The Jar Started Raining at Minute Eight

The jar’s been sealed for eight minutes and already there’s a droplet forming on the lid.

Water cycle speedrun. I’m watching condensation accumulate in real time — a visible proof that the system’s closed-loop feedback is working. The substrate evaporates. The glass condenses. Eventually it’ll rain back down onto the fern, and the cycle repeats. No intervention required. No ammonia strips. No bacterial bloom panic like the aquascaping disaster where I learned the hard way that living systems need time to stabilize.

This one’s different. This one was designed to be hands-off from the start.

I’m at the electronics supply shop picking up trimmer potentiometers when I see it: a glass jar the size of a football, sealed shut, sitting on the counter under a shelf full of voltage regulators. Inside is a miniature forest — moss carpeting what looks like volcanic rock, a fern unfurling from one corner, everything lush and green and clearly thriving. No lid opening. No watering can. Just sitting there, photosynthesizing.

“How long?” I ask the owner.

“Six years this March. Haven’t touched it since I sealed it.”

Six years. The same jar. The same water molecules cycling through evaporation and condensation. A complete hydrological system the size of a teapot, self-regulating since 2020.

He built it with three layers: activated charcoal at the bottom for filtration, then a mix of peat moss and perlite for substrate, then the plants themselves — all species that tolerate high humidity. The trick, he says, is balancing the initial water content. Too much and you get mold. Too little and the plants stress before the cycle establishes. Once it’s right, the system maintains itself indefinitely.

It’s an oscillator. That’s what I kept thinking while he explained it. Get the feedback parameters correct — substrate moisture, air volume, light exposure — and the system finds its own stable operating point. Same principle as designing an RF oscillator where you tune the LC tank and bias until it self-sustains at frequency. Positive feedback through evaporation, negative feedback through condensation, equilibrium somewhere in the middle.

So I bought supplies. Drove to three different garden centres looking for specific plants: Fittonia (nerve plant, tolerates humidity), a small fern whose Latin name I forgot to write down, sheet moss harvested legally from someone’s driveway. The glass container came from a thrift shop — a 4-litre jar with a rubber gasket seal. Sterile substrate because introducing garden soil means introducing pathogens that’ll thrive in the sealed humid environment. This is one lesson the aquascaping tank taught me that I’m not ignoring twice.

Layering took an hour. Activated charcoal first, then the peat-lite mix (peat moss, vermiculite, perlite — the standard formula for closed terrariums, sterilized to prevent microbial overgrowth). Then hardscape: two pieces of lava rock positioned asymmetrically because I can’t stop thinking in Iwagumi ratios after building that tank. Then plants, roots pressed gently into substrate, spacing them to allow growth room.

The critical part: water content. Added it with a spray bottle, watching how the substrate darkened, aiming for “moist but not saturated.” This is where most people fail, according to the terrarium forums I’ve been reading for the past three days. Too wet and you get algae blooms and fungal overgrowth. The fix is opening the lid weekly to vent excess moisture until the system stabilizes, but that defeats the appeal of a sealed ecosystem.

I think I got it right. Sealed the jar. Set it on the windowsill where it’ll get bright indirect light but no direct sun — glass magnifies sunlight, and direct exposure would cook everything inside within hours.

And now: condensation. Clear droplets on the glass, proof that evaporation’s happening, that the water cycle’s already running. In a few hours those droplets will run down the walls and re-enter the substrate. The fern will transpire. The moss will absorb. The cycle continues.

Victorian England had “pteridomania” — a fern-collecting craze enabled by Wardian cases, which were just fancy sealed terrariums built to ship plants across oceans. The British Empire used them to break tea and rubber monopolies by smuggling seedlings from Asia to colonial plantations. They worked so well that Kew Gardens kept using them until 1962.

Six years. That’s how long the electronics shop terrarium has been running. If the water balance holds, if the light stays consistent, if nothing gets introduced that shouldn’t be there — mine could run just as long.

Droplets are forming faster now. The jar’s internal humidity is climbing toward equilibrium. I should probably stop watching and let it do its thing. But there’s something deeply satisfying about seeing a closed-loop system stabilize in front of you, one droplet at a time.