The reorganization of the cartographic section took most of the week.

It was not strictly necessary work -- the catalogue functioned adequately -- but the section had accumulated the minor inefficiencies of decades of individual decisions made in the moment, each reasonable in context, collectively producing a system that required memory to navigate rather than logic. Liora had been meaning to address it for three years and had not had the bandwidth. Elias offered because the structural problem interested him and because he could see it needed doing and because, he admitted to himself honestly, it gave him a reason to work beside rather than across the room from her.

She accepted without making anything of the offer.

They worked together. He had his way of organizing: spatially, by geography and period together, the kind of sorting that produced coherent physical clusters. She had her way: temporally, by date of acquisition layered with date of content, the kind of sorting that produced coherent historical sequences. They discussed the difference and arrived at a hybrid method that neither of them would have come to alone.

He discovered this was pleasant. Not just useful: pleasant. The specific satisfaction of working through a problem with someone whose thinking was different from yours and complementary to it.

She occasionally hummed something he did not recognize, quietly, the kind of involuntary vocal response to focused work that people produced when they forgot they were not alone. He did not mention it. He found it, in some specific way he did not examine closely, one of the most comfortable sounds he had been around in years.

She told him about her grandmother while they worked.

Not formally, not as a briefing. Conversationally, the way information came out during extended shared work: when both hands were occupied and the mind could move alongside the task without the distraction of face-to-face directness.

"My grandmother started the catalogue in 1962," she said, sliding a folder into a new position. "She was fifty-three. She'd been running the archive for twenty years before that on a more informal basis. But in 1962 she decided that informal was no longer adequate and she spent three years building the system." A pause. "She used to say she was cataloguing the future."

He looked up. "What did she mean by that?"

"She meant that the archive contained the record of everything Grayhaven had known, including the Tomorrow Memories of its residents. If you organized it correctly, she said, you had access to a kind of collective future that no individual's Tomorrow Memory could hold." She straightened a stack of folders. "She wasn't wrong. Some of the most useful research we've done has come from reading the archive as a collective document rather than a collection of individual records."

He thought about this. He thought about collective records and individual contributions to them. He thought about what it would mean to leave something in a place so that the place itself held it after you.

"Did she document her own Tomorrow Memories?" he said.

"Yes." Liora was quiet for a moment. "She was unusual in that way. Most people don't want records of their Tomorrow Memories. They feel too private. But she wrote them down. She said the archive couldn't be complete if it excluded the archivist." A pause. "Her records from the 1960s and 1970s are in the back room. I haven't read all of them."

"Why not?"

"Some of them are about my mother," she said. "And some are about me. From before I was born." She said this with the specific evenness of someone who had made a decision about how to hold a thing that was strange.

He was quiet for a moment. Then: "Does it feel strange? Being pre-documented."

"Sometimes," she said. "Less than it used to." She moved to a new section of the shelves. "It depends on how I read it. If I read it as fact, it feels strange. If I read it as the care of someone trying to record what she loved -- it feels like what it is."

He thought about that distinction for a long time while he worked.

The mistake happened in the mid-afternoon.

He was working through a series of land surveys from the 1930s, looking for a specific notation about the coastal formation that matched a reference in Celeste's research notes. He needed a supplementary volume -- a companion index published in 1934, which he had seen in the catalogue but had not yet retrieved.

He asked her for it.

She went to the shelves without hesitation and came back with a folder.

He opened it. It was the wrong one -- a 1934 document, the right year, but a different survey entirely. He looked at it for a moment. He looked at her.

She was still standing near the shelves, and he watched her realize, with the specific expression of someone whose working model had just produced an error.

"That's the wrong one," he said.

"Yes." She came to look at the folder. She said nothing for a moment. "I thought you needed that one first."

"Why?"

"Because -- " She stopped. She was doing the thing he had seen once before: recalibrating in real time, checking her Tomorrow Memory against what was actually in front of her. "Because in the memory I have, you use that one before the companion index." She looked at the folder. "But in the memory, you ask for it differently. And I responded to what I remembered, not what you asked."

He thought about this. He thought about the left hand and the box.

"Your memory is accurate," he said. "I probably will use that one at some point. But it's not what I need right now."

"I know." She took the folder back and retrieved the companion index without being asked. "I'm sorry. I was running ahead of the actual moment."

He accepted the companion index. He opened it. He found what he was looking for on the third page.

"It happens," he said, without looking up.

"Not often," she said. "Not usually with you." A pause. "That's what's interesting."

He looked up at that.

She had said *interesting*, not *troubling* or *worrying*. She had said it with the quality of someone who had found a piece of data they were genuinely curious about rather than threatened by.

He thought about the word *interesting* for the rest of the afternoon.

He thought about what it meant to watch someone discover that their record of you was incomplete, and to respond not with distress but with curiosity. He thought about the particular kind of generosity that required: a willingness to let the record be corrected, to prefer the real version even when the real version was less tidy.

He worked through the 1934 surveys. She worked through the end of the reorganization. The archive was quiet in the way of a shared working space at the end of a task: the specific quiet of completion, of things returned to their correct places.

At four he found the journal.

It was in the back of a folder he had almost finished with: a small notebook, not part of the formal collection, tucked between two survey documents as if someone had placed it there temporarily and forgotten to retrieve it. The cover was unmarked. He opened it.

The handwriting was not Liora's, and not her mother's, and not the careful institutional hand of formal archive work. It was older, looser, the handwriting of someone from a different era of writing tools.

He read the first entry. It was dated 1953.

About two thirds of the way through, a passage caught him: *The man who arrives at the end of summer will not know what he is arriving into until after he has arrived. This is always the way of these things. The arrival must precede the understanding. This is not a failure of foresight -- it is the nature of the gift, which is that understanding is only possible from inside the thing, not from outside it.*

He read it twice.

He looked up. Liora was across the room, returning the last of the reorganized folders to their new positions.

"Liora," he said. "Whose handwriting is in the back section of the 1930s land survey folders?"

She came over. She looked at the journal.

Her face changed. Not dramatically: a slight shift of attention, the quality of recognizing something.

"My grandmother's," she said. "Her personal journal. It got misplaced years ago. I stopped looking for it." She took it carefully. "Where did you find it?"

He told her. She looked at the folder, at the position in the survey documents. She looked at the journal. She opened it to no page in particular and read for a moment.

"I haven't read this," she said. "Since I was a child. I found it once in the house and she caught me and said -- " She stopped. "She said: you can read it when the time is right."

He sat very still.

"Is the time right?" he said.

She looked at him. She looked at the journal. "She would have known," she said quietly. "She would have known exactly when I'd find it again." A pause. "Yes," she said. "I think it is."

She took the journal to her desk. She held it for a moment without opening it. He went back to his work. The archive settled into its late-afternoon quiet.

At five she locked the journal carefully in the back room, where it would not be misplaced again. He walked back to Calder House that evening thinking about the grandmother's journal.

He had not read it -- she had taken it from him before he could -- but the passage he had read stayed with him: *the arrival must precede the understanding.* He turned this over the way he turned over load-bearing problems. The structure of the sentence suggested a specific argument: that arriving was a prerequisite for understanding, not a secondary step. That you could not understand what you were walking into from outside it. You had to be in it first.

He thought about three weeks in Grayhaven. He thought about everything he had understood only after arriving: Tomorrow Memory, the clocks, Liora's eight years of careful management, Clara's choice. He had known some of these things before he came, in the way you knew something from description. He had understood them only after he was inside them.

He thought about what he still did not understand. He thought about the death memory fragment -- the cold air, the grey sky, the hands -- and what it was the prelude to. He thought about Celeste's Convergence Period research and what it implied about the adjustment Liora was making and what that adjustment cost. He thought about the pocket watch and when it would run.

He thought about arrival and understanding as a sequence. Arrival first. Then understanding. The understanding was not yet complete. This was, he decided, all right. He was still arriving, in the sense that mattered.

In the morning he would go back to the archive. They would finish the reorganization. She would have read some of the grandmother's journal, or would not have. He would be there to find out which.

He stood at the window of Room 17 for a while. The harbor was lit. The clock tower showed eleven minutes behind. He had stopped finding this strange and started finding it accurate.

He thought: I am inside the thing now. The understanding will come from here.

The next morning the journal was on the reading table when he arrived.

She had left it there for him, with a note in her handwriting: *read the section from 1961-1964. Then we should talk about it.*

He read it while she was at her desk.

The grandmother's handwriting in that section was more confident than the earlier entries, the writing of someone who had been doing this long enough to know what she wanted to say. The 1961 entries described a period of significant change in the archive: the formal cataloguing, the decision to build a coherent system. But the entries that Liora had marked, the ones from late 1963 and early 1964, were different in character. They were not about the archive. They were about waiting.

*There are events that a Tomorrow Memory shows you clearly,* she had written in November 1963, *and events that it shows you clearly but incompletely. The distinction is important. A clear and complete Tomorrow Memory can be prepared for, can be managed, can be made ready for. A clear but incomplete one cannot be fully prepared for, and the attempt to prepare for it too completely often means you have not left room for the reality to arrive.*

*The reality is always larger than the memory. This is a grace, not a flaw.*

He read this paragraph several times. He thought about Liora's eight years of preparation and the adjustments she had been making since he actually arrived. He thought about the grandmother's observation: the reality is always larger than the memory. He thought about being the reality instead of the memory.

He closed the journal and looked across the room at Liora.

She was watching him.

"She was right," he said.

"She usually was." Liora came to the reading table and sat across from him. "She was writing that for me. She knew I would read it someday. She knew the circumstances." A pause. "She was telling me to leave room."

"Did you?" he said. "Leave room."

She considered this carefully. "I tried to," she said. "I managed the sequence. I tried not to make too many decisions in advance. But -- " She looked at the journal. "There's only so much room you can leave when you've been carrying something specific for eight years. The memory is its own kind of structure. It takes up space."

He thought about structures and the space they took up. He thought about what it would mean to dismantle a structure you had built over years in order to make room for something larger and more real.

"Are you making room now?" he said.

She looked at him directly. "Yes," she said. "I've been making room since the first day you used your left hand instead of your right." A pause, and then: "It's getting easier.

She went back to her desk. He went back to the reading table. They worked through the rest of the afternoon in the comfortable parallel that had become the texture of their working days.

He thought about *getting easier* for a long time.

He had been inside buildings that resisted their own renovation: structures that had been built one way and that pushed back against being made into something different, where every adjustment revealed another resistance, where the process of becoming took longer than anyone expected. He had also been inside buildings that seemed to welcome renovation, that opened up as you worked through them, where each adjustment led naturally to the next and the emerging structure felt like what the building had been trying to be all along.

He thought about that distinction. He thought about the grandmother's sentence: *the reality is always larger than the memory.*

He thought about room, and making it, and how the making of it was not loss but expansion.

At five o'clock she said: "The journal. I'm going to read it tonight. The parts I haven't read."

He looked up.

"I've been waiting until I thought I could read it as information rather than instruction," she said. "I think that's now."

He thought about the difference between instruction and information. He thought about a grandmother who had written about leaving room for the reality to arrive, and a granddaughter who had been carrying the probable future for long enough to have learned, finally, to hold it lightly.

"Will you tell me what you find?" he said.

She considered this. "Yes," she said. "When I know what it means to tell."

He nodded. He put on his jacket. He said good evening. She said good evening, and it had the specific quality of an evening that was not an ending but a pause, a breath between one part of the day and whatever the night would bring to both of them separately before morning brought them back to the same room.

He stood at the window of Room 17 and thought about the grandmother writing for a granddaughter she would not see read it. He thought about the kind of trust that required: to write something across years, to a person who did not yet exist in the form that would need the writing, and to leave it in a place where it could be found at the right time. He thought about what kind of person made that kind of careful provision. He thought about the archive, which was that kind of provision, and the woman who ran it, and the morning when the journal would be read. The direction was toward something specific and near. He felt this the way he felt the approach of the right solution to a structural problem: not as certainty but as the particular quality of momentum that preceded certainty. He walked back through the evening streets of Grayhaven with the clock tower visible above the rooflines and the harbor lights beginning to appear below him and the particular sense of a city that was becoming, incrementally and irrevocably, familiar. He knew the bakery hours and the name of the fisherman at the end of the dock and the specific quality of the light in the archive on different weather days. He knew which questions she would answer immediately and which she would hold briefly before answering. He knew the archive the way you knew a place where significant things had happened and where more significant things were pending.