I was much closer to the ground.
That advantage would become less useful if the ground disappeared beneath water.
The feeder twisted in the basin.
One front hook scraped against mineral. The other remained tangled beneath folded fungus. Its mouthparts closed around a torn living strip and pulled.
The obstruction shifted.
Water surged through the smaller groove.
I stopped watching.
A temporary trap was still temporary, even when it looked impressive from a safe angle.
I crossed toward the far side of the basin.
The first mineral ridge was already slick. I tested it with my right lower tentacle, received a delayed impression of smooth stone, and chose to distrust both the stone and the tentacle.
My front foot held.
The rear slid.
My shell swung toward the water.
I flattened myself, shifted uphill, and waited for the next feeder vibration to pass before moving again.
No Mana.
I had twenty-six points and no confirmed way to recover them.
Falling into the basin would be a poor reason to spend one.
I crossed the ridge in short contractions, used a dead fungal plate as temporary traction, and reached a rougher section above the secondary groove. From there, the remaining retreat became slow.
Slow remained dangerous.
Fungal fragments fell behind me. Water spread across old dry lines. Twice, I stopped while the feeder's movement shook the stone.
Eventually, a narrow crack appeared in the far wall.
The opening was barely wider than my shell.
Promising.
That word had developed a history.
I touched the entrance.
Dry mineral along the upper edge.
Damp stone below.
No visible blue threads.
I pushed my head inside and extended both lower tentacles.
The left reported pressure.
The right found open space, then stone farther ahead.
I entered halfway.
My shell scraped one side but did not wedge.
The floor narrowed, then widened into a shelf above the basin. Water from the secondary groove passed near the lower edge without entering the deepest part of the fissure.
The ceiling held loose mineral flakes.
The feeder's next movement reached the shelf as a dull vibration.
Protected.
Not isolated.
The feeder could not fit through the entrance.
Something smaller could.
Thin fungal growth already had.
Temporary protection.
Acceptable.
I moved onto the elevated shelf.
One edge was unusually straight.
A shallow groove crossed the stone beside it at a nearly constant width. Farther in, one section was flatter than the surrounding mineral.
Worked stone.
Possibly.
Mineral fracture could produce a straight edge. Water could carve a groove. Coincidence could arrange both beside a flat surface.
Coincidence had not yet exceeded its permitted budget.
I did not promote suspicion into history.
The shelf offered resources.
Gray fungus grew near the damp edge. Condensation gathered overhead and fell onto a mineral ridge one droplet at a time. A shallow depression near the back was large enough to withdraw into. The fissure entrance provided a view of part of the basin.
A dry patch near the flat section could support experiments.
The secondary groove crossed too close to all of them.
Laboratory.
Floodplain.
Both descriptions remained accurate.
I tested the gray fungus with one bite.
Bitter.
Soft.
No immediate burning.
Omnivore made it digestible.
It did not make the fungus kind.
I ate a small amount and waited.
No nausea beyond what I already owned.
I ate more.
The hunger weakened.
Strength did not return.
I drank condensation from the ridge.
The water was colder than the basin seep and carried less mineral taste. My foot produced a slightly more consistent mucus line afterward.
Slightly.
The front still hurt.
The rear still hurt differently.
My right lower tentacle continued shaking.
[Health Points: 1/1]
[Mana Points: 26/30]
Food.
Water.
No miracle.
Good laboratory conditions.
I organized the shelf.
The back depression became the resting hollow.
A white mineral flake marked the food-testing position.
Two gray fragments marked the path to condensation.
A scratch made with my radula marked the return route toward the fissure.
The flat mineral patch became the experimental area.
The fissure entrance became the observation point.
Christopher Aspen had once organized equipment by sterilization requirements, hazard category, and experimental priority.
Crypto Asper used fungus crumbs.
Standards had become more flexible.
I was still pleased.
Controlled space mattered.
Even when the control consisted of knowing which puddle was most likely to ruin lunch.
The secondary groove ran beside the food position, then curved below the resting hollow.
At its current flow, it provided water.
If the basin redirected more, it would cross the food area first, wash away the gray fungus, fill the hollow, remove my markers, and turn the experimental patch into a slide.
My laboratory came with automatic cleaning.
Possibly while I was inside it.
I needed warning.
A thin flexible fungal strip grew across one side of the fissure entrance. It bent when I pressed it with my shell, then returned.
Above it, a shallow notch cut across a narrow mineral projection.
Nearby lay a flat stone flake.
Strip.
Notch.
Flake.
Gravity.
I could build an alarm.
Not an intelligent one.
A useful distinction.
If movement bent the strip, the strip could release the flake. The flake could fall onto the shelf and create a vibration I could feel from the resting hollow.
It would not identify the intruder.
It might react to water.
A small creature might pass beneath it.
A large disturbance might destroy the entire mechanism.
Warning.
Not protection.
I pushed the flake toward the notch.
It slipped sideways.
I tried again using my shell.
The angle forced the lower edge away before the upper edge entered.
I needed temporary support.
[Mana Points: 26/30]
I placed the rear edge of my foot against the driest part of the flake.
Heat.
Pressure.
Adhesion.
[Mana Points: 25/30]
The flake stuck to me.
I used my shell to raise it and press one edge into the notch.
The fungal strip moved too early.
The flake tilted.
For one terrible moment, the adhesive held the wrong angle and began pulling me off balance.
I stopped the altered secretion.
Existing mucus remained loyal to the mistake.
I lowered the flake, shifted my shell beneath its center, and tried again.
This time, the upper edge entered the notch. The mineral projection supported the far side. I pushed until the flake rested mechanically between stone and the flexible strip.
The adhesive was no longer necessary.
Unfortunately, it was still present.
Condensation gathered above the projection.
I scraped a tiny channel with my radula.
One droplet entered it.
Then another.
The film reached the exposed adhesive edge.
Nothing happened immediately.
I waited.
The edge softened.
I peeled the rear of my foot free from one side, shifted my weight, and continued until the flake remained in place without me.
The alarm now depended on stone, fungus, gravity, and bad intentions from the universe.
No Mana.
Good.
I tested it by pushing the fungal strip with my shell.
The strip bent.
The flake slipped from the notch.
Tap
It struck the shelf.
The vibration reached my foot clearly.
Success.
Resetting it took three attempts and no additional ability use.
The final arrangement was less elegant than the first.
It also worked.
I retreated to the resting hollow and nudged the entrance strip using a loose gray fragment.
Tap
Still worked.
Anything entering through the fissure might trigger it.
Water might too.
I accepted the flaw.
Perfect detection systems required hands.
And equipment.
And a budget.
I had fungus crumbs.
After resting, I investigated the old adhesive residue visible on a mineral piece I had carried from the fissure edge. I pressed dead fungus against it and waited for the feeder's next movement.
A vibration crossed the shelf.
The dead fragment did nothing distinguishable.
No blue light.
No change in pulse timing I could separate from the surrounding reactions.
I repeated the observation once.
Same result.
The residue had prolonged physical contact when fresh.
That did not prove Mana conduction, fungal recognition, memory, or anything else interesting.
An experiment could fail without becoming useless.
It had protected me from inventing evidence.
I checked my Mana.
[Mana Points: 25/30]
No visible recovery.
Too slow.
Sleep-dependent.
Food-dependent.
Environmental.
Absent.
Unknown.
The answer remained hidden among unpleasant options.
I returned to the observation point.
The feeder still struggled in the basin. Its rear body shifted against torn growth. One hook found the edge, slipped, then struck again.
Alive.
Temporary remained temporary.
The fungal field beyond it pulsed through damaged routes.
I wanted one of those pulses to mean something.
That desire had become easier to admit and harder to trust.
Evidence of another mind would end the silence.
It could also mean the silence had been watching.
I withdrew from the entrance.
My body was transportation, instrument, specimen, construction equipment, and emergency tool.
Christopher Aspen had worked around his body.
Crypto Asper worked with his.
I looked across the shelf.
"My laboratory."
Calling it home felt premature.
Also worse.
Tap
The alarm fired.
I contracted into my shell.
Nothing entered the fissure.
The fungal strip had bent toward the shelf.
Water.
The secondary groove had widened.
More flow crossed the lower edge and pulled the strip far enough to release the flake.
Mechanical alarm.
Correct result.
Wrong threat.
I reset it, then followed the water.
The increased flow passed beneath the experimental area and entered a fungus-covered depression below the shelf.
Water gathered there.
Sediment shifted.
A mineral crust cracked along one edge.
A bubble escaped from beneath the fungal covering.
Then another.
The depression deepened as loose material washed inward.
I remained above it.
Gravity did the work.
Pressure helped.
I had contributed only poor housing decisions.
The fungal curtain over the depression sagged.
Water entered behind it.
The lower edge pulled inward.
A thick section collapsed.
Darkness opened beneath the shelf.
Two stone edges descended into it.
Parallel.
Between them, repeated grooves crossed the surface at even intervals.
One.
Two.
Three.
Four.
Mineral-blocked gaps lined both sides.
Water entered the opening and continued beyond my view.
A natural fracture could be straight.
Erosion could repeat.
Coincidence could do almost anything once.
This had been measured.
The cavern contained something built.