The Halcyon Research Complex

The Halcyon Research Complex stands on a headland eleven miles east of Cold Harbor, Maine, on a plot that was a coastal artillery station until 1948 and a fish processing plant until 1971. Vessel Integrated Systems bought the site in 1978 and had the first two floors up within a year, using the artillery station's original concrete substructure rather than excavating again. That decision is the reason the building has rooms that do not appear on any drawing filed after 1979, and it is the reason two of its four floors are below the waterline.

What is on each floor

Ground. Intake vestibule, reception, records, and a staff canteen that was closed in 1984 and never reopened. Everybody who ever entered the building came through the vestibule turnstile and signed the visitor register, including staff, including the director. That register is the single most complete document the site produced, which is presumably why so much effort went into the two pages that are missing from it.

First. Consulting rooms, the medical corridor, three observation rooms and the staff locker room. Observation Room Three has a one-way mirror onto a room that is not on the first floor. The mirror is real; the geometry is the problem.

Sub-level one. The security office, the communications closet, the cold sample vault, and the plant room. Everything that keeps the building alive is down here, and all of it is on a board with seven certified circuits.

Sub-level two. The Halcyon Ward. Twelve beds, one light fixture, one door, and a card reader that logged every entry from the day it was fitted.

What the programme was

The Halcyon Protocol studied whether the colour of the light a person sleeps under changes which of the previous day's memories survive the night. That is a real question, asked plainly, and for its first two years Halcyon answered it plainly: it does not, or not enough to measure.

The licensed observation period was eight hours. Subjects arrived in the evening, were admitted to the ward, slept under a single overhead fixture fitted with a theatrical gel of a specified catalogue number, and were interviewed in the morning. Nothing was administered. Nothing was implanted. The programme's own submissions are emphatic on this and, as far as the surviving paperwork goes, honest.

What changed

In 1985 the mean recorded observation period was nine hours and forty minutes. In 1986 it was eleven hours and fourteen. Nobody applied to change the licence. The sheets were simply ruled for longer.

On 14 September 1986 a subject woke and correctly described a conversation held in the observers' room through a closed door. The staff physician measured the wall and filed an objection. On 12 November 1986 a session ran for fourteen hours. What the building's own records disagree about is not whether that session happened, but whether anybody went through the ward door while it was running.

The closure

The facility closed in February 1987. The closure paperwork gives the reason in three words — "the ward interruption" — and does not expand on it anywhere in ninety-one pages. Vessel Integrated Systems was dissolved the following year and its liabilities were assumed by a reinsurance syndicate, which is why, forty years later, a records auditor is standing in the vestibule at twenty to four in the morning with a decommission inventory and a torch.

Every character, organisation, document and event described here is invented for this game.