Chapter 3 Six Months in the Wrong Direction

She did not sleep that morning. She went to her room and lay on the bunk and watched the window go from black to a blue that at this height had no warmth in it, and then she got up and went to the archive. The archive room was the coldest place on the station. It held the survey's eyes: hundreds of glass photographic plates in numbered sleeves, made by the Observatory's first astrograph before the digital era, and beside them the modern data drives, so that the whole history of how this mountain had watched the sky stood in one small room and could be read end to end. Viveka wanted the oldest plates of field seven. If RV-417 had been carried in the catalogue for thirty years, it had to appear on the plates from thirty years ago, and if it did, then the source was real and physical and simply strange. If it did not, then the catalogue had been seeded with it, and the seeding was a thing worth understanding. The plates were heavy and cold and smelled of acetic acid. She set them on the light table one at a time and put her eye to the glass, and there in the old emulsion, in the small corner of the field where the catalogue put it, was a faint smudge of an object, present on every plate she checked. She found the earliest one. Her breath fogged the glass and she wiped it with her sleeve. The date on the sleeve was from the first years of the survey. And there, in the same position, was the same faint smudge, unblinking, exactly where the current night's data put it. Thirty-five years of plates, and the object had not moved. This is where a careful person stops and understands what she has been looking at. Everything in the sky moves. The stars move against each other across the plate, slowly, a barely perceptible drift measured in fractions of an arcsecond per year, which is why the survey existed at all: to catch the drift. A star that showed no drift over thirty-five years of plates, no proper motion at all, was already improbable, because the sky is full of stars and they are all going somewhere. But a star that showed no parallax was impossible. Parallax is the geometry of the sky's depth. The Earth goes around the Sun, and so the station, riding the Earth, swings across a baseline of three hundred million kilometers every six months, and every star, no matter how far, appears to shift a little against its background because of that swing. The nearer the star, the bigger the shift. The shift is small, but it is not zero, and the survey measured it precisely, and Viveka had spent her career on those small numbers. A star at ten parsecs shows a shift of about a tenth of an arcsecond. The survey could see a hundredth. Even a distant giant, far across the galaxy, showed something, a whisper, a hundredth of that shift, and the survey's whole purpose was to gather those whispers and turn them into a map. RV-417 had no shift. Its position on the plates from six months apart, six months again, six months again, was identical to a precision the survey was proud of. The object was not merely far away. It was at a distance that geometry could not assign it, because the shift was smaller than the measurement, and the measurement was good. She checked her logic twice. Then she went back through the data and did the thing she had come here to do, the thing that made her the right person for the shock that was coming: she looked for the systematic wobble her old paper had predicted, the one the reference stars should show. And there it was, exactly as she had argued, a faint coherent drift in the residuals of the reference frame. Her signal. Real. And RV-417 sat at the middle of it like a stone in the current, absolutely still. Her eye caught a name while she worked. The early plates were signed in the margin by whoever had exposed them, and the early margin carried the same four letters across the first years of the survey, month after month, in one small careful hand: M. Tornquist. Viveka had read the name in the catalogue notes and had not thought about it. Now she looked at the accumulation of it and felt the presence of a person who had stood where she was standing and looked where she was looking and had seen the same perfect star, and had gone on with the survey anyway. The archive was cold enough that she worked in her coat with the collar up, and the cold had a use, because the plates liked it, and the drone of the building did not reach in here, and the silence let her hear her own pencil. It was the kind of room a person could vanish into for a season, and she understood, sitting in it, why a survey that measured the same half-million stars for thirty-five years needed someone like her, and why a person like her might stay. She sat back on the cold stool and pressed her palms against her eyes. At her old site she had once spent four nights chasing a residual that turned out to be a crack in a filter. She had learned, there, to love the moment when an anomaly collapsed into an ordinary cause. This was not that moment. She did not know yet what this was, but she knew it was not a clerk and not a filter. By that evening she had two facts and no third. Fact one: the object had no proper motion and no parallax over thirty-five years of plates. Fact two: the object was in the catalogue, present and carried forward, and every observer who had ever reduced field seven had seen the same impossible stillness and had written it down as low weight and moved on. She went to the control room and found Emeric running the twilight checklist. "RV-417," she said. He looked up, and his expression did the smallest thing, a flicker she almost missed. "Field seven. The low-weight one." "You know it." "Everyone who has run the Vigil knows it. It was Tornquist's object originally. She used to say it was the perfect star, and that a perfect star is a joke the sky tells a survey." He went back to his checklist. "Some of the old catalogue entries are like that. They survive because nobody bothers to remove them." "You've never measured it properly." "I have measured the survey, Skoglund. Four-thousand nights' worth of it, over my career, at three sites. There are a dozen entries like four-one-seven, and they are all fine, and none of them are the sky. They are us. Our mistakes, carried forward. It's the ordinary cost of a long program." "You're telling me you've seen this exact zero residual before." "I'm telling you I have seen the survey disagree with itself, and the answer was always the survey." He set down his clipboard. "It's your first week. Sleep. Then reduce the field again with a clear head and a clean reference set and see what you find. Nine times in ten, an anomaly at altitude is altitude." She went to her room instead and sat at the desk and wrote the two facts in her own notebook, in pencil, the way she had written every important thing for twenty years. Then she added a third line, because she could not help being exact. *Fact three: it occulted nothing tonight, but it is worth checking whether it ever passes in front of a background star. If it does, it is near. If it is near, everything above is wrong.* She read the line back. Outside the window the ridge went dark, and the first stars came out, and she found herself doing the thing she had never done in twenty years: looking at the sky and waiting for it to move. ---