Roman Belt Fitting

Choose a documented plate type from an excavation report or a museum catalogue and copy its dimensions. Plates of the mid-first century AD are the best recorded; inventing a "Roman-looking" plate produces an object that teaches nothing.

Draw the plate at 1:1 — typically a rectangle around 50 x 45 mm with rivet holes at the corners. Everything downstream is cut to this drawing.

Carve a master pattern slightly oversize to allow for shrinkage. This is the piece the mould is taken from, so every fault in it repeats in every plate.

Press the pattern into casting sand to make a two-part mould, then cut a sprue and vents. Roman fittings were cast in two-part moulds and by lost-wax, and either route is authentic.

Work outdoors or with strong ventilation, in leather boots and gloves and a face shield, with dry tools and a clear path from furnace to mould. Molten metal and any trace of moisture react explosively — every mould and tool must be bone dry.

Melt brass or bronze in a crucible and pour steadily in one continuous stream. Roman belt plates are copper alloys of varying composition — brass, bronze and leaded mixtures all occur.

Let the casting cool fully before opening the mould, then saw off the sprue and file the stub flush.

File the faces flat and true up the edges. A cast plate always comes out with a rough face; the finish is filed on, not cast in.

Drill the corner holes to match your rivets and countersink them lightly on the back so the rivet heads sit down into the leather.

Tin the front face to a bright silvery finish. First-century belt fittings were regularly tinned or silvered — the belts were meant to catch the light, and a plain brown plate is the wrong object.

Engrave a geometric or vegetal pattern with a graver, cutting a clean V-section groove. Cut slowly and let the tool do the work; a slipped graver cannot be undone on a tinned surface.

Fill the cut lines with a black inlay and level it back to the surface. The Roman material was niello — a black silver-sulphide paste — and the effect is a bright plate carrying a black-line design.

Cut a strap of vegetable-tanned leather to the width of your plates and mark the rivet positions in a row along it.

Punch the holes, push the rivets through from the front and peen them over on the back against a steel block. Riveting through pre-drilled holes is exactly how the originals are attached.

Flex the finished strap. Plates set too close make a rigid band that will not sit at the waist — the gaps between plates are what let a metal-covered belt behave like a belt.

A badge, not a fastening. The cingulum militare was worn as a mark of military status by soldiers and officials, and losing the right to wear it was itself a punishment. The decoration is doing social work: this is a uniform component in an army that had no uniform in the modern sense.

How they were actually made. Plates were produced by casting — lost-wax or two-part mould — or by embossing sheet, or a combination of both, in copper alloys of varying composition. In the first century AD in particular they were regularly tinned or silvered, then engraved with geometric or vegetal designs, and the engraved lines filled with niello or enamel. Fixing to the leather was by rivets driven through pre-drilled holes.

The apron. The belt was often completed by a hanging apron of several leather straps, the pteryges, each covered in a row of riveted metal fittings. It offered essentially no protection; it hung, it swung, and it made a noise when the wearer moved. It appears to have been decorative and demonstrative, which is a legitimate function.

Dating by belt plate. Belt fitting styles changed fast enough that archaeologists use them to date deposits — Tiberian to Neronian, or Claudian to Flavian, roughly the middle and later first century AD. A soldier's belt is one of the more precisely datable objects on a Roman military site, which is why so much has been written about a strap.

Where the finds come from. Most surviving fittings come from military sites along the Rhine and Danube frontiers and in Britain, often as single plates lost from the strap. Complete belts are rare; the reconstruction of how a set was arranged is largely inferred from tombstone reliefs, which are pictures rather than objects — a limit worth stating whenever a "typical" Roman belt is described.