Five weeks after Sylvia first noted the formation network's improving coherence, she brought the full measurement set to the north-face alcoves.
It was the first time all four of them had met there. Ryan had confirmed the location was clear five days earlier — no monitoring formations above the campus's third terrace level, and the north face's orientation meant the organisation's east-forest array couldn't reach the angle. They had arrived separately, at different times, through different paths, because Ryan had explained the timing with the particular patience of someone who understood that boring was a survival technique.
Sylvia spread the notebook pages across the flat stone of the alcove's meditation surface. She had been taking measurements every morning for five weeks — the wave pattern of the formation network's pulse, the coherence coefficient at each major node, the rate of improvement.
The charts told a clear story.
"The tower node is at approximately eighty-three percent of its theoretical original capacity," she said. "The east forest node is at sixty-one percent. The south campus node — the one beneath the main cultivation hall — is at fifty-two percent." She paused. "The nodes are improving in the order Aaron worked near them, and at rates proportional to how directly he worked with them." She looked at Aaron. "The bridgehead at forty-eight percent has improved the entire network by approximately ten to fifteen percent of its capacity."
"What happens at sixty percent?" Blake asked.
"The bridgehead starts generating an exterior field," Aaron said. "Subtle — but detectable by a spatial mage who knows what to look for." He looked at Ryan. "Which is where the still-shadow becomes load-bearing."
"I have passive maintenance working," Ryan said. "I've been holding it for six days at full passive — no active direction required. The field covers approximately a four-metre radius around me at current." He paused. "Extending it to cover the tower's exterior foundation will require a significantly larger radius." He said this without apology — it was a technical description of a gap, not an excuse.
"How large?" Sylvia asked.
"The tower's base circumference is roughly sixty metres," Ryan said. "At current, I can generate passive still-shadow to about twelve metres. Scaled to sixty metres —" He paused. "My mage force reserve is the constraint, not the technique. The technique is stable. The volume is the problem."
"How long to develop the mage force reserve?" Aaron asked.
"At my current cultivation rate, rank three advancement will occur within the next six to eight weeks," Ryan said. "Rank three increases my mage force reserve substantially. Combined with the technique —" He paused. "Six to eight weeks."
"The crossing attempt," Aaron said. He had been tracking the disturbance pattern with Ryan. "The acceleration has been consistent. I estimate five to seven weeks until the next event. Ryan's estimate is similar."
"So the still-shadow coverage and the crossing attempt are approximately simultaneous," Sylvia said.
"Yes," Aaron said.
Blake was sitting slightly apart from the group, which was his usual position — not exclusion, just his particular mode of being present. He had the formation paper from their cultivation garden conversation in his bag. He had shown Instructor Vareth. She had asked him three questions about the fourth layer, and he had answered all three correctly, and she had told him to document it for the advanced formation project in the way that meant she was taking it seriously. He had not mentioned this to the group. He mentioned it now.
"Vareth's asked me to demonstrate the lattice in the practical assessment at the end of the semester," he said. "Which means I'll be showing it in front of the full fire element class and likely one or two assessors from outside the academy."
"Association assessors," Ryan said.
"Possibly," Blake said. He said it carefully. "Vareth mentioned that the Association's regional assessment committee does the end-of-year practicals."
Everyone was quiet for a moment.
"You don't have to," Sylvia said.
"I know," Blake said. "But pulling the demonstration now looks worse than going ahead with it. If someone's been watching, a sudden reversal is more suspicious than a student demonstrating a technically sound formation to an assessment committee." He paused. "And the formation is technically sound. There's nothing about it that should attract attention from anyone who doesn't already know to look." He looked at Aaron. "Unless you know something about it that would change that calculation."
Aaron thought about this carefully. The lattice formation was a fire-element innovation — nothing spatial, nothing related to the bridgehead or the Greensiders or the monitoring formations. There was no logical connection between Blake's independent formation work and anything the organisation was watching for.
"It's clean," Aaron said. "Demonstrate it."
Blake nodded. He picked up the formation paper and looked at it. "The fourth layer still bothers me," he said. "I know the compensation is correct. I've tested it. But Vareth asked about the connector angles at the node vertices and I don't have a rigorous mathematical explanation for why I chose those angles — I chose them because they felt right."
"Felt right how?" Sylvia asked.
"They looked like the heat pattern in the core burns cleanly at maximum output," Blake said. "I've been watching the heat pattern in the academy's fire training hall for three months. The distribution when the fire is burning at full output versus the distribution when it's being forced — they're different. Clean fire and forced fire don't look the same." He paused. "The angles in the fourth layer match the clean distribution, not the forced one."
"You're matching the formation geometry to the element's natural expression," Sirath said, through Aaron. "Not to maximum output, but to optimal function." Aaron looked at Blake. "That's not instinct — that's observation. The mathematical explanation is the angular distribution of the element's natural combustion geometry. It can be derived. It's simply that most fire-formation designers work backwards from desired power output rather than forwards from the element's preferred state."
Blake looked at Aaron for a moment. He had encountered the Sirath-through-Aaron communication twice before and had adjusted to it with the pragmatic lack of fuss that characterised his approach to most things that couldn't be helped.
"Tell him I've been looking for that explanation for two months," Blake said.
"I heard," Sirath said, through Aaron. "Tell Blake his approach — working forwards from the element's natural geometry rather than backwards from output — will, at sufficient cultivation level, produce formations that enhance the element's function rather than simply directing it. Most fire formations push the element. His are beginning to invite it."
Blake absorbed this in silence.
"That's what I've been trying to do," he said, after a moment. Not surprised. The tone of someone who has been moving in a direction without a name for it and has just been told the name.
Sylvia had been watching both of them with the particular expression she used when cataloguing information about people. She noted it and moved on.
"The network's coherence improvement has an implication I raised with Aaron three weeks ago," she said. "As the bridgehead advances, the network's recording capacity increases. The formation has been active for three thousand years — it holds impressions. Events with sufficient energy that they left a mark on the formation layer." She looked at Aaron. "At eighty percent network coherence — which occurs when the bridgehead is fully prepared — those impressions may be readable."
"By your Life element," Ryan said.
"By my Life element and Aaron's spatial perception together," Sylvia said. "The Life element reads the formation's living pulse — the quality of the impression as a retained signal. The spatial element can navigate the network's carrier layer — read the location and structure of what the Life element identifies." She paused. "We don't know what's in the impressions. But events significant enough to be retained would include: the Shattering itself, if the network existed at the time —"
"It did," Aaron said.
"The establishment of the Tenders. The first Greensider contact — Aela Voss. And —" She said the next part carefully. "Events in the period before my uncle and father disappeared."
Nobody filled the silence that followed.
"It's not certain," Aaron said. "The records may be too degraded. The resolution may not be sufficient. But at eighty percent coherence, we look."
Sylvia nodded. She did not have more to say on this point. It was the kind of information you either had or you didn't, and you didn't manufacture it with additional words.
"The crossing attempt," Blake said. He had put the formation paper away. "Five to seven weeks. What does it actually look like when it happens?"
"A disturbance in the formation network," Aaron said. "A boundary-crossing attempt from the Greensiders' side that we previously read as spatial instability. Now we know it's a deliberate push — they're trying to cross, and each attempt is more coordinated than the last." He paused. "When the bridgehead is complete, the network will respond to their attempt differently. Instead of reflecting it, it will receive it. The bridgehead formation creates a stable receiving point on our side — the Greensiders push through, the bridgehead catches them, and the crossing completes."
"What happens to the person crossing?" Sylvia asked.
"I don't know precisely," Aaron said. "Sirath's design has never been field-tested. The traveller would be — carried by the formation. Their spatial coherence is maintained by the bridgehead's structural support. It's the difference between jumping across a gap and walking across a bridge." He paused. "In theory."
"In theory," Blake said.
"In theory," Aaron confirmed. "I won't pretend it's been tested."
Blake looked at the mountain. "And if it goes wrong."
"If it goes wrong, the traveller is dispersed in the space between Shards," Aaron said. "Which is not recoverable." He said it plainly.
Silence.
"The Greensiders know this," Sylvia said.
"They've been attempting it for three hundred years," Aaron said. "They know."
Another silence.
"Then we make sure it doesn't go wrong," Blake said. His voice had a quality it sometimes got — not performance, not posture. The real version of his certainty, the one that had nothing to prove. "What do we each need to do when the attempt happens?"
Aaron looked at him.
Then at Ryan.
Then at Sylvia.
He told them.