Sources Memos — Volume II

Bridge — Charleston, February 1714

Charleston, February 1714

The Volume I → Volume II hinge: Barbara Tempest Needham comes off the Hopewell into Charleston harbor, keeps eleven days at Mrs. Brewton's on Tradd Street, and marries Christopher VI de Graffenried at the Cooper Street church on Monday 22 February 1714 — after which Christopher makes the first entry on the fly-leaf of the family Bible in formal French, and writes two words in English beneath it.

Sources memo v1 — audit trail, not for publication. Documented first, section by section; every AR passage flagged; Gaps & Open Items at the end. Tier key follows the volume's spine: P primary / documented on a primary instrument · S secondary or contextual · R reference/aggregator · T tradition/family record · AR authorial reconstruction. Flags only; no corrections made. The documented layer is the volume's own embedded Working Bibliography (section-keyed) and Note on Sources; the Bridge draws its heroine, its "I am coming," and its "She came" from Volume I's already-established close, so this memo also carries the Volume I apparatus forward where it bears.

The Bridge is the thinnest-documented material in the volume and the most continuous with Volume I. Its single load-bearing documentary anchor is the marriage itself — Christopher VI de Graffenried to Barbara Tempest Needham, Charleston, 22 February 1714 — and the object that records it, the Christopher VI de Graffenried family Bible, whose fly-leaf carries the formal French marriage entry. Everything around that anchor — the crossing's arrival scene, the eleven days, the supporting cast, the wedding hour by hour, and the two-word English entry — is authorial reconstruction, most of it built on Volume I's own reconstruction rather than on new record.

§The arrival — the Hopewell comes up the Charleston roads

Documented / disciplined.

• Christopher VI de Graffenried is a documented historical person — son of Baron Christopher V de Graffenried of Worb, of the family that founded New Bern — P/S (the 1925 History of the de Graffenried Family, treated as a transcribing-and-tradition source, S; the family Bible, P). The New Bern half-built farm, the Carolina Land Office obligation, the Yamasee unrest at the south of the colony and the Tuscarora settlement at the north are period context — S.

• The Reverend Gideon Johnston — commissary at Charleston for the Bishop of London — is a documented historical figure of the period, and the officiant named on the marriage record — P/S.

• The family Bible carried from Worb is the documented object the whole volume is structurally about — given to Christopher VI by his father at Worb in 1710, fly-leaf clean, first entry made at the 1714 Charleston marriage — P (Working Bibliography §I).

AR — the arrival scene.

• The entire arrival sequence is authorial reconstruction. The brigantine's half-rigged foremast and the storm of 3 January, the south-southeast wind, Mr. Eldridge at the wheel, the 9:20 a.m. landing at the Cooper Street pier, Barbara at the rail in her wool cloak, Christopher on the bench "every morning since the eighth of January," the hand raised palm-forward at a quarter-mile, "here, here, I am here" — none of this is documented. It is reconstruction on the documented fact that the two married at Charleston in February 1714. The precise arrival date (Thursday 11 February) and the thirty-five-day count on the bench are authorial.

• The supporting cast is AR. Mr. Eldridge (the Hopewell's master), Mrs. Brewton (the Tradd Street widow-landlady), Susannah Fenwick and Mrs. Bellinger, the widow Mrs. Houghton, Mr. Pelletier, the merchant Allston, Hans the Swiss manservant, the runner Wilkes — all are reconstructed figures giving the scene its population. They carry no documentary citation and should be understood as invented texture on the documented marriage.

§The eleven days — Mrs. Brewton's on Tradd Street

Documented / disciplined.

• The "documented things" discipline is stated inside the prose. Barbara's letters home say "no more than the documented facts required" — the Hopewell had landed, she was well, she was at Charleston, she would be married on 22 February. This is the volume's own restraint rule enacted in scene: it is AR that she wrote the letters, but the content named is confined to what the record would bear.

• Christopher's Williamsburg lot. The prose has Christopher speak of "Lot 47, the Pitt-Dixon corner" at Williamsburg. The de Graffenried Williamsburg lot is plausibly a documentable land record but is not cited in the volume's bibliography; treat the lot number and corner as reconstruction/texture pending a Williamsburg land-record check (see Gaps).

AR — the eleven days.

• The whole eleven-day interval is reconstruction. The "floor that holds" motif (the bed that does not move under her, learning land after 106 days at sea), the daily rhythm (Christopher at seven, the harbor master's shed at ten, the parlor at four with the door open for propriety), the fourth-afternoon "I will not be ornamental" exchange, the chapel-garden-at-Wymondley callback — all AR, and much of it draws directly on Volume I's reconstruction of Barbara and the Wymondley courtship. The 106-days-at-sea and the "one hundred and six days" count are authorial.

§The wedding — Monday 22 February 1714, the Cooper Street church

Documented.

• The marriage is the Bridge's load-bearing fact: Barbara Tempest née Needham married Christopher de Graffenried at the commissary's (Cooper Street) church in Charleston, South Carolina, on 22 February 1714, the Reverend Gideon Johnston officiating — P (the family Bible fly-leaf entry, which the prose reproduces; the church register entry named in the prose as made "on the same day"). The formal French entry is transcribed in the 1925 History — S corroboration.

• The formal fly-leaf entry itself is documented text — the volume reproduces it: "Christophe de Graffenried, fils du Baron Christophe V de Bernberg, époux. Barbara Tempest née Needham, de Wymondley en Hertfordshire, épouse. Mariés en l'église du commissaire à Charleston…le vingt-deuxième jour de février, 1714." This is the first entry on the Bible fly-leaf — P.

AR — the wedding as scene.

• The wedding-day hour-by-hour is reconstruction. The 5 a.m. waking and counting the days backward, the altered dark-wool gown, the pearl earrings, the half-past-eight parlor, the witnesses walking to church, the thin cold rain, Eldridge's great-coat over her shoulders, the three-o'clock wedding-supper (Carolina ham, rice with "the new orange spice," custard, Allston's Madeira), the company leaving at six, Barbara asleep in the chair by nine — all AR on the documented fact of the wedding. The named witnesses (Mrs. Houghton and Mr. Pelletier for the bride; Hans and Allston for the groom; Eldridge in the second pew) are reconstruction.

• The physical description of Johnston ("a slight man of thirty-eight…six years and ten months at Charleston…in poor health that winter") is AR/S texture on the documented officiant.

§The fly-leaf — the formal French entry and the two English words

Documented.

• The Bible, the clean fly-leaf, the formal French entry, and the private English entry below it are the documented structure of the object — the volume's Working Bibliography §I states plainly that the first entry on the page is "Christopher VI's own hand at his marriage in Charleston, 22 February 1714 — the formal French entry the Baron had instructed him to write, with a private English entry below it." So the existence of a two-part fly-leaf entry (formal French above, private English below) is documented — P.

AR — and the trilogy-wide apparatus rule.

• The writing scene is reconstruction. The cut quill, "Mrs. Brewton's ink…a slightly browner ink," the ruler he decides not to use, the memory of his father writing his mother's 1697 death entry at Worb, the meditation that "he had never…written anything in English that he meant" — all AR on the documented two-part entry.

• ⚠ "She came" is authorial reconstruction — carried verbatim from the Volume I Chapter 6 memo. The Bridge's closing two-word English entry, "She came," is the same inscription the embedded Volume I Chapter 6 sources memo explicitly flags as invented — verbatim there: "She came, rendered as Christopher's smaller-hand entry on the Bible fly-leaf underneath the formal French inscription, is AR — it is invented," named "the chapter's single most authorially-asserted moment." The documented fact is that a private English entry exists beneath the French; that its words are "She came" is authorial reconstruction. Likewise "I am coming" — the three-word note from New Bern, 17 November 1713 — is the reconstructed close of Volume I, not a surviving document. Neither "She came" nor "I am coming" may ever be cited in any volume's apparatus as documented primary text. This is the single most important consistency rule for the whole trilogy's apparatus, and it originates here and in the Vol I Ch 6 memo.

Gaps and open items

1. ⚠ "She came" / "I am coming" are AR, trilogy-wide. The two-part fly-leaf entry (French above, private English below) is documented; the words of the English line are reconstruction, as Vol I's own Ch 6 memo states. Never cite either phrase as primary text. (Highest-priority apparatus rule; shared with Vol I and the Vol III frame.)

2. The church-register entry — verify it exists as a separate instrument. The prose asserts the marriage "was entered in the church register on the same day." The fly-leaf entry is the certain instrument; a surviving Charleston commissary's-church register entry would be independent corroboration. Confirm whether the register is in hand or is inference from the fly-leaf. (St. Philip's / commissary's-church records, Charleston.)

3. Christopher VI's Williamsburg Lot 47 ("Pitt-Dixon corner"). Named in dialogue as an anchor for the couple's planned move; not in the volume's bibliography. A Williamsburg land/lot record would document it or mark it as texture.

4. Christopher VI's death date (ODQ-V2-06). Not a Bridge fact but structurally attached to this Bible: the volume commits to an October 1742 death (for the Bible's accession to Tscharner) against the 1925 History's soft "ca. 1750." Open; affects Ch 1 §1.5 and Ch 6.

5. The Vol I continuity dates (17 Aug 1710 Baron gives the Bible at Worb; 17 Oct 1710 chapel garden at Wymondley; 17 Nov 1713 the four letters and the note at New Bern) are carried from Volume I's reconstruction, not independently documented here.

End of Bridge Sources Memo v1. Load-bearing result: the Bridge rests on one documented anchor — the 22 February 1714 Charleston marriage of Christopher VI de Graffenried and Barbara Tempest Needham, recorded in the formal French entry on the family Bible fly-leaf — and reconstructs everything around it (the arrival, the eleven days, the wedding as scene, the writing of the fly-leaf). The two carried flags: the whole scene is AR on a thin documentary base largely inherited from Volume I, and — most importantly — "She came" and "I am coming" are authorial reconstruction, never to be cited as documented text in any volume.

Chapter 1 — Lunenburg, Summer 1742

Lunenburg, Summer 1742

The elder generation and the first turning of the Bible: Anton Tscharner DeGraffenreid marries Mary Baker of Chowan (5 July 1742); his father Christopher VI dies at Williamsburg that October; the family Bible and the keeping pass to Tscharner, who enters the late marriage and his father's death on the fly-leaf; the name "Baker" comes into the line, carried forward in the son born 1744.

Sources memo v1 — audit trail, not for publication. Documented first, section by section; every AR passage flagged; Gaps & Open Items at the end. Tier key: P primary · S secondary/contextual · R reference/aggregator · T tradition · AR authorial reconstruction. Flags only; no corrections made. Documentary layer = the volume's embedded Working Bibliography (§§I–V, XIV) and the Searcy Ancestral Workbook v7.11.

Register note the audit must record up front: Chapter 1 is written at deliberate historian's distance. Unlike the scene-built chapters, it repeatedly refuses to reconstruct — "the historian does not stage him," "invents nothing to fill it," "the day is renderable to the country; it is not renderable to the room." So the chapter carries very little scene-level AR. Its reconstruction is almost entirely interpretive framing (the "hinge nobody could see," the meaning drawn from the documented shape), and it is honest about its own gaps in its own text. The chapter's real exposure is not invented scene — it is one load-bearing date the prose states as fact while the apparatus flags it as an unproven commitment. That is item 1 below.

§1.1 The country / §1.2 The Bakers and the match — the documented spine

Documented.

• Anton Tscharner DeGraffenreid, b. 28 November 1722 at Williamsburg, baptized twelve days later by Commissary Blair, the Honorable Nathaniel Harrison a godfather — P/S (family record; parish/Williamsburg record as named). Only child of Christopher VI and Barbara Tempest Needham.

• Christopher VI bought the house in 1717 on the south side of Duke of Gloucester Street, Williamsburg — S (documentable Williamsburg property record; named in the prose, treat as S pending citation).

• The Baker descent and the colonel's standing are carried on primary and reference records — P/R/S: Lawrence Baker of Surry, will proved 1681; Henry Baker's Surry patents 1683–84 and marriage to Mary Blake (daughter of Col. John Blake of Nansemond); Col. Henry Baker II of Chowan — Bucklands (~2,500 acres), militia colonel, sheriff, captain, burgess, JP of Chowan commissioned 1731, the Henry Baker Ferry (in use before 1722). Will of Col. Henry Baker II proved 1739 — names daughter Mary — P (Working Bibliography §II). Will of Edmund Chancey, 1753 — names grandchild "Mary DeGrafinred" — P (§II), fixing Mary's mother as Ruth Chancey, daughter of Edmund Chancey of Pasquotank. Together these two wills "close the elder Mary Baker's parentage on both sides."

• The marriage: Tscharner DeGraffenreid × Mary Baker, 5 July 1742, in Chowan, North Carolina — P (the fly-leaf entry the volume reproduces; also treated as the anchor date in the Workbook). The date-to-the-day and country-to-the-ground are documented; the room and the officiant are not (the prose says so plainly and builds neither).

• The name "Baker" enters the line in 1742 and is carried forward in the first child, Baker DeGraffenreid, b. 6 August 1744 — P (family/Bible record).

AR / disciplined framing.

• The tobacco-frontier country picture (the James drainage, Nottoway/Meherrin/Roanoke, Brunswick County in 1742 → Lunenburg cut from it in 1746) is period context (S), correctly handled — the chapter openly notes the seat "did not exist in the summer of 1742" and that later chapters' lane/oak/table were "neither built nor in a county that had its later name."

• Elder Mary Baker "married at about nineteen" is inference on the family dates, correctly hedged.

• The chapter's refusal to invent the courtship ("how the match was made the record does not say…the historian does not invent the courtday") is discipline enacted, not a flag.

• The "no was hers to say" passage (whether the marriage was the bride's choosing) is interpretive framing that names conditions and invents no feeling — disciplined AR, correctly bounded.

§1.3 The Bible at Williamsburg; the hand-convention — the documented object

Documented.

• The Bible object (octavo ~8×5, dark calf, given clean at Worb 17 August 1710, first entry 1714) and the fly-leaf convention (formal French, keeper's own hand, reserved page) are documented structure — P (Working Bibliography §I). The Bernberg Bible (the Baron's, at Worb; 1697 death entry of Christopher VI's mother) is the convention's origin — S (§I), and the chapter correctly keeps the form (which crossed) distinct from the object (which stayed at Worb).

• The first fly-leaf entry — the reproduced French marriage entry of 1714 — is documented text on the Bible — P.

AR — carried flag.

• ⚠ "She came" beneath the French is authorial reconstruction (Vol I Ch 6 memo; Bridge memo). §1.3 reproduces it — "the smaller English words She came beneath it" — and glosses it as "the family's record" of Barbara's crossing. The two-part entry is documented; the words "She came" are invented. Never cite as primary. (Trilogy-wide rule.)

§1.4 The wedding at distance / §1.5 The October accession — the load-bearing inference

Documented.

• The wedding date and country (5 July 1742, Chowan) — P, rendered "to the country; not to the room," with the officiant and building left unnamed as the record leaves them.

• The colonel was three years dead at the wedding (will proved 1739; Bucklands to the third Henry Baker) — P — so the chapter correctly does not stage him.

• Baker DeGraffenreid b. 6 August 1744 — P (repeated anchor).

• Tscharner's fly-leaf marriage entry (reproduced French: "Tscharner de Graffenried…Marie Baker, fille de feu le colonel Henry Baker…Mariés en Chowan…le cinquième jour de juillet, 1742") — P if a genuine transcribed fly-leaf entry (the 1925 History "transcribes the relevant fly-leaf entries"); see item 2 — verify it is transcribed and not reconstructed in the convention's style.

AR / the chapter's central exposure.

• *⚠ 1. Christopher VI's October 1742 death is a volume commitment, not a documented fact — and §1.5 states it as fact. The prose opens §1.5: "Christopher VI DeGraffenreid died at Williamsburg in October 1742, on a day the surviving record does not name." But the volume's own Working Bibliography (§II and ODQ-V2-06) states the opposite of certainty: the 1925 History "places him settled permanently in Prince George County and gives a soft 'ca. 1750'; the volume commits to an October 1742 death for the Bible's accession. No hard probate date is in hand." The entire §1.5 "October accession" structure — the father dead the same autumn as the July wedding, the two entries (marriage-entered-late and father's-death) falling to the son's hand in one season — rests on this un-documented date. If Christopher VI in fact died ca. 1750 (the printed source's reading), Tscharner did not become keeper in 1742 and the chapter's climactic accession is displaced by up to eight years. This is the single largest load-bearing inference in the chapter, and the prose's flat "died…in October 1742" reads with a certainty the apparatus does not have. → The author's call: either soften the prose to name the commitment (as the bibliography does), or run the Prince George County / ECU (MC 0796) search that ODQ-V2-06 targets and settle it.*

• ⚠ 2. The accession scene — Tscharner taking up the pen, cutting it, entering first the marriage then his father's death — is reconstruction on the commitment above. The chapter is careful ("How the book came from the father's desk to the son's keeping the record does not say…the historian does not invent the journey"), but the sequence and same-hand-same-autumn drama are authorial, and depend on the 1742 death. The death entry's wording is explicitly withheld ("The historian does not give the words of it") — good discipline. But whether Tscharner's 1742 marriage entry exists on the fly-leaf as reproduced, or is an authorial reconstruction in the convention's style, must be verified against the 1925 History's transcription or a Bible image. (Given "She came" is invented, the audit cannot assume every reproduced French entry is transcribed.)

• The "rhyme" between the father's church-and-fly-leaf marriage and the son's roomless one is interpretive framing (AR), correctly not pressed into a claim.

Gaps and open items

1. ⚠ Christopher VI's death date (ODQ-V2-06) — load-bearing and stated as fact. Prose: "died…in October 1742." Apparatus: a commitment against the 1925 History's "ca. 1750," no hard probate. The whole §1.5 accession hangs on it. Retrieval: Prince George County VA will/probate or land record; ECU DeGraffenreid Association Collection (MC 0796). Highest-priority Ch 1 decision.

2. Verify Tscharner's 1742 fly-leaf marriage entry is transcribed, not reconstructed. Check the 1925 History transcription or a Bible fly-leaf image. Applies to the reproduced French entry in §1.5 (and, by extension, the audit's confidence in every reproduced fly-leaf entry).

3. ⚠ "She came" is AR (§1.3). Carried trilogy-wide flag; never cite as primary text.

4. The elder Mary Baker's death (ca. 1756). Not a Ch 1 event, but the bibliography flags that early drafting's "7 June 1773" was "a false precision on a wrong year and was corrected"; a primary death record is still wanted (§VIII). Confirm no residue of the wrong date survives in Ch 1's handling of the namesake-grandmother.

5. Officiant/building of the 5 July 1742 wedding — correctly left open. Named here only to confirm the prose invents neither; no action unless a Chowan parish record surfaces.

6. Williamsburg property (1717 purchase) and the baptism particulars (Blair; Harrison godfather) — name the citations at manuscript-assembly stage; currently stated as fact, plausibly documentable.

End of Chapter 1 Sources Memo v1. Load-bearing result: the documented spine is solid — two Chowan wills (1739, 1753) close the elder Mary Baker's parentage, the 5 July 1742 marriage is fixed to day and country, the Bible and its convention are documented, and the name "Baker" enters the line with the 1744 son. The chapter's discipline of refusing to stage what the record does not hold is exemplary. Its one real exposure is a date, not a scene: the prose states Christopher VI's October 1742 death as fact, while the volume's own apparatus flags it as an unproven commitment against a printed "ca. 1750" — and the chapter's entire closing accession depends on it.

Chapter 2 — The Child at Lunenburg

The Child at Lunenburg

The heroine's childhood: Mary Baker DeGraffenreid ("Polly"), born 3 September 1764 at Tscharner's south-side seat, from the floor up to her eleventh year — the household of three generations (the grandfather Tscharner and his Bible; the father Baker; the mother Sarah Vass; the brother Vincent), the death of an infant sibling, and the father's ride over the Carolina line in the autumn of 1775 that the record shows was the last time he was seen alive.

Sources memo v1 — audit trail, not for publication. Documented first; AR flagged; Gaps & Open Items at the end. Tier key: P primary · S secondary · R reference · T tradition · AR authorial reconstruction. Flags only; no corrections made. Documentary layer = the volume's embedded Working Bibliography (§§II–IV, VIII, XIV) and the Workbook v7.11.

Register note: Chapter 2 is the volume's first fully scene-built, close-third chapter, told from the child's sensory position ("the world came to her as the undersides of things"). It is therefore the most heavily reconstructed chapter so far — but it frames itself honestly with a Historian's Opening and Historian's Closing that carry the documented anchors and name the four-year silence without building across it. The body invents interiority on a documented household — which is the chapter's proper method — with two exceptions that invent events, not just feeling (the infant sibling's death, and the graveside naming scene). Those are the flags.

Historian's Opening & Closing — the documented frame

Documented.

• Mary Baker DeGraffenreid ("Polly"), b. 3 September 1764 — P (Reynolds canonical record; Workbook). Named for her father's mother, the elder Mary Baker.

• The household of three generations at the seat: the keeper-grandfather Tscharner (b. Nov 1722, ~41 in 1764); the father Baker DeGraffenreid (b. 6 Aug 1744), m. Sarah Vass; the brother Vincent ("not yet two" in autumn 1764 → b. ~1762–63); and Polly — all P/R (family record; Workbook).

• Tscharner's marital history: first wife the elder Mary Baker "some eight years dead" by 1764 (→ d. ca. 1756); married twice more; third wife Eliza Allen Embry, m. the year before (1763) — P/R ("the record carries the marriages and the dates and little more").

• The seat (built early 1750s on the ground taken up 1742), Lunenburg cut 1746, tobacco country — S/inference (as in Ch 1).

• ⚠ Load-bearing (well handled): Baker DeGraffenreid's last-alive date. He "put his hand to a deed of trust in Guilford County, North Carolina, on the 20th of October 1775 — the last to set him among the living." Then silence until Sarah Vass acting as administratrix by 10 September 1779 (Lunenburg Order Book 14, p. 38). "Between…lies a silence of nearly four years the surviving sources do not fill" (ODQ-V2-01). Family tradition (Reynolds) holds he was "killed at the outset of the Revolutionary War"; the chapter "carries the tradition and has not found the proof, and names the gap rather than building across it." — P (deed, D; order book, D) + T (tradition, correctly labeled). This is model discipline: the volume's central kept silence is stated as a silence.

§2.1–§2.2 The floor and the house — reconstruction on a documented household

AR (interiority on documented frame).

• The entire child's-eye world is reconstruction: the far ceiling and moving light, the fire, "the hands" (the mother's), being carried between warm and cold, the long table she sits under and then stands at, the oak at the head of the lane, the well, learning the order of the household day. None of the sensory content is documented; all of it is authorial close-third on the documented fact of a child growing up in that household. This is the chapter's proper and declared method.

• The bowl with "Polly" cut on the underside (§2.2) — a lovely, entirely invented object; the name Polly is documented (diminutive of Mary), the bowl and its cut marks are AR.

• Vincent as the elder brother "ahead of her in everything" — documented sibling (P); the characterization is AR.

• Consistency check — Vincent's birth ~1762–63. "Not yet two" in autumn 1764. This must square with Chapter 5, where Vincent makes a substitute enlistment at Lunenburg in January 1777 "at age fourteen" (Working Bibliography §V; ODQ-V2-03). b. 1762–63 → age 14 in Jan 1777 is consistent. Confirm the birth year is fixed the same way in both chapters. (Minor; flagged for internal consistency.)

§2.3 The grandfather's room and the Bible — the graveside naming

Documented.

• The Bible on the desk in the grandfather's front room, in Tscharner's keeping "since his father's death in 1742" — P (the object; Working Bibliography §I).

• The elder Mary Baker's grave in the family burying-ground — P ("Family burying-ground, Lunenburg (D, by use)"; §VIII), the ground "holding the elder Mary Baker's stone, where the child meets the Bible in Chapter 2." So the ground and a grandmother buried in it are documented.

AR — invents an event, flag.

• ⚠ The graveside scene is reconstruction of an event, not just interiority. The grandfather taking Polly out to the rail-enclosed ground, kneeling in the wet grass, letting her hold the Bible against her, telling her the stone is her grandmother's and that the name cut in it is her own — this whole scene is authorial. What is documented: the ground, the grandmother buried there, the child named for her. What is invented: that this specific instruction-at-the-stone happened, and — materially — that a legible cut headstone bearing the elder Mary Baker's name stood in the ground. The bibliography flags the elder Mary Baker's death as "ca. 1756 (P; primary record pending)" and notes an earlier "7 June 1773" was "a false precision on a wrong year and was corrected." A dated, legibly-inscribed stone is an assumption on a grave whose very date is a soft estimate. → Author's call: is the stone (and its being read to the child) supportable, or should the scene rest only on "the ground" and a grandmother in it, without asserting a cut inscription?

§2.4 The littlest — an infant sibling's death (the chapter's largest event-level AR)

AR — invents an event; the audit's principal Ch 2 flag.

• ⚠ "The littlest" — a younger sibling who is born, sickens, dies, is laid out, and is buried in the family ground — is a fully reconstructed event, and the audit cannot confirm the child existed. The chapter builds an entire sequence: the infant in the cradle Polly once lay in; the illness (the closed back room, the kept fire, the sharp smell, the night candle); the death; the neighbor-women and kin arriving to wash and lay out the body; the wake; the burial "short among the older places" with no cut stone; and Polly, drawn into the household's labor, doing "the next thing" through it. It is powerful and it is the chapter's emotional engine — but it asserts a birth and a death. The volume's own Note on Sources sets the rule ("set the known thing down…and do not invent to fill the silence"), and the Kit's worked examples name exactly this hazard ("the birth the record tidies away"). → Author's call: is there a documented DeGraffenreid infant death at the seat ca. 1766–1774 (a Bible entry, a family record) that this scene renders? If yes, cite it. If the sibling is authorial, the reader-facing note must not let this read as a documented death — it is the clearest "invented event on a silent record" in the chapter.

• The laying-out and wake customs, the neighbor-women, the grandfather's words over the small grave — AR on whatever the underlying fact is.

§2.5 The gathering war & the father's ride — the last-alive departure

Documented.

• The coming Revolution reaching the household as news (more riders, court-day changed, the busier road) is period context — S, kept correctly at the child's edge-of-comprehension distance.

• The father's ride "over the line into the Carolina country on the place's business, in the cold coming-on of the year" is the dramatization of the 20 October 1775 Guilford County deed-of-trust trip — his last documented appearance alive (P, §III).

AR / disciplined inference.

• The departure as final farewell is reconstruction on the documented last-alive date. What the record gives: Baker was alive in Guilford on 20 Oct 1775, and by Sep 1779 his widow was administratrix. What the chapter stages: an ordinary cold-morning goodbye (the wide hand on the child's head, the horse down the lane and under the oak) that the reader knows was the last. The Historian's Closing is scrupulous that this is "the last the record holds of Baker DeGraffenreid alive" — the scene is AR, the frame is honest. The emotional freight (that this ordinary errand was the end) rests on the four-year silence, correctly named as a silence. Good discipline; flagged only so the reader-facing note keeps "last seen alive" distinct from "died on this trip" (the record does not say he died on the road, only that he was never documented alive again).

Gaps and open items

1. ⚠ "The littlest" (infant sibling death, §2.4) — verify or mark as authorial. The chapter asserts a sibling's birth and death. Confirm a documented DeGraffenreid infant death ca. 1766–1774, or ensure the reader-facing apparatus does not present this as documented. Highest-priority Ch 2 flag (an invented event, not interiority).

2. ⚠ The grandmother's headstone (§2.3). The ground and the grandmother are documented; a legibly-inscribed, dated stone read to the child is an assumption on a grave whose date is a soft "ca. 1756" (primary record pending). Decide whether the cut stone is supportable.

3. ⚠ Baker DeGraffenreid — where and when he died (ODQ-V2-01). Last alive 20 Oct 1775 (Guilford deed of trust); widow administratrix by 10 Sep 1779 (Lunenburg OB 14 p.38). The volume's central kept silence. Retrieval: Guilford Co. NC; Pittsylvania and Henry Cos. VA — probate/deed/court records 1775–1779. (Handled with model discipline in-text; the open item is documentary, not a prose flag.)

4. Vincent's birth year (~1762–63) — internal consistency with Ch 5. "Not yet two" in autumn 1764 must match the "age fourteen" January 1777 substitute enlistment (§V; ODQ-V2-03). Confirm one fixed birth year across both chapters.

5. Elder Mary Baker's death record (ca. 1756). Primary record still wanted (§VIII); firms both the namesake framing here and the stone question in item 2.

6. Eliza Allen Embry's presence at the seat in 1764. Marriage documented (~1763); her being resident and "keeping the front rooms" in 1764 is inference — minor, correctly light-touch.

End of Chapter 2 Sources Memo v1. Load-bearing result: the documented frame is exemplary — the birth (3 Sept 1764), the three-generation household, and above all the four-year silence around Baker DeGraffenreid (last alive 20 Oct 1775; widow administratrix by 10 Sep 1779) are stated as record and as silence, with the family tradition labeled tradition. The body is close-third reconstruction on that documented household, which is its declared method. Its two exposures both invent events rather than interiority: the infant sibling "the littlest," whose birth and death the audit cannot confirm, and the graveside naming scene, which assumes a legibly-inscribed grandmother's stone on a grave whose date is only estimated.

Chapter 3 — The War in the House

The War in the House

The four-year not-knowing: Baker DeGraffenreid overdue and then unaccounted; the boy Vincent gone as a substitute of fourteen into the line (winter 1777); the girl Mary Baker keeping her brother in letters and learning the family hand; the estate passing to the widow (Sept 1779); the marriage to Gideon Johnson set for November 1779 — and, at the desk the second week of that November, the two words "we continue" written for the first time.

Sources memo v1 — audit trail, not for publication. Documented first; AR flagged; Gaps & Open Items at the end. Tier key: P primary · S secondary · R reference · T tradition · AR authorial reconstruction. Flags only; no corrections made. Documentary layer = the volume's embedded Working Bibliography (§§II–V, XIV).

Register note: Chapter 3 is the volume's emotional center and, like Chapter 2, a scene-built close-third chapter framed by a Historian's Opening and Closing. Its whole interior — the overdue man, the kept place at the table, the crossing-over of the family's waiting — is reconstruction on one documented fact and one documented silence: Baker was alive in Guilford on 20 Oct 1775, and his widow was administratrix by Sept 1779. The chapter's single most important item for the apparatus is that this is where the trilogy refrain "we continue" is invented — and, to its great credit, the chapter's own Historian's Closing says so outright.

Historian's Opening & Closing — the documented frame

Documented.

• Baker DeGraffenreid's last-alive date and the four-year silence — last alive on the Guilford County deed of trust, 20 October 1775 (P, §III); Sarah Vass administratrix by September 1779 (Lunenburg Order Book 14 p. 38, P, §IV). "The record never says he died… leaves the gap a gap." Tradition (Reynolds): killed at the outbreak (T, labeled). ODQ-V2-01. Model discipline, repeated from Ch 2.

• Governor Richard Caswell's grant to Sarah DeGraffenreid, 13 November 1779 — 400 acres, Guilford County (Deed Book 2, pp. 21, 94) — P (§III). The widow made a freeholder in her own name.

• The marriage: Mary Baker DeGraffenreid × Gideon Johnson, 18 November 1779 — P (the Bible marriage entry the grandfather wrote in; DAR corroboration). Gideon Johnson Jr., b. Amelia County, Virginia, 7 November 1754, private of the North Carolina line — P (pension file S4456; DAR Patriot A063204, §§V–VI).

• Vincent's service: a substitute of fourteen who "served in the Virginia line to the autumn of 1781" — P/D (pension file R.2845, §V; see consistency flag below on "Virginia line" vs. the North Carolina engagements).

• The first child of the new household: Sarah Johnson, b. 9 October 1780 — P — "the descent this history follows runs through her." Cross-volume anchor: this is the same Sarah Johnson (b. 9 Oct 1780) who opens the Volume III Bridge's four-daughter chain — the volumes agree.

• The Bible stayed on Tscharner's Lunenburg desk "his to keep until 1794, and then came to Mary" — P (§I; the accession rests on the structural fact per ODQ-V2-06).

§3.1–§3.3 The not-knowing, the boy, the arrangement — reconstruction on the silence

AR (interiority/scene on documented frame).

• The overdue-man interior (the kept place at the head of the table, the mother and child looking up the lane in the evenings, the slow crossing from "late" to "gone") is entirely reconstruction on the documented four-year silence. It invents no fact the record denies — no body, no grave, no date is asserted; the chapter is scrupulous that "nothing came." This is the chapter's method working as designed.

• Vincent's going as a substitute — the cold-morning departure scene is AR; the underlying fact (substitute of fourteen, into the line in the winter of 1777) is documented via pension file R.2845.

• The brother–sister letters and the girl's house-book are AR. No such letters or book survive; the Note on Sources states the volume's rule that "the letters and ledgers…do not survive but [their] habit does." The house-book given "the spring she was thirteen" is an invented object — and it matters, because §3.4 has her write "we continue" into it (see below). Flag that the book itself is reconstruction.

• The courtship/arrangement of the Gideon Johnson match (meeting at a cousin's, the terms corresponded out by post, Gideon asking the girl herself "apart, the supper done and the candle between them") is AR. The marriage (date, Bible entry) is documented; the courtship is invented — and the prose is careful ("What passed there was hers").

• The estate passing to the widow (Sept 1779) is rendered as scene but rests on the documented administratrix appointment (OB 14 p. 38). The "two places parted — the book's given over, the table's kept his" is interpretive framing (AR) on the documented fact.

⚠ Open-question exposure — Gideon's service status at the wedding (ODQ-V2-02). §3.3 presents Gideon as "a private of the Carolina line who had been away in it and was now, in the way of the war, home." This leans toward service concluded by the November 1779 wedding. The bibliography flags ODQ-V2-02 as open: whether Gideon served between his February 1779 discharge and the November wedding, or after it, is undocumented, and "whether his service was understood, that morning, as concluded, paused, or continuing" is called "the instrument behind the volume's central question." Chapter 4 is where this is load-bearing; here it is lighter, but note that the "home from the line" framing commits softly to a reading the record leaves open. → Defer the decision to Ch 4; flag the soft commitment here.

§3.4 The writing-evening — where "we continue" is invented

AR — the trilogy refrain; the audit's principal Ch 3 flag (and the chapter names it itself).

• ⚠ "we continue" is authorial reconstruction, first written here. The whole writing-evening — fifteen-year-old Mary Baker, the second week of November 1779, writing her brother, coming to the foot of the page, finding "the word the doing had wanted, which was two words," writing "we continue" under her name, then writing it a second time in the house-book — is invented. There is no surviving letter and no surviving house-book; the two words are the author's, not a documented inscription. This is consistent with the Volume III Bridge/Frame memo's trilogy-wide rule: the three refrains ("She came," "we continue," "she held the line") are authorial and "must never be cited as documented primary text."

• To the chapter's great credit, its Historian's Closing states this outright: "What the girl had written under her own name twice that November is not in the Bible; the Bible carries the marriage and nothing else of the evening." So the prose itself marks "we continue" as un-recorded. The apparatus rule is therefore already honored inside the text — the memo carries it forward so the reader-facing Note (and any epigraph or marketing use) never presents "we continue" as a quotation from a document. The Epilogue ("We Continue") turns on this same invented phrase — see the Epilogue memo.

Consistency & open items

1. ⚠ "we continue" is AR (trilogy-wide). Invented at §3.4; the chapter's own Closing confirms it is "not in the Bible." Never cite as documented text in any volume's apparatus. (Highest-priority Ch 3 flag; shared rule with "She came" / "she held the line.")

2. ⚠ Vincent's line — "Virginia line" (Ch 3) vs. the engagements audited in Ch 5. Ch 3 says Vincent "served in the Virginia line"; the bibliography's Chapter 5 material names Brandywine, Stony Point, Guilford, Yorktown (Continental service, pension R.2845), and Gideon's file is the 2nd North Carolina line. Confirm whether Vincent was in the Virginia or a North Carolina Continental line, and that Ch 3 and Ch 5 name it the same way. (ODQ-V2-03 — Vincent's full enlistment history is open.)

3. ⚠ Gideon's service status at the November 1779 wedding (ODQ-V2-02). Ch 3 softly frames him as "home" from the line; the question is open and load-bearing in Ch 4. Decide the reading once, apply it to both chapters.

4. The house-book and the brother–sister letters are invented objects. No such items survive; the volume's rule is that the habit is documented, not the artifacts. Ensure the reader-facing note does not imply a surviving letter or diary — especially since "we continue" is staged as written into them.

5. Cross-volume anchor confirmed: Sarah Johnson b. 9 October 1780 matches the Volume III Bridge chain. No action; noted as a point where the volumes agree.

6. The 13 Nov 1779 Caswell grant vs. the 18 Nov wedding vs. the writing-evening "second week of November." The chapter sequences the sealed letter, then the 13 Nov grant "the morning after," then the 18th wedding. Confirm the internal calendar holds (the "second week of November" writing-evening must precede the 13th).

End of Chapter 3 Sources Memo v1. Load-bearing result: the documented frame is strong and consistent across volumes — the 20 Oct 1775 last-alive deed, the Sept 1779 administratrix appointment, the 13 Nov 1779 Caswell grant, the 18 Nov 1779 marriage to Gideon Johnson (b. 7 Nov 1754), and the first daughter Sarah Johnson (b. 9 Oct 1780) who carries the line into Volume III. The interior is reconstruction on the four-year silence and invents no fact the record denies. Its defining apparatus item is that the refrain "we continue" is invented here — and the chapter's own Historian's Closing honorably says the two words are "not in the Bible." Two consistency items for the author: Vincent's line (Virginia vs. North Carolina) against Chapter 5, and Gideon's open service status (ODQ-V2-02) carried into Chapter 4.

Chapter 4 — The Wedding Morning

The Wedding Morning

The marriage of Mary Baker DeGraffenreid and Gideon Johnson Jr., near-real-time from the bride's waking in the dark to the company riding home — Thursday, 18 November 1779, at Sarah Vass DeGraffenreid's borrowed house on the south side of the Dan in Guilford County, North Carolina; the grandfather Tscharner enters the marriage on the Bible fly-leaf in French, and the household signs beneath it.

Sources memo v1 — audit trail, not for publication. Documented first; AR flagged; Gaps & Open Items at the end. Tier key: P primary · S secondary · R reference · T tradition · AR authorial reconstruction. Flags only; no corrections made. Documentary layer = the volume's embedded Working Bibliography (§§I–VI, VIII, XIV).

Register note: Chapter 4 is the volume's most granular real-time reconstruction — the gown sewn from onion-skin dye, the twelve bone buttons, the corn-mush eaten standing, the ceremony wording, the signing order, the supper conversation. Almost the entire chapter is authorial scene on one documented event (the 18 November 1779 marriage, recorded in the Bible). That is the chapter's method and is legitimate. But this chapter carries the audit's first hard factual error — a retired death-date that the prose still prints, twice — and it is the Chapter-4 counterpart of the Volume III Reuben-age error. That is item 1.

The documented spine

Documented.

• The marriage: Mary Baker DeGraffenreid × Gideon Johnson Jr., 18 November 1779, at the bride's mother's house in Guilford County, North Carolina — P (the Bible marriage entry, reproduced in French: "Gédéon Johnson…époux. Marie Baker DeGraffenreid…épouse. Mariés à la maison de la mère de l'épouse dans le comté de Guilford…le dix-huitième jour de novembre, 1779."). The entry's own text supports a home marriage ("à la maison de la mère").

• Gideon Johnson Jr. — private, 2nd North Carolina Continental Line under Captain Armstrong and Colonel Martin, ~15 months 1776–77 — P (pension file S4456; DAR A063204). B. Amelia Co. VA 7 Nov 1754 (Ch 3).

• Tscharner (56 in Nov 1779; keeper since his father's 1742 death; married 1742), Sarah Vass DeGraffenreid (35; married Baker 1762; administratrix; the 13 Nov 1779 Caswell grantee), and Vincent (b. ~Feb 1763; "sixteen years and nine months…as of the sixth of this month") — P/R. Vincent's age here is internally consistent with Ch 2 ("not yet two" in 1764) and the Jan 1777 enlistment at fourteen.

• The Caswell grant discussed at the supper (400 acres, in Sarah's name, filed May, written to her in October, survey to be returned to the Land Office at Hillsborough) — P (§III), rendered as conversation.

The two date errors — the chapter's principal findings

⚠ 1. The elder Mary Baker's death printed as "7 June 1773" — twice — a retired date the bibliography says was corrected to ca. 1756.

• §4.4 (the open Bible at the supper): "the previous page's last entry — the death of his wife the elder Mary Baker at Lunenburg on the seventh of June of 1773 — facing it on the verso."

• §4.4 (the London clock): the clock "the elder Mary Baker had wound every Sunday morning of her keeping of the household at Lunenburg from 1742 to 1773."

• This is a flat contradiction of the volume's own record. Working Bibliography §VIII: "Elder Mary Baker DeGraffenreid — death, ca. 1756… The figure '7 June 1773' carried in early drafting was a false precision on a wrong year and was corrected." The corrected-away date is still in the Chapter 4 prose in two places.

• It also contradicts Chapter 2, which builds the heroine's childhood on a grandmother "eight years in the ground" by 1764 — i.e., dead ca. 1756 — "the grandmother… whom she would never know," "named for someone into the ground before you were born." If the elder Mary Baker lived to 1773, she was alive until the heroine was eight, and Chapter 2's graveside scene and namesake framing collapse. Chapters 2 and 4 cannot both be right; the bibliography sides with Chapter 2 (ca. 1756).

• This is the exact species of error the audit exists to surface — an internal contradiction that also runs against the corrected record, load-bearing for two chapters and for a scene (the Bible "losing and gaining held on the open page" in §4.4 depends on the 1773 death entry facing the 1779 marriage). → Author's call (no correction made): reconcile Chapter 4 to ca. 1756 — the "death entry facing the marriage" and the "1742 to 1773" clock line both need the grandmother's death moved off 1773, or reworded to what the record actually supports (a death ca. 1756, entered on an earlier page). Highest-priority Vol II flag to date.

⚠ 2. Gideon's service status — "home now, since February" — states as fact what ODQ-V2-02 holds open.

• §4.1: Gideon "was home now, since February" (Feb 1779); the supper has the household treating his soldiering as behind him. The bibliography flags ODQ-V2-02 as open: whether Gideon served in the nine-month gap between a February 1779 discharge and the November wedding, or in the southern campaign immediately after, is undocumented, and this is "the instrument behind the volume's central question." The chapter commits to discharged-and-home; the record does not settle it. → Author's call: soften to the open reading, or run the 2nd NC muster-roll retrieval (ODQ-V2-02). (Carried from Ch 3, where it was softer; here it is stated flatly.)

The reconstruction (the chapter's method)

AR — nearly the whole chapter, on the documented marriage.

• The entire morning is invented scene: the bride waking in the dark, the closed journal under her hand, the gown (onion-skin-dyed pale yellow, twelve bone buttons), the washing and dressing, the corn-mush eaten standing, the laid-out parlor, the ceremony wording ("Will you hold him in this life and not put him aside"), the supper (the six-pound ham, the Madeira in trade for a quarter of beef, the boundary-line talk, the laughter). None is documented; all is close-third on the fact of the wedding. This is legitimate and declared.

• The four Johnson witnesses — Joseph (uncle, 320 acres), John (cousin), Elizabeth — and the whole signing sequence (bride signing "Mary Baker DeGraffenreid Johnson," each witness signing beneath) are reconstruction pending confirmation that the Bible entry actually bears witness signatures. The marriage is documented; the witness roster and their signatures on the page are not cited in the bibliography. → Verify against a Bible-page image whether the entry carries signatures; if not, the named witnesses are invented texture.

• The officiant. The entry names a place (the mother's house) but no minister; the chapter has Tscharner lead the vows himself (no clergyman present). Whether a minister officiated a 1779 Guilford home marriage is open; the grandfather-officiant is reconstruction. Mild flag.

• ⚠ "we continue" recap. §4.1 revisits the phrase — "in the letter…in the journal…It was not in the Bible… The phrase was hers." This correctly reiterates the Ch 3 finding: the two words are the heroine's invention and are not in any record. Good internal consistency with the trilogy-wide AR rule; no new exposure, but confirms the apparatus must never cite "we continue."

• Christopher VI's "death at Williamsburg in the autumn of 1742" is referenced again (§4.1) — carries the ODQ-V2-06 Ch 1 flag (the October 1742 date is a commitment, not documented).

Gaps and open items

1. ⚠ Elder Mary Baker's death date — remove the retired "7 June 1773" / "1742 to 1773" from Chapter 4. Contradicts the bibliography's ca. 1756 (which it says "was corrected") and Chapter 2's "eight years in the ground." Two occurrences in §4.4. Highest-priority correction for the author's decision.

2. ⚠ Gideon's service status at the wedding (ODQ-V2-02). §4.1 states "home since February"; the record leaves it open. Decide the reading; apply to Ch 3 and Ch 4 alike.

3. Witness signatures on the Bible entry. Verify the page bears the four Johnson signatures (and the bride's/mother's) before the reader-facing note treats the wedding-signing as documented; otherwise it is reconstructed.

4. The home-ceremony officiant. Entry supports a home marriage; the grandfather-led vows (no clergyman) are reconstruction. Confirm or mark.

5. Vincent's unit ("Virginia Line") and the regiment's Oct 1779 move to South Carolina. The supper tracks "the regiment" to Wilmington then South Carolina; square this with Ch 3's "Virginia line" and Ch 5's engagements (ODQ-V2-03). Internal-consistency item carried to Ch 5.

6. Christopher VI's 1742 death (ODQ-V2-06). Referenced again; carries the Ch 1 commitment flag.

End of Chapter 4 Sources Memo v1. Load-bearing result: the documented event — the 18 November 1779 Guilford home marriage of Mary Baker DeGraffenreid and Gideon Johnson Jr., entered on the Bible fly-leaf in French — is solid, and the granular morning around it is declared reconstruction. The chapter's real exposure is factual, not scenic: it prints the elder Mary Baker's death as "7 June 1773" in two places, a date the volume's own bibliography says was corrected to ca. 1756 and which Chapter 2's whole childhood frame contradicts — the Volume II counterpart of the Reuben-age error, and the top correction for the author's call. Secondary flags: Gideon's open service status stated as settled (ODQ-V2-02), and the reconstructed wedding witnesses/officiant pending a Bible-page check.

Chapter 5 — Vincent's War

Vincent's War

The brother's war, in close third: Vincent DeGraffenreid's substitute enlistment at fourteen (January 1777) and his passage through the four engagements the pension file names — Brandywine (Sept 1777), Stony Point (July 1779), Guilford Court House (March 1781), Yorktown (October 1781) — and the walk up the lane home in December 1781 to a household of women, a keeping-sister, and a niece he had never seen.

Sources memo v1 — audit trail, not for publication. Documented first; AR flagged; Gaps & Open Items at the end. Tier key: P primary · S secondary · R reference · T tradition · AR authorial reconstruction. Flags only; no corrections made. Documentary layer = the volume's embedded Working Bibliography (§V, §VI, §XI, §XIV) — above all pension file R.2845 — and Revolutionary campaign histories.

Register note: Chapter 5 is the volume's longest and most sustained battle reconstruction. Its method is the strongest kind the volume uses: a documented skeleton (a real soldier, a real enlistment, four named actions fixed by his own pension file) carrying invented flesh (the scenes, the comrades, the interior). The audit's job here is to mark clearly which is which — and to surface two items the volume's apparatus does not currently carry: a service-length discrepancy against the pension file, and, most importantly, the enslaved women the chapter names but the bibliography does not document.

The documented skeleton

Documented (pension file R.2845 + DAR + campaign histories).

• Vincent DeGraffenreid's substitute enlistment at Lunenburg, January 1777, at age fourteen — P/D (pension file R.2845; ODQ-V2-03). Consistent with Ch 2's "not yet two" in 1764 (→ b. ~Feb 1763) and Ch 4's "sixteen years and nine months" in Nov 1779.

• The four named engagements — Brandywine, 11 Sept 1777; Stony Point, 15–16 July 1779; Guilford Court House, 15 March 1781; Yorktown, 19 Oct 1781 — are the engagements named in pension file R.2845 (§V). So Vincent's presence at these four actions is documented; each chapter-section is built on a real anchor.

• The campaign facts are period history (S, §XI): Wayne's night bayonet assault on Stony Point (muskets unloaded, the silence order, Wayne's scalp wound); Greene's three-line defense at Guilford; the Yorktown surrender field, 19 Oct 1781. The chapter uses these correctly as the war "as it arrived to the soldier," and even hedges the folklore — "the men afterward said the tune had been The World Turned Upside Down, and it might have been."

• The niece on the Bible page: "Sarah Johnson, born the ninth of October 1780…" reproduced in §5.7 as an entry in Mary's hand — P (Bible; consistent with Ch 3 and with the Volume III Bridge chain). Note the specific claim that this birth entry is in Mary's hand (see item 3).

• Discharge at Williamsburg, third week of November 1781 — P/inference (the chapter's date; see item 2 on the "seven years" discrepancy).

⚠ The enslaved women — named in the prose, absent from the apparatus (the chapter's principal flag)

• §5.2 names Phillis: "Phillis was in the kitchen all the hours the family was, banking the fire and bringing the things and taking them down, owned by the place and doing the work of it… the work of the place that winter… went through Phillis's hands as much as through the mother's, and more." §5.7 names Hetty (the mother calls into the hall, "Sarah, is that Hetty come back?") and "Old Hetty's brother" who cleared a field corner.

• The handling is honest as far as it goes. The prose states the ownership plainly ("owned by the place"), does not launder the slaveholding, and credits Phillis's labor as load-bearing ("and more") rather than erasing it. That is the right instinct and consistent with the trilogy's "not her metaphor" discipline.

• But three things the audit must flag:

1. No documentary citation. The Working Bibliography has no section on enslaved people and never names Phillis or Hetty. Volume III devotes its moral spine to the enslaved people on the Rorie place, each placed against a slave schedule or probate line; Volume II names enslaved individuals with nothing behind them in the apparatus. Are Phillis and Hetty documented names — e.g., in Tscharner's 1794 will and inventory (§II) or Baker's estate — or are they authorial reconstructions of the household's real-but-unnamed enslaved labor? → The author's call: if documented, cite them (the 1794 inventory is the obvious place); if reconstructed, the reader-facing apparatus must say the names are authorial, as Vol III is scrupulous to do.

2. They appear and are not returned to. Phillis carries a paragraph in Ch 5 and Hetty a line; unlike Vol III, the volume gives no accounting of the enslaved people in their own right. This is a volume-level observation, not a Ch 5-only one (see the Ch 6 and Summary memos).

3. Apparatus gap. A volume that documents wills, deeds, order books, pensions, and DAR patriots to this depth, but whose bibliography is silent on the people its own prose says were "owned by the place," has an imbalance the author should decide how to address — at minimum a bibliography note, at most a Vol III-style accounting. This is the most important structural flag in the Vol II audit so far.

⚠ The service-length discrepancy (ODQ-V2-03)

• The chapter has Vincent serve January 1777 → discharge November 1781 = ~4 years 10 months (stated explicitly: "gone four years and ten months and three weeks"). The bibliography, however, says pension file R.2845 "references roughly seven years' Continental service" (§V), and flags the "full muster-roll trail and discharge" as open (ODQ-V2-03). → The ~4y10m the chapter commits to does not match the file's "roughly seven years." Decide which the record supports; the discrepancy is currently unreconciled. (Also carry the earlier "Virginia line" vs. North Carolina question from Ch 3/Ch 4 — confirm the pension file's stated line.)

The reconstruction (the chapter's method)

AR — the scenes and the men.

• Every battle scene is reconstruction on the documented presence: the flints-out silence and paper-on-the-hat at Stony Point, the forlorn hope, the morass at the breast, going over the works, the bayonet "used… not a thing he would set down"; Brandywine (the march to Birmingham Meeting House, Wills hit and carried back); Guilford and Yorktown as lived from the rank. None of the incident is documented; all rests on "he was at this action."

• All named comrades are invented: Doyle, Cole, Wills, Banks, Briggs, Howard (the haversack), Eppes the wagoner. Authorial texture; carry no citation.

• The "I am well" scraps and the carried letter are AR — no such writings survive; they are the "habit" the Note on Sources says the volume renders where the artifacts do not survive. The device is honest but should not be read as a surviving document.

• The homecoming (§5.7) — the walk up the lane, the mother in the hall, Mary at the desk with the quill, the sleeping niece — is reconstruction on the documented facts (Vincent's return; Sarah Johnson b. Oct 1780; Mary married Nov 1779). The "keeping is one keeping" synthesis is authorial interpretation, the volume's thesis rendered as recognition.

Gaps and open items

1. ⚠ Phillis and Hetty — document or mark as authorial; address the apparatus gap. The chapter names enslaved people "owned by the place" with no citation; the bibliography has no section on them. Check Tscharner's 1794 inventory / Baker's estate for their names. Decide whether Vol II gives the enslaved people an accounting (Vol III model) or at least a bibliography note. Highest-priority structural flag.

2. ⚠ Service length (ODQ-V2-03). Chapter commits to ~4y10m (Jan 1777–Nov 1781); pension R.2845 references "roughly seven years." Reconcile. Confirm the line (Virginia vs. North Carolina) named in the file, squaring Ch 3/Ch 4/Ch 5.

3. Sarah Johnson's Bible birth entry "in Mary's hand" (§5.7). Specific claim about the page; verify against a Bible image if available (consistent with the volume's convention, but the "which hand" detail is asserted).

4. The four engagements are documented; the scenes are not. Ensure the reader-facing note distinguishes "present at Brandywine/Stony Point/Guilford/Yorktown" (documented) from the reconstructed incident and comrades.

5. The "I am well" scraps and the carried 1778 letter are invented artifacts. No such writings survive; keep them out of any "documents" list.

End of Chapter 5 Sources Memo v1. Load-bearing result: the chapter's skeleton is genuinely documented — Vincent's 1777 substitute enlistment at fourteen and his presence at Brandywine, Stony Point, Guilford, and Yorktown all rest on pension file R.2845 — and the vivid battle scenes and comrades on it are declared reconstruction. Two items the volume's apparatus does not yet carry: the pension file's "roughly seven years" against the chapter's ~4y10m (ODQ-V2-03), and — most important — the enslaved women Phillis and Hetty, whom the prose names and calls "owned by the place" but the bibliography neither cites nor accounts for, in a volume otherwise documented to the deed and the muster roll. That imbalance is the Vol II counterpart of the moral thread Volume III makes central, and it is the author's to decide how to close.

Chapter 6 — The Keeping

The Keeping

The thirteen years of the keeping, 1785–1794: Mary Baker DeGraffenreid Johnson running her household at Rockingham County, North Carolina — the daughters, the ledger, the drovers' news and the salt-price, the country arriving as fragments — and, at the close, the death of her grandfather Tscharner (second week of February 1794) and the coming of the Christopher VI Bible south into her hands, where she becomes the first woman in the line to keep the record and writes the French death entry and, beneath it, "we continue."

Sources memo v1 — audit trail, not for publication. Documented first; AR flagged; Gaps & Open Items at the end. Tier key: P primary · S secondary · R reference · T tradition · AR authorial reconstruction. Flags only; no corrections made. Documentary layer = the volume's embedded Working Bibliography (§§I–IV, VII, VIII, X–XIV) and the Chadwell/Reynolds records.

Register note: Chapter 6 is the volume's longest (~27,000 words) and closes the book. It alternates domestic/economic reconstruction (Nancy's first birthday, the January letter, the drover and the salt-price, the lamp-reading) with Historian's Opening/Closing and interludes carrying the documented frame and the national context. It is the culmination of the volume's method — and it carries the audit's most significant cluster of findings: the retired "June 1773" grandmother-death date recurs here three times (against the chapter's own Historian's Opening), the closing scene presents the invented refrain "we continue" as an actual Bible inscription, and the enslaved household reaches its fullest named presence with no documentary citation. These are items 1–3.

The documented frame (Historian's Opening / Closing / interludes)

Documented.

• The household and its dates: Mary Baker DeGraffenreid Johnson (20 in July 1785); Gideon holding the cleared land under Sarah Vass's deed of attorney of 19 Dec 1783; the Caswell grant of 13 Nov 1779 (400 acres, Guilford → Rockingham County, created 1785, seat Wentworth); Sarah Vass administering Baker's estate since Sept 1779 — all P/R (§§III, IV).

• The daughters: Sarah (b. 9 Oct 1780), Mildred Baker (b. 11 Nov 1782), Nancy (b. 14 July 1784), Ursula (b. ~13 Dec 1787 — Chadwell Bible p. 93, §I), Polly Baker (b. Apr 1790); ten children to adulthood — P/R.

• Tscharner's death: second week of February 1794, at Lunenburg, age seventy-one, cause and exact day unrecorded — D/P (§VIII). The chapter renders it faithfully: he "went in his sleep… at his own parlor desk," found by Liza; "seventy-one years and two months." Consistent with the bibliography.

• The Bible's passage south to Mary Baker (spring 1794) — P by structural fact, not instrument (§I; ODQ-V2-06): "not by a documented instrument… but by the plain circumstance that Baker DeGraffenreid was twenty years dead with no grave, and his daughter was the one who would keep the record." The Keeping Forward interlude states this honestly.

• The national interludes (the Constitution: signed 17 Sept 1787, Hillsborough declines 184–83 in 1788, Fayetteville ratifies 21 Nov 1789; St. Clair's defeat and the FitzSimons inquiry; the French news of 1793; the Philadelphia yellow fever) are period history — S (§XI) — correctly kept as fragments arriving at the household.

• The Vol II → III handoff (Keeping Forward): the household moves to Williamson County, Tennessee (a state 1796) by 1810; Mary Baker dies there January 1823, age 58, the date disputed by six days (Chadwell 1 Jan / Reynolds 7 Jan) — P (§VIII). Sarah Johnson (→ Sarah Hubbard), Cumberland County NC (§6.7 parenthetical) writes her mother's 1823 death entry — consistent with the Volume III Bridge chain (Sarah Johnson → Hubbard → Rorie). Cross-volume anchors agree.

⚠ 1. The retired "June 1773" grandmother-death date recurs three times — against the chapter's own Historian's Opening

This is the same error flagged in the Chapter 4 memo, and Chapter 6 shows it is systematic, not a one-off — the correction reached the bibliography and some prose but not all of it. Within Chapter 6:

• Correct (Historian's Opening): the elder Mary Baker "had died in the mid-1750s, before the granddaughter who would carry her name was born."

• Wrong (§6.6 dialogue, to Ursula): buried "beside the place where your grandmother Mary Baker DeGraffenreid lies, who died in the June of 1773 a year before I was born." (Doubly broken: the date is the retired one, and "a year before I was born" is impossible — the heroine was born in 1764.)

• Wrong (§6.7): "beneath the entry he had made for his wife Mary Baker in June 1773."

• Wrong (Keeping Forward): "buried in the family ground beside the elder Mary Baker, who had gone before him in June 1773."

Combined with the two Chapter 4 occurrences, the retired date appears at least five times across the volume, while the corrected "ca. 1756 / mid-1750s" appears in Ch 2, the Ch 6 Historian's Opening, and the bibliography (which states outright the "7 June 1773" was "a false precision on a wrong year and was corrected"). → The single highest-priority correction in the Vol II audit: sweep every "1773" reference to the elder Mary Baker's death and reconcile to the record the volume itself commits to (ca. 1756). The §6.6 "a year before I was born" line is additionally self-contradictory and cannot stand as written. No correction made.

⚠ 2. The closing image presents "we continue" as an actual Bible fly-leaf inscription

• The accession scene (§6.7) has Mary write the French death entry and then, beneath a blot, "we continue" in English on the fly-leaf — and the Keeping Forward interlude, in the historian's own voice, treats this as fact: "the two English words she wrote beneath the blot were the first words in any language not French to go onto the family record's leaves at all," and "That she wrote them in the Bible only once, and only there… is the reason the page can carry the weight it carries."

• This is the trilogy's strongest presentation of the refrain as a documented inscription — not a character's thought but the narrator asserting the Bible fly-leaf bears "we continue." Per the established trilogy-wide rule (Vol III Bridge/Frame memo; Ch 3 memo): "we continue" is authorial reconstruction; there is no surviving letter, journal, or Bible page bearing it. The Ch 3 Historian's Closing itself says the two words are "not in the Bible." Ch 6's closing interlude, read straight, says the opposite. → Author's call: this is the most important consistency item for the whole trilogy's apparatus. The reader-facing Note must never cite "we continue" as documented fly-leaf text, and the Keeping Forward interlude's assertion that the Bible carries the words needs reconciling with the Ch 3 statement that it does not. (The dramatic device is legitimate as reconstruction; the danger is the historian-voice interlude reading as documentation.)

⚠ 3. The enslaved household — fullest presence, still uncited (carried from Ch 5)

• Chapter 6 names Old Hetty (at the hearth from five each morning, in nearly every scene), Young George (at the lot), Daniel ("Old Hetty's grandson," who carries letters), and Liza (at Lunenburg, who finds Tscharner dead). With Phillis and Hetty from Ch 5, the volume's enslaved cast is substantial and recurring.

• As in Ch 5: no documentary citation, no bibliography section, no accounting in their own right. These are the people whose labor runs the household the chapter is about, and the apparatus that documents the salt-price to the shilling and the pension file to the captain is silent on them. Tscharner's 1794 will and inventory (§II) is the obvious instrument that would name enslaved people held; it is cited for the estate but not mined for them. → Carried volume-level flag (see the Summary memo): document these names against the 1794 inventory / Baker's estate, or mark them authorial, and decide whether Vol II owes the enslaved people the kind of accounting Volume III makes central.

The reconstruction (the chapter's method)

AR — the domestic and economic scenes.

• §6.1 (Nancy's first birthday), §6.2 (the January letter), §6.3 (the drover and the salt-price error), §6.4 (the lamp-reading), and the §6.5–§6.7 domestic frame are reconstruction on the documented dates. The salt-price sub-plot (Sumner's April figure, Halsey's correction, the thirty-shilling cost absorbed) is invented economic texture illustrating "the household priced each fragment and proceeded." None of the incident is documented; it dramatizes the volume's thesis.

• "we continue" recurs as written in §6.2 (the 1788 letter) and §6.5 (the Oct 1793 letter) as well as §6.7 — all AR (see item 2).

• The grandfather's letters (the Cicero, the French news, "je me porte bien," "your grandfather who loves you") are invented artifacts — no such letters survive; the "habit" device again. Honest, but not documents.

• The accession scene (waking in the small hours, the blot, the two words) is the volume's closing reconstruction on the documented facts (Tscharner's death; the Bible passing to Mary). The "keeping comes up after the hand moves" motif is the volume's authorial thesis.

• Minor: the Keeping Forward interlude calls Mary's entry "the first entry… made by a woman's hand since Barbara" — but the Bridge shows Christopher, not Barbara, writing the 1714 entries (Barbara was asleep). "Since Barbara" implies Barbara wrote in the Bible; confirm the phrasing is intended, or it is a loose reference.

Gaps and open items

1. ⚠ Sweep the retired "June 1773" grandmother-death date (highest priority). Recurs in §6.6, §6.7, and the Keeping Forward interlude — against the chapter's own Historian's Opening and the bibliography's ca. 1756. The §6.6 "a year before I was born" line is also internally impossible. Combined with the two Ch 4 occurrences, five in all. Reconcile the whole volume to ca. 1756.

2. ⚠ "we continue" presented as a Bible inscription (§6.7 + Keeping Forward). The historian-voice interlude asserts the fly-leaf bears the English words; the trilogy-wide rule and Ch 3's own Closing say it does not. Reconcile so the reader-facing apparatus never cites it as documented text.

3. ⚠ The enslaved household (Old Hetty, Young George, Daniel, Liza; + Phillis, Hetty). Named throughout, uncited. Mine Tscharner's 1794 inventory / Baker's estate; decide the volume's accounting. Volume-level structural flag.

4. Vincent's homecoming and service length — internal inconsistency with Ch 5. Ch 6 Historian's Opening: "seven years' service… walked back from Yorktown to Lunenburg in the spring of 1782." Ch 5 §5.7: home first week of December 1781, ~4y10m service. The two chapters disagree on both the date and the length (Ch 6 matches the pension file's "roughly seven years"; Ch 5 does not). Reconcile (ODQ-V2-03).

5. Tscharner's death (documented) vs. the accession scene (reconstructed). The death — second week Feb 1794, age 71, cause/day unrecorded — is documented; the deathbed-by-the-desk detail (found by Liza) and the whole Rockingham accession night are reconstruction. Keep distinct in the reader-facing note.

6. "Since Barbara" (Keeping Forward). Confirm the phrasing — the Bridge has Christopher, not Barbara, making the 1714 entries.

7. Cross-volume handoff confirmed: Williamson Co. TN by 1810; Mary Baker d. Jan 1823 (Chadwell 1 Jan / Reynolds 7 Jan); Sarah Johnson → Hubbard writes the entry — agrees with the Volume III Bridge chain. No action; noted.

End of Chapter 6 Sources Memo v1. Load-bearing result: the documented frame is strong and consistent with Volume III's opening — the deed of attorney, the Caswell grant, the daughters' births, Tscharner's Feb 1794 death, the Bible's structural passage south, and the 1823 Tennessee handoff. But the volume's closing chapter carries its heaviest audit load: the retired "June 1773" grandmother-death date recurs three times here (against the chapter's own Historian's Opening), the closing interlude presents the invented refrain "we continue" as an actual Bible inscription in the historian's own voice, and the enslaved household reaches its fullest named presence (Old Hetty, Young George, Daniel, Liza) with no documentary citation. All three are the author's to resolve; nothing was corrected.

Sources Audit — Summary & Open Items

Date: 21 July 2026. Class: Consolidation of the Volume II per-chapter sources memos (Bridge + Chapters 1–6 + Epilogue) + open-items roll-up. Standing rule for this audit: the memos flag only; no prose was corrected. Every ⚠ below is the author's to decide after matching the source to the sentence himself.

Volume II had a Note on Sources and a fourteen-part Working Bibliography embedded in the manuscript, but no per-chapter AR audit (the old per-chapter memos audited a superseded pre-SolutionF draft and did not map to the finished volume). This cycle produced the missing layer — the documented-vs-reconstruction audit that Volumes I and III have — built on the finished prose section by section, using the manuscript's own section-keyed bibliography as the documented substrate.

A note on the substrate that made this tractable: the final Vol II manuscript's embedded Working Bibliography is keyed to the actual final section numbers (§1.1, §4.4, §6.6, etc.) and marks each source's tier. That is a gift — it meant the documented layer was already reconciled to the finished chapters, and the audit's job was to name the reconstruction and catch where prose outruns the record.

The memos produced (8)

No. · Memo · Documentary spine

Bridge · Charleston, February 1714 · The 1714 marriage; the Bible fly-leaf

Ch 1 · Lunenburg, Summer 1742 · Baker & Chancey wills; the 5 July 1742 marriage; the Bible

Ch 2 · The Child at Lunenburg · Birth 3 Sept 1764; the four-year silence frame

Ch 3 · The War in the House · Guilford deed 1775; administratrix 1779; Caswell grant; 18 Nov 1779 marriage

Ch 4 · The Wedding Morning · The Bible marriage entry, 18 Nov 1779

Ch 5 · Vincent's War · Pension file R.2845 (Brandywine/Stony Point/Guilford/Yorktown)

Ch 6 · The Keeping · Deed of attorney 1783; Tscharner's death Feb 1794; the Bible's passage south

Epilogue · We Continue · (author's voice; recap of the documented spine)

All saved to the project (`claude/Vol_II_Bridge_Sources_Memo_v1_0.md`, `claude/Vol_II_Ch[1–6]_Sources_Memo_v1_0.md`, and this summary).

The Epilogue — We Continue (audited here)

The Epilogue is the author's first-person present-day voice, matching the Vol I and Vol III epilogue register. It is authorial reflection (AR) built on a faithfully recapped documented spine: Mary Baker DeGraffenreid Johnson as the keeper; Baker DeGraffenreid's absent grave ("There is no grave. There is a date he was last seen alive and a date by which his wife was signing herself his administratrix, and between those two dates there is a silence about four years deep") — a scrupulous statement of ODQ-V2-01; the Caswell grant that "stood in a woman's name"; and the line's descent by daughters into the volume that follows. It invents no fact. It does, however, restate "we continue" as inscriptions Mary wrote "twice that night… and… a third time, under his death," carrying the same trilogy-wide flag: the refrain is authorial, and the reader-facing apparatus must never cite it as documented text (see item 2 below).

The overall result

The volume's discipline is real and, in its framing apparatus, exemplary. Across the Bridge, six chapters, and the Epilogue:

• The documentary spine is solid and cross-checks with Volume III. Two Chowan wills close the elder Mary Baker's parentage; the 1742 and 1779 marriages, the 1775 Guilford deed and 1779 administratrix appointment (the four-year silence), the 1779 Caswell grant, the daughters' births, Vincent's pension-file engagements, and Tscharner's Feb 1794 death are documented, and the handoff daughter Sarah Johnson (b. 9 Oct 1780 → Hubbard) matches the Volume III Bridge chain exactly.

• The reconstruction invents scene, not fact — with the flagged exceptions. The child's-eye childhood, the war interiors, the granular wedding morning, the thirteen years of keeping are close-third on documented frames.

• The Historian's Openings/Closings are model discipline. Chapters 2, 3, and 6 state the central silence (Baker's unaccounted death) as a silence, label tradition as tradition, and refuse to build across the gap.

The audit's value is the handful of places where a claim outruns — or contradicts — the record. Those are the ⚠ items.

⚠ Items requiring the author's decision (ranked)

1. The elder Mary Baker's death — a retired date still printed at least five times (the volume's Reuben-age-class error). The prose gives her death as "June 1773" in Ch 4 (§4.4, twice) and Ch 6 (§6.6, §6.7, and the Keeping Forward interlude) — while the bibliography (§VIII) says that exact "7 June 1773" "was a false precision on a wrong year and was corrected" to ca. 1756, and Ch 2 and the Ch 6 Historian's Opening use the correct mid-1750s reading. The §6.6 line — "June of 1773 a year before I was born" — is additionally impossible (the heroine was born 1764). Internal contradiction + against the corrected record + load-bearing for Ch 2's whole childhood frame. → Sweep every "1773" grandmother-death reference to ca. 1756. Top correction.

2. "we continue" is authorial — and the closing presents it as a Bible inscription. The refrain is invented (no surviving letter, journal, or Bible page bears it); Ch 3's own Historian's Closing says it is "not in the Bible." But Ch 6 §6.7 and the Keeping Forward interlude, in the historian's own voice, assert the fly-leaf carries "we continue" ("the first words in any language not French to go onto the family record's leaves"), and the Epilogue restates it as written "twice… and a third time." → Reconcile so no volume's apparatus ever cites "we continue" (or "She came" / "she held the line") as documented primary text. Trilogy-wide rule; this is its most exposed instance.

3. The enslaved household — named throughout, absent from the apparatus. Phillis, Old Hetty, Young George, Daniel, and Liza are named and depicted as "owned by the place," their labor load-bearing — but the Working Bibliography has no section on enslaved people and cites none of them, in a volume documented elsewhere to the deed and the muster roll. → Mine Tscharner's 1794 will and inventory (§II) and Baker's estate for their names; cite where documented, mark where authorial; decide whether Vol II gives the enslaved people the accounting Volume III makes central. The trilogy's "not her metaphor" discipline applies.

4. Christopher VI's October 1742 death — a commitment stated as fact (Ch 1; ODQ-V2-06). §1.5 opens "Christopher VI DeGraffenreid died at Williamsburg in October 1742" as flat fact; the bibliography flags it as the volume's commitment against the 1925 History's "ca. 1750," "no hard probate date in hand." The whole Ch 1 "accession" turns on it. → Soften to the commitment, or run the Prince George Co. / ECU (MC 0796) search.

5. Gideon Johnson's service status at the November 1779 wedding (ODQ-V2-02). Ch 3 (softly) and Ch 4 (flatly, "home since February") present his soldiering as concluded; the record leaves open whether he served in the nine-month gap or in the southern campaign after. → Decide the reading; apply to both chapters.

6. Vincent's service length and homecoming — internal contradiction between Ch 5 and Ch 6 (ODQ-V2-03). Ch 5: home first week of December 1781, ~4 years 10 months. Ch 6 Historian's Opening: "seven years' service… walked back… in the spring of 1782." The pension file (R.2845) references "roughly seven years," so Ch 6 matches it and Ch 5 does not. → Reconcile the length and the date; confirm the line named (Virginia vs. North Carolina) across Ch 3/4/5.

Lesser items to verify (per the memos)

• Ch 2: the infant sibling "the littlest" (a birth and death the audit cannot confirm) and the graveside naming scene (assumes a legible grandmother's headstone on a grave dated only ca. 1756) — verify or mark authorial.

• Ch 4: whether the Bible marriage entry actually bears the four Johnson witness signatures and the bride's/mother's; whether a minister officiated (the prose has Tscharner lead the vows).

• Bridge/Ch 1: whether Tscharner's 1742 fly-leaf marriage entry (and the reproduced French entries generally) are transcribed or reconstructed in the convention's style; the Charleston church-register entry as an independent instrument; Christopher VI's Williamsburg Lot 47.

• Ch 6: the "since Barbara" phrasing in the Keeping Forward interlude (the Bridge has Christopher, not Barbara, making the 1714 entries).

Consistency with the rest of the trilogy

• Vol II → Vol III seam confirmed. Sarah Johnson (b. 9 Oct 1780) → Hubbard, the Williamson County TN move by 1810, and Mary Baker's Jan 1823 death (Chadwell 1 Jan / Reynolds 7 Jan) all match the Volume III Bridge/Frame chain audited in that cycle.

• The three refrains rule is now anchored in all three volumes. "She came" (Bridge/Vol I Ch 6), "we continue" (Vol II Ch 3/6/Epilogue), "she held the line" (Vol III Ch 9) — all authorial; none to be cited as documented text. Volume II is where "we continue" originates and where it most nearly reads as a documented inscription (item 2).

What this is not (yet)

• Not a reader-facing Note on Sources. Volume II already has an embedded Note on Sources and Working Bibliography; these eight memos are the audit trail beneath them (Vol I/III style), naming the reconstruction the reader-facing Note does not itemize.

• Not a set of corrections. Nothing in the manuscript was changed. When the author has matched the sources to the prose and made his calls — starting with the "June 1773" sweep and the "we continue" apparatus rule — the corrections can be applied on his say-so.

End of Volume II Sources Audit Summary v1. The audit Volume II never had now exists: eight memos, the documented spine confirmed and cross-checked to Volume III, every reconstruction flagged, and six decisions plus a set of verifications gathered for the author. The discipline holds where the volume frames itself — the four-year silence stated as silence, tradition labeled tradition — and the audit's real yield is the handful of places the finished prose outran the record: the retired "June 1773" grandmother-death date printed five times against the volume's own corrected reading; the invented refrain "we continue" presented in the closing historian-voice as a Bible inscription; and the enslaved household named in full but documented nowhere in the apparatus. With Volumes I, II, and III now each carrying a per-chapter sources audit, the trilogy is documented end to end for the author's full read and review.