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.
The Fauteuil [armchair] that I am going to describe is one of those that is called a Cabriolet, because of the circular shape of its plan, [which is] different from that of the Queen’s Fauteuils [armchair], which is straight from the side of the back, as one could see in the view of which I made the description of the Queen’s Chair, page 614, etc.
I have chosen this form so that in the description of chairs and fauteuils [side chairs and armchairs], I am not required to repeat myself. What I said of the Queen’s Chairs can be applied to armchairs of the first type. What I am going to say about cabriolets can be applied to side chairs of the second.
Cabriolets are the seating [that is] the most fashionable at present, and at the same time demand the most attention on the part of the Worker, especially with regards to their construction and the cutting of the wood for the back, which being circular in plan and splayed, forms a part of the surface in a tapering shape, which the Joiners call faire la hotte [make the hood. “Unehotte” is a reference to the type of large conical basket splayed toward the top – like the back of the armchair – you cut a cone in half on its center and there you have it, a “hotte” and the back of the armchair].
To make these sorts of armchairs with all the perfection that it is possible to accomplish, one must first begin by taking account of the shape of the plan, which normally is an S in the front, and in a half-circle, or better said, a half-oval in the rear, like in Fig. 5 and Fig. 8, which represent half of the plan Fig. 5, a half larger than the latter, so as to make the operations more sensible.
After having thus drawn this plan of Fig. 8 (the half could be taken for the whole) at about 15 thumbs from the front of the seat on the line from the middle a b, you raise a perpendicular c d, which you set at 11 thumbs in height. Then from point d to point e, which is the center of the part of the circle of the back of the seat, you take a line e f, which represents the middle of the rear leg, to both sides of which line you trace the width of the rear leg parallel to the latter. Whatever be the flare, or to speak like the Workers, the reverse [the angle/tilt] of the back, the face of the upright should always present itself perpendicularly to the curve of the seat where the exterior contour is indicated by lines g, g, g and the interior (at least of the cross-pieces) by h, h, h. Then there remains to draw on the plan the length of the back seat rails and their splay. This cannot be done except after taking into account the height of the back and the form of its contours that you must first draw separately on the surface developed from the back, which is done in the following manner:
The splay of the back being determined, as in Fig. 5 from a b, from these points you lift two perpendiculars on the line of the middle of the seat, which parallels you extend indefinitely outside of the Figure. From point e, (which is the center of the arc of the back of the plan), you lift up likewise a perpendicular parallel to these latter, which you extend indefinitely on both sides. Then, at whatever distance, like in Fig. 4, you lift from this line perpendiculars f g, and g d, of which distance f g is equal to the height of the back. Then, from point d, you pass an oblique line by point h, that you extend until it meets line g f e i at point i (which is found outside the plate), from which point like the center and distances i f and i g, you describe the circular arcs f m, and g n, Fig. 4. This being done, you take on the plan, Fig. 5, the distance a l, that you transfer, Fig. 4, from f to o. From this point and from point i, you pass a line o p, which is the middle of the rear leg. You draw this as usual for both the curves and the meeting of the back rails, whether this curve is of a normal form like side A, on which I just made the demonstration, or even if it is an oval like on side B, which is no matter. The only exception is when the upright must be wider within, which I will speak of in its turn.
The curve of the back being thus drawn, you draw separately, Fig. 7, the upright of the back (which is double the proportion of Fig. 4 in order to correspond to the plan in Fig. 8), that you extend just to the total height of the back. Then you draw on the upright all the locations of rails, both at the top and bottom at their greatest width, as indicated by points a, b, c, d, from which points you lower the line i l, as many perpendiculars as the distances on this line are carried over to the plan of Fig. 8; namely, that of i h, Fig. 7, from I to 2; that of i g, from 1 to 3, which gives the splay of the bottom back rail, that of i f, from 1 to 4; and that of i e, from 1 to 5; which gives the splay of the top rail, which you draw, like the other, with the circular arcs described from center e, Fig. 8.
The lower sections of these rear legs is nothing different from the others of which I already spoke; it is only that the serpentine leg is more splayed to the outside so as to make more of a stable position to the seat, what the Joiners call shoring up [to brace], which should be 2 thumbs at least.
I said above that armchairs differ from side chairs in that the first have these armrests intended for the elbows of those who are seated within. These armrests are composed of an arm a, Fig. 3, of a bracket b, which is assembled at one end to the side seat rail and the others in the arm, which is assembled itself by mortise and tenon on the upright, with which you should take care to make it match in a smooth and gracious manner, as I have noted in Figs. 1, 2 & 3.
The assembly of the arms with the uprights is done squarely, but I believe that whatever the use, one would do very well to make a cut [an angled shoulder], which, by preventing the inconveniences of squared cuts of which I spoke above, renders the work more solid, in that the cut from below would support the arm and would prevent it from dropping further down.
The arms of Fauteuils are drawn on the plan, as are the rails of the back, with the exception that they are not splayed except at the end where they connect with the upright, the other being perpendicular, which gives it an awkward form that you must keep square, as I have indicated by punctuated lines m n and o p, Fig. 7. See also Figs. 5 & 8, where these arms are drawn on the plan, as well as the brackets, of which I will make a more extensive description afterward in speaking of the different sorts of arms of Fauteuils and their brackets.
The Fauteuil of which I am making the description here is prepared to receive a caned seat, as you can see in Fig. 1, which represents it viewed from the side. That of Fig. 2 represents it viewed from the face, the side A completely disassembled and ready for cutting out, and the other side B completely cut out and assembled but for the seat, which is installed only after being finish with the cane, because the tenon of the bracket passes through it to be pegged in the side seat rail.
See also Fig. 6, which represents the rear seat rail of the armchair which receives the seat, as I explained up above, and Fig. 8, where I indicated by punctuated lines i, i, i, the outside of the frame of the seat, of which the projection ends at both uprights, and where the interior indicated by lines l, l, l is wider at the rear to leave solid wood in front of the upright.
I said up above that the frames of seats are assembled en chapeau [capped] from the front. However, I believe that for the neatness of the work, it would be much better to assemble them mitered in the front, like line l i, and at the rear when they are curved, as in this instance by a forked joint [bridle joint], at the space of the notch of the rear feet. The height of Fauteuils is a bit the same as that of side chairs with the exception that the seat should be a bit lower and consequently the back higher in proportion, especially when they are more splayed.
As to their width, they should be more considerable than that of side chairs given that it is necessary that the person who is seated within be contained comfortably with their clothes. That is why you make some width to the seat of Fauteuils from 22 to 26 thumbs by 18 to 20 thumbs of depth, at least for ordinary Fauteuils, that is to say, in public rooms. For those that serve in particular for a single person, one must, as I said above, consult that person’s taste and needs.
The size and the cut of wood for ordinary armchairs is nothing different from that of side chairs, if only in the case of cabriolets, the rails of the backs should be cut according to their tilt, or better said, their splay, which you can do by drawing the top and the bottom with some templates, of which you should have the curve on the plan, in backing them off [to the back] as necessary. What’s more, you could, without any type of loss, take the top and the bottom rails from the same piece of lumber, sawn as nestled patterns, which is very easy to do given that they are of different curves, such that the outside of one can be the inside of the other, at least pretty close.
There you have a bit of the detail of an armchair (and consequently of a cabriolet chair), after which one could construct all sorts of seats, of whatever form they be, given that the method that I just gave for the construction and manner of drawing it here is applicable to all with some minor differences. I have greatly expanded the manner of drawing, both the plan and the elevation, of these sorts of seats, so as to be within reach of the greatest number. They would have not understood me if I would not have been so expansive if I had simply said, as would seem completely natural, that the development of the backs of seats [of cabriolet arm chairs] is only being one part of the surface of a truncated cone, of which the incline is given by the back and is elongated just to its meeting of the center of the seat which represents the axis of the cone, which determines the crown and consequently the center of its development. This simplicity supposes of my readers (at least of ordinary Joiners) some knowledge which they cannot or do not want to acquire, whether I have given the elementary principles in the second part of this work, at the beginning of the Art of Drawing. That is why I believed it necessary, to be available to all, to make all demonstrations that appeared appropriate for saving time of those who would not acquire other knowledge than that of practice, which, for as little as is reasonable, is barely sufficient in the part that I am treating.
What’s more, Chair Joiners do not take all the precautions that I recommend here for drawing the plan or elevation of their works, which [they do by] sawing as accurately as possible, and that they assemble without dressing them, for cutting them out later, after having assembled them, which they do badly. But finally it is their custom and they will not change from that easily.