One Baked Bone Before the Metal Stayed
The eBay seller’s storefront had exactly what I was looking for: sockets for IN-14 Nixie tubes, NOS from Soviet factories, three dollars each plus shipping from Kharkiv. I added six to the cart. Then I kept scrolling.
The bezels caught my attention. Not 3D-printed PLA brackets in off-white or silver filament, but hand-cast metal rings with art deco detailing—stepped geometric patterns that echoed the stacked cathodes visible through the glass. The listing said “pewter alloy, lead-free, hand-finished.” Forty-two dollars for a set of six.
I messaged the seller about the casting process. His reply arrived at 3 a.m. my time: “Cuttlebone. Same bones from pet shop for birds. Very easy technique, works on kitchen stove. Watch this video.” The link pointed to a twenty-minute tutorial where someone carved a pattern into what looked like a chunk of white foam, clamped two halves together, and poured liquid metal into a hole. Out came a ring. No kiln, no foundry, no lost-wax apparatus.
I watched it twice, then ordered a pound of lead-free pewter ingots and a bag of cuttlebones from an aquarium supply company.
Cuttlebone isn’t bone. It’s the internal shell of a cuttlefish—a chambered structure made of aragonite, the same mineral that forms mother-of-pearl. Cuttlefish use it for buoyancy control, adjusting the ratio of gas to liquid in its chambers to hover at precise depths. The material is over 90% porous, which is why it carves so easily with a blunt tool. The same microstructure that lets a cephalopod fine-tune its depth creates the fine surface detail that makes it useful as a mold.
Pet stores sell it as a calcium supplement for birds. They charge two dollars a piece. I bought a bag of twelve.
The technique is simple enough to explain in a paragraph. Cut a cuttlebone in half lengthwise. Sand both cut faces flat so they meet cleanly. Press your pattern into one half—a coin, a carved wax original, anything with relief. Carve a sprue channel from the top edge down to the cavity. Clamp or wire the halves together. Pour in molten pewter. Wait. Split the mold. There’s your casting.
Everything that can go wrong happens in the details.
Pewter melts at 170–230°C depending on the alloy, low enough for a camping stove. The ingots I bought are Britannia metal—92% tin, 6% antimony, 2% copper—which melts at 255°C. The antimony is the interesting part: it causes the alloy to expand slightly during solidification instead of shrinking. Most metals contract as they cool, pulling away from mold surfaces, losing definition. Antimony pushes outward, forcing metal into fine details. The same property made Britannia metal the standard for Oscar statuettes from the 1940s until 2016. Every golden envelope scene for seventy years featured a pewter core plated in copper, nickel, and gold.
My first cast didn’t survive long enough to need finishing.
The cuttlebone had been sitting in open air for a week, which seemed like plenty of time to dry. It wasn’t. Some residual moisture—maybe from the porous structure, maybe absorbed from the basement air—flashed to steam the instant the metal hit. The pewter spat back out of the sprue hole in a fountain of grey droplets. One landed on the workbench and hardened into a tear-shaped blob. The rest scattered across the concrete floor.
The forum posts all warned about this. I thought I’d read them carefully enough.
Second attempt: I baked the cuttlebone halves at 120°C for an hour. The carving I’d pressed into the surface—a test pattern, just a Canadian quarter—survived the heat without distortion. This time the metal flowed. I watched it fill the sprue channel, hesitate at the cavity entrance, then drop out of sight as it found the shape. No spattering. No steam.
The result, once I cracked the mold and brushed away the carbonized cuttlebone residue, was recognizable. A disc with the caribou relief from the quarter, slightly fuzzy around the edges where the mold hadn’t captured fine detail. The surface had a matte grey finish that looked nothing like metal—more like concrete, or weathered stone. But the weight was there, dense and cold in my palm.
The next few hours disappeared into iteration. A deeper sprue cut to feed metal faster. A vent channel on the opposite side to let air escape as the cavity filled. A heavier clamp to prevent the mold halves from shifting. By evening I had three usable castings, four misshapen failures, and a growing pile of carbonized cuttlebone fragments that smelled faintly of fish and burnt toast.
I haven’t polished anything yet. The rock tumbling taught me that finishing is its own process—progressive grits, each stage erasing the scratches left by the last. Pewter follows the same logic: 400, 800, 1200, polishing compound. But polishing erases surface texture. The matte finish from the cuttlebone, the organic grain, the slight asymmetries—all of that disappears under a mirror shine.
I’m not sure yet which version I want. The Nixie bezels that started this entire detour were polished bright, but they also lost something human in the process. The failures from tonight, sitting in a row next to the crude successes, have more character than any of them.
Tomorrow I’ll try carving an original pattern instead of pressing found objects. The cuttlebone takes detail easily—a bent paperclip, a blunt pencil, anything harder than chalk. But for now the workbench is covered in metal and ash, the camping stove is cooling on its tile, and I’ve spent an evening pouring liquid metal into fish bones like it was a reasonable thing to do.
It was.