Chrestomathia I, 'Quattuor prima capita Geneseos', OPENING PAGE: the section head
CHRESTOMATHIA. / I. QUATTUOR PRIMA CAPITA GENESEOS. / CAPUT I. occupies the upper half,
so only 12 lines of Syriac stand here (against 20 on a full page). No Syriac page numeral is
printed on this opening page. Signature E* at the foot. Marginal line numbers 1, 5, 10 in
the right (outer) margin; the piece's line numbering starts here and runs to 55 at the end of
Caput I on p. 70. The page breaks mid-verse 9, after ܠܰܐܬܪܳܐ ܚܰܕ.
⛔ VERSE-NUMBERING FAULT, and it is the page's chief finding. Nestle prints ¹ before ܒܪܺܫܝܬ and ² before ܘܰܐܪܥܳܐ (both correct), then NO numeral at the head of v. 3 (ܘܶܐܡܰܪ ܐܰܠܳܗܳܐ ܢܶܗܘܶܐ ܢܘܗܪܳܐ) — only a faint, plainly under-inked round mark stands there, verified at 10x — and then prints ³ before ܘܰܚܙܳܐ and ⁴ before ܘܰܦܪܰܫ, i.e. against the two halves of what Barnes numbers as the single verse 4. The sequence re-synchronises at ⁵ (ܘܰܩܪܳܐ) and is standard thereafter (⁶ at the firmament, ⁹ at the gathering of the waters). Recorded AS PRINTED; the numerals are not corrected. See [alignment].diff.
Fully vocalized Serto, with seyame and the resh/dalath points. ⭐ This CONTRADICTS the Phase-0 section default (unvocalized Estrangela, taken from the n180 probe) and it also refines MAP.md: the ramp is NOT a one-way drift from Serto to Estrangela across pp. 67–71. See layer_finding on this record and the summary in c070-1.
Barnes, PENTATEUCHUS SYRIACE (post Samuelem Lee), BFBS London 1914 — archive.org ktavadeauritaauk00lees, title page verified at n6. PD. ⭐ PAGINATION CALIBRATED THIS PASS: printed page = leaf − 9 (n10 = p. 1, Gen 1:1–13, signature B; n11 = p. 2, running head ܣܦܪܐ ܕܒܪܝܬܐ + I. 14—28; agreement on two independent pages, per MAP.md's own method). Barnes prints unvocalized Estrangela with seyame and diacritic points, verse numbers in the outer margin against the line on which a verse opens, * as verse divider and ❖ at the close of a paragraph. See research/syriac-peshitta-editions.md
Read from the plate, line by line, off 3.0–5.5x crops cut by horizontal projection (one strip per 2–4 printed lines); the 1.75x half-crops of tools/quarry_fetch.sh are NOT sufficient for pointed Serto and were used only for orientation. Consonantal skeleton, seyame, word division, punctuation, verse numerals, marginal line numerals and square-bracket supplements are read directly. ⚠ HONEST LIMIT, per the never-read-codepoints-by-eye rule: the West-Syriac vowel points are legible AS SHAPES at this resolution (zqofo prints as a small circle, ptoho as an open triangle, and so on — calibrated on ܐܰܠܳܗܳܐ, which recurs on every page), but the finest layer — rukkakha/qushshaya bars, the ܗ̇ point, linea occultans — is recorded where seen and is not warranted point-by-point. The vocalisation below is therefore a plate reading at the word level, not a certified diacritic-level edition; a blind control on a sample ([[reference_arabic-control-rule]]) is still owed before this text is shipped.
⭐ THE RAMP IS BY CAPUT, NOT BY PAGE — this corrects both Phase 0 and MAP.md. Across the whole of Chrestomathia I the script alternates with the chapter: Caput I (pp. 67–70) vocalized Serto · Caput II (pp. 70–73) unvocalized Estrangela · Caput III (pp. 73–76) vocalized Serto · Caput IV (pp. 76–78) unvocalized Estrangela. MAP.md read n156 as Serto and n160 as Estrangela and inferred a one-way drift; the truth is a deliberate alternation, and three pages (70, 73, 76) carry BOTH layers because a Caput boundary falls inside them. Declare the layer per record AND, on those three pages, per section within the record.
¹ ܒܪܺܫܝܬ ܒܪܳܐ ܐܰܠܳܗܳܐ ܝܳܬ ܫܡܰܝܳܐ ܘܝܳܬ ܐܰܪܥܳܐ. ² ܘܰܐܪܥܳܐ ܗܘܳܬ ܬܘܗ ܘܒܘܗ: ܘܚܶܫܘܟܳܐ ܥܰܠ ܐܰܦ̈ܰܝ ܬܗܘܡܳܐ: ܘܪܘܚܶܗ ܕܰܐܠܳܗܳܐ ܡܪܰܚܦܳܐ ܥܰܠ ܐܰܦ̈ܰܝ ܡܰܝ̈ܳܐ. ❖ · ܘܶܐܡܰܪ ܐܰܠܳܗܳܐ: ܢܶܗܘܶܐ ܢܘܗܪܳܐ: ܘܰܗܘܳܐ ܢܘܗܪܳܐ. ³ ܘܰܚܙܳܐ ܐܰܠܳܗܳܐ ܠܢܘܗܪܳܐ ܕܫܰܦܺܝܪ. ⁴ ܘܰܦܪܰܫ ܐܰܠܳܗܳܐ ܒܶܝܬ ܢܘܗܪܳܐ ܠܚܶܫܘܟܳܐ. ⁵ ܘܰܩܪܳܐ ܐܰܠܳܗܳܐ ܠܢܘܗܪܳܐ ܐܺܝܡܳܡܳܐ. ܘܰܠܚܶܫܘܟܳܐ ܩܪܳܐ ܠܺܠܝܳܐ: ܘܰܗܘܳܐ ܪܰܡܫܳܐ ܘܰܗܘܳܐ ܨܰܦܪܳܐ ܝܰܘܡܳܐ ܚܰܕ. ❖ ⁶ ܘܶܐܡܰܪ ܐܰܠܳܗܳܐ: ܢܶܗܘܶܐ ܪܩܺܝܥܳܐ ܒܰܡܨܰܥܰܬ ܡܰܝ̈ܳܐ. ܘܰܗܘܳܐ ܦܳܪܶܫ ܒܶܝܬ ܡܰܝ̈ܳܐ ܠܡܰܝ̈ܳܐ. ⁷ ܘܰܥܒܰܕ ܐܰܠܳܗܳܐ ܪܩܺܝܥܳܐ: ܘܰܦܪܰܫ ܒܶܝܬ ܡܰܝ̈ܳܐ ܕܰܠܬܰܚܬ ܡܶܢ ܪܩܺܝܥܳܐ: ܘܒܶܝܬ ܡܰܝ̈ܳܐ ܕܰܠܥܶܠ ܡܶܢ ܪܩܺܝܥܳܐ. ܘܰܗܘܳܐ ܗܳܟܰܢܳܐ. ⁸ ܘܰܩܪܳܐ ܐܰܠܳܗܳܐ ܠܰܪܩܺܝܥܳܐ ܫܡܰܝܳܐ. ܘܰܗܘܳܐ ܪܰܡܫܳܐ ܘܰܗܘܳܐ ܨܰܦܪܳܐ ܝܰܘܡܳܐ ܕܰܬܪܶܝܢ ❖ ⁹ ܘܶܐܡܰܪ ܐܰܠܳܗܳܐ: ܢܶܬܟܰܢܫܘܢ ܡܰܝ̈ܳܐ ܕܰܠܬܰܚܬ ܡܶܢ ܫܡܰܝܳܐ ܠܰܐܬܪܳܐ ܚܰܕ:
No footnotes on this page. The only apparatus is the running section head and CAPUT I.;
the · printed between ❖ and ܘܶܐܡܰܪ on line 3 is a broken/under-inked sort, not a mark of
Nestle's own (see work_note).
⭐ CHECKED against Barnes p. 1 = leaf n10 (Gen 1:1–13), read at 0.85x and again at page scale.
AGREEMENT, word for word, across the whole page: ܒܪܝܫܝܬ ܒܪܐ ܐܠܗܐ ܝܬ ܫܡܝܐ ܘܝܬ ܐܪܥܐ · ܘܐܪܥܐ ܗܘܬ ܬܘܗ ܘܒܘܗ · ܘܚܫܘܟܐ ܥܠ ܐܦ̈ܝ ܬܗܘܡܐ · ܘܪܘܚܗ ܕܐܠܗܐ ܡܪܚܦܐ ܥܠ ܐܦ̈ܝ ܡ̈ܝܐ · ܘܗܘܐ ܦܪܫ ܒܝܬ ܡ̈ܝܐ ܠܡ̈ܝܐ · ܕܠܬܚܬ / ܕܠܥܠ ܡܢ ܪܩܝܥܐ · ܘܗܘܐ ܗܟܢܐ · ܝܘܡܐ ܕܬܪܝܢ · ܢܬܟܢܫܘܢ ܡ̈ܝܐ … ܠܐܬܪܐ ܚܕ. No lexical or word-order divergence was found.
⭐ ONE DIVERGENCE, and it is in the APPARATUS, not the text: the verse numbering (see work_note). Barnes's margin reads 1 / 3 / 4 / 5 / 6 / 7 / 8 / 9 …, with 3 against the line that opens ܘܐܡܪ ܐܠܗܐ ܢܗܘܐ ܢܘܗܪܐ and 4 against ܘܚܙܐ ܐܠܗܐ ܠܢܘܗܪܐ ܕܫܦܝܪ (whose second half, ܘܦܪܫ ܐܠܗܐ, carries no number of its own). Nestle's 3 and 4 sit one clause later than Barnes's. Since the two editions' TEXT is identical here, this cannot be an edition-level difference in versification carried by the exemplar; it is a compositor's fault in Nestle 1889, and the under-inked mark at the true place of v. 3 is the direct evidence for it.
⚠ Not adjudicated: pointing. Barnes is unvocalized, so it can confirm the consonantal text and the seyame and it cannot check one vowel on this page. The vocalisation here stands on the plate reading alone (see transcription_method).