Diagram verso, 1732/1863 (verso)

The back of Part 1, carrying the servitude-noting box Part 1 points at. Nearly blank: a printed table headed [?SURVEY RECORDS] / [?SERVITUDES], one manuscript line - 'The lines pq and rs represent the new and old flood level line respectively' - and a reference written [?52/68].

THAT REFERENCE IS NOT READ AND IT MATTERS FOR THE GATE. If [?52/68] is a 1968 endorsement then this sheet carries a date after 1936 and should NOT inherit its front's 1862 dates. I cannot read it, so no date is asserted either way and no publish_ok is set. A flood line is a fact about GROUND, not about a person, so 11's principle would release it once somebody has actually read the box. LEAD-0280.

[23 Aug 2026] publish_ok=yes ON HAVING READ THE SHEET, not on its date - 11, 'never release the category, release the sheet'. I opened this file: it holds a printed table box, the single line about the flood level, and the unread reference. IT NAMES NO PERSON AT ALL, and the 1936 cut-off is a proxy for whether a sheet names somebody living. A flood line is a fact about ground. The reading of [?52/68] is still open (LEAD-0280) and publish_ok does not settle it - if it turns out to be a 1968 endorsement the DATE fields must still be corrected, because the row states what is on the paper whatever the gate does.