Identifying the finish: the solvent sequence
The test as a decision table, in the order the solvents go on, with what each answer rules in and rules out, and the one result that means stop rather than carry on.
The sequence is published by Bob Flexner, a finishing writer, in Popular Woodworking in April 2016. The stop condition and the solvent cautions come from the National Park Service's Conserve O Gram leaflets. Finding out what finish is on your furniture is the chapter that runs it.
Before you start
- Pick a place nobody will ever see and where the finish is sound: an inside face, an underside, the back of a leg. Not on damage, and not on a top.
- Use a fresh swab for each solvent. A swab carrying mineral spirits into the alcohol step confuses the result.
- Read the labels and the safety data sheets, and ventilate throughout the work and the drying. That is the Park Service's own instruction to staff. Mineral spirits, sold as paint thinner, is a Class II combustible liquid with a flash point of 102 to 110 °F.
- Decide where the rags go before you open anything. Flat, single layer, on concrete, outdoors.
The sequence
| Step | Solvent | Reaction | What it rules in | What it rules out |
|---|---|---|---|---|
| 1 | Mineral spirits (also naphtha, turpentine) | A waxy smear on the swab; the spot dulls or clouds | Wax on the surface. Keep wiping with fresh swabs until nothing more lifts, then dry and go on | Nothing about the finish underneath yet. Until the wax is off you are testing the wax |
| 1 | Mineral spirits | The finish itself softens or goes tacky | Stop: see the box below | Everything. This result ends the session |
| 1 | Mineral spirits | Dirt lifts, finish unchanged | Mineral spirits is a safe cleaning agent for this surface, which is what the cleaning chapter needs to know | A wax layer, and a fragile finish |
| 2 | Denatured alcohol | The surface goes sticky, or the finish dissolves; color may come off on the swab | Shellac | Lacquer, varnish and the cured coatings |
| 2 | Denatured alcohol | Nothing | Not shellac. Continue | Shellac, in the place you tested |
| 3 | Lacquer thinner | The finish dissolves or goes sticky | Lacquer | Varnish and the newer cured finishes |
| 3 | Lacquer thinner | Goes soft and tacky rather than dissolving | Possibly a water-based finish, which the source says goes soft and tacky in lacquer thinner, toluene and xylene | Little on its own: confirm against the date range below |
| 3 | Lacquer thinner | Nothing, after nothing at steps 1 and 2 | "Varnish or one of the newer more-solvent-resistant finishes," in the source's words. A group, not a name | Wax, shellac and lacquer |
The stop. The Park Service's cleaning leaflet says that if the mineral spirits softens the finish, do not attempt to clean the piece, and consult a furniture conservator for professional assistance. That is a stop rather than an instruction to work more gently or to try a weaker solvent. Record the result and leave the surface alone; Who to hire for furniture work is what follows.
The group. When nothing reacts, the honest entry on your report is the group rather than "varnish". The group spans oil-based varnish, polyurethane, catalyzed conversion finishes, polyester and UV-cured coatings, whose earliest and latest members are separated by most of a century. For every cleaning and waxing decision in this manual, knowing the surface is solvent-resistant is sufficient.
The assumption under everything. The method's own author states it: the reading assumes you are dealing with an original, not refinished, finish. A refinished piece tests as its newest coat, which tells you about the last person who worked on it rather than about the piece.
What each answer suggests about a date
The period dates below are from the same article, and they are this manual's one dating clue with a published source behind it. They date the finish, which on a refinished piece is a different thing from the piece.
| Finish | What the article says | Reading it as a date |
|---|---|---|
| Wax or linseed oil | What makers used in the eighteenth century and earlier, with what was available | Consistent with early work, and with any later piece finished this way |
| Shellac | Alcohol-soluble resins widely available by the 1820s; shellac the overwhelmingly dominant finish for the next hundred years | A floor around the 1820s, and most likely before about the 1920s if original |
| Nitrocellulose lacquer | Available in the 1920s, slowly replacing shellac | A floor in the 1920s |
| Catalyzed finishes | Further resins from the 1960s began replacing lacquer, mainly catalyzed; polyester and UV-cured later, no date given | A floor around the 1960s, within the solvent-resistant group |
| Water-based | Introduced in the 1980s and 1990s to reduce solvent emissions | A floor in the 1980s |
How old your furniture is uses these alongside the construction clues, and explains why this manual prints no date ranges for saw marks, dovetails, screws or nails.
What each finish tolerates
| Finish | How it cures | Redissolves in | Damaged by | Cleaning and waxing |
|---|---|---|---|---|
| Wax | Solvent evaporates; wax remains | Mineral spirits, naphtha, turpentine | Heat, abrasion, its own solvents | Removable, which is why wax is called a sacrificial layer |
| Shellac | Solvent evaporates; the film stays soluble | Denatured alcohol; attacked by lacquer thinner | Alcohol, water, heat | Keep water off it; clean with mineral spirits on a cloth. Waxing suitable on a sound film |
| Nitrocellulose lacquer | Solvent evaporates; the film stays soluble | Lacquer thinner | Solvents; crazes as it ages | The published cleaning routine applies. Treat a crazed surface gently |
| Oil-based varnish and polyurethane | Cures chemically; does not redissolve | Nothing in this test | Abrasion, light, eventual embrittlement | Named by conservation guidance as best maintained with a thin coat of paste wax, no more than once a year |
| Catalyzed, polyester and UV-cured | Cure chemically, often very hard | Nothing in this test | Impact and abrasion rather than solvents | As above; these are the most tolerant surfaces here |
| Water-based | Coalesces as water leaves | Soft and tacky in lacquer thinner, toluene, xylene | Solvents, some cleaning products | Test any wax on a hidden spot first, as the waxing guidance asks for a first waxing |
| Oil finish | Cures in and just below the surface | Not a film to dissolve | Wear through to bare wood | Clean sparingly; mineral spirits keeps lifting for as long as you wipe |
| Paint | Varies | Not tested here | Varies | The cleaning guidance stops at dusting unless you are trained. Read the hazards chapter first |
| Unfinished | No finish | Not applicable | Water raises the grain; dirt enters the pores | No water. Paste wax is not recommended on bare wood: it penetrates the pores and is nearly impossible to remove |
Dana's walnut chest. Tested on the underside of the top rail. Mineral spirits lifted gray dirt and no wax smear, and the finish did not soften. Denatured alcohol made the spot sticky within seconds, with a faint brown on the swab.
Reading the table: row 3 at step 1 (safe to clean with mineral spirits, no wax), then row 1 at step 2 (shellac). Step 3 was not needed.
What it changed: water stays off this piece, which rules out the soap-and-water step of the cleaning routine; and the white ring on the top is probably a mark in a shellac film rather than a stain in the wood, which raking light can test further.
Still assumed: that the rest of the chest carries the same finish as that one rail, and that the shellac is original. One test spot is one test spot.
What to write down
- Where you tested, specifically enough to find the spot again.
- What each solvent did, including the ones that did nothing. A negative is a result.
- Whether the finish softened under mineral spirits, recorded separately, because that is the stop condition rather than an identification.
- Whether there was wax, recorded as a layer rather than as the finish.
- The group rather than a name, where nothing reacted.
All of it goes on the What it is sheet in your condition report, which has a column for the finish and a column for what the solvent test did.