He spent the first hour after lunch working through the theoretical framework before running another experiment.

The Activity resonance field, as best he could characterize it: high-speed frequency resonance liquefied the Activity core material, releasing the locked Activity and creating a field around the core — an outward-radiating structure that concentrated ambient Activity from the surrounding environment. Similar in behavior to a magnetic field, with the distinguishing feature that it actively pulled in external energy rather than simply maintaining a static configuration.

The failure mode he’d been hitting was a closed-field problem.

A magnetic field had two poles — north and south — with field lines that ran from one to the other, forming closed loops. Energy entered through one pole and exited through the other, maintaining a stable circulation. The single-core fields he’d been producing were open: field lines extending outward but with no return path, no circulation, no stability. Energy built up in the core until the system failed catastrophically.

What he needed was a second core to complete the circuit.

North and south. Source and return. Two cores, two poles, field lines running between them in closed loops.

He had a working hypothesis. He had a large supply of low-level Activity cores he’d never found applications for. He started testing.

Three-core triangular configuration first — the triangle being geometrically the most stable shape. Three Short-sting worker ant mandibles, selected for matching frequency signatures and closely aligned Level values.

He ran the frequency acceleration and watched through his perception.

Boom. Boom. Boom.

Three simultaneous failures. No indication of field closure.

He repeated the test more carefully. Same result. A third time.

Triangle model: non-functional. No closure behavior observed across three attempts.

He switched to the dual-core model.

Two poles, not three. If the analogy to magnetic field structure held, two cores were the minimum required for closure. Three might be overdetermined — too many poles for the field geometry to resolve into a stable configuration.

Two Short-sting worker ant mandibles.

Frequency acceleration. Surface deformation beginning. And then — in the instant of failure, in the last fraction of a second before the explosion — the two shockwave fronts interacted. Briefly, incompletely, but visibly through his Active perception: the released energy from each core reaching toward the other rather than simply dispersing outward.

Entanglement. Not the right word technically, but the best available approximation for what he was seeing.

The field had tried to close. It hadn’t sustained long enough, but it had tried.

This is the right structure.

He worked through his Active core collection systematically.

Short-sting soldier ant dual-core: same entanglement effect at failure, slightly more pronounced.

Gnawrat incisor dual-core: stronger entanglement, lasting almost a full second before the explosion overwhelmed it.

Teddy Tyrant fang dual-core: entanglement occurring during the surface liquefaction phase, before failure — the closure attempt happening earlier in the process rather than only at the terminal moment.

Fangwolf fang dual-core: entanglement sustained for a full minute before catastrophic failure. The closures were almost complete, holding together and then breaking apart repeatedly in the seconds before the explosion.

By evening, every dual-core experimental run had produced the same pattern: entanglement behavior that became more sustained and more complete as the Level of the cores increased. The failure threshold was a function of the core material’s capacity to sustain the resonant load, and the capacity scaled with Level.

The progression was clear.

At Level 0.215 — the Bull Demon King horns — the capacity should be sufficient.

He had one pair of horns and no other material at remotely comparable Level. He spent ninety minutes reviewing every piece of experimental data he’d collected, running the theoretical model against the observed entanglement durations, checking the scaling relationship.

The model predicted that at Level 0.215, a dual-core field should be able to close and sustain indefinitely once the closure was established.

Should be.

He noted the uncertainty, acknowledged it, and started eating dinner.

Post-dinner. Mental state review: concentration adequate, fatigue manageable, emotional baseline calm.

He logged the experiment parameters before beginning:

Activity Resonance Field — Dual-Core Model, Trial 6: Materials — Bull Demon King Activity cores, horn pair (Level 0.215). Purpose: achieve and verify stable closed resonance field.

Then he set up the horns.

The frequency acceleration began. At this Level, the response was immediate and more organized than any of the lower-Level tests — the surface liquefaction coming earlier, the characteristic iridescent coloring he’d first seen with the crescent tusk appearing within seconds, the active material flowing across the horn surfaces in visible patterns rather than simply deforming uniformly.

He guided the frequency carefully.

The two horns’ resonance signatures locked together. Not the brief, fragmentary entanglement of the lower-Level tests — a genuine coupling, each horn’s field lines reaching toward the other and finding purchase. The closed loop began to form.

Field lines from one horn to the other. Return lines from the other back to the first. A circulation pattern, completing itself, the Activity within the system cycling rather than building toward failure.

He cut the external frequency acceleration.

The moment of maximum uncertainty.

He brought his arms up over his head, tucked his vital points behind the Baogai Armor’s coverage, and waited for the explosion that might or might not come.

It didn’t come.

The closed resonance field held.

Through his perception, he could see it: a sphere approximately one meter in radius centered between the two horns, field lines running in stable closed loops, the Activity within it cycling continuously without accumulating toward any failure threshold. The resonance maintained itself once established. The external drive frequency had been a catalyst, not a sustaining requirement.

And from the surrounding environment: ambient Activity converging toward the field.

Not slowly, the way passive absorption worked. Not in the burst pattern of the post-explosion pulses. Continuously, steadily, a sustained inflow that showed no sign of stopping — ambient Activity from a radius considerably larger than one meter streaming toward the resonance field and being absorbed by Vajra through the Exchange connection.

Vajra’s reaction came through as something close to joy.

One minute: Level 0.560, plus 0.0002 gained.

Five minutes: Level 0.560, plus 0.0005 total.

The rate was modest — the ambient Activity concentration in the area was still relatively low after the weeks of decline since the last storm. But it was continuous. And continuous, sustained accumulation compounded.

He stayed in position and watched the numbers tick upward.

(End of Chapter 49)