The two weeks it took to build the containment chamber were the two weeks during which the restricted archive's research team produced the most significant update to the Vorash picture.
The pre-mana-event construction records in the deep-history archive section confirmed what the megalithic arch's script notation had suggested: the arch was built using the same construction language as Chronos's temporal anchors, but in reverse application. The temporal anchors were built to preserve the framework's boundary by stabilising the temporal flow at specific geographic points. The arch was built to create a passage through the framework's boundary without destabilising it.
His parents had built a door as well as the wall.
The question of why was currently open. He had sent the documentation to Nyx's established communication channel and received: later. The research first. Which was the response of someone who had priorities and was applying them.
He accepted this and returned to the research.
The Obsidian Mastodon core was the problem and the solution simultaneously.
The problem: the core's mana density was operating at pre-framework levels, which meant the standard Valerian runic engineering's interaction with it produced feedback rather than integration. The framework's compression architecture and the core's uncompressed density were incompatible mediums.
The solution: if the incompatibility was total, then the core could not be channelled through standard runic structures, but it could still be used as a reference frequency. A device that was tuned to the core's specific density signature without trying to carry that density would produce a different interaction than a device trying to contain it.
He worked through this with the Dean of Spatial Geometry, who was, once his initial defensiveness had been addressed, an excellent collaborator — the specific kind of academic who had strong opinions because he had deep expertise and who updated those opinions when evidence required it.
"If we're not trying to channel the core's frequency but to mirror it," the Dean said, sketching on the laboratory board, "then the device is essentially producing a sympathetic resonance at the pre-framework density level without attempting to sustain that resonance. An echo rather than a carrier."
"And at the point where the echo meets an active gate anchor," Markus said, "the density differential between the echo and the anchor's internal structure—"
"Forces the anchor to resolve the differential internally," the Dean completed. He looked at the sketch. "The gate anchor has to handle the collision between its own internal frequency and the echo's external frequency. The anchor becomes the site of conflict rather than the target of external force."
"The anchor closes itself," Markus said.
"If the echo is precise enough," the Dean said. "This is a very specific calibration requirement."
"Yes," Markus said. "That's the difficult part."
The calibration work took four days. Markus and the Dean ran it together, the spatial sense's read of the core's density pattern providing the reference, the Dean's formation engineering producing the aperture blade array that would generate the echo. Elena observed and flagged three calibration errors that would have been significant problems in the field, which was why she was observing.
The trial ran on the fourteenth day.
The containment chamber worked. The core was suspended at the interaction point, the prototype device's aperture blades at the angles the calibration had established. The echo frequency was generated.
The Fate's Eye's read at the interaction point was exactly what the theoretical model had predicted: the core's pre-framework density meeting the echo's mirrored frequency in the containment field produced the specific internal stress signature that matched what a gate anchor would experience.
Not a real gate. The containment field's simulated anchor point. But the physics were real.
"If I'm reading this correctly," the lead researcher at the monitoring array said — a senior practitioner from the Spatial Geometry department, who had been working the analysis for twelve days straight and had the specific haggard precision of someone at the end of an extended sprint — "the anchor point is applying force against itself rather than against the device. The device isn't closing it. The anchor is."
"Yes," Markus said.
"We've built a device that convinces a gate to close itself."
"Essentially."
He looked at the lead researcher with the expression of someone who wanted to be sure the description was accurate rather than euphemistic.
"Essentially yes," Markus confirmed. "The echo creates a pressure the anchor's own physics have to resolve. The resolution is closure."
"That's either the simplest thing I've heard this month," the researcher said, "or the most complicated thing that happens to look simple."
"The calibration is the complicated part," Markus said. "The principle is straightforward once the calibration is right."
Elena and the Dean had been running the same analysis on the separate monitoring station and had arrived at the same reading. Elena's expression was the one she used for things that were genuinely correct rather than merely functional.
"Field test," she said.
"Field test," Markus agreed.
The Shattered Expanse was the target by the time the secondary units were assembled — three more devices, the calibration refined from the prototype trial, each one produced with the specific improvements the trial's data had identified.
The Expanse had the gate the Fate's Eye had been tracking through the Dominion's gate monitoring feed since the survey: Node Epsilon, the gate whose bridge coefficient had been rising at a consistent rate. Not yet permanent, but developing permanent infrastructure. The specific maturation stage where a field test would produce meaningful results — advanced enough to represent a real bridge-formation scenario, not so advanced that the test would require more force than the prototype's calibrated output.
He briefed Elena's three field researchers on what they were looking for: the gate's anchor point characteristics before and after the device's deployment, the specific signature of the self-closure process if it worked correctly, and the failure modes they needed to document if it didn't.
"We bring four devices," he said. "Three for the test, one in reserve. If the first deployment produces unexpected results, we stop, document, and return. We're not committing to field clearance until the prototype confirms the simulation was accurate."
Elena's expression indicated she was going to hold him to this.
"The team for the transit," he said. "Standard formation. Rosanne as commander."
Rosanne, who had been listening from the door, stepped in with the specific quality of someone who had been waiting to be addressed and had been patient about it.
"Briefing?" she said.
"Tonight," he said.
He looked at the four assembled devices, the simulation results on the display, the calibration logs the Dean had documented across fourteen days.
One thing at a time.
The principle was proven in the contained environment. The field would tell them if it held at real scale.
That was what field tests were for.