Yves Scanlon lies on his bunk, not listening. He does not hear any strange sounds filtering in through the hull. There is no reedy, spectral keening from the seabed, no faint sound of howling wind because he knows that, down here, no wind is possible. Imagination, perhaps. A trick of the brain stem, an auditory hallucination. He's not the slightest bit superstitious; he's a scientist. He does not hear the ghost of Karl Acton moaning on the seabed.

And now, concentrating, he's quite certain he hears nothing at all.

It really doesn't bother him, being stuck in a dead man's quarters. After all, where else is there? It's not as though he's going to move in with one of the vampires. And besides, Acton's been gone for months now.

Scanlon remembers the first time he heard the recording. Four lousy words: "We lost Acton. Sorry." Then she hung up. Cold bitch, Clarke. Scanlon once thought something might happen between her and Acton, it was a jigsaw match from the profiles, but you wouldn't know it from that phone call.

Maybe it's her, he muses. Maybe it's not Lubin after all, maybe it's Clarke.

"We lost Acton." So much for eulogy. And Fischer before Acton, and Everitt over at Linke. And Singh before Everitt. And—

And now Yves Scanlon is here, in their place. Sleeping on their bunk, breathing their air. Counting the seconds, in darkness and quiet. In dark—

Jesus Christ, what is—

And quiet. Everything's quiet. Nothing's moaning out there.

Nothing at all.

* * *

TRANS/OFFI/220850:0945

We're all mammals, of course. We therefore have a Circadian rhythm which calibrates itself to ambient photoperiod. It's been known for some time that when people are denied photoperiodic cues their rhythms tend to lengthen, usually stabilizing between twenty-seven and thirty-six hours. Adherence to a regular twenty-four hour work schedule is usually sufficient to keep this from happening, so we didn't expect a problem in the deep stations. As an added measure I recommended that a normal photoperiod be built into Beebe's lighting systems; the lights are programmed to dim slightly between twenty-two hundred and oh seven hundred every day.

The participants have apparently chosen to ignore these cues. Even during 'daytime' they keep ambient lighting dimmer than my suggested 'nocturnal' levels. (They also prefer to leave their eyecaps in at all times, for obvious reasons; although I had not predicted this behavior, it is consistent with the profile.) Work schedules are somewhat — flexible, but this is to be expected given that their sleep cycles are always shifting in relation to each other. Rifters do not wake up in time to perform their duties; they perform their duties whenever two or more of them happen to be awake. I suspect that they also work alone sometimes, a safety violation, but I have yet to confirm this.