Sagging Cane Panel Replaced on a Bergere Chair with 3 mm Pre-Woven Webbing and Reed Spline

A bergere with a pressed-cane seat carries the sheet mesh in a groove cut around the seat opening, typically 4 mm to 6 mm wide and about 5 mm deep. When the seat sags, the strands are rarely the first thing to go. The spline dries, shrinks, and lifts, and the panel edge pulls free. Before you buy anything, press the sag with a flat palm. If the mesh springs back and the give is at the perimeter, you are replacing a panel and a spline, not re-weaving by hand through drilled holes. Hand-caning is a different chair entirely, and this one is not it.

Measure the groove width with a set of feeler gauges or a steel rule read against a good light. Reed spline is sold by width in fractions and millimetres, and a spline 0.5 mm too fat will not seat without splitting the groove wall. Sheet cane comes graded by strand width and hole spacing; 3 mm strand mesh with roughly 12 mm hole centres suits most French seat frames from the last century. Order a metre more than the opening on each axis so you have grip while stretching.

Getting the old panel and spline out without wrecking the groove

The spline is the part that fights you. Old PVA or animal glue under a reed spline goes brittle, and a chisel driven straight down will chip the groove lip, which then shows as a ragged line under the new spline. Warm the spline first. A domestic iron set to medium, pressed onto a damp cotton rag laid over the reed, softens both hide glue and older PVA in 30 to 60 seconds per section. Hide glue in particular releases at around 60 C with moisture present, which is the whole reason it was used before synthetics.

Work a narrow flat screwdriver or a spline-lifting tool under one end of the softened reed and lever along the groove, not across it. The reed usually comes up in 100 mm to 200 mm lengths. Once the spline is out, the sheet cane lifts free because its edge was only trapped, never glued to the frame face. Scrape the groove clean with a bent nail ground flat or a purpose-made groove scraper. Any lump of dried glue left in the bottom stops the new panel seating flush and reappears as a bump. Vacuum the debris, then run a dry finger along the groove; it should read smooth end to end.

If a section of groove lip has chipped, fill it now with a two-part wood epoxy tinted to the frame, cut back flush once cured. A filled lip holds spline as well as original timber. Do not try to glue splinters back; they never sit true and the spline rides over them.

Soaking, positioning and driving the mesh

Dry sheet cane is stiff and cracks at the fold. Submerge the panel in warm water for 20 to 30 minutes, glycerine optional at a capful per few litres to keep it supple longer. Ten minutes is not enough for 3 mm strand; the outer strands soften while the knuckles stay rigid and split when you drive them.

Lay the wet panel over the opening with the shiny side up and the strand pattern square to the seat front. Alignment matters more than most people expect: a mesh rotated 5 degrees reads as a visible skew once dry, because the eye tracks the long diagonals. Tack the panel loosely at four points with the reed itself or with hardwood wedges, then drive the mesh into the groove with a caning chisel or a flat wooden wedge and a mallet. Start at the centre back, then centre front, then the sides, working the panel taut as you go. You are pushing the mesh down into the groove, not stretching it like fabric; overstretching wet cane produces a drum-tight seat that goes slack the moment it dries.

Once the mesh sits in the groove on all four sides, trim the surplus with a sharp knife just below the top edge of the groove so no strand pokes up beside the spline.

The spline goes in wet, over glue

Run a thin bead of PVA into the groove, not onto the mesh. Cut the reed spline to length for each side, soak it five minutes so it bends around the front corners without cracking, and tap it home over the trapped mesh with a mallet and a block. It should sit flush or a hair proud. Wipe squeezed-out glue immediately with a damp cloth, because dried PVA blocks any finish. Leave the chair to dry undisturbed for 24 hours; the mesh shrinks and tightens as it dries, which is what gives a correctly done seat its spring.

Hide glue on the frame joints while the chair is apart

A bergere that has lost its cane has usually also loosened at the seat-rail-to-leg joints, because the same racking that killed the spline works the mortise and tenon. This is the moment to deal with it, since you are already handling the frame.

Hide glue is the correct adhesive for these joints for one practical reason: it is reversible. A joint glued with PVA cannot be taken apart cleanly in 40 years without breaking timber, whereas a hide-glue joint releases with heat and moisture, exactly the trick you used on the spline. Titebond Liquid Hide Glue or hot pearl glue from a heated pot both work; the hot version grabs faster and suits a joint you must clamp quickly.

Knock the loose joint apart with a rubber mallet, protecting the show faces with a folded cloth. Scrape the old glue off both the tenon and the inside of the mortise; hide glue bonds to wood, not to a film of old adhesive, so a tenon still glazed with dried glue will fail again. A sharp chisel and a scraper get it back to raw timber. Test the fit dry. If the tenon now rattles in the mortise from decades of wear, glue a veneer shim to one cheek to take up the slack before you commit glue.

Brush warm glue into the mortise and onto the tenon, assemble, and clamp with a sash cramp across the seat rails, checking the frame sits square by measuring both diagonals; they should match within a millimetre or two. Hide glue has a short open time, so have the clamp ready before you brush. Wipe the squeeze-out with a hot damp rag while it is still soluble. The joint reaches handling strength in a few hours and full strength overnight.

Bringing the frame finish back and cleaning the brass

A cloudy white haze on an old oak or beech frame is usually surface bloom, moisture trapped in the old wax or shellac, not damage to the wood. A wax reviver such as Liberon Wax and Polish Reviver on a soft cloth cuts through most bloom in a few passes, lifting old grime and hazing without stripping the finish underneath. Work a small area, buff, and see whether the bloom clears before committing to the whole frame. Where reviver alone will not shift a heavy white ring, a light rub with 0000 wire wool barely moistened with the reviver takes it back without cutting through to bare timber.

If the frame is bare or thin in patches and you want to build a surface, danish oil is the low-fuss route. Flood the first coat, let it soak 15 minutes, wipe off every trace of surplus, and leave it 24 hours. The wipe-off is the step people skip, and skipping it leaves a sticky film that never fully hardens. Two to three coats, each fully dry and lightly de-nibbed with worn abrasive between coats, gives a low sheen that suits a bergere better than a high-gloss varnish. More than three coats rarely adds anything you can see.

The brass studs and any cabochons or sabots respond to a cream metal polish on a cotton bud, worked into the detail and buffed off before it dries into the crevices. Lacquered brass that has gone patchy needs the old lacquer off first with acetone, or the polish only cleans the bare islands and leaves the lacquered ones dull. Decide whether you want the brass bright or left with its age; a fully polished stud next to an untouched one looks wrong on an old chair.

A quick sanity check on strand width

Hold your ordered mesh against the old panel before you soak anything. A 3 mm strand next to an original 2.5 mm reads coarser and dates the repair instantly. Match the strand, not just the hole spacing.

What the drying tells you

The honest test of the job is the sound the seat makes 48 hours later. A correctly tensioned panel over a sound frame gives a low, even resonance when you tap it and no creak when you sit. A seat that ticks or drums at one corner is telling you the spline lifted there before the glue grabbed, or the mesh went in slack on that side. Whether that failure traces back to the soak time, the groove prep, or a frame that was never quite square is the part you only diagnose by taking that corner apart again, and it is worth asking before you reach for the polish which of those three you actually controlled.