These are the edition's working notes, published as they were written. They record, work by work, where Migne's plate is defective or the reading uncertain: the printed form, what the English does with it, and why. They are not a groomed apparatus. Where a note argues with itself, withdraws an earlier decision, or leaves a question open, that is the record of how the reading was actually reached, and it is left standing — a conjecture that is visible can be checked, and one that has been tidied away cannot.
The English never silently repairs the page. A construable but wrong reading is rendered literally and the conjecture is recorded here, never applied to the text. How the English is produced and checked →
Translated 2026-08-18, one translator, 1 chunk (1,163 Latin words → 1,772 English), on Opus. Structure: ARGUMENTUM, CAPUT UNICUM. Apparatus: 3 [var:] · 0 [sic:] · 0 [cj:] · 0 [d:] · 0 [ed:] · 0 [nt:] · 4 [n:] (exact parity with the Latin twin). 14 column anchors, 45 italic spans, 0 question marks — all at parity. Zero guillemets in either language.
Conventions follow src/english/8968/cruces.md §1 (Ruth, the Glossa pilot): VERS. n.-- passes through verbatim; the lemma is Englished; the sigla ride as ordinary [n:] notes.
§1 · ⛔ MIGNE'S OWN SEQUENCE — VERS. 23 BEFORE VERS. 21. NOT REORDERED.
The lemmata run … 19 · 20 · 23 · 21 · 25. That is his ordering on the plate and it is reproduced exactly: nothing reordered, nothing renumbered, no [ed:] (no text is lost).
⭐ But the "21" is almost certainly a misprint for "24", and the words prove it. The lemma printed under VERS. 21 is In exsultatione — and Jude 21 contains no such phrase (vosmetipsos in dilectione Dei servate, exspectantes misericordiam …). in exsultatione is Jude 24: constituere ante conspectum gloriae suae immaculatos in exsultatione. Two further things point the same way:
- The gloss's own verb is
constituendos— dicit constituendos in exsultatione — which is Jude 24's constituere, not anything in Jude 21. - It opens by looking back at the previous gloss: Quos superius monebat in timore Deo servire — i.e. at VERS. 23's In timore. So the gloss itself asserts it follows 23.
Read as 24, Migne's sequence is 19 · 20 · 23 · 24 · 25 — continuous and in order. The defect is one digit in a verse address, not a displaced gloss.
Nothing was changed in the text. Verse addresses pass through verbatim by the settled convention, and a renumbering would be exactly the silent repair Pattern 7 forbids — the more so because the reader's own Latin column would then disagree with our English. It is recorded here instead.
§2 · THE LEMMA DIVERGENCE LIST — ALL NINETEEN CHECKED AGAINST THE VULGATE
Every verse lemma in the work, plus every second (mid-paragraph) italic lemma. Checked word by word against the Clementine Vulgate of Jude before the English was written. Agreements are listed too, so this can be told apart from a report by someone who only listed findings.
2a · The three that diverge — all marked [var: …]
| col | Migne prints | Vulgate | marker + reasoning |
|---|---|---|---|
| 0707A | VERS. 1 Consolatis et vocatis. | conservatis, et vocatis (Jude 1) | ⭐ [var: Vulg. conservatis et vocatis, "to them that are preserved and called"]. See §3. |
| 0708A | VERS. 9 non est ausus inferre poenam blasphemiae | non est ausus judicium inferre blasphemiae (Jude 9) | [var: Vulg. judicium inferre blasphemiae, "bring in a judgment of blasphemy"]. Material: poena is a penalty imposed, judicium a sentence pronounced — and the whole gloss turns on what Michael/Moses durst not do. Rendered "durst not bring in the penalty of blasphemy". ⚠ Note the launch brief itself quoted this lemma in its Vulgate form (judicium inferre); the plate does not read that. This is the 7a″ trap in its purest shape and the reason the words get named. |
| 0708C | VERS. 12, 2nd lemma Arbores autem. | arbores autumnales (Jude 12) | [var: Vulg. arbores autumnales, "autumn trees"]. Rendered "But trees." — the printed words. Near-certainly a compositor's contraction of autumnales to autem: the gloss's very next clause reads sed sunt arbores autumnales. ⛔ Rendering the lemma "autumn trees" would have been a textbook silent repair, and it is the reading an eye that knows Jude reaches for first. |
2b · The sixteen that agree exactly (checked, no marker)
Misericordia vobis (2) · Charissimi (3) · Gratiam transferentes (4) · Omnia (5) · Quoniam Jesus (5, 2nd) · Angelos vero (6) · Non servaverunt suum principatum (6, 2nd) · Vinculis aeternis (6, 3rd) · Sicut Sodoma (7) · Dominationem autem spernunt (8) · Cum Michael (9) · Vae illis (11) · Maculae (12) · Nubes (12, 2nd) · Bis mortuae (12, 4th) · Sidera (13, twice) · Prophetavit (14) · Enoch (14, 2nd) · Ecce venit Dominus (14, 3rd) · Hi sunt (16) · Spiritum non habentes (19) · Superaedificantes vosmetipsos (20) · In Spiritu sancto orantes (20, 2nd) · Illos vero salvate (23) · In timore (23, 2nd) · Odientes eam (23, 3rd) · In exsultatione (printed under 21; agrees with Jude 24 verbatim — see §1) · Gloria et magnificentia (25).
Notes on three of these that were looked at twice and cleared:
- VERS. 3 Charissimi — Charissimi for Carissimi is an attested orthographic variant, Pattern 9, not a defect. No marker, no italics-carry.
- VERS. 3's four inner lemmata — habui necesse, · scribendi vobis · de vestra salute communi, · faciens omnem sollicitudinem scribendi vobis, — every word is Jude 3's, but the order differs (Vulg. necesse habui; de communi vestra salute). No
[var:]: Pattern 14 fires on a divergence that changes the sense, and a transposition does not. Logged so it is visible that it was seen and let alone. - VERS. 11's Beda excerpt — In viam Cain abeunt where Jude 11 has in via Cain abierunt. Case and tense both shift, but this is Beda's sentence rebuilding the verse into his own syntax, not a proof-text quoted for its words.
[var:]declined.
§3 · 0707A · Consolatis — the work's most interesting word, and the gloss decides it
Migne prints VERS. 1.-- Consolatis et vocatis. The brief flagged the Vulgate as reading dilectis. That is half the story, and the half that matters is the other one: Jude 1 has BOTH — his qui sunt in Deo Patre dilectis, et Christo Jesu conservatis, et vocatis. The word standing immediately before et vocatis is conservatis, and Consolatis is one letter-run away from it.
⭐ The gloss settles it internally, and this is the evidence, not a hunch. The single sentence under this lemma is Vocat, conservat, diligit filios — three verbs answering the verse's three participles, in the verse's own order reversed:
| gloss verb | Jude 1 participle |
|---|---|
| Vocat | vocatis |
| conservat | conservatis |
| diligit | dilectis |
The glossator was reading conservatis. There is no verb answering consolatis anywhere on the page.
What was done, and what was declined:
- Rendered what Migne prints — "To them that are consoled and called." consolatis is a real form (consolor), so it is translatable and Pattern 7 governs; there is nothing for a
[sic:]to wrap (Pattern 12 marks only carried, untranslatable type). [var: Vulg. conservatis et vocatis, "to them that are preserved and called"]stands beside it, per the brief and Pattern 14.[cj: …]was considered and DECLINED. Pattern 18's shape fits — a real word whose faithful English misleads — but its purpose is to put the conjecture where a reader will meet it, and the[var:]already does that, naming conservatis in the same breath. Two markers on four words would be apparatus for its own sake. ⚑ If a later reader thinks the conjecture should carry Migne's own voice rather than the Vulgate's, this is the site.- The gloss's conservat is rendered "preserves" so that the
[var:]'s "preserved" and the gloss's verb are the same English word and the reader can see the join.
§4 · MIGNE'S PRINTING, LEFT STANDING
0707D · coarctabit Vel vinculum — a missing full stop, carried
Vel sententia divina quae eos in perpetuum coarctabit Vel vinculum est cupiditas infimarum rerum… — no stop between the two Vel clauses, and the second Vel is capitalized. The English prints it the same way: "…will confine them for ever Or the chain is the desire…". Pattern 8: never supply a mark the plate lacks. It reads as an error because it is one, and it is his. No [ed:] — nothing is missing, only unpunctuated.
0705D–0706D · the ARGUMENTUM's column jump
The argument opens at [0705D] and takes [0706D] mid-sentence (de sub | jugo). Checked against runbook false-positive 1 (column-band arithmetic) and not reported: no text is broken mid-word, and the anchor falls between preposition and noun exactly as a column break does. The English carries the anchor in the same place.
0709A · qui reprobi … dedicant — a relative with no antecedent
in electorum vero progenie, Enoch, septimus memoratur, qui reprobi in hac vita, quae ante est, seipsos aedificando dedicant. The plural qui has no plural antecedent: the two Enochs are singular. Rendered with reprobi in apposition — "who, the reprobate, in this life, which is before, dedicate themselves by building themselves up" — which is the one construction that uses every printed word. The alternative is qui read as causal (quia), "for the reprobate …"; it construes as well and is NOT excluded. Nothing in the antithesis (reprobi … electi vero) turns on which is right. No marker: the words are legible and Latin, and no marker in this corpus means "a relative pronoun with nothing to attach to".
0709A · ab aliis invident — invideo with an ablative
Avari enim (quibus nihil sufficit) etiam ab aliis invident. Invideo governs the dative; ab + ablative will not construe here on any reading that keeps the sentence's sense (the covetous begrudge other men their good). Rendered "are envious even of others". A literal "envy from others" is not English and asserts nothing. Logged rather than marked, for the Ruth 0534A reason: nothing is broken type, nothing is missing, and inventing a marker for one site is worse than saying so here.
0709A · quo congregatur Ecclesia spiritualis efficitur — two verbs, no conjunction
Printed without et and without a comma between the clauses. Rendered as printed: "(by whom the Church is gathered together, is made spiritual)". The obvious supply is et; it is not supplied, and no conjunction is invented — the 8968 effoetandi precedent, where the first draft's invented "and" was caught by both blind readers.
0709B–0710A · petite — an imperative inside a ut clause
Ut de vestris viribus non praesumatis, sed … non per vos salvari putantes, sed adventum Spiritus sancti in vos petite. The construction breaks off from subjunctive to imperative mid-sentence. Carried as printed: "…not thinking to be saved through yourselves, but ask ye for the coming of the Holy Spirit". Migne's anacoluthon, not smoothed into a fourth subjunctive.
0710B · imputare — infinitive for imperative
Deo, non autem vobis imputare. Rendered as the imperative it functions as ("impute it to God, and not to yourselves"); the infinitive-for-imperative is ordinary medieval usage (Pattern 9), not a defect, and no verb is supplied that the sentence lacks.
0708C–0708D · Sidera. twice — NOT dittography
Two consecutive glosses are keyed to the same lemma Sidera (13). Considered under Pattern 11 and rejected: the two glosses are different (apostasy/fallen angels; then instability and shipwreck), so this is two glosses on one lemma, the Glossa's normal habit, not a doubled run of type. No [d:].
§5 · THE UNEXPRESSED SUBJECTS — EVERY SUPPLIED PRONOUN, NAMED
Jude's glosses shift "they" constantly. Each supplied pronoun below, with what fixes it.
| col | English | who, and why |
|---|---|---|
| 0707C | "he shows how great is the greatness of the Lord" | Jude the apostle. The paragraph describes the epistle's own procedure (Confutatis haereticis … ostendit); the same third-person frame runs through 0708A, 0708C and 0710A. |
| 0707C | "he spared not the angels that sinned, neither will he spare" | God. Si angelis peccantibus non pepercit is 2 Pet 2:4's sentence; the sparer of angels can be nobody else, and the next clause's parcet keeps the subject. |
| 0707D | "he will condemn them grievously before the judgment" | God, continuing pepercit/parcet above; the object is the homines qui non servant principatum of the same sentence. ⚠ Migne shifts subject mid-sentence with no marker (from relinquunt, they, to damnabit, he); the shift is his and is carried, not smoothed. |
| 0708A | "He had given an example of the damnation" | Jude the apostle — Dederat … dat etiam exemplum describes the epistle's argument moving from v. 5–6 to v. 7. |
| 0708B–C | "He compares himself to preachers" (Comparat se) | the heretic, collective singular. The verb is singular and the next clause is plural (sed haeretici sunt nubes). The parallel gloss five lines later prints its subject — Impii et pseudoapostoli comparant se arboribus — which is what settles it: same verb, same figure, subject named. Rendered singular, as printed. |
| 0708C | "The tree dies once which bears no fruit" | arbor, feminine, from the lemma above (Arbores … Bis mortuae, both feminine plural); quae non facit fructum has no other antecedent. Supplied because English cannot leave it bare. |
| 0708C | "but they are also the spots of those" (maculae illorum) | ⚠ left ambiguous on purpose. illorum may be the comessationes et ebrietates just named (Jude 12's in epulis suis maculae) or the alios they destroy in the next clause. Both construe; the gloss does not decide, and neither does the English, which keeps the bare genitive. |
| 0708D | "he signifies the apostates" | Jude the apostle, same frame as 0708A. |
| 0710A | "he says are to be established in exultation" | Jude the apostle; Quos superius monebat refers back to VERS. 23's In timore, so the subject is the one who admonished there. See §1 — this is also the evidence that the gloss belongs after 23. |
§6 · RENDERING DECISIONS WORTH A SECOND EYE
usque ad→as far as. Migne's abridgment formula (etc., usque ad … = quote the head of Beda's sentence, skip, resume at these words). Englished, in its own italic span, per the Glossa convention that the lemma is Englished; four instances, all identical. Leaving it in Latin was considered — it is an editorial formula, not the author's prose — and rejected: "as far as" is a clean equivalent, and the Latin stands opposite in the parallel column. This is the first Glossa book in the corpus to carry the formula in English; if a later book should differ, this is the site that set it.Judas apostolus→ "Jude the apostle." Latin has one name for two men; English has two. Corpus precedent decides it: the shipped English already reads "James and John, and Jude, wrote their Epistles" for exactly this figure, and keeps "Judas" for Iscariot and Maccabeus.animales→ "animal", not "sensual" or "natural." The gloss's whole antithesis is animalis ⟷ spiritualis (Spiritum sanctum … non habent, ideo animales), and "animal" keeps the anima root visible. "Sensual" (Douay) and "natural" (AV) would be conforming the gloss's own keyword to an English Bible — the 7a″ class.coagulum charitatis→ "the rennet of charity." Coagulum is the curdling agent, and the verb it answers is defluunt, "they flow away": without the rennet the mass never sets. "Curd" (the result, not the agent) and "bond" (a paraphrase) were both rejected.pseudo→ "the false ones." propter pseudo qui vobis se ingerunt — pseudo used substantivally, masculine plural by its relative. Attested medieval usage, Pattern 9, so no carry-through and no marker. "False apostles" was declined as importing apostoli, which the gloss does print four lines later (pseudoapostoli) and does not print here.Sodomorum→ "the Sodomites" (genitive of Sodomi, the men), not "of Sodom" — Migne's own incendium Sodomorum against the lemma's place-name Sodoma.ut revocat(0707C) — ut with the indicative, where a purpose clause wants the subjunctive. Rendered "as he calls back", which is what the indicative allows. Rendering it "that he may call back" would silently mend the mood.
§7 · READ CLEAN
Read and cleared with no finding, in both directions (nothing dropped, nothing supplied):
- ARGUMENTUM (0705D–0706D) — one sentence, rendered whole; illicitum … dissertet kept as an argument, not softened to a statement.
- VERS. 2 (0707A) — in nullo carried ("in nothing").
- VERS. 4 (0707B) — the comparative pair tanto licentius … quanto minus checked in both members and rendered "so much the more licentiously as they the less see themselves examined" (Pattern 7a mechanism 4, comparative reversal — the one this shape invites). Not reversed.
- VERS. 5 (0707C) — the negation cluster non habetis opus … non est opus ut audiatis: two negatives, both carried.
- VERS. 6 (0707C–D) — the densest negation run in the work: non sunt imitandi · non verum Deum · non pepercit · nec … parcet · non servaverunt · non servant · non possunt. Seven printed negations, seven in the English. None added.
- VERS. 8 (0708A) — nolendo · nullum · nec: three carried.
- VERS. 9 (0708A–B) — non est ausus and noluit, the two negatives the brief flagged; both carried, and neither smoothed into a positive by the surrounding negative vocabulary.
- VERS. 11 (0708B), VERS. 16 (0709A) — read clean; nihil sufficit carried.
- VERS. 12–13 (0708B–D) — the figure-chain checked one by one and each kept its own image: nubes sine aqua clouds without the water of wisdom, arbores autumnales autumn trees, sidera errantia stars that never continue in the same state (no "planets", no "comets" — Migne says neither).
- VERS. 20 (0709B) — four negations (non praesumatis · non … putantes · non habent · ne … segregemini), four in the English.
- VERS. 23 (0710A) — the salvate/odite antithesis held opposite: Eos qui non sunt damnati … docete (save) against Non corpus nostrum odisse debemus, sed maculatum esse (hate the stain, not the body). Both negations carried; the two attitudes stay opposed.
- VERS. 25 (0710B) — non autem vobis · nolentibus · nihil resistit: three carried; tam volentibus quam nolentibus kept as a pair.
- Punctuation: 0 question marks in the Latin, 0 in the English. Pattern 8b never engaged — the work contains no interrogative of any kind. Sentence-terminal stops are at parity paragraph by paragraph, the 0707D missing stop (§4) included.
§8 · BLIND POLARITY READ — 2026-08-18, and the adjudication
A fenced reader (Latin twin + English only, never this file). 5 sites raised · 3 confirmed and repaired · 2 recorded without repair.
Zero polarity defects, and the census is named — every negative particle checked in both directions: ARGUMENTUM illicitum · VERS. 2 in nullo … dissentiatis · 4 quanto minus vident · 5 non habetis opus / non est opus · 6 non sunt imitandi · non verum Deum · non pepercit, nec … parcet · non servant · non possunt · 8 nullum habentes timorem, nec fidem · nolendo · 9 non est ausus … / noluit aliquam blasphemiam inferre · 12 Non solum … sed etiam · non facit fructum · non bona intentione · 13 nunquam · 14 Non solum, inquit · 16 nihil sufficit · 19 non habent ×2 · non providentes · 20 non praesumatis · non per vos salvari · ne … segregemini · 23 non sunt damnati · Non corpus nostrum odisse debemus · non nostri arbitrii · 25 non autem vobis · nihil resistit. VERS. 9's stacked negation is intact, aliquam included.
Zero silent repairs — and this work was the most exposed in the batch. Every place the memorized English pulls, the translation held to the plate: sine aqua sapientiae → "without the water of wisdom", not "clouds without water"; Sidera. twice → "Stars", not completed to "wandering stars"; in hoc mari vitiorum → "in this sea of vices", no fluctus feri maris imported; prope cadunt in ignem → "almost fall into the fire", no snatching out of it; animales → "animal", not Douay's "sensual"; Non servaverunt suum principatum keeping Migne's suum against Douay's bare "their"; Bede's In viam Cain abeunt keeping in viam and the present tense against the Vulgate's via … abierunt.
✅ 0708B–C · VERS. 12 — CONFIRMED, REPAIRED. The dominant class again, and the plate itself supplies the answer.
Printed: Nubes. Comparat se praedicatoribus pluentibus justitiam, coruscantibus per miracula, sed haeretici sunt nubes obscurantes solem…
Was: "Clouds. He compares himself to preachers raining justice…" Now: "Clouds. The heretic compares himself to preachers raining justice…"
Comparat prints no subject, and the running "he" of this work is Jude in twelve other places (informat, dissertet, Dederat, dat, ostendit ×2, significat, monebat, dicit, inquit, ostendens, noluit) — including the sentence immediately before it. So the default English reading was that Jude likens himself to justice-raining preachers, which his own sed haeretici sunt nubes contradicts three words later.
⭐⭐ What settled it is the corpus-is-its-own-control move, inside a single paragraph: the exactly parallel gloss five lines down prints its subject outright — Impii et pseudoapostoli comparant se arboribus bonum fructum ferentibus, sed sunt arbores autumnales — and our English had rendered that one correctly ("The ungodly and the false apostles compare themselves…"). Same construction, same sed-contrast, one with the subject printed and one without. The author's own repetition is the witness.
✅ 0707C · VERS. 5 lemma — CONFIRMED, REPAIRED. A complementizer read as a causal.
Quoniam Jesus. → was "For Jesus.", now "That Jesus."
In Jude 5 quoniam is the content-clause complementizer after scientes semel omnia — "knowing all things once, that Jesus, having saved the people out of the land of Egypt…". "For Jesus." asserted a causal connective the verse does not carry, and as an English string it reads as a prepositional phrase rather than a clause opener — a lemma that looked like a caption and meant nothing.
✅ 0709A · VERS. 19 — CONFIRMED, REPAIRED. A supplied comma that adjudicated.
Migne prints (quo congregatur Ecclesia spiritualis efficitur) with no mark between Ecclesia and spiritualis efficitur. Our English had supplied a comma. Removed.
⚑ The comma did more than tidy: it forced Ecclesia as subject of both verbs and stranded spiritualis as predicate of efficitur, closing off the reading quo congregatur Ecclesia spiritualis, efficitur — a choice made silently by punctuation. ⭐ The inconsistency was the tell: the same chunk carries 0707D's unpunctuated juncture (coarctabit Vel vinculum) raw and deliberately, so one class was being treated two ways in one work.
⛔ 0708B · VERS. 11 — DECLINED, both readings recorded
Quia per omnia haeretici monstrati sunt a verbo veritatis cecidisse, ostendit eos diversis malis subjectos. Two constructions genuinely construe: (A) ours, cecidisse a verbo veritatis, "fallen from the word of truth"; (B) monstrati sunt a verbo veritatis, "have been shown by the word of truth to have fallen" — the ab-phrase sits in the agent position immediately after the passive.
Left as (A), on the ground the reader did not weigh: under (B), cecidisse is left with no complement at all — fallen from what? — while (A) completes it, and cadere/excidere a fide is the standing idiom. Word order favours (B), sense favours (A). Named here so the choice is on the record rather than invisible.
⬜ 0708C · [var:] #3 — the marker STANDS, with the counter-evidence recorded beside it
The reader audited all three [var:] markers in both directions and found the first two sound (Consolatis for conservatis; poenam for iudicium, sense-changing, English following Migne in each). On the third — Arbores autem against the Vulgate's arbores autumnales — it raised a real objection: Migne's own text prints autumnales one clause later (sed sunt arbores autumnales: tarde enim ferunt bonum fructum sicut arbor in autumno plantata), and the whole exposition depends on it. So the plate may not carry a variant reading of Jude 12 at all, but an abbreviated or mis-set lemma head (aut͞m for autumn.) which the exposition immediately corrects.
The marker stands. Pattern 14 records what the plate prints against what the Vulgate prints, and both halves of that are true here; it never speculates about why they differ, so a compositor's hand is not grounds for removing it. But the reader is right that a [var:] implies a textual divergence where the evidence on the same line points elsewhere, so the counter-evidence is recorded here where a later eye will find it. ⚑ The English rendering itself — "But trees", the shorter form not silently completed — is correct and is exactly the trap the brief named.
Read clean, and named
~45 italic spans checked against the plate; 14 column anchors present in both files at corresponding points (0708C and 0709B shift a few words because English cannot split "their mouth" / "the rennet of charity"); 4 [n: (BEDA.)] verbatim and unanglicized; ? and ! 0 : 0 in both directions with every colon on the plate surviving (VERS. 3 ×3, 5, 7, 8 ×2, 9 ×2, 12, 14); every Pattern 17 site correct (tibi → "thee" at VERS. 9; every vos/vobis/vestris/petite/segregemini → "you/ye").
Migne's verse numbering was checked, not merely left alone: the sequence 19 · 20 · 23 · 21 · 25 stands as printed, and the English around the misnumbered gloss is correct — its lemma In exsultatione expounds Jude 24, which is why his number is off, and the translation neither corrects it nor stumbles over it. Likewise 0707D's unpunctuated coarctabit Vel vinculum rides raw, by design.
Every unexpressed subject traced: informat/dissertet (Jude), impleat/est reconciliatus (God), revocat/ostendit/ostendens (Jude, with ipse = the Lord correctly distinguished), damnabit (God), Dederat/dat, ostendit at VERS. 11, significat, nascitur/memoratur, monebat/dicit, noluit (the Lord Jesus) — all correct but Comparat. One rejected: VERS. 9's ut permitteret illos sacrificare, where the subject must be Pharaoh, not Moses, and our "that he might permit them" is exactly as underdetermined as the Latin — Pattern 7 obeyed, nothing resolved wrongly.
Rejected with reasons: the 0705D → 0706D band jump (false positive 1 — de sub [0706D] jugo … servitutis construes across the anchor, unbroken); purgantur scelera … et eleemosynae fructus (the ambiguity is Migne's, and his comma is preserved exactly); invidere ab (odd Latin, no English exponent); propter pseudo → "the false ones" (attested substantival use, Pattern 9); qui reprobi in hac vita where the sense wants quia (qui is what is printed); and four word-order or connective differences correctly carrying no [var:] under Pattern 14's sense-change threshold — Gloria et magnificentia (VERS. 25), de vestra salute communi (VERS. 3), de corpore Moysi (VERS. 9), and Bede's in viam Cain abeunt.
⚑ Plate: checked zero
The full-width foot bands of leaves n357–n359 of archive.org patrologiaecurs06saingoog (PL 114; leaf n = (column + 9) / 2, calibrated against the running head at n330 = cols 651–652) cover this work's columns, including n357 (the 3 John / Jude division) and n359 (the Jude / Apocalypse division), both falling mid-page — the case that lost three notes on PL 202. Migne prints no foot-of-page apparatus there: no foot rule, no numbered notes, the columns simply end. Recorded in data/plate-notes/coverage.json as tier 3, coverage full, notes_recovered: 0.