Ohio, Wyandot County Probate Court, "Administrators' bonds and letters, v. 1-2, 1860-1895", volume 2 page 109 — administrator's bond and recorded Letters of Administration of C. J. Spencer on the estate of G. G. Cooper, granted 11 November 1886; with the volume 2 decedent index letter C at page [unnumbered]. FamilySearch DGS 5871546 images 413 and 335.

What this document records

The index row verbatim

Volume 2 decedent index, letter C (film image 335), tenth row of twenty-seven: "Cooper G. G. — C. J. Spencer Ad. 109" The index is by decedent, with the administrator's name and the volume page. It is a running index in rough order of page number (12, 19, 37, 130/36, 41, 62, 76, 83, 87, 109, 110, 125, 148, 180, 199, 216, 231, 273, 278, 327, 347, 348, 355, 356, 364, 411, 424), though two rows sit out of sequence.

The caption and granting clause verbatim

Page 109, read in 4× band crops: Caption: "C. J. Spencer / ADMINISTRATOR OF / G. G. Cooper" · then the printed heading BOND. "Know all Men by these Presents, That we, C. J. Spencer, Jordon Gault and Samuel Spencer are held and firmly bound unto the State of Ohio, in the penal sum of one thousand Dollars, to the payment of which we do hereby jointly and severally bind ourselves, or heirs, executors and administrators, if default be made, in the conditions following: Whereas*, Letters of Administration upon the Estate of G. G. Cooper deceased, were granted to the said C. J. Spencer by the Probate Court of Wyandot County, in the State of Ohio, on the 11th day of November, A. D. 1886."* Signed at the foot by C. J. Spencer, Jordon Gault and Samuel Spencer, each against a printed SEAL — three sureties' signatures, no marks. (Signature block read at 0.5× only; the names are those already verified above.)

The letters verbatim

The printed LETTERS. section beneath the bond, being the court's own recorded grant: "The State of Ohio, Wyandot County, ss. To all who shall see these Presents, Greeting: Be it Known, That by the Court of Probate of said County, Administration of all and singular the goods, chattels, rights, credits and estate, which were of G. G. Cooper late of said County, deceased, has been granted unto C. J. Spencer, whose duty it shall be to have all and singular, the said goods, chattels, rights and credits appraised by Samuel Spencer, Peter Babcock & Joe [Cleninger?] and to return, upon oath, within three months, a true inventory thereof…" "Witness, D. D. Clayton, Judge of said Court, at Upper Sandusky Ohio, this 11th day of November, A. D. 1886." — subscribed D. D. Clayton, Probate Judge. Followed by the judge's own certification "that the above is a true copy of the original Letters of Administration, granted in the premises by said Court, and remaining on file in my office", dated at Upper Sandusky the 11th day of November A. D. 1886.

The appointment of three appraisers is the forward pointer

THE MOST ACTIONABLE LINE ON THE PAGE. The Letters direct the administrator to have the estate "appraised by Samuel Spencer, Peter Babcock & Joe [Cleninger?]" and to "return, upon oath, within three months, a true inventory thereof". ⇒ An inventory and appraisement was therefore due by about 11 February 1887, and a separate Inventories/Appraisements series should hold it. An inventory lists the decedent's goods and, where real estate is included, describes it — and the account that follows an administration names the persons paid a distributive share. This bond does NOT name a single heir. That is not a disappointment to be explained away — it is what a bond is: a promise by the administrator and his sureties to the State. The heirs are named in the petition to sell real estate ([K.102] proves one was granted, the sale being recorded 29 Dec 1888) and in the final account.

What this record does not give

No heirs, no widow, no children, no relationship of any kind, and no date of death — a bond recites the grant of letters, not the death. No Cooper is party to it. The administrator, both sureties and all three appraisers are non-Coopers. That is itself a datum worth holding onto: where a widow or eldest son is available and willing, the court commonly appoints them, and here it did not. Susan (his wife per [name.495]) had died in February 1884 ([G.540]), so there was no widow to appoint. Why no child was appointed — his 1880 household held Lydia A., Alvin W. and Angeline W. — is unexplained by this record. ( And "unexplained" is all it is: minority, distance, illiteracy, disinclination and simple local practice are all ordinary reasons, so nothing should be read into it.)

The null this corrects

A PREVIOUSLY RECORDED NULL WAS TRUE BUT WRONGLY SCOPED, AND THIS IS THE LESSON WORTH KEEPING. On 2026-09-14 the C index of this same reel was read and recorded as "no Cooper" — at films/film.5871546.yaml and on Samuel Cooper. That reading was correct. But it was the index of volume 1 (film image 8), and this reel carries volumes 1 AND 2 — each with its own index. Volume 2's index was noted at the time as "not located (not needed for 1871)", which was true for Samuel and became false the moment [K.102] produced an 1888 Cooper administrator's sale. The generalisable rule: a catalog title reading "v. 1-2" means TWO indexes, and a null recorded against "the index" of such a reel is a null against ONE of them. ⇒ when a bounded null is written down, the boundary must be part of the finding — which index, covering which years — or the null will later be read as broader than it was. The volume-1 null still stands for Samuel Cooper Samuel Cooper d. 1871, and volume 1 runs to at least April 1881 (image 290 = page 520), so it covers 1871.

The digit that could not be read

A RULE ZERO-B CASE THAT RUNS THE OPPOSITE WAY FROM THE USUAL ONE, AND IS WORTH MORE THAN THE RECORD IT FOUND. The index page reference was read four times, at four magnifications, and the two highest were both WRONG: at ~0.85× (orientation) it read 109 · at 5× 189 · at 8× 189 · at 20× 199. Escalating magnification produced increasing confidence and a changing answer, because the crop was already past the film's own information limit — beyond that point zoom.py is interpolating pixels, not resolving ink, exactly as the FamilySearch viewer does when zoomed past its tile resolution. Worse, the cross-check pointed the wrong way too. Reference digits were taken from the same page, same hand, same magnification — 148, 180, 199 — and compared stroke-for-stroke; that comparison supported 189. It was a sound method applied to unrecoverable data. What actually settled it was leaving the index alone and going to the register. Page 189 proved to be Benjamin Ranck, administrator of Sarah Ellen Bogard (January 1889); page 109 is the Cooper bond. The rule: when a value is at the resolution limit, do not magnify further — find an INDEPENDENT record that can only be consistent with one reading. An index entry is checkable against the thing it indexes, and that check is worth more than any number of pixels. The ordinal sequence of the index's own page numbers (…87, 109, 110, 125…) had also argued for 109 all along — but that is an inference from context, which RULE ZERO forbids writing into a field. It was properly used as a reason to look harder, not as the reading.

How the page was located

Worth recording as method — the whole hunt was four lookups. 1. The volume boundary, from the thumbnail grid. Image 328 is an END placard and image 331 a BEGIN placard ⇒ volume 1 ends ~326, volume 2 starts at 331. 2. Volume 2's index, one letter per image. Image 333 is A ("Armstrong J R … 274"), 334 B, 335 C. The index runs to about image 350, and images 351-358 are its blank tail. 3. The body start, also from the grid. Image 359 is the first densely written page; 358 and earlier are pale. 4. The page↔image rule, then one probe. Assuming image 359 = page 1 gives image = 359 + (odd page − 1) ÷ 2. Page 189 was predicted at image 453 and proved exact — the opening carries 188 | 189, page numbers at the OUTER top corners. Page 109 was then predicted at image 413 and also proved exact. Two pages per opening, even page on the left. The rule is recorded on the reel file. FamilySearch's /search/film/<DGS> URL redirects to Sign In when the session has lapsed — and so does a known-good ark, which is how the two cases are told apart. See guidance/feedback_drive_the_right_browser.md.

The date conflict this creates

THIS RECORD CONTRADICTS THE DEATH DATE THE CATALOG HAS BEEN CARRYING FOR GEORGE GREEN COOPER, AND THE RECORD IS THE BETTER SOURCE. Letters of administration were granted 11 November 1886 — read at 4× for the day and month and at 11× for the year, in three separate instances on the page. ⇒ he was dead by 11 November 1886. [G.540] and Find a Grave 101434322 both give "12 Dec 1886" — a month after the grant, which is impossible. The year is corroborated by volume sequence, which is an independent control: the facing page 108 is also dated November 188[6], and page 189 is January 1889. So 1886 is not a misread of 1887 or 1888. And [G.540]'s own age statement already disagreed with its own date. The transcription gives b. 13 July 1819 · d. 12 Dec 1886 · 67y 4m 2d. From 13 July 1819, 67y 4m 2d falls on 15 November 1886, not 12 December — so the transcription was internally inconsistent before this record was found, by about four weeks, in the direction this record points. ⇒ The reading that costs least: he died in the first days of November 1886, and "Dec" is a misreading of the stone (or of the transcriber's source). But that is an inference and is NOT written into any header** — the fact recorded here is only the primary one, that he was dead by 11 November 1886. A stone photograph or an Ohio death record would settle it. → [name.495]