I few months ago I embarked on my “Making Roman Jewelry” project. The point was to use wire wrapping to help SCAdians discover the joys of making period adornments in their living rooms. In the course of researching and building that class, I was inspired to learn real metalsmithing. I’ve been assembling earrings and necklaces for decades, but suddenly I really wanted to solder, and set gems, and fabricate! Fortunately, my good friend Fjorlief Inhaga is a brilliant artisan, and she’s been teaching me and letting me use her studio.

The Hercules Knot represents strength and came to be associated with marriage. It’s a common motif in the 1st-3rd centuries CE. This one is dated to the 2nd-3rd century. Christie’s, lot 177, sale 1445.

For my first project, I wanted to duplicate this early Imperial necklace of emerald and gold. For cost purposes, I used brass and glass (a period substitute). The bracelet was my first soldering project, and had some issues. I’m very happy with the necklace, although of course there’s always room for improvement.

The first step was cutting and bending brass wire into tiny loops and soldering them, using a gas torch. Brass is an alloy of copper and zinc. Heating and pickling (a chemical solution used after soldering) brings the copper to the surface, so the brass looks pink or reddish, depending on the light. The loops on the left are my very first solders, and you can see how sloppy they are. 

Here’s how the loops look when intertwined for the necklace. I slid them together and modified the bends so they laid flat.

I soldered both the points where the curved arc crossed the other loop “handle.” This keeps them in position and looking consistent.

I coiled more brass wire around a metal tool (it’s for poking tiny holes or dents, but it worked great for this) to make a spring. Then I locked the spring into a ring clamp and used a jeweler’s saw to cut them into identical round twists.

Once they were cut, I used pliers to bend the tiny twists into circles. You can see the one at the top of the picture hasn’t been closed yet. YAY for remembering to take pictures!

To give the curved edge of the jump ring a smooth connection, I used a rounded file on the trimmed ends of the soldered knots. The bottom right is not yet filed into a curve.

Joining the rings to the knots is tricky, both because it’s a tiny little join, and because you have to careful not to melt your previous solders!

 I tried hard to line up the openings in the rings with the knots ends, but sometimes they moved on me. You can see at this point (the necklace) I am using much less solder and the joins are prettier. YAY for practice!

The finished pieces go into a 2nd pickling bath – 3 parts hydrogen peroxide, 1 part vinegar – for half an hour. See the streams of micro bubbles rising from the metal? That strips off the pink copper coloration. A quick plunge in pure vinegar, a quick scrub with a brass brush, and they are ready for polishing with a motorized whirling scrubber!

 After and before polishing. Note that these two are from the bracelet (only half intertwined).

During the second batch, I accidentally discovered that leaving the knots in the final vinegar bath for 5 minutes makes them come out much brighter. They didn’t need much polishing (below).

I took the completed knots home, where I used wire wrapping to add the glass “emerald” component and made a clasp. I chose African-made glass beads to provide some variation and that “handmade” feel.

Things I would do differently: Mostly I would try to ensure all the rings’ openings stayed centered. Ideally I’d use period metalsmithing techniques, but soldering over fire while using a blowpipe on these tiny pieces frankly sounds like a nightmare. Just finding a way to hold them still without interfering in the metal melting, while maintaining the precise heat to melt the solder but not the main metal… Ug. Much research is coming.

I used wire wrapping for the beads, while most examples show either soldering at the beads or very minimal wrapping, but that’s not necessarily wrong. You can see wrapping in plenty of other Roman artifacts and on the clasps below.

The “tails” on my knots are longer than the extant examples we have (left and below, all from the British Museum). In fact, many are continuous loops. Here’s an extant part of a necklace I’d like to try making. It’ll be a good “smooth solder” challenge. [Someone asked why I chose not to do loops this time. Getting a perfect brass solder at my level of experience would have been difficult – the Romans were working in high karat gold which is much easier. Also, we were working from a pic of one example, and it happened to have tails (or at least looked that way in the pic, blue background below) and so that’s what we did…]

Still, I am very pleased with my necklace, particularly as a second work. I feel confident that (aside from not being gold) it would fit well in a Roman lady’s jewelry chest.

For more details, sources, and historical context, wait for my official paper. And of course you’ll probably see me wearing mine at events. 🙂 I’m wearing it at the office today…