← The land

Source SRC-200

Beaufort West NG Gemeente, Baptisms 1819-1844 (doopregister)

church register · consulted 2026-08-02

Type
church register
Repository
FamilySearch, "South Africa, Dutch Reformed Church Records (Stellenbosch Archive), 1690-2011"; citing Genealogiese Instituut van Suid-Afrika (GISA), Stellenbosch
Reference
South Africa > Cape > Beaufort West > Baptisms 1819-1844 > 302 images. Waypoint wc=M64H-PNL:13850901,13850902,14731201,14762701; cc=1392488. Last image (302 of 302) is ark:/61903/3:1:S3HY-68TW-GKL. Held here: 91 distinct page images, filed under their ark suffix in sources/registers/beaufort-west-baptisms-1819-1844/
Consulted
2026-08-02

Open the record ↗ the permanent identifier, not a search URL

The citation as it was given

Supplied by , 92 downloads of which two were the same image (GS2), so 91 distinct pages. TRANSCRIPTION ON HOLD, entry by entry, into data/transcriptions/bw-baptisms-1819-1844.csv - one row per baptism with child, birth, baptism, both parents, all witnesses, the veldkornetskap and the page. COLUMNS OF THE REGISTER: Kind | Geboren | Gedoopt | Ouders | Getuigen, with the page number written at the head. THE LINE UNDER THE CHILD IS THE VELDKORNETSKAP, not the minister. Page 265 spells it out - "Veldcornetschap van der Walt, District George" - and the same position on page 264 carries "C de Clercq", "C J de Clercq", "S Esterhuizen". So every entry carries the WARD the family lived in, which makes this register a source of PLACES as well as names. SCAN DEFECT, and it is not uniform: some pages are cropped at the RIGHT (page 264, losing witness surnames) and some at the LEFT (page 265, losing the child's first name). Truncations are marked [truncated] or reconstructed in [brackets], never silently completed - 5a. The images have no sequence in their ark suffixes; PAGE NUMBER, written on each sheet, is the ordering key. | [3 Aug 2026] 35 OF 91 PAGES DONE, 419 baptisms. Frans paused the work: 'baptism transcriptions on hold. put files in folder with note that you should not touch it till explicitly asked.' The brief, the remaining batch lists and the raw agent output are in _ON-HOLD-baptism-transcription/, with the rule written down. OPEN QUESTION TO SETTLE BEFORE RESUMING: several sheets carry two page numbers, a printed stamp and a handwritten one consistently +2. Everything transcribed so far uses the STAMP, checked against S3HY-68TW-G25 = page 230, so the corpus is internally consistent - but which one FamilySearch's own image numbering follows is unconfirmed. | THE PAGE NUMBER: THE STAMP, NOT THE HAND. (Settled 3 Aug 2026, before transcription resumed.) Several sheets carry TWO numbers - a printed stamp, and a handwritten number in the top corner that is consistently the stamp PLUS TWO. Every page transcribed so far, by hand and by batch, has used the stamp, and that convention stands. The reason is not tradition: the register runs to 302 IMAGES on FamilySearch, and the stamped numbers on the pages transcribed run 216 to 299 - so the stamp tracks FamilySearch's own image sequence almost one to one. It is therefore the number that actually helps a reader FIND the image, which is the whole job of a citation. The handwritten +2 is a later foliation and is recorded here once rather than on 419 rows. If it is ever shown to be the canonical one, the correction is a single constant applied to the whole corpus, not a re-transcription - which is why this was not allowed to block.

Kept verbatim. This project contradicts published genealogies and corrects a family's own oral tradition, so every claim has to be checkable by someone who was not in the room.

What has been transcribed from this source

50 occasions. 46 of them are not yet tied to a person in this model - the names come off the page half-legible and matching them is research, not filing.

What rests on this source

People