The muster roll
Nine fields in a Dutch East India Company ledger: a young man from Woerden - or Werden, or Werl - sails from Texel owing the Company his passage, sees Table Bay 127 days later, and five years on stops being a servant and becomes a settler. Everyone in this project descends from that decision.
Texel to Table Bay, and what happened to three sons' worth of descendants, 1685 to 1838
One line in a ledger
The Dutch East India Company kept a book of everyone who sailed. Name, home town, the job you signed for, the ship, the day you started, the day you stopped, and why. Nothing else. No age, no description, no reason for going. Nine fields, and every line in this history comes down from them:
Voornaam opvarende Johan Achternaam opvarende Oosthuijsz Herkomst opvarende Woerden Datum indiensttreding 13-05-1685 Functie bij indiensttreding Adelborst Uitgevaren met het schip Land van Schouwen Datum uit dienst 15-08-1690 Waar uit dienst Kaap de Goede Hoop Reden uit dienst Vrijburger Schuldbrief Ja · Maandbrief Nee
What the nine fields say
Adelborst. Officer cadet, in Company usage a junior soldier of some standing, generally a young man in his teens or early twenties. That single word is the only evidence of his age anyone has, and it is why the compiled trees cannot agree: one says he was born in 1665, another 1670, another 1672, another 1675. A boy of ten does not enlist. Somewhere between 1665 and 1672 is as close as the record allows, and it will not go closer.
Schuldbrief ja. Maandbrief nee.
A schuldbrief was a debt letter: the Company advanced your passage and your kit against wages you had not yet earned. A maandbrief was the opposite, an instruction to pay part of your wages to someone at home, a mother, a wife.
He owed the Company money before he set foot on the ship, and there was nobody in Europe to send anything back to.
The Company entered a man's service on the day his ship sailed, so 13 May 1685 is not a paperwork date. It is the day he left.
The ship
The Land van Schouwen was built in 1676 at the Company's yard in Middelburg for the Chamber of Zeeland, 570 last, about 1,140 ton. Her master on this voyage was Kornelis Klaasz.
| Departed Texel | 13 May 1685 |
| Arrived at the Cape | 17 September 1685 |
| Departed the Cape | 13 October 1685 |
| Arrived Batavia | 24 December 1685 |
One hundred and twenty-seven days from the Netherlands to Table Bay. Whether he stayed ashore in September or went on to Batavia and came back, the muster does not say. It records only where a man left the Company's service, and that was the Cape, five years later.
There is a small, cold footnote. In 1690, the same year he stepped off her books for good, the Land van Schouwen lost her fleet on the fifth homeward voyage and went down off the Shetland Islands. The ship that carried him out did not survive the year he decided to stay.
15 August 1690
Reden uit dienst: vrijburger.
Five years and three months, and then he stopped being a servant of a company and became a man with his own name on his own ground. He was perhaps twenty-two, perhaps twenty-five.
The family's own account, carried in the published genealogy, says he came to South Africa in 1691 and settled at Drakenstein, taken up as a free immigrant on the same terms offered to the French Huguenots, who were arriving in those very years: sixty morgen rent-free, passage paid, on condition that everything he grew, wheat, vegetables, wine, was sold to the Company, and that he stayed at least five years.
That 1691 looks at first like a contradiction of the muster's 1690, and it is not. He was discharged in August 1690 and on the land in 1691. The family remembered the year they got the farm; the Company recorded the year he stopped working for them. Both are right, and neither is the year he arrived. That was 1685, and nobody remembered it at all.
It is the same trap as every register town in this history. The date a record was made is not the date the thing happened.
Where was he from?
This is the part that is not settled, and it is worth being honest about how unsettled it is.
| Says | Source |
|---|---|
| Woerden, Utrecht | the VOC opvarenden index, the record above |
| Werden, on the Ruhr, now part of Essen | the soldijboek, the pay ledger the index was compiled from; also J. Hoge, Personalia of the Germans at the Cape, and Linda Zöllner |
| Weerl → most likely Werl, Westphalia | his own will, 1722 (CJ 2601) |
| Weert, Limburg | various published registers |
| Wierden, Overijssel | the 1946 armorial entry |
Five spellings, two or three real places, one man.
Woerden and Werden are one vowel apart, and a Dutch clerk writing down what a German said his home town was is exactly how that happens. The direction of the error matters: an index is copied from a ledger, and the ledger is closer to the man. The will is closer still, because it is the only one of the five in which he is speaking about himself.
Zöllner's framing makes the German reading easy to believe. The Thirty Years' War left the German states wrecked, and for two generations afterwards young men walked west into the wealth of the Netherlands and took the Company's shilling because there was nothing behind them. Schuldbrief ja, maandbrief nee is what that looks like in a ledger.
On the weight of it he was probably a German in Dutch service. Neither the will nor the soldijboek has been read directly, though, and both are quoted here at second hand, so the question stays open.
There is a nice irony sitting underneath it. The 1946 armorial explains the surname as Oosthuizen, a village in North Holland. If the will is right, the name is Dutch and the man was not.
The two baptisms
Two records survive that could be his own baptism, and the compiled trees have picked one each without saying why.
At Amsterdam, father Johannes Oosthuijsz, mother Catrina Bouwmans, witnesses Baltjis Jacobus Bouwman and Maria Matthijs Oosthuijs. This is the one Frans's tree follows, dating it 1672.
At Leiden, 31 March 1675, father Johannes Oosthuysen, mother Antonette de Rede, witnesses Katharina van Nes and Lysbeth Oosthuysz. This is almost certainly where FamilySearch's 1675 comes from.
Two different families, both with a father called Johannes, because Johannes Oosthuysen was not a rare name in the seventeenth-century Netherlands. Choosing the one whose date you like is exactly how a compiled tree goes wrong.
There is a better test, and this family makes it available: look at what he named his children.
What he named his children
Dutch and Cape families named children in a fixed order, first son for the father's father, second for the mother's father, third for the father himself; first daughter for the mother's mother, second for the father's mother, third for the mother.
Of the six slots, three can be checked against people already known. All three hold:
| slot | rule says | he named | check |
|---|---|---|---|
| 2nd son | mother's father | Martinus | the mother is Johanna Martens, daughter of Marten ✓ |
| 3rd son | the father himself | Jan Hendrik | he is Jan ✓ |
| 3rd daughter | the mother | Jannetje | she is Jannetje ✓ |
Three for three. This family used the pattern strictly, which means the two slots that cannot be checked are not blanks. They are evidence.
| slot | rule says | he named | therefore |
|---|---|---|---|
| 1st son | father's father | Gerhardus | his father was Gerhardus |
| 2nd daughter | father's mother | Barbara | his mother was Barbara |
Neither baptism fits
Each has a father named Johannes. If either baptism were his, his eldest son would have been Jan. His eldest son was Gerhardus. And neither mother is Barbara: one is Catrina, the other Antonette.
Neither record is his. The Amsterdam family's own witness list agrees, from a direction nobody was looking in: a Maria Matthijs Oosthuijs standing at the font points to a grandfather called Matthijs, not Gerhardus.
The surname is a place
Oosthuizen is a village in Noord-Holland, near Edam. Names built that way are polyphyletic: they arise independently wherever people live by the place, so unrelated families carry the same surname and never share an ancestor. And the great majority of Dutch Oosthuizens stayed exactly where they were. Only one of them got on a ship.
The two baptisms above are the demonstration rather than the exception: two Oosthuysen households in the Netherlands in the same decade, neither of them ours. There were certainly more.
So the search cannot be run on the surname. Finding a Gerhardus married to a Barbara with a son Johannes in the right window would be suggestive and no more. It would still need corroboration from a place, an onward record tying that child to the muster of 13 May 1685, or Y-DNA, which is the one test that separates same surname from same family.
The same fact cuts against the Dutch reading. In Westphalia, farm names became family names in exactly this pattern, built on the West Low German -hues / -huis for house, and the surnames cluster in Münster, Coesfeld, Steinfurt, Recklinghausen, Borken and Unna, which is the country that holds Werl and Werden. A Low German Osthues or Osthaus, east house, written down by a Dutch clerk as Oosthuysen is precisely the mechanism that turned Werden into Woerden.
The 1946 armorial glosses the name as a North Holland village and puts Holland on the scroll. That is an assumption about a name two countries could have produced. The surname settles nothing about where he came from, in either direction.
Gerhardus and Barbara
Both are Rhineland and Westphalian names. Barbara is the patroness of miners and belongs overwhelmingly to the Catholic Ruhr rather than to Holland.
So the names he chose for his own children, read through a pattern he demonstrably followed, point at Werden or Werl, independently of the pay ledger and independently of the will. Two separate lines of evidence, arrived at from opposite directions, landing in the same country.
The same test throws off a free result on the other side: the stammoeder's parents were Marten and Anna.
The naming pattern generates hypotheses. It has never proved a parentage and does not prove one here. What it does is say which record to go and find.
What he did next
He married Johanna Martens, Jannetje as the register writes her, around 1700 to 1703, and they raised nine children in the Drakenstein valley, of whom seven lived to marry.
Two of them died in 1713, five and three years old, in the year the smallpox came ashore on a Company fleet and killed something like a quarter of the colony.
In 1722, in his fifties, he had a last son, Jacobus, baptised at Drakenstein on 22 November, with a single witness, Susanna Garde. That same year he made his will.
He died before 4 November 1730.
The fork
The lines this record can follow run down three of his sons, and they go three ways.
The eldest, Gerhardus, farmed at Drakenstein and then moved east, to the Uitenhage country, where a son and a grandson and a great-grandson farmed after him on Steenekoppies for sixty-odd years, through the frontier wars. From there the line went on to Tarkastad, Sterkstroom and Steynsburg, and then into the Great Trek with Piet Retief.
The youngest, Jacobus, born 1722, became a burgher at Swellendam and married Dina Carolina Brits at Tulbagh on 10 April 1749. His son Ockert died on the Olifants River in 1809. His grandson Ockert Almero ended up on Middelwater, in the Koup.
And between those two in age was Jan Hendrik, christened at Paarl in 1714, whose children took a third road: north-east onto the Sneeuberg in the 1770s, a generation before Jacobus's branch turned inland. His son Johannes and the widow Anna Botha stand in the frontier records there, and that road is told in The Oosthuizens of the Koup and The Koup before the grants.
One family, three roads out of the same valley: east to the frontier and on to Natal, north-east onto the Sneeuberg, or west to Swellendam and then inland to the Karoo.
16 February 1838
The eastern road produced this.
Days after Piet Retief and his party were killed, and the attacks at Bloukrans and Weenen, a small group was holding Van Rensburg's laager, nineteen men by the family's account, against a Zulu force estimated at fifteen hundred. The powder ran out.
Marthinus Jacobus Oosthuizen got on his horse, rode straight through them to a wagon, took powder and ball, and rode back through them again. The published account names Paul Naudé and Abraham de Beer alongside him. The laager held.
He was born 14 October 1818 and died 2 April 1897.
His genealogical code is b1c4d7e3, first son of the progenitor, then down. Frans's is through b8, the last son. Both are direct descendants of the man in the muster roll, and the two lines separate at his children and never touch again.
A de Beer riding beside an Oosthuizen in Natal in 1838, thirteen years before a de Beer and an Oosthuizen married in the Koup, is the kind of coincidence this family keeps producing.
The distance
| 13 May 1685 | Texel |
| 17 September 1685 | the Cape |
| 15 August 1690 | free burgher |
| 1691 | Drakenstein |
| 1749 | Tulbagh |
| 1783 | Swellendam |
| 1809 | de Modder Fonteyn, on the Olifants River |
| 1821 | the Koup |
| 1833 | Middelwater surveyed |
A hundred and thirty-six years, from a debt letter to a farm.
What this story is about
- about Johannes "Jan" Oosthuijsz (stamvader)
- about Johanna Martens
- about Marthinus Jacobus Oosthuizen (b1c4d7e3)
Sources
Everything this story rests on, with a link to the original wherever one exists. Where a citation is incomplete it says so, an incomplete citation is a research gap, not a formatting problem. If your work is here and the credit is wrong or thin, please tell us and it will be fixed.
- OOSTHUIJSZ Johan - van Woerden, adelborst, 1685 Nationaal Archief, Den Haag (NL-HaNA), VOC Opvarenden; transcription published by the eGGSA Documents CollectionNL-HaNA 1.04.02 (VOC 1602-1811), inventarisnummer 5331, folio 215view it online →
- VOC-schip Land van Schouwen (1676) - voyage of 1685 De VOCsite (vocsite.nl), compiled from VOC shipping recordsschip 10948; outward voyage departing Texel 13 May 1685view it online →
- Suid-Afrikaanse Geslagregisters (SAG), Vol 7 (O-Ph), pp. 136-182 - OOSTHUYZEN, with the family biography and the 1946 coat of arms GISA - Genealogiese Instituut van Suid-AfrikaSAG Vol 7, pp. 136-182
- Will of Jan Oosthuysen, 1722 Cape Archives (KAB), Court of JusticeCJ 2601
- Baptism of Johannes, son of Johannes Oosthuijsz and Catrina Bouwmans - Amsterdam Stadsarchief Amsterdam, toegang 5001 (Archief van de Burgerlijke Stand: doop-, trouw- en begraafboeken)DTB Dopen, archiefnummer 5001, inventarisnummer 44, blad p.349, aktenummer DTB 44view it online → Citation incomplete: Supplied as a screenshot. The ark and film/image numbers were never captured, so the original cannot be ordered from this entry alone.
- Baptism of Johannes, son of Johannes Oosthuysen and Antonette de Rede - Leiden, 31 March 1675 Erfgoed Leiden en Omstreken, Doop-, trouw-, begraaf- en geboorteboeken Leiden (DTB Leiden)DTB Leiden, Deel 239, 31 March 1675view it online →
- Toponymic surnames: the village Oosthuizen (Noord-Holland), and the Westphalian -hues / -huis farm-name pattern WikipediaOosthuizen (Noord-Holland); Niehuesview it online →
- Marthinus Jacobus Oosthuizen (1818-1897) Wikipediaview it online →
- J. Hoge, Personalia of the Germans at the Cape, 1652-1806 (Archives Year Book for South African History, 1946); and Linda Zoellner, Germans who came to the Cape 1652-1806 Union of South Africa, Government Printer / Archives Year BookHoge 1946view it online →
Research notes
Written 29 Jul 2026 from the VOC muster roll THE ORIGIN QUESTION IS OPEN - five candidate home towns across two countries, and his own will of 1722 (CJ 2601) is the document that would settle it. UPDATED 29 Jul 2026: Frans proposed testing the two candidate Dutch baptisms against the children's names. Both fail, and the pattern predicts parents called GERHARDUS and BARBARA - Rhineland names, agreeing independently with the will.
[31 Jul 2026] PUBLISHED. Prose had been finished and sitting at status='researched'; the audit's stage 4 found five such. Checked before publishing: every entity id in the prose resolves, none names a tombstone, and nothing it rests on was contradicted by the day's 55 merges - the ten entities that did change were additive (merge annotations, transcriptions, one new detail).
[6 Aug 2026] 'monsterrol' CHECKED and kept. It is not a Dutch loan smuggled in: the Western Cape Archives' VC series is literally called the Monsterrolle, and Karel Schoeman compiled 'Kaapse monsterrolle, 1656-1662 & 1664'. That is the established term in Afrikaans historiography, which is the test AFRIKAANS.md 1 sets - not whether a word looks Afrikaans, but whether it is established. Recorded in AFRIKAANS.md 4 so it is not re-litigated.
[12 Aug 2026] VOICE REWRITE to STORY-VOICE.md, both languages. en 0/19/1 -> 0/1/2, af 3/20/1 -> 0/1/2 (narration/bold-sentence/punchline). The dominant habit here was bold as mood, not narration: 39 bolded sentences across the two files, including whole arguments set in bold. Dutch terms that were bolded are now italic, which is what the standard reserves italics for - 'Adelborst', 'Schuldbrief ja. Maandbrief nee.', 'Reden uit dienst: vrijburger'. FIRST PERSON REMOVED: 'But I have not read the will or the soldijboek' was the researcher speaking on the page; it now says neither has been read directly and both are cited at second hand, which is the same evidential claim without the narrator. Headings that argued now tell: 'Which baptism is his? Neither of them.' -> 'The two baptisms'; 'Both candidates fail, and fail identically' -> 'Neither baptism fits'; 'A caution that outranks all of this: the surname is a place' -> 'The surname is a place'; 'And the two predicted names point where the will points' -> 'Gerhardus and Barbara'. 'Open [TO VERIFY]' renamed to 'What has not been worked out' so it and the Sources list lift into the research panel and the prose ends on its own last line. Research-priority claims ('the most valuable single request in the whole project', 'the sharpest search in the project', 'the first parcel in the model that is not in the Karoo') moved out of the narrative into that panel and restated about the evidence rather than about us. Bookkeeping line 'Recorded as a tombstone so nobody re-imports it' dropped; the rejection itself is kept in full. No facts, hedges or [TO VERIFY] markers changed. verify_af passing.
[13 Aug 2026] SECOND VOICE PASS, both languages, and THE LINK RULE APPLIED, which the first pass did not finish. en 0/1/2 -> 0/0/1, af 0/1/2 -> 0/0/1 (narration/bold-sentence/punchline).
NINE PEOPLE WERE LINKED IN story_links AND NOT ONE WAS REACHABLE FROM THE PROSE. The page carried no person link at all. Now linked where they are named: PER-000297 Johanna MartensM, PER-000306 Susanna GardeM, PER-000230 Jacobus b.1722, PER-000118 Ockert b.1755, PER-000001 Ockert Almero, PER-000321 Marthinus Jacobus, PER-000319 the rejected Johannes of Grijpskerke, plus the two farms the distance table names, PAR-000001 Middelwater and PAR-000065 de Modder Fonteyn.
SPELLING ALIGNED TO THE PERSON RECORD AND THE REGISTER. The page wrote the stammoeder Johanna MAARTENS; PER-000297's display_name is Johanna MARTENS and its notes record that the 1722 register writes her JANNETJE MARTENS. Neither the record nor the paper says Maartens, so it is changed here, and the naming argument reads 'daughter of Marten' with the free result 'Marten and Anna'. The argument is untouched: Martens is still the patronymic that makes the second son Martinus fit. 5a-i - the person carries one spelling, the transcription keeps the paper's, and *Jannetje as the register writes her* stays.
THE PUNCHLINE THAT WENT AND THE ONE THAT STAYS. 'Nine fields. Every line in this history comes down from them.' stood alone before a heading; the same words now close the paragraph that lists the fields, where they summarise rather than land. NOTE FOR ANYONE CHASING THIS COUNT: moving the sentence BELOW the blockquote does not clear it, because checks.py strips blockquotes and the sentence simply becomes the last block again. It has to join a real paragraph. 'A hundred and thirty-six years, from a debt letter to a farm' is KEPT. It closes the distance table and the whole arc, and STORY-VOICE.md says the fault is the pattern and not the earned line.
The soldijboek panel label carries its archive reference outside the bold.
FLAGGED, NOT FIXED: PER-000296, the stamvader, has TWO story_links rows for this story, one 'subject' and one 'mentioned'. verify_af passing.
The research behind this page, kept separate from it: dated working notes, open questions, and what is not yet settled.