BEYOND RUTH X

Convergent Evidence and the Genealogical Proof Standard of a Mayflower Line

ABSTRACT

This article presents a structured adversarial evaluation of a fourteen-generation Mayflower line of descent from Edward¹ Fuller (d. Plymouth, 1620/21) to Margaret Deborah¹⁴ Page (b. 1948), conducted through dialogue between the author—who constructed the original proof arguments—and Claude (version Sonnet 4.6), a large language model developed by Anthropic, deployed as an analytical interlocutor. The article addresses GPS compliance across each generational junction, constructs and rebuts the strongest available counter-arguments, applies probabilistic reasoning to assess residual uncertainty, and confronts directly the confirmation bias inherent in an author’s self-evaluation of his own proof. It concludes that the line is GPS-compliant throughout and that the one junction resting on convergent indirect evidence satisfies the standard as rigorously as the surviving record permits.

---

I. Introduction: The Problem of Self-Evaluation

Genealogical proof arguments are rarely evaluated by anyone other than their authors and their editors. Peer review at serious journals—the New York Genealogical and Biographical Record, Mayflower Descendant, The American Genealogist—provides one corrective, but a reviewer evaluates the argument as presented. The choices its author made about what to include, what weight to assign competing evidence, and how to characterize the nature of the proof are not fully visible in the finished text.

This article arises from a methodological experiment in managing that problem. The author, Jeffery Allen Record, had constructed two GPS-standard proof arguments bearing on a fourteen-generation Mayflower line: one identifying Ruth⁶ Fuller as the wife of Henry Francisco of Whitehall, New York, published in The New York Genealogical and Biographical Record (vol. 157, July 2026, pp. 231–236)¹; and one identifying Anna Belle¹⁰ Cisco as the daughter of Shelton Clark⁹ Cisco of Madison, Indiana.² The line as a whole was documented in a privately printed monograph.³ Having constructed these arguments, the author subjected them to structured adversarial analysis using Claude (version Sonnet 4.6), a large language model developed by Anthropic, engaged not as a research assistant but as an analytical opponent: reading the finished proof documents, constructing the strongest available counter-arguments, identifying vulnerabilities, and pressing the author to defend or revise positions under questioning.

The primary methodological risk is confirmation bias. The author built the proof; the AI evaluated it with the conclusion already in hand. This article does not pretend that risk was eliminated. It argues that the risk was managed through three structural features of the dialogue: counter-arguments were pressed to their hardest available form before being rebutted; genuine vulnerabilities were named and left standing; and the author treated the AI’s revisions in his favor as signals requiring scrutiny rather than as validation. Whether those measures were adequate is a question the reader must evaluate. The claim of this article is not only that this Mayflower line is valid, but that the process by which genealogical proof arguments are evaluated—and the conditions under which they can be trusted—deserves more explicit attention than the field typically gives it.

II. The Genealogical Proof Standard

The Genealogical Proof Standard, as defined by the Board for Certification of Genealogists, requires five elements: a reasonably exhaustive search of relevant sources; complete and accurate citation of those sources; analysis and correlation of the collected evidence; resolution of any conflicting evidence; and a soundly reasoned written conclusion.⁴ The standard is evidence-type-neutral. It does not require direct evidence or a single document naming the connection. A proof argument resting on convergent indirect evidence—multiple independent lines, none sufficient alone but all pointing to the same answer—satisfies the GPS as fully as one resting on a single direct record, provided the conclusion is supported to the exclusion of all reasonable competing alternatives.

III. The Line: A Generational Survey

Generations One Through Five: The Silver Book Foundation

The first five generations—from Edward¹ Fuller through Aaron⁵ Fuller—are documented in Bruce Campbell MacGunnigle’s Mayflower Families Through Five Generations, Volume Four: Family of Edward Fuller, 3rd edition (Plymouth: General Society of Mayflower Descendants, 2006), the standard GSMD authority for this line,⁵ and require no independent evaluation here. One observation bears on what follows: the Silver Book records Aaron⁵ Fuller’s death as “probably at Cornwall, Conn., ca. 1762.” Morton’s reconstructed tenant list for Skenesborough directly contradicts this, placing both Aaron⁵ Sr. and Aaron⁶ Jr. there between 1759 and 1775.⁶ The Silver Book should be revised accordingly. The revision strengthens the line by placing Aaron⁵ Sr. alive at Skenesborough during the period critical to the Generation 6 identification.

Generation Six: The Critical Junction

Generation 6 is the most difficult junction in the line: the identification of Ruth⁶ Fuller—born 15 November 1747 at Colchester, Connecticut, daughter of Aaron⁵ and Ruth (Sawyer) Fuller, recorded in the Silver Book but left untraced after birth—as the wife of Henry Francisco of Whitehall, Washington County, New York. No single document names the connection directly. The identification rests on convergent indirect evidence. Before presenting that evidence, the two published competing identifications must be addressed.

Frank J. Doherty’s The Settlers of the Beekman Patent attributes Henry’s wife to a daughter of Daniel Fuller of Dutchess County⁷ but cites no primary source. No land, tax, probate, census, or church record connects any child of Daniel Fuller to Skenesborough or Whitehall, and the claim’s ultimate source appears to be Carol Clark Johnson’s Fullers, Sissons, and Scotts, itself without primary citation.⁸ Proponents on WikiTree acknowledge that “this assertion needs proof.” It is dismissed. William Hyslop Fuller’s Genealogy of Some Descendants of Edward Fuller of the Mayflower assigns Aaron’s Colchester daughter a marriage to Job Gould on “Feb. 29, 1767”—an impossible date, as 1767 was not a leap year.⁹ The actual Job Gould Jr.–Ruth Fuller marriage occurred on 24 February 1773 in Sharon, Connecticut, with the bride identified as “of Kent”—not Colchester¹⁰—a demonstrable conflation of two distinct women that Fuller himself corrects elsewhere in the same volume. Both competing identifications are dismissed.

The affirmative case rests on nine independent lines of evidence. Henry Francisco’s schedule of property dated 30 August 1820 names his wife as Ruth, “about seventy four years of age,”¹¹ implying a birth of about 1746—within one year of the Silver Book’s recorded birth date of 15 November 1747. The 1904 DAR membership application of Ada Sophia (Palmer) Westcott, tracing her lineage through Henry and Ruth’s son Jacob⁷ Francisco, explicitly names Ruth’s maiden name as Fuller¹²—a secondary source, but an independent family-source statement formally submitted under DAR institutional scrutiny and unchallenged in over a century. The Solomon Francisco family Bible records the 1787 marriage of Solomon⁷ Francisco as officiated by “Mr. Aaron Fuller, Esq.”¹³—Aaron⁶ Fuller Jr., documented brother of Ruth⁶ Fuller and Justice of the Peace at Whitehall, one of five justices then sitting in Washington County.¹⁴ His selection to officiate a Francisco family event is not explained by necessity; it is explained by kinship.

The Philip Skene Papers ledger of promissory notes at Skenesborough, 1767–1773, names Henry Francisco alongside Aaron Fuller, Judah Fuller, Gershom Fuller, and Ebenezer Vine in a single primary financial document.¹⁵ Morton’s reconstruction of Skene’s tenant roster, 1759–1775, places Aaron⁵ Sr., Aaron⁶ Jr., and four other Fuller brothers, the Sawyer maternal relatives of Ruth⁶ Fuller, and Henry Francisco at the settlement—Francisco the sole individual in an otherwise entirely kin-defined cluster, with no other documented explanation for his specific placement within it.¹⁶ Freelove⁶ Fuller, documented sister of Ruth⁶, is independently confirmed at Whitehall on four evidentiary lines: her calculated birth date matches within one day across the Silver Book, Starr’s History of Cornwall, Connecticut, and a DAR gravestone calculation; her gravestone records death on 28 January 1829 aged 68 years, 8 months, and 28 days; a Washington County deed of 1784 places Solomon Vine’s property immediately adjacent to Aaron⁶ Fuller Jr.’s land; and the 1790 census enumerates Solomon Vine proximate to Aaron⁶ Jr.¹⁷ Female members of the Colchester Fuller family migrated to Whitehall—not merely the brothers.

Ephraim⁶ Fuller named daughters Ruth⁷ and Freelove⁷,¹⁸ the pairing pointing to the sisters rather than to the mother Ruth Sawyer alone. The interlocking pension affidavits of Henry Francisco (no. S44864), Solomon T. Vine (no. S43213), and Thomas Lyon (no. S9687), sworn 15 April 1818 before Judge Jonathan Wood, document a military-social network independent of the land, Bible, and tenant records: Lyon attested to both Vine’s and Francisco’s service; all three served in Captain Burroughs’ Company under Colonel Seth Warner.¹⁹ Finally, a comprehensive search of Washington County census schedules, town meeting minutes, assessment rolls, church and probate records identified no other woman named Ruth of suitable age as a competing candidate for Henry Francisco’s wife—a documented exhaustive search that produced no alternative.

IV. The Counter-Arguments: A Steel-Man Analysis

The Westcott Problem

The Westcott DAR application is the only document that actually names “Ruth Fuller” as Henry Francisco’s wife. Everything else in the proof argument establishes deep and sustained association between Henry Francisco and the Fuller family. It does not, strictly speaking, establish the maiden name. Westcott is a great-granddaughter making a claim four generations removed, with no identified primary source cited. Family tradition, however consistently transmitted, is not primary evidence, and the Westcott naming of “Fuller” is doing significant evidentiary work in the argument.

This critique has genuine force. The response is threefold: the Westcott application was accepted under DAR institutional scrutiny and has not been challenged or contradicted in over a century; the alternative the overlay argument requires—a woman of some other surname whose Fuller connection is entirely undocumented—is not a named candidate but a logical possibility, and the GPS requires the elimination of documented competing candidates, of which there are none; and Ruth⁶ Fuller is a documented individual of exactly the right name, right age, right family, and right geographic trajectory, present in the same community at the relevant period, with no subsequent history in any other record.

The “Ruth X” Hypothetical

A harder version of the overlay argument dispenses with the Fuller connection entirely and posits any Ruth of unknown or unrelated surname—call her “Ruth X”—who left no record and whose family has no connection to the Fuller cluster. The surname is immaterial; the structure of the argument is the same regardless of what name is supplied. For any such candidate to be viable, every one of the following would have to be true simultaneously: the exhaustive search failed because she left no trace; her family’s presence at Skenesborough is entirely undocumented; Francisco’s placement within the Fuller-Sawyer sub-cluster is coincidental; Aaron⁶ Fuller Jr.’s selection as wedding officiant is civic coincidence; Freelove⁶ Fuller’s confirmed presence at Whitehall is unrelated to the Francisco family; the Westcott tradition naming Fuller is fabricated; and Ephraim⁶ Fuller’s naming of daughters Ruth and Freelove is purely coincidental. Each proposition is individually implausible. Together they are not a competing hypothesis; they are the assertion that all the evidence is simultaneously wrong. The GPS does not require the refutation of unfalsifiable hypotheticals.

The Confirmation Bias Problem

The most serious methodological challenge is not about the evidence but about the analyst. Claude read the finished proof before conducting its analysis and revised its assessments twice in the author’s favor—a confirmation bias signal named here as such. The structural features of the adversarial dialogue managed but did not eliminate that risk. What can be said is that both revisions were correct on the merits: the Greene County deed’s sworn dower examination is primary legal evidence of the marriage, not a gap; and the Generation 6 argument, having passed peer review at the most rigorous genealogical journal in the country, is not “genuinely uncertain.” Those corrections favored the author’s position because the initial assessments were wrong in a particular direction—not because the interlocutor was capitulating to preference.

V. The Remaining Generations

Solomon⁷ Francisco’s birth (2 October 1768), death (Cincinnati, 20 January 1844), and marriage (28 December 1787, Mary Freeman) are established by the Solomon Francisco family Bible.²⁰ His identity as Henry’s son rests on the Bible’s deaths page, in Solomon’s own hand: “October the 10, 1820, died henary francisco 134 yrs. father of Solomon francisco.”

John⁸ Sisco, son of Solomon⁷, was born 5 February 1792 and died 5 December 1834, both per the Solomon Francisco family Bible, corroborated by a Hamilton County, Ohio probate entry of 4 October 1837 granting administration of his estate to “Anna B. Sisco.”²¹ His identity as Solomon’s son is established by his brother Elon⁸ Francisco’s 1886 letter: “your Grand Father was my Brother.”²² No primary marriage record for John and Annabelle Kyle has been located, but the marriage is established in primary legal instruments: a Greene County, Ohio deed of 22 June 1826 names “John Sisco and Anna Bella his wife,” with Annabelle separately examined as to her dower interest before Justice James Foster, and the Hamilton County probate grants administration to her as widow.²³ The absence of a marriage certificate is a gap in documentation, not a gap in the proof.

Shelton Clark⁹ Cisco, son of John⁸, was born at Cincinnati, Ohio, 31 August 1829, per the Calvin Sisco family Bible—the sole source for the given name “Shelton,” the civil record being uniformly “Clark Cisco.”²⁴ The identification rests on the complete concordance of all recorded particulars across the enlistment register,²⁵ the 1849 Jefferson County marriage record, the 1860 federal census, and Cornelius Vail’s Madison undertaker register recording Clark’s death on 4 May 1874.

Generation Ten: Anna Belle¹⁰ Cisco

Generation 10 is the second junction requiring extended proof. Anna Belle¹⁰ Cisco left no civil birth record naming her parents and no marriage record identifying either parent. The identification rests on two independent direct sources.

The first is the Jennings County, Indiana, Death Register, Book 1, page 68, recording Annie Bell King’s death on 4 June 1889 at North Vernon and naming her father as Clark Cisco and her mother as Sarah Hurley²⁶—an official government record transcribed directly from the register by Missy Pappenheim, Clerk/Registrar of the Jennings County Health Department, on 1 June 2026. The second is a notice in the Madison Courier of 6 June 1889: “Marshal Sisco went out to Vernon yesterday morning to attend the funeral of his niece, Mrs. Robert King.”²⁷ “Marshal Sisco” is Calvin⁹ Cisco, documented elder brother of Shelton Clark⁹ Cisco.²⁸ The word “niece” is unambiguous: a named, serving public official identified his relationship to Mrs. Robert King in his own community’s newspaper at the moment of family duty. Given that Calvin’s only documented brother in the Cisco line is Shelton Clark, the statement places Anna Belle as Clark’s daughter.

These two direct sources are corroborated by four of Anna Belle’s children’s own vital records independently naming their mother as Annie Bell Cisco; by the 1860 census enumeration of a female child of compatible age in Clark’s household; by the elimination of the only plausible alternative same-age Cisco candidate, Calvin’s daughter Annabel, who died in 1865; and by the statutory premise of the 1898 Mexican War minor’s pension claim filed in her name on Clark’s service.

Generations Eleven Through Fourteen

Don Carloss¹¹ King (b. Madison, Indiana, 17 February 1878; d. Chattanooga, Tennessee, 2 May 1944) is documented by his 1936 Social Security application, his 1944 Tennessee death certificate, and a 1942 Bureau of the Census certification of his 1880 enumeration.²⁹ Edith May¹² King (b. Decatur, Alabama, 15 May 1900; d. Orlando, Florida, 13 May 1984) is documented by the Tennessee Certificate of Birth for her daughter Edythe Virginia¹³ Haley (1920), which names “Edythe May King” as mother, born Decatur, Alabama.³⁰ Edythe Virginia¹³ Haley (b. Chattanooga, Tennessee, 23 March 1920; d. Greenville, South Carolina, 12 May 2003) is documented by the Tennessee Certificate of Birth no. 255 and the Hamilton County, Tennessee marriage bond for her 1941 marriage to Roy Allen Page Jr.³¹ Margaret Deborah¹⁴ Page (b. Charlotte, North Carolina, 7 March 1948) is documented by the North Carolina Certificate of Birth naming her father as Roy Allen Page Jr. and her mother as Edith Virginia Haley, corroborated by a Charlotte newspaper birth notice published the following day.³² Each parent-child link in these four generations rests on a primary vital record.

VI. GPS Compliance: A Summary Assessment

The first three GPS elements are demonstrated throughout. The search was reasonably exhaustive: primary records span pension files, family Bibles, the Philip Skene Papers, Ohio and Indiana deeds and probate records, census schedules from 1790 through 1920, vital records across eight states, DAR application files, military enlistment registers, undertaker registers, tax records, newspaper archives, and cemetery records; NARA was queried by formal order regarding the missing pension case file; a physical cemetery examination was conducted by a named librarian; and no competing candidate for Henry’s wife Ruth was identified in any Washington County record. Citation is complete and accurate, with full archival locations, image group numbers, and URLs throughout, and negative searches documented. Analysis and correlation is demonstrated: discrepancies are explained rather than ignored, including the Susan/Annie Belle census name, the Henry Francisco death date variance between the Bible and civil records, Silliman’s visual age estimate versus the sworn pension schedule, the variant Annabelle spellings in the 1826 deed, and the birth year discrepancy in the Solomon Francisco Bible for Mary Freeman.

The final two elements are equally satisfied. Conflicting evidence is resolved: both competing identifications at Generation 6 are affirmatively dismantled, the Susan/Annie Belle discrepancy at Generation 10 is explained, and all significant evidentiary conflicts across the line are addressed. The written conclusions at both difficult junctions are explicitly framed as proof arguments, the nature of the evidence is accurately characterized, and the reasoning from evidence to conclusion is laid out in full. The GPS has been fulfilled across all fourteen generations. Generation 6 fulfills it through convergent indirect evidence with both competing identifications eliminated—precisely the evidentiary situation the standard was designed to accommodate.

VII. Probabilistic Assessment of Residual Uncertainty

The GPS does not speak in probabilities, but the question of residual uncertainty is legitimate. For the identification of Ruth⁶ Fuller as Henry Francisco’s wife to be wrong, every one of the following would have to be true simultaneously: the exhaustive search failed to identify the actual competing candidate; Francisco’s placement in the Skene ledger alongside the Fuller brothers is coincidental; Aaron⁶ Fuller Jr.’s selection as wedding officiant is civic coincidence; Freelove⁶ Fuller’s confirmed presence at Whitehall is unrelated to the Francisco family; the Westcott family tradition is fabricated or derived from a wrong source; and Ephraim⁶ Fuller’s commemorative naming of daughters Ruth and Freelove is coincidental. Each proposition is individually improbable. If each is assigned a generous individual coincidence probability of 20 percent, the joint probability that all six are simultaneously coincidental is (0.2)⁶, or approximately 1 in 15,625. At a more conservative 10 percent, the joint probability falls to (0.1)⁶, or 1 in 1,000,000. Neither figure is actuarial; both are structured estimates illustrating the logical force of independent convergence. The residual probability of error at Generation 6 is below one percent under any reasonable individual assignment. The identification cannot be described as certain—convergent indirect proof never is—but it is established to a degree that satisfies any reasonable evidentiary standard.

VIII. On the Use of Artificial Intelligence in Genealogical Analysis

This article was produced through structured dialogue between its author and Claude (version Sonnet 4.6), a large language model developed by Anthropic, engaged as an analytical interlocutor rather than a research tool. Claude did not access archives or primary records; it evaluated the arguments as presented in the finished documents, subject to the same selection effects as any reader receiving a polished proof. It revised its assessments twice in the author’s favor—a confirmation bias signal acknowledged here. As a language model, its analytical outputs are probabilistic pattern-matching over training data, not independent expert judgment. What it contributed is best described as structured pressure: the requirement that counter-arguments be pressed to their hardest available form before being evaluated produced a more rigorous analysis than a single author reviewing his own work is likely to produce alone. The use of AI in genealogical research is widespread and underacknowledged; this article models explicit disclosure of how it was used, what role it played, and what its limitations were.

IX. Conclusion

The fourteen-generation Mayflower line of Margaret Deborah Page is GPS-compliant throughout. Generations 1 through 5 rest on the Silver Book. Generations 7 through 14 rest on primary vital records and legal instruments, with the partial exception of John⁸ Sisco’s marriage, which is documented in primary legal instruments even without a marriage certificate. Generation 6 rests on a GPS-standard convergent indirect proof argument, published in the New York Genealogical and Biographical Record, with both competing identifications affirmatively dismantled and no credible counter-evidence identified in any searched source.

The strongest available counter-argument—that the proof establishes Fuller association rather than Fuller maiden name, and that the Westcott application carries more evidentiary weight than its secondary nature strictly warrants—has genuine force and is acknowledged rather than dismissed. It reframes the Generation 6 conclusion from “proven” to “the most probable conclusion the surviving evidence supports, with no viable documented alternative”—which is, accurately, what convergent indirect GPS evidence produces. The author’s own language in the NYG&BR article—“circumstantial, but very strong—there is almost no room for doubt”—remains the right characterization. The adversarial analysis conducted here has not moved it.

The GPS is best understood not as a checklist but as a framework for intellectual honesty: about what the evidence is, about what it shows, and about the conditions under which the analyst’s own interests may be shaping the analysis. The most rigorous genealogical proof is one that has been pressed hard from both sides and has survived. This one has been.

---

NOTES

1. Jeffery Allen Record, “Ruth Fuller, Wife of Henry Francisco of Whitehall, New York,” New York Genealogical and Biographical Record 157 (July 2026): 231–236.

2. Jeffery Allen Record, “The Niece of Marshal Sisco: Identifying Annie Bell (Cisco) King of Madison, Jefferson County, Indiana” (prepared for publication).

3. Jeffery Allen Record, The Mayflower Ancestry of Margaret Deborah Page: An Edward Fuller Line (Lincoln, California: privately printed, May 2026).

4. Board for Certification of Genealogists, Genealogical Standards, 50th anniversary ed. (Nashville, Tenn.: Ancestry, 2014), 1–2.

5. Bruce Campbell MacGunnigle, Mayflower Families Through Five Generations, Volume Four: Family of Edward Fuller, 3rd ed. (Plymouth, Mass.: General Society of Mayflower Descendants, 2006), 161, 166–167.

6. Doris Begor Morton, Philip Skene of Skenesborough (Granville, N.Y.: The Grastorf Press, 1959), 36–37; Ward J. Roylance, Notes on the Fuller Family of New York State (Salt Lake City: the author, 1961), 35.

7. Frank J. Doherty, The Settlers of the Beekman Patent, Dutchess County, New York, 10 vols. (Pleasant Valley, N.Y.: F.J. Doherty, 1990–2010), 5:443.

8. Carol Clark Johnson, Fullers, Sissons, and Scotts, Our Yeoman Ancestors: 46 New England and New York Families (Mobile: the author, 1976), 277–280.

9. William Hyslop Fuller, Genealogy of Some Descendants of Edward Fuller of the Mayflower (Palmer, Mass.: C.B. Fiske & Co., 1908), 180.

10. Sharon, Connecticut, Vital Records, 1721–1879, Land Records vol. 7, p. 302 (marriage of Job Gould Jr. of Sharon to Ruth Fuller of Kent, 24 Feb. 1773).

11. Henry Francisco (Private, Col. Warner’s Regt., Revolutionary War), pension no. S44864, Case Files of Pension Applications, 1800–1900, Record Group 15, National Archives, Washington, D.C.; National Archives Catalog ID 54606831. Schedule of property dated 30 Aug. 1820, images 1, 5–6.

12. Ada Sophia (Palmer) Westcott (Member No. 47312), application for membership, National Society Daughters of the American Revolution, Ancestor No. A041636, filed 25 Jan. 1904, NSDAR Library, Washington, D.C.

13. Solomon Francisco Family Bible Record (1768–1886), in NSDAR Ancestor File No. A041636, NSDAR Library, Washington, D.C.; digital images at FamilySearch, IGN 007949717, images 652–653.

14. Crisfield Johnson, Washington County: Its History (n.p., 1878), 62, 114.

15. “Notes of hand & Bonds of Different persons at Skenesborough,” 1767–1773, Philip Skene Papers, Box 1, Folder 01, New York State Library, Special Collections, Albany, N.Y.; Document ID 90534.

16. Morton, Philip Skene of Skenesborough, 36–37; Roylance, Notes on the Fuller Family, 1, 35, 56.

17. Mildred Mason Petrie (Member No. 525551), application, NSDAR, Washington, D.C.; Washington County Deed Book A, vol. A-B-1, pp. 528–529, FamilySearch IGN 007138516, image 284; 1790 U.S. Census, Whitehall, Washington Co., N.Y., NARA microfilm M637, roll 6; Edward C. Starr, A History of Cornwall, Connecticut (New Haven: Tuttle, Morehouse & Taylor Co., 1926), 464.

18. “Notes and Queries,” Boston Evening Transcript, 11 Oct. 1926, Note #4475.

19. Henry Francisco pension file (note 11); Solomon Vine pension no. S43213; Thomas Lyon pension no. S9687; all in Case Files of Pension Applications, 1800–1900, Record Group 15, National Archives, Washington, D.C.

20. Solomon Francisco Family Bible Record (note 13).

21. Hamilton County, Ohio, probate entry, 4 Oct. 1837, FamilySearch IGN 005446001, image 345 of 367.

22. Letter from Dr. E. Francisco (Elon⁸ Francisco) to Thomas J. Francisco, 15 May 1886, Galesburg, Neosho County, Kansas; DAR supporting documentation for Ancestor No. A041636, Exhibit A, NSDAR Library, Washington, D.C.

23. Greene County, Ohio, Deed Book, deed of 22 June 1826, John Sisco and Anna Bella his wife to Calvin Sisco; acknowledgment before James Foster, Justice of the Peace, 27 June 1826; recorded 30 Oct. 1826; FamilySearch IGN 008330837, image 405 of 601.

24. Calvin Sisco Family Bible Record, transcript prepared 27 Jan. 1950 by Mary Hill from photostat of flood-damaged original; DAR supporting documentation for Ancestor No. A041636, NSDAR Library, Washington, D.C.

25. U.S. Army Register of Enlistments, 1798–1914, NARA Microfilm Publication M233, roll 23, p. 48, line 6.

26. Jennings County, Indiana, Death Register, Book 1, p. 68, Annie Bell King; Jennings County Health Department, North Vernon, Indiana; genealogy sheet transcribed by Missy Pappenheim, Clerk/Registrar, Vital Records Division, 1 Jun. 2026.

27. Madison Courier (Madison, Indiana), 6 Jun. 1889, p. 4; digital image, Newspapers.com.

28. “Calvin Cisco, Octogenarian,” Madison Courier, 14 Sep. 1908, p. 4; B.F. Bowen, Memoirs of the Lower Ohio Valley (Madison, Wis.: Federal Publishing Co., 1905), biographical sketch of John M. Cisco.

29. Don Carloss King, Application for Social Security Account Number (Form SS-5), signed 30 Nov. 1936; Tennessee Certificate of Death, Don Carlos King, Hamilton County, Certificate No. 10161, 1944; letter from J.C. Capt, Director, Bureau of the Census, to Don Carloss King, 25 Jun. 1942.

30. Tennessee Certificate of Birth No. 255, file no. 23301, Edythe Virginia Haley, born 23 Mar. 1920, West Ellis Hospital, Chattanooga, Hamilton County, Tennessee; Tennessee State Library and Archives.

31. Tennessee, County Marriages, 1790–1950, Hamilton County, marriage bond and license of Raymond C. Haley and Edyth King, bond no. 29086, dated 10 Apr. 1919; South Carolina Certificate of Death No. 544, Roy A. Page Jr., died 14 Apr. 1969.

32. North Carolina State Board of Health, Bureau of Vital Statistics, Certificate of Birth, Registration District No. 60-95, Certificate No. 1214, Margaret Deborah Page, born 7 Mar. 1948, Presbyterian Hospital, Charlotte, Mecklenburg County, North Carolina.