Bench / Projects / An 18th-Century Elm Coffer: Six Hours of Work, Almost All of It Cleaning
An 18th-Century Elm Coffer: Six Hours of Work, Almost All of It Cleaning
A boarded elm coffer with its original iron banding and a lid that had been sat on for three centuries. The correct treatment was almost entirely cleaning — no stripping, no sanding of the surface, no attempt to make 1700 look like 2026.
Before
After
| Field | Value |
|---|---|
| Piece | Elm coffer on a plank base, iron banding |
| Estimated date | 1700-1760 |
| Substrate | Solid timber |
| Old finish removal | Mechanical scraping only |
| New finish system | Wax only (microcrystalline) |
| Coats | 2 |
| Dry between coats | 24 h |
| Bench hours | 6 h |
| Materials cost | CNY 90-160 |
The single decision that defined this job
The coffer arrived with a lid that was almost black, a body that was grey, and iron banding that had rusted at every fixing. The owner's request was "make it look like new".
That request cannot be granted, and the reason is not sentimentality. A boarded coffer from 1700 has no veneer, no inlay, no complex joinery — what it has is a single slab of elm 30 mm thick, cut from a tree that was already old when it was felled, and a surface that three centuries of use has polished into something no new piece of elm has. Sanding that off produces a plank. It does not produce an 18th-century coffer.
So the treatment was: clean, stabilise, wax. Six hours spread across two days, of which four and a half were cleaning.
Bench note: The black on the lid was not dirt sitting on the wood. It was the wood. Elm darkens with light and with the oils from hands and clothing, and the lid of a coffer is the most handled surface in any house. What looked like filth was 300 years of that process.
Wardrobe beetles, and why you check before you clean
Before anything else, the whole piece was checked for woodworm. Fresh frass — pale, powdery, the colour of the timber, and it moves when you blow on it — means active infestation. Old frass is dark, compacted and stuck to the surface, and it means the infestation ended long ago.
This coffer had old frass in two places on the underside of the base and nothing fresh. That determined that no treatment was needed. Had there been fresh frass, the piece would have been bagged, and the cleaning would have waited for an assessment.
Take the loose debris out with air and a brush, dry
No water at this stage. 25 mm soft brush along the grain of the boards, working the dust out of the gaps between the planks with the vacuum nozzle held just behind the brush. The gap between two boards on a coffer is where centuries of dust, seed husks and insect debris end up.
The hinges and the banding fixings got a bamboo skewer and a dental pick. This took an hour and produced about 200 g of debris.
Wash the body, and only the body
Linseed oil soap at half strength, on a cloth wrung almost dry, one board at a time, wiped off with a clean damp cloth and dried immediately. Never on the lid; the lid's surface is the most valuable thing about the piece and any water on it risks lifting the black.
The body came up from grey to a pale honey colour. That change is the removal of three centuries of dust and grease, and it is reversible if anyone later decides the grey was better. Nothing was removed that cannot be put back.
Scrape only the raised grain around the lid edge
The lid edge had raised splinters in three places where it had been dragged. These got a straight scraper, worked in the direction of the grain, taking off the minimum needed to stop them catching on hands and clothing.
Nothing else on the whole piece was scraped or sanded. Not one abrasive touched a surface.
Stabilise the ironwork with tannic acid, not with a wire brush
The rust on the banding was surface rust, stable and dry. Wire-brushing it back to bright metal would have removed the forged surface texture along with the rust and would have left iron that rusts faster than before, because clean iron has no protective oxide layer at all.
Instead: a brass wire brush used lightly to lift loose flakes, then a tannic-acid rust converter brushed on the remaining rust and left to go black. Tannic acid converts iron oxide into iron tannate, which is stable, dark and does not continue to rust.
Then microcrystalline wax over the whole ironwork, worked into the fixings with a brush. That wax is the only thing standing between this banding and another 50 years of rust.
Wax the body and the lid with a hard wax, thinly
Renaissance wax or an equivalent microcrystalline wax, applied with a cloth in a very thin film and buffed after 15 minutes with a separate cloth. Microcrystalline wax rather than beeswax because it is harder, less sticky, and does not hold dust the way beeswax does.
Two coats, 24 hours apart. The first coat goes into the dry surface and disappears; the second is the one you can see.
Write down what was done, and pass it on
The last step of the job was a one-page note: what was cleaned, what was scraped, what was converted and what was waxed, with the date. It went into the coffer.
A treatment record inside a piece of furniture is not a romantic gesture. It means that in thirty years, when someone else opens this coffer and wonders whether the black on the lid is original or was applied in the 1970s, they will know. That is worth more than the wax.
A piece of this age should be assessed by a conservator before any intervention. This is a record of a minimal cleaning treatment, not instruction. Information only.
Abrasive ladder
Grit sequence, coarsest first — each step removes the scratches of the one before.