She found the name in the fragments.

It was Day 147, and they were working on the Heliogen records reconstruction together -- she had offered three weeks earlier to help with the fragment decoding, citing her familiarity with the file architecture from the facility. He had accepted. She had been right: she moved through the fragment structure faster than he did, recognizing patterns in the deletion that he would have taken twice as long to identify.

The fragment set was the last major section of the records he had not fully reconstructed. These were the administrative files -- not the research documentation but the operational layer: authorizations, personnel records, the internal communication chain from the years of the program's operation.

She found the name in the authorization chain for the Year 7 deletion.

"There's a signature here," she said. She was at the analysis room table with the fragment display, and he was in the research wing visible through the partition. "On the erasure authorization."

He came to the partition. He looked at the fragment.

The signature was partially corrupted -- a fragment of a digital authorization record, the kind of timestamp-and-identity marker that administrative systems used when recording who had authorized a data deletion. The corruption had eaten most of the identifier, but enough remained: a specific pattern in the authentication hash, and three characters of a surname beginning with P-e-l.

He recognized the pattern. He had seen it in another fragment, weeks earlier, in a research authorization from Year 3.

"Ivara Pell," he said.

She looked at him. "You know that name."

"I've seen the authentication pattern in a Year 3 research authorization. A different document." He looked at the fragment. "The signature matches."

"Who is she?"

"Senior research lead, Heliogen facility. I don't have a physical description or current location. She signed authorizations in Years 3 through 7." He paused. "She signed the Year 7 deletion."

Mirelle was quiet.

"The deletion was for the records she was responsible for," he said. "The Year 3 authorization she signed was for the Vesperin protocol research. She designed it."

Mirelle looked at the fragment display. She did not say anything for a moment. He watched her process this -- the specific quality of someone who had grown up knowing the effects of that design on their body, and was now learning who had made the decision that caused it.

"She designed the Vesperin cascade," Mirelle said.

"Based on the record fragments I've reconstructed, yes."

"And she deleted the records in Year 7."

"Yes."

"Why Year 7?"

He thought about what else had happened in Year 7 -- what the fragments had shown about the program's operational history. He had been reconstructing the timeline for weeks. Year 7 was the year the facility had shifted from the multi-subject approach to a focus on the single viable subject. Year 7 was also the year the Year 5 experiment results would have been reviewed.

He told her what he knew. He told her about the Year 5 documentation he had partially reconstructed -- the record of an experimental procedure, a failed attempt at something the fragments described as harvest protocol, the designation Subject 7.

She stopped him.

"H-07," she said. "Bao."

"I don't have the subject's name in the fragments."

"H-07. He was -- he was there in Year 5." She looked at the display. "He didn't come back from a session. I remember -- he was there and then he wasn't." She stopped. "Year 5. I was thirteen."

He was quiet.

"What happened to Subject 7?" she said.

He told her what the fragments said: an experimental procedure, incomplete, Subject 7 deceased, duration logged as nineteen hours. The fragments did not include the cause of death or the specific nature of the procedure. The procedure was referenced in later documents as the approach that was abandoned in favor of an alternative methodology.

She sat with this.

He did not fill the silence. He stood in the partition doorway and let the room hold what it was holding.

---

After a while she said: "Nineteen hours."

"Yes."

She looked at the display. She was very still in the specific way of someone who is not performing stillness but who has gone to a place inside themselves that required stillness to hold. He recognized this. He had watched her arrive at this place before -- in the thirty-two names session, in the Ivara Pell first call, in the moments when the history she carried arrived in the present with specific weight.

He waited.

"When Eleven gave me the names," she said. "When she told me about H-07 -- about Bao -- she told me he had gone to a session and not come back. She didn't know what had happened. None of us knew what happened when someone didn't come back." She paused. "We assumed things. We told ourselves the ones who didn't come back had gotten out. That was -- we needed that story." She looked at the fragment display. "It wasn't always possible to keep believing it."

He said nothing.

"Nineteen hours," she said again. "He was seven years old in Year 1. He would have been eleven in Year 5."

He had calculated this. An eleven-year-old subject. He had known the ages when he started reconstructing the fragment timeline and he had set the knowledge aside because engaging with the specific age of each subject in the Year 5 record was not something the research process required. Now she was saying it out loud and it was in the room and he could not set it aside.

"I'm sorry," he said.

She looked at him. Not the assessment look. Something else -- the look of someone who has received a statement that they are checking against what they know.

"I know," she said.

She stood up. She went to the Surface Room.

He stood at the partition doorway. He had learned, over months, the map of when to follow and when to stay. This was a staying moment. She needed to take this somewhere alone and he needed to let her take it.

He went back to the research wing.

He opened Bren's most recent message: *Lead confirmed. Target located, northeast sector settlement compound. Can arrange contact, two to three days.*

He wrote back: *Arrange it.*

He put the message down. He looked at the research wing -- the synthesis equipment, the monitoring displays, the characterization data from the failed batch, the files for the formulation rebuild. He looked at the encrypted folder identifier on the file system, which was just text, which was not visible as anything other than a folder name.

He thought about Bao and nineteen hours and an eleven-year-old who had had a language no one else knew.

He thought about what Ivara Pell had known when she signed the Year 7 deletion.

He thought about what she had been trying to make disappear.

He opened the alternative pathway analysis file. He had been working through it methodically. He was at the edge of his current knowledge. He needed Ivara Pell's perspective on the Year 3 design -- the original cascade mechanism, the decisions she had made when she built the system, the pathways she had considered and rejected.

The answer, if it existed, was probably in something she had decided not to do in Year 3. The answer, if it existed, was in the road not taken by the person who had taken a different road and had then spent decades trying to make the record of it disappear.

Bren had found her.

Two to three days.

He went back to the formulation rebuild, which was the current work and which needed to be done. He worked on it until midnight. He went to bed.

In the Surface Room, the light was on.

She was still in the window chair when he passed the Surface Room doorway. He could see the monitoring display from the corridor, and her profile in the window chair against it. He did not go in. He stood in the corridor.

After a moment she said, without turning: "I know you're there."

"Yes."

A pause.

"Nineteen hours," she said.

"Yes."

"He was eleven."

"I know."

She was quiet again. He stood in the corridor.

"I'm going to be in here for a while," she said. "You don't have to stay in the corridor."

"I know."

She turned then -- just slightly, just enough to see him in the doorway without fully facing him.

"You can come in," she said. "If you want."

He went in. He sat in the chair across from the window chair -- not the window chair, the smaller chair at the side, the one that was not hers. He did not say anything. She did not say anything. They sat together in the Surface Room in the late-night monitoring display light and the station was quiet and outside the surface was at its standard state.

After a while she said: "Did Bren find her?"

"Yes. Two to three days to arrange contact."

She nodded. She looked at the monitoring display.

"I want to talk to her," she said.

"I know."

"I don't know what I'll say."

"You'll know when you're there," he said.

She looked at the display. "Maybe." Then: "When Eleven gave me the names, she told me the names were all that was left when everything else was taken. She said: remember them, because remembering is the thing they can't design away from you." She paused. "I've been remembering for twenty-three years. I want to say them to the person who took them."

He sat with this.

"Ivara Pell deleted the Year 7 records," he said. "She deleted what she could access of the program's documentation. She did not delete the names. Because she didn't know them."

Mirelle looked at him.

"The names are yours," he said. "She couldn't delete what she didn't know existed."

She held this for a moment. Then she looked at the monitoring display and said nothing for a while and they sat together in the Surface Room until she was ready to sleep.

---

Two days later Bren confirmed the contact.

The connection would run through the relay network -- not the provisional Bellhaven channel, the older northeast network that ran through the junction infrastructure. Bren had established a signal path and a time window: the next morning, six-thirty, before the transit hour.

They set up in the Surface Room, which was the room with the best relay display.

Mirelle was in the window chair. He was at the monitoring station. The relay connected at six-thirty-two.

The woman on the screen was approximately sixty years old. Greying hair, short. A plain background that suggested a settlement residential unit. She looked at the screen when the connection established, and he watched her read the two of them -- the way her eyes moved to Mirelle and stayed.

He had expected hostility, or defensiveness, or the careful performance of someone who had rehearsed what they would say if they were ever found. He did not see any of those things. He saw someone who had been found and who was not trying to manage it.

"Dr. Pell," he said.

"Yes." Her voice was steady. Looking at Mirelle. "You're the seventeenth."

"Yes," Mirelle said.

The silence that followed had weight. He held it and did not fill it.

Ivara Pell said: "I've been waiting for this contact. Not with -- I have not been waiting with hope. I've been waiting because I owed it." She paused. "I didn't know if you were alive."

"I am," Mirelle said.

"I know that now." Ivara looked at her hands. "I designed the Vesperin cascade. I designed it poorly -- I built a mechanism that was intended as a tracer and the tracer became a weapon because the program didn't need a tracer. They needed a control mechanism."

"I know what they needed it for," Mirelle said.

"I know you do." Ivara looked up. "I left Heliogen in Year 7. I had been leaving incrementally for two years -- refusing certain procedures, reclassifying certain research directions. By Year 7 I was a liability to the program and they were managing me out."

"But you deleted the records before you left," Mirelle said.

"Yes."

"To protect yourself."

"Yes." Ivara said this flatly, without defense. "I told myself it was to protect the program from being reconstructed. I told myself that if the records were gone the research couldn't be repeated. But the primary reason was protection. I was afraid of what they would find if the records survived."

Mirelle was quiet for a moment. Then she said: "Did you know what happened to Subject 7?"

Ivara closed her eyes briefly. "Yes."

"The fragments say nineteen hours."

"I know what the fragments say." She opened her eyes. "I wrote the process note. I was the senior researcher on the experimental procedure. I wrote the process note and I could not make it accurate and I filed it inaccurately and I left the facility two months later." She paused. "I have been living with the accurate version for eighteen years."

Mirelle said: "His name was Bao. He was eleven years old. He had a language that no one else spoke in the facility. He used it when he was afraid."

Ivara was very still.

"I'm telling you his name," Mirelle said, "because you deleted the records and so you never knew it. The names are what's left after everything else is taken. His name was Bao."

---

After a silence, Lucan said: "Dr. Pell. I need to ask you about the cascade mechanism."

Ivara looked at him. The shift was visible: from the weight of what Mirelle had said to the clinical space he was inviting her into. Not relief -- she was not escaping the weight. She was being given a different task to do alongside it.

"The stage one mechanism," he said. "I've been developing a neutralizer using the stage one intervention point. I've confirmed the mechanism in synthesis and in controlled testing. I'm working on a scale-up model for Luminase-R production." He paused. "The current scale-up model requires a biological source that is -- not renewable. I'm looking for an alternative pathway."

She looked at him. He watched her process the phrase *not renewable* with the specific attention of someone who understood exactly what it meant.

"There's no alternative in the standard approach," she said. "I looked. In Year 4 and Year 5, before the program shifted to extraction-based production, we tried several alternative synthesis routes. None of them produced sufficient yield."

"Why not?"

"The protein structure requires a specific biological activation that the standard synthesis environment can't replicate. The only biological systems that produce the correct activation are the subjects themselves."

He had known this. It confirmed the model. He held the confirmation and said: "You said in Year 4 and Year 5. Were there earlier approaches -- before Year 4? Approaches that were discarded in the initial research phase?"

Ivara thought. "Year 2. There was a fungal scaffold approach that one of the junior researchers proposed. It was discarded because the yield was insufficient for the program's needs."

He was quiet.

"What was the yield basis of comparison?" he said. "What was it being compared against?"

Ivara frowned. "The program needed a production scale for -- for the application they had in mind. Mass distribution. The fungal scaffold approach produced yields that were adequate for small-scale treatment but insufficient for the volume the program required."

He sat with this.

"What if the required volume was small-scale?" he said. "Not mass production. A treatment population of hundreds rather than millions."

Ivara looked at him. She looked at Mirelle. She said: "I don't know the specific yield figures from the Year 2 research. The junior researcher left the program before Year 3. I don't know what happened to her notes."

"Do you remember her name?"

"Her first name. Vesti. I don't know her surname. She was in the facility for six months."

He looked at Mirelle. She was looking at the screen.

"I remember Vesti," she said. "She was a junior researcher who was reassigned. I didn't know she had been working on an alternative synthesis approach." A pause. "She was kind."

"She left voluntarily," Ivara said. "She disagreed with the program's direction. By Year 3 there were not many people who left voluntarily."

Lucan said: "The fungal scaffold approach. If it produced sufficient yield for small-scale treatment -- for a clinical trial and a treatment population of several hundred -- it would change the scale-up calculation entirely."

Ivara looked at him. "It might. I don't have the data."

"No," he said. "But the Year 2 research existed. The data might still exist somewhere." He looked at the relay display. "Or someone might remember it."

He looked at Mirelle.

Mirelle was looking at the screen with the specific quality of someone arriving at a memory they had not thought about in a long time.

"She showed me," Mirelle said. "Vesti. In Year 2 -- I would have been nine years old. She was showing me something in one of the abandoned experiment sections. A fungal culture she had been growing." She was quiet. "I remember what it looked like. I remember what she said about it." A pause. "I remember the structure of it. She drew it for me."

He felt something shift in the research framework he had been building for days.

He said: "Tell me what she drew."

She described it for twenty minutes. She was methodical about it -- she had the memory clearly, preserved with the specific intensity of a child who had been shown something kind by a person who was kind in an environment that offered very little kindness. The structure, the growth pattern, the cellular arrangement that Vesti had drawn on a piece of paper and given to a nine-year-old to keep.

She did not have the paper. It had been lost twenty-two years ago.

But she had the image of it.

He listened. He took notes. He asked three clarifying questions and she answered each one with the precision of someone who had been looking at this image in their memory for twenty-two years without knowing it would matter.

When she finished, Ivara said: "That's the scaffold structure. That's the Year 2 approach."

He said: "I think I can rebuild the synthesis model from this description. I need two days."

Mirelle looked at him. He held her eyes.

"Two days," she said.

"Two days."

Ivara looked at both of them. She said: "The junior researcher. Vesti. If you find what she was working on -- I would like to know. I would like to know what became of it." A pause. "If it works, I would like to know."

Mirelle looked at the screen. She said: "I'll tell you."

The connection closed.

He was already at the workstation.