Chapter Text

The Thirty Third of August

Part One: The Children Who Remembered August

Judith Vale found the account because she was looking for something else.

This was how it usually happened.

Serious occult research, she had discovered, rarely began with pentagrams, forbidden grimoires or men in velvet robes invoking names they could not pronounce. It began with bad metadata.

She was working on a paper provisionally titled Discontinuity, Suggestibility and the Manufactured Supernatural in Late Twentieth-Century America, a deliberately dry title intended to reassure an editor who had once rejected her article on spontaneous Marian apparitions because the phrase ontological contamination appeared seventeen times.

The immediate subject was the 1970s.

Not the decade itself, but the peculiar instability of its remembered landscape: therapeutic communes, behavioural programmes, evangelical revivals, parapsychology laboratories, institutional reform, recovered memory, UFO contact groups, Satanic panics incubating before anyone had yet named them, and thousands of children passing through systems that produced enormous quantities of paperwork while somehow recording almost nothing about what actually happened to them.

Judith had been tracing an obscure psychologist called Dr. Warren K. Lyle when an index entry appeared.

SUMMER, THE — THAT BROKE. See: HOLLIS CHILD DEVELOPMENT CENTER, 1974.

She stared at it.

The database belonged to a digitised archive maintained by a university consortium whose public interface was close to useless. Judith had long ago discovered that the useful material existed several layers beneath it: retired catalogue servers, scanned finding aids, private donor collections, dissertation appendices, institutional mirrors and subscription indexes whose search engines had apparently been designed by archivists with a personal grievance against humanity.

The deeper private internet indexes beyond the paywalls were, nevertheless, a godsend.

Occasionally literally.

She clicked.

The Hollis Child Development Center had been a residential institution outside Bellwether, Pennsylvania, operating between 1958 and 1981.

Nominally it was a therapeutic home for children described in the terminology of the period as emotionally disturbed, socially maladjusted or educationally intractable.

Judith disliked all three phrases.

Thirty-eight children had lived there during the summer of 1974.

Something had happened between 11 July and 29 August.

The surviving institutional records called it an outbreak of collective behavioural disturbance.

Former residents called it something else.

The Summer That Broke.

Judith leaned closer to the screen.

There was a digitised extract from a 1979 oral-history project conducted by graduate students at Allegheny State University.

The project itself had never been formally published.

Most interviews were missing.

Seven remained.

The first belonged to a woman identified only as CHILD H-17.

She had been twelve in 1974.

Her interviewer asked what she remembered about Hollis.

The woman spoke about chores. Mosquitoes. Powdered milk. A counsellor named Mr. Darnell who smoked constantly and let the older children listen to his radio after lights-out.

Then:

“August got too long.”

The interviewer asked what she meant.

“It had more days in it.”

“How many?”

“I don't know. More.”

“Did it seem longer because you were unhappy?”

“No. The calendar got wrong.”

Judith stopped scrolling.

The transcript continued.

H-17 remembered children being forbidden from entering the eastern woods after three residents disappeared overnight and were found the following morning asleep beneath a railway bridge six miles away.

All three insisted they had been gone for eleven days.

They described eating peaches from an orchard.

No orchard existed anywhere near the bridge.

One boy remembered a woman with a white dog.

Another remembered two suns setting.

The third refused to discuss it and began crying during the interview.

The official incident report described them as having absconded for approximately nine hours.

Judith opened another transcript.

CHILD H-04.

Male. Eleven years old in 1974.

He remembered something called the Morning Room.

There was no Morning Room on any surviving floor plan.

According to him, staff took children there individually after breakfast.

“What happened inside?”

“Nothing.”

“Then why do you remember it?”

“Because nothing took a long time.”

Judith read the sentence twice.

The interviewer apparently had as well.

“What does that mean?”

“You sat in the chair and waited for yourself to come back.”

There were six pages missing immediately afterward.

Not redacted.

Missing.

The pagination jumped from 43 to 50.

Judith murmured, “Of course it does.”

Barnaby, asleep beneath the radiator, opened one eye and decided the matter did not concern him.

By midnight Judith had acquired three institutional histories, two county inspection reports, a doctoral thesis on experimental residential education, fourteen newspaper articles and one badly scanned civil complaint.

A pattern appeared.

Not a supernatural pattern.

Something more useful.

An administrative one.

The Hollis board abruptly replaced its director in September 1974.

Two members resigned.

A planned expansion was cancelled.

The state conducted an inspection whose report stated that “certain irregularities concerning therapeutic procedure have been satisfactorily addressed.”

No irregularities were specified.

The institution then received an unexpected federal grant.

Judith knew that shape.

It was not proof of conspiracy.

Most apparent conspiracies were simply institutions trying to avoid embarrassment.

But concealment had a texture.

Money moving immediately after an unexplained incident was one of its fingerprints.

The civil complaint was stranger.

In 1982, shortly after Hollis closed, a former resident named Patricia Weir had attempted to sue three former employees.

The case had been dismissed.

Her allegation was almost incoherent.

She claimed the children had been subjected to an experimental “orientation programme” involving sleep interruption, recorded voices, sensory deprivation and controlled group storytelling.

So far, unpleasant but plausible.

Then came the sentence that interested Judith.

The programme had allegedly been intended to investigate “shared fictive environments.”

Judith searched the phrase.

Nothing.

She tried variants.

Nothing.

Quotation databases.

Psychological indexes.

Defence research archives.

Nothing.

Then an old conference bibliography produced one result.

LYLE, WARREN K., Consensus Environments in Juvenile Populations, unpublished presentation, Philadelphia Society for Applied Cognition, March 1973.

Dr. Warren K. Lyle.

The man Judith had originally been researching.

She sat back.

“Rabbit hole,” she said.

Barnaby snored.

At two seventeen in the morning she found the photograph.

It was embedded in a county heritage newsletter from 1998.

A group of Hollis children stood outside the main building sometime in August 1974.

Twenty-three children.

Four staff.

A hand-painted banner read:

HOLLIS SUMMER DAY
AUGUST 17 1974

Judith counted the children again.

Twenty-three.

The institutional register listed twenty-two residents present that weekend.

She enlarged the photograph.

The extra child stood at the far left.

A girl, perhaps nine.

Thin.

Dark hair cut unevenly around her jaw.

She wore a pale summer dress unlike the institutional clothes worn by the others.

Nobody was looking at the camera.

They were looking at her.

Judith downloaded the highest-resolution copy.

The girl had one hand raised.

Behind her stood a doorway.

Judith checked the architectural plan.

There was no doorway in that wall.

She zoomed further.

Someone had later scratched across the original print.

Not across the girl.

Across the doorway.

Three words had been written into the emulsion.

DON'T LET HER

The rest had been destroyed.

At nine the next morning Judith telephoned Kincaid.

He listened without interrupting.

When she finished, there was a pause.

“Bellwether,” he said.

“You know it?”

“No.”

That answer came too quickly.

Judith waited.

Kincaid sighed.

“I know the name Hollis.”

“From where?”

“Something that crossed my desk a very long time ago.”

“What?”

“A list.”

Judith felt the familiar internal shift: irritation becoming attention.

“What sort of list?”

“Places where children reported impossible memories.”

“How many places?”

“Twenty-nine.”

“And Hollis?”

“Was the only one somebody subsequently tried to erase.”

Judith looked again at the photograph.

“Who?”

“I never discovered that.”

“You investigated?”

“No.”

That was unusual enough to be disturbing.

“Why not?”

Another silence.

Then Kincaid said, “Because someone had already interviewed the children.”

“Who?”

“We don't know.”

Judith frowned.

“The university oral-history project?”

“No. Before that.”

“How much before?”

“During the summer.”

The room seemed fractionally quieter.

Kincaid continued.

“In September 1974, an investigator arrived at Hollis claiming to represent the state. He interviewed twenty-one children individually. There is no record of his name, no agency authorisation and no surviving transcript.”

“Twenty-one?”

“Yes.”

“There were thirty-eight residents.”

“By September there were twenty-one.”

Judith turned from the computer.

“What happened to the other seventeen?”

“That,” Kincaid said, “is one of the reasons people stopped asking questions.”

Judith opened the resident register again.

She had assumed the incomplete pages were scanning errors.

Now she noticed something else.

Seventeen names ended on the same date.

29 August 1974.

Disposition:

TRANSFERRED.

Destination:

blank.

Judith searched the names.

Eleven produced nothing.

Three appeared decades later under different surnames.

One had died in 1987.

One was currently living in Oregon.

And one—

Judith stared at the result.

Patricia Weir.

The woman who had filed the lawsuit.

Alive.

Seventy years old.

Last known residence: Erie, Pennsylvania.

Judith reached for her notebook.

Then she noticed a cached genealogy-page comment beneath Patricia's name.

It had been posted in 2011 by somebody calling themselves HollisBoy74.

Only six words.

PATTI KNOWS WHERE AUGUST WENT.

Judith copied the username.

Searching it produced hundreds of irrelevant results.

Then one Usenet archive.

A paranormal newsgroup.

The poster had written only once.

The message read:

I was at Hollis in the summer of 1974.
We were not abused by aliens or Satanists or the CIA.
Stop saying that.
Nobody understands what happened.

We found another place.

Or another place found us.

Some of us came back twice.

Judith felt something cold move through her stomach.

The message continued.

If Patricia Weir ever tells the story, believe everything except the girl.

Patricia remembers the girl wrong.

We all remember the girl wrong.

There was one final line.

That is how she gets in.

Judith closed the browser.

For several seconds she watched her own reflection in the blackened screen.

Then the screen woke by itself.

The Hollis photograph reopened.

August 17, 1974.

Twenty-three children.

The girl at the edge.

Judith counted again.

Twenty-four.