The signal came in on a Tuesday, three weeks after Isolde brought the drive.\n\nElianor was running the standard overnight capture review when she found it — a passive capture from the harbor monitoring array's secondary sensor, logged automatically at two forty-seven in the morning during a minor aurora event that she had not been awake for and had not expected to produce anything significant.\n\nThe signal was not from the harbor district.\n\nShe ran the location analysis three times before she accepted the result. The frequency signature was consistent with equipment from the late period before the Frostline 19 incident — the specific electromagnetic fingerprint of field recording hardware that had been manufactured for approximately three years and then discontinued, the kind that Niven had used. The kind that had been at the bottom of Lysmere Lake for seven years.\n\nThe kind that should not have been transmitting at all.\n\nShe sat at her workstation for a long time. The early morning archive was very quiet — she had come in at six, as she sometimes did when the overnight queue needed review, and the rest of the building would not be active for another hour. The monitoring systems ran their background hum. The narrow windows showed the dark of the early October morning.\n\nShe picked up her phone.\n\nShe typed: *The harbor array captured a signal from Lysmere Lake last night. Equipment-type match to field recorder, late-period pre-incident manufacture. The signal came from under the ice.*\n\nShe set the phone down. She ran the localization analysis again, with the secondary confirmation parameters. The result was the same.\n\nHer phone showed: *Are you certain of the source localization?*\n\nShe typed: *Yes. The harbor array has full directional coverage of the harbor basin and the lake approach. The signal geometry is unambiguous.*\n\nA pause. Then: *The lake sensors. When it's cold enough, the equipment at depth can sometimes restart. If the battery degraded partially rather than completely, the cold can actually — certain battery chemistries will temporarily regain function at low temperatures.*\n\nShe had not known this. She thought about it. A battery that had been at the bottom of a cold lake for seven years, degraded but not dead, temporarily reactivated by the combination of cold temperature and the magnetic conditions of the aurora event.\n\nShe typed: *There's material in the lake that still has the potential to transmit.*\n\nHe replied: *Yes.*\n\nShe looked at the localization data on her screen. The signal had lasted eleven seconds before dropping out. Eleven seconds of a field recorder that had been sitting in the dark cold of Lysmere Lake since January fourteenth, seven years ago.\n\nShe said to her empty workstation: *He was thorough.*\n\nShe thought about Niven recording on the train. She thought about him in the carriage, the recorder running, doing what he always did — documenting the sounds of the world he was moving through. She thought about the recorder surviving the impact and the cold and the water and seven years at depth, holding whatever it had captured in its degraded battery and its sealed housing.\n\nShe thought about all the things that had been held in the cold, waiting for the conditions to be right.\n\nShe picked up her phone again and typed: *We need Rowan's team to do a survey of the lake bottom. If there's recoverable equipment down there — not just the recorder, but the structural evidence of the carriage — it changes the physical evidence available to the panel.*\n\nThe response came after a pause: *I'll contact him this morning. Do you want me to come to the archive first?*\n\nShe looked at the localization data.\n\nShe typed: *Yes. Come at eight.*\n\n---\n\nRowan arrived at nine-thirty, which meant he had been on the phone before eight.\n\nHe had a tablet with preliminary survey maps of Lysmere Lake — the depth charts, the ice condition assessment, the access points that were operable in October before the ice became too thick for diving operations. He spread them on the reading room table and they all stood around them: Elianor, Einar, Rowan.\n\n\"The lake is accessible right now,\" Rowan said. He was using his professional voice — the one Elianor had heard in briefings, the one that was calm and precise and made the dangerous sound manageable. \"We have a window before the serious ice sets in. Late October, early November at the latest. After that, ice diving is possible but significantly more complex and the visibility drops.\"\n\nElianor looked at the depth charts. The area where Frostline 19 had gone in was marked — a notation from the original rescue operation, seven years old. The depth at the impact point was significant. Not impossible for a survey dive, but challenging.\n\nShe said: \"The signal localization puts the source here.\" She pointed to the depth chart. \"That's consistent with the impact geometry from the original accident report.\"\n\nRowan looked at the point. He looked at Einar.\n\nEinar was looking at the map. She had been watching him since he arrived — he was quiet, quieter than usual, the particular stillness of someone managing a weight they had been carrying for a long time and had just been told they might have to go back to where they set it down.\n\nShe said: \"You don't have to come.\"\n\nHe looked at her.\n\n\"The survey is Rowan's team. I can be there as the archive's representative. You've provided the documentation you were asked to provide.\" She paused. \"You don't have to be on the ice.\"\n\nHe was quiet for a moment. He looked at the map.\n\nHe said: \"I've been on that ice many times. In the seven years before Frostline 19, I surveyed the lake three times as part of the aurora monitoring program. The magnetic conditions there were relevant to the tower's baseline data.\" A pause. \"The lake was part of my work before it was part of my history.\"\n\nShe looked at him.\n\n\"I'm going,\" he said. It was not said with force or resistance — just with the evenness of a decision that had been made and did not require argument.\n\nRowan said: \"Survey is Thursday. Weather window is good. I'll need both of you for the documentation chain — the physical evidence recovery will need to be handled correctly for the panel to accept it.\"\n\nElianor looked at the depth chart. At the point where the signal had come from. At the outline on the map of the route Frostline 19 had been running when the navigation system failed.\n\nShe thought about Niven on that train with his recorder. She thought about the recorder at the bottom of the lake, waking for eleven seconds during an aurora event, sending out its signal into the cold dark water and then going quiet again.\n\nShe thought: he was still there, in a way. Still transmitting, still keeping his record, still doing what he had always done — trying to capture the sounds of the world through whatever interference was present.\n\nShe thought: the archive built to listen would hear him, eventually.\n\nThe archive was hearing him now.\n\nShe said: \"Thursday.\"\n\n---\n\nWednesday she did not sleep well. Not badly — not the sleeplessness of crisis, not the kind that left her hollowed out in the morning. Just the kind where the mind kept returning to the same image and she would let herself rest in it for a while before it drifted away and then returned.\n\nThe image was the lake. Lysmere Lake in winter — the way it looked from the ridge path up to the tower, the pale grey expanse of it, the way the ice formed in patterns that Niven had found interesting from an acoustic perspective because the thickness varied and the variation created different resonance properties. He had written about this in the field notebook. *The lake ice sounds different in different sections because of the different thicknesses. The sound moves faster in the thick ice. I want to record this.*\n\nHe had not recorded it. The field notebook entry was from his last winter season, two months before the accident. He had meant to go back.\n\nShe lay in bed and thought about him going back. About the recording he had been planning to make. About all the recordings that had existed as intentions and had never become actualities, the sounds he had meant to capture that no one would ever hear.\n\nShe thought: but the recorder was there. On the train. Running. Whatever it had captured in those minutes was still there, at the bottom of the lake, in the housing that had survived the impact and the cold.\n\nShe would find out, on Thursday, whether it was recoverable.\n\nShe thought about what it would mean to have another recording of him. Something from those last minutes, those last sounds — the train, the passengers around him, whatever he had been doing and thinking and observing. The recorder had been running. That was who Niven was: someone who kept the recorder running, because the sounds of the world were always worth capturing.\n\nShe thought about being ready for that.\n\nShe was not sure she was. She thought she was getting there.\n\nShe fell asleep at some point after two and woke at five-thirty as always, the harbor bell counting the early hour through the dark, Aurevik beginning its Thursday in the ordinary way it always began, indifferent to what the day was going to contain.\n\nShe made coffee. She sat at the kitchen table.\n\nThe lamp was burning in the other room.\n\nShe thought: *not yet.* But she looked at it differently now than she had in the early months. The not-yet felt less like waiting and more like the interval before something she had chosen.\n\nShe dressed. She went downstairs. She walked to meet Rowan's team at the lake.\n\nEinar was already at the staging point when she arrived.\n\nHe looked at her. She looked at him.\n\nShe said: \"Ready?\"\n\nHe said: \"Yes.\"\n\nThey walked to the edge of the ice together, the survey team spreading out around them, the equipment coming off the transport, the pale October morning still dark at the horizon where the lake met the sky.\n\nShe thought: this is the right way to do it. Not alone. With the person who was there.\n\nThey stepped onto the ice.\n\n---\n\nThe ice held.\n\nRowan's team had done the pre-survey the previous day — ice thickness testing at multiple points across the survey route, the standard protocol before any team went onto lake ice in conditions like these. The ice was thick enough. Not comfortable-thick, not walk-on-confidently-thick, but within the operational parameters Rowan's team worked within.\n\nShe had been on ice like this before, though not often. There was a particular quality to it — the sound of it underfoot, slightly different from ordinary ground, slightly hollow in a way that was not alarming but was noticeable. She was aware of the water beneath. She had always been aware, standing over Lysmere Lake, of the water beneath.\n\nShe thought: Niven thought about acoustic properties. He would have been interested in how different the footstep sounds were on ice over water versus ice over land. He would have taken notes.\n\nShe was taking her own kind of notes. She was noticing things and storing them in the way she had learned to store things over the past year — not as grief, not as loss, but as record. He had been here. She was here now. The record was the continuity.\n\nThe dive team went in at the point Rowan had identified from the depth charts and the signal localization. They were down for forty-one minutes — long enough for a preliminary survey of the bottom in the target area, documentation of what was there, assessment of recovery feasibility.\n\nShe and Einar waited on the ice with the surface team.\n\nShe was aware of him the way she was always aware of him in significant situations — not with the cataloguing attention of the early months, when she had been reading every expression and gesture and filing it against the question of who he was and whether she could trust him. She knew who he was. The awareness was different now, more like the awareness of the harbor bell — something that marked the presence of something reliable.\n\nHe was standing at the edge of the access point with his hands in his coat pockets, looking at the hole in the ice where the divers had gone down. She thought about what this place meant to him. About the night he had been pulled out of this water — hypothermic, alone, the only person to come out of that carriage. About the years of other cities and the weight he had described as not going anywhere, weight without direction.\n\nShe thought about what it meant that he was standing here now. Not rescued. Not alone. Standing on the ice in October with a purpose that had taken seven years to arrive at.\n\nShe moved to stand beside him. Not quite touching — just close. In the range where their breath was the same cold air.\n\nHe looked at her briefly. He did not say anything. Neither did she.\n\nThe divers came up at eleven forty-three.\n\nRowan pulled the team leader out and they conferred in low voices while the second diver handed up the documentation equipment. Then Rowan came toward them.\n\nHe said: \"There's material. The carriage frame — most of it, actually, it didn't disperse much. And there's a compartment section that's structurally intact enough to survey.\" He paused. \"And there's equipment. Small electronics, the kind that would have been in passengers' luggage. Some identifiable. The divers documented what they could see.\" He looked at Elianor. \"There's a recorder housing consistent with the equipment type in the signal analysis. We couldn't access it from above — the frame is obstructing the approach — but it's there.\"\n\nShe looked at the hole in the ice.\n\n\"Recovery is possible,\" Rowan said. \"It will take specialized equipment and a more extensive operation than today's survey. Two days of preparation, a full recovery team.\" He paused. \"I'm recommending to the panel that they authorize a full evidence recovery operation.\"\n\nShe looked at Einar.\n\nHe was looking at the ice. His expression was the one she had come to know as processing — not the closed-down management of difficult emotion, but the active receipt of something real and significant that needed to be held carefully before it could be responded to.\n\nShe said: \"Thank you, Rowan.\"\n\nRowan looked at her. He looked at Einar. He said: \"I'll file the recommendation today.\" He turned to coordinate the surface team.\n\nShe and Einar were standing at the edge of the access point with the cold coming up off the water below.\n\nShe said: \"The recorder.\"\n\n\"Yes,\" he said.\n\n\"It was there the whole time.\"\n\n\"Yes.\"\n\nShe looked at the ice. At the dark water below the surface, at the pale October sky above, at the harbor visible in the distance at the edge of the lake.\n\nShe said: \"I'm not ready to hear what's on it. I need to say that now, before anyone starts talking about evidence recovery and panel documentation. I know it will need to be processed as evidence. I know I'll be part of that process. But I'm not ready for what it might contain.\"\n\nHe said: \"I know.\"\n\n\"I wanted to say it first. Before the process starts and it becomes about the documentation.\" She paused. \"It was him. Whatever is on it, it was him, in those last minutes, doing what he always did.\"\n\n\"Yes,\" Einar said.\n\nShe looked at the ice.\n\nShe thought: the recorder at the bottom of the lake, waking for eleven seconds under an aurora event, trying to transmit. Transmitting because that was what it did, because Niven had built the habit into it, because the equipment kept doing what it had been set to do even when the person who had set it could no longer tell it to stop.\n\nShe thought: the archive was designed to receive transmissions from things that could not speak for themselves anymore.\n\nShe thought: I have been preparing for this for seven years without knowing that was what I was doing.\n\nShe said: \"I'll be ready when the time comes.\"\n\nHe looked at her. The particular attention.\n\nShe said: \"I just needed to say it here first.\"\n\nHe said: \"I heard you.\"\n\nThey stood on the ice for a while longer. The surface team was packing up around them, the divers warming up, the documentation equipment being logged and stored. Rowan was on his phone, filing the recommendation. The ordinary practical work of a significant morning continuing around them.\n\nThe harbor bell was too far away to hear from here. But she knew what time it was.\n\nShe had always known what time it was in Aurevik.\n\nShe looked at the ice one more time. She thought about what was down there — the frame, the compartment, the small electronics, the recorder housing consistent with the equipment type that had transmitted for eleven seconds on a Tuesday morning at two forty-seven.\n\nShe thought: he was still transmitting.\n\nShe thought: I am still listening.\n\nShe stepped back from the edge. She turned toward shore.\n\n\"Come on,\" she said to Einar.\n\nHe stepped back too.\n\nThey walked back across the ice together, the pale October light on everything, the lake quiet behind them, the city beginning to appear at the edge of the ice as they neared the shore.\n\nShe thought about Thursday. She thought about how it would be a day she would remember for the rest of her life — the ice, the signal, the divers, the confirmation that the recorder was there. She thought about how, when she told her mother, Maera would be quiet for a long time before she said anything. She thought about how Niven would have found the technical aspects fascinating and would have wanted to understand exactly how the battery had temporarily reactivated and what the atmospheric conditions had been at the precise moment of the transmission.\n\nShe thought: I will tell him, someday. In the way I tell him things now — at the archive, in the monitoring data, in the field notebook she had logged in the collection, in the harbor bell recordings she made every morning with her phone on the kitchen windowsill. In the ways of keeping that she had learned.\n\nShe stepped off the ice onto the shore.\n\nThe ground was solid.\n\nShe was here.\n\nShe had been here the whole time.\n\nShe was glad.\n