Corrections & Errata
We do not quietly fix things. When the catalogue is wrong we say so here, with a date on it, and the wrong version stays in the revision history where anybody can find it. This is not a courtesy. It is the only reason a file from 1952 is worth anything at all.
What is and is not a correction
A correction is a demonstrable factual error in our catalogue: a date, a place name, a transcription, a custody record, a duplicate, a classification that the underlying paperwork does not support.
A correction is not your disagreement with a classification. It is not a new theory about the name. It is not a letter explaining that we have been too sceptical, and it is not a letter explaining that we have not been sceptical enough. I receive roughly one of each per month and I file them together.
Corrections with a file number go to [email protected]. Corrections without a file number go nowhere, slowly.
Classification corrected: Probable to Disputed
The 1993 Hazleton displacement file sat coded "Probable" for twenty-two years on the strength of a 2004 spreadsheet export that nobody checked against the paper. The original homeowner follow-up records a damaged gate latch and a windy night, neither of which survived into the later summary. Now Disputed. The witness statement itself is untouched, as always.
Encounter date corrected
Newsletter Vol. 8 No. 2 printed 12 May 1997. The signed form, the telephone log, and the postmark all say 14 May 1997. Database amended. The newsletter scan is not amended, because scans of things are not amended.
Location label standardised
Early summaries alternated between "Bangor" and "Upper Mount Bethel." A parcel review put the bridge approach outside the boundary one field notebook had assumed. The public database now says Bangor, PA. The original site sheet keeps the precise parcel notation and stays restricted, because it is somebody's address.
Custody status amended
A 2023 evidence list called the Reading fibre bundle "laboratory inconclusive." That was wrong and it mattered. Comparison was discontinued when the packaging was compromised in transit; no finding of any kind was ever issued. "Inconclusive" implies a test happened. Nothing happened. The catalogue now reflects that.
Duplicate photograph removed
Second-generation scan of a frame already catalogued on the same roll. Removed from the public image count, retained in the accession folder with a duplicate marker.
Volume numbering
The 2004 standardisation index listed the January 1994 issue as Vol. 4 No. 5. The masthead and the mailing ledger both say Vol. 5 No. 1. The Newsletter Archive uses the corrected citation. Nine other citations elsewhere on this site were fixed at the same time; if you find a tenth, tell me.
Witness count withdrawn
Somewhere between 1952 and a retyped index in the 1980s, the Upper Lehigh livestock report acquired three direct witnesses. The surviving correspondence supports one named observer and two people who heard about it afterwards. The number has been removed from the public summary and will stay removed until the missing chapter sheet turns up, which after seventy years I do not expect.
I raise this one whenever a member tells me the old records speak for themselves. They do not. They speak for whoever retyped them last.
Submitting a correction
Use the intake form, select an existing case reference, and say where your correction comes from — a document, a postmark, a parcel map, a person who was there. "I think" is not a source.
This page is updated as corrections are issued, not on a schedule. Last entry 14 February 2026.