# Early Glasgow families: manuscript and relationship reassessment Reviewed 10 September 2026. Public synthesis: `/research/transcriptions/early-families/`. ## Scope and durable records This pass revisits all 15 cases in the illustrated collection: 14 testaments and the separately identified George1605 baptism. Three bounded manuscript audits cover the early Edinburgh, Glasgow/Ayrshire, and Lanark/Leith groups. They inspect account names and continuations as well as the featured crops. - `name-audit-2026-09-10-early.json` - `name-audit-2026-09-10-glasgow.json` - `name-audit-2026-09-10-lanark-leith.json` - `originals-2026-09-10-reread/*.json` - `early-families-2026-09-10.json`: maintained public reconstruction and bibliography. Older versions remain available. Confirmed/reconfirmed findings are not all new discoveries. Repeated clauses within one testament and duplicate PDFs are not independent sources. The per-case audit distinguishes a name reading, an explicit relationship and a proposed WikiTree identity. ## Important manuscript results John1577's genuine continuation is image277. It repeats the children/executors and calls the deceased their father. Androw remains the probable second forename. The unrelated third screenshot concerns Andrew Hutcheson and is unresolved, despite its reused viewer header. Margaret's surname and the exact antecedent of the opening child clause require the qualifications recorded in the case audit; do not supply them from a pre-existing tree. Andrew1598's continuation image124 identifies Robert as his son begotten upon Elizat Stevinsoun, his wife. This is direct maternity evidence, not a generic widow inference. Marion, Jonet and Isobel are named; the unborn-child provision is conditional, not evidence of an additional live birth. Corstoun is the published index's place normalization, not a newly certain manuscript reading. Margaret1580's son remains Cuthbert Kaddell/Keddell. John Edstoun is her former husband, not thereby the named father of Cuthbert. Michael Mertyn and Roland Dundas are clearer account associates; neither is made kin. Geillis1603's executors are James, Isobell and Bessie Wilson. A provisional reading of words after James as son was withdrawn on tighter inspection. Preserve the uncertainty. James Glasgow in Stevenston is a creditor of the estate. He supplies a useful local contact, not an automatic father or brother. Jonet Meik1607 directly names Robert as her son. Her husband's forename is probably Niniane; the exact identity deserves a qualification separate from the secure maternity clause. Marioun's following word, a possible Meik sibling clause and another Glasgow legatee remain unresolved. New probable associate readings include Donald Mitchell and Cathrene Boyle; neither creates kinship. Ninian1613's brother John, sister Marioun, son Robert and spouse Jonet Hutchesoun remain secure. The possible William support clause is physically damaged. Missing ink cannot be recovered by repeating the expected phrase his son. Jonet Hutchesoun's maternity of Robert is not established. Saltcoats1637 names widow Isobell Gray and daughters Katherine and Agnes. The earlier Mary reading is rejected. John Broune is a sailor and cautioner. Do not attach this household to Glasgow-1030 merely through his occupation. William1646 names Archibald as his lawful son and sole executor; William died in September1644. This does not revive the old supposed brother/executor John. John1650 supplies a genuine sibling pair: Bessie Glasgow in Quothquan and her deceased brother John. The supplied nested PDF contains both register images 65 and 66, including the opening and inventory. Archibald1662's own family clauses name spouse Agnes Watson and lawful sons Archibald, George and William. Thomas Glasgow the Edinburgh weaver and William Glasgow the Leith weaver have administrative/financial roles, not explicit brotherhood. The full Lanark/Leith audit preserves the revised1673 executor names, Thomas's unresolved1662 descriptor and Jonet's qualified1685 surname. ## What can be reconstructed ### Glasgow burgh The strongest sixteenth-century core is William Glasgow-1098 and Esbell Brade-232 to their son John Glasgow-1093, explicitly in protocol1196 of 31December1534. John has sons Stephen Glasgow-1029 and William Glasgow-3522 through separate later protocols; younger William and Margaret Herbesoun are parents of George Glasgow-1099. Stephen's relationship to George is obtained through these father-son chains, not merely his appointment as curator. Robert the fisher Glasgow-1096, Archibald Glasgow-1097 and Isobella Glasgow-3916 are probable siblings of John / children of William based on the repeated property, trade, witness and Herbesoun network. This is a substantive kin hypothesis. It is neither an explicit proved edge nor mere name similarity. Elizabeth Glasgow-3915 has weaker support and should remain a possible relative. The 1598 marriage contract and the daughters named in1624/1633 burgess admissions extend George's household. They must not be confused with the Mid Calder George1605 baptism. The admissions name the daughters' father, not their mother. ### Ninian The burgess register identifies Ninian on23October1596 as the second son of Robert Glasgow, fisher and burgess. This establishes the documentary father. The maintained identification with Robert-1096 is strongly supported by trade and locality; it does not prove Robert's own father. Ninian's brothers/sisters need not all share both parents, since the testament does not distinguish full from half siblings. The Jonet Meik1607 and Ninian1613 Robert entries support a strong household reconstruction, with the 1607 husband's forename qualified. Keep Jonet Meik's direct maternity separate from the later spouse Jonet Hutchesoun. Ninian's brother John is not identified merely by linking the name to a convenient existing John profile. The1577 decedent cannot ordinarily be the living legatee of1613. The supposed Ninian-to-Leith-William link remains unproved. ### Leith Independent primary-edition texts were retrieved in this pass: - [Edinburgh apprentices, p.71](https://archive.org/details/scottishrecordso28scotuoft/page/71/mode/2up): Archibald, son of William Glasgow, mealmaker in Leith, to Thomas Park, cordiner,23October1639. - [General retour,1May1647](https://archive.org/details/inquisitionumadc00grea_0): Archibald, shoemaker in Leith, heir of William Glasgow, mealmaker in the same town, his father; originalxix.128, indexed General3278. Together with the1646 testament these secure William-3075 to Archibald-3078. The1618 baptism is compatible and is retained in the existing identity case. The1662 testament, however, calls its subject a maltman. Its immediate household is secure, but the occupation differs from the1647 shoemaker. This is not proof of two different men: a change of occupation is possible. It is a reason to require a bridge through Agnes Watson, a property or a guild record before treating the1662 family as an independently established extension of the earlier chain. Preserve the distinction in public prose. ### Ayrshire The indexed NRS description of [GD39/5/59](https://catalogue.nrscotland.gov.uk/nrsonlinecatalogue/details.aspx?reference=GD39%2F5%2F59), 9February1571, names James Glasgow elder and sons James and John among Corsinkell occupiers. This is a genuine father-and-sons group at catalogue abstract level. The original deed was not obtained in this pass. The Stevenston creditor James in Geillis's estate is a concrete lead for connecting that group to the later Wilson/Weir households. Neither a shared place nor a debt identifies him uniquely as a Corsinkell James. Katherine's minor Weir executors are very probably her children, although no secure child term has been recovered. Geillis's compressed executor wording has less secure kinship evidence. Do not flatten those two assessments into one certainty level. Saltcoats John, widow Isobell Gray and Katherine/Agnes remain a distinct documentary household. A daughter's marriage or succession could connect it to the earlier Ayrshire group; no reviewed clause does so now. ### Lothian and Lanark Andrew's widow/son holding and succession at Corstoun provide an unusually specific traceable route. The1641 Edinburgh apprentice entry names James, son of Robert Glasgow, servitor to William Crichtoun in Wester Duddingston, to William Sclaitter, weaver. That record identifies a later household but does not identify its Robert as Andrew's son. Keep separate candidates. In Lanark, preserve the separate1650,1662 and1685 John histories. The first gives Bessie as a sister; the last gives widow Jonet but no newly established children. Executors, procurators fiscal and cautioners are not default sons. The individual audits are authoritative for the current personal-name readings. ## What original families means at this stage Do not merge the1447/1455 Glasgow burgess,1475/1479-80 Stirling witness, 1506 John alias Smith and later Saltmarket title holder into a universal John1475 ancestor. A historical title repeated in a later instrument does not necessarily prove its earlier owner was alive on the later date. These are separate identity and transmission questions. No testament here provides the missing named emigrant to early Irish branches. Genetic clustering cannot select one sixteenth-century man or invent the intervening generations. Retain securely documented roots, clearly express strong probable kin groups, and target the exact joins that lack evidence. ## Access and publication limits The15-case manuscript reread is new; much of the wider property reconstruction is established local research reassessed alongside it. The public JSON lists source-by-source access levels. Published-register OCR was used to recover entries, not presented as a new reading of the handwritten original. The NRS Corsinkell catalogue was available through its indexed description, but direct retrieval failed. WikiTree pages/API were unavailable for a fresh attachment check. No live profiles or relationship settings were changed. Publish with `.venv/bin/python tools/build_transcriptions.py`. The publisher reads the versioned reviews and three name audits, adds each case's audit, and generates the family synthesis and downloadable evidence. Browser and automated code tests are not part of this research pass.