Thomas William Cavins

PROOF CLASS: BOTH PARENTS = ARGUMENT (Owen graded 4 Sep 2026). NO record names Thomas's parents - the corpus that would carry one has been read to exhaustion and does not contain it. This link closes on ten correlated records. Summary: proof-summary-279-cavins-2026-09-04.md THE ARGUMENT. James E. Cavins and Mary Frances Mansfield were married in Randolph Co., Missouri on 24 Nov 1842 by Thomas Fristoe, "a preacher of the Gospel", both of Randolph County, recorded 23 Mar 1843 - read on the manuscript image, register K202, marriage register p. 74. [279] Thomas William Cavins was born 16 Mar 1844, fifteen and a half months later. The 1850 census of Salt Springs, Randolph Co. (dwelling 348) enumerates "J E Cavens 31, Farmer, b. Kentucky" with Mary 22 and four sons - Thomas 6, John 4, Robert 3, Benjn 1 - matching four of the seven children of F363 by age, four for four. The 1860 census of Silver Creek, same county (dwelling 377), enumerates Mary F Cavins 42 b. Virginia as WIDOWED HEAD with six of the seven, Thomas W among them at 16. James's Randolph Co. estate opens with letters of administration 19 Jan 1860, intestate (K203), and a fourth instrument from an unrelated hand - The Randolph Citizen of 21 Mar 1861, a year-in-review roster of Randolph Co. men who died in 1860 - lists "James E. Cavens" first (K426). The strongest single correlating fact is not co-residence at all: in the 1870 census [908] Benjamin Eddins Cavins, a documented member of this sibship, is enumerated as "Benj Cornelius" in his mother's household after her remarriage, a child carrying his stepfather's surname. THE HONEST WEAKNESS, and it is why the class is ARGUMENT rather than DIRECT. Every load-bearing record above is CO-RESIDENCE or identity-continuity, never a filiation statement. Neither the 1850 nor the 1860 nor the 1870 census has a relationship column - that first appears in 1880. The Randolph Co. probate corpus that would carry an heir roster has been read to completion under both spellings (K203, K205) and the county's own general probate index checked (K343-K345); no distribution or final settlement naming these children exists in it, and the working explanation is that the estate was consumed by debts before any child-naming distribution was made - a 1863 order sells the land "for the payment of debts". A CLAIM THAT WAS CHECKED AND REMOVED, 4 Sep 2026: the naming pattern "Thomas WILLIAM, after his grandfather [890] William Cavins" was offered as corroboration in proof-standard.md section 4. It is UNSUPPORTED - this project's register carries 14 namesake-style arguments and NOT ONE is for Cavins, Mansfield or William (positive control: Cavins returns 50 register hits). It is struck, and it is NOT a pillar of this grade. WHAT WOULD OVERTURN IT: a Randolph Co. guardianship or curatorship bond for a minor Cavins child between 1860 and 1865 - all six surviving children were minors when James died - or the Missouri State Archives loose case file for Mary F. Cornelius's 1892 probate (reel C35013, cases 272-399), naming a different set of children as heirs, or omitting Thomas from an otherwise complete roster. That MSA order was e-mailed 21 Aug 2026 and, as of 4 Sep 2026, nothing has come back. STILL OPEN IN THIS SIBSHIP, neither bearing on [279]: [911] Pleasant Madison is the only F363 child with zero citations and is absent from both the 1850 and the 1860 census; and [910] Sarah Elizabeth is enumerated aged 7 in 1860, older than [909] Margaret Jane, reversing the tree's birth order - recorded in [889]'s own note and NOT changed here.