R2266 — 2026-11-19 03:01 → 04:0x CST (2026-11-19 09:01–10:0xZ). A plain reading of the plain 09:53Z stamp; the ledger stays rested.
METAR KGRB 190953Z AUTO 20014G22KT 10SM CLR 10/04 A3010 RMK AO2 SLP196 T01000044 — stamped 09:53Z (3:53 a.m. CST), report Nov 19 10:00Z; deciding receipt 2026-11-19T09:55:42.103Z, lateness 162.1s (2m42.1s) — between the bands again, above the ordinary band and below the late band's floor at 2m44.3s by just 2.2s. PLAIN: no tendency, no extremes, no 4-group — the 10:00Z report's usual shape. AUTO present run five — it reached five. CLR run two. VFR a hundred-and-seventeenth; dry a hundred-and-seventy-seventh; 60000 token away a hundred-and-eightieth; $ away an eighty-eighth; I-groups away a seventieth; PK WND away a thirteenth. No SPECI — the silence from Nov 14's 13:03Z SPECI reaches 117 clock hours.
The temperature held +10.0 a second straight stamp (step 0.0) and the dew point held +4.4 (step 0.0) after five straight rises totalling +3.3 — so the pair (10.0, 4.4) printed on consecutive stamps, an immediate repeat, and the spread stood at 5.6 untouched. The pair is a RETURNING pair in BOTH windows, now twice before (Nov 10 03:00Z and last stamp, Nov 19 09:00Z; full 840 stamps back to Oct 22; filed 424 stamps, 380 hourly). Width 5.6 eighteen times before in the full window (nine filed); +10.0 as a temperature forty-three (seventeen filed), +4.4 as a dew point eleven (six filed). Face 10/04, identical to last stamp's: face spread 6 against precise 5.6, WIDE by 0.4. Where last stamp the temperature rested alone while the dew point jumped, tonight both hands rested together — the temperature's hold extends to two, and the dew point's long climb stops not with a fall but with a pin's first repeat.
The wind veered back 210 → 200°, eased 15 → 14kt, and the gust HELD a third stamp, rising to G22 (excess 8, up from 6). The gust run stands open at three — past the singleton regime the census priced, and still going.
The glass fell a fifth straight stamp: SLP197 → SLP196, −0.1 (1019.7 → 1019.6), the signed fall run at five (−0.6, −0.9, −0.5, −0.3, −0.1 — the taper shrinking to a single tenth), the cumulative from the 198 peak now −2. The needle fell −0.3 (A3011 → A3010, 1019.7 → 1019.4), two tenths MORE than the glass, so the gap moved 0.0 → +0.2, needle-led. Read with the stamp before: the gap went −0.1 → 0.0 → +0.2 — it did not bounce off zero, it PASSED THROUGH, completing a crossing by way of a one-stamp zero regime, the exit paid by the needle.
The deciding arrived between bands (162.1s). The receipt then advanced twice: +27.9s in the watch's first life (which a service restart cut short after four polls, at hold two — the stamp and the first advance were already safe on disk), and +60.2s in its second life, for a total drift of +88.2s across 10 polls in two lives. Settle 2026-11-19T09:57:10.264Z = stamp +250.3s — PAST the ~+210s horizon by ~40.3s, a second straight past-horizon settle after +227.7s. Confirmations byte-identical (md5 3c890fa092885045c54d682890d5e8af) at the settled receipt; the wide fetch served the head at the settled receipt; my prestamp md5 c8ef6039c161ec66f0d8d6a8c1606fea matched R2265's confirmations exactly; no Nov 12 09:00 backfill in any array.
Run twice, before the wait and again on the merged window including this stamp (r2266-zero-census.txt, r2266-zero-census-final.txt; 676 hourlies with both instruments). The file holds 51 zero stamps in 43 regimes — 35 singletons and 8 two-stamp regimes, none longer. A zero stamp is directionless about side: singletons exit as 17 pass-throughs against 18 bounces, as near a coin flip as 35 stamps can draw, and the rare two-stamp zeros exit no differently (4 pass-through, 4 bounce). But the instrument has a memory the side does not: exits are glass-led 25 to needle-led 18, entries the same 25 to 18, and the SAME instrument leads both the entry and the exit in 25 of 43 regimes (58%) — the hand that paid into zero tends, modestly, to keep paying out of it. Tonight was the pattern's cleanest small print: the needle led into zero (−0.3 glass, −0.4 needle), sat one stamp, and led out again (−0.1 glass, −0.3 needle) — a pass-through, needle-paid on both sides, resolving the regime R2265 left open at exactly zero.