 thumbs. After a log is thus squared, you saw it into thin panels or lumber planks, according to its hardness or softness. You can then judge whether it is appropriate for one or the other. [That is,] [u]nless one cuts planks from the entire width of the tree, especially with soft wood, which I will speak of later.
The double doors of main entrances are ordinarily 12, 15 or even 18 feet long, by 1 foot or 15 thumbs wide, for the longest, and by 4–5 thumbs thick. They are almost always of a hard-quality wood. They should have neither knots nor splits. This may be found in the woods of Vosge, but they are very expensive and very rare.
Lumber for framing is of a length from 6, 9, 12 and 15 feet. They are 6 thumbs wide by 3 thumbs thick.
Rafters have the same length as framing, and sometimes more, by 3 to 4 thumbs squared, that is to say, they have as much thickness as their width.
Planks have 6, 9, 12, 15 and even 18 feet of length, by 1 thumb 15 lines, 1–thumb-and-a-half, one thumb-9 lines, and 2 thumbs thickness.
There are also planks of 7 feet length, but they are rarer than the others, and are difficult to find in all thicknesses.
With regards to the width of planks of French wood, they vary from 9 thumbs up to 1 foot. However, those of 1–and-one-half thumbs to 2 thumbs are ordinarily a foot wide, and those below this thickness from 9 up to 10 or 11 thumbs at the most.
There is still another type of thin French oak wood, named Entrevoux, which only has but 9–10 lines of thickness, by 6, 7 or 9 feet in length, which is appropriate for making panels, provided that it is soft and beautiful.
For the wood from Vosge, there are all kinds of lengths and thickness of which I spoke above, except that there are none of 6–7 feet, or even less. There are also those of 3 thumbs thickness by 12 feet in length. With regards to its width, that is not fixed, because in all the different lengths and thicknesses of this wood, there are from 6–7 thumbs of width up to 18, 20 and even 26 to 30 thumbs [width]. That is why Merchants do not sell this wood by measurement [of the individual boards], as with the others, but by each row of the lumber stack, which is 4 feet in width.
To facilitate the understanding of widths and thickness of woods for joinery relative to their different lengths, I have attached a table [on the next page] where all the wood types are distinguished according to their length, width and thicknesses.
The wood from Holland is not included in the number of those I have mentioned here because it is only a thin wood, which is sold by the handful or even by the cord. These lengths are of 6, 7, 9 or 12 feet by a thickness of 6 or 9 lines.
The thickest of these woods is called three quarters, because it should have 9 lines thickness, although often it only has 7 or 8 at the most. (The thinner ones are called feuillet [leaf] and are only 4 to 5 lines thick, while it should be 6 lines thick.)
It is to be noted that French wood is always thicker than the wood from Vosge with each sample; that is to say, that the first always has 2–3 lines more than its thickness, such that the wood of 1 thumb sometimes has 14–15 lines [thickness]. On the contrary, the latter [wood] always has 1 line less than it should have, which is a shortcoming. Also it has the advantage of being straighter than the other, and has less waste.
For the battens made of oak, wood Merchants sell them only rarely. Joiners use wood from Holland for thin panels, and they even saw [resaw] them at their shops while on edge, to the thickness and of the quality that they judge to be appropriate.
Pine is not subject to the rules of thickness of which I just spoke, at least that type used in the woodworking of buildings.
That from Auvergne ordinarily is 12 feet in length by 14–15 lines in thickness. Its width varies from 10 to 14–15 thumbs.
That from Lorraine has only 11 feet in length at the most. Its thickness is the same as that from Auvergne, but the most ordinary thickness is from 10–12 lines. Its width varies thus, from that of the latter.
There are also the little leaves of pine from Lorraine, of the same length as the planks, which have from 6 up to 8 lines thickness.
Walnut and elm are not found cut into planks like the other woods. Whenever Joiners have enough money, they buy whole logs which they cut themselves, namely the elm, into slabs of 5 thumbs’ thickness, and the walnut into slabs of 3 thumbs [thickness]. They still saw black walnut to make the panels for tables of 4 lines thickness, which have a width of the whole log, which is sometimes 2–2.5 feet in width.
Beech is found cut into planks of 15–18 lines, and even 2 thumbs thickness by 7, 9 and 12 feet in length. They also sell slabs of this wood for making woodworking benches, tables for the kitchen, and butcher tables, tables that have a length from 7–12, and even 15 feet, by 18–30 thumbs in width, and 5–6 thumbs’ thickness.
Although the wood which one chooses has by itself all the required qualities, it is still necessary to watch out for its preservation. Because wood for joinery should not be used except very dry, it is of the final consequence to Joiners to always be well provisioned with wood of all types, which they keep and dry in their yard before using them.
They should also take care that their yard not be placed too low, nor planted with grass, because the falling and gathering of leaves will prevent the run off of water, which could ruin the wood pile coverings and also the base of a woodpile.
The terrain occupied by the woodpiles should be higher than the rest of the yard, so that water does not collect there. It must be well set up and leveled, after which you put on top some pieces of wood side A, which we call chantier [beam/timber spacers] which has a length the same as the width of the pile – ordinarily 4 feet, although sometimes they make them wider. You make them the greatest thickness possible, so that they make the pile taller with the most possible space between the boards.
You put the spacers distant from each other about 3 feet. Their topsides should be squared and straight, after which you pile the wood on top, after having taken the precaution of putting the worst planks on the lowest level to save the better woods from ground moisture. You make the piles in two ways, according to whether the wood is being dried or is already dry.
In the first case, you pile them up to see through, which is done in the following two ways:
The first is to place the planks side by side with a space [between them] about equal to two-thirds of their width, and to separate each row of planks by laths [stickers] f–f, which separates them, prevents them from touching and maintains them in a solid position on top of each other, such that one can stack up the piles up to 20–25 feet in height. (Fig. 1).
The second way to make see-through piles is to make them squarely; that is to say, to give them as much width as the planks are long. You first put a row of planks spaced equally, as in the first way, always such that the width of the row of planks and the additional space between them is equal to their length.
After this, you put on top some planks in another row, in the same order and at a right angle, which means that you have no need for stickers/spacers, and that the planks have more air between them. However, you should not leave them thus piled for a long time, for fear that the wood will rot where the pieces sit on top of each other. (Fig. 2.)
Rafters of 6–9 feet are piled in this fashion, without however being see-through.
The way to pile seasoned wood does not differ from the first of these two ways, except that the planks touch each other side by side, instead of being see-through. You separate each row with some spacers which you put at an equal distance to that of the chantiers which are three feet apart, so that the planks are always straight and do not warp. However, this last term signifies more of a hollowed-out plank along its width [cupping] than a warping [bowing].
The top of the pile is covered with planks positioned to overlap one on top of the other; one of the ends of which is positioned on another plank (side a, Fig. 4) which is called l’egout de la couverture [not quite like a gutter but more like a rain diverter] and which lies flat upon the pile. One should note, however, that it overhangs the front of the pile by 3 or 4 thumbs, and that it slopes a bit to the outside, in order to facilitate the runoff of any water. You raise it up a bit at the back to create this effect. The other end of the planks of the cover hold a piece of wood b, which is named chevet [or riser], which is positioned on edge on two pieces of wood c, in which a notch is placed, and which is stopped with wedges d, in order that it not turn. The chevet should be a foot and a half tall at least, so that the water accumulates less on the piles.
The middle of the covering should be supported by a piece of wood e, which is placed above the pile, and the two planks along the sides rs, rs should be wider by 3 or 4 thumbs than the two sides of the pile, so that the water does not fall back along its length.
When you wish to make the piles more than 4 feet thickness, you should take care to put the spacers in concert (that is, you place them such that the total length of two spacers is more than 4 feet, which is the length of the laths), overlapping such that it maintains the solidity of the pile. You will need to take care to keep everything straight and vertical in all directions, so as to avoid accidents that its fall might cause.
For thin wood, like the wood from Holland, the battens of oak and of pine, the custom is not to pile them in the open air in the middle of the yard, but to pile them under sheds and above the shop where the workers work, the reason being, they say, that they are preserved better. I believe, in spite of this practice, that they would be better in the yard, where they will receive the air from all sides, and where they will not be exposed to insects [powder-post beetles etc.].
As to their preservation, I believe they run no danger being out in the air. The piles of wood from Holland, which reside for a long time in the wood lots of the Port of Hopital and of Rapee without any damage, are surely guarantees of the truth of what I advance here.
What I am saying here is only general. I know perfectly well that all woodworkers cannot have
great wood yards nor large provisions of wood. But still, for reasons of economy, they should always do their best to be well prepared with samples, and to watch over their preservation as best as they can, so as not to be obliged to have to buy some from the Merchants. The wood that they sell is almost never dry, and the woodworker will pay dearly for what the wood Merchants have.
The more wood is hard, the more time it takes to dry. That is why one should not reasonably use wood that has not been cut at least 8 years in order to be able to do good work. It is not necessary, however, that it be too dry, especially for pieces of joinery, where the wood has no more sap and where the humidity is totally expunged: this cannot be appropriate. [Once the sap no longer is flowing from the lumber and the moisture has departed there is no need to season the lumber any further.]
This doesn’t help explain why lumber prices are so high . . .
I’m curious about the measurements used. Did Roubo use feet/inches? What preceded metric on the continent?
They used feet and inches. But the French inch is slightly different than the U.S. Customary Units we use today.
The French word “pouce” means both “inch” and “thumb”… like the word “pied” means both “foot” and, um, “foot.” Handy to have your measurements on you at all times.
But not all thumbs (or feet) are created equal. Didn’t Chris have a story somewhere about using a Danish ruler for a project that didn’t end well when he learned the Danish inch was a slightly different than the English inch? Or maybe that was me…
Anyway I still wonder what a ligne (line) is… maybe a 1/16th or so?
Zut alors,
David
A few paragraphs above the tables, Roubo says:
“The thickest of these woods is called three quarters, because it should have 9 lines thickness, although often it only has 7 or 8 at the most.”
By which I infer that a line is 1/12 a thumb. Coincidentally (or maybe not), that’s about the width of a rip kerf.
For the curious: a thumb appears to be 1 inch (I just found that out)
How thick is a line?
A few paragraphs above the tables, Roubo says:
“The thickest of these woods is called three quarters, because it should have 9 lines thickness, although often it only has 7 or 8 at the most.”
By which I infer that a line is 1/12 a thumb. Coincidentally (or maybe not), that’s about the width of a rip kerf.
In the “A key to the text” part of the book from which the above is an excerpt, it says that “[i]instead of converting all of Roubo’s measurements to U.S. Customary Units (or metric), we decided to use his original terms. As such, you will find the units of “thumbs” and “lines”. A thumb is just slightly more than our modern inch – 1.066″. The thumb is further divided into 12 “lines.” Each line is equivalent to .088″ today. The French foot is 12.792″.”
As Wikipedia is happy to tell us the English word inch comes from the Latin uncia, meaning “one-twelfth”, while in many other European languages the corresponding/similar unit of measurement has a name based on that language’s word for thumb.
Thus, in my mother tongue, Swedish, we say tum about the measurement, and call our most useful finger tumme.
To further muddy the waters/make things interesting for the nerdily inclined, an inch is not necessarily an inch. Even we leave local variation and the lack of precision in early standardization attempts (like Edward II official definition from 1324 of one inch being equal to three barleycorns, i.e. to “three grains of barley, dry and round, placed end to end, lengthwise”) to one side, the current standard international inch, where 1″ = 25.4 mm, was only formally adopted in 1959, although it had been preceded by the otherwise identical industrial inch, adopted by the UK in 1930 and the US in 1933, which in turn came out of the 1″ = 25.4 mm compromise between US and UK inches, settled upon by C.E. Johansson for his ‘Johansson’ gage blocks. Before that, from 1866 the US inch was 25.4000508 mm long, while from 1826 the Imperial UK inch was 25.399977 mm. Again, thanks, Wikipedia!
As we already now from the “A key to the text” quote above, the French pouce was different again, as were, I think, many (most?, all?) other similar, duodecimal systems of measurement.
For example, until Sweden went fully metric in 1889, we used up to three inch or inch-like systems concurrently: engelska tum (“English inches”, i.e. UK Imperial inches), primarily for lumber; verktum (“work inches”), which was our domestic duodecimal system, with 1 verktum = 24.74175 mm, laid down by royal decree in 1739, and decimaltum (“decimal inches”), in use between 1855 and 1859, where the same foot as in the verk/work system was divided into ten inches, rather than twelve (and all other units, smaller and larger, where also based on multiples of ten).
While I certainly was nor around when any of those systems were in use, and they had long been consigned to history and a few standing expressions when I first needed to measure something, I don’t think that the Swedish verk/work system, although duodecimal, ever used fractions in the way that the US and UK systems did and do. Instead, like those French pouces, it used lines (verklinje) for smaller measurements, and then kvarter (“quarter”), fot (“foot”), aln (“ell”) and famn (“fathom”), like so:
1 verklinje = 2.0618125 mm = 0.08117″
12 verklinjer = 1 verktum = 24.74175 mm = 0.97408″
6 verktum = 1 kvarter = 148.4505 mm = 5.84451″
2 kvarter = 1 fot = 296.901 mm = 11.68902″
2 fot = 1 aln = 0.593802 m ” = 23.37803″
3 alnar = 1 famn = 1.781406 m = 70.13409″
That said, I’m quite sure at least in formal use would have included expressions like en kvartstum (“a quarter of an inch”) and similar. In any case, the 1880s must have been a fun time to measure stuff in Sweden, when three different inches on their way out were cohabiting with the metric system on its way in …
Cheers,
Mattias
Meant to say “informal”, not “in formal” in that final paragraph; sorry!
When Roubo mentions Dutch wood I am quite sure that means wood that is sawn in The Netherlands, Along the river Zaan. It is known that we imported wood from the Voges to saw it and to sell it as Dutch wood (and imported by the French as Bois Holandais). The sawmills where able to cut oak for boards in the radial plane in thicknesses of 0,5 thumb and even thinner. The planed boards in Dutch antique funiture is 9 or 11 mm.
Here we use now the metric system in meters and centimeters (got that from Napoleon) But it is funny to see that the measures in length and width are still adopted from the measurements in feet and thumb. So we have length of 270, 360, 450 cm, which is 9, 12 and 15 foot.
The reason why lumber is still sold in lnegths, widths and thicknesses that correspond to round numbers in feet and inches is, I think, because it is still traded in feet and inches. Not sold to consumers, in places where those consumers think in metric, but traded. For example, rouigh-sawn planks might be listed by a lumber merchant in thicknesses of, i.a., 26 and 52 mm, which in the trade will be designated as 4/4 and 8/4, i.e. four and eight quarters of an inch respectively.
More on lines …
If you have in your library the Dover reprint of “The First American Furniture Finisher’s Manual, A Reprint of ‘The Cabinet Maker’s Guide’ of 1827” (and who doesn’t?), you’ll find that Table V in the Appendix gives all the conversions for different countries into the English units. The “Explanation of Table V” I here recreate in full:
“Table V will be found very useful for changing the different measures of foreign countries to that of England, and the contrary; and is a selection from a very extensive one given by Dr Hutton in his Mathematical Recreations, and needs no explanation; but that the second column is calculated as the supposition, that the English foot is divided into 12 inches and each inch into 12 parts called lines, and each line into 10 parts, making 1440 parts in the English foot; and the foot of all other countries contains the number of the same parts they are equal to, when compared to the English foot.”
So, for example, Table V gives the English foot as 1440 parts and the Paris foot as 1535 parts, or 1 in 1.065977, confirming what Mattias said above.
Prior to the French Revolution, France had over 800 measurement systems. Each town had its own system. That is why Napolion had the Metric System developed.