Anwen's Stage II was accelerating.

Dr. Ohen's message arrived through the official Bellhaven research registry -- not the provisional relay, the official channel -- which meant the message had been sent as medical communication under the trial protocol authorization, which meant it had been sent because the timeline required it.

Lucan read it in the research wing on Day 155, at seven in the morning.

Anwen's most recent imaging showed accelerated tissue progression. The Stage II to Stage III threshold, which had been projected at approximately twelve months from the trial application date, was now being revised downward. Dr. Ohen's estimate: the current progression rate suggested the threshold could be reached in approximately ten weeks.

Ten weeks. The Luminase-R compound was in transit stabilization. The Bellhaven registry review was in progress. The trial enrollment was three weeks away at minimum. The treatment itself required six weeks of compound administration. The timeline had no margin.

He messaged Dr. Ohen through the official channel -- the first time he had used the official channel since going off-grid, which was a risk and which was necessary: *Received. Is expedited enrollment possible?*

Dr. Ohen's reply: *Requires pre-authorization from the registry. Process is twelve to fourteen days minimum. I will file immediately. No guarantee on outcome.*

Twelve to fourteen days. Two weeks of process before enrollment could begin. Then six weeks of treatment. Plus the buffer for the progression rate. The arithmetic was possible but contained no room for error.

He sat with this.

Then he went to find Mirelle.

---

She was in the analysis room with the relay extension map -- she had been building the extension route to the northeast perimeter sector, the route toward the house they had been describing, doing the planning work because planning work was what she did with things she had decided were real.

He told her about Dr. Ohen's message. He told her the revised timeline. He told her about the expedited enrollment process.

She listened. She did not interrupt. When he finished she was quiet for a moment.

"The compound is in transit," she said.

"Yes. The stabilization was confirmed. It's moving through the relay network."

"And the formulation rebuild?"

"Still in progress. That's the neutralizer batch -- separate from the treatment compound. Anwen's treatment doesn't require the neutralizer."

"I know." She looked at the relay map. "If we move the research operation closer to Bellhaven, we compress the timeline on all of it -- the trial oversight, the treatment monitoring, Ohen's access to the research team. Less delay at every step."

He had been thinking about this. It was the logical response to the timeline compression. It was also a significant change in operational situation: moving the station's research operation toward Bellhaven meant increasing their visible presence in the network, which meant increasing their exposure to the tracking systems that had been looking for the station since Day 1.

"Moving closer means more exposure," he said.

"I know."

"Halden has been searching. The official channel message I sent to Ohen this morning -- I used the research registry credentials. That's traceable."

"I know that too." She looked at him. "The transit timeline has to come first. Anwen's timeline is what it is." She paused. "But before we talk about moving -- I want one thing clear."

He looked at her.

"No hidden results," she said. "I need that to be active, not theoretical. I need to know what you're working on, what you've found, what the problems are, in real time." She held his eyes. "I've noticed your attention is distributed. I don't know what you're distributing it to. I'm asking you to tell me."

He had the encrypted folder.

He had the scale-up model.

He had two days of working on the fungal scaffold model reconstruction from Mirelle's description of Vesti's drawing, which was showing early promise but was not yet confirmed.

He thought about what to say.

He said: "The scaffold model I started after the Ivara Pell call is showing early promise. I've been rebuilding the theoretical framework from your description of Vesti's work. Two more days to know if the structural model holds."

She held his eyes. "That's the alternative pathway."

"Yes."

"You've been working on it since the Ivara Pell call."

"Since before that. I started the alternative pathway search earlier." He paused. "The scaffold approach is the most specific lead I've had."

She looked at him. He could see her deciding whether the answer was complete. He knew it was not complete. He knew she was reading that it was not complete and he was watching her decide whether to press.

She said: "When you know the result of the scaffold model, tell me. Same day."

"Yes," he said.

"Same day," she said again, not repeating for emphasis -- confirming a condition.

"Same day," he said.

She nodded. She looked at the relay map. "I'll plan for the Bellhaven approach. We'll need to establish a new relay anchor point in the mid-sector range and I want Bren's opinion on the northwest exposure before we commit to any route."

She was already moving toward the logistics. She was letting the condition stand and moving forward and trusting that the condition would be honored.

He thought about the encrypted folder.

He thought: two more days. Two days on the scaffold model. If it holds, I will tell her everything -- the model, the scale-up result, the scaffold as the solution. At the same time. The whole picture, with the answer already available.

He thought: that is not the same as the same day. But it is almost the same day.

He told himself this was still within the condition.

He was aware that he was telling himself this.

---

Bren came two days after the Ohen message.

He came with supplies and with his northwest exposure assessment -- he had done a route survey in the days since Lucan had asked for his input. He laid out what he had found on the kitchen table while Mirelle made tea and Lucan ran the morning's synthesis monitoring.

The northwest corridor was passable. There were three Recovery Corps passive monitoring stations in the mid-sector range, but all three were on the standard tracking frequency -- the one the station had pulled from on Day 1. If they maintained the off-grid status and moved through the northeast rather than the northwest, the exposure was manageable.

"The northeast approach is longer," Bren said. "Three more hours of travel time to reach the Bellhaven outer perimeter."

"I know," Mirelle said. She had the route plan out. "The northeast approach uses infrastructure I've already mapped. I know every junction from here to the perimeter. Three more hours is three more hours I can account for."

Bren looked at the route plan. He looked at the detail of it -- forty-seven nodes, eleven relay points, the extension sketch in the margin. He said: "This is complete."

"Yes."

He looked at her. "You built this."

"Yes."

He was quiet for a moment. He had the specific quality of someone who was revising a previous estimate. "I knew you were running transit checks. I didn't know the scope."

"You didn't ask."

"No." He looked at the route plan. "I would have asked earlier if I'd known this existed."

She said: "Ask me now if you need anything."

He looked at the route plan for another moment. Then he looked at Lucan in the research wing visible through the partition. He looked at the two of them in their respective spaces and at the station around them and said: "You're both ready to move?"

Lucan said: "When the scaffold model confirms. Two days."

Bren looked at him. "What's the scaffold model?"

"Alternative synthesis pathway. For the scale-up." He paused. "It's promising. I need two days to confirm."

Bren looked at him in the specific way he looked at things that required further assessment. Lucan met the look. After a moment Bren said: "Two days." Not a question. An acknowledgment that the timeline was what it was.

"Two days," Lucan said.

Bren finished his tea. He went over the route plan with Mirelle for an hour, asking specific questions about junction capacity and relay load. She answered each question with the precision of someone who had built the system and knew it completely. He revised two points on the plan. She incorporated the revisions.

Before he left, he caught Lucan in the corridor.

"The scaffold model," Bren said.

"Yes."

"Two days."

"Yes."

Bren looked at him. "Whatever else is in that research, she needs to know it." He said this in the flat direct way of someone who had known Lucan for eleven years and was not asking a question.

Lucan said: "I know."

Bren left.

---

That evening she came to the research wing partition.

"Anwen's letter," she said. She held it out.

He took it. He read it.

The letter was two pages. Anwen wrote about the imaging results -- not the clinical language of Dr. Ohen's report but her own experience of hearing the revised timeline, which she described with characteristic precision: *I am not afraid of it, exactly. I'm afraid of missing things. There is a difference.* She wrote about the book series, which she was now two volumes into and which she was reading at a faster pace than usual, not from anxiety, she said, but from the specific pleasure of being inside something complete and wanting to see how it ended.

At the second page: *Mirelle told me about the house. The light room with the stone floor. I want to visit it. I want to sit on the stone floor in the transit hour and have her describe the light the way she describes it when you do it together. I am going to hold this as something that is going to happen. Tell me that's allowed.*

He set the letter down.

He looked at the research wing. He looked at the synthesis files. He looked at the encrypted folder on the file system, which he had not opened in six days while he worked on the scaffold model reconstruction.

He thought: she is sitting in a corridor adjusting her schedule to follow the changing angle of the light. She is reading faster because she wants to see how things end. She wrote *tell me that's allowed.*

He thought: the scaffold model has two more days. Then it has a result. Then I tell Mirelle everything at once -- the model, the scale-up, the scaffold as the answer. Then we move.

He wrote back to Anwen. He said: *You'll visit the light room. You'll sit on the stone floor. That's allowed.*

He sealed the letter. He put it with the relay envelopes.

He went back to the scaffold model.

It was Day 155. Two more days.

---

The two days passed.

Not easily. He had been working on the scaffold model reconstruction for six days total -- from the end of the Ivara Pell call, through the Bren visit, through the evening with Anwen's letter. The reconstruction was built from Mirelle's description of Vesti's drawing, which was a twenty-two-year-old memory of a sketch made by a junior researcher who had been building an alternative synthesis approach in year two of a program that had been trying to make the approach unnecessary.

The memory was detailed. More detailed than he had expected -- she had been nine years old, and she had been shown the structure once, but she had the specific retention of someone who had been in an environment that required her to remember everything precisely because she had no other way to protect herself. She had looked at the drawing and stored it as information that might matter. Twenty-two years later it was the most complete record he had of what Vesti had been building.

He was working from a description. He was not working from the original data. The model he was rebuilding was an approximation -- a theoretical reconstruction of a research pathway that had been abandoned and whose original documentation had been deleted.

On the second morning, he ran the scaffold model's first complete structural simulation.

The result showed a synthesis pathway that was theoretically viable.

He ran it again.

The second run showed the same pathway.

He ran three more structural simulations from different starting parameters. All five returned a viable pathway. The yield figures were consistent: significantly lower than the extraction-based scale-up model, sufficient for a treatment population of several hundred in the initial clinical period, scalable to the low thousands with optimized production.

Not the hundreds of thousands the original scale-up model could produce.

But viable. And not fatal.

He sat with the five simulation results.

He thought: this is the same day. I said same day.

He printed the scaffold model results. He took the printout to the analysis room.

She was at the table. She looked at him when he came in, read his face, and waited.

He set the printout on the table.

"The scaffold model is viable," he said. "The theoretical pathway holds. Five structural simulations, consistent results." He paused. "The yield is lower than the extraction-based model but sufficient for clinical trial scale and the first treatment population."

She looked at the printout. She picked it up.

He said: "There's more."

She looked at him.

He went back to the research wing. He opened the encrypted folder. He brought the scale-up model to the analysis room -- the full file, not a printout, open on the analysis room display so she could see all of it.

He said: "This is the scale-up model I built after the synthesis confirmation. The extraction-based model." He stood beside the display and did not sit down. "Page four is the yield calculation. The extraction-based approach is sufficient to produce seed culture for mass treatment. It is not a sustainable or renewable process. The biological source required is--" He stopped. "I want you to read it. I want you to read it yourself and I'll answer any questions after."

She looked at him. She looked at the display.

She read page four.

He watched her read it. He watched her face and he had been watching her face for months and he could read it now, accurately, the way she had been reading his -- the specific movement when she understood a result, the stillness when the result was significant, the particular quality of attention when she was making sure she had understood correctly.

She read it twice.

She set the printout of the scaffold model next to the display showing the extraction model. She looked at both.

She said: "You had this before I asked you to move."

"Yes."

"You had this when I asked you no hidden results."

"Yes."

He stood with this.

She said: "When did you build this model?"

"Day 128. The night the synthesis compound was confirmed viable."

She looked at the display. "Day 128 is the same day you told me the compound was viable."

"Yes."

"You confirmed the compound in the afternoon. You told me. We celebrated. I made dinner." She paused. "And then that night you built this model and read page four and created an encrypted folder."

"Yes."

She was quiet for what felt like a long time. He did not fill the silence. He had told himself, for weeks, that the full picture with the answer would be better than the problem alone. He was standing in the analysis room on Day 157 with the full picture and the answer available and he was understanding, for the first time in a direct way, that the full picture and the answer did not change what had happened in between.

She had asked him no hidden results before any of this. She had asked him on Day 155, during the Bellhaven move discussion, and he had told her: *same day.* He had not told her the scale-up model already existed. He had confirmed a condition he was already failing to meet.

She said: "You told me same day."

"Yes."

"And you had this for twenty-nine days before that conversation."

He could not make that different. "Yes."

She stood up. She went to the Surface Room.

He stood in the analysis room with both models on the table and he did not follow her.

---

He stayed in the analysis room.

He was aware of the specific quality of this -- staying, not following, not explaining, not entering the Surface Room uninvited. He had learned this over months. She would come to him when she was ready or she would not come to him, and either of those outcomes was hers to determine, and the right thing he could do right now was stand with the result of what he had done and not add to it by requiring her to manage his need to explain.

He did not want to explain. He wanted to tell her the truth of it, which was different from explaining -- explaining had a justificatory quality, a structure of reasons leading to a conclusion. The truth was simpler and worse: he had been afraid. He had known the model's result and he had been afraid of what she would do with it and he had hidden it, and when she had asked for the condition he had confirmed it while already violating it, because the fear was still there and the fear had won.

He had not examined the fear this directly before. He had dressed it in other language: *protecting the integrity of the research process, waiting for a complete picture, not raising a problem without having a solution.* All of those things had also been true. None of them had been the primary reason.

The primary reason was that he could not watch her choose.

He had known this since he had closed page four on Day 128 and had not wanted to name it. He knew it now. He had decided for sixteen years what other people's lives were worth in the service of a research goal, and he had told himself this was different, and then he had done the same thing again because the person it involved was one he loved and that had made the fear larger rather than smaller.

He stood in the analysis room and did not move and held this.

After a long time, the Surface Room light went off.

He heard her footsteps in the corridor. Not toward him -- toward her room. The door.

Closed.

He went to the research wing. He did not open any files. He sat at the workstation in the dark and thought about the house and about Anwen reading fast to see how it ended and about twenty-nine days and about what same day meant and what he had said it meant.

In the morning he would tell her everything else he had been holding. All of it. Not strategically sequenced, not held until the answer was available. All of it, in the morning, because she had the right to know everything and he had the right to nothing more than the choice to tell the truth.

He would start with what he should have said on Day 128.

He sat in the research wing in the dark and waited for the morning.