James Thomas "Tommie" Coe

AUDITED 22 Aug 2026 - TWO SEPARATE GRADES ON THIS MAN, AND THEY MUST NOT BE READ TOGETHER. His BIRTH DATE is now SOURCED for the first time (his own WWI draft card). His DEATH DATE, 10 Jan 1920, is ASSERTED AND NOT PROVEN - it carries ZERO citations and no record has been found for it. 1. BIRTH - 22 May 1879, now cited. E559 has carried this exact date all along with NO source. His WWI draft registration card (coll. 6482 rec. 15454122, read on the image) states 'May 22 1879' in what appears to be his own hand, signed 'J. T. Coe', nearest relative 'Mrs Willie Coe' - his wife [128] - at the same Winnsboro address. Transcript and the glyph control are in C986. NOTE THE GRADE: this is CORROBORATION of a date the tree already asserted, not a discovery. CORRECTION ON THE RECORD: the agent that opened this card reported the date as 'the identifier every failed search has lacked'. That is FALSE - E559 already held 22 May 1879. The date was never missing. What was missing was a source for it. AND THE CARD WAS ALREADY CITED ON HIM: C986 was linked to E559 with an empty transcript. Located, never opened - the same pattern as registers K260 and K270. 2. DEATH - 10 Jan 1920, Winnsboro: ASSERTED, NOT PROVEN. E560 carries no citation of any kind; the date is an unsourced inheritance from a copied tree, stated to the day. A controlled search of the Texas death certificates 1903-1982 (coll. 2272) under two independent name forms, 49 and 195 hits with both controls firing, returned no 1920 certificate for him in Franklin or Wood County (registers K259, K267). CAVEAT FIRST, BECAUSE IT CUTS AGAINST THE DOWNGRADE: rural Texas death registration in 1920 was incomplete, so a missing certificate can be a fact about registration compliance rather than about the man. The date has therefore NOT been altered or deleted - an unsourced date is a bookkeeping gap, not a disproof. AN UNRESOLVED TENSION: he is enumerated in the 1920 census at Winnsboro (roll T625_1804, page 21B, ED 59). That is not automatically a contradiction, because the 1920 census reference date is 1 January 1920 and a man who died on the 10th is correctly enumerated - but enumeration ran through January and February and nobody has opened the page. NAMED NEXT, AND IT IS ALREADY ATTACHED AND UNREAD: the 1920 census citations on [127] and on [128] both have empty transcripts. Reading that household image settles it directly - is [128] listed as WIFE or as WIDOW, and is he in the household at all. Do that before hunting another certificate. 3. HIS PARENTAGE REMAINS OPEN and is untouched here. The Find a Grave reading that his father was James Coe 1848-1894 x Mary Hannah Wren, making the tree's John P. Coe his uncle, is UNCORROBORATED by any vital record. Do not act on it. AUDITED 22 Aug 2026 (later) - THE 1920 CENSUS HOUSEHOLD, READ ON THE IMAGE. He is enumerated as HEAD, MARRIED, age 40; [128] is WIFE, not widow. ⛔ THIS DOES NOT PROVE HE LIVED TO 10 JAN 1920, AND THE COINCIDENCE OF DATES WILL TEMPT YOU. The sheet's header reads 'ENUMERATED BY ME ON THE 10 DAY OF January, 1920' - his recorded death date - but the form's printed rubric enumerates every person living on JANUARY 1, so a man who died on the 3rd is written down exactly the same way. WHAT IS ESTABLISHED: he was alive on 1 Jan 1920, and a death before that date is ruled out. E560 keeps its grade - ASSERTED, NOT PROVEN - because no record of the death itself has been found. ★ A SECOND INSTRUMENT ON THE BIRTH DATE, by the same control as the draft card. AGE 40 against a 1 Jan 1920 reference admits birth between 2 Jan 1879 and 1 Jan 1880 - 22 May 1879 fits, 22 May 1880 would give 39. Ancestry's index says 'abt 1880' because it subtracts naively (1920-40); the index is wrong and the age column is right. Independent of the WWI card and agreeing with it. ◎ HIS PARENTS' NATIVITY, AND WHY IT STILL DOES NOT CLOSE F83. The sheet gives his father born ALABAMA and his mother born ALABAMA. The tree's [129] John P. Coe was born Benton, Lowndes, Alabama and [130] Prudence Mae Weatherford at Covington, Alabama - both fit. But a census nativity column NAMES NO PARENT; it is household-grade and sets no proof flag here. And it cannot decide the open question: the Find a Grave alternative makes the father James Coe 1848-1894, brother of the tree's John P. - BROTHERS SHARE A BIRTH STATE, so this evidence is consistent with both candidates. F83 stays 0/0. TWO DAUGHTERS ON THIS SHEET WERE MISSING FROM THE TREE and are added by this write: Mabel (age 4) and Evelyn (age 3). Neither existed anywhere in the database under any spelling before today. --- 31 Aug 2026 (K330) - HIS FIVE CHILDREN ARE NOW ON A SWORN COURT RECORD, AND A FIFTH CHILD IS ADDED. Franklin County, Texas probate case 731, October Term 1923. His widow Willie Coe petitioned for administration and swore that 'the said J.T. Coe left surviving him the following children, to-wit: Melva Coe, a girl 12 years old, Mozelle Coe, a girl 10 years old, and Maybel Coe, a girl 8 years old, Eveline Coe, a girl 7 years old, and Tommy Coe, a boy 3 years old, all of whom reside with your applicant'. WRITTEN: ProofFather=1 on all four existing F82 children, and James Thomas 'Tommy' Coe added as the fifth at the same grade. The four names reconcile without strain (Melva/Melba, Maybel/Mabel, Eveline/Evelyn) and the ages back-calculate to about 1911, 1913, 1915, 1916 and 1920. NOT WRITTEN, DELIBERATELY: ProofMother stays 0 on every row. The petition makes the children HIS and says only that they 'reside with your applicant' - residence is not maternity. And no interval argument is available: F82's marriage E8167 is 'abt 1911' with ZERO citations, an estimate in this database rather than a date any record states. CHEAPEST ROUTE TO ProofMother: a Texas death certificate for any daughter names both parents - the instrument that settled [381] Viola. None of the four carries one today. ALSO FROM THE SAME FILE: the community estate inventoried 18 Oct 1923 - 1 acre of the D.M. Connally survey in Winnsboro and 105 acres of the Warren Davis survey, Franklin County, total $3,175 against $1,300 of debt - now recorded as a Property fact. HIS OWN PARENTAGE IS UNTOUCHED AND REMAINS CONTESTED. F83 stays 0/0. This document names his widow and children, not his ancestors. The open question is whether he is the son of [129] John P. Coe x [130] Prudence Mae Weatherford as the tree has it, or of John P.'s brother James Coe x Mary Hannah Wren as Find a Grave has it; four discriminators have been tested and died, and the named next document is Franklin Co. probate case 254, Vol D pp. 178-337. See the drained queue entry of 23 Aug 2026 and registers K282-K293, K329.