Apocalypsis B. Joannis

PL 114, cols 0709C–0752B · 20 chunks · 16,670 Latin words. Attributed by Migne to Walafrid Strabo; in fact Anselmus Laudunensis et schola. Translated 2026-08-28 by four Opus stints of five chunks each, none of which could see the others.

This file is the four-way MERGE. The range files (cruces-0000.md, cruces-0005.md, cruces-0010.md, cruces-0015.md) are preserved below, unedited, each under its own heading. This head is the work-wide layer: what was reconciled ACROSS the four ranges, and what a later sweep must not churn.

§0a · Apparatus, counted from the English and not summed from four reports

34 [var:] · 2 [sic:] · 2 [cj:] · 0 [d:] · 0 [ed:] · 0 [nt:] — counted from the English after the plate reads and the blind reads, not summed from the stints' own reports. Three markers the stints took were withdrawn on evidence (§0b, §0g); one silent conformation the stints missed was repaired (§0g).

verify-english.mjs 8977 → OK, 20 chunks, columns/notes/sections aligned, no duplicate paragraphs. scan-raw-markers.mjs → no marker rendered as literal bracket text. One ratio warning stands on 0000 (1.72×), reviewed by its stint clause-by-clause: Jerome's prologue is compressed synthetic Latin, and no padding was found.

§0b · ✅ THE PLATE READ — PL 114 p. 377, and why it was worth one page

Two of the range files independently flagged a single-letter difference on a real word in a lemma and correctly declined to settle it from the file. Both sites fall on one page (cols 743/744; corner numbers read before the crop, page map (column + 11) / 2 calibrated this session on p. 381 = cols 751/752). The plate answered them in opposite directions:

• 0743C — lignis was OURS. Migne prints Sicut flamma ignis. Corpus Corporum inserted an l, manufacturing "a flame of timber". There was therefore no divergence from Apc 19:12 at all, and the [var:] had asserted one on the strength of our own digitization's error. Marker withdrawn, English now reads As a flame of fire., TEI patched (data/tei-patches/8977.json), work re-chunked. The gloss corroborates from inside the work: comburit … accendit … illuminat is about fire and never mentions wood.
• 0744C — vide is MIGNE'S. He prints Et vide sedes, the imperative, where the Clementine has Et vidi sedes. The [var:] stands, confirmed — and his own gloss below re-quotes vidi three times (Ligato diabolo, vidi solutos fideles … Vidi in tempore ligationis … vidi statim in gloriam intrare). The plate contradicts its own lemma, which is exactly what this edition exists to show, and is not ours to tidy.

• 0733B — the [cj:] DISSOLVED, on a second page (p. 372). The blind reader raised it as an observation rather than a finding: the conjecture sat oddly for a defect that was purely an anchor splitting a word. The plate shows why. Migne prints infructuosos, hyphenated across a line break (in- / fructuosos), and Corpus Corporum failed to rejoin it — the mirror image of the 8946 Salomomonis case, where the same digitization rejoined a hyphenated word and duplicated the syllable. Our anchor then landed in the gap. The conjecture was right in substance (super arenam requires the privative) and wholly unnecessary. Word made whole, anchor kept at the line turn where band B falls, marker withdrawn, TEI patched. ⚑ Corroborated from the source side as 8946 was: the TEI preserves a newline on both sides of the <pb/>, which is where the printed hyphen and the line turn stand. ⭐⭐ THE GENERAL RULE: an anchor that appears to fall INSIDE a word is a REJOIN candidate before it is a conjecture candidate. Check the plate for a hyphen before reaching for [cj:].

⭐ The lesson worth carrying: these two were indistinguishable from the file — same class, same shape, flagged by the same kind of reader in the same breath. Corpus frequency and Vulgate collation could not separate them. Only the plate did. A single-letter difference on a real word in a lemma is a plate question, always, and it is cheap when the volume's page map is already calibrated.

§0c · VOCABULARY RECONCILED ACROSS THE FOUR RANGES — the merge's real work

The four ranges were diffed for the book's load-bearing vocabulary before merging. No convention drift was found, and the apparent splits are distribution by content, not disagreement: candlestick(s) only in chunks 0000–0004 (Apc 1–2, where they occur), harlot only in 0015–0019 (Apc 17–19), plagues and vials only in 0010–0014 (Apc 15–16), beast only from 0010 on (Apc 13 forward). Counted with word boundaries, after a first pass with bare substrings gave a false seat/throne split by matching seated.

One real distinction, and it is deliberate — do not flatten it. sedes is rendered "throne" where it is God's throne of judgment (the dominant sense, chs. 4–5, 20) and "seat" where it is a dwelling or a base: sunt sedes Dei → "who are the seat of God" (0718B, of the subjects in whom God rests) and sedem bestiae → "upon the seat of the beast" (Apc 16:10). English has no one word that carries both, and the gloss turns on the difference at both sites. A later sweep that normalizes sedes to a single English word will destroy an argument Migne is making.

§0d · STANDING CHECKS — the negatives, recorded so they count

• deservit / deserit: CHECKED ZERO across all twenty chunks, verified mechanically over the whole Latin work, not inherited from the stints' individual reports. Neither spelling, nor any deservi- / deseru- form, occurs anywhere in 0709C–0752B. This pair has fired in both directions elsewhere in the series, so the zero is worth having on record.
• No polarity defect was found by any of the four stints, in either direction. The negation work that did land was hosting, not presence — see §0e.
• Migne's own ( sic ) does not occur in this work; the two [sic:] are ours, both on non-words, both with the corrupt word Englished beside the marker (7a⁗).

§0e · WHAT A BLIND READER SHOULD GO TO FIRST

Listed here because the range files each named their own, and a merged list is what the step-4a reader actually needs. This section is NOT to be given to that reader — the runbook forbids showing the cruces before findings are formed; it is for whoever adjudicates afterwards.

1. 0717C — the lemma counsels buying tried gold; the gloss says talia opera non perducent ad salutem, contradicting its own lemma. Rendered as printed with the negation intact. Likeliest explanation is a gloss slipped from Apc 3:17, but relocating it would be rearranging Migne's page.
2. 0733B — [cj: fructuosos; read infructuosos]. A negative hidden inside a column split (in [0733B] fructuosos); super arenam requires the privative. The class to watch generally: any anchor falling inside a word.
3. 0745A — [cj: augentur; read anguntur].
4. 0719B — [cj: Augusti; read Augustinus], settled by the corpus (this work prints Augustinus at chunk 0019 and Augusti nowhere else).
5. 0739D — In die ortus sui where Ez 16:4 has tui, and Migne's own next lemma keeps the second person four words later.

§0f · PLATE-READ CANDIDATES STILL OPEN

Carried as printed, no marker, awaiting a plate that has not been bought. All are in PL 114 and the page map is calibrated, so each costs one render.

⚑ 0751B is already answered and needs no render — p. 381 was read this session for the plate spot check, and Migne indeed sets no terminal mark there. Recorded so it is not bought twice.

Preserved unedited from cruces-0000.md.

8977 · Glossa ordinaria, Apocalypsis B. Joannis — cruces, chunks 0000–0004

PL 114, cols 0709C–0719D (Prologue + Argument + CAPUT I–IV). Range file: this covers chunks 0000–0004 only. cruces.md is the four-way merge and is written once at the end; nothing here is to be duplicated into it by hand before that merge.

Attributed by Migne to Walafrid Strabo; in fact Anselmus Laudunensis et schola. Conventions followed: src/english/8968/cruces.md §1 (Ruth, the Glossa pilot) and GLOSSA-NT-AGENT-BRIEF.md.

Apparatus taken in this range: 18 [var:] · 1 [cj:] · 0 [sic:] · 0 [d:] · 0 [ed:] · 0 [nt:]. Marker parity against the Latin twins: column anchors 3/14/7/8/8, [n:] notes 0/1/4/7/2, italic spans 0/78/39/50/42, question marks 0/0/2/0/0 — all exact. Punctuation compared band by band (colons, semicolons, ?, !, guillemets) across the verified anchor partition: zero mismatched bands after four supplied semicolons were removed (see §4). Zero guillemets in either language.

§1 · THE 7a″ DELIVERABLE — every lemma in cols 0709C–0719D, collated

Collated against sources/vulgate/clementine-flat.txt, cited from sources/vulgate/clementine/Apc.lat etc. This section lists what was checked, not only what was found — the agreements are the evidence that the divergences were looked for.

1a. Divergences that took a [var:] — 18, one per row

Distribution: 5 in chunk 0001, 4 in 0002, 4 in 0003, 5 in 0004.

1b. Divergences examined and DECLINED, with the reason — the marker would have been noise

• 0709C Joannes septem Ecclesiis quae sunt in Asia. Quam dedit. — one italic span joining Apc 1:4's opening to Apc 1:1's Quam dedit. Both agree with the Clementine verbatim; the join is the gloss's own citation practice, not a textual variant. No marker; the span is kept as one span in the English, as Migne sets it.
• 0709D (prologue) Ego sum Alpha et Omega, initium et finis — agrees verbatim with Apc 21:6. It is not Apc 1:8 (principium et finis), and reading it as a divergent 1:8 would have been the error. Checked clean, no marker. Note it is not italicized in the Latin and is therefore not italicized here.
• 0712A Alpha, principium. / Omega, finis. — Apc 1:8 reads alpha et omega, principium et finis. Migne re-pairs the four words into two glossable couples. A lemma-splitting device, not a variant reading.
• 0711D Videbit omnis terra talem. — Migne himself labels it Alia translatio habet. His own apparatus has already adjudicated it; a [var:] would restate what the page says.
• 0713B aurichalco Libani — likewise introduced In alia translatione habetur. Declined for the same reason. Not italicized in the Latin; not italicized here.
• 0713C Et ex ore. vs Apc 1:16 et de ore. Preposition only; nothing in the gloss turns on it. Cruces, no marker.
• 0716A Dedistis mihi manducare et bibere vs Mt 25:35 dedistis mihi manducare : sitivi, et dedistis mihi bibere. A condensation, not a divergent wording: every word Migne prints stands in the verse in that order. Pattern 14 asks for a sense-changing divergence and there is none. ⚠ Note the 7a″ trap here and that it was avoided: manducare is rendered "to eat" because that is what manducare means, not by the route that produced the 7914 manum → "to eat" defect.
• 0716A charitatem autem non habeam, nihil sum vs 1Cor 13:2 caritatem autem non habuero. Present subjunctive for future perfect; charitatem/caritatem is orthographic. Not material. ⚠ But see §3 on the siglum, which is a separate finding.
• 0717D Ecce ego sto ad ostium et pulso vs Apc 3:20 Ecce sto ad ostium. An added emphatic pronoun; the gloss (Exspecto, vel inspiro) does not use it. Cruces, no marker.
• 0714A Et angelo. at Apc 2:1, where the Clementine has bare Angelo Ephesi ecclesiae scribe. Migne's Et matches his own 2:8 and 2:12. Cruces, no marker.
• 0715A Et dabo ei. vs Apc 2:17 et dabo illi. Pronoun only.
• 0715A Abscisus est lapis sine manibus vs Dn 2:34 abscissus est lapis de monte sine manibus (and Dn 2:45 in the other order). de monte dropped, Abscisus with one s. The gloss argues only from sine manibus, so nothing turns on the omission. Cruces, no marker. ⚠ Migne's own siglum reads (Dan. I) for what is Daniel 2 — passed through verbatim, per convention 4; it is not ours to correct.
• Orthographic only, all checked and all declined: tanquam for tamquam (five sites, 0713A–0713B), coelo for caelo (0716D, 0718A), charitatem for caritatem (0714C, 0716A), Pathmos for Patmos (0709C, 0712B — Migne is consistent in both places), aurichalcum for auricalco (0713B), chrystallus for crystallo (0718D).

1c. Lemmata checked and found to agree with the Clementine EXACTLY

Apc 1: Apocalypsis Jesu Christi · Quae oportet fieri · Cito · Qui testimonium perhibuit · Verbo Dei · Beatus · Qui legit · Servat ea · Septem ecclesiis · septem · In Asia · Gratia vobis · Qui est et qui erat · Throni · Testis fidelis · Primogenitus mortuorum · Princeps regum terrae · et sacerdotes · Ecce venit · Cum nubibus · Et videbit eum · Omnis oculus · Plangent · plangent se · Ego sum · Ego Joannes (×2) · Frater vester · in Christo · Propter verbum Dei · In Dominica die · Audivi post me vocem · Quod vides, scribe · Epheso · Et conversus sum · Septem candelabra · Aurea · Similem Filio hominis · Podere · Praecinctum ad mamillas · ad mamillas · Capilli erant candidi · Nix · Oculi · Et pedes · In camino ardenti · Et vox illius tanquam vox aquarum multarum · Stellas · Sicut sol lucet in virtute · Cecidi · Tanquam mortuus · Dexteram suam · Noli timere · Sum vivens · Et habeo · Mortis et inferni · Quae oportet fieri post haec · Sacramentum.

Apc 2: Ephesi Ecclesiae · Septem · Non potes sustinere (×2) · Tentasti eos qui se · Sustinuisti propter nomen meum · Et non defecisti · Sed habeo adversum te quod charitatem · Movebo candelabrum · Facta Nicolaitarum · Qui habet aurem · Smyrnae · quasi myrrha · Haec dicit primus (×2) · Aliquos ex vobis · Ut tentemini · Diebus decem · A morte secunda · Rhomphaeam · Non negasti fidem meam · Balaam · Doctrinam Nicolaitarum · Manna · Et in calculo · Scriptum · Novi opera tua · Fornicari · Non mittam.

Apc 3: Et angelo Ecclesiae Sardis · Scio opera tua · Esto vigilans · Non enim invenio opera tua plena · In mente ergo habe · Audieris · Non inquinaverunt vestimenta sua · Non delebo · Induravit Dominus cor Pharaonis (Ex 10:20, less its opening Et) · De libro vitae · Confitebor nomen ejus · Venite, benedicti Patris mei (Mt 25:34) · Haec dicit · Claudit · aperit · Ecce dedi coram te · Ostium · Judaeos · Hunc nescimus (Jo 9:29, a fragment of hunc autem nescimus unde sit) · Mentiuntur · Ecce faciam illos ut veniant · Scient · Servasti verbum patientiae meae · Ab hora tentationis · Tentare habitantes in terra · Ecce venio cito · Columnam in templo · Non egredietur · Novae · De coelo · Testis · Et verus · Neque frigidus es, neque calidus · Utinam frigidus · Incipiam te evomere · Ex ore meo · Dives sum · Suadeo tibi emere a me aurum ignitum probatum · Non appareat confusio · Collyrio · Sedere mecum in throno meo · Sicut et ego vici et sedi cum Patre meo in throno ejus.

Apc 4: Post haec vidi · In coelo · Tanquam tubae loquentis mecum · Ostendam tibi · Statim fui · Et ecce sedes · Sedens · Qui sedebat · Similis erat aspectui · Jaspidis et sardinis · Smaragdinae · Viginti quatuor seniores · Coronae · De throno · Septem spiritus · In conspectu · Mare vitreum · In medio · Sedis · Quatuor · Animalia · Plena oculis · Leoni · Et quatuor animalia singula eorum habebant alas senas · Alas senas · In circuitu et intus · Plena sunt oculis · et in circuitu · Qui venturus est · Et benedictionem · Et adorabant · Dicentes · Creasti omnia · Propter voluntatem tuam · Erant · Creata sunt.

Not scripture, so not collated (Hebrew-name etymologies and the gloss's own re-cuts): voluntas mea · lapsus profundus · consilium (0712D, 0714B) · voluntas (0714B) · canticum (0714D) · principium pulchritudinis (0715D) · plena sunt oculis intus (0719B, Migne's own re-division of Apc 4:8 for the gloss, not a variant).

§2 · THE ONE CONJECTURE — 0719B, Augusti

Migne prints, of the sixth wing of the four living creatures:

sexta, quorumcunque decreta, ut Augusti, Gregorii, Ambrosii, per quos Ecclesia alta petit

Augusti is a perfectly good Latin word — the genitive of Augustus — so nothing is broken and Pattern 12 has nothing to wrap. But the faithful English asserts something false: it makes the decrees of the emperor Augustus a wing of the Church, standing between Gregory and Ambrose in a list of doctors, and it is decreta of the Fathers a quibus praedicatores accipiunt totius praedicationis fundamentum. Pattern 18 exactly.

⭐ The corpus is its own control and it is decisive here: this same work prints Augustinus at chunk 0019, and Augusti nowhere else. One occurrence of the truncated form against a full one in the same plate, in the same company of Fathers.

Marker taken, additive, Migne's word kept and rendered: of Augustus [cj: Augusti; read Augustini, "of Augustine"]

The rival considered and rejected: that Migne means the civil decreta Augusti — Roman imperial law as a source of ecclesiastical discipline, which is not absurd in itself. It loses on the sentence's own structure: the three genitives are a single series governed by per quos Ecclesia alta petit and a quibus praedicatores accipiunt … fundamentum, and neither predicate is sayable of imperial legislation in a list whose other two members are Gregory and Ambrose.

§3 · CRUCES — logged, no marker fired

Each of these was examined and left as printed under Pattern 7. A crux asserts nothing about the plate; where I could not settle a thing from the file, the crux is where it stays.

3a. Referents supplied, and what decided them

• 0709C (prologue), commendasset / asciverat / tradidisset. Three unexpressed subjects in one sentence, and they are not the subject of the verb before them. recumberet is John; commendasset, asciverat and tradidisset are all Christ — matrem propriam and virginem are his mother, and ad crucem astanti soli is John in the dative. Rendered so. This is the class the brief names as dominant; naming it here is the record that it was decided rather than defaulted.
• 0709C, ab eodem: the Apocalypse praeostensa by Christ, not by John — a Christo electus atque dilectus is the sentence's own antecedent for idem.
• 0714A, requirit: Id est episcopo scribit, de manu cujus peccata subditorum requirit, et sine cujus consensu … non praesumit ipse Joannes. Two readings both construe: the subject of requirit is God/Christ (who requires the sins at the bishop's hand), or John. ⚠ The explicit ipse Joannes in the second clause is evidence for the first — Migne names John only where the subject changes. The English keeps a bare "he", which is exactly as under-determined as the Latin, so nothing is decided in the text; this entry is the record that the ambiguity is real and was not resolved silently.
• 0714B, (qui lapsi fuerant): the parenthesis attaches to the implicit subject of corrigantur, i.e. those to be corrected, not to quorum operibus Deus delectatur in the clause before. Rendered "that by the example of these they (who had fallen) may be corrected", keeping Migne's parenthesis in his own position.
• 0718B, quos hortatus est ad pugnam: a plural relative with no plural antecedent in its sentence (de quo singular precedes it). Rendered literally as "whom he has exhorted to the fight". A dative quibus would construe; that is a conjecture and it stays here. ⛔ Not a [cj:]: Pattern 18's bar is that the English asserts something false, and this English asserts nothing at all — it is incoherent, and 18a's exception covers only a negation or a correlative gone wrong. This one is neither.

3b. Grammar the plate prints and English cannot mend

• 0714A, ministri ejus sunt infernus: plural subject, singular predicate nominative. Rendered as printed: "his ministers are hell, in whom he has a place." No English exponent for the mismatch, and no emendation to make.
• 0714D, etiam canticum in quibus delectatur: singular canticum, plural in quibus. Rendered "a song also, in which things he delights", carrying the plural.
• 0712C, quae monuit communiter mittere ecclesiae Asiae … per quas: ecclesiae is singular by form and per quas wants a plural (a dative plural would be ecclesiis). ⭐ English "by which" carries no number at all, so the rendering reproduces the Latin's own indeterminacy without choosing. Logged rather than fixed.
• 0718B, In quibus sedet Deus et judicat, et etiam subditos: sedes at Apc 4:2 is singular; In quibus is plural. English "In which" is number-neutral. subditos is an accusative object gapped off judicat; the English repeats the verb ("and judges also"), which is ordinary English gapping and adds no claim.
• 0712B–C, in fervorem fidei / in perditionem Sodomae: in + accusative where sense wants time or occasion. Rendered "unto the fervour of faith" / "unto the destruction of Sodom", literally, rather than smoothed to "at".

3c. Sentences that do not construe, or construe against their own neighbour

• 0711A, Christus dedit haec mihi, beatus autem quia me acceperit. Elliptical: no subject for beatus, and a perfect subjunctive with no governing verb. Rendered as printed, "but blessed because he has received me." No conjecture offered; there are at least two (beatus autem [est] qui…, or a first-person sum) and nothing in the column chooses between them.
• 0711A, quia nec longa mora laboris, et praemia vicina. A nec … et pair where sense wants nec … et to be neque … sed or a bare asyndeton. The nec is carried: "because neither is there a long delay of labour, and the rewards are near." It reads oddly. That is the point of Pattern 7a — I have not converted nec to "there is no" plus a smoothed second limb, which would hide the plate's own awkwardness.
• 0717C, Suadeo tibi emere a me aurum ignitum probatum. Quia talia opera non perducent ad salutem. The gloss contradicts its own lemma: the lemma counsels buying tried gold and the gloss says such works do not lead to salvation. Rendered as printed, negation intact. ⚠ The likeliest explanation is that the gloss belongs to a neighbouring lemma (Apc 3:17's dives sum / nudus) and has slipped; but that is an editorial rearrangement of Migne's page and Pattern 7 forbids it in the text. Flagged here for a blind reader — this is the most conspicuous sense-defect in the range.
• 0712C, Salomon, non reservaturus, sapientiam nocte suscepit. A bare future participle with no object. Rendered "Solomon, not about to keep it, received wisdom by night", the "it" supplied from sapientiam in the same clause and flagged here as supplied.
• 0717A, quidam Eliae monitis legem spiritualiter impleturi, hostem credunt victuri. Two future participles, the second in an accusative-and-infinitive with the copula elided. Rendered "certain, being about to fulfil the law spiritually by the warnings of Elias, believe that they will vanquish the enemy." The construction is elliptical in the plate and is left elliptical.
• 0718C, Per ruborem ejusdem — a redness predicated of an emerald, two clauses after its green is expounded. Real words, coherent syntax, odd content. Rendered as printed. ⚠ Raised under brief §4 only as a negative: nothing here reads suspiciously tidy, and no plate read is requested.

3d. Migne's own sigla, passed through and NOT corrected

Convention 4 governs: the sigla ride verbatim and passing one through is not endorsing it. Three in this range are demonstrably wrong and are recorded rather than repaired.

• 0715A [n: (Dan. I)] on a quotation of Daniel 2 (2:34 / 2:45).
• 0716B [n: (Jac. II)] stands after charitatem autem non habeam, nihil sum, which is 1 Cor 13:2. The James 2:10 matter is the preceding quotation in the same sentence, so the siglum has either slipped one clause or is meant to cover both.
• 0716B [n: (Joan. XIV)] on a quotation whose second half is Jo 16:23 — the conflation already marked [var:] at that column.

3e. Words examined for the digitization-alteration class (brief §4), and the finding

• 0712D, Ephesas. Migne prints Epheso. as the lemma, then glosses Ephesas voluntas mea interpretatur. ⭐ The work's own control refutes any tidying instinct: at 0714B the same etymology is introduced as Ephesus secundum bene persistentium partem — so the book spells the name Ephesus two columns away and Ephesas only here. Rendered as printed ("Ephesas is interpreted…"), because Pattern 7 governs a proper-name form exactly as it governs any other. Not a [sic:]: Ephesas is a well-formed token, not a non-word, and Addendum A is explicit that a real, well-formed word is Migne's until a plate says otherwise. A plate read of col. 0712D would settle it and is cheap.
• ⭐ Recorded negatives. deservit / deserit, the standing confusion pair for this series: grepped across the whole of 8977 and it occurs zero times, in either spelling. Nothing to check. Likewise no split or fused runs (Pattern 10) anywhere in cols 0709C–0719D, no ( sic ) of Migne's own, no dittography, and no non-word of any kind — which is why this range fires no [sic:] at all. A checked zero is worth having; the alternative is a reader who cannot tell this range from one nobody read.

§4 · SELF-AUDIT — the tests the brief requires, and what each returned

• Test 1 (a [sic:] must not swallow the word it marks). Vacuously clean: no [sic:] in this range. The [cj:] at 0719B is additive by construction — "of Augustus" stands in the running text and the bracket is beside it, so stripping every marker leaves a readable sentence. Stripped and read aloud; no stranded article or preposition anywhere in the five chunks.
• Test 2 / 2a (name the host of each negative). All 84 printed negatives in chunks 0001–4 walked one by one, each mapped to the single Latin word it attaches to. Five sat in the trigger population (a negative on a verb with a quantifier or correlative adjacent) and each was proved from the sentence's structure, not from what read well:
• 0711D quia non tantum dolebunt de ipso tormento, quam quod repellentur — host is tantum, proved by the quam correlative. English: "shall not grieve so much over the torment itself, as that…"
• 0712A Ne a me hoc dictum, sed ab ipso Christo credatis — host is a me, proved by sed ab ipso Christo, not credatis. English negates the agent, not the believing.
• 0714C Hoc non fecisti pro humana laude, sed propter nomen meum — host is the prepositional phrase, proved by sed.
• 0715C non enim prius est charitas quam fides — host is prius, proved by quam.
• 0717B quia neque omnino fidem ignoras, neque aperte negas acceptam — the two adverbs omnino ‖ aperte are the pivot, so the scope is "not wholly ignorant" and "not openly denying", which is what makes the man lukewarm rather than either cold or hot. Reading the first as "thou art not ignorant of the faith at all" would collapse the contrast the whole gloss is built on.
• 0717B Non optat eum frigidum simpliciter, sed talem de quo major spes habeatur — host is frigidum simpliciter, proved by sed talem.
• 0718B Ubi tantum boni sunt, non est opus judicare — ⚠ here tantum is in the protasis and is not the host; the negative is on est opus, and the ubi … ubi pivot proves it. Recorded because it is the same shape as the trigger cases and resolves the other way.
• Test 2b (anchor POSITION, not presence). All 40 anchors in the range compared against the word they stand beside in the Latin, not merely counted. All 40 sit at the same point in the construction. Four stand inside an italic lemma and were kept inside it: Ut [0715A] tentemini., …qui loquitur [0713D] in vobis…, Si videris mulierem ad concupiscendam [0715D] eam…, Ecce ego sto ad ostium et pulso. Si [0717D] quis audierit,. Four stand hard against a VERS. address with no space ([0714D]VERS. 7.--, [0716A]VERS. 2.--, [0717B]VERS. 14.--, [0718B]VERS. 2.--, [0718D]VERS. 6.--) and are reproduced that way. One was moved during the audit: 0719A, first set after "gentle" and re-set immediately after "are", which is the word sunt it stands before in the Latin.
• Test 3 (punctuation BY BAND, never by total). Built over the verified anchor partition, markers stripped from the English first. It returned four supplied semicolons where Migne prints a comma — 0711A (Asia est elatio, Ecclesia ergo), 0711B (nemo mala intellexerat, ponit), 0712D (consummatur, et haec), 0714D (delectatur, myrrha) — all four repaired to commas. ⚠ Reported honestly: in this range the totals did NOT cancel — chunk 0001 stood at 16 semicolons against Migne's 13 and chunk 0002 at 3 against 2, so a total-only check would have caught these four too. The band test is still what located them: it named the four columns instantly instead of leaving four sites to be found by re-reading 1,563 words. Final state: zero mismatched bands on : ; ? ! « ». ⛔ No comma→semicolon sweep was attempted in the other direction: this book prints semicolons freely (27 in the range), so the Tobit exception does not apply here.
• Test 5 (does the English ASSERT what he asserts). Run last, counts already clean, each paragraph read back against its Latin. It produced §3c's list — the four sentences that do not construe are recorded there rather than repaired, and 0717C is named as the site a blind reader should go to first.

§5 · CONVENTION CALLS MADE IN THIS RANGE

• Fragmentary lemmata expanded per Ruth's convention 3 at two places, and only two: Tentasti eos qui se. → "Thou hast tried them who say themselves to be apostles" and Sed habeo adversum te quod charitatem. → "But I have against thee, that thou hast left thy charity." Both stumps end mid-clause and are not English. ⚠ Note what was not done: neither expansion imports a word where Migne diverges from the Clementine, and Sed habeo adversum te quod charitatem. deliberately does not carry the Clementine's primam ("thy first charity"), because Migne's fragment stops before it. ⛔ Complete one-word lemmata (Nix. · Oculi. · Manna. · Judaeos.) are a different class and are rendered as they stand; the two are not contrasted here.
• Et videbit eum. (0711D) rendered "And shall see him." — a verb phrase with its subject deliberately withheld, because Migne makes Omnis oculus the next lemma and glosses it separately. Expanding to "And every eye shall see him" would have swallowed the following lemma and is precisely the conformation 7a″ forbids. Logged because the English is visibly incomplete and that incompleteness is Migne's.
• Non possumus quae vidimus non loqui (0716C) → "We cannot but speak the things which we have seen." English "cannot but" is the exponent of the Latin double negative, not the loss of one; both negatives are present in the idiom. Recorded because a negation-counting check will see only one.
• Pattern 17. The addressee of CAPUT II–III is the angel of each church, singular, and takes thou throughout; the plural passages (Apc 2:10 Aliquos ex vobis … ut tentemini, and the Apostle's vos at 0712B, 0713C, 0714A) take you. The distinction is Migne's morphology, not a register choice.
• VERS. n.-- passed through verbatim everywhere, per convention 1. CAPUT PRIMUM. → CHAPTER ONE., CAPUT II./III./IV. → CHAPTER II./III./IV., matching Ruth's shipped English. PROLOGUS HIERONYMI. → THE PROLOGUE OF JEROME.; ARGUMENTUM. → THE ARGUMENT., after Ruth.
• Pattern 4 [f:] NOT applied — this is a lemma-and-gloss commentary, which the runbook's false-positive 4 excludes by name.
• Column-band gaps not reported. 0709C → 0709D → 0710C skips 0710A–B, and 0709C reappears as the colContext of three later chunks in the lemma brief. Runbook false-positive 1: the A–D marks are positional quarter-guides, not four per column. Named here only so a later reader can see it was recognized, not missed.
• Ratio warn on chunk 0000 (1.72×). verify-english.mjs flags it. The chunk is 188 words of Jerome's prologue — ablative absolutes, gerundives and a triple ut-chain — which unpack long in English. Checked for padding clause by clause and none found; nothing is summarized and nothing is doubled. Recorded rather than trimmed, because trimming would cost content.

Preserved unedited from cruces-0005.md.

8977 · Glossa ordinaria, Apocalypsis B. Joannis (PL 114) — cruces, chunks 0005–0009

Range: CAPUT V–IX, cols 0719D–0729A (colFirst 0720A → colLast 0729A). One translator, five chunks (3,640 Latin words). verify-english.mjs clean.

Apparatus taken in this range: 2 [sic:] · 5 [var:] · 0 [cj:] · 0 [d:] · 0 [ed:] · 0 [nt:] · 2 [n:] (parity with the Latin twins). Anchors 37/37, verbatim, in order, and each checked against the WORD it stands beside in the Latin (test 2b). Question marks 0/0 in both columns. Punctuation compared band by band (test 3): zero mismatched bands for colons, semicolons, question marks, exclamation marks and guillemets, after the corrections recorded in §4. Zero guillemets in either column. ## heads 1:1 (CAPUT V. → CHAPTER V., and so through IX, matching Ruth's shipped English).

⚠ Nothing in this file has been checked against Migne's plate. Every candidate below that would need the plate is named as such, and none of them fired a marker.

§1 · THE 7a″ DELIVERABLE — every lemma in the range, collated against the Clementine

Collated against sources/vulgate/clementine-flat.txt; readings cited from sources/vulgate/clementine/Apc.lat etc. The range glosses Apc 5:1–9:19.

1a. Divergences that took a [var: …] (4 of the 5; the fifth is in §1b)

0721D (chunk 0006), Apc 6:4 — Migne's lemma prints Sumeret pacem de terra, et ut indicem, etc.; Clementine Apc 6:4: ut sumeret pacem de terra, et ut invicem se interficiant. indicem ("an informer") is a real, well-formed accusative and takes no [sic:], but it leaves the clause with no verb and it is not the verse. Rendered as printed ("and that an informer, etc.") with [var: Vulg. ut invicem se interficiant, "that they should kill one another"]. ⚑ Plate-read candidate: indicem / invicem is a plausible n→d misreading, and brief §4 warns the digitization alters spelling in both directions. The marker claims only what the Clementine reads, which is checked; it does not claim the plate.

0722B (chunk 0006), Apc 6:6 — Migne's lemma prints Et vivum et oleum ne laeseritis.; Clementine Apc 6:6: et vinum, et oleum ne laeseris. Two divergences in five words: vivum ("the living thing") for vinum ("wine"), and a plural imperative laeseritis for the singular laeseris. Rendered as printed with [var: Vulg. et vinum, et oleum ne laeseris, "and the wine and the oil hurt thou not"]. ⚑ This is the site where the lemma and its own gloss disagree, which is why it matters: the gloss immediately under it expounds vino, id est sanguine meo ("with wine, that is, with my blood"), and the same paragraph has already given Vinum, praelo quidem attriti… and Vinum similiter emitur denario uno. The gloss reads vinum three times; only the lemma reads vivum. Both are rendered as printed — "the living thing" in the lemma, "wine" in the gloss — and the contradiction stands on the page, per 7a. Pattern 17 governs the number: laeseritis is plural, so the English imperative addresses a plural ("do not hurt"). ⚑ Plate-read candidate (vivum / vinum, again a real word, so no marker beyond the [var:]).

0722D (chunk 0006), Apc 6:10 — Migne's lemma prints Judicas et non vindicas sanguinem.; Clementine Apc 6:10: non judicas, et non vindicas sanguinem nostrum. Migne's lemma lacks the first non. Rendered as printed — "Thou judgest and dost not avenge the blood" — with [var: Vulg. non judicas, et non vindicas sanguinem nostrum, "thou dost not judge, and dost not avenge our blood"]. Pattern 17: judicas is singular, so thou. ⚑ Note that the gloss under it, Faciendo discretionem bonorum et malorum ("By making a distinction between the good and the evil"), construes more easily with the affirmative lemma Migne prints than with the Clementine's negative one — so this is not obviously a compositor's slip, and no conjecture is offered. The negative that Migne does print (non vindicas) is carried; no negative is supplied that he does not print.

0724B (chunk 0007), Phil 1:23 — the gloss on Issachar quotes Cupio dissolvi, et esse cum Christo; Clementine Phlp 1:23: desiderium habens dissolvi, et esse cum Christo. [var: Vulg. desiderium habens dissolvi, et esse cum Christo, "having a desire to be dissolved, and to be with Christ"]. The received-text form is the familiar patristic citation form, not the Clementine's; it is recorded, not explained (Pattern 14 forbids speculating why).

1b. The remaining divergences — one marker taken, the rest declined with the reason

• 0723A (0006), Exite de Babylone, populus meus, et immundum ne tetigeritis. This is a conflation and is not any single Clementine verse: the nearest are Apc 18:4 (Exite de illa populus meus), Jr 51:45 (Egredimini de medio ejus, populus meus), and 2 Cor 6:17, whose closing clause it matches; "Babylon" is in none of them at that point. A [var:] was taken here, worded so it names the witness and does not repeat Migne's own words (the verifier's mirror test): it seemed worse to leave a reader believing he was meeting a verse. Recorded here as the one marker in the range whose content is a composite statement rather than a single received reading.
• 0720B (0005), Luke 10:24 — Migne prints Multi reges et prophetae voluerunt videre; Clementine Lc 10:24 has multi prophetae et reges voluerunt videre. Order only, sense unchanged; Pattern 14 fires on divergence "in a way that changes the sense", so declined. Logged because a blind reader will see it.
• 0725C (0008), Apc 8:3 — Migne's lemma Super aureum altare.; Clementine super altare aureum. Word order only; the gloss turns on neither. Declined, logged.
• 0728D (0009), Apc 9:17 — Migne's lemma Procedebat ignis (imperfect); Clementine de ore eorum procedit ignis (present). Rendered as printed ("Fire went forth"). The gloss is three words long (Causa aeternae poenae) and depends on no tense. Declined as immaterial, logged. ⚠ Note that 7914 @1071D treats a tense change as material when the English restores the Vulgate tense; nothing is restored here — Migne's tense is what is rendered.
• Orthography, none material, none marked: tanquam for tamquam throughout (0720C, 0723A, 0727B, 0728D) is this corpus's normal orthography, not a divergence; lacrymam for lacrimam (0725A); Haebraei for Hebraei (0720D); Hebraice lower-cased in the Clementine, capitalized by Migne (0727C).

1c. Lemmata checked and found to agree with the Clementine EXACTLY

Recorded in full, because a findings list alone cannot be distinguished from a report by an agent that never looked. Each of the following is verbatim at the verse it glosses (allowing only the Clementine's own comma-spacing and the tanquam/tamquam orthography noted above):

Apc 5 (chunk 0005): Et vidi in dextera sedentis · Librum · Foris · Intus · Signatum · Sigillis · Septem (all six from 5:1, the last five being the single words librum … intus et foris, signatum sigillis septem glossed one by one) · Angelum · Praedicantem · Aperire librum (5:2) · Flebam multum (5:4) · Et unus de senioribus · Vicit leo (5:5) · Et ecce in medio throni · Tanquam occisum · Cornua septem (5:6) · Ceciderunt coram agno · Habentes singuli citharas (5:8) · Cantabant canticum · Tribu (5:9) · Vidi · audivi · Vocem angelorum · Multorum in circuitu throni et animalium et seniorum · Millia millium (5:11) · Sapientiam, et fortitudinem (5:12) · Benedictio · et honor · et gloria · et potestas (5:13) · Et quatuor animalia · Ceciderunt in facies (5:14).

Apc 6 (chunk 0006): Et vidi · Quod aperuisset · Veni et vide (6:1) · Ecce equus albus · Arcum (6:2) · Et cum aperuisset sigillum secundum (6:3) · Equus rufus · Et datus est ei gladius magnus (6:4) · Et ecce equus niger · Stateram (6:5) · Et audivi tanquam · Bilibris · tritici (6:6) · Ecce equus pallidus · Nomen illi mors · Et infernus · Super quatuor (6:8) · Et cum · Vidi subtus altare · sub · Altare · Interfectorum propter verbum (6:9) · Usquequo (6:10) · Conservi eorum (6:11) · Sol factus est niger · Tanquam saccus (6:12) · Grossos suos (6:13) · Coelum (6:14) · Reges terrae · Divites · Fortes (6:15) · Et dicunt montibus (6:16).

Apc 7 (chunk 0007): Post haec · Quatuor angelos · Angelos · Tenentes · Ventos · Super mare (7:1) · Et vidi alterum · Ascendentem · Signum (7:2) · Centum quadraginta · Ex omni tribu (7:4) · Duodecim (7:5) · Simeon · Levi · Issachar (7:7) · Zabulon · Joseph · Benjamin (7:8) · Post haec · Stantes (7:9) · Salus Deo nostro qui sedet (7:10) · Et omnes (7:11) · Amen · benedictio · claritas (7:12) · Et respondit unus (7:13) · Stolas suas (7:14) · Non esurient, neque sitient (7:16) · Deducet eos ad vitae · Absterget Deus omnem lacrymam (7:17). Also the one full proof-text in the chunk: Rom 10:10, Corde enim creditur ad justitiam, ore autem confessio fit ad salutem — verbatim in the Clementine, punctuation apart.

Apc 8 (chunk 0008): Et cum aperuisset (8:1) · Septem angelos · Et datae sunt (8:2) · Et alius angelus · Stetit ante altare · Habens · Thuribulum · De orationibus (8:3) · Fumus (8:4) · Praeparaverunt (8:6) · Et primus angelus · Et facta est grando · In terram · Arborum (8:7) · Et secundus angelus · In mare (8:8) · Mortua est tertia · Quae habebant animas (8:9) · Et tertius angelus · Fontes (8:10) · Et quartus angelus (8:12) · Unius aquilae (8:13).

Apc 9 (chunk 0009): Et quintus angelus · Et vidi stellam · Stellam · Et data est (9:1) · Aperuit puteum abyssi · Fumus fornacis · Obscuratus est sol et aer (9:2) · Exierunt locustae (9:3) · Praeceptum est illis ne laederent (9:4) · Ne occiderent · Mensibus quinque · Ut cruciatus scorpii (9:5) · Et fugiet mors ab eis (9:6) · Similitudines locustarum · Similes equis · Tanquam coronae (9:7) · Et habebant capillos · Dentes leonum (9:8) · Loricas · Sicut vox curruum (9:9) · Et habebant caudas (9:10) · Et habebant super se regem · Cui nomen Hebraice Abaddon (9:11) · Vae unum abiit (9:12) · Ex quatuor cornibus · Altaris aurei · Vocem unam (9:13) · Solve quatuor · Qui alligati sunt in flumine magno Euphrate · solve (9:14) · Soluti sunt · In horam et diem · horam · diem · mensem · annum (9:15) · Vicies millies · Et audivi · audivi (9:16) · Et ita vidi · Habebant loricas igneas · Vel habebant loricas igneas · Tanquam capita leonum · Vel ignis · fumus · sulphur · Fumus et sulphur (9:17) · Occisa est tertia pars (9:18) · Similes serpentibus (9:19).

⚑ One lemma is the commentator's own citation of a variant, not a divergence: at 0727B he writes Alia littera dicit mensibus sex, propter sex aetates — he is reporting a rival reading against the mensibus quinque of Apc 9:5, and says so. No marker; the English says so too.

1d. Fragmentary lemmata — how each was resolved, and why the two calls differ

Convention 3 renders a fragment as the phrase it stands for; the brief's closing note adds that a complete one-word lemma is rendered as it stands, and that contrasting the two classes manufactures an inconsistency. Two lemmata in this range sit on that boundary and were decided in opposite directions, deliberately:

• 0724D–0725A, Deducet eos ad vitae, etc. — EXPANDED to "He shall lead them to the fountains of the waters of life". "To life's" is not English at all; the fragment breaks inside a genitive phrase (Apc 7:17 ad vitae fontes aquarum), which is precisely Ruth's Non te. class.
• 0723D–0724A, Centum quadraginta. — NOT expanded, rendered "A hundred and forty". The clip lands on a whole numeral that is readable English; nothing is ungrammatical, so convention 3's trigger is not met, and expanding it to "a hundred and forty-four thousand" would import a number into a paragraph where the number is itself in question (see §2, 0724A). The Latin fragment stands opposite in the parallel column either way.
• 0727A, Sicut habent scorpiones. — rendered as printed, "As the scorpions have". This one elides internally: Clementine Apc 9:3 reads sicut habent potestatem scorpiones terrae, so both the object and the qualifier are gone from the middle of the lemma, not clipped off its end. Supplying "power" would have been the only expansion available and it would have put a word inside Migne's phrase rather than after it. Rendered as it stands; the elision logged here.
• Ordinary end-clips, rendered as the phrase they open and needing no note: Cornua septem, etc. · Ceciderunt in facies, etc. · Sapientiam, et fortitudinem, etc. · Sumeret pacem…, etc. · quia bilibris tritici, etc. · Post haec, etc. · Stolas suas, etc. · Non esurient, neque sitient, etc. · Similitudines locustarum, etc. · Sicut vox curruum, etc. · Et audivi, etc. · Soluti sunt, etc. · Quae habebant animas, etc. · Praeceptum est illis ne laederent, etc.

§2 · THE PLATE'S OWN TROUBLE — rendered as printed, logged here

0720A (0005) · intellexeruut — a non-word. [sic:] TAKEN.

Sancti et antiqui Veteris Testamenti patres intellexeruut aliquando… — intellexeruut is no Latin word (u for n). Corpus control: 1 occurrence against 21 intellexerunt across all of src/latin/. Per addendum A a non-word is the class worth raising, and it is carried under Patterns 7/10/12: the English gives the sense and the marker — "…of the Old Testament understood [sic: intellexeruut] at one time…". Strip the marker and the sentence still reads (test 1). ⚑ Plate-read candidate: this may be the digitization's corruption rather than Migne's type; the marker asserts only that the run in our Latin twin is carried, which is what the verifier checks.

0726D–0727A (0009) · baereticorum — a non-word. [sic:] TAKEN.

Discipuli baereticorum locustis comparantur. Corpus control: 1 occurrence of baereticorum (plus one baereticae elsewhere) against 501 haeretic- forms; this work alone prints haeretic- fifteen times in the same chunk. Rendered "The disciples of the heretics [sic: baereticorum] are compared to locusts". Same plate-read caveat.

0722A (0006) · a full stop mid-sentence — PRESERVED, per Pattern 8

Videns diabolus se per apertas persecutiones non profecisse. immittit haereticos… — Migne sets a period and then a lower-case immittit, which is the same class as 11542 @1041D. The English reproduces both: "…had not prevailed by open persecutions. he sends in heretics…". Pattern 8's "never delete"; not tidied, not merged.

0722D (0006) · a paragraph that simply stops — PRESERVED

Interfectorum propter verbum. Vel aperto martyrio, vel aliquibus anxietatibus* — no terminal mark. The English stops in the same place with no mark supplied (Pattern 8).

0725B (0008) · VERS. 3-- — one verse address out of eighty lacks its period

Every other address in the range is set VERS. n.--; this one is VERS. 3--. The address passes through verbatim, untranslated, so the English prints VERS. 3-- too. ⚑ Plate-read / TEI-patch candidate. This is exactly Ruth's 0537D class, where the digitization dropped the space after VERS. at one mark in twenty-three and the plate set it like all the others. Not patched here — a patch is a claim about the plate and this range has had no plate read. If a reader confirms Migne sets VERS. 3.--, it belongs in data/tei-patches/8977.json, not in the English.

0724A (0007) · the arithmetic does not come out — centum sexaginta quatuor. NO MARKER.

duodecim per quatuor multiplicamus, et fiunt quadraginta octo; ut vero ad Trinitatem haec perfectio referatur, quadraginta octo multiplicamus, et sic fiunt centum sexaginta quatuor. 48 × 3 = 144, which is the verse's own number (Apc 7:4 centum quadraginta quatuor millia), and the same column's preceding lemma prints Centum quadraginta. Migne prints 164. Rendered as printed, "a hundred and sixty-four". ⛔ [cj:] considered and DECLINED. sexaginta is a real, well-formed word occupying its slot, so Pattern 12 has nothing to wrap; and Pattern 18a's ruling is scoped to negations and correlatives (with the census closed over verb and word substitutions), while addendum A is explicit that a real word reading oddly is Migne's until a plate says otherwise and must not take a marker on the strength of an argument, however good the argument. This is a good argument. It is still an argument. ⚑ Plate-read candidate, high value: sexaginta / quadraginta. ⚑ Note also that quadraginta octo multiplicamus names no multiplier — "we multiply forty-eight" — the per tria being elided as per quatuor was not. Rendered elliptically, as printed.

0726C (0009) · the fifth angel called quartus. NO MARKER.

Et quintus angelus. Ordine narrationis et ordine temporis hic angelus dicitur esse quartus. The formula recurs three times in the range and elsewhere agrees with the lemma (0725C Et primus angelus … Ordine narrationis; 0726A Et tertius angelus, ordine narrationis; 0726B Quartus dicitur ordine narrationis; and at 0727C Vae unum abiit is glossed primum ordine narrationis et ordine temporis). Against a lemma reading quintus, quartus is the odd man. Rendered as printed. Declined for the same reason as 0724A: a real word in its slot, and the corpus cannot convict it. ⚑ Plate-read candidate.

0726C (0009) · clavis illa sunt — a number disagreement with an English exponent

Vel clavis illa sunt ipsi principes per quos haeretici operantur. A singular noun and demonstrative against a plural verb and plural predicate. English does have an exponent for this (that/those, is/are), so Pattern 12's second test does not apply — rendering it does not hide it. Rendered literally: "Or that key are the very princes through whom the heretics work." No marker; logged. ⚑ Plate-read candidate (claves illae).

0722B (0006) · nec per apertas haereses — apertas against the work's own scheme

Videns diabolus nec per apertas tribulationes, nec per apertas haereses posse proficere… The chapter's own frame two columns earlier makes the third horse haereticorum persecutio occulta and only the second one open, so "open heresies" is what the argument denies. Rendered as printed. ⛔ [cj:] considered and DECLINED on addendum A's ground: apertas is a real word in its slot, the second apertas may simply be Migne repeating the first, and no corpus argument may fire a marker. ⚑ Plate-read candidate. Both negatives (nec … nec) are carried and each is on its own prepositional phrase, not on the verb (test 2/2a).

0728D (0009) · nunquam fuerunt baptizati, nec ab Ecclesia separati — the contradiction stands

omnes Judaei vel pagani, qui nunquam fuerunt baptizati, nec ab Ecclesia separati. Carried whole: "who were never baptized, nor separated from the Church." Under 7a this is rendered as printed even though the two limbs sit oddly together (one who was never baptized was never in the Church to be separated from it). No word translated as its opposite; no negation moved; nothing supplied. Whether the plate reads otherwise is a question for a plate read, not for the English.

0723D (0007) · signatio pectoris, confessio est cordis — rendered as printed

Signum geritur in fronte ne celetur in tribulatione, nam signatio pectoris, confessio est cordis. The very next sentence assigns the breast to fidem cordis and the forehead to confiteatur, so the two do not divide the body the same way. Rendered literally, "for the signing of the breast is the confession of the heart". No conjecture offered.

0723C–D (0007) · Vel a se descendit — rendered as printed

In a paragraph glossing Ascendentem and running ascendit … ascendit ab ortu, the third alternative prints Vel a se descendit. A real verb in its slot; rendered "Or he came down from himself". ⚑ Plate-read candidate; no marker.

Pattern 9 items — checked and NOT marked

• 0727C sagita for sagitta (corpus: 1 against 76 sagitta-). Single-for-double consonant is standard medieval orthography, and Pattern 9 rules that attested medieval spellings are correct Latin for their period. Declined, though it is the closest call in the range: it is a hair's breadth from the non-word class, and a plate read would settle it.
• 0728D Antechristi in a chunk that prints Antichrist- fifteen times. Antechristus is an attested medieval form; corpus-wide it occurs once, which is a reason to raise it and not a reason to mark it. Rendered "Antichrist". ⚑ Plate-read candidate.
• 0727C adherentes for adhaerentes — ordinary ae/e monophthongization. No marker.
• 0720D Haebraei, 0725A lacrymam, 0723A coelum, tanquam throughout — orthography, no marker.

### §3 · REFERENTS AND NEGATION — the sites where two readings construe, or where a supplied ## pronoun had to be named

The brief's dominant defect class. Every "he/she/it/they" not printed in the Latin was named to a specific antecedent before it was written. The ones worth recording:

• 0720A (0005), et doctrina et exemplo munivit — no object is printed. Supplied as the Church: she is the subject of the immediately preceding quae fuit albus equus, and the next relative, quam rufus equus impugnavit, is feminine accusative and can only be the Church. The horse (equus, masculine) is excluded by that quam. English: "which he fortified by doctrine and by example, and which the red horse assailed."
• 0720D (0005), caro sanctorum, quae tenditur in ligno, dum … imitantur — quae is singular and binds caro (the flesh), while reddentes … imitantur is plural and binds the saints. The English keeps both: "the flesh of the saints, which is stretched upon the wood, while by divers torments, rendering the same sound, they imitate the cross of Christ." Collapsing them into one number was the available error and was not made.
• 0723C (0007), ne sui aliquatenus torpeant vel desperent — sui is God's own, not the devil's: the clause explains why the devil's readiness to hurt is reported together with his being bridled, and the consolation is for the faithful. Both readings parse; the sense decides, and the English says "his own" with the ambiguity recorded here rather than resolved silently in a footnote the reader cannot see.
• 0723D (0007), qui est ortus solis, id est Patris — the genitive Patris glosses solis, not ortus; the sun is the Father. Kept on its own noun: "who is the rising of the sun, that is, of the Father." (The brief's named failure — a genitive lifted off its noun — was the risk here.)
• 0728C (0009), sicut intellexeram … sic intellexit — a first-person pluperfect answered by a third-person perfect. Rendered as printed: "as I had understood … so he understood". Under Pattern 18a a person slip is explicitly declined for [cj:], and under 7a the English may not quietly become "I understood". ⚑ Plate-read candidate (intellexi).
• Test 2a sites — the host of the negative proved from the sentence's own structure, not from what reads well:
• 0722A non solum in minores, sed etiam in majores — host is in minores, proved by the sed etiam limb. English: "not only against the lesser, but even against the greater".
• 0722B oleum sunt non alios inebriantes, sed prius orationibus ungentes — host is alios inebriantes, proved by the sed limb; not the copula sunt.
• 0723D Hoc signum etsi non re, fide tamen antiqui patres praetulerunt — host is re, proved by the answering fide tamen. English: "though not in fact, yet in faith".
• 0724C Non per nos, sed per agnum habemus salutem — host is per nos.
• 0725C quia non omnes orationes repraesentat Patri, sed eas exaudit quae… — host is the quantifier omnes, proved by the sed limb, which is what makes the sentence coherent: he presents not all prayers, not "he does not present prayers". English: "because not all prayers does he present to the Father, but he hears those which pertain to salvation." ⚑ This is the one site in the range with 2a's exact trigger shape (negative, verb, quantifier adjacent), and the wrong host reads perfectly well.
• 0727B non aperte tormentant, sed apud saeculi principes accusant — host is aperte, proved by the sed limb (they do torment; not openly).
• Every other printed negative was checked onto its own word and carried: non/nec/ neque/nisi/ne/nihil/nullus/nemo/nunquam/nolentes/nondum at 0720A, 0720B, 0721B, 0721C, 0721D, 0722A ×3, 0722B ×4, 0722C ×2, 0722D ×2, 0723A ×3, 0723B, 0723C ×2, 0723D ×4, 0724B ×2, 0724C, 0725B ×3, 0725C ×2, 0726A, 0726D, 0727A ×4, 0727B ×2, 0727C, 0727D, 0728D ×2. None appears in the English that Migne does not print, and none of his is missing.

§4 · PUNCTUATION — what was corrected in our own draft, and the standing decision

Test 3 was run band by band across all 37 anchor-delimited bands. The first pass found 24 mismatched bands, every one of them the translator's convenience and not the plate's: semicolons and colons supplied where Migne prints commas (and, at 0725D and 0726D, two of his semicolons softened to commas in our English). All were conformed to the printed marks; the second pass returns zero mismatched bands. Recorded because the brief's own warning is that the totals can cancel — here they did not cancel, but the class is exactly the one it names, and it was invisible until the bands were compared.

⚠ No comma→semicolon sweep was performed in the other direction: Migne prints semicolons himself in this range (twelve of them), so Tobit's narrow exception does not apply and was not invoked.

§5 · SECTIONS READ CLEAN, AND CANDIDATES REJECTED

Read clean, no crux: CAPUT V's glosses on 5:8–5:14 (0720D–0721B); CAPUT VII's tribal etymologies at 0724B (Simeon / Levi / Issachar / Zabulon / Joseph / Benjamin — the Hebrew glosses are the commentator's and are rendered as he gives them, not corrected against Jerome); CAPUT VIII entire apart from 0725B's VERS. 3--; CAPUT IX's 0728B chronology (horam = half a year, diem = a year, mensem = a second, annum = a third, totalling the three and a half years Migne then names — internally consistent, checked, no defect).

Candidates raised and rejected, with the reason:

1. sagita (0727C) — Pattern 9 orthography, not a non-word. Declined.
2. Antechristi (0728D) — attested medieval form. Declined.
3. centum sexaginta quatuor (0724A) — real word, corpus argument cannot fire a marker. Declined; plate read requested.
4. quartus for quintus (0726C) — same. Declined; plate read requested.
5. apertas haereses (0722B) — same. Declined; plate read requested.
6. descendit for ascendit (0723D) — same. Declined.
7. intellexit for intellexi (0728C) — a person slip, explicitly outside Pattern 18a. Declined.
8. clavis illa sunt (0726C) — English has an exponent for the disagreement, so rendering it does not hide it and Pattern 12's second test is not met. Declined.
9. Luke 10:24 word order, Apc 8:3 word order, Apc 9:17 tense — divergences that do not change the sense. [var:] declined for each, all three logged in §1b.
10. The A–D band sequence of this range (0719D → 0720A … 0729A) contains no gap or recurrence worth reporting; per runbook false positive 1, no band arithmetic was performed and none is reported. No text is broken mid-word or mid-construction across any anchor.
11. No [f: …] tags: this is a lemma-and-gloss commentary, runbook false positive 4.
12. No em-dash was marked, reported, or engineered around.

Open, for whoever reads the plate: the eight plate-read candidates flagged above, in descending value — centum sexaginta quatuor (0724A) · vivum/vinum (0722B) · indicem/invicem (0721D) · VERS. 3-- (0725B) · quartus (0726C) · apertas haereses (0722B) · clavis illa sunt (0726C) · intellexeruut (0720A) and baereticorum (0726D), these last two only to establish whether the corruption is Migne's or the digitization's, since the [sic:] markers stand either way on the run our Latin twin carries.

Preserved unedited from cruces-0010.md.

Glossa ordinaria, Apocalypsis B. Joannis (PL 114) — cruces, chunks 0010–0014 (cols 0729A–0739B)

Range file for the third of four agent stints (Apc 10–16). Not the merge — the merged cruces.md is written once at the end from all four range files.

Translated 2026-08-28. Chunks 0010, 0011, 0012, 0013, 0014. Marker parity against the Latin twins, verified paragraph by paragraph: column anchors 11 · 6 · 6 · 12 · 6, verbatim and in order; [n:] 2 · 0 · 0 · 0 · 0; italic spans 106 · 60 · 112 · 118 · 62 (exact, with marker contents stripped); question marks 0 in both languages throughout; guillemets 0; ## heads 1:1 (CAPUT X.–XVI. → CHAPTER X.–XVI., Ruth's shipped form). Colons and semicolons compared band by band (paragraph by paragraph, the anchors partitioning both twins identically): zero mismatched bands after the five repairs listed in §4.

Apparatus taken in this range: 7 [var:] · 1 [cj:] · 0 [sic:] · 0 [d:] · 0 [ed:] · 0 [nt:].

§1 · THE 7a″ DELIVERABLE — every lemma in the range, collated

Collated against sources/vulgate/clementine-flat.txt, cited from sources/vulgate/clementine/Apc.lat, Ps.lat, Mt.lat, 1Cor.lat, Zach.lat. This section lists the agreements as well as the divergences, because a findings-only list cannot be told apart from a report by an agent that never looked.

1a. Divergences that took a [var:] (7)

⚠ On the last: enim is a connective, not a sense-word, and the marker was still taken because the insertion is what turns a quoted verse into the gloss's own middle term. A reader who does not know the psalm cannot see that the "for" is Migne's.

1b. Divergences examined and deliberately NOT marked, with the reason

• 0729C Tempus amplius non erit. — Apc 10:6 reads Quia tempus non erit amplius. Word order only; the negation, its host (erit) and the sense are untouched. Pattern 14 wants a divergence "in a way that changes the sense". Logged, not marked.
• 0729C Et ea quae in illo. — Apc 10:6 has ea quae in eo sunt (three times). A bare demonstrative swap, illo for eo, with no change of reference or sense. ⭐ Worth recording that the gloss's own re-quotation four words later reverts to eo (Creavit ea quae in eo sunt), so the divergence is confined to the lemma. Logged, not marked.
• 0735C Et angelus tertius. — Apc 14:9 Et tertius angelus secutus est illos. Word order only.
• 0739A Dare ei calicem. — Apc 16:19 dare illi calicem. Same pronoun swap as 0729C; no sense change.
• 0736C, the Prophet's Lacus ubi non erat aqua — the nearest Clementine text is Zach 9:11 de lacu in quo non est aqua (Gn 37:24 cisternam … quae non habebat aquam is the other candidate). Migne's is a loose allusion, not italicized on his page, and the negation is intact in both. No marker; the English keeps it unitalicized, as he prints it.
• 0729C Nam si Christus resurrexit, et nos resurgemus [n: (I Cor. XV)] — no such sentence stands in 1 Cor 15. It is a free epitome of the chapter's argument (cf. 15:12–16, 20), and Migne's own note cites the chapter, not a verse. ⛔ A [var:] here would have to pick one verse to be "the" received reading, which would misdescribe what he is doing. Declined deliberately; logged instead.
• Orthography, throughout, and none of it marked: coelo/coelum/coeli for caelo etc.; tanquam for tamquam (0729A, 0737A, 0738A); characterem for the Clementine's caracterem (0734B); Citharaedorum for citharoedorum (0735A); terraemotus set solid. These are Pattern 9 items — conventions of the printing, carrying no information about this plate — and marking them would bury the seven real divergences in noise.

1c. Lemmata checked and found to agree with the Clementine exactly

Every remaining italic span in cols 0729A–0739B was walked against the flat file. All of the following agree word for word with the verse they gloss (fragments agreeing with the corresponding stretch of it):

Apc 10 (chunk 0010): Et vidi alium angelum fortem · De coelo · Amictum nube · Et iris in capite · Et habebat in manu sua libellum apertum · Pedem suum · Sinistrum autem super terram · Quemadmodum cum leo rugit · Scripturus eram · Signa quae locuta · Levavit manum suam ad coelum · Juravit · Consummabitur mysterium · Mysterium Dei · Sicut evangelizavit · Et audivi · Vade · Stantis super mare · Et abii · Ad angelum · Et dixit mihi: Oportet. Apc 11: Metire · Atrium · Civitatem sanctam · Et dabo duobus · Duobus testibus · Diebus mille · Super aquas · In sanguinem · Aegyptus · Et munera mittent · Et post dies · Super pedes · Vocem magnam · Terraemotus magnus · Et decima pars · Occisa sunt in terraemotu · Gloriam Deo · Et septimus angelus · Factum · Et regnabit · Et viginti quatuor seniores · Ceciderunt in facies · Qui corruperunt terram · Et apertum est · Arca · Et grando magna. Apc 12 (chunk 0011): Et signum · Luna sub pedibus · Et in capite ejus corona · Draco magnus · Et cornua decem · Et cauda · Stellarum · Stetit ante mulierem · Devoraret · Et peperit · Diebus mille · Michael et angeli ejus praeliabantur · Non valuerunt · Et audivi vocem · Accusabat illos · Die ac nocte · Descendit diabolus · Sciens quod modicum tempus habet · Et postquam vidit · Et datae sunt mulieri · Aquilae · Per tempus, et tempora (all three quotations) · Et adjuvit terra mulierem · Et aperuit · Absorbuit · Iratus est draco in mulierem · Et stetit. Apc 13 (chunk 0012): Et vidi · Capita septem et cornua decem · Nomina blasphemiae · Quasi occisum · Et plaga mortis ejus curata est (and its re-quotation) · Et aperuit os suum · Et tabernaculum · Occisus est ab origine mundi · Qui in captivitatem · Qui in gladio occiderit · Hic est · Et vidi · Cornua · Et loquebatur sicut draco · Et potestatem · Terram (and its re-quotation) · habitantes in ea · Adorare · Ignem · In conspectu · Ut loquatur imago · Numerum · Qui habet. Apc 14–15 (chunk 0013): Et vidi: et ecce Agnus stabat · Centum quadraginta · Tonitrui · Citharizantium · Et cantabant · Hi sunt qui cum mulieribus non sunt · Ex hominibus primitiae · Et vidi · Alterum angelum · Per medium coeli · Evangelium aeternum · Timete Dominum · Quia venit · Cecidit (both) · Vino irae fornicationis · Quod mistum · In calice · Et fumus tormentorum · Ascendet in saecula (both) · Nec habent requiem · Hic patientia · Opera · Et vidi · Similem Filio hominis · Angelus · Exivit de templo · Quia venit hora · Et alius angelus · Alius angelus · Mitte falcem · Et misit in lacum · Et exivit sanguis · Usque ad frenos equorum · Per stadia · Et vidi · Et mirabile · Plagas septem · Quoniam in illis · plagas · novissimas · Tanquam mare vitreum mistum igne · Habentes citharas · Canticum Moysi · Viae · justae · verae · Judicia tua · Et post haec · Apertum est templum · Et exierunt · Et unum · Iracundiae · Fumo a majestate · Introire. Apc 16 (chunk 0014): Vulnus · Et factus est sanguis · Sanguis · tanquam mortui · Omnis anima vivens · Super flumina · Audivi · Qui es et qui eras · Et sanguinem eis dedisti bibere · Et audivi · Etiam, Domine Deus omnipotens · Justa · Et aestuaverunt · Neque egerunt · Super sedem bestiae · Commanducaverunt · Praepararetur via · Vidi · Ranarum · Procedunt · Et congregabit · Et septimus angelus · Et facta est civitas · Civitates gentium ceciderunt · Venit in memoriam · Et omnis insula fugit · Talentum · Et blasphemaverunt.

1d. Fragment lemmata — how each was rendered, per convention 3

• 0729C Signa quae locuta: rendered "Seal up the things which the seven thunders have spoken:". Signa is the Vulgate's imperative (Apc 10:4 Signa quae locuta sunt septem tonitrua), not the noun "signs"; a stump ("Signs which have spoken") is not English and would also invite the wrong part of speech. The trailing colon is Migne's and is carried.
• 0730B Et dabo duobus. expanded to "And I will give to my two witnesses" — the governed noun is elided (Apc 11:3 dabo duobus testibus meis). Its neighbour Duobus testibus. is a complete phrase, nothing elided inside it, and is rendered as it stands, "To two witnesses". ⚑ Per the brief's control-class rule these are two different classes and the pair is not an inconsistency.
• 0730D and 0732B Diebus mille. kept as "A thousand days", NOT expanded to the verse's diebus mille ducentis sexaginta. Reason, and it decides the sense: the gloss argues hi dies non perficiunt tres annos et dimidium. On 1,260 days that is at best arguable (1,260 = 3½ years of 360 days); on the bare thousand Migne prints it is plainly true. Expanding would have imported the very number the gloss is denying.
• Ordinary clipped lemmata (De coelo, Pedem suum, Gloriam Deo, Super pedes, Cornua, Opera, Talentum, Vulnus, Justa) are rendered as the phrase they open, without importing wording from any English Bible. No lemma in this range was conformed to Douay or AV; each was read word by word off the Latin before any English was written.

§2 · CRUCES

0733B · in [0733B] fructuosos — the word is split across the column anchor, and the split hides a negative. [cj:] TAKEN.

Migne, glossing Apc 12:18 Et stetit [supra arenam maris]:

Non tamen omnes illos vicit minores, sed tantum super arenam, id est, in [0733B] fructuosos moram fecit, in quibus apparent signa vestigii.

The dragon did not conquer all the lesser ones, but only super arenam — and sand is barren. in fructuosos, "among the fruitful", says the opposite of what arenam and the whole sentence require; infructuosos, "the unfruitful", says exactly it, and sand is precisely what takes and shows a footprint (in quibus apparent signa vestigii).

This is the one thing the runbook's false-positive item 1 does tell us to report: text broken mid-word across an anchor. The anchor falls between in and fructuosos, which is also a legitimate word boundary if the text really is two words — which is what makes it undecidable from the file.

• Marker: [cj: fructuosos; read infructuosos, "the unfruitful"], additive per Pattern 18 — Migne's word keeps its English ("among the … fruitful") and the conjecture stands beside it. Nothing is removed and the reader sees both.
• Not a [sic:]: fructuosos is a perfectly good Latin form occupying its slot, so Pattern 12 has nothing to wrap. This is the 11081 @0030B munde/mundo shape exactly.
• Corpus control: infructuos- is attested 32 times across src/latin/ (though not in this exact accusative plural), so the conjectured word is a word this corpus uses.
• ⛔ The rival reading, and why it lost: "he made his stay among the fruitful" can be construed — moram facere in + accusative for "among" is loose but medieval. It loses because it contradicts super arenam in the same clause, which is the sentence's own structure and not a matter of what reads well.
• Plate read wanted at col. 0733B. If Migne prints infructuosos broken at the column turn, this is Corpus Corporum's division and belongs in data/tei-patches/8977.json, and the [cj:] is withdrawn in the same commit.

0734A · tundamentum — a NON-WORD. Plate read wanted. No marker fired, deliberately.

Quia et hi, simulata resurrectione, praedicationis suae faciunt tundamentum.

tundamentum is no Latin word. Corpus frequency across src/latin/: 1 × tundamentum against 107 × fundamentum. The shape (f → t, the crossbar) is a classic broken-type or mis-transcription.

⛔ No [sic:]. The brief's addendum A records that every suspicion aimed at a non-word in that batch was confirmed as our corruption, not Migne's — and addendum B records seven false [sic:] withdrawn in one batch, each publicly blaming Migne for the digitization's error. Firing one here on a corpus argument alone would be the same mistake. The Glossa block's own precedent points the same way: Ruth's significaRuth @0534A was resolved by a tei-patch after a plate read, not by a marker.

The English therefore reads "make the foundation of their preaching", and this entry is the trace. ⚑ If the plate reads tundamentum, a [sic:] must be added here and the sense kept beside it (7a⁗ test 1: the word is the clause's object, so quarantining it alone would leave "make the ___ of their preaching").

0734D · centecimum — non-word or attested medieval spelling? No marker; plate read wanted.

Hi reddunt fructum, alii tricesimum, alii sexagesimum, alii centecimum.

Corpus: 1 × centecimum against 21 × centesim- forms. Two readings, and they point different ways:

1. Our corruption (s → c), same class as tundamentum above.
2. An attested medieval spelling. Assibilated c/s before a front vowel is a real convention of the period (cf. nuncius/nuntius), and under Pattern 9 an attested medieval spelling is correct Latin for its period — translate normally, no italics, no crux.

Because reading (2) is live, and because either way the marker would be wrong, none was fired; the English reads "a hundredfold". ⚑ Raised here as a plate candidate. Note also that tricesimum in the same line is not suspect — it is classical, and no divergence.

0735C · corrrigitur — trebled r. Same class, same treatment.

Qui non corrrigitur vindicta ejus ad correctionem data, punietur aeterna poena.

Corpus: 1 × corrrigitur against 8 × corrigitur. A doubled letter is a plate or transcription accident, not a reading; the negation non attaches to it and is carried ("He who is not corrected …"), which is the thing that matters. No marker; plate read wanted at 0735C.

0737B–C · CAPUT XV prints VERS. 18 for what is Apc 15:8. Reproduced as printed.

Apocalypse 15 has eight verses. Migne's last address in the chapter is VERS. 18.-- Fumo a majestate, etc., and the text glossed is unmistakably 15:8 (Et impletum est templum fumo a majestate Dei … nemo poterat introire). The verse number is his, and the brief is explicit that his numbering is not ours to tidy: reproduced verbatim, not corrected to VERS. 8, and logged here. (The following chunk resumes at CAPUT XVI VERS. 2, so nothing downstream depends on it.)

0737B · Cum ostendissent — a plural verb where the sentence has a singular subject.

Cum ostendissent illos angelos habere potestatem plagandi, … prius proponit causam…

ostendissent is 3rd plural, proponit 3rd singular, and both clauses have the same implied subject (John). Rendered as printed — "When they had shown … he first sets forth" — because English does have an exponent for the number, so the disagreement stays visible to the reader (this is not the Pattern 9 gender case, where English has no exponent). No marker: a real word in its own slot, wrong only in number, is Pattern 7's business and not Pattern 12's.

0729C · Cum dicis — second person singular. [cj:] CONSIDERED AND DECLINED.

Cum dicis: Creavit ea quae in eo sunt, haeresim quorumdam destruxit…

dicis is "thou sayest", and the first reflex is that the sense wants dicit, "when he says", since the words quoted are the angel's and destruxit has a third-person subject. Declined, and the reason is that the printed reading construes without help: the commentator is addressing John, whose vision this is and to whom he has been speaking throughout the chunk (monetur, sibi, Ecce revelavi tibi omnia), and John both says the words and, by saying them, destroys the heresy. A [cj:] asserts that the faithful English misleads; here it does not. Rendered "When thou sayest", singular per Pattern 17.

0736D–0737A · consideratione tamen — rendered literally, sense not settled.

Sed omissa prima, quae fuit de correctione ecclesiarum, de aliis tribus dicit, consideratione tamen.

A bare ablative closing the sentence with no governing word. "Yet with consideration" is what the words say and is what the English prints; whether Migne means "though with [some further] consideration", or the phrase is the stub of something longer, cannot be settled from the file. No marker, no conjecture — flagged so a blind reader does not read past it.

0738D · ad quem gloriabuntur — ad for in.

gloriari takes in; ad quem gloriabuntur is not standard. Rendered "or unto whom all the wicked of the diverse sects will glory", as closely as the printed words allow. Real word, real construction, merely unusual — Pattern 7, no marker.

0734B · remunerationem eo minorem — obscure, rendered literally.

eo read as an ablative of degree of difference against minorem: "a reward so much the less, or a middling one, or the greatest", answering the sex / sexaginta / centum ladder that follows. Flagged because the alternative (eo = "from him", already carried by ab eo three words earlier) would be a dittography of sense and is not obviously wrong.

0735A · sua corda devota — a pun English cannot hold.

ut extenderent sua corda devota super lignum crucis.

Under Citharaedorum, "of harpers": corda is strings (of the cithara) and devota is the language of the heart. English must choose one vehicle; "their devout strings" was taken, because the harp governs the whole image and the wood of the cross is the sounding board. The heart-sense is not lost to a reader who has the Latin column opposite. Recorded because a later reader may reasonably prefer "hearts".

§3 · REFERENTS SUPPLIED — every "he/she/it/they" the Latin does not print

The dominant Glossa defect class. Each supplied pronoun in this range, with what fixes it:

• 0729A–0730B, chunk 0010 passim — the unexpressed subject of exsuperavit, fecit, posuit, rugit is the angel of Apc 10:1 identified with Christ by the opening gloss (fortis armatus … exsuperavit, and ita Christus suos in fide concludens). Named in the sentence itself at 0729B, so the "he"s are not free.
• 0729C Cum dicis … destruxit — see §2. Subject of destruxit is John, addressed in dicis, not "the angel"; either reading gives the same English word, but the crux records that the choice was made and not defaulted.
• 0730A monetur quae vidit interim praedicare — John, from Et audivi (Apc 10:8) and from quasi dicatur aperte sibi.
• 0733C ejus in Et tabernaculum. Id est omnem institutionem ejus — God's, from Apc 13:6 (blasphemare nomen ejus, et tabernaculum ejus), not the beast's. The following per quam laudabilis apparet confirms it: it is God who appears praiseworthy through his institution. ⚑ Binding ejus to the nearest noun (the beast) would have made the beast's own apparatus praiseworthy, which is the exact "relative bound to the nearest noun" failure the brief names.
• 0735A ad consortium suum under Tonitrui — "their fellowship" = the hundred and forty-four thousand of the lemma above (Apc 14:1), not the thunder's. The antecedent is in the previous lemma, as the brief warns is usual here.
• 0736B numerus tuorum — thine, singular, addressed to Christ/God (ut Deus separationem faciat two clauses earlier). Pattern 17: singular addressee takes thou.
• 0732D aggreditur expugnare — the object supplied as "her" = the woman/Church (mulierem in the same verse and Ecclesiam throughout the gloss), not the Son.
• 0738B isti malitia diabolum — verb supplied from the parallel (generant): "so these by malice beget the devil". The ellipsis is Migne's and the parallel is his own.

Where two readings both construe, they are named above rather than chosen silently.

§4 · NEGATION AND PUNCTUATION — what was checked, and what was repaired

Negation (test 2 + 2a). Every printed non · nec · neque · nisi · nunquam · nullus · nihil · ne · haud in cols 0729A–0739B was located, its host word named, and the host found under the negative in the English. The population that fires trigger 2a — a negative on a verb with a quantifier or correlative nearby — was worked from the sentence's structure, not from what reads well:

• 0729C non ad detrimentum suae salutis praedicationem subtrahi — host is the phrase ad detrimentum, not subtrahi. English: "preaching is withdrawn not to the detriment of their salvation." (The 8960 @0740B model.)
• 0730A non tanta patieris quanta patientur — host is the quantity tanta, fixed by the quanta correlative. English: "thou shalt not suffer so great things as they shall suffer."
• 0731B Non tantum minores ita glorificant Deum, sed etiam majores — host is tantum minores, fixed by sed etiam. English: "Not only do the lesser thus glorify God, but the greater also." ⛔ Not "the lesser do not glorify".
• 0732A Non de magno effectu, sed de potentia — host is the prepositional phrase.
• 0735C non dicit pro damnatione describenda: sed tantum pro terrore inferendo — host is pro damnatione describenda, fixed by sed tantum.
• 0736B non quales olim fuerant, sed quales in aeternum mansuri — host is quales olim fuerant.
• 0736C Aeterna poena … non autem vinum … ponitur — host is vinum, not ponitur; one verb serves both limbs.
• 0738A non quia aperte persequuntur, sed propter consensum damnabuntur — host is the quia clause, not damnabuntur. English: "they shall be damned not because they openly persecute, but on account of their consent."
• 0738B ii sunt tantum corpora, et non spiritus … corporalia et non spiritualia dabit — hosts are the two nouns.
• 0739A sed non quantum meruerunt — host is quantum.
• 0733B — see §2: a negative that may have been removed by a word division, handled with a [cj:] rather than silently supplied.

⛔ Nothing was rendered as its opposite anywhere in this range, and no negative was supplied that Migne does not print. The sentences that read awkwardly in English (0736D consideratione tamen, 0738D ad quem gloriabuntur, 0734B eo minorem) read awkwardly because he prints them so.

Punctuation (test 3, by band). Compared paragraph by paragraph, the verified column anchors partitioning both twins identically. Question marks: zero in the Latin of all five chunks and zero in the English — nothing supplied, nothing deleted, and Pattern 8b never arose (the indirect questions at 0731C ad quae supplicia ducatur nescit and 0737D quare alios elegit both close with a full stop on his page and do so on ours). Five first-draft mismatches were found by the band comparison and repaired toward Migne, not away:

Semicolon and colon counts now match band for band in all five chunks, including chunk 0012's 25 semicolons, nearly all of them inside the Greek and Latin numeral tallies at 0734B–C, which are reproduced with Migne's own mixture of spelled-out numbers (unum, quinquaginta) and Roman numerals (CCCC, LXX, XL) rather than normalized to one or the other. The unitalicized A, unum at 0734C, sitting among italicized letters, is his inconsistency and is reproduced.

Em-dashes: none introduced. deservit/deserit: neither form occurs anywhere in cols 0729A–0739B — checked, and recorded as a checked zero.

§5 · TEST 5 — does the English ASSERT what he asserts?

Each finished paragraph was read back against the Latin with the counts already known clean, asking only that question. Three sentences that read smoothly were re-examined on that ground alone, and two of the three turned out to need nothing:

• 0730C "Know ye that all the persecutions … proceed from Antichrist, as do those which shall be in his own time." — sicut illas quae erunt in tempore suo. Checked: the comparison is that past and present persecutions proceed from Antichrist just as the ones in his own time will; the English does not soften it into "including". Clean.
• 0736C "just as to a vineyard, that it may bear fruit, greater care is applied than to land" — sicut vineae, ut fructificet major cura quam terrae adhibetur. The dative vineae and terrae are both governed by adhibetur; the English keeps the comparison on the care, not on the fruit. Clean.
• 0733D quia, illo destructo, sciunt se remunerandos — "he being destroyed" is Antichrist, from Qui in captivitatem … scilicet Antichristus two clauses up, not the sword and not the devil. Fixed by the lemma above; recorded in §3's class.

⚠ Not read clean, and named as such: 0736D–0737A consideratione tamen (§2). It reads as a stub, the English reads as a stub, and no reading was chosen.

§6 · WHAT A LATER READER SHOULD LOOK AT FIRST

1. 0733B — the [cj:] on fructuosos/infructuosos. A plate read settles it and may convert it to a tei-patch.
2. 0734A tundamentum, 0734D centecimum, 0735C corrrigitur — three plate candidates in one stint, all in the non-word class, none marked. If the plate is read for any one of them, read the whole line, not the word (addendum C).
3. 0737B CAPUT XV VERS. 18 — Migne's own numbering, reproduced. Do not "fix" it.
4. The seven [var:] — 0729A, 0729C, 0732A, 0734D, 0735B, 0738B, 0739A. Each rests on sources/vulgate/clementine-flat.txt, cited above by book chapter:verse.

Preserved unedited from cruces-0015.md.

8977 · Glossa ordinaria, Apocalypsis B. Joannis — cruces, chunks 0015–0019 (cols 0739C–0752B)

Range file for the fourth and last agent block (CAPUT XVII–XXII, the close of the work). Merge into cruces.md with the other three range files. Conventions are those settled on Ruth (src/english/8968/cruces.md §1) and the Glossa NT brief; they are not re-argued here.

verify-english.mjs 8977: verify OK — 20 chunks, columns/notes/sections aligned, no duplicate paragraphs.

Marker tally for this range: 5 [var:] · 1 [cj:] · 0 [sic:] · 0 [d:] · 0 [ed:] · 0 [nt:] · 10 [n:] (parity with the Latin twins).

Parity checked mechanically before filing, per chunk: column anchors (verbatim, in order, and each re-checked against the word it stands beside in the Latin), italic-span count with marker contents stripped, [n:] count, and punctuation band by band (; : ?), which is the check that actually returned sites — see §5.

§1 · THE 7a″ DIVERGENCE LIST — every lemma in the range, collated

Collated against sources/vulgate/clementine-flat.txt, cited from sources/vulgate/clementine/Apc.lat and the other books named. The whole of Apc 17–22 was pulled and read line by line; the pre-scan in data/briefs/8977-lemmata.txt was used only to order the work, and it is incomplete for this range — it carries no entry for chunk 0015's opening spans (Et venit unus · Angelis · Ostendam tibi damnationem) or for chunk 0018's (Et vidi · Coelum novum et terram novam · Sicut sponsam ornatam · Vocem magnam de throno · Ecce tabernaculum). Those were walked in place.

1a. The five [var:] taken

⚠ Two of these five (lignis, vide) are single-letter differences on a real word, which is the shape brief §4 says the digitization can create in either direction. The [var:] is still the right marker and is safe: it asserts only what the Clementine reads, never that Migne's compositor erred. If a plate read returns ignis / vidi, the fix is a TEI patch and the markers come out in the same commit; nothing on the page has to be retracted as a false accusation.

1b. Divergences found and declined, with the reason

Recorded because a findings-only list cannot be told from a report by an agent that never looked. Each of these was checked against the flat file and is a real difference from the Clementine; none of them changes the sense, which is Pattern 14's own threshold.

• 0739D · Job 40:11. Migne: Virtus ejus … in lumbis … in umbilico. Clementine: Fortitudo ejus in lumbis ejus, et virtus illius in umbilico ventris ejus. Migne substitutes virtus for fortitudo, drops ejus and ventris, and so attaches to the loins the noun the Vulgate attaches to the navel. Declined: virtus and fortitudo are near-synonyms here, and the gloss's whole argument turns on lumbis and umbilico, which Migne prints exactly as the Vulgate does. This is the strongest of the declines and a later editor may reasonably reverse it.
• 0741A · Jo 4:18. Migne: Quem nunc habes non est tuus vir. Clementine: et nunc, quem habes, non est tuus vir. Trimmed opening plus quem nunc for nunc quem. Declined: word order only; the negative and every content word stand.
• 0743A · the Vende omnia quotation. Vende omnia, et sequere me answers to no single verse: Lc 18:22 omnia quaecumque habes vende … et veni, sequere me; Mc 10:21 vade, quaecumque habes vende … et veni, sequere me; Mt 19:21 vade, vende quae habes … et veni, sequere me. Declined: a [var:] names one received reading against one printed one, and there is no single verse this compresses. Migne prints it without a note.
• 0745A · Apc 20:5. Migne: Caeteri vero mortuorum non vixerunt. Clementine: Ceteri mortuorum non vixerunt. An added adversative particle. Declined: connective only, and the gloss builds nothing on it. (Caeteri for ceteri is orthography, not a divergence.)
• 0748B · Phil 3:20. Migne: Nostra conversatio in coelis est. Clementine: Nostra autem conversatio in caelis est. A dropped particle in a quotation the gloss puts in the saints' own mouths (dicentes). Declined, same ground as the item above. Applying the two declines the same way is deliberate: added and dropped particles are one class.
• 0748B · Apc 21:19. Migne heads the stone Chalcedoni,; the Clementine reads tertium, calcedonius. Declined: a stone-name form in a series of lemma heads, all the rest of which Migne sets nominative (Smaragdus · Sardonyx · Sardius · Beryllus · Topazius · Chrysoprasus · Hyacinthus · Amethistus). No sense turns on it. Rendered "Chalcedony,", keeping Migne's comma and the [n: (BED.)] inside the span as he sets it.
• 0749C · Apc 21:21. Migne's second lemma sunt margaritae reorders the verse's own duodecim margaritae sunt. Declined: order only, inside a lemma he has already clipped.
• 0741B · Apc 17:16. Migne Igne concremabunt; Clementine et ipsam igni concremabunt. Declined: igne and igni are both regular ablatives of ignis.
• 0745B · Apc 20:8. Migne circumierunt; Clementine circuierunt. Declined: both are standard forms of circumeo.
• The coel- class is not a divergence and took nothing. Et vidi coelum apertum (0743B) · Coelum novum (0745D) · in coelis (0748B) · Exsultate super eam, coeli (0742D, whose [var:] is about number, not spelling). The pre-scan flags these ⚠ because clementine-flat.txt spells cael-; that is a known false positive and was ignored.

1c. Lemmata checked and found to agree with the Clementine exactly

Listed so a later reader can tell a checked lemma from an unchecked one.

Apc 17 (0015): Et venit unus (17:1) · Angelis (17:1) · Ostendam tibi damnationem (17:1) · Meretricis (17:1) · Quae sedet super aquas multas (17:1) · Inebriati sunt (17:2) · Mulierem sedentem · Capita septem · Cornua decem (17:3) · Circumdata purpura · Auro · Poculum aureum · Abominatione (17:4) · Et in fronte · in fronte (17:5) · Ebriam · Et miratus sum (17:6) · Ego tibi dicam (17:7, = ego dicam tibi, order only) · Bestiae quae portat eam · Fuit (17:7–8) · Ascensura est · Et in interitum ibit · Mirabuntur (17:8) · Septem capita (17:9) · Quinque ceciderunt (17:10) · De septem est · et ipsa octava (17:11) · Populi sunt (17:15) · Desolatam facient · Et nudam · Carnes (17:16) · Donec consummentur verba (17:17) · Ez 16:4 non est praecisus umbilicus tuus · Job 40:11 in lumbis · in umbilico.

Apc 18–19 (0016): Et post haec vidi (18:1) · Cecidit, cecidit · Custodia omnis spiritus immundi (18:2) · De virtute (18:3) · Et audivi · Exite (18:4) · Quoniam pervenerunt peccata (18:5) · Reddite illi · Duplicate duplicia (18:6) · Quantum glorificavit se · Et vidua non sum (18:7) · Mors (18:8) · Et flebunt · Reges terrae · Cum viderint fumum incendii ejus (18:9) · Et negotiatores terrae flebunt (18:11) · Auri et argenti (18:12) · Cinnamomum · Similae et tritici · Mancipiorum et animarum (18:13) · Et sustulit · Lapidem quasi molarem · Hoc impetu (18:21) · Post haec audivi · Alleluia (19:1) · Et ceciderunt seniores · Amen (19:4) · Et vox · Pusilli (19:5) · Et audivi quasi vocem (19:6) · Et uxor ejus (19:7) · Et cecidi · Vide ne feceris · Testimonium enim (19:10) · Et vidi coelum apertum · Equus albus (19:11) · Verbum Dei (19:13) · Vestiti byssino albo et mundo (19:14) · Gladius ex utraque parte acutus · Reget eas (19:15) · Et habet in vestimento · Rex regum (19:16) · Et vidi unum angelum · Omnibus avibus · Venite · Et congregamini (19:17) · Et apprehensa est bestia · Vivi missi sunt hi duo (19:20) · Et caeteri occisi sunt · Et omnes aves (19:21) · Ps 32:9 Ipse dixit, et facta sunt.

⚠ Apc 19:5. The Clementine copy on disk prints Ex vox de throno exivit; Migne's lemma Et vox agrees with the ordinary text and no divergence was logged. sources/vulgate/ README.md warns that later upstream corrections are not applied, and a call turning on one letter should say so: this is one, and it is our file that is odd, not Migne's.

Apc 20 (0017): Et vidi angelum · Abyssi (20:1) · Apprehendit · Ligavit eum (20:2) · Et misit eum · Signavit super illum · Et post haec oportet (20:3) · Et animas decollatorum · Et vixerunt (20:4) · Haec est resurrectio · Prima (20:5) · Qui habet partem · Secunda mors · Sed erunt sacerdotes (20:6) · De carcere · Magog (20:7) · Et circumierunt castra sanctorum (20:8) · Die ac nocte (20:10) · Et vidi thronum (20:11) · Et libri aperti sunt · Alius liber · Ex his quae scripta erant (20:12) · Et dedit mare mortuos (20:13) · Haec est mors secunda (20:14).

Apc 21 (0018): Et vidi · Coelum novum et terram novam (21:1) · Sicut sponsam ornatam (21:2) · Vocem magnam de throno · Ecce tabernaculum · Et ipsi populus ejus (21:3) · Et mors (21:4) · Et dixit · Scribe: quia haec (21:5) · Et dixit mihi · Factum est · Alpha et omega · Sitienti (21:6) · Possidebit haec (21:7) · Timidis (21:8) · Veni et ostendam (21:9) · In montem magnum · mons · altus · Ostendit mihi civitatem (21:10) · Lumen · Jaspidis (21:11) · Murum · Portas duodecim · Nomina inscripta (21:12) · Ab oriente portae tres (21:13) · Murus · Et in ipsis duodecim (21:14) · Et qui loquebatur mecum · Mensuram arundineam auream · Ut metiretur civitatem et portas ejus (21:15) · In quadro posita est · Per stadia · Duodecim millia (21:16) · Mensura hominis (21:17) · Et erat · Structura muri ejus ex lapide jaspide · Aurum (21:18) · Jaspis · jaspis · Sapphirus (21:19) · Sardonyx · Sardius · Beryllus · Topazius · Chrysoprasus · Hyacinthus · Amethistus (21:20; Amethistus for amethystus is i/y orthography) · Et duodecim portae · Per singulas · Singulae portae (21:21) · 2 Cor 3:2 Epistola nostra vos estis scripta in cordibus nostris.

Apc 21–22 (0019): Non eget sole · non eget sole, neque luna (21:23) · Et ambulabunt · ambulabunt gentes · In lumine · Reges terrae (21:24) · Et portae (21:25) · Afferent (21:26) · Non intrabit in eam (21:27) · Et ostendit · Fluvium · De sede Dei (22:1) · Et ex utraque · lignum vitae · Lignum vitae · Fructus · Et folia ligni (22:2) · Et dixit mihi · Dominus · Angelum suum (22:6) · Et postquam audissem, et vidissem (22:8) · Et fratrum tuorum (22:9) · Qui nocet, noceat adhuc (22:11) · Beati qui lavant (22:14) · Ego Jesus · quod ego Jesus misi angelum meum testificari vobis haec. Stella splendida. (22:16 — the span elides in ecclesiis and et matutina, an ordinary clip, not a divergence) · Et spiritus · Et spiritus et sponsa dicunt · Et qui sitit, veniat · Qui vult, accipiat aquam vitae gratis (22:17) · Contestor enim (22:18) · Gratia Domini nostri Jesu Christi (22:21).

1d. The re-quotation check (7a″ step 3)

Every lemma marked [var:] was held fixed and every re-quotation of it in the gloss below was read against the fixed rendering. Two of the five markers were confirmed BY the re-quotation, and they are the strongest sites in the range:

• 0739D ortus sui is contradicted four words later by Migne's own umbilicus tuus and non refrenasti. The English keeps both persons as printed and does not harmonize.
• 0744C vide is contradicted by three first-person vidis in the sentences under it. The English renders the lemma as the imperative Migne prints and the gloss as the first person he prints.
• 0742D Exsultate … coeli is the gloss's own re-quotation of the 0742B lemma; both occurrences carry a marker.
• 0749C Plateae (plural) is carried into the gloss's plural pleni, mundi, so the plural is Migne's throughout, not a stray ending.

§2 · THE [cj:], AND THE RIVAL IT BEAT

0745A · Secunda mors … non apprehendet electos, qui in nullo unquam augentur.

Rendered: …shall not lay hold of the elect, who are never in anything increased [cj: augentur; read anguntur, "are distressed"].

augentur is a perfectly good form, so Pattern 12 has nothing to wrap and Pattern 7 requires the English "increased"; Pattern 18 is exactly the gap this falls into — the printed word is real and the faithful English misleads. The sentence's own structure decides it: the clause is appended to non apprehendet electos, a statement about what the second death does not do to them, and anguntur ("are afflicted") continues that thought where augentur changes the subject to the measure of their glory.

The rival, and why it lost. augentur can stand, meaning that the elect's state admits no increase — a real medieval commonplace. It lost on attachment: the qui clause is subordinate to non apprehendet, and an aside about the fixity of glory has no business there, whereas anguntur is the conclusion the main clause is driving at. The marker is additive, so a reader who prefers the rival still has Migne's word and its English on the page.

§3 · REFERENTS SUPPLIED, AND WHERE TWO READINGS BOTH CONSTRUE

Every "he/she/it/they" the Latin does not print was named before it was written. The ones a later reader should be able to check:

1. 0739D · quos attrahit ad se luxuria. luxuria is nominative or ablative, and both construe. Taken as ablative of means — "whom she draws to herself by wantonness" — because quae regnat has just set the harlot as subject and no new nominative is marked; ad se then points at her, which is what the imagery wants. The rival: luxuria nominative, "whom wantonness draws to itself". Nothing in the sentence excludes it. Not adjudicable from the file; recorded here rather than chosen silently.
2. 0740B–D · the beast and the devil. The lemmata are feminine (Ascensura est, bestia), the gloss masculine (recipiet, ascendet, habebit), because the gloss has identified the beast as diabolus. The English follows each as printed: "it" for the beast-lemmata, "he" for the devil in the comment. Not a defect; the identification is Migne's own.
3. 0741B · Carnes fornicariae manducare dicuntur. Unexpressed subject supplied as "they" = the ten horns of Apc 17:16 (hi odient fornicariam … carnes ejus manducabunt), which is the verse the lemma comes from, not anything in the sentence.
4. 0741D · Reges terrae … dolebunt, quod nunquam similes illis fuerint. illis is not determinable from the sentence: it may be the saints who never fornicated, or the kings who governed their earthliness well. Rendered "like to them", which keeps the Latin's own openness. Flagged, not resolved.
5. 0743B · In veteri lege non prohibuit se adorari. Subject is the angel (Apc 19:10, Vide ne feceris), not Christ; videns super se exaltatum hominem confirms it — the man exalted above him is the reason he now refuses. The English says "he" throughout and the antecedent is the lemma above.
6. 0744C · peccasse ostendit dum se immunem a peccato ille diabolus occidit. se is used loosely for eum: the subject of ostendit is Christ, and se is Christ as object of occidit. Read reflexively to diabolus the sentence says the devil killed himself, which it cannot mean. Rendered "he showed him to have sinned, in that that devil slew him who was free from sin", which keeps the Latin's own two "him"s unresolved on the page.
7. 0744C · Abyssus tenebrosa vocantur corda impiorum; vel ipse diabolus, quos Dominus… quos is plural after a singular diabolus; it reaches back past it to the ungodly. Rendered "whom", carrying the anacoluthon rather than repairing it.
8. 0750C · ipse etiam dat eis qui pararent eos quibus ipse reddat. Three pronouns, no antecedent printed for any of them. Rendered as literally as English allows; the clause is obscure in the Latin and is left obscure. Flagged.

§4 · PLATE-READ CANDIDATES AND CONSTRUCTIONS CARRIED AS PRINTED

No [sic:] was fired anywhere in this range, and that is deliberate — the ship test (brief addendum B) says an unread candidate is safe iff it fires no marker, and none of the candidates below can be settled from the file.

Plate-read candidates, in priority order

1. 0743C lignis for ignis and 0744C vide for vidi — see §1a. Both are one-letter differences on a real word in a scripture lemma; both are the shape of brief §4's normalization/corruption class, which is invisible from the Latin alone. A plate read settles both cheaply.
2. 0749C Vitrum per lucidum (Apc 21:21 tamquam vitrum perlucidum). A word split across a space is Pattern 10's shape and brief addendum A says a non-word raised as a plate candidate has been right every time — but "right" there meant our digitization's corruption, not Migne's, and a [sic:] publicly blames Migne. No marker. Rendered by sense, "Glass thoroughly clear.", which is also what the gloss requires (nullas sordes in se celantes). Note too that perlucidus written open is ordinary medieval practice and may be Pattern 9, not a defect at all.
3. 0748D obscurascit. A hapax in the whole Latin corpus (one occurrence, here); the standard inchoative is obscurescit. But it is a well-formed inchoative on obscurus, and addendum A is explicit that corpus frequency is near-worthless against a real word. Rendered "grows dark", no marker. Raised only so a plate reader can glance at it while reading the topazius paragraph for something else — which is how the normalization class has been found twice (addendum C).
4. 0741B tormentabuntur. Also a corpus hapax. tormentare is well attested medieval Latin; rendered "shall be tormented", no marker, recorded as a checked negative.
5. 0748D vel minori meritis fiunt. minori where minores is wanted. Rendered "or become less in merits". A real form, so no marker; worth a look if the page is opened.

Constructions carried as printed, not repaired

• 0740D–0741A · Adam, si in obedientia perseverasset, sensus corporis … verterentur. A nominative left hanging. Carried: "Adam, if he had persevered in obedience, the senses of the body would have been turned to no pleasure". Not smoothed into "Had Adam persevered".
• 0745A · Et non dico vivent et regnabunt in futuro, sed etiam in hoc praesenti. The non … sed etiam correlative with its solum elided. No "only" was supplied; the English reads "And I say not, they shall live and reign in the time to come, but also in this present time", which is as unbalanced as the Latin.
• 0748A · ut sit defensor aliorum … ut sit defensor aliorum. The same purpose clause twice in one sentence, eleven words apart. Both rendered; no [d:]. Pattern 11 is for a set run repeated; here a full clause intervenes and the second reads as a resumptive purpose clause, which a compositor's doubling would not. Recorded as a possible dittography a plate read could settle.
• 0744C · pseudo apostolorum set open, against pseudoprophetis set closed at 0741C in the same work. Rendered "false apostles" / "false prophets". No marker: Pattern 9 covers period orthography, and the open form is common.
• 0739D · quarum semen est umbilico — ablative of place with no preposition, against quorum seminarium in lumbis est in the same sentence. Rendered "whose seed is in the navel". Not a defect; recorded for the asymmetry.
• 0751B · the chapter's last sentence has NO terminal stop. …per meam resurrectionem in mane factam — Migne prints nothing after factam. Pattern 8: never supply. The English ends the same way. This is the range's most visible oddity on the page and it is his.
• 0750D–0752A · the anchor sequence jumps 0751B → 0752A, skipping 0751C–D. Runbook false positive 1 (band arithmetic): the A–D marks are positional quarter-guides, gaps are normal, not reported as a defect. Recorded only so a reader is not surprised.
• 0748B · Migne glosses no seventh stone. Sardius is the sixth and Beryllus the eighth, with chrysolithus (Apc 21:20) passed over in silence. The ordinals he does print — fourth, fifth, sixth, eighth, ninth, tenth (in denario numero), eleventh, twelfth — all agree with the Clementine's order, so nothing is displaced; the seventh is simply not commented. Abridgment, not loss. No [ed:].

usque ad — decided per occurrence (brief §D)

Four occurrences in 0018 (0748B ×3, 0748C ×1), all four the abridgment formula, all inside Bede excerpts opening etc., and all rendered as far as on the settled precedent (8985 §371, 8976 §191, from 8978 §205). No verb was supplied to complete any of them.

Two look-alikes in 0019 are the words' own, not the formula, and were rendered as ordinary prose: usque ad finem mundi (0749D, "even to the end of the world") and duxit filios Israel usque ad flumen Jordanis (0750B). One more in 0016: Peccata perveniunt usque ad contemptum Dei (0741D). Three ordinary uses, four formulas, seven occurrences, all decided in place.

§5 · THE FOUR TESTS, RUN

Test 1 — no [sic:] may swallow its word. Vacuous here: the range fires no [sic:]. The strip-and-read pass was still run over the [var:] and [cj:] sites, since the same failure is available to them. All six sentences read aloud complete with every marker removed; the [cj:] site keeps its verb ("are never in anything increased"), and each [var:] sits beside a lemma that is already fully Englished.

Test 2 / 2a — name the host, prove it from structure. Every printed negative in the range was walked. The population 2a fires on — a negative on a verb with a quantifier or correlative nearby — has eight members here, and all eight are governed by an explicit correlative, which is the only reason they can be called clean:

Simple negatives carried without incident, named so they are checkable: nihil timet · nec Deum diligant, nec poenas timeant (0740A) · ne desperetis (0740B) · nisi doctis (0740C) · amplius non habebit (0740D) · ad nullam voluptatem (0741A) · nunquam similes (0741D) · et vidua non sum (0742A) · nullus possit resistere (0743C) · nequeunt … nec possunt (0743A) · nisi bene pugnaveritis (0744A) · nunquam ad inferos descendere (0744D) · Caeteri vero mortuorum non vixerunt · omnes perfecti esse non possunt, nec in eadem claritate · in nullo unquam (0745A) · ne pro velle saeviat (0745B) · non fecerat mentionem · se non praedestinatos · quae agere noluerunt (0745C–D) · dubitabilis non est quin (0746A) · in nullo offendentes, quod non esse potest · non erunt ibi lacrymae … non erit causa (0746B) · nihil restare (0746B) · una non debet aliam excedere · ne nimis justus (0747C–D) · fides non marcescens (0748A) · quae non finitur (0749A) · ne fures intrent (0749D) · qui non (0750A) · Non sunt … alieni putandi (0750C) · ne ideo cesset (0751A) · non est labor (0752A). Count in, count out: no negative added, none dropped.

Test 2b — anchor position. All 50 anchors were re-checked against the word they stand beside in the Latin, not merely for presence. Three sit mid-construction and were kept there: Et [0742A] vidua non sum (inside an italic lemma, reproduced as "And [0742A] I am not a widow"), crucifixus [0748C] est ("Christ was crucified [0748C] on the sixth day"), and cum aere [0749B] mutatur. Four are the no-space form Migne sets — [0740A]VERS. 2.--, [0742B]VERS. 12.--, [0744A]VERS. 17.--, [0745A]VERS. 5.--, [0745B]VERS. 7.--, [0752A]VERS. 17.--, [0752B]VERS. 21.-- — reproduced closed up, as printed.

Test 3 — punctuation by band, never by total. This is the check that returned sites, and it returned twenty-three: every one an English semicolon or colon where Migne prints a comma, or the reverse. All twenty-three were repaired; the range now matches band for band on ;, : and ?, verified per chunk after the edits. ⚠ The totals would have hidden most of them — 0018 alone had a supplied colon at 0747C and a destroyed colon at 0747B, which is the Tobit failure exactly. No comma→semicolon sweep was made: Migne prints 33 semicolons in these five chunks himself, so the Tobit exception does not apply to this book.

Both ? marks are direct questions and ride as printed: Et isti, quare damnabuntur? (0743B) and Cum sim malus, quare me poeniteret? (0751A). No ? was supplied and none deleted, and no 8b recasting was needed — the indirect questions in the range (mirantur … quare Deus permittit illos exaltari, 0740C; non fecerat mentionem corporum, si deberent glorificari vel puniri, 0745C; Ne quaereret aliquis: Quid post illam innovationem futurum est, 0746B) all close with Migne's own comma or full stop, so nothing had to be moved behind a colon.

Test 5 — does the English ASSERT what he asserts? Read back paragraph by paragraph against the Latin with the counts already clean. Two things came out of it and neither is a containment failure:

• 0741B · perdet is rendered twice, two different ways — "whom he shall destroy" (quos perdet, of persons) and "this he shall lose" (hanc perdet, of the power). Both are proper senses of perdo and the object decides which; recorded because a reader comparing the columns will see one Latin verb against two English ones.
• 0741B · Desolatam facient / Solatium diaboli. The gloss is a pun: desolatam is glossed by solatium. The English keeps "desolate" against "solace" so the pun survives. Nothing is claimed about whether Migne's source intended it.

Em-dashes: none in this range. Nothing was restructured to avoid one and none is reported.

§6 · SMALLER RENDERING CALLS, RECORDED

• 0739C Angelis (from Apc 17:1 de septem angelis) rendered "Of the angels." — a case-form fragment, Ruth convention 3, the phrase it stands for rather than the bare noun. Meretricis (0739D) likewise "Of the harlot." Neither is a stump-vs-expanded inconsistency with the true one-word lemmata in the range (Auro · Ebriam · Fuit · Carnes · Mors · Amen · Prima · Magog · Lumen · Aurum · Jaspis · the eight stone names · Fluvium · Dominus · Fructus), which have nothing elided and are rendered as they stand. Naming the two classes before contrasting them, per the brief's closing rule.
• 0742D antonomastice rendered "by antonomasia" — a rhetorical term with a settled English form, not italicized as untranslatable Latin.
• 0745B detectos rendered "the uncovered", to hold the detectos / tegentes play the gloss builds Magog on.
• 0749A pan · Chryson · prasos are Greek words Migne sets roman, not italic; the English keeps them roman. Italicizing them would have added three italic spans the Latin does not have.
• 0750B · one connective supplied. …in quibus reficimur ultra flumen, id est, in futuro habebimus ipsum praesentem. Migne prints no mark between reficimur and ultra flumen, where the sense turns from circa flumen to ultra flumen. Rather than supply a mark (Pattern 8) the English supplies the word "and": "in which we are refreshed, and beyond the river…". A word, not a mark; recorded because it is the one place the English adds anything.
• Pattern 17. tibi / thou oughtest / si polis, obscuras are singular and take thou (0740C, 0742B, 0748D, 0750C); vobis, vos, cavete, providete vobis, debetis are plural and take you (0739C, 0742D, 0743A, 0744A, 0752A). The two plural scripture imperatives keep "ye" under the scripture-echo licence: Exsultate → "Rejoice ye" (0742B, 0742D), Venite → "Come ye", Et congregamini → "And be ye gathered together" (0744A).
• Pattern 4 [f:] does not apply to this work (runbook false positive 4): it is a lemma-and-gloss commentary and its [n:] citations are Migne's inline parentheticals. None was tagged.
• Pattern 15. All ten [n:] in the range are citations or attribution sigla — (Job. XL) · (Ezech. XVI) · (Joan. IV) · (Psal. XXXII) · (II Cor. III) · (Phil. III) · (BED.) ×2 · (ID.) ×2. No note in this range is editorial prose, so no [nt:] was made. A checked zero, not a skip.

✅ PLATE READ, 2026-08-28 — the two single-letter candidates of §1a, settled OPPOSITE ways

Both were flagged here as needing one letter decided at the plate. PL 114 p. 377 carries both, and it splits them:

• 0743C lignis was OURS. Migne prints ignis; the [var:] is withdrawn and the TEI patched. See the row in §1a.
• 0744C vide is MIGNE'S. He prints the imperative Et vide sedes, and the [var:] against Clementine Et vidi sedes stands, confirmed. His own gloss below re-quotes vidi three times (Ligato diabolo, vidi solutos fideles … Vidi in tempore ligationis … vidi statim in gloriam intrare), so the plate contradicts its own lemma — exactly what this edition exists to show, and not ours to tidy.

⭐ The lesson, and it is the argument for reading plates at all: these two were indistinguishable from the file. Same class, same shape, same evidence, flagged by the same reader in the same breath — and the answers were opposite. Corpus frequency and Vulgate collation could not have separated them; only the plate did.

PL 114, cols 0709C–0719D (Prologue + Argument + CAPUT I–IV). Range file: this covers chunks 0000–0004 only. cruces.md is the four-way merge and is written once at the end; nothing here is to be duplicated into it by hand before that merge.

Attributed by Migne to Walafrid Strabo; in fact Anselmus Laudunensis et schola. Conventions followed: src/english/8968/cruces.md §1 (Ruth, the Glossa pilot) and GLOSSA-NT-AGENT-BRIEF.md.

Apparatus taken in this range: 18 [var:] · 1 [cj:] · 0 [sic:] · 0 [d:] · 0 [ed:] · 0 [nt:]. Marker parity against the Latin twins: column anchors 3/14/7/8/8, [n:] notes 0/1/4/7/2, italic spans 0/78/39/50/42, question marks 0/0/2/0/0 — all exact. Punctuation compared band by band (colons, semicolons, ?, !, guillemets) across the verified anchor partition: zero mismatched bands after four supplied semicolons were removed (see §4). Zero guillemets in either language.

§1 · THE 7a″ DELIVERABLE — every lemma in cols 0709C–0719D, collated

Collated against sources/vulgate/clementine-flat.txt, cited from sources/vulgate/clementine/Apc.lat etc. This section lists what was checked, not only what was found — the agreements are the evidence that the divergences were looked for.

1a. Divergences that took a [var:] — 18, one per row

Distribution: 5 in chunk 0001, 4 in 0002, 4 in 0003, 5 in 0004.

1b. Divergences examined and DECLINED, with the reason — the marker would have been noise

• 0709C Joannes septem Ecclesiis quae sunt in Asia. Quam dedit. — one italic span joining Apc 1:4's opening to Apc 1:1's Quam dedit. Both agree with the Clementine verbatim; the join is the gloss's own citation practice, not a textual variant. No marker; the span is kept as one span in the English, as Migne sets it.
• 0709D (prologue) Ego sum Alpha et Omega, initium et finis — agrees verbatim with Apc 21:6. It is not Apc 1:8 (principium et finis), and reading it as a divergent 1:8 would have been the error. Checked clean, no marker. Note it is not italicized in the Latin and is therefore not italicized here.
• 0712A Alpha, principium. / Omega, finis. — Apc 1:8 reads alpha et omega, principium et finis. Migne re-pairs the four words into two glossable couples. A lemma-splitting device, not a variant reading.
• 0711D Videbit omnis terra talem. — Migne himself labels it Alia translatio habet. His own apparatus has already adjudicated it; a [var:] would restate what the page says.
• 0713B aurichalco Libani — likewise introduced In alia translatione habetur. Declined for the same reason. Not italicized in the Latin; not italicized here.
• 0713C Et ex ore. vs Apc 1:16 et de ore. Preposition only; nothing in the gloss turns on it. Cruces, no marker.
• 0716A Dedistis mihi manducare et bibere vs Mt 25:35 dedistis mihi manducare : sitivi, et dedistis mihi bibere. A condensation, not a divergent wording: every word Migne prints stands in the verse in that order. Pattern 14 asks for a sense-changing divergence and there is none. ⚠ Note the 7a″ trap here and that it was avoided: manducare is rendered "to eat" because that is what manducare means, not by the route that produced the 7914 manum → "to eat" defect.
• 0716A charitatem autem non habeam, nihil sum vs 1Cor 13:2 caritatem autem non habuero. Present subjunctive for future perfect; charitatem/caritatem is orthographic. Not material. ⚠ But see §3 on the siglum, which is a separate finding.
• 0717D Ecce ego sto ad ostium et pulso vs Apc 3:20 Ecce sto ad ostium. An added emphatic pronoun; the gloss (Exspecto, vel inspiro) does not use it. Cruces, no marker.
• 0714A Et angelo. at Apc 2:1, where the Clementine has bare Angelo Ephesi ecclesiae scribe. Migne's Et matches his own 2:8 and 2:12. Cruces, no marker.
• 0715A Et dabo ei. vs Apc 2:17 et dabo illi. Pronoun only.
• 0715A Abscisus est lapis sine manibus vs Dn 2:34 abscissus est lapis de monte sine manibus (and Dn 2:45 in the other order). de monte dropped, Abscisus with one s. The gloss argues only from sine manibus, so nothing turns on the omission. Cruces, no marker. ⚠ Migne's own siglum reads (Dan. I) for what is Daniel 2 — passed through verbatim, per convention 4; it is not ours to correct.
• Orthographic only, all checked and all declined: tanquam for tamquam (five sites, 0713A–0713B), coelo for caelo (0716D, 0718A), charitatem for caritatem (0714C, 0716A), Pathmos for Patmos (0709C, 0712B — Migne is consistent in both places), aurichalcum for auricalco (0713B), chrystallus for crystallo (0718D).

1c. Lemmata checked and found to agree with the Clementine EXACTLY

Apc 1: Apocalypsis Jesu Christi · Quae oportet fieri · Cito · Qui testimonium perhibuit · Verbo Dei · Beatus · Qui legit · Servat ea · Septem ecclesiis · septem · In Asia · Gratia vobis · Qui est et qui erat · Throni · Testis fidelis · Primogenitus mortuorum · Princeps regum terrae · et sacerdotes · Ecce venit · Cum nubibus · Et videbit eum · Omnis oculus · Plangent · plangent se · Ego sum · Ego Joannes (×2) · Frater vester · in Christo · Propter verbum Dei · In Dominica die · Audivi post me vocem · Quod vides, scribe · Epheso · Et conversus sum · Septem candelabra · Aurea · Similem Filio hominis · Podere · Praecinctum ad mamillas · ad mamillas · Capilli erant candidi · Nix · Oculi · Et pedes · In camino ardenti · Et vox illius tanquam vox aquarum multarum · Stellas · Sicut sol lucet in virtute · Cecidi · Tanquam mortuus · Dexteram suam · Noli timere · Sum vivens · Et habeo · Mortis et inferni · Quae oportet fieri post haec · Sacramentum.

Apc 2: Ephesi Ecclesiae · Septem · Non potes sustinere (×2) · Tentasti eos qui se · Sustinuisti propter nomen meum · Et non defecisti · Sed habeo adversum te quod charitatem · Movebo candelabrum · Facta Nicolaitarum · Qui habet aurem · Smyrnae · quasi myrrha · Haec dicit primus (×2) · Aliquos ex vobis · Ut tentemini · Diebus decem · A morte secunda · Rhomphaeam · Non negasti fidem meam · Balaam · Doctrinam Nicolaitarum · Manna · Et in calculo · Scriptum · Novi opera tua · Fornicari · Non mittam.

Apc 3: Et angelo Ecclesiae Sardis · Scio opera tua · Esto vigilans · Non enim invenio opera tua plena · In mente ergo habe · Audieris · Non inquinaverunt vestimenta sua · Non delebo · Induravit Dominus cor Pharaonis (Ex 10:20, less its opening Et) · De libro vitae · Confitebor nomen ejus · Venite, benedicti Patris mei (Mt 25:34) · Haec dicit · Claudit · aperit · Ecce dedi coram te · Ostium · Judaeos · Hunc nescimus (Jo 9:29, a fragment of hunc autem nescimus unde sit) · Mentiuntur · Ecce faciam illos ut veniant · Scient · Servasti verbum patientiae meae · Ab hora tentationis · Tentare habitantes in terra · Ecce venio cito · Columnam in templo · Non egredietur · Novae · De coelo · Testis · Et verus · Neque frigidus es, neque calidus · Utinam frigidus · Incipiam te evomere · Ex ore meo · Dives sum · Suadeo tibi emere a me aurum ignitum probatum · Non appareat confusio · Collyrio · Sedere mecum in throno meo · Sicut et ego vici et sedi cum Patre meo in throno ejus.

Apc 4: Post haec vidi · In coelo · Tanquam tubae loquentis mecum · Ostendam tibi · Statim fui · Et ecce sedes · Sedens · Qui sedebat · Similis erat aspectui · Jaspidis et sardinis · Smaragdinae · Viginti quatuor seniores · Coronae · De throno · Septem spiritus · In conspectu · Mare vitreum · In medio · Sedis · Quatuor · Animalia · Plena oculis · Leoni · Et quatuor animalia singula eorum habebant alas senas · Alas senas · In circuitu et intus · Plena sunt oculis · et in circuitu · Qui venturus est · Et benedictionem · Et adorabant · Dicentes · Creasti omnia · Propter voluntatem tuam · Erant · Creata sunt.

Not scripture, so not collated (Hebrew-name etymologies and the gloss's own re-cuts): voluntas mea · lapsus profundus · consilium (0712D, 0714B) · voluntas (0714B) · canticum (0714D) · principium pulchritudinis (0715D) · plena sunt oculis intus (0719B, Migne's own re-division of Apc 4:8 for the gloss, not a variant).

§2 · THE ONE CONJECTURE — 0719B, Augusti

Migne prints, of the sixth wing of the four living creatures:

sexta, quorumcunque decreta, ut Augusti, Gregorii, Ambrosii, per quos Ecclesia alta petit

Augusti is a perfectly good Latin word — the genitive of Augustus — so nothing is broken and Pattern 12 has nothing to wrap. But the faithful English asserts something false: it makes the decrees of the emperor Augustus a wing of the Church, standing between Gregory and Ambrose in a list of doctors, and it is decreta of the Fathers a quibus praedicatores accipiunt totius praedicationis fundamentum. Pattern 18 exactly.

⭐ The corpus is its own control and it is decisive here: this same work prints Augustinus at chunk 0019, and Augusti nowhere else. One occurrence of the truncated form against a full one in the same plate, in the same company of Fathers.

Marker taken, additive, Migne's word kept and rendered: of Augustus [cj: Augusti; read Augustini, "of Augustine"]

The rival considered and rejected: that Migne means the civil decreta Augusti — Roman imperial law as a source of ecclesiastical discipline, which is not absurd in itself. It loses on the sentence's own structure: the three genitives are a single series governed by per quos Ecclesia alta petit and a quibus praedicatores accipiunt … fundamentum, and neither predicate is sayable of imperial legislation in a list whose other two members are Gregory and Ambrose.

§3 · CRUCES — logged, no marker fired

Each of these was examined and left as printed under Pattern 7. A crux asserts nothing about the plate; where I could not settle a thing from the file, the crux is where it stays.

3a. Referents supplied, and what decided them

• 0709C (prologue), commendasset / asciverat / tradidisset. Three unexpressed subjects in one sentence, and they are not the subject of the verb before them. recumberet is John; commendasset, asciverat and tradidisset are all Christ — matrem propriam and virginem are his mother, and ad crucem astanti soli is John in the dative. Rendered so. This is the class the brief names as dominant; naming it here is the record that it was decided rather than defaulted.
• 0709C, ab eodem: the Apocalypse praeostensa by Christ, not by John — a Christo electus atque dilectus is the sentence's own antecedent for idem.
• 0714A, requirit: Id est episcopo scribit, de manu cujus peccata subditorum requirit, et sine cujus consensu … non praesumit ipse Joannes. Two readings both construe: the subject of requirit is God/Christ (who requires the sins at the bishop's hand), or John. ⚠ The explicit ipse Joannes in the second clause is evidence for the first — Migne names John only where the subject changes. The English keeps a bare "he", which is exactly as under-determined as the Latin, so nothing is decided in the text; this entry is the record that the ambiguity is real and was not resolved silently.
• 0714B, (qui lapsi fuerant): the parenthesis attaches to the implicit subject of corrigantur, i.e. those to be corrected, not to quorum operibus Deus delectatur in the clause before. Rendered "that by the example of these they (who had fallen) may be corrected", keeping Migne's parenthesis in his own position.
• 0718B, quos hortatus est ad pugnam: a plural relative with no plural antecedent in its sentence (de quo singular precedes it). Rendered literally as "whom he has exhorted to the fight". A dative quibus would construe; that is a conjecture and it stays here. ⛔ Not a [cj:]: Pattern 18's bar is that the English asserts something false, and this English asserts nothing at all — it is incoherent, and 18a's exception covers only a negation or a correlative gone wrong. This one is neither.

3b. Grammar the plate prints and English cannot mend

• 0714A, ministri ejus sunt infernus: plural subject, singular predicate nominative. Rendered as printed: "his ministers are hell, in whom he has a place." No English exponent for the mismatch, and no emendation to make.
• 0714D, etiam canticum in quibus delectatur: singular canticum, plural in quibus. Rendered "a song also, in which things he delights", carrying the plural.
• 0712C, quae monuit communiter mittere ecclesiae Asiae … per quas: ecclesiae is singular by form and per quas wants a plural (a dative plural would be ecclesiis). ⭐ English "by which" carries no number at all, so the rendering reproduces the Latin's own indeterminacy without choosing. Logged rather than fixed.
• 0718B, In quibus sedet Deus et judicat, et etiam subditos: sedes at Apc 4:2 is singular; In quibus is plural. English "In which" is number-neutral. subditos is an accusative object gapped off judicat; the English repeats the verb ("and judges also"), which is ordinary English gapping and adds no claim.
• 0712B–C, in fervorem fidei / in perditionem Sodomae: in + accusative where sense wants time or occasion. Rendered "unto the fervour of faith" / "unto the destruction of Sodom", literally, rather than smoothed to "at".

3c. Sentences that do not construe, or construe against their own neighbour

• 0711A, Christus dedit haec mihi, beatus autem quia me acceperit. Elliptical: no subject for beatus, and a perfect subjunctive with no governing verb. Rendered as printed, "but blessed because he has received me." No conjecture offered; there are at least two (beatus autem [est] qui…, or a first-person sum) and nothing in the column chooses between them.
• 0711A, quia nec longa mora laboris, et praemia vicina. A nec … et pair where sense wants nec … et to be neque … sed or a bare asyndeton. The nec is carried: "because neither is there a long delay of labour, and the rewards are near." It reads oddly. That is the point of Pattern 7a — I have not converted nec to "there is no" plus a smoothed second limb, which would hide the plate's own awkwardness.
• 0717C, Suadeo tibi emere a me aurum ignitum probatum. Quia talia opera non perducent ad salutem. The gloss contradicts its own lemma: the lemma counsels buying tried gold and the gloss says such works do not lead to salvation. Rendered as printed, negation intact. ⚠ The likeliest explanation is that the gloss belongs to a neighbouring lemma (Apc 3:17's dives sum / nudus) and has slipped; but that is an editorial rearrangement of Migne's page and Pattern 7 forbids it in the text. Flagged here for a blind reader — this is the most conspicuous sense-defect in the range.
• 0712C, Salomon, non reservaturus, sapientiam nocte suscepit. A bare future participle with no object. Rendered "Solomon, not about to keep it, received wisdom by night", the "it" supplied from sapientiam in the same clause and flagged here as supplied.
• 0717A, quidam Eliae monitis legem spiritualiter impleturi, hostem credunt victuri. Two future participles, the second in an accusative-and-infinitive with the copula elided. Rendered "certain, being about to fulfil the law spiritually by the warnings of Elias, believe that they will vanquish the enemy." The construction is elliptical in the plate and is left elliptical.
• 0718C, Per ruborem ejusdem — a redness predicated of an emerald, two clauses after its green is expounded. Real words, coherent syntax, odd content. Rendered as printed. ⚠ Raised under brief §4 only as a negative: nothing here reads suspiciously tidy, and no plate read is requested.

3d. Migne's own sigla, passed through and NOT corrected

Convention 4 governs: the sigla ride verbatim and passing one through is not endorsing it. Three in this range are demonstrably wrong and are recorded rather than repaired.

• 0715A [n: (Dan. I)] on a quotation of Daniel 2 (2:34 / 2:45).
• 0716B [n: (Jac. II)] stands after charitatem autem non habeam, nihil sum, which is 1 Cor 13:2. The James 2:10 matter is the preceding quotation in the same sentence, so the siglum has either slipped one clause or is meant to cover both.
• 0716B [n: (Joan. XIV)] on a quotation whose second half is Jo 16:23 — the conflation already marked [var:] at that column.

3e. Words examined for the digitization-alteration class (brief §4), and the finding

• 0712D, Ephesas. Migne prints Epheso. as the lemma, then glosses Ephesas voluntas mea interpretatur. ⭐ The work's own control refutes any tidying instinct: at 0714B the same etymology is introduced as Ephesus secundum bene persistentium partem — so the book spells the name Ephesus two columns away and Ephesas only here. Rendered as printed ("Ephesas is interpreted…"), because Pattern 7 governs a proper-name form exactly as it governs any other. Not a [sic:]: Ephesas is a well-formed token, not a non-word, and Addendum A is explicit that a real, well-formed word is Migne's until a plate says otherwise. A plate read of col. 0712D would settle it and is cheap.
• ⭐ Recorded negatives. deservit / deserit, the standing confusion pair for this series: grepped across the whole of 8977 and it occurs zero times, in either spelling. Nothing to check. Likewise no split or fused runs (Pattern 10) anywhere in cols 0709C–0719D, no ( sic ) of Migne's own, no dittography, and no non-word of any kind — which is why this range fires no [sic:] at all. A checked zero is worth having; the alternative is a reader who cannot tell this range from one nobody read.

§4 · SELF-AUDIT — the tests the brief requires, and what each returned

• Test 1 (a [sic:] must not swallow the word it marks). Vacuously clean: no [sic:] in this range. The [cj:] at 0719B is additive by construction — "of Augustus" stands in the running text and the bracket is beside it, so stripping every marker leaves a readable sentence. Stripped and read aloud; no stranded article or preposition anywhere in the five chunks.
• Test 2 / 2a (name the host of each negative). All 84 printed negatives in chunks 0001–4 walked one by one, each mapped to the single Latin word it attaches to. Five sat in the trigger population (a negative on a verb with a quantifier or correlative adjacent) and each was proved from the sentence's structure, not from what read well:
• 0711D quia non tantum dolebunt de ipso tormento, quam quod repellentur — host is tantum, proved by the quam correlative. English: "shall not grieve so much over the torment itself, as that…"
• 0712A Ne a me hoc dictum, sed ab ipso Christo credatis — host is a me, proved by sed ab ipso Christo, not credatis. English negates the agent, not the believing.
• 0714C Hoc non fecisti pro humana laude, sed propter nomen meum — host is the prepositional phrase, proved by sed.
• 0715C non enim prius est charitas quam fides — host is prius, proved by quam.
• 0717B quia neque omnino fidem ignoras, neque aperte negas acceptam — the two adverbs omnino ‖ aperte are the pivot, so the scope is "not wholly ignorant" and "not openly denying", which is what makes the man lukewarm rather than either cold or hot. Reading the first as "thou art not ignorant of the faith at all" would collapse the contrast the whole gloss is built on.
• 0717B Non optat eum frigidum simpliciter, sed talem de quo major spes habeatur — host is frigidum simpliciter, proved by sed talem.
• 0718B Ubi tantum boni sunt, non est opus judicare — ⚠ here tantum is in the protasis and is not the host; the negative is on est opus, and the ubi … ubi pivot proves it. Recorded because it is the same shape as the trigger cases and resolves the other way.
• Test 2b (anchor POSITION, not presence). All 40 anchors in the range compared against the word they stand beside in the Latin, not merely counted. All 40 sit at the same point in the construction. Four stand inside an italic lemma and were kept inside it: Ut [0715A] tentemini., …qui loquitur [0713D] in vobis…, Si videris mulierem ad concupiscendam [0715D] eam…, Ecce ego sto ad ostium et pulso. Si [0717D] quis audierit,. Four stand hard against a VERS. address with no space ([0714D]VERS. 7.--, [0716A]VERS. 2.--, [0717B]VERS. 14.--, [0718B]VERS. 2.--, [0718D]VERS. 6.--) and are reproduced that way. One was moved during the audit: 0719A, first set after "gentle" and re-set immediately after "are", which is the word sunt it stands before in the Latin.
• Test 3 (punctuation BY BAND, never by total). Built over the verified anchor partition, markers stripped from the English first. It returned four supplied semicolons where Migne prints a comma — 0711A (Asia est elatio, Ecclesia ergo), 0711B (nemo mala intellexerat, ponit), 0712D (consummatur, et haec), 0714D (delectatur, myrrha) — all four repaired to commas. ⚠ Reported honestly: in this range the totals did NOT cancel — chunk 0001 stood at 16 semicolons against Migne's 13 and chunk 0002 at 3 against 2, so a total-only check would have caught these four too. The band test is still what located them: it named the four columns instantly instead of leaving four sites to be found by re-reading 1,563 words. Final state: zero mismatched bands on : ; ? ! « ». ⛔ No comma→semicolon sweep was attempted in the other direction: this book prints semicolons freely (27 in the range), so the Tobit exception does not apply here.
• Test 5 (does the English ASSERT what he asserts). Run last, counts already clean, each paragraph read back against its Latin. It produced §3c's list — the four sentences that do not construe are recorded there rather than repaired, and 0717C is named as the site a blind reader should go to first.

§5 · CONVENTION CALLS MADE IN THIS RANGE

• Fragmentary lemmata expanded per Ruth's convention 3 at two places, and only two: Tentasti eos qui se. → "Thou hast tried them who say themselves to be apostles" and Sed habeo adversum te quod charitatem. → "But I have against thee, that thou hast left thy charity." Both stumps end mid-clause and are not English. ⚠ Note what was not done: neither expansion imports a word where Migne diverges from the Clementine, and Sed habeo adversum te quod charitatem. deliberately does not carry the Clementine's primam ("thy first charity"), because Migne's fragment stops before it. ⛔ Complete one-word lemmata (Nix. · Oculi. · Manna. · Judaeos.) are a different class and are rendered as they stand; the two are not contrasted here.
• Et videbit eum. (0711D) rendered "And shall see him." — a verb phrase with its subject deliberately withheld, because Migne makes Omnis oculus the next lemma and glosses it separately. Expanding to "And every eye shall see him" would have swallowed the following lemma and is precisely the conformation 7a″ forbids. Logged because the English is visibly incomplete and that incompleteness is Migne's.
• Non possumus quae vidimus non loqui (0716C) → "We cannot but speak the things which we have seen." English "cannot but" is the exponent of the Latin double negative, not the loss of one; both negatives are present in the idiom. Recorded because a negation-counting check will see only one.
• Pattern 17. The addressee of CAPUT II–III is the angel of each church, singular, and takes thou throughout; the plural passages (Apc 2:10 Aliquos ex vobis … ut tentemini, and the Apostle's vos at 0712B, 0713C, 0714A) take you. The distinction is Migne's morphology, not a register choice.
• VERS. n.-- passed through verbatim everywhere, per convention 1. CAPUT PRIMUM. → CHAPTER ONE., CAPUT II./III./IV. → CHAPTER II./III./IV., matching Ruth's shipped English. PROLOGUS HIERONYMI. → THE PROLOGUE OF JEROME.; ARGUMENTUM. → THE ARGUMENT., after Ruth.
• Pattern 4 [f:] NOT applied — this is a lemma-and-gloss commentary, which the runbook's false-positive 4 excludes by name.
• Column-band gaps not reported. 0709C → 0709D → 0710C skips 0710A–B, and 0709C reappears as the colContext of three later chunks in the lemma brief. Runbook false-positive 1: the A–D marks are positional quarter-guides, not four per column. Named here only so a later reader can see it was recognized, not missed.
• Ratio warn on chunk 0000 (1.72×). verify-english.mjs flags it. The chunk is 188 words of Jerome's prologue — ablative absolutes, gerundives and a triple ut-chain — which unpack long in English. Checked for padding clause by clause and none found; nothing is summarized and nothing is doubled. Recorded rather than trimmed, because trimming would cost content.

Range: CAPUT V–IX, cols 0719D–0729A (colFirst 0720A → colLast 0729A). One translator, five chunks (3,640 Latin words). verify-english.mjs clean.

Apparatus taken in this range: 2 [sic:] · 5 [var:] · 0 [cj:] · 0 [d:] · 0 [ed:] · 0 [nt:] · 2 [n:] (parity with the Latin twins). Anchors 37/37, verbatim, in order, and each checked against the WORD it stands beside in the Latin (test 2b). Question marks 0/0 in both columns. Punctuation compared band by band (test 3): zero mismatched bands for colons, semicolons, question marks, exclamation marks and guillemets, after the corrections recorded in §4. Zero guillemets in either column. ## heads 1:1 (CAPUT V. → CHAPTER V., and so through IX, matching Ruth's shipped English).

⚠ Nothing in this file has been checked against Migne's plate. Every candidate below that would need the plate is named as such, and none of them fired a marker.

§1 · THE 7a″ DELIVERABLE — every lemma in the range, collated against the Clementine

Collated against sources/vulgate/clementine-flat.txt; readings cited from sources/vulgate/clementine/Apc.lat etc. The range glosses Apc 5:1–9:19.

1a. Divergences that took a [var: …] (4 of the 5; the fifth is in §1b)

0721D (chunk 0006), Apc 6:4 — Migne's lemma prints Sumeret pacem de terra, et ut indicem, etc.; Clementine Apc 6:4: ut sumeret pacem de terra, et ut invicem se interficiant. indicem ("an informer") is a real, well-formed accusative and takes no [sic:], but it leaves the clause with no verb and it is not the verse. Rendered as printed ("and that an informer, etc.") with [var: Vulg. ut invicem se interficiant, "that they should kill one another"]. ⚑ Plate-read candidate: indicem / invicem is a plausible n→d misreading, and brief §4 warns the digitization alters spelling in both directions. The marker claims only what the Clementine reads, which is checked; it does not claim the plate.

0722B (chunk 0006), Apc 6:6 — Migne's lemma prints Et vivum et oleum ne laeseritis.; Clementine Apc 6:6: et vinum, et oleum ne laeseris. Two divergences in five words: vivum ("the living thing") for vinum ("wine"), and a plural imperative laeseritis for the singular laeseris. Rendered as printed with [var: Vulg. et vinum, et oleum ne laeseris, "and the wine and the oil hurt thou not"]. ⚑ This is the site where the lemma and its own gloss disagree, which is why it matters: the gloss immediately under it expounds vino, id est sanguine meo ("with wine, that is, with my blood"), and the same paragraph has already given Vinum, praelo quidem attriti… and Vinum similiter emitur denario uno. The gloss reads vinum three times; only the lemma reads vivum. Both are rendered as printed — "the living thing" in the lemma, "wine" in the gloss — and the contradiction stands on the page, per 7a. Pattern 17 governs the number: laeseritis is plural, so the English imperative addresses a plural ("do not hurt"). ⚑ Plate-read candidate (vivum / vinum, again a real word, so no marker beyond the [var:]).

0722D (chunk 0006), Apc 6:10 — Migne's lemma prints Judicas et non vindicas sanguinem.; Clementine Apc 6:10: non judicas, et non vindicas sanguinem nostrum. Migne's lemma lacks the first non. Rendered as printed — "Thou judgest and dost not avenge the blood" — with [var: Vulg. non judicas, et non vindicas sanguinem nostrum, "thou dost not judge, and dost not avenge our blood"]. Pattern 17: judicas is singular, so thou. ⚑ Note that the gloss under it, Faciendo discretionem bonorum et malorum ("By making a distinction between the good and the evil"), construes more easily with the affirmative lemma Migne prints than with the Clementine's negative one — so this is not obviously a compositor's slip, and no conjecture is offered. The negative that Migne does print (non vindicas) is carried; no negative is supplied that he does not print.

0724B (chunk 0007), Phil 1:23 — the gloss on Issachar quotes Cupio dissolvi, et esse cum Christo; Clementine Phlp 1:23: desiderium habens dissolvi, et esse cum Christo. [var: Vulg. desiderium habens dissolvi, et esse cum Christo, "having a desire to be dissolved, and to be with Christ"]. The received-text form is the familiar patristic citation form, not the Clementine's; it is recorded, not explained (Pattern 14 forbids speculating why).

1b. The remaining divergences — one marker taken, the rest declined with the reason

• 0723A (0006), Exite de Babylone, populus meus, et immundum ne tetigeritis. This is a conflation and is not any single Clementine verse: the nearest are Apc 18:4 (Exite de illa populus meus), Jr 51:45 (Egredimini de medio ejus, populus meus), and 2 Cor 6:17, whose closing clause it matches; "Babylon" is in none of them at that point. A [var:] was taken here, worded so it names the witness and does not repeat Migne's own words (the verifier's mirror test): it seemed worse to leave a reader believing he was meeting a verse. Recorded here as the one marker in the range whose content is a composite statement rather than a single received reading.
• 0720B (0005), Luke 10:24 — Migne prints Multi reges et prophetae voluerunt videre; Clementine Lc 10:24 has multi prophetae et reges voluerunt videre. Order only, sense unchanged; Pattern 14 fires on divergence "in a way that changes the sense", so declined. Logged because a blind reader will see it.
• 0725C (0008), Apc 8:3 — Migne's lemma Super aureum altare.; Clementine super altare aureum. Word order only; the gloss turns on neither. Declined, logged.
• 0728D (0009), Apc 9:17 — Migne's lemma Procedebat ignis (imperfect); Clementine de ore eorum procedit ignis (present). Rendered as printed ("Fire went forth"). The gloss is three words long (Causa aeternae poenae) and depends on no tense. Declined as immaterial, logged. ⚠ Note that 7914 @1071D treats a tense change as material when the English restores the Vulgate tense; nothing is restored here — Migne's tense is what is rendered.
• Orthography, none material, none marked: tanquam for tamquam throughout (0720C, 0723A, 0727B, 0728D) is this corpus's normal orthography, not a divergence; lacrymam for lacrimam (0725A); Haebraei for Hebraei (0720D); Hebraice lower-cased in the Clementine, capitalized by Migne (0727C).

1c. Lemmata checked and found to agree with the Clementine EXACTLY

Recorded in full, because a findings list alone cannot be distinguished from a report by an agent that never looked. Each of the following is verbatim at the verse it glosses (allowing only the Clementine's own comma-spacing and the tanquam/tamquam orthography noted above):

Apc 5 (chunk 0005): Et vidi in dextera sedentis · Librum · Foris · Intus · Signatum · Sigillis · Septem (all six from 5:1, the last five being the single words librum … intus et foris, signatum sigillis septem glossed one by one) · Angelum · Praedicantem · Aperire librum (5:2) · Flebam multum (5:4) · Et unus de senioribus · Vicit leo (5:5) · Et ecce in medio throni · Tanquam occisum · Cornua septem (5:6) · Ceciderunt coram agno · Habentes singuli citharas (5:8) · Cantabant canticum · Tribu (5:9) · Vidi · audivi · Vocem angelorum · Multorum in circuitu throni et animalium et seniorum · Millia millium (5:11) · Sapientiam, et fortitudinem (5:12) · Benedictio · et honor · et gloria · et potestas (5:13) · Et quatuor animalia · Ceciderunt in facies (5:14).

Apc 6 (chunk 0006): Et vidi · Quod aperuisset · Veni et vide (6:1) · Ecce equus albus · Arcum (6:2) · Et cum aperuisset sigillum secundum (6:3) · Equus rufus · Et datus est ei gladius magnus (6:4) · Et ecce equus niger · Stateram (6:5) · Et audivi tanquam · Bilibris · tritici (6:6) · Ecce equus pallidus · Nomen illi mors · Et infernus · Super quatuor (6:8) · Et cum · Vidi subtus altare · sub · Altare · Interfectorum propter verbum (6:9) · Usquequo (6:10) · Conservi eorum (6:11) · Sol factus est niger · Tanquam saccus (6:12) · Grossos suos (6:13) · Coelum (6:14) · Reges terrae · Divites · Fortes (6:15) · Et dicunt montibus (6:16).

Apc 7 (chunk 0007): Post haec · Quatuor angelos · Angelos · Tenentes · Ventos · Super mare (7:1) · Et vidi alterum · Ascendentem · Signum (7:2) · Centum quadraginta · Ex omni tribu (7:4) · Duodecim (7:5) · Simeon · Levi · Issachar (7:7) · Zabulon · Joseph · Benjamin (7:8) · Post haec · Stantes (7:9) · Salus Deo nostro qui sedet (7:10) · Et omnes (7:11) · Amen · benedictio · claritas (7:12) · Et respondit unus (7:13) · Stolas suas (7:14) · Non esurient, neque sitient (7:16) · Deducet eos ad vitae · Absterget Deus omnem lacrymam (7:17). Also the one full proof-text in the chunk: Rom 10:10, Corde enim creditur ad justitiam, ore autem confessio fit ad salutem — verbatim in the Clementine, punctuation apart.

Apc 8 (chunk 0008): Et cum aperuisset (8:1) · Septem angelos · Et datae sunt (8:2) · Et alius angelus · Stetit ante altare · Habens · Thuribulum · De orationibus (8:3) · Fumus (8:4) · Praeparaverunt (8:6) · Et primus angelus · Et facta est grando · In terram · Arborum (8:7) · Et secundus angelus · In mare (8:8) · Mortua est tertia · Quae habebant animas (8:9) · Et tertius angelus · Fontes (8:10) · Et quartus angelus (8:12) · Unius aquilae (8:13).

Apc 9 (chunk 0009): Et quintus angelus · Et vidi stellam · Stellam · Et data est (9:1) · Aperuit puteum abyssi · Fumus fornacis · Obscuratus est sol et aer (9:2) · Exierunt locustae (9:3) · Praeceptum est illis ne laederent (9:4) · Ne occiderent · Mensibus quinque · Ut cruciatus scorpii (9:5) · Et fugiet mors ab eis (9:6) · Similitudines locustarum · Similes equis · Tanquam coronae (9:7) · Et habebant capillos · Dentes leonum (9:8) · Loricas · Sicut vox curruum (9:9) · Et habebant caudas (9:10) · Et habebant super se regem · Cui nomen Hebraice Abaddon (9:11) · Vae unum abiit (9:12) · Ex quatuor cornibus · Altaris aurei · Vocem unam (9:13) · Solve quatuor · Qui alligati sunt in flumine magno Euphrate · solve (9:14) · Soluti sunt · In horam et diem · horam · diem · mensem · annum (9:15) · Vicies millies · Et audivi · audivi (9:16) · Et ita vidi · Habebant loricas igneas · Vel habebant loricas igneas · Tanquam capita leonum · Vel ignis · fumus · sulphur · Fumus et sulphur (9:17) · Occisa est tertia pars (9:18) · Similes serpentibus (9:19).

⚑ One lemma is the commentator's own citation of a variant, not a divergence: at 0727B he writes Alia littera dicit mensibus sex, propter sex aetates — he is reporting a rival reading against the mensibus quinque of Apc 9:5, and says so. No marker; the English says so too.

1d. Fragmentary lemmata — how each was resolved, and why the two calls differ

Convention 3 renders a fragment as the phrase it stands for; the brief's closing note adds that a complete one-word lemma is rendered as it stands, and that contrasting the two classes manufactures an inconsistency. Two lemmata in this range sit on that boundary and were decided in opposite directions, deliberately:

• 0724D–0725A, Deducet eos ad vitae, etc. — EXPANDED to "He shall lead them to the fountains of the waters of life". "To life's" is not English at all; the fragment breaks inside a genitive phrase (Apc 7:17 ad vitae fontes aquarum), which is precisely Ruth's Non te. class.
• 0723D–0724A, Centum quadraginta. — NOT expanded, rendered "A hundred and forty". The clip lands on a whole numeral that is readable English; nothing is ungrammatical, so convention 3's trigger is not met, and expanding it to "a hundred and forty-four thousand" would import a number into a paragraph where the number is itself in question (see §2, 0724A). The Latin fragment stands opposite in the parallel column either way.
• 0727A, Sicut habent scorpiones. — rendered as printed, "As the scorpions have". This one elides internally: Clementine Apc 9:3 reads sicut habent potestatem scorpiones terrae, so both the object and the qualifier are gone from the middle of the lemma, not clipped off its end. Supplying "power" would have been the only expansion available and it would have put a word inside Migne's phrase rather than after it. Rendered as it stands; the elision logged here.
• Ordinary end-clips, rendered as the phrase they open and needing no note: Cornua septem, etc. · Ceciderunt in facies, etc. · Sapientiam, et fortitudinem, etc. · Sumeret pacem…, etc. · quia bilibris tritici, etc. · Post haec, etc. · Stolas suas, etc. · Non esurient, neque sitient, etc. · Similitudines locustarum, etc. · Sicut vox curruum, etc. · Et audivi, etc. · Soluti sunt, etc. · Quae habebant animas, etc. · Praeceptum est illis ne laederent, etc.

§2 · THE PLATE'S OWN TROUBLE — rendered as printed, logged here

0720A (0005) · intellexeruut — a non-word. [sic:] TAKEN.

Sancti et antiqui Veteris Testamenti patres intellexeruut aliquando… — intellexeruut is no Latin word (u for n). Corpus control: 1 occurrence against 21 intellexerunt across all of src/latin/. Per addendum A a non-word is the class worth raising, and it is carried under Patterns 7/10/12: the English gives the sense and the marker — "…of the Old Testament understood [sic: intellexeruut] at one time…". Strip the marker and the sentence still reads (test 1). ⚑ Plate-read candidate: this may be the digitization's corruption rather than Migne's type; the marker asserts only that the run in our Latin twin is carried, which is what the verifier checks.

0726D–0727A (0009) · baereticorum — a non-word. [sic:] TAKEN.

Discipuli baereticorum locustis comparantur. Corpus control: 1 occurrence of baereticorum (plus one baereticae elsewhere) against 501 haeretic- forms; this work alone prints haeretic- fifteen times in the same chunk. Rendered "The disciples of the heretics [sic: baereticorum] are compared to locusts". Same plate-read caveat.

0722A (0006) · a full stop mid-sentence — PRESERVED, per Pattern 8

Videns diabolus se per apertas persecutiones non profecisse. immittit haereticos… — Migne sets a period and then a lower-case immittit, which is the same class as 11542 @1041D. The English reproduces both: "…had not prevailed by open persecutions. he sends in heretics…". Pattern 8's "never delete"; not tidied, not merged.

0722D (0006) · a paragraph that simply stops — PRESERVED

Interfectorum propter verbum. Vel aperto martyrio, vel aliquibus anxietatibus* — no terminal mark. The English stops in the same place with no mark supplied (Pattern 8).

0725B (0008) · VERS. 3-- — one verse address out of eighty lacks its period

Every other address in the range is set VERS. n.--; this one is VERS. 3--. The address passes through verbatim, untranslated, so the English prints VERS. 3-- too. ⚑ Plate-read / TEI-patch candidate. This is exactly Ruth's 0537D class, where the digitization dropped the space after VERS. at one mark in twenty-three and the plate set it like all the others. Not patched here — a patch is a claim about the plate and this range has had no plate read. If a reader confirms Migne sets VERS. 3.--, it belongs in data/tei-patches/8977.json, not in the English.

0724A (0007) · the arithmetic does not come out — centum sexaginta quatuor. NO MARKER.

duodecim per quatuor multiplicamus, et fiunt quadraginta octo; ut vero ad Trinitatem haec perfectio referatur, quadraginta octo multiplicamus, et sic fiunt centum sexaginta quatuor. 48 × 3 = 144, which is the verse's own number (Apc 7:4 centum quadraginta quatuor millia), and the same column's preceding lemma prints Centum quadraginta. Migne prints 164. Rendered as printed, "a hundred and sixty-four". ⛔ [cj:] considered and DECLINED. sexaginta is a real, well-formed word occupying its slot, so Pattern 12 has nothing to wrap; and Pattern 18a's ruling is scoped to negations and correlatives (with the census closed over verb and word substitutions), while addendum A is explicit that a real word reading oddly is Migne's until a plate says otherwise and must not take a marker on the strength of an argument, however good the argument. This is a good argument. It is still an argument. ⚑ Plate-read candidate, high value: sexaginta / quadraginta. ⚑ Note also that quadraginta octo multiplicamus names no multiplier — "we multiply forty-eight" — the per tria being elided as per quatuor was not. Rendered elliptically, as printed.

0726C (0009) · the fifth angel called quartus. NO MARKER.

Et quintus angelus. Ordine narrationis et ordine temporis hic angelus dicitur esse quartus. The formula recurs three times in the range and elsewhere agrees with the lemma (0725C Et primus angelus … Ordine narrationis; 0726A Et tertius angelus, ordine narrationis; 0726B Quartus dicitur ordine narrationis; and at 0727C Vae unum abiit is glossed primum ordine narrationis et ordine temporis). Against a lemma reading quintus, quartus is the odd man. Rendered as printed. Declined for the same reason as 0724A: a real word in its slot, and the corpus cannot convict it. ⚑ Plate-read candidate.

0726C (0009) · clavis illa sunt — a number disagreement with an English exponent

Vel clavis illa sunt ipsi principes per quos haeretici operantur. A singular noun and demonstrative against a plural verb and plural predicate. English does have an exponent for this (that/those, is/are), so Pattern 12's second test does not apply — rendering it does not hide it. Rendered literally: "Or that key are the very princes through whom the heretics work." No marker; logged. ⚑ Plate-read candidate (claves illae).

0722B (0006) · nec per apertas haereses — apertas against the work's own scheme

Videns diabolus nec per apertas tribulationes, nec per apertas haereses posse proficere… The chapter's own frame two columns earlier makes the third horse haereticorum persecutio occulta and only the second one open, so "open heresies" is what the argument denies. Rendered as printed. ⛔ [cj:] considered and DECLINED on addendum A's ground: apertas is a real word in its slot, the second apertas may simply be Migne repeating the first, and no corpus argument may fire a marker. ⚑ Plate-read candidate. Both negatives (nec … nec) are carried and each is on its own prepositional phrase, not on the verb (test 2/2a).

0728D (0009) · nunquam fuerunt baptizati, nec ab Ecclesia separati — the contradiction stands

omnes Judaei vel pagani, qui nunquam fuerunt baptizati, nec ab Ecclesia separati. Carried whole: "who were never baptized, nor separated from the Church." Under 7a this is rendered as printed even though the two limbs sit oddly together (one who was never baptized was never in the Church to be separated from it). No word translated as its opposite; no negation moved; nothing supplied. Whether the plate reads otherwise is a question for a plate read, not for the English.

0723D (0007) · signatio pectoris, confessio est cordis — rendered as printed

Signum geritur in fronte ne celetur in tribulatione, nam signatio pectoris, confessio est cordis. The very next sentence assigns the breast to fidem cordis and the forehead to confiteatur, so the two do not divide the body the same way. Rendered literally, "for the signing of the breast is the confession of the heart". No conjecture offered.

0723C–D (0007) · Vel a se descendit — rendered as printed

In a paragraph glossing Ascendentem and running ascendit … ascendit ab ortu, the third alternative prints Vel a se descendit. A real verb in its slot; rendered "Or he came down from himself". ⚑ Plate-read candidate; no marker.

Pattern 9 items — checked and NOT marked

• 0727C sagita for sagitta (corpus: 1 against 76 sagitta-). Single-for-double consonant is standard medieval orthography, and Pattern 9 rules that attested medieval spellings are correct Latin for their period. Declined, though it is the closest call in the range: it is a hair's breadth from the non-word class, and a plate read would settle it.
• 0728D Antechristi in a chunk that prints Antichrist- fifteen times. Antechristus is an attested medieval form; corpus-wide it occurs once, which is a reason to raise it and not a reason to mark it. Rendered "Antichrist". ⚑ Plate-read candidate.
• 0727C adherentes for adhaerentes — ordinary ae/e monophthongization. No marker.
• 0720D Haebraei, 0725A lacrymam, 0723A coelum, tanquam throughout — orthography, no marker.

## §3 · REFERENTS AND NEGATION — the sites where two readings construe, or where a supplied ## pronoun had to be named

The brief's dominant defect class. Every "he/she/it/they" not printed in the Latin was named to a specific antecedent before it was written. The ones worth recording:

• 0720A (0005), et doctrina et exemplo munivit — no object is printed. Supplied as the Church: she is the subject of the immediately preceding quae fuit albus equus, and the next relative, quam rufus equus impugnavit, is feminine accusative and can only be the Church. The horse (equus, masculine) is excluded by that quam. English: "which he fortified by doctrine and by example, and which the red horse assailed."
• 0720D (0005), caro sanctorum, quae tenditur in ligno, dum … imitantur — quae is singular and binds caro (the flesh), while reddentes … imitantur is plural and binds the saints. The English keeps both: "the flesh of the saints, which is stretched upon the wood, while by divers torments, rendering the same sound, they imitate the cross of Christ." Collapsing them into one number was the available error and was not made.
• 0723C (0007), ne sui aliquatenus torpeant vel desperent — sui is God's own, not the devil's: the clause explains why the devil's readiness to hurt is reported together with his being bridled, and the consolation is for the faithful. Both readings parse; the sense decides, and the English says "his own" with the ambiguity recorded here rather than resolved silently in a footnote the reader cannot see.
• 0723D (0007), qui est ortus solis, id est Patris — the genitive Patris glosses solis, not ortus; the sun is the Father. Kept on its own noun: "who is the rising of the sun, that is, of the Father." (The brief's named failure — a genitive lifted off its noun — was the risk here.)
• 0728C (0009), sicut intellexeram … sic intellexit — a first-person pluperfect answered by a third-person perfect. Rendered as printed: "as I had understood … so he understood". Under Pattern 18a a person slip is explicitly declined for [cj:], and under 7a the English may not quietly become "I understood". ⚑ Plate-read candidate (intellexi).
• Test 2a sites — the host of the negative proved from the sentence's own structure, not from what reads well:
• 0722A non solum in minores, sed etiam in majores — host is in minores, proved by the sed etiam limb. English: "not only against the lesser, but even against the greater".
• 0722B oleum sunt non alios inebriantes, sed prius orationibus ungentes — host is alios inebriantes, proved by the sed limb; not the copula sunt.
• 0723D Hoc signum etsi non re, fide tamen antiqui patres praetulerunt — host is re, proved by the answering fide tamen. English: "though not in fact, yet in faith".
• 0724C Non per nos, sed per agnum habemus salutem — host is per nos.
• 0725C quia non omnes orationes repraesentat Patri, sed eas exaudit quae… — host is the quantifier omnes, proved by the sed limb, which is what makes the sentence coherent: he presents not all prayers, not "he does not present prayers". English: "because not all prayers does he present to the Father, but he hears those which pertain to salvation." ⚑ This is the one site in the range with 2a's exact trigger shape (negative, verb, quantifier adjacent), and the wrong host reads perfectly well.
• 0727B non aperte tormentant, sed apud saeculi principes accusant — host is aperte, proved by the sed limb (they do torment; not openly).
• Every other printed negative was checked onto its own word and carried: non/nec/ neque/nisi/ne/nihil/nullus/nemo/nunquam/nolentes/nondum at 0720A, 0720B, 0721B, 0721C, 0721D, 0722A ×3, 0722B ×4, 0722C ×2, 0722D ×2, 0723A ×3, 0723B, 0723C ×2, 0723D ×4, 0724B ×2, 0724C, 0725B ×3, 0725C ×2, 0726A, 0726D, 0727A ×4, 0727B ×2, 0727C, 0727D, 0728D ×2. None appears in the English that Migne does not print, and none of his is missing.

§4 · PUNCTUATION — what was corrected in our own draft, and the standing decision

Test 3 was run band by band across all 37 anchor-delimited bands. The first pass found 24 mismatched bands, every one of them the translator's convenience and not the plate's: semicolons and colons supplied where Migne prints commas (and, at 0725D and 0726D, two of his semicolons softened to commas in our English). All were conformed to the printed marks; the second pass returns zero mismatched bands. Recorded because the brief's own warning is that the totals can cancel — here they did not cancel, but the class is exactly the one it names, and it was invisible until the bands were compared.

⚠ No comma→semicolon sweep was performed in the other direction: Migne prints semicolons himself in this range (twelve of them), so Tobit's narrow exception does not apply and was not invoked.

§5 · SECTIONS READ CLEAN, AND CANDIDATES REJECTED

Read clean, no crux: CAPUT V's glosses on 5:8–5:14 (0720D–0721B); CAPUT VII's tribal etymologies at 0724B (Simeon / Levi / Issachar / Zabulon / Joseph / Benjamin — the Hebrew glosses are the commentator's and are rendered as he gives them, not corrected against Jerome); CAPUT VIII entire apart from 0725B's VERS. 3--; CAPUT IX's 0728B chronology (horam = half a year, diem = a year, mensem = a second, annum = a third, totalling the three and a half years Migne then names — internally consistent, checked, no defect).

Candidates raised and rejected, with the reason:

1. sagita (0727C) — Pattern 9 orthography, not a non-word. Declined.
2. Antechristi (0728D) — attested medieval form. Declined.
3. centum sexaginta quatuor (0724A) — real word, corpus argument cannot fire a marker. Declined; plate read requested.
4. quartus for quintus (0726C) — same. Declined; plate read requested.
5. apertas haereses (0722B) — same. Declined; plate read requested.
6. descendit for ascendit (0723D) — same. Declined.
7. intellexit for intellexi (0728C) — a person slip, explicitly outside Pattern 18a. Declined.
8. clavis illa sunt (0726C) — English has an exponent for the disagreement, so rendering it does not hide it and Pattern 12's second test is not met. Declined.
9. Luke 10:24 word order, Apc 8:3 word order, Apc 9:17 tense — divergences that do not change the sense. [var:] declined for each, all three logged in §1b.
10. The A–D band sequence of this range (0719D → 0720A … 0729A) contains no gap or recurrence worth reporting; per runbook false positive 1, no band arithmetic was performed and none is reported. No text is broken mid-word or mid-construction across any anchor.
11. No [f: …] tags: this is a lemma-and-gloss commentary, runbook false positive 4.
12. No em-dash was marked, reported, or engineered around.

Open, for whoever reads the plate: the eight plate-read candidates flagged above, in descending value — centum sexaginta quatuor (0724A) · vivum/vinum (0722B) · indicem/invicem (0721D) · VERS. 3-- (0725B) · quartus (0726C) · apertas haereses (0722B) · clavis illa sunt (0726C) · intellexeruut (0720A) and baereticorum (0726D), these last two only to establish whether the corruption is Migne's or the digitization's, since the [sic:] markers stand either way on the run our Latin twin carries.

Range file for the third of four agent stints (Apc 10–16). Not the merge — the merged cruces.md is written once at the end from all four range files.

Translated 2026-08-28. Chunks 0010, 0011, 0012, 0013, 0014. Marker parity against the Latin twins, verified paragraph by paragraph: column anchors 11 · 6 · 6 · 12 · 6, verbatim and in order; [n:] 2 · 0 · 0 · 0 · 0; italic spans 106 · 60 · 112 · 118 · 62 (exact, with marker contents stripped); question marks 0 in both languages throughout; guillemets 0; ## heads 1:1 (CAPUT X.–XVI. → CHAPTER X.–XVI., Ruth's shipped form). Colons and semicolons compared band by band (paragraph by paragraph, the anchors partitioning both twins identically): zero mismatched bands after the five repairs listed in §4.

Apparatus taken in this range: 7 [var:] · 1 [cj:] · 0 [sic:] · 0 [d:] · 0 [ed:] · 0 [nt:].

§1 · THE 7a″ DELIVERABLE — every lemma in the range, collated

Collated against sources/vulgate/clementine-flat.txt, cited from sources/vulgate/clementine/Apc.lat, Ps.lat, Mt.lat, 1Cor.lat, Zach.lat. This section lists the agreements as well as the divergences, because a findings-only list cannot be told apart from a report by an agent that never looked.

1a. Divergences that took a [var:] (7)

⚠ On the last: enim is a connective, not a sense-word, and the marker was still taken because the insertion is what turns a quoted verse into the gloss's own middle term. A reader who does not know the psalm cannot see that the "for" is Migne's.

1b. Divergences examined and deliberately NOT marked, with the reason

• 0729C Tempus amplius non erit. — Apc 10:6 reads Quia tempus non erit amplius. Word order only; the negation, its host (erit) and the sense are untouched. Pattern 14 wants a divergence "in a way that changes the sense". Logged, not marked.
• 0729C Et ea quae in illo. — Apc 10:6 has ea quae in eo sunt (three times). A bare demonstrative swap, illo for eo, with no change of reference or sense. ⭐ Worth recording that the gloss's own re-quotation four words later reverts to eo (Creavit ea quae in eo sunt), so the divergence is confined to the lemma. Logged, not marked.
• 0735C Et angelus tertius. — Apc 14:9 Et tertius angelus secutus est illos. Word order only.
• 0739A Dare ei calicem. — Apc 16:19 dare illi calicem. Same pronoun swap as 0729C; no sense change.
• 0736C, the Prophet's Lacus ubi non erat aqua — the nearest Clementine text is Zach 9:11 de lacu in quo non est aqua (Gn 37:24 cisternam … quae non habebat aquam is the other candidate). Migne's is a loose allusion, not italicized on his page, and the negation is intact in both. No marker; the English keeps it unitalicized, as he prints it.
• 0729C Nam si Christus resurrexit, et nos resurgemus [n: (I Cor. XV)] — no such sentence stands in 1 Cor 15. It is a free epitome of the chapter's argument (cf. 15:12–16, 20), and Migne's own note cites the chapter, not a verse. ⛔ A [var:] here would have to pick one verse to be "the" received reading, which would misdescribe what he is doing. Declined deliberately; logged instead.
• Orthography, throughout, and none of it marked: coelo/coelum/coeli for caelo etc.; tanquam for tamquam (0729A, 0737A, 0738A); characterem for the Clementine's caracterem (0734B); Citharaedorum for citharoedorum (0735A); terraemotus set solid. These are Pattern 9 items — conventions of the printing, carrying no information about this plate — and marking them would bury the seven real divergences in noise.

1c. Lemmata checked and found to agree with the Clementine exactly

Every remaining italic span in cols 0729A–0739B was walked against the flat file. All of the following agree word for word with the verse they gloss (fragments agreeing with the corresponding stretch of it):

Apc 10 (chunk 0010): Et vidi alium angelum fortem · De coelo · Amictum nube · Et iris in capite · Et habebat in manu sua libellum apertum · Pedem suum · Sinistrum autem super terram · Quemadmodum cum leo rugit · Scripturus eram · Signa quae locuta · Levavit manum suam ad coelum · Juravit · Consummabitur mysterium · Mysterium Dei · Sicut evangelizavit · Et audivi · Vade · Stantis super mare · Et abii · Ad angelum · Et dixit mihi: Oportet. Apc 11: Metire · Atrium · Civitatem sanctam · Et dabo duobus · Duobus testibus · Diebus mille · Super aquas · In sanguinem · Aegyptus · Et munera mittent · Et post dies · Super pedes · Vocem magnam · Terraemotus magnus · Et decima pars · Occisa sunt in terraemotu · Gloriam Deo · Et septimus angelus · Factum · Et regnabit · Et viginti quatuor seniores · Ceciderunt in facies · Qui corruperunt terram · Et apertum est · Arca · Et grando magna. Apc 12 (chunk 0011): Et signum · Luna sub pedibus · Et in capite ejus corona · Draco magnus · Et cornua decem · Et cauda · Stellarum · Stetit ante mulierem · Devoraret · Et peperit · Diebus mille · Michael et angeli ejus praeliabantur · Non valuerunt · Et audivi vocem · Accusabat illos · Die ac nocte · Descendit diabolus · Sciens quod modicum tempus habet · Et postquam vidit · Et datae sunt mulieri · Aquilae · Per tempus, et tempora (all three quotations) · Et adjuvit terra mulierem · Et aperuit · Absorbuit · Iratus est draco in mulierem · Et stetit. Apc 13 (chunk 0012): Et vidi · Capita septem et cornua decem · Nomina blasphemiae · Quasi occisum · Et plaga mortis ejus curata est (and its re-quotation) · Et aperuit os suum · Et tabernaculum · Occisus est ab origine mundi · Qui in captivitatem · Qui in gladio occiderit · Hic est · Et vidi · Cornua · Et loquebatur sicut draco · Et potestatem · Terram (and its re-quotation) · habitantes in ea · Adorare · Ignem · In conspectu · Ut loquatur imago · Numerum · Qui habet. Apc 14–15 (chunk 0013): Et vidi: et ecce Agnus stabat · Centum quadraginta · Tonitrui · Citharizantium · Et cantabant · Hi sunt qui cum mulieribus non sunt · Ex hominibus primitiae · Et vidi · Alterum angelum · Per medium coeli · Evangelium aeternum · Timete Dominum · Quia venit · Cecidit (both) · Vino irae fornicationis · Quod mistum · In calice · Et fumus tormentorum · Ascendet in saecula (both) · Nec habent requiem · Hic patientia · Opera · Et vidi · Similem Filio hominis · Angelus · Exivit de templo · Quia venit hora · Et alius angelus · Alius angelus · Mitte falcem · Et misit in lacum · Et exivit sanguis · Usque ad frenos equorum · Per stadia · Et vidi · Et mirabile · Plagas septem · Quoniam in illis · plagas · novissimas · Tanquam mare vitreum mistum igne · Habentes citharas · Canticum Moysi · Viae · justae · verae · Judicia tua · Et post haec · Apertum est templum · Et exierunt · Et unum · Iracundiae · Fumo a majestate · Introire. Apc 16 (chunk 0014): Vulnus · Et factus est sanguis · Sanguis · tanquam mortui · Omnis anima vivens · Super flumina · Audivi · Qui es et qui eras · Et sanguinem eis dedisti bibere · Et audivi · Etiam, Domine Deus omnipotens · Justa · Et aestuaverunt · Neque egerunt · Super sedem bestiae · Commanducaverunt · Praepararetur via · Vidi · Ranarum · Procedunt · Et congregabit · Et septimus angelus · Et facta est civitas · Civitates gentium ceciderunt · Venit in memoriam · Et omnis insula fugit · Talentum · Et blasphemaverunt.

1d. Fragment lemmata — how each was rendered, per convention 3

• 0729C Signa quae locuta: rendered "Seal up the things which the seven thunders have spoken:". Signa is the Vulgate's imperative (Apc 10:4 Signa quae locuta sunt septem tonitrua), not the noun "signs"; a stump ("Signs which have spoken") is not English and would also invite the wrong part of speech. The trailing colon is Migne's and is carried.
• 0730B Et dabo duobus. expanded to "And I will give to my two witnesses" — the governed noun is elided (Apc 11:3 dabo duobus testibus meis). Its neighbour Duobus testibus. is a complete phrase, nothing elided inside it, and is rendered as it stands, "To two witnesses". ⚑ Per the brief's control-class rule these are two different classes and the pair is not an inconsistency.
• 0730D and 0732B Diebus mille. kept as "A thousand days", NOT expanded to the verse's diebus mille ducentis sexaginta. Reason, and it decides the sense: the gloss argues hi dies non perficiunt tres annos et dimidium. On 1,260 days that is at best arguable (1,260 = 3½ years of 360 days); on the bare thousand Migne prints it is plainly true. Expanding would have imported the very number the gloss is denying.
• Ordinary clipped lemmata (De coelo, Pedem suum, Gloriam Deo, Super pedes, Cornua, Opera, Talentum, Vulnus, Justa) are rendered as the phrase they open, without importing wording from any English Bible. No lemma in this range was conformed to Douay or AV; each was read word by word off the Latin before any English was written.

§2 · CRUCES

0733B · in [0733B] fructuosos — the word is split across the column anchor, and the split hides a negative. [cj:] TAKEN.

Migne, glossing Apc 12:18 Et stetit [supra arenam maris]:

Non tamen omnes illos vicit minores, sed tantum super arenam, id est, in [0733B] fructuosos moram fecit, in quibus apparent signa vestigii.

The dragon did not conquer all the lesser ones, but only super arenam — and sand is barren. in fructuosos, "among the fruitful", says the opposite of what arenam and the whole sentence require; infructuosos, "the unfruitful", says exactly it, and sand is precisely what takes and shows a footprint (in quibus apparent signa vestigii).

This is the one thing the runbook's false-positive item 1 does tell us to report: text broken mid-word across an anchor. The anchor falls between in and fructuosos, which is also a legitimate word boundary if the text really is two words — which is what makes it undecidable from the file.

• Marker: [cj: fructuosos; read infructuosos, "the unfruitful"], additive per Pattern 18 — Migne's word keeps its English ("among the … fruitful") and the conjecture stands beside it. Nothing is removed and the reader sees both.
• Not a [sic:]: fructuosos is a perfectly good Latin form occupying its slot, so Pattern 12 has nothing to wrap. This is the 11081 @0030B munde/mundo shape exactly.
• Corpus control: infructuos- is attested 32 times across src/latin/ (though not in this exact accusative plural), so the conjectured word is a word this corpus uses.
• ⛔ The rival reading, and why it lost: "he made his stay among the fruitful" can be construed — moram facere in + accusative for "among" is loose but medieval. It loses because it contradicts super arenam in the same clause, which is the sentence's own structure and not a matter of what reads well.
• Plate read wanted at col. 0733B. If Migne prints infructuosos broken at the column turn, this is Corpus Corporum's division and belongs in data/tei-patches/8977.json, and the [cj:] is withdrawn in the same commit.

0734A · tundamentum — a NON-WORD. Plate read wanted. No marker fired, deliberately.

Quia et hi, simulata resurrectione, praedicationis suae faciunt tundamentum.

tundamentum is no Latin word. Corpus frequency across src/latin/: 1 × tundamentum against 107 × fundamentum. The shape (f → t, the crossbar) is a classic broken-type or mis-transcription.

⛔ No [sic:]. The brief's addendum A records that every suspicion aimed at a non-word in that batch was confirmed as our corruption, not Migne's — and addendum B records seven false [sic:] withdrawn in one batch, each publicly blaming Migne for the digitization's error. Firing one here on a corpus argument alone would be the same mistake. The Glossa block's own precedent points the same way: Ruth's significaRuth @0534A was resolved by a tei-patch after a plate read, not by a marker.

The English therefore reads "make the foundation of their preaching", and this entry is the trace. ⚑ If the plate reads tundamentum, a [sic:] must be added here and the sense kept beside it (7a⁗ test 1: the word is the clause's object, so quarantining it alone would leave "make the ___ of their preaching").

0734D · centecimum — non-word or attested medieval spelling? No marker; plate read wanted.

Hi reddunt fructum, alii tricesimum, alii sexagesimum, alii centecimum.

Corpus: 1 × centecimum against 21 × centesim- forms. Two readings, and they point different ways:

1. Our corruption (s → c), same class as tundamentum above.
2. An attested medieval spelling. Assibilated c/s before a front vowel is a real convention of the period (cf. nuncius/nuntius), and under Pattern 9 an attested medieval spelling is correct Latin for its period — translate normally, no italics, no crux.

Because reading (2) is live, and because either way the marker would be wrong, none was fired; the English reads "a hundredfold". ⚑ Raised here as a plate candidate. Note also that tricesimum in the same line is not suspect — it is classical, and no divergence.

0735C · corrrigitur — trebled r. Same class, same treatment.

Qui non corrrigitur vindicta ejus ad correctionem data, punietur aeterna poena.

Corpus: 1 × corrrigitur against 8 × corrigitur. A doubled letter is a plate or transcription accident, not a reading; the negation non attaches to it and is carried ("He who is not corrected …"), which is the thing that matters. No marker; plate read wanted at 0735C.

0737B–C · CAPUT XV prints VERS. 18 for what is Apc 15:8. Reproduced as printed.

Apocalypse 15 has eight verses. Migne's last address in the chapter is VERS. 18.-- Fumo a majestate, etc., and the text glossed is unmistakably 15:8 (Et impletum est templum fumo a majestate Dei … nemo poterat introire). The verse number is his, and the brief is explicit that his numbering is not ours to tidy: reproduced verbatim, not corrected to VERS. 8, and logged here. (The following chunk resumes at CAPUT XVI VERS. 2, so nothing downstream depends on it.)

0737B · Cum ostendissent — a plural verb where the sentence has a singular subject.

Cum ostendissent illos angelos habere potestatem plagandi, … prius proponit causam…

ostendissent is 3rd plural, proponit 3rd singular, and both clauses have the same implied subject (John). Rendered as printed — "When they had shown … he first sets forth" — because English does have an exponent for the number, so the disagreement stays visible to the reader (this is not the Pattern 9 gender case, where English has no exponent). No marker: a real word in its own slot, wrong only in number, is Pattern 7's business and not Pattern 12's.

0729C · Cum dicis — second person singular. [cj:] CONSIDERED AND DECLINED.

Cum dicis: Creavit ea quae in eo sunt, haeresim quorumdam destruxit…

dicis is "thou sayest", and the first reflex is that the sense wants dicit, "when he says", since the words quoted are the angel's and destruxit has a third-person subject. Declined, and the reason is that the printed reading construes without help: the commentator is addressing John, whose vision this is and to whom he has been speaking throughout the chunk (monetur, sibi, Ecce revelavi tibi omnia), and John both says the words and, by saying them, destroys the heresy. A [cj:] asserts that the faithful English misleads; here it does not. Rendered "When thou sayest", singular per Pattern 17.

0736D–0737A · consideratione tamen — rendered literally, sense not settled.

Sed omissa prima, quae fuit de correctione ecclesiarum, de aliis tribus dicit, consideratione tamen.

A bare ablative closing the sentence with no governing word. "Yet with consideration" is what the words say and is what the English prints; whether Migne means "though with [some further] consideration", or the phrase is the stub of something longer, cannot be settled from the file. No marker, no conjecture — flagged so a blind reader does not read past it.

0738D · ad quem gloriabuntur — ad for in.

gloriari takes in; ad quem gloriabuntur is not standard. Rendered "or unto whom all the wicked of the diverse sects will glory", as closely as the printed words allow. Real word, real construction, merely unusual — Pattern 7, no marker.

0734B · remunerationem eo minorem — obscure, rendered literally.

eo read as an ablative of degree of difference against minorem: "a reward so much the less, or a middling one, or the greatest", answering the sex / sexaginta / centum ladder that follows. Flagged because the alternative (eo = "from him", already carried by ab eo three words earlier) would be a dittography of sense and is not obviously wrong.

0735A · sua corda devota — a pun English cannot hold.

ut extenderent sua corda devota super lignum crucis.

Under Citharaedorum, "of harpers": corda is strings (of the cithara) and devota is the language of the heart. English must choose one vehicle; "their devout strings" was taken, because the harp governs the whole image and the wood of the cross is the sounding board. The heart-sense is not lost to a reader who has the Latin column opposite. Recorded because a later reader may reasonably prefer "hearts".

§3 · REFERENTS SUPPLIED — every "he/she/it/they" the Latin does not print

The dominant Glossa defect class. Each supplied pronoun in this range, with what fixes it:

• 0729A–0730B, chunk 0010 passim — the unexpressed subject of exsuperavit, fecit, posuit, rugit is the angel of Apc 10:1 identified with Christ by the opening gloss (fortis armatus … exsuperavit, and ita Christus suos in fide concludens). Named in the sentence itself at 0729B, so the "he"s are not free.
• 0729C Cum dicis … destruxit — see §2. Subject of destruxit is John, addressed in dicis, not "the angel"; either reading gives the same English word, but the crux records that the choice was made and not defaulted.
• 0730A monetur quae vidit interim praedicare — John, from Et audivi (Apc 10:8) and from quasi dicatur aperte sibi.
• 0733C ejus in Et tabernaculum. Id est omnem institutionem ejus — God's, from Apc 13:6 (blasphemare nomen ejus, et tabernaculum ejus), not the beast's. The following per quam laudabilis apparet confirms it: it is God who appears praiseworthy through his institution. ⚑ Binding ejus to the nearest noun (the beast) would have made the beast's own apparatus praiseworthy, which is the exact "relative bound to the nearest noun" failure the brief names.
• 0735A ad consortium suum under Tonitrui — "their fellowship" = the hundred and forty-four thousand of the lemma above (Apc 14:1), not the thunder's. The antecedent is in the previous lemma, as the brief warns is usual here.
• 0736B numerus tuorum — thine, singular, addressed to Christ/God (ut Deus separationem faciat two clauses earlier). Pattern 17: singular addressee takes thou.
• 0732D aggreditur expugnare — the object supplied as "her" = the woman/Church (mulierem in the same verse and Ecclesiam throughout the gloss), not the Son.
• 0738B isti malitia diabolum — verb supplied from the parallel (generant): "so these by malice beget the devil". The ellipsis is Migne's and the parallel is his own.

Where two readings both construe, they are named above rather than chosen silently.

§4 · NEGATION AND PUNCTUATION — what was checked, and what was repaired

Negation (test 2 + 2a). Every printed non · nec · neque · nisi · nunquam · nullus · nihil · ne · haud in cols 0729A–0739B was located, its host word named, and the host found under the negative in the English. The population that fires trigger 2a — a negative on a verb with a quantifier or correlative nearby — was worked from the sentence's structure, not from what reads well:

• 0729C non ad detrimentum suae salutis praedicationem subtrahi — host is the phrase ad detrimentum, not subtrahi. English: "preaching is withdrawn not to the detriment of their salvation." (The 8960 @0740B model.)
• 0730A non tanta patieris quanta patientur — host is the quantity tanta, fixed by the quanta correlative. English: "thou shalt not suffer so great things as they shall suffer."
• 0731B Non tantum minores ita glorificant Deum, sed etiam majores — host is tantum minores, fixed by sed etiam. English: "Not only do the lesser thus glorify God, but the greater also." ⛔ Not "the lesser do not glorify".
• 0732A Non de magno effectu, sed de potentia — host is the prepositional phrase.
• 0735C non dicit pro damnatione describenda: sed tantum pro terrore inferendo — host is pro damnatione describenda, fixed by sed tantum.
• 0736B non quales olim fuerant, sed quales in aeternum mansuri — host is quales olim fuerant.
• 0736C Aeterna poena … non autem vinum … ponitur — host is vinum, not ponitur; one verb serves both limbs.
• 0738A non quia aperte persequuntur, sed propter consensum damnabuntur — host is the quia clause, not damnabuntur. English: "they shall be damned not because they openly persecute, but on account of their consent."
• 0738B ii sunt tantum corpora, et non spiritus … corporalia et non spiritualia dabit — hosts are the two nouns.
• 0739A sed non quantum meruerunt — host is quantum.
• 0733B — see §2: a negative that may have been removed by a word division, handled with a [cj:] rather than silently supplied.

⛔ Nothing was rendered as its opposite anywhere in this range, and no negative was supplied that Migne does not print. The sentences that read awkwardly in English (0736D consideratione tamen, 0738D ad quem gloriabuntur, 0734B eo minorem) read awkwardly because he prints them so.

Punctuation (test 3, by band). Compared paragraph by paragraph, the verified column anchors partitioning both twins identically. Question marks: zero in the Latin of all five chunks and zero in the English — nothing supplied, nothing deleted, and Pattern 8b never arose (the indirect questions at 0731C ad quae supplicia ducatur nescit and 0737D quare alios elegit both close with a full stop on his page and do so on ours). Five first-draft mismatches were found by the band comparison and repaired toward Migne, not away:

Semicolon and colon counts now match band for band in all five chunks, including chunk 0012's 25 semicolons, nearly all of them inside the Greek and Latin numeral tallies at 0734B–C, which are reproduced with Migne's own mixture of spelled-out numbers (unum, quinquaginta) and Roman numerals (CCCC, LXX, XL) rather than normalized to one or the other. The unitalicized A, unum at 0734C, sitting among italicized letters, is his inconsistency and is reproduced.

Em-dashes: none introduced. deservit/deserit: neither form occurs anywhere in cols 0729A–0739B — checked, and recorded as a checked zero.

§5 · TEST 5 — does the English ASSERT what he asserts?

Each finished paragraph was read back against the Latin with the counts already known clean, asking only that question. Three sentences that read smoothly were re-examined on that ground alone, and two of the three turned out to need nothing:

• 0730C "Know ye that all the persecutions … proceed from Antichrist, as do those which shall be in his own time." — sicut illas quae erunt in tempore suo. Checked: the comparison is that past and present persecutions proceed from Antichrist just as the ones in his own time will; the English does not soften it into "including". Clean.
• 0736C "just as to a vineyard, that it may bear fruit, greater care is applied than to land" — sicut vineae, ut fructificet major cura quam terrae adhibetur. The dative vineae and terrae are both governed by adhibetur; the English keeps the comparison on the care, not on the fruit. Clean.
• 0733D quia, illo destructo, sciunt se remunerandos — "he being destroyed" is Antichrist, from Qui in captivitatem … scilicet Antichristus two clauses up, not the sword and not the devil. Fixed by the lemma above; recorded in §3's class.

⚠ Not read clean, and named as such: 0736D–0737A consideratione tamen (§2). It reads as a stub, the English reads as a stub, and no reading was chosen.

§6 · WHAT A LATER READER SHOULD LOOK AT FIRST

1. 0733B — the [cj:] on fructuosos/infructuosos. A plate read settles it and may convert it to a tei-patch.
2. 0734A tundamentum, 0734D centecimum, 0735C corrrigitur — three plate candidates in one stint, all in the non-word class, none marked. If the plate is read for any one of them, read the whole line, not the word (addendum C).
3. 0737B CAPUT XV VERS. 18 — Migne's own numbering, reproduced. Do not "fix" it.
4. The seven [var:] — 0729A, 0729C, 0732A, 0734D, 0735B, 0738B, 0739A. Each rests on sources/vulgate/clementine-flat.txt, cited above by book chapter:verse.

Range file for the fourth and last agent block (CAPUT XVII–XXII, the close of the work). Merge into cruces.md with the other three range files. Conventions are those settled on Ruth (src/english/8968/cruces.md §1) and the Glossa NT brief; they are not re-argued here.

verify-english.mjs 8977: verify OK — 20 chunks, columns/notes/sections aligned, no duplicate paragraphs.

Marker tally for this range: 5 [var:] · 1 [cj:] · 0 [sic:] · 0 [d:] · 0 [ed:] · 0 [nt:] · 10 [n:] (parity with the Latin twins).

Parity checked mechanically before filing, per chunk: column anchors (verbatim, in order, and each re-checked against the word it stands beside in the Latin), italic-span count with marker contents stripped, [n:] count, and punctuation band by band (; : ?), which is the check that actually returned sites — see §5.

§1 · THE 7a″ DIVERGENCE LIST — every lemma in the range, collated

Collated against sources/vulgate/clementine-flat.txt, cited from sources/vulgate/clementine/Apc.lat and the other books named. The whole of Apc 17–22 was pulled and read line by line; the pre-scan in data/briefs/8977-lemmata.txt was used only to order the work, and it is incomplete for this range — it carries no entry for chunk 0015's opening spans (Et venit unus · Angelis · Ostendam tibi damnationem) or for chunk 0018's (Et vidi · Coelum novum et terram novam · Sicut sponsam ornatam · Vocem magnam de throno · Ecce tabernaculum). Those were walked in place.

1a. The five [var:] taken

⚠ Two of these five (lignis, vide) are single-letter differences on a real word, which is the shape brief §4 says the digitization can create in either direction. The [var:] is still the right marker and is safe: it asserts only what the Clementine reads, never that Migne's compositor erred. If a plate read returns ignis / vidi, the fix is a TEI patch and the markers come out in the same commit; nothing on the page has to be retracted as a false accusation.

1b. Divergences found and declined, with the reason

Recorded because a findings-only list cannot be told from a report by an agent that never looked. Each of these was checked against the flat file and is a real difference from the Clementine; none of them changes the sense, which is Pattern 14's own threshold.

• 0739D · Job 40:11. Migne: Virtus ejus … in lumbis … in umbilico. Clementine: Fortitudo ejus in lumbis ejus, et virtus illius in umbilico ventris ejus. Migne substitutes virtus for fortitudo, drops ejus and ventris, and so attaches to the loins the noun the Vulgate attaches to the navel. Declined: virtus and fortitudo are near-synonyms here, and the gloss's whole argument turns on lumbis and umbilico, which Migne prints exactly as the Vulgate does. This is the strongest of the declines and a later editor may reasonably reverse it.
• 0741A · Jo 4:18. Migne: Quem nunc habes non est tuus vir. Clementine: et nunc, quem habes, non est tuus vir. Trimmed opening plus quem nunc for nunc quem. Declined: word order only; the negative and every content word stand.
• 0743A · the Vende omnia quotation. Vende omnia, et sequere me answers to no single verse: Lc 18:22 omnia quaecumque habes vende … et veni, sequere me; Mc 10:21 vade, quaecumque habes vende … et veni, sequere me; Mt 19:21 vade, vende quae habes … et veni, sequere me. Declined: a [var:] names one received reading against one printed one, and there is no single verse this compresses. Migne prints it without a note.
• 0745A · Apc 20:5. Migne: Caeteri vero mortuorum non vixerunt. Clementine: Ceteri mortuorum non vixerunt. An added adversative particle. Declined: connective only, and the gloss builds nothing on it. (Caeteri for ceteri is orthography, not a divergence.)
• 0748B · Phil 3:20. Migne: Nostra conversatio in coelis est. Clementine: Nostra autem conversatio in caelis est. A dropped particle in a quotation the gloss puts in the saints' own mouths (dicentes). Declined, same ground as the item above. Applying the two declines the same way is deliberate: added and dropped particles are one class.
• 0748B · Apc 21:19. Migne heads the stone Chalcedoni,; the Clementine reads tertium, calcedonius. Declined: a stone-name form in a series of lemma heads, all the rest of which Migne sets nominative (Smaragdus · Sardonyx · Sardius · Beryllus · Topazius · Chrysoprasus · Hyacinthus · Amethistus). No sense turns on it. Rendered "Chalcedony,", keeping Migne's comma and the [n: (BED.)] inside the span as he sets it.
• 0749C · Apc 21:21. Migne's second lemma sunt margaritae reorders the verse's own duodecim margaritae sunt. Declined: order only, inside a lemma he has already clipped.
• 0741B · Apc 17:16. Migne Igne concremabunt; Clementine et ipsam igni concremabunt. Declined: igne and igni are both regular ablatives of ignis.
• 0745B · Apc 20:8. Migne circumierunt; Clementine circuierunt. Declined: both are standard forms of circumeo.
• The coel- class is not a divergence and took nothing. Et vidi coelum apertum (0743B) · Coelum novum (0745D) · in coelis (0748B) · Exsultate super eam, coeli (0742D, whose [var:] is about number, not spelling). The pre-scan flags these ⚠ because clementine-flat.txt spells cael-; that is a known false positive and was ignored.

1c. Lemmata checked and found to agree with the Clementine exactly

Listed so a later reader can tell a checked lemma from an unchecked one.

Apc 17 (0015): Et venit unus (17:1) · Angelis (17:1) · Ostendam tibi damnationem (17:1) · Meretricis (17:1) · Quae sedet super aquas multas (17:1) · Inebriati sunt (17:2) · Mulierem sedentem · Capita septem · Cornua decem (17:3) · Circumdata purpura · Auro · Poculum aureum · Abominatione (17:4) · Et in fronte · in fronte (17:5) · Ebriam · Et miratus sum (17:6) · Ego tibi dicam (17:7, = ego dicam tibi, order only) · Bestiae quae portat eam · Fuit (17:7–8) · Ascensura est · Et in interitum ibit · Mirabuntur (17:8) · Septem capita (17:9) · Quinque ceciderunt (17:10) · De septem est · et ipsa octava (17:11) · Populi sunt (17:15) · Desolatam facient · Et nudam · Carnes (17:16) · Donec consummentur verba (17:17) · Ez 16:4 non est praecisus umbilicus tuus · Job 40:11 in lumbis · in umbilico.

Apc 18–19 (0016): Et post haec vidi (18:1) · Cecidit, cecidit · Custodia omnis spiritus immundi (18:2) · De virtute (18:3) · Et audivi · Exite (18:4) · Quoniam pervenerunt peccata (18:5) · Reddite illi · Duplicate duplicia (18:6) · Quantum glorificavit se · Et vidua non sum (18:7) · Mors (18:8) · Et flebunt · Reges terrae · Cum viderint fumum incendii ejus (18:9) · Et negotiatores terrae flebunt (18:11) · Auri et argenti (18:12) · Cinnamomum · Similae et tritici · Mancipiorum et animarum (18:13) · Et sustulit · Lapidem quasi molarem · Hoc impetu (18:21) · Post haec audivi · Alleluia (19:1) · Et ceciderunt seniores · Amen (19:4) · Et vox · Pusilli (19:5) · Et audivi quasi vocem (19:6) · Et uxor ejus (19:7) · Et cecidi · Vide ne feceris · Testimonium enim (19:10) · Et vidi coelum apertum · Equus albus (19:11) · Verbum Dei (19:13) · Vestiti byssino albo et mundo (19:14) · Gladius ex utraque parte acutus · Reget eas (19:15) · Et habet in vestimento · Rex regum (19:16) · Et vidi unum angelum · Omnibus avibus · Venite · Et congregamini (19:17) · Et apprehensa est bestia · Vivi missi sunt hi duo (19:20) · Et caeteri occisi sunt · Et omnes aves (19:21) · Ps 32:9 Ipse dixit, et facta sunt.

⚠ Apc 19:5. The Clementine copy on disk prints Ex vox de throno exivit; Migne's lemma Et vox agrees with the ordinary text and no divergence was logged. sources/vulgate/ README.md warns that later upstream corrections are not applied, and a call turning on one letter should say so: this is one, and it is our file that is odd, not Migne's.

Apc 20 (0017): Et vidi angelum · Abyssi (20:1) · Apprehendit · Ligavit eum (20:2) · Et misit eum · Signavit super illum · Et post haec oportet (20:3) · Et animas decollatorum · Et vixerunt (20:4) · Haec est resurrectio · Prima (20:5) · Qui habet partem · Secunda mors · Sed erunt sacerdotes (20:6) · De carcere · Magog (20:7) · Et circumierunt castra sanctorum (20:8) · Die ac nocte (20:10) · Et vidi thronum (20:11) · Et libri aperti sunt · Alius liber · Ex his quae scripta erant (20:12) · Et dedit mare mortuos (20:13) · Haec est mors secunda (20:14).

Apc 21 (0018): Et vidi · Coelum novum et terram novam (21:1) · Sicut sponsam ornatam (21:2) · Vocem magnam de throno · Ecce tabernaculum · Et ipsi populus ejus (21:3) · Et mors (21:4) · Et dixit · Scribe: quia haec (21:5) · Et dixit mihi · Factum est · Alpha et omega · Sitienti (21:6) · Possidebit haec (21:7) · Timidis (21:8) · Veni et ostendam (21:9) · In montem magnum · mons · altus · Ostendit mihi civitatem (21:10) · Lumen · Jaspidis (21:11) · Murum · Portas duodecim · Nomina inscripta (21:12) · Ab oriente portae tres (21:13) · Murus · Et in ipsis duodecim (21:14) · Et qui loquebatur mecum · Mensuram arundineam auream · Ut metiretur civitatem et portas ejus (21:15) · In quadro posita est · Per stadia · Duodecim millia (21:16) · Mensura hominis (21:17) · Et erat · Structura muri ejus ex lapide jaspide · Aurum (21:18) · Jaspis · jaspis · Sapphirus (21:19) · Sardonyx · Sardius · Beryllus · Topazius · Chrysoprasus · Hyacinthus · Amethistus (21:20; Amethistus for amethystus is i/y orthography) · Et duodecim portae · Per singulas · Singulae portae (21:21) · 2 Cor 3:2 Epistola nostra vos estis scripta in cordibus nostris.

Apc 21–22 (0019): Non eget sole · non eget sole, neque luna (21:23) · Et ambulabunt · ambulabunt gentes · In lumine · Reges terrae (21:24) · Et portae (21:25) · Afferent (21:26) · Non intrabit in eam (21:27) · Et ostendit · Fluvium · De sede Dei (22:1) · Et ex utraque · lignum vitae · Lignum vitae · Fructus · Et folia ligni (22:2) · Et dixit mihi · Dominus · Angelum suum (22:6) · Et postquam audissem, et vidissem (22:8) · Et fratrum tuorum (22:9) · Qui nocet, noceat adhuc (22:11) · Beati qui lavant (22:14) · Ego Jesus · quod ego Jesus misi angelum meum testificari vobis haec. Stella splendida. (22:16 — the span elides in ecclesiis and et matutina, an ordinary clip, not a divergence) · Et spiritus · Et spiritus et sponsa dicunt · Et qui sitit, veniat · Qui vult, accipiat aquam vitae gratis (22:17) · Contestor enim (22:18) · Gratia Domini nostri Jesu Christi (22:21).

1d. The re-quotation check (7a″ step 3)

Every lemma marked [var:] was held fixed and every re-quotation of it in the gloss below was read against the fixed rendering. Two of the five markers were confirmed BY the re-quotation, and they are the strongest sites in the range:

• 0739D ortus sui is contradicted four words later by Migne's own umbilicus tuus and non refrenasti. The English keeps both persons as printed and does not harmonize.
• 0744C vide is contradicted by three first-person vidis in the sentences under it. The English renders the lemma as the imperative Migne prints and the gloss as the first person he prints.
• 0742D Exsultate … coeli is the gloss's own re-quotation of the 0742B lemma; both occurrences carry a marker.
• 0749C Plateae (plural) is carried into the gloss's plural pleni, mundi, so the plural is Migne's throughout, not a stray ending.

§2 · THE [cj:], AND THE RIVAL IT BEAT

0745A · Secunda mors … non apprehendet electos, qui in nullo unquam augentur.

Rendered: …shall not lay hold of the elect, who are never in anything increased [cj: augentur; read anguntur, "are distressed"].

augentur is a perfectly good form, so Pattern 12 has nothing to wrap and Pattern 7 requires the English "increased"; Pattern 18 is exactly the gap this falls into — the printed word is real and the faithful English misleads. The sentence's own structure decides it: the clause is appended to non apprehendet electos, a statement about what the second death does not do to them, and anguntur ("are afflicted") continues that thought where augentur changes the subject to the measure of their glory.

The rival, and why it lost. augentur can stand, meaning that the elect's state admits no increase — a real medieval commonplace. It lost on attachment: the qui clause is subordinate to non apprehendet, and an aside about the fixity of glory has no business there, whereas anguntur is the conclusion the main clause is driving at. The marker is additive, so a reader who prefers the rival still has Migne's word and its English on the page.

§3 · REFERENTS SUPPLIED, AND WHERE TWO READINGS BOTH CONSTRUE

Every "he/she/it/they" the Latin does not print was named before it was written. The ones a later reader should be able to check:

1. 0739D · quos attrahit ad se luxuria. luxuria is nominative or ablative, and both construe. Taken as ablative of means — "whom she draws to herself by wantonness" — because quae regnat has just set the harlot as subject and no new nominative is marked; ad se then points at her, which is what the imagery wants. The rival: luxuria nominative, "whom wantonness draws to itself". Nothing in the sentence excludes it. Not adjudicable from the file; recorded here rather than chosen silently.
2. 0740B–D · the beast and the devil. The lemmata are feminine (Ascensura est, bestia), the gloss masculine (recipiet, ascendet, habebit), because the gloss has identified the beast as diabolus. The English follows each as printed: "it" for the beast-lemmata, "he" for the devil in the comment. Not a defect; the identification is Migne's own.
3. 0741B · Carnes fornicariae manducare dicuntur. Unexpressed subject supplied as "they" = the ten horns of Apc 17:16 (hi odient fornicariam … carnes ejus manducabunt), which is the verse the lemma comes from, not anything in the sentence.
4. 0741D · Reges terrae … dolebunt, quod nunquam similes illis fuerint. illis is not determinable from the sentence: it may be the saints who never fornicated, or the kings who governed their earthliness well. Rendered "like to them", which keeps the Latin's own openness. Flagged, not resolved.
5. 0743B · In veteri lege non prohibuit se adorari. Subject is the angel (Apc 19:10, Vide ne feceris), not Christ; videns super se exaltatum hominem confirms it — the man exalted above him is the reason he now refuses. The English says "he" throughout and the antecedent is the lemma above.
6. 0744C · peccasse ostendit dum se immunem a peccato ille diabolus occidit. se is used loosely for eum: the subject of ostendit is Christ, and se is Christ as object of occidit. Read reflexively to diabolus the sentence says the devil killed himself, which it cannot mean. Rendered "he showed him to have sinned, in that that devil slew him who was free from sin", which keeps the Latin's own two "him"s unresolved on the page.
7. 0744C · Abyssus tenebrosa vocantur corda impiorum; vel ipse diabolus, quos Dominus… quos is plural after a singular diabolus; it reaches back past it to the ungodly. Rendered "whom", carrying the anacoluthon rather than repairing it.
8. 0750C · ipse etiam dat eis qui pararent eos quibus ipse reddat. Three pronouns, no antecedent printed for any of them. Rendered as literally as English allows; the clause is obscure in the Latin and is left obscure. Flagged.

§4 · PLATE-READ CANDIDATES AND CONSTRUCTIONS CARRIED AS PRINTED

No [sic:] was fired anywhere in this range, and that is deliberate — the ship test (brief addendum B) says an unread candidate is safe iff it fires no marker, and none of the candidates below can be settled from the file.

Plate-read candidates, in priority order

1. 0743C lignis for ignis and 0744C vide for vidi — see §1a. Both are one-letter differences on a real word in a scripture lemma; both are the shape of brief §4's normalization/corruption class, which is invisible from the Latin alone. A plate read settles both cheaply.
2. 0749C Vitrum per lucidum (Apc 21:21 tamquam vitrum perlucidum). A word split across a space is Pattern 10's shape and brief addendum A says a non-word raised as a plate candidate has been right every time — but "right" there meant our digitization's corruption, not Migne's, and a [sic:] publicly blames Migne. No marker. Rendered by sense, "Glass thoroughly clear.", which is also what the gloss requires (nullas sordes in se celantes). Note too that perlucidus written open is ordinary medieval practice and may be Pattern 9, not a defect at all.
3. 0748D obscurascit. A hapax in the whole Latin corpus (one occurrence, here); the standard inchoative is obscurescit. But it is a well-formed inchoative on obscurus, and addendum A is explicit that corpus frequency is near-worthless against a real word. Rendered "grows dark", no marker. Raised only so a plate reader can glance at it while reading the topazius paragraph for something else — which is how the normalization class has been found twice (addendum C).
4. 0741B tormentabuntur. Also a corpus hapax. tormentare is well attested medieval Latin; rendered "shall be tormented", no marker, recorded as a checked negative.
5. 0748D vel minori meritis fiunt. minori where minores is wanted. Rendered "or become less in merits". A real form, so no marker; worth a look if the page is opened.

Constructions carried as printed, not repaired

• 0740D–0741A · Adam, si in obedientia perseverasset, sensus corporis … verterentur. A nominative left hanging. Carried: "Adam, if he had persevered in obedience, the senses of the body would have been turned to no pleasure". Not smoothed into "Had Adam persevered".
• 0745A · Et non dico vivent et regnabunt in futuro, sed etiam in hoc praesenti. The non … sed etiam correlative with its solum elided. No "only" was supplied; the English reads "And I say not, they shall live and reign in the time to come, but also in this present time", which is as unbalanced as the Latin.
• 0748A · ut sit defensor aliorum … ut sit defensor aliorum. The same purpose clause twice in one sentence, eleven words apart. Both rendered; no [d:]. Pattern 11 is for a set run repeated; here a full clause intervenes and the second reads as a resumptive purpose clause, which a compositor's doubling would not. Recorded as a possible dittography a plate read could settle.
• 0744C · pseudo apostolorum set open, against pseudoprophetis set closed at 0741C in the same work. Rendered "false apostles" / "false prophets". No marker: Pattern 9 covers period orthography, and the open form is common.
• 0739D · quarum semen est umbilico — ablative of place with no preposition, against quorum seminarium in lumbis est in the same sentence. Rendered "whose seed is in the navel". Not a defect; recorded for the asymmetry.
• 0751B · the chapter's last sentence has NO terminal stop. …per meam resurrectionem in mane factam — Migne prints nothing after factam. Pattern 8: never supply. The English ends the same way. This is the range's most visible oddity on the page and it is his.
• 0750D–0752A · the anchor sequence jumps 0751B → 0752A, skipping 0751C–D. Runbook false positive 1 (band arithmetic): the A–D marks are positional quarter-guides, gaps are normal, not reported as a defect. Recorded only so a reader is not surprised.
• 0748B · Migne glosses no seventh stone. Sardius is the sixth and Beryllus the eighth, with chrysolithus (Apc 21:20) passed over in silence. The ordinals he does print — fourth, fifth, sixth, eighth, ninth, tenth (in denario numero), eleventh, twelfth — all agree with the Clementine's order, so nothing is displaced; the seventh is simply not commented. Abridgment, not loss. No [ed:].

usque ad — decided per occurrence (brief §D)

Four occurrences in 0018 (0748B ×3, 0748C ×1), all four the abridgment formula, all inside Bede excerpts opening etc., and all rendered as far as on the settled precedent (8985 §371, 8976 §191, from 8978 §205). No verb was supplied to complete any of them.

Two look-alikes in 0019 are the words' own, not the formula, and were rendered as ordinary prose: usque ad finem mundi (0749D, "even to the end of the world") and duxit filios Israel usque ad flumen Jordanis (0750B). One more in 0016: Peccata perveniunt usque ad contemptum Dei (0741D). Three ordinary uses, four formulas, seven occurrences, all decided in place.

§5 · THE FOUR TESTS, RUN

Test 1 — no [sic:] may swallow its word. Vacuous here: the range fires no [sic:]. The strip-and-read pass was still run over the [var:] and [cj:] sites, since the same failure is available to them. All six sentences read aloud complete with every marker removed; the [cj:] site keeps its verb ("are never in anything increased"), and each [var:] sits beside a lemma that is already fully Englished.

Test 2 / 2a — name the host, prove it from structure. Every printed negative in the range was walked. The population 2a fires on — a negative on a verb with a quantifier or correlative nearby — has eight members here, and all eight are governed by an explicit correlative, which is the only reason they can be called clean:

Simple negatives carried without incident, named so they are checkable: nihil timet · nec Deum diligant, nec poenas timeant (0740A) · ne desperetis (0740B) · nisi doctis (0740C) · amplius non habebit (0740D) · ad nullam voluptatem (0741A) · nunquam similes (0741D) · et vidua non sum (0742A) · nullus possit resistere (0743C) · nequeunt … nec possunt (0743A) · nisi bene pugnaveritis (0744A) · nunquam ad inferos descendere (0744D) · Caeteri vero mortuorum non vixerunt · omnes perfecti esse non possunt, nec in eadem claritate · in nullo unquam (0745A) · ne pro velle saeviat (0745B) · non fecerat mentionem · se non praedestinatos · quae agere noluerunt (0745C–D) · dubitabilis non est quin (0746A) · in nullo offendentes, quod non esse potest · non erunt ibi lacrymae … non erit causa (0746B) · nihil restare (0746B) · una non debet aliam excedere · ne nimis justus (0747C–D) · fides non marcescens (0748A) · quae non finitur (0749A) · ne fures intrent (0749D) · qui non (0750A) · Non sunt … alieni putandi (0750C) · ne ideo cesset (0751A) · non est labor (0752A). Count in, count out: no negative added, none dropped.

Test 2b — anchor position. All 50 anchors were re-checked against the word they stand beside in the Latin, not merely for presence. Three sit mid-construction and were kept there: Et [0742A] vidua non sum (inside an italic lemma, reproduced as "And [0742A] I am not a widow"), crucifixus [0748C] est ("Christ was crucified [0748C] on the sixth day"), and cum aere [0749B] mutatur. Four are the no-space form Migne sets — [0740A]VERS. 2.--, [0742B]VERS. 12.--, [0744A]VERS. 17.--, [0745A]VERS. 5.--, [0745B]VERS. 7.--, [0752A]VERS. 17.--, [0752B]VERS. 21.-- — reproduced closed up, as printed.

Test 3 — punctuation by band, never by total. This is the check that returned sites, and it returned twenty-three: every one an English semicolon or colon where Migne prints a comma, or the reverse. All twenty-three were repaired; the range now matches band for band on ;, : and ?, verified per chunk after the edits. ⚠ The totals would have hidden most of them — 0018 alone had a supplied colon at 0747C and a destroyed colon at 0747B, which is the Tobit failure exactly. No comma→semicolon sweep was made: Migne prints 33 semicolons in these five chunks himself, so the Tobit exception does not apply to this book.

Both ? marks are direct questions and ride as printed: Et isti, quare damnabuntur? (0743B) and Cum sim malus, quare me poeniteret? (0751A). No ? was supplied and none deleted, and no 8b recasting was needed — the indirect questions in the range (mirantur … quare Deus permittit illos exaltari, 0740C; non fecerat mentionem corporum, si deberent glorificari vel puniri, 0745C; Ne quaereret aliquis: Quid post illam innovationem futurum est, 0746B) all close with Migne's own comma or full stop, so nothing had to be moved behind a colon.

Test 5 — does the English ASSERT what he asserts? Read back paragraph by paragraph against the Latin with the counts already clean. Two things came out of it and neither is a containment failure:

• 0741B · perdet is rendered twice, two different ways — "whom he shall destroy" (quos perdet, of persons) and "this he shall lose" (hanc perdet, of the power). Both are proper senses of perdo and the object decides which; recorded because a reader comparing the columns will see one Latin verb against two English ones.
• 0741B · Desolatam facient / Solatium diaboli. The gloss is a pun: desolatam is glossed by solatium. The English keeps "desolate" against "solace" so the pun survives. Nothing is claimed about whether Migne's source intended it.

Em-dashes: none in this range. Nothing was restructured to avoid one and none is reported.

§6 · SMALLER RENDERING CALLS, RECORDED

• 0739C Angelis (from Apc 17:1 de septem angelis) rendered "Of the angels." — a case-form fragment, Ruth convention 3, the phrase it stands for rather than the bare noun. Meretricis (0739D) likewise "Of the harlot." Neither is a stump-vs-expanded inconsistency with the true one-word lemmata in the range (Auro · Ebriam · Fuit · Carnes · Mors · Amen · Prima · Magog · Lumen · Aurum · Jaspis · the eight stone names · Fluvium · Dominus · Fructus), which have nothing elided and are rendered as they stand. Naming the two classes before contrasting them, per the brief's closing rule.
• 0742D antonomastice rendered "by antonomasia" — a rhetorical term with a settled English form, not italicized as untranslatable Latin.
• 0745B detectos rendered "the uncovered", to hold the detectos / tegentes play the gloss builds Magog on.
• 0749A pan · Chryson · prasos are Greek words Migne sets roman, not italic; the English keeps them roman. Italicizing them would have added three italic spans the Latin does not have.
• 0750B · one connective supplied. …in quibus reficimur ultra flumen, id est, in futuro habebimus ipsum praesentem. Migne prints no mark between reficimur and ultra flumen, where the sense turns from circa flumen to ultra flumen. Rather than supply a mark (Pattern 8) the English supplies the word "and": "in which we are refreshed, and beyond the river…". A word, not a mark; recorded because it is the one place the English adds anything.
• Pattern 17. tibi / thou oughtest / si polis, obscuras are singular and take thou (0740C, 0742B, 0748D, 0750C); vobis, vos, cavete, providete vobis, debetis are plural and take you (0739C, 0742D, 0743A, 0744A, 0752A). The two plural scripture imperatives keep "ye" under the scripture-echo licence: Exsultate → "Rejoice ye" (0742B, 0742D), Venite → "Come ye", Et congregamini → "And be ye gathered together" (0744A).
• Pattern 4 [f:] does not apply to this work (runbook false positive 4): it is a lemma-and-gloss commentary and its [n:] citations are Migne's inline parentheticals. None was tagged.
• Pattern 15. All ten [n:] in the range are citations or attribution sigla — (Job. XL) · (Ezech. XVI) · (Joan. IV) · (Psal. XXXII) · (II Cor. III) · (Phil. III) · (BED.) ×2 · (ID.) ×2. No note in this range is editorial prose, so no [nt:] was made. A checked zero, not a skip.

✅ PLATE READ, 2026-08-28 — the two single-letter candidates of §1a, settled OPPOSITE ways

Both were flagged here as needing one letter decided at the plate. PL 114 p. 377 carries both, and it splits them:

• 0743C lignis was OURS. Migne prints ignis; the [var:] is withdrawn and the TEI patched. See the row in §1a.
• 0744C vide is MIGNE'S. He prints the imperative Et vide sedes, and the [var:] against Clementine Et vidi sedes stands, confirmed. His own gloss below re-quotes vidi three times (Ligato diabolo, vidi solutos fideles … Vidi in tempore ligationis … vidi statim in gloriam intrare), so the plate contradicts its own lemma — exactly what this edition exists to show, and not ours to tidy.

⭐ The lesson, and it is the argument for reading plates at all: these two were indistinguishable from the file. Same class, same shape, same evidence, flagged by the same reader in the same breath — and the answers were opposite. Corpus frequency and Vulgate collation could not have separated them; only the plate did.