Reading Furniture by Its Joints and Tool Marks

Furniture records its own manufacture on the surfaces nobody finished: drawer sides, backboards, the undersides of tops. That is where to look, and it is where reproductions are least careful.

Dovetails

The joint at the corner of a drawer is the single most useful character in furniture dating.

Hand-cut dovetails are few in number, often only two or three on a drawer side, wide, and slightly irregular. The pins — the narrow wedge shapes — are typically much narrower than the tails, because a hand craftsman minimises the sawing. A scribe line marking the depth is usually visible, and it runs across the whole board because the cabinetmaker scored it before cutting.

Machine-cut dovetails are numerous, evenly spaced and identical, with pins and tails of similar width. Early machine dovetails from the 1870s onward often have a distinctive rounded profile from the cutter.

Neither proves age on its own — hand-cut joints are still made — but machine-cut joints put a lower bound on the date, and that is frequently enough.

Saw marks

Look at unfinished surfaces: the back of a case, the underside of a drawer bottom.

Circular saw arcs are the most useful because they are unmistakable and they set a firm earliest date.

Screws

Screws changed in a well-documented sequence, and they are hard to fake because the whole screw would have to be reproduced.

Hand-made screws before about 1815 have off-centre slots, irregular threads, and a blunt end rather than a point. Machine-made screws from 1815 to about 1850 have regular threads and still a blunt end. Pointed screws, made by a gimlet-point machine, date from about 1850 onward.

A blunt-ended screw with an off-centre slot is genuinely early. A modern Phillips-head screw in an eighteenth-century piece means a repair at best.

Wood shrinkage

Wood shrinks across the grain and barely at all along it, and it continues for a very long time. A round table top made two hundred years ago is now measurably oval, by as much as several millimetres across a large top.

Measure a circular top in two directions. If it is perfectly round, it is either recent or made from a stable modern board.

Where reproductions slip

Reproductions get the visible surfaces right and the hidden ones wrong. A piece with a beautifully aged finish and machine-cut dovetails, or perfect hand-cut dovetails alongside pointed screws, is telling you two contradictory dates.

Contradiction is the finding. A genuine piece is internally consistent because everything about it was made at the same moment, and the cheapest thing to check is whether the hidden work agrees with the visible work. Dating furniture in five checks covers the rest of the sequence.

Secondary woods

The visible timber is chosen to look good and the hidden timber is chosen because it was to hand. That makes drawer sides, backboards and dust boards regionally diagnostic.

American eighteenth-century furniture commonly uses tulip poplar, pine or chestnut as secondary wood. English furniture of the same period uses oak or deal. A piece presented as American with an English secondary wood is worth a second look, and vice versa.

Chestnut is particularly informative in American furniture, because chestnut blight destroyed the species commercially in the early twentieth century. Chestnut as a secondary wood indicates the piece predates that.

Patina inside and out

Wood darkens with light and air, so surfaces that were exposed differ from those that were not. Inside a drawer, the wood should be lighter than the outside. Under a table top, lighter still.

A piece where interior and exterior surfaces are the same colour has either been stripped and refinished, or made recently and coloured all over. Both are worth knowing before buying, and the check takes one open drawer.