Chapter Eleven — William | The Shape of Things

William had not set out to prove anything. He wanted, initially, only to satisfy himself that the eddy timestamps and the outcrop timestamps, now that both existed as real, recoverable data rather than a hedge in a log entry, actually referred to the same physical location and not merely the same general stretch of ocean — a confirmation exercise, nothing more, the kind of thing he did instinctively with any claim that had started to accumulate weight before anyone had checked whether the ground under it was solid.

He pulled three things onto one working sheet: the ship's own precise coordinates for the outcrop, recorded to the meter during the original survey; the storm-cell tracking data Yaz had logged, which gave him altitude and lateral position for each of the six counter-clockwise vortices at the moment of formation; and, because it seemed the obvious next thing to check once the first two were in front of him, the position of the peak spike on the comms monitor's solar feed for each of those same six windows, which the system logged automatically and which nobody, until this moment, had thought to plot against anything at all.

He converted everything to a single coordinate frame, corrected for the ship's own drift over the observation period, and drew the line.

It was very close to vertical.

He assumed, at first, an error in his own conversion — a sign flipped somewhere, a datum mismatch between the outcrop survey and the storm tracking, the kind of small, boring mistake that produced dramatic-looking results purely by accident, the sort of thing he had caught in other people's work a dozen times and had no particular reason to believe he was immune to himself. He redid the conversion from the raw data, independently, without referencing his first pass, and got the same line.

He checked it a third time, differently — not converting coordinates at all this time, working instead from the raw bearing and elevation angles as recorded by the instruments themselves, refusing to let his own arithmetic touch the numbers any more than it had to. The line held.

Six points. Six separate events, each one hundreds of miles removed from the others in time if not in space, and each one, when you dropped a plumb line from where the eddy had formed straight down through the atmosphere, through the ocean, to where the heat signature sat waiting in the rock — landed within a margin so small it was well inside the combined error bars of every instrument that had produced the underlying numbers. Not close. Not roughly aligned, the way two things drawn from noisy data were roughly aligned when someone wanted very badly for them to be. Coincident. As though someone had taken a plumb bob a hundred thousand feet long, hung it from the edge of the atmosphere, and let it settle exactly where the rock was waiting.

He sat with that for longer than he generally allowed himself to sit with anything.

The trouble — and it took him a further hour, working back through half-remembered coursework and then, when memory wasn't enough, the ship's own reference library, to be certain he had this part right — was that a vertical line was not what the geomagnetic field was supposed to hand you at this latitude. Field lines at twenty-four degrees north didn't run straight down. They ran inclined, canted well off true vertical, the same way a compass needle here didn't lie flat but dipped, because you were nowhere near a pole and the geometry of the whole planetary field simply didn't produce plumb lines this far from one. A true vertical connection, vents straight overhead to floor straight below, was the signature you'd expect standing at eighty-some degrees of latitude, not twenty-four.

Which meant that if this line meant anything mechanistic at all — and William was, even now, deeply reluctant to grant that it did — it wasn't behaving like the geomagnetic field behaved anywhere else on the planet that anyone had ever measured it. It was behaving like something that didn't care what the ordinary rules said should happen this far from a pole, and had simply gone ahead and connected the two points anyway, straight down, the shortest distance between two places that had, as far as every instrument aboard this ship could tell him, no ordinary business being connected at all.

He thought, unbidden, of a radio signal arriving faint out of the noise a few nights before, and did not enjoy the comparison, and set it aside without examining why.

He brought it to the table that evening the only way he knew how to bring anything anywhere — not as a speech, not the way Heath would have carried it in, building toward a reveal, but as a printed sheet, set down in the middle of the table between the bread and the water pitcher, six points and one line and the error bars drawn in so small they were barely visible at all.

"I wanted to be wrong about this," he said, which was, for William, close to a confession. "I checked it three separate ways specifically hoping one of them would break it. None of them did."

Amaleto picked up the sheet and studied it for a long moment, and did not, this time, reach immediately for a question that might dismantle it, which told William something about the sheet more clearly than anything he could have said about it himself.

Across the table, Heath hadn't touched his fork.

"Well," he said, quietly, and for once did not follow it with anything at all.

[Talbot, personal log, day eleven, 22:50]

Item one: instrumentation. Field-line reconstruction complete, independently verified three times by separate methods. Vertical alignment confirmed to within combined instrument error. This is not an artifact of my own arithmetic; I have made certain of that, and will make certain of it again tomorrow, because being certain once has never previously stopped me from checking a second time and I see no reason to change that practice now, of all nights.

Item two: crew dynamics. No concerns.

Item three: correspondence. Not tonight.

Item four: personal. I do not have a category for what I am feeling about item one, which is itself worth recording, because I generally have a category for everything, and the absence of one is, I think, the most honest data point in this entire log.

End log.