The Shattered Expanse's volcanic landscape managed, against the competition, to be bleaker than the Iron-Root Glade, the Gale-Glass Desert, and the Toxic Swamp. The ash-choked blizzard was a detail that the local environment had clearly developed a commitment to across whatever geological period had produced this specific combination of volcanic activity and northern latitude.
The grade-V gate at the valley's centre was visible through it.
The same conversion pattern he had been reading across the Dominion's gate network, only here at the maturation stage the survey had flagged as the optimal test window. The bridge coefficient had been confirmed by the monitoring feed that morning: the gate was developing permanent foundation architecture without having completed it. The Catalyst would be engaging the anchor at its most readable and most destructive.
Terros drove the sealing pikes with the efficiency of a Tier 7 practitioner who had been working with formation engineering for fifty years. The confinement matrix that spread across the crater floor was the specific formation type Terros had developed in the second year of the formations course — the fourth-anchor solution, extended to field scale, anchoring the gate's oscillation into a rigid temporary stasis rather than allowing it to continue its conversion cycle during the team's approach.
"Two hours to matrix failure," Terros said. He was not frantic about it. It was a schedule constraint, and schedule constraints were managed.
"We'll need twenty minutes," Markus said.
The Swiss Guard unit moved into the gate in the diamond formation they used for hostile-interior entry — three shields at the front, three strike-position practitioners behind, the Captain managing the tactical cohesion. Markus entered at the formation's centre, which was the correct position for a practitioner whose primary contribution was the spatial sense's read and the Fate's Eye's targeting.
The interior was the specific character of a gate that had been absorbing Vorash's conversion architecture for approximately eighteen months by the maturation estimate. The standard dungeon parameters had been partially overwritten — the tier-classification mana distribution the framework enforced was giving way to the denser, more chaotic pattern that the conversion produced.
The Blood-Gorged Behemoth was the gate's sovereign, and like the Abyssal Dread-Fiend, it was a product of accumulated density rather than natural development.
The Fate's Eye mapped it in the first four seconds.
"Primary load-bearing node three centimetres beneath the central clavicle plate," Markus said, through the team channel. "It's what the conversion architecture is anchored to in the biological structure."
The Captain registered this and integrated it into the Guard's formation. They were good — the specific quality of Tier 6 practitioners who had been operating together long enough that targeting data arrived as a resource rather than as an interruption. The strike came at the moment the formation's positioning gave the Captain the angle the targeting required.
The Behemoth was resolved in two minutes forty seconds.
Not a record. A professional engagement against a genuinely difficult opponent, executed by practitioners who were very good at their specific work.
What replaced the sovereign's collapse was what the research had predicted and the field needed to confirm: the gate's conversion process, interrupted by the boss's fall, completing the transition to its next stage. The crimson boundary went dark, the primordial stone gateway's arch manifesting from the dimensional fabric as the barrier between the framework and the pre-framework space dropped to its thinnest point.
Through it: the triple suns and iron-ore mountains.
Consistent with the Sector 12 arch.
"Recording array," the Captain said, and the three shield practitioners produced the data-capture prisms and positioned them at the angles the laboratory team had specified in the pre-deployment brief.
Markus lifted the prototype from its armoured case.
The aperture blades, calibrated to the Obsidian Mastodon core's density signature over two weeks of laboratory work, registered the arch's pre-framework frequency the moment they came within range and began their rotation.
He placed it at the arch's threshold rather than throwing it — the specific placement position the simulation had identified as producing the cleanest interaction geometry. The device needed to be at the point where the two dimensional environments met, where the density differential was highest.
He stepped back.
"Deploy," he said.
The Catalyst did not produce a visible effect for approximately four seconds, which was within the simulation's predicted range for the anchoring process to begin.
Then the arch began to close inward.
Not from external pressure — from its own structural response to the anharmonic echo the device was projecting. The Fate's Eye read it clearly: the arch's anchor points were experiencing the pressure the simulation had modelled, the primordial density meeting the echo's mirror frequency at the anchor points rather than at the device, the gate applying the force to itself.
Eleven seconds from deployment to complete closure.
The recording prisms captured the millisecond data across the full sequence.
The cavern, when the arch was gone, was stable — the spatial field reading as clean, the gate's conversion architecture absent without the anchor maintaining it. The local dimensional geometry was ordinary.
"We need the prism data analysed before we call it confirmed," Markus said. "The field read looks clean but the simulation comparison is what tells us whether the closure sequence matched the model or achieved the same result through a different mechanism."
"Understood," the Captain said. "Extract?"
"Extract."
The briefing ran for three hours.
Elena, Terros, the Dean of Spatial Geometry, and the lead research team worked through the prism data against the simulation model. Markus watched the analysis and answered questions where the Fate's Eye's field read added information the recording hadn't captured.
The field result matched the simulation model.
Not perfectly — there was a three-second variance in the closure timeline that the simulation hadn't predicted, attributable to a density differential between the laboratory's reference material and the actual arch's anchor characteristics. The variance was documented and the calibration adjustment identified.
The device worked.
"Twelve units," Markus said, when the analysis was complete. "The manufacturing specifications have been updated with the calibration adjustment from the field data. The two-week production window holds."
"We can have the first four units in ten days," Elena said. "If we run three shifts."
"Run three shifts," he said. "The Dominion's remaining compromised gates need to be addressed before Vorash identifies the Trench as a closed channel and reconfigures his approach."
He looked at Valerian.
"The gate network problem is solvable with the Catalyst," he said. "The Vorash problem is not. Closing his harvester gates removes his mana supply from this side, but he has thirty years of accumulated reserves. Whatever he's building in the citadel has thirty years of material in it."
He let this be present without dramatising it.
"We have a tool for the harvester gates," he said. "We don't yet have a tool for the citadel. That's the next research question."
The room held the weight of this accurately.
"The archive records on Chronos's construction language," Valerian said. "The arch I accessed was built with the same notation as the temporal anchors. My parents built a door as well as the wall. Understanding why they built it is part of understanding what Vorash's citadel is actually doing."
"The archive team has been working that thread since you returned," Elena said. "They have preliminary findings. It's going to require your attention this week."
"Tomorrow," he said. He looked at the twelve-unit production schedule, the calibration adjustment documentation, the prism data logs. "Tonight the team debrief, then sleep. The archive findings tomorrow."
Rosanne, from the doorway, gave him the expression she used when she had been waiting to say something and was deciding whether now was the moment.
"The team debrief can be brief," she said. "We held the perimeter and watched you hand the Captain targeting data like a map coordinate. The debrief is: it went correctly."
"That's accurate," Markus said.
"Good," she said. "Then sleep."
He went.