Base text after Brill’s New Jacoby / Jacoby (FGrH). The apparatus records differences against the open-corpus witness (Scaife/Perseus) within the overlapping span; encoding and orthography are folded. Collation codepoint-verified and adversarially audited (2026-06-26).
| Lemma / locus | BNJ base | Witness | Type | Note |
|---|---|---|---|---|
| — | [ἢ Ἑκαταίου] | Perseus (Arrian, Anab. 5.6.5; urn:cts:greekLit:tlg0074.tlg001.perseus-grc2) | seclusion | BNJ editorial seclusion (athetesis): the candidate target_grc brackets ἢ Ἑκαταίου as deleted/doubted by the editor. Perseus prints ἢ Ἑκαταίου UNBRACKETED as part of the running text. Confirmed by codepoint diff: with the brackets removed and iota-adscript folded to subscript, candidate and Perseus are byte-for-byte identical through ὀνομάζουσιν, so the bracket is the sole substantive editorial mark here. Per F301 policy this is type 'seclusion', NOT a supplied/[]-restored text. |
| — | καὶ οὐκ ἀμαυροῖς τεκμηρίοις ὅτι ταύτηι ἔχει ῾Ηροδότωι ἐπιδέδεικται. | Perseus (Arrian, Anab. 5.6.5) | omission | Extent difference: the stored Perseus witness_text ENDS at ὀνομάζουσιν; the entire final clause καὶ οὐκ ἀμαυροῖς τεκμηρίοις ὅτι ταύτῃ ἔχει Ἡροδότῳ ἐπιδέδεικται is absent from the witness. The candidate (BNJ) carries the fuller fragment. This is an omission/abridgement in the retrieved witness, not a word-substitution; classified omission because the witness lacks text the candidate prints. (The candidate's ταύτηι/῾Ηροδότωι in this clause again differ from standard subscript only by adscript encoding, but cannot be collated since the witness does not reach this clause.) |
urn:cts:greekLit. This is a BNJ/Jacoby-based edition with an open-corpus apparatus, not a verbatim copy of the witness edition. Source data: explorers/_build/editions/ecoed-hecat-fgrh1-f301.json.