Kaart se keersy, 922/1872 (verso), 6 Junie 1872

[5 Aug 2026 inbox, re-pulled per LEAD-0095] Verso of the Rondabel Portion 1 diagram: the deduction table, and it NAMES SIX PORTIONS with their diagrams, extents and deeds. Ptn. 2 'Portion A', 2107 mg 267 sr, deed 1900-14.1082; Ptn. 3 'Brand Kraal', 4 mg 53 sr, deed 1910.133.8526; Ptn. 4 'Driehoek', dgm 269/1919, 536 mg 14 sr, deed 10.6.1919.5387; Ptn. 5 'Hartebeest Voedpad', dgm 270/1919, 1606 mg 79 sr, deed 10.6.1919.5388; Ptn. 6 'Elandsfontein', dgm 891/1921, 2705 mg 257 sr, deed 8.7.1921.6443; Ptn. 7 'Klein Elandsfontein', dgm 3273/33, 1544 mg 361 sr, deed 16.5.1934-3875. PUBLISHES on 1934, the newest line in the table, which is inside the 1936 cut-off. The gate reads the newest ink and that is what it read here.

[5 Aug 2026] MERGED DOC-0211 INTO THIS ROW - one sheet, two rows. Byte-identical files (MD5). DOC-0211 was a 3 Aug orphan-scan stub; DOC-0241 transcribed the six-portion deduction table 8 cites. DOC-0211 is retired and its id will never be reused. Moved across: survey_edtf=1872-06-06; image_edtf=1872. DISAGREEMENTS between the two rows, recorded rather than resolved: source_file: kept 'sources/diagrams/DOC-0241-rondabel-124-ptn1-1872-verso.jpg', DOC-0211 said 'sources/diagrams/DOC-0186-rondabel-124-ptn1-1872-verso.jpg'; sg_diagram_no: kept '1872 Rondabel Ptn 1 (verso)', DOC-0211 said '922/1872 (verso)'; title_edtf: kept '1934-05-16', DOC-0211 said '1868-01-21'; source_id: kept 'SRC-283', DOC-0211 said 'SRC-245'. DOC-0211's own note read: THE RONDABEL DEDUCTION TABLE - the sheet 8 cites for naming six portions in one pass and giving DRIEHOEK to PAR-000117, which had carried no name and no extent. It has been on disk and off the site since it arrived, because its filename was never written into any row. Sheet of DOC-0186; dates are that sheet set’s. Row created 3 Aug 2026 after the new orphan-scan check found the file on disk with nothing pointing at it.

[5 Aug 2026] sg_diagram_no rewritten from '1872 Rondabel Ptn 1 (verso)' to '922/1872 (verso)'. 8 makes the string load-bearing, not descriptive: it must be the FRONT'S string plus ' (verso)', because _doc_year() prefix-matches on that stem to find the front. A descriptive string matches no front, so the sheet falls back to its own dates or to nothing. The front here is DOC-0186.