The testing proposal appeared on the kitchen counter on a Wednesday morning, which was how Lucan delivered things that required a decision: in writing, at the place she was most likely to encounter it without it being a confrontation.

She read it over tea.

The document was three pages, printed on the station's small printer in his compact handwriting -- not from the system's template, which told her he had written it out rather than using any standard form. It outlined a series of Vesperin dispersion measurements: controlled distance testing at intervals from eighty centimeters down to thirty, with protocol, safety considerations at each stage, and a third column she read twice. The column was headed *Alternative approach if subject declines* and contained, for each measurement point, a specific alternative methodology that would produce a partial result without requiring her presence.

She set the document down and drank her tea.

He had prepared for her refusal. He had not done this once, not just for the significant decisions -- he had done it consistently, across every request he had made of her since the Terms conversation, every time he needed something from her body or her knowledge or her time. He built in the refusal. He treated it as a real option rather than a formality.

She understood, now, that this was not a research methodology. It was how he thought about people. The refusal was a real option because to him it was a real option, because the arrangement they had made was real, because the conditions she had set were conditions he had actually internalized rather than memorized.

She went to the research wing.

He was at the workstation. He looked up when she came in.

"I'll do it," she said.

He nodded. "Whenever works for you."

"This afternoon," she said. "After lunch."

---

They set up the testing in the examination room, which had the most controlled dimensions and the monitoring equipment already calibrated for Vesperin baseline readings. She stood at the far end. He stood at the entrance, eighty centimeters away, with a portable monitor clipped to the wall beside him.

"Baseline?" he asked.

"Standard." She ran the internal check she had developed over weeks of being around him -- a deliberate inventory of her own physiological state that she had learned to conduct quickly and accurately. Pulse rate, respiratory depth, Vesperin production at resting level. "Eighty centimeters: baseline stable."

He made the note. Moved to seventy centimeters. "Seventy."

"Stable."

Sixty. Fifty. Each measurement confirmed, noted, anode moved. The room was small enough that by the time they reached forty-five centimeters the quality of the space had changed -- not Vesperin, not threat response, something else that she was aware of and was treating as a separate variable for now. The monitor showed her baseline stable. Her internal assessment confirmed it.

"Forty-two," he said, moving to the measurement.

"Stable." She paused. "This is the distance from the power reduction."

He looked at the monitor. "Yes. The Surface Room."

"And it's still stable."

"Yes." He wrote the measurement. He did not make anything of it. "Forty," he said, and moved.

She ran the check. Forty centimeters: stable. She reported it.

Thirty-eight.

She ran the check more carefully at thirty-eight, because thirty-eight was the verified distance from the Surface Room, which had been stable, and she wanted the measurement to be accurate rather than approximate. Stable. She reported it.

Thirty-five.

She knew what would happen at thirty-five. She had felt it in the Surface Room testing -- the small elevation, the specific quality of it. She ran the check and found it: Vesperin baseline with a minor upward pressure. Not a spike. An elevation. She reported: "Minor elevation. Not a spike. Increase of approximately three to four units from baseline."

He wrote it. "That's the anomaly from the first testing series."

"Yes."

He looked at the monitor, which was showing the same thing. "The anomaly is consistent. That's meaningful." He looked at her. "Do you have a hypothesis for what's causing it at thirty-five that isn't present at thirty-eight?"

She had a hypothesis. She had had it since the first testing series. She had been choosing not to share it because sharing it required saying something out loud that she was not yet certain she wanted to say out loud.

"I have a working theory," she said.

"Which is?"

She looked at him across the thirty-five centimeters and thought about the question. She thought about the Surface Room and the transit maps and the notebook with light observations and the kettle that had appeared on the shelf. She thought about the corridor at two in the morning and the thirty-two names written in a small brown notebook.

"The anomaly isn't threat response," she said. "Threat response produces a full spike, bilateral sympathetic activation, the standard Vesperin cascade. This is different -- it's a unilateral elevation, low and sustained rather than acute. The mechanism is different."

"Different trigger," he said.

"Different trigger." She paused. "The standard Heliogen documentation says Vesperin is a defense mechanism. It's not. The defense mechanism was layered on top of something that already existed. The original mechanism was designed for a different activation trigger."

He looked at her. He was very still in the specific way of someone who has received a significant piece of data and is processing its implications before responding.

"What was the original activation trigger," he said. Not a question. The tone of someone completing a sentence that had already been written.

She met his eyes across thirty-five centimeters.

"I don't know," she said. "I'm theorizing based on differential response pattern. I haven't confirmed it."

He held the look for a moment. Then he looked at his notebook and wrote something, and she could not see what he wrote from this angle, and he closed the notebook.

"Let's move to thirty," he said.

She was almost grateful for the clinical redirect. Almost.

"Thirty," she confirmed, and they moved.

---

At thirty centimeters, Vesperin spiked.

Not catastrophic -- the spike was rapid but stayed below the threshold that would have produced the involuntary toxin discharge she was capable of. She reported it immediately and stepped back to thirty-five, and the spike resolved within twelve seconds of the increased distance. He checked his vitals. Nothing. She had controlled it.

"Thirty is the threshold," she said.

He was writing. "For today's testing, yes. The threshold may vary with physiological state." He looked up. "Are you all right?"

"Yes. The spike resolved cleanly."

"The resolution time -- twelve seconds?"

"Approximately."

He wrote it. He was being very precise and very thorough and she was aware of this as deliberate -- the precision was the care, the same way every time. He noted the resolution time and the threshold distance and the conditions, and when he had finished he closed the notebook and looked at her directly.

"Thank you," he said. "This is -- this is very useful data."

"I know."

"The differential response pattern you described. If the original mechanism was designed for a different trigger, that changes the neutralizer research substantially."

"I know that too." She leaned against the examination room wall. "The Heliogen researchers didn't know it. They had thirty-two subjects and they approached the Vesperin as a uniform defensive system without investigating the original mechanism layer. The neutralizer they were developing targeted the wrong stage of the cascade."

He looked at her. "You worked this out yourself."

"I had years to think about it. And I had access to Subject Eleven's notes for a short period before they were destroyed." She paused. "Eleven was more technically trained than the others. She understood what was being done to us better than I did at the time. She left notes in places the facility staff didn't think to look."

He sat quietly for a moment. Then: "Those notes -- do you still have them?"

"No. They were destroyed. But I remember them." She looked at him. "I can tell you what I remember. It's not complete, but it's more than you have."

He picked up his notebook.

"Tell me," he said.

She told him. She talked for forty minutes -- about what Eleven had documented, about the mechanisms as best she understood them, about the layers of the Vesperin system and where the defense mechanism had been grafted onto the original design. He listened and wrote and asked specific technical questions at the points where his understanding needed more detail, and she answered as completely as she could, and when she reached the gaps in her knowledge she said so precisely.

When she finished, he sat with his notes.

"This changes the synthesis direction," he said. "The neutralizer I've been developing targets stage three of the cascade. Based on what you've just told me, stage one is the actual intervention point."

"Yes."

"That's a significant revision. Six to eight days of new framework development, minimum."

She looked at him. "And then?"

"And then we have a better neutralizer. Not just longer duration. A different mechanism entirely." He paused. "A better mechanism."

She thought about forty-two centimeters stable. She thought about thirty-five centimeters elevated but not threatened. She thought about stage one of a cascade that had been designed for a trigger no one had correctly identified.

"A different mechanism," she said. "Yes."

He looked at her across the examination room. The monitor was still running, showing her baseline: elevated slightly, not spiking, holding steady at the level she had been maintaining since the testing at thirty-five.

"How long has the baseline been at this level?" he asked.

She ran the internal check. Reported accurately: "Since approximately thirty-eight centimeters."

He wrote this down. He was very careful about what expression he allowed on his face while he wrote it, which was itself a kind of expression. She watched him be careful and found, in the specific way she had been finding things about him for weeks, that it was something she wanted to note.

"Thank you for the mechanism information," he said. "And for the testing."

"It serves the research," she said.

"It does." He paused. "And more than that."

She looked at him.

"You've been holding this information since before you arrived here," he said. "You didn't have to share it. You chose to."

She considered the various responses available and chose the most accurate one: "I know what good research looks like when I see it. And I know what the wrong research direction looks like. I've watched the wrong research direction run long enough." She picked up the door handle. "Fix the cascade. Fix the neutralizer. Then we'll see what the original mechanism was doing."

She left the examination room.

He stayed in it for a while, she knew, because she heard the specific quality of someone not moving in a room. Then she heard him return to the research wing, and the sound of the synthesis files opening, and the beginning of the framework revision she had just made necessary.

She went to the Surface Room and sat in the window chair and looked at the atmospheric display and thought about stage one and about the Vesperin's original activation trigger and about what she had been choosing not to say when he had asked her for a hypothesis.

She thought about this for a long time.

Then she reached for the transit maps and went back to work, because the maps needed finishing and the route needed planning and the day was still proceeding, and all of these things were true simultaneously and did not require resolution right now.

She would give the hypothesis time. Hypotheses were better for being given time before they were named.

---

The following morning she found his revision notes for the neutralizer framework on the kitchen counter.

Not a full document -- a single page, handwritten, the beginning of the new cascade model. He had been working on it late, probably past midnight by the quality of the handwriting near the bottom of the page, which was slightly more compressed than his morning script. She picked it up and read it with her morning tea.

He had built the new model from the stage one intervention point she had described. The architecture was different from the previous approach in a fundamental way: instead of interrupting the toxin cascade after activation, the new model addressed the activation mechanism itself. If the mechanism was correct, the neutralizer would prevent the cascade from triggering rather than arresting it partway through.

This was a substantially better approach. She had known it when she had described the mechanism to him yesterday. She had also known it earlier than that, in the theoretical sense of having worked out the implications of what Eleven's notes had contained, over years of having nothing to do but think. She had not said this yesterday because yesterday the testing had been enough and the framework revision had been enough and she had reached the limit of what she was prepared to give in one session.

He had worked through the night on it anyway, from what she had given him.

She read the revision to the bottom of the page. At the bottom: a note in different handwriting, more careful, written separately from the working material. *Reference: Mirelle's account of H-11 mechanism notes. Cross-check against available Heliogen literature. Confirm with Mirelle before finalizing pathway -- her input at each stage required, not optional.*

She looked at that last line. *Her input at each stage required, not optional.* Not a protocol instruction. A principle he had written to himself.

She set the page down on the counter. She finished her tea.

She went to the research wing and knocked on the partition frame -- she had started doing this rather than just walking in, a small ceremony that acknowledged that the research wing was his primary space even though she was often in the adjacent room.

He looked up. He had the synthesis files open, the new framework visible on the display. "You read the notes."

"Yes." She looked at the display. "The stage one targeting is correct. The mechanism description I gave you yesterday is accurate to what Eleven documented, and I've confirmed it against my own physiological experience." She paused. "There's one additional element I didn't mention yesterday."

He picked up his pen.

She told him about the threshold timing -- not just the activation distance, but the recovery pattern after the mechanism was triggered. Eleven had noted that the recovery curve was atypical: the cascade, once interrupted at stage one, resolved faster and more completely than standard Vesperin episodes. The implication for the neutralizer was significant: a stage one intervention would produce not just cessation of the active cascade but a return to below-baseline Vesperin production. A restorative effect, not just a preventive one.

He wrote this. He asked two clarifying questions. She answered both.

When she finished, he put the pen down and looked at the screen.

"This changes the expected duration," he said. "The 87-second window was based on arrest-and-hold. If the mechanism is restorative, the duration could be substantially longer."

She had thought this. She had not said it yesterday because she had not wanted to introduce speculation into what she was trying to keep as documented information. "Possibly. That would need to be tested."

"It would." He looked at her. "Do you want to tell me what you think the duration might be?"

She had a number. Eleven had been speculative in her notes -- not a firm figure, a theoretical range. She told him the range.

He was very still for a moment.

"That's not 87 seconds," he said.

"No."

"That's a different treatment protocol entirely."

"Yes."

He turned back to the display and began revising the framework. She watched him work for a moment -- the specific quality of someone whose thinking had just accelerated significantly, the kind of focus that made the work visible even from across the room.

She went back to the analysis room. She sat with the transit maps and listened to the sound of the synthesis work resuming at speed on the other side of the partition and thought about what she had just given him and what it meant and whether she had decided to give it because it was the right time to give it or because she had wanted to.

Both, she thought. Both were true. The fact that they were both true was the part she was sitting with.

She opened the map and worked on the route and let the partition carry the sound of the research from one room to the next, and the morning proceeded into afternoon, and the station held them both.

That evening she found him in the common area with the revised framework printed out, annotated in three colors. He looked up when she came in.

"The duration range you gave me," he said. "It's consistent with the mechanism. I've been cross-checking it against the dispersion data from our testing series, and the numbers align. The stage one intervention point is correct."

She sat down. "When will the revised synthesis be ready to test?"

"Six days, if the framework holds. Possibly five." He set the papers down. "The new batch would be a different compound entirely. Not a modification -- a different mechanism. It will need to be tested from scratch."

"I know."

"The testing protocol I've been using won't be valid for the new compound. We'll need to rebuild it."

"I know that too." She looked at the annotated framework. "You should sleep."

He looked at the clock. It was past midnight. He had been working since midmorning. "I'm finishing one more section."

"You'll be more accurate in the morning."

He looked at her. She looked back at him. This was the third or fourth time she had said a version of *you should sleep* to him, and he had never once simply agreed. She was developing a sense of the specific point in his resistance at which it was worth pressing and the specific point at which it was not.

This was the not-worth-pressing point.

"One more section," she said. "Then sleep."

He nodded. She went to her room.

She heard him in the research wing for another forty minutes, and then the workstation going to standby, and then the station's overnight quiet settling in. She lay on her bed and thought about stage one and about the duration range and about the fact that she had been keeping this information for three years and had given it to him this week, and about what had made this week the week she gave it.

She thought: I gave it because the research needed it. Also: I gave it because I wanted him to have it. Also: I am no longer certain which of those is the larger reason.

She thought about the two of them in the same research being true simultaneously and not collapsing. She thought about how long it had been since she had wanted to give someone something.

She went to sleep.