He spent the night thinking rather than sleeping.

His body worked on repairing itself with the accumulated Activity while his mind worked on the problem. The two processes ran in parallel, neither quite finished by morning, both making measurable progress.

The questions kept expanding.

Imperfect mutation. Vajra’s Activity ceiling. The gas-liquid-solid transition. What happened to other iron people at Level 1. Whether the Level system he’d built was even measuring the right thing.

He’d been confident about the Level framework since Chapter 45, comfortable with its limitations as a reference tool, satisfied that it captured something real about the hierarchy of Activity development. The overload event had introduced a variable he hadn’t accounted for: the iron-person pathway might have a threshold transition that the other mutation types didn’t face.

Or did they?

He ran through the different mutation categories methodically.

Plants: metallic bark transformation, Activity distributed through structural components, no Activity cores in the forms he’d examined except the Man-eater Peach Tree’s roots. How did a plant transition from Level 0 to Level 1? What would that look like for something rooted in place, unable to seek new resources, accumulating gradually through passive absorption?

Animals: evolved beasts with Activity cores going directly to solid-state. Half-beasts: same pathway, failed morphological integration. If they had no liquid phase, no Exchange feedback loop, no partner system to buffer transitions — did they experience a Level 1 crisis, or did the solid-state pathway avoid that specific failure mode?

Iron people: the delayed transition through liquid phase, the Exchange mechanism, the Lv1 overload. Had Wang Gen experienced this? Had anyone else survived it? Had people died from it without anyone understanding why?

He didn’t have the data to answer any of these questions. He needed to talk to someone who’d been through it.

The military had collected evolved people. The provincial military district had come to Juchao before the flood and taken most of the evolved population to Hefei. Those people had been there longer than him, with more resources, and presumably with someone doing systematic research rather than one person working with improvised equipment in an armored truck.

They might have answers. I need to go find them.

But he needed to resolve the immediate problem first.

The immediate problem: the Activity overflow would happen again.

Without the Bull Demon King Totem running, Vajra’s absorption rate had dropped to passive baseline. But passive baseline still accumulated, slowly and continuously. His calculations suggested approximately three to four days before the overflow threshold was reached again — at which point, without a managed discharge mechanism, the Exchange would force another uncontrolled pulse through him.

He had three to four days to figure out a solution.

The good news: at Level 0.954, his body was significantly more capable of absorbing a controlled overload than it had been at 0.836. The damage he’d just experienced would likely be manageable rather than life-threatening if it happened again in his current state. But manageable rather than life-threatening wasn’t the same as acceptable, and he wasn’t going to design a development strategy around repeatedly getting knocked unconscious by his own vehicle.

He needed to understand why it was happening.

The siphon analogy arrived with the sudden clarity of a well-framed problem.

High-side water flows continuously through a siphon tube toward the low side, as long as the differential exists. The flow doesn’t require pressure — it’s driven by the height difference alone.

Vajra: Level 1.000. Him: Level 0.954. Differential: 0.046.

The higher system is pushing toward the lower system through the Exchange connection. Not because of any active mechanism — just because the differential exists and the connection allows flow.

He’d been thinking about the Exchange as a deliberate exchange — Vajra offering Activity to him as part of the bond, a managed transfer. What the siphon model suggested was something more passive: the Activity simply moving down the gradient, from higher concentration to lower, the way heat moved from hot to cold or water from high to low.

If the model was correct, the solution was straightforward: eliminate the differential by bringing his own Level up to match Vajra’s.

I need to reach Level 1 myself.

At 0.954, he was 0.046 below the threshold. The overload event had pushed him 0.12 forward in a single event — brutal and damaging, but effective. If he could close the remaining gap in a more controlled way, the siphon pressure would equalize and the overflow would stop.

How to close 0.046 of Level difference without another catastrophic overload?

He didn’t have Honey Peaches. He didn’t have a high-Level evolved beast nearby to absorb. The Man-eater Peach Tree was thirty kilometers behind him. The Dog-Croc was inaccessible.

He had: the Bull Demon King horns, which were still Level 0.215 cores and the highest-intensity Active material in his possession. The partial crescent tusk remnants from the resonance experiments. The ant larvae, currently growing toward their pupation stage. The slowly-accumulating passive absorption from ambient Activity.

And he had time. Three to four days before the next overflow.

Let the passive absorption run. Let the body continue repairing. Reach the threshold before the overflow reaches the trigger point.

The math was tight. At passive absorption rates, he’d gain approximately 0.002 per day from ambient absorption. Four days was 0.008. That left 0.038 of the gap unaddressed.

Not enough on its own.

What else can I integrate?

The definition problem presented itself almost as a distraction — he was turning over the question of what would close his Level gap when a different question surfaced from underneath it, one he’d never fully confronted.

What exactly is Vajra?

He’d been measuring Vajra’s Level against Vajra’s total Active volume as the denominator. But he’d never precisely defined what counted as part of Vajra.

The truck frame — obviously Vajra. The original components — obviously Vajra. The modifications he’d added: the external armor, the radar system, the camera network, the infrared sensors — all Amalgamated, all functioning as parts of the integrated system, all contributing to the Active volume he’d been measuring.

The Baogai Armor — drawing on Vajra’s Activity, operating as an extension of the system, but not continuously Amalgamated. Reintegrated periodically.

The fang javelins, the crossbow bolts with Activity-core tips — Active when charged, passive otherwise.

The inflatable companions, the laptop, the USB drives, the RC car, the drone — inside Vajra physically but not Amalgamated. Not part of the Active system.

The ant larvae in their container — actively communicating with him through a shared Active bond, growing and developing, eating metal they were processing into biological structures that contained Activity.

Where was the line?

I’ve been measuring against a denominator I never defined precisely. The Level system assumes the denominator is stable. But Vajra’s definition changes every time I Amalgamate a new component or cut one free.

He sat with this for a long time.

If the denominator was unstable, every Level calculation he’d done was approximately correct but not precisely correct. The comparative rankings still held — a Fangwolf was still below a Bull Demon King — but the absolute values might be off.

More importantly: if Vajra’s definition could be changed, then the Level 1 threshold was a moving target. Adding more material to Vajra without Amalgamating it would expand the total Active volume denominator, dropping the saturation percentage even with the same absolute Activity quantity. Adding Amalgamated material would add both denominator and numerator, but the ratio might shift in either direction depending on the Activity Level of the added material.

Could I deliberately expand Vajra’s definition to manage the overflow?

If he added a large mass of non-Amalgamated material to the system, Vajra’s Level 1 state would become a Level 0.something state, and the siphon pressure would reduce accordingly.

What’s Vajra’s Amalgamation limit? Does it have one?

He looked at the question from every angle he could find.

The Amalgamation property had never failed to work on any material he’d tried it with — the difference was always speed, never capability. Metal was fast. Non-metal was slow. Complex electronics were slower still. But he’d never found something Vajra couldn’t eventually integrate.

Test the limit.

He didn’t have the strength to run field experiments today. His body was still in repair mode, his Active reserves still managing the damage from the overload. But he could design the test.

When he was functional again — in two days, three at most — he would start systematically testing whether Vajra’s Amalgamation capacity had an upper bound.

And if it did, that boundary might tell him what Vajra actually was.

(End of Chapter 56)