before 18 June 1965
bound · probable
Reasoning
The same sentence already cited above for the lower bound continues: '... to last Friday night when she told me' -- naming the span's upper edge in Neville's own words, relative to the moment of speaking. Lecture nid 1213's own delivery date, 1965-06-22, is certain (independently corroborated: legacy catalog #287 and a site chrome-header dateline '6/22/65' that explicitly ruled out a nearby WP post-stamp as a trap). Per rule 18, a weekday is computed from a date already in the packet: Zeller's congruence on 1965-06-22 (q=22, m=6, K=65, J=19) gives h = (22 + floor(13*7/5) + 65 + floor(65/4) + floor(19/4) - 2*19) mod 7 = (22+18+65+16+4-38) mod 7 = 87 mod 7 = 3 = Tuesday. Cross-checked two further ways: offset from 1965-06-04 (independently established as a Friday in this run's sibling packet, hist_event 6470): 22-4=18 days, 18 mod 7=4, Friday+4 days=Tuesday, agrees; and direct Zeller's congruence on 1965-06-18 itself (q=18, other terms unchanged) gives h=(18+18+65+16+4-38) mod 7 = 83 mod 7 = 6 = Friday, exactly 4 days before the Tuesday lecture and exactly 14 days (two weeks) after the sibling packet's Friday 4 June anchor -- four independent checks agree. So 'last Friday night', relative to the Tuesday 22 June 1965 lecture, resolves unambiguously to Friday, 18 June 1965. Recorded as a bound (qualifier before) rather than a single-day exact claim, since the event is a multi-year series; this narrows the standing 'before 1965' bound (year precision, npr 95306) to day precision without contradicting it. Confidence held at probable, not certain, for two reasons visible in this same passage, neither touching the arithmetic itself: Neville immediately adds 'She could have told me two weeks before,' showing the disclosure's exact timing was not pinned down tightly even by his own account; and the two now-dated bookends span only about two and a half years, well short of the 'three-and-a-half years' the passage states for the unveiling immediately beforehand, so this date most plausibly marks an approximate personal anecdote rather than a precise measure of that stated duration. Both observations sit in the passage already cited, so they cap this claim below the anchor's own certain ceiling without importing anything from outside the packet.